Andrew D. Austin & Mark Dowton (Editors)
HYMENOPTERA Evolution, Biodiversity and Biological Control
© CSIRO 2000 All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO PUBLISHING for all permission requests. National Library of Australia Cataloguing-in-Publication entry International Hymenopterists Conference (4th :, 1999 : Canberra, A.C.T.). Hymenoptera : evolution, biodiversity and biological control. Bibliography. Includes index. ISBN 0 643 06610 1 (hardback). ISBN 0 643 09008 8 (eBook). 1. Hymenoptera – Congresses. I. Austin, Andrew. II. Dowton, Mark. III. International Society of Hymenopterists. 595.79 Available from: CSIRO PUBLISHING
150 Oxford Street (PO Box 1139) Collingwood VIC 3066 Australia Tel: Fax: Email: Web site:
+61 3 9662 7500 +61 3 9662 7555
[email protected] http://www.publish.csiro.au
Front cover: Microplitis demolitor (Braconidae) SEM (original by P.C. Dangerfield) Set in Adobe Minion, Copperplate and Frutiger Typeset by Desktop Concepts P/L, Melbourne Printed in Australia by Ligare
Foreword
The idea for this book originated from discussions held among members of the organising committee for the Fourth International Hymenoptera Conference, held in Canberra in January 1999. It was thought that an up-to-date account of current research being undertaken on this important group of insects was timely given the rapid progress that has occurred in numerous areas during the last 10 years. Further, the conference and symposia that comprised the scientific program attracted many of the best international researchers in their fields. In all, approximately four-fifths of the papers and posters presented at the conference appear in this volume, although many have been substantially modified in scope and content compared with the original conference presentations. All of them have been extensively refereed using the guidelines that generally apply to the scientific journals published by CSIRO Publishing. The editors have organised the papers into related topics and an overview of these are given in the first section ‘The Hymenoptera – An Introduction’. As a compendium of current reviews and research papers, we hope this volume will stimulate interest among students of the Hymenoptera, researchers from related disciplines, and those who simply wish to learn more about this fascinating group of insects. Andrew D. Austin & Mark Dowton
iii
This Page Intentionally Left Blank
Acknowledgements
The production of this book would not have been possible without the help and financial support of the International Society of Hymenopterists and delegates that attended the Fourth International Hymenoptera Conference held in Canberra in January 1999. The editors would particularly like to thank the following people and organisations who contributed in various ways to the development and production of this book: John LaSalle and Paul DeBarro who acted as conference symposium organisers; CSIRO Entomology and the Australian Quarantine & Inspection Service for sponsoring plenary speakers at the conference; John Jennings and Nick Stevens for their editorial assistance and helpful ideas; Jim Whitfield, John LaSalle, Bob Matthews and Peter Bailey for their constructive comments on several sections; The Department of Applied & Molecular Ecology, The University of Adelaide and the Department of Biology, Wollongong University for access to facilities used by the editors; Emily Shephard, Image and Copy Centre, The University of Adelaide for assistance with illustrations; the numerous people who referred manuscripts, and the authors for their cooperation and patience when the editors insisted on last minute changes.
v
This Page Intentionally Left Blank
Contents
Foreword
iii
Acknowledgements
v
Part 1 The Hymenoptera – an introduction
3
Andrew Austin and Mark Dowton
Part 2 – Development and Physiology The effects of life history on development of the Hymenoptera
11
M. R. Strand Molecular basis of developmental alteration in Heliothis virescens (F.) larvae parasitised by Cardiochiles nigriceps Viereck
17
Francesco Pennacchio, Patrizia Falabella, Paola Varricchio, Rocco Sordetti, S. Bradleigh Vinson, Franco Graziani and Carla Malva Protection by immune disguise: a new lesson from a parasitoid wasp
23
Ulrich Theopold, Dongmei Li, Wanja Kinuthia and Otto Schmidt Host defence manipulation by parasitoid wasps and the problem of assessing host specificity
29
Otto Schmidt, Sassan Asgari, Markus Beck and Ulrich Theopold Two genetically distinct Venturia canescens strains display different reproductive strategies
38
Markus Beck, Ulrich Theopold and Otto Schmidt The response of Cardiochiles nigriceps Viereck (Hymenoptera: Braconidae) larvae to conspecific competitors
46
S. Bradleigh Vinson and Ahmed Kamal Mourad Hypopharyngeal gland funtion, glandular cell senescence and gland reactivation in bees
50
Carminda da Cruz-Landim, Rosiléia Ana Cassia da Costa and Regina Lúcia Morelli Silva de Moraes
Part 3 – Molecular Phylogenetics Molecular systematics of the Chalcidoidea using 28S-D2 rDNA
59
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock
vii
Contents viii
Hymenoptera: Evolution, Biodiversity and Biological Control
Phylogeny of the subfamilies of Ichneumonidae (Hymenoptera): a simultaneous molecular and morphological analysis
74
D. L. J. Quicke, M. G. Fitton, D. G. Notton, G. R. Broad and K. Dolphin Molecular evolution in social wasps
84
J. Schmitz and R. F. A. Moritz Rearrangement of the hymenopteran mitochondrial genome is accelerated relative to orthopteroid insects
90
Mark Dowton, Andrew D. Austin and Paul K. Flook Phylogeny of microgastroid braconid wasps, and what it tells us about polydnavirus evolution
97
James B. Whitfield Evolutionary transitions in Aphidiinae (Hymenoptera: Braconidae)
106
Paul T. Smith and Srinivas Kambhampati Genetic structure of the cypress seed chalcid Megastigmus wachtli (Torymidae) within its Mediterranean distribution
114
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques Systematics of the ant genus Camponotus (Hymenoptera: Formicidae): a preliminary analysis using data from the mitochondrial gene cytochrome oxidase I
131
Seán G. Brady, Jürgen Gadau and Philip S. Ward
Part 4 – Systematics Can braconid classification be restructured to facilitate portrayal of relationships?
143
Robert A. Wharton Higher-level phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea)
154
John T. Jennings and Andrew D. Austin Monophyly and relationship of the genus Coelopisthia Förster (Chalcidoidea: Pteromalidae)
165
Hannnes Baur A preliminary phylogeny for the Baeini (Hymenoptera: Scelionidae): endoparasitoids of spider eggs
178
Muhammad Iqbal and Andrew D. Austin Hymenopteran orbicular sensilla
192
Hasan H. Basibuyuk, Alexandr P. Rasnitsyn, Mike G. Fitton and Donald L. J. Quicke Karyology of parasitic Hymenoptera: current state and perspectives Vladimir E. Gokhman
198
Contents ix
Morphology and biogeography of the north African Ceramius maroccanus-complex (Vespidae: Masarinae): contribution of morphometric analyses to taxonomic decisions Volker Mauss
207
Some problems with the Australian tiphiid wasps with special reference to coupling mechanisms Graham. R. Brown
210
Historical review and current state of the world genera classification of oak gall wasps (Hymenoptera: Cynipidae, Cynipini) George Melika and Warren G. Abrahamson
218
Australian Hymenoptera in the Spinola collection: a list of species M. Generani and P. L. Scaramozzino
231
Part 5 – Biology, Ecology and Behaviour New insights into the foraging behaviour of parasitic wasps Michael A. Keller and Brigitte Tenhumberg
247
The biology of perreyiine sawflies (Hymenoptera: Pergidae) of the Perreyia genus-group Carmen Flores, Jesús Ugalde, Paul Hanson and Ian Gauld
258
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California: methods of introduction and evidence of host shifting E. E. Grissell and K. R. Hobbs
267
Biology of an extant species of the scolebythid genus Dominibythus (Hymenoptera: Chrysidoidea: Scolebythidae), with description of its mature larva Gabriel A. R. Melo
281
Defense adaptations in velvet ants (Hymenoptera: Mutillidae) and possible sources of selection pressure for such Donald G. Manley
285
Introduction and spread of four aculeate Hymenoptera in Italy, Sardinia and Corsica (Hymenoptera: Sphecidae, Chrysididae) Guido Pagliano, Pierluigi Scaramozzino and Franco Strumia
290
Biological notes and larval morphology of Donquickeia (Hymenoptera: Braconidae: Doryctinae) Angélica Maria Penteado-Dias
296
Part 6 – Biodiversity Driving Miss DAISY: the performance of an automated insect identification system I. D. Gauld, M. A. O'Neill and K. J. Gaston
303
Contents x
Hymenoptera: Evolution, Biodiversity and Biological Control
Data warehousing architecture and tools for Hymenoptera biodiversity informatics Norman F. Johnson and Luciana Musetti
313
Preliminary study of pteromalid diversity in China: taxonomic and geographic variation Hui Xiao, Da-Wei Huang and Steven L. Heydon
320
The family Braconidae in China (Hymenoptera) Chen Xuexin, He Junhua and Ma Yun An annotated list of Encyrtidae (Hymenoptera: Chalcidoidea) of Tbilisi (Georgia) George. O. Japoshvili
328
339
Part 7 – Biological Control Predictive and empirical evaluation for parasitoids of Bemisia tabaci (biotype ‘B’), based on morphological and molecular systematics J. A. Goolsby, M. A. Ciomperlik, A. A Kirk, W. A. Jones, B. C. Legaspi, Jr. , J.C. Legaspi, R. A Ruiz, D. C. Vacek and L. E. Wendel
347
Which factors govern the host preference of aphid parasitoids when offered host races of an aphid species? Anja Hildebrands, Thomas Thieme and Stefan Vidal
359
Size and asymmetry as quality control indicators in Trichogramma spp. (Hymenoptera: Trichogrammatidae) D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann
371
The effects of two new insecticides on the survival of adult Trichogramma pretiosum Riley in sweet corn B. C. G. Scholz and M. P. Zalucki
381
Field observations on selective food plants in habitat manipulation for biological control of potato moth by Copidosoma koehleri Blanchard (Hymenoptera: Encyrtidae) L. R. Baggen, G. M. Gurr and A. Meats
388
Understorey management for the enhancement of the leafroller parasitoid Dolichogenidea tasmanica (Cameron) in Canterbury, New Zealand orchards N. A. Irvin, S. D. Wratten and C. M. Frampton
396
Impact and control of introduced Vespula wasps in New Zealand Jacqueline R. Beggs
404
Taxonomic relationships of parasitoids: poor indicators for their suitability or effectiveness as biological control agents D. P. A. Sands
410
Natural population of Aenasius advena Compere (Chalcidoidea: Encyrtidae) and its host preference in Bangladesh Badrul A. Bhuiya, Shafique H. Chowdhury and S. M. Humayun Kabir
417
Contents xi
Studies on Eretmocerus sp. (Hymenoptera: Aphelinidae) – a promising natural enemy of the castor whitefly Trialeurodes ricini (Hemiptera: Aleyrodidae) Seeba Balan and R.W. Alexander Jesudasan
421
Part 8 – Hymenoptera in Education Developing life science instructional materials using a parasitic wasp, Melittobia digitata Dahms (Hymenoptera: Eulophidae): a case study Robert W. Matthews
427
Part 9 – Medical Effects of Hymenoptera Review of bee and wasp sting injuries in Australia and the USA N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel
437
Part 10 – Future Research Hymenopteran research – future directions into the next millennium Mark Dowton and Andrew Austin
451
Indexes Index to authors
457
Index to hymenopteran names
458
Index to other animal names
466
Index to plant and micro-organism names
468
This Page Intentionally Left Blank
PART
1
Introduction
This Page Intentionally Left Blank
The Hymenoptera: An Introduction Andrew D. Austin1 and Mark Dowton1,2 1
Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) 2
Australian Flora and Fauna Research Centre, Department of Biology, Wollongong University, Wollongong, N. S. W. 2522 Australia (email:
[email protected])
The insect order Hymenoptera is one of the dominant life forms on earth, both in terms of number of species and in the diversity of life styles that have evolved within the group. The Hymenoptera contain the vast majority of socially organised insects and parasitoids, as well as a great variety of specialist predators and herbivores. They have emerged as the most speciose group in many studies on terrestrial biodiversity and they are pre-eminent as biological control agents of pest insects. The number of hymenopteran species is unknown and, at present, is almost impossible to estimate with any accuracy. Even the number of described species has not been accurately calculated given that many families do not have available check-lists or catalogues (but see for example Johnson 1992; Bolton 1995; Noyes 1998). LaSalle and Gauld (1993) and Gaston (1993) have estimated the number of described species of Hymenoptera at >115,000 species. However, the total number (including undescribed species) could be 5–10 times this figure given that this is often the proportion of new species that are discovered following taxonomic revision of highly speciose families (Austin 1999). The major difficulty for accurately estimating total hymenopteran species comes from trying to determine the number of species for the ‘megadiverse’ regions of the world. These mostly comprise tropical or subtropical countries (i.e. Australia, India, Malaysia, Indonesia, China, Brazil, Ecuador, Peru, Columbia, Mexico, Zaire and Madagascar; McNeely et al. 1990) that with few exceptions have been poorly surveyed. In recent years, the hymenopteran fauna of Costa Rica has been particularly well-studied compared with other countries (Hanson & Gauld 1995) and this work serves as a useful foundation for future research on the fauna of Costa Rica itself, and for comparison with other regions. The true extent of species richness and biological complexity within the Hymenoptera dictates that the group should be at the centre of studies assessing arthropod diversity. These facets will only be revealed when detailed studies similar to those in Costa Rica are undertaken for other species-rich regions of the world. Gauld and Bolton (1996) identify four biological features that have been of primary importance in the evolution of the Hymenoptera. These are the ovipositional mechanism (used both for oviposition and for venom delivery), parental provision for the larva (i.e. location of food on or in which an egg is deposited), diversification of larval diet (associated with the development of a blindending gut and storage of dietary waste until pupation), and haplo-diploid sex determination. Some of these factors are obviously strongly connected but together they have resulted in a bewildering array of specialised life histories, including ecto- and endoparasitoids, hyperparasitoids,
3
Andrew D. Austin and Mark Dowton 4
Hymenoptera: Evolution, Biodiversity and Biological Control
gall-formers, seed feeders, pollinators (bees and fig wasps), predators (e.g. spider wasps), and eusocial species. In most cases there are major differences between the life histories of juvenile (larval) and adult stages, particularly for predatory and parasitic species. It is inevitably the larvae which are parasitic or predatory, while the adults are phytophagous (feeding on nectar or pollen) or are non-feeding. In the case of most aculeate Hymenoptera, although larval and adult life histories are very different, they are also strongly linked as the adult stage is responsible for providing the developing larvae with food. In the case of many wasps (e.g. Pompilidae, Vespidae, Sphecidae) the adult female uses a sting (modified ovipositor) to subdue prey items with a paralysing venom on which the larvae then develop in a previously constructed nest. In many respects the Hymenoptera have reached the pinnacle of evolutionary complexity within the insects. This is typified by two examples which have had far-reaching influence on the biological and species diversity of the order. These are: 1) the evolution of highly social species and the corresponding development of separate reproductive castes (Wilson 1971) – characteristics which, compared with the termites, have evolved multiple times within several aculeate lineages (i.e. within the Apidae, Formicidae and the Vespidae); and 2) the evolution of endoparasitism which, like eusociality, has evolved separately in numerous parasitic lineages (Dowton & Austin 1994). Significantly, endoparasitic Hymenoptera are challenged by the immune defences of their host and, in response, they have evolved a bewildering array of structural, behavioural, physiological, molecular and symbiotic adaptations to overcome these defences and allow the successful development of the parasitoid larva (Vinson & Iwantsch 1980; Vinson 1990). Without doubt, one of the most intriguing adaptations is the apparent mutualistic association between some ichneumonoid wasps and viruses, where the latter act to neutralise the host’s immune system (Stoltz & Whitfield 1992; Beckage 1998). Developing a comprehensive knowledge of events such as these (eusociality and endoparasitism) is important to understanding and eventually manipulating key processes in natural and agricultural systems, and the use of novel hymenopteran compounds in biotechnology. These include nutrient turn-over by ants, pollination of native plants and crops by bees, the efficiency and host specificity of parasitoids employed as biological control agents, the development of antifungal/antibacterial compounds sequestered by ants and wasps, and the incorporation of insect specific venom genes into plants. Given the extent of research currently being undertaken on the Hymenoptera it is imperative that the results of this research be brought together on a regular basis to assess current directions, and reflect on future needs and potential for new initiatives. Unfortunately, this has not happened for many years, and previous synopses have mostly concentrated on parasitic Hymenoptera (e.g. Waage & Greathead 1986; Bin 1991; Gupta 1993). However, in more recent years this has been partly balanced by the appearance of several excellent books on the biology and/or identification of the Hymenoptera (e.g. Hölldobler & Wilson 1990; Ross & Matthews 1991; LaSalle & Gauld 1993; Goulet & Huber 1993; Godfray 1994; Gauld & Bolton 1996; Quicke 1998). This volume provides a timely overview of the current research being undertaken on the Hymenoptera. It comprises a series of review and research papers by those currently working in the field and covers all major areas of hymenopteran research. The first section deals with the development and physiology of the Hymenoptera. Strand reviews the effect that life history has on early embryonic development and shows that there is substantial variability among wasp groups, particularly the parasitoids, and that they differ from the almost universally accepted pattern for Drosophila. Theopold et. al. and Schmidt et al. examine aspects of host immune
The Hymenoptera: an Introduction 5
disruption by polydnaviruses and the underlying mechanisms involved, while Beck et al. and Vinson & Mourad present new findings on behavioural aspects of host-parasitoid interactions. An understanding of the evolutionary relationships of the Hymenoptera has remained elusive, even though such knowledge impacts on conclusions reached from almost every other study of hymenopteran biology. Rasnitsyn’s (1980, 1988) morphological and fossil-based phylogenies have spawned a series of studies in the last decade of the twentieth century that critically assess his ideas. Many of these studies have used comparative DNA sequences, and in the section molecular phylogenetics, a series of chapters report on the phylogeny of the Hymenoptera from the ordinal to the species level. Dowton et al. present a comparison of gene rearrangement frequencies between the Hymenoptera and orthopteroid insects; Campbell et al. present the first molecular phylogeny for the families of Chalcidoidea; Quicke et al. report on a combined molecular and morphological analysis of ichneumonid relationships; Schmitz & Moritz examine molecular evolution among the social wasps using sequence data from multiple genes, Whitfield shows that the phylogeny of polydnaviruses tracks that of their microgastrine hosts; Smith & Kambhampati examine the molecular phylogeny of the aphidiines and the major biological transitions that have occurred within the group; Brady et al. report on a molecular study of the ant genus Camponotus; while Rasplus et al. examine intraspecific variation within a seed wasp using molecular data. Chapters comprising the systematics section represent both review and research papers that cover the full spectrum of the discipline including classification, phylogenetics, karyology, morphometrics, comparative morphology, taxonomic history and check-listing. Wharton presents a new and somewhat controversial scheme for the reclassification of braconid subfamilies; Jennings & Austin examine the phylogeny of the Aulacidae and Gasteruptiidae; Iqbal & Austin re-evaluate the relationships among the spider-parasitising Scelionidae; Gokhman reviews the role of karyology in hymenopteran systematics; Basibuyuk et al. report on the comparative morphology of orbicular sensilla and their phylogenetic potential within the order; Baur examines the relationships of the pteromalid genus Coelopisthia; Brown discusses the morphology of the coupling mechanism of thynnine wasps and impact of this character system on generic-level taxonomy; Mauss reports on the taxonomy of the masarine genus Ceramius; Melika & Abrahamson present an historical review of the classification of oak gall wasps (Cynipini); and Generani & Scaramozzino discuss the Australian Hymenoptera in the Spinola Collection and present a species check-list. As discussed above, the Hymenoptera exhibit a great diversity of life styles. In the section on biology and behaviour this theme is expanded on by Flores et al. who report on the first instance of possible mycophagy in larvae of a pergid sawfly from Costa Rica; Keller & Tenhumberg provide a detailed overview of foraging behaviour in parasitic wasps; Grissell & Hobbs provide a comprehensive account of host plant switching in a species of Megastigmus; separate accounts by Manley and Melo outline aspects of the biology of mutillids and scolebythids respectively; Pagliano et al. present an historical account of the geographic spread of four species of Sphecidae and Chrysididae introduced into Italy; and Penteado-Dias discusses the biology of a putatively phytophagous braconid from Brazil. It is increasingly clear that the challenge to catalogue and measure the species richness of Hymenoptera will be met with computer-assisted technologies. In the section on biodiversity, Johnson & Musetti discuss a framework that will bring together the fragmented but invaluable information from separate insect collections through a single web-site, while Gauld et al. report
Andrew D. Austin and Mark Dowton 6
Hymenoptera: Evolution, Biodiversity and Biological Control
on the development of an image-analysis system that will ultimately facilitate the ‘hands off’ sorting and identification of Hymenoptera from complex samples. Xiao et al. and Xuexin et al. present accounts of the Pteromalidae and Braconidae of China respectively, while Japoshvili provides a check-list of the Encyrtidae and their hosts from Georgia (ex USSR). The modern application of parasitoids as control agents of agricultural pests combines a detailed knowledge of the ecology of both organisms, as well as of the system which they inhabit. The section on biological control presents a diversity of studies that reflect this approach. The chapter by Goolsby et al. provides a detailed overview of contemporary taxonomic research to characterise parasitoids of Bemisia tabaci which could serve as a model for other such studies; Hieldebrands et al. describe the factors that influence host preference by aphid parasitoids; Bennett et al. assess size and asymmetry characteristics as measures of fitness in Trichogramma, while Scholz & Zulucki describe the effects of two new pesticides on the same genus; effects of habitat manipulation on parasitoid performance are assessed for two different systems by Baggen et al. and Irvin et al.; Sands discusses contentious aspects of predicting host range of parasitoids under quarantine conditions; Beggs provides an account of the control strategies being considered for the invasive environmental pest, Vespula vulgaris in New Zealand; while Bhuiya et al. and Balan & Jesudasan report on aspects of the biology of potential biocontrol agents on the Indian subcontinent. In the section on Hymenoptera in education, Matthews outlines the novel use of a species of Melittobia for biological science instruction at secondary level, while Levick et. al. review the level, treatment and recording of bee and wasp stings in Australia and USA in the section medical effects of Hymenoptera. Finally, in hymenopteran research – future directions into the next millennium, Dowton & Austin reflect on the chapters in this book as a representative overview of contemporary work on the group, and they provide a personal commentary on the research areas that are likely to be most influential during the next decade in providing a deeper understanding of the evolution of the Hymenoptera, their use as model systems in biology, and their application in environmental biology and biological control.
Acknowledgements We wish to thank John Jennings and John LaSalle for their comments on a draft of the manuscript.
References Austin, A. D. (1999) The role of species in biodiversity research – lessons from the parasitic Hymenoptera. pp. 159-165. In W. Ponder & D. Lunney (Eds), The Other 99% – The Conservation and Biodiversity of Invertebrates. Royal Society of New South Wales, Sydney. Beckage, N. E. (1998) Parasitoids and polydnaviruses. Bioscience 48: 305-311. Bin, F. (Ed.) (1991) Insects Parasitoids – Tritrophic Interaction. Redia vol. 74. Bolton, B. (1995) A New General Catalogue of the Ants of the World. Harvard University Press, Cambridge, Massachusetts. Dowton, M. & Austin, A. D. (1994) Molecular phylogeny of the insect order Hymenoptera: Apocritan relationships. Proceedings of the National Academy of Sciences, USA 91: 9911-9915.
The Hymenoptera: an Introduction 7
Gaston, K. J. (1993) Spatial patterns in the description and richness of the Hymenoptera. pp. 277-293. In LaSalle, J. & Gauld, I. D. (Eds) Hymenoptera and Biodiversity. CABI, Wallingford. Gauld, I. & Bolton, B. (Eds) (1996) The Hymenoptera. 2nd edition. Oxford University Press, Oxford and The Natural History Museum, London. Godfray, H. C. J. (1994) Parasitoids – Behavioral and Evolutionary Ecology. Princeton University Press, Princeton. Goulet, H. & Huber, J. T. (Eds) (1993) Hymenoptera of the World: An Identification Guide to Families. Research Branch, Agriculture Canada, Ottawa. Gupta, V, K. (Ed.) (1993) Studies on the Hymenoptera. A Collection of Articles on Hymenoptera Commemorating the 70th Birthday of Henry K. Townes. Contributions of the American Entomological Institute Vol. 20. Hanson, P. E. & Gauld, I. D. (Eds) (1995) The Hymenoptera of Costa Rica. Oxford University Press, Oxford and The Natural History Museum, London. Hölldobler, B. & Wilson, E. O. (1990) The Ants. Belknap Press, Cambridge, Massachusetts. Johnson, N. F. (1992) Catalog of World Species of Proctotrupoidea, Exclusive of Platygastridae (Hymenoptera). The American Entomological Institute: Gainesville. LaSalle, J. & Gauld, I. D. (Eds) (1993) Hymenoptera and Biodiversity. CABI, Wallingford. McNeely, J. A., Miller, K. R., Reid, W. V., Mittermeier, R. A. & Werner, T. B. (1990) Conserving the World’s Biological Diversity. IUCN, Gland, Switzerland; WRI, CI,WWF-US and the World Bank, Washington, D.C. Noyes, J. S. (1998) Catalogue of the Chalcidoidea of the World. Biodiversity Catalogue Database and Image Library CDROM Series. ETI Biodiversity Center, Amsterdam. Quicke, D. L. J. (1998) Parasitic Wasps. Chapman & Hall, London. Rasnitsyn, A. P. (1980) The origin and evolution of the Hymenoptera. Trudy Paleontologicheskogo Instituta Akademiya Nauk SSSR 174: 1-192 [In Russian]. Rasnitsyn, A. P. (1988) An outline of evolution of the hymenopterous insects (Order Vespida). Oriental Insects 22: 115-145. Ross, K. G. & Matthews, R. W. (1991) The Social Biology of Wasps. Comstock Publishing, Ithaca. Stoltz, D. B. & Whitfield, J. B. (1992) Viruses and virus-like entities in the parasitic Hymenoptera. Journal of Hymenoptera Research 1: 125-139. Vinson, S. B. (1990) How parasitoids deal with the immune system of their host: an overview. Archives of Insect Biochemistry & Physiology 13: 3-27. Vinson, S. B. & Iwantsch, G. F. (1980) Host suitability for insect parasitoids. Annual Review of Entomology 25: 397-419. Waage, J. & Greathead, D. (Eds) (1986) Insect Parasitoids. Academic Press, London. Wilson, E. O. (1971) The Insect Societies. Belknap Press, Cambridge, Massachusetts.
This Page Intentionally Left Blank
PART
2
Development and Physiology
This Page Intentionally Left Blank
The Effects of Life History on Development of the Hymenoptera M. R. Strand Department of Entomology, University of Wisconsin-Madison, Madison, WI 53706 USA (email:
[email protected])
Introduction Two key questions in developmental biology are what factors drive changes in developmental programs, and how do these changes lead to novelties in embryogenesis, morphogenesis and adult morphology? Comparison of organisms across diverse phyla suggest that changes in developmental programs likely result from an interplay between phylogenetic history and the developmental constraints that arise within taxa, and the environment in which organisms develop. Intensive study of model organisms like Caenorhabditis, Drosophila, Xenopus, zebrafish and the mouse indicates that conserved regulatory gene families direct development of all metazoans. Comparison of these phylogenetically distant animals has also led to the suggestion that regulatory processes evolve slowly and that when species exhibit differences in embryonic development, they usually also exhibit marked differences in adult morphology (Gould 1977; Buss 1987; Thomson 1988). However, a few comparative studies between closely related species report distinct differences in embryogenesis, yet their adult stages look very similar (del Pino & Ellinson 1983; Scott et al. 1990; Wray & Raff 1990; Jeffery & Swalla 1991; Raff 1992; Wray & Bely 1994). These results suggest that changes in early development may arise in response to shifts in life history (Wray 1995). How widespread punctuated modes of developmental evolution are among taxa, and whether certain life history transformations lead to changes in early development more often than others, are unclear. The study of D. melanagaster Meigen has resulted in a greater understanding of the processes regulating insect embryogenesis than any other group of animals. In contrast, the role life history has played in shaping the developmental evolution of insects is largely unknown. We recently suggested that the Hymenoptera are especially well suited for examining how ancestry and environmental factors interact to affect development, because of the replicate shifts in life history that have arisen in this group (Strand & Grbic 1997; Grbic & Strand 1998; Strand 1999). Here, I briefly discuss the variation seen in the early development of wasps, factors that have potentially contributed to this variation, and how alterations in embryonic programming have influenced the evolution of novel traits.
Insect Embryogenesis To appreciate the diversity seen in hymenopteran development, it is important to first consider the development of insects at-large. Most insects, including D. melanogaster, lay yolky eggs that undergo syncytial cleavage and long germband development whereby all segments of the body are established near simultaneously (Sander 1983). In Drosophila, this patterning process is initiated by maternal factors localised during oogenesis that trigger transcription of gap and
11
M. R. Strand 12
Hymenoptera: Evolution, Biodiversity and Biological Control
pair-rule segmentation genes whose products diffuse within the syncytium to produce gradients of positional information (St. Johnston & Nusslein-Volhard 1992). By the time the blastoderm cellularises, these factors have programmed the cells in different regions of the embryo to express segment polarity and homeotic genes that define segment-specific and regional elements such as antennae or legs (Ingham 1988). The most important question from a life history perspective is whether these events are representative of insects generally. Developmental biologists have generally assumed the answer to be yes. Some differences in patterning have been documented in insects from phylogenetically more primitive groups like grasshoppers (Patel et al. 1992; Dawes et al. 1994), but studies of other holometabolous insects including beetles (Tribolium sp.), moths (Manduca sexta (L.)), other flies (Musca domestica L.) and the honeybee (Apis mellifera L.) agree well with the paradigms established in Drosophila (summarised by Tautz & Sommer 1995; Strand & Grbic 1997).
The Hymenoptera Suggest that Environmental Factors Profoundly Affect Insect Development If ancestry is the primary factor influencing developmental processes, we would predict from the results mentioned above that embryonic development of hymenopterans should proceed similarly to the honeybee and other holometabolous insects. However, if patterns in early development reflect the environment in which eggs develop, then all of the insects discussed above may appear similar to one another because their eggs all develop under similar environmental circumstances. In contrast, the conditions under which many parasitic Hymenoptera develop differ considerably from the terrestrial conditions experienced by most other insects. Modern treatments of hymenopteran phylogeny recognise that the more primitive hymenopterans traditionally placed in the suborder ‘Symphyta’ (sawflies and woodwasps e.g. Xyeloidea, Tenthredinoidea, Megalodontoidea, Cephoidea and Siricoidea) are in actuality a paraphyletic assemblage, whereas the more advanced hymenopterans in the suborder Apocrita are monophyletic (Rasnitsyn 1988; Dowton & Austin 1994; Hanson & Gauld 1995; Whitfield 1998). The Apocrita likely evolved from an ectoparasitic ancestor (Orussoidea), but thereafter free-living, ectoparasitic and endoparasitic species have arisen independently within and/or between each lineage. These replicate shifts in life history have also given rise to situations where relatively closely related species develop in very different environments. At the superfamily level, the sister group to the Aculeata, as represented by the honeybee, is the parasitic Ichneumonoidea whose species are divided into the families Braconidae and Ichneumonidae. Ecto- and endoparasitic species occur in both families. To examine whether shifts between a free-living, ecto-, and endoparasitic existence affects early development, we examined two parasitic wasps from the Ichneumonoidea in relation to the paradigms established through Drosophila and the honeybee (Grbic & Strand 1998). Habrobracon hebetor Say is a braconid ectoparasitoid that lays its eggs on larvae of certain moths, whereas Aphidius ervi Nees is a braconid endoparasitoid whose eggs develop in the hemocoel of selected aphids. Morphological characterisation revealed that H. hebetor lays yolky eggs surrounded by a rigid chorion, and that early embryogenesis proceeds in a syncytium similar to other canonical long germband insects (Fig. 1). In contrast, A. ervi undergoes a very different form of embryogenesis. Except for the first cleavage that proceeds without cytokinesis, all other cleavages are cellular (Fig. 1). Injection fluorescently conjugated dextran tracers into selected blastomeres confirmed that A. ervi undergoes holoblastic (complete) cleavage, and that molecules the size of the transcription factors regulat-
The Effects of Life History on Development of the Hymenoptera 13
ing Drosophila patterning are unlikely to freely diffuse between embryonic cells. This form of cleavage results in development of a morula stage embryo surrounded by an extraembryonic membrane. The morula ruptures from the chorion and, thereafter, undergoes morphogenesis in a manner that resembles short germband development. To compare patterning events at the molecular level, we stained H. hebetor and A. ervi embryos with antibodies that recognise conserved epitopes of Eve, Engrailed (En) and Ultrabithorax/ Abdominal-A (Ubx/Abd-A) in different insect species. Eve, a primary pair-rule gene is expressed in the Drosophila syncytium and forms a characteristic seven-stripe pattern with double segment periodicity (Frasch et al. 1987). En, which is regulated by Eve, is a segment polarity gene that specifies the posterior segmental compartments (MacDonald et al. 1986). Ubx and Abd-A are Drosophila homeotic proteins that specify the posterior thorax and abdomen (Sanchez-Herrero et al. 1985). In H. hebetor, Eve, En and Ubx/Abd-A are expressed in a largely conserved fashion to Drosophila and other long germband insects (Grbic & Strand 1998). In A. ervi, however, we were unable to detect either a pair-rule or segmental pattern for Eve expression although Eve antigen was detected in dorso-lateral mesoderm and neurons. En stripes formed sequentially in A. ervi during germband extension, resulting in a mature pattern of segmentally iterated stripes that localised to the posterior segmental compartments. Ubx/Abd-A was expressed in the posterior thorax and abdomen in the retracted germband stage (Grbic & Strand 1998). The contrasts between early development of H. hebetor and A. ervi are greater than any described previously for insects in the comparative developmental literature. Yet, rather than comparing insects from different orders, these wasps reside in the same monophyletic family. The most striking difference between H. hebetor and A. ervi is the environment in which their eggs develop; an observation suggesting that shifts in life history may induce significant alterations in early development of insects. The similarities between H. hebetor, the honeybee, and Drosophila suggest that the evolution of a parasitic life history per se does not result in significant alterations in early development. However, the shift from an essentially free-living, terrestrial existence (Habrobracon) to development within another organism (Aphidius) has favoured adaptations in Aphidius for survival in a new environment. In particular, the loss of yolk and a chorion in Aphidius would appear to be key alterations in the shift from syncytial to total cleavage, and in the corresponding alterations in expression of genes regulating anterior-posterior axis formation. If so, we would predict that differences in patterning mechanisms should also be seen between species from other higher groups. Indeed, our own studies and the descriptive embryological literature indicate that total cleavage and alterations in expression of patterning genes has arisen among endoparasitoids from other apocritan lineages. In contrast, every ectoparasitic and free-living hymenopteran examined to date undergoes syncytial cleavage and patterns of development similar to H. hebetor and the honeybee (summarised by Strand & Grbic 1997). Our comparative studies further suggest that alterations in early cleavage and embryonic patterning have also been essential preadaptations for the evolution of other traits. Among the most dramatic of these is polyembryony which is defined as the formation of multiple embryos from a single egg. In insects, polyembryony is known only from endoparasitoids in four families of Hymenoptera (Braconidae, Platygasteridae, Dryinidae and Encyrtidae) and the Strepsiptera. Detailed studies of encyrtids like Copidosoma floridanum (Ashmead) and descriptions of polyembryony in other wasp families reveal remarkable similarities in early development of all polyembryonic species. These include the oviposition of yolkless eggs, complete cleavage, and formation of a trophamnion of polar body origin (reviewed by Strand & Grbic 1997). In each polyembyronic
M. R. Strand 14
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Embryogenesis of Habrobracon hebetor and Aphidius ervi. Confocal, fluorescent and Nomarski images of embryonic development: A) After oviposition the H. hebetor egg has a clear polarity corresponding to the dorsal-ventral and anterior-posterior embryonic axes: embryonic nuclei (arrowed) divide without cytokinesis; B) During the first few syncytial cleavages nuclei remain in the yolk (arrowed): after the tenth cleavage nuclei migrate to the periphery of the egg where they undergo two additional division cycles in the syncytium before finally forming a cellular blastoderm; C) Following germband, the germband undergoes retraction and segmentation (anterior and posterior limits of embryo marked by arrows); D) After oviposition, the A. ervi egg is lemon-shaped and does not exhibit any axial polarity (nucleus marked by arrow, chorion by arrowhead); E) The first nuclear division proceeds in a syncytium, without cytoplasmic cleavage, but subsequent cleavages result in nuclei separated by cell membranes (phalloidin staining demarcates the cell cortex underlying the cell membranes): large blastomeres (arrowed) give rise to an extraembryonic membrane while small blastomeres give rise to the future embryo; F) Later in development, the extraembryonic membrane (arrowed) surrounding the morula stage embryo ruptures from the chorion (arrowhead); G) The embryo undergoes germband extension by posterior growth followed by condensation and segmentation (extraembryonic membrane removed to facilitate viewing) (Figures A-G adapted from Grbic & Strand 1998). Scale bars: A-C = 80 µm; D-G = 7 µm.
The Effects of Life History on Development of the Hymenoptera 15
taxon, multiple embryos arise from the simultaneous proliferation of blastomeres and partitioning of these cells by ingrowth of a trophamnion. Strand and Grbic (1997) concluded that syncytial cleavage and the constraints on volume inherent in the architecture of typical insect eggs would prevent polyembryony from ever evolving in most insect groups. However, once the transition from syncytial to complete cleavage arose in monoembryonic endoparasitoids, polyembryony has arisen on several occasions. Among the genetic regulatory changes required for polyembryony is the uncoupling of pattern formation processes from early cleavage events in the egg. Analysis of C. floridanum indicates that this has occurred since blastomeres remain undifferentiated, and no patterning genes are expressed during embryo proliferation (Grbic et al. 1996a, 1966b, 1998). Freed from the constraint of early specification of cell fate, embryo proliferation can then proceed by partitioning of blastomeres over the course of the life cycle of the host. Among the ecological transitions favouring polyembryony would be host shifts toward increased size or conditions in which risks of immature mortality are high. Not surprisingly, all polyembryonic wasps are egglarval parasitoids or larval parasitoids that oviposit in young hosts. In summary, our results indicate that early development of the Hymenoptera can vary greatly among species and that departures from the paradigms established in Drosophila may occur commonly among insects whose eggs develop under conditions different from typical terrestrial species. Important future research goals include the need for quantitative analysis of developmental traits in relation to established phylogenies. This will require more rigorous comparison of embryonic and larval development of species from appropriate taxa to determine how evolutionarily labile specific traits might be. Such studies will also require additional phylogenetic information. Parasitoids arguably exhibit greater developmental variation than any other arthropodan group (Strand and Grbic 1997). As such, future comparative studies of the Hymenoptera offer a real opportunity for enhancing our understanding of insect developmental evolution.
Acknowledgements The work discussed here was conducted with support from the National Science Foundation. I also wish to thank M. Grbic for all of his input and important contributions to the work summarised in this brief review.
References Buss, L. W. (1987) The Evolution of Individuality. Princeton University Press, Princeton, New Jersey. Dawes, R., Dawson, I. Falciani, F., Tear, G. & Akam, M. (1994) Dax, a locust Hox gene related to fushi-tarazu but showing no pair-rule expression. Development 120: 1561-1572. del Pino, E. M. & Ellinson, R. P. (1983) A novel development pattern for frogs: Gastrulation produces an embryonic disc. Nature 306: 589-591. Dowton, M. & Austin, A. D. (1994) Molecular phylogeny of the insect order Hymenoptera: apocritan relationships. Proceedings of the National Academy of Sciences, USA 91: 9911-9915. Frasch, M., Hoey, T., Rushlow, C., Doyle, M. & Levine, M. (1987) Characterization and localisation of even-skipped protein of Drosophila. EMBO 6: 749-759. Gould, S. J. (1977) Ontogeny and Phylogeny. Belknap Press, Cambridge, Massachusetts. Grbic, M., Nagy, L. M., Carroll, S. B. & Strand, M. (1996a) Polyembryonic development: insect pattern formation in a cellularized environment. Development 12: 795-804.
M. R. Strand 16
Hymenoptera: Evolution, Biodiversity and Biological Control
Grbic, M., Nagy L. M.& Strand, M. (1996b) Pattern duplication in the polyembryonic wasp Copidosoma floridanum. Development, Genes and Evolution 206: 281-287. Grbic, M., Nagy L. M. & Strand M. R. (1998) Development of polyembryonic insects: a major departure from typical insect embryogenesis. Development, Genes and Evolution 208: 69-81. Grbic, M. & Strand, M. R. (1998) Shifts in the life history of parasitic wasps correlate with pronounced alterations in early development. Proceedings of the National Academy of Sciences, USA 95: 1097-1101. Hanson, P. E. & Gauld, I. D. (1995) The Hymenoptera of Costa Rica. Oxford University Press, Oxford. Ingham, P. W. (1988) The molecular genetics of embryonic pattern formation in Drosophila. Nature 335: 25-33. Jeffery, W. R. & Swalla, B. J. (1991) An evolutionary change in the muscle lineage of an anural ascidian embryo is restored by the interspecific hybridization with a urodele ascidian. Developmental Biology 145: 328-337. MacDonald, P., Ingham, P. & Struhl, G. (1986) Isolation, structure, and expression of evenskipped: a second pair-rule gene of Drosophila containing a homeobox. Cell 47: 721-734. Patel, N. H., Ball, E. & Goodman, C. S. (1992) Changing role of even-skipped during the evolution of insect pattern formation. Nature 357: 339-342. Raff, R. (1992) Direct-developing sea urchins and the evolutionary reorganization of early development. BioEssays 14: 211-218. Rasnitsyn, A. P. (1988) An outline of the evolution of the hymenopterous insects (Order Vespida). Oriental Insects 22: 115-145. Sanchez-Herrero, E., Vernos, I., Marco, R. & Morata, G. (1985) Genetic organisation of the Drosophila bithorax complex. Nature 313: 108-113. Sander, K. (1983) The evolution of patterning mechanisms: gleanings from insect embryogenesis and spermatogenesis. pp. 137-159. In Goodwin, B. P. (Ed.), Development and Evolution. Cambridge University Press, Cambridge. Scott, L. B., Lennarz, W. J., Raff, R. A. & Wray, G. A. (1990) The “lecitotrophic” sea urchin Heliocidaris erytrogama lacks typical yolk platelets and yolk proteins. Developmental Biology 138: 188-193. St Johnston, D. & Nüsslein-Volhard, C. (1992) The origin of pattern and polarity in the Drosophila embryo. Cell 68: 201-219. Strand, M. R. (1999) Developmental traits and life history evolution in parasitoids. pp. 139–162. In Hochberg, M. & Ives, A. R. (Eds), Parasitoid Population Biology. Princeton University Press, Princeton. Strand, M. R.& Grbic, M. (1997) The development and evolution of polyembryonic insects. Current Topics in Developmental Biology 35: 121-158. Tautz, D. & Sommer R. (1995) Evolution of the segmentation genes in insects. Trends in Genetics 1: 23-27. Thomson, K. S. (1988) Morphogenesis and Evolution. Cambridge University Press, Cambridge. Whitfield, J. B. (1998) Phylogeny and evolution of host-parasitoid interactions in Hymenoptera. Annual Review of Entomology 43: 129-151. Wray, G. A. (1995) Punctuated evolution of embryos. Science 267: 1115-1116. Wray, G. A. & Bely, A. E. (1994) The evolution of echinoderms is driven by several distinct factors. Development Supplement 97-106. Wray, G. A. & Raff, R. A. (1990) Novel origins of lineage founder cells in the direct developing sea urchin Heliocidaris erytrogama. Developmental Biology 141: 41-54.
Molecular Basis of Developmental Alteration in Heliothis virescens (F.) Larvae Parasitised by Cardiochiles nigriceps Viereck Francesco Pennacchio1,3, Patrizia Falabella1, Paola Varricchio1, Rocco Sordetti1, S. Bradleigh Vinson2, Franco Graziani3 and Carla Malva3 1
Dipartimento di Biologia, Difesa e Biotecnologie Agro-Forestali, Università della Basilicata, via N. Sauro, 85 – 85100 Potenza, Italy (email:
[email protected]) 2
Department of Entomology, Texas A&M University, College Station, TX 77843-2475 USA
3
Istituto Internazionale di Genetica e Biofisica, C.N.R., via Marconi, 10–80125 Napoli, Italy
Introduction Cardiochiles nigriceps Viereck is an endophagous parasitoid of the larval stages of the tobacco budworm, Heliothis virescens (F.). Larval development and moulting of parasitised hosts are not inhibited until the last instar is attained. At that time, pupation failure and developmental arrest are observed, associated to a significant increase of host nutritional suitability for the developing parasitoid larva (Pennacchio et al. 1993). This developmental arrest of host mature larvae is induced by an alteration of ecdysone biosynthesis and metabolism. A reduced biosynthetic activity of prothoracic glands, which do not show any gross morphological degeneration (Tanaka & Vinson 1991), along with the conversion of 20-hydroxyecdysone to inactive polar derivatives have been reported (Pennacchio et al. 1994). This metabolic inactivation of 20hydroxyecdysone is partly mediated by teratocytes, cells deriving from the dissociation of the serosal membrane, which grow in size without undergoing division (Pennacchio et al. 1994). Prothoracic glands of host mature larvae are inactivated by a parasitoid-induced disruption of PTTH (prothoracicotropic hormone) signal transduction pathway (Pennacchio et al. 1997, 1998a). More precisely, an evident underphosphorylation of key regulatory proteins, such as ribosomal S6 and ß-tubulin, is observed in response to PTTH stimulation. This probably induces a reduced rate of general protein synthesis and a substantial depression of the cytoskeletonmediated transport of ecdysone precursors and intermediate metabolites (Pennacchio et al. 1997, 1998a). The major host regulatory factors actively interfering with ecdysone biosynthesis in host prothoracic glands are calyx fluid and venom of the parasitoid (Tanaka & Vinson 1991). More recently, indirect experimental evidence indicates that trascriptionally active C. nigriceps polydnavirus (CnPDV) is required to induce host prothoracic gland inactivation in vitro (Pennacchio et al. 1998b). In fact, trioxsalen addition to CnPDV preparations followed by UV irradiation, eliminated the negative effect of the virus on ecdysteroidogenesis, registered when pupally committed prothoracic glands explanted from non-parasitised larvae, are co-incubated in vitro along with CnPDV and venom. 17
F. Pennacchio et al. 18
Hymenoptera: Evolution, Biodiversity and Biological Control
In order to isolate the parasitoid-viral genes involved in the inactivation of host prothoracic glands, a molecular characterisation of CnPDV genome and of the expressed sequences in parasitised host larvae has been undertaken. The most recent molecular data are briefly summarised in this note, along with functional information on the possible mechanism of action of the CnPDV genes expressed in the host and targeting prothoracic glands of the tobacco budworm larvae.
CnPDV Molecular Characterisation and Analysis of the Genes Expressed in Parasitised Hosts Free viral particles were isolated from ovaries of C. nigriceps and purified on a sucrose gradient. Negatively stained nucleocapsids showed evident ‘end structures’ and a membranaceous tail-like appendage. Size fractionation of DNA extracted from viral particles indicated the occurrence of circular dsDNA molecules, approximately ranging from 3 kb to 30 kb. This genome segmentation, typical of all Polydnaviridae, is associated to non-equimolar ratios among the different DNA segments, which are integrated as provirus in the genomic DNA of both females and males and transmitted vertically through the germ line (Varricchio et al. 1999). Northern blot studies were conducted on total or polyA+ RNA extracted from H. virescens larvae, on day 4 of 5th instars, at different time intervals (12, 24 and 48 h) after parasitoid oviposition, using CnPDV genomic DNA as a probe. This experiment indicated the occurrence of 6-8 polyadenylated main transcripts of viral origin. These transcripts were present from 12 to 48 h after parasitoid oviposition, at nearly constant levels (Varricchio et al. 1999). In order to clone viral genes expressed in parasitised hosts, two libraries of CnPDV genome were prepared and screened in reverse Northern, using as probe a labelled cDNA preparation, obtained by using polyA+ RNA extracted from H. virescens larvae, 24 h after C. nigriceps parasitisation. Various genomic viral fragments were identified and further analysed. These fragments, when used as probes in Southern blot experiments, hybridised to different undigested CnPDV DNA molecules and to various HindIII and EcoRI restriction fragments of the digested viral DNA and of the digested genomic DNA, extracted from both females and males of C. nigriceps (Varricchio et al. 1999). A cDNA library was prepared by using polyA+ mRNAs extracted from parasitised host larvae and screened using the isolated viral genomic clones as probes. Several positive cDNA clones were isolated and sequenced. The sequence analysis of two of these cDNAs and of the corresponding genomic sequences, allowed us to define the structure of the isolated CnPDV genes, expressed in parasitised hosts. These genes have introns and short open reading frames coding for small putative translation products with no substantial similarity with known proteins (Varricchio et al. 1999; unpublished). The cDNAs were used as digoxigenin-labelled probe for in situ hybridisation experiments on prothoracic glands explanted from H. virescens larvae, on day 4 of final (5th) instars, 24 h after they were parasitised by C. nigriceps. Hybridisation signals were clearly evident in the outer cytoplasmic layer of prothoracic glands explanted from parasitised hosts, while no signal was detected in glands obtained from non-parasitised controls (Varricchio et al. 1999).
Alteration of PTTH Signal Transduction Pathway The PTTH signal transduction pathway in prothoracic glands of H. virescens last instar larvae parasitised by C. nigriceps is deeply impaired, due to titer reduction of proteins involved in ecdysteroid biosynthesis, associated with a lowered efficiency of cytoskeleton-mediated trans-
Molecular Basis of Developmental Alteration in Heliothis virescens (F.) 19
port of ecdysone precursors and intermediate metabolites (Pennacchio et al. 1997, 1998a). These combined alterations presumably derive from the depressed phosphorylation rate of key regulatory proteins, as described in detail above. To better define the molecular mechanisms of the parasitoid-induced alteration of the PTTH signal transduction pathway, we designed an experiment aiming at evaluating the effect, if any, of the parasitism on the level of S6 kinase activity in prothoracic glands. Cytosolic extracts of prothoracic glands explanted from parasitised host last instar larvae and from synchronous nonparasitised controls were fractionated by HPLC on an anion exchange column (Mono Q HR). S6 kinase activity of each chromatographic fraction was assessed in vitro, using as a substrate a synthetic peptide (S6-21 : AKRRRLSSLRASTSKSESSQK), corresponding to the structural domain of ribosomal S6 phosphorylated at multiple serine sites (Brandon & Masaracchia 1991), and [32P] ATP as phosphate donor. There were no substantial differences in the level of S6 kinase activity in the cytosolic extracts obtained from prothoracic glands explanted from parasitised host larvae compared to those from synchronous non-parasitised controls (Falabella et al. unpublished). This experimental evidence suggests that a possible inhibition of cAMP dependent protein kinases may be active in intact cells.
Discussion and Conclusions CnPDV typically shows a segmented genome and the circular dsDNA segments are integrated as provirus in the wasp’s genome and transmitted vertically to the progeny, through the germ line. The number of viral genes expressed in parasitised hosts is lower to that reported for the ichneumonid Campoletis sonorensis (Cameron) (Blissard et al. 1986). These genes are persistently expressed over time, at least for the intervals considered. If we take into account the number of expressed genes compared to the aggregate genome size, which is around 100 kb, and the fact that the viral genomic clones characterised hybridise with different DNA segments, it is reasonable to conclude that a remarkable genome redundancy may occur. Further study is required to assess whether each superhelix is a mosaic of unique and shared sequence, and if gene families do also occur. The gene duplication in polydnaviruses can be an important evolutionary strategy that allows the accumulation of multiple copies of important functional genes (Fleming & Krell 1993). It is reasonable to consider genome segmentation, repeated sequences, segment nesting (Xu & Stoltz 1993; Cui & Webb 1997; Webb & Cui 1998) and non-equimolar segment ratios as different ways adopted by the virus to abundantly express genes that may have relevant functional roles in parasitised hosts, where virus replication does not occur. CnPDV actively regulates the physiology of H. virescens host larvae, inducing several alterations of both immune and endocrine systems. The molecular mechanisms of these pathological syndromes are poorly investigated, not only in the experimental system considered in the present study. However, the immune suppression mechanisms and the polydnavirus genes involved are by far more deeply investigated in various model systems, with numerous recent interesting studies (Strand & Pech 1995a, 1995b; Asgari et al. 1996, 1998; Yamanaka et al. 1996; Cui et al. 1997; Strand et al. 1997). The work reported here on neuroendocrine host alterations induced by the braconid C. nigriceps is probably the first attempt at a functional genetics approach to the study of the mechanisms regulating the developmental disruption, often observed in parasitised lepidopteran larvae. As visually summarised in Figure 1, the PTTH signal transduction pathway in prothoracic glands of H. virescens last instar larvae is interrupted, due to reduced phosphorylation of regulatory
F. Pennacchio et al. 20
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Proposed model for PTTH signal transduction pathway in prothoracic glands of mature Manduca sexta larvae (after Gilbert et al. 1988), which also applies to Heliothis virescens last (5th) instar larvae. Parasitism by Cardiochiles nigriceps interferes with the steps downstream the phosphorylation of key regulatory proteins controlling the protein synthesis rate and cytoskeleton-mediated transport of ecdysone precursors and intermediate metabolites. Cardiochiles nigriceps polydnavirus seems to exert a negative impact on the cAMP dependent kinase activity in prothoracic glands of host larvae.
proteins. Based on the functional and molecular studies presented here, we may hypothesise that some of the polydnavirus genes we have isolated and characterised could be involved in an inactivation mechanism of cAMP dependent protein kinases. However, further studies are required in order to more directly assess the role of the viral gene products expressed in prothoracic glands of parasitised insects. Research efforts are currently focused on the development of a suitable protocol for expressing foreign genes in prothoracic gland. This will allow us to assess more directly their role in determining the molecular alterations observed in prothoracic glands of naturally parasitised host mature larvae. It is very likely that CnPDV genes precociously expressed in parasitised tobacco budworm larvae may also play a role in the suppression of host immune response. If this is the case, then it will be interesting to compare the effects of the same genes, or even similar genes belonging to gene families, on different cell types. If a common mechanism is involved, we should expect that a conserved and basic metabolic pathway is probably affected. Such an approach could shed some light on possible common key strategies adopted by polydnaviruses for interfering with many different target tissues, eliciting an apparently unrelated array of symptoms, which concur in defining complex parasitism-induced syndrome.
Acknowledgements The work reported in this paper has been financially supported by research grants provided by M.U.R.S.T. (National Project SIDiVVAM) and EU-Regione Basilicata (P.O.P. – FESR: Metodi non convenzionali di difesa di colture di interesse agrario per la Regione Basilicata).
Molecular Basis of Developmental Alteration in Heliothis virescens (F.) 21
References Asgari, S., Hellers, M. & Schmidt, O. (1996) Host haemocyte inactivation by an insect parasitoid: transient expression of a polydnavirus gene. Journal of General Virology 77: 2653-2662. Asgari, S., Theopold, U., Wellby, C. & Schmidt, O. (1998) A protein with protective properties against the cellular defense reactions in insects. Proceedings of the National Academy of Sciences, USA 95: 3690-3695. Blissard, G. W., Fleming , J. G. W., Vinson, S. B. & Summers, M. D. (1986) Campoletis sonorensis virus: expression in Heliothis virescens and identification of expressed sequences. Journal of Insect Physiology 32: 351-359. Brandon, S. D. & Masaracchia, R. A. (1991) Multisite phosphorylation of a synthetic peptide derived from the carboxyl terminus of the ribosomal protein S6. Journal of Biological Chemistry 266: 380-385. Cui, L. & Webb, B. A. (1997) Homologous sequences in the Campoletis sonorensis polydnavirus genome are implicated in replication and nesting of the W segment family. Journal of Virology 71: 8504-8513. Cui, L., Soldevila, A. & Webb, B. A. (1997) Expression and haemocyte-targeting of a Campoletis sonorensis polydnavirus cysteine-rich gene in Heliothis virescens larvae. Archives of Insect Biochemistry & Physiology 36: 251-271. Fleming, J. G. W. & Krell, P. J. (1993) Polydnavirus genome organization. pp. 189-225. In Beckage N. E., Thompson S. N. & Federici B. A. (Eds), Parasites and Pathogens of Insects, Vol. 1 Parasites. Academic Press, New York. Gilbert, L. I., Combest, W. L., Smith, W. A., Meller, V. H. and Rountree, D. B. (1988). Neuropeptides, second messangers and insect molting. BioEssay 8: 153-157. Pennacchio, F., Vinson, S. B. & Tremblay, E. (1993) Growth and development of Cardiochiles nigriceps Viereck (Hymenoptera, Braconidae) larvae and their synchronization with some changes of the hemolymph composition of their host, Heliothis virescens (F.) (Lepidoptera, Noctuidae). Archives of Insect Biochemistry & Physiology 24: 65-77. Pennacchio, F., Vinson, S. B., Tremblay, E. & Ostuni, A. (1994) Alteration of ecdysone metabolism in Heliothis virescens (F.) (Lepidoptera, Noctuidae) larvae induced by Cardiochiles nigriceps Viereck (Hymenoptera, Braconidae) teratocytes. Insect Biochemistry & Molecular Biology 24: 383-394. Pennacchio, F., Sordetti, R., Falabella, P. & Vinson, S. B. (1997) Biochemical and ultrastructural alterations in prothoracic glands of Heliothis virescens (F.) (Lepidoptera: Noctuidae) last instar larvae parasitised by Cardiochiles nigriceps Viereck (Hymenoptera: Braconidae). Insect Biochemistry & Molecular Biology 27: 439-450. Pennacchio, F., Falabella, P., Sordetti, R., Varricchio P., Malva C. & Vinson, S. B. (1998a) Prothoracic gland inactivation in Heliothis virescens (F.) (Lepidoptera: Noctuidae) larvae parasitised by Cardiochiles nigriceps Viereck (Hymenoptera: Braconidae). Journal of Insect Physiology 44: 845-857. Pennacchio, F., Falabella, P. & Vinson, S. B. (1998b) Regulation of Heliothis virescens prothoracic glands by Cardiochiles nigriceps polydnavirus. Archives of Insect Biochemistry & Physiology 38: 1-10. Strand, M. R. & Pech, L. L. (1995a) Immunological basis for compatibility in parasitoid-host relationships. Annual Review of Entomology 40: 31-56. Strand, M. R. & Pech, L. L. (1995b) Microplitis demolitor polydnavirus induces apoptosis of a specific haemocyte morphotype in Pseudoplusia includens. Journal of General Virology 76: 283-291. Strand, M. R., Witherell, R. A. & Trudeau, D. (1997) Two Microplitis demolitor polydnavirus mRNAs expressed in hemocytes of Pseudoplusia includens contain a common cysteine-rich domain. Journal of Virology 71: 2146-2156.
F. Pennacchio et al. 22
Hymenoptera: Evolution, Biodiversity and Biological Control
Tanaka, T. & Vinson, S. B. (1991) Depression of prothoracic gland activity of Heliothis virescens by venom and calyx fluids from the parasitoid, Cardiochiles nigriceps. Journal of Insect Physiology 37: 139-144. Varricchio, P., Sordetti, R., Falabella P., Graziani, F., Malva, C. & Pennacchio, F. (1999) Cardiochiles nigriceps polydnavirus: molecular characterisation and expression in parasitised Heliothis virescens larvae. Insect Biochemistry & Molecular Biology 29: 1087-1096. Webb, B. A. & Cui, L. (1998) Relationship between polydnavirus genomes and viral gene expression. Journal of Insect Physiology 44: 785-793. Xu, D. & Stoltz, D. B. (1993) Polydnavirus genome segment families in the ichneumonid parasitoid Hyposoter fugitivus. Journal of Virology 67: 1340-1349. Yamanaka, A., Hayakawa, Y., Noda, H., Nakashima, N. & Watanabe, H. (1996) Characterisation of polydnavirus-encoded mRNA in parasitised armyworm larvae. Insect Biochemistry & Molecular Biology 26: 529-536.
Protection by Immune Disguise: a New Lesson From a Parasitoid Wasp Ulrich Theopold*, Dongmei Li, Wanja Kinuthia and Otto Schmidt Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) (*present address: Department of Molecular Biology, Stockholm University SE-10691 Stockholm, Sweden)
Introduction Immune protection of parasitoid wasps has been ascribed to a number of mechanisms, which include systemic mechanisms like suppression of the host’s cellular immune response and localised surface protection like egg coating with wasp maternal protein secretions. In this paper, we review recent work describing a novel mechanism of immune protection employed by the parasitoid wasp Venturia canescens Gravenhorst, which we call ‘immune disguise’. During immune disguise, the parasitoid is covered with a specific moiety of host hemolymph components thus gaining protection from further attack by the host, which fails to recognise the components as foreign. The sequestration of hemolymph components by the parasitoid shows similarities with insect hemolymph coagulation.
Proteolytic Cascades Involved in Innate Immunity The innate immune system of vertebrates comprises a number of proteolytic activation cascades, such as the complement system and the coagulation cascade, which help to establish a first line of defence against intruding micro-organisms (Cerenius & Soederhaell 1995). Searching for similar systems in arthropods, two proteolytic cascades have been identified, the prophenoloxidase-activating cascade (PPO cascade) and the coagulation cascade in the horseshoe crab, Limulus (Iwanaga et al. 1998). Both cascades are activated by molecules of microbial origin, including lipopolysaccharide, laminarin and proteoglycan. Although the PPO cascade is mostly known for its capacity to produce microbicidal compounds, it is also involved in several other reactions of the invertebrate immune system including cellular reactions (Cerenius & Soederhaell 1995). Accordingly, a cell-surface form of prophenoloxidase has been identified (Charalambidis et al. 1996). In addition to proteolytic activities, a number of protease inhibitors, which act at different levels of the PPO cascade have been isolated from hemolymph (Cerenius & Soederhaell 1995). It can be proposed that these inhibitors are involved in regulating different steps of the PPO cascade. In this scenario, induction of one type of immune response may occur at the expense of other possible responses. One such decision is between cellular and humoral immune reactions. In contrast to Limulus, little is known about the coagulation reaction of insect hemolymph, which is usually regarded as part of the hemostatic wound response. Nevertheless, just as in Limulus, microbial elicitors also seem to enhance coagulation of insect hemolymph (Duvic & Brehelin 1998). It has also been shown that the resulting coagulum acts as a trap for bacteria, which bind to fibrillar structures formed by hemocyte degranulation (reviewed by Ratcliffe & Rowley 1979). 23
Ulrich Theopold, Dongmei Li, Wanja Kinuthia and Otto Schmidt 24
Hymenoptera: Evolution, Biodiversity and Biological Control
The Coagulation of Insect Hemolymph Clotting in vertebrates is known to depend on the interaction between cellular components, mostly derived from platelets and soluble serum factors, the clotting factors. The factors that form the coagulum exist in an inactive form, called coagulogen and are activated through proteolytic cascades composed mostly of serine proteases. The mobilisation of the cellular coagulogen involves the release of specialised so-called alpha-granules and the formation of microparticles, small membrane vesicles, which help to attract clotting factors and aid in their activation (Dachary-Prigent et al. 1996). As mentioned above, most of our understanding on coagulation reactions in arthropods stems from studies of crustaceans and arachnids where, in some species, most factors involved in hemolymph coagulation have been identified and characterised at a molecular level (Iwanaga et al. 1998). Only a few proteins share complete sequence homology with vertebrate proteins involved in proteolytic cascades. Nevertheless, a number of domains appear to be evolutionarily conserved including serine protease- and epidermal growth factor-like domains, C-type lectinand defensin-like domains, and domains found in a number of complement proteins (Muta & Iwanaga 1996). Hemolymph clotting in insects, like in vertebrates, involves the interaction between cellular and humoral pro-coagulant activity (Bohn 1986). In several species, the humoral pro-coagulant has been identified as lipophorin, a multifunctional abundant hemolymph protein (reviewed by Bohn 1986). This was recently confirmed in locusts where coagulation could be induced by the addition of laminarin, a component of yeast cell walls, to hemolymph (Duvic & Brehelin 1998). The precipitate contained protein bands of 265 and 80 kDa, indicative of apolipohorin I and II. In that regard, insects show similarities to crustaceans, where the major protein recruited by hemolymph clotting is also a lipoprotein (Hall et al. 1995). Interestingly, a proteolytic cascade with a function in insect development, encoded by the genes of the dorso-ventral group, shows striking similarities to coagulation cascades (Muta & Iwanaga 1996). Several members of the dorso-ventral cascade are serine proteases, which are involved in the localised formation of a ligand for the toll receptor. It is tempting to speculate that this cascade might be also multi-functional like the intracellular signal transduction cascade encoded by other members of the dorso-ventral group, which are used both in development and immunity (Lemaitre et al. 1996).
Microparticles as part of the cellular coagulogen in insects We have previously shown that insect hemolymph coagulation leads to the formation of vesicles which show a number of similarities to vertebrate microparticles (Theopold & Schmidt 1997): 1) The formation of microparticles is calcium-dependent, possibly mediated by calpain, a calcium-dependent protease, which we previously identified in insects (Theopold et al. 1995). 2) Microparticles are released from the cell surface through fragmentation of filopodia, and 3) Microparticles expose negatively charged phospholipids on their surface that can be visualised by staining with annexin V (Theopold & Schmidt 1997). Due to the similarities between vertebrate microparticles and the membrane vesicles we could identify in insect, we decided to name the vesicles ‘insect microparticles’ (Theopold & Schmidt 1997).
Protection by Immune Disguise: a New Lesson From a Parasitoid Wasp 25
A Lipophorin
crosslinking activity
Hemomucin
MP surface
B Lipophorin
Hemomucin
crosslinking activity
Figure 1
egg surface
A) Hypothetical scheme of insect hemolymph coagulation involving the surface of microparticles (MP) and the interaction between hemomucin and lipophorin; B) scheme of the formation of a similar complex on the surface of Venturia eggs and larvae (for further explanations see text).
In addition to annexin V, insect microparticles from a number of species can be stained with lectins, which are specific for N-Acetyl Galactosamine (GalNAc) like Helix pomatia L. agglutinin (H.p. lectin). A protein that is labelled by H.p. lectin could be identified in a lepidopteran species (Galleria mellonella L.) and in Drosophila melanogaster Meigen. Using a hemocyte-like cell line (mbn-2 cells), we were able to purify the protein and isolate the corresponding cDNA for the Drosophila protein (Theopold et al. 1996). After its source of isolation and because it contains a typical mucin domain, the protein was named ‘hemomucin’. Pointing towards a possible role in coagulation, hemomucin could be shown to attract lipophorin, the previously identified humoral coagulant, from hemolymph (Theopold & Schmidt 1997). Figure 1A shows a hypothetical model for the role hemomucin plays as a component of insect microparticles in hemolymph
Ulrich Theopold, Dongmei Li, Wanja Kinuthia and Otto Schmidt 26
Hymenoptera: Evolution, Biodiversity and Biological Control
A
B
Figure 2
A) Venturia larva dissected from the chorion and stained with H.p. lectin. The first signs of staining occur at segmental borders, as shown in this larva. At a later stage, the stripe pattern is replaced by an even surface staining as described for the larva at the time of hatching from the egg shell (Kinuthia et al. 1999); B) same section as in (A) but in phase contrast.
clotting. The cross-linking activity needed for the formation of an insoluble coagulum is most likely phenoloxidase. In this context, it is interesting to note that hemomucin shows significant sequence similarity to a class of plant enzymes that are involved in the formation of alkaloid precursors by catalysing a condensation reaction between a monoterpene and an indole ring (Theopold et al. 1996). It can be speculated that hemomucin might play a role in the crosslinking reaction by attracting tyrosine derivatives which share some structural similarity with the substrates of the plant enzyme and are known as substrates for phenoloxidase.
Protection by Immune Disguise: a New Lesson From a Parasitoid Wasp 27
A Hemomucin-like Protein in a Parasitoid Wasp In addition to hemocytes and the lining of the gut, Drosophila hemomucin was also found on the surface of the egg (Theopold et al. 1996). Since mucopolysaccharides on the egg surface were implicated in immune protection (Fuehrer 1972), we propose that this expression pattern might constitute a pre-adaptation for the parasitic lifestyle. By expressing a phylogenetically conserved protein, which is also part of the immune system, on their surface, parasitoid eggs might be protected from attack by their host’s immune system. We were indeed able to identify a possible hemomucin homologue in the parasitoid wasp V. canescens (Kinuthia et al. 1999). Staining with GalNAc-specific lectins showed a pattern similar to Drosophila with strong staining of hemocytes including microparticles, the egg surface and the larval surface. In addition, similar to purified hemomucin, eggs prepared from the oviduct were able to attract lipophorin after incubation with Ephestia hemolymph (Kinuthia et al. 1999). We are thus faced with the paradoxical situation that the freshly deposited parasitoid egg uses part of the host’s humoral immune response to avoid the cellular response. The important difference between a normal coagulation reaction and the reaction on the egg surface may be the limited nature of the latter. A limited coagulation reaction, which involves the parasitoid’s hemomucin and host lipophorin, seems to protect the egg from any further cellular attack. Since we could also see H.p. lectin staining on larval surfaces and identify hemomucin in larval extracts (Kinuthia et al. 1999) (Figs 2A, 2B), the same mechanisms could explain how the larva is protected once it emerges from the eggshell. It has indeed been shown before that lipophorin inhibits hemocyte attachment (Coodin & Caveney 1992; Mandato et al. 1996) and, more recently, that lipophorin complexes derived from a coagulation reaction have inhibitory activity on hemocyte phenoloxidase and on the regulatory protease which activates phenoloxidase (Duvic & Brehelin 1998). In our model, the protection conferred by lipophorin is thus mediated through hemomucin, which is both present on the egg surface and on surfaces that are exposed to the immune system (Fig. 1B). It is possible that one of the protective proteoglycans, which were described earlier in V. canescens (Fuehrer 1972) is in fact identical with hemomucin. The crosslinking activity on the egg surface could be an ovary-specific phenoloxidase or peroxidase, activities which have both been detected in mosquitoes (Li et al. 1996).
Acknowledgements This work was supported by a grant from the Australian Research Council. U.T. is supported by an ARC Research Fellowship.
References Bohn, H. (1986) Hemolymph clotting in insects. pp. 189-207, In Brehelin, M. (Ed.), Immunity in Invertebrates. Springer-Verlag, Heidelberg. Cerenius, L. & Soederhaell, K. (1995) Crustacean immunity and complement; a premature comparison? American Zoologist 35: 60-67. Charalambidis, N. D., Foukas, L. C., Zervas, C. G. & Marmaras, V. J. (1996) Hemocyte surface phenoloxidase (PO) and immune response to lipopolysaccharide (LPS) in Ceratitis capitata. Insect Biochemistry & Molecular Biology 26: 867-874. Coodin, S. & Caveney, S. (1992) Lipophorin inhibits the adhesion of cockroach (Periplaneta americana) haemocytes in vitro. Journal of Insect Physiology 38: 853-862.
Ulrich Theopold, Dongmei Li, Wanja Kinuthia and Otto Schmidt 28
Hymenoptera: Evolution, Biodiversity and Biological Control
Dachary-Prigent, J., Toti, F., Satta, N., Pasquet, J. M., Uzan, A. & Freyssinet, J. M. (1996) Physiopathological significance of catalytic phospholipids in the generation of thrombin. Seminars in Thrombosis & Hemostasis 22: 157-164. Duvic, B. & Brehelin, M. (1998) Two major proteins from locust plasma are involved in coagulation and are specifically precipitated by laminarin, a beta-1,3-glucan. Insect Biochemistry & Molecular Biology 28: 959-967. Fuehrer, E. (1972) Mucopolysaccharide im weiblichen Geschlechtsapparat parasitischer Hymenopteren. Naturwissenschaften 59: 167-168. Hall, M., van Heusden, M. C. & Soederhaell, K. (1995) Identification of the major lipoproteins in crayfish hemolymph as proteins involved in immune recognition and clotting. Biochemical & Biophysical Research Communications 216: 939-946. Iwanaga, S., Kawabata, S. & Muta, T. (1998) New types of factors and defence molecules found in horseshoe crab hemolymph – their structures and functions. Journal of Biochemistry 123: 1-15. Kinuthia, W. , Li, D., Schmidt, O. & Theopold, U. (1999) Is the surface of endoparasitic wasp eggs and larvae covered by a limited coagulation reaction? Journal of Insect Physiology 45: 501-506. Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J.-M. & Hoffmann, J. A. (1996) The dorsoventral regulatory gene cassette spaetzle/toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86: 973-983. Li, J., Hodgeman, B. A. & Christensen, B. M. (1996) Involvement of peroxidase in chorion hardening in Aedes aegypti. Insect Biochemistry & Molecular Biology 26: 309-317. Mandato, C. A., Diehljones, W. L. & Downer, R. G. H. (1996) Insect hemocyte adhesion in vitro – inhibition by apolipophorin I and an artificial substrate. Journal of Insect Physiology 42: 143-148. Muta, T. & Iwanaga, S. (1996) The role of hemolymph coagulation in innate immunity. Current Opinion in Immunology 8: 41-47. Ratcliffe, N. A. & Rowley, A. F. (1979) Role of hemocytes in defence against biological agents. pp. 331-414. In Gupta, A. P. (Ed.), Insect Hemocytes. Cambridge University Press, Cambridge. Theopold, U., Pinter, M., Daffre, S., Tryselius, Y., Friedrich, P., Naessel, D. & Hultmark, D. (1995) CalpA, a Drosophila calpain homolog specifically expressed in a small set of nerve, midgut and blood cells. Molecular & Cellular Biology 15: 824-834. Theopold, U., Samakovlis, C., Erdjument-Bromage, H., Dillon, N., Axelsson, B., Schmidt, O., Tempst, P. & Hultmark, D. (1996) Helix pomatia lectin, an inducer of Drosophila immune response binds to hemomucin, a novel surface mucin. Journal of Biological Chemistry 271: 12708-12715. Theopold, U. & Schmidt, O. (1997) Helix pomatia lectin and annexin V, two molecular probes for insect microparticles: possible involvement in hemolymph coagulation. Journal of Insect Physiology 43: 667-674.
Host Defence Manipulation by Parasitoid Wasps and the Problem of Assessing Host Specificity Otto Schmidt, Sassan Asgari, Markus Beck* and Ulrich Theopold** Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) (present addresses: *Department of Entomology, University of Wisconsin-Madison, Madison, WI 53706 USA; **Department of Molecular Biology, Stockholm University SE-10691 Stockholm, Sweden)
Introduction Hymenopteran parasitoids are increasingly used as biocontrol agents due to a seemingly inexhaustible source of wasp species known to parasitise a wide range of arthropod pests. However, the potential to reduce insect pest numbers in the field is not always realised in practice. In biological control, as in many complex experimental situations, only positive outcomes are likely to reveal testable information, which is why little is known about specific reasons of control failures. Many steps in the introduction of biological control agents are indeed a black box (Waage 1998), including some fundamental questions, like the genetic diversity and ecology of the pest insect which may not correspond to adaptive features of the introduced wasp (see Beck et al. this volume), the rearing process which may have altered essential wasp properties and the release and establishment procedures which may not be appropriate in the field (Hopper & Roush 1993). Although it was long assumed to be a main factor, the application of sophisticated diagnostic tools increasingly demonstrates the importance of genetic variation in both target pests and natural enemies. Genetic variation and genetic diversity are closely associated with important features of parasitoid host interactions, such as host adaptations and specificity. The problem is that some of the most fundamental physiological processes of how insect endoparasitoids interact with the host are not known at the molecular level. This includes two basic questions: 1) how insects recognise and react to foreign objects and distinguish them from self, and 2) how some parasites and pathogens are easily able to overcome the host defence. Recent observations on the cellular induction of humoral defence molecules revealed a cytoplasmatic signalling cascade which resembles those in vertebrates (Hultmark 1994a). In Drosophila melanogaster Meigen the transcription factors involved in the activation of immune genes belong to the rel family of gene products, involved in the mammalian acute response (Sun et al. 1991; Ip et al. 1993; Liou & Batimore 1993; Reichhart et al. 1993). This may be an indication that the intracellular regulation of extracellular immune signals has been highly conserved during evolution. Interestingly, the induction of dorso-ventral polarity in embryonic development is regulated in D. melanogaster by a cascade of cell surface (Hultmark 1994b) and transcription factors (Lemaitre et al. 1995), that are used in later stages again for the activation of defence molecules against fungal invasion (Lemaitre et al. 1996). From these and other studies it emerges that different micro-organisms elicit specific responses and that the intracellular regulation of immune defence molecules is partitioned in separate regulatory cascades. Unfortunately the progress of unravelling the intracellular regulation of immune responses is not matched with the recognition
29
Otto Schmidt, Sassan Asgari, Markus Beck and Uli Theopold 30
Hymenoptera: Evolution, Biodiversity and Biological Control
The Campoletis/Heliothis model Egg
Larval stages
29-36 kDa ovarian proteins
VHv1.1 (cysteine-rich virus protein)
WHv1.0 (cysteine-rich virus protein)
oviposition Figure 1
Schematic depiction of gene activity during egg and larval stages of the parasitoid Campoletis sonorensis inside the host caterpillar Heliothis virescens. Bold lines indicate viral transcripts. Thin lines indicate the presence of proteins (produced from maternal transcripts in the female wasp ovary or from viral transcripts in infected host cells). In this system the immune suppression in the caterpillar is continuous throughout wasp development and mediated by wasp genes coding for a group of ovarian proteins 29-34 kDa in size (Webb & Luckhart 1994, 1996) and polydnavirus-specific gene families (VH and WH denotes different viral plasmids) coding for Cysteine-rich proteins (Dib-Haijj et al. 1993; Li & Webb 1994). The embryonic and larval contours are a schematic representation of the two developmental stages and do not represent accurate time frames.
and signalling at the extracellular level. It is still too early to ask the question, whether secreted components used in recognition and inactivation of foreign objects in insects are in any case related to mammalian recognition molecules. Although immunoglobulin-like molecules exist (Sun et al. 1990), insects lack immunoglobulins as specific recognition molecules and with it the clonal selection that allows to distinguish ‘self’ from ‘non-self’ as well as a memory of previous encounters with micro-organisms (Hoffmann et al. 1996). If at all, any similarities to vertebrate immunity may be at the level of innate immune responses (Hoffmann 1995), which include nonclassical pathways leading to coagulation (Theopold & Schmidt 1997), phagocytosis (Foukas et al. 1998), cell attachment (LanzMendoza et al. 1996), opsonisation (Schmidt et al. 1993) and release of defence molecules (Hoffmann & Hetru 1992). The naive assumption is that the identification of key reactions relevant to host immune defence and its suppression by parasitoids, will allow us to make assessments on parasitoid virulence and host specificity by looking at the relatedness of the corresponding genes in parasitoids and respective pests. However, a prerequisite for this assumption is that parasitoids use the same pathway in the host insect to interfere with defence reactions. Recent studies on the molecular mechanisms of immune evasion and suppression by parasitoids has dramatically changed our understanding of the complexity of insect immunity and the diversity of parasitoid adaptations to overcome host defence reactions. For example, we have identified two completely different mechanisms of immune evasion and suppression in Cotesia
Host Defence Manipulation by Parasitoid Wasps 31
rubecula (Marshall)/Pieris rapae L. and Venturia canescens Gravenhorst/Ephestia kuehniella Zeller, the two parasitoids studied in our laboratory. For historical and comparative reasons, we also include the Campoletis sonorensis (Cameron)/Heliothis virescens (F.) system for discussion. In addition there are a number of other systems where the mechanism of protection are known (Beckage 1995; Strand & Pech 1995). The putative molecular mechanisms of protection of the developing wasp in the three systems are summarised and the implications for possible predictions of host specificity and their evolutionary origins discussed.
The Campoletis/Heliothis Model The hallmark of host manipulation by hymenopteran parasitoids is the presence of polydnaviruses, which are only produced in female ovary tissues and coded by provirus DNA that is integrated into the wasp genome (Stoltz & Vinson 1979; Fleming & Summers 1986, 1991; Xu & Stoltz 1991). At least two types of polydnaviruses are known as ichnoviruses and bracoviruses, depending on whether the parasitoid species belongs to Ichneumonidae or particular subfamilies of Braconidae (Stoltz et al. 1984). The viruses are expressed in host cells within a few hours after parasitisation, producing specific virus products that inactivate the host defence system (Summers & Dib-Hajj 1995). The first experimental demonstration that polydnaviruses are able to suppress the host defence system was performed in C. sonorensis, where purified ichnovirus (CsV) were injected into caterpillars and shown to suppress the encapsulation reaction similar to the situation in parasitised caterpillars (Edson et al. 1981). Molecular characterisation of circular polydnavirus-coded genes revealed a cluster of genes producing a family of cysteine-rich proteins (Dib-Hajj et al. 1993), that are involved in the suppression of the cellular defence (Li & Webb 1994). In this system the protective mechanism during the crucial time immediately after egg deposition and onset of virus protein synthesis, which in C. sonorensis takes at least 5 h (Luckhart & Webb 1996), is accomplished by calyx tissue secretions including a group of virus encoded proteins with similarity to the cysteine-rich proteins that are deposited together with the egg (Luckhart & Webb 1996; Webb & Luckhart 1994, 1996). Thus, the suppression of cellular defence functions is achieved by virus encoded genes that are expressed in the wasp ovary as maternal secretions and in virus-infected host cells and tissues (Fig. 1). Virus-mediated suppression of the host defence may require more than one gene product possibly acting in combination (Webb & Luckhart 1994) and virus expression of a complex set of genes continues for weeks while the parasitoid larvae develops inside the caterpillar hemocoel.
The Cotesia/Pieris Model Compared to the situation in C. sonorensis, the expression of circular bracovirus DNA in C. rubecula (CrV) is restricted to only two transcripts in virus-infected host hemocytes and fat body cells, which are expressed as a single intense peak between four and 12 h after egg deposition (Asgari et al. 1996). One of the two gene products was identified as an immune suppressor (CrV1), secreted into the host hemolymph from infected tissues, where it is bound to hemocytes and inactivates the endocytoskeleton by an unknown mechanism (Asgari et al. 1997). These hemocytes remain round spheres and are unable to spread on a glass surface or phagocytose micro-organisms (Asgari 1997; Asgari et al. 1997), but restore their endoskeleton after several days and subsequently regain cellular defence functions (Asgari 1997). Thus, the CrV1-mediated immune suppression is transient and, importantly, the inactivation extends through a crucial period when the wasp larvae emerge from the egg shell (Fig. 2). Since the larval cuticle constitutes a foreign surface eliciting an
Otto Schmidt, Sassan Asgari, Markus Beck and Uli Theopold 32
Hymenoptera: Evolution, Biodiversity and Biological Control
The Cotesia/Pieris model Egg
Larval stages
Crp32 CrV1
Crp32-like host proteins?
................................................... oviposition Figure 2
Schematic depiction of gene activity during egg and larval stages of the parasitoid Cotesia rubecula inside the host caterpillar Pieris rapae. Bold lines indicate viral transcripts. Thin lines indicate proteins produced from maternal transcripts in the female wasp ovary (e.g. Crp32) or from viral transcripts in infected host cells (e.g. CrV1). Dotted lines indicate proteins of host origin (e.g. Crp32-like protein).
immune response, the absence of an effective cellular defence is crucial for the developing parasitoid. For the wasp, recovery of cellular host defence functions may provide a selective advantage, as it protects the parasitised caterpillar from possible hyperparasitoids and pathogens. However, the question remains, how is the larval cuticle protected against the restored host defence? A possible clue may come from a novel mechanism of surface protection employed by the parasitoid during the first hours after the egg deposition. We found that the C. rubecula egg was protected after injection into the host in the absence of polydnaviruses (Asgari & Schmidt 1994). Since the surface protection was removed after washing with a mild detergent, we concluded that the egg and possibly the viruses are covered by calyx fluid proteins that preclude encapsulation or phagocytosis by host hemocytes (Asgari et al. 1998). The observation, that antisera against purified CrVs cross-reacted with proteins on the egg surface, allowed us to identify possible candidates for protective proteins and the subsequent cloning of the coding DNA of one of these proteins (Crp32), using two independent approaches involving expression library screening as well as peptide micro-sequencing and PCR cloning (Asgari et al. 1998). The molecular mechanism of how Crp32 protects the surface remains to be uncovered. From protection experiments using recombinant Crp32 on sephadex beads it appears that the protein is protective at low surface densities, suggestive that the protein invokes local inhibition reactions rather than surface disguise. This protective feature may be relevant to future application programs using surface protective proteins in genetic improvement programs of biological control agents and biopesticides against lepidopteran pests. When Crp32-specific antibodies were used against non-parasitised caterpillar protein extracts a cross-reactive host protein of similar size was discovered in the hemolymph in low amounts
Host Defence Manipulation by Parasitoid Wasps 33
The Venturia/Ephestia model Egg
Larval stages
PI
chorion-specific
cuticle-specific mucins
lipophorin .. . ........................... .......... ..................................
oviposition Figure 3
Schematic depiction of gene activity during egg and larval stages of the parasitoid Venturia canescens inside the host caterpillar Ephestia kuehniella. Thin lines indicate proteins produced from maternal transcripts in the female wasp ovary (e.g. protease inhibitor, PI) or from embryonic or larval transcripts (e.g. mucin-like proteins). Dotted lines indicate proteins of host origin (e.g. lipophorin).
(Asgari et al. 1998). It is tempting to speculate that similar protective proteins exist in the insect hemocoel and that objects introduced into the hemolymph would eventually be covered if there is enough time to allow the protective protein to accumulate on the surface. Therefore, the suppression of the cellular host defence may provide the emerging larva enough time to accumulate the host-derived protective protein and would not be attacked after the hemocytes have recovered. The cloning of the host-specific Crp32-like protein will allow us to test this assumption.
The Venturia/Ephestia Model Although Venturia is closely related to C. sonorensis its mode of protection is different from any other known wasp system and has intrigued scientists for many decades (Salt 1938, 1964). One of the first peculiarities observed in this wasp is the presence of so-called virus-like particles (VLPs) which resemble ichnoviruses in particle morphology, particle assembly in calyx nuclei and mode of secretion, but they completely lack nucleic acids (Rotheram 1967; Bedwin 1979; Feddersen et al. 1986). Another early observation was the apparent lack of any visible inactivation of host hemocytes (Salt 1976, 1980), although a reduction of melanisation in cell-free hemolymph was known to occur in parasitised caterpillars. Another illustration that most host hemocytes remain intact is seen in cellular capsules formed around Venturia larvae in superparasitised caterpillars. Again, the fact that these capsules are not melanised confirms genetic evidence that melanisation and encapsulation are independent processes (Rizki & Rizki 1990). When Venturia eggs are treated with detergent the protection is lost but regained after incubation in VLP suspensions, suggestive that VLPs are protecting the egg surface (Feddersen et al. 1986; Schmidt & Schuchmann-Feddersen 1989). Given the mode of surface protection in C. rubecula, which covers both virus and egg, it is conceivable that the protective mechanisms in Venturia are
Otto Schmidt, Sassan Asgari, Markus Beck and Uli Theopold 34
Hymenoptera: Evolution, Biodiversity and Biological Control
not specific to VLPs but involve a complex reaction that includes maternal protein secretions and egg chorion-specific components as well (see Theopold et al. this volume). The main reason that the surface protection mechanism is different from the other two systems is probably the hydropic nature of the Venturia egg. After deposition into the host hemocoel, the egg takes up hemolymph from the host increasing in size for up to five times the original volume. Obviously this requires a protective strategy that is dramatically different from the other two wasp systems. The fact that cell-free host hemolymph enters the egg requires immediate inactivation of cellular and melanotic activity and, at the same time, a protective layer against host hemocyte attack without precluding cell-free hemolymph from entering the egg shell. We are still at an early stage of understanding the complex changes that occur on the surface of the Venturia egg, but two elements of a protective mechanism are becoming evident (Fig. 3). Firstly, there is a local hemocyte inactivation around the freshly deposited Venturia egg caused by a viscous calyx fluid which contains a protease inhibitor that inactivates hemocyte spreading (M. Beck unpublished). Since hemocytes are only inactivated by highly enriched inhibitor, the effect probably disappears when calyx fluid dissipates into the hemolymph. Secondly, hemomucin-like glycoproteins (Theopold et al. 1996) found on the Venturia egg chorion react with host lipophorin to form a complex (Theopold & Schmidt 1997), which may be cross-linked by a phenoloxidase-like activity found on the egg surface (Kinuthia et al. 1999). By the time hemocytes are recovered from the protease inhibition, the surface properties of the complex resembling a limited coagulation reaction, may have formed and the egg surface subsequently not recognised or attacked as a foreign object. Several features of the described Venturia egg surface modification are worthy of note: the suppression of host hemocytes is reduced to a short period, probably affecting only local populations of hemocytes around the deposited egg. Again, this is an evolutionary advantage to the parasitoid, since it leaves the parasitised caterpillar still capable of defending against other parasites and pathogens. In this context it is however important to understand how Venturia larvae are protected. Interestingly, the Venturia embryo is covered with hemomucin-like glycoproteins that are similar to those detected on the egg surface (Kinuthia et al. 1999). Since cell-free hemolymph enters the egg and surrounds the developing larva, it is conceivable that by the time the larva hatches these glycoproteins form complexes with lipophorin similar to those observed on the egg surface. Further experiments are required to establish the molecular basis of larval protection in Venturia.
Conclusions A number of conclusions can be drawn from these examples regarding the evolution and adaptive properties of host defence manipulation and host specificity of hymenopteran parasitoids. The first is the unexpected diversity of mechanisms found in closely related wasps. This is not only a reflection on the complexity of invertebrate immune defence reactions, but also illustrates an unexpected genetic flexibility on the side of the wasp to exploit and lock into different mechanisms of host immune regulation during adaptation to a particular host range. Can we explain the genetic flexibility by the presence of a symbiotic virus in the wasp genome? Some observations suggest a larger role of the virus in addition to just being a vehicle for delivering immune suppressor gene products into the host. For example, CrV1 is coded by the circular polydnavirus genome in C. rubecula, C. congregata, and a number of other Cotesia
Host Defence Manipulation by Parasitoid Wasps 35
species (see Whitfield this volume) and is therefore considered a virus product. However, CrV1like DNAs are not detected in other closely related Cotesia spp. The relative relatedness of CrV1 DNA sequences among Cotesia spp. suggest that the virus genes should be conserved enough among Cotesia spp., indicating that the putative CrV1 genes are either deleted or have undergone rapid sequence changes to have escaped PCR detection (J. B. Whitfield pers. comm.). Since circular DNA molecules found in polydnavirus particles have been shown to undergo rapid changes in the form of DNA rearrangements (Webb & Cui 1998), it is likely that CrV1 genes are deleted in some Cotesia spp. In this context it is interesting to discuss the origin of the Crp32 gene. If the operational definition of a polydnavirus gene is its presence on a circular DNA molecule, the Crp32 is not a virus protein. However, from a functional point of view, it is tightly associated with virus particles and protects the virus. Given the degree of DNA rearrangements is it possible that Crp32 is part of a virus-derived gene but not part of the circularised genome of polydnavirus? Likewise, if we accept that virus DNA is rearranged relatively frequently in the wasp genome in such a way that DNA fragments that are not necessary may be abolished, the evolutionary outcome is that Venturia VLPs lack nucleic acids altogether.
References Asgari, S. (1997) Cotesia rubecula polydnavirus-specific gene expression in the host Pieris rapae. Ph.D Thesis, The University of Adelaide. Asgari, S., Hellers, M. & Schmidt, O. (1996) Host haemocyte inactivation by an insect parasitoid: transient expression of a polydnavirus gene. Journal of General Virology 77: 2653-2662. Asgari, S. & Schmidt, O. (1994) Passive protection of eggs from the parasitoid, Cotesia rubecula, in the host, Pieris rapae. Journal of Insect Physiology 40: 789-795. Asgari, S., Schmidt, O. & Theopold, U. (1997) A polydnavirus encoded protein of an endoparasitoid is an immune suppressor. Journal of General Virology 78: 3061-3070. Asgari, S., Theopold, U., Wellby, C. & Schmidt, O. (1998) A protein with protective properties against the cellular defense reactions in insects. Proceedings of the National Academy of Sciences, USA 95: 3690-3695. Beckage, N.E. (1995) Polydnaviruses: Potent mediators of host insect immune dysfunction. Parasitology Today 11: 368-378. Bedwin, O. (1979) An insect glycoprotein: a study of the particles responsible for the resistance of the parasitoid’s egg to the defence reaction of its insect host. Proceedings of the Royal Society London 205: 267-270. Dib-Hajj S. D., Webb, B. A. & Summers, M. D. (1993) Structure and evolutionary implications of a “cysteine-rich” Campoletis sonorensis polydnavirus gene family. Proceedings of the National Academy of Sciences, USA 90: 3765-9. Edson, K. M., Vinson, S. B., Stoltz, D. B. & Summers, M. D. (1981) Virus in a parasitoid wasp: suppression of the cellular immune response in the parasitoid´s host. Science 211: 582-583. Feddersen, I., Sander, K. & Schmidt, O. (1986) Virus-like particles with host protein-like antigenic determinants protect an insect parasitoid from encapsulation. Experientia 42: 1278-1281. Fleming, J. G. W. & Summers, M. D. (1986) Campoletis sonorensis endoparasitic wasps contain forms of C. sonorensis virus DNA suggestive of integrated and extrachromosomal polydnavirus DNAs. Journal of Virology 57: 552-562. Foukas, L. C., Katsoulas, H. L., Paraskevopoulou, N., Metheniti, A., Lambropoulou, M. & Marmaras, V. J. (1998) Phagocytosis of Escherichia coli by insect hemocytes requires both
Otto Schmidt, Sassan Asgari, Markus Beck and Uli Theopold 36
Hymenoptera: Evolution, Biodiversity and Biological Control
activation of the ras/mitogen-activated protein kinase signal transduction pathway for attachment and beta(3) integrin for internalization. Journal of Biological Chemistry 273: 14813-14818. Hoffmann, J. A. (1995) Innate immunity of insects. Current Opinions in Immunology 7: 4-10. Hoffmann, J. A. & Hetru, C. (1992) Insect defensins: inducible antibacterial peptides. Immunology Today 13: 411-415. Hoffmann, J. A., Reichhart, J. M. & Hetru, C. (1996) Innate immunity in higher insects. Current Opinions in Immunology 8: 8-13. Hopper, K. R. and Roush, R. T. (1993) Mate finding, dispersal, number released and the success of biological control introductions. Ecological Entomology 18: 321-331. Hultmark, D. (1994a) Insect immunology. Ancient relationships [news]. Nature 367: 116-117. Hultmark, D. (1994b) Macrophage differentiation marker MyD88 is a member of the Toll/IL-1 receptor family. Biochemistry & Biophysics Research Communications 199: 144-146. Ip, Y. T., Reach, M., Engström, Y., Kadalayil, L., Cai, H., Gonzales-Crespo, S., Tataei, K. & Levine, M. (1993) Dif, a dorsal-related gene that mediates an immune response in Drosophila. Cell 75: 753-763. Kinuthia, W., Li, D., Schmidt, O. & Theopold, U. (1999) Is the surface of endoparasitic wasp eggs and larvae covered by a limited coagulation reaction? Journal of Insect Physiology 45: 501506.. LanzMendoza, H., Bettencourt, R., Fabbri, M. & Faye, I. (1996) Regulation of the insect immune response: The effect of hemolin on cellular immune mechanisms. Cellular Immunology 169: 47-54. Lemaitre, B., Meister, M., Govind, S., Georgel, P., Steward, R., Reichhart, J. M. & Hoffmann, J. A. (1995) Functional analysis and regulation of nuclear import of dorsal during the immune response in Drosophila. EMBO 14: 536-45. Lemaitre, B., Nicolas, E., Michaut, L., Reichhart, J. M. & Hoffmann, J. A. (1996) The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potent antifungal response in Drosophila adults. Cell 86: 973-83. Li, X. S. & Webb, B. A. (1994) Apparent functional role for a cysteine-rich polydnavirus protein in suppression of the insect cellular immune response. Journal of Virology 68: 7482-7489. Liou, H. C. & Baltimore, D. (1993) Current Opinions in Cell Biology 5: 477-487. Luckhart, S. & Webb, B. A. (1996) Interaction of a wasp ovarian protein and polydnavirus in host immune suppression. Developmental & Comparative Immunology 20: 1-21. Reichhart, J. M., Georgel, P., Meister, M., Lemaitre, B., Kappler, C. & Hoffmann, J. A. (1993) Expression and nuclear translocation of the rel/NF-kappa B-related morphogen dorsal during the immune response of Drosophila. Comptes Rendues de l’Academie des Sciences 316: 1218-1224. Rizki, R.M. & Rizki, T.M. (1990) Encapsulation of parasitoid eggs in phenoloxidase-deficient mutants of Drosophila melanogaster. Journal of Insect Physiology 36: 523-529. Rotheram, S. (1967) Immune surface of eggs of a parasitic insect. Nature 214: 700. Salt, G. (1938) Experimental studies in insect parasitism. VI Host suitability. Bulletin of Entomolological Research 29: 223. Salt, G. (1964) The ichneumonid parasite Nemeritis canescens (Gravenhost) in relation to the wax moth Galleria mellonella (L.). Tranactions of the Royal Entomological Society, London 116: 1-14. Salt, G. (1976) The hosts of Nemeritis canescens, a problem in the host specificity of insect parasitoids. Ecological Entomology 1: 63-67.
Host Defence Manipulation by Parasitoid Wasps 37
Salt, G. (1980) A note on the resistance of two parasitoids to the defence reactions of their hosts. Proceedings of the Royal Society London, Series B 207: 351-353. Schmidt, O., Faye, I., Lindstrom Dinnetz, I. & Sun, S. C. (1993) Specific immune recognition of insect hemolin. Developmental & Comparative Immunology 17: 195-200. Schmidt, O. & Schuchmann-Feddersen, I. (1989) The role of virus-like particles in parasitoid-host interaction of insects, pp. 91-119. In Harris, J. R. (Ed.), Subcellular Biochemistry. Plenum Press, New York. Stoltz, D. B., Krell, P., Summers, M. D. & Vinson, S. B. (1984) Polydnaviridae-a proposed family of insect viruses with segmented, double-stranded, circular DNA genomes. Intervirology 21: 1-4. Stoltz, D. B. & Vinson, S. B. (1979) Viruses and parasitism in insects. Advances in Virus Research 24: 125-171. Strand, M. R. & Pech, L. L. (1995) Immunological basis for compatibility in parasitoid-host relationships. Annual Review of Entomology 40: 31-56. Summers, M. D. & Dib-Hajj, S. D. (1995) Polydnavirus facilitated endoparasitoid protection against host immune defenses. Proceedings of the National Academy of Science, USA 92: 29-36. Sun, S. C., Lindstrom, I., Boman, H. G., Faye, I. & Schmidt, O. (1990) Hemolin: an insect-immune protein belonging to the immunoglobulin superfamily. Science 250: 1729-1732. Sun, S.-C., Lindström, I., Lee, J.-Y. & Faye, I. (1991) Structure and expression of the attacin genes in Hyalophora cecropia. European Journal of Biochemistry 196: 247-254. Theopold, U., Samakovlis, C., Erdjument-Bromage, H., Dillon, N., Axelsson, B., Schmidt, O., Tempst, P. & Hultmark, D. (1996) Helix pomatia lectin, an inducer of Drosophila immune response, binds to hemomucin, a novel surface mucin. Journal of Biological Chemistry 271: 12708-12715. Theopold, U. & Schmidt, O. (1997) Helix pomatia lectin and annexin V, molecular markers for hemolymph coagulation. Journal of Insect Physiology 43: 557-674. Waage, J. (1998) Yes, but does it work in the field? The challenge of technology transfer in biological control. Entomophaga 41: 315-332. Webb, B. A. & Cui, L. W. (1998) Relationships between polydnavirus genomes and viral gene expression. Journal of Insect Physiology 44: 785-793. Webb, B. A. & Luckhart, S. (1994) Evidence for an early immunosuppressive role for related Campoletis sonorensis venom and ovarian proteins in Heliothis virescens. Archives of Insect Biochemistry & Physiology 26: 147-163. Webb, B. A. & Luckhart, S. (1996) Factors mediating short- and long-term immune suppression in a parasitized insect. Journal of Insect Physiology 42: 33-40. Xu, D. & Stoltz, D. B. (1991) Evidence for a chromosomal location of polydnavirus DNA in the ichneumonid parasitoid Hyposoter fugitivus. Journal of Virology 65: 6693-6704.
Two Genetically Distinct Venturia canescens Strains Display Different Reproductive Strategies Markus Beck*, Ulrich Theopold** and Otto Schmidt Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) (present addresses: *Department of Entomology, University of Wisconsin-Madison, Madison, WI 53706 USA; ** Department of Molecular Biology, Stockholm University SE-10691 Stockholm, Sweden)
Introduction Endoparasitoids parasitise insect hosts and kill them by feeding on internal tissues to finalise embryonic and larval development. In the case of solitary endoparasitoid wasps, one parasitised host supports only the development of one wasp offspring (Quicke 1997). Therefore, it could be expected that a foraging solitary wasp prefers non-parasitised over already parasitised hosts in order to avoid inter-larval host competition. However, the fact that solitary wasps, capable of discriminating between both host types, nevertheless lay eggs into already parasitised hosts suggests that superparasitism must be considered a potentially successful reproductive strategy (van Alphen & Visser 1990; Sirot & Krivan 1997). Recently, we have discovered two genetically defined strains, designated RP and RM, in a parthenogenetic laboratory population of the solitary endoparasitoid wasp Venturia canescens Gravenhorst (Ichneumonidae) (Hellers et al. 1996). Analysis of the two strains revealed phenotypic differences in the morphology of the respective ovarian calyx tissues, which appear to affect the transport of mature eggs from the ovarioles into the oviduct (Beck et al. 1999). When adult wasps were dissected after being kept without hosts for several days, leading to an accumulation of eggs in the paired oviducts, 80% of all RP egg reservoirs contained more than 40 eggs, whereas in RM wasps such high egg numbers were found only in 12% of all egg reservoirs (Beck et al. unpublished). Consequently, RP wasps should have more mature eggs in the oviducts available for deposition than RM wasps. In this context it is intriguing that RM and RP wasps, in addition to their ovarian phenotypes, also display differences in their reproductive success. When RP wasps were separated from RM wasps and each wasp strain was allowed to parasitise host caterpillars in the absence of the other strain under identical conditions, more offspring were produced by the RP strain as compared to the RM strain (Beck et al. 1999). Surprisingly, when wasps from both strains parasitised hosts together for the same time, the relative number of offspring was reversed and RM wasps were more successful (Beck et al. 1999). The availability of eggs in the respective oviducts of RP and RM wasps might explain the observed reproductive success under conditions where both strains parasitised hosts individually, but not the result obtained when they were ovipositing into caterpillars together. Interestingly, when embryonic development and ovipositing behaviour of RP and RM wasps was studied, additional differences were observed that might provide an explanation for the number of offspring produced by each strain in a situation where they were competing for the same hosts. In order to assess the time needed by RP and RM wasps to complete embryonic development, host caterpillars parasitised once by each wasp type were dissected at distinct time periods after 38
Hymenoptera: Evolution, Biodiversity and Biological Control
Two Genetically Distinct Venturia canescens Strains 39
an ovipositing event. The number of larvae that had already hatched and those still inside the egg shell were recorded. This experiment revealed that offspring from the RM strain delay larval hatching from the egg shell and evidently develop slower than offspring from RP wasps (Beck et al. 1999). Moreover, when the two wasp strains were observed while foraging together on the same host patch, RP wasps immediately started to lay eggs whereas RM wasps initially remained passive and delayed egg deposition until most of the host caterpillars were already parasitised by RP wasps (Beck et al. unpublished). Obviously, RM wasps do not only develop slower than RP wasps, but also tend to lay eggs at a later point in time. It is tempting to assume that the RM strain’s delayed oviposition and slower embryonic development has a synergistic effect, postponing the moment relative to the RP strain, when embryonic RM larvae emerge from the egg shell. Since the age of rival parasitoid larvae constitutes an important factor in host competition (Fisher 1961, 1963; Marris & Casperd 1996) this finding should be of principal importance. In fact, data presented by Marris and Casperd (1996) suggest that for a short period of time, younger Venturia larvae are advantaged over older larvae when physically fighting for host supremacy. This could explain why RM wasps produce more offspring than RP wasps when both wasp types parasitise together and compete for the same hosts. Since the whole laboratory population studied in this investigation consists exclusively of parthenogenetically reproducing females, the RP and the RM strain represent two clonal lines that are reproductively separated. Moreover, since analysis of the corresponding genomes revealed a number of differences, the two lines are also genetically distinct. In addition to the allelic gene, which has been used as a diagnostic marker to isolate the two strains from a mixed laboratory culture (see Fig. 1), they differ in at least one other gene (Beck et al. unpublished). Interestingly, when genomic DNA isolated from individual RP and RM wasps was analysed by RAPD-PCR, wasps of the same strain always produced identical DNA banding patterns with each of the 23 random primers that were used (six individuals per strain were tested for each primer). However, for some primers, the resultant banding patterns differed between the strains. Closer inspection of the RP and RM strain specific DNA banding patterns revealed 90.5% monomorphic (identical in both strains) and 9.5% polymorphic (unique to one strain) scorable DNA fragments (Beck et al. 1999). Although it remains to be shown whether some of the polymorphic DNA fragments derive from coding regions, this level of polymorphism indicates that both strains possess distinct genomes with possible differences in several genes. Given that sexual recombination between these two genetically distinct strains does not occur, it could be assumed that particular gene combinations are conserved and transferred into the next generation. This would also apply to specific genes or gene combinations responsible for differences in ovarian phenotypes, reproductive success and behaviour. Considering this, it is tempting to propose that parthenogenesis has resulted in two genetically different strains leading to two distinct reproductive strategies. In order to learn more about the reproductive strategies employed by RP and RM wasps, and the reasons why both strains have been maintained in the mixed laboratory population, investigations were started to determine the outcome of host competition between the two wasp types under different parasitisation conditions. Preliminary data are presented in the present paper.
Materials and Methods Insect cultures Parthenogenetically reproducing female V. canescens were reared on final instar caterpillars of the Mediterranean flour moth, Ephestia kuehniella Zeller (Phycitidae). Moths were fed on
Markus Beck, Ulrich Theopold and Otto Schmidt 40
Hymenoptera: Evolution, Biodiversity and Biological Control
A)
N
TR
N B)
PHGPX
PHGPX RP
RM
M
C
C
kDa 64 50 36 30
Figure 1
Two different variants of a virus-like particle protein (VLP1) are produced by two alleles of the maternal gene vlp1, which is expressed in the ovaries of the female wasps. In one of the these variants a tandem repeated sequence (TR) is deleted. Since individual wasps in the laboratory population are homozygous for either one of the two vlp1 alleles, they are according to the allele addressed as wasps belonging to the RP (repeat plus) or RM (repeat minus) strain (Hellers et al. 1996): A) domain structure of the two allelic VLP1 variants. The C-terminus is in both alleles identical and displays homology to the antioxidant enzyme phospholipid-hydroperoxide glutathione peroxidase (PHGPX). The position of the deleted TR is indicated (TR); B) western blot analysis of wasp ovaries from individual wasps using anti-VLP1 antiserum. The different molecular weight of the two allelic VLP1 variants allows the distinction between RP and RM wasps. RP = RP wasp ovary; RM = RM wasp ovary; M = SeeBlue® pre-stained molecular weight marker from Novex (San Diego, USA).
crushed oats and V. canescens, when necessary, on a 33% water-honey solution. Both insects were maintained in the laboratory at 25˚C under a 14:10 h (light:dark) photoperiod.
SDS-PAGE and Western blots Protein extracts from wasp ovaries were electrophoretically separated by SDS-PAGE and analysed on Western blots using antiserum specific for VLP1 as described previously (Hellers et al. 1996). VLP1 is the protein encoded by the vlp1 gene. Since vlp1 exists in two alleles, two VLP1 variants of different molecular weights are expressed in the ovaries (Hellers et al. 1996). These allelic VLP1 variants can be used to distinguish the RP and RM strain. Competition experiment Two healthy looking, freshly emerged female wasps, one RP and one RM, were placed together with 25 final instar host caterpillars into a culture vessel (6 cm in diameter, 9 cm high). After 24 h the wasps were removed and their genetic identity as RP or RM wasps confirmed by Western blot analysis. Twenty-five days after the experiment commenced, offspring started to emerge.
Two Genetically Distinct Venturia canescens Strains 41
A
B
Figure 2
Venturia canescens larvae fighting for host dominance: A) fight between two parasitoid larvae showing a younger larva attacking an older one causing an injury; B) encapsulation of a wounded larva. The injury inflicted on the larval cuticle induced host defence reactions which led to cell aggregation and encapsulation at the site of the wound. Scale bars = 300 µm.
Individual offspring were collected and their genetic identity analysed as described above, in order to determine which wasp type won the competition for host supremacy. Abbreviations used in the text and figures are: PAGE, Polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; PHGPX, phospholipid-hydroperoxide glutathione peroxidase; RAPD, randomly amplified polymorphic DNA; RM, repeat minus; RP, repeat plus; SDS, sodium dodecyl sulphate.
Results and Discussion The exact mechanisms that influence the outcome of host competition in superparasitised caterpillars are not known. However, two general processes have been described in the literature involving physical combat among parasitoid larvae (Salt 1966; Marris & Casperd 1996) and physiological suppression (Fisher 1963; Vinson & Hegazi 1998). Physiological suppression seems to be mediated by the nutrient content (Vinson & Hegazi 1998) and/or the concentration of
Markus Beck, Ulrich Theopold and Otto Schmidt 42
Hymenoptera: Evolution, Biodiversity and Biological Control
A) Average no. of offspring
14 12 10 8 6 4 2 0
B)
RP
RM
RP (1)
RM (2)
Average no. of offspring
14 12 10 8 6 4 2 0
14 Average no. of offspring
C)
12 10 8 6 4 2 0
RP (2) Figure 3
RM (1)
Average number of RP and RM offspring when two wasps, one of each strain, were allowed to compete for 25 hosts for 24 h: A) two wasps, one of each strain, were put together with host caterpillars at the same time (RP; RM); B) one RP wasp was allowed to search and parasitise for 7 h [RP (1)], before an RM wasp was added [RM (2)]; C) reversed situation, where one RM wasp was allowed to search and parasitise for 7 h [RM (1)], before an RP wasp was added [RP (2)]. Each experiment was carried out twice. In order to determine whether the emerging wasps were offspring from the RP or the RM wasp, they were individually dissected, and the corresponding ovary samples examined by Western blot analysis (see Fig. 1) (bars = ranges).
Two Genetically Distinct Venturia canescens Strains 43
dissolved oxygen (Fisher 1961, 1963) in the host hemolymph. Since an altered physiology in the parasitised host appears to render conditions unfavourable for the development of newly oviposited eggs (Vinson & Hegazi 1998) and younger larvae (Fisher 1963), physiological suppression affects early developmental stages more severely than older ones. Similarly, when wasp larvae physically fight with each other within the hemocoel of a superparasitised host, older larvae are usually more likely to eliminate younger ones (Vanbaaren et al. 1995). However, V. canescens can be considered as an interesting exception in this context. First instar larvae seem to be more mobile than older larvae and possess a sclerotised head capsule with a strong beak, which is less developed in later larval stages (Marris & Casperd 1996). They are able to seek out supernumerary larvae and use their mandibles to perforate the cuticle of a competitor (Fig. 2A). Any damage inflicted to the cuticle is immediately recognised as foreign and attacked by the host’s immune system (Fig. 2B), which eventually leads to the death of the injured larva (Salt 1966). When Marris and Casperd (1996) staged in vitro contests between different larval V. canescens instars, they observed that first instar larvae competing with second or third instars initiate more fights, inflict more bites, and sustain these bites for longer. From this they concluded that first instar larvae must be advantaged over second and third instars and are more likely to win a physical combat (Marris & Casperd 1996). Indeed, in vivo experiments carried out by Sirot (1996) show that the interval between ovipositions, and therefore the age difference between the first and second laid offspring, is crucial to the success rate of the superparasitising female. In fact, a substantial pay-off from superparasitism could be achieved by these females as long as the first and second ovipositions were less than three days apart (Sirot 1996). These data strongly support the idea that, dependent on the age distribution of larvae inside superparasitised caterpillars, either very old or very young larvae, may have a selective advantage. It seems that younger larvae are more likely to succeed when physically fighting against larvae just a few days older, whereas larvae too far ahead in growth suppress younger larvae physiologically (see also Vinson & Mourad, this volume). The RM strain’s reproductive success when parasitising hosts together with the RP strain at the same time (Fig. 3A), in combination with the delayed timing of oviposition and the extended developmental time (Beck et al. unpublished), suggests that the RM strain has successfully evolved mechanisms to increase the pay-off from superparasitism. Since creating a larger age difference only makes sense if competing with relatively young larvae, it appears that only a small time frame exists for the RM strain to be successful. In order to test this assumption and the initial hypothesis that the reproductive success of the RM strain is associated with changes in behaviour and development, one strain was allowed to parasitise hosts for a period of 7 h before the other strain was added. The number of offspring produced by each strain under these conditions was then analysed. When RP wasps were given a head start of 7 h, they produced more offspring than RM wasps (Fig. 3B). This outcome resembles the situation where each wasp type parasitises hosts separately (Beck et al. 1999). On the basis of our behavioural studies the simplest interpretation for this result is that RP wasps deposit eggs immediately into caterpillars and by the time RM wasps start to superparasitise caterpillars, the oldest RP larva may already be able to suppress RM larvae by physiological means. Interestingly, when RM wasps were given a head start, they were not able use this advantage, and the number of offspring for each wasp strain was about the same (Fig. 3C). When competing inside hosts, RM and RP offspring appear to have been evenly matched regarding their competitive abilities. It seems that RM wasps must have delayed egg laying, but still oviposited before the RP wasps. Only under such conditions should RM and RP parasitoid larvae be about the same age when
Markus Beck, Ulrich Theopold and Otto Schmidt 44
Hymenoptera: Evolution, Biodiversity and Biological Control
competing within the host and therefore have an equal probability of winning the combat. In addition, it could be imagined that during the initial 7 h RM wasps may have depleted their limited egg reserves (Beck et al. unpublished), which would preclude them from depositing eggs into hosts previously parasitised by RP wasps. Whatever the reasons for the observed differences in reproductive success, it appears that this preliminary result corresponds well to our previous observations and the information available in the literature concerning host competition in Venturia. Furthermore, the competitive abilities of both strains were not symmetrical, giving the impression that the RM strain is dependent on the RP strain in order to be reproductively successful. Therefore, this experiment can be considered as additional evidence for the co-existence of two reproductive strategies that were maintained in the laboratory population by two different strains. Since the RP and the RM strain are genetically distinct and sexual recombination does not occur (Beck et al. unpublished), the question can be raised as to whether host competition under such circumstances can be considered conspecific. This sheds new light on behavioural studies dealing with superparasitsm; in particular considering that morphological characters commonly used for taxonomic recognition of ichneumonid wasps fail to distinguish between the RP and RM V. canescens strains (P. Dangerfield, pers. comm.).
Acknowledgements We thank Paul Dangerfield for the taxonomic analysis of the two Venturia strains. This work was supported by a grant from the Australian Research Council (ARC) to OS, an ARC Research Fellowship to UT, and a Postgraduate and an Overseas Postgraduate Research Scholarship from The University of Adelaide to MB.
References Beck, M., Siekmann, G., Li, D., Theopold, U. & Schmidt, O. (1999) A maternal gene mutation correlates with an ovary phenotype in a parthenogenetic wasp population. Insect Biochemistry and Molecular Biology 29: 453-460. Fisher, R. C. (1961) A study in insect multiparasitism. II. The mechanism and control of competition for possession of the host. Journal of Experimental Biology 38: 605-628. Fisher, R. C. (1963) Oxygen requirements and the physiological suppression of supernumerary insect parasitoids. Journal of Experimental Biology 40: 531-540. Hellers, M., Beck, M., Theopold, U., Kamei, M. & Schmidt, O. (1996) Multiple alleles encoding a virus-like particle protein in the ichneumonid endoparasitoid Venturia canescens. Insect Molecular Biology 5: 239-249. Marris, G. C. & Casperd, J. (1996) The relationship between conspecific superparasitism and the outcome of in vitro contests staged between different larval instars of the solitary endoparasitoid Venturia canescens. Behavioural Ecology & Sociobiology 39: 61-69. Quicke, D. L. J. (1997) Parasitic Wasps. Chapman and Hall, London. Salt, G. (1966) Experimental studies in insect parasitism XIV. The haemocytic reaction of a caterpillar to larvae of its habitual parasite. Proceedings of the Royal Society of London, Series B 165: 155-178. Sirot, E. (1996) The pay-off from superparasitism in the solitary parasitoid Venturia canescens. Ecological Entomology 21: 305-307. Sirot, E. & Krivan, V. (1997) Adaptive superparasitism and host-parasitoid dynamics. Bulletin of Mathematical Biology 59: 23-41.
Two Genetically Distinct Venturia canescens Strains 45
van Alphen, J. J. M. & Visser, M. E. (1990) Superparasitism as an adaptive strategy for insect parasitoids. Annual Review of Entomology 35: 59-79. Vanbaaren, J., Boivin, G. & Nenon, J. P. (1995) Intraspecific hyperparasitism in a primary hymenopteran parasitoid. Behavioural Ecology & Sociobiology 36: 237-242. Vinson, S. B. & Hegazi, E. M. (1998) A possible mechanism for the physiological suppression of conspecific eggs and larvae following superparasitism by solitary endoparasitoids. Journal of Insect Physiology 44: 703-712. Vinson, S. B. & Ahmed Kamal Mourad, A. K. (this volume) The response of Cardiochiles nigriceps Viereck (Hymenoptera: Braconidae) larvae to conspecific competitors.
The Response of Cardiochiles nigriceps Viereck (Hymenoptera: Braconidae) Larvae to Conspecific Competitors S. Bradleigh Vinson1 and Ahmed Kamal Mourad2 1
Department of Entomology, Texas A&M University, College Station, TX 77843 USA (email:
[email protected]) 2
Plant Protection Department, Alexandria University, Alexandria, Egypt
Introduction Most parasitic hymenoptera are solitary individuals and as larvae require the complete resources of the host in order to develop. However, it may be advantageous for a female to oviposit more than one egg into a host (van Alphen et al. 1987; van Dijke & Waage 1987). This results in superparasitised hosts and intense intraspecific competition between the developing larvae for possession of the host (van Lenteren 1976). This competition leads to the elimination of competitors by physical attack or physiological suppression. However, little is known in regards to either of these activities (Vinson & Iwantsch 1980; Mackauer 1990).
Physiological Suppression In some super-parasitised hosts either one of the embryos fail to develop and hatch, or the embryos hatch but one of the larvae does not develop. Usually the older is the victor. Physiological suppression has been commonly reported and attributed to a number of mechanisms. These include death due to starvation or anoxia, toxins, cytolytic enzymes, or other secretions which are either released by the developing larvae and associated teratocytes or injected by the female (Mackauer 1990). Of these methods, the release of toxins has been most often cited (Silvers & Nappi 1986; Lawrence 1988). While there is a lot of data to show that female parasitoids inject factors, and larvae and their associated teratocytes release factors into the host, there is no supporting evidence that any of these factors are toxins (Vinson 1990; Beckage 1993). Further, as suggested by Vinson and Hegazi (1998) it would not make much evolutionary sense for a female to evolve the ability to produce a toxin that kills or prevents the development of her youngest progeny. While the evolution of a toxin by the larvae would provide a competitive edge, it is also difficult to see how such a toxin would only effect younger stages. Hu and Vinson (1997) showed that Campoletis sonorensis (Cameron) eggs would not develop in various in vitro media, while development would take place once the embryonic membranes were developed. Similar data was provided by Pennacchio et al. (1992) for Cardiochiles nigriceps Viereck. Vinson and Hegazi (1998) showed that embryonated eggs were able to develop in media with a range of osmotic pressures and in hemolymph from both parasitised and non-parasitised hosts, while younger eggs only showed development in non-parasitised hemolymph. Further, the development of non-embryonated and embryonated eggs singly or in pairs in diluted non-parasitised hemolymph revealed the occurrence of physiological suppression of the younger larvae in
46
Hymenoptera: Evolution, Biodiversity and Biological Control
The Response of Cardiochiles nigriceps Viereck Larvae to Conspecific Competitors 47
all paired cases involving a younger and older larva. This removed any adult involvement. Further, some of the paired young and paired older larvae also failed to develop. These results were not consistent with the presence of a toxin. Vinson and Hagazi (1998) offered an alternate suggestion based on several observations and the available evidence. These are: 1) that embryonated eggs are more effective in regulating the flow of ions and chemicals through the chorion and associated egg membranes than non-embryonated eggs; 2) that as an egg develops its ability to continue to develop in less than a favourable environment, represented by the media, increases; 3) that as the egg enclosed embryo develops, its nutritional needs change; 4) that the hemolymph of a parasitised host is altered and becomes different when compared to the hemolymph of non-parasitised hosts; 5) the hemolymph changes are important to the changing nutritional needs of the developing larvae, and 6) these nutritional changes in the host are regulated by and are one of the major functions of the factors that a female parasitoid injects and the developing larvae release. As a result, any eggs deposited into the hemolymph of a previously parasitised host encounters a hemolymph environment that will no longer support the development of these earlier developmental stages (Vinson & Hagazi 1998). There appear to be two important developmental changes that provide an advantage to the older larvae. This is the development of the embryonic membranes which provide the embryo with the ability to regulate the movement of factors present in the hemolymph environment in and out. Further, the embryonic membrane can alter some of these factors as they pass through the membrane; a situation not available to younger developing eggs. The second is when the embryo hatches. After hatching the embryo is now presumably able to feed and digest large molecular weight proteins which are not available to younger developmental stages. At the same time the protein content of the hemolymph of the host, in this case Heliothis virescens (F.), increases (Pennacchio et al. 1994).
Physical Combat Although the larvae of the Hymenoptera are mandibulate, the first instar of most solitary parasitoid species are considered to be strongly mandibulate (Hagen 1964). This is in contrast to gregarious species. For example, Laing and Corrigan (1987) reported that the gregarious Cotesia glomerata (L.) had poorly developed mandibles in comparison to the solitary species Cotesia rubecula (Marshall). Further, the large head and falcate mandibles typical of many first instar solitary parasitoids are lost in older larvae which become more grub-like (Hagen 1964). These older larvae not only have comparatively smaller mandibles, but are less agile and tend to lose in physical combat with first instars (Chow & Mackauer 1984). Although physical combat has been commonly reported, no one has examined the mobility of larvae. We wanted to determine if newly hatched larvae could move and, if so, could they detect the presence of other larvae and locate them? We initiated some experiments to determine if newly hatched larvae could orientate to and locate competing larvae. Although physical combat is commonly reported (Vinson & Iwantsch 1980) there has been no information regarding the ability of larvae to move and seek out these competitors.
Methods Here we report on recent studies concerning the ability of C. nigriceps to move and impact competing conspecifics (Vinson & Mourad in review). First instar larvae were obtained from dissecting earlier super-parasitised hosts. Two of these larvae were placed 5 to 10 mm apart in a
S. Bradleigh Vinson and Ahmed Kamal Mourad 48
Hymenoptera: Evolution, Biodiversity and Biological Control
Petri dish with insect tissue culture media and were observed to determine if they could move, the time to contact each other, and the behaviour on contact. To determine if larvae could detect and move to a conspecific, we developed a modified aquatic ‘Y’ tube that was a ditch in the form of a ‘Y’ formed out of wax. A reservoir filled with cell culture media fed the two arms of the ‘Y’, the media then flowed through the common base of the ‘Y’ and out to another lower reservoir. Flow was aided with a cotton thread that acted as a capillary from the reservoirs to the arms of the ‘Y’. We also found that the addition of the thread to the ‘Y’ ditch aided the movement of the larvae. One larva was placed in one of the arms and the other larva was placed at the base of the ‘Y’.
Results When two first instar larvae are placed within a few millimeters of each other they move towards each other and attack head to head. However, this movement required several hours. We found that this movement could be facilitated by adding a filter paper substrate. Studies using the ‘Y’ ditch revealed that within an hour significantly more of the larvae placed in the base of the ‘Y’ were found in the arm containing the restricted competitor larvae. We also observed that older larvae placed in media and exposed to a younger larvae were less able to move in a directed way and were less agile compared to the younger larvae. This along with the comparatively reduced mandibles of the older larvae placed them at a disadvantage. As a result the younger larvae generally won the encounter.
Conclusion These studies demonstrate that first instar C. nigriceps larvae can orient to, move towards, and attack conspecifics. These results also suggest that physiological suppression favours older larvae in any competition with conspecifics, while physical combat favours younger larvae. This presents a strong competitive environment that provides for a clear resolution, but does not favour either age. Thus, females electing to super-parasitise submit their larvae to a true contest that favours the survival of the fittest of the offspring.
Acknowledgements The work reported in this paper has been partially supported through a Fulbright Fellowship to AKM.
References Beckage, N. E. (1993) Endocrine and neuroendocrine host-parasite relationships. Receptor 3: 1-13. Chow, F. J. & Mackauer, M. (1984) Inter- and intra-specific competition in Aphidius smithi and Praon pequodorum (Hymenoptera: Aphidiidae). Canadian Entomologist 116: 1097-1107. Hagen, K. S. (1964) Developmental stages of parasites. pp. 168-246. In: DeBach, P. (Ed.), Biological Control of Insects, Pests and Weeds. Chapman & Hall, London. Hu, J. S. & Vinson, S. B. (1998) The in vitro development from egg to prepupa of Campoletis sonorensis (Hymenoptera: Ichneumonidae) in an artificial medium: importance of physical factors. Journal of Insect Physiology 44: 455-462.
The Response of Cardiochiles nigriceps Viereck Larvae to Conspecific Competitors 49
Laing, J. E. & Corrigan, J. E. (1987) Intrinsic competition between the gregarious parasite, Cotesia glomeratus and the solitary parasite, Cotesia rubecula (Hymenoptera: Braconidae) for their host, Artogeia rapae (Lepidoptera: Pieridae). Entomophaga 32: 493-501. Lawrence, P. O. (1988) In vivo and in vitro development of first instars of the parasitic wasp, Biosteres longicaudatus (Hymenoptera: Braconidae). pp. 351-366. In Gupta, V. K. (Ed.), Advances in Parasitic Hymenoptera Research. E. J. Brill, Leiden. Mackauer, M. (1990) Host discrimination and larval competition in solitary endoparasitoids. pp. 41-62. In Mackauer, M, Ehler, L.E. & Rolands, J. (Eds), Critical Issues in Biological Control. Intercept Ltd, Andover. Pennacchio, F., Vinson, S. B. & Tremblay, E. (1992) Preliminary results on in vitro rearing of the endoparasitoid Cardiochiles nigriceps from egg to second instar. Entomologia Experimentalis et Applicata 64: 209-216. Pennacchio, F., Vinson, S. B., Tremblay, E. & Tanaka, T. (1994) Biochemical and developmental alterations of Heliothis virescens (F.) (Lepidoptera, Noctuidae) larvae induced by the endophagous parasitoid Cardiochiles nigriceps Viereck (Hymenoptera, Braconidae). Archives of Insect Biochemistry & Physiology 26: 211-233. Silvers, M. J. & Nappi, A. J. (1986) In vitro study of physiological suppression of supernumerary parasites by the endoparasitic wasp Leptopilina heterotoma. Journal of Parasitology 72: 405-409. van Alphen, J. J. M., van Dijken, M. J. & Waage, J. K. (1987) A functional approach to superparasitism: Host discrimination needs not be learnt. Netherlands Journal of Zoology 37: 167-179. van Dijken, M. & Waage, J. K. (1987) Self and conspecific superparasitism by the egg parasitoid Trichogramma evanescens. Entomologia Experimentalis et Applicata 43: 183-192. van Lenteren, J. C. (1976) The development of host discrimination and the prevention of superparasitism in the parasite Pseudeucoila bachei Weld (Hymenoptera: Cynipidae). Netherlands Journal of Zoology 26: 1-83. Vinson, S. B. (1990) How parasitoids deal with the immune system of their host: an overview. Archives of Insect Biochemistry & Physiology 13: 3-27. Vinson, S. B. & Hegazi, E. M. (1998) A possible mechanism for the physiological suppression of conspecific eggs and larvae following superparasitism by solitary endoparasitoids. Journal of Insect Physiology 44: 703-712 Vinson, S. B. & Iwantsch, G. F. (1980) Host suitability for insect parasitoids. Annual Review of Entomology 25: 397-419. Vinson, S. B. & Mourad, A. K. (in review) The behaviour and the physical response of first instar parasitoid larvae to competitors. Entomologica Experimentalis et Applicata.
Hypopharyngeal Gland Function, Glandular Cell Senescence and Gland Reactivation In Bees Carminda da Cruz-Landim, Rosiléia Ana Cássia da Costa and Regina Lúcia Morelli Silva de Moraes Departamento de Biologia, Instituto de Biociências, UNESP, 13506-900 Rio Claro, SP, Brazil (e-mail:
[email protected])
Introduction The highly eusocial bees (as is the case of Apinae and Meliponinae) have a division of labour in which the young workers undertake tasks in the nest and older workers forage in the field. The exocrine glands of these bees frequently have developmental cycles closely related to this division of labour, since the glandular products are necessary for performance of some of these tasks. As the colony functions as a ‘super-organism’ and must maintain an internal homeostasis, the age at which workers are able to perform their duties is not rigid (Moritz & Southwich 1992). This means that the secretory gland cycles can be changed according to colony requirements. The hypopharyngeal glands are one such case for which the secretory cycle is closely related to the workers’ division of labour. Hypopharyngeal glands are present in all Hymenoptera and has been reported in solitary bees, ants and wasps as having a wide variety of types, from very simple to a complex organisation, but in general their function is unknown. In Apis mellifera L. these glands are well documented as responsible for food production for the brood and queen. In this species the glands are found only in workers and are more developed during nursing activity (Cruz-Landim & Hadek 1969). At the beginning of foraging activity the gland undergoes degeneration although, according to some authors, glands which have already finished food production can be reactivated as a response to colony requirements (Browers 1983; Gracioli et al. 1999). It is also reported that after the end of the nursing phase the hypopharyngeal glands are still functional, but have changed function, to produce only enzymes (Simpson et al. 1968; Takenaka et al. 1990). In the Meliponinae, aside from being present in the workers of all species, the occurrence of hypopharyngeal glands has been reported also in males and queens. In workers, these glands have a secretory cycle similar to A. mellifera, i.e. being more developed in nurses (Cruz-Landim et al. 1986/87; Costa 1997) and undergoing regression in forager workers. Further, the electrophoretic pattern of glandular extracts from this species is similar to A. mellifera (Silva de Moraes et al. 1996; Costa 1997). The aim of the present investigation was to determine whether the morphology of glandular cells of apine and meloponine bees support the idea of glandular reactivation, and the occurrence of a second cycle of secretion after the nursing phase. Further, it aimed to compare the protein patterns of gland extracts to evaluate the possibility that the glands of forager workers return to food production in colonies under natural conditions.
50
Hymenoptera: Evolution, Biodiversity and Biological Control
Hypopharyngeal Gland Function, Glandular Cell Senescence and Gland Reactivation In Bees 51
Material and Methods Hypopharyngeal glands of newly emerged nurse and forager workers of A. mellifera, Melipona quadrifasciata anthidioides Latreille and Scaptotrigona postica Lepeletier collected from normal colonies, were dissected and fixed in 2.5% glutaraldehyde in 0.1M cacodylate buffer, pH 7.2. After washing in buffer, glands were post-fixed in 1% osmium tetroxide in the same buffer. The glands were then contrasted overnight with 2% alcoholic solution of uranyl acetate, dehydrated in a series of increasing concentrations of acetone and embedded in Epon Araldite. The ultrathin sections were stained with lead citrate and examined under a transmission electron microscopy (TEM). Glands of forager workers were also submitted to acid phosphatase reaction using glycerophosphate as a substrate to the enzyme. Extracts of glands of newly emerged, nursing and forager workers of S. postica were used for electrophoresis by SDS-PAGE using extracts of glands of nursing workers of A. mellifera for comparison. The age of bees was not controlled. Specimens were collected while nursing or foraging. The newly emerged workers of S. postica were collect as soon as they left the comb cell. At least ten individuals of each phase was examined.
Results and Discussion The hypopharyngeal glands of A. mellifera, M. q. anthidioides and S. postica are formed by two long bunches of secretory units connected to a main excretory duct by canaliculi coming from each glandular cell. The main excretory canals end laterally on the floor of the buccal cavity, where the secretion is discharged. Glandular cells may be classified as class III according to Noirot and Quennedey (1974), i.e. they possess an intracellular canaliculus or space for secretion collection inside the glandular cell (Beams & King 1933). In A. mellifera each secretory unit comprises several cells, although each one is independently linked to the main excretory duct by its own excretory canaliculus (Fig. 1). In M. q. anthidioides and S. postica the secretory units are unicellular (Fig. 2). Based on similar morphologies of the cells during the worker’s developmental cycle and protein product similarities (Silva de Moraes et al. 1996; Costa 1997), an attempt was made to determine the capacity of hypopharyngeal glands to re-start food production in old workers by correlating ultrastructural features with electrophoretic protein patterns of S. postica. The results indicate that hypopharyngeal gland cells have a second round of protein secretion, similar to those of A. mellifera. Later, at the beginning of the forager phase, most food secretion has been eliminated from the cell (Figs 3, 4) and is sometimes replaced by several myelin figures, indicating a post-secretory phase in the glandular cycle. However, a significant amount of rough endoplasmic reticulum is still present in the gland cells of some workers in this phase which is morphologically different from that present in nursing individuals, and a new type of secretion is evident. This is represented by larger, less electron-dense granules sometimes with a granular content (Figs 5, 6). These two types of secretion may be temporarily separated, as in A. mellifera, but in S. postica they seem to overlap. The second type of secretion has morphological features of granules that contain mucous-like substances. Treatment with the acid periodic-Schiff reacts positively in these granules, confirming their mucous-like nature. During the second secretory cycle the myelin figures are still present and some electron-dense granules around the vacuoles react positively to acid phosphatase. This indicates re-absorption of organelles and the first type of protein secretion (Fig. 7) that is no longer functional. The
Carminda da Cruz-Landim et al. 52
Figures 1-8
Hymenoptera: Evolution, Biodiversity and Biological Control
1) Light micrographs of a hypopharyngeal gland of an A. mellifera forager worker; 2) low magnification transmission electron micrograph of a secretory unit of M. q. anthidioides nurse worker; 3) and 4) transmission electron micrographs of cells of A. mellifera (3) and S. postica (4) from hypopharyngeal glands of an early forager showing many degenerative figures; 5) and 6) transmission electron micrographs of forager glandular cells of S. postica (5) and A. mellifera (6) showing a new type of secretion(s); 7) acid phosphatase reaction (arrows) in a glandular cell of a forager worker; 8) condition of a glandular cell in an old forager (av = autophagic vacuole; ec = excretory canal; ic = intracellular canal; mc = main canal; mf = myelin figures; mr = residual membranes; n = nucleus; nu = nucleolus; rer = rough endoplasmic reticulum; s = secretion; se = secretory cells).
Hypopharyngeal Gland Function, Glandular Cell Senescence and Gland Reactivation In Bees 53
Figure 9.
SDS PAGE profiles of the extracts of hypopharyngeal gland proteins (EW = newly emerged worker of S. postica; FW = forager worker of S. postica; NW1 nurse worker of A. mellifera, NW2 = nurse worker of S. postica).
presence of an extensive autophagy seems to indicate a cellular senescence. However, these cells still seem to be functional but accumulation of the first type of protein secretion is not observed to occur in them. The features mentioned above are described for cells from early forager workers. Older foragers show very degenerated gland cells. Their cell cytoplasm appears full of myelin figures, and their general appearance is of total disintegration, with the organelles being difficult to recognise (Fig. 8). The nucleus shows a high chromatin condensation and very irregular contours. This degeneration is so widespread that cell recuperation seems most improbable. It should be noted that this cell degeneration occurs along the gland, is not synchronous, that some secretory units still remain in good conditions, even in older foragers, but cells that have degenerated are fated to die. Glandular cell degeneration was observed in all three species studied, and was occasionally found among forager bees that still had well-developed glands, but these may be younger bees that have only just left the colony for foraging duties. According to the morphological results, the early foragers are finishing the first secretory phase in which they seem to be involved only or mainly in food production for the brood. Subsequently, the cells start the production of a second type of secretion, possibly digestive enzymes. In this phase the gland cells, although having undergone some autophagic degeneration and
Carminda da Cruz-Landim et al. 54
Hymenoptera: Evolution, Biodiversity and Biological Control
showing signs of senescence, are still active. In older foragers this condition was not observed and, although a few secretory units may appear healthy, it seems improbable that the gland could return to food secretion or produce any other kind of secretion. The contridictory reports about hypopharyngeal gland recovery could be due to a misinterpretation of several events that take place in bee colonies; for instance, the fact that some workers do not pass through all steps of the colony work division. Some forager workers with well developed glands may represent this type of individual. However, it is possible that gland development may respond to environmental conditions in the colony. In workers of A. mellifera from weak colonies, hypopharyngeal glands attain the peak of development early and remain longer in this condition. The stimulus for this glandular behaviour could be nervous or hormonal. Therefore, it appears that the glands of workers that revert to an earlier task have not recovered, but rather have remained in functional condition during the following activity of the worker. The electrophoretic results (Fig. 9) corroborate the morphological findings, indicating that although the glands of foragers remain functional, they produce different patterns of protein secretion. The electrophoretic profiles of S. postica show several similarities with A. mellifera (NW1) and among the different behavioural phases of S. postica. However, some bands are absent or very weak particularly in foragers (FW). The bands corresponding to the food production phase, between 48 and 76 kDa, are practically absent from foragers, while the low molecular weight bands are intensified and those of high molecular weight are completely absent. Therefore, although the electrophoretic results refer to total gland extracts which also contain cellular components, it seems that proteins of low molecular weight, probably enzymes, may be produced by glands of forager workers. These results corroborate the observation that the hypopharyngeal glands at different times and from different individuals produce different types of secretion. In the transition from nurse to forager, a rearrangement of the gland’s secretory apparatus may be necessary to accommodate the synthesis of new protein secretion. This is accounted for by the degenerative processes seen in the glandular cells by the end of the nurse phase. However, it is only at the end of this activity, which almost coincides with death of a worker, that the glandular cells enter in an irreversible process of degeneration. It can be assumed that final cell death occurs progressively, due to exhaustion of the secretory cellular program and that this coincides with the end of the worker’s life. For this reason the great majority of old foragers will be precluded from reassuming any secretory activity and, by consequence, acting as nurses. Only under special conditions of the colony (e.g. weak colonies) or in rare cases in large colonies, the foragers will still have functional hypopharyngeal glands. In older foragers, depending on the number of cell deaths per gland, some secretion production may still be possible, suggesting that the second type of secretion is necessary for foraging, or for processing the foraged products (mainly nectar). In summary, the morphological evidence and eletrophoretic protein patterns for S. postica, indicate that hypopharyngeal glands coming back to food production in older workers seems unlikely in colonies under normal conditions.
Acknowledgements Thanks to Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the financial support.
Hypopharyngeal Gland Function, Glandular Cell Senescence and Gland Reactivation In Bees 55
References Beams, H. W. & King, R. L. (1933) The intracellular canaliculi of the pharyngeal glands of the honeybee. Biological Bulletin 64: 309-314. Browers, E. V. M. (1983) Activation of the hypopharyngeal glands of honey bees in winter. Journal of Apicultural Research 22: 137-141. Costa, R. A. C. (1997) Morfologia e funcionamento das glândulas hipofaríngeas em operárias, rainhas e machos de Scaptotrigona postica Latreille, 1804 (Hymenoptera, Apidae, Meliponinae). Dissertação de Mestrado, IBRC, Unesp. Cruz-Landim, C. & Hadek, R. (1969) Ultrastructure of Apis mellifera hypopharyngeal gland. pp 121-130. In Proceedings of the International Union for the Study of Social Insects, University of Bern, Bern. Cruz-Landim, C., Silva de Moraes, R. L. M. & Costa-Leonardo, A. M. (1986/87) Ultra-estrutura das glândulas hipofaríngeas de Melipona quadrifasciata anthidioides Lep. (Hymenoptera, Apidae). Naturalia 11/12: 89-96. Gracioli, L. F., Silva de Moraes, R. L. M. & Cruz-Landim (1999) Electrophoretical studies on protein hypopharyngeal glands of aged Popis mellifera (Hymenoptera: Apidae): workers induced to return to brood feeding activity. Naturalia 24: 9-17. Mortiz, R. F. A. & Southwick, E. E. (1992) Bees as Super Organisms. An Evolutionary Reality. Spring-Verlag, Berlin Noirot, C. & Quennedey, A. (1974) Fine structure of insect epidermal glands. Annual Review of Entomology 19: 61-80. Silva de Moraes, R. L. M., Braga, M. R. B. & Azevedo, A. (1996) Eletrophoretical studies of the hypopharyngeal glands and of the larval food of Melipona quadrifasciata anthidioides Lep. (Hymenoptera, Meliponinae). Insectes Sociaux 43: 183-188. Simpson, J., Riedel, B. M. & Wilding, N. (1968) Invertase in the hypopharyngeal glands of the honeybee. Journal of Apicultural Research 7: 29-36. Takenaka, T., Miwa, S. & Echrigo, T. (1990) Changes of protein content and enzyme activity in hypopharyngeal glands during lifespan of honeybee workers (Apis mellifera L.). Bulletin of the Faculty of Agriculture, Tamagawa University 30: 1-7.
This Page Intentionally Left Blank
PART
3
Molecular Phylogenetics
This Page Intentionally Left Blank
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA B. Campbell1, J. Heraty2, J.-Y. Rasplus3, K. Chan1, J. Steffen-Campbell1 and C. Babcock2 1
USDA-ARS, Western Regional Research Centre, Albany, CA 94710 USA
2
Department of Entomology, University of California, Riverside, CA 92521 USA
3
INRA-CNRS, Laboratoire de Modélisation et de Biologie Evolutive, 488 rue Croix de Lavit, F-34090 Montpellier, France
Introduction Chalcidoidea are recognised to contain somewhere on the order of 18,500 described species distributed in 20 families and 89 subfamilies (Noyes 1990; Gibson et al. 1999). Estimates of the number of species range somewhere between 60 000 and 100 000 (Noyes 1978, 1990; Gordh 1979). Ecologically and economically, they are one of the most important groups for control of other insect populations (Noyes 1978; LaSalle 1993). However, after more than 200 years of descriptive work, the taxonomy and classification of Chalcidoidea is still unresolved, frequently revised, and largely lacking a consensus in understanding of monophyly at higher taxonomic levels (Gibson et al. 1999). Monophyly of many higher taxonomic groups, including larger family groups such as Eulophidae, Aphelinidae, Pteromalidae and Eupelmidae, has not been determined (Gibson 1989, 1990, 1995; Noyes 1990; Heraty et al. 1997). As the taxon of focus becomes narrowed to either subfamilies or families with few included genera, the problem is not one of defining monophyly but of positing relationships to other groups of Chalcidoidea. New characters for the analysis of relationships are necessary to solve these problems but only after thorough surveys are undertaken to more accurately estimate the real distribution, homology and transformation of each feature (Heraty et al. 1994, 1997). Until recently, classification of Chalcidoidea has been based on morphological similarities and differences rather than on shared apomorphies. As a result, some families are generally regarded as paraphyletic, if not polyphyletic (Woolley 1988; Gibson 1989, 1990; Noyes 1990). Above the family level, inclusion of Mymarommatidae or Mymaridae in Chalcidoidea has been debated (Kozlov & Rasnitsyn 1979; Rasnitsyn 1980; Gibson 1986), with Mymarommatidae currently excluded as a separate superfamily (Goulet & Huber 1993). The number of chalcidoid families has stabilised at 20 (Goulet & Huber 1993), but disagreement still persists over placement of several subfamilies (e.g. Akapalinae, Calesinae, Chrysolampinae, Eriaporinae, Eunotinae, Philomidinae and Sycoryctinae). The inability to resolve the placement of these taxa into family-level groups using available morphological criteria clearly indicates that new additional character systems are needed. Three major groups of taxa are currently considered in the Chalcidoidea. These groups are usually referred to as the mymarid, eulophid and pteromalid lineages, with the pteromalid lineage often subdivided into chalcidid, torymid and encyrtid sub-lineages. Trichogrammatidae and Aphelinidae are usually placed in the ‘eulophid lineage’ along with Elasmidae, Eulophidae and Signiphoridae. This eulophid group has not been characterised as monophyletic based on
59
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 60
Hymenoptera: Evolution, Biodiversity and Biological Control
definitive characters but rather on a preponderance of reductions (e.g. fewer antennal articles, reduced number of tarsomeres). Combined, the eulophid lineage accounts for 675 genera and 5955 nominal species of Chalcidoidea (Noyes 1978; LaSalle & Gauld 1991; Goulet & Huber 1993). Within the ‘pteromalid lineage’, problems centre around lack of resolution within Pteromalidae, which contains between 32 and 42 subfamilies (Gibson et al. 1999). This group is generally regarded as being paraphyletic or polyphyletic and the typical ‘dumping ground’ for unplaced taxa within the Chalcidoidea. Embedded within this lineage are the Eupelmidae and Encyrtidae. The three subfamilies of Eupelmidae are likely a grade taxon closely related to the pteromalid subfamily Cleonyminae according to Gibson (1990). Monophyly of Encyrtidae is demonstrated by several external (Noyes & Hayat 1994) and internal structures (Heraty et al. 1997). Most genera of Encyrtidae can be assigned to one of the two recognised subfamilies; however, the relationships among tribes within Encyrtidae are unclear (Trjapitzin & Gordh 1978; Noyes & Hayat 1994). Similar problems of classification exist for almost all higher level taxa within the Chalcidoidea. Convergence and extreme divergence of morphological traits lead to much of the taxonomic confusion. In the Chalcidoidea, a large number of similar characters are considered non-homologous, even though they have similar structure; for example, presence of antennal rami, reduced number of tarsomeres, loss of the mesofurcal bridge, presence of a linea calva on the fore wing, and enlargement of the acropleuron (Gibson 1986, 1989; Heraty et al. 1997). Some features, such as the expanded mesopleuron of the Eupelmidae and Encyrtidae, can be dismissed as being convergent on the basis of detailed morphological studies (e.g. Gibson 1989), but others such as the linea calva of Aphelinidae and Encyrtidae are structurally almost identical, causing problems with their classification. In some Aphelinidae and Encyrtidae, a peculiar socketed tooth on the mandible is extremely similar in both structure and function, but must be interpreted as being convergent within these two divergent lineages (Heraty & Schauff 1998). Each of the above features is ‘locally’ important for identification of monophyletic lineages, but at the superfamily level, or ‘globally’, they are presumed convergent. Considerable discussion exists on the classification and placement of taxa within the Chalcidoidea (cf. Bouºek 1988; Hayat 1994; Gibson et al. 1999). However, character-based phylogenetic studies have usually been confined to assessment of relationships within families, often with some reference made to characteristics of other closely related families. A few papers focus on relationships at the family level, but again these discuss relationships or characteristics of only a limited number of families (LaSalle & Noyes 1985; Bouºek & Noyes 1987). At the superfamily level, only one study has focused on a character system across a large representation of taxa (Heraty et al. 1997). Noyes (1990) presented the only ‘dendrogram’ of relationships so far (cf. Heraty et al. 1997), but this was admittedly an intuitive hypothesis and not based on synapomorphies. Some major problems facing a morphologically based phylogenetic analysis of the Chalcidoidea are simply the sheer number of taxa, the extreme diversity of form, and the tendency towards reduction of similar characters in unrelated taxa. Even complex morphological features, such as an enlarged mesopleuron and jumping mechanisms in subfamilies of Eupelmidae, some Aphelinidae, Tanaostigmatidae and Encyrtidae, may be convergent (Gibson 1989). Molecular systematics offers a different set of characteristics that may be used to assess hypotheses of monophyly. Major lineages of Hymenoptera have been surveyed using mitochondrial 16S rDNA sequences (Derr et al. 1992a, 1992b; Dowton & Austin 1994, 1995) to assess earlier hypotheses of relationships based on morphological characters and to evaluate hypotheses of
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 61
single or multiple origins of parasitism within Hymenoptera. Inferences of evolutionary affiliations of different studies using nucleotide sequences have been inconclusive in demonstrating relationships among sawflies but appear to be relatively congruent for relationships among Apocrita (Carmean et al. 1992; Dowton & Austin 1994, 1995). Within these analyses Chalcidoidea have been shown to be monophyletic, but this inference was based upon at most three species (Derr et al. 1992a, 1992b; Dowton & Austin 1994, 1995). Within Chalcidoidea, several studies have begun to address higher relationships but usually with only a few taxa. Different species and populations of Nasonia Ashmead and Trichomalopsis Crawford (Pteromalidae) were analysed using nucleotide sequences of the ITS2 and 28S-D2 regions of the rRNA transcript (Campbell et al. 1993). Using either Trichomalopsis as an outgroup for the ITS2 data or Melittobia Westwood (Eulophidae) as an outgroup for the more conserved 28S data, the relationships among species of Nasonia were the same (N. vitripennis (Walker) + (N. giraulti Darling + N. longicornis Darling)). These results were concordant with phylogenetic trees obtained for their cytoplasmic incompatability bacteria of the genus Wolbachia (Breeuwer et al. 1992). Relationships among several species of Trichogramma Westwood were analysed using ITS2 (Frenk et al. 1996). Machado et al. (1996) used mitochondrial 12S to analyse relationships of the subfamilies Agaoninae, Otitisellinae, Sycoryctinae and Sycophaginae (Agaonidae) with Doryctinae (Braconidae) as an outgroup. Similar results were obtained from analyses of the 28S-D2 region using more reasonable outgroups (Eurytomidae, Figitidae and Ichneumonidae) and other representatives of Agaonidae sensu Bouºek (1988) (at least two each of Epichrysomallinae, Otitisellinae, Sycophaginae, Sycoryctinae and Sycoecinae) (Rasplus et al. 1998). The results suggest that Agaonidae is not monophyletic with Agaoninae having a very distant relationship with other subfamilies currently included in Agaonidae. The D2 expansion region of 28S rDNA was shown previously to have one to six substitutional differences between species of Nasonia, 10–11 between Nasonia and Trichomalopsis (both Pteromalinae) and 42–46 between these Pteromalidae and Melittobia (Eulophidae) (Campbell et al. 1993). This degree of genetic variation was considered to provide an appropriate phylogenetic signal at the generic and family levels, and was chosen for a broader molecular phylogenetic analysis of the Chalcidoidea reported here. N.B. The authors for genera and species are given in Appendix 1.
Materials and Methods Samples Voucher specimens of almost all taxa sampled (see Appendix 1) to date are deposited in the Entomology Research Museum at the University of California, Riverside or the collection at INRA, Montpellier, France. A few specimens of species not commonly collected but easily identified do not have vouchers. Sequences for several eulophid and encyrtid taxa were provided by Donald Quicke (Imperial College, London; identified by a ‘q’ following the generic name in Fig. 1), and sequences for Uscana and Trichogramma fuentesi Torre were provided by Richard Stouthammer (Wageningen Agricultural University, The Netherlands). A goal of the taxon sampling was to obtain sequences for two or more representatives of each higher taxonomic group (tribe or subfamily). Of the 109 species analysed to date, six belong to the outgroups Cynipoidea and Scelionidae, and one species, N. vitripennis, was duplicated in the analysis from different populations (United States and France), and sequenced independently in the Campbell and Rasplus
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 62
Hymenoptera: Evolution, Biodiversity and Biological Control
laboratories. Eighteen of the 20 families of Chalcidoidea are represented, but only 39 (43.8%) of the 89 subfamilies.
Methods Specimens were killed and stored by freezing at –85°C or were collected into 80–100% EtOH. DNA extractions followed the methods outlined in Campbell et al. (1993). PCR was performed in 25µl reactions using a GeneAmp® Kit (Perkin Elmer Cetus, Norwalk, Conn.). The majority of samples were cloned using a plasmid vector (Invitrogen TA Cloning™ System), but 28 taxa were sequenced directly. Primer sequences for PCR amplification of the D2 expansion region of 28S rDNA and direct sequencing are: forward primer 5'-CGT GTT GCT TGA TAG TGC AGC-3', and reverse primer 5'-TTG GTC CGT GTT TCA AGA CGG-3'. Sequencing was done initially using 33P-dATP based autorads on a Genomyx thermal sequencer and later using an infrared dye system on a LI-COR 4200 automated sequencer. The universal Sp6 and T7 primers were used to sequence plasmid 28S-D2 inserts. Both top and bottom strands were completely sequenced. Nucleotide sequences were aligned initially using the ClustalW subprogram on MacVector v 6.5 (Oxford Molecular) with the majority of remaining sequences aligned manually. Gaps were treated as missing values. Phylogenetic analyses were performed on PAUP 4.0b2a (Sinauer Associates, Inc.) using the random addition sequence search algorithm with 25 replicates through three iterations beyond the point where there was no change in tree length. Each iteration was started using the seed number from the shortest tree of the previous iteration. Successive approximations character weighting (Carpenter 1988) was performed on the shortest tree using the maximum value of the rescaled consistency index and a base weight of 1000; hereafter referred to as the reweighted analyses. Successive iterations did not produce a stable result (increasing length with each iteration), and the final tree was selected after four iterations, the last three of which produced a tree of the same length (the same tree) when all characters were reweighted to unity.
Results The aligned 28S-D2 data matrix consisted of 863 bases, of which 299 were constant and 158 autapomorphic. A total of 406 informative sites were found among chalcidoids, which is more than found in the mitochondrial 16S region (217 across Apocrita; Dowton & Austin 1995). One tree was recovered from the parsimony analysis with a length of 3437 steps, consistency index of 0.274 and a retention index of 0.529 (Fig. 1). After successive weighting, a single tree of 3457 steps with a slightly lower consistency index (0.272) and retention index (0.525) was obtained. The reweighted tree was different from the most parsimonious tree, however, many of the relationships were recovered in both trees (bold lines, Fig. 1). The two major apical clades indicated in Figure 1 were recovered in both analyses but three taxonomic groups (Cleonyminae + Chalcidinae, Eucharitidae, and Eusandalum) shifted to the other of these two clades in the reweighted tree (marked by an asterisk). In all analyses (including our studies with fewer taxa or different alignments), Mymaridae are the sister group to remaining Chalcidoidea and Cales (Calesinae; currently unplaced at family level) is positioned basally. Of the genera represented by more than one species, none of the included species were misplaced into non-congeneric taxa. Forty-four species representing 18 genera were, respectively, placed as monophyletic in both unweighted and reweighted parsimony analyses. Species of six genera (Brachymeria, Eurytoma, Megastigmus, Nasonia, Podagrion and Spalangia) were sequenced
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 63
Leptopilina Xyalophora CYNIPOIDEA Diplolepis Gryon SCELIONIDAE Telenomus Trissolcus Polynema MYMARIDAE Gonatocerus 1 Gonatocerus 2 CALESINAE Cales Coccophagoides Encarsia aurantii Encarsia luteola Encarsia formosa APHELINIDAE: COCCOPHAGINAE: PTEROPTRICINI Encarsia protransvena Encarsiella Encarsia lutea Encarsia pergandiella PTEROMALIDAE: EUTRICHOSOMATINAE Eutrichosoma Coccophagus rusti APHELINIDAE: COCCOPHAGINAE: COCCOPHAGINI Coccophagus scutellaris APHELINIDAE: AZOTINAE Ablerus Eretmocerus 19 Eretmocerus 05 APHELINIDAE: APHELININAE: ERETMOCERINI Eretmocerus 01 Eretmocerus 03 Eretmocerus 02 Aphelinoidea Oligosita Uscana TRICHOGRAMMATIDAE Trichogramma fuentes Trichogramma pretiosum Tichogramma platneri Glyphomerus TORYMIDAE: TORYMINAE Podagrion 1 Podagrion 2 Tanaostigmodes TANAOSTIGMATIDAE Sycophila EURYTOMIDAE: EURYTOMINAE Eurytoma 1 Eurytoma 2 TORYMIDAE: TORYMINAE Torymus Aphytis melinus APHELINIDAE: APHELININAE: APHYTINI Aphytis yanonensis Aphelinus asychis APHELINIDAE: APHELININAE: APHELININI Aphelinus albipodus Aphelinus varipes Hockeria Psilocharis CHALCIDIDAE: HALTICHELLINAE Schwarzella 1 Schwarzella 2 EUPELMIDAE: EUPELMINAE Eusandalum PTEROMALIDAE: EUNOTINAE Eunotus Megastigmus 1 TORYMIDAE: MEGASTIGMINAE Megastigmus 2 Aenasius q ENCYRTIDAE + Metaphycus SIGNIPHORIDAE Chartocerus Copidosoma Comperiella encyrtid Ooencyrtus
*
*
10 changes
Figure 1
Cleonymus Epistenia Conura 3 Conura 1 Conura 2 Leucospis
PTEROMALIDAE: CLEONYMINAE CHALCIDIDAE: CHALCIDIDINAE LEUCOSPIDAE
Spalangia 1 PTEROMALIDAE: SPALANGIINAE Spalangia 2 Acanthochalcis CHALCIDIDAE: BRACHYMERIINAE Brachymeria 1 Brachymeria 2 Eupelmus EUPELMIDAE: EUPELMINAE Asaphes PTEROMALIDAE: ASAPHINAE Epiclerus TETRACAMPIDAE Cirrospilus Entedon Horismenus q Chrysocharis 1 Chrysocharis 2 EULOPHIDAE Henryana q Melittobia Aprostocetus q Elachertus q Deutereulophus q Elasmus ELASMIDAE Pachycrepoideus PTEROMALIDAE: PTEROMALINAE Colotrechnus PTEROMALIDAE: COLOTRECHNINAE Seres AGAONIDAE: AGAONINAE Aepocerus AGAONIDAE: OTITISELLINAE Muscidifurax Trichomalopsis Nasonia longpetiolata PTEROMALIDAE: PTEROMALINAE Nasonia giraulti Nasonia vitripennis a Nasonia vitripennis b Calosota Brasema 2 Brasema 1 Chrysomalla PERILAMPIDAE: CHRYSOLAMPINAE Dirhinus CHALCIDIDAE: DIRHININAE Rileya 1 EURYTOMIDAE: RILEYINAE Rileya 2 PTEROMALIDAE: EUNOTINAE Scutellista Obeza EUCHARITIDAE Orasema Idioporus PTEROMALIDAE: EUNOTINAE Perilampus 1 PERILAMPIDAE: PERILAMPINAE Perilampus 2
*
Single most parsimonious tree (phylogram) using the 28S-D2 region (Length 3437 steps, consistency index 0.27, retention index 0.53). Bold lines represent branches supported by both parsimony and successive approximations weighting analyses. Thin lines were supported only in the parsimony analysis. Taxonomic groups indicated by shaded bars. Asterisk indicates clades or taxa that join a different major clade in the reweighted tree.
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 64
Hymenoptera: Evolution, Biodiversity and Biological Control
separately as blind tests to check for sequence fidelity in the Campbell and Rasplus laboratories. The molecular data set also placed the six species representing Encarsia as monophyletic, but this clade also included Encarsiella, a relationship which is supported by another study incorporating more species of Encarsia (Babcock & Heraty unpublished). Of 12 families represented by more than one genus, three were monophyletic: Mymaridae, Trichogrammatidae and Eucharitidae. Eulophidae was monophyletic but included Elasmus (Elasmidae) grouped within the Eulophinae. The placement of Elasmus within Eulophinae rather than as a sister group of Eulophidae was proposed initially by John Noyes (pers. comm.) based on similar host characteristics. The inclusion of this and additional species of Elasmus within Eulophinae has been supported in a more complete analysis of eulophid genera using the same genetic region (Gauthier et al. 2000). Encyrtidae was also monophyletic, but included Chartocerus (Signiphoridae). We have attempted to verify the sequence of Chartocerus with that of two other species of Signiphora but have had difficulties with amplifying the region. Until the sequence is verified this placement is tentative. Of subfamilies represented by multiple genera, seven were monophyletic: Aphelininae (based only on Aphytini and Aphelinini), Haltichellinae, Cleonyminae, Brachymeriinae (sensu lato), Tetrastichinae and Entedoninae, and Pteromalinae (excluding Pachycrepoideus). The Eulophinae (Cirrospilus, Deuteroeulophus and Elachertus) was polyphyletic in the unweighted analysis but monophyletic (including Elasmus) in the reweighted analysis. Not all higher taxa were coherently resolved. None of the three genera of Eunotinae (Scutellista, Eunotus and Idioporus) grouped together. Toryminae were separated into two groups (Glyphomerus + Podagrion and Torymus) in both the unweighted and reweighted analyses. Within Eupelmidae, neither Eupelminae (Eupelmus and Brasema) or Calosotinae (Calosota and Eusandalum) formed a subfamily grouping, although Brasema formed a group with Calosota. The Coccophaginae (Aphelinidae) were split into a paraphyletic Pteroptricini (Coccophagoides, Encarsia and Encarsiella) at the base of the tree, and Coccophagini (Coccophagus) which was grouped with Eutrichosoma (Pteromalidae: Eutrichosomatinae) on both trees; the latter grouping being an unlikely hypothesis. Of three families expected to be monophyletic based on a consensus of findings from morphological studies, Torymidae (Toryminae and Megastigminae), Chalcididae (Chalcidinae, Brachymeriinae, Haltichellinae and Dirhininae) and Eurytomidae (Rileyinae and Eurytominae) were not. Also, Chrysolampinae, which are generally assigned to either Pteromalidae or Perilampidae, were not affiliated with either group. Eupelmidae were scattered across the tree and showed no affinities with Cleonyminae, Tanaostigmatidae or Encyrtidae. The various tribes of Aphelinidae also were scattered across the tree, as were the Pteromalidae. However, as might be expected for such a diverse assemblage, the latter result was not unexpected. Surprisingly, the subfamilies of Chalcididae showed no immediate common affiliation with each other, although genera within each subfamily grouped together. The higher level relationships, basically the backbone of the cladogram, are generally unstable and can change if fewer taxa or different alignments are considered. Different rearrangements usually correspond with relationships supported only in the parsimony tree (thin branches, Fig. 1), although even some of the well-supported relationships (bold branches) can change. Some relationships are very stable even with different taxa or alignments. Across different analyses, the relationship of Ablerus (Azotinae) and Eretmocerus + Trichogrammatidae remain unchanged. Also, Cales (Calesinae) and the Pteroptricini are usually placed basally, although in some analyses these are replaced as basal by Perilampus and Idioporus. Some relationships
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 65
change depending on the taxa being included. For example, a sister group relationship between Tanaostigmatidae and Encyrtidae (LaSalle & Noyes 1985) was not recovered in this analysis, but did occur in some of the earlier analyses with fewer taxa. Adding more taxa may recover this relationship. Notably, across the Chalcidoidea, the number of base changes does not always correlate with consistency of relationships. For example, the clade formed by Pteromalinae, Colotrechninae, Agaoninae and Otitisellinae has relatively few substitutional differences but the relationships remain consistent in all analyses, especially in the clade that includes Aepocerus and Nasonia.
Discussion An important goal of this study is to represent all of the terminal taxa of interest (families, subfamilies or tribes) by two or more divergent taxa, as proposed by Wheeler et al. (1993). Where we have a high degree of confidence in the relationships of closely related taxa based on morphological evidence, for example two unequivocally placed genera of the same subfamily, they should group together on trees produced by molecular evidence (cf. Patterson et al. 1993). If not, then we need to either re-evaluate our initial assumptions of relationship or quality of the molecular data. If we can accept our initial assumptions of relationships, then informative nucleotide changes are best if they are shared by all included members of a taxon (synapomorphic) and, potentially, worst if they are shared by few members of that same taxon and a distantly related taxon (homoplastic). Presumably, derived character states shared by divergent groups within a taxon are more likely to be indicative of synapomorphic changes rather than autapomorphic changes. Results from the 28S-D2 data set are encouraging and indicate that the region is appropriate for analysis of chalcidoid taxa, which may have diverged in the Upper Cretaceous or early Tertiary. Including outgroup taxa, 80% (88/110) of the species are placed into some form of realistic grouping (generic or family group taxon) based on morphological evidence. From our analysis, there was no support for the so-called eulophid or pteromalid lineages. Fourteen subfamilies are represented by only a single species and essentially remain unverified in the data set. The 28S-D2 dataset shows greatest support at the subfamily and generic levels, although five families, Mymaridae, Encyrtidae (+ Chartocerus), Eulophidae (+ Elasmus), Eucharitidae, and Trichogrammatidae are placed as monophyletic or paraphyletic with one other taxon. Based on morphological evidence, four families expected to be monophyletic (Chalcididae, Eurytomidae, Perilampidae and Torymidae) were not. Eunotinae was the only subfamily not having any of the species grouped together. In part this lack of a coherent relationship may be justified based on recent discussions on the placement of Idioporus, in which they were put into Eunotinae as a distinct tribe, but only after consideration of shared characteristics with Eriaporinae (unplaced at the family level), Aphelinidae and Eulophidae (LaSalle et al. 1997). The genera of Eulophidae were all placed into a monophyletic group with monophyly of the Tetrastichinae and Entedoninae supported, while Eulophinae were supported only in the reweighted analysis. Elasmidae are usually placed as the sister group of Eulophidae based on a similar reduction of antennal segments and a simple calcar. The association of Elasmidae with Eulophinae has not been proposed on morphological characters, but instead due to a similar habit of attacking leaf-mining Lepidoptera (Noyes pers. comm.). The results using molecular data support use of character reductions for defining Eulophidae.
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 66
Figure 2
Hymenoptera: Evolution, Biodiversity and Biological Control
Secondary structure model for D2 expansion segment of 28S rRNA of Psilochalcis sp. (Chalcididae: Haltichellinae). Overall structure inferred from thermodynamic folding of the full sequence of the D2 region using mfold version 2.3. Individual substructures within subdomains were confirmed by folding respective nucleotide sequences using the RNA folding program in MacDNASIS and examination of compensatory substitutions (Roussett et al. 1991). Nomenclature of subdomains and helices are according to Michot & Bachellerie (1987).
Some groups, such as Aphelinidae, Eupelmidae and Pteromalidae are not supported by morphological synapomorphies (Heraty et al. 1997; Gibson et al. 1999). Thus, absence of monophyly for these groups in an analysis using molecular data is not unwarranted, although Eupelmidae should have been grouped together at least as a paraphyletic assemblage (Gibson 1995). Generally, tribes of Aphelinidae are monophyletic except for Pteroptricini, which was paraphyletic at the base of the tree. Only Aphytini and Aphelinini were monophyletic. A sister group relationship between Eretmocerus and Trichogrammatidae was supported in all molecular analyses to date, including a separate analysis using the highly variable region synonymous to the E21 helix of 18S rRNA (Campbell & Heraty unpublished). Eretmocerus has usually been placed in the Aphelininae, recently within a separate tribe, the Eretmocerini (Hayat 1998). However, Shaffee and Rizvi (1990) proposed a classification in which Eretmocerus was closer to Trichogrammatidae, and Heraty et al. (1997) found shared characteristics of the mesofurca between the two groups. We are unaware of any morphological support for a sister group relationship between Ablerus (Azotinae) and Eretmocerus + Trichogrammatidae as inferred on the molecular tree. Azotinae is usually placed as closer to Signiphoridae (Woolley 1988), Coccophaginae (Hayat 1994) or Aphelininae (Heraty et al. 1997). Analysing the 28S-D2 data within a more restricted taxonomic range may be better for addressing relationships among some of the problematic taxa. For example, Chalcididae are a demonstrably monophyletic group sharing several derived characteristics, including a plate-like labrum, non-overlapping clypeus, reduced prepectus, lateral scutellar arch and enlarged hind
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 67
femora (Wijesekara 1997). Few systematists would doubt the monophyly of included members, yet none of the subfamilies formed a coherent ‘chalcidid’ assemblage in our molecular based analysis including all chalcidoids. However, monophyly of Haltichellinae was supported, and relationships of included genera match those based on morphological hypotheses (Wijesekara 1997). Dirhininae was represented by only one species in this analysis. The three species of Conura in our analysis are members of the side-group of species within Chalcidini and are not representative of the diversity within the tribe. Brachymeria and Acanthochalcis were previously treated as the Brachymeriinae, but were recently allocated to Chalcidinae as Brachymeriini and Cratocentrini (Boucek 1988). Recently each was elevated to subfamily status by Wijesekara (1997). Results of these analyses suggest Acanthochalcis and Brachymeria are closely related, but their relationship to Chalcidini is uncertain. An effort was made to analyse the chalcidid taxa independent of other chalcidoids, but including traditionally accepted sister groups of Eurytomidae and Leucospidae. In this analysis, the 28S-D2 sequences were aligned so that homologous base positions were matched to the greatest extent possible. This alignment was based on homology of secondary structure and substructures of rRNA synonymous to the D2 region. Secondary structures were inferred based upon thermodynamic folding using mfold v.2.3 (Zucker et al. 1999) via the mfold server (http://mfold1. wustl.edu/~mfold/rna/form1.cgi). Confirmation of certain inferred substructures using shorter sequences was made with the RNA-folding subprogram of MacDNASIS® (Hitachi Software). The overall secondary structure of the chalcidid D2 region (Fig. 2) corresponded to three subdomains inferred for Drosophila (Linares et al. 1991; Rousset et al. 1991) and aphidiines (Belshaw & Quicke 1997). The fewer number of taxa examined in this subset of the chalcidoid dataset enabled a more robust alignment of homologous nucleotide positions. These positions could be rigorously ascertained according to homology of subdomain structure and retention of certain substructures (bulges, loops, etc), which show much broader variation among all chalcidoids. Moreover, it was determined that alignment based on ‘similarity’ using ClustalW did not always align homologous positions according to secondary structure. Despite this independent rigor given to sequences of Chalcididae, Leucospidae and Eurytomidae, the same problems of non-monophyly occurred. Eurytomids were polyphyletic and the haltichellines were the sister to all other chalcids and the leucopsid. When these taxa were scored according to the morphological character matrix presented by Wijesekara (1997) a monophyletic eurytomid was sister to Leucospidae + Chalcididae, with internal arrangement of Chalcididae almost concordant with that of Wijesekara (1997) and Haltichellinae as the distal lineage. Character analysis of the morphological and nucleotide datasets provided some explanation for the different topologies generated. While the morphological dataset provided three synapomorphies supporting Chalicididae, there were zero in the nucleotide dataset. The morphological dataset provided one synapomorphy to support Leucospidae + Chalcididae and the nucleotide dataset provided zero. Conversely, the nucleotide dataset provided three molecular synapomorphies supporting Brachymeria + Acanthochalcis (Brachymeriinae sensu lato), while there were no morphological synapomorphies for this clade. Haltichellinae was supported by both datasets, as was also the Hybothoricini, suggesting that current haltichelline taxonomic groupings are probably accurate. Interestingly, when both datasets were combined a currently preferred set of relationships (except for equivocal placement of Brachymeria and Acanthochalcis) was resolved (Campbell & Heraty unpublished). The observed congruence between the molecular hypotheses generated from the 28S-D2 region and the accepted morphological-based classifications for some of the included taxa increase our faith in this region as a reasonable estimator of relationships for some families and almost all
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 68
Hymenoptera: Evolution, Biodiversity and Biological Control
subfamilies. However, we must regard this analysis as preliminary. Better resolution of some groups, such as Chalcididae, Eurytomidae and Torymidae, will be necessary before we can begin to accept relationships postulated for higher level taxa. Is it a case of adding more molecules (new regions), adding morphological data, or adding more taxa? Additional genes, based on the same limited sampling of taxa, are likely to duplicate inadequacies of the 28S data set. Adding morphological characters to the matrix is obviously a necessary step towards fully understanding the relationships of this group. However, less than 45% of chalcidoid subfamilies are represented in the current analysis. For Pteromalidae, only five of 32 subfamilies are included. Within some groups, representation is poor. For example, sequences of Chalcidini are based only on members of one clade, the Conura side-group, and Perilampidae is represented only by two species of the Perilampus fulvicornis-group. Other studies demonstrate that an adequate sampling of taxa is most important to provide a proper estimate of the phylogeny (Graybeal 1998; Poe 1998). Hence, future emphasis on increased taxon sampling, particularly for the Pteromalidae, Chalcididae, Eurytomidae and Torymidae may be the best approach for further resolving higher-level relationships of Chalcidoidea.
Summary Based on cladistic analysis of nucleotide sequences of the 28S-D2 expansion region we find: 1) Chalcidoidea are monophyletic, and Mymaridae are the sister group to the remaining Chalcidoidea. This relationship agrees with recent morphology-based hypotheses (Gibson et al. 1999). 2) Eulophidae (including Elasmus) are monophyletic. This is consistent with morphological data, although hypotheses of monophyly are based only on reductions or losses of characters (number of tarsal segments and flagellomeres and reduced fore tibial spur). 3) Elasmidae are closest to the Eulophinae within the Eulophidae. 4) Eretmocerus and Trichogrammatidae are sister groups. This relationship was proposed by Shafee & Rizvi (1990) without supporting character evidence, and again more recently by Heraty et al. (1997) using similar structure of the mesofurca. 5) The families Mymaridae, Eucharitidae and Trichogrammatidae, and the subfamilies Aphelininae (excluding Eretmocerini), Cleonyminae, Haltichellinae and Brachymeriinae (sensu lato) are all supported as monophyletic using more than one generic exemplar. 6) Encyrtidae is rendered paraphyletic by the genus Chartocerus (Signiphoridae) in all analyses. The Chartocerus sequence has been checked with other specimens, but an additional sequence from another genus of signiphorid is needed to verify its placement. 7) Aphelinidae is never supported as monophyletic. Aphelininae (excluding Eretmocerus) are monophyletic in all analyses, but the monophyly of the Coccophaginae (as represented by Encarsia, Coccophagoides, Coccobius and Coccophagus) is transient. No morphological evidence supports the monophyly of Aphelinidae (Gibson et al. 1999). 8) Pteromalidae are not monophyletic. However, only five of 32 subfamilies are represented. 9) Chalcididae, as represented by Brachymeriinae (Brachymeria, Acanthochalcis), Chalcidinae (Conura) and Haltichellinae (Psilocharis, Hockeria, Schwarzella), is not monophyletic. This was a surprising result considering strong morphological evidence supporting this
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 69
assemblage (Wijesekara 1997). The combination of molecular and morphological data sets resolved relationships to current consensus. 10) Agaonidae is polyphyletic in accordance with recent findings of Rasplus et al. (1998).
Acknowledgements We appreciate Dave Hawks, Jung-Wook Kim and Michael Gates (UCR) for their assistance in acquiring specimens and sequences. Numerous people have helped to supply specimens, but in particular we would like to thank Donald Quicke (Imperial College, London), John LaSalle (IIE, London), Nathan Schiff (USDA, Stoneville, MS), Molly Hunter (University of Arizona, Tucson), Les Ehler (University of California, Davis), and Tom Bellows, Dan Gonzalez, Greg Walker and Robert Luck (University of California, Riverside). We thank Anura Wijesekara (Horticulture Research and Development Institute, Sri Lanka) for insightful discussions.
References Belshaw, R. & Quicke, D. L. J. (1997) A molecular phylogeny of the Aphidiinae (Hymenoptera: Braconidae). Molecular Phylogenetics & Evolution 7: 281-293. Boucˇek, Z. (1988) Australasian Chalcidoidea (Hymenoptera): A Biosystematic Revision of Genera of Fourteen Families, with a Reclassification of Species. C.A.B. International, Wallingford. Boucˇek, Z. & Noyes, J. S. (1987) Rotoitidae, a curious new family of Chalcidoidea (Hymenoptera) from New Zealand. Systematic Entomology 12: 407-412. Breeuwer, J. A. J., Stouthamer, R., Burns, S. M., Pelletier, D. A., Weisberg, W. G. & Werren, J. H. (1992) Phylogeny of cytoplasmic incompatibility microorganisms in the parasitoid was genus Nasonia (Hymenoptera: Pteromalidae) based on 16S ribosomal DNA sequences. Insect Molecular Biology 1: 25-36. Campbell, B., Steffen-Campbell, J. D. & Werren, J. H. (1993) Phylogeny of the Nasonia species complex (Hymenoptera: Pteromalidae) inferred from an internal transcribed spacer ITS2 and 28S rDNA sequences. Insect Molecular Biology 2: 225-237. Carmean, D., Kimsey, L. S. & Berbee, M. L. (1992) 18S rDNA sequences and the holometabolous insects. Molecular Phylogenetics & Evolution 1: 270-278. Carpenter, J. (1988) Choosing among equally parsimonious cladograms. Cladistics 4: 291-296. Derr, J. N., Davis, S. K., Woolley, J. B. & Wharton R. A. (1992a) Variation and the phylogenetic utility of the large ribosomal subunit of mitochondrial DNA from the insect order Hymenoptera. Molecular Phylogenetics & Evolution 1: 136-147. Derr, J. N., Davis, S. K., Woolley, J. B. & Wharton R. A. (1992b) Reassessment of the 16S rRNA nucleotide sequence from members of the parasitic Hymenoptera. Molecular Phylogenetics & Evolution 1: 338-341. Dowton, M. & Austin, A. D. (1994) Molecular phylogeny of the insect order Hymenoptera: Apocritan relationships. Proceedings of the National Academy of Sciences, USA 91: 9911-9915. Dowton, M. & Austin, A. D. (1995) Increased genetic diversity in mitochondrial genes is correlated with the evolution of parasitism in the Hymenoptera. Journal of Molecular Evolution 41: 958-965. Frenk, J. P. M., Silva, I., Schilthuizen, M., Pinto, J. D., & Stouthamer, R. (1996) Use of DNA-based methods for the identification of minute wasps of the genus Trichogramma. Proceedings of Experimental and Applied Entomology N.E.V. Amsterdam 7: 233-237.
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 70
Hymenoptera: Evolution, Biodiversity and Biological Control
Gauthier, N., LaSalle, J., Quicke, D. L. J. & Godfray, H. C. J. (2000) Phylogeny of the Eulophidae (Hymenoptera: Chalcidoidea), with a reclassification of the Eulophinae and the recognition that the Elasmidae are derived eulophids. Systematic Entomology 25: 1-20. Gibson, G. A. P. (1986) Evidence for monophyly and relationships of Chalcidoidea, Mymaridae, and Mymarommatidae (Hymenoptera: Terebrantes). Canadian Entomologist 118: 205-240. Gibson, G. A. P. (1989) Phylogeny and classification of Eupelmidae, with a revision of the world genera of Calosotinae and Metapelmatinae (Hymenoptera: Chalcidoidea). Memoirs of the Canadian Entomological Society 149: 121 pp. Gibson, G. A. P. (1990). A word on chalcidoid classification. Chalcid Forum (newsletter) 13: 7–9. Gibson, G. A. P. (1995) Parasitic wasps of the subfamily Eupelminae: Classification and revision of World genera (Hymenoptera: Chalcidoidea: Eupelmidae). Memoirs of Entomology, International 5: 1-421. Gibson, G. A. P., Heraty, J. M. & Woolley, J. B. (1999) Phylogenetics and classification of Chalcidoidea and Mymarommatoidea – a review of current concepts (Hymenoptera, Apocrita). Zoologica Scripta 28: 87-124. Gordh, G. (1979) Encyrtidae. pp. 890-966. In Krombein, K. V., Hurd, B., Smith, D. R. & Burks, B. D. (Eds), Catalog of Hymenoptera in America North of Mexico. Volume 1. Smithsonian Institution Press, Washington. Goulet, H. & Huber, J. T. (1993) Hymenoptera of the World: An Identification Guide to Families. Agriculture Canada Publication 1894/E. Graybeal, A. (1998) Is it better to add taxa or characters to a difficult phylogenetic problem? Systematic Biology 47: 9-17. Hayat, M. (1994) Notes on some genera of the Aphelinidae (Hymenoptera: Chalcidoidea), with comments on the classification of the family. Oriental Insects 28: 81-96. Hayat, M. (1998) Aphelinidae of India (Hymenoptera: Chalcidoidea): A taxonomic revision. Memoirs on Entomology, International 13: 416 pp. Heraty, J. M. & Schauff, M. E. (1998) Mandibular teeth in Chalcidoidea: function and phylogeny. Journal of Natural History 32: 1227-1244. Heraty, J. M., Woolley, J. B. & Darling, D. C. (1994) Phylogenetic implications of the mesofurca and mesopostnotum in Hymenoptera. Journal of Hymenoptera Research 3: 241-277. Heraty, J. M., Woolley, J. B. & Darling, D. C. (1997) Phylogenetic implications of the mesofurca and mesopostnotum in Chalcidoidea (Hymenoptera), with emphasis on Aphelinidae. Systematic Entomology 22: 45-65. Kozlov, M. A. & Rasnitsyn, A. P. (1979) On the limits of the Serphitidae (Hymenoptera, Proctotrupoidea). Entomologicheskoe Obozrenie 58: 402–416. [in Russian] LaSalle, J. (1993) Parasitic Hymenoptera, biological control and the biodiversity crisis. pp. 197216. In LaSalle, J. & Gauld, I. D. (Eds), Hymenoptera and Biodiversity. C.A.B. International: Wallingford. LaSalle, J. & Gauld, I. D. (1991) Parasitic Hymenoptera and the biodiversity crisis. Redia 74: 315334. LaSalle, J. & Noyes, J. S. (1985) New family placement for the genus Cynipencyrtus (Hymenoptera: Chalcidoidea: Tanaostigmatidae). Journal of the New York Entomological Society 93: 1261-1264. LaSalle, J., Polaszek, A., Noyes, J. S. & Zolnerowich, G. (1997) A new whitefly parasitoid (Hymenoptera: Pteromalidae: Eunotinae), with comments on its placement, and implications for classification of Chalcidoidea with particular reference to the Eriaporinae (Hymenoptera: Aphelinidae). Systematic Entomology 22: 131-150.
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 71
Linares, A. R., Hancock, J. M. & Dover, G. A. (1991) Secondary structure constraints on the evolution of Drosophila 28S ribosomal RNA expansion segments. Journal of Molecular Biology 219: 381-390. Machado, C. A., Herre, E. A., McCafferty, S. & Berminham, E. (1996) Molecular phylogenies of fig pollinating and non-pollinating wasps and the implications for the origin and evolution of the fig-fig wasp mutualism. Journal of Biogeography 23: 531-542. Michot, B. & Bachellerie, J. P. (1987) Comparison of large subunit rRNAs reveal some eukaryotespecific elements of secondary structure. Biochimie 69: 11-23. Noyes, J. S. (1978) On the numbers and species of Chalcidoidea (Hymenoptera) in the world. Entomologist’s Gazette 29: 163-164. Noyes, J. S. (1990) A word on chalcidoid classification. Chalcid Forum (newsletter) 13: 6-7. Noyes, J. S. & Hayat, M. (1994) Oriental Mealybug Parasitoids of the Anagyrini (Hymenoptera: Encyrtidae). CAB International, Wallingford. Patterson, C., Williams, D. M. & Humphries, C. J. (1993) Congruence between molecular and morphological phylogenies. Annual Review of Ecology & Systematics 24: 153-188. Poe, S. (1998) Sensitivity of phylogeny estimation to taxonomic sampling. Systematic Biology 47: 18-31. Rasplus, J. Y., Kerdelhué, C. & Mondor, G. (1998) Molecular phylogeny of fig wasps (Hymenoptera). Agaonidae is not monophyletic. Compte Rendu de l’Académie des Sciences de Paris 321: 21-31. Rasnitsyn, A. P. (1980). Origin and evolution of hymenopterous insects. Trudy Paleontologicheskogo Instituta Akademiya Nauk SSSR 174: 1–192. [in Russian] Rousset, R., Pélandakis, M. & Solignac, M. (1991) Evolution of compensatory substitutions through G•U intermediate state in Drosophila rRNA. Proceedings of the National Academy of Sciences, USA 88: 10032-10036. Shafee, S. A. & Rizvi, S. (1990) Classification and phylogeny of the family Aphelinidae (Hymenoptera: Chalcidoidea). Indian Journal of Systematic Entomology 7: 103–115. Trjapitzyn, V. A. & Gordh, G. (1978) Review of the genera of Nearctic Encyrtidae (Hymenoptera: Chalcidoidea). II. Entomological Review 57: 437-448. Wheeler, W. C., Cartwright, P. & Hayashi, C. (1993) Arthropod phylogeny: a combined approach. Cladistics 9: 1-39. Wijesekara, A. (1997) A phylogenetic analysis of the Chalcididae (Hymenoptera: Chalcidoidea). Memoirs of the American Entomological Institute 29: 1-61. Woolley, J. B. (1988) Phylogeny and classification of the Signiphoridae (Hymenoptera: Chalcidoidea). Systematic Entomology 13: 465-501. Zuker, M., Matthews, D. H. & Turner, D. H. (1999) Algorithms and Thermodynamics for RNA Secondary Structure Prediction: A Practical Guide in RNA Biochemistry and Biotechnology. Barciszewski, J. & Clark, B. F. C. (Eds), NATO ASI Series, Kluwer Academic Publishers.
B. Campbell, J. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell and C. Babcock 72
Appendix 1 List of species used in this study showing their family/subfamily placement. Agaonidae Agaoninae Seres Waterston Otitisellinae Aepocerus Mayr Aphelinidae Coccophaginae Coccophagus rusti Compere Coccophagus scutellaris Dalman Coccophagoides Girault Encarsia formosa Gahan Encarsia lutea Masi Encarsia aurantii Howard Encarsia protransvena Viggiani Encarsia luteola Howard Encarsia pergandiella Howard Encarsiella Hayat Aphelininae Aphelinus varipes Foerster Aphelinus albipodus Hayat & Fatima Aphelinus asychis Walker Aphytis melinus DeBach Aphytis yanonensis Debach & Rosen Eretmocerus Haldeman Azotinae Ablerus Howard Calesinae Cales noacki Howard Chalcididae Chalcidinae Conura Spinola Brachymeriinae Brachymeria Westwood Acanthochalcis Cameron Dirhininae Dirhinus Dalman Haltichellinae Hockeria Walker Psilochalcis Kieffer Schwarzella Ashmead
Hymenoptera: Evolution, Biodiversity and Biological Control
Elasmidae Elasmus Westwood Encyrtidae Metaphycus Mercet Copidosoma Ratzeburg Comperiella Howard Aenasius Walker Ooencyrtus Ashmead Zaomma Ashmead Eucharitidae Oraseminae Orasema Cameron Eucharitinae Obeza Heraty Eulophidae Entedoninae Chrysocharis Foerster Horismenus Walker Entedon Dalman Eulophinae Cirrospilus Westwood Elachertus Spinola Deutereulophus Schulz Tetrastichinae Melittobia Westwood Aprostocetus Westwood Henryana Yoshimoto Eupelmidae Eupelminae Brasema Cameron Eusandalum Ratzeburg Eupelmus Dalman Calosotinae Calosota Curtis Eurytomidae Eurytominae Eurytoma Illiger Sycophila Walker Rileyinae Rileya Ashmead
Molecular Systematics of the Chalcidoidea Using 28S-D2 rDNA 73
Leucospidae Leucospis F.
Tanaostigmatidae Tanaostigmodes Ashmead
Mymaridae Polynema Haliday Gonatocerus Nees
Tetracampidae Epiclerus Haliday
Perilampidae Perilampinae Perilampus Latreille Chrysolampinae Chrysomalla Foerster Pteromalidae Asaphinae Asaphes Walker Colotrechninae Colotrechnus Thomson Eunotinae Scutellista Motschulsky Eunotus Walker Idioporus LaSalle & Polaszek Eutrichosomatinae Eutrichosoma Ashmead Spalangiinae Spalangia Latreille Pteromalinae Trichomalopsis Crawford Nasonia vitripennis Walker Nasonia longicornis Darling Nasonia giraulti Darling Muscidifurax Girault & Sanders Pachycrepoideus Ashmead Cleonyminae Cleonymus Latreille Epistenia Westwood Signiphoridae Chartocerus Motschulsky
Torymidae Toryminae Torymus Dalman Podagrion Spinola Glyphomerus Foerster Megastigminae Megastigmus Dalman Trichogrammatidae Trichogramma pretiosum Riley Trichogramma platneri Nagarkatti Trichogramma fuentesi Torre Aphelinoidea Girault Oligosita Walker Uscana Girault
Outgroups Scelionidae Trissolcus Ashmead Gryon Haliday Cynipidae Diplolepis rosae (L.) Eucoilidae Leptopilina boulardi (Barbotin, Carton, & Kelner-Pillault) Figitidae Xyalophora Kieffer
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera): a Simultaneous Molecular and Morphological Analysis D. L. J. Quicke1,2, M. G. Fitton2, D. G. Notton2,3, G. R. Broad1,4 and K. Dolphin1 1
Unit of Parasitoid Systematics, CABI Bioscience UK Centre (Ascot), Department of Biology, & 4Centre for Population Biology, Imperial College at Silwood Park, Ascot, Berkshire SL5 7PY United Kingdom (email:
[email protected])
2 Department of Entomology, The Natural History Museum, London SW7 5BD United Kingdom 3
Present address: Reading Museum Service, Blagrave Street, Reading RG1 7HB United Kingdom
Introduction The Ichneumonidae is one of the most species rich families of insects and it is now widely recognised that the figure of 60 000 species world-wide (Townes 1969) is an oft-repeated underestimate (Gauld 1991). It is currently divided into some 39 subfamilies; Gauld (1997) listed 37 and Porter (1998) has recognised two more. However, despite recent advances in the systematics of the family, many of the larger polythetic groups are still poorly defined and some are undoubtedly paraphyletic or polyphyletic assemblages. To date, no morphological phylogenetic analysis of the whole family has been published. Phylogenetic analyses of some individual subfamilies and a few groups of subfamilies have been undertaken [Labeninae – Gauld 1983; ‘Oxytorinae’ (Microleptinae sensu Townes) – Wahl 1986; Ophioninae – Gauld 1985; Pimplinae (sensu lato) – Eggleton 1989; ‘Pimpliformes’ – Wahl 1990; Wahl & Gauld 1998; Campopleginae – Wahl 1991; Mesochorinae – Wahl 1993a; Labeninae – Wahl 1993b; Xoridinae – Wahl 1997] but none of these works addressed the structure of the family as a whole. Although some major groupings of subfamilies have been proposed, (Wahl 1990, 1991, 1993a; Gauld 1991, 1997) only two of these have been defined monophyletically (Wahl 1990, 1991), and many subfamilies have remained unplaced in this loose classification. Phylogenetic analyses have hitherto concentrated on morphological characters and have been restricted in their applicability at the subfamily level by extensive homoplasy (Gauld & Mound 1982). The molecular phylogeny of Belshaw et al. (1998) was the first formal analysis to include a wide sample of representative ichneumonid subfamilies. They indicated relationships for some of the ‘unplaced’ subfamilies whilst supporting some of the previously suggested groupings. Quicke et al. (1999) also used the D2 variable region of 28S nuclear rDNA in an attempt to identify the most basal extant subfamilies. With the emergence of higher level phylogenetic hypotheses for the Braconidae (Belshaw et al. 1998; Dowton et al. 1998; Dowton & Austin 1998; Dowton 1999), the lack of any reasonable phylogenetic reconstructions of their sister group, the Ichneumonidae, is conspicuous and hampers any comparative analyses of biological strategies, such as those suggested by Gauld (1988). We
74
Hymenoptera: Evolution, Biodiversity and Biological Control
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera) 75
hope to redress the balance somewhat with this work using both molecular and morphological data sets. Hypotheses concerning the evolutionary transitions within the Ichneumonidae (e.g. Gauld 1988) have rested on a priori assumptions concerning biological and morphological features. It is thought that the ancestral ichneumonid was an idiobiont ectoparasitoid of xylophagous coleopterous or siricid larvae, a view that has been given some support by the molecular analysis of putatively basal subfamilies by Quicke et al. (1999). However, some molecular studies on the sister family, Braconidae, have tentatively suggested that their ancestral biology was endoparasitism (Dowton et al. 1998).
Materials and Methods The character matrix included 786 characters in total and was a combination of 123 morphological and life history characters which covered external and internal anatomy including immature stages (taken from the unpublished matrix of Fitton and Quicke) as well as existing manually aligned sequence data sets for the D2 variable region of the nuclear 28S rDNA gene (Belshaw et al. 1998; Quicke et al. 1999). Gaps were coded as missing, all characters were treated as unordered and assigned equal weight. The combined morphological and molecular character matrix, full descriptions of all the characters, and list of summary taxa for the morphological character set, can be found on the Imperial College web site [http://www.bio.ic.ac.uk/research/dlq/ ich_mat.pau and http://www.bio.ic.ac.uk/research/dlq/ morph_ch.rtf]. The 61 taxa included in this study represented 31 of the 39 or so ichneumonid subfamilies, making this the widest ranging phylogenetic study of the family to date. The morphological data referred to summary terminal taxa (usually tribes or subfamilies but sometimes groups of individual genera) to which the species sequenced belonged (details on www, see above). The names of terminal taxa shown on our trees (Figs 1-3) are given in Table 1 with their subfamily, tribal or genus-group placement as applicable. Although morphological data have been collected for all ichneumonid subfamilies and tribes, only those for which we had molecular data were included in the analyses presented here.
Phylogenetic analyses The matrix was partitioned into molecular and non-molecular components, and conflict between the partitions was investigated using the Incongruence Length Difference test (Farris et al. 1994) as implemented in the pre-release PAUP* version 4.0d61 (written by David L. Swofford). All phylogenetic analyses were conducted using PAUP* 4.0d61 as above. Three analyses were undertaken using (i) the molecular, (ii) non-molecular and (iii) combined data sets. The search strategy was as follows [details of the efficiency of this strategy will be presented elsewhere (Quicke et al. in review)]: (a) Heuristic searches were carried out using tree bisection-reconnection (TBR), 100 random addition sequence replicates and unlimited maxtrees. (b) Characters were reweighted by the maximum Rescaled Consistency Index derived from the cladograms resulting from (a), and a further heuristic analysis was carried out using the trees obtained from (a) as starting trees. (c) Using all of the resulting cladograms as starting points, and with all characters reweighted to unity, further TBR rearrangements were performed.
D. L. J. Quicke, M. G. Fitton, D. G. Notton, G. R. Broad and K. Dolphin 76
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Subfamilial and tribal placements of ichneumonid taxa sequenced for this study. The classification in this table shows only the taxa included in the simultaneous analyses. All taxa are those used for the morphological study and were chosen as far as possible to be monophyletic, without any implications as to their rank. The following were not included in the current analyses or classification: Groteini, Ganodes, Rodrigama, Sphinctini, Ankylophonini, Eclytini, Exenterini, Idiogrammatini, Claseinae, Ateleute, Lycorininae, Agriotypinae, Banchus-group, Glyptini, Townesioninae, Panteles, Olethrodotini, Scolobatini, Euryproctini, Belesica-group, Hellwigia, Skiapus, Tatogastrinae, Apolophus, Bremiella, Ischyrocnemis, Lapton, Anomalonini, Coleocentrini, Cylloceriinae, Oxytorinae, Hyperacmus, Brachyscleroma, Melanodolius-group, Phaeogenini. These taxa being those for which we have no 28S rDNA data for any included species; Pedunculinae were included within the Brachycyrtinae for the morphological character set. There is therefore no implication that they are excluded from the subfamily groups recognised above (1not including Belesica-group; 2not including Apolophus, Bremiella, Ischyrocnemis, Lapton; 3not including Brachyscleroma and Melanodolius-group; 4not including Panteles; 5not including Ganodes and Rodrigama; 6not including Ateleute). Subfamily group
Taxon
Genera with molecular data included in analyses
Ophioniformes
Campoplegini Nesomesochorini Cremastinae1 Ophioninae Gravenhorstiini Tersilochinae Paxylommatinae Atrophini Exetastes-group Mesochorinae Ctenopelmatini Perilissini Mesoleiini Pionini Metopiinae2 Neorhacodinae Tryphonini Oedemopsini Phytodietini Phrudinae3 Stilbopinae4
Venturia, Lathrostizus, Dusona Nonnus Pristomerus, indet. cremastine Enicospilus, Eremotylus, Ophion Agrypon indet. tersilochine Hybrizon Lissonota Exetastes Mesochorus Xenoschesis Perilissus, Absyrtus Alexeter Pion Exochus, Hypsicera, Colpotrochia Neorhacodes Cosmoconus, Monoblastus, Polyblastus, Grypocentrus Oedemopsis Netelia Phrudus Stilbops
Orthopelmatiformes
Orthopelmatinae
Orthopelma
Pimpliformes
Poemeniinae5 Rhyssinae Ephialtini Polysphinctini Pimplini Diacritinae Microleptinae Diplazontinae Orthocentrinae Helictinae Acaenitini Collyriinae
Neoxorides, Poemenia, Pseudorhyssa Megarhyssa Ephialtes, Tromatobia, Dolichomitus Schizopyga Apechthis, Itoplectis Diacritus Microleptes Diplazon Orthocentrus Proclitus Phaenolobus Collyria
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera) 77
Table 1 (continued) Subfamily group
Taxon
Genera with molecular data included in analyses
Ichneumoniformes
Eucerotinae Brachycyrtinae Ichneumoninae Cryptinae6 Adelognathinae
Euceros Brachycyrtus Crypteffigies, Alomya Dichrogaster, Nematopodius Adelognathus
Labeniformes
Labenini Poecilocryptini
Labena Poecilocryptus
Xoridiformes
Xoridinae
Xorides
(d) Finally, all of the shortest topologies recovered up to this stage were used as starting points for a final series of TBR rearrangements. In order to assess levels of support of the clades recovered, bootstrapping was undertaken using 10 random addition sequences and 100 bootstrap replicates.
Results and Discussion Figures 1–3 show the trees resulting from analyses of molecular, morphological and simultaneous analysis data sets respectively. Although there was significant conflict between the morphological and molecular data sets (partition homogeneity test, p < 0.01) the signal of neither data set was swamped by the other and the simultaneous analysis tree (Fig. 3) was better resolved than either tree produced by a single data set (Figs 1 and 2) [see Nixon & Carpenter (1996) for discussion]. Two of the groupings of subfamilies are supported by our analyses, namely the Ophioniformes (sensu Wahl 1991) and the Pimpliformes (sensu Wahl 1990). Our own overall interpretations of the major subfamilial groupings are indicated in Figure 3 and Table 1. Two of Gauld’s (1997) groupings, the Tryphoniformes and Labeniformes, both appear as polyphyletic assemblages with convergent or plesiomorphic life histories respectively. The Tryphoninae, with Stilbops and Phrudus, consistently appeared as a basal member of the Ophioniformes indicating a plesiomorphic biological strategy of ectoparasitism within this group. Placements of the Eucerotinae (obligate hyperparasitoids with planidial larvae) and of the Adelognathinae (primitively koinobiont ectoparasitoids) within the Ichneumoniformes (sensu Wahl 1993b) suggest previously unrecognised evolutionary pathways within this group that would certainly be worthy of further study. The position of Microleptes, an ‘unplaced subfamily’ in Gauld (1997), within the Diptera-parasitising clade is concordant with its biology as a parasitoid of stratiomyiid flies although this does not agree with Wahl’s (1986) findings as regards larval morphology. Besides the relocation of the Tryphoninae, changes to the large scale classificatory system are few. Orthopelma was placed by Gauld (1997) in the Labeniformes group but its appearance in our trees varies in the analyses, being placed (with weak support) as the sister group to the Ophioniformes in the simultaneous analysis. It is not surprising that the abovenamed genera (Phrudus, Stilbops, Euceros, Adelognathus, Microleptes and Orthopelma) should have affinities not in agreement with current classificatory systems, considering the vagabond nature of these taxa in previous classificatory schemes. Four of the currently recognised subfamilies are not demonstrably monophyletic on the basis of our analyses. These are the Pimplinae, Tryphoninae, Ctenopelmatinae and the Campopleginae. Aberrant elements may have been pigeon-holed in the latter three subfamilies (Netelia, Pion and
D. L. J. Quicke, M. G. Fitton, D. G. Notton, G. R. Broad and K. Dolphin 78
Hymenoptera: Evolution, Biodiversity and Biological Control
81
97 100
92
86
60
68
56 100
98
75 53
92
66 55
Figure 1
Venturia Lathrostizus Dusona Exochus Hypsicera Ophion ventricosus Ophion obscuratus Eremotylus Perilissus Absyrtus Alexeter Xenoschesis Colpotrochia Mesochorus Pion tersilochine indet. Phrudus Labena Poecilocryptus Hybrizon Pristomerus cremastine indet. Nonnus Enicospilus Agrypon Neorhacodes Cosmoconus Grypocentrus Monoblastus Polyblastus Oedemopsis Netelia Stilbops Lissonota Exetastes Ephialtes Tromatobia Schizopyga Dolichomitus Orthocentrus Proclitus Diacritus Apechthis Itoplectis Phaenolobus Collyria Neoxorides Pseudorhyssa Poemenia Megarhyssa Diplazon Euceros Brachycyrtus Nematopodius Adelognathus Crypteffigies Alomya Dichrogaster Microleptes Orthopelma Xorides
Strict consensus of the 72 most parsimonious cladograms resulting from analysis of the sequence data. Bootstrap values above 50% are indicated (Length=1228; CI=0.398; RI=0.549).
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera) 79
Oedemopsini Brachycyrtinae Diacritinae Adelognathinae Cryptinae
53
Campoplegini Cremastinae
61
52
Diplazontinae Orthocentrus group Microleptinae Helictes group Orthopelmatinae
Alomya Ichneumoninae Metopiinae Eucerotinae Tryphonini Ephialtini Tersilochinae Phrudinae Perilissini Ophioninae Poemeniinae Rhyssinae Pseudorhyssa Pimplini Nonnus Gravenhorstiini Mesochorinae Mesoleiini Ctenopelmatini Pionini Atrophini Exetastes -group Stilbopinae Neorhacodinae Phytodietini Collyriinae Acaenitini Polysphincta group Labenini Poecilocryptini Xoridinae Paxylommatinae
Figure 2
Strict consensus of the 170 cladograms resulting from analysis of the morphological and life-history data. Bootstrap values above 50% are indicated (Length=628; CI=0.275; RI=0.647).
D. L. J. Quicke, M. G. Fitton, D. G. Notton, G. R. Broad and K. Dolphin 80
Hymenoptera: Evolution, Biodiversity and Biological Control
78
1 00
Gauld 1997 System: ● Ophioniformes ❍ Tryphoniformes ■ Pimpliformes ■ Labeniformes ? Unplaced
10 0
99
76
84
68
89 70
97
55 59
1 00
60
86
87
55
51
100
90 75
Figure 3
Ichneumoniformes
71
Pimpliformes
84
Ophioniformes
99
Venturia ● Lathrostizus ● Dusona ● Pristomerus ● cremastine indet. ● Ophion ventricosus ● Ophion obscuratus ● Eremotylus ● Enicospilus ● Agrypon ● tersilochine indet. ● Hybrizon ? Nonnus ● XXXXXXX ● Exetastes ● Mesochorus ● XXXXXXXX ● Absyrtus ● Alexeter ● Xenoschesis ● Exochus ? Hypsicera ? Colpotrochia ? Pion ● Neorhacodes ? Cosmoconus ❍ Monoblastus ❍ Polyblastus ❍ Grypocentrus ❍ Oedemopsis ❍ Phrudus ? Netelia ❍ Stilbops ? Orthopelma ■ Neoxorides ■ Poemenia ■ Pseudorhyssa ■ Megarhyssa ■ Ephialtes ■ Tromatobia ■ Dolichomitus ■ Schizopyga ■ Diacritus ■ Microleptes ? Diplazon ■ Orthocentrus ■ Proclitus ■ Apechthis ■ Itoplectis ■ Phaenolobus ■ Collyria ? Euceros ❍ Brachycyrtus ■ Alomya ■ Crypteffigies ■ Dichrogaster ■ Nematopodius ■ Adelognathus ❍ Labena ■ Poecilocryptus ■ Xorides ■
Strict consensus of the two cladograms resulting from simultaneous analysis of all data. Bootstrap values above 50% are indicated (Length=1962; CI=0.337; RI=0.556).
Nonnus, respectively, representing the tribes Phytodietini, Pionini and Nesomesochorini). The two major tribes of the Pimplinae, the Pimplini and Ephialtini, were found not to be closely related, despite the subfamily being regarded as monophyletic by Wahl and Gauld (1998); the Perithoini and Delomeristini sensu Wahl and Gauld (1998) were not included. Our result is, however, in agreement with Finlayson’s (1967) findings based on larval morphology. Studies
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera) 81
utilising more taxa from these groups will be necessary to delimit more accurately the limits of the subfamilies involved. An interesting finding is the placement of Paxylommatinae (Hybrizon), a troublesome entity that has been placed variously in the Braconidae, Ichneumonidae and in its own family, the Paxylommatidae (Sharkey & Wahl 1992). It falls within the Ophioniformes lineage in our analyses, its systematic position previously having been clouded by its suite of autapomorphies. In light of this, it would be interesting to know whether it is a koinobiont endoparasitoid, as are the members of its putative sister-group, the Tersilochinae. Suggestions that it is a very basal ichneumonid on molecular considerations (Belshaw et al. 1998; Quicke et al. 1999) result from the lack of phylogenetic signal in the Hybrizon 28S gene sequence, long branch attraction, and the use in the latter paper of a reduced subset of taxa. Because of a lack of braconid morphological data for the characters used in this study, and the difficulty in rooting the cladograms with data from this gene fragment (Belshaw et al. 1998), the trees have been left unrooted. We present the trees with the Xoridinae as basal, following Quicke et al. (1999). If this is the case then our cladograms also support suggestions that the Labeninae (excluding Brachycyrtinae) are a basal subfamily (Gauld 1983). Members of both of these subfamilies possess plesiomorphic life histories. The placement of the Xoridinae and Labeninae as the sister groups to the rest of the Ichneumonidae lends support to previous a priori reconstructions of groundplan morphological and biological states within the family, namely that the ancestral ichneumonid was an idiobiont ectoparasitoid of xylophagous coleopterous or siricid larvae (Gauld 1988). We currently recognise three major groupings of subfamilies, the Pimpliformes, Ichneumoniformes and Ophioniformes, with the Orthopelmatinae, Labeninae and Xoridinae now each regarded as comprising distinct groups. Our inclusion of the Tryphoninae, Mesochorinae and Metopiinae within the Ophioniformes extends this group considerably but is supported, for example, by their putatively apomorphic, long lateral oviducts (Wahl 1993a). Maintenance of the Tryphoniformes in which Gauld (1991) also placed Eucerotinae and Adelognathinae (largely on the basis of egg form and biology) is unwarranted given that the last two subfamilies appear to be associated with the Ichneumoniformes. Further gene sequencing (both missing taxa and additional gene fragments) is required to test the above hypotheses.
Acknowledgements We are grateful to Ian Gauld and Mark Shaw for supplying specimens of several taxa for gene sequencing and to David Swofford for providing a pre-release version of PAUP*. This work was supported by a Natural Environment Research Council (NERC) grant to DLJQ and MGF and by the NERC Initiative in Taxonomy.
References Belshaw, R., Fitton, M. G., Herniou, E., Gimeno, C. & Quicke, D. L. J. (1998) A phylogenetic reconstruction of the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28S ribosomal RNA. Systematic Entomology 23: 109-123. Dowton, M. (1999) Relationships among the cyclostome braconid (Hymenoptera: Braconidae) subfamilies inferred from a mitochondrial tRNA gene rearrangement. Molecular Phylogenetics & Evolution 11: 283-287.
D. L. J. Quicke, M. G. Fitton, D. G. Notton, G. R. Broad and K. Dolphin 82
Hymenoptera: Evolution, Biodiversity and Biological Control
Dowton, M., Austin, A. D. & Antolin, M. F. (1998) Evolutionary relationships among the Braconidae (Hymenoptera: Ichneumonoidea) inferred from partial 16S rDNA gene sequences. Insect Molecular Biology 7: 129-150. Dowton, M. & Austin, A. D. (1998) Phylogenetic relationships among the microgastroid wasps (Hymenoptera: Braconidae): combined analysis of 16S and 28S rDNA genes and morphological data. Molecular Phylogenetics & Evolution 10: 354-366. Eggleton, P. (1989) The Phylogeny and Evolutionary Biology of the Pimplinae (Hymenoptera: Ichneumonidae). 295 pp. Unpublished PhD thesis, University of London. Farris, J. S., Källersjö, M., Kluge, A. G. & Bult, C. (1994) Testing significance of incongruence. Cladistics 10: 315-319. Finlayson, T. (1967) A classification of the subfamily Pimplinae (Hymenoptera: Ichneumonidae) based on final-instar larval characteristics. The Canadian Entomologist 99: 1-8. Gauld, I. D. (1983) The classification, evolution and distribution of the Labeninae, an ancient southern group of Ichneumonidae (Hymenoptera). Systematic Entomology 8: 167-178. Gauld, I. D. (1985) The phylogeny, classification and evolution of parasitic wasps of the subfamily Ophioninae (Ichneumonidae). Bulletin of the British Museum (Natural History), Entomology 51: 61-185. Gauld, I. D. (1988) Evolutionary patterns of host utilization by ichneumonoid parasitoids (Hymenoptera: Ichneumonidae and Braconidae). Biological Journal of the Linnean Society 35: 351-377. Gauld, I. D. (1991) Ichneumonidae of Costa Rica, 1. Memoirs of the American Entomological Institute 47: 1-589. Gauld, I. D. (1997) Ichneumonidae of Costa Rica, 2. Memoirs of the American Entomological Institute 57: 1-485. Gauld, I. D. & Mound, L. A. (1982) Homoplasy and the delineation of holophyletic genera in some insect groups. Systematic Entomology 7: 73-86. Nixon, K. C. & Carpenter, J. M. (1996) On simultaneous analysis. Cladistics 12: 221-241. Porter, C. C. (1998) Guía de los géneros de Ichneumonidae en ela región Neantárctica del sur de Sudamérica. Opera Lilloana 42: 1-234. Quicke, D. L. J., Lopez-Vaamonde, C. & Belshaw, R. (1999) The basal Ichneumonidae (Insecta: Hymenoptera): 28S D2 rDNA considerations of the Brachycyrtinae, Labeninae, Paxylommatinae and Xoridinae. Zoologica Scripta 28: 203-210. Quicke, D. L. J., Taylor, J. & Purvis, A. (in press) Changing the landscape: a new tree searching strategy. Systematic Biology. Sharkey, M. J. & Wahl, D. B. (1992) Cladistics of the Ichneumonoidea (Hymenoptera). Journal of Hymenoptera Research 1: 15-24. Townes, H. K. (1969) Genera of Ichneumonidae, Part 1. Memoirs of the American Entomological Institute 11: 1-300. Wahl, D. B. (1986) Larval structures of oxytorines and their significance for the higher classification of some Ichneumonidae (Hymenoptera). Systematic Entomology 11: 117-127. Wahl, D. B. (1990) A review of the mature larvae of Diplazontinae, with notes on larvae of Acaenitinae and Orthocentrinae and proposal of two new subfamilies (Insecta: Hymenoptera, Ichneumonidae). Journal of Natural History 24: 27-52. Wahl, D. B. (1991) The status of Rhimphoctona, with special reference to the higher categories within Campopleginae and the relationships of the subfamily (Hymenoptera: Ichneumonidae). Transactions of the American Entomological Society 117: 193-213.
Phylogeny of the Subfamilies of Ichneumonidae (Hymenoptera) 83
Wahl, D. B. (1993a) Cladistics of the genera of Mesochorinae (Hymenoptera: Ichneumonidae). Systematic Entomology 18: 371-387. Wahl, D. B. (1993b) Cladistics of the ichneumonid subfamily Labeninae (Hymenoptera: Ichneumonidae). Entomologia Generalis 18: 91-105. Wahl, D. B. (1997) The cladistics of the genera and subgenera of Xoridinae. pp. 454-460. In I. D. Gauld (Ed.), Ichneumonidae of Costa Rica, 2. Memoirs of the American Entomological Institute 57: 485 pp. Wahl, D. B. & Gauld, I. D. (1998) The cladistics and higher classification of the Pimpliformes (Hymenoptera: Ichneumonidae). Systematic Entomology 23: 299-303.
Molecular Evolution in Social Wasps J. Schmitz1 and R. F. A. Moritz2 1
German Primate Centre, Primate Genetics, Kellnerweg 4, 37077 Göttingen, Germany (email:
[email protected]) 2
Institute for Zoologie-Molecular Ecology, Martin-Luther-Universität Halle/Wittenberg, Kröllwitzer Str. 44, 06099 Halle/Saale, Germany
Introduction The Vespidae (Hymenoptera) include the subfamilies Masarinae, Eumeninae, Stenogastrinae, Polistinae and Vespinae. Eusociality is found in Stenogastrinae, Polistinae and Vespinae. Whereas Carpenter (1988) concluded that the eusocial subfamilies have a common ancestor not shared with the solitary vespids, we found strong evidence that the stenogastrine wasps are not closest to the Vespinae + Polistinae cluster (Schmitz & Moritz 1998). Instead, the Eumeninae are closest to Vespinae + Polistinae, supported by 34 informative nucleotide positions in a molecular data set of 28S rDNA and 16S rDNA sequences (583 nucleotide positions in total). This monophyly was confirmed by 100% bootstrap replicates. Using morphological data, Richards (1971), Spradbery (1975), and van der Vecht (1977) also suggest that the Stenogastrinae are not closest to the remaining social vespids. In the present study we further contribute to this topic with a more extensive data set (29 vespid species) including two Masarinae for the mitochondrial 16S rDNA gene.
Materials and Methods Samples For subfamily relationships we analysed rDNA sequences of eleven Vespinae [Vespa crabro (L.), V. orientalis (L.), Provespa anomala (Saussure), P. nocturna (Vecht), Dolichovespula media (Retzius), D. maculata (L.), D. sylvestris (Scopoli), D. saxonica (F.), Vespula vulgaris (L.), V. germanica (F.), V. rufa (L.)], six Polistinae (Polistes dominulus (Christ), P. saggittarius (F.), P. flavus (Cresson), Belonogaster petiolata (Degeer), B. sp. nov., B. juncea colonialis (Kohl)), six Stenogastrinae (Parischnogaster jacobsoni (Du Buysson), P. mellyi (Saussure), P. alternata (Sakagami), Liostenogaster flavolineata (Cameron), L. sp. nov., Eustenogaster calyptodoma (Sakagami & Yoshikawa)), three Eumeninae (Eumenes sp., Ancistrocerus oviventris (Wesmael), A. nigricornis (Curtis)), and two Masarinae species (Metaparagia maculata (Meade-Waldo), Pseudomasaris maculifrons (Fox)). Furthermore, we analysed two Apis species (Apis mellifera L., A. dorsata F.). Sequences can be retrieved from GeneBank under accession numbers AF066893-AF066939 and AF067145. Corresponding DNA sequences from Cotesia glomerata (L.) (Braconidae), Xylocopa virginica (L.) (Xylocopinae), and Lissonata sp. (Ichneumonidae), were retrieved from GeneBank under accession numbers U06958, L22905, and Z97906 respectively, and used as outgroups. The DNA processing has been performed as described in Schmitz and Moritz (1998). We sequenced at least two representatives of each species and could not detect any intraspecific sequence variability.
84
Hymenoptera: Evolution, Biodiversity and Biological Control
Molecular Evolution in Social Wasps 85
Data Analyses Sequence data were aligned by using the CLUSTAL X program (Higgins & Sharp 1989) and improved by comparison of the secondary structure of the rRNA’s (Schmitz & Moritz 1994). One disadvantage of most methods for phylogenetic reconstructions is the slowness of the algorithms, in particular with high numbers of species. Therefore, we reduced the numbers of sequences by building the 60% consensus sequences representative for each subfamily (Consensus: http://coot.embl.heidelberg.de/Alignment/consensus.html). Characters with ambiguities in the consensus sequences and also gap positions were removed from the data set. Maximum Parsimony (MP) analyses were conducted with the computer program PAUP* 4.0b2 (Swofford 1998) including all characters. Branch-and-Bound searches with collapsing zero-length branches and MULPARS option were used to find the most parsimonious trees. Bootstrap values were determined from 1000 replications. Distances (D) were measured in PAUP* 4.0b2 (Swofford 1998) by using the Kimura 3-parameter model of nucleotide substitutions. Tree reconstruction was performed by the Branch-and-Bound searches. 1000 bootstrap replications were undertaken to find the relative support for nodes. Maximum-likelihood (ML) trees were constructed with PHYLIP 3.572 DNAMLK and DNAML programs (Felsenstein 1993). Alternatively, we used the ML option in PAUP* 4.0b2 and PUZZLE 4.02 (Strimmer & von Haeseler 1996). In general, we used the estimation of transitions/transversions and frequencies from the data set with eight rate categories. Furthermore, the gamma rate heterogeneity parameters were estimated from the data set (PUZZLE 4.02). Values lower than 1 indicate a strong rate heterogeneity across sites (Strimmer & von Haeseler 1996). The molecular clock likelihood ratio tests have been done in PUZZLE 4.02. For tree-length calculation we used MacClade 3.05 (Maddison & Maddison 1992).
Results 28S rDNA The AT content of the 28S rDNAs ranges between 35% and 40%. In relative rate tests we could not detect any association between the nucleotide content and the sequence divergence. In total, 246 characters (51 informative) were used for further analyses after omitting the characters with ambiguous states in the 60% consensus sequences and gap positions of the 28S rDNA. In this data set we could detect a significant phylogenetic signal (g1 = –0.60, p < 0.05) by performing a tree-length skewness test (Hillis & Huelsenbeck 1992). The g1 value increased after removing the outgroup sequence (g1= –1.20, p < 0.01). The mean base compositions for the analysed subfamily consensus sequences were 14% A, 33% C, 33% G, and 20% T. We found no base composition heterogeneity between the various consensus sequences (p = 0.99). For the consensus sequences, the mean transition/transversion ratio was 1.74 ± 0.34, and the rate heterogeneity across sites, calculated in PUZZLE 4.02 was a = 0.86 ± 0.31 (without outgroup: a = 0.51 ± 0.26). 16S rDNA The high AT content of the investigated 16S rDNA sequences (84–87%) is characteristic for Hymenoptera. Corresponding to Schmitz and Moritz (1998) three highly variable regions were excluded (64 nt in total). For 16S rDNA 163 characters (28 informatives) remained after exclusion of characters which failed to fulfil the 60% consensus constraints and removal of gap positions. As in the nuclear data set, we detected a significant phylogenetic signal (g1 = –1.26, p < 0.01; without outgroup: g1 = –0.82, p < 0.01) in the mitochondrial data. The mean base compositions for the analysed mt DNA was 44% A, 7% C, 11% G, and 37% T. We found no base composition heterogeneity between the various mitochondrial consensus sequences (p = 0.99).
J. Schmitz and R. F. A. Moritz 86
Hymenoptera: Evolution, Biodiversity and Biological Control
a)
6 7
Vespinae 13
18
3
Eumeninae
26
27
Stenogastrinae
Apinae
4
b) 5
4 5
Polistinae
100
Vespinae Polistinae
71 4
Eumeninae
2 11
16
16 Figure 1
Masarinae Stenogastrinae
Xylocopa
Phylogenetic reconstruction (MP) of the relationships among subfamilies of Vespidae for a) 28S rDNA and b) 16S rDNA. Numbers above the branches indicate the inferred branchlength. Percentage bootstrap values are shown at the internodes.
The mean transition/transversion ratio was 0.42 ± 0.14 and indicates a high frequency of AT transversions. This is reported also for other wasp taxa (Dowton & Austin 1997). The rate heterogeneity across sites is α = 1.12 ± 0.80.
Phylogenetic reconstructions of 28S rDNA The genetic distances between the subfamily 60% consensus sequences were calculated in PAUP* 4.0b2 and are shown in Table 1. The phylogenetic reconstruction of the 28S rDNA consensus sequences is shown in Figure 1a. In this reconstruction, the consensus sequence of A. mellifera and A. dorsata was used as the outgroup. All three reconstruction methods (MP, ML and D) coincide in the tree topology. The monophyly of Eumeninae + (Polistinae + Vespinae) is well supported by 100% bootstrap proportions independent of the method of reconstruction. The tree-length of the most parsimonious tree shown in Figure 1 was 100 (RC = 0.68). Rearrangements of branches to
Molecular Evolution in Social Wasps 87
Table 1 Mean pairwise distances (PAUP* 4.0) for 28S rDNA consensus sequences (above diagonal) and 16S rDNA consensus sequences (below diagonal). Apinae/ Xylocopinae Stenogastrinae Apinae
0.191
Masarinae
Eumeninae
Polistinae
Vespinae
–
0.195
0.224
0.199
–
0.191
0.232
0.191
Stenogastrinae
0.196
Masarinae
0.184
0.184
Eumeninae
0.141
0.160
0.092
Polistinae
0.160
0.172
0.123
0.092
Vespinae
0.160
0.184
0.123
0.080
–
–
0.093
0.065 0.077
0.049
maintain the monophyly of the social wasps required 12 additional steps. However, choosing a taxon outside of the Aculeata results in a trifurcation of Apinae, Stenogastrinae and Eumeninae + Polistinae + Vespinae. In a previous study (Schmitz & Moritz 1998) we analysed a slightly supported (60% bootstrap replications) monophyly of Apinae + (Eumeninae + (Polistinae + Vespinae)) with the stenogastrine wasps as the sister group. Possible reasons for the unusual position of the bees may be homoplastic characters shared with the outgroup species Nasonia vitripennis (Walker) and the Stenogastrinae. A rate heterogeneity in the data set can cause the misinterpretation of sequence data, well known as long branch attraction. The relative rate test rejected a clock like tree at a significance level of 5%.
Phylogenetic reconstructions of 16S rDNA The phylogenetic tree of the 16S rDNA consensus sequences shown in Figure 1b, was reconstructed using the xylocopine bee X. virginica as outgroup. In this phylogenetic tree Masarinae, Eumeninae, and Polistinae + Vespinae build a monophyletic group, with a somewhat smaller bootstrap probability of about 60–70% compared to the nuclear data set. The tree-length was 67 (RC = 0.44). Four additional steps are necessary to support the monophyly of social wasps. Using solitary wasps as the outgroup will result in very strong support for the Masarinae, Eumeninae, (Polistinae + Vespinae) cluster as shown also in Schmitz and Moritz (1998). This entails, possibly due to homoplasies, that the bees are closer to Eumeninae + (Polistinae + Vespinae) than the Stenogastrinae.
Discussion The pairwise genetic distances for 28S rDNA and 16S rDNA consensus sequences are presented in Table 1. The pairwise distances of Stenogastrinae to any other Vespidae sequences turned out to be the highest. The high sequence divergence is also reflected in the phylogenetic reconstructions (Fig. 1). What are possible reasons for this finding? One could argue that the molecular data sets are randomised and therefore not representative for phylogenetic evaluations. We yielded significantly skewed MP-tree distributions (p < 0.01) indicating strong phylogenetic signals in both the 28S and 16S rDNA data. The test of rate heterogeneity in the 28S rDNA (α = 0.86 ± 0.31) and the 16S rDNA (α = 1.22±0.80) indicated some degree of substitution rate differences across sites in the nuclear data. This is not an unusual finding when rDNAs of distantly related taxa are compared (Friedrich & Tautz 1997). More critical is rate heterogeneity among lineages as shown by Carmean and Crespi (1995) and Hwang et al. (1998). Rate heterogeneity among lineages can cause the socalled long branch attraction. By including the Masarinae in the mitochondrial data we could
J. Schmitz and R. F. A. Moritz 88
Hymenoptera: Evolution, Biodiversity and Biological Control
show, that the phylogenetic reconstructions are not effected by attraction of Eumeninae and Polistinae + Vespinae DNA sequences. Compared to Stenogastrinae, the Masarinae also show a closer relationship to Polistinae + Vespinae. The relative rate test indicated a significant departure of the clock-like distribution of branch length for the nuclear data. However, ML reconstructions are most robust when rate heterogeneity among lineages increases (Hwang et al. 1998). In all ML trees, the Stenogastrinae remain outside of Eumeninae + (Polistinae + Vespinae). This holds also when the data are split into the two functional different rRNA structural elements, loops and double stranded regions, and reanalysed (Schmitz & Moritz 1998). Further investigations with additional sequences of conservative genes are called for to verify the position of Stenogastrinae outside of the remaining social Vespidae.
Acknowledgements We thank James M. Carpenter for providing samples of Metaparagia maculata and Pseudomasaris maculifrons. This study was supported by the Deutsche Forschung-gemeinschaft (Mo 373/41 and Mo 373/4-2).
References Carmean, C. & Crespi, J. (1995) Do long branches attract flies? Nature 373: 666. Carpenter, J. M. (1988) The phylogenetic system of the Stenogastrinae (Hymenoptera: Vespidae). Journal of the New York Entomological Society 96: 140-175. Dowton, M. & Austin, A. D. (1997) Evidence for AT-transversion bias in wasp (Hymenoptera: Symphyta) mitochondrial genes and its implications for the origin of parasitism. Journal of Molecular Evolution 44: 398-405. Felsenstein, J. (1993) PHYLIP Manual Version 3.5c. University of Washington, Seattle. Friedrich, M. & Tautz, D. (1997) An episodic change of rDNA nucleotide substitution rate has occurred at the time of the emergence of the insect order Diptera. Molecular Biology & Evolution 14: 644-653. Higgins, D. G. & Sharp, P. M. (1989) Fast and sensitive multiple sequence alignments on a microcomputer. CABIOS 5: 151-153. Hillis, D. M. & Huelsenbeck, J. P. (1992) Signal, noise, and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189-195. Hwang, U. W., Kim, W., Tautz, D. & Friedrich, M. (1998) Molecular phylogenetics at the Felsenstein Zone: Approaching the Strepsiptera problem using 5.8S and 28S rDNA sequences. Molecular Phylogenetics & Evolution 9: 470-480. Maddison, W. P. & Maddison, D. R. (1992) MacClade (v. 3.06): Analysis of Phylogeny and Character Evolution. Sinauer, Sunderland. Richards, O. W. (1971) The biology of the social wasps (Hymenoptera, Vespidae). Biological Reviews 46: 483-528. Schmitz, J. & Moritz, R. F. A. (1994) Sequence analysis of the D1 and D2 regions of 28S rDNA in the hornet (Vespa crabro) (Hymenoptera, Vespinae). Insect Molecular Biology 3: 273-277. Schmitz, J. & Moritz, R. F. A. (1998) Molecular phylogeny of Vespidae (Hymenoptera) and the evolution of sociality in wasps. Molecular Phylogenetics & Evolution 9: 183-191 Spradbery, J. P. (1975) The biology of Stenogaster concina van der Vecht, with comments on the phylogeny of the Stenogastrinae (Hymenoptera: Vespidae). Journal of the Australian Entomological Society 14: 309-318.
Molecular Evolution in Social Wasps 89
Strimmer, K. & von Haeseler, A. (1996) Quartet puzzling: a quartet maximum likelihood method for reconstructing tree topologies. Molecular Biology & Evolution 13: 964-969. Swofford, D. L. (1998) PAUP 4*. Phylogenetic Analysis Using Parsimony *and Other Methods. Version 4. Sinauer Associates, Sunderland, Massachusetts. Tajima, F. & Nei, M. (1984) Estimation of evolutionary distance between nucleotide sequences. Molecular Biology & Evolution 1: 269-285. van der Vecht, J. (1977) Studies of oriental Stenogastrinae (Hymenoptera Vespoidea). Tijdschrift voor Entomologie 120: 55-75.
Rearrangement of the Hymenopteran Mitochondrial Genome is Accelerated Relative to Orthopteroid Insects Mark Dowton1,2, Andrew D. Austin2 and Paul K. Flook3 1
Australian Flora and Fauna Research Centre, Department of Biology, Wollongong University, Wollongong, N.S.W. 2522 Australia (email:
[email protected]) 2
Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia 3
Zoologisches Institut, Rheinsprung 9, Basel 4051 Switzerland
Introduction Examination of a taxonomically diverse sample of mitochondrial genomes within the Metazoa suggests that genes change position rarely (e.g. Jakobs et al. 1989; Boore et al. 1995; Macey et al. 1997). This observation has led to claims that variation in mitochondrial gene organisation represent highly reliable indicators of common ancestry (e.g. Boore et al. 1995, 1998). As such, these characters could be valuable synapomorphies for a few natural groups, just as holometabolism is a useful character for defining a single node in the insect phylogeny. Such characters clearly warrant further investigation. Although sampling of mitochondrial gene orders has covered a taxonomically diverse range of metazoans, the number of taxa sampled has been restricted by the high cost of generating complete mitochondrial sequences. To circumvent this, we recently sampled one mitochondrial region spanning just four genes to assess how frequently rearrangements occur in a broad range of Hymenoptera (Dowton & Austin 1999). This region was particularly prone to rearrangement, having experienced at least five evolutionarily independent rearrangements during the last 180 Myr. Nevertheless, closely related taxa shared novel genome organisations, implying that these arrangements may be useful phylogenetic characters within the Hymenoptera (e.g. Dowton 1999). These data suggest that either gene rearrangements occur much more frequently than previously considered, or that the rate of gene rearrangement is accelerated in the Hymenoptera relative to other insect groups. To distinguish between these two possibilities, we are examining gene organisation in this region in a range of non-hymenopteran insects. In the present study, we focus on gene organisation in the orthopteroid insect orders, as previous evidence indicated that at least one rearrangement has occurred in this gene region during the evolution of these orders (Flook et al. 1995a, 1995b).
Materials and Methods Sequences spanning the junction between the 3’-end of the cytochrome oxidase II (COII) gene and the start codon of the ATPase 8 (A8) gene were generated as described (Dowton & Austin 1999). No additional hymenopteran taxa were sequenced for the present study, but a range of non-hymenopteran taxa were. These were from the insect orders Mantodea, Grylloblattodea,
90
Hymenoptera: Evolution, Biodiversity and Biological Control
Rearrangement of the Hymenopteran Mitochondrial Genome 91
Phasmotodea, Blattodea, Orthoptera, and Plecoptera; specific taxa are listed in Table 1. tRNA genes were identified using tRNAscan-SE (version 1.1, http://genome.wustl.edu/eddy/tRN Ascan-SE: Lowe & Eddy 1997), as described in Dowton and Austin (1999). Abbreviations used in the text and figures are: A8, ATPase 8; COII, cytochrome oxidase II; tRNAD, tRNA-aspartate; tRNAH, tRNA-histidine; tRNAK, tRNA-lysine.
Results We sequenced the junction spanning the COII and A8 genes in 13 non-hymenopteran taxa, representing most of the orthopteroid insect orders. No gene rearrangements were evident in any of the orthopteroids examined (Table 1); i.e. in all cases, the gene order is COII-tRNAK-tRNAD-A8 (Fig. 1). The only orthopteran that does not have this arrangement is the previously sequenced Locusta migratoria (L.) (Flook et al. 1995a), in which the order of the tRNA genes is reversed. Further, the general structure of this region is highly conserved in the orthopteroids, and contrasts sharply with the plasticity seen in the Hymenoptera (Fig. 1). Generally, the two taxonomic groups (Hymenoptera and orthopteroids) have similarly sized COII carboxy termini – between 17 and 24 amino acids from the primer-binding site to the stop codon. All orthopteroids examined have an incomplete stop codon for the cytochrome oxidase II gene (‘T’ in all cases) which is presumably completed (to TAA) after polyadenylation (Ojala et al. 1981). The Hymenoptera exhibit more variation, with some having complete stop codons (TAA, or rarely TAG), others incomplete. The first tRNA gene lies directly next to the stop codon in all orthopteroids sequenced, whereas this intergenic region can have as many as 138 non-coding nucleotides in hymenopteran taxa. The identity of the tRNA gene immediately downstream of the COII gene was invariably tRNAK in the orthopteroids (but not in Locusta, where it is tRNAD; Flook et al. 1995a), but can be tRNAK or tRNAD in the Hymenoptera. Further, this tRNA gene was always encoded on the mitochondrial J-strand in the orthopteroids (including Locusta), whereas this first tRNA gene is sometimes encoded on the mitochondrial N-strand in the Hymenoptera (strand nomenclature after Simon et al. 1994). The anticodon for the tRNAK gene was ‘CTT’ in all orthopteroids examined (Table 1), whereas the Hymenoptera have either ‘CTT’ or ‘TTT’, with the derived ‘TTT’ condition having multiple origins (Dowton & Austin 1999). The intergenic region between the two tRNA genes was less variable when these two groups were considered. Orthopteroids have at most three non-coding nucleotides in this region, whilst the Hymenoptera have at most eight. However, overlaps are more severe in the Hymenoptera. In the orthopteroids, overlaps are limited to at most a single nucleotide (see Table 1), whereas some Hymenoptera have as many as three nucleotide overlaps. In the latter case, production of both tRNA molecules from a single mitochondrial polycistronic transcript can only occur after polyadenylation (for a discussion see Dowton & Austin 1999). The second tRNA gene in this region is always tRNAD in the orthopteroids examined (but not in Locusta, where it is tRNAK; Flook et al. 1995a), whereas it could by tRNAD, tRNAK or tRNAH in the Hymenoptera. Further, in some Hymenoptera there is no second tRNA gene here, while in others there is a third (Dowton & Austin 1999). Interestingly, in both insect groups the second tRNA gene (and third, where it occurs) is always encoded on the mitochondrial J-strand, regardless of whether a rearrangement has occurred. Finally, the intergenic region between the A8 and neighbouring upstream tRNA gene contains no non-coding nucleotides in the orthopteroids examined, although Locusta is a notable exception (see Table 1). Locusta is the only taxon in this sample of orthopteroids that has rearranged. By comparison, Hymenoptera can have as many as
Mark Dowton, Andrew D. Austin and Paul K. Flook 92
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Genetic characteristics of the junction between the cytochrome oxidase II and ATPase 8 genes for various orthopteroid taxa. Taxa with rearranged genes are boxed; ‘aas; stop codon’ indicates the number of amino acids in the cytochrome oxidase II gene between the primer binding site and the carboxy terminus; (T) indicates that a complete stop codon can only be produced after polyadenylation of the transcription product. ‘A nts’ is the number of nucleotides between the cytochrome oxidase II gene and the first tRNA gene, ‘B nts’ is the number of nucleotides between the first and second tRNA genes, and ‘C nts’ is the number of nucleotides between the second tRNA gene and the ATPase 8 gene, as indicated in the scheme above the column titles. Negative numbers indicate overlaps, with the overlap shown in brackets; ‘ND’ indicates not determined, due to incomplete sequence data.
COI I
A
t RNA1
B
t RNA2
C
ATP ase 8
COII (aas; stop codon)
A nts
B nts
C nts
Taxon
Choeradodis rhombicollis Wood-Mason
20 (T)
0
–1 (A)
0
KD (CTT)
Grylloblatta rothi Gurney
21 (t)
0
+2
0
KD (CTT)
Pseudophasmatidae
Agatheromera crassa (Blanchard)
17 (T)
0
–1 (A)
0
KD (CTT)
Phasmatidae
Megacrania apheus (Westwood)
ND
0
–1 (A)
0
KD (CTT)
Gromphadorhina portentosa (Schaum)
20 (T)
0
0
0
KD (CTT)
Gryllotalpa gryllotalpa (L.)
18 (T)
0
–1 (A)
0
KD (CTT)
Gryllus campestris (L.)
18 (T)
0
0
0
KD (CTT)
Acheta domesticus (L.)
19 (T)
0
0
0
KD (CTT)
Eneopterus sp.
19 (T)
0
+3
0
KD (CTT)
Taxonomic affiliation
Arrangement (tRNAK anticodon)
Mantodea
Grylloblattodea Grylloblattidae Phasmotodea
Blattodea Panchloridae Orthoptera Ensifera Gryllidae
Rearrangement of the Hymenopteran Mitochondrial Genome 93
Stenopelmatidae
Penalva sp.
21 (T)
0
–1 (A)
0
KD (CTT)
Anastostamatidae
Hemideina crassidens (Blanchard)
22 (T)
0
–1 (A)
0
KD (CTT)
Cylindraustralia kochii (Saussure)
18 (T)
0
–1 (A)
0
KD (CTT)
Locusta migratoria (L.)
19 (T)
0
+3
+17
DK (CTT)
Protonemura meyeri (Pictet)
21 (T)
0
–1 (A)
0
KD (CTT)
Caelifera Cylindrachetidae Acrididae Plecoptera Nemouridae
Mark Dowton, Andrew D. Austin and Paul K. Flook 94
Hymenoptera: Evolution, Biodiversity and Biological Control
A. Orthopteroids
COII
T
17-22 aas
stop codon
intergenic region
COII
T,TA TAA
-4, tRNA-K or D +138
17-24 aas
0
tRNA- K or D
stop intergenic region codon
-1, +3
tRNA-D or K
intergenic region
-3, +8
tRNA-D, K or H
intergenic region
0, +17
A8
intergenic region
-9, +99
A8
intergenic region
B. Hymenoptera Structural organisation of the COII-A8 mitochondrial junction in A) orthopteroids, and B) Hymenoptera. Arrows indicate direction of translation; thus rightward arrows indicate the gene is encoded on the J-strand, leftward arrows on the N-strand. Numbers in ‘intergenic regions’ refer to minimum and maximum length of non-coding nucleotides, respectively, with negative numbers indicating that the genes overlap.
Figure 1
99 non-coding nucleotides in this region, but only in taxa having a rearrangement. Taxa that retain the plesiomorphic arrangement (i.e. KD) have at most a single nucleotide here (Dowton & Austin 1999). Thus there are generally intergenic nucleotides at gene boundaries in rearranged genomes, but not in unrearranged ones. Others have suggested that the appearance of intergenic nucleotides after rearrangement is consistent with postulated mechanisms of genome reorganisation (Kumazawa & Nishida 1995; Macey et al. 1998). Following duplication of the region, one of the duplicate genes is randomly rendered non-functional by mutation. The non-functional gene then evolves without selective constraints as an intergenic region. However, it is not clear why these intergenic nucleotides should (a) be maintained over long periods of evolution, given the general drive towards minimisation of the mitochondrial genome size, and (b) accumulate at protein/tRNA boundaries, but not tRNA/tRNA boundaries. In both rearranged and non-rearranged genomes, there are at most 12, but rarely more than 10 non-coding nucleotides at the tRNA/tRNA boundary (Dowton & Austin 1999, and present study).
Discussion Previous work indicated at least one mitochondrial gene rearrangement occurred at the COII-A8 gene junction during the evolution of the Orthoptera; two representatives from the suborder Ensifera had unrearranged genomes, while three members of the suborder Caelifera had rearranged genomes (Flook et al. 1995b). The present study similarly suggested that this orthopteran rearrangement occurred some time during the evolution of the Caelifera, as we found that one basally derived caeliferan retains the plesiomorphic arrangement [Cylindraustralia kochii (Saussure)]. Present evidence suggests that this rearrangement character will be a useful synapomorphy for superfamily level relationships in the Orthoptera: Caelifera, with representatives from the Acridoidea and Pamphagoidea apparently sharing the rearrangement (Flook et al. 1995b). Consistent with this observation, recent molecular analyses of orthopteran relationships (Flook & Rowell 1998) recover the Acridoidea + Pneumoroidea + Trigonopterygoidea + Pamphagoidea as
Rearrangement of the Hymenopteran Mitochondrial Genome 95
a natural group. Future studies will focus on the arrangement of these genes in representatives from the Tetrigoidea and Eumasticoidea, which are thought to fall outside of this clade (Flook & Rowell 1998). We previously observed that the junction of the COII and A8 genes was prone to rearrangement in the Hymenoptera (Dowton & Austin 1999), contrary to expectations that mitochondrial genome organisation is highly conserved. We had some evidence that this accelerated rate of rearrangement was restricted to the hymenopteran suborder Apocrita, as three representatives from distinct symphytan lineages were not rearranged. The present study supports the contention that mitochondrial gene rearrangements are rare evolutionary events, with the Hymenoptera representing a lineage which uncharacteristically rearranges more frequently. These observations strengthen the case that the Hymenoptera represent an ideal model system for discovering the underlying mechanisms of gene rearrangement in the mitochondrial genome.
Acknowledgements This work was supported by grants from the Australian Research Council, the Linnean Society of NSW, and the Mark Mitchell Foundation.
References Boore, J. L., Collins, T. M., Stanton, D., Daehler, L. L. & Brown, W. M. (1995) Deducing the pattern of arthropod phylogeny from mitochondrial DNA rearrangements. Nature 376: 163-165. Boore, J. L., Lavrov, D. V. & Brown, W. M. (1998) Gene translocation links insects and crustaceans. Nature 392: 667-668. Dowton, M. (1999) Relationships among the cyclostome braconid (Hymenoptera: Braconidae) subfamilies inferred from a mitochondrial tRNA gene rearrangement. Molecular Phylogenetics & Evolution 11: 283-287. Dowton, M. & Austin, A. D. (1999) Evolutionary dynamics of a mitochondrial rearrangement “hotspot” in the Hymenoptera. Molecular Biology & Evolution 16: 298-309. Flook, P. K. & Rowell, C. H. F. (1998) Inferences about orthopteroid phylogeny and molecular evolution from small subunit nuclear ribosomal DNA sequences. Insect Molecular Biology 7: 163-178. Flook, P., Rowell, H. & Gellissen, G. (1995a) Homoplastic rearrangements of insect mitochondrial tRNA genes. Naturwissenschaften 83: 336-337. Flook, P. K., Rowell, C. H. F. & Gellissen, G. (1995b) The sequence, organisation, and evolution of the Locusta migratoria mitochondrial genome. Journal of Molecular Evolution 41: 928-941. Jakobs, H. T., Asakawa, S., Araki, T., Miura, K.-I., Smith, M. J. & Watanabe, K. (1989) Conserved tRNA gene cluster in starfish mitochondrial DNA. Current Genetics 15: 193-206. Kumazawa, Y. & Nishida, M. (1995) Variations in mitochondrial tRNA gene organization of reptiles as phylogenetic markers. Molecular Biology & Evolution 12: 759-772. Lowe, T. M. & Eddy, S. R. (1997) tRNAscan-SE: A program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Research 25: 955-964. Macey, J. R., Larson, A., Ananjeva, N. B., Fang, Z. & Papenfuss, T. J. (1997) Two novel gene orders and the role of light-strand replication in rearrangement of the vertebrate mitochondrial genome. Molecular Biology & Evolution 14: 91-104. Macey, J. R., Schulte, J. A., II, Larson, A. & Papenfuss, T. J. (1998) Tandem duplication via lightstrand synthesis may provide a precursor for mitochondrial genomic rearrangement. Molecular Biology & Evolution 15: 71-75.
Mark Dowton, Andrew D. Austin and Paul K. Flook 96
Hymenoptera: Evolution, Biodiversity and Biological Control
Ojala, D., Montoya, J. & Attardi, G. (1981) tRNA punctuation model of RNA processing in human mitochondria. Nature 290: 470-474. Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook, P. (1994) Evolution, weighting, and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved polymerase chain reaction primers. Annals of the Entomological Society of America 87: 1-51.
Phylogeny of Microgastroid Braconid Wasps, and What It Tells Us About Polydnavirus Evolution James B. Whitfield Department of Entomology, 321 Agriculture Building, University of Arkansas, Fayetteville, AR 72701 USA (email:
[email protected])
Introduction One of the most remarkable interactions known within the Animal Kingdom is that between parasitoid wasps and hereditary viruses known as polydnaviruses (Whitfield 1990, 1994, 1997; Fleming 1992). It is likely to be one of the only known mutualisms between viruses and eukaryotes (Whitfield 1990; Fleming 1992). The parasitoid wasps live as larvae within the bodies of other larval insects and depend on these viruses to both suppress the immune system and alter the physiology of the host to allow their own survival (Edson et al. 1981; Beckage 1997, 1998). The polydnaviruses (PDV’s), in turn, are fully integrated into the chromosomal DNA of their wasp carriers, and are thus vertically transmitted (with 100% efficiency) in Mendelian fashion (Fleming & Summers 1986; Stoltz 1990). They are not known to exist independently of the wasps (Stoltz & Whitfield 1992), and indeed no longer possess a recognisable complete viral life cycle (Stoltz 1993). The PDV can be considered analogous to a ‘venom’ system used by the wasp against its host organisms (typically caterpillars). There is evidence, in fact, that some venom genes of wasp origin are now packaged within PDV genomes for export into host caterpillars (Webb & Summers 1990). Much of the PDV gene expression is thus targeted toward a different host (the caterpillar) than that within which the virus replicates (the wasp). Two groups of PDV’s are known and are currently treated unofficially as ‘genera’ of the Polydnaviridae (Stoltz & Whitfield 1992; Stoltz 1993). The first (the ichnovirus group) has been found only within the ichneumonid subfamily Campopleginae and a few putatively related genera. The second group comprises the bracoviruses, which have been found within only a few subfamilies of Braconidae (Stoltz & Whitfield 1992; Whitfield 1997). Both PDV groups are found only in relatively derived lineages within the two separate wasp families. While the two viral groups appear to share some characteristics of genome organisation and functional significance, they differ in gross structure and packaging, and in their sub-cellular sites of replication in the wasps (Stoltz 1993; Beckage 1998). It is therefore likely that they originated twice independently among the wasps. This paper focuses upon the evolution of the bracoviruses. Currently, this group is more promising than the ichnoviruses for evolutionary study because the phylogenetic relationships among the braconid carriers are better understood. Also, several functional PDV genes are currently under comparative study (see below). Recently, braconids carrying bracoviruses were determined to comprise a monophyletic lineage within the Braconidae (Fig. 1 in Whitfield 1997). This monophyly has subsequently been corroborated by independent analyses of 16S data (Dowton et al. 1998) and of 28S sequences (Belshaw et al. 1998; Dowton & Austin 1998). Based
97
James B. Whitfield 98
Hymenoptera: Evolution, Biodiversity and Biological Control
on these inferred relationships and known fossils, the common ancestor of this lineage is likely to have lived over 60 million years ago (Whitfield 1997). Thus, if the bracoviruses and wasps have been associated throughout their history in the current manner, the PDV/wasp relationship must be at least this old. The availability of the braconid phylogeny and several known polydnaviral genes allows me to explore and test hypotheses concerning the evolution of the bracovirus/wasp relationship. I hypothesise that the phylogenies of the wasps and viruses will mirror one another. That is, we expect that the phylogeny of the viruses, determined by analysis of sequence data from genes of viral origin, will be co-phylogenetic with the wasps. Several lines of indirect evidence (see below) support this hypothesis, but the expectation of co-phylogeny has never been tested directly with decisive evidence.
Indirect evidence for co-phylogeny between braconid wasps and PDV’s 1) As mentioned briefly above, it has been demonstrated that PDV’s are integrated into wasp chromosomal DNA as proviruses (Fleming & Summers 1986; Xu & Stoltz 1991),which are inherited in a Mendelian fashion (Stoltz 1990). There is no evidence that PDV’s can be transferred horizontally between parasitoids, and the PDV’s do not appear to replicate once the wasp inserts them into its host caterpillar (Stoltz 1993). 2) Braconid wasps that are known to carry PDV’s are themselves closely related and, based on current evidence, form a monophyletic lineage (Whitfield 1997). 3) All investigated members of this lineage are known to carry PDV’s (Stoltz & Whitfield 1992) including the previously unstudied Miracinae (R. Wharton pers. comm.). 4) Results from preliminary studies of relatedness among PDV’s from different wasp species using Southern blotting and comparative serology indicate that more closely related wasps carry more genetically similar PDV’s (summarised in Stoltz & Whitfield 1992). Thus, I have hypothesised that PDV’s from related wasps will be genetically related in a manner that reflects the phylogeny of their wasp carriers. Few PDV studies have directly analysed interspecific genetic variation and evolution of specific genes or viral sequences. This is primarily because so few genes of PDV origin have been characterised other than the family of omegaconotoxin-like sequences found in ichneumonids (see Dib-Hajj et al. 1993). However, within the last four years this situation has changed significantly. Our laboratory is currently investigating comparative DNA sequences of two PDV genes from a set of 20 wasp species in the genus Cotesia, as well as several outgroup genera. These genes are the early protein 1 (EP1) gene indentified and characterised originally in Cotesia congregata (Say) (Harwood & Beckage 1994; Savary et al. 1997) and the glycosylating secreted protein (CrV1) gene originally characterised from C. rubecula (Marshall) (Asgari et al. 1996, 1997). The products of both genes have been implicated in immune suppression of the host, and both genes appear to originate from the virus genome rather than that of the wasp. The goal of this study is to compare the phylogeny estimated from the two PDV genes to that estimated from mtDNA 16S and NADH1 genes, to directly test the question of co-phylogeny between the wasps and the PDV’s they carry. Below I present preliminary results from comparisons based on one of the genes of viral origin (CrV1) and the two wasp mtDNA genes, for the taxa for which our lab has so far been able to sequence both PDV and wasp genes. The ultimate goal is to compare results from all four genes for all 20+ taxa. At the end of this report I discuss some technical reasons why it has been difficult to rapidly achieve this goal.
Phylogeny of Microgastroid Braconid Wasps 99
Table 1 PCR and sequencing primers used in this study (* = developed by P. T. Smith from C. rubecula PDV sequences in Asgari et al. (1996); they are based on positions 40-59, 1143-1162, 562-581, 562-581 (reverse complement), 951-970 and 194-213, respectively, of the C. rubecula CrV1 sequence). Primer name
Sequence
Source
16S 16S B
5’-CACCTGTTTATCAAAAACAT-3’
Dowton and Austin (1994)
16S outer
5’-CTTATTCAACATCGAGGTC-3’
Whitfield (1997)
ND1-F
5’-GATAAATCAAA/TGG/T –3’
Smith et al. (1999)
ND1-R
5’-CAACCTTTTAGTGATGC-3’
Smith et al. (1999)
CrV1-F
5’-CTCCTGAGTCAATCATGTAC-3’
This study*
CrV1-R
5’-GCTAGAACATTTAGATTGCA-3’
This study*
NADH1
CrV1
CrV1-int
5’-CGTGAAGATTTGCTTTCTGA-3’
This study*
CrV1-irc
5’-TCAGAAAGCAAATCTTCACG-3’
This study *
CrV1R-II
5’-CATACTTTCATAAGTAGACT-3’
This study *
CrV1F-III
5’-CAAACGACTTCGATGAATCT-3’
This study *
Materials and Methods Either living, freshly frozen or ethanol-preserved wasps of 20 species of Cotesia and a variety of potential outgroup species of Microgastrinae were obtained from laboratory cultures, field collecting or other collaborators (many of these sources are reported in more detail in the generic study by Mardulyn & Whitfield 1999). Contributors of material to this study are identified in the Acknowledgements. DNA from whole wasps (with appendages and wings removed as voucher material) was extracted using a standard phenol/chlorophorm extraction protocol. The wasp and PDV genes were amplified using standard PCR protocols (Palumbi 1996) and the primers listed in Table 1. PCR products were purified using Promega Wizard PCR Preps, and 1/2 volume sequencing reactions were run using the ABI dRhodamine Terminator Cycle Sequencing Ready reaction kit. Sequencing reaction products were sent to the University of Florida Core Sequencing Facility to be run on an ABI 377 automated sequencer. Resulting sequences were checked, aligned using Clustal W (Thompson et al. 1994) and entered into NEXUS format using MacClade 3.07 (Maddison & Maddison 1992). All analyses reported here were run using the maximum parsimony criterion (PAUP* – Swofford 1998), and employed branch-and-bound searches to ensure finding the shortest trees. Tree comparison statistics (especially the partition metric of Penny and Hendy 1985) were calculated using Component version 2.0 (Page 1993). Corresponding branch lengths from the resulting trees were entered into Kaleidagraph 3.0.5 (Abelbeck Software 1994) for calculation of correlations.
Results Sequences from the 16S (373 aligned bp) and NADH1 (496 aligned bp) genes were obtained from 18 species of Cotesia and two outgroup species (Glyptapanteles porthetriae (Muesebeck) and Microplitis demolitor Wilkinson). Results from only six of the ingroup species are reported
James B. Whitfield 100
Hymenoptera: Evolution, Biodiversity and Biological Control
Fig. 1
Comparison of the trees resulting form maximum parsimony analysis of sequence data from wasp and PDV DNA sequences (see text for details of the sequences). Phylogenetic analyses were run using PAUP*4.0 test version b1 (Swofford 1998), and tree comparison statistics were obtained using COMPONENT 2.0 (Page 1993). Wasp DNA and PDV DNA trees are identical except for one unresolved node in the PDV tree.
here, since it is for these six ingroup species that CrV1 sequences (approximately 800 bp, 349 of which have been unambiguously aligned) have also been obtained for comparison. Sequences from these six species are being deposited in GenBank, along with the mtDNA data for the outgroup species. Parsimony analysis of the 16S (100 informative sites) and NADH1 (55 informative sites) genes arrived at the same topology when only the six ingroup Cotesia and two outgroup species were analysed (Figs 1, 2) and, not surprisingly, this same tree resulted from combined analysis of the two genes (Fig. 1). The tree resulting from parsimony analysis of the CrV1 data (96 informative sites) was topologically identical to that from the wasp genes, except that the position of C. marginiventris (Cresson) relative to the other taxa was unresolved. The chance that the wasp and PDV trees would be this similar (as measured by the partition metric) by chance alone is less than .05, based on comparisons of sets of 10 000 random trees. Since the trees from the wasp and PDV gene data were topologically identical except for one unresolved node, it was possible to directly compare the corresponding branch lengths from the wasp and PDV trees, to determine if lengths tended to be correlated (i. e. whether the trees correspond in terms of inferred relative amounts of change on the branches, in addition to the topological similarity). The lengths from the 16S + NADH1 tree and those from the CrV1 tree are indeed significantly correlated (p < .05, see Fig. 3 for details). Considering the relatively small amount of data so far applied to this testing of co-phylogeny, the degree of correspondence between wasp and PDV trees is remarkable.
Discussion So far the data obtained from the wasp genes and genes of PDV origin indicate a strong degree of parallel pattern between the wasp phylogeny, as estimated from mtDNA, and the PDV
Phylogeny of Microgastroid Braconid Wasps 101
C. rubecula C. congregata C. marginiventris C. glomerata C. melitaearum C. orobenae Fig. 2
Phylograms (showing branch lengths obtained from parsimony analysis) of the three analyzed genes. Note similarity in relative lengths of corresponding branches between genes.
James B. Whitfield 102
Hymenoptera: Evolution, Biodiversity and Biological Control
Fig. 3
Plots of branch lengths for corresponding branches between genes (branch lengths from Fig. 2). All are significantly correlated, and note especially the significant correlation between branch lengths from the two combined wasp genes and those from the PDV gene.
phylogeny, as estimated from CrV1. A more convincing demonstration of co-phylogeny will come from; 1) analysis of more taxa, including not only the other Cotesia species for which we have wasp sequence data, but also other outgroup genera; 2) the addition of data from the EP1 gene – so far we have alignable sequence from only three Cotesia species; and 3) data from structural virus genes, which can be demonstrably of only viral origin. Why has it been so difficult to obtain these additional data? Firstly, at the beginning of this study, the viral genes had not been sequenced from more than one species, so the design of conserved PCR primers was problematical. The redesign of these primers is still ongoing, since even the current ones will not amplify all of the taxa. Secondly, some current work in other laboratories suggests that some of the functional viral genes, including EP1 and CrV1, may be non-functional in some species (S. Asgari, M. R. Strand, pers. comm.). This may explain why some of the sequences we have already obtained but not analysed appear so divergent, and may also help to explain the difficulty in amplifying some taxa. In addition, the non-functional genes could pose some problems for phylogenetic analysis in future, in that some taxa would have a completely different pattern and rate of change in a given gene from others – this is a classic case where phylogenetic analysis using most current methods can fail (Steel et al. 1993; Lockhart et al. 1994). In summary, the current data suggests that wasp and PDV gene phylogenies will match perfectly, at least within Cotesia. Combined with the establishment of monophyly of the PDV-bearing
Phylogeny of Microgastroid Braconid Wasps 103
group of braconid wasps, the evidence certainly suggests a long-term co-phylogeny between the wasps and PDV’s. It is still too early to say this with complete confidence, however, and more work lies ahead to understand the comparative sequence data from the PDV genes.
Acknowledgements I would especially like to thank Otto Schmidt and Nancy Beckage for information concerning the PDV genes identified and characterised in their laboratories, to Don Stoltz for stimulating some of the ideas developed here, and to Paul Smith for beginning the sequencing studies of EP1 and CrV1, as well as of the Cotesia mtDNA genes. Jasa Holt provided technical assistance with PCR and sequencing, while Sassan Asgari and Mike Strand provided some useful unpublished information about PDV genes. Patrick Mardulyn provided some microgastrine 16S sequences from another project in our laboratory. Nancy Beckage, Tim Herman, Judy Pell, Don Stoltz, Mike Strand and John Ruberson provided specimens of various Cotesia species to us. This work was funded by NSF grants BSR 9111938 and INT-9605091, USDA grant 95- 01893 and grant 94-B04 from the Arkansas Science and Technology Authority.
References Abelbeck Software (1994) Kaleidagraph, version 3.0.5. Data analysis/graphics application for the Macintosh. Synergy Software, Reading, PA. Asgari, S., Hellers, M. & Schmidt, O. (1996) Host haemocyte inactivation by an insect parasitoid: transient expression of a polydnavirus gene. Journal of General Virology 77: 2653-2662. Asgari, S., Schmidt, O. & Theopold, U. (1997) A polydnavirus-encoded protein of an endoparasitoid wasp is an immune suppressor. Journal of General Virology 78: 3061-3070. Beckage, N. E. (1997) The parasitic wasp’s secret weapon. Scientific American Nov. 1997: 82-87. Beckage, N. E. (1998) Parasitoids and polydnaviruses. Bioscience 48: 305-311. Belshaw, R., Fitton, M., Herniou, E., Gimeno, C. & Quicke, D. L. J. (1998) A phylogenetic reconstruction of the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28S ribosomal RNA. Systematic Entomology 23: 109-123. Dib-Hajj, S. D., Webb, B. A. & Summers, M. D. (1993) Structure and evolutionary implications of a “cysteine-rich” Campoletis sonorensis polydnavirus gene family. Proceedings of the National Academy of Sciences, USA 90: 3765-3769. Dowton, M., Austin, A. D. & Antolin, M. F. (1998) Evolutionary relationships among the Braconidae (Hymenoptera: Ichneumoidea) inferred from partial 16S rDNA gene sequences. Insect Molecular Biology 7: 129-150. Dowton, M. & Austin, A. D. (1998) Phylogenetic relationships among the microgastroid wasps (Hymenoptera: Braconidae): combined analysis of 16S and 28S rDNA genes, and morphological data. Molecular Phylogenetics & Evolution 10: 354-366. Edson, K. M., Vinson, S. B., Stoltz, D. B. & Summers, M. D. (1981) Virus in a parasitoid wasp: suppression of the cellular immune response in the parasitoid’s host. Science 211: 582-583. Fleming, J. G. W. (1992) Polydnaviruses: mutualists and pathogens. Annual Review of Entomology 37: 401-426. Fleming, J. G. W. & Summers, M. D. (1986) Campoletis sonorensis endoparasitic wasps contain forms of C. sonorensis virus DNA suggestive of integrated and extrachromosomal polydnavirus DNAs. Journal of Virology 57: 552-562.
James B. Whitfield 104
Hymenoptera: Evolution, Biodiversity and Biological Control
Harwood, S. H. & Beckage, N. E. (1994) Purification and characterization of an early-expressed polydnavirus-induced protein from the hemolymph of Manduca sexta larvae parasitized by Cotesia congregata. Insect Biochemistry & Molecular Biology 24: 685-698. Lockhart, P. J., Steel, M. A., Hendy, M. D. & Penny, D. (1994) Recovering evolutionary trees under a more realistic model of sequence evolution. Molecular Biology & Evolution 11: 605-612. Maddison, W. P. & Maddison, D. R. (1992) MacClade, Version 3. Analysis of Phylogeny and Character Evolution. Sinauer Associates, Sunderland, MA. Mardulyn, P. & Whitfield, J. B. (1999) Phylogenetic signal in the COI, 16S and 28S genes for inferring relationships among genera of Microgastrinae (Hymenoptera: Braconidae); evidence of a high diversification rate in this group of parasitoids. Molecular Phylogenetics & Evolution 12: 282-294. Page, R. D. M. (1993) COMPONENT, Version 2.0. Tree comparison software for use with Microsoft Windows. Biogeography and Conservation Laboratory, The Natural History Museum, London. Palumbi, S. R. (1996) Nucleic acids II: the polymerase chain reaction. pp. 205-247. In Hillis, D. M, Moritz, C. & Mable, B. K. (Eds), Molecular Systematics, 2nd edition, Sinauer Associates, Sunderland, MA. Penny, D. & Hendy, M. D. (1985) The use of tree comparison metrics. Systematic Zoology 34: 75-82. Savary, S., Beckage, N., Tan, F., Periquet, G. & Drezen, J.-M. (1997) Excision of the polydnavirus chromosomal integrated EP1 sequence of the parasitoid wasp Cotesia congregata (Braconidae, Microgastrinae) at potential recombinase binding sites. Journal of General Virology 78: 3125-3134. Smith, P. T., Kambhampati, S., Völkl, W. & Mackauer, M. (1999) A phylogeny of aphid parasitoids (Hymenoptera: Braconidae: Aphidiinae) inferred from mitochondrial NADH1 dehydrogenase gene sequence. Molecular Phylogenetics & Evolution 11: 236-245. Steel, M. A., Lockhart, P. J. & Penny, D. (1993) Confidence in evolutionary trees from biological sequence data. Nature 364: 440-442. Stoltz, D. B. (1990) Evidence for chromosomal transmission of polydnavirus DNA. Journal of General Virology 71: 1051-1056. Stoltz, D. B. (1993) The polydnavirus life cycle. pp. 167-187 In Beckage, N. E., Thompson, S. N. & Federici, B. A. (Eds), Parasites and Pathogens of Insects, Vol. 1: Parasites. Academic Press, San Diego. Stoltz, D. B. & Whitfield, J. B. (1992) Viruses and virus-like entities in the parasitic Hymenoptera. Journal of Hymenoptera Research 1: 125-139. Swofford, D.L. (1998) PAUP* version 4.0 b1 – Phylogenetic Analysis Using Parsimony (and Other Methods). Software and Manual. Sinauer Associates, Sunderland, MA. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673-4680. Webb , B. A. & Summers, M. D. (1990) Venom and viral expression products of the endoparasitic wasp Campoletis sonorensis share epitopes and related sequences. Proceedings of the National Academy of Sciences, USA 87: 4961-4965. Whitfield, J. B. (1990) Parasitoids, polydnaviruses and endosymbiosis. Parasitology Today 6: 381-384. Whitfield, J. B. (1994) Mutualistic viruses and the evolution of host ranges in endoparasitoid Hymenoptera, pp. 163-176. In Hawkins, B. A. & Sheehan, W. (Eds), Parasitoid Community Ecology. Oxford University Press, Oxford.
Phylogeny of Microgastroid Braconid Wasps 105
Whitfield, J. B. (1997) Molecular and morphological data suggest a single origin of the polydnaviruses among braconid wasps. Naturwissenschaften 84: 502-507. Whitfield, J. B., & Mason, W. R. M. (1994) Mendesellinae, a new subfamily of braconid wasps (Hymenoptera, Braconidae) with a review of relationships within the microgastroid assemblage. Systematic Entomology 19: 61-76. Xu, D. & Stoltz, D. B. (1991) Evidence for a chromosomal location of polydnavirus DNA in the ichneumonid parasitoid, Hyposoter fugitivus. Journal of Virology 65: 6693-6704.
Evolutionary Transitions in Aphidiinae (Hymenoptera: Braconidae) Paul T. Smith and Srinivas Kambhampati Department of Entomology, Kansas State University, Manhattan, Kansas 66506-4004 USA (email:
[email protected])
Introduction Aphidiine wasps (Braconidae: Aphidiinae) are solitary koinobiont endoparasitoids of aphids. The subfamily is currently considered to include approximately 50 genera and 400 species which are divided into four tribes: Aclitini, Aphidiini, Ephedrini and Praini (Mackauer & Stary 1967; Stary 1988). The Aphidiini is the largest of the four tribes and includes a majority of known genera and species; it is sub-divided into three subtribes, Aphidiina, Monoctonina and Trioxina. Although aphidiines are defined by a number of synapomorphies (e.g. host specialisation and presence of a flexible suture between the second and third tergites of the gaster), significant differences exist in morphology, biology and behaviour among tribes, genera and species. Because of varied interpretation of these differences, there have been disagreements concerning the phylogenetic placement of some aphidiine taxa. A number of different phylogenies have been proposed previously for aphidiines based on embryology, morphology and DNA sequences that differ in the postulated relationships among taxa [see Smith et al. (1999) for a review of the various phylogenetic proposals for aphidiines]. For example, each of the four tribes mentioned above have been suggested as being basal in independent studies of morphology (Mackauer 1961, 1968; Tobias 1967; Tobias & Kyriak 1971; Edson & Vinson 1979; Chou 1984; Gärdenfors 1986; Finlayson 1990) and DNA (Belshaw & Quicke 1997; Smith et al. 1999; Kambhampati et al. 2000). Determination of which group is basal is important because of its strong implications for ingroup character state polarisation. Recently, Smith et al. (1999) and Kambhampati et al. (2000) examined the phylogenetic relationship among aphidiine genera. Smith et al. (1999) proposed a phylogenetic tree for Aphidiinae based on 465 bp of the mitochondrial NADH1 dehydrogenase gene. Their study included 39 taxa representing 14 genera and three braconid outgroups. The length of the NADH1 dehydrogenase gene fragment was invariant among the 39 taxa and alignment of the sequence was straightforward. Smith et al. (1999) found that Praini was basal, with strong quantitative support among aphidiines included in their study. However, the study of Smith et al. (1999) did not include a potentially basal genus, Aclitus Foerster, which possesses a number of presumed plesiomorphic characters (Mackauer 1961; Takada & Shiga 1974). Kambhampati et al. (2000) proposed a phylogenetic tree for aphidiines based on DNA sequence of a portion of the mitochondrial 16S rDNA gene for 47 ingroup taxa representing 24 genera and seven outgroup taxa. Their study indicated the following relationships: Aclitus + (Praini + (Ephedrini + Aphidiini)). Although the genus Aclitus was basal, it was only represented by a single species (A. obscuripennis Foerster) and its position did not have strong quantitative support. The 106
Hymenoptera: Evolution, Biodiversity and Biological Control
Evolutionary Transitions in Aphidiinae (Hymenoptera: Braconidae) 107
finding of Kambhampati et al. (2000) that Praini was basal relative to Ephedrini is congruent with the results reported by Smith et al. (1999). A combined analysis of the NADH1 and 16S rDNA data partitions for 27 ingroup taxa representing 14 genera (Aclitus not included) by Kambhampati et al. (2000) indicated a topology nearly identical to that which was reported by Smith et al. (1999), but with even stronger quantitative support. In the present study we assess whether evolutionary transitions of various morphological and behavioural character states are compatible with the phylogenetic tree inferred from a combined analysis of the mitochondrial NADH1 and 16S rDNA genes by Kambhampati et al. (2000). The mapping of various characters onto the combined evidence tree indicated that many of these characters have a simple evolutionary trajectory, with either a single transition in the case of twostate characters or multiple sequential transitions in the case of multi-state characters.
Materials and Methods A list of taxa and the phylogenetic tree on which this study is based was presented by Kambhampati et al. (2000). The phylogenetic tree was inferred using maximum parsimony methods in PAUP* Ver 4d64 (written by D. L. Swofford). Branch support was assessed by bootstrapping (fast stepwise addition-10,000 replications; Felsenstein 1985) and decay index (Bremer 1994). Maclade v. 3.04 (Maddison & Maddison 1992) was used to examine the evolutionary transitions of the following characters by mapping each state onto the phylogenetic tree: A) egg shape, B) shape of first instar mandibles, C) distribution of first instar abdominal spines, D) final instar hypostomal spur, E) venom apparatus, F) shape of ovipositor sheaths, G) pupation behaviour, and H) emergence hole position. These eight characters were selected because they represent a range in morphology from egg to larva to adult, and a range in behaviour from immediately prior to pupation to immediately following pupal development.
Results and Discussion Egg shape and first instar larval structures Tremblay and Calvert (1971) examined the systematic position of aphidiines based on embryology. In their study they identified two distinct differences in egg shape associated with aphidiines, those that are prolongately oval and those that are lemon shaped. Mapping each state onto our phylogenetic tree indicated that prolongately oval eggs constitutes the plesiomorphic state, and lemon-shaped eggs, in Aphidiini, the apomorphic state (Fig. 1A). Similarly, first instar larval aphidiines were examined for differences in morphology by O’Donnell (1989). Two characters associated with the head and abdomen (i.e., mandible shape and abdominal spine distribution) were identified to each have two distinct states. With respect to mandible shape, O’Donnell (1989) found that the first instar larvae exhibited either hook-shaped mandibles or sickle-shaped mandibles; and that the abdominal spines were arranged either in a continuous or discontinuous fashion (i.e. no apparent pattern to arrangement). Our phylogenetic tree indicated two identical evolutionary lineages with respect to the two characters (Figs 1B, 1C). In this regard, hook-shaped mandibles and a continuous arrangement of the abdominal spines are plesiomorphic with a single evolutionary transition for both characters in Aphidiini to sickle-shaped mandibles and a discontinuous arrangement of the abdominal spines.
Fourth instar hypostomal spur Finlayson (1990) conducted a systematic study of aphidiines in which various cephalic structures of fourth instar larvae were examined for differences. Of particular interest is the presence/
Paul T. Smith and Srinivas Kambhampati 108
Hymenoptera: Evolution, Biodiversity and Biological Control
(A) Egg Shape
Prolonged, Oval Lemon Shape Equivocal
(B) 1st Instar Mandible Shape Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
(C) 1st Instar Abdominal Spine Distribution
Continuousl Discontinuous Equivocal
Figure 1
Sickle Shape Hook Shape Equivocal
Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
(D) 4th Instar Hypostomal Spur Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
Present Absent
A phylogenetic tree for Aphidiinae based on DNA sequence of portions of the 16S rDNA and NADH1 dehydrogenase genes. The tree shown is a majority rule consensus of two equally parsimonious trees for 27 aphidiine taxa representing 14 genera and an outgroup (monophyletic genera reduced to a single taxon name). The evolutionary transitions of the following characters were examined: A) egg shape; B) first instar mandible shape; C) first instar abdominal spine distribution; D) fourth instar hypostomal spur;
Evolutionary Transitions in Aphidiinae (Hymenoptera: Braconidae) 109
(E) Venom Apparatus
Type 2
(F) Ovipositor Sheath Shape Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
Curved Up
Type 1
Short, Curved Up
Type 1-basal filament attachment
Straight Curved Down Equivocal
Type 1-separate filament attachment Equivocal
(G) Pupation Behavior
Outside Inside
Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
(H) Emergence Hole Position Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
Cocoon Emergence
Aphidius Diaeretiella Lysiphlebus Euaphidius Pauesia Monoctonus Falciconus Paramonoctonus Lipolexis Trioxys Binodoxys Ephedrus Dyscritulus Praon Perilitis
EH at cauda EH at cornicles cap bears cornicle(s) EH between cornicles and cauda EH between thorax and cornicles Equivocal
Figure 1 (cont’d)
E) venom apparatus; F) ovipositor sheath shape; G) pupation behavior, and H) emergence hole position. These characters represent a range in morphology from egg to larva to adult and a range in behavior from immediately prior to undergoing pupation to immediately following pupal development.
Paul T. Smith and Srinivas Kambhampati 110
Hymenoptera: Evolution, Biodiversity and Biological Control
absence of a hypostomal spur. Finlayson (1990) found that only a sub-set of aphidiine genera (i.e. the Trioxina) possess this structure. Since most braconids possess a hypostomal spur, Finlayson (1990) reasoned that those aphidiines that have lost this structure are more ‘derived’ (i.e. she assumed the sister group to aphidiines possessed a hypostomal spur and that possession of this character is the apomorphic condition). If Finlayson’s (1990) hypothesis of aphidiine evolution is correct, it would imply that there has been a considerable amount of convergence during the course of aphidiine evolution (e.g. egg shape, first instar mandible shape, and abdominal spine distribution). In marked contrast, our phylogenetic tree indicates that there was an initial loss and secondary acquisition of this character within a monophyletic group of genera within the Aphidiini (Fig. 1D). Although our explanation is less parsimonious than Finlayson’s (1990), it requires only one additional step and is more parsimonious when the evolutionary transitions of many other characters are considered.
Venom apparatus Braconid wasps exhibit one of two different types of venom apparatus (VA) morphology following the classification scheme of Edson and Vinson (1979). In general, the VA consists of a venom reservoir, gland filaments, and a venom duct which extends into the ovipositor. Following Edson and Vinson’s (1979) classification scheme, one VA type (Type 1) is characterised by a cone shaped reservoir surrounded by many muscles and two or more gland filaments, and is associated with ecto- or endo-parasitoids which pupate inside the host remains. The Type 1 VA can be further sub-divided based on structure (e.g. attachment site of the gland filaments and/or appearance of the venom reservoir). A second type of VA (Type 2) is characterised by a thin walled reservoir surrounded by few muscles and only two gland filaments, and is associated with endoparasitoids which pupate outside the host remains. Unlike other braconid subfamilies which are monomorphic for either type, the Aphidiinae exhibit both types. The VA classification scheme of Edson and Vinson (1979) is highly compatible with our phylogenetic tree with four evolutionary lineages being recognised (Fig. 1E). The plesiomorphic VA structure within Aphidiinae is a Type 2 structure with no apparent reservoir and is exhibited by members of the tribe Praini which pupate outside and underneath the exoskeleton of the host (Fig. 1E). The apomorphic state is a Type 1 structure and is exhibited by members of Ephedrini and Aphidiini, both of which pupate inside the host. However, the Type 1 VA of Ephedrini differs from that of Aphidiini in that the gland filaments are attached at the base of the venom reservoir and the cuticular lining of venom reservoir lacks a spiral-like appearance. Within Aphidiini, the genera Monoctonus Haliday, Falciconus Mackauer and Paramonoctonus Stary (=Monoctonina) can be further distinguished by the separate attachment of the two gland filaments to the venom reservoir and the cuticular lining of the venom reservoir lacks a spiral-like appearance (Fig. 1E).
Ovipositor sheath shape Female aphidiines exhibit variation in the shape and length of the ovipositor sheaths (3rd valvulae). The ovipositor sheaths can be sparsely or densely pubescent, but they always bear several sensory hairs at their apex. The variation in ovipositor sheath shape ranges from short to long and either straight, or curved up or down. According to Stary (1981), the ovipositor sheaths and their functional differentiation are one of the key characters in the evolution of aphidiine groups. Our phylogenetic tree indicates the presence of four evolutionary lineages with respect to shape and size of the ovipositor sheaths (Fig. 1F). The morphological features of the ovipositor sheaths
Evolutionary Transitions in Aphidiinae (Hymenoptera: Braconidae) 111
are apparent adaptive strategies in parasitoid/host interactions. For example, the downward curved ovipositor sheaths of Monoctonus, Falciconus, Paramonoctonus, Lipolexis Foerster, Trioxys Haliday, and Binodoxys Mackauer apparently work in conjunction with a pair of prongs on the last abdominal sternum to hold the attacked aphid and prevent its escape (Stary 1981).
Pupation behaviour Following the completion of larval development, aphidiines will pupate either inside the dead host (most extant species) or outside and underneath the empty exoskeleton of the host (most members of Praini). Mackauer (1961) proposed that outside pupation in Praini evolved by way of secondary loss of internal pupation and may represent an adaptation in response to hyperparasitoids. Our results indicated that this behaviour is plesiomorphic with a single transition to internal pupation among members of Ephedrini and Aphidiini (Fig. 1G). In contrast, Mackauer’s (1961) proposal would imply multiple transitions (i.e. from internal pupation in Ephedrini to external pupation in Praini and then back to internal pupation in Aphidiini). It is possible, however, that internal pupation is plesiomorphic despite which aphidiine lineage is basal. Central to this issue is the phylogenetic position of the genus Areopraon Mackauer (a rare genus currently classified in Praini not included in our study) which contains species which pupate internally and others which pupate externally or both depending on the circumstances (Stary 1970). If an internally pupating Areopraon species is basal to all other Paini, then Mackauer’s (1961) proposal of internal pupation being plesiomorphic is possible and would be the most parsimonious explanation as it implies only a single transition. However, the relationship of Areopraon to other aphidiines (including Praini) remains to be vigorously tested with cladisitc analysis. Emergence hole position Following the completion of immature development, the adult aphidiine uses its mandibles to cut an emergence hole in the cocoon (Praini) or host exoskeleton (Aclitini, Aphidiini, Ephedrini). Among the aphidiines that pupate inside the host, some species cut an emergence hole between the thorax and cornicles, some species cut an emergence hole at the level of the cornicles (in such case the cap may contain one or both cornicles), some species cut an emergence hole posterior to the cornicles (in such case the cap never contains a cornicle), and finally some species cut an emergence hole perpendicular to the longitudinal axis of the host exoskeleton (in such case the cap bears the cauda and is relatively large; Stary 1970). With respect to emergence hole position, our phylogenetic tree indicates the presence of four distinct evolutionary lineages (Fig. 1H). Of those aphidiines that cut an emergence hole in the exoskeleton of the host, there appears to be a gradual transition in emergence hole position from the cauda to a dorsal position just posterior to the thorax (Fig. 1H). In summary, our results indicated that the phylogenetic tree inferred from parsimony analysis of the mitochondrial NADH1 and 16S rDNA genes was useful with respect to tracing the evolutionary transitions of various morphological, biological, and behavioural characters in Aphidiinae. Our trees are generally compatible with a smooth transition from one character state to another and in most cases, represent the most parsimonious solution. Some characters are evolutionarily more labile than others (e.g. egg shape, mandible shape, pupation behaviour). An examination of these characters in a broader range of taxa within Braconidae may provide insights into which characters are generally more labile and which are conserved.
Paul T. Smith and Srinivas Kambhampati 112
Hymenoptera: Evolution, Biodiversity and Biological Control
Acknowledgements We thank W. Völkl, R. Belshaw, J. Obrycki, and P. Stary for providing some of the specimens used in this study, and M. Mackauer for discussion of aphidiine evolution. Financial support for this study was provided by United States Department of Agriculture NRI grant 9401865 to S. K and M. M and Hatch grants H-28 and H-497 to S.K. This is proceedings/book article no. 99-296A of the Kansas Agriculture Experiment Station.
References Belshaw, R. & Quicke, D. L. J. (1997) A molecular phylogeny of the Aphidiinae (Hymenoptera: Braconidae). Molecular Phylogenetics & Evolution 7: 281-293. Bremer, K. (1994) Branch support and tree stability. Cladistics 10: 295-304. Chou, L-Y. (1984) The phylogeny of Aphidiidae (Hymenoptera). Journal of Agricultural Research (China) 33: 437-446. Edson, K. M. & Vinson, S. B. (1979) A comparative morphology of the venom apparatus of female braconids (Hymenoptera: Braconidae). Canadian Entomologist 111: 1013-1024. Felsenstein, J. (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791. Finlayson, T. (1990) The systematics and taxonomy of final instar larvae of the family Aphidiidae. Memoirs of the Entomological Society of Canada 152: 1-74. Gärdenfors, U. (1986) Taxonomic and biological revision of Palaearctic Ephedrus (Haliday) (Hymenoptera, Braconidae, Aphidiinae). Entomologica Scandinavica Supplement 27: 1-95. Kambhampati, S., Völkl, W. and Mackauer, M. (2000) Phylogenetic relationships among genera of Aphidiinae (Hymenoptera: Braconidae) based on DNA sequence of the mitochondrial 16S rRNA gene. Systematic Biology 25: 1-9. Mackauer, M. (1961) Die Gattungen der Familie Aphidiidae und ihre verwandtschaftliche Zuordnung (Hymenoptera: Ichneumonoidea) Beiträge zur Entomologie 11: 792-803. Mackauer, M. (1968) Hymenopterorum Catalogus. Pars 3. Aphidiidae, Junk, The Hage. Mackauer, M. & Stary, P. (1967) World Aphidiidae (Hymenoptera: Ichneumonoidea). Le Francois, Paris. Maddison, W. P. & Maddison, D. R. (1992) MacClade: Analysis of Phylogeny and Character Evolution. Version 3.04. Sinauer Associates, Sunderland, MA. O’Donnell, D. J. (1989) A morphological and taxonomic study of first instar larvae of Aphidiinae (Hymenoptera: Braconidae). Systematic Entomology 14: 197-219. Smith, P. T., Kambhampati, S., Völkl, W. & Mackauer, M. (1999) A phylogeny of aphid parasitoids (Hymenoptera: Braconidae: Aphidiinae) inferred from mitochondrial NADH 1 dehydrogenase gene sequence. Molecular Phylogenetics & Evolution 11: 236-245. Stary, P. (1970). Biology of Aphid Parasites (Hymenoptera: Aphidiidae) with Respect to Integrated Control. Series Entomologica 6, Junk, The Hague. Stary, P. (1981) Biosystematical classification of Trioxys Hal. and related genera (Hymenoptera, Aphidiidae). Bollettino del Laboratoria di Entomologia Agraria “Filippo Silvestri”, Portici 38: 84-93. Stary, P. (1988) Aphidiidae, pp. 171-184. In Minks, A. K. & Harrewijn, P. (Eds). Aphids, Their Biology, Natural Enemies and Control. Vol 2B. Elsevier, Amsterdam. Takada, H. & Shiga, M. (1974) Description of a new species and notes on the systematic position of the genus Aclitus (Hymenoptera: Aphidiiidae). Kontyû 42: 283-292.
Evolutionary Transitions in Aphidiinae (Hymenoptera: Braconidae) 113
Tremblay, E. & Calvert, D. (1971) Embryosystematics in the aphidiines (Hymenoptera: Braconidae). Bollettino del Laboratoria di Entomologia Agraria “Filippo Silvestri”, Portici 29: 223-249. Tobias, V. I. (1967). A review of the classification, phylogeny and evolution of the family Braconidae (Hymenoptera) Entomologicheskoe Obozrenie 56: 646-659. Tobias, V. I. & Kyriak, I. G. (1971) Areopraon pilosum Mackauer, 1959, and problems concerning phylogeny and distribution of the family Aphidiidae (Hymenoptera). Entomologicheskoe Obozrenie 50: 11-16.
Genetic Structure Of The Cypress Seed Chalcid Megastigmus wachtli (Torymidae) within its Mediterranean Distribution J. Y. Rasplus1, E. Carcreff2, J. M. Cornuet1 and A. Roques3 1
INRA, Centre de Biologie et de Gestion des Populations, 488 rue Croix de Lavit, 34090 Montpellier, France (email:
[email protected]) 2
INRA, Unités de Recherches Forestières, Pierroton, France
3
INRA, Zoologie Forestière, Ardon Olivet, France
Introduction The evergreen cypress (Cupressus sempervirens L.) is widespread in the Mediterranean region. Its natural range extends from North Iran to Crete and the Dodecanese Islands (Vidacovic 1991). This species has been introduced in the western part of the Mediterranean Basin (Western Europe and North Africa), first by the Ancient Greeks and then by the Romans (Baumann 1982). Once established, the species propagated spontaneously and spread along the Mediterranean coast. Species belonging to the cosmopolitan genus of seed chalcid Megastigmus Dalman have long been considered as important pests of commercially grown conifers. In Europe, the cypress seed chalcid Megastigmus wachtli Seitner mostly develops within the seeds of C. sempervirens. Occasionally the species can develop within seeds of introduced Californian species of Cupressus (C. ambramsiana C.B. Wolf, C. arizonica Greene, C. bakeri Jepson and C. goveniana Gordon) (Roques et al. 1999a, 1999b). Megastigmus wachtli occurs both in the natural (Canakcioglu 1959; Roques et al. 1997) and the introduced range of C. sempervirens (Ben Jamaa & Roques 1997; Guido et al. 1995; Roques & Raimbault 1986). The species is also thought to be associated with C. atlantica Gaussen, an endangered species endemic to the Atlas Mountains (Morocco) (Fernandes 1979), but wasps from this host tree show slight differences and are thought to represent different species (see below). Due to low variability among the loci analysed, it has been difficult to characterise Megastigmus populations developing on related host plants with allozyme markers (Roux & Roques 1996). Microsatellites increasingly show their potential for genetic studies of the Hymenoptera, a group where haplodiploidy is a probable cause of the low level of allozyme polymorphisms (Pamilo & Crozier 1981). Until now, these single locus co-dominant markers have been reported from social Hymenoptera (Aculeata) but never from the large and economically important parasitic wasps which include both pests and beneficial insects. For the first time we use highly polymorphic microsatellites to assess the genetic structure of populations of the cypress seed chalcid. We also compare populations associated with the evergreen cypress in its supposed native range and in areas where the cypress has recently been introduced by humans.
114
Hymenoptera: Evolution, Biodiversity and Biological Control
Genetic Structure of the Cypress Seed Chalcid 115
Figure 1
Geographic locations of sampling sites and number of collected individuals of the cypress seed chalcid, Megastigmus wachtli.
Material and Methods Sampling and sequencing In each sampling site (Fig. 1), 100 mature 2-year-old cones were collected in late spring on 10 different trees of C. sempervirens. The cones were stored in boxes and exposed to outdoor conditions until wasps emerged. They were then stored at -80˚C until DNA analysis. Because of the haplodiploid sex determination in Hymenoptera, only females were used for the genetic study. We collected 191 females from 10 localities. Five within the supposed natural range of C. sempervirens (Imbros and Zourva – both in Crete near Lévka Óri, Samos, Kos and Rhodes) and five in the area where the species is supposed to have been introduced (mainland Greece, France, Italy and Tunisia) (Fig. 1). We also included five specimens from Morocco developing in seeds of C. atlantica, a rare plant species localised to small patches in the Atlas Mountains and consequently difficult to sample. The isolation and characterisation of microsatellite loci followed Carcreff et al. (1998). Primer sequences and PCR conditions are detailed for each locus in Table 1. Statistical analyses For each locus and each population sample, the gene diversity was estimated by [HE = (n/n-1) (1-∑ipi2)] where pi is the frequency of allele i estimated over n chromosomes sampled at this locus in this population (Nei 1987). Gene diversity was also quantified by the number of alleles per locus (A) and observed heterozygosity (Ho). Deviation from Hardy-Weinberg equilibrium, linkage disequilibrium, differences in allele frequencies and isolation by distance were analysed using GENEPOP version 3.1a (Raymond & Rousset 1995). Population structure was analysed by FST (θ) and RST (ρ). FSTAT (Goudet 1995) was used to calculate unbiased estimates for analogues of Wright’s (1951) F-statistics : θIS, θIT, θST, following Weir and Cockerman (1984). Permutation procedures (N=1000) were used to test whether values were significantly greater than zero by permuting multilocus genotypes among samples. ρ, an estimator of RST, was calculated using RSTCALC 2.1 (Goodman 1997). This unbiased estimator is analogous to θ, and takes into
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 116
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Characteristics of six polymorphic loci in Megastigmus wachtli. Optimal annealing temperatures (T°m) and MgCl2 concentrations (mM) are also given. Annealing temp (°C)
MgCl2 (mM)
GeneBank N°
TGCTGAGCCCCTCTATACCC AGAACCAAAAGGCGTGCG
54°C
1.2
AJ101065
(TC)11
ACCCGCTTTGCTCATCC TGTGCTGCTACACCGAGG
52°C
1.2
AJ101066
MW 22
(CA)8 TA (CA)7
CCATCCTCGAACCTTTTTACC GCTGCTGCTGCTTCTACCTC
58°C
1
AJ101067
MW 34
(AG)17 G10
CCCCGCCTCTACCAAATC TTGAAATTGCTCGGACCG
58°C
0.8
AJ101068
MW 47
(CT)13
CATGGGGTTTCAAGTGCTG CTCTAGCTTTTGCCTGAGCC
58°C
1
AJ101069
Locus
Core sequence
Primer sequences (5’ 3’)
MW 10
(CT)20
MW 21
account differences in sample size between populations and differences in variance between loci. Migration rates (Nm) were calculated as Nm = 1/4((1/θ)-1). Nm for r were calculated as for q. We compared multilocus FST/(1-FST) and RST/(1-RST) estimates over geographical distances for all pairs of populations within the natural distribution range of the species. Geographical distances between populations were the shortest distances measured on a map. Neighbour-joining (NJ) (Saitou & Nei 1987) trees were constructed using two different distances: Cavalli-Sforza and Edwards’ chord distance (DCE) (Cavalli-Sforza & Edwards 1967) and Chakraborty and Jin’s distance (DAS) (Chakraborty & Jin 1993). Bootstrap values were computed by resampling loci and are given as percent values of 2000 replications. Chakraborty and Jin’s distance, based on the proportion of shared alleles (DAS), was also used to assess relationships among individuals. NJ trees relating populations and individuals were constructed using programs written by J. M. Cornuet.
Results Variability of microsatellite loci For each of the five loci across the nine sampled populations, allele frequencies, proportions of heterozygotes and gene diversities are given in Table 2. The number of alleles detected at the five loci analysed varied from six (locus M21) to 35 (locus M34) and the number of alleles detected per population varied from one (in Greece – Locus M21; Rhodes and Samos – Locus M22; Italy – Locus M47) to 19 (for both Cretan populations at Locus M34). The mean number of alleles per locus ranged from 2.0±0.707 in Italy to 10.2±6.38 in Crete (Imbros), resulting in an overall expected heterozygosity ranging from 0.118±0.102 in France to 0.816±0.146 in Crete (Zourva). A total of 85 alleles were observed but two loci (M10 and M34) accounted for about two thirds (58 alleles) of the total. The remaining three loci show relatively lower allelic variability. The average number of alleles per locus is higher for the two Cretan populations (Table 3). When adjusted for equal population size (Ewens 1972), these two populations have significantly more alleles than all the other populations except the Moroccan population (Wilcoxon’s signed rank test, 0.028
Genetic Structure of the Cypress Seed Chalcid 117
Hardy-Weinberg and linkage disequilibriums The deviations of genotype frequencies from Hardy-Weinberg equilibrium expectations were significant for three populations (France, Tunisia and Greece). The Italian population was only just non-significant (P=0.06). These deviations mostly appear to derive from the presence of individuals which share the same five-locus genotype. Linkage disequilibrium measures the departure of the observed association of alleles of different loci from expected values calculated on the basis of random association. Exact tests for genotypic linkage disequilibrium between microsatellite loci gave nine significant adjusted P values at the 5% level out of 67 comparisons (13.4%), but six of these values occurred in the French population. These deviations appear to derive mostly from inclusion of a high proportion of identical copies within the French population (N. B. 12 individuals out of the 17 are copies of a single genotype). When the French population was removed, exact tests gave three significant adjusted P values at the 5% level out of 57 comparisons (5.3%), a proportion slightly higher than expected by chance alone (2.9% expected). This result suggests low or no physical linkage between loci.
Population differentiation and structure Log-likelihood (G)-based exact tests for differentiation of all pairwise combinations of populations at each locus (n=225 tests) were performed. Most of these comparisons (188 out of 225) yielded significant differences. Fourteen of the 37 non-significant results were at the least variable locus (M21), and 11 of them were between Kos, Samos and Rhodes populations. Fixation indices are a measure of differentiation which incorporates information on both the frequency and the identity of alleles. We calculated both q (FST) and r (RST ) values to describe the overall differences between pairs of populations (Table 3). Both, q and r values, were significantly different from zero. q ranged from low values between geographically close populations within the natural range of C. sempervirens (Dodecanese Islands, Cretan and Greek populations), to very high values between the Moroccan population and all other. On average, ρ values tend to be greater than θ values (Table 3). This result is consistent with the assumption that mutations are expected to play a major role when the time scale of interest is long (Slatkin 1995). Excluding the Moroccan population, the mean θ and ρ averaged across all pairwise population comparisons were 0.324 and 0.365, respectively. Pairwise gene flow estimates are obtained from both FST and RST and ranged from Nm (θ)=0.125, Nm (r)=0.09 between Italy and Morocco populations; to Nm (θ)=40.07 and Nm (ρ) = 19.12 between Kos and Samos. Values of Nm less than one are sufficient to allow strong genetic differentiation to develop by permitting alternative alleles to drift to fixation in different populations. Where Nm exceeds one, some differentiation may occur, but alternative alleles rarely become fixed. Values strongly exceeding one indicate panmixia. The observed values indicate strong levels of differentiation for most populations except some which are geographically close (Kos-Samos-Rhodes, Cretan and Greek populations) (Table 4). Matrices of pairwise multilocus FST estimates computed for the east Mediterranean populations were not significantly correlated to geographic distances (P=0.06, Mantel’s test). In contrast, significant correlations were detected between matrices of pairwise multilocus RST estimates and geographic distances (P=0.003, Mantel’s test), indicating that isolation by distance does influence the degree of population differentiation within the natural range of M. wachtli.
Genetic relationships among populations NJ trees based on DCE and DAS distances are given in Figure 2. They show an overall similar topology, although there are some differences between them. Both trees show two distinct clusters.
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 118
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Allelic variability at five microsatellite loci in Megastimus wachtli populations. Number of alleles per locus (A), allelic size in bp (S), allelic frequencies, gene diversity (HE), heterozygote proportion (HO) and number of samples successfully genotyped (N) are given for each population and locus. Locus M21
S
Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
N
10
40
48
34
14
42
50
54
50
40
A
5
2
3
2
1
3
4
2
3
2
142
0.100
-
-
-
-
-
-
-
-
-
146
0.200
-
-
-
-
-
-
-
-
-
Means 2.7
148
0.300
0.900
0.208
0.971
1.000
0.690
0.480
0.870
0.860
0.825
150
0.300
-
-
0.029
-
0.167
0.360
0.130
0.100
0.175
152
0.100
0.100
0.771
-
-
0.143
0.140
-
0.040
-
154
-
-
0.021
-
-
-
0.020
-
-
-
Ho
0.600
0.200
0.333
0.059
-
0.524
0.680
0.259
0.280
0.350
HE
0.844
0.185
0.370
0.059
-
0.487
0.633
0.230
0.254
0.296
0.336
Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
Means
N
10
40
48
34
14
42
50
54
50
40
A
3
2
3
3
2
6
5
1
1
2
93
-
-
-
-
-
0.024
-
-
-
-
95
0.400
-
0.042
0.088
0.643
0.548
0.580
1.000
1.000
0.975
103
0.500
-
-
-
-
-
-
-
-
-
105
0.100
-
-
-
-
-
-
-
-
-
107
-
-
-
-
-
0.071
0.140
-
-
-
109
-
-
-
0.029
-
0.214
0.180
-
-
0.025
Locus M22
S
0.329
2.8
111
-
0.250
0.417
-
0.357
0.119
0.080
-
-
113
-
0.750
0.542
0.882
-
0.024
0.020
-
-
-
Ho
0.800
0.500
0.417
0.059
0.143
0.619
0.640
-
-
0.050
0.323
HE
0.644
0.385
0.543
0.219
0.495
0.649
0.617
-
-
0.050
0.36
Genetic Structure of the Cypress Seed Chalcid 119
Locus M10
S
N
Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
10
40
48
34
14
42
50
54
48
40
A
Ho HE
3
2
5
3
5
14
14
7
7
7
101
-
-
-
-
-
0.024
-
-
-
-
105
-
-
-
-
-
0.095
-
-
-
-
109
-
-
-
-
-
0.024
-
-
-
-
119
-
-
-
-
-
0.024
-
-
-
-
126
-
-
-
-
-
0.024
-
-
-
-
155
0.800
-
-
-
-
-
-
-
-
-
157
0.100
-
-
-
-
-
0.020
-
-
-
163
0.100
-
-
-
-
0.024
0.020
-
-
-
165
-
-
-
-
-
0.048
-
-
-
-
169
-
-
-
-
-
-
0.020
-
-
-
171
-
-
-
-
0.643
0.214
0.320
0.222
0.125
0.050
173
-
0.975
0.625
0.882
-
0.286
0.200
0.296
0.167
0.250
175
-
0.025
0.042
-
0.143
0.071
0.180
-
0.021
-
177
-
-
0.125
-
0.071
-
0.040
0.074
0.021
0.050
179
-
-
0.167
-
-
0.024
0.040
0.056
-
0.025
181
-
-
-
0.029
0.071
0.024
0.020
0.019
0.042
0.075
183
-
-
0.042
0.088
-
0.071
0.020
0.259
0.479
0.450
185
-
-
-
-
-
-
-
0.074
0.146
0.100
187
-
-
-
-
-
-
0.040
-
-
-
189
-
-
-
-
-
0.048
-
-
-
-
193
-
-
-
-
0.071
-
0.040
-
-
-
195
-
-
-
-
-
-
0.020
-
-
-
199
-
-
-
-
-
-
0.020
-
-
-
0.400 0.378
0.050 0.050
0.333 0.574
0.059 0.219
0.286 0.593
0.857 0.865
0.920 0.833
0.815 0.796
0.667 0.718
0.700 0.732
Means 6.7
0.509 0.576
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 120
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Allelic variability at five microsatellite loci in Megastimus wachtli populations. Number of alleles per locus (A), allelic size in bp (S), allelic frequencies, gene diversity (HE), heterozygote proportion (HO) and number of samples successfully genotyped (N) are given for each population and locus. (continued) Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
N
Locus M47
S
10
40
48
34
14
42
50
54
50
40
A
4
1
3
3
2
9
6
5
6
5
121
0.200
-
-
-
-
-
-
-
-
-
129
0.600
-
-
-
-
-
-
-
-
-
131
0.100
-
-
-
-
0.024
-
-
-
-
133
-
-
-
-
-
0.119
0.100
0.093
0.080
0.025
135
-
1.000
0.896
0.912
-
0.143
0.260
0.019
0.120
0.050
137
-
-
0.042
0.029
0.786
0.238
0.240
0.611
0.360
0.425
139
-
-
0.063
0.059
0.214
0.143
0.100
0.259
0.340
0.450
141
-
-
-
-
-
0.262
0.280
-
0.080
0.050
143
-
-
-
-
-
-
-
0.019
0.020
-
145
-
-
-
-
-
0.024
0.020
-
-
-
151
0.100
-
-
-
-
-
-
-
-
-
161
-
-
-
-
-
0.024
-
-
-
-
163
Means 4.4
-
-
-
-
-
0.024
-
-
-
-
Ho
0.400
-
0.125
0.118
0.143
0.857
0.880
0.556
0.760
0.500
0.434
HE
0.644
-
0.196
0.169
0.363
0.837
0.792
0.560
0.742
0.627
0.493
Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
Means
10
40
48
34
14
42
50
54
50
40
Locus M34
S
N A
7
3
5
6
6
19
19
9
11
8
102
-
-
-
0.029
-
-
0.080
-
-
0.025
104
-
-
-
-
0.071
0.048
0.040
-
-
-
106
-
-
-
-
-
0.095
-
-
-
-
107
-
-
-
-
-
-
0.020
-
-
-
108
-
-
-
-
-
0.024
-
-
-
-
9.3
Genetic Structure of the Cypress Seed Chalcid 121
159
-
-
-
-
-
0.024
-
-
-
160
0.100
-
-
-
-
-
-
-
-
-
164
-
-
-
-
-
0.024
0.020
0.019
0.040
-
165
-
-
-
-
-
-
0.020
-
0.020
-
166
0.100
-
-
-
0.143
0.048
0.100
0.241
0.460
0.275
167
0.200
-
-
-
0.357
0.071
0.180
0.019
0.040
-
168
-
-
-
-
-
0.048
0.040
-
0.020
-
169
-
0.075
0.083
-
0.286
0.071
0.120
-
-
-
170
-
0.050
-
-
-
0.048
0.060
-
-
-
171
-
0.875
0.375
-
-
0.095
0.060
-
0.040
-
173
0.100
-
0.083
0.029
-
0.095
0.040
-
0.020
0.100
174
-
-
-
-
0.071
0.024
0.060
-
-
-
175
-
-
-
0.029
-
-
0.020
-
-
0.025
176
-
-
-
-
-
0.048
-
-
-
-
177
-
-
-
-
-
0.024
0.020
-
0.040
-
178
-
-
-
-
-
-
0.060
0.019
-
-
179
-
-
-
-
-
-
-
0.556
0.220
0.400
180
-
-
-
0.029
-
-
0.020
0.019
-
0.075
181
-
-
-
-
-
-
-
0.093
0.040
0.075
182
-
-
-
-
-
0.048
-
-
-
-
184
-
-
-
-
-
0.071
-
-
-
-
185
-
-
-
-
0.071
-
-
0.019
-
-
186
0.300
-
0.063
-
-
-
0.020
-
-
-
188
-
-
0.396
0.794
-
0.048
-
-
0.060
-
190
-
-
-
0.088
-
-
-
-
-
0.025
191
0.100
-
-
-
-
-
-
-
-
-
192
-
-
-
-
-
-
-
0.019
-
-
195
-
-
-
-
-
0.048
-
-
-
-
197
0.100
-
-
-
-
-
-
-
-
-
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 122
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Allelic variability at five microsatellite loci in Megastimus wachtli populations. Number of alleles per locus (A), allelic size in bp (S), allelic frequencies, gene diversity (HE), heterozygote proportion (HO) and number of samples successfully genotyped (N) are given for each population and locus. (continued) 199
-
-
-
-
-
-
0.020
-
-
-
Ho
0.600
0.100
0.625
0.294
0.429
0.952
0.960
0.519
0.840
0.550
HE
0.911
0.232
0.699
0.369
0.813
0.959
0.933
0.635
0.742
0.760
0.705
Morocco
Italy
Tunisia
France
Greece
Imbros
Zourva
Rhodes
Samos
Kos
Means
Mean Allele Number
4.400
2.000
3.800
3.400
3.200
10.200
9.600
4.800
5.600
4.800
Allele Number Standard Deviation
1.673
0.707
1.095
1.517
2.168
6.380
6.580
3.347
3.847
2.775
Mean Heterozygote proportion
0.560
0.170
0.367
0.118
0.200
0.762
0.816
0.430
0.509
0.430
Heterozygotes proportion Standard Deviation
0.167
0.199
0.180
0.102
0.163
0.181
0.146
0.311
0.357
0.246
Mean Gene Diversity
0.684
0.170
0.476
0.207
0.453
0.760
0.761
0.444
0.491
0.493
Gene Diversity Standard Deviation
0.209
0.153
0.196
0.112
0.302
0.190
0.135
0.323
0.345
0.309
All loci
0.587
Genetic Structure of the Cypress Seed Chalcid 123
Imbros Zourva
74 53 66
97 A
Figure 2
98
43 67
Morocco 58
Greece Rhodes Samos Kos France Italic Tunisia
96
Imbros Zourva 44 90
74
Morocco
96
Greece Samos Rhodes 24 Kos Tunisia Italy France
B
Neighbour-Joining (NJ) trees relating the 10 populations of Megastigmus wachtli. Trees were reconstructed using A) Cavalli-Sforza and Edwards’ (1967) chord distance (DCE) and B) Chakraborty and Jin’s (1993) shared allele distance (DAS). Bootstrap values have been computed by resampling loci and are given as percentage over 2000 replications. All trees are unrooted.
The first one groups the western populations of M. wachtli (Tunisia, France and Italy) with very high bootstrap values (>95%). The second one clusters the Moroccan and the east Mediterranean populations (Crete, Greece, Kos, Samos and Rhodes) (with the same bootstrap values). The Greek population, which was previously thought to be introduced, groups strongly with the other eastern Mediterranean populations. Within this last group, the two Cretan populations cluster together in the DCE tree and the remaining populations (Greece, Kos, Samos and Rhodes) form a closely related group, supported by relatively high bootstrap values, despite the large geographic separation between continental Greece and the Dodecanese Islands.
Genetic relationships among individuals The dendrogram of all 191 individuals based on DAS distance (Fig. 3) shows some differences compared to the population trees. Individuals collected in Morocco from cones of C. atlantica cluster together and are separated from the individuals of M. wachtli developing in C. sempervirens, which in turn are divided in two clusters. The first cluster groups a majority of individuals from Zourva (15 out of 25), a few individuals from Imbros (6 out of 21) and all the individuals from western populations supposed to have been introduced, including the population from Tunisia. These last individuals cluster together, on relatively long basal branches, in a monophyletic assemblage which is grouped with the Cretan individuals. This is the most striking difference between the individual and population trees. The lower genetic variability of Megastigmus individuals occurring within the introduced range of the evergreen cypress is confirmed by the shorter branch length of this group compared with that for the eastern Mediterranean cluster. French and Italian populations are clustered with Tunisian individuals and show comparatively lower genetic variability. Identical copies of individuals are characterised by rake arrangements. The second cluster groups all individuals from the Dodecanese Islands, Greece and some Cretan individuals. Within this lineage, populations appear relatively mixed except for most individuals from Crete (10 from Zourva and 15 from Imbros) which are grouped in the basal part of the cluster. It should be noted that one individual reared in France from C. goveniana cones (a cypress species introduced from the Nearctic) clusters with the eastern Mediterranean individuals. Discrepancy between population and individual trees is mostly due to differences between the reconstruction methods. Population trees are built using methods based on allele frequencies, and are highly sensitive to change in frequencies due to strong founder effects or to changes in type of reproduction (sexual versus asexual). The method used to build the individual tree is
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 124
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 3 Pairwise multi locus ρ estimates (above the diagonal) and θ estimates (below the diagonal). Gene flow estimates in terms of Nm obtained from both FST estimates are given in brackets. Morocco Morocco Italy
Italy
Tunisia
France
Greece
Imbros Crete
ZourvaCrete
Rhodes
Samos
Kos
0.736
0.718
0.699
0.690
0.402
0.596
0.768
0.773
0.774
(0.090)
(0.098)
(0.108)
(0.113)
(0.371)
(0.169)
(0.075)
(0.074)
(0.073)
0.591
0.255
0.462
0.235
0.360
0.769
0.736
0.727
(0.173)
(0.094)
0.667 (0.125)
Tunisia France Greece Imbros Zourva Rhodes Samos Kos
(0.731)
(0.291)
(0.813)
(0.445)
(0.075)
(0.090)
0.428
0.309
0.566
0.619
0.310
0.385
0.745
0.722
0.714
(0.334)
(0.560)
(0.191)
(0.154)
(0.557)
(0.400)
(0.086)
(0.096)
(0.100)
0.613
0.441
0.327
0.459
0.271
0.410
0.660
0.656
0.640
(0.158)
(0.317)
(0.514)
(0.295)
(0.672)
(0.360)
(0.129)
(0.131)
(0.141)
0.335
0.668
0.477
0.617
0.072
0.071
0.194
0.160
0.162
(0.496)
(0.124)
(0.274)
(0.155)
(3.208)
(3.254)
(1.041)
(1.309)
(1.294)
0.184
0.385
0.276
0.348
0.105
0.061
0.170
0.177
0.172
(1.106)
(0.400)
(0.655)
(0.468)
(2.122)
(3.884)
(1.219)
(1.162)
(1.201)
0.180
0.400
0.277
0.370
0.119
0.008
0.215
0.169
0.174
(1.143)
(0.375)
(0.652)
(0.426)
(1.853)
(29.512)
(0.913)
(1.232)
(1.187)
0.383
0.591
0.488
0.541
0.183
0.137
0.165
0.046
0.035
(0.402)
(0.173)
(0.262)
(0.212)
(1.113)
(1.575)
(1.269)
(5.211)
(6.900)
0.342
0.564
0.454
0.507
0.198
0.111
0.140
0.053
-0.013
(0.482)
(0.193)
(0.301)
(0.243)
(1.011)
(1.998)
(1.540)
(4.503)
-(19.116)
0.335
0.580
0.459
0.522
0.205
0.113
0.145
0.019
0.006
(0.496)
(0.181)
(0.295)
(0.229)
(0.967)
(1.970)
(1.480)
(12.703)
(40.073)
Genetic Structure of the Cypress Seed Chalcid 125
based on comparison of individual multilocus genotypes, which measure the genotypic similarity of individuals and consequently is not sensitive to change in allele frequencies. This latter method also gives a better assessment of the exact relationships of individuals belonging to bottlenecked populations.
Discussion The status of the Moroccan population Moroccan individuals collected on C. atlantica differ from other M. wachtli by a few clear morphological characters (Roques & Skrzypczynska in review). Moroccan females differ from M. wachtli females by 1) having the ovipositor sheaths clearly shorter (about as long as the thorax and gaster combined versus longer than the whole body in M. wachtli), 2) different distal end to the dorsal ovipositor valve, and 3) slight differences in colour of the thoracic pubescence and of the body. Males can be separated by the number of teeth on the digitus (Roques pers. comm.). Furthermore, partial sequences (631bp) of the cytochrome b gene (mtDNA) show relatively strong divergence between these two Megastigmus taxa (3.3%) while intraspecific variability is very low (on average 0.49%). Both populations are in reciprocal monophyly for the mtDNA haplotypes which also provide evidence of long term isolation of populations. The high level of differentiation showed by the fixation indices suggests, along with substantial biological, morphological, and mitochondrial and nuclear sequence evidence, that the population from Morocco belongs to different species strictly associated with C. atlantica. Using a rough estimate of rates of mtDNA sequence evolution (Brower 1994), our data would infer a divergence time for these species of 1.4–1.9 Mya. This range may correspond to late Cenozoic glaciation events (Donau), the only one to have periglacial influences in the Mediterranean Basin (Roques et al. 1999a). There are good reasons to regard this estimate with caution as the published rate estimates are based on different portions of the mtDNA genome and do not incorporate potential rate differences among lineages. However, while estimates of the timing of cladogenetic events within the group of western Palaearctic Cupressus (C. sempervirens, C. atlantica and C. dupreziana Camus) had been attributed, without any objective results, to relatively recent glaciation events (i.e. late Weichselian 20 000–12 000 years BP) (Fernandes 1979), our molecular estimation places the split for these species, inferred from cladogenesis of mtDNA, more than two orders of magnitude earlier.
Risk of introgression of the species associated with C. atlantica Introduced species can generate extinction of native fauna by hybridisation and introgression (Rhymer & Simberloff 1996). Whether such hybridisation may lead to lowered fitness of the introgressed genotypes is still uncertain. However, it is known that in some cases, hybridisation induces a decline in adaptation to the local environment and causes outbreeding depression through breakdown of co-adapted gene complexes. Introduction of C. sempervirens, and consequently of M. wachtli, in the native range of the endemic C. atlantica may lead to a threat of extinction of the native Megastigmus associated with C. atlantica. Hybridisation may induce genetic mixing of the two gene pools of wasps and appearance of less fit populations. However, the fact that M. wachtli mates with individuals associated with C. atlantica does not automatically mean that introgression may occur. The hybrids could all be sterile or the introgression could be limited or biased. Nevertheless, it is worth attempting to prevent deliberate contact between the two species of Megastigmus. Studies are urgently needed to evaluate the risk of their genetic
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 126
Figure 3
Hymenoptera: Evolution, Biodiversity and Biological Control
Neighbour-Joining (NJ) phenogram of individual based on Chakraborty and Jin’s (1993) shared allele distance (DAS). The geographic origin of the cypress seed chalcid is given by the first three letters of the name of location and by the shape of the arrows.
Genetic Structure of the Cypress Seed Chalcid 127
mixing and to prevent potential extinction by introgression of the native Megastigmus associated with the endangered C. atlantica.
Native range versus introduced range The native range of C. sempervirens is thought to extend from northern Iran to Crete. Timbal (1975) supposed that Tunisian old growth forests of cypress (e.g. Makthar Valley) also belong to natural populations. Our results are in general agreement with these assumptions but contradict Timbal’s hypothesis. The high level of polymorphism found in the populations of M. wachtli in the eastern Mediterranean Basin suggests that these populations belong to the native range of the wasp. Wasps collected in Crete showed a very high level of variability. They were collected in an area (Lévka Óri Mountains) where the cypress forest is considered to be natural and contains among the oldest trees in Europe (more than 1000 years old) (Rackham 1998). The Megastigmus population from Tunisia exhibits a relatively low level of polymorphism and a few clonal copies of the same genotype. From the genetic data presented here, it is likely that this population came from Crete and originated from an introduction of seed by humans. The relatively higher level of genetic variability compared to other introduced populations (France, Italy) is probably linked to the age of the introduction. The writings of Theophraste in the year 287 BC reports the presence of large cypress forests in Crete. Cypresses were subsequently introduced by the Greeks across the whole Mediterranean Basin, especially in Tunisia where they propagated spontaneously. In 1414, introductions stopped when an edict of the senate from Venice prohibited cypress trade. Megastigmus populations from Tunisia could have been constituted from recurrent introductions during Greek and Roman times and, at most, could be 20 centuries old (#2000 generations), a period long enough to show some genetic variability. Western European populations show a significant trend towards monomorphism at loci having high allelic diversity in the native range. While this may appear from founder events and genetic drift (Stone & Sunnucks 1993), it is more likely due to the apparition of non-sexual reproduction within these populations. Genetic differentiation and dispersal ability The high differentiation observed between the Crete and the other eastern populations strongly suggests that migration events between these areas are rare. We currently lack data to better assess the dispersal capacity of Megastigmus. However, a distance of about 100 km over the sea may constitute a geographical barrier for significant migration events and, hence, gene flow between continental and insular populations. In contrast, populations from the Dodecanese Islands, which are separated by the same geographical distance but are interconnected by continental populations of cypress, show very low levels of genetic differentiation. Relatively strong genetic differentiation is observed between sampled sites separated by 200-300 km. This result is in agreement with the level of genetic differentiation observed in other phytophagous Hymenoptera attacking woody hosts (e.g. Herbst & Heitland 1994; Stone & Sunnucks 1993), and indicates a relatively low level of dispersal ability for a winged insect. Furthermore, the pattern of relationships observed within the eastern Mediterranean populations of M. wachtli also occurs among the cypress populations (Papageorgiou et al. 1994).
Thelytoky The low level of genetic variability found in western Mediterranean populations suggests either a strong bottleneck effect due to local importation of C. sempervirens and/or a change in the reproductive system of M. wachtli. Microsatellite markers show that the western populations are
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 128
Hymenoptera: Evolution, Biodiversity and Biological Control
partly composed of exact copies of the same genotype and, consequently, that thelytoky occurs. These parthenogenetic individuals are not different from the sexual ones in their mtDNA sequence (cytochrome b) and are recently derived from them (Roques et al. 1999a). This is also confirmed by the identity found at potentially informative allozyme loci (G. Roux pers. comm.). Thelytoky enables virgin females to produce diploid females from unfertilised eggs (White 1973). Thelytokous reproduction is frequent within the Hymenoptera, occurring in seven superfamilies and probably more than 2000 species (Luck et al. 1993). Several other species of Megastigmus are known to exhibit thelytokous parthenogenesis (M. aculeatus (Swederus), M. brevicaudis Ratzeburg, M. pictus Foerster (Skrzypczynska 1981), M. pinsapinis Hoffmeyer (Pintureau et al. 1991), M. pistaciae Walker, M. rosae Boucek and M. suspectus Borries). In the Hymenoptera, thelytokous lineages may appear under different circumstances. In several species, extra-chromosomal elements (Wolbachia, Arsenophonus, microsporidia) have been found to be associated with thelytoky. In a few other cases, appearance of such parthenogens seems to be of hybrid origin. Lastly, some species of parasitic wasps reproduce thelytokously in the area where they were introduced for biological control purposes (Aeschlimann 1990). Using Wolbachia-specific primers, we were unable to reveal the presence of the symbiotic bacteria in the populations which show clonal genotypes. Furthermore, the presence in Italy of heterozygous clones excludes the possibility of gamete duplication at the first mitotic division and, consequently, the intervention of the Rickettsia-like proteobacteria. Hybrid origin can also be excluded as an explanation for the appearance of thelytokous forms within introduced populations. Thelytokous strains could be selectively advantaged under several conditions. For example, low population density of the host-plant and/or of Megastigmus may advantage asexual strains which do not need to find mates. This could be the case when the host-plant is either associated with disturbed non-climax habitats or scattered in small groups. Furthermore, genetic homogeneity of the host-plant, with associated low levels of spatial and temporal variability, may also favour selection of asexual populations of phytophagous insects. Populations of evergreen cypress which occur in western Mediterranean countries mostly grow in patches within disturbed habitats. Moreover, most individuals derive from few cultivars with low levels of genetic variability (Papageorgiou et al. 1994). These conditions, associated with loss of genetic variability of M. wachtli due to their introduction by humans, may have favoured selection of thelytokous forms in the areas where cypress trees have been introduced. The evolution of obligate parthenogens often involves distinct selective bottlenecks, resulting in highly coadapted genomes with strong epistatic interactions throughout the genome (Templeton 1979).
Acknowledgements The study was supported by an AIP (Action Incitative Programmée): ‘Structuration génétique des populations naturelles’ from BRG and INRA. We would like to thank Isabelle Le Clainche, Gwenaëlle Mondor and Jean-Paul Raimbault for technical assistance.
References Aeschlimann, J. P. (1990) Simultaneous occurrence of thelytoky and bisexuality in Hymenopteran species, and its implications for the biological control of pests. Entomophaga 35: 3-5. Baumann, H. (1982) Die grieschische Pflanzenwelt in Mythos, Kunst und Literatur. Hirmer, München.
Genetic Structure of the Cypress Seed Chalcid 129
Ben Jamaa, M. L. & Roques, A. (1997) Survey of impact on seed cones of two species of Cupressaceae, Cupressus sempervirens L. and Tetraclinis articulata Mast. in Tunisia. Proceedings of the 6th Arabian Congress for Vegetal Protection, Beyrouth. Brower, A. V. (1994) Rapid morphological radiation and convergence among races of the butterfly Heliconius erato inferred from patterns of mitochondrial DNA evolution. Proceedings of the National Academy of Sciences, USA 91: 6491-6495. Canakcioglu, H. (1959) Studies on insects which are injurious to the Turkish forest tree seeds and control of some of the important enemies. Oman Fakültesi Dergisi, Series A 9: 126-165. Carcreff, E., Rasplus, J. Y., Roques, A., Mondor, G., Vautrin, D. & Solignac, M. (1998) Isolation and characterization of microsatellite loci in the seed chalcid Megastigmus wachtli (Hymenoptera). Molecular Ecology 7: 251-253. Cavalli-Sforza, L. L. & Edwards, A. W. F. (1967) Phylogenetics analysis: models and estimation procedures. American Journal of Human Genetics 19: 233-257. Chakraborty, R. & Jin, L. (1993) A unified approach to study hypervariable polymorphisms: Statistical considerations of determining relatedness and population distances. pp. 153-175. In Pena, S. D. I., Chakraborty, R., Epplen, J. T. & Jeffreys, A. J. (Eds), DNA Fingerprinting: State of the Science. Birkauser Verlag, Basel. Ewens, W. J. (1972) The sampling theory of selectively neutral alleles. Theoretical Population Biology 3: 87-112. Fernandes, F. (1979) Les cyprès Africains. pp. 45-47. In C. E. E. (Ed.). Il Cipresso: Malatti e Difesa. Agrimed, Florence. Goodman, S. J. (1997) RSTCALC: collection of computer programs for calculating unbiased estimates of genetic differentiation and gene flow from microsatellite data and determining their significance. Molecular Ecology 6: 881-885. Goudet, J. (1995) FSTAT version 1.2 : a computer program to calculate F-statistics. Journal of Heredity 86: 485-486. Guido, M., Battisti, A. & Roques, A. (1995) A contribution to the study of cone and seed pests of the evergreen cypress (Cupressus sempervirens L.) in Italy. Redia 78: 211-227. Herbst, J. & Heitland, W. (1994) Genetic differentiation among populations of the sawfly species Platycampus luridiventris associated with different alder species (Hymenoptera: Tentredinidae). Entomologia Generalis 19: 39-48. Luck, R. F., Stouthamer, R. & Nunney, L. (1993) Sex determination and sex ratio patterns in parasitic Hymenoptera. pp. 442-476. In Wrench, D. L. & Ebbert, M. A. (Eds), Evolution and Diversity of Sex-Ratio in Haplodiploid Insects and Mites. Chapman and Hall, London. Nei, M. (1987) Molecular Evolutionary Genetics. Columbia University Press, New-York. Pamilo, P. & Crozier, R. H. (1981) Genetic variation in male haploids under deterministic selection. Genetics 98: 198-214. Papageorgiou, A. C., Panetsos, K. P. & Hattemer, H. H. (1994) Genetic differentiation of natural Mediterranean Cypress (Cupressus sempervirens L.) population in Greece. Forest Genetic 1: 1-12. Pintureau, B., Fabre, J. P. & Oliviera, M. L. (1991) Etude de deux formes de Megastigmus suspectus Borries. (Hym. Torymidae). Bulletin de la Société Entomologique de France 95: 277-290. Rackham, O. (1998) Implications of historical ecology for conservation. pp. 152-175. In Sutherland, W. J. (Ed.), Conservation Science and Action. Blackwell, Abington. Raymond, M. & Rousset, F. (1995) GENEPOP (version 1.2): Population genetics software for exact tests and ecumenism. Journal of Heredity 86: 248-249.
J. Y. Rasplus, E. Carcreff, J. M. Cornuet and A. Roques 130
Hymenoptera: Evolution, Biodiversity and Biological Control
Rhymer, J. M. & Simberloff, D. (1996) Extinction by hybridization and introgression. Annual Review of Ecology & Systematics 27: 83-109. Roques, A., Carcreff, E. & Rasplus, J. Y. (1999a) Cupressus sempervirens vs. cypress seed chalcid, Megastigmus wachtli: genetic and evolutionary relationships. pp. 65-77. In Lieutier, F., Mattson, W. J. & Wagner, M. R. (Eds), Physiology and Genetics of Tree Phytophage Interactions, INRA Editions, Versailles. Roques, A., Markalas, S., Roux, G., Pan, Y., Sun, J. H. & Rimbault, J. P. (1999b) Impact of insect damaging seed cones of cypress (Cupressus sempervirens) in natural stands and plantation of south eastern Europe. Annals of Forest Science 56: 167-177. Roques, A. & Raimbault, J. P. (1986) Cycle biologique et répartition de Megastigmus wachtli (Seitn.) (Hym., Torymidae), chalcidiens ravageurs des graines de cyprès dans le bassin méditerranéen. Zeitschrift für Angewandte Entomologie 101: 370-381. Roques, A., Roux, G. & Markalas, S. (1997) Entomofauna of seed cones evergreen cypress, Cupressus sempervirens, in natural strands of the eastern Aegean region. pp. 175-191. Proceedings of the 5th International Conference on Cone and Seed Insects IUFRO Working Party (WP S7.03-01), Sept. 96, Monte Bondone, Italy. Roques, A. & Skrzypczynska, M. (in review) Seed-infesting chalcids of the genus Megastigmus Dalman (Hymenoptera: Torymidae) native and introduced to Europe: taxonomy, host specificity and distribution. Journal of Natural History. Roux, G. & Roques, A. (1996) Biochemical genetic differentiation among seed chalcid of genus Megastigmus (Hymenoptera : Torymidae). Experientia 52: 522-530. Saitou, N. & Nei, M. (1987) The neighbor-joining method: A new method for reconstructing phylogenetic trees. Molecular Biology & Evolution 4: 406-425. Skrzypczynska, M. (1981) Males of Megastigmus pictus (Förster) in Poland (Hymenoptera, Torymidae). Polskie Pismo Entomologiczne 51: 207-208. Slatkin, M. (1995) A measure of population subdivision based on microsatellite allele frequencies. Genetics 139: 457-562. Stone, G. N. & Sunnucks, P. (1993) Genetic consequences of an invasion through a patchy environment – the cynipid gallwasp Andricus quercuscalicis (Hymenoptera: Cynipidae). Molecular Ecology 2: 251-268. Templeton, A. R. (1979) The unit of selection in Drosophila mercatorum. II. Genetic revolution and the origin of coadapted genomes in parthenogenetic strains. Genetics 92: 1265-1282. Timbal, J. (1975) Chorologie des espèces ligneuses. Tome 2: Essences de reboisement dans la zone méditerranéenne. INRA-CNRF, Nancy. Vidacovic, M. (1991) Conifers: Morphology and Variation. Graficki zavod Hvrastke, Zagreb. Weir, B. S. & Cockerham, C. C. (1984) Estimating F-statistic for the analysis of population structure. Evolution 38: 1358-1370. White, M. J. D. (1973) Animal Cytology and Evolution. Cambridge University Press, Cambridge. Wright, S. (1951) The genetical structure of populations. Annals of Eugenics 15: 323-354.
Systematics of the Ant Genus Camponotus (Hymenoptera: Formicidae): a Preliminary Analysis Using Data from the Mitochondrial Gene Cytochrome Oxidase I Seán G. Brady, Jürgen Gadau* and Philip S. Ward Department of Entomology and Center for Population Biology, University of California, Davis CA 95616 USA (e-mail:
[email protected]) (*present address: Universität Wuerzburg-Biozentrum, Zoologie II, Am Hubland 97074 Wuerzburg, Germany)
Introduction Camponotus Mayr is the largest genus currently recognised in the family Formicidae (ants). The 931 described species in this genus represent 10% of all known ant species. Camponotus displays a cosmopolitan distribution with substantial numbers of species occurring in every major biogeographical region (Bolton 1995b). Many Camponotus species have featured prominently in ecological, evolutionary and behavioural research (Hölldobler & Wilson 1990). In spite of this predominance, the higher systematics of Camponotus remains poorly understood. The last attempt at a genus-wide systematic treatment was made by Emery (1925), who recognised 38 subgenera. This total has risen to 46 currently valid subgenera (Bolton 1995a). Although this scheme certainly helps to manage such a diverse group, the question remains open as to how well these subgeneric divisions reflect the underlying evolutionary history of the genus. Many of the subgenera are defined by characters that may be prone to homoplasy (e.g. pilosity and head shape) and may not be indicative of monophyly. One particular subgenus, Colobopsis Mayr, is characterised by having the anterior portion of the head sharply truncated in the soldier and queen castes. These plug-shaped heads serve a phragmotic function, i.e. they are used to block the entrances to the arboreal nests inhabited by these ants (e.g. Szabó-Patay 1928). Phragmotic heads have evolved independently in other ant genera (Hölldobler & Wilson 1990). Other members of Camponotus (e.g. subgenera Hypercolobopsis Emery and Pseudocolobopsis Emery) also show a morphological trend toward phragmosis, in the form of partial truncation of the anterior end of the head. Thus, we would like to know if the extreme truncation of the head of Colobopsis represents a single origin of this character, or whether Colobopsis itself is polyphyletic. The monophyly of the genus as a whole may also be viewed in doubt. Camponotus appears to be closely related to two other genera: Polyrhachis F. Smith (477 described species) and Dendromyrmex Emery (7 spp.). All species examined so far from these three genera, with only two exceptions, lack the metapleural gland in workers (Hölldobler & Engel-Siegel 1984). This structure is found in nearly all other ant species, and is considered one of the strongest synapomorphies for Formicidae. The loss or extreme reduction of this gland in these three genera suggests monophyletic status for this group as a whole. Within this putative clade, however, Camponotus may
131
Seán G. Brady, Jürgen Gadau and Philip S. Ward 132
Hymenoptera: Evolution, Biodiversity and Biological Control
be a paraphyletic assemblage from which Polyrhachis and Dendromyrmex arose. Camponotus is distinguished from these other genera by three traits: 1) worker polymorphism; 2) lack of spines or teeth on the pronotum, propodeum or petiole; and 3) a relatively short first gastral (=abdominal segment III) tergite (Bolton 1994). These characteristics are found in numerous other ant taxa and may not be particularly indicative of Camponotus monophyly. Specifically, taxonomists have considered the possibility that Polyrhachis has arisen from within one or more Camponotus lineages (Dorow 1995). Thus, although a close relationship between the three genera seems wellsupported, the monophyly of Camponotus itself may be questionable. The purpose of our research is to address the following questions: 1) is Camponotus monophyletic? 2) are the subgenera of Camponotus monophyletic? and 3) was the evolution of a sharply truncate, phragmotic head a unique event in this group? In the current paper we report our initial data set gathered from mitochondrial DNA sequences, with which we begin to examine these issues. N.B. The author for genera, subgenera and species are given in Table 1.
Materials and Methods We sequenced a portion of the mitochondrial gene cytochrome oxidase I (COI) from 43 individuals representing 32 Camponotus species, six Polyrhachis species, one Formica species, one Dendromyrmex species, and one Oecophylla species (Table 1). All of these genera except Oecophylla are members of the Formica-genus group (Agosti 1991), so Oecophylla was used as the outgroup for all phylogenetic analyses. Our sampling involved representatives from 10 Camponotus subgenera, including the monotypic subgenera Dinomyrmex and Myrmepomis. We sequenced samples from two separate colonies for two species (C. atriceps and C. rufipes) and treated them as individual taxa. Voucher specimens have been deposited in the Bohart Museum of Entomology, University of California, Davis. Sequences for F. fusca and F. truncorum were obtained from GenBank (accession nos. AB0101925, AB0101929). Genomic DNA was isolated from ethanol preserved adult worker individuals using either a CTAB-phenol extraction method (Hunt & Page 1994) or a DNA extraction kit (ID Pure). A fragment of 385 base pairs (excluding primers) from COI was amplified under the following conditions: 0.5 U Taq DNA polymerase (GIBCO/BRL); standard 10X buffer (GIBCO/BRL); final concentration of MgCl2 2 mM. PCR amplification consisted of 35 cycles of 1 min at 95˚C, 1 min at 47˚C, and 1 min 30 sec at 72°C. A single set of primers was used for all amplifications. ‘Jerry’ (sense strand) is a universal insect primer (Simon et al. 1994) which corresponds to positions 2161-2183 of COI (numbered according to the Drosophila yakuba (Burla) sequence in Clary & Wolstenholme 1985). ‘Ben3R’ (anti-sense strand) was designed by T. R. Schultz (Smithsonian Institution) and corresponds to positions 2568-2591. Primer sequences are as follows: Jerry (5' – CAA CAT TTA TTT TGA TTT TTT GG – 3'); Ben3R (5'-GCW ACW ACR TAA TAK GTA TCA TG-3'). PCR product was purified using microcon 100 microconcentrators (Amicon), followed by automated sequencing of both strands on an ABI Prism 377 DNA Sequencer (Perkin Elmer) using the same primers as above. Sequences will be deposited in GenBank. The resulting sequences were unambiguously aligned by eye, with no resulting gaps. All sequence variation and phylogenetic analyses were performed using the computer program PAUP* 4.0b1 (Swofford 1998). Maximum parsimony (MP) trees were inferred using TBR branch swapping from 100 random addition replicates. The nonparametric bootstrap (Felsenstein 1985) with 1000 replicates was used to assess support of individual clades. The use of log-determinant distances can help compensate for differences among taxa in base composition bias (Lockhart et al.
Systematics of the Ant Genus Camponotus (Hymenoptera: Formicidae) 133
1994). For this reason, we also calculated bootstrap neighbour-joining (NJ) trees using log-determinant distances. All findings reported herein were robust to both tree-building methods.
Results And Discussion Sequence variation We obtained COI sequences totalling 385 base pairs from 45 ant specimens. Of these sites, 154 were variable and 144 informative for parsimony. The absolute number of variable sites at the first (nt1), second (nt2), and third (nt3) codon positions was 3, 29, and 122, respectively. Because many uncorrected pairwise distance values at nt3 positions were quite high (e.g. a mean of 0.36 between the outgroup and ingroup species), the historical signal at these sites may be obscured due to saturation (see below). However, the skewness statistic (g1: Hillis & Huelsenbeck 1992) calculated using 104 random trees indicated significant phylogenetic signal when only nt3 positions were included in the test (g1 = -0.45, P < 0.05). The entire data set (nt1 + nt2 + nt3) also displayed significant structure (g1 = -0.48, P < 0.05). Overall base composition was biased in favour of A+T (69.6%). Base composition bias was more pronounced in nt3 positions (83.8%) compared to nt1+nt2 (62.5%). Between-species comparisons of the homogeneity of base frequencies showed that composition did not vary at nt1 or nt2 (χ2, P >> 0.05), but differed significantly at nt3 (χ2, P << 0.05). Strong AT-bias and site-to-site variation have been demonstrated to be generally prevalent in the hymenopteran mitochondrial genome (Dowton & Austin 1997; Whitfield & Cameron 1998). Several phylogenies of other ant groups (Chenuil & McKey 1996; Wetterer et al. 1998) have used different portions of mitochondrial cytochrome oxidase genes and have reported patterns of sequence variation similar to our study. One interesting difference, however, is that AT-bias at nt1+nt2 positions was considerably higher in these studies (Chenuil & McKey 1996: 77%-79%; Wetterer et al. 1998: 74%) compared to the present study (62.5%). It is unclear whether this contrast was due to the different taxa involved, or to the fact that different fragments of these genes were employed.
Phylogenetic analyses Unweighted parsimony analysis resulted in 4 most parsimonious trees (tree length = 1107; consistency index excluding uninformative characters = 0.22; retention index = 0.44). The strict consensus of these trees (Fig. 1) showed modest support for a clade consisting of Camponotus + Polyrhachis + Dendromyrmex (68% bootstrap support) and strong support for the monophyly of Formica (99%). The data also point to monophyly among groups of Camponotus species in the same subgenus from the same region, e.g. Neotropical Myrmothrix (C. atriceps, C. floridanus, C. rufipes – 90% bootstrap support), Holartic Camponotus (s.str.) excluding C. quercicola (C. herculeanus, C. laevigatus, C. ligniperdus, C. modoc, C. pennsylvanicus – 73%), and two South American species of Tanaemyrmex (C. balzani, C. silvicola – 78%). Yet three Nearctic species of Myrmentoma became disassociated: one species (C. clarithorax) clustered with Camponotus (Myrmaphaenus) yogi (69% bootstrap support) while the other two species examined (C. essigi, C. hyatti) were sister taxa (100%) in another part of the tree. The genus Polyrhachis and the camponotine subgenera Colobopsis, Myrmobrachys, and Tanaemyrmex each appeared polyphyletic, although this was not well supported by bootstrap results. With regard to Colobopsis, at the present time we cannot reliably assess whether the evolution of phragmotic heads in this subgenus represents a unique event or parallel evolution.
Seán G. Brady, Jürgen Gadau and Philip S. Ward 134
Table 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Collections of specimens used in this study. Two different specimens were used from C. atriceps and C. rufipes.
Species
Subgenus
Locality
Collector
Genus Camponotus Mayr C. herculeanus (L.)
Camponotus Mayr
Montana, USA
Ward #12116
C. laevigatus (F. Smith)
Camponotus
California, USA
Ward #12669
C. ligniperdus (Latreille)
Camponotus
Bavaria, GERMANY
Gadau
C. modoc W.M. Wheeler
Camponotus
California, USA
Brady #203
C. pennsylvanicus (De Geer)
Camponotus
eastern USA
G. Gräf
C. quercicola M.R. Smith
Camponotus
California, USA
Brady #328
C. sp. nr. gasseri (Forel)
Colobopsis Mayr
ACT, AUSTRALIA
Brady #437
C. melanocephalus Roger
Colobopsis
NSW, AUSTRALIA
Ward #9763
C. impressus (Roger)
Colobopsis
Florida, USA
M. Deyrup & Z. Prusak
C. papago Creighton
Colobopsis
Sonora, MEXICO
Ward #13458
C. sp. nr. saundersi Emery
Colobopsis
MALAYSIA
U. Maschwitz
C. gigas (Latreille)
Dinomyrmex Ashmead
MALAYSIA
U. Maschwitz
C. yogi W.M. Wheeler
Myrmaphaenus Emery
California, USA
Ward #13535
C. clarithorax Creighton
Myrmentoma Forel
California, USA
Ward #13291
C. essigi M.R. Smith
Myrmentoma
California, USA
Brady #159
C. hyatti Emery
Myrmentoma
California, USA
Ward #12534
C. sericeiventris (Guérin-Méneville)
Myrmepomis Forel
Misiones, ARGENTINA
F. Roces
C. mus Roger
Myrmobrachys Forel
Buenos Aires, ARGENTINA
F. Roces
C. planatus Roger
Myrmobrachys
Florida, USA
S. Cover
C. cf. suffusus (F. Smith)
Myrmosaulus W.M. Wheeler
ACT, AUSTRALIA
Brady #441
C. atriceps (F. Smith) 1
Myrmothrix Forel
TRINIDAD
O. Rüppel
C. atriceps (F. Smith) 2
Myrmothrix
MEXICO
O. Rüppel
C. floridanus (Buckley)
Myrmothrix
Florida, USA
Gadau
C. rufipes (F.) 1
Myrmothrix
Rio de Janeiro, BRAZIL
F. Roces
C. rufipes (F.) 2
Myrmothrix
El Bagual, ARGENTINA
F. Roces
C. balzani Emery
Tanaemyrmex Ashmead
Cusco, PERU
S. Cover
Systematics of the Ant Genus Camponotus (Hymenoptera: Formicidae) 135
C. castaneus (Latreille)
Tanaemyrmex
Florida, USA
Gadau & Schilder
C. consobrinus (Erichson)
Tanaemyrmex
Vic., AUSTRALIA
R. Crozier
C. semitestaceus Snelling
Tanaemyrmex
California, USA
Brady #325
C. silvicola Forel
Tanaemyrmex
Cusco, PERU
S. Cover
C. socius Roger
Tanaemyrmex
Florida, USA
S. Cover
C. vicinus Mayr
Tanaemyrmex
California, USA
Brady #171
C. sp. nr. vicinus
Tanaemyrmex
California, USA
Ward #13044
C. sp. nov.
unknown
Baja California, MEXICO
Ward #13592
—
Santa Cruz, BOLIVIA
Ward #12168
Genus Dendromyrmex Emery D. cf. traili (Mayr) Genus Formica L. F. fusca L.
—
unknown
GenBank no. AB0101925
F. truncorum F.
—
unknown
GenBank no. AB0101929
F. moki W.M. Wheeler
—
California, USA
Brady #168
—
Centre-Sud, CAMEROON
D. M. Olson
Genus Oecophylla F. Smith O. longinoda (Latreille) Genus Polyrhachis F. Smith P. sp. nr. inconspicua Emery
Campomyrma W.M. Wheeler NT, AUSTRALIA
B. B. Lowery
P. flavibasis Clark
Campomyrma
Ward #9857
Qld, AUSTRALIA
P. sp. nr. fuscipes Mayr
Campomyrma
SA, AUSTRALIA
Brady #432
P. hostilis gp. sp.
Chariomyrma Forel
East Sepik, PNG
Ward #10202
P. sp. nr. crawleyi Forel
Hagiomyrma W.M. Wheeler
WA, AUSTRALIA
B. B. Lowery
P. dives F. Smith
Myrmhopla Forel
East Sepik, PNG
Ward #10175
Seán G. Brady, Jürgen Gadau and Philip S. Ward 136
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Strict consensus of four most parsimonious trees for species belonging to Camponotus, Dendromyrmex, Formica and Polyrhachis, with Oecophylla longinoda as the outgroup. Numbers above branches are bootstrap percentage values for clades found in both the most parsimonious and bootstrap majority rule trees.
The maximum parsimony (MP) analysis also suggested that the genus Camponotus was not monophyletic, but again bootstrap support for this result was poor (Fig. 1). Constraining for the monophyly of Camponotus required a minimum of 15 extra steps relative to the MP trees. However, these constraint trees were not significantly different from the MP trees by either the Templeton (1983) Wilcoxon signed-ranks test (0.06 ≤ P ≥ 0.29) nor the Kishino-Hasegawa (1989) parametric test (0.06 ≤ P ≥ 0.21). So given the current data, we hesitate to reject the hypothesis that Camponotus is a monophyletic group. Because the nt3 position is biased in base composition and contributes the majority of characters informative for parsimony, these data may lead to misleading results under MP (Lockhart et al. 1994; Eyre-Walker 1998). However, all findings reported above from the MP tree also occurred
Systematics of the Ant Genus Camponotus (Hymenoptera: Formicidae) 137
60
Observed transitions
50
40
30
20
10
0 0
5
10
15
20
25
30
35
Observed transversions
Figure 2
Observed number of transitions plotted as a function of the observed number of transversions from representative pairwise comparisons within the ingroup.
with similar bootstrap support in the log-determinant neighbour-joining (NJ) tree (not shown), a method which can compensate for such biases (Lockhart et al. 1994). The low bootstrap values over much of the MP tree (Fig. 1) may be due to several sources. Because there are 144 parsimony-informative characters distributed over 45 taxa, some bootstrap replicates may not retrieve clades due to sampling error. Given even a complete lack of homoplasy, three synapomorphies are required to retrieve a clade with 95% bootstrap support (Felsenstein 1985). This is more problematic for our data set, whose consistency index of 0.22 indicates a substantial amount of homoplasy, perhaps largely due to saturation. A scatter plot of uncorrected pairwise transitions versus transversions (Fig. 2) indicates saturation over portions of our data set, which may contribute to the lack of resolution in parts of our analysis. In summary, our phylogenetic analyses revealed that some Camponotus subgenera (e.g. Myrmothrix) accurately reflect monophyletic groupings, at least with respect to the species sampled in this study. Other subgenera, however, may not be monophyletic, but in most cases (e.g. Colobopsis) bootstrap support for these conclusions was lacking. We found reasonably strong support for the monophyly of Camponotus + Polyrhachis + Dendromyrmex . There was an indication from the MP and log-determinant NJ trees that Camponotus itself may not be monophyletic, but this result lacked statistical support, so a more reliable assessment of this issue must await the accumulation of further data.
Acknowledgements We thank Mark Deyrup, Christina Elsishans, Uli Raub, and Steve Shattuck for providing some of the specimens used in this study, Ted Schultz for PCR primers, and Robert Page and Michael
Seán G. Brady, Jürgen Gadau and Philip S. Ward 138
Hymenoptera: Evolution, Biodiversity and Biological Control
Sanderson for laboratory facilities. This work was supported by a Mildred E. Mathias Graduate Student Research Grant and a Center for Population Biology Graduate Research Award to S.G.B., a Feodor-Lynen Forschungsstipendium to J.G., and National Science Foundation funding to P.S.W.
References Agosti, D. (1991) Revision of the oriental ant genus Cladomyrma, with an outline of the higher classification of the Formicinae (Hymenoptera: Formicidae). Systematic Entomology 16: 293-310. Bolton, B. (1994) Identification Guide to the Ant Genera of the World. Harvard University Press, Cambridge, Massachusetts. Bolton, B. (1995a) A New General Catalogue of the Ants of the World. Harvard University Press, Cambridge, Massachusetts. Bolton, B. (1995b) A taxonomic and zoogeographical census of the extant ant taxa (Hymenoptera: Formicidae). Journal of Natural History 29: 1037-1056. Chenuil, A. & McKey, D. B. (1996) Molecular phylogenetic study of a myrmecophyte symbiosis: did Leonardoxa/ant associations diversify via cospeciation? Molecular Phylogenetics & Evolution 6: 270-286. Clary, D. O. & Wolstenholme, D. R. (1985) The mitochondrial DNA molecule of Drosophila yakuba: nucleotide sequence, gene organization, and genetic code. Journal of Molecular Evolution 22: 252-271. Dorow, W. H. O. (1995) Revision of the ant genus Polyrhachis Smith, 1857 (Hymenoptera: Formicidae: Formicinae) on a subgenus level with keys, checklist of species and bibliography. Courier Forschungsinstitut Senckenberg 185: 1-113. Dowton, M. & Austin, A. D. (1997) Evidence for AT-transversion bias in wasp (Hymenoptera: Symphyta) mitochondrial genes and its implications for the origin of parasitism. Journal of Molecular Evolution 44: 398-405. Emery, C. (1925) Hymenoptera. Fam. Formicidae. Subfam. Formicinae. Genera Insectorum 183: 1-302. Eyre-Walker, A. (1998) Problems with parsimony in sequences of biased base composition. Journal of Molecular Evolution 47: 686-690. Felsenstein, J. (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791. Hillis, D. M. & Huelsenbeck, J. P. (1992) Signal, noise, and reliability in molecular phylogenetic analyses. Journal of Heredity 83: 189-195. Hölldobler, B. & Engel-Siegel, H. (1984) On the metapleural gland of ants. Psyche 91: 201-224. Hölldobler, B. & Wilson, E. O. (1990). The Ants. Harvard University Press, Cambridge, Massachusetts. Hunt, G. J. & Page, R. E. Jr. (1994) Linkage analysis of sex determination in the honey bee (Apis mellifera). Molecular Genetics 244: 512-518. Kishino, H. & Hasegawa, M. (1989) Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170-179. Lockhart, P. J., Steel, M. A., Hendy, M. D., & Penny, D. (1994) Recovering evolutionary trees under a more realistic model of sequence evolution. Molecular Biology & Evolution 11: 605-612. Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H., & Flook, P. (1994) Evolution, weighting, and the phylogenetic utility of mitochondrial gene sequences and a compilation of conserved chain reaction primers. Annals of the Entomological Society of America 87: 651-701.
Systematics of the Ant Genus Camponotus (Hymenoptera: Formicidae) 139
Swofford, D. L. (1998) PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associates, Sunderland, Massachusetts. Szabó-Patay, J. (1928) A kapus-hangya. Természettudományi Kozlony 60: 215-219. Templeton, A. R. (1983) Phylogenetic inference from restriction endonuclease cleavage site maps with particular reference to the humans and apes. Evolution 37: 221-244. Wetterer, J. K., Schultz, T. R., & Meier, R. (1998) Phylogeny of fungus-growing ants (tribe Attini) based on mtDNA sequence and morphology. Molecular Phylogenetics & Evolution 9: 42-47. Whitfield, J. B. & Cameron, S. A. (1998) Hierarchical analysis of variation in the mitochondrial 16S rRNA gene among Hymenoptera. Molecular Biology & Evolution 15: 1728-1743.
This Page Intentionally Left Blank
PART
4
Systematics
This Page Intentionally Left Blank
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? Robert A. Wharton Department of Entomology, Texas A&M University, College Station, Texas 77843 USA (email:
[email protected])
Introduction The family Braconidae is exceptionally diverse. It is the second largest family within the Hymenoptera, and contains over 15 000 described species. A considerable amount of attention has been given to the classification of the Braconidae in recent years, including the production of comprehensive catalogs and regional synopses, and the publication of several treatises on higher order relationships within the family (e. g. Telenga 1955; ¯apek 1965, 1970; Fischer 1971, 1987; Shenefelt 1978; Mason 1981, 1983; van Achterberg 1984; Tobias 1986; Maetô 1987; Quicke & van Achterberg 1990; Shaw & Huddleston 1991; Wharton et al. 1992; van Achterberg & Quicke 1992; Dowton et al. 1998; Belshaw et al. 1998). There has been relatively little agreement to date concerning the subfamily classification, and shifts in rank are commonplace. For example, in the same year, Sharkey (1993) and van Achterberg (1993) recognised 29 and 43 braconid subfamilies, respectively. The works of Nees von Esenbeck (1812, 1814, 1816) represent the first attempt to establish a hierarchical framework for the Braconidae, and the names Bracones and Bassi, proposed by Nees in 1812, are the oldest valid family group names known to me for members of what is now the family Braconidae. The classifications of Nees, Haliday, and Wesmael, though more original and in some ways superior to that of Blanchard (1845) and other early workers, unfortunately commonly employed informal group names (such as Cyclostomi, Areolarii and Cryptogastres) that were not based on generic names and therefore are not available in the meaning of the Code (ICZN 1985). Most of the better-known subfamily and tribal names in use today date from the work of Foerster (1862) who used numerous family group names for the Braconidae. Foerster proposed the following as new: Chelonoidae and Microgasteroidae (validating the species group names used by Nees), Blacoidae, Brachistoidae, Dacnusoidae, Diospiloidae, Doryctoidae, Eumicrodoidae, Euphoroidae, Euspathioidae, Exothecoidae, Hecaboloidae, Helconoidae, Hormioidae, Ichneutoidae, Liophronoidae, Macrocentroidae, Perilitoidae, Rhyssaloidae, and Rogadoidae. All except Liophronoidae, Euspathioidae and Eumicrodoidae are currently in use as either tribes or subfamilies within the Braconidae. We began this century with two classifications derived from Marshall’s (1885) embellishment of Foerster (1862). The first was that of Ashmead (1900), who recognised 17 subfamilies. The second was that of Szépligeti (1904) who, in proposing 31 subfamilies, elevated several of Ashmead’s tribes to subfamily rank and incorporated a few additions of his own. With the possible exception of Telenga’s (1955) work, this pattern of rearrangement and expansion has occurred
143
Robert A. Wharton 144
Hymenoptera: Evolution, Biodiversity and Biological Control
repeatedly, resulting in a classification today that consists of 50 taxa that have been used at the subfamily rank within the last few years (van Achterberg 1993, 1995a, 1995b; Wharton 1997). The present classification is a result of the enormous effort over the last 15–20 years in characterising both the morphological and biological diversity of the Braconidae. I argue here, however, that a classification composed of 50 subfamilies no longer facilitates communication between specialists and non-specialists, nor even among specialists. In recent years, rigorous analyses of various character systems (including many newly discovered ones) have provided clear evidence of relationships among certain subfamilies and tribes. Yet, the current trend towards proliferation of subfamilies provides us with a very limited ability to portray these relationships within the traditional Linnaean hierarchy. If we accept (and try to work within) the limitations of the Linnaean system, we have two equally effective options: make more families or greatly reduce the number of subfamilies. An examination of other large families in Hymenoptera as well as those in other major orders of insects, reveals that both options have been employed. As a result, almost all large families in other insect orders have less than half the number of subfamilies as in the Braconidae, and most have very few. I therefore propose a reduction in number of subfamilies as the more viable means of communicating our ideas about biologically and morphologically meaningful groupings within the Braconidae. Splitting the Braconidae into several families is also feasible (and has been proposed by both Mason and Sharkey unpublished) but is nomenclaturally more radical and thus less likely to be acceptable, especially for the major users of classifications: the non-specialists.
Materials and Methods Specimens representing all taxa recognised as subfamilies in recent years were examined to confirm external morphological features. Most of these taxa are represented in the Insect Collection of the Department of Entomology at Texas A&M University. A few extremely rare taxa were studied at the American Entomological Institute, Gainesville, The British Museum (Natural History), London, The Canadian National Collection, Ottawa, and the Nationaal Natuurhistorisch Museum, Leiden. Representative taxa were also dissected to verify internal morphological features associated with the male and female reproductive tracts. Terminology for wing venation follows Sharkey and Wharton (1997). In order to trace the origin of certain wing vein features in the Braconidae, it has been necessary to use Xiphidriidae for comparison. More immediate outgroups such as Ichneumonidae and Orussidae exhibit varying degrees of independent reduction, and are thus less useful for understanding braconid venation.
Proposed Classification There is growing evidence of an extreme basal split within the Braconidae, separating at least two major clades (Wharton et al. 1992; Wharton 1997). These two clades have often been referred to as the cyclostomes and non-cyclostomes, and these informal categories will be used here to divide the family into two, roughly equal, parts. Although there has not been unanimous agreement in the past regarding the monophyly of these two clades (see cladograms in Quicke & van Achterberg 1990 and Dowton et al. 1998), support for their monophyly comes from several sources, and is briefly discussed in the following three paragraphs. The term cyclostome refers to an indentation of the labrum and lower portion of the clypeus, producing a concavity in the face between the upper, outer portion of the clypeus and the dorsal margin of the mandibles. The groundplan state of a distinctly concave labrum and clypeus
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? 145
defines the cyclostomes as monophyletic, although some of the more derived cyclostomes, such as the alysiines, have lost this feature in association with other modifications to the lower region of the face. A combination of internal characters associated with the male and female reproductive systems has been used to further differentiate the cyclostomes from the non-cyclostomes (Edson & Vinson 1979; Maetô 1987; Quicke & van Achterberg 1990; Wharton et al. 1992). The morphology of the poison glands and the attachment of the vas deferens in particular provide support for the monophyly of these two groups. The hind wing venation has also developed along different lines to some extent. A well-developed hind wing m-cu (= posternervellus) is present in many cyclostomes but is virtually absent in non-cyclostomes. The origin of this vein from hind wing m-cu + 3CU is clearly seen in doryctines such as Zombrus Marshall. Recently, Belshaw et al. (1998) also provided molecular evidence for monophyly of the cyclostomes. Biologically, all non-cyclostomes are endoparasitoids, but about half of the cyclostomes are also endoparasitoids. All cases of endoparasitism within the cyclostome lineage are convincingly derived from ectoparasitoids with typical cyclostome facial features (Whitfield 1992; Quicke 1993). Endoparasitism is thus a derived condition within the cyclostomes, with ectoparasitism the putatively primitive condition. Non-cyclostomes can therefore be characterised by endoparasitism that evolved separately from a different common ancestor. This separate derivation of endoparasitism is a putative synapomorphy for the non-cyclostomes. These two groups, the cyclostomes and non-cyclostomes, are thus each defined as monophyletic, and are so considered here. A 10-fold reduction of the bewildering array of subfamilies considerably facilitates recognition of this basal split, and can be supported on biological and morphological grounds. I therefore propose two cyclostome subfamilies, the Braconinae and Alysiinae, and two non-cyclostome subfamilies, the Agathidinae and Microgastrinae. A fifth subfamily, the Aphidiinae, is tentatively placed with the cyclostomes, though I admit to some uncertainty regarding its placement (see Table 1). This places our classification in line with explicit hypotheses for the evolution of idiobiont and koinobiont strategies within the Braconidae. Specifically, it treats koinobiont endoparasitism in the newly defined Agathidinae + Microgastrinae as a separate evolutionary event, independent of and consequently not derived from evolution of a similar lifestyle within the newly defined Braconinae + Alysiinae. I further hypothesise that the transition from ectoparasitism to endoparasitism may have occurred in two basically different ways in these groups (although supporting data is limited): through gradual internalisation of the egg in older instars in the Braconinae + Alysiinae (e.g. Shaw 1983), and initially through avoidance of the host immune system in the Agathidinae + Microgastrinae by ovipositing in host eggs or internal organs.
Braconinae + Alysiinae: the Cyclostomes The Braconinae and Alysiinae represent the two branches of what I believe to be a basal split within the cyclostomes. I have included a suite of former subfamilies within the Alysiinae, as this reflects our current uncertainty regarding the relationships among the various groups that are either intermediate between or basal to the more typical doryctines and rogadines. Alysiinae is the oldest available family group name for these taxa. I also admit to uncertainty regarding the placement of the previously recognised subfamily Telengaiinae, and consequently I leave it as an unplaced tribe within the cyclostomes.
Robert A. Wharton 146
Table 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Proposed reclassification of Braconidae.
Cyclostomes Braconinae Alysiinae Alysiini (includes Opiinae) Apozygini Doryctini (includes Ypsistocerinae and Mesostoinae) Exothecini Gnamptodontini Histeromerini Hormiini (includes Lysiterminae) Pambolini Rhysipolini Rhyssalini Rogadini (includes Betylobraconinae) Aphidiinae Unplaced Telegaiini Non-cyclostomes Microgastrinae Chelonidii Adeliini Chelonini Dirrhopidii Ecnomiidii Mendesellidii Microgastridii Cardiochilini Khoikhoiini Miracini Microgastrini Agathidinae Agathidini (includes Sigalphinae and Pselaphanus) Blacini Cenocoeliini Euphorini (includes Meteorinae) Helconini Homolobini Ichneutini Macrocentrini (includes Amicrocentrinae) Meteorideini Neoneurini Orgilini Trachypetini Xiphozelini Unplaced Masoninae
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? 147
Subfamily Braconinae I would tentatively place vaepellines in with the more typical braconines (in part because of the structure of the petiole), but otherwise keep the Braconinae unchanged from its more traditional characterisation (e.g. van Achterberg 1984; Quicke & van Achterberg 1990) as it is well defined morphologically. As a whole, this subfamily is composed of idiobiont ectoparasitoids with one small group having become endoparasitoids of lepidopterous pupae. Members of some genera, and even some species, readily cross ordinal lines and are thus broadly polyphagous. Several groups of genera appear to form well supported monophyletic units that should be formally recognised as tribes. Some such taxa are completely or nearly completely host specific at the ordinal level, forming suites of either lepidopterous or coleopterous parasitoids (e. g. groups of genera containing Vipio L. and Atanycolus Foerster, respectively). Subfamily Alysiinae The subfamily Alysiinae is monophyletic relative to the Braconinae if we hypothesise that all members are derived from typical doryctines which primitively used their fore tibial spines to assist them in escaping from tunnels in wood, then gradually lost them as they radiated onto different hosts in different habitats. This group includes at least 14 taxa that have been treated as subfamilies over the last few years, including the Doryctinae, Apozyginae, Mesostoinae, Ypsistocerinae, Rogadinae, Betylobraconinae, Rhysipolinae, Exothecinae, Hormiinae, Lysiterminae, Pambolinae, Alysiinae, Opiinae, and Gnamptodontinae. Recent work suggests that rhyssalines and/or histeromerines may be basal members of the cyclostome assemblage. Although I currently include them as tribes within the Alysiinae pending further analysis, the classification proposed here could just as easily accommodate them as separate subfamilies. Support for separate recognition of the rhyssalines comes from retention of the second anal cross-vein of the fore wing in several of its members, a feature lost in almost all other cyclostomes. This is a relatively weak character, however, requiring a single loss in the Alysiinae + Braconinae and an independent series of multiple losses in what would be the Rhyssalinae, if we were to recognise Rhyssalinae and place it as basal to Alysiinae + Braconinae. An alternative hypothesis to that taken here is that the Braconinae, even though well characterised, are derived from within the proposed Alysiinae, making the latter paraphyletic (see molecular studies of Dowton et al. 1998 and Belshaw et al. 1998). The following tribes are proposed for the Alysiinae: Apozygini, Doryctini (including the former Mesostoinae and Ypsistocerinae), Rogadini (including the former Betylobraconinae), Rhysipolini, Hormiini (including the former Lysitermini), Exothecini, Pambolini, Alysiini (including the former Opiinae), Gnamptodontini, Rhyssalini, and Histeromerini. All have been characterised elsewhere, although often inadequately. Apozyx Mason poses an interesting problem since it has typical doryctine features of the cyclostome cavity and fore tibial spines, but retains both fore wing 2r-m and hind wing CUb while apparently losing all traces of hind wing m-cu. It is accommodated within Alysiinae if the fore tibial spines are treated as defining the Alysiinae rather than the Doryctini sensu stricto. I place Mesostoa van Achterberg and related genera in the Doryctini on the basis of ovipositor morphology, though the placement of the metasomal spiracles, the presence of the second anal cross vein in the fore wing, and the absence of hind wing m-cu suggests an alternative placement in the Rhyssalini. I include Betylobracon Tobias in the Rogadini (the latter defined by the biological features of koinobiont endoparasitism and eventual mummification of their lepidopterous hosts) because of similarities between Betylobracon and Yelicones Cameron, including the form of the mandible. Although it has been proposed that such
Robert A. Wharton 148
Hymenoptera: Evolution, Biodiversity and Biological Control
similarities represent homoplasies, the examination of a great deal of new material not heretofore available, particularly of key genera such as Mesocentrus Szépligeti, leads me to propose this synonymy. Van Achterberg (1995a) provides a detailed discussion of characters along with an alternative arrangement. Similarly, I define the tribe Alysiini very strictly in terms of koinobiont endoparasitism of cyclorrhaphous Diptera, and include the former Opiinae therein. Within this redefined subfamily Alysiinae, we find multiple transitions to a koinobiont endoparasitic lifestyle, a few apparent transitions from idiobiont to koinobiont ectoparasitism, and all known instances of braconid phytophagy. This classification does not solve the relationships of the majority of the cyclostomes, and in particular does not provide unequivocal support for the monophyly of most tribes included in the Alysiinae (nor is that the intention). By containing all the relevant, putatively interrelated elements within a single subfamily, however, it does help us to focus our attention on one of the most critical problems in cyclostome classification: that of rooting the rogadine and alysiine clades relative to the doryctines. Improvement in our understanding of the cyclostomes should eventually lead to the break up of the doryctines (in the broad sense as used by Belokobylskij 1992) into separate doryctine, alysiine, and rogadine clades. Once this is accomplished, I envision a relatively minor change in the classification that will give us two to four additional subfamilies. In the meantime, the proposed classification enables us to portray higher-level relationships among cyclostomes without the undue proliferation of subfamilies that precludes illustrating these relationships. It places all potentially related taxa in a single subfamily until their relationships are clarified, and provides available categories for depicting those relationships. Problems remain at a lower rank, and elevation in rank should only occur after resolution of the problems.
Agathidinae + Microgastrinae: the Non-cyclostomes The non-cyclostomes are contained within two major clades, the microgastroids, and a group that is frequently referred to as the helconoids. Using available family group names, these would translate to the two subfamilies: Microgastrinae and Agathidinae. Use of the family group name Agathidinae (vs. Bassinae) is controversial (Simbolotti & van Achterberg 1995), but there is insufficient space to deal with that complex problem here.
Subfamily Microgastrinae Relationships among the members of the Microgastrinae are perhaps better known than any other group of equivalent size in the Braconidae, thanks largely to the works of Nixon, Mason, Whitfield, and others. Whitfield and Mason (1994) discussed 11 subfamilies as possible members of this assemblage, with eight of these appearing consistently as a monophyletic group following their analyses. Using this work as a basis, I recognise five supertribes each of which has been previously defined as monophyletic: the Chelonidii (including adeliines), Microgastridii (including typical microgastrines, cardiochilines, khoikhoiines and miracines), Mendesellidii, Dirrhopidii, and Ecnomiidii. Although Whitfield and Mason (1994) were unable to assign the ecnomiines to their microgastroid assemblage unequivocally, the unequal mid-tibial spurs, fore wing venation, and distinct vannal lobe support their placement here. Retention of the occipital carina and placement of the T1 spiracle on the median tergite suggest that ecnomiines are closer to the Chelonidii than to the Microgastridii. The supertribal classification proposed here reflects the following relatively well-corroborated set of relationships: (Ecnomiidii + ((Chelonidii) + (Dirrhopidii) + (Microgastridii + Mendesellidii))). Recovery of these relationships has been problematic using
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? 149
molecular data (Whitfield 1997; Belshaw et al. 1998; Dowton & Austin 1998; Dowton et al. 1998), in part because some rare taxa have not been available for sequencing. When defined thusly, excluding ichneutines and neoneurines, this is a clade of koinobiont endoparasitoids of Lepidoptera. I therefore predict that taxa with unknown biologies (such as the Ecnomiidii) will have this host relationship. Further clarification of basal lineages and the nature and distribution of polydnaviruses will help better delineate the microgastrines relative to the agathidines, and possibly confirm a separate origin from that of the agathidines. The rank of supertribe is rarely employed, and I have done so here only with reluctance. The Microgastridii, however, is a large, diverse group with clearly defined subunits. Given the size of the Microgastridii as proposed here, it is therefore deemed more suitable to employ the rank of supertribe rather than tribe for these five taxa, providing an additional category to facilitate depiction of relationships within this larger taxon.
Subfamily Agathidinae The Agathidinae are defined by endoparasitism, and treated as a series of tribes of uncertain relationship to one another. The group as a whole is poorly defined, unless endoparasitism of Coleoptera is hypothesised as the groundplan, which has problems relative to recent suggestions of a basal position for the Agathidini. Principle tribes that I recognise are: the Agathidini (including sigalphines and Pselaphanus Szépligeti), Orgilini, Macrocentrini (including amicrocentrines), Helconini, Ichneutini (including proteropines), Blacini, Cenocoeliini, Euphorini (including meteorines), Neoneurini, Homolobini, Meteorideini, Trachypetini, and Xiphozelini. I also support the further break-up of the former Helconinae, since this will facilitate establishment of relationships in this assemblage. These taxa have all been defined elsewhere (e.g. van Achterberg 1984; Quicke & van Achterberg 1990), although usually at a different rank. Within the Agathidinae transitions from coleopterous to lepidopterous hosts have occurred at least twice. The current classification does not reflect this very well, and it is one of the major areas needing attention, similar to the situation with the Alysiinae. As with the Alysiinae, I expect further work will ultimately enable us to split the Agathidinae into 3-4 subfamilies. However, until relationships are clarified and the major groupings within the Agathidinae are thereby revealed, any further recognition of subfamilies is unwarranted. Some molecular data (e.g. Belshaw et al. 1998; Dowton et al. 1998) also suggest that Agathidinae as defined here are not monophyletic if Microgastrinae are excluded, and this alternative should be explored in studies addressing relationships among all non-cyclostomes.
Taxa of Uncertain Placement I am unable to place the Masoninae or Aphidiinae with confidence. Both have recently been treated as subfamilies within the Braconidae (van Achterberg 1993, 1995a). The aphidiines are especially interesting because some have the typical cyclostome poison gland and others do not. This, in combination with other features such as form of the fore wing subbasal cell, leads me to include them as a separate subfamily within the cyclostomes, despite the lack of an obvious cyclostome condition. The form of the clypeus and labrum is highly variable in the aphidiines, but those species with a more exposed, relatively bare labrum resemble species with similarly reduced labra among both euphorines and several small cyclostomes.
Robert A. Wharton 150
Hymenoptera: Evolution, Biodiversity and Biological Control
Discussion and Conclusions A major goal of systematics is to translate phylogenies into classifications that are sufficiently stable to facilitate rather than confound communication. But an exceptionally large number of subfamilies precludes depiction of within-family relationships under the framework of our Linnaean system. The classification presented here results in a drastic reduction in the number of subfamilies to facilitate portrayal of higher order relationships. By reducing most of the problems to the tribal level, the proposed classification also provides an explicit series of outgroups that can be used to address these problems. Replacing the large number of subfamilies with a few, large monophyletic groupings provides a useful framework for exploring fundamental issues (such as basal apocritan relationships, patterns of host usage, and physiological interactions among hosts and their parasitoids). This holds true even if some of the larger groupings proposed here (most notably the Agathidinae) are only weakly supported as monophyletic. This new proposal has some weaknesses, and one might argue, for example, that I am adding structure at higher levels while sacrificing our ability to portray lower level relationships. But classifications are dynamic, not static, and this one will change with addition of new information (including new taxa) just like any other. A major difference between this and more recent attempts to classify braconids is that an increase in number of subfamilies occurs only after increased understanding of relationships, rather than increased recognition of differences. I have proposed six subfamilies and five supertribes. If the philosophy underlying this classification is followed, I envision at most four additional subfamilies and perhaps wider use of the supertribal category once higher level relationships are more firmly established. Another potential weakness of this proposal is that I have not accorded formal rank to the primary division within the family: the cyclostomes vs. non-cyclostomes. It is logical to do so, given that I accept the hypothesis that these two clades are strongly supported. Recognising cyclostomes and non-cyclostomes as subfamilies, however, necessitates extensive use of the rank of supertribe in order to accommodate subordinate taxa.
Acknowledgements Most of the ideas on relationships proposed here are not new, and I have relied heavily on the published literature in this regard. I would like to acknowledge in particular discussions I have had with M. Sharkey and especially W. R. M. Mason, as well as detailed character assessments published by D. Quicke, C. van Achterberg, and J. Whitfield. I am also grateful to A. Gillogly, E. Riley, and S. Weller for comments on Coleoptera and Lepidoptera classifications, and to an anonymous reviewer for pointing out areas in need of clarification. This work would not have been possible without the generosity of curators and colleagues who provided rare material or access to such material over the years. I am particularly grateful to T. Huddleston and P. Marsh in this regard, but I also owe a debt of gratitude to D. Johnson, W. Mason, N. Schiff, M. Sharkey, H. Townes, C. van Achterberg, and D. Wahl. Finally, and not least of all, I thank F. Ronquist for stimulating me to think about this problem. This work was supported in part by the National Science Foundations (DEB9712543).
References Achterberg, C. van (1984) Essay on the phylogeny of Braconidae (Hymenoptera: Ichneumonoidea). Entomologisk Tidskrift 105: 41-58.
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? 151
Achterberg, C. van (1993) Illustrated key to the subfamilies of the Braconidae (Hymenoptera: Ichneumonoidea). Zoologische Verhandelingen 283: 1-189. Achterberg, C. van (1995a) Generic revision of the subfamily Betylobraconinae (Hymenoptera: Braconidae) and other groups with modified fore tarsus. Zoologische Verhandelingen Leiden 298: 1-242. Achterberg, C. van (1995b) Additions and corrections to van Achterberg, C. 1993. Illustrated key to the subfamilies of the Braconidae (Hymenoptera: Ichneumonoidea). pp. 1-2. Achterberg, C. van & Quicke, D. L. J. (1992) Phylogeny of the subfamilies of the family Braconidae: a reassessment assessed. Cladistics 8: 237-264. Ashmead, W. H. (1900) Classification of the Ichneumon flies, or the superfamily Ichneumonoidea. Proceedings of the United States National Museum 23: 1-220. Belokobylskij, S. A. (1992) On the classification and phylogeny of the braconid wasp subfamilies Doryctinae and Exothecinae (Hymenoptera, Braconidae). Part I. On the classification, 1. Entomologischeskoe Obozrenie 71: 900-928. [in Russian] Belshaw, R., Fitton, M., Herniou, E., Gimeno, C., & Quicke, D. L. J. (1998) A phylogenetic reconstruction of the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28S ribosomal RNA. Systematic Entomology 23: 109-123. Blanchard, E. (1845) Histoire des insects traitant de leurs moeurs et de leurs métamorphoses en général et comprenant une nouvelle classification fondée sur leurs rapports naturels. 1. Didot, Paris. Cˇapek, M. (1965) The classification of Braconidae in relation to host specificity. Proceedings of the 12th International Congress of Entomology. 1964: 98-99. Cˇapek, M. (1970) A new classification of the Braconidae (Hymenoptera) based on the cephalic structures of the final instar larva and biological evidence. Canadian Entomologist 102: 846-875. Dowton, M. & Austin, A. D (1998) Phylogenetic relationships among the microgastroid wasps (Hymenoptera: Braconidae): combined analysis of 16S and 28S rDNA genes and morphological data. Molecular Phylogenetics & Evolution 10: 354-366. Dowton, M., Austin, A. D. & Antolin, M. F. (1998) Evolutionary relationships among the Braconidae (Hymenoptera: Ichneumonoidea) inferred from partial 16S rDNA gene sequences. Insect Molecular Biology 7: 129-150. Edson, K. M. & Vinson, S. B. (1979) A comparative morphology of the venom apparatus of female braconids (Hymenoptera: Braconidae). Canadian Entomologist 111: 1013-1024. Fischer, M. (1971) Hym. Braconidae. World Opiinae. Index of Entomophagous Insects. Le Francois, Paris. Fischer, M. (1987) Hymenoptera: Opiinae III – aethiopische, orientalische, australische und ozeanische Region. Das Tierreich 105: 1-734. Foerster, A. (1862) Synopsis der Familien und Gattungen der Braconen. Verhandlungen des Naturhistorischen Vereines der preussischen Rheinlande und Westphalens 19: 225-288. ICZN (1985) International Code of Zoological Nomenclature, 3rd Ed. Adapted by the XX General Assembly of the International Union of Biological Science. International Trust for Zoological Nomenclature, London. Maetô, K. (1987) A comparative morphology of the male internal reproductive organs of the family Braconidae (Hymenoptera, Ichneumonoidea). Kontyû 55: 32-42. Marshall, T. A. (1885) Monograph of British Braconidae. Part I. Transactions of the Royal Entomological Society of London 1885: 1-280.
Robert A. Wharton 152
Hymenoptera: Evolution, Biodiversity and Biological Control
Mason, W. R. M. (1981) The polyphyletic nature of Apanteles Foerster (Hymenoptera: Braconidae): a phylogeny and classification of Microgastrinae. Memoirs of the Entomological Society of Canada 115: 1-147. Mason, W. R. M. (1983) A new South African subfamily related to Cardiochilinae (Hymenoptera: Braconidae). Contributions of the American Entomological Institute 20: 49-62. Nees von Esenbeck, C. G. (1812) Ichneumonides adsciti in genera et familias divisi a Dre. Nees ab Esenbeck. Magazin Gesellschaft Naturforschender Freunde zu Berlin 5: 3-37. Nees von Esenbeck, C. G. (1814) Ichneumonides adsciti in genera et familias divisi a Dre. Nees ab Esenbeck. Continuatio. Magazin Gesellschaft Naturforschender Freunde zu Berlin 6: 183-221. Nees von Esenbeck, C. G. (1816) Ichneumonides adsciti in genera et familias divisi a Dre. Nees ab Esenbeck. Continuatio. Magazin Gesellschaft Naturforschender Freunde zu Berlin 7: 243-277. Quicke, D. L. J. (1993) The polyphyletic origin of endoparasitism in the cyclostome lineages of Braconidae (Hymenoptera): a reassessment. Zoologische Mededelingen 67: 159-177. Quicke, D. L. J. & Achterberg, C. van (1990) Phylogeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneumonoidea). Zoologische Verhandelingen 258: 1-95. Sharkey, M. J. (1993) Family Braconidae. pp. 362-395. In Goulet H. & Huber, J. T.(Eds), Hymenoptera of the World: An Identification Guide to Families. Agriculture Canada, Ottawa. Sharkey, M. J. & Wharton, R. A. (1997) Morphology and terminology. pp. 19-37. In Wharton, R. A., Marsh, P. M. & Sharkey, M. J. (Eds), Manual of the New World Genera of the Family Braconidae (Hymenoptera). Special Publication No. 1, International Society of Hymenopterists, Washington, D. C. Shaw, M. R. (1983) On evolution of endoparasitism: the biology of some genera of Rogadinae (Braconidae). Contributions of the American Entomological Institute 20: 307-328. Shaw, M. R. & Huddleston, T. (1991) Classification and biology of braconid wasps. pp. 3-125. In Dolling, W. R. & Askew, R. R. (Eds), Handbooks for the Identification of British Insects vol. 7, part 11. Royal Entomological Society of London, London. Shenefelt, R. D. (1978) Braconidae 10 Braconinae, Gnathobraconinae, Mesostoinae, Pseudodicrogeniinae, Telengainae, Ypsistocerinae plus Braconidae in general, major groups, unplaced genera and species. pp. 1425-1872. In Achterberg, C. van & Shenefelt, R. D. (Eds), Hymenopterorum Catalogus (nova editio) pars 15. Dr. W. Junk, The Hague. Simbolotti, G. & van Achterberg, C. (1995) Revision of the Euagathis species (Hymenoptera: Braconidae: Bassinae) from the Sunda Islands. Zoologische Verhandelingen 293: 1-62. Szépligeti, G. (1904) Hymenoptera Fam. Braconidae (Première partie). Genera Insectorum 22 and 23: 1-253. Telenga, N. A. (1955) Fauna of the USSR. Hymenoptera. Vol V. No. 4. Braconidae Microgasterinae and Agathinae. Akademia Nauk SSSR, Moscow. Tobias, V. I. (1986) Keys to the Insects of the European Part of the USSR. III. Hymenoptera. 4. Akademia Nauk, St. Petersburg. [in Russian] Wharton, R. A. (1997) Introduction. pp. 1-18. In Wharton, R. A., Marsh, P. M. & Sharkey, M. J. (Eds), Manual of the New World Genera of the Family Braconidae (Hymenoptera). Special Publication No. 1, International Society of Hymenopterists, Washington, D. C. Wharton, R. A., Shaw, S. R., Sharkey, M. J., Wahl, D. B., Woolley, J. B., Whitfield, J. B., Marsh, P. M. & Johnson, J. W. (1992) Phylogeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneumonoidea): a reassessment. Cladistics 8: 199-235.
Can Braconid Classification be Restructured to Facilitate Portrayal of Relationships? 153
Whitfield, J. B. (1992) The polyphyletic origin of endoparasitism in the cyclostome lineages of Braconidae (Hymenoptera). Systematic Entomology 17: 273-286. Whitfield, J. B. (1997) Molecular and morphological data suggest a single origin of the polydnaviruses among braconid wasps. Naturwissenschaften 84: 502-507. Whitfield, J. B. & Mason, W. R. M. (1994) Mendesellinae, a new subfamily of braconid wasps (Hymenoptera, Braconidae) with a review of relationships within the microgastroid assemblage. Systematic Entomology 19: 61-76.
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) John T. Jennings and Andrew D. Austin Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected])
Introduction The Evanioidea are one of the most easily recognised and distinct group of parasitic Hymenoptera. All members of the superfamily have the metasoma inserted high on the propodeum, although this character state is found in some Ichneumonoidea, Cynipoidea and Chalcidoidea (Naumann 1991). Since Hedicke (1930) divided the superfamily into three families, the Aulacidae, Gasteruptiidae and Evaniidae, this classification has remained unchanged. However, the grouping of the Evaniidae with the Aulacidae and Gasteruptiidae has been questioned by numerous authors (e.g. Townes 1950; Crosskey 1951; Crosskey 1962; Carlson 1979; Naumann 1991; Gauld & Bolton 1996), the high insertion of the metasoma possibly having been acquired independently by the Evaniidae, as it has within other groups of Hymenoptera, albeit less obviously. A close relationship between the Aulacidae and Gasteruptiidae is less problematic and is at least putatively supported by several morphological characters. Gasteruptiidae and Aulacidae have a similar, rigid abutment of the pronotum and mesepisternum and a similar fusion or partial fusion of the first and second metasomal segments (Naumann 1991). Quicke et al. (1994) found some similarities between the ovipositor in the Aulacidae and Gasteruptiidae, but not Evaniidae. Both aulacids and gasteruptiids have a medial thickening of the ventral wall of the upper valve, but the latter has a mid-dorsal longitudinal ridge which is absent in aulacids. Evaniidae, however, differ in that their ovipositor is dorso-ventrally compressed rather than diverging as in aulacids and gasteruptiids. Apart from the high point of insertion of the metasoma, the only other putative synapomorphy for the three families is the loss of functional spiracles from all segments of the metasoma except the eighth (Gauld & Bolton 1996). In contrast, Dowton and Austin (1994), in a phylogenetic analysis of the mitochondrial 16S rRNA gene, have shown some tentative support for the Evaniidae being closely related to the Gasteruptiidae, but this finding was based on results for three exemplar species and did not include representatives of the Aulacidae. The Aulacidae is world-wide in distribution and comprises about 150 species arranged in two genera, Aulacus Jurine and Pristaulacus Kieffer (Gauld & Bolton 1996). The Gasteruptiidae comprises six genera currently arranged in two subfamilies: the Gasteruptiinae, containing the cosmopolitan Gasteruption L., and the Hyptiogastrinae which have a Gondwanan distribution (Crosskey 1962; Jennings & Austin 1997b). In the latter subfamily, Aulacofoenus Kieffer is known from Australia and South America (Jennings & Austin 1997b), Crassifoenus Crosskey and Hyptiogaster Kieffer are restricted to Australia (Jennings & Austin 1994a, 1997a), Eufoenus Szépligeti is predominantly Australian but extends into New Guinea and several south-west Pacific islands (Jennings & Austin 1994a, 1997a), and Pseudofoenus Kieffer is endemic to New Zealand (Jennings & Austin 1994b). 154
Hymenoptera: Evolution, Biodiversity and Biological Control
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) 155
The relationships within the Gasteruptiidae, particularly the phylogenetic status of the subfamilies, have not been examined previously. Here we present the results of a cladistic analysis aimed at confirming the monophyly of the Aulacidae and Gasteruptiidae, and the two subfamilies Gasteruptiinae and Hyptiogastrinae, as a prelude to more detailed studies revising the taxonomy of the latter subfamily and assessing competing hypotheses to explain its Gondwanan distribution.
Materials and Methods Specimens were examined using a Zeiss DR stereomicroscope, a Cambridge Stereoscan 250 (Mk 3B), a Phillips XL20 scanning electron microscope, or a Phillips XL30 field emission scanning electron microscope. They were first cleaned by removing obvious dirt and other debris and, for SEM, were either sputter-coated with gold or examined uncoated which was the case for the majority of specimens. Those used for examination of genitalia were partly dissected prior to the removal of the metasoma, hydrated in distilled water for up to 24 h, and the genitalia teased out with a fine needle. Terminology for general morphology follows Jennings and Austin (1994a), and that for wing venation follows the modified Comstock-Needham system after Sharkey (1988), but with some modifications, and using the nomenclature of van Achterberg (1979) for cells (see Jennings & Austin 1994a). A total of 54 taxa (see Appendix 1) were included in the phylogenetic analysis. Schlettererius cinctipes (Cresson) (Stephanidae) was used as an outgroup because it is considered basal to the Evanioidea (see Dowton & Austin 1994). Six exemplar species of Aulacidae from Australia were selected, three each from Aulacus and Pristaulacus. Four Australian species of Gasteruption and one unidentified South American species were also included. Forty-two hyptiogastrines were included because of their higher generic diversity, viz. nine species of Aulacofoenus, three Crassifoenus, twenty-three Eufoenus (including nine undescribed species), five Hyptiogaster and two Pseudofoenus. A total of 58 characters (see Appendix 1) were employed in the analysis. Female-based characters were used unless otherwise specified. Primary absence of a character was given a character state number ‘0’. Of the 58 characters, seven were multistate (characters 1, 3, 14, 16, 25, 37 and 38), while the remainder were binary. Two separate analyses were conducted using identical PAUP options: (a) S. cinctipes designated as the only outgroup taxon, and (b), where outgroups were not constrained (i.e. no outgroups specified). In each case, an heuristic search of 100 random replicates was undertaken using PAUP UNIX 4.0.0d63 (Swofford 1998) with random addition sequence, tree-bisection-reconnection branch swapping, steepest descent and MULPARS options. Because of the time required to conduct the analyses, the nchuck = 100 and chucklen = 1 options were invoked. All character states were unordered, i.e. there was no a priori assumption made regarding character evolution (see Ponder & Lindberg 1997).
Results and Discussion Heuristic analysis of the data set (see Appendix 2), found 20 000 equally parsimonious trees each 124 steps in length. In both analyses, the strict consensus tree (Fig. 1) generated from these 20 000 trees was identical and had a CI of 0.32. The low CI indicates a high level of homoplasy (Wiley et al. 1991), and is clearly expressed in the comb-like structure of the clade defining the Hyptiogastrinae. Characters 2, 10, 19–23, 26, 49, 53, and 55 proved uninformative and were excluded from
John T. Jennings and Andrew D. Austin 156
Hymenoptera: Evolution, Biodiversity and Biological Control
Aulacidae
Gasteruptiinae
Gasteruptiidae
Hyptiogastrinae
Figure 1
Schlettererius cinctipes Aulacus atriceps Aulacus sp. 1 Aulacus sp. 2 Pristaulacus cinguiculatus Pristaulacus variegatus Pristaulacus sp. 1 Gasteruption brachyurum Gasteruption fluviale Gasteruption spinigerum Gasteruption ?paradoxale Gasteruption sp. 1 Aulacofoenus deletangii Aulacofoenus fallax Aulacofoenus fletcheri Aulacofoenus infumatus Aulacofoenus kurmondi Aulacofoenus marionae Aulacofoenus perenjorii Aulacofoenus thoracicus Aulacofoenus whiani Crassifoenus grossitarsis Crassifoenus houstoni Crassifoenus macronyx Eufoenus antennalis Eufoenus australis Eufoenus crassitarsis Eufoenus darwini Eufoenus spinitarsis Eufoenus sp. 9 Eufoenus sp. 2 Eufoenus extraneus Eufoenus ferrugineus Eufoenus floricolus Eufoenus inaequalis Eufoenus minimus Eufoenus patellatus Eufoenus pilosus Eufoenus rieki Eufoenus ritae Eufoenus sp. 1 Eufoenus sp. 3 Eufoenus sp. 4 Eufoenus sp. 5 Eufoenus sp. 6 Eufoenus sp. 7 Eufoenus sp. 8 Hyptiogaster humeralis Hyptiogaster kalbarii Pseudofoenus crassipes Pseudofoenus unguiculatus Hyptiogaster arenicola Hyptiogaster pinjarregaensis Hyptiogaster rufus
Strict consensus tree of 20 000 equally most parsimonious trees each of 124 steps generated from a heuristic analysis of the data set in Appendix 2. CI = 0.32.
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) 157
the analyses. Despite this level of homoplasy, the strict consensus tree shows several important features; specifically that Aulacidae, Gasteruptiidae, Gasteruption (Gasteruptiinae) and Hyptiogastrinae are all monophyletic The monophyly of Aulacidae is supported by seven unequivocal character state changes in this study, viz. the antennae are inserted level with base of eyes (character 1:state 1); antennal sockets wide apart (4:1); eye small and circular/sub-circular (7:1); notauli percurrent, Y-shaped (25:2) (see Crosskey 1962); propleuron not elongate (30:1); propodeum not pyramidal (33:1) (see Crosskey 1962); and the presence of a groove on the inner surface of the hind coxae of females (41:1) (see Crosskey 1962). However, some Aulacus not included in this study, mostly those with short ovipositors, lack this groove (Townes 1951). Within Aulacidae, Aulacus is supported by the absence of the occipital carina (16:0) whilst in Pristaulacus it is not complete (absent medially) (state 1) (see Townes 1950). Pristaulacus is supported by the presence of pectinate hind claws (43:1). This character has been used by Townes (1950), Oehlke (1983) and others to distinguish between Aulacus and Pristaulacus (Aulacidae). In Aulacus, the claws are apparently simple, but each has a single basal tooth which is difficult to see (Townes 1950). The monophyly of the Gasteruptiidae is supported unequivocally by the antennae inserted approximately half-way up the eyes (character 1: state 2). Within Gasteruptiidae, Gasteruptiinae is supported by the mandibles overlapping only slightly (13:0), whereas they broadly overlap in Hyptiogastrinae (state 1) (see Crosskey 1962), and fore wing vein 1-Rs+M joins at M+Cu, 1-R (37:0) (see Jennings & Austin 1994b). The monophyly of the Hyptiogastrinae is supported unequivocally by the lack of a trochantellus on the hind leg (50:0) and an incised apical sternum (51:0). In S. cinctipes, Aulacidae and Gasteruptiinae, the trochantellus is present and the apical sternum is not incised. Relationships within the Hyptiogastrinae are largely unresolved. Of the five currently recognised hyptiogastrine genera, only the monophyly of Crassifoenus and Pseudofoenus are supported in this analysis. Crassifoenus is supported by the presence of hind wing vein 1-Cu (39:1), a mid tibial notch in females (48:1), and ovipositor sheath margins that are either undulate or serrate (58:1) (see also Jennings & Austin 1994a). Pseudofoenus is supported by a reduction in two characters; the absence of discal cells in the fore wing (34:0) and fore wing vein 1-Rs+M fused to form Rs+M+Cu(b) (37:2). Jennings and Austin (1994b) discuss possible evolutionary pathways for these character states. Although not presented here, further manipulation of the data set to take into account the influence of multistate characters, i.e. successive weighting, did not change the overall conclusion outlined above, although relationships within the Hyptiogastrinae were slightly more resolved. However, the included species of Aulacofoenus, Eufoenus and Hyptiogaster were still not resolved as monophyletic. This is the first cladistic evidence that the Hyptiogastrinae is monophyletic, a result that will allow future work to more realistically focus on the internal relationships of the subfamily by increasing the number of taxa and characters within this group. This is important for two reasons; 1) to provide an assessment of likely factors that are involved in determining the present continental distribution of hyptiogastrine wasps, a group showing a restricted Gondwanan distribution (Australasia and South America), and 2) to determine likely pathways for the development of observed host relationships. Whereas members of the Aulacidae are endoparasitoids of wood boring wasps (Xiphydriidae) or beetles (Cerambycidae and Buprestidae) (Carlson 1979; Gauld &
John T. Jennings and Andrew D. Austin 158
Hymenoptera: Evolution, Biodiversity and Biological Control
Bolton 1996), a wide range of solitary bees and wasps act as hosts for Gasteruption including members of the families Anthophoridae, Colletidae, Megachilidae, and Vespidae (Crosskey 1962; Jennings & Austin unpublished data). Whilst there is only limited information on the host relationships of the Hyptiogastrinae, they appear to be predator-inquilines in the nests of solitary bees or wasps. Aulacofoenus is recorded as a predator-inquiline of halictid bees (Jennings & Austin 1997b), Crassifoenus of stenotritid bees (Houston 1984b, 1987; Jennings & Austin 1994a), Eufoenus of colletid and halictid bees (Jennings & Austin unpublished data), Hyptiogaster of masarid wasps (Houston 1984a; Naumann & Cardale 1987; Jennings & Austin 1997a) and stenotritids (Houston 1984b, 1987; Jennings & Austin 1997a), and Pseudofoenus of colletids (Valentine & Walker 1991; Jennings & Austin 1994b). A robust phylogeny for the Hyptiogastrinae is necessary to further assess the level of host group specialisation for the various clades and the likely pathways that led to the transition from one host group to another.
Acknowledgements The authors wish to acknowledge the curators of a large number of collections from which the specimens used in the analyses were borrowed. We also wish to thank Dr Peter Cranston of the Australian National Insect Collection, Canberra, for valuable discussions on interpretation of the PAUP analyses.
References Carlson, R. W. (1979) Superfamily Evanioidea. pp. 1109-1118. In Krombein, K. V., Hurd, P. D., Smith, D. R. & Burks, B. D. (Eds), Catalog of Hymenoptera in America North of Mexico. Vol. 1. Symphyta and Apocrita (Parasitica). Smithsonian Institution Press, Washington, D.C. Crosskey, R. W. (1951) The morphology, taxonomy, and biology of the British Evanioidea (Hymenoptera). Transactions of the Royal Entomological Society, London 102: 247-301. Crosskey, R. W. (1962) The classification of the Gasteruptiidae (Hymenoptera). Transactions of the Royal Entomological Society, London 114: 377-402. Dowton, M. & Austin, A. D. (1994) Molecular phylogeny of the insect order Hymenoptera: Apocritan relationships. Proceedings of the National Academy of Sciences, USA 91: 99119915. Gauld, I. & Bolton, B. (Eds) (1996) The Hymenoptera (2nd Ed.). British Museum (Natural History), London and Oxford University Press, Oxford. Hedicke, H. (1930) Gasteruptiidae. Hymenopterorum Catalogus 11: 1-54. Houston, T. F. (1984a) Bionomics of a pollen-collecting wasp, Paragia tricolor (Hymenoptera: Vespidae: Masarinae), in Western Australia. Records of the Western Australian Museum 11: 141-151. Houston, T. F. (1984b) Biological observations of bees in the genus Ctenocolletes (Hymenoptera: Stenotritidae). Records of the Western Australian Museum 11: 153-172. Houston, T. F. (1987) A second contribution to the biology of Ctenocolletes bees (Hymenoptera: Apoidea: Stenotritidae). Records of the Western Australian Museum 13: 189-201. Jennings, J. T. & Austin, A. D (1994a). Revision of the genus Crassifoenus Crosskey (Hymenoptera: Gasteruptiidae: Hyptiogastrinae), with a description of a new species from Western Australia. Records of the Western Australian Museum 16: 575-91. Jennings, J. T. & Austin, A. D. (1994b) Revision of Pseudofoenus Kieffer (Hymenoptera: Gasteruptiidae), a hyptiogastrine wasp genus endemic to New Zealand. Invertebrate Taxonomy 8: 1289-1303.
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) 159
Jennings, J. T. & Austin, A. D. (1997a) Revision of the Australian endemic genus Hyptiogaster Kieffer (Hymenoptera: Gasteruptiidae), with descriptions of seven new species. Journal of Natural History 31: 1533-1562. Jennings, J. T. & Austin, A. D. (1997b) Revision of Aulacofoenus Kieffer (Hymenoptera: Gasteruptiidae), hyptiogastrine wasps with a restricted Gondwanic distribution. Invertebrate Taxonomy 11: 943-976. Naumann, I. D. (1991) Hymenoptera. pp. 916-1000. In CSIRO, The Insects of Australia, Vol. II. Melbourne University Press, Melbourne. Naumann, I. D. & Cardale, J. (1987) Notes on the behaviour and nests of an Australian masarid wasp Paragia (Paragia) decipiens decipiens Shuckard (Hymenoptera: Vespoidea: Masaridae). Australian Entomological Magazine 13: 59-65. Oehlke, J. (1983) Revision der europäischen Aulacidae (Hymenoptera-Evanioidea). Beiträge zur Entomologie 33: 439-447. Ponder, W. F. & Lindberg, D. R. (1997) Towards a phylogeny of gastropod molluscs: an analysis using morphological characters. Zoological Journal of the Linnean Society 119: 83-265. Quicke, D. l. J., Fitton, M. G., Tunstead, J. R., Ingram, S. N. & Gaitens, P. V. (1994) Ovipositor structure and relationships within the Hymenoptera, with special reference to the Ichneumonoidea. Journal of Natural History 28: 635-682. Sharkey, M. (1988) Ichneumonoid wing venation. Ichnews 11: 2-12. Swofford, D. L. (1998) PAUP – Phylogenetic Analysis Using Parsimony, Version 4.0.0d. Computer program distributed by Laboratory of Molecular Systematics, Smithsonian Institution, Washington D.C. Townes, H. K. (1950) The Nearctic species of Gasteruptiidae (Hymenoptera). Proceedings of the United States National Museum 100: 85-145. Townes, H. K. (1951) Family Gasteruptiidae. pp. 655-661. In Muesebeck, C. F. W., Krombein, K. V. Townes, H. K., et al. (Eds), Hymenoptera of America North of Mexico, Synoptic Catalog. United States Department of Agriculture, Agricultural Monograph. van Achterberg, C. (1979) A revision of the subfamily Zelinae auct. (Hymenoptera, Braconidae). Tijdschrift voor Entomologie 122: 241-479. Valentine, E. W. & Walker, A. K. (1991) Annotated catalogue of New Zealand Hymenoptera. DSIR Plant Protection Report 4: 1-84. Wiley , E. O., Siegel-Causey, D., Brooks, D. R., Fink, V. A. et al. (1991) The Compleat Cladist. A Primer of Phylogenetic Procedures. The University of Kansas, Museum of Natural History Special Publication.
John T. Jennings and Andrew D. Austin 160
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 1 Characters and character states employed in the phylogenetic analysis. Character 1. Antennal insertion relative to eye: 0) well below eyes; 1) level with base of eyes; 2) approximately half-way up eyes. Character 2. Antennal segment number, female: 0) multi-segmented; 1) 14. Character 3. Antennal segment number, male: 0) multi-segmented; 1) 13; 2) 14. Character 4. Width between antennal sockets: 0) wide apart; 1) close. The latter state is found in aulacids (Crosskey 1962). Character 5. Clypeal margin: 0) sinuate; 1) truncate lobe. Character 6. Clypeal ridge medially: 0) absent; 1) present. Character 7. Eye size and shape: 0) large and elliptical; 1) small and circular/sub-circular. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 8. Flagellomere 1 length relative to flagellomere 2: 0) first flagellomere greater in length than second; 1) first flagellomere less than or equal to length second. Character 9. Frontal carina: 0) absent; 1) present. Character 10. Head width:length when viewed dorsally: 0) quadrate to lateral; 1) elongate (i.e. longer than wide). Character 11. Lateral epistomal suture: 0) absent; 1) present. Character 12. Malar space width:height eye: 0) ≤ 0.15; 1) > 0.15. Character 13. Mandibles broadly overlap: 0) no; 1) yes. In Gasteruption, the mandibles overlap only slightly whereas they broadly overlap in the Hyptiogastrinae (Crosskey 1962). Character 14. Mandibular median teeth number: 0) one tooth; 1) two teeth; 2) three teeth (2). Character 15. Mouthparts extendible: 0) fixed; 1) extendible. Character 16. Occipital carina: 0) absent; 1) incomplete (absent medially); 2) complete. In Aulacidae, Townes (1950) distinguished between Pristaulacus and Aulacus by the latter lacking an occipital carina. Character 17. Occipital margin sculpturing: 0) sculptured; 1) smooth. Character 18. Scape in lateral view: 0) convex; 1) parallel-sided. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 19. Scape width relative to pedicel: 0) scape much wider than pedicel; 1) scape slightly wider than pedicel. Character 20. Subantennal groove: 0) absent; 10 present. Character 21. Dorsal tentorial pits on head: 0) absent; 1) present. Character 22. Functional abdominal spiracles: 0) other than 1 & 8; 1) 1 and 8. Character 23. Median sulcus of mesoscutum: 0) absent; 1) present.
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) 161
Character 24. Mesothorax anterior face in lateral view: 0) truncate; 1) not so. Character 25. Notauli percurrent: 0) not percurrent; 1) percurrent, not Y-shaped; 2) percurrent, Y-shaped (2). This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Crosskey distinguished between scutum ‘divided’ (not Y-shaped) or ‘not fully divided by prescutum’ (Y-shaped). Character 26. Prepectus: 0) absent; 1) present. Character 27. Antero-dorsal pronotal processes: 0) absent; 1) present. Jennings and Austin (1997b) described these processes in some Aulacofoenus species. Character 28. Dorso-lateral pronotal processes: 0) absent; 1) present. Jennings and Austin (1997b) described these processes in some Aulacofoenus species. Character 29. Propleural carina: 0) absent; 1) ventro-lateral. Character 30. Propleuron shape: 0) elongate; 1) not elongate. Character 31. Propodeal carina: 0) absent; 1) present. Character 32. Propodeal spiracle: 0) glabrous or almost so; 1) fringed with setae. When setae are present, they are generally long and found on both the anterior and posterior margins of the spiracle – only occasionally are the setae short. If only one or two setae are present, this was considered as state 0. Character 33. Shape of propodeum: 0) not pyramidal; 1) pyramidal. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 34. Fore wing discal cell number: 0) 2 cells; 1) 1 cell. Jennings and Austin (1994a, 1994b, 1997a, 1997b) figured the discal cells of many included taxa. Character 35. Fore wing plication at rest: 0) no; 1) yes. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 36. Fore wing vein ‘r-m’: 0) absent; 1) present. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 37. Fore wing vein 1-Rs+M: 0) joins at M+Cu, 1-R; 1) joins 1-M and 1-Rs; 2) fused – forms Rs+M+Cu(b). Jennings and Austin (1994b) discussed possible evolutionary pathways for this character. Character 38. Fore wing vein 2-M colour: 0) even; 1) pale apically; 2) pale basally. With state 2, the vein also becomes thinner apically. Character 39. Hind wing vein 1-Cu: 0) absent; 1) present. Character 40. Submarginal cell number in fore wing: 0) 2 or 3 cells; 1) 1 cell. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 41. Groove on hind coxa: 0) absent; 1) present. This character was used by Crosskey (1962) to distinguish between Aulacidae and Gasteruptiidae. Character 42. Groove on hind trochanter: 0) absent; 1) present. This character has been used by Crosskey (1962) and Jennings and Austin (1994b) to distinguish between Pseudofoenus and Eufoenus which lack a dorso-ventral groove on the hind trochanter and other hyptiogastrine
John T. Jennings and Andrew D. Austin 162
Hymenoptera: Evolution, Biodiversity and Biological Control
genera which have the groove. However, several described species of Eufoenus have a weak illdefined groove on either the dorso-lateral surface or restricted to the inner lateral surface. There is some doubt as to whether this is analogous to the groove in genera such as Crassifoenus, but in these species this state has been coded as 1. Character 43. Hind claw: 0) simple; 1) pectinate. These character states have been used by Townes (1950), Oehlke (1983) and others to distinguish between Aulacus and Pristaulacus (Aulacidae). In Aulacus, the claws are apparently simple, but each has a single basal tooth which is difficult to see (Townes 1950). In Pristaulacus, two or more teeth are found. Character 44. Hind tarsal segment 1: 0) without projection, symmetrical; 1) with lateral projection, highly asymmetrical. Character 45. Lateral projections on hind tarsal segments 2-4: 0) absent; 1) present. Character 46. Length of hind tarsal segments: 0) normal; 1) shortened. Jennings and Austin (1997b) have described shortened hind tarsal segments found in a small number of taxa. Character 47. Hind tibia with ventro-apical pecten of stout spines: 0) absent; 1) present. Character 48. Mid tibial notch: 0) absent; 1) present. The mid tibial notch is present in females of Crassifoenus (Jennings & Austin 1994a). Character 49. Prefemur on hind leg: 0) absent; 1) present. Character 50. Trochantellus on hind leg: 0) absent; 1) present. Character 51. Apical sternum: 0) incised; 1) not incised. Character 52. Digitus length compared with length of basiparameres (male): 0) digitus length < basiparameres; 1) digitus length ≥ basiparameres. Character 53. Metasomal insertion on propodeum: 0) low; 1) high. Character 54. Metasomal shape: 0) not sub-clavate; 1) sub-clavate. Character 55. Metasomal T1 and T2: 0) not fused; 1) fused. Character 56. Metasomal T1 longitudinal medial ridge or line: 0) absent; 1) present. Character 57. Ovipositor exsertion: 0) exserted; 1) not exserted. Of the hyptiogastrine wasps, only Hyptiogaster has exserted ovipositors (Jennings & Austin 1997a). Character 58. Ovipositor sheath margin: 0) smooth; 1) undulate or serrate. Crassifoenus species have ovipositor sheath margins that are either undulate or serrate (Jennings & Austin 1994a).
Higher-level Phylogeny of the Aulacidae and Gasteruptiidae (Hymenoptera: Evanioidea) 163
Appendix 2 Data matrix for 54 included taxa and 58 characters. Inapplicable data indicated by ‘-’. Taxon/Character
1
11111
11112
22222
22223
33333
33334
44444
44445
55555
555
12345
67890
12345
67890
12345
67890
12345
67890
12345
67890
12345
678
Schlettererius cinctipes
00010
00100
01100
21001
01001
10001
10110
01000
00000
01011
10000
110
Aulacus atriceps
11100
11000
01110
0-010
10102
00011
00010
01000
11000
00001
11101
010
Aulacus sp. 1
11100
11000
01110
0-010
10112
00011
00010
01000
11000
00001
11101
010
Aulacus sp. 2
11100
11000
01110
0-010
10102
00011
00010
01000
11000
00001
10101
010
Pristaulacus cinguiculatus
11100
01000
01110
10010
10102
00011
00010
01000
11100
00001
10101
010
Pristaulacus variegatus
11100
01000
01110
10010
10102
00011
00010
01000
11100
00001
10101
010
Pristaulacus sp. 1
11100
01000
01110
10010
10102
00011
00010
01000
11100
00001
11101
010
Gasteruption brachyurum
21110
00000
00001
20110
10110
00010
10111
10101
01000
01001
10111
010
Gasteruption fluviale
21110
00000
00001
20110
10111
00000
10111
10101
01000
01001
10111
010
Gasteruption spinigerum
21110
00000
01001
20110
10111
00000
11111
10101
01000
00001
10111
010
Gasteruption ?paradoxale
21110
00010
00001
20110
10111
00000
11111
10101
01000
01001
10111
010
Gasteruption sp. 1
21110
00100
00001
21110
10101
00010
10111
10101
01000
01001
10111
110
Aulacofoenus deletangii
21110
00100
00111
20110
10110
00010
11111
11201
01000
10000
00111
000
Aulacofoenus fallax
21110
00110
00111
21110
10110
00010
11111
11101
01000
00000
00111
000
Aulacofoenus fletcheri
21110
10111
00111
20110
10110
00000
10111
11101
01000
00000
00111
100
Aulacofoenus infumatus
21110
00100
00111
20110
10110
00010
10111
11201
01000
00000
00111
000
Aulacofoenus kurmondi
21110
00110
00111
21110
10100
00010
11111
11101
01000
10000
00111
000
Aulacofoenus marionae
21110
10110
00111
21110
10110
00000
10111
11101
01000
00000
00111
000
Aulacofoenus perenjorii
21110
00110
00111
20110
10110
00000
11111
11101
01000
00000
00111
100
Aulacofoenus thoracicus
21110
00110
00111
20110
10110
00010
11111
11101
01011
00000
00111
100
Aulacofoenus whiani
21110
00110
01111
21110
10110
00000
11111
11101
01000
00000
00111
000
Crassifoenus grossitarsis
21110
00110
11111
20110
10110
00010
11111
11111
01000
10100
00111
001
Crassifoenus houstoni
21110
00110
11111
20110
10110
00010
11111
11111
01000
10100
00111
001
Crassifoenus macronyx
21110
00110
11111
20110
10110
00010
11111
11111
01000
10100
00111
001
Eufoenus antennalis
21110
00110
00111
21110
10110
00010
11111
11101
01000
01000
00111
000
Eufoenus australis
21110
00110
00111
21110
10110
00010
11111
11101
00000 00&1000
00111
000
John T. Jennings and Andrew D. Austin 164
Hymenoptera: Evolution, Biodiversity and Biological Control
Taxon/Character
1
11111
11112
22222
22223
33333
33334
44444
44445
55555
555
12345
67890
12345
67890
12345
67890
12345
67890
12345
67890
12345
678
Eufoenus crassitarsis
21110
00110
00111
21110
10110
00000
10111
11101
00010
01000
00111
000
Eufoenus darwini
21110
00100
00111
21110
10110
00000
10111
11101
00011
00000
00111
000
Eufoenus extraneus
21210
00100
00111
21110
10110
00010
11111
11001
00000
01000
00111
000
Eufoenus ferrugineus
21110
00110
00111
21110
10110
00000
11111
11101
00000
01000
00111
000
Eufoenus floricolus
21110
00100
00111
21110
10110
00010
10111
11101
00000
01000
00111
000
Eufoenus inaequalis
21110
00110
00111
21110
10110
00000
11111
11101
00000
10000
00111
100
Eufoenus minimus
21110
00100
01111
20110
10110
00010
11111
11101
00000
01000
00111
000
Eufoenus patellatus
21110
00110
00111
21110
10110
00010
11111
11101
00000 00&1000
00111
000
Eufoenus pilosus
21110
00100
00111
21110
10110
00010
00111
11101
00000
00000
00111
000
Eufoenus rieki
21210
00110
00111
21110
10110
00010
10111
11101
00000
01000
00111
000
Eufoenus ritae
21110
10100
00111
21110
10110
00000
11111
11101
00000
01000
00111
000
Eufoenus spinitarsis
21110
00100
01111
21110
10110
00000
11111
11101
00011
00000
00111
000
Eufoenus sp. 1
21110
00100
00111
21110
10110
00000
11111
11101
00000
00000
00111
000
Eufoenus sp. 2
21110
00100
01111
21110
10110
00010
10111
11101
00010
00000
00111
000
Eufoenus sp. 3
21110
00100
01111
21110
10110
00010
10111
11101
00000
00000
00111
000
Eufoenus sp. 4
21110
00100
00111
21110
10110
000?0
??111
11101
00000
01000
00111
000
Eufoenus sp. 5
21110
00100
00111
21110
10110
00010
10111
11101
00000
01000
00111
000
Eufoenus sp. 6
21110
00110
00111
20110
10110
00010
11111
11001
00000
00000
00111
100
Eufoenus sp. 7
21110
10110
00111
21110
10110
00010
11111
11101
00000
00000
00111
000
Eufoenus sp. 8
21110
00110
00111
21110
10110
00000
10111
11001
00000
00000
00111
000
Eufoenus sp. 9
21110
10100
01111
21110
10110
00000
11111
11101
00011
00000
00111
000
Hyptiogaster arenicola
21110
00110
10111
21110
10100
00010
10111
11101
01000
01000
00111
010
Hyptiogaster humeralis
21111
00110
10111
21110
10110
01110
11111
11101
01000
10000
00111
010
Hyptiogaster kalbarrii
21111
00010
10121
21110
10100
01110
11111
11101
01000
00000
00111
010
Hyptiogaster pinjarregaensis
21110
00110
11111
21110
10100
00000
10111
11101
01000
01000
00111
010
Hyptiogaster rufus
21110
00010
10111
21110
10100
00000
10111
11101
01000
01000
00111
010
Pseudofoenus crassipes
21110
10110
00111
21110
10110
00010
11101
12101
00010
00000
00111
000
Pseudofoenus unguiculatus
21110
10110
00111
21110
10110
00010
10101
12101
00000
10000
00111
000
Monophyly and Relationship of the Genus Coelopisthia Foerster (Chalcidoidea: Pteromalidae) Hannes Baur Department of Invertebrates, Natural History Museum, Bernastrasse 15, CH-3005 Bern, Switzerland (email:
[email protected])
Introduction The genus Coelopisthia Foerster, 1856, belongs to the subfamily Pteromalinae and is known from Europe (Askew 1980) through Middle Asia (Dzhanokmen 1978) to China (Yang 1996), and North America (Peck 1963). Twelve species are considered as valid by Noyes (1998 sub Kranophorus Graham) but examination of a large amount material from the Holarctic region revealed that there may be at least 15 species. Host records are rather scarce and include various Lepidoptera (Arctiidae, Geometridae, Noctuidae, Tortricidae) and Coleoptera (Curculionidae, Scolytidae) (Peck 1963; Graham 1969; Dzhanokmen 1978; Askew 1980; Yang 1996). A few species possibly attack their hosts as secondary parasitoids via other Hymenoptera (Ichneumonidae) (cf. Huber et al. 1996). Species of Coelopisthia have been readily recognised in the past by a combination of features (Graham 1956, 1969; Bouºek & Rasplus 1991; Bouºek & Heydon 1997) which include a protruding face at the level of the toruli, a large head with conspicuous temples, enlarged anelli with at least the second one subquadrate, reduced wing pilosity, and an almost circular gaster. Although these characters may well be diagnostic for the genus they have proved to be critical with respect to phylogenetic considerations. There are for instance many other genera within the pteromalines with which individual features are shared, e. g. enlarged anelli (Rhopalicus Foerster), a protruding face (Conomorium Masi, Diglochis Foerster), a roundish gaster (Cyclogastrella Bukowskii, Schizonotus Rutzeburg). Moreover, some species, like Coelopisthia pachycera Masi, deviate considerably from the above pattern in that the anelli are quite strongly transverse and the face is much less protruding. In the context of a revision of Holarctic species of Coelopisthia these findings thus led to open questions about monophyly and relationships. In order to determine the natural limits of the genus a cladistic analysis was performed of which the results are presented here. The study concentrated on identifying synapomorphies of Coelopisthia whereas the cladistic structure among its species was of secondary interest. It was also not intended to solve problems of monophyly and relationship of any other genera mentioned below, as these should be addressed in a more comprehensive study of the entire subfamily. N.B. The authors for genera and species are given in Appendix 1.
Materials and Methods The material examined in this study is deposited in the institutions listed in the Acknowledgements. Specimens were examined either under a stereo-microscope with a magnification up to 100 × or a scanning electron microscope (SEM). Individual parts were mounted on slides according to the method described by Noyes (1982) but with the use of Euparal instead of Canada balsam. Morphology and terminology follow Gibson (1997).
165
Hannes Baur 166
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Strict consensus tree of 203 most parsimonious trees of a length of 67 steps (CI 0.67, RI 0.84). Taxon names abbreviated using the first three letters of genus and species name (cf. Appendix 3).
For the cladistic analysis 17 discrete morphological characters were scored for females of 44 species in 16 genera (see Appendices 1–3). These taxa could be united into 30 groups by the «search & merge» routine in MacClade (Maddison & Maddison 1992) before the analysis. Multistate characters were not polarised and treated as unordered, and multistate taxa as polymorphic. A parsimony analysis was performed using PAUP 3.1.1 (Swofford 1993). Because of the large size of the data matrix the heuristic search method was employed using random addition sequences with 50 replicates and TBR branch swapping. The search was repeated five times and always found the same set of most parsimonious trees. Consensus trees were computed with PAUP but the final trees were edited in MacClade.
Phylogenetic Analysis The selection of taxa for the ingroup and outgroup was difficult, since the phylogenetic relationships within and among Pteromalinae have not yet been investigated. Graham’s (1969) monumental revision of the species of north-western Europe is the last comprehensive treatment of the subfamily at the species level but dates from a pre-cladistic era. Although the work offers plenty of information, and the genera are apparently arranged according to some supposed relationship, no statements can be found with regard to their particular classification. The Dibrachysgroup of Wallace (1973) simply lumps several genera of Pteromalinae in a more or less artificial manner. Bouºek’s (1988) account on the Australasian fauna which I will follow here is more informative with respect to some lower levels (tribes) but covers few of the potentially useful genera. In an attempt to find a sound starting point for the analysis, exemplars of an array of genera within the tribe Pteromalini were therefore chosen for the ingroup (Appendices 1, 3). These taxa are characterised by having the postmarginal vein only about as long as the stigmal vein. Considering the rest of Pteromalinae and the probably closely related Miscogastrinae (Boucek 1988; Bouºek & Rasplus 1991) as outgroups, a reduced postmarginal vein may appear as a possible synapomorphy of those genera. There are, however, some exceptions in pteromalines: species of Meximalus Bouºek, Ptinocida Bouºek, Hemadas Bouºek (cf. Bouºek & Heydon 1997), and certain Mesopolobus Westwood (e.g. M. adrianae Gijswijt) also have a shortened postmarginal vein but are likely to be only distantly related and were not included in the analysis. Coe-
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 167
Figure 2
Section of strict consensus tree (Fig. 1) showing distribution of character state changes in Coelopisthia s.l. and Diglochis.
lopisthia s.l. (cf. Appendix 1) comprised 12 species of which C. pachycera, C. sp. D, and C. sp. F were tentatively placed in the genus. The latter deviate more or less from the general pattern of Coelopisthia and their position on the tree was therefore subjected to a test in the analysis. Following Yeates (1995) the choice of representatives for the other large genera consisted of a mixed sample of supposed basal and derived taxa. The same applied to the outgroup taxa with respect to the ingroup. Pachyneuron is possibly close to the ingroup and shows many derived traits, while Rhopalicus would be expected to split off at the base of Pteromalini. Janssoniella and Plutothrix are currently classified in the tribe Trigonoderini, while Rhicnocoelia belongs to the Miscogastrinae. Adult females of each species (see Appendix 1) were examined for discrete morphological characters. Males were not available for many species and were therefore excluded from the analysis. For some of the features it was not possible to define discrete states, since the underlying structure varied continuously. This led to the exclusion of much potentially useful information, sometimes to fusion of states. Some of these characters were nevertheless retained for the analysis and were broken arbitrarily into different states (marked with *). Furthermore, where characters appeared to be correlated they were treated as one character with a number of states (e. g. characters 4, 12). Eventually, a total of 17 characters were scored and are described in Appendix 2. Analysis of the reduced data matrix yielded more than 200 most parsimonious trees. This large number was probably caused by insufficient data available to resolve some of the taxa. These groups show numerous polytomies, some of them with many branches, in the strict consensus tree (Fig. 1). However, the latter is most revealing with regard to the initial questions about monophyly and relationship of Coelopisthia (Fig. 2). Diglochis appeared as the sister group of Coelopisthia s.l. based on the mouth margin being straight before its junction with the malar sulcus (2.0, Fig. 4), and the anterior part of the plicae strongly bent inwards (12.2, Fig. 10). Both character states appeared only once on the tree. It should be noted, however, that these states are independently derived in C. pachycera (12.1, Fig. 9) and some species of Diglochis (12.4, Fig. 11).
Hannes Baur 168
Figures 3–12
Hymenoptera: Evolution, Biodiversity and Biological Control
3) mouth margin (MM) and malar sulcus (MS) of Conomorium amplum; 4) mouth margin (MM) and malar sulcus (MS) of Coelopisthia pachycera; 5) occipital carina absent in Conomorium amplum; 6) occipital carina (OC) present in Coelopisthia extenta; 7, 8) distal edge (DE) of first left antennal anellus: 7) Dibrachys cavus; 8) Coelopisthia sp. L.; 9) anterior plica (AP) and posterior plica (PP) of Coelopisthia pachycera; 10) anterior plica and posterior plica of Coelopisthia extenta; 11) anterior plica and posterior plica of Diglochis sylvicola; 12) anterior plica and posterior plica of Cyclogastrella clypealis. Scale lines = 0.1 mm.
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 169
Coelopisthia s.l. (i. e. the above mentioned doubtfully placed species C. pachycera, C. sp. D, and C. sp. F plus the rest of the genus) is characterised by the presence of a laminate edge at the distal end of the first anellus (4.1, Fig. 8) and the particular shape of the third flagellar segment (7.1), both uniquely derived. The latter species except C. sp. D, finally shows a laterally weakly developed and medially almost effaced occipital carina (Fig. 6) as a synapomorphy.
Discussion The above results confirm to some extent the works of earlier studies (Bouºek & Rasplus 1991, Graham 1956, 1969). On the other hand there are some differences which may reflect a shift in focus. One aim of this study was then to transform earlier claims into some precise hypothesis on character state distribution and sister group relationship using cladistic methodology. This is particularly illuminating with regard to characters and the formulation of separate states. For instance Graham’s (1969) mentioning of the anelli being large is certainly true as these are larger in Coelopisthia than in many other genera and somehow ‘different’. But as discussed above, size alone was not the decisive criterion, since other and sometimes similar species shared this feature, too. What matters was the laminate edge on the distal end of the first anellus, evidently a synapomorphy of those species. Moreover, some important characters have also gone unnoticed in the past. One of them concerns the occipital carina, an autapomophy of Coelopisthia s.str. According to Graham (1956) those species show no carina but only a slight edge due to the abrupt angle by which the vertex turns into the occiput. Close examination nevertheless revealed that a weak carina was always indicated laterally below the transition from the vertex to the occiput (Fig. 6). The condition described by Graham is present in species of Conomorium where no carina was discernible (Fig. 5). The species which were initially regarded as doubtfully placed in the genus must now be considered differently. C. pachycera and C. sp. F were unified in a polytomy with Coelopisthia s.str. and therefore caused no problem. On the other hand, C. sp. D, an undescribed North African species, was shown to be the single sister taxon of the latter clade. That species differed considerably from all other species and showed many highly derived features. It might therefore be considered as belonging to a separate genus by many workers. However, it still shared two synapomorphies (4.1, 7.1) and was equally supported as the sister group on all of the trees. This and a certain discomfort with the creation of a generic name for a single species, particularly in a group where according to Grissell & Schauff (1990) about half of the Nearctic genera alone were monotypic, led to the inclusion of C. sp. D in Coelopisthia. Based on this broader concept the genus is thus characterised as follows: body generally squat; head large with conspicuous temples, more or less protuberant at level of toruli; antenna with both anelli enlarged, also the first anellus with a laminate edge at the distal end, third flagellar segment with one row of longitudinal sensilla confined to distal half; fore wing broad, wing disk slightly to very distinctly vaulted, sometimes darkened, sparsely pilose, without marginal setae; marginal and postmarginal vein slightly removed from wing margin, postmarginal vein from slightly shorter than to as long as stigmal vein; plica of propodeum with anterior part often strongly bent inwards and sometimes reaching adpetiolar strip, posterior part distinct and sometimes reaching anterior part or spiracles; median carina of propodeum effaced in the middle or irregular; gaster almost circular, broader than mesosoma. Finally, the sister group relationship of Coelopisthia and Diglochis is most interesting concerning their hosts. Species of Coelopisthia, Conomorium and Cyclogastrella are known to attack mostly
Hannes Baur 170
Hymenoptera: Evolution, Biodiversity and Biological Control
Lepidoptera (occasionally Coleoptera in some Coelopisthia) while Diglochis was reared exclusively from tabanid flies (Noyes 1998). The strict consensus tree suggests that there was a host switch from Lepidoptera to Diptera in the ancestor of Diglochis. How this different preferences came into existence can not be the subject of this study but probably led to the apparent differentiation of many features in those taxa (cf. Graham 1969).
Acknowledgements I am indebted to Karl Babl, Peter Eggli, and Werner Graber, Institute of Anatomy, University of Bern, Bern, Switzerland, for access to the SEM and their kind help with mounting specimens and taking micrographs. I am most grateful to my colleagues Elsa Obrecht and Christian Kropf, both of the Natural History Museum, Bern (NMBE), for useful comments on the manuscript. Finally, I thank the following persons and institutions for the loan of specimens: John S. Noyes, The Natural History Museum, London, UK (BMNH); John T. Huber, Canadian National Collection of Insects and Arachnids, Ottawa ON, Canada (CNCI); Csaba Thuróczy, Hungarian Natural History Museum, Budapest, Hungary (HNHM); Da-Wei Huang and Hui Xiao, Institute of Zoology, Chinese Academy of Sciences, Beijing, China (IZAS); Ivan Löbl, Muséum d’histoire naturelle, Geneva, Switzerland (MHNG); Stefan Schödl, Naturhistorisches Museum, Vienna, Austria (NMW); Lars-Åke Janzon, Naturhistoriska Riksmuseet, Stockholm, Sweden (NRS); Kazuaki Kamijo, Laboratory of Systematic Entomology, Hokkaido University, Sapporo, Japan (SEHU); E. Eric Grissell, National Museum of Natural History, Washington DC, USA (USNM); Zdenæk Bouºek, Flackwell Heath, UK (ZB).
References Askew, R. R. (1980) The European species of Coelopisthia (Hymenoptera: Pteromalidae). Systematic Entomology 5: 1-6. Boucˇek, Z. (1988) Australasian Chalcidoidea (Hymenoptera) – A Biosystematic Revision of Genera of Fourteen Families, with a Reclassification of Species. C.A.B. International, Wallingford. Boucˇek, Z. & Heydon, S. L. (1997) Pteromalidae. pp. 541-692. In Gibson, G. A. P., Huber, J. T. & Woolley, J. B. (Eds) Annotated Keys to the Genera of Nearctic Chalcidoidea (Hymenoptera). NRC Research Press, Ottawa. Boucˇek, Z. & Rasplus, J.-Y. (1991) Illustrated Key to West-Palearctic Genera of Pteromalidae (Hymenoptera: Chalcidoidea). Institut National de la Recherche Agronomique, Paris. Dzhanokmen, K. A. (1978) Hymenoptera III. Chalcidoidea 5. Pteromalidae. Opredelitel’ Nasekomikh Evropeyskoy Chasti SSSR57-228 [in Russian]. Gibson, G. A. P. (1997) Morphology and terminology. pp. 16-44. In Gibson, G. A. P., Huber, J. T. & Woolley, J. B. (Eds) Annotated Keys to the Genera of Nearctic Chalcidoidea (Hymenoptera). NRC Research Press, Ottawa. Graham, M. W. R. de V. (1956) A revision of the Walker types of Pteromalidae (Hym., Chalcidoidea). Part 2 (including descriptions of new genera and species). Entomologist’s Monthly Magazine 92: 246-263. Graham, M. W. R. de V. (1969) The Pteromalidae of North-Western Europe. Bulletin of the British Museum (Natural History) Entomology, Supplement 16: 1-908. Grissell, E. E. & Schauff, M. E. (1990) A Handbook of the Families of Nearctic Chalcidoidea (Hymenoptera). Entomological Society of Washington, Handbook No. 1.
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 171
Huber, J. T., Eveleigh, E., Pollock, S. & McCarthy, P. (1996) The chalcidoid parasitoids and hyperparasitoids (Hymenoptera: Chalcidoidea) of Choristoneura species (Lepidoptera: Tortricidae) in America North of Mexico. Canadian Entomologist 128: 1167-1220. Maddison, W. P. & Maddison, D. R. (1992) MacClade: Analysis of Phylogeny and Character Evolution. Version 3.07. Sinauer Associates, Sunderland, Massachusetts. Noyes, J. S. (1982) Collecting and preserving chalcid wasps (Hymenoptera: Chalcidoidea). Journal of Natural History 16: 315-334. Noyes, J. S. (1998) Catalogue of the Chalcidoidea of the World. Biodiversity Catalogue Database and Image Library CDrom series. ETI Biodiversity Center, Amsterdam. Peck, O. (1963) A catalogue of the Nearctic Chalcidoidea (Insecta: Hymenoptera). Canadian Entomologist, Supplement 30: 1-1092. Swofford, D. L. (1993) PAUP: Phylogenetic Analysis Using Parsimony. Version 3.1.1. Illinois Natural History Survey, Champain. Wallace, G. E. (1973) New Pteromalidae of the Dibrachys group (Hymenoptera: Chalcidoidea) with a key to genera. Annals of the Carnegie Museum 44: 171-181. Yang, Z. (1996) Parasitic Wasps on Bark Beetles in China. Science Press, Beijing. Yeates, D. K. (1995) Groundplans and exemplars: paths to the tree of life. Cladistics 11: 343-357.
Hannes Baur 172
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 1 List of material used in the analysis. Information on each species is arranged as follows: species, number of females, distribution, depository (N = Nearctic, P = Palearctic; see acknowledgements for abbreviation of acronyms). Coelopisthia Foerster areolata Askew [>10; P – Austria, Czech Rep., Switzerland; MHNG, NMBE, NMW, ZB] sp. B = sp. indet. [>10; N – Arkansas, Tennessee; CNCI, USNM] bicarinata Girault [3; N – Maryland, Montana; USNM] sp. C = sp. indet. [3; N – Florida, Virginia; USNM] caledonica Askew [>10; P – Scotland, Switzerland; BMNH, NMBE] sp. D = sp. indet. [2; P – Algeria; MHNG] extenta (Walker) [>10; P – Germany, England, Switzerland; BMNH, NMBE, NMW] sp. F = sp. indet. [3; P – Nepal, China; CNCI, IZAS] fumosipennis Gahan [10; N – Ontario, Florida, Montana; CNCI, USNM, ZB] sp. L = sp. indet. [6; P – Japan; SEHU] pachycera Masi [>10; P – England, Italy, Switzerland; BMNH, MHNG, NMBE] suborbicularis (Provancher) [>10; N – Canada, USA; CNCI, USNM, ZB] Conomorium Masi amplum (Walker) [10; P – Italy, Switzerland; MHNG, NMBE] patulum (Walker) [9; P – Sweden, Switzerland; NMBE, NRS] sp. indet. [2; P- Morocco; MHNG] Cyclogastrella Bukowskii clypealis Bouºek [>10; P – France, Morocco, Switzerland; NMBE] simplex (Walker) [= deplanata (Nees)] [8; P – Switzerland; NMBE] flavius (Walker) [4; P – Switzerland; NMBE] Dibrachoides Kurdjumov dynastes (Foerster) [9; P – Morocco, Sweden; NMBE, NRS] Dibrachys Foerster affinis Masi [4; P – Serbia; MHNG] boarmiae (Walker) [6; P – Switzerland; NMBE] cf. braconidis (Ferrière & Faure) [7; P – Switzerland; NMBE] cavus (Walker) [6; Cosmopolitan; Switzerland; NMBE] confusus (Girault) [2; N – Wisconsin; USNM] pelos Grissell [2; N – Connecticut; USNM] Diglochis Foerster sp. A = sp. indet. [1; P – Hungary; HNHM] sp. B = sp. indet. [1; P – Hungary; HNHM] occidentalis (Ashmead) ) [2; N – Colorado; USNM] sylvicola (Walker) [8; P – Switzerland, Hungary; HNHM, MHNG, NMBE]
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 173
Duartea Boucˇek daphne (Girault) [3; N – California; BMNH, USNM] Janssoniella Kerrich sp. indet [1; P – Switzerland; NMBE] Pachyneuron Walker formosum (Walker) [>10; P – England, Switzerland; MHNG, NMBE] muscarum (L.) [5; P – Switzerland; NMBE] Plutothrix Foerster bicolorata (Spinola) [2; P – Switzerland; NMBE] obtusiclava Graham [1; P – Switzerland; NMBE] Rhicnocoelia Graham constans (Walker) [8; P – Switzerland; NMBE] Rhopalicus Foerster tutela (Walker) [>10; P – Switzerland; NMBE] Schizonotus Ratzeburg sieboldi Ratzeburg [6; P – Hungary, Switzerland; HNHM, NMBE] Stichocrepis Foerster armata Foerster [3; P – Hungary; HNHM] Systellogaster Gahan gahani Wallace [2; N – Montana; USNM] ovivora Gahan [3; N – USA, Vienna, Va; USNM] Tritneptis Girault doris Burks [4; N – Arizona; USNM] hemerocampae Girault [3; N – New Jersey, Connecticut; USNM] sp. indet = ?klugii (Ratzeburg) [6; P – Switzerland; NMBE]
Hannes Baur 174
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 2 Characters and character states († and †† refer to SEM micrographs published in Bouºek & Rasplus (1991) and Huber et al. (1996), respectively; see comments in text for asterisks). 1.
*Level of toruli (frontal view with lower edge of median ocellus and of toruli equidistant to the objective): upper edge below (0); upper edge at or above (1); lower edge at or above ocular line (2).
2.
Mouth margin before conjunction with malar sulcus: straight with weak fovea above (0) (Fig. 4); curving with weak fovea above (1) (Fig. 3); curving with fovea reaching halfway along malar sulcus (2).
3.
Occipital carina: absent (0) (Fig. 5); weakly developed laterally, more or less effaced medially, curving, high (1) (Fig. 6); distinct throughout, curving, high (2) (Fig. †29); distinct throughout, straight, low (3) (Fig. †30); weakly developed medially along edge formed by vertex and occiput (4); as 1 but traceable medially, longer, and more curving (5); absent but distinct edge in the middle quarter formed by vertex and occiput (6); distinct throughout, curving, low (7).
4.
First anellus: laminate edge on distal end absent (0) (Fig. 7); laminate edge on distal end present (1) (Fig. 8).
5.
Third flagellar segment, shape: conical (0); cylindrical (1); with constricted neck at base (2).
6.
Third flagellar segment, arrangement of longitudinal sensilla: absent (0); in one row (1); in two or more rows (2).
7.
Third flagellar segment, distribution of longitudinal sensilla: distributed over distal two thirds to base (0); confined to distal half (1); confined to distal quarter (2).
8.
Micropilosity on clava: present on third segment only (0); present on all segments (1); present on second and third segment (2).
9.
Pronotal collar: bluntly ridged medially (0); sharply carinate medially (1); evenly rounded (2).
10. Position of marginal and postmarginal vein: right along wing margin (0); slightly removed from wing margin (1). 11. *Length of postmarginal vein: about as long as stigmal vein (0); distinctly longer than stigmal vein (1). 12. Plica, anterior (AP) and posterior part (PP): AP sharp, moderately curving, reaching halfway along propodeum (PPD), PP hardly traceable (0); like 0 but PP distinctly developed (Fig. 9) (1); AP sharp, strongly curving, often reaching adpetiolar strip, PP distinct, sometimes reaching AP or spiracles (Fig. 10, ††20) (2); AP and PP sharp and smoothly joining to form a moderately sinuate edge, AP sometimes forming a costula (3) (Fig. 12); like 2 but AP pointing backwards (4) (Fig. 11); AP blunt, weakly curving, PP distinct (5) (Fig. ††21); AP hardly developed, PP distinct (6); like 6 but PP almost reaching spiracles (7); AP sharp, strongly bent inwards and joining median carina at base, PP distinct (8). The above anterior and posterior parts of plica are collectively referred to as one ‘plica’ or ‘plical carina’ in all major textbooks (e. g. Graham 1969; Boucek 1988; Gibson 1997). However, close examina-
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 175
tions revealed that it consists of an anterior and a posterior part. Even where both parts form a seemingly continuous edge (Fig. 12), a small break was always detectable. On the other hand, where a cross carina known as ‘costula’ was indicated, that was formed by the anterior part of plica curving inwards. 13. Median carina of PPD: straight, distinct throughout (0) (Figs 11, 12); irregular and/or effaced medially (1) (Figs 9, 10). 14. Pilosity on callus of PPD: reaching below spiracles (0); not reaching below spiracles (1). 15. Petiole surface structure: alutaceous to weakly reticulate without median carina (0); smooth with indistinct median carina (1); smooth with strong median carina (2). 16. Proximal edge of gastral tergum one: lateral laminate borders fused medially (0); borders just touching medially (1); borders separate (2). 17. Placement of cerci on metasomal tergum eight: ventro-laterally (0); dorsally (1).
Hannes Baur 176
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 3 Data matrix of character states for 44 species of 16 genera. Taxa arranged according to their appearance on the strict consensus tree (Fig. 1). Coelopisthia areolata
1 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. sp. B
1 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. sp. F
1 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. extenta
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. bicarinata
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. sp. C
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 ?
C. fumosipennis
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
0
C. sp. L
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. suborbicularis
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0
C. caledonica
0 0 1 1 0 1 1 0 0 1 0 2 1 0 1 0 0 1
C. pachycera
1 0 1 1 0 1 1 0 0 1 0 1 1 0 1 0 0
C. sp. D
1 0 0 1 0 1 1 0 1 1 0 2 1 0 1 0 0
Diglochis occidentalis
1 0 5 0 0 0 ?
1 0 0 0 2 0 1 2 0 0
D. sp. A
1 0 5 0 0 0 ?
1 0 0 0 4 1 1 2 0 0
D. sp. B
1 0 5 0 0 0 ?
1 0 0 0 4 1 1 2 0 0
D. sylvicola
1 0 5 0 0 0 ?
1 0 0 0 4 0 1 2 0 0
Conomorium amplum
1 1 0 0 2 1 2 0 0 0 0 0 0 1 0 0 0
C. patulum
2 1 0 0 2 1 2 0 0 0 0 0 0 1 0 0 0
C. sp.
1 1 0 0 2 1 2 0 0 0 0 0 0 1 0 0 0 2 1
Cyclogastrella clypealis
2 1 0 0 0 0 ?
0 2 0 0 3 0 0 1 0 0 1
C. flavius
2 1 0 0 0 0 ?
0 2 0 0 3 0 0 1 0 0 1
C. simplex
1 1 0 0 0 0 ?
0 2 0 0 3 0 0 1 0 0 1
Dibrachys confusus
1 1 2 0 1 1 0 0 2 0 0 3 0 1 1 2 0
Dibrachoides dynastes
2 2 4 0 1 1 0 0 1 0 0 5 1 1 1 1 0
Dibrachys affinis
2 1 3 0 1 1 0 0 0 1 0 3 1 1 1 2 0
D. boarmiae
2 1 2 0 1 1 0 0 2 1 0 5 1 1 1 2 0
D. cavus
2 1 2 0 1 1 0 0 2 1 0 5 1 1 1 2 0
D. braconidis
2 1 2 0 1 1 0 0 2 0 0 5 1 1 1 2 0
D. pelos
1 1 2 0 1 1 0 0 2 1 0 5 1 1 1 2 0
Duartea daphne
2 2 7 0 1 1 0 2 2 0 0 5 1 1 1 2 0
Schizonotus sieboldi
2 2 0 0 1 1 0 0 2 1 0 3 0 1 1 1 1 1
Stichocrepis armata
0 1 0 0 1 1 0 0 2 1 0 8 1 1 1 0 0
Systellogaster ovivora
2 1 6 0 1 1 0 0 2 1 0 5 1 1 1 2 0
S. gahani
2 1 6 0 1 1 0 0 2 1 0 5 1 1 1 2 0
Tritneptis doris
0 1 0 0 1 1 0 0 2 1 0 3 1 1 1 2 0
Monophyly and Relationship of the Genus Coelopisthia Förster (Chalcidoidea: Pteromalidae) 177
T. hemerocampae
1 1 0 0 1 1 0 0 2 1 0 5 1 1 1 2 0
T. sp.
1 1 0 0 1 1 0 0 2 1 0 3 1 1 1 2 0
Janssoniella sp.
2 1 0 0 1 2 0 0 2 0 1 7 0 1 1 2 0
Rhicnocoelia constans
2 1 0 0 1 2 0 0 2 0 1 7 0 1 1 ?
Plutothrix bicolorata
2 1 0 0 1 2 0 0 1 0 1 7 0 1 1 2 0
P. obtusiclava
2 1 0 0 1 2 0 0 1 0 1 7 0 1 1 2 0
Rhopalicus tutela
2 1 0 0 1 2 0 0 2 0 1 6 0 1 1 2 0
0
Pachyneuron formosum
2 1 0 0 1 1 0 0 1 0 1 5 1 1 0 2 1
P. muscarum
2 1 0 0 1 1 0 0 1 0 1 5 1 1 0 2 1
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae): Endoparasitoids of Spider Eggs Muhammad Iqbal and Andrew D. Austin Department of Applied and Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected])
Introduction The Scelionidae is a highly speciose family of small parasitic wasps that exclusively oviposit into the eggs of their hosts. They have exploited most orders of insects and spiders, and show a high level of host group specificity in that most tribes of scelionids are associated with specific host groups, e.g. Orthoptera, Mantodea, Heteroptera, Embioptera, Coleoptera, Lepidoptera (Austin & Field 1997). One tribe, the Baeini, many of which are less than 1 mm in length (Fig. 1), are obligate endoparasitoids of the eggs of spiders and are the only group of scelionids to utilise this host group (Austin 1985). They are ubiquitous in most habitats and are postulated to be important regulating agents of spider populations (Austin 1984). The tribe is one of the most easily recognised groups of scelionid wasps, because of the strongly clavate antenna in females, and the often bizarre, elongate metasomal horn of some genera, viz. Odontacolus Kieffer and Ceratobaeus Ashmead (Figs 2, 3). Taxonomically, baeine wasps are well-known at the generic level (e.g. Galloway & Austin 1984), but the species are poorly documented with only 10–20% being described for Australasia (Iqbal & Austin 1997) and less than this on a world-wide basis. The monophyly of the Baeini is supported by several characters, none of which are unequivocal. These include the terminal four segments of the female antenna being fused into a compact clava, the mandible being tridentate, and the antennal scape not reaching to the level of the vertex (Austin & Field 1997). A number of genera are putatively monophyletic and have identifiable synapomorphies: for instance Baeus Haliday and Mirobaeoides Dodd are apterous, with a fusiform body and greatly reduced metanotum (Figs 1 & 8); Hickmanella Austin has long, sparse pilosity covering the body; while Odontacolus and Cyphacolus Priesner have a laterally compressed metasomal horn. However, the monophyly of most genera is questionable, particularly the two largest genera, Idris Foerster and Ceratobaeus, which comprise the majority of described species. To date, the Baeini have not been exposed to phylogenetic analysis. The present study arose from a detailed taxonomic revision of Australasian Ceratobaeus (Iqbal & Austin 2000). Previous authors have treated this genus synonymously with or as a subgenus of Idris (Huggert 1979; Johnson 1992; Masner & Denis 1996). Although Ceratobaeus is easily recognised by the development of a hump or cylindrical horn on the first metasomal tergite, it is also possible that this character has arisen independently within several lineages of Idris. This study aimed to test the monophyly of Ceratobaeus in relation to other genera of Baeini, and generate a preliminary phylogeny for the tribe that could be used to support future work on the systematics, zoogeography and host relationships of the group. Note that the authors of species referred to in the text are given in Appendix 2.
178
Hymenoptera: Evolution, Biodiversity and Biological Control
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 179
Materials and Methods Material examined and terminology Specimens for SEM were cleaned in a dilute pure soap solution (5%), rinsed in distilled water and dehydrated in an alcohol series. They were then mounted on card-points with water based seccotine glue; cards were secured to SEM stubs with carbon-based plasticine (Leitz-C-Plast), coated with gold to a thickness of 300–400 angstroms, and examined under a Phillips XL30 field emission SEM using secondary electron imaging at 10 kv. The external morphology of scelionids is outlined in Masner (1979, 1980), Galloway & Austin (1984), and Masner and Denis (1996). Terminology and measurements for body parts and wing venation are detailed in Iqbal (1998) and Iqbal and Austin (2000). Abbreviations used in the text are as follows: ANIC, Australian National Insect Collection, Canberra; CNCI, Canadian National Collection of Insects, Ottawa; WARI, The University of Adelaide Collection, Adelaide. Phylogenetic analysis Exemplar taxa from all recognised baeine genera, including the type-species of some of these genera, comprised the in-group. Species of Ceratobaeus and Idris were selected to represent at least part of the apparent morphological variation evident with these two very large genera. Seventeen species of Ceratobaeus, eight species of Idris, four species each of Baeus, Mirobaeoides, Odontacolus, two species of Hickmanella, and one species each of Mirobaeus Dodd, Neobaeus Austin, Apobaeus Masner, Anabaeus Oglobin, Cyphacolus and Echthrodesis Masner, were included (voucher material in WARI). Five additional species were also included that could not be reliably placed into existing genera in an attempt to determine their phylogenetic status. These species are referred to as Genus 1–5, of which Genus 1–3 are from Australia (voucher material in ANIC, WARI), and Genus 4 and 5 are from South Africa (voucher material in CNCI and WARI). Four out-group taxa were employed separately for initial parsimony analyses. These were Sparasion L. and Nixonia Masner, which are plesiomorophic and postulated to be basal to most if not all other scelionid genera, and Embidobia Ashmead and Gryon Haliday which represent members of putative sister tribes (Masner 1976; Austin & Field 1997). Thirty-five morphological characters were selected as potentially informative (Appendix 1). Characters were scored only for females as this sex displays substantially more morphological variability than males. Males from different genera are very similar to each other and generally lack characters informative at higher taxonomic levels (Galloway & Austin 1984). Further, it has not been possible to associate the sexes for most species of Baeini. Where possible, characters were divided into binary states to avoid hierarchical linkage and the problem of scoring of nonapplicable states. The morphometric characters 32, 33, and 34 were coded using gap coding while character 35 was coded using segment coding (Chappill 1989). Gap coding was used when there were obvious gaps present in the character distribution among taxa. The procedure adopted for gap coding was as follows: measurements were undertaken for a particular character and the mean value calculated for each species. If only a single specimen was available its measurement was used in place of the mean. A graph of mean values for all taxa was then prepared, significant gaps identified in the distribution of means (see Iqbal 1998), and different states assigned to taxa by considering these gaps as boundaries to separate states. For continuous data, segment coding was adopted. The mean values of measurements were arranged in ascending order and then one standard deviation value was added to the minimum mean value in the table. All taxa equal or less than the added value were coded with the one state. This process was continued until the last mean value was assigned to a segment.
Muhammad Iqbal and Andrew D. Austin 180
Hymenoptera: Evolution, Biodiversity and Biological Control
PAUP (Phylogenetic Analysis Using Parsimony) version 3.1.1 (Swofford 1993) and various test versions of PAUP* 4 for Power Macintosh and UNIX (Swofford 1998) were used for all parsimony-based analyses, while MacClade 3.07 (Maddison & Maddison 1997) was employed to input the data matrix (Appendix 2) in spreadsheet format and to trace character distribution on trees. AutoDecay 2.9.8 (Eriksson 1997) was used to calculate decay values and the resulting trees were viewed and printed using TreeView 1.4 (Page 1997). Due to the large size of the data matrix only heuristic searches were undertaken, using the Random Addition Sequence and TBR branch swapping options in PAUP. Uninformative characters (constant and autapomorphic) were excluded before analysis. Characters were weighted equally irrespective of the number of states. Ten trees were held at each step to minimise the effect of ties early in the stepwise addition process. Trees were rooted using the out-group taxa discussed below, so that the in-group was always monophyletic. Bootstrap (Felsenstein 1985) and Bremer support (Bremer 1994) analysis were undertaken as a measure of tree fitness. Bootstraps were calculated by using 10 000 replicates and saving a single tree in each bootstrap using ‘fast’ swap option of PAUP* 4. This strategy was used to decrease analysis times, and is likely to have underestimate bootstrap support. As such, it is a conservative measure of clade support. To explore the effect of grouping particular species or characters, they were constrained using PAUP* and the corresponding tree lengths and topology compared.
Results Analyses using either Sparasion or Nixonia as the out-group were unsatisfactory given they reached a tree buffer overflow and so there was no certainty that the most parsimonious trees had been attained. Those derived from Embidobia as the out-group did reach a parsimonious solution (length 151 steps) but the resulting strict consensus tree (not shown here) was largely comb-like and uninformative for discussing relationships among the Baeini. The reason for this is not clear, given that the Embidobiini is the postulated sister tribe to the Baeini (Austin & Field 1997). However, possibly this genus is too derived within the tribe and convergently shares too many character states with some members of the in-group, a problem that could only be overcome by expanding the number of informative characters and/or including a wider range of taxa from the Embidobiini. Analysis of the data-set using Gryon as the out-group generated 80 shortest trees of 147 steps (CI = 0.25; RI = 0.66). The resulting strict consensus tree (Fig. 9) is relatively well-resolved but few clades are supported by unequivocal characters states and, accordingly, bootstrap and Bremer support is low for most branches on the tree. However, the results serve as a framework to discuss a number of postulated relationships and as a basis for further studies. Significantly, there is no support for the monophyly of Idris s. str. or for Idris s. l. (i.e. with the inclusion of Ceratobaeus spp.). Species of Idris s. str. fall out in several major clades resolved in this analysis, together with other baeine genera (e.g. clades 2, 4 and 12: Fig. 9), or separately in the two clades below node 7 (i.e. Idris theridii, I. seminitidus and I. helpidid). Members of Ceratobaeus are mostly contained within clade 8 which is supported by two homoplasious characters; T1 having a metasomal horn (Character 21:1) and humeral sulcus being crenulate (character 11:1). However, two species fall outside this clade: C. setosus is the sister to all reduced-winged and apterous baeines (clade 12), while Ceratobaeus sp. 1 is the sister to Idris sp. 2 + Hickmanella (clade 2). The position of C. intrudae is unclear in that this species forms a tricotomy with clades 8 and 12. Constraining all Ceratobaeus, all Idris s. str. and all Idris s. str. + Ceratobaeus (i.e. Idris s. l.)
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 181
Figures 1-8
Micrographs showing different characters in Baeini: 1) Baeus sp., female showing attachment of metasoma and aptery; 2) Ceratobaeus flavipes, female showing wing flaps and metasomal horn; 3) Ceratobaeus leai, female showing wing flaps and metasomal horn; 4) Genus 1, female showing metasomal horn; 5) Genus 3, female showing wing pads; 6) Genus 5, female showing attachment of metasoma; 7, 8) Mirobaeoides pecki: 7) female showing hind leg femoral spines; 8) male showing wing development. Scale lines: 1, 2, 4–6, 8 = 100 µm; 3 = 200 µm; 7 = 50 µm.
Muhammad Iqbal and Andrew D. Austin 182
Hymenoptera: Evolution, Biodiversity and Biological Control
requires 3, 4 and 5 additional steps, respectively, indicating that none of these hypotheses can be supported over the shortest tree depicted in Figure 9. However, these analyses all resulted in a tree buffer overflow and so there is no guarantee that the shortest trees were obtained. Hickmanella (clade 3) is supported by one unequivocal character state, the frontal carina reaching the median ocellus (character 2:2), while Hickmanella + Ceratobaeus sp. 1 and Idris sp. 2 (clade 2) is supported by having five funicle segments (character 5:0) and the antennal clava segmented (character 7:0). However, these latter two states are also found independently in many of the apterous species in clade 16. Odontacolus is monophyletic but only with the inclusion of Cyphacolus sp. (clade 5) and, together, these taxa are supported by the unequivocal character state, metasomal horn compressed laterally (character 25:1). Surprisingly, the apterous taxon Genus 3 is the sister to I. pulcher + Odontacolus + Cyphacolus, but there is little or no support for this placement within the tree. All reduced-winged female Baeini, except for C. leai, Ceratobaeus sp. 2 and Genus 3, are contained within a single large clade (12), which includes seven currently recognised genera (viz. Anabaeus, Apobaeus, Baeus, Echthrodesis, Mirobaeoides, Mirobaeus and Neobaeus) as well as Genus 1, 2, 4 and 5, and I. flavicornis and Idris sp. 1. Within this group, the five basal taxa, I. flavicornis, Idris sp. 1 and Genus 2, 4 and 5 (except Anabaeus sp.) are brachypterous in that the greatly shortened wings are still membranous at least in part (as in Figs 2, 3), while those species contained within clade 16 are supported by the state of being apterous (i.e. having the wings reduced to minute sclerites (character 13:1; Figs 1, 4, 8)). Of the two genera for which multiple species have been included, Baeus and Mirobaeoides, neither are resolved as monophyletic in the strict consensus tree. However, both are monophyletic in a majority of the 80 most parsimonious trees, where M. pecki is the sister to other Mirobaeoides spp. or to Neobaeus + Baeus + Apobaeus (clade 17) , and Apobaeus is the sister to Baeus or the latter is rendered paraphyletic. Interestingly, Neobaeus is sister to these two genera (clade 18), a relationship which is supported by the metasomal laterotergites being free (i.e. not inserted into a submarginal groove) (character 31:0), however this state is reversed in Apobaeus. Significantly, Mirobaeoides is the only baeine genus known where the male is also apterous, a condition associated with it inhabiting a small oceanic island (Lord Howe Is; Fig. 8). The presence of hind femoral spines (character 29; Fig. 7) is a putative synapomorphy for Mirobaeoides (Austin 1986), however in the analysis this character is homoplasious as it also occurs in Genus 2 and 3, at the base of clades 14 and 4, respectively. Several characters previously used to diagnose genera were constrained to determine their effect on both tree length and topology. These were wing reduction in females (characters 13, 14; Figs 1–6, 8), presence of hind femoral spines (character 29; Fig. 7), presence of a metasomal hump or horn on the first metasomal tergite (characters 2–4; Figs. 2–4), and the antennal clava being fused (i.e. unsegmented clava; character 6). In all cases parsimony analysis generated trees 2-8 steps longer than the shortest tree shown in Figure 9 (Table 1). The tree which was only two steps longer resulted from constraining apterous and brachypterous species into a monophyletic clade, which is not surprising given that, except for three species, they already form a monophyletic group in the most parsimonious tree (clade 12). The longest tree of eight additional steps resulted from constraining all taxa with a metasomal hump or horn. These analyses had variable affects on tree topology but did not result in any relationships among taxa, other than those within the constrained clade, that are not discussed above. Wing reduction is often homoplasious in insects associated with soil and leaf-litter, as is the case for many baeines (Iqbal & Austin 1997). However in the most parsimonious trees obtained here, most
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 183
X31:1
Apobaeus sp. Baeus leai Baeus seminulum Baeus machodoi Baeus saliens Neobaeus novazealandensis X35:1 17 Mirobaeoides barbarae =29:1 Mirobaeoides scutellaris 2 Mirobaeoides tasmanicus =29:1 Mirobaeoides pecki =13:1 Echthrodesis sp. X35:1 16 Mirobaeus bicolor X35:1 Genus 1 =29:1 X35:1 Genus 2 Genus 4 Genus 5 Anabaeus sp. Idris flavicornis Idris sp. 1 Ceratobaeus setosus =18:0 Ceratobaeus cuspicornutus X11 :0 11 Ceratobaeus sp. 6 =34:1 =19:1 Ceratobaeus reiki 10 Ceratobaeus mirabilis 2 Ceratobaeus fasciatus Ceratobaeus cornutus =28:1 2 Ceratobaeus sp. 5 2 Ceratobaeus sp. 4 Ceratobaeus sp. 2 9 =14:1 =19:1 X11: 0 2 Ceratobaeus leai Ceratobaeus laeviventris Ceratobaeus giraulti Ceratobaeus sp. 3 Ceratobaeus sp. 7 Ceratobaeus intrudae Idris helpidid Idris seminitidus Idris theridii =15:1 Odontacolus sp. 1 50 6 2 25:1 Cyphacolus sp. =21:1 2 Odontacolus longiceps 5 =28:1 =11 :1 Odontacolus sp. 2 X4:1 2 Odontacolus sp. 3 4 Idris pulcher =1:1 =14:1 =29:1 2:2 =28:1 X8:0 Genus 3 75 Hickmanella holoplatysa =5:0 =15:1 3 3 Hickmanella intrudens =7:0 =18:0 2 2 Idris sp. 2 2 Ceratobaeus sp. 1 Idris niger Gryon sp. X7:1 X5:1 =20:1 19 X31:0 18 2
X5:0 X7:0
=1:1 X35:0 =10:1 15 =19:1 =14:1
14
2
12 =10:1
13 =18:0
7
8
2
=28:1
2
=10:1
1
Figure 9
X4:1 X8:1
=21:1 X11: 1
=13:1
Strict consensus of 80 most parsimonious trees (length 147; CI = 0.25; RI = 0.66). Bootstrap values are given above and Bremer (decay) values below the nodes (N.B. only Bremer values ≤ 2 are shown). • unequivocal synapomorphy; = homoplasy; X reversal.
Muhammad Iqbal and Andrew D. Austin 184
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Effect of constraining particular morphological characters and species (L = tree length; N = number of trees saved; CI = consistency index; RI = retention index; RC = rescaled consistency index). Constraint
L
N
CI
RI
RC
f-ratio
Unconstrained (Gryon sp. out-group)
147
80
0.25
0.66
0.16
0.73
Ceratobaeus spp. constrained
150
o/flow
0.24
0.65
0.16
0.71
Idris spp. constrained
151
o/flow
0.24
0.66
0.16
0.41
Idris + Ceratobaeus spp. constrained
152
o/flow
0.24
0.65
0.16
0.58
Apobaeus position constrained
149
64
0.25
0.66
0.16
0.78
Wing reduction constrained
149
o/flow
0.24
0.66
0.16
0.46
Femoral spines constrained
150
57002
0.24
0.65
0.16
0.49
Unsegmented clava constrained
151
o/flow
0.24
0.65
0.16
0.53
Metasomal horn constrained
155
o/flow
0.23
0.64
0.15
0.88
Exclusion of characters 13 & 14
141
98696
0.24
0.65
0.16
0.56
apterous and brachypterous species are found within a single clade. To examine whether wing reduction (characters 13 and 14) was over-riding otherwise informative characters in the analysis, the wing reduction characters were excluded and the data matrix reanalysed. This resulted in over 98,000 shortest trees (Table 1), and the strict consensus of these was largely comb-like in structure (not shown here), particularly among the species contained within clade 12, indicating that wing reduction is not over-riding any other informative pattern. Also evident from this result is that much of the structure in the tree is reliant upon the inclusion of these characters.
Discussion Although the results presented here are preliminary in that the number of characters scored is relatively low compared with the number of taxa, they do confirm the monophyly (or near monophyly) of several of the smaller baeine genera and the polyphyly of the two largest genera, Idris and Ceratobaeus. Even though the relationships among many species included in this study would likely change with the addition of new characters and/or taxa, it also seems probable that neither of the latter two genera are ever likely to be resolved as monophyletic, given that no unequivocal synapomorphies have been yet found for them. Further, the synonymy of Ceratobaeus with Idris (sensu Huggert 1979) or its reduction in rank to subgeneric level within Idris s. l. (sensu Johnson 1992; Masner & Denis 1996) is no more tenable than recognising these genera as separate. There seems little or no advantage in making such generic-level changes until a more complete understanding of relationships within the tribe can be achieved. Hence, for the time being we would advocate that it is better to treat Idris and Ceratobaeus as separate genera for
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 185
reasons of taxonomic convenience, i.e. they are both very speciose genera and can be readily identified using existing keys (e.g. Masner 1976, 1980; Galloway & Austin 1984). The presence of a hump or horn on the first metasomal tergite is a diagnostic character for Ceratobaeus and Odontacolus/Cyphacolus, but some dorsal expansion of this tergite is also known in at least one species of Mirobaeoides (M. barbarae) and in Genus 1, 2 and 4. Given that this structure occurs independently in a number of platygastroid genera (see Austin & Field 1997), it is also likely to have evolved independently within separate lineages of Baeini. In this tribe, the presence of a metasomal horn, which acts as a recess for a longer ovipositor, has been associated with species that oviposit through the silk wall of spider egg-sacs (Austin 1985). Several relationships in this analysis seem worthy of immediate study. The position of the single representative of Cyphacolus within Odontacolus and the fact that this group is supported by a strong, unequivocal synapomorphy (i.e. laterally compression of the metasomal horn) indicates that the status of the former genus should be reconsidered. This will require an analysis including a much larger number of species, but already it seems from examination of the available Old World fauna (about 20 mostly undescribed species for both genera) that Cyphacolus represents a derived species-group within Odontacolus which is defined by strongly patterned wings, and the marginal and stigmal veins being spectral, at least in part. Also worthy of future study is the relationships among taxa in clade 17 (comprising Apobaeus, Baeus, and Mirobaeoides) with Genus 1 and 2 and, related to this, the distribution of two character states, presence of free lateral tergites and hind femoral spines. Although neither character unequivocally defines a monophyletic group in the present study, they would seem to have significant phylogenetic potential. Within the Baeini, free lateral tergites and the corresponding absence of a metasomal submarginal groove occurs only in Baeus and Neobaeus, although this state also occurs in the Telenominae and some genera of Thoronini (e.g. Masner 1972, 1980), while the presence of hind femoral spines is unique to the tribe but apparently restricted to taxa from the Australian region (including Genus 1 and 2). Interestingly, the putative plesiomorphic state of the antennal clava being clearly divided into four segments is also unique to taxa from Australia (viz. Mirobaeoides, Mirobaeus, Hickmanella, some Ceratobaeus), although a somewhat intermediate state is found in the more cosmopolitan Odontacolus (Galloway & Austin 1984). The most concerning aspect of this analysis is the dependence on two intuitively homoplasious character states for much of the topology seen within the tree generated here. These are wing reduction/aptery and the presence of a hump or horn on metasomal tergite 1. Aptery in females is a diagnostic character for several genera, viz. Anabaeus, Apobaeus, Baeus, Echthrodesis, Mirobaeoides, Mirobaeus and Neobaeus. However, wing reduction also occurs in several species of Ceratobaeus, Idris and Genus 1–5, but is unknown in Hickmanella, Odontacolus and Cyphacolus. Given that wing reduction also occurs in many platygastroid genera, e.g. Dyscritobaeus Perkins, Probaryconus Kieffer, Trimorus Foerster, Austromerus Masner & Huggert, Parabaeus Kieffer, Platygastoides Dodd (Galloway & Austin 1984; Masner & Huggert 1989; Austin & Field 1997; Austin unpublished) and is undoubtedly an adaptation for living in soil and leaf-litter and/ or is associated with island faunas, it seems very probable that this trait has evolved within several baeine lineages that have independently radiated into spider hosts that live in these habitats. However, in the absence of other informative characters, wing reduction and aptery support an unconvincing monophyletic group in the analysis conducted here. Also related to this, but representing a wider problem for the taxonomy of baeines, is what to do with the significant number of new taxa that cannot be assigned easily to any existing baeine genera (here represented by the
Muhammad Iqbal and Andrew D. Austin 186
Hymenoptera: Evolution, Biodiversity and Biological Control
brachypterous/apterous species Genus 1–5)? It would not seem sensible to describe them as new genera and proliferate the number of monospecific genera in the tribe, although using current generic concepts this could be justified. An alternative strategy might be to describe them as new species-groups of Idris. This approach at least makes the descriptions formally available so the species can be assessed more widely in the future. However, it would broaden the limits of this genus to the point where it was almost synonymous with the limits of the tribe.
Acknowledgements We wish to thank Paul Dangerfield, Lubomir Masner and Mark Harvey for their help during this project. We also wish to acknowledge the assistance of CEMMSA and ICC staff at Adelaide University for help with SEM techniques and photography, respectively. We are grateful for the financial support provided by AusAID to M.I. and Adelaide University to A.D.A.
References Austin, A. D. (1984) The fecundity, development and host relationship of Ceratobaeus spp. (Hymenoptera: Scelionidae), parasites of spider eggs. Ecological Entomology 9: 125-138. Austin, A. D. (1985) The function of spider egg sacs in relation to parasitoids and predators, with special reference to the Australian fauna. Journal of Natural History 19: 359-376. Austin, A. D. (1986) A taxonomic revision of the genus Mirobaeoides Dodd (Hymenoptera: Scelionidae). Australian Journal of Zoology 34: 315-337. Austin, A. D. & Field, S. A. (1997) The ovipositor system of scelionid and platygastrid wasps (Hymenoptera: Platygastroidea): comparative morphology and phylogenetic implications. Invertebrate Taxonomy 11: 1-87. Bremer, K. (1994) Branch support and tree stability. Cladistics 10: 295-304. Chappill, J. A. (1989) Quantitative characters in phylogenetic analysis. Cladistics 5: 217-234. Eriksson, T. (1997) AutoDecay, Version 2.9.8. Computer program distributed by the author, Botaniska Institutionen, Stockholm University, Stockholm. Felsenstein, J. (1985) Confidence limits on phylogenies: an approach using bootstrap. Evolution 39: 783-791. Galloway, I. D. & Austin, A. D. (1984) Revision of the Scelioninae (Hymenoptera: Scelionidae) in Australia. Australian Journal of Zoology, Supplementary Series 99: 1-138. Huggert, L. (1979) Revision of the west Palaearctic species of the genus Idris Foerster, s.l. (Hymenoptera, Proctotrupoidea: Scelionidae). Entomologica Scandinavica, Supplement 12: 1-60. Iqbal, M. (1998) Systematics and phylogeny of the Baeini (Hymenoptera: Scelionidae) with special reference to Australasian fauna. Ph.D. Thesis, The University of Adelaide, Adelaide, South Australia. Iqbal, M. & Austin, A. D. (1997) Species richness and endemism of baeine wasps (Hymenoptera: Scelionidae) in Australia. Memoirs of the Museum of Victoria 56: 455-459. Iqbal, M. & Austin, A. D. (2000) Systematics of Ceratobaeus Ashmead (Hymenoptera: Scelionidae) from Australasia. Records of the South Australian Museum Monographic Series 6: 1-164. Johnson, N. F. (1992) Catalog of World species of Proctotrupoidea, Exclusive of Platygastridae (Hymenoptera). The American Entomological Institute, Gainesville.
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 187
Maddison, W. P. & Maddison, D. R. (1997) MacClade: Interactive analysis of phylogeny and character evolution. Version 3.07. Computer program distributed by Sinauer Associates, Sunderland. Masner, L. (1972) The classification and interrelationships of Thoronini (Hymenoptera: Proctotrupoidea, Scelionidae). Canadian Entomologist 104: 833-49. Masner, L. (1976) Revisionary notes and keys to world genera of Scelionidae (Hymenoptera: Proctotrupoidea). Memoirs of the Entomological Society of Canada 97: 1-87. Masner, L. (1979) Pleural morphology in scelionid wasps (Hymenoptera: Scelionidae) – an aid to higher classification. Canadian Entomologist 111: 1079-1087. Masner, L. (1980) Key to genera of Scelionidae of the Holarctic region, with descriptions of new genera and species (Hymenoptera: Proctotrupoidea). Memoirs of the Entomological Society of Canada 113: 1-54. Masner, L. & Denis, J. (1996) The Nearctic species of Idris Foerster, Part 1: The mellus-group (Hymenoptera: Scelionidae). Canadian Entomologist 128: 85-114. Masner, L. & Huggert, L. (1989). World review and keys to genera of the subfamily Inostemmatinae with reassignment of the taxa to the Platygastrinae and Sceliotrachelinae (Hymenoptera: Platygastridae). Memoirs of the Entomological Society of Canada 147: 1-214. Page, R. D. M. (1997) TreeView, Version 1.4. Computer program distributed by the author, Division of Environmental and Evolutionary Biology, University of Glasgow, Glasgow. Swofford, D. L. (1993) PAUP: Phylogenetic Analysis Using Parsimony, Version 3.1.1. Computer program distributed by Illusions State Natural History Survey, Champaign. Swofford, D. L. (1998) PAUP*: Phylogenetic Analysis Using Parsimony, Version 4.0 (PPC, test). Computer program distributed by the author, Smithsonian Institution: Washington, D. C.
Muhammad Iqbal and Andrew D. Austin 188
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 1 Characters and character states employed in the phylogenetic analysis. Character 1. Attachment of metasoma to mesosoma: 0) metasoma relatively small and not broadly abutted against mesosoma, 1) metasoma sub-sessile against mesosoma, 2) metasoma sessile against mesosoma. Character 2. Frontal carina: 0) frontal carina absent or rudimentary, 1) frontal carina fine or strong but not reaching to median ocellus, 2) frontal carina reaching to median ocellus. Character 3. Sculpturing on cheek and lower frons: 0) cheek and lower frons smooth, 1) cheek and lower frons striate. Character 4. Speculum development: 0) speculum present, 1) speculum absent (i.e. frons uniformly sculptured). Character 5. Funicle segment number: 0) five or more funicle segments, 1) four funicle segments. Character 6. Antennal clava: 0) distinct antennal clava present, 1) antennal clava absent. Character 7. Antennal clava segmentation: 0) antennal clava segmented, 1) antennal clava appearing fused. Character 8. Eyes pilosity: 0) eyes with long hairs, 1) eyes with minute hairs (visible only at high magnification, > x 80). Character 9. Eye size: 0) eyes size normal, height more than half that of head, 1) eyes small, height less than half that of head. Character 10. Axillar sculpturing: 0) axillar crenulae present, 1) axillar crenulae absent. Character 11. Humeral sulcus: 0) humeral sulcus smooth, 1) humeral sulcus crenulate. Character 12. Notauli development: 0) scutum with notauli, 1) scutum lacking notauli. Character 13. Wing development: 0) wings present (i.e. brachypterous or full-winged), 1) apterous (i.e. wings reduced to tiny sclerites, no more than about twice the length of the tegula). Note that species coded (1) for character 13, are coded as missing data for character 14-18. Character 14. Degree of wing development: 0) wings fully developed, 1) brachypterous. Character 15. Wing venation: 0) wing venation tubular and clear, 1) wing venation blurred. Character 16. Bristles on submarginal vein: 0) bristles on submarginal vein absent or short, 1) bristles on submarginal vein long, reaching beyond anterior margin of wing. Character 17. Fore wing marginal fringe: 0) fore wing marginal fringe present, 1) fore wing marginal fringe absent. Character 18. Basal vein development: 0) basal vein present and pigmented, 1) basal vein absent. Character 19. Scutellum posterior margin: 0) scutellar rim with single row of foveae, 1) scutellar rim smooth. Character 20. Metanotum exposure: 0) metanotum visible medially, 1) metanotum concealed medially.
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 189
Character 21. T1 development: 0) T1 flat, 1) T1 with a broad hump or horn. Note that species coded (0) for character 26, are coded as missing data for character 22-25. Character 22. Metasomal horn length: 0) distal metasomal horn reaching to or past level of posterior margin of scutellum (in lateral view), 1) distal metasomal horn not reaching to posterior margin of scutellum (in lateral view). Character 23. Metasomal horn shape: 0) metasomal horn uniformly slender, 1) metasomal horn broad at base. Character 24. Distal metasomal horn shape: 0) distal metasomal horn convex and narrowly rounded, 1) distal metasomal horn flattened. Character 25. Metasomal horn lateral shape: 0) metasomal horn circular in cross-section, 1) metasomal horn compressed laterally. Character 26. Effect of metasomal horn on scutellum: 0) posterior scutellum rounded or straight in dorsal view, 1) posterior scutellum indented medially. Character 27. Scutum groove/indentation: 0) scutum normal, 1) scutum with medial groove or deep emargination. Character 28. Propodeal lamellae: 0) dorsal propodeal lamellae blunt, 1) dorsal propodeal lamellae sharply pointed. Character 29. Hind femoral spines: 0) distal hind femur simple, 1) distal hind femur with two spines. Character 30. T2 anterior margin sculpturing: 0) T2 anterior margin striate, 1) T2 anterior margin smooth. Character 31. Laterotergite position: 0) laterotergites free, 1) laterotergites inserted into submarginal groove. Character 32. First funicle segment (F1) length : width ratio: 0) first funicle segment ≤ 2 x as long as wide, 1) first funicle segment 3–3.5 x as long as wide, 2) first funicle segment ≥ 4 x as long as wide. Character 33. Postmarginal vein : stigmal vein ratio: 0) postmarginal vein ≤ 0.72 x as long as stigmal vein, 1) postmarginal vein 1 – 1.1 x as long as stigmal vein, 2) postmarginal vein ≥ 2 x as long as stigmal vein. Note that brachypterous species are coded as missing data for this character. Character 34. Metasoma length : width ratio: 0) metasoma ≤ 2.05 x as long as wide, 1) metasoma 2.4 – 2.85 x as long as wide, 2) metasoma 3.72 – 3.81 x as long as wide, 3) metasoma ≥ 5 x as long as wide. Character 35. T3 : T2 ratio: (measured in dorsal mid-line) 0) T3 ≤ 0.96 x as long as T2, 1) T3 0.97 – 1.85 x as long as T2, 2) T3 1.86 – 2.74 x as long as T2, 3) T3 2.75 – 3.63 x as long as T2, 4) T3 ≥ 3.64 x as long as T2.
Muhammad Iqbal and Andrew D. Austin 190
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 2 Data matrix of character states for representative species of baeine genera 1 1111111112 2222222223 33333 1234567890 1234567890 1234567890 12345 Sparasion sp.
000101-101 0100000100 0----00100 11221
Nixonia sp.
000101-101 0100000100 0----00000 11021
Embidobia metoligotomae Dodd
0100000000 0100010100 1110000000 10201
Gryon sp.
0001100100 0100010100 0----00000 10201
Anabaeus sp.
2101101001 0111----10 1000000001 10-01
Apobaeus sp.
2101101101 011-----11 0----00001 10-00
Baeus leai Dodd
2001101101 011-----11 0----00001 00-00
Baeus seminulum Haliday
2101101111 011-----11 0----00001 00-00
Baeus machodoi (Risbec)
2101101001 011-----11 0----00001 00-00
Baeus saliens (Hickman)
2101101011 011-----11 0----00001 00-00
Ceratobaeus cornutus Ashmead
01?1101?0? ?0000?0?00 1000010?00 10012
Ceratobaeus sp. 1
0101000100 0000000100 1000000000 10003
Ceratobaeus sp. 2
0110000100 0101-1—-10 1000010000 10-02
Ceratobaeus leai Dodd
0111101000 0101-1—-10 1000010000 10-03
Ceratobaeus giraulti Dodd
0001101100 1100001100 1000010000 10002
Ceratobaeus mirabilis Dodd
0000101100 1100010010 1000011000 10011
Ceratobaeus reiki Austin
0000101000 0100010010 1000010000 10032
Ceratobaeus cuspicornutus Austin
0001101000 0100010010 1000010000 10011
Ceratobaeus fasciatus Dodd
0110101100 1100010110 1000010000 10011
Ceratobaeus laeviventris Dodd
0110101100 1100010100 1000010000 10001
Ceratobaeus sp. 3
0101101100 1100000110 1000010000 10003
Ceratobaeus sp. 4
0110101100 1100000100 1000010100 10002
Ceratobaeus setosus Dodd
0110101001 0100010000 1000000000 10001
Ceratobaeus intrudae Austin
0100101000 0100010100 1000000100 10112
Ceratobaeus sp. 5
0111101100 1000010100 1000000100 10002
Ceratobaeus sp. 6
0101101000 0100010010 1000000000 10011
Ceratobaeus sp. 7
0000101100 1100010100 1100000000 10002
Cyphacolus sp.
0001101100 0000101100 1000100101 10-01
Echthrodesis sp.
1001000001 011-----10 0----00001 10-00
Hickmanella holoplatysa Austin
0201000000 0000110100 0----00100 10-02
Hickmanella intrudens (Hickman)
0201000000 0000100100 0----00000 10-02
Idris flavicornis Foerster
0100101001 0001010100 0----00000 10003
Idris helpidid (Hickman)
0101101000 0100010100 0----00000 10002
Idris niger (Hickman)
0101101100 0100000100 0----00000 10003
Idris pulcher (Dodd)
0101101101 0100000100 0----00100 10004
Idris seminitidus (Dodd)
0001101001 0000010100 0----00000 10002
Idris theridii (Hickman)
0001101001 0100010100 0----00000 10002
Idris sp. 1
0100101001 0101----00 0----00000 10-03
A Preliminary Phylogeny for the Baeini (Hymenoptera: Scelionidae) 191
Idris sp. 2
0101000101 0100101000 0----00000 10002
Mirobaeoides barbarae Austin
2101000101 011-----10 1010000010 10-01
Mirobaeoides pecki (Austin)
2000000001 011-----11 0----00010 10-00
Mirobaeoides scutellaris Austin
2101000101 011-----10 0----00011 10-00
Mirobaeoides tasmanicus Dodd
2011000101 011-----10 0----00011 10-00
Mirobaeus bicolor Dodd
1001000001 011-----10 0----00000 10-01
Neobaeus novazealandensis Austin
2100000001 011-----11 0----00001 00-00
Odontacolus longiceps Kieffer
0001101100 0100010100 1000100100 10101
Odontacolus sp. 1
0111101100 0100101100 1000100100 10001
Odontacolus sp. 2
0011101101 1100010100 1000100101 10001
Odontacolus sp. 3
0100101101 1100000100 1000100101 10001
Genus 1
1001000101 011-----10 1010010000 10-03
Genus 2
1100000001 0101----10 1010010010 10-02
Genus 3
1001101101 0101----00 0----00010 10-01
Genus 4
2110101001 0101----11 1010010000 10-00
Genus 5
1100101001 0101----10 0----00000 10-00
Hymenopteran Orbicular Sensilla Hasan H. Basibuyuk1,2,3, Alexandr P. Rasnitsyn4, Mike G. Fitton2 and Donald L. J. Quicke1,2 1
Unit of Parasitoid Systematics, CABI Bioscience UK Centre (Ascot), Department of Biology, Imperial College at Silwood Park, Ascot, SL5 7PY United Kingdom (email:
[email protected]) 2
Department of Entomology, The Natural History Museum, London SW7 5BD United Kingdom 3
Department of Biology, Cumhuriyet University, 58140 Sivas, Turkey
4
Palaeontological Institute Russian Academy of Sciences, Profsoyuznaya Str. 123, 117647 Moscow, Russia
Introduction The sensory systems of insects have evolved numerous specialisations that allow them to detect important features of the external environment, monitor their own internal states and provide information on position of their appendages. Considering the vast numbers of insect species and variety of habitats they successfully occupy, it is not surprising that they posses equally diverse sensory systems (Frazier 1985). However, most studies have only covered antennal sensilla, while sensilla located in other body parts have not been thoroughly investigated. Studies on the sensory systems of the Hymenoptera show this general pattern and, apart from several studies on the ovipositor sensilla, there have only been a few papers reporting sensilla from other body parts (Schmidt & Smith 1987; Navasero & Elzen 1991; Meyhöfler et al. 1997). Sensilla from other body parts may also be useful potentially for phylogenetic reconstruction. Although there has been increasing effort to reconstruct the evolutionary history of the Hymenoptera (e.g. Rasnitsyn 1988; Brothers & Carpenter 1993; Dowton & Austin 1994; Vilhelmsen 1997; Ronquist et al. 1999), there is still no one robust higher level phylogenetic hypothesis for the order (Ronquist et al. 1999). Rasnitsyn’s (1988) most comprehensive and fully resolved higher level phylogeny has recently been shown (Ronquist et al. 1999) to have no convincing support for any of its major groupings. The relationships among symphytan (Vilhelmsen 1997) and aculeate families (Brothers & Carpenter 1993) have been analysed independently and are relatively robust in comparison with the parasitoid groups. Exploration of novel character systems is essential if we are to test existing hypotheses and obtain a more stable phylogeny. In this paper, we present some preliminary findings from one novel character system, orbicular sensilla and some associated structures. The orbicula is a dorsal cuticular sclerite at the base of the membranous arolium in the distal most tarsal segment of insect legs (Figs 1–4). Phylogenetic and functional implications, especially for non-aculeate apocritans, are discussed.
Material and Methods A total of 137 species belonging to 121 genera, representing all superfamilies and most families of non-aculeate apocritans, were studied. Only one species was examined for some relatively rare taxa, but in most cases the number of species examined was three or more. Standard SEM
192
Hymenoptera: Evolution, Biodiversity and Biological Control
Hymenopteran Orbicular Sensilla 193
Table 1
Explanation of characters and character states.
Code
Character
States
A
Sensilla trichodea A (STA) on orbicula
0 = absent; 1 = present
B
Number of STA (when present)
0 = one; 1 = two; 2 = three or more
C
Sensilla trichodea B (STB)
0 = absent; 1 = present
D
Number of STB (when present)
0 = one; 1 = two; 2 = three or more
E
STB (if two or more)
0 = scattered; 1 = forming a single row
F
STB
0 = simple; 1 = basally widened
G
STB
0 = without a basal spur;1 = with a basal spur
H
Sensilla campaniformia (SC)
0 = none; 1 = with encircling ridge (type A); 2 = without an encircling ridge (type B)
I
Number of SC (when present)
0 = one; 1 = two; 2 = three or more
J
Position of SC relative to STA
0 = anterior (distal); 1 = posterior (basal)
techniques were employed on dry or ethanol preserved specimens. The identifications of orbicular sensilla are based on Frazier (1985).
Results Four types of morphologically discernible sensilla are found. Of these two are setiform, termed here sensilla trichodea A (STA) and sensilla trichodea B (STB), and the other two are dome-like sensilla, sensilla campaniformia A (SCA) and sensilla campaniformia B (SCB). Variations in these and associated structures are summarised in Tables 1 and 2. We have not specifically investigated intraspecific variation but there appears to be no major variation in several cases we have examined. STA arise from a flexible socket with the hair trunk gradually tapering towards the tip (Figs 1–4). STB can be differentiated from the STA in usually being shorter and more slender, positioned posterior to the STA and usually with a less pronounced socket (Figs 1–3). STB are usually simple cylindrical hairs with an irregular distribution but, in some cases, they are arranged into a single transverse row (Figs 2–3) and may be basally compressed and widened (Fig. 2). Some taxa additionally possess a spur on the sensilla socket (Fig. 2). Of the two types of sensilla campaniformia, SCA are usually in the shape of a rounded dome encircled by a pronounced cuticular ridge (Fig. 1); their number and placement are variable. SCB, however, usually have no distinct encircling ridge but consistent in number and are always located laterally (Fig. 2).
Discussion The orbicular sensilla are arguably all mechanoreceptors (McIver 1985). In one eulophid, they have been suggested as having a role in detecting vibration produced by its leaf miner host (Meyhöfer et al. 1997). However, the presence of orbicular sensilla in early phytophagous lineages (Basibuyuk et al. 2000), as well as in males, suggest that they must have other roles. It is most likely that they evolved primarily to monitor the state of stress caused by locomotive behaviour. Host detection, in this context, can be seen as a secondary role assumed after the evolution of parasitism.
Hasan H. Basibuyuk, Alexandr P. Rasnitsyn, Mike G. Fitton and Donald L. J. Quicke 194
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Orbicular sensilla and associated structures in the non-aculeate apocritan Hymenoptera. Symbols: (—) not applicable; (?) state is unknown due to unavailability of data; (x?) a specified state with question mark shows our best judgement but not absolutely certain (see Table 1 and text for explanation of characters). A
B
C
D
E
F
G
H
I
J
Stephanidae
0/1
1
0/1
2
0
0
0
0/1
1
0
Trigonalidae
1
1/2
1
2
0
0
0
1
1
0
Megalyridae
0/1
2
1
2
0
0
0
1
0
0
Evanioidea
1
1
1
2
0
0
0
0
—
—
Ibaliidae
1
1
1
2
0
0
0
0
—
—
other Cynipoidea
1
1
0/1
1/2
1
1
0?
0
—
—
Megaspilidae
1
2?
0/1
2?
0
0
0
0
—
—
Ceraphronidae
1
0
0/1
2?
0/1
?
0
0
—
—
Diapriidae
1
0
1
1
1
1
0
2
1
0
Heloridae
1
1
1
2
0
0
0
0
—
—
Vanhorniidae
1
1
1
2
1
?
0
0
—
—
Pelecinidae
1
1
1
2
0
0
0
0
—
—
Proctotrupidae
0/1
1?
1?
1/2
0
?
0
0
—
—
Roproniidae
1
1
1
2
0
0
0
0
—
—
Monomachidae
1
1
1
2
0
0
0
0?
—
—
New Zealand fam n.
1
0
1
1
1
1
1
2?
?
?
Scelionidae
1
1
0/1
0/2
0
?
0
0/2
1
0
Platygastridae
0?
—
1?
1
1
1
0
0
—
—
Chalcididae
1
0
1
2
0
1
0
2
1
1
other Chalcidoidea
1
0
1
1/2
1
1
1
2
1
1
Mymarommatidae
1
0
1
1
1
?
?
?
?
?
Braconidae
1
1/2
1
2
0
0
0
1
1
1
Ichneumonidae
1
1/2
0/1
0/1/2
0
0
0
1
1/2
0/1
Variation in the number, shape, position and arrangement of orbicular sensilla and associated structures are found to be particularly informative for relationships of several microhymenoptera. Most basal lineages of Apocrita have two STA and a change in number is considered to be derived. Presence of a single STA with a well-developed socket and collar is a putative synapomorphy for the Chalcidoidea, Diapriidae, Mymarommatidae, the undescribed New Zealand family of Proctotrupoidea and possibly the Ceraphronidae (Figs 2, 4). Arrangement of the STB into a single transverse row (Figs 2–4) supports a larger putative clade comprising the Chalcidoidea (except Chalcididae), Mymarommatidae, Ceraphronidae,
Hymenopteran Orbicular Sensilla 195
Figures 1-4
Orbicular sensilla in various Hymenoptera: 1) Megalyra fasciipennis (Megalyridae), scale bar = 10 mm; 2) Gastracanthus pulcherrimus (Pteromalidae), scale bar = 8 mm; 3) Callaspidia defonscolombei (Figitidae), scale bar = 7 mm; 4) Coptera occidentalis (Diapriidae), scale bar = 3 mm. B = basal spur; SCA = sensilla campaniformia A; SCB = sensilla campaniformia B; STA = sensilla trichodea A; STB = sensilla trichodea B.
Cynipoidea, Diapriidae, Platygastridae, Vanhorniidae and the undescribed New Zealand family of Proctotrupoidea. STB are laterally compressed and widened basally (Fig. 2–4) in the Chalcidoidea, Cynipoidea, Diapriidae, Platygastridae and the undescribed New Zealand family of Proctotrupoidea, suggesting another clade within the above larger grouping. A spur is present at the base of each STB in the Chalcidoidea (except Chalcididae) and in the undescribed New Zealand family of Proctotrupoidea (Fig. 2), and a similar structure, but situated far from the sensillum, is also present in some Cynipoidea.
Hasan H. Basibuyuk, Alexandr P. Rasnitsyn, Mike G. Fitton and Donald L. J. Quicke 196
Hymenoptera: Evolution, Biodiversity and Biological Control
SCB are present in almost all members of the Chalcidoidea (Fig. 2) and Diapriidae (Fig. 4), and in some Scelionidae and possibly in the New Zealand family of Proctotrupoidea. However, its presence in the new family has yet to be confirmed. Several recent investigations of novel character systems have suggested alternative relationships among parasitic taxa (e.g. Basibuyuk & Quicke 1997, 1999; Quicke et al. 1998; Gibson 1999), but no single character system is likely to verify the ‘true’ relationships. A formal parsimony analysis including all the new findings is needed to evaluate these alternative hypotheses and is currently in preparation.
Acknowledgements We wish to thank the staff of electron microscopy and photographic units, NHM, London for their assistance. Kees van Achterberg, Nando Bin, Barry Bolton, John LaSalle, Massimo Olmi, Andy Polaszek, Frank van Veen and Annette Walker donated some of the material studied. This work was supported in part by a Leverhulme Trust research grant to DQ and MF and by the NERC Initiative in Taxonomy.
References Basibuyuk, H. H. & Quicke, D. L. J. (1997) Hamuli in the Hymenoptera (Insecta) and their phylogenetic implications. Journal of Natural History 31: 1563-1585. Basibuyuk, H. H. & Quicke, D. L. J. (1999) Grooming behaviours in the Hymenoptera (Insecta): potential phylogenetic significance. Zoological Journal of the Linnean Society 125: 349-382. Basibuyuk, H. H., Quicke, D. L. J., Rasnitsyn, A. P. & Fitton, M. G. (2000) Morphology and sensilla of the orbicula, a sclerite between the tarsal claws in Hymenoptera. Annals of the Entomological Society of America 93: 625-636. Brothers, D. J. & Carpenter, J. M. (1993) Phylogeny of Aculeata: Chrysidoidea and Vespoidea (Hymenoptera). Journal of Hymenoptera Research 2: 227-304. Dowton, M. & Austin, A. D. (1994) Molecular phylogeny of the insect order Hymenoptera: Apocritan relationships. Proceedings of the National Academy of Sciences, USA 91: 99119915. Frazier, J. L. (1985) Nervous system: sensory system. pp. 287-356. In Blum, M. S. (Ed.), Fundamentals of Insect Physiology. John Wiley and Sons, New York. Gibson, G. A. P. (1999) Sister-group relationships of the Platygastroidea and Chalcidoidea (Hymenoptera) – an alternate hypothesis to Rasnitsyn (1988). Zoologica Scripta 28: 125-138. McIver, S. B. (1985) Mechanoreception. pp. 71-132. In Kerkut, G. A. and Gilbert, L. I. (Eds), Comprehensive Insect Physiology Biochemistry and Pharmacology, Volume 6. Pergamon Press, Oxford. Meyhöfler, R., Casas, J. & Dorn, S. (1997) Mechano- and chemoreceptors and their possible role in host-location behavior of Sympiesis sericeicornis (Hymenoptera: Eulophidae). Annals of the Entomological Society of America 90: 208-219. Navasero, R. C. & Elzen, G. W. (1991) Sensilla on the antennae, foretarsi and palpi of Microplitis croceipes (Cresson) (Hymenoptera: Braconidae). Proceedings of the Entomological Society of Washington 93: 373-347. Quicke, D. L. J., Wyeth, P. , Fawke, J. D., Basibuyuk, H. H. & Vincent, J. (1998) Manganese and zinc in the ovipositors and mandibles of hymenopterous insects. Zoological Journal of the Linnean Society 124: 387-396.
Hymenopteran Orbicular Sensilla 197
Rasnitsyn, A. P. (1988) An outline of evolution of the hymenopterous insects (Order Vespida). Oriental Insects 22: 115-145. Ronquist, F., Rasnitsyn, A. P., Roy, A., Eriksson, K. & Lindgren, M. (1999) Phylogeny of the Hymenoptera: A cladistic reanalysis of Rasnitsyn’s (1988) data. Zoologica Scripta 28: 13-50. Schmidt, J. M. & Smith, J. B. (1987) The external sensory morphology of the legs and hairplate system of female Trichogramma minutus Riley (Hymenoptera: Trichogrammatidae). Proceedings of the Royal Society of London Series B 232: 323-366. Vilhelmsen, L. (1997) The phylogeny of lower Hymenoptera (Insecta), with a summary of the early evolutionary history of the order. Journal of Zoological Systematics and Evolutionary Research 35: 49-70.
Karyology of Parasitic Hymenoptera: Current State and Perspectives Vladimir E. Gokhman Botanical Garden, Moscow State University, Moscow 119899 Russia (email:
[email protected])
Introduction Recent progress in molecular genetics provides substantial independent information on phylogenetic and taxonomic relationships of various insect groups. However, more traditional methods of contemporary taxonomy, such as chromosomal analysis, still have an important role to play in entomological research. This is because karyological techniques have certain advantages in studying taxonomic structure of many insect taxa, including those of parasitic wasps. First, karyotypic features of many groups of parasitic Hymenoptera are relatively diverse and stable, thus offering an important source of knowledge valuable for making taxonomic decisions, especially at lower levels. Second, most chromosomal characters are in fact morphological and may therefore be studied and analysed similarly to those of external morphology. Finally, karyotyping may be effectively used for detecting sibling species and population polymorphisms as a rapid and inexpensive screening method. The last detailed review of the karyology of parasitic Hymenoptera was published by Gokhman & Quicke (1995). Since that time, however, chromosomes of several dozen additional species have been examined. Moreover, new conclusions have been forthcoming on the basis of accumulated karyological information (Gokhman 1997a). In this paper I will present a brief overview of the current state and perspectives of chromosome studies in parasitic wasps.
Historical Review There are two conditions necessary for a successful karyotaxonomic study: 1) the presence of a considerable number of dividing cells in a certain tissue, and 2) the possibility for reliable identification of the individuals examined. Chromosomes of parasitic wasps from laboratory stocks have been traditionally investigated using immature stages (mostly prepupae or early pupae), but this method can be hardly applied to natural populations of parasitic Hymenoptera, because it is very difficult to locate these stages in nature as well as to recognise their taxonomic position in the absence of adults. However, prepupae and pupae of certain phytophagous Hymenoptera, such as Cynipidae, may be easily identified by their galls or after rearing adults from those galls. Thus, it is not surprising that the first data on chromosomes of the parasitic Hymenoptera were obtained more than 100 years ago by Henking (1892) who examined meiosis in the cynipid Diplolepis rosae (L.). He observed nine bivalents in D. rosae, and this information has been confirmed by modern cytogenetics (Stille & Dävring 1980). Unfortunately, results of many karyological studies of parasitic wasps made at the beginning of the century were unreliable, mainly because the histological protocols of the time were largely 198
Hymenoptera: Evolution, Biodiversity and Biological Control
Karyology of Parasitic Hymenoptera: Current State and Perspectives 199
unsuitable for karyotyping purposes. These dubious results were often omitted in later reviews, e.g. in that by Crozier (1975) who cited chromosome numbers for only 20 parasitic wasps karyotyped after 1932, whereas about 40 species were examined prior to this time. Even the advent of more progressive cytological methods in the early 1930’s, such as squash techniques, had no considerable impact on the number of wasp species studied. Furthermore, an apparent uniformity of chromosome number in the parasitic Hymenoptera (e.g. Dodds 1938) also inhibited karyological research for this group of insects. Even several decades later Crozier (1975) wrote: ‘The Parasitica are characterised by highly restricted variation in chromosome number within major groups and by a much smaller overall range of chromosome number than the Symphyta and Aculeata. However, the sample available was both small and unrepresentative’. Contrary to Crozier’s evaluation, Goodpasture (1974) in his pioneering Ph.D. thesis, ‘Cytological data and classification of the Hymenoptera’, demonstrated considerable karyotypic diversity in 21 species of parasitic wasps. For example, he found n = 4, 5 and 6 in different members of the genus Monodontomerus Westwood (Torymidae). Regretfully, much of Goodpasture’s data, except those on Torymidae and Cynipidae (Goodpasture 1975a, 1975b; Goodpasture & Grissell 1975), remained unpublished. Several papers appeared during the next decade (Hunter & Bartlett 1975; Hung 1982, 1986) which contained additional karyotypic information for several chalcidoid species belonging to the families Chalcididae, Encyrtidae and Trichogrammatidae. A new period of chromosomal research in the parasitic Hymenoptera is associated with the transition from squash to air-drying preparation techniques. As mentioned above, it was the use of immature stages of parasitic wasps for karyotyping which imposed considerable restrictions on the process of chromosomal investigation in natural populations of parasitic Hymenoptera. Using air-drying techniques for making chromosome preparations from ovaries of adult females, it has been possible to examine more than 90 species of Ichneumonidae, mainly belonging to the subfamily Ichneumoninae (Gokhman 1985 onwards). However, many other authors (Sanderson 1988; Dijkstra 1986; Dijken 1991; Baldanza et al. 1991a, 1991b, 1994) used immature stages for the karyological investigation of chalcidoids and cynipoids. The total number of species reached 190 after a detailed review was published by Gokhman and Quicke (1995). A further 60 species have been added during the last four years (Baldanza 1996; Quicke & Gokhman 1996; Gokhman & Kolesnichenko 1996, 1997, 1998a, 1998b, 1998c; Abe 1998; Quicke pers. comm.; Gokhman unpublished), including the first chromosome records for several major taxa, such as the Megaspilidae (n = 9), Figitidae (n = 11) and Dryinidae (n = 5), the ichneumonid subfamilies Ctenopelmatinae (n = 11 and 12) and Metopiinae (n = 11; Figs 1, 2), and the braconid groups Agathidinae (n = 11) and Cheloninae (n = 6). Therefore, about 250 species have been studied to date, which is approximately five times more than in 1975, but this is still minute when compared to the 50 000 described species of parasitic Hymenoptera world-wide (LaSalle & Gauld 1991). Nevertheless, a number of conclusions can be made even on the basis of this restricted sample.
Karyology of Parasitic Wasps: Recent Update and New Hypotheses The haploid number in the parasitic Hymenoptera ranges from 3 in Brachymeria intermedia (Nees) (Chalcididae), Encarsia protransvena viggiani (= transvena auct.) (Aphelinidae) and Aphidius sp. (Braconidae) (Hung 1986; Baldanza 1996, pers. comm.; Quicke pers. comm.) to 21 in Perithous scurra (Panzer) (Ichneumonidae) (Gokhman & Kolesnichenko 1997) and Chrysis viridula L.
Vladimir E. Gokhman 200
Figures 1–4
Hymenoptera: Evolution, Biodiversity and Biological Control
Karyograms of parasitic wasps: 1) Triclistus podagricus (Gravenhorst) (Ichneumonidae: Metopiinae), mitosis, 2n = 22; 2) same individual, first metaphase of meiosis, n = 11; 3) Aptesis puncticollis (Thomson) (Ichneumonidae: Cryptinae), mitosis, 2n = 16; 4) Aphidius matricariae Haliday (Braconidae: Aphidiinae), mitosis, 2n = 14. Scale bar = 10 µm.
(Chrysididae: Quicke & Gokhman 1996), its frequency distribution being apparently bimodal with peaks at 5 and 11 (Fig. 5). However, the latter mode may change to 10 if identical chromosome numbers are scored only once per genus (Fig. 6) according to the ‘genus-karyotype’ approach developed by Crozier (1975) in order to avoid uneven sampling errors. Since the first peak comprises derived members of the parasitic Hymenoptera, such as many Chalcidoidea, n = 10 or 11 were suggested to be the norm for all parasitic wasps (e.g. Gokhman & Quicke 1995). Recently, however, a karyotypic study of several less specialised groups, such as the Gasteruptiidae, Pimplinae (Ichneumonidae), Doryctinae, Opiinae and Alysiinae (Braconidae), demonstrated that the members of these groups (with a few exceptions) have n = 14–17 (Quicke & Gokhman 1996; Gokhman & Kolesnichenko 1997, 1998a), and thus the initial haploid number is probably closer to those values. Therefore, an independent parallel reduction in chromosome number is common in various lineages, e.g. in the Ichneumonidae and Braconidae having respective modes at n = 11 and 10. A similar number also occurs in some less advanced Chalcidoidea (for example, in the Eurytomidae n = 9–10), as opposed to the majority of chalcidoids with n = 5–6. Even in the high-numbered taxa there are several species with low haploid numbers, e.g. Alysia manducator (Panzer) (Braconidae) with n = 11 and Polysphincta tuberosa (Gravenhorst) (Ichneumonidae: Pimplinae) with n = 9, the other Alysiinae and Pimplinae having n = 16–17 and 14–21, respectively. Moreover, similar situations may be observed in some genera and species groups, and in these cases low n values are considered as synapomorphies. Specifically, the ichneumonid genera Patrocloides Heinrich (n = 8) and Pseudoamblyteles Heinrich (n = 9) are synapomorphic for their lower chromosome numbers, as opposed to the other members of the subtribe Ichneumonina with n = 10–17 (Gokhman 1997a). Similarly, two closely related cynipid
Karyology of Parasitic Hymenoptera: Current State and Perspectives 201
60 Chrysidoidea
50
Ceraphronoidea
40
Evanioidea Proctotrupoidea s.l.
30
Cynipoidea
20
Chalcidoidea Ichneumonoidea
10 0 2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22
Haploid chromosome number Figure 5
Histogram of haploid chromosome numbers for parasitic Hymenoptera.
35 Chrysidoide
30
Ceraphronoidea Evanioidea
25
Proctotrupoidea s.l.
20
Cynipoidea
15
Chalcidoidea Ichneumonoidea
10 5 0 2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22
Haploid chromosome number Figure 6
Histogram of haploid chromosome numbers for parasitic Hymenoptera, each n value occurring in a genus being represented only once.
species, Andricus kashiwaphilus Abe (n = 5) and A. mukaigawae (Mukaigawa) (n = 6) are synapomorphic for their n values, since all other members of the genus Andricus have n = 10 (Abe 1998). Chromosome numbers are known for all superfamilies of the parasitic Hymenoptera except Stephanoidea, Megalyroidea and Trigonalyoidea. However, in many others, such as Evanioidea and Ceraphronoidea, only one or two species have been examined. The Ichneumonoidea (n = 3–21; Fig. 3), Cynipoidea (n = 5–11) and Chalcidoidea (n = 3–12) are the best studied, the first group occupying the whole range of n values found in the parasitic wasps. The Chrysidoidea almost have as wide a variation in chromosome number (n = 5–21) as the Ichneumonoidea, although karyotypes of only six species have been investigated to date. Chrysidoids therefore
Vladimir E. Gokhman 202
Hymenoptera: Evolution, Biodiversity and Biological Control
represent a very divergent group where variation for n values in all families examined have n values which do not over lap (Bethylidae 10–14, Chrysididae 19–21, Dryinidae 5) (Gokhman & Kolesnichenko 1998b). These data also demonstrate that the apparent problem concerning numerical aspects of chromosomal evolution in higher hymenopterans postulated by Crozier (1975: ‘The most puzzling aspect of the overall karyotypic evolution of the Hymenoptera is the occurrence of low chromosome numbers in the Parasitica followed by the reappearance of high number in aculeates’) is now largely resolved. Groups with well studied karyotypes belong to the Ichneumonoidea, Cynipoidea or Chalcidoidea, namely Ichneumonidae (n = 8–21), Braconidae (n = 3–17), Cynipidae (n = 5–10), Aphelinidae (n = 3–11), Pteromalidae (n = 5–7), Torymidae (n = 4–6) and Trichogrammatidae (n = 5). Again, although chromosome numbers are now studied in 23 wasp families, as compared to eight listed by Crozier (1975), this is less than half of all known parasitic families. Karyological data obtained for separate tribes and lower taxa may also be of use in parasitic Hymenoptera taxonomy (see Gokhman 1997a for examples), although chromosome studies usually provide the most valuable information for taxonomic research at the species level. Except for a few unpublished cases of cryptic species detected in the Aphidiinae (Braconidae: Quicke pers. comm.), eight groups of sibling species belonging to the families Ichneumonidae, Encyrtidae, Pteromalidae and Torymidae are known to date (summarised by Gokhman & Quicke 1995). Recently, however, a further two such species groups were discovered in the Pteromalidae and Cynipidae. Chromosomes of 10 members of the Pteromalidae have been studied (Gokhman & Quicke 1995; Gokhman unpublished), all of them having n = 5 except for a local population (perhaps sibling species) of Nasonia vitripennis (Walker) from California with n = 6 (Goodpasture 1974). This family therefore has highly restricted variation in chromosome number. For example, although different karyotypes do occur in N. vitripennis which normally has n = 5, two other species belonging to the same complex, namely N. longicornis Darling and N. giraulti Darling, have the same n value. Moreover, karyotypes of the three Nasonia species studied under routine and differential (C-) staining appear to be very similar, if not identical (but see Gokhman & Westendorff 2000). Recent study of a species complex found in the pteromalid Anisopteromalus calandrae (Howard), represents a highly convincing example of using karyological methods for detecting sibling species (Gokhman et al. 1998). Anisopteromalus calandrae is widely known as an effective cosmopolitan parasitoid of stored-product pests. It has been thoroughly studied (see Gokhman et al. 1998 for review) but the presence of sibling species was previously not suspected. Recently, the karyotype of this species from a laboratory population maintained at the Imperial College (Silwood Park, U.K.) (Gokhman & Quicke 1995) was investigated. All individuals examined invariably showed a haploid number of seven, the value found only among chalcidoids in the aphelinid Encarsia tricolor Foerster (Baldanza et al. 1994). Therefore, populations of A. calandrae from other regions were likely to have other chromosome numbers. Karyological investigation of the laboratory stock originated from an indoor population from Moscow has revealed n = 5. Subsequent analysis demonstrated that the two populations were reproductively isolated from each other. They also appeared to differ in some morphological details and ecological attributes, thus supporting the proposal that they represent two distinct sibling species.
Karyology of Parasitic Hymenoptera: Current State and Perspectives 203
The example of the A. calandrae species complex shows that chromosomal investigation may be of taxonomic use even in groups presumed to be karyologically uniform and well-studied by other methods. It also demonstrates that cytogenetic information often has some prognostic features. Specifically, if an unusual karyotype is found in a particular taxon at the species level, other populations of this group should be intensively studied because they may well differ in chromosomal characters, thus making possible the detection of sibling species. A similar situation is known for A. mukaigawae s.l. (Abe 1998, see above), but two host races previously revealed in this complex appear to have different chromosome numbers, n = 5 and 6. Thus, the two races have acquired the status of sibling species differing in host range, shape of unisexual galls and karyotype structure. Again, n = 10 was previously found in all members of the Cynipoidea except for Diplolepis species (Cynipidae) with n = 9. Recently, however, karyological study of the first examined member of the Figitidae, Callaspidia defonscolombei Dahlbom showed n = 11 in this species (Gokhman 1999). Karyological analysis is the only direct method for studying the number of linkage groups in various species of parasitic wasps. This is a very simple and effective technique which sometimes cannot be replaced by formal or molecular genetics methods, and therefore it should not be neglected. The history of studying the PSR (paternal sex ratio) factor in N. vitripennis provides an excellent example in this respect. In the early 1980’s a particular sex ratio distorter was discovered in N. vitripennis which caused complete male offspring and was found to be transmitted paternally. Although some microscopic preparations were made from fertilised eggs, they were unsuitable for cytogenetic purposes and showed only chromosome clumping and subsequent elimination of a haploid set from those eggs. Moreover, the PSR factor was considered extrachromosomal, despite the fact that factors of that kind in animals are not transmitted by males (Werren et al. 1987). Crossing experiments seemed to support this assumption because they involved genetic markers for all linkage groups known that time in N. vitripennis. Only after DNA sequencing of the PSR region was undertaken and revealed satellite chromosomal DNA, a thorough karyological study was initiated which demonstrated the presence of a smaller B chromosome responsible for the PSR effect (Nur et al. 1988; Werren 1991).
Future Chromosome Studies on the Parasitic Hymenoptera Despite the general progress in studies on the karyology of the parasitic Hymenoptera, there are some prospective areas which deserve more intensive research. In support of higher-level phylogenetic studies within the Apocrita, more attention could be paid to elucidating the karyology of major groups, such as Megalyroidea, Trigonalyoidea and Stephanoidea, which have not yet been examined. This also applies to phylogenetic investigations at family level for groups where there are conflicting hypotheses. For example, studies on the Chalcidoidea could profitably target information on chromosome number and other characters for families which are believed to be relatively basal, such as the Myrmaridae. Chromosomal analysis of species held as laboratory stocks may represent a useful direction for karyotaxonomic research, because of their easy availability for interpreting differences in karyological characters and because they are often of economic importance. This is perhaps also true for parasitic Hymenoptera attacking patchy hosts. It was recently suggested by Gokhman et al. (1998) that parasitoids of such hosts may have higher inbreeding rates, as proposed by Askew
Vladimir E. Gokhman 204
Hymenoptera: Evolution, Biodiversity and Biological Control
(1968) for many chalcidoids, and therefore these taxa may have a larger variation in chromosomal characters. Indeed, karyotypes of Encarsia (Aphelinidae) (n = 3, 5, 6, 7, 8, 9 and 10) and Aphidius (Braconidae) (n = 3, 5, 6, and 7 – Fig. 4) (Baldanza 1996; Gokhman & Quicke 1995; Quicke pers. comm.; Gokhman unpublished) parasitising aggregated hosts (scales, whiteflies and aphids) support this assumption. Similarly, chromosome studies on the Encyrtidae or those attacking stored-product pests, may be extremely useful in solving taxonomic problems, as discussed above for the A. calandrae species complex. Modern methods of differential staining may provide important information, especially at lower taxonomic levels. For example, karyotypes of three species belonging to the genus Dirophanes Foerster (Ichneumonidae) have been studied using chromosomal C-banding, namely: D. callopus (Wesmael), D. fulvitarsis (Wesmael) and D. invisor (Thunberg). Although D. fulvitarsis and D. invisor both have 2n = 20, as opposed to D. callopus with 2n = 18, strong differences in heterochromatin distribution were found between the first two species (Gokhman 1997b). In an analogous way, Baldanza (1996) revealed notable differences between several members of the genus Encarsia having the same chromosome number with the help of the Ag-NOR banding. Karyological research can now also be incorporated into molecular studies on parasitic wasps. Study of the PSR factor in N. vitripennis is perhaps the best example of this. Another promising area of investigation involves entomopathogenic viruses associated with many species of parasitic wasps (Stoltz & Whitfield 1992). Since certain polydnaviruses may be integrated in the wasp genome, e.g. as in the braconid Chelonus inanitus (L.) (Gruber et al. 1996), the FISH (fluorescent in situ hybridisation) technique could become an important step in localising the virus on chromosomes. However, its karyotype must be examined in detail prior to such molecular studies in order to render the chromosome recognisable. Studies of this kind have been undertaken recently on C. inanitus (Gokhman & Kolesnichenko 1998c), with this braconid being the first parasitic wasp for which chromosomal localisation of a symbiotic virus can be demonstrated.
Acknowledgements I am very grateful to Fulvio Baldanza and Donald Quicke for their kind permission of citing their unpublished results, as well as to Beatrice Lanzrein and John Huber for helpful discussion. Donald Quicke, Leo Beukeboom, Petr Stary and Ulrich Schwörer also provided laboratory stocks of several parasitic wasp species for karyological studies.
References Abe, Y. (1998) Karyotype differences and speciation in the gall wasp Andricus mukaigawae (s.lat.) (Hymenoptera: Cynipidae), with description of the new species A. kashiwaphilus. Entomologica Scandinavica 29: 131-135. Askew, R. R. (1968) Considerations on speciation in Chalcidoidea (Hymenoptera). Evolution 22: 642-645. Baldanza, F. (1996) Studi citotassonomici ed analisi dei profili proteici di alcune specie del genere Encarsia Foerster (Hym.: Aphelinidae). Tesi di dottorato. Università degli studi di Napoli “Federico II”, Napoli (Portici). Baldanza, F., Gaudio, L. & Viggiani, G. (1991a) Ricerche cariologiche sull’Archenomus orientalis Silvestri (Hymenoptera: Aphelinidae), parassitoide di Pseudaulacapis pentagona (Targioni Tozzeti) (Homoptera: Diaspididae). pp. 457-461. In Atti XVI Congresso nazionale italiano di Entomologia. Bari – Martina Franca 23-28 settembre 1991.
Karyology of Parasitic Hymenoptera: Current State and Perspectives 205
Baldanza, F., Odierna, G. & Viggiani, G. (1991b) A new method for studying chromosomes of parasitic Hymenoptera, used on Encarsia berlesei (Howard) (Hymenoptera: Aphelinidae). Bollettino del Laboratorio di Entomologia Agraria Filippo Silvestri 48: 29-34. Baldanza, F., Odierna, G. & Viggiani, G. (1994) Studi cariologici comparati su alcune specie del genere Encarsia Foerster (Hymenoptera: Aphelinidae). pp. 153-157. In Atti XVII Congresso nazionale italiano di Entomologia. Udine 13-18 giugno 1994. Crozier, R. H. (1975) Animal Cytogenetics 3 (7). Gebrüder Borntraeger, Berlin-Stuttgart. Dijken, M. J. van. (1991) A cytological method to determine primary sex ratio in the solitary parasitoid Epidinocarsis lopezi. Entomologia Experimentalis et Applicata 60: 301-304. Dijkstra, L. J. (1986) Optimal selection and exploitation of hosts in the parasitic wasp Colpoclypeus florus (Hym., Eulophidae). Netherlands Journal of Zoology 36: 177-301. Dodds, K. S. (1938) Chromosome numbers and spermatogenesis in some species of the hymenopterous family Cynipidae. Genetica 20: 67-84. Gokhman, V. E. (1985) Chromosome sets in some Ichneumoninae (Hymenoptera: Ichneumonidae). Zoologichesky Zhurnal 64: 1409-1413. [in Russian] Gokhman, V. E. (1997a) Chromosome number and other karyotypic features of parasitic wasps as a source of taxonomic information. Boletín de la Asociacíon Española de Entomología (Suplemento) 21: 53-60. Gokhman, V. E. (1997b) Differential chromosome staining in parasitic wasps of the genus Dirophanes (Hymenoptera, Ichneumonidae). Zoologichesky Zhurnal 76: 65-68. [in Russian] Gokhman, V. E. (1999) Chromosomes of Callaspidia defonscolombei (Hymenoptera: Figitidae) Zoologichesky Zhurnal 78: 1476-1477. [in Russian] Gokhman, V. E. & Kolesnichenko, K. A. (1996) New data on karyology of the Ichneumonoidea (Hymenoptera). pp. 25-27. In Gokhman, V. E. & Kuznetsova, V. G. (Eds), Karyosystematics of the Invertebrate Animals 3. Botanical Garden, Moscow State University, Moscow. [in Russian] Gokhman, V. E. & Kolesnichenko, K. A. (1997) Chromosomes of ichneumon flies of the subfamily Pimplinae (Hymenoptera, Ichneumonidae). Folia Biologica (Kraków) 45: 139-141. Gokhman, V. E. & Kolesnichenko, K. A. (1998a) Chromosomes of parasitic wasps of the subfamily Alysiinae (Hymenoptera, Braconidae). Zoologichesky Zhurnal 77: 1197-1199. [in Russian] Gokhman, V. E. & Kolesnichenko, K. A. (1998b) First chromosome record for the family Dryinidae: the karyotype of Anteon brevicorne Dalman (Hymenoptera: Chrysidoidea). Journal of Hymenoptera Research 7: 116-117. Gokhman, V. E. & Kolesnichenko, K. A. (1998c) Karyotype of Chelonus inanitus (L.) (Hymenoptera, Braconidae). Entomologicheskoye Obozreniye 77: 663-666. [in Russian] Gokhman, V. E., Timokhov, A. V. & Fedina, T. Yu. (1998) First evidence for sibling species in Anisopteromalus calandrae (Hymenoptera: Pteromalidae). Russian Entomological Journal 7: 157-162. Gokhman, V. E. & Quicke, D. L. J. (1995) The last twenty years of parasitic Hymenoptera karyology: An update and phylogenetic implications. Journal of Hymenoptera Research 4: 41-63. Gokhman, V. E. & Westendorff, M. (2000) The chromosomes of three species of the Nasonia complex (Hymenoptera: Pteromalidae) Beiträge zür Entomologie 50: 193-198. Goodpasture, C. (1974) Cytological Data and Classification of the Hymenoptera. Unpublished Ph.D. Thesis. University of California, Davis. Goodpasture, C. (1975a) Comparative courtship behaviour and karyology in Monodontomerus (Hymenoptera: Torymidae). Annals of the Entomological Society of America 68: 391-397.
Vladimir E. Gokhman 206
Hymenoptera: Evolution, Biodiversity and Biological Control
Goodpasture, C. (1975b) The karyotype of the cynipid Callirhytis palmiformis (Ashmead). Annals of the Entomological Society of America 68: 801-802. Goodpasture, C. & Grissell, E. E. (1975) A karyological study of nine species of Torymus (Hymenoptera: Torymidae). Canadian Journal of Genetics and Cytology 17: 413-432. Gruber, A., Stettler, P., Heiniger, P., Schümperli, D. & Lanzrein, B. (1996) Polydnavirus DNA of the braconid wasp Chelonus inanitus is integrated in the wasp’s genome and excised only in later pupal and adult stages of the female. Journal of General Virology 77: 2873-2879. Henking, H. (1892) Untersuchungen über die ersten Entwicklungsvorgänge in der Eiern der Insekten. III. Spezielles und Allgemeines. Zeitschrift für wissenschaftliche Zoologie 54: 1-274. Hung, A. C. F. (1982) Chromosome and isozyme studies in Trichogramma (Hymenoptera: Trichogrammatidae). Proceedings of the Entomological Society of Washington 84: 791-796. Hung, A. C. F. (1986) Chromosomes of three Brachymeria species (Hymenoptera: Chalcidoidea). Experientia 42: 579-580. Hunter, K. W., Jr., & Bartlett, A. C. (1975) Chromosome number of the parasitic encyrtid Copidosoma truncatellum (Dalman). Annals of the Entomological Society of America. 68: 61-62. LaSalle, J. & Gauld, I. D. (1991) Parasitic Hymenoptera and the biodiversity crisis. Redia 74: 315-334. Nur, U., Werren, J. H., Eickbush, D. G., Burke, W. D. & Eickbush, T. H. (1988) A ‘selfish’ B chromosome that enhances its transmission by eliminating the paternal genome. Science 240: 512-514. Quicke, D. L. J. & Gokhman, V. E. (1996) First chromosome records for the superfamily Ceraphronoidea and new data for some genera and species of Evanioidea and Chrysididae (Hymenoptera: Chrysidoidea). Journal of Hymenoptera Research 5: 203-205. Sanderson, A. R. (1988) Cytological investigation of parthenogenesis in gall wasps (Cynipidae: Hymenoptera). Genetica 77: 189-216. Stille, B. & Dävring, L. (1980) Meiosis and reproductive strategy in the parthenogenetic gall wasp Diplolepis rosae (L.) (Hymenoptera: Cynipidae). Hereditas 92: 353-362. Stoltz, D. B. & Whitfield, J. B. (1992) Viruses and virus-like entities in the parasitic Hymenoptera. Journal of Hymenoptera Research 1: 125-139. Werren, J. H. (1991) The paternal sex-ratio chromosome of Nasonia. American Naturalist 137: 392-402. Werren, J. H., Nur, U. & Eickbush, D. (1987) An extrachromosomal factor causing loss of paternal chromosomes. Nature 327: 75-76.
Morphology and Biogeography of the North African Ceramius Maroccanus-complex (Vespidae: Masarinae): Contribution of Morphometric Analyses to Taxonomic Decisions Volker Mauss Institut für Landwirtschaftliche Zoologie und Bienenkunde, Rheinische Friedrich-WilhelmsUniversität, Melbweg 42, D-53127 Bonn, Germany (email:
[email protected])
Introduction The genus Ceramius comprises a monophyletic subtaxon within the Masarinae (Carpenter 1993). Twenty species exist in the Afrotropical region (Gess 1996; Gess 1997), while 13 species are reported to occur in the Palaearctic (Richards 1962; Mauss 1998). Knowledge about the taxonomy and bionomy of the Afrotropical taxa of Ceramius is extensive (e.g. Gess 1996; Gess 1997; Gess et al. 1997), while the Palaearctic species are comparatively poorly known. The following study on the Ceramius maroccanus-complex is part of a revision of the Palaearctic taxa of the genus. The C. maroccanus-complex constitutes a monophyletic subtaxon of species group 7 sensu Richards of Ceramius (Mauss 1999). It is endemic to south-western Morocco. Gusenleitner (1990) distinguished three taxa: C. m. maroccanus (Giordani Soika), C. m. rubripes Gusenleitner and C. montanus Gusenleitner. New material from Morocco which was received by the ‘Biologiezentrum des Oberösterreichischen Landesmuseums Linz’ in Austria suggests that the situation is more complicated.
Materials and Methods Dry specimens were investigated under a stereomicroscope (51 females, 52 males). All males were genitalised; male genitalia were removed after resoftening the specimens in a wet chamber and studied in 70% ethanol. Fourteen parameters of the exoskeleton of the specimens were measured using a WILD M3 stereomicroscope in monaxial position with calibrated ocular micrometer. Wing venation was examined by use of digital pictures which were analysed semiautomatically with WinBee 6.0 (cf. Schröder et al. 1999). Twenty-two lines and 34 angles in the fore wing were measured. A large number of dimensionless quotients was calculated from the measured lines of fore wing and exoskeleton. Then a discriminant analysis was calculated separately for each sex by SPSS using the dimensionless quotients and the angles of the fore wing as discriminating variables. The geographical distribution of the discriminant function scores were investigated following a method proposed by Fuchs (1998). By this method the position of each specimen within the discriminant space is represented by the colour of a symbol via the RGB-mode, while its geographical origin is indicated by plotting the symbol on a distribution map.
207
Volker Mauss 208
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Records of members of the Ceramius maroccanus-complex in south-western Morocco (number of specimens per site is not taken into account).
Results The morphological investigation revealed that the complex consists of four taxa. They can be separated by characters of the male genitalia, especially the aedeagus and the volsella, the shape of the abdominal sternum IX of the males, and the form of the scutellum and the pronotal furrow of the females. Moreover, the taxa show differences in the integumental sculpture of various parts of the exoskeleton and in the variability of coloration. After comparing these taxa with the types or paratypes they were assigned to C. m. maroccanus, C. m. rubripes, C. montanus and an unknown taxon. The four taxa differ in their geographical distribution (Fig. 1), altitudinal dispersal and flight period. Males and females of the four taxa can be separated completely by three discriminant functions each. The discriminant function scores do not show any discernible geographical variation within the taxa and no clinal variation could be observed between them (Mauss & Schröder unpublished). Therefore, it is assumed that there is no gene flow between the taxa.
Discussion The morphological, biogeographical and morphometric analyses indicates that the C. maroccanus-complex consists of four well separated taxa. Probably no gene flow exists between the populations. Nevertheless, conclusions about the status of the four taxa can only be obtained by indirect indications of reproductive isolation mechanisms between them (cf. Willmann 1985). Such indications could be distinctive differences in characters which probably influence the
Morphology and Biogeography of the North African Ceramius Maroccanus-complex 209
mating success of individuals. Field observations of mating behaviour in some species of Ceramius (Mauss & Berger 2000) indicate that relevant characters of the males are probably the coloration of the head, the form of abdominal sternum IX and the configuration of male genitalia, among others. In the females the shape of the scutellum and possibly also of the pronotal furrow seem to be important. The four recognised taxa differ especially in these characters indicating that they are reproductively isolated. Furthermore, the more-or less sympatric occurrence of C. montanus and C. m. maroccanus or C. m. maroccanus and C. m. rubripes, respectively can be used as an additional criterion for the existence of reproductive isolation mechanisms between them (cf. Mayr 1975). For these reasons the four taxa are hypothesised to represent isolated biospecies. They are described or redescribed by Mauss (1999).
References Carpenter, J. M. (1993) Biogeographic patterns in the Vespidae (Hymenoptera): Two views of Africa and South America. pp. 139-155 In Goldblatt, P. (Ed.), Biological Relationships Between Africa and South America. Yale University Press, New Haven and London. Fuchs, S. (1998) Die Oberurseler Datenbank in Farbe. ADIZ / die biene 8: 17. Gess, F. W. (1997) Contributions to the knowledge of Ceramius Latreille, Celonites Latreille, Jugurtia Saussure and Masarina Richards (Hymenoptera: Vespidae: Masarinae) in South Africa. Journal of Hymenoptera Research 6: 36-74. Gess, S. K. (1996) The Pollen Wasps – Ecology and Natural History of the Masarinae. Harvard University Press, Cambridge, Massachusetts. Gess, S. K, Gess, F. W. & Gess, R. W. (1997) Update on the flower associations of southern African Masarinae with notes on the nesting of Masarina strucki Gess and Celonites gariepensis Gess (Hymenoptera: Vespidae: Masarinae). Journal of Hymenoptera Research 6: 75-91. Gusenleitner, J. (1990) Die bisher bekannten Nordafrikanischen Arten der Gattung Ceramius Latreille 1810 (Hymenoptera, Vespoidea, Masaridae). Linzer biologische Beiträge 22: 565570. Mauss, V. (1998) The identity and distribution of Ceramius auctus (Fabricius, 1804) Vecht, 1970 (Ceramius spiricornis Saussure, 1854 syn. nov.) and Ceramius beaumonti (Giordani Soika, 1957) Richards, 1962 (Hymenoptera, Vespidae, Masarinae). Annales de la Société Entomologique de France 34: 163-183. Mauss, V. (1999) Taxonomy, biogeography and phylogenetic position of the North African Ceramius maroccanus-complex (Hymenoptera: Vespidae, Masarinae). Entomologica Scandinavica 30: 323-348. Mauss, V. & Berger, A. (2000) Funktion von Strukturen des Exoskeletts und des Genitalapparates der Männchen verschiedener Ceramius-Arten (Hymenoptera, Vespidae, Masarinae) während der Kopulation. Mitteilungen der Deutschen Gesellschaft für allgemeine und angewandte Entomologie 12: 485-488. Mayr, E. (1975) Grundlagen der zoologischen Systematik. Parey, Hamburg. Richards, O. W. (1962) A revisional Study of the Masarid Wasps (Hymenoptera, Vespoidea). British Museum (Natural History), London. Schröder, S., Wittmann, D., Roth, V. & Steinhage, V. (1999) Automated identification system for bees. p. 427. In Schwarz, M.P. & Hogendoorn, K. (Eds), Social Insects at the Turn of the Millennium, Proceedings of the XIII International Congress of IUSSI. Flinders University Press, Adelaide. Willmann, R. (1985) Die Art in Raum und Zeit: Das Artkonzept in der Biologie und Paläontologie. Parey, Hamburg.
Some Problems with Australian Tiphiid Wasps, with Special Reference to Coupling Mechanisms G. R. Brown Museum and Art Gallery of the Northern Territory, GPO Box 4646, Darwin, NT 0801 Current Address: Department of Primary Industry and Fisheries, GPO Box 990, Darwin, NT 0801, Australia (email:
[email protected])
Introduction The predominant element of Australian Tiphiidae is the Thynninae, although four other subfamilies are represented in the fauna: Tiphiinae and Diamminae each with one species, Myzininae with eight species, and Anthoboscinae with about 60 species. The Thynninae is virtually limited to Australia and South America and is most easily distinguished from all other Australian tiphiids, except Diamma bicolor Westwood, by the presence of complete sexual dimorphism including full aptery in females. Little is known of the biology of the group although almost 600 species have been described. A few species have been recorded as parasitoids of scarab larvae (e.g. Given 1953; Ridsdill-Smith 1970) while males of others are specific orchid pollinators (e.g. Adams & Lawson 1993; Bower 1996; Brown 1997). Because females are wingless, they are dependant on males to fly them to and from a food source (such as nectar) whilst remaining in copula. This results in pairs remaining in a copulatory position for prolonged periods and has favoured the development of spines, grooves and other more complex structures in both sexes. There are many problems associated with the study of this group. These include the near impossible task of rearing these wasps in the laboratory, the large number of undescribed and unrecognised higher taxa, and an estimated 2000+ undescribed species. In addition, there are numerous cryptic species that are more easily detected using biological methods, such as baiting with orchids, than by morphological characters alone. Because of the extreme sexual dimorphism in the group, pairs can only be associated by collecting them in copula. However, it appears that females will readily couple with males of other species, if conspecific males are rare or absent. This makes the association of congeneric sexes tenuous, especially if based on a single pair, but is the only method available. However, non-congeneric pairs involving described genera can be recognised on the basis of published morphological differences (e.g. Turner 1910; Brown 1998). If this misassociation of pairs is a common occurrence as suggested by the high number of miscoupled pairs in collections (Given 1954), it may be a survival mechanism that ensures an individual female is fed in preference to successful mating and fertilisation occurring. Association of the sexes could then only be reliably undertaken using morphological rather than behavioural methods. In a search for structures that may associate the dimorphic sexes of thynnines, I approached the problem with the assumption that any such structures would represent adaptations for prolonged coupling. These structures are discussed below and interpreted in relation to adaptations that enable pairs to remain in copula for prolonged periods whilst in flight and while feeding.
210
Hymenoptera: Evolution, Biodiversity and Biological Control
Some Problems with Australian Tiphiid Wasps 211
Materials and Methods Most specimens in Australian institutions as well as many holdings of Australian species held in overseas institutions were examined. Male genitalia were dissected from the metasoma, cleared in 10% KOH, rinsed in distilled water, dehydrated through an ethanol series, and then stored in glycerine filled vials that are attached to the pin of the original specimen.
Results and Discussion Coupling and copulatory position Typically, when females are ready to mate, they climb low plants and rest in a characteristic headup position with their antennae erect. Males are attracted from down-wind, and often fly in a zigzag manner. A male will couple with a female by curving the tip of the abdomen under, exerting the genitalia and clasping her. In this position, the male and female are both dorsum up with the male on top of the female. At some point the male genitalia become inverted along their longitudinal axis so that both are dorsum up, but facing in opposite directions, thus allowing both individuals to feed with more freedom. This rotation may simply occur from movement by the female relative to the male. Different taxa appear to have different positions before and after rotation of the male genitalia. These positions are difficult to quantify because captured pairs quickly separate, and there is only a handful of specimens in collections in which the pairs are still coupled. Some of these are illustrated in Figures 1-6.
Male genitalia and associated structures The male genitalia occur internally at the apex of the metasoma and are enclosed by the hypopygium (i.e. the last visible sternum, S8) ventrally and the epipygium (i.e. the last visible tergum, T7) dorsally. The genitalia consist of a fused pair of basiparameres with a proximal basal ring, an apical aedeagus, pairs of parameres, volsellae and, in some genera, parapenal lobes. Normally the dorsal surface is longer than the ventral surface (especially that of the basiparameres), but after the initial coupling with a female the genitalia become rotated along their longitudinal axis and thus inverted so that the dorsal surface becomes the ventral surface and vice versa. After uncoupling from the female, the genitalia appear to revert to their original orientation. The basal ring is usually short and ring-like. However, for example, in Phymatothynnus Turner it is long and narrow while in Tachynomyia Guérin it is long and curved. This increased length enables the apex of the genitalia to be exerted further than in other genera, and is assisted by the metasomal segments being relatively weakly sclerotised and therefore more flexible in allowing large genitalia to be exerted. The basiparameres have a long dorsal surface and a short ventral surface medially. When the genitalia are inverted after coupling, the now antero-ventral surface forms a ball and socket joint with the basal concavity of the hypopygium and the apex butts up against the apex of the female S6 preventing too close a coupling. The now dorsal surface is short medially and, together with the now dorso-basal margins of the parameres and volsellae, forms a concavity into which the apex of the epipygium and/or apex of the female pygidium fit (although this cavity is not species specific). In some genera such as Iswaroides Ashmead, Epactiothynnus Turner and Encopothynnus Turner, the dorso-basal angle of the basiparamere (viewed in profile) is pronounced. This is
G. R. Brown 212
Figures 1-6
Hymenoptera: Evolution, Biodiversity and Biological Control
Male-female coupling mechanisms (males stippled): 1) Zaspilothynnus interruptus Westwood, dorsal view, male genitalia inverted; 2) Zaspilothynnus interruptus, lateral view; 3) Rhagigaster stradbrokensis Given, dorsal view, male genitalia inverted; 4) Rhagigaster stradbrokensis, lateral view; 5) Aeolothynnus sp., dorsal view, male genitalia inverted; 6) Phymatothynnus nr monilicornis (Smith), dorsal view, male and male genitalia, ventral view of female (a = basiparameres, c = cuspis, d = digitus, e = epipygium, h = hypopygium, p = paramere, s = tergum 5, t = tergum 6, w = sternum 5, x = sternum 6, y = pygidium). Scale lines = 0.5 mm.
Some Problems with Australian Tiphiid Wasps 213
correlated with the presence of a transverse carina on the posterior margin of the epipygium which prevents the genitalia being moved too far dorsally when in copula. The parameres are the main grasping organs and as such their apical margins are shaped to fit against the apex of the female metasoma. The presence of setae, especially apically, increases the hold on the female. In most of the highly evolved genera such as Thynnus F., Thynnoides Guérin and Lestricothynnus Turner, there is a dense subapical internal brush of stout setae. As these setae are internal, their function may be sensory or stimulatory (although no corresponding features are found in females), rather than for grasping, particularly since some females such as those of Zaspilothynnus Ashmead and Catocheilus Guérin have highly modified pygidia to increase stability in coupling. The parameres may be narrowed where the apex of the female metasoma is modified, usually variously excavate, so that the sexes fit tightly together. This is particularly evident in Catocheilus, some Zaspilothynnus and, to some extent, in Thynnus. The volsellae are highly variable, and consist of basal cuspides and dorso-apical (when present) digiti. The cuspides probably help open the female and guide the genitalia into position. In Catocheilus the pygidium is notched laterally for reception of the cuspides. In some Rhagigasterini the cuspides each have a curved lobe which is apically setose. Their function, indicated by the presence of these setae, is probably sensory. Digiti are present in most of the more evolved genera, but may also occur in more primitive groups. They typically have a granular surface which is used to grasp the raised lip of the female S6. The granulations on the cuspides of Dimorphothynnus Turner may serve a similar purpose. The digiti are also apically setose and often vary considerably between closely related species. The parapenal lobes are probably involved in separating the female pleura for penetration by the aedeagus. This is suggested by the rounded and often enlarged apices. The aedeagus is variable in length among species and may differ significantly between closely related species. This variation may be related to the closeness of the pairs when in copula. Male hypopygium. The hypopygium (S8) has a basal concavity on the upper surface which encloses the genitalia ventrally, and an apical flat area which is especially well-developed in the more advanced Thynnini. The apical flat area is visible dorsally, and typically has an apical spine with a smaller lateral (i.e. close to the apical spine) or basal (i.e. close to the base of the hypopygium) spine. When the genital cavity is closed the hypopygium is contiguous with the epipygium and its shape somewhat reflects that of the epipygium, at least internally. The width of the apical spine and the presence of basal lobes or spines appears to be associated with the ventral (uninverted) distance between the parameres. Where a distinct angle is formed between the base of the apical spine and basal lobes, a notch is formed which guides the genitalia when moving outwards and downwards. When the genitalia are rotated along their longitudinal axis, the hypopygium is moved up and the apex may rest on the apical sterna of the female and the guiding mechanism is no longer engaged. The margins of these spines and lobes are often carinate and may serve to strengthen the hypopygium. The basal concavity supports the basiparameres in a ball and socket joint which provides mostly up and down movement (lateral movement being restricted by the shape of the epipygium). The contact area between female and male is increased and strengthened in some Rhagigaster Guérin by the presence of a subapical spine which is formed by the raised internal margins of the hypopygium. This spine may also make contact with the subgenital plate of the female during coupling to assist in penetration of the female.
G. R. Brown 214
Hymenoptera: Evolution, Biodiversity and Biological Control
Downward movement of the female is prevented by contact between the apical spine and S5 of the female. Where this spine is strongly developed, as in the more advanced Thynnini and some Rhagigasterini, S5 is rugose or carinate (compared with punctate in the more primitive Thynnini) and movement between the apical spine and S5 is restricted. Downward movement may be further restricted by friction between carinae on the flattened apex of the male hypopygium dorsally and the raised lip of S6 on the female. In many genera, the basal concavity of the hypopygium has a group of long erect setae. The function of these setae is unknown but it may be sensory in nature with the setae being used to determine the position of the genitalia. There are no other apparent alternative setae which may fulfil this function in those genera in which the hypopygial setae are lacking. Male epipygium. The epipygium is concave internally with an inner lip which contacts the hypopygium to close the genital cavity. In the more advanced Thynnini the hypopygium is located more posteriorly than in other tiphiid subfamilies and tribes so that much of the hypopygium is visible dorsally beyond the epipygium. In these genera the epipygium is produced posteriorly into a membranous apical plate which covers the genital cavity posteriorly. This membrane is bent upwards by the basiparameres when the genitalia are exerted. In many genera the epipygium restricts the rotation of the genitalia. In Rhagigaster the epipygium is often postero-laterally excavate for the reception of the parameres. There is some correlation between the shape of the epipygium apically and the gap between the parameres dorsally when the male genitalia is inverted so that, where the epipygium is broadly produced, the gap may also be broad. Strongly carinate lateral margins or other carinae are often present to strengthen the epipygium. The posterior margin may be transversely carinate with the carina often being further strengthened by the presence of upturned or subtuberculate postero-lateral angles. In the more advanced Thynnini the epipygium is usually variously carinate, especially longitudinally. In some genera such as Zaspilothynnus, Thynnus, Guerinius Ashmead and Oncorhinothynnus Salter, it is produced into a protruding subapical plate. This plate is often strengthened by longitudinal carinae while, posterior to this plate, the epipygium is transversely carinate, allowing some upward deflection of the sclerite, but with excess upward movement of the genitalia being prevented by the subapical plate. However, in Oncorhinothynnus the epipygium has a strongly developed median sagittal carina between the subapical plate and the apical margin. This strengthens the apical margin which is consequently unable to flex. In those genera in which the epipygium is simple and non-carinate (Eirone Westwood and some primitive Thynnini such as Phymatothynnus) the genitalia are often large compared with the size of the genital cavity opening, and the epipygium is weakly sclerotised so that it can flex when the genitalia are exerted.
Metasoma In males the metasoma is variously modified so that its flexibility and thus the movement of the female in copula is increased or decreased. In some genera (e.g. Lestricothynnus and Oncorhinothynnus) the metasomal segments are strongly sclerotised, strongly constricted, and the intersegmental muscles are large providing great flexibility. In extreme cases (e.g. Doratithynnus Turner and Zaspilothynnus) the sterna and terga may be produced into tubercles or spines. These processes are hollow and filled with muscle fibres thus extending the pulling distance of the muscle. Such modifications do not occur in the female.
Some Problems with Australian Tiphiid Wasps 215
Movement of the metasoma is limited in some genera, especially in the male, by modifications to S1, S2 and T1. The most flexible ventral suture is that between S1 and S2. These sterna are separated by a deep groove and are often obliquely truncate at the suture. This allows greater movement than other ventral sutures. In many Rhagigaster spp. the metasoma is longer than in other genera, and S1 and S2 are medially tuberculate near the suture. These tubercules limit downward movement. In genera that have S1 medially raised (e.g. Agriomyia Guérin) downward movement is less than in those which have S1 flat (e.g. Elidothynnus Turner). The upward movement of the metasoma is similarly restricted by modifications to T1 anteriorly and the propodeum posteriorly. In males of those genera in which T1 is wider than long, the propodeum is usually almost flat and oblique so that the two surfaces can come into contact. This is common in males of Thynnini where the females are often relatively much larger than in other tribes. The greatest flattening of the propodeum occurs in the males of Thynnus where the propodeum is almost vertically truncate. In the females of most species T1 is usually truncate anteriorly and the propodeum truncate posteriorly. A greater degree of movement is possible where the propodeum is excavate posteriorly, as in Rhagigaster castaneus Smith, but restricted where T1 is produced antero-medially, as in R. auriceps Turner. Female pygidium. The pygidium is variously modified to increase adhesion between the sexes. Modifications include lines or brushes of setae, a rugose surface, carinae, notches or grooves into which the cuspides or parameres of the male may fit, and the narrowing of the pygidium so that it can be grasped laterally. Female sternum 5. Sternite 5 is rugose or carinate in most groups where the apical spine of the hypopygium of the male is long (especially the Thynnini). Contact between S5 and the hypopygium restricts the downward movement of the female relative to the male while in copula. Female sternum 6. The margin of S6 is lipped in some genera so that it can be: 1) grasped by the digitus of the male (when present); 2) butted against by the apex of the basiparameres ventrally (when inverted); and 3) stabilised by touching the carinae on the apical flat area of the hypopygium dorsally.
Associations between structures The above modifications can be summarised as follows: 1) those which associate sexes; 2) those which are associations between the male genitalia and surrounding sclerites; 3) those which favour movement of pairs whilst in copula; or 4) those which restrict movement of pairs whilst in copula (Table 1). None of these structures appear to be restrictive enough to ensure that species are always correctly paired. More associations were found between the genitalia and the external apical sclerites of the male (Table 2). However, there is a point where too much movement will place strain on the coupling, and there must be some mechanism(s) which restrict or limit movement (Table 3). How species specific is this coupling mechanism? This study has identified some relationship between certain morphological structures in both sexes. However, it has not been possible to demonstrate a specific ‘lock and key’ mechanism that prevents pairs from different species or genera coupling. Further, there may be an alternative pheromonal or behavioural mechanism that enables females to miscouple when males of the same sex are rare or absent. The existence of such a mechanism may rely on the ability of females
G. R. Brown 216
Table 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Morphological structures that associate conspecific pairs of thynnine wasps.
Male
Female
parameres narrow (Fig. 1)
pygidium grooved laterally (Fig. 1)
cuspis large and triangular (Fig. 2)
pygidium with lateral grooves/spines
basiparameres truncate or emarginate (Fig. 6)
S6 prominent (Fig. 6)
digitus large (Fig. 2)
S6 with lip (Fig. 2)
hypopygium dorsal carinae
S6 with lip (Fig. 2)
hypopygium with stout apical spine (Fig. 2)
S5 carinate/rugose (Fig. 2)
paramere shape and setal patterns (Figs 1, 3, 4)
pygidial shape, carinae, setae (Figs 1, 5)
Table 2 wasps.
Morphological association between the genitalia and surrounding sclerites in male thynnine
Genitalia
Sclerites
basiparameres
internal surface of the hypopygium basally
ventral margin of parameres
apical margin of hypopygium
basal angle of basiparameres
apex of epipygium
ventral margin of parameres
apex of epipygium
size of genitalia
degree of scleritisation of adjacent sclerites
internal margin of hypopygium
internal margin of epipygium
Table 3
Structures that favour increased or decreased movement whilst in copula.
Increased
Decreased
lengthening of metasoma (especially in the male)
presence of tubercles on sterna 1 and 2
presence of a ball and socket joint with the basiparamere and hypopygium of the male
shape of the propodeum and tergum 1
weakly sclerotised metasomal segments in the male
presence of longitudinal carinae on epipygium
large gap between sterna 1 and 2 transverse carinae on epipygium (especially if below a preapical plate)
to manipulate the pheromones they release to attract males of other species. However, as yet nothing is known about thynnine pheromones. One advantage to the female of a miscoupling is that it would allow her to be transported to a food source. It does not matter to the individual female if successful insemination occurs or not as unsuccessfully mated wasps produce males. However, it is unclear what benefits, if any, the males would receive from this behaviour. If such a mechanism does exist, then miscoupling could be a common occurrence within the subfamily. This has been suggested by several authors (Turner 1907, 1910; Rohwer 1910; Given 1959) and includes a number of non-conspecific type pairs as listed by Given (1954), however, more recently, only one instance of miscoupling has been confirmed (Brown 1993).
Some Problems with Australian Tiphiid Wasps 217
References Adams, P. B. & Lawson, S. D. (1993) Pollination in Australian orchids: a critical assessment o the literature 1882-1992. Australian Journal of Botany 41: 553-75. Bower, C. C. (1996) Demonstration of pollinator-mediated reproductive isolation in sexually deceptive species of Chiloglottis (Orchidaceae: Caladeniinae). Australian Journal of Botany 44: 15-33. Brown, G. R. (1993) A new species of Lestricothynnus with notes on miscoupling in Thynninae (Hymenoptera: Tiphiidae). Journal of the Australian Entomological Society 32: 197-199. Brown, G. R. (1997) Arthrothynnus, a new genus of orchid-pollinating Thynninae (Hymenoptera: Tiphiidae). The Beagle, Records of the Museums and Art Galleries of the Northern Territory 13: 73-82. Brown, G. R. (1998) Revision of the Neozeleboria cryptoides species group of thynnine wasps (Hymenoptera: Tiphiidae): pollinators of native orchids. Australian Journal of Entomology 37: 193-205. Given, B. B. (1953) General report on a search for parasites of Melolonthinae in Australia. New Zealand Journal of Science and Technology (B) 34: 322-340. Given, B. B. (1954) A catalogue of the Thynninae (Tiphiidae, Hymenoptera) of Australia and adjacent areas. New Zealand Department of Scientific and Industrial Research Bulletin 109: 1-89. Given, B. B. (1959) Notes on Australian Thynninae. II. The genera Dimorphothynnus, Rhagigaster and Eirone. Proceedings of the Linnean Society of New South Wales 83: 309-26. Ridsdill-Smith, T. J. (1970) The biology of Hemithynnus hyalinatus (Hymenoptera: Tiphiidae), a parasite on scarabaeid larvae. Journal of the Australian Entomological Society 9: 183-195. Rohwer, S. A. (1910) Turner’s Genera of Thynnidae with notes on Ashmeadian Genera. Entomological News 21: 345-351. Turner, R. E. (1907) A revision of the Thynnidae of Australia (Hymenoptera). Pt I. Proceedings of the Linnean Society of New South Wales 32: 206-290. Turner, R. E. (1910) Additions to our knowledge of the fossorial wasps of Australia. Proceedings of the Zoological Society of London 1910: 253-356.
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps (Hymenoptera: Cynipidae: Cynipini) George Melika1 and Warren G. Abrahamson2 1
Systematic Parasitoid Laboratory, Kelcz-Adelffy St. 6, Köszeg 9730 Hungary (email:
[email protected]) 2
Department of Biology, Bucknell University, Lewisburg, PA 17837 USA
Introduction The Cynipoidea includes approximately 3000 described species worldwide. Both structural and biological characteristics have been used to divide the superfamily into two groups, ‘macrocynipoids’ and ‘microcynipoids’ (Ronquist 1995). The latter group includes the Cynipidae which comprises six tribes (Ronquist 1994, 1995; Nieves Aldrey 1994; Fergusson 1995), all of which except the Aylacini are demonstrably monophyletic (Nieves Aldrey 1994; Ronquist 1994, 1995; Liljeblad & Ronquist 1998). Members of the other five tribes (Aylacini, Diplolepidini, Eschatocerini, Pediaspidini and Cynipini) induce galls while the Synergini comprises inquilines, the larvae of which are phytophagous and utilise the galls of other cynipids. Kinsey (1920) first considered the Cynipini (oak gall wasps) to be monophyletic, the members of which are associated only with Quercus or closely related genera in the Fagaceae. They are also defined by having a short pronotum and opened radial cell (Weld 1952; Ronquist 1994; Liljeblad & Ronquist 1998), synapomorphies that also occur in the Diplolepidini and Eschatocerini. The clear definition of genera is a significant problem within the Cynipidae, particularly for the Cynipini which comprises 85% (about 750–800 species) of the family, because of the presence of alternate unisexual and bisexual generations in many genera, which are morphologically different among adults. The existing keys to the world genera of Cynipini (Dalla Torre & Kieffer 1910; Weld 1952) are out of date and unreliable in that they often employed characters now known to be variable. These included the structure and location of a gall or novel host association which were more heavily weighted in the description of new genera than were morphological characteristics of the adult. For example, the North American genus Heteroecus Kinsey was established on the basis of its association with Quercus chrysolepis, thus assuming gall structure and/or host-plant association were unique features. While this is the case for many species, our research shows that different species can induce galls with similar structure, especially for those that cause catkin, bud and stem swelling-like galls. Many North American species of Callirhytis, for example C. cornigera (Osten-Sacken), C. quercusclavigera (Ashmead) and C. quercuspunctata (Bassett), induce stem swelling-like galls that are similar in size and shape, although the species can be separated on their adult morphology. Many Atrusca spp. in south-western USA and Mexico (e.g. A. brevipennata (Gillette), A. bella (Bassett), A. capronae (Weld) and A. catena (Kinsey)) induce similar leaf galls on the same species of oaks (Kinsey 1936; Weld 1960). Many European species of Andricus
218
Hymenoptera: Evolution, Biodiversity and Biological Control
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 219
induce catkin and bud galls in their bisexual generations that are structurally similar. Furthermore, the structure of galls caused by the same species can vary with the host oak species attacked or their geographic location. For example, galls of A. solitarius (Fonscolombe) in Europe are densely pubescent when they develop on Quercus pubescens and are smooth and glabrous on Q. robur. The unisexual leaf galls of Cynips quercusfolii L. are small with a smooth surface when development occurs on Q. petraea, but they are larger with a course irregular surface on Q. robur. Galls of A. lucidus (Hartig) on Corfu (Greece) and in Turkey are usually twice as large as those in Central Europe. Here we review the history of higher level taxonomic research on the Cynipidae and particularly the Cynipini, and in so doing outline problems associated with previous studies. We contend that the principal criteria for the diagnosis of genera and species should be characters derived from adult morphology. Accordingly, the taxonomic potential of several character systems is reviewed. N.B. Unless included in the text, the authors of genera are given in Table 1.
Historical Review of Cynipidae From the time of Linnaeus to the 1830’s, the systematics and classification of the Cynipidae was confused to the point of mixing species from different superfamilies, viz. Cynipoidea and Chalcidoidea. The generic name Cynips originated in ‘Systema Naturae’ (Linnaeus 1758) with eight described gall-inducing species. The remainder of his Cynips species were chalcid wasps. Without doubt, Linnaeus erected Cynips to accommodate gall-inducing wasps in order to separate them from ‘Tenthredo’ where they were placed originally along with other large Hymenoptera (Linnaeus 1746). Various chalcids were also placed under the name Cynips because of their morphological similarity to cynipids rather than to other ‘macro’ Hymenoptera. Later, Geoffroy (1762) erected the genus Diplolepis for the true cynipid gall-inducing wasps, with Cynips applied to parasitic species. In all editions of Linnaeus’ ‘Systema Naturae’, the genus Cynips began with gall-inducing cynipids and were followed by the parasitoids of gall inducers. Unlike Geoffroy’s use of Diplolepis, Linnaeus never used the latter name to differentiate gall-inducers from parasitoids. Fourcroy (1785), Olivier (1790, 1791) and Latreille (1805, 1810) followed Geoffroy by placing gall-inducers into Diplolepis and parasitoids into Cynips, while Fabricius (1804), Panzer (1806) and Spinola (1808) used the name Cynips for gall-inducing cynipids, and Diplolepis for chalcid parasitoids. Westwood (1829) strongly criticised the cynipid-chalcid classification, particularly the placement of chalcids and cynipids in the same genus. The keys of Foerster (1869), Mayr (1881) and Ashmead (1903) did not mention Diplolepis. Dalla Torre (1893) considered Diplolepis as a synonym to Cynips, however this was not accepted and Dalla Torre and Kieffer (1910) restored Diplolepis as a valid genus. Rohwer and Fagan (1917) synonymised Rhodites Hartig with Diplolepis, while Belizin (1961) and Kinsey (1929) considered Diplolepis a synonym of Cynips. Foerster (1869) later designated Cynips rosae L. as the type species of Rhodites. All the species inducing galls on roses were placed into this genus, while Kinsey and Ayres (1922) were the first to use Diplolepis for rose gall-inducing cynipids. Given this convoluted history, the validity of the name Diplolepis must be carefully considered and requires special study. Westwood (1840) was the first author who separated the genus Biorhiza Westwood from Cynips. Later, Hartig (1840a, 1840b) added eight new genera of gall-inducing cynipids: Andricus, Aylax
George Melika and Warren G. Abrahamson 220
Hymenoptera: Evolution, Biodiversity and Biological Control
Hartig, Diastrophus Hartig, Neuroterus, Rhodites, Spathegaster Hartig, Teras Hartig and Trigonaspis, as well as two genera of inquilines, Ceroptres Hartig and Synergus Hartig. Giraud (1859) erected Dryocosmus, Foerster (1869) established several other new genera from Europe, while Fitch (1859), Reinhard (1865) and Ashmead (1881, 1887, 1897a, 1897b) described 14 new genera of Cynipini from North America. Adler’s (1881) discovery of alternating generations in cynipids necessitated a taxonomic revision of genera. Mayr (1881) included a key to 29 genera of Cynipidae including 14 genera that belonged to the Cynipini and descriptions of seven new genera. Dalla Torre (1893) systematised the classification of cynipids in his ‘Catalogue Hymenopterorum – Cynipidae’. Later, Ashmead (1903) provided a key to 33 genera of Cynipini and a synopsis of world Cynipidae. ‘Das Tierreich – Cynipidae’ was completed by Dalla Torre and Kieffer (1910) and, although part of their classification has stood the test of time, a large number of taxonomic changes have subsequently been made to their work. Many of these changes to the generic classification of the Cynipini have been made by Kinsey (e.g. 1929, 1936, 1937) and Weld (e.g. 1921, 1951, 1952, 1957, 1959, 1960). However, the taxonomic approach of these two workers was antipodal: Kinsey was a ‘lumper’ at both the generic and specific levels, while Weld tended to split taxa. For instance, Kinsey’s concept of Cynips (Kinsey 1929, 1936) was treated by Weld as eight distinct genera. The current generic classification of world Cynipini follows Weld (1952). However, we consider Weld’s classification to be artificial and in need of substantial alteration, given that very little was known about the alternation of generations in North American cynipids at that time. Detailed studies of the alternation of generations for North American cynipids have been completed during the past several decades (Doutt 1959, 1960; Dailey 1969; Dailey & Sprenger 1973a, 1973b; Dailey et al. 1974; Evans 1967, 1972; Lyon 1959, 1963, 1964, 1969, 1970). This work has increased our understanding of gall-inducing cynipids and provided a solid background to establish a more natural classification for the Cynipini, particularly for taxa restricted to North America which were previously less well understood than for the Palaearctic fauna. Weld (1952) distinguished 39 genera of Cynipini. Lyon (1993) synonymised Xystoteras with Phylloteras and described Euxystoteras which differs from Phylloteras only by having simple tarsal claws. Maisuradze (1961) reported a new genus, Repentinia Belizin & Maisuradze from Azerbaijan, and Monzen (1954) described Neoneuroterus from Japan. Both genera are closely related to Neuroterus and/or Trichagalma, however their status remains uncertain as we have been unable to examine the types. Kovalev (1965) described two new genera, Belizinella and Ussuraspis, from Far East Russia which are closely related to the Nearctic Xanthoteras and Xystoteras, respectively, as well as to Trigonaspis from the Palaearctic. Later, Melika and Abrahamson (1997) described Eumayriella from eastern USA. Since Linnaeus, 74 genera of Cynipini have been described, of which 44 are currently recognised as valid.
Morphology and Classification of Cynipini The classification of cynipids has become confused in part by the use of incorrect terms for morphological structures. The nomenclature used to describe taxonomically important structures prior to the 1960’s is inconsistent (Dailey & Menke 1980). Various authors have used different terminology to define the same morphological structure and have employed the same term for different structures. As a consequence, species have been misunderstood and misidentified. Weld’s publications on oak gall cynipids are the most important for North America. Unfortunately, his terminology is seldom descriptive and sufficiently precise. Much of the most recent
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 221
Table 1 Division of the Cynipini based on the presence or absence of a basal lobe on the tarsal claws (after Weld 1952) (*genera which include species in both groups). Genera with toothed claw
Genera with a simple claw
Acraspis Mayr
Aphelonyx Mayr
Adleria Rohwer & Fagan
Bassettia Ashmead
Amphibolips Reinhard
Belonocnema Mayr
Andricus Hartig
Biorhiza Westwood
Antron Kinsey
Callirhytis Foerster
Atrusca Kinsey
Chilaspis Mayr
Besbicus Kinsey
Dryocosmus Giraud
Cynips L.
Erythres Kinsey
Disholcaspis Dalla Torre & Kieffer
Eumayria Ashmead
Dros Kinsey
Euxystoteras Lyon
Liodora Foerster
Fioriella Kieffer
Paracraspis Weld
Heteroecus Hartig
Parandricus Kieffer*
Holocynips Kieffer
Philonix Fitch
Loxaulus Mayr
Phylloteras Ashmead
Neuroterus Hartig*
(=Xystoteras Ashmead)
Odontocynips Kieffer
Trichoteras Ashmead
Parandricus Kieffer*
Trigonaspis Hartig*
Plagiotrochus Mayr
Xanthoteras Ashmead
Sphaeroteras Ashmead Trichagalma Mayr Trigonaspis Hartig* Trisoleniella Rohwer & Fagan Zopheroteras Ashmead
Genera described after 1952 Belizinella Kovalev
Belizinella Kovalev
Neoneuroterus Monzen
Eumayriella Melika & Abrahamson
Repentinia Belizin & Maisuradze
Ussuraspis Kovalev
research has followed ‘Weldian terminology’ (Dailey & Menke 1980) and, consequently, must be re-examined in light of modern concepts for the group (i.e. Menke 1993; Gibson 1985; Fergusson 1988; Ronquist & Nordlander 1989). The Cynipini can be easily distinguished from other Cynipidae by the following characters: pronotum dorsally very short, without truncation or pits; mesopleuron without longitudinal furrow; radial cell open; ventral spine of hypopygium never ploughshare-shaped (as in Diplolepidini) (Weld 1952; Eady & Quinlan 1963). With the exception of Neuroterus, the systematic arrangement of world genera is based on the presence or absence of a tooth on the tarsal claws and this divides the tribe into two groups (Table 1) (Figs 1–9). Use of this character within the Cynipini dates from Ashmead’s (1903) key to genera, although Mayr (1881), Ashmead (1885), Dalla Torre (1893) and Kieffer (1897–1901)
George Melika and Warren G. Abrahamson 222
Figures 1–13
Hymenoptera: Evolution, Biodiversity and Biological Control
1-9, tarsal claws: 1) Amphibolips quercuscinerea (Ashmead); 2) Andricus quercusfoliatus (Ashmead); 3) Cynips quercusfolii L.; 4) Cynips divisa Hartig; 5) Neuroterus quercusbaccarum (L.); 6) Neuroterus tricolor (Hartig), unisexual female; 7) N. tricolor, bisexual female; 8) Belonocnema quercusvirens (Osten Sacken); 9) Sphaeroteras ocala (Weld); 10-13, mesosoma, dorsal view: 10, Andricus caputmedusae (Wachtl); 11) N. tricolor (scutellar foveae absent); 12) Loxaulus masneri Melika & Abrahamson (notauli incomplete); 13) Andricus seckendorffi (Wachtl) (notauli complete) (apl = anterior parallel line (anteroad median line); mms = median mesoscutal sulcus (median mesoscutal line); n = notaulus; pl = parapsidal line; sc = scutum (mesoscutum); scf = scutellar fovea; scl = scutellum; tsa = transscutal articulation). Scale lines: 1-8, 13 = 0.1 mm; 9 = 0.5 mm; 10 = 1.0 mm; 12 = 0.25 mm.
used the presence/absence of tarsal claws to recognise subgenera of Andricus. However, this character has been treated too simply in the past, and the term ‘basal lobe’ is more precise. In some cases the claws have a strong basal lobe, for example in Amphibolips and many Andricus. However, in numerous species, including those that Weld (1952) treated as possessing a tooth, no distinct tooth in fact occurs on the claws. Further, a number of genera are variable for this character. For instance, the majority of Neuroterus spp. have tarsal claws without a basal lobe,
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 223
however N. quercusbaccarum L., N. numismalis Olivier and N. petioliventris Hartig have a basal lobe in unisexual generations but simple claws in the alternate bisexual generations. The same is true for Callirhytis, in which both generations of the European species C. glandium (Giraud) have a basal lobe as does the unisexual generation of C. bella (Dettmer) (Nieves Aldrey 1992). One species of Trigonaspis, T. megaptera (Panzer), has a weak basal lobe, while other representatives of this genus have simple claws. The bisexual generations of the North American genus Xystoteras have a very weak basal lobe, and in Parandricus mairei Kieffer females have obvious lobed claws while the males have simple claws. In Belizinella one species, B. gibbera Kovalev, has claws with a basal lobe while other species, such as B. vicina Kovalev have simple claws (Kovalev 1965). The presence or absence of a basal lobe is a likely homoplasy and probably evolved separately in different cynipine genera. The ancestral condition is a simple tarsal claw without a basal lobe, based on its appearance in the majority of genera from the Aylacini. However, even in this tribe, Xestophanes Foerster, Diastrophus Hartig and Gonaspis Ashmead have a basal lobe. Furthermore, the tarsal claws of Synergini inquilines can be simple or lobed, as in Synergus umbraculus Olivier and Periclistus brandtii Ratzeburg. Consequently, we propose to avoid the use of this character to diagnose genera of Cynipini.
Problem characters Weld utilised a number of primary characters to separate genera in his classification. These included: 1) the completeness of the notauli (Figs 10–13); 2) the number of antennal flagellomeres; 3) the pubescence of the thorax and the shape of the scutellar foveae, and 4) the shape of the terga and their pubescence. Our analysis of these characters below shows considerable variability at both the specific and intraspecific levels, and very often their use alone is insufficient for diagnoses of genera. Completeness of the notauli. The genus Andricus is supposed to comprise species with complete notauli that distinctly reach the pronotum, however many European species have incomplete notauli (cf. Figs 10–13). For instance, notauli are absent or very indistinct over the anterior onethird to a quarter of the scutum in the unisexual females of A. lignicolus (Hartig), A. gallaetinctoriae (Olivier), the bisexual females of A. grossulariae Giraud, and A. kollari (Hartig). Moreover, the completeness of the notauli can vary intraspecifically as it does in the previous species. The bisexual females of Acraspis and Cynips were separated by Weld (1952) on the basis of the notauli. However, some specimens of A. gemula (Bassett) have obvious complete notauli and, thus, are congeneric with Cynips on the basis of this character. Number of antennal flagellomeres. The suture between the terminal flagellomeres in representatives of various genera can be distinct or very indistinct so that the number of flagellomeres appears to vary. Pubescence of thorax and shape of scutellar foveae. The pubescence of the thorax can vary strongly among species, for instance within the large genus Andricus. Typically, unisexual females are more strongly pubescent than bisexual forms. The same is true of the shape of the scutellar foveae which can vary intraspecifically, for example in Sphaeroteras ocala (Weld) where some specimens have distinct scutellar foveae while others have them completely lacking. Shape of terga and their pubescence. Both of these characters vary strongly at the intraspecific level and, thus, cannot be used for the separation of genera. The shape of the terga depends on egg production in females and consequently its shape is an artifact. Tergal pubescence can vary strongly at the specific level as, for instance, among Andricus spp.
George Melika and Warren G. Abrahamson 224
Hymenoptera: Evolution, Biodiversity and Biological Control
Figures 14–24 14, metasoma in lateral view, Loxaulus huberi Melika & Abrahamson; 15-22, ventral spine of the hypopygium: 15) Cynips quercusfolii L., unisexual female; 16) Cynips longiventris Hartig, bisexual female; 17) Neuroterus quercusbaccarum (L.), unisexual female; 18) Neuroterus numismalis (Geoffroy), bisexual female; 19) Neuroterus laeviusculus (Schenck), bisexual female; 20) Andricus solitarius Fonscolombe; 21) Andricus quercusramuli (L.); 22) Sphaeroteras ocala (Weld); 23, 24, mesosoma, dorsal view: 23) Bassettia pallida Ashmead (scutum transversely rugose); 24) Loxaulus huberi Melika & Abrahamson (scutum reticulate) (hyp = hypopygium; ov = ovipositor; sas = subapical setae; vsp = ventral spine of the hypopygium). Scale lines = 0.5 mm.
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 225
Figures 25–30 25-29, propodeum: 25) Loxaulus huberi Melika & Abrahamson (with median propodeal carina); 26) Andricus quercuscorticis (L.); 27) Andricus quercusradicis (F.); 28) Andricus testaceipes Hartig; 29) Bassettia pallida Ashmead; 30) forewing, L. huberi (length = length of radial cell; w = width of radial cell). Scale lines: 25 = 0.2 mm; 26-28 = 0.25; 29 = 0.4 mm; 30 = 1.9 mm.
Informative characters Our preliminary research has shown a number of characters are not intraspecifically variable and have significant potential for distinguishing genera and/or groups of genera. These are 1) the shape of the ventral spine of the hypopygium and the direction and length of subapical setae on the spine (Figs 14–22); 2) the shape and sculpturing of the mesosoma (Figs 12, 13, 23, 24); 3) the structure of the propodeum (Figs 25–29), and 4) the shape of the radial cell and Rs of the fore wing (Fig. 30). Also potentially informative for some genera, are the ratio of lengths of the hind tarsomeres, the shape of F1, and the ratio of F1 to the scape and pedicel.
George Melika and Warren G. Abrahamson 226
Hymenoptera: Evolution, Biodiversity and Biological Control
We can divide the Cynipini into several groups of genera on the basis of the shape of the ventral spine of the hypopygium and the length and direction of the subapical setae on the spine. We differentiate three principal states of the ventral spine of the hypopygium and its setae as follows: 1) ventral spine short, broadest at apex or equally broad through its entire length, the prominent portion of the spine shorter or equal to its width; subapical setae dense and long, directed backwards and reaching far beyond the apex of the spine, which in some genera form a dense truncate tuft; 2) ventral spine thin, slender, needle-like, if short then distinctly but gradually tapering to a point, prominent portion longer than broad; subapical setae long and reaching far beyond the apex of the spine, and 3) ventral spine needle-like, long; prominent portion much longer than broad; subapical setae sparse, short, directed at a nearly straight angle to the spine, usually present ventrally only, never reaches beyond the apex of the spine. The shape and sculpturing of the mesosoma are important characters for the separation of generic groups. A dorso-ventrally compressed mesosoma separates Bassettia, Callirhytis (in part, the unisexual generations only), Eumayria, Eumayriella, and Plagiotrochus (unisexual generations mainly). The scutum is transversely rugose in Bassettia, Callirhytis and Plagiotrochus; it is coarsely rugose and dull in Amphibolips, while in the other genera the scutum is reticulate or smooth and shiny. A median propodeal carina is present only in Loxaulus, Plagiotrochus, and some species of Callirhytis (sometimes a very indistinct, incomplete and fragmented median propodeal carina is visible in some specimens of Bassettia), while this characteristic is absent in all other genera. The shape of the radial cell and Rs of the fore wing are also important diagnostic characters and have been used previously by Weld (1952). In Atrusca, Belonocnema and Loxaulus the 2nd abscissa of Rs is strongly angulate, the radial cell is only 2.0–2.5 times longer than broad, the fore wing has dark spots and/or dark stripes along the veins, while all other genera have the 2nd abscissa of Rs straight or very slightly angulate, and the radial cell is more elongate, at least 2.7–3.0 times longer than broad. The only exceptions to this are three species of Holocynips, H. badia (Bassett), H. hartmanni (Weld) and H. maxima (Weld), which have a short radial cell. Additional diagnostic characters to those discussed above will need to be employed to convincingly separate genera and demonstrate their monophyly. The fact that bisexual females and males are morphologically more uniform than unisexual females makes it difficult to adequately separate genera on the former generation alone. Clearly, a stable generic classification that reflects phylogenetic relationships within the Cynipini will only be forthcoming after a comprehensive dataset is assembled which comprises informative characters from both generations.
Acknowledgements We express our deepest appreciation to A. Menke for his valuable suggestions during our work in the USNM, Washington, D. C. and R. Lyon for his suggestions on the classification of oak gallinducing cynipids. We thank J. Abrahamson, C. Abrahamson, R. Bowman, G. Csóka, R. Hammer, A. Johnson, I. Kralick, R. Peet, R. Roberts, P. Schmaltzer, A. Schotz, R. Scrafford, C. Winegarner, and M. Winegarner for field and technical assistance. We also thank D. Notton, N. Fergusson, G. McGaven, S. Schödl, L. Zombori, J. Papp, O. Kovalev, L. Djakonchuk, J. Nieves Aldrey, J. Pujade i Villar, G. Stone, Gy. Csóka, C. Thuróczy for loans and/or gifts of research material. We also thanks Ms. Eva Foki (Systematic Parasitoid Laboratory, Koszeg, Hungary) for the illustrations. Support was provided to G.M. and W.G.A. by Bucknell University’s David Burpee Endowment and the
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 227
Archbold Biological Station, to W.G.A. by N.S.F. Grant BSR-9107150, and to G.M. by the Smithsonian Institution National Museum of Natural History.
References Adler, H. (1881) Über den Generationswechsel der Eichen-Gallwespen. Zeitschrift fur wissenschaft im Zoologie 35: 151-246. Ashmead, W. H. (1881) On the cynipidous galls of Florida. Transactions of the American Entomological Society 9: 9-28. Ashmead, W. H. (1885) On the cynipidous galls of Florida with descriptions of new species. Transactions of the American Entomological Society 12: 5-9. Ashmead, W. H. (1887) On the cynipidous galls of Florida, with descriptions of new species and synopses of the described species of North America. Transactions of the American Entomological Society 14: 125-158. Ashmead, W. H. (1897a) Description of some new genera in the family Cynipidae. Psyche 8: 67-69. Ashmead, W. H. (1897b) Descriptions of five new genera in the family Cynipidae. Canadian Entomologist 29: 260-263. Ashmead, W. H. (1903) Classification of the gall-wasps and the parasitic cynipoids, or the superfamily Cynipoidea. III. Psyche 10: 140-155. Belizin, V. I. (1961) The oak gall wasps of the genus Cynips (Hymenoptera, Cynipidae). Zoologicheskij Zhurnal 40: 207-213. [in Russian] Dailey, D. C. (1969) Synonymy of Dryocosmus attractans (Kinsey) and Callirhytis uvellae Weld. Pan-Pacific Entomologist 45: 132-134. Dailey, D. C. & Menke, A. S. (1980) Nomenclatorial notes on North American Cynipidae (Hymenoptera). Pan-Pacific Entomologist 56: 170-174. Dailey, D. C. & Sprenger, C. M. (1973a) Unisexual generation of Andricus atrimentus (Hymenoptera: Cynipidae). Pan-Pacific Entomologist 49: 171-173. Dailey, D. C. & Sprenger, C. M. (1973b) Synonymy of Andricus gigas and the bisexual generation of Andricus crenatus (Hymenoptera: Cynipidae). Pan-Pacific Entomologist 49: 188-191. Dailey, D. C., Perry, T. & Sprenger C. M. (1974) Biology of three Callirhytis gall wasps from Pacific Slope Erythrobalanus oaks (Hymenoptera: Cynipidae). Pan-Pacific Entomologist 50: 61-67. Dalla Torre, de C. G. (1893) Catalogus Hymenopterorum Hucusque Descriptorum Systematicus et Synonymicus. Cynipidae. Lipsiae. Sumptibus Guilelmi Engelmann 2: 1-140. Dalla Torre, K. W. von & Kieffer, J. J. (1910) Cynipidae. Das Tierreich 24: 1-891. Doutt, R. L. (1959) Heterogeny in Dryocosmus (Hymenoptera, Cynipidae). Annals of the Entomological Society of America 52: 69-74. Doutt, R. L. (1960) Heterogeny in Andricus crystallinus Bassett (Hymenoptera: Cynipidae). PanPacific Entomologist 36: 167-170. Eady, R. D. & Quinlan, J. (1963) Hymenoptera. Cynipoidea. Handbooks for the Identification of British Insects. Royal Entomological Society, London. Vol. 8, Part I(a): 1-81. Evans, D. (1967) The bisexual and agamic generations of Besbicus mirabilis (Hymenoptera: Cynipidae), and their associate insects. Canadian Entomologist 99: 187-196. Evans, D. (1972) Alternate generations of gall cynipids (Hymenoptera: Cynipidae) on Garry oak. Canadian Entomologist 104: 1805-1818. Fabricius, I. C. (1804) Systema Piezatorum. Ordines, Genera, Species. Brunsvigae, Apud Carolum Reichard.
George Melika and Warren G. Abrahamson 228
Hymenoptera: Evolution, Biodiversity and Biological Control
Fergusson, N. D. M. (1988) A comparative study of the structures of phylogenetic importance of female genitalia of the Cynipoidea (Hymenoptera). Systematic Entomology 13: 13-30. Fergusson, N. D. M. (1995) The cynipoid families. pp. 247-265. In Hanson, P. E. & Gauld, I. D. (Eds.), The Hymenoptera of Costa Rica. Oxford University Press, Oxford. Fitch, A. M. D. (1859) Fifth report on the noxious and other insects of the state of New York. Insects infesting deciduous forest trees. Annual Reports of the New York State Agricultural Society 18: 781-784. Foerster, A. (1869) Ueber die Gallwespen. Verhandlungen der Zoologische-Botanische Gesselschaft 19: 327-370. Fourcroy, A. F., de. (1785) Entomologia Parisiensis; sive Catalogus Insectorum quae in Agro Parisiensi reperiuntur; Secundum Methodum Geoffraeanam in Sectiones, Genera & Species Distributus. Pars. 2. Sectio 3. Classis Insectorum. LXXXVI. Cynips. Le Cinips. LXXXVII. Diplolepis. Le Diplolepe. Hymenoptera Parisiis, Via et Aedibus Serpentineis. pp. 379-393. Sub Privilegio Academiae. Geoffroy, L. (1762) Histoire abrÈgée des insectes qui se trouvent aux environs de Paris. Dans laquelle ces Animaux sont rangés suivant un Ordre méthodique. 2. Paris. Gibson, G. A. P. (1985) Some pro- and mesothoracic structures important for phylogenetic analysis of Hymenoptera, with a review of terms used for the structures. Canadian Entomologist 117: 1395-1443. Giraud, J. (1859) Signalements de quelques especes nouvelles de Cynipides et de leurs galles. Verhandlungen der Zoologischen-Botanischen Gesellschaft 9: 337-374. Hartig, T. (1840a) Ueber die Familie der Gallwespen. III. Zeitschrift für Entomologie 2: 176-209. Hartig, T. (1840b) Erster Nachtrag zur Naturgeschichte der Gallwespen. Zeitschrift für Entomologie 2: 322-358. Kieffer, J. J. (1897-1901) Monographie des Cynipides d’Europe et d’Algerie. Ibalynae et Cynipinae. Paris. Librairie Scientifique, A. Hermann. Kinsey, A. C. (1920) Phylogeny of cynipid genera and biological characteristics. Bulletin of the American Museum of Natural History 42: 357-402. Kinsey, A. C. (1929) The gall wasp genus Cynips: A study in the origin of species. Indiana University Studies 84-86: 1-517. Kinsey, A. C. (1936) The origin of higher categories in Cynips. Indiana University Publications. Science Series 4. Entomological Series 10: 1-334. Kinsey, A. C. (1937) New Mexican gall wasps (Hymenoptera, Cynipidae). Revue de Entomologia 7: 39-79. Kinsey, A. C. & Ayres, K. D. (1922) Varieties of a rose gall wasp (Cynipidae, Hymenoptera). Indiana University Studies 53: 142-171. Kovalev, O. V. (1965) Gall wasps (Hymenoptera, Cynipidae) from the south of the Soviet Far-East. Revue d’Entomologie de l’URSS 44: 46-73. Latreille, P. A. (1805) Histoire Naturelle, generale et particuliere, des Crustaces et des Insectes. Famille Soixante-Deuxieme. Diplolepaires; diplolepariae. Famille Soixante-Troisieme. Cinipseres; cinipsera. Paris, De l’Imprimerie de F. Dufart 3: 196-225. Latreille, P. A. (1810) Considerations Generales sur l’ordre Naturel des Animaux. Composant les Classes des Crustaces, des Arachnides, et des Insectes. Ordre V. Hymenopteres. Hymenoptera. pp. 279-437. Paris, F. Schoell. Liljeblad, J. & Ronquist, F. (1998) A phylogenetic analysis of higher-level gall wasp relationships (Hymenoptera: Cynipidae). Systematic Entomology 23: 229-252.
Historical Review and Current State of the World Generic Classification of Oak Gall Wasps 229
Linnaeus, C. (1746) Fauna Svecica. Sistens Animalia Sveciae Regni: Quadrupedia, Aves, Amphibia, Pisces, Insecta, Vermes, Distributa per Classes & Ordines, Genera & Species. V. Hymenoptera. pp. 282-297. Stockholmiae, Sumtu & literis Laurentii Salvii. Linnaeus, C. (1758) Systema Naturae per Regna tria Naturae, Secundum Classes, Ordines, Genera, Species, cum characteribus, differentiis, synonymis, locis. Classis V. Insecta. V. Hymenoptera. Tomus I. Editio Decima. pp. 553-583. Holmiae, Impensis Direct. Laurentii Salvii. Lyon, R. J. (1959) An alternating, sexual generation in the gall wasp Callirhytis pomiformis (Ashm.) (Hymenoptera, Cynipidae). Bulletin of the California Academy of Sciences 58: 33-37. Lyon, R. J. (1963) The alternate generation of Heteroecus pacificus (Ashmead) (Hymenoptera, Cynipoidea). Proceedings of the Entomological Society of Washington 65: 250-254. Lyon, R. J. (1964) The alternate generation of Callirhytis agrifoliae (Ashmead) (Hymenoptera: Cynipoidea). Proceedings of the Entomological Society of Washington 66: 193-196. Lyon, R. J. (1969) The alternate generation of Callirhytis quercussuttonii (Bassett) (Hymenoptera: Cynipoidea). Proceedings of the Entomological Society of Washington 71: 61-65. Lyon, R. J. (1970) Heterogeny in Callirhytis serricornis (Kinsey) (Hymenoptera: Cynipoidea). Proceedings of the Entomological Society of Washington 72: 176-178. Lyon, R. J. (1993) Synonymy of two genera of cynipid gall wasps and description of a new genus (Hymenoptera: Cynipidae). Pan-Pacific Entomologist 69: 133-140. Maisuradze, N. L. (1961) Notes on the gallflies (family Cynipidae) which are harmful to oaks in the Lenkoran zone. Uchenije Zapiski Azerbaijanskoho Universiteta. Serija Biologija 1: 21-30. [in Russian] Mayr, G. (1881) Die Genera der gallenbewohnenden Cynipiden. Jahresberichte der CommunalOberrealschule im I. Bezirke 20: 1-38. Melika, G. & Abrahamson, W. G. (1997) Synonymy of two genera (Eumayria and Trisoleniella) of cynipid gall wasps and description of a new genus, Eumayriella (Hymenoptera: Cynipidae). Proceedings of the Entomological Society of Washington 99: 665-674. Menke, A. (1993) Notauli and parapsidal lines: just what are they? Sphecos 24: 9-12. Monzen, K. (1954) Revision of the Japanese gall wasps with the Description of New Genus, Subgenus, Species and Subspecies. Cynipidae (Cynipinae). II. Annual Reports of the Gakugei Faculty, Iwate University 6: 24-38. Nieves Aldrey, J. L. (1992) Revision of the European species of the genus Callirhytis Foerster (Hymenoptera, Cynipidae). Graellsia 48: 171-183. Nieves Aldrey, J. L. (1994) Revision of West-European genera of the tribe Aylacini Ashmead (Hymenoptera, Cynipidae). Journal of Hymenoptera Research 3: 175-206. Olivier, M. (1790) Encyclopedie Methodique. Histoire Naturelle. Insectes. Cinips. Cynips. Paris, Panckoucke 5: 772-792. Olivier, M. (1791) Encyclopedie Methodique. Histoire Naturelle. Insectes. Diplolepe. Diplolepis. Paris, Panckoucke 6: 276-282. Panzer, W. F. (1806) Kritische Revision der Insekten Faune Deutschlands. Cynips. Diplolepis. Nurnberg, Felsseckerschen Buchhandlung. Bd. 2: 92-95. Reinhard, D. (1865) Die Hypothesen über die Fortpflanzungweise bei den eingeschlechtigen Gallwespen. Berlinger Entomologische Zeitung 9: 1-13. Rohwer, S. A. & Fagan, M. M. (1917) The type-species of the genera of the Cynipoidea, or the gall wasps and parasitic cynipoids. Proceedings of the United States National Museum 53: 357-380.
George Melika and Warren G. Abrahamson 230
Hymenoptera: Evolution, Biodiversity and Biological Control
Ronquist, F. (1994) Evolution of parasitism among closely related species: phylogenetic relationships and the origin of inquilinism in gall wasps (Hymenoptera, Cynipidae). Evolution 48: 241-266. Ronquist, F. (1995) Phylogeny and early evolution of the Cynipoidea (Hymenoptera). Systematic Entomology 20: 309-335. Ronquist, F. & Nordlander, G. (1989) Skeletal morphology of an archaic cynipoid, Ibalia rufipes (Hymenoptera: Ibaliidae). Entomologica Scandinavica, Supplement 33: 1-60. Spinola, M. (1808) Insectorum Liguriae. Species Novae aut Rariores. Genuae. Sumptibus Auctoris. Tom II. Fasciculus 2: 1-25. Fasciculus 3. V. G. Cynips (Lin. Fab.) G. Diplolepis (Latr.): 157-167. Fasciculus 4. De Diplolepibus: 209-238. Weld, L. H. (1921) American gallflies of the Family Cynipidae producing subterranean galls on oaks. Proceedings of the United States National Museum 59: 187-246. Weld, L. H. (1951) Superfamily Cynipoidea. pp. 594-654. In Muesebeck, C. F. W., Krombein, K. V. & Townes, H. K. (Eds), Hymenoptera of America North of Mexico. Agricultural Monograph No. 2. United States Government Printing Office, Washington, D. C. Weld, L. H. (1952) Cynipoidea (Hym.) 1905-1950 being a Supplement to the Dalla Torre and Kieffer monograph – the Cynipidae in Das Tierreich, Leiferung 24, 1910 and bringing the systematic literature of the world up to date, including keys to families and subfamilies and list of new generic, specific and variety names. Ann Arbor, Michigan (124 pp. Privately Printed). Weld, L. H. (1957) Cynipid galls of the Pacific slope. Ann Arbor, Michigan (64 pp. Privately Printed). Weld, L. H. (1959) Cynipid galls of the Eastern United States. Ann Arbor, Michigan (124 pp. Privately Printed). Weld, L. H. (1960) Cynipid galls of the southwest. Ann Arbor, Michigan (34 pp. Privately Printed). Westwood, J. O. (1829). On the Chalcididae. The Zoological Journal 4: 3-31. Westwood, J. O. (1840) An introduction to the modern classification of Insects: founded on the natural habits and corresponding organization of the different families. (Synopsis of the Genera of British Insects) 2: 1-587. London.
Australian Hymenoptera in the Spinola Collection: a List of Species M. Generani1,2 and P. L. Scaramozzino1,2 1
Museo Regionale di Scienze Naturali, Via Giolitti 36, I-10123 Torino, Italy
2
present address: Museo di Storia Naturale e del Territorio, Università degli Studi di Pisa, Via Roma 79, I-56011 Calci, Italy (email:
[email protected])
Introduction In the Hymenoptera collection of the Marquis Massimiliano Spinola, which is one of the most important at the Museo Regionale di Scienze Naturali, Torino, there are 377 specimens of Australian origin belonging to 22 families. These specimens were given to Spinola by numerous wellknown naturalists of his time, including Deyrolles, Dupont, Jekel, Reiche and Veranì. A few were also handed down from the collections of Latreille, Serville and Klug. On the pink labels (indicating Australia) on the bottom of each insect-box, Spinola marked the names he had given to each specimen. These names in litteris were later published by Casolari and Casolari Moreno (1978, 1979a, 1979b) but are nomina nuda (see ICZN 1985, Article 51D). The aim of this paper is to provide a brief synopsis of Spinola’s life, describe the Hymenoptera part of his collection which has significant historical importance, and list the material within the collection as an aid to current taxonomic studies.
Life of Massimiliano Spinola Marquis Massimiliano Spinola, count of Tassarolo, Senator of Regno Sardo (Fig. 1), belonged to one of the most noble families of Liguria. A dedicated student of the arts and sciences, he excelled above all in entomology publishing 53 works mainly on Hymenoptera, Coleoptera and Hemiptera, but also on ichthyology, e.g. the fish of the Gulf of Genoa (Gestro 1915; Vidano & Arzone 1978; Passerin d’Entrèves 1980). His life can be summarised as follows: 1780 (10 July): Massimiliano Spinola is born at Pezenas, Hérault, France. 1789: He settles in Paris with his family but the outbreak of the French Revolution forces them to flee to Genoa. 1801: He marries the Marquise Clelia Durazzo who dies of tuberculosis two years later, having given him a son, Agostino. 1805: Spinola publishes his first entomological work Faunae Ligusticae Fragmenta, Decas Ia. 1806: He publishes the first volume of his work Insectorum Liguriae species novae aut rariores and marries his cousin Maria Giulia Spinola. 1808: He publishes the second volume of Insectorum Liguriae. 1822: His second wife passes away, having given birth to three sons, Cristoforo, Massimiliano and Bendinello, and three daughters, Enrichetta, Anna and Maria. 231
M. Generani and P. L. Scaramozzino 232
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Portrait of Marquis Massimiliano Spinola [from Passerin d’Entrèves (1980)].
1833: Jailed for political reasons at the fortress of Alessandria. After his release, Spinola spends some time under surveillance at Tassarolo castle. 1834 to 1851: Spinola devotes himself to his entomological studies and publishes most of his significant works during this time. 1841: He attends the 3rd Meeting of Italian Scientists in Florence. 1842: At the 4th Meeting of Italian Scientists in Padoa, Spinola is elected Vice-President of the Zoology Session and reads his paper on the genus Sirex Fabricius (Hymenoptera). 1846: He attends the 8th Meeting of Italian Scientists at Genoa. His heath is deteriorating and he is unable to participate actively in the meeting. 1848: Spinola is proclaimed Senator of Regno Sardo. 1853: Because of cataracts, he becomes practically blind and gives up his studies to spend the following years at Tassarolo with his children. He bequeaths his Hymenoptera collection to his grand children (of his son Agostino, who had died previously), but he keeps it himself on loan until his death. 1854: Publication of Spinola’s last scientific work. 1857 (12 November): After a cataract operation to avoid total blindness Spinola dies of a stroke.
The Hymenoptera Collection In 1858, a few months after Spinola’s death, his Hymenoptera collection was handed over by his grandchildren to Filippo De Filippis, Director of the Zoological Museum of Turin, to be sold at the best price. As no acceptable offers were forthcoming, Professor De Filippis purchased the collection for the Museum of Turin. In the 1970’s the Spinola’s collection of
Australian Hymenoptera in the Spinola Collection: a List of Species 233
Figure 2
A draw (N. 73) of Australian thynnine wasps (Tiphiidae) from the Spinola Collection, including a number of unpublished ‘new’ species recognised by Spinola.
Figure 3
The lectotype of Scolia soror Smith (Scoliidae) from the Spinola Collection showing the original label in Spinola’s hand-writing.
M. Generani and P. L. Scaramozzino 234
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Original collections or donors and localities of origin of the Australian Hymenoptera in the Spinola Collection. Donors or original collection (C)
Geographic localities
Latreille (C)
Australia, New Holland
19
Serville (C)
New Holland, Van Diemen
7
De Schmidt
New Holland
2
Deyrolles
New Holland, Sidney
73
Dupont
New Holland, Van Diemen’s Land
19
Dupont & Verani
Van Diemen’s Land
2
Jekel
New Adelaide, New Holland, Tasmania
23
Klug
New Holland
5
Rambur
New Holland
2
Reiche
New Holland, Swan River
13
Sturm
New Holland
1
Verani
New Holland, Van Diemen’s Land
11
Voyage de l’Astrolabe
Australasia
1
unknown origin
New Holland, New Adelaide, Sydney
33
Total
No. of species
130
Hymenoptera, originally arranged in 69 containers was reorganised into 139 new boxes by Dr Carlo Casolari (Figs 2, 3). The specific names, authors, the collections of origin or donor’s names, and the geographic localities where the specimens had been collected had been written by Spinola himself on labels on the bottom of each drawer. The majority of Australian specimens had been determined to specific level by Spinola himself. Unfortunately his increasingly precarious health did not allow him to complete his work and its publication. The Australian material of the Spinola collection was probably among the first to reach Europe. The donors and places of origin of the Hymenoptera in his collection are shown in Table 1. Appendix 1 provides a detailed account of the material in the collection.
Acknowledgements We wish to acknowledge with gratitude the assistance of Guido Pagliano for scientific advice and Marina Spini, librarian at Museo Regionale di Scienze Naturali of Turin, for bibliographic research.
References Casolari, C. & Casolari Moreno, R. (1978) Catalogo della collezione imenotterologica di Massimiliano Spinola. Parte 1. Bollettino dei Musei di Zoologia Università di Torino 5: 27-74. Casolari, C. & Casolari Moreno, R. (1979a) Catalogo della collezione imenotterologica di Massimiliano Spinola. Parte 2. Bollettino dei Musei di Zoologia Università di Torino 2: 19-82. Casolari, C. & Casolari Moreno, R. (1979b) Catalogo della collezione imenotterologica di Massimiliano Spinola. Parte 3. Bollettino dei Musei di Zoologia Università di Torino 4: 91-130.
Australian Hymenoptera in the Spinola Collection: a List of Species 235
Gestro, R. (1915) Res Ligusticae LXII. Ricordo di Massimiliano Spinola. Annali del Museo Storia Naturale di Genova 7: 33-53. International Commission on Zoological Nomenclature (1985) International Code of Zoological Nomenclature. Third Edition. International Trust for Zoological Nomenclature, London. Passerin d’Entrèves, P. (1980) La collezione Spinola di Tassarolo. Guide alle Mostre Temporanee Museo Regionale di Scienze Naturali, Torino. Vidano, C. & Arzone, A. (1978) Sulla collezione Spinola conservata nel Castello di Tassarolo. pp. 253-260. Atti XI Congresso Nazionale Italiano di Entomologia, Portici-Sorrento, 10-15 Maggio 1976. Yu, D. S. & Horstmann, K. (1997) A catalogue of world Ichneumonidae. Parts 1 and 2. Memoirs of the American Entomological Institute 58: 1-1558.
M. Generani and P. L. Scaramozzino 236
Hymenoptera: Evolution, Biodiversity and Biological Control
Appendix 1 List of Australian species of Hymenoptera in the Spinola Collection. The information provided is arranged as follows: first, the name given to the species by Spinola and the author who described it (if mentioned); in brackets, the number of the new box in which the species is stored. The complete transcription of the original labels on the bottom of the box, bears the following indications compiled by Spinola and always in the same order: 1st line, name of genus, species and author, possible synonym and supplementary notes; 2nd line, original collection or donor (sometimes missing) and the geographic locality where the material was collected; number of specimens and their sex; description and complete transcription of the labels pinned on the specimens or placed at their side. In the collection sometimes there are many series of specimens of the same species marked with their own label on the bottom of the box. Each of these series is marked with the specific name given by Spinola. When the information given on the label is identical for two or more series, they are not repeated, but only the number and sex of the specimens are shown. Illegible words on labels are shown as (......). Typical series are marked with an asterisk before the name of the species. FAMILY PERGIDAE Perga dorsalis Leach (box 11) Perga dorsalis Leach D. Deyrolles – Sidney 3 Perga ferruginea Leach (box 11) Perga ferruginea Leach D. Deyrolles – Sidney 2 Perga lewisii Westwood (box 11) Perga lewisii Westw. D. Dupont – Van Diemen 3
FAMILY GASTERUPTIIDAE Foenus gigas Spinola in Casolari & Casolari Moreno, 1979 nomen nudum (box 52) Foenus gigas m. D. Jekel 1851 – Nov. Holl. 1 FAMILY MEGALYRIDAE Megalhira fasciipennis Westwood (box 47) Megalhira fasciipennis Westwood D. Verany – N. Holl. 3 FAMILY ICHNEUMONIDAE Anomalon flavitarsus Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 29) Anomalon flavitarsus m. D. Deyrolles – Sidney 1
Pterygophorus cinctus Leach (box 13) Pterygophorus cinctus var. 1.a Leach D. Deyrolles – Sidney 1 the same as above 3, 1, the latter with a round pink label “Pterigophorus N.lle Holl” Pterygophorus cinctus var. 2.da Leach D. Deyrolles – Sidney 3 , one with a round pink label “Pterigophorus N.lle Holl” the same as above 1 Pterygophorus cinctus var. 3.a D. deyrolles – Sidney 3
Ichneumon australis Brullè (box 16) Ichneumon australis Brullè ? D. Deyrolles – Sidney 2
Pterygophorus interruptus Leach (box 13) Pterygophorus interruptus Leach D. Deyrolles – Sidney 3
Ichneumon ischioleucos Brullè (box 16) Ichneumon ischioleucos Brull. ? D. Deyrolles – Sidney 1
Anomalon hollandiae Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 29) Anomalon Hollandiae m. D. Verany – Nouv. Holl. 2, 2
Australian Hymenoptera in the Spinola Collection: a List of Species 237
Mesostenus luperus Brullè (box 33) Mesostenus luperus ? Br. D. Deyrolles – Sidney 2 Ophion australasiae Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 37) Ophion australasiae m. D. Deyrolles – Sidney 2 Ophion bicallosus Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 36) Ophion bicallosus m. D. Deyrolles – Sidney 1 specimen without abdomen Ophion dorsatus Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 37) Ophion dorsatus m. D. Deyrolles – Sidney 1, 2 Ophion merdarius Gravenhorst (box 36) Ophion merdarius Grav. D. Jekel 1851 – Nov. Holl. 1, 2 Paniscus australasiae Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 37) Paniscus australasiae m. D. Deyrolles – Nouv. Holl. 2, one of which with a pink label “productus Brullè” Paniscus australasiae var. D. Dupont – Van-Diemen 2, 1
Pimpla subpetiolata m. Nouv. Holl. 1, 3 Rhyssa maculipennis Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 23) Rhyssa maculipennis m. n. sp. D. Deyrolles – Sidney No specimens *Westwoodia ruficeps Brullè (box 36) Westwoodia ruficeps Br. Coll. Serville D. Deyrolles – N. Holl. 2 Note: The type of this species does not exist at the Natural History Museum in Paris (Yu & Horstmann, 1997). The specimens in the Spinola Collection (originally from the Serville Collection) could be the missing types. FAMILY BRACONIDAE Agathis dimidiata Brullè (box 39) Agathis dimidiata Br. Coll. Serville – Van Diemen 1 Bracon argenteociliatus Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 25) Bracon argenteo-ciliatus D. Jekel 1851 – Nouv. Holl. 1 Bracon capitator Fabricius (box 27) Bracon capitator Fab. Nouvelle Holl. 2, 1
Paniscus difficilis Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 37) Paniscus difficilis m. – Paniscus testaceus var ! D. Deyrolles – Nouv. Holl. 1 with a pink label “productus Brullè”
Bracon dimidiatocinctus Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 27) Bracon dimidiato-cinctus m. inedit. D. Deyrolles – Sidney 1
Pimpla crenator Fabricius (box 22) Pimpla crenator Fab. Nouv. Holl. 2
Bracon hyalinipennis Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 27) Bracon hyalinipennis m. inedit. D. Deyrolles – Sidney 1
Pimpla subpetiolata Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 22) Pimpla subpetiolata D. Jekel 1851 – Nov. Adelaid. 1
Bracon mutator Fabricius (box 27) Bracon mutator Fab. D. Deyrolles – Sidney 1
M. Generani and P. L. Scaramozzino 238
Bracon sericatum Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 25) Bracon sericatum m. D. Jekel 1851 – N. Holl. 1 Bracon tasmaniae Spinola in Casolari & Casolari Moreno, 1978 nomen nudum (box 25) Bracon Tasmaniae m. D. Jekel 1851 – Tasmania 1 FAMILY CHRYSIDIDAE Chrysis lyncea Fabricius (box 51) Chrysis lyncea var. Fab. Nov. Holl. 2 specimens, one with round white label “6608” FAMILY MUTILLIDAE Mutilla dorsigera Westwood (box 62) Mutilla dorsigera Westw. D. Reiche – Swan-Riv. 1 labelled “Ephutomorpha edmondi André” (det. Zavattari) Mutilla D. Deyrolles – Sidney 1 labelled “Ephutomorpha edmondi André “ (det. Zavattari)” Mutilla rugicollis Westwood (box 62) Mutilla rugicollis Westw. D. Deyrolles – Sidney 1 labelled “Ephutomorpha rugicollis Westw.” (det. Zavattari) Mutilla sp. (=Ephutomorpha multicostata André) (box 62) Mutilla D. Deyrolles – Sidney 1 labelled “Ephutomorpha multicostata André” (det. Zavattari) Mutilla sp. (=Ephutomorpha venusta Smith) (box 60) Mutilla D. Reiche – Swan Riv. 1 labelled “Ephutomorpha venusta Smith” (det. Zavattari) *Mutilla tricarinata Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (=Mutilla tricarinata tricarinata Zavattari ) (box 63) Mutilla tricarinata m. Coll. Latr. – Nouv. Holl. 1 with four labels “Mutilla tricarinata Zav.”, “Trogaspidia tricarinata tricarinata (Zav.) B. Petersen det. 1978”, “Holotype tricarinata”, and “Holotype teste B. Petersen 1985”
Hymenoptera: Evolution, Biodiversity and Biological Control
FAMILY SCOLIIDAE *Scolia cyanipennis Fabricius (=Scolia soror Smith) (box 67) Scolia cyanipennis St. Farg. an Fab.? D. de Schmidt – Nouv. Holl. 2, 1, one labelled: “59”, “Cyanipennis Fab. (St. Fargeau olim)” and on a red label “Lectotype soror Smith” Scolia cyanipennis Lepell. an Fab. D. de Schmidt – N. Holl. 1, 1, the labelled: “Scolia cyanipennis foemina”, “Scolia soror f Smith”; the male with a red label “Allotype soror Smith, Betrem 1969” Scolia glabrata Hagemb. (box 67) Scolia glabrata Hagemb., Scolia cyanipennis, Le P. v.? D. Sturm – Nouv. Holl.da 1 labelled “Austroscolia varifrons det. Betrem 1969” Scolia javana Lepeletier (box 70) Scolia javana var. ? D. Reiche – Swan-Riv. 2, one labelled “Trielis anthracina Burm.” Scolia javana, var. D. Reiche – Swan-Riv. 2 Scolia sp. (=Elis tasmaniensis Saussure) (box 68) Scolia, G. Colpa, St. Farg. D. Reiche – Swan-Riv. 3 one labelled “tasmaniensis” Scolia (box 71) D. Reiche – N. Holl.de 2, one labelled “tasmaniensis” Scolia sp. (=Scolia zonata Smith) (box 71) Scolia, Capsomeris, Le Pell. ad colpas accadens Nouv. Holl.da 2, one labelled “zonata” Scolia verticalis Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 71) Scolia verticalis m. n. sp.? D. Jekel 1851 – Nov. Holl. 1 FAMILY TYPHIIDAE Diamma bicolor Westwood (box 73) Diamma bicolor Westw. D. Klug et Dupont – Nouv. Holl.da 4 Psamattra chalybea Shuckard (box 74) Psamattra chalybea Shuck. D. Verani – Nouv. Holl. 1
Australian Hymenoptera in the Spinola Collection: a List of Species 239
Thynnus annulatus (box 73) Thynnus annulatus, Thynnus melleus Westw. D. Deyrolles – Sidney 2
Thynnus rubripes Guérin (box 73) Thynnus rubripes Guér. D. Dupont – Terre de Van-Diamen 3
Thynnus femoralis Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 73) Thynnus femoralis m. n.sp.? Thynnus depressus Westw. D. Verani – Nouv. Holl.de 5
Thynnus senilis Er. (box 73) Thynnus senilis Er. D. Reiche – Swan-Riv. 2, one with a white label with a black border “9”
Thynnus festivus Erichson (box 73) Thynnus festivus Er., Aelurus abdominalis Westw. D. Verani – Nouv. Holl.de 3 Thynnus festivus (box 74) Thynnus festivus, var.? D. Deyrolles – Nouv. Holl., Sidney 2, one of which labelled “Thynnus n.sp.?, Nouv. Holl.” Thynnus flavomaculatus Latreille (box 73) Thynnus Flavomaculatus Latr 1 labelled “6497” and “Thynnus Flavimaculatus Latr. male” Thynnus gravidus Westwood (box 74) Thynnus gravidus Westw. Sidney 1 Thynnus haemorrhoidalis Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 73) Thynnus haemorrhoidalis m. n.sp.? D. Dupont et Verani – Terre de Van-Diemen 4 Thynnus obscuripennis Guérin (box 73) Thynnus obscuripennis, var. Guér. Nouv. Holl.de 2 Thynnus obscuripennis Guér. D. Dupont – Nouv. Holl.de 1 labelled “Is. .......bank, New Holland” Thynnus octomaculatus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 73) Thynnus octomaculatus m. D. Dupont et Verani – Terre de Van-Diamen 1 Thynnus ramburi Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 73) Thynnus Ramburi m. n.sp.? D. Rambur – Nouv. Holl. 3
Thynnus trifidus (box 73) Thynnus trifidus, Agriomya maculata, Guér.? D. Rambur – Nouv. Holl. 2 Thynnus variabilis (box 73) Thynnus variabilis, var. D. Deyrolles – Nouv. Holl. Sidney 2, 1 specimens labelled “........ Deyrolles” Thynnus variabilis, Myrmecoda maculata, Latr Coll. Latr. – Nouv. Holl. de 2 Thynnus variabilis D. Reiche – Swan-Riv. 2 Thynnus variabilis, var. Sidney 3 Thynnus villosus Klug (box 73) Thynnus villosus Kl. Nouv. Holl.de 4, one labelled “” Thynnus xanthognathus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 74) Thynnus xanthognathus m., Thynnus picipes, Wesm. D. Deyrolles – N. Holl.de Sidney 1 FAMILY POMPILIDAE Agenia ornatipennis Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 78) Agenia ornatipennis m., n.sp. D. Dupont – Nov. Holl. 2 Agenia ornatipennis m. D. Dupont – Nov. Holl. 1 Agenia croceocera (!) Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 78) Agenia crocescens m., n.sp.? (in Casolari & Casolari Moreno, 1979 as A. croceocera) D. Deyrolles – N. Holl. 2
M. Generani and P. L. Scaramozzino 240
Agenia crocescens m. (in Casolari & Casolari Moreno, 1979 as A. croceocera) D. Deyrolles – N. Holl. 1 Aporus marginatus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 81) Aporus marginatus m. D. Veranì – Nouv. Holl. 1 Aporus marginatus m. D. Dupont – Terre de Van Diemen 2 Pompilus australasiae (box 80) Pompilus australasiae Voyage de l’Astrolabe – Australasie 1
Hymenoptera: Evolution, Biodiversity and Biological Control
2 labelled “Vespa tropica trisignata Pér. det. J.V.D. Vecht 1957” Polystes facialis De Saussure (box 104) Polystes facialis De Sauss. D. Deyrolles – Nouv. Holl. 1 Polystes flaveola (box 104) Polystes flaveola De Hann – Nov. Adelaida 1, 1 Polistes ……leti De Saussure (box 105) Polistes ....leti De Sauss. Coll. Latr. D. Deyrolle – Australie ………... 1, 1
Pompilus morio Dahlbom (box 80) Pompilus morio Dahlbom et Fab. D. Jekel, 1851 – Nov. Holl. 1
Paragia shuckarti Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 107) Paragia Shuckardti m., n.sp. Nouv. Holl. 2
Pompilus sericeocinctus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 80) Pompilus sericeo-cinctus m. D. Veranì – Nouv. Holl. 3
Eumenes campaniformis Fabricius (box 109) Eumenes campaniformis Fab. Nouv. Holl. 1
Pompilus xanthocerus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 80) Pompilus xanthocerus m. D. Veranì – Nouv. Holl. 1 (two specimens as listed by Casolari & Casolari Moreno, l. c.) Priocnemis hollandiae Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 82) Priocnemis hollandiae m. Nova Hollandia 1 Priocnemis hollandiae m. Nova Hollandia 2, 2 Priocnemis ruficeps Lepeletier (box 84) Priocnemis ruficeps Lepeletier (Pallosoma Lep.) (Les onglettes trapeus .........) Coll. Serville – Nov. Holl. 1 FAMILY VESPIDAE Vespa australis (box 103) Vespa australis D. Klug (M. B.) – Nouv. Holl
Eumenes latreillei De Saussure (box 108) Eumenes Latreillei De Saussure D. Reiche – Swan-Riv. 1, 2 Odynerus albifrons Fabricius (box 110) Odynerus G. Alastor Lep. albifrons (Vespa) Fabr. Nouv. Holl. 1 labelled “Alastor ........ de l’A. tuberculatus Sp. mais periment distincto Chepirm.. brisn. male” Odynerus atripes (box 109) Odynerus (Alastor) ?atripes (last name crossed out) Lepel. Nouv. Holl. 1 labelled “Alastor lachesis Saus.” (det. Zavattari) Odynerus clotho Lepeletier (box 109) Odynerus (Alastor) Clotho Lep.?, Vespa albifrons Fab.? D. Dupont – Van Diemen 3 Odynerus (Alastor) Clotho var. Nov. Holl. 1
Australian Hymenoptera in the Spinola Collection: a List of Species 241
Odynerus lepidus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 110) Odynerus lepidus m., n.sp. D. Jekel – Nov. Adelaide 1, 2 damaged specimens Odynerus sp. (vernalis) (box 112) Odynerus D. Dupont – Van-Diemen 1, 1, one labelled “vernalis ?” Odynerus swanii De Saussure (box 112) Odynerus Swanii De Saussure D. Reiche – Swan-River 1, 1 Odynerus tamarinus De Saussure (box 109) Odynerus tamarinus (sic.) De Saussure D. Reiche – Swan-Riv. 1 Rhygchium ephippium Fabricius (box 109) Rhygchium vel Odynerus ephippium (Vespa) Fab. D. Klug – Nouv. Holl. 1 labelled “Abispa ephippium Fab.” (det. Zavattari) FAMILY FORMICIDAE Eciton anale Klug (box 54) Eciton (Myrmecia) anale Klug D. Klug – Nouv. Holl.de 1 Eciton duponti Spinola in Casolari & Casolari Moreno, 1979 nomen nudum (box 54) Eciton Duponti m. D. Dupont – Terre de Van Diemen 3 workers with label “com. a Emery n.6” Eciton forficatum Klug (box 54) Eciton (Myrmecia) forficatum Klug D. Klug – Nouv. Holl.de 2 workers, one with white round label “6581” and white square label “Myrmica gulosa Latr. Nova-holl.” Eciton nigridens (box 54) Eciton nigridens Nouv. Holl.de 3 workers and 1 (?), one labelled “var. picidens N. Holl.”, another with a printed white square label “529” Eciton pallidens (box 54) Eciton pallidens Nouv. Holl.de 2 workers, 2
Eciton posticum (box 54) Eciton posticum , ..... rufum, var. ? Nouv. Holl.de 3 workers Eciton rufum (box 54) Eciton rufum , Myrmecia asmiens, Klug in coll. Latr. Nouv. Holl.de 3 workers Formica ammon Fabricius (box 56) Formica Ammon Fab. Coll. Latr. – Nouv. Holl. 3 workers, one labelled “n. h.” Formica Ammon, v? Coll. Latreille – Nouv. Holl. 2 workers *Formica argentata Fabricius (box 56) Formica argentata Fab.; Formica 6-spinosa Latr. Coll. Latr. Typus – Nouv. Holl. 1 Formica australis Latreille (box 56) Formica australis Latr? Coll. Latr. – Nouv.Holl. 2 workers, 1, one labelled “Polyrhachis femorata F. Sm.” Formica carinata Fabricius (box 56) Formica carinata ? Fab. Coll. Latreille – Nouv. Holl. 1 worker Formica hastata (Latreille) (box 56) Formica hastata Latr. Coll. Latr. – Nouv.Holl. 3 workers, one specimens with two square labels, the white one “Formica hastata coll.Spin.N. Holl. N 1” , the other, orange “Polyrhachis hastata Latr.”; another specimen with two white square labels “Port Natal, M.Delagoigne” and “Polyrhachis caffrorum Forel” -Formica hastata ? Coll. Latr. – Nouv.Holl. 1 Formica herculeana (box 57) Formica herculeana, v.? D. Deyrolles – Nouv. Holl. 1 Formica intricans Kirby (box 56) Formica intricans K.by Coll. Latr. – Nouv. Holl.
M. Generani and P. L. Scaramozzino 242
Hymenoptera: Evolution, Biodiversity and Biological Control
1 worker, with a square white label “Camponotus intrepidus Kirby”
1,1, the labelled “I S Bowerbank New Holland” “80”
Formica metalliceps Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 57) Formica metalliceps m. n.sp.? Coll. Latr. – Nouv. Holl. 2 workers labelled “Iridomyrmex purpureus F. Sm.”
Hogardia rufescens Lepeletier (box 93) Hogardia rufescens Lep. Stizus Hogardi Latr. M. Dupont – Nouv. Holl.de 2 Hogardia rufescens Nouv. Holl.de 1
Formica sp. (=Polyrrhachis presta Mayr) (box 56) Formica Coll. Latr. – Nouv.Holl. 1 worker, on its side a square white label “Polyrrhachis presta Mayr, la patria pare sbagliata (= country of origin probably wrong)” Formica sp. (=Polyrrhachis laboriosa F. Smith) (box 56) Formica Coll. Latr. – Nouv. Holl. 1 worker , with a square white label “Polyrrhachis laboriosa F. Smith (specie africana) (= African species)” FAMILY SPHECIDAE Ammophila incana Klug (box 75) Ammophila incana Kl.? Coll. Latr. – Nouv. Hol.de 2 Bembex adelaidae Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 93) Bembex Adelaidae m. D. Jekel, 1851 – Nov. Adelaida 1,1 Bembix furcata Erichson (box 92) Bembix furcata Erich. D. Reiche – Swan-Riv. 3 females Bembix furcata Erich. D. Dupont – Isle de Van-Diemen 1,1 Crabro australasiae Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 96) Crabro australasiae m., G. Solenius Lep. D. Dupont – Isle de Van Diemen 2 Exeirus lateritius Shuckard (box 86) Exeirus lateritius Shuck. D. Reiche – Swan-Riv. 1 Exeirus lateritius Shuck. Coll. Serville – Nouv. Holl.
Larra australasiae Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 90) Larra vel Lyrops Australasiae m. D. Deyrolles – Nouv. Holl. / Sidney 2 Physoscelus australasiae Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 98) Physoscelus australasiae m. n.sp.? D. Deyrolles – Sidney 1 Physoscelus australasiae var. D. Deyrolles – Sidney 1 Pison spinolae Shuckard (box 91) Pison spinolae Shuck D. Deyrolles – Nouv. Holl 2 Pison spinolae Shuck D. Verani – Nouv. Holl.da 1, 1 Sphex distincta Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 76) Sphex distincta m. D. Deyrolles – Sidney 1 Sphex pubiventris Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 76) Sphex pubiventris m. D. Deyrolles – Sidney 2 Sphex rufipennis Fabricius (box 76) Sphex rufipennis ? D. Deyrolles – N. Holl. 1 FAMILY ANDRENIDAE Colletes (Andrena) cicalybenla Erichson (box 136) Colletes (Andrena) cicalybenla Erichs D. Deyrolles – Sidney 3 specimens
Australian Hymenoptera in the Spinola Collection: a List of Species 243
FAMILY ANTHOPHORIDAE Crocisa emarginata Latreille (box 130) Crocisa emarginata Latr. Nouv. Holl. 2, one labelled: “Thyreus n. nitidulus (F.) Det. M.A.Lieftinck 1958”; the other labelled: “D. Watermann 1850. Fulodinang(?)”, “MT” and “Thyreus himalayensis Radosz. Det. M.A.Lieftinck” Crocisa guttata Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 130) Crocisa guttata m. n.sp.? D. Jekel – Nov. Holl. 2, one, white spotted, with a white label “Thyreus lugubris Sm.Det. M.A.Lieftinck 1958, cpd. with type”, a red label “MT” and another white label “5”; the other specimen with blue spots labelled “Thyreus caeruleopunctatus (Bl.) det M.A.Lieftinck 1958”. On the label on the bottom of drawer is another label placed on “Crocisa coeruleopunctata Blan” FAMILY APIDAE Allodapes cinea Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 132) Allodapes cinea m. n.sp. D. Jekel 1851- Nouv. Holl. 1 Allodapes rufiventris Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 132) Allodapes rufiventris m. D. Deyrolle – Sidney 1 Xylocopa aestuans Fabricius ?(box 128) Xylocopa aestuans Coll. Serville – Nouv. Holl. 1 Xylocopa muscaria Olivier (box 129) Xylocopa muscaria (Apis) Oliv., – (Centris) Fab., (Lestis) Lep. female Nouv. Holl. 2 Xylocopa muscaria male, Centris bombilans Fab. Nouv. Holl. 2 FAMILY COLLETIDAE Colletes rubricollis Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Colletes rubricollis m. D. Deyrolles – Sidney 2
Colletes rubricollis D. Deyrolles – Sidney 4 Colletes rufipes Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Colletes rufipes m. D. Deyrolles – Sidney 1 Colletes rufiventris Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Colletes rufiventris m. D. Dupont – Nouv. Holl. 2 Colletes unicolor Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Colletes unicolor m. inedit. Nouv. Holl. 1 Colletes vandiemenii Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Colletes vandiemenii m. D. Dupont – Isle de Van-Diemen 3, 2 Prosopis alcyone Erichson (box 132) Prosopis alcyone Erichs Nouv. Holl. 1, 2 FAMILY HALICTIDAE Halictus distinguendum Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Halictus distinguendus m. n.sp. D. Jekel 1851 – Nov. Holl. 2 Halictus nigritarsus (box 135) Halictus an nigritarsus Coll. Serville – Van Diemen 1 Halictus orbitus Spinola, in Casolari & Casolari Moreno, 1979 nomen nudum (box 136) Halictus orbitus m. D. Jekel 1851- Nov. Holl. 1 FAMILY MEGACHILIDAE Megachile chrysura (box 126) Megachile chrysura D. Deyrolles – Sidney 2, 2, one labelled “Nuova Olanda”
This Page Intentionally Left Blank
PART
5
Biology, Ecology and Behaviour
This Page Intentionally Left Blank
New Insights into the Foraging Behaviour of Parasitic Wasps Michael A. Keller and Brigitte Tenhumberg* Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, P.M.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) (*present address: Department of Zoology & Entomology, The University of Queensland, Brisbane, Qld 4064 Australia)
Introduction The foraging behaviour of parasitic wasps has received considerable attention from experimental and theoretical biologists for several reasons (Godfray 1994). There is a direct link between foraging behaviour and reproductive success, so the fitness associated with behaviour can be evaluated. Also, the link between behaviour and rates of parasitism allows direct study of one of the mechanisms driving population dynamics, especially the dynamics of pests under biological control. More recently, there have been concerns about the specificity of biological control agents. Understanding the factors that influence the susceptibility of non-target organisms to parasitism has further stimulated interest in foraging behaviour (Keller 1999). In order to link behaviour with reproductive success, one must understand the factors that influence the behavioural processes of a foraging wasp. Of particular importance are the decisionmaking processes that determine where a wasp will search and how long it will do so in each location (Charnov & Skinner 1985). Until recently, theoretical foraging models have driven research into these decision-making processes and the resulting levels of parasitism. The advent of new statistical methods, in particular Cox’s Proportional Hazards Model (Cox & Oakes 1984), has revolutionised the study of foraging behaviour in parasitic wasps. In this paper, one classical foraging model that is at the foundation of modern foraging theory is reviewed. Then the use of Cox’s Proportional Hazards Model and its application to the empirical study of decision-making by parasitic wasps is described. Finally, the foraging behaviour of the parasitic wasp Cotesia rubecula (Marshall) (Braconidae) at several spatial scales is described and analysed using the Proportional Hazards Model. This analysis shows how classical foraging models can guide empirical research as well as exposing their deficiencies when applied to a real organism.
Classical Foraging Theory Two questions are at the core of understanding how a foraging individual should respond to the patchy distribution of resources (Godfray 1994). Firstly, how long should an individual forager spend searching in a given patch upon encountering it? Secondly, how should a population of competing foragers be distributed among patches of varying quality? Classical foraging theory addresses these questions from an evolutionary perspective. It is assumed that foragers act in such a way that they maximise their individual fitness. Then, given various assumptions such as the spatial distribution of patches, energetic costs associated with travelling and searching, and the fitness gain from parasitising hosts, behavioural rules governing foraging have been derived. Charnov’s Marginal Value Theorem was one of the first models to address the question of how long a forager should search within a patch (Charnov 1976). In patches where resources are
247
Michael A. Keller & Brigitte Tenhumberg 248
Hymenoptera: Evolution, Biodiversity and Biological Control
40.0
Fitness gain
0.0 -40
0
Travel time
40
Patch time Optimum leaving time
Figure 1
Patch time as a function of travel time and fitness gain when a forager is searching in a patchy environment as predicted by the Marginal Value Theorem (Charnov 1976). A forager should leave the patch when the instantaneous rate of gain is maximised.
depleted over time, it is assumed that the rate of resource harvest decreases monotonically over time. The costs of foraging and travelling are uniform over time. In this environment, foragers are assumed to have perfect knowledge of the distribution of resources within patches and the associated costs and benefits of searching. Several predictions follow from these assumptions (Fig. 1). In order to maximise lifetime fitness, a forager within a patch should leave it when the rate of resource harvest, calculated from the time of departure from the last patch visited, is maximised. No forager should leave a patch before this maximum is reached, but leave each patch when the instantaneous harvest rate has dropped to the average maximum that can be achieved in the environment. Foragers should spend more time in patches with abundant resources, and avoid searching in patches where the expected rate of harvest is less than the average maximum. As travel time between patches increases, the time spent foraging within a patch should also increase. Such behaviour is only possible if the forager would be omniscient. Clearly real foragers are not omniscient. Lack of prior information about the distribution of patches and resources in space limits their ability to behave in the theoretically optimal way predicted by the Marginal Value Theorem and other models. This leads to two questions: 1) which information is used by real foragers to decide how long to search in a patch, and 2) how do foragers process this information? These questions can be answered by using Cox’s Proportional Hazards Model (Cox and Oakes 1984).
Application of Cox’s Proportional Hazards Model to Behavioural Data Two seminal papers on the application of Cox’s Proportional Hazards Model to ecological data in general and the foraging behaviour of parasitic wasps in particular have led to considerable
New Insights into the Foraging Behaviour of Parasitic Wasps 249
research on the factors that influence this behaviour (Haccou & Hemerik 1985; Haccou et al. 1991). The Proportional Hazards Model is an analytical tool from statistical survival analysis. This class of statistical methods considers the likelihood of a particular type of failure occurring over time. It has been applied in medicine to the study of cancer mortality and in engineering to the longevity of car parts. In foraging behaviour, the ‘failure’ of interest is the time when a forager leaves a patch containing hosts. In practice, many similar behavioural events could be analysed, e.g. the time until the first oviposition is observed within a patch. The analytical calculations used to fit the Proportional Hazards Model are complex, but the results are easily interpreted. The model assumes that there is a survival curve which describes the decreasing number of searching wasps that remain in a patch over time. This curve describes the relationship between elapsed time and the proportion of searching individuals remaining within a patch. Normally this proportion of individuals is plotted on a logarithmic scale. The log transformation produces a linear curve if the rate of departure per individual is constant over time, i.e. if the curve exhibits exponential decay. Non-linear plots indicate the tendency to leave changes over time and these can be easily interpreted by visual inspection. The Proportional Hazards Model is used to compare survival curves when initial conditions vary. An important assumption of the Proportional Hazards Model is that the differences between the survival curves can be described by a constant proportional factor. If this assumption is invalid, then the model cannot be fitted to the data. When this proportionality among survival curves does occur, they appear to be pivoted about the starting point. The Proportional Hazards analysis produces a series of coefficients that describe how each variable, or ‘covariate’, affects the tendency of a wasp to leave the patch. The Proportional Hazards Model is considered to be non-parametric because there is no statistical curve fitting. Rather, the points along the survival curves are compared directly. One important advantage of using survival analysis is that all data can be included in the analysis, even if some individuals have not left the patch before the end of the observation. These incomplete observations are said to be censored. Observations can be censored in other ways. Using classical foraging theory as a guide, various foraging events can be considered as censoring an observation. For example, when a wasp oviposits in a host, it may reassess its perception of how many hosts are present in the patch. In survival analysis, events like these can be considered as censoring times at which the covariates used in the analysis are updated. The time spent searching until the oviposition is included in the analysis because we know that the wasp would have searched in the patch at least as long had it not oviposited. However, as it did not leave the patch, it selfcensored this portion of the observation. Such self censoring enables researchers to analyse how the leaving tendency changes as a wasp searches within a patch. Foraging theory plays an important role in this type of analysis (Haccou et al. 1991). It is used to guide behavioural scientists in hypothesising censoring points when the wasp’s tendency to continue searching within a patch may change. Note that even though circumstances may change during foraging, the covariates used in the analysis are set only at the start of the foraging period when the wasp entered the patch or when the wasp self-censors the observation by ovipositing or performing some other act. This is an important point. The Proportional Hazards Model does not indicate the factors that finally trigger patch leaving. Rather, the factors or covariates that are found to influence the leaving tendency are those which exist at the start of each measured period of searching. Some of these may change during the course of a patch visit, but these changes in covariates are not included in the analysis. Unfortunately, it is probably impossible to determine factors that finally trigger
Michael A. Keller & Brigitte Tenhumberg 250
Hymenoptera: Evolution, Biodiversity and Biological Control
4m Figure 2
Arrangement of plants in an experimental evaluation of the patch leaving tendencies of Cotesia rubecula. Wasps were released into the central patch of one or four plants and observed until they reached one of the six peripheral plants.
patch leaving, hence an interplay between experimental and theoretical studies is important for furthering the understanding of foraging behaviour in parasitic wasps.
The Foraging Behaviour of Cotesia rubecula The braconid wasp C. rubecula parasitises the first four instars of the cabbage white butterfly, Pieris rapae L. Its behaviour has been studied at three different spatial scales: within patches of plants, on whole plants and on leaves. These studies have revealed how various factors influence searching decisions by C. rubecula at different spatial scales. They illustrate how the Proportional Hazards Model can be used to elucidate the mechanisms underlying decision-making by C. rubecula and other parasitoids. An experiment was conducted to investigate how C. rubecula searches for hosts within patches of plants. Wasps were released into patches of 1 or 4 plants separated by 75 cm (Fig. 2). Host density was 1 or 3 hosts/plant. Nine to 13 wasps were observed for each treatment combination. Six plants with 1 host/plant were situated at equal distances around a circle 4 m away from the centre
New Insights into the Foraging Behaviour of Parasitic Wasps 251
Table 1 Results of analyses of foraging data for Cotesia rubecula using Cox’s Proportional Hazards Model. Note that coefficients (β) for each covariate analysed indicate the effect on patch leaving tendency. Negative values indicate that the leaving tendency decreases as covariate values increase and vice versa. (a) Residence time on the first plant visited in patches of one or four cabbage plants. Covariate
β
Host density
–0.54
No. of plants in patch
0.34
No. encounters with damage or hosts
0.16
Host stung? (0 = no, 1 = yes)
ns
No. of stings
ns
Rate of stinging
ns
(b) Residence time in the whole patch of one or three cabbage plants. Covariate
β
Host density
–0.90
No. of plants in patch
–0.35
No. encounters with damage or hosts
ns
No. of plant visits
ns
No. of stings
ns
Rate of stinging
ns
(c) Residence time on individual plants when searching among cabbage plants bearing 0, 1, 2 or 5 hosts/plant. Covariate
β
Host density
–0.20
Rate of Stinging (no./min.)
0.07
First plant visited? (0 = no, 1 = yes)
-0.60
Host density
-0.90
(d) Residence time on Brussels sprouts leaves (data from Vos et al. 1998). Covariate
β
Encounter with host damage
–0.84
Oviposition
ns
Host encounter
ns
Elapsed time
ns
No. of stings
ns
Past encounters with damage or hosts
ns
Michael A. Keller & Brigitte Tenhumberg 252
Hymenoptera: Evolution, Biodiversity and Biological Control
100
% foraging on plant
1 plant, 1 host/plant
1 plant, 3 hosts/plant
4 plants, 1 host/plant 4 plants, 3 hosts/plant
10 0
25
50
Searching time (min) Figure 3
Survival curves for patch residence times of Cotesia rubecula searching for second instar Pieris rapae on the first plant visited in patches of one or four cabbage plants. At the start, all individuals are searching on the plant, but the numbers decline over time. The percentage that continue to search until the given time are plotted on a logarithmic scale. If the numbers decay exponentially, these lines would be straight.
of the experimental plot. When the wasps arrived at the outer circle or had disappeared for at least 10 min, they were considered to have left the experimental patch. Wasps were observed continuously by two observers and their behaviour was recorded using an event recorder. In this experiment, the residence times of the wasps was considered at two spatial scales: on the first plant visited and within the entire patch. If the first plant visited was considered to be the patch, the Marginal Value Theorem predicts that the time spent searching on the plant should be shorter when there is more than one plant present within the patch because travel time to reach another plant is shorter. Likewise, if one assumes that wasps will achieve a steeper gain curve and parasitise hosts more quickly when densities are higher, then the Marginal Value Theorem predicts that wasps should search longer on plants with higher host densities. When host density is higher, some hosts appear to be more easily located than others so this assumption seems reasonable. These predictions were compared to the results of a Proportional Hazards analysis (Table 1). The behaviour of the wasps is in general agreement with the theoretical predictions. The leaving tendency of the wasps increased when patch size was 4 plants compared to single plant patches. Thus, wasps left the first plant more quickly when other plants were nearby. Also, the leaving tendency of wasps was lower, i.e. they searched longer on the plants, when the host density was higher, which is in agreement with the prediction. However, note that some wasps left the first plant soon after arriving on it (Fig. 3). This is contrary to the prediction of the Marginal Value Theorem that they should remain on the plant until an overall optimum rate of oviposition is achieved. Another factor that affected the leaving tendency on the first plant was the numbers of encounters with hosts or plant damage. As these increased, so too did the leaving tendency. When numbers of such encounters continued to increase, it was an indication that the foraging wasp had been unsuccessful at attacking a host. However, no statistical influence on the leaving tendency was detected for the
New Insights into the Foraging Behaviour of Parasitic Wasps 253
100
% Remaining in patch
1 plant, 1 host/plant 1 plant, 3 hosts/plant 4 plants, 1 host/plant
4 plants, 3 hosts/plant
10 0
50
100
Time (min) Figure 4
Survival curves for patch residence times of Cotesia rubecula searching for second instar Pieris rapae for the whole patch of one or four cabbage plants.
act of stinging a host, the rate of stinging and the numbers of times hosts were stung. The absence of an influence of stinging has been interpreted as an indication that the tendency to leave the patch is re-set each time an oviposition occurs (Vos et al. 1999; see below). When foraging by wasps in this experiment was considered at a larger spatial scale of the whole patch, similar results emerged from the Proportional Hazards analysis (Fig. 4). As density increased from 1 to 3 larvae per plant or as the number of plants in the patch increased from 1 to 4, the leaving tendency decreased. Thus wasps spent longer times within the patches that were larger or had higher host densities. However, no statistical effect was detected for four other factors: the number of encounters with feeding damage or hosts, the number of plants visited before leaving the patch, the rate of stinging, or the number of hosts stung. The absence of an effect of encounters with feeding damage or hosts is contrary to what was found when behaviour on the first plant was analysed. This suggests that factors that influence wasp behaviour are different at different spatial scales. The absence of any effects of stinging hosts was consistent with results found on the first plant. Both the size of patches and the density of hosts affected rates of parasitism (Table 2). As host density increased the number of hosts stung increased, while more hosts were stung in the larger patches. However, the percentage of hosts parasitised was greatest on single plants that bore only one host, and this percentage decreased with increased density and increased patch size. Although these results suggest that mortality due to parasitism may be inversely density-dependent, this may not be so. If there is aggregation of wasps in patches with higher host density, or there are changes in host attractiveness over time as a consequence of parasitism, the relationship between density and parasitism at the level of the population could differ from the trends observed in this experiment.
Michael A. Keller & Brigitte Tenhumberg 254
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Effects of patch size and host density on parasitism of second-instar Pieris rapae by Cotesia rubecula. Host Density (No./Plant)
n
No. Stung (Mean + 95% CI)
% Parasitised
1
1
9
0.7 + 0.38
67
1
3
13
1.6 + 0.68
54
4
1
9
1.7 + 1.09
42
4
3
9
4.4 + 1.89
37
No. Plants
Superparasitism was commonly observed. Overall, each parasitised host was stung 1.53 times. Dissections of other larvae superparasitised by C. rubecula have indicated that an egg is laid at each sting (Sassan Asgari pers. comm.). The observations in this experiment were used to test one of the assumptions of the Marginal Value Theorem, specifically the assumption that fitness gains are a monotonically decreasing function of time spent foraging in a patch. Inspection of curves illustrating the numbers of hosts parasitised versus time spent foraging in patches indicated that this was commonly not the case (Fig. 5). Note that superparasitism occurred before all hosts within the patch were parasitised. Such departures from model assumptions suggest that an alternative to the Marginal Value Theorem must explain the foraging behaviour of C. rubecula. One such model might be a countdown mechanism as proposed by Driessen et al. (1995). However, no increase in leaving tendency was indicated as the number of ovipositions increased, so this model does not apply to C. rubecula. Wasps did respond to host density. The initial leaving tendency of C. rubeucla is determined by their perception of the concentration of synomones in the patch (Thomsen 1999). When wasps forage on plants from which hosts have been removed, the leaving tendency decreases with increasing levels of plant damage. Another important conclusion derived from this experiment was that searching for hosts is not systematic. Wasps may search a plant, leave it and then return again, in some instances many times. Some wasps left the focal patch and flew away from it briefly only to return and resume searching later, also many times in some instances. An omniscient forager would not behave in this way, nor would a forager be likely to do so if it remembered where it had searched. This suggests that wasps are constrained by their perception of cues released by host insects and host plants. Variation in host activity causing differential production of kairomones, the various positions of hosts on plants, changing wind speed and direction, combined with the chaotic turbulence inherent in air flows around plants, all combine to limit a wasp’s perception of hosts. A second experiment was conducted to determine how host density affected the time spent by C. rubecula searching among plants bearing different densities of hosts (Tenhumberg & Keller unpublished). Individual wasps that had previous oviposition experience were released into patches containing 16 plants, 4 plants each bearing densities of 0, 1, 2 and 5 hosts per plant arranged in a Latin square. In a proportional hazards analysis of this experiment, three factors were shown to affect the leaving tendency of the wasp: host density, the rate of oviposition (=1/ time since last oviposition) and the first plant visited. The leaving tendency decreased as host density increased. As the rate of oviposition increased, the leaving tendency increased. Thus, successful wasps moved more quickly from one plant to another. Wasps were slower to leave the first plant visited after release than when they searched on other plants. As a result, wasps laid more
New Insights into the Foraging Behaviour of Parasitic Wasps
No. Different Hosts Stung
255
9
6
3
0 0
20
40
60
80
100
Time in Patch (min) Figure 5
Gain curves for three Cotesia rubecula that searched in patches of four cabbage plants with three second-instar Pieris rapae on each plant. Symbols indicate the times when hosts were stung. Note that the curves do not increase monotonically and that in one instance the first superparasitism occurred after the first host was stung.
eggs (mean = 0.8/plant) on the first plant than on subsequent patches (0.3/plant). This suggests that the wasps were adjusting in some way to their environment following release. The nature of such an adjustment is unknown. This study highlighted an important methodological problem in defining patch boundaries. When the wasp was considered to have left the patch when she took flight, no effect of host density on the leaving tendency could be detected (P=0.28). However, if wasps were considered to have left the patch when they flew past a line marking the mid-point between plants, then the effect of host density on the leaving tendency was found (P<0.0001). Thus, the observer’s perception of the patch boundary can influence the outcome of the analysis. In this case the wasps intermittently flew around plants when searching, but the exact point at which their focus shifted from one plant to another was impossible to measure. The time at which wasps passed the gridlines provided the best estimate of the patch departure times. In a third study, the behaviour of C. rubecula foraging on single leaves of Brussels sprouts plants was analysed by Vos et al. (1998). No effects of oviposition, host encounter, numbers of hosts stung, past encounters with hosts or damage, and elapsed time searching were detected in a Proportional Hazards Analysis. Only encounters with host damage were shown to decrease the leaving tendency of wasps. The absence of an effect of oviposition on the leaving tendency suggested that wasps ‘re-set’ their leaving tendency when a host is parasitised. It should be noted that the Brussels sprouts leaves in this study were much larger than those of the pre-heading cabbage plants in the preceding two studies. This difference in size makes it difficult to compare the results among all three scales of reference. Nevertheless, different factors were shown to affect the leaving tendency of C. rubecula at each scale. At the smallest scale, that of the leaf, only host damage was shown to influence the leaving tendency. Cotesis rubecula is known to walk more slowly and turn more frequently when it encounters host damage (Nealis 1986). This searching behaviour increases the probability of finding a host because hosts often
Michael A. Keller & Brigitte Tenhumberg 256
Hymenoptera: Evolution, Biodiversity and Biological Control
rest away from feeding sites. At the level of the whole plant, the numbers of encounters with host damage was shown to have the opposite effect. This is not surprising because wasps moved from leaf to leaf, and greater numbers of encounters with host damage were associated with unsuccessful searching. At the largest scale of groups of plants, no effect of the number of encounters with host damage was detected. Host density was found to affect the leaving tendency when C. rubecula searched on plants and on groups of plants. It is possible that rather than assessing host density directly, the wasps are influenced by the amounts of synomones released from plants. The wasp Venturia canescens Gravenhorst is known to search longer in patches that have higher concentrations of semiochemicals which vary as a direct consequence of varying host density (Driessen et al. 1995). Cotesia rubecula tends to leave plants more quickly when the rate of oviposition is high. This may be an adaptive response to avoid superparasitism. This parasitoid does not seem to avoid superparasitising hosts directly. By leaving patches when oviposition rates are high, the probability of superparasitism should decrease. The model of adjustable termination rates proposed by Vos et al. (1998) best describes the behaviour of C. rubecula, but this model needs to be reformulated to account for different spatial scales. In particular, wasps exhibit subtle differences in behaviour when foraging is studied at the level of the leaf, the whole plant and group of plants. The factors that influence the leaving tendency changes according to scale of reference.
Conclusions The case studies of the foraging behaviour of C. rubecula indicate that the spatial scale of experimental focus affects how we interpret foraging decisions. Studies of foraging behaviour should be conducted at different spatial scales in order to fully understand the factors that influence decisions by foraging wasps. Cox’s Proportional Hazards Model was an important tool used in each study of C. rubecula. This statistical method permits analysis of the factors that influence the leaving tendency, and indirectly the time spent searching in patches, at different spatial scales. The Proportional Hazards Model has generated renewed interest in foraging behaviour because its output can give direct insight into the factors that affect the decision-making processes of searching wasps. The interplay between foraging theory and analysis of observations of foraging wasps should lead to a much better understanding of the mechanisms that govern foraging behaviour.
Acknowledgements We thank Jacques van Alphen for his comments on a draft of the manuscript. This research was supported by grants from the Australian Research Council.
References Charnov, E. L. (1976) Optimal foraging: the marginal value theorem. Theoretical Population Biology 9: 129-136. Cox, D. R & Oakes, D. (1984) Analysis of Survival Data. Chapman and Hall, London. Driessen, G., Bernstein, C, van Alphen, J. J. M. & Kacelinik, A.. (1995) A count-down mechanism for host search in the parasitoid Venturia canescens. Journal of Animal Ecology 64: 117-125.
New Insights into the Foraging Behaviour of Parasitic Wasps 257
Haccou, P. & Hemerik, L. (1985) The influence of larval dispersal in the cinnabar moth (Tyria jacobaeae) on predation by the red wood ant (Formica polyctena): An analysis based on the proportional hazards model. Journal of Animal Ecology 54: 755-769. Haccou, P., de Vlas, S. J., van Alphen, J. J. M. & Visser, M. E. (1991) Information processing by foragers: effects of intra-patch experience on the leaving tendency of Leptopilina heterotoma. Journal of Animal Ecology 60: 93-106. Keller, M. A. (1999) Understanding host selection behaviour: the key to more effective host specificity testing. pp. 105-116 In Withers, T. M., Barton Brown, L. & Stanley, J. (Eds), Host Specificity Testing in Australiasia: Towards Improved Assays for Biological Control. CRC for Tropical Pest Management, Brisbane. Nealis, V. G. (1986) Responses to host kairmones and foraging behaviour of the insect parasite Cotesia rubecula (Hymenoptera: Braconidae). Canadian Journal of Zoology 64: 2393-2404. Thomsen, D. (1999) The effects of host density, oviposition and experience on patch leaving decisions in the parasitoid Cotesia rubecula (Hymenoptera: Braconidae). Diploma Thesis, University of Hannover. Vos, M. Hemerik, L. & Vet, L. E. M. (1998) Patch exploitation by the parasitoids Cotesia rubecula and Cotesia glomerata in multipatch environments with different host distributions. Journal of Animal Ecology 67: 774-783.
The Biology of Perreyiine Sawflies (Hymenoptera: Pergidae) of the Perreyia Genus-group Carmen Flores1, Jesús Ugalde2, Paul Hanson1 and Ian Gauld3,1 1
Escuela de Biologia, Universidad de Costa Rica, San Perdo, Costa Rica
2
Instituto Nacional de Biodiversidad de Costa Rica, Santo Domingo de Heredia, Costa Rica
3
Department of Entomology, The Natural History Museum, Cromwell Rd, London SW7 5BD United Kingdom (email:
[email protected])
Introduction The Pergidae is one of the more species-rich families of Tenthredinoidea, the principal clade of folivorous Hymenoptera (Gauld & Bolton 1988; Hanson & Gauld 1995; Vilhelmsen 1996). It comprises some 500 species in 14 subfamilies and 57 genera (Smith 1993). Unlike the majority of the other tenthredinoid families, which are most species-rich in northern temperate regions, the Pergidae is a Gondwanan group. In the Old World, it is restricted to the Australian plate (Australia, New Guinea, Sulawesi and associated islands), where it is numerically the dominant sawfly group (Naumann 1991). In the New World isolated species have ranges extending north to south-eastern Canada (Smith, in Krombein et al. 1979), but the overwhelming majority of taxa (32 genera, and over 350 species) are restricted to South and Central America (Smith 1990). In the Tenthredinoidea, only the Argidae are similarly most species-rich in the Neotropics but, unlike the Pergidae, the Argidae is a diverse, cosmopolitan family. Like the overwhelming majority of other tenthredinoids the Pergidae were thought to be herbivorous as larvae (Naumann 1991; Smith 1993), and to feed on the foliage of a range of green plants. In Australia, where the habits of a number of pergids are well known, most species are external feeders on the leaves of woody plants, especially on species of Myrtaceae (notably Eucalyptus, Callistemon, Tristania, Melaleuca and Leptospermum), although a few also attack Rosaceae (Rubus), Elaeocarpaceae (Elaeocarpus) and herbaceous Polygonaceae (Rumex, Emex), and one species feeds on a fern (Marsilea: Marsileaceae). Although most are external folivores, the larvae of Phylacteophaginae are miners in leaves (Naumann 1991; MacDonald & Ohmart 1993). A few Australian species of Euryinae are unusual as they have been recorded as feeding on dead and dying leaves (Moore 1957). Comparatively less is known about the larval food-plants and habits of New World species, but in the Neotropics a species of Syzygonia Klug (Syzygoniinae) feeds on Tibouchina (Melastomataceae), Tequus Smith (Acordulecerinae) on cultivated Solanum (Solanaceae), Haplostegus Konow (Pergulinae) on Psidium (Myrtaceae), Heteroperreyia Schrottky (Perreyiinae) on Schinus (Anacardiaceae), Cerospastus Schulz on Nothofagus (Fagaceae), Incalia Cameron (Syzygoniinae) may feed on Vitis (Vitaceae), and a species of Enjijus Smith (Acordulecerinae) is sometimes a pest of guava (Psidium; Myrtaceae) (Pyenson 1940; Smith 1990; Hanson & Gauld 1995; Naumann & Groth 1998). A North American Acordulecera Enderlein (Acordulecerinae) feeds on the leaves of Fagaceae and Juglandaceae (Smith in Krombein et al. 1979). It has generally been assumed all New World pergids similarly feed on green plants (e.g. Smith & Middlekauf 1987).
258
Hymenoptera: Evolution, Biodiversity and Biological Control
The Biology of Perreyiine Sawflies (Hymenoptera: Pergidae) of the Perreyia Genus-group 259
Most of the 14 pergid subfamilies are restricted to either Australasia or to the New World, but three, the Pergulinae, Perreyiinae and Philomastiginae, occur in both areas. It is the second of these that is the subject of this paper. The Perreyiinae comprises seven genera, six in the New World (Camptoprium Spinola, Heteroperreyia, Barilochia Malaise, Perreyia Brullé, Perreyiella Conde and Decameria Lepeletier and Serville), and one in New Guinea and Sulawesi, Cladomacra Smith. These genera fall into two very distinctive groups (Smith 1990): Group I (Camptoprium, Heteroperreyia, Barilochia), which may be a paraphyletic assemblage with a strongly sclerotised, ‘typical’ ovipositor, and the monophyletic clade, Group II (Perreyia, Perreyiella, Decameria and Cladomacra) (hereinafter referred to as the Perreyia genus-group), with very reduced, needle-like ovipositors. One species in Group I, Heteroperreyia jorgenseni (Jörgensen), has been reared on Anacardiaceae, but virtually nothing is known about the biology of any species in the Perreyia genus-group. Data on specimens in museum collections seem to support the supposition that, like other tenthredinoids, the larvae are phytophagous. For example, specimens of Decameria similis (Enderlein) are labelled as ‘on Ficus carica (Moraceae)’ and ‘on mango (Anacardiaceae)’, and Perreyia tropica (Norton) has purportedly been reared ‘from grass roots’ (Smith 1990). However, as Smith remarks, it is seldom clear what exactly old specimen data means. For example, the adult specimen of D. similis may simply have been collected off Ficus, and P. tropica may well have been reared from a cocoon found in a grass root tussock. Neither species may have any real biological association with these plants. We suspect that the Perreyia genus-group have a rather different larval biology from other tenthredinoids, because information has gradually been accruing about them, that hints there may be very unusual features about the habits of these sawflies. First is the actual morphological adaptation defining the group – the semi-vestigial ovipositor. This is an extraordinary modification, for the ovipositors of most other tenthredinoids (and indeed other leaf-feeding sawflies such as Pamphiliidae) are robust, highly modified saw-like structures that are used for cutting a slit in plant tissue in which eggs are laid (Benson 1950; Carne 1962; Naumann & Groth 1998). The fact that these perreyiines have a virtually functionless ovipositor suggests they have a very different ovipositional biology from other tenthredinoids. Second, two species of Cladomacra have been discovered that have apterous females (Naumann 1984; Shinohara 1986), a very unusual condition amongst the sawflies (Naumann 1997). The wingless female of C. terricola Naumann was found guarding its eggs under a log (Naumann 1984), and although egg-guarding is not an uncommon trait in tenthredinoids, it is usual for the female to oviposit on or into plant tissue that will be consumed by the larva (Azevedo Marques 1933; Benson 1950; Dias 1975, 1976; Naumann 1991). Although aptery is unknown in Neotropical sawflies, in Central America male perreyiines are more commonly collected in flight interception traps than females; females are much larger than males, and are very clumsy fliers so they probably tend to be sedentary. There also remains the possibility that some New World species may have apterous females, in that for many species this sex is unknown. In a recent revision of the family (Smith 1990), the females of 22 of the 81 Neotropical perreyiines (27%) remain unknown, compared with just 6 of the 87 species of the related subfamily Acordulecerinae (7%). Third, was a recent observation in Costa Rica that the larvae of Decameria rufiventris Cameron are mycophagous (Smith, in Hanson & Gauld 1995). This was the first unequivocal record of fungus feeding in the entire Tenthredinoidea. About a dozen larvae of D. rufiventris were found at Monteverde (1600 m) feeding on a jelly fungus (Auricularia sp., Auriculariaceae) growing on rotten wood (Olsen & Joyce pers. comm.). Two were reared to adulthood on this diet.
Carmen Flores, Jesús Ugalde Paul Hanson and Ian Gauld 260
Hymenoptera: Evolution, Biodiversity and Biological Control
In this paper we present our observations on the larvae of another member of the Perreyia genusgroup, P. tropica. The larvae of this species are very common during the rainy season in coffee plantations and forests of Costa Rica. The migrating masses of black larvae are well-known among rural people and are given numerous common names in Spanish: ‘gusanos de octubre’, ‘gusanos de invierno’, ‘gusanos de San Francisco’, ‘gusanos de San Lorenzo’, ‘siete cueros’, ‘retoños’. Despite their conspicuousness, the identity of these larvae was previously unknown to biologists, since they had never been reared to the adult stage. We succeeded in rearing these caterpillars to maturity on a diet of leaf litter.
Materials and Methods Between October 1994 and August 1996, 14 groups of larvae were observed in the following locations in Costa Rica (all in the central valley, except those where a province is indicated): San Antonio de Belén (950 m altitude), Estación Biológica San Ramón (700 m and 1200 m, Alajuela Province), Santo Domingo de Heredia (1200 m), Sabanilla de Montes de Oca (1220 m), San Antonio de Escazú (1300 m), Estación Biológica Las Alturas (1500 m, Puntarenas Province) and Cerros de Escazú (1900 m). Eight of these groups were captured and placed in glass aquaria or large plastic containers filled with soil and humid leaf litter, for further observation. In the laboratory some larvae were marked with nail polish to monitor individual movement, aggregation and the division of groups. During experiments the soil was kept humid and changed periodically to avoid accumulation of faeces. Pupae were placed in clean Petri dishes with a small amount of sifted soil. Reared specimens were identified by D. Smith as P. tropica and deposited in INBio, Costa Rica.
Results Altitudinal and seasonal distribution Perreyia tropica has only been found between 700 and 1900 m in Costa Rica, even though intensive inventorying has been undertaken at all altitudes between sea-level and 3000 m (Hanson & Gauld 1995). Adults have been field collected during March and April, and in the laboratory adults emerged in April and May. Larvae have been found in the field from mid April until the end of October and, larvae collected in the field and brought into the laboratory, passed the months of October through March as cocooned pupae. This suggests that there is a single generation per year, with larvae active during the rainy season (May – November), while most of the dry season is passed in cocoons in the soil. Adults, fecundity and oviposition Adults are sexually dimorphic, with the males being considerably smaller (fore wing length 7.1–8.3 mm) than the females (fore wing length 9.6–10.4 mm). Almost all specimens collected in flight interception traps were males, suggesting males are more active fliers than females. Fecundity is unusually high. One of the females reared in the laboratory was dissected, and contained approximately 300 eggs, whilst a female captured in the field contained 230 eggs. Similar-sized species of the Perginae generally have about 50–180 eggs (Carne 1962; MacDonald & Ohmart 1993). The eggs are roughly oval in shape, bright yellow, smooth and approximately 0.82 × 0.56 mm. Their high fecundity, coupled with a relatively large egg size presumably, accounts for the fact that females are rather massive, and clumsy, poor fliers.
The Biology of Perreyiine Sawflies (Hymenoptera: Pergidae) of the Perreyia Genus-group 261
Figure 1
Larval mass of Perreyia tropica.
Although we have not observed oviposition in the field, indirect evidence suggests that eggs are deposited in masses on or in the soil. On two occasions in the laboratory, females laid eggs in a single mass at the bottom of a tube, whilst a field-collected female was observed to have mud on her ovipositor.
Larval behaviour The gregarious larvae are blackish in colour and, at least in later instars, they move around on the ground in highly visible groups (Fig. 1). Although the larvae have a slightly warty integument (Fig. 2), they are not defended by obvious spines as occur in, for example, the blennocampine tenthredinids (Smith & Middlekauf 1987). Group size varies from about 10 to over 250 larvae. Occasionally one may find an isolated larva, and one of us (C. F.) once observed an aggregation of several thousand individuals, although these were locally dispersed on the ground and not in a single mobile group. The mobile larval masses are somewhat cylindrical in form, with four to five layers of caterpillars, one on top of another. The mass proceeds by the top layer of larvae moving forward at the greatest velocity, the middle layers moving more slowly and the bottom layer not moving at all. As the posterior end of the bottom layer is exposed, the larvae comprising it begin to crawl up on top of the mass from behind. The velocity of the larval mass during the day was measured on four occasions (once in Santo Domingo, once in Escazú and twice in Las Alturas), and varied from 0.72 to 1.3 metres per hour. Limited observations suggest that larval masses may move more at night than during the day. Upon disturbance the larval mass becomes quiescent. This response occurs instantly when the front end of the mass is disturbed, but less rapidly when the rear end is stimulated. Movement is only resumed after an average of four minutes (n = 6). Field observations suggest that the larval mass is not adept at climbing over
Carmen Flores, Jesús Ugalde Paul Hanson and Ian Gauld 262
Figure 2
Hymenoptera: Evolution, Biodiversity and Biological Control
Final instar larva, Perreyia tropica, lateral view of head and thorax, abdominal segment 3, and caudal segments.
The Biology of Perreyiine Sawflies (Hymenoptera: Pergidae) of the Perreyia Genus-group 263
obstacles. One mass (Cerros de Escazú, 7 July 1996) was observed spreading out in front of a fallen tree trunk; some larvae were successful in climbing up the rough, moss-covered bark, but those reaching around a metre above the ground eventually fell off. In another case (San Antonio de Escazú, 17 July 1996), when rocks were placed in front of the larval mass, the larvae went around the obstacles. When a tree trunk was placed in front of this same larval mass, it succeeded in passing underneath, through a very small opening, in the course of which the height of the mass (i.e. number of individuals on top of one another) was reduced. The larvae are apparently distasteful, and it is well-known to local people that domestic chickens will not eat them. We observed no incidences of attack by predators or parasitoids, and the only larval parasitoid we reared from P. tropica was the tachinid Vibrissina sp. Larvae apparently do not feed whilst moving as a group. Two field observations and numerous observations in the laboratory suggest that the larvae disperse slightly in order to feed. Larvae feed primarily below and within the leaf litter, and are rarely exposed on the surface. They feed concurrently on dead leaves, and presumably fungal hyphae. One group of larvae (Santo Domingo, September 1996) maintained in captivity was given different types of dead leaves: coffee (Coffea arabica; Rubiaceae), cas (Psidium friedrichsthalianum; Myrtaceae), mango (Mangifer indica; Anacardiaceae) and poró (Erythrina poeppigiana; Fabaceae). All were consumed, but they appeared to devour less of the tougher leaves (cas and mango) than the softer leaves. Almost the entire dead leaf is usually consumed. As most dead leaves in the litter are colonised by fungi, it is possible that fungi comprise an important part of the diet. Fungal enzymes may aid in digesting cellulose. Future studies using fungus-free dead leaves are required to determine the significance of fungi in the diet. Most larval masses contain similar sized individuals, probably because they comprise siblings originating from the same egg mass. However, we also observed occasional masses containing very different sized larvae. For example, on August 1996 in Santo Domingo, a mass of approximately 250 individuals was observed, containing larvae varying from 1.2 to 3.3 cm long. These heterogeneous masses probably result from larvae becoming separated, then reuniting with a different group. To test the possibility that membership in a particular larval mass may change we marked 65 individuals captured in San Antonio de Belén and placed them in an aquarium containing a group from Santo Domingo. Later the same day the two groups were united into a single mass. Larvae do not moult synchronously and the moulting larvae (yellowish in colour) continue to move with the mass. When maximum size is achieved (about 3 cm long) the larvae disperse slightly and construct cocoons in the soil. At least some of their time in the cocoon seems to be passed as a prepupa, like most other tenthredinoid sawflies. Cocoons of the females are generally larger than those of males, though too few adults were reared to allow statistical confirmation of this to be undertaken.
Discussion In the New World the phenomenon of mobile symphytan larval masses is not confined to Costa Rica. Masses of ‘migrating’ larvae (possibly Perreyia species) have been observed in Mexico, Bolivia, Brazil and Guyana (Smith 1990), so it is probable that the observations reported here apply to other species in the genus. Although it is unequivocal that they consume dead leaves, more research is needed to determine the presence and possible significance of fungi in the diet
Carmen Flores, Jesús Ugalde Paul Hanson and Ian Gauld 264
Hymenoptera: Evolution, Biodiversity and Biological Control
of P. tropica. However, mycophagy is now established as a characteristic of at least some of the Perreyia genus-group (Decameria), and from the evidence available concerning the unusual traits of others, we suggest that mycophagy and/or dead-leaf-feeding may be a widespread characteristic of the clade. Although we have not found eggs in the field, Naumann (1984) reported that one species of the related genus, Cladomacra, guards eggs concealed under a log. We suspect P. tropica has a similar ovipositional biology. The use of oviposition sites on the ground, in leaf litter, etc explains why species of this group lack a fully-functional, sawing ovipositor. These observations strongly suggest that earlier records of a range of possible green plants as food for the larvae of the Perreyia genus-group are erroneous. Egg-guarding occurs in Cladomacra (Naumann 1991) and may possibly occur in all member of the Perreyia genus-group. This, in conjunction with an epigeal or hypogeal oviposition site, would explain the tendency towards female aptery, a condition that elsewhere in the Hymenoptera is often associated with an hypogeal existence (Hanson & Gauld 1995). The more gracile males actively fly in search of the females. The adaptive significance of gregarious larval behaviour in P. tropica is presently speculative, although there is an extensive literature on this subject in a wide range of insects (see Vulinec 1990, and references therein). The larvae of P. tropica are almost certainly either distasteful, and/ or quite toxic, as they are avoided by chickens in Costa Rica, and there are records of cattle being killed by consuming perreyiines in Brazil (Camargo 1956). Possibly masses are more memorable, and thus more likely to be avoided by a predator than single larvae. However, experimental manipulations are needed to determine whether larvae in masses do have a greater chance of surviving attack than do isolated individuals. Gregariousness may also serve as a defence against parasitoids (Weinstein 1989; Codella & Raffa 1993). Despite their apparancy, and the sympatric occurrence of many species of Ichneumonidae that attack sawfly larvae (Gauld 1997), we have not reared a single hymenopteran parasitoid from P. tropica. To date the only confirmed incidences of mycophagy and detritivory in Tenthredinoidea all occur within the Perreyiinae (Smith, in Hanson & Gauld 1995), or in Australia, in the related pergid subfamily Euryinae (Naumann 1991). Such dietary preferences are elsewhere uncommon in larval Hymenoptera (Hammond & Lawrence 1989; Hanson, in Hanson & Gauld 1995), although mycophagy is the normal mode of nutrition in siricoids and leaf-cutter ants. However, adults of both these groups, in their different ways ensure that fungal growth is available for their young. Siricoid females inoculate the ovipositional substrate with fungal spores, whilst leafcutter ants cultivate ‘fungal gardens’. Only in the Pergidae mentioned here do larvae have to seek out and find their own fungal food source. In conclusion, perhaps the most salutary lesson that emerges from this study is the danger of extrapolating from observations made on north temperate organisms, and assuming their tropical relatives have similar biologies. There is now conclusive evidence for the occurrence of mycophagy and detritivory in tropical Tenthredinoidea – dietary habits that do not occur in their north temperate relatives.
Acknowledgements We thank David Smith for confirming the identity of these insects and for information about pergids generally, and Manuel Zumbado for identifying the tachinid. We also thank David Olsen,
The Biology of Perreyiine Sawflies (Hymenoptera: Pergidae) of the Perreyia Genus-group 265
Frank Joyce, Gerardo Rojas, Edgar Quiros, Gerardo Carballo and Ana Leon for observations in the field, and Dick Vane-Wright for comments on the manuscript.
References Azevedo Marques, L. A. de (1933) Tenthredinidae con hecida por “mosca de serra” cuja larva, ou “falsa lagarta” é nociva a varias espécies de Tibouchina. Ministerio da Agricultura, Instituto de Defesa Agricola, Rio de Janiero 12: 1-11 Benson, R. B. (1950) An introduction to the natural history of British sawflies. Transactions of the Society for British Entomology 10: 45-142. Camargo, O. R. (1956) As larvas “mata-porco” no Rio Grande do Sul. Boletim da Directoria da Produccão Animal 13: 23-29. Carne, P. B. (1962) The characteristics and behaviour of the saw-fly Pergia affinis affinis (Hymenoptera). Australian Journal of Zoology 10: 1-34. Codella, S. G. & Raffa, K. F. (1993) Defense strategies of folivorous sawflies. pp. 261-294. In Wagner, M. R. & Raffa, K. F. (Eds), Sawfly Life History Adaptations to Woody Plants. Academic Press, San Diego. Dias, B. F. de Souza (1975) Comportamento pré-social de Sínfitas do Brasil Central. I. Themos olfersii (Klug) (Hymenoptera, Argidae). Studia Entomologica (N.S.) 18: 401-432. Dias, B. F. de Souza (1976) Comportamento pré-social de Sínfitas do Brasil Central. II. Dielocerus diasi Smith, 1975 (Hymenoptera, Argidae). Studia Entomologica (N.S.) 19: 461-501. Gauld, I. D. (1997) The Ichneumonidae of Costa Rica, 2. Memoirs of the American Entomological Institute 57: 1-487. Gauld, I. D. & Bolton, B. (Eds) (1988) The Hymenoptera. Oxford University Press and British Museum (Natural History), Oxford. Hammond, P. M. & Lawrence, J. F. (1989) Mycophagy in insects: a summary. pp 275-324. In Wilding, N., Collins, N. M., Hammond, P. M. & Webber, J. F. (Eds), Insect-Fungus Interactions. Academic Press, London. Hanson, P. E. & Gauld, I. D. (Eds) (1995) The Hymenoptera of Costa Rica. Oxford University Press and British Museum (Natural History), Oxford. Krombein, K. V., Hurd, P. D. Jr., Smith, D. R. & Burks, B. D. (Eds) (1979) Catalog of Hymenoptera in America North of Mexico, 3 Vols. Smithsonian Institution Press, Washington, D. C. MacDonald, J. & Ohmart, C. P. (1993) Life history strategies of Australian pergid sawflies and their interactions with host plants. pp. 485-502. In Wagner, M. R. & Raffa, K. F. (Eds), Sawfly Life History Adaptations to Woody Plants. Academic Press, San Diego. Moore, K. M. (1957) Notes on the biology of the sawfly Polyclonus atratus Kirby 1882 (family Pergidae, subfamily Euryinae) and some of its parasites. Proceedings of the Royal Zoological Society of New South Wales 1955-56: 74-81. Naumann, I. D. (1984) An apterous female sawfly (Hymenoptera: Symphyta) from Papua New Guinea. Systematic Entomology 9: 339-349. Naumann, I. D. (1991) Hymenoptera (Wasps, Bees, Ants, Sawflies). pp 916-1000. In The Insects of Australia, Volume II. Melbourne University Press, Melbourne. Naumann, I. D. (1997) A remarkable new Australian sawfly with brachypterous, nocturnal or crepuscular females (Hymenoptera: Symphyta: Pergidae). Journal of Natural History 31: 1335-1346. Naumann, I. D. & Groth, H. (1998) A revision of Philomastigine sawflies of the world (Hymernoptera: Pergidae). Journal of Hymenoptera Research 7: 127-148.
Carmen Flores, Jesús Ugalde Paul Hanson and Ian Gauld 266
Hymenoptera: Evolution, Biodiversity and Biological Control
Pyenson, L. (1940) Notes on the biology of three tenthredinid (Hym.) pests of the guava. Bulletin of Entomological Research 30: 467-469. Shinohara, A. (1986) A new apterous sawfly from Sulawesi, Indonesia (Hymenoptera: Pergidae: Perreyiinae), and the pleural origin of the ventral region of the sawfly mesothorax. Systematic Entomology 11: 247-253. Smith, D. R. (1990) A synopsis of the sawflies (Hymenoptera, Symphyta) of America south of the United States: Pergidae. Revista Brasileira de Entomologia 34: 7-200. Smith, D. R. (1993) Systematics, life history and distribution of sawflies. pp. 3-32. In Wagner, M. R. & Raffa, K. F. (Eds), Sawfly Life History Adaptations to Woody Plants. Academic Press, San Diego. Smith, D. R. & Middlekauf, W. W. (1987) Suborder Symphyta. pp. 618-649. In Stehr, F. W. (Ed.), Immature Insects. Kendall Hunt, Dubuque. Vilhelmsen, L. (1996) The preoral cavity if lower Hymenoptera (Insecta): comparative morphology and phylogenetic significance. Zoologica Scripta 25: 143-170. Vulinec, K. (1990) Collective security: aggregation by insects as a defence. pp. 251-288. In Evans, D. L. & Schmidt, J. O. (Eds) Insect Defenses. State University of New York Press, New York. Weinstein, P. (1989) Cycloalexy in an Australia pergid sawfly (Hymenoptera: Pergidae). Bulletin et Annales de la Société Royale Belgique de Entomologique 125: 53-60.
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California: Methods of Introduction and Evidence of Host Shifting E. E. Grissell1 and K. R. Hobbs2 1
USDA, ARS, PSI, Systematic Entomology Laboratory, National Museum of Natural History, Washington, D.C. 20560-0168 USA 2
Entodendrex, P.O. Box 3895, Palm Desert, CA 92261-3895 USA
Introduction In 1961 the internal, obligate seed-feeding wasp Megastigmus transvaalensis (Hussey) was reported in California where it was reared from seeds of Schinus molle L. (Anacardiaceae) (Harper & Lockwood 1961). This discovery was odd in two respects: first, M. transvaalensis was known only from South Africa (Hussey 1956) where it was reared from S. molle, and second, S. molle was an endemic South American tree. This conjunction of facts immediately raised the question: how did a putative South African wasp end up infesting a South American tree in California? At first glance the answer to this question seemed obvious. Schinus molle, along with its seedwasp, was introduced into South Africa where the wasp was discovered and named by Hussey (1956). The wasp also was introduced into California where it was discovered in 1961. Obviously, M. transvaalensis was of South American origin and had been moved around by some unknown means to two widely different areas of the world. Unfortunately, as with many simple answers, this one did not fit any of the established facts known about M. transvaalensis. This paper attempts to answer the above question, or at least parts of the question, based on data that the senior author gathered during a three month research trip to South Africa in 1998 and on information uncovered by the junior author in California in 1999. The answer, if correct, now points to an even more critical question that needs to be answered; a question that will be posed at the conclusion of this paper. Megastigmus Dalman contains 126 described species, one third of which are obligate seedfeeders. The species are predominantly Old World (85%), mainly Australian and Palaearctic, with a small representation in the Nearctic (15%) (Grissell 1999). Until 1998 no species was known from South America (Grissell and Heydon 1999), but late in that year we saw the first examples of M. transvaalensis from Brazil (Perioto pers. comm.) and Argentina (Wheeler pers. comm.). The senior author has examined these specimens and confirmed their identity. Three species of Megastigmus have been named and identified based more upon their host preference than upon their morphological distinctness. If, in fact, these are reproductively isolated taxa, they would be called sibling species. These taxa are M. transvaalensis reared from Schinus spp. originally in South Africa (though the host is South American in origin), M. rhusi (Hussey) reared from Rhus spp. in South Africa, and M. pistaciae Walker reared from Pistacia spp. (and
267
E. E. Grissell and K. R. Hobbs 268
Hymenoptera: Evolution, Biodiversity and Biological Control
cultivars) originally in France and Italy (now also found in California). The hypothesised hostspecificity in these taxa is more readily accepted because, essentially, all other seed-feeding species of Megastigmus are known to be host-specific at the genus level of the plants they attack (Grissell 1999). The consequences of recognising host-based taxa, especially those that cannot otherwise be identified, are that associated data such as geographic distribution, reports of exotic introduction, true host preferences, and phylogenetic relationships, ironically become blurred due to the precise ability to define the host without regard for biological reality. In the case of the Megastigmus species treated here, it appears as if host-based identifications might be responsible for obscuring a potentially radical case of host-shifting and subsequent geographic movement. The findings are of practical importance because they impinge directly upon the potential biological control of both ornamental and invasive noxious weeds (Schinus spp.), as well as an important food crop (Pistacia spp. and cultivars). Available information concerning both plant and insect taxa is complex and diffuse. Below we outline this information first by insect seed-feeder and then by plant host. For purposes of discussion, recognition of the seed-feeder is based on its host association.
Megastigmus Seed-feeders: History, Distribution and Hosts Megastigmus transvaalensis This wasp was described from Pretoria, South Africa in 1956, where it was reared from seeds of S. molle (Hussey 1956). In 1998 the senior author collected and studied thousands of specimens from South Africa all reared and/or associated with S. molle and S. terebinthifolius Raddi. The first New World record for the wasp was published in 1961 when it was reported from San Diego County, California attacking S. molle (Harper & Lockwood 1961). In 1989 M. transvaalensis was reported from Florida and Hawaii attacking seeds of S. terebinthifolius (Habeck et al. 1989). In 1998 the senior author examined specimens of M. transvaalensis discovered in seeds of S. terebinthifolius in Brazil (Perioto unpublished; specimens collected in 1998) and in seeds of S. molle from Argentina (Wheeler unpublished; specimens collected in 1996). Megastigmus rhusi (Hussey) This wasp was first described from Bloemfontein, South Africa in 1956, where it was reared from seeds of Rhus lancea L. (Hussey 1956). The senior author, during three months of collecting in South Africa, reared M. rhusi (as defined by host association) from two additional species of Rhus (R. laevigata L. and R. angustifolia L.) and collected it on R. chirindensis Baker, R. rehmanniana glabrata (Sonder) Moffett, and R. pendulina Jacquin. The species was collected throughout Western Cape Province in every habitat sampled including seaside nature reserves, sand dunes, inland agricultural areas (i.e. highly disturbed areas), rocky lowland fields, high deserts, and mountain forests. In short, essentially everywhere Rhus grew, no matter how isolated, M. rhusi was found. Additionally, the senior author saw a specimen reared from seeds of Rhus viminalis of authors (now = R. lancea or R. pendulina) collected in Stellenbosch in 1895 (Stellenbosch University Museum). Thus, the species has been known from South Africa for over 100 years. In 1999 the junior author discovered M. rhusi (as defined by host plant) in seeds of introduced R. lancea in Portola County, California. This was the first collection of M. rhusi in the New World. Megastigmus pistaciae Walker The pistachio seed chalcid was described from France and Italy in 1871, where it was collected on Pistacia lentiscus L. and P. terebinthus L. This species occurs naturally in regions of the world
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 269
where Pistacia species are endemic, namely the Mediterranean, north Africa, and the Middle East. The wasp was first discovered in the New World in northern California in September 1967 attacking Pistacia vera L. (Robinson 1968). Later the wasp was reported attacking an ornamental pistachio (P. chinensis Bunge) as well as P. atlantica Desfontaines (Vettel & Harper 1969), also in California. Between 1969 and 1983 the pistachio seed chalcid was collected in Orange and San Diego counties, far to the south of the initial report in Chico (Rice & Michailides 1988). Within 14 years M. pistaciae had spread nearly the entire length of California. Rice and Michailides (1988) hypothesised that the pistachio seed chalcid was introduced into the USDA Tree Improvement Center (at Chico) via infested and improperly fumigated Pistacia seed imported from Iran in 1965. The same authors also hypothesised that the apparently rapid movement of the pistachio seed chalcid from the northern to southern areas of California was hastened by the planting of ornamental P. chinensis trees along major interstate and federal highway systems in the state. Rice and Jones (1996) reported finding the pistachio seed chalcid in a single, isolated P. atlantica seed tree, used for nursery stock nuts, located in Kern County, California. With a virtual highway system of host trees distributed throughout the state of California, it could be hypothesised that M. pistaciae is adept at finding its host tree no matter where it grows. However, as we point out at the conclusion of this paper, it may be that this species is not as host specific as is currently imagined.
Plant Hosts: History and Distribution The two main genera of trees and shrubs involved in this study are Rhus and Schinus. A third genus, Pistacia, also is involved, but only in an ancillary way at present. All taxa are in the family Anacardiaceae. Each of these genera produces (or historically has produced) some product or service that has made at least some of its taxa of value to humans and thus has created a reason to move the host to non-endemic areas of the world. In fact, the seeds of each of these genera have been moved about so much that their precise movements are essentially impossible to document. Below we summarise information about each plant based on historical published records.
Schinus spp. Two species of Schinus are known hosts for Megastigmus: S. terebinthifolius and S. molle. Both are endemic to South America and have been introduced into other portions of the New and Old Worlds. The history of the spread of this genus within foreign areas is virtually undocumented, but there is evidence (demonstrated below) that trees or seeds were being introduced into the Old and New Worlds at least from the mid to late 1800’s. During these times, unspecified species were cultivated in North Africa, the Cape Verde Islands, the Mascarenes (Thonner 1915), and the United States (Barkley 1944). The reasons these trees were cultivated and moved about are many. In South Africa, Schinus trees yielded ‘...timber, resin used industrially and medicinally, tanning and dyeing materials, vinegar, syrup, and medicaments’ (Thonner 1915). Schinus terebinthifolius is considered an excellent hedge plant and S. molle is an ornamental tree good for dry areas (Sim 1927). The senior author has seen numerous specimens of this tree growing in subdivisions in Somerset West, Western Cape Province, and it is planted frequently at roadside rests in the hot, dry parts of Western Cape Province. Despite the fact that fruits of Schinus are considered somewhat toxic, its ‘pink peppercorns’ are produced commercially in Réunion and shipped throughout the world (Habeck et al. 1989; pers. obser.). Fruits of S. terebinthifolius (and perhaps S. molle) may be purchased in bulk as whole ‘peppercorns’ from spice companies in South Africa and the United States. These fruits also are
E. E. Grissell and K. R. Hobbs 270
Hymenoptera: Evolution, Biodiversity and Biological Control
mixed with white, green, and black peppercorns (i.e. ‘true’ pepper, Piper nigrum L.) to produce commercially available mixtures sold under such names as ‘Rainbow Peppercorns’ (The Cape Herb and Spice Corp., Cape Town), or ‘Pepper Melange’ (McCormick Spice Co., Baltimore). In addition to their culinary use, the mature red berries of either species are used in wreaths during holiday seasons. We have seen several types of wreaths advertised in different mail-order garden catalogs distributed throughout North America. These items apparently can be sent anywhere in the world and might serve as one effective method of national and international distribution of living wasps via seeds. Also, the berries are available in craft stores (Wheeler pers. comm.).
Schinus terebinthifolius (Brazilian pepper, Florida holly, Christmas berry) This species is indigenous to Argentina, Paraguay, and Brazil (Hall & Vandiver 1997). Published records show that S. terebinthifolius has been used and moved about the New World for at least 150 years. The species is said to have been introduced into Florida in the early 1840’s (Barkley 1944), and then reintroduced into Florida in 1898 by a USDA plant explorer (Morton 1978). In the northern hemisphere of the New World, S. terebinthifolius is now reported from California, Texas, Florida, all the main Hawaiian Islands (HEAR 1997), Puerto Rico, and the Virgin Islands (USDA, NRCS 1997). The tree is having negative environmental impacts in Florida and Hawaii where it is considered a noxious weed. In south Florida alone, the tree has invaded some 800 000 acres of land, and one year of control cost $60 million (Devine 1998). This tree is known to displace native plants due to its aggressive growth and allelopathic effects, and can produce toxic effects on humans and animals, especially birds that feed excessively on the berries (NPSIPM 1997). The seeds are distributed by birds in Hawaii (HEAR 1997) and Florida (McCann et al. 1996). Trees of this species can produce seeds in as few as three years from germination (Hall & Vandiver 1997). Schinus molle (Peruvian peppertree, peppertree, California peppertree) This species is indigenous to Argentina, Brazil, Chile, Ecuador, Paraguay, Peru, and Uruguay (GRIN 1998). Relatively little information has been published for S. molle relative to S. terebinthifolius. In California during the 1840’s, ‘Chilean pepper’ is reported to have been planted in Los Angeles and approximately 300 miles northward in San Jose (Wickson 1921). By current terminology, Chilean pepper would refer to S. polygamus (Cav.) Cabr., but this common name is probably not now applied in the same context as it was in the mid-eighteenth century. In 1876 S. molle, the only species then reported in California, was called ‘Chili Pepper’ by Gray (1876), and until 1965 this species was the only Schinus reported in California (Munz 1965). It is known from Chile, so it is likely that this species is what was being referred to as Chilean pepper in the 1840’s. Schinus molle is also grown as an ornamental tree in Texas, Puerto Rico, and Hawaii (USDA NRCS 1997). In California S. molle has not been reported as invasive, although it does ‘naturalise’ (Munz 1965). Rhus spp. In the Old World Rhus is the most speciose genus of Anacardiaceae, and it is abundantly represented in South Africa with nearly 75 species (47 species in Western Cape Province alone) (Archer 1993). Although Rhus is generally not thought of in terms of economic value, its species have proved useful to humans and have thus been introduced into non-indigenous areas. Thonner (1915) reported that some species yielded ‘timber, tanning and dyeing materials ..., condiments, medicaments, and edible fruits’. Sim (1927) mentioned R. coriaria L. as a good tanning agent for leather (e.g. gloves, book bindings) and stated that 14 million pounds were
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 271
imported into the United States in 1918. He also mentioned that wax was produced from the Palaearctic species R. vernicifera DC (now = Toxicodendron vernicifluum (Stokes) F. A. Barkley). In South Africa several species provide hedging plants for houses and trees for roadside rest stops (pers. obser.). Breitenbach (1984) listed four species of Rhus introduced into South Africa from abroad: R. glabra L., R. succedanea L., R. typhina L. and R. verniciflua Stokes. Thus, Rhus has been moved about the world in several directions. A few species are indigenous to the New World and are found throughout the Nearctic, but no Megastigmus have yet been reared from them (Grissell 1999).
Rhus lancea (African Sumac, Willow Rhus, Karee) This species is endemic to Africa and is reported from Botswana, Lesotho, Namibia, South Africa and Zimbabwe (GRIN 1998). Although the species was introduced into the United States and ‘... has been used for decades in Southwest landscaping’ (Ellis Farms 1999), its presence has scarcely been noted in the printed literature. This species is listed as a good fire resistant tree in California (CDFFP 1999) and is recommended for planting in dry areas (SDCLUE 1999). According to Andy Sanders (pers. comm.), a botanist at University of California, Riverside, the species has probably become naturalised in Riverside County, and in Arizona the species is considered an escapee from cultivation (Rondeau et al. 1999) and an invasive weed (Campbell 1999). The junior author discovered R. lancea growing in Palm Desert (Riverside County, May 1999). Pistacia spp. In the Old World species of this genus occur naturally in ‘North Africa’ and ‘northern East Africa’ (Thonner 1915). According to Breitenbach (1984), P. chinensis Bunge (Chinese pistachio) and P. vera L. (‘real’ pistachio) were introduced into South Africa. The genus is primarily known today for its valuable edible nuts based on hybrid cultivars but several species are also widely planted as ornamentals. Historically the genus yielded ‘... timber, tanning and dyeing materials, resins (mastic and turpentine) ... used industrially, in medicine, as fumigatories, masticatories, or condiments, and for preparing spirituous drinks, also edible oily fruits and seeds and various medicaments’ (Thonner 1915). Pistacia vera, from which the commercial cultivars have been developed, ‘is a small tree native to Syria’, which is ‘... used as nuts and flavouring in ice cream and candy’ (Hill 1937). In the New World, both ornamental and commercial Pistacia species have been introduced. Pistacia atlantica Desfontaines is reported from California and Utah, P. chinensis from California, Texas, Alabama and Georgia (USDA NRCS 1997), and P. integerrima Stewart from California (Rice and Jones 1996). Pistacia chinensis is an ornamental tree planted along major north-south highway systems in the state of California (Rice & Michailides 1998). Also in California, P. atlantica and P. integerrima J. Stewart provide seed for rootstock of the commercial pistachio cultivar P. vera ‘Kerman’, which has been planted extensively in California and much less so in Arizona (Rice & Michailides 1988).
Megastigmus transvaalensis: Hypotheses for Center of Origin, Evidence for Host Shifting and Methods of Distribution Habeck et al. (1989) suggested that M. transvaalensis was possibly transported to the New World via the commercial trade in ‘pink peppercorns.’ The world’s centre of distribution for this product is Réunion off the eastern coast of Africa. By itself, this hypothesis leaves many basic questions unanswered. When discussing the potential movement (or introduction) of species from one area to another (in this case continents), three basic factors must be considered: first
E. E. Grissell and K. R. Hobbs 272
Hymenoptera: Evolution, Biodiversity and Biological Control
the centre or point of origin; second the direction of the movement, and third the method of movement (i.e. how was the movement accomplished). To put this in context for the case of M. transvaalensis, what is the most likely point of origin for the species; what were the possible pathways of movement between the United States, South Africa and Brazil, and what was the method of movement? All questions are difficult, but the first is perhaps the easiest to answer. Below we present discussions and evidence for the centre of origin of this wasp and for host shifting and possible methods of distribution.
Center of origin Based upon present information, there appear to be three primary centres of origin for M. transvaalensis: 1) it is endemic to South America, occurring naturally in the seeds of species of Schinus; 2) it is endemic to the United States (or North America) and host-shifted to Schinus from an endemic plant host, and 3) it is endemic to South Africa (or Africa) and host-shifted to Schinus from an endemic plant host. The first hypothesis of a South American origin is the simplest to accept because it eliminates the need for biological inventiveness or ‘host-shifting’. In this hypothesis, M. transvaalensis was introduced into the United States and/or Africa from South America in the seeds of its natural host, Schinus (NB. The name was first used by a taxonomist in Africa, but if a taxonomist in California, for example, had discovered the wasp first it could as easily have been called californensis). This hypothesis, however, is suspect for several reasons. First, historically no species of Megastigmus (126 known) have been reliably reported from South America (for a discussion see Habeck et al. 1989; Grissell 1999; Grissell & Heydon 1999). Second, all other known species of anacardiaceous seed-feeders (viz., M. rhusi, M. thomseni (Hussey), and M. pistaciae), are demonstrably Old World in origin. Third, M. transvaalensis is morphologically indistinct from either M. rhusi or M. pistaciae and can be assumed to share at best a sibling species relationship with them. Fourth, extensive collecting, over at least a 10 year period by several biological control workers in Brazil, produced no wasps (Habeck pers. comm.). This last point admittedly is not much in the way of proof because the wasp’s absence could be the result of true (i.e. evolutionary) absence or poor sampling (i.e. artificial absence). Recently specimens of M. transvaalensis have come to our attention from Argentina, collected in 1996 (Wheeler pers. comm.), and from Brazil, collected in 1998 (Perioto pers. comm.). Both collections were from Schinus seed. Although an hypothesis of geographic origin in South America cannot be positively ruled out, the above reasons cast doubt on this region as a point of origin. The second hypothesis of a North American origin and host-shifting is perhaps the least likely to accept for the following reason. Although Megastigmus spp. have been heavily sampled and are well-known in North American (Breland 1949), not a single species except M. transvaalensis, M. rhusi, and M. pistaciae have been associated with plants in the family Anacardiaceae. That none of these wasps are associated with native North American plant species, and that all three are found on exotic, introduced plant species of Neotropical and Afrotropical origin, indicates the likelihood that they are not endemic to North America. For this reason, we tend to dismiss the probability of a North American origin. The third hypothesis of an Afrotropical origin for M. transvaalensis is the easiest to explain based on the systematics of Megastigmus species, their known distribution, and their known feeding habits. It does require, however, the biologically awkward step of an endemic seed wasp shifting from a native host plant to the introduced host Schinus.
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 273
Based upon the last hypothesis and the possibility that M. transvaalensis and M. rhusi were at best sibling (if not conspecific) species, the senior author undertook a study in 1998 to examine the indigenous South African species of Megastigmus associated with Rhus and to investigate the potential of host-shifting of Megastigmus between species of Schinus and Rhus. If host-shifting could be demonstrated between these two species, it was believed that further clues as to the possible movement of M. transvaalensis into the New World could be found.
Evidence for host shifting If an endemic South African species of Megastigmus attacking Rhus has host-shifted to the introduced South American Schinus, as suggested in the final hypothesis above, then it should be possible to demonstrate reciprocal movement of Megastigmus species between the two plant genera in either direction. Because there is no way yet known to distinguish morphologically between the Megastigmus ‘species’ from Rhus or from Schinus, we refer to the former as M. ‘rhusi’ and the latter as M. ‘transvaalensis’ as if they were, indeed host-restricted species. To perform experiments that would substantiate these hypotheses absolutely would require uninfested plants of Rhus and Schinus in fruit, a controlled environment capable of excluding all but the test subjects, the ability to rear and isolate sexes of the wasp on demand (to study pre- and post-mating effects), and many months or years of work. Given the lack of time available for this study, we could not meet these requirements. Therefore, we devised a simple procedure to score ritualised behaviour of wasps based on the immediate and simultaneous availability of alternate host plants and wasps. These are acceptance tests that merely indicate that wasps will or will not attempt to oviposit into a given host under similar conditions. The results are offered only as an ‘indication’ of possible host behaviour and not as proof positive. Methods and materials
Host acceptance tests were performed outdoors in a 9 cm Petri dish under daylight conditions but out of direct sunlight. Air temperatures varied between 75–85˚F. The Petri dish was swabbed with alcohol, rinsed with water, and dried between each newly introduced host tested. Fruit samples were collected from branches of Schinus and Rhus and tested either the same or next day. A small cluster of fruit containing from 30 to 40 drupes was cut from a branch and all leaves were stripped from it. Within a cluster all fruits were of the same age, and all were post petal fall, just barely beginning to swell, and still green. A prepared cluster of fruit from the same host plant as the wasp was provided first. Biological observations based on the female on her own host plant were thought to be the best indication of how a particular female might be reacting in general to her surroundings and was one attempt at standardisation. A single female wasp was placed on this fruit and her activities monitored, second by second, for 15 min, then the fruit cluster was removed and a cluster of the alternate host genus was placed for an additional 15 min. This represented one test, and the procedure was repeated 10 times with 10 different female wasps. Earlier tests, first using periods of 1 h and then 30 min, did not provide more useful data than the 15 min tests. The final test resulted in monitoring the behaviour of 10 wasps for a total of 2.5 h on each host. In each of these tests 10 females were chosen at random from a sample of wasps. In each case a female was placed directly on the fruit sample by aspiration but the 15 min timing was not begun until the female spent at least 10 sec on the sample. If she immediately flew off the sample, she was replaced on the sample until she remained there for 10 sec. This was enough time to establish contact and an immediate interest in exploring her surroundings. Once the 10 sec was passed, no
E. E. Grissell and K. R. Hobbs 274
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Summary of host acceptance test: Females of Megastigmus ‘rhusi’ reared from drupes of Rhus angustifolia and placed on drupes of Schinus terebinthifolius. Ten females were tested, each of which was placed for 15 min on drupes of the same species from which she emerged and then for 15 min on drupes of the alternate potential host species. Total testing time was 18000 sec [15 min (900 sec) x 2 plant hosts (= 1800 sec) x 10 females tested]; during unrecorded time, females were walking about on the fruit and stems or wandering about the test chamber. Average time on drupe (sec) (No. of observ.)
Average time per ovipositor probe (sec) (No. of observ.)
Rhus angustifolia
15.8 (160)
5 (1)
0 (0)
Schinus terebinthifolius
14.5 (276)
10.6 (17)
75.5 (9)
Plant Taxon
Average time per ovipositor insertion (sec) (No. of observ.)
additional prompting was attempted. In several cases females were placed on fruit and immediately observed to be morphologically impaired (i.e. legs or antennae aberrant); these females were replaced by undamaged ones. A standard of stereotyped biological behaviours was established using field observations and the initial 1 h or 30 min tests. These behaviours fell into three categories: 1) seconds spent on the fruit; 2) seconds spent probing the fruit (probing is a behaviour in which the female arches the abdomen and touches the surface of the fruit with her ovipositor); and 3) seconds spent inserting the ovipositor into the fruit. In each test the number of instances of each of these behaviours was noted as well as their length of time. No time was recorded when the wasp was off the fruit cluster, but the summations of the 3 time periods subtracted from 900 sec (15 min × 60 sec) equals the ‘down-time’, or time off the fruit cluster. For experiment 1 (Table 1) fruits of R. angustifolia were collected at about 1000 m elevation (Franschhoek Pass, Western Cape Province) and emerging females were tested on a cluster of the same fruit and then on a cluster of S. terebinthifolius fruit collected in Cape Town (Western Cape Province) at about sea level. Fruits from the Schinus tree were heavily infested with M. transvaalensis, so there was no way to know if they were already infested with eggs or larvae without dissecting specific fruits used in the test. Dissections would have made the fruits useless for testing. A single fruit cluster (with 36 drupes) was used for all tests, and each fruit that received a ‘complete insertion’ of the ovipositor was marked to determine if an egg was actually deposited. All fruits were dissected after the tests were completed, and a total of eight (of 36) ovules were found to have either an egg or larva in it. The fruit clusters, even if previously infested, appeared to provide ample opportunity for oviposition by females from Rhus. For experiment 2 (Table 2) adult female wasps were collected on a single, isolated tree of S. terebinthifolius in Cape Town (Western Cape Province). Although these females were not reared for the tests, six weeks of prior experience working with this tree (and others) demonstrated that wasps were emerging routinely from its fruits and were reinfesting other fruits on the tree. Females were first tested on a single fruit cluster (with 40 drupes) of the same Schinus tree and then switched to a cluster of 35 drupes of R. laevigata collected at approximately 0.5 km from the Schinus site (at 400 m elevation). No dissections were made. In both experiments, females had an opportunity to mate prior to the test (i.e. males were present), but it was not known if tested females were mated or unmated. None of the wasps were fed.
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 275
Table 2 Summary of host acceptance test: females of Megastigmus ‘transvaalensis’ collected on drupes of Schinus terebinthifolius and placed on drupes of Rhus laevigata. Ten females were tested, each of which was placed for 15 min on drupes of the same species from which she was collected and then for 15 min on drupes of the alternate potential host species. Total testing time was 18000 sec [15 min (900 sec) x 2 plant hosts (= 1800 sec) x 10 females tested]; during unrecorded time, females were walking about on the fruit and stems or wandering about the test chamber. Average time on drupe (sec) (Number of observ.)
Average time per ovipositor probe (sec) (No. of observ.)
Average time per ovipositor insertion (sec) (No. of observ.)
Schinus terebinthifolius
9.5 (324)
24.7 (17)
63.6 (7)
Rhus laevigata
11.7 (314)
37.6 (29)
21.7 (9)
Plant Taxon
Experimental results and discussion
Megastigmus ‘rhusi’ reared from Rhus, tested on Schinus Results (Table 1). The average time females spent on fruit of either the plant host (Rhus angustifolia) or ‘non-host’ (Schinus terebinthifolus) was about equal (15.8 sec on Rhus; 14.5 sec on Schinus). However, the number of encounters (160) for females on their own host, Rhus, was only about half that on the non-host (276). On Rhus, females averaged half as much time probing (5 sec) as on Schinus (10.6 sec), and they were nearly 20 times more likely to probe the non-host (17) than their own host (1). On Rhus no females reared from Rhus even attempted ovipositor insertion, but on the non-host, Schinus, at least 9 insertions were made with an average time of 75.5 sec per insertion. The question of whether any of these insertions resulted in successful oviposition is unknown despite dissections of fruit. Discussion. The data indicate that Rhus ‘rhusi’ females more readily inserted their ovipositors into the non-host (Schinus) than into their own host (Rhus). This argues in favour of alternate host acceptance, although completed oviposition into ovules could not be demonstrated. Acceptance of a non-host indicates that M. ‘rhusi’ is potentially adaptable to an alternative host. This adaptability could simply have been a result of the non-suitability of their own host (Rhus) as presented. But because the fruits were taken from the same plants from which the wasps emerged, and were the only alternative available to the wasps within flying distance, it would appear that non-suitability is an unlikely explanation. Although this evidence is largely circumstantial it demonstrates that host shifting (from Rhus to Schinus) is entirely possible.
Megastigmus ‘transvaalensis’ from Schinus, tested on Rhus Results (Table 2). Based on averages alone, time spent on each of the fruit choices was about equal at about 10 (Schinus terebinthifolius) to 12 (Rhus laevigata) sec. The number of encounters for females on their own Schinus host (324) and the related Rhus host (314) were also about equal. Schinus females spent more time on average probing fruits of Rhus (37.6 sec) than probing their own host (24.7 sec) and also probed the Rhus fruits more often (29 times). Schinus females spent three times longer per insertion into their own host (63.6 sec) than into Rhus (21.7 sec), but the number of insertions was about the same for both hosts (7 and 9 respectively). Discussion. The data indicate that M. ‘transvaalensis’ females inserted their ovipositors into the nonhost (Rhus) just as readily, if not more often, as into their own host. As with the reciprocal test (above), this again argues in favour of alternate host acceptance. Acceptance of a non-host indicates
E. E. Grissell and K. R. Hobbs 276
Hymenoptera: Evolution, Biodiversity and Biological Control
that M. ‘transvaalensis’ is potentially adaptable to an alternative host. Although the evidence is largely circumstantial it demonstrates that host shifting (from Schinus to Rhus) is possible.
Methods of distribution The methods of past distribution of M. transvaalensis are open to speculation and may never be known. We do know that M. transvaalensis occurs in the New World in California, Florida, Hawaii, Brazil, and Argentina. The only current hypothesis proposed for the movement of this seed-feeding wasp is through the world-wide trade in pink peppercorns proposed by Habeck et al. (1989). The fact that this trade is centred in Réunion reconfirms an African connection for the wasp, but no one has actually gone to this area and surveyed the fauna. The basic problem with this method of introduction or distribution is that it presupposes a number of not-soprobable events. The wasps would have to be transported into a new location carried inside the imported seed, then they would have to emerge from this seed, escape from the container holding the seed, escape from the building holding the container, and finally find a suitable host susceptible to oviposition. We believe we have enough direct and circumstantial evidence to propose a simple hypothesis for the initial introduction of M. transvaalensis into the New World, particularly into California. The occurrence of the wasp in other areas poses a secondary question for which we also suggest a methodology of movement. In South Africa, we have demonstrated that female Megastigmus attacking Rhus (called M. rhusi) and Schinus (called M. transvaalensis) appear to accept their reciprocal hosts, at least as sites of possible oviposition. Because the two wasp taxa are morphologically indistinguishable we propose that a single species is involved (the nomenclature and taxonomy of these names will be explained in a paper in preparation by Grissell and Prinsloo), and that this species host-shifted from endemic Rhus spp. to introduced Schinus spp. In California, African sumac (Rhus lancea), the type host of M. rhusi, was introduced in some past (but undetermined) time, and we believe that its seed-feeding wasp was introduced with the plants (or their seeds). These plants then produced wasps that host-shifted to the fruit of Schinus trees. Thus, host-shifting occurred independently at least twice in two different regions of the world. In all likelihood there is no biological barrier between Rhus and Schinus, and the wasps most likely move freely between these hosts. If this method of introduction to the New World is the correct one, then the next problem that needs to be resolved is how M. transvaalensis has managed to spread to such diverse areas as Hawaii, Florida, and South America. The most direct method, as with California, would be the importation of Rhus lancea into these areas. We know of no records, however, of this plant in those areas, but this possibility should be investigated. The second most direct method would be the transport of infested Schinus plants or seed. The simplest method would surely have been the transport of stock trees complete with fruit. While today this may seem unlikely, there is no way to know how Schinus stock plants were transported from the mid 1800’s to early 1900’s, a time when the tree was being planted in California and Florida. Still, even accepting the argument that this could have happened, it seems odd that no adult wasps would be found until the 1960’s in the United States. Although Habeck et al. (1989) suggested that transport might be by seeds sold as spices, we suggest another method, not previously proposed. If seeds (as opposed to trees) were used for propagation, a common situation even today with forest tree species, it seems likely that infested seeds would be collected and exported along with viable ones – there is no external difference between the two. If
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 277
all seeds were planted in nursery beds, the viable seeds would sprout whereas the infested seeds would remain undisturbed in the soil. If this ‘used’ soil was then thrown into bins (or heaps) for later sterilisation, reuse, or discard, it is entirely possible that infested seeds would reside in the soil for some time and that wasps could eventually manage to emerge and fly at least a short distance. It is well known that in various species of Megastigmus emergence from seed may take up to one or two years (Milliron 1949). Thus, if ungerminated seeds were disposed of along with soil used to germinate Schinus seeds, it is entirely possible that these seeds could linger in places in the more unkempt sorts of plant propagation efforts. The crux of the matter appears to be whether or not suitable host plants would be nearby to which the emerging wasps could fly. It seems likely that such a propagation operation would support the production and growth of trees over a series of at least a few years. A nursery set-up with several years’ worth of seedling production and subsequent growth, along with concurrent emergence from newly planted seed and/or emergence from accumulated discarded seeds and soil might possibly provide the basis for an infestation of nearby growing trees in various stages of development. According to Hall and Vandiver (1997), S. terebinthifolius can produce trees that set seed in as few as three years from germination. Thus, it is possible in a nursery situation, that seedlings held as few as three years could be infested by wasps from seeds of a concurrent or previous year’s planting.
Conclusions In a survey of Rhus and Schinus in South Africa, we found M. transvaalensis in every habitat in which Rhus grew in the Western Cape Province from sea level to 1000 m, from isolated pristine nature reserve to agricultural fence rows, from seaside to desert, and from xeric to mesic habitats. The species is known to be reared from four Rhus species and has been collected extensively on three additional species, all of which are endemic to Africa. The senior author examined museum specimens from the South African Museum (Cape Town), National Collection of Insects (Pretoria), and Stellenbosch University Museum (Stellenbosch) that demonstrate M. transvaalensis is widespread throughout South Africa and has been collected there as early as 1894 (Carnarvon, South African Museum), and the earliest reared specimen dated from 1895 (Stellenbosch, Stellenbosch University Museum). Thus we can date the existence of the wasp in South Africa at least back 100 years, but the oldest records in the New World date back only to 1961. Given the evidence presented above, it appears likely that M. transvaalensis is an endemic widespread South African species that host-shifted from the seeds of Rhus spp. into the seeds of introduced Schinus trees, which are endemic to South America. The wasp has apparently achieved the same result in California, only in this case both of the host plants were introduced: a Rhus species from Africa, and two Schinus species from South America. Still unresolved is the question of number of introductions. Was this a single introduction into California, with subsequent spread, or multiple introductions into different regions of the New World? And if this was a single introduction, how did the wasp reach its other New World destinations? Finally, attempting to solve the question of origin, host shifting, and distribution of M. transvaalensis leads to one additional and important problem needing resolution. This has to do with M. pistaciae and its true identity. This species was discovered in California in 1968 attacking seeds of introduced, cultivated Pistacia (Robinson 1968), just six years after M. transvaalensis was discovered attacking Schinus, also in California (Harper & Lockwood 1961). The pistachio seed chalcid is endemic to the Old World, with a known distribution in coastal Mediterranean areas (Bouºek 1977) where it extends into Iran (Roques & Skrzypczynska in prep.), the Crimea,
E. E. Grissell and K. R. Hobbs 278
Hymenoptera: Evolution, Biodiversity and Biological Control
Transcaucasia and Turkmenia (Bouºek 1977). Megastigmus pistaciae is morphologically identical to M. transvaalensis except for its larger size. Historically it has been identified based essentially upon its host association. However, Pistacia is a genus closely related to Schinus and Rhus (see Furth 1985; Furth & Young 1988), and there is no reason to believe that if a species shifted from Rhus to Schinus that it could not also shift from Rhus or Schinus to Pistacia (or vice versa). Molecular studies are currently underway to determine the degree of genetic relatedness between M. transvaalensis (i.e. rhusi + transvaalensis) and M. Pistaciae from several different parts of the world. It would appear likely that they are minimally host races of the same species. If they turn out to be conspecific, then the potential pathways of introduction and spread are considerably increased.
Acknowledgements During the course of this study much help and sound advice were rendered by the following individuals to whom we express our sincerest thanks: Simon van Noort, Vincent Whitehead, Hamish Robertson and Brian Fisher (South African Museum, Cape Town); David McDonald and John Donaldson (Kirstenbosch Botanic Garden, Cape Town); Jan Giliomee and Henk Geertsema (University of Stellenbosch, Stellenbosch); Brian Barnes (Stellenbosch Institute for Fruit Technology, Stellenbosch); Gail Littlejohn and Mark Wright (Agricultural Research Council, Elsenburg); John Marshall (De Muel Farm, Ceres); Percy Watkins (Green Valley Nuts, Prieska); Gerhard Prinsloo (Agricultural Research Council, Pretoria); Chris Desjardins (University of Maryland, College Park, Maryland); Fred Roth (California State Polytechnic University, Pomona, California), and Nelson Perioto (Instituto Biologico, Ribeirao Preto, Brazil). We also thank Gregory Wheeler (USDA, ARS, Ft. Lauderdale, Florida), and John Huber (Canadian Forest Service, Ottawa, Ontario) who reviewed the manuscript and provided much helpful advice. EEG would like to thank the USDA for a Beltsville Area Fellowship Award, which allowed him to spend three months in South Africa. He also is especially grateful to Julie and Steven Sloan of Sloan’s Guest House, Cape Town, who allowed their premises and business to be invaded by all sorts of foreign objects, both living and dead, and still maintained their sense of humour.
References Archer, R. H. (1993) Anacardiaceae, pp. 474-480. In Arnold, T. H. & de Wet, B. C. (Eds), Plants of Southern Africa: Names and Distribution. Memoirs of the Botanical Survey of South Africa 62: 1-825. Barkley, F. A. (1944) Schinus L. Brittonbia 5: 160-198. Boucˇek, Z. (1977) A faunistic review of the Yugoslavian Chalcidoidea (Parasitic Hymenoptera). Acta Entomologica Jugoslavica (Supplement) 13: 1-145. Breitenbach, F. von. (1984) National List of Introduced Trees. (Dendrological Foundation) Promedia Publications, Silverton, S.A. CDFFP. (1999) Fire Resistive Landscaping Can Save Your House And Your Life: http:// www.firesafe.com/firescape.html. Cambell, F. (1999) Plants that hog the garden -- invasive plants in the United States; Fine Gardening on Line: http://www.taunton.com/fg/features/plants/invasive/2.html. Devine, R. (1998) Alien Invasion. The National Geographic Society, Washington, D. C. Ellis Farms. (1999) Ellis Farms Catalog: http://www.ellisfarms.com/trees.html. Furth, D. B. (1985) The natural history of a sumac tree, with an emphasis on the entomofauna. Transactions of the Connecticut Academy of Arts and Sciences 46: 137-234.
Megastigmus transvaalensis (Hussey) (Hymenoptera: Torymidae) in California 279
Furth, D. B. & Young, D. A. (1988) Relationships of herbivore feeding and plant flavonoids (Coleoptera: Chrysomelidae and Anacardiaceae: Rhus). Oecologia 74: 496-500. GRIN. (1998) Germplasm Resources Information Network, National Plant Germplasm System, ARS, USDA: http://www.ars-grin.gov/npgs/tax/index.html. Gray, A. (1876) Geological Survey of California, Botany, Vol. 1. Welch, Bigelow, & Co., University Press, Cambridge. Grissell, E. E. (1999) An annotated catalog of world Megastigminae (Hymenoptera: Chalcidoidea: Torymidae). Contributions of the American Entomological Institute. 31: 1-92. Grissell, E. E. & Heydon, S. (1999) The identity of two unplaced New World Megastigminae (Hymenoptera: Torymidae). Proceedings of the Entomological Society of Washington. 101: 611-613. Habeck, D. H., Bennett, F. D., & Grissell, E. E. (1989) First record of a phytophagous seed chalcid from Brazilian peppertree in Florida. Florida Entomologist 72: 378-379. Hall, D. W. & Vandiver, V. V. (1997) Brazilian pepper-tree. Florida Agricultural Information Retrieval System: http://hammock.ifas.ufl.edu/txt/fairs/4954. Harper, R. W. & Lockwood, S. (1961) Bureau of Entomology. Forty-first Annual Report. California Department of Agriculture Bulletin 2: 127-129. HEAR (1997) Hawaii Ecosystems at Risk Project. University of Hawaii, Department of Botany: http://www.hear.org. Hill, A. F. (1937) Economic Botany. McGraw-Hill, New York and London. Hussey, N. W. (1956) A new genus of African Megastigminae (Hymenoptera: Chalcidoidea). Proceedings of the Royal Entomological Society of London, Series B 25: 157-162. McCann, J. A., Arkin, L. N. & Williams, J. D. (1996) Nonindigenous aquatic and selected terrestrial species of Florida. Aquatic and Wetland Plant Information Retrieval System (APIRS): http:// aquat1.ifas.ufl.edu/mctitle.html. Milliron, H. E. (1949) Taxonomic and biological investigations in the genus Megastigmus with particular reference to the taxonomy of the Nearctic species (Hymenoptera: Chalcidoidea: Callimomidae). American Midland Naturalist 41: 257-420. Morton, J. F. (1978) Brazilian pepper — its impact on people, animals, and the environment. Economic Botany 32: 353-359. Munz, P. A. (1965) A California Flora. University of California Press, Berkeley and Los Angeles. NPSIPM. (1997) Identification and biology of kudzu, saltcedar and Brazilian pepper. The National Park Service Integrated Pest Management Manual, National IPM Network: http:// www.colostate.edu/Depts/IPM/natparks/exweeds1.html. Rice, R. E. & Jones, R. (1996) Seasonal monitoring of the pistachio seed chalcid. Kearney Plant Protection Group, Plant Protection Quarterly 6(1): 1-3. Rice, R. E. & Michailides, T. J. (1988) Pistachio seed chalcid, Megastigmus pistaciae Walker (Hymenoptera: Torymidae), in California. Journal of Economic Entomology 81: 1446-1449. Robinson, D. W. (1968) California Department of Agriculture Pistachio Seed Chalcid Progress Report 68-1. Rondeau, R. J., Van Devender, T. R., Bertelsen, C. D., Jenkins, P. D., Van Devender, R. K. & Dimmitt, M. A. (1999) Flora and Vegetation of the Tucson Mountains, Pima County, Arizona: http://eebweb.arizona.edu/herb/Tucsons/. SDCLUE. (1999) County of San Diego acceptable plants for a defensible space in fire prone areas: http://www.co.san-diego.ca.us/cnty/cntydepts/landuse/plantlist.html. Sim, T. R. (1927) Treeplanting in South Africa. Natal. Witness, Limited. Thonner, F. (1915) The Flowering Plants of Africa. Dulau and Co., London.
E. E. Grissell and K. R. Hobbs 280
Hymenoptera: Evolution, Biodiversity and Biological Control
USDA, NRCS (1997) The PLANTS database. National Plant Data, Baton Rouge: http:// plants.usda.gov. Vettel, W. G. & Harper, R. W. (1969) California Department of Agriculture Pistachio Seed Chalcid Progress Report No. 69-1. Wickson, E. J. 1921. California Nurserymen and the Plant Industry 1850-1910. The California Association of Nurserymen, Los Angeles.
Biology of an Extant Species of the Scolebythid Genus Dominibythus (Hymenoptera: Chrysidoidea: Scolebythidae), with Description of its Mature Larva Gabriel A. R. Melo Departamento de Biologia, FFCLRP, Universidade de São Paulo, Av. Bandeirantes 3900, 14040-901, Ribeirão Preto, SP, Brazil (e-mail:
[email protected])
Introduction Scolebythids, a rare and little diverse group of chrysidoid wasps, are known only from tropical areas of the New World, Africa (including Madagascar) and Australia. Three extant genera have been described: Clystopsenella Kieffer (New World and Australia), Scolebythus Evans (Madagascar) and Ycaploca Nagy (Africa and Australia). These three genera contain only one described species each. Two additional genera were recently described from fossil material: Dominibythus Prentice & Poinar (Dominican amber) and Libanobythus Prentice & Poinar from Lebanese amber (Prentice et al. 1996). Very little is known of the biology of scolebythids and all the information available is based on circumstantial evidence (summarised in Brothers 1981). Immatures of Ycaploca evansi Nagy have been assumed to develop gregariously on larvae of cerambycid beetles and a group of five females and one male of Scolebythus madecassus Evans was found within a rotten stick in forest litter. This paper presents preliminary data on the biology of a recently described species, Dominibythus strictus Azevedo, found in south-eastern Brazil (Azevedo 1999).
Material and Methods Females of D. strictus were found walking on wood posts infested with anobiid beetles. To observe the interactions of the adult wasp with the beetle larvae and the later development of the wasp immatures, anobiid larvae taken from these posts were individually placed in small chambers carved in soft pith of the pedicel of Cecropia leaves (the pieces of pedicel were cut in half, and the chambers carved in the exposed pith). This chamber was connected to the outside by a narrow tunnel filled with loose frass taken from the channels made by the anobiid larvae. The whole set-up was covered with red cellophane and placed inside a small glass vial (closed with a plastic lid).
Results and Discussion The results obtained in this study confirm previous suggestions that scolebythids develop as gregarious parasitoids on wood-boring beetles. Dominibythus strictus proved to be an idiobiont ectoparasitoid of anobiid larvae. These beetle larvae were readily attacked by D. strictus females and the wasp immatures developed successfully on them. The female wasp, once in contact with the frass filling up the tunnel, immediately starts to dig through it using her mandibles, and pulling the loose frass under her body (using her legs,
281
Gabriel A. R. Melo 282
Hymenoptera: Evolution, Biodiversity and Biological Control
especially the fore and mid ones). After reaching the host chamber, the female starts to move more slowly and upon touching the beetle larva with her antennae, she turns around immediately and starts going back into the tunnel, pulling frass inside the chamber. She repeats this sequence of behaviours a few times before entering the chamber and stinging the beetle larvae. Stinging was observed only once. The host was stung ventrally, but there seems to be no precise spot for stinging. During stinging, the apex of the female metasoma is directed upward, but not over her body. The suggestion of Prentice et al. (1996) that scolebythid females would curve the metasoma over their body during stinging was not confirmed. After stinging the host, the female bites its integument and feeds on hemolymph coming out from the punctures. Dominibythus strictus females probably depend on host feeding for production of a full batch of eggs, since the eggs are laid only after about two days (n=8) from the first contact with the host. Five to seven eggs (n=5) are laid per host (the batch with seven eggs is perhaps atypical since two of the seven larvae were cannibalised at the end of the feeding phase). Gregarious larval feeding is also present in Sclerogibbidae and several Bethylidae and might be part of the Chrysidoidea groundplan. There seems to be no specific position on the host for egg laying, but the eggs seem to be preferentially laid along the posterior half of the host’s body. The eggs took about four days (n=7) to eclose and feeding lasted from 5 (n=3) to 6 days (n=3). The larvae have a distinct drop-like shape when young, becoming progressively elongated toward the second half of their development (Fig. 1). Their head capsule is very reduced (see description below) and they apparently feed only through sucking. After eclosion and throughout the feeding phase, the larvae remain in the same position on the host’s body. The fully-grown larvae remain close to each other and spin their white, cottony cocoons contiguously. The developmental time from egg to adult lasted about 34 days (n=2). As in other scolebythids, males of D. strictus are distinctly smaller than females. This difference in size is already noticeable at the end of the larval stage. Only one male is produced in each batch of immatures. Preliminary data revealed that the male emerges first and opens the female cocoons (at this stage, the females had moulted only the posterior half of their metasoma), which suggests sib mating. Two females having had contact only with their sib mates were able to produce females, as well as males, providing additional evidence for occurrence of sib mating. Sib mating, associated with parasitism of beetle larvae living in wood, has probably facilitated successful establishment of scolebythids in new regions after inadvertent introductions by humans. The presence in Africa and Australia of what has been considered the same species of Ycaploca (Nagy 1975) might represent an accidental introduction from one of these places to the other, as it has already been suggested by Naumann (1990). The peculiar morphology of the prothorax of adult scolebythids seems to represent an adaptation for making lateral turns in narrow tunnels, and not for holding the prey as suggested by Prentice et al. (1996). Females are able to turn around within the tunnels of the artificial set-ups by simply moving laterally their head and prothorax and then pulling around the rest of the body.
Mature Larva of D. strictus Description Small, whitish, unpigmented (including mandibles), fusiform larvae (Fig. 1); anus small, terminal; pleural lobes not well developed; body segmentation somewhat indistinct; ten pairs of apparently functional spiracles, 1st and 4th–7th the largest and approximately 23 µm in diameter, 3rd,
Biology of an Extant Species of the Scolebythid Genus Dominibythus 283
Figures 1–6
Mature larva of Dominibythus strictus: (1) body, lateral view; (2) sixth pair of spiracles, frontal (left) and lateral views (right); (3) ninth spiracle, lateral view; (4) mandible, antero-dorsal view; (5) head, lateral view; (6) head, frontal view.
8th and 9th and 18 µm, and 2nd and 10th pairs the smallest and about 14 µm in diameter; atrium simple, opening into subatrium unarmed (Fig. 2); subatrium well developed, in some spiracles as wide as atrium (Fig. 3); integument smooth except for rows of spinules on anterior thoracic segments; spinules directed caudad and arranged, side by side, in rows forming rings around the body, the rings forming two large bands, one immediately behind the head capsule and the other immediately anterior to 1st pair of spiracles, the spinules being absent in between. Head small and poorly differentiated posteriorly (Fig. 5); coronal suture and parietal bands absent; antenna represented by a pair of sensilla, each pair on a weak protuberance but not forming a distinct orbit (Figs 5 and 6); lower frons with a sensilla-like structure on each side; clypeus and clypeo-labral suture only weakly indicated; labrum broadly convex apically, with two lateral pairs of sensilla; labral margin not spinulose; mandibles small, somewhat hidden by labrum and maxillae (Fig. 6), tridentate apically (Fig. 4); maxilla with a conspicuous, button-like galea and a very reduced, conical palpus (Fig. 6); galea bearing four sensilla; surface of maxilla apparently without any spinulose areas; labium not protuberant, largely fused to maxillae, and with a few, scattered setae; labial palpi similar to maxillary galeae; spinneret apparently a short, tubular projection (the spinneret could not be studied in more detail because this area deformed during KOH clearing).
Comments The larvae of D. strictus are somewhat similar to those of Bethylidae described by Evans (in Evans et al. 1987). They have in common a small and incompletely formed head, bands of spinulose integument on the thoracic segments, 2nd pair of spiracles reduced, antennae weakly differentiated, small mandibles, and short, button-like labial palpi. In Dominibythus, the large sensory
Gabriel A. R. Melo 284
Hymenoptera: Evolution, Biodiversity and Biological Control
structure on the maxilla is being interpreted as the galea, since a smaller structure, here interpreted as the palpus, is also present. However, it is possible that this latter structure is not the palpus, and therefore a galea would be absent in Dominibythus. In Bethylidae, the galea is considered absent. Reduction of the galea seems to be the rule in apocritans, and not the contrary, i.e. reduction of the palpus. Among chrysidoids, tridentate mandibles are present in some Bethylidae and in Chrysididae, and bands of spinulose integument are also found in some Dryinidae, besides Bethylidae.
References Azevedo, C. O. (1999) A key to world species of Scolebythidae (Hymenoptera: Chrysidoidea), with description of a new species of Dominibythus from Brazil. Journal of Hymenoptera Research 8: 1-5. Brothers, D. J. (1981) Note on the biology of Ycaploca evansi (Hymenoptera: Scolebythidae). Journal of the Entomological Society of South Africa 44: 107–108. Evans, H. E., Smith, D. A., Middlekauff, W. W., Finlayson, T. & McGinley, R. J. (1987) Order Hymenoptera. pp. 597–710. In Stehr, F. W. (Ed.), Immature Insects. Kendal/Hunt, Dubuque. Nagy, C. G. (1975) A new genus of Scolebythidae (Hymenoptera) from South Africa and Australia. Journal of the Entomological Society of South Africa 38: 75–78. Naumann, I. D. (1990) The aculeate wasps and bees (Hymenoptera) of Norfolk and Philip Islands. Australian Entomological Magazine 17: 17–28. Prentice, M. A., Poinar, G. O., Jr., & Milki, R. (1996) Fossil scolebythids (Hymenoptera: Scolebythidae) from Lebanese and Dominican amber. Proceedings of the Entomological Society of Washington 98: 802–811.
Defense Adaptations in Velvet Ants (Hymenoptera: Mutillidae) and their Possible Selective Pressures Donald G. Manley Department of Entomology, Clemson University, Pee Dee Research and Education Center, 2200 Pocket Rd., Florence SC 29506-9706 USA (email:
[email protected])
Introduction Insects make up a large part of the diet of many animals, both vertebrates and invertebrates. Most velvet ants (Mutillidae) are relatively large, brightly coloured and conspicuous. They possess a wide variety of defensive adaptations, some of which are found only in females, some only in males, and some in both sexes. These include the sting, warning coloration (both aposematic and pseudaposematic), cryptic coloration, ant mimicry, stridulation, exudate production, cursorial speed, hard cuticle, flight and powerful mandibles. It is reasonable to assume that such measures evolved under selection pressures from certain predators. The purpose of this study was to determine what organisms might represent potential predators to velvet ants and may have led to the development of these defense mechanisms.
Overview of Mutillid Defenses The most significant defense mechanism exhibited by velvet ants is their sting, which is formed from a modified ovipositor and, therefore, is found only in females. It may be as long as the metasoma itself, nearly half the length of the entire body (Figs 1, 2). Males have evolved elaborate ‘pseudostinging’ behaviours. They possess two postero-lateral, sclerotised processes, the paraprocts. When held in the fingers, a male will manipulate the abdomen into a position so that it can be thrust down to contact the skin of the holder, much like the female when stinging (pers. observ.). The combination of the visual image of the abdomen pumping and the pricking sensation of the paraprocts is often enough to effect the release of the captured male. Mutillids possess a formidable pair of mandibles. Although the primary defensive role of the mandibles would appear to be grasping and holding, this action is capable of breaking the skin and causing pain (pers. observ.). Velvet ants also have a very heavy, deeply pitted cuticle, which is effective in repelling the stings and bites of other insects (Evans & Eberhard 1970), and which makes these insects difficult to pin out for insect collections. The bright, conspicuous colour of most mutillids may serve as a warning signal to prospective predators (i.e. aposematic coloration), the latter presumably learning to avoid the painful or lethal sting (Metcalf & Flint 1932). Some males, although harmless, appear to mimic the colour pattern of females or other aculeate Hymenoptera (i.e. pseudaposematic coloration) and thus may avoid predation.
285
Donald G. Manley 286
Figure 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Dasymutilla occidentalis (L.), female, with sting extended (scale in mm).
A few species of mutillids have cryptic colour patterns. For example, Dasymutilla gloriosa (Saussure), sometimes referred to as the ‘thistle down mutillid’ (Cockerell 1895), inhabits the deserts of New Mexico, Arizona and southern California, and is covered with long, white hairs. Unless moving, it is very difficult to see against a light, sandy background. Furthermore, it seems to be found only in association with creosote bush (Larrea tridentata Coville). This mutillid looks very much like the fruit of the creosote bush, and even moves in a jerky pattern that resembles the creosote fruit blowing in the wind (pers. observ.). Velvet ants possess a file on the base of the third metasomal tergite which, when rubbed against the underside of the second metasomal tergite, produces a ‘squeaking’ sound (Hinton et al. 1969). This stridulation is produced by both sexes and is known to serve a sexual role (Spangler & Manley 1978). It is also produced when individuals are disturbed and is suspected of increasing the effectiveness of warning coloration (Kirkpatrick 1957). Schmidt and Blum (1977) demonstrated that both male and female mutillids release an exudate, apparently from the mandibular glands, which acts in a defensive capacity, much like alarm pheromones in ants. For at least some groups of velvet ants, this exudate is reported to elicit a pungent odour (Hennessee pers. comm.). Because all female mutillids lack wings, escape by flight is an option available only to males. Males of some genera (e.g. Timulla Ashmead) are known to carry females during mating (i.e. phoretic copulation), and this may protect both sexes from potential predators, as is reported for some tiphiids (see Brown this volume). Cursorial speed and erratic movement patterns can also serve in a defensive capacity, particularly for females. Anyone who has tried to capture a velvet ant knows that they can move rapidly, often in a zig-zagging pattern that makes capture difficult. Schmidt and Blum (1977) reported relatively fast movement for Dasymutilla occidentalis (L.) which has a mean ground speed of about 0.5 km/hr.
Materials and Methods Five potential predators were used in laboratory experiments: two southern grasshopper mice (Onychomys torridus (Coues)), a scorpion (Hadrurus arizonensis (Ewing)), two lizards (Scelo-
Defense Adaptations in Velvet Ants (Hymenoptera: Mutillidae) 287
Figure 2
Dasymutilla magnifica Mickel, female, with sting extended (scale in mm).
porus magister Hallowell and Dipsosaurus dorsalis (Baird & Girard)), and an undetermined mantid. Experiments were conducted by introducing velvet ants into a container with the predator. A glass jar was used in the experiment with the desert mantid, and a 20 × 30 cm glass aquarium covered by a coarse mesh wire screen was employed in all other experiments. In all trials except for the lizards, the predator was offered other arthropods after a velvet ant to determine if it would respond to other potential prey. The first experiment was undertaken on 23 August 1974, when a female Dasymutilla clytemnestra (Fox) was placed in a glass jar with a desert mantid captured at light the previous night. Following removal of the velvet ant, two moths and a fly were introduced into the jar (one at a time) to test the predatory response of the mantid. The second experiment was undertaken on 6 September 1974, using two laboratory-born southern grasshopper mice. Since the mice had previously been fed on commercial rat pellets, they were conditioned to live food for one week prior to the experiment using two live crickets per day. The mouse that had shown the keenest predatory response was transferred to the glass aquarium and allowed to become acclimated for one hour, and then a female Dasymutilla magnifica Mickel was introduced into the aquarium. In order to test the mouse’s predatory drive, a bumblebee was then placed in the aquarium with the mouse. The second mouse was then tested with a velvet ant, and a milkweed bug (Oncopeltus fasciatus (Dallas)) was used to test ‘post-mutillid’ predatory response. About one hour later, the first mouse was again introduced into the aquarium and exposed to a female D. magnifica, followed by a female Dasymutilla satanas Mickel. The third experiment was conducted on 13–20 December 1974. On the first day, a female D. magnifica was placed in the aquarium with a scorpion (H. arizonensis) that had been captured the previous August. Observations were made periodically for the next seven days. Cockroaches and mealworms were subsequently placed in the tank to test the scorpion’s ‘post-mutillid’ predatory response. The final experiment took place on 14–19 May 1975 using two lizards caught three days before and which were not fed after capture. On 14 May the first lizard (D. dorsalis) was placed in
Donald G. Manley 288
Hymenoptera: Evolution, Biodiversity and Biological Control
the aquarium and allowed to acclimate for one hour. At that time a thread was tied around a female Dasymutilla nocturna Mickel so that its movements were not restricted. It was then lowered in front of the lizard. The second lizard (S. magister) was subsequently placed in the aquarium and the procedure repeated. On 16 May, these trials were repeated using the same two lizards. This time each lizard was presented first with a female D. nocturna, then with a female D. clytemnestra. A final test was conducted on 19 May by lowering another D. nocturna in front of S. magister. Additional observations were made during 1974 on two birds, a thrasher (Toxostoma sp.) and a roadrunner (Geococcyx californianus (Lesson)), and on various species of ants. Further tests conducted at the Southwestern Research Station near Portal, Arizona in 1991 (Manley & Sherbrooke in press) used the Texas horned lizard (Phrynosoma cornutum (Harlan)). Two horned lizards were first fed harvester ants (Pogonomyrmex spp.), their normal prey, to induce feeding. They were then offered female velvet ants representing four species of Dasymutilla: D. chiron (Blake) (orange and black), D. dilucida (black and white, ant-like in appearance), D. foxi (Cockerell) (red and white) and D. gloriosa (white). Two velvet ants of each species were used in the trials.
Results Over many years of fieldwork, the only capture of a velvet ant ever observed under natural conditions was made by a thrasher. This occurred near Scissors Crossing in San Diego County, California. The bird picked up a female mutillid (later determined to be D. magnifica) from the ground and flew off. However, a few meters into the air, the velvet ant was released by the bird and, upon examination, did not appear to be harmed. No other visible reaction by the bird was apparent. Although grasshopper mice routinely consume prey items such as stink beetles (Eleodes spp.) and scorpions, and they repeatedly fed on all other introduced prey items during the laboratory trials, at no time did they consume a velvet ant. One of the mice was finally able to kill a velvet ant, despite numerous stings. However, it never ate the velvet ant, even after several days. Much the same was true of the lizard species tested, although an individual desert spiny lizard (S. magister) finally swallowed one of the velvet ants (D. nocturna) and exhibited no ill effects, it did so only after considerable prompting and annoyance. None of the other potential predators tested made an attempt to capture any of the mutillids presented, except for the Texas horned lizard. One of the horned lizards did capture and eat two female velvet ants, but only D. dilucida, the ant mimic, and subsequently showed no adverse reaction. All of the other more colorful mutillids were completely ignored by the lizards.
Discussion Velvet ants are obviously well equipped defensively. The reasons for this have been previously examined by Schmidt and Blum (1977) and Manley (1984). Schmidt and Blum suggested that diverse defense mechanisms were necessary to ward off a broad range of potential predators, and that the various defense mechanisms may synergize with one another. These authors selected potential predators (e.g. grasshopper mice and gerbils) on the basis of their highly insectivorous nature and predatory drive. Because these organisms are not necessarily likely to encounter velvet ants in nature, the reasoning was that if they were able to successfully prey upon velvet ants, then other similar species may be able to do so as well. Conversely, if they could not successfully cope with mutillids, then other predators probably could not either. Warning coloration would only be of benefit if the potential predator possessed colour vision and was active
Defense Adaptations in Velvet Ants (Hymenoptera: Mutillidae) 289
during daylight hours. Since brightly coloured velvet ants are exclusively diurnal, they would be unlikely to encounter grasshopper mice or gerbils, which are nocturnal. However, if they did encounter them, other mechanisms, such as the sting, hard integument, cursorial speed, flight and sound, would presumably still be effective defenses. The fact that few other species are known to be successful predators on velvet ants suggests that the mutillids’ arsenal of defense mechanisms constitutes a formidable barrier. The one predator that has been shown to capture and consume a velvet ant, the Texas horned lizard, will do so only for those mutillids that mimic ants, their normal prey.
References Cockerell, T. D. A. (1895) A mutillid which resembles thistle-down. Psyche, Supplement 7(233): 6. Evans, H. E. & West Eberhard, M. J. (1970) The Wasps. University of Michigan Press, Ann Arbor. Hinton, H. E., Gibbs, D. F. & Silberglied, R. (1969) Stridulatory files as diffraction gratings inmutillid wasps. Journal of Insect Physiology 15: 549-552. Kirkpatrick, T. W. (1957) Insect Life in the Tropics. Longmans, Green & Co., London. Manley, D. G. (1984) Predation upon velvet ants of the genus Dasymutilla in California (Hymenoptera: Mutillidae). Pan-Pacific Entomologist 60: 219-226. Manley, D. G. & Sherbrooke (in press) Predation upon velvet ants (Hymenoptera: Mutillidae) by Texas horned lizards (Phrynosoma cornutum). The Southwestern Naturalist. Metcalf, C. L. & Flint, W. P.(1932) Fundamentals of Insect Life. McGraw-Hill , New York. Schmidt, J. O. & Blum, M. S.(1977) Adaptations and responses of Dasymutilla occidentalis (Hymenoptera: Mutillidae) to predators. Entomological Experiments and Applications 21: 99-111. Spangler, H. G. & Manley, D. G.(1978) Sounds associated with the mating behavior of a mutillid wasp. Annals of the Entomological Society of America 71: 389-392.
Introduction and Spread of Four Aculeate Hymenoptera In Italy, Sardinia and Corsica Guido Pagliano1, Pierluigi Scaramozzino2,3, Franco Strumia1 1
Museo di Storia Naturale-Pisa University- via Roma, 79 – 56011 Calci (Pisa), Italy
2
Museo Regionale di Scienze Naturali via Giolitti, 36, 10123 Torino, Italy
3 present address: Aluseo di Storia Naturale-Pisa University-via Roma, 79-560 m Calci (Pisa), Italy
Introduction In recent years we have observed the arrival in Italy of four conspicuous wasps species: Sceliphron caementarium (Drury), Sceliphron curvatum (Smith), Isodontia mexicana (Saussure) (Sphecidae), and Chrysis marginata Mocsáry (Chrysididae). Here we describe the spread of these four species in Italy and in the islands of Corsica and Sardinia based on 30 years of field observations together with an examination of material in the largest insect collections in Italian institutions.
Results and Discussion Sceliphron (Sceliphron) caementarium (Drury) Sceliphron caementarium is a Nearctic species which has been accidentally introduced into Europe. It was first established in southern France in 1970 where it is now widespread and apparently threatening native Sceliphron species (Hamon et al. 1989), in particular S. spirifex (L.) (Piek 1986). In Italy S. caementarium was found for the first time in 1990 in Tuscany, near the port at Leghorn and is now a common species on the Pisa plain and surrounding hills (Pagliano 1992; Strumia 1996). Our field observations in West Tuscany began in 1970 and, since that time, S. caementarium has not been observed outside of the Pisa plain (Fig. 1), supporting the idea of an accidental introduction from the port of Leghorn. Field observations show the rate of spread of this species to be very slow, only about 30 km in eight years. A second independent introduction of S. caementarium was discovered near the port at Ravenna (Campadelli et al. 1999). Compared with the population at Pisa, S. caementarium appears to be in competition with the native S. destillatorium (Illiger) at Ravenna, and its spread into surrounding areas appears to be even slower than on the Pisa plain. A third introduction of S. caementarium was discovered near the coast of western Liguria (Garlenda and Albenga (Imperia); Pagliano 1995) about 60 km from the French border (Fig. 1). Given the proximity, this population is likely to have arisen from the long established population in southern France. Sceliphron caementarium was also previously known from a single locality on Corsica (Solaro, June 1986) (Bitsch et al. 1997). In 1998, we found several individuals at various localities (viz. Etang d’Urbino, 18 May (east coast); St. Florant, 24 August (north coast); Liamone river-mouth, 21 August (west coast; Fig. 1), indicating that the species is widespread across the island. At all localities it was coexisting with two native species, S. destillatorium and S. spirifex.
290
Hymenoptera: Evolution, Biodiversity and Biological Control
Introduction and Spread of Four Aculeate Hymenoptera in Italy, Sardinia and Corsica 291
Figure 1
Known distribution of Sceliphron caementarium in Italy and Corsica.
Sceliphron (Hensenia) curvatum (Smith) Sceliphron curvatum was described from India and is also known from Pakistan, Nepal, Tadzhikistan, Kazakhstan and Afghanistan (Hensen 1987). In 1979 it was first recorded outside this range in south-eastern Austria (Fig. 2) (van der Vecht 1984), probably as a result of an accidental introduction (see van der Vecht 1984; Gepp 1995). In 1991 it was reported from Slovenia (Gogala 1995), while in 1995 it was recorded for the first time in northern Italy (Veneto, Emilia and Piedmont) (Scaramozzino 1995, 1996) (Fig. 2). In June 1996, S. curvatum was recorded from Sardinia and in 1998 from central Italy (Pisa and Rome). From the few available records it appears that this species is rapidly dispersing westwards and southwards, after being confined to south-eastern Austria for more than 10 years. Unlike S. caementarium, S. curvatum is associated with urban environments in Italy, often inside houses (e.g. Torino, Bologna, Pisa and Rome) (N.B. Hensenia Pagliano & Scaramozzino (1990) is a replacement name for Prosceliphron van der Vecht & Breugal (1968)). Isodontia mexicana (Saussure) Isodontia mexicana is a Nearctic species and probably accidentally introduced into Europe. It was first discovered in Hérault (southern France) in 1960 and is now distributed over the Mediterranean part of France and Catalonia (Bitsch et al. 1997). The species was not reported from Italy until 1985 (Scaramozzino & Pagliano 1987), but since then it has spread rapidly into northern and central Italy where it is now common and widespread (Fig. 3). More recently I. mexicana has reached southern Switzerland (Vernier 1995), Slovenia (Gogala 1995) and Croatia (Gusenleitner 1996). In 1997 it was discovered in Corsica (Folelli Beach and San Pellegrino) and was still present
Guido Pagliano, Pierluigi Scaramozzino, Franco Strumia 292
Figure 2
Hymenoptera: Evolution, Biodiversity and Biological Control
Known distribution and spread of Sceliphron curvatum in Italy including Sardinia, and Austria.
there in 1998. However, following extensive fieldwork in the late spring of 1997 and summer of 1998, I. mexicana appears to be absent from Sardinia and southern Italy (respectively).
Chrysis marginata Mocsáry Chrysis marginata is a medium sized, brightly coloured chrysidid, described from Caucasus, and known also from Palestine, Asia Minor and south-eastern Europe. In 1915 it was reported from Hungary (Kiss 1915) and in 1959 from Dalmatia (Island Krc) (Linsenmaier 1959). The species was first discovered in Italy in 1962–64 in Emilia (Rimini and around Bologna) and, a year later, in Tuscany (near Pisa). Since 1975, C. marginata has been discovered at several localities in northern Italy and Tuscany, where it has become a widespread and common species in different habitats. The southern-most known locality is near Rome (Fusano, August 1989). More recently (1996-1998), C. marginata was discovered from Western Liguria, Garlenda near Albenga (Savona), indicating its continued westward spread. If this movement continues, the species will reach southern France in the next few years. Further, Bregant (1998) has reported C. marginata from north-western Austria, near the German border, thus showing the species is spreading westward in the region north of the Alps.
Summary The absence of such large and conspicuous wasps in major collections prior to the dates discussed above, confirms their absence in Italy until recent times. Our observations show that three
Introduction and Spread of Four Aculeate Hymenoptera in Italy, Sardinia and Corsica 293
Figure 3
Known distribution and spread of Isodontia mexicana in Italy, Corsica, Switzerland, Croatia and Slovenia.
Figure 4
Known distribution and spread of Chrysis marginata in Italy and Dalmatia.
Guido Pagliano, Pierluigi Scaramozzino, Franco Strumia 294
Hymenoptera: Evolution, Biodiversity and Biological Control
species, S. caementarium, I. mexicana and C. marginata are well established in northern and central Italy, but have not yet reached southern Italy and Sicily. The fourth species, S. curvatum which is the most recent introduction, requires further study to confirm its distribution and rate of spread. This study reveals the importance of long-term field observations in documenting the introduction and spread of introduced wasps (and other insects).
References Bitsch, J., Barbier, Y., Gayubo, S., Schmidt, K. & Ohl, M. (1997) Hyménopteres Sphecidae d’Europe Occidentales vol. 2. Faune de France 82. Bregant, E. (1998) Hymenopterologische Notizen aus Österreich. 10 (Hymenoptera, Chrysididae). Linzer Bilogische Beiträge 30: 623-628. Campadelli, G., Pagliano, G., Scaramozzino, P. L. & Strumia F. (1999) Parassitoidi ed inquilini di Scheliphron caemetarium (Drury, 1773) (Hymenoptera Sphecidae) in Romagna, Bollettino Museo Regionale di Storia Nataturale, Torino 16: 225-240. Gepp, J. (1995) Die orientalischeMauerwespe Scheliphron curvatum (Smith, 1870) (Hymenoptera Sphecidae). Stapfia 37: 153-166. Gogala, A. (1995) Two non-european species of digger-wasps recorded also in Slovenia (Hymenoptera Sphecidae). Acta Entomologica Slovenica 3: 73-75. Gusenleitner, J. (1996) Kurzericht über Sphecinae in Istrien (Croatia) (Hymenoptera Sphecidae). Linzer biologische Beiträge 28: 817-819. Hamon, J., Bitsch, J., Schwartz, F., Maldès, J., Delmas, R., Adamski, A. & Tussac, H. (1989) Quelques observations sur la distribution en France d’un insecte américain Scheliphron caemetarium (Drury, 1773) (Hymenoptera Sphecidae). L’Entomologiste 45: 115-120. Hensen, R. V. (1987) Revision of the subgenus Prosceliphron van der Vecht (Hymenoptera, Sphecidae). Tijdschrift voor Entomologie 129: 217-262. Kiss, E. Z. (1915) Ujabb adatok Magyarrorszag Hymenoptera-faunajahoz. Rovartani Lapok 22: 76-86. Linsenmaier, W. (1959) Revision der Familie Chrisididae (Hymenoptera). Mitteilungen der Schweizerischen. Entomologischen Gesellschaft 32: 1-240. Pagliano, G & Scaramozzino, P. L. (1990) Elenco dei generi di Hymenoptera del mondo. Memorie della Società Entomologica Italiana 68: 3-210. Pagliano, G. (1992) Scheliphron caemetarium (Drury, 1773) (Hymenoptera Sphecidae) specie nuova della penisola italiana. Hy-Men 3: 5. Pagliano, G. (1995) Ampliamento dell’area di diffusione in Italia di Scheliphron caemetarium (Drury). Hy-Men 6: 11. Piek, T. (1986) Scheliphron caemetarium (Drury) supersedes S. spirifex Linnaeus in the Provence, France (Hymenoptera Sphecidae). Entomologische Berichten 46: 77-79. Scaramozzino, P. L. (1995) Nuovi arrivi da Est: Scheliphron (Hensenia) curvatum (Smith) (Hymenoptera Sphecidae). Hy-Men 6: 9-11. Scaramozzino, P. L. (1996) Nuova località di cattura di Scheliphron (Hensenia) curvatum (Smith) (Hymenoptera Sphecidae). Hy-Men 7: 9. Scaramozzino, P. L. & Pagliano, G. (1987) Note sulla presenza in Italia di Isodontia mexicana (Saussure, 1867) (Hymenoptera Sphecidae). Rivista Piemontese Storia Naturale 8: 155-159. Strumia, F. (1996) Accidental introduction of the Nearctic Scheliphron caemetarium (Drury) in Pisa Province, Italy (Hymenoptera: Sphecidae). Frustula Entomologica 19: 176-179. van der Vecht, J. (1984) Die orientalische Mauerwespe Sceliphron curvatum (Smith, 1870) in der Steiermark, Oesterreich (Hymenoptera, Sphecidae). Entomofauna 6: 213-219.
Introduction and Spread of Four Aculeate Hymenoptera in Italy, Sardinia and Corsica 295
van der Vecht, J. & van Breugel, F. M. A. (1968) Revision of the nominate subgenus Scheliphron Klug (Hymenoptera, Sphecidae). Tijdschrift voor Entomologie 111: 185-255. Vernier, R. (1995) Isodontia mexicana un Sphecini américain naturalisé en Suisse. Mitteilungen Schweizerische. Entomologische Gesellschaft 69: 169-177.
Biological Notes and Larval Morphology of Donquickeia (Hymenoptera: Braconidae: Doryctinae) Angélica Maria Penteado-Dias Universidade Federal de São Carlos, Departmento de Ecologia e Biologia Evolutiva, CP 676, CEP 13 565-905, São Carlos, SP, Brazil (e-mail:
[email protected])
Introduction The doryctine braconid genus Donquickeia Marsh is known only from two described species from Brazil (Marsh 1993, 1997a). It is similar to Semirhytus Szépligeti but differs in that vein 1m-cu of the fore wing meets 2-m beyond 2-RS, and that the oral opening is small and circular rather than wide and oval. From other New World doryctine genera it can be identified using the key in Marsh (1997b). During studies on the biodiversity of the brazilian braconid fauna we have found three new species of Donquickeia, all of which have been reared from the galls of cecidomyiids. This study presents the first description of the larvae of Donquickeia and discusses its biology, particularly in reference to the possibility that the genus is phytophagous.
Material and Methods Galls were collected from three plant species: Mikania sp. (Asteraceae), Eugenia rotundifolia Casar (Myrtaceae), both from Rio de Janeiro, RJ and from an unidentified species from a gallery forest at Luís Antônio, SP. For larval morphology, material was obtained by cutting galls transversely with a razor blade and the larvae removed. Terminology for larval morphology follows Capek (1970).
Results and Discussion The larval features of Donquickeia are most unlike those of other doryctines (¯apek 1970). They can be characterised by having small, toothless mandibles in proportion to the size of the head, a labial sclerite which is deep and complete ventrally, an epistome which is not well-developed, and the body surface with pointed papulas (Figs 1, 2). In this respect the morphology of the larval head is more similar to members of the endemic Australian Mesostoinae (Quicke & Huddleston 1989) than it is to other Doryctinae. In total, three species of Donquickeia were reared from cecidomyiid galls from three different plant species. One species, for which the larva is described above, was reared from leaf galls of an unknown cecidomyiid on an unidentified plant (Fig. 3). Two additional species of Donquickeia were reared from cecydomyiid galls on Mikania sp. and E. rotundifolia. The cecidomyiid responsible for the galls on the latter plant is Stephomyia rotundifoliorum Maia. The galls are formed at the bases of the leaves (Fig. 4), are unilocular, white inside, have a brown surface, and are cylindrical in shape (Maia 1993). For the first species of Donquickeia (from the unidentified plant), only one larva was found inside each cecidomyiid gall.
296
Hymenoptera: Evolution, Biodiversity and Biological Control
Biological Notes and Larval Morphology of Donquickeia 297
Figures 1–2
Donquickeia sp.: 1) cephalic structures of final instar larva; 2) larval integument.
At this stage it is not known whether members of Donquickeia are parasitic on the resident cecidomyiids or whether they are phytophagous and inquilines in the galls. The blunt, toothless form of the larval mandibles, which is unlike that of entomophagous larvae, and the fact that several other braconid species have recently been shown to have phytophagous larvae (Infante et al. 1995; Austin & Dangerfield 1998), means that the possibility of Donquickeia larvae also being phytophagous cannot be easily dispelled. Clearly, additional research is required to elucidate the complete biology of this unusual genus.
Acknowledgements I thank Ricardo Monteiro and Valéria Cid Maia for sending the material for study, and for providing photographs of E. rotundifolia galls. I also thank Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for financial support.
Angélica Maria Penteado-Dias 298
Figure 3–4
Hymenoptera: Evolution, Biodiversity and Biological Control
3) Galls (arrowed) on leaves of an unidentified plant collected in gallery forest area in Luís Antônio, SP, Brazil. 4) Galls (arrowed) in the base of leaves of Eugenia rotundifolia (Myrtaceae).
References Austin, A. D. & Dangerfield, P. C. (1998) Biology of the Mesostoa kerri Austin and Wharton (Insecta: Hymenoptera: Braconidae: Mesostoinae), an endemic Australian wasp that causes stem galls on Banksia marginata Cav. Australian Journal of Botany 46: 559-569. Capek, M. (1970) A new classification of the Braconidae (Hymenoptera) based on the cephalic structures of the final instar larva and biological evidence. Canadian Entomologist 102: 846-875.
Biological Notes and Larval Morphology of Donquickeia 299
Infante, F., Hanson, P. & Wharton, R. (1995) Phytophagy in the genus Monitoriella (Hymenoptera: Braconidae), with a description of a new species. Annals of the Entomological Society of America 88: 406-415. Maia, V. C. (1993) Considerações sobre Stephomyia Tavares (Diptera, Cecidomyiidae, Asphondyliidi), com descrição de quatro espécies novas associadas com Eugenia L. e Neomithranthes obscura (DC.) Legr. (Myrtaceae). Revista Brasileira de Zoologia 10: 521-530. Marsh, P. M. (1993) Descriptions of new western hemisphere genera of the subfamily Doryctinae (Hymenoptera: Braconidae). Contributions of the American Entomological Institute 28: 1-58. Marsh, P. M. (1997a) Replacement names for Western Hemisphere genera of the subfamily Doryctinae (Hymenoptera: Braconidae). Proceedings of the Entomological Society of Washington 99: 586. Marsh, P. M. (1997b) Subfamily Doryctinae. pp. 206-233. In Wharton, R. A., Marsh, P. M. & Sharkey, M. J. (Eds), Manual of the New World Genera of the Family Braconidae (Hymenoptera). Special Publication No. 1, International Society of Hymenopterists, Washington, D. C. Quicke, D. L. J. & Huddleston, T. (1989) The Australian braconid wasp subfamily Mesostoinae (Hymenoptera: Braconidae) with a description of a new species of Mesostoa. Journal of Natural History 23: 1309-1317.
This Page Intentionally Left Blank
PART
6
Biodiversity
This Page Intentionally Left Blank
Driving Miss DAISY: the Performance of an Automated Insect Identification System I. D. Gauld1, M. A. O’Neill2 and K. J. Gaston3 1
Department of Entomology, The Natural History Museum, Cromwell Rd, London SW7 5BD United Kingdom (email:
[email protected]) 2
Digital Vision, 11 Edmonds Court, Didcot, Oxfordshire OX11 8QY United Kingdom
3
Department of Animal and Plant Sciences, University of Sheffield, Sheffield S10 2TN United Kingdom
Introduction Accurate and repeatable species-level identification of organisms is a fundamental part of almost all biological work. Species names are the nouns in the language of biology. They are the labels via which one accesses the great array of information available about organisms. The growth in Information Technology (IT), and widespread development of electronic data bases, makes having the correct name for one’s study organisms ever more important. But accurate identifications are of broader concern than to just research biologists, they are vital in many applied areas. Whilst it is well known that correct identification underpins integrated pest management programmes (Lattin & Knutson 1982; Hawksworth 1994), and is essential for monitoring the spread of pollution and disease vectors (Chalmers 1996), it is perhaps less well-known that correct species’ names are vital for the successful implementation of the Convention on Biological Diversity. Identification is a fundamental part of conservation evaluation (Vane-Wright 1996); it underpins monitoring the effects of anthropogenic disturbance (Brown 1997) and is crucial for formulating conservation legislation (May 1990), and for issues relating to intellectual property rights (Bousouf et al. 1996). Accurate and repeatable identifications also underpin pharmaceutical prospecting (Reid et al. 1993), are vital in quarantine inspection, and have importance in many other activities of both an economic and educational nature (Whitten 1996). A mechanism whereby organisms can be reliably and repeatably identified is thus of broad sectoral interest. Despite the importance of accurate identification, the body of taxonomic expertise available to carry it out for insect pests, pathogens and environmental indicators is being steadily eroded world-wide (Gaston & May 1992). The demand for routine identifications now far outstrips the capabilities of the dwindling biosystematics community (Holden 1989; House of Lords 1991). This steadily worsening situation has attracted considerable international attention, initially from the inter-governmental Subsidiary Body for Scientific, Technical and Technological Advice (SBSTTA) to the Convention on Biological Diversity, and subsequently from the Conference of Parties to the Convention itself. It is now generally accepted that taxonomy is a sine qua non for the implementation of the Convention, and that the infrastructure for taxonomy in biodiversityrich tropical countries should be strengthened in order to provide a basis for the monitoring, inventorying and sustainable utilisation of biological diversity (di Castri et al. 1992; Janzen 1993; UNEP/CBD/SBSTTA/2 1996).
303
I. D. Gauld, M. A. O’Neill and K. J. Gaston 304
Hymenoptera: Evolution, Biodiversity and Biological Control
The problem caused by the lack of resources available for arthropod identification is aggravated both by the taxonomic community themselves, and by their traditional funding agencies. Efforts and resources are often primarily focused on areas where intellectual debate can easily be conducted, such as phylogenetic reconstruction. The more mundane tasks of monographic, floristic and faunistic studies are less attractive, both to scientists and to funding agencies, even though it is the results of such work that are in greatest demand by the global community in general. Compounding this problem is the undeniable fact that many of the traditional ‘applied’ taxonomic products (e.g. printed keys) are often almost impossible to use without both adequate reference collections and an extensive knowledge of arcane specialist terminology. Consequently, even where the literature to identify organisms exists, many biologists, let alone other potential users, cannot and do not use it (Greenslade 1985; Gauld 1986; Tilling 1987; Alberch 1993). In the jargon of the marketplace, the products of the taxonomic community are generally not appropriate for the needs of the potential user community. In attempts to rectify this situation and overcome the resulting ‘taxonomic impediment’, the traditional products of taxonomists are beginning to be augmented by electronic technology. First, was the development of computerised multi-access keys, starting with text-based keys (e.g. Pankhurst 1978) and culminating recently in multimedia works such as CABIKEY (White & Scott 1994). Whilst undoubtedly an advance over hard copy works, computerised keys still rely on the ability of users to compare pictorial information with specimens. Such skills are honed by years of practice in taxonomists, but other biologists often experience great difficulty in appreciating the subtle differences in shape and form which discriminate taxa, particularly amongst many groups of invertebrates. Using computers to present taxonomic characters, while relying on users to compare specimens to images or illustrations, represents a failure to utilise the full potential offered by information technology. Second has been the development and application of image analysis techniques. This area of technology has undergone great advances in recent years, raising the possibility of automating, or at least semi-automating, much of the process of routine taxonomic identification. However, such an approach has to date only been used in a very limited fashion (Weeks & Gaston 1997). For example, Daly et al. (1982) used image analysis to measure 25 morphometric characters of honey bees and then by discriminant analysis determined whether the bees were European or Africanised. Yu et al. (1992) measured the wings of ichneumonids by semi-automated image analysis and used discriminant analysis to accurately identify five species on the basis of differences in their wings. Such studies encourage belief that image analysis techniques may represent a way forward towards a large-scale taxonomic identification system based on computer vision. In an attempt to more fully automate identification, we developed a prototype novel automated identification system (DAISY – the Digital Automated Identification SYstem) the technical details of which have been fully described elsewhere (Weeks et al. 1997, in press).
DAISY: an Automated Identification System DAISY is a holistic vision based system which has been designed to classify fuzzy visual sets. It uses fuzzy matching techniques such as, for example, principal component analysis (PCA) based on human facial recognition work pioneered by Turk and Pentland (1991), or nearest neighbour classifiers (NNC) (e.g. Alexander 1984; Lucas 1997), in order to identify objects presented to it via a CCD camera attached to a PC. With such systems it is important to present objects in a standardised way. As alignment is important, we are principally working with planar, two-
Driving Miss DAISY: the Performance of an Automated Insect Identification System 305
Table 1 Mean Kendall-τ correlations between the test images of the five Costa Rican pimplines plus an extralimital species, and the five species classifiers. Species of test images
N. abramsae
N. lineata
N. mellosa
P. croceiventris
P. sumichrasti
Neotheronia. abramsae
0.861
0.693
0.751
0.541
0.774
N. lineata
0.564
0.865
0.725
0.513
0.682
N. mellosa
0.630
0.768
0.899
0.527
0.744
Pimpla croceiventris
0.487
0.543
0.505
0.832
0.485
P. sumichrasti
0.623
0.624
0.736
0.473
0.917
Pimpla sp. (Europe)
0.345
0.384
0.336
0.434
0.380
dimensional objects as it is relatively simple to eliminate misclassification errors due to poorly standardised pose. In the prototype this was achieved by aligning the object being imaged against a standard line on the CCD monitor. However, we are implementing more robust methods for standardising pose recovery within the DAISY system, using snake or caterpillar algorithms, processes that effectively extract the image of the object being imaged from the background, and allow this image to be rotated into a mathematically pre-defined position. Another advantage of using thin planar objects is that illumination can be easily standardised. This is important as optimal performance is only achieved using a uniform intensity of illumination throughout. In order to classify objects using computer vision technology, a classifier needs to be trained by providing it with examples of the objects we wish it to recognise. A critical difference between matching non-variable objects, such as fingerprints or retinal scans, and identifying biological species is, of course, that the latter exhibit ranges of intraspecific variation, not shown by the former. Thus it was necessary to train DAISY with a series of individuals encompassing the statistical variation of the species. This set of reference individuals is called the training set. Training set data are used to establish a species-classifier which codes for a given type of fuzzy object, the species. These species-classifiers operate as associative memories which correlate an unknown pattern (i.e. a specimen submitted for identification) with the training sets using either an NNC or PCA classification scheme. Each species-classifier returns a number a, the affinity, which is a measure of the similarity of the unknown to the material in the training set. A working DAISY system is thus a network of PCA or NNC based associative memories. When presented with an unknown specimen each of the n species-classifiers returns an affinity vector A = {a1, a2, … ai … an}, which is a measure of similarity of the object to the training set.
Performance of the Prototype The prototype DAISY system uses a first-past-the-post (FPTP) classification algorithm. That is to say the identity of the unknown is determined as that of the training set of the winning classifier – the species-classifier with the highest value of A. Typical results using the prototype DAISY system to identify five species of Costa Rican pimpline ichneumonids are shown in Table 1. For this, the Kendall-τ (affinity) value shown is the mean of 35 individual tests using different test images. In each case the mean correlation with the correct species-classifier is strikingly higher than the mean correlation with the classifiers of other species. Overall, the proportion of test images correctly identified to species was 0.94. However, correct identification can obviously only be achieved if the appropriate species-classifier for the test specimen is present in the classifier data base. A specimen for which no species-
I. D. Gauld, M. A. O’Neill and K. J. Gaston 306
Hymenoptera: Evolution, Biodiversity and Biological Control
0.9
Proportion of correct identifications
0.85
0.8
0.75
0.7
0.65 0
Figure 1
2
4
6
8
10
12
DAISY identifies correctly a greater proportion of specimens as the number of images used to train the species classifiers increases. (• ) Kendall-τ metric, (■) Vector-Difference metric (results based on iterative analysis of data set comprising 49 species of Culicoides and Forcipomyia (Diptera: Ceratopogonidae); see Weeks et al. 1999).
classifier is present still has a similarity to the species-classifiers. For example, a European species of Pimpla F. (P. hypochondriaca (Retzius)) would be misidentified as P. croceiventris (Cresson) if an identification was attempted using the available species-classifiers (Table 1). In most cases this does not present major difficulties as extralimital species were found to have much lower Kendall-t values (italics in Table 1) with the Costa Rican classifiers than do the correct species. In such cases misidentifications can be prevented by establishing a cut-off threshold value below which an unknown is rejected as not being conspecific with any of the species-classifiers. These results are encouraging, but if such an automated system is to have general applicability, then the properties of the system need to be more rigorously examined. How, for example, does accuracy change as the number of specimens in the training set is altered, or if physical conditions of image-grabbing change? And if the number of species-classifiers in the data-base increases, does accuracy change?
Effect of training set size on accuracy of the system As the number of images incorporated in the training set was increased the accuracy of identifications improved. With low numbers, the addition of one further image resulted in a marked
Driving Miss DAISY: the Performance of an Automated Insect Identification System 307
1.0 0.9 0.8 0.7 h l
0.6 0.5 0.4 0.3 0.2
Neotheronia lineata Neotheronia mellosa
0.1 0
1.0V
1.5V
2.0V
2.5V
3.0V
3.5V
4.0V
4.5V
5.0V
Voltage of illumination source Figure 2
Variation in illumination intensity, and corresponding change in the Kendall-metric of two species of Costa Rican Pimplinae correlated with the lineata classifier.
increase in accuracy. However, this rate in the increase of accuracy declined as the size of the training set was increased, until an asymptote was approached with approximately 15 images per training set (Fig. 1). With increasing numbers of specimens in the training set the variance in the proportion of correct identifications was also found to decline. These properties suggest that, for routine identifications, only relatively small training sets need to be constructed.
Effect of illumination intensity on accuracy of the system Throughout our tests a standardised illumination was used when imaging specimens for the training sets. To give optimal results when making an identification, a similar illumination was found to be necessary. Deviation from this optimum resulted in a decrease in the Kendall-τ value. However, the magnitude of this decrement differed from species to species. Consequently, at extremes of light intensity misidentifications may occur. In the example shown (Fig. 2), the specimen of Neotheronia mellosa (Cresson) has a greater correlation coefficient with the N. lineata L. species-classifier than does a specimen of N. lineata at very high and at very low illumination intensities. These results are likely to be of considerable importance when imaging an unknown is done on a different image-grabbing system from that used to construct the training set. Effect of number of training sets on accuracy of the system Any system intended to facilitate the routine identification of a group of insects needs to be expandable to accommodate large numbers of taxa. It is therefore of considerable practical interest to examine the behaviour of the system as the number of species-classifiers available to it increases. The behaviour of the DAISY prototype was examined using a set of 49 species of Ceratopogonidae, with each species-classifier trained on 11 images. Accuracy declined from 98% with only two species-classifiers, to 85% with 49 species-classifiers (Fig. 3). However, this decline appears not to be linear. An increase from 5 to 15 species-classifiers resulted in a 5%
I. D. Gauld, M. A. O’Neill and K. J. Gaston 308
Hymenoptera: Evolution, Biodiversity and Biological Control
1
0.98
0.96
0.94
0.92
0.9
0.88
0.86
0.84 0
10
20
30
40
50
Number of species classifiers Figure 3
DAISY misidentifies an increasing proportion of specimens as the number of species classifiers to which specimens belong increases. (• ) Kendall-τ metric, (■) Vector-Difference metric (results based on iterative analysis of data set comprising 49 species of Culicoides and Forcipomyia (Diptera: Ceratopogonidae); see Weeks et al. 1999).
decrease in correct identification, but a similar increase from 35 to 45 species-classifiers caused only a 2% drop. These data suggest accuracy may level off around 80% for large sets of ceratopogonid species. Although this does not sound impressive, it is difficult to find comparable data using traditional identification methods. Experts readily confuse females of at least four of the 49 ceratopogonid species, and use other (non-wing) characters to separate many others (J. Boorman pers. comm.). The success of non-experts using keys is more readily assessed. Twenty students in a class of second-year biology undergraduates at Sheffield University each attempted to identify ten
Driving Miss DAISY: the Performance of an Automated Insect Identification System 309
1
0.98
0.96
0.94
0.92
Figure 4 0.9
0.88
0.86
0.84 0
2
4
6
8
10
Minimum rank of the correlation coefficient of the ‘correct’ individual Figure 4
The performance of DAISY increases as “a correct identification” is progressively extended to include the 2, 3, 4… n highest ranking species-classifiers. (• ) Kendall-τ metric, (■) VectorDifference metric (results based on iterative analysis of data set comprising 49 species of Culicoides and Forcipomyia (Diptera: Ceratopogonidae); see Weeks et al. 1999).
specimens selected from 26 species of British Culicoides L. Using a specially simplified key they achieved accuracies ranging from 0 to 43% correct (mean = 13.1%). This compared very unfavourably with results of 87-89% correct obtained by non-specialists using DAISY.
Limitation of the Prototypical DAISY System Using an optimum number of specimens of 15 species of aphidiine braconids, DAISY was capable of achieving an overall accuracy of between 40 and 60%. This was perhaps not surprising as visual
I. D. Gauld, M. A. O’Neill and K. J. Gaston 310
Hymenoptera: Evolution, Biodiversity and Biological Control
identification techniques tend to perform poorly if the objects presented have few differentiating characters. However, it is noteworthy that this result is still much better than chance. Examination of misidentifications in this and various other data sets suggested that even when a wrong identification is made, the correct species-classifier is often ranked closely behind the winning species-classifier. For the ceratopogonid data set the performance of DAISY increased from 85% (highest ranking species-classifier only is correct) to 94% as the definition of ‘a correct identification’ is extended to include the correct classifier as being one of the top three – i.e. the identity of the unknown is narrowed down from one of 49 species to one of just three. As a simple FPTP system ignores much of the holistic information within the affinity vector A, alternative classifiers which use all of the information presented to the system were therefore explored.
DAISY Mark II: the Incorporation of a Second Order Meta-Classifier This leads to the development of a meta-classifier which compares affinity vectors as opposed to raw imagery. These classifiers can bootstrap themselves, using statistical clustering techniques in conjunction with arbitrary training data to build species classifiers in an emergent fashion. Details of the algorithms underlying meta-classification and a statistical evaluation of its performance will be given elsewhere (O’Neill et al. in prep.). However, tests suggest second order classifiers somewhat improve the performance of DAISY. For example, the classifier was able to distinguish Africanised from non-Africanised honey-bees 100% correctly, and Culex pipiens pipiens L. from C. pipiens molestus Forskal with >98% accuracy (Fig. 4) (although human experts cannot discriminate these forms on morphological features). The system also achieved high accuracies when classifying four sibling species of Colletes L. (C. hederae Schmidt & Westrich, C. succinctus L., C. perforator Smith and C. halophilus Verhoeff) and, with the aphidiine data set, accuracy was increased by approximately 25%, which is probably the best that is possible as the training sets hyperspacially intersect.
Conclusion Results obtained so far suggest that machine vision techniques are very applicable to problems of insect identification, and can deliver systems capable of accuracies approaching or exceeding that of traditional human based identification services. Whilst there are problems to be resolved, especially regarding handling large data sets and in dealing with 3-dimensional objects, machine vision technology may eventually yield rapid on-line identification services, at least for groups of organisms where there is a high user demand. For groups of visually very similar organisms, such as species of Aphidius Nees, totally reliable species identifications are unlikely to be achieved on wings alone, but the system does have potential to eliminate many species as ‘improbables’, and thus narrow the field of ‘possibles’ to one of a very few species. We envisage one of the main constraints in the future will be obtaining funding for populating both the classifier database, and the associated biological information database that actually presents the user with both an identification and some information about the organism.
Acknowledgements We thank Sondra Ward for providing technical support of the highest standard. Jon Boorman was kind enough to prepare a special key to ceratopogonids, and throughout acted as an invaluable source of information about these organisms. Wilf Powell (Rothampsted) donated reared
Driving Miss DAISY: the Performance of an Automated Insect Identification System 311
material of several Aphidiinae, John Ismay (Hope Entomological Collections, Oxford) provided source material of Culex pipiens. Africanised and non-Africanised honeybee wings and the Colletes material were supplied by Chris O’Toole (Bee Systematics and Biology Unit, Oxford). We are grateful to Jill Taber for volunteering her Christmas vacation to run a series of tests on image illumination.
References Alberch, P. (1993) Museums, collections and biodiversity inventories. Trends in Ecology & Evolution 8: 372-375. Alexander, I. (1984) WISARD – a radical step forward in image recognition. Sensor Review 17: 1507-1512. Bousouf, C., Gauld, I., Jordan, J., MacGillivray, A., Mackenzie, R., Shankleman, J., Sunman, H., Swanson, T., tenKate, K., Tomkins, R., Walden, I., Ward, H. & Yamin, F. (1996) Towards Implementation of Articles 15 and 16 of the Convention on Biological Diversity. ERM, London. Brown, K. S. (1997) Diversity, disturbance and sustainable use of Neotropical forests: insects as indicators for conservation monitoring. Journal of Insect Conservation 1: 25-42. Chalmers, N. R. (1996) Monitoring and inventorying biodiversity: collections, data and training. pp. 171-179. In di Castri, F. & Younès, T. (Eds), Biodiversity, Science and Development. Towards a New Partnership. CAB International, Wallingford. Daly, H. V., Hoelmer, K., Norman, P. & Allen, T. (1982) Computer-assisted measurement and identification of honey bees (Hymenoptera: Apidae). Annals of the Entomological Society of America 75: 591-594. di Castri, F, Robertson Vernhes, J. & Younès, T. (1992) Inventorying and Monitoring Biodiversity. Biology International, Special Issue 27: 1-28. Gaston, K. J. & May, R. M. (1992) Taxonomy of taxonomists. Nature 356: 281-282. Gauld, I. D. (1986) Taxonomy, its limitations and its role in understanding parasitoid biology. pp. 1-21 In Waage, J. & Greathead, D (Eds), Insect Parasitoids. Academic Press, London. Greenslade, P. J. N. (1985) Pterygote insects and the soil: their diversity, their effects on soils and the problems of species identification. Quaestiones Entomologicae 21: 571-585. Hawksworth, D. L. (1994)The Identification and Characterisation of Pest Organisms. CAB International, Wallingford. Holden, C. (1989) Entomologists wane as insects wax. Science 246: 754-756. House of Lords (1991) Systematic Biology Research. Report of the Select Committee on Science and Technology. HMSO, London. Janzen, D. H. (1993) Taxonomy: universal and essential infrastructure for development and management of tropical wildland biodiversity. pp 100-113. In Sandlund, O. T. & Schei, P. J. (Eds), Proceedings of the Norway/UNEP Expert Conference on Biodiversity, Trondheim, Norway. NINA, Trondheim. Lattin, J. D. & Knutson, L. (1982) Taxonomic information and services on arthropods of importance to human welfare in Central and South America. FAO Plant Protection Bulletin 30: 92-95. Lucas, S. M. (1997) Face recognition with the continuous n-tuple classifier. British Machine Vision Conference Proceedings 1: 222-231. May, R. M. (1990) Taxonomy as destiny. Nature 347: 129-130. Pankhurst, R. J. (1978) Biological Identification. Arnold, London.
I. D. Gauld, M. A. O’Neill and K. J. Gaston 312
Hymenoptera: Evolution, Biodiversity and Biological Control
Reid, W. V., Laird, S. A., Meyer, C. A., Gámez, R., Sittenfeld, A., Janzen, D. H., Gollin, M. A. & Juma, C. (1993) Biodiversity Prospecting: Using Genetic Resources for Sustainable Development. World Resources Institute, Washington, D. C. Tilling, S. M. (1987) Education and taxonomy: the role of the Field Studies Council and AIDGAP. pp 87-96. In Berry, R. J. & Crothers, J. H. (Eds), Nature, Natural History and Ecology. Academic Press, London. Turk, M. & Pentland, A. (1991) Eigenfaces for recognition. Journal of Cognative Neurosciences 3: 71-86. UNEP/CBD/SBSTTA/2 (1996) Report of the Subsidiary Body on Scientific, Technical and Technological Advice on the Work of its Second Meeting. Secretariat to the Convention on Biological Diversity, Montreal. Vane-Wright, R. I. (1996) Systematics and the conservation of biological diversity. Annals of the Missouri Botanical Garden 83: 47-57. Weeks, P. J. D. & Gaston, K. J. (1997) Image analysis, neural networks, and the taxonomic impediment to biodiversity studies. Biodiversity & Conservation 6: 263-274. Weeks, P. J. D., Gauld, I. D., Gaston, K. J. & O’Neill, M. A. (1997) Automating the identification of insects: a new solution to an old problem. Bulletin of Entomological Research 87: 203-211. Weeks, P. J. D., O’Neill, M. A., Gaston, K. J. & Gauld, I. D. (1999). Species-identification of wasps using principal component associative memories. Image and Vision Computing 17: 861-866. White, I. M. & Scott, P. R. (1994) Computer information resources for pest identification: a review. pp 129-137. In Hawksworth, D. L. (Ed.), The Identification and Characterisation of Pest Organisms. CAB International, Wallingford. Whitten, A. (1996) Field guides: useful tools in environmental planning and management. World Bank Environment Department, Dissemination Notes 51: 1-4. Yu, D. S., Kokko, E. G., Barron, J. R., Schaalje, G. B. & Gowen, B. E. (1992) Identification of ichneumonid wasps using image analysis of wings. Systematic. Entomology 17: 389-395.
Data Warehousing Architecture and Tools for Hymenoptera Biodiversity Informatics Norman F. Johnson and Luciana Musetti Department of Entomology, The Ohio State University, 1315 Kinnear Road, Columbus, Ohio 43212-1192 USA (email:
[email protected],
[email protected])
Introduction The natural history collections of the world are estimated to have some 2.5 billion specimens in their holdings (Duckworth et al. 1993). This material forms the physical documentation of much of our knowledge of biological diversity, both past and present. The intrinsic characteristics of the specimens, from morphological features to DNA sequences, are used to recognise taxa and to reconstruct their phylogenetic history. Equally important are the associated data which document the geographic and temporal distribution of taxa as well as ecological associations. Specimens in entomological collections may be preserved by one of a variety of methods: they may be individually pinned, one or more may be stored in vials, or they may be mounted on microscope slides. In each of these formats, the specimens are expected to have information relating to their collection written on a label affixed to the pin or slide, or included in the same vial. Minimally, this information includes the place at which the specimens were collected and the date or range of dates of collection. Other data commonly include the names of the collectors, the method of collection, brief notes on the habitat in which the specimens were found, and associations with other species. All of these data vary greatly in their detail and completeness. Entomological material also is somewhat restricted in that the information is usually constrained to appear on the small bits of paper attached to insect pins. The individual pieces of data associated with single specimens are of limited value themselves: the information may be anomalous or simply wrong. In the aggregate, however, these data usually are the only hard facts that we have in hand to support our ideas about the geographic distribution of a taxon, the time at which adults may be found on the wing, the habitats in which the species is likely to be found, or the hosts (plants or other arthropods) upon which the larvae feed. Voucher specimens in natural history collections allow us to reassess our understanding of the biology and distribution of taxa, even as taxonomic concepts develop and mature through time. The holdings of most collections, today, are accessible to bona fide researchers. Specimens are routinely sent on loan around the world, and some of the major museums annually host hundreds of visiting scientists. The physical accessibility of specimens is a major advance from the 19th century, when workers had to rely on incomplete or even inaccurate published descriptions as a means of recognising taxa. However, in a practical sense, this ‘accessibility’ is often a mirage. One must discover that the desired specimens actually exist before loan requests can be made or travel arranged. This discovery process is very time-consuming and expensive, involving communication with curators of the institutions known or suspected to have the material of interest and, often, personal travel to some of these museums. And at the end of the day, the
313
Norman F. Johnson and Luciana Musetti 314
Hymenoptera: Evolution, Biodiversity and Biological Control
material thus harvested is only a subset of the total actually available. In a practical sense, the remaining specimens, and their data, are inaccessible or non-existent. The discovery process described may well be sufficient to meet the needs of many taxonomic research projects. People know, at least generally, which institutions have large holdings in their group of interest, and resources of these collections are invariably tapped. The chances that a small collection will have a significant number of specimens of an important new taxon are rather small, and can easily be incorporated into the context of a revision at a later time. However, the specimens are more than just a source of taxonomic information. Small, regional collections are likely to be very important in filling in gaps in our knowledge of geographic distribution, seasonal phenology, and biological interactions. It is precisely these important data that we lose by focusing on only the major collections. Recent developments in information technologies offer researchers and curators the opportunity to provide real and affordable access to the specimen data that has been gathered over the years at great cost in time and funds. This ‘access’ means not only providing raw data, although this aspect is critically important, but also packaging this information in forms that facilitate interpretation and understanding. The emerging field of biodiversity informatics seeks to develop the tools for visualisation, summarisation, and generalisation of these data. Current progress in networking, databases, and geographical information systems is opening exciting new avenues for the application of specimen data beyond the fields of taxonomy and systematics. We describe here an information system that leverages World Wide Web technologies to provide real-time access to data on the taxonomy, distribution, phenology, and biology of taxa derived from both natural history collections and the published literature.
Needs A needs analysis is the first step in the development of new software applications. The target audiences must be defined: these may range from scientific researchers to governmental agencies, NGOs, and the general public. The perceived needs of these groups typically develop over time, especially as the new application is used and its potential better understood. From our perspective, at the highest level an information system should be widely accessible and platform independent, i.e. function on the widest range of hardware and operating systems. The information retrieved from the system should be as current as possible, preferably the latest information available. A wide range of visualisation and summarisation techniques should be provided to the user, but ultimately the actual data from specimens and publications must be available. Finally, the sources of all information must be documented, thus giving a relative measure of accuracy or reliability. More specifically, the kinds of tools that users will desire may be gleaned from the types of data summarisations that authors commonly use in traditional publications, especially distribution maps, graphs of seasonal flight periods, images, and keys. Typical avenues for access to data would be based upon the identity of a taxon, biogeographical region, country or author. Response time to information requests is important (the faster the better), and the speed of response to the most common query types must be optimised. We believe that World Wide Web technologies effectively address the issues of accessibility and platform independence. Relational databases are an established and proven tool for storing large amounts of information and maximising the efficiency of its retrieval. A well-designed
Data Warehousing Architecture and Tools for Hymenoptera Biodiversity Informatics 315
database system can provide for an almost infinite range of ad hoc queries, and results of common but complex queries can be extracted and stored separately to reduce response times to an absolute minimum. The common gateway interface scripts (cgi scripts) that effect the gateway between the Web and the database can also invoke auxiliary programs, such as mapping and plotting functions.
System description Hymenoptera On-Line is our implementation of such an information system. The primary target audience is the scientific community. Our goal is to provide this audience with access to the raw data as well as summarisations and visualisations of this information. The URL for access to the system (as of March 1999) is http: //iris.biosci.ohio-state.edu/hymenoptera. We describe below the basic elements and functionality of the system. More details may be obtained at http: //iris.biosci.ohio-state.edu/projects/tpp/clusters.html or from the authors. Database structure The underlying database structure is an implementation of the 1993 information model for biological collections developed by the Association of Systematics Collections (ASC 1993). We have made a number of extensions to that model, especially in the areas dealing with publications and unvouchered records. The data are stored in a total of 57 tables, which we group into seven functional clusters for the purposes of discussion. The basic idea is that the primary elements of a collection (Collecting Units) are derived from the act of collecting (Collecting Events), a unique combination of time, place, people and method. Collecting Unit Cluster. The original ASC model was devoted to natural history collections, and the Collecting Unit naturally formed the fundamental item of interest. Collecting Units may be individual specimens, groups of specimens (lots), or items derived from specimens (e.g. DNA extracts or genitalia). This idea has been somewhat expanded to incorporate more intangible constructs such as observations. Information about such units includes the number of specimens in a lot, the sex of a specimen, the life stage (e.g. egg, larva, pupa, adult), and its identification code. This code, the Collecting Unit ID, is a unique combination of collection coden (from Arnett et al. 1993) and a number. Some collections, such as Instituto Nacional de Biodiversidad and the Peabody Museum, have already provided unique identifiers for their specimens, and these codes are used when available. Collecting Event Cluster. A Collecting Unit is acquired as a result of a Collecting Event, that is, a unique combination of place, time, method and collectors. Other attributes could reasonably be used to further define an event, such as habitat or substrate. We continue to use separate tables to handle information about collecting dates, although a subsequent version of the ASC model has abandoned this in favour of a single text entry (ASC 1997). Collecting locality information leads to a small set of tables that distinguish among types of localities; store information on elevation, latitude and longitude; and place the locality within the political hierarchy of place names. Agent Cluster. The information about the entities that take actions, such as collectors or authors, is stored in the tables of this cluster. These include both people and organisations. We also have recently defined groups as separate entities, thus allowing an unlimited number of people to be associated as, e.g. authors or collectors.
Norman F. Johnson and Luciana Musetti 316
Hymenoptera: Evolution, Biodiversity and Biological Control
Taxonomy Cluster. The system of names that are applied to organisms is surprising in its complexity. These tables distinguish between taxonomic concepts, i.e. hypotheses about the existence of species, genera, etc., and the names that are applied to these concepts. We adopt the position that taxa at all levels are real individuals in nature and that these are the entities we are interested in modelling within the database. An alternative viewpoint is that the taxonomic concepts of each individual author are entities that should be modelled separately. This, in our view, is a legitimate position, but confusion arises because the same set of taxonomic names are applied to both kinds of concepts. Our database structure incorporates only the first. The taxonomic hierarchy is embodied by a recursive relationship within a single table. The ASC model attempted to include capacity to store alternative classifications, but we have not implemented this aspect. The tables of the taxonomy cluster and those of the collecting unit cluster are connected through a determination table, i.e. the assignment of a taxonomic name to a collecting unit. Documentation Cluster. The original ASC model had only the rudiments of the database structure needed for literature and associating this with specimens. The most recent version (ASC 1997) has considerably expanded in the area of literature, but we were forced to develop our own structure independently. The Documentation cluster of tables can treat both published and unpublished sources of information. The only form of the latter that we have found necessary to incorporate at this point is data recorded on specimen labels; other possibilities, though, include field notes and letters. Different forms of publication, e.g. journal articles versus books, traditionally are cited in different manners. Hence, we have a small proliferation of tables that store data particular to these different formats. Data from the literature is associated with collecting units, localities, and taxon names through specific citation tables. Authority Cluster. This is a small set of tables that serve as authority files for static information and includes data on collecting methods, categories of types (holotypes, lectotypes, etc), and the kinds of biological associations between taxa or individuals. Integration Cluster. Some classes of information are stored outside of the database itself, some in the file system of our servers, others on computers elsewhere on the Internet. These tables provide the means to retrieve this information. For example, sequence information is stored in GenBank databases. We have implemented the capability to automatically invoke the cgi scripts used at GenBank to retrieve information on hymenopteran data.
Web interface Our fundamental window into the information stored in the database is based upon the names of organisms. The user may enter a taxonomic name directly or step through the taxonomic hierarchy until reaching the taxon of interest. The various taxa have differing amounts and types of information associated with them: the categories of information available are indicated in abbreviated form on the list of names. These include 1) taxonomic data, 2) literature on taxonomy and systematics, 3) geographic distribution, 4) collecting dates, 5) images, 6) biological associations with other taxa, and 6) connections to other databases on the Internet. This interface is highly susceptible to changes as we incorporate new options and respond to user requests, so we will not dwell on the existing functionality here. Except for the initial entry Web page, all the html pages are produced dynamically using a combination of cgi scripts (written in Perl) and stored database procedures (written in PL/SQL, the procedural language extensions to the database standard structured query language). Distribution maps are created by extracting from the database the latitude and longitude for specimens of the selected taxon and painting circles on
Data Warehousing Architecture and Tools for Hymenoptera Biodiversity Informatics 317
pre-existing base maps at these points. Graphs of collecting dates are created and displayed using external Java applets. Connections to other databases are made by mimicking the native cgi scripts of their Web gateways.
Discussion Our database is designed to store and retrieve information from two primary and overlapping sources: the published literature and the specimens in collections. Although we have limited our universe of interest to a single order of arthropods, the magnitude of a complete database of information on specimens and literature of Hymenoptera would be overwhelming. We do not propose that time and money be invested in databasing every last specimen of honey bee in every collection in the world. Clearly there are priorities and points of diminishing returns. For some topics, the goal of comprehensiveness is both desirable and achievable. For example, a complete listing of all described species of Hymenoptera is a daunting, but attainable goal. Documenting geographic distribution, on the other hand, perhaps may be best achieved by recording only new county, state or country records. A number of problems are encountered in developing such a database. The most critical bottleneck in converting specimen information from the printed label to a digitally stored form is not the keyboard strokes involved in the transcription process itself, but the exercise of geo-referencing the collecting locality. Some countries have published or on-line gazetteers that greatly facilitate this process; for others no such information is available. Labels are often insufficiently specific to allow determination of latitude and longitude. This may be because only a large geographical area is cited as the locality (e.g. very old specimens may simply record that the specimen came from ‘Brazil’); a single place name may be used several times within a political entity (e.g. there are at least three places called Sugar Grove, Ohio); or the data may be ambiguous (e.g. 60 km W of Cairns: is this 60 km on a particular road, 60 km in a straight line, is it really due west or just generally toward the west). The importance of the third type of ambiguity is scale-dependent. Most of our questions are directed at continental-scale patterns of distribution, and the different interpretations have little practical effect. However, if one is interested in fine-scale patterns, such ambiguity is critical. Finally, political instability can cause difficulties, not so much in the changes of names or boundaries of countries, but in reorganisations of political subunits within countries. Data curation is a relatively expensive process in terms of both time and funds. Curation includes both the aspects of data input as well as quality control. These costs must be included in the overall budget of an institution or individual and will compete with the expenses of necessary activities such as physical curation of specimens, acquisition of new specimens, and mailing costs. A museum may well decide that its priorities cannot include databasing. We believe that it is imperative that alternative strategies be developed to provide access to the information contained in such collections, particularly those with geographically or taxonomically important holdings. Data warehousing and collective data capture by individual researchers are examples of such strategies. Our first gateway into the databased information relies upon the user selecting a taxon and then selecting the type of information desired. However, many other avenues of entry can be envisioned. The focus of interest may be geographical: What species are known from Costa Rica? What region has more species of braconids, the Neotropics or the Oriental region? Which taxa of Hymenoptera exhibit austral disjunct distributions? Chapada dos Guimarães is a classic
Norman F. Johnson and Luciana Musetti 318
Hymenoptera: Evolution, Biodiversity and Biological Control
collecting locality visited by H. H. Smith: what species are known from this site; what does it look like; what is its climate like? Or perhaps, the interest may be on the biological associates: What parasitoids attack the eggs of Carabidae? Science also has a social dimension: What species and genera were described by William H. Ashmead? How many of those are still considered valid? Who described the greatest number of synonyms and homonyms? All of these questions, and many more, can be addressed using our existing database structure. One of the values of recording information at the level of specimens is that it opens the possibility of using the data to model the geographic distribution of taxa. Maps with dots indicating known collecting localities and shaded areas indicating predicted distribution are a common element in the taxonomic literature. The basis upon which such shadings are made is almost never explicitly defined and, we believe, is simply arbitrarily determined by the author. It is possible, though, to use the geographic coordinates of collecting localities, and the number of specimens from each locality and, conceivably, the absence of specimens from a locality as inputs into modelling applications for which the parameters can be explicitly defined. Distribution predictions may make use of climatic variables such as precipitation and temperature, land cover, distributions of other organisms, human population density, and so on, to the satisfaction of the modeller. These have the advantage that the underlying data and assumptions are explicit and can be challenged or modified by others. With such distribution models in hand it is possible to imagine a number of interesting applications. The ultimate range of invasive species can be predicted. The search for potential biological control agents can be narrowed to match not only the targeted host, but also the climatic regime in which control is desired. Summaries of the distribution of entire groups can be used to precisely identify areas of high species richness or endemism. Observed biogeographical patterns may be quantitatively compared, i.e. is the distribution observed for one species unusual? How unusual is it? Which other species share this pattern? This is an exciting area in which to contemplate the application of data that too often lie unused in dusty collections or within unread monographs. Finally, we can see two other clear and relatively accessible applications for the database and Web gateway we have described. One is the development of a host-parasite index. In an electronic online format, newly acquired data can be easily added and existing data updated where necessary. Second, taxonomic names are, by definition, published and ‘available’. However, the literature is littered with homonyms and errors in spelling and citation of authorities. A complete nomenclature for the order is an attainable goal, one that would provide an up-to-date resource for the most basic piece of information about any taxon: its name. We do not, of course, believe that these ideas exhaust the range of applications of such databases. We stand ready to collaborate with the rest of our community in developing an exciting resource to advance our knowledge of the Hymenoptera.
Acknowledgements This material is based in part upon work supported by the National Science Foundation under Grant No. DEB-95221648.
References Arnett, R. H., Jr., Samuelson, G. A. & Nishida, G. M. (1993) The Insect and Spider Collections of the World. Second edition. Flora & Fauna Handbook No. 11. Sandhill Crane Press, Inc., Gainesville.
Data Warehousing Architecture and Tools for Hymenoptera Biodiversity Informatics 319
Association of Systematics Collections (1993) An information model for biological collections. Report of the Biological Collections Data Standards Workshop, August 18-24, 1992. URL: gopher: //kaw.keil.ukans.edu:70/11/standards/asc. Association of Systematics Collections (1997) The ASC Reference Model. URL: http:// gizmo.lbl.gov/DM_TOOLS/OPM/BCSL/ LIB/ASC.html. Duckworth, W. D., Genoways, H. H. & Rose, C. L. (1993) Preserving Natural Science Collections: Chronicle of Our Environmental Heritage. National Institute for the Conservation of Cultural Property, Inc., Washington, D. C.
Preliminary Study of Pteromalid Diversity in China: Taxonomic and Geographic Variation Hui Xiao1, Da-Wei Huang1 and Steven L. Heydon2 1
Institute of Zoology, Chinese Academy of Sciences, Beijing 100080 China (email:
[email protected]) 2
Bohart Museum of Entomology, Department of Entomology, University of California, Davis, CA 95616-8584 USA
Introduction The Pteromalidae is the most morphologically variable family within the Chalcidoidea, and this variability reflects the diverse life histories of the included taxa. Pteromalidae have been associated with approximately 100 families of plants and their hosts include members of most insect orders as well as spiders (Bouºek 1988; Bouºek & Rasplus 1991). They behave as ectoparasitoids, endoparasitoids, koinobionts, idiobionts, solitary or gregarious, primary or secondary parasitoids, while a few species can be classified as predators. Pteromalidae is one of the largest families of the Chalcidoidea, containing 588 genera and 3364 valid species (Noyes 1998). It was traditionally divided into 14–19 subfamilies, but Bouºek (1988) erected 12 new subfamilies, so current classifications include up to 31 subfamilies. The taxonomy of the Pteromalidae in China has made rapid progress since the 1970’s, with the number of recorded and described species having increased greatly in recent years. This work reviews the current knowledge of the Chinese pteromalid fauna by examining its diversity at generic and subfamily levels and assessing geographical variation within the group. The data come from material in the major international collections with substantial holdings from China: viz. the Zoological Museum, Institute of Zoology, Chinese Academy of Sciences, the Biocontrol Institute, Fujian Agriculture University, the Taiwan Agricultural Research Institute, the Bohart Museum of Entomology, University of California, Davis, the Natural History Museum, London, and the National Museum of Natural History, Smithsonian Institution. N.B. The authors for genera in the text are given in Appendix 1.
Subfamily Diversity and Proportional Representation of Subfamilies We recognise 17 subfamilies of Pteromalidae from China (including Taiwan and Macao) (Table 1), and have identified 155 genera, virually all of which are recorded from China for the first time (Appendix 1). Clearly this number will increase with further collecting and research. The Chinese fauna is dominated by the Pteromalinae (65% of genera) and, to a lesser degree, the Miscogasterinae (13%) (Miscogasterini sensu Graham 1969) (Table 1). The remaining 15 subfamilies comprise only 22% of genera (34), with nine subfamilies only being represented by a single genus. A comparison of these proportions with the world pteromalid fauna is given in Table 2. These data indicate that the Chinese fauna may be depauperate for the subfamilies Diparinae and Ormocerinae, and very diverse (at least at the generic level) for Miscogasterinae and Pteromalinae.
320
Hymenoptera: Evolution, Biodiversity and Biological Control
Preliminary Study of Pteromalid Diversity in China: Taxonomic and Geographic Variation 321
Table 1 Summary of the Chinese pteromalid fauna showing the number genera and their proportion of the total for each subfamily.
Subfamilies Austroterobiinae
No. genera included
Proportion %
1
0.65
No. genera included
Proportion %
Pireninae
2
1.29
Subfamilies
Asaphinae
1
0.65
Cerocephalinae
3
1.94
Colotrechninae
1
0.65
Diparinae
4
2.58
Herbertiinae
1
0.65
Ormocerinae
4
2.58
Macromesinae
1
0.65
Eunotinae
5
3.23
Neodiparinae
1
0.65
Cleonyminae
8
5.16
Panstenoninae
1
0.65
Miscogasterinae
20
12.90
Spalangiinae
1
0.65
Pteromalinae
100
64.52
Storeyinae
1
0.65
Table 2 Comparison of the Chinese and world pteromalid faunas (species’ numbers are given in parentheses; *includes unpublished data from the authors; ** data from Noyes 1998).
Subfamily
Chinese genera and species*
World genera and species**
Subfamily
Chinese genera and species*
World genera and species**
Austroterobiinae
1 (2?)
1 (2?)
Miscogasterinae
20 (101)
34 (319)
Asaphinae
1 (6)
4 (20)
Neodiparinae
1 (1)
1 (3)
Cerocephalinae
2 (3)
13 (41)
Ormocerinae
4 (20)
40 (165)
Cleonyminae
8 (20)
43 (261)
Panstenoninae
1 (5)
2 (10)
Colotrechninae
1 (1)
19 (41)
Pireninae
1 (4)
17 (184)
Diparinae
3 (7)
31 (102)
Pteromalinae
100 (160)
317 (1964)
Eunotinae
5 (12)
21 (77)
Spalangiinae
1 (14)
2 (51)
Herbertiinae
1 (1)
1 (7)
Storeyinae
1 (1)
1 (1)
Macromesinae
1 (1)
1 (11)
Distribution of Subfamilies The various pteromalid subfamilies show different patterns of distribution even when this is indicated only by a relatively crude measure such as ‘presence’ or ‘absence’ at particular localities. Asaphinae, Pteromalinae, Cleonyminae, Eunotinae, Panstenoninae, Ormocerinae, Pireninae and Spalangiinae are distributed throughout China. Other subfamilies show more restricted distributions. The Diparinae and Cerocephalinae are found mainly in central and southern China, but the genus Netomocera (Diparinae) can be found as far north as Beijing. In contrast, the Macromesinae is limited to the north of China and the Miscogasterinae (Miscogasterini sensu Graham 1969) to the north and west. However, a few genera in the latter subfamily, such as Glyphognathus, Lamprotatus, Miscogaster and Stictomischus, only occur further south. Several subfamilies may have very limited distributions but more intense sampling for these rarer subfamilies is needed. The Herbertiinae is recorded only from Taiwan, Hong Kong and Hainan Island; Neodiparinae is known only from Fujian Province in south-eastern China; Storeyinae
Hui Xiao, Da-Wei Huang and Steven L. Heydon 322
Table 3
Hymenoptera: Evolution, Biodiversity and Biological Control
Number of pteromalid genera recorded from the provinces of China.
Provinces
No. Genera
Provinces
No. Genera
Heilongjiang (HL)
26
Anhui (AH)
5
Inner Mongolia (NM)
24
Jiangsu (JS)
8
Liaoning (LN)
18
Sichuan (SC)
39
Jilin (JL)
33
Shanghai (SH)
6
Shaanxi (SN)
42
Zhejiang (ZJ)
12
Gansu (GS)
22
Fujian (FJ)
52
Qinghai (QH)
11
Jiangxi (JX)
3
Ningxia (NX)
19
Hunan (HN)
30
Xingiang Uygur (XJ)
29
Guangdong (GD)
21
Tibet (XZ)
23
Guangxi (GX)
20
Beijing (BJ)
49
Hong Kong (HK)
6
Hebei (HEB)
53
Taiwan (TW)
34
Shanxi (SX)
13
Macao (MC)
4
Shandong (SD)
22
Hainan (HAN)
17
Henan (HEN)
30
Guizhou (GZ)
7
Hubei (HB)
21
Yunnan (YN)
56
from Sichuan Province, and Austroterobiinae from Hong Kong. Generally, there appears to be an increase in subfamily diversity from north to south through the country.
Variation of Generic Diversity in Provinces Considerable variation is found in the numbers of pteromalid genera recorded from the various municipalities, provinces and autonomous regions (Table 3; Fig. 1). Tianjin and Chongqing municipalities have no recorded pteromalids so far, seven provinces have 10 or fewer recorded genera, seven provinces have 10–20 genera, 10 provinces 20–30, three provinces 30–40, two provinces 40–50, and three provinces have greater than 50 recorded genera. Of the latter three provinces, two are in south China and one is in the north. This fact provides only weak support for the idea that there is an increase in generic diversity from north to south. Attempting to find gradients in diversity by examining records for all the provinces together is also less than satisfactory. The seven provinces with less than 10 recorded genera are all located in central-southern China, but this probably does not indicate that pteromalids are depauperate there, but rather that the region has been less intensively sampled. These provinces should be the primary focus of future surveys.
Interesting Distribution Patterns of Some Special Genera Many pteromalid genera from Australia, the Pacific region and Africa are also found in China, particularly in the south. For example, Parurios (Diparinae) was previously recorded from Australia with one additional species in India. We have recently found Parurios in Hainan, Yunnan, Fujian and Hunan, and as far north as latitude 34˚N. Grahamisia (Diparinae) was previously recorded from Africa and Sri Lanka, but is now also recorded from Hubei Province, central China. Agiommatus (Pteromalinae), known from Australia, Indonesia, Malaysia, Sri Lanka, India and Madagascar, has recently been recovered from numerous localities in southern China (Fujian, Guangxi and Guangdong Provinces) and from as far north as Hunan Province in central
Preliminary Study of Pteromalid Diversity in China: Taxonomic and Geographic Variation 323
Figure 1
Distribution and diversity of pteromalid genera in the provinces of China (see Table 3 for abbreviations).
China. Zolotarewskya (Cleonyminae) is recorded from Australia (Queensland), Madagascar, Singapore and Algeria, but we have also found it in northern China (Beijing). One genus, probably the undescribed taxon from Australia and referred to the Austroterobiinae by Bouºek (1988) is now known to occur in Hong Kong. The pteromalid fauna of northern China is similar to that of temperate Palaearctic Europe, but the ranges of some Palaearctic genera are found to extend into southern China. For example, Neodipara (Neodiparinae) was previously known from Europe, but we have recorded it from Fujian Province in south-eastern China. Lyubana (Pteromalinae) was described from Yugoslavia (Bouºek & Rasplus 1991) but is now known form Fujian Province and appears to be widely distributed throughout south-east Asia. In the Miscogasterinae, generic diversity appears to increase from the north-eastern towards the south-west of mainland China.
Discussion Based on the above findings, the following preliminary comments and conclusions can be made about the Pteromalidae of China: 1.
Genera characteristic of the Palaearctic region can usually be found in northern China; genera characteristic of the Oriental region predominate in southern China.
Hui Xiao, Da-Wei Huang and Steven L. Heydon 324
Hymenoptera: Evolution, Biodiversity and Biological Control
2.
Genera mainly distributed in Australia, can be found in southern China, but also in northern and central parts. This needs to be further studied.
3.
Based on the distribution of genera documented here, it is reasonable to separate the Palaearctic from the Oriental Regions in central China along the Qinling Mountains.
4.
Whilst diversity increases towards the tropics, it is not uniform across all subfamilies, with some showing a reverse trend.
5.
The seven provinces that have less than 10 pteromalid genera recorded are all located in central-southern China. This may not reflect the real situation but rather be an artefact of collecting, which should be addressed by concentrated collecting efforts in future surveys.
Although the above findings are based on an extensive amount of material, the data are still far from complete. Most regions of China have been relatively poorly sampled for Chalcidoidea, given the size of the country and diversity of habitats. Our work to date indicates that further collecting effort could be profitably focused in the Qinghai-Xizang Plateau and the downstream area of the Yangtze River. This effort should employ a wide array of collecting techniques to maximise the chalcidoid diversity in samples.
Acknowledgements The project supported by National Natural Science Foundation of China, (NSFC grant No. 39625004). We would like to thank the Chinese Academy of Sciences, which provided Da-Wei Huang the opportunity of studying at the Bohart Museum of Entomology, Department of Entomology, University of California. Space and facilities during this time were kindly provided to Da-Wei Huang by the Bohart Museum of Entomology.
References Boucˇek, Z. (1988) Australasian Chalcidoidea (Hymenoptera): A Biosystematic Revision of Genera of Fourteen Families, with a Reclassification of Species. CAB. International, Wallingford. Boucˇek, Z. & Rasplus, J.-Y. (1991) Illustrated Key to West-Palearctic Genera of Pteromalidae (Hymenoptera: Chalcidoidea). Institut National de la Recherche Agronomique, Paris. Graham, M. W. R. de V. (1969) The Pteromalidae of North-Western Europe (Hymenoptera: Chalcidoidea). Bulletin of the British Museum Natural History (Entomology) Supplement 16: 1-908. Noyes, J. S. (1998) Catalogue of the Chalcidoidea of the World (CD-product, windows version 1.0). ETI/The Natural History Museum, London.
Preliminary Study of Pteromalid Diversity in China: Taxonomic and Geographic Variation 325
Appendix 1 Pteromalid subfamilies and genera recorded from China. Subfamily
Genus
Subfamily
Genus
Subfamily
Genus
Asaphinae (1)
Asaphes Walker
Pteromalinae (101)
Ablaxia Delucchi
Pteromalinae (101)
Miscogasteriella Girault
Cerocephalinae (3) Acerocephala Gahan Cerocephala Westwood Cleonyminae (8)
Acroclisoides Girault
Mokrzeckia Mokrzecki
Acrocormus Förster
Nasonia Ashmead
Theocolax Westwood
Agiommatus Crawford
Norbanus Walker
Anacallocleonymus Yang
Allocricellius Yang
Notoglyptus Masi
Callocleonymus Masi
Amblyharma Huang & Tong
Oxysychus Delucchi
Cleonymus Latreille
Anisopteromalus Ruschka
Pachycrepoideus Ashmead Pachyneuron Walker
Heydenia Förster
Apsilocera Boucˇek
Notanisus Walker
Arthrolytus Thomson
Paracarotomus Ashmead
Oodera Westwood
Callimerismus Graham
Paracroclisis Girault
Zolotarewskya Risbec
Callitula Spinola
Paroxyharma Huang & Tong
Solenura Westwood
Capellia Delucchi
Peridesmia Förster
Colotrechninae (1)
Colotrechnus Thomson
Catolaccus Thomson
Pezilepsis Delucchi
Diparinae (4)
Dipara Walker
Cecidostiba Thomson
Platecrizotes Ferrière
Grahamisia Delucchi
Cheiropachus Westwood
Platneptis Boucˇek
Netomocera Boucˇek
Chlorocytus Graham
Platygerrhus Thomson
Parurios Girault
Coelopisthia Thosmon
Plutothrix Förster
Cephaleta Motschulsky
Conomorium Masi
Propicroscytus Szelényi
Eunotus Walker
Coruna Walker
Pseudocatolaccus Masi
Moranila Cameron
Cryptoprymna Förster
Psilocera Walker
Eunotinae (5)
Hui Xiao, Da-Wei Huang and Steven L. Heydon 326
Subfamily
Hymenoptera: Evolution, Biodiversity and Biological Control
Genus
Subfamily
Genus
Subfamily
Genus
Ophelosia Riley
Cyclogastrella Bukovskii
Scutellista Motschulsky
Cyrtogaster Walker
Pteromalus Swederus
Herbertiinae (1)
Herbertia Howard
Cyrtoptyx Delucchi
Pterosemigastra Girault & Dodd
Mactomesinae (1)
Macromesus Walker
Dibrachys Förster
Pterosemopsis Girault
Miscogasterinae (19)
Glyphognathus Graham
Dinarmus Thomson
Pycnetron Gahan
Neodiparinae (1)
Psychophagus Mayr
Ammeia Delucchi
Dinotiscus Ghesquicre
Rakosina Boucˇek
Drailea Huang
Dinotoides Boucˇek
Rhaphitelus Walker
Halticoptera Spinola
Eulonchetron Graham
Rhopalicus Förster
Halticopterina Erdös
Eumacepolus Graham
Roptrocerus Ratzeburg
Lamprotatus Westwood
Euneura Walker
Schizonotus Ratzeburg
Merismus Walker
Eurydinota Förster
Sphegigaster Spinola
Miscogaster Walker
Eurydinotomorpha Girault
Stenomalina Ghesquicre
Nodisoplata Graham
Gastracanthus Westwood
Synedrus Graham
Rhicnocoelia Graham
Glyptosticha Masi
Syntomopus Walker
Seladerma Walker
Gugolzia Delucchi & Steffan
Thinodytes Graham
Skeloceras Delucchi
Habritys Thomson
Tomicobia Ashmead
Sphaeripalpus Förster
Hobbya Delucchi(?)
Toxeumorpha Girault
Stictomischus Thomson
Holcaeus Thomson
Trichomalopsis Crawford
Thektogaster Delucchi
Homoporus Thomson
Trichomalus Thomson
Toxeuma Walker
Inkaka Girault
Trigonoderoides Kamijo
Tricyclomischus Graham
Ischyroptyx Delucchi
Trigonoderus Westwood
Tumor Huang
Isocyrtus Walker
Tritneptis Girault
Xestomnaster Delucchi
Kaleva Graham
Trychnosoma Graham
Neodipara Erdös
Lariophagus Crawford
Tsela Boucˇek
Preliminary Study of Pteromalid Diversity in China: Taxonomic and Geographic Variation 327
Subfamily
Genus
Ormocerinae (4)
Semiotellus Westwood
Subfamily
Genus
Subfamily
Genus
Lyubana Boucˇek
Uniclypea Boucˇek
Systasis Walker
Makaronesa Graham
Vrestovia Boucˇek
Ormocerus Walker
Manineura Boucˇek
Zdenekiana Huggert
Oxyglypta Förster
Meraporus Walker
Panstenoninae (1)
Panstenon Walker
Merismomorpha Girault
Pireninae (2)
Gastrancistrus Westwood
Merisus Walker
Macroglenes Westwood
Mesopolobus Westwood
Spalangiinae (1)
Spalangia Latreille
Metacolus Förster
Storeyinae
Storeya Boucek
Metastenus Walker
Austroterobiinae
Genus ?
Micradelus Walker
The Family Braconidae in China (Hymenoptera) Chen Xuexin, He Junhua & Ma Yun Institute of Applied Entomology, Zhejiang University, Huajiachi Campus, Hangzhou 310029 China (email:
[email protected])
Introduction The braconid wasp fauna of China is a rich and diverse one. This is consistent with the diversity of habitats found in such a large area of the eastern Palaearctic and northern Oriental region, spanning some 25 degrees of latitude and, related to this, that the country represents one of several ‘megadiverse’ regions of the world. This is reflected in the insect fauna China in a general way, and also in its parasitic Hymenoptera, including the Braconidae. However, compared with some other regions of the world, the comprehensive study of Chinese braconids was only begun relatively recently, but has accelerated at an increasing pace. In the last 20 years the number of described species has increased 2.5 times and, including the fauna of Taiwan, represents about 1100 recognised taxa. The aim of this paper is to outline the history of taxonomic studies undertaken on the braconids of China, to document the people, institutions and taxonomic groups involved, and to provide a detailed listing of the literature as an aid to braconid and other systematists from outside the country.
Brief History of Braconids in China About 280 species belonging to 80 genera and 18 subfamilies of the family Braconidae (Hymenoptera) were listed for China in the multi-volume monograph ‘Hymenopterorum Catalogus’ (Shenefelt 1969-1978; Shenefelt & Marsh 1976). This compendium clearly shows that the majority of studies on the Braconidae of China prior to the 1970’s were undertaken mainly by foreign experts, and that those by Chinese taxonomists themselves mostly started later than this time. However, there are some notable exceptions. Professors Chu Joo-Tso, Li Feng-Swen and Chin Shin-Foon [=Zhao Shan-Huan] were among the earlier pioneers of braconid studies for the region, particularly on the natural enemies (parasitoids) of agricultural pests and their application for biological control. During the period 1933–1937, Chu published a series of papers on the parasitoids of agricultural and forestry pests, such as Parnara guttata (Bremer & Grey), Pieris rapae L., Dendrolimus punctatus Walker, and the insect pests of mulberry (e.g. Chu 1934, 1937). Li (1935) reported on the parasitoids of cotton pests, while Chin (1937) focused on the parasitoids of one of the insect pests of rice, Scirpophaga incertulas (Walker). A number of braconid species are mentioned in each of these studies. However, the earliest works on the taxonomy of Chinese braconids are Chu’s papers ‘Preliminary notes on the ichneumon-flies in Kiangsu and Chekiang Provinces, China’ (Chu 1935) and ‘Notes on Cheloninae of China...’ (Chu 1936). During the 1940’s to 1960’s, little work was undertaken on the Braconidae of China. However, two authors who published during this time deserve mentioned. One is Professor Hsich Sun-Yun
328
Hymenoptera: Evolution, Biodiversity and Biological Control
The Family Braconidae in China (Hymenoptera) 329
[=Xia Songyun], who continued his preliminary work on parasitic Hymenoptera attacking important pests of rice in Hunan Province, reporting many species of braconids, especially of the genus Apanteles Foerster s.l. (Hsich 1957). The other is Professor Chao Hsiu-Fu [=Zhao Xiufu], who worked on the genera Spathius Nees and Platyspathius Viereck (Doryctinae) (Chao 1957; Chao & Chen 1965) and continued these studies into the 1970’s (Chao 1977, 1978). It was during the 1970’s that research on biological control, and associated studies on braconids and other groups of parasitic Hymenoptera, began in earnest. In the early part of this decade, Chu and He published a series of four papers dealing with the identification of parasitoids of rice pests, two of them dealing with the common species of Braconidae (Chu & He 1973; Chu et al. 1976). In 1978, the ‘Atlas of Natural Enemies of Economic Insects’ was compiled by the Institute of Zoology, Academia Sinica, Zhejiang Agricultural University, etc. This was the first publication that enabled workers to identify common natural enemies in China, and it included 40 species of braconids. Soon after the book ‘Atlas of Natural Enemies of Rice Pests in Zhejiang Province, China’ edited by He (1979) was published. During this time and into the early 1980’s several other provinces put out similar publications dealing with natural enemies. Concurrently, a large number of individual papers on the biology of natural enemies appeared. In 1986 a general book ‘Atlas of Natural Enemies of Rice Pests in China’ edited by He and Pang (1986) appeared. As a result of these publications substantial information on the common species of natural enemies, both parasitoids and predators, was documented that greatly encouraged systematic studies, including those of braconids. In 1973, a meeting sponsored by the Academia Sinica was held to plan and coordinate studies on the biology and taxonomy of the fauna and flora of China. In 1986, this led to the project ‘Economic Insect Fauna of China: Hymenoptera-Braconidae’, headed by Professor He Junhua and sponsored by the Academia Sinica. In 1992 this project became a sub-project called ‘Fauna of China: InsectaHymenoptera-Braconidae (I)’ of the key project ‘Fauna of China’ sponsored by China National Science Foundation during the period of the 8th Five-Year-Plan. Since the resumption of recruiting graduate students in 1978 to undertake advanced studies on biological control and insect systematics, several students were enrolled for Masters and PhD’s to work on the systematics of parasitic Hymenoptera. Subsequently, the first Master and PhD degrees on the taxonomy of the Braconidae were conferred in 1987 and 1994, respectively at Zhejiang University. Coincident with an increased focus on the taxonomy and biology of parasitic Hymenoptera, was the growth and development of collections. The Zoological Institute of the Academia Sinica, Fujian Agricultural University, the Taiwan Agricultural Research Institute, and the Zhejiang University are the four leading institutions which have built up collections of Chinese parasitic Hymenoptera. The total number of pinned specimens in these four collections probably exceeds two million. The Shanghai Institute of Entomology, the Hunan Agricultural University in Changsa, the Northwestern Agricultural University in Xi’an, the Zhongshan University in Guangzhou, and the Henan Agricultural University in Zhengzhou also have moderately large collection of parasitic Hymenoptera.
Institutions and Individuals Currently, the following persons and institutions are involved in the study of the Braconidae in China: He Junhua and Chen Xuexin at the College of Agriculture and Biological Technology, Zhejiang University (Huajiachi Campus) in Hangzhou (previously called the Zhejiang Agricultural University from 1952 to 1998); Chao Hsiu-Fu, Chen Jiahua and Wu Zhishan at
Chen Xuexin, He Junhua & Ma Yun 330
Hymenoptera: Evolution, Biodiversity and Biological Control
the Fujian Agricultural University in Fuzhou; and You Lanshao at the Hunan Agricultural University in Changsa. The Parasitic Hymenoptera Research Center at Zhejiang University started as the Entomological Bureau of Zhejiang Province where Professor Chu Too-tsu begun studies on the Braconidae in the 1930’s. It now holds a collection of about 250 000 pinned specimens of parasitic Hymenoptera from all parts of China, plus as many specimens again in alcohol, and more than 600 identified species from other regions of the world. Chu was the founder of this center and served as a professor until 1981 when he passed away. He Junhua is the second professor at this institution and is still actively involved in the study of parasitic Hymenoptera, including Braconidae after he completed the monograph ‘Economic Insect Fauna of China: Hymenoptera-Ichneumonidae’ in 1993 (He 1996). Chen Xuexin is currently in charge of the sub-project ‘Fauna of China: InsectHymenoptera-Braconidae (II)’ of the project ‘Fauna of China’ which commenced in early 1998 during the 9th Five-Year-Plan.
Present Status of Each Group During the period 1987 to 1992 Chen and He published a series of papers on the genus Aleiodes Wesmael (Rogadinae). From 1992 onwards they have conducted work on rogadines supported by the National Science Foundation of China, the results of which were published in a monograph of the group which treats 19 genera and 116 species (Chen & He 1997). Quicke et al. (1997) reviewed the species of Yelicones Cameron for the East Palaearctic and Oriental regions, in which all the Chinese species were included. Two genera, Pentatermus Hedquist and Aulosaphes Muesebeck (including Aulosaphoides van Achterberg) (Lysiterminae) were reviewed by He and Chen (1995) and Chen et al. (1996b), respectively, while Chen et al. (1995) revised Acanthormius Ashmead. The large genus Macrocentrus Curtis (Macrocentrinae) has been studied by numerous Chinese authors prior to 1997 (You & Luo 1988; He & Lou 1993; He et al. 1996; Chen et al. 1997b). He and van Achterberg (1994) reviewed the Chinese species Aulacocentrum Ashmead and He et al. (1996) reported on the Chinese species of Rectizele van Achterberg. The small subfamily Sigalphinae is represented in China by two widely distributed genera, Acampsis Wesmael and Sigalphus Latreille. He and Chen (1992) studied the former genus while He and Chen (1993), He et al. (1994) and You et al. (1991) revised the latter. Only the nominal tribe of the Homolobinae, represented by the near cosmopolitan genus Homolobus Foerster, has been treated for China (Chen 1991; Chen et al. 1991; Chou & Hsu 1995). Chou and Hsu (1995) and Chen et al. (1996a) studied the small genus Charmon Haliday (Charmontine) and described four species from China. Two species of Distilirella van Achterberg (Xiphozelinae) were reported by He (1985) and Chou and Hsu (1995), and one species of Xiphozele Cameron by You et al. (1990). He et al. (1997a) described the new genus Sinoneoneurus from China which belongs to the somewhat aberrant subfamily Neoneurinae. Papp and Chou (1996) and Chen et al. (1997a) studied the Miracinae, reporting two new species of Centistidea Rohwer and Mirax Haliday, respectively. Of the three tribes of Ichneutinae, only the Ichneutini have been studied for China. He et al. (1997b) described three new species of Ichneutes Nees and one species of Pseudichneutes Belokobylskij. Chou and Lee (1991) and He and Chen (1996) studied two genera of Cenocoeliinae, Cenocoelius Haliday and Rattana van Achterberg, respectively.
The Family Braconidae in China (Hymenoptera) 331
The results of the above studies have been compiled into a single monographic treatment entitled ‘Fauna of China: Insecta, Hymenoptera, Braconidae (I)’, the publication of which is imminent (He et al. 2000). This work represents the first comprehensive book dealing with the systematics of Chinese braconids. It provides a general account of the morphology, biology, history of study, phylogeny and biogeography of Chinese braconids, and a key to subfamilies. The taxonomic part covers 13 subfamilies (i.e. Rogadinae, Lysiterminae, Miracinae, Neoneurinae, Ichneutinae, Adeliinae, Charmontinae, Macrocentrinae, Cenocoeliinae, Meteorideinae, Sigalphinae, Homolobinae and Xiphozelinae) and treats 49 genera (one of which is described as new) and over 296 species (77 of which are new). This work presents a major part of the sub-project of the ‘Fauna of China’ (see above). Among the remaining subfamilies not covered by He et al. (2000), little comprehensive taxonomic research has been conducted for China, with the exception of the Euphorinae, Helconinae (Bruleiini) and Alysiinae (Alysiini). There were 10 genera and 69 species of Euphorinae known from China prior to 1997. Among these, Aridelus Marshall and Streblocera Westwood have been relatively well-studied (Chao 1964, 1993; Chen & van Achterberg 1997), while Chou (1986) treated a species of Chrysopophthorus Goidanich, Papp and Chou (1995) three species of Wesmaelia Foerster from Taiwan, and Yang (1996) two species of Cosmophorus Ratzeburg from mainland China. In 1997, Chen and van Achterberg published a revision of the Euphorinae for China, which included 24 genera and 150 species. However, the Meteorini, an important group including parasitoids of many agricultural lepidopteran pests, was not included in this work and the group remains largely unworked, except for Zele Curtis which has been revised for mainland China (Chen et al. 1987) and Taiwan (Chou & Chou 1993b). The genus Meteorus Haliday has yet to be revised, however He (1982) provides notes on four common species while Wang (1984b) has described a new species. Both the Opiinae and Alysiinae, important parasitoids of pest Diptera are poorly known for China. However, the Alysiini was recently studied by Chen and Wu (1994), who cover 101 species in 19 genera, while Wharton and Chou (1983, 1985) have revised Heratemis Walker and Alloea Haliday from Taiwan. You Lanshao and colleagues have studied the microgastrine genus Apanteles s.l. (i.e. sensu Nixon 1965), and have described a number of new species and recorded others from China for the first time (e.g. You & Xiong 1983; You et al. 1985, 1987, 1994). You et al. (1988) published a list of 60 Apanteles species recorded from China and reviewed previous work on the genus. Unfortunately, they failed to include many species, even some previously described by themselves, and so the work is of only limited use. Chen et al. (1994) revised Fornicia Brullé, describing five new taxa, and Chou (1985b) described a new species of Buluka De Saeger. However, the Microgasterinae of China remains poorly studied and is in urgent need of taxonomic attention. The Helconinae of mainland China has only been studied for the subtribe Bruleiina (Bruelliini) (Chen et al. 1993), but recently Chou and Hsu (1998) have dealt with the Taiwanese species. Since Chu’s early work on Chinese Cheloninae in the 1930’s, little has been done since then on this subfamily. Tang and Marsh (1994) reviewed the species of Ascogaster Wesmael, while He et al. (1994, 1997c) described a new genus, Siniphanerotomella, and a new subgenus of Chelonus Panzer. Likewise, the Doryctinae of China are poorly known, with work on Spathius (and Platyspathius) by Chao Hsiu-Fu (1957, 1977) and on Zombrus Marshall by He and Ma (1982) being the only published studies on this subfamily. Chou (1995) has covered the species of
Chen Xuexin, He Junhua & Ma Yun 332
Hymenoptera: Evolution, Biodiversity and Biological Control
Cardiochilinae from Taiwan, and Chen et al. (1998) have reviewed the genus Hartemita Cameron from China, but again this group requires study. Two large subfamilies that remain poorly are the Agathidinae and Braconinae. Wang (1984a) described a species each from Gyrochus Enderlein and Isoptronotum Enderlein (Agathidinae), and Yang (1996) has described several species of Coeloides Wesmael (Braconinae). Liang-Yih Chou has been undertaking a systematic treatment of the Taiwanese fauna since the late 1970’s, and because of his work and the much smaller size of the region, the Braconidae of Taiwan is better known than that of mainland China. To present he has published a series of eight papers covering 11 subfamilies (Chou 1981-1995; Chou & Sharkey 1989; Chou & Lee 1991; Chou & Chou 1991, 1993a, 1993b; Chou & Hsu 1995, 1996, 1998; Papp & Chou 1996). In total, 440 species are known from Taiwan as follows: Agathidinae (38), Alysiinae (14), Aphidiinae (36), Betylobraconinae (1), Blacinae (10), Braconinae (28), Cardiochilinae (5), Cenoceliinae (4), Charmontinae (1), Cheloninae (27), Doryctinae (44), Euphorinae (51), Exothecinae (11), Helconinae (42), Histeromerinae (1), Homolobinae (11), Hormiinae (7), Lysiterminae (5), Macrocentrinae (4), Microgastrinae (45), Miracinae (3), Neoneurinae (1), Opiinae (11), Orgilinae (18), Pambolinae (2), Rogadinae (18) and Xiphozelinae (2) (Chou pers. comm.).
Studies on Biology, Ecology and Biological Control As discussed above, a nationwide survey of natural enemies of insect pests of major crops in China was initiated in 1978. Rice being the most important crop in China has been extensively surveyed for natural enemies and this has resulted in numerous publications. The following are pests from which braconid parasitoids have been investigated: rice stem borers [Chilo suppressalis Walker, Scrirpophage incertulas (Walker), Sesamia inferens (Walker)), rice leaf folder (Cnaphalocrocis medinalis Guenée), green rice caterpillar (Naranga aenescens Moore), rice skippers (Parnara spp.), Mycalesis gotama Moore, and rice leaf beetle (Oulema aryzae (Kuwayama)] (e.g. Hsich 1957; He 1979; He & Pang 1986). The natural enemies of major insect pests of cotton [such as Pectinophora gossypiella (Saunders) (e.g. Li 1935)], forests (such as pine caterpillars, Dendrolimus spp. and Lymantira dispar L.) (e.g. Yang 1996), citrus, apples, tea trees, and vegetables have also been moderately well-studied. Numerous studies on the biology, ecology and application of braconids in biological control have also been undertaken in China. For example, Ontsira palliatus (Cameron) has been widely studied and used extensively to control a range of hosts, mainly species of Cerambycidae (Coleoptera), including Semanelus sinoauster (Gressitt), Stromaltum lingicorna (Neumman), Monochamus alternatus Hope, Callidium villosulum Fairmaire, Xystrocera globosa (Olivier), Saperda populnea (L.) and Anoplophora chinensis Foerster, all of them being major pests of fruit trees and timber in China. Innundative release of O. palliatus has been used for many years and inoculative release has been conducted in several southern provinces. In the last decade, the cotton bollworm, Helicoverpa armigera (Hübner) has became the major pest in the cotton-cultivated area of northern China. This insect is attacked by a number of parasitoids, including five species of Microplitis, viz. M. mediator (Haliday), M. tuberculifer (Wesmael), M. erythrogaster Abdinbekova, and two undescribed species. Among them, one of the new species is dominant in attacking young larvae, and several studies have been conducted on its biology and ecology, and potential as a control agent. Leucania separata Walker is used as a host in the laboratory for this parasitoid after a set of techniques for its mass rearing were
The Family Braconidae in China (Hymenoptera) 333
developed, and the inoculative release of this braconid has been carried out in some provinces for several years. The following are some more important braconids that have been well investigated, some of which have been used as biological control agents. They are Cotesia glomerata (L.) and C. rubecula (Marshall) on cabbage worm (P. rapae); Cotesia plutellae (Kurdjumov) on Plutella xylostella L.; Macrocentrus cingulum Bricchke (misidentified as M. linearis (Nees)) on Onstrinia furnacalis Guenée; Bracon greeni Ashmead on Eublemma amabilis Moore; Amyosoma chinensis (Szépligeti), Bracon onukii Watanabe and Chelonus munakatae Matusmura on rice stem borers; Apanteles cypris Nixon on Cnaphalocrocis medinalis Guenée; Aleiodes narangae (Rohwer) on Naranga aenescens Moore; Bracon isomera (Cushman), B. nigrorufum (Cushman) on P. gossypiella; Habrobracon hebetor Say on Sitotroga cereatella Olivier (as well as Cadra cautella Walker and Plodia interpunctella Hübner); Aleiodes esenbeckii (Hartig) on Dendrolimus spp.; Meteorus rubens Nees on Agrotis spp.; Cotesia kariyai (Watanabe) on L. separata, and Cotesia ruficrus (Haliday) on a number of lepidopteran pests.
Acknowledgements We wish to thank Dr Kees van Achterberg, National Museum of Natural History, Leiden for his critical reading of the manuscript . The first author would also like to thank Dr Andy Austin, University of Adelaide for his help with arrangements to visit Canberra in January 1999 which was supported by a grant from China National Science Foundation (Grant No. 39810212021-Co2).
References Chao, H. F. (1957) On south-eastern Chinese braconid-flies of the subfamily Spathinae (Braconidae). Journal of Fujian Agricultural College 4: 1-18. [in Chinese with English summary] Chao, H. F. (1964) Descriptions of four new species of braconid wasps of the genus Streblocera Westwood (Hymenoptera). Acta Zootaxonomia Sinica 1: 153-162. [in Chinese with English summary] Chao, H. F. (1977) A study on Chinese braconid wasps of the tribe Spathini (Hymenoptera: Braconidae: Doryctinae). Acta Entomologia Sinica 20: 205-216. [in Chinese with English summary] Chao, H. F. (1978) A study on Chinese braconid wasps of the tribe Spathini (Hymenoptera: Braconidae: Doryctinae) [cont.]. Acta Entomologia Sinica 21: 173-184. [in Chinese with English summary] Chao, H. F. (1993) Two new species and synonyms of known species of Streblocera Westwood from Fujian province, China (Hymenoptera: Braconidae: Euphorinae). Wuyi Science Journal 10: 61-69. Chao, H. F. & Chen, J. H. (1965) A new species of the genus Spathius Nees (Hymenoptera: Braconidae) from Fujian province, China. Acta Zootaxonomia Sinica 2: 106-108. [in Chinese with English summary] Chen, X. X. (1991) A new species of the genus Homolobus Foerster from China, with a checklist of the Chinese species (Hym.: Braconidae). Entomologische Berichten 51: 47-49. Chen, X. X. & He, J. H. (1997) Revision of the subfamily Rogadinae (Hymenoptera: Braconidae) from China. Zoologische Verhandelingen 308: 1-187.
Chen Xuexin, He Junhua & Ma Yun 334
Hymenoptera: Evolution, Biodiversity and Biological Control
Chen, X. X., He, J. H. & Ma, Y. (1987) Five new records of species of Zele Curtis (Hymenoptera: Braconidae: Meteorinae) from China. Wuyi Science Journal 7[1990]: 94-96. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1991) The genus Homolobus Foerster (Hymenoptera: Braconidae) of China. Acta Agriculturae Universitatis Zhejiangensis 17: 192-196. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1994) Five new species of the genus Fornicia Brullé (Hymenoptera: Braconidae) from China. Entomotaxonomia 16: 127-134. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1995) Hymenoptera: Braconidae. pp .256-263. In: Insects and Mushrooms of Mt. Gutian Nature Reserve, Zhejiang, China. Zhejiang Science and Technology Publishing House, Hangzhou. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1996a) The genus Charmon Haliday (Hymenoptera: Braconidae: Charmontinae) from China. Entomotaxonomia 18: 59-64. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1996b) The genus Aulosaphes Muesebeck (Hymenoptera: Braconidae: Lysiterminae) from China. Entomotaxonomia 18: 223-229. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1997a) Two new species of the subfamily Miracinae (Hymenoptera: Braconidae) from China. Wuyi Science Journal 13: 63-69. Chen, X. X., He, J. H. & Ma, Y. (1997b) Hymenoptera: Braconidae. pp. 1647-1668. In Insects of the Three Gorge Reservoir Area of Yangtze River, China. Chongqing Press. [in Chinese with English summary] Chen, X. X., He, J. H. & Ma, Y. (1998) Revision of the genus Hartemita Cameron (Hymenoptera: Braconidae: Cardiochilinae) from China. Entomotaxonomia 20: 208-218. Chen, X. X., He, J. H. & van Achterberg, C. (1993) A revision of the subtribe Brulleiina van Achterberg (Hym.: Braconidae: Helconinae) from China. Zoologische Mededelingen 67: 375-395. Chen, X. X. & van Achterberg, C. (1997) Revision of the subfamily Euphorinae (Hymenoptera: Braconidae) from China. Zoologische Verhandelingen 311: 1-217. Chen, J. H. & Wu, Z. S. (1994) The Alysiini of China (Hymenoptera: Braconidae: Alysiinae). China Agriculture Press, Beijing. [in Chinese with English summary] Chin, S. F. (1937) Notes on the natural enemies of the paddy borer, Schoenobius incertellus Walker in Canton with a list its natural enemies in the world. Entomology & Phytopathology 5: 442-457. Chou, L. Y. (1981a) A preliminary list of Braconidae (Hymenoptera) of Taiwan. Journal of Agricultural Research China 30: 71-88. Chou, L. Y. (1981b) The genera of Aphidiidae (Hymenoptera: Ichneumonidae) in Taiwan. Journal of Agricultural Research China 30: 308-323. Chou, L. Y. (1985a) A new species of Neoneurinae (Hymenoptera: Braconidae) from Taiwan. Journal of Agricultural Research China 30: 477-480. Chou, L. Y. (1985b) A new species of Buluka (Hymenoptera: Braconidae) from Taiwan. Chinese Journal of Entomology 5: 85-88. Chou, L. Y. (1986) A new species of Chrysopophthorus from Taiwan (Hymenoptera: Braconidae). Chinese Journal of Entomology 6: 215-217. Chou, L. Y. (1987) The genus Aridelus of Taiwan (Hymenoptera: Euphorinae). Taiwan Agricultural Research Institute Special Publication 22: 19-39. Chou, L. Y. (1990) The Braconidae (Hymenoptera) of Taiwan, II. The genus Streblocera (Euphorinae). Journal of the Taiwan Museum 43: 89-148.
The Family Braconidae in China (Hymenoptera) 335
Chou, L. Y. (1995) The Braconidae (Hymenoptera) of Taiwan. V. Cardiochilinae and Orgilinae. Journal of Agricultural Research China 44: 174-220. Chou, L. Y. & Chou, K. C. (1991) The Braconidae (Hymenoptera) of Taiwan. IV. Histeromerinae. Journal of Agricultural Research China 40: 472-474. Chou, L. Y. & Chou, K. C. (1993a) A new species of Trioxys (Hymenoptera: Braconidae) from Taiwan. Chinese Journal of Entomology 13: 375-378. Chou, L. Y. & Chou, K. C. (1993b) The genus Zele of Taiwan (Hymenoptera: Braconidae). Journal of Agricultural Research China 42: 446-448. Chou, L. Y. & Hsu, T. C. (1995) The Braconidae (Hymenoptera) of Taiwan. VI. Charmontinae, Homolobinae and Xiphozelinae. Journal of Agricultural Research China 44: 357-378. Chou, L. Y. & Hsu, T. C. (1996) The Braconidae (Hymenoptera) of Taiwan. VII. Subtribe Triaspina. Journal of Agricultural Research China 45: 436-497. Chou, L. Y. & Hsu, T. C. (1998) The Braconidae (Hymenoptera) of Taiwan. VIII. Brulleiini, Diospilini and Helconini. Journal of Agricultural Research China 47: 283-314. Chou, L. Y. & Lee, P. P. (1991) The Braconidae (Hymenoptera) of Taiwan, III. Cenocoeliinae. Chinese Journal of Entomology 11: 49-57. Chou, L. Y. & Sharkey, M. J. (1989) The Braconidae (Hymenoptera) of Taiwan. I. Agathidinae. Journal of the Taiwan Museum 42: 147-223. Chu, J. T. (1934) An investigation on the parasites of Parnara guttatus Brem. from Hangchow. Entomology & Phytopathology 2: 662-663. Chu, J. T. (1935) Preliminary notes on the ichneumon flies in Kiangsu and Chekiang Provinces China. Yearbook of the Bureau of Entomology of Chekiang Province 4, 7: 32. Chu, J. T. (1936) Notes on Cheloninae of China, with description of a new species. (Hymen., Braconidae). Entomology & Phytopathology 4: 682-685. Chu, J. T. (1937) Notes on the hymenopterous parasites of the pine caterpillar Dendrolimus punctatus Walker in China. Entomology & Phytopathology 5: 56-103. Chu, J. T. & He, J. H. (1973) The identification of common parasitic wasps on the rice stem borers from China (I) – Braconidae. Utilization and Control of Animal 1973 (3): 9-14. Chu, J. T., He, J. H. & Yun, J. X. (1976) The Notes of parasitic wasps on the green rice caterpillar Naranga aenescerns Moore from China (I) – Braconidae. Kunchong Zhishi (Entomological Knowledge) 13: 145-147. He, J. H. (Ed.) (1979) Atlas of Natural Enemies of Rice Pests in Zhejiang Province, China. Zhejiang People’s Publishing House. [in Chinese] He, J. H. (1982) Four common species of Meteorus in China. Kunchong Zhishi (Entomological Knowledge) 19: 31-34. [in Chinese] He, J. H. (1985) Description of a new species of Distrilirella van Achterberg (Hymenoptera: Braconidae). Acta Agriculturae Universitatis Zhejiangensis 11: 327-329. [in Chinese with English summary] He, J. H. & Chen, X. X. (1992) The genus Acampsis Wesmael (Hymenoptera: Braconidae: Sigalphinae) from China. Entomotaxonomia 14: 217-221. [in Chinese with English summary] He, J. H. & Chen, X. X. (1993) The genus Sigalphus Latreille (Hymenoptera: Braconidae: Sigalphinae) from China. Acta Entomologica Sinica 36: 90-93. [in Chinese with English summary] He, J. H. & Chen, X. X. (1995) One new species of the genus Pentatermus Hedqvist (Hymenoptera: Braconidae: Lysiterminae) from China. Entomotaxonomia 17: 225-227. [in Chinese with English summary]
Chen Xuexin, He Junhua & Ma Yun 336
Hymenoptera: Evolution, Biodiversity and Biological Control
He, J. H. & Chen, X. X. (1996) One new species of the genus Rattana van Achterberg (Hymenoptera: Braconidae: Cenocoeliinae) from China. Entomotaxonomia 18: 219-222. [in Chinese with English summary] He, J. H., Chen, X. X. & Ma, Y. (1989) The braconid (Hymenoptera) parasites of Adoxophyes oana Fischer von Rosbergstamm (Lepidoptera: Tortricidae) from China. Acta Agriculturae Universitatis Zhejiangensis 15: 437-439. [in Chinese with English summary] He, J. H., Chen, X. X. & Ma, Y. (1994) Revision of the Sigalphus species from China with descriptions of two new species (Hymenoptera: Braconidae: Sigalphinae). Journal of Zhejiang Agricultural University 20: 441-448. He, J. H., Chen, X. X. & Ma, Y. (1996) Five new records of species of Macrocentrus Curtis (Hymenoptera: Braconidae) from China. Entomotaxonomia 18: 77-78. [in Chinese with English summary] He, J. H., Chen, X. X. & Ma, Y. (2000) Fauna of China: Insecta,Hymenoptera, Braconidae (I). Science Press, Beijing (in press). [in Chinese with English summary] He, J. H., Chen, X. X. & van Achterberg, C. (1994) Siniphanerotomella gen. nov., a new genus of the subfamily Cheloninae Nees (Hymenoptera: Braconidae) from China. Zoologische Mededelingen 68: 191-195. He, J. H., Chen, X. X. & van Achterberg, C. (1997a) One new genus of the subfamily Neonurinae (Hymenoptera: Braconidae) from China. Wuyi Science Journal 13: 70-75. He, J. H., Chen, X. X. & van Achterberg, C. (1997b) Five new species of the subfamily Ichneutinae (Hymenoptera: Braconidae) from China. Zoologische Mededelingen 71: 9-23 He, J. H., Chen, X. X. & van Achterberg, C. (1997c) Scarichelonus, a new subgenus of the genus Chelonus Panzer (Hymenoptera: Braconidae) from China. Zoologische Mededelingen 71: 5356. He, J. H., Chen, X. X., Zhou, Z. H. & Liu, Z. B. (1990) Two new records of species of Braconidae (Hymenoptera) parasitic on rice insect pests from China. Acta Agriculturae Universitatis Zhejiangensis 16: 217. [in Chinese with English summary] He, J. H. & Lou, X. M. (1993) Description of a new species of Macrocentrus Curtis parasitic on Cnaphalocrocis medinalis Guenée (Hymenoptera: Braconidae: Macrocentrinae). Entomological Journal of East China 2: 12-16. [in Chinese with English summary] He, J. H., Lou, X. M. & Ma, Y. (1996) Notes on Rectizele van Achterberg from China (Hymenoptera: Braconidae: Macrocentrinae). Journal of Zhejiang Agricultural University 22: 33-36. [in Chinese with English summary] He, J. H. & Ma, Y. (1982) Two common species of the genus Zombrus (Braconidae) from China. Insect Natural Enemies 4: 18-19. [in Chinese] He, J. H. & Pang, X. F. (Eds) (1986) Atlas of Natural Enemies of Rice Pests in China. Shanghai Science and Technology Publishing House. [in Chinese] He, J. H. & van Achterberg, C. (1994) A revision of the genus Auloacocentrum Brues (Hymenoptera: Braconidae: Macrocentrinae) from China. Zoologische Mededelingen 68: 159-171. Hsich, S. Y. (1957) Preliminary notes on parasitic Hymenoptera (Hymenoptera, Parasitica) attacking important pests of rice in the province of Hunan, China. Acta Entomologia Sinica 7: 295-319. Institute of Zoology of Academia Sinica, Zhejiang Agricultural University, ect. (1978) Atlas of Natural Enemies of Economic Insects. Science Press, Beijing. [in Chinese] Li, F. S. (1935) A list of the parasitic and predacious insects of cotton pests in Kiangsu and Chekiang. Entomology & Phytopathology 3: 304-307. Nixon, G. E. J. (1965) A reclassification of the tribe Microgasterini (Hymenoptera: Braconidae). Bulletin of the British Museum (Natural History), Entomology Supplement 2: 1-288.
The Family Braconidae in China (Hymenoptera) 337
Papp, J. & Chou, L. Y. (1995) The genus Wesmaelia Foerster of Taiwan (Hymenoptera: Braconidae: Euphorinae). Chinese Journal of Entomology 15: 345-354. Papp, J. & Chou, L. Y. (1996) The genus Mirax from Taiwan (Hymenoptera: Braconidae: Miracinae). Chinese Journal of Entomology 16: 107-115. Quicke, D. L. J., Chishti, M. J. K., Chen, X. X. & Kruft, R. A. (1997) Revision of Yelicones (Hymenoptera: Braconidae: Rogadinae) from the East Palaearctic and Oriental regions with descriptions of four new species. Journal of Natural History 41: 779-797. Shenefelt, R. D. (1969) Braconidae 1. Hymenopteroeum Catalogus 4: 1-176. Dr W. Junk N. V., 'sGravenhage. Shenefelt, R. D. (1970) Braconidae 2. Hymenopteroeum Catalogus 5: 177-306. Dr W. Junk N. V., 's-Gravenhage. Shenefelt, R. D. (1971) Braconidae 3. Hymenopteroeum Catalogus 6: 307-428. Dr W. Junk N. V., 's-Gravenhage. Shenefelt, R. D. (1972) Braconidae 4. Hymenopteroeum Catalogus 7: 429-668. Dr W. Junk N. V., 's-Gravenhage. Shenefelt, R. D. (1973) Braconidae 5. Hymenopteroeum Catalogus 9: 669-812. Dr W. Junk N. V., 's-Gravenhage. Shenefelt, R. D. (1973) Braconidae 6 Hymenopteroeum Catalogus 10: 813-936. Dr W. Junk B. V., 's-Gravenhage. Shenefelt, R. D. (1974) Braconidae 7. Hymenopteroeum Catalogus 11: 937-1114. Dr W. Junk B. V., 's-Gravenhage. Shenefelt, R. D. (1975) Braconidae 8. Hymenopteroeum Catalogus 12: 1115-1262. Dr W. Junk B. V., 's-Gravenhage. Shenefelt, R. D. (1978) Braconidae 10. Hymenopteroeum Catalogus 15: 1425-1872. Dr W. Junk B. V., The Hague. Shenefelt, R. D. & Marsh, P. M. (1976) Braconidae 9. Hymenopteroeum Catalogus 13: 1263-1424. Dr W. Junk B. V., 's-Gravenhage. Tang, Y. Q. & Marsh, P. M. (1994) A taxonomic study of the genus Ascogaster in China (Hymenoptera; Braconidae: Cheloninae). Journal of Hymenoptera Research 3: 279-302. Wang, J. Y. (1984a) Two new species of the subfamily Agathidinae from China (Hymenoptera: Braconidae). Acta Zootaxonomia Sinica 9: 151-154. [in Chinese with English summary] Wang, J. Y. (1984b) A new species of the genus Meteorus Haliday from Sichuan province (Hymenoptera: Braconidae). Acta Zootaxonomia Sinica 9: 321-323. [in Chinese with English summary] Wharton, R. A. and Chou, L. Y. (1983) The genus Heratemis Walker, with a review of the Taiwanese species (Hymenoptera: Braconidae). Journal of the Taiwan Museum 36: 7-13. Wharton, R. A. and Chou, L. Y. (1985) Revision of the species of Alloea Haliday (Hymenoptera: Braconidae, Alysiinae). Journal of Agricultural Research. China 34: 352-367. Yang, Z. Q. (1996) Parasitic Wasps on Bark Beetles in China. Science Press, Beijing. You, L. S. & Luo, H. W. (1988) New records of the genus Macrocentrus Curtis from China. Journal of Hunan Agricultural College 14: 37-38. [in Chinese] You, L. S., Quicke, D. L. & Zhou, Z. H. (1994) Notes on eighteen braconid species (Hymenoptera: Braconidae) from China. Wuyi Science Journal 11: 120-125. [in Chinese] You, L. S. & Xiong, S. L. (1983) Two new species of Apanteles Foerster (Hymenoptera: Braconidae) from China. Entomotaxonomia 5: 225-229. [in Chinese with English summary] You, L. S., Xiong Shulin & Wang, Z. D. (1988) Annotated list of Apanteles Foerster (Hymenoptera: Braconidae) from China. Entomologica Scandinavica 19: 35-42.
Chen Xuexin, He Junhua & Ma Yun 338
Hymenoptera: Evolution, Biodiversity and Biological Control
You, L. S., Xiong, S. L. & Zhou, Z. H. (1985) A new species of Apanteles Foerster (Hymenoptera: Braconidae) from China. Acta Zootaxonomia Sinica 10: 421-423. [in Chinese with English summary] You, L. S., Xiong, S. L. & Zhou, Z. H. (1987) On a new species of Apanteles Foerster from Yunnan province (Hymenoptera: Braconidae: Microgasterinae). Acta Zootaxonomia Sinica 12: 424426. [in Chinese with English summary] You, L. S., Zhou, Z. H. & Tong, X. W. (1990) Description of Xiphozele Cameron in Hunan and Guangxi (Hymenoptera: Braconidae). Journal of Hunan Agricultural College 16: 150-152. [in Chinese with English summary] You, L. S., Zhou, Z. H. & Tong, X. W. (1991) Two new species of the genus Sigalphus Latreille from Hunan province (Hymenoptera: Braconidae: Cheloninae). Acta Entomologia Sinica 34: 225229. [in Chinese with English summary]
An Annotated List of Encyrtidae (Hymenoptera: Chalcidoidea) of Tbilisi (Georgia) G. O. Japoshvili Institute of Zoology, Tbilisi, Republic of Georgia (email:
[email protected])
Introduction Georgia is a relatively small mountainous country in the Caucasus which has a relatively rich flora and fauna. The region around the border city of Tbilisi has a particularly rich assemblage of insects. There are many species of economically important sucking insects viz. Coccoidea, Psylloidea and Aleyrodoidea, and a corresponding large number of parasitoids associated with them. The Encyrtidae is the most speciose group of parasitoids attacking scale and psyllid insects. In the Palaearctic there is aproximately 1260 described species of encyrtids, predominantly from southern regions (Trjapitzin 1989), and 248 species are known from the Caucasus (Trjapitzin & Doganlar 1997). This is a relatively rich fauna, given that only 50 species are known for Turkey, an area twice that of the Caucasus (Trjapitzin & Doganlar 1997). The aim of this investigation was to document the encyrtid species of Tbilisi, and record their insect hosts and associated host plants. The authors for all insect species are given in Table 1. The material was collected from 1994 to 1998 in different parts of Tbilisi, and voucher material is lodged in the Institute of Zoology, Tbilisi. Taxonomic concepts mostly follow Trjapitzin (1989) and Noyes and Woolley (1994).
Results and Discussion Table 1 presents a detailed list of the encyrtid species recorded from Tbilisi. Previously, there were only 20 encyrtid species recorded from Tbilisi, however the present study has expanded this to 42, including a new species, Psyllaephagus georgicus Yasnosh & Japoshvili (1999). In addition to the 22 new records, eight of these species are also new to the Caucasus. The most commonly collected species from Tbilisi were Blastothrix longipennis, B. nikolskajae, Cheiloneurus claviger, Encyrtus lecaniorum and Microterys sylvius; the other species being encountered more rarely. The extensive rearing undertaken during this study also allows a number of generalisations to be made about the host relationships of the recorded encyrtids. The majority of the 42 species have been recorded from a single host, with only five species being reared from multiple hosts (Table 1). Of the 27 host species listed, 17 only have a single species of encyrtid associated with them, while 10 have between two and five parasitoid species. For example, the coocid Sphaerolecanium prunastri has had four encyrtids reared from it at Tbilisi: Cerapterocerus mirabilis, Discodes coccophagus, Metablastothrix truncatipennis and Microterys hortulanus. Although the above results are comprehensive, given the timeframe of the study and extent of the rearings, it is also likely that additional encyrtid species and host associations will be documented for Tbilisi and the Caucasus as work continues in the future.
339
G.O. Japoshvili 340
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 List of encyrtid species recorded from Tbilisi, with their insect host(s) and food plant(s) (nt = new record for Tbilisi; nc = new record for Caucasus; APH = Aphalaridae; AST = Asterolecaniidae; COC = Coccidae; DIA = Diaspididae; ER = Eriococcidae; KER = Kermesidae; PS = Pseudococcidae; PY = Psyllidae; TRI = Triozidae). Parasitoid
Host
Host Plant
Anagyrus pseudococci (Girault)
Planococcus ficus Signoret (PS)
Hedera, Ficus carica
Reference Yasnosh 1972
Anthemus funicularis (Bakkendorf)
Leucaspis loewi Colvee (DIA)
Pinus
nt/nc
Anthemus pini Ferriére
Leucaspis pusilla Löw (DIA)
Pinus
Trjapitzin 1968
Aphycoides clavellatus (Dalman)
Physokermes hemicriphus (Dalman) (COC)
Picea
nt
Blastothrix hungarica Erdös
Rhodococcus spiraeae Borchsenius (COC), Parthenolecanium persicae F. (COC)
Spiraea
Yasnosh 1972
Blastothrix longipennis Howard
Parthenolecanium corni Bouché (COC)
Acacia, Fraxinus, Cercis siliquastrum, Ulmus foliacea, Crataegus, Malus, Thelicranium australis, Corylus
nt
Blastothrix nikolskajae Sugonjaev
Parthenolecanium rufulum Cockerell (COC)
Spiraea
Yasnosh 1972
Blastothrix sp.nr nikolskaiae
Rhodococcus spiraeae
Spiraea
nt/nc
genus & sp. indet. nr Bureshiella Hoffer
Parthenolecanium corni
Cercis siliquastrum
nt/nc
Cerapterocerus mirabilis Westwood
Sphaerolecanium prunastri Fonscolombe (COC)
Prunus
nt; Japoshvili in press
Cheiloneurus claviger (Thomson)
Ceroplastes japonicus Green (COC) Rhodococcus spiraeae
Ilex, Spiraea
nt
Cheiloneurus kollari (Mayr)
Planchonia arabidis Signoret (AST)
Hedera
Yasnosh 1972
Cheiloneurus paralia (Walker)
Eriopeltis festucae Fonscolombe (COC)
Festucae, Arrenaterium elatius, Lolium
Trjapitzin 1968
Choreia maculata (Hoffer)
Eriopeltis festucae
Arrhenatherium elatius
Akhvlediani 1966
Discodes coccophagus (Ratszeburg)
Sphaerolecanium prunastri
Prunus
nt; Japoshvili in press
Encyrtus lecaniorum (Mayr)
Coccus hesperidum L. (COC)
Diospyros
nt
An Annotated List of Encyrtidae (Hymenoptera: Chalcidoidea) of Tbilisi (Georgia) 341
Epitetracnemus zetterstedtii (Westwood)
Lepidosaphes ulmi L. (DIA)
Populus tremula
Akhvlediani 1966
Hadzhibeylia physococci Myartseva & Trjapitzin
Ritsemia pupifera Lichtenstein (PS)
Ulmus foliacea
Myartseva & Trjapitzin 1981
Homalotylus quaylei Timberlake
Planococcus ficus
–
Trjapitzin 1968
Mahencyrtus coccidiphagus (Mercet)
Coccus hesperidum
Crataegus
nt/nc
Metablastothrix truncatipennis Ferriére
Sphaerolecanium prunastri
Prunus divaricata
nt/nc
Metaphycus asterolecanii (Mercet)
Asterodiaspis quercicola Bouché (AST) Parthenolecanium corni, Pulvinaria populi Signoret (COC)
–
Trjapitzin 1968
Fraxinus Populus transcaucasica, P. gracilis –
Yasnosh 1972
Metaphycus insidiosus (Mercet)
” ”
Metaphycus zebratus (Mercet)
Planchonia arabidis
Microterys clauseni Compere
Ceroplastes japonicus
Ilex, Laurus nobilis, Hedera
nt; Yasnosh & Japoshvili 1998
Microterys duplicatus (Nees)
Parthenolecanium corni
Cercis siliquastrum
nt
Microterys ferrugineus (Nees)
Kermes roboris Fourcroy (KER)
Quercus
nt
Microterys hortulanus (Erdös)
Sphaerolecanium prunastri
Prunus
nt; Japoshvili in press
Microterys sylvius (Dalman)
Rhodoccus spiraeae
Spiraea
nt
Microterys tricoloricornis (De Stefani)
Coccus hesperidum
Hedera
Yasnosh 1972
Microterys trjapitzini Jasnosh
Acanthococcus aceris Signoret (ER)
Acer
Yasnosh 1969
Monodiscodes itermedius (Mayr)
Rhizopulvinaria armeniaca Borchsenius (COC)
Herniaria
Yasnosh 1972
Prionomitus mitratus (Dalman)
Psylla crataegi Schrunk (PY)
Crataegus
nt
Pseudaphycus phenacocci Jasnosh
Phenacoccus mespili (Signoret) (PS)
–
Yasnosh 1957
G.O. Japoshvili 342
Hymenoptera: Evolution, Biodiversity and Biological Control
Parasitoid
Host
Host Plant
Reference
Psyllaephagus bachardenicus Myartseva
Psylla ramnicola Scott (PY) Trioza magnisetosa Loginova (TRI) Trioza magnisetosa
Rhamnus pallasii, Elaeagnus angustifolia Elaeagnus angustifolia
nt/nc; Yasnosh & Japoshvili 1999
Psyllaephagus tokgaevi Myartseva Psyllaephagus georgicus Yasnosh & Japoshrili
Crastina tamaricina Loginova (APH) Trioza magnisetosa
Tamarix Elaeagnus angustifolia
Trichomasthus albimanus Thomson
Parthenolecanium corni Luzulaspis luzulae Dufour (COC) Planchonia arabidis Eriopeltis festucae Eriopeltis festucae
Prunus divaricata
Psyllaephagus sp. nr rubriscutellatus Myartseva
Trichomasthus cyaneus (Dalman) Trichomasthus ivericus Jasnosh
Gramineae
nt/nc; Yasnosh & Japoshvili 1999 nt/nc; Yasnosh & Japoshvili 1999 nt/nc; Yasnosh & Japoshvili 1999
Trjapitzin 1968; Yasnosh 1972 Trjapitzin 1968
Festucae Acanthococcus aceris
Yasnosh 1969 Acer
Zaomma lambinus (Walker)
Adiscodiaspis tamaricicola Malenotti (DIA)
Tamarix
nt
An Annotated List of Encyrtidae (Hymenoptera: Chalcidoidea) of Tbilisi (Georgia) 343
Acknowledgements The author wishes to thank Dr V. Yasnosh, Dr V. Trjapitzin and Dr E. Kvavadze for their help and advice during this project.
References Akhvlediani, M. (1966) Materials for the fauna of chalcids (Hymenoptera: Chalcidoidea) in East Georgia. Materials for fauna of Georgia. pp. 78-84 [in Russian]. Japoshvili G.O. (in press) On the population dynamics of the plum scale Spaerolecanium prunastri Fonscolombe (Coccoidea, Coccidae) in Georgia. Proceedings of VIII ISSIS. Wye College, University of London. Myartseva, S. N. & Trjapitzin, V. A. (1981) A new genus of encyrtids (Hymenoptera) from Georgia. Zoologichesky Zhurnal 60: 621-623 [in Russian]. Noyes, J. S. & Woolley, J. B. (1994) North American encyrtid fauna (Hymenoptera: Encyrtidae): taxonomic changes and new taxa. Journal of Natural History 28: 1327-1401. Trjapitzin, V. A. (1968) On the Encyrtidae fauna of Caucasus. Proceedings, All-Union Entomological Society (Leningrad) 52: 43-125 [in Russian]. Trjapitzin, V. A. (1989) Parasitic Hymenoptera of the Fam. Encyrtidae of Palaearctics. Leningrad “Nauka”, Leningrad division [in Russian]. Trjapitzin, V. A. & Doganlar, M. (1997) A review of encyrtids (Hymenoptera, Encyrtidae) of Turkey. Revue d’Entomologie de I’URSS 86: 213-222 [in Russian]. Yasnosh, V. A. (1957) New parasites (Hymenoptera: Aphelinidae, Encyrtidae) reared from Coccoidea in Georgia, Caucasus. Revue d’Entomologie de I’URSS 36: 715-720 [in Russian]. Yasnosh, V. A. (1969) New species of Ecyrtidae (Chalcidoidea) in Georgia. Zoologichesky Zhurnal 6: 931-935 [in Russian]. Yasnosh, V. A. (1972) Chalcids (Hymenoptera, Chalcidoidea) – parasites of coccids in arid forests of Georgia. Proceedings, All-Union Entomological Society (Leningrad) 55: 217-247 [in Russian]. Yasnosh, V. A. & Japoshvili, G. O. (1998) Japanese wax scale and natural enemies in Tbilisi. Bulletin of the Georgian Academy of Sciences 157: 132-134. Yasnosh, V.A. & Japoshvili, G.O. (1999) Parasitoids of the genus Psyllaephagus Ashmead (Hymenoptera: Chalcidoidea: Encyrtidae) in Georgia with the description of P. georgicus sp. nov. Bulletin of the Georgian Academy of Sciences 159: 516-519.
This Page Intentionally Left Blank
PART
7
Biological control
This Page Intentionally Left Blank
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci (Biotype ‘B’), Based on Morphological and Molecular Systematics J. A. Goolsby1, M. A. Ciomperlik2, A. A Kirk4, W. A. Jones3, B. C. Legaspi, Jr.3, J.C. Legaspi5, R. A Ruiz2, D. C. Vacek2 and L. E. Wendel2 1
USDA-ARS, Australian Biological Control Laboratory, 120 Meiers Rd, Indooroopilly, QLD 4068 Australia (email:
[email protected]) 2
USDA-APHIS-PPQ Mission Plant Protection Centre, P. O. Box 2140, Mission, TX 78573 USA
3
USDA-ARS, Beneficial Insects Research Unit, Weslaco, TX 78596 USA
4
USDA-ARS, European Biological Control Laboratory, Montpellier, France
5
Texas A & M University, Texas Agricultural Experiment Station, Weslaco, TX 78596 USA
Introduction In support of the United States Department of Agriculture (USDA) National Research and Action Plan for Management of silverleaf whitefly, the Animal and Plant Health Inspection Service (USDA-APHIS) collaborated with USDA, Agricultural Research Service (ARS), state departments of agriculture and universities to implement biological control strategies for Bemisia tabaci (Gennadius) (Biotype ‘B’) (Homoptera: Aleyrodidae) (= Bemisia argentifolii Bellows & Perring). As the primary quarantine facility in the USA for the importation of exotic natural enemies of the silverleaf whitefly, the USDA-APHIS-PPQ, Mission Plant Protection Centre (MPPC) in Texas processed over 80 shipments of predators, parasitoids and pathogens sent by collectors world-wide from 1992–1998. During this time period B. tabaci was a serious pest of vegetables, cotton and ornamentals across the United States subtropical growing areas and in greenhouses throughout the country. MPPC imported and cultured over 56 populations of Encarsia spp. and Eretmocerus spp. (both Aphelinidae), several of which were new species. Parasitoids were categorised in quarantine using RAPD-PCR and morphologically-based systematics. Integration of the two techniques proved to be useful in capturing the maximum amount of species diversity with a minimum amount of duplication in cultures. Additionally, the two methods were integrated for identifying indigenous and imported parasitoids in field evaluation efforts. A combination of predictive and empirical evaluation methods was developed which analysed the performance of the imported parasitoid species under laboratory, caged and free release conditions. This paper discusses the evaluation of multiple species in a large-scale classical biological control program, and how it was made possible by the integration of morphological and molecular systematics. We hope this program will serve as a model for future biological control programs. Evaluations of the parasitic Hymenoptera were performed in Texas, Arizona and California during 1994–96. The work done in Texas is the primary focus of this paper. Researchers in
347
J. A. Goolsby et al. 348
Hymenoptera: Evolution, Biodiversity and Biological Control
Arizona and California in collaboration with MPPC (Hoelmer 1998; Hoelmer et al. 1998; Pickett et al. 1999) conducted similar work. Evaluation of the various geographic strains or species of parasitoids studied at MPPC began in quarantine with an assessment of their fecundity on selected crop plants (Goolsby et al. 1996). Promising parasitoid populations were then reared and released onto these same crops in the field in order to measure rates of parasitism under field conditions (Goolsby et al. 1998). All species approved for release were further tested using free releases in an establishment evaluation. To determine if released populations became widely established, whitefly populations were sampled periodically during 1997–98 using the ‘sentinel plant’ technique for recruitment of parasitoids.
Quarantine Methods Identification of new organisms being imported is one of the critical functions of quarantine work. Separation of natural enemies into distinct taxa should occur as cultures are initiated. Characterisation of natural enemies in quarantine often requires a blend of observations of their morphology, mating behaviour, host plant associations, etc. Taxonomic keys may be available, but are often difficult to use with live material. In the Bemisia biological control program, taxonomic keys to the Eretmocerus Haldeman and Encarsia Foerster were not available for either the indigenous or imported species. Eretmocerus species indigenous to the southern United States and south-west desert were poorly understood but, based on biosystematic studies, were determined to be a complex of species (Hunter & Rose 1996). Foreign material arriving into quarantine showed extensive diversity, especially in the genus Eretmocerus (Legaspi et al. 1996; Goolsby et al. 1998). To best handle the issues of cryptic species, species complexes, and the need to initiate pure cultures representing the maximum available diversity of natural enemies, a unique quarantine protocol was developed which integrated biosystematics and molecular techniques. Foreign collections were categorised in quarantine by plant type, site location and the macrocharacters of the parasitic Hymenoptera and Aleyrodidae. Only parasitoids reared from individuals of the B. tabaci complex met the requirements for permitting, as stated in the Environmental Assessments of the genera Eretmocerus and Encarsia (APHIS 1995). Further, the imported species must have had a biology described as uniparental, biparental or autoparasitoid. Species which displayed obligate hyperparasitism of other taxa were not considered suitable for release (Hunter et al. 1996). The requirements of the Environmental Assessments were intended to identify the parasitoid species with the most specificity to the B. tabaci complex. Species which met these criteria were acceptable for processing using our quarantine protocol. Eretmocerus and Encarsia were separated into distinct groups using the morphology of the pupae and adult females. Individuals from each unique accession were immediately characterised at the MBCC Genetics Laboratory using RAPD-PCR with primers CO4 and A10 (Black et al. 1992; Vacek et al. 1996). Detailed methodology and representative electrophoretic gel patterns for Eretmocerus and Encarsia parasitoids are contained in Legaspi et al. (1996). Cohorts of the original parental material were sent to co-operating systematists. Information from the collaborating systematists and geneticists allowed for characterisation of quarantine material while the original parental cohort was still alive. Typically, material was characterised using both methods within two to three days after acceptance into quarantine. Unique parasitoid accessions were set up in pure cultures reared on the local B. tabaci with hibiscus var. kona pink, Hibiscus rosasinensis L. as the plant host. Duplicate accessions were combined or, in the later stages of the program, processed only for reference purposes. Representatives of all the accessions were cryogenically stored at the MPPC Genetics Laboratory, and vouchered at the Texas A&M University, Department of Ento-
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci 349
mology Collection, College Station, Texas and the USDA-ARS, Systematic Entomology Laboratory, Washington, D. C. From 1992 to 1998, 235 accessions were processed, which resulted in 56 cultures of parasitic Hymenoptera, 17 of which were determined to be distinct species (Appendix 1). Some species were widely distributed in the Old World, such as Eretmocerus mundus Mercet, and displayed the same RAPD’s banding pattern from all eighteen locations. By comparison, Eretmocerus emiratus Rose & Zolnerowich was collected from only one locality and displayed a unique banding pattern. Apart from differences in RAPD’s patterns, new species were only distinguishable from each other by minute differences in the first funicular antennal segment of the female. Screening with RAPD’s was really the only method available for distinguishing such cryptic species. Without RAPD’s, multiple cultures of widely distributed species, such as E. mundus, would have used most of the quarantine resources. These were conserved which allowed for new species to be imported later in the program. With both RAPD’s and morphological identifications available, it became apparent that unique RAPD’s patterns directly corresponded to distinct species for both Eretmocerus and Encarsia accessions. The one exception appeared to be Encarsia transvena Timberlake. This species was collected in many different geographic regions and was characterised by five separate banding patterns (Appendix 1). However, Viggiani (pers. comm.) has found some crossing incompatibility between the Spanish and Pakistan populations of E. transvena that is correlated with morphometric differences. It is possible this could reflect species level differences in the two E. transvena populations just as the RAPD’s patterns suggested. In summary, from our experience with these two genera, using RAPD’s in quarantine as a way of measuring and identifying diversity appears to be an excellent means of making preliminary separations of geographic populations or strains, and possible new species. Biological control programs may be well served if RAPD-PCR could be included as part of the protocol in the quarantine phase of the program.
Laboratory and Field Evaluation of Parasitic Hymenoptera In order to screen the full diversity of natural enemies being imported, we devised a three-tiered system of experiments, starting with laboratory tests and moving to more realistic field studies. The field experiments were divided into replicated field cage tests and free releases into long-term garden plots. The free release test, which we called ‘establishment evaluation’, provided an opportunity for ‘poor’ performing agents to demonstrate their potential in a non-agricultural setting. Establishment evaluations were conducted in garden plots in the LRGV which had a diversity of host plants with silverleaf whitefly, planted in a continuous rotation. The combination of this three-tiered system gave us both predictive and empirical data for decision-making in the action program. A total of 38 exotic and two native parasitoids were evaluated in laboratory and field experiments (Goolsby et al. 1998). Numbers of B. tabaci parasitised were counted in sleeve cages on cantaloupe melons (Cucumis melo L. cv ‘Perlita’), cotton (Gossypium hirsutum L. cv ‘Delta Pine 51’), and broccoli (Brassica oleracea L. cv ‘Patriot’). Highest attack rates were found for Encarsia nr. pergandiella (Brazil) and E. mundus (Spain) on melons; for Eretmocerus hayati Rose & Zolnerowich (Pakistan) on cotton; and for E. mundus (Spain) on broccoli. In the laboratory, these three exotic parasitoids attacked significantly greater numbers of hosts than the native species of Encarsia pergandiella Howard and Eretmocerus tejanus Rose & Zolnerowich. Selected exotic parasitoids were evaluated in the field using sleeve cages on melons, cotton and kale (Brassica
J. A. Goolsby et al. 350
Hymenoptera: Evolution, Biodiversity and Biological Control
oleracea L. cv ‘Siberian kale’). Eretmocerus spp. from Spain and India performed well in all crop types. Encarsia nr pergandiella performed well on melons, but not on kale or cotton. In the establishment evaluation, 29 species/populations were released in the garden plots. During the two years of monitoring, 11 of these populations were recovered. Eretmocerus mundus and E. hayati were most commonly recovered throughout the evaluation period. Results from the laboratory, field and establishment evaluation were used to prioritise the E. mundus and E. hayati for mass rearing and development of augmentative biological control programs against B. tabaci (Goolsby & Ciomperlik 1999). Research co-operators in Arizona and California to prioritise candidates for field evaluation also used these findings. Identification of parasitoid species was critical to the field evaluation program. Morphological characters of the immature parasitoids were used to separate the two genera encountered in field samples. Native Encarsia pergandiella were distinguished from the Eretmocerus spp. by the presence of meconia in the host remains of the fourth instar whitefly. Encarsia transvena was never common in the field collection, but could be clearly identified by its black pupal skin. Pupal Eretmocerus were removed from the leaf and held in vials placed within a humiditron for emergence of the adults (DeBach & Rose 1985). Two primers, A-10 and C0-4, were used in the RAPDPCR procedure to determine the percentages of Eretmocerus spp. The second primer, CO-4, produced the most useful DNA profiles. The cost of running specimens through the RAPD’s process was the limiting factor. To overcome the cost and time limitations of RAPD’s, research was initiated to develop specific DNA probes from the satellite DNA for development into a squash blot kit. The use of this technique is discussed in the next section.
Recovery Survey of Parasitoids Using Sentinel Plants During 1997 a survey program for establishment of exotic parasitoids was implemented using ‘sentinel’ plants. Sentinel plants are pre-infested with immature whitefly and placed in field locations to sample parasitoid species composition. Plants stay in the field for two days and are then returned to the laboratory for rearing and identification of the parasitoids. This technique has two major advantages over conventional leaf sampling methods: 1) sentinel plants provide a standardised test unit across a broad range of locations and crop types, and 2) sentinels give a true measure of primary parasitism without the masking effects of hyperparasitism by the autoparasitoid, E. pergandiella. Ten sentinel locations were selected which represented a varied mix of agricultural and urban sites across the LRGV. Muskmelon var. perlita, followed by cotton in the summer months were used for sentinel plants. Plants were grown in the greenhouse until the three-leaf stage and then infested with whitefly. The root balls of the plants were placed in sealed plastic containers to retain moisture. Ten sentinel plants were placed monthly in each location for a period of two days. Plants were then returned to the laboratory for removal of plant pests and live parasitoid adults. After recovery and cleaning, each plant was first placed in a 150-mm diameter ventilated tissue culture dish and held in a reach-in environmental growth chamber at 27˚C for development of the parasitoids. Encarsia pergandiella pupae on each leaf were counted and removed prior to emergence to avoid hyperparasitism of the Eretmocerus pupae. Eretmocerus males and females were removed daily upon emergence and separated for identification. Males were slide mounted to determine if they were exotic or native. The pedicels of males of the introduced species are uniformly fuscous as compared to an amber coloration in the native E. tejanus males. This character proved to be very useful in determining the overall percentage of native versus
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci 351
Table 1 Percentages and total numbers of exotic and native male Eretmocerus recovered from sentinel plants across time.
Date
Jun 97
% Exotic
1.5
% Native
98.5
N
225
141
Table 2
Aug 97
Sept 97
Oct 97
Jan 98
Feb 98
Mar 98
May 98
June 98
July 98
Aug 98
64.2
91.1
86.3
40.2
70.0
19.9
50.0
25.0
90.0
100.0
35.7
8.9
13.7
59.8
30.0
80.1
50.0
75.0
10.0
0
130
209
201
90
151
8
24
56
100
Identification of female Eretmocerus from sentinel plant recovery surveys
Technique
Date
N
E. mundus
E. hayati
E. tejanus
RAPD-PCR
Nov. 97
56
0%
64%
36%
Morphology
July 98
18
38%
62%
0%
Satellite DNA
Aug. 98
67
34%
38 %
28%
exotic Eretmocerus spp. The percentage assumes a standard female to male sex ratio of 60:40 across Eretmocerus spp. Female Eretmocerus were used for species determinations. These were made using slide mounts and keys to the imported species of Eretmocerus, RAPD-PCR, or Satellite DNA probes (Rose & Zolnerowich 1998). Species specific probes were developed from satellite DNA to allow for bulk identification of Eretmocerus specimens collected from field evaluation studies (Heilmann 1997). Specific patterns of positive and negative hybridisation distinguished E. mundus, E. hayati, and E. emiratus; however, the absence of hybridisation with any of the probes was interpreted to be the native E. tejanus. Both live pupae and adults were suitable for identification using the probe. The ability to use parasitoid pupae eliminated the need to rear adults from leaf samples for identification. Pupal Eretmocerus were affixed to a membrane for shipment to USDA-ARS (Fargo, ND) where the DNA probes were utilised. Further development of this technique is underway to detect additional species and to make a kit suitable for use in field settings. Results of the sentinel plant recovery survey showed a dramatic increase in the numbers of introduced Eretmocerus spp. (Table 1). At the beginning of the survey in June 1997, native E. tejanus represented greater than 95% of Eretmocerus spp. Three months later, during the fall of the 1997 (August- October), exotic populations began to increase, representing 85% (n = 700) of the Eretmocerus spp. recovered. Analysis of the female Eretmocerus using both morphological techniques and molecular genetics identified three species, E. tejanus, E. mundus, and E. hayati (Table 2). The two exotic species, E. mundus and E. hayati can now be considered to be widely established across the Rio Grande River flood plain. Further analysis of this data is currently underway by Jones and Goolsby.
Conclusions From 1993 to 1998, 17 species of Encarsia and Eretmocerus from 19 countries were imported and evaluated at MBCC for biological control of B. tabaci in the USA. Two introduced species, E. mundus and E. hayati are now established and changing the species complex attacking the silverleaf whitefly in Texas. Additional work is needed to quantify the impact of these biological control agents. Extensive pre-release studies were conducted that should provide a baseline
J. A. Goolsby et al. 352
Hymenoptera: Evolution, Biodiversity and Biological Control
comparison for future researchers (Moomaw 1995; Legaspi et al. 1997; Riley & Ciomperlik 1997). The impact of the introduced parasitoids on non-target whitefly species should also be investigated. Other whitefly species, such as Trialeurodes abutilonea (Haldeman), Aleurocanthus woglumi Ashby and Aleurothrixus floccosus Maskell, are common in the agricultural and urban habitats where the E. mundus and E. hayati have been colonised. The silverleaf whitefly biological control program provides an opportunity to study the field specificity of entomophagous biological control agents. The integration of the morphology-based systematics and molecular techniques has optimised the efficient characterisation of natural enemies in quarantine, and allowed each new cryptic species to be fully evaluated and maintained as a pure culture. The techniques were also equally valuable in field-testing. It was possible to test multiple species simultaneously in the field and sort out species in the recovery samples by their DNA profile. Combining the use of RAPDPCR and classical systematics allowed for an increase in the numbers of individuals, which could be characterised to species in the evaluation effort. Ultimately, this led to the best possible determination of which species showed the greatest efficacy in the field. The molecular methods were only valuable when combined with classical systematics. Properly identified and curated specimens will provide the best permanent record of the species released and established. Taxonomic keys, developed as part of this program, will provide a means for other researchers, without access to molecular expertise, to identify the known world-wide diversity of Bemisia parasitoids (Rose & Zolnerowich 1997, 1998). One of the unique aspects of the Bemisia biological control program was the availability of so many species of natural enemies from intensive directed foreign explorations, combined with extensive mass rearing facilities. This unique situation provided the opportunity to evaluate, on a large scale, many different parasitoid species simultaneously, and predict which species would be most effective. Many authors have proposed that biological control become more of predictive science rather than an empirical method (Ehler 1990; Harris 1998). Despite considerable discussion of the topic few biological control programs have attempted to test the value of predictive tests. In the Bemisia biological control program we endeavoured to predict efficacy of the imported parasitoids and then test resulting hypotheses. Two species, E. mundus and E. hayati, performed significantly better in laboratory and field evaluations conducted in Texas. We predicted that these two species would be the most efficacious and selected them for augmentative biological control field evaluations. To test our predictions, the other species/populations that were not selected were reared in substantial numbers and released (Goolsby et al. 1998). The realities of an applied biological control program did not allow us to rear and release equal numbers of all the species permitted. Because of this, it was not possible to test our hypothesis as rigorously as we would have liked. However, the recoveries of the E. mundus and E. hayati across a broad geographical area suggests that the laboratory and field tests were valuable in predicting the ‘success’ of these biological control agents. If the predictive methods used in a biological control program are accurate, the time taken to identify an effective natural enemy can be shortened. Reducing the time it takes to evaluate a collection of biological control agents also has value in terms of conserving resources in a research program and reducing the short-term impact of the target pest. We hope that these results will encourage other biological control programs to develop predictive methods and test their predictions in field settings. The information gathered from the combined predictive and empirical method that we propose might further the theoretical aspects of our science, and in turn increase the likelihood of success in future biological control programs.
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci 353
Acknowledgements We would like to thank the following systematists for their expertise: Mike Rose (Montana State University, Bozeman, MT), James B. Woolley and Greg Zolnerowich (Department of Entomology, Texas A & M University, College Station, TX), Mike Schauff (USDA-ARS, Systematic Entomology Laboratory, Smithsonian, Washington, D. C), John Heraty (Department of Entomology, University of California, Riverside, CA) and Ray Gill (California Department of Agriculture, Sacramento, CA). The satellite DNA probes were developed by Larry Heilmann (USDA-ARS, Fargo, ND). We also wish to thank Don Sands (CSIRO, Indooroopilly, Queensland) and Andy Austin (University of Adelaide, Adelaide, South Australia) for their helpful comments.
References APHIS – Animal and Plant Health Inspection Service, United States Department of Agriculture (1995) Field releases of nonindigenous parasitic wasps in the genus Eretmocerus and Encarsia (Hymenoptera: Aphelinidae) for biological control of whitefly pests (Homoptera: Aleyrodidae). Environmental Assessment. Riverdale, Maryland. Black, W. C. IV, DuTeau, N. M., Puterka, G. J., Nechols, J. R., & Pettorini, J. N. (1992) Use of the random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR) to detect DNA polymorphisms in aphids (Homoptera: Aphididae). Bulletin Entomological Research 82: 151-159. DeBach, P. & Rose, M. (1985) Humidity control during shipment and rearing of parasitic Hymenoptera. Chalcid Forum 4: 11-13. Ehler, L. E. (1990) Introduction strategies in biological control of insects. pp. 111-130. In Mackauer, M., Ehler, L. E. & Roland, J. (Eds), Critical Issues in Biological Control. Intercept, Andover, Hants. Goolsby, J. A, Legaspi, J. C., & Legaspi, Jr, B. C. (1996) Quarantine evaluation of exotic parasites of the sweetpotato whitefly, Bemisia tabaci (Gennadius). Southwestern Entomologist 21: 13-21. Goolsby, J. A., Ciomperlik, M. A., Legaspi, Jr, B. C., Legaspi, J. C. & Wendel, L. E. (1998) Laboratory and field evaluation of exotic parasitoids of Bemisia tabaci (Biotype ‘B’) in the Lower Rio Grande Valley of Texas. Biological Control 12: 27-135. Goolsby, J. A. & Ciomperlik, M. A. (1999) Development of parasitoid inoculated seedling transplants for augmentative biological control of silverleaf whitefly (Homoptera: Aleyrodidae). Florida Entomologist 4: 532-545. Harris, P. (1998) Evolution of Classical Weed Biocontrol: Meeting Survival Challenges. Bulletin of the Entomological Society of Canada 30: 134-143. Heilmann, L. (1997) Development of species specific DNA probes for Eretmocerus species. p. 87. In Henneberry, T. J., Toscano, N. C., Perring, T. M. & Faust, R. M. (Eds), Silverleaf Whitefly: National Research, Action and Technology Transfer Plan, 1997-2001: Fifth Annual Review. United States Department of Agriculture ARS 1997-01, San Diego. Hoelmer, K. A. (1998) Comparative field cage evaluations of top-performing introduced parasitoids in desert cantaloupes. p. 68. In Henneberry, T. J., Toscano, N. C., Perring, T. M. & Faust, R. M. (Eds), Silverleaf Whitefly: National Research, Action and Technology Transfer Plan, 1997-2001: First Annual Review. United States Department of Agriculture ARS 199801, Charleston. Hoelmer, K. A., Roltsch, W. J. & Simmons, G. S. (1998) Establishment of introduced Eretmocerus species in Imperial Valley, CA. p. 70. In Henneberry, T. J., Toscano, N. C., Perring, T. M. & Faust, R. M. (Eds), Silverleaf Whitefly: National Research, Action and Technology Transfer
J. A. Goolsby et al. 354
Hymenoptera: Evolution, Biodiversity and Biological Control
Plan, 1997-2001: First Annual Review. United States Department of Agriculture ARS 199801, Charleston. Hunter, M. S. & Rose, M. (1996) Courtship behavior, reproductive relationships, and allozyme patterns of three North American populations of Eretmocerus nr. californicus (Hymenoptera: Aphelinidae) parasitizing the whitefly Bemisia sp., tabaci complex (Homoptera: Aleyrodidae). Proceedings of the Entomological Society of Washington 98: 126-137. Hunter, M. S., Rose, M. & Polazek, A. (1996) Divergent host relationships of males and females in the parasitoid Encarsia porteri (Hymenoptera: Aphelinidae). Annals of the Entomological Society of America 89: 667-675. Legaspi, J. C., Legaspi, Jr., B. C., Carruthers, R. I., Goolsby, J. A., Jones, W. A., Kirk, A. A., Moomaw, C., Poprawski, T. J., Ruiz, R. A., Talekar, N. S. & Vacek, D. (1996) Foreign exploration for natural enemies of Bemisia tabaci from Southeast Asia. Subtropical Plant Science 48: 48-53. Legaspi, Jr, B. C., Legaspi, J. C., Carruthers, R. I., Goolsby, J. A., Hadman, J., Jones, W., Murden, D. & Wendel, L. E. (1997) Areawide population dynamics of silverleaf whitefly (Homoptera: Aleyrodidae) and its parasitoids in the Lower Rio Grande Valley of Texas. Journal Entomological Science 32: 445-459. Moomaw, C. (1995) Survey of the indigenous parasitoids of Bemisia tabaci in the Lower Rio Grande Valley of Texas. Masters Thesis, Department of Entomology, Texas A&M University, College Station. Pickett, C. H., Simmons, G. S., Goolsby, J. A. & Overholt, D. (1999) Fall releases of parasites into citrus. p. 83. In Henneberry, T. J., Toscano, N. C., Perring, T. M. & Faust, R. M. (Eds), Silverleaf Whitefly: National Research, Action and Technology Transfer Plan, 1997-2001: Second Annual Review. United States Department of Agriculture ARS 1999-01, Albuquerque. Riley, D. G. & Ciomperlik, M. A. (1997) Regional population dynamics of whitefly (Homoptera: Aleyrodidae) and associated parasitoids (Hymenoptera: Aphelinidae). Environmental Entomology 26: 1049-1055. Rose, M. & Zolnerowich, G. (1997) Eretmocerus Haldeman (Hymenoptera: Aphelinidae) in the United States, with descriptions of new species attacking Bemisia (tabaci complex) (Homoptera: Aleyrodidae). Proceedings of the Entomological Society of Washington 99: 1-27. Rose, M. & Zolnerowich, G. (1998) Eretmocerus Haldeman (Hymenoptera: Aphelinidae) imported and released in the United States for control of Bemisia (tabaci complex) (Homoptera: Aleyrodidae). Proceedings of the Entomological Society of Washington 100: 31-323. Vacek, D. C., Ruiz, R. A. & Wendel, L. E. (1996) RAPD-PCR identification of natural enemies of SPWF. p. 139. In Henneberry, T. J., Toscano, N. C., Perring, T. M. & Faust, R. M. (Eds), Silverleaf Whitefly: National Research, Action and Technology Transfer Plan, 1997-2001: Fourth Annual Review. United States Department of Agriculture ARS 1996-01, San Antonio.
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci 355
Appendix 1 Parasitic Hymenoptera imported into the USA for biological control of Bemisia tabaci (Biotype ‘B’), 1992 to 1998.
M#
DNA Pattern
Collection Locality
Encarsia sp. nr. strenua
M92018
EN- 1
Encarsia sp. nr. strenua
M93010
Encarsia formosa
Name
Collector
Date
Tax Id
Host
Host Plant
Biology
India, Parbhani Nguyen
I-92
Woolley & Schauff
Bemisia tabaci complex
Autoparasitoid
EN- 1
India, Parbhani Nguyen
I-92
Woolley & Schauff
Bemisia tabaci complex
Autoparasitoid
M92017
EN-2
Greece, Angelohori
Kashefi
I-92
Woolley & Schauff
Trialeurodes vaporarium
Bean
Uniparental
Encarsia formosa
M92030
EN-2
Egypt, Nile Delta
Kirk & Lacey
I-92
Schauff
Bemisia tabaci complex
Lantana
Uniparental
Encarsia formosa
M94051
EN-2
Thailand, Saen Kirk & Lacey
III-94
Woolley & Schauff
Bemisia tabaci complex
Snakeweed
Uniparental
Encarsia formosa
M94089
EN-2
Italy, Borgo Corso
Kirk & Campobasso
IX-94
Woolley & Schauff
Trialeurodes vaporarium
Tomato
Uniparental
Encarsia transvena
M94017
EN-3
Taiwan, ShanHua
Legaspi, Carr., Popr.
III-94
Woolley & Johnson
Bemisia tabaci complex
Soybean/ Tomato
Autoparasitoid
Encarsia transvena
M94019
EN-4
Taiwan, ShanHua
Legaspi, Carr., Popr.
III-94
Woolley & Johnson
Bemisia tabaci complex
Soybean/ Tomato
Autoparasitoid
Encarsia transvena
M94041
EN-5
Thailand, Chiang Mai
Kirk & Lacey
III-94
Woolley & Johnson
Bemisia tabaci complex
Poinsettia
Autoparasitoid
Encarsia transvena
M94047
EN-5
Malaysia, Kuala Lumpur
Kirk & Lacey
III-94
Woolley & Johnson
Bemisia tabaci complex
Mussaenda sp. Autoparasitoid
Encarsia transvena
M95023
EN-5
Thailand, Doi Suthep
Carruthers & Legaspi
V-95
Woolley & Johnson
Bemisia tabaci complex
uknown woody plant
Autoparasitoid
Encarsia adrianae
M94024
EN-6
Thailand, Kirk & Lacey Kampang Saen
III-94
Woolley & Johnson
Bemisia tabaci complex
Snakeweed
Autoparasitoid
Encarsia transvena
M93003
EN-7
Spain, Murcia
I-93
Woolley & Schauff
Bemisia tabaci complex
Lantana
Autoparasitoid
Encarsia spp.
Kirk & Lacey
J. A. Goolsby et al. 356
Hymenoptera: Evolution, Biodiversity and Biological Control
Name
M#
DNA Pattern
Collection Locality
Encarsia transvena
M95107
EN-5
Pakistan, Multan
Encarsia transvena
M96065
EN-5
Encarsia lutea
M93064
Encarsia lutea
Collector
Date
Tax Id
Host
Host Plant
Biology
Kirk & Lacey
XI-95
Goolsby
Bemisia tabaci complex
Cotton
Autoparasitoid
Pakistan, Jalari Kirk
X-96
Goolsby
Bemisia tabaci complex
Cotton
Autoparasitoid
EN-10
Cyprus, Mazotos
Kirk & Lacey
I-93
Woolley & Johnson
Bemisia tabaci complex
Lantana
Autoparasitoid
M94096
EN-10
Italy, Testa Di Lespe
Kirk & Campobasso
IX-94
Woolley & Johnson
Bemisia tabaci complex
Eggplant
Autoparasitoid
Encarsia lutea
M94107
EN-10
Israel, Givat Haim
Kirk & Lacey
X-94
Woolley & Johnson
Bemisia tabaci complex
Cotton
Autoparasitoid
Encarsia lutea
M94115
EN-10
Israel, Ein Gedi, Kirk & Lacey Dead Sea
X-94
Woolley & Johnson
Bemisia tabaci complex
Lantana
Autoparasitoid
Encarsia lutea
M94129
EN-10
Spain, Mazarron Casas Nuevas
Kirk & Lacey
XI-94
Woolley & Johnson
Bemisia tabaci complex
Ipomea sp.
Autoparasitoid
Encarsia lutea
M96044
EN-10
Sicily, Ragusa
Kirk & Campobasso
IX-96
Johnson
Bemisia tabaci complex
Solanaceous weed
Autoparasitoid
Encarsia transvena
M94014
EN-11
Philippines, Benguet
Legaspi, Carr., Popr.
III-94
Woolley & Johnson
Trialeurodes sp. White potato
Autoparasitoid
Encarsia transvena
M94016
EN-11
Taiwan, ShanHua
Legaspi, Carr., Popr.
III-94
Woolley & Johnson
Bemisia tabaci complex
Poinsettia
Autoparasitoid
Encarsia nr. pergandiella
M94055
EN-15
Brazil, Sete Lagoas
Rose
II-94
Rose & Woolley
Bemisia tabaci complex
Poinsettia/ Soybean
Uniparental
Encarsia nr. hispida
M94056
EN-16
Brazil, Sete Lagoas
Rose
II-94
Rose & Woolley
Bemisia tabaci complex
Poinsettia/ Soybean
Uniparental
Encarsia sp. M95001 (parvella group)
EN-18
Dominican Ciomperlik Republic, Azua
I-95
Schauff
Bemisia tabaci complex
Tomato
Autoparasitoid
ERET-1
Spain, Murcia
I-92
Schauff
Bemisia tabaci complex
Cotton
Biparental
Eretmocerus spp. Eretmocerus mundus
M92014
Kirk , Chen & Sobhain
Predictive and Empirical Evaluation for Parasitoids of Bemisia tabaci 357
Eretmocerus mundus
M92019
ERET-1
India, Padappai Kirk & Lacey
I-92
Rose & Zolnerowich
Bemisia tabaci complex
Eggplant
Biparental
Eretmocerus mundus
M92027
ERET-1
Egypt, Cairo
Kirk & Lacey
I-92
Rose & Zolnerowich
Bemisia tabaci complex
Lantana
Biparental
Eretmocerus mundus
M93004
ERET-1
Spain, Murcia
Kirk & Lacey
I-93
Woolley & Schauff
Bemisia tabaci complex
Sonchus
Biparental
Eretmocerus mundus
M93058
ERET-1
Taiwan, Tainan Moomaw
XII-93
Rose & Zolnerowich
Bemisia tabaci complex
Tomato
Biparental
Eretmocerus mundus
M94085
ERET-1
Italy, Frascati
Kirk & Campobasso
IX-94
Rose & Zolnerowich
Bemisia tabaci complex
Hibiscus
Biparental
Eretmocerus mundus
M94092
ERET-1
Italy, Castel Gondolfo
Kirk & Campobasso
IX-94
Rose & Zolnerowich
Bemisia tabaci complex
Ipomea sp.
Biparental
Eretmocerus mundus
M94097
ERET-1
Italy, Testa Di Lespe
Kirk & Campobasso
IX-94
Rose & Zolnerowich
Bemisia tabaci complex
Eggplant
Biparental
Eretmocerus mundus
M94103
ERET-1
Israel, Gat
Kirk & Lacey
X-94
Rose & Zolnerowich
Bemisia tabaci complex
Kohlrabi
Biparental
Eretmocerus mundus
M94105
ERET-1
Israel, Gat
Kirk & Lacey
X-94
Rose & Zolnerowich
Bemisia tabaci complex
Sonchus sp.
Biparental
Eretmocerus mundus
M94125
ERET-1
Israel, Golan Kibutz
Kirk & Lacey
X-94
Rose & Zolnerowich
Bemisia tabaci complex
Euphorbia spp. Biparental
Eretmocerus mundus
M94120
ERET-1
Israel, Golan Kirk & Lacey Ma’Aleh Gamla
X-94
Rose & Zolnerowich
Bemisia tabaci complex
Melons
Biparental
Eretmocerus mundus
M94124
ERET-1
Israel, Negev Desert
Kirk & Lacey
X-94
Rose & Zolnerowich
Bemisia tabaci complex
Cucumber
Biparental
Eretmocerus mundus
M96028
ERET-1
Sicily, Santa Groce
Kirk & Campobasso
IX-96
Goolsby
Bemisia tabaci complex
Eggplant
Biparental
Eretmocerus mundus
M97046
ERET-1
Cyprus, Nicosia Kirk
VII-97
Goolsby
Bemisia tabaci complex
Lantana
Biparental
Eretmocerus hayati
M93005
ERET-2
India, Thirumala
Kirk & Lacey
I-93
Rose & Zolnerowich
Bemisia tabaci complex
Eretmocerus melanoscutus
M94036
ERET-3
Thailand, Chiang Mai
Kirk & Lacey
III-94
Rose & Zolnerowich
Bemisia tabaci complex
Biparental Chromolaena odorata
Biparental
J. A. Goolsby et al. 358
Hymenoptera: Evolution, Biodiversity and Biological Control
Name
M#
DNA Pattern
Collection Locality
Date
Tax Id
Host
Host Plant
Biology
Eretmocerus melanoscutus
M94040
ERET-3
Thailand, Kirk & Lacey Kampang Saen
III-94
Rose & Zolnerowich
Bemisia tabaci complex
Cotton
Biparental
Eretmocerus melanoscutus
M94023
ERET-8
Thailand, Sai Noi Klong Ha Roi
III-94
Rose & Zolnerowich
Bemisia tabaci complex
Eggplant/ melon
Biparental/ Uniparental
Eretmocerus melanoscutus
M95097
ERET-3
Taiwan, Tainan Talekar & Jones X-95
Rose & Zolnerowich
Bemisia tabaci complex
Tomato
Biparental
Eretmocerus eremicus
M94001
ERET-4
Brawley, CA
Hoelmer
I-94
Rose & Zolnerowich
Bemisia tabaci complex
Okra
Biparental
Eretmocerus staufferi
M94002
ERET-5
College Station, TX
Stauffer
I-94
Rose & Zolnerowich
Bemisia tabaci complex
Tomato
Uniparental
Eretmocerus tejanus
M94003
ERET-6
USA, Mission, TX
Rodriquez
I-94
Rose & Zolnerowich
Bemisia tabaci complex
Cabbage
Biparental
Eretmocerus hayati
M95012
ERET-10 Pakistan, Multan
Kirk & Akey
IV-95
Rose & Zolnerowich
Bemisia tabaci complex
Eggplant
Biparental
Eretmocerus hayati
M95105
ERET-10 Pakistan, Multan
Kirk & Lacey
IX-95
Rose & Zolnerowich
Bemisia tabaci complex
Eggplant
Eretmocerus hayati
M96064
ERET-10 Pakistan, Jalari Kirk
X-96
Goolsby
Bemisia tabaci complex
Cotton
Biparental
Collector
Kirk & Lacey
Eretmocerus sp. M95098 nr. furuhashii
ERET-11 Taiwan, Tainan Talekar & Jones X-95
Rose & Zolnerowich
Bemisia tabaci complex
Tomato
Biparental
Eretmocerus sp. M95026 nr. furuhashii
ERET-11 Taiwan, Chiuju Kirk
V-94
Goolsby
Bemisia tabaci complex
Cabbage
Biparental
Eretmocerus emiratus
ERET-12 United Arab Emirates
Porter, Romadon
XI-95
Rose & Zolnerowich
Bemisia tabaci complex
Okra
Biparental
ERET-13 Ethiopia, Melka Werer
Gerling, Terefe XI-96
Goolsby
Bemisia tabaci complex
Cotton
Biparental
M95104
Eretmocerus sp. M96076
Which Factors Govern the Host Preference of Aphid Parasitoids When Offered Host Races of an Aphid Species? Anja Hildebrands1, Thomas Thieme2 and Stefan Vidal3 1
Institute for Plant Diseases and Plant Protection, University, Herrenhaeuser Str. 2, 30419 Hanover, Germany 2
BTL Biotestlabor GmbH, 18184 Sagerheide, Germany
3
Institute for Plant Pathology, University, Grisebachstr. 6, 37077 Goettingen, Germany (email:
[email protected])
Introduction The evolution of complex life histories in aphids, which often involve polymorphism, is regarded as a prerequisite for the success of several aphid species in exploiting variable environments. Polymorphism within aphid species may refer to morph differentiation during different phases of their life cycle, colour polymorphism, or the ability to use different host plants (e.g. Hille Ris Lambers 1966; Moran 1992; Dixon 1996). This latter feature of aphids favours the formation of subspecies, biotypes or races, a phenomenon observed in several species (Eastop 1973). However, the transition of an aphid species to new host plants by the formation of demes may not only be advantageous with respect to the exploitation of new resources, but also with respect to lowered parasitisation of aphid populations by aphidiine braconid wasps. These wasps are able to use, beside other traits, visual or chemical cues emitted by the host plants for host location (Grasswitz & Paine 1993; Mackauer et al. 1996). Thus, new aphid/host plant associations may not fit within the searching templates of the female wasp and release the aphids, at least in part, from parasitisation pressure. Moreover, by physiologically adapting to new host plants, aphids could become less suitable for the development of parasitoid larvae. The black bean aphid Aphis fabae Scopoli is known to comprise several taxa which are barely distinguishable by morphological characters, but differ in their preferred host plants. Müller (1982) elaborated a simple biological test for the differentiation of A. fabae fabae (AFF), A. f. solanella Theobald (AFS), A. f. evonymi F. (AFE) and A. f. cirsiiacanthoidis Scopoli (AFC) by their performance on the marking of host plants Vicia faba, Solanum nigrum, Euonymus europaeus and Cirsium arvense, respectively. Thieme (1988) extended this biological test to include A. f. mordwilkowi Börner & Janisch (AFM; on Tropaeolum majus) and A. f. armata Hausmann (AFA; on Digitalis purpurea). Additionally, we used the closely related species A. rumicis L. (AR; on Rumex obtusifolius) to include a host plant which is suitable for all taxa in this context. One of the aims of the study reported here was to test the hypothesis that the morphological, physiological and/or behavioural variability in members of the black bean aphid-complex (Aphis fabae-Aggr.) affects the host acceptance behaviour of aphidiine parasitoids. We used two polyphagous species of the genus Lysiphlebus, L. fabarum (Marshall), a common parasitoid of the black bean aphid on agricultural crops (Völkl & Stechmann 1998), native to Europe and L. testaceipes Cresson. The latter species has been introduced to Europe in recent times (Stary et al. 1988) and is therefore not adapted to the aphid taxa tested. We compared the effects of visual cues of
359
Anja Hildebrands, Thomas Thieme and Stefan Vidal 360
Hymenoptera: Evolution, Biodiversity and Biological Control
aphid movement on the behaviour of the two parasitoid species. Additionally, we tested whether host preference was related to larval performance.
Material and Methods Insect cultures Laboratory stocks of AFA, AFC, AFE, AFM and AR, which originated from material cultured by Thieme in Rostock, and AFF and AFS, which originated from taxa collected at Hanover, were maintained either on their potted marking host plants or on potted Rumex crispus in wooden framed cages closed with plexiglas on top and fine-meshed gauze on the sides. They were kept at 21 ± 1˚C under continuous light conditions in a greenhouse. The solitary parasitoids L. testaceipes and L. fabarum were reared on the nominal species AFF under the same conditions as the aphid stocks. Lysiphlebus testaceipes originated from material of ÖRE-Bioprotect, Raisdorf, Germany. Lysiphlebus fabarum was initially reared from A. fabae mummies collected at Papenburg, Lower Saxonia. In all experiments we used mated female wasps that had no previous experience with the hosts. Host acceptance We investigated host examination and attack behaviour in experiments that were designed as dichotomous choice tests. In order to evaluate the influence of visual cues and host movement on host acceptance, experiments were repeated under dark conditions and with anaesthetised aphids, following the methodology of Michaud and Mackauer (1994). Each experiment was repeated ten times. In experiment 1 single female wasps were provided with 10 L3 nymphs of each of two host taxa in a petri-dish (5.7 cm) for 20 min (L. testaceipes) or 30 min (L. fabarum). During this time we recorded four patterns of behaviour: a) examination of the host by antennation without subsequent ovipositor contact; b) probing with the ovipositor without oviposition; c) oviposition, and d) no contact with the host. Because probing and actual oviposition could not be separated by visual inspection alone, attacked aphids were removed from petri-dishes, reared on R. crispus for 4 d and then dissected to check for parasitoid larvae. To separate aphid taxa in the dichotomous combinations, one tarsus was amputated from an aphid ahead of the experiments, alternating between the two taxa tested within the replications. In experiment 2 (host acceptance in the dark), single female wasps were provided with 10 L3nymphs of each of two host taxa for 4 h in a petri-dish in complete darkness. The experiment was set between 9.00 a.m. and 2.00 p.m., when wasps experienced their highest activity. Aphids were reared on R. crispus for 4 d and were then dissected to check for parasitoid larvae. Experiment 3 (host acceptance of immobilised aphids) was designed as in experiment 1. However, in this experiment aphids were anaesthetised by a 5 min exposure to CO2. This treatment kept the instars immobilised for about 25 min.
Host suitability One hundred larvae of each aphid taxon which had been previously attacked by L. testaceipes were maintained on R. crispus plants in separate cages. To assess the current developmental stage of the parasitoid larvae, five aphid larvae of each taxon were dissected during the following days until adult eclosion. Larval stages were determined by the characters given in Couchman & King (1977).
Which Factors Govern the Host Preference of Aphid Parasitoids? 361
L. testaceipes
L. fababarum
L. testac. on Rumex
9
Oviposition Index
8 7 6 5 4 3 2 1 0 AFC AFC AFC AFC AFF AFF AFF AFF AFM AFM AFM AFM AFA AFE AR AFS AFA AFE AR AFS AFA AFE AR AFS
Aphid Combination Figure 1
Oviposition indices (higher oviposition rate/lower oviposition rate) of Lysiphlebus testaceipes (black columns), L. fabarum (grey columns) and L. testaceipes with all aphid taxa reared on Rumex crispus (white columns) for aphid combinations with significant preference for one taxon under daylight conditions. Aphid taxa were reared on their marking host plants. Names of preferred aphid taxa are bold and underlined. AFA = Aphis fabae armata; AFC = A. f. cirsiiacanthoidis; AFE = A. f. evonymi; AFF = A. f. fabae; AFM = A. f. mordwilkowi; AR = A. rumicis; AFS = A. f. solanella. All indices shown are significant at P < 0.05 (G-test with Williams correction).
Statistical analysis The G-test for goodness-of-fit with Williams’ correction (Sokal & Rohlf 1995) was used to determine significant differences between attack rates or oviposition rates in single dichotomous choice tests. The Man-Whitney U-Test was used to compare the development time of parasitoid larvae in different aphid taxa, and mortality rates in the larval stages were compared using ANOVA and a Tukey-Test.
Results Host acceptance Both L. fabarum and L. testaceipes consistently preferred parasitising AFC, AFF and AFM in comparison to AFE, AFA, AR and AFS. However, the oviposition indices (the ratio of higher to lower oviposition rates) were higher in all but one case for L. fabarum (Fig. 1). No preferences were found when pairwise combinations of the three former or the four latter aphid taxa were offered. In this experiment all aphid taxa were reared on their marking host plants. Therefore, the observed
Anja Hildebrands, Thomas Thieme and Stefan Vidal 362
Hymenoptera: Evolution, Biodiversity and Biological Control
9 L. testaceipes
L. fabarum
8
Oviposition Index
7 6 5 4 3 2 1 0 AFC AFC AFC AFC AFF AFF AFF AFF AFM AFM AFM AFM AFA AFE AR AFS AFA AFE AR AFS AFA AFE AR AFS
Aphid Combination Figure 2
Oviposition indices of Lysiphlebus testaceipes and L. fabarum for aphid combinations with significant preference for one taxon under dark conditions (see Figure 1 for further explanation).
differences in preference could have been due to an influence of some nutritional factors of the marking host plant alone. We repeated the experiment using L. testaceipes and aphid taxa which had been reared on Rumex crispus, a plant suitable for all taxa of the A. fabae-complex. Again, L. testaceipes significantly preferred the same aphid taxa as in the first experiment, the oviposition indices being significant at least at P < 0.01 (G-test with Williams correction). The preference of these parasitoids for AFC, AFF and AFM in comparison to AFA, AFE, AR and AFS corresponded with differences in the colour of these two groups, the former appearing greenish-black, the latter appearing brownish-black when reared under greenhouse conditions (elimination of UV-radiation). To test for the effects of visual cues, especially of the colour of the aphid integument, on the preference of parasitoids for particular aphid taxa, the former experiment was repeated under conditions of complete darkness. Again, both L. testaceipes and L. fabarum consistently showed the same ovipositional preferences as in the first experiment (Fig. 2). All oviposition indices for the aphid taxa AFC, AFF and AFM were significantly higher in comparison to the taxa AFE, AFA, AR, or AFS. However, especially in L. fabarum, most indices were less pronounced than in the first experiment, but still significant at P < 0.01 (G-test with Williams correction). These results suggest that visual cues may have influenced host recognition but did not affect host acceptance. The acceptance of a potential aphid taxon for oviposition may also be affected by the defensive behaviour of the host. In experiment 3 we therefore tested host acceptance of the two parasitoid
Which Factors Govern the Host Preference of Aphid Parasitoids? 363
9 L. testaceipes
L. fabarum
8
Oviposition Index
7 6 5 4 3 2 1 0 AFC AFC AFC AFC AFF AFF AFF AFF AFM AFM AFM AFM AFA AFE AR AFS AFA AFE AR AFS AFA AFE AR AFS
Aphid Combination Figure 3
Oviposition indices of Lysiphlebus testaceipes and L. fabarum for aphid combinations with significant preference for one taxon when aphids were anaesthetised (see Figure 1 for further explanation).
species using immobilised aphids. Again both parasitoid species preferred to lay their eggs in the aphid taxa AFC, AFF and AFM in comparison to AFA, AFE, AR and AFS, resulting in significant oviposition indices for the same taxon combinations as in experiment 1 (Fig. 3). Moreover, immobilisation of aphid larvae increased oviposition in the preferred taxa, resulting in oviposition indices that were more pronounced than in experiment 1 (compare scaling of the vertical axis in Figs 1–3). However, when attack rates (contact with the ovipositor) of female parasitoids were examined when aphids were immobilised, we found no significant differences (G-test with Williams’ correction, P > 0.05; Fig. 4). These results indicate that the two parasitoid species responded to physiological differences within the aphid taxa after contact with their ovipositor rather than to differences in their defensive behaviour.
Host suitability In experiment 4 we examined whether the differences in the preferences for certain aphid taxa shown by L. testaceipes in experiments 1–3 corresponded with a different suitability of these aphids for the development of the parasitoid larvae. Suitability of a host as an oviposition site is considered to be an important aspect of host selection by female parasitoids (Vinson 1976). The larvae of L. testaceipes developed much faster in the preferred taxa AFC, AFF and AFM, corroborating the oviposition preferences in the first experiments (Fig. 5). The total development times from the egg stage to eclosion of the adults was significantly longer in AFA ( 19.75 d), AFE (15.47 d) and AR (19.0 d) than in AFF (12.53 d), AFC (12.25 d) and AFM (14.04 d)
Anja Hildebrands, Thomas Thieme and Stefan Vidal 364
Hymenoptera: Evolution, Biodiversity and Biological Control
2,0 L. testaceipes
L. fabarum
Attack Index
1,5
1,0
0,5
0,0 AFC AFC AFC AFC AFF AFF AFF AFF AFM AFM AFM AFM AFA AFE AR AFS AFA AFE AR AFS AFA AFE AR AFS
Aphid Combination Figure 4
Attack indices (higher attack rate/lower attack rate) of Lysiphlebus testaceipes and L. fabarum for aphid combinations with significant preference for one taxon in the former experiments when aphids were anaesthetised. Attack rates were not significantly different from 1 (G-test with Williams correction; P > 0.05) (see Figure 1 for further explanation).
(Mann-Whitney U-test, P < 0.001). The total developmental time for AS was somewhat, but not significantly, longer (16.0 d) than in the three most preferred taxa. Moreover, the mortality rate during the larval stages paralleled the developmental time (Fig. 6). The parasitoid larvae experienced a significantly lower mortality rate in the three taxa AFC, AFF and AFM compared to the taxa AFA and AR (Tukey-Test, P < 0.05).
Discussion Studies on host-parasitoid interactions often concentrate on the question of why some insects are accepted for oviposition and others ignored or rejected (Mackauer et al. 1996). Host acceptance by female parasitoids is governed by the quality, and host suitability by the performance of larvae (Michaud & Mackauer 1994). By using an aphid species which is known to have distinct biotypes or host races adapted to certain host plants, we tested the consequences of intraspecific host plant adaptation on the preference and performance of natural enemies. In our study the two aphidiine parasitoids L. testaceipes and L. fabarum consistently preferred to oviposit in the aphids A. f. fabae, A. f. cirsiiacanthoidis and A. f. mordwilkowi compared with A. f. armata, A. f. evonymi, A. f. solanella and A. rumicis. These preferences were not due to the colour or the defensive behaviour of the aphids nor to influences mediated by the nutritional quality of the host plant. The preferred aphid taxa were more suitable for growth of the parasitoid larvae, resulting in significantly
Which Factors Govern the Host Preference of Aphid Parasitoids? 365
I
Developmental Stage
M L4 AFF AFC AFM AFS AFE AR AFA
L3 L2 L1 E 0
2
4
6
8
10
12
14
16
18
20
Days after Oviposition Figure 5
Total developmental time from oviposition to eclosion of adults for Lysiphlebus testaceipes in taxa of the Aphis fabae-complex. For the larval stages means of five dissected aphids per taxon and day are shown; for the mummy stage and the adult eclosion means of 10 mummified aphids are shown.
shorter total developmental times than in the rejected taxa. Moreover, the parasitoid larvae experienced a higher mortality when developing in the rejected taxa. Visual characteristics of host size, shape or colour are important cues for host location by aphidiine parasitoids. For example, Ankersmit et al. (1986) found that Aphidius rhopalosiphi DeStefani parasitised green coloured morphs of Sitobion avenae F. more successfully than brown morphs. Michaud and Mackauer (1994, 1995) showed that several Aphidius spp., Praon pequodorum Viereck and Monoctonus paulensis (Ashmead) used visual cues for the evaluation of the aphid hosts Macrosiphon creelii Davis (pink and green morphs) and Acyrthosiphon pisum (Harris) (green morph). Some of these parasitoid species changed their attack and host acceptance behaviour when aphids were offered under dark conditions (Michaud & Mackauer 1995), indicating that visual cues are important but not the only host signals, that are used by female parasitoids. If visual cues were responsible for the two Lysiphlebus species tested to differentiate between the members of the A. fabae-complex, we would have expected to find differences in oviposition indices when the aphids were offered under dark conditions. However, the preferences for oviposition in certain taxa were the same whether in light or dark conditions. The acceptance of a potential host may be related to its defensive behaviour such as kicking, jerking or walking during attack by a parasitoid (Klingauf 1967; Gardner et al. 1984; Völkl 1991; Weisser 1994; Michaud & Mackauer 1994, 1995). By anaesthetising aphid nymphs we were able to show that this host manipulation did not result in differences in the preference ranking of the
Anja Hildebrands, Thomas Thieme and Stefan Vidal 366
Hymenoptera: Evolution, Biodiversity and Biological Control
60
a
a
Mortality (%)
50 40 30
b
b
20
c
c
c
10
15
82
51
89
59
48
88
AFA
AFC
AFE
AFF
AFM
AR
AFS
0
Aphid Taxon Figure 6
Mortality rate of Lysiphlebus testaceipes in different taxa of the Aphis fabae-complex during their larval development. Numbers in columns give the numbers of parasitised aphids which could be used for analysis. Different letters above columns denote significant differences between the taxa (Tukey-Test, P < 0.05).
aphid taxa compared to normal host behaviour. Rather, an even more pronounced oviposition index was observed in both Lysiphlebus species in the most preferred aphid taxa. Olfactory cues from honeydew and cuticular lipids, originating from the host plants on which the aphids had been cultured, can have a stimulating effect on attack behaviour by aphidiines (Bouchard & Cloutier 1985; Sheehan & Shelton 1989; Budenberg 1990; Grasswitz & Paine 1992,1993; Battaglia et al. 1993; Liepert & Dettner 1993; Pennacchio et al. 1994). Lysiphlebus testaceipes may also perceive kairomones from cornicle waxes that stimulate oviposition during attack (Grasswitz & Paine 1992). Cornicle waxes are a mixture of wax and haemolymph (Fröhlich 1991) and contain species-specific mixtures of triglycerides (Callow et al. 1973; Greenway & Griffith 1973). However, if one or a combination of the above parameters had triggered the attack behaviour of L. testaceipes or L. fabarum, we would have expected to find different attack rates when the defensive behaviour of the aphid taxa was excluded by anaesthetising them. Contrary to our expectations, the attack indices were still all positive but not significantly different for the preferred taxa, indicating an inherent preference related to one or more factors discussed above. Moreover, these findings emphasise the importance of internal factors of the aphid taxa tested, which cannot be perceived by visual inspection or antennation alone, but may become functional when the ovipositor is inserted into the host. The ovipositor of female Aphidiinae bears numerous chemosensilla, which could serve as a tool to evaluate the chemical quality of the potential hosts. The differences in host acceptance may thus be due to different kairomones in the haemolymph of the host. Srivastava and Singh (1988) showed a stimulating effect of host haemolymph on ovi-
Which Factors Govern the Host Preference of Aphid Parasitoids? 367
position behaviour of Trioxys indicus Subba Rao & Sharma on A. carccivora Koch. This effect was related to water soluble kairomones in the haemolymph. Kainoh and Brown (1994) reported the haemolymph of the larvae of Trichoplusia ni (Hübner) (Noctuidae) as an inducer of oviposition in Chelonus sp. near curvimaculatus (Cameron) (Braconidae). Artificial eggs, which were filled with 11 different amino acids, induced oviposition especially when they contained arginine, histidine and lysine. However, the highest ovipositional response was achieved, when eggs containing host haemolymph were provided, suggesting that specific mixtures of amino acids are responsible for host acceptance in this parasitoid. Biotypes of A. pisum differ in the structure of their symbionts (Srivastava 1987) resulting in the ability to synthesise different amino acids. Recent findings by Adams and Douglas (1997) suggest that this is also the case in the A. fabaecomplex, emphasising the potential role of amino acid composition in host preference behaviour of L. fabarum and L. testaceipes. The quantity of resources available for larval koinobionts is dependent on the growth and nutrient uptake of the host after its parasitisation (Mackauer 1986; Sequeira & Mackauer 1992; Harvey et al. 1995). The total time of development of L. testaceipes in A. f. fabae, A. f. cirsiiacanthoidis and A. f. mordwilkowi (12-14 d) is comparable to that reported for other aphid host species (Hight et al. 1972; Salto et al. 1983; van Steenis 1994), suggesting these three taxa are optimal hosts for this aphidiine parasitoid. In addition, the higher mortality rates experienced by parasitoid larvae in A. f. armata, A. f. evonymi and A. rumicis compared with A. f. cirsiiacanthoidis, A. f. fabae and A. f. mordwilkowi suggest that the nutritional quality of these hosts is low, since physiological defensive reactions like encapsulation could not be detected during dissections. The duration of development depends on either the nutritional quality (Sequeira & Mackauer 1992) or the quantity of resources (Wilbert 1965; Elliot et al. 1994; Harvey et al. 1994). The quantity of resources in aphids may vary according to size and, because size may be associated with oviposition preferences, we excluded host plant influences by rearing all taxa on the same host plant R. crispus. This plant species represents an alternative host for most taxa in the A. fabae-complex, whereas it is the preferred host plant of A. rumicis. However, although A. rumicis was larger in size than the other aphid taxa, the larger amount of resources in this species did not positively influence larval development of L. testaceipes. In contrast, total time of development lasted about 5 d longer than in the most preferred taxa. Therefore, quantity of resources can be ruled out as responsible for the observed differences in the development of larvae of L. testaceipes. In conclusion, patterns of host acceptance and host suitability of different members of the A. fabae-complex for L. testaceipes and L. fabarum are related to differences in the physiology of these hosts and not to visual or external chemical cues. Host plant shifts in aphid taxa may be accompanied by different physiological states. Aphidiine parasitoids which are adapted to particular aphid hosts may not be able to locate the same hosts on new host plants, or immediately adapt to the different internal host environment . If this assumption is true, not only for the parasitoid species tested in our study, but also for other species normally encountered in colonies of A. fabae, the ability to use new host plants may provide these aphid taxa with an enemy free space (Jeffries & Lawton 1984).
Acknowledgements We thank Dr. A. F. G. Dixon and an anonymous reviewer for comments on an earlier draft of this paper. A. H. was funded by the Deutsche Forschungsgemeinschaft (Vi 117/4-1) and S. V. by a travel grant (Vi 117/7-1) to Canberra.
Anja Hildebrands, Thomas Thieme and Stefan Vidal 368
Hymenoptera: Evolution, Biodiversity and Biological Control
References Adams, D. & Douglas, A. E. (1997) How symbiotic bacteria influence plant utilisation by the polyphagous aphid, Aphis fabae. Oecologia 110: 528-532. Ankersmit, G. W., Bell C., Dijkman H., Mace N., Rietstra S., Schröder J. & de Visser C. (1986) Incidence of parasitism by Aphidius rhopalosiphi in colour forms of the aphid Sitobion avenae. Entomologica Experimentalis et Applicata 40: 223-229. Battaglia, D., Pennacchio F., Marincola, G. & Tranfaglia, A. (1993) Cornicle secretion of Acyrthosiphon pisum (Homoptera: Aphididae) as a contact kairomone for the parasitoid Aphidius ervi (Hymenoptera: Braconidae). European Journal of Entomology 90: 423- 428. Bouchard, Y. & Cloutier C. (1985) Role of olfaction in host finding by aphid parasitoid Aphidius nigripes (Hymenoptera: Aphidiidae). Journal of Chemical Ecology 11: 801-808. Budenberg, W. J. (1990) Honeydew as a contact kairomone for aphid parasitoids. Entomologica Experimentalis et Applicata 55: 139-148. Callow, R. K., Greenway, A. R. & Griffith, D. C. (1973) Chemistry of the secretion from the cornicles of various species of aphids. Journal of Insect Physiology 19: 737-748. Couchman, J. R. & King, P. E. (1977) Morphology of the larval stages of Diaeretiella rapae (M’Intosh) (Hymenoptera: Aphidiidae). International Journal of Insect Morphology & Embryology 6: 127-136. Dixon, A. F. G. (1996) Aphid Ecology. Chapman & Hall, London. Elliot, N. C., French, B. W., Reed, D. K., Burd, J. D. & Kindler, S. D. (1994) Host species effects on parasitization by a syrian population of Diaeretiella rapae M’Intosh (Hymenoptera: Aphidiidae). The Canadian Entomologist 126: 1515-1517. Eastop, V. F. (1973) Deductions from the present day host plants of aphids and related insects. Symposium of the Royal entomological Society of London 6: 157-178. Fröhlich, G. (1991) Phytopathologie und Pflanzenschutz. 2. überarbeitete Auflage. Gustav Fischer, Jena. Gardner, S. M., Ward, S. A. & Dixon, A. F. G. (1984) Limitation of superparasitism by Aphidius rhopalosiphi: a consequence of aphid defense behaviour. Ecological Entomology 9: 149-155. Grasswitz, T. R. & Paine, T. D. (1992) Kairomonal effect of an aphid cornicle secretion on Lysiphlebus testaceipes. Journal of Insect Behavior 5: 447-457. Grasswitz, T. R. & Paine, T. D. (1993) Influence of physiological state and experience on the responsiveness of Lysiphlebus testaceipes (Cresson) (Hymenoptera, Aphidiidae) to aphid honeydew and to host plants. Journal of Insect Behavior 6: 511-528. Greenway, A. R. & Griffith, D. C. (1973) A comparison of triglycerides from aphids and their cornicle secretions. Journal of Insect Physiology 19: 1649-1655. Harvey, J. A., Harvey J. F. & Thompson, D. J. (1994) Flexible larval growth allows use of a range of host sizes by parasitoid wasps. Ecology 75: 1420-1428. Harvey, J. A., Harvey, J. F. & Thompson, D. J. (1995) The effect of host nutrition on growth and development of the parasitoid wasp Venturia canescens. Entomologica Experimentalis et Applicata 75: 213-220. Hight, S. C., Eikenbary, R. D., Miller, R. J. &. Starks, K. J. (1972) The greenbug and Lysiphlebus testaceipes. Environmental Entomology 1: 205-209. Hille Ris Lambers, D. (1966) Polymorphism in Aphididae. Annual Review of Entomology 11: 47-78. Jeffries, M. J. & Lawton J. H. (1984) Enemy free space and the structure of ecological communities. Biological Journal of the Linnean Society 23: 269-286.
Which Factors Govern the Host Preference of Aphid Parasitoids? 369
Kainoh, Y. & Brown J. J. (1994) Amino acids as oviposition stimulants for the egg-larval parasitoid, Chelonus sp. near curvimaculatus (Hymenoptera: Braconidae). Biological Control 4: 22-25. Klingauf, F. (1967) Abwehr- und Meidereaktionen von Blattläusen (Aphididae) bei Bedrohung durch Räuber und Parasiten. Zeitschrift für Angewandte Entomologie 60: 269-317. Liepert, C. & Dettner, K. (1993) Recognition of aphid parasitoids by honeydew-collecting ants: The role of cuticular lipids in a chemical mimicry system. Journal of Chemical Ecology 19: 2143-2153. Mackauer, M. (1986) Growth and developmental interactions in some aphids and their hymenopterous parasites. Journal of Insect Physiology 32: 275-280. Mackauer, M., Michaud, J. P. & Völkl, W. (1996) Host choice by aphidiid parasitoids (Hymenoptera: Aphidiidae): host recognition, host quality, and host value. The Canadian Entomologist 128: 959-980. Michaud, J. P. & Mackauer, M. (1994) The use of visual cues in host evaluation by aphidiid wasps. I. Comparison between three Aphidius parasitoids of the pea aphid. Entomologica Experimentalis et Applicata 70: 273-283. Michaud, J. P. & Mackauer, M. (1995) The use of visual cues in host evaluation by aphidiid wasps. II. Comparison between Ephedrus californicus, Monoctonus paulensis, and Praon pequodorum. Entomologica Experimentalis et Applicata 74: 267-275. Moran N. A. (1992) The evolution of aphid life cycles. Annual Review of Entomology 37: 321-348. Müller, F. P. (1982) Das Problem Aphis fabae. Zeitschrift für Angewandte Entomologie 94: 432446. Pennacchio, F. M., Digilio, C. & Tremblay, A. (1994) Host recognition and acceptance behaviour in two aphid parasitoid species: Aphidius ervi and Aphidius microlophii. Bulletin of Entomological Research 84: 57-64. Salto, C. E., Eikenbary, R. D. & Starks, K. J. (1983) Compatibility of Lysiphlebus testaceipes (Hymenoptera: Braconidae) with Greenbug (Homoptera: Aphididae) Biotypes “C” and “E” reared on susceptible and resistent oat varieties. Environmental Entomology 12: 603-604. Sequeira, R. & Mackauer, M. (1992) Nutritional ecology of an insect host-parasitoid association: the pea aphid-Aphidius ervi system. Ecology 73: 183-189. Sheehan, W. & Shelton, A. M. (1989) The role of experience in plant foraging by the aphid parasitoid Diaeretiella rapae (Hymenoptera: Aphidiidae). Journal of Insect Behavior 2: 743-759. Sokal, R. R. & Rohlf, F. J. (1995) Biometry. 3rd Edition. Freeman & Co., New York. Srivastava, M. (1987) Nutritional physiology. pp. 99-121. In Minks, A. K. & Harrewijn, P. (Ed.), World Crop Pests, 2A. Aphids: Their Biology, Natural Enemies and Control. Elsevier, Amsterdam. Srivastava, M. & Singh, R. (1988) Bionomics of Trioxys indicus, an aphidiid parasitoid of Aphis craccivora. 26. Impact of host extract on the oviposition response of the parasitoid. Biological Agriculture and Horticulture 5: 169-176. Stary P., Lyon, J. P. & Leclant, F. (1988) Biocontrol of aphids by the introduced Lysiphlebus testaceipes (Cress.)(Hym., Aphidiidae) in Mediterranean France. Journal of Applied Entomology 105: 74-87. Thieme, T. (1988) Zur Biologie von Aphis fabae mordwilkowi Börner und Janisch, 1922 (Hom., Aphididae). Journal of Applied Entomology 105: 510-515. van Steenis, M. J. (1994) Intrinsic rate of increase of Lysiphlebus testaceipes Cresson (Hym.; Braconidae), a parasitoid of Aphis gossypii Glover (Hom., Aphididae) at different temperatures. Journal of Applied Entomology 118: 399-406.
Anja Hildebrands, Thomas Thieme and Stefan Vidal 370
Hymenoptera: Evolution, Biodiversity and Biological Control
Vinson, S. B. (1976) Host selection by insect parasitoids. Annual Review of Entomology 21: 109133. Völkl, W. (1991) Species-specific larval instar preferences and aphid defence behaviour in three parasitoids of Aphis fabae. pp. 73-78. In Polgar, L., Chambers, R. J., Dixon, A. F. G. & Hoder, I. (Eds), Behaviour and Impact of Aphidophaga. SPB Academic, The Hague. Völk, W. & Stechmann D. H. (1998) Parasitism of the black bean aphid (Aphis fabae) by Lysiphlebus fabarum (Hym., Aphidiidae): the influence of host plant and habitat. Journal of Applied Entomology 122: 201-206. Weisser, W. W. (1994) Age-dependent foraging behaviour and host-instar preference of the aphid parasitoid Lysiphlebus testaceipes. Entomologica Experimentalis et Applicata 70: 1-10. Wilbert, H. (1965) Die Größenvariabilität von Aphelinus semiflavus Howard (Hym., Aphelinidae) und ihre Ursachen. Zeitschrift für Pflanzenkrankheiten und Pflanzenschutz 72: 670-684.
Size and Asymmetry as Quality Control Indicators in Trichogramma spp. (Hymenoptera: Trichogrammatidae) D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann Centre for Environmental Stress and Adaptation Research, La Trobe University, Bundoora, Victoria 3083 Australia (email:
[email protected])
Introduction Wasps of the genus Trichogramma Westwood are effective in parasitising eggs of many lepidopteran pest species, and have been subjected to extensive use as an inundative release agent, but not always with consistent results. Overseas experience indicates that the inadequacy of mass rearing procedures (Bigler et al. 1987; Hassan & Guo 1991) and the use of inappropriate strains may have been the reason for failures in control programs. Therefore, identification of a suitable strain and determining ways in which the quality of mass reared wasp is maintained in a commercial production facility is particularly important (Wajnberg & Hassan 1994; Smith 1996). Here we consider ways of identifying quality of wasps in two species, Trichogramma nr. brassicae (Bezdenko) and Trichogramma (Trichogrammanza) carverae (Oatman & Pinto). Trichogramma nr. brassicae (previously known as T. nr. ivelae Pang & Chen) is an egg parasitoid of budworm, Helicoverpa punctigera (Wallengren) (Noctuidae) in tomatoes, and effectively controls the host toward the end of the crop season (Bentley et al. 1996). However, control during the early season is inadequate, highlighting the importance of developing a suitable wasp strain as an inundative release agent. Early field trials conducted using commercial strains of T. nr. brassicae in Australia have given mixed results, sometimes with extremely poor rates of parasitism. Trichogramma carverae is an effective control agent for light brown apple moth, Epiphyas postvittana (Walker) (Tortricidae), particularly in vineyards (Glenn & Hoffmann 1997). Field trials have indicated that high rates of parasitism are possible (Glenn & Hoffmann 1997), although variable results have been achieved commercially. An obvious quality measure potentially useful in parasitoid wasps is adult body size. Many studies have established positive associations between size and standard laboratory fitness measures (see Godfray 1994). More recently, studies have addressed the impact of size on field fitness by comparing the size distributions of wasps at emergence sites and at oviposition sites (Visser 1994; Kazmer & Luck 1995; West et al. 1996). As the ability to find hosts may depend on size, the difference between emergence and ovipositing female size distributions could give an indication of the degree of selection on size. Kazmer and Luck (1995) used this approach to show that in the egg parasitoid, Trichogramma pretiosum (Riley), host location success increased with size in smaller females but reached a plateau in larger females. Another potentially useful indicator of quality is fluctuating asymmetry. Fluctuating asymmetry refers to the absolute difference in measures taken on both sides of a bilaterally symmetrical organism and is assumed to reflect the degree of developmental buffering against environmental
371
D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann 372
Hymenoptera: Evolution, Biodiversity and Biological Control
disturbances. Previous studies of fluctuating asymmetry in animals have shown that it may be associated with components of fitness (Møller 1990, 1992; Thornhill 1992; Thornhill & Sauer 1992; Harvey & Walsh 1993). We are unaware of previous studies on parasitoids linking fluctuating asymmetry to fitness, although it has been postulated (Clarke & McKenzie 1992) that asymmetry should be a good quality indicator for insects such as Trichogramma spp. that are mass reared commercially. Here the relationship between fitness and both size and asymmetry measurements is considered in T. carverae and T. nr. brassicae. We ask the following questions: 1) is size and asymmetry correlated with the successful location in the field of egg masses of E. postvittana by T. carverae? These experiments are only described briefly because details are provided elsewhere (Bennett & Hoffmann 1998); and 2) is size and asymmetry correlated with the location of H. punctigera eggs in a glasshouse by T. nr. brassicae? We focused on a glasshouse because with high background parasitism levels, it is almost impossible to examine quality control measures in this species in field tomatoes.
Materials and Methods Trichogramma nr. brassicae Strains of T. nr. brassicae were collected from H. punctigera eggs placed in tomato paddocks in Victoria and southern New South Wales. Nineteen isofemale lines were collected and combined to form a genetically heterogeneous mass-bred population. This stock was raised on Sitotroga cerealella (Olivier) (Gelechiidae) for two generations before a sample was taken from the population prior to releasing wasps into the glasshouse. Females found ovipositing on H. punctigera eggs laid on the potted tomato plants were collected for comparison with the sample of wasps taken pre-release. Using a dissecting microscope, individual wasps from both samples were mounted dorso-ventrally on slides with Hoyer’s mounting medium and held in place with a cover slip. An image analysis system (Trace) linked to a Zeiss compound microscope fitted with a Panasonic digital colour video camera (Model WV-CP610/A) was used to measure size and asymmetry. We obtained six size measurements from females. These were length of the apical antennal segment (ANL), length of the hind tibia (HTL), two indices of forewing length (FWLa, FWLb), width of the forewing (FWW), and width of the head (HDW). The wing measures are illustrated in Figure 1. All of these size measures except HDW and ANL were recorded for both sides of the body to provide asymmetry data. In addition, five meristic traits were obtained from the forewings for FA analyses. These consisted of counts of setae along the wing veins and wing margins (see Fig. 1). The asymmetry of each trait was measured as the signed (R-L) difference. FA is the absolute value of this measure. The repeatability of all FA measures was first examined using the procedure outlined in Palmer (1994). All asymmetry measures were also used to construct a total FA for each individual. Following Woods et al. (1998), this composite index involved standardising all FA measures (leading to a mean of 0 and variance of 1) prior to summing the standardised measures. This approach ensures that all traits are weighted equally when estimating the total FA of an individual.
Trichogramma carverae Parasitised eggs of E. postvittana were used to collect T. carverae from Chardonnay vines. Fourteen isofemale strains from Sunraysia founded the heterogeneous stock, which was maintained on eggs
Size and Asymmetry as Quality Control Indicators in Trichogramma spp. 373
1 2 3
S5 S4
S3 S1 S2 Figure 1
Forewing of female Trichogramma nr. brassicae showing traits measured (the head of the wasp is facing towards the bottom of the page): A) 1 = measure of wing width (FWW), 2 = measure of wing length (FWLa), 3 = measure of wing length (= FWLb); B) setal counts, along the wing veins (S1, S2) and along the wing margin (S3, S4, S5). All traits were scored on both wings to provide a measure of wing FA.
of the grain moth S. cerealella. To test effects of size and asymmetry on field parasitism, releases were conducted in 12-year old, non-irrigated Chardonnay vines with each spot release surrounded by 45 oviposition sites, each containing an E. postvittana egg mass. Prior to release, we collected and measured a sample of wasps emerging from the parasitised S. cerealella eggs. Natural parasitism in the vineyard was tested by placing a total of 120 egg masses laid on strips of plastic by the E. postvittana laboratory culture (see Glenn & Hoffmann 1997; Bennett & Hoffmann 1998), in the vines before and during the trial. During the trials these egg cards, identical to those used in the experiment as collection points, were placed in the opposite half of the vineyard to the wasp release points. As no parasitism was recorded, it can be assumed that the only Trichogramma spp. in the vineyard were the released T. carverae. Wasps ovipositing on the eggs surrounding the release points were collected and measured for size and asymmetry. We obtained six size measures from the female wasps, but these were not all identical to the traits measured on T. nr. brassicae. The traits were HTL, HDW, FWW (all as above), FWLc (similar to FWLa), hind wing length (HWL), and body length from the anterior edge of the thorax to the posterior edge of the abdomen (BDL). Repeat measures on these traits for 30 wasps
D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann 374
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 Results of separate variance t-tests on size measures on Trichogramma spp. to investigate the effects of size on fitness (avalues remain significant at the 5% level even when corrections are made for multiple comparisons with the Dunn-Sidák method (Sokal & Rohlf 1995).
Trait
Pre-release mean
Ovipositing mean t-stat (d.f.)
P (2-tail)
F-stat (d.f.)
P
7.62 (52,45)
0.007a
T. nr. brassicae HTL
0.1410
0.1409
-0.11 (79.28)
0.915
HDW
0.1742
0.2064
13.39 (79.36)
<0.001
0.44 (38,42)
0.508
FWW
0.2282
0.2291
0.64 (89.67)
0.526
2.99 (54,49)
0.087
FWLa
0.3254
0.3304
1.46 (60.82)
0.151
2.76 (54,49)
0.100
FWLb
0.2410
0.2424
0.99 (93.79)
0.323
0.84 (55,49)
0.363
ANL
0.0681
0.0686
0.85 (55.48)
0.401
3.15 (45,33)
0.080
HTL
0.1362
0.1447
3.43 (60.27)
0.001a
5.49 (53,145)
<0.001a
HDW
0.2126
0.2265
2.98 (56.83)
0.004a
4.33 (49,147)
<0.001a
FWW
0.2213
0.2400
4.03 (56.97)
<0.001a
13.11 (54,153)
<0.001a
FWLc
0.4087
0.4475
4.38 (56.53)
<0.001
15.39 (54,153)
<0.001a
HWL
0.2672
0.2849
3.21 (50.91)
0.002
11.34 (48,143)
<0.001a
BDL
0.4065
0.4273
2.77 (53.38)
0.008a
2.39 (37,73)
<0.001a
a
T. carverae
a
a
indicated that they could be measured with a high degree of repeatability (in all cases, r > 0.90, P < 0.001). Asymmetry was measured on six bilateral traits, forewing width (FWW) and three setae counts are directly comparable to earlier measurements on T. nr. brassicae, while the measure of forewing length (FWLc) and fourth setae count are not. More detailed information about these experiments and measurements is provided in Bennett and Hoffmann (1998).
Results Trichogramma nr. brassicae We compared pre-release and ovipositing samples for the six size traits. We used t-tests to compare the groups for these measures. The results (Table 1) indicate that one trait (HDW) showed a highly significant difference between the oviposition and pre-release groups. Most measures of body size were therefore not associated with parasitism success, but there was an extremely strong association with head width. F-tests indicate that the variances of females collected pre-release and those captured while ovipositing did not differ except for HTL (Table 1), for which there was a decreased variance for the ovipositing wasps. A histogram for head width data (Fig. 2) shows the clear difference between the ovipositing and pre-release wasps, the former having much wider heads. There is only a small amount of overlap between the two groups. We compared pre-release and oviposition samples for the fluctuating asymmetry (FA) of each of the nine traits, as well as for a composite measure of FA. Mann-Whitney tests were used to compare the groups for these measures. There was no evidence for significant differences between the ovipositing and pre-release groups for FA of individual traits (Table 2). All probabilities are greater than 0.05 with the exception of FA in S1. However, this value is no longer significant once an adjustment is made for multiple comparisons due to the large number of traits measured.
Size and Asymmetry as Quality Control Indicators in Trichogramma spp.
frequency
375
20 18 16 14 12 10 8 6 4 2 0
pre-release ovipositing
<0.17
0.17-0.18
0.18-0.19
0.19-0.20
0.20-0.21
>0.21
head width (mm) Figure 2
Distribution of head width in Trichogramma nr. brassicae from the pre-release sample and wasps ovipositing in the glasshouse.
35 30
frequency
25 20
pre-release
15
ovipositing
10 5 0 <-1
-1 to -0.5
-0.5 to 0
0 to 0.5
0.5 to 1
>1
composite index of asymmetry Figure 3
Distribution of composite asymmetry for Trichogramma nr. brassicae from the pre-release sample and wasps ovipositing in the glasshouse.
We also examined the composite FA because this might provide a more sensitive indicator of developmental stability, constituting a combination of FA measures for all traits. Because not all measures of FA could be scored on all individuals, sample sizes available for comparing groups for total FA were smaller than for the individual FA measures. Nevertheless, the Mann-Whitney test on composite FA (Table 2) shows that there is a difference between the pre-release and ovipositing groups. The Kolmogorov-Smirnov statistic (testing differences between the distributions) was also significant. The distribution of composite FA values (Fig. 3) shows that prerelease values are higher than those from the ovipositing group. Total FA therefore appears to be a sensitive indicator of wasp fitness in the glasshouse.
Trichogramma carverae The t-tests (Table 1) indicated that all the size traits differed between the ovipositing and prerelease samples. In each case, ovipositing wasps were larger than the pre-release wasps. Because all size traits were positively correlated, we focused on hind tibia length, which is commonly used to assess size in parasitoids. The histogram (Fig. 4) indicates that extremely large wasps were
D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann 376
Hymenoptera: Evolution, Biodiversity and Biological Control
40 35
frequency
30 25 20
pre-release
15
ovipositing
10 5
0. 17 5
0. 17
0. 16 5
0. 16
0. 15 5
0. 15
0. 14 5
0. 14
0. 13 5
0. 13
0. 12 0. 12 5
0. 11 5
0
hind tibia length (mm)
Figure 4
Distribution of hind tibia length in Trichogramma carverae from the pre-release sample and wasps ovipositing in the field.
Table 2 Comparisons of pre-release and ovipositing samples for fluctuating asymmetry in Trichogramma spp., using Mann-Whitney tests and two-sample Kolmogorov-Smirnov tests [aKolmogorov-Smirnov statistics were not computed for bristle asymmetries falling into discrete categories; bvalues remain significant at the 5% level even when corrections are made for multiple comparisons with the Dunn-Sidák method (Sokal & Rohlf 1995)].
Trait
Pre-release Ovipositing mean mean
MannWhitney (z statistic)
P
KolmogorovSmirnov statistic
P
T. nr. brassicae FWW (x100 )
0.25
0.23
-0.45
0.65
0.47
0.98
FWLa (x100 )
0.30
0.27
-1.03
0.31
0.75
0.62
FWLb (x100 )
0.29
0.22
-1.36
0.17
0.65
0.79
HTL (x100 )
0.24
0.20
-0.64
0.52
0.53
0.94
S1a
1.89
1.28
-2.39
0.02
S2a
0.76
0.83
-0.41
0.68
S3
a
1.11
1.09
-0.21
0.83
S4a
1.36
1.06
-1.98
0.05
S5a
1.07
0.82
-1.10
0.27
Composite FA
0.95
-1.30
-2.70
<0.01b
1.68
<0.01b
FWW (x100 )
0.32
0.28
0.12
0.90
0.53
0.94
FWLc (x100 )
1.16
0.83
3.23
<0.001
1.61
<0.01b
FWW (adj) (x100 )
1.42
1.16
0.45
0.65
0.52
0.95
FWLc (adj) (x100 )
2.87
2.05
3.16
<0.001b
1.85
<0.01b
S2a
1.59
1.45
0.27
0.79
S3a
1.38
1.28
0.18
0.85
S4a
1.00
0.96
1.19
0.23
S6a
1.23
1.57
1.37
0.17
Composite FA
0.09
-0.30
-1.22
0.22
-0.22
1.21
T. carverae b
Size and Asymmetry as Quality Control Indicators in Trichogramma spp. 377
25
frequency
20
15 pre-release ovipositing
10
5
4
8 0. 05
0. 05
6
05 0.
2
04 0.
8
04 0.
4 03
03 0.
0.
6
03 0.
2
02 0.
4
8
02 0.
0. 01
0. 01
6
01 0.
00 0.
0.
00
2
0
hind tibia length (mm)
Figure 5
Distribution of fluctuating asymmetry in wing length for Trichogramma carverae from the pre-release sample and wasps ovipositing in the field.
particularly successful at finding E. postvittana egg masses in the field. This resulted in a larger variance for ovipositing females as reflected in the F-test comparing variances (Table 1). For asymmetry, only forewing length asymmetry differed between the samples after adjustment for trait size (Table 2). More asymmetric wasps had a lower oviposition success (Fig. 5). Unlike in T. nr. brassicae, there was no difference between the samples for an index of composite asymmetry.
Discussion Size and fitness Numerous laboratory experiments on parasitoids have found positive correlations between body size and fitness components such as fecundity, mortality, rate of search and flight ability (Godfray 1994) and these include studies on Trichogramma spp. (e.g. Frei & Bigler 1993; Dutton & Bigler 1995). Our data on T. carverae are consistent with three previous studies on field fitness with parasitoids (Visser 1994; Kazmer & Luck 1995; West et al. 1996) indicating that larger wasps are relatively more successful at locating host eggs. Two previous studies on parasitoids (Kazmer & Luck 1995; West et al. 1996) have found that size and field fitness are linearly related until an intermediate size is reached and there is a plateau. The pattern in T. carverae reared on S. cerealella is different; wasps that are intermediate in size, tend to perform poorly and there is an increase in fitness only for wasps having hind tibia lengths that are more than 0.14 mm. This length is only exceeded in 30–40% of the pre-release sample. The data for T. carverae also suggest that wasps in the smallest size classes perform better than those in intermediate classes, although additional releases are needed to verify this further. The size data on T. nr. brassicae shows a different pattern. There is no association between successful egg location and hind tibia length, unlike in T. carverae. However, the association between successful egg location and head width is extremely strong. Thus, while overall size is not correlated to parasitism success, wasps with a relatively wide head are at an advantage. We have no idea why this is the case. However, field trials are needed to confirm the usefulness of head width as a quality indicator. Such trials will need to be undertaken in areas where T. nr. brassicae is rare to
D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann 378
Hymenoptera: Evolution, Biodiversity and Biological Control
avoid the problem of background parasitism. Parts of Queensland where T. pretiosum predominates in tomato crops may be suitable.
Fluctuating asymmetry and fitness In the case of T. nr. brassicae, trait asymmetries were not related to parasitism success, while the total asymmetry was a good predictor of success. This suggests that developmental stability as measured by total asymmetry can be linked to fitness, and that selection probably does not act directly on the asymmetry of the traits, but on the overall level of developmental disturbance experienced by an organism. As in the case of head width, this finding needs to be validated in field releases. However, the data from T. carverae suggest a different story. In this species, females that located host eggs in the field exhibited a lower degree of fluctuating asymmetry in forewing length when compared to an emergence sample, whereas overall developmental disturbance was not important. This finding suggests that wing length asymmetry reflects the fitness level of an organism. Asymmetry in wing length may have directly contributed to the ability of wasps to reach the eggs. This asymmetry could influence flight ability (e.g. Swaddle et al. 1996), which is thought to be an important predictor of field parasitism rates in Trichogramma spp. (Cerutti & Bigler 1995). These findings suggest that asymmetry in traits that are the target of selection may be a better predictor of fitness than the traits themselves.
Implications for quality control The ability of parasitic wasps to discover hosts is an important factor determining their effectiveness and success as a biological control agent (Pak et al. 1991). Traits identified as controlling fitness in this study are likely to contribute to the success of inundative releases of Trichogramma. Our findings indicate that only a small fraction of released wasps have a high fitness. For instance, in the case of T. carverae, wasps having a hind tibia length of 0.16 mm or greater had the highest fitness but would have comprised only 0.04% of the release sample based on the mean and standard deviation of lengths in the emergence sample. Yet these wasps comprised 27% of the ovipositing sample. Wasps reaching oviposition sites may therefore be an extremely select group of those released. Thus, appropriate measures of quality can act as important tools in ensuring high rates of parasitism. Unfortunately, the data presented here also indicate that different quality measures may be pertinent to different Trichogramma spp. and/or different settings. This means that a thorough study is required of individual Trichogramma-target systems to identify quality indicators relevant to particular situations. There is probably no universal quality indicator.
Acknowledgements We are grateful to DeAnn Glenn for collecting T. carverae strains. We also thank Reg Egan and Lou Chirnside for providing access to their properties. This work was supported by grants from the Grape and Wine Research and Development Corporation, the Australian Processing Tomato Research Council, the Horticultural Research and Development Corporation and the Australian Research Council.
References Bennett, D. M. & Hoffmann, A. A. (1998) Effects of size and fluctuating asymmetry on field fitness of the parasitoid Trichogramma carverae (Hymenoptera: Trichogrammatidae). Journal of Animal Ecology 67: 580-591.
Size and Asymmetry as Quality Control Indicators in Trichogramma spp. 379
Bentley, J., Ridland, P., Walker, B. & Hind, M. (1996) Final Report – Development of IPM strategies for processing tomatoes. Agriculture Victoria, Melbourne. Bigler, F., Meyer, A. & Bosshart, D. (1987) Quality assessment in Trichogramma maidis Pintureau et Voegle’ reared from eggs of the factitious hosts Ephestia kueniella Zell. and Sitotroga cerealella. Journal of Applied Entomology 104: 340-353. Cerutti, F. & Bigler, F. (1995) Quality assessment of Trichogramma brassicae in the laboratory. Entomologia Experimentalis et Applicata 75: 19-26. Clarke, G. M. & McKenzie, L. J. (1992) Fluctuating asymmetry as a quality control indicator for insect mass rearing processes. Journal of Economic Entomology 85: 2045-2050. Dutton, A. & Bigler, F. (1995) Flight activity assessment of the egg parasitoid Trichogramma brassicae (Hymenoptera: Trichogrammatidae) in laboratory and field conditions. Entomophaga 40: 223-233. Frei, G. & Bigler, F. (1993) Fecundity and host acceptance tests for quality control of Trichogramma brassicae. pp. 81-95. Proceedings of the 7th Workshop of the Global IOBC Working Group “Quality Control of Mass Reared Arthropods”. IOBC, Rimini. Glenn, D. C. & Hoffmann, A. A. (1997) Developing a commercially viable system for biological control of Epiphyas postvittana (Lepidoptera: Tortricidae) in grapes using endemic Trichogramma (Hym: Trichogrammatidae). Journal of Economic Entomology 90: 370-382. Godfray, H. C. J. (1994) Parasitoids, Behavioural and Evolutionary Ecology. Princeton University Press, Princeton, New Jersey. Harvey, I. F. & Walsh, K. J. (1993) Fluctuating asymmetry and lifetime mating success are correlated in males of the damselfly Coenagrion puella (Odonata: Coenagrionidae). Ecological Entomology 18: 198-202. Hassan, S. A. & Guo, M. F. (1991) Selection of effective strains of egg parasites of the genus Trichogramma (Hym., Trichogrammatidae) to control the European corn borer Ostrinia nubilalis Hb. (Lep., Pyralidae). Journal of Applied Entomology 111: 335-341. Kazmer, D. J. & Luck, R. F. (1995) Field tests of the size-fitness hypothesis in the egg parasitoid Trichogramma pretiosum. Ecology 76: 412-425. Møller, A. P. (1990) Fluctuating asymmetry in male sexual ornaments may reliably reveal male quality. Animal Behaviour 40: 1185-1187. Møller, A. P. (1992) Patterns of fluctuating asymmetry in weapons: evidence for reliable signalling of quality in beetle horns and bird spurs. Proceedings of the Royal Society of London, Series B 248: 199-208. Pak, G. A., Berkhout, H. & Klapwijk, J. (1991) Do Trichogramma look for hosts? pp. 77-80. Third International Symposium on Trichogramma and Other Parasitoids, INRA, San Antonio. Palmer, A. R. (1994) Fluctuating asymmetry analysis: a primer. pp. 335-364. In Markow, T. A. (Ed.), Developmental Instability: its Origins and Evolutionary Implications. Kluwer Academic, Dordrecht. Smith, S. M. (1996) Biological control with Trichogramma: advances, successes and potential of their use. Annual Review of Entomology 41: 375-406. Sokal, R. R. & Rohlf, F. J. (1995) Biometry, 3rd Ed. Freeman, New York. Swaddle, J. P., Witter, M. S., Cuthill, I. C., Budden, A. & McCowen, P. (1996) Plumage condition affects flight performance in common starlings – implications for developmental homeostasis, abrasion and moult. Journal of Avian Biology 27: 103-111. Thornhill, R. (1992) Fluctuating asymmetry and the mating system of the Japanese scorpionfly, Panorpa japonica. Animal Behaviour 44: 867-879. Thornhill, R. & Sauer, P. (1992) Genetic sire effects on the fighting ability of sons and daughters and mating success of sons in a scorpionfly. Animal Behaviour 43: 255-264.
D. M. Bennett, S. Hewa-Kapuge and A. A. Hoffmann 380
Hymenoptera: Evolution, Biodiversity and Biological Control
Visser, M. E. (1994) The importance of being large: the relationship between size and fitness in females of the parasitoid Aphaereta minuta (Hymenoptera: Braconidae). Journal of Animal Ecology 63: 963-978. Wajnberg, E. & Hassan, S. A. (1994) Biological Control with Egg Parasitoids. CAB International. West, S. A., Flanagan, K. E. & Godfray, H. C. J. (1996) The relationship between parasitoid size and fitness in the field, a study of Achrysocharoides zwoelferi (Hymenoptera: Eulophidae). Journal of Animal Ecology 65: 631-639. Woods, R., Hercus, M. & Hoffmann, A. A. (1998) Estimating the heritability of fluctuating asymmetry in field Drosophila. Evolution 52: 816-824.
The Effects of Two New Insecticides on the Survival of Adult Trichogramma Pretiosum Riley in Sweet Corn B. C. G. Scholz1 and M. P. Zalucki2 1
Farming Systems Institute, Queensland Department of Primary Industries, P.O. Box 102, Toowoomba, Qld 4350 Australia (e-mail:
[email protected]) 2
Department of Zoology and Entomology, University of Queensland, Brisbane, Qld 4072 Australia
Introduction The corn earworm or heliothis caterpillar, Helicoverpa armigera (Hübner), is a major pest of sweet corn in the Lockyer Valley region of south-east Queensland. This pest has been managed by chemical insecticides for many years, but is now becoming more difficult to control due to widespread resistance problems. Growers are keen to adopt new pest management practices to reduce the current dependence on chemicals. The Queensland Department of Primary Industries (QDPI) has been evaluating control tactics that have potential for use in integrated pest management (IPM) programs. A key component of the IPM programs being investigated is the conservation of naturally occurring egg parasitoids and predators. A species of egg parasitoid, Trichogramma pretiosum Riley, from Kununurra in Western Australia was released in the Lockyer Valley in 1995 (Scholz unpublished). This species became established in the region and by 1997 was causing significant H. armigera mortality in summer planted sweet corn (Scholz et al. 1998). The use of existing chemical insecticides, notably deltamethrin and methomyl, is highly disruptive to the activity and survival of Trichogramma. Alternative products that have minimal impact on non-target organisms are being sought for inclusion in the IPM programs. Some of the newest insecticides being developed act as heliothis larvicides, but are reported to have minimal, or reduced, impact on predators and parasitoids. IPM programs incorporating such products can use the natural mortality caused by beneficial invertebrates to supplement the mortality caused by the insecticide. Two new insecticides, spinosad and indoxacarb, have recently become available for evaluation against lepidopteran larvae in Australia. Spinosyn will be marketed as Success® in vegetables and Tracer® in cotton, and acts as a nerve toxin on a range of lepidopterous larvae, while not adversely affecting most predatory insects and spiders (Sparks et al. 1995). The primary mode of action of spinosad is via ingestion, although it also has some contact action (Holloway & Forrester 1998). spinosad has potential for inclusion in IPM programs due to its’ selective nature, and has effectively controlled heliothis larvae in cotton (Murray & Lloyd 1997) and sweet corn (Scholz 1998). Indoxacarb (DPX-MP062) is another novel insecticide, likely to be marketed as Avatar® in vegetables and Steward® in field crops. Indoxacarb is a stomach poison and is effective against a range of caterpillar pests, including heliothis, with minimal impact on non-target organisms (Harder
381
B. C. G. Scholz and M. P. Zalucki 382
Hymenoptera: Evolution, Biodiversity and Biological Control
et al. 1997). The effects of applications of these new products against naturally occurring populations of T. pretiosum in sweet corn were evaluated at the QDPI Gatton Research Station in the Lockyer Valley.
Materials and Methods Two sweet corn trials were planted at QDPI Gatton Research Station as part of on-going evaluations of IPM practices against heliothis. The work reported here was part of two trials evaluating the effects of new insecticides as heliothis management tools in sweet corn, which will be reported elsewhere. Both trials consisted of 0.9 ha plantings of sweet corn. Trial 1 was planted on 4 February 1998 with a tropical sweet corn hybrid (H5), at a rate of 80 000 plants/ha. Trial 2 was planted on 25 August 1998 with a temperate hybrid (Golden Sweet), at a rate of 60 000 plants/ha. The study plots were 25 m long and 10 rows wide, and were arranged as randomised complete blocks separated by 15 m of buffer corn longitudinally and six rows of buffer corn laterally that were not treated. Three buffer rows were sown on both edges of each trial. Spinosad was available during Trial 1, and both spinosad and indoxacarb were available during Trial 2.
Effects of insecticides on foraging T. pretiosum Cards containing approximately 20 H. armigera eggs less than 24 h old (referred to as heliothis egg cards), were used to assess levels of H. armigera egg parasitism following field applications of spinosad and deltamethrin to sweet corn. Laboratory reared heliothis moths were placed in oviposition chambers where they laid eggs onto paper towelling (see Teakle & Jensen 1985). Each card was made by stapling pieces of paper towelling containing H. armigera eggs to white paper strips measuring 1.5 × 7 cm. They were collected 48 h after being placed in the field. The numbers of eggs on each card before and after collection were recorded. At collection all whole eggs remaining on cards were ‘tagged’ by placing a small ink dot beside each egg. The numbers of eggs that collapsed or were infertile were also recorded when the cards were collected. Eggs that were white, and showed no visual sign of aging after two days exposure in the field, were assumed to be infertile. Collected egg cards were held in a constant temperature room at approximately 22°C and 70% R.H. to determine the levels of egg parasitism. Three days after collection, the cards were inspected again to determine the fate of all ‘tagged’ eggs. The numbers of parasitised (black) eggs, hatched eggs, collapsed eggs and unhatched eggs were recorded. A total of 12 egg cards/plot (48 per treatment) were stapled onto the upper surface of corn leaves at silk height in the crop. Egg cards were placed in plots in a grid of four cards per row, 5 m apart, in every third row. The cards were inspected hourly and checked for foraging Trichogramma with a hand lens. The percentage of viable eggs parasitised was calculated according to the following formula: %VEP = (NBE/ (NBE + NH)) × 100 where
%VEP = percentage viable eggs parasitised NBE = number of black (parasitised) eggs NH = number of eggs hatched
This study was carried out during Trial 1. All plots were sprayed twice with a Stihl SR400 blower mister in a spray volume of 50 L/ha. The study was undertaken one day after the second spray
The Effects of Two New Insecticides on the Survival of Adult Trichogramma Pretiosum 383
when the crop was silking. There were three treatments: 1) unsprayed control, 2) spinosad at 96 g.a.i./ha (800 mL product/ha), and 3) deltamethrin at 12.5 g.a.i./ha (500 mL product/ha).
Effects of insecticide sprays on parasitoid survival The effect of directly spraying Trichogramma was evaluated by exposing caged wasps to insecticide sprays in the field. Fine stainless steel mesh cages were used to contain adult T. pretiosum at silk height in plots of sprayed sweet corn. The cages were 5 × 10 cm cylinders constructed from very fine stainless steel mesh (47 strands/cm; 0.125 mm aperture; 0.08 mm diameter wire), with a removable lid at each end. A 4 cm diameter hole was cut in each lid and covered with the same wasp proof stainless steel mesh. Approximately 20 T. pretiosum adults, less than 24 h old, were introduced into each cage. The cages were tied around a silk at random locations throughout plots just prior to spraying, and collected immediately after spraying. They were placed in a cooler, returned to the laboratory, and the numbers of dead and surviving wasps in each cage were recorded. The effects of methomyl, spinosad and indoxacarb applications on wasp survival were compared to unsprayed control plots on three separate occasions during Trial 2. The application rates of products were: 1) indoxacarb at 75 g.a.i./ha (250 g product/ha), 2) spinosad at 96 g.a.i./ha (800 mL product/ha), and 3) methomyl at 450 g.a.i./ha (2.0 L product/ha). The indoxacarb and spinosyn were applied in a spray volume of 50 L/ha using a Stihl SR400 blower mister, and the methomyl was applied to a nearby commercially managed planting of sweet corn using a ground rig at 555 L/ha.
Effects of leaf residues on parasitoid survival A series of bioassay chambers were used to evaluate the effects of insecticide residues on sweet corn leaves on parasitoid survival during Trial 2. The apparatus was described by Scholz (1994), and consisted of 60 ventilated glass tubes 150 mm long and 40 mm in diameter. Sweet corn plants were sprayed in the field and leaves were collected 1, 3 and 5 days after spraying. Approximately 20 newly emerged adult T. pretiosum, less than 24 h old, were exposed to the leaves for 4 h, and the proportion of wasps that died in each chamber were recorded. Ten replicates were set up for each insecticide on each day. Two 10 m row lengths of sweet corn (in a buffer region) were sprayed with each insecticide in a spray volume of 50 L/ha using a hand held rotary cage atomiser. The treatments were: 1) unsprayed control, 2) indoxacarb at 75 g.a.i./ha (250 g product/ha), 3) spinosad at 96 g.a.i./ha (800 mL product/ha), 4) endosulfan at 735 g.a.i./ha (2.1 L product/ha), and 5) methomyl at 450 g.a.i./ha (2.0 L product/ha).
Effects of plot spraying on natural heliothis egg parasitism It was not possible to carry out regular collections of naturally laid eggs from all plots throughout the trial due to constraints on time and resources. However, 25 brown H. armigera eggs were collected from the control, standard chemical (methomyl), spinosad and indoxacarb treatments on one occasion. The proportion of eggs parasitised were recorded, and are presented here.
Results The synthetic pyrethroid deltamethrin was highly disruptive to foraging Trichogramma, with wasps observed on one egg card only, at 11:00 am (Fig. 1). Foraging Trichogramma were readily
B. C. G. Scholz and M. P. Zalucki 384
Hymenoptera: Evolution, Biodiversity and Biological Control
100
Control
Deltamethrin Spinosad
60
40
20
0 7:00
Figure 1
8:00
9:00
10:00
11:00
12:00
13:00
14:00
15:00
16:00
The proportion (%) of H. armigera egg cards containing foraging T. pretiosum in sweet corn sprayed with different insecticides at QDPI Gatton Research Station. The egg cards were inspected hourly for foraging Trichogramma on 7 April 1998. The deltamethrin and spinosad treatments were sprayed twice, and the observations occurred one day after the second spray.
observed in the unsprayed control and spinosad plots from 10:00 am onwards, although the activity in the spinosad plots was approximately one third that in the control plots (Fig. 1). No survival of T. pretiosum in cages was recorded in methomyl sprayed plots (Fig. 2), indicating that it was highly toxic to the parasitoids. The caged wasps spent 2 h longer in the crop during the methomyl evaluation, due to a delay in spraying. This may explain why the control survival in the methomyl evaluation (67%) was lower than the control survival for the other two evaluations (94–99%). There was moderate survival of Trichogramma in the spinosad treatment (54%) and high survival in the indoxacarb treatment (93%). There was an obvious residual affect of methomyl on Trichogramma survival one day after spraying (DAS) (Table 1), with only 13% of wasps surviving the 4 h exposure to sprayed leaves. Almost half of the wasps exposed to spinosad sprayed leaves 1 DAS survived (45%), and nearly all of the wasps exposed to indoxacarb sprayed leaves 1 DAS (95%) survived. Endosulfan was not as disruptive as methomyl or spinosad 1 DAS (Table 1). There were no significant differences between the survival of Trichogramma exposed to leaves 3 DAS and 5 DAS (Table 1). The natural egg parasitism data supported the other findings, i.e. indoxacarb did not greatly affect the levels of egg parasitism, spinosad was moderately disruptive, and standard chemicals were highly disruptive. The parasitism levels recorded were: 48, 64, 12 and 0% for the control, indoxacarb, spinosad and standard chemical treatments respectively. Two species of egg parasitoids were recorded from naturally laid heliothis eggs, i.e. T. pretiosum and a species of Telenomus (Scelionidae).
Discussion The data suggest that indoxacarb is very safe on Trichogramma. Only minor mortality due to residues on leaves were recorded, while direct spraying caused little mortality and natural levels
The Effects of Two New Insecticides on the Survival of Adult Trichogramma Pretiosum 385
* P = 0.0181
100
98.5 93.4
Control Treatment
** P < 0.0001 93.5
** P = 0.0016
80
67.2
%
54.4
60 40 20 0
0 Indoxacarb Figure 2
Spinosad
Methomyl
The survival (%) of T. pretiosum in cages in sweet corn after plots were sprayed with different insecticides. The cages were placed in plots immediately before spraying, and collected immediately after spraying. Data are the means (and S.E.) of 10 cages and, for a given insecticide, the means were compared using an unpaired t-test.
of egg parasitism were not reduced. Spinosad was moderately disruptive to Trichogramma. Approximately half of the wasps exposed to leaf residues or direct spraying in the field died, and natural levels of egg parasitism in the field were reduced. In contrast, methomyl was highly disruptive to Trichogramma survival and activity, with leaf residues causing nearly 90% mortality 1 DAS and direct spraying killing all caged wasps. Bull and House (1983) also found that methomyl was toxic to T. pretiosum, and that the residual action of the product on leaves was high one day after spraying. Waite (1981) suggested that methomyl was responsible for reducing the natural levels of heliothis egg parasitism in cotton in central Queensland from 60% to 7%. Methomyl has been widely used by sweet corn growers, but is clearly disruptive to Trichogramma. Endosulfan appears to be one of the least disruptive of the existing products. Jacobs et al. (1984) found that endosulfan was much safer on T. pretiosum than permethrin, and that residues on tomato leaves only caused significant parasitoid mortality for one day after spraying. Although endosulfan is regarded as being the ‘softest’ (on beneficials) of the existing products registered for use against heliothis on sweet corn, there are environmental concerns about the product (Connolly 1998). Ideally products that are benign on non-target organisms are most suited for IPM programs. Sweet corn growers are keen to adopt IPM and use ‘selective’ insecticides for heliothis management, however there are currently no ‘selective’ insecticides registered for use in sweet corn in Queensland, including the pathogens Bacillus thuringiensis and Baculovirus heliothis. High natural levels of heliothis egg parasitism (95%) have been reported in summer planted sweet corn in the Lockyer Valley (Scholz et al. 1998). Every attempt should be made to conserve these natural populations of parasitoids by using selective insecticides only when needed. The two new products evaluated here, indoxacarb and spinosad, do not affect T. pretiosum as much as existing products and would be valuable tools in an IPM program. Indoxacarb appears reasonably safe against Trichogramma and spinosad is only moderately toxic to Trichogramma. Both products are much more ‘Trichogramma friendly’ than existing chemical
B. C. G. Scholz and M. P. Zalucki 386
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 1 The survival (%) of T. pretiosum after 4 h exposure to sweet corn leaves sprayed with different insecticides, at various times after spraying. Values represent the mean ± standard error of ten replicates. Column means followed by the same letter are not significantly different (P ≤ 0.05, ANOVA, Fisher’s LSD comparison). Absence of letters indicates non-significance. All data were arcsine transformed for analyses. Days After Spraying Treatment
1
3
5
Control
98.2 ± 1.2 a
94.1 ± 2.0
93.2 ± 2.4
Indoxacarb
95.3 ± 1.9 a
96.5 ± 1.8
94.7 ± 1.8
Endosulfan
74.2 ± 7.5 b
90.6 ± 3.3
93.4 ± 2.9
Spinosyn
44.8 ± 6.7 c
86.7 ± 4.3
85.5 ± 3.8
Methomyl
13.3 ± 6.6 d
81.8 ± 5.4
92.3 ± 3.6
insecticides. In addition they do not adversely affect predatory beetles, bugs or spiders (Murray & Lloyd 1997; Scholz 1998). Consequently, they are ideal for IPM programs, where their action can complement mortality due to naturally occurring Trichogramma and/or predators.
Acknowledgements We thank Dave Schofield and the farm hands at QDPI Gatton Research Station for planting and maintaining the trials. Kristin Latimer, Sue MacLean (QDPI) and Rob Annetts (Dow AgroSciences) provided technical support. Tim Hammond, Geoff Cornwell (DuPont) and Rob Annetts (Dow AgroSciences) provided product for the research. This work was funded by the Queensland Horticulture Institute, the Horticultural Research and Development Corporation, and DuPont Agricultural Products, and is part of a national sweet corn project (HRDC Project No. VG97036).
References Bull, D. L. & House, V. S. (1983) Effects of different insecticides on parasitism of host eggs by Trichogramma pretiosum Riley. Southwestern Entomologist Supplement 8: 46-53. Connolly, R. (1998) Farm management to minimise endosulfan movement. The Australian Cottongrower 19: 62-64. Harder, H. H., Riley, S. L., McCann, S. F. & Sherrod, D. W. (1997) DPX-MP062: A novel broad spectrum, environmentally soft, insect control compound Proceedings of the 1997 Beltwide Cotton Conferences, January 6-10, New Orleans, Louisiana 1: 48-50. Holloway, J. & Forrester, N. (1998) New insecticide chemistry for Australian cotton IPM. The Australian Cottongrower 19: 29-34. Jacobs, R. J., Kouskolekas, C. A. & Gross Jr., H. R. (1984) Responses of Trichogramma pretiosum (Hymenoptera: Trichogrammatidae) to residues of permethrin and endosulfan. Environmental Entomology 13: 355-358. Murray, D. A. H. & Lloyd, R. J. (1997) The effect of spinosad (Tracer) on arthropod pest and beneficial populations in Australian cotton. Proceedings of the 1997 Beltwide Cotton Conferences, January 6-10, New Orleans, Louisiana 2: 1087-1091. Scholz, B. C. G. (1994) The effects of insecticides on the survival of heliothis egg parasitoids. Proceedings of the Seventh Australian Cotton Conference. Broadbeach, Queensland, 10-12 August, 1994. pp. 69-73.
The Effects of Two New Insecticides on the Survival of Adult Trichogramma Pretiosum 387
Scholz, B. (1998) IPM of heliothis in Lockyer Valley sweet corn. pp. 30-35. In Williams, B. & Beckingham, C. (Eds), Proceedings of the New South Wales Vegetables Conference, 5 August 1998, Bathurst, NSW. Scholz, B. C. G., Monsour, C. J. & Zalucki, M. P. (1998) An evaluation of selective Helicoverpa armigera control options in sweet corn. Australian Journal of Experimental Agriculture 38: 601-607. Sparks, T. C., Thompson, G. D., Larson, L. L., Kirst, H. A., Jantz, O. K., Worden, T. V., Hertlein, M. B. & Busacca, J. D. (1995) Biological characteristics of the spinosyns: A new naturally derived insect control agents. Proceedings of the 1995 Beltwide Cotton Conferences, January 4-7, San Antonio, Texas 2: 903-907. Teakle, R. E. & Jensen, J. M. (1985) Heliothis punctiger. pp. 313-322. In Singh, P. & Moore, R. F. (Eds), Handbook of Insect Rearing, Vol. 2. Elsevier Science Publishers, Amsterdam. Waite, G. K. (1981) Effect of methomyl on Heliothis spp. eggs on cotton in central Queensland. Protection Ecology 3: 265-268.
Field Observations on Selective Food Plants in Habitat Manipulation for Biological Control of Potato Moth by Copidosoma koehleri Blanchard (Hymenoptera: Encyrtidae) L. R. Baggen1, G. M. Gurr2 and A. Meats3 1,2
Faculty of Rural Management, The University of Sydney, PO Box 883, Orange, NSW 2800 Australia (2corresponding Author – email:
[email protected])
3
School of Biological Sciences, The University of Sydney, NSW 2006 Australia
Introduction The potato moth, Phthorimaea operculella (Zeller), remains a key pest of potatoes in Australia (Dillard et al. 1993; Spooner-Hart & Redgrove 1995). This is despite the establishment of several exotic hymenopteran parasitoids, including Copidosoma koehleri Blanchard which is a dominant species in inland south-east Australia (Horne 1990). The BIOCAT database of classical biological control attempts involving insects (Greathead & Greathead 1992) indicates that this is not unusual. From the 1880’s to the present day fewer than half of the agents which established gave successful control of the target pest. It has been argued that a factor likely to have contributed to this low level of success is that the ecological requirements of agents are seldom considered (Gurr & Wratten 1999). Adult C. koehleri, benefit from nectar (Baggen & Gurr 1998) but opportunities to feed under field conditions are poor as a result of efficient weed control and use of large potato monocultures. Conservation biological control methods (see recent reviews by Wratten et al. 1998; Gurr et al. 1998a; Landis et al. 2000) seek to remedy such paucities of key ecological resources, often by using habitat manipulation to introduce nectar rich plants into the agricultural landscape. However, some such attempts can exacerbate crop damage as a result of the pest utilising the resource (e.g. Collins & Johnson 1985; Baggen & Gurr 1998). Laboratory screening of a range of plants has identified several species with flowers which, whilst conferring a benefit to adult C. koehleri, conferred no such benefit to the pest, P. operculella (Baggen & Gurr 1998; Baggen et al. 1999). Phacelia (Phacelia tanacetifolia Benth), nasturtium (Tropaeoleum majus L.) and borage (Borago officinalis L.) were such ‘selective’ food plants. The aim of this study was to test whether such selectivity applied under field conditions or was an artefact of the small scale of previous flight cage testing. The effects of growing borage along the margin of a commercial potato crop were compared with those of buckwheat (Fagopyrum esculentum Benth.), a non-selective plant which, under laboratory conditions, was fed upon by both parasitoid and pest.
Materials and Methods Field experiments took place in two commercial potato crops on the New South Wales Tablelands close to the town of Blayney. Each crop was divided into two approximately equal halves
388
Hymenoptera: Evolution, Biodiversity and Biological Control
Field Observations on Selective Food Plants in Habitat Manipulation 389
separated by an unplanted ‘laneway’ approximately 10 m wide. On one site, borage was sown in the field margin along one half of one edge such that it extended from one corner to as far as the ‘laneway’. On the second site, buckwheat was planted in the same manner. The resulting strips of borage and buckwheat were approximately 200 m long × 2 m wide. All monitoring took place during the period when both these plants were flowering.
Oocyte counts On four dates, 21 and 28 February and 7 and 14 March 1997, three samples of moths were taken in each of six 25 m long × 3 m wide zones within the crop. In the half of the trial with a border of flowers, zones were 3 m from this margin, 3 m from the opposite edge (100 m and 150 m distant in the borage and buckwheat experiments, respectively) and equidistant between the two edges. In the half of the trial without flowers, zones were in equivalent positions. Collections were made in late afternoon/early evening using a sweepnet and applying a uniform sampling effort to each of the zones within both experiments. After sampling each zone, the contents of the net were emptied into a labeled plastic bag, returned to the laboratory and frozen. Subsequently, five female moths were randomly selected from within each sample and dissected under a binocular microscope using × 10–20 magnification. Dissection involved removing the abdomen with a scalpel, teasing it apart with fine tweezers and counting chorionated oocytes. No stains were used. Insect trapping Sticky intercept traps were constructed by screwing four rigid 2 mm thick transparent acetate boards measuring 210 mm × 300 mm onto a wooden tomato stake. A sheet of clear acetate sheeting was then attached to one face of each board and held in place by bulldog clips. The outer surface of acetate sheets were than treated with Tanglefoot® aerosol spray. These boards were mounted on each stake such that when placed in the field one pair of opposite facing boards was immediately above crop canopy height, approximately 65 cm high, and the second pair corresponded to the height of the buckwheat or borage flowers, approximately 1.0 m high. Four such trap assemblies were placed in each of fifteen positions within each of the two experiments. Five trap positions occupied the half of the field with a floral edge such that one was 2 m from the flowers, a second 2 m from the opposite margin, a third in the centre, and a further two equidistant between the centre and outer traps. Stakes were orientated so traps faced towards or against the flowering margin. The same trap positions applied within the half of each experiment without a floral edge, with traps facing either towards or against the unsown (control) edge. The remaining five trap positions occupied the unplanted ‘laneway’ between the two halves of each experiment. These were spaced in the same pattern as the other traps but were orientated at 90° so as to catch insects moving between the halves of the experiments. Trap catches were recovered on 21 and 28 February and 7 and 14 March 1997, each date marking the end of a one-week trapping period. On each occasion a fresh sticky acetate sheet was used, the old one being placed inside a folded sheet of clear plastic. This allowed samples to be readily transported, stored and for the moths and wasps subsequently to be counted.
Data analysis Oocyte counts for the 15 moths caught from each of the six positions within each experiment were meaned for each of the four sample dates. Because relatively few moths and wasps were trapped, data from all four weeks were pooled. These were analysed by graphical plots to show spatial trends evident for wasps and moths over the 15 positions within each experiment.
L. R. Baggen, G. M. Gurr and A. Meats 390
Hymenoptera: Evolution, Biodiversity and Biological Control
The value used for each position was the total number of moths caught on all four boards on each wooden stake and was taken as an indication of trivial (i.e. non-directional) insect movement.
Results Mean numbers of chorionated oocytes in adult female P. operculella caught from close to the buckwheat flowers ranged between 55 and 62 on the four dates. Elsewhere within this trial numbers were lower, ranging from 22 to 46. In the borage-bordered field, chorionated oocyte means ranged from 22 to 41 with values for the position closest to the flowers no higher than elsewhere within the experiment (Table 1). Mean catches of adult P. operculella on sticky traps placed in the control half and intersection areas of the buckwheat experiment ranged from one to six with no spatial trend (Fig. 1A). Catches from the buckwheat-planted half of the experiment were within this range for traps 75 m or more from the flowering strip but higher (8.5) for the traps 36 m from the flowers, and virtually three times this number (24) for traps 2 m from buckwheat. In the borage experiment, catches ranged from 2 to 10 and no elevation of numbers was evident close to the borage flowers (Fig. 1B). Mean catches of adult C. koehleri on sticky traps placed in the control half and intersection areas of the buckwheat experiment ranged from 0.5 to 2.0 with no spatial trend (Fig. 2A). Catches from the buckwheat-planted half of the experiment were within this range for traps 75 m or more from the buckwheat strip, but several times this (4.5) for the traps 36 m from the flowers and more than twice this (9.75) for traps 2 m from buckwheat. In the borage experiment, mean catches of adult C. koehleri on sticky traps placed in the control half and intersection areas ranged from 0.25 to 2.0 with no spatial trend (Fig. 2B). Catches from the borage-planted half of the experiment were 2.25 or lower for traps 50 m or more from the flowering strip, but more than twice this (5.5) for the traps 27 m from the flowers and 8.25 for traps 2 m from borage (Fig. 2B).
Discussion Under laboratory conditions, adult longevity and fecundity of P. operculella is greater when caged with honey solution or with flowers of buckwheat than when caged with borage flowers with water or with no provision (Baggen & Gurr 1998). Measuring either of these parameters under open field conditions is intractable but counts of chorionated oocytes within trapped females provide a good indication of relative fecundity. In the present study, moths caught close to flowering buckwheat consistently contained considerably more chorionated oocytes than did those caught from elsewhere. No such effect was evident for the borage field. This agreement of field observations with earlier laboratory studies is consistent with moths being able to feed upon buckwheat flowers but unable to access nectar from the ‘selective’ flowers of borage. If adult P. operculella feed from particular flower species, trapping should indicate a greater level of activity in the vicinity of these than is apparent for other positions within the same field where no such resources are available. This would result from moths encountering available nectar but having to return to the potato crop to oviposit on the host plant. This effect was apparent for the buckwheat field, with more than double the number of moths caught on traps 2 m from the buckwheat than on other traps.
Field Observations on Selective Food Plants in Habitat Manipulation 391
Table 1 Mean (±S.E.) numbers of chorionated oocytes per female Phthorimaea operculella caught from different zones of field experiments with borage or buckwheat floral edges (n = 15). Buckwheat Planted on ‘Floral Edge’
Borage Planted on ‘Floral Edge’
21 Feb ‘97
28 Feb ‘97
7 Mar ‘97
14 Mar ‘97
21 Feb ‘97
28 Feb ‘97
7 Mar ‘97
14 Mar ‘97
Floral Edge
55.00(6.04)
57.07(4.90)
61.47(4.93)
59.33(3.98)
39.93(5.83)
34.93(6.45)
32.00(5.82)
28.67(4.29)
Centre
22.07(5.69)
32.80(5.31)
33.40(6.65)
30.87(5.98)
36.53(6.86)
33.40(5.94)
39.87(5.42)
27.87(4.76)
Opposite Floral Edge
23.60(5.00)
32.13(5.45)
41.07(5.38)
34.07(6.20)
37.40(6.73)
32.73(4.50)
30.47(5.67)
33.73(3.46)
Control Edge
27.07(5.74)
30.33(5.13)
26.73(4.81)
32.60(6.47)
40.00(6.63)
33.47(3.82)
31.27(4.85)
28.60(5.79)
Centre
33.07(6.68)
31.27(5.66)
36.13(6.23)
32.33(5.98)
33.20(5.13)
35.13(4.58)
30.53(6.92)
31.73(5.02)
Opposite Control Edge
27.33(5.45)
39.33(6.43)
46.00(4.65)
34.27(5.52)
41.47(5.89)
25.00(5.21)
22.20(4.78)
29.07(5.93)
L. R. Baggen, G. M. Gurr and A. Meats 392
Hymenoptera: Evolution, Biodiversity and Biological Control
A 30
Mean moths/trap
25
20
15
10
5
0 0
50
100
150
Distance from margin treatment (m)
B
30
Mean moths/trap
25
20
15
10
5
0 0
50
100
150
Distance from margin treatment (m) Figure 1
Effect of field margin treatment and distance from margin on trap catches of adult Phthorimaea operculella in potato crops bordered by: A) buckwheat; and B) borage (◆ traps on transect away from margin with flowers; ▲ traps on transect away from control margin without flowers; ■ traps in ‘laneway’ on transect away from intersection between flowers and control).
Results show that buckwheat is fed upon by moths and support the ‘selective’ nature of borage. Unlike the buckwheat flower, which is a shallow cup with exposed nectaries, borage flowers have a narrow corolla opening through which P. operculella is unable to crawl and a corolla depth greater than the length of the moth’s proboscis (Baggen et al. 1999). Crane et al. (1984) described borage as a morning producer of nectar, and its presence was observed on the exterior of these flowers, its escape being facilitated by their pendulous orientation. A similar build-up of nectar, resulting in its availability to Hymenoptera has previously been noted by Orr and Pleasants
Field Observations on Selective Food Plants in Habitat Manipulation 393
A 14 12
Mean wasps/trap
10 8 6 4 2 0 0
50
100
150
Distance from margin treatment (m)
B 14 12
Mean wasps/trap
10 8 6 4 2 0 0
50
100
150
Distance from margin treatment (m) Figure 2
Effect of field margin treatment and distance from margin on trap catches of adult Copidosoma koehleri in potato crops bordered by: A) buckwheat; and B) borage (◆ traps on transect away from margin with flowers; ▲ traps on transect away from control margin without flowers; ■ traps in ‘laneway’ on transect away from intersection between flowers and control).
(1996). However, because P. operculella is inactive during the day (Atherton 1936) this nectar is likely to have crystallised or have become too viscous for the moth to feed upon by the time adults become active at dusk (Traynier 1983). In contrast, hymenopteran adults with mandibulate mouthparts, such as C. koehleri, are able to feed on such foods so borage nectar, even when solid, is likely to benefit this important parasitoid. This effect was apparent in observations of parasitism within the borage field which were greater close to the flowering borage than elsewhere (Gurr et al. 1998b). The spatial pattern evident in trap catches in the present study also support the use
L. R. Baggen, G. M. Gurr and A. Meats 394
Hymenoptera: Evolution, Biodiversity and Biological Control
of both borage and buckwheat by C. koehleri since catches of wasps were high for traps close to flowers. The decline in wasp catches to levels equivalent to the ‘background’ level observed in areas of the crop remote from nectar sources occurred within approximately 50 m of the flowers. This indicates that the needs of this biological control agent may be met in fields of up to 1 ha (i.e. 100 m × 100 m) by flowers planted around the entire margin. For larger fields, however, additional strips spaced at approximately 100 m intervals through the crop itself may be required. Results from the present study also suggest that choice of plant in habitat manipulation attempts is more important than generally realised. Plants like buckwheat, with exposed nectaries may be fed upon by some pests, leading to exacerbation rather than alleviation crop damage. Cultural practices can reduce tuber damage caused by P. operculella, and these include mechanically ridging soil over tubers and an irrigation practice that avoids water uncovering tubers, yet is sufficient to prevent drying soil cracking. However, these procedures are not easy in practice and, even when achieved, cannot prevent foliar damage. Consequently, biological control is important. Because no benefit to the moth occurs from borage, selective plants such as this have potential in integrated management of P. operculella, via the conservation biological control contribution of C. koehleri.
Acknowledgements We thank W. Kingham and J. Plantinga for hosting field work and H.I. Nicol (NSW Agriculture) for statistical advice. This project was supported by a University of Sydney Postgraduate Award to LRB and a Department of Industry Science and Tourism grant to GMG.
References Atherton, D. O. (1936) Leaf miner and stem borer of tobacco in North Queensland. Queensland Agriculture Journal 45: 12-31. Baggen, L. R. & Gurr, G. M. (1998) The influence of food on Copidosoma koehleri, and the use of flowering plants as a habitat management tool to enhance biological control of potato moth, Phthorimaea operculella. Biological Control 11: 9-17. Baggen, L. R., Gurr, G. M. & Meats, A. (1999) Flowers in tri-trophic systems: mechanisms allowing selective exploitation by insect natural enemies for conservation biological control. Entomologia Experimentalis et Applicata 91: 155-161. Collins, F. L. & Johnson, S. J. (1985) Reproductive response of caged adult velvetbean caterpillar and soybean looper to the presence of weeds. Agriculture Ecosystems and Environment 14: 139-149. Crane, E., Walker, P. & Day, R. (1984) Directory of Important Honey Sources. International Bee Research Association, London. Dillard, H. E., Wicks, T. J. & Philip, B. (1993) A grower survey of diseases, invertebrate pests, and pesticide use on potatoes grown in South Australia. Australian Journal of Experimental Agriculture 33: 653-661. Greathead, D. J. & Greathead, A. H. (1992) Biological control of insect pests by insect parasitoids and predators: the BIOCAT database. Biocontrol News and Information 13: 61N-68N. Gurr, G. M. & Wratten, S. D. (1999) “Integrated biological control”: a proposal for enhancing success in biological control. International Journal of Pest Management 45: 81-84.
Field Observations on Selective Food Plants in Habitat Manipulation 395
Gurr G. M., van Emden, H. F. & Wratten, S. D. (1998a) Habitat manipulation and natural enemy efficiency: implications for the control of pests. pp 155-183. In Barbosa, P. (Ed.), Conservation Biological Control. Academic Press, San Diego. Gurr, G. M., Wratten, S. D., Irvin, N. A., Hossain, Z. , Baggen, L. R. , Mensah, R. K. & Walker, P. W. (1998b) Habitat manipulation in Australasia: recent progress and prospects for adoption. pp. 225-235. In Zaluki, M. P., Drew R. A. I. & White G. G. (Eds), Pest Management – Future Challenges: Proceedings of the Sixth Australian Applied Entomological Research Conference. 29 September-2 October 1998 Volume 2. The University of Queensland, Brisbane. Horne, P. A. (1990) The influence of introduced parasitoids on the potato moth, Phthorimaea operculella (Lepidoptera: Gelechiidae) in Victoria, Australia. Bulletin of Entomological Research 80: 159-163. Landis, D. A., Wratten, S. D. & Gurr, G. M. (2000) Habitat management to conserve natural enemies of arthropod pests in agriculture. Annual Review of Entomology 45: 175-201. Orr, D. B. & Pleasants, J. M. (1996) The potential of native prairie plant species to enhance the effectiveness of the Ostrinia nubilalis parasitoid Macrocentrus grandii. Journal of the Kansas Entomological Society 69: 133-143. Spooner-Hart, R. & Redgrove, H. (1995) IPM Report. Potato News 2: 6. Traynier, R. M. M. (1983) Influence of plants and adult food on the fecundity of potato moth, Phthorimaea operculella. Entomologia Experimentalis et Applicata 33: 145-154. Wratten, S. D., van Emden, H. F. & Thomas, M. B. (1998) Within field and border refugia for the enhancement of natural enemies. pp. 375-403. In Picket, C. H. & Bugg, R.L. (Eds), Enhancing Natural Control of Arthropod Pests Through Habitat Management. University of California Press, Berkeley.
Understorey Management for the Enhancement of the Leafroller Parasitoid Dolichogenidea tasmanica (Cameron) in Orchards at Canterbury, New Zealand N. A. Irvin1, S. D Wratten1 and C. M. Frampton2 1
Ecology and Entomology Group, P. O. Box 84, Lincoln University, New Zealand (email:
[email protected]) 2
Applied Computing, Mathematics and Statistics Group, P.O. Box 84, Lincoln University, New Zealand.
Introduction Leafrollers (Lepidoptera: Tortricidae) are major pests of pipfruit and other crops in New Zealand and elsewhere and have a high pest status in ‘conventionally managed’ crops, mainly due to the zero tolerance of their presence in export markets and the high requirement for blemish-free fruit (Wratten et al. 1998). Natural enemies have not exhibited commercially useful effects on leafroller populations in orchards because the intensive use of conventional insecticides kills them and frequent herbicide use has removed the plants that could provide the habitats, and pollen and nectar resources required by many (van Driesche & Bellows 1996; Barbosa & Wratten 1998). Recently, the adverse effects of intensive pesticide use have become the focus of a major concern for the New Zealand fruit industry. Increasing demands for non-detectable or extremely low levels of residues in exported New Zealand fruit, combined with quarantine requirements for virtually pest-free produce, provide a ‘double-bind’ for producers and exporters (MacIntyre et al. 1989; Wratten et al. 1998). Therefore, the New Zealand fruit industry needs to move towards a more integrated pest management approach. In fact, the first priority of ENZAFRUIT New Zealand is to give preference to non-chemical methods of pest and disease control, leading to a decrease in agrichemical usage (Wearing 1996). Integrated Fruit Production (IFP) is becoming a requirement of ENZAFRUIT New Zealand and the markets that it supplies (Wratten et al. 1998). This programme was described by Batchelor et al. (1997) and requires growers to use monitoring procedures and action thresholds to determine when pesticide application is essential. The introduction of the insect growth regulator Mimic® (tebufenozide) into IFP programmes has allowed a ‘window’ for utilising natural enemies for leafroller biological control. This compound is both selective and able to meet export quarantine requirements (Walker et al. 1991). However, information on understorey management is still required for the IFP-P manual (Batchelor et al. 1997; Walker et al. 1997) which, if successful, may result in enhanced control through the additive effects of biological control agents and tebufenozide. Conservation biological control (the provision of resources to natural enemies to improve their effectiveness at controlling pests; Bugg & Pickett 1998) has potential in orchards and recent reviews emphasise this (Bugg & Pickett 1998; Barbosa 1998; Landis et al. 2000). One mechanism is the
396
Hymenoptera: Evolution, Biodiversity and Biological Control
Understorey Management for the Enhancement of Dolichogenidea Tasmanica 397
enhancement of nectar and pollen resources. For example, buckwheat (Fagopyrum esculentum Moench), a herbaceous dicotyledonous annual in the family Polygonaceae with small white flowers with shallow corollae, provides pollen and nectar to a wide range of beneficial insects (Lövei et al. 1993), including short-tongued parasitoids. Buckwheat has a short sowing-to-flowering time (Bowie et al. 1995) and its seeds are cheap and available in New Zealand and elsewhere. Work with buckwheat as an understorey option in a Canterbury orchard led to significantly higher levels of parasitism by Dolichogenidea tasmanica (Cameron) (Braconidae) of released leafroller larvae in buckwheat plots than in controls (Stephens et al. 1998). Also, higher numbers of D. tasmanica were captured on yellow sticky traps in buckwheat plots. However, the replicate number and plot sizes used in the study were small and further investigation is needed into other understorey options, in conjunction with larger plot sizes and an increase in replicate number, to determine the potential for enhancing leafroller parasitoids and their effects. The nectar of the flowers of broad (faba) bean (Vicia faba L.) is probably inaccessible to shorttongued parasitoids. However, a rich assemblage of ichneumonids attends extra-floral nectaries of broad bean, including parasitoids of agricultural and forest pests (Bugg et al. 1989). Broad bean has the added advantages of being able to be planted at any time of the year and the extrafloral nectaries are present soon after germination when the seedlings are only a few centimetres tall. Therefore they may be useful for early-spring enhancement of natural enemy populations. This research aimed to investigate the influence of buckwheat and broad bean as floral resources on leafroller parasitoid abundance, and consequent leafroller parasitism by D. tasmanica in apple orchards not receiving organo-phosphorous pesticide treatments in Canterbury, New Zealand. The work extends that of Stephens et al. (1998) by using broad bean as an additional understorey option, by separating the floral effect that the buckwheat provides from the potential shelter effect (by removing buckwheat flowers as one treatment) and by using more replicates and larger plot sizes.
Methods Site description and experimental design Four replicates of each of four treatments (buckwheat with flowers, buckwheat without flowers, broad beans and herbicide-treated control) were set up in a randomised block design in an apple orchard at the Lincoln University Horticultural Research Area, Canterbury, New Zealand. Each replicate was 7.4 m × 2 m and consisted of 8-year-old apple trees (cvs. Braeburn, Royal Gala and Cox’s Orange). Buckwheat (cv. Shinano Natsu) seeds were sown on December 6, 1996 and on January 6, 1997 to ensure continual flowering. Broad beans (cv. Exhibition Long Pod) were sown on November 11, 1996 and on January 6, 1997. In these plots, a 50 mm wide strip of buckwheat or a row of broad beans was sown each side of the trees at a depth of approximately 30–40 mm. Buckwheat was sown at two seeds/cm and broad beans were spaced approximately 3 cm apart. The orchard had a history of broad-spectrum, persistent herbicide use, with simazine being applied each year. The orchard, therefore was virtually devoid of ground vegetation. Because of the simazine use, the top 20 mm of soil was removed from each of the areas which were due to receive seeds prior to sowing. Approximately six weeks after sowing (mid-January, 1997), the buckwheat began flowering and parasitoid trapping and leafroller release began. Flowers on the ‘buckwheat without flowers’ treatment were removed by hand every 3–5 d, and on February 14 the top third of the plants in the first buckwheat drilling in the ‘buckwheat with flowers’ plots was
N. A. Irvin, S. D Wratten and C. M. Frampton 398
Hymenoptera: Evolution, Biodiversity and Biological Control
cut off to encourage lateral flowering and to ensure a longer flowering time, while the second drilling was left uncut.
Parasitoid trapping One yellow ‘Trappit’ (Agrisense-BCS-Ltd, Treforest Industrial Estate, Pontypridd, Mid-Gladmorgan, U. K.) trap (catching surface: 200 × 245 mm) was placed 1 m above the ground, in the centre of each plot, facing the middle of the row. After 9-14 d between 31 January and 19 March, 1997, these were collected and replaced. The numbers of D. tasmanica were counted under a binocular microscope (20× magnification) in the laboratory. Egg release Eggs of the lightbrown apple moth, Epiphyas postvittana (Walker), were placed on a branch in the centre of each plot every 15–20 d from 14 January to 19 March, 1997. Stephens et al. (1998) released leafroller larvae rather than eggs. However, batches of eggs are easier to count and handle and because the larvae hatch under field conditions they may be better acclimatised to the field environment. Eggs laid on paper in the laboratory were divided into batches, of approximately 150 eggs, by counting the eggs and cutting the paper under a binocular microscope. The paper pieces were stapled to the underside of a leaf of a branch on the middle tree in each plot (150 eggs per branch). A nylon material sleeve (600 × 200 mm) was placed over the branch and closed with string to allow hatched larvae to settle and to protect them and the eggs from predators and egg parasitoids, such as Trichogramma spp. Sleeves were removed after 2–5 d, leaving the larvae exposed to parasitoids. The branches were collected 6-8 weeks after egg placement and the number of parasitoid cocoons and leafroller larvae and pupae present were recorded. Parasitism rate was expressed as: number of parasitoid cocoons recovered/(number of parasitoid cocoons recovered + number of leafroller larvae and pupae recovered) × 100. Trap catches and parasitism rates were compared between treatments and times using repeated measures ANOVA. Where the ANOVA indicated significant main or interaction effects, these were further explored using Fisher’s least significant difference test. Trap catch data were logtransformed (logex+1) before analysis to stabilise variances and are therefore reported as geometric means.
Results Parasitoid trap catches There was an overall significant (F = 7.75, df = 3,9, P < 0.01) effect of understorey management treatment on the abundance of leafroller parasitoids; their numbers were significantly (P < 0.001) higher in the ‘buckwheat with flowers’ treatment (1.27/trap/d) and in the broad bean treatment (0.52) compared with the control (0.18). The ‘buckwheat with flowers’ treatment enhanced the abundance of leafroller parasitoids significantly (P < 0.001) more than did the broad beans. There was no significant difference (P > 0.05) in leafroller parasitoid trap catches between the ‘buckwheat without flowers’ treatment (0.24/trap/d) and the control (0.18) (see Fig. 3 in Gurr et al. 1998). Leafroller parasitism rate On one of the leafroller release dates (13 February, 1997) there was an overall significant (F = 8.32, df = 3,6, P < 0.05) effect of treatment on parasitism rate (Fig. 1). There was a significantly (P < 0.05) higher parasitism rate in the ‘buckwheat with flowers’ treatment (86%), the
Understorey Management for the Enhancement of Dolichogenidea Tasmanica 399
Figure 1
Leafroller parasitism rate by Dolichogenidea tasmanica in each understorey management treatment for each egg release date (* = P < 0.05, ** = P < 0.01, BW = buckwheat).
broad beans treatment (75%) and the ‘buckwheat without flowers’ treatment (96%) compared with the control (0%). These results were based on 34 cocoons and 15 larvae recovered from the 13 February, 1997 release date. There was a trend of a higher parasitism rate in the ‘buckwheat with flowers’ treatment compared with the control on the 29 January, 1997 and 5 March, 1997 release dates. However, this was not significant (29 January, 1997, F = 2.22, df = 3,9, P > 0.05; 5 March, 1997, F = 0.054, df = 3,10, P > 0.05). Combining all release dates, there were no significant (F = 1.46, df = 3,30, P > 0.05) differences in parasitism rate between treatments.
Discussion Parasitoid trap catches More than seven times as many leafroller parasitoids were trapped in the buckwheat plots compared with controls, suggesting that buckwheat may enhance parasitoid abundance for leafroller biocontrol. These results support the work of Stephens et al. (1998). The ‘buckwheat with flowers’ treatment significantly enhanced parasitoid numbers compared with the controls, whereas there was no significant difference between control plots and ‘buckwheat without flowers’. This suggests that it is the buckwheat floral resources that lead to enhanced parasitoid numbers, and not the shelter, which the buckwheat also presumably provides to some natural enemies. Broad beans also enhanced parasitoid abundance, although to a lesser extent than buckwheat. This may be because buckwheat flowers (with their shallow corollae) provide both pollen and nectar to the parasitoid (Lövei et al. 1993), whereas broad bean plants provide only nectar, via extra-floral nectaries, because their flowers are too large for short-tongued parasitoids to gain access to floral nectar or pollen (Bugg et al. 1989). Although the plot size used in this study was increased to extend the work of Stephens et al. (1998), it may not have been large enough to prevent parasitoids moving between nectar resources and between all plots/treatments. However, it has shown that trap catches do not differ
N. A. Irvin, S. D Wratten and C. M. Frampton 400
Hymenoptera: Evolution, Biodiversity and Biological Control
significantly between small (7.4 × 2 m) and large (40 × 5 m) buckwheat plots, and that catches from traps placed at a gradient of distances from a buckwheat plot show that D. tasmanica trap catches are significantly higher than in controls for up to 2 m from the nectar source (Irvin unpublished). Buffer zones used in the current study extended beyond 10 m and therefore would have negligible numbers of parasitoids moving between treatments.
Leafroller parasitism rate Buckwheat significantly enhanced leafroller parasitism by up to 68% compared with the control, indicating that the control potential of these parasitoids is high. In fact, Zandstra and Motooka (1978) showed that only 3.5% of apples were infested with codling moth larvae where buckwheat was present, compared with 1.5% with full chemical control and 54% with no treatment. Orchards in Russia are commonly undersown with buckwheat because its flowers are a food source for adult Trichogramma spp., which attack codling moth eggs (Zandstra & Motooka 1978). For D. tasmanica, Dumbleton (1935) found that parasitism in Nelson, New Zealand, ranged from 20 to 50%, and Thomas (1965) found rates ranging between 4 and 28% in unsprayed Canterbury orchards. The differences between studies may be accounted for by the differences in the methods of leafroller release and the techniques for calculating parasitism rate. For instance, different release densities, sleeve types and leafroller developmental stages may have affected establishment. In the current work, trap catch data indicates that floral resources are the dominant factor influencing parasitoid numbers. However, parasitism rate data were variable and most leafroller parasitism rates did not differ between treatments. On the release date where treatments did differ, all treatments were significantly different from the control. For example, the buckwheat without flowers treatment appeared to give parasitism rates as high as those for the ‘with flowers’ plots. This is probably an artefact associated with the methodology. The high dispersal rates of the leafroller larvae (Penman 1984) will tend to make recovery of parasitoid cocoons from the release sites variable, making it difficult to detect differences between treatments. Of those cocoons and larvae that were recovered, a high proportion had been parasitised, giving apparent high parasitism rates. This may have been because those larvae that are not parasitised are more likely to disperse or drop to the ground to pupate, whereas those that have been parasitised may be less mobile. Currently, research is under way to determine which leafroller release technique is the best method to achieve a realistic assessment of parasitism rate. These methods involve the release of leafroller eggs in sleeves, as above, and the recovery of individual released larvae, after two and four weeks. Larvae are then reared individually on artificial diet and the proportion subsequently parasitised is recorded (Irvin unpublished). Understorey management potential and future work Buckwheat shows great potential as an understorey management option for leafroller biocontrol in apple orchards in the Canterbury region. It establishes well, has a short sowing-to-flowering time (Bowie et al. 1995) and is cheap and available in New Zealand. The introduction of Mimic® (tebufenozide), an insect growth regulator, into IFP programmes will enhance the role of natural enemies, given the minimal non-target effects of this compound (Walker et al. 1991). Using buckwheat in orchard understoreys could be a vital part of IFP programmes, by enhancing the effectiveness of natural enemies for leafroller biocontrol. However, future research into the practical use of buckwheat and/or other plant species in this way is required. The number, size and spacing of the plots of conservation biocontrol plantings need to be determined (Landis et al. 2000). Also, some mark/release and recapture work may be beneficial to track the scale of movement of D. tasmanica and other biological control agents, and may therefore help to determine the optimal buffer zone to use between treatments in future experiments.
Understorey Management for the Enhancement of Dolichogenidea Tasmanica 401
The mechanisms by which buckwheat may enhance leafroller parasitoid abundance and increase parasitism rates of leafrollers are unknown. Results from this study indicate that it is the flowers that influence parasitism rate. However, the question remains whether providing buckwheat in the orchard understorey influences other factors, such as increasing the ‘fitness’ of parasitoids. Field trials are currently being conducted to investigate the effects of understorey management on the sex ratio of emerging parasitoids. Also, choice and no-choice laboratory experiments are being conducted to rank D. tasmanica ‘preference’ for various floral resources, and to investigate the influence of these resources on fecundity and longevity. Buckwheat is an annual plant and would therefore have to be resown each year. Self-sowing is not a viable option as the plant is frost-tender (Bowie et al. 1995) and pre-emergent herbicide use in the understorey inhibits germination. Also, annual-plant understorey options may not suite some orchard growers. Therefore, a large replicated experiment has been set up at Lincoln University, to determine whether alyssum (Lobularia maritima (L.) Desv.) is a suitable perennial understorey option for the enhancement of leafroller parasitism by D. tasmanica. Twenty-two flowering plant species, including phacelia (Phacelia tanacetifolia Benth.), buckwheat, broad bean and coriander (Coriandrum sativum L.), were recently ranked in California for their potential use as in-field insectaries; numbers of beneficial insects and pest species being recorded (Chaney 1998). Alyssum showed the greatest potential because no other plant tested flowered as quickly from seed or attracted as many beneficial insect species. The abundance of Anacharis sp., a parasitoid of Tasmanian lacewing, Micromus tasmaniae (Walker), can also be enhanced by undersowing buckwheat (Stephens et al. 1998). Baggen et al. (2000) demonstrated that buckwheat enhanced the fecundity of the potato moth (Phthorimaea operculella (Zeller)), whereas no such benefit occurred when borage (Borago officinalis L.) was used. Therefore, it is important to determine the influence of understorey plants on pest species and on hyperparasitoids of natural enemies. Future research into the potential of understorey management to enhance other parasitoids of leafrollers is required. For example, the egg parasitoids Trichogramma spp. may be a better agent than the larval parasitoid D. tasmanica since the pest is killed before any feeding damage occurs and therefore may suit the zero tolerance requirements by ENZA. However, perhaps the biggest challenge is to determine whether conservation biocontrol can enhance leafroller parasitism to an extent that it reduces leafroller populations below economic thresholds for local and export apple markets.
Acknowledgements We thank the Agricultural and Marketing Research Development Trust (AGMARDT), New Zealand and Lincoln University, New Zealand for financial support and Keith MacIntosh, Lincoln University, for his assistance in trial establishment and maintenance. We are also grateful to Graham Burnip and Max Suckling, of The Horticultural and Food Research Institute of New Zealand Ltd. for advice and support.
References Baggen, L. R., Gurr, G. M. & Meats, A. (2000) Field observations on selective food plants in habitat manipulation for biological control of potato moth, Phthorimaea operculella (Lepidoptera: Gelechiidae) by Copidosoma koehleri (Hymenoptera: Encyrtidae). pp. 388-395 this volume.
N. A. Irvin, S. D Wratten and C. M. Frampton 402
Hymenoptera: Evolution, Biodiversity and Biological Control
Barbosa, P. (Ed.) (1998) Conservation Biological Control. Academic Press, London. Barbosa, P. & Wratten, S. D. (1998) Influence of plants on invertebrate predators: Implication to conservation biological control. pp. 83-100. In Barbosa, P. (Ed.), Conservation Biological Control. Academic Press, San Diego. Batchelor T. A., Walker, J. T. S., Manktelow, D. W. L., Park, N. M. & Johnson, S. R. (1997) New Zealand Integrated Fruit Production for pipfruit – Charting a new course. Proceedings of the Fiftieth New Zealand Plant Protection Conference 1997: 14-19. Bowie, M. H., Wratten, S. D. & White, A. J. (1995) Agronomy and phenology of “companion plants” of potential for enhancement of insect biological control. New Zealand Journal of Crop and Horticultural Science 23: 423-427. Bugg, R. L., Ellis, R. T. & Carlson, R. W. (1989) Ichneumonidae (Hymenoptera) using extrafloral nectar of faba bean (Vicia faba L., Fabaceae) in Massachusetts. Biological Agriculture and Horticulture 6: 107-114. Pickett, C. H. & Bugg, R. L. (Eds) (1998). Enhancing Biological Control. University of California Press, Berkley. Chaney, W. E. (1998) Biological control of aphids in lettuce using in-field insectaries. pp. 73-84. In Pickett, C. H. & Bugg, R. L. (Eds), Enhancing Biological Control – Habitat Manipulation to Promote Natural Enemies of Agricultural Pests. University of California Press, Berkeley. Driesche, R. G. van & Bellows, T. S. Jr. (1996) Biological Control. Chapman and Hall, New York. Dumbleton, L. J. (1935) Apanteles tasmanica Cam.: a braconid parasite of leafroller larvae. The New Zealand Journal of Science and Technology 18: 572-576. Gurr, G. M., Wratten, S. D., Irvin, N. A., Hossain, Z., Baggen, L. R., Mensah, R. K. & Walker, P. W. (1998) Habitat manipulation in Australasia: recent progress and prospects for adoption. Proceedings of the 6th Australasian Applied Entomological Research Conference, Brisbane, Australia: Pest Management – Future Challenges, September 1998. 1: 225-235. Landis, D. B., Wratten, S. D. & Gurr, G. M. (2000) Habitat manipulation for natural enemies. Annual Review of Entomology 45: 175-201. Lövei, G. L., Hodgson, D. J., MacLeod, A. & Wratten, S. D. (1993) Attractiveness of some novel crops for flower-visiting hoverflies (Diptera: Syrphidae): Comparisons from two continents. pp. 368-370. In Corey, S., Dall, D. & Milne, W. (Eds), Pest Control and Sustainable Agriculture. CSIRO Publications, Melbourne. MacIntyre, A. A., Allison, N. & Penman, R. (1989) Pesticides: Issues and Options for New Zealand. Ministry for the Environment, Auckland. Penman, D. R. (1984) Deciduous fruit tree pests. pp. 33-49. In Scott R. R. (Ed), New Zealand Pests and Beneficial Insects, Lincoln University College of Agriculture, Caxton Press, Christchurch. Stephens, M. J., France, C. M., Wratten, S. D. & Frampton, C. (1998) Enhancing biological control of leafroller (Lepidoptera: Tortricidae) by sowing buckwheat (Fagopyrum esculentum) in an orchard. Biocontrol, Science and Technology 8: 547-558. Thomas, W. P. (1965) Studies on Epiphyas postivttana (Walker) (Lepidoptera: Tortricidae) and its parasitoid complex. Master of Agricultural Science thesis, Lincoln University. Walker, J. T. S., Baynon, G. T. & White, V. (1991) Insect control on apples with RH-5992, a novel insect growth regulator. Proceedings of the 44th Conference of the Plant Protection Society of New Zealand: 66-69. Walker, J. T. S., Hodson, A. J., Wearing, C. H., Bradley, S. J., Shaw, P. W., Tomkins, A. R., Burnip, G. M., Stiefel, H. E. & Batchelor, T. A. (1997) Integrated Fruit Production for New Zealand pipfruit: Evaluation of pest management in a pilot programme. Proceedings of the 50th New Zealand Plant Protection Conference 1997: 258-263.
Understorey Management for the Enhancement of Dolichogenidea Tasmanica 403
Wearing, C. H. (1996) Evaluation of pests and their natural enemies in conventional, integrated and biological fruit production systems in Central Otago 1995-96. HortResearch Client Report No. 96/191. Wratten, S. D., Wearing, C. H., King, D. & Davis, S. (1998) Market-driven IPM in New Zealand – has it delivered? Proceedings of the Sixth Australasian Applied Entomological Research Conference, Brisbane, Australia: Pest Management – Future Challenges, September 1998. 1: 153-159. Zandstra, B. H. & Motooka, P. S. (1978) Beneficial effects of weeds in pest management – a review. PANS 24: 333-338.
Impact and Control of Introduced Vespula Wasps in New Zealand Jacqueline R. Beggs Landcare Research, Private Bag 6, Nelson, New Zealand (email:
[email protected])
Introduction Throughout the world, social insects have been highly successful invaders. These insects can pose a massive threat to their host community (Howarth 1985; Wojcik 1994; Moller 1996), as well as an economic and health threat to human communities. The reproductive and dispersal strategies of these invaders mean they are a formidable foe. They often become abundant and efficient competitors and/or predators in a receiving community. They have the potential to extirpate prey species because they are polyphagous with abundant, alternative food resources (Holt & Lawton 1994). Therefore, social insects can markedly restructure communities (Paine 1974). New Zealand now has two accidentally introduced Vespula species, the German wasp (V. germanica (F.)) and the common wasp (V. vulgaris (L.)). The German wasp arrived in 1945, about 30 years before the common wasp. Both species are now distributed throughout New Zealand in a wide range of habitats (Clapperton et al. 1989b, 1994) and over a wide range of altitudes (Beggs 1991; Fordham 1991). New Zealand has a very limited social insect fauna. Indeed, there are no native social wasps or bees (Valentine & Walker 1991). Thus, the arrival of alien social wasps met with limited biotic resistance. There were very few predators, competitors or diseases, and an abundant supply of food – a wasp paradise. Both species of social wasp are considered to be an economic pest of primary industries such as beekeeping, forestry and horticulture. For example, 80% of beekeepers who responded to a questionnaire considered wasps to be a nuisance (Clapperton et al. 1989a). Wasps totally destroyed or seriously affected almost 10% of hives, which translates to a significant financial loss (Clapperton et al. 1989a). Furthermore, wasps can be a major social pest as they disrupt people’s enjoyment of the outdoors, the operation of some schools, and are a health threat.
Honeydew Beech Forests The common wasp has almost completely displaced the German wasp from some habitats, particularly beech forests infested with honeydew (Harris et al. 1991; Clapperton et al. 1994), hereafter referred to as honeydew beech forest. Wasps can become very abundant in these forests (Thomas et al. 1990) and there is increasing concern about their status as a conservation pest. Honeydew is a sugary fluid produced by endemic scale insects (Ultracoelostoma spp.). The scale insects live in hard capsules in the bark of trees (protected from the wasps), and insert their mouthparts into the sap vessels of the tree (Morales et al. 1988). They siphon the sap and excrete
404
Hymenoptera: Evolution, Biodiversity and Biological Control
Impact and Control of Introduced Vespula Wasps in New Zealand 405
the sugary excess via a long waxy filament. Millions of the insects can infest a hectare of beech forest, thus providing an abundant sugar supply for wasps. Wasps are more prevalent where there is honeydew. Wasp numbers were indexed using 20 traps baited with sardine catfood at 25 sites in the northern South Island. Traps in honeydew beech forest caught much higher numbers of wasps than traps at sites without honeydew (Beggs in press). A Geographic Information System (GIS) was used to model the distribution of the scale insect using published geographic limits, altitudinal limits and estimated rainfall limits. Areas of forest containing beech, beech-scrub mixtures, and beech-grassland mixtures were filtered in a sequence of overlays to exclude those areas outside the geographic or ecological range of the insect. This provided a crude estimate that honeydew occurs over about 1 000 000 ha of land in New Zealand (Beggs in press). Thus, honeydew beech forest represents about 15% of indigenous forest cover in New Zealand (6 300 000 ha, Newsome 1987). This figure is only an approximation, and survey work is required to determine the real extent of this resource. The mean density of wasp nests at 19 honeydew beech forest sites was 12 nests/ha and there were about 1500 wasps per nest at the peak of the wasp season (March; Thomas et al. 1990). Thus, at the peak of the wasp season, honeydew beech forests contain about 18 × 109 wasps. The average biomass of wasps at the peak of the wasp season is about 3.8 kg/ha (this includes brood and workers) in honeydew beech forest (Thomas et al. 1990). This equates to about 4 million kg of wasps. However, wasp abundance can vary by an order of magnitude in space and time, e.g. in one year, nest density at 19 sites ranged from 1 to 33 nests/ha (Thomas et al. 1990). Between 50 and 65% of foraging wasps collect honeydew, and they are able to carry loads of about 15 µl of honeydew at a time (Harris 1991). Thus, the estimated intake of honeydew was between 80 and 340 litres/ha/season at two sites, depending on the density of wasps. Exclosure experiments have demonstrated that wasps reduced the standing crop of honeydew by more than 99% for about 4 months, and 90% for a further 2 months (Moller et al. 1991b). This loss of honeydew means there is little left for native organisms, including birds, reptiles, invertebrates and micro-organisms, all of which feed on honeydew (Beggs in press). A lack of honeydew may lead to a range of responses by birds from simply switching to an alternative food source (if there is one available), to reducing reproductive output. Mortality rates may also increase because of insufficient food. Kaka (Nestor meridionalis meridionalis) are a threatened endemic forest parrot that feed on honeydew. It is energetically unprofitable for them to collect honeydew when wasps are numerous (Beggs & Wilson 1991). Thus, it was suggested that wasps were one of a combination of introduced predators and competitors that were causing a decline in reproductive success of this parrot (Beggs & Wilson 1991). While subsequent research has shown predation by introduced stoats (Mustela erminea) is likely to be a primary cause in the decline of kaka (Wilson et al. 1998), it is possible that wasps still play an important role. It is estimated that wasps remove by direct predation about 1–8 kg of native invertebrates/ha/ year in honeydew beech forests, mostly spiders (30%), caterpillars (20%), ants and bees (20%), and flies (15%) (Harris 1991). As well as a potential impact on populations from these groups, wasps could have a flow-on effect to insectivorous birds. Predation rates have been measured for a species of free-living caterpillar (Beggs & Rees 1999), and for an orb-web spider (Toft & Rees 1998). In both cases rates were so high that the probability
Jacqueline R. Beggs 406
Hymenoptera: Evolution, Biodiversity and Biological Control
of a caterpillar surviving to an adult in the wasp season was virtually nil (10-78 to 10- 40) and the probability of a spider surviving to the end of the wasp season was also virtually nil (10-18). Both studies concluded that wasps needed to be reduced by about 80 to 90% in order to conserve such vulnerable species. Not all species will be equally vulnerable to wasp predation, so these estimates provide a maximum level of control required to conserve some species. Achieving a lower level of control is likely to be beneficial to less vulnerable species.
Wasp Control Wasp numbers may need to be reduced by 80–90% for conservation purposes, but there are no estimates of the reduction required for economic or social reasons. Nevertheless, there are repeated requests for effective control tools from a variety of sectors. There are two main strategies being developed: poison-baiting for short-term, localised control; and biological control for self-sustaining, widespread control.
Poison-baiting Toxins such as 1080 (Spurr 1991) and sulfluramid (Spurr 1993) mixed with a sardine catfood bait can be effective in controlling wasps. Poison-baiting can kill 80–100% of the colonies within a site, but reinvasion by foraging workers means that total wasp biomass is not reduced by as much (Beggs et al. 1998). The extent of reinvasion will depend on the size of the site poisoned, but since wasps have been recorded foraging up to 4 km from their nest (Coch 1972), a site would have to be very large (perhaps 2000 ha) before reinvasion at the core was not a problem. Some queen wasps are estimated to fly up to 30–70 km before establishing a nest (Moller et al. 1990), so reinvasion in spring of even a 2000 ha site would be likely. Furthermore, in some habitats, particularly non-honeydew beech forest sites, wasps are not greatly attracted to a proteinbased bait. Some poison-baiting operations have failed to reduce wasp abundance for this reason. It is not feasible to use a carbohydrate bait because of the risk of poisoning honeybees. Even in honeydew beech forest, it is difficult to poison wasps early in the season (before January), because they are not attracted to bait.
Biological control Since 1987, more than 200 000 overwintering cocoons of a wasp parasitoid (Sphecophaga vesparum vesparum) have been released in 968 release boxes in 12 out of 13 regions of New Zealand (Beggs et al. 1996). In addition, almost 270 wasp nests at 65 sites had a piece of wasp comb inserted which contained pre-emergent parasitoid cocoons. The parasitoid has established in two out of 26 sites checked (Moller et al. 1991a; Beggs et al. 1996). Despite being established at one site for more than 10 years, there has been no measurable reduction in wasp abundance (Beggs et al. 1996; unpublished). Mathematical modelling predicts that nest density will be reduced by up to 25%, but only if the parasitoid kills about 50% of early season nests (Barlow et al. 1996; Toft et al. 1999). If fewer early season nests are killed, then the predicted reduction rapidly reduces to around zero (Toft et al. 1999). The parasitoid can take up to 4 years to emerge from a cocoon (86% emerge in year 3), so it could be at least 20 years before the parasitoid population reaches an equilibrium and a reduction in wasp numbers is detectable (Toft et al. 1999). Two other strains of S. vesparum have been released since 1996 (Harris & Read 1998), but neither have yet established, and it is not known whether they will have a greater or lesser impact on wasp populations than the original strain.
Impact and Control of Introduced Vespula Wasps in New Zealand 407
New developments The challenge now is to identify and develop additional wasp control tools. There are a range of fungi, bacteria, nematodes, protozoa, and viruses which have been associated with nest material of wasps (Rose et al. 1999). Wasps are susceptible to some of these pathogens, such as the fungus Beauveria bassiana (Harris et al. 2000). Fungal spores can be transferred between workers, and between workers and larvae (Harris et al. 2000), a pre-requisite if a pathogen is to control wasps. A technique is being developed to incorporate a pathogen into a bait. The challenge is to find a way for the pathogen to overcome the microbial defences of a colony so that it can spread between nest mates (Harris pers. comm.). The advantage over using a toxin is that not as much would have to be taken back to the nest, so it would work at much lower doses. A new toxin, fipronil®(Aventis), is also being investigated for use in poison-baiting. Trials to date have shown that fipronil is fast acting and effective at low concentrations (Harris unpublished). In 1999, a large scale trial (300 ha) in honeydew beech forest conducted by the New Zealand Department of Conservation indicated that it is effective even when wasp density is relatively low. Preliminary results indicate that more than 90% of nests were killed (Butler pers. comm.). Much has been achieved in quantifying the wasp problem and in developing control strategies. However, there is still much work to be done. It is likely that a range of control tools will be necessary to achieve an adequate reduction in wasp abundance in a range of habitats.
Acknowledgements This review is only possible because of more than a decade of work by a team of researchers. Much of the work was initiated by Henrik Moller (ecology), Eric Spurr (poison-baiting) and Barry Donovan (Sphecophaga vesparum). Richard Harris has continued to develop many of the control strategies, including initiating the work on pathogens. Special thanks are owed to Jocelyn Tilley, Jo Rees, Richard Toft, Peter Read and Jason Malham for years of dedicated collection of data (and stings!). The Foundation for Research, Science and Technology and the Department of Conservation have funded much of the research this review is based on. An early draft of this manuscript was improved by Richard Harris.
References Barlow, N. D., Moller, H. & Beggs, J. R. (1996) A model for the effect of Sphecophaga vesparum vesparum as a biological control agent of the common wasps in New Zealand. Journal of Applied Ecology 33: 31–44. Beggs, J. R. (1991) Altitudinal variation in abundance of common wasps (Vespula vulgaris). New Zealand Journal of Zoology 18: 155–158. Beggs, J. R. (in press) The ecological consequences of social wasps (Vespula spp.) invading an ecosystem which has an abundant carbohydrate resource. Biological Conservation. Beggs, J. R., Harris, R. J. & Read, P. E. C. (1996) Invasion success of the wasp parasitoid Sphecophaga vesparum vesparum (Curtis) in New Zealand. New Zealand Journal of Zoology 23: 1–9. Beggs, J. R. & Rees, J. S. (1999) Restructuring of Lepidoptera communities by introduced Vespula wasps in a New Zealand beech forest. Oecologia 119: 565-571. Beggs, J. R., Toft, R. J., Malham, J. P., Rees, J. S., Tilley, J. A. V., Moller, H. & Alspach, P. (1998) The difficulty of reducing introduced wasp (Vespula vulgaris) populations for conservation gains. New Zealand Journal of Ecology 22: 55–63.
Jacqueline R. Beggs 408
Hymenoptera: Evolution, Biodiversity and Biological Control
Beggs, J. R. & Wilson, P. R. (1991) The kaka, Nestor meridionalis, a New Zealand parrot endangered by introduced wasps and mammals. Biological Conservation 56: 23–38. Clapperton, B. K., Alspach, P. A., Moller, H. & Matheson, A. G. (1989a) The impact of common and German wasps (Hymenoptera: Vespidae) on the New Zealand beekeeping industry. New Zealand Journal of Zoology 16: 325–332. Clapperton, B. K., Moller, H. & Sandlant, G. R. (1989b) Distribution of social wasps (Hymenoptera: Vespidae) in New Zealand in 1987. New Zealand Journal of Zoology 16: 315–323. Clapperton, B. K., Tilley, J. A. V., Beggs, J. R. & Moller, H. (1994) Changes in the distribution and proportions of Vespula vulgaris (L.) and Vespula germanica (Fab.) (Hymenoptera: Vespidae) between 1987 and 1990 in New Zealand. New Zealand Journal of Zoology 21: 295–303. Coch, F. (1972) Probleme der Wespenbekämpfung in Bäckereien un Konditoreien. Bäcker Konditor 26: 246–248. Fordham, R. A. (1991) Vespulid wasps at the upper forest margin in Tongariro National Park – a threat to the native biota? New Zealand Journal of Zoology 18: 151–153. Harris, R. J. (1991) Diet of the wasps Vespula vulgaris and V. germanica in honeydew beech forest of the South Island, New Zealand. New Zealand Journal of Zoology 18: 159–170. Harris, R. J., Harcourt, S. J., Glare, T. R., Rose, E. A. F. & Nelson, T.J. (2000) Susceptibility of Vespula vulgaris and V. germanica (Hymenoptera: Vespidae) to generalist entomopathogenic fungi and their potential for wasp control. Journal of Invertebrate Pathology 75: 251-258. Harris R. J. & Read P. E. C. (1998) Enhanced biological control of wasps. Unpublished contract report to Department of Conservation. No LC9798/134. Harris, R. J., Thomas, C. D. & Moller, H. (1991) The influence of habitat use and foraging on the replacement of one introduced wasp species by another in New Zealand. Ecological Entomology 16: 441–448. Holt, R. D. & Lawton, J. H. (1994) The ecological consequences of shared natural enemies. Annual Review of Ecological Systems 25: 495–520 Howarth, F. G. (1985) Impacts of alien land arthropods and molluscs on native plants and animals in Hawaii. pp 149-173. In Stone, C. P. & Scott, J. M. (Eds), Hawaii’s Terrestrial Ecosystems Preservation and Management. University of Hawaii Press, Honolulu. Moller, H. (1996) Lessons for invasion theory from social insects. Biological Conservation 78: 125–142. Moller, H., Beggs, J. R., Tilley, J. A. V., Toft, R. J., Wilson, N. J. & Alspach, P. A. (1990) Ecology and control of wasp populations in New Zealand. Unpublished DSIR Land Resources Contract Report. Moller, H., Plunkett, G. M., Tilley, J. A. V., Toft, R. J. & Beggs, J. R. (1991a) Establishment of the wasp parasitoid, Sphecophaga vesparum (Hymenoptera: Ichneumonidae), in New Zealand. New Zealand Journal of Zoology 18: 199–208. Moller, H., Tilley, J. A. V., Thomas, B. W. & Gaze, P. D. (1991b) Effect of introduced social wasps on the standing crop of honeydew in New Zealand beech forests. New Zealand Journal of Zoology 18: 171–179. Morales, C. F., Hill, M. G. & Walker, A. K. (1988) Life history of the sooty beech scale (Ultracoelostoma assimile)(Maskell), (Hemiptera: Margarodidae) in New Zealand Nothofagus forests. New Zealand Entomologist 11: 24–37. Newsome, P. F. J. (1987) The Vegetative Cover of New Zealand. Water and Soil Miscellaneous Publication 112. Ministry of Works & Development, Wellington. Paine, R. T. (1974) Intertidal community structure. Experimental studies on the relationship between a dominant predator and its principal predator. Oecologia 15: 93–120.
Impact and Control of Introduced Vespula Wasps in New Zealand 409
Rose, E. A. F., Harris, R. J. & Glare, T. R. (1999) Possible pathogens of social wasps (Hymenoptera: Vespidae) and their potential as biological control agents. New Zealand Journal of Zoology 26: 89-96. Spurr, E. B. (1991) Reduction of wasp (Hymenoptera: Vespidae) populations by poison-baiting; experimental use of sodium monofluoroacetate (1080) in canned sardine. New Zealand Journal of Zoology 18: 215–222. Spurr, E. B. (1993) The effectiveness of sulfluramid in sardine bait for control of wasps (Hymenoptera: Vespidae). Proceedings 46th New Zealand Plant Protection Conference: 307–312. Thomas, C. D., Moller, H., Plunkett, G. M. & Harris, R. J. (1990) The prevalence of introduced Vespula vulgaris wasps in a New Zealand beech forest community. New Zealand Journal of Ecology 13: 63–72. Toft, R. J., Malham, J. P. & Beggs, J. R. (1999) Mortality and emergence pattern of over-wintering cocoons of the wasp parasitoid Sphecophaga vesparum vesparum (Hymenoptera: Ichneumonidae) in New Zealand. Environmental Entomology 28: 9–13. Toft, R. J. & Rees, J. S. (1998) Reducing predation of orb-web spiders (Araneidae) by controlling common wasps (Vespula vulgaris) in a New Zealand beech forest. Ecological Entomology 23: 90–95. Valentine, E. W., & Walker, A. K. (1991) Annotated catalogue of New Zealand Hymenoptera. Unpublished DSIR Plant Protection Report No. 4. Wilson, P. R., Karl, B. J., Toft, R. J, Beggs, J. R. & Taylor, R. H. (1998) The role of introduced predators and competitors in the decline of kaka (Nestor meridionalis) populations in New Zealand. Biological Conservation 83: 175–185. Wojcik, D. P. (1994) Impact of the red imported fire ant on native ant species in Florida. pp 269–281. In Williams, D. F. (Ed.), Exotic Ants: Biology, Impact, and Control of Introduced Species. Westview Press, Boulder.
Taxonomic Relationships of Parasitoids: Poor Indicators for Their Suitability or Effectiveness as Biological Control Agents D. P. A. Sands CSIRO Entomology, Private Bag No. 3, Indooroopilly Qld, 4068 Australia (email:
[email protected])
Introduction Hymenoptera are important natural enemies of arthropods and, as predators or parasitoids, influence the abundance of their prey and hosts. Hymenoptera (12%) is the order of agents most effective for classical biological control of arthropod pests although others, especially Coleoptera (9%), Diptera (8%) are also well represented (Greathead & Greathead 1992). The Braconidae, Aphelinidae, Encyrtidae, Ichneumonidae and Eulophidae are important families for controlling species of Hemiptera, Coleoptera, Diptera, Lepidoptera and Hymenoptera. Not all biological control agents are effective agents and a proportion fail to control target organisms after they have become established. Julien (1989) indicated for biological control of weeds, that whereas the rate of establishment of agents released to 1980 was 63%, the rate of effectiveness was only 24%. The proportion of agents effectively contributing to control of arthropods has been estimated to be less than 10% (Gurr & Wratten 1999). The rate of effectiveness for biological control of target insect orders favours Hemiptera (Homoptera 66%), Lepidoptera (18%), Coleoptera (7%) and Diptera (5%) over the other orders (4%) (BIOCAT in Way 1990). The scale insects, white flies, aphids, plant hoppers and mealybugs rank highest but other successes include leaf mining beetles and moths, weevils, fruit flies, and Sirex wood wasp (Waterhouse & Sands unpublished). Groups least effectively controlled include insect borers, terrestrial and soil insects (BIOCAT in Way 1990). In addition to achieving effective control of pests, an important consideration before releasing exotic natural enemies is to ensure that they will not have any undesirable impact on non-target organisms once they have become established (Sands 1997). Concerns that some agents have had considerable impacts on non-target species have been raised (Howarth 1991; Simberloff 1992), but some of the examples have been disputed due to lack of convincing evidence (Funasaki et al. 1988). The need to balance any risks from introducing exotic agents with the benefits from controlling pests, was discussed by Cullen (1997). In this paper some of the constraints and applications of taxonomy, to selecting and testing hymenopterous biological control agents are discussed.
Identifying Effective Exotic Agents When selecting biological control agents to introduce from one country to another, their history of performance overseas provides an indication for their potential effectiveness. However, if no agents are already known, exploration is most likely to be successful in the country of origin for
410
Hymenoptera: Evolution, Biodiversity and Biological Control
Taxonomic Relationships of Parasitoids 411
the target species (Waterhouse 1991). Early in the exploratory stage it is useful to examine taxonomic groups of agents that have proven to be the most effective against the particular orders or suborders of target organisms. When selecting hymenopterous parasitoids certain families often predominate, for example Encyrtidae, Aphelinidae, Eulophidae and Pteromalidae are well represented for Sternorrhyncha; Encyrtidae and Scelionidae for Heteroptera and Trichogrammatidae, Braconidae, Ichneumonidae and Chalcididae for Lepidoptera. For hymenopterous agents, the origin and distribution of a target pest, the taxonomic relationships if known, and environmental adaptation may be important. Parasitoids that are taxonomically closely related, tend to develop in taxonomically-related hosts at similar stages of development. However, taxonomically related parasitoids can only be used as a guide for selecting agents, and other taxa should not be excluded when planning biological control programs. Agents sometimes prove to be effective if parasitising a close relative of the target pest from another region. Potential agents are cultured for at least one generation before release, to ensure they are compatible with their target taxon, and free of diseases or hyperparasitoids (Waterhouse 1991). Whereas the potential impacts of introducing diseases such as microsporidia are not always addressed, action to avoid or unwittingly introduce hyperparasitoids is considered to be of major importance. Microsporidia can be removed by selecting uninfected pairs for laboratory cultures, or specific antibiotics may be used to ensure that infections do not develop and to prevent the release of infected agents (Sheetz 1997). Primary or hyperparasitic parasitoids may sometimes be morphologically similar. Hyperparasitoids often contaminate cultures of primary parasitoids and are easily overlooked, sometimes mistaken for primary species and have on occasions, been accidentally released (Essig 1931). The primary parasitoid Aprostocetus ceroplastae (Girault), for example, is an important primary agent and is sometimes host for the hyperparasitoid, Baryscapus ceroplastophilus (Domenichini) (Noyes 1998). Both are macroscopically similar in appearance and are associated with coccids, and the latter species has been overlooked in cultures of the former. Similarly, when cultures of Encarsia spp. were introduced into the Pacific for biological control of Pseudaulacaspis pentagona (Targioni-Tozzettii) (Liebregts et al. 1989), emerging parasitoids included hyperparasitoids. If they had not been microscopically examined, hyperparasitoids would have been released with the primary species and, if they became established, may have had detrimental impacts on the subsequently effective primary parasitoids. Often microhymenoptera lack sufficient morphological characters to allow separation of closelyrelated species. Resulting lack of systematic data can lead to some potential agents appearing to be polyphagous, whereas they actually represent a cluster of sibling taxa in which individual species are more narrowly specific. For example, biotypes or ‘host races’ of Diversinervis elegans Silvestri (Encyrtidae) are adapted to different species of scale insects and each contributes to their control. Parasitoid biotypes with adaptation to particular climates may also influence their effectiveness, for example Anicetus communis (Annecke) and A. nyasicus (Compere) (Encyrtidae) were adapted to different climatic regions when introduced into Australia for biological control of white wax scale, Ceroplastes destructor Newstead (Sands et al. 1986). Cultures of both parasitoids originated from the same region of South Africa where there was no evidence of their climatic preferences. Each species on its own would probably have been ineffective after their establishment over a major geographical area occupied by C. destructor in eastern Australia. Specificity is very difficult to evaluate if they oviposit on and develop in a range of hosts in the confines of a laboratory, different to the field. For example, host biotypes of the pteromalid eggpredator, Scutellista caerulea (Fonscolombe) developed on three Ceroplastes spp. in the laboratory
D. P. A. Sands 412
Hymenoptera: Evolution, Biodiversity and Biological Control
but in the field only one biotype utilised C. destructor as a host (Sands unpublished). Host utilisation by different biotypes may be easily overlooked in the native range and attributed to polyphagy by one biotype. Although potentially important, there are no ways to detect host or climate-adapted biotypes other than by extensive field and laboratory evaluation. Percentage parasitism may be interpreted as an indication of the effectiveness of a parasitoid but on its own is likely to lead to bias (van Driesche 1983). Moreover, percentage parasitism in the native range of a parasitoid and its host is not a reliable indication for effectiveness of a parasitoid, after it has been introduced into another country. The effectiveness of parasitoids may also be compounded by the different plant substrates on which the target hosts feed. For example, Eretmocerus spp. from Spain and India performed well as a parasitoid for Bemisia tabaci (Gennadius) (Biotype B) on all crops, whereas Encarsia sp. nr pergandiella Howard performed well on melons but not on cotton and kale (Goolsby et al. 1998; Goolsby et al. this volume). Understanding tri-trophic interactions may be very important when selecting parasitoids that are effective in certain cropping systems but very difficult to recognise during the exploration phase of a biological control program.
Tests with Parasitoids and Non-target Taxa before their Release After a biological control agent has been identified, tests to ensure it is unlikely to have undesired impacts on non-target organisms are usually required before it is released. This evaluation has been a standard practice for many years for evaluating weed agents (Wapshere 1974), but only recently has this approach been applied to arthropod agents (Sands 1997; van Driesche & Hoddle 1997). Kuhlmann et al. (1998) recommended a phylogenetic approach, similar to that used for weeds, to select non-target species for testing with potential agents. He also recommended testing some species of economic or conservation significance. Most hymenopterous parasitoids are not monospecific in their native ranges, but parasitise closely-related taxa, for example in the same genus or in very closely related genera. However, the range of suitable hosts is usually much smaller in the country where a parasitoid is considered for introduction. Narrowly specific agents are given priority but the methods to accurately determine their specificity are often difficult to devise. Some potentially effective parasitoids may be automatically excluded from consideration when they are part of a taxonomic complex, when the host range of species in that complex is relatively broad. Parasitoids introduced for controlling pests are known to have utilised native species without having undesired impacts on them. For example, the scelionid Trissolcus basalis (Wollaston) is an important egg parasitoid of green vegetable bug, Nezara viridula (L.) (Pentatomidae), and has not been reported to have any impact of non-target taxa in Australia. However, Howarth (1991) has disputed the safety of several biological control introductions and has linked the introduction into Hawaii of T. basalis with the decline of native pentatomid bugs. If evaluated now for introduction into Australia, T. basalis would probably not be approved for release, due to its parasitisation of non-target native bugs. In the past, parasitism of non-target taxa by agents has been encouraged. The ‘lying in wait’ approach to establishing exotic agents, in native hosts where they are ready to transfer to attack target pests when they appear in a cropping situations, has been widely promoted (Murdoch et al. 1985). Attempts were made to establish strains of the polyphagous Aphelinus varipes
Taxonomic Relationships of Parasitoids 413
Table 1 Guidelines for selecting non-target taxa for testing with potential agents. Choose taxonomic level appropriate from: • Economically-important taxa Predators, pollinators, biocontrol agents, cultural or commercial value • Taxa of conservation significance Rare or threatened taxa • Representatives of indigenous fauna
Selection based on phylogenetic relationships to target using: • Morphological systematics Family, subfamily, tribe, genus • Molecular systematics Centrifugal relationships
Evaluation • Parasitoid behaviour may reflect host acceptance • Target taxa must support complete development of agent • Non-target taxa fail to support complete development of agent Accept: agent eligible for release • Non-target taxa support incomplete / reduced development of agent Agent requires further evaluation • Non-target taxa support complete development of agent Agent requires further evaluation / risk analysis or Reject: agent not eligible for release
(Foerster) on the wheat aphid, Rhopalosiphum padi (L.) in Australia, as a lying-in-wait strategy for control of Russian wheat aphid, Diuraphis noxia, but the parasitoids apparently failed to become established (Hughes et al. 1994). Many generalists that have the potential to have an impact on non-target taxa are undoubtedly eliminated from consideration for introduction during the process of laboratory evaluation. For example, Ooencyrtus papilionis Ashmead, O. trinidadensis Crawford and O. nezarae Ishii were not released in the USA for biological control of N. viridula, because they had a wide host range and effectiveness in their native range had not been demonstrated (Jones 1988). The host specificity of exotic parasitoids is usually determined by exposing them to selected organisms related to the target. A major impediment to testing agents for control of arthropods is lack of systematic knowledge to enable this selection, particularly the taxonomic relationships of arthropods when compared with plants (Kuhlmann et al. 1998). Without available phylogenies it may be very difficult to select for host range testing, those representatives most closely related to the target in the country being proposed for introduction. Often the degree of relatedness of targets and non-targets is not well known; for example the systematic relationships between the different genera of white flies is not known, making difficult the process of selecting non-target species for exposure to potential agents (DeBarro pers. comm.). Molecular methods for identifying relationships may be suitable when conventional phylogenies have not been developed, or when complementary information is required (Maley & Marshall 1998).
D. P. A. Sands 414
Hymenoptera: Evolution, Biodiversity and Biological Control
Other impediments include obtaining and culturing the stages of the non-target taxa most similar to those of the target host utilised, for exposure to parasitoids being evaluated. Moreover, methods for testing parasitoid behaviour for interpreting their host interactions have rarely been developed. Behaviour of parasitoids often reflects host recognition and may be an indication of host acceptance. Palpation and probing by parasitoids, particularly on host eggs or scale insects may indicate pre-ovipositional behaviour. Pre-ovipositional behaviour on larvae may be important, for example display by Cotesia erionotae (Wilkinson) (Braconidae) when offered larvae of the target species, Erionota thrax (L.) (Hesperiinae), was quite different when offered larvae of a related, non-target species, Cephrenes augiades (C. Felder). The display followed by normal parasitoid development in the banana skipper, but no parasitoid development in the non-target species, was interpreted as an indication of specificity (Sands et al. 1993). Tests conducted in a laboratory to determine the host range of parasitoids, may not accurately reveal their host recognition when disruption of behaviour occurs. For example, the egg parasitoid Ooencyrtus erionotae Ferriere, oviposited in non-target species in the laboratory but was apparently specific to its target host, E. thrax, when the same non-target taxa were present in the field (Sands 1997). Results from choice tests in which target organisms are caged with non-target taxa need careful design and interpretation of results. Kairomones from a target in close proximity may disrupt host recognition and induce oviposition by a parasitoid on non-target taxa, and are also prone to absorption onto cage materials. No-choice tests and renewed cage materials avoid these anomalies (Sands 1997).
Discussion Although the Hymenoptera are likely to continue to be most effective classical biological control agents for arthropod pests, the selection of those effective and tests to ensure that they are safe, are likely to become more complex, resulting in delays and some rejections. Effective and hostspecific agents cannot be selected on the basis of their taxonomic affinities with other agents although groups, particularly certain genera, are likely to be the best candidates. Biotypes of morphologically similar parasitoids, especially taxa previously overlooked, are more likely to be of value in the future than previously recognised. Selection of non-target taxa for testing with potential agents requires improved guidelines (see Table 1). The number of non-target taxa included in tests on a parasitoid should not be regarded as a priority, rather their relationships with the target taxon. The utilisation of non-target taxa, a characteristic sometimes considered to be beneficial as ‘lying in wait’ for exotic agents, should not be readily accepted without risk assessment. Conversely, agents that utilise native non-target taxa in the laboratory should not automatically be excluded from further assessment or risk analysis. Testing the safety of candidate species prior to their being released, will require refinement as concerns for threats to non-target utilisation gain increasing attention. When non-target taxa are selected on the basis of their relationships with the targets, especially species of conservation significance, it may be difficult or impossible to rear them when their life-histories are unknown. The logistics of testing agents with appropriate non-target taxa is certain to be a major constraint for the future assessment of parasitic Hymenoptera.
References Cullen, J. M. (1997) Biological control and impacts on non-target species. pp 195-201. In Proceedings of the 50th New Zealand Plant Protection Conference.
Taxonomic Relationships of Parasitoids 415
Essig, E. O. (1931). A History of Entomology. Macmillan, New York. Funasaki, G. Y., Lai, P. Y., Nakahara, L. M., Beardsley, J. W. & Ota, A. K. (1988) A review of biological control introductions in Hawaii: 1890-1985. Proceedings of the Hawaiian Entomological Society 28: 105-160. Goolsby, J. A., Ciomperlik, M. A., Legaspi, B. C., Legaspi, J. C. & Wendel, L. E. (1998) Laboratory and field evaluation of exotic parasitoids of Bemisia tabaci (Gennadius) (Biotype “B”) (Homoptera: Aleyrodidae) in the lower Rio Grande valley of Texas. Biological Control 12: 127-135. Greathead, D. J. & Greathead, A. H. (1992) Biological control of insect pests by insect parasitoids and predators: the BIOCAT database. Biocontrol News and Information 13: 61N-68N. Gurr, G. M. & Wratten, S. D. (1999) “Integrated biological control”: a proposal for enhancing success in biological control. International Journal of Pest Management 45: 81-84. Howarth, F. G. (1991) Environmental Impacts of classical biological control. Annual Review of Entomology 36: 485-509. Hughes, R. D., Hughes, M. A., Aeschlimann, P. P., Woolcock, L. T. & Carver, M. (1994) An attempt to anticipate biological control of Diuraphis noxia (Hom., Aphididae) Entomophaga 32: 211-223. Jones, W. A. (1988) World review of the parasitoids of the southern green stink bug, Nezara viridula (L.) (Heteroptera: Pentatomidae). Annals of the Entomological Society of America 81: 262-273. Julien, M. H. (1989) Biological control of weeds worldwide: trends, rates of success and the future. Biocontrol News & Information 10: 299-306. Kuhlmann, U., Mason, P. G. & Greathead, D. J. (1998) Assessment of potential risks for introducing European Peristenus species as biological control agents of native Lygus species in north America: a cooperative approach. Biocontrol News & Information 19: 83N-90N. Liebregts, W. M. M., Sands, D. P. A. & Bourne, A. S. (1989) Population studies and biological control of Pseudaulacaspis pentagona (Targioni-Tozzetti) (Hemiptera: Diaspidae), on passion fruit in Western Samoa. Bulletin of Entomological Research 79: 163-171. Maley, L. E. & Marshall, C. R. (1998) The coming of age of molecular systematics. Science 279: 505-506. Murdoch,W. W., Chesson, J. & Chesson, P. L. (1985) Biological control in theory and practice. American Naturalist 125: 344-509. Noyes, J. S. (1998) Catalogue of the Chalcidoidea of the world. Biodiversity catologue database and image library, CD-ROM Series. The Natural History Museum, London. Sands, D. P. A. (1997) The ‘safety’ of biological control agents: assessing their impact on beneficial and other non-target hosts. Memoirs of the Museum of Victoria 56: 611-615. Sands, D. P. A., Bakker, P. & Dori, F. M. (1993) Cotesia erionotae (Wilkinson) (Hymenoptera: Braconidae), for biological control of banana skipper, Erionota thrax (L.) (Lepidoptera: Hesperiidae) in Papua New Guinea. Micronesica, Supplement 4: 99-105. Sheetz, R., Goolsby, J. & Poprawski, T. (1997) Antibiotic treatment of a Nosema sp. (Protozoa: Microsporida) infecting the ovaries of a parasitic Encarsia wasp (Hymenoptera: Aphelinidae). Subtropical Plant Science 49: 50-52. Simberloff, D. & Stiling, P. (1996) How risky is biological control? Ecology 77: 1965-1974. van Driesche, R. G. (1983). Meaning of “percent parasitism” in studies of insect parasitoids. Environmental Entomology 12: 1611-1622. van Driesche, R. G. & Hoddle, M. (1997) Should arthropod parasitoids and predator be subject to host range testing when used as biological control agents? Agriculture & Human Values 14: 211-226.
D. P. A. Sands 416
Hymenoptera: Evolution, Biodiversity and Biological Control
Wapshere, A. J. (1974) A strategy for evaluating the safety of organisms for biological control. Annals of Applied Biology 77: 201-211. Way, M. J. (1990) Biological control introductions. A review and justification of the need for a code of conduct with a recommendation for consideration at a FAC expert consultation. 31 pp, unpublished report IIBC, Silwood Park, U.K. Waterhouse, D. F. (1991) Guidelines for biological control projects in the Pacific. Information document No. 57, South Pacific Commission, Noumea.
Natural Population of Aenasius Advena Compere (Chalcidoidea: Encyrtidae) and its Host Preference in Bangladesh B. A. Bhuiya1, S. H. Chowdhury2 and S. M. H. Kabir3 1,2
Department of Zoology, University of Chittagong, Chittagong 4331 Bangladesh (1email:
[email protected])
3
Department of Zoology, University of Dhaka, Dhaka 1000 Bangladesh
Introduction Ferrisia virgata (Cockerell) and Planococcus pacificus Cox (Pseudococcidae) are two important mealybug pests of guava (Psidium guayava L.) in different parts of Bangladesh and elsewhere on the Indian subcontinent. Aenasius advena Compere is a cosmopolitan encyrtid parasitoid which is known to attack F. virgata in India (Rawat & Modi 1968a), Bangladesh (Kerrich 1967; Bhuiya et al. 1997), South Africa (Prinsloo 1983), Hawaii (Compere 1937), and Central and South America (Noyes & Hayat 1994; Noyes 1995). In Bangladesh A. advena was also found to parasitise P. pacificus at the same locality. Given that natural populations of parasitoids are dependent on the population size of their hosts (Doutt 1964), the seasonal fluctuation in rates of parasitism by A. advena was studied for these two mealybug hosts to determine its efficiency as a potential biological control agent for guava in Bangladesh (Chowdhury & Ullah 1984).
Materials and Methods Natural populations of A. advena; F. virgata and P. pacificus were monitored in a large guava garden at Kanchan Nagar, near Patiya, about 45 km from Chittagong city. Populations were assessed at fortnightly intervals for a period of three years (January 1983–January 1986) to determine the rate of parasitism by A. advena on the two hosts. . The guava garden was divided into 10 sections with 10 trees in each section. At least 10–15 leaves infested with host insects were selected randomly from each tree and were brought to the laboratory in polythene bags. Mealybugs in each bag were counted and the leaves containing different host species were kept in separate emergence boxes. They were observed twice daily and parasitoid emergence from each individual was recorded.
Results Over the three years of the study, fluctuations in mealybug numbers varied between species and years. Ferrisia virgata was present for the whole year or virtually so, but numbers were generally lowest during the hot part of the year (June–August) and highest during the cooler months (September–March) (Figs 1–3). However, there was substantial difference in population size for the same month, from year to year. For instance, in the first year numbers were reasonably constant between months, except for a significant drop in July–August (Fig. 1), while in the second year F. virgata declined steadily from January to August, numbers peaked in September
417
B. A. Bhuiya, S. H. Chowdhury and S. M. H. Kabir 418
Hymenoptera: Evolution, Biodiversity and Biological Control
1
80
Host number & % parasitism
70 60
NHFV NHPP
50
%PFV %PPP
40 30 20 10 0 Jan
Feb Mar
Apr
May Jun Jul 1st year
Aug Sep
Oct
Nov Dec
2 100
Host number & % parasitism
90 80 70
NHFV 60
NHPP
50
%PFV %PPP
40 30 20 10 0 Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
2nd year
Host number & % parasitism
3
70
60
50
NHFV NHPP
40
%PFV %PPP
30
20
10
0 Jan
Feb
Mar
Apr
May
Jun Jul 3rd year
Aug
Sep
Oct
Nov
Dec
Figures 1-3 Monthly populations of the mealybugs Ferrisia virgata (NHFV) and Planococcus pacificus (NHPP) and level of parasitism by Aenasius advena (on F. virgata = %PFV) and on P. pacificus = %PPP): 1) in the 1st year (1983); 2) in the 2nd year (1984); and 3) in the 3rd year (1985) of the study.
Natural Population of Aenasius Advena Compere (Chalcidoidea: Encyrtidae) 419
and declined thereafter (Fig. 2). In contrast, in the third year numbers were high for January– March and September–December (Fig. 3). Populations trends for P. pacificus over the three years differed substantially between years and from those of F. virgata. In the first year there were significant peaks in numbers in January and September (Fig. 1); in year two there was a peak in March followed by a steep decline to May, with numbers remaining low for the next seven months (Fig. 2); while in year three the population was high from February to August (with peaks in April and August), and then a decline to lower numbers in October–December (Fig. 3). Adults of the parasitoid A. advena were present in all months of the year. In the first and second years they parasitised F. virgata from January to June and from September onwards, while in July and August they switched to parasitising P. pacificus. At this time populations of F. virgata had dropped to zero or almost so (Figs 1, 2). However, in the third year F. virgata was still present in the field during July and August (Fig. 3), and consequently were parasitised by A. advena albeit at a low level. In all three years this parasitoid showed a bimodal pattern of parasitism with peaks in March and October, and low levels of parasitism in July–August and December–January. Although P. pacificus was present throughout the year, often in high numbers (see above), it was not utilised as a host by A. advena in the first half of the year, from December to May–June. Parasitism of P. pacificus began in June and continued during July–September, whereafter it declined. However, during the second six months of the year the parasitoid was utilising both hosts.
Discussion The results of the present study show rates of parasitism by A. advena are consistently higher on F. virgata than on P. pacificus, except during mid summer (July–August) when F. virgata is absent or at very low numbers. During this period P. pacificus acts as an alternate host, but at other times of the year rates of parasitism are usually very low on P. pacificus, even though this mealybug can be very common in the field. Overall rates of parasitism are highest in March and October and this is generally correlated with a greater abundance of hosts at the same time or slightly earlier. Further, the decline in rates of parasitism on F. virgata leading up to mid summer and mid winter track fairly closely the decline in host numbers. However, the situation for P. pacificus is quite different in that there appears to be little or no correlation between population size of this host and rates of parasitism. Of particular interest is that A. advena does not parasitise P. pacificus during late winter and spring but does so, albeit at low levels, during autumn, even though host numbers are often similar at both times of the year. The results of Rawat and Modi (1968b) differ markedly from the present study. They recorded the highest levels of parasitism by A. advena on F. virgata between September and December. However, they did not present any results for the remaining part of the year, either because the period of the study was restricted or because the host was absent or at very low numbers. The above results strongly indicate that the performance of A. advena differs significantly on the two hosts present on guava. Planococcus pacificus might be a less suitable host for a variety of reasons which leads the parasitoid to lay fewer eggs in it. For instance, host size is known to influence ovipositional rate of Nasonia vitripennis (Walker) (Pteromalidae) in that it lays fewer eggs in small versus large pupae of Musca domestica L. (Wylie 1958). Alternatively, ovipositional rates may be similar between the two hosts, but developmental success of the parasitoid differs between them. Competition with other parasitoids may result in different rates of parasitism on the two hosts. However, this does not appear to be the case here given that other parasitoid species (Blepyrus
B. A. Bhuiya, S. H. Chowdhury and S. M. H. Kabir 420
Hymenoptera: Evolution, Biodiversity and Biological Control
insularis Cameron and Anagyrus mangicola Noyes – both encyrtids) were only found to utilise F. virgata and only at low levels. At present, the prospect for biological control of F. virgata and P. pacificus on guava is unclear. Although rates of parasitism on F. virgata can reach 50% at times, they are generally much lower than this. However, if a better understanding of the ovipositional behaviour of A. advena and the effect of extrinsic environmental factors on its development can be achieved in the future, then it may be possible to manipulate this parasitoid so that its performance as a control agent is increased.
Acknowledgements The authors wish to thank Dr B. R. Subba Rao for confirming the identification of the parasitoid species. Thanks are also due to Dr D. J. Williams for identification of the host species, and to Dr Mike Fitton for his valuable comments on the final manuscript.
References Bhuiya, B. A., Chowdhury, S. H. & Kabir, S. M. H. (1997) An annotated list of chalcidoid parasitoids (Hymenoptera) of Coccoidea (Homoptera) on guava in Bangladesh. Bangladesh Journal of Zoology 25: 53-63. Chowdhury, S. H. & Ullah, G. M. R. (1984) Coccoids and their host plants in Bangladesh – A checklist (Homoptera: Coccidae) Bangladesh Journal of Zoology 12: 39-45. Compare, H. (1937) The species of Aenasius, encyrtid parasites of mealybugs. Proceedings of the Hawaiin Entomological Society 9: 383-404. Doutt, R. L. (1964) Biological characteristics of entomophagous adults. pp. 145-167. In DeBach, P. (Ed.), Biological Control of Insect Pests and Weeds. Chapman & Hall, London. Kerrich, G. J. (1967) On the classification of the anagyrine Encyrtidae, with a revision of some of the genera (Hymenoptera: Chalcidoidea) Bulletin of the British Museum (Natural History), Entomology 20: 143-250. Noyes, J. S. (1995) Encyrtidae of Costa Rica (Hymenoptera: Chalcidoidea): The genus Aenasius Walker, parasitoids of mealybugs (Homoptera: Pseudococcidae). Bulletin of the Natural History Museum, Entomology 64: 117-163. Noyes, J. S. & Hayat, M. (1994) Oriental Mealybug Parasitoids of the Anagyrini (Hymenoptera: Encyrtidae). CAB International, Wallingford. Prinsloo, G. L. (1983) A parasitoid-host index of Afrotropical encyrtidae (Hymenoptera: Chalcidoidea). Entomology Memoirs of the Department of Agriculture Republic of South Africa 60: 1-35. Rawat, R. R. & Modi, B. N. (1968a) First record of Aenasius advena Compere (Encyrtidae: Hymenoptera) from India as a parasite of Ferrisia virgata Ckll. Indian Journal of Entomology 30: 85. Rawat, R. R. & Modi, B. N. (1968b) A record of natural enemies of Ferrisia virgata Ckll. in Madhya Pradesh (India). Mysore Journal of Agricultural Sciences 2: 51-53. Wylie, H. G. (1958) Factors that affect host finding by Nasonia vitripennis (Walk.) (Hymenoptera: Pteromalidae). Canadian Entomologist 90: 597-608.
Studies on Eretmocerus sp. (Hymenoptera: Aphelinidae) – a Promising Natural Enemy of the Castor Whitefly Trialeurodes ricini (Hemiptera: Aleyrodidae) Seeba Balan and R. W. Alexander Jesudasan Department of Zoology, Madras Christian College (Autonomous), Tambaram, Chennai-600 059 Tamil Nadu, India (email:
[email protected])
Introduction The castor whitefly, Trialeurodes ricini (Misra) infests a wide range of plants, viz. Gossypium hirsutum, Murraya koenigii, Annona glabra, Ipomoea batata, Achras zapota, Arbutus sp., Rosa spp. and Melanthesa rhamnoides. However, the primary plant hosts of T. ricini are Ricinus communis and Phyllanthus acidus, both of which are economically important. Ricinus communis is a notable oil-seed crop in India, most of which is exported, while the fruits of P. acidus are edible and also have possible medicinal properties. Various insect pests are associated with these two plants of which T. ricini, inflicts substantial damage. Although investigations on the taxonomy, host plant correlated variation and control of the castor whitefly have been undertaken, there is limited knowledge in regard to its natural enemies. The aim of this study was to document aspects of the biology of its major parasitoid, Eretmocerus sp.
Material and Methods Parasitoids that emerged from nymphs of T. ricini were collected using an aspirator and identified as Eretmocerus sp. (Aphelinidae). Studies on the life cycle of this species were carried out from June 1994 to May 1995. Three to four pairs of adults representing both sexes of Eretmocerus sp. were released onto unparasitised nymphs of T. ricini secured using clip-on cages. This set-up was left undisturbed for 2–3 days to ensure complete parasitism. Subsequently, parasitised nymphs were removed from leaves, mounted on glass slides, dissected under a stereomicroscope (Nikon Binocular, SMZ 1D), and the developmental stages of the parasitoid recorded. Studies were conducted on the ovipositional behaviour of Eretmocerus sp. between November 1994 and May 1995 by releasing wasps onto leaves infested with T. ricini and continuously monitored them using a hand-lens. Observations on ovipositional efficiency of the parasitoid on nymphs of T. ricini were made in the morning and evening for one hour on every alternate day.
Results and Discussion Life cycle Eggs of Eretmocerus sp. are laid either inside the host body through the vasiform orifice or underneath the body. They are elongate measuring 0.13–0.15 mm in length and 0.05–0.07 mm in
421
Seeba Balan and R. W. Alexander Jesudasan 422
Table 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Morphometrics of the life history stages of Eretmocerus sp.
Stage
Length (mm)
Width (mm)
Egg
0.13 – 0.15
0.05 – 0.07
I instar
0.30 – 0.35
0.11 – 0.14
II instar
0.51 – 0.58
0.23 – 0.34
III instar
0.52 – 0.55
0.26 – 0.32
IV instar
0.53 – 0.59
0.28 – 0.31
Adult female
0.61 – 0.87
-
Adult male
0.59 – 0.82
-
width (Table 1). The development time of the egg stage was virtually the same throughout the year on the two host plants, R. communis and P. acidus (Table 2). The first instar larva is pear-shaped and measures 0.30–0.35 mm in length and 0.11–0.14 mm in width. Its development time was the same on the two host plants but varied between months from 1.5 to 2.5 days. The second instar is spherical and globular in shape, measuring 0.51–0.58 mm in length and 0.23–0.34 mm in width. Development time showed a greater range between months on P. acidus (2.5–4.5 days) compared with R. communis which was constant (Table 2). The third instar larva is globular and measures 0.52–0.55 mm in length and 0.26–0.32 mm in width (Table 1), and occupied three-quarters of the host’s body. Again, development time showed a greater range between months on P. acidus (3.5–5.5 days) compared with R. communis (4.5–5.6 days) (Table 2). The fourth instar is dark brown in colour and measures 0.53–0.59 mm in length and 0.28–0.31 mm in width (Table 1). The range in development time between months for this stage was longer on P. acidus (6.5–11.5 days) compared with R. communis (5.5–8.5 days) (Table 2), but data for the first host plant is partly skewed due to the longer development time for two months (August and September 1994; 11.5 and 10.5 days, respectively). The adult is dark-yellow in colour. The female measures 0.61–0.87 mm and the male 0.59–0.82 mm in length (Table 1). Adult longevity varied from 5.5 to 7.5 days when larvae fed on P. acidus and 4.5 to 11.5 days when on R. communis, while the total life cycle varied from 22 to 29 days and 23 to 28 days on these two plants, respectively (Table 2). Overall, the life cycle of Eretmocerus sp. was shortest from March to June and longest during the coolest months (September–December) on both host plants. Eretmocerus sp., like other members of the genus, is a solitary, internal parasitoid (Clausen 1940). After oviposition through the vasiform orifice of the whitefly nymph, the first instar larva penetrates the host’s body from underneath. Both the second and third instars have recessed mouthparts. No meconium is cast at the end of larval development. The pupa is orientated anterodorsally of the empty host skin and emerges through an oval or elliptical emergence hole. The emergence hole of T. ricini is different from that of the parasitoid: the former emerges through an inverted T-shaped opening, while the latter emerges through a circular opening. Generally, these observations are in agreement with the findings of Foltyn & Gerling (1985), and the information derived during this study is consistent with that for other species in the genus, e.g. Eretmocerus corni Haldeman on T. packardi (Morrison) (Kunezel 1977) and Eretmocerus sp. infesting T. vaporariorum (Westwood) (Vet & Lenteren 1981; Kajitha 1981; Taylor 1981).
Studies on Eretmocerus sp. (Hymenoptera: Aphelinidae) 423
Table 2 Development times for life history stages of Eretmocerus sp. parasitising Trialeurodes ricini on two host plants, Phyllanthus acidus and Ricinus communis for the period June 1994 to May 1995 (mean days ± S. D.) Phyllanthus acidus
Ricinus communis
Stage
Lowest month
Highest month
Lowest month
Highest month
Egg
1.2 ± 0.7
1.5 ± 0.7
1.5 ± 0.7
1.5 ± 0.7
I instar
1.5 ± 0.7
2.5 ±0.7
1.5 ± 0.7
2.5 ± 0.7
II instar
2.5 ± 0.7
4.5 ± 2.1
3.5 ± 0.7
3.5 ± 0.7
III instar
3.5 ± 0.7
5.5 ± 0.7
4.5 ±0.7
5.6 ± 0.7
IV instar
6.5 ± 2.1
11.5 ± 0.7
5.5 ± 0.7
8.5 ± 0.7
Adult
5.5 ± 0.7
7.5 ±0.7
4.5 ± 0.7
7.5 ± 0.7
Total
22.5 ± 4.9
29.0 ± 4.2
23.0 ± 5.6
28.0 ± 5.6
Parasitism rates on different host plants During July 1994, 34.8% and 41.6% parasitism of T. ricini were recorded on the host plants P. acidus and R. communis, respectively. However, in January 1995 maximum parasitism was 73.9% and occurred on R. communis, while it reached only 18.8% on P. acidus in July of the same year. Even though the population of T. ricini was higher on P. acidus, the rate of parasitism on this host plant was lower than on R. communis. The reasons for this are unclear but are possibly affected by characteristics of the leaves of the two host plants or differential performance of T. ricini on the different plants (see below). However, the data available show no clear pattern from year to year, and further research is required to determine the factors that affect rates of parasitism of T. ricini on different host plants. Levels of parasitism in excess of 70% by Eretmocerus sp. are in accordance with rates of parasitism recorded for other aphelinids, such as Encarsia formosa Gahan on T. vaporariorum (Jordaan-Halspas et al. 1987), and indicate that this parasitoid has real potential as a biocontrol agent of T. ricini.
Ovipositional behaviour Female parasitoids appear to search randomly for hosts, with leaves having lower host densities being preferred compared with high densities. Although all host stages were present on leaves, the parasitoid selected late second instar T. ricini in preference to other stages. Antennating and ovipositor probing are the probable mechanisms by which Eretmocerus sp. determines host suitability, including avoiding already parasitised hosts. Only one egg is laid per host, the process of oviposition taking 5–10 min, which is then repeated when the next suitable host is encountered. Approximately 10–15 eggs are oviposited per day, with oviposition occurring mostly at dawn and dusk. Eretmocerus sp. does not have a rigid ovipositor and presumably this is why eggs are deposited either underneath the host’s body or through the vasiform orifice (van Lenteren et al. 1980). Host searching success is apparently influenced by the morphology of the host plant leaves, availability of hosts, and host defence mechanisms. Plants with more hairs or trichomes seem to hinder searching and probing behaviour by the parasitoid, as reported by Li et al. (1987) for En. formosa which has its host searching ability impeded on leaves that have a large number of hairs. Copious secretion of honey and wax produced by aleyrodid nymphs can also deter oviposition by parasitoids as it hinders movement, antennation and oviposition. These secretions are
Seeba Balan and R. W. Alexander Jesudasan 424
Hymenoptera: Evolution, Biodiversity and Biological Control
maximal in the late third and fourth instar nymphs of T. ricini and it is these stages that are least preferred for oviposition by Eretmocerus sp. Similar findings have been reported by Shimron et al. (1992) who showed that excess honey dew decreased walking speed of En. formosa. Although the above observations provide some insight into the interactions between Eretmocerus sp. and T. ricini, further research will be required to elucidate a more complete understanding of the factors that influence the ovipositional success of this parasitoid.
Acknowledgements We wish to thank Dr. M. Hayat, Aligarh Muslim University, Uttar Pradesh, India for identifying the parasitoid.
References Clausen, C. P. (1940) Entomophagous Insects. Mc Graw Hill , New York. Foltyn, S. & Gerling, D. (1985) The parasitoid of a aleyrodid Bemisia tabaci in Israel. Development, host preference and discrimination of the aphelinid, Eretmocerus mundus. Entomologia Experimentalis et Applicata 38: 255-260. Jordaan-Hulspas, P. M., Christochowitz, E. E., Woets, J. & van Lenteren, J. C. (1987) The Parasitehost relationship between Encarsia formosa Gahan (Hymenoptera : Aphelinidae) and Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae) XXIV. Effectiveness of Encarsia formosa in the green house at low temperature. Journal of Applied Entomology 103: 368-378. Kajitha, H. (1981) Native parasites of the greenhouse whitefly Trialeurodes vaporariorum (Westwood) in Japan and results of first use as biocontrol agents. Zeitschrift fur Angewandte Entomologia 92: 457-464. Kuenzel, N. T. (1977) Population dynamics of a protelean parasite (Hymenoptera: Aphelinidae) attacking a natural population of Trialeurodes packardi (Homoptera: Aleyrodidae) and new host record for two species. Proceedings of the Entomological Society of Washington 79: 400-404. Li, Z. H., Lammes, F., van Lenteren, J. C., Huisman, P. W. T., van Vianen, A. & De Pointi, O. M. B. (1987) The Parasite-host relationship between Encarsia formosa Gahan (Hymenoptera: Aphelinidae) and Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae) XXV. Influence of leaf structure on the searching activity of Encarsia formosa. Journal of Applied Entomology 104: 297-304. Shimron, O., Hefetz, A. & Gerling, D. (1992) Arrestment responses of Eretmocerus sp. and Encarsia deserti (Hymenoptera: Aphelinidae) to Bemisia tabaci honey dew. Journal of Insect Behaviour 5: 517-526. Taylor, D. E. (1981) Whiteflies . Zimbabwe Agricultural Journal 78: 25. van Lenteren, J. C., Nell, H. W., Sevenster, Vander & Lelie, L. A. (1980) The Parasite-host relationship between Encarsia formosa (Hymenoptera: Aphelinidae) and Trialeurodes vaporariorum (Homoptera : Aleyrodidae) IV. Ovipositional behaviour of the parasite with aspects of host selection, host discrimination and host feeding. Zeitschrift fur Angewandte Entomologie 89: 442-454. Vet, L. E. M. & van Lenteren, J. C. (1981) Parasite-host relationship between Encarsia formosa Gahan (Hymenoptera: Aphelinidae) and Trialeurodes vaporariorum (Westwood) (Homoptera: Aleyrodidae) X. A comparison of three Encarsia spp. and one Eretmocerus sp. to estimate their potentialities in controlling whitefly on tomatoes in greenhouse with a low temperature regime. Zeitschrift Fur Angewandte Entomologie 91: 327-348.
PART
8
Hymenoptera in Education
This Page Intentionally Left Blank
Developing Life Science Instructional Materials Using a Parasitic Wasp, Melittobia digitata Dahms (Hymenoptera: Eulophidae): a Case Study Robert W. Matthews Department of Entomology, University of Georgia, Athens, GA 30602 USA (email:
[email protected])
Introduction Science education in the United States and elsewhere in the world is at a crossroads. Science teachers face increasing demands for accountability, because scores on standardised science tests are unacceptably low compared to expectations. Science text materials for children all too often contain either encyclopedic compilations of fact or a whir of simplistic inaccuracies. At the same time, students in science classes often seem disinterested (Talton & Simpson 1985). Educational research shows that middle schools are especially critical for stimulating a life-long interest in science and, for some time, many have been failing badly at this job (Simpson & Oliver 1985). As working scientists, we cannot afford to ignore this situation. I believe that if this state of affairs is to improve, there is an urgent need for research scientists and university teachers to become more involved in curriculum materials development for pre-college science. Furthermore, it is in our best interests to do so. As one slogan phrases it, ‘Tomorrow’s scientists are in today’s schools’.
The Case for Insects in the Classroom Sheer numbers, diversity, adaptability, and evolutionary success ought to be reasons enough for including insects in all pre-college biology classrooms. One can easily add practical reasons ranging from fiscal attractiveness to relatively short life cycles and fewer restrictive regulations regarding their laboratory use. However, in truth, although insects offer a nearly inexhaustible well of stimulating educational material, it has barely been drawn upon (Matthews et al. 1997). Although many activities have been published (particularly for children in the younger grades), the insect species they represent are disproportionately few. Furthermore, with a few notable exceptions, most of the literature on insects produced for school educators is authored by writers, other educators and teachers, with little or no input from professional entomologists. Hymenoptera, in particular, have considerable potential as classroom organisms, yet there is little curriculum innovation that uses them. Even ants, perennially popular in fascinating generations of children and teachers with their cooperative excavations in ‘ant farms’, are subject only for a few casual observational activities, and relatively little formal curricular material. Beyond this, there are activities on honeybee biology and communication, and Trichogramma as an agent of biological control, and little else.
427
Robert W. Matthews 428
Hymenoptera: Evolution, Biodiversity and Biological Control
Might you be interested in venturing into this void? Since 1995, with funding from the U. S. National Science Foundation, I have headed a project to transform a common parasitic wasp from a laboratory insect used with a few graduate students into the ‘WOWBug’, a new instructional organism now being employed to teach a variety of life science concepts and process skills to thousands of middle school and high school biology students. With the hope of encouraging you to join me in the arena, I would like to share my experiences as a fledgling in science curriculum materials development.
What is a WOWbug? WOWBugs (Melittobia spp.) are parasitic (or parasitoid) wasps in the chalcidoid family Eulophidae. The species used in the United States is Melittobia digitata Dahms. For the past 30 years, it has been a research animal in my insect behaviour laboratories. The genus is found worldwide, and includes 14 described species (Dahms 1984). In Australia and throughout much of South and Central America, M. australica Girault is often locally abundant. Its behaviour is very similar to that of M. digitata. Melittobia lay their eggs on prepupae or pupae of other insects, mud dauber wasps being common natural hosts. Development is gregarious, with complete metamorphosis. The entire life cycle takes 17 to 21 days under normal room temperatures. Each stage, including the egg, is readily observable on the exterior of the host. Sexes are easily distinguishable. Black, winged females vastly outnumber caramel-colored males (about 25:1) in the population. In addition, males lack eyes, have short stubby wings, and enlarged antennae. Males are extremely aggressive toward other males, often killing one another after ferocious fights. This is all the more remarkable since males are totally blind. Matthews et al. (1996) and Matthews (1997) provide overviews of the biology and habits of the WOWBug. Additional information can be found on the world wide web at http://entomology.ent.uga.edu/wowbugs. The first part of the name WOWBug is an acronym which can stand for Working On Wasps or for Wild Over Wasps, but the second part may puzzle you. When we began our curriculum development effort, these organisms had no common name, and herein hangs a tale. At first we dubbed them Fast Wasps, in homage to Fast Plants, the rapid-cycling brassicas of University of Wisconsin fame (Williams 1995). It was a public relations disaster. To the average person, wasp equates to fear and stings. It quickly became clear that even science teachers were a bit squeamish at the initial thought of using wasps in the classroom. However, once they had met Melittobia first-hand, they were both receptive to and intrigued by its classification. Of course, parasitoid wasps are not true bugs in the entomological sense, but marketing concerns cannot be ignored when dealing with the public. The teachers’ initial reticence evaporated when we changed the name of our insect to WOWBugs. Since additional dictionary definitions of ‘bug’ include both ‘an insect’ and ‘a sudden enthusiasm’, we felt justified.
Desirable Attributes for a Classroom Insect In making an initial decision about an organism’s suitability for classroom use, a number of practical concerns must be met (Table 1). About the size of the Drosophila fruit fly, Melittobia are harmless and cannot sting humans. They pose no environmental risks, should a few escape. With their rapid, easily observable life cycle and fascinating behaviours, WOWBugs are highly engaging for students. The task for us as curriculum material developers has been to make them instructional and easy to use as well.
Developing Life Science Instructional Materials Using a Parasitic Wasp 429
Table 1
Ten practical questions to ask about a potential classroom insect.
Concern
Relevant Attribute of Melittobia
1. Are they abundant?
A single mud dauber cocoon can give rise to 300–900 WOWBugs, enough for several classes to do several activities.
2. Will they survive in a typical classroom?
WOWBugs thrive at typical room temperature and humidity. As adults, WOWBugs do not need to be fed or provided with water.
3. Are they inexpensive?
Cultures are available commercially, and the insects can usually be readily obtained from the nests of mud dauber wasps, their natural hosts.
4. Do they take up much space?
Even if each student is given their own culture, this amounts only to a row of vials or pill bottles tucked away in a corner.
5. Are they easily reared?
Continuous rearing requires only placing a few newly emerged adults on a new host once a month, a process that takes only a few minutes. Inexpensive suitable hosts are blow fly puparia, available from biological supply houses.
6. Do they pose any environmental risks?
Cosmopolitan in distribution, Melittobia have been so spread about the world via their hosts that their actual geographic origin is unknown. Should they escape from cultures or experiments, there is no risk of introducing an unwanted alien insect. Moreover, their dispersal ability on their own is quite limited.
7. Are they harmless and easy to handle?
Males are blind and flightless; females have wings, but seldom fly. Although they are wasps, their tiny stinger cannot sting humans. They have no odor or other objectionable features.
8. Are they reliable?
Neither sex is dependent on photoperiod or emergence rhythms for behaviour expression.
9. Are the sexes distinct?
Even as pupae, WOWBugs can be separated by sex because of the absence of compound eyes in males. Adults exhibit extreme sexual dimorphism.
10. Do they allow students Female reproductive behaviour is extremely flexible, including the to ask meaningful ability to delay oviposition, to produce either males parthogenetically or scientific questions? normal mixed-sex clutches. WOWBug males attack one another ferociously, and may be cannibalistic. WOWBug courtship is elaborate and reliably observable, and involves an unusual chemical communication system. All stages of the rapid life cycle are readily observable. Skewed sex ratios and high inbreeding suggest interesting genetic questions. And this is just the beginning.
WOWBugs are extremely low maintenance in the classroom. Although their natural preferred hosts are the young of mud dauber wasps, they readily accept substitutes such as blow fly pupae (Sarcophaga bullata). A single culture in a container the size of a matchbox provides 300+ adult WOWBugs, enough to supply several classes. Adult WOWBugs require no feeding unless an ongoing culture will be maintained. In this case, a new host must be provided so females can develop their eggs.
Development of Curricular Materials In today’s educational world, successful curriculum development is a widely cooperative process. As principal investigator and project director, I have been privileged to have had the help of a talented team of science educators who have proven invaluable in coordinating and running the various stages of what has ultimately become a comparatively large project. We began with a grant from the U. S. National Science Foundation for a four-year project with three major phases.
Robert W. Matthews 430
Hymenoptera: Evolution, Biodiversity and Biological Control
Table 2 Results of WOWBugs materials development effort funded by U.S. National Science Foundation, 1995-1998. Materials
Description
WOWBugs: New Life for Life Science
Basic 320-page manual, with over 20 classroom-tested activities and experiments for grades 5-12, plus extensive support materials for teachers. [ISBN 1-888499-06-0}
What’s Inside a Mud Dauber’s Nest?
Illustrated 32-page teacher’s guide to the most common WOWBug host, with identification keys, directions for dissecting nests, biological information about common parasites, predators and scavengers. [ISBN 1-888499-07-9].
WOWBug Biology
17-minute videotape detailing life cycle, adaptation, locomotion, competition, and reproduction, with illustrated teacher’s guide.
WOWBug Rearing and Maintenance
14-minute videotape detailing diet, housing, hosts, and tips for success, with illustrated teacher’s guide.
Organ Pipe Mud Dauber Biology
21-minute videotape of ecological interactions, including concepts of food chains and webs, parasitism, and predation, with illustrated teacher’s guide.
Organ Pipe Mud Dauber Slide Set
20 full-color 35-mm slides, numbered narrative, and Teacher’s guide with keys to nest contents. Introduces predator/prey relationships through mud dauber wasp nests.
WOWBugs Bulletin
Multi-page illustrated newsletter with new ideas from classroom activities to science fair projects, board games to book reviews, newest research, and tips from teachers successfully using WOWBugs. Ongoing publication, four times a year.
Phase I involved recruiting a core group of middle school life science teachers through presentations given at the Georgia Science Teachers Association annual meeting. This core group of 12 teachers (the ‘Development Team’) was then introduced to Melittobia. Each teacher was given a culture of the insects to take back to the classroom. They were encouraged to expose their students to these new organisms, and to come up with creative ways to use them. There was continual feedback as we shared ideas back and forth. The teachers were extremely creative and enthusiastic, although sometimes naive about basic entomological facts. In one instance a teacher asked in all seriousness if the Melittobia would get bigger as they got older. This sort of query made us realize that it would be critically important to include extensive entomological background for both teachers and students along with any activities that were generated. From these initial experiences, a series of prototype activities was developed. For Phase II, we recruited a larger group of nearly 100 teachers who attended a two day workshop on WOWBugs, and then agreed to pilot test one or more of the prototype activities in their classrooms and share their experiences with us. With their input and feedback, the initial set of prototype activities was refined. This phase culminated in the production of various teacher support materials (Table 2). The centerpiece of these is a comprehensive 320-page manual (Matthews et al. 1996) with 20 investigations that use the WOWBug, plus practical suggestions for concept elaboration and ideas for extended independent investigations. The manual also includes detailed information on WOWBug biology, management, culture and rearing tips, transparency masters, and reproducible student pages. Because classroom educators face an array of administrative pressures to justify and document all aspects of their teaching, each activity in the manual was keyed and cross-indexed to relevant science process skills and the new U. S. National Science
Developing Life Science Instructional Materials Using a Parasitic Wasp 431
Education Standards (National Research Council 1995). Finally, a professional editor was engaged to polish the text and provide a uniform format for the activities. Concurrently, ancillary supporting materials were developed, including professional quality videotapes, a set of Kodachrome slides, and teachers’ guides to these materials. An on-going quarterly newsletter, WOWBugs Bulletin, was established to serve as a mechanism for maintaining communication with teachers to share ideas and new developments. Phase III of the project was dissemination. The National Science Foundation, as a condition for funding this sort of project, mandates that a mechanism for dissemination of the developed materials be included and implemented. This was a completely new area for me. Fortunately our university research foundation, sensing a potential source of revenue, came forward to work out licensing agreements with publishers and biological supply companies. This took far longer than we ever imagined, with lawyers for the parties concerned negotiating terms acceptable to all. The outcome was that the materials needed for successfully using WOWBugs became commercially available in the United States through several channels, including directly from Riverview Press and (with living cultures and kits) from Carolina Biological Supply Company. Merely making something available in the marketplace is not sufficient in itself, of course. As part of the dissemination phase, we found it crucially important to be willing to attend professional meetings of science teachers where teachers could learn about the new materials and be introduced to WOWBugs through hands-on workshops. In addition, a variety of popular articles targeted science teachers (e.g. Matthews 1997).
The Great WOWbug Round-up: a Sample Activity to Introduce WOWbugs Innovative new activities using living animals that fire student’s imagination and interest can be a powerful teaching tool, but they must be designed to fit into existing curricula. They should illustrate important biological principles. They should provide a framework for guided inquiry, rather than ‘cookbook’ lesson, and should encourage critical thinking extending beyond a single laboratory session (Lawson 1995). And as a practical note, they must be low in cost and make efficient use of time, or many teachers will be unable to implement them. However, this need not be a daunting task. Published WOWBug activities range from the life history observations to sensory investigations without requiring any materials beyond those found in the average household. As an example, here is one of the simplest, an introductory exercise that has proven both entertaining and instructional for students from age 11 to adult. Minimal materials for this activity include a culture of living adult WOWBugs, a sheet of white paper, and a box or bag of ‘capture tools’ such as pipe cleaners, cotton swabs, small watercolor brushes, plastic vials, and clear soda straws. After the instructor dispenses three to eight WOWBugs onto each team’s sheet of paper on their desktop, students must figure out a way to get all the WOWBugs back into captivity without death or injury, using any of the ‘tools’ provided. After about 10 minutes, class members share methods and techniques tried, and discuss the advantages, disadvantages, and relative efficiency of each. ‘The Great WOWBug Roundup’ helps students appreciate that a major part of every animal’s behaviour involves orientating towards or away from factors like food, mates, predators, light, etc. Students discover basic concepts about animal orientation and develop a familiarity with a new organism as they attempt to solve a simple but important task, the job of trying to recapture
Robert W. Matthews 432
Hymenoptera: Evolution, Biodiversity and Biological Control
WOWBugs. An added element of surprise is provided as they are confronted with the discovery that, despite functional wings, Melittobia crawl and hop rather than fly, and play possum when disturbed. Very soon at least some students discover that the insects react in a directed way to light, and classroom discussion naturally moves to influences of light and gravity upon insect orientation. Like the other activities in the manual, ‘The Great WOWBug Roundup’ focuses on concepts developed by students through a simple inquiry-centered investigation. Creativity is encouraged. No one way of corralling the insects is more correct; some ways are simply more efficient. The teacher is free to introduce formal terminology such as taxes (= oriented movements in response to stimuli) as appropriate, but it arises in a natural context rather than being forced. Encouraged to explore reasons why certain oriented responses might occur, students come to formulate various hypotheses, and these in turn stimulate further experimentation.
Conclusion To address evident needs in science instruction, teachers need appropriate tools. Because insects of all sorts are inherently interesting, we in entomology in particular, are especially well positioned to abandon our traditional apathy and become more involved in science curricular development. We have no right to complain that things need to be repaired while keeping our own hands in our pockets. It is not a painful process: WOWBugs have been a real adventure for me, and the experience of working with teachers and young students has been one of the highlights of my career. It is particularly gratifying to see the metamorphosis of one’s ‘laboratory rat’ into a widely used, dare I say ‘model’, organism. As colleagues, I urge you to consider reaching beyond the traditional narrow confines of academia, and share your entomological knowledge with those upon whom our future depends. You will gain as much as you give, and will learn as much as you teach. As has been said before (Matthews et al. 1997), ‘The grins, gasps and exclamations of wide-eyed students will be ample reward for the modest effort required’.
References Dahms, E. C. (1984) A review of the biology of species in the genus Melittobia (Hymenoptera: Eulophidae) with interpretations and additions using observations on Melittobia australica. Memoirs of the Queensland Museum 21: 361-385. Lawson, A. E. (1995) Science Teaching and the Development of Thinking. Wadsworth, Belmont. Matthews, R. W., Koballa, T. R., Jr., Flage, L. R. & Pyle, E. J. (1996) WOWBugs: New Life for Life Science. Riverview Press, Athens. Matthews, R. W. (1997) Weird wonderful WOWBugs. Carolina Tips 60 (2): 1-4. Matthews, R. W., Flage, L. R., & Matthews, J. R. (1997) Insects as teaching tools in primary and secondary education. Annual Review of Entomology 42: 269-289. National Research Council. (1995) National Science Education Standards. National Academy of Sciences, Washington, D.C. Simpson, R. D. & Oliver, J. S. (1985) Attitudes toward science and achievement motivation profiles of male and female science students in grades six through ten. Science Education 69: 511-26.
Developing Life Science Instructional Materials Using a Parasitic Wasp 433
Talton, E. L. & Simpson, R. D. (1985) Relationships between peer and individual attitudes toward science among adolescent students. Science Education 69: 19-24. Williams, P. (1995) Exploring with Wisconsin Fast Plants. A Resource for Primary and Secondary Teachers. Kendall-Hunt, Dubuque.
This Page Intentionally Left Blank
PART
9
Medical Effects of Hymenoptera
This Page Intentionally Left Blank
Review of Bee and Wasp Sting Injuries in Australia and the USA N. R. Levick1, J. O. Schmidt2, J. Harrison3, G. S. Smith4 and K. D. Winkel5 1
Division Pediatric Emergency Medicine, Johns Hopkins University, School of Medicine, Baltimore, MD 21287-3144 USA 2
USDA-ARS Carl Hayden Bee Research Center, 2000 East Allen Rd, Tucson, Arizona 85719 USA 3
Research Centre for Injury Studies, Mark Oliphant Building, Flinders University, Bedford Park, SA 5042 Australia 4
Centre for Injury Research and Policy, Johns Hopkins School of Public Health, Baltimore, MD 21205 USA. 5
Australian Venom Research Unit, Department of Pharmacology, University of Melbourne, Victoria 3052 Australia (email:
[email protected])
Introduction The surge in numbers of the introduced European wasp (Vespula germanica F.) in south-eastern Australia during the summer of 1997–98 prompted the Victorian Government to call for a national wasp control strategy, and the South Australian Government to fund research into potential control measures. An increase in the distribution of these wasps had been predicted (Spradbery & Maywald 1992), and public health problems anticipated (Levick & Braitberg 1996; Levick et al. 1997). Unfortunately, whilst the toxins of the honey bee (Apis mellifera L.) (Schmidt 1995) and some vespids (King 1996) have been extensively studied, much less is known about the broader health care impact of this type of injury. In particular, little has been documented about the hazards of wasp stings in Australia. Previous reports, however, do indicate that bee and wasp stings are established as a major cause of bite and sting injuries in both Australia and the U.S.A. (Harvey et al. 1984; Southcott 1988; Langley & Morrow 1997). This paper reviews the extent of the bee and wasp sting problem in these two countries, the bee and wasp species involved, and discusses the limitations of current information sources.
Bees and Wasps in the U.S.A. Various native and introduced vespid wasps, termed ‘yellowjackets’ are abundant and conspicuous in the U.S.A. Seventeen native species in two genera (Vespula Thomson and Dolichovespula Rohwer), plus the exotic European hornet (Vespa crabro L.) and European wasp (often referred to as the German yellowjacket) are present (see Akre et al. 1980). The European wasp has rapidly become the most medically important of these species. Yellowjackets are most abundant in the northern and central temperate regions of the U.S.A. and most of Canada. They are infrequent in southern U.S.A. and, until recently, were absent in the desert habitats of the south-west. A single vespid species is found in Hawaii (V. pennsylvanica Saussure). In addition to yellowjackets, paper wasps (Polistes spp.), bumblebees (Bombus spp.), sweat bees (Halictidae), and especially honey bees represent other medically significant stinging insects in
437
N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel 438
Hymenoptera: Evolution, Biodiversity and Biological Control
the U.S.A. (Schmidt 1992). Although the absolute importance of wasps versus honey bees in causing mortality and morbidity cannot yet be determined, yellowjackets are considered to represent a greater hazard (Schmidt 1992). The European wasp is a particular problem as it has a predilection for nesting in urban areas, including in the walls or cavities of houses and other man-made structures (Morse et al. 1977). Vespula germanica is native to Europe, western Asia and northern Africa, where it is often abundant and a nuisance pest. It was introduced into North America several times during the last 100+ years, but did not become firmly established until 1968 when it was reported from Maryland (Morse et al. 1977; also see Menke & Snelling 1975). Since that time, V. germanica has rapidly expanded its range in the U.S.A. to include virtually all of the northern half of the country, southern Ontario, the Pacific region including British Columbia, but is apparently absent from the Rocky Mountains, and the southern and south-east states. It now dominates the wasp fauna in much of North America, often displacing or reducing populations of native wasp species in the process. The success of this species relates, in part, to its ability to thrive amongst human surroundings and because its young queen-founded colonies develop faster than other species (Matthews et al. 1982). It also forages more aggressively at feeding sites (Parrish & Fowler 1983).
Bees and Wasps in Australia Various native and introduced bees and wasps represent a potential health hazard in Australia. However, due to their wide distribution and cultivation by humans (Gibbs & Muirhead 1998), honey bees have been the major medical problem. Feral honey bees have caused particular problems when the density of their colonies is high, such as in the Wyperfeld National Park, northern Victoria, which has a feral bee population that may be the highest in the world (up to 148 colonies per km2) (Honan 1997a). Most of the >3000 species of native Australian bees (CSIRO 1991) have been presumed to be harmless and nothing is known of their venom. All female bees, with the exception of Trigona spp., have a sting and venom which has the potential to trigger allergic reactions, including anaphylaxis. However, such events are rare because the great majority of these species are solitary (i.e. not social) and therefore not aggressive around a nest site. However, there has been at least one fatality attributable to a native bee (Morris et al. 1988). In most temperate and some tropical parts of Australia native halictid bees are abundant, and four out of five native bee stings, including the fatality reported by Morris et al. (1988), were caused by members of this family. Also a single case of a large local skin reaction has been reported from the sting of a colletid bee (Morris et al. 1988), another diverse family of native bees in Australia. Several medically troublesome wasps have been introduced to Australia, but of these V. germanica is the most important. This species first arrived in Australia in 1954 but only became established in 1959 in Hobart. Although quickly becoming widespread in Tasmania, it only reached the mainland in 1977 (around Melbourne). Over the next 10 years V. germanica spread throughout Victoria and into N.S.W., the A.C.T., South Australia (S.A.) and Western Australia (W.A.). Isolated nests have also been detected in Queensland and a single queen was reported in the Northern Territory (Spradbery & Maywald 1992). By 1991 many thousands of nests were estimated to be destroyed in metropolitan Melbourne annually, with wasp densities of up to 40 per km2 being reported (Crosland 1991). Despite this, ecological modeling suggests the potential for a further significant increase in this wasp’s distribution in Australia (Spradbery & Maywald 1992). Potentially important in Australia is the overwintering of wasp colonies, a phenomenon rare in the endemic range of Europe and North Africa (Spradbery 1973). These perennial
Review of Bee and Wasp Sting Injuries in Australia and the USA 439
colonies may be 4 m in length and contain millions of wasps (Spradbery 1988). Whilst the current geographic range and abundance of V. germanica in Australia is only approximately known, they are now considered established in the capital cities and large county towns of Victoria, South Australia, Tasmania and eastern N.S.W. (Honan 1997b). By contrast, the English wasp, V. vulgaris (L.), which arrived in Victoria in 1958, has not spread beyond specific areas of greater Melbourne. There are also various species of introduced and native paper wasps throughout Australia, including members of Polistes Latreille and Ropalidia Guérin. In some states, such as Queensland, native paper wasps are the major cause of severe allergic reaction attributable to wasps (Solley 1990).
Bees Versus Wasps: Appearance, Behaviour and Venom Chemistry The differences between honey bee and European wasp behaviour are summarised in Table 1. Honey bees are predominantly drab brown or yellow-brown in colour, have a pilose body, and leave their sting in the victim. European wasps have a distinctive black and yellow colour pattern, a much less hairy surface, are slightly larger and more elongate than bees, and virtually never leave their stings in the victim. Individual honey bees have between 150-250 µg venom/bee while European wasps have about 60–100 µg/wasp (Schmidt 1990, 1992; Schmidt et al. 1990). When a bee stings, the vast majority of the venom is injected via the prolonged pumping action of the automised sting. However, a wasp sting often lasts less than a second, and it is unknown how much venom is injected. Swarms of wasps and ‘killer’ africanised bees have been observed to sting mostly on the head and neck of their victim, presumably as these are the most exposed areas of skin. Hence, when a victim presents for emergency care, it is important for the clinician to check over the scalp for stings hidden under the hair. In a well documented pediatric fatality (Korman et al. 1990), the extent of envenomation was not appreciated until post mortem when >100 sting lesions were found on the victim’s scalp. The venoms of the two species share little in common biochemically. Bee venom contains mainly a pain-inducing peptide called melittin, plus apamin, mast-cell-degranulating peptide and several enzymes including hyaluronidase and a phospholipase A2. European wasp venom contain predominantly kinins, peptides similar in structure to the cardiac-active bradykinin, antigen 5, and enzymes including hyaluronidase and phospholipase A1B (Schmidt 1992). Further, the enzymes of the two species are structurally different and, consequently, their venoms exhibit very little antigenic cross-reactivity (Hoffman et al. 1988). These differences in venom composition also mean that the toxic syndrome induced by bees and wasps should, ideally, be treated with specific therapy. However, there are no antivenoms available to neutralise the potentially lethal effects of these venoms, and at present both are managed by non-specific supportive care.
Bee and Wasp Sting Injuries in the U.S.A. Mortality Bee, wasp and hornet stings now constitute the majority of fatalities from specified animal attacks in the U.S.A. (Langley & Morrow 1997). Data from 1979-90 death certificates attributed an average of 44 deaths/year to bee, wasp or hornet stings (a fatality rate of 0.184/106/year) which is three times that for dog bites and 10 times that of the snake or spider bites, and an apparent increase from the last survey (0.14/106/year) (Parrish 1963). It is unclear what proportion of these deaths were due to bees versus wasps and what proportion of these recent fatalities were due
N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel 440
Table 1
Hymenoptera: Evolution, Biodiversity and Biological Control
Summary of the behavioural characteristics of honey bees and European wasp.
European Wasp
Honey Bee
Predatory on insects except in autumn when they Pollen and nectar feeders forage for honey dew and other natural or human sugar sources Will swarm and attack but only near nest site; more aggressive in late summer than at other times of year
Swarm at specific times but not attack (except Africanised ‘killer’ bees, which are rare in the U.S.A. and are not found in Australia)
Are scavengers, attracted to food around domestic environs
Not attracted to food or trash
Nests often at ground level, camouflaged, often mistaken for old cardboard or obscured by soil or rocks where they may be accidentally disturbed
Nests are seldom at ground level (except professional beehives)
Capable of stinging multiple times; do not leave the stinging apparatus in the victim
Individual bees sting only once; usually leave the stinging apparatus in the victim
Venom may result in more severe local tissue reaction, possibly due to more severe local chemical toxicity; cause infections from the sting
Cause less severe local reactions than do wasps; rarely cause infection from the sting
to allergic reactions compared with venom toxicity. A previous survey of 400 hymenopteran stings in the U.S.A. found that 44% were due to wasps and hornets and 44% due to bees (Barnard 1973). Most fatalities occurred in males older than 40 years suggesting vulnerability of this group due to underlying cardiovascular and/or pulmonary disease.
Morbidity There are a number of databases that capture information about non-fatal bee, wasp and hornet stings in the U.S.A. Toxic Exposure Surveillance System (TESS) data is compiled and published annually by the American Association of Poison Control Centers, representing the majority of American Poison Control Centers (PCC) (Litovitz et al. 1998). This system represents an estimated 94% of human poison exposures in the U.S. that precipitate PCC contact, including bee and wasp sting calls made by the general public. In 1996, bee, wasp and hornet stings represented 16% of all bite and sting calls (total 16 336 calls) and were the largest single bite and sting call category. Information on more serious morbidity is available via the 1992–95 National Ambulatory Medical Care Survey, a large survey sampling the U.S. population (~268 million). This survey estimated that there were at least 314 500 Emergency Department (ED) bee, wasp and hornet presentations annually in the U.S.A. resulting in at least 1800 hospital admissions during the survey period. The estimated cost of these emergency department visits and hospital admissions was in excess of $72 million (Rebecca Spicer, Pacific Institute of Research and Evaluation, Maryland, pers. comm.). This estimate excludes lost economic productivity and reduced quality of life.
Bee and Wasp Sting Injuries in Australia There is a lack of population-based data on the incidence of bee and wasp stings in Australia. Estimates can be made from cases admitted to hospitals and by extrapolating from the small number of recorded deaths. Information is also available for emergency department attendances
Review of Bee and Wasp Sting Injuries in Australia and the USA 441
in south-eastern Victoria, although these data have substantial limitations (see limitations of data sources, below).
Mortality Nation-wide, 43 deaths were reported to the International Classification of Diseases (9th revision, ICD-9), cause category (E905.3) by the Australian Bureau of Statistics (ABS) for the 19 year period from January 1979 to December 1997. Only venomous snake bite was a commoner cause of death due to venomous creatures in Australia during this period (n=48). The 43 cases can be summarised as follows. Most victims were male (37/43), a bias which has been documented in previous studies and probably reflects higher bee and wasp exposure and underlying susceptibility for males. Two deaths were recorded amongst those less than 30 years of age, and highest for ages 70 and older. Deaths occurred in every state and territory with N.S.W. (13 deaths), Victoria (11) and Queensland (9) having the largest number. The small numbers make interpretation difficult, however, there is apparently no clear correlation with the incidence of European wasp, given that the species is virtually absent in Queensland, and often highly abundant in cities and towns in South Australia and Tasmania (the same also seems to be true for data on hospital admissions – see below under Morbidity). There are indications that mortality from bee and wasp stings rose for several years in the mid-1990’s. Harvey et al. (1984) argued that ABS mortality data available in the early 1980’s was likely to underestimate the true mortality from bee and wasp stings due to under-reporting. It remains an open question whether the apparently higher rates seen more recently are the result of more accurate estimates, or the increased incidence of stings, or both. Either way, it is likely that the public health burden from this condition is somewhat higher than was estimated by Harvey et al. (1984). Clearly, there is a need to better document the incidence and species responsible for stings which lead to serious medical effects and death. Morbidity One indication of the medical significance of wasp stings comes from an analysis of calls received by the Poisons Information Centres (PIC). For example, in 1997, 864 wasp sting calls were received by the PIC (NSW PIC 1998). This was the most frequent call for identified insect stings (bees were second with 427 calls). The only state-wide injury dataset in Australia is the Victorian Emergency Minimum Dataset (VEMD), managed by Monash University’s Accident Research Centre, which provides information for approximately 80% of injury cases attending 25 emergency departments. No other such comprehensive state-wide emergency department injury data collection is currently available. However, even the VEMD, which is still under development, failed to ‘capture’ most of the patients that were reported to be admitted to hospital with bee and wasp stings in 1996–7 (Winkel et al. 1998). Within these limitations it was notable that bee and wasp stings represented 41% of all bite and sting presentations identified for the period October 1995 to December 1997. Bees were attributed to 30% and wasp 11% of all sting presentations. A coding scheme used widely for hospital discharge diagnoses is the clinical modification of the ICD-9 (ICD-9-CM) and the following version, ICD-10. Unfortunately, neither version distinguishes between bee or wasp related fatalities, however, a clinical modification of ICD-10 has been developed for use in Australia (ICD-10-AM). The second edition of ICD-10-AM provides separate categories for coding cases due to wasps and bees. The proportion of cases in which information is available from clinical records and allows the use of specific taxon categories
N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel 442
Hymenoptera: Evolution, Biodiversity and Biological Control
remains to be seen. This will be of particular importance as there is often confusion between bees and various native and introduced wasp species. Data are available on hospital admissions attributed to bee and wasp stings. Comparisons of hospital morbidity data over time and between jurisdictions have many limitations and therefore great care should be taken before drawing conclusions about apparent differences (Mclntyre et al. 1997). Data for a single year (July 1996 to June 1997) are presented here. Over this period, 977 new cases of bee and wasp stings resulted in admission to an Australian hospital, or 53 admissions per 106 population. These cases made up 30% of 3288 cases attributable to injury by venomous animals and plants. Case numbers were highest in summer with 33% of cases occurred in January and February. The largest number of cases was recorded in Queensland (308), followed by N.S.W. (246), S.A. (152), W.A. (147) and Victoria (119). The other States and Territories recorded fewer than 10 cases each. Corresponding rates were highest in S.A. (103 per million population), Queensland (91 pmp) and W.A. (83 pmp). Conservative direct inpatient hospital costs, assuming a one day admission for all patients, are estimated to be approximately $0.5 million.
Limitations of the Data Collection of more useful data on mortality and morbidity due to bee and wasp stings is hindered by 1) the lack of formal coding categories to record the exact species involved; 2) difficulties in ensuring that cases due to such stings are reliably and completely identified as such in clinical records, and 3) difficulties attributing a particular clinical case to an exact species. As noted above, the first of these problems is being addressed in the second Australian edition of the ICD10 (ICD-10-AM). However, the ability to code cases accurately will depend upon the quality of information in the medical record. There are currently no obvious solutions to the second and third problems. Clinical education concerning case diagnosis may reduce the second problem, although some cases of anaphylaxis due to a hymenopteran sting may escape attribution as such, even if there is an appropriate level of clinical suspicion. Reliable identification of the stinging species is the most difficult problem. While a sting in the skin can be used to identify A. mellifera, there is no equivalently specific criterion for identifying cases due to various species of wasps. Only rarely is the specimen available for later professional identification.
Envenomation Versus Allergy Bees and wasps inject various substances with their sting, some of which are directly toxic causing an ‘envenomation’ (if in a sufficient dose) whilst others do harm by triggering dangerous allergic reactions (even in small doses). Thus, several distinct health problems can result from these stings. A single sting can cause local pain, redness, swelling and itch which, if it occurs in the mouth or throat, can obstruct the airway and be life threatening. In allergic individuals a single sting to the limbs can also trigger a potentially lethal reaction (Glaspole et al. 1997). This is in contrast to the direct toxic effects of the venom (see Table 2) (Wong 1970; Barss 1989; Korman, et al. 1990; Levick & Braitberg 1996; Kolecki 1999).
Injury Prevention and Control It is important to consider the prevention and control issues that impact on the healthcare burden resulting from bee and wasp strings. Principle issues are the awareness and education of the risk of stings, eradication of wasp nests in areas where there is a risk of sting exposure,
Review of Bee and Wasp Sting Injuries in Australia and the USA 443
Table 2 Clinical manifestations of bee and wasp stings: a comparison of venom allergy with venom toxicity. Allergy
Toxicity
Itch, urticaria, wheeze, oedema, erythema, dizziness, weakness
Encephalopathy, CNS depression, agitation, optic neuritis, acute and chronic extrapyramidal syndromes, cerebral infarction, hypertension, tachycardia, dysrrhythmias, thirst, nausea & vomiting, nephrotic syndrome, hepatotoxicity, haemolysis, coagulopathy, thrombocytopenia, rhabdomyolysis, acute renal failure, hyperkalaemia, hyperglycaemia, hypocalcaemia
Progressing to: anaphylaxis: respiratory distress, generalised erythema, nausea & vomiting, abdominal pain, syncope, hypotension, shock
particularly to children, and advice on safe methods of nest destruction. Involvement of local councils in nest eradication programs has had some success in Australia (Honan 1997b). Continued education of the health professions and the public as to the risks and management of bee and wasp envenomation and allergy is important for secondary and tertiary injury prevention. This issue is discussed in more detail by Winkel et al. (1998).
Injury Surveillance Public health surveillance is crucial for measuring and characterising the health impact of bee and wasp stings and for monitoring changes over time. The most promising way to determine this is by means of surveys. The practicability of doing so depends on the incidence of the condition and the period during which recall of the condition is reliable. Given the sudden and dramatic nature of stinging events, recall by patients may be better than for many other conditions. Opportunities should be sought to include appropriate questions in large general population surveys. Death records and hospital admissions are probably the best practical criteria for defining ‘severe’ cases resulting from bee and wasp stings. The ABS mortality data provide the only routine historical information on deaths attributed to bee and wasp stings. Completion of the National Coronial Information System in Australia promises more detailed case information on this uncommon but important group of cases and the ability to follow-up with in-depth analysis. In the U.S.A., data from state medical examiners could be collected and employed in the same way, but the processes are not as well developed as in Australia. In Australia inpatient data are state-based and compiled nationally. In recent years their scope has become more-or-less complete, and their quality and timeliness has improved. ICD-10-AM diagnosis codes and external cause codes are applied to all cases. While they cover only a small proportion of sting cases, these are likely to include those that have life-threatening consequences. In the U.S.A. similar state-based hospital data exists. However, ‘Admission to a hospital’ is not an ideal criterion, as it may be influenced by the availability of services, clinical practice, economic factors and clinical condition. A case severity criterion should be sought which can be applied independently of these factors. The VEMD can provide important information on the circumstances and some short term consequences of hymenopteran stings, including cases that do not result in admission to a hospital. Such data is also a potential information source for case-control studies of risk factors for stings. It has limited potential for assessing rates and trends, mainly because of the poorly defined proportion of incident cases that attend an emergency department and are captured by the VEMD.
N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel 444
Hymenoptera: Evolution, Biodiversity and Biological Control
The U.S. Consumer Product Safety Commission (CPSC) currently only deals with productrelated injury surveillance and does not address cases attributable to bites and stings. However, the 1999 National Academy of Sciences report on ‘Injuries in America’ (Bonnie et al. 1999) recommended expanding its scope to include all injuries which would then provide valuable information on a national basis.
Other data sources Australian PIC and the U.S. PCC may be a useful resource for obtaining information that helps to characterise the circumstances of hymenopteran stings and related morbidity in the community, as most patients do not require medical attention. They are not a useful source for estimating incidence rates or trends in incidence rates because of the poorly defined relationship between incident cases and the cases that result in calls. Other useful organisational resources include: The Australian Venom Research Unit (AVRU) which is uniquely positioned in Australia and internationally to be a focal point for the interdisciplinary surveillance of hymenopteran sting-related healthcare problems and research on venom and toxicity; The Australian Institute of Health and Welfare National Injury Surveillance Unit (NISU) at the Research Center for Injury Studies which is the national centre for developing and reporting on data collection useful for surveillance of injury, including bites and stings, and Organisations such as the Australia’s CSIRO, Departments of Agriculture, and University Departments which are useful sources of information for monitoring and predicting changes in the distribution and abundance of hymenopteran species implicated in causing medical problems.
Future Research, Policy and Recommendations Further research is required both in Australia and the U.S.A. to determine the true incidence of wasp sting injuries, the species which cause them, the nature of the presenting problem (allergy or envenomation), and to review current medical practices and assess outcomes of injury prevention strategies. The mechanism involved in fatalities is also important from an injury control perspective. International experience with wasp stings suggests that many deaths occur quickly after one or a small number of stings from a single insect (Barnard 1973; Mosbech 1983; Meier 1995). Furthermore, many such deaths occur in patients that do not have a past history of wasp venom allergy (Mosbech 1983). Therefore, prophylactic therapy for people with known life threatening wasp allergies by immunotherapy may not prevent many wasp sting deaths, although it may reduce the overall healthcare burden. Reducing the risk of death through measures aimed at preventing sting events, and heightened awareness of symptoms and management of allergic reactions and envenomation syndromes, may be more successful. In this respect it is important to distinguish an allergy related single sting from envenomation associated with multiple stings. Recognition by patients and emergency department staff of the distinguishing characteristics of the various stinging species involved is essential to accurate data collection and optimal case management. Evaluation of the knowledge level of medical staff about these insects should be undertaken, and the relative merits of education evaluated. The introduction in the second edition of ICD-10-AM of separate categories for coding bee stings and wasp sting cases is a first step and should be encouraged in revisions of ICD-10. However, the latter category still includes
Review of Bee and Wasp Sting Injuries in Australia and the USA 445
numerous species of wasps (both native and introduced) and therefore does not allow for accurate reporting of the incidence of stings and symptoms associated with particular species whose distribution and abundance may vary geographically. Information sheets and posters specifically developed for surgeries, out-patient clinincs and emergency departments, similar to those produced by CSIRO Entomology and the S.A. Local Government Association, will obviously help in this regard. The next step is to evaluate the extent to which it is feasible to identify species-related factors using current clinical records, and to identify potentially correctable barriers. An integral aspect of surveillance, research and policy with respect to hymenopteran stingrelated injury is interdisciplinary collaboration. This approach requires an interface between health care (public health, toxicology/immunology, emergency medicine) and entomology (CSIRO, Departments of Agriculture, Universities). Each of these disciplines is essential to our understanding of the magnitude and key determinants of the healthcare burden caused by hymenopteran sting-related injury and determining ways to minimise this overall healthcare burden. In Australia the following strategies have been recommended by the AVRU specifically to address European wasp related injuries and a similar approach is applicable for the U.S.A.: 1) encourage federal emergency department injury surveillance funding; 2) provide dedicated envenomation injury surveillance funding, including in depth follow-up of all deaths using coroners and other records; 3) apply the collected data to the primary, secondary and tertiary prevention of the European wasp health care burden; 4) reduce the total wasp population around sites of significant human activity; 5) promote wasp sting prevention and allergy awareness; 6) promote early effective acute envenomation and allergy management; 7) promote appropriate use of immunotherapy to reduce allergy risks; 8) recommend adoption of envenomations as notifiable diseases; 9) fund evaluation of medical and public health interventions, and 10) target a 50% reduction in wasp-related morbidity and mortality.
Acknowledgements This work was supported by the Victorian Department of Human Services, Snowy Nominees and the BHP Community Trust.
References Akre, R. D., Greene, A., MacDonald, J. F., Landolt, P. J., & Davis, H. G. (1980) Yellowjackets of America north of Mexico. United States Department of Agriculture, Agriculture Handbook No. 552, 102 pp. U.S. Government Printing Office Washington, D. C. Barnard, J. (1973) Studies of 400 Hymenopteran sting deaths in the United States. Journal of Allergy and Clinical Immunology 52: 259-264. Barss, P. (1989) Renal failure and death after multiple stings in Papua New Guinea. Ecology, prevention and management of attacks by vespid wasps. Medical Journal of Australia 151: 659-663. Bonnie R. J., Fulco C. F., Liverman, C. T. (Eds) (1999) Reducing the Burden of Injury. Advancing Prevention and Treatment. Committee on Injury Prevention and Control, Division of Health Promotion and Disease Prevention, Institute of Medicine. National Academy Press, Washington, D.C. CSIRO. (1991) The Insects of Australia. 2 vols, 2nd edition. Melbourne University Press, Melbourne.
N. R. Levick, J. O. Schmidt, J. Harrison, G. S. Smith and K. D. Winkel 446
Hymenoptera: Evolution, Biodiversity and Biological Control
Crosland, M. W. J. (1991) The spread of the social wasp, Vespula germanica, in Australia. New Zealand Journal of Zoology 18: 375-388. Gibbs, D. M. H. & Muirhead, I. F. (1998) The economic value and environmental impact of the Australian Beekeeping Industry. Unpublished report for the Australian Beekeeping Industry. Glaspole, I., Douglass, J., Czarny, D. & O’Hehir, R. (1997) Stinging insect allergies: assessing and managing. Austalian Family Physician 26: 1395-1399. Harvey, P., Sperber, S., Kette, F., Heddle, R. J. & Roberts-Thomson, P.J. (1984) Bee sting mortality in Australia. Medical Journal of Australia 140: 209-211. Hoffman, D. R., Dove, D. E., Moffitt, J. E. & Stafford, C. T. (1988). Allergens in Hymenoptera venom XXI. Cross-reactivity and multiple reactivity between fire ant venom and bee and wasp venoms. Journal of Allergy and Clinical Immunology 82: 828-834. Honan, P. (1997a) Management of Feral Bees in Wyperfeld National Park. Issues and Recommendations for Research. Keith Turnbull Research Institute, Agriculture Victoria. Honan, P. (1997b) Proceedings of the European Wasp Strategy Meeting. 11 September, Victorian Government. King, T. P. (1996) Immunochemical studies of stinging insect venom allergens. Toxicon 34: 14551458. Kolecki, P. (1999) Delayed toxic reaction following massive bee envenomation. Annals of Emergency Medicine 33: 114-116. Korman, S. H., Jabbour, S. & Harari, M. D. (1990) Multiple hornet stings with fatal outcome in a child. Journal of Paediatric and Child Health 26: 283-285. Langley, R. L. & Morrow, W. E. (1997) Deaths resulting from animal attacks in the United States. Wilderness & Environmental Medicine 8: 8-16. Levick, N. R. & Braitberg, G. (1996) Massive European wasp envenomation of a child. Emergency Medicine 8: 239-245. Levick, N. R., Winkel, K. D. & Smith, G. S. (1997) European wasps: an emerging hazard in Australia. Medical Journal of Australia 167: 650-651. Litovitz T. L., Klein-Schwartz, W., Dyer, K. S., Shannon, M., Lee, S. & Powers, M. (1998) 1997 Annual Report of the American Association of Poison Control Centers Toxic Exposure Surveillance System. American Journal of Emergency Medicine 165: 443. Matthews, R. W., Ross, K. G. & Morse, R. A. (1982) Comparative development of queen nests of four species of yellowjackets (Hymenoptera: Vespidae) reared under identical conditions. Annals of the Entomological Society of America 75: 123-29. Mclntyre, C. R., Ackland, M. J., Chandraraj, E. J. & Pilla, J. E. (1997) Accuracy of ICD-9-CM codes in hospital morbidity data, Victoria: implications for public health research. Australian and New Zealand Journal of Public Health 21: 477-482. Meier, J. (1995) Biology and distribution of hymenopterans of medical importance, their venom apparatus and venom composition. pp 331-348. In Meier, J. & White, J. (Eds), Handbook of Clinical Toxicology of Animal Venoms and Poisons. CRC Press, Boca Raton, Florida. Menke, A. S. & Snelling, S. (1975) Vespula germanica (Fabricius), an adventive yellow jacket in the northeastern United States (Hymenoptera: Vespidae). U. S. Department of Agriculture Coop Economic Insect Report 25: 193-200. Morris, B., Southcott, R. V. & Gale, A. E. (1988) Effects of stings of Australian native bees. Medical Journal of Australia 149: 707-709. Morse, R. A., Eickwort, G. C. & Jacobson, R. S. (1977) The economic status of an immigrant yellowjacket, Vespula germanica (Hymenoptera: Vespidae), in Northeastern United States. Environmental Entomology 6: 109-110.
Review of Bee and Wasp Sting Injuries in Australia and the USA 447
Mosbech, H. (1983) Deaths caused by wasp and bee stings in Denmark 1960-1980. Allergy 38: 195-200. New South Wales Poisons Information Centre (1998) Annual Report. The New Children’s Hospital, Royal Alexandra Hospital for Children, Westmead. Parrish, H. M. (1963) Analysis of 460 fatalities from venomous animals in the United States. American Journal of Medical Science 245: 129-141. Parrish, M. D. & Fowler, H. G. (1983) Contrasting foraging related behaviours in two sympatric wasps (Vespula maculifrons and V. germanica). Ecological Entomology 8: 185-90. Schmidt, J. O. (1990) Hymenoptera venoms: striving toward the ultimate defense against vertebrates. pp. 387-419. In Evans, D. L. & Schmidt, J. O. (Eds), Insect Defenses. SUNY Press, Albany, New York. Schmidt, J. O. (1992) Allergy to venomous insects. pp. 209-69. In Graham, J. M. (Ed), The Hive and the Honey Bee. Dadant & Sons, Hamilton, Ilinois. Schmidt, J. O. (1995) Toxinology of venoms from the honeybee genus Apis. Toxicon 33: 917-927. Schmidt, J. O., Blum, M. S. and Overal, W. L. (1990) Comparative lethality of venoms from stinging Hymenoptera. Toxicon 18: 469-474. Solley, G. (1990) Allergy to stinging and biting insects in Queensland. Medical Journal of Australia 153: 650-654. Southcott, R. V. (1988) Some harmful Australian insects. Medical Journal of Australia 149: 656654. Spradbery, J. P. (1973) The European wasp Paravespula germanica (F.) (Hymenoptera: Vespidae) in Tasmania, Australia. pp.375-380. Proceedings of the VII Congress of the International Union for the Study of Social Insects, London. Spradbery, J. P. (1988) The European wasp in Australia: present status and future prospects. Proceedings of the Sydney Allergen Group 6: 78-86. Spradbery, J. P. & Maywald, G. F. (1992) The distribution of the European or German wasp in Australia, past, present and future. Australian Journal of Zoology 40: 495-510. Winkel, K., Hawdon, G. & Ashby, K. (1998) Venomous bites and stings. Victorian Injury Surveillance System Hazard 35: 1-16. Wong, H. B. (1970) Wasp and bee stings in children. Journal of the Singapore Pediatric Society 12: 126-134.
This Page Intentionally Left Blank
PART
10
Future Research
This Page Intentionally Left Blank
Hymenopteran Research – Future Directions into the Next Millennium Mark Dowton1,2 and Andrew D. Austin1 1
Department of Applied & Molecular Ecology, Waite Campus, The University of Adelaide, PM.B. 1 Glen Osmond, S. A. 5064 Australia (email:
[email protected]) 2
Australian Flora and Fauna Research Centre, Department of Biology, Wollongong University, Wollongong, N. S. W. 2522 Australia (email:
[email protected])
The Hymenoptera will be a focus for biologists in the 21st century for the same reasons they have been during the 20th century – they occupy a pivotal position in terrestrial ecology and biodiversity studies, and because their highly specialised biologies provide unlimited scope for natural history and evolutionary studies. Further, their continued use as biological control and pollinating agents will ensure sustained interest from applied biologists. In this section, we briefly examine some areas of research that have the potential to influence the direction of hymenopteran research into the beginning of the next millennium. In so doing, we would point out the obvious, that it becomes very difficult to speculate on research outcomes more than a few years into the future. This is particularly so when research directions are driven by technological and/or methodological advances. For example, who could have predicted the current outcomes in molecular biology prior to the advent of the polymerase chain reaction? This said, however, several areas of hymenopteran research that are currently very active have the potential over the next few years to change our way of thinking about this fascinating and important group of insects. One theme to emerge from this volume is that the parasitoid-host association is central to understanding the evolution and biodiversity of the Hymenoptera. In particular, the intricate biology of endoparasitism, and the evolutionary forces that have shaped the various parasitic lifestyles continue to dominate the attention of hymenopterists. Recent developments in our understanding of polydnavirus biology, together with resolution of the phylogenetic relationships among the polydnavirus-bearing wasps, will make possible the dissection of this complex interaction within a framework of comparative biology. The microgastroid braconids represent one of the best ‘model groups’ to examine the evolution of parasitism. All microgastroid subfamilies appear to contain polydnavirus DNA (Whitfield 1997; this volume), molecules that apparently facilitate avoidance/suppression of the host’s immune system. This association has been a long one (at least 60 Myr), and is likely to have been crucial in the ability of these wasps to parasitise a variety of lepidopteran hosts. Further, recent studies are beginning to elucidate mechanisms by which polydnaviruses facilitate parasitoid survival (e.g. Pennacchio et al.; Schmidt et al., this volume). Although such studies have been performed on isolated species, as more information accumulates it should become possible to assess the phylogenetic spread of these mechanisms. For example, it will be of particular interest to determine whether various microgastroids with variant parasitoid lifestyles utilise different mechanisms of immune system avoidance. Given that the biology of microgastroids at the
451
Mark Dowton and Andrew D. Austin 452
Hymenoptera: Evolution, Biodiversity and Biological Control
subfamily level is relatively constant, once a relatively robust phylogeny is available, it will be possible to dissect the evolution of host immune system avoidance strategies using the methods of comparative biology (Harvey & Pagel 1991). The relationships among microgastroid subfamilies are now relatively well understood (Whitfield 1997; Dowton & Austin 1998). However, to more fully understand the co-evolutionary dynamics of the wasp/polydnavirus association, a more detailed understanding of the phylogenetic relationships within microgastroid subfamilies will be required, particularly for the Microgastrinae which comprises the vast majority of species within this assemblage. Such studies using both morphological and sequence data from multiple genes are currently underway (J. Whitfield, pers. comm.). Other ‘big’ questions in parasitoid evolution concern host switching between higher level groups and transitions between various modes of parasitism (particularly between ecto- and endoparasitism). Because of their diversity of parasitic lifestyles, the braconids have much potential for furthering progress in these areas (see Gauld 1988), although such work continues to be frustrated by the lack of a phylogenetic framework within which to interpret evolutionary transitions. Recent molecular phylogenetic studies have consistently resolved certain (but by no means all) braconid affiliations (Dowton et al. 1998; Belshaw et al. 1998, 2000). Nevertheless, many areas of the braconid tree are poorly known, and will require additional information for their elucidation. Not surprisingly, the same is also true of other major parasitoid lineages such as the Ichneumonidae and Chalcidoidea (see Quicke et al. and Campbell et al., this volume). Greatly improved phylogenetic resolution of these groups should result from analysis of new character systems emerging from work presented in this volume. These include mitochondrial gene rearrangement data, molecular data from multiple genes, morphological data coded for specific exemplars, and possibly nuclear genome organisational data. Although the latter character has traditionally been considered an unreliable phylogenetic indicator, Gokhman (this volume) has determined that chromosomal rearrangements have occurred relatively frequently during the evolution of the Hymenoptera. Other recent developments in the detection and analysis of chromosomal rearrangements (e.g. Müller et al. 2000) render this a promising character for phylogenetics at lower levels, such as among tribes. Accurate knowledge of evolutionary history is crucial in many areas of hymenopteran biology, yet remains elusive. In spite of the tantalising hope that comparative molecular data would make great inroads into phylogenetics generally, the last 10 years have seen very few convincingly resolved phylogenies. Although a number of studies have recommended that dense taxonomic sampling is crucial to break up long branches, it is also clear that multi-gene studies will be required to overcome the inherent ‘noisiness’ of molecular data. An important lesson may be learnt from the recent resolution of the angiosperm phylogeny (Soltis et al. 1999). Earlier phylogenetic analyses, employing hundreds of taxa for a single gene, suffered from poor resolution and/or weak support in many parts of the tree. The inclusion of additional data, sequences from three extra genes in the angiosperm case, greatly improved both resolution and support. Perhaps what is now emerging is a realisation of just how much data are necessary to resolve phylogeny. This will require a substantially increased sequencing effort, but one now made possible by recent developments in sequencing and computer technology that allow for the generation and analysis of data-sets several times larger than were possible just a few years ago – presumably this trend will continue. Comparative embryology is another area that has enjoyed a resurgence of interest due to the discovery of highly conserved genes that appear to critically direct embryo- and morphogenesis.
Hymenopteran Research – Future Directions into the Next Millennium 453
The pattern of expression of these developmental genes promises to allow the critical dissection of morphological homology (e.g. Averof & Cohen 1997), and moreover will be informative about the biology of insect development generally. It has been claimed by some (e.g. Tautz et al. 1994) that the pattern of embryogenesis described for Drosophila is likely to apply to insects generally. Although patterns of gene expression during embryogenesis have been studied only in a few Hymenoptera, it appears that significant alterations to this pattern occur, and that such changes may be driven by the conditions under which eggs develop, including environmental effects (Strand, this volume). During the next few years, it should be possible to make detailed comparisons among developmental patterns between sister groups that differ in biologically well-defined ways. Again, the resolution of phylogenetic relationships among various superfamilies, families and subfamilies will significantly inform such work. One area in which there is likely to be significant advances during the next decade is in documenting biodiversity of Hymenoptera. Work to date in this area can be separated into three distinct but related activities: 1) documenting new species, 2) surveying and collecting, and 3) making available already existing information. Unfortunately, anecdotal evidence indicates that the first two of these may not be sustained at the same level of activity that has been evident during the last 20 years or so. As we enter the new millennium, there is an urgent need to increase (not decrease) taxonomic research on hymenopteran and other insect faunas of the poorly studied regions of the world. Many of the major international insect collections, such as the Natural History Museum (London), the Smithsonian (Washington, D.C.), the Canadian Insect Collection (Ottawa) and the Australian National Insect Collection (Canberra), have reduced staff numbers, while other staff have shifted away from taxonomic studies to phylogenetic research. The recognition of ‘megadiverse’ countries (see Introduction), and the knowledge that many unique habitats are rapidly disappearing or being irreversibly degraded, appears not to have caused a shift in institutional or governmental priorities. Indeed, descriptive taxonomic research is little rewarded within universities and even in some museums. Quite sensibly, conservation programs world-wide are rapidly moving from a focus on individual species to one on habitat preservation. In parallel with this, the major museums and other scientific institutions of the world have a real opportunity to raise the profile of ‘habitat’ preservation by revealing and advertising the extreme biodiversity such habitats harbour. As probably the largest group of insects, studies on hymenopteran diversity can play pivotal roles in this process, particularly if they are linked to research on their function in ecosystems as pollinators, specialist herbivores, predators and parasitoids. As highlighted in the Introduction to this volume, the current biodiversity program being undertaken on the Hymenoptera and other insects of Costa Rica (see Gámez & Gauld 1993; Hanson & Gauld 1995) can serve as a useful model for such studies in other megadiverse countries. Making already existing information available to a wider scientific community has become an increasing priority for many institutions and individuals. The advent of sophisticated databasing programs, coupled with the global adoption of internet technology (see Johnson & Musetti, this volume), is providing the means to make vast amounts of information available to the scientific and general community. Within the last few years many provincial and national museums have developed internet home pages that list the holdings in their collection, allow for electronic request of material, and/or provide access to more detailed data-bases that provide information on species distribution, biology and bibliographic information. Some programs, such as the Australian Biodiversity Information Facility (ABIF) run by the Australian Biological Resources Study, Canberra (http://www.anbg.gov.au/abrs/) are developing on-line check-lists
Mark Dowton and Andrew D. Austin 454
Hymenoptera: Evolution, Biodiversity and Biological Control
and catalogues to all groups of Australian animals including Hymenoptera. Other taxon-focused projects and the home pages of individual researchers are providing internet access to information on specific groups of Hymenoptera (e.g. Gibson, http://res.agr.ca/ecorc/apss/chalhome.htm and Johnson et al., http://iris.biosci.ohio-state.edu/home/hym_ongoing.html). In the future, as data-bases are linked either directly or indirectly and go on-line, the vast amount of information previously locked up in collections and other institutions will be available to everyone. However, what use is made of this information will be up to us! Predicting future developments in biological control using Hymenoptera is difficult because of the heterogeneity of parasitoid and host biologies, habitats and climatic regimes. It is possible that research on the behaviour and ecology of parasitoids (e.g. Keller & Tenhumberg; Hildebrands et al. this volume) may lead to more effective biological control of specific targets than has occurred in the past. However, there is an apparent lack of ecological theory derived from the practice of biological control (see Peters 1991). As seems to have been the case over the last two decades, future research in this area is likely to focus on better application of what we already know, rather than new scientific and/or technological break-throughs in insect ecology. In this respect, the work of Goolsby et al. (this volume) on the selection, characterisation and evaluation of parasitoids of Bemisia tabaci provides a good general model for procedures of best practice during the quarantine phase of any program (also see Sands, this volume). Further, research that examines the efficiency and survival of parasitoids in various agro-ecosystems is likely to provide the means by which long-term establishment and success of biological control programs can be gauged. However, in many cases both classical and augmentative control programs world-wide are run on drastically inadequate budgets. This often means that important elements to what might otherwise be well designed programs are reduced or left out – for instance taxonomic characterisation of the agents being used (Schauff & LaSalle 1998), and/or post release field evaluation of the success of the program. In the future development of effective biological control programs we might also expect to see more rigorous application of population genetics and metapopulation theory. Although elements of these disciplines have appeared in some studies in recent years, they have been more commonly associated with conservation biology than with the management of beneficial entomophagous insects.
Acknowledgements We would like to thank Jim Whitfield, Bob Matthews, John Jennings and Peter Bailey for their constructive criticism of this chapter.
References Averof, M. & Cohen, S. M. (1997) Evolutionary origin of insect wings from ancestral gills. Nature 385: 627-630. Belshaw, R., Dowton, M., Quicke, D. L. J. & Austin, A. D. (2000) Estimating ancestral geographical distributions: a Gandwanan origin for aphid parasitoids? Proceedings of the Royal Society of London B 267: 491-496. Belshaw, R., Fitton, M. G., Herniou, E., Gimeno, C. & Quicke, D. L. J. (1998) A phylogenetic reconstruction of the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28S ribosomal RNA. Systematic Entomology 23: 109-123. Dowton, M. & Austin, A. D. (1998) Phylogenetic relationships among the microgastroid wasps (Hymenoptera: Braconidae): combined analysis of 16S and 28S rDNA genes, and morphological data. Molecular Phylogenetics & Evolution 10: 354-366.
Hymenopteran Research – Future Directions into the Next Millennium 455
Dowton, M., Austin, A. D. & Antolin, M. F. (1998) Evolutionary relationships among the Braconidae (Hymenoptera: Ichneumonoidea) inferred from partial 16S rDNA gene sequences. Insect Molecular Biology 7: 129-150. Gámez, R. & Gauld, I. D. (1993) Costa Rica: An innovative approach to the study of tropical biodiversity. pp. 329-336. In LaSalle, J. & Gauld, I. D. (Eds) Hymenoptera and Biodiversity. CABI, Wallingford. Gauld, I. D. (1988) Evolutionary patterns of host utilization by ichneumonoid parasitoids (Hymenoptera: Ichneumonidae and Braconidae). Biological Journal of the Linnean Society 35: 351-377. Harvey, P. H. & Pagel, M. D. (1991) The Comparative Method in Evolutionary Biology. Oxford University Press, Cambridge. Hanson, P. E. & Gauld, I. D. (Eds) (1995) The Hymenoptera of Costa Rica. Oxford University Press, Oxford and The Natural History Museum, London. Müller, S., Stanyon, R., Finelli, P., Archidiaconon N. & Wienberg, J. (2000). Molecular cytogenetic dissection of human chromosomes 3 and 21 evolution. Proceedings of the National Academy of Science, USA 97: 206-211. Peters, R. H. (1991) A Critique for Ecologists. Cambridge University Press, Cambridge. Schauff, M. E. & LaSalle, J. (1998) The relevance of systematics to biological control: protecting the investment in research. pp. 425-436. In Zalucki, M. P., Drew, R. A. I. & White, G. G. (Eds) Pest Management – Future Challenges. Proceedings of the Sixth Australasian Applied Entomological Research Conference, Brisbane. Soltis, P. S., Soltis, D. E. & Chase, M. W. (1999) Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature 402: 402-404. Tautz, D., Friedrich, M. & Schröder, R. (1994) Insect embryogenesis - what is ancestral and what is derived? Development Supplement: 193-199. Whitfield, J. B. (1997) Molecular and morphological data suggest a single origin of the polydnaviruses among braconid wasps. Naturwissenschaften 84: 502-507.
This Page Intentionally Left Blank
Index to Authors Abrahamson, W. G. 218 Asgari, S. 29 Austin, A. D. 3, 90, 154, 178, 451 Babcock, C. 59 Baggen, L. R. 388 Balan, S. 421 Basibuyuk, H. H. 192 Baur, H. 165 Beck, M. 29, 39 Beggs. J. R. 404 Bennett, D. M. 371 Bhuiya, B. A. 417 Brady, S. G. 131 Broad, G. R. 74 Brown, G. R. 210 Campbell, B. 59 Carcreff, E. 114 Chan, K. 59 Chowdhury, S. H. 417 Ciomperlik, M. A. 347 Costa, R. A. C. 50 Cornuet, J. M. 114 Cruz-Landim, C. 50 Dolphin, K. 74 Dowton, M. 3, 90, 451 Falabella, P. 17 Fitton, M. G. 74, 192 Flook, P. K. 90 Flores, C. 258 Frampton, C. M. 396 Gadau, J. 131 Gaston, K. J. 303 Gauld, I. D. 258, 303 Generani, M. 231 Gokhman, V. E. 198 Goolsby, J. A. 347 Graziani, F. 17 Grissell, E. E. 267 Gurr, G. M. 388 Hanson, P. 258
Harrison, J. 437 Heraty, J. 59 Hewa-Kapuge, S. 371 Heydon, S. L. 320 Hildebrands, A. 359 Hobbs, K. R. 267 Hoffmann, A. A. 371 Huang, D.-W. 320 Iqbal, M. 178 Irvin, N. A. 396 Japoshvili, G. O. 339 Jennings, J. T. 154 Jesudasan, R. W. A. 421 Johnson, N. F. 313 Jones, W. A. 347 Junhua, H. 328 Kabir, S. M. H. 417 Kambhampati, S. 107 Keller, M. A. 247 Kinuthia, W. 23 Kirk, A A. 347 Legaspi Jr., B. C. 347 Legaspi, J. C. 347 Levick, N. R. 437 Li, D. 23 Malva, C. 17 Manley, D. G. 285 Matthews, R. W. 427 Mauss, V. 207 Meats, A. 388 Melika, G. 218 Melo, G. A. R. 281 Moraes, R. L. M. S. 50 Moritz, R. F. A. 84 Mourad, A. K. 46 Musetti, L. 313
Pennacchio, F. 17 Penteado-Dias, A. M. 296 Quicke, D. L. J. 74, 192 Rasnitsyn, A. P. 192 Rasplus, J.-Y. 59, 114 Roques, A. 114 Ruiz, R. A. 347 Sands, D. P. A. 410 Scaramozzino, P. L. 231, 290 Schmidt, J. O. 437 Schmidt, O. 23, 29, 38 Schmitz, J. 84 Scholz, B. C. G. 381 Smith, G. S. 437 Smith, P. T. 106 Sordetti, R. 17 Steffen-Campbell, J. 59 Strand, M. R. 11 Strumia, F. 290 Tenhumberg, B. 247 Theopold, U. 23, 29, 38 Thieme, T. 359 Ugalde, J. 258 Vacek, D. C. 347 Varricchio, P. 17 Vidal, S. 359 Vinson, S. B. 17, 46 Ward, P. S. 131 Wendel, L. E. 347 Wharton, R. W. 143 Whitfield, J. B. 97 Winkel, K. D. 437 Wratten, S. D. 396
Notton, D. G. 74
Xiao, H. 320 Xuexin, C. 328
O’Neill, M. A. 303
Yun, M. 328
Pagliano, G. 290
Zalucki, M. P. 381
457
Index to Hymenopteran Names Ablaxia 325 Ablerus 63-4, 66, 72 Absyrtus 76 Acampsis 330 Acanthochalcis 63, 67-8, 72 Acanthormius 330 Acerocephala 325 Aclitus 106 curipennis 106 Acordulecera 258 Acraspis 221, 223 gemula 223 Acroclisoides 325 Acrocormus 325 Adelognathus 77 Adleria 221 Aenasius 63, 72 advena 417-20 Aeolothynnus 212 Aepocerus 63, 65, 72 Agathis dimidiata 237 Agenia croceocera 239 ornatipennis 239 Agiommatus 322, 325 Agriomyia 215 Agrypon 76 Aleiodes 330 esenbeckii 333 narangae 333 Alexeter 76 Allocricellius 325 Allodapes cinea 243 rufiventris 243 Alloea 331 Alomya 77 Alysia manducator 200 Amblyharma 325 Ammeia 326 Ammophila incana 242 Amphibolips 221-2, 226 quercuscinerea 222 Amyosoma chinensis 333 Anabaeus 179, 182-3, 185 Anacallocleonymus 325 Anacharis 401 Anagyrus mangicola 420 psendococci 340
458
Ancistrocerus nigricornis 84 oviventris 84 Andricus 201, 218-9, 221, 223 caputmedusae 222 gallaetinctoriae 223 grossulariae 223 kashiwaphilus 201 kollari 223 lignicolus 223 lucidus 219 mukaigawae 201, 203 quercuscorticis 225 quercusfoliatus 222 quercusradicis 225 quercusramuli 224 seckendorffi 222 solitarius 219, 224 testaceipes 225 Anicetus communis 411 nyasicus 411 Anisopteromalus 325 calandrae 202-4 Anomalon flavitarsus 236 hollandiae 236 Anthemus funicularis 340 pini 340 Antron 221 Apanteles 329, 331 cypris 333 Apechthis 76 Aphelinoidea 63, 73 Aphelinus 221-2, 226 albipodus 63, 72 asychis 63, 72 varipes 63, 72, 412 Aphelonyx 221 Aphidius 13, 108, 199, 204, 310, 365 ervi 12-4 matricariae 200 rhopalosiphi 365 Aphycoides clavellatus 340 Aphytis melinus 63, 72 yanoensis 63, 72 Apis dorsata 84
mellifera 121, 50-4, 84, 437, 442 Apobaeus 179, 182-5, 190 Apolophus 76 Aporus marginatus 240 Apozyx 147 Aprostocetus 63 ceroplastae 411 Apsilocera 325 Aptesis puncticollis 200 Areopraon 111 Aridelus 331 Arthrolytus 325 Asaphes 63, 73, 325 Ascogaster 331 Atanycolus 147 Ateleute 76 Atrusca 218, 221, 226 bella 218 brevipennata 218 capronae 218 catena 218 Aulacocentrum 330 Aulacofoenus 154-5 deletangi 156, 163 fallax 156, 163 fletcheri 156, 163 infumatus 156, 163 kurmondi 156, 163 marionae 156, 163 perenjorii 156, 163 thoracicus 156, 163 whiani 156, 163 Aulacus 154-7 atriceps 156, 163 Aulax 219 Aulosaphes 330 Aulosaphoides 330 Austromerus 185 Baeus 178-9, 181-2, 185 leai 183, 190 machodoi 183, 190 saliens 183, 190 seminulum 183, 190 Banchus 76 Barilochia 259 Baryscapus ceroplastophilus 411 Bassettia 221, 226 pallida 224-5 Belesica 76
Hymenoptera: Evolution, Biodiversity and Biological Control
Index to Hymenopteran Names 459
Belizinella 220-1, 223, gibbera 223 vicina 223 Belonocnema 221, 226 quercusvirens 222 Belonogaster juncea colonialis 84 petiolata 84 Bembex adelaidae 242 Bembix furcata 242 Besbicus 211 Betylobracon 147 Binodoxys 108, 111 Biorhiza 221 Blastothrix hungarica 340 longipennis 339-40 nikolskajae 339-40 Blepyrus insularis 419-20 Bombus 437 Brachycyrtus 77 Brachymeria 62-3, 67-8, 72 intermedia 199 Brachyscleroma 76 Bracon argenteociliatus 237 capitator 237 dimidiatocinctus 237 hyalinipennis 237 greeni 333 isomera 333 mutator 237 nigrorufum 333 onukii 333 sericatum 237 tasmaniae 237 Brasema 63-4, 72 Bremiella 76 Buluka 331 Cales 62-4, 72 Callaspidia defonscolombei 195, 203 Callimerismus 325 Callirhytis 218, 221, 223, 226 bella 218, 223 cornigera 218 glandium 223 quercusclavigera 218 quercuspunctata 218 Callitula 325 Callocleonymus 325 Calosota 63-4, 72
Campoletis 19, 30-1 sonorensis 30-1, 33, 46 Camponotus 131-4, 136-7 atriceps 132-4, 136 balzani 133-4, 136 castaneus 135 consobrinus 135 clarithorax 133-4, 136 Colobopsis 131, 133-4, 136 Dinomyrmex 132-4, 136 essigi 133-4, 136 floridanus 133-4, 136 gigas 134 herculeanus 133-4, 136 hyatti 133-4, 136 Hypercolobopsis 131 impressus 134 laevigatus 133-4, 136 ligniperdus 133-4, 136 modoc 133-4, 136 melanocephalus 134 mus 134 Myrmaphaenus 133-4, 136 Myrmentoma 133-4, 136 Myrmepomis 132 Myrmobrachys 133-4, 136 Myrmothrix 133-4, 136 papago 134 pennsylvanicus 133-4, 136 planatus 134 Pseudocolobopsis 131 quercicola 133-4, 136 rufipes 132-4, 136 semitestaceus 135 sericeiventrus 134 silvicola 133-4, 136 socius 135 Tanaemyrmex 133-4, 136 vicinus 135 yogi 134, 136 Camptoprium 259 Capellia 325 Cardiochiles nigriceps 17-20, 46-48 Catocheilus 213 Catolaccus 325 Cecidostiba 325 Cenocoelius 330 Centistidea 330 Cephaleta 325 Ceramius 5 maroccanus-complex 207-9 m. maroccanus 207-8 m. rubripes 207-8 montanus 207-8 Cerapterocerus mirabilis 339-40
Ceratobaeus 178-80, 182, 184-5 cornutus 183, 190 cuspicornutus 183, 190 fasciatus 183, 190 flavipes 181, 190 giraulti 183, 190 intrudae 180, 183, 190 laeviventris 183, 190 leai 181-3, 190 mirabilis 183, 190 reiki 183, 190 setosus 180, 183, 190 Cerocephala 325 Ceroptres 220 Cerospastus 258 Charmon 330 Chartocerus 63-5, 68, 73 Cheiloneurus claviger 339-40 kollari 340 paralia 340 Cheiropachus 325 Chelonus 330 inanitus 204 munakatae 333 nr curvimaculatus 367 Chilaspis 221 Chlorocytus 325 Choreia maculata 340 Chrysis lyncea 238 marginata 290, 292-4 viridula 199 Chrysocharis 63, 72 Chrysomalla 63, 73 Chrysopophthorus 331 Cirrospilus 63, 64 Cladomacra 259, 264 terricola 259 Cleonymus 63, 73, 325 Clystopsenella 281 Coccobius 68 Coccophagoides 63-4, 68, 72 Coccophagus 64, 68 rusti 63, 72 scutellaris 63, 72 Coeloides 332 Coelopisthia 5, 165-7, 169-70, 172, 325 areolata 172, 176 bicarinata 172, 176 caledonica 172, 176 extenta 168, 172, 176 fumosipenni 172, 176 Kranophorus 165
Index to Hymenopteran Names 460
pachycera 165, 167-9, 172, 176 suborbicularis 172, 176 Colletes (Andrena) cicalybenla 242 halophilus 310 hederae 310 perforator 310 rubricollis 243 rufipes 243 rufiventris 243 succinctus 310 unicolor 243 vandiemenii 243 Collyria 76 Colotrechnus 63, 73, 325 Colpotrochia 76 Comperiella 63, 72 Conomorium 165, 169, 172, 325 amplum 168, 172, 176 patulum 172, 176 Conura 63, 67-8, 72, 325 Copidosoma 63, 72 floridanum 13, 15 koehleri 388, 390, 393-4 Coptera occidentalis 195 Coruna 325 Cosmoconus 76 Cosmophorus 331 Cotesia 31-2, 34-5, 99, 102 congregata 34, 98, 100, 101 erionotae 414 glomerata 47, 100-1, 333 kariyai 333 marginiventris 100-1 melitacarum 100-1 orobenae 100-1 plutellae 333 rubecula 30-32, 34, 47, 98-101, 250-6, 333 ruficrus 333 Crabro australasiae 242 Crassifoenus 154-5, 157 Crocisa emarginata 243 guttata 243 Crypteffigies 27 Cryptoprymna 325 Cyclogastrella 165, 169, 172, 326 clypealis 168, 172, 176 deplanata 172 flavius 172, 176 simplex 172, 176 Cynips 219, 221, 223 divisa 222 longiventris 224
Hymenoptera: Evolution, Biodiversity and Biological Control
quercusfolii 222, 224 rosae 219 Cyphacolus 178-9, 182, 185, 190 Cyrtogaster 326 Cyrtoptyx 326 Dasymutilla chiron 288 clytemnestra 286, 288 dilucida 288 foxi 288 gloriosa 286, 288 magnifica 286, 288 nocturna 288 occidentalis 286 satanas 286 Decameria 259, 264 rufiventris 259 similis 259 Dendromyrmex 131-3, 135-7 Deuteroeulophus 63-4, 72 Diacritus 76 Diaeretiella 108 Diamma bicolor 210, 238 Diastrophus 220, 223 Dibrachoides 172 dynastes 172-176 Dibrachys 166, 172, 326 affinis 172, 176 boarmiae 172, 176 braconidis 176 cf. braconidis 172 cavus 168, 172, 176 confusus 172, 176 pelos 172, 176 Dichrogaster 77 Diglochis 165, 167, 169-70, 172 occidentalis 172, 176 sylvicola 168, 172, 176 Dimorphothynnus 213 Dinarmus 326 Dinotiscus 326 Dinotoides 326 Dipara 325 Diplazon 76 Diplolepis 63, 203, 219 rosae 73, 198 Diratithynnus 214 Dirhinus 63, 72 Dirophanes 204 callopus 204 fulvitarsis 204 invisor 204 Discodes coccophagus 339-40 Disholcaspis 221 Distilirella 330
Diversinervis elegans 411 Dolichogenidae tasmanica 396-401 Dolichomitus 99 Dolichovespula 437 maculata 84 media 84 saxonica 84 sylvestris 84 Dominibythus 281, 283-4 strictus 281-2 Donquickeia 296-7 Doratithynnus 214 Drailea 326 Dros 221 Dryocosmus 221 Duartea 173 daphne 173, 176 Dusona 76 Dyscritobaeus 185 Dyscritulus 108 Echthrodesis 179, 182-3, 185, 190 Eciton anale 241 duponti 241 forficatum 241 nigridens 241 pallidens 241 posticum 241 rufum 241 Eirone 214 Elachertus 63-4, 72 Elasmus 64-5, 68, 72 Elidothynnus 215 Embidobia 179-80 metoligotomae 190 Encarsia 64, 68, 204, 347-9, 351, 355, 411 adrianae 355 aurantii 63, 72 formosa 63, 72, 352, 423-4 nr hispida 356 luteola 63, 72 lutea 63, 72, 356 parvella group 356 pergandiella 63, 72, 349-50 nr pergandiella 349-50, 356, 412 protransvena 63, 72, 199 nr strenua 355 transvena 349-50, 355-6 tricolor 202 Encarsiella 63-4, 72 Encopothynnus 211
Index to Hymenopteran Names 461
Encyrtus lecaniorum 339-40 Enicospilus 76 Enjijus 258 Entedon 63, 72 Epactiothynnus 211 Ephedrus 108 Ephialtes 76 Epiclerus 63, 72 Epistenia 63, 72 Epitetracnemus zetterstedtii 341 Eremotylus 76 Eretmocerus 63-4, 66, 68, 72, 347-50, 358, 412, 421-4 corni 422 emiratus 349, 351, 358 eremicus 358 nr furuhashii 358 hayati 349-52, 358 melanoscutus 356-8 mundus 349-52, 356 staufferi 358 tejanus 349-51, 358 Erythres 221 Euaphidius 108 Euceros 72 Eufoenus 154-5, 157 antennalis 156, 163 australis 156, 163 crassitarsi 156, 164 darwini 156, 164 extraneus 156, 164 ferrugineus 156, 164 floricolus 156, 164 inaequalis 156, 164 minimus 156, 164 patellatus 156, 164 pilosus 156, 164 rieki 156, 164 ritae 156, 164 spinitarsis 156, 164 Eulonchetron 326 Eumacepolus 326 Eumayria 221, 226 Eumayriella 220-1, 226 Eumenes 84 campaniformis 240 latreillei 240 Euneura 326 Eunotus 63-4, 73, 325 Euplemus 63-4, 72 Eurydinota 326 Eurydinotomorpha 326 Eurytoma 62-3, 72 Eusandalum 62-4, 72 Eustenogaster calyptodoma 84 Eutrichosoma 63-4, 73
Euxystoteras 220-1 Exeirus lateritus 242 Exetastes 76 Exochus 76 Falciconus 108, 110, 111 Fioriella 221 Foenus gigas 236 Formica 132-3, 135 ammon 241 argentata 241 australis 241 carinata 241 fusca 132, 135 hastata 241 herculeana 241 intricans 241 metalliceps 242 moki 135 truncorum 132, 135 Ganodes 76 Gasteruption 154-5, 157 brachyurum 156, 163 fluviale 156, 163 spinigerum 156, 163 paradoxale 156, 163 Gastracanthus 326 pulcherrimus 195 Gastrancistrus 327 Glyphognathus 321, 326 Glyphomerus 63-4, 73 Glyptapanteles porthetriae 99 Glyptosticha 326 Gonaspis 223 Gonatocerus 63, 73 Grahamisia 322, 325 Gryon 63, 73, 179-80, 183-4, 190 Grypocentrus 76 Guerinius 214 Gugolzia 326 Gyrochus 332 Habritys 326 Habrobracon hebetor 12-4, 333 Hadzhibeylia physococci 341 Halictus distinguendum 243 nigritarsus 243 orbitus 243 Halticoptera 326 Halticopterina 326 Haplostegus 258 Hartemita 332 Hellwigia 76
Hemadas 166 Henryana 63, 72 Hensonia 291 Heratemis 331 Herbertia 326 Heteroecus 218, 221 Heteroperreyia 258-9 jorgenseni 259 Heydenia 325 Hickmanella 178-80, 182, 185, 190 holoplatyse 183, 190 intrudens 183, 190 Hobbya 326 Hockeria 63, 68 Holcaeus 326 Holocynips 221, 226 badia 226 hartmanni 226 maxima 226 Homalotylus quaylei 341 Homolobus 330 Homoporus 326 Horismenus 63, 72 Hybrizon 76, 81 Hyperacmus 76 Hypsicera 76 Hyptiogaster 154-5, 157 humeralis 156, 164 kalbarii 156, 164 arenicola 156, 164 pinjarregaensis 156, 164 rufus 156, 164 Ichneumon australis 236 ischioleucos 236 Ichneutes 330 Idioporus 63-5, 73 Idris 178-80, 182, 184-6 flavicornis 182-3, 190 helpidid 180, 183, 190 niger 183, 190 pulcher 182-3, 190 seminitidus 180, 183, 190 theridii 180, 183, 190 Incalia 258 Inkaka 326 Ischyrocnemis 76 Ischyroptyx 326 Isocyrtus 326 Isodontia mexicana 290-4 Isoptronotum 332 Iswaroides 211 Itoplectis 76
Index to Hymenopteran Names 462
Janssoniella 167, 173, 176 Kaleva 326 Kranophorus 165 Labena 77 Lamprotatus 321, 326 Lapton 76 Lariophagus 326 Larra autralasiae 242 Lathrostizus 76 Leptopilina 63 boulardi 73 Lestricothynnus 213-14 leucospis 63, 73 Libanobythus 281 Liodora 221 Liostenogaster flavolineata 84 Lipolexis 108, 111 Lissonota 76, 84 Loxaulus 221, 226 huberi 224, 225 masneri 222 Lysiphlebus 108, 359, 365 fabarum 359-64, 366-7 testaceipes 359-67 Lyubana 323, 327 Macrocentrus 330 cingulum 333 linearis 333 Macroglenes 327 Macromesus 326 Mahencyrtus coccidiphagus 341 Makaronesa 326 Manineura 327 Megachile chrysura 243 Megalhira fasciipennis 236 Megalyra fasciipennis 195 Megarhyssa 76 Megastigmus 5, 62-3, 73, 114, 123, 125, 127-8, 267-73, 276 aculeatus 128 brevicaudis 128 pictus 128 pinsapinis 128 pistaciae 128, 267, 269, 272, 277-8 rhusi 267-8, 272-6, 278 rosae 128 suspectus 128 thomseni 272 transvaalensis 267-8, 271-6, 278 wachtli 114-8, 120, 122-3, 125, 128
Hymenoptera: Evolution, Biodiversity and Biological Control
Melanodolius 76 Melipona quadrifasciata anthidioides 51-2 Melittobia 6, 61, 63, 72, 428-30, 432 australica 428 digitata 427-8 Meraporus 327 Merismomorpha 327 Merismus 326 Merisus 327 Mesocentrkus 148 Mesochorus 76 Mesopolobus 166, 327 adrianae 166 Mesostenus luperus 237 Mesostoa 147 Metablastothrix truncatipennis 339, 341 Metacolus 327 Metaparagia maculata 84 Metaphycus 63, 72 asterolecanii 341 insidiosus 341 zebratus 341 Metastenus 327 Meteorus 331 rubens 333 Meximalus 166 Microdelus 327 Miscogaster 321, 326 Microgasteriella 325 Microleptes 76-7 Microplitis 332 demolitor 99 erythrogaster 332 mediator 332 tuberculifer 332 Microterys clauseni 341 duplicatus 341 ferrugineus 341 hortulanus 339, 341 sylvius 339, 341 tricoloricornis 341 trjapitzini 341 Mirax 330 Mirobaeoides 178-9, 182, 185 barbarae 183, 185, 190 pecki 181-3, 190 scutellaris 183, 190 tasmanicus 183, 190 Mirobaeus 179, 182, 185 bicolor 183, 190 Mokrzeckia 325 Monoblastus 76
Monoctonus 76 paulensis 365 Monodiscodes itermedius 341 Monodontomerus 199 Moranila 325 Muscidifurax 63, 73 Mutilla dorsigera 238 rugicollis 238 tricarinata 238 Nasonia 61-2, 65, 202, 325 giraulti 61, 63, 73, 202 longicornis 61, 73, 202 longpetiolata 63 vitripennis 61, 63, 73, 87, 202-4, 419 Nematopodius 77 Neobaeus 179, 182, 185 novazealandensis 183, 190 Neodipara 323, 326 Neoneuroterus 220-1 Neorhacodes 76 Neotheronia ambramsae 305 lineata 305, 307 mellosa 305, 307 Neoxorides 76 Netelia 76-7 Netomocera 321, 325 Neuroterus 220-1 laeviusculus 224 numismalis 223-4 petioliventris 223 quercusbaccarum 222-4 tricolor 222 Nixonia 179-80, 190 Nodisoplata 326 Nonnus 76 Norbanus 325 Notanisus 325 Notoglyptus 325 Obeza 63, 72 Odontacolus 178-9, 182-3, 185 longiceps 183, 190 Odontocyips 221 Odynerus albifrons 240 atripes 240 clotho 240 lepidus 241 swanii 241 tamarinus 241 Oecophylla 132 longinoda 135-6
Index to Hymenopteran Names 463
Oedemopsis 76 Oligosita 63, 72 Oncorhinothynnus 214 Ontsira palliatus 332 Oodera 325 Ooencyrtus erionotae 63, 72 nezarae 413 papilionis 413 trinidadensis 413 Ophelosia 326 Ophion 76 australasiae 237 bicallosus 237 dorsatus 237 merdarius 237 Orasema 72 Ormocerus 327 Orthocentrus 76 Orthopelma 76-7 Oxyglypta 327 Oxysychus 325 Pachycrepoideus 63-4, 73, 325 Pachyneuron 167, 173, 325 formosum 173, 176 muscarum 173, 176 Panicus australasiae 237 difficilis 237 testaceus 237 Panstenon 327 Panteles 76 Parabaeus 185 Paracarotomus 325 Paracraspis 221 Paracroclisis 325 Paragia shuckarti 240 Paramonoctonus 108, 110-1 Parandricus 221 mairei 223 Parischnogaster alternata 84 jacobsoni 84 mellyi 84 Paroxyharma 325 Parurios 322, 325 Patrocloides 200 Pauesia 108 Pentatermus 330 Perga dorsalis 236 ferruginea 236 lewisii 236 Periclistus brandtii 223 Peridesmia 325
Perilampus 63-4, 73 fulvicornis 68 Perilissus 76 Perilitis 108 Perithous scurra 199 Perreyia 259-60, 263-4 tropica 259-64 Perreyiella 259 Pezilepsis 325 Phaenolobus 76 Philonix 221 Phrudus 76-7 Phylloteras 220-1 Phymatothynnus 211, 214 nr monilicornis 212 Physoscelus australasiae 242 Pimpla 305-6 crenator 237 croceiventris 305-6 hypochondriaca 306 subpetiolata 237 sumichrasti 305 Pion 76-7 Pison spinolae 242 Plagiotrochus 221, 226 Platecrizotes 325 Platneptis 325 Platygastiodes 185 Platygerrhus 325 Platyspathius 329, 331 Plutothrix 167, 173, 325 bicolorata 173, 176 obtusiclava 173, 176 Podagrion 62-4, 73 Poecilocryptus 77 Poemenia 76 Pogonomyrmex 288 Polistes 437, 439 dominulus 84 flavus 84 leti 240 saggittarius 84 Polyblastus 76 Polyrema 63, 73 Polyrhachis 131-3, 135-7 flavibasis 135-6 hostilis 135-6 dives 135-6 Polysphincta tuberosa 200 Polystes faciali 240 flaveola 240 Pompilus 240 australasiae 240 morio 240
sericeocinctus 240 xanthocerus 240 Praon 108 pequodorum 365 Priocnemis 240 hollandiae 240 ruficeps 240 Prionomitus mitratus 341 Pristaulacus 154-5, 157 cinguiculatus 156, 163 variegatus 156, 163 Pristomerus 76 Probaryconus 185 Proclitus 76 Propicroscytus 325 Prosceliphron 291 Prosopis alcyone 243 Provespa anomala 84 nocturna 84 Psamattra chalybea 238 Pselaphanus 146, 149 Pseudaphycus phenacocci 341 Pseudichneutes 330 Pseudoamblyteles 200 Pseudocatolaccus 325 Pseudofoenus 154-5, 157 crassipes 156, 164 unguiculatus 156, 164 Pseudomasaris maculifrons 84 Pseudorhyssa 76 Psilocera 325 Psilochalcis 66, 72 Psilocharis 63, 68 Psychophagus 326 Psyllaephagus bachardenicus 342 georgicus 339, 342 nr rubriscutellatus 342 tokgaevi 342 Pteromalus 326 Pterosemigastra 326 Pterosemopsis 326 Pterygophorus cinctus 233 interruptus 233 Ptinocida 166 Pycnetron 326 Rakosina 326 Rattana 330 Rectizele 330 Repentinia 220-1 Rhagigaster 213-215 auriceps 215
Index to Hymenopteran Names 464
castaneus 215 stradbrokensis 212 Rhaphitelus 326 Rhicnocoelia 167, 173, 326 constans 173, 176 Rhodites 219-20 Rhopalicus 165, 167, 173, 326 tutela 173, 176 Rhygchium ephippium 241 Rhyssa maculipennis 237 Rileya 63, 72 Rodrigama 76 Ropalidia 439 Roptrocerus 326 Scaptotrigona postica 51-4 Sceliphron caementarium 290-1, 294 curvatum 290-2, 294 destillatorium 290 spirifex 290 Schizonotus 165, 173, 326 sieboldi 173, 176 Schizopyga 76 Schlettererius cinctipes 155-6, 163 Schwarzella 63, 68, 72 Scolebythus 281 madecassus 281 Scolia cyanipennis 238 glabrata 238 javana 238 soror 233 verticalis 238 Scutellista 63-4, 73, 326 caerulea 411 Seladerma 326 Semiotellus 327 Semirhytus 296 Seres 65, 72 Sigalphus 330 Signiphora 64 Siniphanerotomella 331 Sinoneoneurus 330 Sirex 232, 410 Skeloceras 326 Skiapus 76 Solenura 325 Spalangia 62-3, 73, 327 Sparasion 179-80, 190 Spathegaster 220 Spathius 329, 331
Hymenoptera: Evolution, Biodiversity and Biological Control
Sphaeripalpus 326 Sphaeroteras 221 ocala 222-4 Sphecophaga vesparum 406 vesparum vesparum 406 Sphegigaster 326 Sphex distincta 242 pubiventris 242 rufipennis 242 Stenomalina 326 Stichocrepis 173 armata 173, 176 Stictomischus 321, 326 Stilbops 76-7 Storeya 327 Streblocera 331 Sycophila 63, 72 Synedrus 326 Synergus 220 umbraculus 223 Syntomopus 326 Systasis 327 Systellogaster 173 gahani 173, 176 ovivora 173, 176 Syzygonia 258 Tachynomyia 211 Tanaostigmodes 63, 73 Telenomus 63, 384 Tequus 258 Teras 220 Thektogaster 326 Theocolax 325 Thinodytes 326 Thynnoides 213 Thynnus 213-5 annulatus 239 depressus 239 femoralis 239 festivus 239 flavomaculatus 239 gravidus 239 haemorrhoidalis 239 melleus 239 obscuripennis 239 octomaculatus 239 picipes 239 ramburi 239 rubripes 239 senilis 239
trifidus 239 variabilis 239 villosus 239 xanthognathus 239 Timulla 286 Tomicobia 325 Torymus 63-4, 73 Toxeuma 326 Toxeumorpha 326 Trichagalma 220 Trichogramma 6, 61, 371-2, 378, 381-6, 398, 400-1, 427 nr brassicae 371-8 carverae 370-8 Trichogrammanza carverae 371 fuentesi 71 nr ivelae 371 platneri 63, 73 pretiosum 63, 73, 371, 378, 381-6 Trichomalopsis 61, 63, 73, 326 Trichomalus 326 Trichomasthus albimanus 342 cyaneus 342 ivericus 342 Trichoteras 221 Triclistus podagricus 200 Tricyclomischus 326 Trigona 438 Trigonaspis 220-1, 223 megaptera 223 Trigonoderoides 326 Trigonoderus 326 Trimorus 185 Trioxys 108, 111 indicus 367 Trisoleniella 221 Trissolcus 63, 73 basalis 412 Tritneptis 173, 326 doris 173, 176 hemerocampae 173, 176 ? klugii 173 Tromatobia 76 Trychnosoma 326 Tsela 326 Tumor 326 Uniclypea 326 Uscana 61, 63 Ussuraspis 220-1
Index to Hymenopteran Names 465
Venturia 33-5, 44, 76 canescens 23, 25-7, 31, 33, 38-41, 43-4, 256 Vespa australis 240 crabro 84, 437 orientalis 84 pennsylvanica 437 Vespula 404, 437 germanica 84, 404, 437-9 rufa 84 vulgaris 6, 84, 404, 439 Vipio 147 Vrestovia 326
Wesmaelia 331 Westwoodia ruficeps 237 Xanthoteras 220-1 Xenoschesis 76 Xestomnaster 326 Xestophanes 223 Xiphozele 330 Xorides 77 Xyalophora 63, 73 Xylocopa aestuans 243 muscaria 243 virginica 84, 87
Xystoteras 220-1, 223 Ycaploca 281-2 evansi 281 Yelicones 147, 330 Zaomma lambinus 342 Zaspilothynnus 213-4 interruptus 212 Zdenekiana 326 Zele 331 Zolotarewskya 323, 325 Zombrus 145, 331 Zopheroteras 221
Index to Other Animal Names Acanthococcus aceris 341-2 Acheta domesticus 92 Acyrthosiphon pisum 365 Adiscodiaspis tamaricicola 342 Agatheromera crassa 92 Agrotis 333 Aleurocanthus woglumi 352 Aleurothrixus floccosus 352 Anoplophora chinensis 352 Aphis carccivora 367 pisum 367 rumicis 359-64, 366-7 Aphis fabae-complex 359-67 armata 359-64, 367 cirsiiacanthoidis 359-64, 366-7 evonymi 359-64, 367 fabae 359-64, 366-7 mordwilkowi 359-64, 366-7 solanella 359-64, 366-7 Arsenophonus 128 Asterodiaspis quercicola 341 Bemisia 348, 352 argentifolii 347 tabaci 6, 347-8, 351, 355-8, 412, 454 Cadra cautella 333 Caenorhabditis 11 Callidium villosulum 332 Cephrenes augiades 414 Ceroplastes 411 destructor 411-2 japonicus 340-1 Chilo suppressalis 332 Choeradodis rhombicollis 92 Cnaphalocrocis medinalis 332-3 Coccus hesperidum 340-1 Crastina tamaricina 342 Culicoides 306, 308-9 Culex pipiens-complex molestus 310 pipiens 31 Cylindraustralia kochii 93-4 Dendrolimus 332-3 punctatus 328 Dipsosaurus dorsalis 287 Diuraphis noxia 413
466
Drosophila 4, 11-3, 15, 27, 67, 428, 453 yakuba 132 melanogaster 11, 25, 29 Eleodes 288 Eneopterus 92 Ephestia 27, 33 kuehniella 31, 33, 39 Epiphyas postvittana 371-3, 377, 398 Erionota thrax 414 Eriopeltis festucae 340, 342 Eublemma amabilis 333 Ferrisia virgata 417-420 Forcipomyia 306, 308-9 Galleria mellonella 25 Geococcyx californianus 288 Gromphadorhina portentosa 92 Grylloblatta rothi 92 Gryllotalpa gryllotalpa 92 Gryllus campestris 92 Hadrurus arizonensis 286-7 Helicoverpa armigera 332, 381-3 punctigera 371-2 Heliothis 30 virescens 17-20, 30-1, 47 Helix pomatia 25-27 Hemideina crassidens 93 Kermes roboris 341 Lepidosaphes ulmi 341 Leucania separata 332-3 Leucaspis loewi 340 pusilla 340 Limulus 23 Locusta 91 migratoria 91, 93 Luzulaspis luzulae 342 Lymantira dispar 332 Macrosiphon creelii 365 Manduca sexta 12, 20 Megacrania apheus 92 Micromus tasmaniae 401 Monochamus alternatus 332
Musca domestica 12, 419 Mustela erminea 405 Mycalesis gotama 332 Naranga aenescens 332-3 Nestor meridionalis meridionalis 405 Nezara viridula 412-3 Oncopeltus fasciatus 287 Onstrinia furnacalis 333 Onychomys torridus 286 Oulema aryzae 332 Parnara 332 guttata 328 Parthenolecanium corni 340-1 persicae 340 rufulum 340 Pectinophora gossypiella 332-3 Peiris 31-2 rapae 31-2, 250, 252-5, 328, 333 Penalva 93 Phenacoccus mespili 341 Phrysonoma cornutum 288 Phthorimaea operculella 388, 390-4, 401 Physokermes hemicriphus 340 Planchonia arabidis 340-2 Planococcus ficus 340-1 pacificus 417-20 Plodia interpunctella 333 Plutella xylostella 333 Protonemura meyeri 93 Pseudaulacaspis pentagona 411 Psylla crataegi 341 ramnicola 342 Pulvinaria populi 341 Rhizopulvinaria armenica 341 Rhodococcus spiraeae 340-1 Rhopalosiphum padi 413 Rickettsia 128 Ritsemia pupifera 341 Saperda populnea 332 Sarcophaga bullata 429
Hymenoptera: Evolution, Biodiversity and Biological Control
Index to Other Animal Names 467
Sceloporus magister 286-8 Scirpophaga incertulas 328, 332 Semanelus sinoauster 332 Sesamia inferens 332 Sitobion avenae 365 Sitotroga cereatella 333, 372-3, 377 Sphaerolecanium prunastri 339-41 Stephomyia rotundifoliorum 296
Stromaltum lingicorna 332
Trioza magnisetosa 342
Toxostoma 288 Trialeurodes 356 abutilonea 352 packardi 422 ricini 421-4 vaporariorum 355, 422-3 Tribolium 12 Trichoplusia ni 367
Ultracoelostoma 404 Vibrissina 263 Wolbachia 61, 128 Xenopus 11 Xystrocera globosa 332
Index to Plant and Micro-organism Names Acacia 340, 342 Acer 341 Achras zapota 421 Annona glabra 421 Arbutus 421 Arrenaterium elatius 340 Auricularia 259 Bacillus thuringiensis 385 Baculovirus heliothis 385 Beauveria bassiana 407 Borago officinalis 388, 401 Brassica oleracea 349-50 Callistemon 258 Cecropia 281 Cercis siliquastrum 340-1 Chromolaena odorata 357 Cirsium arvense 359 Coffea arabica 263 Coriandrum sativum 401 Corylus 340 Crataegus 340-1 Cucumis melo 349 Cupressus ambramsiana 114 arizonica 114 atlantica 114-5, 123, 125, 127 bakeri 114 dupreziana 125 goveniana 114, 123 sempervirens 114-5, 117, 123, 125, 127 Digitalis purpurea 359 Diospyros 340 Elaeagnus angustifolia 342 Elaeocarpus 258 Emex 258 Erythrina poeppigiana 263 Eucalyptus 258 Eugenia rotundifolia 296, 298 Euonymus europaeus 359 Euphorbia 357 Fagopyrum esculentum 388, 397 Festucae 340, 342 Ficus carica 259, 340 Fraxinus 340-1
468
Gossypium hirsutum 349, 421 Gramineae 342 Hedera 340-1 Herniaria 341 Hibiscus rosasinensis 348 Ilex 340-1 Ipomoea 356-7 batata 421 Larrea tridentata 286 Laurus nobilis 341 Leptospermum 258 Lobularia maritima 401 Lolium 340 Malus 340 Mangifer indica 263 Marsilea 258 Melaleuca 258 Melanthesa rhamnoides 421 Mikania 296 Murraya koenigii 421 Mussaenda 355 Nothofagus 258 Phacelia tanacetifolia 388, 401 Phyllanthus acidus 421-3 Picea 340 Pinus 340 Piper nigrum 270 Pistacia 267-9, 271, 277-8 atlantica 269, 271 chinensis 269, 271 integerrima 271 lentiscus 268 terebinthus 268 vera 269, 271 Populus gracilis 341 transcaucasica 341 tremula 341 Prunus 340-1 divaricata 341 Psidium 258 friedrichsthalianum 263 guayava 417
Quercus 218, 341 chrysolepis 218 petraea 219 pubescens 219 robur 219 Rhamnus pallasii 342 Rhus 267-71, 273-6, 278 angustifolia 268, 274 chirindensis 268 coriaria 270 glabra 271 laevigata 268, 274-5 lancea 268, 271, 276 pendulina 268 rehmanniana glabrata 268 succedanea 271 typhina 271 vernicifera 271 viminalis 268 Ricinus communis 421-3 Rosa 421 Rubus 258 Rumex 258 crispus 360, 362 obtusifolius 359 Schinus 258, 268-70, 272-6, 278 polygamus 270 molle 267-70 terebinthifolius 268-70, 274-5 Solanum 258 nigrum 359 Sonchus 357 Spiraea 340-1 Tamarix 342 Thelicranium australis 340 Tibouchina 258 Toxicodendron vernicifluum 271 Tristania 258 Tropaeoleum majus 359, 388 Ulmus foliacea 340-1 Vicia faba 359, 397 Vitis 258
Hymenoptera: Evolution, Biodiversity and Biological Control