Wild Crop Relatives: Genomic and Breeding Resources: Millets and Grasses

Wild Crop Relatives: Genomic and Breeding Resources

.

Chittaranjan Kole Editor

Wild Crop Relatives: Genomic and Breeding Resources Millets and Grasses

Editor Prof. Chittaranjan Kole Director of Research Institute of Nutraceutical Research Clemson University 109 Jordan Hall Clemson, SC 29634 [email protected]

ISBN 978-3-642-14254-3 e-ISBN 978-3-642-14255-0 DOI 10.1007/978-3-642-14255-0 Springer Heidelberg Dordrecht London New York # Springer-Verlag Berlin Heidelberg 2011 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: deblik, Berlin printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

Dedication

Dr. Norman Ernest Borlaug,1 the Father of Green Revolution, is well respected for his contributions to science and society. There was or is not and never will be a single person on this Earth whose single-handed service to science could save millions of people from death due to starvation over a period of over four decades like Dr. Borlaug’s. Even the Nobel Peace Prize he received in 1970 does not do such a great and noble person as Dr. Borlaug justice. His life and contributions are well known and will remain in the pages of history of science. I wish here only to share some facets of this elegant and ideal personality I had been blessed to observe during my personal interactions with him. It was early 2007 while I was at the Clemson University as a visiting scientist one of my lab colleagues told me that “somebody wants to talk to you; he appears to be an old man”. I took the telephone receiver casually and said hello. The response from the other side was – “I am Norman Borlaug; am I talking to Chitta?” Even a million words would be insufficient to define and depict the exact feelings and thrills I experienced at that moment!

1

The photo of Dr. Borlaug was kindly provided by Julie Borlaug (Norman Borlaug Institute for International Agriculture, Texas A&M Agriculture) the granddaughter of Dr. Borlaug.

v

vi

I had seen Dr. Borlaug only once, way back in 1983, when he came to New Delhi, India to deliver the Coromandal Lecture organized by Prof. M.S. Swaminathan on the occasion of the 15th International Genetic Congress. However, my real interaction with him began in 2004 when I had been formulating a 7-volume book series entitled Genome Mapping and Molecular Breeding in Plants. Initially, I was neither confident of my ability as a series/book editor nor of the quality of the contents of the book volumes. I sent an email to Dr. Borlaug attaching the table of contents and the tentative outline of the chapters along with manuscripts of only a few sample chapters, including one authored by me and others, to learn about his views as a source of inspiration (or caution!) I was almost sure that a person of his stature would have no time and purpose to get back to a small science worker like me. To my utter (and pleasant) surprise I received an email from him that read: “May all Ph.D.’s, future scientists, and students that are devoted to agriculture get an inspiration as it refers to your work or future work from the pages of this important book. My wholehearted wishes for a success on your important job”. I got a shot in my arm (and in mind for sure)! Rest is a pleasant experience – the seven volumes were published by Springer in 2006 and 2007, and were welcome and liked by students, scientists and their societies, libraries, and industries. As a token of my humble regards and gratitude, I sent Dr. Borlaug the Volume I on Cereals and Millets that was published in 2006. And here started my discovery of the simplest person on Earth who solved the most complex and critical problem of people on it – hunger and death. Just one month after receiving the volume, Dr. Borlaug called me one day and said, “Chitta, you know I cannot read a lot now-a-days, but I have gone through only on the chapters on wheat, maize and rice. Please excuse me. Other chapters of this and other volumes of the series will be equally excellent, I believe”. He was highly excited to know that many other Nobel Laureates including Profs. Arthur Kornberg, Werner Arber, Phillip Sharp, G€ unter Blobel, and Lee Hartwell also expressed generous comments regarding the utility and impact of the book series on science and the academic society. While we were discussing many other textbooks and review book series that I was editing at that time, again in my night hours for the benefit of students, scientists, and industries, he became emotional and said to me, “Chitta, forget about your original contributions to basic and applied sciences, you deserved Nobel Prize for Peace like me for providing academic foods to millions of starving students and scientists over the world particularly in the developing countries. I will recommend your name for the World Food Prize, but it will not do enough justice to the sacrifice you are doing for science and society in your sleepless nights over so many years. Take some rest Chitta and give time to Phullara, Sourav and Devleena” (he was so particular to ask about my wife and our kids during most of our conversations). I felt honored but really very ashamed as I am aware of my almost insignificant contribution in comparison to his monumental contribution and thousands of scientists over the world are doing at least hundred-times better jobs than me as scientist or author/editor of books! So, I was unable to utter any words for a couple of minutes but realized later that he must been too affectionate to me and his huge affection is the best award for a small science worker as me! In another occasion he wanted some documents from me. I told him that I will send them as attachments in emails. Immediately he shouted and told me: “You know, Julie (his granddaughter) is not at home now and I cannot check email myself. Julie does this for me. I can type myself in type writer but I am not good in computer. You know what, I have a xerox machine and it receives fax also. Send me

Dedication

Dedication

vii

the documents by fax”. Here was the ever-present child in him. Julie emailed me later to send the documents as attachment to her as the ‘xerox machine’ of Dr. Borlaug ran out of ink! Another occasion is when I was talking with him in a low voice, and he immediately chided me: “You know that I cannot hear well now-a-days; I don’t know where Julie has kept the hearing apparatus, can’t you speak louder?” Here was the fatherly figure who was eager to hear each of my words! I still shed tears when I remember during one of our telephone conversations he asked: “You know I have never seen you, can you come to Dallas in the near future by chance?” I remember we were going through a financial paucity at that time and I could not make a visit to Dallas (Texas) to see him, though it would have been a great honor. In late 2007, whenever I tried to talk to Dr. Borlaug, he used to beckon Julie to bring the telephone to him, and in course of time Julie used to keep alive all the communications between us when he slowly succumbed to his health problems. The remaining volumes of the Genome Mapping and Molecular Breeding in Plants series were published in 2007, and I sent him all the seven volumes. I wished to learn about his views. During this period he could not speak and write well. Julie prepared a letter based on his words to her that read: “Dear Chitta, I have reviewed the seven volumes of the series on Genome Mapping and Molecular Breeding in Plants, which you have authored. You have brought together genetic linkage maps based on molecular markers for the most important crop species that will be a valuable guide and tool to further molecular crop improvements. Congratulations for a job well done”. During one of our conversations in mid-2007, he asked me what other book projects I was planning for Ph.D. students and scientists (who had always been his all-time beloved folks). I told him that the wealth of wild species already utilized and to be utilized for genetic analysis and improvement of domesticated crop species have not been deliberated in any book project. He was very excited and told me to take up the book project as soon as possible. But during that period I had a huge commitment to editing a number of book volumes and could not start the series he was so interested about. His sudden demise in September 2009 kept me so morose for a number of months that I could not even communicate my personal loss to Julie. But in the meantime, I formulated a 10-volume series on Wild Crop Relatives: Genomic and Breeding Resources for Springer. And whom else to dedicate this series to other than Dr. Borlaug! I wrote to Julie for her formal permission and she immediately wrote me: “Chitta, Thank you for contacting me and yes I think my grandfather would be honored with the dedication of the series. I remember him talking of you and this undertaking quite often. Congratulations on all that you have accomplished!” This helped me a lot as I could at least feel consoled that I could do a job he wanted me to do and I will always remain grateful to Julie for this help and also for taking care of Dr. Borlaug, not only as his granddaughter but also as the representative of millions of poor people from around the world and hundreds of plant and agricultural scientists who try to follow his philosophy and worship him as a father figure. It is another sad experience of growing older in life that we walk alone and miss the affectionate shadows, inspirations, encouragements, and blessings from the fatherly figures in our professional and personal lives. How I wish I could treat my next generations in the same way as personalities like Mother Teresa and Dr. Norman Borlaug and many other great people from around the world treated me!

viii

Dedication

During most of our conversations he used to emphasize on the immediate impact of research on the society and its people. A couple of times he even told me that my works on molecular genetics and biotechnology, particularly of 1980s and 1990s, have high fundamental importance, but I should also do some works that will benefit people immediately. This advice elicited a change in my thoughts and workplans and since then I have been devotedly endeavoring to develop crop varieties enriched with phytomedicines and nutraceuticals. Borlaug influenced both my personal and professional life, particularly my approach to science, and I dedicate this series to him in remembrance of his great contribution to science and society and for all his personal affection, love and blessings for me. I emailed the above draft of the dedication page to Julie for her views and I wish to complete my humble dedication with great satisfaction with the words of Julie who served as the living ladder for me to reach and stay closer to such as great human being as Dr. Borlaug and express my deep regards and gratitude to her. Julie’s email read: “Chitta, Thank you for sending me the draft dedication page. I really enjoyed reading it and I think you captured my grandfather’s spirit wonderfully. . .. So thank you very much for your beautiful words. I know he would be and is honored.” Clemson, USA

Chittaranjan Kole

Preface

Wild crop relatives have been playing enormously important roles both in the depiction of plant genomes and the genetic improvement of their cultivated counterparts. They have contributed immensely to resolving several fundamental questions, particularly those related to the origin, evolution, phylogenetic relationship, cytological status and inheritance of genes of an array of crop plants; provided several desirable donor genes for the genetic improvement of their domesticated counterparts; and facilitated the innovation of many novel concepts and technologies while working on them directly or while using their resources. More recently, they have even been used for the verification of their potential threats of gene flow from genetically modified plants and invasive habits. Above all, some of them are contributing enormously as model plant species to the elucidation and amelioration of the genomes of crop plant species. As a matter of fact, as a student, a teacher, and a humble science worker I was, still am and surely will remain fascinated by the wild allies of crop plants for their invaluable wealth for genetics, genomics and breeding in crop plants and as such share a deep concern for their conservation and comprehensive characterization for future utilization. It is by now a well established fact that wild crop relatives deserve serious attention for domestication, especially for the utilization of their phytomedicines and nutraceuticals, bioenergy production, soil reclamation, and the phytoremediation of ecology and environment. While these vastly positive impacts of wild crop relatives on the development and deployment of new varieties for various purposes in the major crop plants of the world agriculture, along with a few negative potential concerns, are envisaged the need for reference books with comprehensive deliberations on the wild relatives of all the major field and plantation crops and fruit and forest trees is indeed imperative. This was the driving force behind the inception and publication of this series. Unlike the previous six book projects I have edited alone or with co-editors, this time it was very difficult to formulate uniform outlines for the chapters of this book series for several obvious reasons. Firstly, the status of the crop relatives is highly diverse. Some of them are completely wild, some are sporadically cultivated and some are at the initial stage of domestication for specific breeding objectives recently deemed essential. Secondly, the status of their conservation varies widely: some have been conserved, characterized and utilized; some have been eroded completely except for their presence in their center(s) of origin; some are at-risk or endangered due to genetic erosion, and some of them have yet to be explored. The third constraint is the variation in their relative worth, e.g. as academic model, breeding resource, and/or potential as “new crops.” ix

x

The most perplexing problem for me was to assign the chapters each on a particular genus to different volumes dedicated to crop relatives of diverse crops grouped based on their utility. This can be exemplified with Arabidopsis, which has primarily benefited the Brassicaceae crops but also facilitated genetic analyses and improvement in crop plants in other distant families; or with many wild relatives of forage crops that paved the way for the genetic analyses and breeding of some major cereal and millet crops. The same is true for wild crop relatives such as Medicago truncatula, which has paved the way for in-depth research on two crop groups of diverse use: oilseed and pulse crops belonging to the Fabaceae family. The list is too long to enumerate. I had no other choice but to compromise and assign the genera of crop relatives in a volume on the crop group to which they are taxonomically the closest and to which they have relatively greater contributions. For example, I placed the chapter on genus Arabidopsis in the volume on oilseeds, which deals with the wild relatives of Brassicaceae crops amongst others. However, we have tried to include deliberations pertinent to the individual genera of the wild crop relatives to which the chapters are devoted. Descriptions of the geographical locations of origin and genetic diversity, geographical distribution, karyotype and genome size, morphology, etc. have been included for most of them. Their current utility status – whether recognized as model species, weeds, invasive species or potentially cultivable taxa – is also delineated. The academic, agricultural, medicinal, ecological, environmental and industrial potential of both the cultivated and/or wild allied taxa are discussed. The conservation of wild crop relatives is a much discussed yet equally neglected issue albeit the in situ and ex situ conservations of some luckier species were initiated earlier or are being initiated now. We have included discussions on what has happened and what is happening with regard to the conservation of the crop relatives, thanks to the national and international endeavors, in most of the chapters and also included what should happen for the wild relatives of the so-called new, minor, orphan or future crops. The botanical origin, evolutionary pathway and phylogenetic relationship of crop plants have always attracted the attention of plant scientists. For these studies morphological attributes, cytological features and biochemical parameters were used individually or in combinations at different periods based on the availability of the required tools and techniques. Access to different molecular markers based on nuclear and especially cytoplasmic DNAs that emerged after 1980 refined the strategies required for precise and unequivocal conclusions regarding these aspects. Illustrations of these classical and recent tools have been included in the chapters. Positioning genes and defining gene functions required in many cases different cytogenetic stocks, including substitution lines, addition lines, haploids, monoploids and aneuploids, particularly in polyploid crops. These aspects have been dealt in the relevant chapters. Employment of colchiploidy, fluorescent or genomic in situ hybridization and Southern hybridization have reinforced the theoretical and applied studies on these stocks. Chapters on relevant genera/species include details on these cytogenetic stocks. Wild crop relatives, particularly wild allied species and subspecies, have been used since the birth of genetics in the twentieth century in several instances such as studies of inheritance, linkage, function, transmission and evolution of genes. They have been frequently used in genetic studies since the advent of molecular markers. Their involvement in molecular mapping has facilitated the development of mapping

Preface

Preface

xi

populations with optimum polymorphism to construct saturated maps and also illuminating the organization, reorganization and functional aspects of genes and genomes. Many phenomena such as genomic duplication, genome reorganization, self-incompatibility, segregation distortion, transgressive segregation and defining genes and their phenotypes have in many cases been made possible due to the utilization of wild species or subspecies. Most of the chapters contain detailed elucidations on these aspects. The richness of crop relatives with biotic and abiotic stress resistance genes was well recognized and documented with the transfer of several alien genes into their cultivated counterparts through wide or distant hybridization with or without employing embryo-rescue and mutagenesis. However, the amazing revelation that the wild relatives are also a source of yield-related genes is a development of the molecular era. Apomictic genes are another asset of many crop relatives that deserve mention. All of these past and the present factors have led to the realization that the so-called inferior species are highly superior in conserving desirable genes and can serve as a goldmine for breeding elite plant varieties. This is particularly true at a point when natural genetic variability has been depleted or exhausted in most of the major crop species, particularly due to growing and promoting only a handful of so-called high-yielding varieties while disregarding the traditional cultivars and landraces. In the era of molecular breeding, we can map desirable genes and polygenes, identify their donors and utilize tightly linked markers for gene introgression, mitigating the constraint of linkage drag, and even pyramid genes from multiple sources, cultivated or wild taxa. The evaluation of primary, secondary and tertiary gene pools and utilization of their novel genes is one of the leading strategies in present-day plant breeding. It is obvious that many wide hybridizations will never be easy and involve near-impossible constraints such as complete or partial sterility. In such cases gene cloning and gene discovery, complemented by intransgenic breeding, will hopefully pave the way for success. The utilization of wild relatives through traditional and molecular breeding has been thoroughly enumerated over the chapters throughout this series. Enormous genomic resources have been developed in the model crop relatives, for example Arabidopsis thaliana and Medicago truncatula. BAC, cDNA and EST libraries have also been developed in some other crop relatives. Transcriptomes and metabolomes have also been dissected in some of them. However, similar genomic resources are yet to be constructed in many crop relatives. Hence this section has been included only in chapters on the relevant genera. In this book series, we have included a section on recommendations for future steps to create awareness about the wealth of wild crop relatives in society at large and also for concerns for their alarmingly rapid decrease due to genetic erosion. The authors of the chapters have also emphasized on the imperative requirement of their conservation, envisaging the importance of biodiversity. The importance of intellectual property rights and also farmers’ rights as owners of local landraces, botanical varieties, wild species and subspecies has also been dealt in many of the chapters. I feel satisfied that the authors of the chapters in this series have deliberated on all the crucial aspects relevant to a particular genus in their chapters. I am also very pleased to present many chapters in this series authored by a large number of globally reputed leading scientists, many of whom have contributed to the development of novel concepts, strategies and tools of genetics, genomics and breeding and/or pioneered the elucidation and improvement of particular plant

xii

genomes using both traditional and molecular tools. Many of them have already retired or will be retiring soon, leaving behind their legacies and philosophies for us to follow and practice. I am saddened that a few of them have passed away during preparation of the manuscripts for this series. At the same time, I feel blessed that all of these stalwarts shared equally with me the wealth of crop relatives and contributed to their recognition and promotion through this endeavor. I would also like to be candid with regard to my own limitations. Initially I planned for about 150 chapters devoted to the essential genera of wild crop relatives. However, I had to exclude some of them either due to insignificant progress made on them during the preparation of this series, my failure to identify interested authors willing to produce acceptable manuscripts in time or authors’ backing out in the last minute, leaving no time to find replacements. I console myself for this lapse with the rationale that it is simply too large a series to achieve complete satisfaction on the contents. Still I was able to arrange about 125 chapters in the ten volumes, contributed by nearly 400 authors from over 40 countries of the world. I extend my heartfelt thanks to all these scientists, who have cooperated with me since the inception of this series not only with their contributions, but also in some cases by suggesting suitable authors for chapters on other genera. As happens with a megaseries, a few authors had delays for personal or professional reasons, and in a few cases, for no reason at all. This caused delays in the publication of some of the volumes and forced the remaining authors to update their manuscripts and wait too long to see their manuscripts in published form. I do shoulder all the responsibilities for this myself and tender my sincere apologies. Another unique feature of this series is that the authors of chapters dedicated to some genera have dedicated their chapters to scientists who pioneered the exploration, description and utilization of the wild species of those genera. We have duly honored their sincere decision with equal respect for the scientists they rightly reminded us to commemorate. Editing this series was, to be honest, very taxing and painstaking, as my own expertise is limited to a few cereal, oilseed, pulse, vegetable, and fruit crops, and some medicinal and aromatic plants. I spent innumerable nights studying to attain the minimum eligibility to edit the manuscripts authored by experts with even life-time contributions on the concerned genera or species. However, this indirectly awakened the “student-for-life” within me and enriched my arsenal with so many new concepts, strategies, tools, techniques and even new terminologies! Above all, this helped me to realize that individually we know almost nothing about the plants on this planet! And this realization strikingly reminded me of the affectionate and sincere advice of Dr. Norman Borlaug to keep abreast with what is happening in the crop sciences, which he used to do himself even when he had been advised to strictly limit himself to bed rest. He was always enthusiastic about this series and inspired me to take up this huge task. This is one of the personal and professional reasons I dedicated this book series to him with a hope that the present and future generations of plant scientists will share the similar feelings of love and respect for all plants around us for the sake of meeting our never-ending needs for food, shelter, clothing, medicines, and all other items used for our basic requirements and comfort. I am also grateful to his granddaughter, Julie Borlaug, for kindly extending her permission to dedicate this series to him. I started editing books with the 7-volume series on Genome Mapping and Molecular Breeding in Plants with Springer way back in 2005, and I have since

Preface

Preface

xiii

edited many other book series with Springer. I always feel proud and satisfied to be a member of the Springer family, particularly because of my warm and enriching working relationship with Dr. Sabine Schwarz and Dr. Jutta Lindenborn, with whom I have been working all along. My special thanks go out to them for publishing this “dream series” in an elegant form and also for appreciating my difficulties and accommodating many of my last-minute changes and updates. I would be remiss in my duties if I failed to mention the contributions of Phullara – my wife, friend, philosopher and guide – who has always shared with me a love of the collection, conservation, evaluation, and utilization of wild crop relatives and has enormously supported me in the translation of these priorities in my own research endeavors – for her assistance in formulating the contents of this series, for monitoring its progress and above all for taking care of all the domestic and personal responsibilities I am supposed to shoulder. I feel myself alien to the digital world that is the sine qua non today for maintaining constant communication and ensuring the preparation of manuscripts in a desirable format. Our son Sourav and daughter Devleena made my life easier by balancing out my limitations and also by willingly sacrificing the spare amount of time I ought to spend with them. Editing of this series would not be possible without their unwavering support. I take the responsibility for any lapses in content, format and approach of the series and individual volumes and also for any other errors, either scientific or linguistic, and will look forward to receiving readers’ corrections or suggestions for improvement. As I mentioned earlier this series consists of ten volumes. These volumes are dedicated to wild relatives of Cereals, Millets and Grasses, Oilseeds, Legume Crops and Forages, Vegetables, Temperate Fruits, Tropical and Subtropical Fruits, Industrial Crops, Plantation and Ornamental Crops, and Forest Trees. This volume “Wild Crop Relatives – Genomic and Breeding Resources: Millets and Grasses” includes 16 chapters dedicated to Agrostis, Bromus, Cenchrus, Cynodon, Dactylis, Dichanthium, Eleusine, Eragrostis, Festuca, Lolium, Panicum, Paspalum, Pennisetum, Phleum, Setaria and Zoysia. The chapters of this volume were authored by 48 scientists from 11 countries of the world namely Algerie, Argentina, Australia, France, India, Japan, New Zealand, Poland, Portugal, Turkey, and the USA. It is my sincere hope that this volume and the series as a whole will serve the requirements of students, scientists and industries involved in studies, teaching, research and the extension of millets and grasses with an intention of serving science and society. Clemson, USA

Chittaranjan Kole

.

Contents

1

Agrostis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 B.S. Ozdemir and H. Budak

2

Bromus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . W.M. Williams, A.V. Stewart, and M.L. Williamson

15

3

Cenchrus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Goel, H.D. Singh, and S.N. Raina

31

4

Cynodon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yanqi Wu

53

5

Dactylis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alan V. Stewart and Nicholas W. Ellison

73

6

Dichanthium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vishnu Bhat, C. Mahalakshmi, Shashi, Sunil Saran, and Soom Nath Raina

89

7

Eleusine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Susana S. Neves

8

Eragrostis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Mahmoud Zeid, Vivana Echenique, Marina Dı´az, Silvina Pessino, and Mark E. Sorrells

9

Festuca . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Toshihiko Yamada

10

Lolium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Hongwei Cai, Alan Stewart, Maiko Inoue, Nana Yuyama, and Mariko Hirata

11

Panicum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Hem S. Bhandari, Masumi Ebina, Malay C. Saha, Joseph H. Bouton, Sairam V. Rudrabhatla, and Stephen L. Goldman

12

Paspalum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 W.M. Williams, M.L. Williamson, and D. Real

xv

xvi

Contents

13

Pennisetum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Thierry Robert, Nadra Khalfallah, Evelyne Martel, Franc¸oise Lamy, Valerie Poncet, Cle´mentine Allinne, Marie-Stanislas Remigereau, Samah Rekima, Magalie Leveugle, Ghayas Lakis, Sonja Siljak-Yakovlev, and Aboubakry Sarr

14

Phleum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257 Alan V. Stewart, Andrzej J. Joachimiak, and Nicholas W. Ellison

15

Setaria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 Henri Darmency and Jack Dekker

16

Zoysia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Shin-ich Tsuruta, Makoto Kobayashi, and Masumi Ebina

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311

Abbreviations

5S A/L ABA ACCase AFLP ALS ANTC APO ASGR ASGR AusPGRIS BAC BADH BC BP BSA CAD CCoAOMT CCR cDNA CENARGEN CGIAR CHCH CIAT CID CISP cM CMS COMT COMT cpDNA CPPSE CP-sHSP Crs-1 CTAB cTBP

rDNA gene coding for the 5S rRNA Accumulation/loss mechanism Abscisic acid Acetyl coenzyme A carboxylase Amplified fragment length polymorphism Acetolactate synthase Anthocyanin Aposporous locus Apospory-specific genomic region Apomixis-related gene region Australian Plant Genetic Resource Information Service Bacterial artificial chromosome Betaine-aldehyde dehydrogenase Backcross Before present Bulked segregant analysis Cinnamyl alcohol dehydrogenase Caffeoyl-CoA O-methyltransferase Cinnamoyl CoA reductase Complementary DNA Centro Nacional de Pesquisa em Recursos Gene´ticos e Biotecnologia Consultative Group on International Agricultural Research Coiled-coil-helix–coiled-coil-helix International Centre for Tropical Agriculture Carbon isotope discrimination Conserved intron scanning primers Centi-Morgan Cytoplasmic male sterility Caffeic acid O-methyltransferase Caffeic acid/5-hydroxyferulic acid O-methyltransferase Chloroplast-DNA Centro de Pesquisa de Pecua´ria do Sudeste Chloroplast-localized small HSP Creeping bentgrass specific-1 gene Cetyl trimethyl ammonium bromide Combinatorial tubulin-based polymorphism xvii

xviii

cv. CWR DAF DAPI DDBJ DEF DM EMBL EMBRAPA EST FAA FAME FISH GAI GBSSI GC GFP GI GIS GMM GR GRIN GSS hph HSP IBC IBONE ICRISAT IGER IGFRI ILRI IMI INDEL INTA ISSR ITS IVDMD LEA LG LLR MAS MDA mRNA NADP NADP-ME NBPGR NBS NCBI NDF

Abbreviations

Cultivar Crop wild relative DNA amplification fingerprinting 40 ,6-Diamidino-2-phenylindole DNA Data Bank of Japan Diferentially expressed fragment Dry matter European Molecular Biology Laboratory Empresa Brasileira de Pesquisa Agropecua´ria Expressed sequence tag Formalin: acetic acid: ethyl alcohol Fatty acid methyl ester Fluorescence in situ hybridization Gene of agronomic interest Granule-bound starch synthase I Gas chromatography Green fluorescent protein Genes index Genomic in situ hybridization Genotype matrix mapping Glyphosate resistant Germplasm Resource Information Network (USDA-ARS) Genome survey sequence Hygromycin phosphotransferase (gene) Heat shock protein Institute of Biodiversity Conservation Instituto de Bota´nica del Nordeste International Crops Research Institute for the Semi-Arid Tropics Institute of Grassland and Environmental Research Indian Grassland and Fodder Research Institute International Livestock Research Institute Imidazolinone Insertion/deletion National Institute of Agriculture of Argentina Inter-simple sequence repeat Internal transcribed spacer In vitro dry matter digestibility Late embryogenesis abundant Linkage group Leucine-rich repeat Marker-assisted selection Malondialdehyde Messenger-RNA Nicotinamide adenine dinucleotide phosphate NADP-malic enzyme National Bureau of Plant Genetic Resources Nucleotide binding site National Center for Biotechnology Information Neutral detergent fiber

Abbreviations

xix

ndhF nirK, nirS NMDH NOR nosZ NPGS NPK OMT PAC PACC PAL PC PCA PCR PD PGR PHYB PIC PSII Q-PCR QTA QTL QTN R gene RAPD rDNA RFLP RGL RIL RNAi RT-PCR RWC s.str. SC SCAR SDRF SINGER SNP SRAP SSR STS TBP TDF TNL tRNA uidA UNED UPGMA

NADH dehydrogenase F Nitrite reductase genes NADP-dependent malate dehydrogenase Nucleolar organizer region Nitrous oxide reductase gene National Plant Germplasm System Nitrogen phosphorus potassium O-Methyl transferase Plasmid artificial chromosome Panicoid-arundinoid-chloridoid-centothecoid Phenylalanine ammonialyase Principal component Principal component analysis Polymerase chain reaction Power of discrimination Plant growth regulator Phytochrome B Polymorphic information content Photosystem II Quantitative real-time PCR Quantitative trait allele Quantitative trait loci Quantitative trait nucleotide Resistance gene Random(ly) amplified polymorphic DNA Ribosomal DNA Restriction fragment length polymorphism R-gene-like sequence Recombinant Inbred Line RNA-interference Real-time quantitative PCR Relative water content sensu stricto Secondary constriction Sequence-characterized amplified region Single dose restriction fragment System-wide Information Network for Genetic Resources Single nucleotide polymorphism Sequence-related amplified polymorphism Simple sequence repeat Sequence tagged site Tubulin-based polymorphism Transcript derived fragment TIR-NBS-LRR Transfer-RNA b-Glucuronidase (gene) United Nations Conference on Environment and Development Unweighted pair group method with arithmetic mean

xx

USDA USDA-APHIS USDA-SCS USGA WUE

Abbreviations

United States Department of Agriculture USDA-Animal and Plant Health Inspection Service USDA Soil Conservation Service United States Golf Association Water use efficiency

List of Contributors

Cle´mentine Allinne Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91405 Orsay, France, Clementine. [email protected] Hem Bhandari Forage Improvement Division, The Samuel Roberts Noble Foundation, Inc., 2510 Sam Noble Parkway, Ardmore, OK 73401, USA, [email protected] Vishnu Bhat Department of Botany, University of Delhi, Delhi 110007, India, [email protected] Joseph Bouton Forage Improvement Division, The Samuel Roberts Noble Foundation, Inc., 2510 Sam Noble Parkway, Ardmore, OK 73401, USA, [email protected] H. Budak Engineering and Natural Sciences, Biological Sciences and Bioengineering Program, Sabanci University, Orhanli, 34956 Tuzla, Istanbul, Turkey, [email protected] Hongwei Cai Forage Crop Research Institute, Japan Grassland Agriculture and Forage Seed Association, 388-5, Higashiakata, Nasushiobara, Tochigi 329-2742, Japan, [email protected] Henri Darmency INRA, Weed Biology and Management, 17 rue Sully, BP 86510, 21065 Dijon, France, [email protected] Jack Dekker Weed Biology Laboratory, Agronomy Department, Iowa State University, Ames, IA 50011, USA, [email protected] Marina Dı´az Departamento de Biologı´a, Bioquı´mica y Farmacia, Universidad Nacional del Sur, San Juan 670, 8000 Bahia Blanca, Argentina, [email protected] Masumi Ebina Forage Plant Breeding and Biotechnology, The National Institute of Livestock and Grassland Science, 768 Sembonmatsu, Nasushiobara, Tochigi 329-2793, Japan, [email protected] Viviana Echenique Departamento de Agronomı´a, Universidad Nacional del Sur, CERZOS-CONICET, San Andre´s 800, 8000 Bahia Blanca, Argentina, [email protected] Nicholas W. Ellison AgResearch, Private Bag 11008, Palmerston North, New Zealand, [email protected]

xxi

xxii

S. Goel Department of Botany, University of Delhi, Delhi 110007, India, [email protected] Stephen L. Goldman Department of Environmental Science, The University of Toledo, 2801 W. Bancroft, Toledo, OH 43606, USA, [email protected] Mariko Hirata Forage Crop Research Institute, Japan Grassland Agriculture and Forage Seed Association, 388-5, Higashiakata, Nasushiobara, Tochigi 329-2742, Japan, [email protected] Maiko Inoue Department of Plant, Soil, and Insect Science, University of Massachusetts, 8 French Hall, 230 Stockbridge Road, Amherst, MA 01003, USA, [email protected] Andrzej J. Joachimiak Department of Plant Cytology and Embryology, Institute of Botany, Jagiellonian University, Grodzka 52, 31-044 Cracow, Poland, [email protected] Nadra Khalfallah Laboratoire de Ge´ne´tique, Biochimie et Biotechnologies ve´ge´tales, Universite´ Mentouri Constantine, Constantine, Alge´rie, khalfallah@ wissal.dz Makoto Kobayashi National Institute of Livestock and Grassland Science, Forage Plant Breeding and Biotechnology, 768 Sembonmatsu, Nasushiobara, Tochigi 329-2793, Japan, [email protected] Ghayas Lakis Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Franc¸oise Lamy Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Magalie Leveugle Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] C. Mahalakshmi Department of Botany, University of Delhi, Delhi 110007, India, [email protected] Evelyne Martel Universite´ Lyon 1, UMR CNRS 5023 ‘Ecology of Fluvial Hydrosystems’, 43 Bd du 11 novembre, F-69622 Villeurbanne Cedex, France, [email protected] Susana S. Neves Plant Cell Biotechnology Laboratory,ITQB, Instituto de Tecnologia Quı´mica e Biolo´gica, Universidade Nova de Lisboa, Apartado 127, 2781-901 Oeiras, Portugal, [email protected] B.S. Ozdemir Engineering and Natural Sciences, Biological Sciences and Bioengineering Program, Sabanci University, Orhanli, 34956 Tuzla, Istanbul, Turkey, [email protected]

List of Contributors

List of Contributors

xxiii

Silvina Pessino Facultad de Ciencias Agrarias, Universidad Nacional de RosarioArgentina, Parque Villarino, S2125ZAA Zavalla, Santa Fe, Argentina, [email protected] Soom Nath Raina Amity Institute of Biotechnology, Amity University, Noida 201303, Uttar Pradesh, India, [email protected] Daniel Real Department of Agriculture and Food, Western Australia (DAFWA), South Perth, WA 6151, Australia, [email protected] Samah Rekima Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Marie-Stanislas Remigereau Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Thierry Robert Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Malay Saha Forage Improvement Division, The Samuel Roberts Noble Foundation, Inc., 2510 Sam Noble Parkway, Ardmore, OK 73401, USA, [email protected] R.V. Sairam Pennsylvania State University-Harrisburg, The School of Science Engineering and Technology, Middletown, PA 17057, USA, [email protected] Sunil Saran Amity Institute of Biotechnology, Amity University, Noida 201303, Uttar Pradesh, India, [email protected] Aboubakry Sarr Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] Shashi Department of Botany, University of Delhi, Delhi 110007, India, [email protected] Sonja Siljak-Yakovlev Laboratoire Ecologie Syste´matique et Evolution UMR (8079) CNRS, Agroparitech, Universite´ Paris XI, B 360, 91 405 Orsay, France, [email protected] H.D. Singh Department of Botany, University of Delhi, Delhi 110007, India, [email protected] Mark E. Sorrells Department of Plant Breeding and Genetics, Cornell University, 240 Emerson Hall, Ithaca, NY 14853-1902, USA, [email protected] Alan V. Stewart PGG Wrightson Seeds, PO Box 175Lincoln, Christchurch 7640, New Zealand, [email protected] Shin-ich Tsuruta University of Miyazaki, Faculty of Agriculture, 1-1 Gakuen Kibanadai-nishi, Miyazaki 889-2192, Japan, [email protected]

xxiv

Warren M. Williams AgResearch, Grasslands Research Centre, Private Bag 11008, Palmerston North 4442, New Zealand, [email protected] Michelle L. Williamson AgResearch, Grasslands Research Centre, Private Bag 11008, Palmerston North 4442, New Zealand, michelle.williamson@ agresearch.co.nz Yanqi Wu Department of Plant and Soil Sciences, Oklahoma State University, 368 Ag Hall, Stillwater, OK 74078, USA, [email protected] Toshihiko Yamada Field Science Center for Northern Biosphere, Hokkaido University, Kita 11, Nishi 10, Kita-ku, Sapporo 060-0811, Japan, yamada@fsc. hokudai.ac.jp Nana Yuyama Forage Crop Research Institute, Japan Grassland Agriculture and Forage Seed Association, 388-5, Higashiakata, Nasushiobara, Tochigi 329-2742, Japan, [email protected] Mahmoud Zeid Department of Plant Breeding and Genetics, Cornell University, 418 Bradfield Hall, Ithaca, NY 14853-1902, USA, [email protected]

List of Contributors

.

Chapter 1

Agrostis B.S. Ozdemir and H. Budak

1.1 Introduction The grass family (Poaceae) emerged 60 million years ago (Kellogg 2001) and it is one of the largest families that include various species with high economical importance, especially the essential cereal crops that are obligate in the daily diet. Grasses are used for many purposes such as in food production, industry, lawns, and sports fields. Turfgrasses are used in sports fields and recreation areas whereas they prevent soil erosion in natural habitats. In the United States, turfgrass constitutes the second place in the seed market (Lee 1996). Agrostis spp., bentgrass, contains more than 200 perennial turfgrass species as a genus in the Poaceae family (Hitchcock 1971), but five species of this genus are mainly used as turfgrass, which are all outcrossing, perennial, and cool-season grasses: colonial (Agrostis capillaris L.), velvet (Agrostis canina L.), creeping (Agrostis stolonifera L.), redtop (Agrostis gigantea Roth), and dryland (Agrostis castellana Boiss. and Reut.). Agrostis spp. is taxonomically classified under Gramineae (Poaceae) family, Pooideae subfamily, Aveneae tribe, and Agrostis genus (Warnke 2003). Bentgrasses are cool-season grasses that are widely used on golf courses (tees, fairways, and greens) at

H. Budak (*) Engineering and Natural Sciences, Biological Sciences and Bioengineering Program, Sabanci University, Orhanli, 34956 Tuzla, Istanbul, Turkey e-mail: [email protected]

temperate regions due to their dense nature, low mowing heights without damage, and green appearance. They are also used for parks and forage. They are cross-pollinating, self-incompatible, and pollinated by wind. Agrostis can exhibit both clonal growth and seed reproduction. It is hard to classify Agrostis genus taxonomically. Due to the similar morphological characters among this genus, others features are needed to be included for identification of new germplasms. Laser flow cytometry for the determination of ploidy level was found to be effective in differentiating between diploid, tetraploid, and hexaploid forms by evaluating six Agrostis species; A. canina L. subsp. canina, A. canina L. subsp. montana (Hartm.) Hartm., A. stolonifera var. palustris (Huds.) Farw., A. capillaris L., A. castellana Boiss. & Reut., and Agrostis alba L. (Bonos et al. 2002).

1.2 Three Major Bentgrass Species 1.2.1 Creeping Bentgrass (A. stolonifera L.) Creeping bentgrass is a cool-season grass species that is native to Western Europe. It is not only adapted to cool and humid areas but also preferred to be used in warmer places at golf courses due to its fine texture. Creeping bentgrass has high density and mowing height as low as 3 mm; these features make it suitable to be used especially in greens though it is also used in tees and fairways of the golf courses (Warnke 2003). It requires high maintenance, so it is not ideal for home

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_1, # Springer-Verlag Berlin Heidelberg 2011

1

2

lawns. It has low growth habit and aggressive spreading (Casler 2006). Creeping bentgrass is commonly referred to as A. stolonifera L., but other synonyms are also used such as Agrostis palustris Huds. and A. stolonifera L. var. palustris (Huds.). It is a strict allotetraploid with a genome A2A2A3A3 (2n ¼ 4x ¼ 28) (Warnke 2003). Complete chloroplast genome sequence of A. stolonifera (136,584 bp) was presented and compared with other grass species. The chloroplast genome contains 53.6% coding regions (44.7% protein coding genes and 8.9% RNA genes), and the rest is noncoding regions (Saski et al. 2007). Creeping bentgrass requires vernalization, and flowering occurs generally in late spring to early summer (Warnke 2003). Interspecific hybridization of creeping bentgrass is possible with five different Agrostis species (A. canina L., A. castellana Boiss. & Reut, A. gigantea Roth, A. capillaris L., and A. vinealis Schreb). Due to its stoloniferous growth, thick thatch layers can be formed, if it is not managed properly. This then serves as a good environment for pathogens and insects. It is susceptible to many diseases, but dollar spot (Sclerotinia homeocarpa), brown patch (Rhizoctonia solani), Typhula blight (Typhula incarnata or Typhula ishikariensis), the so-called gray snow mold, and Fusarium blight (Fusarium roseum and Fusarium tricinctum), also named as pink snow mold, are the main concerns. A study with rhizobacteria, isolated from roots of bentgrass and bermudagrass located from USGA golf putting greens, showed that Pseudomonas was the dominant genus in the roots of bentgrass for denitrification. For both species, 17% of the isolates were identified by GC-FAME (gas chromatography – fatty acid methyl ester) and 16S rDNA analyses. Besides, nitrous oxide reductase (nosZ) and nitrite reductase genes (nirK and nirS) were recognized (Wang and Skipper 2004).

1.2.2 Colonial Bentgrass (A. capillaris L.) Colonial bentgrass is both referred as A. capillaris L. and Agrostis tenuis Sibth. Origin of colonial bentgrass is Europe and temperate Asia. It is a cool-

B.S. Ozdemir and H. Budak

season grass and also named as brown top. While it is used in lawns and sports fields such as tennis courts and golf courses, it is also a choice for erosion control (Hubbard 1984). It is compatible with other species (Casler 2006). Colonial bentgrass is preferred on fairways and tees at golf courses with a mowing height of 1.0–2.5 cm; hence, it is not much suitable for greens. It has lower maintenance needs than creeping bentgrass; instead it has more limitations (Ruemmele 2003). Colonial bentgrass can adapt to different environments with forming specific ecotypes (Busey 2003). The colonial bentgrass (A. capillaris L.) with the genome A1A1A2A2 and creeping bentgrass (A. stolonifera L.) with the genome A2A2A3A3 are both allotetraploid and have 14 chromosome pairs (2n ¼ 4x ¼ 28) (Ruemmele 2003; Warnke 2003). Since they are sexually compatible, hybrids can be produced. Genomics studies facilitate analysis of many species for their origin and evolution with the use of many tools it provides. Expressed sequence tag (EST) sequence analysis of colonial and creeping bentgrass A2 genomes suggested that these genomes diverged from a common ancestor at about 2.2 million years ago and supported the previous studies that A2 genomes were common for both species. Their findings also showed that both creeping and colonial bentgrasses were closer to tribe Poeae than tribe Aveneae that are actually closely related tribes (Rotter et al. 2007).

1.2.3 Velvet Bentgrass (A. canina L.) Velvet bentgrass is referred as A. canina L., which is native to Europe. It is diploid (2n ¼ 2x ¼ 14) with genome designation of A1A1. It is confused widely with A. vinealis (2n ¼ 4x ¼ 28) which is an autotetraploid of A. canina. Velvet bentgrass is a cool-season grass with very fine texture, maintaining a good putting surface at golf courses, and it has stoloniferous growth habit. It is the most shade-tolerant among other Agrostis spp. It has good heat, low temperature, and drought tolerance. Besides, it is more tolerant to acidic soils then all other bentgrasses. However, it needs high maintenance (Brilman 2003) and it has low growth habit (Casler 2006).

1 Agrostis

3

1.3 Three Minor Bentgrass Species

1.4 Marker Systems

1.3.1 Redtop (A. gigantea Roth)

The origin of colonial bentgrass is temperate Asia and Europe. It is well adapted to low temperatures and partial shade and mostly used for tees and fairways of the golf courses. In order to study genetic variation and relation among colonial bentgrass populations, random amplified polymorphic DNA (RAPD) marker analysis was applied, revealing high degree of genetic diversity. The accessions introduced from Europe to USA had genetic resemblance with the European cultivars; however, accessions from Bulgaria and Turkey showed the lowest genetic similarity, though these two regions are not geographically distant from each other (Rajasekar et al. 2007). Twenty-two different colonial bentgrasses (A. capillaris L.) from different countries, including few commercial cultivars, were analyzed with 128 amplified fragment length polymorphism (AFLP) markers, resulting in high level of genetic biodiversity. Diverse gene pool availability was especially investigated in the accession rather than the commercial cultivars (Zhao et al. 2006). It is natural that during domestication events, through breeding, cultivated species loses its genetic diversity and gains uniformity. So that wild relatives or other species play an important role in plant improvement programs. Colonial bentgrass is a good source with its diverse germplasm for the improvement of other Agrostis species or turfgrass types. Creeping bentgrass (A. stolonifera L.), colonial bentgrass (A. capillaris L.), and velvet bentgrass (A. canina L.) are the extensively used bentgrasses (Agrostis spp.) on putting greens, tees, and fairways of the golf courses at temperate regions. They are highly difficult to be identified using their morphological features. However, it was investigated that RAPD markers were effective in differentiation and identification of bentgrass species at molecular level (Hollman et al. 2005). Besides the use of RAPD markers, sequence characterized amplified region (SCAR) markers were designed for colonial and creeping bentgrasses. They were found to be effective to differentiate between these two bentgrasses and these from other Agrostis spp. This technology was also recommended to be used for selecting the progenies produced from interspecific hybridization between colonial and creeping bentgrasses (Scheef et al. 2003).

Redtop is a cool-season, perennial grass, which is referred as A. gigantea Roth or A. alba. Redtop bentgrass (A. gigantea Roth) is allopolyploid with the genome A1A1A2A2A3A3 (2n ¼ 6x ¼ 42), which are the inclusion of the genomes of both creeping and colonial bentgrasses. It can be crossed with creeping bentgrass to produce infertile hybrids. Redtop is used as forage or turf or for reclamation purposes for revegetation of disturbed areas. It has low maintenance (Brede and Sellmann 2003). It has rapid establishment (Casler 2006). It is tolerant to high metal concentrations and is the most salt-tolerant among other bentgrasses. It can be used in breeding programs for enhanced tolerance. It has the potential to be improved as forage or turf (Brede and Sellmann 2003).

1.3.2 Highland Bentgrass (A. castellana Boiss. and Reuter) Highland bentgrass is a dense turf that has good color in winter times. It is tetraploid, but its hexaploid types can also be found. It is adapted to warmer and less humid areas than A. capillaris. A. castellana has mowing heights ranging from 1.0 to 1.9 cm and can be used for turf purposes. But it is better if used as a low maintenance grass or for reclamation purposes (Brede and Sellmann 2003); it is a drought-tolerant species (Casler 2006).

1.3.3 Idaho Bentgrass (Agrostis idahoensis Nash) Idaho bentgrass is a cool-season, perennial grass species, which is native to North America. Depending on its mowing height, it shows more tolerance to diseases than any other bentgrass types. It can be used for improvement of Agrostis germplasm in pest resistance (Brede and Sellmann 2003).

4

Few creeping bentgrass cultivars were identified using restriction fragment length polymorphism (RFLP) markers (Caceres et al. 2000), which is one of the first marker types to be used in linkage studies and an important tool for genetic mapping. Moreover, molecular marker-based linkage map was developed for creeping bentgrass using AFLP, RAPD, and cDNA-RFLP markers. Further use of this map would be useful in detection of quantitative trait loci (QTL) and marker-assisted breeding of agronomically important traits like disease resistance (Chakraborty et al. 2005). Transcript polymorphism analysis using SRAP system in three different bentgrasses (colonial, creeping, and velvet bentgrass) was the first study to be reported on turfgrasses. The ESTs obtained from this study could be used in turfgrass improvement programs (Dinler and Budak 2008). AFLP analysis for tetraploid creeping bentgrass and hexaploid redtop bentgrass was performed to investigate genetic diversity of the old and new cultivars, European plant introductions, and gray snow mold (T. incarnata Lasch) resistant genotypes of creeping bentgrass and could be used in plant improvements and gene mapping studies (Vergara and Bughrara 2004). For dollar spot disease, QTL analysis was performed to deeply understand the genetics of the resistance to this disease. It was found that dollar spot resistance is a quantitative trait and highly heritable. RAPD markers tightly linked to this QTL could be a useful tool in breeding programs of Agrostis spp. (Chakraborty et al. 2006).

1.5 Heat Stress and Drought Tolerance Plants are affected from temperature changes since all metabolic activities take place at a certain temperature. Heat stress tolerance level differs according to the plant species. Some plants can become tolerant to high levels of heat stress by time, and that can be exhibited as a genetic character. In cool-season grass species, heat stress causes leaf senescence that is regulated by cytokinins. The decrease in cytokinin content causing an increase in leaf senescence process was reported in Agrostis species (Xu and Huang 2007). Chlorosis followed by leaf senescence causes a decline in chlorophyll content and photosynthetic capacity resulting in reduced plant

B.S. Ozdemir and H. Budak

growth (John et al. 1995) and, especially for turfgrass, loss of deep green color, which is an important parameter for its economic value. SAG12-ipt (ligation of ipt, adenine isopentenyl transferase gene to SAG12, senescence-activated promoter) gene integration into creeping bentgrass resulted in transgenic lines that had faster growth under normal temperatures and increased ipt gene expression and cytokinin production suppressed heat-induced leaf senescence (Xu et al. 2009) resulting in an improvement for coolseason grasses exposed to heat stress. The main problem of growing cool-season grasses in temperate areas is the increasing temperature in summer time causing a decline in plant productivity. Using differential display analysis, upregulated genes (18 in A. stolonifera and 22 in A. scabra) in response to heat stress were identified and most genes were found in both species. A. scabra is a heat-tolerant species for which the significant decline in soluble protein content was observed at 40 C, whereas A. stolonifera is nontolerant under heat stress, and protein content conservation was up to 30 C (Xu and Huang 2008a, b). Only three of these genes were expressed in A. scabra, and it was reported that these genes might have been serving for the heat tolerance. Previous analysis of thermal and nonthermal A. scabra with 60 RAPD loci showed that they were not clustered independently but they were distantly related although they were morphologically similar (Tercek et al. 2003). If it is the case, this might enlighten the mechanism of plants’ tolerance to heat stress and facilitate development of in heat tolerant cultivars (Xu et al. 2008). Heat shock proteins (HSPs) in plants are synthesized in response to heat stress for protection. The small HSPs that are localized in chloroplasts (CP-sHSPs) are essential in heat tolerance expressing varying levels of its isoforms. Two additional isoforms of CP-sHSPs in heat-tolerant variants of creeping bentgrass (Penncross cultivar), which were generated under heat stress and selected for heat tolerance, were accumulated with respect to heat-sensitive variants that were not subjected to heat stress. These isoforms were genetically linked to heat tolerance (Park et al. 1996; Luthe et al. 2000). The study of CP-sHSP encoding genes isolated from heat-tolerant and heat-sensitive variants showed that the differences and variation in the expression of CP-sHSPs were related with the environmental adaptation and suggested that the amount of CP-sHSPs

1 Agrostis

might have been the source of heat tolerance (Wang et al. 2003; Wang and Luthe 2003). During the growth of creeping bentgrass species, the temperature was increased gradually. Following heat acclimation, the creeping bentgrass showed an improved tolerance under heat stress. In consecutive studies, it was found that heat acclimation was associated with suppressed lipid peroxidation (Larkindale and Huang 2004), enhanced heat shock protein expression (He et al. 2005), and enhanced photosynthetic activity by higher pigment content and rubisco activity (Liu and Huang 2008). A. scabra is a thermal species found to be tolerant to high temperatures like 40–45 C and to be living at the Yellowstone Natural Park (Tercek et al. 2003). For heat tolerance capacity, cool-season, perennial grasses, A. scabra with two genotypes and commercially important A. stolonifera with ten genotypes were investigated for the expression of AsEXP1 gene under heat stress. Heat tolerance of the treated plants differed according to the Agrostis species and genotypes and fell into three classes as being most tolerant to most sensitive. Heat tolerance level was positively correlated with the level of AsEXP1 gene expression. Totally, four genotypes, two from A. scabra and two from A. stolonifera, exhibited the highest level of gene expression with response to heat stress, being the most heat-tolerant ecotypes. This expansin gene, AsEXP1, was found to be highly upregulated in shoots and affected the heat tolerance of both species as “wholeplant.” The identification of this gene as response to heat tolerance in C3 Agrostis species for the first time was suggested to be important both for the study and investigation of the heat tolerant germplasm of grasses (Xu et al. 2007). In a further study, novel heat responsive genes were identified in A. scabra by subtractive suppression hybridization approach. The differentially expressed genes were classified based on their role in stress and defense mechanism, signaling and transcription, and protein or carbon metabolism (Tian et al. 2009) For plants, the soil temperature is more important than temperature of the air. The soil temperature primarily affects the roots. Heat-tolerant A. scabra is adapted to high soil temperatures at geothermal locations, whereas heat-sensitive A. stolonifera, adapted to cool climatic regions, cannot survive at high temperatures. When compared to A. stolonifera, whole-plant carbon balance and root carbon utilization was maintained positively,

5

and the root respiration rate was low or downregulated in A. scabra at high temperatures. Both factors were suggested to be the reason for root thermotolerance adaptation of cool-season grasses (Lyons et al. 2007). Both short-term and long-term respiratory acclimation was investigated to be accompanied with root thermotolerance A. scabra at increasing temperatures. A. scabra was found to be giving less response, in terms of root respiration, than A. stolonifera. The adjustment of root carbon utilization was stated to be caused by respiratory acclimization that the need for increased respiratory energy was lowered at high temperatures, resulting in increased root survival (Rachmilevitch et al. 2008). Thermal A. scabra has higher root viability than heat-sensitive A. stolonifera. The protein and phosphoprotein patterns of A. scabra and A. stolonifera were mainly different under heat stress. It was proposed that sucrose synthase, glutathione S-transferase, superoxide dismutase, stress-inducible heat shock protein, and aldolase phosphorylation might have been associated with root thermotolerance in cool-season grasses under heat stress (Xu and Huang 2008a, b). Since carbon utilization and accumulation are thought to have a role in heat tolerance and enhancement of plant survival in perennial grass species, three species of cool-season turfgrasses, colonial bentgrass (A. capillaris L.), creeping bentgrass (A. stolonifera L.), and velvet bentgrass (A. canina L.), were examined based on their shoot and root carbon partitioning and carbohydrate accumulation. Measuring the turf quality and relative leaf water content revealed that velvet bentgrass was the most tolerant under drought conditions and colonial bentgrass had the most amendatory potential. According to total nonstructural carbohydrate content measurements, increased carbon accumulation in roots after initial drought application and then its gathering in stems and leaves at increased drought durations suggested an adaptive response for drought survival (DaCosta and Huang 2006a, b, c). Other study with creeping bentgrass with different irrigation applications resulted in enhanced carbohydrate levels in leaves and roots when irrigating at wilt (Fu and Dernoeden 2008). Velvet bentgrass was found to be the most droughttolerant bentgrass species among others. In colonial, creeping, and velvet bentgrasses, prolonged drought stress caused a decrease in antioxidant enzyme activities and an increase in lipid peroxidation. However, in velvet bentgrass, oxidative damage was prevented for

6

a longer period of time under drought stress compared to other two bentgrasses that resulted in maintenance of higher turf quality and leaf relative water content (DaCosta and Huang 2007a, b). Besides, velvet bentgrass had higher degree of osmotic adjustment, which is crucial for cell turgor pressure maintenance, and this osmotic adjustment was accompanied with increased proline content and deposition of water-soluble carbohydrates both in velvet and creeping bentgrasses (DaCosta and Huang 2006a, b, c). Velvet bentgrass exhibited less injury and more ABA (abscisic acid) accumulation under drought conditions with respect to colonial and creeping bentgrass. On the other hand, creeping bentgrass exhibited higher recovery from drought stress and, at the same time, showed rapid decline in ABA and incline in cytokinin content, where ABA and cytokinin are regulators in decreased water availability for plant consumption (DaCosta and Huang 2007a, b).

1.6 Salt Tolerance Efficient irrigation management results in higher turf quality, water saving, and less cost on golf courses by evaluating the water requirements of the species that were found to be depending on species and time of the year (DaCosta and Huang 2006a, b, c). However, water availability and quality is decreasing in today’s world and at the same time, the demand for salt-tolerant cultivars is increasing. Salt-tolerant and sensitive cultivars were determined among creeping, colonial, and velvet bentgrass species (McCarty and Dudeck 1993; Marcum 2001), and red top bentgrass (McCarty and Dudeck 1993).

1.7 Invasion Properties and Weed Control Invasion Cultural management of weed in turfgrass is important in the fight of weed invasion since it is a more combined system. Healthy plants are less prone to weed colonization. Environmental stresses, both abiotic and biotic, cause weed colonization. The use of mowing, cultivation, fertilization, irrigation, turfgrass selection,

B.S. Ozdemir and H. Budak

adaptation to biotic and physiological stresses, and planting in cultural management of weed in turfgrass is reviewed by Busey (2003). Bentgrasses are grown as monoculture on golf courses. Weed control is the main problem of golf course maintenance. Poa annua L. is the major weed that shows similar tolerance responses to herbicides like creeping bentgrass. There is no effective control of this weed, so it causes severe problems. So, genetic manipulation of turfgrass species for glyphosate or glufosinate type resistance is mainly to fight with P. annua in greens and unwanted grasses in other areas of golf courses (Duncan 2004). A. capillaris grows in patches with its dense tillers. Leaf growth of A. capillaris is postponed in spring time. It is a rhizomatous species. A. capillaris was found to be vegetatively the most competitive and invasive temperate grass species when compared with Festuca rubra, Holcus lanatus, Lolium perenne, and Poa trivialis, though their time of vegetative spreading differed during the growing season (Barthram et al. 2006). It was proposed that the varied invasion properties of different grass species and the physical resistance was not only related with lamina density but also other lamina characteristics should have been involved (Barthram et al. 2005).

1.8 Diseases in Agrostis Species Fungal diseases are tried to be prevented with extensive fungicide application, which generates both environmental problems and economical losses. So, improvement of turfgrass species for resistance to fungal diseases such as dollar spot or brown patch, which cause serious problems in turfgrass management, has crucial importance (Chakraborty et al. 2006). Dead spot disease is a newly occurring disease caused by Ophiosphaerella agrostis and effective in creeping bentgrass. It is a major problem of greens in the golf courses. The disease occurs within the first and sixth years of the established greens, and in older greens, it might be seen after fumigation with methyl bromide. This disease appears first in 1–2 cm diameter area with reddish brown spots and later the patches become 8–10 cm in diameter (Dernoeden et al. 1999; Caˆmara et al. 2000; Kaminski and Dernoeden 2006).

1 Agrostis

Gray snow mold is another fungal disease to which the creeping bentgrass is susceptible and caused by T. incarnata (Wu and Hsiang 1998). It is favored by cold and humid conditions like under snow cover. Dollar spot (Sclerotinia homoeocarpa F.T. Bennett) is one of the major diseases of turfgrasses. Fungal diseases cause serious problem in all grasses. Disease control is mostly done through fungicide application (Dai et al. 2003). Treatment trials of creeping bentgrass with fungicides and fungicides plus plant growth regulators (PGRs) resulted in a better quality of creeping bentgrass after treatment with fungicides plus PGRs rather than fungicides alone (Fidanza et al. 2006). Differences in day and night temperatures and high humidity fortify the dollar spot disease. Though the disease resistance in creeping bentgrass has been genetically identified, its tolerance varies partially with environmental conditions (Bonos et al. 2003; Bonos 2005). In creeping bentgrass species, it was proposed that the dollar spot disease was most probably a quantitatively inherited trait, and parent selection was crucial for improvements in disease resistance (Bonos 2006). While creeping bentgrass is susceptible to dollar spot disease and shows different levels of sensitivity among its cultivars, colonial bentgrass species has resistance to this severe disease. Interspecific hybrids between these two species were performed and in some of the hybrids produced, high resistance to dollar spot disease and even no disease symptoms were observed (Belanger et al. 2004). During studies with resistance to dollar disease with creeping and colonial bentgrass species, it was found that a gene loss was occurring in the Agrostis genus. The creeping bentgrass-specific-1 gene, Crs-1, was examined in creeping bentgrass plants conferring the loss in most of the individuals, which occurs rarely, but the function of the protein has not been investigated yet (Li et al. 2005). Resistance (R) genes are important for genetic improvement of plant species. There are more than five different classes of R-genes. In a study of Budak and co-workers, constitutively expressed R-gene-like sequences (RGLs) from different Agrostis species (creeping, colonial, and velvet bentgrasses) were isolated and characterized by PCR-based motif-directed RNA fingerprinting. It was found that RGLs from these Agrostis species were highly conserved and some shared conserved motifs with other disease-resistant genes. Two TNL (TIR-NBS-LRR)-type RGLs were isolated

7

conferring the presence of these types of genes in grasses (Budak et al. 2006).

1.9 Plant Transformation in Agrostis Species In order to achieve improvements in agricultural traits, plant transformation technologies are important to use with classical breeding methods. Today’s molecular biology and plant biotechnology techniques enable us to study the structure and function of desired gene, isolate it from its wild relatives, other cultivars, and even from other species, and transfer that gene of interest to the existing cultivar that is in the process of improvement. Application and integration of the modern science into conventional breeding methods reduce the time spent for plant improvement studies and also provide us the ability of transferring defined gene or set of genes without contamination of undesired genes. The desired gene(s) to be integrated into the existing cultivar can be obtained from its domestic cultivars, wild relatives, and related genera (or any other organism) by expanding the availability of gene pool. In turfgrass species, although microprojectile bombardment and protoplast transformation are the major gene transfer systems to achieve transgenic plants, Agrobacterium-mediated transformation is used to get low copy number of the transgene in the transformed cultivars. Agrobacterium-mediated transformation is efficiently used in dicot plants since they occur as a natural host for Agrobacterium tumefaciens. However, by the improvements and modifications of the bacterial strains in this system, many monocots, especially the important cereal crops, can be transformed. Turfgrass was also transformed by using silicon carbide fibers or whiskers, electroporation, and polyethylene glycol-mediated techniques. Herbicide resistance, disease or insect resistance, and stress tolerance were the main traits to be improved in plant transformation studies. The transgenic Agrostis species with different transformation methods, marker systems, and transgene types are outlined in Table 1.1. By using green fluorescent protein (GFP) as a reporter gene, the first study of creeping bentgrass (A. stolonifera L.) by Agrobacterium-mediated transformation was done by Yu et al (2000). In a latter report, higher efficiency with large number of

8

B.S. Ozdemir and H. Budak

Table 1.1 Genetic transformation of bentgrass species with different transgene expression [modified from Wang and Ge (2006)] Plant species Transgenes Method Outcome References Agrostis alba (redtop) npt2 (neomycin phosphotransferase II) Protoplasts Transgenic plants Asano and Ugaki (1994) gusA (b-glucuronidase) Biolistics Transgenic plants Zhong et al. (1993) Agrostis stolonifera (creeping bar (phosphinothricin acetyltransferase) Biolistics Transgenic plants Hartman et al. bentgrass) (1994) bar (phosphinothricin acetyltransferase) Protoplasts Transgenic plants Lee et al. (1996) hph (hygromycin phosphotransferase), Biolistics Transgenic plants Xiao and Ha (1997) gusA (b-glucuronidase) bar (phosphinothricin acetyltransferase) Protoplasts Transgenic plants Asano et al. (1998) hph (hygromycin phosphotransferase), Whiskers Transgenic plants Dalton et al. (1998) gusA (b-glucuronidase) gfp (green fluorescent protein) Agrobacterium Transgenic plants Yu et al. (2000) Biolistics Transgenic plants Chai et al. (2002) hs2 (chitinase-like protein from American elm), bar (phosphinothricin acetyltransferase) PAPII, PAP-Y (pokeweed antiviral proteins), Biolistics Transgenic plantsa Dai et al. (2003) hph (hygromycin phosphotransferase) PR5K (receptor protein kinase gene from Biolistics Transgenic plantsa Guo et al. (2003) Arabidopsis thaliana) Biolistics Transgenic plantsa Wang et al. (2003) RCH10 (chitinase gene from rice), ALG (glucanase gene from alfalfa), bar (phosphinothricin acetyltransferase) bar (phosphinothricin acetyltransferase) Agrobacterium Transgenic plants Luo et al. (2004) TLPD34 (thaumatin-like protein from rice), Agrobacterium Transgenic plantsa Fu et al. (2005a, b) bar(phosphinothricin acetyltransferase) hph (hygromycin phosphotransferase), Agrobacterium Transgenic plants Han et al. (2005) gusA (b-glucuronidase) bar (phosphinothricin acetyltransferase), Agrobacterium Transgenic plants Wang and Ge (2005) gusA (b-glucuronidase) Agrostis tenuis hph (hygromycin phosphotransferase), Agrobacterium Transgenic plants Chai et al. (2004) (colonial bentgrass) gusA (b-glucuronidase) a Transgenic plants transformed with agronomically important traits

transgenic herbicide (bar gene integration)-resistant creeping bentgrass plants was achieved with low copy number of transgene integration by using super binary vector system (Luo et al. 2004). Transgenic herbicideresistant creeping bentgrass plants were produced recently using Agrobacterium-mediated transformation (Kim et al. 2007). Transgenic turfgrass was first produced via bombardment of A. palustris Huds. embryogenic callus tissues (Zhong et al. 1993). On the other hand, A. alba L. (redtop) plants were regenerated from protoplasts that were achieved from suspension culture of embryogenic calli (Asano and Sugiura 1990). The importance of tissue culture parameters in the efficiency of plant transformation system is well known. In order to obtain higher number of transgenic lines, first step is to establish an efficient tissue culture system. In genetic transformation studies, especially in monocots, immature or mature embryo-derived callus

cultures or directly the embryos are used as explant sources. But either way, calli formation is required before or after the transformation process, which is time-consuming. Mature embryo-derived embryogenic calli was used to transfer uidA (b-glucuronidase) and hph (hygromycin phosphotransferase) gene to creeping bentgrass with Agrobacterium-mediated transformation (Han et al. 2005). For creeping bentgrass species, stolon nodes were used with Agrobacterium-mediated transformation. This procedure maintained shoot formation directly from infected nodes by-passing the callus formation step (Wang and Ge 2005). Also, different tissues of creeping bentgrass cultivar Penn A4 were compared for their nuclease activity and its effect on transient GUS expression efficiency after bombardment (Basu et al. 2003). Using the whole vector construct during transformation results in integration of vector backbone sequences

1 Agrostis

into the genome. Using a simple cassette rather than the whole plasmid in transformation of creeping bentgrass by the bombardment of tissues with GFP reporter gene increased the low copy number of transgene integration since the system eliminates the nonessential insertions (Jayaraj et al. 2008) by time eliminating the disadvantage of biolistic transformation. Plant biotechnology is also widely used in improvement of disease resistance. Creeping bentgrass is sensitive to many insects, weeds, disease-causing fungi, and bacteria; hence, this species is used in many genetic improvement studies. Biolistic transformation of creeping bentgrass expressing three forms of pokeweed antiviral proteins (ribosome-inactivating proteins) was established via biolistics (Dai et al. 2003). Creeping bentgrass cultivar was transformed with the CP4 EPSPS gene that confers resistance to glyphosate herbicide. This Roundup Ready® glyphosate-resistant creeping bentgrass is under USDA-APHIS regulated status and for production control studies. It was planted for a four-year trial. Since creeping bentgrass is an outcrossing species with wind dispersal mechanism, at the end of the fourth year its dispersal area was heavily widened. Though in situ hybrids were not found, the potential transgene flow from creeping bentgrass and the possibility of its dispersion via water ways due to its vegetative growth by stolon formation was expressed. The importance of wind for pollen dispersal was emphasized and reported. Moreover, the study of Watrud et al. (2004) emphasized the importance of pollen dispersal by wind and demonstrated the long distance of viable pollen movement and gene flow using CP4 EPSPS as a marker. Also, Reichman et al. (2006) reported the identification of glyphosate-resistant creeping bentgrass plants outside the control area and the transgene escape from cultivated area to the native population area. Not for all transgenic crop species but especially the outcrossing ones were suggested to be evaluated for risk assessment more strictly (Zapiola et al. 2008). Factors affecting gene flow and the status of glyphosate-resistant (GR) crops including the GR creeping bentgrass that are commercially available or under deregulation in the USA were all reviewed by Mallory-Smith and Zapiola (2008). Overexpression of late embryogenesis abundant (LEA) proteins was shown to cause improved tolerance to water deficit in various species. Transformation of creeping bentgrass with barley hva1 gene, member of LEA protein family, using constitutive or

9

stress-inducible promoters was done, and transgenic plants with improved tolerance to drought were developed (Fu et al. 2007). Creeping bentgrass differs in color from a range of olive green to pale green. For ornamental purposes, creeping bentgrass was transformed with maize flavonoid/anthocyanin biosynthetic pathway transcription factor genes, Lc (leaf color) and Pl (purple leaf) via Agrobacterium-mediated gene transfer technique to obtain purple colored creeping bentgrass and the result was three different phenotypes (Han et al. 2009). The RTS gene expressed in tapetal cells was isolated from rice panicles. It was shown that this gene had a role in pollen development and its promoter conferred cell-specific expression. Transformation of rice, Arabidopsis, and creeping bentgrass with RTS promoter and cytotoxic barnase gene induced male sterility showing the promoter’s anther-specific expression, its role in pollen development, and use in male sterility both in dicot and monocot plants (Luo et al. 2006). The use of FLP/FRT site-specific DNA recombination system was evaluated using transgenic creeping bentgrass for yeast FLP recombinase expression and site-specific recombination was suggested to be a useful system for genome modification and transgene manipulation in turfgrass by altering the gene escape due to outcrossing and vegetative growth (Hu et al. 2006). Agrobacterium-mediated transformation was done with mature embryo-derived calli of A. stolonifera. PMI/mannose selection and GFP screenable marker system was used in this study by transferring GFP gene and E. coli manA (also termed as pmi and codes for phosphomannose isomerase) gene. This system was suggested to be environment and ecosystem friendly when compared with the use of herbicide or antibiotic selection systems (Fu et al. 2005a, b).

1.10 Hybridization Studies and Gene Escape Interspecific and intergeneric hybridization between glyphosate-resistant transgenic creeping bentgrass and nontransgenic species from both Agrostis and Polypogon genus was performed. From A. capillaris and P. fugax fertile hybrids were achieved, which makes the transgene movement possible within and

10

between species. This increased the risk of contamination of the weeds with herbicide resistance like genes and the possibility of new weed occurrence (Zhao et al. 2007). Interspecific hybridization between transgenic creeping bentgrass (A. stolonifera L.) that conferred resistance to glufosinate herbicide and four other nontransgenic Agrostis species (A. castellana Boiss. and Reut., A. gigantea Roth, A. canina L., and A. capillaris L.) were examined under field conditions. Interspecific transgenic hybrid frequency was at a lower rate with respect to intraspecific hybridization (0.631%). While the interspecific transgenic hybrids were formed between creeping bentgrass and A. capillaris (0.044%) and A. castellana (0.0015%), no transgenic hybrids with A. gigantea or A. canina were observed. The flowering time difference between the Agrostis species and the only substantial overlap of flowering between A. stolonifera and A. capillaris was the main concern of these hybridization rates (Belanger et al. 2003). Using the data from the previous study, transgene pollen dispersal model was performed. It was shown that pollen dispersal varied from site to site with the environmental conditions. It was stated that there was risk of transgene escape, but the persistence of this trait was possible only with herbicide selection, and the herbicide application was also causing natural resistance in weed populations. They suggested that the results of such gene flow could cause minimum problem at ecological aspect, and mismanagement of herbicide application could cause more serious problems (Meagher et al. 2003). Herbicide-resistant transgenic A. stolonifera was produced by engineering the CP4 EPSPS gene. It would have been the first transgenic wind-pollinated perennial crop grown in open areas, but it was reported that pollen-mediated transgene flow was causing a serious risk. Cytoplasmic male sterility or plastid engineering might be the solution in these cases to prevent transgene escape through pollen dispersal.

References Asano Y, Sugiura K (1990) Plant regeneration from suspension culture-derived protoplasts of Agrostis alba L. (Redtop). Plant Sci 72(2):267–273 Asano Y, Ugaki M (1994) Transgenic plants of Agrostis alba obtained by electroporation-mediated direct gene transfer into protoplasts. Plant Cell Rep 13:243–246

B.S. Ozdemir and H. Budak Asano Y, Ito Y, Fukami M, Sugiura K, Fujiie A (1998) Herbicide-resistant transgenic creeping bentgrass plants obtained by electroporation using an altered buffer. Plant Cell Rep 17:963–967 Barthram GT, Elston DA, Mullins CE (2005) The physical resistance of grass patches to invasion. Plant Ecol 176:79–85 Barthram GT, Elston DA, Griffiths JH, Bolton GR, Wright G (2006) Within-year variation in the vegetative spread of five temperate grasses. J Veg Sci 17:315–322 Basu C, Luo H, Kusch A, Chandlee J (2003) Transient reporter gene (S) in creeping bentgrass (Agrostis palustris) is affected by in vivo nuclease activity. Biotechnol Lett 25:939–944 Belanger FC, Meagher TR, Day PR, Plumley K, Meyer WA (2003) Interspecific hybridization between Agrostis stolonifera (creeping bentgrass) and related Agrostis species under field conditions. Crop Sci 43:240–246 Belanger FC, Bonos S, Meyer WA (2004) Dollar spot resistant hybrids between creeping bentgrass and colonial bentgrass. Crop Sci 44(2):581–586 Bonos SA (2005) Creeping bentgrass cultivars with improved dollar spot resistance. Golf Course Manag 73:96–100 Bonos SA (2006) Heritability of dollar spot resistance in creeping bentgrass. Phytopathology 96(8):808–812 Bonos SA, Plumley KA, Meyer WA (2002) Ploidy determination in Agrostis using flow cytometry and morphological traits. Crop Sci 42(1):192–196 Bonos SA, Casler MD, Meyer WA (2003) Inheritance of dollar spot resistance in creeping bentgrass. Crop Sci 43:2189–2196 Brede DA, Sellmann MJ (2003) Three minor Agrostis species: Redtop, Highland Bentgrass, and Idaho Bentgrass. In: Casler M, Duncan RR (eds) Turfgrass biology, genetics and breeding. Wiley, Hoboken, NJ, USA, pp 207–223 Brilman LA (2003) Velvet bentgrass (Agrostis canina L.). In: Casler M, Duncan RR (eds) Turfgrass biology, genetics and breeding. Wiley, Hoboken, NJ, pp 201–205 Budak H, Su S, Ergen N (2006) Revealing constitutively expressed resistance genes in Agrostis species using PCR-based motifdirected RNA fingerprinting. Genet Res 88:165–175 Busey P (2003) Cultural management of weeds in turfgrass: a review. Crop Sci 43:1899–1911 Caceres ME, Pupilli F, Piano E, Arcioni S (2000) RFLP markers are an effective tool for the identification of creeping bentgrass (Agrostis stolonifera L.) cultivars. Genet Resour Crop Evol 47:455–459 Caˆmara MPS, O’Neill NR, van Berkum P, Dernoeden PH, Palm ME (2000) Ophiosphaerella agrostis sp. nov. and its relationship to other species of Ophiosphaerella. Mycologia 92:317–325 Casler MD (2006) Centenary review: perennial grasses for turf, sport and amenity uses: evolution of form, function and fitness for human benefit. J Agric Sci 144:189–203 Chai B, Maqbool SB, Hajela RK, Green D, Vargas JM Jr, Warkentin D, Sabzikar R, Sticklen MB (2002) Cloning of a chitinaselike cDNA (hs2), its transfer to creeping bentgrass (Agrostis palustris Huds.) and development of brown patch (R. solani) disease resistant transgenic lines. Plant Sci 163:183–193 Chai ML, Senthil KK, Kim DH (2004) Transgenic plants of colonial bentgrass from embryogenic callus via Agrobacterium-mediated transformation. Plant Cell Tissue Organ Cult 77:165–171

1 Agrostis Chakraborty N, Bae J, Warnke S, Chang T, Jung G (2005) Linkage map construction in allotetraploid creeping bentgrass (Agrostis stolonifera L.). Theor Appl Genet 111: 795–803 Chakraborty N, Curley J, Warnke S, Casler MD, Jung G (2006) Mapping QTL for dollar spot resistance in creeping bentgrass (Agrostis stolonifera L.). Theor Appl Genet 113: 1421–1435 DaCosta M, Huang B (2006a) Changes in carbon partitioning and accumulation patterns during drought and recovery for colonial bentgrass, creeping bentgrass, and velvet bentgrass. J Am Soc Hortic Sci 131(4):484–490 DaCosta M, Huang BR (2006b) Osmotic adjustment associated with variation in bentgrass tolerance to drought stress. J Am Soc Hortic Sci 131(3):338–344 DaCosta M, Huang BR (2006c) Minimum water requirements for creeping, colonial, and velvet bentgrasses under fairway conditions. Crop Sci 46(1):81–89 DaCosta M, Huang BR (2007a) Changes in antioxidant enzyme activities and lipid peroxidation for bentgrass species in response to drought stress. J Am Soc Hortic Sci 132(3):319–326 DaCosta M, Huang BR (2007b) Drought survival and recuperative ability of bentgrass species associated with changes in abscisic acid and cytokinin production. J Am Soc Hortic Sci 132(1):60–66 Dai WD, Bonos S, Guo Z, Meyer WA, Day PR, Belanger FC (2003) Expression of pokeweed antiviral proteins in creeping bentgrass. Plant Cell Rep 21:497–502 Dalton SJ, Bettany AJE, Timms E, Morris P (1998) Transgenic plants of Lolium multiflorum, Lolium perenne, Festuca arundinacea and Agrostis stolonifera by silicon carbide fibermediated of cell suspension cultures. Plant Sci 132:31–43 Dernoeden PH, O’Neill NR, Caˆmara MPS, Feng Y (1999) A new disease of Agrostis palustris incited by an undescribed species of Ophiosphaerella. Plant Dis 83:397 Dinler G, Budak H (2008) Analysis of expressed sequence tags (ESTs) from Agrostis species obtained using sequence related amplified polymorphism. Biochem Genet 46:663–676 Duncan RR (2004) Exchange of science between turf & field crops: commonalities between turf & field crops and lessons. “New directions for a diverse planet”. Proceedings for the 4th International Crop Science Congress, 26 Sep–1 Oct 2004, Brisbane, Australia, pp 1–7 Fidanza MA, Wetzel HC III, Agnew JE, Kaminski JE (2006) Evaluation of fungicide and plant growth regulator tank-mix programmes on dollar spot severity of creeping bentgrass. Crop Prot 25:1032–1038 Fu JM, Dernoeden PH (2008) Carbohydrate metabolism in creeping bentgrass as influenced by two summer irrigation practices. J Am Soc Hortic Sci 133(5):678–683 Fu D, Tisserat Ned A, Xiao Y, Settleb D, Muthukrishnanc S, Liang GH (2005a) Overexpression of rice TLPD34 enhances dollar-spot resistance in transgenic bentgrass. Plant Sci 168:671–680 Fu D, Xiao Y, Muthukrishnan S, Liang GH (2005b) In vivo performance of a dual genetic marker, manA-gfp, in transgenic bentgrass. Genome 48:722–730 Fu D, Huang B, Xiao Y, Muthukrishnan S, Liang GH (2007) Overexpression of barley hva1 gene in creeping bentgrass for improving drought tolerance. Plant Cell Rep 26:467–477

11 Guo ZF, Bonos S, Meyer WA, Day PR, Belanger FC (2003) Transgenic creeping bentgrass with delayed dollar spot symptoms. Mol Breed 11:95–101 Han N, Chen D, Bian H-W, Deng M-J, Zhu M-Y (2005) Production of transgenic creeping bentgrass Agrostis stolonifera var. palustris plants by Agrobacterium tumefaciensmediated transformation using hygromycin selection. Plant Cell Tissue Organ Cult 81:131–138 Han Y-J, Kim Y-M, Lee Y-J, Kim SJ, Cho K-C, Chandrasekhar T, Song P-S, Woo Y-M, Kim J-I (2009) Production of purplecolored creeping bentgrass using maize transcription factor genes Pl and Lc through Agrobacterium-mediated transformation. Plant Cell Rep 28:397–406 Hartman CL, Lee L, Day PR, Tumer NE (1994) Herbicide resistant turfgrass (Agrostis palustris Huds.) by biolistic transformation. Biotechnology 12:919–923 He Y, Liu X, Huang B (2005) Protein changes in response to heat stress in acclimated and nonacclimated creeping bentgrass. J Am Soc Hortic Sci 130:521–526 Hitchcock AS (1971) Manual of the grasses of the United States, 2nd edn. Dover Publications, New York, USA Hollman AB, Stier JC, Casler MD, Jung G, Brilman LA (2005) Identification of putative velvet bentgrass clones using RAPD markers. Crop Sci 45(3):923–930 Hu Q, Nelson K, Luo H (2006) FLP-mediated site-specific recombination for genome modification in turfgrass. Biotechnol Lett 28:1793–1804 Hubbard JCE (1984) Grasses: a guide to their structure, identification, uses and distribution in the British Isles, 3rd edn. Viking Penguin, New York, USA Jayaraj J, Liang GH, Muthukrishnan S, Punja ZK (2008) Generation of low copy number and stably expressing transgenic creeping bentgrass plants using minimal gene cassette bombardment. Biol Plant 52(2):215–221 John I, Drake R, Farrell A, Cooper W, Lee P, Horton P, Grierson D (1995) Delayed leaf senescence in ethylene deficient ACCoxidase antisense tomato plants-molecular and physiological analysis. Plant J 7:483–490 Kaminski JE, Dernoeden PH (2006) Dead spot severity, pseudothecia development, and overwintering of Ophiosphaerella agrostis in creeping bentgrass. Phytopathology 96:248–254 Kellogg EA (2001) Evolutionary history of the grasses. Plant Physiol 125:1198–1205 Kim SJ, Lee JY, Kim YM, Yang SS, Hwang OJ, Hong NJ, Kim KM, Lee HY, Song PS, Kim JII (2007) Agrobacterium-mediated high-efficiency transformation of creeping bentgrass with herbicide resistance. J Plant Biol 50(5):577–585 Larkindale J, Huang B (2004) Change of lipid composition and saturation level in leaves and roots for heat-stressed and heat-acclimated creeping bentgrass (Agrostis stolonifera). Environ Exp Bot 51:57–67 Lee L (1996) Turfgrass biotechnology. Plant Sci 115:1–8 Lee L, Laramore CL, Day PR, Tumer E (1996) Transformation and regeneration of creeping bentgrass (Agrostis palustris Huds.) protoplasts. Crop Sci 36:401–406 Li HM, Rotter D, Bonos SA, Meyer WA, Belanger FC (2005) Identification of a gene in the process of being lost from the genus Agrostis. Plant Physiol 138:2386–2395 Liu X, Huang B (2008) Photosynthetic acclimation to high temperatures associated with heat tolerance in creeping bentgrass. J Plant Physiol 165:1947–1953

12 Luo H, Hu Q, Nelson K, Longo C, Kausch AP, Chandlee JM, Wipff JK, Fricker CR (2004) Agrobacterium tumefaciensmediated creeping bentgrass (Agrostis stolonifera L.) transformation using phosphinothricin selection results in a high frequency of single-copy transgene integration. Plant Cell Rep 22:645–652 Luo H, Lee J-Y, Hu Q, Nelson-Vasilchik K, Eitas TK, Lickwar C, Kausch AP, Chandlee JM, Hodges TK (2006) RTS, a rice anther-specific gene is required for male fertility and its promoter sequence directs tissue-specific gene expression in different plant species. Plant Mol Biol 62:397–408 Luthe DS, Krans JV, Park S-Y, Wang D (2000) The presence and role of heat-shock proteins in creeping bentgrass. In: Wilkinson RE (ed) Plant environment interactions, 2nd edn. Marcel Dekker, New York, USA, pp 283–319 Lyons EM, Pote J, DaCosta M, Huang B (2007) Whole-plant carbon relations and root respiration associated with root tolerance to high soil temperature for Agrostis grasses. Environ Exp Bot 59:307–313 Mallory-Smith C, Zapiola M (2008) Gene flow from glyphosate-resistant crops. Pest Manag Sci 64:428–440 Marcum MB (2001) Salinity tolerance of 35 bentgrass cultivars. HortScience 36(2):374–376 McCarty LB, Dudeck AE (1993) Salinity effects on bentgrass germination. HortScience 28(1):15–17 Meagher TR, Belanger FC, Day PR (2003) Using empirical data to model transgene dispersal. Philos Trans Roy Soc Lond B Biol Sci 358:1157–1162 Park S-Y, Shivaji R, Krans JV, Luthe DS (1996) Heat shock response in heat tolerant and non-tolerant variants of Agrostis palustris Huds. Plant Physiol 111:515–524 Rachmilevitch S, Lambers H, Huang B (2008) Short-term and long-term root respiratory acclimation to elevated temperatures associated with root thermotolerance for two Agrostis grass species. J Exp Bot 59(14):3803–3809 Rajasekar S, Fei S-Z, Christians NE (2007) Analysis of genetic diversity in colonial bentgrass (Agrostis capillaris L.) using randomly amplified polymorphic DNA (RAPD) markers. Genet Resour Crop Evol 54:45–53 Reichman JR, Watrud LS, Lee EH, Burdick CA, Bollman MA, Storm MJ, King GA, Mallory-Smith C (2006) Establishment of transgenic herbicide-resistant creeping bentgrass (Agrostis stolonifera L.) in nonagronomic habitats. Mol Ecol 15:4243–4255 Rotter D, Bharti AK, Li HM, Luo C, Bonos SA, Bughrara S, Jung G, Messing J, Meyer WA, Rudd S, Warnke SE, Belanger FC (2007) Analysis of EST sequences suggests recent origin of allotetraploid colonial and creeping bentgrasses. Mol Genet Genomics 278:197–209 Ruemmele BA (2003) Agrostis capillaris (Agrostis tenuis Sibth.) colonial bentgrass. In: Casler M, Duncan RR (eds) Turfgrass biology, genetics and breeding. Wiley, Hoboken, NJ, USA, pp 187–200 Saski C, Lee S-B, Fjellheim S, Guda C, Jansen R, Luo H, Tomkins J, Rognli OA, Daniell H, Clarke JL (2007) Complete chloroplast genome sequences of Hordeum vulgare, Sorghum bicolor and Agrostis stolonifera, and comparative analyses with other grass genomes. Theor Appl Genet 115:571–590 Scheef EA, Casler MD, Jung G (2003) Development of speciesspecific SCAR markers in bentgrass. Crop Sci 43:345–349

B.S. Ozdemir and H. Budak Tercek MT, Hauber DP, Darwin SP (2003) Genetic and historical relationships among geothermally adapted Agrostis (bentgrass) of North America and Kamchatka: evidence for a previously unrecognized, thermally adapted taxon. Am J Bot 90:130–1312 Tian J, Belanger FC, Huang B (2009) Identification of heat stress-responsive genes in heat-adapted thermal Agrostis scabra by suppression subtractive hybridization. J Plant Physiol 166:588–601 Vergara GV, Bughrara SS (2004) Genetic differentiation of tetraploid creeping bentgrass and hexaploid redtop bentgrass genotypes by AFLP and their use in turfgrass breeding. Crop Sci 44(3):884–890 Wang Z-Y, Ge Y (2005) Rapid and efficient production of transgenic bermudagrass and creeping bentgrass bypassing the callus formation phase. Funct Plant Biol 32:769–776 Wang Z-Y, Ge Y (2006) Invited review: recent advances in genetic transformation of forage and turf grasses. In Vitro Cell Dev Biol Plant 42:1–18 Wang D, Luthe DS (2003) Heat-sensitivity in a bentgrass variant: failure to accumulate a chloroplast heat shock protein isoform implicated for heat tolerance. Plant Physiol 133:319–327 Wang G, Skipper HD (2004) Identification of denitrifying rhizobacteria from bentgrass and bermudagrass golf greens. J Appl Microbiol 97:827–837 Wang Z-Y, Bell J, Ge YX, Lehmann D (2003) Inheritance of transgenes in transgenic tall fescue (Festuca arundinacea Schreb). In Vitro Cell Dev Biol Plant 39:277–282 Warnke S (2003) Creeping bentgrass (Agrostis stolonifera L.). In: Casler M, Duncan RR (eds) Turfgrass biology, genetics and breeding. Wiley, Hoboken, NJ, USA, pp 175–185 Watrud LS, Lee EH, Fairbrother A, Burdick C, Reichman JR, Bollman M, Storm M, King G, Van de Water PK (2004) Evidence for landscape level pollen-mediated gene flow from genetically modified creeping bentgrass using CP4 EPSPS as a marker. Proc Natl Acad Sci USA 101: 14533–14538 Wu C, Hsiang T (1998) Pathogenicity and formulation of Typhula phacorrhiza, a biocontrol agent of gray snow mold. Plant Dis 82:1003–1006 Xiao L, Ha SB (1997) Efficient selection and regeneration of creeping bentgrass transformants following particle bombardment. Plant Cell Rep 16:874–878 Xu Y, Huang B (2007) Heat-induced leaf senescence and hormonal changes for thermal bentgrass and turf-type bentgrass species differing in heat tolerance. J Am Soc Hortic Sci 132:185–192 Xu J, Huang B (2008a) Differential protein expression for geothermal Agrostis scabra and turf-type Agrostis stolonifera differing in heat tolerance. Environ Exp Bot 64:58–64 Xu C, Huang B (2008b) Root proteomic responses to heat stress in two Agrostis grass species contrasting in heat tolerance. J Exp Bot 59(15):4183–4194 Xu J, Tian J, Belanger FC, Huang B (2007) Identification and characterization of an expansin gene AsEXP1 associated with heat tolerance in C3 Agrostis grass species. J Exp Bot 58(13):3789–3796 Xu J, Belanger F, Huang B (2008) Differential gene expression in shoots and roots under heat stress for a geothermal and non-thermal Agrostis grass species contrasting in heat tolerance. Environ Exp Bot 63:240–247

1 Agrostis Xu Y, Tian J, Gianfagna T, Huang B (2009) Effects of SAG12-ipt expression on cytokinin production, growth and senescence of creeping bentgrass (Agrostis stolonifera L.) under heat stress. Plant Growth Regul 57:281–291 Yu TT, Skinner DZ, Liang GH, Trick HN, Huang B, Muthukrishnan S (2000) Agrobacterium-mediated transformation of creeping bentgrass using GFP as a reporter gene. Hereditas 133:229–233 Zapiola ML, Campbell CK, Butler MD, Mallory-Smith CA (2008) Escape and establishment of transgenic glyphosate-resistant creeping bentgrass Agrostis stolonifera in

13 Oregon, USA: a 4-year study. J Appl Ecol 45 (2):486–494 Zhao H, Bughrara SS, Oliveira JA (2006) Genetic diversity in colonial bentgrass (Agrostis capillaris L.) revealed by EcoRIMseI and PstI-MseI AFLP markers. Genome 49:328–335 Zhao H, Bughrara SS, Wang Y (2007) Cytology and pollen grain fertility in creeping bentgrass interspecific and intergeneric hybrids. Euphytica 156:227–235 Zhong H, Bolyard MG, Srinivasan C, Sticklen MB (1993) Transgenic plants of turfgrass (Agrostis palustris Huds.) from microprojectile bombardment of embryogenic callus. Plant Cell Rep 13:1–6

.

Chapter 2

Bromus W.M. Williams, A.V. Stewart, and M.L. Williamson

2.1 Introduction Bromus L. is a genus of approximately 150 C3 grass species (Clayton and Renvoize 1986; Watson and Dallwitz 1992) that can be considered to be intermediate between the Festuceae and the Triticeae. The genus is distributed widely in Asia, Europe, Africa, and the Americas, and today, introductions are widespread in the temperate world. There is a high incidence of polyploidy, species ranging from 2x to 12x. Many of the high polyploids are of allopolyploid (hybrid) origin (Stebbins 1981; Armstrong 1991). There is a diversity of annuals and perennials with a range of bunchgrass and rhizomatous morphologies.

2.2 Evolution and Systematics The genus is taxonomically difficult with several unresolved species complexes, especially in section Ceratochloa. For the purposes of this chapter, the classification of Bromus into seven sections (Smith 1970, 1985) will be used (Table 2.1). These sections are: Bromus (30–40 species), Genea Dumort. (seven species), Pnigma Dumort. (about 60 species), Ceratochloa (P. Beauv.) Griseb. (10–16 species), Neobromus (Shear.) Hitchcock (two species), and Nevskiella

W.M. Williams (*) AgResearch, Grasslands Research Centre, Private Bag 11008, Palmerston North 4442, New Zealand e-mail: [email protected]

(Krecz & Vved.) Tournay. (one species). Section Boissiera (Hochst. ex Steudel) P. M. Smith was included in section Bromus by Smith (1970) and separated as a separate section by Smith (1985). The major agricultural species come from sections Pnigma (B. inermis Leyss.) and Ceratochloa (B. catharticus Vahl, B. sitchensis Trin. in Bong). Polyploidy and hybridization feature strongly in the evolution of many species (Stebbins 1981). The sections are distinguished morphologically according to numbers of nerves in the glumes, spikelet shape, and lemma and awn morphology, as well as karyotypes, genome relationships (chromosome pairing), ploidal levels, and serological differences. Because of its complexity, no worldwide taxonomic treatment exists, but many regional descriptions and identification keys have been published, including those of Alaska (Mitchell 1967), Mexico and Central America (Soderstrom and Beaman 1968), North America (Wagnon 1952; Allred 1993; Pavlick 1995), South America (Pinto-Escobar 1981, 1986; Matthei 1986; Gutierrez and Pensiero 1998; Planchuelo and Peterson 2000), Malesia (Veldkamp et al. 1991), New Zealand (Forde and Edgar 1995), South-East Asia (Chen and Kuoh 2000), and Europe and North Africa (Spalton 2002, 2004). Stebbins (1981) hypothesized that Bromus probably arose in Eurasia, when the Festuceae and Triticeae were separating. The original Bromus species are extinct and were probably wiped out during the dramatic climatic fluctuations of the Pliocene and Pleistocene. During the Pliocene, differentiation of sections Neobromus, Ceratochloa and Pnigma occurred. Sections Neobromus and Ceratochloa spread to North America and later to South America. These sections became extinct in Eurasia, and even

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_2, # Springer-Verlag Berlin Heidelberg 2011

15

16 Table 2.1 Genus Bromus taxonomic delimitation by different authors [modified from Armstrong (1991)] Tzvelev (1976) Smith (1970, 1985) Stebbins (1981) 7 Genera 7 Sections 7 Subgenera Anisantha Genea Stenobromus Bromus Bromus Bromus Bromopsis Pnigma Festucaria Ceratochloa Ceratochloa Ceratochloa – Neobromus Neobromus Boissiera Boissiera Boissiera Nevskiella Nevskiella Nevskiella Littledalea

diploids and tetraploids became extinct in the New World, with the build-up of octoploids in Ceratochloa. In the Pleistocene, section Pnigma spread to Africa and North America, while polyploids increased in Eurasia. Sections Bromus and Genea have evolved more recently. The basic chromosome number is x ¼ 7, like other Poaceae. Chromosome and nuclear genome sizes have changed considerably during speciation (Armstrong 1991). Stebbins (1981) proposed that the trend was towards larger genomes. On this basis, section Ceratochloa would have the most primitive genome size. Comparisons of chloroplast restriction site maps of B. inermis with the cereals have shown that Bromus is closer to the Triticeae than to the Aveneae and, within the Triticeae, it is closer to barley than to wheat and rye (Pillay 1993, 1995). Attempts to intercross species from different sections indicate that the reproductive barriers between sections are strong. Although some intersection hybrids have been produced, none has shown pairing of the chromosomes from the different sections, indicating wide differentiation of the chromosomes (Armstrong 1991). Molecular marker and DNA sequence phylogenies are revealing that some of the old sections are artificial groupings that require revision. A study of 46 species, representing a wide sample of the species diversity in Bromus, was conducted by Saarela et al. (2007), using nuclear and chloroplast DNA sequences. This study showed that current systematic classifications do not fully reflect phylogeny within the genus. Some differences between nuclear and chloroplast sequence

W.M. Williams et al.

phylogenies require clarification, and much wider sampling of the species is also needed.

2.3 Agricultural Status About ten species are used in agriculture and many more have weed status. Two sections contribute important agricultural species: Section Pnigma has the winterhardy Bromus inermis group, B. inermis and B. biebersteinii R. & S. (2n ¼ 56), B. riparius Rehm. (2n ¼ 70), and B. erectus Huds. (2n ¼ 28) with more than 40 cultivars in North America and Europe. Additionally, this section also contains B. auleticus Trin. ex Nees (2n ¼ 42) used to a limited extent in southern Brazil and Uruguay. Section Ceratochloa has the hexaploid B. catharticus complex (2n ¼ 42), including B. stamineus E. Desv. and B. valdivianus Phil. with over 30 cultivars in Argentina, France, Chile, New Zealand, Australia, and USA. The octoploid B. carinatus Hook. and Arn. complex (2n ¼ 56), including B. sitchensis and B. marginatus Nees ex Steud., has over five cultivars in France and USA, while the South American octoploid B. coloratus Steud. has one cultivar in Australia. The duodecaploid B. arizonicus (Shear) Stebbins (2n ¼ 84) has one very winter active annual cultivar, Cucamonga used in USA. In addition, some annual species of section Bromus have had cultivars developed for revegetation uses. For example, B. arvensis Guss. (2n ¼ 14) has a cultivar Dos in Russia and B. mollis L. (2n ¼ 28) has the cultivar Blando in USA. From section Genea, B. rubens L. has the cultivar Panache (Alderson and Sharp 1995). B. mango E.Desv. (section Ceratochloa) was used as a cereal grain crop in Chile. This plant was cultivated as a biennial cereal by the Araucana Indians of Chile until at least the middle of the 1800s. Thought to have become extinct (Scholz and Mos 1994), it is claimed to have been rediscovered and has been classified as a form of B. catharticus ssp. catharticus (Vahl) Herter (Massa et al. 2004). The potential agricultural value of many of the Bromus species for New Zealand was assessed by Rumball (1968) and Rumball and Forde (1976).

2 Bromus

2.4 Mediterranean and SW Asian Annual Species of Section Genea Section Genea (subgenus Stenobromus; Stebbins 1981) consists of seven annual species that are self-fertilizing, with ploidies ranging from 2x to 8x. The diploids are B. sterilis L., B. tectorum L., and B. fasciculatus C. Presl, and the tetraploids B. madritensis L. and B. rubens. In addition, there is a hexaploid B. rigidus Roth and an octoploid B. diandrus Roth. Among the diploid species, B. sterilis and B. tectorum are distinct and widespread in Europe and Eurasia while B. fasciculatus is restricted to the eastern Mediterranean region. However, the widespread tetraploids have variously been classified as subspecies (Sales 1994) or separate species (Oja 2002). The hexaploid and octoploid species occur in the Mediterranean region and southern and western Europe. They have also been variously classified as separate species, subspecies, or varieties, and as a polyploid species complex. Most of the species of this section have become weeds in many countries. An elegant molecular phylogenetic study (Fortune et al. 2008) has shown that all of the polyploid species of the section have hybrid origins and, apart from one unidentified lineage, all three other parental lineages were closely related to the three diploid species. Designating the diploid genomes as B. sterilis (SS), B. fasciculatus (FF), and B. tectorum (TT), the polyploids have been proposed (female donor first) as: B. madritensis (FFSS), B. rubens (FFTT), B. rigidus (XXFFTT), and B. diandrus (XXFFTTSS), where XX is an unidentified parent (Fortune et al. 2008). The two 4x species were demonstrated to be distinct species, and the 8x B. diandrus appears to have originated as a hybrid between 6x B. rigidus and 2x B. sterilis. This last observation has been supported by earlier research (Fortune et al. 2008). Species of this section have apparently donated genomes via hybridization with species of section Bromus to form the tetraploid B. pectinatus Thunb. complex (Stebbins 1981; Saarela et al. 2007) (see Sect. 2.6). The main characteristic of this group of species is their propensity for weediness. B. tectorum is generally considered to be cleistogamous and has become an invasive noxious weed in dry, open areas of North America, where it is called cheatgrass or downy chess. Simple sequence repeat (SSR) marker research has

17

indicated that B. tectorum can outcross, giving populations the capacity to absorb new genetic variation and potentially to increase their invasive capacity (Ashley and Longland 2007). The high polyploids, B. rigidus (6x) (ripgut brome) and B. diandrus (8x) (great brome) are severe weeds of crops in Australia (Kon and Blacklow 1990) and New Zealand (Dastgheib et al. 2003). These weeds form acceptable early spring forage in the early growth stages, but the presence of awns on mature seeds makes them a danger to the health of grazing animals in late spring and summer.

2.5 Section Pnigma Species of this large section (about 60) are native to Eurasia, Africa, and the Americas. Two distinct species groups were proposed by Armstrong (1981, 1983). The first was a group of long-lived perennials (rhizomatous or densely tufted) from Eurasia that were mostly polyploids (4x–10x) with large anthers and small chromosomes. In this group was B. pumpellianus Scribn., which also occurs in North America, as well as the Eurasian species B. inermis, B. erectus, B. variegatus M. Bieb., and B. riparius. The second group included mainly diploid, relatively short-lived, loosely tufted, non-rhizomatous species from North America characterized by small anthers and large chromosomes. This group included the B. ramosus Huds. complex from Eurasia (Armstrong 1981, 1983, 1991; Stebbins 1981). The Eurasian and North American species were reported to be difficult to hybridize (Armstrong 1983), supporting the existence of the groups and suggesting that they may have quite different evolutionary histories. However, further analyses of these species, while confirming the existence of groups with very different chromosome sizes, have not supported a simple evolutionary situation. Analysis has not been helped by difficult taxonomy in which similar taxa with widely different ploidy levels have been given the same species designations. Examples include B. variegatus, which is said to exist in a range of ploidy levels from diploid to decaploid. However, a decaploid is very unlikely to be an autoploid form derived from the diploid. It is far more likely that hybridization has been involved and that these taxa are

18

morphologically similar but probably have very different genomic constitutions. The existence of the two species groups with different chromosome sizes was established by hybridization studies and confirmed by nuclear DNA analyses. Hybrids were used to show that the chromosomes of the Eurasian B. ramosus complex were larger than the other Eurasian species B. inermis, B. variegatus, B. erectus, B. pumpellianus, and B. riparius. However, they were similar in size to B. pacificus Shear and B. richardsonii Link from North America. The North American species may all have variations of a single (L) genome. Interspecific hybrids were made between several of the diploid North American species and, despite good chromosome pairing, they were sterile. This was consistent with these species all containing the same (L) genome and in the process of differentiating by cryptic chromosomal rearrangements (Armstrong 1991). The Eurasian species have so far had two genomes (A and B) identified by cytogenetic analyses (Armstrong 1991) and there may be a third (C) (see Sect. 2.5.1). The A and B genomes are the two genomes in tetraploid forms of B. inermis and B. pumpellianus, one of which (A) is also close to that of tetraploid B. erectus. The B genome has some homology (homeology) with the first but is of uncertain origin. Armstrong (1983) suggested that the species of the B. ramosus complex may be the progenitors, or be related to the progenitors, of the American species. Sutkowska and Mitka (2008) also hypothesized that this Eurasian complex is the source of the L genome. These authors suggested that the L genome may have originated from the B genome, in line with the suggestion (Stebbins 1981; Armstrong 1983) that the large chromosome species were derived from small chromosome species, possibly before the migration to America, and before the evolution of the polyploid species currently found in Eurasia. In support of the existence of two groups of species, there appeared to be little or no pairing homology in hybrids between the large chromosomes of B. ramosus and B. pacificus and the small chromosomes of the Eurasian polyploids (Armstrong 1984). However, there was also little pairing homology between the chromosomes of B. ramosus and B. pacificus, possibly emphasizing the difficulties of using pairing affinities in tetraploids, where genetic control of chromosome pairing can occur (Armstrong 1984).

W.M. Williams et al.

Sutkowska and Mitka (2008) used random amplified polymorphic DNA (RAPD) analyses to obtain evidence consistent with B. erectus and its close relatives being the oldest group of species in the section. This evidence was also consistent with B. variegatus having the B genome. The inconsistencies in chromosome behavior are reflected by the results of molecular phylogenetic analyses. Nuclear DNA (internal transcribed spacer, ITS) sequence analysis has shown that section Pnigma is far from being monophyletic (Saarela et al. 2007). The Eurasian species of section Pnigma did not occur in one clade and the American species formed several well defined clades. The molecular data indicated that section Pnigma is probably an artificial group and that its current status as a section may be inappropriate. This was supported by the results of C-banding and DNA content analyses (Tuna et al. 2005). However, care is needed as reticulate evolution may have mixed the ancestral genomes such that they are apparently now distributed between continents (Sutkowska and Mitka 2008). There are many conflicting observations and there is a clear need to use further molecular analyses to resolve the species relationships in this artificial group so that more effective use of the potentially valuable wild relatives can be achieved. Analysis of chloroplast DNA using species-specific restriction fragment length polymorphisms (RFLPs) following interspecific hybridization has indicated uniparental maternal inheritance in Bromus species so far investigated (Pillay and Armstrong 2001). In crosses between B. arvensis (2n ¼ 14) as female and B. inermis (2n ¼ 28, 56), and B. erectus (2n ¼ 42, 56, 70), all F1 plants had the female chloroplast restriction patterns. No paternal or biparental inheritance was detected. In summary, there are about 35 Eurasian and 25 American species in this section. To-date, very few species have been studied, and conclusions about the evolution of the groups may be clarified as more species are given close research attention (Armstrong 1984).

2.5.1 B. inermis (Smooth Bromegrass, Russian Brome) One of the most important agricultural species, B. inermis, belongs to this section. ITS DNA sequence

2 Bromus

analysis (Saarela et al. 2007) placed this species in a small clade with four other Eurasian species, B. erectus, B. korotkoyi Drob., B. pumpellianus (also native to North America), and B. riparius. Somewhat surprisingly, these formed a sister clade to a larger group comprising all of the monophyletic sections Genea, Neobromus, and Ceratochloa, as well as B. brachyanthera Doll. from South America. The form commonly grown in North America is autoallooctoploid (2n ¼ 8x ¼ 56). There are also allotetraploid (2n ¼ 4x ¼ 28) and hexaploid (2n ¼ 6x ¼ 42) forms (Tan and Dunn 1977). The >250 accessions held by USDA Plant Germplasm System are predominantly 8x, with a few 4x and no 6x forms (Tuna et al. 2001). The tetraploid has regular meiosis (Carnahan and Hill 1960). The octoploid is irregular at meiosis, forming mostly quadrivalents and bivalents. A chlorophyll mutant has exhibited tetrasomic inheritance and an intermediate chromosome-chromatid type of segregation pattern (Ghosh and Knowles 1964). The genomic constitutions of the octoploid and tetraploid have been proposed to be AAAABBBB and AABB, respectively, and the A and B genomes appear to be closely related (Armstrong 1979). Giemsa C-banding has confirmed the alloploid nature of the tetraploid. However, karyotype analysis and Giemsa C-banding has indicated that the octoploid cytotype is not the same as a doubled version of the tetraploid cytotype (Tuna et al. 2004). There is evidence that the genomic constitution could be AAAABBCC, based on karyotype analyses showing that B. inermis has two pairs of chromosomes with large satellites and only one pair with small satellites (Ghosh and Knowles 1964; Wilton 1965; Armstrong 1973). However, this characteristic appears to be either inconsistent or polymorphic (Rychlewski 1970; Armstrong 1981). Tuna et al. (2004) confirmed the inconsistency of expression of the small satellite chromosomes and generally also confirmed the karyotypic analysis of Armstrong (1977b) showing two large satellite pairs and one small satellite pair. These authors indicated that the AAAABBBB hypothesis of Armstrong (1977b) was not supported by their chromosome analyses. To-date, the genome of B. inermis has remained relatively intractable to cytogenetic analyses using karyotype and C-banding. There is a need for the application of fluorescence in situ hybridization (FISH) and genomic in situ hybridization (GISH) techniques.

19

Interspecific hybrids indicate that the A genome may come from B. erectus (2n ¼ 4x ¼ 28) (Armstrong 1991; Walton 1980). However, a similar genome also occurs in a diploid form of B. variegatus Bieb. and a diploid that resembles B. inermis (Armstrong 1991). Hybrids between diploid B. variegatus and tetraploid and octoploid B. inermis had chromosome pairing patterns that suggested that the B. variegatus genome was not the same as either the A or B genome but was similar to both (Armstrong 1984). Karyotype and Giemsa C-banding analyses have indicated that the chromosomes of both B. erectus and B. variegatus differ from each other and from those of B. inermis. Therefore, if either species is a progenitor of B. inermis, significant chromosomal change should have occurred post-hybridization and polyploidization (Tuna et al. 2006). Natural B. erectus x B. inermis hybrids have been reported from Ukraine (Sutkowska et al. 2002). Tetraploid and octoploid forms of B. inermis have 2C DNA contents of 11.74 pg and 22.15 pg (Tuna et al. 2001). These values are significantly lower than those that would have resulted from multiples of the proposed diploid progenitors. Therefore, it is likely that significant DNA loss has occurred during polyploidization. Such loss of DNA has been recorded for the Triticeae (Vogel et al. 1999). Hybrids with B. pumpellianus ssp. dicksonii indicated that the two species have similar chromosomes, differing only by inversions and translocations (Armstrong 1982). 2.5.1.1 Genetic Diversity in B. inermis Based on molecular analyses, the present distribution of B. inermis in Poland was attributed to post-glacial migrations from two separate refuges (Sutkowska et al. 2002). North American breeders of B. inermis generally classify the natural variation of the species into two ecotypes – a northern or “meadow” type adapted to valleys and moist regions of eastern Europe and temperate Asia, and a southern, or “steppe” type adapted to dry steppe regions. A third, intermediate group has developed apparently by intermixing of the first two. The ecotypes differ morphologically in root depth and leaf size (Fernandez and Coulman 2004). Amplified fragment length polymorphism (AFLP) analyses of 14 cultivars revealed that the

20

older (pre-1980) varieties based on southern germplasm were distinctive and less diverse than varieties based on northern germplasm and more recent southern or mixed varieties. These results suggest that intermixing of southern and northern ecotypes has occurred recently in North American breeding programs (Fernandez and Coulman 2004). A diversity analysis of cultivars based on morphological characteristics (Casler et al. 2000) indicated that similar morphological types had been developed from very diverse genetic backgrounds. 2.5.1.2 Breeding Progress In recent years, there has been very little private sector breeding of B. inermis and relatively little cultivar development by public sector plant breeders. Emphasis on selection for improved quality has significantly improved digestibility and reduced fiber concentrations. Yield improvement has been slow, not only because of the low effort but also because of the complex polyploid genetics and research emphasis on genetics rather than breeding (Casler et al. 2000). There has been almost no molecular breeding in Bromus. One of very few studies investigated the prospective use of marker-assisted selection for forage quality using neutral detergent fiber (NDF) concentration as a predictor of animal intake (Diaby and Casler 2005; Stendal et al. 2006). Although RAPD markers were difficult to use because of low repeatability, the association of markers with NDF was established. This has opened the way for marker-assisted selection for NDF concentration using more repeatable marker systems. B. inermis is one of the most freezing-tolerant perennial grass species. Consequently, it is the subject of considerable research into the genetics of freezing tolerance, as well as being a potential genetic resource for isolation of important freezing tolerance genes (see Sect. 2.10).

2.5.2 B. riparius Rehm. (Meadow Bromegrass) This species is often called B. biebersteinii and is confused with B. erectus, and its taxonomy is still to

W.M. Williams et al.

be resolved (Pavlick 1995). However, its description fits that for B. riparius according to Smith (1980). It is a long-lived perennial with chromosome numbers of 2n ¼ 2x ¼ 14, 2n ¼ 8x ¼ 56, and 2n ¼ 10x ¼ 70. In addition to being a useful species itself, it is a potential source of germplasm for the improvement of B. inermis, it has lower growing points and so recovers better from defoliation, is less aggressively rhizomatous, and has a longer growing season into the autumn than B. inermis (Armstrong 1991). The common cultivars have 2n ¼ 10x ¼ 70 and are believed to have the same chromosome constitution as B. inermis plus an additional genome of unknown origin (Armstrong 1991). Sutkowska and Mitka (2008) obtained evidence using RAPDs that this additional genome could be closely related to, or ancestral to, the L genome. Diploid B. riparius has a somatic cell (2C) DNA content of 6.14  0.09 pg (Tuna et al. 2001) – very similar to B. erectus (Tuna et al. 2006). Common B. riparius has a 2C DNA content of 22.15 pg (Tuna et al. 2001). The diploid form of B. riparius was found in Kazakhstan and is quite similar to the tetraploid B. inermis from the same region. This led Armstrong (1987) to suggest that diploid B. riparius might be a progenitor of polyploid B. inermis. A C-banding analysis of diploid B. riparius was completed by Tuna et al. (2001), and Tuna et al. (2004) analyzed the respective karyotypes and showed that B. riparius is an unlikely progenitor of B. inermis. Sterile hybrids between B. variegatus and diploid B. riparius have been produced and were reported to have normal chromosome pairing behavior (Armstrong 1991). B. riparius is widely grown in North-central USA and in Canada. It resembles B. inermis in appearance but has shorter rhizomes, awned seeds, and pubescent leaves (Knowles et al. 1993). The cultivar “Regar” was released in 1966 by the Colorado Experiment Station. Introgression of B. riparius traits (e.g., lower growing points and less vigorous rhizomes) into B. inermis would be useful. To this end, Armstrong (1990) obtained fertile hybrids between B. inermis (2n ¼ 8x ¼ 56), as female, and B. riparius (2n ¼ 10x ¼ 70) without difficulty. The F1 plants were backcrossed to B. inermis as male and produced seed when open-pollinated. They had 2n ¼ 63 chromosomes. The F2 population ranged from 2n ¼ 56 to 72 and tended to be in the 63–70 range. Chromosome

2 Bromus

pairing results were equivocal and there was no clear evidence for or against recombination between the genomes of the two species. However, there was a suggestion that reversion to the parental types occurred, as had been observed earlier by Nielsen et al. (1965). Backcross progenies rapidly reverted to the parental chromosome numbers and appeared to lose B. riparius traits as they did so. Cytoplasmic effects were also apparent, and no clear strategy emerged for the use of B. riparius germplasm in the improvement of B. inermis. Nevertheless, a hybrid population was selected and interpollinated for several generations by Knowles and Baron (1990). Selections from this population were analyzed using RAPD markers and compared with the parental cultivars (Fernandez et al. 2001). A hybrid population was genetically intermediate between the parents, although closer to B. inermis, possibly because it had been selected for B. inermis traits. Thus, it is possible that recombination had occurred between B. inermis and B. riparius genomes in that population and that similar hybrids may be useful for the genetic improvement of smooth bromegrass.

21

and other disturbed places, as well as in pastures. It is often confused with B. riparius. Because it probably shares genomes with B. inermis, B. erectus is a potential source of genetic diversity for B. inermis breeding. It also shares a genome with B. pumpellianus (Armstrong 1981).

2.5.4 B. variegatus B. variegatus forms part of the meadow bromegrass complex of decaploid (2n ¼ 70) species, along with B. riparius and B. biebersteinii. A diploid population of this species is one of very few diploids among the Eurasian species of the section. It has a somatic cell DNA content of 6.76  0.05 pg (Tuna et al. 2006). As already discussed, the role of the diploid form in the ancestry of the polyploid species remains unresolved.

2.5.5 B. pumpellianus (Arctic Bromegrass) 2.5.3 B. erectus (Erect Bromegrass) This diploid perennial species has a somatic nucleus (2C) DNA content of 6.19  0.08 pg (Tuna et al. 2006). Armstrong (1973) used B. erectus (2n ¼ 4x ¼ 28) to make 6x hybrids with B. inermis (2n ¼ 8x ¼ 56). On the basis of chromosome pairing and karyotype analysis, he concluded that the B. erectus genome was present in tetrasomic condition in B. inermis. Armstrong (1977b) carried out further analysis of the karyotypes of these hybrids and B. inermis and favored the view that the B. erectus genome constituted the A genome in 8x (AAAABBBB) B. inermis. However, he was unable to rule out the AAAABBCC hypothesis of Ghosh and Knowles (1964). B. erectus is naturally distributed in Europe, Britain, Ireland, and North Africa (Meusel et al. 1965) and is introduced and widespread in the North-Eastern states of USA and sporadically elsewhere (Pavlick 1995). The present distribution of B. erectus in Poland was studied by polymerase chain reaction (PCR) and attributed to a single post-glacial expansion (Sutkowska et al. 2002). It is a tufted perennial that occurs on roadsides

This octoploid (2n ¼ 8x ¼ 56) species has populations that are native to Asia, as well as the mountains of western North America, where it provides useful high quality forage in high altitude grasslands up to 3,350 m altitude (Casler and Carlson 1995). Hybrids between 8x B. inermis and 8x B. pumpellianus are fertile, suggesting that they are subspecies (Armstrong 1991). However, there is confusion because, at the tetraploid level, they appear to be different species. Armstrong (1985) suggested that these are very heterogeneous taxa that involve allopolyploidy and have undergone introgression. A variety, “Polar” was developed from such interspecific hybrids (Hodgson et al. 1971).

2.5.6 Other North American Species in Section Pnigma Nodding brome (B. anomalus Rupr. ex Fourn.) and tetraploid (2n ¼ 28) fringed brome (B. ciliatus L.)

22

provide native grazing in the western mountain regions of North America (Casler and Carlson 1995). A diploid form of B. ciliatus is considered to be the ancestral American species (Armstrong 1991). A detailed Giemsa C-banding chromosome analysis of B. ciliatus by Tuna et al. (2005) revealed that the genomes in this American species are quite different from those of any European species in the section Pnigma. In addition, the 2C nuclear DNA content of B. ciliatus was 19.13 pg, contrasting strongly with that of tetraploid B. inermis (11.74 pg). The chromosomes of B. ciliatus were almost double the length of those of Eurasian species – consistent with previous observations that the two groups of species differ in chromosome size. Additionally, the C-banding indicated quite different patterns of constitutive heterochromatin. These observations, along with the strong genetic isolation between the two groups, support the contention that they are not closely related. B. ciliatus is of interest in North America as a potential outcrossing native grass species for restoring vegetation to ecologically important sites. Genetic diversity of Canadian populations was studied by Fu et al. (2005) using AFLPs. This identified regional Canadian forms. However, there was surprisingly little intrapopulation variation, suggesting that perhaps more self-fertilization occurred than was expected. This emphasized the need for proper breeding system studies to be carried out where genetic diversity is an important element in the end-use of a species. Although most North American species are diploid, B. frondosus (Shear) Woot. and Standl., B. richardsonii, and B. pacificus all behave like allotetraploids (Armstrong 1984). B. mucroglumis Wagnon from Mexico and adjacent areas of the USA is similar to B. richardsonii (Peterson et al. 2001), but the other three tetraploid species in North America do not appear to have been studied. Hybrids between tetraploid B. ciliatus and B. frondosus indicated similar genomic structures (Barnett 1957).

2.5.7 South American Species in Section Pnigma Native to southern Brazil and Uruguay, B. auleticus Trin. ex Nees is a perennial allogamous hexaploid (2n ¼ 6x ¼ 42) that provides outstanding forage in

W.M. Williams et al.

native pastures (Martinello and Schifino-Wittmann 2003). It has caught the attention of agronomists and plant breeders, and some isozyme and RAPD characterizations have been done (Yanaka 2002). B. urugayensis is also hexaploid, while B. macranthus has been reported as having 4x, 8x, 10x, and 16x chromosomes (Stebbins 1981). It is uncertain how these species are related to the remainder of section Pnigma.

2.6 Section Bromus This section of about 40 species is considered to be the most advanced. It is native to Europe and Asia and consists of annual or biennial diploid and tetraploid species. The tetraploids are likely to have hybrid origins (Stebbins 1981). Southwest Asia and the eastern Mediterranean regions are the centers of diversity (Stebbins 1981). Several annual or biennial species have spread widely beyond their native regions and have become significant weeds, e.g., B. briziformis Fisch. and C.A. Mey., B. commutatus Schrad. (meadow brome), B. hordeaceus L. (B. mollis), B. japonicus Thunb. in Murr., B. racemosus L., B secalinus L., and B. squarrosus L. are now widespread throughout North America (Pavlick 1995). Only a small number of artificial interspecific hybrids have been made within section Bromus (Armstrong 1991). B. mollis was crossed with B. arenarius (Knowles 1944), leading to an indication that B. mollis contains two distinct genomes and that the same two occur in B. racemosus. One of the genomes in B. mollis was partly homologous with one of the genomes in B. arenarius (4x). B. commutatus, B. macrostachys, and B. secalinus are allotetraploids. Hybrids of B. arvensis (2x) x B. commutatus (4x) and B. secalinus (4x) showed some chromosome pairing, indicating a relationship between the genome of B. arvensis and those in the tetraploid species (Jahn 1959). There were also pairing homologies between some chromosomes of B. macrostachys and B. mollis. Sections Boissiera (one species, B. pumilio (Trin.) P.M. Smith) and Triniusia (two species, B. danthoniae Trin. ex C.A. Mey., B. pseudodanthoniae Drobow), recognized by some authors (Smith 1985; Scholz 1998; Saarela et al. 2007), were included in Section Bromus by Smith (1970). They comprised annual

2 Bromus

species from Asia and the eastern Mediterranean region. ITS and chloroplast DNA sequence analysis (Saarela et al. 2007) confirmed that the species of section Triniusia belong in section Bromus. The B. pectinatus complex (five 4x species ranging from Africa to Asia) is intermediate in morphology between sections Bromus and Genea, leading Stebbins (1981) to suggest that these species arose as hybrids between species of the two sections. The DNA sequence data of Saarela et al. (2007) support this idea. If this is correct, then section Bromus is monophyletic.

2.6.1 B. arvensis (Field Bromegrass) and Its Close Relatives This diploid (2n ¼ 14) winter annual species and its close relatives, B. japonicus and B. squarrosus, form a complex that is taxonomically difficult and has been subject to several different treatments. B. arvensis is allogamous (Oja et al. 2003) unlike the other two, which are almost exclusively self-fertilizing. Serological analysis (Smith 1972), isozyme analysis (Oja et al. 2003), and DNA analyses (Ainouche and Bayer 1997; Ainouche et al. 1999) indicated that B. japonicus and B. squarrosus might be sister species or a species complex distinct from B. arvensis. Oja et al. (2003) suggested that B. japonicus and B. squarrosus are possibly self-fertilizing derivatives of B. arvensis. However, a later morphological separation, using discriminant analysis to determine the most useful characters for distinguishing the taxa (Oja and Paal 2007), indicated that the three species could be fairly reliably separated using floral morphological traits, especially anther length and lemma width. The Mediterranean self-fertilizing diploid B. intermedius Guss. is also very similar in morphology to B. japonicus var. villosus (Oja 2005) and apparently also belongs to this species complex. This is supported by isozyme data (Oja 2005) and ITS sequence analysis (Ainouche and Bayer 1997). It may also be a selffertilizing derivative of B. arvensis (Oja 2005). Isozyme analyses (Ainouche et al. 1995) indicated that North African populations of B. intermedius and its close relative B. squarrosus were much less variable than the tetraploids B. hordeaceus and B. lanceolatus. Nevertheless, heterozygotes were present in higher

23

than expected frequencies, suggesting that outcrossing was a significant evolutionary factor in these predominantly self-fertilized diploids. B. arvensis is native to the Mediterranean and into southern and central Europe and is widespread in Eurasia (Oja and Paal 2007). It has been introduced to Australia and the USA in the 1920s where it became adapted to the corn-belt region and eastward. It provides excellent winter cover, soil stabilization, and green manure as a result of its extensive root system. B. japonicus and B. squarrosus are common in the Mediterranean and Southwest Asia and have become significant weeds in several countries. B. arvensis (2n ¼ 14) has been hybridized with the perennial species B. inermis (2n ¼ 56) and B. erectus (2n ¼ 28) of section Pnigma, but the hybrids were sterile (Armstrong 1977a). This intersectional hybridization does not appear to be a fruitful source of germplasm for the widening of the B. inermis or B. erectus gene pools (Armstrong 1977a). However, Armstrong (1977b) was able to use 5x hybrids between B. arvensis and B. inermis to analyze the alloploid gametic (AABB) chromosome complement of B. inermis because the B. arvensis chromosomes were much larger and could be clearly distinguished.

2.6.2 B. hordeaceus Also known as B. mollis, this is an aggressive species that has a wide climatic range that has enabled it to spread widely within Eurasia, Africa, the Americas, and Australia. It is a probable allotetraploid (2n ¼ 28, Stebbins 1981). It shows wide morphological variation indicative of ecotype differentiation and, in the Mediterranean region, may hybridize with B. lanceolatus., a closely related tetraploid (Ainouche et al. 1995). On the basis of its morphological variation, Pavlick (1995) separated the species into four subspecies. It has been the subject of analyses using isozymes to relate population genetic diversity to habitat diversity in Australia (Brown et al. 1974). This led to the identification of fitness-related polymorphism at the alcohol dehydrogenase locus (Brown et al. 1976). Isozyme analyses have also indicated quite high outcrossing rates in some populations (Brown et al. 1974). However, Ainouche et al. (1995) found high levels of intragenomic homozygosity in North African

24

populations, suggesting high selfing rates in this species (and B. lanceolatus). High levels of intergenomic heterozygosity in these two species were consistent with their allotetraploid genomic structure.

2.7 Section Ceratochloa Section or subgenus Ceratochloa is a small section consisting of up to 16 polyploid perennial and annual species. No diploid or tetraploid species are found in this section. The section Ceratochloa has the smallest genome size in the genus (Armstrong 1991). Hybridization is rife in this section, making species boundaries obscure and the taxonomy very difficult. Section Ceratochloa appears to be monophyletic and most species display almost identical chloroplast DNA sequences (Pillay and Hilu 1990, 1995), suggesting a similar maternal genome. Similarly, the ITS sequence and other nuclear DNA sequence variation is trivial (Saarela et al. 2007), a situation found in most post glacial grass expansions. This suggests that much of Ceratochloa is very recent and post glacial, a conclusion drawn by Stebbins (1981). The agricultural value of many of the species in section Ceratochloa was reviewed by Stewart (1996). The species of section Ceratochloa can be divided into four genomically distinct classes:

2.7.1 The Hexaploid B. catharticus Complex (2n ¼ 42) This group of species is endemic to South America (Massa et al. 2004), with members also present in Africa, B. leptocladis in South Africa and B. runssoroensis in highland Central Africa. The hexaploid species of section Ceratochloa are all found to be strict allopolyploids with genomic formula AABBCC (Stebbins and Tobgy 1944). These genomes are almost indistinguishable and have a nuclear DNA content of 12.7–15.1 pg (mean 1C ¼ 2.32 pg) (Klos et al. 2009). Morphological variation in these hexaploids is large and it is fair to say this group is the oldest of the four groups in Ceratochloa. Originally, this complex was separated into a number of species but, over

W.M. Williams et al.

time, many of these so-called species have become amalgamated into a hyper-variable species. In this sense, it may be fairer to say that the hexaploids are just over the “cusp” of speciating. It is unclear how the African and American species relate. The commonly named taxa that constitute this hexaploid group from South America (following Planchuelo and Peterson 2000) are B. catharticus, B. coloratus, B. lithobius, B. stamineus, B. mango, and B. tunicatus Phil. Using a multivariate analysis of this group, including morphological and molecular data, Massa et al. (2001, 2004) indicated that the variation was consistent with the existence of a single hexaploid species, B. catharticus, with two nearly continuous subspecies, B. catharticus ssp. catharticus (Vahl) Herter and B. catharticus ssp. stamineus (E. Desv.) Massa. B. catharticus ssp. catharticus includes, among others, the following older taxa: B. unioloides Kunth, B. tunicatus, and B. burkartii Munoz, as well as B. mango, while B. catharticus ssp. stamineus includes B. stamineus, B. coloratus, B. lithobius, B. fonkii Phil., and B. valdivianus Phil. These authors identified octoploid forms of B. coloratus and B. lithobius in South America, which they reclassified as B. coloratus Steud. These three newly defined taxa are separable using a simple three-step morphological key. There are also differences in adaptation, the populations of ssp. stamineus being perennial, predominantly cleistogamous and from humid places, while those of ssp. catharticus are facultatively cleistogamous and are from high altitudes in open woodlands of the precordillera as well as the lower slopes of the Andes. Species from this complex have been introduced to many countries and there are more than 30 cultivars marketed in Argentina, France, Chile, New Zealand, Australia, and USA.

2.7.2 Two Disjunct Octoploid Groups These are the B. carinatus Hook. and Arn. complex (2n ¼ 56), found in higher latitudes of North America, and the 8x B. coloratus referred to above, found in the higher latitudes of South America (Massa et al. 2004). They have a genomic constitution ABCL where L is a larger genome, probably from section Pnigma. DNA contents of B. carinatus (20.9–22.9 pg/cell) were more

2 Bromus

than 70% larger than B. catharticus (13.0–14.9 pg/ cell) (Joachimiak et al. 2001). Therefore, the octoploid species of the B. carinatus complex are considered to be intersectional amphidiploids between diploids of section Pnigma and hexaploid species of section Ceratochloa (Stebbins 1956). This was supported by Pillay (1996) who found rDNA sequences common between this group of species and B. inermis. The seven species of the B. carinatus complex are B. aleutensis Trin. ex Griseb., B. carinatus sens. str., B. marginatus, B. maritimus (Pip.) A.S. Hitch., B. polyanthus Scribn., B. sitchensis, and B. subvelutinus Shear. Based on intergrading morphologies and partial interfertilities, some taxonomists believe that all of these taxa belong to a single diverse species, B. carinatus sens. lato (e.g., Soderstrom and Beaman 1968; Stebbins 1981). However, Pavlick (1995) argued that the major differences in morphology and barriers of hybrid sterility justify separation into separate species. A revision of this group using molecular phylogenetic methods is required (Saarela et al. 2007). The main agricultural species of the group are as below.

2.7.2.1 B. carinatus (California Brome) This is annual, biennial, or a short-lived perennial with deep roots, strong leafy growth, and good seed production that is used for grazing animals. It is naturally distributed along the west coast of North America from the Canadian border region to Baja California. Two major varieties are often distinguished (e.g. Pavlick 1995) – var. carinatus generally west of the Sierra Nevada and Cascade mountain ranges and the more perennial var. hookerianus (Thurb.) Shear. to the east in the Columbia River Basin. The two intergrade and var. hookerianus intergrades with B. marginatus and B. subvelutinus at higher altitudes to the east.

2.7.2.2 B. marginatus (Mountain Brome) This perennial occurs generally at high altitudes throughout the western half of North America from Canada to N. Mexico. It intergrades with B. carinatus and B. subvelutinus to the West, B. aleutensis to the North, and B. polyanthus to the Southeast. Several

25

cultivars have been developed and are used in the NW USA and Canada, as well as in New Zealand.

2.7.2.3 B. sitchensis (Sitka or Alaska Brome) Native to mountain meadows of the Pacific coast of North America from Washington state northwards to Alaska. It is closely related to B. aleutensis and is used for pastures in northern Europe.

2.7.3 B. arizonicus (2n ¼ 84) (Arizona Brome) This duodecaploid is native to California and east to Arizona. Morphological and cytological analyses performed by Stebbins et al. (1944) indicated that B. arizonicus is an allopolyploid derived from B. catharticus and hexaploid B. trinii Desv. in Gay (B. berterianus Colla) or an unknown close relative of this species. B. berterianus is the representative of the section Neobromus, native to the Pacific coast of North and South America. The genomic constitution of B. arizonicus was postulated as ABCC’DE (Stebbins 1947). Klos et al. (2009) have shown that all twelve genomes in B. arizonicus are approximately the same average size (2.1–2.3 pg) and the same size as the genomes in the hexaploid Ceratochloa species. A winter active annual cultivar, Cucamonga, is used in California.

2.7.4 Duodecaploid (2n ¼ 84) Accessions Found in Andean Regions of South America Two 12x accessions investigated by Klos et al. (2009) differed markedly from B. arizonicus in nuclear DNA content and in chromosome size. In particular, these South American duodecaploids had some large chromosomes, making up about 37% of the genome. This group potentially has the genomic constitution ABCLLL (Klos et al. 2009) and is clearly different in constitution from the North American duodecaploid, B. arizonicus. These accessions may belong to the species B. ayacuchensis (Saarela et al. 2006), for which a chromosome number has not been determined.

26

W.M. Williams et al.

2.8 Section Neobromus

2.10.3 Protoplast Fusion

This section consists of two annual hexaploids native to the western coasts of North and South America., including B. gunckelli Matthei and B. berterianus (B. trinii, Pavlick 1995).

Fusion of protoplasts from hexaploid (2n ¼ 42) wheat cv. 99P embryo-derived callus and UV-irradiated protoplasts from tetraploid (2n ¼ 28) B. inermis resulted in asymmetric somatic hybridization (Xiang et al. 1999). Three albino hybrid explants were obtained with 42–54 chromosomes, including small chromosomes and chromosome fragments from B. inermis.

2.9 Section Nevskiella This section comprises one species, B. gracillimus Bunge from Asia and the eastern Mediterranean.

2.10 Role of Bromus Species in Crop Improvement Using Biotechnology 2.10.1 Tissue Culture Callus cultures from immature inflorescences of B. inermis have been grown and regenerated into green plants (Wattanasiri and Walton 1993). Significant somaclonal variation was detected for several plant traits, and although none of 10 somaclones was taller than the parent plant in a replicated trial, some had more tillers and larger leaf-stem ratios. It was concluded that this approach should aid the development of superior clones of B. inermis.

2.10.2 Cell Culture Non-embryogenic suspension cell cultures of B. inermis have been used for over 20 years and this system has been thoroughly characterized (Ishikawa et al. 1990; Robertson et al. 1994; Wilen et al. 1996). It has proved to be useful for analysis of the mechanisms of stress tolerance induced either by drought or cold treatment or by abscisic acid (ABA) in warm conditions (e.g., Lee and Chen 1993; Wu et al. 2005). Nakamura and Ishikawa (2006) have successfully transformed B. inermis cell cultures using Agrobacterium tumefaciens, opening the way for the analysis of gene functions through reverse genetics.

2.10.4 Genome Mapping The B. inermis chloroplast genome was physically mapped by Pillay (1993) using barley and wheat chloroplast DNA probes to map restriction sites. The organization of the Bromus chloroplast genome was similar to other grasses, with a pair of inverted repeat regions flanked by single copy regions. DNA sequence colinearity was shared with wheat and other cereal grasses.

2.11 Bromus Species in Genetic Research 2.11.1 Herbicide Resistance B. tectorum (downy brome) has developed resistance to herbicides based on acetolactate synthase (ALS) – inhibitors and PS II inhibitors and ACCase inhibitors. This multiple herbicide resistance was traced, in part, to a single amino acid substitution in protein product of the psbA gene, the target site of PS II inhibitors (Park and Mallory-Smith 2005).

2.11.2 Genetic Diversity and Weediness Molecular markers (mainly AFLP and SSR) have been used to study the population genetics of the spread of B. tectorum in North America. As noted earlier, although this species is treated as an obligate selfpollinator, molecular evidence indicates some outcrossing. Nevertheless, strong associations have been

2 Bromus

found between marker allele presence and phenotypic variation in several adaptive traits (Ramakrishnan et al. 2004, 2006), suggesting that marker diversity can be used to infer adaptive variation.

2.12 Endophytic Fungi in Bromus Many grasses host choke forming Clavicipitaceae endophytic fungi of the genus Epichlo¨e or their asexual derivatives Neotyphodium. The sexual stroma of Epichlo¨e causes “choke disease,” which prevents seedheads emerging. Non-choke inducing asexual Neotyphodium endophytes occur in many grasses such as perennial ryegrass (Lolium perenne) and tall fescue (Festuca arundinacea), and they provide the host with advantages such as improved insect resistance and grazing deterrence (Schardl and Phillips 1997). The Bromus species at present known to host endophytes appear to be restricted to section Pnigma, with reports for B. auleticus, B. brachyanthera, B setifolius, B. benekenii, B. erectus, B. ramosus, B. tomentellus, and B. anomalus (Groppe et al. 2002; Cabral et al. 2007). These endophytes can have significant ecological effects and may well provide insect resistance and other important agricultural effects (Novas et al. 2007). The full relevance of endophyte fungi in Bromus is not known. However, it will be important that endophytes are collected and studied. Germplasm collections will need to monitor and maintain endophytes where they are present.

2.13 Recommendations for Future Action Both partially domesticated groups identified in this review consist of allopolyploid hybrid species. Therefore, there are enormous untapped opportunities for using hybridization and ploidal manipulations among members of the various species groups to achieve new grass varieties for temperate agriculture. It is recommended that grass breeders investigate this approach to using the wealth of diversity present among the wild

27

species. However, this work should be supported by new DNA sequence analyses to resolve species relationships, especially in the large section Pnigma, where hybridization and polyploidy, followed by genomic changes, have mixed the genomes and blurred species relationships. There is an on-going need to collect Bromus species from the wild in both Eurasia and in the Americas in order to realize the potential offered by the genus for agricultural purposes.

References Ainouche ML, Bayer RJ (1997) On the origins of the tetraploid Bromus species (section Bromus, Poaceae): insights from internal transcribed spacer sequences of nuclear ribosomal DNA. Genome 40:730–743 Ainouche ML, Misset MT, Huon A (1995) Genetic diversity in Mediterranean diploid and tetraploid Bromus L. (section Bromus Sm.) populations. Genome 38:879–888 Ainouche ML, Bayer RJ, Gourret JP, Defontaine A, Misset MT (1999) The allotetraploid invasive weed Bromus hordeaceus L. (Poaceae): genetic diversity, origin and molecular evolution. Folia Geobot 34:405–419 Alderson J, Sharp WC (1995) Grass varieties in the United States. Lewis Publ, Boca Raton, FL, USA Allred KW (1993) Bromus section Pnigma in New Mexico, with a key to the bromegrasses of the state. Phytologia 74:319–345 Armstrong KC (1973) Chromosome pairing in hexaploid hybrids from Bromus erectus (2n ¼ 28) x B. inermis (2n ¼ 56). Can J Genet Cytol 15:427–436 Armstrong KC (1977a) Hybrids of the annual Bromus arvensis with perennial B. inermis and B. erectus. Z Pflanzenz€ucht 79:6–13 Armstrong KC (1977b) Karyotypic models for the A and B genomes of Bromus inermis. Z Pflanzenz€ucht 78:244–252 Armstrong KC (1979) A and B genome homologies in tetraploid and octaploid cytotypes of Bromus inermis. Can J Genet Cytol 21:65–71 Armstrong KC (1981) The evolution of Bromus inermis and related species of Bromus sect. Pnigma. Bot Jahrb Syst Pflanzengesch Pflanzengeogr 102:427–443 Armstrong KC (1982) Hybrids between the tetraploids of B. inermis and B. pumpellianus. Can J Bot 60:476–482 Armstrong KC (1983) The relationship between some Eurasian and American species of Bromus section Pnigma as determined by the karyotypes of some F1 hybrids. Can J Bot 61:700–707 Armstrong KC (1984) Chromosome pairing affinities between Old and New World species of Bromus. Can J Bot 62: 581–585 Armstrong KC (1985) Chromosome pairing failure in an intersectional amphiploid of Bromus altissimus x B. arvensis. Can J Genet Cytol 27:705–709

28 Armstrong KC (1987) Chromosome numbers of perennial Bromus species collected in the USSR. Can J Plant Sci 67:267–269 Armstrong KC (1990) Cytology of F1 hybrids and chromosome number of F2 and BC1 progeny of the cross Bromus riparius x B. inermis. Theor Appl Genet 79:137–142 Armstrong KC (1991) Chromosome evolution of Bromus. In: Armstrong KC, Tsuchiya T, Gupta P (eds) Chromosome engineering in plants: genetics, breeding, evolution, Part B. Elsevier Science Publ, Amsterdam, Netherlands, pp 366–377 Ashley MC, Longland WS (2007) Microsatellite evidence of facultative outcrossing in cheatgrass (Bromus tectorum): implications for the evolution of invasiveness. Plant Species Biol 22:197–204 Barnett FL (1957) Cytogenetics of interspecific hybrids in the Bromopsis Section of Bromus. I. Diploid and tetraploid hybrids. Agron J 49:77–82 Brown AHD, Marshall DR, Albrecht L (1974) The maintenance of alcohol dehydrogenase polymorphism in Bromus mollis L. Aust J Biol Sci 27:545–559 Brown AHD, Marshall DR, Munday J (1976) Adaptedness of variants at an alcohol dehydrogenase locus in Bromus mollis L. Aust J Biol Sci 29:389–396 Cabral D, Iannone LJ, Stewart AV, Novas MV (2007) The distribution and incidence of Neotyphodium endophytes in native grasses from Argentina and its association with environmental factors. 6th international endophyte symposium, 26–28 March 2007, Christchurch, Australia, pp 79–82 Carnahan HL, Hill HD (1960) The nature of polyploidy in smooth bromegrass: Bromus inermis Leyss. J Hered 51:43–44 Casler MD, Carlson IT (1995) Smooth bromegrass. In: Barnes RF, Nelson CJ, Collins M, Moore KJ (eds) Forages, vol I, An introduction to grassland agriculture. Iowa State Univ Press, Ames, IA, USA, pp 313–324 Casler MD, Vogel KP, Balasko JA, Berdahl JD, Miller DA, Hansen JL, Fritz JO (2000) Genetic progress from 50 years of smooth bromegrass breeding. Crop Sci 40:13–22 Chen C-H, Kuoh C-S (2000) The genus Bromus L. (Poaceae) in Taiwan: a DELTA database for generating key and descriptions. Taiwania 45:311–322 Clayton WS, Renvoize SA (1986) Genera Graminum: grasses of the world. Kew Bull Add Ser 13:1–389 Dastgheib F, Rolston MP, Archie WJ (2003) Chemical control of brome grasses (Bromus spp.) in cereals. NZ Plant Protec 56:227–232 Diaby M, Casler MD (2005) RAPD marker variation among divergent selections for fiber concentration in smooth bromegrass. Crop Sci 45:27–35 Fernandez YSN, Coulman BE (2004) Genetic relationships among smooth bromegrass cultivars of different ecotypes detected by AFLP markers. Crop Sci 44:241–247 Fernandez YSN, Somers DJ, Coulman BE (2001) Estimating the genetic relationship of hybrid bromegrass to smooth bromegrass and meadow bromegrass using RAPD markers. Plant Breed 120:149–153 Forde MB, Edgar E (1995) Checklist of pooid grasses naturalised in new Zealand. 3. Tribes Bromeae and Brachypodiaceae. NZ J Bot 33:35–42

W.M. Williams et al. Fortune PM, Pourtau N, Viron N, Ainouche ML (2008) Molecular phylogeny and reticulate origins of the polyploidy Bromus species from section Genea (Poaceae). Am J Bot 95:454–464 Fu YB, Coulman B, Ferdinandez Y, Cayouette J, Peterson PM (2005) Genetic diversity of fringed brome (Bromus ciliatus) as determined by amplified fragment length polymorphism. Can J Bot 83:1322–1328 Ghosh AN, Knowles RP (1964) Cytogenetic investigations of a chlorophyll mutant in bromegrass, Bromus inermis Leyss. Can J Genet Cytol 6:221–231 Groppe K, Steinger T, Sanders I, Schmid B, Wiemken A, Boller T (2002) Interaction between the endophytic fungus Epichloe¨ bromicola and the grass Bromus erectus: effects of endophyte infection, fungal concentration and environment on grass growth and flowering. Mol Ecol 8:1827–1835 Gutierrez HE, Pensiero JF (1998) Sinopsis de las species Argentinas del genero Bromus (Poaceae). Darwiniana 35:75–114 Hodgson HJ, Wilton AC, Taylor RL, Klebesadel LJ (1971) Registration of Polar bromegrass. Crop Sci 11:939 Ishikawa M, Robertson AJ, Gusta LV (1990) Effect of temperature, light, nutrients and dehardening on abscisic acid induced cold hardiness in Bromus inermis Leyss suspension cultured cells. Plant Cell Physiol 31:51–59 Jahn A (1959) Cytologische untersuchungen an k€unstlichen artbastarden in der gattung Bromus L. Z Pflanzenz€ucht 42:25–50 Joachimiak A, Kula A, Sliwinska E, Sobieszczanska A (2001) C-banding and nuclear DNA amount in six Bromus species. Acta Biol Crac Bot 43:105–115 Klos J, Sliwinska E, Kula A, Golczyk H, Grabowska-Joachimiak A, Ilnicki T, Szostek K, Stewart A, Joachimiak AJ (2009) Karyotype and nuclear DNA content of hexa-, octo-, and duodecaploid lines of Bromus subgen. Ceratochloa. Genet Mol Biol 32(3):528–537 Knowles PF (1944) Interspecific hybridization of Bromus. Genetics 29:128–140 Knowles RP, Baron VB (1990) Performance of hybrids of smooth bromegrass (Bromus inermis Leyss.) and meadow bromegrass (B. riparius Rehm.). Can J Plant Sci 70:330–331 Knowles RP, Baron VS, McCartney DH (1993) Meadow bromegrass. Agri Canada Publ 188/E Kon KF, Blacklow WM (1990) Polymorphism, outcrossing and polyploidy in Bromus diandrus and B. rigidus. Aust J Bot 38 (6):609–618 Lee SP, Chen THH (1993) Molecular cloning of abscisic acidresponsive mRNAs expressed during the induction of freezing tolerance in bromegrass (Bromus inermis Leyss) suspension culture. Plant Physiol 101:1089–1096 Martinello GE, Schifino-Wittmann MT (2003) Chromosomes of Bromus auleticus Trin. ex Nees (Poaceae). Genet Mol Biol 26:369–371 Massa AN, Larson SR, Jensen KB, Hole DJ (2001) AFLP variation in Bromus section Ceratochloa germplasm of Patagonia. Crop Sci 41:1609–1616 Massa AN, Jensen KB, Larson SR, Hole DJ (2004) Morphological variation in Bromus section Ceratochloa germplasm of Patagonia. Can J Bot 82:136–144 Matthei O (1986) The genus Bromus L. (Poaceae) in Chile. Gayana Bot 43:47–110

2 Bromus Meusel H, Jager E, Weinert E (1965) Vergleichende Chlorologie der Zentraleuropaischen Flora 2. Karten, Veb Gustav Fischer, Jena, Germany Mitchell WW (1967) Taxonomic synopsis of Bromus section Bromopsis (Gramineae) in Alaska. Can J Bot 45:1309–1313 Nakamura T, Ishikawa M (2006) Transformation of suspension cultures of bromegrass (Bromus inermis) by Agrobacteriun temefaciens. Plant Cell Tissue Organ Cult 84:293–299 Nielsen EL, Drolsom PM, Jalal SM (1965) Evaluation of backcross and self-pollination progenies from interspecific hybridization of Bromus. Crop Sci 5:339–342 Novas VM, Collantes M, Cabral D (2007) Environmental effects on grass-endophyte associations in the harsh conditions of south Patagonia. Microb Ecol 61:164–173 Oja T (2002) Bromus fasciculatus Presl – a third diploid progenitor of Bromus section Genea allopolyploids (Poaceae). Hereditas 137:113–118 Oja T (2005) Isozyme evidence on the genetic diversity, mating system and evolution of Bromus intermedius (Poaceae). Plant Syst Evol 254:199–208 Oja T, Paal J (2007) Multivariate analysis of morphological variation among closely related species Bromus japonicus, B. squarrosus and B. arvensis (Poaceae) in comparison with isozyme evidences. Nord J Bot 24:691–702 Oja T, Jaaska V, Vislap V (2003) Breeding system, evolution and taxonomy of Bromus arvensis, B. japonicusand B. squarrosus (Poaceae). Plant Syst Evol 242:101–117 Park KW, Mallory-Smith CA (2005) Multiple herbicide resistance in downy brome (Bromus tectorum) and its impact on fitness. Weed Sci 53:780–786 Pavlick LE (1995) Bromus L. of North America. Royal British Columbia Museum, Victoria, Australia Peterson PM, Cayouette J, Ferdinandez YSN, Coulman B (2001) Recognition of Bromus richardsonii and B. ciliatus: evidence from morphology, cytology, and DNA fingerprinting (Poaceae: Bromeae). Aliso 20:21–36 Pillay M (1993) Chloroplast genome organization of bromegrass, Bromus inermis Leyss. Theor Appl Genet 86:281–287 Pillay M (1995) Chloroplast DNA similarity of smooth bromegrass with other pooid cereals: implications for plant breeding. Crop Sci 35:869–875 Pillay M (1996) Genomic organization of ribosomal RNA genes in Bromus (Poaceae). Genome 39:198–205 Pillay M, Armstrong KC (2001) Maternal inheritance of chloroplast DNA in interspecific crosses of Bromus. Biol Plant 44:47–51 Pillay M, Hilu KW (1990) Chloroplast DNA variation in diploid and polyploid species of Bromus (Poaceae) subgenera Festucaria and Ceratochloa. Theor Appl Genet 80: 326–332 Pillay M, Hilu KW (1995) Chloroplast-DNA restriction site analysis in the genus Bromus (Poaceae). Am J Bot 82: 239–249 Pinto-Escobar P (1981) The genus Bromus in northern South America. Bot Jahrb Syst 102:445–457 Pinto-Escobar P (1986) El genero Bromus en Los Andes Centrales de Suramerica. Caldasia 15:15–34 Planchuelo AM, Peterson PM (2000) The species of Bromus (Poaceae: Bromeae) in South America. In: Jacobs SWL, Everett J (eds) Grasses: systematics and evolution. CSIRO, Melbourne, Australia, pp 89–101

29 Ramakrishnan AP, Meyer SE, Waters J, Stevens MR, Coleman CE, Fairbanks DJ (2004) Correlation between molecular markers and adaptively significant genetic variation in Bromus tectorum (Poaceae), an inbreeding annual grass. Am J Bot 91:797–803 Ramakrishnan AP, Meyer SE, Fairbanks DJ, Coleman CE (2006) Ecological significance of microsatellite variation in western North American populations of Bromus tectorum. Plant Species Biol 21:61–73 Robertson AJ, Reaney MJT, Wilen RW, Lamb N, Abrams SR, Gusta LV (1994) Effect of abscisic acid metabolites and analogs on freezing tolerance and gene expression in bromegrass (Bromus inermis Leyss) cell cultures. Plant Physiol 105:823–830 Rumball W (1968) Patterns of variation in Bromus L. introductions. NZ J Agric Res 11:277–285 Rumball W, Forde MB (1976) Plant introduction trials. Performance of Bromus species at Palmerston North. NZ J Exp Agric 5:93–95 Rychlewski J (1970) Karyology of species of the genus Bromus. Acta Biol Crac Bot 13:23–35 Saarela JM, Peterson PM, Refulio-Rodriguez NF (2006) Bromus ayacuchensis (Poaceae: Pooideae: Bromeae), a new species from Peru, with a key to Bromus in Peru. Sida 22:915–926 Saarela JM, Peterson PM, Keane RM, Cayouette J, Graham SW (2007) Molecular phylogenetics of Bromus (Poaceae: Pooideae) based on chlorplast and nuclear DNA sequence data. Aliso 23:379–396 Sales F (1994) Evolutionary tendencies in some annual species of Bromus (Bromus L. sect. Genea Dum. (Poaceae)). Bot J Linn Soc 115:197–210 Schardl C, Phillips T (1997) Protective grass endophytes: where are they from and where are they going? Plant Dis 81: 430–438 Scholz H (1998) Notes on Bromus danthoniae and relatives (Gramineae). Willdenowia 28:43–150 Scholz H, Mos U (1994) Status and kurze Geschichte des ausgestorbenen Kulturgetreides Bromus mango E. Desv. – und die Genese des Bromus secalinus L. Flora 189:215–222 Smith P (1970) Taxonomy and nomenclature of the bromegrasses (Bromus L. s.l.). Notes Roy Bot Gard Edinb 30: 361–375 Smith PM (1972) Serology and species relationships in annual Bromes (Bromus L. sect. Bromus). Ann Bot 36:1–30 Smith PM (1980) Bromus L. In: Tutin TG, Heywood VH, Burges NA, Valentine DH (eds) Flora Europae, vol 5. Cambridge Univ Press, UK, pp 182–189 Smith PM (1985) Bromus L. In: Davis PH (ed) Flora of Turkey and the East Aegean Islands, vol 9. Edinburgh Univ Press, UK, pp 272–301 Soderstrom TR, Beaman JH (1968) The genus Bromus (Gramineae) in Mexico and Central America. Publ Mus Michigan State Univ Biol Ser 3:465–520 Spalton LM (2002) A new key to the tribe Bromeae in Britain and Ireland. BSBI Newsl 91:22–26 Spalton LM (2004) A key to Bromeae in Mediterranean climatic zones of southern Europe, southwest Asia, and North Africa. BSBI Newsl 95:22–26 Stebbins GL (1947) The origin of the complex of Bromus carinatus and its phylogeographic implications. Contrib Gray Herb Harv Univ 165:42–55

30 Stebbins GL (1956) Cytogenetics and evolution of the grass family. Am J Bot 43:890–905 Stebbins GL (1981) Chromosomes and evolution in the Genus Bromus (Gramineae). Bot Jahrb Syst 102:359–379 Stebbins GL, Tobgy HA (1944) The cytogenetics of hybrids in Bromus. 1. Hybrids within the section Ceratochloa. Am J Bot 31:1–11 Stebbins GL, Tobgy HA, Harlan JR (1944) The cytogenetics of hybrids in Bromus II. Bromus carinatus and Bromus arizonicus. Proc Calif Acad Sci 25:307–322 Stendal C, Casler MD, Jung G (2006) Marker-assisted selection for neutral detergent fiber in smooth bromegrass. Crop Sci 46:303–311 Stewart AV (1996) Potential value of some Bromus species of the section Ceratochloa. NZ J Agric Res 39:611–618 Sutkowska A, Mitka J (2008) RAPD analysis points to old world Bromus species as ancestral to new world subgen. Festucaria. Acta Biol Crac Bot 50:117–125 Sutkowska A, Wilk L, Mitka J, Joachimiak A (2002) DNA polymorphism in Bromus erectus and Bromus inermis from Poland’s territory. Zesz Probl Post Nauk Rol 488:187–195 (Polish with English summary) Tan GY, Dunn GM (1977) Mitotic instabilities in tetraploid, hexaploid, and octoploid Bromus inermis. Genome 19: 531–536 Tuna M, Gill KS, Vogel KP (2001) Karyotype and C-banding patterns of mitotic chromosomes in diploid bromegrass (Bromus riparius Rehm). Crop Sci 41:831–834 Tuna M, Vogel KP, Gill KS, Arumuganathan K (2004) C-banding analyses of Bromus inermis genomes. Crop Sci 44:31–37 Tuna M, Vogel KP, Arumuganathan K (2005) Genome size and Giemsa C-banded karyotype of tetraploid Bromus ciliatus L. Euphytica 146:177–182 Tuna M, Vogel KP, Arumuganathan K (2006) Cytogenetic and nuclear DNA content characterization of diploid Bromus erectus and Bromus variegatus. Crop Sci 46:637–641

W.M. Williams et al. Tzvelev NN (1976) Poaceae URSS Tribe 4 Bromeae Dum. USSR Academy of Science Press, Leningrad, USSR, pp 530–608 Veldkamp JE, Eriks J, Smit SS (1991) Bromus (Gramineae) in Malesia. Blumea 35:483–497 Vogel KP, Arumuganathan K, Jensen KB (1999) Nuclear DNA content of perennial grasses of the Triticeae. Crop Sci 39:661–667 Wagnon HK (1952) A revision of the genus Bromus, section Bromopsis, of North America. Brittonia 7:415–480 Walton PD (1980) The production characteristics of Bromus inermis Leyss and their inheritance. Adv Agron 33:341–369 Watson L, Dallwitz MJ (1992) The grass genera of the world. CAB Int, Wallingford, UK Wattanasiri C, Walton PD (1993) Effects of growth regulators on callus cell growth, plant regeneration, and somaclonal variation of smooth bromegrass (Bromus inermis Leyss). Euphytica 69:77–82 Wilen RW, Fu P, Robertson AJ, Abrams SR, Low NH, Gusta LV (1996) An abscisic acid (ABA) analog inhibits ABAinduced freezing tolerance and protein accumulation, but not ABA-induced sucrose uptake in a bromegrass (Bromus inermis Leyss) cell culture. Planta 200:138–143 Wilton AC (1965) Phylogenetic relationships of an unknown tetraploid. Bromus, section Bromopsis. Can J Bot 43:723–730 Wu G, Robertson AJ, Zheng P, Liu X, Gusta LV (2005) Identification and immunogold localization of a novel bromegrass (Bromus inermis Leyss) peroxisome channel protein induced by ABA, cold and drought stresses, and late embryogenesis. Gene 363:77–84 Xiang F-N, Xia G-M, Zhou A-F, Chen H-M, Huang Y, Zhai XL (1999) Asymmetric somatic hybridization between wheat (Triticum aestivum) and Bromus inermis. Acta Bot Sin 41:458–462 (Chinese with English abstract) Yanaka FY (2002) Caracterizacao molecular e isoenzimatica de acessos de Bromus auleticus Trinius ex Nees. MSc Dissert, Univ Fed Rio Grande do Sul, Porto Alegre

Chapter 3

Cenchrus S. Goel, H.D. Singh, and S.N. Raina

3.1 Introduction Grasses (Poaceae) feed the world either directly as food crops, such as wheat rice, millets, and other grains, or indirectly as primary fodder for the livestock. It is estimated that grasslands cover around 20% of the earth’s total land area. Cenchrus, a member of the tribe paniceae of the Poaceae family, is also one of the important components of major grass cover of the world. It is distributed throughout the arid and semiarid tropical regions of the world (Bogdan 1977). Considering the diversity and present-day distribution, Cenchrus probably originated in the eastern tropical Africa and tropical Asia and is widely naturalized in new world countries. Cenchrus ciliaris and two closely related species, C. setiger and C. pennisetiformis, are three main species of Cenchrus having the potential for forage production across the world. They are well adapted to harsh climatic conditions and flourish vigorously with the availability of soil moisture. C. ciliaris is more widespread and more valued as forage grass for dry areas because of its high biomass production and tolerance to low rainfall conditions. Some of the species of Cenchrus are also potential invader weeds. After invasion, they replace the native species by changing not only the ecophysiological characteristics but also the fire regime. The grass base of Royal Botanical Garden, Kew has recorded 25 species of Cenchrus. However,

S.N. Raina (*) Amity Institute of Biotechnology, Amity University, Sector 125, Noida, Uttar Pradesh, India e-mail: [email protected]

Germplasm Resource Information Network (GRIN) of USDA recorded 35 different species of Cenchrus. This genus belongs to the tribe Paniceae of Poaceae family. Majority of the taxa in the tribe Paniceae are polyploids with basic chromosome no x ¼ 9 or 10; however, basic chromosome number determination is almost impossible due to the presence of an aneuploid series in this taxa. This aneuploid series results in abnormal meiotic chromosome behavior during sexual reproduction, resulting in sterility and subsequently culminating in predominance of aposporic embryo development (Snyder et al. 1955). But both sexual and apomictic modes of reproduction have been reported in C. ciliaris.

3.2 Morphology, Taxonomy, and Geographical Distribution of Genetic Diversity The name Cenchrus comes from the Greek word “Kenchros” meaning millet. The Latin word Cenchrus was used by Plinius, for an Arabian diamond or an unknown kind of piece, big as a grain of millet (Quattrocchi 2008). Cenchrus belongs to family Poaceae (Gramineae), subfamily Panicoideae, tribe Paniceae. The subfamily Panicoideae includes approximately 208 genera and ~3,300 species. Cenchrus is an extremely variable genus having both annual and perennial species. It is tufted, sometimes shortly rhizomatous grass with branching culms. Culms are ascending, erect, or decumbent and stoloniferous, ranging from 68 to 300 cm in length (Clayton et al. 2006 onwards). Some species reach up to 9 feet in

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_3, # Springer-Verlag Berlin Heidelberg 2011

31

32

height. Leaves are with ligules, either with a ciliate membrane or fringe of hairs. Leaf blades are flat, linear, or lanceolate. It is a bisexual plant with an erect or nodding cylindrical spike-like panicle inflorescence with angular rachis. Panicle has an axis bearing deciduous spikelet clusters. Each cluster has one to eight spikelets subtended by a deciduous involucre of flattened and often spiny bristles, which are connate at the base. Spikelets are sessile, lanceolate or ovate, closely compressed, and unawned. Each spikelet has one basal sterile or male floret and one fertile (hermaphodite) floret. Each floret has one or two glumes, which are unequal, shorter, or longer than the adjacent lemmas. Lemmas are as long as the spikelet, membranous to coriaceous, the flat margins of which cover much of palea. Lower glume is minute or is sometimes suppressed. Lodicules are absent in some species and are with two lodicules in some other species. Flowers are with three anthers and two stigmas. Fruit is small, typical of grass, i.e., caryopsis with adherent pericarp, dorsally compressed. The shape of the fruit varies from ellipsoidal, oblong, or ovoid to obovoid. Endosperm is hard, is without lipids, and contains only simple starch grains. Morphologically, Cenchrus is very similar to Pennisetum with few species of Setaria namely Setaria grisebachii and Setaria macrostachya, which often lead to confusion for identification but can be differentiated from each other by the morphology of the involucral bristles. In Cenchrus, bristles are fused with each other at the base, forming a spiny cup, but they are free in Pennisetum (Stieber and Wipff 2000; Tu et al. 2002). In C. ciliaris, the extent of fusion is minimal, which led to its placement in the genus Pennisutum as Pennisetum ciliare (Fisher et al. 1954; Taliaferro and Bashaw 1966; Vielle-Calzada et al. 1996). The membership of the C. ciliaris in genus Cenchrus is justified by its close relationship with C. setiger. In Setaria, the spikelets are borne above the bristles, and the bristles remain attached to the rachis of the inflorescence even after the spikelets fall off, whereas in Cenchrus, the bristles fall off with the spikelet as a unit at maturity. This species comprises many ecotypes differing greatly in their morphogenetic characters across accessions collected from the same locality or natural distribution (Arshad et al. 2007). Even differing number of chromosomes has been discovered between the ecotypes (Hignight et al. 1991).

S. Goel et al.

All species of Cenchrus are very much similar in their morphology. The genus Cenchrus has been separated further to different species level mainly based on the nature of bristles and their degree of fusion. 1. Cenchrus argimonioides Trin., commonly known as agrimony sandbur in US and as kamanomano in Hawaii, is native to North Pacific regions of the world, i.e., western coast of the US and Hawaii Islands (GRIN Taxonomy USDA) and is endemic to this region (Quattrocchi 2008). This is a perennial species with robust culms, which are 25–200 cm long. Leaf blades are 16–41 cm long, 6–16 mm wide, green to bluish green in color. Panicle spiciform, loose, 7–28 cm long. Deciduous spikelets in clusters of 1–2 spikelets subtended by an involucre of numerous deciduous bristles, which connate 2–3 mm forming a cup at the base. Bristles are rigid, retrorsely scaberulous, ciliate and spiny (Clayton et al. 2006 onwards). Fertile spikelets with 1 basal sterile and 1 fertile floret. Basal sterile floret barren and is with significant palea. Lemma of lower sterile floret with 5 veins. 2. Cenchrus biflorus Roxb., also known as Indian Sandbur, is native to Indian subcontinent and tropical African countries (GRIN Taxonomy USDA; Clayton et al. 2006 onwards; Quattrocchi 2008). It is an annual species with tufted habit. The culms are 5–90 cm long, ascending to erect and are unequal in length. The upper internodes are longer than the lower (Clayton et al. 2006 onwards). Leaf blades 2–25 cm long, 2–7 cm wide. Panicle 2–15 cm long with axis bearing deciduous spikelet in clusters of 1–3 spikelets subtended by a rigid involucre of two whorls of bristles, which connate at the base for 0.5–1 mm, forming a diamond-shaped shallow disc at the base. Involucral bristles ovoid in shape, 4–11 mm long retrorsely barbellate and tenaciously prickly. Inner spines flattened and their abaxial surface with 1–3 grooves. Outer bristles numerous and spiny, shorter than the inner, rarely suppressed. Fertile spikelets with 1 basal sterile and 1 fertile floret. Basal sterile floret barren and is with significant palea. Lemma of lower sterile floret with 5 veins. This species is well adapted to hot and dry tropical areas and it might be a paleotropical invasive weed. Impressions of its seeds were found in

3 Cenchrus

the archeological site of Tichit Chekka III (Mauritania, west Africa) (Quattrocchi 2008). 3. Cenchrus brownii Roem. and Schult is native to coastal plains of southeastern US, Central America, and northern coast of southern America (GRIN Taxonomy USDA; Quattrocchi 2008). This species is distributed in tropical and temperate regions of Asia as well as Australia (Clayton et al. 2006 onwards). It is also commonly called as Green Sandbur or Slimbristle Sandbur. This is an annual species with tufted herbaceous habit. Culms 25–100 cm long, ascending or erect. Rooting at the decumbent base, culm nodes are sometimes found in this species (Quattrocchi 2008). Leaf blades 8–30 cm long, 4–11 mm wide. Panicle linear or oblong, 3–12 cm long. Deciduous spikelets in clusters of 2–3 spikelets subtended by involucral bristles connate at the base up to 2.5–4 mm above to form a cup. Inner bristles subequal to outer bristles with the longest slightly emergent, flattened, rigid retrorsely scaberulous and spinose. Fertile spikelets with one basal sterile and one fertile floret. Basal sterile floret barren and is with significant palea. Lemma of lower sterile floret with 5 veins. This species grows in sandy waste places, forest borders, savannah, on beaches, and near the ocean (Quattrocchi 2008). 4. Cenchrus calicalatus Cav., commonly known as Hillside burrgrass or large burrgrass, is native to tropical Southeast Asia, Australia, New Zealand, and Pacific Islands (GRIN Taxonomy USDA; Quattrocchi 2008). This perennial species grows in clumps forming dense mats with its short woody rhizome trailing or scrambling and rooting at the lower nodes. Culms erect or ascending, 30–200 cm long, sometimes reaching up to 300 cm. Leaf blade flat and narrowly lanceolate, 5–56 cm long and 1–19 mm wide. Panicle spiciform, loose and 4–23.5 cm long with axis bearing deciduous spikelet clusters of 1–3 spikelets. Involucral bristles connate at the base to form a disc, which extends above for 0.5 mm above. Bristles numerous, 8–12 in principle whorl. The inner bristles terete, rigid and spinose and are without grooves. Basal sterile florets male or barren, with or without significant palea. Lodicules are absent in this species. This species is found growing on poor soils, on rocky coasts and on open sunny areas.

33

5. Cenchrus ciliaris (L) Link, commonly known as Buffelgrass, is native to tropical Africa and Asia and is naturalized widely in subhumid and semiarid tropics and subtropics (GRIN Taxonomy USDA; Clayton et al. 2006 onwards; Quattrocchi 2008). It is also known with other synonyms “Pennisetum ciliare” and “Pennisetum cenchroids” (GRIN Taxonomy USDA). C. ciliaris is perennial species with tufted habit. Culms are 10–50 cm long at maturity and are ascending to erect, profusely branched, shrub-like in growth habit. Leaf blades 3–25 cm long, 4–10 mm wide and green or bluish green in color. Panicle spiciform, 2–14 cm long, bears spikelets in clusters of 1–4 spikelets. Involucral bristles are numerous and united at the base to form a shallow disc up to 0.5 mm above the rim. The bristles in C. ciliaris are antrorsely scaberulous and not prickly. The inner bristles are flexuous, filiform in shape, and exceed much in length than spikelets, one of them longer and stouter than the rest (DeLisle 1963). Basal sterile florets are barren and are without significant palea. Lemma of the sterile floret with 5 veins. This species grows mainly on sandy soils and alluvial plains, on shallow marginally fertile soils, along roadsides, rocky hillsides, hot and dry areas, and denuded arid lands (Quattrocchi 2008). 6. Cenchrus distichophyllus Griseb. is distributed in Caribbean Islands (Clayton et al. 2006 onwards). It is a perennial species with distinct rootstock habit. Culms are erect, 15–40 cm long. Leave sheaths are longer than adjacent internodes. Leaf blade 2.6–3 cm long, 1.5 mm wide. Panicle 2.5–4 cm long with axis bearing spikelet clusters of one spikelet. Involucral bristles connate at the base to form a cup. Basal sterile florets barren and are with palea. Lemma of lower sterile floret ovate and is with 3 veins (Clayton et al. 2006 onwards). 7. Cenchrus echinatus L., commonly known as Sandbur or hedhog sandbur, is found in tropical and subtropical regions of the world. This is an annual species with herbaceous to subshrubby habit. Culms are erect and 15–90 cm long. Leaf blades 4–25 cm long, 3–10 mm wide. Panicle linear and interrupted. Axis bears spikelet clusters of 2–3 spikelets. Involucral bristles connate for 2–5 mm at the base into a cup. The inner bristles flattened and outer whorl thinner than the inner

34

and are tereted. Basal serile florets barren and are with palea. Lemmas of the lower sterile floret are with 5 veins. Seed heads of this species are composed of spiny burrs. These spines cause problems in cultivated lands and are very irritating to humans. This species commonly grows on waste ground, coastal dunes near the ocean, open areas, roadsides, and sandy or limestone soils (Quattrocchi 2008). 8. Cenchrus elymoides is found distributed in Australia (Clayton et al. 2006 onwards). It is a perennial species with robust habit. Culms are 60–150 cm long. Leaf blades 14–34 cm long and 3.3–10.2 mm wide. Panicle spiciform, dense and 10.5–16 cm long. Spikelet clusters with 1–3 spikelets in a cluster. Involucral bristles connate for 0.5 mm below to form a disc at the base. The inner bristles are longer than the outer bristles with one conspicuously longer than the others, which expand at the base. Bristles are flexible, antrorsely scaberulous, and attenuate. Basal florets barren and are with palea. Lemma of the lower sterile floret is with 3–5 veins. 9. Cenchrus gracillimus Thunb. is commonly called as Slender sandbur and is a native to southeastern US and Caribbean Islands. It is a perennial species (Stieber and Wipff 2000; Quattrocchi 2008) with dense tufted growth habit. Culms are 15–60 cm in length. Leaf blades 4.5–17 cm long, 1.1–3.3 wide. Panicle spiciform, loose and 2–6.8 cm long. Axis bears spikelet clusters of 1–3 spikelets. Involucral bristles connate at the base to form a cup. Bristles rigid, flattened, retrorsely scaberulous, and emerged irregularly from the body of the burr. Basal florets barren and are with palea. Lemma of the lower sterile floret is with 3–5 veins. This weed species grows in sandy soils of pinelands, wet prairies, and river flats, behind the sandy dunes. 10. Cenchrus incertus M.A. Curtis., commonly known as Coastal Sandbur grass or Coastal Sanbur, is native to northern and southern America (GRIN Taxonomy USDA). It is found distributed to other parts of the world – in southwestern, southeastern, and eastern Europe, in tropical Africa, and temperate Asian countries. It is an annual or short lived perennial grass species with tufted, erect with shallow root system. The lower nodes of the prostrate stem give rise to secondary roots. Leaf blade 2–18 cm long and 2–6 mm wide.

S. Goel et al.

Panicle linear, 2–8.5 cm long, and often partially enclosed by the upper leaf sheath (Quattrocchi 2008). Spikelets in clusters of 2–4 spikelets. Involucral bristles connate for 2–7 mm into a cup below with clefts on two sides. Bristles rigid, flattened, and retrorsely scaberulous, with the longest bristle scarcely emergent. Basal florets barren and are with palea. Lemma of the lower sterile floret is with 5–7 veins. This species grows mostly in disturbed areas, dry sites, dry sand in saline meadows, sand hills, road verges, and paddocks. This species often acts as indicator for the poor fertility of the soil (Quattrocchi 2008). 11. Cenchrus longispinus (Hackel) Fernald., commonly known as long spine sandbur or spiny burrgrass, is native to Argentina, Mexico, and US, and is naturalized widely in Australia (GRIN Taxonomy USDA). This is an annual or short-lived perennial spiny species with coarse and tufted habit. Culms are ascending or sometimes decumbent and ranges from 10 to 90 cm in length. Leaf blades 6.3–18.7 cm long and 3–7.2 mm wide. Panicle spiciform and dense, 4.1–10.2 cm in length. Spikelets in clusters of 2–3 spikelets. Involucral bristles connate to form a cup below. Bristles flattened, without grooves, rigid, retrorsely scaberulous, pubescent, and spinose. Basal sterile florets barren and are with palea. Lemma of the lower sterile floret is with 3–7 veins. This is an invasive weed species with low forage value. It is a restricted grass species and is included in aquatic or terrestrial noxious weed and/or noxious-weed seed in U.S. This species grows in disturb soils, dry road side gravel, railroads, and in waste grounds. This species is very similar in morphology and habit with Cenchrus echinatus, hence it is difficult to distinguish from each other (Quattrocchi 2008). 12. Cenchrus melanostachyus A. Camus. is found distributed in eastern Africa at western coastline of Indian ocean and Madagascar (Clayton et al. 2006 onwards). This is a perennial species with short rhizomes. Culms are ascending, 40–60 cm in length, and swollen at the base. Leaf blades 25–35 cm long and 4–5 cm wide. Panicle spiciform and 3–8 cm long. Spikelets in clusters of 3 spikelets. Involucral bristles connate for 1 mm into a disc below. Bristles are flattened, rigid, and antrorsely scaberulous and spiny. Basal sterile

3 Cenchrus

floret male and are with palea. Lemma with 5 veins. 13. Cenchrus mitis Anderss. is native to Africa, Tanzania, Mozambique, and Kenya (GRIN Taxonomy USDA; Clayton et al. 2006 onwards; Quattrocchi 2008). It is an annual species. Culms are 20–100 cm long and geniculately ascending or sometimes decumbent. Leaves flat 5–30 cm long and 2–10 mm wide. Panicle spiciform, linear and 4–18 cm long. Spikelets in clusters of 2 sometimes 3 spikelets. Involucres globose to ovoid, body of involucre pubescent but not grooved. Bristles connate for 2–6 mm or 1/3–2/3 of their length to form a cup at the base. Bristles rigid, antrorsely scaberulous, and spiny. Basal sterile floret male and is with palea. Lemma with 5–7 veins. This species grows in bushland, coastal bushland, sandy soils, and abandoned cultivated lands (Quattrocchi 2008). 14. Cenchrus multiflorus J. Presl. is native to Mexico and Mesoamericana (Clayton et al. 2006 onwards). It is an annual species with tufted habit. Culms 60–140 cm long. Leaf blades 16–25 cm long and 8–20 mm wide. Panicle linear, dense, and 7–18 cm long. Axis bears spikelets in clusters of 2–4 in spikelets. Involucral bristles connate at the base for 3 mm to form a cup below. Inner bristles longer than outer. Bristles flattened and without grooves, rigid, antrorsely scaberulous, and attenuate. Basal sterile floret male and is with palea. Lemma with 5–7 veins. One distinct character of this species is a reduced or scarcely developed leaf in the early development stage of the plant life (cataphyll) (Clayton et al. 2006 onwards). 15. Cenchrus myosuroides Kunth. is commonly called as Big Sandbur and is native to South-central US and northern South American countries and Caribbean Islands (GRIN Taxonomy USDA; Clayton et al. 2006 onwards; Quattrocchi 2008). This is a perennial tall bunchgrass species. Culms are 50–200 cm long, tough, stout, and erect with decumbent base. Leaf blades 12–38 cm long and 4–13 mm wide. Panicle linear 6.5–23 cm long and dense. Spikelets in clusters of 1(-3) spikelets. Bur consists of single whorls of bristles. Bristles united at the base to form a disc below. Bristles as long as the spikelets, flattened, rigid, retrorsely scaberulous, and spiny. Basal sterile floret male

35

and is with palea. Lemma with 3–7 veins. This species is very palatable and grows mostly along roadsides and in other waste places (Quattrocchi 2008). 16. Cenchrus palmeri Vasey. is an annual species found growing in Mexico (Clayton et al. 2006 onwards). Culms are solitary, 9–35 cm long. Leaf blades 4.4–10 cm long and 3.8–6.8 mm wide. Panicle 2–4.3 cm long. Spikelets in clusters of 4–8 spikelets. Involucral bristles connate to form a cup below. Bristles numerous, terete, rigid, retrorsely scaberulous, and spiny. Basal sterile floret male and is with palea. Lemma with 5–7 veins (Clayton et al. 2006 onwards). 17. Cenchrus pennisetiformis (Hochst. and Steud), commonly known as Cloncurry buffelgrass, White or slender buffelgrass, is native to African countries including Kenya, Ethiopia, Somalia, Sudan, and Asian countries including India, Pakistan, Yemen, and southern Iran (Quattrocchi 2008). It is naturalized in the Mediterranean region, Southeast Asia and northern Australia (GRIN Taxonomy USDA; Clayton et al. 2006 onwards). It is a perennial species behaving occasionally as an annual. Culms are 10–40 cm long and are ascending. Leaf blades linear, 2–20 cm long and 2–5 mm wide. Panicle 2–7 cm long. Axis bears spikelets in clusters of 1–3. Involucral bristles connate at the base for 1–2.5 mm into a cup below. Bristles are numerous, flexible, antrorsely scaberulous, and ciliate. Basal sterile florets are with anthers and palea. Lemma with 5 veins. Morphologically it is similar to small-medium type C. ciliaris. It is differentiated from C. ciliaris by its extent of fusion of the involucral bristles. In C. pennisetiformis, the fusion of inner bristles extends more above the rim of the basal disc. Apart from it, plants of C. pennisetiformis are of smaller stature, usually annual, and favor subdesert habitats (Quattrocchi 2008). 18. Cenchrus pilosus Kunth. It is found distributed in Mexico, northern South America, and southwestern America (Clayton et al. 2006 onwards). This is an annual species with dense and tufted habit. Culms 30–60 cm in length. Leaf blades 6–30 cm long and 4–11 mm wide. Panicle spiciform, dense, and 2–13 cm long. Spikelets in clusters of 2–3 spikelets. Involucral bristles connate for 2–4 mm at the base to form a cup. Bristles flattened,

36

19.

20.

21.

22.

S. Goel et al.

flexible, smooth, and attenuate. Basal sterile florets barren and with palea. Lemma with 3–7 veins. Cenchrus platyacanthus. Annual species distributed in western South America (Clayton et al. 2006 onwards). Clumps weak, either erect or ascending, 20–90 cm in length. Leaf blades elongate, 4–24 cm long and 2–8 mm wide. Panicle spiciform, dense, and 3.8–8.6 cm long. Spikelet clusters with single spikelet. Involucral bristles connate at the base for 2.8–6 mm into a cup below. Bristles terete, rigid, retrorsely scaberulous, and spiny. Basal sterile florets barren and with palea. Lemma with 3 veins. Cenchrus prieurii (Kunth) Maire. This species is distributed in India and tropical African region (Clayton et al. 2006 onwards). It is an annual or short species like perennial tufted species, with erect to ascending habit. Culms 30–75 cm long. Leaf blades 8–30 cm long and 5–10 mm wide. Panicle 6–14 long. Spikelets in clusters of 2. Involucral bristles connate at the base for 1 mm into a disc below. Bristles numerous and are flattened, grooved on the face, rigid, antrorsely scaberulous, and ciliate. Basal sterile florets barren and without significant palea. Lemma with 3 veins. Cenchrus robustus R.D. Webster is found distributed in Australia (Clayton et al. 2006 onwards). This is a perennial species with short rhizome. Distinct cataphyll. Culms are 40–100 cm long and erect. Leaf blades 6–22 cm long and 3–7 mm wide. Panicle 4–8 cm long. Axis bears spikelets in clusters of 1–2 spikelets. Bristles connate for 0.5 mm below to form a disc. Bristles numerous, rigid, terete, retrorsely scaberulous, and spiny. Basal sterile florets barren and are with palea. Lemma with 5 veins. Cenchrus setiger Vahl. is commonly known as Birdwood grass and is native to East and Northeast Africa, Northwest India, Arabia, Yemen, Australia, and Brazil (GRIN Taxonomy USDA; Quattrocchi 2008). It is a perennial species, which grows vigorously and rapidly into bunch, more or less erect, forming clumps from a bulbous base. Culms are 40–100 cm long. Leaf blades 6–22 cm long and 3–7 mm wide. Panicle spiciform, 4–8 cm long. Axis bears spikelets in clusters of 1–3 spikelets. Involucral bristles connate below for 1–3 mm into a cup at the base. Bristles numerous, rigid,

terete, retrorsely scaberulous, and spiny. This species can be distinguished from others by its stiff inner spiny bristles and vestigial outer bristles. 23. Cenchrus somalensis Clayton. is a perennial species native to Somalia (Clayton et al. 2006 onwards). Plants densely tufted in habit. Culms erect and 25–45 cm long. Leaf blades narrow and convolute, 5–15 cm long and 1 mm wide. Panicle linear to oblong, 2.5–6 cm long. Axis bears spikelets in clusters of 1–2. Involucral bristles connate for 2–4 mm into a cup below. Bristles flattened and flexible, antrorsely scaberulous, and ciliate. Basal sterile florets with anthers and palea. Lemma with 5 veins. This species grows mainly under shade of trees and among the bushes. 24. Cenchrus spinifex Cav. is commonly known as Coastal sandbur. This species is found distributed in US and South America (Clayton et al. 2006 onwards; Quattrocchi 2008). Plants are annual or short-lived perennial, culms 30–100 cm long, geniculately ascending, or decumbent. Panicle 3–5 cm long. Spikelets in clusters of 2–4 and are with sharp spines. Bristles connate at the base for 1.5–2.5 mm into a cup. Bristles numerous, rigid, flattened, retrorsely scaberulous, and spiny. Basal sterile florets are with anthers or barren. Florets are with palea or sometimes absent. Lemma with 5–7 veins. It is a noxious weed growing on sandy soil in both open areas and thin wood, along roadsides, and waste fields and places. It is very similar in morphology with C. longispinus. But it has shorter spikelets, fewer bristles, wider inner bristles, and flattened outer bristles (Stieber and Wipff 2000). 25. Cenchrus tribuloides L., commonly known as American burrgrass, is native to northeastern and southeastern US, Caribbean Islands, and Brazil (Clayton et al. 2006 onwards; Quattrocchi 2008). This is an annual species with sharp spines, which are painful to touch. Culms are 10–70 cm long, erect or ascending, branching and rooting at the lower nodes. Leaf blades 2–14 cm long and 3–14 mm wide and gray green in color. Spikelets hard and prickly, infrequent, not in clusters, articulated, solitarily on the panicle axis, and densely pubescent. Bristles connate into a cup below. Spikelets numerous, flattened, rigid, retrorsely scaberulous, ciliate, and spinose. This is a weed species, which is a useful stabilizer of sand dunes and grows

3 Cenchrus

in moist and sandy dunes, on coastal sands, open dunes, tropical salt marshes, and open grasslands to dense shrub-cultivated fields. The young plants are used as fodder in dry areas. These species of Cenchrus are mainly distributed in the sub-Saharan central and eastern Africa, arid and semi-arid south-central to southwestern US to northern Mexico on the west, temperate Asia (Middle East countries), tropical Asia (India, Pakistan, Indonesia) on the east and semi-arid regions of Australia (Queensland, northern South Wales). Eastern Africa is considered as the center of diversity for tropical grasses with over 90% of the cultivated forage grasses including Cenchrus originated in this region (Suttie et al. 2005). Out of the 25 species, two species are reported in Europe, 11 in Africa, eight in temperate Asia, eight in tropical Asia, 11 in Australia, seven species in North America, and 13 species in South America (Clayton et al. 2006 onwards). Out of the 25 species, 13 species are perennial, 10 are annual, and two behave as both annual and perennial in habit. Cenchrus grows mainly on sandy soils and alluvial plains, on shallow marginally fertile soils, along roadsides, rocky hillsides, hot and dry areas, and denuded arid lands, growing when soil moisture is available.

3.3 Cytology and Cytogenetics The most widespread species of Cenchrus, C. ciliaris, shares the tertiary gene pool of pearl millet (Pennisetum glaucum) along with other apomictic species of Pennisetum (Jessup et al. 2000; Martel et al. 2004). This grass species has a small genome size of ~ 3,200 MB with 1.33 pg of DNA per haploid genome, which is low for a polyploid grass species (Jessup et al. 2000; Roche et al. 2002). However, limited studies have been done so far to identify other species belonging to the primary, secondary, and tertiary gene pool of C. ciliaris. C. setiger may be in the secondary gene pool of this grass species. Cytological examination of meiotic chromosome behavior in tetraploid (4x ¼ 36) accessions of both species show similar pairing patterns (normal disjunction) during metaphase I (Fisher et al. 1954). These two species can cross with each other, producing fertile hybrids. Meiotic chromosome spreads of the hybrids show more bivalents and fewer

37

quadrivalents, but there is no significant difference in the size of the paired chromosomes (Read and Bashaw 1969). This relationship between these two species is well supported by isozyme and DNA-based marker analyses (Chandra and Dubey 2007, 2008; Kellogg et al. 2009). Cenchrus is a much variable genus. It has base chromosome numbers x ¼ 9, 10, 15, 16, 17, 18 (Fisher et al. 1954; DeLisle 1963; Vij and Chaudhary 1981; Morrone et al. 2006). Majority of the species possess chromosome numbers, which are multiple of the original basic number, i.e., are mainly polyploids and also show high degree of aneuploidy. Somatic (2n) chromosome numbers of different species ranges from 18 to 70 (Table 3.1). The available cytological data suggested that majority of the species are stable cytologically with the exception of a few species. Greater instability in chromosome numbers is observed in C. ciliaris, C. Browni, C. biflorus, Cenchrus mysuroides, and Cenchrus setiger (Vij and Chaudhary 1981). Lagging of chromosomes during first meiotic anaphase due to precocious meiotic division is a common phenomenon in cytotypes with multiple diploid chromosome numbers (Fisher et al. 1954). Such

Table 3.1 Chromosome numbers of different species of the genus Cenchrus Species Chromosome numbers C. biflorus 30, 32, 34, 36 C. brownie 34, 36, 66ca C. cathcarticus 34 C. ciliaris 18, 32, 34, 36, 36+0-2Bs, 40, 44, 50, 52, 54, 56 C. echinatus 34, 68 C. glaucus 36, 38 C. gracillimus 34 C. incertus 32, 34 C. inflexus 34 C. longispinus 34 C. mitis 34 C. multiflorum 34 C. mysuroides 54, 68 ca 70 C. palmeri 34 C. parviceps 34 C. pauciflorus 34, 36 C. pennisetiformis 35, 42, 54 C. pilosus 34 C. prieuri 34 C. setiger 34, 36, 37, 54 C. tribuloides 34

1 2 3 4 5 6 7 8 9

ASGR

variation in chromosome number in different species of Cenchrus may have aroused due to the natural interspecific or intergenic hybridizations, which occurs frequently in grasses. Majority of these species showing cytological instability are facultative apomicts with the exception of C. ciliaris that reproduces through sexual fertilization also. Therefore, there is more chance of preserving the new chromosome combinations. On the other hand, an increase in chromosome number may induce meiotic abnormalities resulting in sterility and subsequently may have evolved apomictic mode of reproduction to escape the sterility. These variations in chromosome number or genome size might have contributed to the variations in certain morphological features, developmental variations, and productive capacity (Mnif et al. 2005). C. ciliaris alone shows variable ploidy levels with 2n ¼ 18, 27, 32, 36, 43, 45, 48, 54 (Fisher et al. 1954; Snyder et al. 1955; Bashaw and Hignight 1990; Hignight et al. 1991; Mnif et al. 2005), but tetraploid with chromosome no. 36 (2n ¼ 4x ¼ 36) is presumably the most widespread ecotype (Hignight et al. 1991). It behaves like a segmental allotetraploid, showing at the average of two to four quadrivalents and 10–14 bivalents during metaphase I (Bashaw 1962; Hignight et al. 1991). This segmental behavior of the buffelgrass genome is verified genetically by Jessup et al. (2003). Analysis of the polymorphic probes in genetic linkage maps reveal repulsion phase associations of 26 linkage groups indicating preferential pairing of linkage groups. The paternal and maternal chromosomes behave differentially from each other, with the maternal parent having higher frequency of crossing over against the paternal parent (Jessup et al. 2003). The chromosomes are very similar in morphology as well as in size and difficult to distinguish. Akiyama et al. (2005) constructed a quantitative ideogram of tetraploid Buffelgrass (C. ciliaris) using fluorescent in situ hybridization (FISH) with 5S rDNA, 18s rDNA, and bacterial artificial chromosome (BAC) clones containing apomixis-linked markers (C101) as probes to DAPI stained spreads (Fig. 3.1). Thirty-six chromosomes were classified into nine morphologically similar chromosome types with two additional chromosomes with no similarity with other chromosomes. All chromosomes are of metacentric and submetacentric types with arm ratio ranging from 1.1 to 1.7 (Akiyama et al. 2005).

S. Goel et al.

Strange

38

Fig. 3.1 Quantitative ideogram of Mitotic Cenchrus ciliaris chromosome. Red, green, and yellow circles indicate C101, 18S rDNA, and 5S rDNA probes, respectively. Source: Akiyama et al. 2005

Many authors had reported Cenchrus setiger with 2n chromosome number of 34, 36, 37, and 54 (Fisher et al. 1954; Vij and Chaudhary 1981; Ahsan et al. 1994). However, 2n ¼ 34 seems to be most widespread. The chromosomes are of metacentric and submetacentric types with length ranging from 3.61 to 1.00 m, and pollen mother cell meiosis shows variable number of quadrivalents (1–5) and bivalents (Vij and Chaudhary 1981). The somatic chromosome complement of Cenchrus biflorous ranges from 2n ¼ 30, 32, 34, 36 with the base chromosome numbers x ¼ 16, 17, 18 (Vij and Chaudhary 1981; Ahsan et al. 1994). The chromosomes are predominated by submetacentric types with length ranging from 2.73–1.07 m. The chromosomes can be grouped into two types with relatively larger group of size range of 2.72–1.74 m and smaller group of 1.50–1.07 m. The meiotic behavior of the chromosomes differs with different cytotypes showing difference in the number of bivalents of the smaller group of chromosomes. However, cytotype with n ¼ 17, which shows more stability during pollen meiosis, seems to be predominant (Vij and Chaudhary 1981). Majority of the species of Cenchrus reproduce apomictically, but low level of sexuality has been reported in some species. The reduced female gametophyte is of polygonum type with two synergids, an egg cell, two polar nuclei in the central cell, and three antipodals toward the chalazal end (Bashaw 1962). Aposporous unreduced female gametophytes can be identified cytologically by the absence of antipodals in mature embryo sac or presence of multiple embryo sacs in cleared ovules. Ovules can be effectively cleared in methyl salicylate after fixing for 24 h in FAA (formalin: acetic acid: ethyl alcohol) and storing in 70% ethyl

3 Cenchrus

alcohol (Young et al. 1979). In both sexual and apomictic plants, the differentiation of the archesporial cell and subsequent reductional division to form a linear tetrad is similar, but in apomictic ovules, all members of linear tetrad degenerate, although rarely, and the chalazal megaspore develops into 2-nucleate stage (Bashaw 1962). In the meantime, the unreduced nucellar cells undergo two successive equational division and enlarge rapidly to form multiple 4-nucleate unreduced embryo sacs; one egg, two synergids, and one diploid polar nuclei and occupy a large portion of the ovule. Despite its unreduced egg cell reproduction, fertilization of the central cell through cross pollination is needed to develop endosperm to nourish the developing embryo and hence to ensure seed set (Hignight et al. 1991). Presence of such dual mode of reproduction may be reason for the high levels of polyploidy and aneuploidy in different collections of Cenchrus (Fisher et al. 1954). An increase in chromosome number in aneuploid may induce meiotic abnormalities resulting in sterility and subsequently may have evolved apomictic mode of reproduction to escape the sterility. These variations in chromosome number or genome size might have contributed to the variations in certain morphological features, developmental variations, and productive capacity (Mnif et al. 2005).

3.4 Phylogenetic Relationship The subfamily Panicoideae of the large family Poaceae includes approximately 208 genera and ~3,300 species. This large subfamily shows monophyly, which is well supported morphologically by the presence of bifloral spikelets (lower sterile or staminate flower and upper fertile spikelets) and anatomically by the presence of simple starch grains in caryopsis (Aliscioni et al. 2003). The molecular level analysis of nuclear and chloroplast genes also supported the monophyly of this large subfamily (Giussani et al. 2001; Doust and Kellogg 2002; Doust et al. 2007; Kellogg et al. 2009). Panicoideae comprises two tribes: Paniceae and Andropogoneae. The tribe Paniceae is a highly diverse and species-rich assemblage of >200 species comprising one-fifth of the total species of grass (Duvall et al. 2001) and about 60% of panicoideae. This tribe is paraphyletic with two clades that

39

differ in the base chromosome number x ¼ 9 and x ¼ 10. Cenchrus belongs to the x ¼ 9 clade along with the type genera of the tribe, Panicum, and other species-rich genus Pennisetum and Setaria. Most of the relative genera are important forage grasses except one species P. glaucum (Pearl Millet), which is regarded as a major cereal crop of the semiarid tropics. This clade is known informally as “Bristle clade” (Doust and Kellogg 2002) because of the presence of sterile inflorescence bristle, which is unique among grasses. Based on morphological similarity, Cenchrus was earlier placed under genus Pennisetum (Jauhar 1981). Different species have been described in both genera at different times, and several authors also indicated that the two genera may be merged into a single genus (DeLisle 1963; Correlld and Johnston 1970). Presently, two genera are still maintained as separate genus (Doust and Kellogg 2002; Kellogg et al. 2009). Both of them show very similar pattern of early inflorescence development, but the difference appears late in development. In many species of Cenchrus, except C. ciliaris and C. myosuroides, the lateral branches or inflorescence bristles become broadened laterally, while those of Pennisetum remain more or less terete (Doust and Kellogg 2002). Their separation into two separate genera is also supported by the extent of fusion of the inflorescence bristles. In Cenchrus, the lateral edges fuse to variable extent while they are free in Pennisetum and elongate more just before anthesis (Doust and Kellogg 2002). Phylogenetic analysis shows that, within the paraphyletic Pennisetum, Cenchrus forms a derived monophyletic subclade (Giussani et al. 2001; Doust and Kellogg 2002; Martel et al. 2004; Doust et al. 2007; Kellogg et al. 2009; Fig. 3.2). The genus Cenchrus consists of both perennial and annual species. The knotted 1 marker gene analysis of the members of panicoideae including six species of the Cenchrus reveals that three species of Cenchrus, C. ciliaris, C. setiger, and C. echinatus, are very closely related, forming a close monophyletic clade, and other three species C. calycalatus, C. myosuroids, and C. pilosus form a separate group (Doust et al. 2007). Another study with chloroplast ndhF marker shows closer relationship of C. echinatus with C. calycalatus, C. myosuroids, and C. pilosus group and also reconfirms the close relationship between C. ciliaris and C. setiger (Kellogg et al. 2009). The close relationship among these three species, C. ciliaris,

40

S. Goel et al.

Fig. 3.2 ML phylogram for full kn1 dataset. Support values above branches are ML bootstrap/MP bootstrap; below branches are Bayesian posterior probabilities

C. setiger and C. pennisetiformis, may be because of chance hybridization among the three species (Doust et al. 2007). Although separated into two different

species, C. ciliaris and C. setiger are morphologically very similar and a few workers have placed C. setiger as a variety of C. ciliaris. Both apomictic and sexual

3 Cenchrus

forms are reported in C. ciliaris, but sexual reproduction has not been reported in C. setiger (Read and Bashaw 1969). These two species can also cross with each other, producing fertile hybrids. The growth habit of the hybrid is intermediate of the two parents: ranging from upright as C. ciliaris and procumbent as C. setiger. The seed set under open-pollination reach up to 99% and up to 84% under self-pollination, which are comparable to the progenies of sexual buffelgrass. The closer relationship between C. ciliaris and C. setiger is also supported by isozyme, sequence tagged site (STS) and random amplified polymorphic DNA (RAPD) markers (Chandra and Dubey 2007, 2008). It seems that the variation between species is affected by the type of habit of the species. The RAPD, isozyme and STS markers clearly separated the annual forms from perennial species with the exception of C. mysuroides, which is a perennial species grouping with annual species. Two perennial species C. ciliaris and C. setiger are shown to be most closely related and they are most distant from the strictly annual species C. prieurri. This indicates that C. mysuroides may be acting as a bridging species between annual and perennial forms (Chandra and Dubey 2008).

3.5 Ecological Behavior The available distribution data suggested that many of the species of Cenchrus originated from tropical and subtropical India and Africa, where humid warm seasonal bimodial (distributed in two summer growing season) rainfall predominates. Mean minimum and maximum temperatures of the region vary from 21 to 24 C and 31 to 36 C, respectively (Cox et al. 1988). It is a xerophytic C4 grass, which grows in bunches and more often with creeping rhizomes. It spreads predominantly in areas where soil is of coarse texture sandy loam type with marginal productivity on clay soils. Poor soil aeration and limited availability of phosphorous limits the survival and productivity of this grass in clay soils (Ibarra-F et al. 1995). These grass species can be established under minimal level of precipitation ranging from as low as 100 mm to 1,000 mm with a few exceptions to more than 2,500 mm annual average. However, seed germination and establishment of the plants during extended wet periods are affected by allelopathy, infection by insects and pathogens, and

41

competition with other grasses, thus reducing its productivity (Cox et al. 1988). Growth initiated in the spring season continues to grow during the hot summer temperature and produces new growth with seasonal rain. Many species of Cenchrus are also potential invasive weeds, which displace native species. They modified to a good extent natural plant communities in California, Texas, and Hawaii in US, Sonora in Mexico, and in Australia (Martin et al. 1995; Tix 2000; Tunyalee 2000; Chen and Kuoh 2004; Franklin et al. 2006). Spines on the surface and the seeds of the C. echinatus, C. tribuloides, and C. biflorus are very irritating, and one cannot walk bare footed in the areas naturalized by these plant (Quattrocchi 2008). Like other invasive grasses, Cenchrus is very efficient in water usage. Its rhizomatous root draws soil moisture rapidly as compared to other plants. C. ciliaris also exudes some allelopathic chemicals, which inhibit the germination of other plants (Tunyalee 2000), and as a result, C. ciliaris once established in an area rapidly modifies the physiochemical properties of natural habitat and makes the soil unfit for others. Following invasion, this weed species forms dense thicklets. The dense growth habit of this rhizomatous grass has the potential to carry the fire rapidly across broad areas and change the fire regime of the ecosystem, where fire does not normally play important role. Like most others, this grass penetrates the soil, and hence, fire has minimal negative effect on this grass and it recovers very quickly. Recurrent fire in the area maintains C. ciliaris populations and the ecological result is the conversion of native scrub communities to African type Savannah with reduced native biological diversity (Tu et al. 2002). It also imposes economic cost through need to manage fire and to protect biodiversity and also the infrastructural establishments.

3.6 Traits of Agronomic Importance and Scope for Domestication Since new to cultivation, C. ciliaris has not yet gone through bottleneck of domestication, and extreme variability exists in wild at intraspecific level also. During 1880–1930, frequent droughts in Africa, North America, and Australia resulted in reduction of

42

livestock numbers. Many botanists, even the military personnel, were sent to many parts of the world in search for grass species, which produced abundant forage in the minimum level of precipitation. Many seeds from harsh climate were collected that undergone many trials. From these thousands of trials, a few grass species were collected for their ease of establishment and forage production in harsh environments (Cox et al. 1988) and C. ciliaris was one of them. This grass species occupies the major component of Dichanthium–Cenchrus–Lasiurus grass cover of India. It was one of the most commercially significant sown pasture grasses in Australia during 1960s and also planted as a valued pasture grass in Texas. Altogether, 34 different cultivars with different eco-physiological characteristics such as high forage quality and high yield have been released for pasture improvement in US, Australia, Brazil, South Africa, Mexico, India, and tropical African countries (Cook et al. 2005). This grass can grow in minimum amount of water as compared to other grass species (Osman et al. 2008), suggesting that the potential of buffelgrass could be harnessed in dry Arabian Peninsula to develop grasslands and reclamation of degraded rangelands with minimum irrigation. Various traits are known to exist in wild species of genus Cenchrus, although due to lack of enough studies, full potential of these species in improvement of cultivated C. ciliaris has not been realized. C. ciliaris although adapted to arid zone conditions, is susceptible to frost and various biotic stresses. More resistance to frost has been reported in C. setiger, but studies are required to establish the source of resistance to biotic stresses at intra and interspecific level.

3.7 Breeding and Crop Improvement C. ciliaris is a perennial warm season grass species quite tolerant to drought conditions and with good forage characteristics. On the contrary, buffelgrass establishment in other regions is limited due to lack of winter hardiness and salt tolerance in this plant. It can adapt to only a narrow range of soil types and is susceptible to several diseases (Jessup et al. 2000), the most destructive being a rampant leaf spot disease (Rust) caused by the fungus, Magnaporte grisea (anamorph Pyricularia grisea), which reduces quality and

S. Goel et al.

production of forage and in some cases destroys stands (Cook et al. 2005). In addition to C. ciliaris, other two closely related species, C. pennisetiformis and C. setiger (birdwood grass), are also potential grass for pasture plantation, but they are less common than C. ciliaris (Weed Management 2008). C. pennisetiformis is adapted to arid conditions and slightly alkaline soils. This grass species shows more tolerance to drought and dry conditions and survives seasonal flooding. It grows commonly on deep sandy soils of semi-desert habitats and is very useful for the stabilization of sand dunes. C. setiger can adapt to more wide range of soil types as compared to C. ciliaris; however, it prefers light textured sandy soils. C. Setiger shows more resistance to heat and frost compared to C. ciliaris; however, it has lower yield as compared to C. ciliaris and C. pennisetiformis. This species is well adapted to arid and semiarid regions with a long dry season and is very effective in stabilizing the moving sand. These agronomic traits if incorporated in C. ciliaris can yield rich dividends. Importance of wild species in improvement of C. ciliaris can be realized with these two species (C. pennisetiformis and C. setiger), although more studies are warranted in wild species of the genus Cenchrus.

3.8 Application of Molecular Techniques in Crop Improvement Compared to other members of the family Poaceae, limited studies have been done so far to assess the diversity at genetic level in tropical grasses such as Cenchrus. RAPD, AFLP (amplified fragment length polymorphism), ISSR (inter-simple sequence repeat), and STS markers have been used to reveal genetic distance at interspecific and intraspecific level (Gustine et al. 1996, Griffa et al. 2006; Chandra and Dubey 2007, 2008; Gutierrez-Ozuna et al. 2009). RAPD analysis with 106 primers in 10 populations of C. ciliaris consisting of a sexual line, an apomictic line, and eight progenies of the cross between two lines generated 569 distinct bands, out of which 495 bands showed differential expression among the lines (Gustine et al. 1996). In another RAPD study (Chandra and Dubey 2008), a higher degree of polymorphism was recorded among eight species of Cenchrus collected from

3 Cenchrus

different regions of the world. One hundred and eighty-seven decamer primers generated a total of 1,296 scorable bands, out of which 1,204 (92.9%) showed polymorphism among the eight species (Chandra and Dubey 2008). This result is the indication of a broad genetic base in the genus. Gutierrez-Ozuna et al. (2009) investigated the genetic diversity among pasture and colonized roadside populations of northwestern part of Mexico using ISSR markers. The genotypic parameters analysis demonstrated that populations have low level of genetic diversity between the roadside populations and pasture populations. These high frequencies indicate that the genotype of the region may be maintained by agamospermy or probably due to founders effect, or the populations sampled may represent the original introduced genotypes (Gutierrez-Ozuna et al. 2009). However, low level of polymorphism shown in vegetative characters and at the DNA level may be a result of occasional sexual reproduction or chance fertilization of unreduced gametes (Hignight et al. 1991). Somatic variation can also be responsible for the low level variation. Griffa et al. (2006) employed AFLP markers in assessing genetic diversity in buffelgrass. According to the authors, it was the first report of utilization of the technique in buffelgrass. Their analysis using three primer combinations produced 152 scorable bands among six apomictic cultivars, one sexual line and 15 putative hybrids of the cross between apomictic male cultivar and sexual female lines. These markers segregate most of the putative F1 hybrids towards the female parent with the exception of two putative F1 hybrids, which cluster neither with their parents nor with the other putative hybrids. The closeness of the hybrids to the female parent indicates that these hybrids may have originated through self-pollination of a sexual line. The two putative hybrids, which form a separate cluster, may represent the true hybrids. The hybrid nature of the two F1 putative hybrids was confirmed cytologically with observation of aposporous embryo sacs and with the amplification of sequencecharacterized amplified region (SCAR) marker “ugt 197”. Chandra and Dubey (2007) demonstrated that the STS markers developed on the basis of functional gene sequences of other non-related grass species can be successfully employed to analyze the level of genetic diversity between eight different species of Cenchrus.

43

Among the 17 Stylosanthes STS primers tested, 14 STS primers amplified 195 scorable bands, out of which 162 showed polymorphism between five species, indicating high amplification success of 82%. These bands show reasonable genetic distance ranging from 0.51 to 0.79 between species. This result provides an opportunity to use the primer sets in deciphering the diversity at both inter and intraspecific level. The authors also demonstrated that the isoenzyme variation between different species of Cenchrus is affected by the breeding system of the species. The perennial species show more polymorphism than the annual species. Out of nine isoenzymes tested, esterase produced more variable loci showing three and two polymorphic loci in perennial and annual species, respectively. Chandra and Dubey (2008) also assessed the differential response of eight species of Cenchrus using drought tolerant physiochemical parameters such as malondialdehyde (MDA), proline content, specific leaf area, and carbon isotope discrimination (CID). Generally, the level of MDA and proline accumulation increases in response to water stress in wheat (Zhang and Kirkham 1994), indicating a higher peroxidation of lipids. The same result of MDA accumulation was observed in most of the eight species subjected to water stress, with the exception of C. ciliaris, which shows significant drop in MDA accumulation. There is also significant increase in proline content with the exception of C. prieuri, which is an annual species. The percent increase in the MDA and proline accumulation was maximum in C. pennisetiformis, thus indicating higher tolerance of this species to water stress conditions.

3.9 Genetic Map Genetic map has been reported in C. ciliaris, but no reports could be found in other species. A full genetic map, which covers approximately 70–80% of the genome, has been reported in buffelgrass (Jessup et al. 2003; Fig. 3.3). A total of 100 cDNA probes consisting of clones from a cDNA library from the pistils of an apomictic buffelgrass and other cDNA clones, which show differential expression between sexual and apomictic ovaries in a previous study by Vielle-Calzada et al. (1996), and 360 heterologous polymorphic

44

S. Goel et al. P6C07c

8.8

3.3 7.7

P3C80a

RZ478i P8B05c P6B06c P1H06e P9G08c P9G04a P10A04b

P9F03c

P1H05j

P10H05g P9D10i P7C10o

12.8 CSU410h 25.5

19.4

8.7 5.4

P7A01e

P10H05e P3C08b 21.9 P12C05g 1b 117.5 cM

P12C03d P4F04a P9B07c

10.7 1.7 9.2

P8F01a 26.7 CSU523c CSU742e P10E03e 7b, LOD-4.90 11.7

12.5

P3H07d

23.0

P4A06d CSU5391 P7C03c P10C08c P7H05e P3B106

5.4 CDSR94g

16.5 P7G03a 17.7

P7F01h

15.0 7.5

6.0 8.1 CDSR94i

8.0 4.0 4.9

P1C12f

14.7

14.3

2.1

Pc08b

15.9 P2E06d

P5G09a P8F01c

13.6

P9D09b P7F01e

22.5

P9B09c

18.0 P1C12c P10E03a 2a 171.2 cM

24.7

P9C09c M848d P8H10b P9A09g

P10C12c

26.4

P3A06n

P7G03c P8A02d

Pcs1d

17.5 P12D07a

P4A06b

27.7

P9D10c

19.7

P10F10c P1C12b

10.9

CDSR94J

19.2

P1D01f P9B09a

P7D07a 2a 228.8 cM

23.4

28.0

36.4

P7A01d

20.8

P5F11h

M651d P7A01f

14.0

13.5 3.5 3.1 14.6

20.8

21.2

P9B07e P10C12a

P7G03e

P6B07g 20.7

P9B07d

P10D08f

18.1

P9B07b

Pcs1c 15.9

16.6

C356d

CSU360i

P1C05d C746b

12.9

P8H10c P10H08h

P8A02e P10H05d CSU410i P8C08i

P9C09a

14.6

31.1

P8A05b M891b 19.4 6.5 C356g 3.5 P3H07k P5H09l 3.4 P3B10c 16.8 10.7 CDSR94k 1.0 1.2 1b 1.6 116.2 cM 12.9

13.2

P7H04h

P2C06n

27.8

P3A04k

CSU742f

25.3 26.2

23.3

19.0

P6G04c P1C09g

15.6

P7A08g 16.4

17.6

12.5

22.2

20.5

19.4

Pca3b 23.6

27.4

P11B11f P10C12b

P6B06a P10H05b

23.0

16.4

20.7

19.6

20.7

13.9 4.0 6.3 4.7

P7G01d

16.0 P9H05f

12.1

P12C02d P2E04a

MS48x

P2C03d

P9H08h

P10H05c

11.3

16.1 P10F11d P11C07a

25.2 10.3 1.2 8.3 6.4

SB1541

CSU706b

18.7

P5G09c 2b 190.9 cM

P10D03a 2b 192.2 cM

P10H05f

18.5

1a 293.0 cM

CSU527e 12.4 P3C08c 1a 241.9 cM

P3E08i

CSU455d 12.7

26.7

P1D01i P8A05g

12.7 P9D10f

31.9 2.0

P9H05g P9H05e

24.7

14.7

17.2 2.9 5.2 5.0

30.7 P1D01h

P1B11b 13.3 P9A10c P10H12d 5.5 5.0

P2B11b P9A06d P9B06d CDSR7f

34.2 SB289d

P9D06f P2D11a CDSR7g

6.1 8.1 1.8 5.0

19.3

M891a P4E01b 13.0 P4A06e P4c06e CDSR94l

P3C01g 8.7 P7D12c P9D04d P1C03h 14.3 P1C03e P6A08e 5.5

P7H07k

14.1 3.7 7.5

21.3

23.9

12.5

8.9 P6H03a

CSU360l

8.8 7.6

16.8

11.0

24.8

5.7 6.5

13.4 3.8 5.4 2.2 7.3

P7C03e 13.5 P4C06d P10G06i 11.5 CSU539b CDSR94d

P8D04a 3a 222.4 cM

CDSR7j P7E03g

P6A08c P10C08a

8.9

P5D01m P3C01a

23.5

11.1 3.0 4.6 7.6

PSF11a P1C03d

13.7

P10F06e

7.2 1.5 3.3 6.0

P11D10c

16.9

P8D06c

24.1 P9F06j P9F06h C250d 5a, LOD = 3.17 15.5 9.4 P1D03h 5a, LOD = 4.20 1.4 P3G10a 5a, LOD = 3.18 12.3 3.6 P10C05d 7.8 4.1 P7B12f 8.8 9.9 M737d 8.4 10.2 3.8 CDSR7a 4b 19.5 93.7 cM

P8D06e PSH07e P12E06c

P1D04d P1D04i P1D03f P1D03g P10C05a P10G02k M737b

P7B12g 4a 141.6 cM

6.9 2.7 2.7 5.7 3.1

P1D03a P7C04f P4G04h P10G12b SB289c 34.9 C250b

P10H12c 4b 74.9 cM

12.5

14.0 1.9 1.9 3.5 2.4 2.0 7.1

3b 153.2 cM

P7H01d 4a 200.2 cM

P5H07d P9A10e P11D10e P5H02e P3G08g

16.8 P1D04a P10B12c 27.9

C250c 4b, LOD–4.20

P1A09b 5b 57.2 cM

P5D10g P3F02d CSU460j P7C04g P4B06c P4G04f CSU479g P1D03c P10D04e P3G10c CDSR7i P9D02c M737e P6A06b P10C05f P10B12a

21.1 P4G04b

33.2

P4E03b 12.8 P9B03h 16.7 M248b 5a 186.0 cM

Fig. 3.3 (continued)

CSU351aS

P1B11d

32.3

P7A03g

P7C04h

P9A10i 13.5

P8D04b

17.6

21.9

P9F06i P8G08g P1D03b

11.6 P9D04b

P6A08b

P7D12g

Pca2d

3 Cenchrus

45

P10B12d

CSU351b

P1F011

P9E10g

P10F12h

30.1 C147d

P10G02b P10G07h CDSR7c 1.5 2.9 P7C04c 10.4 P1C01b P3F02c CSU460h 10.7 P9A10j

7.2

15.2 14.5

P6H01b

16.4 P4F01a P8B08a Sub31e

5.8 7.7

P10F04c P3B11d

17.0

P6A08d

P7G10e P9F02c

Sub31f P10D08h

P9G09d

9.6 3.7 2.8

P7H07i P8C02c

14.0 25.3 P7H08c

HHU27a M466a

P8C08h 7b 78.8 cM

8.0 5.6 7.7 5.7

P10C11h P10C11e P9C04a P7E03j P3G11h P3G11b

3.7

P6D01b

P3G11l P3G11a

3.4 3.5

P2B07b P9A08c P10B07c

9.9 P10C11c

12.6 7.3

16.5

P7H09d

P10C11g 10b 48.0 cM

P3A07e P13G07a 8a 59.6 cM

P1F02a 9a 67.9 cM

11.7 P6E08g P6C04c 11a 48.2 cM

3.0

4.2 7.8 8.0 3.8

P9A08g P2B07c

P7H09c M568c

P1H11b 14.5 P9A08k P2B07f

P9C04i

P9C04g 11b 46.4 cM

7.1

P9A08i

21.4

23.4

P2F07c

M568a 11b 34.3 cM

22.6 P8G09e P3E01g CSU428e

9.5 P4D08b 17.8

P1E05i P1H04d

P1F12b P4F02b

14 83.5 cM

13 100.6 cM

Fig. 3.3 (continued)

20.3

P3B11g 21.8

26.5

3.2

2.5

16 57.6 cM

P4D08c 25.0

26.7 P10H08e

9.6 P4E12f 17 54.2 cM

P8C02d 18 48.5 cM

7.6

P2C02d P7G01c

11.9 P7B03a 12b 75.7 cM

P6C10g P10A08f

17.7

C250e 23.0

P1B07e

P7D12f

P6B07c P9B09d

14.9 P1H04c

15 69.3 cM CSU525n

Sub31b

23.0 P10F11e

23.6 P10D08a

12a 84.2 cM

16.0

9.3

36.1

P6G04b

23.7

11.0

18.5

P13F03d RZ_478_2

P10E03c

13.3 P9G04f

13.9

12.1

P1H08e

CSU413a

P11C07c P2A09e P9G04c

13.6

P9C04h

P6C07d

22.6

P10H3d

27.5

3.7

8.1

P7H05f P10H08i

Pca8a 9b 61.0 cM

17.3

18.3

11.8

P5H10h

P5D01d

2.6 2.3 13.0

M466f

P7A05f

P9A08e

8.8

1.5

14.2 8.4

P11F10d 15.6

P9E10f

P9C04f 11a 76.0 cM

21.7

P8F02f

17.2

24.6

9.1

24.6

26.1

CSU479i

Pca5a

P8G08i

P9H09i

P9E02h

17.6

9.0

10a 59.7 cM

P9H09g

2.8 2.1 5.9 20.2

14.2

P2F06e 8b 63.2 cM

16.6

24.4 P7H09f

P9C10b P9c10c P11B11d C746c P1C05a

P8C03c 8a 82.1 cM

P6H01a

P6E08f 21.1

CSU423d

CSU781b 8b 16.9 cM

P5G10b P4D05h Pca5c

28.2

21.7

P12C05b

P3F08e 12.4

9.6

28.0

16.8

PAp01 P8C08j P3A07g

25.1

M466b Pcs4a

P8H05h

7.3

P4D05c 8.1 2.6

19.7

P3A07a 16.9

19.5 P3E01d

12.3 CSU423f

6.3 5.2

18.7 P4D05e

23.4

2.1 8.6 1.4

Pcs5b 7a 100.5 cM

P10E11a

CSU410b

CSU410f

P8H05i

18.8

P8F03a

M812g 6a 131.2 cM

SB1773e

CDSR94a 7a 74.4 cM

8.5

CSU781c 2a, LOD-4.90 7b 72.9 cM

6a 111.6 cM

22.5

P1H02f 5b 171.4 cM

P9F11f

17.6

P8D08b

P8C06g

21.8

P8B12b

P4B11b

P4F01c P8B08c P10D08i

Pcs4d P7E10f P7G12b

18.8

17.0

15.4

15.9

6.2 P6C02c

P1H11a P7G04a 6b 76.8 cM

P4D05f 8.3

14.5

16.4

24.1

M466c

P4D05d

M568b

P9H09c

HHU27d

15.8

15.8

21.5

8.2

9.9

P1F08B

HHU27b Pcs2b SB1773d

15.5

26.4

21.9

SB1773f 8.1

6.4 1.7

26.4

18.6

11.5 5.8 1.1 5.0 8.5

7.1

M869d 13.4

22.8

P1D03d

CDSR7e 5a 22.3 cM

P10G03i

13.4

24.4

22.3

25.8 M869b 19 46.0 cM

P1D02k 22.7 C250f 20 45.7 cM

46

S. Goel et al.

P7E03m

CSU428a 18.4

19.9 4.4 1.7

P4G04d P8B04f P7C06g

P12E05a P5H01a 22.0

P5H01b P12E05b

6.1

P4G01l P4G01e

21.8

18.4 RZ166c 21 44.4 cM

4.4 10.4

P8C02e 22 40.4 cM

P2C03f

15.8 Pca12e 23 32.9 cM

P3E09c 6.4 5.8

P5D01c

2.4

14.0

18.2 P5H10a

P8A07d

P10A08h

P8G01g 17.1

P9E02i RZ478b P1B07f

24 30.4 cM

25.0

24.6

15.1

P1A04c

P8C08e 25 29.1 cM

18.1

P11F08c 26 27.0 cM

P6A07b P8A07j P3E09e P8H11a

1.5 3.5 8.3

15.3

P5H08c M443b

M180a

P9H09h 10.6 5.5

P3B04b P7C07e

26.3 RZ166f 25 82.7 cM

26.6

P7A05k

P1C12d

8.4 7.4 1.3 9.5

P11F08g P10H07g P9H02h

M180f 27 26.6 cM

P12B03b

RG463a

11.8 P10F12c

26.5

24.8

23.7

P8H07d 22 131.1 cM

CSU539d

CSU705c

P6G02e

29 24.8 cM

8.1

22.8

28.1 CSU377l 21.3 P8D02c

5.1 4.6 4.4 2.5 5.5

17.9

CDSR18b 36 50.1 cM

36.0

CSU527a

11.7 P7D12d

M869c 38 40.4 cM

8.8 4.3

29.5

P8G09b P8G09f P8H06a

C250g 18.3

P1E06b 35 108.1 cM

SB41a 42 29.5 cM

CSU413c 39 39.1 cM

7.0 5.0

P9B03g 43 27.3 cM

9.9

P10G2g M834c

18.2

7.3

P7D12i P1E09g

M869e 41 34.2 cM

P2D06b P12F04n

P1A10a

P9H08e 13.8 P12D02c

24.1

12.6 P1F12c 45 26.4 cM

M248c 46 24.1 cM

P8C02b

19.8 CSU460c 47 19.8 cM

P8C02b

P10E11b

SubEe P4E03h

CSU409g 14.3

12.6

P6H09d 40 37.8 cM

P9H08j

P3C04d 44 27.1 cM

CSU399e 16.4 8.0

Pca8d 16.3

15.1

14.2 C250h

15.8

P4H07a

19.2

P9G04d 33 13.4 cM

M466d

P9H10h 19.9

P7G10c 15.9

P2A06f 37 48.0 cM

P7A05c 32 18.1 cM

9.8 P9H09d

13.2

25.6

P9H09b

P6G10l P6B12c P9G01i

27.3

P12B03f

13.4

17.2

12.1 9.3

P10F12f P8H07c P7D05c P12C03e P8H04f P1D02j

P7A05j 31 20.8 cM

11.3

M568e

P3H12d

C250k

6.3

14.2

32.3

30 23.7 cM

P2A09d P3G08f

30.7

RZ166d

CSU706e

P1F01m

P2C02e 18.1

20.8

18.6 P7A09c 28 26.5 cM

P4G01m

P7C12c

18.9 P12C05a 48 18.9 cM

15.5 P9E02k 49 15.5 cM

7.3

P7G01a P7E03k

50 7.3 cM

P1A02c P6G05a 51 6.1 cM

6.1

P8G01h 29.0 P12D02f 34 241.2 cM

Fig. 3.3 Maternal (white linkage groups) and paternal (black linkage groups) RFLP maps of Pennisetum ciliare. Framework map (LOD > 2.0) markers and map distances (Kosambi map units) are shown to the right and left of the horizontal lines, respectively

cDNAs and gDNAs from across the Poaceae were hybridized to 87 F1 hybrids derived from a cross between heterozygous sexual genotype with another highly heterozygous apomictic genotype. The 460 polymorphic probes yielded segregation data for 400 single dose restriction fragments (SDRF) in female parent and 298 SDRF in male parent. MAPMAKER EXP v3.0 assembled 322 SDRFs into 47 linkage groups spanning over 3,464 cm in the mater-

nal parent with an average marker interval of 10.8 cm, while 78 SDRFs did not show any linkage. In the paternal map, 245 SDRFs were assembled into 42 linkage groups spanning over 2,757 cm with an average marker interval of 11.3, while 53 SDRFs did not show any linkage. These linkage maps give an overall estimate of 4,309 Centi Mozgan and 3,679 Centi Mozgan maternal and paternal genome sizes, respectively. More refined genetic maps can be

3 Cenchrus

developed by inclusion of more markers, which will assist in locating quantitative trait loci (QTLs) controlling the major agronomically important traits.

3.10 BAC Libraries Roche et al. (2002) constructed BAC libraries from the apomictic line B-12-9. This buffelgrass bacterial artificial chromosome (BAC) library contains 68,736 clones with an average insert size of 109 kb, which covers 4.8 haploid genome equivalents. These BAC clones will serve as a useful tool for more refined mapping of buffelgrass genome through the application of more advanced techniques such as FISH, which will ultimately help in resolving physical location of genome complexes.

3.11 Genetics and Molecular Mechanism of Apomixis in Buffelgrass The genus Cenchrus has been studied extensively due to the existing trait of apomixis and hence needs a special mention. This trait is promising and can play rich economic dividends, if deciphered. Studies focusing on this trait have led to interesting information in the genus Cenchrus. The discovery of sexual lines in C. ciliaris provides an opportunity to study the genetics of apomixis in this grass species. Taliaferro and Bashaw (1966) assayed the inheritance pattern of apomixis in selfed progenies of sexual buffelgrass (B-1s) and F1 progenies of a cross between sexual (B-1s) and two apomictic lines. It was concluded in this study that the gene controlling apomixis is dominant and sexuality is controlled by a gene epistatic to gene controlling apomixis. The genotype of apomictic plants will be “A-bb” and those for sexual will be “aabb,” suggesting that the sexual plant originated as a result of natural mutation of the dominant apomictic gene. A subsequent study by Sherwood et al. (1994) from selfing and intercrossing of sexual B-2s lines and progeny of open-pollinated sexual B-2s and other two apomictic lines concluded that inheritance pattern in buffelgrass shows two loci tetrasomic model. These authors also suggested that the sexual line used by Taliaferro and Bashaw in 1966 was highly sexual but facultatively apomictic because selfing of these lines

47

provided progenies segregating for sexuality and apomixis, although results of such crosses do not seem to be compatible with the models of disomic and allotetrasomic inheritance of apospory locus (Gustine et al. 1997). Segregation analysis of isozyme, protein, and RAPD markers within half-sib progenies of both open-pollinated sexual and apomictic lines by Gustine et al. (1996) does not reveal any marker co-segregating with apospory, although subsequent bulked segregant analysis using RAPD markers in two populations raised through crossing of sexual B-2s  apomictic Higgins and B-2s  B-12-9 identified two RAPD markers (J16-800, M02-680) tightly linked to aposporus locus (Gustine et al. 1997). Ozias-Akins et al. (1998) reported 12 co-segregating SCAR markers linked to apomixis in Pennisetum squamulatum. Roche et al. (1999) analyzed these markers in two Cenchrus populations raised earlier by Gustine et al. (1997). Out of these 12 markers, 9 markers (UGT-197, P16R, Q8M, U12H, V4, X18R, A10H, C4, and 07M) and another putative marker C16 were found to be associated only with aposporous mode of reproduction in buffelgrass populations. All markers are segregated as dominant locus and these markers are also conserved between P. squamulatum and apomictic C. ciliaris. They concluded that all the markers clustered together at what was called “apospory specific genomic region (ASGR)”. ASGR has been shown to be hemizygous with suppressed recombination in P. squamulatum (Ozias-Akins et al. 1998). Jessup et al. (2002) reported two markers in C. ciliaris, HHU27 and UGT-197, flanking the aposporus locus at a distance of 1.4 cm and 10.7 cm, respectively, although Goel et al. (2006) did not find any recombination between these markers and ASGR in a population of 94 individuals from a cross between P. glaucum (induced tetraploid, 2n ¼ 28) and P. squamulatum (2n ¼ 56). Roche et al. (2002) isolated the BAC clones from the library of two apomictic grasses, P. squamulatum and C. ciliaris, using six low copy SCAR markers tightly linked with ASGR. Fingerprinting analysis of the BAC clones indicates that the markers, although clustered together in a single genomic region, are duplicated several times within the ASGR in both the species (Roche et al. 2002). FISH analysis using the BAC clones containing SCAR markers as probes confirmed the results from the earlier genetic and

48

molecular mapping that apomixis is controlled by a single dominant, hemizygous locus in both P. squamulatum and C. ciliaris (Goel et al. 2003; Akiyama et al. 2005). The FISH results revealed that ASGR in C. ciliaris is located on the short arm, close to the centromere of a chromosome that also contains the 18S rDNA locus (Goel et al. 2003; Akiyama et al. 2005). The ASGR carrying chromosome does not appear to have a morphologically similar chromosome in C. ciliaris genome. Despite its difference in morphology, ASGR carrying chromosome pairs with other chromosomes during meiosis I forms a quadrivalent (Akiyama et al. 2005). This suggests that ASGR carrying chromosome have at least one homeologous chromosome that may undergo synapsis and recombination during meiosis (Akiyama et al. 2005). Physical mapping studies also suggest that suppression of recombination is mainly due to hemizygocity of ASGR (Goel et al. 2006). Lack of recombination in ASGR hampered the construction of a genetic map and further isolation of novel genes controlling the apomixis in buffelgrass because it is unlikely that further recombination studies will provide a finer map of ASGR (Akiyama et al. 2005). An alternative approach to identify the apomictic genes in C. ciliaris has been to analyze genes differentially expressed during the sexual and apomictic reproduction (Vielle-Calzda et al.1996). Two cDNA clones (Pca-2 and Pca-3) were found to be expressed specifically in apomictic ovaries, one clone (Pca-2) only in sexual ovaries and one fragment (Pca-1) in both sexual and apomictic ovaries. Two more genes Pca 21 and Pca 24 that expressed preferentially (Pca 21) and specifically (Pca 24) in apomictic ovaries have been isolated through subtracted cDNA microarray analysis (Singh et al. 2007). Pca 21 shows homology with two wheat genes expressed specifically in flowers and in inflorescence and Pca 24 with a maize cDNA that encode putative CHCH (coiled-coil-helix–coiledcoil-helix) gene, which have developmental roles (Singh et al. 2007), but none of the genes have yet shown to be linked with ASGR in buffelgrass (OziasAkins et al. 2003; Singh et al. 2007). These results show that even the gene expression is conserved between sexual and apomictic ovaries. It may be because of the same pattern of megaspore mother cell differentiation and meiosis in sexual and aposporus ovules up to the tetrad stage (Vielle-Calzada et al. 1996).

S. Goel et al.

3.12 Genomic Database The NCBI database has the record of 87 core nucleotide sequences of Cenchrus (36 of C. ciliaris, 12 of C. echinatus, 9 of C. setiger, 9 of C. pilosus, 8 of C. myosuroides, 5 of C. calyculatus, 4 of C. incertus, 2 of C. argimonoides and 2 of C. brownii), 21,729 expressed sequence tag (EST), 2,145 genome survey sequence (GSS) starting from Genebank accession no ED544199 to ED 546344 and 45 amino acid sequences (28 of C. ciliaris, 9 of C. echinatus and 8 of C. pilosus). The core nucleotide sequences consist of published and unpublished diverse sequences ranging from apomixis-associated mRNA, tRNA, ribosomal protein gene, chloroplast ndhF gene, knotted 1 gene, rbcL, phytochrome B genes to signaling proteinlike mRNA sequences. Out of 36 nucleotide sequences in C. ciliaris, 14 are of apomoixis-associated mRNA submitted by Gustine et al. (1996), Conner et al. (2008), and Singh et al. (2007), and six are of chloroplast marker genes ndhF (Giussani et al. 2001) and knotted 1 (Doust et al. 2007) . The unpublished 21,729 ESTs in the database represent the total transcriptome of the apomictic pistils of C. ciliaris, submitted by A. H. Paterson and S. R. Schulze, Plant Genome Mapping Laboratory, University of Georgia. In addition to these ESTs, A. Hussey Mark submitted eight ESTs that show differential expression between apomictic and sexual ovaries in P. ciliare. These sequences appear in GeneBank, EMBL and DDBJ under the accession numbers U65082, U65383, U65384, U65385, U65386, U65387, U65388, and U65389 (VielleCalzada et al. 1996). GSS data in the GeneBank are the short-gun sequenced high quality BAC clone reads of “ASGR” of C. ciliaris submitted by Peggy Ozias-Akins, University of Georgia, Tifton Campus. These sequences are also available in FUNGEN ASGR database at the web portal http://asgr.uga.edu (Conner et al. 2008). After assembling these sequences in contigs, BLASTX analysis with the Uniport TrEMBL database shows 25 sequence contigs from C. ciliaris similar to proteins with diverse functions reported from other species (Conner et al. 2008). These putative protein coding regions were mostly from the BACs mapped to low copy region of ASGR. The EMBL database has 26,011 nucleotide sequence entry out of which 63 sequences are of

3 Cenchrus

coding sequence. This database also has 178 amino acid sequence entry. These database also include the sequences of the related genus of Cenchrus and nucleotide sequences of the pathogens, which infects these plants.

3.13 Germplasm Banks National Bureau of Plant Genetic Resources (NBPGR), New Delhi, is the nodal organization in India for exchange, quarantine, collection, conservation, evaluation, and the systematic documentation of plant genetic resources. The genebanks of NBPGR and its regional research centers including Indian Grassland and Fodder Research Institute (IGFRI), Jansi, Central Arid Zone Research Institute, Jodhpur, and Agricultural Research Institute, Bikaner, conserve the germplasm collected from different regions of the world, and the samples of the conserved germplasm are made available to the interested sectors through Germplasm Exchange Division of NBPGR. A total 750 accessions of wild species of the genera Pennisetum and Cenchrus, assembled from 50 countries are conserved at the genebank of ICRISAT (India). Many of the wild relatives have evolved surviving drought, floods, extreme heat and cold, and in the process, they have become adapted or developed resistance to the pests and diseases, which cause heavy losses to the crops. These accessions can be a valuable source for research and development. International Livestock Research Institute (ILRI) situated in Kenya and Ethiopia held 294 accessions of Cenchrus. Most of the accessions were collected from the tropical African countries and are predominated by C. ciliaris. The list of germplasm held at this institute can be accessed through the website http:// www.ilri.org/forage/Cenchrus.pdf. The National Germplasm Resources Laboratory, Beltsville, Maryland (GRIN, USDA), has listed 600 accessions of P. ciliare (syn. C. ciliaris) and 12 accessions of other species of Cenchrus available in its genebank. GRIN provides information of the accessions and its availability for distribution. AusPGRIS – the Australian Plant Genetic Resource Information Service holds 453 accessions of Cenchrus collected from 33 countries of the world. The accessions represent ten species; however, the number of

49

C. ciliaris accessions predominates. The list of germplasm hold can be accessed through the website http:// www2.dpi.qld.gov.au/extra/asp/auspgris/. It provides passport information on germplasm collections related to the identity, origin, characteristics, and evaluation of accessions held. Germplasm samples are conserved under long-term storage conditions and made available throughout the world for research, breeding, and conservation. Genetic Resources Program of the International Centre for Tropical Agriculture (CIAT) is another genebank, which is among the top germplasm holdings of tropical forages obtained or collected from over 141 countries. Germplasms are conserved under an agreement with the International Treaty on Plant Genetic Resources for Food and Agriculture of the FAO. The genebank has listed 21 accessions of different species of Cenchrus available for distribution. The accessions can be accessed through the website http:// www.ciat.cgiar.org/urg. A total of 444 accessions of different species of Cenchrus have been listed the Genetic Resources Unit of the Institute of Grassland and Environmental Research (IGER). This unit is primarily concerned with the collection, storage, characterization, and documentation of temperate forage grasses and legumes. IGER is also responsible for the European Central Crop Databases for Lolium (rye-grass) and Trifolium repens (white clover), and the information of the crops can be accessed from IGER’s web page.

3.14 Potential and Pitfalls Because of its low cost of establishment in low nutrient habitats and high forage production and relative resistance to drought conditions and overgrazing, C. ciliaris serves as an important fodder grass for the arid and semiarid tropical regions of the world. It has also been reported that the biomass production and water use efficiency of this grass is better than other arid zone grasses like Lasiurus scindicus, Panicum turgidum, and Coelachyrum piercei (Osman et al. 2008). Another aspect prevailing in the genus Cenchrus is the apospory trait. At the first conference on apomixis held in September 1995 at the College Station of the Texas A&M University, it was stated that “harnessing apomixis genes for plant improvement

50

offers the potential for quantum in agriculture production – an ‘asexual revolution’, the benefits of which could dwarf those of green revolution,” C. ciliaris along with a relative species P. squamulatum is more extensively studied to harness the benefit of apomixis mechanism. Most studies in the genus Cenchrus has been conducted considering economic impacts limiting most studies to C. ciliaris and little has been done to study other species in genus. This leaves a huge scope to enrich the information in other species of this genus. Many species in genus Cenchrus are potential invasive weeds threatening many keystone habitats of many plants and animal species. More in-depth knowledge of these species will also be helpful in controlling spread of these weeds. In conclusion, C. ciliaris is a crop, which has not gone through the bottleneck of domestication and still open for exploitation of naturally occurring variation in the species, but occurrence of both sexual and apomictic genotypes provide an excellent opportunity to exploit the variation occurring in other closely related species from the genus. For exploiting these species, a systematic study involving various aspects and carried out on all the species is the need of hour.

References Ahsan SMN, Vahidy AA, Ali SI (1994) Chromosome numbers and incidence of polyploidy in Panicoideae (Poaceae) from Pakistan. Ann MO Bot Gard 81:775–783 Akiyama Y, Hanna WN, Ozias-Akins P (2005) High-resolution physical mapping reveals that the apospory-specific genomic region (ASGR) in Cenchrus ciliaris is located on a heterochromatic and hemizygous region of a single chromosome. Theor Appl Genet 111:1042–1051 Aliscioni SS, Giussani LM, Zuloaga FO, Kellog EA (2003) A molecular phylogeny of Panicum (Poaceae: Paniceae): tests of monophyly and phylogenetic placement within the Panicoideae. Am J Bot 90(5):796–821 Arshad M, Ashraf MY, Ahamad M, Zaman F (2007) Morphogenetic variability potential of Cenchrus ciliaris L., from Cholistan desert, Pakistan. Pak J Bot 39(5):1481–1488 AusPGRIS (2003) Australian plant genetic resource information service. http://www2.dpi.qld.gov.au/extra/asp/auspgris/. Accessed 20 Apr 2009 Bashaw EC (1962) Apomixis and sexuality in buffelgrass. Crop Sci 2:412–415 Bashaw EC, Hignight KW (1990) Gene transfer in apomictic buffelgrass through fertilization of an unreduced egg. Crop Sci 30:571–575

S. Goel et al. Bogdan AV (1977) Tropical pastures and fodder plants. Longman, London, UK Chandra A, Dubey A (2007) Transferability of STS markers for studying genetic diversity within the genus Cenchrus (Poaceae). Curr Sci 92:961–967 Chandra A, Dubey A (2008) Evaluation of genus Cenchrus based on malondialdehyde, proline content, specific leaf area and carbon isotope discrimination for drought tolerance and divergence of species. Acta Physiol Plant 30:53–61 Chen C-H, Kuoh C-S (2004) Cenchrus ciliaris L., a newly naturalized grass in Taiwan. Taiwania 49(4):232–236 Clayton WD, Harman KT, Williamson H (2006 onwards) GrassBase – the online world grass flora. http://www.kew.org/ data/grasses-db.html. Accessed 3 Feb 2009 Conner JA, Goel S, Gunawan G, Marie-Michele C-P, Virgil Ed J, Chun L, Haiming W, Lee HP, John EM, Jeremy D, Lixing Y, Jeffrey LB, Patricia EK, Ozias-Akins P (2008) Sequence analysis of bacterial artificial chromosome clones from the apospory-specific genomic region of Pennisetum and Cenchrus. Plant Physiol 147:1396–1411 Cook BG, Pengelly BC, Brown SD, Donnelly JL, Eagles DA, Franco MA, Hanson J, Mullen BF, Partridge IJ, Peters M, Schultze-Kraft R (2005) Tropical forages: an interactive selection tool. CSIRO, DPI&F (Qld), CIAT and ILRI, Brisbane, Australia. http://www.tropicalforages.info/. Accessed 16 Mar 2009 Correlld S, Johnston MC (1970) Manual of the vascular plants of Texas. Texas Research Foundation, Renner, TX, USA Cox JR, Martin-R MH, Ibarra-F FA, Fourie JH, Rethman JFG, Wilcox DG (1988) The influence of climate and soils on the distribution of four African grasses. J Range Manag 41 (2):127–139 DDBJ: DNA Data Bank of Japan (2009). http://www.ddbj.nig. ac.jp/. Accessed 17 Apr 2009 DeLisle DG (1963) Taxonomy and distributon of the genus Cenchrus. Iowa State Coll J Sci 37:259 Doust AN, Kellogg EA (2002) Inflorescence diversification in the Panicoid “Bristle Grass” clade (Paniceae, Poaceae): evidence from molecular phylogenies and developmental morphology. Am J Bot 89(8):1203–1222 Doust AN, Penly AM, Jacobs SLW, Kellog EA (2007) Congruence, conflict, and polyploidization shown by nuclear and chloroplast markers in the monophyletic “Bristle Clade” (Paniceae, Panicoideae, Poaceae). Syst Bot 32(3):531–544 Duvall MR, Noll JD, Minn AH (2001) Phylogenetics of Paniceae (Poaceae). Am J Bot 88(11):1988–1992 EMBL-EBI Nucleotide Sequence Database (2009). http://www. ebi.ac.uk/embl/. Accessed 17 Apr 2009 Fisher WD, Bashaw EC, Holt EC (1954) Evidence of apomixis in Pennisetum ciliare and Cenchrus setigerus. Agron J 46:401–404 Franklin KA, Lyons K, Nagler PL, Lampkin D (2006) Buffelgrass (Pennisetum ciliare) land conversion and productivity in the plains of Sonora, Mexico. Biol Conserv 127:62–71 Genetic Resources Program, International Centre for Tropical Agriculture (CIAT) (2009). http://isa.ciat.cgiar.org/urg. Accessed 20 Apr 2009 Giussani LM, Nchez JHC-S, Zuloaga FO, Kellogg EA (2001) A molecular phylogeny of the grass subfamily Panicoideae (Poaceae) shows multiple origins of C4 photosynthesis. Am J Bot 88(11):1993–2012

3 Cenchrus Goel S, Chen Z, Conner JA, Akiyama Y, Hanna WW, OziasAkins P (2006) Comparative physical mapping of the apospory-specific genomic region in two apomictic grasses: Pennisetum squamulatum and Cenchrus ciliaris. Genetics 173:389–400 Goel S, Chen Z, Akiyama Y, Conner JA, Basu M, Gustavo G, Hanna WW, Ozias-Akins P (2003) Delineation by fluorescence in situ hybridization of a single hemizygous chromosomal region associated with aposporous embryo sac formation in Pennisetum squamulatum and Cenchrus ciliaris. Genetics 163:1069–1082 Griffa SDD, Ribotta A, Castelli SL, Munoz N, Colomba EL, Luna C, Grunberg K, Biderbost E (2006) Molecular genetic discrimination of Buffel grass genotypes and F1 hybrids for breeding purposes using amplified fragment length polymorphism analyses. Grass Forage Sci 61:454–458 Gustine DL, Sherwood RT, Gounaris Y, Huff D (1996) Isozyme, protein, and markers within a half-sib family of buffelgrass. Crop Sci 36:723–727 Gustine DL, Sherwood RT, Huff D (1997) Apospory-linked molecular markers in buffelgrass. Crop Sci 37:947–951 Gutierrez-Ozuna R, Eguiarte LE, Molina-Freaner F (2009) Genotypic diversity among pasture and roadside populations of the invasive buffelgrass (Pennisetum ciliare L. Link) in north-western Mexico. J Arid Environ 73:26–32 Hignight KW, Bashaw EC, Hussey MA (1991) Cytological and morphological diversity of native apomictic buffelgrass, Pennisetum ciliare (L.) Link. Bot Gaz 152(2): 214–218 Ibarra-F FA, Cox JR, Martin-R MH, Crowl TA, Call CA (1995) Predicting buffelgrass survival across a geographical and environmental gradient. J Range Manag 48:53–59 ILRI-International Livestock Research Institute (2009). http:// www.ilri.org/. Accessed 20 Apr 2009 International Crops Research Institute for the Semi-Arid Tropics (ICRISAT). http://www.icrisat.org/. Accessed 20 Apr 2009 Jauhar PP (1981) Cytogenetics and breeding of pearl millet and related species. Alan R. Liss, New York Jessup RW, Burson BL, Paterson AH, Hussey MA (2000) Breeding apomictic forage grasses: Molecular strategies. Proceedings of the 55th southern pasture and forage crop conference, Raleigh, NC, USA Jessup RW, Burson BL, Burow GB, Wang YW, Hussey MA (2002) Disomic inheritance, suppressed recombination, and allelic interactions govern apospory in buffelgrass as revealed by genome mapping. Crop Sci 42:1688–1694 Jessup RW, Burson BL, Burow GB, Wang YW (2003) Segmental allotetraploidy and allelic interactions in buffelgrass (Pennisetum ciliare (L.) Link syn. Cenchrus ciliaris L.) as revealed by genome mapping. Genome 46:304–313 Kellogg EA, Aliscioni SS, Morrone O, Pensiero J, Zuloaga FO (2009) A phylogeny of Setaria (Poaceae, Panicoideae, Paniceae) and related genera based on the chloroplast gene ndhF. Int J Plant Sci 170(1):117–131 Martel E, Poncet V, Lamy F, Siljak-Yakovlev S, Lejeune B, Sarr A (2004) Chromosome evolution of Pennisetum species (Poaceae): implications of ITS phylogeny. Plant Syst Evol 249:139–149 Martin-R M, Cox JR, Ibarra-F FA (1995) Climatic effects on buffelgrass productivity in the Sonoran Desert. J Range Manag 48:60–63

51 Mnif L, Belgacem AO, Cortina J (2005) A comparative analysis of establishment of Cenchrus ciliaris provenances in arid zone of Tunisia. Arid Land Res Manag 19:341–351 Morrone O, Escobar A, Zuloaga FO (2006) Chromosome studies in American Panicoideae (Poaceae). Ann Mo Bot Gard 93:647–657 National Bureau of Plant Genetic Resources (NBPGR). http:// www.nbpgr.ernet.in/. Accessed 20 Apr 2009 NCBI GenBank. http://www.ncbi.nlm.nih.gov/Genbank/. Accessed 17 Apr 2009 Osman AE, Makawi M, Ahmed R (2008) Potential of indigenous desert grasses of the Arabian Peninsula for forage production in a water-scare region. Grass Forage Sci 63: 495–503 Ozias-Akins P, Roche D, Hanna WW (1998) Tight clustering and hemizygocity of apomixis-linked molecular markers in Pennisetum squamulatum implies genetic control of apospory by a divergent locus that may have no allelic form in sexual genotypes. Proc Natl Acad Sci U S A 95:5127–5132 Ozias-Akins P, Akiyama Y, Hanna WW (2003) Molecular characterization of the genomic region linked with apomixis in Pennisetum/Cenchrus. Funct Integr Genomics 3:94–104 Quattrocchi U (2008) CRC world dictionary of grasses. Taylor & Francis Group, Boca Raton, London, New York Read JC, Bashaw EC (1969) Cytotaxonomic relationships and the role of apomixis in speciation in buffelgrass and birdwoodgrass. Crop Sci 9(6):805–806 Roche D, Cong P, Sherwood RT, Ozias-Akins P (1999) An apospory-specific genomic region is conserved between buffelgrass (Cenchrus ciliaris) and Pennisetum squamulatum Fresen. Plant J 19(2):203–208 Roche D, Conner JA, Budiman MA, Frisch D, Wing R, Hanna WW, Ozias-Akins P (2002) Construction of BAC libraries from two apomictic grasses to study the microcolinearity of their apospory-specific genomic regions. Theor Appl Genet 104:804–812 Sherwood RT, Berg CC, Young BA (1994) Inheritance of apospory in buffelgrass. Crop Sci 34:1490–1494 Singh M, Burson BL, Finlayson SA (2007) Isolation of candidate genes for apomictic development in buffelgrass (Pennisetum ciliare). Plant Mol Biol 64:673–682 Stieber MT, Wipff JK (2000) Cenchrus L. http://herbarium.usu. edu/webmanual/info2.asp?name¼Cenchrus_longispinus&type¼treatment. Accessed 25 Mar 2009 Suttie JM, Reynold SG, Batello C (2005) Grasslands of the world. Food and Agriculture Organisation, Rome, Italy Snyder LA, Hernandez AR, Warmke HE (1955) The mechanism of apomixis in Pennisetum ciliare. Bot Gaz 116:209–221 Taliaferro CM, Bashaw EC (1966) Inheritance and control of apomixis in breeding buffelgrass, Pennisetum ciliare. Crop Sci 6:473–476 Tix D (2000) Cenchrus ciliaris Invasion and Control in Southwestern U.S. Grasslands and Shrublands. http://horticulture. cfans.umn.edu/vd/h5015/00papers/tix.htm. Accessed 25 Mar 2009 Tu M, Randall JM, Rice B (2002) Element stewardship abstract for Cenchrus ciliaris L.(African foxtail, buffelgrass, anjangrass). http://www.imapinvasives.org/GIST/ESA/esapages/ documnts/cenccil.pdf. Accessed 25 Mar 2009 Tunyalee M (2000) Cenchrus ciliaris (syn. Pennisetum ciliare L. Link) (buffelgrass, anjagrass, African foxtail grass).

52 Wildland invasive species program, the nature conservancy. http://www.invasive.org/gist/alert/alrtcenc.html. Accessed 13 Apr 2009 USDA (2009) A National Genetic Resources Program. Germplasm resources information network – (GRIN). National Germplasm Resources Laboratory, Beltsville, Maryland, USA. http://www.invasive.org/gist/alert/alrtcenc.html. Accessed 3 Feb 2009 Vielle-Calzada J-P, Nuccio ML, MdA B, Thomas TL, Burson BL, Hussey MA, Wing RA (1996) Comparative gene expression in sexual apomictic ovaries of Pennisetum ciliare (L.) Link. Plant Mol Biol 32:1085–1092

S. Goel et al. Vij SP, Chaudhary JD (1981) Cytological investigation in three species of Cenchrus L. (Gramineae). Cytologia 46: 661–669 Weed Management Guide (2008) Buffelgrass (Cenchrus ciliaris). http://www.weedscrc.org.au/documents/wmg_buffel%20grass.pdf. Accessed Mar 2009 Young BA, Sherwood RT, Bashaw EC (1979) Cleared-pistil and thick sectioning techniques for detecting aposporous apomixis in grasses. Can J Bot 57:1668–1672 Zhang J, Kirkham MB (1994) Drought-stress-induced changes in activities of superoxide dismutase, catalase, and peroxidase in wheat species. Plant Cell Physiol 35(5):785–791

Chapter 4

Cynodon Yanqi Wu

4.1 Introduction Cynodon L. C. Rich. is a small genus of warm-season species in the tribe Cynodonteae, subfamily Chloridoideae of the grass family Poaceae (Gramineae). Though small in number of species, the genus contains species that are economically immensely important as grazed or stored forages, for soil stabilization, and turf. The focus of this chapter is on use of Cynodon species for forage. Although Cynodon species have incidentally provided forage for herbivores for many centuries, it was not until the middle of the twentieth century that large scale monoculture use developed to provide forage for domesticated livestock. This development was in the southern USA and occurred in large measure due to breeding gains, leading to widely adapted cultivars with greatly enhanced forage yield potential. The most important of the Cynodon species as a forage is C. dactylon (L.) Pers. This species consists of six taxonomic varieties that vary substantially in magnitude of natural distribution, genetic diversity, and value for forage. The taxon of greatest importance is C. dactylon (L.) Pers. var. dactylon. This taxon is distributed across all continents and on most ocean islands between latitudes of approximately 45
North and South. It is enormously polymorphic with forms adapted to a range of climatic and edaphic conditions. It is generally known as “common bermudagrass” because of its prevalence.

Y. Wu Department of Plant and Soil Sciences, Oklahoma State University, 368 Ag Hall, Stillwater, OK 74078, USA e-mail: [email protected]

Common bermudagrass was introduced to colonial America at least by the mid-1700s and spread rapidly with the development of agriculture initially in the southeast and subsequently in southcentral and southwest regions. Early reactions of farmers to bermudagrass were mixed as the grass was a serious weed in row crops, especially cotton (Nelson and Burns 2006). Many southern and central states listed it as a noxious weed. It was not until the middle of the twentieth century that attitudes changed regarding its value to agriculture, and circumstances led to its widespread adoption as a cultivated pasture and hay crop. Harlan (1970) described “the development of C. dactylon var. dactylon as an improved hay and pasture grass suitable for intensive management as almost entirely an American phenomenon.” He cited several factors contributing to bermudagrass becoming a major perennial grass for animal grazing and hay production in the southern USA. Prominent among the factors was the release of “Coastal” bermudagrass in 1943. Coastal was developed by USDA-ARS Geneticist G.W. Burton at the Coastal Plains Experiment Station, Tifton, GA. Release of this high yielding variety coincided with other factors, leading to its widespread adoption in the southern USA. These factors included (1) a decrease in the cotton industry in the southern USA, (2) a concurrent increase in the cattle industry in the same region, and (3) greater availability of nitrogen fertilizer. These circumstances led to greatly increased research effort by state agricultural experiment stations and other research entities on the use of bermudagrass as a pasture and hay crop. The development of specialized sprig harvesting and planting equipment also increased the use of bermudagrass (Taliaferro et al. 2004a). Currently, forage bermudagrass is grown on about 10–12 million ha in the USA as a major

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_4, # Springer-Verlag Berlin Heidelberg 2011

53

54

perennial grass feed base for animal husbandry (Taliaferro et al. 2004). Bouton (2007) estimated the current economic value of the forage systems in the southeastern 14 States to be approximately US$11.6 billion annually, in which forage bermudagrass is a major component along with tall fescue [Festuca arundinacea Schreb., recently Schedonorus phoenix (Scop.) Holub] and bahiagrass (Paspalum notatum Flugge). The remainder of this chapter focuses on the taxonomy, distribution, origin and evolution, and germplasm conservation of C. dactylon var. dactylon and related species relative to their use in the development of new forage cultivars.

4.2 Basic Botany of the Species 4.2.1 Taxonomy, Morphology, and Distribution The genus Cynodon consists of nine species in the revised taxonomic classification of Harlan et al. (1970b). The species are C. aethiopicus Clayton et Harlan, C. arcuatus J. S. Presl ex C. B. Presl, C. barberi Rang. Et Tad., C. dactylon (L.) Pers., C. incompletus Nees, C. nlemfuensis Vanderyst, C. plectostachyus (K. Schum.) Pilg., and C. transvaalensis Burtt-Davy (de Wet and Harlan 1970). de Wet and Harlan (1970) did not include C. x magennisii in the revised system while they listed the taxon as a species in “A guide to the species of Cynodon (Gramineae)” (Harlan et al. 1970b). C. x magennisii is a naturally occurring triploid hybrid between C. dactylon and C. transvaalensis (Harlan et al. 1970a). Clayton et al. (2009) lists 10 Cynodon species in Kew’s online grass database. In this list, C. coursii is ranked as a species, while it was a variety of C. dactylon in the classification of Harlan et al. (1970b); C. radiatus Roth ex Roem. and Schult was described as an annual species and as a synonym of C. arcuatus (Clayton and Harman 2002), while Harlan et al. (1970b) described it as a perennial. C. parviglumis Ohwi is listed as a species by Clayton et al. (2009) but not described by Harlan et al. (1970b). C. aethiopicus is a perennial sod-forming species without rhizomes (Harlan et al. 1970b). The species has low tolerance to freezing temperatures. Its inflorescence is composed of one to three whorls of

Y. Wu

racemes, which usually are dark red in color, similar to those of C. plectostachyus, but relatively smaller. Bright red racemes are digitate from a few to 15, with each raceme being straight and 4–8 cm in length. Spikelets are sessile, normally two rows on a rachis. Glumes are persistent and similar but about 3/4 spikelet length. Florets are laterally compressed, 2.5–3.0 mm long, and red or purple. Lemma is keeled and hairy on the keel. Leaf blades of C. aethiopicus is 3–25 cm long and 3–7 mm wide, linear lanceolate, coarse, stiff and harsh, glabrous or sparsely hairy, and conduplicate (Harlan et al. 1970b; Clayton et al. 2009). Culms are rather woody at maturity, can be 40–100 cm long, and 2–6 mm in diameter. Stolons have long internodes and are stout (Harlan et al. 1970b). Harlan et al. (1970b) indicated that C. aethiopicus was distributed in wet habitat on coastal plains to highlands from Ethiopia to Transvaal of South Africa. In addition to its distribution in Africa, it is distributed in Australia (Clayton et al. 2009). Taliaferro et al. (2004a) indicated that stargrass species, including C. aethiopicus, are indigenous to tropical Africa but have been exported to many other tropical and semitropical regions of the world. C. aethiopicus cultivars have been developed by biotype selection (Taliaferro et al. 2004a). Available evidence indicates some hybridization potential between C. aethiopicus and C. dactylon (Harlan et al. 1969). Harlan et al. (1969) attempted many crosses between plants of the two species with only six triploid hybrids produced from diploid (2n ¼ 2x ¼ 18) C. aethiopicus by tetraploid (2n ¼ 4x ¼ 36) C. dactylon. No hybrids were generated by crossing diploid C. dactylon with diploid and tetraploid C. aethiopicus (Harlan et al. 1969). de Wet and Harlan (1970) subsequently reported that tetraploid C. aethiopicus crossed readily with tetraploid C. dactylon var. dactylon to produce hybrids, which were partially fertile. Based on this information, it would be possible to use C. aethiopicus as a genetic source to improve C. dactylon as a forage grass using traditional breeding procedures. C. arcuatus is a sod-forming grass without rhizomes (Harlan et al. 1970b). Its inflorescence normally has 4–5 racemes, which are long, slender, and flexuous. Spikelets are rather densely arrayed on racemes. Plants of C. arcuatus in our greenhouse are perennial and readily set seed. Glumes are subequal and about one-half the spikelet length. Lemma is hairy on keel. Stolons are slender and have short internodes.

4 Cynodon

One distinct character of C. arcuatus is that its leaf blade is very broadly lanceolate. The grass is distributed in northern Madagascar, adjacent islands to Sri Lanka (Ceylon), South and East India, Southeast Asia, Southern China, Philippines, Indonesia, and Northern Australia (Harlan et al. 1970b). Habitat of the species is in lowland wet areas. C. arcuatus is genetically completely isolated from all other Cynodon species (Harlan et al. 1969). The grass has limited forage value, but its genetic isolation from other Cynodon species precludes its use in interspecific hybridization as a means of breeding (Harlan 1970). C. barberi plants are small in size, rare in abundance, and confined in wet areas and edges of permanent water courses. The species is endemic to southern India and of little value as forage, but it has some local use as turf (Harlan 1970). The grass is perennial and sod-forming, but has no rhizomes, and has low tolerance to freezing temperatures. Culms can ascend up to 25–40 cm tall (Clayton et al. 2009), but stolons form loose mats less than 10 cm in height (Harlan et al. 1970b). An inflorescence normally consists of 2–5 short, slender and delicate racemes. Glumes are unequal, and longest glumes can slightly exceed the length of the lemma, which is somewhat hairy on keel. Leaf blades of C. barberi are short, broadly lanceolate with long, sparse hairs. Harlan et al. (1970b) indicated that this species is somewhat a miniature version of C. arcuatus, except that inflorescences of the two species are very different. C. barberi is genetically well isolated from other species in the genus Cynodon (Harlan et al. 1970c). C. dactylon is the species of greatest genetic diversity, widest geographic distribution, and greatest economic value within the genus. Harlan et al.’s (1970b) taxonomic classification includes six varieties within the species. They are var. afghanicus Harlan et de Wet, var. aridus Harlan et de Wet, var. coursii Harlan et de Wet, var. dactylon (L.) Pers., var. elegans Rendle, and var. polevansii (Stent) Harlan et de Wet (Harlan et al. 1970b). The intraspecific classification was based on data for natural distribution, morphological distinctness, cytogenetic behavior, and ecology (Harlan and de Wet 1969; Harlan 1970). In addition to a living collection of 453 accessions from 28 countries grown at Stillwater, Oklahoma, for the data collection, specimens were examined in Kew, British Museum, Edinburgh, Paris, Honolulu, and Washington (Harlan and de Wet 1969). Among the six varieties, var. dactylon

55

has worldwide distribution in warm and temperate climates, var. aridus is relatively widely distributed, and vars. afghanicus, coursii, elegans, and polevansii are narrow endemics (Harlan and de Wet 1969; Harlan et al. 1970b; de Wet and Harlan 1970). Var. afghanicus is an Afghanistan endemic taxon recognized by Harlan and de Wet. The characteristics of var. afghanicus include no rhizomes in the diploid race but very short rhizome-like structures in tetraploid race, and its racemes bear closely imbricated spikelets. Even without rhizomes, living plants of var. afghanicus overwintered in Stillwater, OK, for several years, demonstrating tolerance to freezing temperatures (Harlan et al. 1970b). The variety is found only in lowland sites of steppes and along irrigation ditches. It seems that tetraploids have a wider natural distribution than diploids. Var. aridus is diverse in morphology, but it has less genetic variability than var. dactylon. A race with relatively small plants is found in southern India while a race with much larger plants occurs in South Africa. Two to several racemes constitute an inflorescence with one or rarely two whorls. Glumes are about 2/3 the length of spikelets, which are loosely arranged on the racemes. Var. aridus plants have little winter hardiness in Stillwater, OK, but their rhizomes can grow deep enough in soil to escape freezing temperatures (Harlan et al. 1970b). Harlan and de Wet (1969) indicated that the variety is distributed in a vast geographic area from South Africa, Zambia, Tanzania, to Israel and then to India and Sri Lanka (Ceylon). The variety is called “giant” bermudagrass in the USA where seed is produced in Arizona and California. Var. coursii is relatively coarse. It can grow 40–50 cm tall. Morphology of its inflorescences, racemes, spikelets, and glumes is similar to var. dactylon while lemmas are pilose on keel (Harlan et al. 1970b). Var. coursii has no rhizomes, and leaves are often bunched and appressed at the tip of culms. The variety is distributed in the central plateau of Madagascar. It has no winter hardiness in Stillwater, OK. Var. dactylon is enormously variable with plants ranging in size from very small, fine textured types suitable for turf use to large and robust ones best suited for forage production (Harlan and de Wet 1969). The variety is distributed across all continents and on most ocean islands between North and South latitudes of approximately 45
. In Europe, it can extend to locations, such as Yorkshire in England, Friesland in

56

Netherlands, Potsdam in Germany, and southern Siberia in the Russian Federation (formerly USSR), all at approximately 53
N latitude (Harlan and de Wet 1969). Inflorescences contain racemes arranged in one whorl, rarely two. Glumes are usually ¾ the spikelet length, occasionally slightly longer. The stolons of var. dactylon are never as large as those of the stargrass species, C. plectostachyus, C. aethiopicus, or C. nlemfuensis (Harlan et al. 1970b). The variety has rhizomes and likes disturbed habitats. Harlan and de Wet (1969) reported finding natural spread of var. dactylon plants in valleys about 3,000 m above sea levels in the southern Tibetan region but not on the high plateau of the northern Tibetan region. Harlan and de Wet (1969) recognized three races as tropical, temperate, and seleucidus on the basis of appearance, adaptation, and geographic distribution. This variety is extremely valuable and extensively used directly or as a parent to develop interspecific varieties as commercial products for forage and turf. Var. elegans is a medium coarse variety confined to southern Africa (Harlan and de Wet 1969). It has a distinctive appearance from sympatric var. aridus. Var. elegans has green foliage rather than glaucous one of var. aridus (Harlan and de Wet 1969). Plants of var. elegans can form a lax, loose sod, and its culms are decumbent with ascending appressed leaves (Harlan et al. 1970b). The variety has no winter hardiness in Stillwater, OK. It is valuable in breeding forage bermudagrass. Var. polevansii is small in plant size. Stolons are rather fine with short internodes forming a dense turf, and leaves are very harsh, rigid, and erect (Harlan and de Wet 1969). The variety has only been collected from one site near Barberspan, South Africa (Harlan et al. 1970b). It has good winter hardiness in Oklahoma and is dark green in color and therefore could be valuable in the development of turf type cultivars. C. incompletus is a South African species. Like several other African Cynodon species, C. incompletus has no rhizomes. The species is subclassified into two varieties, var. incompletus and var. hirsutus (Stent) de Wet et Harlan. de Wet and Harlan (1971) indicated that var. incompletus plants have leaves, which are at most moderately pilose, and lower and upper glumes are about 1/4 and 1/2 the spikelet length, respectively, while var. hirsutus has hirsute leaves and relatively longer glumes, which are 1/2 to 3/4 as long as the spikelet. Lemmas are often somewhat blunt with hairs on keels (Harlan et al. 1970b). Morphologically, plants

Y. Wu

of C. incompletus have fine, slender, reddish stolons, linear-lanceolate somewhat hairy leaf blades, and one whorl of racemes in inflorescence. The two varieties are winter hardy in Stillwater, OK (Harlan et al. 1970b). The species is distributed in southern Africa, primarily in South Africa as far north as 23
South latitude. Var. incompletus predominates in the arid regions, while var. hirsutus predominates in the Bushveld, but their distribution ranges overlap extensively in the western Transvaal (de Wet and Harlan 1971). C. nlemfuensis is an important forage species with robust plants capable of rapid growth and high biomass yield. It is one of the stargrass species from which forage cultivars have been developed, and it has been used in interspecific hybridization with C. dactylon to produce forage cultivars. Typical plants of the species have inflorescences in one or rarely two or three whorls of racemes, which are very long, spreading, and flexuous, usually 5 or more, about 4 or more centimeters in length (Harlan et al. 1970b). Glumes are subequal, about ¾ of the length of a spikelet. The species has no rhizomes and has low tolerance to freezing temperatures. Two botanical varieties were recognized in Harlan et al.’s (1970) taxonomical classification, var. nlemfuensis and var. robustus. The nlemfuensis variety is widely distributed in the wetter parts of East and Central Africa from Ethiopia to Zimbabwe (Rhodesia) and Angola. Compared to variety nlemfuensis, var. robustus is taller and coarser and has bigger stolons. The robustus variety is mainly distributed in high rainfall areas of eastern Africa, ranging from Ethiopia to Zimbabwe and westward to Angola. C. plectostachyus is another non-rhizomatous large East African Cynodon species included in the stargrass grouping. It is distinct from all other Cynodon species because of its minute glumes, which are at most 1/4 to 1/5 of the length of the spikelet (Clayton and Harlan 1970). Its inflorescence is normally arranged in 2–5 whorls. It has robust, erect culms reaching heights of 30–90 cm. The species has low tolerance to freezing temperatures. Its habitat is in areas of relatively high rainfall, particularly along the highlands of East Africa. The species is distributed from Ethiopia to Kenya, Tanzania, Zambia, Malawi, and Uganda to Eastern Congo (Harlan et al. 1970b). C. transvaalensis is an endemic species in South Africa. Typical plants of the species are small in size with yellowish-green, erect leaves and small inflorescences with the spikelets loosely arranged on the

4 Cynodon

racemes (de Wet and Harlan 1970). Its leaf blades are rarely more than 2 mm wide, and its stolons fine, slender, with short internodes. Plants of C. transvaalensis are usually found in damp areas around permanent waterholes and along stream banks (Harlan et al. 1970b). Its natural distribution is in a range from southwestern Transvaal, Orange Free State to the northern part of the central Cape Province of South Africa. It is known as “African bermudagrass” and has been widely used for turf and as a parent in crosses with common bermudagrass in the development of F1 hybrid cultivars. African bermudagrass has little value for forage due to its small size and management requirements.

4.2.2 Cytology, Karyotype, and Genome Size Hurcombe (1947) reported chromosome numbers of several cultivated Cynodon species in South Africa using root tip sections to count chromosomes. Her investigations indicated that chromosomes of C. bradleyi (now considered a hybrid between C. incompletus var. incompletus and var. hirsutus), C. transvaalensis, C. dactylon, and C. magennisii were 18, 20, 40, and 30, respectively. She believed that the base chromosome number of Cynodon was 10. She speculated that the determination of 18 somatic chromosomes for C. bradleyi was due to aneuploidy (Hurcombe 1947). Forbes and Burton (1963) reported chromosome numbers of six Cynodon species using root tip smears. Their study confirmed the base number of Cynodon to be 9 rather than 10. Their results indicated that C. bradleyi, C. incompletus, C. plectostachyus, C. transvaalensis, and C. x magennisii had respective somatic chromosome numbers of 18, 18, 18, 18 and 27, while C. dactylon had both diploid (2n ¼ 2x ¼ 18) and tetraploid (2n ¼ 4x ¼ 36) plants. They indicated that some chromosome satellites could cause misinterpretation of chromosome numbers. Harlan et al. (1970b) published a taxonomic revision of the genus Cynodon, which is presented in Table 4.1. Their revision was based on comprehensive cytotaxonomic investigations and has been widely adopted. Malik and Tripathi (1968) reported three cytotypes of naturally occurring C. dactylon in Udaipur, India.

57 Table 4.1 Cynodon species and their somatic chromosomes (Harlan et al. 1970b) Taxon Somatic chromosome number C. aethiopicus Clayton et Harlan 18, 36 C. arcuatus J. S. Presl. Ex C. B. Presl. 36 C. barberi Rang. Et Tad. 18 C. dactylon (L.) Pers. var. dactylon 36, 45a, 54a var. afghanicus Harlan et de Wet 18, 36 var. aridus Harlan et de Wet 18 var. coursii (A. Camus) Harlan et de Wet 36 var. elegans Rendle 36 var. polevansii (Stent) Harlan et de Wet 36 C. incompletus Nees var. incompletus 18 var. hirsutus (Stent) de Wet et Harlan 18, rarely 36 C. nlemfuensis Vanderyst var. nlemfuensis 18, rarely 36 var. robustus Clayton et Harlan 18, 36 C. plectostachyus (K. Schum.) Pilger 18, rarely 36a C. transvaalensis Burtt-Davy 18 a

Sources: Assefa et al. (1999) and Wu et al. (2006a)

They found that diploid (2n ¼ 2x ¼ 18), tetraploid (2n ¼ 4x ¼ 36), and hexaploid (2n ¼ 6x ¼ 54) races were sympatric. de Silva and Snaydon (1995) studied chromosome numbers in C. dactylon of Sri Lanka. They reported that most populations were tetraploids, while plants from roadsides and lawns in the wet region and from forests in the hill country were diploids. They also added that populations from paddy fields in the wet region contained both diploid and tetraploid plants. Surprisingly, they did not find triploids in any environments (de Silva and Snaydon 1995). Wu et al. (2006a) reported four ploidy levels, triploid, tetraploid, pentaploid, and hexaploid in 119 Chinese C. dactylon accessions. The few triploid plants in the collection were considered likely interspecific hybrids that had been introduced to the collection area. Tetraploids were the most common (88%) cytotype in the collection, which was consistent with previous reports of ploidy levels in Chinese Cynodon germplasm. Pentaploids and hexaploids in the collection were only from eastern provinces of China. They did not find diploids in the collection. Similarly, Kang et al. (2008) reported triploid, tetraploid, pentaploid, and hexaploid cytotypes among 43 Korean Cynodon accessions collected in the southern regions of the Korean peninsula between 34 and 36
N latitude and from 126 to 129
E longitude. It was speculated that triploids were introduced since no diploid

58

plants were ever reported in the nation. More recently, Gulsen et al. (2009) reported diploids, triploids, tetraploids, pentaploids, and hexaploids in 182 naturally occurring Cynodon accessions in the southern provinces in Turkey. They further commented that diploids occurred in three provinces, indicating their indigenousness. Their results supported the statement made by Harlan (1970) that the region from West Pakistan to Turkey is the evolutionary center of C. dactylon. Ourecky (1963) reported the pachytene chromosome morphology of a diploid C. dactylon plant from Afghanistan, which probably was var. afghanicus. Pachytene chromosomes of the Cynodon plant varied from 16.8 to 36.3 mm in length. The shortest one in the genome complement was designated as chromosome 1 and the longest as chromosome 9 (Ourecky 1963). Chromosome 4 contains the nucleolar organizing region. Brilman et al. (1982) reported pachytene chromosome morphology of diploid C. dactylon var. aridus and plants of var. afghanicus x var. aridus. The chromosomes ranged in mean length from 20.5 to 44.6 mm, which was reasonably consistent with the results of Ourecky (1963). Arm ratios were from about 1.5 to 2.14 (Brilman et al. 1982). Brilman et al. (1982) suggested that the genomes of var. afghanicus and var. aridus are completely homologous. Knowledge of genome size is important in plant species being investigated with molecular genetic techniques. The use of flow cytometry to estimate nuclear genome size, or genomic DNA content, of Cynodon species was first reported by Taliaferro et al. (1997). The mean nuclear DNA contents of diploids, triploids, tetraploids, and hexaploids in their study were 1.11  0.04 (1,087.8  39.2 Mbp), 1.60  0.04 (1,568  39.2 Mbp), 2.25  0.13 (2,205  127.4 Mbp), and 2.80  0.14 (2,744  137.2 Mbp) pg/2C, respectively. Their research indicated that nuclear DNA content can be used to estimate ploidy level in Cynodon. Genome sizes of diploid African, triploid Tifway, triploid Tifgreen, and tetraploid “Savannah” as reported by Arumuganathan et al. (1999) were 1.03  0.01 (1,009.4  9.8 Mbp) pg/2C, 1.61  0.00 (1,577.8  0.0 Mbp), 1.37  0.01 (1,342.6  9.8 Mbp) pg/2C, and 1.95  0.01 (1,911  9.8 Mbp) pg/2C, respectively. Recently, Wu et al. (2006a), Kang et al. (2008), and Gulsen et al. (2009) reported genome sizes for various ploidy levels of native Cynodon accessions in China, Korea, and Turkey using flow cytometry (Table 4.2).

Y. Wu

4.2.3 Agriculture Uses Several of the Cynodon species have economic value derived from their agricultural use, principally as grazed or conserved fodder for livestock. The most important of the species is C. dactylon, which includes the ubiquitous common bermudagrass, C. dactylon var. dactylon that is of paramount importance. Stargrasses are also of substantial importance in tropical and subtropical regions. In addition to livestock fodder, these grasses have value in controlling soil erosion and in the remediation of spoil soils. They are also effective sinks for removing excess nutrients from soils receiving large amounts of animal effluent. They also have potential as bioenergy feedstock crops because of their high biomass production capability (Casler et al. 2009). The use of bermudagrass as turf rivals or exceeds its use as livestock fodder. Bermudagrass has had a dramatic effect on the livestock industry in the southern USA (Nelson and Burns 2006). Bermudagrass pastures are managed for grazing or haying or dual purpose in the South. Stargrass is cultivated and used in Florida, Central and South America, the Caribbean, and tropical Africa (Taliaferro et al. 2004a). Taliaferro et al. (2004a) pointed out many characteristics making bermudagrass valuable for use as grazed or harvested forages. These include their relative ease of establishment, high biomass production potential per unit area, relatively good palatability, and high tolerance to biotic and abiotic stresses. Improved cultivars coupled with advanced management techniques enable bermudagrass forage production to be a low cost system. Bermudagrass is a major crop of the turf industry in southern USA and many other countries in the world, where climates are tropical, subtropical, and warm Table 4.2 Genome size for Cynodon native in and Turkey Ploidy (2n) Genome size range (pg/2C) Chinese Korean Cynodon by Cynodon Wu et al. by Kang et al. (2006a) (2008) Diploid (18) N/A N/A 1.42–1.56a Triploid (27) 1.55–1.65a Tetraploid (36) 1.96–2.30 1.94–2.19 Pentaploid (45) 2.37–2.49 2.54 Hexaploid (54) 2.90–3.13 2.77–2.85 a

Likely introduced plants

China, Korea,

Turkish Cynodon by Gulsen et al. (2009) 1.03–1.14 1.44–1.62 1.95–2.36 2.56–2.75 3.13–3.44

4 Cynodon

temperate. Taliaferro (2003) indicated the use of bermudagrass as turf rivals its use as a forage grass. As a matter of fact, bermudagrass used as a turfgrass is much more extensive than for forage on a global scale. Because of its relatively high drought tolerance (Beard 1973), bermudagrass has replaced other turf grasses, particularly cool-season species, in turf applications as the availability of irrigation water becomes limited in many parts of the world. Currently, most bermudagrass turf is established using clonal hybrid cultivars, such as “Tifway”, “Patriot”, etc., derived from crossing C. dactylon with C. transvaalensis. However, seeded cultivars of C. dactylon have been bred and commercialized in an accelerated pace in the recent years. Much of this latter effort is by private industry (Taliaferro 2003). Turf bermudagrass is used on lawns, parks, roadsides, sports fields, and golf courses, to name a few. Bermudagrass has an exceptional ability to develop a dense vegetative cover, which can protect the underlying soil from erosion caused by wind, water, or other physical forces. The sod-forming ability of bermudagrass comes from aggressive growth of stolons and rhizomes, which work together with its fibrous root system in binding the topsoil tightly against erosion movement of soil particles. Bermudagrass can withstand inundation for an extended period so that it is an ideal plant species to grow along water ways, on river banks, and other flooded areas. Bermudagrass generally grows well in disturbed habitats, such as right-ofways, roadsides, and urban areas including parks. The ability of bermudagrass to provide an attractive turf with minimal adverse effects due to biotic and abiotic stresses and protect the environment through prevention of soil erosion, removal of greenhouse gasses, and release of oxygen make it a very valuable grass.

4.3 Conservation Initiatives 4.3.1 Species for Forage As previously mentioned, the value of Cynodon species and varieties as grazed or stored forage for ruminant animals, and their potential value in genetic improvement programs, varies from high to negligent. The taxa C. barberi, C. arcuatus, C. transvaalensis, and C. incompletus are of minor value in grazing and have no value in hay production according to Harlan

59

(1970). Additionally, they have little potential for contributing to the development of improved forage cultivars as genetic sources. It should be reemphasized that C. transvaalensis is crucial in the breeding of hybrid turf bermudagrass cultivars and also has direct use as for turf. The most valuable species for forage use include the stargrass species, C. aethiopicus, C. nlemfuensis, and C. plectostachyus, in addition to the cosmopolitan and ubiquitous species C. dactylon. The Stargrass species are generally well adapted to tropical climates receiving 800 mm or more annual rainfall (Taliaferro et al. 2004a). C. aethiopicus has wider distribution than the other two robust East African species. Harlan (1970) noted that the species is very robust and can grow up to 2 m tall. At maturity, woody culms of C. aethiopicus rattle in the wind. Because of its abundance and apparent productivity, the species had probably been used considerably as a forage grass before it was recognized as a species (Harlan 1970). Under heavily grazed conditions, C. aethiopicus and C. dactylon may look similar, but the absence of rhizomes in C. aethiopicus and their presence in C. dactylon can be used to distinguish them. Genetically, the species is well isolated from the other species, but at the diploid level, C. aethiopicus crosses with C. nlemfuensis Vanderyst var. nlemfuensis, and at the tetraploid level, it crosses readily with C. dactylon var. dactylon to produce partially fertile hybrids (de Wet and Harlan 1970). Accordingly, C. aethiopicus has potential to contribute to the genetic improvement of bermudagrass and stargrass via its hybridization potential with forage C. dactylon and C. nlemfuensis var. nlemfuensis. Forage cultivars of C. aethiopicus were released from ecotype selections of natural collections (Table 4.3). C. nlemfuensis is the most promising of the stargrasses used directly for forage (Harlan 1970) and is highly valuable in the development of hybrid cultivars from interspecific crossing with C. dactylon (Burton 1972; Burton et al. 1993). Harlan (1970) noted that C. nlemfuensis var. robustus was used to establish pastures at the Muguga Research Station near Nairobi and had high yield in field trials at the Henderson Research Station. Compared to var. robustus, var. nlemfuensis is finer, less robust, and less hairy. “IB-8” is a cultivar of C. nlemfuensis var. nlemfuensis, selected and released by H. R. Chheda at the University of Ibadan, Nigeria, in 1968 (Harlan 1970; Crowder and Chheda 1982). Characteristics of IB-8 are high forage

60

Y. Wu

Table 4.3 Forage cultivars of stargrass and hybrid cultivars from bermudagrass by stargrass crosses [after Taliaferro et al. (2004a)] Cultivar Taxon Year of Origin References release Cv No 2 (Rhodesian Cynodon aethiopicus Unknown Zimbabwe Bogdan (1977), No. 2 stargrass) Clatworthy (1985) McCaleb C. aethiopicus 1975 Florida Agricultural Experiment Hodges et al. (1975) Station (FAES) IB-8 C. nlemfuensis var. nlemfuensis 1968 University of Ibadan, Nigeria Crowder and Chheda (1982) Muguga C. nlemfuensis var. nlemfuensis Unknown Zimbabwe Clatworthy (1985) Ona C. nlemfuensis var. nlemfuensis 1979 FAES Hodges et al. (1984) Tifton 68 C. nlemfuensis 1984 USDA-ARS and Georgia AES Burton and Monson (1984) Florico C. nlemfuensis var. nlemfuensis 1988 Florida and Puerto Rico AES Mislevy et al. (1989a) and USDA-ARS Florona C. nlemfuensis var. nlemfuensis 1988 FAES Mislevy et al. (1989b) Coastcross-1 C. dactylon  C. nlemfuensis 1967 USDA-ARS and Ga. AES Burton (1972) var. robustus 1982 Tex, La., and Okla. AESs Eichhorn et al. (1984), Brazos C. dactylon var. Taliaferro (1986) afghanicus  C. nlemfuensis var. robustus Tifton 85 C. dactylon  C. nlemfuensis 1992 USDA-ARS and Ga. AES Burton et al. (1993)

yield, good drought tolerance, and outstanding animal weight gain compared to local bermudagrass. Later in the USA, several C. nlemfuensis cultivars and hybrids derived from crosses of C. dactylon by C. nlemfuensis were developed/selected and released (Table 4.3). C. dactylon is the most valuable and important species as a forage grass for grazing and/or hay production and in the development of improved cultivars using intraspecific and interspecific hybridizations. However, according to Harlan’s evaluation, four varieties in C. dactylon, vars. afghanicus, coursii, elegans, and polevansii, are of relatively minor value as herbage grasses, as they are endemic when compared to var. dactylon and aridus (Harlan 1970). Plants of var. afghanicus are relatively robust, vigorous, and winter hardy but are endemic to the lowland and moist sites in Afghanistan (Harlan 1970). Var. coursii and polevansii are endemic to Madagascar and South Africa, respectively, containing no germplasm attributes that will likely be of importance in breeding new forage cultivars (Taliaferro 1986). Var. elegans is indigenous in southern Africa and relatively common in the south of latitude 13
S. Plants of var. elegans have good yield potential but have moderate to low winter hardiness (Taliaferro 1986). The development of Coastal having var. elegans as a parent and “Midland 99” and “Brazos” both having genetic contributions from var. afghanicus indicates that vars. elegans and afghanicus are of high valuable potential in breeding new forage

cultivars. Var. aridus ranges from small plants in South India to large ones in South Africa. Forms of var. aridus from East Africa are larger and more vigorous (Harlan 1970). Cooper and Burton (1965) evaluated forage and turf potential of giant bermudagrass (var. aridus) in southeastern USA. They reported that giant bermudagrass yielded no more forage than common bermudagrass and only around 55% of Coastal bermudagrass (Cooper and Burton 1965). Var. dactylon is highly important as a herbage grass as it is so widely distributed and has been used to develop numerous improved cultivars for forage use. Harlan (1970) stated that the release of Coastal in 1943 represented a hallmark in the records of plant breeding. Coastal, yielding 2–4 times the biomass yield of naturalized common strains, changed the attitude towards bermudagrass among the southern row crop farmers around the middle of twentieth century (Harlan 1970). More efforts have been initiated on the forage bermudagrass breeding since then, and consequently more cultivars have been released and used in commercial production in the USA (Table 4.4).

4.3.2 Germplasm Collection and Conservation The availability of a genetically diverse germplasm collection is often key to the success of a breeding

4 Cynodon

61

Table 4.4 Forage bermudagrass cultivars released in the USA [after Taliaferro et al. (2004a)] Cultivar Year of release Origin Coastal 1943 USDA-ARS and Ga. AES Midland

1953

USDA-ARS and Okla. AES

Suwannee Greenfield NK-37 Alicia Hardie

1953 1954 1957 Mid-1960s 1974

USDA-ARS and Ga. AES Okla. AES Northrup, King & Co Cecil Greer Grass Farm, Edna, TX Okla. AES

Callie Tifton 44

1974 1978

Miss. AES USDA-ARS and Ga. AES

Guymon Tifton 78

1982 1984

Okla. AES and USDA-ARS USDA-ARS and Ga. AES

Grazer Cheyenne

1985 1989

La. AES and USDA-ARS Jacklin Seed Co.

Russell Florakirk Wrangler Midland 99

1994 1994 1999 1999

Goodwell

2007

Ala. and La. AESs Fla. AES Johnston Seed Co. Okla., Kan. Mo., Ark. AESs, Noble Foundation, and USDA-ARS Okla. AES

program. Genes influencing a range of performance traits can often be found in relatives of commercial cultivars or the breeding populations from which those cultivars originated. Though giant strides have been made in Cynodon breeding, the scope of the breeding effort and the degree to which the total germplasm base has been utilized are small. According to Harlan (1970), the most extensive collections of Cynodon germplasm in East Africa were assembled at the Kitale Research Station in Kenya from about 1951 to 1962 by A.V. Bogdan. Bogdan described his and other’s work with herbage Cynodon germplasm in Tropical Pasture and Fodder Plants (Bogdan 1977). The collections of Bogdan included C. nlemfuensis, C. plectostachyus, and C. aethiopicus. Other significant stargrass collections were housed at the Frankenwald Botanical Research Station in South Africa and the Henderson Research Station in Harare, Zimbabwe (Taliaferro et al. 2004a). However, very little information is available on the magnitudes of phenotypic and genetic diversity in the three East African Cynodon species. Although endemic to Africa, each of the species is not widely distributed in natural environments. Genetic erosion within the respective species has not been

References Myers (1951), Burton (1948, 1954) Hein (1953), Harlan et al. (1954) Burton (1962) Elder (1955) Hanson (1972) Taliaferro et al. (2004a) Taliaferro and Richardson (1980) Watson (1974) Burton and Monson (1978) Taliaferro et al. (1983) Burton and Monson (1988) Eichhorn et al. (1986) Samudio and Brede (1998) Ball et al. (1996) Mislevy et al. (1999) Taliaferro et al. (2004a) Taliaferro et al. (2002) Wu et al. (2008)

characterized, but their limited distributions and population sizes likely make them vulnerable, certainly more so than in C. dactylon. Morphological and genetic diversity of C. dactylon var. dactylon is well documented (Harlan and de Wet 1969; Harlan 1970; Harlan et al. 1970a; Wu et al. 2007). Native collections of C. dactylon from China, Korea, and Turkey studied by Wu et al. (2006a, b, 2007), Kang et al. (2008), and Gulsen et al. (2009), respectively, all provided results indicating tremendous genetic variation within the taxon. Additional reports indicating substantial variation within C. dactylon var. dactylon, indigenous to China, were made by Liu and colleagues at the Institute of Botany, Nanjing (Liu et al. 2003), by Abulaiti and colleagues at Xingjiang Agricultural University (Abulaiti et al. 2003), and by Wu before 2001, and now Zhang and colleagues at Sichuan Agricultural University (Wu et al. 2001; Yi et al. 2008). Cynodon plants may have been introduced to the Americas soon after the discovery of the new world by Columbus in 1492 (Taliaferro et al. 2004a). The first written report of bermudagrass in the contiguous USA was made in 1751 by the then Georgia Governor

62

Henry Ellis as noted by Kneebone (1966) and Burton and Hanna (1995). By the early 1800s, bermudagrass had become one of the most important grasses in the South (Taliaferro et al. 2004a). Today, bermudagrass is the major warm-season perennial used in the southern US for forage and turf, and it has spread to the central and northern states. However, organized collections of Cynodon germplasm for cultivar development and genetic research in the USA occurred in the twentieth century, more intensively after the1930s. In the bermudagrass breeding program initiated by Glenn Burton in 1936 at the Coastal Plain Experiment Station of the Crops Research Division, USDA-ARS in Tifton, GA, bermudagrass and stargrass germplasm from Africa and other countries were assembled via collection trips, gifts, and exchange (Burton 1951). The efforts of Burton and subsequently Wayne Hanna and Bill Anderson and their colleagues have accumulated a worldwide collection of more than 600 forage bermudagrass accessions (Anderson 2005). The collection is highly genetically diverse. It has contributed to the development and release of numerous hybrid cultivars over the last half century and continues to be used for improvement in forage traits. Anderson (2005) selected a core collection from the large forage bermudagrass accessions using phenotypic data sampled in the field in cluster analysis. The core collection was to be used for the assessment of chemical composition or stress tolerance. Jack R. Harlan, Wayne W. Huffine, J.M.J. de Wet, and colleagues assembled a worldwide bermudagrass collection at the Oklahoma State University (OSU) in the late 1950s and early 1960s by making germplasm collection trips to Africa, Asia, Australia, and southern Europe (Harlan and de Wet 1969). The collection contained about 700 accessions and was used in an extensive and comprehensive biosystematic investigation of Cynodon (Wu and Taliaferro 2009). Charles M. Taliaferro initiated and directed the bermudagrass breeding program at OSU from 1968 to 2006 in the development of forage and turf bermudagrass cultivars. He and colleagues added more Cynodon accessions from Africa, Australia, Europe, and Asia to the OSU germplasm pool by Harlan (Taliaferro et al. 2004b). The OSU bermudagrass germplasm has been used to develop and release numerous elite forage and turf cultivars by Taliaferro and colleagues. A total of 435 Cynodon accessions are currently maintained in the National Plant Germplasm System

Y. Wu

(NPGS) of the United States Department of Agriculture (USDA) as indicated by the Germplasm Resources Information Network (GRIN) database records as of June 2009 (GRIN 2009). Compared to the number of accessions in the collection in 1995, the current USDA NPGS Cynodon collection contains 22% fewer accessions. Among the US Cynodon collection, 379 accessions can be listed as forage germplasm (Table 4.5) as discussed above. C. aethiopicus, C. nlemfuensis, and C. plectostachyus are represented by one, three, and 32 accessions, respectively, while C. dactylon is represented by 318 accessions. It is evident that the national collection does not contain ample genetic diversity in the East African Cynodon species. Of the 318 C. dactylon accessions, 272 belong to var. dactylon and have been collected from 36 countries. C. dactylon varieties afghanicus, aridus, coursii, elegans, and polevansii are represented by 1, 1, 5, 2, and 4 accessions, respectively, and accessions from each taxon originated from only one country. Even with C. dactylon var. dactylon, the potential to increase the diversity by more comprehensive germplasm collection is large. The data further indicate that comprehensive germplasm collection in major geographic regions is required to sample the full extent of the available variation. In both South and North Americas, introduced Cynodon plants have spread over wide geographic expanses and have become naturalized to specific climatic and edaphic conditions. Consequently, genetic diversity has increased in these populations due to gene recombination, gene mutation, new gene generation, and natural selection under diverse environments since and during their introduction and dispersion, respectively. Thus, germplasm collected from these new world environments is of potential value to genetic enhancement efforts. Certainly, there should be effort to

Table 4.5 Forage Cynodon collections in the US National Plant Germplasm System by species and origin as of June 2009 Species Accessions Continental origin C. aethiopicus 1 Africa C. dactylon 318 Africa, Asia, Australia, Europe, N. America, S. America C. nlemfuensis 3 Africa, N. America C. plectostachyus 32 Africa, Asia C. species 25 Africa, Asia, Australia Total 379

4 Cynodon

increase the sample size of Cynodon taxa now poorly represented in the NPGS collection, but even the number of C. dactylon var. dactylon accessions in the collection likely poorly samples the total germplasm pool of that taxon. Preservation and maintenance of a large bermudagrass germplasm collection is essential to a breeding program and provide research materials for scientific efforts from basic to applied investigations. Widely used techniques for the preservation include maintenance of living clonal plants in the greenhouse or field germplasm nursery, or both, and storage of seed samples in a dry and cold room at 5–10
C. Viability of seed can be maintained for a long term if stored at 18
C or lower. However, genetic preservation of clonal germplasm requires daily care of potted plants, such as watering, trimming, and fertilizing in the greenhouse, or careful management of individual plots in the field to avoid contamination due to encroachment of plants from adjacent plots or by dissemination of seed from plot to plot. Herbicides such as glyphosate are effective in maintaining alleys between neighboring plots. More challenging is to control the contamination from volunteer seedlings derived from seed produced on the plant through outcrossing. One way to reduce or eliminate seed set is mowing the plants before seed is produced on the plant. Alternative forms of preservation techniques have been studied. Reed et al. (2005) reported a cryopreservation protocol for storing bermudagrass shoot tips in liquid nitrogen (196
C). The protocol is most effective when combined with a 1–4 weeks cold acclimation period and dehydration to 19–23% moisture before shoot tips are exposed to liquid nitrogen (Reed et al. 2005). The protocol was used to store 25 bermudagrass germplasm accessions at the National Clonal Germplasm Repository, Corvallis, Oregon, and at the National Center for Germplasm Resources Preservation in Fort Collins, Colorado.

4.4 Origin and Evolution The geographic center of origin of common bermudagrass [C. dactylon (L.) Pers. var. dactylon] has long been of scientific interest because the grass is so widely distributed. About 200 years ago, bermudagrass was recorded in Mease’s Geological Account of

63

the United States, as cited by Tracy (1917), as one of the most important grasses in the southern States. Tracy believed bermudagrass was a native grass in the “Old World,” probably India. Tracy (1917) speculated that bermudagrass probably was introduced to the American colonies by way of the Bermuda Islands (Tracy 1917). Later, Burton in 1951 noted that Thomas Spalding of Sapaloe Island documented in his diary that Henry Ellis, the then Governor of Georgia, introduced bermudagrass to Savannah, Georgia, in 1751. But Spalding did not indicate where Henry Ellis obtained the bermudagrass from (Burton 1951). Burton (1951) assumed that bermudagrass came to Americas much earlier than 1751, probably in the hay brought by the Spanish conquistadors to feed their horses. He believed Africa was the primary center of origin as the bermudagrass introductions from Africa showed greater diversity. Harlan et al. (1970b) noted that Africa has more Cynodon species than other continents (Table 4.6). Clayton et al. (2009) reported that there are seven Cynodon species in the African continent, surpassing the number found on all other continents. Cynodon diversity at species level is obviously highest in Africa and appears to be second highest in Asia, third in Australia, and least in Europe. Beard and Watson (1982) stated that lower East Africa is generally recognized as the center of origin of bermudagrass. Cosmopolitan tetraploid C. dactylon var. dactylon was likely derived from diploid progenitors (Wu and Taliaferro 2009). Diploid Cynodon taxa include C. aethiopicus, C. barberi, C. dactylon var. afghanicus, C. dactylon var. aridus, C. incompletus, C. plectostachyus, and C. transvaalensis. Collectively, those diploids are distributed in a range from South Africa, to East Africa, then Eastern Mediterranean region, and then West and South Asia (Fig. 4.1). Among the diploid Cynodon species, C. aethiopicus, C. barberi,

Table 4.6 Continental distribution of native Cynodon species Continent Number Species name of species Africa 7 C. aethiopicus, C. dactylon, C. incompletus, C. x magennisii, C. nlemfuensis, C. plectostachyus, C. transvaalensis Asia 3 C. arcuatus, C. barberi, C. dactylon Australia 2 C. arcuatus, C. dactylon Europe 1 C. dactylon

64

C. nlemfuensis, C. plectostachyus, and C. incompletus may not contribute to the formation of rhizomatous tetraploid C. dactylon var. dactylon in any way due to the absence of rhizomes in the five species. The research of Harlan and colleagues indicated that diploid C. aethiopicus, C. barberi, C. nlemfuensis vars. nlemfuensis and robustus, and C. plectostachyus have limited or no hybridization potential with C. dactylon var. dactylon, further indicating that these likely did not contribute to the formation and diversity of tetraploid C. dactylon var. dactylon. Harlan and de Wet (1969) summarized that C. dactylon is completely isolated genetically from C. barberi and C. plectostachyus and that it is difficult to cross C. dactylon with C. aethiopicus. Harlan and de Wet (1969) also believed that C. nlemfuensis was an unlikely contributor to the variability of C. dactylon. Diploid C. transvaalensis can readily hybridize with C. dactylon artificially and in nature (Harlan and de Wet 1969). But the species may not have contributed much to the formation of C. dactylon var. dactylon and its enormous diversity since the South African species is narrowly endemic in distribution. However, C. dactylon var. aridus is different. The taxon has well devel-

Fig. 4.1 Geographic distribution of diploid Cynodon species

Y. Wu

oped rhizomes and its natural distribution ranges from South Africa, Zambia, Tanzania, Israel, to India and Sri Lanka. The contributions of var. aridus to the formation of tetraploid var. dactylon are basic (Harlan and de Wet 1969). Harlan and de Wet (1969) believed that introgression between species was not significant in the evolution of C. dactylon var. dactylon. Vars. aridus and afghanicus are the only two diploid forms within C. dactylon. As mentioned above, var. aridus is rhizomatous, relatively widely distributed, and ranges from small plants in India to tall plants in South Africa, while var. afghanicus is endemic and has no rhizomes. Presumably, var. aridus or its direct ancestor is the diploid progenitor of the cosmopolitan tetraploid var. dactylon, and var. afghanicus also contributed to the great variability in var. dactylon (Harlan and de Wet 1969) (Fig. 4.2). According to its variation patterns, morphological variation, adaptation, geographic distribution, and possible genetic origin, the cosmopolitan var. dactylon was subdivided into tropical, temperate, and seleucidus races by Harlan and de Wet (1969) (Fig. 4.2). The tropical race presumably originated from diploid

4 Cynodon

65 Union of unreduced gametes

Var. aridus (2x = 18)

Tropical race of var. dactylon (4x = 36) Union of unreduced and reduced gametes

Evolution Union of reduced gametes

Diploid Cynodon ancestor (2x = 18)

Temperate race of var. dactylon (4x = 36)

Triploids (3x = 27)

Hexaploid var. dactylon (6x = 54)

Union of reduced gametes

Pentaploid var. dactylon (5x = 45) Seleucidus race of var. dactylon (4x = 36)

Evolution

Var. afghanicus (2x = 18)

Union of unreduced gametes

Var. afghanicus (4x = 36)

Fig. 4.2 Cynodon dactylon var. dactylon and its genetic variability are derived from two putative diploid progenitors, C. dactylon var. aridus and var. afghanicus, and genetic events, such as union of unreduced gametes of diploid ancestors, union of reduced and unreduced gametes of tetraploid var.

dactylon, union of reduced gametes of diploid var. aridus and tetraploid var. dactylon, and union of reduced gametes of hexaploid and tetraploid var. dactylon (Harlan and de Wet 1969; Wu et al. 2006a; Gulsen et al. 2009)

var. aridus. Therefore, the geographic origin of tropical var. dactylon might be in the region along the Indian Ocean from southern India to East Africa. The temperate race possibly was derived from evolution of tropical var. dactylon as it spread beyond tropical climates. The seleucidus race may have originated from hybridization between temperate tetraploid var. dactylon and tetraploid var. afghanicus. Tetraploid var. afghanicus likely originated directly from diploid var. afghanicus by union of unreduced gametes of parent plants. Harlan observed various intermediate forms between temperate var. dactylon and var. afghanicus in the region from West Pakistan to Turkey, where the Seleucid Empire existed (Harlan et al. 1970a). From this evolutionary center, the seleucidus race spread outward and merged into a temperate race in Europe (Harlan and de Wet 1969). More recently, hexaploid cytotypes of the var. dactylon were reported in the southeast part of China and the southern region of Korea (Wu et al. 2006a;

Kang et al. 2008). The origin of the hexaploid forms most likely resulted from the union of reduced and unreduced gametes of tetraploid parents (Wu et al. 2006a). Obviously, the evolutionary events in producing hexaploid var. dactylon are internal within C. dactylon var. dactylon. Numerous hexaploid Cynodon plants were reported in collections from a region south of the Taurus Mountains along the Mediterranean coast in Turkey as well (Gulsen et al. 2009).

4.5 Cynodon Genetic Diversity Revealed by Molecular Markers DNA molecular markers, neutral to environment, have been extensively used to measure genetic diversity of Cynodon species. Assefa et al. (1999) reported high genetic diversity in the genus using DNA

66

amplification fingerprinting (DAF) to assess molecular marker polymorphisms of eight species and seven varieties, including C. aethiopicus, C. arcuatus, C. barberi, C. dactylon var. dactylon, C. dactylon var. afghanicus, C. dactylon var. aridus, C. dactylon var. coursii, C. dactylon var. elegans, C. incompletus var. incompletus, C. nlemfuensis var. robustus, C. plectostachyus, and C. transvaalensis. Of 539 bands scored, 92% (496) bands were polymorphic, indicating enormous genetic diversity in Cynodon at the genomic DNA sequence level. Respective accessions of C. dactylon var. dactylon and C. dactylon var. afghanicus were clustered in separate groupings suggesting higher genetic diversity in the two varieties while other accessions from the same taxa were generally clustered. They reported accessions within taxa, differing in ploidy (2x vs. 4x) clustered in all instances indicating that diploid and tetraploid forms are closely related. Using amplified fragment length polymorphisms (AFLP), Wu et al. (2004) assessed genetic diversity and relatedness among 28 disparate C. dactylon var. dactylon accessions originating from 11 countries (Australia, Bulgaria, China, Germany, France, Italy, Japan, South Africa, Spain, Zimbabwe, and the United Arab Emirates) on 4 continents. Of the 590 bands scored, 75% (443) were polymorphic. Genetic similarity coefficients ranged from 0.53 to 0.98 for the 28 accessions. Accessions originating from Australia, Asia, Africa, and Europe were placed in distinct groups, indicating that geographic origin was a significant factor in their genetic differentiation. This suggests that the genetic isolation of plant populations would provide opportunities for the respective populations to be further genetically differentiated by selective forces for adaptation to specific environments. Later, their expanded AFLP study continued on a C. dactylon var. dactylon collection of 119 accessions from 11 provinces in China (Wu et al. 2006a). Among 763 scored AFLP bands generated by 13 primer combinations, 61.1% (466) were polymorphic. Their results indicated the Chinese tetraploid accessions contained much higher genetic variation than hexaploid accessions while pentaploids had very low genetic diversity. Though the accessions from the same or nearby regions tended to cluster, genetic differentiation among accessions from distinct regions was not evident. The results may be associated with the cross-pollination and self-incompatibility reproductive behavior of bermudagrass. Cross-pollination results in

Y. Wu

gene flow between natural populations, which likely prevents formation of distinctly differentiated genetic groups. Although morphologically, hexaploids and pentaploids were similar to tetraploids in the Chinese C. dactylon collection, genetic differentiation among the three ploidy levels was discernable but not fully separate. The close genetic relatedness of the three different cytotypes indicates that they have a recent common ancestry. Kang et al. (2008) reported genetic diversity of 40 Korean bermudagrass (Cynodon spp.) accessions as assessed using the AFLP marker system. They scored 2,256 bands generated from PCR reactions of 29 selective primer combinations. Among the scored AFLP markers, 87.8% (1,982) were polymorphic. Genetic diversity was high in the Korean Cynodon accessions as evidenced by their genetic similarity coefficients ranging from 0.42 to 0.94. As the triploids included in the Korean collection likely are introduced, the true genetic diversity of indigenous Korean bermudagrass could be much lower. There were no native diploid Cynodon reported in Korea (Kang et al. 2008). Genetic separation of tetraploids, pentaploids, and hexaploids did not occur in the AFLP marker data, though their genome sizes were significantly different. Gulsen et al. (2009) reported genetic diversity in 182 Turkish Cynodon accessions and its variance partitioning for ploidy, geographic region, and province. The Cynodon accessions were genotyped by combinational use of four molecular marker systems, encompassing sequence related amplified polymorphism, peroxidase gene polymorphism, intersimple sequence repeat, and random amplified polymorphic DNA. The accessions were considerably diverse and their genetic similarity coefficients ranged from 0.50 to 0.98. Major portions of the genetic variation in the large collection resided within ploidy level (91%) and provinces (94%), respectively. Genetic differentiation was not significant among ploidy levels. Similarly, genetic variation pattern in adjacent regions was not significant as well. The results were basically consistent with those of Wu et al. (2006a). Interestingly, several diploid Cynodon accessions existed in the Turkish collection of Gulsen et al. (2009) but not in the Chinese and Korean collections by Wu et al. (2006a) and Kang et al. (2008), respectively. Karaca et al. (2002) assessed genetic diversity for some released forage bermudagrass cultivars and

4 Cynodon

related selections using four molecular marker systems including AFLP, chloroplast-specific simple sequence repeat length polymorphism, random amplified polymorphic DNA, and directed amplification of minisatellite-region DNA. Among a total of 1,423 DNA fragments scored, 472 (33%) were polymorphic, indicating low genetic diversity in the forage cultivars. The narrow genetic base of forage bermudagrass cultivars was also evidenced by the high genetic similarity coefficients ranging from 0.608 to 0.977. Comparing the genetic diversity of the forage bermudagrass cultivars with the above mentioned genetic diversity in Cynodon germplasm, we can see that high potential exists to develop new cultivars if the naturally occurring diverse germplasm is used in fuller extent.

4.6 Breeding Interspecific Hybrid Cultivars Development of improved bermudagrass cultivars for forage and turf utilization has been tremendously successful since the 1940s. In the review report on the 50-year grassland science achievements, Nelson and Burns (2006) note that improved forage bermudagrass has had a dramatic effect on the livestock industry in the southern USA. In a sister report of the 50-year turfgrass research in the USA, Shearman (2006) indicates that improved turf bermudagrass is the most widely grown warm-season turf grass internationally. The breeders have extensively and effectively used interspecific and intraspecific crosses to create segregating populations for selection and evaluation. For forage cultivar development, elite parents from C. dactylon and C. nlemfuensis have been used in interspecific crosses. Similarly, selected plants of C. dactylon and C. transvaalensis have been routinely used in interspecific crosses to develop interspecific hybrid turf bermudagrass cultivars. Bermudagrass can reproduce asexually and sexually. Vegetative propagation of bermudagrass is realized using its shoot cuttings, stolons, rhizomes, or combinations of those. In the turf industry, bermudagrass sod is an important asexual form used in the planting of improved vegetatively propagated turf bermudagrass cultivars. Over the years, equipment has been invented and techniques developed to facilitate

67

the use of clonal bermudagrass cultivars. Vegetative reproduction in bermudagrass maximized profitability of the forage industry for cattle (Bos taurus) by using the very best individual genotypes for farmers and ranchers to plant for grazing and hay (Hanna and Anderson 2008). Burton (1956) clearly indicated using the F1 hybrids provided the means to take advantage of the heterosis exhibited by superior plants in segregating populations. F1 hybrids are fixed in genetic make up, if they are reproduced clonally. The systematic interspecific hybridization research among Cynodon species of Harlan and colleagues in the 1960s provided detailed insights and useful information to guide the future development of interspecific forage bermudagrass cultivars. As discussed above, C. aethiopicus, C. dactylon, C. nlemfuensis, and C. plectostachyus are deemed to be valuable forage germplasm resources, since they are productive in biomass. Since the remaining Cynodon species are not productive, they will not be included in the following discussion. Harlan et al. (1969) indicated that C. plectostachyus is genetically well isolated from C. aethiopicus, C. dactylon, and C. nlemfuensis. The conclusion was made on the basis of extensive hybridizations. More than 500 inflorescences containing 18,555 spikelets emasculated by hand were pollinated with pollen sources of other Cynodon species and those produced just two presumptive hybrid plants (Harlan et al. 1969). Therefore, the potential of using C. plectostachyus in an interspecific breeding program would be quite low. The latter C. aethiopicus, C. dactylon, and C. nlemfuensis species are genetically related but have varying hybridization barriers, which, however, do not completely prevent intercrossing among them. In the hybridization research of Harlan et al. (1969), only five hybrid plants were obtained from numerous crosses between a diploid C. aethiopicus and a tetraploid C. dactylon. They reported six hybrid plants from two pollinated inflorescences between a tetraploid C. aethiopicus and diploid C. nlemfuensis. The research indicates the great difficulty in crossability of C. dactylon with C. aethiopicus but relative ease with C. nlemfuensis. However, the crossability between diploid and tetraploid C. nlemfuensis and tetraploid C. dactylon is relatively high. Crosses of 91 emasculated inflorescences of diploid C. nlemfuensis var. nlemfuensis pollinated with pollen of tetraploid C. dactylon produced 78 hybrid plants, and crosses

68

of 124 inflorescences of both diploid and tetraploid C. nlemfuensis var. robustus with tetraploid C. dactylon produced 126 hybrid plants, indicating weak crossing barriers between the two taxa (Harlan et al. 1969). Their study indicated that diploid interspecific hybrids between the two species were very vigorous and productive. The substantial value of C. nlemfuensis in interspecific hybridization with C. dactylon has been clearly demonstrated in the development of several cultivars, most notably “Coastcross 1” and “Tifton 85”. Tifton 85 has largely supplanted Coastal as the cultivar of choice in the southern USA (34
S latitude) and is being extensively used in many countries in mild climatic areas. Released forage cultivars of bermudagrass (C. dactylon), stargrass (C. nlemfuensis, C. aethiopicus, and C. plectostachyus) and interspecific hybrids were derived from ecotype selection and organized breeding procedures (Taliaferro et al. 2004a). “Coastal” bermudagrass, the most widely used forage bermudagrass cultivar, was developed from the bermudagrass breeding program by Burton at the USDA-ARS, Coastal Plain Experiment Station in Tifton, GA. “Coastal” produced more than twice as much forage as naturalized common bermudagrass (Myers 1951). The performance (body weight gain) of animals grazing bermudagrass including Coastal and other cultivars clearly indicated their high production potential per unit area but relatively low potential on an individual animal basis due to relatively low nutritive value (Elder and Murphy 1961; McCormick et al. 1964). Forage dry matter digestibility is significantly correlated with body weight gains on pasture, with correlation coefficient being 0.797 (McCullough and Neville 1959). Burton et al. (1967) reported that bermudagrass can be improved for dry matter digestibility by using interspecific crosses between C. dactylon with highly digestible C. nlemfuensis. The best F1 hybrid derived from “Coastal” bermudagrass by “Kenya 56 #14” (PI 255445) released as “Coastcross-1” had 12.3% better dry matter digestibility than “Coastal” while both had similar forage yields (Burton et al. 1967; Burton 1972). Consequently, results from a 3-year grazing study indicated that average daily gain of steers on “Coastcross-1” bermudagrass was 29% more than average daily gain of steers on “Coastal” bermudagrass (Burton 1972). “Kenya 56 #14” (PI 255445) was obtained from A. V. Bogdan of Grassland Research Station, Kitale,

Y. Wu

Kenya, in 1958 (Burton 1972). PI 255445 was identified as C. nlemfuensis var. robustus by Harlan (1970) (Taliaferro et al. 2004a). “Tifton 85” bermudagrass is a high yielding and highly digestible hybrid cultivar derived from a cross of C. dactylon var. dactylon PI 290884 by a C. nlemfuensis selection, “Tifton 68” (Burton et al. 1993). Tifton 85 produced 26% more dry matter that was 11% more digestible than Coastal bermudagrass in two replicated small plots tests over 3 years (Burton et al. 1993). Hill et al. (2001) noted lower concentrations of either-linked ferulic acid in Tifton 85, explaining the higher digestibility of Tifton 85 than of Coastal hays. Higher dry matter yield and greater forage digestibility of Tifton 85 than many other forage bermudagrasses allows increased stocking rates and better body weight gain per hectare. The forage cultivar has received increased acceptance for hay production and pastures in the southern US and been grown on more than 1 million ha in Brazil (Hill et al. 2001). The same breeding protocol using crosses between C. dactylon by C. nlemfuensis was and will be used to develop nutritive-value-enhanced bermudagrass cultivars.

4.7 Recommendations for Future Actions Plant germplasm is among the most precious assets on the earth and should be conserved for the future generations. Bermudagrass is among the major perennial warm-season forage grasses in the world. Continuing efforts in the genetic improvement of bermudagrass and related stargrass will be greatly facilitated by a much larger, better evaluated, and carefully preserved germplasm collection. Individual research programs have collected and used Cynodon germplasm in scientific studies and for breeding, but such programs are generally not charged with, or funded for, long-term preservation of the germplasm. Effective collection, evaluation, and long term preservation can best be accomplished by agencies like the USDA NPGS, which has those dedicated responsibilities. However, inadequate funding has limited the ability of such agencies to fulfill their mandate, especially for crops with lesser value than the major food crops. Forty years ago, Harlan noted that only a small fraction of the Cynodon germplasm pool had been studied and

4 Cynodon

even less used in plant improvement programs (Harlan 1970). Taliaferro recently noted that the extent of Cynodon germplasm collected, studied, and used in breeding new cultivars remains basically the same as many years ago (Taliaferro et al. 2004a). This clearly indicates the need of greater efforts in collecting, evaluating, and preserving Cynodon germplasm at local, national, and international levels. Governments and their agencies should make larger and sustained investments in support of Cynodon germplasm research. As Cynodon forage cultivars are used in many countries and Cynodon forage germplasm is not distributed evenly across the world, international collaborations are necessary to facilitate collection and exchange of Cynodon germplasm, communication of research and breeding information, and distribution of improved cultivars.

Acknowledgments The author dedicates the humble chapter to his mentor, Dr. Charles M. Taliaferro, Regents Professor Emeritus of the Oklahoma State University, and also thanks him for his review and comments for improving the manuscript.

References Abulaiti SDS, Yang ZM, Li PY, Zhao Q, Sun ZJ (2003) Xingjiang no 1 bermudagrass. Pratacult Sci 20:30–31 Anderson WF (2005) Development of a forage bermudagrass (Cynodon sp.) core collection. Grassl Sci 51:305–308 Arumuganathan K, Tallury SP, Fraser ML, Bruneau AH, Qu R (1999) Nuclear DNA content of thirteen turfgrass species by flow cytometry. Crop Sci 39:1518–1521 Assefa S, Taliaferro CM, Anderson MP, de los Reyes BG, Edwards RM (1999) Diversity among Cynodon accessions and taxa based on DNA amplification fingerprinting. Genome 42:465–474 Ball DM, Eichhorn MM Jr, Burdett RA Jr, Bice DM (1996) Registration of Russell’ bermudagrass. Crop Sci 36:467 Beard JB (1973) Turfgrass: science and culture. Prentice Hall, New Jersey Beard JB, Watson JR (1982) Recent turfgrass plant explorations in Africa. USGA Green Section Record, July/August: 6–8 Bogdan AV (1977) Tropical pasture and fodder plants (grasses and legumes). Longman Group, London, UK Bouton J (2007) The economic benefits of forage improvement in the United States. Euphytica 154:263–270 Brilman LA, Kneebone WR, Endrizzi JE (1982) Pachytene chromosome morphology of diploid Cynodon dactylon (L.) Pers. Cytologia 47:171–181 Burton GW (1948) Coastal bermudagrass. Cir 10. Georgia Agricultural Experiment Station, Georgia

69 Burton GW (1951) The adaptability and breeding of suitable grasses for the southeastern States. In: Norman AG (ed) Advances in agronomy, vol 3. Academic, New York, pp 197–241 Burton GW (1954) Coastal Bermuda grass for pasture, hay and silage. Bull N.S. 2. Georgia Coastal Plain Experiment Station, Tifton, GA Burton GW (1956) Utilization of heterosis in pasture plant breeding. In: Neale GJ (ed) Proceedings of the 7th international grassland congress, Palmerston North, NZ, 6–12 Nov 1956, Wellington, New Zealand, pp 3–12 Burton GW (1962) Registration of varieties of bermudagrass. Crop Sci 2:352–353 Burton GW (1972) Registration of ‘Coastcross-1’ bermudagrass. Crop Sci 12:125 Burton GW, Hanna WW (1995) Bermudagrass. In: Heath ME, Barnes RF, Metcalfe DS (eds) Forages: the science of grassland and agriculture. Iowa State Univ Press, Ames, IA, pp 421–429 Burton GW, Monson WG (1978) Registration of ‘Tifton 44’ bermudagrass. Crop Sci 18:911 Burton GW, Monson WG (1984) Registration of ‘Tifton 68’ bermudagrass. Crop Sci 24:1211 Burton GW, Monson WG (1988) Registration of ‘Tifton 78’ bermudagrass. Crop Sci 28:187–188 Burton GW, Hart RH, Lowrey RS (1967) Improving forage quality in bermudagrass by breeding. Crop Sci 7:329–332 Burton GW, Gates RN, Hill GM (1993) Registration of ‘Tifton 85’ bermudagrass. Crop Sci 33:644–645 Casler MD, Heaton E, Shinners KJ, Jung HG, Weimer PJ, Liebig MA, Mitchell RB, Digman MF (2009) Grasses and legumes for cellulosic bioenergy. In: Wedin WF, Fales SL (eds) Grassland: quietness and strength for a new American agriculture. ASA, CSSA, and SSSA, Madison, WI, p 213 Clatworthy JN (1985) Pasture research in Zimbabwe: 1964–84. In: Kategile JA (ed) Pasture improvement research in Eastern and Southern Africa. Proceedings of the 1st PANESA workshop, Harare, Zimbabwe, 17–21 Sept 1984. Int Dev Res Center, Ottawa, Canada Clayton WD, Harlan JR (1970) The genus Cynodon L.C. Rich. in tropical Africa. Kew Bull 24:185–189, Her Majesty’s Stationery Office, London, UK Clayton WD, Harman KT (2002) World grass species: synonyms. Royal Botanic Gardens, Kew, UK Clayton WD, Harman KT, Williamson H (2009) GrassBase – the online world grass flora. http://www.kew.org/data/ grasses-db.html. Accessed 30 Apr 2009 Cooper RB, Burton GW (1965) Forage and turf potential of Giant bermudagrass in the southeastern United States. Agron J 57:239–240 Crowder LV, Chheda HR (1982) Tropical grassland husbandry. Longman Group, Essex, England de Silva PHAU, Snaydon RW (1995) Chromosome number in Cynodon dactylon in relation to ecological conditions. Ann Bot 76:535–537 de Wet JMJ, Harlan JR (1970) Biosystematics of Cynodon L.C. Rich. (Gramineae). Taxon 19:565–569 de Wet JMJ, Harlan JR (1971) South African species of Cynodon (Gramineae). J S Afr Bot 37(1):53–56

70 Eichhorn MM Jr, Nelson BD, Montgomery CR, Davis AV, Hallmark WB (1984) Brazos – a new bermudagrass for pasture and hay in Louisiana. In: Didier DET (ed) Louisiana agriculture. Louisiana Agricultural Experiment Station, Baton Rouge, FL, pp 18–19 Eichhorn MM Jr, Oliver WM, Hallmark WB, Young WA, Davis AV, Nelson BD (1986) Registration of ‘Grazer’ bermudagrass. Crop Sci 26:835 Elder WC (1955) Greenfield bermudagrass. Bull B-455. Oklahoma Agricultural Experiment Station, Stillwater, OK Elder WC, Murphy HF (1961) Grazing characteristics and clipping responses of bermudagrass. B-577. Oklahoma Agricultural Experiment Station, Stillwater, OK Forbes I, Burton GW (1963) Chromosome numbers and meiosis in some Cynodon species and hybrids. Crop Sci 3:75–79 GRIN (2009) Summary statistics for GRIN. http://www.ars-grin. gov/npgs/stats/. Accessed 6 Jun 2009 Gulsen O, Sever-Mutlu S, Mutlu N, Tuna M, Karaguzel O, Shearman RC, Riordan TP, Heng-Moss TM (2009) Polyploidy creates higher diversity among Cynodon accessions as assessed by molecular markers. Theor Appl Genet 118:1309–1319 Hanna WW, Anderson WF (2008) Development and impact of vegetative propagation in forage and turf bermudagrasses. Agron J 100:S103–S107 Hanson AA (1972) Grass varieties in the United States. USDA-ARS Agri Handbook 170. US Gov Print Office, Washington, DC Harlan JR (1970) Cynodon species and their value for grazing and hay. Herbag Abstr 40:233–237 Harlan JR, de Wet JMJ (1969) Sources of variation in Cynodon dactylon (L.) Pers. Crop Sci 9:774–778 Harlan JR, Burton GW, Elder WC (1954) Midland bermudagrass, a new variety for Oklahoma pastures. Bull B-416. Oklahoma Agricultural Experiment Station, Stillwater, OK Harlan JR, de Wet JMJ, Richardson WL (1969) Hybridization studies with species of Cynodon from east Africa and Malagasy. Am J Bot 56:944–950 Harlan JR, de Wet JMJ, Rawal KM (1970a) Origin and distribution of the seleucidus race of Cynodon dactylon (L.) Pers. var. dactylon (Gramineae). Euphytica 19:465–469 Harlan JR, de Wet JMJ, Huffine WW, Deakin JR (1970b) A guide to the species of Cynodon (Gramineae), Bull B-673. Oklahoma Agricultural Experiment Station, Stillwater, OK Harlan JR, de Wet JMJ, Rawal KM, Felder MR, Richardson WL (1970c) Cytogenetic studies in Cynodon L. C. Rich (Gramineae). Crop Sci 10:288–291 Hein MA (1953) Registration of varieties and strains of bermudagrass, II [Cynodon dactylon (L.) Pers.]. Agron J 45: 572–573 Hill GM, Gates RN, West JW (2001) Advances in bermudagrass research involving new cultivars for beef and dairy production. J Anim Sci 79(E suppl):E48–E58 Hodges EM, Boyd FT, Dunavin LS, Kretschmer AE Jr, Mislevy P, Stanley RL Jr (1975) ‘McCaleb’ stargrass. Circ S-231. Florida Agricultural Experiment Station, Gainesville, FL Hodges EM, Mislevy P, Dunavin LS, Ruelke OC, Stanley RL Jr (1984) ‘Ona’ stargrass. Circ S-268A. Florida Agricultural Experiment Station, Gainesville, FL Hurcombe R (1947) A cytological and morphological study of cultivated Cynodon species. J S Afr Bot 13:107–116

Y. Wu Kang SY, Lee GJ, Lim KB, Lee HJ, Park IS, Chung SJ, Kim JB, Kim DS, Rhee HK (2008) Genetic diversity among Korean bermudagrass (Cynodon spp.) ecotypes characterized by morphological, cytological and molecular approaches. Mol Cells 25:163–171 Karaca M, Saha S, Zipf A, Jenkins JN, Lang DJ (2002) Genetic diversity among forage bermudagrass (Cynodon spp.): evidence from chloroplast and nuclear DNA fingerprinting. Crop Sci 42:2118–2127 Kneebone WR (1966) Bermuda grass – worldly, wily, wonderful weed. Econ Bot 20:94–97 Liu JX, Guo AG, Guo HL (2003) Morphological variation and types of Cynodon dactylon. Acta Pratacult Sin 12:99–104 Malik CP, Tripathi RC (1968) Cytological evolution within the Cynodon dactylon complex. Biol Zent Bl 87:625–627 McCormick WC, Marchant WH, Southwell BL (1964) Effect of stocking level on gains of steers grazing coastal bermudagrass. Georgia Agricultural Experiment Station, Georgia McCullough ME, Neville WE (1959) Some factors affecting weight gains of dairy heifers fed all rough rations. J Dairy Sci 42:1698–1702 Mislevy P, Brown WF, Caro-Costas R, Vicente-Chandler J, Dunavin LS, Hall DW, Kalmbacher RS, Overman AJ, Ruelke OC, Sonoda RM, Sotomayor-Rios A, Stanley RL Jr, Williams MJ (1989a) ‘Florico’ stargrass. Circ S-361. Florida Agricultural Experiment Station, Gainsville, FL Mislevy P, Brown WF, Dunavin LS, Hall DW, Kalmbacher RS, Overman AJ, Ruelke OC, Sonoda RM, Stanley RL Jr, Williams MJ (1989b) ‘Florona’ stargrass. Circ S-362. Florida Agricultural Experiment Station, Gainsville, FL Mislevy P, Brown WF, Kalmbacher RS, Dunavin LS, Judd WS, Kucharek TA, Ruelke OC, Noling JW, Sonoda RM, Stanley RL Jr (1999) Registration of Florakirk bermudagrass. Crop Sci 39:587 Myers WM (1951) Registration of varieties and strains of bermudagrass [Cynodon dactylon (L.) Pers.]. Agron J 43: 240 Nelson C, Burns JC (2006) Fifty years of grassland science leading to change. Crop Sci 46:2204–2217 Ourecky DK (1963) Pachytene chromosome morphology in Cynodon dactylon (L.) Pers. Nucleus 6:63–82 Reed BM, Schumacher L, Wang N, D’Achino J, Barker RE (2005) Cryopreservation of bermudagrass germplasm by encapsulation dehydration. Crop Sci 46:6–11 Samudio SH, Brede AD (1998) Registration of ‘Cheyenne’ bermudagrass. Crop Sci 38:279 Shearman RC (2006) Fifty years of splendor in the grass. Crop Sci 46:2218–2229 Taliaferro CM (1986) Bermudagrass germplasm resources for beef production. In: Visiting scholar lectures, special report 122. Arkansas Agricultural Experiment Station, Fayetteville, AR, pp 5–19 Taliaferro CM (2003) Bermudagrass. In: Casler MD, Duncan R (eds) Turfgrass biology, genetics and breeding. Wiley, New York, pp 235–256 Taliaferro CM, Richardson WL (1980) Registration of Hardie bermudagrass. Crop Sci 20:413 Taliaferro CM, Ahring RM, Richardson WL (1983) Registration of Guyman bermudagrass. Crop Sci 23:1219

4 Cynodon Taliaferro CM, Hopkins AA, Henthorn JC, Murphy CD, Edwards RM (1997) Use of flow cytometry to estimate ploidy level in Cynodon species. Int Turfgrass Soc Res J 8:385–392 Taliaferro CM, Anderson JA, Richardson WL, Baker JL, Coleman SW, Phillips WL, Sandage JL, Moyer JL, Hanson TL, Kallenbach RL, Crawford RJ (2002) Registration of ‘Midland 99’ bermudagrass. Crop Sci 42:2212–2213 Taliaferro CM, Rouquette FM Jr, Mislevy P (2004a) Bermudagrass and stargrass. In: Moser LE, Burson BL, Sollenberger LE (eds) Warm-season (C4) grasses, vol 45, Agron Monogr. ASA, CSSA, and SSSA, Madison, WI, pp 417–475 Taliaferro CM, Martin DL, Anderson JA, Anderson MP, Guenzi AC (2004b) Broadening the horizons of turf bermudagrass. USGA Turfgrass Environ Res Online 3(2):1–9 Tracy SM (1917) Bermuda grass, farmer’s bulletin 814. United States Department of Agriculture, US Gov Print Office, Washington, DC Watson VH (1974) A chronological review of the selection, research, and distribution of a bermudagrass called Callie. Rep Mississippi Agricultural Experiment Station, Starkville, MS Wu YQ, Taliaferro CM (2009) Bermudagrass. In: Singh RJ (ed) Genetic resources, chromosome engineering, and crop improvement, vol 5, Forage crops. CRC Press, New York, pp 229–273

71 Wu YQ, Liu LL, Xiong X, Xu XG, Wang Z, Yan Y (2001) Utilization and evaluation of native germplasm of bermudagrass in Sichuan. Grassl Turf 94:32–34 Wu YQ, Taliaferro CM, Bai GH, Anderson MP (2004) AFLP analysis of Cynodon dactylon (L.) Pers. var. dactylon genetic variation. Genome 47:689–696 Wu YQ, Taliaferro CM, Bai GH, Martin DL, Anderson JA, Anderson MP, Edwards RM (2006a) Genetic analyses of Chinese Cynodon accessions by flow cytometry and AFLP markers. Crop Sci 46:917–926 Wu YQ, Taliaferro CM, Martin DL, Goad CL, Anderson AJ (2006b) Genetic variability and relationships for seed yield and its components in Chinese Cynodon accessions. Field Crops Res 98:245–252 Wu YQ, Taliaferro CM, Martin DL, Anderson AJ, Anderson MP (2007) Genetic variability and relationships for adaptive, morphological, and biomass traits in Chinese bermudagrass accessions. Crop Sci 47:1985–1994 Wu YQ, Taliaferro CM, Kochenower R (2008) Development of ‘Goodwell’ forage bermudagrass. Joint Annual Meeting of GSA, SSSA-ASA-CSSA, and GCAGS, Oct 5–9, Houston, TX, USA Yi YJ, Zhang XQ, Huang LK, Ling Y, Ma X, Li W (2008) Genetic diversity of wild Cynodon dactylon germplasm detected by SRAP markers. Hereditas (Beijing) 30(1): 94–100

.

Chapter 5

Dactylis Alan V. Stewart and Nicholas W. Ellison

5.1 Introduction The genus Dactylis consists of a single perennial forage species, Dactylis glomerata L. known as orchardgrass or cocksfoot. The basic chromosome number is 2n ¼ 14 and, although the species consists of diploids, tetraploids and a hexaploid, the commonly used forage form is tetraploid. Dactylis species have only rarely been crossed with species outside this genus (Nakazumi et al. 1997) and the genus is not closely related to any major crop. Therefore, the story of Dactylis wild relatives is largely one of the diversity within the genus. Although the distribution and diversity of Dactylis is complex, much of this can be readily explained by its evolutionary, migrational and genomic history over the last few hundred thousand years. The understanding of the diversity within the wild relatives of Dactylis has considerable implications for germplasm collection and conservation as well as for the breeding techniques employed. There is considerable urgency to make sure that all the geographically diverse diploid subspecies are collected and represented in genebanks.

fourth most widely used grass genus with 3.3% of the world’s temperate seed, following Lolium, Festuca and Phleum (Bondesen 2007). It is almost never used for turf despite fine dense turf forming types being available. This may be because of the limitations imposed by seed yield and subsequent seed price as well as suitable cultivars.

5.3 Basic Botany of the Species 5.3.1 Morphology D. glomerata is in general a perennial bunchgrass. Forms adapted to more humid habitats usually have large tillers and broad leaves while those adapted to drier conditions exhibit a more xeromorphic form with a greater number of smaller tillers with narrower glaucous leaves. Some forms growing on coastal cliffs have branching stems, which sprawl along the ground, even hanging in some situations.

5.3.2 Taxonomy of D. glomerata L 5.2 Agricultural Status D. glomerata is an important forage grass in the temperate areas of the world. Each year approximately 14,000 tons of seed is harvested making it the A.V. Stewart (*) PGG Wrightson Seeds, PO Box 175, Lincoln, Christchurch 7640, New Zealand e-mail: [email protected]

As a genus Dactylis is subject to differing taxonomic interpretations in different regions of its natural range (Europe, Asia, North Africa and the Canary Islands). There are no modern taxonomic treatments which interpret all forms on the same basis. Here we interpret D. glomerata to be monotypic consisting of one diverse species complex (Jogan 2002). The genus however, is clearly on the verge of speciation with many diploid, tetraploid and a hexaploid subspecies

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_5, # Springer-Verlag Berlin Heidelberg 2011

73

74

A.V. Stewart and N.W. Ellison

altaica

G 510 T

28 ITS 1 cp

Lamarckia aurea

C 90 T A 613 G

C 55 T A 425 G

G 92 C

judaica

A 571 G lusitanica T 190 C

A 601 G C 649 A C cp444 T

ibizensis G 610 A

izcoi juncinella

G 475 T

592 T insert santai sinensis aschersoniana castellata smithii

parthiana

himalayensis

cpDNA Del 287

C 627 T mairei

woronowii

Fig. 5.1 Molecular descent in diploid Dactylis glomerata using ITS and trnL intron chloroplast sequences (Numbers refer to mutations in the ITS or chloroplast (cp) sequences)

and some reduction in fertility when different diploid forms are crossed (Borrill 1977).

5.3.2.1 Diploid D. glomerata Forms Today almost all diploid populations have restricted ranges, indicative of a remnant distribution from a former widespread range. The populations are usually described as subspecies, or sometimes even species, and publications vary as to their interpretation. Here we will not enter into any taxonomic debate but recognize them as discreet remnant geographic populations. They occur from China across to Portugal and North Africa as well as on to nearby north Atlantic islands such as the Canary Islands and Cape Verde.

5.3.2.2 Origin of Diploids It is very clear that Dactylis originated as a diploid and that the polyploids have developed from them. Molecular information on diploids using the nuclear internal transcribed spacer (ITS) and trnL intron chloroplast sequences, reported here for the first time, support the

hypothesis that the diploid progenitor originated in, and migrated out from Central Asia (Fig. 5.1). This progenitor is probably very similar to subsp. altaica. We have observed no more than 2–7 ITS mutations and only 0 or 1 trnL intron chloroplast mutation in any of the diploid Dactylis lineages and this allows an estimate of timing of the original divergence. The rate at which ITS and chloroplast sequences mutate depends on many factors including effective population size, annual or perennial habit, and the breeding system. In annual grasses, such as barley and maize, the ITS mutation rate has been calculated to be approximately one mutation every 23,000 years (Zurawski et al. 1984), while the rate in perennial grasses is slower and potentially similar to that observed in perennial Phleum, where Stewart et al. (2008) observed the most rapidly changing lineage to be one in 30,000 years. Chloroplast genome trnL intron mutation rates are usually 3–8 times slower, with one per 200,000 years in rice and maize (Yamane et al. 2006) and one per 90,000 years for the most rapidly changing perennial Phleum lineage (Stewart et al. 2008). These rates for Dactylis would suggest that the first divergence occurred in the order of 60,000–210,000 years BP, a date which is far more recent by orders of

5 Dactylis

magnitude than the 10 million years, or more, dating back to the tertiary period postulated by Stebbins and Zohary (1959) who used geological events as the explanation. However, it is only with the recent advent of molecular data that the powerful influence of historical glaciation events on plant migration and evolution have become to be fully appreciated (Hewitt 1999). This relatively recent origin for Dactylis is much more consistent with the perceived interglacial age of many of the tetraploids (Lumaret 1986) and there is no reason to believe that the process of tetraploidy was not initiated early in the development of Dactylis. This time period is also consistent with the observation that, while there is sufficient genetic differentiation to result in reduced fertility among crosses of the diploid subspecies, there are no major structural chromosomal differences (Borrill 1977). One of the closest genera to Dactylis is Lamarckia, and L. aurea diverged from it with more than 28 ITS sequences differences and 1–2 trnlL intron chloroplast differences (Fig. 5.1). As L. aurea is an annual, these differences suggest a divergence time in the order of 200,000–750,000 years BP. The first molecular divergence within Dactylis (Fig. 5.1) divides judaica, himalayensis and the paternal side of parthiana from the remaining diploids. This is consistent with an origin in Central Asia for Dactylis and an early geographic split leaving the ancestors of judaica, himalayensis and parthiana spread widely over the southern regions of Western Asia, and the ancestors of the remaining diploids in the north-west of this region. The remaining diploids exhibit a second divergence (Fig. 5.1) with the Spanish forms lusitanica, izcoi, juncinella and ibizensis splitting off from the group. This is consistent with a migration through the Eurasian temperate forest regions from the Caucasus to Portugal. This most probably occurred during a warm interglacial period, probably 75,000–150,000 years BP, when conditions allowed continuous distribution of temperate forest in Europe. This European form has since migrated into North Africa and over to China as the molecular profiles of aschersoniana, santai, “castellata”, smithii and sinensis are identical, with both woronowii and mairei derived from this group. The lack of molecular differences suggests a relatively recent migration, probably no older than the last glacial period. Glaciation in Europe forced the European flora south into dissected

75

glacial refugia at lower altitudes or latitudes (Hewitt 1999) and allowed Dactylis to expand into the extensive North African grasslands. Subsp. woronowii shows a very similar flavonoid composition to aschersoniana (Lumaret 1988) and exhibits a molecular sequence derived from it, but its morphological characteristics are more similar to other xeromorphic Mediterranean types. It is likely that this entity originated from the forest margin dwelling aschersoniana to become the first xeromorphic form adapted to drier grassland conditions during the interglacial period. Ancestors of smithii would have easily migrated the 100 km from North Africa to the Canary Islands, and it is likely that bird migration was responsible for the larger 1,500 km movement to the Cape Verde Islands. The origin of smithii from aschersoniana and/or woronowii is one of the options put forward by Stebbins. In his opinion smithii was either a very old remnant equally as old as aschersoniana, or a recent migrant via North Africa (Stebbins and Zohary 1959). Its unique morphological characteristics made him favor the former option whereas our molecular data support a more recent migration via North Africa. Sahuquillo and Lumaret (1999) also provide evidence of tetraploid Dactylis migrations from North Africa to the Canary Islands, they used chloroplast molecular markers supported by morphological, allozyme and phenolic indicators. In the post-glacial period as the climate became warmer and drier North African grasslands contracted leaving dissected remnant populations, mairei in the Kerrata Gorge of Algeria, santai and “castellata” in Algeria and Morocco, and smithii in the Canary and Cape Verde Islands. As the glaciated conditions of northern Europe contracted, the northern temperate forms could migrate from their glacial refugia to high altitudes or latitudes. The European progenitor of aschersoniana in particular was able to recolonize northern Europe from a probable refuge in the Caucasus region, even extending to the east to China as sinensis. It seems probable also that the tetraploids also recolonized Europe to cover northern regions so successfully that they excluded the expansion of the remaining diploids. The Iberian lusitanica progenitor was able to migrate throughout Iberia so that today we have a dissected distribution with lusitanica in central Portugal,

76

izcoi in Galicia and juncinella in the Sierra Nevada Mountains. The ancestral form also migrated to the Balearic Isle, where ibizensis occurs today. The Iberian Peninsula is composed of a complex set of different glacial refugia which, if understood, may help explain the relationships between the different forms present today (Go´mez and Lunt 2006). It is also apparent from hybrid ITS sequences of tetraploids in North Africa, the Canary Islands and Spain that many of these forms have resulted from hybridization between North African and Spanish forms. It is clear also that parthiana has been influenced by genetic introgression from overlapping woronowii populations. The probable migration of the diploids determined from the molecular research is outlined in Fig. 5.2. As yet we have been unable to obtain samples of hyrcana, reichenbachii, metlesicsii or smithii from Cape Verde to determine their origins.

A.V. Stewart and N.W. Ellison

Himalayensis D. glomerata subsp. himalayensis Domin occurs in the western Himalayas at altitudes of 1,800–4,000 m in cool temperate forest zones.

Sinensis D. glomerata subsp. sinensis Camus occurs in Sichuan, Hubei, Guizhou, Xinjiang and Yunnan in central China, at altitudes of 1,000–3,800 m in cool temperate forest zones. Although some authors have classified these Chinese forms into himalayensis (Stebbins and Zohary 1959; Borrill 1977) the molecular differences reported here and the genetic differences reported by Lumaret (1987) provide considerable support to maintaining sinensis as separate from himalayensis.

5.3.2.3 The Diploid Subspecies

Altaica

The following are the known diploid remnant populations:

The diploid D. glomerata subsp. altaica (Bess.) Domin occurs in the Alatau Mountains of Kazakhstan (Czerepanov 1981; Mizianty 1991). The one sample

Fig. 5.2 Probable migration routes of diploid Dactylis based on molecular results. (black lines – before glaciation, dashed lines – North Africa during the glaciation; dotted lines – post-glacial, Northern Europe and China)

5 Dactylis

that we have obtained from this region had an ITS sequence basal to all other diploid Dactylis forms (Fig. 5.1). No collections are available in genebanks and collection must be a priority.

77

can not discount the possibility that this genebank accession of parthiana was collected in the zone of hybridization and that those higher altitude forms which may be collected less frequently may be more “pure”.

Aschersoniana D. glomerata subsp. lobata (Drejer) Lindb.  subsp. aschersoniana (Graebn.) Thell.  D. aschersoniana Graebn.  D. polygama Horv.  subsp. polygama (Horv.) Domin occurs mostly in deciduous forests from the Caucasus Mountains across central Europe, into Sweden in the north, western Russia in the east, south into Yugoslavia and northern Macedonia. This distribution is typical of a species recolonizing Europe from a glacial refuge in the Caucasus region (Hewitt 1999). The habitats of aschersoniana and himalayensis are similar as they both occur in cool temperate forests in continental climates often at high altitudes and they bear a close resemblance in isozyme allelic patterns (Lumaret 1988), flavone phytochemistry (Lumaret 1988), DNA content (Tuna et al. 2007) and morphology, so much so that the geographic origin is often necessary to correctly assign them to the appropriate taxa (Stebbins and Zohary 1959). This form has potentiality to be used in agriculture (Stuczynski 1992) and there is at least one commercial cultivar, Tosca, bred in the Czech Republic (Mı´ka et al. 1999).

Parthiana D. glomerata subsp. parthiana Parker et Borrill. occurs in the moist oak woodlands on the northern slopes of the Elburz Mountains of Iran between 1,600 and 1,900 m. At lower altitudes its distribution overlaps with woronowii and hybrid plants occur (Borrill and Carroll 1969). Parker and Borrill (1968) showed that artificial hybrids of this cross are fully fertile. Our molecular results provide confirmation of this hybridization in the only genebank accession available, which combines a himalayensis type ITS sequence with a woronowii type chloroplast sequence. Similarly, flavone phytochemistry has shown that both parental pathways are present (Ardouin et al. 1985), and electrophoretic patterns are also more variable than other subspecies (Lumaret 1988). However, we

Reichenbachii D. glomerata subsp. reichenbachii (Dalla Torre et Sarnth.) Stebbins et Zohary  D. reichenbachii (Dalla Torre et Sarnth.) Meusel occurs on south-facing meadows on dolomite soils of the European Alps and French Massif Central (Speranza and Cristofolini 1987). Although we were unable to obtain a sample for molecular analysis, flavonoid phytochemistry places this between aschersoniana and lusitanica (Fiasson et al. 1987), while isozyme data places it very close to aschersoniana and himalayensis (Lumaret 1988). Stebbins and Zohary (1959) note that this form is highly variable exhibiting characteristics of both woronowii and aschersoniana, and is likely to be an edaphic remnant of the original European temperate forest flora.

Lusitanica D. glomerata subsp. lusitanica Stebbins et Zohary is restricted to the Serra de Sintra, and a few other localities in central-southern Portugal (Stebbins and Zohary 1959). It appears to be a remnant temperate forest species of the mildest temperate forest zone.

Iscoi D. glomerata subsp. izcoi Ortiz et Rodriguez-Oubina, has also often been referred to as “Galician diploid”. This temperate form occurs in Galicia, Spain, persisting in forest habitats at altitudes of 400–650 m (Ortiz and Rodriguez-Oubin˜a 1993; Lindner et al. 2004). It shows the greatest molecular similarity with juncinella and both must have developed from populations in the same complex glacial refugia of the Iberian Peninsula (Go´mez and Lunt 2006). The UK commercial cultivar Conrad is a diploid izcoi type (Borrill 1977).

78

Woronowii D. glomerata subsp. woronowii (Ovcz.) Stebbins et Zohary  D. woronowii Ovcz. is a xeromorphic grassland taxon in the mountain steppes and steppe forests in northern and north-eastern Iran and adjacent Turkmenistan (Stebbins and Zohary 1959). It occurs on dry base-rich limestone outcrops in open thorn scrub up to 1,500 m in the Elburz Mountains (Borrill and Carroll 1969). Sympatric tetraploids are also common in this region (Doroszewska 1963).

Hyrcana D. glomerata subsp. hyrcana Tzvelev is an endemic subspecies of deciduous woods in the Talish plains of Azerbaijan and Iran at 200–300 m altitude (Tzvelev 1983). We were unable to obtain samples so it is unclear how this relates to the others but its habitat appears to be an extension of the range of woronowii and it is likely to be closely related to it.

Mairei D. glomerata subsp. mairei Stebbins et Zohary is confined to shaded limestone cliffs of the Kerrata Gorge in Algeria with a relatively high annual rainfall of 1,100–2,000 mm (Stebbins and Zohary 1959). It is highly probable that since the last glaciation its range has become restricted as the North African grasslands have declined (Borrill and Lindner 1971). Molecular, isozyme (Lumaret 1988) and flavone data (Ardouin et al. 1985) show that it exhibits close phylogenetic affinity to “castellata”, santai, smithii and woronowii. This form is an isolated remnant of a diploid widespread in the North African grasslands since the last glaciation (Borrill and Lindner 1971).

Santai and “castellata” D. glomerata subsp. santai Stebbins et Zohary and subsp. “castellata”. occur in western Algeria to western Morocco at 150–1,500 m in the Tell Atlas, in humid and subhumid bioclimates (Amirouche and Misset 2007). Flavonoid phytochemistry and enzyme data fail to differentiate between “castellata” and santai,

A.V. Stewart and N.W. Ellison

as do morphological comparisons, so these two probably do not deserve separate subspecies status (Lumaret 1988; Amirouche and Misset 2007). The names, however, may remain useful as indicators of geographic origin with “castellata” from Algeria and santai from Morocco. The Moroccan forms of santai show characteristics of the nearby ibizensis diploid (Mousset 2000) and this is also apparent in hybrid ITS sequences which suggest introgression from nearby southern Spanish forms.

Smithii D. glomerata subsp. smithii  D. smithii Link. occurs in the wetter sides of the Canary Islands at altitudes of 100–700 m. It has a distinctive growth habit with branching culms with numerous nodes (Stebbins and Zohary 1959). Today its habitat has reduced because of construction in this zone. The form on the Cape Verde Islands at 1,200 m is most probably extinct (Lumaret pers. comm.). Isozyme studies group smithii near mairei, santai and “castellata” (Lumaret 1988). In hybridization studies there was evidence for cytoplasmic and nuclear differentiation in smithii (Parker and Borrill 1968). Clearly smithii is the most differentiated in a range of features. Its habitat is more subtropical and this has influenced its flavonoid chemistry and morphology greatly (Ardouin et al. 1985). Its flowering habit has changed so that it requires little vernalization for flowering and its trailing habit may be due to a lack of apical dominance.

Metlesicii Dactylis metlesicsii Scho¨nfelder et D. Ludwig occurs at a high altitude in the Canary Islands and is potentially a high altitude form of smithii (Scho¨nfelder and Ludwig 1996) or results from a separate migration into the Canary Islands. We have not been able to obtain a sample of this form for analysis. This form is listed as an endangered species in the Spanish Red List, although one could question its species status. Analysis of phenolic compounds showed that high altitude tetraploid forms from Gran Canary Island were well differentiated from the lower altitude forms in the region (Jay and Lumaret 1995). This

5 Dactylis

suggests that diploid metlesicsii may well be substantially different from lower altitude smithii.

79

D. glomerata subsp. juncinella (Bory) Boiss occurs in the Sierra Nevada range in Spain at altitudes of 2,200–2,900 m in the subalpine and alpine herb belt. Its flavonoid phytochemistry exhibits similarity to lusitanica, ibizensis, and less so to izcoi (Ardouin et al.1985). Hybridization studies suggest a close relationship between juncinella and ibizensis and a reasonably close relationship between smithii and juncinella (Wei-Lin Hu and Timothy 1971).

This supports a recent post-glacial change from the temperate form to the more summer dormant xeromorphic Mediterranean form. Judaica is also most probably a relic population of the Eurasian temperate flora that extended into the ancient forests of Lebanon and Israel. However, the rapid post-glacial warming and drying of the climate in this region has forced it to adopt a xeromorphic growth pattern and morphology in response to the Mediterranean climate. This is supported by a heterogeneous flavone phytochemistry with some individuals showing characteristics of “temperate” phenolics (Ardouin et al. 1985), a phenomenon attributed to evolutionary lag between a temperate electrophoretical pattern (Lumaret 1984, 1986) and Mediterranean morphology (Borrill 1977).

Ibizensis

5.3.2.4 Tetraploid D. glomerata Forms

D. glomerata subsp. nestorii Rossello et L. Sa´ez  D. ibizensis Gandoger  subsp. ibizensis (Gandoger) Stebbins et Zohary. Although renamed subsp. nestorii by Castro et al. (2007), it is still commonly known as ibizensis in almost all of the literature. It occurs on the Balearic Isle near Spain and is derived from the original Spanish forms but with some molecular introgression from nearby North African mairei. Sympatric tetraploid forms also occur on this Island (Castro et al. 2007).

Tetraploids occupy almost the full range of the species and consist of both autotetraploids based on single diploid subspecies, and interecotypic hybrids based on more than one diploid subspecies. These interecotypic tetraploids often have a selective advantage over their diploid counterparts.

Juncinella

Judaica D. glomerata subsp. judaica Stebbins et Zohary, and subsp. lebanotica. Subsp. judaica occurs in the hills of Israel while lebonotica occurs in Lebanon (Stebbins and Zohary 1959; Apiron and Zohary 1961) and potentially also in Syria. Whether these deserve taxonomic differentiation is doubtful but the names at least have value as indicators of geographic origin. It has a Mediterranean summer dormant growth pattern yet the molecular results show that it has developed from a temperate himalayensis progenitor. The heterogeneous flavone phytochemistry of judaica shows some individuals having characteristics of “temperate” phenolics (Ardouin et al. 1985). This heterogeneity could be related to the evolutionary lag between temperate patterns (Lumaret 1984, 1986) and Mediterranean morphology (Borrill 1977).

5.3.2.5 Origin of Tetraploids It seems likely that the process of tetraploidization was initiated early in the development of Dactylis. Unreduced gametes are known to occur frequently in diploid populations and gene flow is largely from diploids to tetraploid with triploids generally being aborted at an early stage (Lumaret and Barrientos 1990; Lumaret et al. 1992). Early tetraploids were most probably autotetraploids in the sense that they were based upon only one diploid population and as such they may never have had any selective advantage over their diploid counterparts. However, with the forced retreat of both diploids and tetraploids into confined glacial refugia during the last glaciation, some tetraploids were able to benefit from gene exchange with a number of ecologically divergent diploid populations. These interecotypic hybrid tetraploids often gain considerable evolutionary success by their hybrid vigor and their enhanced genetic variation (Stebbins 1971). According to Soltis and Soltis (1993) tetraploids have more polymorphic loci than diploids, 0.80 compared

80

to 0.70, higher heterozygosity, 0.43 compared to 0.17, and a greater number of alleles per locus, 2.36 compared to 1.51. The hybrid vigor obtained from these interecotypic tetraploids readily allowed them to dominate the post-glacial expansion to such an extent that we have an almost continuous distribution of tetraploids over the full range of the species today (Lumaret 1986). Autotetraploids frequently have a sympatric distribution with the diploid population that they are based upon (Borrill and Lindner 1971; Lumaret et al. 1989) while interecotypic hybrid tetraploids based on multiple diploid populations are more widespread. It is commonly reported that almost all the diploid entities have a sympatric association of autotetraploids or at least tetraploids derived predominantly from a single diploid. Often this coexistence is dependent upon each occupying different ecological niches (Maceira et al. 1993). The taxonomically distinguishable tetraploid subspecies include the widespread glomerata, the taller calciphilous slovenica, the xeromorphic hispanica, the seaside forms of marina  hackelii and oceanica and the cliffside hylodes. However, it is probably fair to conclude that the tetraploid entities merge into one complex species where characters vary along mutually independent clines (Speranza and Cristofolini 1986).

5.3.2.6 The Tetraploid Subspecies

A.V. Stewart and N.W. Ellison

Stebbins also notes that as reichenbachii is similar to woronowii in many features it is possible that some glomerata have developed from the hybridization of reichenbachii and aschersoniana in the Alps where they occur together. Domin (1943) notes that the primary distribution of subsp. glomerata is in Northern Europe but that it has a secondary distribution in the temperate forest regions of Algeria and Morocco. However, in this case it is likely that the ancestral diploids of the two forms may be different. Although subsp. glomerata is usually considered to be summer active and hispanica summer dormant, glomerata forms with summer dormancy are known to occur in the Languedoc-Roussillon Mediterranean climatic region of France (Mousset 1995).

Slovenica D. glomerata subsp. slovenica (Domin) Domin  D. slovenica Domin occurs in the mountains of Central Europe mostly at altitudes between 600 and 1,300 m and usually on calcareous or dolomitic soils (Fig. 5.3; Mizianty 1997). One of the most notable characteristics of slovenica is its plant height which can be up to 1.5 m (Domin 1943). It is likely that diploid aschersoniana has contributed to this form as they are both tall and have a hairless keel and margin (or few hairs) (Mizianty and Cenci 1995). The current distribution suggests that it has spread from a glacial refuge in the Alps or Carpathians.

Five tetraploid subspecies are recognized. Hispanica Glomerata D. glomerata L. subsp. glomerata is the most widespread tetraploid entity of temperate forest regions and is the form most widely used in agriculture. Stebbins and Zohary (1959) suggest that it may have developed from the interecotypic hybridization between aschersoniana and woronowii, as artificial hybrids are intermediate between the two and indistinguishable from subsp. glomerata. These subspecies occur in close geographic proximity and both could indeed have provided the very broad interecotypic tetraploids necessary for successful colonization of Europe from the Caucasus glacial refugia.

D. glomerata subsp. hispanica (Roth) Nyman  D. hispanica Roth is widespread in the Mediterranean region from the Iberian Peninsula to the Crimea and Caucasus. It is xeromorphic exhibiting a smaller plant size with a compact panicle of limited branching (Speranza and Cristofolini 1986). It usually exhibits early flowering and a summer dormant growth pattern which provides good drought tolerance. Domin (1943) notes that hispanica is a variable complex with one stable feature, having lemmas divided into two blunt lobes from the middle of which is a short awnlet, or a mere point which sometimes does not reach the size of an awn. This lobed

5 Dactylis

81

Fig. 5.3 Distribution of Dactylis glomerata subsp. slovenica [after Mizianty (1997)]

characteristic is less pronounced in the north where it shows a transition towards the large growing glomerata plants used in agriculture. The summer dormant forms in the LanguedocRoussillon Mediterranean region of France represent hybrids, or introgression hybrids, of hispanica and glomerata (Lumaret 1986). A number of summer dormant cultivars developed of hispanica types from Mediterranean germplasm including Uplands, Sendace, Berber, Kasba, (Australia), Jana (Italy), Medly (France), Perouvia and Chrysopigi (Greece), and germplasm of this form offer considerable potentiality for breeding cultivars suited to Mediterranean climates.

provide the plant with an advantage under coastal conditions. The smooth leaf margin trait is believed to enhance palatability (van Dijk 1961) while the herbage has superior digestibility in the order of 2–6%. For this reason marina has attracted the attention of plant breeders (Tyler and Borrill 1983) but as yet this feature has not resulted in any cultivar, presumably because marina crosses have not had adequate yields. Marina is believed to have arisen from diploids closely associated with smithii and ibizensis (Borrill 1961). However, it is also possible that the characteristic smooth leaf margins, papillose epidermis and glaucous appearance could arise in material of different origins as well.

Marina

Oceanica

D. glomerata subsp. hackelii (Asch. et Graeb.) Cif. et Giacom.  D. marina Borrill  D. glomerata L. subsp. marina (Borrill) Greuter  D. smithii Link subsp. marina (Borrill) Parker is localized on sea cliffs of the Mediterranean extending along the coast of Southwest Europe (Portugal and Galicia) and the Atlantic Islands (Borrill 1961; Fig. 5.4). Marina is characterized by smooth leaf margins devoid of silicified teeth, papillose epidermis and a particularly glaucous appearance, all features which

D. glomerata subsp. oceanica G. Guignard occurs on the coast of southern France and the Atlantic and the Channel coasts of north-western France (Guignard 1985). This form could potentially be viewed as an extension of the range of the marina type into northern Europe, however, the marina types of the Mediterranean exhibit many hispanica features while oceanica exhibits many glomerata features. Forms similar to oceanica have been observed along the coast of Cornwall and an Irish island (Lumaret 1988).

82

A.V. Stewart and N.W. Ellison

Fig. 5.4 Distribution of Dactylis glomerata subsp. marina [after Borrill (1961)]

Hylodes D. glomerata subsp. hylodes P.F. Parker is confined to inland cliffs of Madeira, Canary Islands, and Cape Verde whereas marina occurs on the coastal cliffs (Parker 1972). This subspecies is closely related to the diploid smithii with persistent, trailing, woody stems and papillose leaves. Parker (1972) noted that both hybridization and polyploidy have provided a physiological versatility allowing this subspecies to maintain a wider distribution than smithii. Our molecular results show that these tetraploids have profiles from both smithii and Spanish diploids.

5.3.2.7 Hexaploid D. glomerata Hexaploid populations have been reported from a restricted range within Libya and western Egypt (Jones et al. 1961; Jones 1962). Potentially these are also interecotypic polyploids, but little is known of their origin. Two hexaploid plants are reported from the coast near Pontevedra in Galicia, Spain but these are probably rogue individuals within the local tetraploid rather than a self maintaining hexaploid population (Horjales et al. 1995).

5.3.3 Climatic Races The development of climatic races has been part of the continuing process of adaptive radiation with climatic races adapted to drier conditions and also to warmer subtropical conditions. The oldest diploid forms such as altaica, himalayensis and aschersoniana adapted to temperate forest margins, while their derivative woronowii developed xeromorphic features as it expanded into the drier grasslands of western and central Asia. These xeromorphic features are also typical of hispanica tetraploids in the drier Mediterranean, and include smaller, more glaucous leaves and tillers as well as smaller inflorescences. Independently, xeromorphic forms of judaica have developed in the Mediterranean from temperate pre-himalayensis types. The seaside dwelling forms marina and oceanica have developed glaucous characteristics and in the case of marina have papillate epidermises. Those on coastal cliffs (hylodes, smithii) have developed a sprawling branching habit.

5.3.4 Edaphic Adaptation The development of edaphic races has also been part of the continuing process of adaptive radiation so that

5 Dactylis

today many commercial glomerata forms are noted for their tolerance of acidic soils with aluminum, and subsp. slovenica grows on calcareous soils while reichenbachii is adapted to dolomitic soils.

5.3.5 Genome Size Genome size in Dactylis has been reported to decrease with altitude in France and Italy, often by as much as 30% (Reeves et al. 1998). There is also approximately a 15% reduction in the DNA mass of diploids forms, which have developed in southern Iberia, North Africa and the Canary Islands when compared to those in northern Iberia, Europe, Middle East and central Asia (Tuna et al. 2007). Interestingly, these are probably more recent in origin.

5.4 Genetic Resources For plant improvement it is appropriate to divide genetic resources into primary, secondary, tertiary and quaternary gene pools as follows.

83

many Islands. These resources are often explored by plant breeders but they are of lesser significance than the primary pool simply because adaptation to agricultural environments is usually inferior. In general, these resources still occur in nature and in genebank collections. Many of the wild populations are under threat with climate warming and humaninduced habitat changes, while even those in genebanks will only be maintained if adequate resources are available. The secondary gene pool includes tetraploid forms sympatric with diploid populations. The Australian cultivar Porto and the Spanish cultivar Adac 1 are based on natural tetraploids with lusitanica influence. A number of cultivars have been based on tetraploid forms of izcoi from Galicia, such as Grasslands Wana (New Zealand), Cambria (UK) and Artibro (Spain). It is also possible that other cultivars are based on tetraploid forms of diploid populations. The tetraploid forms of many of these diploid populations represent an enormous resource for breeders and representative collections are required. Although many have been collected, few are identified as tetraploid forms of the remnant diploid population.

5.4.3 Tertiary Gene Pool 5.4.1 Primary Gene Pool These can be defined as cultivars and elite breeding lines adapted to the region of agricultural use. This is the primary activity of most effective plant breeders and consists almost exclusively of tetraploids, although the diploids izcoi and aschersoniana have been used commercially. In general, these resources are well used by breeders and their “working” collections largely represent this gene pool. The largest threat to this gene pool would be any decrease in the number of breeding programs in Dactylis.

5.4.2 Secondary Gene Pool This can be defined as those tetraploid populations outside the regions of main agricultural use of the species in Europe, Asia and North Africa and the

These can be defined as those populations from other ploidy levels where introgression into commercial breeding programs is possible. These would largely consist of diploids, but potentially the hexaploids could also be used for tetraploid breeding. Breeders have crossed germplasm from diploid lusitanica into tetraploid material with examples being the two UK commercial cultivars Saborto and Calder, as well as the NZ cultivar Grasslands Kara. Tetraploidy can be induced from the diploids by colchicine treatment as in Saborto, or via unreduced gametes as in Calder (Lewis 1975). In general, these resources are poorly represented in genebanks and are seldom used by breeders. Yet because of the foundation upon which modern tetraploids are based, the remnant diploid populations represent an enormous resource. Many are under threat with climate warming and human induced habitat changes. Even those in genebanks will only survive if adequate funding is provided.

84

5.4.4 Quaternary Gene Pool This can be defined as those species outside Dactylis, which may be a source of genes for breeding. However, only a few species have ever been hybridized with Dactylis, such as Lolium multiflorum (Oertel et al. 1996), Festuca arundinacea (Matzk 1981), and Phleum pratense (Nakazumi et al. 1997), and successful embryos were formed in cereals (wheat, barley, rye) pollinated with D. glomerata (Zenkteler and Nitzsche 1984). As fertility is extremely low these hybrids have never been fully explored in breeding programs and their development does not appear to be a priority.

5.4.5 Conservation Recommendations We can conclude that genetic resources of the primary and secondary tetraploid gene pools are moderately well represented in genebanks and breeders’ collection. The tertiary diploid gene pool is in urgent need of collection as many of the geographic forms are not present in international genebanks or are represented by no more than two accessions (Table 5.1). In particular, altaica, hyrcana, metlesicsii, reichenbachii and smithii from Cape Verde are absent from collections, and sinensis is difficult to obtain for anyone outside China. It is also important to collect and store the autotetraploid forms of the diploid populations such as smithii, woronowii, izcoi, reichenbachii and mairei, and also of any others that occur. Although in situ conservation should be encouraged, it is important to maintain ex situ collections of each form. These can be stored as seed collections in the international agricultural genebanks.

5.4.6 Genomics Resources Bushman et al. (2007) and Robins et al. (2008) are generating Dactylis expressed sequence tag (EST) libraries using four tissues: cold acclimated crowns, etiolated seedlings, salt/drought stressed shoots, and salt/drought stressed roots. They are also using genetic association mapping populations, and trait evaluation field plots to help overcome winter hardiness and to provide tolerance to dry summers. Simple sequence

A.V. Stewart and N.W. Ellison Table 5.1 Diploid accessions of Dactylis glomerata in genebanks of Europe (Eurisco), North America (USDA and Canada) and New Zealand (Margot Forde Germplasm Center) with priority for collection Diploid population Number of Priority accessions for collection altaica 0 Urgent aschersoniana 26 Low “castellata” 3 High himalayensis 4 High hyrcana 0 Urgent nestorii ¼ ibizensis 2 High izcoi 40 Low judaica 10 Moderate juncinella 4 High lusitanica 11 Moderate mairei 10 Moderate metlesicsii 0 Urgent parthiana 3 High reichenbachii 0 High santai 9 Moderate sinensis 0 Urgent smithii (Canary Islands) 10 Moderate smithii (Cape Verde) 0 Urgent woronowii 21 Low Note: these collections are likely to include some duplication

repeats (SSRs) identified from these libraries will be aligned to rice chromosomes to determine predicted locations of the SSR markers. PCR primers designed from 30 UTR regions from other species contain high degrees of polymorphism, and the same is expected for this library. Additionally, the sequence data from the library will be used to identify single nucleotide polymorphisms (SNPs). The USDA-ARS Forage and Range Research Laboratory have an extensive orchardgrass improvement program, and their objective is to identify phenotypic associations, and apply the markers to a marker-assisted selection program for increased salt tolerance and winter hardiness. In order that molecular resources can be applied in an effective and non-dominating balanced way, it is important to ensure that pragmatic field breeding programs continue in all major regions.

5.4.7 Karyotype The karyotypes of many diploid Dactylis forms has been reviewed by Wetschnig (1991). There has been no

5 Dactylis

research on cytogenetic stocks, such as addition or substitution lines. Williams and Barclay (1972) described the transmission of B chromosomes in Dactylis.

5.4.8 Inheritance A number of inheritance studies in tetraploid Dactylis have shown that it essentially behalves as an autopolyploid. The genomes are not greatly differentiated and gene exchange occurs between genomes.

5.4.9 Role in Crop Improvement Through Traditional and Advanced Tools While the majority of D. glomerata cultivars have been developed by traditional breeding, some have been developed using introgression from diploids, notably lusitanica. Some tetraploid lines may be based on interecotypic hybrids while others may be largely single origin tetraploids. Transformation systems to develop transgenic plants have been developed for Dactylis (Cho et al. 2001), but at this stage no transgenic cultivars have been released.

5.4.10 The Future Improvement of Dactylis Using Wild Relatives Germplasm of many of the diploid forms is under serious threat from habitat degradation and climate warming in situ. Unfortunately, many of these forms are absent from or are not well represented in ex situ genebanks and it is critical that a wide range of these forms be collected for storage. It is also crucial that viable large scale breeding programs are maintained internationally to allow adequate cultivar development, germplasm collection, introgressions of wild germplasm and application of molecular resources. Acknowledgements The authors wish to acknowledge the support of PGG Wrightson Seeds and AgResearch, and the numerous people who supplied germplasm.

85

References Amirouche N, Misset MT (2007) Morphological variation and distribution of cytotypes in the diploid-tetraploid complex of the genus Dactylis L. (Poaceae) from Algeria. Plant Syst Evol 264:157–174 Apiron D, Zohary D (1961) Chlorophyll lethal in Natural populations of the Orchard Grass (Dactylis glomerata L.). A case of balanced polymorphism in plants. Genetics 46:393–399 Ardouin P, Fiasson JL, Jay M, Lumaret R, Hubac JM (1985) Chemical diversification within Dactylis glomerata L. polyploid complex (Graminaceae). In: Jacquard P, Heim G, Antonovics J (eds) Proceedings of the NATO symposium on genetic differentiation and dispersal in plants, vol G5. Springer, Basel, Switzerland Bondesen OB (2007) Seed production and seed trade in a globalised world. Seed production in the northern light. Proceedings of the 6th international herbage seed conference in Norway, pp 9–12 Borrill M (1961) Patterns of morphological variation in diploid and tetraploid Dactylis. Bot J Linn Soc 56:441–459 Borrill M (1977) Evolution and genetic resources in Cocksfoot. Annu Rep Welsh Plant Breed Stat, pp 90–209 Borrill M, Carroll CP (1969) A chromosome atlas of the genus Dactylis (part two). Cytologia 34:6–17 Borrill M, Lindner R (1971) Diploid-Tetraploid sympatry in Dactylis (Gramineae). New Phytol 70:1111–1124 Bushman BS, Larson SR, Wang R, Mott IW (2007) New genomic resources for pasture and range grasses. Poster presentation, 5th international symposium on the molecular breeding of forage and turf, 1–6 July 2007, Sapporo, Japan Castro M, Fraga P, Torres N, Rossello JA (2007) Cytotaxonomical observations on flowering plants from the Balearic Islands. Ann Bot Fenn 44:409–415 Cho MJ, Choi HW, Lemaux PG (2001) Transformed T0 orchardgrass (Dactylis glomerata L.) plants produced from highly regenerative tissues derived from mature seeds. Plant Cell Rep 20:318–324 Czerepanov SK (1981) Sosudistye Rasteniia SSSR. Nauka, Leningradskoe Otdnie, Leningrad, Nauka, USSR, p 509 (In Russian) Domin K (1943) Monografica studie o rodu Dactylis L. Acta Bot Bohem 14:3–147 Doroszewska AA (1963) An investigation on the diploid and tetraploid forms of Dactylis glomerata L. subsp. woronowii (Ovczinn.) Stebbins et Zohary. Acta Soc Bot Pol 32: 113–130 Fiasson JL, Ardouin P, Jay M (1987) A phylogenetic groundplan of the specific complex Dactylis glomerata. Biochem Syst Ecol 15:225–230 Go´mez A, Lunt DH (2006) Refugia within refugia: patterns of phylogeographic concordance in the Iberian Peninsula. In: Weiss S, Ferrand N (eds) Phylogeography of southern European refugia. Springer, The Netherlands, pp 155–188 Guignard G (1985) Dactylis glomerata ssp. oceanica, a new taxon of the Atlantic coast. Bull Soc Bot France Let Bot 132:341–346 Hewitt GM (1999) Post glacial recolonisation of European biota. Biol J Linn Soc 68:87–112

86 Horjales M, Redondo N, Pe´rez B, Brown S (1995) Presencia en Galicia de Dactylis glomerata L. hexaploide. Bol Soc Brot 67:223–230 Jay M, Lumaret R (1995) Variation in the subtropical group of Dactylis glomerata L. 2. Evidence from phenolic compound patterns. Biochem Syst Ecol 23:523–531 Jogan J (2002) Systematics and chorology of Cocksfoot Group (Dactylis glomerata agg.) in Slovenia. Doctoral Dissert, Univ of Llubljana, Slovenia Jones K (1962) Chromosomal status, gene exchange and evolution in Dactylis. Genetica 32:272–295 Jones K, Carroll CP, Borrill M (1961) A chromosome atlas of the genus Dactylis. Cytologia 26:333–343 Lewis EJ (1975) An alternate technique for the production of amphidiploids. Annu Rep Welsh Plant Breed Stn 1974:15 Lindner R, Lema M, Garcı´a A (2004) Extended genetic resources of Dactylis glomerata subsp. izcoi in Galicia (northwest Spain). Genet Resour Crop Evol 51:437–442 Lumaret R (1984) The role of polyploidy in the adaptive significance of polymorphism at the GOT 1 locus in the Dactylis glomerata complex. Heredity 52:153–169 Lumaret R (1986) Doubled duplication of the structural gene for cytosolic phosphoglucose isomerase in the Dactylis glomerata L. polyploid complex. Mol Biol Evol 3:499–521 Lumaret R (1987) Differential degree in genetic divergence as a consequence of a long isolation in a diploid entity of Dactylis glomerata L. from the Guizhou region (China). Presentation to the 2nd symposium Paleoenvironment East Asia, Hong Kong, Jan 9–14,1987 Lumaret R (1988) Cytology, genetics and evolution in the genus Dactylis. Crit Rev Plant Sci 7:55–91 Lumaret R, Barrientos E (1990) Phylogenetic relationships and gene flow between sympatric diploid and tetraploid plants of Dactylis glomerata (Gramineae). Plant Syst Evol 169: 1615–6110 Lumaret R, Bowman CM, Dyer TA (1989) Autotetraploidy in Dactylis glomerata L.: further evidence from studies of chloroplast DNA variation. Theor Appl Genet 78:393–399 Lumaret R, Bretagnolle F, Maceira NO (1992) 2n gamete frequency and bilateral polyploidization in Dactylis glomerata. In: Mariana A, Tavoletti S (eds) Gametes with somatic chromosome number in the evolution and breeding of polyploid polysomic species: achievements and perspectives. Forage Plant Breeding Institute, Perugia, Italy, pp 15–21 Maceira NO, Jacquard P, Lumaret R (1993) Competition between diploid and derivative autotetraploid Dactylis glomerata L. from Galicia. Implications for the establishment of novel polyploid populations. New Phytol 124:321–328 Matzk F (1981) Successful crosses between Festuca arundinacea Schreb. and Dactylis glomerata L. Theor Appl Genet 60:119–122 Mı´ka V, Kohoutek A, Smrz J (1999) Dactylis polygama, a nonaggressive cocksfoot for grass/clover mixtures. Grassland ecology V. Proceedings of the 5th ecological conference, Banska´ Bystrica, Slovakia, 23–25 Nov 1999 Mizianty M (1991) Biosystematic studies on Dactylis (Poaceae). 2. Original research. 2.2. Cytological differentiation of the genus in Poland. Fragm Flor Geobot 36:301–320 Mizianty M (1997) Distribution of Dactylis glomerata subsp. slovenica (Poaceae) in Europe. Fragm Flor Geobot 42: 207–213

A.V. Stewart and N.W. Ellison Mizianty M, Cenci CA (1995) Dactylis glomerata L. subsp. slovenica (Dom.) Dom. (Gramineae), a new taxon to Italy. Webbia 50:45–50 Mousset C (1995) Les dactyles ou le genre Dactylis. In: Prosperi JM, Balfourier F, Guy P (eds) Ressources Ge´ne´tiques des Gramine´es Fourrage`res et a` Gazon. INRA-BRG, Paris, France, pp 28–52 Mousset C (2000) Rassemblement, utilisation et gestion des ressources ge´ne´tiques de dactyle a` l’INRA de Lusignan. Fourrages 162:121–139 Nakazumi H, Furuya M, Shimokouji H, Fujii H (1997) Wide hybridization between timothy (Phleum pratense L.) and orchardgrass (Dactylis glomerata L.). Bull Hokkaido Prefectural Agric Exp Stn (Jpn) 72:11–16 Oertel C, Fuchs J, Matzk F (1996) Successful hybridization between Lolium and Dactylis. Plant Breed 115:101–105 Ortiz S, Rodriguez-Oubin˜a J (1993) Dactylis glomerata subsp. izcoi, a new subspecies from Galicia NW Iberian peninsula. Ann Bot Fenn 30:305–311 Parker PF (1972) Studies in Dactylis II Natural variation, distribution, and systematics of the Dactylis smithii Link. Complex in Madeira and other Atlantic Islands. New Phytol 72:371–378 Parker PF, Borrill M (1968) Studies in Dactylis. 1 Fertility relationships in some diploid subspecies. New Phytol 67:649–662 Reeves G, Francis D, Davies MS, Rogers HJ, Hodkinson TR (1998) Genome size is negatively correlated with altitude in natural populations of Dactylis glomerata. Ann Bot 82(suppl A):99–105 Robins JG, Bushman BS, Jensen KB (2008) New genomic resources for orchardgrass. Proceedings of the 27th EUCARPIA symposium on improvement of fodder crops and amenity grasses, 19–23 Aug 2007, Copenhagen, Denmark, pp 202–203 Sahuquillo E, Lumaret R (1999) Chloroplast DNA variation in Dactylis glomerata L. taxa endemic to the Macaronesian islands. Mol Ecol 8:1797–1803 Scho¨nfelder P, Ludwig D (1996) Dactylis metlesicsii (Poaceae), eine neue Art der Gebirgsvegetation von Tenerife, Kanariische Inseln. Willdenowia 26:217–223 Soltis DE, Soltis PS (1993) Molecular data and the dynamic nature of polyploidy. Crit Rev Plant Sci 12:243–273 Speranza M, Cristofolini G (1986) The genus Dactylis in Italy. 1. The tetraploid entities. Webbia 39:379–396 Speranza M, Cristofolini G (1987) The genus Dactylis in Italy. 2. The diploid entities. Webbia 41:213–224 Stebbins GL (1971) Chromosomal evolution in higher plants. Edward Arnold, London, UK Stebbins GL, Zohary D (1959) Cytogenetic and evolutionary studies in the genus Dactylis. I Morphological, distribution and inter relationships of the diploid subspecies. Univ Calif Publ Bot 31:1–40 Stewart AV, Joachimiak A, Ellison N (2008) Genomic and geographic origins of timothy (Phleum sp.) based on ITS and chloroplast sequences. In: Yamada T, Spangenberg G (eds) Proceedings of the 5th international symposium on the molecular breeding of forage and turf, 1–6 July 2007, Sapporo, Japan, pp 71–81 Stuczynski M (1992) Estimation of suitability of the Ascherson’s cocksfoot (Dactylis aschersoniana Graebn.) for field cultivation. Plant Breed Acclim Seed Prod 36:7–42

5 Dactylis Tuna M, Teykin E, Buyukbaser A, Budak H (2007) Nuclear DNA variation in the grass genus Dactylis L. Poster presentation, 5th international symposium molecular breeding of forage and turf, 1–6 Jul 2007, Sapporo, Japan Tyler B, Borrill M (1983) The use of wild species in forage grass breeding. Genetika 15:387–396 Tzvelev NN (1983) Grasses of the Soviet Union. Amerind Publishing, New Delhi, India Van Dijk GE (1961) The inheritance of harsh leaves in tetraploid cocksfoot. Euphytica 13:305–313 Wei-Lin Hu W, Timothy DH (1971) Cytological studies of four diploid Dactylis subspecies, their hybrids and induced tetraploid hybrids. Crop Sci 11:203–207

87 Wetschnig W (1991) Karyotype morphology of some diploid subspecies of Dactylis glomerata L. (Poaceae). Phyton Horn 31:35–55 Williams E, Barclay PC (1972) Transmission of B-chromosomes in Dactylis. NZ J Bot 10:573–584 Yamane K, Yano K, Kawahara T (2006) Pattern and rate of indel evolution inferred from whole chloroplast intergenic regions in sugarcane, maize and rice. DNA Res 13:197–204 Zenkteler M, Nitzsche W (1984) Wide hybridization experiments in cereals. Theor Appl Genet 68:311–315 Zurawski G, Clegg MT, Brown HD (1984) The nature of nucleotide sequence divergence between barley and maize chloroplast DNA. Genetics 106:735–749

Chapter 6

Dichanthium Vishnu Bhat, C. Mahalakshmi, Shashi, Sunil Saran, and Soom Nath Raina

6.1 Introduction The genus Dichanthium Willemet belongs to the tribe Andropogoneae of the family Poaceae, having wide distribution across the tropics and the subtropics of the Old World. It is one of the major components of SehimaDichanthium and Dichanthium–Cenchrus–Lasiurus grass covers of India (Dabadghao and Shankarnarayan 1973). Various species of Dichanthium such as D. annulatum, D. sericeum, D. caricosum and D. aristatum are of economic importance because of their high palatability and nutritional value. D. aristatum is useful for stabilization of waterway and banks and for suppressing invasive weeds such as Phyla canescens. Most of the Dichanthium species are used as pastures for grazing and are suitable for silage and hay. Recent studies have shown that D. annulatum has the highest relative preference value than any other grasses of northern grasslands of Pakistan (Sultan et al. 2008). Dichanthium forms an agamic complex with Bothriochloa Kuntze and Capillipedium Stapf. (de Wet and Harlan 1968, 1970a). It has sexually reproducing diploid representatives and apomictic polyploids with pseudogamous type of apomixis (Brown and Emery 1957; Harlan et al. 1958; Harlan and de Wet 1963a; de Wet and Harlan 1970a; D’Cruz and Reddy 1971). Around 20 species have been reported in this genus from tropical to the subtropical region. The basic chromosome number is 10 and different ploidy levels

V. Bhat (*) Department of Botany, University of Delhi, Delhi 110 007, India e-mail: [email protected]

have been reported in the D. annulatum complex (Mehra 1961). Diploid species such as D. armatum, D. maccannii, and D. panchaganiense are endemic to India (Harlan 1963), whereas other diploid species such as D. humilis, D. setosum, D. sericeum, and D. superciliatum are of Australian origin but widely distributed (de Wet and Harlan 1962; Fig. 6.1). The tetraploid agamospecies such as D. annulatum, D. aristatum, and D. caricosum have diploids, which are sexually reproducing and are well adapted polyhaploids (de Wet 1968). D. caricosum produce hybrids with both D. annulatum and D. aristatum naturally because of which these species are grouped as one large agamospecies. Development of efficient genetic improvement methods requires comprehensive analysis of various information on the taxonomy, cytology, reproductive biology, and molecular genetics of cultivated Dichanthium species and their wild relatives. The information provided in this chapter can be utilized for systematic exploitation of the gene pool for commercial breeding.

6.2 Taxonomy and Distribution Dichanthium first described by Willemet (1976) belongs to the family Poaceae, Subfamily Panicoideae, Group Andropogonodae, Tribe Andropogoneae. The genus Dichanthium Willemet is closely associated with the genera Bothriochloa Kuntze and Capillipedium Stapf. because of which they were all included under Andropogon Linn. by Hackel (1889), forming Dichanthium–Bothriochloa–Capillipedium agamic complex (de Wet and Harlan 1968, 1970a;

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_6, # Springer-Verlag Berlin Heidelberg 2011

89

90

V. Bhat et al.

Fig. 6.1 Diagram showing distribution of four important species, namely D. annulatum (a), D. aristatum (b), D. caricosum (c), and D. sericeum (d), major forage crop species of Dichanthium across the world [From: Cook et al. (2005)]. Leg-

end shows the suitable areas (dark red), i.e., the locations where a species is highly adapted and marginal areas (orange), where the species occurrence is uncertain but can survive in that location

Clayton 1977). It is sometimes very difficult to treat agamic complex taxonomically as every adaptive genotype is maintained as a distinct morphological unit (Baker 1959; Lo¨ve 1960; Valentine 1960). The hybrids and their derivatives are usually included with a species that bridges the gap between the members of agamic complex. Such a species is termed as “compilospecies” (Harlan and de Wet 1963b). In case of Dichanthium–Bothriochloa–Capillipedium, de Wet and Harlan (1970b) suggested Bothriochloa intermedia as a compilospecies. For example, hybrids of B. intermedia and either Capillipedium parviflorum or D. annulatum behave as apomictic allotetraploids and

are recognized taxonomically as B. glabra and B. grahamii, respectively (de Wet and Stalker 1974). Majority of the species of Dichanthium are perennial, herbaceous, branched or simple, erect, sometimes decumbent or geniculate; rhizomatous or stoloniferous, tufted grasses, forming upright tussocks, sometimes with extensive creeping stolons; leaves usually glaucous, sometimes aromatic; hairy or bearded culm nodes; auricles absent; ligule membranous and usually ciliate or fringed; plants bisexual; inflorescence terminal or also axillary, solitary or compound racemes digitately or subdigitately arranged at the summit of the culm; homogamous pairs and obtuse, sessile spikelets, lower

6 Dichanthium

or basal spikelets either male or sterile; sessile spikelets: bisexual and dorsally compressed or imbricate, callus obtuse; fragile rachis, 2 glumes more or less equal or subequal, lower glume boat-shaped and chartaceous to cartilaginous, upper lemma entire and awned, glabrous awn bent about the middle; pedicellate spikelets: male or sterile and unawned, palea absent or present, 2 lodicules free and fleshy, 1–3 stamens, ovary glabrous, 2 stigmas; and fruits compressed, ornamental, and attractive (Jacobsen 1981; Shouliang and Phillips 2006; Quattrocchi 2008; http://www.efloras.org/florataxon.aspx?flora_id¼2& taxon_id¼109887). The diversity in the type of spikelets and the ornamentations of the lower glume of various Dichanthium species has been shown in Fig. 6.2. The taxonomic features of various species based on Quattrocchi (2008) are described as follows: 1. D. affine (R. Br.) A. Camus, commonly known as bluegrass, is native to Australia with its distribution in southern and western Australia,

Fig. 6.2 Different types of inflorescences of few Dichanthium species. (a–d) D. annulatum; (e) D. aristatum; (f) D. armatum; (g–i) D. caricosum; (j–k) D. foveolatum; (l) D. oliganthum;

91

Queensland, Northern Territory, and New South Wales and was introduced to Pakistan as a fodder grass (eFloras). It is a perennial, erect, and geniculate grass with tall and slender clumps, bluish green leaves, nodes hairy or bearded, racemes silky greenish and purplish, very similar to D. sericeum (R. Br.) A. Camus. It is distinguished from D. annulatum by its sessile or subsessile racemes and a subapical ciliate fringe on the lower glume of the sessile spikelet. 2. D. annulatum (Forssk.) Stapf, commonly known as Marvel grass, is native to Southeast Asia, India, China, and tropical Africa and also was introduced to Australia and America. It has been introduced in Texas, South and Southwest America, where it persists as casual weed (Gould 1967). It is very well adapted to most textures of soil, i.e. from coralline sands to heavy black clay soil preferably with neutral to alkaline pH. They show moderate to poor frost tolerance but moderate intolerance to shade (Cook et al. 2005). Plant has twisting and

(m–n) D. panchaganiense and (o) D. sericeum (Courtesy: Dr. Anil Kumar and Prof. S.R.Yadav)

92

ascending to semi-decumbent to prostrate, creeping habits; paniculate inflorescence of compound and several racemes subdigitate and short and glabrous peduncle, racemes densely villous, spikelets subimbricate or closely overlapping, sessile spikelet bisexual, pedicelled spikelets male or neuter, lower floret barren, upper floret hermaphrodite, pedicellate spikelet without a pit, sessile spikelets narrowly oblong and awned, lower glume obtuse or truncate, keel not winged median nerve present, sheath terete, larger ligule, upper lemma awned, 3 stamens, ovary glabrous (Bor 1960; Celarier et al. 1962; Shouliang and Phillips 2006; http://www. efloras.org/florataxon.aspx?flora_id¼2&taxon_id¼ 200025242). Figure 6.3 represents a diagrammatic sketch of D. annulatum showing different taxonomic characters. (a) D. annulatum (Forssk.) Stapf var. annulatum is found mostly in tropical Africa, Southwest Asia, India, and Indonesia. Racemes subdigitate, sessile spikelet with lower glume pubescent to villous and not pitted. (b) D. annulatum (Forssk.) Stapf var. papillosum (Hochst. Ex A. Rich.) de Wet and J.R. Harlan is mostly distributed in southern Africa, Ethiopia, Tanzania, and Kenya. Rhizomatous, ring of hairs around the nodes, leaf blade flattened, leaf sheath glabrous, spikelets paired, sessile spikelet with lower glume not pitted villous and fringed, pedicelled spikelet awnless (Mehra 1964a). 3. D. aristatum (Poir.) C.E. Hubb., commonly known as Angleton grass, is indigenous from tropical India to Indonesia. It has been introduced to many tropical regions and has been naturalized in Australia (Bisset and Sillar 1984), America (Gould 1967), South Africa (Bor 1960; Celarier et al. 1962), and Taiwan and Yunnan region of China (Shouliang and Phillips 2006; http://www.efloras.org/florataxon.aspx?flora_id¼2&taxon_id¼200025243). It is mainly adapted to dark or red soils with neutral to alkaline pH, moderately tolerant to drought, shade, and grazing, poor salt tolerance, and moderate shade tolerance (Cook et al. 2005). It is weakly tufted, characterized by hairy peduncle, culm nodes bearded by very short spreading hairs or glabrous when old, compound racemes sparsely villous, spikelets imbricate and paired, sessile

V. Bhat et al.

4.

5.

6.

7.

spikelets elliptic to obovate, pitless, lower glume pilose and fringed. D. armatum (Hook.f.) Blatter and McCann has its roots of origin in India and is endemic to Western Ghats, Maharashtra (Bor 1960). It is an annual species with characteristic lower glume of pedicelled spikelet armed with marginal bulbousbased bristles and not pitted, lower glume of sessile spikelet pitted (Bor 1960; de Wet and Harlan 1961). D. caricosum (L.) A. Camus Hochst. ex Miq., commonly known as Nadi bluegrass, was first collected at Suva, Fiji Islands in 1936 (Bisset and Sillar 1984). It shows wide distribution in India, Burma, and Ceylon, extending to southern China, Indonesia, and Malaya (Bor 1960; Celarier et al. 1962), where it grows on slightly acidic to alkaline soil with heavy black clay soil texture, intolerant to frost and salinity and moderately tolerant to shade (Cook et al. 2005). It forms small clumps, prostrate habit, peduncle of raceme glabrous, sessile spikelets hermaphrodite, lower floret empty, upper floret bisexual, pedicelled spikelets male, lower glume of the sessile spikelet obovate or oblongtruncated, without a median nerve, sheath compressed, ligule short ciliate membrane, and margins with long and short hairs interspersed, 3 stamens, ovary glabrous (Bor 1960; Celarier et al. 1962; Clayton and Renvoize 1982; Shouliang and Phillips 2006; http://www. efloras.org/florataxon.aspx?flora_id¼2&taxon_id¼ 242317870). Figure 6.4 shows the diagrammatic representation of various taxonomic features of this species. D. compressum (Hook.f.) Jain and Deshpande is a species endemic to India, especially to Western Ghats in Maharashtra, Lonavala, and Khandala, Poona District (Jain and Deshpande 1978). It is a rare and threatened species, with leaf blades linear-lanceolate and finely acuminate, aromatic, and lower glume of sessile spikelets hairy. D. fecundum S.T. Blake, commonly called as Gulf bluegrass, has Australian origin with its distribution in northeastern (Queensland) and western Australia (Jacobsen 1981). It has erect or geniculate habit, forming erect tussocks, glaucous leaves, green or purplish racemes solitary or in group.

6 Dichanthium

Fig. 6.3 Dichanthium annulatum Stapf. (a) Part of the plant with inflorescence; (b) Sessile and Pedicellate spikelet; (c–h) Sessile spikelet: (c) Lower glume; (d) Upper glume; (e) Lower lemma; (f) Upper lemma; (g) Stamens; (h) Pistil and lodicules;

93

(i–l) Pedicellate spikelet: (i) Pedicellate spikelet; (j) Lower glume; (k) Upper glume; (l) Lemma; (m) Collar (Courtesy: Dr. Anil Kumar and Prof. S.R. Yadav)

94

Fig. 6.4 Dichanthium caricosum (L.) A. Camus (a) Habit; (b) Sessile and Pedicellate spikelet; (c–i) Sessile spikelet: (c) Lower glume; (d) Upper glume; (e) Lower lemma; (f) Upper lemma; (g) Palea; (h) Stamens; (i) Pistil and lodicules; (j–o) Pedicellate

V. Bhat et al.

spikelet: (j) Lower glume; (k) Upper glume; (l) Lower Lemma; (m) Upper Lemma; (n) Palea; (o) Stamens; (p) Collar (Courtesy: Dr. Anil Kumar and Prof. S.R. Yadav)

6 Dichanthium

8. D. foveolatum (Delile) Roberty, commonly known as Gandhel is a native species of Africa and tropical and temperate Asia including India, Ceylon, and Pakistan. It is a wiry, very slender, and much-branched plant, with spathiform leaf sheaths, basal sheaths silky, ligule membrane-like, leaves narrow almost filiform, terminal and axillary racemes solitary and long-exserted, awn on one side of spike, lower glume of sessile spikelet pitted, pedicelled spikelet pitted or not, and upper lemma sometimes bidentate (Clayton and Renvoize 1982; GRIN, eFloras). 9. D. humilius (Benth.) J.M. Black is widely distributed in Central Australia. The plant is characterized by sessile or subsessile raceme, hairs on lower glume of sessile spikelets not strictly arranged in a subapical transverse fringe, racemes 2–7 per inflorescence (de Wet and Harlan 1962; GRIN). 10. D. maccannii Blatter is a vulnerable and threatened species, endemic to India, especially to Panchgani Plateau in Western Ghats, Maharashtra. It has racemes solitary or two and characteristic not pitted lower glume of sessile spikelet and marginal bulbous-based bristles on lower glume of pedicellate spikelet (Bor 1960; de Wet and Harlan 1961; Naik 1982). 11. D. mucronulatum Jansen has its distribution in the tropical Asia mainly Indo-China and Malaysia. It is a rare species, ligule eciliate membrane, hairy awn, upper lemma sometimes bidentate (Kew). 12. D. oliganthum (Hochst. ex Steud.) Cope is mainly found in India. It is an annual species (Kew). 13. D. pallidum (Hook.f.) Stapf ex C.E. is reported to be endemic to India, especially Nilgiris, Tamil Nadu, by Bor (1960), and its occurrence was reported in other parts of India namely, Mirzapur, 24 Parganas in West Bengal and Nellore by Bhattacharyya and Uniyal (1973). It is characterized by its small creeping nature, solitary racemes of tightly overlapped spikelets, peduncle glabrous and lower glume of sessile spikelet with shallow longitudinal furrow (GRIN; Bor 1960; Celarier et al. 1962). 14. D. panchaganiense Blatter and McCann is a species endemic to India particularly to Panchgani Plateau, Maharashtra. It is a rare and threatened, annual species, with racemes solitary or digitate, lower glume of pedicelled spikelet always pitted

95

15.

16.

17.

18.

19.

and with marginal bulbous-based bristles (Bor 1960; de Wet and Harlan 1961). D. queenslandicum B. Simon, commonly known as king bluegrass, is a native of Queensland, Australia. It is a vulnerable species (Kew). D. sericeum (R. Br.) A. Camus, called as Queensland bluegrass, has a wide distribution in tropical Asia, Papua New Guinea, western Australia, Queensland, Northern Territory, Victoria, and New South Wales and has been introduced to US (Gould 1967). It is usually prevalent in neutral to alkaline soil conditions and very well adapted to heavy black clay soil moderate to poor tolerance to shade and grazing but intolerant to shade (Cook et al. 2005). It is identifiable with its tussocky nature, bluish green and purplish, culm nodes bearded or with long white hairs, racemes 2–7 with white and bluish silky hairs, pedicels usually bilaterally ciliate, spikelets paired and crowded, lowest pairs of spikelets sterile and reduced to glumes, callus shortly bearded, pedicellate spikelets sterile, hairs on lower glume of sessile spikelet arranged in distinct subapical transverse fringe (de Wet and Harlan 1962, Kew). (a) D. sericeum (R. Br.) A. Camus subsp. humilius New South Wales. (b) D. sericeum (R. Br.) A. Camus subsp. polystachyum Australia, Asia. (c) D. sericeum (R. Br.) A. Camus subsp. sericeum New South Wales. D. setosum S.T. Blake is an Australian species, mainly distributed in Queensland, New South Wales, and western Australia. It differs from other species by the presence of few, never strictly digitate racemes, awned pedicellate spikelet, hairs on the lower glumes arranged in a distinct subapical fringe, lower glume of sessile spikelet 5–6 nerved (de Wet and Harlan 1961, 1962; Jacobsen 1981). D. tenue (R. Br.) A. Camus, commonly called as small bluestem, is a species from Australia now confined to Malaya, Polynesia, Queensland, and New South Wales. It has prostrate habit, culm nodes essentially glabrous, glabrous peduncle of raceme, racemes 1–3, lower glume of sessile spikelet with short hairy margins (Celarier et al. 1962). D. tenuiculum (Steudel) S.T. Blake is an Australian species, commonly known as tassel bluegrass. It is widely spread around western Australia,

96

Queensland, Northern Territory, Philippines, Timor and New Guinea. It is a robust grass, leaves bluish green to blue and very scabrous, culms robust and branched, racemes in a terminal and digitate or tasseled inflorescence (Jacobsen 1981). 20. D. woodrowii (Hook.f.) S.K. Jain and Deshpande is a rare grass endemic to India, confined to Mawal and Paud of Pune District, Maharashtra. It is a vulnerable and threatened species with aromatic, tufted, woody rootstock, sessile spikelets oblong-lanceolate (Jain and Deshpande 1978; Mehrotra 1979).

6.3 Cytology and Hybridization Cytogenetic diversity and chromosomal behavior play an important component of any breeding program. The occurrence of natural interspecific and intergeneric hybrids has been reported in Gramineae, and a thorough study of these hybrids enrich our understanding on the fields of genetics, cytogenetics, biosystematics, and phylogenetics. Many workers including Myers (1947) and Carnahan and Hill (1961) have reported many interspecific and intergeneric hybrids, especially in forage grasses. Stebbins (1947) and Lo¨ve (1964) have proposed that natural polyploids, which arise usually, fall between autopolyploids and allopolyploids. Polyploidy played an important role in the evolution of species, especially the grasses. Allopolyploids played significant role in speciation than autopolyploids.

6.3.1 Cytological Characters Dichanthium, like many other members of agamic complex, has autotetraploidy along with haploidy as a mechanism to bridge gap and ensure contact with sexuality (de Wet 1965, 1968). Mehra (1961), de Wet and Harlan (1968) and many others have reported chromosome numbers in the sexually reproducing and agamospecies of Dichanthium. The basic number for the tribe Andropogoneae is 5 or 10 and Dichanthium species show multiples of 10 in sexual and apomictic types (Celarier 1956, 1957; Celarier and Harlan 1957). The sexually reproducing diploid species of Dichanthium, namely D. sericeum, D. affine, D. setosum, D. tenue,

V. Bhat et al.

D. acutiusculum, D. humilius, D. panchganiense, D. maccannii, and D. tenuiculum, and diploid races of D. annulatum, D. aristatum, and D. caricosum have chromosome number 2n ¼ 20. The tetraploid races of facultative apomicts such as Dichanthium annulatum, D. aristatum, D. caricosum, and obligate apomict D. fecundum have 2n ¼ 40, whereas the hexaploid obligate apomict D. papillosum and hexaploid race of D. aristatum have 2n ¼ 60. D. armatum is cytologically unknown (Oke 1950; Gould 1956; Celarier and Harlan 1957; Bor 1960; Bogdan 1977). Morphological data have suggested hybridization and chromosome doubling as the major phenomena in the evolution of Dichanthium Willemet. The meiotic behavior is regular in case of diploids, whereas polyploids show irregularities, which suggest meiotic irregularities in addition to morphological variations in these species. de Wet and Borgaonkar (1963) found that naturally occurring hybrids were euploids and facultative apomicts, but artificial hybrids were either euploid or aneuploid, and very few were strictly sexual. Depending on these observations, it was proposed that apomixis is a dominant trait over sexuality and also that euploids have selective advantage over aneuploids in nature (de Wet and Borgaonkar 1963; Harlan et al. 1964). By cytoecological studies, it was hypothesized that the tetraploids have arisen from doubling of the related diploid species, which, over a period of time, evolved and differentiated cytologically, whereas hexaploids originated through hybridization of tetraploids and diploids. These theories have been postulated by Celarier and his coworkers in 1958, especially in case of D. annulatum. In D. annulatum, four different morphological types can be distinguished, viz. Tropical, Mediterranean, Senegal, and South African types (Harlan et al. 1958, 1961; Mehra 1964b). Tropical ones include prostrate or decumbent, diploid as well as tetraploid types. The Mediterranean types are generally erect with longer inflorescences, tetraploid, and more adapted to desert conditions. The Senegal type, represented by single collection from northwest Africa, is considered to be extreme form of Mediterranean type. The erect-growing South African types are all hexaploids and obligate apomicts (Mehra and Celarier 1958; Mehra 1962, 1964b; de Wet 1968). Morphological and cytological studies of these races and their hybrids show that they are not genetically isolated. It was, therefore, suggested to group three

6 Dichanthium

races, namely the Tropical, Mediterranean, and Senegal types under D. annulatum var. annulatum.

6.3.2 Ploidy Cycles in Dichanthium Various embryological, cytogenetical, and hybridization studies indicated that diploid species of all the three genera of the agamic complex reproduce sexually, while the polyploids (2x, 4x, 6x) are facultative apomicts, although there are a few exceptions (de Wet and Borgaonkar 1963; de Wet 1965, 1968). Diploid Dichanthium spp. are fully sexual and fertile, but occasionally produce autotetraploids as a result of fertilization of the unreduced male and female gametes. The present day diploid races of tetraploid agamospecies of D. annulatum, D. aristatum, and D. caricosum have been found to represent well adapted polyhaploids as they resemble few sexually reproducing artificial polyhaploids (de Wet and Singh 1964; de Wet 1965; de Wet and Harlan 1970a; D’Cruz and Reddy 1971). Polyhaploidy is rare in nature as they are usually completely sterile and poorly adapted. But a few polyhaploids produced from natural facultative apomicts or from natural or artificial hybrids between D. caricosum and D. annulatum, or D. aristatum resemble natural diploids at morphological and chromosomal level (de Wet 1965). Haploids obtained from Mediterranean ecotype (Mehra 1964b) of D. annulatum do not survive and are completely sterile, whereas those from Tropical ecotypes were found to be vigorous, reproduce sexually but sterile, indicating that apomixis is not functional in haploids. Presence of such fertile, sexually reproducing haploids have phylogenetic significance as it helps in maintaining contact with sexuality in predominantly polyploid agamospecies like Dichanthium (de Wet 1968; de Wet and Harlan 1970a). Triploids derived from hybridization within species of sexually reproducing diploid races of tetraploid D. annulatum, D. aristatum, and D. caricosum and between sexual tetraploid D. annulatum and other diploids or from fertilization of diploid female gametes with haploid male gametes have been reported. These triploids were all sterile but quite vigorous and cytologically irregular (de Wet and Singh 1964). Hexaploid D. papillosum, which has hybrid origin from two tetraploid D. annulatum (Mehra 1962), is an obligate apomict.

97

6.3.3 Hybridization Crossing experiments in Dichanthium were started earlier by many workers in order to study the phenomenon of apomixis in this genera. Several combinations of crossings between different species of Dichanthium, Capillipedium, and Bothriochloa and analyses of different characters of the progenies from such crosses revealed that they were morphologically same as maternal types, which were due to apomictic mode of reproduction and not due to maternal inheritance. Celarier and Harlan (1957) were able to establish this when they crossed among diploid and polyploid species of D. annulatum (2n ¼ 20, 40, 60) and D. caricosum (2n ¼ 40); D. annulatum (2n ¼ 40) with diploid and polyploid Bothriochloa sp. [B. intermedia (2n ¼ 40, 60), B. ischaemum (2n ¼ 50, 60), B.venusta (2n ¼ 40)] and Capillipedium sp. [C. parviflorum (2n ¼ 40), C. spicigerum (2n ¼ 40)]. Cytologically reduced or unreduced female gametes can function sexually or develop parthenogenetically in this genera (Harlan and de Wet 1963a). Introgressive hybridization between different species of Dichanthium, for example, D. annulatum with other species such as D. aristatum, D. caricosum, D. papillosum, and D. fecundum (Celarier et al. 1961; Chheda and Harlan 1963; Borgaonkar and de Wet 1964; Mehra 1966) or within different ecotypes as available in case of D. annulatum (Mehra and Celarier 1958; Mehra 1962), or even across genera, for example, Bothriochloa and Capillipedium (Harlan et al. 1958; Harlan 1963; de Wet and Harlan 1968, 1970a) provide a wide platform of variability in the genus Dichanthium Willemet. Studies in this direction were carried out by Harlan et al. (1964), wherein crosses between Bothriochloa grahmii and D. annulatum yielded hybrids, which were apomictic when both parents were apomicts and sexual when both were sexual. Similar results were also obtained by Harlan et al. (1961) and Singh (1968) when they crossed tetraploid D. annulatum with tetraploid B. grahmii; and diploid and tetraploid D. caricosum, respectively. In crosses between apomictic  sexual and vice versa, hybrids were mostly apomicts, but frequency of sexuals was high compared to apomict  apomict crosses. Harlan and de Wet (1975) were even able to obtain sexual types by crossing facultative apomicts.

98

Hybrids with different ploidy levels were obtained in crosses between facultative apomictic tetraploid D. annulatum and diploid sexuals of D. aristatum and D. caricosum, and their progenies were categorized as sexual and apomictic types (D’Cruz and Reddy 1971). In another study, D. caricosum was found to resemble polyhaploids from F1 hybrids of D. caricosum and D. annulatum, even though diploid D. annulatum and D. caricosum are reproductively isolated. D. aristatum was also indistinguishable from polyhaploid derived from F1 hybrids of D. aristatum and D. caricosum. Thus, D. caricosum acts as a genetic bridge between D. annulatum and D. caricosum at tetraploid level. In a tetraploid cross, when an unreduced female egg cell is fertilized by a reduced male gamete, the resultant progeny is a hexaploid. Best example of this being D. papillosum, which is the result of fertilization of unreduced egg cell of Tropical type with reduced male gamete of Mediterranean type of D. annulatum (Borgaonkar and Singh 1962; Mehra 1964b). It was, therefore, suggested to classify D. papillosum as D. annulatum var. papillosum (Pilger 1954). Similarly, tetraploid D. fecundum show morphological and cytological characters same as D. annulatum and is thus classified as D. annulatum var. fecundum (Hackel 1889). These three species together form an interrelated agamospecies, i.e., the D. annulatum complex (Singh and Mehra 1965). Celarier et al. (1961, 1962) also studied morphological and cytological characteristics of the hybrids between D. tenue (2n ¼ 40) and D. aristatum (2n ¼ 40), which differ in their morphology by the presence of glabrous peduncles and prostrate habit in the former. Three of the hybrids with 2n ¼ 40 chromosomes were decumbent with wooly peduncles, whereas two of them had 2n ¼ 60, which resembled D. tenue when it was used as female parent. The hexaploid was found to be the result of cross between cytologically unreduced female gamete from D. tenue and reduced male gamete from D. aristatum. The reciprocal crosses also yielded a hexaploid with the unreduced gamete coming from D. aristatum and resembled D. aristatum morphologically. Hybrids from such crosses also yielded morphologically intermediate types with semi-decumbent habit and slightly pubescent peduncles. These morphological variations in the hybrids indicate hybridization with limited genetic segregation in facultative apomicts. Celarier

V. Bhat et al.

et al. (1962) along with previous works by de Wet (1954) and Celarier et al. (1961) proposed that species such as D. aristatum, D. caricosum, D. pallidum, and D. tenue form an interrelated agamic complex wherein cytologically reduced and unreduced female gametes function sexually or parthenogenetically to form a series of hybrids with chromosome numbers ranging from 2n ¼ 20 to 2n ¼ 60. The natural tetraploids of D. annulatum are regarded as segmental allopolyploids (Stebbins 1947) and probably originated from hybridization between closely related species or partially isolated races of same species. de Wet et al. (1961) studied the chromosomal association in hybrids of Dichanthium. Hybridizations between two plants of D. annulatum with chromosome complement AAA1A1 and A2A2A3A3 result into hybrids (AA1A2A3), which associate cytologically into 20 bivalents, sometimes as trivalents or tetravalents. Hexaploids have AAA1A1A2A3, thus bringing together whole genome of one parent with half of the other parent. Hexaploid hybrids from D. aristatum (DDD1D1) and D. caricosum (CCC1C1) have chromosome compliment as CCC1C1DD1 and DDD1D1CC1 and behave cytologically as D. annulatum. Hybrids (AA1FF1) between D. annulatum and D. fecundum (FFF1F1) behave like one of the parents. Therefore, de Wet et al. (1961) were able to prove that the hybrids show autosyndetic pairing of the chromosomes in Dichanthium Willemet and also that chromosome pairing can take place within basic genome (n ¼ 10), indicating the original basic number as 5, as earlier suggested by Celarier et al. (1960).

6.4 Embryology The embryological studies and apomixis in Dichanthium have been conducted in detail by many workers (Brown and Emery 1957; Celarier and Harlan 1957; Reddy and D’Cruz 1969a) using microtomy and ovule squash method. Microsporogenesis was found to be slightly irregular with the formation of a few univalents and multivalents at metaphase I and bridges and laggards at anaphase I. Second meiosis was comparatively more regular and forms 3-nucleate pollen grain with thick wall and single germ pore. The female gametophyte development initiates with the differentiation of megaspore mother cell from the

6 Dichanthium

nucellus, which later degenerates (Bhanwra 1988). One or more cells of the nucellus show enlargement and vacuolation and form the aposporous embryo sacs. D. annulatum, being a facultative apomict, has sexual and apomictic embryo sac occurring together in the same ovule or different ovules of same inflorescence. The frequency of apomictic embryo sacs are generally higher than the sexual ones (Brooks 1958). The embryo sacs can be of sexual, reduced type with 8-nuclei or aposporous, unreduced with 4-nuclei. Apart from genotype, environment also plays a role in regulating the type of embryo sac development. According to Heslop-Harrison (1961), photo- and thermo-periods have considerable role in embryo sac development. In D. aristatum, photoperiodic conditions during inflorescence development can be correlated to apomixis and pollen fertility. Knox (1967) correlated the presence of apomictic embryo sacs and day-length in six localities in Australia, which was earlier shown by Knox and Heslop-Harrison (1963) under controlled conditions. The pollen fertility was also found to be affected under short day conditions, which favor apomictic embryo sac production (Knox and Heslop-Harrison 1966). Similarly, in case of D. annulatum, the frequencies of sexual and apomictic embryo sacs fluctuate with season (Gupta et al. 1969). Saran and his coworkers made worldwide collections of different Dichanthium species and studied the variation in these species. They found that all the species were uniform in their characters when collected along the equatorial region, whereas the species showed differences when moved away from equator. They reasoned the kind of uniformity shown among the species along the equator due to the photoperiod, i.e., equal day-length. In this way, the breeders can induce the type of embryo sac development by manipulating the environmental conditions. Experimentally, the shift from sexual to apomictic embryo sac has been tried. For example, Asker (1966) was able to increase the sexual embryo sacs in facultative apomict Potentilla. Apart from the 8-nucleate and 4-nucleate embryo sac distinction for detecting sexuality and apomixis, respectively, Saran and de Wet (1969) found a peculiar opening, which was consistently found in 8-nucleate embryo sacs of D. intermedium [equivalent to Bothriochloa bladhii (Retz.) S.T. Blake according to Quattrocchi 2008] and never in the aposporous embryo sacs. The opening was the result of unequal thickening in the wall layers of the embryo sac. The

99

function of the opening was to provide the least resistance to the growing pollen tube and therefore enhance the chances of fertilization. The presence of such an opening in the facultative apomicts favors the chances of sexuality. Harlan et al. (1964) extensively studied the embryology of six accessions representing different ploidy levels and four species belonging to two genera, viz. Bothriochloa and Dichanthium. They also studied three diploid species of Bothriochloa and 15 collections representing four diploid species of Dichanthium. They found sexuality at the diploid level and apomixis at higher levels and concluded sexuality and asexuality to be dependent on synchronization of different embryological events. With respect to fertilization and post-fertilization events, in case of 4-nucleate embryo sac, only one of the male gametes fuses with the polar nuclei and the other one either enters egg or remain outside. In most of the ovules, single aposporous embryo sac is functional and therefore embryo and endosperm development was observed in only one embryo sac. In case of apomicts, there was precocious development of embryos in the ovules, and embryo formation was observed even before anthesis. Reddy and D’Cruz (1969a) also described the apomictic mode of development of embryo by various mechanisms for elimination of syngamy mechanisms are gamtes either getting degenerated inside the egg cell or fail to enter it or are restricted to the suspensor cell. Other mechanisms include fertilization of both polar nuclei by twomale gametes or development of adventitive embryos from nucellus. Polyembryony in Dichanthium has also been reported in the apomictic species, D. annulatum. The development of twin embryos occurs in two ways. Either the two embryos develop independently in two different embryo sacs or in the same embryo sac. The former one is more prevalent than the latter. In the latter case, one of the embryos enlarges and pushes the other, which remains as a rudimentary structure (Reddy and D’Cruz 1969b).

6.5 Nature and Inheritance of Apomixis in Dichanthium Apomixis is generally defined as asexual reproduction through seeds (apo ¼ away from, and mixis ¼ act of mixing; Asker and Jerling 1992). The increase in

100

number of chromosomes due to polyploidy many a times induces meiotic abnormalities, resulting in sterility. To overcome sterility, species have evolved an alternate mode of reproduction, namely apomixis. The phenomenon of apomixis in Poaceae is quite frequent as reported by Brown and Emery (1957, 1958) and Pullaiah and Febulaus (2000). Apomixis in Dichanthium sp. along with its two closely related members of the tribe Andropogoneae, viz. Bothriochloa O. Kuntze and Capillipedium Stapf., was first studied by Celarier and Harlan (1957). Using hybridization and cytological studies, they were able to establish apomictic reproduction in 12 species of these three genera, excluding the possibilities of maternal inheritance and selffertilization (Celarier and Harlan 1957; Celarier et al. 1958). The type of apomixis in Dichanthium Willemet is of aposporous pseudogamous type (Gustafsson 1947a, b; Brown and Emery 1957; Celarier and Harlan 1957; Harlan et al. 1958; de Wet and Harlan 1970a; D’Cruz and Reddy 1971). In aposporous type of apomixis, all the four megaspores developed from megaspore mother cell degenerate and functional embryo sac is formed from another nucellar cell (called aposporous initial), which is unreduced; pseudogamy is the development of endosperm after fertilization of male gamete with the polar nuclei or the secondary nucleus. Apomixis was reported to be dominant over sexuality and two pairs of genes were found to be involved at tetraploid level, which are heterozygous in case of apomicts and homozygous in sexuals. The tetraploid apomicts were assigned the genotype AaAa and sexual ones aaaa, in the sense that the “A” allele induces formation of 4-nucleate apomictic embryo sac and “a” allele does not (Harlan et al. 1964; de Wet and Harlan 1970a). D’Cruz and Reddy (1971) also concluded sexuality and apomixis to be independent phenomena, associated with ploidy but found its inheritance as complex compared to earlier report by Harlan et al. (1964). They, like many other workers (Harlan et al. 1958 in Dichanthium sp., Chheda et al. 1961 and Borgaonkar and de Wet 1961 in Bothriochloa sp.), also emphasized on delicate nature of genetic balance required for the expression of apomixis in Dichanthium spp. due to sexual reversion of hybrids in F2 generation. Therefore, agamospecies, which were once considered to be genetic dead ends, are now characterized by dynamic genetic system, several mechanisms to facilitate recombination, and capable of high degree of adaptive polymorphism and progressive evolution (de Wet and Harlan 1970a).

V. Bhat et al.

6.5.1 Bothriochloa–Dichanthium– Capillipedium Agamic Complex Grass species have a typical agamic complex, which consists of several basic diploid sexuals and many apomictic polyploids (Harlan and de Wet 1963a, b). The genus Dichanthium forms an interrelated agamic complex (Celarier and Harlan 1957; de Wet and Harlan 1970a; Saran and de Wet 1970; de Wet and Stalker 1974) with two other genera of the tribe Andropogoneae, viz. Bothriochloa and Capillipedium (Fig. 6.5). Gene flow within this agamic complex occurs in different directions but sometimes there are a few incompatibility barriers. Srivastava and Purnima (1990) studied seven polyploid taxa involving Bothriochloa–Dichanthium complex and Heteropogon contortus and Paspalum paspaloides to establish different aspects of apomixis in these species. Species of the genus Bothriochloa, especially B. intermedia (R.Br.) A. Camus [equivalent to B. bladhii (Retz.) S.T. Blake according to Quattrocchi 2008], acts as a bridging species for crosses between Dichanthium and Capillipedium spp. (de Wet and Harlan 1970b; Berthaud 2001). B. intermedia (R.Br.) A. Camus introgresses with B. ewartiana (Domin) C.E. Hubbard, B. ischaemum (Linn.) Keng, Capillipedium parvifolium (R.Br.) Stapf, and D. annulatum (Forssk.) Stapf. Therefore, it was suggested that B. intermedia acts as a compilospecies and is carrying genes from several other species of Bothriochloa O. Kuntze, as well as species of Capillipedium Stapf. and Dichanthium Willemet (Harlan and de Wet 1963a, b; de Wet and Harlan 1966, 1970a, b). It was also suggested by de Wet and Harlan (1968) to combine all the three taxa under Dichanthium genera.

6.6 Breeding and Genomics 6.6.1 Germplasm Collections The online software project namely Germplasm Resources Information Network (or GRIN) of the National Genetic Resources Program of USDA provides germplasm information about plants, animals, microbes, and invertebrates important for food and agricultural production. This project lists different

6 Dichanthium

101

DICHANTHIUM

CAPILLIPEDIUM

parviflorum 2x, 4x

caricosum 2x, 4x

papillosum 6x

aristatum 2x, 4x

annulatum 2x, 4x

assimile 2x, 4x

grahamii 4x, 6x

glabra 4x, 6x

ewartiana 5x, 6x

intermedia 4x

ischaemum 4x, 6x

concanensis 5x

insculpta 5x, 6x

woodrowii 5x

kuntzeana 2x

pertusa 4x

radicans 4x

compressa 2x

BOTHRIOCHLOA

Fig. 6.5 Diagrammatic representation showing relationship among the members of Bothriochloa–Dichanthium–Capillipedium agamic complex [Redrawn from Berthaud (2001)]

species and varieties of genus Dichanthium, and its available accessions are in the National Plant Germplasm System (NPGS). It shows that there are around 80 accessions of D. annulatum with majority of the accessions with Pakistan (29 accessions) and India (28 accessions). There are seven accessions of D. aristatum with majority in US and 14 of D. sericeum with majority available in Australia. There are only two accessions available for D. foveolatum and one each of D. caricosum and D. maccannii.

6.6.2 Development of Cultivars Breeding of apomictic species was a serious challenge due to difficulty in obtaining recombination. With the increasing advances in understanding of the genetics of apomixis and availability of various tools, this task now seems to be achievable (Bashaw 1975, 1980). The

basic methods for cultivar development focuses on selection of useful commercial genotypes from a collection of apomictic genotypes or breeding of synthetic tetraploids with natural tetraploid apomicts. With the available germplasm collections, the major job is to identify the best genotype among the candidate genotypes. The current and previous proposals for breeding involved the perpetuation of heterozygous genotypes but failed to include means to exploit heterosis (Miles 2007). The success of any breeding program depends on the range of variability present in the germplasm collection, the proportion of sexual and apomictic seed production, and variability from crossbred progenies. Like many other grass genera, Dichanthium lacks a collection of different genotypes, and there is a need to look for mechanisms that could create variability. A wide platform of variability is available in case of Dichanthium among its different species such as D. aristatum, D. caricosum, D. papillosum, and

102

D. fecundum and among different ecotypes (as available in case of D. annulatum) or even across related genera, for example, Bothriochloa intermedia, B. ischaemum, B. pertusa, and C. parviflorum (Harlan et al. 1958; Harlan 1963; de Wet and Harlan 1968, 1970a). Also, through genetic segregation and recombination, new types can be developed by the breakdown of genomic relationships among the chromosomes introgressed from the species of Bothriochloa and Dichanthium (Mehra and Singh 1968). Variability can be introduced by the use of mutagenic agents as well. For example, conversion of pedicellate spikelet to bisexual flower in D. fecundum, a character governed by a single gene is an outcome of mutation of the gene governing it (Borgaonkar and de Wet 1960). Dichanthium species are usually the freely available rangeland grasses with high nutritive value. Rai et al. (1981) undertook a study to evaluate the available material and to select the best variety for forage production. They identified two cultivars, S-32 and S-65, of D. annulatum for maximum forage production, whereas the cultivars S-123 and S-128 were found to be suitable for soil conservation purposes. Seeds of native grass like D. sericeum are in great demand for the rehabilitation of landscapes affected by mining or urban development. Many cultivars have also been released in many countries for these forage species of Dichanthium. Following are the promising accessions reported by Cook et al. (2005) that have been released for cultivation: 1. D. annulatum accessions IGFRI-S-495-1 and IGFRI-S-495-5 from Indian Grassland and Fodder Research Institute IGFRI, Jhansi, India, and CPI 84146B from Queensland, Australia. 2. D. aristatum accession CPI 104839 from Australia. 3. D. caricosum accessions from Paraguay and northern Argentina 4. D. sericeum accessions ES-100 and ES-200 from Australia. The cultivar names and countries of origin, along with their features, have been tabulated in Table 6.1.

6.6.3 Genomics Little is known about the genome of Dichanthium, and it is difficult due to the occurrence of Dichanthium– Bothriochloa complex and their hybrids. The genetic

V. Bhat et al.

diversity can be evaluated on the basis of morphological, phenotypic, and genotypic characters. The last is more convenient and authentic than the first two, which are usually time-consuming and under the influence of environmental changes (Tanksley et al. 1989; Powell et al. 1996).

6.6.3.1 DNA Extraction Methodology The preliminary requirement of genetic screening studies is good quality DNA. The common problems observed in tropical grasses are high content of polysaccharides and lignin, which act inhibitory to polymerase chain reaction (PCR) analysis and other molecular techniques (Pandey et al. 1996). Chandra and Saxena (2007) were able to optimize rapid protocol for isolating genomic DNA from five tropical grass species, viz., D. annulatum, Heteropogon contortus, Sehima nervosum, Chrysopogon fulvus, and Cenchrus glaucus. The method involved modified CTAB procedure based on Murray and Thompson (1980).

6.6.3.2 Protein or Isozyme Markers Although a facultative apomict, D. annulatum shows high level of polymorphism at molecular level. Gupta et al. (2003) studied the phylogenetic relationships in four tetraploid species of Dichanthium– Bothriochloa complex based on isozyme phenotypes. They found D. annulatum to be evolving continuously and D. aristatum and D. caricosum to be isolated from D. annulatum over a period of time. D. caricosum was found to be more closely related to D. annulatum than D. aristatum according to protein marker studies.

6.6.3.3 DNA Markers A high level of polymorphism in the agromorphological characters has been reported in D.annulatum, independent of its geographical distribution (Chandra et al., 2004, 2006; Agrawal et al., 1999). Random amplified length polymorphism (RAPD) markerbased studies in different accessions of marvel grass collected from two contrasting regions including the

6 Dichanthium

103

Table 6.1 Detailed list of released cultivars of four major forage species of Dichanthium [reported by Cook et al. (2005)] S. no. Name of the Name of the cultivar Country of origin Features species 1. Dichanthium “Kleberg” USA (1944) Excellent drought tolerance, good seed annulatum producer, moderate salinity tolerance, relished by cattle. “Pretoria 90” USA (1954) Selected for seedling vigor, rapid growth (PI 188926, and yield. Good drought tolerance, BN-6730, T-20090) leafy; used for pasture, hay and silage. “T 587”, “PMT-587” USA (1981) Derived from a composite of 80 accessions. Suitable for pasture, hay, revegetation of disturbed areas and salt scalds, erosion control, and range reseeding. More cold tolerant than “Gordo”, “Medio”, and “Kleberg”. High forage producer. “Marvel 8” India 40–100% better dry matter yields than (CPI 106073) local strain. 2. Dichanthium “Medio” (NSL 20670, USA (1954) Selected from naturalized ecotype in aristatum NSL 22695) Texas. Best adapted to heavy soils USA (1957) Good seed producer. “Gordo” (PI 190302, BN-6851, T-20062, NSL 22694) “Alabang X” Philippines Slow to establish and susceptible to weed (PI 297430) competition when young. Best on poorer soils Erroneously referred to as var. heteropogonoides, but may in fact be an ecotype of D. caricosum. “T 587”, “PMT-587” USA (1981) Derived from a composite of 50 Dichanthium accessions established in 1962. Suitable for pasture, hay, revegetation of disturbed areas and salt scalds, erosion control, and range reseeding. More cold tolerant than “Gordo”, “Medio” and “Kleberg”. Resistant to leaf rust. High forage producer. “Floren” (CPI 106374) Australia (1995) Leafy, very palatable ecotype. 3. Dichanthium “Marvel 40” India (1971) Produce 40–100% more dry matter than caricosum (CPI 106073) the naturally occurring strain. “Marvel 93” India (1971) Produce 40–100% more dry matter than (CPI 106075) the naturally occurring strain. “Alabang X” Philippines Not available 4. Dichanthium “Scatta” Pending release in Not available sericeum Australia

North Indian plateau and South India revealed South Indian accessions to be more polymorphic, whereas those from central plateau were found to be more suitable for drought stress. Thus, Saxena and Chandra (2006) were able to identify three accessions of D. annulatum suitable for genetic resource management and for isolating genes controlling drought tolerance.

6.6.3.4 Phylogenetic Studies Andropogoneae is a member of monophyletic subfamily Panicoideae and forms a part of large and well-supported panicoid–arundinoid–chloridoid– centothecoid (PACC) clade of the family Poaceae. Andropogoneae is monophyletic based on data from morphology (Kellogg and Watson 1993), chloroplast

104

V. Bhat et al. Andropogon spp. - AN

100

Core Andropogoneae

78

Hyparrhenia hirta - AT Schizachyrium scoparium - AN

64

Bothriochloa odorata - SO

100 98

Capiltlpedium parviflorum - SO Dichanthium aristatum - SO Cymbopogon flexuosus - AN

99

Cymbopogon jwarancuss - AN Heteropogon contortus - AT Cleistachne sorghoides - SO 72

Microstegium nudum - SA Miscanthus japonicus - SA

Origin of Awns

Saccharum officinarum - SA 100

Chrysopogon futvus - SO Chrysopogon gryllus - SO Ischaemum spp. - IS Apluda mutica IS Sorghum bicolor - SO Coelorachis selloana - RO Elionurus muticus - RO 95

Origin of Disarticulating Rachis

Tripsacum dactyloides - TR Zea mays - TR

95

Chionachne koenigii - CH Phacelurus digitatus - RO

100

Coix aqua - CO 52

100

Arundinella hirta Arundinella nepalensis Paspalum spp.

Panicum spp.

71 100

100

Tristachya superba Urochloa mutica Pennisetum alopecuroides Danthoniopsis dinteri

10 changes

Fig. 6.6 Single-most parsimonious tree based on combined sequences of PHYB, ndhF, and GBSSI. Abbreviations following taxon names indicate respective subtribes according to Clayton and Renvoize (1986). AN: Andropogoninae; AT:

Anthistiriinae; SO: Sorghinae; SA: Saccharinae; IS: Ischaeminae; RO: Rottboelliinae; TR: Tripsacinae; CH: Chionachninae; CO: Coicinae. (Mathews et al. 2002)

gene ndhF (Spangler et al. 1999), and nuclear genes, viz., granule-bound starch synthase I (GBSSI) (MasonGamer et al. 1998) and phytochrome B (Mathews et al. 2002). Mathews et al. (2002) combined the three matrices namely PHYB, GBSSI, and ndhF to generate a data

set of different members of Andropogoneae (Fig. 6.6). Among the awned taxa, they identified a poorly supported clade named as “core Andropogoneae,” which included Andropogon, Bothriochloa, Capillipedium, Cymbopogon, Dichanthium, Heteropogon,

6 Dichanthium

Hyparrhenia, Schizachyrium, and Sorghastrum. In case of individual trees, they were able to resolve a clade of Bothriochloa, Dichanthium, and Capillipedium, but in PHYB and GBSSI trees, Bothriochloa was found to be united with Capillipedium, while in ndhF tree, it grouped with Dichanthium. This confirms their relationships and interfertilities as suggested by Harlan and de Wet (1963a) and Kellogg (2000). Giussani et al. (2001) studied the molecular phylogeny of the subfamily Panicoideae and also found Dichanthium– Bothriochloa–Capillipedium to represent together in cladogram when they mapped base chromosome number, anatomical characters including chloroplast structure, position, and number of bundle sheaths, and physiological characters including type of decarboxylating enzyme, viz. NADP-maleic enzyme. Rondeau et al. (2005) studied phylogenetic relationship among the members of this tribe using an enzyme, NADPdependent malate dehydrogenase (NMDH). This enzyme belongs to multigene family and is highly conserved among plants. They found two distinct genes (NMDH-I and NMDH-II) encoding this enzyme, which would have arisen due to gene duplication associated with selection pressure. They also speculated presence of both genes in few species analyzed, e.g., D. aristatum, H. contortus, etc. but could not detect the transcripts of the latter because of its scarcity.

6.6.4 Biotechnological Studies in Dichanthium Tissue culture and genetic transformation tools could be effectively used to generate somaclonal variation and targeted gene transfers, respectively. Due to predominant apomictic mode of reproduction, variability in this taxon can be heritably fixed in the subsequent generations. Hence, optimization of tissue culture protocol was a very essential initial step toward genetic improvement of D. annulatum. In vitro plant regeneration has been reported from immature inflorescence (Gupta et al. 1997), nodal explant (Gupta et al. 1998), shoot tip (Dalton et al. 2003), and mature seeds (Kumar et al. 2005) of D. annulatum. Genetic transformation is a means to alter the genome by introgression of genes from distant or unrelated grass species or by introduction of alien genes. The desired genes for improving the nutritional quality

105

of Dichanthium could be for lowering the lignin content while increasing the protein quality. The high lignin content of tropical forage plants reduces in vitro digestibility and voluntary feed intake by ruminants. Dalton et al. (2003) reported three transgenic plants of D. annulatum containing hygromycin resistance gene, while two of them contained b-glucuronidase gene (GUS) after embryogenic calli were co-transformed with two plasmids encoding either of these genes. Hygromycin resistance varied from 68 to 100% in the progeny of these three transformants (Fig. 6.7). Although DNA gun-mediated genetic transformation is the method of choice for range grasses (Dalton et al. 2003), which are recalcitrant to Agrobacterium tumefaciens, it generally results in transgenics with multiple copies of a gene that may cause gene silencing, whereas the Agrobacterium-mediated transformation method is cheaper, easier, more reliable, while only a single copy of a gene is delivered. Hence, a comparison between these two methods in terms of their transformation efficiencies would help in improving the level of Agrobacterium-mediated transformation efficiency to that of the gene gun-mediated transformation. Towards this, Kumar et al. (2005) compared frequency of GUS expression between both methods of transformation using binary vectors pCAMBIA1305 and pCAMBIA1301. Among two binary vectors used for Agrobacterium-mediated transformation, pCAMBIA1301 showed higher frequency of GUS expression while pCAMBIA1305 recorded more of the GUS spots per callus. Particle inflow gun-mediated transformation resulted in higher GUS expression compared to the Agrobacterium method using pCAMBIA1305. Further optimization is necessary before using Agrobacteriummediated transformation of Dichanthium.

6.7 Primary Productivity and Biomass Production Production, compartment transfer of dry matter, and aboveground production are higher in case of tropical grasslands compared to their temperate counterpart (Misra and Mall 1975). Studies on dry matter, organic matter, and crude protein content for rangeland grasses in Pakistan show the highest relative preference for D. annulatum than other rangeland grasses (Sultan et al. 2008). D. annulatum has been shown to produce

106

V. Bhat et al.

Fig. 6.7 Tissue culture, gene-gun-mediated transformation, regeneration, and progeny screening of transgenic Dichanthium plants. (a) Shoot tip-derived embryogenic calli (8 weeks old) on maltose-based medium (W1). (b) Shoot regeneration in 47 weeks old callus on regeneration medium. (c) Callus after 5 weeks of growth on Hygromycin (0, 50 and 100 mg/l) containing callus medium (top to bottom). (d) Transient GUS expression in callus 48 h after bombardment with plasmid pACT1-F. (e) GUS expression 11 days after bombardment of callus with plasmid

pACT1-F. (f) Bombarded tissue surviving after 6 weeks of selection on callus medium containing 75 mg/l hygromycin. (g) Root growth in transformed plants established in soil: tillers of R1 and R10 (A) and non-transformed tillers (B) grown hydroponically in a solution containing 50 mg/l hygromycin. (h) Flowering in transformed Dichanthium plants. (i) Floral head of transformed plant. (j) Survival of seedlings of T1 progeny of plant R2 germinated on medium containing 50 mg/l hygromycin (right) compared with non-transformed control (left)

maximum organic matter at Magan grassland at Dhakarwara (Trivedi and Mishra 1979). With reference to dry matter yields, data has been collected for the major forage crops of Dichanthium

sp. of D. annulatum range from 2–6 t/ha and even upto 17 t/ha, whereas for D. aristatum and D. caricosum, it is in the order of 10–12 t/ha. In case of D. sericeum, the yield is dependent upon the soil and climatic

6 Dichanthium

condition as D. sericeum has a wide distribution range (Cook et al. 2005). A good D. annulatum stand yields 3 t/ha hay compared to an average of 9 t/ha in case of D. aristatum (Cook et al. 2005). With regard to the animal production or carrying capacity, the four major cultivated crop species of Dichanthium have been studied thoroughly. D. annulatum stand can support up to 7 sheeps/ha in semiarid environment. In case of D. aristatum and D. caricosum, the live-weight gain of an animal is 270 kg/ha/yr and 150 kg/ha/yr, respectively, when grown in combination with legumes. On the other hand, D. sericeum has a comparatively low carrying capacity compared to these species (Cook et al. 2005). The next section discusses various factors, which affect the productivity and biomass production in Dichanthium sp.

6.7.1 Various Factors Affecting Productivity and Biomass Production in Dichanthium 6.7.1.1 Biotic Factors Grazing and Herbage Removal The effect of increase in the density or crowding of D. annulatum and little or no grazing showed initial increase in yield but subsequent decrease in aboveground herbage, biomass, and higher mortality (Tripathi and Gupta 1980). Studies on D. annulatum show high degree of palatability at all stages of its growth, and therefore, it is considered to be grazed preferentially compared to other species. This study on relative palatability is useful for studying the impact of grazing on natural grasslands and also in evolving suitable practices required for their management (Dabadghao and Marwaha 1962). However, recent studies have shown that grazing in rangelands causes weakening and death of native and introduced desirable species and allow less desirable invasive plants. For example, D. annulatum competes with native desirable species as it can persist grazing pressure, high temperature, and less water. Ortega-Santos et al. (2007) found that D. annulatum affects primary productivity of grazing lands for cattle

107

as it is less desirable than the native species such as Cynodon nlemfuencis, Panicum maximum, Brachiaria decumbens, and Digitaria decumbens. The growth performance like seed weight of D. annulatum has also been found to decrease with increase in biotic disturbances, whereas the number of vegetative buds on rhizomes is found to increase, which results in higher biomass and productivity. It has also been found that the habit of the plant changes from erect to prostrate forms in case of protected and grazed grounds, respectively (Ambasht and Maurya 1970).

Diseases and Pests Many fungi have been reported to attack Dichanthium sp., but major damage has not been reported. Balansia sclerotica, Cerebella andropogonis, Cochliobolus cymbopogonis, Curvularia andropogonis, C. lunata, C. robusta, C. gudauskasii, Phyllachora ischaemi (tar spot), Pithomyces graminicola, Puccinia cesatii, P. duthiae, P. propinqua, P. kenmorensis (rust), Sclerospora dichanthicola, Sphacelotheca annulata, S. andropogonis-annulati, Striga lutea, Tolyposporella obesa, Uredo susica, Uromyces andropogonisannulati, and U. clignyi to name a few. Downy mildew fungus, namely Sclerospora sorghi, has been identified as pathogens of D. caricosum in Thailand (GRIN). In 1966, an undescribed species of Curvularia from Texas (later named as C. robusta due to the large size of conidia) was identified to be pathogenic on D. annulatum (Kilpatrick and Luttrell 1967). A list of various smut fungi including Sporisorium dichanthii affecting D. aristatum; S. dichathicola pathogenic on D. caricosum; and S. sahahyai (Wang and Piepenbring 2002), Ustilago duthiei and U. sabourieana on D. annulatum has been reported Va´nky (2004). Claviceps sp., which causes ergot disease pose a significant hindrance in seed production in D. annulatum and D. aristatum. D. caricosum has been found to be susceptible to root-knot nematodes whereas D. aristatum acts as an alternate host for sheath blight of rice caused by Rhizoctonia solani (Cook et al., 2005). Apart from pathogenic fungi, grasslands support a rich fungal population, which exists in the mantle around the roots. These rhizosphere fungi were found to effect seed germination and root growth of the seedlings (Leelavathy 1969a, b).

108

6.7.1.2 Abiotic Factors Seasonal Burning Fire is considered to be a powerful factor to design community structure in rangelands and to manipulate plant population. Burning has shown to stimulate the shoot–root growth of D. annulatum but decrease in belowground biomass. It has been found that burnt plots show higher productivity as a result of improved nutrient status, less shading, and presence of young tillers (Pandey 1988). Fire thus influences the uptake and release of nutrients like NPK and in turn increases the mineral pool (Paulsamy et al. 1995, 2005) encouraging higher productivity.

Factors affecting Seed Germination Various physical factors such as light, temperature, and habitat condition affect seed germination of D. annulatum. Seeds of D. annulatum are photosensitive and show high germination rate during rainy season. Seed cover and increasing depth of sown seeds affects germination capability, whereas continuous light and optimum temperature of 30 C shows maximum germination percentage in D. annulatum (Maurya 1972). On the other hand, the warm season native perennial grass D. sericeum germinates usually during mid-summer to early autumn period.

V. Bhat et al.

growth of D. annulatum seedlings have been found to decrease with increase in salinity level (Akhtar and Hussain 2008; Khan et al. 2009). It was found that the moisture content in seedlings grown under saline conditions was more than control seedlings, which indicating adaptive mechanism in grasses by becoming succulents in saline soils.

Soil Moisture Soil moisture stress along with clipping intensity has shown to increase the water use efficiency (WUE) in D. annulatum. As soil moisture stress increases, the transpiration rate drops down. The drop in transpiration rate is more pronounced in winters than in summers, which indicates that a strong moisture conservation mechanism operates in D. annulatum (Singh and Misra 1985). The root:shoot ratio was found to be higher with increase in soil moisture stress, and studies have indicated that dry matter production was maximum when water potential is between field capacity and saturation percentage (Singh and Sinha 1983). Except for nitrogen, concentration of all other nutrients decrease with increase in soil moisture stress. Nitrogen, potassium, and phosphorus use efficiencies increase with high soil moisture stress (Misra and Singh 1982).

Shade Seed Dormancy In Dichanthium, seed dormancy is imposed by covering of the embryo (i.e., the seed coat) or within the embryo. Smoke-stimulated germination and scarification help to release dormancy due to seed coat, whereas exposure of seeds to fluctuating temperatures releases dormancy imposed by the embryo in D. sericeum (Read and Bellaris 1999; Adkins et al. 2002).

Salinity The germination percentage, plumule–radicle length, shoot–root dry weight, and in turn establishment and

Shade affects chlorophyll accumulation in D. annulatum, i.e., chlorophyll b increases but chlorophyll a decreases. An increase in chlorophyll (a þ b) shows adaptability of this grass under shade. Shade also increases the accumulation of carotenoids, which increases the quality of fodder (Baig et al. 2005). Shade has no effect on C and N influxes in D. aristatum, but there is a preferential allocation of C and N under low irradiance, which changes shoot: root ratio in turn affecting its photosynthetic capacity. K and P accumulation also increases under shade. D. aristatum was not able to use radiation efficiently, which indicated that it is not adapted to silvopastoral system where shade is more (Cruz 1997a, b).

6 Dichanthium

6.8 Conclusion Dichanthium forms a major component of grass covers of tropics and subtropics. So far, around 20 species are known 1200 in this genus, which are widely distributed. Four major species are popular forage crops, whose cultivars are commercially utilized. This genus also encompasses high levels of cytomorphological diversity, which has been utilized using selection methodologies. However, the occurrence of predominantly apomictic mode of reproduction restricts the rate of breeding in these cultivated forage species. There have not been concerted efforts to compile the available information on the taxonomy, cytology, hybridization, breeding, and genomic approaches, adopted for the genetic improvement of this grass for achieving higher productivity and increased forage quality. Dichanthium is also a reservoir of genes for bioprospecting such as apomixis, high tillering, perenniality, and stress tolerance. Little attempt has been made towards its genetic improvement using advanced scientific tools. Comprehensive data on genetic diversity at molecular level and biotechnological interventions are lacking in Dichanthium, although it forms a major component of grasslands. By increasing the level of drought tolerance in the cultivated Dichanthium species, the area under their cultivation can be increased. This requires development of effective genetic engineering strategies. Similarly, improvement in protein quality of forages can lead to higher animal productivity. Lowering the lignin content using RNA interference (RNAi) approaches can result in increased dry matter digestibility having repercussions on animal productivity and health. More research efforts are also needed to understand the gene flow among the members of agamic complex formed by Dichanthium with Bothriochloa and Capillipedium.

References Adkins SW, Bellairs SM, Loch DS (2002) Seed dormancy mechanisms in warm season grass species. Euphytica 126:13–20 Agarwal DK, Gupta S, Roy AK, Gupta SR (1999) Study on agro-morphological variations vis-a`-vis geographical distri-

109 bution in Marvel grass [Dichanthium annulatum L. (Stapf)]. Plant Genet Resour Newsl 118:27–29 Akhtar P, Hussain F (2008) Salinity tolerance of three range grasses at germination and early growth stages. Pak J Bot 40 (6):2437–2441 Ambasht RS, Maurya AN (1970) Reproductive capacity of Dichanthium annulatum Stapf. in relation to biotic factors. Trop Ecol 11(2):186–193 Asker S (1966) Effects of mutagen treatment on some apomictic Potentilla species. Hereditas 55:249–265 Asker S, Jerling L (1992) Apomixis in plants. CRC Press, Boca Raton, FL Baig MJ, Anand A, Mandal PK, Bhatt RK (2005) Irradiance influences contents of photosynthetic pigments and proteins in tropical grasses and legumes. Photosynthetica 43(1):47–53 Baker HG (1959) Reproductive methods as factors in speciation in flowering plants. Cold Spring Harb Symp Quant Biol 24:177–191 Bashaw EC (1975) Problems and possibilities of apomixis in the improvement of tropical forage grasses. In: Doll EC, Mott GO (eds) Tropical forages in livestock production systems. Am Soc Agron Spl Publ 24. Madison, WI, 23–30 Bashaw EC (1980) Registration of Neuces and Llano buffelgrass. Crop Sci 20:112 Berthaud J (2001) Apomixis and the management of genetic diversity. In: Savidan Y, Carman JG, Dresselhaus T (eds) The flowering of apomixis: from mechanisms to genetic engineering. CIMMYT, IRD, Mexico DF, pp 8–23 Bhanwra RK (1988) Embryology in relation to systematics of Gramineae. Ann Bot 62:215–233 Bhattacharyya UC, Uniyal BP (1973) On the occurrence of Dichanthium pallidum (Hook. F.) Stapf. ex C.E.C. Fischer in northern India. Bull Bot Surv India 15(1 and 2):167 Bisset WJ, Sillar DI (1984) Angleton grass (Dichanthium aristatum) in Queensland. Trop Grassl 18(4):161–173 Bogdan AV (1977) Tropical pasture and fodder plants. Longman, London, New York Bor NL (1960) The grasses of Burma, Ceylon, India and Pakistan. Pergamon, Oxford Borgaonkar DS, de Wet JMJ (1960) A cytogenetic study of hybrids between Dichanthium annulatum and Dichanthium fecundum. Phyton (Argentina) 13:137–144 Borgaonkar DS, de Wet JMJ (1961) Segregation among partial apomictic Bothriochloa hybrids. Am J Bot 48:545 (Abstr) Borgaonkar DS, de Wet JMJ (1964) Cytogenetical study of hybrids between Dichanthium annulatum and D. fecundum III. Phyton (Argentina) 21:55–66 Borgaonkar DS, Singh AP (1962) Species relationships in Dichanthium IV. D. annulatum and D. papillosum. Phyton 19:101–107 Brooks MH (1958) A study of the reproductive mechanisms in certain species of the Bothriochloa and Dichanthium complexes. PhD Thesis, Okla State Univ, USA Brown WV, Emery WHP (1957) Some South African apomictic grasses. J S Afr Bot 23:123–125 Brown WV, Emery WHP (1958) Apomixis in the Gramineae, Panicoideae. Am J Bot 45:253–269 Carnahan HL, Hill HD (1961) Cytology and genetics of forage grasses. Bot Rev 27:1–162

110 Celarier RP (1956) Additional evidence for five as the basic chromosome number of the Andropogoneae. Rhodora 58:135–143 Celarier RP (1957) Elyonurus argenteus, a South African grass with five chromosome pairs. Bull Torrey Bot Club 84:157–162 Celarier RP, Harlan JR (1957) Apomixis in Bothriochloa, Dichanthium and Capillipedium. Phytomorphology 7:93–102 Celarier RP, Mehra KL, Wulf ML (1958) Cytogeography of the Dichanthium annulatum complex. Brittonia 10:59–72 Celarier RP, de Wet JMJ, Borgaonkar DS, Harlan JR (1960) Intergeneric hybrids in Bothriochloninae I. Bothriochloa intermedia and Dichanthium annulatum. Cytologia 26:170–175 Celarier RP, de Wet JMJ, Richardson WL (1961) Species relationships in Dichanthium I. Hybridizations between D. caricosum, D. arsitatum and D. annulatum. Phyton (Argentina) 16:63–67 Celarier RP, de Wet JMJ, Bakshi JS (1962) Species relationships in Dichanthium. II. The taxonomy of D. caricosum and D. aristatum. Phyton (Argentina) 18(1):1–9 Chandra A, Saxena R (2007) A rapid protocol for isolating genomic DNA from tropical grass species suitable for RAPD, ISSR and STS. Cytologia 72(2):427–434 Chandra A, Saxena R, Roy AK, Pathak PS (2004) Estimation of genetic variation in Dichanthium annulatum genotypes by RAPD technique. Trop Grassl 38:245–252 Chandra A, Saxena R, Roy AK (2006) Polymorphism and genotype-specific markers for Dichanthium identified by random amplified polymorphic DNA. Genet Resour Crop Evol 53:1521–1529 Chheda HR, Harlan JR (1963) A cytogenetical study of intergeneric hybrids between Bothriochloa intermedia and Dichanthium fecundum. Cytologia 27:418–423 Chheda HR, de Wet JMJ, Harlan JR (1961) Aneuploidy in Bothriochloa hybrids. Caryologia 14:205–217 Clayton WD (1977) New grasses from eastern Africa. Studies in Gramineae: 42. Kew Bull 32(4):1–4 Clayton WD, Renvoize SA (1982) Dichanthium foveolatum and Dichanthium caricosum. In: Polhill RM (ed) Flora of tropical east Africa. Balkema AA, Rotterdampp, p 451 Clayton WD, SA Renvoize (1986) Genera graminum. Her Majesty’s Stationery Office, London Cook BG, Pengelly BC, Brown SD, Donnelly JL, Eagles DA, Franco MA, Hanson J, Mullen BF, Partridge IJ, Peters M, Schultze-Kraft R (2005) Tropical forages: an interactive selection tool. CSIRO, DPI&F (Qld), CIAT and ILRI, Brisbane, Australia. http://www.tropicalforages.info Cruz P (1997a) Effect of shade on the carbon and nitrogen allocation in a perennial tropical grass, Dichanthium aristatum. J Exp Bot 48(306):15–24 Cruz P (1997b) Effect of shade on the growth and mineral nutrition of a C4 perennial grass under field conditions. Plant Soil 188:227–237 D’Cruz R, Reddy PS (1971) Inheritance of apomixis in Dichanthium. Indian J Genet Plant Breed 31(3):451–460 Dabadghao PM, Marwaha SP (1962) Relative palatability on important indigenous grass species of Western Rajasthan. Indian J Agron 6:323–327

V. Bhat et al. Dabadghao PM, Shankarnarayan K (1973) The grass cover of India. ICAR, India Dalton SJ, Bettany AJE, Bhat V, Gupta MJ, Bailey K, Timms E, Morris P (2003) Genetic transformation of Dichanthium annulatum (Forssk) – an apomictic tropical forage grass. Plant Cell Rep 21:974–980 de Wet JMJ (1954) Chromosome numbers of a few South African grasses. Cytologia 19:97–103 de Wet JMJ (1965) Diploid races of tetraploid Dichanthium species. Am Nat 99:167–171 de Wet JMJ (1968) Diploid-tetraploid-haploid cycles and the origin of variability in Dichanthium agamospecies. Evolution 22:394–397 de Wet JMJ, Borgaonkar DS (1963) Aneuploidy and apomixis in Bothriochloa and Dichanthium. Bot Gaz 124(6):437–440 de Wet JMJ, Harlan JR (1961) Mode of reproduction and morphological variations in Dichanthium. Am J Bot 48:545–546 de Wet JMJ, Harlan JR (1962) Species relationships in Dichanthium. III D.sericeum and its allies. Phyton 18 (1):11–14 de Wet JMJ, Harlan JR (1966) Morphology of the compilospecies Bothriochloa intermedia. Am J Bot 53:94–98 de Wet JMJ, Harlan JR (1968) Taxonomy of Dichanthium section Dichanthium (Gramineae). Bol Soc Argent Bot 12:206–227 de Wet JMJ, Harlan JR (1970a) Apomixis, polyploidy and speciation in Dichanthium. Evolution 24:270–277 de Wet JMJ, Harlan JR (1970b) Bothriochloa intermedia – a taxonomic dilemma. Taxon 19:339–340 de Wet JMJ, Singh AP (1964) Species relationships in Dichanthium V. The diploid species. Caryologia 17:153–160 de Wet JMJ, Stalker HT (1974) Gametophytic apomixis and evolution in plants. Taxon 23(5/6):689–697 de Wet JMJ, Mehra KL, Borgaonkar DS (1961) Chromosome association in Dichanthium hybrids. Cytologia 26:78–82 eFloras. http://www.efloras.org Giussani LM, Cota-Sa´nchez JH, Zuloaga FO, Kellogg EA (2001) A molecular phylogeny of the grass subfamily Panicoideae (Poaceae) shows multiple origins of C4 photosynthesis. Am J Bot 88(11):1993–2012 Gould FW (1956) Chromosome counts and cytotaxonomic notes on grasses of the tribe Andropogoneae. Am J Bot 43:395–404 Gould FW (1967) The grass genus Andropogon in the United States. Brittonia 19:70–76 GRIN. USDA, ARS, National Genetic Resources Program. Germplasm Resources Information Network – (GRIN) [Online Database]. National Germplasm Resources Laboratory, Beltsville, Maryland. http://www.ars-grin.gov/cgi-bin/ npgs/html/genform.pl Gupta PK, Singh AP, Roy RP (1969) Aposporous apomixis, seasonal variation in tetraploid Dichanthium annulatum (Forssk.) Stapf. Acta Biol 11:253–260 Gupta MG, Gupta S, Bhat V, Bhat BV, Bhat V (1997) In vitro regeneration and somaclonal variation in tropical pasture grass Dichanthium annulatum (Forssk). Indian J Range Manage Agrofor 18:25–30 Gupta MG, Gupta S, Bhat BV, Gupta S (1998) Plant regeneration from callus cultures initiated from nodes of a grass Dichanthium annulatum (Forssk) Stapf. Acta Bot Indica 26:31–36

6 Dichanthium Gupta MG, Bhat V, Bhat BV, Neeraja CN, Gupta S (2003) Phlylogenetic relationships in tetraploid agamospecies of Dichanthium complex based on isozyme phenotypes. J Plant Biol 30(1):61–64 ˚ (1947a) Apomixis in higher plants II. The casual Gustafsson A ˚ rsskr II 43:71–179 agent of apomixis. Lunds Univ A ˚ (1947b) Apomixis in higher plants III. Gustafsson A ˚ rsskr II Biotype and species formation. Lunds Univ A 43:183–370 Hackel E (1889) Monographia andropogonearum. In: DeCandolle (ed) Monographiae phanerogamarum, vol 6. G. Masson, Paris, France Harlan JR (1963) Two kinds of gene centers in Bothriochloininae. Am Nat 97:91–98 Harlan JR, de Wet JMJ (1963a) The role of apomixis in the evolution of the Bothriochloa–Dichanthium complex. Crop Sci 3:314–316 Harlan JR, de Wet JMJ (1963b) The compilo-species concept. Evolution 17:497–501 ¨ . Winge and a prayer: the Harlan JR, de Wet JMJ (1975) On O origins of polyploidy. Bot Rev 41(4):361–390 Harlan JR, Celarier RP, Richardson WL, Brooks MH, Mehra KL (1958) Studies on Old World bluestems II. Okla Agri Exp Stn Tech Bull T-72:1–23 Harlan JR, de Wet JMJ, Richardson WL, Chheda HR (1961) Studies on Old World bluestems III. Okla Agri Exp Stn Tech Bull T-92:5–30 Harlan JR, Brooks MH, Borgaonkar DS, de Wet JMJ (1964) The nature and inheritance of apomixis in Bothriochloa and Dichanthium. Bot Gaz 125:41–46 Heslop-Harrison J (1961) The function of the glume pit and the control of cleistogamy in Bothriochloa decipiens (Hack.) C. E. Hubbard. Phytomorphology 11:378–383 Jacobsen CN (1981) A review of the species of Dichanthium native to Australia with special reference to their occurrence in Queensland. Trop Grassl 15(2):84–95 Jain SK, Deshpande UR (1978) Transfer of some Indian species of Bothriochloa and Capillipedium to Dichanthium. Bull Bot Surv India 20(1–4):133–135 Kellogg EA (2000) Molecular and morphological evolution in Andropogoneae. In: Jacobs SWL, Everett JE (eds) Grasses: systematics and evolution. CSIRO, Collingwood, Victoria, pp 149–158 Kellogg EA, Watson L (1993) Phylogenetic studies of a large data set. I. Bambusoideae, Andropogonodae, and Pooideae (Gramineae). Bot Rev 59:273–343 Kew: Clayton WD, Harman KT, Williamson H (2006 onwards) GrassBase – The Online World Grass Flora. http://www. kew.org/data/grasses-db.html. Accessed 08 Nov 2006 Khan ZH, Qadir I, Yaqoob S, Khan RA, Khan MA (2009) Response of range grasses to salinity levels at germination and seedling stage. J Agric Res 47(2):179–184 Kilpatrick RA, Luttrell ES (1967) An undescribed species of Curvularia pathogenic to Dichanthium annulatum. Mycologia 59:888–892 Knox RB (1967) Apomixis: seasonal and population differences in a grass. Science 157:325–326 Knox RB, Heslop-Harrison J (1963) Experimental control of aposporous apomixis in a grass of the Andropogoneae. Bot Notis 116:127–141

111 Knox B, Heslop-Harrison J (1966) Control of pollen fertility through the agency of the light regime in the grass Dichanthium aristatum. Phyton (Australia) 11:256–267 Kumar J, Shukla SM, Bhat V, Gupta S, Gupta MG (2005) In vitro plant regeneration and genetic transformation of Dichanthium annulatum. DNA Cell Biol 24:670–679 Leelavathy KM (1969a) Effect of some common rhizosphere fungi on root growth of seedlings. Plant Soil 30(2):335–338 Leelavathy KM (1969b) Effect of rhizosphere fungi on seed germination. Plant Soil 30(3):473–476 ` (1960) Biosystematics and classification of apomicts. Lo¨ve A Feddes Rep 63:136–149 ` (1964) The biological species concept and its evolutionLo¨ve A ary structure. Taxon 13:33–45 Mason-Gamer RJ, Weil CF, Kellogg EA (1998) Granule-bound starch synthase: structure, function, and phylogenetic utility. Mol Biol Evol 15:1658–1673 Mathews S, Spangler RE, Mason-Gamer RJ, Kellogg EA (2002) Phylogeny of Andropogoneae inferred from Phytochrome b, GBSSI, and ndhF. Int J Plant Sci 163(3):441–450 Maurya AN (1972) Effect of some physical factors on the seed germination in a herbage grass – Dichanthium annulatum stapf. Trop Ecol 13(2):165–175 Mehra KL (1961) Chromosome number, geographical distribution and taxonomy of the Dichanthium annulatum complex. Cytologia 17:176 Mehra KL (1962) The Dichanthium annulatum complex. I. Morphology. Phyton (Argentina) 18:87–93 Mehra KL (1964a) The Dichanthium annulatum complex III. Origin and artificial synthesis of Dichanthium papillosum Stapf. Caryologia 17(3):545–556 Mehra KL (1964b) The Dichanthium annulatum complex. II. Relationships between the Tropical and Mediterranean types. Phyton 21(2):119–126 Mehra KL (1966) The Dichanthium annulatum complex. IV. Study of natural hybridization from herbarium specimens. Rev Biol 5(3–4):295–302 Mehra KL, Celarier RP (1958) Cytotaxonomic notes on the Dichanthium annulatum complex. Proc Okla Acad Sci 38:22–25 Mehra KL, Singh AP (1968) Chromosome manipulation and genotypic control of chromosome manipulation in Dichanthium annulatum complex and hybrids. Nucleus 11:372–384 Mehrotra A (1979) A few rare Indian grasses. Bull Bot Surv India 21(1–4):237–238 Miles JW (2007) Apomixis for cultivar development in tropical forage grasses. Crop Sci 47(S3):S238–S249 Misra CM, Mall LP (1975) Production and compartment transfer of dry-matter in a tropical grassland community. 41B(5):452–457. http://www.new.dli.ernet.in/rawdataupload/upload/insa/ INSA_1/20005b7e_452.pdf Misra G, Singh KP (1982) Effect of soil moisture and clipping on the nutrient (N, P and K) concentration, uptake and use efficiency in one temperate and two tropical grasses. Plant Soil 69:413–421 Murray MG, Thompson WF (1980) Rapid isolation of high molecular weight DNA. Nucleic Acids Res 8:4321–4325 Myers WM (1947) Cytology and genetics of forage grasses. Bot Rev 13:318–421

112 Naik VN (1982) Notes on two Indian grasses. Bull Bot Surv India 24(1–4):192–193 Oke JG (1950) Chromosome number in some species of Dichanthium Willemet and Bothriochloa O.Kuntze. Proc Indian Acad Sci 32:227–230 Ortega-Santos JA, Avila-Curiel JM, Gonzalez-Valenzuela EA, Gonzalez-Padro´n MA (2007) Grazing intensity and nitrogen fertilization to manage invasive Kelberg Bluestem on Pangola grass pastures in Northern Mexico. Tex J Agric Nat Resour 20:109–115 Pandey AN (1988) Short-term study of recovery of a tropical grassland following seasonal burning. Trop Ecol 29 (2):159–170 Pandey RN, Adams RP, Flournoy LE (1996) Inhibition of random amplified polymorphic DNAs (RAPDs) by plant polysaccharides. Plant Mol Biol Rep 14:17–22 Paulsamy S, Manian S, Udaiyan K (1995) A strategy to evaluate community dynamics through uptake, release rate of nutrients. Range Manag Agrofor 16(1):15–20 Paulsamy S, Manian S, Kil BS (2005) Fire and rangeland’s management in India. Korean J Ecol 28(1):55–61 Pilger R (1954) Das system der Gramineae. Bot Jahrb 76:281–384 Powell W, Morgante M, Andre C, Hanafey M, Vogel J, Tingey S, Rafalski A (1996) The comparison of RFLP, RAPD, AFLP and SSR (microsatellites) marker for germplasm analysis. Mol Breed 2:225–238 Pullaiah T, Febulaus GNV (2000) Embryology and apomixis in grasses. Regency, New Delhi, India Quattrocchi U (2008) CRC World dictionary of grasses: common names, scientific names, Eponyms, Synonyms and Etymology. Taylor and Francis Group, Boca Raton, FL, pp 632–637 Rai P, Pathak PS, Kanodia KC, Patil BD (1981) Comparative assessment of Dichanthium annulatum (Fotssk.) Stapf cultivars for forage production. Indian J Ecol 8(1):17–24 Read TR, Bellaris SM (1999) Smoke affects the germination of native grasses of New South Wales. Aust J Bot 47:563–576 Reddy PS, D’Cruz R (1969a) Mechanisms of apomixis in Dichanthium annulatum (Forssk.) Stapf. Bot Gaz 130(2):71–79 Reddy PS, D’Cruz R (1969b) Polyembryony in Dichanthium annulatum (Forssk.) Stapf. Bot Gaz 130(3):162–165 Rondeau P, Rouch C, Besnard G (2005) NADP-malate dehydrogenase gene evolution in Andropogoneae (Poaceae): gene duplication followed by sub-functionalization. Ann Bot 96:1307–1314 Saran S, de Wet JMJ (1969) A structural peculiarity observed in the sexual embryo sacs of Dichanthium intermedium. Can J Bot 47:1205–1206

V. Bhat et al. Saran S, de Wet JMJ (1970) The mode of reproduction in Dichanthium intermedium (Gramineae). Bull Torrey Bot 97(1):6–13 Saxena R, Chandra A (2006) RAPD and cytological analyses and histological changes caused by moisture stress in Dichanthium annulatum accessions. Cytologia 71(2):197–204 Shouliang C, Phillips SM (2006) Dichanthium Willemet. In: Zhegyi W, Raven PH, Hong D (eds) Flora of China, vol 22, Poaceae. Science Press, Beijing, pp 604–605 Singh AP (1968) Hybridization range of Dichanthium annulatum (Forsk.) Stapf. Intl J Cytol 33:568–574 Singh AP, Mehra KL (1965) Cytological analysis of hybrids within the Dichanthium annulatum complex. Cytologia 30:307–316 Singh KP, Misra G (1985) Water-use efficiency of one C3 and two C4 grasses in response to varying soil moisture and herbage-removal levels in a seasonally dry tropical region. Plant Soil 88:171–180 Singh SK, Sinha P (1983) Influence of soil moisture stress on dry matter production of Dichanthium annulatum. Indian J Ecol 10(2):210–214 Spangler R, Zaitchik B, Russo E, Kellogg EA (1999) Andropogoneae evolution and generic limits in Sorghum (Poaceae) using ndhF sequences. Syst Bot 24:267–281 Srivastava AK, Purnima (1990) Agamospermy in some polyploidy grasses. Acta Bot Indica 18:240–246 Stebbins GL (1947) Types of polyploidy: their classification and significance. Adv Genet 1:403–429 Sultan JI, Inam-ul-rahim YM, Nawaz H, Hameed M (2008) Nutritive value of free rangeland grasses of northern grasslands of Pakistan. Pak J Bot 40(1):249–258 Tanksley SD, Young ND, Paterson AH, Bonierbate MW (1989) RFLP mapping in plant breeding: new tools for an old science. Biotechnology 7:257–264 Tripathi RS, Gupta GP (1980) The growth of Bothriochloa pertusa and Dichanthium annulatum in relation to crowding and herbage removal. Oikos 34:219–226 Trivedi BK, Mishra GP (1979) Seasonal variations in species composition, plant biomass and net community productivity of two grasslands in Sehima-Dichanthium cover type. Trop Ecol 20(1):114–125 Valentine DH (1960) The treatment of apomictic groups in Flora Europaea. Feddes Rep 63:114–127 Va´nky K (2004) The smut fungi (Ustilaginomycetes) of Bothriochloa, Capillipedium and Dichanthium (Poaceae). Fung Divers 15:221–246 Wang S, Piepenbring M (2002) New species and new records of smut fungi from China. Mycol Prog 1(4):399–407 Willemet P (1976) Herbarium mauritianum. In: Usteri P (ed) Annalen der Botanik 18:11–13

.

Chapter 7

Eleusine Susana S. Neves

7.1 Introduction Eleusine Gaertn. is a small genus (eight species) of the grass family (Poaceae, subfamily Chloridoideae) that includes the finger millet or ragi (Eleusine coracana subsp. coracana), a highly valuable crop that over the centuries has guaranteed the survival of millions of people, particularly in eastern and southern Africa and India (National Research Council 1996). In subSaharan Africa and South Asia, millets are survival crops for poor people, and nearly 4 million hectares (ha) of land are devoted to the cultivation of finger millet: 3 million ha in India and 1 million ha in Africa (Naylor et al. 2004). The grain of finger millet is used for preparing diverse types of food (porridge, soup, bread, cakes, malt, alcoholic beverages), is highly nutritious and tasty, and has excellent storage qualities (Hilu and De Wet 1976a; Barbeau and Hilu 1993; National Research Council 1996; Nout 2009). Finger millet and some other Eleusine species are also useful as forage grasses. Two species, E. floccifolia and E. jaegeri, are also used in Ethiopia for making baskets and other household items (Phillips 1995; Grassland Index 2009). However, the genus also includes a very problematic and cosmopolitan weed, commonly known as goosegrass (E. indica), considered one of the worst weeds in the world (Holm et al. 1977; Radosevich et al. 2007) and one of the ten most

S.S. Neves Plant Cell Biotechnology Laboratory, ITQB, Instituto de Tecnologia Quı´mica e Biolo´gica, Universidade Nova de Lisboa, Apartado 127, 2781-901 Oeiras, Portugal e-mail: [email protected]

economically important herbicide-resistant weeds (Basu et al. 2004; Beckie 2007). In spite of the importance of finger millet for so many human populations, the development of genetic resources and breeding efforts in this crop or its allied species have been very limited until recent years. This chapter gives an overview on current knowledge and resources thus far developed in Eleusine, including species characterization, details on research achievements, and some limitations associated with some of the species, and a discussion on future prospects in conservation and utilization of these valuable plants.

7.2 Taxonomy and Species Characterization The genus Eleusine includes eight species of annual and perennial herbs, mostly of African origin, with only one species being native to South America (E. tristachya). Eleusine is essentially a tropical and subtropical genus, with several of its species (E. floccifolia, E. intermedia, E. jaegeri, E. kigeziensis, E. multiflora) being adapted to upland habitats (grassland and open forest or bushland), growing in altitudes well above 1,000 m (Phillips 1972, 1995). Other species of Eleusine (E. coracana, E. indica, and E. tristachya) grow in a wider range of open habitats, from sea level to the highlands. The adaptation to highland habitats is also evident in the crop (E. coracana subsp. coracana), which grows better in higher altitudes than other tropical cereals – finger millet is cultivated in Africa at altitudes between 1,000 and 2,000 m and in Nepal up to at least 2,400 m (National Research Council 1996).

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_7, # Springer-Verlag Berlin Heidelberg 2011

113

114

Eleusine species are herbaceous plants with flattened culms (stems) and a digitate or subdigitate inflorescence (more rarely shortly racemose), which is formed by (1–)2–10(–20) spikes, arranged into a terminal whorl or clustered at the top of the flowering stem, sometimes with one or two spikes placed a little lower in the axis (the finger-like appearance of the terminal inflorescence is the reason why the crop is known as “finger millet”). Each spike has many biseriate and overlapping (imbricate) spikelets that are laterally compressed and usually disarticulate at maturity, except in the crop where spikelets are persistent. The fruit (grain) of Eleusine is ornamented and is enclosed by a thin pericarp that is easily removed from the grain when soaked in a drop of water (Phillips 1972, 1995). The grain of grasses is usually smooth and has a pericarp that is fused to the seed surface (a true caryopsis). Several grass genera have smooth grains with free pericarp (e.g., Brachychloa S.M.Phillips, Chloris Sw., Ochtochloa Edgew., Sclerodactylon Stapf, Sporobolus R.Br.). However, the character combination of an ornamented grain with free pericarp is highly unusual in grasses (Phillips 1972), and it is found only in Eleusine and four other chloridoid genera, Acrachne Chiov., Dactyloctenium Willd., Coelachyrum Hochst. and Nees, and a few species of Sporobolus (for descriptions of grass genera, see Clayton et al. 2006-onwards). These genera, however, can be easily distinguished from Eleusine based on inflorescence characters. Eleusine has spikes that typically terminate in fertile spikelets, and glumes and lemmas (bracts) that are always awnless (i.e., do not have any awns or bristle-like prolongations). In contrast, the spikes of Acrachne terminate in an abortive spikelet, and those of Dactyloctenium terminate in a bare pointed extension of the rhachis (axis of the spike); these two genera also have awned bracts. Coelachyrum has a paniculate or racemose inflorescence (not digitate) and a grain that is concave-convex in cross-section; the grain in Eleusine is trigonous in section or globose (Phillips 1972, 1995). Sporobolus has paniculate inflorescences with 1-flowered spikelets, while Eleusine always have several-flowered spikelets. As it is so uncommon among grasses, the ornamentation of the Eleusine fruit is thus a very useful character for identification of the genus, but the ornate grain surface (finely granular or striate) is visible only with some magnification (10
or more)

S.S. Neves

and after the thin pericarp that encloses the grain has been removed. Keys for identification of Eleusine species are provided by Phillips (1972, 1974, 1995). The online database and software available through GrassBase (Clayton et al. 2006-onwards) are also useful for interactive identification of the world’s grass species. General descriptions of each Eleusine species are given below and in Table 7.1; emphasis is given to characters that help to discriminate species. Table 7.1 summarizes information on species names, chromosome numbers, morphological characters, geographical distribution, uses and economic importance. Authors of species/taxon names given in Table 7.1 are not repeated in the text.

7.2.1 E. coracana E. coracana is an annual allotetraploid (2n ¼ 4x ¼ 36) that includes two distinct subspecies: subsp. coracana (finger millet) and subsp. africana (wild finger millet) (Hilu and De Wet 1976a; Hilu 1994). Morphological, cytogenetic, and molecular evidence suggests that cultivated finger millet (subsp. coracana) was domesticated from wild populations of E. coracana subsp. africana (Chennaveeraiah and Hiremath 1974; Hilu and De Wet 1976a; Hilu and Johnson 1992; Dida et al. 2008). The crop, with its stout spikes and large and globose grain, is easily distinguished from wild taxa of Eleusine, and, for that reason, some authors prefer to treat cultivated and wild finger millet as separate species: E. coracana (the crop) and E. africana (e.g., Hiremath and Salimath 1991; Phillips 1995; Bisht and Mukai 2002). However, hybridization between wild (subsp. africana) and cultivated (subsp. coracana) populations occurs naturally and frequently, giving rise to many morphological intermediates wherever the two populations meet (Kennedy-O’Byrne 1957; Mehra 1962; De Wet et al. 1984; Phillips 1995). Dida et al. (2008) have also recently demonstrated by genotypic analysis of microsatellites that there is clear evidence of gene flow between subsp. africana and subsp. coracana. Therefore, as previously discussed (Neves et al. 2005), the rank of subspecies more adequately reflects the biological relationship between the cultivated “E. coracana” and the wild “E. africana”.

Table 7.1 The species of Eleusine: names, chromosome numbers, morphological characters, geography, uses, and economic impact Taxon Vernacular names Chromosome Diagnostic characters Geographical number distribution Mainly East and 2n (4x) ¼ 36 Annual; culms robust; leaves Finger millet, ragi, Eleusine coracana (L.) southern Africa and herbaceous (soft), glabrous, nachni, koddo, Gaertn. subsp. India, also in Nepal, sometimes with pilose margins; dagussa, dagusha, coracana China, and other inflorescence digitate or tocussa, barancia, South Asian subdigitate; spikes 3–10(–20), thick baranika, uembe, countries and stout, incurved or straight, ulezi, wimbi, rapoko, (cultivated) 4–14 cm long, (7–)9–15 mm wide; etc grain globose, black, brown, reddish, or whitish Tropical and southern E. coracana (L.) Gaertn. Wild finger millet 2n (4x) ¼ 36 Annual; culms moderately robust; Africa, mainly subsp. africana leaves  soft, usually glabrous; eastern and southern (Kenn.-O’Byrne) inflorescence digitate or uplands Hilu and de Wet subdigitate; spikes (2–)4–17, [Synonym: slender, straight, 4–17 cm long, E. africana Kenn.4–8 mm wide; grain ovate-oblong, O’Byrne] trigonous in section, black or brownish E. floccifolia (Forssk.) Akirma, akrma, dagoo, 2n (2x) ¼ 18 Perennial; culms moderately robust; Ethiopia, Eritrea, Spreng. garrgorr leaves  tough, with small tufts of Yemen, Somalia, Kenya white hairs scattered along the (1,500–3,200 m margins; inflorescence subdigitate; altitude) spikes 2–8(–10), slender, straight, 5–12 cm long, 3.5–6 mm wide; grain elliptic to oblong, trigonous, blackish E. indica (L.) Gaertn. Goosegrass, crowsfoot 2n (2x) ¼ 18 Annual; culms slender; leaves soft, Cosmopolitan weed of grass, silver glabrous to sparsely pilose; African origin; crabgrass, wiregrass, inflorescence digitate or mostly tropics and paragis, grama de subdigitate; spikes (1–)2–10(–17), subtropics caballo, pata de slender, straight, 3.5–15.5 cm long, gallina, etc. 3–6 mm wide; grain elliptic, trigonous, blackish Ethiopia, Kenya E. intermedia (Chiov.) – 2n (2x) ¼ 18 Perennial; culms moderately robust; (1,100–1,800 m S.M. Phillips leaves herbaceous (not tough), altitude) glabrous to pilose (no hair tufts); inflorecence subdigitate or shortly racemose; spikes 4–15, slender, straight, 5–12 cm long, 4–8 mm wide; grain broadly elliptic, trigonous, blackish (continued)

Unknown uses but possibly of value as forage grass

Very problematic weed, difficult to eradicate and resistant to multiple herbicides. On the positive side: used as forage and in traditional medicine in Africa and Asia

Used for making baskets (Ethiopia). Frequent in heavily grazed pastures of mid-altitude in Ethiopia, but unpalatable to cattle

Forage grass. Also a weed that when associated to finger millet fields may seriously reduce the yields of the crop

Cereal crop. Grain used to prepare porridge, soup, bread, cakes, malt, and alcoholic beverages (beer, liquors). Also used for forage, soilretention, papermaking, and in traditional medicine

Uses and economic impact

7 Eleusine 115

Vernacular names

Chromosome Diagnostic characters Geographical Uses and economic impact number distribution E. jaegeri Pilg. Manyata grass, 2n (2x) ¼ 20 Perennial; culms robust; leaves tough Kenya, Tanzania, Used for making baskets. Frequent mafutiana, akirma, (like leather) with rough (scabrid) Uganda, Ethiopia invader of highland pastures in East dagoo, titima margins, glabrous; inflorescence (1,800–3,300 m Africa but avoided by livestock subdigitate or shortly racemose; altitude) (unpalatable) spikes 2–10(–13), slender, straight, 4–17 cm long, 3–7 mm wide; grain elliptic-oblong, trigonous, blackish Uganda, Congo, Possibly of value as forage grass E. kigeziensis S.M. – 2n (4x) ¼ 38 Perennial; culms moderately robust; Rwanda, Burundi, Phillips leaves  soft, glabrous beneath, Ethiopia sparsely pilose on the upper surface; (2,000–2,700 m inflorescence digitate; spikes 2–7, altitude) slender, straight, 7.5–14 cm long, 4.5–5.5 mm wide; grain elliptic, trigonous, blackish Ethiopia, Eritrea, Potentially valuable as forage grass. E. multiflora Hochst. ex – 2n (2x) ¼ 16 Annual; culms slender; leaves soft, Kenya, Tanzania Occasional weed A.Rich. sparsely pilose (no hair tufts); (1,500–3,000 m inflorescence racemose; spikes altitude) (2–)3–8, more or less clustered at the top of the axis, oblong to ovate, straight or slightly curved, 1–4 cm long, 8–16 mm wide; grain oblong in profile, laterally compressed, blackish E. tristachya (Lam.) Three-spike goosegrass 2n (2x) ¼ 18 Annual; culms slender; leaves soft, South America Forage grass (Argentina). Also a weed Lam. glabrous; inflorescence digitate; (introduced in other with increasing occurrence in many spikes (1–)2–3, oblong, straight, regions) regions of the world 1–6(–8) cm long, 5–16 mm wide; grain oblong-globose, trigonous in section, blackish Notes: Species/taxon information was obtained from Kennedy-O’Byrne (1957), Phillips (1972, 1974, 1995), Hilu and De Wet (1976a), Hansen (1980), Hilu (1980, 2003), National Research Council (1996), Sisay and Baars (2002), Agnew (2006), Clayton et al. (2006-onwards), Chen and Phillips (2006), Lovisolo and Galati (2007), Agyare et al. (2009), Grassland Index (2009), USDA NRCS (2009), and herbarium specimens (S. Neves, pers. observ.). Chromosome numbers were obtained from Hiremath and Chennaveeraiah (1982) and Hiremath and Salimath (1991)

Table 7.1 (continued) Taxon

116 S.S. Neves

7 Eleusine

117

7.2.1.1 E. coracana subsp. coracana

7.2.1.2 E. coracana subsp. africana

The taxon E. coracana subsp. coracana corresponds to cultivated finger millet and, in the literature and current databases, is more frequently referred simply as E. coracana. Finger millet is easily recognized by its typical digitate or subdigitate inflorescences, with thick and stout spikes (Fig. 7.1a), which are often incurved, and its globose grain that may be blackish, brown, reddish, or even whitish; the grain of other Eleusine species/taxa is typically blackish (visible after pericarp removal) (Phillips 1972, 1974, 1995). Other diagnostic characters of cultivated finger millet are the persistent (non-shattering) spikelets and the exposed grains (visible in the gaping floret when ripe); in other Eleusine species/taxa (including subsp. africana), spikelets disarticulate (between the florets) at maturity and grains remain enclosed (Phillips 1972, 1995; Chennaveeraiah and Hiremath 1973; Chen and Phillips 2006). Morphological variation of the inflorescence in finger millet is considerable and, for that reason, several races have been described in the literature (Hilu and De Wet 1976b; De Wet et al. 1984; Dida and Devos 2006; Upadhyaya et al. 2007). The five races more frequently recognized are: “coracana,” “elongata,” “plana,” “compacta,” and “vulgaris”; for descriptions and illustrations of these races, see De Wet et al. (1984) and Dida and Devos (2006). A review on the origin and evolution of the crop is presented in Sect. 7.5 of this chapter.

This taxon was first recognized as a distinct species: E. africana (Kennedy-O’Byrne 1957). Later, “E. africana” was considered a subspecies of E. indica in view of morphological similarities that were thought to result from hybridization between these two taxa [see: “E. indica subsp. africana” in Phillips (1972, 1974) and Lye (1999)]. However, this information was not confirmed as attempts to produce hybrids between the tetraploid “E. africana” and the diploid E. indica resulted in sterile plants (Chennaveeraiah and Hiremath 1974; Hiremath and Salimath 1992). In contrast, as mentioned above, “E. africana” readily hybridizes with finger millet (subsp. coracana) and, for that reason, is regarded here as a subspecies of E. coracana. The tetraploid E. coracana subsp. africana (2n ¼ 36) may be confused with the diploid E. indica (2n ¼ 18). As Kennedy-O’Byrne (1957) pointed out, most of the morphological differences between these two closely related species are differences in size, with the tetraploid (E. coracana subsp. africana) usually being a more robust plant, with longer leaves, thicker and longer spikes, larger bracts, etc., than its diploid ally (E. indica). However, identification can be difficult when these two taxa present similar sizes. Chromosome numbers and genome size clearly distinguish E. coracana subsp. africana from E. indica, but

Fig. 7.1 Types of inflorescence in Eleusine: (a) digitate: E. coracana subsp. coracana (finger millet). (b) racemose: E. multiflora. Photos: (a) S. Neves; (b) G. Swire-Clark and V. Baird

118

cytogenetic analysis is usually not an option for regular identification. Phillips (1972) suggested that the ligule and grain characters may help to differentiate the two species (the ligule is a leaf structure in the shape of a membrane or fringe of hairs that is located at the base of the leaf blade). In E. coracana subsp. africana, the ligule has a clear ciliate fringe and the grain surface is finely granular with low ridges (barely visible), whereas in E. indica, the ligule is a truncate, scarcely ciliolate membrane and the grain has well-marked, oblique ridges/striae (magnification of 50
or more is needed to observe the fine details of grain surface in any Eleusine species) – drawings of the ligules and SEM (scanning electron microscope) photos of the grains of these species are shown in Phillips (1972) and Hilu et al. (1979).

7.2.2 E. floccifolia E. floccifolia is a perennial diploid (2n ¼ 2x ¼ 18) that can be easily recognized within genus by its unique small tuffs of white hairs that are scattered along the smooth margins of the leaves (Phillips 1972, 1974, 1995). The species is found in mid and high altitude in NE Africa and Arabia, and is reported as dominant in heavily grazed pastures of mid-altitude in Ethiopia (Sisay and Baars 2002), possibly a consequence of being unpalatable and avoided by livestock. The species, however, is much used by local populations, particularly in Ethiopia, for making baskets and other household craft articles (Phillips 1995).

7.2.3 E. indica E. indica (goosegrass) is an annual diploid (2n ¼ 2x ¼ 18), which is a major weed problem in many countries in the world, mostly in the tropics, but now affecting temperate countries as well (see Sect. 7.8.2 for further discussion on this topic). E. indica may sometimes be confused with the closely related E. coracana subsp. africana, an annual tetraploid that is usually a more robust plant; see Sect. 7.2.1.2 above for details on how to differentiate these two species.

S.S. Neves

7.2.4 E. intermedia E. intermedia is a perennial diploid (2n ¼ 2x ¼ 18) with very restricted distribution in the uplands of northern Kenya and southern Ethiopia (Phillips 1972). This species was first described as a variety of E. indica but later recognized by Phillips (1972) as a distinct species. E. intermedia can be easily distinguished from E. indica because of its perennial habit (it has a stout rhizome), more laxly arranged spikelets, and 3-nerved lemmas with 1-nerved keel, whereas E. indica is an annual, with tightly arranged spikelets, and 3-nerved lemmas with a 3-nerved keel (Phillips 1972). E. intermedia may be confused with other perennials of the genus, particularly E. jaegeri, but the latter has tough, glabrous leaves with rough margins, whereas E. intermedia has softer leaves, slightly pilose, with generally smooth margins (Phillips 1972). Uses of E. intermedia are unknown, but the species may have some value as a forage grass when other grazing is scarce (Grassland Index 2009).

7.2.5 E. jaegeri E. jaegeri is a perennial diploid (2n ¼ 2x ¼ 20) and the most robust species of Eleusine. It can be easily distinguished from other species in the genus by its robust habit and stiff, pale green leaves, with rough (sawedged) margins; the plants of the species form dense and coarse tussocks (clumps), with its culms (stems) branching to form thick bunches of whitish, overlapping leaf-sheaths (Phillips 1972, 1974, 1995; Grassland Index 2009). E. jaegeri is common in the grasslands of the East African highlands and is used by locals for basket making; however, the species is unpalatable to livestock and is invasive in heavily grazed pastures of high altitude (Grassland Index 2009).

7.2.6 E. kigeziensis E. kigeziensis is a perennial tetraploid (2n ¼ 2x ¼ 38) with a restricted distribution in the East African highlands (Phillips 1972). The inflorescence of E.

7 Eleusine

119

E. multiflora is an annual diploid (2n ¼ 2x ¼ 16) that is easily recognizable within the genus because of its short, oblong-ovate spikes that alternate at the top of the inflorescence axis (Fig. 7.1b). E. multiflora has been sometimes confused with E. tristachya, the only

other species in the genus with similar short and broad spikes (Phillips 1972, 1974, 1995). However, E. tristachya has a digitate inflorescence with (1–)2–3 spikes that are tightly clustered to each other at the top of the axis, whereas E. multiflora usually has three or more spikes that clearly alternate in the axis, the lower ones often separated from each other by a distance of 1 cm or more. Phillips (1972) and Clayton and Renvoize (1986) considered that E. multiflora represented a link between the genus Acrachne and Eleusine, because the species presents some morphological features that appear intermediate between the two genera. The placement of E. multiflora in the genus was also questioned by other authors based on: flavonoid content (Hilu et al. 1978), restriction fragment length polymorphisms (RFLPs) of ribosomal DNA (Hilu and Johnson 1992), genome size when compared to that of other diploids (Mysore and Baird 1997; see also Sect. 7.3), and cytogenetic pattern of ribosomal DNA sites (Bisht and Mukai 2000). On the other hand, isozymes (Werth et al. 1994) and chloroplast DNA restriction site data (Hilu and Johnson 1997) supported the inclusion of E. multiflora in the genus. If any

Fig. 7.2 Phylogeny of the genus Eleusine. Simplified phylogenetic tree of the combined sequences of the nuclear ITS ribosomal DNA and plastid trnT-trnF regions; data from Neves et al.

(2005). The B genome sequences of E. coracana refer only to ITS data. Eleusine kigeziensis (1) corresponds to the accession “Kew 12560”; E. kigeziensis (2) corresponds to “K. Hilu 2505”

kigeziensis resembles that of the annuals E. coracana subsp. africana and E. indica, with which sometimes is confused; but it is distinguished from these taxa by its perennial habit (E. kigeziensis has a short ascending rhizome that is absent in the annuals). E. kigeziensis is separated from other perennials by its slender rhizome (not stout), fairly soft leaves, which are sometimes pilose, but with no tuffs of hairs in the leaf margins, and by its lemmas with a central 3-nerved keel (i.e., the lemma has a central nerve with two lateral, inconspicuous nerves); the midnerve of the lemmas is always simple (1-nerved) in other perennial species (E. floccifolia, E. intermedia, E. jaegeri, and E. jaegeri) (Phillips 1972, 1974, 1995).

7.2.7 E. multiflora

120

doubts remained, recent sequence analyses of plastid and nuclear DNA have unequivocally demonstrated that E. multiflora is a member of the Eleusine lineage (Neves et al. 2005); see the placement of E. multiflora in the phylogeny of the genus in Fig. 7.2.

7.2.8 E. tristachya E. tristachya is an annual diploid (2n ¼ 2x ¼ 18) that is easily recognizable in the genus by its digitate inflorescence with (1–)2–3 oblong spikes (see spike measurements in Table 7.1), which are tightly clustered at the top of the axis, and its neatly arranged spikelets that are perpendicular to the spike axis (Hilu 1980). The species may occasionally be confused with E. multiflora (see Sect. 7.2.7, above). E. tristachya is the only non-African member of the genus, being native to South America, where it is important as a forage grass, particularly in Argentina (Lovisolo and Galati 2007). However, this species has now a much larger distribution and is naturalized in many regions of the world, including parts of North America, Europe, Africa, and Australia (see e.g., Hansen 1980; Hilu 1980, 2003; Sanz Elorza et al. 2001; Ellis et al. 2004; Clayton et al. 2006-onwards; USDA NRCS 2009).

7.2.9 Species of Uncertain Status or Now Placed in Other Genera The list of Eleusine species addressed in this chapter does not include E. semisterilis S.M.Phillips, a species name published based on a single herbarium specimen (holotype) located at the Kew herbarium, in England (Phillips 1972). This specimen presents a number of unusual features for an Eleusine species (abortive spikelet at the end of the spikes and laxly arranged spikelets), and even Phillips (1972) suggested that the atypical inflorescence of the specimen may be the result of anomalous development. All subsequent references to this species (e.g., Clayton et al. 2006-onwards and Grassland Index 2009) continue to be based on that single specimen. Furthermore, attempts by K.W. Hilu to find more material in the Kenyan locality where the type was collected have failed (Werth et al. 1994;

S.S. Neves

Neves et al. 2005). Therefore, there is currently no evidence that this species exists in the wild. Some species that were originally described in Eleusine are now classified in other genera, such as Acrachne and Ochthochloa. For example, Eleusine racemosa B.Heyne ex Roem. and Schult. and Eleusine verticillata Roxb. are presently recognized as Acrachne racemosa (B.Heyne ex Roem. and Schult.) Ohwi; whereas Eleusine compressa (Forssk.) Asch. and Schweinf. ex C.Chr. and Eleusine flagellifera Nees currently correspond to Ochthochloa compressa (Forssk.) Hilu (Hilu 1981; Phillips 1974, 1995).

7.3 Morphological, Classical Genetic and Cytogenetic Studies in the Genus Eleusine From the late nineteenth century until the 1950s, researchers supported the notion that E. indica was the direct ancestor of cultivated finger millet, and that domestication had likely occurred in India, with the crop being later introduced in Africa (e.g.,. Greenway 1944). However, the finding of a new African tetraploid species (“E. africana”), with close resemblance to the crop, led to the review of the theories of origin of finger millet. Kennedy-O’Byrne (1957) was the first to suggest that domestication may have occurred in North East Tropical Africa, with the diploid E. indica given origin to the wild tetraploid “E. africana” by chromosome doubling and, subsequently, to the development of the crop by selection and cultivation of a large grain mutant of “E. africana.” Supporting the close relationship between “E. africana” and E. coracana sensu stricto (s.str.) were the large number of morphological intermediates between these taxa that were found in cultivated fields in Africa (Kennedy-O’Byrne 1957; Mehra 1962). Initially, E. indica was thought as the only contributor to the genome of finger millet (e.g., Greenway 1944; Kennedy-O’Byrne 1957). Later, Mehra (1963a, b), in a comparative morphological analysis of E. coracana (s.str.), “E. africana,” and E. indica, supported the view that E. coracana had originated in Africa from “E. africana” and that the latter had evolved by hybridization of E. indica and another taxon. Subsequent morphological and archeological studies further confirmed the origin of finger

7 Eleusine

millet in East Africa, and that E. coracana subsp. africana is wild finger millet (Harland 1971; Hilu and De Wet 1976a; Hilu et al. 1979). Until the early 1970s, most cytogenetic work in Eleusine concerned chromosome counts [reviewed by Hiremath and Chennaveeraiah (1982)]. Chennaveeraiah and Hiremath (1973, 1974) started a series of comparative cytogenetic studies intended to elucidate the genomic origin and relationships of finger millet with its wild relatives. Chennaveeraiah and Hiremath (1973) analyzed chromosome pairing during meiosis of a hybrid between the annual E. tristachya and the perennial E. floccifolia, two diploid species with the same chromosome number (2n ¼ 18). These authors observed good pairing between the chromosomes of these species, but the resulting hybrid was completely sterile. Chennaveeraiah and Hiremath (1974) used similar cytogenetic methods to analyze genome homology in two sets of hybrids, E. coracana s.str.
“E. africana” and E. coracana s.str.
E. indica; these authors demonstrated that the genomes of “E. africana” and E. coracana (s.str.) are basically the same and that these two taxa are allotetraploids with AABB genomes. Chennaveeraiah and Hiremath (1974) also suggested that, due to lack of homology, E. indica may not be a genome donor to finger millet, but these authors were cautious about their own results, which were based on analyses of a single hybrid and a small number of cells. After chloroplast DNA evidence (see Sect. 7.4.1) demonstrated that E. indica is the “A” (maternal) genome donor to the crop (Hilu 1988), new cytogenetic analyses (chromosome pairing in hybrids) were carried out by Hiremath and Salimath (1992). These authors performed several crosses between tetraploids (E. coracana s.str. or “E. africana”) and diploids (E. indica, E. floccifolia, E. intermedia, E. multiflora, and E. tristachya) to obtain triploid hybrids (found to be sterile). Hiremath and Salimath (1992) then confirmed that the genome of E. indica has strong homology to (half of) the chromosomes of E. coracana (s.str.), contradicting the initial finding of lack homology between these species (Chennaveeraiah and Hiremath 1974). Hiremath and Salimath (1992) thus confirmed that E. indica is the “A” genome donor of the crop but found no evidence for a possible “B” genome donor among the species they analyzed. In a subsequent study, Salimath et al. (1995b) attempted to produce hybrids between all the diploid species of Eleusine, but crosses were only successful in two combinations:

121

E. indica
E. floccifolia and E. tristachya
E. indica. The resulting hybrids (F1) were analyzed cytogenetically, with demonstration of strong homology among the genomes of E. indica, E. tristachya, and E. floccifolia. However, in spite of the genomic similarity of the parental species, all the hybrids obtained were once again completely seed sterile (Salimath et al. 1995b). More recently, a similar study of hybridization between diploid species was performed by Devarumath et al. (2005), the main difference being the inclusion of E. intermedia and the production of F1 hybrids between this species and E. indica, E. tristachya, and E. floccifolia. This study reported high degree of homology among the four species but, as in earlier studies, reported a lack of fertility in the F1 hybrids.

7.3.1 FISH and GISH Analyses Bisht and Mukai (2000) used for the first time fluorescence in situ hybridization (FISH) in Eleusine species to map and compare the distribution of ribosomal DNA (rDNA) sites in their genomes. This work further confirmed the similarity of the genomes of cultivated E. coracana (s.str.) and wild “E. africana” as the two taxa presented similar location and number of 18S5.8S-26S and 5S rDNA sites on their chromosomes. Bisht and Mukai (2000) also found some similarities in the patterns of rDNA sites of E. indica and E. floccifolia when compared to those seen in E. coracana (s.str.) and “E. africana” and suggested that E. indica and E. floccifolia may be the two genome donors to the crop. In another study, Bisht and Mukai (2001a) used genomic in situ hybridization (GISH) techniques to hybridize genomic DNA of six diploid species of Eleusine (E. indica, E. floccifolia, E. tristachya, E. internedia, E. multiflora, and E. jaegeri) to the chromosomes of the tetraploid E. coracana s.str. (finger millet). The genomic DNA of E. multiflora and E. jaegeri did not produce any hybridization signals in the chromosomes of E. coracana s.str., suggesting that the two species are not closely related to finger millet and may be ruled out as genome donors (Bisht and Mukai 2001a). This study also indicated a close genomic relationship among four diploids (E. indica, E. floccifolia, E. tristachya, E. internedia) and that the genomes of E. indica and E. tristachya are very similar, with nearly identical patterns of hybridization

122

to the chromosomes of E. coracana s.str. (Bisht and Mukai 2001a). In a separate publication, Bisht and Mukai (2001b) reported the results of a similar GISH study using the four diploids they previously found to be closer to finger millet but whose genomic DNA was now hybridized to the chromosomes of “E. africana” (wild finger millet). The two studies (Bisht and Mukai 2001a, b) reported similar hybridization patterns between the genome of the diploids and the chromosomes of E. coracana s.str. and “E. africana,” supporting the view that these two tetraploids are highly similar at the genomic level. Taking into account the results of double GISH on chromosomes of E. coracana (of the two subspecies) using the genomic DNA of E. indica and E. floccifolia, Bisht and Mukai (2001a, b) suggested that these two species are the genome donors to finger millet. However, as discussed in Sect. 7.4.1, recent sequence data contradicts the hypothesis that E. floccifolia is a genome donor to finger millet (Neves et al. 2005). The contradiction between the results of GISH and sequence analyses is only apparent, because GISH results cannot be reliably used for phylogenetic inference (Neves et al. 2005). GISH techniques are certainly useful for assessment of chromosome genetic similarity, but genomic comparisons can only be performed using two or three samples (species) at a time, and as the results cannot be expressed quantitatively (no measurements of genetic distance), they are problematic to interpret at the phylogenetic level, particularly among closely related species that naturally have some degree of genomic similarity (Neves et al. 2005). Bisht and Mukai (2001a) obtained almost identical results of double GISH with E. coracana and the species pairs, E. indica–E. floccifolia and E. tristachya–E. floccifolia. If only the GISH results were taken into account, E. tristachya could as easily be considered one of the genome donors as E. indica (Neves et al. 2005). But, of course, this was not suggested by Bisht and Mukai (2001a) who relied on the results of earlier studies (e.g., Hilu 1988) to exclude E. tristachya as a potential genome donor.

7.3.2 Genome Size Analyses of nuclear DNA content in the species of Eleusine were first carried by Hiremath and Salimath

S.S. Neves

(1991) using Feulgen microspectrophotometry. Later, Mysore and Baird (1997), using laser flow cytometry, determined that there had been an overestimation on nuclear DNA contents for most species analyzed by Hiremath and Salimath (1991). The new genomic sizes determined by Mysore and Baird (1997) for the Eleusine species/taxa were as follows (1 pg ¼ 980 Mbp ): E. coracana subsp. coracana, 2C ¼ 3.36–3.87 pg (2n ¼ 36); E. coracana subsp. africana, 2C ¼ 3.34 pg (2n ¼ 36); E. indica, 2C ¼ 1.61–1.76 pg (2n ¼ 18); E. tristachya, 2C ¼ 1.51 pg (2n ¼ 18); E. floccifolia, 2C ¼ 2.0 pg (2n ¼ 18); E. multiflora, 2C ¼ 2.65 pg (2n ¼ 16); and E. jaegeri, 2C ¼ 1.90 pg (2n ¼ 20). Mysore and Baird (1997) highlighted the fact that, in spite of having the smallest chromosome number in the genus, E. multiflora has a considerably larger genome than any of the other diploids. Although the genome size of the Eleusine species is small when compared with that of many other plants, including other grasses and important crops (see e.g., Bennett and Leitch 2005), it is nonetheless considered relatively large from a genetics or genomics point of view (Dida et al. 2007). Nevertheless, a genetic map for finger millet has already been produced (Dida et al. 2007), demonstrating that genome size is not an impediment for detailed understanding of the genome of the Eleusine species.

7.4 Evolution of the Genus Eleusine: Molecular Evidence 7.4.1 Eleusine Evolution: The DNA Data Previous work in Eleusine using morphology (Mehra 1962, 1963a; Hilu and De Wet 1976a), phytochemistry (Hilu et al. 1978), and cytogenetics (Chennaveeraiah and Hiremath 1974) had demonstrated high similarity between cultivated finger millet (E. coracana s.str.) and wild “E. africana,” that these two taxa hybridize in nature, and that the crop was probably an allotetraploid, with the diploid E. indica a possible parental species. However, Chennaveeraiah and Hiremath (1974) had raised some questions about the status of E. indica as a genome donor to the crop. With the advent of molecular biology techniques that could be easily used by plant scientists, it became possible to

7 Eleusine

use data directly from DNA to address questions on relationships and genome homology that had remained unsolved by more classical approaches. The first DNA analyses in the genus were carried out by Hilu (1988), who used chloroplast DNA (cpDNA) restriction fragment length polymorphism (RFLP) to examine the relationships among cultivated and wild finger millet and two other species in the genus: E. indica and E. tristachya. Hilu (1988) confirmed the close genetic relationship between E. coracana subsp. coracana and its putative wild progenitor E. coracana subsp. africana and demonstrated that E. indica is the maternal (“A”) genome donor of finger millet, with the three taxa sharing a common chloroplast genome (maternally inherited in most plants). Later, Hilu and Johnson (1992) used RFLPs to study the variation of ribosomal DNA (rDNA) among six species of Eleusine; the rDNA data confirmed the close relationship among the two subspecies of E. coracana and E. indica, with E. tristachya being the next closest species to the crop and E. multiflora the most distant. Similar results were obtained by Hilu (1995) using random amplified polymorphic DN (RAPDs), although the samples of wild finger millet (E. coracana subsp. africana) grouped closer to E. indica than to the cultivated cereal. Salimath et al. (1995a) used three different DNA marker techniques, RFLP, RAPD, and inter-simple sequence repeats (ISSR), to analyze five species of Eleusine; their results supported the close affinity of E. coracana, E. indica, and E. tristachya, with E. floccifolia being a more distant relative. Hilu and Johnson (1997) extended earlier restriction site analyses of cpDNA in the genus to demonstrate the monophyly of Eleusine, with the inclusion of E. multiflora, a species whose placement in the genus had been previously questioned (see Sect. 7.2.7 for more details). Monophyly of Eleusine has been conclusively demonstrated by analyses of DNA sequences of the trnT–trnF (chloroplast genome) and the internal transcribed spacer (ITS) region of nuclear rDNA in the eight species that constitute the genus (Neves et al. 2005). A simplified molecular phylogenetic tree representing data and results of Neves et al. (2005) is shown in Fig. 7.2. This study further confirmed the allotetraploid origin of finger millet, since two putative ITS homeologs, corresponding to the “A” and “B” genomes, were detected in all the accessions of E. coracana

123

(subsp. coracana and subsp. africana), with the diploid E. indica sharing an ITS locus “A” sequence closer to that found in both wild and cultivated finger millet. The ITS locus “B” sequence of E. coracana was not found in any other Eleusine species. As sequences of the two ITS loci are clearly nested within the Eleusine lineage (see Fig. 7.2), the two genome donors must be members of the genus; this leads to the conclusion that the “B” genome donor may now be extinct (Neves et al. 2005). E. indica shares with E. coracana highly similar chloroplast sequences, confirming that the former is the maternal genome donor of the crop; the high sequence similarity of these taxa also suggests a relatively recent allopolyploidization event (Neves et al. 2005). However, this study contradicts the hypothesis of Bisht and Mukai (2000, 2001a, b) that E. floccifolia is the second genome donor. The ITS sequences obtained by Neves et al. (2005) for E. floccifolia are very distinct from those of the putative “B” genome donor (see Fig. 7.2). Earlier, Salimath et al. (1995a) had also considered E. floccifolia, an unlikely “B” genome donor based on DNA fingerprinting analyses. The results of Neves et al. (2005) also show good support for a “CAIK” group (E. coracana subsp. coracana and subsp. africana, E. indica, and E. kigeziensis), with the South American E. tristachya as its sister lineage. The perennials E. jaegeri and E. multiflora are shown as the earliest diverging lineages in the genus and the most distant to the crop (Neves et al. 2005). In this study, several Eleusine species (for which more than one accession was sequenced) were shown as monophyletic taxa, with the exception of E. kigeziensis, where an accession from Uganda (Kew 12560; see Fig. 7.2) did not group with the others (KH2505, KH2506, and KH2507) (see Neves et al. 2005). Hiremath and Salimath (1991) and Bisht and Mukai (2002) suggested an allotetraploid origin for E. kigeziensis (2n ¼ 38), possibly from hybridization of the diploids E. indica (2n ¼ 18, x ¼ 9) and E. jaegeri (2n ¼ 20, x ¼ 10). However, in spite of extensive clone sequencing, Neves et al. (2005) only found one type of ITS sequence in E. kigeziensis, raising the possibility of this species being an autopolyploid. E. kigeziensis belongs to a clade (lineage) of species that has x ¼ 9 as basic chromosome number (E. floccifolia, E. intermedia, E. tristachya, E. indica, and E. coracana; see Fig. 7.2). If E. kigeziensis is an autopolyploid, it may have originated from polyploidization of an ancestral species with x ¼ 9 that subsequently

124

underwent chromosome rearrangement and/or aneuploidy, with a slight increase on chromosome number.

S.S. Neves

7.5 The Origin of Finger Millet: Current Understanding Cultivated finger millet (E. coracana subsp. coracana)

7.4.2 Phylogenetic Placement of Eleusine is believed to have been domesticated in East Africa, Among Other Grasses probably in the highland region that extends from The placement of Eleusine in subfamily Chloridoideae is undisputed (e.g., Clayton and Renvoize 1986; Hilu and Alice 2001), but its exact position in the chloridoid clade or its relationships to other chloridoid genera remain uncertain (Hilu and Alice 2001; Columbus et al. 2007; Bouchenak-Khelladi et al. 2008). Due to some morphological similarities, Eleusine was previously considered a close ally of Acrachne and Dactyloctenium (Phillips 1972, 1982, 1995; Clayton and Renvoize 1986). However, current molecular data indicates that Dactyloctenium (Hilu and Alice 2001; Roodt-Wilding and Spies 2006; Columbus et al. 2007) and Acrachne (Neves et al. 2005) are not closely related to Eleusine. As mentioned in Sect. 7.2.9, the genus Ochthochloa now includes a species (O. compressa ¼ “E. compressa”) previously placed in Eleusine. But Ochthochloa and Eleusine are morphologically quite distinct (Hilu 1981; Phillips 1982, 1995); Hilu et al. (1978) also reported a very distinct flavonoid composition for “E. compressa”. Neves et al. (2005) attempted to obtain seeds for O. compressa for cultivation and DNA sequencing, but all accessions received had been incorrectly identified (examples like this highlight the importance of confirming the identification of plants before analyses are carried out). As there is no sequence data yet for Ochthochloa, its relationship to Eleusine still needs be assessed. In a numerical analysis of morphological data, Phillips (1982) identified a group comprised by Eleusine, Acrachne, Dactyloctenium, Coelachyrum, and Sclerodactylon. Coelachyrum is another grass genus with ornamented grain and free pericarp (see also introductory text in Sect. 7.2). However, sequence data indicates that Coelachyrum is not closely related to Eleusine (Ingram and Doyle 2004). Currently, there is no sequence data available in GenBank for Sclerodactylon, but this genus is morphologically quite distinct from Eleusine. In summary, current evidence suggests that Eleusine is a monophyletic group that is relatively isolated from other chloridoid grass genera (Hilu and Johnson 1997; Neves et al. 2005).

Ethiopia to Uganda, where populations of wild finger millet (subsp. africana) are particularly abundant (Harland 1971; Hilu and De Wet 1976a, b; Hilu et al. 1979; De Wet et al. 1984); E. coracana subsp. africana rarely extends to the lowlands of East Africa, and hence the suggestion of an upland origin for the crop (Hilu and De Wet 1976b). The oldest archeological record for finger millet (well preserved grains obtained from a prehistoric site in Ethiopia) has been tentatively dated to the third millennium BC (Hilu et al. 1979), and this suggests that domestication of the crop may have occurred in East Africa at about 5,000 years ago or earlier (De Wet et al. 1984). The crop was subsequently introduced to India, possibly during the second or first millennium BC, where a secondary center of diversity was developed as a consequence of prolonged artificial and natural selection (Hilu and De Wet 1976a, b; De Wet et al. 1984). Conclusive data from morphology, hybridization studies, cytogenetics, and molecular biology indicate that the wild E. coracana subsp. africana (native in Africa and originally absent in India) is the direct progenitor of cultivated finger millet (Chennaveeraiah and Hiremath 1974; Hilu and De Wet 1976a; Hilu and Johnson 1992; Dida et al. 2008). Both cultivated and wild E. coracana have been confirmed as allotetraploids, with the genome notation given as AABB (Chennaveeraiah and Hiremath 1974; Hilu 1988; Hiremath and Salimath 1992; Werth et al. 1994; Bisht and Mukai 2001a, b; Neves et al. 2005; Dida et al. 2007). E. coracana subsp. africana (2n ¼ 36) is an allotetraploid that resulted from natural hybridization between the diploid E. indica (2n ¼ 18), considered the maternal (“A”) genome donor (Hilu 1988; Hiremath and Salimath 1992; Bisht and Mukai 2001a, b), and an unknown diploid (probably also 2n ¼ 18), followed by polyploidization. Sequences of both “A” and “B” genomes have been identified in finger millet (see Sect. 7.4.1); the “A” genome sequences are highly similar to those of E. indica, but the “B” genome sequences have not been identified in any of the known Eleusine species (Neves et al. 2005); see also Fig. 7.2. For that reason, the “B”

7 Eleusine

genome progenitor of E. coracana may correspond to a species that is now extinct (Neves et al. 2005).

7.6 Genome Analysis in Eleusine: Molecular Tools and Genomic Resources 7.6.1 Molecular Tools: SSRs, AFLPs, ESTs, and Others Several fingerprinting techniques have been used to analyze the species of Eleusine. RFLPs were the first DNA markers to be used in the genus (Hilu 1988; Hilu and Johnson 1992; Salimath et al. 1995a) and are still used in current studies, namely for genetic mapping (Dida et al. 2007). The RFLP technique (e.g., Jones et al. 1997) is a reliable technique that produces codominant markers but is time-consuming and may sometimes show little or no variation among closely related individuals (accessions), highly inbred lines, or in germplasm collections with low genetic diversity. This limitation was evident on analyses of different accessions of E. coracana subsp. coracana, where RFLPs revealed very low levels of polymorphism (Muza et al. 1995; 23 lines identical out of 26) or no variation at all (Salimath et al. 1995a; Dida et al. 2007). RFLPs are potentially more useful for interspecific analyses, but, in those cases, DNA sequencing of variable regions of the plastid or nuclear genome will most likely be preferred, mainly due to the relative simplicity of DNA sequencing when compared to RFLP analyses. RAPDs have also been used to study the genetic diversity and species relationships in the genus (Hilu 1995; Salimath et al. 1995a; Babu et al. 2007). RAPD analysis produces dominant markers and is considered inexpensive and fast but has poor reproducibility and is sensitive to experimental conditions; for those reasons, RAPDs are generally considered an unreliable fingerprinting technique (e.g., Salimath et al. 1995a; Jones et al. 1997; Harris 1999). RAPDs are rarely used nowadays, especially considering that far more reliable markers, such as ISSRs or microsatellites (see below), have become relatively inexpensive and easy to use for analyses of diversity in plant populations.

125

Salimath et al. (1995a) used ISSRs in the genus and found the technique to be particularly promising for analysis of plant diversity, including differentiation of accessions within the same species of Eleusine. The ISSR technique produces dominant markers and is considered useful due to its low cost, good reproducibility, and its ability to detect high levels of variation at the population level (e.g., Salimath et al. 1995a; Wolfe et al. 1998). ISSRs may be particularly useful for a first assessment of genetic diversity in species where microsatellites have not yet been developed. Breviario et al. (2007) have recently used a modified version of the TBP (tubulin based polymorphism) method for analyses of diversity and species relationships in Eleusine. The new “combinatorial TBP” (cTBP) technique was evaluated as reliable, reproducible, fast, and easy to use, and able to produce enough polymorphisms to discriminate species and varieties (Breviario et al. 2007). The cTBP technique was able to detect variability in individual accessions of finger millet but failed to detect variation in other crops, suggesting that the number of polymorphisms generated by the technique may be insufficient for analyses of highly inbred species (Breviario et al. 2007) or closely related lines, limiting their applicability for general germplasm characterization. The amplified fragment length polymorphism (AFLP) technique, which also produces dominantly expressed markers, has been extensively used in plants; it is particularly useful in the absence of genomic sequence information, when no specific DNA markers (like microsatellites) have been developed, and for the production of very large numbers of polymorphic markers, suitable for discrimination and unambiguous identification in populations and germplasm collections, even when genetic variability is low (Meudt and Clarke 2007). AFLPs were recently used in Eleusine as one of the sources of markers for the construction of the genetic map of E. coracana (Dida et al. 2007). SSRs (simple sequence repeats) or microsatellites (see e.g., Goldstein and Schlo¨tterer 1999) are currently the most useful DNA markers for studies of genetic diversity and structure of natural and cultivated populations, mainly due to their high variability, abundance in the genomes, and the reliability, ease of use, and affordability of the technique. The only limiting factor of the use of simple sequence repeats (SSRs) is the

126

need for development of specific primers for each new species or group intended to be studied. However, numerous SSRs are already available for E. coracana (subsp. coracana and africana), 45 SSRs of which have been successfully used for the construction of the genetic map of the species (Dida et al. 2007) and analysis of genetic diversity and population structure (Dida et al. 2008). Additional SSRs markers are available from the work of Wang et al. (2005), who effectively transferred more than 100 SSRs from major cereal crops (wheat, rice, maize, and sorghum) to other “minor” grass species, including E. coracana, with an average of 58% success rate of cross-genus amplification in the case of finger millet. Wang et al. (2005) have also analyzed the level of polymorphism detected with the new SSRs and, in the case of finger millet, polymorphism was detected within the species in 21% of the transferred SSRs. These results suggest that many of the SSRs obtained from major grasses with well studied genomes can be successfully amplified in other grass genera and that a significant portion of these markers will show adequate variation for characterization and evaluation of diversity of populations in new grass species (Wang et al. 2005). The high success rate of cross-genus amplification detected in the work of Wang et al. (2005) also suggests that transferring SSRs between closely related grass species (for instance, from those available for E. coracana to other Eleusine species) should be relatively easy with expected high levels of success. Dida et al. (2008) have already used the SSRs of E. coracana in samples of E. indica and E. kigeziensis, which are the two closest species to E. coracana (see Sect. 7.4.1 and Fig. 7.2). Direct use of the available E. coracana SSRs in more distantly related species, like E. multiflora or E. jaegeri, may not be as straightforward as with closely related species/taxa of the so-called “CAIK” group (“coracana”, “africana”, “indica,” and “kigeziensis”). Nevertheless, it is expected that crossspecies amplification will be possible for a number of those markers. There has also been great progress in the development of expressed sequenced tags (ESTs) in Eleusine. Buell (2009) reports that 1,749 ESTs are already available for E. coracana. ESTs represent transcribed fragments of the genome, which are particularly useful for designing new genetic markers (e.g., SSRs) and for studies of functional genomics (Buell 2009). ESTs were also used in the

S.S. Neves

construction of the genetic map of E. coracana (Dida et al. 2007).

7.6.2 The Genetic Map of E. coracana and Comparative Genomics Comparative genetic mapping has revealed conserved gene order (colinearity) among many grass species, but studies at the gene level have demonstrated that microcolinearity of genes is less conserved and that smallscale rearrangements and deletions have disturbed the microcolinearity that was initially predicted, even between relatively close species, such as sorghum and maize (Keller and Feuillet 2000) – Sorghum and Zea are both members of the Andropogoneae in subfamily Panicoideae, see e.g., Bouchenak-Khelladi et al. (2008). After it was realized that the level of synteny between grass genomes was lower than anticipated, it became clear that genomic platforms needed to be established for each crop or species of interest (Powell and Langridge 2004). A first step in the development of molecular tools that could assist the breeding (marker-assisted selection) of any plant crop is the construction of a genetic linkage map. Fortunately, the molecular knowledge of the Eleusine genome has progressed considerably in recent years, and the first genetic map of E. coracana (subsp. coracana and subsp. africana) is already available (Dida et al. 2007). This map was generated using a cross between E. coracana subsp. coracana (a cultivar from Nepal: Okhale-1) and its wild progenitor E. coracana subsp. africana (an accession from Kenya: MD-20). An intersubspecific cross had to be used for development of this linkage map due to the low levels of variation detected in lines of cultivated finger millet (Dida et al. 2007); see also Sect. 7.6.3, below. Comparative analysis of colinearity between the finger millet and rice genomes have been recently completed (Srinivasachary et al. 2007). These authors reported that, other than the expected rearrangements to explain the difference on chromosome numbers between finger millet (2n ¼ 36) and rice (2n ¼ 2x ¼ 24), only 10% of the markers were found in non-syntenic positions and that the finger millet and rice genomes have remained relatively conserved since the divergence of the two lineages

7 Eleusine

from a common ancestor around 60 million years ago (Srinivasachary et al. 2007).

7.6.3 Genetic Diversity and Structure of Populations Several molecular and chemical studies that included multiple accessions or lines of cultivated finger millet (E. coracana subsp. coracana) have consistently reported low levels of diversity in the crop. Analyses of flavonoid content presented no qualitative differences in the cultivated races of finger millet (Hilu et al. 1978). This uniformity in “coracana” accessions was also evident in restriction fragment analyses of rDNA (Hilu and Johnson 1992). Werth et al. (1994) reported an isozyme genotype that was identical in nearly all accessions of finger millet (19 were examined). Similar results were obtained by Muza et al. (1995) in RFLPs analyses (23 lines out of 26 were identical). In contrast, much greater variation was found in both the rDNA pattern (Hilu and Johnson 1992) and isozyme genotypes (Werth et al. 1994) of accessions of wild finger millet (E. coracana subsp. africana). Salimath et al. (1995a), using multiple DNA markers (RFLPs, RAPDs, and ISSRs), also detected very low levels of genetic variation in the crop (17 accessions from India, Nepal, and several African countries), but they were still able to discriminate each of the lines with the few polymorphisms found. Salimath et al. (1995a) did not include any samples of wild finger millet, but they found more variation in two accessions of E. floccifolia (another wild relative) than in all the 17 lines analyzed for finger millet. The pattern of low genetic diversity in domesticated finger millet was further confirmed by Dida et al. (2007). However, Dida et al. (2007) highlighted the fact that traditional finger millet breeding have been mostly carried out by selection of pure lines from local landraces, suggesting that the material that have been examined for genetic diversity may represent only a fraction of the total variation of the landraces. Various authors have also suggested that the narrow genetic pool of cultivated finger millet may be a natural consequence of domestication: the crop resulted from selection of a small subset of the diversity of wild populations and, subsequently, the gene pool remained restricted by limited introgression with wild relatives due to the highly inbred (self-pollinating) nature of the crop (Hilu and

127

Johnson 1992; Werth et al. 1994; Salimath et al. 1995a; Dida et al. 2007). In contrast with the low genetic diversity detected in cultivated finger millet, large morphological variation has been described in the crop (Hilu and De Wet 1976b; De Wet et al. 1984; Upadhyaya et al. 2007; Dida et al. 2008). High diversity has also been reported in protein and calcium contents in the screening of 36 genotypes of finger millet (Vadivoo et al. 1998). The morphological and chemical data indicate that genetic variation exists in the crop and that it is necessary to considerably increase the number of DNA markers and accessions to detect the existing variation in the genome of the crop. This is exactly what Dida et al. (2008) have done with the multiple SSRs that had been previously designed for the construction of the genetic map of E. coracana (Dida et al. 2007). Dida et al. (2008) carried out a comprehensive genotypic analysis using 45 SSRs and 96 accessions of Eleusine: 79 accessions of E. coracana subsp. coracana; 14 of E. coracana subsp. africana; two of E. indica and one accession of E. kigeziensis, the latter used as an outgroup for analyses. Dida et al. (2008) described three fairly distinct subpopulations in E. coracana that essentially corresponded to: (1) subsp. africana, (2) subsp. coracana from Africa, and (3) subsp. coracana from India. The limits of these subgroups are not clear, as it would be expected in populations that naturally interbreed. Dida et al. (2008) also observed clear evidence of gene flow between wild and cultivated African subpopulations. Further genetic diversity analyses are underway in both cultivated and wild accessions (lines) of finger millet, and the new study will include screening of important agronomic traits, such as grain chemical composition and resistance to blast disease (Dida and Devos 2006; Dida et al. 2008). There has been no genetic screening of populations in any other Eleusine species. But E. coracana SSRs can be successfully used in E. indica and E. kigeziensis (Dida et al. 2008).

7.7 Conservation of Genetic Resources In the past few decades, national breeding programs in India and Africa and several international initiatives have contributed to a considerable expansion

128

of plant genetic collections (both ex situ and in situ). As a result, a very large number of germplasm accessions of finger millet are currently preserved and are being characterized for future use in breeding programs, namely at International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) (Upadhyaya et al. 2008). The ICRISAT genebank at Patancheru, India, holds more than 5,000 accessions of finger millet (mean seed viability of 96.2%), and a safety back-up of 4,580 of these accessions are also preserved at the ICRISAT Regional Genebank at Niamey, Niger (Upadhyaya et al. 2008). ICRISAT has also entered an agreement with the Royal Norwegian Ministry of Agriculture and Food to deposit many thousands of germplasm seed samples, including those of finger millet, at the Svalbard Global Seed Vault, Norway (Upadhyaya et al. 2008). Funding provided by the McKnight Foundation (Minnesota, USA) (see also Sect. 7.9) has also helped to increase germplasm collections of finger millet in Africa, namely of local landraces in Ethiopia, Uganda, and Kenya. In a recent project report, it is referred that, during more than three decades, National Semi-Arid Resources Research Institute (NaSARRI), in Uganda, maintained the largest collection of finger millet in Eastern and Southern Africa, with over 2,500 accessions. However, this collection has diminished considerably due to poor conditions of preservation, diseases, pests and civil unrest, with only 965 accessions of finger millet remaining at the NaSARRI genebank (McKnight Foundation Project No. 06-448, Annual Progress Report 2008). A more recent report of the same project indicates that those 965 accessions have now been planted and are being evaluated for agronomic traits such as resistance to blast and drought and that nearly 120 new finger millet accessions have been collected in Uganda, including 19 local landraces (McKnight Grant No. 06448, Annual Progress Report 2009). It is clear that valuable accessions of cultivated finger millet are being preserved, but attention should also be given to collections of wild relatives of the crop, including wild finger millet (E. coracana subsp. africana) and the close ally E. kigeziensis; accessions of the latter continue to be very rare in germplasm collections.

S.S. Neves

7.8 Problems and Limitations of the Eleusine Species 7.8.1 Diseases and Pests Data on diseases and pests is available mainly for cultivated E. coracana. However, finger millet seems to be little affected by diseases and insects, with the exception of blast disease (National Research Council 1996). Blast disease is caused by the fungus Magnaporthe grisea (T.T. Hebert) M.E. Barr, a species closely related to M. oryzae B.C. Couch (Couch and Kohn 2002), the causal agent of rice blast disease, responsible for crop losses in rice of up to 30% a year (Skamnioti and Gurr 2009) – the name Pyricularia grisea (Cooke) Sacc., commonly used in the literature, corresponds to an “anamorph” (different stage in the life cycle) of M. grisea (see e.g., Couch and Kohn 2002). Naylor et al. (2004) highlighted that the main problems of the finger millet are its susceptibility to blast disease and drought and that blast disease may reduce the yields of the crop by 35% or more. For that reason, Naylor et al. (2004) recommends considerable investment in the development of molecular markers and genetic maps, and the use of marker-assisted selection (MAS) tools, which could facilitate the selection of blast-resistant varieties of finger millet. A successful breeding program may increase yields in finger millet by at least 15% (on half of the area of cultivation), representing more than US$ 38 million of gross annual benefits (Naylor et al. 2004). Dida and Devos (2006) have already reported that, in a small-scale study of diversity of E. coracana populations (in progress), they have already detected accessions of E. coracana subsp. africana with good levels of resistance to the blast fungus. Such blast-resistant accessions could be very useful in breeding programs. It is likely that other diseases affect Eleusine species, but information is lacking. A recent report indicates that the wheat streak mosaic virus (WSMS), which can cause severe crop losses in several cereal crops, has been identified in weed populations of E. tristachya in Australia (Ellis et al. 2004). The WSMS is transmitted by a mite species, but the presence of the disease in weeds such as E. tristachya, which is common in pastures in Australia, is a cause of

7 Eleusine

concern as it may facilitate viral infection of neighboring crops.

7.8.2 E. indica: An “Intractable” Weed E. indica or goosegrass (see Table 7.1, for other common names) is considered one of the worst weeds in the world, with large distribution in the tropics, but now affecting temperate countries as well, and reported as a problematic weed for 46 different crop species in more than 60 countries (Holm et al. 1977); it is also one of the ten most important herbicide-resistant weeds (Basu et al. 2004; Beckie 2007). One of the reasons of the reproductive success of goosegrass is its capability of producing large quantities of seeds; a single plant has been reported to produce up to 140,000 seeds (Lee and Ngim 2000). As a result, farmers often rely on intensive use of herbicides to control this weed (Lee and Ngim 2000). Goosegrass has evolved or had pre-existing resistance against a number of important herbicides, such as dinitroanilines and inhibitors of acetohydroxyacid synthase, acetyl-CoA carboxylase and acetolactate synthase, which have been used to control the spreading of this and other major weeds among crops (Baerson et al. 2002; Ng et al. 2004a). The genetics and molecular mechanisms involved in dinitroaniline resistance in E. indica have been extensively studied (e.g., Yamamoto and Baird 1999; Zeng and Baird 1999) and are reviewed by Anthony and Hussey (1999). Goosegrass has also been very problematic as a turfgrass weed but controllable, for many years, by applications of the herbicides metribuzin plus MSMA (monosodium methanearsonate) (Brosnan et al. 2008). However, new metribuzin-resistant goosegrasss biotypes have now been documented in bermudagrass turf (Cynodon spp.), which were also resistant to another herbicide combination: simazine plus MSMA (Brosnan et al. 2008). Nevertheless, the metribuzinresistant biotypes were susceptible to herbicides such as glyphosate and foramsulfuron, which can still be used to control goosegrass in bermudagrass turf (Brosnan et al. 2008). Glyphosate is the active ingredient in the broad spectrum and widely used herbicide known as “Roundup.” Glyphosate acts by inhibition of the enzyme EPSPS (5enolpyruvylshikimate-3-phosphate synthase) that cata-

129

lizes a critical step in the metabolism of plastids, being thus toxic to plants (and some microorganisms) but not animals (Baerson et al. 2002, and references therein). Since the introduction in 1996 of “Round-up Ready” transgenic crops from the Monsanto company, these crops have been widely used and grown under single usage or repeated applications of glyphosate (Sandermann 2006). Historical use and initial tests suggested that glyphosate was a low risk herbicide for the evolution of weed resistance (Baerson et al. 2002). However, the repeated use of the herbicide led to the appearance (selection) of new glyphosate-resistant biotypes in two weed grasses, Lolium rigidum Gaudin (rigid ryegrass) and E. indica (Lee and Ngim 2000). The goosegrass glyphosate-resistant biotype, initially found in Malaysia, was determined to be 8- to 12-fold more resistant to the herbicide than susceptible goosegrass, rendering glyphosate completely ineffective in the control of the weed (Lee and Ngim 2000). The source of resistance to glyphosate in the Malaysian biotypes was later identified as a single nucleotide substitution in the gene encoding the EPSPS enzyme (target for the herbicide) that resulted in an amino acid substitution (proline for either serine or threonine at position 106), with the altered enzyme being no longer affected by the presence of glyphosate (Baerson et al. 2002; Ng et al. 2003). Resistance of E. indica to the herbicide glyphosate was determined to be inherited as a single, nuclear and incompletely dominant gene (Ng et al. 2004a). However, Owen and Zelaya (2005) and Ng et al. (2004b) have evidence that indicates that there is a second resistance mechanism non-targeted to EPSPS, meaning that E. indica has developed alternative resistant mechanisms to glyphosate. This is a cause of major concern given that glyphosate resistance is the dominant transgenic trait being used, corresponding to 114 million ha of transgenic crops in 23 countries (Green 2009). The widespread use and often sole reliance on glyphosate has stimulated the evolution by selection of resistant weeds (Green 2009). Long-term sustainability of glyphosate is thus at risk and, for that reason, companies are currently developing crops that combine glyphosate resistance with resistance to herbicides with other modes of action; the new transgenic crops will give farmers more options in the fight against weeds (Green 2009). Herbicide-resistant weeds will continue to evolve, and only judicious use of herbicides, including the alternation of herbicides with different modes of

130

action, integration of cultural and mechanical weed management methods, as well as rotation of transgenic crops with conventional cultivars, will need to be followed if widespread selection of glyphosate- or other type of herbicide-resistant weeds is to be prevented (Lee and Ngim 2000). E. indica will likely continue to be a major weed problem for many years. But its remarkable capability of resistance to adverse environmental conditions may have some positive applications. The use of E. indica in phytoremediation of petroleum-polluted soil is currently under investigation (Merkl et al. 2005; Wang et al. 2008).

7.9 Future Prospects and Recommendations There has been recently great progress in the genomic knowledge of Eleusine, and the construction of a genetic map for E. coracana (Dida et al. 2007) was a crucial step for future breeding. Comparative analysis of colinearity between the finger millet and rice genomes determined that 90% of markers were found in syntenic positions (Srinivasachary et al. 2007), which suggests that other better known grass genomes can provide invaluable information for future breeding in finger millet. In the case of this crop, yields are limited mainly by blast disease and drought (Naylor et al. 2004). The mechanisms of blast resistance and the location of resistance genes are better understood in rice (see e.g., Hittalmani et al. 2000; Skamnioti and Gurr 2009; Yang et al. 2009). Therefore, the synteny between finger millet and rice genomes can be exploited through marker-assisted selection to breed blast-resistant varieties of finger millet (Naylor et al. 2004; Dawson et al. 2009). Future research in finger millet could also be directed to identify ESTs linked to drought stress or salt tolerance; these markers could then be used for selection of drought- or salt-tolerant crop varieties that will result in increased production in degraded soils and areas normally affected by droughts (Dawson et al. 2009). Many new molecular markers are now available, such as SSRs, which can be easily used for genetic screening and complete characterization of existing germplasm collections. But the effort to expand these collections should continue, particularly to sample

S.S. Neves

wild populations of finger millet and other Eleusine species. Special attention should be given to the collections of E. kigeziensis, which appear to be scarce. Variability in this species is likely to be high as shown from sequencing work on just a few accessions. In addition to the more extensive genetic screening of populations and lines of wild finger millet (E. coracana subsp. africana), it will be important to carry out a comprehensive genetic diversity assessment in E. kigeziensis, a wild tetraploid species that is very closely related to E. coracana (Neves et al. 2005; see also Fig. 7.2) and, potentially, a source of useful traits for the breeding of cultivated finger millet (Dida and Devos 2006). Hybridization between the crop and wild finger millet could also significantly improve the nutritional value of the domesticated plants (Barbeau and Hilu 1993). For many decades, the funds directed to research on “orphan crops,” such as finger millet, had been meager, particularly when compared with those invested on major cereal crops, such as maize, rice, and wheat (Naylor et al. 2004). Even recently, Naylor et al. (2004) considered that aid agencies and science funding organizations should increase the investment on crops such as finger millet. For nearly 15 years now, the McKnight Foundation and its Collaborative Crop Research Program (CCRP) (http://mcknight.ccrp.cornell.edu/) has funded research in finger millet. In 2006, this foundation awarded a new grant of $920,000 to the project: “Genetic improvement, technology dissemination and seed system development in African Chloridoid cereals” (Grant No. 06-448 Tef/finger millet). The project is directed by Dr. Getachew Belay (Ethiopian Institute of Agricultural Research) and involves a network of scientists from Ethiopia, Uganda, Kenya, and the USA. The main goal of this project is to increase the productivity in tef (Eragrostis tef (Zucc.) Trotter) and finger millet crops and eventually benefit farmers in underprivileged regions of Africa. In December 2008, the Bill and Melinda Gates Foundation awarded $26.7 million over 5 years (2009–2013) to the McKnight Foundation for research on crops such as sorghum and finger millet (Star Tribune, Dec. 8, 2008). The ultimate goal of the grant is to help agricultural research and development in poor regions of the world by increasing yields and income of small farmers.

7 Eleusine Acknowledgments I extend my thanks to the herbaria and staff of Virginia Tech (USA), Kew Royal Botanic Gardens (UK), Sweden Museum of Natural History, Botanical Museum Berlin-Dahlem (Germany), Field Museum, Chicago, and INRB-INIA, Estac¸a˜o Agrono´mica Nacional, Oeiras (Portugal), for granting me access to their collections during visits, or for publishing in the web images of herbarium specimens (especially types), which I was then able to examine. Ginger Swire-Clark and Vance Baird (Clemson University) kindly provided photos of Eleusine for this publication. My work is currently supported by a postdoctoral fellowship (SFRH/BPD/26669/2006) from Fundac¸a˜o para a Cieˆncia e a Tecnologia, Portugal.

References Agnew ADQ (2006) A field key to upland Kenya grasses. J East Afr Nat Hist 95:1–86 Agyare C, Asase A, Lechtenberg M, Niehues M, Deters A, Hensel A (2009) An ethnopharmacological survey and in vitro confirmation of ethnopharmacological use of medicinal plants used for wound healing in Bosomtwi-AtwimaKwanwoma area, Ghana. J Ethnopharmacol 125:393–403 Anthony RG, Hussey PJ (1999) Dinitroaniline herbicide resistance and the microtubule cytoskeleton. Trends Plant Sci 4:112–116 Babu BK, Senthil N, Gomez SM, Biji KR, Rajendraprasad NS, Kumar SS, Babu RC (2007) Assessment of genetic diversity among finger millet (Eleusine coracana (L.) Gaertn.) accessions using molecular markers. Genet Resour Crop Evol 54:399–404 Baerson SR, Rodriguez DJ, Tran M, Feng Y, Biest NA, Dill GM (2002) Glyphosate-resistant goosegrass. Identification of a mutation in the target enzyme 5-enolpyruvylshikimate-3phosphate synthase. Plant Physiol 129:1265–1275 Barbeau WE, Hilu KW (1993) Protein, calcium, iron, and amino acid content of selected wild and domesticated cultivars of finger millet. Plant Foods Hum Nutr 43:97–104 Basu C, Halfhill MD, Mueller TC, Stewart CN Jr (2004) Weed genomics: new tools to understand weed biology. Trends Plant Sci 9:391–398 Beckie HJ (2007) Introduction to the symposium grass weed resistance: fighting back. Weed Technol 21:289 Bennett MD, Leitch IJ (2005) Plant DNA C-values database (release 4.0). Royal Botanic Gardens, Kew, UK. http://data. kew.org/cvalues/. Accessed 10 Nov 2009 Bisht MS, Mukai Y (2000) Mapping of rDNA on the chromosomes of Eleusine species by fluorescence in situ hybridization. Genes Genet Syst 75:343–348 Bisht MS, Mukai Y (2001a) Genomic in situ hybridisation identifies genome donor of finger millet (Eleusine coracana). Theor Appl Genet 102:825–832 Bisht MS, Mukai Y (2001b) Identification of genome donors to the wild species of finger millet, Eleusine africana by genome in situ hybridization. Breed Sci 51:263–269 Bisht MS, Mukai Y (2002) Genome organization and polyploid evolution in the genus Eleusine (Poaceae). Plant Syst Evol 233:243–258

131 Bouchenak-Khelladi Y, Salamin N, Savolainen V, Forest F, van der Bank M, Chase MW, Hodkinson TR (2008) Large multi-gene phylogenetic trees of the grasses (Poaceae): progress towards complete tribal and generic level sampling. Mol Phylogenet Evol 47:488–505 Breviario D, Baird WV, Sangoi S, Hilu K, Blumetti P, Gianı` S (2007) High polymorphism and resolution in targeted fingerprinting with combined b-tubulin introns. Mol Breed 20:249–259 Brosnan JT, Nishimoto RK, DeFrank J (2008) Metribuzinresistant goosegrass (Eleusine indica) in bermudagrass turf. Weed Technol 22:675–678 Buell CR (2009) Poaceae genomics: going from unattainable to becoming a model clade for comparative plant genomics. Plant Physiol 149:111–116 Chen S, Phillips SM (2006) Eleusine. In: Wu ZY, Raven PH, Hong DY (eds) Flora of China, vol 22, Poaceae. Science Press/PRC and Missouri Botanical Garden Press, Beijing/St. Louis, MO, p 481 Chennaveeraiah MS, Hiremath SC (1973) Genome relationships of Eleusine tristachya and E. floccifolia. Indian J Cytol Genet 8:1–5 Chennaveeraiah MS, Hiremath SC (1974) Genome analysis of Eleusine coracana (L.) Gaertn. Euphytica 23:489–495 Clayton WD, Renvoize SA (1986) Genera graminum: grasses of the world, vol 13, Kew Bulletin Additional Series. Her Majesty’s Stationery Office, London, UK Clayton WD, Harman KT, Williamson H (2006-onwards) GrassBase – the online world grass flora (version 29 Jan 2008). http://www.kew.org/data/grasses-db.html. Accessed 10 Nov 2009 Columbus JT, Cerros-Tlatilpa R, Kinney MS, Siqueiros-Delgado ME, Bell HL, Griffith MP, Refulio-Rodriguez NF (2007) Phylogenetics of Chloridoideae (Gramineae): a preliminary study based on nuclear ribosomal internal transcribed spacer and chloroplast trnL-F sequences. Aliso 23:565–579 Couch BC, Kohn LM (2002) A multilocus gene genealogy concordant with host preference indicates segregation of a new species, Magnaporthe oryzae, from M. grisea. Mycologia 94:683–693 Dawson IK, Hedley PE, Guarino L, Jaenicke H (2009) Does biotechnology have a role in the promotion of underutilised crops? Food Policy 34:319–328 De Wet JMJ, Prasada Rao KE, Brink DE, Mengesha MH (1984) Systematics and evolution of Eleusine coracana (Gramineae). Am J Bot 71:550–557 Devarumath RM, Hiremath SC, Rao SR, Kumar A, Bewal S (2005) Genome analysis of finger millet E. coracana by interspecific hybridization among diploid wild species of Eleusine (Poaceae). Cytologia 70:427–434 Dida MM, Devos KM (2006) Finger millet. In: Kole C (ed) Genome mapping and molecular breeding in plants, vol 1, Cereals and millets. Springer, Heidelberg, pp 333–343 Dida MM, Srinivasachary RS, Bennetzen JL, Gale MD, Devos KM (2007) The genetic map of finger millet, Eleusine coracana. Theor Appl Genet 114:321–332 Dida MM, Wanyera N, Dunn MLH, Bennetzen JL, Devos KM (2008) Population structure and diversity in finger millet (Eleusine coracana) germplasm. Trop Plant Biol 1:131–141

132 Ellis MH, Rebetzke GJ, Kelman WM, Moore CS, Hyles JE (2004) Detection of wheat streak mosaic virus in four pasture grass species in Australia. Plant Pathol 53:239 Goldstein DB, Schlo¨tterer C (1999) Microsatellites: evolution and applications. Oxford Univ Press, New York, USA Grassland Index (2009) Grassland species profiles. Food and Agriculture Organization of the United Nations, Plant Production and Protection Division, Rome, Italy. http://www.fao.org/ ag/AGP/AGPC/doc/GBASE/. Accessed 10 Nov 2009 Green JM (2009) Evolution of glyphosate-resistant crop technology. Weed Sci 57:108–117 Greenway PJ (1944) Origins of some East African food plants. East Afr Agric J 10:1–29 Hansen A (1980) Eleusine. In: Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, Webb DA (eds) Flora Europaea, vol 5. Cambridge Univ Press, Cambridge, UK, pp 258–259 Harland JR (1971) Agricultural origins: centers and noncenters. Science 174:468–474 Harris SA (1999) RAPDs in systematics – a useful methodology? In: Hollingsworth PM, Bateman RM, Gornall RJ (eds) Molecular systematics and plant evolution. Taylor and Francis, London, UK, pp 211–228 Hilu KW (1980) Eleusine tristachya (Lam.) Lam. (Poaceae). Madrono 27:177–178 Hilu KW (1981) Taxonomic status of the disputable Eleusine compressa (Gramineae). Kew Bull 36:559–563 Hilu KW (1988) Identification of the “A” genome of finger millet using chloroplast DNA. Genetics 118:163–167 Hilu KW (1994) Validation of the combination Eleusine coracana subspecies africana (Kennedy-O’Byrne) Hilu & DeWet. Phytologia 76:410–411 Hilu KW (1995) Evolution of finger millet: evidence from random amplified polymorphic DNA. Genome 38:232–238 Hilu KW (2003) Eleusine Gaertn. In: Barkworth ME, Capels KM, Long S, Piep MB (eds) Flora of North America North of Mexico, vol 25. Oxford Univ Press, New York, Oxford, pp 109–111 Hilu KW, Alice LA (2001) A phylogeny of Chloridoideae (Poaceae) based on matK sequences. Syst Bot 26:386–405 Hilu KW, De Wet JMJ (1976a) Domestication of Eleusine coracana. Econ Bot 30:199–208 Hilu KW, De Wet JMJ (1976b) Racial evolution in Eleusine coracana ssp. coracana (finger millet). Am J Bot 63:1311–1318 Hilu KW, Johnson JL (1992) Ribosomal DNA variation in finger millet and wild species of Eleusine (Poaceae). Theor Appl Genet 83:895–902 Hilu KW, Johnson JL (1997) Systematics of Eleusine Gaertn. (Poaceae: Chloridoideae): chloroplast DNA and total evidence. Ann Mo Bot Gard 84:841–847 Hilu KW, De Wet JMJ, Seigler D (1978) Flavonoid patterns and systematics in Eleusine. Biochem Syst Ecol 6:247–249 Hilu KW, De Wet JMJ, Harlan JR (1979) Archaeobotanical studies of Eleusine coracana ssp. coracana (finger millet). Am J Bot 63:330–333 Hiremath SC, Chennaveeraiah MS (1982) Cytogenetical studies in wild and cultivated species of Eleusine (Gramineae). Caryologia 35:57–69 Hiremath SC, Salimath SS (1991) Quantitative nuclear DNA changes in Eleusine (Gramineae). Plant Syst Evol 178:225–233

S.S. Neves Hiremath SC, Salimath SS (1992) The ‘A’ genome donor of Eleusine coracana (L.) Gaertn. (Gramineae). Theor Appl Genet 84:747–754 Hittalmani S, Parco A, Mew TV, Zeigler RS, Huang N (2000) Fine mapping and DNA marker-assisted pyramiding of the three major genes for blast resistance in rice. Theor Appl Genet 100:1121–1128 Holm LG, Plucknett DL, Pancho JV, Herberger JP (1977) Eleusine indica (L.) Gaertn. In: The World’s worst weeds: distribution and biology. University Press of Hawaii, Honolulu, pp 47–53 Ingram AL, Doyle JJ (2004) Is Eragrostis (Poaceae) monophyletic? Insights from nuclear and plastid sequence data. Syst Bot 29:545–552 Jones N, Ougham H, Thomas H (1997) Markers and mapping: we are all geneticists now. New Phytol 137:165–177 Keller B, Feuillet C (2000) Colinearity and gene density in grass genomes. Trends Plant Sci 5:246–251 Kennedy-O’Byrne J (1957) Notes on African grasses: XXIX. A new species of Eleusine from tropical and South Africa. Kew Bull 11:65–72 Lee LJ, Ngim J (2000) A first report of glyphosate-resistant goosegrass (Eleusine indica (L) Gaertn) in Malaysia. Pest Manag Sci 56:336–339 Lovisolo MR, Galati BG (2007) Ultrastructure and development of the megagametophyte in Eleusine tristachya (Lam.) Lam. (Poaceae). Flora 202:293–301 Lye KA (1999) Nomenclature of finger millet (Poaceae). Lidia 4:149–151 Mehra KL (1962) Natural hybridization between Eleusine coracana and E. africana in Uganda. J Indian Bot Soc 41:531–539 (printed in 1963) Mehra KL (1963a) Differentiation of the cultivated and wild Eleusine species. Phyton 20:189–198 Mehra KL (1963b) Considerations on the African origin of Eleusine coracana (L.) Gaertn. Curr Sci 32:300–301 Merkl N, Schultze-Kraft R, Infante C (2005) Phytoremediation in the tropics – influence of heavy crude oil on root morphological characteristics of graminoids. Environ Pollut 138:86–91 Meudt HM, Clarke AC (2007) Almost forgotten or latest practice? AFLP applications, analyses and advances. Trends Plant Sci 12:106–117 Muza FR, Lee DJ, Andrews DJ, Gupta SC (1995) Mitochondrial DNA variation in finger millet (Eleusine coracana L. Gaertn). Euphytica 81:199–205 Mysore KS, Baird V (1997) Nuclear DNA content in species of Eleusine (Gramineae): a critical re-evaluation using laser flow cytometry. Plant Syst Evol 207:1–11 National Research Council (1996) Lost crops of Africa, vol 1, Grains. National Academy Press, Washington, DC Naylor RL, Falcon WP, Goodman RM, Jahn MM, Sengooba T, Tefera H, Nelson RJ (2004) Biotechnology in the developing world: a case for increased investments in orphan crops. Food Policy 29:15–44 Neves SS, Swire-Clark G, Hilu KW, Baird WV (2005) Phylogeny of Eleusine (Poaceae: Chloridoideae) based on nuclear ITS and plastid trnT-trnF sequences. Mol Phylogenet Evol 35:395–419 Ng CH, Wickneswari R, Salmijah S, Teng YT, Ismail BS (2003) Gene polymorphisms in glyphosate-resistant and -susceptible biotypes of Eleusine indica from Malaysia. Weed Res 43:108–115

7 Eleusine Ng CH, Wickneswary R, Salmijah S, Ismail BS (2004a) Inheritance of glyphosate resistance in goosegrass (Eleusine indica). Weed Sci 52:564–570 Ng CH, Wickneswary R, Salmijah S, Teng YT, Ismail BS (2004b) Glyphosate resistance in Eleusine indica (L.) Gaertn. from different origins and polymerase chain reaction amplification of specific alleles. Aust J Agric Res 55:407–414 Nout MJR (2009) Rich nutrition from the poorest – cereal fermentations in Africa and Asia. Food Microbiol 26:685–692 Owen MDK, Zelaya IA (2005) Herbicide-resistant crops and weed resistance to herbicides. Pest Manag Sci 61:301–311 Phillips SM (1972) A survey of the genus Eleusine Gaertn. (Gramineae) in Africa. Kew Bull 27:251–270 Phillips SM (1974) Eleusine. In: Clayton WD, Phillips SM, Renvoize SA (eds) Gramineae (Part 2). In: Polhill RM (ed) Flora of Tropical East Africa. Crown Agents for Overseas Governments and Administrations, pp 260–267 Phillips SM (1982) A numerical analysis of the Eragrostideae (Gramineae). Kew Bull 37:133–162 Phillips S (1995) Poaceae (Gramineae). In: Hedberg I, Edwards S (eds) Flora of Ethiopia and Eritrea, vol 7. The National Herbarium, Addis Ababa University/Department of Systematic Botany, Uppsala University, Addis Ababa/Uppsala, Sweden Powell W, Langridge P (2004) Unfashionable crop species flourish in the 21st century. Genome Biol 5:233. doi:10.1186/gb2004-5-7-233 Radosevich SR, Holt JS, Ghersa CM (2007) Ecology of weeds and invasive plants: relationship to agriculture and natural resource management, 3rd edn. Wiley, Hoboken, NJ Roodt-Wilding R, Spies JJ (2006) Phylogenetic relationships in southern African chloridoid grasses (Poaceae) based on nuclear and chloroplast sequence data. System Biodivers 4:401–415 Salimath SS, De Oliveira AC, Godwin ID, Bennetzen JL (1995a) Assessment of genome origins and genetic diversity in the genus Eleusine with DNA markers. Genome 38:757–763 Salimath SS, Hiremath SC, Murthy HN (1995b) Genome differentiation patterns in diploid species of Eleusine (Poaceae). Hereditas 122:189–195 Sandermann H (2006) Plant biotechnology: ecological case studies on herbicide resistance. Trends Plant Sci 11:324–328 Sanz Elorza M, Dana E, Sobrino E (2001) Aproximacio´n al listado de plantas alo´ctonas invasoras reales y potenciales en Espan˜a. Lazaroa 22:121–131 Sisay A, Baars RMT (2002) Grass composition and rangeland condition of the major grazing areas in the mid Rift Valley, Ethiopia. Afr J Range Forage Sci 19:161–166

133 Skamnioti P, Gurr SJ (2009) Against the grain: safeguarding rice from rice blast disease. Trends Biotechnol 27:141–150 Srinivasachary DMM, Gale MD, Devos KM (2007) Comparative analyses reveal high levels of conserved colinearity between the finger millet and rice genomes. Theor Appl Genet 115:489–499 Upadhyaya HD, Gowda CLL, Gopal Reddy V (2007) Morphological diversity in finger millet germplasm introduced from Southern and Eastern Africa. J SAT Agric Res 3. http:// ejournal.icrisat.org/ Upadhyaya HD, Gowda CLL, Sastry DVSSR (2008) Plant genetic resources management: collection, characterization, conservation and utilization. J SAT Agric Res 6. http:// ejournal.icrisat.org/ USDA NRCS (2009) The PLANTS Database. US Department of Agriculture, National Plant Data Center, Baton Rouge, Louisiana. http://plants.usda.gov/. Accessed 10 Nov 2009 Vadivoo AS, Joseph R, Ganesan NM (1998) Genetic variability and diversity for protein and calcium contents in finger millet (Eleusine coracana (L.) Gaertn) in relation to grain color. Plant Foods Hum Nutr 52:353–364 Wang ML, Barkley NA, Yu JK, Dean RE, Newman ML, Sorrells ME, Pederson GA (2005) Transfer of simple sequence repeat (SSR) markers from major cereal crops to minor grass species for germplasm characterization and evaluation. Plant Genet Resour 3:45–57 Wang J, Zhang Z, Su Y, He W, He F, Song H (2008) Phytoremediation of petroleum polluted soil. Petrol Sci 5: 167–171 Werth CR, Hilu KW, Langner CA (1994) Isozymes of Eleusine (Gramineae) and the origin of finger millet. Am J Bot 81:1186–1197 Wolfe AD, Xiang QY, Kephart SR (1998) Diploid hybrid speciation in Penstemon (Scrophulariaceae). Proc Natl Acad Sci USA 95:5112–5115 Yamamoto E, Baird WV (1999) Molecular characterization of four b-tubulin genes from dinitroaniline susceptible and resistant biotypes of Eleusine indica. Plant Mol Biol 39:45–61 Yang Q, Lin F, Wang L, Pan Q (2009) Identification and mapping of Pi41, a major gene conferring resistance to rice blast in the Oryza sativa subsp. indica reference cultivar, 93-11. Theor Appl Genet 118:1027–1034 Zeng L, Baird WV (1999) Inheritance of resistance to antimicrotubule dinitroaniline herbicides in an “intermediate” resistant biotype of Eleusine indica (Poaceae). Am J Bot 86:940–947

.

Chapter 8

Eragrostis Mahmoud Zeid, Vivana Echenique, Marina Dı´az, Silvina Pessino, and Mark E. Sorrells

8.1 Introduction Eragrostis is the largest genus within the subfamily Chloridoideae of the Poaceae. The first description and naming of the genus was made by von Wolf (1776) on a specimen of Eragrostis minor. Although it is not clear why Wolf selected this name, “lovegrass” is the accepted name for the genus (from eros, love, and agrostis grass). Today, about 350 species have been described in the genus. Most species occupy open habitats with poor soils and many occur in ruderal sites (Van den Borre and Watson 1994). Their distribution extends across wide altitudinal and moisture gradients from pluvial to xeric habitats. Most of the species (50%) appear to be native of Africa (Cufodontis 1974). Fifty five species are native in Australia (Lazarides 1997), 25 in United States and Canada, and 36 in Mexico (Peterson and Valdes-Reyna 2005). Despite the large number of Eragrostis species, they are mostly unknown weedy grasses, and only a few became known as forage grasses. The adverse effects of the frequent droughts that hit Africa, Australia, and North America in the period between 1880 and 1930 led to a decrease in livestock production on grassland. The need for new grass species that could produce good quality forage under limited precipitation and heavy grazing was realized. This motivated a worldwide search for the “miracle grass” [see Cox et al. (1988) for more details]. The search was led

M.E. Sorrells (*) Department of Plant Breeding and Genetics, Cornell University, Ithaca, NY 14853, USA e-mail: [email protected]

by botanists, ranchers, and even military personnel from Australia, England, South Africa, and the United States. After the evaluation of thousands of worldwide trials, the “winners” were four African grass species that were characterized by the ease of establishment relative to native grasses, persistence, and high productivity as forage species. These grasses were weeping lovegrass [Eragrostis curvula (Schrad.) Nees], Lehmann lovegrass (Eragrostis lehmanniana Nees), buffelgrass (Cenchrus ciliaris L.), and kleingrass (Panicum coloratum L.). E. curvula and E. lehmanniana have been heavily utilized mainly in the USA and Argentina for soil conservation and forage production purposes. Another well known species is E. pilosa, a common weed in many parts of the world and the most closely related to the cultivated species E. tef. Because this species is considered the progenitor of the cultivated tef, many studies were initiated on this weed, and it is now part of the breeding programs for tef improvement in Ethiopia. The three species E. curvula, E. lehmanniana, and E. pilosa will be the focus of this chapter.

8.2 Basic Botany of the Species 8.2.1 Taxonomic Position Eragrostis Wolf, consisting of more than 350 species, is the largest genus in the Eragrostideae, a tribe of 80 genera and about 1,000 species (Peterson et al. 1997) mainly distributed in tropical and warm-temperate regions all around the world, where the centers of diversity are assumed (Hartley and Slater 1960).

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_8, # Springer-Verlag Berlin Heidelberg 2011

135

136

The tribe is regarded as an unnatural grouping of convenience, and Eragrostis itself is a large polyphyletic assemblage (Voigt et al. 2004). Members of Eragrostis are generally characterized by paniculate inflorescence, multifloreted spikelets, glabrous threenerved lemmas and unawned, longitudinally bowedout paleae with ciliolate keels, and leaves with ciliate ligules (Peterson et al. 1997). Attempts at within genus classification have focused largely on characters such as the manner of spikelet disarticulation or C4 photosynthesis type. However, a cladistic analysis of morphological and anatomical characters has suggested that the mode of spikelet disarticulation has little to do with natural relationships in the group (Van den Borre and Watson 1994). Furthermore, relationships may have been obscured by allopolyploidy (Ingram and Doyle 2004). The polyphyletic nature of Eragrostis based on the analysis of conserved sequences has been a point of debate (Hilu and Alice 2000; Ingram and Doyle 2004; Roodt-Wilding and Spies 2006). Ingram and Doyle (2004), using the plastid locus rps16 and the nuclear gene waxy, argue that Eragrostis is monophyletic with the inclusion of several segregates, including Acamptoclados, Diandrochloa, and Neeragrostis. Roodt-Wilding and Spies (2006) on the other hand, using data from chloroplast trnL (UAA) 50 exon-trnF (GAA) region and the nuclear ribosomal internal transcribed spacer (ITS) regions, see the genus as polyphyletic. Peterson and Sanchez Vega (2007) have noted that the debate is not yet over and what is lacking is a definitive treatment of the infrageneric classification of the entire genus.

M. Zeid et al.

E. pilosa is a very widely distributed species with extensive morphological diversity. It is an annual, reproduces by seed that is a caryopsis, ellipsoid, laterally compressed, 0.6–1 mm long (Fig. 8.1). Culms are erect or geniculately ascending, 8–70 cm long. The ligule is a fringe of hairs, and leaf-blades are 2–20 cm long, 1–4 mm wide. The inflorescence is an open panicle, elliptic or ovate, 4–25 cm long with spikelets comprising 4–14 fertile florets. Glumes are deciduous with the lower glume 0.5–0.7  the length of the upper glume. On the other hand, both E. curvula and E. lehmanniana are perennials, reproducing by seeds, although in the former, reproduction is mostly through apomixis (facultative). Seed is an ellipsoid caryopsis with the adaxial surface flattened, 1.4 mm long, 0.6 mm wide, and translucent reddish brown in E. curvula, dark brown or black and smaller (0.6–0.8 mm long, 0.4 mm wide) in E. lehmanniana. Culms are tufted, 40–100 cm tall, erect or ascending, and in E. lehmanniana, some culms are decumbent or geniculate at lower nodes, often stolon-like. The leaf blades are longer in E. curvula (20–30 cm), 1–1.4 mm wide, with ligules densely ciliate, shorter (5–15 cm), 1–3 mm wide, with truncate ciliate ligules in E. lehmanniana. The inflorescence is an open panicle, oblong, 20–40 cm long in E. curvula while oblong to lanceolate and much shorter (7.5–16 cm) in E. lehmanniana. Spikelets are grayishgreen with multiflorets (6–15), glumes deciduous, and the lower glume shorter than the upper one. E. curvula was initially described by Schrader like Poa curvula being included in the genus Eragrostis by

8.2.2 Morphology of Eragrostis There are many detailed reports on the morphology of Eragrostis upon which plants could be identified from various parts of the world, from Africa (Cope 1998), Australia (Lazarides 1997), Bolivia (Renvoize 1998), Brazil (Boechat and Longhi-Wagner 2000), Ecuador (Laegaard and Peterson 2001), Malesia (Veldkamp 2002), Mexico (Peterson and Valdes-Reyna 2005), Peru (Peterson and Sanchez Vega 2007), and the United States and Canada (Peterson 2003). Here, we will give a very brief description of the three species with which this chapter is concerned, based mainly on Clayton et al. (2006).

Fig. 8.1 The small size of Eragrostis pilosa grains in comparison to a wheat grain

8 Eragrostis

Nees. Weeping lovegrass, E. curvula (Schrad.) Nees, is a vigorous grass native of South Africa that is well adapted for forage production and soil conservation in different countries. This species constitutes a polymorphic group that is morphologically diverse, poorly understood, and not well circumscribed. For those reasons, it is frequently referred to as the E. curvula complex or E. curvula sensu lato (Voigt et al. 2004). The difficulty in defining the structure of this group is caused by the lack of taxonomic knowledge. Although there are marked morphological differences between varieties, the existence of ecotypes with intermediate characteristics hinders the classification. In addition, the apomictic reproduction can limit any new variation. This has led to considerable confusion among ecotypes and cultivars that are currently available to breeders and producers (Poverene 1988; Covas 1991). The E. curvula complex includes intermediate forms that overlap with E. chloromelas and E. lehmanniana; so some authors do not consider these species to be separate from E. curvula sensu lato. Several publications in the United States have classified E. chloromelas as Boer grass, but now it is identified as a botanical variety E. curvula var. conferta. Lehmann grass, not always included in the complex E. curvula, is more robust and resistant to cold and has morphological characteristics that differentiate it from weeping lovegrass (Voigt et al. 2004). The morphological diversity of this complex leads to the creation of morphological types that are used to describe the germplasm of the species but with severe limitations. Leigh (1960) and Leigh and Davidson (1968) recognized within E. curvula five agronomic types: curvula, robusta blue, robusta green, robusta intermediate, and chloromelas. Later, another group was recognized, conferta (Jacobs 1982). This classification was based primarily on morphological characteristics of leaves and inflorescences, plant size, and growth habit. Such characteristics are influenced by growth conditions and fertilization, and although they are generally distinguishable as mature plants, they are difficult to recognize as plantlets. The high morphologic diversity existing in the complex is accompanied by a similar change in physiological characteristics and agronomic traits such as cold resistance, photoperiod response, adaptation to various soil types, forage quality, and response to grazing. Covas (1991) did not recognize all types defined above and discarded chloromelas, which was seen as a different species

137

(E. chloromelas). He proposed a classification based on seeds, leaf blades, and ear characteristics for materials grown in Argentina. The new classification and the most important Argentinean cultivars are the following: curvula (cvs. Tanganyika, Ermelo, Don Arturo INTA and Morpa), pilosa (cv. Don Juan INTA), conferta (cv. Don Walter INTA). and robusta (cvs. Don Carlos INTA, Don Pablo INTA and Don Eduardo INTA). The reproductive mode of weeping lovegrass (apomixis) seems to have played a fundamental role in producing a complex superstructure of polyploid forms, as has happened in other grasses. While sexuality would allow the emergence of polyploid types that are progressively more complex, apomixis would ensure the survival of these meiotically unstable new lines. These genotypes could, in turn, lead to other genotypes through the fusion of unreduced gametes and facultative apomixis. The E. curvula complex consists of a set of populations, most of which are clonal and a smaller proportion that is sexual. The contact between both populations continuously adds new genotypes to the complex, which in this way constitutes a true collective species (Poverene 1988). The intermediate types encountered by hybridization are expressed in the cultivars as a combination of morphological and agronomic characteristics of different varieties. There is currently no precise description of each cultivar or a suitable method for characterization of the seed. The seed lacks a sufficient number of distinctive characters to identify cultivars belonging to the same botanical variety (Poverene 1988). That is to say, the color, shape, and size of the seeds of various cultivars belonging to the same morphological type are remarkably constant, but there are significant differences between types.

8.2.3 Cytology and Karyotype The small size of the chromosomes, characteristic of the Eragrostoideae, and the difficulty in obtaining sufficiently clear preparations contribute to the lack of information about the lovegrass karyotype. The chromosome sizes of E. pilosa, E. aethiopiaca, and E. bicolor, as an example, range between 0.8 and 2.2 mm (Tavassoli 1986). The basic chromosome number characteristic of the genus is 10 (x ¼ 10) with

138

various ploidy levels (2n ¼ 20, 40, 60, 80, 120), [see Voigt et al. (2004) for more details]. E. pilosa is a tetraploid (2n ¼ 40) similar to its cultivated relative E. tef (Jones et al. 1978), while the ploidy level in E. lehmanniana varies, i.e., 2n ¼ 40 and 2n ¼ 60 (Halvorsen and Guertin 2003b). Ploidy levels within the E. curvula complex, on the other hand, vary from diploid to octoploid (Spies and Gibbs Russell 1988; Poverene and Voigt 1997). Although a few sexual diploid (2n ¼ 2x ¼ 20) plants were identified, they are very infrequent (so far only four have been collected) and do not occur in all forms of E. curvula (Poverene and Voigt 1997). Diploids are always sexual and self-incompatible (need cross-pollination), while polyploids reproduce by diplosporous apomixis and are self-compatible. Although many of the South African collections are fully apomictic, a few of the last collections were described as facultative (Vorster and Liebenberg 1977). The only report of triploidy in the E. curvula complex comes from Leigh (1960) for a plant belonging to the robusta type. Later, Vorster and Liebenberg (1977) reported the existence of one pentaploid plant. While the heptaploid level (2n ¼ 7x ¼ 70) is more frequent, the even number polyploidy types appear to be much more abundant than odd-polyploidy types in the complex (Poverene 1988). It is clear that the persistence of polyploids should be due to the apomictic reproduction, but the meiotic instability that characterized them necessarily represents an adaptive disadvantage (Poverene and Curvetto 1991). Several studies have also found aneuploids, and the spread of such cytotypes is also possible because of apomixis (De Winter 1955; Leigh 1960; Jones et al. 1978). Poverene (1988) and Cardone et al. (2006) reported the presence of metacentric and submetacentric chromosomes with an average size of 2–3 mm in E. curvula. Meiotic chromosomes in anthers pair mainly as bivalents, but univalents and multivalents are also common. Within the multivalents, tetravalents were the most common configurations. Tanganyika (curvula type 2n ¼ 4x ¼ 40) showed a high frequency of multiple configurations (four tetravalent and one trivalent on average). The irregularities were higher in the accessions of higher ploidy levels (Streetman 1970; Vorster and Liebenberg 1977). Poverene (1988), based on observations in Tanganyika (higher frequency of multiple configurations, the presence of chromatin bridges joining these figures, and the

M. Zeid et al.

frequencies of ring-shaped quadrivalents), hypothesized that there was a high homology of the four chromosomes of the association so that it would be an autopolyploid. Recently, Cardone et al. (2006) analyzed the meiosis of a tetraploid plant obtained by chromosome doubling of an artificial diploid-derived form Tanganyika by in vitro culture from inflorescences with colchicine and obtained an average of 12.2 bivalents, 3.1 quadrivalents, 0.8 trivalents, and 0.8 univalents. These results and those from Poverene (1988) would support the hypothesis of the autopolyploid origin of cv. Tanganyika. In species of Paspalum, a low frequency of tetravalents in tetraploid cytotypes has been considered as the result of segmental allopolyploidy (Quarin et al. 1996). The structural alterations found in Eragrostis are also considered to be an indication of segmental allopolyploidy (Vorster and Liebenberg 1977). Pollen viability evaluated in different plant materials of different ploidy levels in E. curvula is high, with at least 50% of viable pollen grains (Poverene 1988; Echenique and Polci 1994; Polci 2000). This value is typical of a pseudogamous grass and guarantees a normal seed production. Estimates of pollen viability confirmed that none of the accessions was highly sterile, despite the high frequency of meiotic abnormalities observed. Several authors have tried to relate the number of chromosomes and the morphological types or species that make up the complex, but the data are mixed. Each of the species or morphological types includes a number of different cytotypes. Intraspecific variation in chromosome number is very common in grasses (Quarin and Ferna´ndez 1982; Poverene 1988; Keeler 2004). E. tef, the only cereal member of the Eragrostideae, is cultivated for its grain (Bekele et al. 1995). Tef is an allotetraploid (2n ¼ 4x ¼ 40) whose origin within the genus Eragrostis is unknown. Studies based on morphological, cytological, and biochemical characters have suggested as many as 14 wild Eragrostis species as potential progenitors of the crop [see Ingram and Doyle (2003) for further details]. E. pilosa, however, appears to be the most likely candidate for the direct wild progenitor of E. tef. Morphologically, the two species are very similar, although tef plants are generally larger with fewer tillers and later maturing than E. pilosa. Another important difference shows at maturity, where the lemmas, paleas, and the caryopses of E. tef remain attached to the rachis, making it possible for farmers to harvest the grains. On the contrary,

8 Eragrostis

E. pilosa shatters its grains due to the breaking apart of the spikelets at maturity. From the cytological point of view, both species have a similar karyotype (Tavassoli 1986). In addition, evidence from the nuclear gene waxy and the plastid locus rps16 (Ingram and Doyle 2003) clearly indicated that E. pilosa is the closest relative of tef and also supports the hypothesis that E. pilosa is its progenitor. As to the genome size of Eragrostis Wolf, an estimation by flow cytometry rendered 595.35 Mbp for E. curvula (Echenique et al. unpub results) and 714–733 Mbp for the cultivated E. tef (Ayele et al. 1996). The DNA content per haploid nucleus was estimated for E. tef to be 0.68 pg (Bennett and Smith 1976).

8.2.4 Agricultural Status The three introduced species, weeping lovegrass, boer lovegrass, and Lehmann lovegrass have been heavily utilized in the United States’ agricultural system. Weeping lovegrass is unique in that it has been an important resource for soil conservation and forage in semiarid regions since its introduction. It exhibits water-saving strategies, including leaf waxing and rolling in response to increasing water demand without necessarily exhausting its water supply (Echenique et al. 1986a, b; Colom and Vazzana 2001), thus showing good performance in relation to forage production under water stress conditions (Ruiz et al. 2008). In addition, its leaf growth is relatively insensitive to soil water drying (Puliga et al. 1996). However, different water stress responses have been detected among other types within the complex including boer lovegrass (cv. Consol) and robusta that are more resistant than the curvula type (cv. Ermelo) (Colom and Vazzana 2003). These authors reported that during drought, relative water content (RWC) decreased 65% in cv. Ermelo and drought stress caused severe decreases in photosynthetic rates, while cv. Consol showed low variation in these parameters. Echenique and Curvetto (1986) found similar results in materials from Argentina. Balsamo et al. (2006) reported a correlation between drought tolerance of three Eragrostis species and leaf tensile properties (behavior during mechanical stress). E. curvula (drought tolerant) had higher tensile strength values than E. tef (moderately drought tolerant) and E. capensis (drought intolerant).

139

E. curvula tolerates several soil conditions that are harmful to other species including extreme pH. It grows well on acid soils and mine soils at pH 4.0 as well as calcareous soils with pH near 8.0. It has been used for the consolidation of erodible soils through the incorporation of organic matter (Taliaferro et al. 1975; Busso and Brevedan 1991). Gherbin et al. (2007) compared 23 non-native grasses along with Tall Fescue (Festuca arundinacea, a native grass in the Mediterranean environment) for their adaptability to the coastal plains of southern Italy in terms of their productivity and nutritional quality. In the second year after establishment, warm-season perennial grasses showed higher dry matter (DM) production, with E. curvula producing the highest annual DM yield (21.1 t ha1). In Argentina, E. curvula is the most widely cultivated perennial grass with an estimated area of more than 5,000,000 ha (Covas and Cairnie 1985). Weeping lovegrass has a remarkable capacity for the production of forage in environments unsuitable for most of the other species used for feeding cattle. This could be attributed to its drought resistance, its capacity to produce in very loose, low fertility soils, and its tolerance to poor management. In addition, it has outstanding potential longevity that allows pastures to be considered permanent (Covas 1991). The La Pampa province native grasses are the most important forage resource for cattle in Argentina. However, in places where agriculture is possible, weeping lovegrass has increased the receptivity of fields. The weeping lovegrass cycle (spring-summer) allows the combination of its use with the best native species, which have an autumn-winter cycle, like Piptochaetium napostaense, Poa ligularis, Stipa tenuis, Bromus brevis, etc., thus avoiding the loss of these species by overgrazing. The plasticity of weeping lovegrass allows its introduction in different production systems as it adapts to various alternatives, providing greater efficiency in all cases (Herna´ndez 1991).

8.3 Conservation Initiatives Data from the United States Department of Agriculture Germplasm Resources Information Network (USDA-GRIN), (USDA 2009) provide a list of 180 Eragrostis species. Seeds for a large number of those

140

species from a worldwide collection are available for distribution, including the cultivar “A67” from which most of the weeping lovegrass varieties available today in the United States were derived. There are currently 548 E. curvula, 54 E. lehmanniana, and 10 E. pilosa accessions listed in GRIN. In Europe, the System-wide Information Network for Genetic Resources (SINGER, http://singer.cgiar.org, accessed 18 Nov 2009) shows 191 Eragrostis accessions deposited at the International Center for Tropical Agriculture (CIAT) and the International Livestock Research Institute (ILRI). Another major conservation location is the Institute of Biodiversity Conservation (IBC) in Ethiopia where most of the cultivated E. tef accessions and other wild Eragrostis species from Ethiopia were deposited. Weeping lovegrass was first introduced into Argentina from the United States. Predominant cultivars were Tanganyika, Ermelo, and Don Pablo INTA (Covas 1991). The introduced cultivars established in the north of San Luis and Co´rdoba, east of La Pampa, and west of Buenos Aires provinces. Today, an active Germplasm Bank located in the Anguil experimental station (La Pampa), where the cultivars were first introduced, is established. The National Institute of Agriculture of Argentina (INTA), part of the Red Nacional de Recursos Gene´ticos (National Network for Genetic Resources), is the location where forage species are conserved, including Eragrostis species (http://servicios.inta.gov.ar/bancos/ampliacion.html).

8.4 Role in Elucidation of Origin and Evolution of Allied Crop Plants Attempts to quantify genetic variation in cultivated E. tef have usually included the wild E. pilosa and E. curvula as reference species. Results from Costanza et al. (1979), Kefyalew et al. (2000), and Tefera et al. (1990) demonstrated that, although tef showed high levels of phenotypic diversity, even more variation could be observed between species including its closest relative E. pilosa. Molecular marker techniques, however, were essential to overcome the problems associated with the effect of environment on morphological traits as well as seed mixture. Amplified fragment length polymorphism (AFLP) analysis was used to study tef accessions and elite lines as well as

M. Zeid et al.

a number of diverse accessions from E. pilosa and E. curvula (Bai et al. 1999b). Results indicated that variability among E. tef accessions on the DNA level was much lower (18%) than that estimated from morphological traits (Bai et al. 1999b). Furthermore, a poor correlation between relationships based on AFLP data and morphological data was reported by Ayele and Nguyen (2000) and Ayele et al. (1999). E. pilosa appeared more similar to E. tef than was E. curvula based on marker data, but the morphological data failed to differentiate the three species, thus emphasizing the need to employ molecular markers for such studies. Random amplified polymorphic DNA (RAPD) markers were employed by Bai et al. (2000) to estimate relationships among the same accessions from Bai et al. (1999b), with additional new accessions from E. pilosa and E. curvula. Results were similar to those in the study using AFLP markers, although genetic similarity between species was much lower. The first linkage map in the genus Eragrostis was reported by (Bai et al. 1999a) using AFLP markers. A striking observation from that study, which involved a set of recombinant inbred lines from a cross between two cultivated tef varieties (Kaye Murri  Fesho), was the very low level of polymorphism (6.1%) for two morphologically diverse varieties. To overcome this problem of low polymorphism from intraspecific crosses, Zhang et al. (2001) used restriction fragment length polymorphism (RFLP) markers and a population of 116 F8 recombinant inbred lines (RILs) from an interspecific cross between the cultivar Kaye Murri (E. tef ) and its wild relative E. pilosa accession (30-5). E. pilosa was chosen because it was possible to cross with E. tef and because it was characterized by a shorter stature as compared to the cultivated tef. The source of DNA probes was a tef cDNA library, and grass anchor probes from other grass species (barley, rice, oat and wheat). This study realized a substantial increase in the level of polymorphism (67%) and initiated comparative mapping of Eragrostis with other members of the Poaceae, since 40% of the mapped markers were probes from other grass species. Utilizing 94 RILs of the same interspecific cross, Yu et al. (2006b) updated the map by mapping a set of Eragrostis-specific expressed sequence tag (EST)derived simple sequence repeats (SSR), commonly known as EST-SSR markers (Yu et al. 2006a) and heterologous markers based on EST sequences from finger millet, rice, and wheat. Markers were grouped

8 Eragrostis

141

into 21 linkage groups with a mean distance of 12.3 cm between markers (Fig. 8.2). A second linkage map from an interspecific cross between another variety (E. tef cv. DZ-01-2785) and the same E. pilosa

1

RZ15 TCD230 TB RZ909 RZ27 A TCD134 RZ38 TCD182 CSU60 RZ519 BCD20 TCD327 BCD34 BCD1087 BCD944 RZ467 RZ49 TCD99 CDO116 TCD45* TCD99 CNLT11 KSUM15 RM15

0. 15. 26. 33. 43. 53. 66. 79. 95. 119. 134. 146. 159. 177. 190. 200. 209. 218. 229. 248. 264. 271. 296. 298. 305.

2 ISSR842

CNLT49 CNLT49 ISSR548b RM170 RM17 TCD30 ISSR549 CNLT8 ISSR842 ISSR841 TCD21

ISSR842e RZ166 TCD415 TCD397a

ISSR811 RM10

217. 233.

TCD27

257.

0. 6. 32. 43. 73. 85. 111. 114. 117. 127. 135.

0. 14. 32. 40. 56. 68. 94. 113. 119. 137. 162. 168. 179.

0. 0. 23. 30. 38. 61. 88. 101. 106. 130.

12

0. 22. 31. 33. 36. 42. 60. 64. 89. 107.

0.0 RM124c 23.9 RZ467a RZ69* 46.6 RZ251* 60.5 CNLT65 85.4

ISSR840b** DupW124

CNLT146-

158.

0. 18.

RM110a* RM110 CNLT6

40. 53. 68.

CNLT11

91.

22.8

0. 20.

TCD197 RM14 TCD24

59. 81. 92.

CNLT14 RZ44 TCD230

122. 137. 155.

TCD5 CNLT7 CNLT41 CNLT41 ISSR842 TCD327 TCD327

0. 25. 49. 54. 65. 72. 88.

15 0.0 CNLT137-T13 15.9 CNLT154-Sa14* 27.0 RM185* 37.8 ISSR841a* 49.3

16 RZ395 0.0 RM134 23.1 29.6 ISSR836b* 46.7

0.0 11.4 34.6

76.6 19

0.0

ISSR811 RZ460

10

ISSR842 KSUM2

14 0.0 CDO1387 RZ329 20.5 RZ141 32.6 RZ123 50.3 RZ204

18 0.0 9.4 16.3 32.6

0. 6. 28. 53. 62. 72. 88. 113. 134.

9

PAL ISSR81 TCD227b TCD31 TCD11 TCD227 ISSR548 CDO3 TCD32 ISSR81

13

5

SRSC2_02 ISSR842 CNLT151inf30 CSU7 RZ962b TCD306 CDO139 SRSC3_00

8

ISSR811 ISSR840 CNL78 ISSR841 RM124 CNLT14 RZ698 KSUM195 inf14 TCD230

0.0 CNLT61 28.0 RZ962a 54.3 RZ413 63.2 TCD205 87.8 TCD5

4

TCD3 RM170 CDO2 TCD248 TCD95 ISSR549 CSU3 PRSC1_02 RM124 RZ444 RZ909 CNLT13 RZ444

7

17 CNLT149* RZ214b* CNL100 RZ214a

0. 32. 39. 65. 92. 93. 97. 109. 114. 128. 149. 161. 186.

11 DupW4

3

BCD1087b BCD1087a ISSR836a CNL78 CNL5 ISSR54 CDO78 CNLT146BCD88 PALa RZ87 RZ962

6

accession (30-5) was published by Chanyalew et al. (2005). For this linkage map, 120 RILs were screened using AFLP, EST-SSRs from wheat and Eragrostis, SSR markers from rice, and a set of intersimple

TCD503 RZ698b** RZ519b

Fig. 8.2 Genetic linkage map derived from 94 recombinant inbred lines from a cross between E. tef cv. Kaye Murri and E. pilosa (30-5). Loci names with an asterisk indicate significant

21

20 0.0 8.1 13.8

RZ588** RM142

0.0 17.7

lfm256 TCD424

0.0 6.5

distorted segregation (P < 0.05) and single asterisk and double asterisks indicate the preferential transmission from the alleles of E. tef, or E. pilosa, respectively (Yu et al. 2006b)

142

sequence repeats (ISSR) markers. The linkage map covered 78% of the genome, and the average distance between markers amounted to 12.7 cm. To update the existing maps, heterologous markers from pearl millet, finger millet, and rice are being screened to insure that more anchor markers are included on the maps for comparative analysis purposes (M. Zeid and M. E. Sorrells, unpub data). Most recently, genomic SSR markers have been developed to improve genome coverage and to identify quantitative trait loci (QTL) for various traits, mainly yield and lodging resistance in the cultivated tef species (M. Zeid and M. E. Sorrells, unpub data). The recent results from Yu et al. (2007) on QTL analysis in tef have indicated that alleles from the wild relative, E. pilosa (30-5), had an increasing effect for 29% of the QTL detected.

8.5 Crop Improvement Through Traditional and Advanced Tools 8.5.1 Forage Crop Improvement 8.5.1.1 Traditional Tools The E. curvula complex contains a vast diversity of genes preserved and maintained by apomictic reproduction. Variation in the establishment capacity, drought resistance response, forage quality, forage yield, and iron efficiency, among others, has been detected. This variation can be manipulated through apomictic breeding, i.e., by hybridizing rare sexual plants with apomictic selections. In this way, the variation can be released and genotypes with new combinations of traits can be produced. Those hybrids, if sexual, can become the basis of further hybridization (Voigt 1991). Apomictic lovegrass hybrids were developed from crosses of sexual boer  apomictic weeping lovegrass (Voigt 1984) with significant differences in forage vigor and quality found among the hybrids. Selection for palatability gave rise to a variety of weeping lovegrass, “Morpa” (more palatable), which was released by the USDA (Voigt et al. 1970). Morpa was selected for its higher palatability relative to other weeping lovegrass strains, such as Common and Ermelo in summer grazing. However, it is very similar to Ermelo in appearance. The apomictic cultivar Morpa produced

M. Zeid et al.

hybrids that were higher in their in vitro dry matter digestibility (IVDMD) than common weeping lovegrass hybrids. Boer lovegrass genotype 40-9-69FQ (40-9) was inferior as a female parent because its hybrids tended to be lower in vigor than those from other boer lovegrass genotypes. Several hybrids from the cross 40-9 x Morpa were significantly higher in IVDMD than Morpa or Ermelo, the most digestible cultivars (Voigt 1984). Breeders were successful in improving boer lovegrass through direct selection. “A-84” boer lovegrass was first introduced to the United States in 1932 and was officially released in 1950 (Hanson 1965). Selection for seedling drought tolerance in A-84 yielded the variety “Catalina” boer lovegrass. Catalina also produced more and better quality forage than Lehmann lovegrass and was suggested as a replacement for A-84 boer lovegrass and to partially replace A-68 Lehmann lovegrass (Voigt et al. 1970). In order to identify the best morphological traits to be considered in the evaluation of E. curvula hybrids, Di Renzo et al. (2000) estimated the degree of influence of permanent effects on the phenotypic variation. Crown diameter, leaf length, dry matter, and panicle number were considered in this experiment. Repeatability estimates for the vegetative characteristics indicated that environmental effects were small. For vegetative traits, two harvests provided 98% of the accuracy of the total obtained with four cuts, and for panicle number, the same percentage was obtained for three harvests. This stability of performance is a desirable characteristic for grass cultivars. These authors also studied the trait association patterns and found that leaf length was closely associated with dry matter, with a high repeatability, allowing the use of leaf length as an indirect criterion for determining the aerial biomass yield. This study was continued by Di Renzo et al. (2003) using 18 hybrids that were evaluated in three different environments in the semiarid region of Argentina (Villa Mercedes, Rı´o Cuarto and Bahı´a Blanca) in order to identify environments under which the efficiency of indirect selection could be maximized. In general, the heritability obtained was higher in Villa Mercedes and Bahı´a Blanca than in Rı´o Cuarto. The estimated decrease in efficiency of selection for all the traits measured in Rı´o Cuarto was higher than in the other locations, suggesting that this place was not a good choice for carrying out indirect selection. Conversely, genetic progress

8 Eragrostis

would be faster if selection was carried out in Villa Mercedes, because the evaluations provided higher and more accurate estimates of the heritabilities than in the other environments.

8.5.1.2 Advanced Tools Tissue culture protocols for E. curvula have been established in order to generate somaclonal variation and to assist genetic transformation (Echenique et al. 1996, 2001). Plant regeneration from four genotypes of E. curvula, cvs. Kromdraii (6x), Tanganyika (4x), Morpa (4x), and Don Pablo (7x), was reported by three developmental pathways: embryogenesis, organogenesis, and direct regeneration (Echenique et al. 1996). Immature inflorescences were the best explant source, and fertile plants were obtained from all the genotypes evaluated. The less variable apomictic cultivars, such as Morpa, Tanganyika, and Don Pablo (Poverene 1988), showed a lower regeneration response than Kromdraii, which has the highest level of sexual reproduction and is a hexaploid (2n ¼ 60). The high ploidy level could have a “buffering” effect that would minimize genetic abnormalities; the extra chromosome copies having the capacity to compensate for it. Kromdraii and Don Pablo are high polyploids, and the latter is a heptaploid with a high basal level of chromosome aberrations (Poverene 1988), which may have increased during in vitro culture (Echenique et al. 1996). Somaclones were characterized at the morphological and molecular level using randomly amplified polymorphic DNA (RAPD) markers (Polci 2000). After evaluation, three materials were selected from the apomictic tetraploid cv. Tanganyika (2n ¼ 40) (Polci 2000; Cardone et al. 2006): (1) a sexual diploid plant (2n ¼ 20) registered as Victoria (RC9192, 2006–2026, Argentina), (2) a tetraploid line (2n ¼ 40) with a high level of sexual reproduction obtained after the treatment with colchicine of the diploid line (seeds) and registered as Bahiense (RC9193, 2006–2026, Argentina), and (3) two polyploid highly apomictic plants, one of which gave rise to the apomictic cv. Don Luis (RC9191, 2006–2026, Argentina). This cultivar differs from other materials of E. curvula since the chromosomal number is 2n ¼ 64. It is an apomictic cultivar classified as robusta blue type. It differs from cvs. Ermelo, Morpa, and Don Juan because of leaf width and color (blue instead of

143

green) and differs from cv. Don Eduardo in leaf color, chromosome number, and cold resistance. The tetraploid line obtained using colchicine was grouped within the weeping type and linked to Tanganyika using molecular markers (Zappacosta 2009). The case of the diploid somaclone is peculiar, because it presents morphological and molecular differences with the other cultivars of the weeping group. Using RAPDs, AFLPs, and SSRs, it was classified within the curvula type, but at the morphological level, it exhibits similarities with the robusta type. Transformation of an apomictic clone is an attractive strategy, as the transgene is immediately fixed in a highly adapted genetic background capable of largescale clonal propagation. Some efforts have been directed towards developing a genetic transformation protocol for the genus. Ncanana et al. (2005) reported the development of plant regeneration and transformation protocols for E. curvula cv. Ermelo. Callus was generated from leaf and seed tissues and transformed by particle bombardment with the yeast Saccharomyces cerevisiae Hsp12 gene under the maize ubiquitin promoter. Although successful transformation and transcription of the Hsp12 gene occurred, no Hsp12 protein was found present in tissue extracts of the transformed grass. Using embryogenic callus obtained from immature inflorescences, Dı´az (2006) evaluated the transient expression of uidA gene introduced by particle bombardment. Three different promoters were analyzed, and the promoter of the ubiquitin gene from maize produced the best response. Other physical conditions of the bombardment device and mannitol/sorbitol concentration in the pretreatment medium were studied. Forage species showing summer growth, in general, tend to have higher fiber and lower protein contents when compared to those of temperate climates. This implies low digestibility and low animal productivity. E. curvula does not escape from this model, so that one of the main aspects to be considered in breeding programs is forage quality. Implementation of molecular technologies to downregulate lignin using both transgenic (gene technology) and non-transgenic (molecular marker technology) strategies are attractive for Eragrostis. Gene technology approaches for targeted downregulation of key enzymes [e.g., caffeic acid Omethyltransferase (OMT), cinnamyl alcohol dehydrogenase (CAD), cinnamoyl CoA reductase (CCR)] involved in lignin biosynthesis in transgenic species

144

leading to altered lignification and significantly enhanced dry matter digestibility have been developed (Chen et al. 2003; He et al. 2003; Marita et al. 2003). On the other hand, using cDNA libraries of E. curvula (Cervigni et al. 2008a) to identify genes involved in the lignin biosynthetic pathway, five transcripts from putative enzymes were identified (M. Dı´az et al. unpub results): phenylalanine ammonialyase (PAL), caffeic acid/5-hydroxyferulic acid O-methyltransferase (COMT), caffeoyl-CoA O-methyltransferase (CCoAOMT), hydroxycinnamoyl-CoA reductase (CCR), and cinnamyl alcohol dehydrogenase (CAD). The characterization of the full-length CCoAOMT cDNA, a key enzyme of the pathway, and the later genomic sequence isolation in cv. Tanganyika, allowed the identification of four alleles with a tissue-specific expression. Three alleles, A1, A2, and A4, were found to be expressed in leaves at an equal level. A1 does not represent a good candidate to be downregulated, considering that it is highly expressed in roots. The manipulation of A2 may cause an important alteration in inflorescences essential for seed production. The third allele, A4, constitutes a potential candidate to modified lignin (quality or quantity) in weeping lovegrass plants. Thus, molecular genetic markers may be used for the detection and permutation of agronomically important genetic variation in sexual populations, followed by linked marker based transfer of the apomixis character to produce new clonal varieties.

8.5.2 Improvement in Cultivated E. tef Utilizing Its Wild Relatives 8.5.2.1 Traditional Tools Phylogenetic analysis of sequence data from the nuclear gene waxy and the plastid locus rps16 (Ingram and Doyle 2003) strongly support the widely held hypothesis of a close relationship between tef and the wild allotetraploid E. pilosa. Because of this close relationship between the cultivated tef and E. pilosa, the latter has received more attention, especially from plant breeders, as compared to other wild species in the genus. Efforts to improve the yield of the cultivated E. tef have started in the 1950s, basically through pure line

M. Zeid et al.

selection from landraces, since crossing techniques were not available. Because of the small size of the flowers of tef and the lack of knowledge concerning the time of flower opening and pollen shedding, early attempts to produce hybrids were not successful (Ketema 1991). In 1974, the first successful crossing in tef was achieved (Berhe 1975) based on the observation that flowers open earlier in the morning than previously thought. Breeders have utilized this breakthrough in the hybridizing techniques to increase the yield of the crop by providing farmers with improved varieties. The urgent need for interspecific crosses between E. tef and its closest wild relatives was recognized in the 1990s, when yield improvement using intraspecific crosses began to plateau. Crosses between E. cilianensis, E. minor, E. pilosa, and E. tef were attempted (Tavassoli 1986; Gugsa and Mengiste 1999), but success rates in producing hybrid seeds were limited. E. pilosa was seen as the best candidate for such crosses because it has the same ploidy level and chromosome number as E. tef and is characterized by much shorter culms, making it a good candidate for transferring the short stature trait to reduce lodging in cultivated tef. The first crosses were developed at the Debre Zeit Agricultural Research Center in Ethiopia, and recombinant inbred lines from these cross are constantly being evaluated for useful agronomic traits (Tefera et al. 2003), and a candidate variety is currently under evaluation (G. Belay, pers. comm.). Those populations also provided the raw material for molecular marker studies aimed at linkage mapping and quantitative trait loci (QTL) analyses.

8.5.2.2 Advanced Tools The development of tissue culture techniques is an essential tool for embryo rescue for crosses involving E. tef and its close relatives. Breeding new improved E. tef varieties is becoming very difficult due to the plateau reached in yield using the available breeding material and the lack of variation in the available gene pool for lodging resistance. While crossing with E. pilosa has been successful, other interspecific crosses are desired for future improvement in E. tef. The crossing between E. tef and E. curvula as an example would be a major breakthrough, for E. curvula possesses desirable traits such as stem strength, rust resistance, and large florets to hold large seeds

8 Eragrostis

(Berhe 2001). Few in vitro plant regeneration techniques from roots, leaf bases, and seeds of E. tef were reported (Bekele et al. 1995; Mekbib et al. 1997; Kebebew et al. 1998). However, the most recent work by Gugsa et al. (2006) reports successful results from unpollinated flower explants in both E. tef and E. mexicana. Furthermore, flow-cytometry of the tissue culture-derived tef plants have indicated the regeneration of haploids, triploids, tetraploids, and octaploids with varying morphological traits as compared to the allotetraploid control plants. This is another promising tool for creating useful variation in E. tef. Further studies on E. tef and its wild relatives are needed for improvement in breeding strategies and techniques in the cultivated tef.

8.6 Genomics Resources Developed Genomic resources may be used both for gene isolation for transgenesis and for the development of geneassociated molecular markers, such as SSRs and single nucleotide polymorphisms (SNPs). Currently, only about 16,000 expressed sequence tags (ESTs) are available in Genbank (http://www.ncbi.nlm.nih.gov/, accessed 18 Nov 2009) for the genus Eragrostis. Seventy-seven percent of the ESTs were identified in E. curvula (Cervigni et al. 2008a) and the rest in E. pilosa and E. tef (Yu et al. 2006a, b). The E. curvula EST collection was generated from four cDNA libraries, three of which were obtained from panicles of near-isogenic lines with different ploidy levels and reproductive modes [described by Cardone et al. (2006)] and one obtained from 12 day-old plant leaves. A total of 12,295 high-quality ESTs were clustered and assembled, rendering 8,864 unigenes, including 1,490 contigs and 7,394 singletons, with a genome coverage of 22%. A total of 7,029 (79.11%) unigenes were functionally categorized by BLASTX analysis against sequences deposited in public databases, but only 37.80% could be classified according to Gene Ontology. Sequence comparison against the cereals genes indexes (GI) revealed 50% significant hits. A total of 254 EST-SSRs were detected from 219 singletons and 35 from contigs. Di- and trinucleotide motifs were similarly represented with percentages of 39 and 40%, respectively. In addition, 190 SNPs and insertion/deletion (INDEL) polymorph-

145

isms were detected in 18 contigs generated from three to four libraries. In addition to EST sequences, about 1,000 other nucleotide sequences are also available for 49 Eragrostis species, the majority of which are from studies on E. tef and E. pilosa aiming to develop new molecular markers, namely, SSRs (M. Zeid and M. E. Sorrells, unpub results) and conserved intron scanning primers (CISP) markers (Feltus et al. 2006) designed to utilize the information from sorghum and rice to develop genome analysis tools for orphan species. Fifty-seven protein sequences are available for 19 species mainly from phylogeny and evolutionary studies (c.f. Ingram and Doyle 2003; Bell and Columbus 2008; Christin et al. 2009). In E. curvula, a comparative expression analysis was performed based on EST sequencing and differential display using the libraries from the euploid series (Cervigni et al. 2008b). From a total of 8,884 unigenes sequenced, 112 (1.26%) showed significant differential expression in individuals with different ploidy levels and/or variable reproductive mode. Independent comparisons between plants with different reproductive modes (same ploidy) or different ploidy levels (same reproductive mode) allowed the identification of genes modulated in response to diplosporous development or polyploidization, respectively. EST sequencing allowed the identification of approximately 100 genes with differential expression associated with apomixis or ploidy level (Cervigni et al. 2008a). Interestingly, a significant number of genes were similarly expressed in the 2x sexual and the 4x apomictic lines, but differentially in the 4x sexual line. Most of those were silenced in the 2x sexual and the 4x apomictic lines. Based on these results, it was proposed that apomixis in E. curvula could be the consequence of failure in activation of a group of genes that must be expressed when ploidy level increases (Cervigni et al. 2008b).

8.7 Scope for Domestication and Commercialization A number of Eragrostis species is widely known for being grazed by domestic animals in many parts of the world, especially in Africa. Burkill (1994) listed 27

146

Eragrostis species as fodder plants that are in some cases eaten by animals for lack of better feed sources in the west tropical Africa. Furthermore, not all the species are safe for grazing. In southern and eastern Africa, some species (for example, E. plana) contain hydrocyanic acid, which is fatal for some farm animals (Watt and Breyer-Brandwijk 1962). Weeping lovegrass could be an attractive source of biomass for biofuels due to the amount of biomass it generates. McMurphy et al. (1975) found that weeping lovegrass compared favorably to switchgrass across all N rates. The maximum yield of each species was 9.4, 9.3, 6.4, and 8.4 t ha1 for switchgrass, weeping lovegrass, Indian grass, and big bluestem, respectively. Weeping lovegrass and switchgrass were the two most productive species at 82 kg of N and 18 kg of P. Furthermore, Klemp (1981) reported weeping lovegrass three years’ average yield as high as 13.3 t ha1. Other uses of Eragrostis species besides being a feed or biofuel include the use of culms for thatching and matting (E. cilianensis, E. tremula), broom making (E. gangetica, E. ciliaris), and weaving into cordage (E. ciliaris, E. pilosa, E. tremula) (Abbiw 1990; Burkill 1994). Medicinal uses are generally limited to a few species. E. ciliaris is useful as a stomachic, E. japonica and E. tremula as lactation stimulants, and E. scotelliana and roots of E. tremula are sources of aromatic substances (Burkill 1994). E. lehmanniana was used by the Europeans in today’s South Africa as a remedy for colic, diarrhea, and typhoid fever. Also, the Zulu used a decoction of E. plana roots for treating profuse menstruation, impotency, and barrenness, while the Southern Sotho used the plant as a tonic and as one of the charm ingredients in a preparation used for treating fractures (Watt and Breyer-Brandwijk 1962). In Ghana, the ash of the burnt E. ciliaris plants is mixed with shea-butter and smeared over skin burns (Dokosi 1998). Plants of E. pilosa are an effective cure for contusions, while infused leaves of E. japonica are used for headache (Duke and Ayensu 1985). Grains of many Eragrostis species including E. cilianensis, E. ciliaris, E. minor, E. pilosa, E. tenella, E. tremula, and E. turgida are usually consumed by humans in times of severe famine in many parts of Africa (Watt and Breyer-Brandwijk 1962; Burkill 1994; Dokosi 1998). Recent studies have shown that flour from grains of the cultivated E. tef lacked gluten and gluten-like proteins (Spaenij-

M. Zeid et al.

Dekking et al. 2005), making it a safe replacement for wheat flour for celiac disease patients suffering from intolerance to gluten. Tef flour also adds value to the gluten-free products due to its high vitamin and fiber content (Hopman et al. 2008). To our knowledge, no similar studies were done on the wild species of Eragrostis.

8.8 Some Dark Sides 8.8.1 The Invasive Nature of Some Eragrostis Species Seeds of E. lehmanniana originating from South Africa were first introduced to Arizona, USA, in 1932 to reseed/revegetate livestock grazing areas, construction sites, golf courses, and wildfire areas (Anable et al. 1992). At the same time, E. curvula from Tanganyika was introduced to the same region (Cox et al. 1988) and was seeded extensively for erosion control along banks and slopes of highways and mine spoils, on revegetated sites, range, and pasture sites (Alderson and Sharp 1995). The newly created Eragrostis pastures produced forage more consistently than native grass pastures, and ranchers often devoted part of their ranches to these African grasses to ensure forage supply in years with low precipitation (Cable 1971). Recent reports showed that the introduced Eragrostis species have spread beyond initially targeted areas and invaded grassland sites that had no human or animal disturbance (Williams and Baruch 2000). Adverse effects of E. lehmanniana and E. curvula on the areas it invaded were slowly being recognized. Anable et al. (1992) reported that E. lehmanniana has changed/ transformed the structure and function of large areas of the semiarid grassland in Arizona since it was introduced. It reduced faunal diversity (Bock et al. 1986) and affected faunal pedoturbation (Hupy et al. 2004). The species invaded arid grasslands, forming dense monocultures that become fire-prone in dry summer conditions (Cox 1999). These monocultures changed fire frequency and size, competed with and replaced native species, and altered geomorphological processes and hydrology (Marshall et al. 2000). Similarly, many reports, ca. 25, have considered E. curvula as a weed in countries like Australia, New Zealand,

8 Eragrostis

Mexico, Spain, and England (Global Compendium of Weeds 2007). In Japan, E. curvula has been introduced to prevent erosion. This species has established large populations, forming highly productive seed sources, especially in mountainous regions (Washitani 2004). Once riparian habitats were invaded by E. curvula, the original conditions and vegetation of gravelly floodplains, i.e., sparse vegetation cover consisting principally of riparian endemics growing in gravelly sandy conditions, were rapidly lost (Muranaka and Washitani 2003). Nakayama et al. (2007) predicted the expansion range of the species due to seed dispersal from source populations in order to plan effective measures for maintaining both ecosystem integrity and the biodiversity. While E. curvula is recognized in Argentina as a promising forage crop, it has a mixed reputation in Australia as a noxious weed, pasture grass, and soil binder (Lazarides 1997). In the United States, both E. curvula and E. lehmanniana are no longer on the conservation plant release list, and the latter was labeled “potentially invasive” (http://www.plant-materials. nrcs.usda.gov/releases/discontinued.html, accessed 18 Nov 2009). Eragrostis cumingii Steud (Cuming’s lovegrass) and E. cilianensis (Stinking lovegrass), also introduced to the United States from Africa, were declared “invasive” in natural areas (http://www. invasive.org/weedus/grass.html, accessed 18 Nov 2009). The government in New Zealand declared E. curvula among the “unwanted organisms” along with 140 other plant species (http://www.biosecurity.govt. nz/pests-diseases/plants/accord/amending-list.htm, accessed 18 Nov 2009). Control of E. curvula plants through prescribed fire was not successful. E. curvula benefits from fire, generally increasing (Wright et al. 1978) or remaining stable in numbers after fire (Walsh 1994), deep roots and dense crown of growth protects the plant from fire damage (Phillips et al. 1991). E. Lehmanniana also has high potential to reestablish after fire (Sumrall et al. 1991). High seedling emergence after fire could be the result of a greater range in diurnal soil temperatures and increases in red light reaching the seedbed, both of which stimulate seed germination. The Southwest Biological Science Center of the University of Arizona, USA, recommended some control methods and management strategies for Eragrostis. Control could be achieved by digging out the root system in the late winter to early spring. A prescribed fire following this

147

effort will kill the remaining seeds (Invasive Alien Plants Species of Virginia 1999). Selective herbicides or herbicides accompanied by timed and controlled burns following the reseeding of native species are other options for the control of Eragrostis species and promoting native vegetation (Campbell et al. 1985; Halvorsen and Guertin 2003a, b).

8.9 Recommendations for Future Actions Eragrostis species have made many important contributions to human kind, and the wide range of genetic variation in the genus is an indication that there is much potential for novel uses in the future. Clearly, the extensive variation has played an important role in the ability of these species to tolerate environmental extremes and to adapt to forage cropping and perhaps eventually biofuel production. Eragrostis has also been used for many medicinal, nutritional, and functional purposes. The exploitation of this variation can be facilitated through apomictic breeding to release new genetic variation, thus producing genotypes with new combinations of traits. In addition, transformation of an apomictic clone is a useful approach because the transgene is immediately fixed in an elite genotype that is capable of large-scale clonal propagation. Finally, it may eventually be possible to transfer some of the genetic variation underlying useful traits in these species, such as apomixis, to other cultivated species. However, to realize some of these potential benefits, much additional research is needed to develop the necessary information base and molecular tools.

References Abbiw DK (1990) Useful plants of Ghana: West African uses of wild and cultivated plants. Intermediate Technology Publications and the Royal Botanic Gardens, Kew, UK Alderson JS, Sharp WC (1995) Grass varieties in the United States. Agricultural handbook, No 170. Soil Conservation Service, United States Department of Agriculture, Washington, DC Anable ME, McClaran MP, Ruyle GB (1992) Spread of introduced Lehmann lovegrass Eragrostis lehmanniana Nees in southern Arizona, USA. Biol Conserv 61:181–188

148 Ayele M, Nguyen HT (2000) Evaluation of amplified fragment length polymorphism markers in tef, Eragrostis tef (Zucc.) Trotter, and related species. Plant Breed 119:403–409 Ayele M, Dolezel J, Van Duren M, Brunner H, Zapata-Arias FJ (1996) Flow cytometric analysis of the nuclear genome of the Ethiopian cereal Tef [Eragrostis tef (Zucc.) Troter]. Genetica 98:211–215 Ayele M, Tefera H, Assefa K, Nguyen HT (1999) Genetic characterization of two Eragrostis species using AFLP and morphological traits. Hereditas 130:33–40 Bai GH, Tefera H, Ayele M, Nguyen HT (1999a) A genetic linkage map of tef [Eragrostis tef (Zucc.) Trotter] based on amplified fragment length polymorphism. Theor Appl Genet 99:599–604 Bai GH, Ayele M, Tefera H, Nguyen HT (1999b) Amplified fragment length polymorphism analysis of tef [Eragrostis tef (Zucc.) Trotter]. Crop Sci 39:819–824 Bai GH, Ayele M, Tefera H, Nguyen HT (2000) Genetic diversity in tef [Eragrostis tef (Zucc) Trotter] and its relatives as revealed by random amplified polymorphic DNAs. Euphytica 112:15–22 Balsamo R, Vander Willingen C, Bauer A, Farrant J (2006) Drought tolerance of selected Eragrostis species correlates with leaf tensile properties. Ann Bot 97:985–991 Bekele E, Klock G, Zimmermann U (1995) Somatic embryogenesis and plant regeneration from leaf and root explants and from seeds of Eragrostis tef (Gramineae). Hereditas 123:183–189 Bell HL, Columbus JT (2008) Proposal for an expanded Distichlis (Poaceae, Chloridoideae): support from molecular, morphological, and anatomical characters. Syst Bot 33: 536–551 Bennett MD, Smith JB (1976) Nuclear DNA amounts in angiosperms. Philos Trans R Soc Lond B 274:227–274 Berhe T (1975) A breakthrough in tef-breeding technique. FAO Information Bulletin, The Near East Cereal Improvement and Production Project 12:11–13 Berhe T (2001) Closing remarks. In: Tefera H, Belay G, Sorrells M (eds) Narrowing the Rift: Tef research and development. Ethiopian Agricultural Research Organization, Addis Ababa Bock CE, Bock JH, Jepson KL, Ortega JC (1986) Ecological effects of planting African lovegrass in Arizona. Natl Geogr Res 2:456–463 Boechat SC, Longhi-Wagner HM (2000) Padroes de distribuicao geografica dos taxons eiros de Eragrostis (Poaceae, Chloridoideae). Revista Brasil Bot (Sao Paulo) 23:177–194 Burkill HM (1994) The useful plants of West Tropical Africa, vol 2, Families E–I. Royal Botanic Gardens, Kew, UK Busso C, Brevedan R (1991) Nutricio´n mineral. In: Ferna´ndez OA, Brevedan RE, Gargano A (eds) El pasto Iloro´n. Su biologı´a y manejo. CERZOS, Bahı´a Blanca, pp 57–126 Cable DR (1971) Lehmann lovegrass on Santa Rita experimental range, 1937–1968. J Range Manage 24:17–21 Campbell MH, Dellow JJ, Keys MJ, Gilmour AR (1985) Use of herbicides for selective removal of Eragrostis curvula (Schrad.) Nees from a Phalaris aquatica pasture. Aust J Exp Agric 25:665–671 Cardone S, Polci P, Selva JP, Mecchia M, Pessino S, Hermann P, Cambi V, Voigt P, Spangenberg G, Echenique V (2006) Novel genotypes of the subtropical grass Eragrostis curvula

M. Zeid et al. for the study of apomixes (diplospory). Euphytica 151: 263–272 Cervigni GD, Paniego N, Diaz M, Selva JP, Zappacosta D, Zanazzi D, Landerreche I, Martelotto L, Felitti S, Pessino S, Spangenberg G, Echenique V (2008a) Expressed sequence tag analysis and development of gene associated markers in a near-isogenic plant system of Eragrostis curvula. Plant Mol Biol 67:1–10 Cervigni GD, Paniego N, Pessino S, Dı´az ML, Selva JP, Zappacosta D, Spangenberg G, Echenique VC (2008b) Gene expression in diplosporous and sexual Eragrostis curvula genotypes with differing ploidy levels. Plant Mol Biol 67:11–23 Chanyalew S, Singh H, Tefera H, Sorrels M (2005) Molecular genetic map and QTL analysis of agronomic traits based on a Eragrostis tef x E-pilosa recombinant inbred population. J Genet Breed 59:53–66 Chen L, Auh P, Dowling J, Bell F, Chen A, Hopkins R, Dixon R, Wang Z (2003) Improved forage digestibility of tall fescue (Festuca arundinacea) by transgenic down-regulation of cinamyl alcohol dehydrogenase. Plant Biotechnol J 1:1–13 Christin PA, Petitpierre B, Salamin N, Buchi L, Besnard G (2009) Evolution of C4 phosphoenolpyruvate carboxykinase in grasses, from genotype to phenotype. Mol Biol Evol 26:357–365 Clayton WD, Harman KT, Williamson H (2006) GrassBase – The Online World Grass Flora. http://www.kew.org/data/ grasses-db.html. Accessed 5 Mar 2009 Colom MR, Vazzana C (2001) Drought stress effects on three cultivars of Eragrostis curvula: photosynthesis and water relations. Plant Growth Regul 34:195–202 Colom MR, Vazzana C (2003) Photosynthesis and PSII functionality of drought-resistant and drought-sensitive weeping lovegrass plants. Environ Exp Bot 49:135–144 Cope TA (1998) A synopsis of Eragrostis Wolf (Poaceae) in the flora Zambesiaca area. Kew Bull 53:129–164 Costanza SH, Dewet JMJ, Harlan JR (1979) Literature-review and numerical taxonomy of Eragrostis tef (t’ef). Econ Bot 33:413–424 Covas G (1991) Taxonomı´a y morfologı´a del pasto Iloro´n, Eragrostis curvula (Schrad. Nees) con referencia sobre otras especies cultivadas de Eragrostis. In: Ferna´ndez OA, Brevedan RE, Gargano A (eds) el pasto Iloro´n. Bahı´a Blanca, Su biologı´a y manejo, CERZOS, pp 7–17 Covas G, Cairnie AG (1985) El pasto lloro´n. Manual con informacio´n ba´sica y normas para su cultivo y utilizacio´n. Hemisferio Sur, Buenos Aires, Argentina Cox GW (1999) Alien species in North America and Hawaii: impacts on natural ecosystems. Island Press, Washington, DC Cox JR, Martin MH, Ibarra FA, Fourie JH, Rethman NFG, Wilcox DG (1988) The influence of climate and soils on the distribution of four African grasses. J Range Manage 41:127–139 Cufodontis G (1974) Enumeratio plantarum aethiopiae spermatophyta. Jard Bot, Brussels De Winter B (1955) Eragrostis Beauv. In: Chippindall LK (ed) The grasses and pastures of South Africa. Central News Agency, Johannesburg, S Africa, pp 132–184 Di Renzo M, Iban˜ez M, Bonamico N, Poverene M (2000) Estimation of repeatability and phenotypic correlations in Eragrostis curvula. J Agric Sci 134:207–212

8 Eragrostis Di Renzo M, Iban˜ez M, Bonamico N, Faricelli M, Poverene M, Echenique V (2003) Effect of three environments on the efficiency of indirect selection in Eragrostis curvula (lovegrass) genotypes. J Agric Sci 140:427–433 Dı´az M (2006) Biotecnologı´a aplicada al mejoramiento del pasto lloro´n, Eragrostis curvula (Schrad.) Nees. Tesis de Doctorado en Biologı´a. Universidad. Nacional del Sur. Argentina, pp 174 Dokosi OB (1998) Herbs of Ghana. Ghana Unive Press, Accra Duke JA, Ayensu ES (1985) Medicinal plants of China, vol 2. Reference Publ, Algonac, MI Echenique V, Curvetto N (1986) Efectos del de´ficit hı´drico en cinco cultivares de pasto lloro´n, Eragrostis curvula (Schrad.) Nees sens. lat. Niveles de clorofila y prolina y permeabilidad de membranas celulares. Phyton 46:195–206 Echenique V, Polci P (1994) Efecto del estre´s hı´drico en el estado reproductivo en pasto lloro´n, Eragrostis curvula (Schrad.) Nees. Turrialba 44:189–204 Echenique V, Poverene M, Curvetto N (1986a) Epicuticular wax and water stress in Eragrostis curvula cultivars. I. Morphology and distribution. Revista de Microscopı´a Electro´nica y Biologı´a Celular 10:77–89 Echenique V, Poverene M, Curvetto N (1986b) Epicuticular wax and water stress in Eragrostis curvula cultivars. II. Effect of water stress on the quantity and morphology of epicuticular wax. Revista de Microscopı´a Electro´nica y Biologı´a Celular 10:91–99 Echenique V, Polci P, Mroginski L (1996) Plant regeneration in weeping lovegrass (Eragrostis curvula) through inflorescence culture. Plant Cell Tissue Organ Cult 46: 123–130 Echenique V, Dı´az M, Polci P, Mroginski L (2001) Embryogenic cell suspensions from different explants and cultivars of Eragrostis curvula (Schrad.) Nees. Biocell 25: 131–138 Feltus FA, Singh HP, Lohithaswa HC, Schulze SR, Silva TD, Paterson AH (2006) A comparative genomics strategy for targeted discovery of single-nucleotide polymorphisms and conserved-noncoding sequences in orphan crops. Plant Physiol 140:1183–1191 Gherbin P, De Franchi AS, Monteleone M, Rivelli AR (2007) Adaptability and productivity of some warm-season pasture species in a Mediterranean environment. Grass Forage Sci 62:78–86 Gugsa L, Mengiste T (1999) The crossability of Eragrostis tef (Zucc.) trotter with its wild close relatives. In: Grima B, Alemayehu N, Gebeyehu G, Dibabe A, Wakjira A, Dheressa A (eds) Proceedings of the annual conference of the crop science society of Ethiopia, Addis Abebe, pp 57–60 Gugsa L, Sarial AK, Lo¨rz H, Kumlehn J (2006) Gynogenic plant regeneration from unpollinated flower explants of Eragrostis tef (Zuccagni) Trotter. Plant Cell Rep 25:1287–1293 Halvorsen WL, Guertin P (2003a) Fact sheet for: Eragrostis curvula (Schrad.) Nees and Eragrostis curvula var. conferta Stapf. USGS Weeds in the West project: status of introduced plants in Southern Arizona Parks Sonoran Desert Field Station University of Arizona, Tuson, USA. http://sdrsnet.srnr. arizona.edu/data/sdrs/ww/docs/eraglehm.pdf Halvorsen WL, Guertin P (2003b) Fact sheet for: Eragrostis lehmanniana Nees. USGS Weeds in the West project: status of introduced plants in Southern Arizona Parks Sonoran

149 Desert Field Station University of Arizona, Tuson, USA. http://sdrsnet.srnr.arizona.edu/data/sdrs/ww/docs/eraglehm.pdf Hanson A (1965) Grass varieties in the United States, vol 170, USDA-ARS, Agri Handbook. US Govt Print Office, Washington, DC Hartley W, Slater C (1960) Studies on the origin, evolution, and distribution of the Gramineae. III. The tribes of the subfamily Eragrostoideae. Aust J Bot 8:256–276 He X, Hall M, Gallo-Meagher M, Smith R (2003) Improvement of forage quality by downregulation of maize O-methyltransferase. Crop Sci 43:2240–2251 Herna´ndez O (1991) Manejo del cultivo y respuesta al pastoreo. In: Ferna´ndez OA, Brevedan RE, Gargano A (eds) El Pasto Iloro´n. Su Biologı´a y Manejo. CERZOS, Bahı´a Blanca, pp 277–324 Hilu KW, Alice LA (2000) Phylogenetic relationships in subfamily Chloridoideae (Poaceae) based on matK sequences. A preliminary assessment. In: Jacobs SWL, Everett J (eds) Grasses: systematics and evolution. CSIRO, Melbourne, Australia, pp 184–188 Hopman E, Dekking L, Blokland ML, Wuisman M, Zuijderduin W, Koning F, Schweizer J (2008) Tef in the diet of celiac patients in The Netherlands. Scand J Gastroenterol 43: 277–282 Hupy CM, Whitford WG, Jackson EC (2004) The effect of dominance by an alien grass species, Lehmann lovegrass, Eragrostis lehmanniana, on faunalpedoturbation patterns in North American Desert grasslands. J Arid Environ 58: 321–334 Ingram AL, Doyle JJ (2003) The origin and evolution of Eragrostis tef (Poaceae) and related polyploids: evidence from nuclear waxy and plastid rps16. Am J Bot 90:116–122 Ingram AL, Doyle JJ (2004) Is Eragrostis (Poaceae) monophyletic? Insights from nuclear and plastid sequence data. Syst Bot 29:545–552 Invasive Alien Plants Species of Virginia (1999) Weeping lovegrass (Eragrostis curvula (Schrader) Nees). Department of Conservation and Recreation and the Virginia Jacobs SWL (1982) Classification of the Eragrostis curvula complex in Australia. Aust Plant Intr Rev 15:5–14 Jones B, Ponti J, Tavassoli A, Dixon P (1978) Relationships of the Ethiopian cereal Tef (Eragrostis tef ( Zucc.) Trotter): evidence from morphology and chromosome number. Ann Bot 42:1369–1373 Kebebew A, Gaj MD, Maluszynski M (1998) Somatic embryogenesis and plant regeneration in callus culture of tef (Eragrostis tef [Zucc.] Trotter). Plant Cell Rep 18:156–158 Keeler K (2004) Impact of intraspecific polyploidy in Andropogon gerardii (Poaceae) Populations. Am Midl Nat 152:63–74 Kefyalew T, Tefera H, Assefa K, Ayele M (2000) Phenotypic diversity for qualitative and phenologic characters in germplasm collections of tef (Eragrostis tef). Genet Resour Crop Evol 47:73–80 Ketema S (1991) Germplasm evaluation and breeding work on tef Eragrostis tef in Ethiopia. In: Engels JMM, Hawkes JG, Worede M (eds) Plant genetic resources of Ethiopia. Cambridge Univ Press, Cambridge, England, pp 323–328 Klemp RL (1981) Does weeping lovegrass fit into your scheme? Forage Production Cattlemen. Fort Worth, Texas and Southwestern Cattle Raisers Assoc 68(3):124–126

150 Laegaard S, Peterson PM (2001) Eragostis Wolf. In: Harling G, Andersson L (eds) Gramineae (part 2) subfam. Chloridoideae, Flora of Ecuador, pp 25–55 Lazarides M (1997) A revision of Eragrostis (Eragrostideae, Eleusininae, Poaceae) in Australia. Aust Syst Bot 10:77–187 Leigh JH (1960) Some aspects of anatomy, ecology and physiology of Eragrostis. PhD Thesis. Univ of Witwatersrand, Johannesburg, South Africa, 259 p Leigh JH, Davidson RL (1968) Eragrostis curvula (Schrad). Nees and some other African lovegrasses. Aust Plant Intr Rev 5:21–44 Marita J, Ralph J, Hatfield R, Guo D, Chen F, Dixon R (2003) Structural and compositional modifications in lignin of transgenic alfalfa down-regulation in caffeic acid 3-Omethyltransferase and caffeoyl coenzyme A 3-O-methyltransferase. Phytochemistry 62:53–65 Marshall RM, Anderson S, Batcher M, Comer P, Cornelius S, Cox R, Gondor A, Gori D, Humke J, Aguilar RP, Parra IE, Schawartz S (2000) An Ecological analysis of conservation priorities in the Sonoran desert ecoregion. The Nature Conservancy Arizona Chapter, Sonoran Institute, and Instituto Del Medio Ambiente Y El Desarrollo Sustentable Del Estado De Sonora McMurphy WE, Denaman CE, Tucker BB (1975) Fertilization of native grasses and weeping lovegrass. Agron J 67:233–236 Mekbib F, Mantell SH, Buchanan-Wollaston V (1997) Callus induction and in vitro regeneration of tef [Eragrostis tef (Zucc.) Trotter] from leaf. J Plant Physiol 151:368–372 Muranaka T, Washitani I (2003) The population expansion predicted by a simulation model of an invasive alien species, Eragrostis curvula, in a middle-reach floodplain (in Japanese). Jpn J Conserv Ecol 8:51–62 Nakayama N, Nishihiro J, Kayaba Y, Muranaka T, Washitani I (2007) Seed deposition of Eragrostis curvula, an invasive alien plant on a river floodplain. Ecol Res 22:696–701 Native Plant Society. http://www.dcr.virginia.gov/natural_ heritage/documents/fsercr.pdf Ncanana S, Brandt W, Lindsey G, Farrant J (2005) Development of plant regeneration and transformation protocols for the desiccation-sensitive weeping lovegrass Eragrostis curvula. Plant Cell Rep 24:335–340 Peterson PM (2003) Eragrostis. In: Barkworth ME, Capels KM, Long S, Piep MB (eds) Magnoliophyta: Commelinidae (in part): Poaceae, part 2, vol 25, Flora of North America north of Mexico. Oxford University Press, New York, pp 65–105 Peterson P, Sanchez Vega I (2007) Eragrostis (Poaceae: Chloridoideae: Eragrostideae: Eragrostidinae) of Peru. Ann Mo Bo Gard 94:745–790 Peterson PM, Valdes-Reyna J (2005) Eragrostis (Poaceae: Chloridoideae: Eragrostideae: Eragrostidinae) from northeastern Mexico. Sida 21:1365–1420 Peterson P, Webster R, Valdes-Reyna J (1997) Genera of new World Eragrostideae (Poaceae: Chloridoideae). Smithsonian Contrib Bot 87. Smithsonian Inst Press, Washington, DC Phillips S, Brown C, Cole C (1991) Weeping lovegrass, Eragrostis curvula (Schrader) Nees Von Esenbeck, as a harborage of arthropods on the Texas high plains. Southwest Nat 36:49–53 Polci P (2000) Cultivo de tejidos para la obtencio´n de variantes somaclonales de pasto lloro´n, Eragrostis curvula (Schrad.) Nees. Tesis de Doctorado en Biologı´a. Universidad Nacional del Sur, Argentina, p 234

M. Zeid et al. Poverene M (1988) Contribucio´n Citogene´tica y Quimiosistema´tica a la Taxonomı´a del Pasto Lloro´n, Eragrostis curvula (Schrad.), Nees. Tesis de Doctorado en Biologı´a. Universidad. Nacional del Sur, Argentina, p 218 Poverene M, Curvetto N (1991) Citogene´tica. In: Ferna´ndez OA, Brevedan RE, Gargano A (eds) El Pasto Iloro´n. Su Biologı´a y Manejo. CERZOS, Bahı´a Blanca, pp 19–38 Poverene MM, Voigt PW (1997) Isozyme variation and germplasm relationships in the Eragrostis curvula Complex. Biochem Syst Ecol 25:21–32 Puliga S, Vazzana C, Davies W (1996) Control of crops leaf growth by chemical and hydraulic influences. J Exp Bot 297:529–537 Quarin CL, Ferna´ndez A (1982) Genetic studies in diploid and tetraploid Paspalum species. J Hered 76:254–256 Quarin CL, Pozzobon MT, Valls JFM (1996) Cytology and reproductive behavior of diploid, tetraploid and hexaploid germplasm accessions of a wild forage grass: Paspalum compressifolium. Euphytica 90:345–349 Renvoize SA (1998) Gramineas de Bolivia. The Royal Botanic Gardens, Kew, UK Roodt-Wilding R, Spies J (2006) Phylogenetic relationships in southern African chloridoid grasses (Poaceae) based on nuclear and chloroplast sequence data. Syst Biodiver 4: 401–415 Ruiz MA, Golberg AD, Martı´nez O (2008) Water stress and forage production in Tetrachne dregei Nees, Panicum coloratum L. and Eragrostis curvula (Schrad) Nees. Phyton 77:7–20 Spaenij-Dekking L, Kooy-Winkelaar Y, Koning F (2005) The Ethiopian Cereal Tef in Celiac Disease. N Engl J Med 353:1748–1749 Spies JJ, Gibbs Russell GE (1988) Variation in important pasture grasses: II. Cytogenetic and reproductive variation. J Grassland Soc S Afr 5:22–25 Streetman LJ (1970) Cytogenetics of Eragrostis. In: Dalrymple RL (ed) Proceedings of the first weeping lovegrass symposium. The Samuel Roberts Noble Foundation, Ardmore, OK, pp 10–13 Sumrall LB, Roundy BA, Cox JR, Winkel VK (1991) Influence of canopy removal by burning or clipping on emergence of Eragrostis lehmanniana. Int J Wildland Fire 1:35–40 Taliaferro CM, Horn FP, Tucker BB, Totusek R, Morrison RD (1975) Performance of three warm-season perennial grasses and a native range mixture as influenced by N and P fertilization. Agron J 67:289–292 Tavassoli A (1986) The cytology of Eragrostis with special reference to E. tef and its relatives. PhD Thesis, Royal Holloway College, University of London, UK Tefera H, Ketema S, Tesemma T (1990) Variability, heritability and genetic advance in tef (Eragrostis tef (Zucc) Trotter) cultivars. Trop Agric 67:317–320 Tefera H, Assefa K, Belay G (2003) Evaluation of interspecific recombinant inbred lines of Eragrostis tef  E. pilosa. J Genet Breed 57:21–30 USDA-ARS (2009) National Genetic Resources Program, Germplasm Resources Information Network (GRIN Online Database) National Germplasm Resources Laboratory, Beltsville, MD, USA. http://www.ars-grin.gov. Accessed 19 Nov 2009

8 Eragrostis Van den Borre A, Watson L (1994) The infrageneric classification of the Eragrostis (Poaceae). Taxon 43:383–422 Veldkamp JF (2002) Revision of Eragrostis (Gramineae, Chloridoideae) in Malesia. Blumea 47:157–204 Voigt P (1984) Breeding apomictic lovegrasses: forage potential of boer x weeping hybrids. Crop Sci 24:115–118 Voigt P (1991) Eragrostis curvula: sus caracterı´sticas y potencial para el mejoramiento a trave´s de la hibridacio´n. In: Ferna´ndez OA, Brevedan RE, Gargano A (eds) El Pasto Iloro´n. Su Biologı´a y Manejo. CERZOS, Bahı´a Blanca, pp 39–56 Voigt PW, Kneebone WR, McIlvain EH, Shoop MC, Webster JE (1970) Palatability, chemical composition, and animal gains from selections of weeping lovegrass, Eragrostis curvula (Schrad.) Nees. Agron J 62:673–676 Voigt P, Rethman N, Poverene M (2004) Lovegrasses. Warmseasons (C4) grasses. Agron Monogr 45:1027–1055 von Wolf NM (1776) Genera plantarum vocabulis characteristicis definita. Published by the author, Konigsberg Vorster T, Liebenberg H (1977) Cytogenetic studies in the Eragrostis curvula complex. Bothalia 12:215–221 Walsh R (1994) Eragrostis curvula. In: Fire effects information system. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fire Sciences Laboratory, 2002. http://www.fs.fed.us/database/feis/plants/graminoid/eracur/ all.html Washitani I (2004) Invasive alien species problems in Japan: an introductory ecological essay. Glob Environ Res 8:1–11

151 Watt JM, Breyer-Brandwijk MG (1962) The medicinal and poisonous plants of Southern and Eastern Africa. E. and S. Livingstone, Edinburgh Williams DG, Baruch Z (2000) African grass invasion in the americas: ecosystem consequences and the role of ecophysiology. Biol Invasions 2:123–140 Wright HA, Bailey AW, Thompson RP (1978) The role and use of fire in the Great Plains: a state-of-the-art review. In: Proceedings: Prairie prescribed burning symposium and workshop. The Wildlife Society, North Dakota Chapter, Jamestown, North Dakota, pp 8:1–29 Yu J, Sun Q, La Rota M, Edwards H, Tefera H, Sorrells ME (2006a) Expressed sequence tag analysis in tef (Eragrostis tef (Zucc) Trotter). Genome 49:365–372 Yu JK, Kantety RV, Graznak E, Benscher D, Tefera H, Sorrells ME (2006b) A genetic linkage map for tef [Eragrostis tef (Zucc.) Trotter]. Theor Appl Genet 113:1093–1102 Yu JK, Graznak E, Breseghello F, Tefera H, Sorrells ME (2007) QTL mapping of agronomic traits in tef [Eragrostis tef (Zucc) Trotter]. BMC Plant Biol 7:13 Zappacosta D (2009). Contribucio´n al conocimiento de la taxonomı´a y del modo reproductivo del pasto lloro´n Eragrostis curvula (Schrad.) Nees. Doctoral Thesis, Universidad Nacional del Sur. Bahia Blanca, Argentina Zhang D, Ayele M, Tefera H, Nguyen HT (2001) RFLP linkage map of the Ethiopian cereal tef [Eragrostis tef (Zucc) Trotter]. Theor Appl Genet 102:957–964

.

Chapter 9

Festuca Toshihiko Yamada

9.1 Basic Botany of the Species The genus Festuca L. is one of the largest in Gramineae and is, along with Poa L., the largest genus of the tribe Poeae. Festuca L. and its closely allied genus Lolium L. have long fascinated agronomists, evolutionists, and plant breeders, and these genera are among the most widely studied of the non-cereal grasses. Recent evidence from molecular phylogenetic studies shows that the genus lacks monophyly. As a result, several species, including the forage grasses, tall fescue (Fig. 9.1), Festuca arundinacea Schreb. [¼syn. Lolium arundinaceum (Schreb.) Darbysh.] and meadow fescue, Festuca pratensis Huds. [¼syn Lolium pratense (Huds.) Darbysh.], formerly belonging to the genus Festuca, have been recently placed into the genus Lolium (Darbyshire 1993). However, these species are described here as belonging to the genus Festuca. Festuca L. is a large, diverse genus whose members are widely adapted to a variety of ecogeographical regions. It comprises some 450 species (Clayton and Renvoize 1986) that range from diploid (2n ¼ 2x ¼ 14) to dodecaploid (2n ¼ 12x ¼ 84) in chromosome number (Sˇmarda and Stancˇ´ık 2006). A genus of Festuca L. is characterized with morphological features as follows: perennial, tufted to something short-rhizomatous; leaf-blades typically parallel-veined, mostly rolled, something flat; leaf sheaths open; ligules short, membranous; auricles present or absent; inflorescence a compressed or lax

T. Yamada Field Science Center for Northern Biosphere, Hokkaido University, Kita 11, Nishi 10, Kita-ku, Sapporo 060-0811, Japan e-mail: [email protected]

panicle; spikelets two – many-flowered, disarticulating above the glumes; glumes shorter than lemma; upper glume usually three-nerved; lemmas membranous to thinly coriaceous, usually glabrous, fivenerved, acute, awnless or awn-tipped; palea almost equal to lemma, scabrous, or ciliate; lodicules two; stamens three; ovary sometimes hairy on the top (Jauhar 1993). Classification of the almost cosmopolitan genus Festuca has varied through the last two centuries. Hackel (1882) subdivided Festuca into six sections: Ovinae, Bovinae, Sub-bulbosae, Variae, Scariosae, and Montanae based on vegetative and floral characters. Within sect. Ovinae and Variae, subgroups were characterized by intravaginal versus extravaginal innovation. Saint-Yves (1922) largely adopted this system for his worldwide revisions of the genus. However, Cvelev (1971, 1976) and Alexeev (1977, 1981, 1982, 1984, 1988) proposed a substantially new system of classification with 11 subgenera, many of which were further divided into sections. The molecular systematic studies using ITS/trnL-F provided new insights in the phylogeny of subtribe Loliinae of tribe Poeae (Catala´n et al. 2004; Torrecilla et al. 2004). The subtribe can be roughly classified into two major lineages: (1) the poorly supported “broad-leaved” fescues, falling into several lineages of unclear relations and (2) the highly supported “fine-leaved” fescues containing the majority of Festuca species. Two subgroups of Festuca recognized by Hackel (1882) fall into different lineages, raising the need to recognize two new sections. M€uller and Catala´n (2006) described two sections of Lojaconoa and Dimorphae based on results of recent phylogenetic studies. Because of the diversification of the ploidy levels in various taxonomically intricate groups, ploidy level is

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_9, # Springer-Verlag Berlin Heidelberg 2011

153

154

T. Yamada

Fig. 9.1 Tall fescue (Festuca arundinacea)

an important species character in recent Festuca systematic treatment. Recent studies have shown that flow cytometry is useful for the determination of DNA ploidy level in fresh and herbarium materials of Festuca (Sˇmarda and Kocˇ´ı 2005; Sˇmarda et al. 2005, 2008b; Sˇmarda 2006, 2008; Sˇmarda and Stancˇ´ık 2006). The measurement of DNA content (2C-value), monoploid genome size (Cx-value), average chromosome size (C/n-value), and cytosine + guanine (GC) content using flow cytometry in 101 Festuca taxa and 14 of their close relatives facilitated understanding the long-term processes of genome evolution, testing evolutionary hypotheses, and their usefulness for largescale genomic projects, suggesting that there is an evolutionary advantage for small genomes in Festuca (Sˇmarda et al. 2008a). The genus Festuca contains two agriculturally important forage crops, hexaploid tall fescue (F. arundinacea) (2n ¼ 6x ¼ 42) and diploid meadow fescue (F. pratensis) (2n ¼ 2x ¼ 14). Other fescues of some importance are red fescue, F. rubra L. (2n ¼ 6x ¼ 42 or 8x ¼ 56), and sheep fescue, F. ovina L. (2n ¼ 4x ¼ 28). The agriculturally

undeveloped species, giant fescue, F. gigantea (L.) Vill. [¼syn. Lolium giganteum (L.) Darbysh.; 2n ¼ 6 x ¼ 42], has acquired importance because it produces a mass of very large soft leaves of high nutritive value (Thomas and Humphreys 1991). Further, grass breeders have interest on some species, such as diploid F. altissima All., F. donax Lowe and F. drymeja Mert. and W.D.J. Koch, tetraploid F. mairei St. Yves, F. pratensis var. apennina (De Not.) Hack. and F. arundinacea var. glaucescens Boiss., and decaploid F. arundinacea var. letourneuxiana St. Yves (Thomas and Humphreys 1991). Tall fescue is the most important forage species worldwide of the Festuca genus. It is indigenous to Europe and also occurs naturally on the Baltic coasts throughout the Caucasus in western Siberia and extending into China (Jauhar 1993). Tall fescue introductions have been made into North and South America, Australia, New Zealand, Japan, and South and East Asia (Barnes 1990). It was introduced into USA from Europe in the early 1800s. It is widely grown in southern Europe and is the predominant cool-season perennial grass species in the USA,

9 Festuca

especially in the transition between cool-temperature and subtropical zones (Sleper and West 1996). Meadow fescue is one of the most widely used forage grasses in the Scandinavia countries, due to its generally high level of winter stress tolerance. It is distributed natively in Europe and western Asia (Fjellheim et al. 2007). Although not widely used as a forage grass species outside Scandinavia, there has been an increasing interest in meadow fescue during the last decade in Europe, North America, and Japan. In North America, the acreage of meadow fescue has not been significant since the early twentieth century (Casler and van Santen 2000), because extensive testing of forage yield capacity around the turn of the nineteenth century (Buckner et al. 1979) indicated that tall fescue was superior to meadow fescue in vigor and resistance to crown rust (caused by Puccinia coronata). However, recent research has shown that meadow fescue may be more useful than tall fescue in intensive grazing management systems in North America (Casler et al. 1998). In eastern Hokkaido, Japan, which has severe winter climate, meadow fescue is promising as a grazing species. Therefore, breeding program of meadow fescue was started and winter-hardy and high-yielding variety “Makibasakae” has been released recently (Tase et al. unpub). Fine-leaved Festuca spp. such as F. rubra (red fescue) and F. ovina (sheep fescue) are used for turf species (Ruemmele et al. 2003). Fine-leaved Festuca spp. are generally characterized by their fine to very narrow leaves, usually less than 1 mm (Beard 1973). They have been grown on greens of golf course in Scotland for centuries (Beard 1998). Use on fairway increases with the desire for more environmentally sustainable golf courses (Christians 2000).

9.2 Conservation Initiatives Temperate grasses including Festuca species are the main component of permanent grasslands. Genetic diversity present in permanent grasslands is threatened by intensification of forage production. This is basically a consequence of an increasing use of external inputs of fertilizers and large-scale resowing with improved cultivars of only few species (Brown 1992; Tscharntke et al. 2005). The importance of genetic

155

diversity for future breeding objectives as well as for long-term stability of ecosystems gave rise to two conservation strategies for genetic resources (CBD 2005). Ex situ conservation includes the storage in gene banks of germplasm collections intended to best represent genetic diversity. In breeding station, superior genotypes are maintained with vegetative propagation. Ex situ conservation can be effective for protecting species or small number of threatened varieties or ecotypes. However, it is not very effective for maintaining the vast genetic diversity characteristic of the species and are very expensive, especially when live plants are maintained (Peeters 2004). In contrast, in situ conservation includes the maintenance of plants at their specific habitats (e.g., permanent grassland), allowing continuing evolutionary adaptation (Frankel et al. 1995; Maxted et al. 1997). Ecological factors and agricultural practices have created a vast biodiversity that can only be conserved by protecting the habitats and using management methods close to those that created that diversity (Peeters 2004). Natural or semi-natural permanent grasslands harbor forage grass ecotype populations, which are highly adapted to their individual macro- and micro-habitats, showing high variation for traits of adaptive significance such as ear emergence, growth habit, or resistance to various diseases (Wilkins 1991; Fjellheim et al. 2007). This wealth of variation has been used as the main source of genetic variation for forage crop breeding since its beginning in Europe in the early twentieth century (Humphreys 2005). Several studies have investigated ecotype populations of different forage grass species by means of molecular markers (Fjellheim and Rognli 2005; Peter-Schmid et al. 2008a), mostly showing a high variability within the populations, as it is expected for obligate outbreeders with a gametophytic self-incompatibility system. Recent studies indicated that ecotype populations may still allow to improve new varieties with favorable alleles, which do not exist in present cultivars (Fjellheim et al. 2007; Peter-Schmid et al. 2008a). Peter-Schmid et al. (2008b) demonstrated that ecotype populations and cultivars were clearly separated, and there was a significant correlation between diversity of morphological and geographic location of samplings sites. Maintenance of permanent grassland in contrasting environments appears to be a promising strategy for preserving valuable genetic variation of forage grasses (Peter-Schmid et al. 2008b).

156

T. Yamada

9.3 Role in Elucidation of Origin and Evolution of Allied Crop Plants

9.4 Role in Development of Cytogenetic Stocks and Their Utility

Hexaploid tall fescue is endemic to much of Europe and North Africa (Borrill et al. 1971). Tall fescue is an outcrossing allohexaploid with 2n ¼ 6x ¼ 42 chromosomes. Its genome size is approximately 6  103 Mbp (Seal 1983). The genomic formula of tall fescue is considered to be PPG1G1G2G2 (Chandrasekharan and Thomas 1971a, b). The P genome is from the diploid (2n ¼ 2x ¼ 14) meadow fescue, and the G1 and G2 genomes are from the tetraploid (2n ¼ 4x ¼ 28) F. arundinacea var. glaucescens (Sleper 1985). Genome constitution of Festuca species has been considered by Bowman and Thomas (1976), Thomas et al. (1997), and Harper et al. (2004) (Table 9.1). Further genome analyses are needed to prove it.

Festuca and Lolium species hybridize naturally and as hybrids regularly exchange genes at high frequency. Some Festulolium cultivars have been developed as novel temperate forage grasses in both Europe and USA (Table 9.2) (Yamada et al. 2005). Intergeneric hybrids between closely related Festuca and Lolium species are being used to broaden the gene pool and provide plant breeders with options to combine high quality traits with broad adaptation to a range of environmental constraints (Humphreys et al. 2003). Therefore, intergeneric hybridization has been carried out to develop a novel cultivar (Thomas and Humphreys 1991). Many cytological researches have been published. Jauhar (1993) reviewed the studies on cytogenetics of the Festuca–Lolium complex.

Table 9.1 Chromosome number and DNA contents of some Festuca and Lolium species Species Common name Chromosome number/genome Festuca arundinacea Tall fescue 2n ¼ 6x ¼ 42; AABBCC F. arundinacea var. glaucescens 2n ¼ 4x ¼ 28; BBCC Festuca gigantea Giant fescue 2n ¼ 8x ¼ 56; AAXXYYZZ Festuca mairei Atlas fescue 2n ¼ 4x ¼ 28; BBDD Festuca pratensis Meadow fescue 2n ¼ 2x ¼ 14; AA Festuca scariosa 2n ¼ 2x ¼ 14; BB Lolium perenne Perennial ryegrass 2n ¼ 2x ¼ 14 Lolium multiflorum Italian ryegrass 2n ¼ 2x ¼ 14

1C DNA content (pg) 6.05 4.28 7.23 3.95 2.20 2.68 2.08 4.10

Modified with Kopecky´ et al. (2008) Table 9.2 Festulolium cultivars developed in Europe and USA Hybrid combination Cultivar name L. multiflorum  F. pratensis Elmet Perun Rakopan F. pratensis  L. multiflorum Paulita Paulena Punia Felopa Sulino Agula L. perenne  F. pratensis Prior Spring Green L. multiflorum  F. arundinacea Kenhy Johnstone Felina Hykor Korina Becˇva From Yamada et al. (2005)

Type Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Amphidiploid Introgression Introgression Introgression Introgression Introgression Introgression

Country United Kingdom Czech Republic Poland Germany Germany Lithuania Poland Poland Poland United Kingdom USA USA USA Czech Republic Czech Republic Czech Republic Czech Republic

Year 1973 1991 2001 1986 1995 1997 1998 1998 2002 1973 2001 1977 1983 1988 1991 1997 1989

9 Festuca

Recently, Yamada et al. (2005) and Kopecky´ et al. (2008) reviewed the cytogenetics of grass species, which are used in breeding of improved cultivars, including interspecific and intergeneric hybrids, and the main attention was given to Lolium/Festuca hybrids. The Lolium/Festuca introgression system is based on a series of seven monosomic substitution lines. In each of these seven lines, one of the chromosomes of Lolium perenne has been replaced by its homeologous equivalent from F. pratensis (King et al. 1998, 2002a, b). An introgression map of L. perenne/F. pratensis chromosome 3, homeologous to rice chromosome 1, has been generated. The introgression map is composed of 16 individuals, each of which carries different-size Festuca chromosome segments. Alignment of overlapping Festuca chromosome segments effectively divides L. perenne/F. pratensis chromosome 3 up into 18 physically demarcated bins. Screening the individuals that make up the introgression map of Lolium/Festuca chromosome 3 for the presence or the absence of Festuca polymorphisms allows genetic markers including restriction fragment length polymorphism (RFLP), amplified fragment length polymorphism (AFLP), simple sequence repeat (SSR), and single nucleotide polymorphism (SNP) to be assigned to one of the 18 introgression bins (King et al. 2002a, b). King et al. (2007) described the exploitation of the published rice genome sequence to bin map sequences from functionally annotated gene models on approximately every fifth to tenth bacterial artificial chromosome/plasmid artificial chromosome (BAC/PAC) clone from rice chromosome 1 to the Lolium/Festuca chromosome 3 introgression map to elucidate the syntenic relationship between rice and the large-genome monocots, to determine the physical location of rice chromosome 1 genes in large-genome monocots and to determine the relationship between gene distribution and recombination in large-genome crop species.

9.5 Role in Classical and Molecular Genetic Study Intergeneric hybrids between diploid L. multiflorum, L. perenne, and F. pratensis have very low fertility. Doubling of the chromosome number of the F1 hybrids

157

leads to restoration of fertility. Partially male and female fertile F1 hybrids can be also obtained by crossing autotetraploid forms of Lolium sp. and F. pratensis. The first amphidiploid Festulolium cultivars, “Elmet” (L. multiflorum  F. pratensis ) and “Prior” (L. perenne  F. pratensis) were bred at the Welsh Plant Breeding Station, UK (now Aberystwyth University), in the early 1970s (Thomas and Humphreys 1991). A major problem for amphidiploid breeding is the high level of homoeologous pairing between the different genomes that leads to genetic instability and loss of hybridity in later generations. To overcome these problems and reduce transfer of deleterious Festuca traits, selective introgression of genes for desirable traits from Festuca into Lolium has become a favored methodology. This process involves the transfer of small segments of alien Festuca chromatin into the recipient Lolium genome (Thomas et al. 1988; Humphreys 1989; Humphreys and Pasˇakinskiene 1996) and has been successfully employed to produce novel Festulolium lines (Humphreys and Thomas 1993; Humphreys et al. 2005). Intergeneric hybrids between L. multiflorum (2x) and F. arundinacea (6x) were used in the first putatively successful introgression breeding program conducted at the University of Kentucky, USA, in which two cultivars, “Kenhy” (Buckner et al. 1977) and “Johnstone” (Buckner et al. 1983), were developed, although in both cases, the inclusion of Lolium genes was never confirmed. “Kenhy” showed improved palatability and a lower fiber content than tall fescue “Kentucky 31” (Buckner et al. 1979). Introgression procedures for the transfer of genes for drought resistance from F arundinacea var. glucescens (2n ¼ 4x ¼ 28) into L. multiflorum (2n ¼ 2x ¼ 14) using DNA markers derived from F. glucescens were described by Humphreys et al. (2005). Androgenesis was found to be an effective procedure for selection of Lolium–Festuca genotypes comprising gene combinations rarely or never recovered by conventional backcross breeding programs (Les´niewska et al. 2001; Humphreys et al. 2003). Androgenesis from Festuca–Lolium complex had been studied using different parental hybrids such as F. pratensis  L. multiflorum (Les´niewska et al. 2001; Rapacz et al. 2004), L. multiflorum  F. arundinacea (Humphreys et al. 1997, 1998b; Pasˇakinskiene˙ et al. 1997; Zwierzykowski et al. 1999; Zare et al. 1999),

158

L. perenne  F. pratensis (Guo et al. 2005), and L. perenne  F. mairei St. Yves (Atlas fescue) (Cao and Sleper 2001). Superior plants with high freezing tolerance have been found in the backcrossed progeny (Yamada et al. 2007). Kosmala et al. (2006) used florescence in situ hybridization (FISH) and genomic in situ hybridization (GISH) to identify a segment of F. pratensis chromosome 2, which was introgressed into L. multiflorum and was located terminally on a non-NOR arm in two of the three most freezingtolerant plants. Chromosome 4 of F. pratensis with the translocated terminal segments of L. multiflorum on both arms was detected in the third plant. F. mairei has exceptional drought tolerance but poor turf and forage quality. Some progeny in a high droughttolerant group derived from the intergenic hybridization between F. mairei and L. perenne rated better than the F. mairei parent (Wang and Bughrara 2008). These valuable breeding materials will be used in Festulolium introgression breeding programs to accelerate the breeding process and provide novel robust new forage grasses for cultivation in marginal areas. The GISH technique provides the means to identify segments of alien chromosomes introduced into the recipient species and has proved to be a powerful tool for determining their chromosome location (Thomas et al. 1994; Humphreys et al. 1995; Humphreys and Pasˇakinskiene 1996). However, there are potential difficulties in the identification of very small introgressed chromosome segments using the GISH technique (Humphreys et al. 1998a). Molecular DNA markers based on Southern hybridization such as RFLP as well as PCR-based markers such as random amplified polymorphic DNA (RAPD), AFLP, and SSR have been developed for grass species. Yamada and Kishida (2003) applied rice cDNARFLP probes from the activity of the Rice Genome Program (RGP) of Japan to forage grasses in order to investigate genetic variation within and between varieties of grasses and to identify variety-specific RFLP markers for use in breeding programs exploiting intergeneric hybridization of Lolium and Festuca. RFLP analysis is a highly labor-intensive methodology compared to the PCR-based methods. Recently, SSR markers have been developed in tall fescue (Saha et al. 2004). SSR markers provide the current marker system of choice due to their abundance, ubiquitous distribution in plant genomes, high level of reproducibility, ease of PCR-based analysis, and detection of

T. Yamada

codominant multiallelic loci. Momotaz et al. (2004) have analyzed the genetic polymorphism of multiple genotypes derived from taxa of the Lolium/Festuca complex using these distinct sets of SSR markers and applied these data to investigate introgression and genetic relatedness in Festulolium accessions. RAPD and SSR markers were used to detect the parental genome composition of F1 hybrids and backcross, generated from crosses between F. mairei St. Yves and L. perenne (Wang et al. 2009). Recently, Tamura et al. (2009) have developed the intron-flanking EST markers for the genetic analysis and molecular breeding of Lolium, Festuca, and their intergeneric hybrid, Festulolium. In this marker system, primer sets are designed from ESTs of the related species showing high similarity to unique rice genes. Thus, comparative genomic analyses allow the genomic loci of markers in the related species to be estimated from the corresponding rice loci. High species specificities of these markers based on the intron polymorphisms are advantageous for the genetic analysis and molecular breeding of interspecific or intergeneric hybrid.

9.6 Role in Crop Improvement Through Traditional and Advanced Tools Tall fescue is a cross-fertilizing species and is largely self-sterile (Sleper 1985). As a result, individuals are highly heterozygous and populations are heterogeneous (Sleper and West 1996). Source of breeding materials include plant introductions, commercial cultivars, planned crosses among selected clones including progeny from interspecific and intergeneric crosses, and populations resulting recurrent selection (Sleper and West 1996). Breeding objects of tall fescue includes forage yield, seed yield, feeding quality for improved animal performance, and resistance to pest and diseases (Sleper and West 1996). Genetic improvement of forage grasses such as fesues by conventional plant breeding is slow since many species are predominantly, if not completely, allogamous wind-pollinated grasses. Gene technology and the production of transgenic plants offer the opportunity to generate unique genetic variation. Application of transgenesis to forage plant improvement has been focused on the development of transformation events with unique genetic variation and in

9 Festuca

studies on the molecular dissection of plant biosynthetic pathways and developmental processes of high relevance for forage production (Spangenberg et al. 2001). In recent years, biolistic transformation and Agrobacterium-mediated transformation have become the main methods for producing transgenic grasses. Biolistic methodology, based on particle bombardment, employs high-velocity gold or tungsten particles to deliver DNA into living cells for stable transformation. Because biolistic methodology is a physical process that involves only one biological system, it is a fairly reproducible method that can be easily adapted from one laboratory to another laboratory. Transgenic forage plants have been obtained by particle bombardment of embryogenic cell in tall fescue and red fescue (Spangenberg et al. 1995; Cho et al. 2000; Wang et al. 2001, 2003; Chen et al. 2003, 2004) Agrobacteriummediated transformation has the advantage of allowing for low copy number integration of the transgenes into the plant genome. In recent years, significant progress has been made in developing transformation protocols using Agrobacterium tumefaciens as a vector. Transgenics have been obtained by Agrobacterium-mediated transformation in tall fescue (Dong and Qu 2005; Wang and Ge 2005) and Festololium (Guo et al. 2009). Gao et al. (2008) described that Agrobacterium-mediated transformation appears to be the preferred method for producing transgenic tall fescue plants by comparative analysis using both methods. For downregulation of endogenous genes, the target gene is normally isolated from a species and transferred back to the same species or closely related species. To improve forage digestibility by downregulation of lignin biosynthesis, genes encoding two key enzymes involved in lignin biosynthesis, cinnamyl alcohol dehydrogenase (CAD) and caffeic acid O-methyltransferase (COMT), were cloned from tall fescue (Chen et al. 2002). Transgenic tall fescue plants were produced using antisense and sense CAD and COMT gene constructs under the control of maize ubiquitin promoter. Severely reduced mRNA levels and significantly decreased enzymatic activities were found in some transgenic lines. These transgenic tall fescue plants had reduced lignin content, altered lignin composition, and increased in vitro dry matter digestibility (Chen et al. 2003, 2004). Molecular marker system is powerful tool of plant breeding. Some works have been done in tall fescue and meadow fescue. A total of 157 expressed

159

sequenced tag (EST)-SSR primer pairs were designed from tall fescue cDNA library sequences and tested for cross-species amplification in seven different grass and cereal species (Saha et al. 2004). A subset of these markers was used for more intensive phylogenetic analysis of cool-season grasses (Mian et al. 2005). ESTSSR and AFLP primer pairs were used to construct a tall fescue map based on PCR-generated marker genotyping of the F1 (HD28-56  R43-64) two-way pseudo-testcross population (Saha et al. 2005). Integration of the parental genetic maps produced a consensus structure of 17 linkage groups (LGs) and a cumulative map length of 1,841 cm. Homoeologous relationships between six of the seven expected groups were observed. Further map enhancement would be expected to identify the remaining five homologous LGs to the anticipated total of 22. Tall fescue RFLP markers have also been used to efficiently detect RFLPs in the diploid F. pratensis. Comparative genetic mapping between the two species of fescue (Chen et al. 1998) generated a map containing 66 markers on seven LGs with a total map length of 280 cm. A high level of conservation for the orders of most of the genetic markers was observed, with 23 of 33 common markers located in corresponding LGs. A comprehensive genetic map has been constructed for meadow fescue using a two-way pseudo-testcross F1 population from the pair-cross of a single genotype from a Norwegian population that has been selected for frost tolerance (HF2/7) and a genotype from a Yugoslavian variety (B14/16). The combined data for homologous and heterologous RFLP, AFLP, SSR, and isoenzyme markers from the two parental maps defined 466 loci with a total map length of 658 cm (Alm et al. 2003). Conserved synteny was analyzed through the use of heterologous RFLP anchor probes derived from perennial and Italian ryegrass, wheat, barley, oat, rice, maize, and sorghum, demonstrating a high degree of conserved synteny and colinearity with both perennial ryegrass and the Triticeae consensus map. The F1 (B14/16  HF2/7) mapping population has been used for quantitative trait loci (QTL) analysis of vernalization requirement, heading time and number of panicles (Ergon et al. 2006), following phenotypic evaluation (Fang et al. 2004). Molecular genetic markers have also been used to monitor the introgression of a crown rust resistance determinant from F. pratensis into L. multiflorum (Armstead et al. 2006).

160

Somatic hybridization is a technique based upon protoplast fusion, which enables hybridization between sexually incompatible species or genera. This technique may also provide an alternative approach in order to create novel genetic combination between Festuca and Lolium species. A specific objective is the creation of novel cytoplasmic combinations, e.g., a mixture of both parental cytoplasms. Symmetric somatic hybrids (Takamizo et al. 1991) and asymmetric somatic hybrids (Spangenberg et al. 1994) have been produced between F. arundinacea and L. multiflorum, but no further research regarding their progeny with respect to agronomic traits were reported.

T. Yamada

meadow fescue, as well as Lolium species including perennial ryegrass and Italian ryegrass. Introgression program could be valuable to transfer useful genes into cultivated species for forage and turf.

9.9 Some Dark Sides and Their Addressing Some of species of Festuca, especially F. arundinacea, have the potential to become weeds. For forage use, we should consider this invasiveness, especially when we sow seeds of productive and vigorous cultivars near conserved semi-natural or natural pasture.

9.7 Genomics Resources Developed A total of 41,516 ESTs were generated from nine cDNA libraries of tall fescue representing tissues from different plant organs, developmental stages, and abiotic stress conditions. The Festuca Gene Index (FaGI) has been established (Mian et al. 2008). FaGI provides a useful resource for genomics studies of tall fescue and other closely related forage and turf grass species. Comparative genomic analyses between tall fescue and other grass species, including ryegrasses (Lolium spp.), meadow fescue, and tetraploid fescue (F. arundinacea var. glaucescens) will benefit from this database. These ESTs are an excellent resource for the development of SSR and SNP-PCRbased molecular markers. Information on tall fescue ESTs is available in the site: http://www.plantgdb.org/ download/download.php?dir¼/Sequence/ESTcontig/ Festuca_arundinacea/previous_version/151a. A total of 11,346 transcript-derived fragments (TDFs) were detected in F. mairei (Wang and Bughrara 2007). Analysis of the 163 differentially expressed fragments (DEFs) provides a first glimpse into the transcripts of F. mairei during drought stress treatment.

9.8 Scope for Domestication and Commercialization The main potential economic use for wild species of Festuca considered to date is as a genetic resource for improvement of cultivated species such as tall fescue,

9.10 Recommendations for Future Actions Molecular breeding is important and will be used extensively in future forage and turf improvement. Transformation techniques using Agrobacterium or biolistics-based method and many available molecular markers such as SSR markers and some functionallyassociated genetic markers are now developed in Festuca species. We still have remaining challenges for the successful implementation of molecular breeding in practical varietal development. Molecular breeding needs to develop from a platform of good conventional breeding and include supporting agronomic research and partnering with commercial industry where appropriate. With the availability of more sequencing information, such as on ESTs, gene isolation has become much easier than ever before. We should focus on functional characterization of genes and their regulatory elements. With a widening range of traits, techniques for more accurate, rapid and non-invasive phenotyping and genotyping become increasingly important. The large amounts of data involved require good bioinfomatics support. Data of various kinds must be integrated from an increasingly wide range of sources such as genetic resources and mapping information for plant populations through to the transcriptome and metabolome of individual tissues. The merging of data from diverse sources and multivariate data mining across datasets can reveal novel information concerning the biology of complexity.

9 Festuca

Sustainable system for production and maintenance of forage and turf is the most important issue. Some of the Festuca species have abiotic tolerance such as against freezing, heat, and drought. We should evaluate many accessions of Festuca species for the materials of breeding now.

References Alexeev EB (1977) Systematics of Asian fescues (Festuca). I. Subgenera Drymanthele, Subulatae, Schedonorus, Leucopoa. Bjull Moskovsk Obsˇc Isp Prir, Otd Biol 82:95–102 (in Russian) Alexeev EB (1981) The new taxa of the genus Festuca (Poaceae) from Mexico and Central America. Bot Zhurn 66:1492–1501 (in Russian) Alexeev EB (1982) A new section and three new species of the genus Festuca (Poaceae) from Mexico and Central America. Bot Zhurn 67:1289–1292 (in Russian) Alexeev EB (1984) New taxa of the genus Festuca (Poaceae) from Colombia and Ecuador. Bot Zhurn 69:1543–1552 (in Russian) Alexeev EB (1988) Genus Festuca L. (Poaceae) in Japan, Korea and Taiwan. Novosti Sist Vyssˇ Rast 25:5–27 Alm V, Fang C, Busso CS, Devos KM, Vollan K, Grieg Z, Rognli OA (2003) A linkage map of meadow fescue (Festuca pratensis Huds.) and comparative mapping with other Poaceae species. Theor Appl Genet 108:25–40 Armstead IP, Harper JA, Turner LB, Skøt L, King IP, Humphreys MO, Morgan WG, Thomas HM, Roderick HW (2006) Introgression of crown rust (Puccinia coronata) resistance from meadow fescue (Festuca pratensis) into Italian ryegrass (Lolium multiflorum): genetic mapping and identification of associated molecular markers. Plant Pathol 55:62–67 Barnes RF (1990) Importance and problems of tall fescue. In: Kasperbauer MJ (ed) Biotechnology in tall fescue improvement. CRC, Boca Raton, FL, pp 1–12 Beard JB (1973) Turfgrass: science and culture. Prentice-Hall, Englewood Cliffs, NJ Beard JB (1998) Traditional fine-leaf fescue putting greens. Turfax 6:3 Borrill M, Tyler B, Lloyd-Jones M (1971) Studies in Festuca. I. A charomosome atlas of bovinae and scariosae. Cytologia 36:1–14 Bowman JG, Thomas H (1976) Studies in Festuca. 8. Cytological relationships between F. glaucescens (2n ¼ 28), F. mairei (2n ¼ 28) and F. scoriosa (2n ¼ 14). Z Pflanzenzuecht 76:250–257 Brown AHD (1992) Human impact on plant gene pools and sampling for their conservation. Oikos 63:109–118 Buckner RC, Burrus PB II, Bush LP (1977) Registration of Kenhy tall fescue. Crop Sci 17:672–673 Buckner RC, Bush LP, Burrus PB II (1979) Succulence as a selection criterion for improved forage quality in Lolium–Festuca hybrids. Crop Sci 19:93–96

161 Buckner RC, Boling JA, Burrus PB II, Bush LP, Hemken RA (1983) Registration of Johnstone tall fescue. Crop Sci 23:399–400 Cao M, Sleper DA (2001) Use of genome-specific repetitive DNA sequences to monitor chromatin introgression from Festuca mairei into Lolium perenne. Theor Appl Genet 103:248–253 Casler MD, van Santen E (2000) Patterns of variation in a collection of meadow fescue accessions. Crop Sci 40:248–255 Casler MD, Undersander DJ, Fredericks C, Combs DK, Reed JD (1998) An on-farm test of perennial forage grass varieties under management intensive grazing. J Prod Agric 11:92–99 Catala´n P, Torrecilla P, Lo´pez Rodrı´guez JA, Olmstead RG (2004) Phylogeny of the festucoid grasses of subtribe Loliinae and allies (Poeae, Pooideae) inferred from ITS and trnL-F sequences. Mol Phylogenet Evol 31:517–541 CBD (2005) Handbook of the convention on biological diversity (CBD), 3rd edn. http://www.cbd.int/handbook/ Chandrasekharan P, Thomas H (1971a) Studies in Festuca. V. Cytogenetic relationships between species of Bovinae and Scariosae. Z Pflanzenzucht 65:345–354 Chandrasekharan P, Thomas H (1971b) Studies in Festuca. VI. Chromosome relationships between species of Bovinae and Scariosae. Z Pflanzenzucht 66:76–86 Chen C, Sleper DA, Johal GS (1998) Comparative RFLP mapping of meadow and tall fescue. Theor Appl Genet 97:255–260 Chen L, Auh C, Chen F, Cheng XF, Aljoe H, Dixon RA, Wang Z-Y (2002) Lignin deposition and associated changes in anatomy, enzyme activity, gene expression and ruminal degradability in stems of tall fescue at different developmental stages. J Agric Food Chem 50:5558–5565 Chen L, Auh C, Dowling P, Bell J, Chen F, Hopkins A, Dixon RA, Wang Z-Y (2003) Improved forage digestibility of tall fescue (Festuca arundinacea) by transgenic down-regulation of cinnamyl alcohol dehydrogenase. Plant Biotechnol J 1:437–449 Chen L, Auh C, Dowling P, Bell J, Lehmann D, Wang Z-Y (2004) Transgenic down-regulation of caffeic acid O-methyltransferase (COMT) led to improved digestibility in tall fescue (Festuca arundinacea). Funct Plant Biol 31:235–245 Cho MJ, Ha CD, Lemaux PG (2000) Production of transgenic tall fescue and red fescue plants by particle bombardment of mature seed-derived highly regenerative tissues. Plant Cell Rep 19:1084–1089 Christians N (2000) Fairway grass for Midwest golf course: diseases and tough climatic conditions make choices tricky. Golf Course Manag 68:49–56 Clayton WD, Renvoize SA (1986) Genera Graminum. Grasses of the world. Kew Bull Add Ser 13:1–389 Cvelev NN (1971) The taxonomy and phylogeny of genus Festuca L. of the U.S.S.R. flora. I. The system of the genus and main trends of evolution. Bot Zhurn 56:1252–1262 (in Russian) Cvelev NN (1976) Grasses of the Soviet Union. Nauka, Leningrad, USSR (in Russian) Darbyshire SJ (1993) Realignment of Festuca subgenus Schedonorus with the genus Lolium (Poaceae). Novon 3:239–243 Dong S, Qu R (2005) High efficiency transformation of tall fescue with Agrobacterium tumefaciens. Plant Sci 168:1453–1458

162 Ergon A, Fang C, Jorgensen O, Aamlid TS, Rognli OA (2006) Quantitative trait loci controlling vernalisation requirement, heading time and number of panicles in meadow fescue (Festuca pratensis Huds.). Theor Appl Genet 112:232–242 Fang C, Aamlid TS, Jorgensen O, Rognli OA (2004) Phenotypic and genotypic variation in seed production traits within a full-sib family of meadow fescue. Plant Breed 123:241–246 Fjellheim S, Rognli OA (2005) Molecular diversity of local Norwegian meadow fescue (Festuca pratensis Huds.) populations and Nordic cultivars – consequences for management and utilisation. Theor Appl Genet 111:640–650 Fjellheim S, Blomlie AB, Marum P, Rognli OA (2007) Phenotypic variation in local populations and cultivars of meadow fescue – potential for improving cultivars by utilizing wild germplasm. Plant Breed 126:279–286 Frankel OH, Brown AHD, Burdon JJ (1995) The conservation of plant biodiversity. Cambridge University Press, Cambridge Gao C, Long D, Lenk L, Nielsen KK (2008) Comparative analysis of transgenic tall fescue (Festuca arundinacea Schreb.) plants obtained by Agrobacterium-mediated transformation and particle bombardment. Plant Cell Rep 27:1601–1609 Guo Y-D, Mizukami Y, Yamada T (2005) Genetic characterization of androgenic progeny derived from Lolium perenne  Festuca pratensis cultivars. New Phytol 166:455–464 Guo Y-D, Hiroshi H, Shimamoto Y, Yamada T (2009) Transformation of androgenic-derived Festulolium plants (Lolium perenne L.  Festuca pratensis Huds.) by Agrobacterium tumefaciens. Plant Cell Tissue Organ Cult 96:219–227 Hackel E (1882) Monographia Festucarum Europaeum. T. Fischer, Kassel, Berlin Harper JA, Thomas ID, Lovatt JA, Thomas HM (2004) Physical mapping of rDNA sites in possible diploid progenitors of polyploid Festuca species. Plant Syst Evol 245:163–168 Humphreys MW (1989) The controlled introgression of Festuca arundinacea genes into Lolium multiflorum. Euphytica 42: 105–116 Humphreys MO (2005) Genetic improvement of forage crops – past, present and future. J Agric Sci 143:441–448 Humphreys MW, Pasˇakinskiene I (1996) Chromosome painting to locate genes for drought resistance transferred from Festuca arundinacea into Lolium multiflorum. Heredity 77: 530–534 Humphreys MW, Thomas H (1993) Improved drought resistance in introgression lines derived from Lolium multiflorum  Festuca arundinacea hybrids. Plant Breed 111:155–161 Humphreys MW, Thomas HM, Morgan WG, Meredith MR, Harper JA, Thomas H, Zwierzykowski Z, Ghesquiere M (1995) Discriminating the ancestral progenitors of hexaploid Festuca arundinacea using genomic in situ hybridization. Heredity 75:171–174 Humphreys M, Thomas HM, Harper J, Morgan G, James A, Ghamari-Zare A, Thomas H (1997) Dissecting drought- and cold-tolerance traits in the Lolium–Festuca complex by introgression mapping. New Phytol 137:55–60 Humphreys MW, Pasˇakinskiene˙ I, James AR, Thomas H (1998a) Physically mapping quantitative traits for stressresistance in the forage grasses. J Exp Bot 49:1611–1618 Humphreys MW, Zare AG, Pasˇakinskiene˙ I, Thomas H, Rogers WJ, Collin HA (1998b) Interspecific genomic rearrangements

T. Yamada in androgenic plants derived from a Lolium multiflorum  Festuca arundinacea (2n ¼ 5x ¼ 35) hybrid. Heredity 80:78–82 Humphreys MW, Canter PJ, Thomas HM (2003) Advances in introgression technologies for precision breeding within the Lolium–Festuca complex. Ann Appl Biol 143:1–10 Humphreys J, Harper JA, Armstead IP, Humphreys MW (2005) Introgression-mapping of genes for drought resistance transferred from Festuca arundinacea var. glaucescens into Lolium multiflorum. Theor Appl Genet 110:579–589 Jauhar PP (1993) Cytogenetics of the Festuca–Lolium complex. Springer, Heidelberg King IP, Morgan WG, Armstead IP, Harper JA, Hayward MD, Bollard A, Nash JV, Forster JW, Thomas HM (1998) Introgression mapping in the grasses. I. Introgression of Festuca pratensis chromosomes and chromosome segments into Lolium perenne. Heredity 81:462–467 King J, Roberts LA, Kearsey MJ, Thomas HM, Jones RN, Huang L, Armstead IP, Morgan WG, King IP (2002a) A demonstration of a 1:1 correspondence between chiasma frequency and recombination using a Lolium perenne/ Festuca pratensis substitution. Genetics 161:307–314 King J, Armstead IP, Donnison IS, Thomas HM, Jones RN, Kearsey MJ, Robersts LA, Thomas A, Morgan WG, King IP (2002b) Physical and genetic mapping in the grasses Lolium perenne and Festuca pratensis. Genetics 161:315–324 King J, Armstead IP, Donnison SI, Roberts LS, Harper JA, Skøt K, Elborough K, King IP (2007) Comparative analyses between Lolium/Festuca introgression lines and rice reveal the major fraction of functionally annotated gene models is located in recombination-poor/very recombination-poor regions of the genome. Genetics 177:597–606 Kopecky´ D, Lukaszewski AJ, Dolezˇel J (2008) Cytogenetics of Festulolium (Festuca/Lolium hybrids). Cytogenet Genome Res 120:370–383 Kosmala A, Zwierzykowski Z, Gasior D, Rapacz M, Zwierzykowska E, Humphreys MW (2006) GISH/FISH mapping of genes for freezing tolerance transferred from Festuca pratensis to Lolium multiflorum. Heredity 96:243–251 Les´niewska A, Ponitka A, Slusarkiewicz-Jarzina A, Zwierzykowska E, Zwierzykowski Z, James AR, Thomas H, Humphreys MW (2001) Androgenesis from Festuca pratensis  Lolium multiflorum amphidiploid cultivars in order to select and stabilize rare gene combinations for grass breeding. Heredity 86:167–176 Maxted N, Ford-Lloyd BV, Hawkes JG (1997) Plant genetic conservation: the in-situ approach. Chapman & Hall, London Mian MAR, Saha MC, Hopkins AA, Wang ZY (2005) Use of tall fescue EST-SSRs in phylogenetic analysis of coolseason forage grasses. Genome 48:637–647 Mian MAR, Zhang Y, Wang ZJ-Y, Cheng X, Chen L, Chekhovskiy K, Dai X, Mao C, Cheung F, Zhao X, He J, Scott AD, Town CD, May GD (2008) Analysis of tall fescue ESTs representing different abiotic stresses, tissue types and developmental stages. BMC Plant Biol 8:27 Momotaz A, Forster JW, Yamada T (2004) Identification of cultivars and accessions of Lolium, Festuca and Festulolium hybrids through the detection of simple sequence repeat (SSR) polymorphism. Plant Breed 123:370–376 M€uller J, Catala´n P (2006) Notes on the infrageneric classification of Festuca L. (Gramineae). Taxon 55:139–144

9 Festuca Pasˇakinskiene˙ I, Anamthawat-Jonsson K, Humphreys MW, Jones RN (1997) Novel diploids following chromosome elimination and somatic recombination in Lolium multiflorum  Festuca arundinacea hybrids. Heredity 78:464–469 Peeters A (2004) Wild and sown grasses. Blackwell Pub, Rome Peter-Schmid MKI, Boller B, Ko¨lliker R (2008a) Habitat and management affect genetic structure of Festuca pratensis but not Lolium multiflorum ecotype populations. Plant Breed 127:510–517 Peter-Schmid MKI, Ko¨lliker R, Boller B (2008b) Value of permanent grassland habitats as reservoirs of Festuca pratensis Huds. and Lolium multiflorum Lam. populations for breeding and conservation. Euphytica 164:239–253 Rapacz M, Gasior D, Zwierzykowski Z, Lesniewska-Bocianowska A, Humphreys MW, Gay AP (2004) Changes in cold tolerance and the mechanisms of acclimation of photosystem II to cold hardening generated by anther culture of Festuca pratensis  Lolium multiflorum cultivars. New Phytol 162:105–114 Ruemmele BA, Wipft JK, Brilman L, Hignight KW (2003) Fine-leaved Festuca species. In: Casler MD, Duncan RR (eds) Turfgrass biology, genetics and breeding. Wiley, New Jersey, pp 129–174 Saha MC, Mian MAR, Eujayl I, Zwonitzer JC, Wang LJ, May GD (2004) Tall fescue EST-SSR markers with transferability across several grass species. Theor Appl Genet 109:783–791 Saha MC, Mian R, Zwonitzer JC, Chekhovskiy K, Hopkins AA (2005) An SSR- and AFLP-based genetic linkage map of tall fescue (Festuca arundinacea Schreb.). Theor Appl Genet 110:323–336 Saint-Yves A (1922) The Festuca (subg. Eu-Festuca) of North Africa and of the Atlantic Islands. Candollea 1:1–63 (in French) Seal AG (1983) DNA variation in Festuca. Heredity 50:225–236 Sleper DA (1985) Breeding tall fescue. Plant Breed Rev 3: 313–342 Sleper DA, West CP (1996) Tall fescue. In: Moser LE, Buxton DR, Casler MD (eds) Cool-season forage grasses. American Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI, pp 471–502 Sˇmarda P (2006) DNA ploidy levels of Romanian fescues (Festuca L., Poaceae), measured in living plants and herbarium specimens. Folia Geobot 41:417–432 Sˇmarda P (2008) DNA ploidy level variability of some fescues (Festuca subg. Festuca, Poaceae) from Central and Southern Europe measured in fresh plants and herbarium specimens. Biologia 63:349–367 Sˇmarda P, Kocˇ´ı K (2005) Festuca alpina, a new species for the flora of Slovakia. Biologia 60:383–385 Sˇmarda P, Stancˇ´ık D (2006) Ploidy level variability in South American fescues (Festuca L., Poaceae): use of flow cytometry in up to 5 1/2-year-old caryopses and herbarium specimens. Plant Biol 8:73–80 Sˇmarda P, M€uller J, Vra´na J, Kocˇ´ı K (2005) Ploidy level variability of some Central European fescues (Festuca L. subg. Festuca, Poaceae). Biologia 60:25–36 Sˇmarda P, Bures P, Horova L, Foggi B, Rossi G (2008a) Genome size and GC content evolution of Festuca: ancestral expansion and subsequent reduction. Ann Bot 101:421–433 Sˇmarda P, Bures P, Horova L, Rotreklov O (2008b) Intrapopulation genome size dynamics in Festuca pallens. Ann Bot 102:599–607 Spangenberg G, Valles MP, Wang Z-Y, Montavon P, Nagel J, Potrykus I (1994) Asymmetric somatic hybridization

163 between tall fescue (Festuca arundinacea Schreb) and irradiated Italian ryegrass (Lolium multiflorum Lam) protoplasts. Theor Appl Genet 88:509–516 Spangenberg G, Wang Z-Y, Wu XL, Nagel J, Iglesias VA, Potrykus I (1995) Transgenic tall fescue (Festuca arundinacea) and red fescue (F. rubra) plants from microprojectile bombardment of embryogenic suspension cells. J Plant Physiol 145:693–701 Spangenberg G, Kalla R, Lidgett A, Sawbridge T, Ong EK, John U (2001) Breeding forage plants in the genome era. In: Spangenberg G (ed) Molecular breeding of forage crops. Kluwer Academic Publishers, Dordrecht, pp 1–39 Takamizo T, Spangenberg G, Suginobu K, Potrykus I (1991) Somatic hybridization in Gramineae: somatic hybrid plants between tall fescue (Festuca arundinacea Schreb.) and Italian ryegrass (Lolium multiflorum Lam.). Mol Gen Genet 231:1–6 Tamura T, Yonemaru J, Hisano H, Kanamori H, King J, King IP, Tase K, Sanada Y, Komatsu T, Yamada T (2009) Development of intron-flanking EST markers for the Lolium/ Festuca complex using rice genomic information. Theor Appl Genet 118:1549–1560 Thomas H, Humphreys MO (1991) Progress and potential of interspecific hybrids of Lolium and Festuca. J Agric Sci 117:1–8 Thomas H, Morgan WG, Humphreys MW (1988) The use of a triploid hybrid for introgression in Lolium species. Theor Appl Genet 76:299–304 Thomas HM, Morgan WG, Meredith MR, Humphreys MW, Thomas H, Leggett JM (1994) Identification of parental and recombined chromosomes of Lolium multiflorum  Festuca pratensis by genome in situ hybridization. Theor Appl Genet 88:909–913 Thomas HM, Harper JA, Meredith MR, Morgan WG, King IP (1997) Physical mapping of ribosomal DNA sites in Festuca arundinacea and related species by in situ hybridization. Genome 40:406–410 Torrecilla P, Lo´pez-Rodrı´guez JA, Catala´n P (2004) Phylogenetic relationships of Vulpia and related genera (Poeae, Poaceae) based on analysis of ITS and trnL-F sequences. Ann Mo Bot Gard 91:124–158 Tscharntke T, Klein AM, Kruess A, Steffan-Dewenter I, Thies C (2005) Landscape perspectives on agricultural intensification and biodiversity – ecosystem service management. Ecol Lett 8:857–874 Wang JP, Bughrara SS (2007) Monitoring of gene expression profiles and identification of candidate genes involved in drought responses in Festuca mairei. Mol Genet Genomics 277:571–587 Wang JP, Bughrara SS (2008) Evaluation of drought tolerance for Atlas fescue, perennial ryegrass, and their progeny. Euphytica 164:113–122 Wang Z-Y, Ge Y (2005) Agrobacterium-mediated high efficiency transformation of tall fescue (Festuca arundinacea Schreb.). J Plant Physiol 162:103–113 Wang Z-Y, Ye XD, Nagel J, Potrykus I, Spangenberg G (2001) Expression of a sulphur-rich sunflower albumin gene in transgenic tall fescue (Festuca arundinacea Schreb.) plants. Plant Cell Rep 20:213–219 Wang Z-Y, Bell J, Ge YX, Lehmann D (2003) Inheritance of transgenes in transgenic tall fescue (Festuca arundinacea Schreb.). In Vitro Cell Dev Biol Plant 39:277–282 Wang JP, Bughrara SS, Mian RMA, Saha MC, Sleper DS (2009) Parental genome composition and genetic classifications of

164 derivatives from intergeneric crosses of Festuca mairei and Lolium perenne. Mol Breed 23:299–309 Wilkins PW (1991) Breeding perennial ryegrass for agriculture. Euphytica 52:201–214 Yamada T, Kishida T (2003) Genetic analysis of forage grasses based on heterologous RFLP markers detected by rice cDNAs. Plant Breed 122:57–60 Yamada T, Forster JW, Humphreys MW, Takamizo T (2005) Genetics and molecular breeding in Lolium/Festuca grass species complex. Grassl Sci 51:89–106 Yamada T, Guo Y-G, Mizukami Y, Tamura K, Tase K (2007) Introgression breeding program in Lolium/Festuca complex

T. Yamada using androgenesis. In: Xu Z et al (eds) Proceedings 11th IAPTC&B congress – biotechnology and sustainable agriculture 2006 and beyond. Springer, Dordrecht, pp 447–450 Zare A-G, Humphreys MW, Rogers WJ, Collin HA (1999) Androgenesis from a Lolium multiflorum  Festuca arundinacea hybrid to generate extreme variation for freezingtolerance. Plant Breed 118:497–501 Zwierzykowski Z, Zwierzykowska E, Slusarkiewicz-Jarzina A, Ponitka A (1999) Regeneration of anther-derived plants from pentaploid hybrids of Festuca arundinacea  Lolium multiflorum. Euphytica 105:191–195

Chapter 10

Lolium Hongwei Cai, Alan Stewart, Maiko Inoue, Nana Yuyama, and Mariko Hirata

10.1 Introduction The genus Lolium contains two widely cultivated temperate species, L. multiflorum (Italian or annual ryegrass) and L. perenne (perennial ryegrass), and it is against this background that their wild relatives will be covered in this paper. The majority of L. multiflorum sown are annual or Westerwold ryegrass forms, but very large numbers of biannual or Italian ryegrass forms, having potentially 1,000 or more cultivars, are also used. L. perenne or perennial ryegrass is the most widely sown perennial grass of the temperate world, both as forage and a turf with potentially over 2,000 cultivars. In addition, there have been many hybrids made between L. multiflorum and L. perenne and these are termed hybrid ryegrass of which there are potentially over 250 cultivars. Many artificial tetraploids have also been developed in these two species and they are used commercially as forage along with diploids. A number of the annual Lolium species are recognized as serious weeds in cultivated wheat and barley fields. These include three inbreeders, L. temulentum, L. persicum, and L. remotum, as well as an outbreeder, L. rigidum. Apart from its weed status, L. rigidum is also cultivated as it is naturalized in annual pastures in Medi-

H. Cai (*) Forage Crop Research Institute, Japan Grassland Agriculture and Forage Seed Association, 388-5, Higashiakata, Nasushiobara 329-2742 Tochigi, Japan Department of Plant Genetics and Breeding, College of Agronomy and Biotechnology, China Agricultural University, 2, Yuanmingyuan West Road, 100193 Beijing, China e-mail: [email protected], [email protected]

terranean climates of the world. Largely, these pastures are allowed to regenerate naturally, but at times they are also sown. At least four cultivars are recognized in Australia, some are simply regional ecotypes while the cultivar Guard has been bred for resistance to Anguina sp. nematode, a pivotal organism along with a bacterium, Corynebacterium sp., causing the annual ryegrass toxicity syndrome in Australia (Stynes and Bird 1993). In addition, another resistant cultivar, SafeGuard, has been developed from hybridization between Guard and L. multiflorum. L. rigidum has also become notorious in Australia for the development of a wide range of herbicide tolerances, often as few as five generations (Broster and Pratley 2006). However, herbicide resistance is also developing in naturally reseeding pastures of L. multiflorum in other regions of the world. There are now hundreds of cases of herbicide resistances not only in L. rigidum (56) but also many in L. multiflorum (34), L. perenne (5), and L. persicum (2) (http:// www.weedscience.org/In.asp), and many of these involve multiple resistance. The Lolium genus has clearly derived from the closely related section Schedonorus of the genus Festuca, the broad leaf fescues. Indeed, some taxonomic revisions have included the broad leaf fescues in Lolium (Darbyshire 1993), while others have split the broad leaf fescues into a separate genus, Schedonorus (Soreng and Terrell 1997). Here, we will not enter into this debate but restrict ourselves to the traditional Lolium species. It is clear though that the broad leaf fescues can be considered as a secondary gene pool for Lolium breeding. Indeed there have been many Festulolium cultivars developed and are also being used for introgression of genes back into Lolium (Humphreys et al. 2003).

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_10, # Springer-Verlag Berlin Heidelberg 2011

165

166

H. Cai et al.

Like many grass fungal endophytes, the genus Epichloe¨ or their Neotyphodium anamorphs have coevolved with Lolium and some are now used commercially to provide protection against insects, particularly in New Zealand (Easton 2006).

4.

10.2 Basic Botany of the Species

5.

The genus Lolium consists of nine diploid species with a chromosome number of 2n ¼ 2x ¼ 14 (Terrell 1968; Clayton and Renvoize 1986; Jauhar 1993), and these are commonly referred to as ryegrasses. L. edwardii has only recently been named, although Terrell (1968) recognized it as a distinct high altitude form of L. canariense (Scholz et al. 2000). 1. L. perenne L.; the widely used perennial ryegrass of commercial importance for forage and turf (Fig. 10.1) 2. L. multiflorum Lam. (L. italicum A. Braun, nom. inval.); The widely used Italian or annual or Westerwold ryegrass of commercial importance 3. L. rigidum Gaud. [L. stricture C. Presl; L. loliaceum (Bory et Chaub.) Hand.-Mazz.]; the rigid or annual

Fig. 10.1 The plant of Lolium perenne

6. 7.

8.

9.

ryegrass, now distributed widely in naturally regenerating annual pastures in the Mediterranean climatic zones, including Australia where it is a valuable forage L. canariense Steud.; an annual restricted to the Canary Islands where it occurs at low altitudes in poor conditions L. edwardii Scholz, Stierstorfer et Gaisberg; an annual with a limited distribution at higher altitudes on the Canary Islands L. temulentum L.; darnel ryegrass, widely distributed as an annual weed of cereal crops (Fig. 10.2) L. persicum Boiss. et Hohen.; widely distributed as an annual weed of cereal crops predominantly in Western Asia L. remotum Schrank [L. tinicolum A. Br.; L. linicola Sond. Ex Koch]; an annual species known largely as a weed in flax L. subulatum Vis; mainly grows under poor and maritime conditions of Cyprus, Israel, Lebanon, Syria, and former Yugoslavia, and it is not considered a crop weed.

The genus is distributed in temperate regions of Central Asia, Europe, and North Africa and has

10 Lolium

167

Fig. 10.2 The panicle of Lolium temulentum Table 10.1 DNA contents and breeding species Species Mode of Habit breeding L. perenne Outbreeder Perennial L. multiflorum

Outbreeder

Biennial

L. rigidum

Outbreeder

Annual

L. canariense

Outbreeder

Annual

L. edwardii

Inbreedera

Annual

L. temulentum

Inbreeder

Annual

L. remotum

Inbreeder

Annual

L. persicum

Inbreeder

Annual

L. subulatum

Inbreeder

Annual

system in Lolium DNA content (pg/2C) 4.15 (Evans et al. 1972) 4.31(Rees and Jones 1967) 4.33 (Hutchinson et al. 1979) 4.23 (Hutchinson et al. 1979) 6.28 (Stewart Unpublished) 6.23 (Hutchinson et al. 1979) 6.04 (Hutchinson et al. 1979) 6.35 (Hutchinson et al. 1979) 5.49 (Hutchinson et al. 1979)

a

Probably inbreeding as no crossing was apparent when adjacent to L. multiflorum

been introduced to almost all other temperate regions by man. The genus is commonly divided into four outbreeders and five inbreeders (Table 10.1). Chromosome karyotypes do not differ greatly between species despite the approximately 40% larger chromosome mass of the inbreeding species (Thomas 1981). The differentiation between the four outbreeders and five inbreeders was also very apparent from

molecular marker studies using random amplified polymorphic DNA (RAPD), restriction fragment length polymorphism (RFLP) and internal transcribed spacer (ITS)-rDNA studies (Charmet et al. 1997). Naylor (1960) postulated that the inbreeding group as evolved from the outbreeding group. Hybridization among species in Lolium is often very easy with L. perenne, L. multiflorum, and L. rigidum crossing very freely when and where their flowering times overlap. The outbreeders, L. perenne and L. multiflorum, also cross with the inbreeders, L. temulentum, L. subulatum, and L. remotum, although some reciprocal differences exist (Jenkins 1954; Terrell 1966; Charmet et al. 1996). A number of the Lolium species also cross very easily with Festuca pratensis and F. arundinacea to form Festulolium of which there have been over 20 cultivars developed (Charmet et al. 1996; Yamada et al. 2005). Some crosses have also resulted in the development of a cytoplasmic male sterility system in L. perenne (Wit 1974). L. temulentum, or darnel ryegrass, has been proposed as a model species for genomic studies of cool-season forage and turf grasses due to its close relationship to the important cultivated Lolium species, its short life cycle (2–3 months and lack of vernalization requirement), and its diploid self-fertile nature (Baldwin and Dombrowski 2006). L. temulentum has become valuable for the study of self-incompatibility in Lolium. The outbreeders, perennial ryegrass and Italian ryegrass, have the multiallelic two-locus (S, Z) system of self-incompatibility (Cornish et al. 1979), while the self-compatibility of L. temulentum is controlled in a gametophytic manner, derived by a single gene mutation of either the Z-locus or a locus tightly linked to it (Thorogood and Hayward 1992). The ease of hybridization between L. temulentum and the cultivated Loliums, perennial ryegrass and Italian ryegrass (Jenkin 1935), has also allowed the possibility of developing inbred lines in perennial ryegrass and Italian ryegrass in order to capture heterosis in breeding (Yamada 2001).

10.3 Genetic Resources There are a number of genetic resource collections of Lolium throughout the world, 95% of which are L. perenne and L. multiflorum. The three main

168 Table 10.2 Number of Lolium accessions in major Genebanks Species Number of accessions L. perenne 11,565 L. multiflorum 2,660 L. rigidum 328 L. temulentum 291 L. remotum 47 L. persicum 26 L. canariense 16 L. subulatum (L. Loliaceum) 5 L. edwardii 2

international collections of Lolium are in the USA with the National Plant Germplasm System of the USDAARS (http://www.ars-grin.gov/npgs/index.html), in Europe in the Eurisco system (http://eurisco.ecpgr.org/), and in the Margot Forde Germplasm Center in New Zealand (http://www.agresearch.co.nz/seeds). Table 10.2 lists the total numbers of accessions of the nine species, although the actual numbers of different accessions will be less than this because of planned duplication.

10.4 Genetic Diversity of L. temulentum, L. rigidum, and L. persicum Two studies of the diversity with L. temulentum have been carried out with similar results. Senda et al. (2004) studied the genetic diversity and relationships of 48 accessions from eight countries using seven simple sequence repeat (SSR) markers and 44 amplified fragment length polymorphism (AFLP) polymorphic loci. Kirigwi et al. (2008) studied 41 L. temulentum, three F. arundinacea, two F. glaucescens, and two F. pratensis genotypes with selected 30 tall fescue employing expressed sequence tag (EST)-SSRs and 32 Festuca  Lolium (F  L) genomic SSRs. The study by Senda and coworkers resulted in four distinct geographic clusters: Pakistan–Nepal complex, the Mediterranean region, Ethiopia, and Japan, while Kirigwi’s study resulted in three clusters and one lone accession. This clustering and levels of genetic diversity within each group is consistent with the hypothesis that the origin of L. temulentum lies in the general region of Southwest Asia and the Mediterranean basin and it also reflects its weedy origin. The clear groupings of accessions from each region with high coefficient of gene differentiation (Gst ¼ 0.688) indicate that genetic exchange between them is limited.

H. Cai et al.

The genetic structure of the two outbreeders, L. perenne and L. rigidum, was studied by Balfourier et al. (1998) using 12 polymorphic isozyme loci on starch gel electrophoresis with 120 wild accessions of L. perenne and 50 of L. rigidum. Genetic diversity indices [number of alleles (A), observed (Ho) and expected (He) heterozygosity] were significantly higher in L. rigidum than in L. perenne. For both species, most of the diversity appeared to be within the populations. In L. perenne, all three genetic indices (A, Ho, He) showed the same trend of geographical variation, with the lowest values in the Northwest part of the distribution area (UK, Ireland) and the highest ones in the Southeast (Turkey, Lebanon, Cyprus, Iraq, Iran). In the same way, as indicated by logistic regression analyses between allelic frequencies and geographical data of L. perenne populations, the latitudinal gradient of allelic frequencies appeared to be more pronounced, although significant relationships also existed with longitude. In contrast, no spatial organization of the diversity was detected in L. rigidum. They concluded that L. perenne could be derived from a small bottleneck of L. rigidum populations in the Middle East, and its present distribution area in Europe could be explained either by the extension of primitive agriculture from the Fertile Crescent or as a consequence of postglacial recolonization from southern refugia. Further chloroplast studies by Balfourier et al. (2000) and McGrath et al. (2007) are also consistent with this conclusion. Genetic diversity of Iranian natural populations of L. persicum was studied using 12 genomic and 20 EST-SSR primer pairs from tall fescue (Sharifi Tehrani et al. 2008). The analysis showed that the samples could be clustered into three regional groups relating to their geographic origin (N, NW, and SW).

10.5 The Classical and Molecular Genetic Studies of L. temulentum and Other Lolium species L. temulentum has been used as a model species for many studies. In seed shattering studies, results of its hybridization with L. persicum indicated that L. temulentum and L. persicum exhibited typical nonshattering (1.6% shattering) and shattering phenotypes (70.8%), respectively. The F1 hybrids of

10 Lolium

L. temulentum  L. persicum shatter, while the F2s segregate into 15 shattering: 1 non-shattering, indicating that non-shattering trait is controlled by two recessive nuclear-encoded genes (Senda et al. 2006). Paraquat resistance in L. rigidum population (AFLR1) has been attributed to reduced paraquat translocation. Genetic inheritance of paraquat resistance was reported by Yu et al. (2009). The results suggested that paraquat resistance in AFLR1 is inherited as a dominant or partially dominant nuclear-encoded trait. Based on the results of pseudo-F2 generation seedlings treated with multiple dose rates sufficient to control the susceptible parental population, and further confirmed by individual phenotyping of cloned plant genotypes. Therefore, the evolved paraquat resistance in AFLR1 is likely to be controlled by a single major nuclear gene. L. temulentum was also used for many plant physiological studies, including salt stress (Baldwin and Dombrowski 2006; Dombrowski et al. 2008), gibberellin structure, flowering (Evans et al. 1990; Gocal et al. 1999, 2001; King et al. 2001, 2008), and others (Gay and Thomas 1995; Gallagher and Pollock 1998, 2004; Baldwin et al. 2007). Dombrowski and Martin (2009) identified and evaluated reference genes for use in real-time quantitative PCR (RT-PCR) for abiotic stress studies in L. temulentum. The analysis found that among nine L. temulentum housekeeping genes, eEF-1a and UBQ5 were the most stable and ACT11 was the least stable of the genes tested. Analysis by geNorm software indicated that the two most stably expressed housekeeping genes (eEF-1a and UBQ5) should be utilized and that are sufficient for normalization of gene expression during stress-related studies in L. temulentum. Gocal et al. (2001) has also reported the expressions of two APETALA1 (AP1)-like genes, LtMADS1 and LtMADS2, and of L. temulentum LEAFY (LtLFY) gene during floral induction.

169

bridging species L. multiflorum using a program of backcrossing and selfing. Wang et al. (2002) developed an efficient regeneration and transformation system after screening for tissue culture responses using 46 L. temulentum accessions. Embryogenic callus formation frequency ranged from 1 to 11% across all accessions tested. Embryogenic calluses of a few responsive accessions were used to establish cell suspension cultures. The regeneration frequency of green plantlets from the established cell suspension ranged from 15 to 39%. After transferring the regenerants to the greenhouse, fertile plants were readily obtained without any vernalization treatment. In addition, Wang et al. (2005b) also attempted to improve tissue culture response of L. temulentum to anther culture and doubled haploid production. Ge et al. (2007) reported Agrobacterium tumefaciens-mediated transformation of L. temulentum. A. tumefaciens strain EHA105 harboring pCAMBIA1301 and pCAMBIA1305.2 vectors were used to infect embryogenic callus pieces. Hygromycin was used as a selection agent in stable transformation experiments. The transgenic nature of the regenerated plants was checked by PCR, Southern hybridization analysis, and b-glucuronidase (GUS) staining. Progeny analysis showed Mendelian inheritance of the transgenes. Dalton et al. (1999) reported the cotransformation in Lolium including L. temulentum produced by microprojectile bombardment. In their report, 37 plants (30 L. multiflorum, six L. perenne, and one L. temulentum) were regenerated from cell suspension colonies bombarded with plasmid DNAs encoding a hygromycin resistance gene expressed under a CaMV35S promoter and a GUS-gene expressed under a truncated rice actin1 promoter and first intron, or a maize ubiquitin promoter and first intron. PCR analysis showed that the cotransformation frequency of the GUS-gene varied from 33 to 78% of transformants, while histochemical staining of leaf tissue from soil-grown plants showed that the coexpression frequency varied from 37 to 50%.

10.6 Introgression Studies and Genetic Transformation of L. temulentum

10.7 Genomics Resources: ESTs and SSR Markers Thomas et al. (1999) reported the transfer of the staygreen phenotype from F. pratensis into L. temulentum. A mutant allele at the nuclear locus sid confers indefinite greenness on senescing leaves of the pasture grass F. pratensis. The mutant allele (designated sidy) was introgressed into L. temulentum strain “Ceres” via the

Unlike the two cultivated species of Lolium, there are only limited EST sequences of other Lolium species. In PlantGBD (http://www.plantgdb.org/), 6,336 EST sequences of L. temulentum can be found (May 2009).

170

Baldwin and Dombrowski (2006) constructed a salt-stressed subtraction library of L. temulentum to evaluate its utility as a model grass to study salt stress. A total of 528 unique sequences were identified, among which 167 corresponded to orthologs of previously identified plant stress response genes. Baldwin et al. (2007) has also constructed a post harvest subtraction library of L. temulentum, and a total of 598 unique sequences were identified. Gallagher and Pollock (1998) also reported a cDNA library from leaf tissues harvested at different times. The genomes of all eukaryotes contain a class of sequences termed SSRs (Tautz et al. 1986) or microsatellites (Litt and Luty 1989). Microsatellites with tandem repeats of a basic motif of <6 bp have emerged as an important source of ubiquitous genetic markers for many eukaryotic genomes (Wang et al. 1994). Microsatellite or SSR markers are widely used in plant genome analysis. Compared with other marker types, SSR markers have a number of advantages: they are co-dominant PCR-based markers that occur at a high frequency in eukaryotic genomes (Li et al. 2002) and are generally associated with non-repetitive DNA regions (Morgante et al. 2002). They are also highly reproducible across populations within the same species, and to some extent across species and genera (Eujayl et al. 2004; Wang et al. 2004, 2005a; Mian et al. 2005). At present, a large numbers of SSR markers have been developed in perennial ryegrass (Jones et al. 2002) and Italian ryegrass (Hirata et al. 2006), and the related tall fescue species (Festuca arundinacea) (Saha et al. 2004, 2005). Senda et al. (2003) have screened a genomic library of L. temulentum with synthetic (GT)15 and (GA)15 oligonucleotides to identify microsatellite-containing clones. After sequencing, primer design, and testing, 15 primer pairs were found to amplify single fragments and five of them showed polymorphism in L. temulentum and L. persicum accessions. Fifteen primer pairs reported in L. perenne were also tested, nine of them amplified the microsatellite alleles, and two of the nine showed polymorphic alleles in L. temulentum and L. persicum. For the polymorphic markers, polymorphic information content (PIC) values ranged from 0.60 to 0.86 with an average of 0.76 among 10 L. temulentum accessions. Tall fescue SSRs were also tested for using in L. temulentum. A total of 40 tall fescue (TF) ESTSSRs and 60 Festuca–Lolium (F  L) genomic SSRs

H. Cai et al.

were screened on a subset of eight genotypes of L. temulentum. The selected 30 tall fescue EST-SSRs and 32 F  L genomic SSRs were used for further analysis of genotypes. The TF-EST- and the F  L genomic-SSRs identified 10.3 and 9.3 alleles per marker, respectively, with an average PIC value of 0.66 (Kirigwi et al. 2008). Sharifi Tehrani et al. (2008) also tested the transferability of 12 genomic and 20 EST-SSR primer pairs from F. arundinacea to representatives of Iranian natural populations of L. persicum. The percentage of transferable polymorphic loci were about 75% for both genomic SSRs and EST-SSRs, and the PIC values of utilized markers were 0.278 for EST-SSRs and 0.219 for genomic-SSRs. Overall, in wild species of Lolium, L. temulentum has been the subject of most of the studies. To date, only a limited number of SSR markers have been screened, and there has not been any genetic linkage maps constructed. However, it is closely related to the two important cultivated Lolium sp. (Italian and perennial ryegrass) those that have rich genomic recourses, such as bacterial artificial chromosome (BAC) libraries (Fujimori et al. 2004; Farrar et al. 2007), high density RFLP and SSR linkage maps (Jones et al. 2001, 2002; Inoue et al. 2004; Hirata et al. 2006), and more than 50,000 ESTs (Sawbridge et al. 2003; Ikeda et al. 2004), and their basic genome information should be easily established. In addition, the genetic transformation and its advantages in plant physiological studies will allow L. temulentum to be widely used in genomic studies of cool-season forage and turf grasses as a model species.

10.8 Endophytic Fungi There are a number of fungal endophytes of the genus Epichloe¨ or their Neotyphodium anamorphs, which coevolved with Lolium (Schardl et al. 2004). The annual species, L. multiflorum, L. rigidum, L. temulentum, and L. canariense contain N. occultans. L. edwardii has, however, been shown to have an endophyte N. typhinum var. canariense, a derivative of E. typhina (Moon et al. 2000; Stewart 2004). L. perenne contains N. lolii and also another type known as LP-2. These endophytes can have very significant effects on the ecology of the plants through some very potent

10 Lolium

chemistry. N. occultans produces lolines, which are known to be potent insecticides but with little effect on grazing animals. The endophytes of L. perenne can produce a number of alkaloids. These include lolitrems, peramine, ergovaline, and janthitrems. Lolitrems are known to cause staggers in grazing animals; peramine is known to be insecticidal to some insects; and ergovaline can have both insecticidal effects and animal health effects. Ergovaline causes vascular constriction and heat stress in grazing animals. Janthitrem has very potent insecticidal effect and possible livestock tremogenic activity. Recently, New Zealand endophytes, which produce the insecticidal compound but are safe for grazing animals, have been incorporated into cultivars. The endophyte in L. perenne might well have assisted the evolutionary development of perenniality in this genus of annual species.

10.9 Conclusion The wild relatives of Lolium are of considerable importance as a resource for the improvement of the widely cultivated grasses, L. multiflorum and L. perenne. They contain endophytic diversity as well as valuable genetic traits, such as non-shattering, selfcompatibility, larger grain sizes, and potentially many other features as well. Until now, only 715 wild relatives of Lolium have been conserved in ex situ genebanks, and more collections will be needed, especially from East Asia, particularly China and Japan.

References Baldwin JC, Dombrowski JE (2006) Evaluation of Lolium temulentum as a model grass species for the study of salinity stress by PCR-based subtractive suppression hybridization analysis. Plant Sci 171:459–469 Baldwin JC, Dombrowski JE, Martin RC, Banowetz GM (2007) Differentially expressed genes associated with postharvest processing in Lolium temulentum L. Plant Sci 173:73–83 Balfourier F, Charmet G, Ravel C (1998) Genetic differentiation within and between natural populations of perennial and annual ryegrass (Lolium perenne and L. rigidum). Heredity 81:100–110

171 Balfourier F, Imbert C, Charmet G (2000) Evidence for phylogeographic structure in Lolium species related to the spread of agriculture in Europe. A cpDNA study. Theor Appl Genet 101:131–138 Broster JC, Pratley JE (2006) A decade of monitoring herbicide resistance in Lolium rigidum in Australia. Aust J Exp Agric 46:1151–1160 Charmet G, Balfourier F, Chatard V (1996) Taxonomic relationships and interspecific hybridization in the genus Lolium (grasses). Genet Resour Crop Evol 43:319–327 Charmet G, Ravel C, Balfourier F (1997) Phylogenetic analysis in the Festuca–Lolium complex using molecular markers and ITS rDNA. Theor Appl Genet 94:1038–1046 Clayton WD, Renvoize SA (1986) Genera Graminum. Grasses of the world. Her Majesty’s Stationery Office, London Cornish MA, Hayward MD, Lawrence MJ (1979) Self incompatibility in ryegrass. I. Genetic control in diploid Lolium perenne L. Heredity 43:95–106 Dalton SJ, Bettany AJE, Timms E, Morris P (1999) Cotransformed, diploid Lolium perenne (perennial ryegrass), Lolium multiflorum (Italian ryegrass) and Lolium temulentum (darnel) plants produced by microprojectile bombardment. Plant Cell Rep 18:721–726 Darbyshire SJ (1993) Realignment of Festuca subgenus Schedonorus with the genus Lolium (Poaceae). Novon 3:239–243 Dombrowski JE, Martin RC (2009) Evaluation of reference genes for quantitative RT-PCR in Lolium temulentum under abiotic stress. Plant Sci 176:390–396 Dombrowski JE, Baldwin JC, Martin RC (2008) Cloning and characterization of a salt stress-inducible small GTPase gene from the model grass species Lolium temulentum. J Plant Physiol 165:651–661 Easton HS (2006) Grasses and Neotyphodium endophytes: co-adaptation and adaptive breeding. Euphytica 154: 295–306 Eujayl I, Sledge MK, Wang L, May GD, Chekhovskiy K, Zwonitzer JC, Mian MA (2004) Medicago truncatula ESTSSRs reveal cross-species genetic markers for Medicago spp. Theor Appl Genet 108:414–422 Evans GM, Rees H, Snell CL, Sun S (1972) The relationship between nuclear DNA amount and the duration of the mitotic cycle. Chromosomes Today 3:24–31 Evans LT, King RW, Chu A, Mander LN, Pharis RP (1990) Gibberellin structure and florigenic activity in Lolium temulentum, a long-day plant. Planta 182:97–106 Farrar K, Asp T, L€ubberstedt T, Xu M, Thomas AM, Christiansen C, Humphreys MO, Donnison IS (2007) Construction of two Lolium perenne BAC libraries and identification of BACs containing candidate genes for disease resistance and forage quality. Mol Breed 19:15–23 Fujimori M, Hayashi K, Hirata M, Ikeda S, Takahashi Y, Mano Y, Sato H, Takamizo T, Mizuno K, Fujiwara T, Sugita S (2004) Molecular breeding and functional genomics for tolerance to biotic stress. In: Hopkins A, Wang ZY, Mian R, Sledge M, Barker RE (eds) Molecular breeding of forage and turf. Kluwer, Dordrecht, pp 21–36 Gallagher JA, Pollock CJ (1998) Isolation and characterization of a cDNA clone from Lolium temulentum L. encoding for a sucrose hydrolytic enzyme which shows alkaline/neutral invertase activity. J Exp Bot 49:789–795

172 Gallagher JA, Cairns AJ, Pollock CJ (2004) Cloning and characterization of a putative fructosyltransferase and two putative invertase genes from the temperate grass Lolium temulentum L. J Exp Bot 55:557–569 Gay AP, Thomas H (1995) Leaf development in Lolium temulentum L. – photosynthesis in relation to growth and senescence. New Phytol 130:159–168 Ge Y, Cheng X, Hopkins A, Wang ZY (2007) Generation of transgenic Lolium temulentum plants by Agrobacterium tumefaciens-mediated transformation. Plant Cell Rep 26: 783–789 Gocal GFW, Poole AT, Gubler F, Watts RJ, Blundell C, King RW (1999) Long-day up-regulation of a GAMYB gene during Lolium temulentum inflorescence formation. Plant Physiol 119:1271–1278 Gocal GFW, King RW, Blundell CA, Schwartz OM, Andersen CH, Weigel D (2001) Evolution of floral meristem identity genes: analysis of Lolium temulentum genes related to APETALA1 and LEAFY of Arabidopsis. Plant Physiol 125:1788–1801 Hirata M, Cai H, Inoue M, Yuyama N, Miura Y, Komatsu T, Takamizo T, Fujimori M (2006) Development of simple sequence repeat (SSR) markers and construction of an SSR-based linkage map in Italian ryegrass (Lolium multiflorum Lam.). Theor Appl Genet 113:270–279 Humphreys MW, Canter PJ, Thomas HM (2003) Advances in introgression technologies for precision breeding within the Lolium: Festuca complex. Ann Appl Biol 143:1–10 Hutchinson J, Rees H, Seal AG (1979) Assay of the activity of supplementary DNA in Lolium. Heredity 43:411–421 Ikeda S, Takahashi W, Oishi M (2004) Generation of expressed sequence tags from cDNA libraries of Italian ryegrass (Lolium multiflorum Lam.). Grassl Sci 49:593–598 Inoue M, Gao Z, Hirata M, Fujimori M, Cai H (2004) Construction of a high-density linkage map of Italian ryegrass (Lolium multiflorum Lam.) using restriction fragment length polymorphism, amplified fragment length polymorphism, and telomeric repeat associated sequence markers. Genome 47:57–65 Jauhar PP (1993) Cytogenetics of the Festuca–Lolium complex. Relevance to breeding. In: Frankel R, Grossman M, Linskens HF, Maliga P, Riley R (eds) Monographs on theoretical and applied genetics, vol 18. Springer, Berlin, pp 12–19 Jenkin TJ (1935) Interspecific and intergeneric hybrids in herbage grasses. II. Lolium perenne  L. temulentum. J Genet 31:379–412 Jenkin TJ (1954) Interspecific and intergeneric hybrids in herbage grasses. II. Lolium perenne with other Lolium species. J Genet 52:300–317 Jones ES, Dupal MP, Kolliler R, Drayton MC, Forster JW (2001) Development and characterization of simple sequence repeat (SSR) markers for perennial ryegrass (Lolium perenne L.). Theor Appl Genet 102:405–415 Jones E, Dupal M, Dumsday J, Hughes L, Forster J (2002) An SSR-based genetic linkage map for perennial ryegrass (Lolium perenne L.). Theor Appl Genet 105:577–584 King RW, Moritz T, Evans LT, Junttila O, Herlt AJ (2001) Long-day induction of flowering in Lolium temulentum involves sequential increases in specific gibberellins at the shoot apex. Plant Physiol 127:624–632 King RW, Mander LN, Asp T, MacMillan CP, Blundell CA, Evans LT (2008) Selective deactivation of gibberellins

H. Cai et al. below the shoot apex is critical to flowering but not to stem elongation of Lolium. Mol Plant 1(2):295–307 Kirigwi FM, Zwonitzer JC, Rouf Mian MA, Wang ZY, Saha MC (2008) Microsatellite markers and genetic diversity assessment in Lolium temulentum. Genet Resour Crop Evol 55:105–114 Li YC, Korol AB, Fahima T, Beiles A, Nevo E (2002) Microsatellites: genomic distribution, putative functions and mutational mechanisms: a review. Mol Ecol 11:2453–2465 Litt M, Luty JA (1989) A hypervariable microsatellite revealed by in vitro amplification of a dinucleotide repeat within the cardiac muscle actin gene. Am J Hum Genet 44:397–401 McGrath S, Hodkinson TR, Barth S (2007) Extremely high cytoplasmic diversity in natural and breeding populations of Lolium (Poaceae). Heredity 99:531–544 Mian MA, Saha MC, Hopkins AA, Wang ZY (2005) Use of tall fescue EST-SSR markers in phylogenetic analysis of coolseason forage grasses. Genome 48:637–647 Moon CD, Scott B, Schardl CL, Christensen MJ (2000) The evolutionary origins of Epichloe¨ endophytes from annual ryegrasses. Mycologia 92:1103–1118 Morgante M, Hanafey M, Powell W (2002) Microsatellites are preferentially associated with nonrepetitive DNA in plant genomes. Nat Genet 30:194–200 Naylor B (1960) Species differentiation in the genus Lolium. Heredity 15:219–233 Rees H, Jones GH (1967) Chromosome evolution in Lolium. Heredity 22:1–18 Saha MC, Mian MA, Eujayl I, Zwonitzer JC, Wang L, May GD (2004) Tall fescue EST-SSR markers with transferability across several grass species. Theor Appl Genet 109:783–791 Saha MC, Mian R, Zwonitzer JC, Chekhovskiy K, Hopkins AA (2005) An SSR- and AFLP- based genetic linkage map of tall fescue (Festuca arundinacea Schreb.). Theor Appl Genet 110(2):323–336 Sawbridge T, Ong EK, Binnion C, Emmerling M, McInnes R, Meath K, Nguyen N, Nunan K, O’Neill M, O’Toole F, Rhodes C, Simmonds J, Tian P, Wearne K, Webster T, Winkworth A, Spangenberg G (2003) Generation and analysis of expressed sequence tags in perennial ryegrass (Lolium perenne L.). Plant Sci 165:1089–1100 Schardl CL, Leuchtmann A, Spiering MJ (2004) Symbiosis of grasses with seedborne fungal endophytes. Annu Rev Plant Biol 55:315–340 Scholz H, Stierstorfer CH, Gaisberg MV (2000) Lolium edwardii sp. Nova (Gramineae) and its relationship with Schedonorus sect. Plantynia DUMORT. Feddes Repert 111: 561–565 Senda T, Kubo N, Hirai M, Tominaga T (2003) Development of microsatellite markers and their effectiveness in Lolium temulentum. Weed Res 44:136–141 Senda T, Saito M, Ohsako T, Tominaga T (2004) Analysis of Lolium temulentum geographical differentiation by microsatellite and AFLP markers. Weed Res 45:18–25 Senda T, Hiraoka Y, Tominaga T (2006) Inheritance of seed shattering in Lolium temulentum and L. persicum hybrids. Genet Resour Crop Evol 53:449–451 Sharifi Tehrani M, Mardi M, Saeidi H, Gharehyazi B, Assadi M (2008) Transferability of genomic and EST-microsatellites from Festuca arundinacea Schreb. to Lolium persicum Boiss. and Hohen. ex Boiss. Int J Bot 4(4):476–480

10 Lolium Soreng RJ, Terrell EE (1997) Taxonomic notes on Schedonorus, a segregate genus from Festuca or Lolium, with a new Nothogenus x Schedololium, and new combinations. Phytologia 83:85–88 Stewart AV (2004) The Canary Islands endophytes of Lolium – a clarification. 5th international symposium on Neotyphodium/ Grass interactions, May 23–26, Layfayette, Arkansas, USA Stynes BA, Bird AF (1993) Development of annual ryegrass toxicity. Aust J Agric Res 34:653–660 Tautz D, Trick M, Dover GA (1986) Cryptic simplicity in DNA is a major source of genetic variation. Nature 322:652–656 Terrell EE (1966) Taxonomic implications of genetics in ryegrasses (Lolium). Bot Rev 32:138–164 Terrell EE (1968) A taxonomic revision of the genus Lolium. USDA Tech Bull 1392:65 Thomas HM (1981) The giemsa C-band karyotypes of six Lolium species. Heredity 46:263–267 Thomas H, Morgan WG, Thomas AM, Ougham HJ (1999) Expression of the stay-green character introgressed into Lolium temulentum Ceres from a senescence mutant of Festuca pratensis. Theor Appl Genet 99:92–99 Thorogood D, Hayward MD (1992) Self-compatibility in Lolium temulentum L.: its genetic control and transfer into L. perenne L. and L. multiflorum Lam. Heredity 68: 71–78 Wang Z, Weber JL, Zhong G, Tanksley SD (1994) Survey of plant short tandem DNA repeats. Theor Appl Genet 88:1–6

173 Wang Z-Y, Scott M, Hopkins A (2002) Plant regeneration from embryogenic cell suspension cultures of Lolium temulentum. In Vitro Cell Dev Biol Plant 38:446–450 Wang ML, Gillespie AG, Newman ML, Dean RE, Pittman RN, Morris JB, Pederson GA (2004) Transfer of simple sequence repeat (SSR) markers across the legume family for germplasm characterization and evaluation. Plant Genet Res 2:107–119 Wang ML, Barkley NA, Yu J-K, Dean RE, Newman ML, Sorrells ME, Pederson GA (2005a) Transfer of simple sequence repeat (SSR) markers from major cereal crops to minor grass species for germplasm characterization and evaluation. Plant Genet Resour 3:45–57 Wang Z-Y, Ge Y, Mian R, Baker J (2005b) Development of highly tissue culture responsive lines of Lolium temulentum by anther culture. Plant Sci 168:203–211 Wit F (1974) Cytoplasmic male sterility in ryegrasses (Lolium spp.) detected after intergeneric hybridization. Euphytica 23:31–38 Yamada T (2001) Introduction of a self-compatible gene of Lolium temulentum L. to perennial ryegrass (Lolium perenne L.) for the purpose of the production of inbred lines of perennial ryegrass. Euphytica 122:213–217 Yamada T, Forster JW, Humphreys MW, Takamizo T (2005) Genetics and molecular breeding in Lolium/Festuca grass species complex. Grassl Sci 51:89–106 Yu Q, Han H, Nguyen L, Forster JW, Powles SB (2009) Paraquat resistance in a Lolium rigidum population is governed by one major nuclear gene. Theor Appl Genet. doi:10.1007/ s00122-009-1008-3

.

Chapter 11

Panicum Hem S. Bhandari, Masumi Ebina, Malay C. Saha, Joseph H. Bouton, Sairam V. Rudrabhatla, and Stephen L. Goldman

11.1 Introduction The position of the grasses and their impact on the entire globe cannot be undervalued. Indeed, the grasses are virtually ubiquitous and have been calculated to cover approximately 20% of the Earth’s surface (Mathews 1983). Given the role the grasses play in the nourishment for both animals and humans, nearly 50% of the planetary digestible energy can be ascribed to these plants (Evans 1993). Moreover, with increasing demands for cheaper sources of energy, grasses are likely to play a key role, given the opportunity to derive cellulosic-derived biofuels. Whether the Panicum species will figure prominently energy in economics remains to be seen. However, some Panicum species reproduce by apomixis and it should be possible to isolate and clone those genes responsible for the stabilization of genotypes through maternal inheritance. This in turn, would open the possibility of transferring clonal propagation to sexually reproducing plants. The importance of transfer of those genes encoding apomixis from Panicum on the economics of caloric ingestion remains to be determined.

11.2 Botany Description of over 460 species of grasses previously classified under Panicum genera are available in Royal Botanical Garden, Kew “GrassBase” database S.L. Goldman (*) Department of Environmental Science, The University of Toledo, 2801 W. Bancroft, Toledo, OH 43606, USA e-mail: [email protected]

(Clayton et al. 2006), which include 68 species of Panicum under revised taxonomic delimitation. The summary of the description for these 68 species is found in Table 11.1. Panicum species are either annual or perennial in growth habit with rhizomatous or stoloniferous roots. Lateral branches are lacking or sparsely present arising from lower culms. Culms are erect or geniculately ascending or decumbent, which can be as short as 2–6 cm (Panicum mohavense) to as long as 400 cm (P. altum; P. decidum) (Hitchcock 1971; Watson and Dallwitz 1994; Clayton et al. 2006). Culms are comprised of internodes and nodes. These species have well developed terminal or axillary inflorescence called a panicle. Axillary inflorescences can either be compact or enclosed. In species such as P. campestre, P. cayennense, and P. rudgei, axillary inflorescences are intermingled with terminal inflorescences. Panicles are well exerted or embraced at the base by subtending leaf or are subtended by the inflated leaf sheath. The peduncle can be smooth or hispid as well as glabrous, pubescent, hirsute, or pilose above and up to 50 cm long (i.e., P. longissimum). Panicles are open or effuse in addition to being ovate, oblong, or elliptic and can be 1 cm (P. mohavense) to 65 cm (P. ligulare) long. Primary branches are appressed, ascending, or spreading. Panicle axis is either smooth or scaberulous. Spikelets comprise one basal sterile floret and one fertile floret without rachilla extension. Fertile florets are complete and are self-fertile (many diploids) or self-incompatible (almost all tetraploids). Spikelets are ovate, lanceolate, or elliptic as well as dorsally compressed, obtuse, acute, or acuminate, and as short as 1.2 cm (P. philadelphicum) to 9 cm (P. cervicatum) long. Spikelets are usually appressed and are often

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_11, # Springer-Verlag Berlin Heidelberg 2011

175

176

H.S. Bhandari et al.

Table 11.1 Morphological variability and growth habit of different Panicum species Species Growth Culm Culm Leaf length Leaf width habit typea length (cm) (cm) (mm) Section: Dichotomiflora Panicum aquaticum Annual E or GA 25–200 6–30 4–8 Panicum coloratum Perennial E or D 15–100 12–30 5–10 Panicum Annual GA or D 100–200 12–50 3–12 dichotomiflorum Panicum Perennial na 80–100 30–40 11–20 elephantipes Panicum gouinii Perennial E 10–50 5–13 2–4 Panicum lacustre Panicum pedersenii Panicum repens Panicum sublaeve Panicum vaseyanum Panicum alatum

Panicle length (cm)

Panicle type

4–27 4–30 12–40

Open/ovate Open/oblong Open/ovate

15–30

Open/ovate

5–10

Contracted/ lanceolate Open/ovate Open/ovate/effuse Open/oblong Open/ovate Open/Laneolate

Perennial Perennial Perennial Perennial Annual

E na E or GA E of GA D

100–150 70–100 30–100 15–20 50–70

15–30 11–27 7–25 8–10 5–10

2 4–7 2–8 5–10 3–8

20–30 8–23 5–20 5–8 4–7

Annual

E or GA

13–95

2–18

3–17

4–23

Open/pyramidal/ effuse

Section: Panicum Panicum aquarum Annual Panicum aztecanum Perennial

E E

90–130 50–100

11–18 14–30

2–4 1–3

35–40 15–18

Panicum bergii

Perennial

na

30–60

10–37

1–6

10–20

Panicum capillare

Annual

E of GA

20–80

7–30

5–14

15–50

Panicum capillarioides Panicum chaseae Panicum decolorans Panicum diffusum Panicum ephemeroides Panicum exiguum

Perennial

E of GA

30–60

20–50

2–12

10–30

Open/pyramidal Open/pyramidal/ effuse Open/pyramidal/ effuse Open/oblong/ovate/ effuse open/oblong/effuse

Perennial Annual

na na

40 30–60

na 8–15

2–3 7–10

15–23 10–18

Open/ovate Open/Lanceolate

Perennial Annual

GA or D E

20–100 60–70

5–220 20–30

1–5 2

5–25 30

Open/Pyramidal Open/elliptic/effuse

Annual

D

10–30

2–6

2–4

3–8

Panicum fauriei Panicum flexile Panicum furvum Panicum ghiesbreghtii Panicum hallii Panicum hirsutum Panicum hirticaule Panicum hispidifolium Panicum lepidulum Panicum magnispicula Panicum miliaceum Panicum mohavense

Annual Annual Perennial Perennial

na E E GA or D

2–30 12–75 10–21 35–110

2–12 6–25 2–7 11–30

1–4 2–7 2–4 4–8

1–11 7–20 4–6 20–35

Perennial Perennial Annual Annual

E or GA na E/GA/D na

20–80 100–130 25–80 50

4–30 35–70 7–27 15

2–10 10–30 6–15 8

6–20 15–50 13–35 23

Open/pyramidal/ effuse Open/linear Open/oblong Open/ovate Open/ovate/ pyramidal Open/pyramidal Open/elliptic/dense Open/ovate Open/ovate

Perennial Perennial

E na

25–70 20–30

7–30 10–12

5–10 2

7–20 3–6

Open/elliptic Open/ovate

Annual Annual

E or GA na

20–100 2–6

10–20 1–4

6–15 2–3

5–10 1–2

Contracted/obovate Open/oblong/dense

Perennial

na

30–90

25–60

5–15

30–40 (continued)

11 Panicum

177

Table 11.1 (continued) Species Growth habit Panicum mucronulatum Panicum Perennial nephelophilum Panicum Perennial pampinosum Panicum parcum Annual Panicum Perennial peladoense Panicum Annual philadelphicum Panicum Perennial quadriglume Panicum Annual stramineum Panicum Perennial tamaulipense Panicum campestre Perennial Section: Rudgeana Panicum cayennense Panicum cervicatum Panicum ligulare Panicum rudgei Panicum chloroleucum

Panicum decidum Panicum glabripes Panicum longissimum Panicum tricholaenoides Panicum virgatum

Culm length (cm)

Leaf length (cm)

Leaf width (mm)

Panicle length (cm)

Panicle type

na

45–240

10–40

8–25

15–60

Open/oblong/ obovate Open/elliptic/ovate

GA

NA

5–12

5–10

6–15

Open/ovate

E E

50–125 20–60

11–35 7–22

6–10 2–4

20–35 7–18

Open/ovate Open/ovate

E

20–50

5–15

2–7

7–15

Open/ovate

na

40–120

10–40

4–10

7–20

Open/ovate

GA

20–50

10–30

4–15

5–15

Open/ovate

E

50–150

10–50

3–8

15–40

Open/pyramidal

E or GA

30–74

7–25

5–7

5–15

Open/oblong/effuse

Annual

E or GA

18–40

9–13

5–8

5–16

Open/ovate

Perennial

na

50–150

12–40

7–18

30–50

Open/oblong/ovate

Perennial Perennial Perennial

E E or GA E

130–200 75–125 40–70

30–55 20–40 15–30

9–19 8–14 1–4

47–65 30–40 10–30

Open/ovate/effuse Open/ovate/effuse Open/ovate/effuse

na

45–55

23–43

1–4

11–25

Contracted/linear

E

50–100

10–20

4–7

25–30

Open/elliptic

D

200–400

30–45

8–15

20–30

Open/ovate

Perennial

E

30–200

10–50

3–15

10–35

Perennial Perennial Perennial

na E E

200–400 70–130 150–200

7–9 25–45 50–80

4–8 4–5 5–9

3–7 20–26 10–30

Contracted/ lanceolate Open/ovate/dense Open/ovate Contracted/linear

Perennial

na

150–200

25–35

7

30–40

Open/pyramidal

Perennial

E, GA, D

30–300

10–60

3–15

15–55

Open/ovate

Section: Urvilleana Panicum Perennial racemosum Panicum Perennial urvilleanum Panicum altum Perennial Section: Virgata Panicum amarum

Culm typea

Source: Aliscioni et al. (2003) and Clayton et al. (2006) E erect, GA geniculately ascending, D decumbent

a

solitary or occasionally in pairs. Fertile spikelets are pedicelled. Pedicel can be clavate, scabrous, glabrous, or hairy at tip. Fruit is in the form of caryopsis with adherent pericarp. Plant growth stages of Panicum species follow that of most perennial forage grasses (Moore et al. 1991). Five major growth stages including germination,

vegetative leaf development, stem elongation, development of reproductive structures (panicle initiation to anthesis), and seed development and ripening are all established. These stages are further broken down into several substages. The duration of each stage, however, varies according to species, ecotypes within species, genotype within each ecotype, and responses

178

to growth conditions such as photoperiod, temperature, etc. For example, upland ecotypes of P. virgatum L. attain maturity several days earlier than lowland ecotypes. Yet, maturity may become delayed in both ecotypes as one goes to northern latitudes. Traditionally, different species in Panicum were believed to possess either C3 or C4 or C3–C4 intermediate photosynthetic pathway (Brown et al. 1985; Bouton et al. 1986; Leonardos and Grodzinski 2000). The differences between C3, C4 and C3–C4 intermediate species of Panicum and other grasses in leaf anatomy, physiology, and biochemical processes have been documented (Bouton et al. 1986; Warner et al. 1987; Dengler et al. 1994). Existence of C3/C4 pairs of closely related species as well as their molecular phylogeny results suggests the possible two-way evolutionary reversal between C3 and C4 grasses (Duvall et al. 2003; Christin et al. 2009; Ibrahim et al. 2009). The C3–C4 intermediate types possess Krantz like leaf anatomy that show reduced photorespiration and reduced O2 inhibition of apparent photosynthesis compared to their C3 counterpart (Brown and Morgan 1980). It is noteworthy that closely related pair of C3 and C3–C4 species can hybridize and produce viable progeny (Brown et al. 1985; Bouton et al. 1986). The interspecific hybrid between C3 and C3–C4 type produced F1, which showed leaf anatomy, morphology, O2 inhibition, and CO2 compensation intermediate of the two parent types. Several species of C3 or C3–C4 intermediate types of photosynthesis pathway that were previously described under Panicum are now assigned to other genera (Aliscioni et al. 2003; Zuloaga et al. 2006). Apparently, almost all of the Panicum species under revised description follow C4 type of NAD–me photosynthetic pathway (Giussani et al. 2001; Aliscioni et al. 2003; Zuloaga et al. 2006; Christin et al. 2009). These species have specialized, centrifugally placed chloroplasts in the bundle sheath, without well-developed grana, and have lost the parenchymatous outer sheath around the vascular bundles. Deviation from these structures has also been reported in Panicum species, i.e., P. prionitis preserves the outer bundle sheath (Giussani et al. 2001). The C4 photosynthesis consists of specialized anatomical structures and biochemical novelties that create CO2 pump and build up CO2 concentration around Rubisco and reduce photorespiration (Christin et al. 2009). Due to these features, the C4 species are well known for competitive adaptive advantage over C3 species under

H.S. Bhandari et al.

various stresses, such as heat and drought. Data from 14 different Panicum and other grass species indicated that the C4 type has higher rate of photosynthesis and immediate export compared to C3 species (Leonardos and Grodzinski 2000). This has made C4 species a crop of choice in most of the tropical and subtropical regions of the world.

11.3 Allied Crops The Paniceae exhibits great diversity with respect to morphology, anatomy, photosynthetic physiology and karyology (Clayton and Renvoize 1986; Watson and Dallwitz 1994). Indeed, significant attempts were made to use this information prior to the development of molecular markers to understand the taxonomic relationships among related plants. On this basis, at one time, Panicum was divided into six subgenera. These include Panicum, Dicanthelium, Steinchisma, Agrostoides, Megathyrus, and Phanopyrum (Zuloaga et al. 2000; Giusanni et al. 2001; Aliscioni et al. 2003; Simon and Jacobs 2003; Reinheimer et al. 2005). The cytolological, anatomical, and morphological differences and, for that matter, similarities are summarized in Table 11.2. Today, each has been elevated to an independent genus and more are being added (Morrone et al. 2008).

11.4 Distribution Panicum species are adapted to tropical, subtropical, and warm temperate regions throughout the world. Some species are distributed more widely than others. For example, P. miliaceum and P. capillare of section Panicum are found adapted to tropical, subtropical and temperate regions of countries in all the continents, while P. virgatum is found primarily in the US, Mexico, and Brazil in addition to other countries of South America, regions of Mesoamerica, and the Caribbean. Three species, P. coloratum, P. dichotomiflorum, and P. repens, of section Dichotomiflora are distributed in tropical, subtropical, and temperate environments in countries in North America, South America, Africa,

Erect and cespitose

2 sheaths around bundles outer sheath contains specialized centripede chloroplasts Pyramid like

Ellipsode to lancelolate 7–9 nerved

Compound or simple papillae

C4

Culm

Vascular bundles

Spikelet

Anthecium

Photosynthesis

Glumes

Inflorescence

x¼9

Chromosome #

–

–

–

–

–

North American types express distinctive basal rosettee –

x ¼ 10

–

–

Mesotome sheath specialized chloroplasts

Section dependent x ¼ 8 or 9 –

Agrostoides

Upper glume and – lower lemma 3–5 nerved – Entire upper surface Upper/transverse rugose þ covered with cpd. papillae on papillae wrinkles C3/C4 intermediate C4 C4

Open or contracted panicles –

–

Lax pyramid – like many flowered panicle

–

–

–

Polymorphic x ¼ 8 or 9 –

x ¼ 10

Table 11.2 Anatomical, morphological, and cytological similarities among Genera related to Panicum Trait Panicum Dicanthelium Steinchisma Megathyrus

–

Upper glume and lower lemma 7–9 nerved –

–

Upper anthecium is indurate with simple pillae –

Lower glume is 1/2 length

Lower flower, lower palea absent –

–

–

Lax and open

–

Parodiophyollochloa x ¼ 10

Section dependent x ¼ 9 or 10 –

Phanopyrum

11 Panicum 179

180

Europe, Asia, and Australasia. Interestingly, the adaptation of large number of Panicum species is restricted to countries in the western hemisphere; US, Mexico, the Caribbean, Mesoamericana, Brazil, and other countries in South America. Brazil and its neighboring regions possess diverse species of Panicum. Of the 100 species of Panicum, 50 are of American origin, which are distributed from Canada through Argentina (Guglieri et al. 2006); 29 were found in Brazil alone (Guglieri et al. 2004).

11.5 Conservation Initiatives Two major global initiatives have been started to appropriately track and catalog the world’s germplasm and derivative genetic resources. These are the Systemwide Information Network for Genetic Resources (SINGER) and the Consultative Group on International Agricultural Research (CGIAR). SINGER and CGIAR provide a resource to determine where particular seed lines may be found and, in addition, provide critical information related to the collection source, the status of the sample, and whether the seed is derived from domesticated or wild lines. The location of an abundance of molecular resources is also listed. The precise location of Panicum germplasm and accompanying molecular capital are indicated in the body of this review as appropriate. Unlike corn, soybean, and wheat, for example, the databases and seed resources that are highly centralized, Panicum assets are more widely dispersed in smaller collections, those in India, Japan, and along the Pacific Rim being among the more notable.

11.6 Cytology and Karyotype As in most species of the Poaceae, the karyotype of most Panicums is symmetrical and uniform with metacentric or submetacentric chromosomes (Zuloaga et al. 1989). Chromosomes paired predominantly as bivalent. However, few monovalent and quadrivalent in tetraploids and univalent, quadrivalent, or hexavalent in hexaploid cytotypes were also reported

H.S. Bhandari et al.

(Nakagawa and Hanna 1990; Hamoud et al. 1993; Morrone et al. 1995, 2006; Zuloaga et al. 2006).

11.7 Origin of the Genera and Molecular Phylogeny Phylogeny of the tribe Paniceae, subfamily Panicoideae, was analyzed with sequence data obtained from the chloroplast gene ndhF (Giussani et al. 2001). Most of the proteins involved in C4 photosynthesis are encoded by nuclear genes. Therefore, a chloroplast gene should provide an independent history from the genes selected for C4 photosynthesis. The ndhF gene indicates a relatively high rate of molecular evolution in the grasses (Clark et al. 1995). Phylogenetic analysis of 103 species of subfamily Panicoideae clearly grouped into three clades, corresponding largely to the same basic chromosome number (Andropogoneae [x ¼ 10], Paniceae [x ¼ 10], and Paniceae [x ¼ 9]). With a few exceptions, most members of the x ¼ 9 Paniceae clade have nine chromosomes. The x ¼ 9 clade includes members of five subgenera of Panicum (subg. Panicum [C4], Dichanthelium [C3], Agrostoides [C4], Phanopyrum [C3], and Megathyrus [C4]). The subgenus Panicum was characterized as monophyletic, and Dichanthelium appeared as polyphyletic. P. sabulorum and P. koolauense of the subgenus Dichanthelium were unrelated to the other member, P. ovuliferum, of this subgenus. Many of the Panicum species (P. virgatum, P. elephantipes, P. nephelophilum, P. fauriel, P. rudgel, P. repens, P. pedersenil, P. aquaticum, P. bulbosum) fall into a group called Setaria/Urochloa/Panicum clade (supported by 88% bootstrap with 6 bp substitutions). All species in this clade exhibit C4 photosynthesis (Giussani et al. 2001). This study was conducted with large number of species and further expanded the work of Gomez-Martinez and Culham (2000) who divided the tribe Paniceae into two clades corresponding to chromosome number correlated with biogeography (American Paniceae [x ¼ 10] and Pantropical Paniceae [x ¼ 9]). Aliscioni et al. (2003) further expanded the classification study of Giussani et al. (2001). A total of 123 sequences were analyzed, of which 111 represent the tribe Paniceae (66 representing Panicum s.l.). Panicum phylogeny was estimated from a parsimony analysis based on all available chloroplast ndhF gene

11 Panicum

181

Table 11.3 Chromosome number and ploidy level of different Panicum and other related grass species and their configuration during dikenensis of metaphase Ia Taxon Basic chr. # Ploidy Configuration during dikenensis Source of metaphase P. Capillaire 9 2x 8 II þ 2 I Hamaud et al. (1994) P. coloratum 9 2x, 4x 9 II “or” 18 II Hamaud et al. (1994) P. cyanescens 18 4x 18 II or 17 II þ 2 I Morrone et al. (1995) P. hirticaule 9 2x 9 II Morrone et al. (2006) P. miliaceum 9 4x 18 II þ few univalent Hamaud et al. (1994) P. virgatum 9 4x, 6x, 8x 18 II “or” 35 II þ 2 I Lu (1995) P. cipoens 20 4x 20 II Morrone et al. (1995) P. euprepes 20 4x 1 IV þ 18 II “or” 19 II þ 2 I Morrone et al. (1995) P. molinioides 20 4x 19 II þ 2 I or 1 IV þ 17 II þ 2 I Morrone et al. (1995) or 2 IV þ 16 II P. panicillatum ca. 27 6x 27 II Morrone et al. (1995) P. poliophyllum 20 4x 20 II or 1 IV þ 18 II Morrone et al. (1995) or 1 IV þ 17 II þ 2 I P. stenophyllum 10 2x 10 II Morrone et al. (1995) P. stoloniferum 10 2x 10 II Morrone et al. (2006) P. trichanthum 18 4x 18 II Morrone et al. (2006) P. validum 10 2x 10 II Zuloaga et al. (1989) P. wettsteinii 9 2x 9 II Morrone et al. (1995) P. assurgens 9 2x 9 II Morrone et al. (1995) P. cabrerae 9 2x 9 II Morrone et al. (1995) P. heliophilum ca.30 6x Ca. 30 Morrone et al. (1995) P. maximum 10 2x 10 II Nakagawa and Hanna (1990) P. laevifolium 9 2x 8 II þ 2 I Hamaud et al. (1994) P. antidotala 9 2x 8 II þ 2 I Hamaud et al. (1994) P. miliare 9 4x 16 II þ 4 I Hamaud et al. (1994) a

Previously classified under genus Panicum, #7–23 are now assigned to other genera

sequences of the tribe Paniceae. Panicum sensu stricto (s.s.) was restricted to the former subgenus Panicum and supported the recognition of separate genera for Dichanthelium, Phanopyrum, and Steinchisma. The subgenus Panicum was clearly monophyletic (supported by 97% bootstrap) and completely included in the x ¼ 9 Paniceae clade.

11.8 Cytogenetics Panicum species vary widely in chromosome number as well as in ploidy levels (Nielsen 1944; McMillan and Weller 1959; Zuloaga et al. 1989; Hamoud et al. 1993; Morrone et al. 2006) (Table 11.3). Ploidy levels vary from diploid to dodecaploid. The basic chromosome numbers reported for the Panicum genus are predominantly 9 or 10 with a few reported to have 7 or 8 or 11 (Giussani et al. 2001). In switchgrass, all

different ploidy levels from diploid (2n ¼ 2x ¼ 18) to dodecaploid (2n ¼ 12x ¼ 108) have been reported (Church 1940; Burton 1942; Nielsen 1944). However, only two major ploidy levels, the tetraploids (2n ¼ 4 x ¼ 36) and the octoploids (2n ¼ 8x ¼ 72), are predominant (Hopkins et al. 1996). Two cytotypes of switchgrass, U and L, associated with upland and lowland ecotypes have been recognized (Hultquist et al. 1996). The lowland cytotypes are entirely tetraploids and the upland cytotypes are either tetraploids, hexaploids, or octoploids (Barnett and Carver 1967). P. miliaceum and P. sumatrense are tetraploids (2n ¼ 4x ¼ 36) species. P. capillare and P. sonorum are reported as diploid (2n ¼ 2x ¼ 18) species (Tsvelev and Zhukova 1974; Hiremath et al. 1990). Ploidy level of several P. virgatum populations have been estimated using microscopy and flow cytometry techniques (Hopkins et al. 1996; Lu et al. 1998). These studies have demonstrated that plants with average nuclear DNA contents of 2.7 to 3.3 pg are

182

tetraploid populations and 4.7–6.0 pg are octoploid populations. Nuclear DNA contents of the 31 GRIN switchgrass populations were estimated using flow cytometry (Narasimhamoorthy et al. 2008). Twenty accessions were classified as tetraploids, with average DNA content ranging from 2.4 to 3.4 pg. Four cultivars, Trailblazer, Caddo, Blackwell, and Argentina, were clearly classified as octoploids with average DNA content ranging from 4.5 to 5.8 pg. The remaining seven accessions were identified with mixed ploidy levels, which include well characterized cultivars such as Cave-in-rock, Blackwell, Shawnee, and Alamo (Narasimhamoorthy et al. 2008). However, plants obtained from commercial seed lots of Alamo were all tetraploids (mean DNA content of 2.97 pg) while Blackwell and Shawnee were octoploids (mean

Fig. 11.1 (continued)

H.S. Bhandari et al.

DNA content of 4.76 pg), respectively. Within Cavein- rock, only one seedling each was designated as tetraploid and hexaploid with a mean DNA content of 2.96 and 3.68 pg, respectively. Mixed ploidy levels were only detected in Cave-in-rock with tetraploid, hexaploid, and octoploid genotypes being reported (Narasimhamoorthy et al. 2008). P. maximum jacq. (guineagrass), a tropical grass native to Africa, has 2n ¼ 16, 32, and 48 chromosomes at diploid, tetraploid, and hexaploid levels, respectively (Nakagawa and Hanna 1990). Diploid plants of this species follow sexual mode of reproduction, while its tetraploid counterpart has shown mainly apomictic mode of reproduction. Aneuploidy, such as 2n ¼ 18, 40, and 44, are also reported for this species. Experimental evidence suggests that the hexaploid

11 Panicum

183

cytotypes resulted from fertilization of unreduced 2n ¼ 32 egg from apomictic plants by a reduced 16chromosome male gamete. Interestingly, the tetraploid and hexaploid cytotypes are similar in most of the morphological traits, yet 2n ¼ 40 plants show reduced plant height (Nakagawa and Hanna 1990). The nuclear DNA content of diploid guineagrass genome is estimated 500 Mbp (Akiyama et al. 2008). In Panicum maximum Jack., linkage studies have been most effective in analyzing the transmission of apomixis. These molecular linkage maps were constructed by analyzing the progeny derived from a cross between the sexual, colchicine induced, autotetraploid “Noh PL1” and the apomict “Natsukaze” (Ebina et al. 2005). Using various combinations of 56 amplified fragment length polymorphism (AFLP) primers and 41 random amplified polymorphic DNA (RAPD) primers, a total of 360 simplex marker loci were assigned to 39 linkage groups of the cultivar “Natsukaze” as shown in Fig. 11.1. The frequency of

the simplex marker loci in AFLP primer combinations was 6.4 in “Natsukaze” compared to 2.6 in the maternal “Noh PL1”. Thus, the sexual tetraploid line accumulates duplex or multiplex loci, as compared to “Natsukaze.” In contrast, “Natsukaze” has frequent simplex loci. These simplex markers spread the entire linkage map, which covered 1,703.5 cm of the autotetraploid genome (2n ¼ 4x ¼ 32), with an average spacing of 4.7 cm. The number of map units per linkage group correlated with the number of markers per linkage group (R ¼ 0.68), confirming that the loci are randomly distributed across the map. In a mapping population designed to determine the number of segregating genes regulating apomixis, the ratio of apomicts to sexual was 1:1. Among some apomictic segregants, polyembryo was observed, and conjugated cellular structure was observed in the embryosac as shown in Fig. 11.2. Carman (1997) suggested that the polyspory of adventitious embryony and polyembryony of gametophytic apomixis have a

Fig. 11.1 An AFLP and RAPD linkage map of aposporous guineagrass cultivar “Natsukaze” pollen patent. Map distances are in cm in left side of the linkage group. The name of marker

loci is on the right side. The size of the linkage group is shown in the bottom of the each group. The aposporous locus (Apo) was mapped in the group 8

184

common origin. Each is a function of asynchronous expressions of sexual embryo construction genes as a consequence of multiple gene copies (Carman 1997). As predicted by this model, both the sexual progeny of the mapping population in addition to natural sexual accessions never exhibit polyembryony in guineagrass. In linkage group 8, shown in Fig. 11.1, the apomixis trait cosegregated with nine AFLP markers mapping near the center of the linkage group. An expanded version of this linkage group 8 is shown in Fig. 11.3. Thirteen AFLP markers flanked the apomixis region at 1.4 cm, and three measured 4.2 cm to the opposite side. The marker composition of all individuals of the mapping progeny around the aposporous

Fig. 11.2 Embryo sac analyses of the mapping population. GM64 is the name of the population. An individual named GM64-20 presents a typical sexual embryo sac. GM64-50 presents a typical apomixis embryo sac, however, which appears as a twin embryo and cellar conjugation form

Fig. 11.3 An expanded version of linkage group 8 containing the aposporous locus (Apo) and tightly linked markers. Markers cosegregating and tightly linked to Apo are shown in the boxes at the right of the corresponding marker

H.S. Bhandari et al.

apomixis region is consistent with their mode of reproduction. This study is now being expanded by converting these cosegregating AFLP markers into sequence tagged markers. Large numbers of molecular markers, including those completely linked to apospory, have also been annotated in Paspalum notatum (Martinez et al. 2003; Stein et al. 2004; Carneiro et al. 2006). In two species, a small genetic region is conserved due to the suppression of crossing over in spite of the fact that physical interval marking the chromosome is large. This is likely to be the first instance of recombination restriction in Pennisetum squamulatum. The physical size of the apomixis-related gene region (ASGR) is 50 Mbp in P. squamulatum (Akiyama et al. 2004) and 40 Mbp in P. notatam (Carneiro et al. 2006). The magnitude of ASGR in guineagrass is also similar with aposporycontrolling locus in P. maximum and P. notatum with restricted recombination across a large physical region of the chromosome. By analyzing putative polymorphisms focused on the ASGR region, an attempt will be made to dissect the phylogeny of aposporous apomixes among these grasses. In addition to these regional polymorphisms, the next generation sequencers can directly read the huge physical region of the ASGR (Conner et al. 2008). In this study, one-tenth of the region has been

11 Panicum

185

Table 11.4 Nuclear DNA contents determined by flow cytometry and the number of ribosomal DNA (rDNA) loci in Panicum maximum Strains Reproduction No. of Analyzed times Mbp 5S 45S chromosomes of DNA content Mode Ploidy C1 SD rDNA rDNA GR-297 Sexual 2 16 12 499 3 2 4 Noh PL1 Sexual 4 32 5 995 7 4 8 Natsukaze Apospory 4 32 6 1,023 10 4 6 Natsuyutaka Apospory 4 32 5 1,045 11 4 4

sequenced by using bacterial artificial chromosome (BAC) contigs from P. squamulatum and C. cilliaris. In the process, many classes of transposable elements in the ASGR were discovered. Equally notable, this sequence has no large-scale synteny with Oryza and Sorghum genomes. Therefore, not only further comparisons of the sequence between apomictic relatives of P. squamulatum and C. cilliaris but also comparisons made with P. maximum and P. notatum need definition. In principle, this technique should elucidate common sequences important to the expression of apospory.

11.9 Ploidy Level and Apomixis-Related Gene Region Cytological analysis was performed after collecting natural accessions of guineagrass (Hanna et al. 1973; Savidan 1980; Nakagawa and Hanna 1990). Guineagrass has a basic chromosome number of x ¼ 8; most plants are apomictic tetraploid. Using colchicines treatment, the chromosome number of several diploid lines was doubled to obtain tetraploid plants (Savidan and Pernes 1982; Nakagawa and Hanna 1992). Given the importance of tropical forage and apomictic lines, the basic genomic information such as genome size and ribosomal DNA (rDNA) ideogram has only recently been reported (Akiyama et al. 2008). The genome of the diploid sexual line of “GR-297” was determined to be approximately 500 Mbp. As such, it is equal in range to rice, and the three sexual and apomictic tetraploids accordingly corresponded to 1,000 Mbp. A detailed comparison of genome size between the sexual and apomictic tetraploids detected a 30–50 Mbp difference in genome size specific to the

Fig. 11.4 Physical mapping of ribosomal DNA (rDNA) on mitotic chromosome in P. maximum (a) GR-297, (b) Noh PL1, (c) Natsukaze and (d) Natsuyutaka. Red and Green arrow heads indicate 45S and 5S rDNA, respectively. Bars, 10 mm (Akiyama et al. 2008)

apomicts (Table 11.4). Although this difference in genome size is not necessarily congruent with the size of apomixis-related region of guineagrass, a similar number of megabases have been reported in P. sequamulatum (Akiyama et al. 2004). Fluorescent in situ hybridization (FISH) analysis with rDNA probes was performed (Fig. 11.4), which shows the number of 5S rDNA loci in diploid and the tetraploids were two and four, respectively. The number of 45S rDNA loci varied from four to eight among the three tetraploid plants. Quantitative ideogram of diploid strain by image analysis, as shown in Fig. 11.5, indicated that the difference in the number of 45S rDNA is dependent on the presence or absence of locus on chromosome 2.

186

H.S. Bhandari et al.

0

50

DAPI intensity 100 150 200 250

Gray region

50

0

50

0

50

Chromosome 1 100

100

150 200 250

0

50

0

50

100

150 200 250

0

50

0

50

100

150 200 250

Black region

0

150 200 250

Chromosome 2

Chromosome 5

100

150 200 250

Chromosome 6

Chromosome 3 100

150 200 250

Chromosome 7

Chromosome 4 100

150 200 250

Chromosome 8

Fig. 11.5 Quantitative ideograms and distribution patterns 4,6diamino-2-phenylindole (DAPI) intensities on chromosomes constructed from nine mitotic chromosome spreads of strain

GR-297. Closed and open circle indicate loci of 5S and 45S rDNA, respectively (Akiyama et al. 2008)

11.10 Molecular Markers

cytotype despite its classification as lowland cytotype by USDA-SCS (Gunter et al. 1996; Hultquist et al. 1996). The genetic variation in the nuclear gene encoding plastid acetyl-coA carboxylase (ACCase) from six switchgrass cultivars was investigated using DNA sequences (Huang et al. 2003). Tetraploid and octoploid switchgrass ecotypes formed two distinct groups and suggested that the most recent polyploidization events that established modern switchgrass lineages occurred less than two million years ago. RFLP markers were also used to study the genetic variation among 21 switchgrass genotypes randomly selected from two lowland (Alamo and Kanlow) and one upland (Summer) synthetic cultivars (Missaoui et al. 2006). The upland cultivar showed a higher degree of genetic variation than the lowlands. Analysis of chloroplast non-coding region of the trnL DNA sequences indicated a deletion of 49 nucleotides that was specific to lowland cytotypes (Missaoui et al. 2006). A large set of microsatellite markers have been developed from the switchgrass EST sequences (Tobias et al. 2008). Several of these markers were

Molecular markers have become a reliable tool for the detection of genetic variability because they directly reveal changes at the DNA sequence level. In the past two decades, several studies have been conducted to find within and among population variation in P. virgatum using restriction fragment length polymorphism (RFLP), random amplification of polymorphic DNA (RAPD), and simple sequence repeat (SSR) molecular markers. Two distinct cytoplasm types, L and U, associated with lowland and upland ecotypes, respectively, were identified using cpDNA polymorphisms (Hultquist et al. 1996). The lowland cultivars contained a restriction site that was not present in the upland cultivars. A broad assessment of the genetic relationship among 137 genotypes representing 14 switchgrass populations were assayed with 91 polymorphic RAPD markers (Gunter et al. 1996). The individual genotypes within each population were clustered into discrete groups within the ecotypes. Both RAPD and cpDNA analyses suggested that Cave-in-rock is an upland

11 Panicum

used for the genetic diversity analysis and mapping of switchgrass populations. Microsatellite markers have also been developed from the (GA/CT)n enriched switchgrass genomic libraries. DArT markers have been developed and used for mapping a switchgrass population developed by crossing between a lowland and an upland genotype (Katrien Devos, UGA personal communication). A large set of SNP markers are being developed for switchgrass at the Cornell University (Ed Buckler, personal communication). The genetic structure in guineagrass was assessed by the development of SSR markers (Ebina et al. 2007). In this report, 13 SSR markers were successfully developed from genomic DNA library enriched for (AG)/(TC) repeats and SSR motifs survey in expressed sequence tag (EST) sequences. From Japanese accessions of guineagrass, 77 randomly selected accessions were screened and their phylogenetic relationship investigated. These 13 SSR primers produced 1–4 amplified fragment(s) within accessions corresponding to the autotetraploidy of guineagrass. Although the limiting number of SSRs is caused by the low redundancy of SSR in the genomic region and in the ESTs, the power of discrimination (PD) and polymorphic information content (PIC) were relatively high comparing to the other self-compatible species. The SSR motifs of guineagrass were 54% perfect, 31% interrupted, and 14% compound. The ratio of the motif variations is likely to the self-compatible allotetrapoid grass species of zoysia grass (Tsuruta et al. 2005). The limited frequency of the SSR appearance in guineagrass is an open question. The high value of PD and PIC supports the hypothesis of intercrossing among the sexual plants or facultative apomicts to create genetic variability in the natural habitat. Thirty-nine putative alleles of the SSR primer GNK02-02 segregates in the 77 guineagrass accessions where PD equals 0.968. The average number of the putative alleles with 13 SSR primers was 14.9, and the average PD value was 0.859. These values are higher than those reported in zoysia grass (Tsuruta et al. 2005) and perennial ryegrass (Jones et al. 2001). Using these developed SSR primers, a phenogram was comprised of a total of 190 putative alleles in the 77 guineagrass accessions as shown in Fig. 11.6. The phenogram contained six major groups of the guineagrass. Although two of the six major groups were not contained in any South African accessions, all of the

187

groups contained Kenyan accessions. These results are consistent with the hypothesis that guineagrass originates from East Africa, Kenya, and Tanzania. A principle coordinate analysis (PCA) scatter-plot indicated that the 77 guineagrass accessions also were divided into six major groups, as is shown in as shown in Fig. 11.7, hinting that the cultivars selected in different breeding sites belong to different major guineagrass groups. Whereas Japanese cultivars selected in warm temperate zone belong to the group I, the new cultivars of semitropical Okinawa belong to the group III. The importance of these SSR markers is that they should be represented in new accessions and will be key in establishing new breeding lines and cultivars of guineagrass. Indeed, new alleles of the foundation of 13 SSR markers are anticipated, thus facilitating measures of phylogenetic relatedness among newly isolated accessions.

11.11 Crop Improvement Important observations using guineagrass as a model plant have contributed to our understanding of cytology, quantitative/population genetics, and numerical taxonomy during the last 40 years (Combes and Pernes 1970; Combes 1975). Indeed, not only were the diploid and tetraploid sexually reproducing cultivars discovered but also the apomictic accessions (Combes and Pernes 1970; Hanna et al. 1973; Nakajima et al. 1979; Nakagawa and Hanna 1992). Hence, not only could the inheritance patterns of important agronomic genes be determined, but these in turn could also be readily introgressed in both diploid and polyploid lines. Similarly, effective crossing protocols have been designed to determine the number of segregating genes regulating apomixes in a significant number of genera (Nakajima 1991; Blakey et al. 1997, 2001; Ebina and Nakagawa 2001). Gametophytic apomixis results from asexual reproduction, and the resulting embryo is genetically identical to the mother from which it was derived. It can be classified into two forms, that is, aposporous and diplosporous apomixis (Nogler 1984). In addition to guineagrass, many other aposporous apomictic genera have been identified among them such tropical grasses as Pennisetum, Paspalum, and Brachiaria species (Asker and Jerling 1992; Sherwood 2001). The inheritance of

188

H.S. Bhandari et al. Genetic similarity

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0 GR 190 GR 293 GR 243 GR 373 214 GR 193 73-473 211 73-790 73-1006 73-804 GR 161 GR 248B 453 GR 374 GR 173 GR 138 hozon 35 185 K 59104 GR 134B GR 133B GR 131 GR 218 GR 233A GR 144 GR 434 GR 45 GR 377 72-192 GR 238 GR 367 GR 371 GR 470B 73-919B GR 458 GR 422-1 GR 464 GR 208 GR 376 GR 336 GR 440 K 6343 GR 206 GR 37 GR 124 GR 123 GR 329 GR 299 CP 1210-o-8 72-62 GR 50 Monbasa GR 239A GR 101A GR 334 GR 250 GR 120 GR 214 GR 215 GR 441 GR 432-1 GR 470A Natsuyutaka GR 452 K 6348 GR 186 Natsukaze GR 453 GR 456 M70-82 GR 467 Gatton GR 88 GR 365 M70-81 Meidai 9

Kenya Tanzania Kenya Ethiopia Kenya Kenya Kenya Kenya Kenya Kenya Kenya Unknown Kenya Okinawa Morocco Ethiopia Uganda Okinawa Kenya Kenya Uganda Uganda Uganda Unknown Kenya Kenya South Africa Unknown South Africa Kenya Kenya South Africa South Africa South Africa Kenya South Africa America Ghana Kenya Unknown South Africa Zimbabwe Kenya Tanzania ‘Zambezi River’ Kenya Kenya South Africa Tanzania Unknown Nigeria Unknown Unknown Kenya Nigeria South Africa Kenya Ethiopia Kenya Kenya Australia South Africa South Africa Unknown Unknown Kenya Kenya Cultivar Unknown Unknown South Africa Australia Australia Australia South Africa South Africa Unknown

I

II-i

II-ii II

II-iii II-iv II-v

III

IV

V

VI

Fig. 11.6 Phenogram of the 77 accessions of guineagrass based on the UPGMA clustering of 190 alleles obtained with 13 SSR markers (Ebina et al. 2007)

11 Panicum

189 PC2 cv. Zambezi River 64, K 8

14, 15, K

26, T

41, S

cv. Makueni 36, K 12, N

10

2.0

Group IV

75

33, K

40, S

cv. Tanzania

Group V 48, S 52, Z

1.0 47 27, K

–1.0

58, G

PC1

2, K

17,U 20,K 21 5, K 69, K 22, E

53 28, K

32, K 31, K

29, K

13, E 62

1.0

2.0

3.0

54 65,K

4, K 16, U 46, M

7, K 35, K

18,U

24, K 37, K 30

43,S

45,E 71 cv. Paikaji

19, U 70, K

68, K

3, K 25, K

76

23, K

–1.0

77

55 56

67,S 66, S

cv. Natsukaze

60, S

11,74 73

cv. Natsuyutaka

34, K

72 bl. Hozon35

6, K 44, S 50, S 51, S

9

1, N

38, T 49 39, S

57, S

Group III –2.0

61, S

42, S

cv. Gatton

Group I

Group VI

59

bl. Meidai9

63, K bl. Ryukyu-3

–2.0

Group II

Fig. 11.7 Principal component analysis (PCA) scatter-plot of the 77 accessions of guineagrass based on similarity matrix of Jaccard’s coefficient. PC: principle component. The doted circles indicate different groups showing as Fig. 11.1. The symbols

indicate group I: open circle, group II: open triangle, group III: open square, group IV: closed circle, group V: closed triangle, and group VI: closed square (Ebina et al. 2007)

the apomixis in guineagrass is simple and has been ascribed to a single gene (Hanna et al. 1973; Savidan 1980; Nakajima 1991; Ebina and Nakagawa 2001). In addition to guineagrass, apomixes in other grasses follow single gene transmission, thus raising the future possibility of introducing apomixis into new species following recombinant gene transfer. The development of highly efficient gene transfer protocol has been reported for P. virgatum following Agrobacterium tumefaciens-mediated transformation (Somileva 2006). The effectiveness of DNA conveyance has been proven to be a function of genotype and choice of explants. Given the simplicity of apomictic inheritance, attempts have been made to introduce this gene into compatible sexually reproducing lines by breeding. In pearl millet, sexual species of Pennisetum glaucum have been successfully introgressed with aposporous apomixis derived from P. squamulatum (Dujardin and Hanna 1983; Kaushal et al. 2007). The detailed molecular mapping of ASGR, in addition to studies on chromosome structure and the physical chromosome relationships among the agamic complex of Pennisetum and Cenchrus have been reported

(Ozias-Akins et al. 1998; Akiyama et al. 2004; Goel et al. 2006). In contrast, attempts to transfer apomixis from Tripsacum to maize have proven difficult. Some of these difficulties have been ascribed to the male or female sterility and to epigenetic patterning that impose constraints for apomictic seed development (Kindiger et al. 1996; Leblanc et al. 2008). These observations are further complicated by evidence suggesting that the regulation of apomixis in Gama grass is polygenic. Blakey et al. (1997, 2001) have shown that RFLPs linked to multiple chromosomal regions cosegregate with apomixis. Under this scenario, all of the regions encoding genes necessary for apomixis would have to be transferred simultaneously in order for this trait to be expressed. To date, this transfer is conspicuously absent. No apomictic maize has ever been recovered from breeding programs designed to produce maize-Tripsacum hybrids. Apomictics derived following mutation in sexual diploid Tripsacum population has also never been observed. The ecological significance of variability among apomictically reproducing species with respect to population size and the number of genes regulating

190

apomixis has been discussed by Adolfsson and Bengtsson (2007). In this connection, guineagrass has been found to express a large number of variants in nature. Seven percent of the accessions are sexual and it is these that create variability by generating segregating populations or alternatively as a function of facultative apomixis (Savidan and Pernes 1982). Facultative apomixis has inherent difficulties in guineagrass due to self-incompatibility. However, between 1.3% and 4.7% off-type plants arose from facultative apomixis (Warmke 1954). A similar pattern is observed in Brachiaria in which both sexual and apomictic plants have been found (Miles et al. 2004). Guineagrass along with Eriochloa and Urochloa belong to a group of closely related plants whose boundaries are poorly defined (Renvoize et al. 1996). Nevertheless, it is possible that the common ancestor plant to each of these related genera is likely one of the original aposporous apomixis plants among this large group of Gramineae. In addition to the evaluation of a large number of variable apomixis accessions, diploid and teteraploid sexual lines have also been employed in breeding programs designed to elucidate the mechanism of sexuality (Muir and Jank 2004). Due to inherent heterozyogosity associated with cross-pollination and polyploidy, recurrent sexual selection will be a new powerful instrument for any guineagrass breeding program (Vogel and Bruson 2004). Advanced molecular tools, for example, those widely available apomixis markers, should make it easy to construct and select highly pure recurrent sexual pools. While the region linked to the apomixis gene is extremely large, the AFLP markers obtained from “Natsukaze” have not been tested with other cultivars and populations (Akiyama et al. 2008). While the common region among the apomixis accessions remains available to discriminate asexual accessions, the cytological embryo sac clearing method remains useful, for the selection of recurrent sexual pool. This said, more simple diagnostic molecular methods of the pure sexual lines require enhancement. The prospect of accelerating recurrent sexual selection may be made through application of genotype matrix mapping (GMM) selection (Isobe et al. 2007). While autotetraploidy and high heterogeneity in guineagrass have been prevented in conventional breeding, GMM seeks quantitative trait loci (QTLs) and specifically QTL interactions in genotype pools. The detection of

H.S. Bhandari et al.

QTLs defining, for example, yield, abiotic tolerance, etc. will directly lead to the breeding of high performance sexual lines and the production of new, valueadded cultivars of the guineagrass. Recurrent sexual selection pools were obtained following hybridization between the doubled sexual line “Noh PL1” and apomixis cultivar “Natsuyutaka” (Sato et al. 1992; Kouki and Ebina 2007). In more than 100 F1 progeny, apomixis segregated for sex in an 1:1 ratio. Polyembryony was notably absent in the sexual progeny as determined by embryo sac analysis. At least three of the F2 plants of each open-pollinated F1 plants were planted and scored for agronomic traits in addition to whether they reproduced sexually or by apomixis. In F2 population, 41 superior plants were identified with respect to agronomic superiority and, of these, seven plants were sexual. Three plants were also found that are perennial. As expected, all F2 sexual plants exhibit monoembryony, with polyembryony restricted to apomicts. The selected sexual plants exhibit high yield and digestibility in comparison with several apomictic cultivars. For useful application of recurrent selection, additional agronomically superior sexual plants are necessary.

11.12 Genomics Switchgrass cultivars and “prairie-remnant” populations were analyzed with RAPD markers (Casler et al. 2007). The structural patterns and spatial variation suggested that there are different switchgrass gene pools in different regions of the northern and eastern US. Wider within-population variation was reported in comparison to those among-population variation. Thus, a relatively small number of collection sites are considered sufficient to maintain genetic variability within the gene pool. A collection of 31 unique switchgrass accessions from 168 available collections in GRIN were analyzed with expressed sequence tag–simple sequence repeat (EST-SSR) markers (Narasimhamoorthy et al. 2008). Greater within-population (79.6%) rather than among-population (20.4%) variation was reported, which appears to be a common feature in many outcrossing species. The results suggested that the recurrent selection within a population would serve as breeding strategy for switchgrass. The genotypes were also broadly grouped into different

11 Panicum

adaptive zones based on the geographical locations of collections. Recently, cDNA libraries have been constructed from callus, crown, and seedling tissues of P. virgatum, cultivar “Kanlow” (Tobias et al. 2008). The libraries were end-sequenced. An excess of 61,500 high quality ESTs were produced. The consensus sequences were identified and 73% could be aligned with Sorghum bicolor, the vast majority of these being identified with gene ontology. Given the scope of the library, it is likely that the information gleaned from scoring sequence diversity and similarity among wild Panicum species could provide insight on important biosynthetic functionalities as cell wall biosynthesis. Public EST database for guineagrass have recently become available on the DDBJ at accession DB999756–DB999951. These 196 ESTs are genes that were expressed in immature apomixis pistils. These are surveyed by microarray from 4,608 apomictic spikes expressed genes (Yamada-Akiyama et al. 2009). Candidates for the apomixis initiation genes were selected from the 196 ESTs. Quantitative PCR was performed with selected primers by using mRNA isolated from only a few immature pistils in different developmental stages Fig. 11.8. Due to the fact that some guineagrass lines are facultative apomicts, the small number of pistils screened is particularly sensitive with respect to the detection of genes involved in the expression of apomixis initiation. One novel gene homologous to a gene for immature panicle in Sorghum bicolor exhibits higher expression in apomictic immature pistils of guineagrass. Another gene homologous to the apomictic Cenchrus ciliaris pistil-specific gene also demonstrated higher expression in apomictic immature pistils of guineagrass.

191

11.13 Invasive Species Several species of Panicum such as P. capillaire, P. dichotomoflorum, P. repens, and P. virgatum are notoriously invasive. For example, P. repens is a perennial that forms large dense colonies. Concomitant with this configuration long, creeping rhizomes characterize this structure and, as a result, long floating mats are frequently observed. This precise organization can result in the failure of water to flow properly, making the control and management of plant spread difficult. Cutting-back is ineffective; only the application of herbicides can ameliorate the uncontrolled proliferation and growth. This said, it is precisely the formation of this habit that can be used to organize effective soil erosion strategies (www.heap.org./pier/ species/panicum_repens.html). The lesson here is that the use and focused application determine whether a plant is an invader or a friend.

11.14 Economic Advantages Given this, many of these species are also used as forage for livestock. Species such as P. coloratum and P. miliaceum have been cultivated to be used as fodder, while P. coloratum, P. repens, and P. virgatum are important constituents of native pasture (Watson and Dallwitz 1994). P. miliaceum, P. sumatrenes, and P. sonorum are important cultivated species for food for human consumption (Hitchcock 1971; Watson and Dallwitz 1994). In countries such as Asia, Europe, and the Middle East, dehulled grain of P. miliaceum (proso

Fig. 11.8 The ovaries from early and middle immature panicles of P. maximum. Each five panicles were subjected to the Q-PCR analysis. The scale bar indicates 1 mm

192

millet) can be directly cooked like rice or ground to make porridge. The grains are also used as birdfeed. In US, this crop is also used as forage. The species can thrive well under severe heat stress, poor soils, and drought. P. virgatum L. (switchgrass), is a native warm season grass of North American tallgrass prairies and has received widespread attention due to its potential for use as feedstock for cellulose-based bioethanol production (Sanderson et al. 1996; Vogel and Jung 2001; McLaughlin and Kszos 2005; Bouton 2007, 2008). Traditionally, this crop was used primarily as livestock forage, restoration of range land, or as a component of conservation reserve program (Bouton 2007). Earlier work on genetic modification of this crop was focused mainly on the improvement of its forage yield and quality (Vogel and Jung 2001). After this crop was chosen as a main herbaceous crop for production of bioethanol, the research has been shifted to improve the biomass production and quality for the conversion efficiency (Parrish and Fike 2005). Research is now underway on modification of biochemical pathways to impact the recalcitrance and enhance ethanol yield (Zeng-Yu Wang, pers. comm.). The discovery of sexual plants of P. maximum at Tifton led to the development of winter-hardy clones (Hanna et al. 1973; Hanna 1986). The progeny of the same sexual lines generated new cultivars. The offspring of these lines also generated excellent successions, and new cultivars were bred in several countries of South America and Japan (Nakagawa and Hanna 1992; Muir and Jank 2004). A French collection was studied in the Ivory Coast, and several accessions, such as “T58”, “K187,” and “C1,” were evaluated based on agronomic traits and subsequently distributed to some additional African countries (Mandret et al. 1989). This collection, including the sexual materials, was transferred to Brazil EMBRAPA, where a robust breeding program ensued. Three new cultivars were released, namely “Tanzania,” “Monbasa,” and “Massai” (Muir and Jank 2004). Another sexual diploid “GR 297” was discovered and first apomictic hybrid cultivar “Natsukaze” generated from crosses between it and an unidentified tetraploid apomictic line through open-pollination (Nakajima et al. 1979; Sato et al. 1993). Since “Natsukaze” is an apomictic tetraploid, this strain is thought to have originated from an unreduced egg of a diploid sexual plant or a hexaploid apomictic pollen donor. Japan’s Panicum breeding program has been newly

H.S. Bhandari et al.

introduced to the Japanese semitropical islands of Okinawa prefecture, and new hybrid cultivars continue to be released.

11.15 Recommendations for Future Resource Utilization Several major global problems still to be faced relate directly to the inequitable distribution of natural resources, pollution, and biodiversity [for review see, Mahmud et al. (2008) and Rai (2008)]. China has more than 20% of the world’s population but only seven percent of the planet’s arable land. This country, in addition to others in Asia, faces serious threats in their food chain as a consequence of heavy metal contamination of the region’s major food source, specifically rice. This problem is likely to become acerbated with time. India, with the world’s second largest area of tillable land, no longer exports rice but imports wheat, while Ganges River carries arsenic contaminants into Bangladesh. Not only are the land resources notoriously unbalanced with respect to utilization but also is potable water. In fact, 20% of the Earth’s fresh water is to be found in the US Great Lakes System. Complicating this picture further is the fact that, congruent with high selection pressure, plants native to these regions have evolved and reproduce successfully under conditions of abiotic challenges. Ferns are capable of taking up high concentrations of arsenic known as are rice lines that quite successfully package this poison in the grain, thus allowing contaminated cereals into the food chain. Then too, there are plants based on folklore, which are believed to have some pharmaceutical properties based on empirical observation. The putative treatment has been identified to a tissue or organ and anything from poultices to teas have been effectively used from one generation to the next. While the distribution of land and water cannot be re-equilibrated, it is a certainty that the value of native germplasm should at least, in portion, return to the country from which it derives. To address this potential source of exploitation, the United Nations Conference on Environment and Development (UNED) made available for signature The Convention on Diversity. The Biodiversity Treaty, as it is commonly known, has been signed by an excess of 170 countries.

11 Panicum

One of its primary purposes calls for the “fair and equitable sharing of the benefits arising out of the utilization of genetic resources.” In essence, it calls for the recognition and subsequent protection of property rights devolving from biological and genetic assets. Interestingly, a number of countries home to a huge biotechnology industry, including the US, have not signed this treaty. Thus, issues of equity and fairness arise prominently where countries lacking the infrastructure cannot even exploit its genetic resources. This has the potential of allowing well funded enterprises from engaging in the worse kind of gene exploration/abuse. For example, a gene that allows for arsenic packaging in the aerial parts of plant could be isolated, cloned by a foreign enterprise, introduced in rice, and sold back, for example, to Bangladesh, the very source of the plant material from which it is derived. It is to the correction of these manipulations that principled scientists should address themselves, speaking for those not in a position to speak for themselves.

References Adolfsson S, Bengtsson BO (2007) The spread of apomixis and its effect on resident genetic variation. J Evol Biol 20:1933–1940 Akiyama Y, Conner JA, Goel S, Morishige DT, Mullet JE, Hanna WW, Ozias-Akins P (2004) High-resolution physical mapping Pennisetum squamulatum reveals extensive chromosomal heteromorphism of the genomic region associated with apomixis. Plant Physiol 134:1733–1744 Akiyama YY, Yamada-Akiyama H, Yamanouchi H, Takahara M, Ebina M, Takamizo T, Sugita S-I, Nakagawa H (2008) Estimation of genome size and physical mapping of ribosomal DNA in diploid and tetraploid guineagrass (Panicum maximum Jacq.). Grassl Sci 54:89–97 Aliscioni SS, Giussani LM, Zuloaga FO, Kellogg EA (2003) A molecular phylogeny of Panicum (Poaceae: Paniceae): tests of monophyly and phylogenetic placement within the Panicoideae. Am J Bot 90:796–821 Asker SE, Jerling L (1992) Apomixis in plants. CRC Press, Boca Raton, FL Barnett FL, Carver RF (1967) Meiosis and pollen stainability in switchgrass, Panicum virgatum L. Crop Sci 7:301–304 Blakey CA, Dewald CL, Goldman SL (1997) Co-segregation of DNA markers with Tripsacum fertility. Maydica 42:363–369 Blakey CA, Goldman SL, Dewald CL (2001) Apomixis in Tripsacum – comparative mapping of amultigene phenomenon. Genome 44:222–230

193 Bouton JH (2007) Molecular breeding of switchgrass for use as a biofuel crop. Curr Opin Genet Dev 17:553–558 Bouton JH (2008) Improvement of switchgrass as a bioenergy crop. In: Vermerris W (ed) Genetic improvement of bioenergy crops. Springer, Heidelberg, Germany, pp 295–308 Bouton JH, Brown RH, Evans PT, Jernstedt JA (1986) Photosynthesis, leaf anatomy, and morphology of progeny from hybrids between C-3 and C-3/C-4 Panicum species. Plant Physiol 80:487–492 Brown RH, Morgan JA (1980) Photosynthesis of grass species differing in carbon di-oxide fixation pathways 6. differential effects of temperature and light intensity on photo respiration in 3 carbon 4 carbon pathway and intermediate species. Plant Physiol 66:541–544 Brown RH, Bouton JH, Evans PT, Malter HE, Rigsby LL (1985) Photosynthesis, morphology, leaf anatomy, and cytogenetics of hybrids between C-3 and C-3/C-4 Panicum species. Plant Physiol 77:653–658 Burton G (1942) A cytological study of some species in the tribe Paniceae. Am J Bot 29:355–359 Carman JG (1997) Asynchronous expression of duplicate genes in angiosperms may cause apomixis, bispory, tetraspory, and polyembryony. Biol J Linn Soc 61:51–94 Carneiro VTC, Dusi DMA, Orits JPA (2006) Apomixis: occurrence, applications and improvements. In: Teixeira da Silva JA (ed) Floriculture, ornamental and plant biotechnology. Global Science Books, Middlesex, pp 564–571 Casler MD, Stendal CA, Kapich L, Vogel KP (2007) Genetic diversity, plant adaptation regions, and gene pools for switchgrass. Crop Sci 47:2261–2273 Christin PA, Salamin N, Kellogg EA, Vicentini A, Besnard G (2009) Integrating phylogeny into studies of C4 variation in the grasses. Plant Physiol 149:82–87 Church GL (1940) Cytotaxonomic studies in the gramineae Spartina, Andropogon, and Panicum. Am J Bot 27:267–271 Clark LG, Zhang WP, Wendel JF (1995) A Phylogeny of the grass family (Poaceae) based on Ndhf – sequence data. Syst Bot 20:436–460 Clayton WD, Renvoize SA (1986) Panicea: genera Graminum. Grass of the World. Royal Botanic Gardens, Kew, London, pp 256–296 Clayton WD, Harman KT, Williamson H (2006) GrassBase – The Online World Grass Flora. http://www.kew.org/data/ grasses-db.html/. Accessed 1 Jun 2009 Combes D (1975) Polymorphisme et modes de reproduction dens la section des Maximae du genre Panicum (Graminee) en Afrique. Mem ORSTOM 77:99 Combes D, Pernes J (1970) Variations dens le nombres chromosomiques du Panicum maximum Jacq. en relation avec le mode de reproduction. C R Acad Sci D 270:782–785 Conner JA, Goel S, Gunawan G, Cordonnier-Pratt MM, Johnson VE, Liang C, Wang H, Pratt LH, Mullet JE, Debarry J, Yang L, Bennetzen JL, Klein PE, Ozias-Akins P (2008) Sequence analysis of bacterial artificial chromosome clones from the apospory-specific genomic region of Pennisetum and Cenchrus. Plant Physiol 147:1396–1411 Dengler NG, Dengler RE, Donnelly PM, Hattersley PW (1994) Quantitative leaf anatomy of C3 and C4 grasses (Poaceae) – bundle-sheath and mesophyll surface-area relationships. Ann Bot 73:241–255

194 Dujardin M, Hanna WW (1983) Apomictic and sexual pearl millet pennisetum americanum x pennisetum squamulatum hybrids. J Hered 74:277–279 Duvall MR, Saar DE, Grayburn WS, Holbrook GP (2003) Complex transitions between C-3 and C-4 photosynthesis during the evolution of paniceae: a phylogenetic case study emphasizing the position of Steinchisma hians (Poaceae), a C-3-C4 intermediate. Int J Plant Sci 164:949–958 Ebina M, Nakagawa H (2001) RAPD analysis of apomictic and sexual lines in guineagrass (Panicum maximum Jacq.). Jpn J Grassl Sci 47:251–255 Ebina M, Nakagawa H, Yamamoto T, Araya H, Tsuruta S, Takahara M, Nakajima K (2005) Co-segregation of AFLP and RAPD markers to apospory in guineagrass (Panicum maximum Jacq.). Grassl Sci 51:71–78 Ebina M, Kouki K, Tsuruta S, Akashi R, Yamamoto T, Takahara M, Inafuku M, Okumura K, Nakagawa NH (2007) Genetic relationship estimation in guineagrass (Panicum maximum Jacq.) assessed on the basis of simple sequence repeat markers. Grassl Sci 53:115–164 Evans LT (1993) Crop evolution, adaptation and yield. Cambridge University Press, Cambridge Giussani LM, Cota-Sanchez JH, Zuloaga FO, Kellogg EA (2001) A molecular phylogeny of the grass subfamily Panicoideae (Poaceae) shows multiple origins of C-4 photosynthesis. Am J Bot 88:1993–2012 Goel S, Chen Z, Akiyama Y, Conner JA, Basu M, Gualtieri G, Hanna WW, Ozias-Akins P (2006) Cmparative physical mapping of the apospory-specific genomic region in two apomictic grasses, Pennisetum squamulatum and Cenchrus ciliaris. Genetics 163:1069–1082 Gomez-Martinez R, Culham A (2000) Phylogeny of the subfamily Panicoideae with emphasis on the tribe Paniceae: evidence from the trnL-F cpDNA region. In: Jacobs S, Everett J (eds) Grasses, systematics and evolution. CSIRO, Collingwood, Victoria, pp 136–140 Guglieri A, Zuloaga FO, Longhi-Wagner HM (2004) Synopsis of Panicum subg. Panicum (Poaceae, Paniceae) in Brazil. Acta Bot Bras 18:359–367 Guglieri A, Longhi-Wagner HM, Zuloaga FO (2006) Panicum L. subg. Panicum sect. Panicum (Poaceae) in Brazil. Hoehnea 33:185–217 Gunter LE, Tuskan GA, Wullschleger SD (1996) Diversity among populations of switchgrass based on RAPD markers. Crop Sci 36:1017–1022 Hamaud MA, Haroun SA, Macleod RD, Richards AJ (1994) Cytological relationships of selected species of Panicum L. Biol Plant 36:37–45 Hanna WW (1986) Resistration of 12 winter-hardy Panicum maximum germplasm clones. Crop Sci 26:389–390 Hanna WW, Powell JB, Millot JC, Burton GW (1973) Cytoloy of obligate sexual plants in Panicum maximum Jacq. and their use in controlled hybrid. Crop Sci 13:695–697 Hiremath SC, Patil GNV, Salimath SS (1990) Genome homology and origin of Panicum sumatrense Gramineae. Cytologia (Tokyo) 55:315–320 Hitchcock AS (1971) Manual of the grasses of the United States. Dover Publication, New York Hopkins AA, Taliaferro CM, Murphy CD, Christian D (1996) Chromosome number and nuclear DNA content of several switchgrass populations. Crop Sci 36:1192–1195

H.S. Bhandari et al. Huang SX, Su XJ, Haselkorn R, Gornicki P (2003) Evolution of switchgrass (Panicum virgatum L.) based on sequences of the nuclear gene encoding plastid acetyl-CoA carboxylase. Plant Sci 164:43–49 Hultquist SJ, Vogel KP, Lee DJ, Arumuganathan K, Kaeppler S (1996) Chloroplast DNA and nuclear DNA content variations among cultivars of switchgrass, Panicum virgatum L. Crop Sci 36:1049–1052 Ibrahim DG, Burke T, Ripley BS, Osborne CP (2009) A molecular phylogeny of the genus Alloteropsis (Panicoideae, Poaceae) suggests an evolutionary reversion from C-4 to C-3 photosynthesis. Ann Bot 103:127–136 Isobe S, Nkaya A, Tabata S (2007) Genotype matrix mapping: searching for quantitative trait loci interactions in genetic variation in complex trait. DNA Res 14:217–225 Jones ES, Dupal MP, Kolliker R, Drayton MC, Foster JW (2001) Development and characterization of simple sequence repeats (SSR) markers for perennial ryegrass (Lolium perenne L.). Theor Appl Genet 102:405–415 Kaushal P, Roy AK, Khare A, Malaviya DR, Zadoo SN, Choubey RN (2007) Crossability and characterization of interspecific hybrids between sexual Pennisetum glaucum (pearl millet) and a new cytotype (2n ¼ 56) of apomictic P. squamulatum. Cytologia 72:111–118 Kindiger B, Bai D, Sokolov V (1996) Assignment of a gene(s) conferring apomixis in Tripsacum to achromosome arm: cytological and molecular evidence. Genome 39:1133–1141 Kouki K, Ebina M (2007) Selections of sexual lines of guineagrass from 1st and 2nd generation crossed between doubled sexual line and apomixis cultivar. Bull Okinawa Prefect Res Center Livestock Grassl 44:123–127 (in Japanese) Leblanc O, Grimanelli D, Hernandez-Rodriguez M, Galindo PA, Soriano-Martinez AM, Perotti E (2008) Seed development and inheritance studies in apomictic maize-Tripsacum hybrids reveal barriers for the transfer of apomixis into sexual crops. Int J Dev Biol 53:585–596 Leonardos ED, Grodzinski B (2000) Photosynthesis, immediate export and carbon partitioning in source leaves of C-3, C-3-C-4 intermediate, and C-4 Panicum and Flaveria species at ambient and elevated CO2 levels. Plant Cell Environ 23: 839–851 Lu K, Kaeppler SW, Vogel KP, Arumuganathan K, Lee DJ (1998) Nuclear DNA content and chromosome numbers in switchgrass. Great Plains Res 8:269–280 Lu (1995) Cytogenetic analyses in switchgrass (Panicum virgatum L.), Thesis, University of Nebraska, Lincoln Mahmud R, Inoue N, Kasajima S-Y, Shaheen R (2008) Assessment of potential indigenous plant species for the phytoremediation of arsenic-contaminated areas of Bangladesh. Int J Phytoremed 10:117–130 Mandret G, Roberge G, Ourry A (1989) Le Panicum maximum dens l’association maraichageelevage au Senegal. The French Grassl Soc, Versailles Cedex Martinez EJ, Hopp EH, Stein J, Ortiz JP, Quarin CL (2003) Genetic characterization of apospory in tetraploid Paspalum notatum based on the identification of linkage molecular markers. Mol Breed 12:319–3227 Mathews E (1983) Global vegetation and land use: new high resolution data bases for climate studies. J Clim Appl Methodol 22:474–487

11 Panicum McLaughlin SB, Kszos LA (2005) Development of switchgrass (Panicum virgatum) as a bioenergy feedstock in the United States. Biomass Bioenergy 28:515–535 McMillan C, Weller J (1959) Cytogeography of Panicum virgatum in central North America. Am J Bot 46:590–593 Miles JW, Valle CB, Rao IM, Euclides VPB (2004) Brachiariagrasses. In: Moser LE, Burson BL, Sollenberger LE (eds) Warm-season (C4) grasses. American Society of Agronomy, Madison, WI, pp 745–783 Missaoui AM, Paterson AH, Bouton JH (2006) Molecular markers for the classification of switchgrass (Panicum virgatum L.) germplasm and to assess genetic diversity in three synthetic switchgrass populations. Genet Resour Crop Evol 53:1291–1302 Moore KJ, Moser LE, Vogel KP, Waller SS, Johnson BE, Pedersen JF (1991) Describing and quantifying growthstages of perennial forage grasses. Agron J 83:1073–1077 Morrone O, Hunziker JH, Zuloaga FO, Escobar A (1995) Chromosome numbers in South American Paniceae (Poaceae: Panicoideae). Darwiniana (San Isidro) 33:53–60 Morrone O, Escobar A, Zuloaga FO (2006) Chromosome studies in American Panicoideae (Poaceae). Ann MO Bot Gard 93:647–657 Morrone O, Denham SS, Aliscioni SS, Zuloaga FO (2008) Parodiophyllochloa, a new genus segregated from Panicum (Paniceae, Poaceae) based on morphological and molecular data. Syst Bot 33:66–76 Muir JP, Jank L (2004) Guineagrass. In: Moser LE, Burson BL, Sollenberger LE (eds) Warm-season (C4) grass, vol 45. American Society Agronomy, Madison, WI, pp 589–621 Nakagawa H, Hanna WW (1990) Morphology, origin and cytogenetics of a 48-chromosome panicum maximum. Cytologia 55:471–474 Nakagawa H, Hanna WW (1992) Induced sexual tetraploids for breeding guineagrass (Panicum maximum Jacq.). J Jpn Soc Grassl Sci 38:152–158 Nakajima K (1991) Apomixis and its application to plant breeding, vol 29, Gamma field symposia. National Institute of Agrobiological Resources, Tsukuba, Ibaraki, pp 71–94 Nakajima K, Komatsu T, Mochizuki N, Suzuki S (1979) Isolation of diploid and tetraploid sexual plants in guineagrass (Panicum maximum Jacq.). Jpn J Breed 29:228–238 Narasimhamoorthy B, Saha MC, Swaller B, Bouton JH (2008) Genetic diversity in switchgrass collections assessed by EST-SSR markers. Bioenergy Res 1:136–146 Nielsen EL (1944) Analysis of variation in Panicum virgatum L. J Agric Res 69:327–353 Nogler GA (1984) Gametophytic apomixis. In: Johri BM (ed) Embriology of angiosperms. Springer, Berlin, pp 475–518 Ozias-Akins P, Roche D, Hanna WW (1998) Tight clustering and hemizygosity of apomixis-linked molecular markers in Pennisetum squamulatum implies genetic control of apospory by a divergent locus that may have no allelic form in sexual genotypes. Proc Natl Acad Sci USA 95:5127–5132 Parrish DJ, Fike JH (2005) The biology and agronomy of switchgrass for biofuels. Crit Rev Plant Sci 24:423–459 Rai PK (2008) Heavy metal pollution in aquatic ecosystems and its phytormediation using wetland plants: an ecosystem approach. Int J Phytoremed 10:133–160

195 Reinheimer R, Pozner R, Vegetti AC (2005) Inflorescence, spikelet, and floral development in Panicum maximum and Urochloa plantaginea (Poaceae). Am J Bot 92:565–575 Renvoize SA, Clayton WD, Kabuye CHS (1996) Morphology, taxonomy, and natural distribution of Brachiaria (Trin.) Griseb. In: Miles JW, Maass BL, Valle CBD (eds) Brachiaria: biology, agronomy, and improvement. CIAT and CNPGC/EMBRAPA, Cali, Colombia and Campo Grande, MS, Brazil, pp 1–15 Sanderson MA, Reed RL, McLaughlin SB, Wullschleger SD, Conger BV, Parrish DJ, Wolf DD, Taliaferro C, Hopkins AA, Ocumpaugh WR, Hussey MA, Read JC, Tischler CR (1996) Switchgrass as a sustainable bioenergy crop. Bioresour Technol 56:83–93 Sato H, Shimizu N, Nakagawa H, Natsuoka H (1992) A new registered cultivar ‘Natsuyutaka’ of guineagrass, Panicum maximum Jacq. Jpn Agric Res Q 25:259–266 Sato H, Shimizu N, Nakagawa H, Nakajima K, Ochi M (1993) A new registered cultivar ‘Natsukaze’ guineagrass (Panicum maximum Jacq.). Bull Kyushu Nat Agric Exp Stn 24:417–437 Savidan Y (1980) Chromosomal and embryological analyses in sexual  apomictic hybrids of Panicum maximum Jacq. Theor Appl Genet 58:153–156 Savidan YH, Pernes J (1982) Diploid-tetraploid-dihaploid cycles and the evolution of Panicum maximum Jacq. Evolution 36:596–600 Sherwood RT (2001) Genetic analysis of apomixis. In: Savidan Y, Carman JG, Dresselhaus T (eds) The flowering of apomixis: from mechanisms to genetic engineering. CIMMYT, Mexico, pp 64–82 Simon BK, Jacobs SWL (2003) Megathyrsus, a new generic name for Panicum subgenus Megathyrsus. Austrobaileya 6:571–574 Somileva M (2006) Switchgrass (Panicum virgatum). Agrobacterim protocals. Methods Mol Biol 2:65–74 Stein J, Quarin CL, Martinez EJ, Pessino SC, Orits JPA (2004) Tetraploid races of Paspalum notatum show polysomic inheritance and preferential chromosome pairing around the apospory-controlling locus. Theor Appl Genet 109:186–191 Tobias CM, Sarath G, Twigg P, Lindquist E, Pangilinan J, Penning BW, Barry K, McCann MC, Carpita NC, Lazo GR (2008) Comparatice genomics in switchgrass using 61, 585 high-quality expressed sequence tags. Plant Genome 1:111–142 Tsuruta S, Hashiguchi M, Ebina M (2005) Development and characterization of simple sequence repeat markers in Zoysia japonica Steud. Grassl Sci 51:249–257 Tsvelev NN, Zhukova PG (1974) The minimal basic chromosome number in the family Poaceae. Bot Z (St Petersburg) 59:265–269 Vogel K, Bruson B (2004) Breeding and genetics. In: Moser LE, Burson BL, Sollenberger LE (eds) Warm-season (C4) grasses. American Society Agronomy, Madison, WI, pp 51–94 Vogel KP, Jung HJG (2001) Genetic modification of herbaceous plants for feed and fuel. Crit Rev Plant Sci 20:15–49 Warmke HE (1954) Apomixis in Panicum maximum. Am J Bot 41:5–11

196 Warner DA, Ku MSB, Edwards GE (1987) Photosynthesis, leaf anatomy, and cellular constituents in the polyploid C-4 grass Panicum-virgatum. Plant Physiol 84:461–466 Watson L, Dallwitz MJ (1994) The grass genera of the World, 2nd edn. CAB International, Wallingford, 1081 Yamada-Akiyama H, Akiyama Y, Ebina M, Xu Q, Tsuruta S, Yazaki J, Kishimoto N, Kikuchi S, Takahara M, Takamizo T, Sugita S, Nakagawa H (2009) Analysis of expressed sequence tags in apomictic guineagrass (Panicum maximum). J Plant Physiol 166:750–761

H.S. Bhandari et al. Zuloaga F, Morrone O, Dubcovsky J (1989) Exomorphological, anatomical, and cytological studies in Panicum validum (Poaceae, Panicoideae, Paniceae) – its systematic position within the genus. Syst Bot 14:220–230 Zuloaga FO, Morrone O, Giussani LM (2000) A cladistic analysis of the Paniceae: a preliminary approach. In: Jacobs SWL, Everett J (eds) Grasses: systematics and evolution. CSIRO, Collingwood, pp 123–135 Zuloaga FO, Giussani LM, Morrone O (2006) On the taxonomic position of Panicum aristellum (Poaceae: Panicoideae: Paniceae). Syst Bot 31:497–505

Chapter 12

Paspalum W.M. Williams, M.L. Williamson, and D. Real

12.1 Introduction Paspalum L. (tribe Paniceae R. Br.) is one of the largest C4 grass genera with approximately 400 species that are native to predominantly tropical and subtropical regions (Clayton and Renvoize 1986; Watson and Dallwitz 1992), with the center of diversity in South America (Moraes Fernandes et al. 1968). The C4 decarboxylation is catalyzed by NADP malic enzyme (NADP-ME), the most common type in the Paniceae. There are a few semi-domesticated species. P. dilatatum Poir. (dallisgrass), P. pauciciliatum (Parodi) Herter (prostrate dallisgrass), P. urvillei Steud. (Vasey grass, Venuto et al. 2003), P. atratum Swallen (Atra paspalum, Kretschmer et al. 1994), P. glaucescens Hackel, P. guenoarum Arechav. (pasto rojo), P. nicorae Parodi (Brunswick grass), and P. plicatulum Michx. (brownseed paspalum) have been commercialized for forage production, while P. notatum Fl€ ugge (bahiagrass ) is used for both forage and turf (Bennett 1962). P. vaginatum Swartz (seashore paspalum) is used for dune stabilization and turf on saline soils while P. distichum L. (knotgrass) is used in wet places as a feed for water birds. P. scrobiculatum L. (kodo millet) is used in Asia as a cereal grain. Many other species are used locally in native grasslands for forage and feed (Allem and Valls 1987; Filgueiras 1992). The genus Thrasya Kunth (20 species) is very close to Paspalum in

W.M. Williams (*) AgResearch, Grasslands Research Centre, Private Bag 11008, Palmerston North, New Zealand e-mail: [email protected]

chloroplast DNA sequence (Duvall et al. 2001) and morphology (Clayton and Renvoize 1986) and subsequently was absorbed into Paspalum as subgenus Harpostachys (Denham 2005). To date, there have been very few plant breeding programs. There is considerable interest in evaluating Paspalum species in Brazil because there are about 220 species native to that country (Valls and Pozzobon 1987). Workers are researching on a range of native material with a view toward variety development and interspecific hybridization (Pagliarini et al. 2001). Several germplasm collecting trips have been undertaken in the last 30 years, resulting in a collection of over 1,500 accessions at CENARGEN/EMBRAPA by the early 1990s (Takayama et al. 1998; Adamowski et al. 2005). Part of this collection was duplicated at CPPSE/EMBRAPA and at IBONE at Corrientes, Argentina. The predominant chromosome number is tetraploid (2n ¼ 4x ¼ 40), but diploids are also common (Pozzobon et al. 2008). Almost all species have a basic chromosome number of x ¼ 10. However, as new collections are analyzed, the number of exceptions has progressively increased (Penaloza et al. 2008). Those known to date are listed in Table 12.1. It has been suggested that, based on secondary chromosome pairing observations in P. notatum, the basic number of x ¼ 10 is derived from a lower ancestral number (x ¼ 5 or 6, Forbes and Burton 1961). Many species have multiple cytotypes (Table 12.2), with the diploids reproducing sexually and the polyploids apomictically (Quarin and Norrmann 1990). Examples include P. dilatatum (2n ¼ 40, 50, 60; Burson 1991b) and P. notatum, P. thunbergii Kunth ex. Steud., and P. plicatulum (2n ¼ 20, 40; Darlington and Wylie 1961). The mean diploid 2C DNA value is

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_12, # Springer-Verlag Berlin Heidelberg 2011

197

198

W.M. Williams et al.

Table 12.1 Paspalum species with basic chromosome numbers different from x ¼ 10 Species Group Chromosomes P. almum Alma 2n ¼ 12, 24 P. filgueirasii 2n ¼ 24 P. convexum Plicatula 2n ¼ 32 P. lanciflorum Ceresia n¼9 P. stellatum Ceresia 2n ¼ 52 P. reduncum Gardneriana 2n ¼ 18 P. ammodes Eriantha 2n ¼ 36 P. trachycoleon Ceresia 2n ¼ 36 P. burmanii Ceresia 2n ¼ 48

1.6 pg and the species variation is approximately threefold as determined by laser flow cytometry (Watson and Dallwitz 1992). Because many of the apomictic species require fertile pollen for apomixis to succeed (pseudogamy), they are able to be used as male parents in hybridization. Consequently, hybridization programs have involved crosses between diploids and crosses of polyploids with diploids. Before more targeted breeding programs can commence, a much better understanding of species phylogenies is needed. Because of apomixis, it has been possible for species hybrids and even intergeneric hybrids to become established (Quarin 1992).

12.2 Taxonomy The systematic treatment of the genus is complex with a large amount of interspecific hybridization and polyploidy, and many taxonomic groupings are still to be fully resolved. Indeed, the taxonomy is complicated by the presence of non-sexual polyploid complexes with sexual diploid and partially sexual tetraploid populations contributing new genetic diversity via hybridization. Under these circumstances, the concept of “species” is difficult because continued interspecific hybridization, with the hybrids fixed by apomixis, creates continuous intermixing and intergrading of taxa, resulting in agamic complexes (Pozzobon et al. 2008). Various authors have used external morphological characteristics to subdivide the genus into subgenera or sections and informal species groups (e.g., Chase 1929). However, these classifications have been repeatedly revised by various researchers and no single classification has been reached and the informal

References Quarin and Hanna (1980) Penaloza et al. (2008) Selva (1976), Reeder (1984) Davidse and Pohl (1974) Honfi et al. (1990) Penaloza et al. (2008) Penaloza et al. (2008) Penaloza et al. (2008) Penaloza et al. (2008)

groups are not always phylogenetically informative (Souza-Chies et al. 2006). Based on morphological classification, Paspalum species have been classified into four subgenera Anachyris (6 species), Ceresia (approximately 25 species), Harpostachys (39 species), and Paspalum (more than 250 species; Chase 1929; Zuloaga and Morrone 2005; Aliscioni and Denham 2008). Subgenus Paspalum is, in turn, divided into informal groups, e.g., Notata with 21 species (Zuloaga et al. 2004; Table 12.2). Phylogenetic relationships were initially analyzed using chromosome pairing in interspecific hybrids. This led to the identification of several distinct genomes that were annotated (Burson et al. 1991). The occurrence of polyploidy and apomixis makes chromosome analysis of phylogeny very difficult and also hinders conventional breeding (Jarret et al. 1998). Very few species have undergone meiotic analyses. In addition to chromosome pairing abnormalities, Pagliarini et al. (2001) observed non-oriented bivalents at metaphase I of meiosis and early chiasma terminalization, resulting in univalents and micronuclei at anaphase I. The latter abnormalities occurred even when chromosome pairing appeared to be completely normal. To date, sexual forms that have been studied have normal meiosis. The meiotic abnormalities are therefore largely a characteristic of apomictic plants (Burson and Bennett 1971a; Pagliarini et al. 2001).

12.3 Sexual Species of Paspalum Because of the high frequency of apomictic species and biotypes in Paspalum, the sexual species (diploids and facultative apomicts) represent an important genetic resource as generators of genetic diversity for

12 Paspalum

199

Table 12.2 Approximate species numbers in subgenera, sections, and informal groups of the genus Paspalum Subgenera Anachyris Ceresia

Section or informal group Ceresia Pectinata

Harpostachys Paspalum Dilatata

Ungrouped a

Number of speciesa 6 19 8 39 4

Livida Paniculata Quadrifaria Virgata Notata Alma/Notata Ovalia Alterniflora Barbinodia Bertoniana Bifida Bonplandiana Brevia Caespitosa Conjugata Corcovadensia Decumbentes Dimorphostachys Dissecta Disticha

13 8 10 10 21 1 1 4 2 2 1 6 2 17 1 8 4 1 7 4

Eriantha Falcata Fasciculata Filiformia Fimbriata Floridana Gardneriana Inaequivalvia Lachnea Laevia Macrophylla Maculosa Orbiculata Palustrine Parviflora Plicatula

6 1 1 5 3 2 9 2 2 6 2 1 6 2 8 21

Racemosa Recta Reimaria Rupestria Saccharoidea Setacea

10 4 1 3 4 3 34

Cultivated species

P. dilatatum, P. urvillei, P. pauciciliatum

P. notatum

plant breeding. To date, little more than one quarter of the genus has been characterized cytologically. Some of the diploid sexual species and biotypes are summarized in Table 12.3. It is clear that sexual taxa are numerous and occur in almost all subgenera and species groups, so that potentially useful genetic resources are available throughout the genus. There have been a few molecular analyses of phylogeny of the Paspalum species. Jarret et al. (1998) used restriction fragment length polymorphisms (RFLPs) to analyze 51 accessions of 29 species mainly from South America but including one from Japan (P. thunbergii Steud.) and P. scrobiculatum from India. Species relationships generally supported those obtained from morphological and cytological considerations.

12.4 Paspalum as a Model for Apomixis

P. distichum, P. vaginatum

P. atratum, P. guenoarum, P. nicorae, P. plicatulum

Chase (1939), Renvoize (1972), Cialdella et al. (1995), Morrone et al. (1995, 1996, 2000, 2004), Denham et al. (2002), Zuloaga et al. (2004), Denham (2005), Zuloaga and Morrone (2005), Aliscioni and Denham (2008), Hojsgaard et al. (2008)

Paspalum is potentially a very useful genus to research the genetics of apomixis because of the existence of sexual diploids, apomictic polyploids, and facultative apomictic tetraploids. The genetics of apomixis in tetraploid P. notatum was investigated by Martinez et al. (2001) by pollinating an emasculated fully sexual mother plant with apomictic plants and verifying hybrids by RFLP analysis. F2 and backcross families were produced from both sexual and apomictic F1 plants. The F1 produced 157 sexual plants and 56 plants that were apomictic (many of these also showed meiotic embryo sacs), a ratio close to 3:1. Sexual F1  apomictic F1 and backcrosses of sexual  apomictic plants also produced approximate 3:1 ratios. The results were interpreted to be consistent with determination of apospory by tetrasomic inheritance of a dominant allele at one locus, as is likely in other tropical grasses. Failure to obtain a near 1:1 ratio in the F1 was attributed to segregation distortion against apospory, possibly due to a pleiotropic lethal effect of an allele with incomplete penetrance or to closely linked genes with partial lethality. An earlier analysis (Burton and Forbes Jr 1960; Forbes and Burton 1961), concluding that apospory was controlled by a few recessive genes, was rejected on technical grounds (Martinez et al. 2001). Subsequently, a number of molecular markers completely linked to the apospory locus have been found and analyzed (Martinez et al. 2003;

200 Table 12.3 List of some species with diploid sexual biotypes. As indicated, many of these also have polyploid (apomictic) biotypes Subgenus or Species Chromosome Group numbera Subgenus P. malachophyllum 20 (most 40, Anachyris some 60) P. procurrens 20 (also 40) P. simplex 20 (also 40, 60) Subgenus Ceresia P. carinatum 20 (also 80) P. cordatum 20 P. humboldtianum 20 P. lanciflorum 18 (P. contractum) P. malmeanum 20 P. pectinatum 20 (also 40, 60) Subgenus P. decumbens 20 Harpostachys P. cinerascens 20 Subgenus Paspalum Notata P. notatum var. saurae 20 P. approximatum 20 P. barretoi 20 P. cromyorhizon 20 (also 40) P. nummularium 20 P. minus 20 (also 40, 50) P. pumilum 20 Notata/ P. maculosum 20 (also 40) Maculosa Notata/Alma P. almum 12 (also 24) Linearia/Notata P. crispulum 20 P. dedeccae 20 (also 40) P. ellipticum 20 (also 80) P. filifolium 20 P. lineare 20 (also 40, 60) P. pallens 20 Fasciculata/ P. equitans 20 Notata Plicatula P. glaucescens 20 (also 40, 60) P. limbatum 20 (also 40) P. guenoarum 20 (also 40) P. hydrophilum 20 (also 30, 40) P. chaseanum 20 P. compressifolium 20 (also 40, 60) P. plicatulum 20 (also 40) P. wrightii 20 (also 30, 40) Modesta/ P. modestum 20 Plicatula Caespitosa/ P. arenarium 20 Setacea Setacea P. setaceum 20 Falcata/ P. falcatum 20 Lachnea Caespitosa P. indecorum 20 P. chacoense 20 Paniculata P. juergensii 20 (continued)

W.M. Williams et al. Table 12.3 (continued) Subgenus or Species Group P. paniculatum P. palustre P. umbrosum Recta P. rectum P. vescum Parviflora P. scalare P. clavuliferum P. multicaule Disticha P. vaginatum Corcovadensia P. corcovadense Quadrifaria P. quarinii P. haumanii P. brunneum P. intermedium P. quadrifarium Bertoniana P. bertonii Racemosa P. candidum P. prostratum Conjugata P. conjugatum Fasciculata P. fasciculatum Orbiculata P. orbiculatum Livida P. plenum P. trichophyllum Dissecta P. repens Virgata P. rufum Gardneriana P. reduncum

Chromosome numbera 20 (also 40) 20 20 20 20 20 20 20 (also 40) 20 (also 40, 60) 20 20 (also 40) 20 20 20 (also 40) 20 (also 30, 40) 20 20 (also 40, 60) 20, (also 40, 60) 20 (also 40, 80) 20 20 20 (also 40) 20 20 20 (also 40) 18

a

Main sources were Zuloaga and Morrone (2005) and Pozzobon et al. (2008)

Stein et al. 2004), suggesting that the locus is in a region of suppressed recombination. The latter study obtained evidence that the apospory locus was associated with a chromosome region that showed disomic inheritance (while the rest of the genome was largely tetrasomic). The possibility of a chromosome rearrangement (e.g., an inversion) being associated with the segregation distortion as well as the suppressed recombination was discussed. Repression of recombination in the chromosome region controlling apospory, parthenogenesis, and pseudogamy was also observed in studies of P. simplex Morong (subgenus Anachyris) by Pupilli et al. (1997, 2001) and Caceres et al. (2001), and the region showed strong colinearity with a rice chromosome segment. P. notatum and P. simplex were used by Martinez et al. (2007) to investigate the question as to why apospory is associated with polyploidy. This work

12 Paspalum

showed that apomixis was not transmitted by monoploid gametes, explaining why hybridization does not produce apomictic diploids. Gametes higher than monoploid transmitted apospory but with apparently strong segregation distortion. Despite this, a low number of extra aposporic embryo sacs (always associated with normal meiotic embryo sacs) were found in several diploid species by Norrmann et al. (1989). The species were P. intermedium Monro ex. Morong & Britton P. quadrifarium Lamk., P. haumanii Parodi, P. brunneum Mez, and P. rufum Nees ex. Steud. These observations, along with observations of apospory in P. cromyorhizon Trin. ex. Doell (Quarin 1986) and P. equitans Mez (Quarin and Norrmann 1987) and the finding of facultative apospory in colchicine-doubled P. hexastachyum Parodi (Quarin and Hanna 1980), indicate that the potential for apomixis exists in some of the sexual diploid species. Apparently, the expression of this potential can vary with environmental conditions or season (Quarin 1986; Norrmann et al. 1989). The molecular genetic basis of apomixis in P. notatum was investigated by Pessino et al. (2001). A family segregating for sexual and apomictic plants was generated by pollinating a sexual tetraploid with an apomictic tetraploid. Spikelet mRNA was extracted from three sexual and three apomictic progenies, and the two lots were compared in differential display experiments. Three transcripts that were expressed at higher levels in the apomictic plants were cloned and proved to show high homology. However, they showed no homology with genes of known function, although a probable homology was detected with a KSP multiphosphorylation domain that is a phosphorylation site for cdc-2-kinases that are known to be associated with cell division. A genetic linkage map of diploid P. notatum var. saurae (Pensacola bahiagrass) was developed by Ortiz et al. (2001) using a mapping population from a cross between parents from Cayasta, Argentina, and Tifton, Georgia. RFLP clones from maize, rice, and oat were used for the framework, while random amplified polymorphic DNA (RAPD) and amplified fragment length polymorphism (AFLP) markers provided refinement. Ten linkage groups were established, with 112 markers averaging about 9 cM distance. Several RFLP markers associated with apomixis in Brachiaria and maize–Tripsacum hybrids were located on the map, opening the way for comparative genomics of apomixis and many other traits.

201

Because almost all apomictic species are polyploid (predominantly tetraploid), it has been postulated that apomixis evolved as a mechanism by which sexually infertile wide hybrids have been rescued and restored to fertility (Darlington 1939). There is evidence that 5x P. dilatatum arose from interspecific hybridization of sexual species (Burson 1989). However, many apomictic grasses appear to be autopolyploids (Quarin 1992). In Paspalum, there are many species that have sexual diploid and apomictic tetraploid biotypes that are morphologically identical, suggesting that autotetraploidy is perhaps more likely than allotetraploidy. An experiment was carried out by Quarin et al. (1998) in which resynthesis of the tetraploid form of P. rufum (Virgata group) from the diploid form strongly indicated an autopolyploid origin of the wild biotype. It is apparent that apomixis, while predominantly related to polyploidy, is not confined to allo- or auto-polyploids, and that some explanation other than rescue of hybrids is needed for the preponderance of polyploidy in apomictic populations. Although apospory is by far the main type of apomixis in Paspalum, diplospory has occasionally been found. P. minus E. Fourn., a widely spread species with 2x, 4x, and 5x races, has been shown to have both types of apomixis in the 5x form (Bonilla and Quarin 1997).

12.5 Subgenus Anachyris Chase Subgenus Anachyris consists of six species (Morrone et al. 2000) distributed from Mexico to Argentina and Uruguay, characterized by boat-shaped spikelets. Three of these species (P. simplex, P. procurrens Quarin, and P. malachophyllum Trin.) have sexual diploid and facultative apomictic polyploid cytotypes (Table 12.3), while two (P. usterii Hack. and P. volcanensis Zuloaga, Morrone & Denham) are known only as tetraploids (Hunziker et al. 1998; Hojsgaard et al. 2008). P. simplex has been used in studies of the genetics of apomixis, as discussed above. Tetraploid forms of P. malachophyllum, P. procurrens, and P. usterii show chromosome pairing behavior that is consistent with autotetraploidy (Hojsgaard et al. 2008). Tetraploid interspecific hybrids between a colchicine-doubled plant of P. simplex  P. procurrens and  P. usterii and in triploid and diploid hybrids

202

between P. simplex and P. malachophyllum indicate that these species share the same homologous genomes. Hojsgaard et al. (2008) concluded that the entire subgenus is probably an agamic complex, sharing a common genome and speciating at the diploid level before polyploidization, apomixis, and geographic separation, led to the diversity apparent among the “species” now found.

12.6 Subgenus Paspalum 12.6.1 Informal Group Dilatata Chase This group, which includes the agronomically important P. dilatatum Poir. (dallisgrass), is usually considered to be closely related to the Livida, Paniculata, Quadrifaria, and the Virgata groups, which, along with Dilatata, comprise the “Dilatata complex.” Because the Dilatata are hybrid in origin, tracing their origin has been problematic, although studies of genome relationships and preliminary phylogenetic studies are predominantly in agreement (Souza-Chies et al. 2006). Four species make up the Dilatata group: P. dilatatum, P. pauciciliatum, P. urvillei, and P. dasypleurum Kunze ex. Desv. These species show similarities at both the morphological and molecular levels (Chase 1929; Jarret et al. 1998; Vaio et al. 2005, 2007; Aliscioni and Denham 2008). The Dilatata group has sexual tetraploids and apomictic pentaploid and hexaploid cytotypes. The tetraploid forms are hypothesized to be hybrids between P. intermedium (contributing the “I” genome) and P. juergensii Hackel (contributing the “J” genome) from groups Quadrifaria and Paniculata, respectively (Burson et al. 1973; Burson 1983; Espinoza and Quarin 2000; Speranza 2005). Burson et al. (1973) hybridized P. dilatatum ssp. flavescens (2n ¼ 40) with P. intermedium (2n ¼ 20), P. juergensii (2n ¼ 20), P. notatum (2n ¼ 20, Notata), and P. vaginatum (2n ¼ 20, Disticha) to determine their genomic relationships. P. dilatatum ssp. flavescens  P. intermedium and P. dilatatum ssp. flavescens  P. juergensii hybrids were triploid (2n ¼ 30) and showed irregular chromosome pairing at meiosis with ten univalents and ten bivalents, indicating that P. dilatatum shares a common genome with P. intermedium and with P. juergensii. Three hybrids between P. juergensii (2n ¼ 20) and P. intermedium (2n ¼ 20) were gener-

W.M. Williams et al.

ated to determine the genomic relationships between these two species (Burson 1981a). The hybrids were diploid with irregular chromosome behavior, typically having only two bivalents and 16 univalents at meiosis. This result indicated that the P. juergensii and P. intermedium genomes were largely unrelated and that each genome probably paired with different P. dilatatum ssp. flavescens genomes. Hybrids were male and female sterile. The sexual members of the group are P. urvillei, P. dasypleurum, P. dilatatum ssp. flavescens, and biotypes P. dilatatum “Virasoro” and P. dilatatum “Vacaria,” all sharing the “I” and “J” genomes. The pentaploid and hexaploid cytotypes are P. dilatatum “Common” and biotypes P. dilatatum “Chiru” and “Uruguaiana,” which contain additional genome(s) of uncertain origin, designated X. The genomic constitutions of these closely related groups are given by Speranza (2005) and are summarized in Table 12.4. The close relationship of the Livida group with Dilatata is unclear and needs more attention. P. denticulatum Trin. (syn. P. lividum Trin. ex. Schltdl.) has been Table 12.4 Ploidy level and genomic formula for Dilatata, Quadrifaria, Virgata, and Paniculata groups Species or biotype 2n Genomic Formula Dilatata Group P. dilatatum ssp. dilatatum 50 IIJJX P. dilatatum ssp. flavescens 40 IIJJ P. dilatatum “Virasoro” 40 IIJJ P. dilatatum “Vacaria” 40 IIJJ P. dilatatum “Chiru” 60 IIJJXX P. dilatatum “Uruguaiana” 60 IIJJX1X2 P. urvillei 40 IIJJ P. dasypleurum 40 IIJJ Quadrifaria Group P. intermedium P. haumanii P. brunneum P. quadrifarium P. densum P. durifolium

20 20 20 20 20 60

II II II II II IIJ2J2XaXa

Virgata Group P. conspersum P. virgatum P. rufum

40 40 20

I2I2JJ IIJ2J2 II

Paniculata Group P. juergensii P. paniculatum

20 20

JJ JJ

a

Unknown genome not related to other “X” genomes. Table extracted from Speranza (2005)

12 Paspalum

hypothesized to be the source of a genome in 4x P. pauciciliatum (Valls 2000). P. dilatatum and P. lividum showed close phylogenetic similarities in DNA (ITS and chloroplast) sequences (Souza-Chies et al. 2006). In contrast, RFLP data showed a marked separation between P. lividum accessions and the Dilatata group (Jarret et al. 1998), although P. pubiflorum Rupr. ex. Fourn. (Livida group) showed a stronger affinity with P. dilatatum and its relatives than with P. lividum. Paspalum dilatatum ssp. dilatatum (5x) seems to have originated from a natural hybridization between the sexual tetraploid P. dilatatum ssp. flavescens, and the apomictic hexaploid P. dilatatum “Uruguaiana” (Machado et al. 2005). As the original carrier of the “X” genome, the pentaploid also seems to have been involved in the origin of the other apomictic species in these groups (Speranza 2005). Chromosomes of the pentaploid pair as 20 bivalents and 10 univalents at meiosis (Bashaw and Holt 1958a; Burson 1983). Using RAPDs, Miz and Souza-Chies (2006) performed a phylogenetic analysis of the species of the Dilatata group, including seven P. dilatatum biotypes. Also included were the possible donors of the “I” and “J” genomes (P. intermedium, P. juergensii, and P. paniculatum L.). Apart from P. dilatatum “Virasoro” (2n ¼ 40), the taxa were not separable into single clades but formed a group consistent with the existence of a large agamic complex. This highlighted the difficulty of using the biological species concept for a genus in which the combination of sexual and apomictic reproduction can produce large, continuously variable, and intermingled populations in which enough genomic mixing occurs to blur morphological, cytogenetic, and molecular boundaries between populations. Microsatellite (SSR) markers may provide a tool to enable the further resolution of this problem. To this end, Speranza and Malosetti (2007) have developed a highly variable set of nuclear and chloroplast SSR markers that may serve this purpose.

12.6.1.1 P. dilatatum Poir.: Dallisgrass There are several P. dilatatum biotypes, but desirable forage biotypes are obligate apomicts and have low fertility, while sexual biotypes are remarkably uniform with high fertility but have poor forage qualities

203

(Bashaw and Forbes 1958; Bashaw and Holt 1958a; Venuto et al. 2003; Vaio et al. 2007). Garcia et al. (2007) investigated the morphological and molecular (RAPDs) phylogeny of 4x and 5x P. dilatatum accessions. Molecular phylogeny could, to a certain extent, be used to identify germplasm from different origins but could not distinguish between tetraploid and pentaploid accessions. Morphology could distinguish between most tetraploid and pentaploid accessions but could not distinguish between germplasm from different origins. It is possible that different genomes were involved in the origins of the different populations. “Common” dallisgrass: The “Common” biotype (P. dilatatum ssp. dilatatum) is a purple-anthered pentaploid (2n ¼ 50, IIJJX, 2C ¼ 2.96 pg) native to the Buenos Aires province of Argentina, Uruguay and S Brazil. It reproduces apomictically (apospory and parthenogenesis), but the apomictic mode of reproduction is a barrier to genetic improvement (Bashaw and Holt 1958a; Caponio and Quarin 1990; Vaio et al. 2007). It is highly valued as permanent pasture in southeastern USA due to its adaptability to different soil types, ability to grow well with other grasses and legumes, long growing season (earlier in spring and later in autumn than other warm season perennial grasses), good nutrition during all phases of growth, and persistence under grazing and cutting (Venuto et al. 2003). However, its commercial success has been limited by establishment problems due to poor seed quality (less than 50%), slow germination, and difficulties in seed production due to poor seed set and ergot (Claviceps paspali) infection (Venuto et al. 2003). Burson (1991a, b) suggested that the “Common” biotype originated from hybridization between tetraploid P. dilatatum ssp. flavescens and hexaploid P. dilatatum “Chiru.” Despite the apomictic nature of common dallisgrass, genetic improvement has been possible. Intraspecific hybridization was carried out between “Common” (2n ¼ 50) and yellow-anthered P. dilatatum ssp. flavescens (2n ¼ 40; Bennett et al. 1969). When “Common” was used as female, heptaploid (2n ¼ 70) hybrids with 10 univalents and 20 trivalents resulted, indicating that an unreduced female gamete had been fertilized. Hybrids had reduced fertility compared to the female parent. However, when the yellowanthered biotype was used as a female, the resultant plants had 45 chromosomes, which associated as 20

204

bivalents and 5 univalents at meiosis. Subsequent generations (F2, F3, and F4) had only 40 chromosomes, which associated as 20 bivalents. The F2 plants were sexual and showed variety for vigor, growth habit, and fertility. Bennett et al. (1969) were able to select for highly fertile (up to 90.8% seed set) and decumbent biotypes, thereby significantly improving the agricultural qualities of tetraploid P. dilatatum. This was the only study identified, which resulted in combined high fertility from the tetraploid and valuable agronomic traits from the pentaploid. P. dilatatum ssp. flavescens: P. dilatatum ssp. flavescens Roseng., B.R. Arril. & Izag. is often referred to as yellow-anthered dallisgrass, although two distinct yellow-anthered allotetraploid dallisgrass biotypes are mentioned in the literature (Caponio and Quarin 1990; Quarin and Caponio 1995). It is a sexually and autogamously reproducing allotetraploid (2n ¼ 40, IIJJ, 2C ¼ 2.43 pg) with high fertility and good seed production (Bashaw and Holt 1958b; Vaio et al. 2007), which has made it a good subject for research. However, it is not suitable for grazing as it is upright and has poor vigor. It is native to Argentina, Uruguay, and southern Brazil. Bashaw and Patrick (1966) irradiated yellow-anthered dallisgrass and obtained promising prostrate and decumbent plants. However, plants had reduced fertility, and vigor declined in later generations. Following crosses of P. dilatatum ssp. flavescens (2n ¼ 40) with P. notatum (2n ¼ 20, Notata) and P. vaginatum (2n ¼ 20, Disticha) to determine their genomic relationships, P. dilatatum ssp. flavescens  P. notatum and P. dilatatum ssp. flavescens  P. vaginatum hybrids were triploid (2n ¼ 30) with 28–29 univalents at meiosis (Burson et al. 1973). This confirmed that P. dilatatum ssp. flavescens was allotetraploid, with two unrelated genomes, which are unrelated to either P. notatum or P. vaginatum genomes. All hybrids were male and female sterile. P. dilatatum “Vacaria” and P. dilatatum “Virasoro”: The “Vacaria” and “Virasoro” biotypes are tetraploid (2n ¼ 40, IIJJ) biotypes from Brazil and Argentina (Vaio et al. 2007). DNA contents (2C ¼ 2.33 pg and 2.29 pg, respectively) were determined by flow cytometry (Vaio et al. 2007). P. dilatatum “Chiru”: The “Chiru” biotype, also known as “Uruguayan,” is more upright (up to 1 m at flowering) and more robust than other hexaploid biotypes. It is a pseudogamous, facultative apomict with limited sexual reproduction. It has relatively regular

W.M. Williams et al.

chromosome behavior at meiosis, with more than half of the pollen mother cells showing 30 bivalents (2n ¼ 60, 2C ¼ 3.57). Remaining cells had up to eight univalents and rare quadrivalents (Vaio et al. 2007). Burson et al. (1991) hypothesized an IIJJXX genomic configuration for this biotype. “Chiru” (referred to as “Uruguayan”) had significantly higher forage yield and higher fertility than “Common,” but similar nutritional qualities (Venuto et al. 2003). It has reasonable pollen stainability (42–84%) and variable seed set (6–66%) (Burson et al. 1991). P. dilatatum “Torres”: The “Torres” biotype originated from the northeastern corner of the state of Rio Grande do Sul, Brazil. It is semi-decumbent in habit with a significantly lower forage yield than “Common” (Venuto et al. 2003). “Torres” is a hexaploid (2n ¼ 60, 2C ¼ 3.58 pg) with very irregular meiosis with 59 univalents on average (Burson et al. 1991; Vaio et al. 2007). Burson et al. (1991) hypothesized that the lack of chromosome pairing was genetically controlled rather than because of a lack of homology between chromosomes. “Torres” has poor agronomic qualities, low pollen stainability, and poor seed set (Burson et al. 1991; Venuto et al. 2003). P. dilatatum “Uruguaiana”: The “Uruguaiana” biotype, from a small area in the western Rio Grande do Sul, Brazil, has a semi-decumbent growth habit, similar to “Common” dallisgrass (Venuto et al. 2003). It is hexaploid (2n ¼ 60, 2C ¼ 3.69) with variable chromosome configurations at meiosis, typically with 10 univalents, 25 bivalents, and rare quadrivalents (Burson et al. 1991; Vaio et al. 2007). Burson et al. (1991) hypothesized an IIJJX1X2 genomic configuration, where there is only partial homology between the “X” genomes. “Uruguaiana” has low pollen stainability and poor seed set (Burson et al. 1991). P. dilatatum “Bajada de Pena”: P. dilatatum “Bajada de Pena” is a hexaploid biotype (2n ¼ 60, IIJJX1X2, 2C ¼ 3.79 pg) from Uruguay (Vaio et al. 2007). Synthetic pentaploid hybrids (2n ¼ 50, IIJJX) were generated by intraspecific hybridization between tetraploid yellow-anthered P. dilatatum ssp. flavescens (2n ¼ 40, IIJJ) and hexaploid P. dilatatum “Chiru” (2n ¼ 60, IIJJXX; Burson 1991a, b). Hybrids had an average of 10 univalents and 20 bivalents at meiosis, suggesting that the “I” and “J” genomes in each parent were very similar. Some of the hybrids resembled “Common” dallisgrass, and Burson (1991b) speculated that the “Common” biotype might

12 Paspalum

have arisen from hybridization between tetraploid (e.g., P. dilatatum ssp. flavescens or P. urvillei) and hexaploid (e.g., P. dilatatum “Chiru”) cytotypes. However, it remains to be determined if the “X” genome in the “Chiru” race is similar to the “X” genome in common dallisgrass (Burson 1991a, b). Similar results were obtained when Burson (1992) hybridized P. urvillei with P. dilatatum “Chiru” (2n ¼ 60, IIJJXX) and with P. dilatatum “Uruguaiana” (2n ¼ 60, IIJJX1X2). Hybrids were pentaploid with similar meiotic chromosome behavior to “Common” dallisgrass. Thus, the “I” and “J” genomes of these two hexaploid races are the same as those in P. dilatatum ssp. flavescens and P. urvillei, and by extension, common dallisgrass, although the origins of the “X” genomes remain uncertain.

205

seven hybrids. All hybrids were triploid with ten univalents and ten bivalents at meiosis (Burson 1979). This indicated that one of the P. urvillei genomes was similar to the P. intermedium genome. Several P. dilatatum ssp. flavescens (2n ¼ 40)  P. urvillei (2n ¼ 40) hybrids were generated by Caponio and Quarin (1990). Hybrids were tetraploid and formed 20 bivalents during meiosis, indicating that they had similar genomes and confirming the IIJJ genome configuration for P. urvillei. Open pollinated, backcross and F2 seed were produced from the hybrids, therefore indicating that the hybrids would be useful to a breeding program.

12.6.1.4 P. dasypleurum Kunze ex E. Desv 12.6.1.2 P. pauciciliatum (Parodi) Herter: Prostrate Dallisgrass P. pauciciliatum (syn. P. dilatatum var. pauciciliatum Parodi) is a decumbent perennial with forage potential. It has outyielded and was more persistent than “Common” dallisgrass in field trials, and it has greater resistance to foliage diseases (Venuto et al. 2003). However, it is more susceptible to ergot and less fertile than “common” dallisgrass. It is tetraploid (2n ¼ 40, 2C ¼ 2.40 pg) with up to ten bivalents at meiosis (Burton and Jackson 1962; Vaio et al. 2007). A radiation breeding program was carried out in P. pauciciliatum and, although mutants were generated, the target attributes were not altered (Burton and Jackson 1962). One cultivar has been released (Burton and Wilson 1991).

12.6.1.3 P. urvillei Steud.: Vasey Grass P. urvillei is an allotetraploid (2n ¼ 40, IIJJ, 2C ¼ 2.40 pg) that reproduces sexually and is usually selfpollinated (Vaio et al. 2007). It is a warm season bunch-grass indigenous to southern Brazil and northern Argentina but is now grown in warm regions around the world. It is a robust species growing up to 2 m in height, which makes excellent hay and is an acceptable forage. Hybridization between P. urvillei (2n ¼ 40) and P. intermedium (2n ¼ 20, Quadrifaria) generated

P. dasypleurum has the southernmost distribution of South American Paspalum species. It is native to central and southern Chile and southwestern Argentina. It has desirable traits in that it is leafy and is a potential source of cold-tolerance (Quarin and Caponio 1995). The species shows variation for anther color (yellow and purple). P. dasypleurum is tetraploid (2x ¼ 40, 2C ¼ 2.39 pg) and showed 20 bivalents at meiosis (Vaio et al. 2007). P. dasypleurum behaves sexually, is selffertile, and is predominantly autogamous (Quarin and Caponio 1995). Quarin and Caponio (1995) intercrossed species of the Dilatata group to determine the genomic relationship among these species. Pollination of 253 emasculated P. dilatatum ssp. flavescens (2n ¼ 40, IIJJ) spikelets with P. dasypleurum (2n ¼ 40) yielded two hybrid plants. Both hybrids were indehiscent and self-incompatible but produced seed (8.2%) when backcrossed to the P. dilatatum ssp. flavescens parent. Pollination of 102 emasculated purple-anthered P. dasypleurum florets with P. urvillei (2n ¼ 40, IIJJ) generated 55 hybrids. Hybrids segregated for purple (28 plants) and yellow (27 plants) anthers. Anthers of all hybrids were highly indehiscent and no self-seed was recovered, although seeds were obtained when hybrids were backcrossed to the female (4.8%) and male (18.6%) parents. All hybrid plants tested were tetraploid (2n ¼ 40) with 20 bivalents at meiosis, indicating an IIJJ genome configuration for P. dasypleurum. The success of backcrossing indicated that hybridization among these species may be a useful breeding method for genetic improvement within the Dilatata group.

206

12.6.2 Informal Group Livida Chase 12.6.2.1 P. denticulatum Trin.: Longtom Paspalum P. denticulatum (including P. lividum; Zuloaga and Morrone 2003) is a stoloniferous perennial with forage potential, which is often used for erosion control. It is an allopolyploid (2n ¼ 40, 70) with chromosomes typically forming bivalents with few univalents and quadrivalents. P. denticulatum reproduces via apospory, followed by pseudospory (Burson and Bennett 1971a).

12.6.3 Informal Group Virgata Chase 12.6.3.1 P. conspersum Schrad.: Scattered Paspalum This is a leafy, robust, warm-season perennial species, native to South America. It grows up to 2-m tall and is distributed from Mexico to Argentina, primarily in open, moist regions along streams (Chase 1929). A sexual segmental allotetraploid (2n ¼ 40), it resembles the more common P. virgatum (Burson 1978). Burson and Bennett (1976) investigated the feasibility of using P. conspersum in interspecific hybridization. Eighty P. dilatatum ssp. flavescens (2n ¼ 40, Dilatata)  P. conspersum (2n ¼ 40) hybrids, and a single P. conspersum  P. malachophyllum (2n ¼ 40, Anachyris) hybrid, were developed to study the genomic relationships of P. conspersum. Analysis of meiotic chromosome pairing in P. conspersum and P. conspersum  P. malachophyllum led to the conclusion that P. conspersum genomes were partially autosyndetic, i.e., show partial homology between genomes. P. dilatatum  P. conspersum hybrids were tetraploid (2n ¼ 40) with an average of 20 univalents, 10 bivalents, and occasional tri- and quadrivalents. Burson and Bennett (1976) deduced that P. dilatatum and P. conspersum share one genome, and their second genomes had a small amount of homology. To determine which genomes were homologous, P. juergensii (2n ¼ 20, Paniculata) and P. intermedium (2n ¼ 20, Quadrifaria) were crossed to P. conspersum (2n ¼ 40), giving rise to two and 13 hybrids, respectively (Burson 1978). Hybrids were triploid and showed differing

W.M. Williams et al.

chromosomal pairing behavior during meiosis. P. juergensii  P. conspersum hybrids had an average of ten univalents and ten bivalents, suggesting that one of the P. conspersum genomes shared homology with the P. juergensii genome. P. intermedium  P. conspersum hybrids typically had 20 univalents and five bivalents suggesting that there were some similarities between the genomes. Burson (1978) proposed “J” for the P. juergensii genome and “I” for the P. intermedium genome. He further proposed “IIJJ” for P. dilatatum ssp. flavescens, and I2I2JJ for P. conspersum, indicating partial homology between P. intermedium and the other P. conspersum genome.

12.6.3.2 P. rufum Nees ex. Steud P. rufum is a robust, erect perennial grass native to Paraguay, southern Brazil, Uruguay, and northeastern Argentina, where it is normally found in marshes, wet savannahs, or low, open grounds. Diploid (2n ¼ 20) and segmental tetraploid (2n ¼ 40) cytotypes exist (Quarin et al. 1998). The diploid reproduces sexually, while the tetraploid has irregular meiosis and reproduces by aposporous apomixis.

12.6.3.3 P. virgatum L.: Talquezal P. virgatum is a densely clumped perennial, reaching 2 m at flowering. It is distributed throughout the Americas and Australia (Chase 1929).

12.6.4 Informal Group Quadrifaria Barreto The Quadrifaria group was extracted from the Virgata group by Barreto (1954) based on morphological characteristics. There are approximately ten Paspalum species in the Quadrifaria Group (Adamowski et al. 2005; Vaio et al. 2005; Zuloaga and Morrone 2005). Many species have both diploid and polyploid cytotypes. Diploid cytotypes are sexual and self-incompatible, while tetraploids reproduce apomictically. It is widely considered that an ancestral species from the Quadrifaria group was the donor of the Dilatata group “I” genome.

12 Paspalum

Hybrids were produced between diploid cytotypes of P. haumanii, P. intermedium, P. brunneum (reclassified as P. coryphaeum Trin., Zuloaga and Morrone 2003), P. quadrifarium (all Quadrifaria), and P. rufum (Virgata) to investigate their genomic relationships (Quarin and Norrmann 1990). All progeny were diploid and all showed some bivalent pairing during meiosis. Quarin and Norrmann (1990) concluded that all five species shared a common genome (I) that had undergone some changes over time, and the Quadrifaria group of species could be included within the Virgata group. More recently, Vaio et al. (2005), using fluorescence in situ hybridization (FISH), rDNA, and cpDNA sequences, have confirmed differentiation among the Quadrifaria group of species.

12.6.4.1 P. coryphaeum Trin.: Emperor Crowngrass P. coryphaeum (including P. brunneum; Zuloaga and Morrone 2003) is native to tropical northeastern regions of South America and the Caribbean where it grows on small hills or well-drained soils.

207

(2n ¼ 20, II, 2C ¼ 1.51 pg; Vaio et al. 2007) and its chromosomes pair as ten bivalents at meiosis (Burson 1981a). It was considered a possible progenitor of one of the P. dilatatum genomes.

12.6.4.5 P. quadrifarium Lamk.: Tussock Paspalum P. quadrifarium is native to Uruguay, central and eastern Argentina, and Rio Grande do Sul, Brazil, where it is found in low, humid, and generally salty areas. It has low forage value but was probably one of the main species of the “Flooding Pampa” (Perelman et al. 2003). P. quadrifarium-dominated regions of the Flooding Pampa had higher species richness but lower alien invasion than areas from which P. quadrifarium had been removed (Perelman et al. 2003). Diploid (2n ¼ 20, II, 2C ¼ 1.86 pg), triploid (2n ¼ 30, III0 ), tetraploid (2n ¼ 40, III0 I0 , 2C ¼ 3.22 pg), and hexaploid (2n ¼ 60) accessions have been identified (Speranza et al. 2003; Vaio et al. 2007).

12.6.5 Informal Group Paniculata Chase 12.6.4.2 P. exaltatum J. Presl

12.6.5.1 P. juergensii Hack

P. exaltatum (including P. arechavaletae Hack. ex. Arechav., Zuloaga and Morrone 2003) is an allotetraploid (2n ¼ 40, IIII, 2C ¼ 2.95 pg; Vaio et al. 2007) that reproduces via apospory, followed by pseudogamy. P. arechavaletae was shown to reproduce sexually, with limited apomixis (Burson and Bennett 1971a). It is native to Brazil and Paraguay.

P. juergensii is a sexual diploid (2n ¼ 20, JJ, 2C ¼ 1.24 pg; Vaio et al. 2007) indigenous to South America (Brazil and Uruguay). It has been used widely in genomic composition studies and it is a proposed progenitor of P. dilatatum, donating the “J” genome (Burson and Bennett 1971b, 1972; Burson et al. 1973, among others). Hybridization studies between P. juergensii and P. vaginatum (2n ¼ 20, Disticha) and with P. setaceum var. ciliatifolium (2n ¼ 20, Setacea) showed that the latter species did not contain “J” genomes (Burson 1981a). Resulting hybrids were diploid but had very irregular chromosome behavior at meiosis, often with only a single bivalent chromosome pair. Hybrids were male and female sterile.

12.6.4.3 P. haumanii Parodi P. haumanii is a diploid (2n ¼ 20, II, 2C ¼ 1.50 pg; Vaio et al. 2007), native to Argentina, Brazil, Paraguay, and Uruguay.

12.6.4.4 P. intermedium Monro ex Morong & Britton: Intermediate Paspalum

12.6.5.2 P. paniculatum L.: Arrocillo

P. intermedium is a coarse leaved, erect bunchgrass native to South America (Brazil). It is a sexual diploid

P. paniculatum is a semi-decumbent weedy species native to South America and now distributed in

208

warm regions around the world, including Central and South America, the Caribbean, and Australia. It is a sexual diploid (2n ¼ 20, JJ) and shares a common genome with P. dilatatum (Burson 1979). Hybridization between P. dilatatum ssp. flavescens (2n ¼ 40) and P. paniculatum (2n ¼ 20) generated a single triploid hybrid (2n ¼ 30; Burson 1979). Chromosomes associated as ten univalents and ten bivalents at meiosis. This indicated that one of the P. dilatatum genomes was similar to the P. paniculatum genome and led Burson (1979) to conclude that a species or ancestor of the Paniculata group was one of the progenitors of P. dilatatum.

12.6.6 Informal Group Notata Chase This group intergrades with the groups Linearia and Alma and with P. equitans Mez (Fasciculata). The boundaries of these groups remain unclear and controversial (Ferreira and Souza-Chies 2005; Souza-Chies et al. 2006; Essi and Souza-Chies 2007), and reference to the Notata/Linearia group is frequent. For example, Zuloaga et al. (2004) used a single combined subgroup in classifying species of the Notata group. However, there have been progressive indications that the groupings based on morphological traits are not monophyletic when tested by molecular data. On the basis of morphological and DNA sequence analyses, Souza-Chies et al. (2006) identified a “core” Notata group as P. conduplicatum Canto-Dorow, Valls & Longhi-Wagner, P. notatum, P. minus, P. pumilum Nees, and P. subciliatum Chase. Essi and Souza-Chies (2007) used DNA sequence data to identify a core monophyletic group of eight species making up a Notata group senso stricto. These were P. conduplicatum, P. notatum, P. minus, P. pumilum, and P. subciliatum, along with P. barretoi CantoDorow, P. nummularium Chase, and P. flaccidum Nees, which is morphologically distinct and previously classified as part of the Recta group (Zuloaga et al. 2004). P. barretoi, P. nummularium, and P. pumilum are known only as sexual autogamous diploids, whereas P. conduplicatum is hexaploid, P. minus is pentaploid, and P. flaccidum is tetraploid (Pozzobon et al. 2008). P. cromyorhizon Trin. ex Doell was found to be not close enough in DNA sequence to be regarded as a

W.M. Williams et al.

subspecies of P. notatum, as had previously been suggested. This species showed chromosome pairing consistent with being a segmental allotetraploid (Burson and Bennett 1971a). The morphological affinities of P. dedeccae Quarin, P. ionanthum Chase, and P. filifolium Nees ex Steud. with the Notata group were not supported by DNA results. Despite its morphological similarity, P. almum Chase (x ¼ 6) was distant from the Notata group (Souza-Chies et al. 2006).

€ gge: Bahiagrass 12.6.6.1 P. notatum Flu P. notatum is one of the most important species of the Notata group. It is a rhizomatous perennial and is a major constituent of the native grasslands from Mexico to Argentina and Uruguay. Most biotypes of P. notatum var. notatum are self-compatible apomictic tetraploids (2n ¼ 4x ¼ 40) with a genome formula of NNNN (Burson 1983; Quarin et al. 1984). However, self-incompatible sexual diploid (2n ¼ 2x ¼ 20) populations are native to northeastern and central-eastern Argentina and are collectively classified as P. notatum var. saurae Parodi (Dahmer et al. 2008). This variety has become very popular as a forage grass in southern USA, where it is known as Pensacola bahiagrass. It has a small genome with a 1C DNA content of 0.6 pg (Jarret et al. 1995) and a genome formula of NN (Burson 1981b). Some occasional findings of triploid and pentaploid individuals have been reported (Quarin 1992; Tischler and Burson 1995; Daurelio et al. 2004). There is variability reported within the apomictic tetraploid P. notatum, but no clear separation allows the division into different botanical varieties (Cidade et al. 2008). Bahiagrass is also used as a turfgrass in warm, humid areas. It is characterized by slow seedling establishment and slower vegetative growth than other subtropical grasses (Busey 1992). It performed extremely well as a forage grass in subtropical Australia in a trial conducted by Shaw et al. (1965). Cidade et al. (2008) used inter-simple sequence repeat (ISSR) markers to show that considerable genetic diversity exists among tetraploid populations collected near the likely center of origin in South Brazil, despite the apomictic breeding system. This diversity was attributed to the presence of occasional sexual diploids and/or tetraploids with facultative apomixis. Thus, the species has a powerful evolutionary

12 Paspalum

system of generation of new genetic diversity followed by rapid fixation of superior combinations by apomixis. Pillay and Mazzella (1997) compared the chloroplast genome of tetraploid P. notatum (2n ¼ 4x) with that of “Common” P. dilatatum (2n ¼ 5x) using restriction site analysis. The two species had clearly different mutations that distinguished the chloroplast genomes, a result consistent with morphological and nuclear cytological data, indicating that there is a considerable genetic distance between them. The overall Paspalum chloroplast genome was organized in a similar way to those of wheat and barley. Apomixis in P. notatum: In P. notatum, and most apomictic Paspalum species, seeds develop following apospory, parthenogenesis, and pseudogamy. In ovules, the meiotic products break down and are replaced by embryo sacs (usually multiple) developed from nucellar cells. These have the sporophytic chromosome number and are distinguished from meiotic embryo sacs by a lack of the proliferated antipodal cells characteristic of the latter. In facultative apomictic plants that have a high degree of sexuality, some ovules produce both kinds of embryo sacs (Martinez et al. 2001). Autotetraploids produced by colchicinedoubling of sexual diploids are often of this facultative apomictic type (Quarin et al. 2001). Apomixis in tetraploid P. notatum was found to be environmentally stable by Burton (1982a). Breeding apomictic P. notatum: Although most populations of P. notatum are tetraploid, it has been noted that the diploid type (Pensacola bahiagrass) has several more desirable agronomic characteristics, including more rapid spread, longer leaves, greater winter-hardiness, better ergot resistance, and small seeds that do not require scarification (Burton and Hanna 1986). It was, therefore, important to be able to hybridize the diploid and tetraploid to combine the features of both. This could be done by chromosome doubling the diploid to obtain sexual tetraploids and crossing these with apomictic tetraploids to obtain sexual and apomictic tetraploid families (Burton and Forbes Jr 1960; Forbes and Burton 1961). Burton and Hanna (1986) achieved the same objective by using a spontaneous male sterile triploid, which produced tetraploid progeny when pollinated with the diploid population. These tetraploids were facultative apomicts that, when reproducing sexually, generated variable families that could form base populations for selection

209

and breeding. Further development of sexual tetraploids from the above material led to the registration of two germplasm lines for breeding purposes (Quarin et al. 2003). A progeny plant from one of these lines was shown by molecular marker analysis to reproduce totally by sexual means, confirming the potential of this material for plant breeding (Ortiz et al. 1997). Time of pollination can be altered to achieve biparental progeny from apomictic tetraploid P. notatum. Martinez et al. (1994) achieved (2n þ n) progenies by pollinating tetraploid P. notatum 2 or 3 days before anthesis with either the same cytotype or with diploid Pensacola bahiagrass. Apparently, early pollination preempted the parthenogenic development of the unreduced egg, allowing fertilization to occur. Thus, despite apomixis, it is possible to create new genome combinations among wild species and to breed within this complex. Acuna et al. (2007) have confirmed Burton (1955) in finding that some genotypes of diploid Pensacola bahiagrass can show a significant amount of selffertility, and it should not be assumed that all are fully outcrossing. The same applied to artificial tetraploids produced from chromosome doubling of the diploids. Some artificial tetraploids also occasionally produced apomictic embryo sacs (Quarin et al. 2001; Acuna et al. 2007), indicating that the alleles for apomixis must be present in at least some diploids. Breeding diploid sexual “Pensacola bahiagrass”: Burton (1955) showed that this diploid material was predominantly cross-fertilizing and highly variable. It showed considerable (25%) loss of vegetative yield with inbreeding. Two of the classical forage plant breeding methods, recurrent restricted phenotypic selection, RRPS (Burton 1974), and the improved version (Burton 1982b) were refined on Pensacola bahiagrass. Nine cycles of RRPS resulted in the development of the cultivar Tifton 9 (Burton 1989). This material continues to be a source of useful sexual genetic diversity for the improvement of the species.

12.6.6.2 P. subciliatum Chase This tetraploid species has been little studied cytogenetically. One Brazilian accession analyzed by Adamowski et al. (1998) showed 20 bivalents at metaphase I of meiosis, suggesting an allotetraploid origin. After this stage, one of the parental chromosome sets

210

appeared to be out of phase with the other, resulting in the elimination of chromosomes from one set. Pollen fertility was greatly reduced, but seed viability was maintained. Chromosome elimination during gametogenesis is unusual in plants as it usually occurs in somatic cells during early embryogenesis.

12.6.7 Informal Group Disticha Chase The Disticha group consists of three species, including P. distichum L. and P. vaginatum Swartz. The latter is the most important among them (Hojsgaard et al. 2009) and can also be found in a diploid, tetraploid, and hexaploid forms (Quarin and Burson 1983).

12.6.7.1 P. vaginatum Swartz: Seashore Paspalum Native to tropical and subtropical America, this species is adapted to halophytic coastal environments and can stand a very wide soil fertility and pH range (4–10). It is used for forage production and is extensively used as a turfgrass for golf courses, lawns, and parks and for soil conservation where salt tolerance is required (Dudeck and Peacock 1985). It is also used for land reclamation under saline conditions and drainage water recycling schemes. It is hard-wearing due to the shoot density, has general pest resistance, salt tolerance, and some drought, waterlogging, and cold tolerance (Dudeck and Peacock 1985; Huang et al. 1997a, b; Trenholm et al. 1999, 2000; Lee et al. 2005). It exists in dwarf and semi-dwarf forms with short leaves, generally less than 10 cm. The species is more tolerant of low light intensities than hybrid Cynodon cultivars (Jiang et al. 2004). Several cultivars have been developed in SE USA and in Australia based on dwarf and semi-dwarf material. More than 300 populations have been collected that show considerable diversity in leaf texture, from coarse to very fine (Cardona and Duncan 1997). The diploid forms are self-incompatible and produce seeds, but yields are poor and seeds are generally low in viability. Crosses between sexual tetraploid P. dilatatum and P. vaginatum produced triploid hybrids only when P. vaginatum was used as the male parent. There was virtually no chromosome pairing and the hybrids were

W.M. Williams et al.

completely sterile (Burson et al. 1973). Similarly hybridization of P. juergensii and P. vaginatum was successful only when P. vaginatum was the male parent. Again, there was almost no chromosome pairing and none of the hybrids produced seed (Burson 1981a). P. vaginatum does not have either “I” or “J” genomes found in the Dilatata complex, and the genomic formula “DD” was proposed (Burson 1981a). The existence of significant genetic variation in salt tolerance among populations of P. vaginatum (Lee et al. 2005) has led to the use of this species as a tool for investigation of salt tolerance mechanisms (e.g., Lee et al. 2008). It has increased osmotic potential and may also have the ability to take up potassium and lose excess salt through old leaves (Shonubi and Okusanya 2007). High frequency plant regeneration from embryogenic callus cultures was achieved by Cardona and Duncan (1997). This opened the way for the potential use of somaclonal variation to generate variants for genetic improvement.

12.6.7.2 P. distichum L.: Knot Grass This species is native to the Americas and has been introduced widely around the world. It is a rhizomatous perennial adapted to wetlands, both saline and fresh water. It is palatable with moderate spreading ability, high salt tolerance, but lower shade tolerance than P. vaginatum. It is used on contaminated soils where it takes up minerals such as lead, zinc, and cadmium from the soil (Lan et al. 1992; Yu et al. 2005).

12.6.8 Informal Group Plicatula Chase The Plicatula group consists of up to 20 species (Zuloaga and Morrone 2005) and shows a lot of potential for agriculture. Central and western Brazil, Bolivia, and Paraguay are the center of variation for this group (Quarin et al. 1997). The most important species known to be cultivated include the apomictic tetraploids (2n ¼ 4x ¼ 40) P. atratum Swallen, P. plicatulum Michx., P. glaucescens Hack., and P. guenoarum Arechav. (Quarin et al. 1997; Takayama et al. 1998; Pozzobon and Valls 2000). The ancient Asian cereal P. scrobiculatum L. also occurs in this group.

12 Paspalum

12.6.8.1 P. atratum Swallen: Atra Paspalum, Capim-pojuca, Pasto Pojuca This native of Brazil and Bolivia has only relatively recently become used in agriculture. It is naturalized in the subtropics of Australia, USA, and Southeast Asia. It is in use in Thailand, Bhutan, Australia, and USA for forage or hedgerows for erosion control. It grows up to two meters tall when flowering and is suitable for cut and carry forage systems and shows promise for hay making. It can tolerate soils of low fertility in regions of high rainfall (more than 1,000 mm) and tolerates waterlogging. It has limited cool season growth and is damaged by frost but recovers quickly with onset of warm conditions. It is shade-tolerant and useful in agroforestry. It is highly palatable with very good forage quality but is better suited to cutting or lenient grazing managements (Quarin et al. 1997; Kalmbacher and Martin 1999). Relative to many other apomictic Paspalum species, P. atratum is an exceptionally good seeder (Quarin et al. 1997). Nevertheless, poor seed quantity (110 kg/ha) and quality limits its commercialization (Quarin et al. 1997; Kalmbacher et al. 1999). Cultivars have been developed in USA, Brazil, and Thailand, but all are derived from a single accession from Campo Grande, Brazil. P. atratum is a pseudogamous obligate apomictic tetraploid with an average of 19 bivalents and very few uni- or quadrivalents, suggesting that it is a probable segmental allotetraploid (Quarin et al. 1997, Takayama et al. 1998). Unless a sexual diploid (or tetraploid) is found in future germplasm collections, population improvement will be limited to screening of wild accessions (Quarin et al. 1997).

12.6.8.2 P. plicatulum Michx.: Brownseed Paspalum Native to southeastern USA, Central America, and the Caribbean, this species is now widespread throughout South America, China, Australia, and temperate Asia. Its natural adaptation is to moist habitats in high rainfall regions, but it can tolerate as little as 750 mm per year. It is perennial and grows up to 1.3-m tall. P. plicatulum is used as permanent pasture in poorly drained low fertility soils, tolerating high aluminum concentrations. It is tolerant of waterlogging and flooding for short periods and is moderately drought hardy. It is burned

211

by frost, but the crown survives (Shaw et al. 1965). It is less tolerant of grazing than P. atratum and is not shade-tolerant. P. plicatulum is less palatable than other grasses and declines rapidly in quality with maturity. It is established from seed or cuttings. It is free of ergot attack, and dry seed can last for up to 2 years. It is also free of other diseases and has no weedy tendencies. An accession from Guatemala performed extremely well in trials in Queensland, Australia (Shaw et al. 1965). Three cultivars have been developed in Queensland, Australia, from different materials (Oram 1990). This is predominantly an apomictic allotetraploid species, but there are occasional self-incompatible sexual diploids (Pagliarini et al. 2001; Adamowski et al. 2005; Pozzobon et al. 2008). Chromosome doubling of the diploids to produce sexual tetraploids may provide a way of breeding at the tetraploid level.

12.6.8.3 P. guenoarum Arechav.: Pasto Rojo P. guenoarum is a leafy, upright perennial, which can reach 2 m when flowering. It originated from western South America and is currently used in South America, Zimbabwe, and Australia. P. guenoarum is used for grazing, cut-and-carry, hay, and silage. It is moderately tolerant of water-logging and drought and grows well in areas with annual rainfalls between 1,200 and 2,000 mm per year. Natural populations of P. guenoarum are found in an unusually narrow temperature range (19–21 C); although “Bela Vista” (Brazil) performs well in areas with average annual temperatures of 27 C. Genotypes tested in Australia have shown good frost tolerance. P. guenoarum is palatable to cattle and horses and performs best under lenient management. Several cultivars are available in South America, and promising accessions have been identified in material imported into Australia. All accessions tested are tetraploids (2n ¼ 40; Burson and Bennett 1971a; Takayama et al. 1998; Espinoza et al. 2001; Adamowski et al. 2005) and likely to be segmental allotetraploids (Burson and Bennett 1971a). Pollination is needed for endosperm development, indicating pseudogamy in this species, and Burson and Bennett (1971b) showed that pollen stainability was reasonable (47.1%) and suggested that P. guenoarum pollen could be used in Paspalum breeding programs.

212

12.6.8.4 P. glaucescens Hackel P. glaucescens (including P. yaguaronense Henrard; Pozzobon and Valls 2000) populations are from Savannas and ill-drained soils in Argentina, Brazil, and Paraguay. Material was imported into Australia in 1948 and was shown to have high palatability, better annual yields than naturalized P. dilatatum, and good frost tolerance (Shaw et al. 1965). Diploid, tetraploid, and hexaploid accessions have been identified (Pritchard 1962; Takayama et al. 1998; Pozzobon and Valls 2000; Pozzobon et al. 2000; Pagliarini et al. 2001). Diploids (2n ¼ 20) reproduce sexually and are typically selfincompatible (Pritchard 1962; Pozzobon and Valls 2000), although Pritchard (1962) identified a single self-fertile plant. Tetraploids (2n ¼ 40) and hexaploids (2n ¼ 60) are apomictic and are thought to be segmental allotetraploids and interspecific hybrids, respectively (Pozzobon and Valls 2000).

W.M. Williams et al.

tula group. In many respects, this complex typifies the situation with regard to the potential use of wild species in breeding new populations within the genus.

12.6.8.7 Other Species in the Plicatula Group A number of diploid species and cytotypes of the Plicatula group might be useful for interspecific breeding programs. Brazilian species P. hydrophilum Henrard (2n ¼ 20, 30, 40), P. modestum Mez (2n ¼ 20, 40), and P. palustre Mez (2n ¼ 20) were identified as species with potential for use in wetlands (Pozzobon and Valls 2003). Espinoza et al. (2001) speculated that P. limbatum Henrard (2n ¼ 20, sexual) and P. lenticulare Kunth (2n ¼ 40, apomictic) were diploid and tetraploid accessions of the same species.

12.6.8.8 P. scrobiculatum L.: Kodo Millet 12.6.8.5 P. nicorae Parodi: Brunswick Grass P. nicorae is native to eastern South America and naturalized in USA and Australia. It is a variable perennial with long, deep, and vigorous rhizomes and is short-statured, reaching only 40 cm at heading (Barreto 1956). It is used for permanent pasture in the tropics, being tolerant of heavy grazing, palatable, and drought and frost-tolerant. It is an excellent seed producer, sometimes producing over 1 ton of seed/ha. It is a tetraploid and facultative apomict (Burson and Bennett 1970; Pagliarini et al. 2001) that may be interfertile with P. plicatulum. Cultivars have been developed for pasture and turf in Georgia, USA, and Queensland, Australia.

12.6.8.6 P. compressifolium Swallen P. compressifolium Swallen is an important perennial forage contributor to native grasslands in South America, especially South Brazil, East Paraguay, and Northeast Argentina. It has been shown to be a probable agamic complex composed of occasional diploid, predominant tetraploid, and rare hexaploid races (Quarin et al. 1996; Adamowski et al. 2005). The existence of sexual diploid populations provides a way to breed with this complex and within the Plica-

One of the minor millets, this tropical species is grown in fertility-depleted soils in Africa and Asia (Clayton 1975). Its origin is uncertain, but it appears to have been domesticated in India in Neolithic and postNeolithic times (Vishnu-Mittre 1978) and is now mainly of local tribal use. It is found in the wetter parts of Nigeria in moist and shady places but is not halophytic and so is found in non-saline regions (Shonubi and Okusanya 2007). It is very invasive and is a noxious weed in North America. Somatic embryogenesis has been achieved (Vikrant 2002).

12.6.9 Recommendations for Further Action The powerful evolutionary system of Paspalum needs to be better understood, if it is to be exploited by plant breeders. Great genetic power is generated by the occurrence of occasional sexual hybridization of diverse plants, followed by fixation of successful polyploid genotypes by apomixis. If breeders could exploit this system, then considerable advances would be made, not only in Paspalum, but also in other species that combine sexual and apomictic modes of reproduction. In this context, the identification and conservation of the sexual diploid Paspalum

12 Paspalum

species requires priority action as these species will provide the engine for generation of new and useful genetic diversity that can be used by plant breeders.

References Acuna CA, Blount AR, Quesenberry KH, Hanna WW, Kenworthy KE (2007) Reproductive characterization of bahiagrass germplasm. Crop Sci 47:1711–1717 Adamowski ED, Pagliarini MS, Batista LAR (1998) Chromosome elimination in Paspalum subciliatum (Notata group). Sex Plant Reprod 11:272–276 Adamowski ED, Pagliarini MS, Bonato ABM, Batista LAR, Valls JFM (2005) Chromosome numbers and meiotic behavior of some Paspalum accessions. Genet Mol Biol 28:773–780 Aliscioni SS, Denham SS (2008) Rachis of the genus Paspalum L. (Poaceae: Panicoideae: Paniceae): anatomy and taxonomic significance of the primary branches of the inflorescences. Flora 203:60–76 Allem AC, Valls JFM (1987) Recursos forrageiros nativos do Pantanal mato-grossense. EMBRAPA, Centro Nacional de Recursos Gene´ticos e Biotecnologia, Brası´lia, Brazil Barreto IL (1954) Species related to Paspalum virgatum in South America. Rev Argent Agron 21:125–142 Barreto IL (1956) The species related to Paspalum plicatulum in Rio Grande do Sul (Brazil). Rev Argent Agron 23:53–70 Bashaw EC, Forbes I Jr (1958) Chromosome numbers and microsporogenesis in Dallisgrass Paspalum dilatatum Poir. Agron J 50:441–445 Bashaw EC, Holt EC (1958a) Megasporogenesis, embryo sac development and embryogenesis in Dallisgrass, Paspalum dilatatum, Poir. Agron J 50:753–756 Bashaw EC, Holt EC (1958b) Reproduction in dallisgrass, Paspalum dilatatum. In: Annual meeting of American society of agronomy, Lafayette, Indiana, USA, p 43 Bashaw EC, Patrick KH (1966) Radiation studies with sexual yellow-anthered dallisgrass, Paspalum dilatatum Poir. Crop Sci 6:195–198 Bennett HW (1962) Dallisgrass, bahiagrass, and vaseygrass. In: Hughes HD, Heath ME, Metcalfe DS (eds) Forages: the science of grassland agriculture. Iowa State University Press, Ames, Iowa, USA, pp 281–285 Bennett HW, Burson BL, Bashaw EC (1969) Intraspecific hybridization in dallisgrass, Paspalum dilatatum Poir. Crop Sci 9:807–809 Bonilla JR, Quarin CL (1997) Diplosporous and aposporous apomixis in a pentaploid race of Paspalum minus. Plant Sci 127:97–104 Burson BL (1978) Genome relations between Paspalum conspersum and two diploid Paspalum species. Can J Genet Cytol 20:365–372 Burson BL (1979) Cytogenetics of Paspalum urvillei  P. intermedium and P. dilatatum  P. paniculatum hybrids. Crop Sci 19:534–538 Burson BL (1981a) Cytogenetic relationships between Paspalum juergensii and P. intermedium, P. vaginatum, and P. setaceum var. ciliatifolium. Crop Sci 21:515–519

213 Burson BL (1981b) Genome relations among four diploid Paspalum species. Bot Gaz 142:592–596 Burson BL (1983) Phylogenetic investigations of Paspalum dilatatum and related species. In: Proceedings of 14th international grassland congress, Lexington, Kentucky, USA, 15–24 June 1981, pp 170–173 Burson BL (1989) Phylogenetics of apomictic Paspalum dilatatum. In: Proceedings of 16th international grassland congress, Nice, France, Oct 1989, pp 413–414 Burson BL (1991a) Genome relationships between tetraploid and hexaploid biotypes of dallisgrass, Paspalum dilatatum. Bot Gaz 152:219–223 Burson BL (1991b) Homology of chromosomes of the X genomes in common and Uruguayan dallisgrass, Paspalum dilatatum. Genome 34:950–953 Burson BL (1992) Cytology and reproductive behavior of hybrids between Paspalum urvillei and two hexaploid P. dilatatum biotypes. Genome 35:1002–1006 Burson BL, Bennett HW (1970) Cytology, method of reproduction, and fertility of Brunswick-grass, Paspalum nicorae Parodi. Crop Sci 10:184–187 Burson BL, Bennett HW (1971a) Chromosome numbers, microsporogenesis, and mode of reproduction of seven Paspalum species. Crop Sci 11:292–294 Burson BL, Bennett HW (1971b) Cytogenetic relationships between vaseygrass, Paspalum urvillei, and two diploid species, Paspalum juergensii and Paspalum vaginatum. In: Proceedings, abstracts of papers and addresses presented at the 68th Annual Convention of the Association of Southern Agricultural Workers, Inc., Jacksonville, Fla., USA, p 49 Burson BL, Bennett HW (1972) Cytogenetics of Paspalum urvillei  P. juergensii and P. urvillei  P. vaginatum hybrids. Crop Sci 12:105–108 Burson BL, Bennett HW (1976) Cytogenetics of Paspalum conspersum and its genomic relationship with yellow-anthered P. dilatatum and P. malachophyllum. Can J Genet Cytol 18:701–708 Burson BL, Lee HS, Bennett HW (1973) Genome relations between tetraploid Paspalum dilatatum and four diploid Paspalum species. Crop Sci 13:739–742 Burson BL, Voigt PW, Evers GW (1991) Cytology, reproductive behavior, and forage potential of hexaploid dallisgrass biotypes. Crop Sci 31:636–641 Burton GW (1955) Breeding pensacola bahiagrass. Paspalum notatum: I. Method of reproduction. Agron J 47:311–314 Burton GW (1974) Recurrent restricted phenotypic selection increases forage yields of Pensacola bahiagrass. Crop Sci 14:831–835 Burton GW (1982a) Effect of environment on apomixis in Bahiagrass. Crop Sci 22:109–111 Burton GW (1982b) Improved recurrent restricted phenotypic selection increases bahiagrass forage yields. Crop Sci 22:1058–1061 Burton GW (1989) Registration of ‘Tifton 9’ Pensacola bahiagrass. Crop Sci 29:1326 Burton GW, Forbes Jr I (1960) The genetics and manipulation of obligate apomixis in common Bahia grass Paspalum notatum Flugge. In: Proceedings of the 8th international grassland congress, University of Reading, England, UK, pp 66–71

214 Burton GW, Hanna WW (1986) Bahiagrass tetraploids produced by making (apomictic tetraploid  diploid)  diploid hybrids. Crop Sci 26:1254–1256 Burton GW, Jackson JE (1962) Radiation breeding of apomictic prostrate dallisgrass. Paspalum dilatatum var. pauciciliatum. Crop Sci 2:495–497 Burton GW, Wilson JP (1991) Registration of prostrate dallisgrass germplasm #1. Crop Sci 31:1392 Busey P (1992) Seedling growth, fertilization timing, and establishment of bahiagrass. Crop Sci 32:1099–1103 Caceres ME, Matzk F, Busti A, Pupilli F, Arcioni S (2001) Apomixis and sexuality in Paspalum simplex: characterization of the mode of reproduction in segregating progenies by different methods. Sex Plant Reprod 14:201–206 Caponio I, Quarin CL (1990) Intra- and interspecific hybridization between dallisgrass and vaseygrass. Crop Sci 30:362–364 Cardona CA, Duncan RR (1997) Callus induction and high efficiency plant regeneration via somatic embryogenesis in paspalum. Crop Sci 37:1297–1302 Chase A (1929) The North American species of Paspalum. Contrib US Nat Herb 28:1–310 Chase A (1939) Paspalum of South America, unpublished manuscript. Hitchcock and Chase Library, Botany Department, Smithsonian Institute, Washington DC, USA Cialdella AM, Morrone O, Zuloaga FO (1995) Revision de las especies del genero Paspalum (Poaceae: Panicoideae: Paniceae), grupo Bonplandiana. Darwiniana 33:67–95 Cidade FW, Dall’Agnol M, Bered F, Souza-Chies TT (2008) Genetic diversity of the complex Paspalum notatum Flugge (Paniceae: Panicoideae). Genet Resour Crop Evol 55:235–246 Clayton WD (1975) The Paspalum scrobiculatum complex in tropical Africa. Studies in the Gramineae: XXXVIII. Kew Bull 30:101–105 Clayton WD, Renvoize SA (1986) Genera Graminum: grasses of the world. Royal Botanical Garden, Kew, London, UK Dahmer N, Schifino-Wittmann MT, Dall’Agnol M, de Castro B (2008) Cytogenetic data for Paspalum notatum Flugge accessions. Sci Agric 65:381–388 Darlington CD (1939) The evolution of genetic systems. Cambridge University Press, Cambridge, UK Darlington CD, Wylie AP (1961) Chromosome atlas of flowering plants, 2nd edn. George Allen & Unwin, London, UK Daurelio LD, Espinoza F, Quarin CL, Pessino SC (2004) Genetic diversity in sexual diploid and apomictic tetraploid populations of Paspalum notatum situated in sympatry or allopatry. Plant Syst Evol 244:189–199 Davidse G, Pohl RW (1974) Chromosome numbers, meiotic behaviour, and notes on tropical American grasses (Gramineae). Can J Bot 52:317–328 Denham SS (2005) Revision sistematica del subgenero Harpostachys de Paspalum (Poaceae: Panicoideae: Paniceae). Ann MO Bot Gard 92:463–532 Denham SS, Zuloaga FO, Morrone O (2002) Systematic revision and phylogeny of Paspalum subgenus Ceresia (Poaceae: Panicoideae: Paniceae). Ann MO Bot Gard 89:337–399 Dudeck AE, Peacock CH (1985) Effects of salinity on seashore paspalum turfgrasses. Agron J 77:47–50 Duvall MR, Noll JD, Minn AH (2001) Phylogenetics of Paniceae (Poaceae). Am J Bot 88:1988–1992

W.M. Williams et al. Espinoza F, Quarin CL (2000) 2n þ n hybridization of apomictic Paspalum dilatatum with diploid Paspalum species. Int J Plant Sci 161:221–225 Espinoza F, Urbani MH, Martinez EJ, Quarin CL (2001) The breeding system of three Paspalum species with forage potential. Trop Grassl 35:211–217 Essi L, Souza-Chies TT (2007) Phylogeny of Linearia and Notata groups of Paspalum L. (Poaceae, Panicoideae, Paniceae) and related species. Genet Resour Crop Evol 54:779–791 Ferreira TF, Souza-Chies TT (2005) Genetic diversity among Paspalum L. species (Poaceae) belonging to the Notata and Linearia groups based on restriction fragment length polymorphism analyses. Genetica 125:133–140 Filgueiras TS (1992) Native forage grass species from the Federal District, Brazil. Pesqui Agropec Bras 27:1103–1111 Forbes I Jr, Burton GW (1961) Cytology of diploids, natural and induced tetraploids, and intraspecies hybrids of bahiagrass, Paspalum notatum Flugge. Crop Sci 1:402–406 Garcia MV, Balatti PA, Arturi MJ (2007) Genetic variability in natural populations of Paspalum dilatatum Poir. analyzed by means of morphological traits and molecular markers. Genet Resour Crop Evol 54:935–946 Hojsgaard D, Schegg E, Valls JPM, Martinez EJ, Quarin CL (2008) Sexuality, apomixis, ploidy levels, and genomic relationships among four Paspalum species of the subgenus Anachyris (Poaceae). Flora (Jena) 203:535–547 Hojsgaard D, Honfi AI, Rua G, Davina J (2009) Chromosome numbers and ploidy levels of Paspalum species from subtropical South America (Poaceae). Genet Resour Crop Evol 56:533–545 Honfi AI, Quarin CL, Valls JFM (1990) Karyological studies in South American grasses. Darwiniana 30:87–94 Huang B, Duncan RR, Carrow RN (1997a) Drought-resistance mechanisms of seven warm-season turfgrasses under surface soil drying: I. Shoot response. Crop Sci 37:1858–1863 Huang B, Duncan RR, Carrow RN (1997b) Drought-resistance mechanisms of seven warm-season turfgrasses under surface soil drying: II. Root aspects. Crop Sci 37:1863–1869 Hunziker JH, Zuloaga FO, Morrone O, Escobar A (1998) Estudios cromosomicos en Paniceae Sudamericanas. (Poaceae: Panicoideae). Darwiniana 35:29–36 Jarret RL, Ozias-Akins P, Phatak S, Nadimpalli R, Duncan R, Hiliard S (1995) DNA contents in Paspalum spp. determined by flow cytometry. Genet Resour Crop Evol 42:237–242 Jarret RL, Liu ZW, Webster RW (1998) Genetic diversity among Paspalum spp. as determined by RFLPs. Euphytica 104:119–125 Jiang YW, Duncan RR, Carrow RN (2004) Assessment of low light tolerance of seashore paspalum and bermudagrass. Crop Sci 44:587–594 Kalmbacher RS, Martin FG (1999) Comparative yield and nutritive value of atra paspalum and dwarf elephantgrass. In: 58th Annual meeting on soil and crop science society of Florida, Daytona Beach, Florida, USA, pp 97–102 Kalmbacher RS, West SH, Martin FG (1999) Seed dormancy and aging in atra paspalum. Crop Sci 39:1847–1852 Kretschmer AE, Kalmbacher RS, Wilson TC (1994) Preliminary evaluation of Paspalum atratum Swallen (atra paspalum) – a high-quality, seed-producing perennial forage grass for Florida. Soil Crop Sci Soc Florida Proc 53:60–63

12 Paspalum Lan CY, Chen G, Li L, Wong MH (1992) Use of cattails in treating wastewater from a Pb/Zn mine. Environ Manage 16:75–80 Lee GJ, Carrow RN, Duncan RR (2005) Criteria for assessing salinity tolerance of the halophytic turfgrass seashore paspalum. Crop Sci 45:251–258 Lee GJ, Carrow RN, Duncan RR, Eiteman MA, Rieger MW (2008) Synthesis of organic osmolytes and salt tolerance mechanisms in Paspalum vaginatum. Environ Exp Bot 63:19–27 Machado ACC, Valls JFM, Penaloza APS, dos Santos S (2005) New pentaploid biotypes of the Dilatata group of Paspalum L. (Gramineae) from Southern Brazil. Ciencia Rural 35:56–61 Martinez EJ, Espinoza F, Quarin CL (1994) BIII progeny (2n þ n) from apomictic Paspalum notatum obtained through early pollination. J Hered 85:295–297 Martinez EJ, Urbani MH, Quarin CL, Ortiz JPA (2001) Inheritance of apospory in bahiagrass, Paspalum notatum. Hereditas 135:19–25 Martinez EJ, Hopp HE, Stein J, Ortiz JPA, Quarin CL (2003) Genetic characterization of apospory in tetraploid Paspalum notatum based on the identification of linked molecular markers. Mol Breed 12:319–327 Martinez EJ, Acuna CA, Hojsgaard DH, Tcach MA, Quarin CL (2007) Segregation for sexual seed production in Paspalum as directed by male gametes of apomictic triploid plants. Ann Bot 100:1239–1247 Miz RB, Souza-Chies TT (2006) Genetic relationships and variation among biotypes of dallisgrass (Paspalum dilatatum Poir.) and related species using random amplified polymorphic DNA markers. Genet Resour Crop Evol 53:541–552 Moraes Fernandes MIB, Barreto IL, Salzano FM (1968) Cytogenetic, ecologic and morphologic studies in Brazilian forms of Paspalum dilatatum. Can J Genet Cytol 10:131–138 Morrone O, Zuloaga FO, Carbono E (1995) Revision of group Racemosa of the genus Paspalum (Poaceae: Panicoideae: Paniceae). Ann MO Bot Gard 82:82–116 Morrone O, Vega AS, Zuloaga FO (1996) Revision del grupo Dissecta (s. str.) del genero Paspalum (Poaceae, Panicoideae, Paniceae). Candollea 51:2–34 Morrone O, Denham SS, Aliscioni SS, Zuloaga FO (2000) Revision del grupo Anachyris del genero Paspalum (Poaceae, Panicoideae, Paniceae). Candollea 55:105–155 Morrone O, Denham SS, Zuloaga FO (2004) Taxonomic revision of the genus Paspalum Eriantha group (Poaceae, Panicoideae, Paniceae). Ann MO Bot Gard 91:225–246 Norrmann GA, Quarin CL, Burson BL (1989) Cytogenetics and reproductive behavior of different chromosome races in six Paspalum species. J Hered 80:24–28 Oram RN (1990) Register of Australian herbage plant cultivars. CSIRO, Parchment, Melbourne, Australia Ortiz JPA, Pessino SC, Leblanc O, Hayward MD, Quarin CL (1997) Genetic fingerprinting for determining the mode of reproduction in Paspalum notatum, a subtropical apomictic forage grass. Theor Appl Genet 95:850–856 Ortiz JPA, Pessino SC, Vishnu B, Hayward MD, Quarin CL (2001) A genetic linkage map of diploid Paspalum notatum. Crop Sci 41:823–830 Pagliarini MS, Carraro LR, Freitas PM, Adamowski EV, Batista LAR, Valls JFM (2001) Cytogenetic characterization of Brazilian Paspalum accessions. Hereditas 135:27–34

215 Penaloza APS, Cortes AL, Pozzobon MT, Santos S, Rua GH, Valls JFM (2008) Unusual chromosome numbers in Paspalum L. (Poaceae: Paniceae) from Brazil. Genet Mol Res 7:399–406 Perelman SB, Burkart SE, Leon RJC (2003) The role of a native tussock grass (Paspalum quadrifarium Lam.) in structuring plant communities in the Flooding Pampa grasslands, Argentina. Biodivers Conserv 12:225–238 Pessino SC, Espinoza F, Martinez EJ, Ortiz JPA, Valle EM, Quarin CL (2001) Isolation of cDNA clones differentially expressed in flowers of apomictic and sexual Paspalum notatum. Hereditas 134:35–42 Pillay M, Mazzella C (1997) Chloroplast genome differences between Paspalum dilatatum Poir and the related species P. notatum Flugge. Theor Appl Genet 95:696–701 Pozzobon MT, Valls JFM (2000) Cytogeography and variation of stomatal size of Paspalum glaucescens (Gramineae; Paniceae) in southern Brazil. Euphytica 116:251–256 Pozzobon MT, Valls JFM (2003) Chromosome number in Brazilian germplasm accessions of Paspalum hydrophilum, P. modestum and P. palustre (Gramineae; Paniceae). Genet Mol Biol 26:365–368 Pozzobon MT, Valls JFM, dos Santos S (2000) Chromosome counts in Brazilian species of Paspalum L. (Gramineae). Acta Bot Bras 14:151–162 Pozzobon MT, Machado ACC, Vaio M, Valls JFM, Penaloza APS, dos Santos S, Cortes AL, Rua GH (2008) Cytogenetic analyses in Paspalum L. reveal new diploid species and accessions. Cienc Rural 38:1292–1299 Pritchard AJ (1962) The cytology and reproduction of Paspalum yaguaronense (Henr.). Aust J Agric Econ 13:206–211 Pupilli F, Caceres ME, Quarin CL, Arcioni S (1997) Segregation analysis of RFLP markers reveals a tetrasomic inheritance in apomictic Paspalum simplex. Genome 40:822–828 Pupilli F, Labombarda P, Caceres ME, Quarin CL, Arcioni S (2001) The chromosome segment related to apomixis in Paspalum simplex is homoeologous to the telomeric region of the long arm of rice chromosome 12. Mol Breed 8:53–61 Quarin CL (1986) Seasonal changes in the incidence of apomixis of diploid, triploid, and tetraploid plants of Paspalum cromyorrhizon. Euphytica 35:515–522 Quarin CL (1992) The nature of apomixis and its origin in Panicoid grasses. Apomixis Newsl 5:7–15 Quarin CL, Burson BL (1983) Cytogenetic relations among Paspalum notatum var. saurae, P. pumilum, P. indecorum, and P. vaginatum. Bot Gaz 144:433–438 Quarin CL, Caponio I (1995) Cytogenetics and reproduction of Paspalum dasypleurum and its hybrids with P. urvellei and P. dilatatum ssp. flavescens. Int J Plant Sci 156:232–235 Quarin CL, Hanna WW (1980) Effect of three ploidy levels on meiosis and mode of reproduction in Paspalum hexastachyum. Crop Sci 20:69–75 Quarin CL, Norrmann GA (1987) Cytology and reproductive behavior of Paspalum equitans, P. ionanthum, and their hybrids with diploid and tetraploid cytotypes of P. cromyorrhizon. Bot Gaz 148:386–391 Quarin CL, Norrmann GA (1990) Interspecific hybrids between five Paspalum species. Bot Gaz 151:366–369 Quarin CL, Burson BL, Burton GW (1984) Cytology of intraand interspecific hybrids between two cytotypes of Paspalum notatum and P. cromyorrhizon. Bot Gaz 145:420–426

216 Quarin CL, Pozzobon MT, Valls JFM (1996) Cytology and reproductive behavior of diploid, tetraploid and hexaploid germplasm accessions of a wild forage grass: Paspalum compressifolium. Euphytica 90:345–349 Quarin CL, Valls JFM, Urbani MH (1997) Cytological and reproductive behaviour of Paspalum atratum, a promising forage grass for the tropics. Trop Grassl 31:114–116 Quarin CL, Norrmann GA, Espinoza F (1998) Evidence for autoploidy in apomictic Paspalum rufum. Hereditas 129: 119–124 Quarin CL, Espinoza F, Martinez EJ, Pessino SC, Bovo OA (2001) A rise of ploidy level induces the expression of apomixis in Paspalum notatum. Sex Plant Reprod 13:243–249 Quarin CL, Urbani MH, Blount AR, Martinez EJ, Hack CM, Burton GW, Quesenberry KH (2003) Registration of Q4188 and Q4205, sexual tetraploid germplasm lines of bahiagrass. Crop Sci 43:745–746 Reeder JR (1984) Chromosome number reports LXXXII. Taxon 33:126–164 Renvoize SA (1972) Studies in the Gramineae XXX. Kew Bull 27:451–455 Selva SB (1976) Some preliminary cytological observations on a new basic number in Paspalum convexum (Gramineae). Can J Bot 54:385–394 Shaw NH, Elich TW, Haydock KP, Waite RB (1965) A comparison of seventeen introductions of Paspalum species and naturalized P. dilatatum under cutting at Samford, southeastern Queensland. Aust J Exp Agric Anim Husb 5:423–432 Shonubi OO, Okusanya OT (2007) The growth and physiological responses of Paspalum vaginatum S.W. and Paspalum scrobiculatum Linn. in relation to salinity. Asian J Plant Sci 6:949–956 Souza-Chies TT, Essi L, Rua GH, Valls JFM, Miz RB (2006) A preliminary approach to the phylogeny of the genus Paspalum (Poaceae). Genetica 126:15–32 Speranza P (2005) The challenges of the exploration of genetic resources in apomictic plants: lessons from Paspalum dilatatum. V Simposio de Recursos Geneticos para Americana Latina y el Caribe, Montevideo, Uruguay, pp 73–76 Speranza P, Malosetti M (2007) Nuclear and cytoplasmic microsatellite markers for the species of the Dilatata group of Paspalum (Poaceae). Plant Genet Resour Charact Util 5:14–26 Speranza P, Vaio M, Mazzella C (2003) Karyotypes of two cytotypes of Paspalum quadrifarium Lam. (Poaceae). An alternative technique for small chromosomes in plants. Genet Mol Biol 26:499–503 Stein J, Quarin CL, Martinez EJ, Pessino SC, Ortiz JPA (2004) Tetraploid races of Paspalum notatum show polysomic inheritance and preferential chromosome pairing around the apospory-controlling locus. Theor Appl Genet 109:186–191 Takayama SY, Freitas PM, Pagliarini MS, Batista LAR (1998) Chromosome number in germplasm accessions of Paspalum

W.M. Williams et al. (Plicatula group) from different regions in Brazil. Euphytica 99:89–94 Tischler CR, Burson BL (1995) Evaluating different bahiagrass cytotypes for heat tolerance and leaf epicuticular wax content. Euphytica 84:229–235 Trenholm LE, Duncan RR, Carrow RN (1999) Wear tolerance, shoot performance, and spectral reflectance of seashore paspalum and bermudagrass. Crop Sci 39:1147–1152 Trenholm LE, Carrow RN, Duncan RR (2000) Mechanisms of wear tolerance in seashore paspalum and bermudagrass. Crop Sci 40:1350–1357 Vaio M, Speranza P, Valls JF, Guerra M, Mazzella C (2005) Localization of the 5S and 45S rDNA sites and cpDNA sequence analysis in species of the Quadrifaria group of Paspalum (Poaceae, Paniceae). Ann Bot 96:191–200 Vaio M, Mazzella C, Porro V, Speranza P, Lopez-Carro B, Estramil E, Folle GA (2007) Nuclear DNA content in allopolyploid species and synthetic hybrids in the grass genus Paspalum. Plant Syst Evol 265:109–121 Valls JFM (2000) Impacto do conhecimento citogenetico na taxonomia de Paspalum e Axonopus, Gramineae. Topicos atuais em botanica: palestras convidadas do 51st Congresso Nacional de Botanica, Embrapa Recursos Geneticos e Biotecnologia, Brasilia, pp 57–60 Valls JFM, Pozzobon MT (1987) Variacao apresentada pelos principais grupos taxonomicos de Paspalum com interesse forrageiro no Brazil. Encontro Internacional Sobre Melhoramento Genetico de Paspalum, Nova Odessa, IZ Venuto BC, Burson BL, Hussey MA, Redfearn DD, Wyatt WE, Brown LP (2003) Forage yield, nutritive value, and grazing tolerance of dallisgrass biotypes. Crop Sci 43: 295–301 Vikrant RA (2002) Somatic embryogenesis from immature and mature embryos of a minor millet Paspalum scrobiculatum L. Plant Cell Tiss Org Cult 69:71–77 Vishnu-Mittre (1978) Origins and history of agriculture in the Indian sub-continent. J Hum Evol 7:31–36 Watson L, Dallwitz MJ (1992) The grass genera of the world. CAB International, Wallingford, Oxford, UK Yu YH, Feng WS, Shen YF, Ye ZH, Wong MH (2005) Use of microbial community to evaluate performance of a wetland system in treating Pb/Zn mine drainage. Environ Manage 36:842–848 Zuloaga FO, Morrone O (2003) Paspalum. III. Subfamilies Panicoideae, Aristidoideae, Arundinoideae and Danthonioideae. Contrib US Nat Herb, Catalogue of New World Grasses (Poaceae), pp 554–557 Zuloaga FO, Morrone O (2005) Revision de las especies de Paspalum para America del Sur austral (Argentina, Bolivia, Southern Brazil, Chile, Paraguay and Uruguay) Zuloaga FO, Pensiero J, Morrone O (2004) Systematics of Paspalum grupo Notata (Poaceae-Panicoideae-Paniceae). Syst Bot Monogr 71:1–75

Chapter 13

Pennisetum Thierry Robert, Nadra Khalfallah, Evelyne Martel, Franc¸oise Lamy, Valerie Poncet, Cle´mentine Allinne, Marie-Stanislas Remigereau, Samah Rekima, Magalie Leveugle, Ghayas Lakis, Sonja Siljak-Yakovlev, and Aboubakry Sarr

13.1 Structure and Evolutionary Relationships Within Pennisetum Complex of Species 13.1.1 Systematic Considerations of the Genus The genus Pennisetum belongs to the family Poaceae, group Paniceae of the subfamily Panicoideae and is closely related to the genera Cenchrus and Setaria. It includes approximately 140 species (Brunken 1977) distributed in tropical and subtropical regions. The genus Pennisetum has been given various taxonomic status (see Jauhar 1981 for a review). The Stapf and Hubbard (1934) classification of Pennisetum based on morphological characters divides the genus into five sections: Penicillaria, Brevivalvula, Gymnothrix, Heterostachya, and Eu-Pennisetum. Other infrageneric classifications showing some discrepancies were proposed by different authors (Pilger 1940; Brunken 1977; Clayton and Renvoize 1986). The section Pennicillaria includes the domesticated crop Pennisetum glaucum (pearl millet), which is a very important cereal crop in Africa and India but serves also as fodder in other areas. The other species are of multiple purpose uses (forage, pasture grasses, or ornamentals, Table 13.1 and Fig. 13.1). A gene pool classification based on the level of reproductive isolation of the pearl A. Sarr (*) Laboratoire Ecologie Syste´matique et Evolution UMR (8079) Universite´ Parsi-Sud, CNRS, AgroParisTech, Universite´ ParisSud, B 360 91 405 Orsay, France e-mail: [email protected] This chapter has been contributed by lead contributors, Thierry Robert, Valerie Poncet, and Aboubakry Sarr.

millet with its wild relatives was proposed by Harlan and de Wet (1971). Later Perne`s (1984) extended this gene pool classification to the concept of “complex of species”. This taxonomic representation is based not only on the crossability of wild species with the domesticated form but also on the amount of gene flow occurring between all the members of the gene pool. This classification includes three gene pools. The primary gene pool consists of the domesticated form (P. glaucum ssp. glaucum), the wild annual and weedy forms (P. glaucum ssp. monodii represented by two ecotypes: P. violaceum and P. mollissimum). These wild forms cross easily with the cultivated form to produce viable seeds and fertile hybrids. The secondary gene pool includes a perennial fodder wild relative to P. glaucum, P. purpureum, named as Elephant or Niaper grass. Recently, P. squamulatum was included in the secondary gene pool. These two species are easily crossable with pearl millet, but the hybrids are highly sterile. The tertiary gene pool includes true biological species compared to the primary and secondary gene pool species. Indeed, strong reproductive barriers impede natural gene flow and occurrence of hybrids between the members of the tertiary gene pool and the forms belonging to primary and secondary gene pools. The tertiary gene pool includes the majority of the species presented in Table 13.1.

13.1.2 Chromosome Evolution The species of the Pennisetum genus present different basic chromosome numbers x ¼ 5, 7, 8, or 9 (Jauhar 1981), variable genome sizes, and ploidy levels ranging from diploid to octoploid.

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_13, # Springer-Verlag Berlin Heidelberg 2011

217

218

T. Robert et al.

Table 13.1 List of 82 Pennisetum species with some characteristics. Kingdom: Plantae, Subkingdom: Tracheobionta, Superdivision: Spermatophyta, Division: Magnoliophyta, Class: Liliopsida, Subclass: Commelinidae, Order: Cyperales, Family: Poaceae, Genus Pennisetum Rich Species

Life cycle type

Geographic range

Chromosome number and ploidy level (x) n 2n

2C value (pg)

Pennisetum advena (Wipff and Veldkamp) P. alopecuroides (L.) Spreng.

P P

– 18

A P P A P P P P P

– – – – – – – – 18

– 18 (2x) 22 – – 36 54 – – 36 – 36

– 1.90

P. annuum Mez P. articulare Trin ex Spreng. P. bambusiforme (Fourn) Hemsl ex B. D. Jacks. P. basedowii Summerh. and C. E. Hubb. P. beckeroides Leeke P. caffrum (Bory) Leeke P. centrasiaticum Tsvelev P. chilense (E. Desv.) B. D. Jacks. ex R. E. Fr. P. clandestinum Hochst. ex Chiov.

P. complanatum (Nees) Hemsley P. crinitum (HBK) Spreng P. divisum (Forssk. ex J. F. Gmel.) Henrard

P P P

– – 18, 27

– – 36

– – –

P. domingense (Spreng. ex Schult.) Spreng. P. durum Beal P. flaccidum Griseb P. foermeranum Leeke P. frutescens Leeke P. glaucifolium Hochst. ex A. Rich. P. glaucum (L.) R. Br P. gracilescens Hochst. P. henryanum F. Br. P. hohenackeri Hochst. ex Steud P. hordeoides (Lam.) Steud P. humile Hochst. ex A. Rich. P. intectum Chase P. lanatum Klotzsch P. latifolium Spreng P. laxior (Clayton) Clayton P. ledermannii Mez P. longissimum S.L. Chen & Y.X. Jin P. longistylum Hochst. ex A. Rich P. macrostachyum (Brongn.) Trin. P. macrourum Trin

P P P P P P A P P P A P P P P A P P P P P

– – 9, 18, 27 – – – 7 – – 9 – – – 18 – – – – – – –

– – – – – – 14 (2x) – – 18 (2x) – – – – 36 – – 54 45 54 36, 54

– – – – – – 4.71 – – 1.69 – – – – – – – – – – –

P. massaicum Stapf P. mezianum Leeke P. mildbraedii Mez P. mollissimum Hochst. P. monostigma Pilg. P. montanum (Griseb.) Hack. P. nervosum (Nees) Trin. P. nodiflorum Franch P. nubicum (Hochst.) K. Schum. ex Engl. P. occidentale Chase P. orientale Rich

P P P A P P P P A P P

North America Asia-temperate, Asia-tropical, Australasia South America Pacific America Australasia Africa Africa Indian ocean Asia-temperate South America Africa, Asia-temperate, Asiatropical, Australasia, Pacific, America, Antarctic. Pacific, North America North America, Mexico Africa, Asia-temperate, Asiatropical South America North America, Mexico. Asia-temperate, Asia-tropical Southern Africa South America Africa, Asia-temperate Wide Africa Pacific Africa, Asia-tropical Africa, Asia-tropical Africa South America Asia-temperate, Asia-tropical Australasia, South America Africa Africa Asia-temperate Africa Asia-tropical, Pacific Africa, Asia-temperate Australasia Africa Africa, Asia-tropical Africa

– – – – – – – – – – 7, 18

32 16, 32 – 14 (2x) 18 32 36, 72 – – – 36 (4x) 27, 56

– 3.01 (4x) – 4.51 – – – – – – 3.77

P. pauperum Nees ex Steud.

P

–

–

–

Africa South America America Africa Africa, Asia-temperate South America Africa, Asia-temperate, Asia-tropical, Pacific, South America South America

– – – – – – – – 2.30

(continued)

13 Pennisetum

219

Table 13.1 (continued) Species

Life cycle type

Geographic range

Chromosome number and ploidy level (x) n 2n

2C value (pg)

P. pedicellatum Trin.

A

Africa, Asia-tropical, Australasia, South America

14, 18, 27

54 (6x)

5.61

P. peruvianum Trin. P. petiolare (Hochst.) Chiov. P. pirottae Chiov. P. polystachion (L.) Schult.

P A P A or P

South America Africa Africa Wide

– – – 27

4.40 – – – 5.66 4.25

P. procerum (Stapf) Clayton P. prolificum Chase P. pseudotriticoides A. Camus P. pumilum Hack. & Engl. P. purpureum Schumach.

P P P P P

Africa North America Africa Africa Wide

– – – – 7, 14

P. qianningense S.L. Zhong P. ramosum (Hochst.) Schweinf. & Asch.

P A

Asia-temperate Africa

– 5

P. rigidum (Griseb.) Hack. P. riparium Hochst. ex A. Rich. P. rupestre Chase P. sagittatum Henrard P. schweinfurthii Pilg. P. setaceum (Forssk.) Chiov.

P P P P A P

– – – – – 18

P. shaanxiense S.L. Chen & Y.X. Jin P. sichuanense S.L. Chen & Y.X. Jin P. sieberianum (Schltdl.) Stapf & C.E. Hubb. P. sphacelatum (Nees) T. Durand & Schinz P. squamulatum Fresen.

P P A P P

South America Africa South America South America Africa Europe, Africa, Asia-temperate Australasia, Pacific Asia-temperate Asia-temperate Africa, Asia-temperate Africa Africa.

P. stramineum Peter P. tempisquense R.W. Pohl P. thulinii S.M. Phillips P. thunbergii Kunth

P P P P

– – – 18

P. trachyphyllum Pilg. P. trisetum Leeke P. tristachyum (Kunth) Spreng. P. uliginosum Hack. P. unisetum (Nees) Benth. P. villosum R. Br. Ex Fresen.

P P P P P P

Africa, Asia-temperate South America Africa Africa, Asia-temperate, Asiatropical, Australasia Africa Africa America Africa Africa, Asia-temperate Cosmopolitan

36 (4x) – – – 54 (6x) 36 (4x) 34, 63 – – – – 28 (4x) 21 36 10 (2x) 20 – – – – 14 (2x) 27 (3x) 54 (6x) – – – 18 54 (6x) 56 (8x) – – – –

A P P

Africa South America Africa, Asia-temperate

– – – – – 36 (4x) 45, 54 14 (2x) – –

– – – – – 3.47

P. violaceum (Lam.) L. Rich. P. weberbaueri P. yemense

– 36 – – – 27 – 7 – –

– – – – 34

– – – – 4.59

4.04 – – – – 4.97 2.78 5.28 – – – – 9.56 – – – –

4.52 – –

P Perennial, A annual Sources: Clayton WD, Harman KT, Williamson H (2006 onwards). GrassBase – The Online World Grass Flora: http://www.kew. org/data/grasses-db.html (accessed 08 March 2009) Index to Plant Chromosome Numbers Data Base, Missouri Botanical Garden-w3TROPICOS: http://mobot.mobot.org/W3T/search/ ipcn.html (accessed 26 Feb 2009) Botanica sistematica Italia: http://www.homolaicus.com/scienza/erbario/utility/botanica_sistematica/hypertext/1524.htm#000000 (accessed 21 April 2009) Bennett MD, Leitch IJ (2005) Angiosperm DNA C-values database (release 4.0), Royal Botanic Gardens, Kew: http://data.kew.org/ cvalues/ (accessed 28 Feb 2009) Fedorov (1969) Editor, Chromosome numbers of flowering plants. Academy of sciences of the USSR, V.L. Komarov Botanical Institute, Nauka, Leningrad

220

Fig. 13.1 Some of the studied Pennisetum species (spike, rachis and spikelets): P. alopecuroides (a), P. hohenackeri (b), P. mollissimum (c), P. mezianum (d), P. orientale

T. Robert et al.

(e), P. pedicelletum (f), P. polystachyon (g), P. purpureum (h), P. ramosum (i), P. schweinfurthii (j), P. setaceum (k), P. squamulatum (l), P. villosum (m) (Original illustrations)

13 Pennisetum

The diversity in basic chromosome number and cytological observation of different chromosome sizes between species, with the larger size for the lower numbers (Jauhar 1981), reflect chromosome repatterning during evolution within the genus. The dysploidy process, which has occurred in many plants, involves structural modification of the karyotype by change in the basic chromosome number during phylogenetic divergence of species (Stebbins 1956; Siljak-Yakovlev 1996). Based on chromosome pairing observations in Pennisetum, an increase in chromosome number from a basic chromosome number of x ¼ 5 has been suggested by Jauhar (1970, 1981) and Rao et al. (1989). However, the frequent occurrence of the basic chromosome number x ¼ 9 in the Pennisetum genus and in closely related genera Cenchrus, Panicum, and Setaria may suggest an alternative scenario to the proposed ancestral basic chromosome number of x ¼ 5. In this section, we present the chromosome evolution within Pennisetum gene pools focusing on basic chromosome number.

13.1.2.1 Species with x ¼ 7 Basic Chromosome Number Cultivated pearl millet, P. glaucum ssp. glaucum, and its wild relative P. glaucum ssp. monodii belong to the Penicillaria section. These species with 2n ¼ 2x ¼ 14 chromosomes (AA genome) are part of the same primary gene pool (Brunken 1977). They are not reproductively isolated and can cross-hybridize when they grow in sympatry, forming fertile hybrids with normal chromosome pairing. Despite this gene flow, several mechanisms such as linkage, gametophytic competition, and phenology (Robert et al. 1991, 1992; Robert and Sarr, 1992) have been shown to contribute to the maintenance of the genetic structure of these species. The cultivated forms and the wild species from which they evolved (Perne`s 1983) have conserved a high level of similarity of their genomes. Cytogenetic investigations of cultivated and wild stocks of this gene pool have revealed high karyotype similarity (Fig. 13.2) with identical relative lengths of each chromosome pair between these species (Khalfallah et al. 1993). The position of secondary constrictions (SC) is, nevertheless, different among cultivars. Indeed, the majority of stocks that have been studied

221

present an SC on terminal position of the seventh pair, while the cultivar Ligui, originating from Chad, exhibits an intercalary SC on the sixth chromosome pair (Fig. 13.2a–d). In situ hybridization analysis revealed that the localization of the tandemly repeated multigene family coding for the 18S–5.8S–25S rRNAs (rDNA) and genes coding for the 5S rRNA (5S rDNA) were similar among members of the primary gene pool (Martel et al. 1996). Within the primary gene pool (P. glaucum, P. molissimum and P. violaceum), rDNA sequences were detected on the telomeric part of the short arm of pair VI and at the nucleolar organizing region (NOR) of the satellited pair (VII). The 5S rDNA was detected in the telomeric position of the short arm of metacentric chromosome pair IV of the three species of the primary gene pool (Fig. 13.3). Analysis of the genome size and base composition of species in the primary gene pool made using flow cytometric techniques has shown that there is a similar amount of DNA per basic chromosome set among these species (Martel et al. 1997). The species with x ¼ 7 have monoploid genome size (1Cx-value) ranging from 2.25 to 2.36 pg and are relatively GC-rich (between 44.9% and 47.2%) compared with the other Pennisetum species analyzed (Table 13.2). Within the genus Pennisetum, P. schweinfurthii is the only other diploid species with x ¼ 7. This species belongs to the tertiary gene pool and exhibits a strong reproductive barrier with the cultivated P. glaucum (Hanna and Dujardin 1986). A small number of artificial hybrids have been obtained between P. schweinfurthii and cultivated P. glaucum, but these are completely sterile (Hanna and Dujardin 1986; Marchais and Tostain 1997). This feature is consistent with the differences in chromosome configuration observed between these species. Their karyotypes differ notably in the relative size of chromosome pair II, which is larger in P. schweinfurthii (Martel et al. 1996). This difference could be attributed to the haploid genome size of 2.49 pg observed for these species, which is the largest in the genus. The rDNA loci have been observed on the telomeric part of the short arm of two satellited pairs of chromosomes: in P. schweinfurthii, pair I (the longest) and pair IV present distinctly different physical locations with P. glaucum. Despite having the same basic chromosome number, these species are separated into different gene pools presenting important differences in their chromosomal organization and locations of their rDNA loci (Fig. 13.3). These observations may

222 Fig. 13.2 Metaphase plate (a) and karyotype (a0 ) of Pennisetum glaucum (Ligui), P. violaceum (b and b0 ) with B-chromosomes (arrows), 23 DB (c and c0 ) and Thiotande (d and d0 )

T. Robert et al.

a

b

10 µm

a’ m

m

m

m

m

m-sc

sm

m

m

m

m

m

m

sm-sat

b’ Bs

d

c

c’

m

m

m

m

m

m

sm-sat

m

m

m

m

m

sm-sat

d’ m

show that chromosome rearrangements, such as translocations, may have occurred during the evolution of these species, which is consistent with the existence of reproductive barriers occurring in these gene pools. The forage crop “elephant grass,” P. purpureum, has the same basic chromosome number x ¼ 7 and has been considered until now as the only species in the secondary gene pool. This allotetraploid species (2n ¼ 4x ¼ 28) with A0 A0 BB genomes show strong reproductive barriers with P. glaucum, but some partially sterile male and female triploid hybrids occur in

natural conditions. The A0 genome of P. purpureum has been shown to be highly homeologous to the A genome of P. glaucum (Jauhar 1968, 1981), while the donor of the B genome is not known. Meiotic analysis of these hybrids reveals formation of seven bivalents, where a maximum of five are the heteromorphic association of chromosomes belonging to each parent. There are also seven univalent chromosomes that display irregular segregation causing the formation of aneuploid gametes (Jauhar 1981; Techio et al. 2006). Despite their differences in ploidy level, P. purpureum

13 Pennisetum

223

Fig. 13.3 Idiograms showing the number and position of rDNA ribosomal genes 18S–5.8S–26S and 5S in four Pennisetum species: P. violaceum, P. mollissimum, P. glaucum cv. “Massu,” and P. schweinfurthii

P. violaceum

P. mollissimum

P. glaucum cv. “Massue”

P. schweinfurthii

18S-5.8S-26S

and P. glaucum have almost the same genome size (2C ¼ 4.59 pg and 4.71 pg, respectively) but quite different monoploid genome size (1Cx-value), 1.15 pg and 2.35 pg, respectively. The chromosomes of P. purpureum (mean chromosome size of 80 Mbp) are half the size of the chromosome of P. glaucum (mean chromosome size of 165 Mbp) showing important chromosomal changes linked to divergence of these species. Another Pennisetum species may be considered to have x ¼ 7 as its basic chromosome number. On cytological observations, P. squamulatum was initially reported as an allohexaploid species with x ¼ 9 and 2n ¼ 6x ¼ 54 chromosomes (Raman et al. 1959; Patil et al. 1961). However, despite the multiple reports of 54 chromosomes in P. squamulatum, recent studies reveal 56 chromosomes (Goel et al. 2003). Using a fluorescent in situ hybridization method with centromeric and rDNA probes, Akiyama et al. (2006) observed 56 centromeric signals and detected eight chromosomes with rDNA loci in different accessions of P. squamulatum. These results tend to confirm the octoploid nature of this species (Akiyama et al. 2006) with 2n ¼ 8x ¼ 56 chromosomes and a basic number of x ¼ 7. According to this data, the genome size of the species, with 9.56 pg/2C, has 1.19 pg DNA per mono-

5S

ploid genome, which is similar to P. purpureum. This observation is consistent with the hybrid zygotes obtained by crossing tetraploid pearl millet with P. squamulatum (Marchais and Tostain 1997) and the success recombining germplasms from P. glaucum (pearl millet), P. purpureum, and P. squamulatum (Dujardin and Hanna 1984). According to the evidence detailed above, P. squamulatum and P. schweinfurthii are the only tertiary gene pool species with x ¼ 7 and belong to the same morphological section, Heterostachya.

13.1.2.2 Species with x = 5, 8, or 9 as Basic Chromosome Number All other species of the genus Pennisetum present basic chromosome numbers of x ¼ 5, 8, or 9. They are reproductively highly isolated from the pearl millet and belong to the tertiary gene pool. P. ramosum belonging to the Gymnothrix section is the only species with x ¼ 5 and has a 1Cx ¼ 2.02 pg DNA per haploid chromosome set (Martel et al. 1997). P. mezianum is also from Gymnothrix section with x ¼ 8 and presents the lowest monoploid genome

224

T. Robert et al.

Table 13.2 Chromosome number, genome size, and base composition for 15 Pennisetum species representing the taxonomic sections (Stapf et Hub) and gene pools (Harlan and Dewet) classifications. Adapted from Martel et al. (1997) GC 2C Mean Species Section Pool Ploidy level Basic 1C DNA 1CxDNA (chromosome chromosome value (in valueb chromosome (%) value size (in Mbp) formula) (pg) number (x) Mbp) (pg)a 5 P. ramosum Gymnothrix III 2n ¼ 2x ¼ 10 4.04 1,976 2.02 395 43.7 (8m þ 2sm) 7 P. mollissimum Penicillaria I 2n ¼ 2x ¼ 14 4.51 2,205 2.25 315 46.5 P. violaceum Penicillaria I 2n ¼ 2x ¼ 14 4.52 2,210 2.26 316 47.2 (12m þ 2sm-sat) P. glaucum Penicillaria I 2n ¼ 2x ¼ 14 4.71 2,303 2.35 329 44.9 (12m þ 2sm-sat) P. purpureum Penicillaria II 2n ¼ 4x ¼ 28 4.59 2,245 1.15 160 44.3 P. schweinfurthii Heterostachya III 2n ¼ 2x ¼ 14 4.97 2,430 2.48 347 44.0 (12m þ 2sm-sat) P. squamulatum Heterostachya III 2n ¼ 8x ¼ 56 9.56 4,675 1.19 166 44.1 8 P. mezianum Gymnothrix III 2n ¼ 4x ¼ 32 3.01 1,472 0.75 92 43.0 9 P. alopecuroides Gymnothrix III 2n ¼ 2x ¼ 18 1.90 929 0.91 103 43.5 P. hohenackeri Gymnothrix III 2n ¼ 2x ¼ 18 1.69 826 0.84 92 42.7 (18m) P. setaceum Eu-pennisetum III 2n ¼ 3x ¼ 27 2.78 1,359 0.93 101 43.6 2n ¼ 6x ¼ 54 5.28 2,582 0.88 96 43.9 P. orientale Heterostachya III 2n ¼ 4x ¼ 36 3.77 1,844 0.94 102 44.6 P. villosum Eu-pennisetum III 2n ¼ 4x ¼ 36 3.47 1,697 0.87 94 43.6 P. pedicelletum Brevivalvula III 2n ¼ 6x ¼ 54 5.61 2,743 0.94 102 43.7 P. polystachyon Brevivalvula III 2n ¼ 6x ¼ 54 5.66 2,768 0.94 103 44.7 a

1 pg ¼ 978 Mbp (Dolezˇel et al. 2003) 1 Cx-value: DNA content of monoploid genome (Greilhuber et al. 2005)

b

size of the genus: 1Cx ¼ 0.75 pg. Its close phylogenetic and morphological relationship with P. ramosum, which has x ¼ 5, large chromosomes and a large genome size, suggests that major chromosomal repatterning may have occurred during the divergence of P. ramosum, with an increase in the amount of DNA. The other Pennisetum species with the basic chromosome number x ¼ 9 (Table 13.2) have a similar amount of DNA per monoploid genome, independent of their ploidy levels (from 0.85 to 0.95 pg).

13.1.2.3 Phylogeny and Dysploidy The genomic relationships between Pennisetum species have been analyzed using molecular approaches, through variations in chloroplast DNA (Clegg et al. 1984; Renno et al. 2001), mitochondrial DNA (Chowdhury and Smith 1988), or comparisons of repetitive DNA sequences (Ingham et al. 1993). Significant relationships have been noticed between the genomes of P. glaucum and the wild species

P. purpureum and P. squamulatum, suggesting that these three may have a common origin. Close relationships were also observed among species within the Brevivalvula section (Schmelzer and Renno 1999). Basic chromosome number evolution within the genus was recently assessed by molecular phylogenies using the internal transcribed spacer (ITS) sequences of the ribosomal RNA locus (Martel et al. 2004) and chloroplast DNA sequence data (rpl16, trnL-F region; Donadio et al. 2009). The phylogenetic relationships observed are consistent with an ancestral x ¼ 9 small chromosome characters for the genus Pennisetum (Fig. 13.4a). Species with the derived basic chromosome numbers x ¼ 5, 7, or 8 appear in the most recently divergent clades, indicating that the genome structure in Pennisetum might have evolved towards a reduced chromosome number (decreasing dysploidy) and an increased chromosome size (Martel et al. 2004). Therefore, instead of an ascendant dysploidy (Jauhar 1981; Jauhar et al. 2006), we propose a chromosome evolution trend in the genus Pennisetum via a decreasing

13 Pennisetum

225

dysploidy mechanism from a common ancestral progenitor with basic chromosome number x ¼ 9. The evolutionary reconstruction is congruent with the gene pool classification and in agreement with some morphological grouping. Species of the primary gene pool are clustered together with a bootstrap support of 100%. The allotetraploid species P. purpureum representing the secondary gene pool of pearl millet form a highly supported monophyletic clade with

P. glaucum cultivars. The x ¼ 7 character observed in this clade is derived from an ancestral x ¼ 9 status. The species of the tertiary gene pool with x ¼ 7, P. schweinfurthii and P. squamulatum, form a highly supported clade but are not closely related to the other x ¼ 7 species (Fig. 13.4a). This indicates that genomes with x ¼ 7 evolved twice within the Pennisetum genus. The derived species P. mezianum with x ¼ 8 small chromosomes and P. ramosum with x ¼ 5 large

a

Eragrostis dielsii Coix sp. Sorghum nitidum

97 29

Zea mays

100 42

Tripsacum australe Digitaria ciliaris Setaria parviflora a

P. polystachyon P. pedicellatum*

56 8

a a a, s

92 23

84 7

a a s, a

100 11 51 5

63 13

s a

99 9

s s s

100 15 perennial annual / biannual

69 9

s 100 8

s s a

Fig. 13.4 (continued)

P. setaceum

B

E

P. villosum P. ramosum G P. mezianum Cenchrus ciliaris P. hohenackeri G P. alopecuroides P. squamulatum H P. schweinfurthii P. purpureum P. g. ssp. monodii, mollissimum P. g. spp. glaucum cv Massue P. g. spp. glaucum cv Souna P. g. ssp. monodii, violaceum P. orientale

P

H

226

T. Robert et al.

b

Eragrostis dielsii Coix sp. Sorghum nitidum Zea mays Tripsacum australe Digitaria ciliaris Setaria parviflora P. polystachyon P. pedicellatum* P. setaceum

B

E

P. villosum x=5

x=9

x=8

P. ramosum G P. mezianum Cenchrus ciliaris P. hohenackeri

G

P. lopecuroides x=7

P. squamulatum

H

P. schweinfurthii P. purpureum x=7 x=9 increase in genome size

P. g. ssp. monodii, mollissimum P. g. spp. glaucum cv Massue P. g. spp. glaucum cv Souna P. g. ssp. monodii, violaceum P. orientale

P

H

Fig. 13.4 Character evolution traced with ITS phylogeny. The strict consensus tree of the 18 most parsimonious trees from parsimony analysis was generated on ITS rDNA sequences from 16 Pennisetum species and seven outgroups. (a) Sexual expressions optimized onto the consensus tree, with s and a: predominant sexual and apomictic mode of reproduction, respectively. (b) Chromosome number optimized onto the consensus ITS tree. Morphological sections are given: Penicillaria (P), Brevivalvula

(B), Gymnothrix (G), Heterostachya (H), and Eu-Pennisetum (E). Bold numbers above branches are bootstrap values (%) from 1,000 replicates of the PAUP analysis. Bold numbers under branches are the number of characters state changes forming the corresponded consensus branches. Asterisk: Inclusion of P. pedicellatum as sister species of P. polystachyon is given by the ITS2 analysis (bootstrap value ¼ 99%)

chromosomes are closely related in a same clade, having evolved from an ancestral x ¼ 9 status. The evolutive divergence of P. mezianum may be associated with the loss of one chromosome pair while major chromosomal rearrangement may have occurred with an

increase in the amount of DNA during the divergence of P. ramosum (Fig. 13.4b). These molecular phylogenies indicate that the genus Pennisetum is paraphyletic (Martel et al. 2004; Donadio et al. 2009) with the species of Cenchrus

13 Pennisetum

ciliaris nested within it, as previously found by other authors (Go´mez-Martı´nez and Culham 2000; Giussani et al. 2001; Doust and Kellogg 2002; Doust et al. 2007). Close genomic relationships have been observed between the aposporous apomictic species Buffelgrass C. ciliaris (syn. Penissetum ciliare L. Link) and P. squamulatum. These species, used in programs to introgress apomixis from wild relatives into major crops species, showed almost complete macrosynteny at the apospory-specific genomic region (ASGR) governing apomixis in both species (Goel et al. 2006). Species of the genus Cenchrus present basic chromosome numbers of x ¼ 9 and 17, with the x ¼ 17 species occurring in a most recently divergent position. These x ¼ 17 species could be derived from an ancestral tetraploid form with x ¼ 9 (Avdulov 1931) through direct loss of one pair of chromosomes, or by fusion and loss of part of a chromosome containing a centromere (Darlington 1956). The chromosome numbers in C. ciliaris vary from 18 to 56 (Visser et al. 1998). Most genotypes are tetraploid (2n ¼ 4x ¼ 36) with a genome size of approximately 1,500 Mbp/C (Roche et al. 2002), which is roughly the same as the tetraploid species P. mezianum (1,472 Mbp/C; Table 13.2) closely related to C. ciliaris in the ITS phylogeny (Fig. 13.4a, b). These chromosomal and genomic characteristics combined with phylogenetic relationships favor of the inclusion of Cenchrus species within the genus Pennisetum. Accordingly, we propose to reconsider the taxonomic position of Cenchrus species and to rename them as Pennisetum species as they were previously known.

13.1.2.4 B-Chromosomes The occurrence of B-chromosomes (Bs) in Pennisetum typhoides (Burro.) S. & H. (now P. glaucum) was first reported by M€ untzing (1958) and Pantulu (1960). Venkateswarlu and Pantulu (1970) described the cytological behavior of these supernumerary chromosomes. The Bs associate to form bivalents, trivalents, quadrivalents, or multivalents. Their impact on meiosis was outlined by Pantulu and Manga (1975) who found that a reduction of mean chiasma frequency on A chromosomes is observed when the number of Bs reaches a threshold of 4, but that the variance of mean A-chiasma frequency increases with the number of Bs.

227

Different types of Bs were observed by Subba Rao and Pantulu (1978) and Jayalakhsmi and Pantulu (1984): “Standard” Bs, which have no effects on mean A-chiasma frequency but which do affect its variance; “Deficient” Bs, whose presence is associated with a depressing effect; and “Iso-Bs,” which are associated with enhanced phenotypes. They suggested that a balance between the amount of heterochromatin and euchromatin present in the B-chromosomes is responsible for these different effects. Subba Rao (1980) studied the transmission patterns of these three types of B-chromosome, using crosses of a B-chromosome donor line and a male sterile line followed by recurrent backcrossing. The three types of Bs exhibit an accumulation trend in the oosphere, but this trend was not observed in the pollen grain, except for standard Bs. Indeed, trend of B-chromosome elimination is more frequently observed in transmission via the male gametophyte (see Jauhar 1981 for a comprehensive review on B-chromosomes in pearl millet up to the early 1980s). The fascinating behavior, origin, and the transmission patterns of Bs were subsequently studied in crosses involving wild relatives of pearl millet.

Behavior of B-Chromosomes in Confrontations Between “Wild and Domesticated” Genomes B-chromosomes were described by Amouroux-Pezas (1985) in a population of P. violaceum, the wild ancestor of pearl millet, originating from Niger. The behavior of Bs was studied in crosses and recurrent backcrosses of this P. violaceum population with pearl millet lines not carrying Bs (Ligui originating from Chad, Thiotande from Senegal and 23 DB from Tifton, USA). In F1 and segregating backcross (BC) populations, a wide range (0–13 per mitotic cell) of Bchromosomes was observed. These Bs like those described previously are transmissible both through pollen grains and megaspores. A correlation between their occurrence and the variability of some morphological traits was also reported. In a subsequent study using inbred progenies (BC1 first and second generation of selfing: BC1S1, BC1S2), Khalfallah (1990) confirmed the high intra and interindividual variability in B numbers both in mitotic (Fig. 13.5a, b) and meiotic (Fig. 13.6) cells. In meiosis of P. violaceum (wild parent bearing B-chromosomes), the Bs occurred as

228 Fig. 13.5 Eight (a, arrows) and five (b, arrows) Bs in mitotic metaphase cells of hybrid individual (Ligui  violaceum). Presence of Bchromosome in androgenetic doubled haploid plantlets of pearl millet (c and d, arrows). Adapted from Khalfallah et al. (1993) (a and b) and Le Thi et al. (1994) (c and d)

T. Robert et al.

a

b

c

d

10 µm

univalents, bivalents (Fig. 13.6a), trivalents (Fig. 13.6b), and tetravalents (Fig. 13.6c), while in BCS2 progenies (Thiotande  violaceum) above a threshold of 4 Bs frequently formed multivalents composed of 5 or 6 B-chromosomes (Fig. 13.6d–f). The persistence of interchromosomic connections (Fig. 13.6g) between A and Bs chromosomes was also observed. These connections could be due to shared sequences between A- and B-chromosomes acting as end-to-end recognition sites. It is also notable that the behavior of Bs is dependent on the recipient cultivated genotype. Indeed, within Ligui derived families, the mean number of Bs is higher in BC1S1 progenies than in BC1S2 (Fig. 13.7a). In contrast to this trend, no significant difference was observed in mean B numbers between BC1S1 and BC1S2 (Fig. 13.7b) derived from the Thiotande  violaceum cross. This result is consistent with a growing number of findings demonstrating that Bs now cannot be considered only as selfish or parasitic DNA as stated by some ¨ stergren 1947; Kayano 1961; Nur 1977; authors (O Jones 1985; Shaw and Hewitt 1985). Indeed, in rye (Secale cereale), a genotype-dependent B-transmission

rate has led to differing rates (high and low) within accessions (Puertas et al. 1986, 1993). Furthermore, the inheritance of transmission rate of B-chromosomes has been elucidated in both maize (Zea mays subsp. mays) and rye (Chiavarino et al. 1998, 2001; Puertas et al. 1998; Puertas 2002). In maize, the genes involved in Bs transmission rate are located on the A chromosome set. The accumulation of Bs resulting from non-disjunction events during meiosis is ruled by a process involving A chromosome centromeric region activity (Han et al. 2007). In rye, the genes influencing B-chromosome transmission rate are located on the Bs themselves. These genes are also active in the process of chiasma initiation. Accordingly, the differences in transmission rate of Bs in progenies derived from crosses between forms of the P. glaucum primary gene pool (wild relative  domesticated) could reflect a genetic differentiation between pearl millet cultivars. Khalfallah (1990) hypothesized the occurrence of a differential “accumulation/loss” (A/L) mechanism in recipient pearl millet line. The Ligui and 23 DB lines exhibit an early B (A/L) figure, resulting in a reduction of B number from the BC1S1 generation, whereas in the Thiotande line

13 Pennisetum

229

a

c

b

e

d

10 µm

f

g

Fig. 13.6 B-chromosome behavior in meiosis: In P. violaceum, the Bs form two bivalents (a), one trivalent (b), or one tetravalent (c); in the progenies BCS2 (Thiotande  violaceum) B-chromosomes frequently form a star shape multivalent (d, diacinese, arrow). The multivalent composed by five (e) or

six (f) chromosome Bs do not follow the chromosomes A during meiosis and were probably eliminated from cell at the stage of telophase I. Persistence of interchromosomal connections (g, arrow) between A and Bs chromosomes (Original data)

genetic background, this mechanism generates a high number of Bs in late BCS generations. Conversely, the “elimination” component of the process sets in from a threshold number of Bs. Indeed, in the BCS2 progenies (Thiotande  violaceum), B-chromosomes form a visible star shape multivalent in telophase I of meiosis (Fig. 13.6e, f) and are probably eliminated from the cell at this stage. This mechanism, which was shown for the first time by Khalfallah et al. (1988) and Khalfallah (1990), probably underlies the regulation of B number in the confrontation of genomes from wild relatives and domesticated forms. The peculiar expression of the A/L mechanism in the Thiotande line

is consistent with previous genetic abnormalities (male sterility and segregation distortion) observed in the offspring of Wild  Thiotande crosses (Marchais and Perne`s 1985; Sarr and Perne´s 1988; Sarr et al. 1988).

The Effect of B-Chromosomes In pearl millet, the occurrence of B-chromosomes could interfere with plant fertility. Indeed, the drive mechanism described above as “accumulation/loss” induced perturbations of meiosis (Khalfallah 1990). The same effect was observed in many animal and

230

Fig. 13.7 Variation of B-chromosomes within inbred families (BC1 first and second generation of selfing: BC1S1, BC1S2); Ligui  violaceum BC1S1 and BC1S2 (a) Thiotande  violaceum BC1S1 and BC1S2 (b) (Original data)

plant species; in Dactylis (Williams and Barclay 1968) and rye (Jones and Puertas 1993), for example. Amouroux-Pezas (1985) assumed that the presence of Bs could stimulate plant growth at early phenological stages (larger and higher number of leaves), but a depressing effect is observed during heading (narrow flag leaves and reduced productive basal tillers). In the above mentioned study, Khalfallah (1990) described the pattern of morphological trait diversity in segregating progenies derived from crosses between a wild progenitor B donor and domesticated lines empty of Bs (BC1S2 and BC1S3). This analysis (see Fig. 13.8 for BC1S2) revealed high diversity in terms of recombinant phenotypes for a set of 16 morphological characters. The progenies deriving from the line Thiotande as female exhibit phenotypical

T. Robert et al.

diversity and a higher number of Bs. Actually, the global pattern indicate a closer genetic similarity between wild parent and the selfed BC progenies. It is worth mentioning that in BC1S3 progenies, the phenotypical diversity is still high and the recovery of domesticated like progenies is also weak. This pattern is specific to crosses involving P. violaceum (carrying Bs) from Niger as male parent. Indeed, in progenies derived from crosses between P. mollisimum, another wild relative (without Bs), and the same domesticated lines, the recovery of cultivated type in BC progenies is easier (Joly-Ichenhauser 1984; Joly-Ichenhauser and Sarr 1985; Niangado 1981). These original data laid prospects as suggested by Jones et al. (2008) for investigations on genome organization and the dynamics of biodiversity with Bs as linchpin specifically around their supposed origin from interspecific hybridization or genome response to both biotic and abiotic environmental stresses. For example, some specific environments like in vitro tissue culture could induce chromosome abnormalities and the apparition of B-chromosome. Indeed, in doubled haploids plantlets obtained from androgenesis of pearl millet (Fig. 13.5c, d), B-chromosome was observed (Le Thi et al. 1994). This phenomenon was also reported for Nicotiana sylvestris from in vitro culture (De Paepe et al. 1983; Lespinasse et al. 1987).

13.1.3 Domestication Process Hallmarks in the Genus Pennisetum 13.1.3.1 Where, When, and How Pearl Millet (P. glaucum) Was Domesticated? Origin of Pearl Millet Center or Non-Center Domestication of Pearl Millet? In South America, Southeast Asia, and Subsaharian Africa, the high and geographically structured phenotypic diversity that has been observed in numerous crops originated from these areas has led several authors to hypothesize that domestication occurred several times for the same species within large geographical regions, called non-center of domestication (Harlan 1971, 1975). The question whether pearl

13 Pennisetum Fig. 13.8 PCA first plan showing diversity of 4 BC1S2 progenies and parents (wild and domesticated lines) (a); Path discriminant analysis first plan showing high diversity in BC1S2 progenies from the cross Thiotande (Thiotande  violaceum) (b). (Original data)

231

a

b

millet, which had been originated from Africa, was domesticated or not according to a non-center model is still debated. This review proposes to enlighten this question in regards to old and recent data researches in archeology, paleobotanic, and genetics. The Beginning of Agriculture in Africa and Pearl Millet Domestication Although pearl millet (P. glaucum ssp. glaucum) is cultivated in every arid regions of Africa and Southwest Asia, its wild closest relative P. glaucum ssp. monodii has never been observed outside of the Sahelian region where the phenotypic diversity of pearl millet is the highest. This is one of the strong arguments favoring the hypothesis of its domestication

in this area of Africa. Classifications of pearl millet landraces has been defined on the basis of the size and the shape of candels and the color and the shape of seeds (Porte`res 1950; Brunken 1977). Because the distribution of the main pearl millet types shows a clear geographic pattern, it has been supposed that the domestication of this crop has occurred several times along the Sahelian region of Africa (Porte`res 1950; Harlan 1975; Brunken et al. 1977). This hypothetical domestication non-center overlaps the actual distribution of the wild millets (P. glaucum ssp. monodii). The humid climate is not propitious to the conservation of carbonized remains. Conversely, the conditions prevailing in the Sahara desert prevented the formation of humus necessary to trap these remains to protect from oxidation, which have led to

232

a major erosion. The reconstruction of an archeological sequence in relation to the domestication is complicated by the scarcity of carbon remains in spite of the intensification of excavations in recent decades. However, a general pattern emerges, which may question the assumption of multiple domestication for pearl millet. Plant and animal domestication has been a central element of the Neolithic revolution during which agriculture was invented. Farming in Africa probably started in the area, which corresponds to the actual Sahara desert. Indeed, around 18000 BP, the establishment of an intertropical rainy season has erased the hyper-arid desert and provides the opportunity for a wooded savannah to develop in the south part and a more Sahelian vegetation in the North (Adams 1997). A dense hydrographic network of rivers and lakes was established, around which the density of human occupation has continued to increase until the middle of the Neolithic, through the colonization of the Sahara by people from South (Camps 1974; Petit-Maire 2002). However, the climate was dry enough, locally variable, and sometimes caused droughts particularly in the eastern regions where the Atlantic monsoon was short (Wendorf et al. 1992). The climate became progressively degraded from 6000 BP, and the formation of the Sahara was achieved around 3500 BP. Until 8000 BP, Saharian people were exclusively hunter-gatherer. Ceramics was used from 9000 BP in Niger, namely 2,000 years earlier than in the Middle East according to the current state of knowledge (Fig. 13.9). Harvesting of wild grasses, especially Panicum, Digitaria, Pennisetum, and Sorghum species, was intensive as witnessed by the numerous lithic tools found in archeological investigations of this period and by the remains of seeds conserved in tanks used as granaries and dated as early as 9000 BP (at Nabta Playa) (Wendorf et al. 1992; Wendorf and Schild 1998). Archeological reconstructions show that agrarian societies using the Neolithic way of life existed on the African continent since at least the eighth millennium BC. The first signs of bovine domestication (Bos Taurus) found in the west of Egypt and then in the Central Sahara are dated from the same period (Wendorf and Schild 1994; Fig. 13.9). From the seventh millennium BP, the palynologic studies from this period indicate a rapidly changing landscape due not only to pastoral practices but also clearing, which is usually linked to

T. Robert et al.

crop farming (Camps 1974; Ballouche and Neumann 1995). They are consistent with the abundant lithic tools, and footprints of pearl millet seeds that are larger in the domestic form than in the wild Pennisetum, found in two sites dated to 6000 BP (Camps 1974; Klichowska 1978; Fig. 13.9). However, these indications of the existence of domestic Pennisetum are still very tenuous and indirect. However, it seems likely that when the period begins to be more arid in Sahara, people who migrated and settled on the banks along the Nile and in the Sahel are likely to have started agriculture. Probably, pearl millet cultivation was already adopted all along the southern edge of the Sahara, from Mauritania to Sudan, as early as 3500 BP. Fields and granary were numerous at Dhar Tichitt and Dhar Oualata in the actual Mauritania, where among the most old domestic pearl millet remains have been found (Amblard and Perne`s 1989). The most old pearl millet seeds, found in northern Togo, have also been dated to 3500 BP (Birimi) (Fig. 13.9; D’Andrea et al. 2001; D’Andrea and Casey 2002). The Metallurgy was already known in central Sahara as early as 3500 BP and has spread after that to the whole of West Africa, therefore, contributing to an improvement of agricultural practices and to the development of Bantu’s people in the region corresponding to actual Nigeria and North Camerun and to their demographic and geographic expansion (Cavalli-Sforza et al. 1994; Bellwood 2001). However, up to date, no trace of undoubtedly domestic forms of pearl millet have been found in the former sedimentary strata in these sites except on the shores of Lake Chad where pearl millet is nevertheless abundant from 3000 BP (Zach and Klee 2003; Klee et al. 2004). This gap in the sedimentary sequence indicates that the domestication of millet could have taken place elsewhere, perhaps further north under the pressure of the arid conditions, where wild populations were in number (Wetterstrom 1993). Because pearl millet was not introduced as a crop in Egypt by eastern Saharian people who had however migrated there at this period, one could imagine that pearl millet domestication was restricted to the west-center region of the Sahara. The fact that pearl millet was already grown as a crop in India as early as 3000 BP (Fuller et al. 2004) is also an argument in favor of an older domestication of this species. Indeed, the geographic distribution of the wild form of P. glaucum is restricted to the Sahelian

13 Pennisetum

233

Fig. 13.9 Archeological evidences of domestication in Subsaharian Africa

region of Africa so that the presence of this crop species in Asia is the result of a very ancient diffusion outside Africa. It is obviously impossible to assess whether pearl millet was domesticated once or several times from archeological data only. So, genetic tools have also been used with the hope to shed light on the domestication history of pearl millet. Contribution of the Genetic Tools in Order to Enlighten the History of Pearl Millet Domestication Huge collections of domestic and wild forms of pearl millet from the whole area of cultivation of this crop have been analyzed for the allozyme polymorphisms (Sandmeier et al. 1981; Pilate-Andre´ 1992; Tostain 1993). These studies have shown that the genetic diversity was similar in the forms of pearl millet and that the genetic diversity of wild millets displays a clear East/West geographic pattern in the Sahelian

region (Tostain et al. 1987; Tostain and Marchais 1989; Tostain 1992, 1993). However, genetic differentiation between domestic and wild pearl millet could be under-evaluated with this kind of molecular marker since isoenzyme polymorphism is generally low. More recent data have been obtained with microsatellite markers, which are much more polymorphic on a sample representative of the Sahelian and Soudano-Sahelian region (Oumar et al. 2008). The results have shown that only 74% of the genetic diversity of the wild millets is found in the domestic groups. This suggests that the domestication of pearl millet has led to a decrease in the effective population size and thus to a bottleneck effect. Nevertheless, this decrease in pearl millet genetic diversity due to the domestication may be stronger than suggested by this data. The gene flow between domestic and wild forms could later have contributed to the increase of the genetic diversity of the domestic gene pool. The simple sequence repeat (SSR) polymorphism has also

234

T. Robert et al.

revealed a high genetic differentiation between domestic and wild populations of pearl millet (Oumar et al. 2008) (Fst ¼ 0.10), which was higher than the level of genetic differentiation that has been found between maize and teosinte, its wild ancestor. The analyses of the isoenzyme polymorphism on domestic and wild samples have yielded two main results. Firstly, the varieties from the Subsaharian region, the likely area of domestication of this crop according to archeological data, are clearly distinct from the East African varieties. Moreover, two main regions within the Subsaharian region can be defined on the basis of population differentiation (Fig. 13.10). East Africa cannot be considered as a domestication center on the basis of archeological data and because wild forms of pearl millet are totally absent from this region. Secondly, a more in-depth analysis considering domestic and wild samples from the west part of the Subsaharian region shows that domestic and wild forms from the same area clustered together (Pilate-Andre´ 1992). Genetic distances between populations are dependent on the geographic distance rather than on the taxonomic status. This has been considered as an argument in favor of several events of domestication for this crop in this region according to a non-center model (PilateAndre´ 1992) as proposed by Porte`res (1950) and by Harlan (1975). This is congruent with previous results obtained on West African populations (from Senegal to Niger) (Tostain et al. 1987). Data from restriction frag-

a

ment length polymorphism (RFLP) analyses on the nuclear adh1 gene also showed a parallel pattern in domestic and wild forms of pearl millet (Gepts and Clegg 1989). However, the RFLP analysis on the nuclear rDNA gene revealed a contrasted pattern since the diversity in the domestic sample was extremely low while it was high in the wild sample. According to these data, the domestication events of early-flowering pearl millet could have been followed by a strong secondary diversification in an area located at the west side of the actual Lake Chad (Fig. 13.10). Late flowering, new early-flowering varieties, and pearl millet varieties of East and South Africa and of India could have originated from this secondary center of diversification. Indeed, the low genetic distances between late-flowering varieties and early-flowering varieties from this region argue in favor of this scenario as well as the low genetic diversity of late-flowering varieties relatively to earlyflowering ones (Tostain et al. 1987; Tostain and Marchais 1989; Tostain 1992, 1993). More recent data obtained on the basis of SSR polymorphisms (27 loci) within a sample of domestic and wild pearl millet covering the entire geographic distribution area of this species in Africa and India has shown a different pattern. Bayesian analysis of the SSR polymorphisms, taking into account the estimated proportion of domestic/wild admixture within the domestic sample suggests a single domestication center

b Senegal-wild Senegal-dom. Niger-dom. Niger-dom. Cameroon-dom. Mali-dom. Mali-dom. Burkina-dom. Burkina-wild Niger-wild Niger-wild Mali-wild Mali-wild Cameroon-wild

Fig. 13.10 Isoenzyme polymorphism in pearl millet. (a) analysis based on 14 domestic (mainly early varieties) and wild populations from the west part of the Subsaharian region (15 loci). (dom.: domestic accession) (b) Analysis on accessions from the all areas of pearl millet cultivation in Africa (except Maghreb). These data rather support the hypothesis of several domestication primary events of early varieties in a huge noncenter covering the West Sahel. This first phase would have been

c Sudan-wild Niger-wild West-Chad-wild West-Mali-dom. Centre-Sahel-wild West-Sahel-wild West-Sahel-early-dom.

I

II III

III

Togo-Ghana-early-dom. East-Sahel-dom. West-Sahel-late-dom. India-dom East and South Africa-dom.

I: Domestication II: Diversification III: Diffusion

III

Senegal-dom (off season)

followed by a secondary diversification in a much restricted area (at the west side of the actual Lake Chad) and then a diffusion of varieties to East and South Africa, probably mainly at Bantu’s migration period; (c) more likely hypothesis for pearl millet domestication and diffusion in Africa. (1) a primary domestication (single or multiple events?) in West Africa; (2) center of secondary diversification (late varieties could be originating from this center); (3) diffusion of new early and late varieties

13 Pennisetum

for pearl millet located in a region corresponding to the actual West Niger and Mali. Although no clear consensus about the history of pearl millet domestication finally emerges either from genetic data or from archeological sequences, it seems, however, that the data reported from the most recent studies using very informative molecular markers (SSR markers) are in favor of a pearl millet domestication localized in a restricted region of West Africa. This domestication could have been followed by a spread of domestic populations in the all Subsaharian region. A secondary diversification of domestic populations could have occurred in the eastern part of the Sahel (Tostain et al. 1987; Tostain 1992). Pearl millet from Eastern and Southern Africa and late varieties are probably originated from this center of diversification (Fig. 13.10). Inferences of domestication history from genetic data are, however, to be made cautiously mainly because of gene flow that have occurred between domesticated and wild populations since the beginning of the domestication process and is still operating now (Perne`s 1986). Indeed, since its domestication, pearl millet populations have been spread sometimes far from the area where they have been domesticated. Cultivated populations can, therefore, be found in parapatric situations with other wild populations than their wild ancestors so that introgression of new genetic material can modify the original genetic composition of the domestic gene pool. Another point to be considered is that present domestic populations strongly differ from a morphological point of view, from the wild phenotypes. Numerous traits contribute to the domestication syndrome (cf. Sect. 13.1.3.2). It is very unlikely that all the mutations needed to produce the domestic phenotype observed nowadays were present in a unique wild population. For example, the increase in seed size in domestic pearl millet compared to the wild forms seems to have occurred progressively and lately, far after the beginning of the domestication, according to the archeological data (Fuller 2007). A progressive and slow evolution of the loss of shattering ability in domestic rice, wheat, and barley has also been suggested (Purugganan and Fuller 2009). Moreover, domestic traits often display multifactorial inheritance (see Sect. 13.1.3.2 for the genetic basis of the domestication syndrome in pearl millet). One can, therefore, imagine that several mutations present in wild popula-

235

tions located in remote areas of the original domestication center have contributed significantly to the evolution of these traits in domestic populations. In other words, different regions of the genomes of domestic pearl millets could, therefore, originate from different ancestor wild gene pools and thus they could have different histories. Therefore, the domestication of pearl millet in particular, and perhaps of crops in general, can be viewed as a reticulate evolutionary process in which several wild gene pools have contributed. If this did really occur for pearl millet, the reconstruction of the history of its domestication can hardly be achieved using molecular markers. Yet, the identification of genes involved in the domestication syndrome and the analysis of their nucleotide polymorphism pattern among wild and domestic populations could be a powerful strategy in order to achieve this goal. Dynamics of Spontaneous Wild–Weed–Pearl Millet Complex Non-intensive agrosystems often display remarkable occurrence of wild, weedy, and domestic forms of the same species in sympatric/paraptric situations. Gene flow between domesticated and spontaneous relatives has been considered by some authors to contribute favorably to the evolution of crops in these agrosystems. Indeed, it can lead to the production of new genotypes upon which farmers’ selection can operate to produce enhanced phenotypes (Jarvis and Hodgkin 1999) and to the spread of advantageous alleles (Morjan and Rieseberg 2004). On the other hand, experimental and theoretical population genetic models have shown that strong gene flow can be responsible for a decrease in local adaptation through introgression of genes from non-adapted genotypes (Antonovics and Bradshaw 1970; Slatkin 1987). The impact of gene flow from the wild relatives on crop diversity has been questioned (Wood and Lenne´ 1997), although in some instances, it could produce aggressive weedy forms, as clearly documented in beet (Boudry et al. 1993) and common bean (Papa and Gepts 2003). Areas where domestic species and their spontaneous relatives coexist can, therefore, be viewed as in situ laboratories for the study of the effect of biological factors and of anthropologic practices on the level of gene flow and introgression between these different forms of the same species.

236

Weedy millets (Shibra, N’doul in Haoussa and Ouofof language, respectively), with intermediate domesticated/wild phenotypes are observed in most pearl millet fields in West Africa (Brunken et al. 1977) including those where wild forms are totally absent (Busso et al. 2000). Contrary to the welldomesticated forms, Shibra forms exhibit an intense shattering habit at maturity. Their presence in drills (holes in the ground made by farmers in which they generally sow a few dozens of seeds) decreases the production of good quality seeds (large and easy to thresh). Although farmers consider sometimes Shibra as invasive plants and are willing to reduce their numbers, they do not eliminate them completely. Indeed, in harsh conditions or to bridge the gap before the main harvests, Shibra are harvested and consumed, even though their yield is very low and their seeds very hard to prepare for consumption. Weeding is supposed to discard all the weeds, including Shibra, from the field. However, it is frequent that some Shibra plants escape this elimination so that the density of Shibra in pearl millet fields after weeding is sometimes still very high (more than 30 Shibra/100 m2 could be observed; C. Allinne personal communication). Shibra plants are early maturing plants. Although the flowering of their main tiller is generally very early, their flowering period largely overlaps the flowering time of the domestic millet, either early or even late flowering varieties, because they mostly produce a large amount of non-synchronous axillary flowering tillers. They are fully interfertile with domestic and wild pearl millet. In parapatric situations, weedy plants have been considered as derived from recurrent pollen flow between domesticated and wild populations (Marchais 1994; Busso et al. 2000). Their presence in the fields results from sowing or spontaneous seedlings developed from shattered seeds. In the former case, the seeds come from spikes selected by farmers to produce seeds for the next growing season. Even when farmers are able to select their seeds on plants with the best phenotypes (well domesticated), there is no way to recognize seeds formed from pollination by weedy and wild plants. Obviously, the dynamics of weedy populations is dependent on farmers’ practices (especially seed management and weeding). The relative impact of these practices and gene flow from wild populations on these dynamics would be a key question to address from an evolutionary point of view.

T. Robert et al.

The very huge amounts of Shibra in numerous pearl millet fields in Africa and their presence even in the areas where wild forms do not exist anymore or even have probably never existed raise questions on the real contribution of gene flow from wild populations to domestic millets in order to explain their maintenance and even, in some areas, their expansion. In fact, very few studies have focused on the level of genetic introgression of these weedy plants by genes from wild or domesticated plants (Marchais 1994; Miura and Terauchi 2005; Mariac et al. 2006). The studies of both Marchais (1994) and of Mariac et al. (2006) on the in situ morphological diversity of weedy millets (Shibra) based on the traits defining the domestication syndrome have shown a large continuous phenotypic distribution within pearl millet fields. Among weedy millets, plants showing different levels of combination of traits from domesticated and wild phenotypes are found, ranging from nearly domesticated to nearly wild phenotypes (Mariac et al. 2006). However, Shibra that grow in drills are significantly more similar to domestic plants, which suggests that most of them are indeed developed from seeds sown by farmers, and thus, in most cases, from selected candels at the harvest time (selection on the female contribution to the next generation). The high diversity observed among Shibra plants and the multi-factorial inheritance of the traits defining the domesticated phenotypes (Poncet et al. 2000) suggest that Shibra are developed from recombination events between the genomes of domestic and wild millets in hybrid plants. The marked phenotypic differences between Shibra found in drills and those found outside drills (Mariac et al. 2006) could thus reflect different introgression levels by wild genes at loci involved in “domestication” traits. Investigations carried out in several pearl millet fields in Niger have shown that Shibra phenotypes could represent up to 46% of the plants found in drills. However, bulked segregant analysis (BSA) of progenies obtained from selfing Shibra plants suggests that a unique region of the genome could be involved as a major factor explaining the morphological differentiation between Shibra and plants with a truly domestic phenotype (Miura and Terauchi 2005). These apparently different conclusions are, however, in accordance with data from domestication quantitative trait loci (QTL) mapping in pearl millet (see Sect. 13.1.3.2), since a region on the linkage group (LG) 6 of the pearl millet genetic map harbors a high

13 Pennisetum

number of major domestication QTLs, while mostly minor QTLs were detected on other linkage groups (Poncet et al. 2000). The main question emerging from these data concerns the roles of the different factors controlling the abundance and the maintenance of Shibra forms. The existence of a strong genetic differentiation between domestic and wild pearl millets at the continental level based on neutral polymorphisms has been shown by recent studies on SSR polymorphisms (see above). Moreover, genetic analyses using amplified fragment length polymorphism (AFLP) markers on plants sampled in situations where domestic, Shibra, and wild forms are found in parapatric situations have also shown a high level of genetic differentiation between wild forms on one part and the domestic on the other (Mariac et al. 2006), even if assignment analyses also show the existence of introgression of Shibra plants by genes from the wild. In addition, analyses of the genetic composition of the pollen pools, which contribute to fertilize domestic, weedy, and wild forms found in a parapatric situation, also show a clear genetic differentiation of the latter. The overall picture drawn by these data is in favor of a low occurrence of hybridization between domestic and wild pearl millets despite mechanisms favoring potential gene flow between wild and cultivated (domesticated and weedy plants) populations, such as outcrossing and phenological similarity, as already suggested by Marchais (1994). Pre-zygotic (Robert et al. 1991; Renno et al. 1997) or post-zygotic (Amoukou and Marchais 1993) isolation mechanisms already documented in pearl millet could at least partly explain this mating behavior. Then direct hybridization between wild relatives (P. violaceum) and the domesticated forms could not be considered as the major process explaining the abundance and maintenance of weedy Shibra in the crop fields. Shibra themselves obviously contribute, through pollination, to further introgression of traits characteristic of wild plants, such as shattering or a high tillering habit, into the selected seeds for next sowing. The intermediate phenotypes of Shibra plants, despite their highest genetic similarity with domestic rather than wild millets, could be explained by the dominance of alleles conferring the wild phenotype, as it has been demonstrated in QTL studies in pearl millet (Poncet et al. 2000 and also see Sect. 13.1.3.2). Shibra found outside drills and developed from soil seed bank recruitment are mainly

237

the product of crosses between Shibra or between Shibra and domestic millets. However, selection against dominant phenotypes should lead to a rapid decrease in the frequency of Shibra phenotypes in the absence of gene flow from wild populations. Quantification of the farmers’ selection procedure is necessary to gain insight into the mechanisms underlying Shibra persistence in fields. A sizeable proportion (11.4%) of plants identified by farmers to be domesticated was in fact morphologically closer to Shibra plants found in drills (Mariac et al. 2006). Seeds originated from candels of Shibralike plants can, therefore, be unconsciously added to the seed stock through the harvest of not fully domesticated phenotypes, and then be sown in the following season. In addition, the harsh agricultural conditions in most African regions, where pearl millet is cultivated, lead to chronic deficits in seed production. In Niger, for example, farmers are used to get their seeds not only from their own production but also from neighbors, local sellers, or regional markets, where seeds harvested throughout the region and even other regions of Niger are sold (Allinne et al. 2008). A similar situation has been observed in Rajasthan, India (Vom Brocke et al. 2003), where frequent importation of seed lots by farmers explains the very low differentiation between pearl millet populations cultivated in distant villages. It has demonstrated that seeds sold to farmers on the markets are sometimes developed from non-selected plants, even Shibra. The contribution of external seed sources could, therefore, account for the wide presence of Shibra in drills, which in turn could contribute, through their progenies, to the seed bank and therefore to produce populations of Shibra outside drills. It is, therefore, likely that seed flow due to farmer’s strategy to get seeds is the main factor explaining the spread of Shibra plants in pearl millet fields in Sahel. Quantitative genetic models have shown that population differentiation can be high at the trait (phenotypic) level and low at the allelic frequency level for the majority of QTLs underlying the trait, especially when diversifying selection is strong (Le Corre and Kremer 2003). Phenotypes such as Shibra found outside the drills could be favored by natural selection. This probably occurs since Shibra outside the drills generally have a higher shattering ability than plants found in drills. They should, therefore, contribute more than plants growing in drills to the seed bank. Furthermore, progenies developed from these plants

238

will also tend to have a high shattering ability as this phenotype is dominant in pearl millet (Poncet et al. 1998). Consequently, if this selective advantage of shattering phenotypes is effective, a hitch-hiking process may lead to an increase in other wild characters such as candle morphology. This would be reinforced by linkage of domesticated genes in pearl millet (Poncet et al. 2002 and see also Sect. 13.1.3.2). Conclusion The data from genetic studies and investigations on farmers’ practices suggest that the abundance of intermediate phenotypes in fields in southwestern Niger is mainly because of farmers’ practices and especially seed flow due to seed exchanges between farmers rather than the direct pollen flow and hybridization between domesticated and wild pearl millets. It is well known that seed flow due to seed exchanges between farmers regularly increases for several years. Moreover, farmers have no means or even willingness to specifically eliminate Shibra in drills. This provides an opportunity for reinforcing seed banks through shattering plants and could lead to long-term persistence of Shibra plants in the fields. Thus, Shibra plants, rather than wild ones, provide a regular source of traits from the wild to be further introgressed in domesticated phenotypes. However, it is difficult to really appreciate, through the evocation of the farmers whether we are facing a new and evolving process or whether this situation corresponds to a state of dynamic equilibrium that has lasted for decades. In the first hypothesis, the process of homogenization of pearl millet populations through gene flow could result in a rapid degradation of the level of adaptation of local varieties. Hence, it is crucial to analyze the role of changes in farmers’ practices and the economical and ecological constraints in order to better understand the evolution of the pearl millet crop/weed complexes.

13.1.3.2 Evolutionary Genomics and the Wealth of Wild Relatives for Deciphering the Genetic Basis of Quantitative Traits in Pearl Millet The genetic maps of many cereals are becoming increasingly dense, thanks to the dramatic development

T. Robert et al.

of accurate molecular markers (e.g., RFLP, AFLP, SSR, SNP) during the last two decades (Liu et al. 1994; Devos et al. 1995; Qi et al. 2001; Budak et al. 2003; Bertin et al. 2005). The genomic resources resulting from sequencing technologies together with comparative mapping studies highlighted the genomic organization among species of the Poaceae family (Devos et al 2000; Qi et al. 2004; Senthilvel et al. 2008). In pearl millet like in many other species, these maps constitute the linchpin of QTL analysis for the genetic dissection of complex traits. For example, in pearl millet, dwarfing genes have been mapped as well as QTLs involved in tolerance to biotic or abiotic stresses (see Yadav et al. 2002, 2004 for drought tolerance; Jones et al. 1995, 2002; Morgan et al 1998 for resistance to pathogens; Yadav et al 2003; Bidinger et al 2007 for enhanced productivity under diverse environments). Marker-assisted selection has been employed to improve efficiency of pearl millet genetic breeding schemes (Hash et al. 2003; Khairwal and Hash 2007). As far as wild relatives of pearl millet are concerned, two main domains using evolutionary genomics and molecular genetics approaches have been addressed: dissection of domestication process and interspecific hybridization (using tertiary gene pool species, P. squamulatum and C. ciliaris, for example) for transferring single genes of interest or unraveling genetic basis of complex traits like apomixis (Ozias-Akins et al. 1998, 2003; Conner et al. 2008).

The Genetic Basis of Pearl Millet Domestication The QTL analysis and evolutionary genomic studies (Doebley and Stec 1993; Poncet et al. 2000, 2002) have generated a strong body of knowledge on domestication process. These findings laid the foundation to conceptual inferences tracing back to the origin of domesticated plants and the role of evolutionary forces including selection (both natural and anthropogenic), mutation, gene flow, and genetic drift (founder effect) on the success of domestication process and the dynamics of agrobiodiversity. This section focuses on the current understanding of the domestication process in Pennisetum gene pool based on genetics and genomic investigations (see Sect. 13.1 for general frame of pearl millet domestication). The perusal of literature on the analysis of genetic basis of domestication syndrome (i.e., traits differentiating

13 Pennisetum

239

domesticated forms from their wild progenitors as resulting from the effects of the evolutionary forces mentioned above) clearly indicates that a few genes have been recruited during the domestication process. This figure holds for the main cereal species (maize, rice, wheat) despite differences in the architecture of their genomes and their mode of domestication (see modes of domestication scheme by Small 1984; Hillman and Davies 1990; Doebley and Stec 1993; Dorweiler et al. 1993). Studies on pearl millet domestication process are consistent with this general trend.

Table 13.3 List of morphological traits used for QTL mapping Code Units Trait description Traits measured 1 month after sowing (juvenile stage) NL Number of leaves on the primary tiller (shoot) NT Number of basal tillers LS cm Length of the sheath from base to the first leaf, on the primary tiller DS mm Diameter of the sheath LL cm Length of the fourth appearing leaf WiL mm Width of the fourth appearing leaf TNL Total number of leaves

Genetic Dissection of Domestication Syndrome Traits

Flowering time Head Days after sowing

Pearl millet, P. glaucum ssp. glaucum (L.) R. Br., among the species of the genus Pennisetum, is the only domesticated one. This species is one of the main staple cereal crops of the semi-arid regions of Africa and India. The wild forms of pearl millet, P. mollissimum and P. violaceum (ecotypes of P. glaucum ssp. monodii), are distributed only in Africa, where they have been involved in the domestication of the crop for several thousand years (Brunken et al. 1977; Appa Rao and de Wet 1999; see also Sect. 13.2). The main traits composing the domestication syndrome in Pennisetum are listed in Table 13.3. As for many cereals, spike “gigantism” and non-shattering of the spikelets at maturity are among the key traits. The wild relative forms of pearl millet are still found in sympatric or parapatric distribution with the cultivated form in some areas of the Sahelian regions of Africa. Both P. glaucum and its two wild relatives (P. mollissimum and P. violaceum), mainly outcrossers, are fully interfertile and have the same genome size (see Sect. 13.1.2 on Chromosome evolution). This peculiar situation in the pearl millet gene pool offers a remarkable framework for addressing the evolutionary questions about the number of genes involved in the domestication process and their genomic organization. The inheritance of domestication traits distinguishing pearl millet (P. glaucum) from one of its wild relatives viz. P. mollissimum has been assessed by several authors (Joly-Ichenhauser 1984; Perne`s 1986). For example, Poncet et al. (1998) tackled this issue in F2 progenies derived from a cross between a typical early flowering landrace of pearl millet (Souna

HHead cm PI (BMF-BFF)/(DO)

Days to heading of the primary tiller Plant height at first heading date Protogyny index (with BFF: beginning of female flowering, BMF: beginning of male flowering on the primary tiller, DO; duration of the female flowering)

Traits measured at maturity NNTP

Number of nodes on the primary tiller NS Total number of spikes/plant NTM Number of basal tillers Hmax cm Plant height HPT cm Height of the primary tiller LLM cm Length of the fourth leaf at maturity WiLM mm Width of the fourth leaf at maturity Traits measured on the primary mature spike WeS LoS WiS PL

g cm cm mm

BL Ct

mm Scaling points (0 as cultivated phenotype)

AL

Presence (1) or not (0) Presence (1) or not (0) mm

PB GL

Weight of the spike Length of the spike Width of the spike Pedicel length of the floral involucre measured from the abscission layer (or rachis) to the base of the involucre Length of the involucre bristles Seed coating (exposed seeds as the cultivated phenotype or enclosed seeds as the wild phenotype) Functional abscission layer Existence of a longer bristle Length of the upper floret lemma

240

from Mali) and a wild ecotype P. mollisimum near Gao, Mali.

Traits of the Spikelet Structure Displayed a Monogenic or Digenic Inheritance The most critical adaptive differences between wild and cultivated pearl millet involve the spikelet structure (Table 13.3). In wild pearl millet, the presence of an abscission layer at the basis of the involucre allows fertile spikelets to be shed spontaneously at maturity. Abundant bristles and the presence of a long awn, all ornamented with backward-pointing bristles, enhance the dispersal abilities. Once on the ground, these bristles and the arrow-shaped morphology of the sessile spikelets allow deeper penetration in the soil and thus contribute to the survival of the seed until its germination. Table 13.4 presents a comprehensive review of Mendelian segregation of genes involved in the inheritance of spikelet traits. The inheritance of the length of the involucral pedicel (PL) exhibit differences according to the cultivated and wild parent used in the crosses. Indeed, crosses involving the wild ecotype, P. mollissimum, and other cultivated forms, from different geographical origins had always revealed a monogenic inheritance, whereas a cross between the line 23d2B and P. violaceum showed a digenic inheritance. The presence of an abscission layer (AL) and the seed coating have a digenic inheritance in most cases. Using the order and distances between genes, Joly-Ichenhauser (1984) has

T. Robert et al.

hypothesized that two chromosome segments (segment I and segment II) were involved in the change of the spikelet architecture during the domestication process. QTls mapping from the analysis of two (cultivated  wild) F2 populations (Poncet et al. 2000, 2002) has confirmed the implication of major genes controlling the differences in spikelet structure between domestic and wild phenotypes (Table 13.5). It has also shown that the segments I and II previously identified correspond to LG 6 and 7 on the pearl millet genetic map. Moreover, seed coating that indicates the extent to which the glumes enclose the seed depends in part on the length of the glumes (GL). Three unlinked genetic factors were identified for this trait. A major QTL, gl7, which explained 49.9% of the phenotypic variance, mapped to the same region as ct7 on group 7, and two other QTLs with minor effects, gl5 and gl6, were located on LG 5 and LG 6, respectively (Table 13.5). The wild parental alleles of QTLs, gl6 and gl7, increased glume length with overdominant and dominant effects, respectively. Some Insights in the QTL Dissection of Other Domestication Syndrome Traits Spike Morphology Distributions for the length of the spike (LoS) (Fig. 13.11) and the width of the spike (WiS) revealed transgressive segregants among the F2 progenies. The weight of the spike (WeS) was controlled by at least three QTLs: wes2 (LG 2), wes5 (LG 5), and wes7

Table 13.4 Number of genes involved in the variation of the seed coating (Ct), the abscission layer (AL), the length of the pedicel (PL), and their localization on the two main segments of the Pearl millet’s genetic map (segments I or II indicated within brackets) in cultivated  wild pearl millet crosses involving plant material from different regions in Africa Traits Cultivated/wild combinationsa (countries) Ct AL PL 2 (I? + II) 1 (II) 1 + minor (II) cv. Souna  P. mollissimumb (Mali  Mali) i.l. Massue/P. mollissimumc (Mauritania  Mali) 1 (I) 1 (I) 1 (I) cv. Tiotande´/P. mollissimumc (Senegal  Mali) 2 (I + II) 2 (I + II) 1 (I) 2 (I + II) 2 (I + II) 1 (I) cv. Droˆo/P. mollissimumc (Tunisia  Mali) i.l. 23 d2B/P. mollissimumc (Ghana  Mali) 2 (I + II) 2 (I + II) 1 (II) cv. Zongo/P. violaceumc (Niger  Niger) 2 (I + II) 2 (I + II) 1 (I) 1 (II) 2 (I + II) 1 (II) cv. Droˆo/P. violaceumc (Tunisia  Niger) i.l. 23d2B/P. violaceumc (Ghana  Niger) 2 (I + II) 1 (II) 2 (I + II) cv. cultivar, i.l. inbred line The same accessions of P. mollissimum and P. violaceum were used in the different crosses b Poncet et al. (1998) c Joly-Ichenhauser (1984), backcrosses and F2 populations are analysed in each combination a

13 Pennisetum

241

Table 13.5 Biometrical parameters of individual QTLs and cumulated QTLs (tot) affecting the domestication syndrome traits Variable QTL Group Locus LOD % var a d d/a d allele NL / NT nt7 7 655 2.78 16.4 0.14 0.7 5.09 + LS / DS ds7 7 655 2.3 14.7 0.17 1.2 6.99 + LL ll5 5 345 2.8 14.7 2.61 0.18 0.07 / WiL / log(Hmax) hmax7 7 655 3.13 18.9 0.04 0.054 0.02 c hmax6 6 Est-E-al6 0.04 0.01 0.13 + tot 5.08 27.2 HPT htp5 5 345-731 5.58 29.9 21.41 27.46 1.28 c htp2 2 176-592 10.44 16.5 1.58 + tot 7.66 38.6 HPT-HHead (htp-hhead)1 1 515-756 2.68 15.7 4.61 11.57 2.5 c NNPT nnpt2 2 856-176 3.5 17.6 0.29 0.86 2.97 c nnpt6 6 Est-E-al6 2.7 13.9 0.29 0.82 2.82 c tot 6.0 28.5 sr(NS) ns1 1 461-858 2.38 20.4 0.24 1.37 5.73 + ns2 2 856-176 2.78 14.3 0.82 0.06 0.08 ns6 6 Est-E-al6 3.37 25.0 1.12 0.1 0.09 ns7 7 655 2.8 15.4 0.63 0.74 1.17 + tot 11.4 61.3 NTM nbt2 2 332 2.68 13.7 3.31 1.24 0.37 / nbt5-1 5 (1) 345 4.31 21.4 4.54 0.85 0.19 / nbt5-2 5 (2) 735 2.79 17.2 3.78 2.26 0.60 c tot 11.7 57.6 LLM / WiLM wilm2-1 2 (1) 332 5.72 27.6 0.42 0.041 0.1 / wilm2-2 2 (2) 592 4.8 24.3 0.41 0.15 0.37 + wilm4 4 464-515B 3.36 30.9 0.36 0.35 0.99 + wilm6 6 al6-713 2.34 20.8 0.27 0.35 1.28 c tot 10.53 54.9 HHead hhead5 5 651-735 6.65 34.7 24.75 26.13 1.06 c hhead7 7 812 2.47 14.2 16.16 8.04 0.50 c tot 9.15 42.9 Head head5 5 651-735 10.75 59.0 10.93 10.23 0.94 c head7 7 655 3.68 24.2 5.52 7.05 1.28 + tot 15.16 67 BFF bff2 2 176-592 3.2 3.04 0.95 + bff5 5 651-735 12.78 64.0 10.24 11.1 1.08 c bff6 6 202 2.4 14.6 2.28 8.07 3.53 c bff7 7 655 2.62 0.57 0.22 / tot 19.21 73.5 BFF-Head (bff-head)1 1 858-515 5.04 26.2 1.88 0.12 0.06 / (bff-head)7 7 655-812 9.4 43.4 1.96 1.38 0.7 + tot 13.67 58.1 BMF-BFF / FFF-BFF / PI / log(WeS) wes2 2 176-592 15.43 59.7 0.28 0.00 0.00 wes5 5 651-735 2.63 15.8 0.14 0.1 0.72 + wes6 6 Est-E-al6 2.41 18.8 0.13 0.12 0.9 c tot 24.5 82 log(LoS) los1 1 858-515 2.34 12.7 0.04 0.00 0.07 (continued)

242

T. Robert et al.

Table 13.5 (continued) Variable QTL los2 los7 WiS

wis5 wis7

sr(PL)

pl6 pl7

log(BL)

bl6 bl7

GL

gl5 gl6 gl7

Group 2 7 tot 5 7 tot 6 7 tot 6 7 tot 5 6 7 tot

Locus 592-738 655 735 655 Est-E- al6 655 Est-E- al6 655 651-735 Est-E-al6 655

LOD 7.11 2.41 12.3 2.62 15.58 22.2 11.86

% var 34.7 13.7 54 15.4 62.0 75.8 59.5

17.22 3.75 8.90 10.02 3.24 2.87 10.72 18.48

73.4 54.7 44.6 71.4 17.9 16.4 49.9 69.5

a 0.07 0.03

d 0.02 0.04

d/a 0.26 1.61

d allele

0.34 0.54

0.10 0.21

0.31 0.39

/ /

0.62 0.28

0.15 0.13

0.24 0.48

/ +

0.09 0.15

0.3 0.12

3.01 0.77

c +

0.44 0.23 0.52

0.02 0.39 0.37

0.05 1.70 0.71

/ + +

+

% var percentage of phenotypic variance explained, a additive effect of the wild allele (when the positive effect is attributed to the wild allele, figures are in italic), d dominance effect, d allele the dominant allele is c cultivated or þ wild allele

(LG 7), which accounted for 82% of the phenotypic variation (Table 13.5). The major QTL, wes2, alone explained more than 50%. For all the QTLs, the cultivated alleles increased weight, as predicted by the parental means. The length of the spike was mainly determined by three QTLs. The major QTL, los2 (34.7% of the phenotypic variance) was located on LG 2. The additive effect of the cultivated allele increased the length. In contrast, the positive effect of the two minor QTLs located on LG 1 (los1) and LG 7 (los7) was attributed to the wild alleles (Table 13.5). Thus, when positive alleles of the three QTLs are accumulated in the same genotype, transgressive phenotypes with longer spikes than the cultivated parent could be obtained. This can occur quite easily as these QTLs are located on different chromosomes. One of the QTLs for both weight and LoS, wes2 and los2, coincided on LG 2. For these two QTLs, the magnitude and direction of allele effects were similar. Two QTLs were detected for the WiS. The effect of wis7 (LG 7) was the largest and explained 62% of the phenotypic variance. The additive effect of the cultivated allele of wis7 reduced the width. A minor QTL on LG 5 was also detected where the wild allele exerted a positive effect (Table 13.5). Plant Architecture Differences between the wild and cultivated phenotypes were observed at the juvenile stage, especially

for sheath (LS) and leaf (LL) length, and to a less extent for the number of leaves (NL, TNL), the width of the leaf (WiL), and the number of basal tillers (NT). A few QTLs were detected (Table 13.5) for the juvenile traits, nt7, dg7, and lof5. The QTLs nt7 and dg7 were mapped in the same interval on LG 7. In both cases, the wild allele exhibited a positive additive effect and overdominance in their favor. The maximum height of the plant (Hmax) was influenced by a QTL on LG 7, which was responsible for 18.9% of the phenotypic variation. The additive effect of the cultivated allele of hmax7 increased the height. Four unlinked regions contributing to variation in the total number of fertile tillers (NS) were identified: ns1, ns2, ns6, and ns7. All the QTLs were of similar effect, each explained between 14.3% and 25% of the phenotypic variation, totaling 61.3%. As expected, all wild alleles increased the number of fertile tillers. The QTLs detected for the number of basal tillers at maturity (NTM) were located on LG 2 and LG 5. They were distinct from the QTLs for the total number of spikes (NS). However, ns2 and ns6 were closely linked to two QTLs controlling the number of nodes on the primary tiller (NNPT). The QTLs ns7 and nt7 may correspond to the same QTL, whereas ns1, ns2, and ns6 could be involved in axillary tillering.

13 Pennisetum

Number of observations

Fig. 13.11 Frequency distribution for spike length in the F2 population. The means and standard deviations (in brackets) are given for parents, F1 and F2 populations

243 15 F2 F1 Cultivated 14 _ _ _ X = 14.4 (±1.9) 13 X = 12.9 (±2.6) X = 14.6 (±1.8) 12 11 10 9 8 7 Wild 6 X = 7.1 (±1.5) 5 4 3 2 1 0 7 8 9 10 11 12 13 14 15 16 17 18 19 Length of the spike (LoS) in cm

Flowering The two (cultivated  wild) F2 progenies analyzed shows a variation in flowering time. The dates of heading and female and male flowering were all associated with a major QTL on LG 5 that was responsible for a minimum of 50% of the phenotypic variation observed for each trait (Table 13.5). The cultivated alleles at this QTL delayed heading and flowering in a dominant manner. LG 6 and 7 also carried QTLs with weaker effects on these characters. Variation in the delay between heading (Head) and beginning of stigma emergence (BFF) was observed. Two QTLs on LG 7 and LG 1 jointly accounted for 58% of the phenotypic variation for this trait.

Concluding Remarks on the Wealth of Wild Relatives for Unraveling the Evolutionary Mechanisms Underlying the Domestication Process The genetic control of the domestication syndrome in Pennisetum sp. at the spikelet level involved factors with large effects that were located on LG 6 (Fig. 13.12) and/or on LG 7. The evolution from wild to cultivated spikelet morphology occurred mostly through recessive mutations. QTL analysis revealed that a few loci mapped on few genomic regions are involved in the inheritance

20

21

22

23

of the major quantitative traits of the domestication syndrome (Fig. 13.12). This finding is consistent with the current data on the domestication of the principal cereal species. Indeed, comparative mapping studies suggest that domestication was achieved by selection (along with the other evolutionary forces) on only few genomic regions (Peng et al. 2003; Li et al. 2006; Purugganan and Fuller 2009). Nevertheless, this general statement could be relative by considering the methodological aspects of QTL mapping. The usual sizes of the mapping population might be too small for the detection of scattered loci with low effect on domestication quantitative traits. The very importance of these types of QTL as well as the pleiotropic and/or epistatic effects of linked genes on QTL expression are still to be investigated in Pennisetum by means of QTL fine-mapping.

Transgressive Phenotypes Resulting from Wild QTL Effects: An Evolutionary Fancy and Basis for New Genetic Resources Enhancement Strategies Transgressions and Genetic Resource For most of the major QTLs, the directions of the effects of the parental alleles were as predicted by the parental phenotypic means. However, for some traits, Hmax, LoS, WiS, BL, and GL, positive alleles

244

T. Robert et al. GL1

Xpsm461

GL4

GL2

GL6

Xpsm464

Xpsm322

GL3

Xpsm202

R2 > 20%

Xpsm592b

Xpsm856

R2 < 20% Xpsm176

Generative structures Panicle length Panicle weight Glume length Panicle width Bristle length Peduncle length Vegetative structures 4th leaf width Number of tillers Number of fertile tillers Main culm height Plant height Nb of nodes on main culm

Xpsm476 Xpsm592

Xpsm515b Xpsm568 Xpsm84

Xpsm858 Xpsm515 Xpsm756 Xpsm336 Xpsm386

Xpsm416

Xpsm738

Xpsm588

Xpsm696 EST-E

Xpsm416b

GL5

Xpsm345 GL7

Xpsm749 Xpsm731

Xpsm655

Xpsm651 Xpsm812

Xpsm713

Xpsm735

Fig. 13.12 Consensus map of domestication QTLs identified in two cultivated  wild F2 progenies. Most of the traits analyzed involved major QTLs. The linkage group 6 and 7 harbors a high number of QTLs, especially those involved in the variation of spikelet structure

were contributed by both the parents. When an individual received positive alleles from both the parents for a particular trait, its phenotypic value exceeds the parental ones. Transgressive phenotypes occurred even easier when a trait is controlled by unlinked QTLs (e.g., LoS). The existence of wild alleles that reinforce the cultivated phenotype in the F2 background (WiS, LoS, Hmax, etc.) raises questions on the mechanism allowing their maintenance in the wild populations while they are expected to decrease the adaptive value of the wild plant in natural conditions. This paradox could be explained by epistasis of the wild genetic background on the effect of the QTL leading to its silencing. This hypothesis is still to be tested. According to the evolutionary history of a species gene pool, the bottleneck affecting the genetic diversity and the differentiation are more or less important. However, such results open promising possibilities to enhance pearl millet landraces by using genetic resources from the wild. Indeed, despite their overall lower value in terms of agronomic performance, the closest wild relatives of crops can be sources of novel

genes not only to enhance stress tolerance but also to substantially increase yield. Such favorable genes have already been detected for fruit in tomato (Fulton et al. 1997) and for the number of grains per panicle in rice (Xiao et al. 1996). Introgression of Oryza rufipogon wild alleles at two specific loci into a top-performing rice variety was associated with yield increases of 18% and 17%. Based on this finding, new strategies were designed for massive introgression of QTL alleles from wild relatives (Tanksley and McCouch 1997; McCouch et al. 2006). Improvement of yield components in pearl millet, such as LoS, could thus be obtained by introgression of wild alleles. But this strategy should be carefully designed because of the strong gene flow supported by the allogamous breeding system and the sympatric or parapatric distribution of pearl millet with its wild relatives (P. mollissimum or P. violaceum). In addition, this transgression phenomenon could be the source of vigorous introgressed intermediate forms, which are often considered as aggressive weeds by farmers (see Sect. 13.1.3.2). This point was raised by Rieseberg et al. (2007) in sunflower.

13 Pennisetum

245

13.2 Enhancement of Genetic Resources van Dijk 2007). Molecular markers and genomic tools Using Genes from Wild Pennisetum have greatly enhanced the understanding of his trait. Apomictic meiosis has for long been regarded as byRelatives As shown in previous sections, the Pennisetum complex of species provides a favorable biological context for addressing fundamental questions on the evolution of grasses. As for other cereal species belonging to the Poaceae family, wild relatives have been targeted as either sources of genes of interest for tolerance to biotic and abiotic stresses or raw material for direct use. Many wild Pennisetum species were used in grass lands of tropical areas or as fodder crop. For example, P. purpureum, P. squamulatum, P. orientale, P. polystachion, P. flaccidum, and P. pedicellatum contribute significantly to non-cultivated forage production in the tropics. Some species (P. alopecuroides, P. orientale, P. setaceum) are grown as ornamental plants. A sustainable use of wild genetic resources depends by far on the accuracy and the thoughtfulness of fundamental knowledge about the structure of genetic diversity and genome organization as revealed by evolutionary genomics studies on the gene pools. In this section, we put emphasis on some meaningful strategies developed for optimizing the use of wild Pennisetum genetic resources.

13.2.1 Wild Pennisetum Species as Sources of Major Genes of Agronomic Interest (GAI) 13.2.1.1 Apomixis Apomixis is an asexual reproduction system that produces progenies via seeds without sperm–egg fusion. Embryos arise from cells of the maternal genotype. This trait occurs in wild relatives of crop plants and can be introduced by crossing. The inheritance of this breeding system has been studied in many plant families both in monocots and dicots (for example: Poaceae, Liliaceae, Brassicaceae, Compositeae, etc.). Many reviews and books have recently been devoted to understanding apomixis (Savidan 2001; OziasAkins 2006; Hoerandl et al. 2007; Ozias-Akins and

passed or non-functional. Recently, using genomic approaches, it has been demonstrated in Arabidopsis mutants that the gene sw1 (DYAD) causes synapsis to fail in female meiosis and yields two unreduced gametes (Noyes 2008). The genus Pennisetum contains many apomictic species (see Tables 13.1 and 13.2), for example: P. setaceum and P. squamulatum. C. ciliaris (previously P. ciliare), a close relative to Pennisetum species (see Fig. 13.4), also deserves a mention. A large (>50 Mb) ASGR was observed for P. squamulatum (Goel et al. 2003; Akiyama et al. 2004, 2005). Gametophytic apomixis in C. ciliaris is also controlled by ASGR (Akiyama et al. 2005). This trait is highly conserved and macrosyntenic with ASGR of P. squamulatum. Recently, Conner et al. (2008) have sequenced and analyzed a portion of ASGR in Pennisetum and Cenchrus. ASGR sequences reveal the presence of transposable elements and contain multiple regions of microsynteny with Oryza sativa. Forty potentially transcribed genes were identified, some of them may play a role in apomictic development. Given the large non-recombining property of ASGR and the presence of repetitive DNA, the identification of apomixis gene by genetic mapping seems particularly difficult. Apomictic trait has already been transferred from wild progenitors to cultivated forms for the purpose of propagating heterozygous genotypes and hopefully heterosis. In breeding for fodder, the aposporous apomictic trait from P. squamulatum is being transferred to pearl millet by means of backcrosses. Using a molecular cytogenetic method genomic in situ hybridization (GISH), it was demonstrated that in BC8, one to three chromosomes from P. squamulatum are present in the apomictic plants. This finding evidences for many hopes for the release of a apomictic pearl millet variety for fodder purpose. Nevertheless, many difficulties are still to be resolved before producing desired apomictic pearl millet varieties for fodder purpose (Ozias-Akins et al. 2003; Akiyama et al. 2004; Jauhar et al. 2006).

246

Male Sterility When Natural Reproductive Barriers Turn to be Source of GAI The gene pool classification is based on natural crossability between the wild relatives and the domesticated form. The hybrid breeding strategy widely uses cytoplasmic male sterility either naturally occurring in the cultivated form or induced using interspecific hybridization, for example. In pearl millet breeding for F1 varieties, A1 cytoplasmic male sterility (CMS) has been widely used. Nevertheless, in order to avoid cytoplasmic uniformity and the risks of vulnerability to disease and insect pests, other CMS sources were identified as alternative to A1 CMS. Two new CMS systems have been identified in wild species P. glaucum ssp. violaceum (¼ monodii) from Senegal: Av CMS (Marchais and Perne`s 1985) and A4 CMS (Hanna 1989). Evaluation of three CMS systems (Av, A4, and A5; Rai 1995) indicates that the A4 CMS system provides a better alternative to the A1 CMS system than the Av system (Rai et al. 2001). Chandra-Shekara et al. (2007) confirms the potentiality of A4 and A5 CMS sources as potential alternatives to diversify the male sterile cytoplasm in pearl millet hybrid breeding. Rai et al. (2009), comparing the three CMS systems (A1, A4 and A5) suggested that seed parents breeding efficiency will be better with the A5 CMS system, followed by the A4 CMS system, and lesser with the currently commercial A1 CMS system.

Genes Involved in Tolerance to Biotic and Abiotic Stress Despite the drawbacks of hybrids sterility, species from secondary and tertiary gene pools are also potential donors of favorable traits (Hanna 1986, 1990; Rao et al. 2003). Evaluation of wild accessions for useful traits, such as resistance and tolerance, is needed for incorporation of these species in breeding strategies. For instance, P. orientale L. C. Rich (2n ¼ 2x ¼ 18) has been assessed as potential donor for some characters like perenniality, drought tolerance, pest resistance, and winter hardiness (Dujardin and Hanna 1987). Trispecific hybrids between pearl millet, P. squamulatum, and napier grass have also been obtained and could be used to transfer germplasm across species (Dujardin and Hanna 1985). The traits of agronomic interest identified in wild species are usually transferred to pearl millet via back-

T. Robert et al.

crosses. Wild P. glaucum ssp. violaceum from primary gene pool and species from tertiary gene pool were evaluated for disease resistance, mainly for Pyricularia grisea, Puccinia substriata var. indica (Wilson and Hanna 1992). Many resources of germplasms present at least resistance to one of the five pathogens tested and offer global valuable diversity for disease resistance. Resistance to P. grisea (Cke) Sarc. leaf spot of pearl millet was discovered in P. glaucum subsp. monodii originating from Senegal, and was transferred effectively to pearl millet (Hanna and Wells 1989). So, both a resistant inbred line and a resistant commercial forage hybrid were produced. Another important pathogen is a parasitic weed Striga hermonthica that infects grain fields resulting in an annual grain yield losses. Resistance to S. hermonthica was evaluated in more than 250 wild P. glaucum subsp. monodii and stenostachyum accessions (Wilson et al. 2000, 2004) in diverse regions in West Africa. Four accessions were classified as resistant and are useful sources of striga resistance for pearl millet breeding in West Africa. Tolerance to abiotic stresses like drought or salinity have been tested in wild relatives. For example, accessions of P. purpureum were evaluated for tolerance to salinity but no tolerance was detected in this material (Kulkarni et al. 2006).

13.2.2 Wide Use of Wild Pennisetum Species 13.2.2.1 Breeding for Fodder Genetic improvement of Napier grass itself for the increase of biomass yield is usually achieved using ecotypic selection. An extensive breeding and yield evaluations were performed in Florida during the 1980s. Two high yielding improved varieties “Merkeron” Napier grass (a vegetatively propagated F1 clone) and “Mott” Napier grass (a selfed progeny of Merkeron Napier grass) have been registered in 1989. In tropical grasses, numerous interspecific and some intergeneric hybrids were produced (see Burson and Young 2001). For Napier grass, in addition to conventional breeding programs, the interspecific hybridization between P. purpureum and pearl millet

13 Pennisetum

is an alternative to obtain superior fodder cultivars. Indeed, P. purpureum, one of the two members of the secondary gene pool, is easily crossable with P. glaucum, but sterility is predominant in the interspecific triploid (2n ¼ 3x ¼ 21) hybrids. Nevertheless, these triploid hybrids show high heterosis for fodder quality and yield, due to introgression of favorable traits from P. glaucum, such as forage quality. Allohexaploids expressing male- and female-fertility have been produced by means of colchicine chromosome doubling (Gonzalez and Hanna 1984; Hanna et al. 1984). In addition, Hanna (1990) showed the potentiality for gene transferring from P. purpureum to P. glaucum through a mechanism producing monoploid gametes (see Sotomayor-Rios and Pitman 2001 and Jauhar et al. 2006, for review). Hybrids between pearl millet and wild P. glaucum subsp. monodii accessions from three African Sahelian countries were tested during 2 years for production of dry matter (Hanna 2000). This study and others later confirm that genes for enhancing pearl millet yield are present in the wild grassy subspecies.

13.2.2.2 Miscellaneous Uses of Wild Relatives BioEnergy With progress made on converting cellulosic feedstocks to ethanol, Napier grass, due to its high biomass producing, could be used for ethanol production (Anderson et al. 2008a). Three species (Bermuda grass, Napier grass, and giant reed) were compared for ethanol production, the results indicate that Bermudagrass yielded more ethanol compared to napier grass (see Anderson et al. 2008b for more details).

Hedge Vegetative Barriers for Monitoring Soil Erosion in Fields P. purpureum is also currently used as vegetative hedge barrier. For instance, it was tested as buffer strips on row crop production fields (Anderson et al. 2008c). The Efficiency of narrow strips of Napier grass (P. purpureum) and Vetiver (Vetiveria zizanioides) in reducing runoff and nutrient loss from the field have been tested in Kenya (Owino et al 2006). In this study and others (Mutegi et al. 2008), Napier grass was more

247

effective than tree hedges for reduction of soil erosion and thereby for yield improvement of maize. It has been widely adopted among indigenous grasses as the more popular in Kenya for soil conservation.

Push–Pull In Subsaharan African regions, stem borers and parasitic weeds like striga are among the main constraints affecting cereal production. Therefore, research programs on food security concerns have been launched for the improvement of farmers’ livelihood. The management of these two pests is central in these programs. For instance, ICIPE (International Center of Insect Physiology and Ecology) have developed an integrated management system named: Push–Pull in mitigation. This management technology consists in intercropping a fodder legume like Desmodium uncinatum with maize and planting a perimeter of P. purpureum (Niaper grass) around the plot. Stem borers are repelled away from the maize crop (Push) and trapped by the Niaper grass plants (Pull). In addition, Desmodium roots produce several polyphenolic compounds, some of which stimulated striga seeds’ germination while some others inhibit haustorial formation thereby suppressing striga. The adoption of this technology by farmers has been tested in Kenya (Khan et al. 2008). Hari and Jindal (2009) tested the efficiency of P. purpureum  P. glaucum as trap crop for the management of the stem borer Chilo partellus on maize. According to their data, Niaper millet hybrids were preferred for oviposition over maize by the stem borer moths. But subsequently, this material were found to be poor larval hosts and less eaten by borer larvae.

13.2.2.3 Pre-breeding Strategies Bramel-Cox et al. (1986) showed that progeny derived from pearl millet  wild species had higher growth rates. As mentioned in the Sect. 13.1.3.2, the occurrence of transgressive phenotypes in F2 P. glaucum  P. violaceum population (Poncet et al. 2000) suggests that the wild forms can be used to improve pearl millet and be integrated into breeding programs. Hajjar and Hodgkin (2007) reviewed information on the presence of useful genes from crop wild relative (CWR) in

248

cultivars of crops and noted an increase in the rate of cultivars containing genes from CWR, compared to the figures previously reported by Prescott-Allen and Prescott-Allen (1986, 1988). Potentiality of wild Pennisetum species for genetic improvement of pearl millet is well documented and reviewed in Rao et al. (2003) and Jauhar et al. (2006). All the above mentioned strategies for the use of genes from wild relatives are often restricted to the transfer through backcrossing of genes involved in simply inherited characters (e.g., genes encoding for resistance to disease or tolerance to stresses). However, this procedure may lead to genetic erosion because of the large amount of wild relatives collected and the genetic uniformity resulting from backcrosses on the recurrent cultivated parents. Pre-breeding has been identified as an appropriate solution to avoid this drawback and for the conservation and massive use of genes from wild relatives. A project aiming at the implementation of pre-breeding in pearl millet has been developed (Lamy et al. 1994; Sarr et al. 2001). In the frame of an integrated paradigm developed for the study of domestication process in pearl millet, recombinant populations (P. glaucum  P. violaceum) were created and tested in two African countries. The levels of introgression were assessed using molecular markers (allozymes, RLFP, STS, AFLP) and cytogenetics combined with reproductive biology and quantitative trait segregation. This strategy was efficient in partitioning the recombinant offsprings according to their levels of introgression with wild genome. Three pools were identified according to the number of introgressed linkage groups. These populations are potentially interesting as both conservation units and breeding material.

13.3 Conclusion and Future Scope of Research In this chapter, we put emphasis on the domestication process as it occurred in the primary gene pool of P. glaucum. Indeed, it is currently understood that the challenges and hopes for sustainable uses of genetic resources are by far dependent on the understanding of the dynamics of biodiversity, i.e., the factors involved in the origin, maintenance, or loss of diversity consid-

T. Robert et al.

ered from genes to eco/agrosystems. As noted by many authors (see Purugganan and Fuller 2009 for a review), domestication is one of the most important technological innovations in human history. Humans and Poaceae family are interdependent. Poaceae family includes about 10,000 species, which provides cereals, forages, sugarcane. The basis of this interdependence is to be unraveled both through anthropological studies and biological studies aiming at understanding the special features that put Poaceae species on the trail of domestication. Thus, domestication process has inspired many scholars including Charles Darwin. The appropriate paradigm for the study of domestication process is necessarily based on a multidisciplinary approach (genetics, cytogenetics, genomics, physiology, archeology, botany, anthropology, economic botany, agronomy, etc.). The key question is actually to understand the genetic basis of adaptation. Wild relatives of crop plants are central to the assessment of this question. Thanks to the dramatic progress in understanding genome organization made possible by molecular genetics and genomics, the mechanisms underlying plant adaptation and plasticity can now be dissected. The ultimate goal is to understand from an evolutionary point of view the architecture of key adaptative characters. Comparative QTL mapping and subsequent molecular investigations leading to QTA and QTN (individual nucleotide that causes particular phenotype) suggest parallel molecular evolution in Poaceae family at some few key loci involved in the domestication process (e.g., ZmTB1 gene). These studies bear future hopes for understanding how some grass members of the Poaceae family were recruited successfully in the domestication process. This issue could significantly help to uncloak the genetic architecture of the useful traits and then to monitor efficiently wide transfer of genes from wild relatives. As discussed in this chapter, the genus Pennisetum offers favorable conditions to address this issue. Indeed, the occurrence of the domesticated P. glaucum in sympatric or parapatric situations with wild relatives in the Sahelian region is a unique situation for the implementation of the above mentioned paradigm. Climatic pejoration hardly impacts this region generating, in conjunction with human activities, accelerated desertification process. Thus, unraveling the genetic basis of adaptation of pearl millet and its wild relatives to these difficult ecological conditions is a major issue. Indeed, gene

13 Pennisetum

flow, farmers’ strategies to monitor the evolution of domesticated varieties, and folk uses of wild relatives can be addressed on a multidisciplinary basis (anthropological, genetics, and genomic approaches, for instance). These studies are under progress and need strong support in the frame of ecological rehabilitation and biodiversity management programs. Conservations initiatives for Pennisetum species have been launched by many international organizations and national bodies. Many gene banks hold accessions of Pennisetum species. But the in situ conservation initiatives involving farmers’ knowledge and know-how should be prioritized. The on-farm conservation concept is consistent with these initiatives. As for industrial or commercial uses of Pennisetum species, many initiatives are under progress. Nevertheless, the improvement in these fields could benefit from the above mentioned assessment of genetic basis of adaptation and dissection of the architecture of characters using evolutionary genomics. Indeed, these approaches are expected to bridge the genotype–phenotype divide.

References Adams JM (1997) Preliminary vegetation maps of the world since the last glacial maximum: an aid to archaeological understanding. J Archaeol Sci 24:623–647 Akiyama Y, Conner JA, Goel S, Morishige DT, Mullet JE, Hanna WW, Ozias-Akins P (2004) High-resolution physical mapping in Pennisetum squamulatum reveals extensive chromosomal heteromorphism of the genomic region associated with apomixis. Plant Physiol 134:1733–1741 Akiyama Y, Hanna WW, Ozias-Akins P (2005) High-resolution physical mapping reveals that the apospory-specific genomic region (ASGR) in Cenchrus ciliaris is located on a heterochromatic and hemizygous region of a single chromosome. Theor Appl Genet 111:1042–1051 Akiyama Y, Goel S, Chen Z, Hanna W, Ozias-Akins P (2006) Pennisetum squamulatum: is the predominant cytotype hexaploid or octoploid? J Hered 97:521–524 Allinne C, Mariac C, Vigouroux Y, Bezanc¸on G, Couturon E, Moussa D, Tidjani M, Pham JL, Robert T (2008) Role of seed flow on the pattern and dynamics of pearl millet (Pennisetum glaucum [L.] R. Br.) genetic diversity assessed by AFLP markers: a study in south-western Niger. Genetica 133:167–178 Amblard S, Perne`s J (1989) The identification of cultivated pearl millet (Pennisetum) amongst plant impressions on pottery from Oued Chebbi (Dhar Oualata, Mauritania). Afr Archaeol Rev 7:117–126

249 Amoukou AI, Marchais L (1993) Evidence of a partial reproductive barrier between wild and cultivated pearl millets (Pennisetum glaucum). Euphytica 67:19–26 Amouroux-Pezas C (1985) B Chromosomes occurrence in a wild Pennisetum form (P. violaceum) : Behaviour and transferring strategies into domesticated pearl millet inbred lines. PhD Thesis. Paris XI University, France. (Orsay) ( Les chromosomes B du mil : leur gestion dans une forme spontane´e Pennisetum violaceum et leur transfert dans des ligne´es cultive´es. The`se docteur 3e`me cycle Orsay, France Anderson WF, Dien BS, Brandon SK, Peterson JD (2008a) Assessment of bermudagrass and bunch grasses as feedstock for conversion to ethanol. Appl Biochem Biotechnol 145:13–21 Anderson WF, Casler MD, Baldwin BS (2008b) Improvement of perennial forage species as feedstock for bioenergy. In: Vermerris W (ed) Genetic improvement of bioenergy crops. Springer, Berlin Anderson WF, Hubbard RK, Strickland TC (2008c) Napier grass as a field border and bioenergy source. In: Proceedings of joint annual meeting of the Georgia Chapters SWCS, ASABE, and IECA, Athens, GA, USA Antonovics J, Bradshaw AD (1970) Evolution in closely adjacent plant populations. VIII. Clinal patterns at a mine boundary. Heredity 25:349–362 Appa Rao S, de Wet JMJ (1999) Taxonomy and evolution. In: Khairwal IS, Rai KN, Andrews DJ, Harinarayana G (eds) Pearl millet breeding. Science, Enfield, NH, USA, pp 29–47 Avdulov NP (1931) Karyo-systematische untersuchungen der familie gramineen. Bull Appl Bot Genet Plant Breed Suppl 44:1–428 Ballouche A, Neumann K (1995) A new contribution to the Holocene vegetation history of the West African Sahel: pollen from Oursi, Burkina Faso and charcoal from three sites in Northeast Nigeria. Veg Hist Archaeobot 4:31–39 Bellwood P (2001) Early agriculturalist population diasporas? Farming, languages, and genes. Annu Rev Anthropol 30: 181–207 Bertin I, Zhu JH, Gale MD (2005) SSCP-SNP in pearl millet-a new marker system for comparative genetics. Theor Appl Genet 110(8):1467–1472 Bidinger FR, Nepolean T, Hash CT, Yadav RS, Howarth CJ (2007) Quantitative trait loci for grain yield in pearl millet under variable postflowering moisture conditions. Crop Sci 47:969–980 Boudry P, Mo¨rchen M, Saumitou-Laprade P, Vernet P, Van Dijk H (1993) The origin and evolution of weed beets: consequences for the breeding and release of herbicide resistant transgenic sugar beets. Theor Appl Genet 87:471–478 Bramel-Cox P, Andrews DJ, Frey KJ (1986) Exotic germplasm for improving grain yield and growth rate in pearl millet. Crop Sci 26:687–693 Brunken JN (1977) A systematic study of Pennisetum sect Pennisetum (Gramineae). Am J Bot 64(2):161–176 Brunken JN, de Wet JMJ, Harlan JR (1977) The morphology and domestication of pearl millet. Econ Bot 31:163–174 Budak H, Pedraza F, Cregan PB, Baenziger PS, Dweikat I (2003) Development and utilization of SSRs to estimate the degree of genetic relationships in a collection of pearl millet germplasm. Crop Sci 43:2284–2290

250 Burson BL, Young BA (2001) Breeding and improvement of tropical grasses. In: Sotomayor-Rios A, Pitman WD (eds) Tropical forage plants: development and use. CRC, Boca Raton, FL, USA, pp 59–80 Busso CS, Devos KM, Ross G, Mortimore M, Adams WM, Ambrose MJ, Alldrick S, Gale MD (2000) Genetic diversity within and among landraces of pearl millet (Pennisetum glaucum) under farmer management in West-Africa. Genet Resour Crop Evol 47:561–568 Camps G (1974) Northen Africa and Saharan Prehistoric Civilisations(Les civilisations pre´historiques de l’Afrique du Nord et du Sahara). Doin Ed, Paris Cavalli-Sforza LL, Menozzi P, Piazza A (1994) The history and geography of human genes. Princeton University Press, Princeton, NJ, USA Chandra-Shekara AC, Prasanna BM, Singh BB, Unnikrishnan KV, Seetharam A (2007) Effect of cytoplasm-nuclear interaction on combining ability and heterosis for agronomic traits in pearl millet {Pennisetum glaucum (L.) R. Br.}. Euphytica 153:15–26 Chiavarino AM, Rosato M, Rosi P, Poggio L, Naranjo CA (1998) Localization of the genes controlling B chromosome transmission rate in maize (Zea mays ssp. mays, Poaceae). Am J Bot 85:1581–1585 Chiavarino AM, Gonza´lez-Sa´nchez M, Poggio L, Puertas MJ, Rosato M, Rosi P (2001) Is maize B chromosome preferential fertilisation controlled by a single gene? Heredity 86:743–748 Chowdhury MKV, Smith RL (1988) Mitochondrial DNA variation in pearl millet and related species. Theor Appl Genet 76:25–32 Clayton WD, Renvoize SA (1986) Genera Graminum: grasses of the world. HMSO Book, London, UK, 389 p Clegg MT, Rawson JRY, Thomas K (1984) Chloroplast DNA variation in pearl millet and related species. Genetics 106:449–461 Conner JA, Goel S, Gunawan G, Cordonnier-Pratt MM, Johnson VE, Liang C, Wang H, Pratt LH, Mullet JE, Debarry J, Yang L, Bennetzen JL, Klein PE, Ozias-Akins P (2008) Sequence analysis of bacterial artificial chromosome clones from the apospory-specific genomic region of Pennisetum and Cenchrus. Plant Physiol 147:1396–1411 D’Andrea AC, Casey J (2002) Pearl millet and Kintampo subsistence. Afr Archaeol Rev 19:147–173 D’Andrea AC, Klee M, Casey J (2001) Archaeobotanical evidence for pearl millet (Pennisetum glaucum) in sub-Saharan West Africa. Antiquity 75:341–348 Darlington CD (1956) Chromosome botany. George Allen and Unwin, London, UK De Paepe R, Prat D, Knight J (1983) Effects of consecutive androgeneses on morphology and fertility in Nicotiana sylvestris. Can J Bot 61(7):2038–2046 Devos KM, Pittaway TS, Busso CS, Gale MD, Witcombe JR, Hash CT (1995) Molecular tools for the pearl millet nuclear genome. Int Sorghum Millets Newsl 36:64–65 Devos KM, Pittaway TS, Reynolds A, Gale MD (2000) Comparative mapping reveals a complex relationship between the pearl millet genome and those of foxtail millet and rice. Theor Appl Genet 100:190–198 Doebley J, Stec A (1993) Inheritance of the morphological differences between maize and teosinte: comparison of results for two F2 populations. Genetics 134:559–570

T. Robert et al. Dolezˇel J, Bartosˇ J, Voglmayr H, Greilhuber J (2003) Nuclear DNA content and genome size of trout and human. Cytometry 51:127–128 Donadio S, Giussani LM, Kellogg EA, Zuloaga FO, Morrone O (2009) A preliminary molecular phylogeny of Pennisetum and Cenchrus (Poaceae-Paniceae) based on the trnL-F, rpl16 chloroplast markers. Taxon 58:392–404 Dorweiler J, Stec A, Kermicle J, Doebley J (1993) Teosinte glume architecture 1: a genetic locus controlling a key step in maize evolution. Science 262(5131):233–235 Doust AN, Kellogg EA (2002) Inflorescence diversification in the panicoid “bristle grass” clade (Paniceae, Poaceae): evidence from molecular phylogenies and developmental morphology. Am J Bot 89:1203–1222 Doust AN, Penly AM, Jacobs SWL, Kellogg EA (2007) Congruence, conflict, and polyploidization shown by nuclear and chloroplast markers in the monophyletic “bristle clade” (Paniceae, Panicoideae, Poaceae). Syst Bot 32: 531–544 Dujardin M, Hanna WW (1984) Cytogenetics of double cross hybrids between Pennisetum americanum – P. purpureum amphiploids and P. americanum  Pennisetum squamulatum interspecific hybrids. Theor Appl Genet 69:97–100 Dujardin M, Hanna WW (1985) Cytology and reproductive behavior of pearl millet – Napier grass hexaploids  Pennisetum squamulatum trispecific hybrids. J Hered 76:382–384 Dujardin M, Hanna WW (1987) Inducing male fertility in crosses between pearl millet and Pennisetum orientale Rich. Crop Sci 27:65–68 Fuller DQ (2007) Contrasting patterns in crop domestication and domestication rates: recent archaeobotanical insights from the Old World. Ann Bot (Lond) 100:903–924 Fuller D, Korisettar R, Vankatasubbaiah PC, Jones MK (2004) Early plant domestications in southern India: some preliminary archaeobotanical results. Veg Hist Archaeobot 13:115–129 Fulton TM, Beck-Bunn T, Emmatty D, Eshed Y, Lopez J, Petiard V, Uhlig J, Zamir D, Tanksley SD (1997) QTL analysis of an advanced backcross of Lycopersicon peruvianum to the cultivated tomato and comparisons with QTLs found in other wild species. Theor Appl Genet 95:881–894 Gepts P, Clegg M (1989) Genetic diversity in pearl millet (Pennisetum glaucum [L.] R. Br.) at the DNA sequence level. J Hered 80:203–208 Giussani LM, Cota-Sa´nchez JH, Zuloaga FO, Kellogg EA (2001) A molecular phylogeny of the grass subfamily Panicoideae (Poaceae) shows multiple origins of C4 photosynthesis. Am J Bot 88:1993–2012 Goel S, Chen Z, Conner JA, Akiyama Y, Hanna WW, OziasAkins P (2003) Delineation by fluorescence in situ hybridization of a single hemizygous chromosomal region associated with aposporous embryo sac formation in Pennisetum squamulatum and Cenchrus ciliaris. Genetics 163:1069–1082 Goel S, Chen Z, Akiyama Y, Conner JA, Basu M, Gualtieri G, Hanna WW, Ozias-Akins P (2006) Comparative physical mapping of the aposporyspecific genomic region in two apomictic grasses, Pennisetum squamulatum and Cenchrus ciliaris. Genetics 173:389–400 Go´mez-Martı´nez R, Culham A (2000) Phylogeny of the subfamily Panicoideae with emphasis on the tribe Paniceae: evidence from the trnL-F cpDNA region. In: Jacobs SWL,

13 Pennisetum Everett J (eds) Grasses: systematics and evolution. CSIRO, Melbourne, Australia, pp 136–140 Gonzalez B, Hanna WW (1984) Morphological and fertility responses in isogenic triploid and hexaploid pearl millet  napier grass hybrids. J Hered 75:317–318 Greilhuber J, Dolezˇel J, Lysa´k M, Bennett MD (2005) The origin, evolution and proposed stabilization of the terms ‘genome size’ and ‘C-value’ to describe nuclear DNA contents. Ann Bot 95:255–260 Hajjar R, Hodgkin T (2007) The use of wild relatives in crop improvement: a survey of developments over the last 20 years. Euphytica 156:1–13 Han F, Lamb JC, YuW GZ, Birchler JA (2007) Centromere functional nondisjonction are independent components of the maize B chromosome accumulation mechanism. Plant Cell 19:524–533 Hanna WW (1986) Utilization of wild relatives of pearl millet. In: Proceedings of the international pearl millet workshop, 7–11 Apr 1986, ICRISAT, Pattancheru, India Hanna WW (1989) Characteristics and stability of a new cytoplasmic-nuclear male-sterile source in pearl millet. Crop Sci 29:1457–1459 Hanna WW (1990) Transfer of germplasm from the secondary to the primary gene pool in pennisetum. Theor Appl Genet 80:200–204 Hanna WW (2000) Total and seasonal distribution of dry matter yields for pearl millet  wild grassy subspecies hybrids. Crop Sci 40:1555–1558 Hanna WW, Dujardin M (1986) Cytogenetics of Pennisetum schweinfurthii Pilger and its hybrids with pearl millet. Crop Sci 26:449–453 Hanna WW, Wells HD (1989) Inheritance of Pyricularia leaf spot resistance in pearl millet. J Hered 80:145–147 Hanna WW, Gaines TP, Gonzalez B, Monsoon WG (1984) Effect of ploidy on yield and quality of pearl millet  Napiergrass hybrids. Agron J 76:969–971 Hari NS, Jindal J (2009) Assessment of Niaper millet (Pennisetum purpureum  P. glaucum) and sorghum (Sorghum bicolor) trap crops for the management of Chilo partellus on maize. Bull Entomol Res 99:131–137 Harlan JR (1971) Agricultural origins: centers and non-centers. Science 14:468–474 Harlan JR (1975) Crops and man. American Society of Agronomy and Crop Science Society of America, Madison, WI, USA Harlan JR, de Wet JMJ (1971) Toward a rational classification of cultivated plants. Taxon 20:509–517 Hash CT, Bhasker Raj AG, Lindup S, Sharma A, Beniwal CR, Folkertsma RT, Mahalakshmi V, Zerbini E, Bl€ ummel M (2003) Opportunities for marker-assisted selection (MAS) to improve the feed quality of crop residues in pearl millet and sorghum. Field Crop Res 84:79–88 Hillman GC, Davies MS (1990) Domestication rates in wildtype wheats and barley under primitive cultivation. Biol J Linn Soc 39:39–78 Hoerandl E, Grossniklaus U, van Dijk P, Sharbel T (eds) (2007) Apomixis: evolution, mechanisms and perspectives. Gantner Verlag, Rugell, Liechtenstein, Germany Ingham LD, Hanna WW, Baier JW, Hannah LC (1993) Origin of the main class of repetitive DNA within selected Pennisetum species. Mol Genet 238:350–358

251 Jarvis DI, Hodgkin T (1999) Wild relatives and crop cultivars: detecting natural introgression and farmer selection of new genetic combinations in agroecosystems. Mol Ecol 8: S159–S173 Jauhar PP (1968) Inter- and intra-genomal chromosome pairing in an inter-specific hybrid and its bearing on the basic chromosome number in Pennisetum. Genetica 39:360–370 Jauhar PP (1970) Chromosomes behaviour and fertility of the raw and ’evolved’ synthetic tetraploids of pearl millet, Pennisetum typhoides Stapf and Hubb. Genetica 41:407–424 Jauhar PP (1981) Cytogenetics and breeding of pearl millet and related species. Alan R Liss, New York, USA, 289 p Jauhar PP, Rai KD, Ozias-Akins P, Chen Z, Hanna WW (2006) Genetic improvement of pearl millet for grain and forage production: cytogenetic manipulation and heterosis breeding. In: Singh RJ, Jauhar PP (eds) Genetic resources, chromosome engineering, and crop improvement, vol 2, Cereals. CRC, Taylor & Francis, Boca Raton, FL, USA, pp 281–307 Jayalakhsmi K, Pantulu JV (1984) The effect of B-chromosome on A chromosome chiasma formation in pearl millet. Cytologica 49:635–643 Joly-Ichenhauser H (1984) Inheritance of domestication syndrome traits in pearl millet Pennisetum typhoides: comparative analysis of F2 and BC progenies from crosses between wild and domesticated forms. PhD Thesis, University of Paris, Orsay (He´re´dite´ du syndrome de domestication chez le mil Pennisetum typhoides: Etude compare´e de descendances (F2 et BC) issues de croisements entre plusieurs ge´niteurs cultive´s et spontane´s. The`se de 3e`me cycle, Univ Paris, Sud Orsay, France Joly-Ichenhauser H, Sarr A (1985) Preferential associations among characters in crosses between pearl millet (Pennisetum typhoides) and its wild relatives. In: Jacquard P, Heim G, Antonovics J (eds) Genetic differentiation and dispersal in plants. Springer, Berlin, Germany, pp 95–111 Jones RN (1985) Are B-chromosomes selfish? In: Cavalier-Smith T (ed) The evolution of genome size. Wiley, Chichester, UK, pp 397–425 Jones RN, Puertas MJ (1993) The B-chromosomes of rye (Secale cereale L.). In: Dhir KK, Sareen TS (eds) Frontiers in plant science research. Bhagwati Enterprises, Delhi, India, pp 81–112 Jones ES, Liu CJ, Gale MD, Hash CT, Witcombe JR (1995) Mapping quantitative trait loci for downy mildew resistance in pearl millet. Theor Appl Genet 91:448–456 Jones ES, Breese WA, Liu CJ, Singh SD, Shaw DS, Witcombe JR (2002) Mapping quantitative trait loci for resistance to downy mildew in pearl millet: field and glasshouse screens detect the same QTL. Crop Sci 42:1316–1323 Jones RN, Viega W, Houben A (2008) A century of B chromosomes in plants: so what? Ann Bot 101:767–775 Kayano H (1961) Cytogenetic studies in Lilium callosum III. Preferential segregation of a supernumerary chromosome in EMCs. Proc Jpn Acad 33:553–558 Khairwal IS, Hash CT (2007) HHB 67-improved – The first product of marker-assisted crop breeding in India. Asia-Pacific Consortium on Agricultural Biotechnology (APCoAB) e-News. http://www.apcoab.org/special_news.html Khalfallah N (1990) Genetic relationships among wild and cultivated forms belonging to the primary genepool of

252 pearl millet Pennisetum typhoı¨des Stapf et Hubb: Assessment of variability using combine cytogenetic and biometrical approaches. Doctor of Science Thesis. Constantine, Algeria 75 p (Les relations ge´ne´tiques entre formes sauvages et cultive´es du pool primaire du mil, Pennisetum typhoı¨des Stapf et Hubb. Analyses cytoge´ne´tiques et biome´triques conjointes de l’organisation de la variabilite´. The`se de doctorat d’Etat Constantine, Algeria 75 p) Khalfallah N, Sarr A, Barghi N, Siljak-Yakovlev S (1988) Evidence of karyotypic particulatities in Pearl Millet (Pennisetum typhoı¨des Stapf & Hubb) and their incidence on chromosomal behaviour. Abstract of XVIth International Congress of Genetics, August 20–27, Toronto, Canada. Genome 30 (Suppl 1):284 Khalfallah N, Sarr A, Siljak-Yakovlev S (1993) Karyological study of some cultivated and wild stocks of pearl millet from Africa (Pennisetum typhoı¨des Stapf et Hubb. and P. violaceum (Lam) L. Rich.). Caryologia 46(2–3):127–138 Khan ZR, Amudavi DM, Midega CAO, Wanyama JM, Pickett JA (2008) Farmers’ perceptions of a ‘push–pull’ technology for control of cereal stemborers and Striga weed in western Kenya. Crop Prot 27:976–987 Klee M, Zach B, Stika HP (2004) Four thousand years of plant exploitation in the Lake Chad basin (Nigeria), part III: plant impressions in potsherds from the final stone age Gajiganna culture. Veg Hist Archaeobot 13:131–142 Klichowska M (1978) Preliminary results of palaeoethnobotanical studies on plant impressions on potsherds from the Neolithic settlement at Kadero, Sudan. Nyame Akuma 12:42–43 Kulkarni VN, Rai KN, Dakheel AJ, Ibrahim M, Hebbara M, Vadez V (2006) Pearl millet germplasm adapted to saline conditions. ISMN 47:103–106 Lamy F, Martel E, Ricroch A, Robert T, Sarr A (1994) An integrated strategy, including the use of RFLP markers, to optimise the use of genetic resources of the primary gene pool of pearl millet. In: Witcombe JR, Duncan RR (eds) Use of molecular markers in sorghum and pearl millet breeding for developing countries. Overseas Development Administration, London, UK, pp 86–89 Le Corre V, Kremer A (2003) Genetic variability at neutral markers, quantitative trait loci and trait in a subdivided population under selection. Genetics 164:1205–1219 Le Thi K, Lespinasse R, Siljaj-Yakovlev S, Robert T, Khalfallah N, Sarr A (1994) Karyotypic modifications in androgenetic plantlets of pearl millet, Pennisetum glaucum (L.) R. Brunken: occurrence of B chromosomes. Caryologia 47(1):1–10 Lespinasse R, De Paepe R, Koulou A (1987) Induction of B chromosomes formation in androgenetic lines of Nicotiana sylvestris. Caryologia 40(4):327–338 Li CB, Zhou AL, Sang T (2006) Rice domestication by reducing shattering. Science 311:1936–1939 Liu CJ, Witcombe JR, Pittaway TS, Nash M, Busso CS, Hash CT, Gale MD (1994) An RFLP-based genetic map of pearl millet (Pennisetum glaucum). Theor Appl Genet 89:481–487 Marchais L (1994) Wild pearl millet population (Pennisetum glaucum, Poaceae) integrity in agricultural Sahelian areas. An exemple from Keita (Niger). Plant Syst Evol 189: 233–245

T. Robert et al. Marchais L, Perne`s J (1985) Genetic divergence between wild and cultivated pearl millets (Pennisetum typhoides). I-Male sterility. Z Pflanzenz€ucht 95:103–112 Marchais L, Tostain S (1997) Analysis of reproductive isolation between pearl millet (Pennisetum glaucum (L.) R.Br. and P. ramosum, P. schweinfurthii, P. squamulatum, Cenchrus ciliaris. Euphytica 93:97–105 Mariac C, Robert T, Allinne C, Remigereau MS, Luxereau A, Tidjani M, Seyni O, Bezancon G, Pham JL, Sarr A (2006) Genetic diversity and gene flow among pearl millet crop/ weed complex: a case study. Theor Appl Genet 113 (6):1003–1014 Martel E, Ricroch A, Sarr A (1996) Assessment of genome organization among diploid species (2n¼2x¼14) belonging to primary and tertiary gene pools of pearl millet using fluorescent in situ hybridization with rDNA probes. Genome 39(4):680–687 Martel E, De Nay D, Siljak-Yakovlev S, Brown S, Sarr A (1997) Genome size variation and basic chromosome number in pearl millet and fourteen related Pennisetum species. J Hered 88:139–143 Martel E, Poncet V, Lamy F, Siljak-Yakovlev S, Lejeune B, Sarr A (2004) Chromosome evolution of Pennisetum species (Poaceae): implication of ITS phylogeny. Plant Syst Evol 249:139–149 McCouch SR, Sweeney M, Li J, Jiang H, Thomson M, Septiningsih E, Edwards J, Moncada P, Xiao J, Garris A, Tai T, Martinez C, Tohme J, Sugiono M, McClung A, Yuan LP, Ahn SN (2006) Through the genetic bottleneck: O. rufipogon as a source of trait-enhancing alleles for O. sativa. Euphytica 154:317–339 Miura R, Terauchi R (2005) Genetic control of weediness traits and the maintenance of sympatric crop-weed polymorphism in pearl millet (Pennisetum glaucum). Mol Ecol 14(4): 1251–1261 Morgan RN, Wilson JP, Hanna WW, Ozais-Akins P (1998) Molecular markers for rust and pyricularia leaf spot disease resistance in pearl millet. Theor Appl Genet 96:413–420 Morjan CL, Rieseberg LH (2004) How species evolve collectively: implications of gene flow and selection for the spread of advantageous alleles. Mol Ecol 13:1341–1356 M€untzing A (1958) Accessory chromosomes. Trans Bose Res Inst 22:1–15 Mutegi JK, Mugendi DM, Verchot LV, Kung’u JB (2008) Combining Niaper grass with leguminous schrubs in contour hedgerows controls soil erosion without competing with crops. Agrofor Syst 74:37–49 Niangado O (1981) Backcross strategy in pearl millet Pennisetum americanum (L.) Leeke improvement: use for cycle adjustement and the enhancement of genetic resources using wild forms of Pennisetum. PhD Thesis, University of Paris-XI, France, (Utilisation de re´trocroisements chez le mil: 1. Pour changer le re´gime de floraison. 2. Pour exploiter la variabilite´ ge´ne´tique des formes spontane´es. PhD thesis, Universite´ Paris-XI, Orsay, France) Noyes RD (2008) Sexual devolution in plants: apomixis uncloaked? Bioessays 30:798–801 Nur U (1977) Maintenance of a ‘parasitic’ B chromosome in the grasshopper Melanoplus femur-rubrum. Genetics 87: 499–512

13 Pennisetum ¨ stergren G (1947) Heterochromatic B chromosomes in AnthoxO anthum. Hereditas 33:261–296 Oumar I, Mariac C, Pham JL, Vigouroux Y (2008) Phylogeny and origin of pearl millet (Pennisetum glaucum [L.] R. Br) as revealed by microsatellite loci. Theor Appl Genet 117:489–497 Owino JO, Owido SFO, Chemelil MC (2006) Nutrients in runoff from a clay loam soil protected by narrow grass strips. Soil Till Res 88:116–122 Ozias-Akins P (2006) Apomixis: developmental characteristics and genetics. Crit Rev Plant Sci 25:199–214 Ozias-Akins P, van Dijk PJ (2007) Mendelian genetics of apomixis in plants. Annu Rev Genet 41:509–537 Ozias-Akins P, Roche D, Hanna WW (1998) Tight clustering and hemizygosity of apomixis-linked molecular markers in Pennisetum squamulatum implies genetic control of apospory by a divergent locus which may have no allelic form in sexual genotypes. Proc Natl Acad Sci USA 95:5127–5132 Ozias-Akins P, Akiyama Y, Hanna WW (2003) Molecular characterization of the genomic region linked with apomixis in Pennisetum/Cenchrus. Funct Integr Genom 3:94–104 Pantulu JV (1960) Accessory chromosomes in Pennisetum typhoides. Curr Sci 29:28–29 Pantulu JV, Manga V (1975) Influence of B-chromosomes on meiosis in pearl millet. Genetica 45:237–251 Papa R, Gepts P (2003) Asymmetry of gene flow and differential geographical structure of molecular diversity in wild and domesticated common bean (Phaseolus vulgaris L.) from Mesoamerica. Theor Appl Genet 106(2):239–250 Patil BD, Hardas MW, Joshi AB (1961) Auto-alloploid nature of Pennisetum squamulatum Fresen. Nature 189:419–420 Peng J, Ronin Y, Fahima T, Ro¨der MS, Li Y, Nevo E, Korol A (2003) Domestication quantitative trait loci in Triticum dicoccoides, the progenitor of wheat. Proc Natl Acad Sci USA 100:2489–2494 Perne`s J (1983) Some genetic statements on domestication process of cereals. La Recherche 146:910–919 (Points de vue ge´ne´tiques sur la domestication des ce´re´ales. La Recherche 146:910–919) Perne`s J (1984) Plant genetic resources management, vol 1. ACCT, Paris, France (Gestion de ressources ge´ne´tiques des plantes. vol 1. ACCT, Paris, France) Perne`s J (1985) Crop plant evolution: the case of cereals. Life Sci Proc Gen Ser II 5:429–447 (Evolution des plantes cultive´es: l’exemple des cereales.La Vie des Sciences Comptes rendus se´rie ge´ne´rale 5: 429–447) Perne`s J (1986) Outbreeding and domestication process in cereals: the case of maize (Zea mays L. and pearl millet Pennisetum americanum L. K. Schum). Bull Soc Bot Fr 133 (1):27–34 (L’allogamie et la domestication des ce´re´ales: l’exemple du maı¨s (Zea mays L.) et du mil (Pennisetum americanum L.) K. Schum. Bull Soc Bot Fr 133(1): 27–34) Petit-Maire N (2002) Sahara: under the sand . . .lakes. A journey through time. CNRS, Paris, France (Sahara: sous le sable... des lacs. Un voyage dans le temps. CNRS edn, Paris, France) Pilate-Andre´ S (1992) Study of genetic diversity organization within Pennisetum complexe of species using allozymes markers and adh gene region molecular analysis. PhD Thesis, University of Paris-XI, Orsay, France, 220 p (Etude de l’organisation de la diversite´ ge´ne´tique du complexe des mils pe´nicillaires (Pennisetum spp.) par les marqueurs

253 enzymatiques et par l’analyse mole´culaire de la re´gion Adh. Univ Paris-Sud, Orsay, France, 220 p) Pilger RKF (1940) Gramineae III. Unterfamilie Panicoideae. In: Engler A, Prantl K (eds) Die Nat€urlichen Pflanzenfamilen, edn 2, vol 14e. Engelmann, Leipzig, Germany, pp 1–208 Poncet V, Lamy F, Enjalbert J, Joly H, Sarr A, Robert T (1998) Genetic analysis of the domestication syndrome in pearl millet (Pennisetum glaucum L., Poaceae): inheritance of the major characters. Heredity 81:648–658 Poncet V, Lamy F, Devos K, Gale M, Sarr A, Robert T (2000) Genetic control of domestication traits in pearl millet (Pennisetum glaucum L., Poaceae). Theor Appl Genet 100:147–159 Poncet V, Martel E, Allouis S, Devos KM, Lamy F, Sarr A, Robert T (2002) Comparative analysis of QTLs affecting domestication traits between two domesticated  wild pearl millet (Pennisetum glaucum L., Poaceae) crosses. Theor Appl Genet 104:965–975 Porte`res R (1950) Vieilles agricultures de l’Afrique intertropicale; centre d’origine et de diverisfication varie´tale primaire et berceaux d’agriculture ante´rieurs au XVIe`me sie`cle. Agron Trop 5:489–507 Prescott-Allen C, Prescott-Allen R (1986) The first resource: wild species in the North American economy. Yale University, New Haven, USA Prescott-Allen C, Prescott-Allen R (1988) Genes from the wild: using wild genetic resources for food and raw materials. International Institute for Environment and Development, London, UK Puertas MJ (2002) Nature and evolution of B chromosomes in plants: a non-coding but information-rich part of plant genomes. Cytogenet Genome Res 96:198–205 Puertas MJ, Baeza F, De La Pena A (1986) The transmission of B chromosomes in populations of Secale cereale and Secale vavilovii 1. Offspring obtained from OB and 2B plants. Heredity 57:389–394 Puertas MJ, Jime´nez MM, Romera F (1993) Rye B chromosome transmission depends on the B, the carrier of the B and the mother of the carrier. In: Sumner AT, Chandley AC (eds) Chromosomes today, vol 11. Chapman and Hall, London, UK, pp 391–399 Puertas MJ, Gonzalez-Sanchez M, Manzanero S, Romera F, Jimenez MM (1998) Genetic control of the rate of transmission of rye B chromosomes. IV. Localization of the genes controlling B transmission rate. Heredity 80:209–213 Purugganan MD, Fuller DQ (2009) The nature of selection during plant domestication. Nature 457(12):843–848 Qi X, Lindup S, Pittaway TS, Allouis S, Gale MD, Devos KM (2001) Development of simple sequence repeat markers from bacterial artificial chromosomes without subcloning. Biotechniques 31(2):358–362 Qi X, Pittaway TS, Lindup S, Liu H, Waterman E, Padi FK, Hash CT, Zhu J, Gale MD, Devos KM (2004) An integrated genetic map and a new set of simple sequence repeat markers for pearl millet, Pennisetum glaucum. Theor Appl Genet 109 (7):1485–1493 Rai KN (1995) A new cytoplasmic-nuclear male sterility system in pearl millet. Plant Breed 114:445–447 Rai KN, Anand Kumar K, Andrews DJ, Rao AS (2001) Commercial viability of alternative cytoplasmic-nuclear male sterility system in pearl millet. Euphytica 121:107–114

254 Rai KN, Khairwal IS, Dangaria J, Singh AK,·Rao AS (2009) Seed parent breeding effciency of three diverse cytoplasmicnuclear male-sterility systems in pearl millet. Euphytica 165:495–507 Raman VS, Chandrasekharan P, Krishanaswami D (1959) A note on some chromosome numbers in Gramineae. Curr Sci 29:127–128 Rao YS, Rao SA, Mengesha MH (1989) New evidence on the phylogeny of basic chromosome number in Pennisetum. Curr Sci 58(15):869–871 Rao NK, Reddy LJ, Bramel PJ (2003) Potential of wild species for genetic enhancement of some semi-arid food crops. Genet Resour Crop Evol 50:707–721 Renno JF, Winkel T, Bonnefous F, Bezanc¸on G (1997) Experimental study of gene flow between wild and cultivated Pennisetum glaucum. Can J Bot 75:925–931 Renno JF, Mariac C, Poteaux C, Bezancon G, Lumaret R (2001) Haplotype variation of cpDNA in the agamic grass complex Pennisetum section Brevivalvula (Poaceae). Heredity 86:537–544 Rieseberg LH, Seung-Chul K, Randell RA, Whitney KD, Gross BL, Lexer C, Clay K (2007) Hybridization and the colonization of novel habitats by annual sunflowers. Genetica 129:149–165 Robert T, Sarr A (1992) Multivariate analysis of recombination between wild and cultivated genomes within the primary gene pool of pearl millet (Pennisetum typhoides). Genome 35:208–219 Robert T, Lespinasse R, Perne`s J, Sarr A (1991) Gametophytic competition as influencing gene flow between wild and cultivated forms of pearl millet (Pennisetum typhoides). Genome 34:195–200 Robert T, Lamy F, Sarr A (1992) Evolutionary role of gametophytic selection in the domestication of Pennisetum typhoides (pearl millet): a two-locus asymmetrical model. Heredity 69:372–381 Roche DR, Conner JA, Budiman MA, Frisch D, Wing R (2002) Construction of BAC libraries from two apomictic grasses to study the microcolinearity of their aposporyspecific genomic regions. Theor Appl Genet 104: 804–812 Sandmeier M, Beninga M, Perne`s J (1981) Analysis of the genetic relationships between wild and cultivated forms of pearl millet III Inheritance of esterases ansd peroxydase isozyme markers. Agronomie 1:486–494 (Analyse des relations entre formes spontane´es et cultive´es chez le mil a` chandelles. III- Etude de l’he´re´dite´ des este´rases et des peroxydases anodiques. Agronomie 1:486–494) Sarr A, Perne`s J (1988) Unravelling segregation distortions for quantitative traits in pearl millet (Pennisetum typhoı¨des (Burm) Stapf et Hubb using multivariate statistical analysis. Genome 30:411–422 (Analyses multivarie´es de descendances de re´trocroisements et mise en e´vidence de distorsions de se´gre´gation de caracte`res quantitatifs chez le mil (Pennisetum typhoides (Burm.) Stapf et Hubb.). Genome 30:411–422) Sarr A, Sandmeier M, Pernes J (1988) Gametophytic competition in pearl millet (Pennisetum Typhoı¨des (Burm) Stapf et Hubb.). Genome 30(6):924–929 Sarr A, Gale MD, Beninga M, Zangre R, Renno JF, Bezanc¸on G (2001) New strategy including the use of molecular markers

T. Robert et al. to enhance Pennisetum glaucum genetic resources (wild relatives and cultivated forms). Final report, EU Project (ERBTS3*CT940280) Savidan Y (2001) Transfer of apomixis through wide crosses. In: Savidan Y, Carman J, Dresselhaus T (eds) The flowering of apomixis: from mechanisms to genetic engineering. CIMMYT, Mexico, DF; IRD, European Commission DG VI (FAIR), pp 153–167 Schmelzer GH, Renno JF (1999) Genotypic variation in progeny of the agamic grass complex Pennisetum section Brevivalvula in West Africa. Plant Syst Evol 215:71–83 Senthilvel S, Jayashree B, Mahalakshmi V, Kumar PS, Nakka S, Nepolean T, Hash C (2008) Development and mapping of simple sequence repeat markers for pearl millet from data mining of expressed sequence tags. BMC Plant Biol 8:119 Shaw MW, Hewitt GM (1985) The genetic control of meiotic drive acting on the B-chromosome of Myrmeleotettix maculatus (Orthoptera: Acrididae). Heredity 54:187–194 Siljak-Yakovlev S (1996) La dysploı¨die et l’e´volution du caryotype. Bocconea 5:210–220 Slatkin M (1987) Gene flow and the geographic structure of natural populations. Science 236:787–792 Small E (1984) Hybridization in domesticated-weed-wild complex. In: Grant WF (ed) Plant biosystematics. Academic, Toronto, ON, Canada, pp 195–210 Sotomayor-Rios A, Pitman WD (2001) Tropical forage plants: development and use. CRC, Boca Raton, FL, USA Stapf O, Hubbard CE (1934) Pennisetum. In: Prain D (ed) Flora of tropical Africa, vol 9, Part 6. Reeve, Ashford, Kent, UK, pp 954–1070 Stebbins GL (1956) Cytogenetics and evolution in the grass family. Am J Bot 43:890–905 Subba Rao MV (1980) Inherance of B-chromosomes in pearl millet. Heredity 45(1):1–6 Subba Rao MV, Pantulu JV (1978) The effects of derived B-chromosomes on meiosis in pearl millet Pennisetum typhoides. Chromosoma (Berl) 69:121–130 Tanksley SD, McCouch SR (1997) Seed banks and molecular maps: unlocking genetic potential from the wild. Science 277:1063–1066 Techio VH, Davide LC, Pereira AV (2006) Meiosis in elephant grass (Pennisetum purpureum), pearl millet (P. glaucum) (Poaceae, Poales) and their interspecific hybrids. Gen Mol Biol 29(2):353–362 Tostain S (1992) Enzyme diversity in pearl millet (Pennisetum glaucum), wild millet. Theor Appl Genet 83:733–742 Tostain S (1993) Evaluation de la diversite´ ge´ne´tique des mils pe´nicillaires diploı¨des (Pennisetum glaucum (L.) R.Br.) au moyen de marqueurs enzymatiques. Etude des relations entre formes sauvages et cultive´es. Orsay, France, p 331 Tostain S, Marchais L (1989) Enzyme diversity in pearl millet (Pennisetum glaucum) Africa and India. Theor Appl Genet 77:634–664 Tostain S, Riandey MF, Marchais L (1987) Enzyme diversity in pearl millet (Pennisetum glaucum). West Africa. Theor Appl Genet 74:188–193 Venkateswarlu J, Pantulu JV (1970) The cytological behaviour of B-chromosomes in Pennisetum typhoides. Cytologia 35:444–448

13 Pennisetum Visser NC, Spies JJ, Venter HJT (1998) Aneuploidy in Cenchrus ciliaris (Poaseae, Panicoideae, Paniceae): truth or fiction? S Afr J Bot 64:337–345 Vom Brocke K, Christinck A, Weltzien ER, Presterl T, Geiger HH (2003) Farmer’s seed systems and management practices determine pearl millet genetic diversity patterns in semiarid regions of India. Crop Sci 43:1680–1689 Wendorf F, Schild R (1994) Are the early Holocene cattle in the eastern Sahara domestic or wild? Evol Anthropol 3:118–128 Wendorf F, Schild R (1998) Nabta Playa and its role in Northeastern African prehistory. J Anthropol Archaeol 17:97–123 Wendorf F, Close AE, Schild R, Wasylikowa K, Housley RA et al (1992) Saharan exploitation of plants 8,000 years BP. Nature 359:721–724 Wetterstrom W (1993) The origins of agriculture in Africa: with particular reference to sorghum and pearl millet. Rev Archaeol – Spl Issue: the transition to agriculture in the Old World 19:30–46 Williams E, Barclay P C (1968) The effects of B chromosomes on vigor and fertility in Dactylis hybrids. NZ J Bot 6(4):405–416 Wilson JP, Hanna WW (1992) Disease resistance in wild Pennisetum species. Plant Dis 76:1171–1175 Wilson JP, Hess DE, Hanna WW (2000) Resistance to Striga hermonthica in wild accessions of the primary gene pool of Pennisetum glaucum. Phytopathology 90:1169–1172 Wilson JP, Hess DE, Hanna WW, Kumar KA, Gupta SC (2004) Pennisetum glaucum subsp. monodii accessions with Striga resistance in West Africa. Crop Prot 23:865–870 Wood D, Lenne´ JM (1997) The conservation of agrobiodiversity on-farm: questioning the emerging paradigm. Biodivers Conserv 6:109–129 Xiao J, Grandillo S, Ahn SN, McCouch SR, Tanksley SD, Li J, Yuan L (1996) Genes from wild rice improve yield. Nature 384:223–224 Yadav RS, Hash CT, Bidinger FR, Cavan GP, Howarth CJ (2002) Quantitative trait loci associated with traits determin-

255 ing grain and stover yield in pearl millet under terminal drought-stress conditions. Theor Appl Genet 104:67–83 Yadav RS, Bidinger FR, Hash CT, Yadav YP, Yadav OP, Bhatnagar SK, Howarth CJ (2003) Mapping and characterization of QTL  E interactions for traits determining grain and stover yield in pearl millet. Theor Appl Genet 106:512–520 Yadav RS, Hash CT, Bidinger FR, Devos KM, Howarth CJ (2004) Genomic regions associated with grain yield and aspects of post-flowering drought tolerance in pearl millet across stress environments and tester background. Euphytica 136:265–277 Zach B, Klee M (2003) Four thousand years of plant exploitation in the Chad Basin of NE Nigeria II: discussion on the morphology of caryopses of domesticated Pennisetum and complete catalogue of the fruits and seeds of Kursakata. Veg Hist Archaeobot 12:187–204

Databases Clayton WD, Harman KT, Williamson H (2006 onwards) GrassBase – The Online World Grass Flora: http://www.kew.org/ data/grasses-db.html (accessed 08 March 2009) Index to Plant Chromosome Numbers Data Base, Missouri Botanical Garden-w3TROPICOS: http://mobot.mobot.org/ W3T/search/ipcn.html (accessed 26 Feb 2009) Botanica sistematica Italia: http://www.homolaicus.com/ scienza/erbario/utility/botanica_sistematica/hypertext/1524. htm#000000 (accessed 21 April 2009) Bennett MD, Leitch IJ (2005) Angiosperm DNA C-values database (release 4.0), Royal Botanic Gardens, Kew: http:// www.rbgkew.org.uk/cval/homepage.html (accessed 28 Feb 2009)

.

Chapter 14

Phleum Alan V. Stewart, Andrzej J. Joachimiak, and Nicholas W. Ellison

14.1 Introduction The genus Phleum contains one major perennial forage species, Timothy (P. pratense) and approximately 13 other species. Timothy is commonly used as a forage and hay crop in cold temperate regions of the world, where it is noted for its winter hardiness and high forage quality. Phleum species have only rarely been crossed with species outside this genus and they are not closely related to any major crop. Therefore, the story of wild relatives for Phleum is largely one of species and genomic diversity within the genus. Although the distribution and diversity of Phleum is complex, much of this can be readily explained by its evolutionary, migrational and genomic history over the last few hundred thousand years. The understanding of the diversity within wild relatives of Timothy has considerable implications for germplasm collection and conservation as well as for breeding techniques employed.

14.2 Agricultural Status Phleum is the third most important forage grass genus of the temperate world following Lolium and Festuca. Seed production of Timothy makes up over 8% of world production of temperate forage species (Bondesen 2007). Its production and use is concentrated in the

A.V. Stewart (*) PGG Wrightson Seeds, PO Box 175, Lincoln, Christchurch 7640, New Zealand e-mail: [email protected]

northern latitudes of Europe and North America with approximately 34,000 ton produced each year (Anonymous 2007). Use of Timothy in the southern hemisphere is minimal with the major user, New Zealand, using less than 100 ton annually (Charlton and Stewart 2000; AgriQuality 2007). Outside its original Eurasian range, P. pratense has now expanded to most suitable habitats around the world. It is now widespread in North America, where it is both a major and valuable forage species and, along with many introduced grasses, is of ecological concern for displacing native grasses.

14.3 Morphology and Flowering Behavior Timothy is a perennial bunchgrass that grows up to 1.5m tall with flat leaves and is characterized by corms, bulb-like swellings of stem bases that serve as storage organs for carbohydrate reserves. New tillers form from axillary buds at the base of the corms, resulting in a relatively non-aggressive spreading habit. At times elongated stems trail along the ground and root at the nodes, thus exhibiting a semi-stoloniferous habit. This is particularly apparent in diploid forms from central Europe (Kovats 1976), but it is also common in P. pratense forms as a result of high temperature preventing reproductive initiation but not internode elongation (Cooper 1958). Commercial hexaploid Timothy lacks a vernalization requirement for flowering unlike many other temperate grasses. It, therefore, has no strong mechanism to prevent reproductive development during summer

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_14, # Springer-Verlag Berlin Heidelberg 2011

257

258

and the growing point becomes elevated and vulnerable to removal by grazing. This makes the plant susceptible to grazing damage in summer. However, a vernalization response is known in diploid germplasm (Cooper and Calder 1964) and some Mediterranean hexaploid germplasm, but breeders have not introduced this feature into commercial cultivars.

14.4 Taxonomy Species of the genus Phleum have been divided into four sections as outlined by Joachimiak 2005 (modified from Humphries 1980; Dogan 1988, 1991; Dogan and Us 1996).

14.4.1 Section Phleum Today eight “entities” from diploid to octoploid are recognized in three species, two perennials and one annual (Joachimiak 2005; Kula 2005a; Stewart et al. 2008). In this review, we use the widely accepted nomenclature of Humphries (1978, 1980) for European species and of Barkworth et al. (2007) for American species as follows:

14.4.1.1 P. alpinum L. P. alpinum L. is an alpine species differentiated into diploid or tetraploid cytotypes with ciliate or glabrous awns: 1. A glabrous awned tetraploid; P. alpinum L.
P. commutatum Gaudin. with a circumpolar northern hemisphere and South American distribution. 2. A ciliate awned diploid form, known from the Rhaetic Alps of Italy and the Balkans; P. alpinum subsp. rhaeticum Humphries
P. rhaeticum (Humphries) Rauschert. 3. A glabrous awned diploid, also known as P. alpinum L., which cannot be differentiated morphologically from the widespread tetraploid; for ease of presentation we will refer to this by the informal name “commutatum” following Joachimiak and Kula (1993).

A.V. Stewart et al.

14.4.1.2 P. pratense L. P. pratense L. is a lowland species represented by a diploid to octoploid polyploid series 1. Diploids occur throughout much of Europe and parts of North Africa; P. pratense subsp. bertolonii (DC.) Bornm.
P. bertolonii DC.
P. nodosum L. 2. Less common tetraploid forms in southern Europe; P. pratense subsp. pratense. 3. Widespread agricultural hexaploid forms; P. pratense subsp. pratense. 4. An octoploid form restricted to southern Italy; P. pratense subsp. pratense.

14.4.1.3 Phleum echinatum Host. Phleum echinatum Host. is the most unusual species within the section Phleum as it is a winter active annual grass of eastern Mediterranean mountains.

14.4.2 Section Chilochloa (Beauv.) Dum. This contains four perennial and three annual species representing the polyploid series from diploid to octoploid.

14.4.2.1 P. phleoides (L.) Karsten, P. phleoides (L.) Karsten, is a perennial grass occurring in dry grasslands, often in calcareous soils in parts of Asia, Europe and Morocco (Fig. 14.1) at altitudes of 150–2,800 m (Dogan 1991). This species and P. hirsutum are seen as a core species of section Chilochloa (Joachimiak 2005). Diploid and tetraploid forms occur (Humphries 1980). This species exhibited little agricultural potential in New Zealand and was noted as having poor productivity, an open growth habit and an early flowering behavior with few seedheads (Caradus 1978).

14.4.2.2 P. hirsutum Honckeny P. hirsutum Honckeny is a diploid perennial of grasslands, open woods, stony places and mountain pastures

14 Phleum

259

Fig. 14.1 Distribution of P. phleoides (after Conert 1998)

at 900–2,200 m in South and Central Europe from Bulgaria to Italy (Humphries 1980). This species exhibited little agricultural potential in New Zealand (tested as P. michelii All.) and was noted as having poor productivity with an early flowering behavior with few seedheads (Caradus 1978).

14.4.2.3 P. arenarium L. P. arenarium L. is a diploid annual on sand dunes, dry limestone verges from 0 to 1,700 m, distributed in southern and western regions of Europe as well as Morocco (Humphries 1980). This species exhibited little agricultural potential in New Zealand and was noted as having poor productivity but had an early flowering behavior with a prolific number of seedheads (Caradus 1978).

14.4.2.4 P. montanum C Koch P. montanum C Koch is a variable perennial grass of dry mountain slopes in southern and eastern Europe and Asia. Its chromosome numbers are recorded as 2x ¼ 14, 4x ¼ 28, and 6x ¼ 42 and it is divided into two subspecies: Subsp. montanum occurs on dry mountain slopes, grassy igneous sandstone slopes, edges of beech forests

at 1,500–2,000 m in Asia from Russia to Turkey, in North Africa in Morocco and in Europe from Bulgaria and Spain. Subsp. serrulatum (Boiss) Dogan occurs on mountain slopes, in tree plantations, mixed deciduous woodlands on sandy soils at 430–2,800 m, in Asia from Iran to Turkey and in Europe from Bulgaria to Italy. This species exhibited little agricultural potential in New Zealand and was noted as having poor productivity with an early flowering behavior with few seedheads (Caradus 1978). 14.4.2.5 P. paniculatum Hudson P. paniculatum Hudson is a tetraploid annual (Humphries 1980), which is commonly divided into two subspecies: Subsp. ciliatum (Boiss.) Dogan occurs on mountain slopes, sandy places and dried mud from 30 to 1,800 m, distributed in Europe from Greece to Britain. Subsp. paniculatum occurs in fields, dry rocky slopes and open places, often on limestone, 300–2,440 m, and is distributed in Asia from Iran to Turkey. 14.4.2.6 P. himalaicum Mez. P. himalaicum Mez. is an octoploid (8x ¼ 56) annual on mountain slopes and steep rocky hillsides,

260

1,300–2,200 m, distributed from the Himalayas of India through to Afghanistan (Joachimiak 2005).

14.4.2.7 P. iranicum Bornm. et Gauba P. iranicum Bornm. et Gauba is a perennial with creeping rhizomes occurring on open slopes, stony clays from 2,130 to 2,800 m, distributed in Iran. This species exhibited little agricultural potential in New Zealand and was noted as having poor productivity, and an early flowering behavior with few seedheads (Caradus 1978).

A.V. Stewart et al.

80 to 1,650 m, distributed from Iran to Britain. The chromosome size of this form was 1.6 pg (studied as P. graecum Kula 2005b). Subsp. aegaeum (Vierh.) Greuter occurs on sand dunes at sea level, distributed from Palestine to Greece.

14.4.4 Section Maillea (Parl.) Horn af Rantzien This contains a single annual species P. crypsoides (d’Urv.) Hackel distributed on sandy soils near the coast in Cyprus, Italy, Greece, and Crete.

14.4.3 Section Achnodon (Nees) Griseb. This section contains three diploid annual species.

14.4.5 Cytology and Karyotype

14.4.3.1 P. subulatum (Savi) Asch. et Graelon.

Phleum has a basic chromosome number of seven. Species range from diploid (2n ¼ 14) to octoploid (2n ¼ 56), with the exception of P. echinatum, with a reduced chromosome number of 2n ¼ 10. Karyotypes of P. pratense and P. alpinum taxa are slightly asymmetric and composed of medium-sized chromosomes. The chromosome complement of P. echinatum is much bigger and highly asymmetric. Heterochromatin is mostly located near centromeres and telomeres. Joachimiak (2005) provides a comprehensive review of the karyotypes of the majority of these species. The genomes of species in the different sections appear to be substantially different since molecular probes designed to detect species in section Phleum failed to detect those of P. montanum in section Chilochloa (Cai and Bullen 1994).

P. subulatum (Savi) Asch. et Graelon. is a diploid species considered to be a core species for this section (Joachimiak 2005). It is divided into two subspecies: Subsp. subulatum occurs on sandy clay cliffs on coast, limestone mountain slopes, grasslands, oak and pine forest from 0 to 1,100 m, in the region from Turkey to Israel and Cyprus. Subsp. ciliatum (Boiss.) Humphries occurs on grassy slopes, lawns, stony ground near the coast, on limestone 0–700 m, in Greece and Crete

14.4.3.2 P. boissieri Bornm. P. boissieri Bornm. is a diploid occurring in oak scrub, rocky limestone slopes, sand dunes, fallow fields 450–1,300 m, from Afghanistan across to Britain (Dogan 1991). The behavior of B-chromosomes in this species has been studied by Joachimiak (1982).

14.4.3.3 P. exaratum Hochst. P. exaratum Hochst. is a diploid of dry open habitats divided into two subspecies: Subsp. exaratum Hochst. Ex Griseb. occurs in open places, pine forest, saltflats and calcareous soils from

14.4.6 Genome Size Phleum has a relatively small genome size and seems particularly prone to developing higher ploidies (Kellogg and Bennetzen 2004). In the majority of species the basal genome size (1Cx) does not exceed 2 pg (Sliwinska et al. 2003; Kula 2005a). The only exception is P. echinatum with 1Cx ¼ 3.64 pg. Sliwinska et al. (2003) showed that as expected some DNA mass

14 Phleum

is lost with increased ploidy so that tetraploids have slightly less than expected for twice the diploid, an effect due to the partial loss of heterochromatin (Joachimiak 2005). Section Chilochloa contains at least two groups based on genome size, P. hirsutum (1.9 pg) and P. phleoides 1.8 pg are large, while both P. arenarium and P. montanum (1.4 pg) are smaller (Kula 2005b). The genome size of P. subulatum (1.9 pg) in section Achnodon is similar to that of P. phleoides and P. hirsutum (Kula 2005b).

261

– Autotetraploid pratense in France BSBSBNBN – Allotetraploid pratense in the Italian Alps BNBN RGRG – Tetraploid hybrid of bertolonii and hexaploid pratense BNBNBNRG – Common agricultural hexaploid pratense BNBNBN BNRGRG – Hexaploid pratense in southern Italy RGRGXXXX – Hexaploid pratense in Morocco BSBSXXXX – Octoploid pratense in southern Italy R8R8XXXX XX

14.4.7.3 P. echinatum

14.4.7 Genomic Formula P. echinatum has been assigned EE. Genomic formulae have been assigned to 22 genomic forms within section Phleum (Fig. 14.2). 14.4.7.4 Other Sections 14.4.7.1 P. alpinum – – – – – – – – – – –

Ancestral paleo-diploid rhaeticum in Asia RARA Diploid rhaeticum in the Alps RSRS Diploid rhaeticum in the Pyrenees RPRP Diploid rhaeticum in Italy RIRI Diploid rhaeticum in Greece RGRG Diploid “commutatum” in the Carpathian mountains CC Diploid hybrids of rhaeticum and “commutatum” RC Tetraploid rhaeticum “commutatum” hybrids, Italy RSRSCC Tetraploid rhaeticum “commutatum” hybrids, Caucasus CCRGRG Tetraploid alpinum of Europe across to Iceland REREXX Tetraploid alpinum East Asia, the Americas, to Iceland RWRWXX

14.4.7.2 P. pratense – Diploid subsp. bertolonii in northern Europe BNBN – Diploid subsp. bertolonii in Spain and Portugal BSBS – Diploid subsp. bertolonii in Greece and the Balkans BGBG

The only species from any other section for which it is possible to assign a partial genomic formula is RARAXX for P. paniculatum of section Chilochloa.

14.5 Distribution in Relation to Historical Glaciations and Migrations In order to understand the distribution and relationships among species in section Phleum, it is important to understand how historical glaciation events have driven migrations that allowed a range of entities to come into contact and hybridize and create today’s allopolyploids. Using molecular sequences, it has been possible to reconstruct the origin of the species within section Phleum (Stewart et al. 2008). The chloroplast trnL intron and the nuclear internal transcribed spacer (ITS) sequences have provided a valuable methodology for tracking the adaptive radiation of diploid Phleum species as well as up to two of the genomes in polyploid forms. The results show that an ancestor of P. alpinum subsp. rhaeticum RARA is the progenitor of section Phleum. The widespread circumpolar allotetraploid P. alpinum RRXX is a hybrid of this paleogenome with

262

A.V. Stewart et al.

Deschampsia cespitosa P. hirsutum, P. phleoides, P. iranicum

Ancestral rhaeticum central Asia RARA

P. paniculatum RARAXX (1 unknown genome)

ITS mutation event

Chloroplast mutation event

rhaeticum Alps of Austria, Italy, Switzerland RSRS

alpinum 4x Japan, America RWRWXX (1 unknown genome)

alpinum 4x Europe, Iceland REREXX (1 unknown genome)

alpinum 4x Iceland

rhaeticum, Alps of Italy and Austria

alpinum 4x Taiwan

paternal

alpinum 4x Italian Alps rhaeticum Greece RGRG rhaeticum Greece

diploid RC and tetraploid RRCC hybrids of rhaeticum and "commutatum" , Caucusus, Carpathian and Altau Mountains

awns become glabrous

diploid "commutatum" CC Austria, Romania

4x pratense, Italy BBRR and commercial 6x pratense BBBBRR maternal

diploid "commutatum" Austria, Switzerland tetraploid hybrid of 2x bertolonii BB and 6x commercial pratense BBBR Alps Italy

tetraploid BNBNBNBN bertolonii France

maternal

reduction in centromeric heterochromatin and awn length

bertolonii BSBS Portugal, Spain, and 6x pratense Morocco BSBSXXXX

bertolonii BNBN Northern Europe hubbardii Poland

echinatum EE Mediterranean mountains

8x pratense Southern Italy (3 unknown genomes) R8R8XXXXXX

reduction in telomeric heterochromatin and anther length paternal

rhaeticum Pyrenees RPRP

4x pratense Southern France BNBNBSBS

Fig. 14.2 The development of Phleum species from an Asian ancestor of diploid P. alpinum subsp. rhaeticum RARA showing ITS and cpDNA mutations, hybridizations and polyploidizations

14 Phleum

an unknown genome (probably another Phleum species from a different section). The three other diploids, P. alpinum form “commutatum” CC, P. echinatum EE, and P. pratense subsp. bertolonii BB, are all derived from P. alpinum subsp. rhaeticum. In each case these have undergone molecular, cytological and morphological changes as well as changes in adaptation to environmental conditions.

14.5.1 Phleum alpinum P. alpinum L. occurs in most mountains of Europe, northern Asia and North and South America. At 30 latitudes, it occurs at altitudes over 4,000 m but this reduces to sea level at 60 . Tetraploids are widespread, while diploids are restricted to Europe (Fig. 14.3). P. alpinum is considered to be a glacial relict and

Fig. 14.3 Distribution of P. alpinum complex (Conert 1998)

263

would have had a more continuous distribution at lower altitudes during the most recent glaciation event (Thorn 1960).

14.5.1.1 Diploid Phleum alpinum On the basis of morphology and molecular form, diploid P. alpinum can be divided into three groups, subsp. rhaeticum RR, “commutatum” CC, and their hybrids RC.

Subsp. rhaeticum RR The diploid subsp. rhaeticum RR is recognized by the presence of ciliate awns, a feature where intermediate forms can make this distinction difficult. Subsp. rhaeticum is the dominant species in Switzerland, but it

264

occurs throughout the European Alps RSRS and in most central and southern European mountain ranges including the Pyrenees RPRP and the Balkans RGRG. It usually occurs in the subalpine and alpine belt from 1,000 to 2,500 m in fertile and humid habitats (Zernig 2005). The molecular results show that rhaeticum has migrated out from a base population in the Alps along adjacent mountain chains to the Pyrenees, Apennines and the Balkans. Associated with these migrations are small changes in the genome as typified by the reduction in centromeric heterochromatin observed in the Greek molecular form (Kula 2005a). The rhaeticum molecular form present in the Alps RSRS is the progenitor of all the forms within section Phleum, “commutatum” CC, bertolonii BB, and polyploid pratense. The notable exception to this is the widespread paleo-allotetraploid P. alpinum RRXX. In particular, the molecular form of rhaeticum from the eastern Alps and Greece identifies itself as an ancestor of all agricultural hexaploid pratense sharing the RGRG ITS sequence.

Form “commutatum” The diploid CC identified by its glabrous awns is found in the Alps and north into Germany and the Czech Republic, as well as the Carpathian mountains from Poland to Romania and Sweden (Joachimiak and Kula 1996). Diploid “commutatum” CC originates from rhaeticum of the Alps RSRS but differs from it by a chloroplast DNA insertion glabrous awns, reduced telomeric heterochromatin (Joachimiak 2005), reduced anther length and by growing at higher altitudes among the snow-bed vegetation (Zernig 2005). From the geographic location of molecular derivatives, it is apparent that there has been a general northern and eastern radiation outwards from the Alps towards Germany and the Carpathian mountains of Poland and Romania.

A.V. Stewart et al.

Of the 21 hybrids discovered, five were a new tetraploid form.

14.5.1.2 Tetraploid Phleum alpinum On the basis of morphology and molecular form, tetraploid P. alpinum can be divided into two groups, a widespread allotetraploid RRXX, as well as the tetraploid hybrids of “commutatum” CC, rhaeticum RR and their hybrids RRCC or CCRR.

Tetraploid Euro-American P. alpinum RRXX This form has the widest distribution of any of the Phleum species. Not only is it present on many of the mountains of Europe and North Asia, it is the only species of the genus to successfully migrate to both North and South America (Conert 1998). The molecular data reveal a central Asian origin with a divergent migration east to coastal Asia, Japan and to the Americas RWRWXX and west into Europe REREXX. Once in America it migrated rapidly to South America. A recent divergence of the European and American forms is supported by the lack of significant karyological differences between them (Kula et al. 2006) and only minor ecotypic differences (Heide and Solhaug 2001). Its circumpolar migration was completed in Iceland where derivatives of both the European and American molecular forms occur. However, we cannot discount the possibility that the European forms may have been introduced into Iceland with human colonization. The formation of its allotetraploid ancestor RRXX is likely to have been in Asia over 300000 years BP, before it diverged into the two forms. The diploid behavior of the species and the difficulty in crossing with P. pratense also suggest an ancient formation of this allotetraploid (No¨rdenskiold 1945).

Hybrids Between rhaeticum and “commutatum”

Tetraploid Hybrids Between rhaeticum and “commutatum” RRCC, CCRR

Of the European diploid P. alpinum analyzed, approximately a third each were rhaeticum RR, “commutatum” CC, and their hybrids RC or CR. A high frequency of hybridization has occurred over time.

We report a new form of tetraploid P. alpinum RRCC and CCRR based on hybrids of rhaeticum and “commutatum,” which adds further complexity to the many P. alpinum forms. Five populations were discovered,

14 Phleum

one in the European Alps, two in the Caucasus Mountains and two in the subalpine meadows of the Altau Mountains. This suggests these hybrids are part of a continuing eastward migration, extending at least as far as Kazakhstan.

14.5.2 P. pratense P. pratense is a lowland species distributed naturally throughout Europe, parts of North Africa and Asia (Fig. 14.4). Hexaploids are used for agricultural purposes in all cool temperate regions of the world. 14.5.2.1 Diploid subsp. bertolonii Diploid subsp. bertolonii occurs throughout Europe often sympatric with hexaploid forms (Perny´ et al.

265

2008). Compared with the hexaploid, it is less common in northern areas, being absent from the Romanian and East Carpathian regions and more common in the south (Humphries 1980), except for Italy where it is uncommon (Cenci et al. 1984). It also occurs in the mountains of North Africa (Maire 1953). Løhde (1977) and Foerster (1968) describe taxonomic differences between diploid subsp. bertolonii and hexaploid P. pratense but Løhde concedes that the only reliable taxonomic differences are chromosome number and ligule hairiness. Subsp. bertolonii BB is derived from P. alpinum subsp. rhaeticum RR and associated with this derivation is the loss of awns, loss of centromeric heterochromatin (Joachimiak 2005), and most importantly, a change in adaptation from subalpine to lowland conditions. Two major molecular forms exist within bertolonii, one restricted to Spain and Portugal, BSBS, and a

Fig. 14.4 Distribution of P. pratense diploids and hexaploids (after Conert 1998)

266

second widespread across northern Europe, BNBN. It is likely that these have diverged as a result of glaciation events in Europe. The northern European molecular form is likely to have reinvaded northern Europe after the last glaciation from a refuge in Italy or the Balkans. A particularly stoloniferous form of bertolonii in central Europe has been named Phleum hubbardii (Kovats 1977, 1980). Its molecular ITS sequence is derived from the northern European bertolonii BNBN, but differs with a mutation at one position. This minor difference and the lack of any significant differences in karyotype (Kula 2003) suggest that P. hubbardii is not a very well differentiated taxon.

14.5.2.2 Tetraploid P. pratense Tetraploid P. pratense has been reported to occur in central and northern Italy, France, Belgium, Spain, Poland, Slovakia and Bulgaria (Kozuharov and Petrova 1991; Uhrı´kova´ and Kra´lik 2000; Joachimiak 2005). We have found four molecular forms of tetraploid pratense, two different allotetraploids and two autotetraploids. Two allotetraploid pratense were discovered in the Alps, one a hybrid of bertolonii with rhaeticum BBRR and the other a hybrid of bertolonii with hexaploid pratense BBBR. One autotetraploid pratense from southern France has molecular characteristics of both northern European and Spanish bertolonii BSBSBNBN, while the second French population exhibits only northern European bertolonii BNBNBNBN molecular characteristics.

A.V. Stewart et al.

hybridization (GISH) studies also suggest the presence of two bertolonii genomes (Joachimiak unpublished) and on the basis of geographic origin these two bertolonii genomes are most probably both the northern European form making hexaploid pratense BNBNBNBNRGRG. Chromosome pairing in triploid plants of hexaploid pratense (7II þ 7I) (No¨rdenskiold 1945) suggests at least minor differences in the structure of the bertolonii B and rhaeticum R genomes. The small difference between these two genomes readily allows a synthesized auto-hexaploid bertolonii to cross with natural hexaploid P. pratense (No¨rdenskiold 1957). The uniformity of the molecular profile in agricultural P. pratense suggests that the formation of this hexaploid is probably post-glacial. Its distribution suggests it has expanded throughout Europe from a glacial refuge, most likely from the Balkans/Italy glacial refuge (Hewitt 1999). Two different hexaploid molecular forms were also found in southern Italy RGRGXXXX and Morocco BSBSXXXX. Our results show that hexaploid pratense is polyphyletic in origin, having formed at least three times from different diploid ancestors, a situation very common in polyploid species (Soltis and Soltis 2000).

14.5.2.4 Octoploid P. pratense Octoploid pratense R8R8XXXXXX is reported only from southern Italy (Cenci et al. 1984). Two samples with the very short stature of diploid bertolonii have the maternal genome derived from rhaeticum R, but it is not possible to determine from our results the origin of the other three genomes.

14.5.2.3 Hexaploid P. pratense Hexaploid pratense can be divided into three molecular forms. The most common is the agricultural hexaploid with the cytoplasmic molecular pattern of bertolonii BB and an ITS molecular pattern from the Balkans rhaeticum RGRG present from Greece to the eastern Alps. Although it is not possible from our results to determine the origin of the third genome, there is cytological and molecular evidence to suggest that this third genome is homologous with one of the two genomes (M€ untzing and Prakken 1940; Levan 1941; No¨rdenskiold 1945; Cai and Bullen 1991, 1994; Joachimiak 2005). More recent genomic in situ

14.5.3 P. echinatum The annual P. echinatum EE occurs in eastern Mediterranean mountains ranging from Sicily to Crete (Humphries 1980). Our results show that this species has developed from hybridization between two different derivatives of rhaeticum. Furthermore, it is likely that it originated from a single hybridization event as this species has undergone genetic reconstruction from the genus norm of 14 chromosomes to 10 (Ellestrom and Tijo 1950). Surprisingly, dysploid reduction from x ¼ 7 to x ¼ 5 was accompanied with the considerable

14 Phleum

increase of genome size in this species. The species exhibits some features of P. alpinum subsp. rhaeticum, but has a longer awn and reduced centromeric heterochromatin (Joachimiak 2005).

14.5.4 Migration History in Relation to Glaciation Events The molecular results show an Asian origin for the section Phleum and identify two separate migrations into Europe. The first migration into Europe was of an ancestor of diploid P. alpinum subsp. rhaeticum RR. The penultimate Riss glaciation 130000–150000 year BP provided ample opportunity for this alpine species to migrate vast distances through lowland areas to eventually become isolated on the Alps RSRS during subsequent warmer interglacial periods. Subsequent migration along mountain ranges has occurred so that today rhaeticum occurs in the Alps RSRS, Pyrenees RPRP, Apennines RIRI, and the Balkans RGRG.

267

Migration also occurred onto the colder mountain ranges to the north into Germany and to the Carpathian Mountains of Poland and Romania but was associated with micro-evolutionary changes in morphology and cytology to develop into diploid “commutatum” CC. The overlap of the range of rhaeticum and “commutatum” has since allowed considerable hybridization so that a swarm of hybrids RC or CR overlaps the range of “commutatum” and part of the rhaeticum range. Occasional tetraploid hybrids RRCC or CCRR have developed and have migrated east at least as far as Kazakhstan. Migration of rhaeticum populations back into the lowlands as a result of climate cooling eventually resulted in the first lowland species of this group, P. pratense subsp. bertolonii BB. This was also accompanied by micro-evolutionary changes in cytology, morphology and adaptation. As the climate cooled during the last glaciation (the W€urm 22000 to 13000 years BP) this lowland species retreated into southern European glacial refugia. Upon warming these subsequently reinvaded northern Europe from the Balkan/ Italy refugia as molecular form BNBN, with a second molecular form BSBS remaining restricted to the

Fig. 14.5 Generic glacial refugia of southern Europe (shaded areas) (after Hewitt 1999) and potential post-glacial migration route of diploid ssp. bertolonii and agricultural hexaploid pratense

268

Spanish glacial refuge. Hybridization occurred when these forms met at the interface in France resulting in a recent autotetraploid BSBSBNBN. Hybrids formed in the Italian Alps where subsp. bertolonii and the Balkans rhaeticum overlapped resulting in an allotetraploid pratense BNBNRGRG. It is probable that a further hybridization with the adjacent northern European subsp. bertolonii BNBN lead to the formation of the agricultural hexaploid pratense BNBNBNBNRGRG. Upon warming in the holocene these subsequently reinvaded northern Europe from the Balkan/Italy refugium, a refugium common to a wide range of European biota (Hewitt 1996, 1999) (Fig. 14.5). Two further hexaploid forms and an octoploid occur within glacial refugia, all based on local diploids but in all cases the remaining genomes are unknown, a hexaploid in southern Italy RGRGXXXX, another hexaploid in Morocco BSBSXXXX and an octoploid in southern Italy R8R8XXXXXX. Hybridization among rhaeticum forms led to the Mediterranean mountain annual P. echinatum EE with a reconstructed genome of only 10 chromosomes. This event probably took place prior to the last glaciation enabling it to spread throughout the Balkan/Italy glacial refuge and to subsequently recolonize the mountains of the eastern Mediterranean as the climate warmed. The very widespread allotetraploid P. alpinum formed over 300000 years BP in Asia from hybridization of an ancestral rhaeticum with another unknown genome, RARAXX. This form remained in Asia until eventually migrating into Europe during the last glaciation (the W€urm 22000 to 13000 years BP) when conditions were suitable. At the same time many species including this one were able to migrate into the Americas via the Bering/Aleutian route, although probably not completing their entry until this route became open around 8,000 years ago (Hong et al. 1999; Weber 2003). This divergent migration has lead to a divergence in molecular forms, one in northern Europe REREXX and the other in Japan and the Americas RWRWXX. The circumpolar migration was completed in Iceland where derivatives of both forms occur. Dogan (1991) describes the Mediterranean and western Asia as the center of origin for the genus Phleum. This remains true for P. pratense today,

A.V. Stewart et al.

although it is unexpected for many people that a species so common in high latitude cold temperate zones has originated within southern European glacial refugia and that these areas still retain high genetic diversity today. Although it has been suggested that northern Europe is a center of diversity (Guo et al. 2003), this appears unlikely as it has only been free of glaciation in the last 12,000 years and any diversity must be recent, or of migratory origin.

14.5.5 Gene Exchange Between Genomes and Inheritance Hexaploid P. pratense has many examples of tetradisomic or hexasomic inheritance (No¨rdenskiold 1953; Nielsen and Smith 1959; Nath 1967). This means that there is considerable gene exchange between genomes and that, while genomes can be identified and an allohexaploid genomic formula can be assigned BNBNBN BNRGRG, the genomes are not well differentiated. This is consistent with many studies concluding that P. pratense is largely an autopolyploid (Nielsen and Nath 1961; Nath 1967; No¨rdenskiold 1953, 1957, 1960; Wilton and Klebesadel 1973) and that autohexaploids developed from diploid P. pratense subsp. bertolonii are very similar to the agricultural hexaploid P. pratense (No¨rdenskiold 1949). From our new molecular knowledge on genomic origins of P. pratense hexaploids we now know that all three genomes are derived from the rhaeticum RSRS genome marooned in Europe around 150,000 years ago. The differentiation of this subalpine genome RSRS into the Balkans form RGRG and the lowland derivative BB genome may be much less than 50,000 years. For this reason there is only minor differentiation between these genomes and it is to be expected that P. pratense behaves largely as an autopolyploid. On the other hand the widespread paleo-allotetraploid P. alpinum RRXX had both greater differentiation between the genomes and a longer period of time, probably over 300,000 years, to become stabilized as an allotetraploid (No¨rdenskiold 1945). However other more recent tetraploid P. alpinum forms based on hybrids of rhaeticum and “commutatum” should behave more like autotetraploids.

14 Phleum

14.6 Wild Relatives as Genetic Resources for Phleum pratense P. pratense genetic resources can be divided into primary, secondary, tertiary and quaternary resources as follows.

14.6.1 Primary Gene Pool This is the major pool of hexaploid P. pratense germplasm used by breeders, consisting of those adapted cultivars, elite breeding lines and local ecotypes within the regions where the species is used. This is almost exclusively the common genomic form BNBNBNBNRGRG used in northern Europe, northern Asia and North America and to a much lesser extent New Zealand. The molecular diversity of the European lines exhibits a geographic pattern and derivatives in Asia, Americas and New Zealand can be related to these (Yang-Dong et al. 2003). The vast majority of hexaploid P. pratense accessions in genebanks, potentially over 99%, would be of this common genomic form. As there is considerable exchange of germplasm among breeders in different regions it would probably be fair to conclude that this pool of germplasm has been extensively exploited by breeders. Yet by and large it represents only the winterhardy northern genomic form of P. pratense and fails to utilize the material from the center of origin in southern regions of Europe and North Africa.

14.6.2 Secondary Gene Pool This pool includes hexaploid germplasm of P. pratense from the center of origin in the Balkans/ Italy region of southern Europe, the Mediterranean mountains and North Africa, a region largely outside the region of commercial use. It would also include other genomic constitutions of hexaploid P. pratense such as the southern Italian RGRGXXXX and Moroccan BSBSXXXX (potentially an autohexaploid BSBSBSBSBSBS). These regions have never been targeted for genebank collection because this material is not considered

269

sufficiently winter-hardy for use in more northern European climates. However, such material deserves special attention as it represents unique genomic hexaploid constitutions, and genetic erosion in these southern regions is occurring at an alarming rate as many in situ populations are under threat from climate warming and human-induced habitat degradation. Genebank material from these southern regions is very limited and we know of only one sample available from North Africa and only two from southern Italy. The senior author has found that the Moroccan BSBSXXXX and the southern Italian RGRGXXXX both readily cross with commercial hexaploid pratense BNBNBNBNRGRG.

14.6.3 Tertiary Gene Pool This consists of germplasm derived entirely from the original P. alpinum subsp. rhaeticum genome RARA but of ploidy levels other than hexaploid. P. pratense diploid BB, tetraploid BBBB and BBRR, octoploid RRXXXXXX as well as diploid rhaeticum RR, “commutatum” CC, and the diploid and tetraploid hybrids of rhaeticum and “commutatum” RC and RRCC and CCRR. The widespread paleo-allotetraploid P. alpinum RRXX would be excluded from this pool and classified in the quaternary pool. With the current knowledge of the genomic constitution of P. pratense and other forms it should be possible to either resynthesize P. pratense from different forms of the same genomes, or use genomes from other forms and ploidy levels of section Phleum for introgression into P. pratense hexaploids. While there do not appear to be any major barriers to hybridization between Phleum species or forms (Nath 1967), in practice the major barrier is one of ploidy and/or allopolyploidy. Crosses between diploid bertolonii and hexaploid P. pratense are easy to produce artificially (No¨rdenskiold 1945), and although not always tetraploid (Løhde 1977), BNBNBNRG are found in nature (M€untzing 1935; Foerster 1968, 2005). These readily cross back to hexaploid P. pratense. Hybrids between tetraploid P. pratense BNBNRG RG and the octoploid R8R8XXXXXX have been made by the senior author and these form fertile hexaploid plants. These plants have been crossed with regular agricultural hexaploids BNBNBNBNRGRG and these

270

also form fertile progeny. Some sterility is apparent from low seed yield and from ergot in the seed samples, a factor encouraged by the sterile florets remaining open for very long periods of time. Hybrids of tetraploid and hexaploid forms are pentaploid as expected (Nielsen and Nath 1961) and these may be backcrossed to hexaploids quite readily. The tertiary genepool potentially represents an enormous pool of unexplored material for breeders, but collections are necessary.

14.6.4 Quaternary Gene Pool By definition these would consist of the more difficultto-cross material such as allotetraploid P. alpinum RRXX as well as species in other sections, such as P. phleoides in section Chilochloa (of which there is an herbarium sample from Iceland in the University of Leiden, National Herbarium, Netherlands) and P. hirsutum in section Chilochloa (which the senior author has crossed with hexaploid P. pratense to form tetraploids), and P. subulatum of section Achnodon. The hybrid with P. subulatum formed a male sterile tetraploid morphologically similar to P. pratense (Myers 1941). There is also a report of a sterile cross between tetraploid Dactylis glomerata and hexaploid P. pratense (Nakazumi et al. 1997). At this stage the quaternary gene pool offers much less potential for breeders than the secondary and tertiary gene pools and collection and crossing efforts would be much better targeted at these.

14.6.5 The Conservation of Genetic Resources The conservation of a full range of genetic resources is pivotal to future improvements in P. pratense. Ultimately the best method of conservation would be in situ in the natural and agricultural world. However, in many regions Phleum species face an enormous threat from changing land use as well as climate change. In particular, these two factors place a serious threat to the many isolated mountain populations of P. alpinum and to P. pratense in the Mediterranean mountains of

A.V. Stewart et al.

former glacial refugia. Even in many of the northern European agricultural regions older fields are replaced as agriculture intensifies (Weibull et al. 2007). These changes make the ex situ conservation in genebanks and breeders’ seed collections very critical for the future.

14.6.6 Germplasm Banks Today there are approximately 120 cultivars of P. pratense certified under the OECD seed scheme around the world. There could be potentially another 30 or more outside this scheme as well as an estimated 50 older cultivars no longer used, making probably no more than 200 named cultivars of Timothy. Most of these cultivars are preserved in genebanks along with collections from old pastures in the regions where Timothy is used commercially. Germplasm collections generally preserve seeds of Phleum under refrigerated conditions and cryopreservation has not been carried out to any extent. There are more than 6,000 accessions of Phleum around the world in genebank collections. Of these over 98% are P. pratense and less than 2% are P. alpinum and other species. Major collections occur in Europe (>4,800), USA (>550), New Zealand (>300), Canada (>80) and Japan (>40). It is impossible to determine the number of unrelated accessions due to high rates of duplication within and between genebanks. It is of concern that there are so few P. pratense subsp. bertolonii and diploid P. alpinum accessions from Europe as well as so few southern European hexaploid P. pratense accessions.

14.6.7 Development of Core Collections An important process of germplasm conservation is to prioritize the collection into a core collection to allow maintenance of a relatively full range of diversity using a more manageable number of accessions. There are many factors to consider in the development of a core collection. These include morphological,

14 Phleum

geographic, edaphic and molecular diversity as well as an extreme variation of key individual traits. Casler (2001) showed very strong relationships between geographic origin of germplasm and a series of morphological traits and he was able to develop a core collection on this basis. Now that we have more understanding of the genomic constitutions and diversity, as well as the geographic variation due to glacial driven migrations, it should be possible to assemble a more representative core collection. However, to do this it will require the collection of Phleum resources in the glacial refugia of southern Europe and North Africa. These forms are seriously lacking in any collection.

14.6.8 Endophytic Fungi Many grasses host choke forming endophytic fungi of the genus Epichlo¨e or their non-choking asexual derivatives Neotyphodium. The sexual stroma of Epichlo¨e cause “choke disease,” which prevents seedheads emerging. Non-choke inducing asexual Neotypodium endophytes occur in many grasses such a perennial ryegrass (Lolium perenne) and tall fescue (Festuca arundinacea) and they provide the host with advantages such as improved insect resistance and grazing deterrence (Schardl and Phillips 1997). Timothy hosts a partial choke forming Epichlo¨e typhina endophyte and in this situation infected plants suffer from less infection with the leaf disease Cladosporium phlei (Seto et al. 2005). In Finland, 33% of plants were found to be infected with endophyte (Saikkonen et al. 2000). In Patagonia P. alpinum is known to host the nonchoking asexual Neotyphodium tembladerae (Moon et al. 2004; Gentile et al. 2005). This provides infected plants with an advantage over uninfected plants in some circumstances (Cabral et al. 2007). In this situation the infection with Neotyphodium represents a symbiotic relationship whereby the plant gains protection from insects and overgrazing while the endophyte obtains nutrition. The full relevance of endophyte fungi in Phleum is not known. However, it is important that endophytes are collected and studied. Germplasm collections will need to monitor and maintain endophytes where they are present.

271

14.7 Cytological, Genomic and Molecular Resources in Phleum There have been numerous cytological studies in Phleum, largely in an effort to understand the relationship between the genomes in P. pratense, for example as described by Joachimiak (2005). The B-chromosomes have also been studied (Bosemark 1967; Fro¨st 1969; Joachimiak 1982, 1986). There have been no attempts to develop chromosome substitution or deletion lines. Anther culture methodology has been developed and triploid and hexaploid regenerants have been obtained (Abdullah et al. 1994). Molecular techniques have been developed for phylogenetic and genomic identification (Stewart et al. 2008). Cai et al. (2003) developed a set of simple sequence repeat (SSR) markers and a team in Finland announced its intention in 2006 (Manninen et al. 2006) to use a candidate gene approach, bulked segregant analysis and gene expression chips to find modern genetic tools applicable to Timothy breeding. This has resulted in a publication on the development of markers for feeding quality and gray snow mould resistance (Tanhuanp€a€a et al. 2007, 2008). A study on genetic resources of Nordic Timothy is underway in Norway using molecular techniques (Fjellheim et al. 2007). Although no genetically modified Timothy cultivar has been developed, Timothy has been included in a number of patents for genes for disease resistance, herbicide tolerance and allergy studies.

14.8 Recommendations for Future Actions The current understanding of the genomic constitution within Phleum should allow breeders to utilize the genetic resources more effectively than previously. It should now be possible to resynthesize new forms and to introgress a much wider range of diversity into P. pratense than has occurred naturally. However, many of the genetic resources of wild relatives are under threat from climate warming and human induced habitat degradation (t’Mannetje 2007).

272

There is an urgent need to collect hexaploid P. pratense germplasm from Mediterranean mountain glacial refuge areas as well as a wide range of genetically diverse diploid, tetraploid and octoploid P. pratense and of the readily crossable forms of P. alpinum. These include diploid subsp. rhaeticum and “commutatum” as well as their diploid and tetraploid hybrids. These collections should be integrated into core collections to maximize molecular diversity of the available genomes. It is also important that each of the major regions where Timothy is used maintains strong functional field breeding programs to allow adequate cultivar development, germplasm collection, introgressions of wild germplasm and exploration of molecular resources. Acknowledgments The authors wish to acknowledge the support of PGG Wrightson Seeds and AgResearch, and the numerous people who supplied germplasm.

References Abdullah AA, Pedersen S, Andersen SB (1994) Triploid and hexaploid regenerants from hexaploid Timothy (Phleum pratense) via anther culture. Plant Breed 112:342–345 AgriQuality (2007) Seed certification statistics 2006–2007. AgriQuality, New Zealand Anonymous (2007) Appendix: Total world seed production (tonnes) of various grasses and clovers in the world and EU countries 1993–2006. In: Proceedings of the 6th international herbage seed conference, Norway, 2007 Barkworth ME, Capels KM, Long S, Anderton LK (2007) Flora of North America north of Mexico, vol 24: Magnoliophyta; Commelinidae (in part); Poaceae, part 1. Oxford University Press, Oxford, UK Bondesen OB (2007) Seed production and seed trade in a globalised world. Seed production in the northern light. In: Proceedings of the 6th international herbage seed conference, Norway, pp 9–12 Bosemark NO (1967) Edaphic factors and the geographic distribution of accessory chromosomes in Phleum phleoides. Hereditas 57:239–262 Cabral D, Iannone LJ, Stewart AV, Novas MV (2007) The distribution and incidence of Neotyphodium endophytes in native grasses from Argentina and its association with environmental factors. In: Proceedings of the 6th international endophyte symposium, Christchurch, New Zealand, 26–28 Mar 2007, pp 79–82 Cai Q, Bullen MR (1991) Characterization of genomes of timothy (Phleum pratense L.) 1. Karyotypes and C-banding patterns in cultivated timothy and two wild relatives. Genome 34:52–58

A.V. Stewart et al. Cai Q, Bullen MR (1994) Analysis of genome specific sequences in Phleum species: identification and use for study of genomic relationships. Theor Appl Genet 88:831–837 Cai HW, Yuyama N, Tamaki H, Yoshizawa A (2003) Isolation and characterization of simple sequence repeat markers in the hexaploid forage grass timothy (Phleum pratense L.). Theor Appl Genet 107:1337–1349 Caradus JR (1978) Plant introduction trials. Performance of Timothy cultivars and lines in New Zealand as spaced plants. NZ J Exp Agric 6:11–17 Casler MD (2001) Patterns of variation in a collection of Timothy accessions. Crop Sci 41:1616–1624 Cenci CA, Pegiati MT, Falistocco E (1984) Phleum pratense (Gramineae): chromosomal and biometric analysis of Italian populations. Willdenowia 14:343–353 Charlton JFL, Stewart AV (2000) Timothy, the plant and its use on New Zealand farms. Proc NZ Grassl Assoc 62:147–153 Conert HJ (1998) Phleum. In: Hegi G (ed) Illustrierte Flora von Mitteleuropa. Verlag Paul Parey, Berlin, pp 190–206 Cooper JP (1958) The effect of temperature and photoperiod on inflorescence development in strains of timothy (Phleum spp.). J Br Grassl Soc 13:81–91 Cooper JP, Calder DM (1964) The inductive requirements for flowering of some temperate grasses. Grass Forage Sci 19:6–14 Dogan M (1988) A scanning electron microscope survey of the lemma in Phleum, Pseudophleum and Rhizocephalus (Gramineae). Notes R Bot Gard Edinburgh 45:117–124 Dogan M (1991) A taxonomical revision of the genus Phleum L. (Gramineae). Karaca Arbor Mag 1:53–70 Dogan M, Us J (1996) Infrageneric classification of the genus Phleum L. (Gramineae) estimated by numerical taxonomy. ¨ zt€urk M, Sec¸men O ¨ , Go¨rk G (eds) Plant life in SouthIn: O west and Central Asia. Ege University Press, Izmir, Turkey, pp 160–165 Ellestrom S, Tijo JH (1950) Note on the chromosomes of Phleum echinatum. Bot Not 4:463–465 Fjellheim S, Pedersen AJ, Andersen JR, Antonius-Klemola K, Bondo L, Brantestam AK, Dafga˚rd L, Helgadottir A, Isolahti M, Jensen LFB, L€ubberstedt T, Mannien O, Marum P, Merker A, Tanuanp€a€a P, Weibull J, Weibull P, Rognli OA (2007) Phenotypic and molecular characterization of genetic resources of Nordic timothy (Phleum pratense L.). In: Abstracts book of 27th EUCARPIA symposium on improvement of fodder crops and amenity grasses, Copenhagen, Denmark, 19–23 Aug 2007, p 61 Foerster E (1968) Ein beitrag zur untercheidung von Phleum pratense und Phleum nodosum. Go¨ttinger Flor Rundbreife 1:9 Foerster E (2005) Nat€urliche Hybriden zwischen Phleum pratense und Phleum bertolonii (Natural hybrids between Phleum pratense and Phleum bertolonii). Gr€undungstagung der Vereinigung zur Erforschung der Flora Deutschlands, University of Vechta, Germany, 29–30 Oct 2005 Fro¨st S (1969) The inheritance of accessory chromosomes in plants, especially in Ranunculus acris and Phleum nodosum. Hereditas 61:317–326 Gentile A, Rossi MS, Cabral D, Craven KD, Schardl CL (2005) Origin, divergence, and phylogeny of Epichloe¨ endophytes of native Argentine grasses. Mol Phylogenet Evol 35: 196–208

14 Phleum Guo Y-D, Yli-Matilla T, Pulli S (2003) Assessment of genetic variation in timothy (Phleum pratense L.) using RAPD and UP-PCR. Hereditas 138:101–113 Heide OM, Solhaug KA (2001) Growth and reproduction capacities of two bipolar Phleum alpinum populations from Norway and South Georgia. Arct Antarct Alp Res 33:173–180 Hewitt GM (1996) Some genetic consequences of ice ages, and their role in divergence and speciation. Biol J Linn Soc 58:247–276 Hewitt GM (1999) Post glacial recolonisation of European biota. Biol J Linn Soc 68:87–112 Hong Q, White P, Klinka K, Chourmouzis C (1999) Phytogeographical and community similarities of alpine tundras of Changbaishan Summit, and Indian Peaks, USA. J Veg Sci 10:869–882 Humphries CJ (1978) Notes on the genus Phleum L. Bot J Linn Soc 76:337–340 Humphries CJ (1980) Phleum. In: Tutin TG, Heywood VH, Burgess NA, Moore DM, Valentine DH, Walters SM, Webb DA (eds) Flora Europeaea. Alismataceae to Orchidaceae (Monocotyledones), vol 5. Cambridge University Press, Cambridge, UK, pp 239–241 Joachimiak A (1982) Cyto-genetics of standard B-chromosomes in Phleum boehmeri from Poland. Acta Biol Cracov Ser Bot 24:63–77 Joachimiak A (1986) B-chromosome condensation in Phleum pollen grains. Genetica 68:169–174 Joachimiak A (2005) Heterochromatin and microevolution in Phleum. In: Sharma AK, Sharma A (eds) Plant genome: biodiversity and evolution, vol 1, Part B: Phanerogams. Science, Enfield, NH, USA, pp 89–117 Joachimiak A, Kula A (1993) Cytotaxonomy and karyotype evolution in Phleum sect. Phleum (Poaceae) in Poland. Plant Syst Evol 188:11–25 Joachimiak A, Kula A (1996) Karyosystematics of the Phleum alpinum polyploid complex (Poaceae). Plant Syst Evol 203:11–25 Kellogg EA, Bennetzen JL (2004) The evolution of nuclear genome structure in seed plants. Am J Bot 91:1709–1725 Kovats D (1976) Phleum studies I. Data of the taxonomy and morphology of Phleum bertolonii DC. and Phleum pratense L. Acta Bot Acad Sci Hungar 22:107–126 Kovats D (1977) Phleum studies II. Phleum hubbardii a new species of Poaceae (Gramineae). Acta Bot Acad Sci Hungar 23:119–142 Kovats D (1980) Distribution and diversity of Phleum hubbardii and Phleum pratense (Poaceae) in the Carpathian Basin. Stud Bot Hungar 14:107–116 Kozuharov SI, Petrova AV (1991) Chromosome numbers of Bulgarian angiosperms. Fitologija 39:72–77 Kula A (2003) Morphology and cytogenetics of Phleum hubbardii. In: Frey L, Szafer W (eds) Problems of grass biology. Institute of Botany, Polish Academy of Sciences, Krakow, Poland, pp 299–312 Kula A (2005a) Kariologia i morfologia gatunko´w z rodzaju Phleum. Zesz Nauk AR Krak 418:1–172 Kula A (2005b) Searching for a primeval Phleum karyotype. In: Frey L, Szafer W (eds) Problems of grass biology. Institute of Botany, Polish Academy of Science, Krakow, Poland, pp 197–206

273 Kula A, Dudziak B, Sliwinska E, Grabowska-Joachimiak A, Stewart AV, Golczyk H, Joachimiak A (2006) Cytomorphological studies on American and European Phleum commutatum Gaud. (Poaceae). Acta Biol Cracov Ser Bot 48:99–108 Levan A (1941) Syncyte formation in the pollen mother-cells of haploid Phleum pratense. Hereditas 27:243–252 Løhde JJH (1977) Phleum pratense and Phleum bertolonii hybridisation, morphology and ecology in Denmark. Dissertation, Kongelige Veterinær- og Landbohøjkole, Copenhagen, Denmark, 80p Maire R (1953) Flore de l’Afrique du Nord, vol 2. Lechevalier, Paris, France Manninen O, Erkkil€a M, Isolahti M, Nissinen O, P€arssinen P, Rinne M, Tanhuanp€a€a P (2006) Biotechnological tools for breeding feeding quality and optimal growth rhythm in timothy, Phleum pratense. Timothy productivity and forage quality – possibilities and limitations. NJF Seminar 384, Akureyri, Iceland, 10–12 Aug 2006, pp 119–120 Moon CD, Craven KD, Leuchtmann A, Clement SL, Schardl CL (2004) Prevalence of interspecific hybrids amongst asexual fungal endophytes of grasses. Mol Ecol 13:1455–1467 M€untzing A (1935) Cyto-genetic studies on hybrids between two Phleum-species. Hereditas 20:103–136 M€untzing A, Prakken R (1940) The mode of chromosome pairing in Phleum twins with 63 chromosomes and its cytogenetic consequences. Hereditas 26:463–501 Myers WM (1941) Meiotic behaviour of Phleum pratense, Phleum subulatum and their F1 hybrid. J Agric Res 63:649–655 Nakazumi H, Furuya M, Shimokouji H, Fujii H (1997) Wide hybridization between timothy (Phleum pratense L.) and orchardgrass (Dactylis glomerata L.) Bull. Hokkaido Prefectural Agric Exp Stn (Japan) 72:11–16 Nath J (1967) Cytogenetical and related studies in the genus Phleum L. Euphytica 16:267–282 Nielsen EL, Nath J (1961) Cytogenetics of a tetraploid form of Phleum pratense L. Euphytica 10:343–350 Nielsen EL, Smith DC (1959) Chlorophyll inheritance patterns and extent of natural self-pollination in Timothy. Euphytica 8:169–179 No¨rdenskiold H (1945) Cyto-genetic studies in the genus Phleum. Acta Agric Suec 1:1–138 No¨rdenskiold H (1949) Synthesis of Phleum pratense L. from P. nodosum L. Hereditas 35:190–202 No¨rdenskiold H (1953) A genetical study in the mode of segregation in hexaploid Phleum pratense. Hereditas 39:469–488 No¨rdenskiold H (1957) Segregation ratios in progenies of hybrids between natural and synthesized Phleum pratense. Hereditas 43:525–540 No¨rdenskiold H (1960) The mode of segregation in a family of hexaploid Phleum pratense. Hereditas 46:504–510 Perny´ M, Kolarcik V, Majesky´ L, Ma´rtonfi P (2008) Cytogeography of the Phleum pratense group (Poaceae) in the Carpathians and Pannonia. Bot J Linn Soc 157:475–485 Saikkonen K, Ahlholm J, Helander M, Lehtim€aki S, Niemel€ainen O (2000) Endophytic fungi in wild and cultivated grasses in Finland. Ecography 23:360–366 Schardl C, Phillips T (1997) Protective grass endophytes: where are they from and where are they going? Plant Dis 81: 430–438

274 Seto Y, Kogami Y, Shimanuki T, Takahashi K, Matsuura H, Yoshihara T (2005) Production of Phleichrome by Cladosporium phlei as Stimulated by Diketopiperadines of Epichloe typhina. Biosci Biotechnol Biochem 69:1515–1519 Sliwinska E, Kula A, Joachimiak A, Stewart AV (2003) Genome size in seven Phleum species. In: International workshop on application of novel cytogenetic and molecular techniques in genetics and breeding of grasses, Poznan, Poland, 1–2 Apr 2003, p 19 Soltis PS, Soltis DE (2000) The role of genetic and genomic attributes in the success of polyploids. Proc Natl Acad Sci USA 97:7051–7057 Stewart AV, Joachimiak A, Ellison N (2008) Genomic origins of subgenus Phleum based on ITS and chloroplast sequences. In: Proceedings of the 5th international sympsoium on molecular breeding of forage and turf, Sapporo, Japan, 1–6 July 2008, pp 71–81 t’Mannetje L (2007) Climate change and grasslands through the ages: an overview. Grass Forag Sci 62:113–117 Tanhuanp€a€a P, Erkkil€a M, Nissinen O, Rinne M, Manninen O, Isolahti M, P€arssinen P (2007) Developing DNA markers for feeding quality and gray snow mold resistance in timothy, Phleum pratense. In: Plant GEM 6: Plant Genomics European Meeting, Tenerife, Spain, 3–6 Oct 2007, p 94 Tanhuanp€a€a P, Isolahti M, Nissinen O, P€arssinen P, Kalendar R, Schulman A, Manninen O (2008) Identification of DNA markers for gray snow mold resistance in timothy, Phleum

A.V. Stewart et al. pratense, using bulked segregant analysis. In: Molecular mapping and marker assisted selection in plants, Vienna, Austria, 3–6 Feb 2008, p 68 € Thorn K (1960) Bemerkungen zu einer Ubersichtskarte vermutlicher Glazialre-liktpflanzen Deutschlands. Mitt Florist 8:81–85 Uhrı´kova´ A, Kra´lik E (2000) Karyologicke sˇtu´dium slovenskej flo´ry XX1X. Acta Fac Rerum Nat Univ Comen Bot 40:17–22 Weber WA (2003) The Middle Asian element in the Southern Rocky Mountain flora of the western United States: a critical biogeographical review. J Biogeogr 30:649–688 Weibull J, Ottosson F, Kolodinska Brantestam A, Dafgard L, Weibull P, Merker A (2007) Vanishing variation – the diversity of Timothy (Phleum pratense L.) in historical grasslands. In: 18th EUCARPIA genetic resource section meeting – plant genetic resources and their exploitation in the plant breeding for food and agriculture, Piestany, Slovakia, 23–26 May 2007 Wilton AC, Klebesadel LJ (1973) Karyology and phylogenetic relationships of Phleum pratense, P. commutatum, and P. bertolonii. Crop Sci 13:663–665 Yang-Dong G, Tapani Y, Seppo P (2003) Assessment of genetic variation in Timothy (Phleum pratense L.) using RAPD and UP-PCR. Hereditas 138:101–113 Zernig K (2005) Phleum commutatum and Phleum rhaeticum (Poaceae) in the Eastern Alps: characteristics and distribution. Phyton 45:65–79

Chapter 15

Setaria Henri Darmency and Jack Dekker

15.1 Introduction Foxtail millet, Setaria italica (L.) Beauv., is a small grain C4 cereal of the Paniceae tribe of Poaceae family (Fig. 15.1). It is one of the few economically important species of Setaria. It has long been important in China and India for human consumption. It is also grown in small quantities throughout Eurasia for some traditional uses and for feeding birds, and in Europe and America for hay and silage. Grains have been discovered in Neolithic relics (circa 8,000 years ago) in various places of China, along the upper and middle valleys of the Yellow river and the Yangtze River (Li and Wu 1996; Nasu et al. 2007) and in Europe (Naciri and Belliard 1987). A small number of Setaria species are closely related and constitute a narrow gene pool potentially useful for foxtail millet genetic improvement. Besides S. italica, other Setaria are grown as crop or forage (De Wet et al. 1979; Austin 2006). It includes Setaria pumila partially domesticated and cultivated in mixed stand with other millets in India (Kimata et al. 2000), S. sphacelata is used as wild cereal in Africa and planted as a tropical pasture species (Jank et al. 2007), and S. parviflora is used as a cereal and is found in the pre-historic remains in Central America (Austin 2006). The present chapter

H. Darmency (*) UMR 1210 Biologie et Gestion des Adventices, Institut National de la Recherche Agronomique, 17 rue Sully, BP 86510, 21065 Dijon, France e-mail: [email protected]

focuses on foxtail millet only. It is generally classified into three races: maxima, indica and moharia (Prasada Rao et al. 1987). Races maxima and indica are typically used as a cereal, while race moharia is used as forage. As far as we know, the use of wild Setaria in breeding programs to improve foxtail millet is very scarce. In contrast, wild Setaria were invaluable resources to gain better knowledge of the genome organization of the crop. Exploiting wild and weedy Setaria species has not gained the interest of crop breeders for several reasons. Perhaps, the fact that most of the research and achievements have been carried out in China, India, and Japan led to the release of the corresponding reports written in local languages and thus unidentified in recent worldwide databases (Gu 1987; Li 1995). Another reason of the rare use of the wild germplasm could be the structure of the wild–weed–crop complex providing enough genetic diversity through “off-types” to satisfy new breeding plans. In addition, the huge diversity of landraces collected in China (23,400; Li and Wu 1996), India (1,951; Seetharam 1998) or Japan (NIAS 2009) made available a quantity of interesting material to modern breeders, which resulted in poor interest in wild Setaria. Collection and characterization of landraces were carried out through China (Li and Wu 1996), India (Upadhyaya et al. 2008), Russia (Varadinov 1986), France (Nguyen Van and Perne`s 1985), and Japan (Nakayama et al. 2005). Finally, the crop foxtail millet is not traded on world grain exchanges, while it remains an important grain crop for indigenous, often poorer, populations in Eurasia, Africa and the Americas, and is therefore not of interest to mainstream crop improvement projects.

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_15, # Springer-Verlag Berlin Heidelberg 2011

275

276

H. Darmency and J. Dekker

Fig. 15.1 Mature infrutescences (ears) of foxtail millet and green foxtail

15.2 The Complex of Species of Foxtail Millet 15.2.1 A Brief Glance Worldwide there are 125 Setaria species divided among several subgenera (Hubbard 1915; Rominger 1962). The genus, Setaria P. Beauv. (the foxtails), is widely distributed in tropical, subtropical, and temperate

parts of the world and is of significant agricultural importance as crops and also as noxious weeds (Rominger 1962; Prasada Rao et al. 1987). The taxonomy of the genus is very complex, and an accurate classification has been confounded by the high degree of overlapping morphological characters both within and between species, and the diverse polyploidy levels within the genus. One of the most coherent taxonomic treatments was given by Rominger (1962). There is evidence that the genus Setaria is actually divided into

15 Setaria

277

two groups, or clades (Doust and Kellogg 2002). This reassessment of the genus would place S. pumila and S. parviflora into a common (S. pumila) clade, and S. italica, S. viridis, S. faberi and S. verticillata in another (S. viridis). Observations drawn from karyological studies and interspecific hybridization suggest that the complex of species of foxtail millet is organized in three gene pools (Fig. 15.2). Closest to S. italica is its putative wild ancestor, S. viridis (L.) P. Beauv., the green foxtail, which forms the primary gene pool at a diploid level with the A genome (Li et al. 1945; Harlan and De Wet 1971; Darmency et al. 1987b; Benabdelmouna et al. 2001a). The secondary gene pool corresponds to the species carrying the B genome. It includes the diploid S. adhaerans (Forsskal) Chiov. (bristly grass), and the tetraploid species S. verticillata (L.) P. Beauv. (bristly foxtail) and S. faberi F. Hermann (giant foxtail). A tertiary gene pool could tentatively contain S. pumila (Poiret) Roemer & Schultes (yellow foxtail) in addition to many other wild species with unidentified genomic composition, but no link has been established yet. All these species are annual, typically weedy, and originated from the temperate regions of the northern hemisphere. Reviews of the weedy Setaria species-group (Dekker 2003, 2004), as well as three of individual species have been published previously in the Biology of Canadian Weeds series (S. pumila and S. verticillata, Steel et al. 1983; S. viridis, Douglas et al. 1985; S. faberi, Nurse et al. 2009).

15.2.2 Taxonomy The genus Setaria belongs to the tribe Paniceae, subfamily Panicoideae and family Poaceae. Linnaeus (1753) was the first to recognize foxtail millet and S. viridis as two independent species, Panicum italicum (foxtail millet) and P. viride (S. viridis). However, numerous observations and experiments on their S. adhaerens S. italica S. viridis

Primary gene pool Genome A

?

S. verticillata S. faberi Secondary gene pool Genomes A and B

Tertiary gene pool

Fig. 15.2 Putative gene pools of foxtail millet

continuous and overlapping genetic variation, the evidence of interfertility, and the weedy origins of the crop (see Sect. 15.6.1) greatly support the hypothesis of a common “biological species” status, namely S. viridis, with two subspecies: spp. italica (foxtail millet), and spp. viridis (green foxtail). Beside a long history of nomenclatural changes, including that both taxa be considered as subspecies of S. italica, which implied the primacy of the crop over that of its wild Setaria ancestors, the correct binomial to be used now maintains the two taxa at the species level, S. viridis, and S. italica. Present nomenclatural authorities are presented in Table 15.1 for the complex of species of foxtail millet. Besides, there are many nomenclaturally designated forms or racial taxonomies below the subspecies level that have been proposed based on morphology rather than genetic distance (Rominger 1962; Douglas et al. 1985). For instance, S. viridis ssp. pycnocoma should not be considered as a fixed narrow taxon since numerous weedy forms may originate independently through natural crossing, as indicated in Sect. 15.3.3.2.

Table 15.1 Karyological data of Eurasian taxa of the complex of species of the foxtail millet Species Most frequent Genome ploidy level S. italica (L.) P. Beauv. 2n ¼ 2x ¼ 18 AA S. viridis (L.) P. Beauv. 2n ¼ 2x ¼ 18 AA 2n ¼ 2x ¼ 18 AA S. viridis spp. pycnocoma (Steudel) Tzveleva S. adhaerans (Forsskal) Chiov. 2n ¼ 2x ¼ 18 BB S. adhaerans var. antrorsa 2n ¼ 2x ¼ 18 BB (A. Braun) H. Scholz Spontaneous F1 hybrid S. viridis 2n ¼ 2x ¼ 18 AB  S. adhaerans S. verticillata (L.) P. Beauv. 2n ¼ 4x ¼ 36e,f AABB 2n ¼ 4x ¼ 36 AABB S. verticillata var ambigua (Guss.) Parl.d Spontaneous F1 hybrid 3x ¼ 27 AAB S. verticillata  S. italica S. faberi F. Hermann 2n ¼ 4x ¼ 36 AABB 2n ¼ 4x ¼ 36g,h,i ? S. pumila (Poiret) Roemer & Schultesc S. parviflora (Poir.) Kergue´lend 2n ¼ 36 j ? Frequent synonymous: aS. viridis var. major, bS. verticilliformis, S. lutescens and S. glauca, dS. geniculata. Variant 2n numbers: e 2n ¼ 18 (Pohl 1962; Osada 1989), f2n ¼ 54 (Khosla and Sharma 1973; Singh and Gupta 1977; Osada 1989), g2n ¼ 18 (de Wet 1954; Wang et al. 2007), h2n ¼ 72 (Brown 1948; Rominger 1962; Khosla and Sharma 1973; Wang et al. 2007), i aneuploidy, hypertetraploidy n ¼ 22, n ¼ 19 þ 03B (Khosla and Sharma 1973); j2n ¼ 72 (Brown 1948; Rominger 1962; Osada 1989; Wang et al. 2007) c

278

Darmency (2004) discussed all the conditions that have been met to establish a dynamic and evolving system known as a “wild–weed–crop complex” and comprising S. viridis, S. viridis ssp. pycnocoma (as the descendants of interspecific hybrids) and S. italica. It is likely that varieties such as robusta-alba and robusta-purpurea described in America belong to that dynamics. Indeed, early morphological and chemical methods indicated that “robust” foxtails are more closely related to the giant green foxtail than to the green foxtail (Williams and Schreiber 1976). In addition, all these subspecies and varieties are completely embedded within the S. viridis range of variation in a multiple isozyme loci analysis (Wang et al. 1995a). Another case of taxonomic trouble arises with S. adhaerens and S. verticillata, both species involving the majority of vouchers with retrorsely barbed bristles, but a few display antrorse barbed bristles, which lead to various taxa levels (for instance var. ambigua ¼ S. verticilliformis) although it is a simple genetic attribute. The different ploidy levels also, because they are not apparent in the phenotype, add trouble in the correspondence between the botanical descriptions and the actual genetic nature.

15.2.3 Phylogeny Harlan and de Wet (1971) suggested that S. italica and its assumed wild ancestor S. viridis are taxonomically the same species. Indeed, several studies showed that a limited number of differential traits have been involved in domestication (Li et al. 1945; Darmency and Perne`s 1987). Such a strong link is also supported by results of comparative studies between S. italica and S. viridis, which showed that the storage proteins and isozyme differentiation between the two types was regional rather than taxonomic: S. italica and S. viridis from the same geographic region were more closely related than S. italica accessions from different regions (De Wet et al. 1979; Jusuf and Perne`s 1985; Gao and Chen 1988). More recently, molecular data also support the close relationship of the cultivated species S. italica and the wild species S. viridis. Using random amplified polymorphic DNA (RAPD) analysis among and within species, Li et al. (1998) concluded that S. italica is more closely related to S. viridis than to other species of Setaria and appears

H. Darmency and J. Dekker

to be conspecific. The analysis of ribosomal genes’ sequences confirmed the closest relationship of S. italica and S. viridis accessions from the same geographical region (Benabdelmouna et al. 2001b). Wang et al. (1998) compared both intra and interspecific restriction fragment length polymorphism (RFLP) based maps of S. italica and found that these maps were very similar in length, show the same order of the RFLP markers and also the same genetic distances between the loci. This strongly indicates that there have been no major chromosomal rearrangements between the genomes of S. italica and S. viridis. Finally, a genomic-based chromosome analysis, genomic in situ hybridization (GISH), showed that the genomes of S. viridis and S. italica were closely related and could not be separated, while there was a clear absence of homology between S. italica and S. adhaerans (Benabdelmouna et al. 2001a). Whether S. italica and S. viridis should be considered as conspecific or as subspecies is a matter of taxonomy rules. The only justification for keeping the two taxa as separate species is the domestication syndrome that makes morphology, growth and agronomic use of foxtail millet very different in both taxa. The GISH techniques also provided a powerful tool to revise the phylogeny of the secondary gene pool of the foxtail millet complex. S. adhaerens was confirmed to belong to that secondary gene pool, S. faberi was confirmed as allotetraploid (after the hypothesis of Li et al. 1942), but S. verticillata appeared to share the same allotetraploid genome as S. faberi (Benabdelmouna et al. 2001a). Thus, the fundamental difference between the two allotetraploids, S. faberi and S. verticillata, seems not to be linked to their genomic composition. Perhaps, the difference between the two allotetraploid species could be a matter of geographical origin or genetic difference among parents for a few genes encoding morphological traits. The comparison of the sequences of the ribosomal genes adds evidence of the phylogenetic relationships between the diploid and the allotetraploid Setaria (Benabdelmouna et al. 2001b). The origins of S. parviflora (Poir.) Kergue´len (syn. S. geniculata P. Beauv.) are of particular note. S. parviflora, a perennial, is the only weedy foxtail native to the New World, and very closely resembles S. pumila, an annual species of Eurasian origin (Rominger 1962; Wang et al. 1995b). S. parviflora and S. pumila are frequently misidentified by weed

15 Setaria

managers. The primary morphological differences between S. parviflora and S. pumila are found in the spikelet length and lower floret, palea, and the presence of short rhizomes (Hitchcock 1971). S. parviflora has a high degree of sterilization in the fascicle, unlike any other species native to North America, but similar to tropical Setaria species. It has been suggested that the cause of this enigma was a very ancient dispersal event westward from the Old to New World (Rominger 1962).

15.3 Genetics Attributes 15.3.1 Cytogenetics and Karyotypes The genome size of Setaria is one of the smaller one of cereals, with rice (Devos et al. 1998). S. italica and S. viridis have the same DNA content, 2C ¼ 0.51 pg (Le Thierry d’Ennequin et al. 1998). The most frequently reported number base of chromosomes ever reported is x ¼ 9 (Singh and Gupta 1977). Setaria species of the foxtail millet complex includes the polyploid series of 2x (S. italica, S. viridis, S. adhaerens), 4x (S. faberi, S. pumila, S. parviflora, S. verticillata), 6x (S. verticillata), and 8x (S. pumila, S. parviflora) possibilities (Table 15.1). The tetraploid level contains the majority of counts recorded for S. verticillata and S. pumila. Diploid counts of S. verticillata could be attributed to the diploid S. adhaerens with similar morphology (Wang et al. 2009b). However, the availability of chromosome races in the same taxon raises important questions about the taxonomy and speciation processes, in particular about the reproductive isolation of these different chromosome races. For instances, samples recognized as probably S. verticillata were in fact spontaneous hybrids between S. italica and S. verticillata, confirmed at 3x ¼ 27 chromosomes, so that they can serve as bridge between the ploidy levels (PoirierHamon and Perne`s 1986). Aneuploidy in the genus has been reported, but is considered insignificant. GISH, a technique of painting chromosomes according to genomic affinity, showed clear differentiation between the A genome set of chromosome of S. viridis and the B set of S. adhaerens. Metaphase of hand crossed hybrids between these two species had one set of 9 chromosomes of each parent, which confirmed that

279

clear discrimination between the A and the B genomes was efficient (Benabdelmouna et al. 2001a). A similar pattern was observed between S. italica and S. adhaerens, but no difference was detected between S. italica and S. viridis, thus confirming their genetic proximity. This technique was also applied on polyploid Setaria species and enabled to confirm the assumed AABB allotetraploid nature of S. faberi (Li et al. 1942) and to demonstrate that both S. verticillata and S. verticillata var. ambigua were also allotetraploids of the same genomes AABB (Benabdelmouna et al. 2001a). Hexaploid accessions of S. verticillata have not been included in this study so that the question remains open about a third genome or partial autoploidy, which could explain the abnormal meiosis observed in these plants (Gupta and Yashvir 1973). Interestingly, S. queenslandica, a species originating from Australia, where both S. viridis and S. italica were imported, showed to be the first recorded autotetraploid composed of genome AAAA (Wang et al. 2009b). The polyploid S. pumila was shown to bear an unknown genomic composition, which is not closely related either to genome A or to genome B (Benabdelmouna et al. 2001a), but no study was carried out to investigate a possible homology to some part of genomes A and B, which should be helpful to determine if S. pumila could be used in breeding foxtail millet. A third, different, genome, designated as genome C, was recently discovered in S. grisebachii that originated from Mexico, but there is no information about its potential crossing to foxtail millet as yet (Wang et al. 2009b). The characterization of Setaria species originating from Africa and America could bring new data about the genetic proximity within the genus. For instance, the close similarity between the annual S. pumila of the Old world and the S. parviflora of the New World leads us to expect some possible relationships and crossing potential in a foreign tertiary gene pool (Wang et al. 1995b).

15.3.2 Mating System All the Setaria discussed in this chapter are purported to be highly self-fertilized. The weedy foxtails are primarily a self-pollinated species (Pohl 1951; Mulligan and Findlay 1970). Wind pollination (anemophily) is the mode in those rare circumstances of outcrossing (Pohl 1951), with pollen probably moving

280

at most a few dozen meters (Wang et al. 1997). The panicle has many very short branches (fascicles), each bearing a number of branchlets and consisting of several spikelets and a cluster of setae (bristles). S. pumila panicle morphology differs from that of the secondary gene pool: only a single spikelet is found in each S. pumila fascicle. A spikelet comprises two flowers; one is hermaphrodite, the other is sterile for S. viridis, S. faberi and S. verticillata. In S. pumila, the lower floret has no pistil but three well developed anthers, which open 3–7 days later than the first floret (Willweber-Kishimoto 1962). Although the anthers are visible at flowering, and then feathery stigmates also exert from the glumes, the fertilization principally occurs within the flower. Stigmates maturity is generally synchronous with anther dehiscence, which results in a high probability of self-pollination. Some rare cases of protogyny, or in contrast, of protandry, can be observed according to the genotype and the environmental conditions, but in both cases the more proximate pollen belongs to the other spikelets of the same panicle, thus leading again to self-pollination. Enclosing one single panicle of S. viridis and S. viridis spp. pycnocoma in a bag causes no more than 5% of empty spikelets, which could be due to the bag and the greenhouse effects (Darmency and Perne`s 1985). The former experiments to report on quantitative estimates were carried out for foxtail millet using red pigmentation of the stem as a marker. When sowed in rows 0.60 m apart, outcrossing was 0.59%, but it could be up to 2.26% in the case of mixed planting (Takahashi and Hoshino 1934). Thick bristled strains showed lower percentage of natural crossing than the rough bristled, which could predict lower outcrossing in wild Setaria because they are generally more bristly. Till-Bottraud et al. (1992) found 0.74% outcrossing for S. viridis plants spaced every 0.25 m. Similarly, a selfing rate exceeding 99% was reported in Jasieniuk et al. (1994) for S. viridis. Using a dominant herbicideresistance marker, Volenberg and Stoltenberg (2002) found an outcrossing rate ranging from 0 to 2.4% for S. faberi planted at 0.36 m interval. There was very little report of intraspecific crosses to study the genetic nature of the traits of the wild Setaria. In two cases, they concern herbicide resistance. Jasieniuk et al. (1994) crossed trifluralin-resistant and susceptible S. viridis plants, which showed that the resistance was under the control of a recessive nuclear gene. Volenberg and Stoltenberg (2002) crossed fluazifop

H. Darmency and J. Dekker

resistant and susceptible S. faberi and found a single codominant allele endowing the resistance. Asexual reproduction is not a common mode of reproduction in weedy foxtails. The only weedy foxtail relying on vegetative asexual reproduction is S. parviflora, a perennial species with short, branched, knotty rhizomes (Rominger 1962). Apomixy in S. viridis has been noted (Mulligan and Findlay 1970), and has also been suggested for S. pumila and S. verticillata (Steel et al. 1983). Apomixis has also been reported in other Setaria species within the Setaria subgenus (Emery 1957; Chapman 1992). Emery (1957) reports both sexual (autogamy) and apomictic seed production in S. leucopila, S. machrostachya and S. texana, a highly unusual condition. S. faberi and S. pumila tillers readily root in soil when cut, separated from the plant, and buried in moist soil, an important trait allowing weedy foxtails to reestablish themselves after cultivation and mowing (Santleman et al. 1963; Schreiber 1965).

15.3.3 Interspecific Hybridization 15.3.3.1 Hand Crossing Beside spontaneous crosses reported or analyzed in some instances, most of the direct knowledge on the hybridizations relies on hand crossing experiments in controlled conditions, as described in Sect. 15.7.1. Hybridization between foxtail millet and all the other species of the group were successful, but to different extent: it was rather easy with S. viridis (Kihara and Kishimoto 1942; Li et al. 1945; Darmency and Perne`s 1985), but it was apparently less evident with S. adhaerens, S. faberi, and S. verticillata. In addition, further viable progeny was not always obtained, which, however, depends on the efforts made for the number of crosses carried out and the genotypes used. Benabdelmouna et al. (2001a) obtained an average of one hybrid per emasculated spike of S. viridis when the pollen donor belongs to S. adhaerens. These interspecific hybrids were weak, grew slowly, and produced very few flowers, most of them were sterile. The reciprocal cross gave no hybrid, but from a lesser number of spike. Li et al. (1942), after many attempts, obtained two seeds of the S. faberi  S. italica cross. These hybrids looked like S. faberi but they were

15 Setaria

almost sterile. Poirier-Hamon and Perne`s (1986) obtained a few sterile hybrids between S. verticillata and S. italica, but a colchicine treatment got a very variable progeny. Although attempted, no natural or artificial crosses between S. viridis and S. pumila succeeded. Willweber-Kishimoto (1962) successfully pollinated both S. italica and S. viridis with S. faberi and hybrid seed was formed. Willweber-Kishimoto (1962) successfully pollinated S. pumila (n ¼ 18) with S. faberi (n ¼ 18), no hybrid seed were formed.

15.3.3.2 Spontaneous Crosses In experimental plots, as much as 3% outcrossing was recorded in experiments where the two species were grown together (De Wet et al. 1979). Lower frequencies were generally found in most experiments, values differing according to the cultivar, the ratio of target versus pollen donor, the respective location of target plants in the experimental design, and which species was the target (Darmency et al. 1987b, 1992). In conditions that mimic a foxtail millet field, seeds produced by S. viridis growing between the rows contained 0.2% hybrids, while only 0.002% of the grains produced by the crop were hybrids (Till-Bottraud et al. 1992). This unequal balance is due to the taller habit of the crop that produces a lot of pollen grains directly above the wild plants, while the pollen of the shorter S. viridis remains close to the soil even with air turbulence. The rate of hybrids decreases with the distance to the plot of foxtail millet used as pollen donor (Wang et al. 2001). Under commercial field conditions, the rate of hybrid produced by S. viridis is still lower: 0.039, 0.007 and 0.002% inside the field, 5 and 20 m away, respectively (Shi et al. 2008). Since both S. italica and its wild relatives are very variable in morphology, it is quite normal that botanists rarely described hybrid forms. For instance, shape and growth habit of hybrids between S. viridis and S. italica correspond to the gross description of S. viridis spp. pycnocoma that is often still considered a genuine taxon (Darmency et al. 1987b; Darmency 2004). In the wild, such hybrids could suffer fitness penalty due to a mix of antagonistic wild and domesticated characters, lack of seed dormancy, needs of more resources, etc. Although no direct estimate of the relative fitness of hybrids has ever been carried out,

281

it is likely that the contribution of one hybrid to the seed production of a field population would be about 30 times less than for a green foxtail. Hybridization studies showed that hybrids between green foxtail and foxtail millet had three times fewer tillers and spikes (Darmency et al. 1987a), and four times less spikelet fertility for the main spike (Darmency and Perne`s 1985; Wang et al. 2001), and 12 times less fertility for secondary tillers (Wang et al. 2001) than their green foxtail parents. Similar results were found by Li et al. (1945) and reported by Gu (1987). Hybrids using the giant green foxtail, S. viridis ssp. pycnocoma, had less spikelet sterility and were rather close to the cultivated type, thus indicating a closer proximity to the domesticated species (Darmency and Perne`s 1985). This closer proximity was confirmed by lower variance of the traits measured on the F2 compared to that of a cross with a typical S. viridis (Darmency et al. 1987b). Some spontaneous hybrids, or descendants of hybrids, were directly identified in the field in few instances. Sterile hybrids between S. italica and S. verticillata were observed repeatedly in millet field in France (Poirier-Hamon and Perne`s 1986). Fertile hybrids between S. italica and S. viridis also occurred in both directions in the same region, as determined by chloroplast DNA analysis (Till-Bottraud et al. 1992). Stace (1975) found a naturally occurring S. viridis  S. verticillata hybrid called S.  ambigua Guss (2n ¼ 36; Clayton 1980) (syn. S.  ambigua Guss emend. Koyama; S.  decipiens Schimp. (Osada 1989)) throughout southern Europe. Finally, the occurrence of allotetraploids of AB genomes indicates that spontaneous interspecific crosses were successful in a remote past and have resulted in fertile progeny. In particular, the relatively homogeneous population allozyme data suggest this polyploidization event in S. faberi was a relatively recent evolutionary event (Wang et al. 1995b). A comprehensive picture of the relationships among the Setaria species of the foxtail millet complex has been presented in Darmency (2004).

15.4 Phenotypic Variations 15.4.1 Morphological Variants S. viridis varies greatly in morphology and growth characteristics. A number of morphological variants

282

of S. viridis have been reported, including named varieties: S. viridis var. viridis; S. viridis var. weinmanni (R. & S.) Brand.; S. viridis var. ambigua; S.viridis var. major (Gaud.) Posp., giant green; S. viridis var. robusta-alba Schreiber, robust white; S. viridis var. robusta-purpurea Schreiber, robust purple; S. viridis var. pycnocoma (Steud.) Henrard & Nakai; and S. viridis var. pachystachys (Franch & Savat.) Makino & Nemoto (Hubbard 1915; Schreiber and Oliver 1971; Kawano and Miyake 1983). The taxonomic validity of formally recognized morphological varieties within S. viridis varieties has been questioned because they do not differ isozymatically from typical forms of S. viridis (Wang et al. 1995a). It was observed that these varieties show no genetic differentiation from the common type, and are completely embedded in common S. viridis, S. viridis var. viridis allozyme principal component analysis (PCA) plots. Confusion with identification is highlighted by the observation that presence or absence of hairs on upper leaf surface, normally thought of as diagnostic for S. faberi, are found in both S. viridis (spp. pycnocoma, giant green) and S. faberi (green giant; Wang et al. 1995a, b). However, S. parviflora and S. pumila, natives to two different continents, were found to be so similar morphologically that the 24 accessions studied by Wang et al. (1995b) were initially identified as S. pumila. Allozyme data proved to be diagnostic in the identification of these cryptically different species. Wide morphological and physiological variations among accessions of yellow and S. faberi were observed (Santelmann and Meade 1961). Abnormal forked, divided or segmented panicles in S. viridis have been observed, although these traits apparently are not heritable. Reproductive developmental mutants have been observed in which spikelets proliferate as green scales, named S. viridis, var. vivipara (Bertol.) Parl. (Dore and McNeill 1980). Seed (spikelet) morphology and shape vary widely among individual foxtail species and genotypes. The degree of rugosity of the lemma is a valuable taxonomic diagnostic character. Rugosity varies from smooth and shiny in foxtail millet, to finely ridged in S. viridis, to very coarse rugose seed in other foxtail species (Rominger 1962). This rugosity may play an important role in interactions with the soil environment determining their behavior in soil seed banks, especially at the soil particle-seed interface affecting the uptake of water and gases into the foxtail seed placental pore

H. Darmency and J. Dekker

(see dormancy below). Similarly, the growth kinetics parameters have been seldom taken into account in the description of the variability, but they can have major impact on the competitive and reproductive output of the variants. Thus, experiments in growth cabinet suggest higher potential for competition of the giant type S. viridis spp. pycnocoma at the seedling and juvenile stage (Darmency et al. 1999).

15.4.2 Physiological Variants The very rapid increase in frequency of herbicide resistant variants in populations is the most apparent behavioral variation among Setaria phenotypes observed in management situations in the last 25 years. Cases of resistance to dalapon (Santelmann and Meade 1961), dinitroanilines, photosystem II, ACCase and ALS inhibitors were reported in Canada, USA and Europe (see references in Heap 2009). Besides mutations (see Sect. 15.9.3), the mechanisms of resistance present in weedy foxtail populations include exclusion and metabolism (e.g., Thornhill and Dekker 1993; Wang and Dekker 1995). They resulted in several marked differences with respect to the wild type, for instance chloroplast anatomy and functioning (Gasquez and Compoint 1981; Ricroch et al. 1987), seed germination (Reschly et al. 1996; Tranel and Dekker 2002), growth and reproduction (Wang et al. 2010), although not in all the cases (Wiederholt and Stoltenberg 1996). Physiological variation in dormancy and germinability induced in seed during embryogenesis is one of the most important phenotypic traits leading to colonizing ability and is discussed in detail below (see Sect. 15.6.1.4 on seed dormancy below; Dekker et al. 1996). Variation in drought tolerance among the foxtails has been observed (Blackshaw et al. 1981; Manthey and Nalawaja 1987, 1982; Taylorson 1986). Potentially salt tolerant genotypes of S. viridis, var. pachystachys have been observed along the seacoasts of Japan (Kawano and Miyake 1983; Jack Dekker personal observation 1992, 2000, data not reported). Salt tolerance exists in other Setaria species (Chapman 1992). These studies indicate that there may be an intimate physiological and morphological relationship

15 Setaria

between drought tolerance, salt tolerance and seed dormancy. This is most apparent in extreme habitats where foxtail millet remains an important crop and cultivated cereal (e.g., Central Asia including Afghanistan, India, sub-Saharan Africa). “Primitive” variants of foxtail millet (e.g., race moharia) are grown in marginal areas of Asia (e.g., Afghanistan) and are well adapted to high elevation, semi-arid conditions on poor soil (de Wet et al. 1979). These variants yield a crop in areas with as little as 125 mm of rainfall per annum.

15.4.3 Geographic Variants Locally adapted populations of Setaria occur throughout its range of distribution, yet almost nothing is known of this type of adaptation or how this local adaptation occurs. Wang et al. (1995a, b) observed north–south genotypic variation in North American populations of S. viridis and S. parviflora. Differences in S. pumila seed dormancy and germination requirements were found among five S. pumila accessions from California, Iowa, Pennsylvania, Connecticut, and Massachusetts (Schoner et al. 1978; Norris and Schoner 1980). A more complete presentation of global geographic variation is presented in the section on population genetic structure below (Sect. 15.5.2).

283

culms per plant. They have small (5–20-cm long), erect or slightly nodding panicles with short (1–3-cm long) fascicle branches on which the spikelets are arranged in clusters (de Wet et al. 1979; Prasada Rao et al. 1987). The setae are well developed. They differ from weedy variants by having spikelets that persist after maturity, aiding harvest. Variants of the race maxima are 45–100-cm tall, 1-8 (av. 2) unbranched culms, and large, pendulous panicles (to 35-cm long, 8-cm wide) with elongated secondary (fascicle) branches (2–6-cm long) forming lobed clusters (de Wet et al. 1979; Prasada Rao et al. 1987). Race maxima is highly variable and totalizes the main part of landrace accession in genetic resources bank. Variants of race indica are intermediate to the other two races, 1-25 (av. 3) culms, but are unique due to panicles 6–30 cm long whose fascicle branches are loosely arranged along the main axis, and their occurrence in India and Pakistan. In addition, one may add among the domesticated variant the giant green foxtail, S. viridis spp. pycnocoma because it is a side-product of the domestication and believed to be a hybridization product of foxtail millet and S. viridis (see Sect. 15.3.3.2). This variant is robust (1.5–2.5 m height), with larger panicles (to 20 cm), compared to other S. viridis variants. It also has larger, lobed panicle branch fascicles (to 4 cm).

15.4.5 Plasticity of Weedy Setaria 15.4.4 Domesticated Foxtail Millet Variants Hubbard (1915) found foxtail millet to be exceedingly variable, and describes in detail three subspecies, five varieties, six subvarieties, and six formae; as well as many different cultural names for the different variants. “Primitive” races of foxtail millet resemble giant S. viridis inflorescences, while more “highly evolved” (domesticated) races possess extremely large panicles (de Wet et al. 1979). A more recent classification of foxtail millet recognizes two or three different morphological races based on plant and panicle structure (de Wet et al. 1979; Prasada Rao et al. 1987). Variants of race moharia are 25–100-cm tall, with 5-52 (av. 9)

One of the most striking observations of foxtail behavior is the utilization of phenotypic heterogeneity within a single genotype or individual, rather than genetic diversity among the individuals of a population, as a means of exploiting a locality (Scheiner 1993; Wang et al. 1995a, b). This phenotypic heterogeneity takes the form of phenotypic plasticity and somatic polymorphism in many of its most important traits, especially during reproduction. For instance, the number of fertile spikelet per fascicle within each giant or green foxtail panicle is variable and more plastic in response to environmental conditions (Clark and Pohl 1996; Doust and Kellogg 2006). A large individual variability on a range of 1–20 was found for the size and reproductive output among the progeny of a single green foxtail plant in a maize

284

field, mainly dependent on earliness of emergence (Darmency et al. 1999). Most of the phenotype space of the foxtails has not been characterized. Only the most obvious behaviors and morphologies have been described (e.g., plant size and habit, color morphology, abundance of bristles, herbicide resistance), and these are usually the characters that are most apparent in management situations.

15.5 Biology and Habitat of the Wild Setaria 15.5.1 Habitat and Distribution The wild Setaria of the foxtail millet complex are well known weeds and colonize waste places (Dekker 2003, 2004). The biology of S. viridis was summarized by Douglas et al. (1985), that of S. pumila and S. verticillata by Steel et al. (1983), and that of S. faberi by Nurse et al. (2009). Frequently more than one Setaria species coexist together in a locality; possibly allowing a more complete exploitation of resources left available by human disturbance and management. Several weedy foxtail species often coexist in a single field (most commonly with S. viridis), each exploiting a slightly different “opportunity space” or niche. Their presence is not reported in natural types of plant community. Since S. viridis may be the pre-agricultural wild colonizer that humans first domesticated, it is likely that its habitat turned to be completely cultivated or managed by human activities. This probably applies for the other species of the secondary gene pool. In addition, gene flow within the wild–crop–weed complex probably has changed the original genetics of the present S. viridis, such that there are no longer any wild progenitors left in the world capable to establish in other communities. Discovering wild specimens participating stably or repeatedly in natural communities could be invaluable to enlarge the environmental scale of habitat adaptation of the germplasm resources of foxtail millet. Wild allies of foxtail millet in Eurasia are distributed in subtropical to temperate areas, and they are found under similar climates in other places of the world where they have been introduced by man. However, the species do not exactly match together:

H. Darmency and J. Dekker

S. adhaerens needs a warmer climate than S. viridis that in turn needs a warmer climate than S. verticillata. For instance, S. verticillata is found all over France, while S. viridis is recorded in the southern half part of France, and S. adhaerens in the Mediterranean border only (Tela-Botanica 2009). In contrast, foxtail millet cultivars are adapted to a wide range of climatic conditions, including in northern latitudes (North China, Mongolia, Russia) where its wild ancestor S. viridis is seldom found.

15.5.2 Population Genetic Structure The pattern of genetic diversity within an individual weedy Setaria sp. is characterized by unusually low intrapopulation genetic diversity, and unusually high genetic diversity between populations, compared to an “average” plant (Wang et al. 1995a, b). These two patterns of population genetic structure appear to typify introduced, self-pollinated weeds that are able to rapidly adapt to local conditions after invasion and colonization. Although relative genetic diversity within each of the several Setaria species is very low, differences between homogeneous populations are high, indicating a strong tendency for local adaptation by a single genotype. Nearly all populations analyzed in America consisted of a single multilocus genotype, while more diversity could be found within European and Chinese populations. Genetic bottlenecks associated with founder events may have strongly contributed to that genetic structure. The founder effect has been observed in S. viridis to a certain degree. S. viridis accessions from North America have reduced allelic richness compared to those of Eurasia. Genetic drift probably has occurred in S. viridis, as indicated by the many fixed alleles in North American accessions. Multiple introductions of S. viridis, in the absence of local adaptation, should have produced a random, mosaic pattern of geographic distribution among North American accessions. Instead, a strong intracontinental differentiation is observed in S. viridis populations, both in Eurasia and North America (Jusuf and Perne`s 1985; Wang et al. 1995a). S. viridis populations in North America are genetically differentiated into northern and southern groups separated on either side of a line at about 43.5 N latitude. The northern type is less variable than southern type. This regional divergence

15 Setaria

suggests that natural selection has partitioned S. viridis along a north–south gradient. These observations imply that the present patterning among S. viridis populations in North America is the consequence of multiple introductions into the New World followed by local adaptation and regional differentiation. S. pumila populations are genetically clustered into overlapping Asian, European, and North American groups (Wang et al. 1995b). S. pumila populations from the native range (Eurasia) contain greater genetic diversity and a higher number of unique alleles than those from the introduced range (North America). Within Eurasia, Asian populations have greater genetic diversity than those from Europe, indicating that S. pumila originated in Asia, not Europe. These observations indicate there have been numerous introductions of S. pumila from Eurasia to North America, the majority from Asia. This pattern may also explain the enigma of the origins of S. pumila and S. parviflora. The pattern of S. pumila genetic variability in North American was unexpected: nearly the entire diversity of this species appears to be encompassed by accessions from Iowa, whereas populations collected from other North American locations were nearly monomorphic for the same multilocus genotype (Wang et al. 1995b). In this respect, it is significant or coincidental that this pattern was repeated in the diversity data for S. viridis, also a native of Eurasia (Wang et al. 1995a). Iowa possesses a surprising Setaria genetic diversity: all five weedy Setaria species are present. Typically two or more Setaria species occur in the same field at the same time. Iowa is the center of the north–south agro-ecological gradient in North America, perhaps leading to greater environmental heterogeneity. Despite originating on different continents, the genetic diversity patterns for S. parviflora parallel those for S. pumila and S. viridis: greater genetic diversity occurs in accessions from the New World compared to those from the introduced range (Eurasia) (Wang et al. 1995b). This most likely reflects genetic bottlenecks associated with sampling a limited number of founding propagules and the history of multiple introductions from the Americas to Eurasia. The population genetic structure of S. parviflora consists of three nearly distinct clusters, groups from Eurasia, northern United States, and southern United States. Accessions from Eurasia and North America are approximately equally diverse genetically. Within

285

North America, S. parviflora accessions were strongly differentiated into southern and northern groups at about the Kansas-Oklahoma border (37 N latitude); indicating greater genetic differentiation within North American populations than between North American and Eurasian populations. S. faberi contains virtually no allozyme variation. Fifty of the fifty-one accessions surveyed by Wang et al. (1995b) were fixed for the same multilocus genotype.

15.6 Role of Wild Setaria in Elucidation of Origin and Evolution of Foxtail Millet 15.6.1 Analysis of the Domesticated Traits 15.6.1.1 Seed Shedding On the basis of visual estimates of two classes (shedding or not), Li et al. (1945) showed that seed shedding is controlled by two pairs of complementary genes, with shedding being dominant, and a 9:7 ratio in F2 populations. However, the frequencies obtained in the F3 were significantly different from the expected ones. Other studies showed different ratios, from 14:2 to 11:5, and around 35% of non-shedding F2 segregating in F3 (Darmency and Perne`s 1987). Finally, a full quantitative approach estimating seed shedding as the proportion of seed with attached glumes allowed to calculate a minimum number of four genes involved (Darmency et al. 1987a), probably two major genes as observed above and two minor genes. Several quantitative trait loci (QTLs) related to seed shattering, stem height, node number and length of panicle were identified using an F2 population (Kun et al. 2008).

15.6.1.2 Inflorescence Architecture The morphology of Setaria inflorescences consists of organs and tissues organized in hierarchical nested structural sets: tiller culm, panicle, fascicle, spikelet and floret (Dekker et al. 1996). The domesticated species has more primary fascicle (branch) number

286

and more spikelet number than its wild ancestor, which negatively correlates with primary branch density. No classical genetic interpretation has been given. Eleven QTLs of five linkage groups were significantly associated to these traits (Doust et al. 2005). Several candidate genes identified to have similar action in maize and rice were proposed to control these wild versus crop differences.

15.6.1.3 Tillering The profuseness of tillering due to the wild parent of the cross between S. viridis and S. italica makes it quite difficult to classify the progeny. On a visual scale, low tillering seems to be somewhat recessive and high tillering quantitatively dominant. Earlier works suggested that only one co-dominant gene was involved, meaning that 3/4th of the F2 progeny have various degrees of profuseness of tillering, and 1/4th have few tillers resembling the crop parent (Li et al. 1945). In a quantitative analysis of the parents, of the F1, F2 and F3 populations, a broad heritability estimate of h2 ¼ 0.65 was calculated (Darmency et al. 1987a). A QTL approach allowed the identification of four QTL markers associated with tillering difference between S. italica and S. viridis, and four associated with axillary branching: candidate genes could be researched at corresponding places on the gene map of other cereals (Doust et al. 2004). A further research increased the number of QTLs associated to the phenotypic distribution of the traits in F3 families, but some of these QTLs were also associated to phenotypic variation according to the environment (Doust and Kellogg 2006).

15.6.1.4 Germination No inheritance study was directly dedicated to germination and seed dormancy, probably because seed germination in weedy Setaria is controlled by many gene complexes involved in the formation and variation in seed structures. The wide geographic range of adaptation, heterogeneity in dormancy phenotypes, and genotypic diversity raise the question of what mechanisms in weedy Setaria spp. seeds drive seed behaviors (e.g., induction of dormancy, after-ripening, germination, induction of summer dormancy,

H. Darmency and J. Dekker

and seedling emergence; Dekker 1999). There is evidence that soil water, temperature and oxygen play a unifying role in regulating both the global bio-geographic distribution of weedy Setaria spp., as well as the responses of individual seeds in a soil microsite (Dekker 2000; Dekker et al. 2001; Dekker and Hargrove 2002). The morphology of weedy Setaria spp. seeds provides clues about which environmental factors limit germination and maintain dormancy. The weedy Setaria spp. seed interior (embryo, endosperm, and aleurone layer) is enveloped by several layers controlling its behavior, notably the hull (lemma, palea) and the caryopsis coat (Rost 1973; Dekker et al. 1996; Dekker and Luschei 2009). The caryopsis coat is water- and gas-tight, and continuous except at the narrow placental pore opening on the basal end of the seed. The mature weedy Setaria spp. seed is capable of freely imbibing water and dissolved gases, but entry is restricted and regulated by the placental pore tissues, notably membrane control by the transfer aleurone cell layer (Rost and Lersten 1970). This morphology strongly suggests that seed germination is restricted by water availability in the soil, and by the amount of oxygen dissolved in water reaching the inside of the seed interior to fuel metabolism. Thus, the signal stimulating Setaria behavior in an individual seed is oxygen mass per water volume of symplast (caryopsis) per time (e.g., hour, day): mass O2 volume H2O1 time1 seed1 (oxy-hydro time; Dekker et al. 2001; Dekker and Atchison 2003).

15.6.1.5 Grain Size There is no more than one single grain or seed per spikelet. This seed is indurate, usually transversely wrinkled, lemma and palea (hull) of similar marking and texture, which tightly encloses the caryopsis within at maturity (see dormancy regulation below for caryopsis morphology). Spikelet length for Setaria spp. varies both within and between species (Rominger 1962). S. viridis seeds are 1.8–2.4-mm long (var. giant S. viridis, 1.8–2.2 mm; var. robust purple and robust white, 2.0–2.4 mm), 1.0–1.3-mm wide, and weigh 0.71–0.87 mg (var. giant S. viridis, 1.08 mg; var. pachystachys 0.87 mg; var. robust purple 1.27 mg; var. robust white 1.39 mg). S. faberi seeds are

15 Setaria

2.5–3.0 mm in length, and weigh between 2.00 and 2.169 mg (Schreiber and Oliver 1971; Kawano and Miyake 1983). S. pumila seeds are 2.5–3.4-mm long, 1.5–2.2-mm wide, 1.0–1.5-mm thick, and weigh between 2.7 and 4.5 mg (Stevens 1932; Povilaitis 1956; Schreiber and Oliver 1971; Kawano and Miyake 1983; Steel et al. 1983). S. verticillata spikelets are 1.8–2.3-mm long, 1.0–1.5-mm wide, 0.6–1.3-mm thick, and weigh between 1.1 and 1.3 mg (Steel et al. 1983). Foxtail millet of maxima race are 2.8–3.5 mg. Quantitative analysis showed that grain size was highly heritable (broad heritability h2 ¼ 0.84), but it was impossible to propose any genetic system (Darmency et al. 1987b).

15.6.1.6 Grain Quality Two grain attributes appear to be tightly associated to the wild Setaria status and are rarely found among the crop cultivars, probably because they confer bitter taste to the flour: the presence of polyphenoloxidases in the seed coat, and the presence of phenolic pigments in the pericarp. The phenol color reaction is controlled by a single gene, with the positive (i.e., the wild) phenotype being dominant (Kawase and Sakamoto 1982; Darmency and Perne`s 1987; Till-Bottraud and Brabant 1990). As for pericarp pigmentation, F1 hybrids of interspecific crosses have the crop phenotype (unpigmented pericarp). Various F2 segregation ratios were observed in different studies, but all the authors propose a single gene for pigmentation with its action augmented, modified, or suppressed by the action of a second gene (Li et al. 1945; Darmency and Perne`s 1987; Till-Bottraud and Brabant 1990).

15.6.1.7 Miscellaneous Other information not directly associated to the domestication process was also obtained from the interspecific crosses. For instance, the stem base pigmentation (one dominant gene) and molecular markers including isozymes and SSR and also QTLs were used to identify linkage groups (Darmency and Perne`s 1987; Kun et al. 2008).

287

15.6.2 Application of Morphotaxonomy, Chemotaxonomy, Biochemical and Molecular Markers The evolution and domestication history of S. italica have been studied using different characters such as phenol color reaction, esterase isozymes, hybrid pollen sterility, morphological data, storage protein, RFLP, RAPD and amplified fragment length polymorphism (AFLP) molecular markers, and transposon DNA sequence. A few to several S. viridis were generally included in the studies that comprised large geographical samples of landraces and cultivars. It has been suggested that a unique domestication center in China was followed by local adaptation (Nakayama et al. 1999; Le Thierry d’Ennequin et al. 2000). Alternately, up to three domestication centers have been suggested in China, Europe and a geographical area ranging from Afghanistan to Lebanon (Kawase and Sakamoto 1982, 1987; Jusuf and Perne`s 1985; Darmency and Perne`s 1987; Li et al. 1995, 1998; Fukunaga et al. 1997; Schontz and Rether 1999; Hirano et al. 2008). A different picture is provided by the study of the heterogeneity of copia-like retrotransposons. Their sequencing from five species revealed high sequence heterogeneity even between clones from a single species (Benabdelmouna and Darmency 2003). Since the degree of sequence divergence is linked generally to phylogenetic relationships, it could have been expected that bottlenecks of diversity occurring along with domestication of the cultivated S. italica would have resulted in a much lower genetic variability in the cultivated crop than in its wild ancestor S. viridis and allied species. However, the sequence divergence was found to be as high between the two cultivars of S. italica as among the four wild Setaria species. It is possible that the high level of divergence revealed in the cultivated species resulted from numerous introgressions from related wild species. Other studies suggested that genetic differentiation occurred after domestication and that the lack of structure of the wild gene pool might be a consequence of gene flow between the crop and its wild relatives (Nakayama et al. 1999; Le Thierry d’Ennequin et al. 2000). Alternatively, multiple independent domestication events, from different S. viridis accessions, could have occurred, thus generating copia

288

sequence variability at the very origin of the crop. The variability and differences among ribosomal DNA gene sequences of foxtail millet landraces support such a multiple, independent, domestication (Fukunaga et al. 2006).

15.7 Role of Wild Setaria in Molecular Genetic Studies 15.7.1 Development of Cytogenetic Stocks and Genetic Maps In parallel to constructing a RFLP-based map from an intraspecific S. italica F2, Wang et al. (1998) constructed another map from an interspecific S. italica  S. viridis F2, in order to reveal wider polymorphism. As expected, the second map showed much more polymorphism (75% relative to 44% polymorphic loci). The map based on the interspecific cross was subsequently used as a reference for comparative studies, in particular to investigate the synteny with rice and provide researchers with large pools of known markers for gene identification (Devos et al. 1998). The map comprises nine linkage groups, which were assigned to the corresponding nine chromosomes using trisomic lines (Wang et al. 1999). One of the first outcome of this genetic approach was to allow QTL analyses to determine regions containing candidates genes in maize or rice, such as tb1 (teosinte branched 1) and ba1 (barren stalk 1) for the control of tillering and branching (Doust and Kellogg 2006). In addition, because of the plant plasticity of the wild parent according to the growth conditions, QTL regions variably associated to a given trait could be both involved in its genetic control and its phenotype expression, which potentially provides new insight in the genome organization. This map from the S. italica  S. viridis F2 is now enriched with simple sequence repeat (SSR) markers (Jia et al. 2009). A green foxtail bacterial artificial chromosome (BAC) library is under construction (X Diao personal communication). Trisomic lines and other aneuploids to identify linkage groups were obtained by crossing colchicine-induced autotetraploid S. italica crossed to normal line (Wang et al. 1999); wild tetraploid was never successfully used.

H. Darmency and J. Dekker

15.7.2 Genes and Genome Sequencing A few DNA sequences are already available. They correspond to the few genomics resources developed in studies including wild Setaria and are available at EMBL and DDBJ GenBanks. They include ribosomal DNA genes (Benabdelmouna et al. 2001b), ACCase gene (De´lye et al. 2002), copia-like retrotransposons (Benabdelmouna and Darmency 2003), a and b-tubulin genes (De´lye et al. 2004), and ribosomal intergenic spacer subrepeats (Fukunaga et al. 2006). The genome of S. italica is being sequenced. Together with a crop cultivar, an accession of S. viridis will be sequenced too and compared to that of the domesticated plant (Doust et al. 2009). It will facilitate comparative genomic analysis, assignment of domestication traits along the chromosome, thus allowing better knowledge of the genome organization and constituting a helpful tool to manage introgression when desirable genes will have to be introgressed from wild relatives.

15.8 Methods for Hybridization and Introgression of Traits from Wild Setaria 15.8.1 Hybridization Anther emasculation by hand is possible, but it is a careful, boring and time-consuming method. In many cases, emasculation was performed by dipping the entire inflorescence just starting to bloom in hot water at 42 C for 20 min (Li et al. 1935). The inflorescence was then enclosed in a glassine bag along with an inflorescence of the male parent, and the bag was shaken several times a day to allow the pollen reaching the castrated flowers. A second round of the emasculation technique 1 or 2 days later can allow up to 60% of cross-hybridization, but the hybrid status of the released seeds must be confirmed because of abortion or destruction of all the pollen grains of the inflorescence used as the female cannot be warranted. Different morphological or molecular markers are necessary to confirm the hybridity, for instance the dominant red collar pigmentation easily observed at the two-leaf stage (Darmency and Perne`s 1985), or isozymes

15 Setaria

markers (Wang and Darmency 1996), or panicle characteristics (size, anther color, bristle color; Li et al. 1945). When male sterile lines of foxtail millet were made available, hybridization became very simple by enclosing one ear of each parent in the same paper bag (Wang and Darmency 1997), but no such male sterile plants were found among the wild relatives.

15.8.2 Introgression Mass selection among the huge variability created in the F2 progeny of interspecific hybrids is a first possibility to detect favorable genetic recombination. Multivariate analyses showed that the differing characters between the crop and the wild millet seem to be distributed along three independent groups (1) the plant tillering, (2) the seed size and shedding, and (3) the size of the vegetative and reproductive organs. Other traits, such as flowering time, behave independently (Darmency et al. 1987a, b). Since important characters to differentiate the wild and the cultivated millet seems to be grouped, a rapid return to the cultivated type can be achieved in the breeding plans while new independent traits are identified and used. This fact has an important outcome on crop breeding as it opens the way for a larger use of wild plants in improving the agronomic quality of foxtail millet. High genetic variation remained within F3 families for most characters, which may provide further gain in further selection (Zangre and Darmency 1993). However, if autogamy allows easy selection of F2 and F3, backcrossing is a more difficult process. Backcrossing to the same parent can be done easily in the case of a male sterile line (Wang et al. 2009a). Otherwise, accurate hand crossing involving very few flowers and availability of differential parental markers are necessary. Alternatively, screening suitable plants among selfed F2s retaining a given marker and then backcrossing them to the cultivar can allow a rapid return to the cultivated type within only two backcross generations (Naciri et al. 1992).

15.9 Role in Crop Improvement Very little achievements of foxtail millet breeding using wild relatives have been reported. They include cases of male sterility, tetraploidy and herbicide resistance.

289

15.9.1 Male Sterility A cytoplasmic male sterile (CMS) line was obtained in the progeny of backcrossed plants after crossing a S. verticillata with an autotetraploid S. italica (Zhu et al. 1991). However, the research of restorer genes among the wild collections to setup restorer lines has not been documented yet. Nuclear encoded male sterile genes are now used in millet breeding, but they apparently belong to crosses between geographically distant landrace accessions (China and Australia) whose pedigree could have, perhaps, included wild relatives.

15.9.2 Tetraploidy Numerous crop species are polyploid so that attempts were performed to obtain tetraploid cultivars exhibiting higher productivity. However, autotetraploids resulted in poor values compared to diploids, and we are not aware of any tetraploid commercial cultivar. In a few instances, a wild species was involved in a breeding plan in order to stabilize the valuable characteristics of the tetraploid (small plant and high thousand-grain weight) and to counterbalance the deficiencies (in particular a reduced fertility), or simply to explore the potential outcome of interspecific breeding or serve as a bridge to use wild related polyploids. For instance, an autotetraploid S. italica was created to cross with S. yunnanensis, a tetraploid wild relative. The resulting hybrids were sterile, but some progeny at 2n ¼ 36 could be obtained through in vitro culture of immature inflorescences (Zhou et al. 1988). Similarly, crosses have been done between an autotetraploid S. italica line and S. faberi, a wild tetraploid species that already presents large seeds (Luo et al. 1993). Alternatively, selection of autotetraploid could be performed by colchicine induction after the production of an interspecific hybrid between S. viridis and S. italica (Ahanchede et al. 2004).

15.9.3 Herbicide Resistance Research for herbicide resistance (or tolerance) is a very recent goal in plant breeding (Darmency 2003).

290

In the case of foxtail millet, weed control remains one of the main problems to be solved as no selective herbicide has been developed for it because it does not represent a market large enough to justify the research expenses. Thus, no satisfactory weed control herbicide program exists, while hand control is more and more expensive and less and less attractive for farmers. Herbicide resistant millet cultivars were designed to overcome this situation.

15.9.3.1 Triazine Resistance Triazine is a herbicide class that inhibits the electron transfer at the photosystem II level. In Europe, atrazine at low dose was often used for weed control in foxtail millet, but its effect on weeds does not last enough to maintain an efficient protection during the whole growth season. A few populations of green foxtail resistant to 1,000 times the field doses of atrazine were found in France (Gasquez and Compoint 1981), and subsequently in other Setaria species and other countries (Heap 2009). Since no resistant cultivar was observed among a world collection of foxtail millet, the resistant green foxtail was used in order to introgress the resistance trait. Only hybrids from the green foxtail used as the female parent were confirmed to be resistant, and all their F2 progeny by self-pollination were resistant, which indicated maternal inheritance (Darmency and Perne`s 1985). Further studies showed that the Ser264–Gly mutation of the chloroplast psbA gene was the molecular basis of atrazine resistance in Setaria (Tian and Darmency 2006). This mutation at the photosystem II, however, was shown to incur a high fitness cost. When resistant BC2 were compared to the original susceptible cultivar, both had similar developmental characteristics, but the rate of photosynthesis (CO2 fixation) was lower for the resistant at 27 C while remaining not different than that of the susceptible plants at 17 C (Ricroch et al. 1987). Chlorophyll fluorescence analysis showed that electron transfer in photosystem II was slower in the resistant plants, thus representing a limiting factor at temperatures allowing higher biological activity. When both plant materials were grown in the field, this resulted in a 22% grain yield reduction for the resistant plants (Darmency and Perne`s 1989). Nevertheless, high yield lines could be derived from these crosses (Ji et al. 2006; Shi et al. 2008).

H. Darmency and J. Dekker

15.9.3.2 Trifluralin Resistance Several populations of green foxtail evolved resistance to dinitroanilines, a group of chemicals that inhibits cell division. The resistant plants were seven times more resistant to trifluralin and caused trouble in wheat fields (Beckie and Morrison 1993). The resistance was inherited as a recessive nuclear gene (Jasieniuk et al. 1994), but a strong bias was observed in the progeny of the interspecific hybrids, with on average 14% homozygous resistant instead of 25% (Wang et al. 1996b). The discovery of the mutations responsible for the resistance, a Thr239–Ile and a Leu136–Phe mutation of the gene encoding a2-tubulin (De´lye et al. 2004), allowed to check that only one structural gene was involved. The bias was inherited in the next generations, but after further crosses and generations, a normal inheritance pattern could be observed in some lines, which suggested the possibility of a linkage between the a2-tubulin and a modifier gene (Tian et al. 2006). It was inferred from the alignment of rice and millet genetic maps that the a2-tubulin gene belongs to the linkage group IX that showed distorted segregation also in an RFLP study (Wang et al. 1998). Resistant lines were further selected, but because the intensity of the resistance is not high enough to warrant good weed control in the field, registered cultivars displaying this property could hardly have been developed. However, it is a helpful tool to maintain pure male sterile lines used to produce hybrid seeds as it allows discriminating homozygous recessive resistant from heterozygous susceptible descendants developed from uncontrolled pollination (Wang et al. 1996a). An allele-specific polymerase chain reaction (PCR) routine protocol was set up to allow quick discrimination of the different alleles (De´lye et al. 2005).

15.9.3.3 Sethoxydim Resistance Several populations of green foxtail evolved resistance to sethoxydim in Canada (Heap and Morrison 1996), a herbicide of the cyclohexanedione group that blocks fatty acid biosynthesis in the Gramineae. Crosses were carried out with fertile and male sterile millet lines. Results in F1, F2 and BC1 showed that the resistant progeny was 700 times more resistant than the susceptible cultivars and that a single, completely dominant,

15 Setaria

nuclear gene controlled the resistance (Wang and Darmency 1997). Materials derived from these crosses were used in a breeding program to release a series of resistant lines, of which some (e.g., variety SR3522) were registered as elite cultivar at the Seed Supervising Bureau of Zhangjiakou in China (2005). The resistance was demonstrated to be due to a point mutation in the carboxyltransferase domain of the nuclear gene encoding plastidic ACCase isoform, a Ile1781–Leu residue substitution (De´lye et al. 2002). It was not clear whether the resistant allele has fitness cost because it was possibly linked to genes boosting juvenile growth, which resulted in being advantageous in stressing field conditions (Wang et al. 2009a). Close AFLP markers are now available to facilitate selection (Niu et al. 2002).

15.9.3.4 Other Sources of Resistance Further sources of differential selectivity of herbicide were demonstrated to occur in various wild Setaria for dalapon and butylate (Santelmann and Meade 1961; Oliver and Schreiber 1971), and for PSII inhibitors (Thornhill and Dekker 1993; Wang and Dekker 1995). Perhaps, other mutant alleles, expressing different pattern of cross-resistance, could be found in the giant foxtail resistant to ACCase inhibitors (Stoltenberg and Wiederholt 1995; Volenberg and Stoltenberg 2002). High resistance against imidazolinone and sulfonylurea herbicides, two groups of molecules that block the biosynthesis of branched-chain amino acids, was found in several populations of green, yellow and giant foxtail (Volenberg et al. 2001; Heap 2009) in Canada and USA. In the case of S. viridis, the resistance was demonstrated to be due to several mutations at the acetolactate synthase genes: Ser653–Thr, Ser653–Asn, Ser653–Ile, and Gly654–Asp (Laplante et al. 2009). Crosses between male sterile millet and plants of one of these IMIresistant populations were undertaken in China (T. Wang personal communication).

15.9.4 Other Desirable Agricultural Traits Beside the use as a cereal for food, grain to feed poultry and spring forage for cattle, there are varieties

291

of uses of foxtail millet (Kimata et al. 1998). This diversity of use, the need to get a new taste or a new color for food, the increasing importance of plants as sources of therapeutic compounds and dietary supplements, could encourage searching for new quality attributes among wild relatives. Because the native gene pool for Setaria is very large its potential is vast, especially as a crop in poor, dry and salty environments.

15.10 Conclusion The Setaria (foxtail) wild–crop–weed species-group is one of the most successful terrestrial plant complexes on earth. The genotypic and phenotypic biodiversity of this species-group has conferred on it the ability to invade, colonize, adapt and endure in a wide range of disturbed habitats in temperate, tropical, and subtropical regions. Over 100 Setaria species were distributed throughout the world before the advent of agriculture. With crop domestication and land cultivation about 10,000 years ago, wild, pre-agricultural, colonizing foxtails could adapt and thrive in the newly created agro-ecosystems. The success of the foxtails is due to the intentional and unintentional consequences of weed management for over 10,000 years. Therefore, the future of the foxtail complex is to continuously adapt to the management systems of human-imposed selection forces, and also to changing climate and soil conditions. Evolution of a complex seed dormancy system has given the foxtails the ability to precisely time seedling emergence on time scales from hours to decades, providing them with the ability for enduring occupation of a locality. Foxtail soil seed banks are the source of all future annual weed infestations in a locality, spreading the risk of continued development over time, a bet-hedging strategy balancing mortality against future fitness. Phenotypic plasticity in growth and development has allowed the foxtails to optimize the match between the immediate resources available and vegetative size and reproductive output, avoiding pre-mature mortality of the individual or its parts. The foxtails are endowed with the ability to tolerate high concentrations of salt in the soil, low soil water, tillage, and extremes in temperature. Because of their ability to resist herbicides and adapt to most advanced and powerful farming practices, they cannot be

292

considered as endangered wild resources. As a consequence, the best germplasm bank for preserving future potentiality of this wild genetic pool is the agroecosystem itself. Besides, it should be necessary to prospect populations adapted to extreme conditions in order to preserve a specific case of unique genetic adaptation, in particular, among the species that do not belong to the primary and secondary gene pool, and originate from abroad, Eurasia.

References Ahanchede A, Poirier-Hamon S, Darmency H (2004) Why no tetraploid cultivar of foxtail millet? Genet Resour Crop Evol 51:227–230 Austin DF (2006) Fox-tail millets (Setaria: Poaceae) - abandoned food in two hemispheres. Econ Bot 60:143–158 Beckie HJL, Morrison IN (1993) Effect of ethalfluralin and other herbicides on trifluralin-resistant green foxtail (Setaria viridis). Weed Technol 7:6–14 Benabdelmouna A, Darmency H (2003) Copia-like retrotransposon in the genus Setaria: sequence heterogeneity, species distribution and chromosomal organization. Plant Syst Evol 237:127–136 Benabdelmouna A, Shi Y, Abirached-Darmency M, Darmency H (2001a) Genomic in situ hybridization (GISH) discriminates between the A and B genomes in diploid and tetraploid Setaria species. Genome 44:685–690 Benabdelmouna A, Abirached-Darmency M, Darmency H (2001b) Phylogenetic and genomic relationships in Setaria italica and its close relatives based on the molecular diversity and chromosomal organization of 5S and 18S-58S-25S rDNA genes. Theor Appl Genet 103:668–677 Blackshaw RE, Stobbe EH, Shayewich CF, Woodbury W (1981) Influence of soil temperature and soil moisture on S. viridis (Setaria viridis) establishment in wheat (Triticum aestivum). Weed Sci 29:179–184 Brown WV (1948) A cytological study in the Gramineae. Amer J Bot 35:382–395 Chapman GP (1992) Apomixis and evolution. In: Chapman GP (ed) Grass evolution and domestication. Cambridge University Press, Cambridge, UK, pp 138–156 Clark LG, Pohl RW (1996) Agnes Chase’s first book of grasses, 4th edn. Smithsonian Institution Press, Washington, pp 68–69 Clayton WD (1980) Setaria. In: Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, Webb DA (eds) Flora Europea, Vol 5. Cambridge Univ Press, Cambridge, pp 263–264 Darmency H (2003) Transgenic herbicide-resistant crops: what makes the difference? In: Lelley T, Balazs E, Tepfer M (eds) Ecological impact of GMO dissemination in agroecosystems. OECD, Facultas Verlag, Wien, Austria, pp 77–88 Darmency H (2004) Incestuous relations of foxtail millet (Setaria italica) with its parents and cousins. In: Gressel J (ed) Crop ferality and volunteerism: a threat to food security in the transgenic era. CRC, Boca Raton, pp 81–96

H. Darmency and J. Dekker Darmency H, Perne`s J (1985) Use of wild Setaria viridis (L) Beauv to improve triazine resistance in cultivated S. italica (L) by hybridization. Weed Res 25:175–179 Darmency H, Perne`s J (1987) An inheritance study of domestication in foxtail millet using an interspecific cross. Plant Breed 99:30–34 Darmency H, Perne`s J (1989) Agronomic performance of a triazine resistant foxtail millet (Setaria italica (L) Beauv). Weed Res 29:147–150 Darmency H, Ouin C, Perne`s J (1987a) Breeding foxtail millet (Setaria italica) for quantitative traits after interspecific hybridization and polyploidization. Genome 29: 453–456 Darmency H, Zangre GR, Perne`s J (1987b) The wild-weed-crop complex in Setaria: a hybridization study. Genetica 75:103–107 Darmency H, Lefol E, Chadoeuf R (1992) Risk assessment of the release of herbicide resistant transgenic crops: two plant models. In: 9th Colloque international sur la biologie des mauvaises herbes, Dijon, France, pp 513–523 Darmency H, Assemat L, Wang T (1999) Millet as a model-crop to assess the impact of gene flow toward weed populations. In: Lutman P (ed) Gene flow and agriculture, relevance for transgenic crops. British Crop Protection Council Symposium Proceedings No 72, Keele, pp 261–267 De Wet JMJ (1954) Chromosome numbers of a few South African grasses. Cytologia 19:97–103 De Wet JMJ, Oestry-Stidd LL, Cubero JI (1979) Origins and evolution of foxtail millets (Setaria italica). J Agric Trad Bot Appl 26:53–64 Dekker J (1999) Soil weed seed banks and weed management. In: Buhler D (ed) Expanding the context of weed management. Haworth, New York, USA; J Crop Prod 2:139–166 Dekker J (2000) Emergent weedy foxtail (Setaria spp.) seed germinability behavior. In: Black M, Bradford KJ, VasquezRamos J (eds) Seed biology: advances and applications. CAB International, Wallingford, UK, pp 411–423 Dekker J (2003) The foxtail (Setaria) species-group. Weed Sci 51:641–646 Dekker J (2004) The evolutionary biology of the foxtail (Setaria) species-group. In: Inderjit K (ed) Weed biology and management. Kluwer Academic, Dordrecht, The Netherlands, pp 65–113 Dekker JB, Atchison JK (2003) Setaria spp. seed pool formation and initial assembly in agro-communities. Asp Appl Biol 69:247–259 Dekker J, Hargrove M (2002) Weedy adaptation in Setaria spp.: V. Effects of gaseous atmosphere on giant foxtail (Setaria faberii) (Gramineae) seed germination. Am J Bot 89: 410–416 Dekker J, Luschei EC (2009) Water partitioning between environment and Setaria faberi seed exterior-interior compartments. Agric J 14604:66–76 Dekker J, Dekker BI, Hilhorst H, Karssen C (1996) Weedy adaptation in Setaria spp.: IV. Changes in the germinative capacity of S. faberi embryos with development from anthesis to after abscission. Am J Bot 83:979–991 Dekker J, Lathrop J, Atchison B, Todey D (2001) The weedy Setaria spp. phenotype: how environment and seeds interact from embryogenesis through germination. In: Proceedings

15 Setaria of British Crop Protection Conference–Weeds, Brighton, UK, pp 65–74 De´lye C, Wang T, Darmency H (2002) An isoleucine-leucine substitution in chloroplastic acetyl-CoA carboxylase from green foxtail (Setaria viridis L Beauv) is responsible for resistance to the cyclohexanedione herbicide sethoxydim. Planta 214:421–427 De´lye C, Menchari Y, Michel S, Darmency H (2004) Molecular bases for sensitivity to tubulin-binding herbicides in green foxtail. Plant Physiol 136:3920–3932 De´lye C, Menchari Y, Michel S (2005) A single polymerase chain reaction-based assay for simultaneous detection of two mutation conferring resistance to tubulin-binding herbicides in Setaria viridis. Weed Res 45:228–235 Devos KM, Wang ZM, Beales J, Sasaki T, Gale MD (1998) Comparative genetic maps of foxtail millet (Setaria italica) and rice (Oryza sativa). Theor Appl Genet 96:63–68 Dore WG, McNeill J (1980) Grasses of Ontario. Research Branch, Agriculture and Agri-Food Canada, Monograph 26, Hull, pp 482–494 Douglas BJ, Thomas AG, Morrison IN, Maw MG (1985) The biology of Canadian weeds. 70. Setaria viridis (L.) Beauv. Can J Plant Sci 65:669–690 Doust AN, Kellogg EA (2002) Inflorescence diversification in the Panicoid “Bristle Grass” clade (Paniceae, Poaceae): evidence from molecular phylogenies and developmental morphology. Am J Bot 89:1203–1222 Doust AN, Kellogg EA (2006) Effect of genotype and environment on branching in weedy green millet (Setaria viridis) and domesticated foxtail millet (Setaria italica) (Poaceae). Mol Ecol 15:1335–1349 Doust AN, Devos KM, Gadberry MD, Gale MD, Kellogg EA (2004) Genetic control of branching in foxtail millet. Proc Natl Acad Sci 101:9045–9050 Doust AN, Devos KM, Gadberry MD, Gale MD, Kellogg EA (2005) The genetic basis for inflorescence variation between foxtail millet and green millet (Poaceae). Genetics 169: 1659–1672 Doust AN, Kellogg EA, Devos KM, Bennetzen JL (2009) Foxtail millet: a sequence-driven grass model system. Plant Physiol 149:137–141 Emery WHP (1957) A study of reproduction in Setaria macrostachya and its relatives in the southwestern United States and Northern Mexico. Bull Torrey Bot Club 84: 106–121 Fukunaga K, Domon E, Kawase M (1997) Ribosomal DNA variation in foxtail millet, Setaria italica (L.) Beauv., and a survey of variation from Europe and Asia. Theor Appl Genet 95:751–756 Fukunaga K, Ichitani K, Kawase M (2006) Phylogenetic analysis of the rDNA intergeneric spacer subrepeats and its implication for the domestication history of foxtail millet, Setaria italica. Theor Appl Genet 113:261–269 Gao M, Chen J (1988) Isozymic studies on the origin of cultivated foxtail millet. Acta Agric Sin 14:131–136 Gasquez J, Compoint JC (1981) Observation de chloroplasts re´sistants aux triazines chez une panicoide´e, Setaria viridis L. Agronomie 1:923–926 Gu SL (1987) Foxtail millet cultivation in China. China Agriculture Press, Beijing, PRC

293 Gupta PK, Yashvir AL (1973) Abnormal meiosis in hexaploid Setaria verticillata. Phyton 15:31–36 Harlan JR, de Wet JMJ (1971) Towards a rational taxonomy of cultivated plants. Taxon 20:509–517 Heap I (2009) The international survey of herbicide resistant weeds. www.weedscience.com. Accessed 19 Nov 2009 Heap I, Morrison I (1996) Resistance to aryloxyphenoxypropionate and cyclohexanedione herbicides in green foxtail (Setaria viridis). Weed Sci 44:25–30 Hirano R, Naito K, Fukunaga K, Okumoto Y, Tanisaka T, Watanabe KN, Kawase M (2008) Genetic diversity demonstrated by transposon display in foxtail millet (Setaria italica (L.) P. Beauv.). In: 114th Meeting on Japanese Society on Breeding, Ikushugaku Kenkyu 10:238 Hitchcock AS (1971) Manual of the grasses of the United States, 2nd edn, vol 2. Dover, New York, pp 724–726 Hubbard FT (1915) A taxonomic study of Setaria and its immediate allies. Am J Bot 2:169–198 Jank L, Quesenberry KH, Sollenberger LE, Wofford DS, Lyrene PM (2007) Selection of morphological traits to improve forage characteristics of Setaria sphacelata grown in Florida. NZ J Agric Res 50:73–83 Jasieniuk M, Bruˆle´-Babel A, Morrison I (1994) Inheritance of trifluralin resistance in green foxtail (Setaria viridis). Weed Sci 42:123–127 Ji G, Du R, Hou S, Ceng R, Wang X, Zhao X (2006) Genetics, development and application of cytoplasmic herbicide resistance in foxtail millet. Sci Agric Sin 39:879–885 Jia X, Zhang Z, Liu Y, Zhang C, Shi Y, Song Y, Wang T, Li Y (2009) Development and genetic mapping of SSR markers in foxtail millet (Setaria italica (L.) P. Beauv.). Theor Appl Genet 18:821–829 Jusuf M, Perne`s J (1985) Genetic variability of foxtail millet (Setaria italica P. Beauv.). Electrophoretic study of five isoenzyme systems. Theor Appl Genet 71:385–391 Kawano S, Miyake S (1983) The productive and reproductive biology of flowering plants. X. Reproductive energy allocation and propagule output of five congeners of the genus Setaria (Gramineae). Oecologia 57:6–13 Kawase M, Sakamoto S (1982) Geographical distribution and genetic analysis of phenol color reaction in foxtail millet, Setaria italica (L.) P. Beauv. Theor Appl Genet 63:117–119 Kawase M, Sakamoto S (1987) Geographical distribution of landrace groups classified by intraspecific hybrid pollen sterility in foxtail millet, Setaria italica (L.) P. Beauv. Jpn J Breed 37:1–9 Kihara H, Kishimoto E (1942) Bastarde zwischen Setaria italica und S. viridis. Bot Mag 56:62–67 Kimata M, Kanoda M, Seetharam A (1998) Traditional and modern utilizations of millets in Japan. Environ Educ Stud 8:21–29 Kimata M, Ashok EG, Seetharam A (2000) Domestication, cultivation and utilization of two small millets, Brachiaria ramosa and Setaria glauca (Poaceae), in south India. Econ Bot 54:217–227 Khosla PK, Sharma ML (1973) Cytological observations on some species of Setaria. Nucleus 26:38–41 Kun Y, Ma L, Lin L, Hui Z, Wei L, Wang Y, Li H, He W, Shang Z, Diao X (2008) Construction of SSR based linkage map and QTL analysis of several important traits in foxtail millet,

294 Setaria italica Beauv. http://www.plantgenomics.cn/abslist. cgi?absid¼602. Accessed 19 Nov 2009 Laplante J, Rajcan I, Tardif FJ (2009) Multiple allelic forms of acetohydroxyacid synthase are responsible for herbicide resistance in Setaria viridis. Theor Appl Genet 119: 577–585 Le Thierry d’Ennequin M, Panaud O, Brown S, Siljak-Yakovlev SA (1998) First evaluation of nuclear DNA content in Setaria genus by flow cytometry. J Hered 89:556–559 Le Thierry d’Ennequin M, Panaud O, Toupance B, Sarr A (2000) Assessment of genetic relationships between Setaria italica and its wild relative S. viridis using AFLP markers. Theor Appl Genet 100:1061–1066 Li Y (1995) Foxtail millet breeding. China Agriculture Press, Beijing, PRC Li Y, Wu S (1996) Traditional maintenance and multiplication of foxtail millet (Setaria italica (L.) P. Beauv.) landraces in China. Euphytica 87:33–38 Li HW, Meng CJ, Liu TN (1935) Problems in the breeding of millet (Setaria italica (L.) Beauv.). J Am Soc Agron 27:963–970 Li CH, Pao WK, Li HW (1942) Interspecific crosses in Setaria. II. Cytological studies of interspecific hybrids involving: 1, S. faberii and S. italica, and 2, a three way cross, F2 of S. italica  S. viridis and S. faberii. J Hered 33:351–355 Li HW, Li CH, Pao WK (1945) Cytological and genetical studies of the interspecific cross of the cultivated foxtail millet, Setaria italica (L.) Beauv., and the green foxtail millet, S. viridis L. J Am Soc Agron 37:32–54 Li Y, Wu S, Cao Y (1995) Cluster analysis of an international collection of foxtail millet (Setaria italica (L.) P. Beauv.). Euphytica 83:79–85 Li Y, Jia J, Wang Y, Wu S (1998) Intraspecific and interspecific variation in Setaria revealed by RAPD analysis. Genet Resour Crop Evol 45:279–285 Linneaus C (1753) Species plantarum. Holmiae, L. Salvius, Stockholm Luo X, Guo F, Zhou J, Ma H, Wu Q, Zhu G, Ma Y (1993) Immature embryo culture of a cross between Setaria italica (Ch4n) and S. faberii and studies on the morphological and cytological characteristics of the F1 plant. Acta Agric Sin 19:352–358 Manthey DR, Nalawaja JD (1982) Moisture stress effects on foxtail seed germination. Proc North Central Weed Control Conf 37:52–53 Manthey DR, Nalawaja JD (1987) Germination of two foxtail (Setaria) species. Weed Technol 1:302–304 Mulligan GA, Findlay JN (1970) Reproductive systems and colonization in Canadian weeds. Can J Bot 48:859–860 Naciri Y, Belliard J (1987) Le millet Setaria italica une plante a rede´couvrir. J Agric Trad Bot Appl 34:65–87 Naciri Y, Darmency H, Belliard J, Dessaint F, Perne`s J (1992) Breeding strategy in foxtail millet, Setaria italica (LP Beauv), following interspecific hybridization. Euphytica 60:97–103 Nakayama H, Namai H, Okuno K (1999) Geographical variation of the alleles at the two prolamin loci, Pro1 and Pro2, in foxtail millet, Setaria italica (L.) P. Beauv. Genes Genet Syst 74:293–297 Nakayama H, Nagamine T, Hayashi N (2005) Genetic variation of blast resistance in foxtail millet (Setaria italica (L.)

H. Darmency and J. Dekker P. Beauv.) and its geographic distribution. Genet Resour Crop Evol 52:863–868 Nasu H, Momohara A, Yasuda Y (2007) The occurrence and identification of Setaria italica (L.) P. Beauv. (foxtail millet) grains from the Chengtoushan site (ca. 5800 cal B.P.) in central China, with reference to the domestication centre in Asia. Veg Hist Archaeobot 16:481–494 Nguyen Van E, Perne`s J (1985) Genetic diversity of foxtail millet (Setaria italica). In: Jacquard P (ed) Genetic differentiation and dispersal in plants. NATO ASI Ser G5, Springer, Berlin, pp 113–128 NIAS (2009) Genebank. http://www.gene.affrc.go.jp/databasesplant_search_en.php. Accessed 19 Nov 2009 Niu Y, Li Y, Shi Y, Song Y, Ma Z, Wang T, Darmency H (2002) AFLP mapping for the gene conferring sethoxydim resistance in foxtail millet (Setaria italica L Beauv). Acta Agric Sin 28:359–362 Norris RF, Schoner CA Jr (1980) Yellow foxtail (Setaria lutescens) biotype studies: dormancy and germination. Weed Sci 28:159–163 Nurse RE, Darbyshire SJ, Bertin C, DiTommaso A (2009) The biology of Canadian weeds. 141. Setaria faberi Herrm. Can J Plant Sci 89:379–404 Oliver LR, Schreiber MM (1971) Differential selectivity of herbicides on six Setaria taxa. Weed Sci 19:428–431 Osada T (1989) Illustrated grasses of Japan. Heibonsha Pub, Tokyo Pohl RW (1951) The genus Setaria in Iowa. Iowa State J Sci 25:501–508 Pohl RW (1962) Notes on Setaria viridis and S. faberi (Gramineae). Brittonia 14:210–213 Poirier-Hamon S, Perne`s J (1986) Instabilite´ chromosomique dans les tissus somatiques des descendants d’un hybride interspe´cifique Setaria verticillata (P. Beauv.)  Setaria italica (P. Beauv.). CR Acad Sci (Paris) 302:319–324 Povilaitis B (1956) Dormancy studies with seeds of various weed species. Proc Int Seed Testing Assoc 21:87–111 Prasada Rao KE, De Wet JMJ, Brink DE, Mengesha MH (1987) Infraspecific variation and systematics of cultivated Setaria italica, foxtail millet (Poaceae). Econ Bot 41:108–116 Reschly B, Dekker JH, Stoltenberg DE (1996) Comparison of seed germinability between acetyl coenzyme A carboxylase inhibitor resistant and susceptible giant foxtail. American Society of Agronomy, Madison, WI, USA, Agronomy Abstracts 6 Ricroch A, Mousseau M, Darmency H, Perne`s J (1987) Comparison of triazine-resistant and -susceptible cultivated Setaria italica L (PB): growth and photosynthetic capacity. Plant Physiol Biochem 25:29–34 Rominger JM (1962) Taxonomy of Setaria (gramineae) in North America. Illinois Biol Monogr 29:78–98 Rost TL (1973) The anatomy of the caryopsis coat in mature caryopses of the yellow foxtail grass (Setaria lutescens). Bot Gaz 134:32–39 Rost TL, Lersten NR (1970) Transfer aleurone cells in Setaria lutescens (Gramineae). Protoplasma 71:403–408 Santelmann PW, Meade JA (1961) Variation in morphological characteristics and dalapon susceptibility within the species Setaria lutescens and S. faberii. Weeds 9:406–410 Santleman PW, Meade JA, Peters RA (1963) Growth and development of yellow foxtail and giant foxtail. Weeds 11:139–142

15 Setaria Scheiner SM (1993) Genetics and evolution of phenotypic plasticity. Annu Rev Ecol Syst 24:35–68 Schoner CA, Norris RF, Chilcote W (1978) Yellow foxtail (Setaria lutescens) biotype studies: growth and morphological characteristics. Weed Sci 26:632–636 Schontz D, Rether B (1999) Genetic variability in foxtail millet, Setaria italica (L.) P. Beauv.: Identification and classification of lines with RAPD markers. Plant Breed 118:190–192 Schreiber MM (1965) Effect of date of planting and stage of cutting on seed production of giant foxtail. Weeds 13: 60–62 Schreiber MM, Oliver LR (1971) Two new varieties of Setaria viridis. Weed Sci 19:424–427 Seetharam A (1998) Small millets research: achievements during 1947–97. J Agric Sci 68:431–438 Shi Y, Wang T, Li Y, Darmency H (2008) Impact of transgene inheritance on the mitigation of gene flow between crops and their wild relatives: the example of foxtail millet. Genetics 180:969–975 Singh RV, Gupta PK (1977) Cytological studies in the genus Setaria (gramineae). Cytologia 42:483–493 Stace CA (1975) Hybridization and the flora of the British Isles. Academic Press, London Steel MG, Cavers PB, Lee SM (1983) The biology of Canadian weeds. 59. Setaria glauca (L.) Beauv. and Setaria verticillata (L.) Beauv. Can J Plant Sci 63:711–725 Stevens OA (1932) The number and weight of seeds produced by weeds. Am J Bot 19:784–794 Stoltenberg DE, Wiederholt RJ (1995) Giant foxtail (Setaria faberi) resistance to aryloxyphenoxypropionate and cyclohexanedione herbicides. Weed Sci 43:527–535 Takahashi N, Hoshino T (1934) Natural crossing in Setaria italica (Beauv.). Proc Crop Sci Soc Jpn 6:3–19 Taylorson RB (1986) Water stress induced germination of giant foxtail (Setaria faberi) seeds. Weed Sci 34:871–875 Tela-Botanica (2009) http://www.tela-botanica.org/page:eflore. Accessed 09 Mar 2009 Thornhill R, Dekker J (1993) Mutant weeds of Iowa: V. S-triazine resistant giant foxtail (Setaria faberii Hermm.). J Iowa Acad Sci 100:13–14 Tian X, Darmency H (2006) Rapid bidirectional allele-specific PCR identification for triazine resistance in higher plants. Pest Manag Sci 62:531–536 Tian X, De´lye C, Darmency H (2006) Molecular evidence of biased inheritance of trifluralin herbicide resistance in foxtail millet. Plant Breed 125:254–258 Till-Bottraud I, Brabant P (1990) Inheritance of some Mendelian factors in intra- and interspecific crosses between Setaria italica and Setaria viridis. Theor Appl Genet 80:687–692 Till-Bottraud I, Reboud X, Brabant P, Lefranc M, Rherissi B, Vedel F, Darmency H (1992) Outcrossing and hybridization in wild and cultivated foxtail millets: consequences for the release of transgenic crops. Theor Appl Genet 83: 940–946 Tranel D, Dekker J (2002) Differential seed germinability in triazine-resistant and -susceptible giant foxtail (Setaria faberii). Asian J Plant Sci 1:334–336 Upadhyaya HD, Pundir RPS, Gowda CLL, Reddy VG, Singh S (2008) Establishing a core collection of foxtail millet to

295 enhance the utilization of germplasm of an underutilized crop. Plant Genet Resour 7:177–184 Varadinov SG (1986) Initial material for breeding foxtail millet. Sb Nauch Trud Prikl Bot Genet 105:103–106 Volenberg DS, Stoltenberg DE (2002) Giant foxtail (Setaria faberi) outcrossing and inheritance of resistance to acetylcoenzyme A carboxylase inhibitors. Weed Sci 50:622–627 Volenberg DS, Stoltenberg DE, Boerboom CM (2001) Biochemical mechanism and inheritance of cross-resistance to acetolactate synthase inhibitors in giant foxtail. Weed Sci 49:635–641 Wang T, Darmency H (1996) Comparison of growth and yield of foxtail millet (Setaria italica) resistant and susceptible to acetyl-coenzyme A carboxylase inhibiting herbicides. In: 10th Colloque international sur la biologie des mauvaises herbes, Dijon, France, pp 203–210 Wang T, Darmency H (1997) Inheritance of sethoxydim resistance in foxtail millet, Setaria italica (L) Beauv. Euphytica 94:69–73 Wang RL, Dekker J (1995) Weedy adaptation in Setaria spp. II. Variation in herbicide resistance in Setaria spp. Pestic Biochem Physiol 51:99–116 Wang RL, Wendel JF, Dekker JH (1995a) Weedy adaptation in Setaria spp. I. Isozyme analysis of genetic diversity and population genetic structure in Setaria viridis. Am J Bot 82:308–317 Wang RL, Wendel JF, Dekker JH (1995b) Weedy adaptation in Setaria spp. III. Genetic diversity and population genetic structure in S. glauca, S. geniculata and S. faberii. Am J Bot 82:1031–1039 Wang T, Du R, Chen H, Darmency H, Fleury A (1996a) A new way of using herbicide resistant gene on hybrid utilization in foxtail millet. Sci Agric Sin 29:96 Wang T, Fleury A, Ma J, Darmency H (1996b) Genetic control of dinitroaniline resistance in foxtail millet (Setaria italica). J Hered 87:423–426 Wang T, Chen HB, Reboud X, Darmency H (1997) Pollenmediated gene flow in an autogamous crop: Foxtail millet (Setaria italica). Plant Breed 116:579–583 Wang Z, Devos KM, Liu CJ, Wang RQ, Gale MD (1998) Construction of RFLP-based maps of foxtail millet, Setaria italica (L.) P. Beauv. Theor Appl Genet 96:31–36 Wang R, Gao J, Liang GH (1999) Identification of primary trisomics and other aneuploids in focxtail millet. Plant Breed 118:59–62 Wang T, Zhao Z, Yan H, Li Y, Song Y, Ma Z, Darmency H (2001) Gene flow from cultivated herbicide-resistant foxtail millet to its wild relatives: a basis for risk assessment of the release of transgenic millet. Acta Agric Sin 27:681–687 Wang T, Li Y, Shi Y, Reboud X, Darmency H, Gressel J (2004) Low frequency transmission of a plastid-encoded trait in Setaria italica. Theor Appl Genet 108:315–320 Wang Y, Zhi H, Li W, Li H, Wang Y, Diao X (2007) Chromosome number identification of some wild Setaria species. J Plant Genet Resour 8:159–164 Wang T, Picard JC, Tian X, Darmency H (2009a) A herbicideresistant ACCase 1781 Setaria mutant shows higher fitness than wild type. Heredity. doi: 10.1038/hdy.2009.183 Wang Y, Zhi H, Li W, Li H, Wang Y, Diao X (2009b) A novel genome of C and the first autotetroploid species in the

296 Setaria genus identified by genomic in situ hybridization. Genet Resour Crop Evol 56:843–850 Wang T, Shi Y, Li Y, Song Y, Darmency H (2010) Population growth rate of Setaria viridis in absence of herbicide and resulting yield losses in foxtail millet. Weed Res 50:228–234 Wiederholt RJ, Stoltenberg DE (1996) Absence of differential fitness between giant foxtail (Setaria faberi) accessions resistant and susceptible to Acetyl-Coenzyme A Carboxylase inhibitors. Weed Sci 44:18–24 Williams RD, Schreiber MM (1976) Numerical and chemotaxonomy of the green foxtail complex. Weed Sci 24:331–335

H. Darmency and J. Dekker Willweber-Kishimoto E (1962) Interspecific relationships in the genus Setaria. Contrib Biol Kyoto Univ 14:1–41 Zangre GR, Darmency H (1993) Potential for selection in the progeny of an interspecific hybrid in foxtail millet. Plant Breed 110:172–175 Zhou J, Luo X, Guo F, Ma H, Zhu G (1988) Plant regeneration in tissue culture of Setaria yunnanensis  S. italica (4n) F1 plants. Acta Agric Sin 14:227–231 Zhu G, Wu Q, Ma Y (1991) Breeding of new type of male sterility “Ve” in foxtail millet. J Shanxi Agric 1:7

Chapter 16

Zoysia Shin-ichi Tsuruta, Makoto Kobayashi, and Masumi Ebina

16.1 Basic Botany of the Species and Conservation Initiatives The members of the genus Zoysia Willd. (family Poaceae, subfamily Chloridoideae, tribe Zoysieae) are mat-forming perennial grasses, consisting of about ten species, which are indigenous to the Pacific Rim (Clayton and Renvoize 1986; Engelke and Anderson 2003). The native distribution of the genus extends from New Zealand to the Hokkaido Island of Japan, including Polynesia and Southeast Asian coastal countries (Engelke and Anderson 2003). Two of the species, Z. japonica and Z. machrostachya, are scattered in the some parts of Hokkaido (Ishida 1990; Osada 1993). Hokkaido has a subarctic climate, and the genus Zoysia is one of the C4 grasses most well adapted to cold climatic regions (Hatch and White 2004). Therefore, the genus Zoysia is one of the most salt and cold tolerant C4 grass species in the family Poaceae. Three commercially important species are recognized in the genus Zoysia and referred to as zoysia grass, Z. japonica Steud., Z. matrella (L.) Merrill, and Z. tenuifolia Will. ex Trin. At least five species of the genus, including the three commercially important species, are native to the Japanese islands (Fig. 16.1), and they exhibit morphological and ecological differences (Ishida 1990). Therefore, Japan is considered to be one of the origins of the diversity of the genus M. Ebina (*) National Institute of Livestock and Grassland Science, Forage Plant Breeding and Biotechnology, 768 Sembonmatsu, Nasushiobara, Tochigi 329-2793, Japan e-mail: [email protected] Shin-ichi Tsuruta and Makoto Kobayashi contributed equally and should be viewed as first authors.

Zoysia. Because the five Japanese species have the same chromosome number (2n ¼ 4x ¼ 40) and there is no chromosomal barrier of interspecific crossing, intermediate types are found along species’ borders in their natural habitats in Japan and have allotetraploidy (Forbs 1952; Yaneshita et al. 1999). Pistils are mature for 7–10 days prior to stamens, thus allogamous crossing is considered to be the preferred mode of reproduction. The genus is self-compatible, however, and has very low canopy height and comparatively heavy pollen, so the ratio of allogamous crossing in natural habitats is open to question. Z. japonica is distributed in Japan, Korea, and eastern China. The species is perennial and forms large mats in sunny fields or is cultivated as lawn, with long creeping rhizomes. Culms are 5–15-cm tall. Leaf blades are 3–10-cm long and 2–5 (or 6) mm wide, flat, with no ligule, but with long hairs at the mouths of sheaths. Spikelets are obliquely ovate, about 3-mm long (two to three times that of the breadth), and one-flowered; they consist of the upper glume and lemma and lack the lower glume and palea. The upper glume is coriaceous, smooth, glossy, and keeled on the back, with five faint, barely visible nerves; the lemma is papery, slightly shorter than the glume, one-nerved, and sharply keeled. Anthers are 1.5-mm long (Osada 1993). Along the southern ends of Tanegashima and Yakushima Islands in Japan, the biogeographical line of Watashe is recognized at a latitude 30 N. Z. japonica is found north of this latitude. Also, Blakiston’s line is recognized between Honshu and Hokkaido Islands at latitude 40 N (Tamate et al. 1998). Because the distribution of Z. japonica on Hokkaido is limited and scattered, the northern edge of the species’ original natural habitat is considered to be Honshu (Ishida 1990). Z. japonica adapts to low-input sustainable

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0_16, # Springer-Verlag Berlin Heidelberg 2011

297

298

S. Tsuruta et al.

Fig. 16.1 The five species of Zoysia indigenous to Japan: Z. japonica, Z. matrella, Z. tenuifolia, Z. macrostachya, and Z. sinica

grazing fields and is commonly used in the borders of paddy fields (Yamamoto et al. 1988; Otani et al. 1999; Ogura et al. 2001). The other four species, native to Japan, have strong salt tolerance, and only Z. japonica has a weak tolerance to salt (Oishi and Ebina 2005). Z. japonica is distributed naturally in mountainous areas, along riversides, and in coastal area (Otani et al. 1999). It has a moderate to weak shade tolerance and subsists in soils ranging from infertile sands to clays (Ishida 1990). Growth is better in soils that are weakly alkaline, but this species tolerates acidic soils as well (Ishida 1990). The hard seed easily germinates in the manure of ruminant animals (Takahashi et al. 1995). Z. matrella is morphologically intermediate between Z. japonica and Z. tenuifolia and is distributed in the southeastern part of Honshu Island, southern region of South Korea, and China. It is a low-growing, rhizomatous perennial with culms up to 5-cm tall. Leaf blades are glabrous, up to 7-cm long and <3-mm wide, and involute. Racemes are 1–4-cm long and have 10 to 40 spikelets. The spikelets are 2.1–3.2-mm long, each with one floret that is smaller than but quite similar to that of Z. japonica. First glumes are usually absent. Second glumes have indurate bases, and basal margins are connate and awnless. Lemmas are 1.6–2.5-mm long, and the

paleas are usually absent (Osada 1993; Engelke and Anderson 2003). Z. tenuifolia is a low-growing rhizomatous perennial found in subtropical and tropical regions and is widely distributed in Japan (southern Kyushu and Okinawa Islands), Taiwan, southern China, Indonesia, Micronesia, New Guinea, Madagascar, and India. It is adapted to coastal zones buffeted by high winds and waves that deliver substantial amounts of salt. Leaf blades are completely involute and acerate. Racemes have fewer than 12 spikelets (Osada 1993; Engelke and Anderson 2003). Z. macrostachya Franch. et Savat. is a creeping perennial that grows only on sandy soil along the seashore. Culms are tufted, erect on the nodes of creeping rhizomes, and 15–40-cm tall. Leaf blades are 3–8-cm long and 2–5-mm wide, involute when dry, and spiny at the apex. Racemes are dense and spike-like, 3–4-cm long, and 5–7-mm wide. The lower half of each raceme is enclosed by the uppermost leaf sheath. Spikelets are 6–8-mm long and 2–4-mm wide (Osada 1993). The rhizome is too prostrate to create a mat. Therefore, no commercial use has been reported. The species is distributed widely from the Ryukyus to Honshu, Hokkaido, and southern region of South Korea, but the natural habitats are very narrow strips

16 Zoysia

along the seashore. Natural hybrids with Z. japonica are found. Z. macrostachya has a unique trait of mature seeds that can be peeled easily. Several attempts have been made to introduce this trait to construct a new cultivar of seed-propagating zoysia grass. This trait appears to be controlled by very complex quantitative trait loci (QTL), however, no significant segregation occurs in F2 progeny. Therefore, breeding efforts for constructing seed-propagating zoysia grass are still underway. Zoysia grass collection and release in the US were described in detail by Engelke and Anderson (2003). The first organized collection expedition supported by the US Department of Agriculture (USDA) was undertaken in 1982 by Murray and Engelke. From the Pacific Rim countries of Japan, Korea, Taiwan, and the Philippines, approximately 700 accessions from five of Zoysia species were collected. The original collection is maintained as vegetative propagules in greenhouses at Texas A&M University (Engelke and Anderson 2003). The evaluation of the original accessions and continuous breeding efforts have been maintained to improve zoysia grass, particularly the traits of seed propagation, turf density, drought resistance, salinity tolerance, cold hardiness, and insect resistance (Engelke and Anderson 2003). With the aim of identifying zoysia grasses well suited for low-input sustainable grazing, two collections of the Japanese ecotypes have been performed. The first collection occurred on all the Japanese islands in the 1980s and included all five of the species native to Japan. Materials from this collection were evaluated at the Japanese National Agricultural Research Center for the Tohoku Region and conserved in the Japanese National Livestock Breeding Center (Fukuoka 2000; Fukuoka et al. 2009). Several ecotypes were selected for good turf quality, and one of them has been released as “Tsukuba Green.” The second collection was organized by the Japanese Ministry of Agriculture, Forestry and Fisheries and focused intensively on grasses for low-input sustainable grazing. More than 1,200 accessions of Z. japonica were collected and evaluated (Ebina et al. 2000a), and these are maintained at the Japanese National Institute of Livestock and Grassland Science, supported by the Genebank of the National Institute of Agrobiological Science in Japan. Z. japonica “Asagake” was selected for its ability to rapidly establish grazing fields, and “Asamoe”

299

was selected for its ability to form a dense turf and the rapid establishment of turf and grazing fields.

16.2 Salt Tolerance Mechanisms and Phylogenetic Relationships In their natural habitats, zoysia grasses (except Z. japonica) have strong salt tolerance. Because they have a low canopy but moderate to weak shade tolerance, in the humid Pacific Rim zoysia grasses prefer costal habitats, where no other plant can survive. The salt tolerance among the coastal species of zoysia grass is associated with shoot saline ion exclusion via leaf salt glands and the overall density of leaf salt glands (Marcum et al. 1998). In addition to these morphological characteristics for salt tolerance, Oishi and Ebina (2005) reported biochemical salt protection from the osmotic balancers glycine betaine and polyamine. The gene coding for the enzyme involved in the synthesis of glycine betaine (BADH) was isolated and the enzymatic properties were determined. The BADH enzyme from the salt-tolerant species Z. tenuifolia has bifunctional enzyme activities: it functions as betaine aldehyde dehydrogenase to accumulate betaine aldehyde and as aminoaldehyde dehydrogenase to accumulate polyamines. The betaine aldehyde is a comparatively small molecule that maintains osmotic balance (Nuccio et al. 1999), and the polyamine has a huge molecular weight and acts as an osmoprotectant (Trossat et al. 1997). In addition to the elucidation of the salt-tolerance mechanisms, isozyme patterns and genomewide polymorphisms also have been studied to estimate phylogenetic relatedness and trace the evolutionary path of zoysia grass (Table 16.1). Yaneshita et al. (1997) used the restriction fragment length polymorphism (RFLP) markers to estimate the genetic variation among five species of Zoysia collected from natural populations in Japan by analyzing nuclear and chloroplast DNA. Based on the analysis of genetic variation by RFLP of ribosomal and chloroplast DNA of five Zoysia species, Akiyoshi et al. (1998) reported that salt tolerance of zoysia grass originated from the nuclear DNA of Z. matrella and was not maternally inherited. Amplified fragment length polymorphism (AFLP) variation in zoysia grass also has been investigated, and genetic diversity and its geographical variation among natural populations have been characterized using AFLP

300

S. Tsuruta et al.

Table 16.1 Molecular marker-based isozyme and DNA profiling studies in zoysia grasses Target Marker system No. of markers No of accession Species Protein Isozyme 1 24 Zj Isozyme 2 182 Zm, Zs Isozyme 2 131 Zm, Zt, Zs Nuclear RFLP 20 17 Zj, Zm, Zt, Zs, Zmac RAPD 12 131 Zm, Zt, Zs AFLP 4 46 Zj, Zm, Zt, Zs, Zmac AFLP 10 20 Zj SSR 12 24 Zj, Zm, Zt, Zs, Zmac SSR 12 41 Zj, Zm, Zt SSR 30 20 Zj Chloroplast RFLP 10 17 Zj, Zm, Zt, Zs, Zmac Sequence 6 3 Zj, Zm, Zt SSR 21 3 Zj, Zm, Zt

References Yamada and Fukuoka (1984) Weng (2002) Weng et al. (2007) Yaneshita et al. (1993, 1997) Weng et al. (2007) Ebina et al. (2000a) Hong et al. (2008) Tsuruta et al. (2005) Hashiguchi et al. (2007) Ma et al. (2007) Yaneshita et al. (1997) Tsuruta et al. (2008) Tsuruta et al. (2008)

Species: Zj: Z. japonica, Zm: Z. matrella, Zt: Z. tenuifolia, Zs: Z. sinica, Zmac: Z. macrostachya

markers (Fig. 16.2). Recently, random amplified polymorphic DNA (RAPD) analysis (Weng et al. 2007), simple sequence repeat (SSR) or microsatellite analysis (Tsuruta et al. 2005; Ma et al. 2007), and sequence variation of chloroplast DNA (Tsuruta et al. 2008) were applied to clarify the phylogenetic relatedness among Zoysia species. However, the phylogeny and taxonomy of this genus remain unclear, and leaf width, traditionally used as an indicator for species discrimination, shows continuous variation among species, making species identification difficult. More recently, microsatellite markers have been developed based on chloroplast DNA sequences of Zoysia (Tsuruta et al. 2008), and phylogenetic clarification utilizing these molecular markers from the organellar genome is now in progress.

16.3 Classical and Molecular Genetic Studies Among the members of the genus Zoysia, Z. japonica, Z. matrella, and Z. tenuifolia are used commonly in parks and as sports turf (Ishida 1990; Engelke and Anderson 2003). Also, in the coastal and cultivated volcanic mountainous areas of Japan, Zoysia species have been used successfully for livestock grazing (Ishida 1990; Akamine et al. 2005). The natural populations have different distributions, reflecting the genus’ environmental adaptability. Z. japonica is adapted to warm climates and is more resistant to low

temperatures than the other species. Z. matrella and Z. tenuifolia are native to coastal areas in tropical and subtropical regions and have poor cold tolerance, but their tolerance of drought, heat, and salinity is excellent. The three species of Zoysia are allotetraploids, with a chromosome number of 2n ¼ 4x ¼ 40, and they can generate interspecific hybrids in all combinations (Forbs 1952; Arumuganathan et al. 1999). The species are also protogynous hermaphroditic plants with both a pistil and three stamens in a single flower. The pistil matures 7–10 days earlier than stamens, and the green or yellow lemma turns dark brown at the tip as the pistil matures (Honda and Kono 1963). The difference in maturation stage between the pistil and stamens promotes outcrossing, although plants can be artificially self-pollinated at the stage when both the pistil and stamens are mature because of the lack of physiological self-incompatibility mechanisms (Asano and Aoki 1998). 1. Interspecific hybrids of Zoysia have proven to be highly effective in the genetic analysis of commercially important traits, such as environmental adaptability. Genetic linkage maps of Zoysia have been constructed on the basis of RFLP (Yaneshita et al. 1999), AFLP (Ebina et al. 1999; Cai et al. 2004), and microsatellite markers (Cai et al. 2005) using the progeny obtained from interspecific hybrids of Z. japonica and Z. matrella. Yaneshita et al. (1999) and Cai et al. (2004, 2005) used the self-pollinated progenies of the ecotype or F1 clone of Z. japonica  Z. matrella hybrids as the mapping populations, whereas Ebina et al. (1999) employed F1 progenies

16 Zoysia

301 SIMILARITY

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1.0

16 26 24 25 17 21 19 18 20 48 45 46 47 2 15 3 10 11 6 5 1 8 9 4 12 13 36 38 35 37 14 43 44 41 40 39 7 31 33 34 32 42 22 23 30 29 28 27

Z. macrantha

I

Z. sinica var . sinica

II

Z. macrostachya

Z. japonica

Z. matrella

III

IV

V

Z. matrella

Z. tenuifolia

VI

Z. matrella

Z. sinica var. nipponica

Z. matrella Z. tenuifolia VII Z. sinica var. nipponica

Z. tenuifolia

Fig. 16.2 Phenogram of the 46 accessions including five species of Zoysia based on the unweighted pair-group method average using 206 bands obtained with four AFLP markers

produced from crossing between Z. japonica and Z. matrella ecotypes (Fig. 16.3). In the linkage analysis using the F1 population, 267 polymorphic bands from the pollen parent Z. matrella were observed by using 30 AFLP primer combinations, of which 252 bands (94.4%) were segregated at a 1:1 ratio (chisquare test, P >0.01). In contrast, 41 polymorphic

bands were obtained from the maternal parent Z. japonica, and most of them did not segregate in the F1 population, indicating that the maternal ecotype used in this study had higher homozygosity compared to the pollen parent Z. matrella. This accession of Z. japonica was collected from an extremely isolated, high-elevation natural population in

302

S. Tsuruta et al.

Fig. 16.3 The parents of the F1 progeny. Accession P29 is an ecotype that is typical of Z. japonica but lacks anthocyanin in the runner. Accession P272 is a Z. matrella ecotype with a narrow leaf

Japan, suggesting that the high homozygosity in this accession was associated with geographical factors. Using 252 markers derived from 30 primer combinations, Ebina et al. (1999) successfully constructed the AFLP linkage map of zoysia grass with 32 linkage groups covering a total of 1,600 cM, with the average distance between markers being 6.4 cM. Furthermore, the F2 population, consisting of 93 individuals and produced by selfing the individuals in the F1 segregating population, was analyzed using 159 AFLP markers derived from 24 primer combinations. The linkage map consisted of 27 linkage groups with a total map distance of 1,418 cM and an average distance of 8.9 cM between markers. Segregation of the main characteristics of zoysia grass was observed in the F1 and F2 populations (Ebina et al. 2000b). Pigmentation in the stolon from anthocyanin, which was absent in the maternal parent Z. japonica, segregated at a 3:1 ratio for the AFLP markers, indicating that the trait is determined by a single major dominant gene. The anthocyanin locus was found to be closely linked to AFLP markers B4-182 and E5-42 (Fig. 16.4). Because Zoysia species are vegetatively propagated by stolons and rhizomes, F2 seeds should be easily obtained by self-fertilization of propagated F1 clones. Therefore, genetic approaches using large-scale mapping populations are available to identify and isolate the major genes governing important traits in zoysia grass. Recently, the availability of the inclusion of anthocyanin in the abovementioned F1 and F2 populations has stimulated an interest in the study of their potential as functional

ingredients. However, the characterization and identification of major genes governing such qualitative traits can also be accomplished in Zoysia.

16.4 Crop Improvement Using Traditional and Advanced Tools Zoysia grasses have been used as economically and environmentally desirable landscape and forage grasses. In Japan, the use of zoysia grass was recorded in the Sakuteiki, a Japanese gardening book published in 1156, and its commercial use began in the 1700s (Kitamura 1970). However, the selection and breeding of zoysia grass is a relatively recent phenomenon. Zoysia grass was introduced to the US from Japan in 1902 (Meyer and Funk 1989). Z. japonica “Meyer” was the first variety developed jointly by the USDA and the US Golf Association. “Meyer” was originally selected from a population of plants grown from seed in 1941 and was released in 1951 (Grau and Radko 1951), and it is the most widely used zoysia grass cultivar in the US. In 1955 “Emerald,” selected from an interspecific hybrid between Z. japonica and Z. tenuifolia, was released by the USDA and the Georgia Agricultural Experiment Station (Forbs 1962). It combines the fine texture of Z. tenuifolia with the cold tolerance and faster rate of spread of Z. japonica. Subsequently, some commercial cultivars of zoysia grass such as “Midwest” (released in 1963), “El-Toro” (released in 1986), and “Belair” (released

16 Zoysia

303

Dis cM

Marker name

Distance from ANTC locus

Group 1 B1-112 28.2 A4-92 11.3 1.2 1.7 5.5 10.6 6.1

A4-80 A8-96 E2-105 D6-65 A7-65 A2-341

12.8 5.3

D2-151 D1-236

9.3 C4-172

75.4

E5-42

12.9 cM

ANTC B4-182 B4-157

0.0 cM 5.2 cM

12.9 5.4 4.2

38.8

E4-139

Fig. 16.4 Linkage group 1, which contains the anthocyanin locus (ANTC) and tightly linked AFLP markers

in 1987) have been selected from collected genetic resources and populations of plants grown from seed. Additional cultivars based on zoysia grass collections have been developed and released since the 1990s (Asano and Aoki 1998). By the 1980s, the breeding of zoysia grass had started in Japan as well. “Miyako,” a naturally interspecific hybrid between Z. japonica and Z. matrella that was originally selected, was the first cultivar registered in Japan (Asano and Aoki

1998). Z. matrella “Winter Carpet” and “Winter Field” were released by Nagatomi et al. (1993, 1998) in 1995 and 1996. These cultivars were first produced by gamma-irradiation mutagenesis in the sod; they exhibit a completely distinct character: retaining green leaves under the same seasonal conditions in which other varieties of Z. matrella lose their green leaves. Thus far, 41 zoysia grass varieties have been released in Japan. Most cultivated zoysia grasses have been improved through conventional breeding methods such as clonal selection, hybridization, and mutagenesis. Because zoysia grasses show great natural variation, these will likely be the main methods used for the development of new varieties of zoysia grass. As noted above, DNA-based molecular markers such as RFLPs, AFLPs, and SSRs have been developed and are being exploited increasingly in the construction of genetic linkage maps of zoysia grass. In the near future, breeders may be assisted by DNA analysis, which confirms the degree of genetic variation in zoysia grass. Ebina et al. (2000b) have conducted a genetic linkage analysis of several important traits such as leaf width, salt tolerance, and freezing hardiness using F1 and F2 populations derived from an interspecific hybrid of Z. japonica and Z. matrella. Leaf width is used to discriminate among Zoysia species and is one of the most visible determinants of turf quality (Kitamura 1970; Turgeon 1996). According to QTL analysis of several morphological traits, leaf width is controlled by two major QTLs and one minor QTL. The major QTLs, which are located in linkage groups 7 and 15, have significant effects, and the region around the D2-443, AFLP marker has an especially large effect (Fig. 16.5a; Table 16.2). These results suggest that the genetic factors controlling leaf width in zoysia grass might be relatively simple, although most morphological traits for turf species show continuous phenotypic variation and are controlled by QTLs. In addition, a linkage peak at a similar position to that for leaf width has been identified with leaf length, runner length, and the length of the stem at heading, suggesting that these traits are at least partially controlled by the same QTL. Although salt tolerance segregated in the F1 population, QTLs with LOD scores higher than 3.0 were not detected (Ebina et al. 2000b). Two QTLs with LOD scores greater than 2.0 were detected in linkage groups 15 and 28, respectively (Fig. 16.5b, Table 16.2). In addition, a peak identified with leaf width is at a

304

S. Tsuruta et al.

Fig. 16.5 LOD scores from the QTL analysis of leaf width (a) and salt tolerance (b) of the F1 population derived from the cross between Z. japonica and Z. matrella

Table 16.2 Location of QTLs affecting leaf width and salt tolerance Character Peak marker Locus LOD Contribution ratio Leaf width D2-433 7 3.55 0.266 E5-97 7 3.37 0.246 D6-278 7 3.15 0.236 B6-256 15 2.62 0.208 A3-165 7 2.62 0.201 A3-59 7 2.62 0.201 Salt tolerance A6-383 15 –2.08 –0.163 H2-158 28 –2.04 –0.160 B6-256 15 –1.78 –0.146 B3-202 28 –1.33 –0.104 D3-172 15 –1.23 –0.098 B5-202 28 –1.19 –0.094

similar position in linkage group 15 as a peak for salinity tolerance. In a study of barley, Mano and Takeda (1997) reported that the QTLs for the most effective abscisic acid response, which is correlated with leaf growth, were located very close to those for salt tolerance. Therefore, genetic linkage between salinity tolerance and leaf width may be a characteristic common among members of the Poaceae. This finding should assist in the elucidation of salt tolerance mechanisms in Zoysia and related genera. Zoysia grass is one of the most tolerant of freezing among the C4 turf grasses. Freezing tolerance is often evaluated by measuring electrolyte leakage or regrowth of plant tissues after freezing. A QTL for freezing tolerance of zoysia grass was detected in linkage group 25 in the F2 population (Table 16.3),

16 Zoysia Table 16.3 Location of QTLs affecting freezing tolerance Peak marker Locus LOD Contribution ratio B1-227 25 –6.49 –0.294 E2-177 6 –1.64 –0.084 E1-478 24 –1.24 –0.065 B4-124 18 1.18 0.060

but no significant QTL was detected in the F1 population (Ebina et al. 2000b). This result suggests that freezing tolerance in zoysia grass is conferred by a dominant gene, which was homozygous in the female parent, Z. japonica. The QTL was located in the region of B1-227 in linkage group 25, with a 0.294 contribution rate. In zoysia grass, morphological characters that can be assessed visually are important factors in determining turf quality, and QTLs for tolerance of environmental stressors such as salinity and freezing are genetic elements that may be widely distributed among species. Zoysia grass is widely used in many situations, such as sports fields and golf courses, as well as for livestock grazing. Information on the association between genetic markers and important traits will be available for the diversification and advancement of Zoysia varieties in the near future. Transgenic approaches to genetic manipulation offer the opportunity to generate unique genetic variation. Plant regeneration systems for zoysia grass have been established from protoplasts (Asano 1989; Khayri et al. 1989; Inokuma et al. 1996) and embryogenic calluses (Bae et al. 2001). Recently, transgenic zoysia grass plants were produced by polyethylene glycolmediated (Inokuma et al. 1998) and Agrobacteriummediated transformation (Toyama et al. 2003; Ge et al. 2006; Zhang et al. 2007). Rahman et al. (2003) established a plant regeneration system from calluses derived from the apical meristems of seedlings and successfully produced transgenic zoysia grass plants using Agrobacterium-mediated transformation of embryogenic calluses (Fig. 16.6). In addition, some useful genes have been introduced into zoysia grasses by Agrobacterium-mediated transformation to improve their resistance to a variety of biotic and abiotic stressors (Li et al. 2006; Zhang et al. 2007). The transgenic plants can be grown by vegetative propagation of stolons and rhizomes. Although there is some concern that large-scale releases of transgenic plants may cause serious ecological and environmental

305

problems, genetic engineering represents a new avenue for the rapid production of unique varieties of Zoysia.

16.5 Controlling Weeds in Zoysia Grass Weed invasion is a potentially great problem when establishing and maintaining turf. Because zoysia grass is slow to propagate and establish, early weed control is especially important. Weed control methods for turf grasses, including zoysia grass, can be divided into two categories: cultural and chemical. Cultural management of weeds in zoysia grass includes the use of mowing, fertilization, and irrigation (Turgeon 1996). In addition, varieties of zoysia grass that are genetically resistant to biotic stress can be utilized for weed control. Busey (1982) reported that genotypes of zoysia grass with lower densities of stinging nematode infestation also had less abundant weed populations. Since the discovery of the selective toxicity of synthetic organic herbicides such as 2,4-dichlorophenoxyacetic acid (Marth and Mitchell 1944), there has been considerable research on chemical management of weeds in turf grass. Herbicides are applied prior or subsequent to the emergence of target weed species; these are called pre- and post-emergence herbicides, respectively. Although both pre- and post-emergence herbicides effectively control weeds in zoysia grass (Johnson 1996a, b), the use of pre-emergence herbicides is time consuming and expensive and often affects the ability to reestablish zoysia grass. Consequently, post-emergence herbicides are often used to remove weeds. Post-emergence herbicides are of two types: non-selective and selective. A nonselective herbicide provides control of most plants, whereas a selective herbicide is designed to control a specific type of plant, usually annual grasses or broadleaf weeds. Currently, there are no herbicides with adequate selectivity for controlling most perennial grasses. Therefore, non-selective herbicides must be used for controlling these weeds (Turgeon 1996). Herbicide-tolerant zoysia grass also has been developed by Agrobacterium-mediated transformation (Toyama et al. 2003). All Zoysia species are sexually compatible, and thus it is essential to consider the environmental impact of using transgenic Zoysia. In confined and unconfined test fields, Bae et al. (2008)

306

Fig. 16.6 In vitro apical meristematic culture and genetic transformation of Z. japonica mediated by Agrobacterium tumefaciens. (a) Highly regenerative embryogenic callus induced from seed in Z. japonica; (b) initiation of plant regeneration; (c) established regenerated plant from callus; (d) embryogenic

S. Tsuruta et al.

callus induced from apical meristems of regenerated plants via seed-derived callus; (e) plants regenerated from callus; (f, g) transient GUS expression on hygromycin-resistant callus after infection with A. tumefaciens; GUS expression in leaves of transgenic (h) and control plant (i)

16 Zoysia

assessed the potential environmental and biodiversity effects by investigating major traits equivalence, the ability to cross-pollinate, and gene flow between transgenic zoysia grasses containing a herbicide resistance gene and wild type grasses. The transgenic zoysia grass showed tolerance to Basta, a broad-spectrum glufosinate-based herbicide, whereas the wild type grasses stopped growing and died. Bae et al. (2008) concluded that the transgenic zoysia grass would not pose a significant risk when cultivated outside of the test field, because unintended cross-pollination with and gene flow from transgenic zoysia grass were not detected in the neighboring weed species they examined, but these were observed in wild type zoysia grass. Weeds in zoysia grass could be effectively managed by utilizing such an herbicide resistant transgenic plant.

16.6 Recommendations for Future Actions Germplasm is the basic material for plant breeding, and the most important foundation for a breeding program is an extensive collection of genetically diverse germplasm. Characterizing and recording all the details of the plant material in a collection is also important to its subsequent management and use. Thus, a germplasm collection is becoming increasingly important in the breeding and all phases of biological research of zoysia grasses. However, zoysia grass ecotypes are gradually disappearing due to changes in their natural habitats. In New Zealand, for example, habitats of endemic species including Zoysia are being lost as human activities expand and invasive foreign species encroach (Esler 1991). Fukuoka (2000) reported that populations of rare species in Japan, such as Z. sinica and Z. macrostachya, are declining at a rapid rate due to development and erosion of coastal areas. These Zoysia species are potential genetic resources for improving resistance to environmental stressors and seed propagation (Fukuoka 2000; Fukuoka et al. 2009). Every effort should be made to maintain populations of these native plants that are useful breeding materials through habitat preservation and plant collection.

307

In addition to maintaining the genetic resources of Zoysia, a robust means of varietal identification is also required for the effective management of collected genetic resources and the assessment of turf quality. Traditional varietal identification of zoysia grass has been performed primarily by examining morphological characteristics; however, this is often difficult with closely related cultivars due to ambiguous differences and phenotypic modifications caused by environmental factors (Yaneshita et al. 1997). In addition, the ease with which zoysia grass propagates vegetatively adds another burden with regard to variety protection. Consequently, the enforcement of plant variety protection regulations is difficult, and an efficient and effective method is necessary for the enforcement of plant breeders’ rights. Because zoysia grass is primarily propagated by stolons and rhizomes, there should be no genotypic differences within a given cultivar. Therefore, the development of molecular markers would be a valuable tool for plant protection as well as a vehicle for facilitating breeding improvements and estimating genetic relatedness among various zoysia grass genotypes. Several DNA-based molecular markers, including RFLP (Yaneshita et al. 1997), RAPD (Weng et al. 2007), AFLP (Ebina et al. 2000b), and microsatellites (Tsuruta et al. 2005; Ma et al. 2007), have been developed for the estimation of genetic variation in zoysia grass. These markers can be useful for evaluating phylogenetic relatedness objectively, performing genetic analyses of commercially important traits, promoting the development of new varieties, and ensuring the protection of zoysia grass.

References Akamine H, Kawamoto Y, Ishimine Y, Kuramoti H, Ichizen N (2005) Morphological characteristics of Zoysia tenuifolia Willd. J Jpn Soc Turfgrass Sci 33:122–126 (In Japanese with English summary) Akiyoshi M, Yaneshita M, Nagasawa R, Endo N (1998) Sea water tolerance of zoysiagrass in relation to morphological and genetic classification. Grassl Sci 44:7–13 Arumuganathan K, Tallury SP, Fraser ML, Bruneau AH, Qu R (1999) Nuclear DNA content of thirteen turfgrass species by flow cytometry. Crop Sci 39:1518–1521 Asano Y (1989) Somatic embryogenesis and protoplast culture in Japanese lawngrass (Zoysia japonica). Plant Cell Rep 8:141–143

308 Asano Y, Aoki K (eds) (1998) Turfgrasses and the cultivars. Soft Science, Tokyo (in Japanese) Bae CH, Toyama K, Lee SC, Lim YP, Kim HI, Song PS, Lee HY (2001) Efficient plant regeneration using mature seed-derived callus in zoysiagrass (Zoysia japonica Steud.). Kor J Plant Tissue Cult 28:61–67 Bae TW, Vanjildorj E, Song SY, Nishiguchi S, Yang SS, Song IJ, Chandrasekhar T, Kang TW, Kim JI, Koh YJ, Park SY, Lee J, Lee YE, Ryu KH, Riu KZ, Song PS, Lee HY (2008) Environmental risk assessment of genetically engineered herbicidetolerant Zoysia japonica. J Environ Qual 37:207–218 Busey P (1982) Cultural management of weeds in turfgrass: a review. Crop Sci 43:1899–1911 Cai H, Inoue M, Yuyama N, Nakayama S (2004) An AFLPbased linkage map of zoysiagrass (Zoysia japonica). Plant Breed 123:543–548 Cai H, Inoue M, Yuyama N, Takahashi W, Hirata M, Sasaki T (2005) Isolation, characterization and mapping of simple sequence repeat markers in zoysiagrass (Zoysia spp.). Theor Appl Genet 112:158–166 Clayton WD, Renvoize SA (1986) Genera Graminum, Grasses of the World. Her Majesty’s Stationary Office and Royal Botanic Gardens, Kew, London Ebina M, Kobayashi M, Kasuga S, Araya H, Nakagawa H (1999) An AFLP-based genome map of zoysiagrass. In: Plant animal genome VII conference, San Diego, CA, USA, pp 17–21 Ebina M, Abe A, Kobayashi M, Kasuga S, Araya H, Nakagawa H (2000a) Phylogenetic analysis of genus Zoysia for improvement of indigenous grazing grassland. In: Proceedings of the international workshop integration of biodiversity and genome technology for crop improvement, Tsukuba, Japan, pp 133–134 Ebina M, Kobayashi M, Muraki M, Kikawada T, Araya H, Nakagawa H (2000b) Molecular mapping of zoysiagrass for some QTL analysis. In: Plant animal genome VIII conference, San Diego, CA, USA, pp 204 Engelke MC, Anderson S (2003) Zoysiagrass (Zoysia spp.). In: Casler MD, Duncan RR (eds) Turfgrass biology and breeding. Wiley, Hoboken, NJ, pp 271–285 Esler AE (1991) Changes in the native plant cover of urban Auckland, New Zealand. NZ J Bot 29:177–196 Forbs IJ (1952) Chromosome numbers and hybrids in Zoysia. Agron J 44:147–151 Forbs IJ (1962) Registration of emerald zoysiagrass. Crop Sci 2:533–534 Fukuoka H (2000) Breeding of zoysia grass. 1. Collection of genetic resources and general view of their characteristics. J Jpn Soc Turfgrass Sci 29:11–21 (in Japanese with English summary) Fukuoka H, Murata T, Shibata K, Shinodad K, Takahashi Y (2009) Breeding of zoysia grass. 2. A breeding method for seeded type variety. J Jpn Soc Turfgrass Sci 37:91–97 (in Japanese with English summary) Ge Y, Norton T, Wang ZY (2006) Transgenic Zoysia (Zoysia japonica) plants obtained by Agrobacterium-mediated transformation. Plant Cell Rep 25:792–798 Grau FV, Radko AM (1951) Meyer (Z-52) zoysia. USGA J Turf Manag 4:30–31 Hashiguchi M, Tsuruta S, Matsuo T, Ebina M, Kobayashi M, Akamine H, Akashi R (2007) Analysis of genetic resources

S. Tsuruta et al. in Zoysia spp. 2. Evaluation of genetic diversity in zoysiagrass indigenous to southwest islands of Japan based on simple sequence repeat markers. Jpn J Grassl Sci 53:133–137 (in Japanese with English summary) Hatch SL, White RH (2004) Additional C4 turf and forage grasses. In: Moser LE, Burson BL, Sollenberger LE (eds) Warm-season (C4) grasses, vol 45, Agronomy. American Society of Agronomy, Madison, WI, USA, pp 1081–1119 Honda H, Kono M (1963) Morphological and anatomical studies of the lawn grasses with special reference to the Japanese lawn grass, Zoysia japonica Steud. Tech Bull Fac Hort Chiba Univ 11:1–22 (in Japanese with English summary) Hong J, Liebao H, Zhang Y (2008) AFLP analysis on genetic diversity of Zoysia japonica. Acta Hort 783:265–272 Inokuma C, Sugiura K, Cho C, Okawara R, Kaneko S (1996) Plant regeneration from protoplasts of Japanese lawngrass. Plant Cell Rep 15:737–741 Inokuma C, Sugiura K, Imaizumi N, Cho C (1998) Transgenic Japanese lawngrass (Zoysia japonica Steud.) plants regenerated from protoplasts. Plant Cell Rep 17:334–338 Ishida R (1990) General remarks on the research works of Japanese lawn grass (Zoysia japonica Steud.) and zoysia type grasslands in Japan. J Jpn Grassl Sci 36:210–217 (in Japanese with English summary) Johnson BJ (1996a) Reduced rates of preemergence and postemergence herbicides for large crabgrass (Digitaria sanguinalis) and goosegrass (Eleusine indica) control in bermudagrass (Cynodon dactylon). Weed Sci 44:585–590 Johnson BJ (1996b) Effect of reduced dithiopyr and prodiamine rates on large crabgrass (Digitaria sanguinalis) control in common bermudagrass (Cynodon dactylon) and tall fescue (Festuca arundinacea) turf. Weed Technol 10:322–326 Khayri JM, Huang FH, Thompson LF, King JW (1989) Plant regeneration of zoysiagrass from embryo-derived callus. Crop Sci 29:1324–1325 Kitamura F (1970) Studies on the horticultural classification and development of Japanese lawn grasses. Bull Kemigawa Arboretum Fac Agric Univ Tokyo 3:1–60 (in Japanese) Li RF, Wei JH, Wang HZ, He J, Sun ZY (2006) Development of highly regenerable callus lines and Agrobacterium-mediated transformation of Chinese lawngrass (Zoysia sinica Hance) with a cold inducible transcription factor, CBF1. Plant Cell Tissue Org Cult 85:297–305 Ma KH, Jang DH, Dixit A, Chung JW, Lee SY, Lee JR, Kang HK, Kim SM, Park YJ (2007) Characterization of 30 new microsatellite markers, developed from enriched genomic DNA library of zoysiagrass, Zoysia japonica Steud. Mol Ecol Notes 7:1323–1325 Mano Y, Takeda K (1997) Mapping quantitative trait loci for salt tolerance at germination and the seeding stage in barley (Hordeum vulgare L.). Euphytica 94:263–272 Marcum KB, Anderson SJ, Engelke MC (1998) Salt gland ion secretion: a salinity tolerance mechanism among five zoysiagrass species. Crop Sci 38:806–810 Marth PC, Mitchell JW (1944) 2, 4-dichlorophenoxyacetic acid as a differential herbicide. Bot Gaz 106:224–232 Meyer WA, Funk CR (1989) Progress and benefits to humanity from breeding cool-season grasses for turf. In: Sleeper DA, Asay KA (eds) Contribution from breeding forage and turf grasses. Crop Science Society of America, Madison, WI, USA, pp 31–48

16 Zoysia Nagatomi S, Mitsui K, Miyahara K (1993) Selection of evergreen mutant variety in Manila grass (Zoysia matrella Merr.). Institute of Radiation Breeding Technique News, No 44 Nagatomi S, Mitsui K, Miyahara K, Nakagawa K, Yamagishi T (1998) A new variety of Manila grass, ‘winter field’: frostresistant and dwarf mutant. Institute of Radiation Breeding Technique News, No 63 Nuccio ML, Rhodes D, McNeil SD, Hanson AD (1999) Metabolic engineering of plants for osmotic stress resistance. Curr Opin Plant Biol 2:128–134 Ogura S, Kosako T, Hayashi Y, Dohi H (2001) In sacco ruminal degradation characteristics of chemical components in fresh Zoysia japonica and Miscanthus sinensis growing in Japanese native pasture. Asian-Australas J Anim Sci 14:41–47 Oishi H, Ebina M (2005) Isolation of cDNA and enzymatic properties of betaine aldehyde dehydrogenase from Zoysia tenuifolia. J Plant Physiol 162:1077–1086 Osada T (1993) Illustrated grasses of Japan. Heibonsha, Tokyo, Japan Otani I, Yamamoto N, Entsu S (1999) Effect of spring-sowing of herbage species on the establishment of Zoysia japonica Steud. in the coast of the Japan Sea in Chugoku districts. Grassl Sci 45:257–263 Rahman SML, Mackay WA, Ebina M, Nakagawa H, Mesbahuddin ASM, Quebedeaux B (2003) Transient gene expression in Zoysia japonica using Agrobacterium tumefaciens. Subtropic Plant Sci 55:11–17 Takahashi T, Saito S, Otani I, Hagino K (1995) Spreading of the zoysiagrass seeds by manure of ruminant animals in zoysia type grazing field. J Jpn Grassl Sci 41(suppl):15–16 (in Japanese) Tamate HB, Tatsuzawa S, Suda K, Izawa M, Doi T, Sunagawa K, Miyahira F, Tado H (1998) Mitocondrial DNA variations in local populations of the Japanese Sika Deer, Cervus nippon. J Mammal 79:1396–1403 Toyama K, Bae C-H, Kang J-G, Lim Y-P, Adachi T, Rui K-Z, Song P-S, Lee H-Y (2003) Production of herbicide-tolerant zoysiagrass by Agrobacterium-mediated transformation. Mol Cell 16:19–27 Trossat C, Rathinasabapathi B, Hanson AD (1997) Transgenically expressed betaine aldehyde dehydrogenase efficiently

309 catalyzes oxidation of dimethylsulfoniopropinaldehyde and omega-aminoaldehydes. Plant Physiol 113:1457–1461 Tsuruta S, Hashiguchi M, Ebina M, Matsuo T, Yamamoto T, Kobayashi M, Takahara M, Nakagawa H, Akashi R (2005) Development and characterization of simple sequence repeat markers in Zoysia japonica Steud. Grassl Sci 51: 249–257 Tsuruta S, Hosaka F, Otabara T, Hashiguchi M, Yamamoto T, Akashi R (2008) Genetic diversity of chloroplast DNA in Zoysia and other warm-season turfgrasses. Grassl Sci 54:151–159 Turgeon AJ (1996) Turfgrass management, 4th edn. PrenticeHall, Upper Saddle River, NJ, USA Weng JH (2002) Genetic variation of Zoysia in Taiwan as analysed by isozyme patterns and salinity tolerance. Plant Prod Sci 5:236–241 Weng JH, Fan MJ, Lin CY, Liu YH, Huang SY (2007) Genetic variation of Zoysia as revealed by random amplified polymorphic DNA (RAPD) and isozyme pattern. Plant Prod Sci 10:80–85 Yamada T, Fukuoka H (1984) Variations in peroxidase isozyme of Japanese lawn grass (Zoysia japonica Steud.) population in Japan. Jpn J Breed 34:431–438 Yamamoto Y, Yagi T, Saito Y, Kirita H (1988) Changes in the species diversity, H’, of Miscanthus-type grassland in relation to vegetation change by grazing. Grassl Sci 44:122–126 Yaneshita M, Ohmura T, Sasakuma T, Ogihara Y (1993) Phylogenetic relationships of turfgrasses as revealed by restriction fragment analysis of chloroplast DNA. Theor Appl Genet 87:129–135 Yaneshita M, Nagasawa R, Engelke MC, Sasakuma T (1997) Genetic variation and interspecific hybridization among natural populations of zoysiagrasses detected by RFLP analyses of chloroplast and nuclear DNA. Genes Genet Syst 72:173–179 Yaneshita M, Kaneko S, Sasakuma T (1999) Allotetraploidy of Zoysia species with 2n¼40 based on RFLP genetic map. Theor Appl Genet 98:751–756 Zhang L, Wu D, Zhang L, Yoyang C (2007) Agrobacteriummediated transformation of Japanese lawngrass (Zoysia japonica Steud.) containing a synthetic cryIA(b) gene from Bacillus thuringiensis. Plant Breed 126:428–432

.

Index

b-Glucuronidase, 8 2C-value, 154 A Abiotic stress, 58 tolerance, 190 Abscisic acid (ABA), 6, 26, 304 Abscission, 239 Accumulation/loss (A/L), 228 Adaptability, 139, 203, 300 Adaptation, 64, 82–83, 155, 267 Adaptive polymorphism, 100 Adaptive radiation, 82 Agamic, 90 complex, 97, 203 Agrobacterium, 7, 26, 105, 159, 169 transformation, 305 Agroforestry, 211 Agronomic traits, 127, 204 Agrostis A. alba, 3 A. canina, 2 A. capillaris, 2, 3 A. gigantea, 3 A. palustris, 2 A. scabra, 4 A. stolonifera, 2, 9 Allergy, 271 Allogamous, 22, 158 Allohexaploid, 223 Allopolyploidization, 123 Allopolyploidy, 3, 98 segmental, 138 Allotetraploid, 19, 22, 23, 114, 144, 209, 225 segmental allotetraplod, 38, 211 Allotetrasomic, 47 Allozyme, 282 Amphidiploid, 25, 157 Amplified fragment length polymorphism (AFLP), 3, 19–20, 42, 66, 125, 140, 157, 168, 183, 201, 237, 287, 299 Anamorph, 128 Ancestral species, 123–124 Androgenesis, 157 Aneuploid, 31, 96, 222 Aneuploidy, 39

Angleton grass, 92 Anguina, 165 Anther culture, 169 Anthocyanin, 302 Apomicts facultative apomict, 97, 99, 102, 204, 212 obligate apomict, 97 Apomictic, 37, 138, 142, 182, 198, 208, 227 gene, 47 clone, 143 pseudogamous obligate apomict, 211 trait, 245 Apomixis, 47, 97, 99, 100, 201, 245 aposporous apomixis, 190, 206 diplosporous apomixis, 138, 187 facultative apomixis, 208 gene, 49 Apomixis-related gene region (ASGR), 184 Apospory, 47, 199, 209 Apospory-specific genomic region (ASGR), 47, 48, 227, 245 Arabidopsis, 9, 245 Aromatic, 96 Asexual revolution, 50 Association mapping, 84 Autoallooctoploid, 19 Autogamous, 205 Autogamy, 280 Autoploidy, 279 Autopolyploid, 123, 138, 201 Autotetraploid, 79, 96, 183, 268 Aveneae, 16 Aveneae tribe, 1 B Backcrossing, 157, 199, 227 Bacterial artificial chromosome (BAC), 157, 170, 288 clone, 47 contig, 185 library, 47 Bahiagrass, 208 Bayesian analysis, 234 B-chromosome, 227 Bentgrass colonial bentgrass, 2 creeping bentgrass, 1, 9

C. Kole (ed.), Wild Crop Relatives: Genomic and Breeding Resources, Millets and Grasses, DOI 10.1007/978-3-642-14255-0, # Springer-Verlag Berlin Heidelberg 2011

311

312 Bentgrass (cont.) highland bentgrass, 3 Idaho bentgrass, 3 redtop bentgrass, 3 velvet bentgrass, 2 Bermudagrass African bermudagrass, 57 coastal bermudagrass, 68 Biodiversity, 147, 248 Bioenergy, 58 Bioethanol, 192 Biofuel, 147 Biolistics, 8, 159 Biomass, 56, 105, 192, 246 Biosystematics, 96 Biotic stress, 42, 305 Biotype, 204 Birdwood grass, 36 Blast disease, 127 BLASTX, 145 Bluegrass gulf bluegrass, 92 king bluegrass, 95 nadi bluegrass, 92 Queensland bluegrass, 95 Bothriochloa, 97 Bridging species, 100 Brome Alaska brome, 25 Arizona brome, 25 California brome, 25 Mountain brome, 25 Russian brome, 18–20 Sitka brome, 25 Bromegrass arctic bromegrass, 21 erect bromegrass, 21 field bromegrass, 23 meadow bromegrass, 20–21 Bromus B. arizonicus, 16 B. arvensis, 23 B. auleticus, 16 B. biebersteinii, 16 B. carinatus, 16, 25 B. catharticus, 15, 16 B. erectus, 16, 21 B. inermis, 15, 16, 20, 26 B. mango, 16 B. marginatus, 16, 25 B. mollis, 16 B. pumpellianus, 21 B. riparius, 16 B. rubens, 16 B. sitchensis, 15, 25 B. valdivianus, 16 B. variegatus, 21 Brunswick grass, 212 Buffelgrass, 47, 135 cloncurry buffelgrass, 35 Bulked segregant analysis (BSA), 236, 271

Index Burrgrass American burrgrass, 36 hillside burrgrass, 33 large burrgrass, 33 C Candidate gene, 271, 286 Capillipedium, 97 Capim-pojuca, 211 Carbon isotope discrimination (CID), 43 C-banding, 22 C3 species, 178 C4 cereal, 275 C4 grasses, 178, 297 C4 photosynthesis, 136, 178, 180 C4 species, 178 cDNA clone, 48 library, 43, 140, 144, 159, 160, 170, 191 microarray, 48 Cenchrus, 31 C. argimonioides, 32 C. biflorus, 32 C. brownii, 33 C. calicalatus, 33 C. ciliaris, 33, 42, 48 C. distichophyllus, 33 C. echinatus, 33 C. elymoides, 34 C. gracillimus, 34 C. incertus, 34 C. longispinus, 34 C. melanostachyus, 34 C. mitis, 35 C. multiflorus, 35 C. myosuroides, 35 C. palmeri, 35 C. pennisetiformis, 35 C. pilosus, 35 C. platyacanthus, 36 C. prieurii, 36 C. robustus, 36 C. setiger, 36 C. somalensis, 36 C. spinifex, 36 C. tribuloides, 36 Center of origin, 208 Centromere, 227 Chiasma, 227 Chilochloa, 258 Chloridoid, 130 Chloridoideae, 113, 135, 297 Chloroplast, 66–67, 103, 168, 197, 264 DNA (cpDNA), 18, 23, 123, 299 gene, 180 genome, 26, 74 Chlorosis, 4 Chromatin bridge, 138 Chromosome deletion, 271 doubling, 96, 138, 211

Index elimination, 210 organization, 221 pairing, 18, 98, 210 rearrangement, 200, 222 substitution, 271 Cladosporium phlei, 271 Clustering, 168 Coelachyrum, 49 Colchicine, 183 Cold hardiness, 299 tolerance, 210 Colinearity, 126, 200 Colonization, 284 Comparative genetic mapping, 126, 159 Comparative mapping, 243 Compilospecies, 90 Conservation, 180, 192, 270 Conserved intron scanning primers (CISP), 145 Cool-season forage, 170 Cool-season grass, 4, 159 Core collection, 62, 270 Corynebacterium, 165 Cross-pollination, 66, 307 Crown rust, 155, 159 Cryopreservation, 63, 270 Cynodon C. aethiopicus, 54 C. arcuatus, 54, 55 C. barberi, 54, 55 C. dactylon, 53, 54 C. incompletus, 54 C. nlemfuensis, 54 C. plectostachyus, 54 C. transvaalensis, 54 C. x magennisii, 54 Cytogenetic diversity, 96 Cytokinin, 6 Cytoplasmic male sterility (CMS), 10, 167, 246 Cytotypes, 65, 181, 183, 206, 258 D Dactylis, 270 D. glomerata, 73 Dallisgrass, 203–205 prostrate dallisgrass, 205 Dead spot, 6 Desmodium, 247 Dichanthium D. affine, 91 D. annulatum, 90, 91 D. aristatum, 92 D. armatum, 92 D. caricosum, 92 D. compressum, 92 D. fecundum, 92 D. foveolatum, 95 D. humilius, 95 D. maccannii, 95 D. mucronulatum, 95 D. oliganthum, 95

313 D. pallidum, 95 D. panchaganiense, 95 D. queenslandicum, 95 D. sericeum, 95 D. setosum, 95 D. tenue, 95 D. tenuiculum, 95 D. woodrowii, 95 Digestibility, 68, 143 Dilatata complex, 202 Diploid, 17, 74 Disomic, 47 DNA amplification fingerprinting (DAF), 65–66 content, 58 Data Bank of Japan (DDBJ), 288 insertion, 264 sequence, 27, 125, 208, 224, 300 Dollar spot, 7 Domestication, 41, 145–146, 230, 238 history, 287 syndrome, 238, 278 Dormancy, 80, 108, 281 Doubled haploid, 169, 230 Downy mildew, 107 Drought, 42, 192 hardy, 211 resistance, 299 stress, 139, 160 tolerance, 59, 103, 139, 158, 246, 282 tolerant, 3 Dry matter (DM), 139 Duodecaploid, 25 Dysploidy, 224–227, 266 E Ecosystem, 147, 155 Ecotype, 97, 155, 177, 299 Edaphic race, 82 Elephant grass, 222 Eleusine, 113 E. africana, 117 E. coracana, 114 E. floccifolia, 118 E. indica, 118 E. intermedia, 118 E. jaegeri, 118 E. kigeziensis, 118–119 E. multiflora, 119–120 E. tristachya, 120 Embryo rescue, 144 sac, 184, 209 Embryogenesis, 169, 210 Embryology, 98 Endophyte, 27, 170, 271 Endophytic, 170–171 fungi, 27, 271 Epichlo¨e, 271 Epigenetic, 189 Eragrostis, 130

314 Eragrostis (cont.) E. aethiopiaca, 137 E. bicolor, 137 E. curvula, 136, 143 E. japonica, 146 E. lehmanniana, 136, 146 E. pilosa, 136 E. scotelliana, 146 E. tef, 138 E. tremula, 146 Ergot disease, 107 Ergovaline, 171 Ethanol, 247 Euploid, 96 European Molecular Biology Laboratory (EMBL), 288 database, 48 Evolution, 75, 221 progressive, 100 Evolutionary center, 65 event, 65 lag, 79 reconstruction, 225 system, 212 Expressed sequence tag (EST), 2, 48, 84, 126, 140, 145, 168 sequence, 187 F Fertility, 99, 121 Fescue meadow fescue, 155 tall fescue, 158 Festuca, 73 Festuca Gene Index (FaGI), 160 Festuca, 153 Festulolium, 157 F1 hybrid, 157 Finger millet, 113 Flavone, 77 Flavonoid, 78 Flow-cytometry, 145, 182 Flowering behavior, 257, 258 Fluorescence in situ hybridization (FISH), 19, 38, 47, 158, 207 Fodder, 145–146 Forage, 56, 73, 160 Foxtail bristly foxtail, 277 giant foxtail, 277 giant green foxtail, 281 millet, 275 F1 population, 301 F2 population, 302 Freezing, 54 hardiness, 303 tolerance, 158 G Gametophytic, 183 Gandhel, 95 Gene

Index differentiation, 168 expression, 169 expression chip, 271 flow, 100, 109, 217, 235, 238, 284 ontology, 145 pool primary, 83, 221, 246, 269, 277 quaternary, 84, 270 secondary, 83, 269, 277 tertiary, 37, 83, 246, 269, 277 wild, 235 restorer gene, 289 technology, 143–144, 158 GeneBank, 48, 49, 77, 83, 124, 128, 155, 171, 270 Genetic background, 143 bottleneck, 284 differentiation, 66, 237 distance, 234 diversity, 19–20, 26–27, 62, 65, 66, 126, 155, 168, 209, 233, 284, 299 drift, 238, 284 engineering, 305 erosion, 61 instability, 157 isolation, 66 linkage map, 141, 201 map, 43, 125, 126, 159, 236, 238, 288 polymorphism, 158 relatedness, 66 screening, 127 similarity, 140 structure, 168, 221, 283 transformation, 8, 105, 143 variation, 66, 79, 140, 155, 158 Genome, 19, 206 donor, 121 evolution, 154 mapping, 26 sequence, 124 sequencing, 288 size, 16, 57–58, 83, 221, 260 Genomic, 126, 238 constitution, 25 DNA, 102 organization, 238 Genomic in situ hybridization (GISH), 19, 121, 158, 245, 266, 278 Genotype matrix mapping (GMM), 190 Genotyping, 160 Germination, 286 Germplasm, 27, 85, 155, 180, 269, 307 bank, 49 collection, 60–63 conservation, 60–63 Germplasm Resource Information Network (GRIN), 33, 49, 62, 100, 139, 190 Giemsa C-banding, 19 Gigantism, 239 Glacial relict, 263 Glaciation, 75, 261–263

Index Glyphosate, 129 Golf course, 155, 302 Goosegrass, 129 Grain quality, 287 size, 286–287 Gramineae, 96, 153 GrassBase, 175 Grazing, 118, 135, 155 Guineagrass, 183 H Haploid, 221 Heat shock protein (HSP), 4 stress tolerance, 4 Herbicide, 305 resistance, 26, 129, 282, 289 tolerance, 165, 271 Heritability, 287 Heterochromatin, 260 Heterosis, 245 Heterozygosity, 80, 190 Hexaploid, 19, 24, 57, 82, 97, 181, 204, 257 Homoeologous pairing, 157 Homology, 138, 204 Housekeeping gene, 169 Hybridity, 157 Hybridization, 54, 79, 124, 130, 160, 167, 212, 238, 268, 288–289 Hybrid breeding, 246 trispecific hybrid, 246 Hygromycin, 169 I Ideogram, 38, 185 Inbred line, 167 Indigenous grasses, 247 Inflorescence, 117 architecture, 285–286 Inheritance, 85, 169 digenic inheritance, 240 maternal inheritance, 97 monogenic inheritance, 240 Insect resistance, 299 Insertion/deletion (INDEL), 145 Interecotypic hybridization, 80 Intergeneric hybrid, 96, 157, 158 Intergenetic hybridization, 9, 38, 158 Internal transcribed spacer (ITS), 18, 74, 123, 136, 167, 224, 261 International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), 128 Inter-simple sequence repeat (ISSR), 42, 123, 141–142, 208 Interspecific crossing, 59, 287, 297 hybridization, 3, 10, 60, 197, 230, 246, 280–281 hybrid, 19, 67, 96, 212, 300

315 Intraspecific hybridization, 203 Intraspecific variation, 138 Introgression, 20, 64, 235, 269, 289 hybrid, 81 map, 157 Introgressive hybridization, 97 Intron, 158 Invasion, 6, 284, 305 Invasive, 17, 147 species, 191 weed, 41 Invasiveness, 160 Inversion, 200 In vitro culture, 143 In vitro dry matter digestibility (IVDMD), 142 Isoenzyme, 43, 234 Isozyme, 102, 299 K Karyotype, 19, 84–85, 137, 221, 260 Kleingrass, 135 L Lamarckia aurea, 75 Landrace, 287 Landscape, 302 Lasiurus, 49 Leaf spot, 42 Lignin, 102 biosynthesis, 143–144, 159 Linkage, 221 group (LG), 140, 141, 183, 287, 302 mapping, 144 map, 126, 141, 183, 300 Lolium, 49, 73, 153 L. canariense, 166 L. edwardii, 166 L. multiflorum, 165 L. perenne, 157, 165 L. persicum, 165 L. remotum, 165 L. rigidum, 165 L. subulatum, 166 L. temulentum, 165 Love grass boer love grass, 139 Lehmann love grass, 135 stinking love grass, 147 weeping love grass, 135 M Macrosynteny, 227 Magnaporte, 42 Maillea, 260 Major gene, 302 Male sterility, 289 Manodialdehyde (MDA), 43 Mapping, 26, 48 population, 159, 183, 201, 302 Marker-assisted selection (MAS), 84, 128 Marvel grass, 91

316 Mating system, 279 Megaspore, 98 Meiotic instability, 138 Microcolinearity, 126 Microsatellites, 125, 170, 186, 203, 300 Microsporogenesis, 98 Microsynteny, 245 Migration, 75, 261–263 Model species, 167 Molecular breeding, 160 derivative, 264 descent, 74 diversity, 269 form, 264 mapping, 48 profile, 266 Monoculture, 146 Monoembryony, 190 Monophyletic, 39, 103, 208 Monoploid, 221, 247 Monosomic substitution line, 157 mRNA, 201 Mutant, 169 Mutation, 62, 74, 102, 266 N NADP-dependent malate dehydrogenase (NMDH), 105 Napier grass, 246 National Plant Germplasm System (NPGS), 62 National Semi-Arid Resources Research Institute (NaSARRI), 128 Natural selection, 124 Neotyphodium, 271 Nicotinamide adenine dinucleotide phosphate (NADP), 105, 197 Nuclear DNA, 18, 182, 299 O Ochthochloa, 124 Octoploid, 181, 182, 258, 266 Off-type, 275 O2 inhibition, 178 Ophiosphaerella, 6 Orphan crop, 130 species, 145 Oryza sativa, 245 Outbreeder, 167 Outcrossing, 237, 279 Overdominance, 242 Overexpression, 9 Overgrazing, 271 P Palatability, 157 Paleo-allotetraploid, 264 Paniceae, 31 Panicoideae, 31 Panicum, 49 P. capillare, 175

Index P. coloratum, 175 P. dichotomiflorum, 175 P. maximum, 182 P. miliaceum, 175 P. repens, 175 P. virgatum, 175 Paraphyletic, 39, 226 Parthenogenesis, 200, 209 Particle bombardment, 159 Paspalum P. atratum, 211 P. compressifolium, 212 P. conspersum, 206 P. coryphaeum, 207 P. dasypleurum, 205 P. denticulatum, 206 P. dilatatum, 201, 202 P. distichum, 210 P. equitans, 208 P. exaltatum, 207 P. glaucescens, 212 P. guenoarum, 211 P. haumanii, 207 P. intermedium, 202 P. juergensii, 202, 207 P. malachophyllum, 201 P. nicorae, 212 P. notatum, 200, 202, 208 P. paniculatum, 207 P. pauciciliatum, 205 P. plicatulum, 211 P. procurrens, 201 P. rufum, 206 P. scrobiculatum, 212 P. simplex, 200, 201 P. subciliatum, 209–210 P. urvillei, 205 P. usterii, 201 P. vaginatum, 202, 210 P. virgatum, 206 Pearl millet, 232, 238 Pedoturbation, 146 Pennisetum, 32, 39, 184 P. glaucum, 39, 221 P. mezianum, 223 P. ramosum, 223 Pentaploid, 57, 203 Peroxidation, 5 Phenogram, 187 Phenotypic heterogeneity, 283 plasticity, 283 selection, 209 variation, 242 Phenotyping, 160, 169, 237 Phleum, 73, 257 P. alpinum, 258, 261, 263–265 P. arenarium, 259 P. echinatum, 258, 261, 266–267 P. himalaicum, 259–260 P. hirsutum, 258–259

Index P. iranicum, 260 P. montanum, 259 P. paniculatum, 259 P. phleoides, 258 P. pratense, 257, 258, 265, 266 Photosynthesis, 178, 290 Phylogenetic, 96 affinity, 78 analysis, 144 divergence, 221 relatedness, 187, 300 relationship, 39, 278 tree, 123 Phylogeny, 178, 180, 198, 203, 224, 278 Phylogram, 40 Physical mapping, 185 Phytoremediation, 130 PlantGBD, 169 Plasmid artificial chromosome (PAC), 157 Pleiotropic, 199 Ploidy, 66 level, 138, 154, 181, 217 Poa, 136, 153 P. annua, 6 Poaceae, 1, 31, 135, 297 Pollen, 99 flow, 238 Polyembryo, 183 Polyembryony, 99 Polyethylene glycol, 305 Polyhaploid, 97 Polymerase chain reaction (PCR), 102, 290 Polymorphic information content (PIC), 170, 187 Polyploidization, 123–124, 145, 202, 281 Polyploidy, 39, 82, 96, 97, 137, 138, 201 Polyspory, 183 Pooideae, 1 Population structure, 126 Power of discrimination (PD), 187 Pre-breeding, 247–248 Principle coordinate analysis (PCA), 187 Protandry, 280 Protogyny, 280 Protoplast, 8 Pseudo-F2, 169 Pseudogamous, 89, 100 Pseudogamy, 200, 209 Pseudomonas, 2 Pseudo-testcross, 159 Pyricularia, 42 Q Quantitative trait loci (QTL), 4, 47, 142, 144, 159, 190, 236, 285, 299 expression-QTL, 243 R Random amplified polymorphic DNA (RAPD), 3, 102, 123, 140, 158, 167, 186, 201, 278, 300 Real-time quantitative PCR (RT-PCR), 169 Reciprocal cross, 280 Recombination, 102

317 Refugia, 80, 168, 267 Regeneration, 210, 305 Relative water content (RWC), 139 Reporter gene, 9 Resistance (R), 7 Restorer, 289 Restriction fragment length polymorphism (RFLP), 4, 123, 140, 157, 167, 186, 199, 234, 278, 299 R-gene-like sequence (RGL), 7 Rhizobacteria, 2 Ribosomal DNA (rDNA), 121, 123, 185, 223 RNA interference (RNAi), 109 Ryegrass darnel ryegrass, 166 Italian ryegrass, 160, 165 perennial, 271 rigid, 153 S Saccharomyces, 143 Salinity, 108 tolerance, 299 Salt tolerance, 6, 42, 130, 210, 282 Sandbur agrimony sandbur, 32 big sandbur, 35 Coastal sandbur, 34, 36 hedhog sandbur, 33 Indian sandbur, 32 Slender sandbur, 34 Sclerotinia, 7 Secale cereale, 228 Seed bank, 237 flow, 238 shedding, 285 Segregation, 47, 102, 222 distortion, 199, 229 Mendelian, 240 Selection, 237 natural, 237 Selectivity, 305 Self-compatibility, 138, 167, 187, 208, 297 Self-fertility, 167, 175, 209 Self-incompatibility, 66, 155, 175, 190, 206 Self-sterile, 158 Senescence, 4 Sequence characterized amplified region (SCAR), 3, 43 Sequencing, 124, 145 Setaria, 32, 39, 275 S. adhaerans, 278 S. faberi, 277 S. geniculata, 278 S. italica, 277 S. leucopila, 280 S. machrostachya, 280 S. parviflora, 277 S. pumila, 277 S. texana, 280 S. verticillata, 277 S. verticilliformis, 278

318 Setaria (cont.) S. viridis, 277 S. yunnanensis, 289 Sethoxydim, 290–291 Sexual diploid, 211 tetraploid, 211 Shattering, 168, 238 Sheath blight, 107 Shibra, 236, 237 Simple sequence repeat (SSR), 17, 84, 125, 140, 157, 168, 186, 233, 271, 288, 300 Simplex loci, 183 marker, 183 Single nucleotide polymorphism (SNP), 84, 145, 157 Somatic, 160 hybrid, 160 polymorphism, 283 Sorghum, 191 Southern hybridization, 169 Speciation, 96, 279 Stargrass, 59 Substitution, 180 Supernumerary chromosome, 227 Switchgrass, 186, 190 Synteny, 130, 159 T Tall fescue, 139 Tetraploid, 4, 37, 54, 75, 79, 118, 181, 197, 258, 266, 289 sympatric tetraploid, 78 Tetrasomic, 21 Threatened species, 92 Tillering, 286 Timothy, 257 Tonic, 146 Transcript, 4 derived fragment, 160 Transformation, 7, 85, 105, 143, 159, 169, 305 Transgene, 143 Transgenesis, 158, 159, 169 Transgenic, 143–144, 305 cultivar, 85 plant, 8, 85, 158 Transgressive, 242 Triazine, 290

Index Trifluralin, 290 Trifolium, 49 Triploid, 97 Tripsacum, 189 Trisomic, 288 Turf, 56, 160 Turfgrass, 208 U US Department of Agriculture (USDA), 33, 49, 62, 100, 120, 139, 186, 299 US Department of Agriculture-Agricultural Research Service (USDA-ARS), 62, 84 V Vasey grass, 205 Vernalization, 2, 167, 169 Vulnerable species, 96 W Water-logging, 210 Weed, 305 aggressive, 244 Weediness, 26–27 Wheat streak mosaic virus (WSMS), 128 Wild alleles, 244 form, 237 phenotype, 236 population, 236 progenitor, 239 relative, 246 Winter, 23 hardiness, 42, 56, 246 X Xeromorphic, 80 Z Zea mays, 228 Zoysia, 297 Z. japonica, 297 Z. macrostachya, 298 Z. matrella, 298 Z. tenuifolia, 298 Zoysieae, 297