>) Ka KNIE VERA a a Le n U a 0 1 ’ var TUE KEN sr Dr ah, nr - Br ‘ ne Nenn an “ N ni BUNDES VEIT TERN eher an 3 f Ai Ye Ama BAGS Ru wine I onen. h 5 es, ; 10 Kaly EL E Anl am £ Urne zn, R a arena“ de 2 ee A En gaun Du ul Ysard HAIR EDER CH \ ya i \ Da aan aan araen Fa uudhung, Mirens e Bultanlurhgt u R elncin. $ vru re anıamun, a a Ylralıc, een a ld, ER Le beta, Pe Det [6 BDA NEN hie, ten, NE Mn uritg an, nn rent et beige neu ante he N erahnen ag Ver a, arte a hr, Enger run a mn. sem er ; pr; Learn a NE ren Pr RER ESALNREN FIN Bay Pa NN EREN! oasn Far - . ana arg HT Te > Fe We Es te Fi i er r en 2 c UPFLIEL FETTE 4 rd I TE EUER FAGR TE DE ESIRE Nie r Bi Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Säugetierkunde (German Society of Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel R. Hutterer, Bonn E. Kalko, Ulm P. Lüps, Bern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Volume 66 2001 URBAN & FISCHER II Contents volume 66 Reviews Hewison, A. J. M.; Danilkin, A.: Evidence for separate specific status of European (Capreolus capreolus) and Siberian (C. pygargus) roe deer — Beweise für den Artstatus von Europäischem (Capreolus capreols und. Sibiuschemi(€pygargus))RehwildE27 were Smith, M. H.; Novak, J. M.; Peles, J. D.; Purdue, J. R.: Genetic heterogeneity of white-tailed deer: man- agement lessons from a long-term study — Genetische Heterogenität beim Weißwedelhirsch: Für die Wildbewirtschaftung relevante Erkenntnisse aus einer Langzeitstudie. ........... 2.2222 2220.. Original investigations Alves, P. C.; Ferrand, N.; Suchentrunk, F: Developmental stability and protein heterozygosity in a local population of Iberian hares (Lepus granatensis) — Heterozygotie und Entwicklungshomöostase bei Iberischen Hasen (Lepus granatensis) aus einer lokalen Population in Portugal................. Caro, T. M.; Brock, R.; Kelly, M.: Diversity of mammals in the Bladen Nature Reserve, Belize, and fac- tors affecting their trapping success — Diversität von Säugern im Bladen Naturreservat, Belize, und Faktoren, die einen Fangerfole’beeinflussen”. .... 2.202. ee Conroy, C. J.;, Hortelano, Y.; Cervantes, F. A.; Cook, J. A.: The phylogenetic position of southern relictual species of Microtus (Muridae: Rodentia) in North America —- Die phylogenetische Stellung südlicher Reliktarten von Microtus. (Muridae: Rodentia) in Nordamerika... 2.2.22... 2 E de Moraes, D. A.; Lemos, B.; Cerqueira, R.: Supernumerary molars in neotropical opossums (Didelphi- morphia, Didelphidae) — UÜberzählige Molaren bei neotropischen Beuteltieren (Didelphimorphia, Didelphidae) ....2% ».. 0. a. Sera. ar a ee ee N Del Valle, J. C.; Lohfelt, M. I.; Comparatore, V. M.; Cid, M. S.; Busch, C.: Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco) — Auswahl und Bevorzugung von Nahrung bei Crenomys talarum (Tuco-Tuco) :: :.....2.22 8022 22.000 re ee Dubost, G.: Comparison of the social behaviour of captive sympatric peccary species (genus Tayassu); corre- lations with their ecological characteristics — Vergleiche des Sozialverhaltens zweier sympatrischer Pe- kariarten (Genus Tayassu) in Menschenobhut; Beziehungen mit ihren ökologischen Merkmalen ..... Glitzner, I.; Gossow, H.: Kleinsäuger auf forstwirtschaftlich unterschiedlich behandelten Windwurfflä- chen eines Bergwaldes — Small mammal distribution on differently managed storm areas of a moun- tain forest. . „u. mel ee ee 2 US Pe Gloor, S.; Bontadina, F.; Hegglin, D.; Deplazes, P.; Breitenmoser, U.: The rise of urban fox populations in Switzerland — Die Entstehung urbaner Fuchspopulationen in der Schweiz... ..........2222202.. Good, T. C.; Hindenlang, K.; Imfeld, S.; Nievergelt, B.: A habitat analysis of badger (Meles meles L.) setts in a semi-natural forest - Eine Analyse der Habitatcharakteristika von Dachsbauten (Meles meles L.) in.einem naturnahen Wald... .. u... ... eu. sauce ec ee en Goodman, S. M.; Hutterer, R.; Ngnegueu, P. R.: A report on the community of shrews (Mammalia: Sorici- dae) occurring in the Minkebe Forest, northeastern Gabon - Bericht über die Artengemeinschaft von Spitzmäusen (Mammalia: Soricidae) im Minkebe Regenwald, Nord-Ost Gabun ................ Heide-Jorgensen, M. P.; Lockyer, Ch.: Age and sex distribution in the catches of belugas, Delphinapterus leucas, in West Greenland and in western Russia — Alters- und Geschlechtsverteilung von Belugafän- gen, Delphinapterus leucas, in Westgrönland und Westrußland ..............2.2.2seeeeeeenn Holley, A. J. F: The daily activity period of the brown hare (Lepus europaeus) — Die tägliche Aktivitäts- periode. des Europäischen Feldhasen(Zepusieuropaeus). 2.2.2.2. 2.2.2. 22 202 Holzhaider, J.;, Zahn, A.: Bats in the Bavarian Alps: species composition and utilization of higher alti- tudes in summer — Fledermäuse in den Bayerischen Alpen: Artenspektrum und Nutzung von höheren Lagen.im’Sommer 2. aa: Ha. er a Bee SE Kittlein, M. J.; Vassallo, A. 1.; Busch, C.: Differential predation upon sex and age classes of tuco-tucos (Ctenomys talarum, Rodentia: Octodontidae) by owls - Alters- und geschlechtsbedingte Unterschiede bei Prädation von Tuco-Tucos (Ctenomys talarım, Rodentia: Octodontidae) durch Eulen.......... 13 238 90 332 193 165 65 290 155 204 22 215 397 144 Contents volume 66 Krystufek, B.; Spitzenberger, F.; Kefelioglu, H.: Description, taxonomy, and distribution of Talpa davidi- ana — Beschreibung, Taxonomie und Verbreitung von Talpa davidiana ............222222c220.. Nordvig, K.; Reddersen, J.; Jensen, T. S.: Small mammal exploitation of upper vegeation strata in non-for- est, mixed farmland habitats - Ausnutzung höherer Vegetationsstrata durch Kleinsäuger in gemisch- ten®Xckerland-Habitaten 8.8 4... a 2: 2.2 rn ie ses. Nupp, T. E.; Swihart, R. K.: Assessing competition between forest rodents in a fragmented landscape of midwestern USA - Abschätzungen zur Konkurrenz von im Wald lebenden Rodentiaspecies in einer kracmentertenKandschateimsmittleren\Westendell USA... 0... 0... .uunnnouesceun. Ochoa G., J.; Aguilera, M.; Pacheco, V.; Soriano, P. J.: A new species of Aepeomys Thomas, 1898 (Roden- tia: Muridae) from the Andes of Venezuela - Eine neue Art von Aepeomys Thomas, 1898 (Rodentia: NMunidao)lausden»Anden’vonVenezuelatmrn2. 1. Di EN rss Reutter, B. A.; Brünner, H.; Vogel, P.: Biochemical identification of three sympatric Apodemus species by protein electrophoresis of blood samples -— Biochemische Bestimmung dreier sympatrisch vorkom- mender Apodemus-Arten mittels Elektrophorese von Blutproteinen ...................220200. Röhrs, M.; Ebinger, P.: Welche quantitativen Beziehungen bestehen bei Säugetieren zwischen Schädelka- pazität und Hirnvolumen? — How is cranial capacity related to brain volume in mammals?........ Scandura, M.; Apollonio, M.; Mattioli, L.: Recent recovery of the Italian wolf population: a genetic inves- tigation using microsatellites - Neues Anwachsen der italienischen Wolfspopulation: eine genetische löintensuchungsmittels»Mikrosatelliten. .. . ze See Sana an saeasne nen een Steffen, K.; Kruska, D.; Tiedemann, R.: Postnatal brain size decrease, visual performance, learning, and discrimination ability of juvenile and adult American mink (Mustela vison: Carnivora: Mammalia) — Postnatale Hirngrößenabnahme, visuelle Leistung, Lernen und Unterscheidungsvermögen von juveni- len und adulten amerikanischen Minken (Mustela vison: Carnivora: Mammalia)................ Suchentrunk, F.; Jaschke, C.; Haiden, A.: Little allozyme and mtDNA variability in brown hares (Lepus europaeus) from New Zealand and Britain — A legacy of bottlenecks? — Geringe Allozym- und mtDNA-Variabilität bei neuseeländischen und britischen Feldhasen (Zepus europaeus) — eine Folge Ulevicius, A.; Sidorovich, V.; Lauzhel, G.: Specificity of non-metric parameters of American mink (Mus- tela vison) populations in relation to habitat differences in Belarus - Spezifität nicht-metrischer Para- meter von Mink-Populationen (Mustela vison) im Verhältnis zu Habitat-Unterschieden in \WVerbrußland; oe. hun role SA Be EN ren Be Ru BR De Wöhrmann-Repenning, A.; Bergmann, M.: The vomeronasal complex in strepsirhine primates and Tar- sius -— Der Vomeronasalkomplex bei strepsirhinen Primates und Tarsius ......... 222222. 2200. Short communications Bortoluzzi, A.; Gutierrez, M.; Baldo, J.; Gardenal, C. N.: Protein polymorphism in two species of Cre- nomys (Rodentia, Ctenomyidae) from Cördoba province, Argentina — Proteinpolymorphismus bei zwei Ctenomys-Arten (Rodentia, Ctenomyidae) aus der Provinz Cördoba, Argentinien .......... Dunnum, J. L.; Salazar-Bravo, J.; Yates, T. L.: The Bolivian bamboo rat, Dactylomys boliviensis (Roden- tia: Echimyidae), a new record for chromosome number in a mammal - Die bolivianische Bambus- ratte, Dactylomys boliviensis (Rodentia: Echimyidae), ein neuer Rekord für die Chromosomenzahl bei einem Saumeller u werds ee RE a Endepols, S.; Dietze, H.; Endepols, H.: The occurrence of roof rats (Rattus rattus L., 1758) in Germany during the late 20 century - Vorkommen der Hausratte (Rattus rattus L., 1758) in Deutschland am Enderdes@ 02 Jahchundentsag au: re a ee ee Eee ee ae WEM 00060000 000.00000000000000680860000002000000000000000000000000000000000000 Hice, Ch. L.: Records of a few rare mammals from northeastern Peru — Nachweis von einigen seltenen SangetierenfausdemnerdostlichemgBerunas 2. man. ee ee ee III 135 129 345 228 84 102 32 269 48 35 257 308 al 301 60 IV Contents volume 66 Koh, H. S.; Randi, E.: Genetic distinction of roe deer (Capreolus pygargus Pallas) sampled in Korea — Genetische Eigenständigkeit von Rehen (Capreolus pygargus Pallas) aus Korea................ Syali Martinkovä, N.; Ziak, D.; Kocian, L’.: Response of Apodemus flavicollis to conditions at the altitude limit in the western Tatra Mountains — Reaktionen von Apodemus flavicollis auf Lebensbedingungen an der Höhengrenze in der westlichen Watra 2.1. ee NE RRER I 185 Sabatini, V.; Paranhos de Costa, M. J. R.: Caecotrophy in pacas (Agouti paca Linnaeus, 1766) — Caecotro- phieibewBRakası(Agoun pacalEinnaeus, 17606), 2 a NE a 305 Sänchez-Villagra, M. R.; Nummela, S.: Bullate stapedes in some phalangeriform marsupials — Bullaför- mige Stapedesbeiiphalangeritormen Marsupialian 2 222 Se 174 Scharff, A.; Macholan, M.; Zima,J.; Burda, H.: A new karyotype of Heliophobius argenteocinereus (Bathyergidae, Rodentia) from Zambia with field notes on the species - Ein neuer Karyotyp von He- liophobius argenteocinereus (Bathyergidae, Rodentia) aus Sambia und Freilanddaten zur Art...... 376 Sommer, $.: Reproductive ecology of the endangered monogamous Malagasy giant jumping rat, Hypo- geomys antimena — Reproduktionsökologie der bedrohten, monogamen madagassischen Springratte, Hypogeomys antimena.:... 3.22 KIN TER TR N EEN EE RRERN R ul Stevens, R. D.: Twinning in the big fruit-eating bat Artibeus lituratus (Chiroptera: Phyllostomidae) from eastern Paraguay - Zwillingsgeburt bei der großen fruchtfressenden Fledermaus Artibeus lituratus (Chiroptera: Phyllostomidae) aus demröstlichen Paraguay "mr... 222.2. BE 178 Vargas, J.; Simonetti, J. A.: New distributional records of small mammals at Beni Biosphere Reserve, Bolivia — Neue Nachweise zur Verbreitung von Kleinsäugern aus dem Beni Biosphere Reservat, Bolivien! 2, 1 MERKE BEA. DUDEN HEREIN ON 379 Wallner, B.; Huber, S.; Achmann, R.: Non-invasive PCR sexing of rabbits (Oryctolagus cuniculus) and hares (Lepus europaeus) — Nicht-invasive Geschlechtsbestimmung mittels PCR bei Kaninchen (Oryctolagus' cunieulus)'und Eeldhasen’(Zepusieuropaeus)". 2.22. ES nn 190 Walston, J.,; Veron, G.: Questionable status of the “Taynguyen civet”, Viverra tainguensis Sokolov, Rozh- nov and Pham Trong Anh, 1997 (Mammalia: Carnivora: Viverridae) — Fraglicher Status der „Tayn- guyen Zibetkatze“ Viverra tainguensis Sokolov, Rozhnov und Pham Trong Anh, 1997 (Mammalia: Carnivora:Viverndae).... Ya WLMHADT IM TE IR SR MR LM SER ER 181 Wang, G.; Zhou, Q.; Zhong, W.; Wang, D.: Comparative food preference of Microtus brandti and Ochoto- na daurica in grassland of Inner Mongolia, China — Vergleichende Nahrungspräferenz von Microtus brandti und Ochotona daurica im Grasland der Inneren Mongolei, China ...............2.22000. Sr Wang, H.; Fuller, T. K.: Notes on the ecology of sympatric small carnivores in southeastern China — Bemerkungen zur Ökologie von sympatrischen kleinen Carnivoren im südlichen China .......... 251 Weinandy, R.; Hofmann, S.; Gattermann, R.: Mating behavior during the estrus cycle in female Mongo- lian gerbils (Meriones unguiculatus) — Paarungsverhalten während des Östruszyklus bei der weibli- chen Mongolischen Wüstenrennmaus (Meriones unguiculatus) .........--»==22222ee een nnnn 116 Wittmer, H. U.: Home range size, movements, and habitat utilization of three male European wildcats (Felis silvestris Schreber, 1777) in Saarland and Rheinland-Pfalz (Germany) — Wohngebietsgröße, Ortsveränderung und Habitatnutzung von drei männlichen Wildkatzen (Felis silvestris Schreber, 1777) im Saarland und.in Rhemland-Pfalz (Deutschland) .2.....2... 2.2. ee 365 Mitteilung der Gesellschaft... 2.2... Weser. aa a. Se SE 63 Bookireviewsan.a. 2 SR er ab Be 127, 256, 320, 383 Mammalian Biolog; Zeitschrift für Säugetierkunde 1.2001 | mr oa )\ ISSN 1616-5047 Mamm. biol. - 66(2001)1 - 1-64 www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Säugetierkunde (German Society for Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel R. Hutterer, Bonn E. Kalko, Ulm P. Lüps, Bern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repennning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.O. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, : Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/46 35 70, E-mail: l.still@nature.com Subscribers in the USA can call toll free on ++1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 493.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipients name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by persönal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. Ifno cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH &) As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANSI/NISO 239.48-1992 (Permanence of Paper). ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Printed in Germany Mamm. biol. 66 (2001) 1-12 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Review Genetic heterogeneity of white-tailed deer: management lessons from a long-term study By M. H. SMITH, J. M. Novak, J. D. PELES, and J. R. PURDUE University of Georgia’s Savannah River Ecology Laboratory, USA, Institute of Ecology, University of Georgia, Athens, GA, USA, Department of Genetics and School of Forest Resources, University of Georgia, Athens, GA, USA, Ostermayer Laboratory, Pennsylvania State University, McKeesport, PA, USA, and Zoology Section, Illinois State Museum, Springfield, USA Receipt of Ms. 24. 07. 2000 Acceptance of Ms. 20. 10. 2000 Abstract Genetic data from a long-term (16-year) study of white-tailed deer (Odocoileus virginianus) on the U.S. Department of Energy’s Savannah River Site (SRS) were examined to evaluate spatial and tem- poral genetic heterogeneity in this species. Based on our analyses of the long-term data set, three major findings emerged, all of which have important implications for management of white-tailed deer: (1) There exists significant spatial genetic heterogeneity in white-tailed deer based on ana- lyses of allozyme frequencies and mtDNA haplotypes. This heterogeneity exists on a much smaller spatial scale than would be expected for such a large and potentially mobile species as 0. virginia- nus. (2) The genetic structure of white-tailed deer at SRS is temporally dynamic and significant het- erogeneity exists within demographic units such as age and sex classes. (3) Levels of genetic varia- tion, as measured by multilocus heterozygosity, are frequently correlated to characteristics that are important determinants of ecological function in white-tailed deer populations. These findings are evaluated in the context of a general management model for O. virginianus that is also applicable to other wildlife species. Key words: Odocoileus virginianus, allozymes, mtDNA, spatio-temporal heterogeneity, demographic heterogeneity Introduction For most of this century, population geneti- cists and evolutionary biologists have as- sumed that populations consist of a large number of randomly breeding individuals (panmixia). This view made it easier to mathematically describe the behavior of po- pulations and resulted in a relatively static concept of their genetic characteristics. Lit- tle effort was expended in linking genetic and demographic changes in populations. 1616-5047/01/66/01-001 $ 15.00/0. Wildlife biologists considered changes in po- pulation numbers, quality of individuals within them, and other demographic para- meters as being due to environmental ef- fects, and genetic differences were often not considered at all. Despite this, the envi- ronmental or habitat model, which became the almost exclusive population dynamics paradigm in wildlife biology, was very suc- cessful in explaining population differences. 2 M. H. SMITH et al. The term “genetics” was not even men- tioned in most wildlife management texts during the first two thirds of this century. Technological advances in the 1950s and 1960s made it much easier to describe char- acter variation among individuals and to determine the genetic basis of this varia- tion. There was a virtual explosion in the number of studies that provided estimates of genetic varıation in natural vertebrate populations (SMITH et al. 1982, 1994). As a result of these studies, it became clear that the model of a large panmictic population was not correct for most terrestrial and freshwater vertebrates (e.g. SMITH et al. 1978; AvısE 1994). However, most of the data, especially for mammals, were from small relatively short lived forms (e.g. Kregs et al. 1973). Data from the white- tailed deer summarized here support the view that genetic heterogeneity over short distances may be common even in large, va- gile vertebrates. Temporal genetic heterogeneity over short time predicts the need for further refine- ment of habitat management models used in wildlife management. Characteristics of concern to natural resource management, including conservation, need to be thought of as being due to the influences of Envi- ronment (E; Habitat) + Genetics (G; Geno- type) + Environment-Genetic Interactions (E*G). A holistic perspective would dictate that the enviromnent-genetic interactions would be at least as important in determin- ing the characteristics of wildlife species as the main effects of genotype and environ- ment. Studies that document differential po- pulation responses to similar environmental changes may indicate the importance of en- vironment-genetic interaction and/or differ- ences in the genetic composition of the ref- erence populations. This interpretation stresses the importance of genetic factors in formulating management programs for both game and nongame species. Genetics is most likely to be important if management units have different genetic characteristics from each other and/or they show temporal variations in their genetic characteristics. Our primary objective is to examine existing genetic evidence to see how common spatial and temporal heteroge- neity is in white-tailed deer (Odocoileus vir- ginianus, Zimmermann). Our purpose is to review the literature on the genetics of the white-tailed deer, present the results of some new analyses of data from a long-term study of this species, and to propose a new per- spective on the important conceptual issues. Sampling considerations Management decisions based upon data collected from public hunts need to be viewed with caution. Such data must be ex- amined to determine if inferences can be expanded beyond the limits of the available data in time and/or space. Basically this re- quires that anımals are collected randomly with respect to variables of interest such as sex, age, antler morphology, genotype, etc. Deer collected on the Savannah River Site (SRS) in the southeastern United States, because of the limited public access and the details of the hunting methods used, can generally be considered to represent a random sample of individuals from the herd for most variables of interest (Novak et al. 1991). Novak et al. (1991) found no hunter selectivity based upon sex but some selec- tivity based upon age (older deer being pre- ferentially selected) thus slightly biasing the distribution of ages upwards. Thus age-re- lated genetic changes may be harder to de- tect than genetic changes related to sexual differences. Spatial heterogeneity Many genetic studies have shown that white-tailed deer populations are subdi- vided spatially. The effect is most noticeable in analyses that encompass large geographic areas (CRonIN 1989; ELLSWORTH 1994a, b; HıLLestap 1984; KENNEDY et al. 1987). In these studies Fsr (or a similar statistice that estimates the proportion of variance among populations) for both diploid (allozymes) and haploid (mitochondrial DNA [mtDNA]) genetic markers is large, indicat- ing strong differentiation between local populations. On a small geographic scale, it is possible that spatial subdivision would not exist for a large, potentially mobile mammal, such as the white-tailed deer. However, a number of studies reject this notion. Spatial differentia- tion of populations for allozyme frequencies was readily apparent in white-tailed deer from the Adirondack Mountains of New York (MATHEws and PorTER 1993), north- eastern Minnesota (Cronin et al. 1991), and on an even smaller scale, the SRS, South Car- olina (MAnLovE et al. 1976; Ramsey et al. 1979), and Cumberland Island, Georgia (RowrLAnD 1989). When studied, mtDNA markers usually, but not always show greater differentiation than those representing the nuclear genome. For example, Cronin et al. Frequency Frequency Allele Genetic heterogeneity of white-tailed deer 3 (1991) found the Fsr value for mtDNA to be 9 times greater than the Fsr for allozymes in mule deer from Montana but found no sig- nificant difference between mtDNA and al- lozyme-derived Fs7r values for white-tailed deer from Minnesota. Generally, genetic differentiation of popula- tions is attributed to reduced gene flow, his- toric events and/or genetic drift (Cronin et al. 1991; ELLSwoRTH et al. 1994 a, b; LEBERG et al. 1994). In white-tailed deer, gene flow is influenced strongly by the species’ mating system, females being philopatric and males doing the majority of movement among breeding groups (NELSon and Mech 1987). The effect of extirpation in the late 1800s and subsequent restocking have had a pro- found effect on the spatial pattern of genetic differentiation of white-tailed deer popula- tions over most of their range. However, in Haplotype Fig. 1. Comparison of haploid (mtDNA) and diploid (allozyme) genetic markers for white-tailed deer populations collected in 1992 from the Savannah River Site (SRS; Nmtona = 215, Natozym = 737) and Webb Wildlife Center (WEBB, Nmtona = 31, Natlozym = 32). The populations are separated by approximately 100 km. Shown are aconitate hydratase (AH), adenosine deaminase (ADA), and L-iditol dehydrogenase (IDDH) (also known as sorbitol dehy- drogenase [SORDH]), the three most variable of the 13 loci sampled. Designations for alleles refer to relative mo- bility in electrophoretic starch gels. Only haplotypes and alleles with frequencies > 0.01 are shown. 4 M. H. SMITH et al. the coastal plain of South Carolina and Georgia, native herds were not hunted to extinction and restocking was minimal. Re- cent analyses of deer from SRS and Webb Wildlife Center, located 100 km apart on the coastal plain of South Carolina, docu- ment significant spatial heterogeneity in both nuclear and mtDNA genomes. Deer sampled from SRS and Webb center display markedly different genetic profiles for nu- clear and mitochondrial genes (Fig. 1). This and other studies (KENNEDY et al. 1987) in- dicate that for allozymes all alleles at a locus are present in most samples, although shifts in frequencies are often observed. In con- trast, mtDNA types, which are haploid and maternally inherited, are much more local- ized. Sometimes, sampling locations sepa- rated by only 20 km share no mtDNAÄ types. Female white-tailed deer thus may be extremely philopatric (PuURDUE et al. 2000). The role of female philopatry in the mainte- nance of genetic structure of white-tailed deer can be seen in an inadvertent “experi- ment” provided by the restocking of deer in Greene county on the piedmont of Geor- gia. Early in the twentieth century, native deer were extirpated from Greene and sur- rounding counties and never recolonized the area. In the late 1980s, extensive re- stocking was undertaken in the area. North- ern Greene county was supplied with 60 deer from Ossabaw Island and 7 from adjacent DBlackbeard island, Georgia (BLACKARD 1971). The Ossabaw Island deer carry a mtDNA type unique to the island and a few mainland localities on the lower coastal plain. In counties adjacent to Greene, deer were transplanted from Texas and Wisconsin. In 1994, the mtDNA of 20 deer from Greene county were exam- ined. Seven of ten deer sampled in the northern part of the county carried the Os- sabaw island mtDNÄ type. The other three, plus 10 additional individuals from southern Greene county, displayed mtDNA types characteristic of deer from the Midwestern United States. After 40 years and 10- 20 generations, female deer from Ossabaw Island have apparently dispersed little be- yond their release site. These results rein- force the idea that white-tailed deer are ge- netically subdivided on a finer geographic scale than is apparent based upon their body size and vagility. Demographic heterogeneity Management decisions are usually made for a herd or larger grouping of individuals. However, smaller subsets of individuals (age or sex classes) may be progressing along separate evolutionary trajectories sub- ject to differing ecological challenges. These demographic groups may exhibit different spatial or temporal patterns for both indivi- duals and genotypes. Thus, genetic variabil- ity must be analyzed with respect to demo- graphic classes of age and/or sex within a spatio-temporal context. The SRS deer herd provides a unique opportunity to analyze such data because of the size of the data set within years (Minimum =409, Maxi- mum = 1999, Total = 14221 deer), number of years for which data are available (16) and limited public access to the site. Demographic heterogeneity in the SRS deer herd was analyzed for the years 1974- 1989 based upon 7 polymorphic locı avail- able in all years. Data for two highly poly- morphic loci, ß-hemoglobin and transferrin, were not available for the year 1980, so that year was not included in the analysis. Thus, all deer were categorized for multilocus heterozygosity class based upon 7loci (HCI was 0, 1, 2, 3 and 4+ heterozygous loci, and H [arcsine of square root HC/Total number loci scored]), year of collection (TIME), age class (AGE)(0921 5735, 3.5+ years), sex (SEX), and spatial unit (SPACE) (swamp or upland herd). Ex- panded definitions of the above variables can be found in SCRIBNER et al. (1985) and Novak et al. (1991). Probabilistic regression (PROBIT) analysis indicates that the distribution of AGE is a function of both "TIME TandSPACE (x?=61.65, P<0.0001 and x”= 13.09, P = 0.0003, respectively). However, the dis- tribution of SEX is a function of TIME but not SPACE (x’=48.24, P<0.0001 and x” =0.69, P= 0.4075, respectively). Thus, analyses of genetic heterogeneity in rela- tion t0 AGE and SEX must be performed with the appropriate spatial and temporal variables in the analysis. Probabilistic regression using a Gompertz distribution for HC (GOMPIT) analysis in- dicates that there are significant SPACE (X=7.32, P=0.0068) and TIME (x? = 101.64, P < 0.0001) effects, a marginal AGE (x? = 6.59, P = 0.0863) effect and no SEX (x? = 0.02, P = 0.8989) effect. Unfortu- nately, interactions among dependent vari- ables cannot be analyzed using a probabilis- tic regression approach to account for TIME and/or SPACE heterogeneity of SEX and AGE. Therefore, an ANOVA was performed with H as the dependent variables and the main effect of SEX (E0537 2 PE10.4676), ’AGE: 1 (F=.0.82, P = 0.4799), TIME (F=3.84, P< 0.0001), and SPACE (F = 4.19, P = 0.0406), and the two-way interactions of SEX and AGE (a0 3417), 1SEX and "TIME = Female —ı—- Male Mean Multilocus Heterozygosity 1973 1974 1975 1976 1977 1978 1979 1981 Genetic heterogeneity of white-tailed deer 5 (F=1.87, P= 0.0242), AGE and TIME (F= 1.17, P= 0.2066), AGE and SPACE (F=0.34, P=0.7930), and TIME and SPACE (F = 1.64, P = 0.0621). No higher order interactions were significant, and were therefore not included in the model. The significant interaction of SEX and TIME is due to differenees in H between males and females in different years (Fig. 2). There is no consistent sexual bias in H, 6 years show no significant difference, 5 years show a male bias for higher H, and 4 years show a female bias (Fig. 2). Previous analysis for the effects of age, sex, year and spatial location on single locus heterozygosity (h) for ß-hemoglobin by CHESSER et al. (1982) revealed slightly dif- ferent results. Sex was not found to be an important variable although it is unclear whether a sex by year interaction was tested. This analysis was performed over only a three year time span, for only a sin- gle locus and used simple tests of indepen- dence that did not analyze variables concur- 1982 1984 1983 1986 1985 1988 1987 1990 1989 Year Fig. 2. Multilocus heterozygosity values for male and female deer for the years 1974 through 1989. The year 1980 is not included as indicated in the text. 6 M.H. SMITH et al. rently. As indicated by the analyses per- formed here, there is a much larger range of variation in all variables when analyzed over a longer time span. In addition, longer time series are more likely to include peri- ods of environmental stress. Thus, results based upon data that are limited in time, space or number of loci should be viewed with caution. Differences in results can also be seen in the studies of SMITH et al. (1990) where a significant spatial effect was seen and SCRIBNER et al. (1985) where a signifi- cant effect of space was not seen. The first study included data from a longer time se- ries (13 years) than the second (6 years) but both estimated H using the same seven loci used here. The above analyses illustrate the need to examine demographic effects on genetic heterogeneity in light of spatial and tempor- al variation of both demographic and genet- ic variables. Management decisions based upon only the main effect, SEX, would not be the same as those based upon the inter- action of SEX and TIME. The interaction of SEX and TIME is not surprising for the SRS white-tailed deer herd given the rela- tionships between male body mass and fat levels (SCRIBNER et al. 1989), female fat lev- els and their relationship to pregnancy (CoTHRAN et al. 1987), conception date of females (RHopDes and JoHnns 1993) and fe- male age specific body mass (RHoDESs et al. 1991). It is unclear if white-tailed deer are unusual for mammals in how they partition genetic variation in space and time. Although other studies have analyzed de- mographic heterogeneity, few have looked at the interaction of age and/or sex with space and none have analyzed differences over a comparable time span (SMITH et al. 1994). The interaction of SEX and TIME has direct consequences for the estimation of genetically effective population sizes and minimum viable population sizes. If dif- ferent demographic units are present in a population and each is progressing along in- dependent or semi-independent evolution- ary trajectories then management plans need to encompass this heterogeneity. Man- agement decisions must be based upon in- formation gathered to assess the additional ecological and genetic dynamics that such population substructuring introduces. Fitness correlates and energetics Fitness correlates A fitness correlate may be defined as a phe- notypic characteristic in which the degree of expression is related to the survival and/ or reproductive success (fitness) of an indivi- dual. Numerous relationships between mul- tilocus heterozygosity (H) and fitness corre- lates have been demonstrated in a long- term study of white-tailed deer on the SRS (reviewed by RHoDEs and SMITH 1992). Within age classes of male deer, H is related to (a) body mass and fat levels (SCRIBNER et al. 1989), (b) antler size (SCRIBNER et al. 1989), (ce) antler symmetry and Boone and Crocket scores (SMITH et al. 1991), (d) fre- quency of spike antlers (SCRIBNER et al. 1984), and (e) testicle size in fawns (URB- ston 1976). H in female deer is correlated with (a) the frequency of twin fetuses (CHES- SER and SMITH 1987; JoHns et al. 1977), (b) age-specific body mass (RHoDes et al. 1991), (ce) conception date and fetal growth rate (CoTHRAN et al. 1983; RHODES and Jonns 1993), and (d) body fat levels prior to conception and loss of fat during pregnancy (CoTHRAN et al. 1987). Fetal growth rate is also related to the overall H of the fetus (Co- THRAN et al. 1983; LEBERG et al. 1990). SMITH and RISENHOOVER (1993) demon- strated a positive association between H and production of offspring in eight species of cervids. In addition, relationships between H and fitness correlates have been observed in many other organisms (ALLENDORF and LEARY 1986; MıTTon and GRANT 1984). Thus, H likely integrates many important genetic characteristics of forest organisms. The general trend of these relationships de- scribed for white-tailed deer is for expres- sion of the reference character to increase (e.g., antler size) or decrease (e.g., inci- dence of spiked antlers) with increasing number of heterozygous loci. However, the functional relationship varıies depending on both the specific character and the age of the deer. In addition, there is evidence to suggest that expression of a reference char- acter may decrease slightly at high H levels compared to that of intermediate levels (e.8., CHESSER and SmıtH 1987) although this may be an artifact of small sample size at older age classes. In most cases, H explains only a small per- centage of the variability in characteristics. For example, H is responsible for only 10- 15% of the variability in main beam length and diameter of antlers, number of antler points, and incidence of spiked antlers (SCRIBNER and SMITH 1990). Therefore, fac- tors such as age, body condition, habitat, and resource quality, as well as their inter- action with H, must be considered when ex- plaining the expression of fitness-related characteristics in individual deer. Although H may only account for a small amount of the variability in characters, deer with high H generally grow faster, have higher body fat levels and higher reproduc- tive rates than deer with low H. These rela- tionships suggest that deer with various le- vels of H may partition their energy differently. The potential relationship of H to energetics requires further consideration. Heterozygosity and energetics An organism’s energy budget can be de- scribed by I=A +E, where I is the total amount of energy (Kcal *gbody mass'') ingested, A is assimilated energy, and E is egested energy (egestion). Assimilated en- ergy is partitioned into three categories with A=M+G+R where M is mainte- nance energy and G + R represents assimi- lated energy used for growth or reproduc- tion (1. e., secondary productivity). A number of investigations have demon- strated a relationship between H and ener- getic parameters (reviewed by Mırtton and GRANT 1984). H has been correlated with decreased rate of oxygen consumption (KoEHN and SHUMWwAY 1982; MıTTon and KoEnHn 1985; Mıtton et al. 1986) and a low- Genetic heterogeneity of white-tailed deer 7 er rate of protein turnover (HAwkıins et al. 1986). These findings suggest differences in maintenance metabolism among individuals with varying levels of H. We hypothesize that increased energetic ef- ficiency could explain the effects of H on fit- ness-related characteristics in white-tailed deer. Hypothetical energy budgets for an or- ganism with varying H are depicted in Fig. 3. In both homozygous and heterozygous indi- viduals, a portion of assimilated energy must be utilized for maintenance metabolism (M) which includes energy used for normal ac- tivity. The remaining energy can be used for secondary productivity (G+R). How- ever, in the more heterozygous individual, increased energetic efficiency as a result of higher H could reduce the amount of assıimi- lated energy required for maintenance me- tabolism (M). A slight decrease in the amount of energy needed for maintenance could permit heterozygous individuals to partition much more energy for growth and reproduction (G+R, Fig. 3a). The above hypothesis assumes that ingested energy (I) is relatively constant among indi- viduals. However, individuals with higher H may be able to ingest more energy as a re- sult of aggressive behavior (GARTEN 1976) or an increased scope of activity (MITTON and GRANT 1984). Consequently, assimi- lated energy would be greater among more heterozygous individuals, providing more energy for growth and reproduction, even if energetic efficiency is not affected by H (Fig. 3b). The effect of H on energetics is most likely to result in a selective advantage during periods of stress (KoEHN and SHUMWAY 1982; RODHOUSE and GAFFNEY 1984; TESKA et al. 1990). TEsKkA et al. (1990) demon- strated that old-field mice of varying H dif- fer regarding feeding efficiency only as food quality is decreased. These results suggest that the effects of temporal variation of H may be to decrease the ability to detect dif- ferences in H among individuals during non-stressful periods. These findings may explain the inconsis- tency of some relationships between H and fitness correlates observed in white-tailed 8 M. H. SMITH et al. deer. For example, a relationship between et al. 1977) whereas no such relationship H and the frequency of twin fetuses was ob- was found during the 1980s (RHoDes et al. served among does from the SRS during 1991). Future investigations concerned with the 1970s (CHESSER and SMITH 1987; Jouns documenting H effects in white-tailed deer A LOW H HIGH H ZN (el a B HIGH H BE Fig. 3. Hypothetical energy budgets for an organism with relatively low and high levels of heterozygosity (H). High H may increase the amount of energy available for growth (G) and reproduction (R) by: (A) Reducing the percentage of assimilated energy needed for maintenance (M) via effects on metabolic efficiency or: (B) Increas- ing the amount of assimilated energy via effects on foraging and ingestion. The size of each circle is related to the amount of ingested energy. should take into account spatial and tem- poral variation in environmental quality as well as in H. The influence of H on energetics is related to individual fitness and quality of indivi- duals in a population. Genetic variability could be especially important in allowing forest organisms to persist with increasing levels of anthropogenic and non-anthropo- genic stress. Understanding the role of ge- netic variation has important implications _ for both conservation and management practices of forest wildlife species. General management model Genetic analyses of white-tailed deer popu- lations, as well as other animal populations, have provided insights about their function- ing that need to be incorporated in future management plans (SMITH et al. 1976). The results of these analyses are especially im- portant to the formulation of management plans. They are as follows: 1) animal popula- tions, especially white-tailed deer, show ge- netic heterogeneity over relatively short dis- tances and among demographic units within populations, 2) white-tailed deer popula- tions, and probably those of other species, are generally dynamic over short time peri- ods, and 3) levels of genetic variability are frequently correlated to many characteris- tics that are important determinants of eco- logical functioning of populations and of concern to natural resource managers. Althoush the correlation of genetic variabil- ity and phenotypic characteristics do not usually explain a large proportion of the to- tal variation, each correlation may be some- what independent such that the overall ef- fects on the ecologieal dynamics of the population function are very important. White-tailed deer show a surprising amount of spatial genetic heterogeneity even in areas like the SRS where the habitats are not severely fragmented. In areas where forested habitats are becoming even more fragmented (Harrıs 1984), spatial hetero- geneity in gene frequency may be further increased. Spatial genetic heterogeneity needs to be taken into account in defining Genetic heterogeneity of white-tailed deer 9 boundaries of management units. In addi- tion, conservation efforts need to recognize that many forms of a species having unique combinations of genes may occur in subpop- ulations separated by short distances. Spa- tial heterogeneity in gene frequencies has been recognized in a wide diversity of ani- mals, and its management implications have been recognized as important in fisheries management (RyMmaAn and UTTER 1987). Wide scale fragmentation of forested habi- tat can lead to reduction of census and ef- fective population sizes, which may fall be- low the minimum viable size (SouULE 1987). One of the most important long-term ef- fects of falling below the minimum viable population size is stochastic loss of genetic variability, which is important for both the future evolution and the ecological func- tioning of populations. Small populations may also be more susceptible to the effects of inbreeding, especially if population num- bers are reduced quickly and kept low for an extended period of time (THORNHILL 1993). Although we do not know whether genetic variability causes changes in popu- lation parameters and/or is a result of them, it would seem prudent to manage popula- tions in a way that minimizes the chance of losing genetic varıabiılity. The genetic structure of populations is tem- porally dynamic over time periods that in- clude the length of typical studies (SMITH et al. 1990). This dynamic behavior of popu- lations may result from the interactions from smaller groups that differ from each other genetically. Animals that disperse among these subpopulations to breed may have relatively outbred offspring with high- er levels of genetic variability and different phenotypic characteristics than those that breed within the subpopulation in which they were born. Management of forest ha- bitats (e.g., maintaining corridors) to allow this type of dispersal among subpopulations may be essential to the long-term health of many of forest animals (HArrıs 1984), espe- cially large vertebrates. One measure of the success of various man- agement programs could be the degree to which we maintain the genetic integrity of 10 M. H. SMITH etaal. the species. Genetic integrity must not be based on a static concept of the genetic characteristics of the species. Populations are extremely dynamic through space and time, and it seems prudent to manage biolo- gical resources so that they continue to ex- hibit their normal variation in both space and time (Norse et al. 1986). Thus, we are trying to manage species that are likely to be genetically different in both space and time, and these genetic differences are likely to have direct relationships with bio- logical characteristics important to both the survival of the species and the produc- tion of benefits for humans. As human so- Zusammenfassung ciety continues to increase its impact on every habitat on earth, it will be challenging to devise management and conservation strategies for our precious life support sys- tems, especially forests. Acknowledgements We wish to thank all of the people, and especially PAUL E. JoHns, involved in collecting the electro- phoretic data from the SRS herd. This research was supported by contract DE-FC09-96 SR 18546 between the University of Georgia and the U. S. Department of Energy. Genetische Heterogenität beim Weißwedelhirsch: Für die Wildbewirtschaftung relevante Erkenntnisse aus einer Langzeitstudie Daten aus einer Langzeitstudie (16 Jahre) an Weißwedelhirschen (Odocoileus virginianus) aus dem Sa- vannah River Site (SRS) des U.S.. Department of Energy wurden im Hinblick auf das Vorkommen von räumlicher und zeitlicher genetischer Heterogenität bei dieser Art analysiert. Die Untersuchung er- brachte drei wesentliche Befunde, die auch für die Bewirtschaftung des Weißwedelhirsches von Be- deutung sind: (1) Wie aus der Analyse von Allozymfrequenzen und mtDNA-Haplotypen hervorging, besteht in Populationen des Weißwedelhirsches eine ausgeprägte räumliche genetische Heterogeni- tät, und zwar auf wesentlich geringerem Raum, als man dies bei einer potentiell so mobilen Art erwar- ten würde. (2) Die genetische Struktur der Weißwedelhirsche am SRS ist zeitlich unterschiedlich und es gibt eine ausgeprägte Heterogenität zwischen demographischen Entitäten wie Alters- und Geschlech- terklassen. (3) Die in elektrophoretischen Untersuchungen ermittelte Heterozygotierate ist häufig mit Merkmalen korreliert, die für die ökologischen Beziehungen in Weißwedelhirschbeständen bedeutsam sind. Diese Befunde wurden im Rahmen eines generellen Bewirtschaftungsmodells für O. virginianus evaluiert, das auch für andere Wildtierarten anwendbar ist. References ALLENDORF, F. W.; LEARY, R. F. (1986): Heterozyg- osity and fitness in natural populations of ani- mals. In: Conservation Biology: The Science of Scarcity and Diversity. Ed by M. E. SouLe£. Sunderland, Massachusetts: Sinauer Associ- ates. Pp. 57-76. Management of the Cervidae Ed. by C. M. WEMMER. Washington, D.C.: Smithso- nian Institution Press. Pp. 168-177 CHESSER, R. K.; SMITH. M. H.; JoHns, P. E.; MAN- LOVE, M.N.; STRANEY, D. O.; Baccus, R. (1982): Spatial, temporal, and age-dependent AVISE, J. C. (1994): Molecular Markers, Natural History and Evolution. New York: Chapman and Hall. BLACKARD, J. J. (1971): Restoration of the white- tailed deer in the Southeastern United States. M.S.thesis, Louisiana State University, Baton Rouge, Louisiana. CHESSER, R. K.; SMITH, M. H. (1987): Relationship of genetic variation to growth and reproduc- tion in the white-tailed deer. In: Biology and heterozygosity of beta-hemoglobin in white- tailed deer. J. Wildl. Manage. 46, 983-99. COTHRAN, E.G.; CHESSER,R.K.; SMITH, M.H. (1983): Influences of genetic variability and maternal factors on fetal growth in white- tailed deer. Evolution 37, 282-291. COTHRAN, E.G.; CHESSER,R.K.; SMITH, M.H.; Jonns, P. E. (1987): Fat levels in female white- tailed deer during the breeding season and pregnancy. J. Mammalogy 68, 111-118. Genetic heterogeneity of white-tailed deer 11 CRoNIN, M. A. (1989): Molecular evolutionary ge- netics of cervids. PhD. dissertation, Yale Uni- versity, New Haven, Connecticut. CRoNIN, M. A.; NELSoN, M. E.; Pac, D. F. (1991): Spatial heterogeneity of mitochondrial DNA and allozymes among populations of white- tailed deer and mule deer. J. Heredity 82, 118-127. ELLSWORTH, D. L.; HONEYCUFF, R. L.; SıLvy, N. J.; SMITH, M.H.; BicKHAM, J. W.; KLım- STRA, W.D. (1994 a): White-tailed deer res- toration to the southeastern United States: evaluating genetic variation. J. Wildl. Manage. 58, 685-697. ELLSWORTH, D. L.; HONEYCUTT, R. L.; SıLvy, N. J.; BICKHAM, J. W.; KLıMsTRA,W.D. (1994b): Historical biogeography and contemporary patterns of mitochondrial DNA variation in white-tailed deer from the southeastern Uni- ted States. Evolution 48, 122-136. GARTEN, C. T. JR. (1976): Relationships between aggressive behavior and genetic heterozygos- ity in the oldfield mouse, Peromyscus poliono- tus. Evolution 30, 59-72. HARRIS, L. D. (1984): The Fragmented Forest: Is- land Biogeography Theory and the Preserva- tion of Biotic Diversity. Chicago: University of Chicago Press. HAwRINS, A.J.S.; BAYNnE, B. L.; Day, A.J. (1986): Protein turnover, physiological energetics and heterozygosity in the blue mussel, Mytilus edu- lis: the basis of variable age-specific growth. Proc. Royal Soc. London 229, 161-176. HiLLESTAD, H. ©. (1984): Stocking and genetic variability of white-tailed deer in the south- eastern United States. PhD dissertation, Uni- versity of Georgia, Athens, Georgia. JOHNS, P.E.;, BAccus, R.; MANLOVE, M.N.; Pın- DER, J. E.; SMITH, M.H. (1977): Reproductive patterns, productivity and genetic variability in adjacent white-tailed deer populations. Proc. Southeastern Assoc. Game Fish Comm. 31, 167-172. KENNEDY, P.K.; KENNEDY, M.L.; Beck, M.L. (1987): Genetic variability in white-tailed deer (Odocoileus virginianus) and its relation- ship to environmental parameters and herd origin (Cervidae). Genetica 74, 189-201. KoEHn, R.K.; SHUMwAy, S.R. (1982): A genetic/ physiological explanation for differential growth rate among individuals of the Ameri- can oyster Crassostrea virginica (Gmelin). Marine Biol. Lett. 3, 33-42. LEBERG, P.L.; SMITH, M.H.; RHoDes, OÖ. E. JR. (1990): The association between heterozygos- ity and growth of deer fetuses is not explained by effects of the loci examined. Evolution 44, 454-458. LEBERG, P. L.; STANGEL, P. W.; HILLESTAD, H. O.; MARCHINTON, R. L.; SMITH, M. H. (1994): Ge- netic structure of reintroduced wild turkey and white-tailed deer populations. J. Wildl. Manage. 58, 698-711. MANLOVE, M.N.; SMITH, M. H.; HiLLESTAD, H. ©.; FULLER, S.E.; JOoHNS,P.E.; STRANEY, D. O. (1976): Genetic subdivision in a herd of white-tailed deer as documented by spatial shifts in gene frequencies. Proc. Ann. Conf. SE Assoc. Game Fish Comm. 30, 487-492. MATHEWwS, N. E.; PORTER, W.F. (1993): Effect of social structure on genetic structure of free- ranging white-tailed deer in the Adirondack Mountains. J. Mammalogy 74, 33-43. MıtTon, J. B.; GRANT, M. C. (1984): Associations among heterozygosity, growth rate and devel- opmental homeostasis. Ann. Rev. Ecol. Syst. 15, 479-499. Mıtton, J. B.; KoEHN, R.K. (1985): Shell shape variation in the blue mussel, Mytilus edulis, and its association with enzyme heterozygos- ity. Exp. Marine Biol. 10, 73-80. Mıtton, J. B.; CAREY, C.; KOCHER, T. D. (1986): The relation of enzyme heterozygosity to standard and active oxygen consumption and body size of tiger salamanders, Ambystoma ti- grinum. Physiol. Zool. 59, 574-582. NELSoN, M. E.; MEcH, L. D. (1987): Demes within a northeastern Minnesota deer population In: Mammalian Dispersal Patterns. Ed. by B. D. CHEPKO-SADE and A. T. HArpın. Chica- go: University of Chicago Press. Pp. 2740. NORSsE, E. A.; ROSENBAUM, K.L.; WILCOVE, D. S.; WiLcox, B. A.; ROMME, W. H.; JoHNSTON, D. W.; Stout, M.L. (1986): Conserving Biological Di- versity in our National Forests. Wilderness So- ciety, Alexandria, Virginia: Global Printing. NOVAK, J. M.; SCRIBNER,K.T.; Dwupront, W. D.; SMITH, M. H. (1991): Catch-effort estimation of white-tailed deer population size. J. Wildl. Manage. 55, 31-38. PURDUE, J. R.; SMITH, M. H.; PATToN, J. C. (2000): Female philopatry and extreme spatial het- erogeneity in a large mammal. J. Mammalogy 81, 179-185. RAMSEYy, P.R.; AviıseE, J. C.,;, SMITH, M.H.; URB- STON, D. F. (1979): Biochemical variation and genetic heterogeneity in South Carolina deer populations. J. Wildl. Manage. 43, 136-142. RHOoDES, O. E. JR.; SMITH, M.H. (1992): Genetic perspectives in wildlife management: the case of large herbivores. In: Wildlife 2000 Popula- tions. Ed. by D. McCurLocH and R. BARRET. New York: Elsevier. Pp. 985-99. 12 M. H. SMITH etaal. RHODES, O. E. JR.; JOHNS, R.S. (1993): Relation- ships between heterozygosity and conception date in white-tailed deer from South Carolina. In: Forests and Wildlife: Towards the 21st Century. Ed. by I. D. Thompson. Halıfax, Canada, International Union of Game Biolo- gists. Pp. 119-125. RHODES, O. E. JR.; SMITH, M. H.; CHESSER, R.K. (1991): Prenatal losses in white-tailed deer. In: Biology of Deer. Ed by R. D. Brown. New York: Springer. Pp. 390-397. RODHOUSE, P. G.; GAFFNEY, P.M. (1984): Effect of heterozygosity on metabolism during starva- tion in the American oyster Crassostera virgi- nica. Marine Biol. 80, 179-187. ROWLAND, R.D. (1989): Population genetics of white-tailed deer on Cumberland Island, Georgia. M.S. thesis, University of Georgia, Athens, Georgia. SCRIBNER, K.T.; SMITH,M.H. (1990): Genetic variability and antler development. In: Horns, pronghorns, and antlers. Ed. by G. A. BUBE- NIK and A. B. BUBENIK. New York: Springer. Pp. 460-473. SCRIBNER, K. T.; SMITH, M. H.; Jonns, P.E. (1984): Age, condition, and genetic effects of inci- dence of spike bucks. Proc. Ann. Conf. South- eastern Fish Wild. Agencies 38, 23-32. SCRIBNER, K. T.; SMITH, M. H.; JoRns, P. E. (1989): Environmental and genetic components of antler growth in white-tailed deer. J. Mam- malogy 70, 284-291. SCRIBNER, K.T.; WOoTEn,M.C.; SMITH,M.H.; JoHns, P.E. (1985): Demographic and genetic characteristics of white-tailed deer populations subjected to different harvest methods. In: Pro- ceedings of the Symposium on Game Harvest Management. Ed. by S.L.BEAsom and S. F. ROBERTSoON. Caesar Kleberg Foundation, Wildlife Research Institute, Kingsville, Texas. SMITH, M. H.; RISENHOOVER, K.L. (1993): Asso- ciation between production of offspring and varlability in cervids. In: Forest and Wildlife: Towards the 21st Century. Ed. by I. D. THomP- son. Halifax, Canada, International Union of Game Biologists. Pp. 113-118. SMITH, M.H.; MANLOVE,M.N.; JOULE,J. (1978): Spatial and temporal dynamics of the genetic organization of small mammal populations. In: Populations of Small Mammals Under Natural Conditions. Ed. by P.SnYDer. Special Publica- tion Series, Pymatuning Laboratory of Ecology, Vol. 5, Linesville, Pennsylvania. Pp. 99-113. SMITH, M.H.; Wıruıs, K.B.; Jonns, P.E. (1990): Spatial-genetic variation in a white-tailed deer herd. In: Proceedings of the 19th Con- gress of the International Union of Game Biologists. Vol. 1: Population Dynamics. Nor- wegian Institute for Nature Research. Ed. by S. MYRBERGET. Trondheim: Norway. Pp. 80-84. SMITH, M. H.; HILLESTAD, H. ©.; MANLOVE, M.N.; MARCHINTON, R. L. (1976): Use of population genetic data for the management of fish and wildlife populations. Trans. N. Amer. Wildl. Nat. Res. Conf. 41, 119-130. SMITH, M. H.; SCRIBNER, K. T.; JoHns, P. E.: RHo- DES, O. E. JR. (1991): Genetics and antler de- velopment. In: Proceedings of the 18th Con- gress of the International Union of Game Biologists. Ed. by B. BoBeEk. Jagiellonian Uni- versity Krakow, Poland. Pp. 323-325. SMITH, M.W.; AQUADRO,C.F; SMITH, M.H.; CHESSER, R.K.; ETGES, W. T. (1982): Biblio- graphy of Electrophoretic Studies of Bio- chemical Variation in Natural Vertebrate Po- pulations. Lubbock, Texas: Texas Tech Press. SMITH, M. H.; HERNANDEZ-MARTICH, J. D.; Novak, J. M.; STANGEL, P. W.; LoweEry,J.G. (1994): Bibliography of Electrophoretic Studies of Biochemical Variation in Natural Vertebrate Populations. Vol. 2, 1981-1989. SouL£, M.E. (1987): Viable Populations for Con- servation. Cambridge, UK: Cambridge Uni- versity Press. TESKA, W. R.; SMITH, M. H.; Novak, J. M. (1990): Food quality, heterozygosity, and fitness cor- relates in Peromyscus polionotus. Evolution 44, 1318-1325. THORNHILL, N. W. (1993): The Natural History of Inbreeding and Outbreeding. Chicago, Ili- nois: University of Chicago Press. URrBsTon, D. F. (1976): Descriptive aspects of two fawn populations as delineated by reproduc- tive differences. PhD Dissertation, Virginia Polytechnic Institute, Blacksburg, Virginia. Authors’ adresses: MICHAEL H. SMITH and JamEs M. Novak, Univer- sity of Georgia’s Savannah River Ecology La- boratory, P. O. Drawer E, Aiken, SC, 29802, USA and Institute of Ecology, University of Georgia, Athens, GA, 30602, USA; MICHAEL H. SMITH, Department of Genetics and School of Forest Re- sources, University of Georgia, Athens, GA, 30602, USA (e-mail: smith@srel.edu); JoHn D. PELEs, 207 Ostermayer Laboratory, Pennsylvania State University McKeesport, 4000 University Drive, McKeesport, PA 15132, USA; JamESR. PURDUE, Zoology Section, Illinois State Museum, 1011 E. Ash St., Springfield, IL, 62703, USA. Mamm. biol. 66 (2001) 13-21 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Review Evidence for separate specific status of European (Capreolus capreolus) and Siberian (C. pygargus) roe deer By A. J. M. Hewison and A. DANILKIN Institut de Recherche sur les Grands Mammiferes, Institut National de la Recherche Agronomique, Toulouse, France and Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia Receipt of Ms. 14. 06. 2000 Acceptance of Ms. 13. 09. 2000 Abstract Two forms of roe deer, the European (Capreolus capreolus) and the Siberian (Capreolus pygargus), are widely recognised. Some authors consider these two forms as separate species, while others classify them as merely subspecies or races which are closely related. In this study, we compare the geographic distribution, morphological characteristics, karyotypes, biochemical variability, and potential for hybridisation of European and Siberian roe deer, addressing the question of their phylogenetic status. For most of historical times, the ranges of these two forms have been indepen- dent due to physical barriers such as glaciers or flooding. Overlap occurred for a time in the Middle Ages and again more recently, for the last few decades, but even then, the potential hybrid zone was small and hybrids are not thought to have persisted. The Siberian roe deer is substantially lar- ger than its European counterpart in all body measurements, with only the very smallest Siberian individuals and the very largest European deer of approximately equivalent size. Furthermore, the two forms can be reliably distinguished on the basis of cranial shape, due to differential rates of growth of the skull, illustrating the hiatus in morphology between the two forms. All European roe have a karyotype of 2n = 70, while Siberian roe possess between 1 and 14 additional accessory B-chromosomes, increasing clinally from west to east. Changes in karyotype seem to occur at phys- ical boundaries, suggesting the differences are due to partial or total absence of gene flow. On the basis of polymorphism of several enzymes as well as blood and muscle proteins, the genetic dis- tance between the two forms is characteristic of fairly reliable species. A series of hybridisation experiments have illustrated that, although successful crosses can be achieved, they more often result in stillbirths or birth complications leading to the death of both mother and kid, and reduced or complete infertility among F1 hybrid bucks. It is likely therefore that hybridisation in the wild would be rare or absent, and that hybrids would not persist in the face of immigration of either pure form. We conclude that by all the criteria of classical systematics, the European and Siberian roe deer are separate, good, species, albeit phylogenetically closely re- lated. Key words: Capreolus capreolus, Capreolus pygargus, species status, systematics 1616-5047/01/66/01-013 $ 15.00/0. 14 A. J. M. Hewison et al. Introduction Although roe deer (Capreolus sp.) were once classified as belonging to the Cervinae sub- family, it now seems clear that they are in fact part ofthe Odocoileinae (GRoVES and GRUBB 1987, Gruß 1993). However, taxonomic Te- lationships within this group are far less evi- dent, in particular the status of the various geographical forms of the genus Capreolus. Roe deer cover an enormous geographical distribution, ranging from Great Britain and Spain to the Far East and from Kazakhstan and central Asia to northern Scandinavia and Siberia, and a large amount of data has now accumulated which reveals great varia- tion of form over this range. This has lead certain authors to suggest that the genus con- tains more than one species and perhaps sev- eral subspecific forms (CoRrBET 1978; DA- NILKIN 1986a; LEHMANN and SÄGESSER 1986). Here, we review published data on geographic distribution, morphometry, and genetics of Capreolus to conclude whether this genus is monospecific or not. Geographic range Fossil records suggest that both the Euro- pean and Siberian roe deer forms have ex- isted since the Pleistocene period (DANILK- ın and Hewıson 1996). However, it seems that their geographical ranges remained in- dependent due to the glaciation of the Rus- sian plains and the Caspian Sea floods which extended far northward along the Volga. Once these barriers receded, the Si- berian roe deer moved west, colonising the plains up to the Dneiper and possibly reaching the northern Caucasus in the Mid- die Ages (Frerov 1952). Thus, more re- cently, prior to the twentieth century, the ranges of European and Siberian roe deer overlapped in a small part of their overall distribution, in the northern Caucasus and possibly also in the Dnepropetrovsk, Kiro- vograd and Orel regions (Fig. 1). Hybridisa- tion may well have occurred here, but due to reproductive barriers (see below) and the predominance of the European form, they almost certainly did not persist. In SE SI———— 27 > m 2 FI SE seen) iy= =— Pen TI . f a es: I gr! BR | ? Ei L \ — =) RER UNE 7 E mM =S \— \ m SS \ Q 14 WR I ) / NG: v @ Im en / / R IB 76-80 / ee ee 7 76-80 a ao r 75-78 EEE = Ä 9: 16,18 II 16,700 ln: 1 ——— HS Fig. 1. The distribution of Capreolus, showing variation in chromosome number (70-84) across its present geo- graphical range (black line), the extent of historical maximal overlap in range between the European and Siberian forms ([7]) and of present day overlap (3). The range of the European roe is to the left of this overlap zone and the range of the Siberian roe is to the right. Adapted from DanıLkın and Hewison (1996). modern times, reduction in range and num- bers of the Siberian roe due to excessive hunting and the abundance of predators re- sulted in discontinuities in geographical range and isolation of the European and Si- berian populations (DAnıLkın and HEWI- son 1996). However, numbers started to re- cover from the 1930s due to moderation of hunting and a warmer climate and the over- all range increased once again. As recently as the 1960s, the advance westward of Si- berian roe deer reached the Volga and sub- sequently the Khoper and Don rivers in the Volgograd region, bringing European and Siberian deer into contact once more during the last couple of decades (Fig. 1). Thus, the geographic ranges of the Euro- pean and the Siberian roe deer have been largely independent for much of history, overlapping only in a restricted area during certain periods. The complete isolation of the ranges of these two forms has only very recently been bridged again and the poten- tial hybrid zone remains very small with re- spect to the total geographic range. Further- more, there is little evidence that hybrids have persisted in any area, probably due to reproductive isolation between the two forms. Despite the fact that a large number of Siberian deer have been used for intro- duction programmes within the European roe deer’s range, only those introductions that took place where the European form was present in very low numbers or entirely absent have proved successful (DANILKIN and Hewıson 1996). Morphology Despite the fact that there is clearly sub- stantial environmental influence on overall body size and weight of roe deer (e.g. GAIL- LARD et al. 1996; Hewiıson et al. 1996, b), the European form is markedliy smaller than the Siberian form in all body dimen- sions (Fig. 2), including size of antlers (Eu- ropean: length 17-26 cm, span 7-14 cm, Si- berian: length >27 cm, span 17-20 cm) and skulls (condylobasal length: European 180- 200 mm, Siberian 201-231 mm). Some over- Specific status of roe deer 15 lap in size may occur between the very largest individuals of the European form and the very smallest Siberian roe deer, but more generally there is discontinuity in average size between adjacent populations at the range limits between these two forms. This discontinuity is due to differential rates of early growth and development: kids aver- aged 4 kg weight gain per month for Euro- pean roe and 6kg per month for Siberian roe when the two were kept together under identical environmental conditions (Da- NILKIN and Hewıson 1996). The difference persisted through to adulthood, when the Siberian roe weighed about 20% more in all seasons. In addition to simple size variation, Euro- pean and Siberian roe deer can be distin- guished on the basis of cranial shape. Multi- variate analyses of 905skulls from populations over the entire geographical range have identified two well-differen- tiated morphs, the Siberian and the Euro- pean (SoKoLov et al. 1985a). Again, this discontinuity appears early in life due to differential growth rates of the skull (SoKo- Lov et al. 1985b). There are also some indi- cations from this type of analysis that further discrimination within each main group may be possible, particularly for the Siberian morph (northern Siberia and the Far East), perhaps supporting the designa- tion of two or more subgroups (MARKOV et al. 1985 a; SoKoLov et al. 1986a; see also Hewıson 1997). An analogous analysis of antler characteristics was unable to distin- guish clearly between European and Sibe- rıan forms, presumably because of the pro- nounced influence of age, condition and environmental factors on these structures which are regrown annually (DAnILKın and HEwıson 1996). Genetic and biochemical variability The karyotypes (chromosomal morphol- ogy) of European and Siberian roe deer dif- fer dramatically. All populations of the European form are characterised by an identical Karyotype of 2n = 70, while all Si- 16 A. J. M. Hewison et al. 70 160 60 140 = 120 2 8) S S un << 8 100 40 3 = 3 2 30, u 5 e) 60 ae) 20 40 et 20 10 6) 2 d Q ‘go 6) ® Body Body Hind foot Body length girth length mass Character Fig. 2. Variation in total body length, body girth, hind foot length and body mass of European (white bars) and Siberian (black bars) male and female roe deer (adapted from Danıukın and Hewıson, 1996). The central dot de- notes the average value for each group, the bar gives the limits for population averages and the vertical line shows the range of extreme values of single individuals. berian populations have karyotypes which contain 1 to 14 additional accessory B-chro- mosomes, 2n = 70 + (1-14) (DAnıLkın 1985; SOKOLOV et al. 1986 b). Furthermore, the Si- berian roe deer exhibits mosaicism, particu- larly in the Far East, where different num- bers of B-chromosomes occur within different tissues ofthe same animal, as well as among individuals of the same popula- tion. In addition, all 35 pairs of the main set of chromosomes differ in length be- tween the European and Siberian groups. Ihe number of B-chromosomes present shows clear clinal variation, increasing steadily from west to east (Fig. 1). However, changes in karyotype across the geographic distribution of roe deer are abrupt and seem to occur at physical barriers such as mountain ranges. Hybridisation (see below) leads to inheritance of some B-chromo- somes among offspring, but the number in- herited is usually less than half the number of the Sıberian parent, probably due to un- equal segregation during meiosis. At the notional boundary between the two forms in the Ukraine and the northern Caucasus, individuals both with and without B-chro- mosomes have been identified (DAnNILKIN and Hrwıson 1996). At the biochemical level, electrophoresis of certain enzymes has revealed differences in protein polymorphism between European and Siberian roe. Of 14 systems tested, 3 were polymorphic in the European sample, while only two were polymorphic in the Si- berian sample and frequency differences between the two forms were found at one particular enzyme locus (SoKoLov et al. 1986 c). Isoelectric focusing of blood plasma proteins has identified differences in the pre-albumin zone of the spectrum which are fixed, i.e. all Siberian individuals are different from all European roe. Similarly, differences are also present in the IEF spec- tra of soluble proteins of the muscle tissue. These different protein fractions probably represent the products of alternative alleles for particular loci. Additionally, immuno- chemical investigations have indicated that the blood serum of European roe deer con- tains certain antigens which are characteris- tic of this group only and may also include two accessory antigens with very different molecular weights (MARKov et al. 1985 b). Hybridisation A large number of introductions of Siberian deer into areas inhabited by European roe have been carried out with the aim of in- creasing body weights and improving tro- phy quality (DAnıLkın and Hewıson 1996). Indeed, those hybrids that are able to sur- vive are heavier and have larger antlers than the pure European form. However, it seems probable that such operations have proved unsuccessful (see above), with even the introduction of a substantial number (several dozen) of Siberian animals result- ing in gradual but complete loss of the Si- berian form. Hybrid populations have not developed in the wild due to rather high level of repro- ductive isolation between the European and Siberian groups, ıillustrated by a series of experiments on captive deer. In the first experiments (STUBBE and BRUCHHOLZ 1979, 1980), two Siberian bucks were mated with a group of European does a total of 32 times. Of these matings, 13 did not result in pregnancy while 19 births were recorded. Caesarean delivery was necessary in 9 cases and another 3 required manual assistance due to the large size of the kid. The level of reproductive isolation between the Siberian and the European roe deer is clearly de- monstrated by the fact that 10 subsequent matings between two Fl hybrid bucks and a group of hybrid does did not produce a sin- gle offspring. Indeed, it seems that many hy- brid bucks are sterile, however, back-crosses between hybrid does and pure bucks of either form. did produce viable offspring. Si- milarly, SOKOLOV and GRoMov (1985) found Specific status of roe deer 17 Table 1. Some outcomes of experimental hybridization of Siberian and European roe deer (adapted from Da- NILKIN and Hewıson 1996) Cross! Normal Delivery’ Successful Mating“ Female x Male Sib. x Eur. Eur. x Sib. EISGEUR F1x Sib. EuroxaEl Sib> --., Pm are the frequencies of the m phens in the variability of the i-non-metric parameter for the first sample (p;< 1), and qı, G25 2, Qmaaressihe frequeneieszof thezsame m phens in the variability of the i-non-metric parameter for the second sample (q; < 1). If the samples are compared by k non-metric para- meters, then r is calculated as: r=(1+nT+...+T%)JK. The identity criterion, as a tool for evaluation of significance of phenetic distances, has been de- fined as follows: I=8nıns(1-r-(po + go)/A)/nı + 3, where nı and n; are the sizes of the samples com- pared; po is the sum of frequencies of phens that are presented in the first sample but not in the second one, gu — accordingly, is the sum of fre- quencies of phens that are presented in the sec- ond sample but not presented in the first one. The identity criterion I is distributed as the well known x-square criterion with the degrees of free- dom df=m-]1. By involving k non-metric para- meters for the pairwise comparison of samples, I has been defined as: I=hı, +b,+...+ 1, with the degrees of freedom calculated as d=mı +m»+...+mx -K. Twenty-two non-metric skull parameters were used for the phenetic study of the American mink. Their variability gives 80 variants i.e. phens as typical states of the non-metric skull para- meters (Figs.2, 3). Number, presence/absence, shape and location of foramına on a particular bone and other bony structures were the basic ca- tegories of these non-metric skull variables. In case of bilateral parameters, only the right side of the cranıum was taken into account. In total, 80 phens were revealed (Fig. 3) using the following non-metric skull parameters: 1. Shape of the foramen infraorbitale (front view of the skull): 1.1 - oval; 1.2 - side bend; 1.3 -— bottom bend; 1.4 - triangle; 2. Foramen occipitale superior (back view): 2.1 — one foramen directly below crista occipitalis; 2.2 — absent; 2.3 — one foramen located be- tween crista occipitalis and foramen magnum; 2.4 - one foramen directly above crista occipi- talis; 2.5 - two foramina located separately hor- izontally directly below crista occipitalis; 2.6 — two foramina located separately, one - directly below crista occipitalis, another — directly above foramen magnum; 2.7 - three and more foramina directly below crista occipitalis; Fig. 2. Location of the non-metric parameters on an American mink skull. Specificity of non-metric parameters of American mink 39 Be Peer Tee Te SERFIE SE AIE ee ee - aEIRIKSKAESESES 87 | sı ) 92 | DERBBSBMERK BE nn VE re u; [N 0 Spereg: 2 103) 201 | 202 | 2 Fig. 3. Variability of non-metric parameters of American mink skulls in Belarus. fio - foramen infraorbitale, fm - foramen magnum, cd - condylus occipitalis, bt - bulla tympanicum, m" - first upper molar, orb - orbita, mae - meatus acusticus externus, m, - first lower molar. 3. Profile shape of the processus jugularis ossis 8. Foramen on the processus postorbitalis ossis OC- oceipitalis (side view): 3.1 — proximal bend; cipitalis (side view): 8.1 - one foramen located 3.2 — lower angle turned up; 3.3 - straight; in front of the processus; 8.2 — two foramina, 3.4 -— upper angle turned down; one - ın front , another - at the back of the pro- 4. Shape of the bony micro eminencies (esti- cessus; 8.3 — two foramina, one - in front, an- mated by rubbing with an aluminium ruler) other — oblong - at the back of the processus; on the os sphenoidale in front of the bulla 8.4 — foramen absent; 8.5 — one foramen at the tympani (bottom view): 4.1 — V-shaped; 4.2 — back of the processus; 8.6 — two foramina, both bow-shaped; 4.3 — U-shaped; 4.4 — crown- located in front of the processus; 8.7 - one fora- shaped; 4.5 — dash-shaped; men located on top of the processus; 5. Shape of the bony micro eminencies (esti- 9. Foramen temporale (side view): 9.1 — one mated by rubbing with an aluminium ruler) foramen; 9.2 - two foramina, the front one is on the os sphenoidale between the bulla tym- significantly smaller: 9.3 - two foramina of pani (bottom view): Sa y-shaped; Sr Ve the same sıze:; 9.4 - foramen absent: row backwards; 5.3 — arrow forwards; 5.4 - 10. Foramen canalis condylaris (back view): 10.1 V-shaped; - one foramen; 10.2 - two foramina; 10.3 — 6. Number of the foramina palatinum posterior one foramen with a rudiment of the horizon- (bottom view): 6.1 - one; 6.2 - two; 6.3 - three tal partition; 10.4 -— one foramen with a rudi- and more foramina; ment of the vertical partition; 7. Number of small foramina located above the 11. Presence or absence of the foramen hypoglos- foramen infraorbitalis (front view): 7.1 — one; sus (bottom view): 11.1 — present; 11.2 — ab- 7.2 — absent; 7.3 - iwo foramina; sent; 40 A. ULEVICIUS et al. 12. Number of additional foramina located in front of the foramen incisivum (bottom view): 12.1 — one foramen; 12.2 -— absent; 12.3 - two forami- na; 12.4- three and more small foramina; 13. Number of the foramina mandibulae (front view): 13.1- one foramen; 13.2 - two foramina; 13.3- three and more foramina; 13.4 — absent. 14. Foramen located on the os sphenoidale in front of the bulla tympani (bottom view): 14.1 - one foramen; 14.2 — absent; 14.3 — two foramina; 15. Number of the foramina opticum (side view): 15.1 - one foramen; 15.2 - two foramina (par- tition is a little bit deeper); 16. Shape of the processus coronoideus mandibu- lae (side view): 16.1 — pyramid-shaped with oval apex; 16.2 — with angular hinder margin; 16.3 — with acute and turned back apex; 16.4 —- pyramid-shaped with cut off apex; 17. Position of front margin of the fossa masseter- ica mandibulae with respect to the hinder margin of the M} tooth (side view): 17.1 - in front; 17.2 - on the same level; 17.3 — behind; 18. Shape of the foramen magnum (back view): 18.1 — round-shaped; 18.2 — pyramıd-shaped; 18.3 — pear-shaped; 19. Shape of bony vault above the foramen mag- num (top view): 19.1 - straight; 19.2 — with two eminencies; 19.3 — with three eminencies; 20. Presence or absence of foramen located be- tween the condylus occipitalis (bottom view): 20.1 — present; 20.2 — absent; 21. Position of additional foramen located behind the foramina incisivum with respect to the hinder margin of these (bottom view): 21.1 — on the same level; 21.2 - in front; 21.3 — be- hind; 22. Shape of bend between the condylus occipita- lis (bottom view): 22.1 — V-shaped; 22.2 — with eminencies on both sides. Skull non-metric variability related to sex was tested using Zhivotovsky’s test, and sex-depen- dent parameters were excluded from further ana- lysis. Out of 22non-metric parameters of the American mink’s skull, only 5 were significantly related to sex (Tab. 2). These were not used in the analysis below. The effect of age was not tested. Distinct changes in size and proportion of mink skulls occur during the first year of life (KruskA 1979), thus only skulls belonging to adult American mink aged one year and older Table 2. Differences between sexes (Zhivotovsky’s test) according to the skull non-metric parameters of Ameri- can mink in Belarus Non-metric n of males n of females r parameter oo wmrHr * differences between sexes are statistically significant 3 6 3 4 3 2 2 6 3 3 1 3 3 2 1 3 1 2 2 1 2 1 Specificity of non-metric parameters of American mink 41 (1+) were used for this study. All skulls having closed sutures (STUBBE 1973 for review) were ad- ditionally tested for age using histological sections of the canine teeth (KLEvEsAaL and KLEINENBERG 1969). Results Differences between founders and current local population fragments Significant differences were found between the founder sample and all other samples derived from current local population frag- ments of naturalised American mink (Tab. 3). Especially substantial differences Table 3. The skull non-metric differences (by complex of all non-metric parameters, df=39) between the founder and the present local population fragments of the American mink, Belarus Founders compared r with the sample number: PrrroeH NWwvrocosnsoauRhNwm hr were established by comparing the founder sample with the sample 12 from Volka sam- pling area, Neman river catchment (a :0377:=10-12036,p2 20.000) and. with the sample 1 from Ushacha sampling area, Western Dvina river catchment (r = 0.793; I = 128.71; p = 0.000). Lower but significant phenetic differences were discovered be- tween founders and the sample 14 from Svi- sloch sampling area, Dnepr river catchment (0.9391 377703:5p 0.000) Differences among local population fragments There were no significant differences in mean level of phenetic similarıty by analys- ing both groups of samples (Tab. 4). Mean index of similarity was only slightly higher among local population fragments inhabit- ing the same river basin (0.928 vs. 0.912; p = 0.128). Also, there was no difference in rate of significantly dissimilar pairs of sam- ples. Approximately one half of the pair- wise comparisons exhibited statistically sig- nificant non-metric differences in both groups of samples (45% vs. 55%, p = 0.5). However, by comparing local samples be- longing to the same basin of a large river, a significant negative correlation between the index of phenetic similarity and spatial distance between two samples was found (coefficient of correlation, r = -0.77,n = 20, p = 0.000; Fig. 4). This correlation was very low and not significant (coefficient of corre- lation, r=-0.24, n= 71, p= 0.842) when samples belonging to the different basıns of large rivers were analysed. Table 4. Phenetic differences between pairwise compared local samples of American mink from the same (A) and the different (B) basins of large rivers in Belarus Indicator A (n = 20 pairs of samples) Rate of pairs of samples with 45 significant difference, % Mean r+SD Mean I+SD 0.923 # 0.0324 64.51 + 15.680 B Significance of (n= 71 pairs of samples) difference, p 55 0.912 # 0.0430 7\62055333.7103 42 A. ULEVICIUS et al. A Coeff. of correlation, r = -0.77; n=20; p=0.000 0.98 0.96 0.94 0.92 0.90 0.88 0.86 0.84 0.82 Zhivotovsky's index of similarity, r 0 40 80 120 160 200 240 Distance between samples, km B Coeff. of correlation, r = -0.02; n=71; p=0.842 1.00 Ä Duke) « O0 0 oo. . .. 0. 01000 D ee ie er er are AO N 00005000 -50 50 150 250 350 450 550 Distance between samples, km Fig. 4. Correlation between the phenetic similarity and spatial distance among local population fragments from the same (A) and different (B) large river basins (geographical populations) in Belarus. Specificity of non-metric parameters of American mink 43 Table 5. The non-metric skull differences (r; I; p, by complex of all non-metric parameters, df = 39) among pair- wise compared geographical populations of American mink inhabiting basins of large rivers in Belarus Basins Lovat (n = 79) 0.976; 150.8; 0.000 W. Dvina (n = 222) Neman + Dnepr (n = 100) Differences among geographical populations The non-metric differences were statisti- cally significant, and sufficiently high to be characterised by substantial non-metric spe- cificity between all the 4 geographical po- pulations inhabiting the basins of the large rıvers — Lovat, W. Dvina, Neman-Dnepr, and Pripjat (Tab. 5). The greatest difference was found between populations from the W. Dvina and Pripjat river basins (7 0.955.217 1182; p< 0.001). These river basins are characterised by considerably different habitat conditions for American mink. Discussion Substantial intraspecific phenetic differen- tiation was found among American mink within the fairly small area of Belarus (204000 km’). After approximately 30- 40 years, established local populations were also markedly different in comparison with the founder population. This suggests a high level of adaptability of this naturalised pre- dator to new habitat conditions. High levels of non-metric plasticity could be one of the basic factors which enabled the American mink to adapt to different ecological condi- tions and spread throughout Europe (PAv- Lov et al. 1974; GERELL 1967 a,b, 1968; DA- nıLov and TumAnov 1976; DunsTonE 1993; SiDorovicH 1993, 1997). In domesticated American mink popula- tions, the absence of strict natural selection as well as deliberate artificial selection could lead to a certain partial “packing up” of the gene pool. In several European coun- tries significantly lower levels of sexual di- morphism were found in domestic Ameri- Neman + Dnepr 0.977; 102.3; 0.000 0.990; 68,8; 0.020 Pripjat (n = 28) 0.963; 71,2; 0.010 0.955; 118.2; 0.000 0.968; 72.3; 0.010 can mink in comparison with the feral ones (Lynch and Haven 1995). It was inter- preted as weak sexual selection, absence of competition, and purposeful artificial selec- tion for larger specimens of both sexes. De- crease in size of brain and some other or- gans in the domesticated American mink may result from reductions of central ner- vous and circulatory functions in the do- mesticated organism (KruskA 1996; KRUS- KA and SCHREIBER 1999). More diverse selection started when domes- tic minks were placed in completely differ- ent feral conditions. The gene pool of the newiy formed populations of American mink was affected by different pressures of natural selection in comparison with ranch conditions. Consequently, the phenetic structure of these populations should change. This could explain our finding that the founder population differs substantially phenetically from all the current local po- pulations. Non-metric skull differences among con- temporary local population fragments of American mink might also be affected by stochastic changes in frequencies of variants of non-metric parameters in small spatial groups of individuals (processes of genetic drift: the bottleneck, founder effect), espe- cially when small samples from different river catchments were compared. Spatially remote local groups of individuals belong- ing to different geographical populations might be phenetically similar, whereas the neighbouring ones could be phenetically very different. Absence of a correlation be- tween non-metric and spatial distances de- monstrate that phenetic relations between population fragments from the different geographical populations are rather sto- chastic. Such stochastic differentiation has 44 A. ULEVICIUS et al. been reported for many other species (e. g. GREWwAL and DascuptTA 1967; MCLELLAN and FiINNEGAN 1990; KoZAkKIEWIcCZ and KA- NOPKA 1991; LoRENZINI et al. 1993; Ryan et al. 1996). However, presence of a significant correla- tion between the non-metric differences and the spatial distances among samples from the same geographical population sug- gests another interpretation. This finding demonstrates a certain regularity in the in- trapopulational non-metric (possibly also genetic) divergence rather than the pre- sence of a stochastic factor. Our results also demonstrate that all geographical popula- tions of American mink (inhabiting catch- ments of large rivers) were phenetically specific, thus having their own “general” vector of selection. Different selective pres- sures within one geographical population likely result in formation of phenetically different groups of individuals. The rate of gene flow between such groups would de- pend on the degree of spatial isolation. Spa- tial distances among intraspecific groups of- ten correlate with phenetic or genetic differences (REES 1969; McLELLANn and FINNEGAN 1990; Urevicıus 1992). Thus, spa- tial isolation can influence genetic and phe- netic structures. The social intraspecific structure can lead to considerable genetic differentiation of adja- cent social groups, too. It has been estab- lished in primates (SCHEFFRAHN et al. 1996). Both genetic and environmental factors might be important for the control of phe- netic variability (e.g. PETRAS 1967; Howe and PArsons 1967; BERRY and BERRY 1972; see also HArTL et al. 1993b, for review). Some authors have argued that the genetic variation explains more than 50% of phe- notypic variation (SOULE and ZEGERS 1996). A significant part of phenetic varia- tion can be influenced by phenotypic plas- ticity as a function of the environment. Ge- netically, plasticity is likely due to both differences in allelic expression across en- vironments, and changes in interactions among loci (SCHEINER 1993). Results of our study might be interpreted in connection with a very high phenetic plasti- city of American mink occupying new and diverse habitat conditions. Other ecological characteristics of naturalised American mink populations in our study area con- firmed the distinct ecological plasticity of this species (Sıporovich 1993, 1997). It should be emphasised that our data are not in accord with the data from some other po- pulations of American mink. For example, American mink from Norway exhibited re- latively little geographic variation in either the metrical measurements or the non-me- trical traits thus indicating little genetic var- iation (WırG and LıE 1979). Electrophoreti- cal investigations on wild and ranch mink from Canada and Germany, respectively, showed low protein heterozygosity in both groups (KrUSKA and SCHREIBER 1999). These authors also reviewed works of other investigators showing low allozyme hetero- zygosity of mustelids. In this respect an ex- planation of phenetic differentiation of American mink in Belarus due to the phe- netic plasticity would also be reasonable be- cause the phenetic expression of mono- morphic locı may be unequal in different environments. For a more detailed study of mechanisms of the non-metric differentiation of American mink in Belarus biochemical-genetic inves- tigations are needed. However, the pre- sently discovered substantial non-metric differences in temporal and geographical scales show that an influence of genetic fac- tor was very important. Acknowledgements We are very thankful to Dr. F. TATTERSALL for English correction of the manuscript and K.-A. NITscHE for translation into German. Specificity of non-metric parameters of American mink 45 Zusammenfassung Spezifität nicht-metrischer Parameter von Mink-Populationen (Mustela vison) im Verhältnis zu Habitat-Unterschieden in Weißrußland Insgesamt 418 Schädel des Mink, Mustela vison, wurden durch eine nicht-metrische (phänetische) Studie geprüft um die Spezifität der Einzelheiten von 15 auseinander liegenden und für 4 geogra- phische Populationen der großen Flußbecken in Weißrußland zu untersuchen. Verglichen wurden auch 14 Proben von den derzeitigen Populations-Fragmenten mit einer Probe der regionalen Grün- derpopulation dieser Art. Die Distanzen zwischen den Proben wurden unter Benutzung von 22 nicht-metrischen Merkmalen abgeschätzt. Ein hohes Niveau phänetischer Divergenz wurde in den natürlichen Mink-Populationen von Weilsrußland festgestellt. Die Gründertiere zeigten bedeu- tende phänetische Unterschiede, die mit jedem der 14 entfernten Populations-Fragmenten vergli- chen wurden. Beträchtliche phänetische Unterschiede werden in der Hälfte des paarweisen Ver- gleiches zwischen entfernten Populations-Fragmenten gezeigt. Es gab außerdem eine signifikante negative Korrelation zwischen der phänetischen Ähnlichkeit und räumlicher Entfernung unter den Populations-Fragmenten an einem einzelnen Flußbecken. Es gab keine derartige Korrelation unter Populations-Fragmenten von unterschiedlichen Flußbecken. Statistisch bedeutungsvoll waren die phänetischen Distanzen zwischen allen 4 geographischen Populationen, welche die großen Flußbecken bewohnen. Die nicht-metrische Differenzierung bei den natürlich lebenden Tieren wird unter dem Aspekt der vielfältigen Habitat-Bedingungen, in der die Population vorkommt, bespro- chen. Die in unserer Studie dargestellte phänetische Plastizität des Mink (welche die genetische Plastizität kennzeichnet) ist eine Anpassung, die über den hohen demographischen Erfolg dieser freilebenden Art entscheidet, was in vielen Regionen Europas und Asiens gezeigt werden kann. References BERRY, A. C.; BERRY, R. J. (1972): Epigenetic poly- morphism in the primate skeleton. In: Com- parative Genetics in Monkeys, Apes and Man. Ed. by A.B. CHIArRELLI. New York: Academic Press. Pp. 13-41. BoRISov, V.I.; BARANOoV, A.S.; VALETSKY, A. V.; ZAKHAROV, V. M. (1997): Developmental stabi- lity of the mink Mustela vison under the im- pact of PCB. Acta Theriol. 42, Suppl. 4, 17-26. CHANIN, P.R.F.; Linn, 1. J. (1980): The diet of the feral mink (Mustela vison) in south-west Brit- ain. J. Zool. (London) 192, 205-223. DANILov, P. I.; TUMANov, I. L. (1976): [Mustelids of north-eastern USSR.] Leningrad: Nauka publ. [In Russian] DunsTone,N. (1993): The Mink. London: T and AD Posyer Ltd. DunsTOone, N.; BiIRKs, J. D. S. (1987): The feeding ecology of mink (Mustela vison) in a coastal habitat. J. Zool. (London) 212, 69-83. GERELL, R. (1967 a): Dispersal and acclimatiza- tion of the Mink (Mustela vison Schreb.) in Sweden. Viltrevy 5, 3-38. GERELL, R. (1967 b): Food selection in relation to habitat in mink (Mustela vison Schreber) in Sweden. Oikos 18, 233-246. GERELL, R. (1968): Population studies on the mink (Mustela vison Schreber), in Sweden. Viltrevy 8, 83-114. GREWAL, M.S.; DaAasGuPpTA,S. (1967): Skeletal polymorphism and genetic drift in a Delhi Frog, Rana cyanophlictis. Genet. Res. 9, 299- 307. HARTL, G. B.; MARKOWSKI, J.; PUCEK, Z. (1993 a): The Meeting “Ecological Genetics in Mam- mals — Current Research and Future Perspec- tives”. Foreword. Acta Theriol. 38, Suppl. 2, 5-0. HARTL, G. B.; SUCHENTRUNK, F.; NADLINGER, K.; WirLing, R. (1993b): An integrative analysis of genetic differentiation in the brown hare Lepus europaeus based on morphology, allo- zymes, and mitochondrial DNA. Acta Ther- iol. 38, Suppl. 2, 33-57. Howe, W.L.; Parsons, P. A. (1967): Genotype and environment in the determination of min- or skeletal variants and body weight in mice. J. Embryol. Exp. Morph. 17, 283-292. KLEVESAL, G. A.; KLEINENBERG, S. E. (1969): Age determination of mammals by layered struc- tures of teeth and bones. Jerusalem: Israel Progr. Sci. transl. 46 A. ULEVICIUS et al. KOZAKIEWICZ, M.; KAnoPkA,J. (1991): Effect of habitat isolation on genetic divergence of bank vole populations. Acta Theriol. 36, 363-367. Kruska, D. (1979): Vergleichende Untersuchun- gen an den Schädeln von subadulten and adulten Farmnerzen, Mustela vison f. dom. (Mustelidae; Carnivora). Z. Säugetierkunde 44, 360-375. KRrUSKA, D. (1996): The effect of domestication on brain size and composition in the mink (Mus- tela vison). J. Zool. (London) 239, 645-661. KRrUSKA, D.; SCHREIBER, A. (1999): Comparative morphometrical and biochemical-genetic in- vestigations in wild and ranch mink (Mustela vison: Carnivora: Mammalia). Acta Theriol. 44, 377-3932. LORENZINI, R.; PALATANO, M.; APOLONIO, M.; MAZZARONE, V. (1993): Genetic variability of roe deer Capreolus capreolus in Italy: electro- phoretic survey on populations of different origin. Acta Theriol. 38, Suppl. 2, 141-151. LynchH, J. M.; HAYDEN, T. J. (1995): Genetic influ- ence on cranial form: Variation among ranch and feral American mink Mustela vison (Mammalia: Mustelidae). Biol. J. Linn. Soc. 55, 293-307. MARKOWSKI, J.; MARKOWSKA, M. (1988). Non-me- trical variation in three populations of roe deer. Acta Theriol. 33, 519-536. MCLELLAN, L. J.; FINNEGAN, M. (1990): Geo- graphic variation, asymmetry, and sexual di- morphism of nonmetric characters in the deer mouse (Peromyscus maniculatus). J. Mammal- ogy 71, 524-533. PAvLov, M. P.; KoRSAKOVA, 1. B.; TIMOFEEV, V. V.; SAFONOV, V.G. (1974): [Acclimatization of game mammals and birds in USSR.] Kirov: Kirov Dep. Volgo-Vyatsk. Book Publ. [In Russian] PETRAs, M.L. (1967): Studies of natural popula- tions of Mus. IV. Skeletal variations. Can. ]J. Genet. Cytol. 4, 575-588. Rees, J. W. (1969): Morphological variation in the cranıum and mandible of the white tailed deer (Odocoileus virginianus): acomparative study of geographical and four biological distances. J. Morph. 128, 95-112. RUIZ-GARCIA, M. (1998): Genetic structure of dif- ferent populations of domestic cat in Spain, Italy, and Argentina at a micro-geographic le- vel. Acta Theriol. 43, 39-66. Ryan, A.; Duke, E.; FAIRLEY, J.S. (1996): Mito- chondrial DNA in bank voles Clethrionomus glareolus in Ireland: evidence for a small founder population and localızed founder ef- fects. Acta Theriol. 41, 45-50. SCHEFFRAHN, W.; DE RUITER, J. R.; Van Hoff, - J. A.R.A.M. (1996): Genetic relatedness within and between populations of Macaca fascicularis on Sumatra and of-shore islands. In: Evolution and Ecology of Macaque Socie- ties. Ed. by J. E. FA and D. G. LinDBURG. Cam- bridge: Univ. Press. Pp. 20-42. SCHEINER, S. M. (1993): Genetics and evolution of phenotypic plasticity. Annual Review of Ecol- ogy and Systematics 24, 35-68. SCRIBNER, K. T. (1993): Conservation genetics of managed ungulate populations. Acta Theriol. 38, Suppl. 2, 89-101. SIDOROVICH, V. E. (1993): Reproductive plasticity of the American mink (Mustela vison) in Be- larus. Acta Theriol. 38, 175-183. SIDOROVICH, V. E. (1995): [Minks, otter, weasel, and other mustelids]. Minsk: Uradzhai publ. [In Russian] SIDOROVICH, V. E. (1997): Mustelids in Belarus. Evolutionary Ecology, Demography and In- terspecific Relationships. Minsk: Zolotoy uley publ. [In Russian with English summary] SıövoLD, T. (1977): Non-metrical divergence be- tween skeletal populations. Ossa 4, Suppl. 1, 1-133. SMITH, M. F. (1981): Relationships between genet- ic variability and niche dimensions among co- existing species of Peromyscus. J. Mammalogy 62, 273-285. SOULE, M. E.; ZEGERS, G. P. (1996): Phenetics of natural populations. V. Genetic correlates of phenotypic variation in the pocket gopher (Thomomys bottae) in California. J. Heredity 87, 341-350. STUBBE, M. (1973): Buch der Hege. Band 1: Haar- wilde Berlin: VEB Deutscher Land- wirtschaftsverlag. TERNOVSKY, D. V. (1977): [Biology of mustelids (Mustelidae).] Novosibirsk: Nauka publ. [In Russian] TERNOVSKY, D. V.; TERNOVSKAJA, Y.G. (1994): [Ecology of mustelids.] Novosibirsk: Nauka publ. [In Russian] ULeviıcıus, A. (1992): Phenetic structure of the beaver Castor fiber population in Lithuania. Acta et Commentationes Universitatis Tar- tuensis 955, 120-132. Wir, ©.; LiE, R. W. (1979): Metrical and non-me- trical skull variations in Norwegian Wild Mink (Mustela vison Schreber). Zool. Scr. 8, 297-300. ZHIVOTOVsKY, L. A. (1979): Population similarity measure for polymorphic characters. J. Gener- al Biol. (Moscow) 40, 587-602. [In Russian with English summary] Specificity of non-metric parameters of American mink 47 YABLOKOV, A. V.;; PERRIN, W. F.; MiınA, M. V. (1983): Evaluation of phenetic relations among groups of dolphins using analysis of non-metric cranial variation. Zool. Zhurn. 62, 1887-1897. [In Russian with English sum- mary] Authors’ addresses: A. ULrevicıus, Institute of Ecology, Akademijos 2, LI-2600 Vilnius, Lithuania (e-mail: aliusu@takas.lt), V.Sıporovich and G. LAUZHEL, Institute of Zoology, National Acad- emy of Sciences of Belarus, Akademicheskaja Str. 27, Minsk-220072, Belarus Mamm. biol. 66 (2001) 48-59 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Original investigation Little allozyme and mtDNA variability in brown hares (Lepus europaeus) from New Zealand and Britain - A legacy of bottlenecks? By F. SUCHENTRUNK, CLAUDIA JASCHKE, and ANITA HAIDEN Research Institute of Wildlife Ecology, Vienna Veterinary University, Vienna, Austria Reiceipt of Ms. 07. 02. 2000 Acceptance of Ms. 07. 03. 2000 Abstract We studied cross nuclear and mitochondrial gene pools of brown hares (Lepus europaeus) from three local populations in Britain and two in New Zealand, to test the hypothesis of reduced genet- ic variability in hares from New Zealand resulting from few founders originating from Britain. Multi- locus allozyme electrophoresis of 52 protein loci and analysis of restriction fragment length poly- morphisms of total mitochondrial DNA based on 16 hexanucleotid-cleaving restriction enzymes were employed in 119 and 36 hares, respectively. Observed and expected average heterozygosities, rates of polymorphism, average numbers of alleles per locus, Shannon-Weaver information indices of allelic diversity, as well as values of haplotype and nucleotide diversity were similar in all regio- nal samples. But hares from both New Zealand and Britain had significantly lower genetic diversity than brown hares from continental Europe studied earlier. Thus, gene pool erosion likely occurred already in British hares, perhaps associated with their probable introduction in Roman times. The- oretically, the small number of alleles found in British brown hares could have been sampled by the few hares that were reported as having constituted the founder stock in New Zealand in the nineteenth century. As expected, rare alleles of British brown hares were absent in New Zealand. But drift had only a slight effect on the gene pool composition of hares in New Zealand. Key words: Lepus europaeus, allozymes, mtDNA, genetic bottleneck Introduction In New Zealand, brown hares (Lepus euro- paeus) have higher rates of ovarian tumors and missing posterior upper molars (M’°) than in Europe (Frux 1965, 1980; PARKES 1988; SUCHENTRUNK et al. 1992). This might result from low genetic variability as a con- sequence of a small number of founder in- dividuals (Frux 1965). Historical docu- ments suggest that brown hares released in New Zealand in the 19" century by various Acclımatization Societies were mostly ta- 1616-5047/01/66/01-048 $ 15.00/0. ken from Phillip Island, Victoria, Australia (LEvER 1985; FLux 1990). There, only six hares had built up a population of 200 indi- viduals by 1865, few years after introduc- tion (FLux 1990; FLux et al. 1990; see also Rorıs 1969 and LEvER 1985). In Australia, brown hares were probably first success- fully introduced in 1859 by W.IyArr on the shores of Western Port Bay, Victoria (LEVER 1985), and afterwards on Phillip Is- land (MAHoop 1983). In February 1864 an- Little allozyme and mtDNA variability in brown hares 49 other nine hares were released into an en- closure near Geelong, Victoria by T. Aus- TIN, who imported them from England (cf. LEVER 1985). All Australian hares are con- sidered originating from Britain (cf., FLux 1990), but no details as to specific regions are given in the available literature (LEVER 1985; FLux pers. com.). The exact numbers of hares that have successfully bred after their naturalization in Australia and New Zealand remain unknown. However, the list of importations to New Zealand pre- sented by LEvEr (1985) suggests limited ge- netic variability in the founder gene pool (see also FLux 1990). In this study we compared levels of genetic variability of hares from New Zealand and Britain, to test this genetic bottleneck hy- pothesis (FLux 1965). Theory and empirical findings (e. g., FUERST and MARUYAMA 1986; LEBERG 1992; HARTL and PuUcEK 1994; TIE- DEMANN et al. 1997) predict a smaller effect of bottlenecks on multi-locus allozyme het- erozygosity than on other indicators of allo- zymic variability, such as the rate of poly- morphism (P), and mean number of alleles per locus (A). Allozyme heterozygosity may even increase after bottlenecks (e. g., LEBERG 1992). Thus, we expected lower P- and A-values for hares from New Zealand than for British brown hares, whereas het- erozygosities might be similar. Particularly alleles with low frequencies in British brown hares might have not been sampled by the few founders in New Zealand. Drift effects could have caused shifts in allele fre- quencies and consequently increased genet- ic divergence between hares from Britain and New Zealand. We also expected a pro- nounced reduction of variability in the mi- tochondrial DNA (mtDNA), because of the lack of recombination in this maternally inhereted genome in post-bottleneck popu- lations (e.g., GILEs et al. 1980; LANSMAN et al. 1981; AvısE 1994; AvısE and HAMRICK 1996; see e.8., GYLLENSTEN et al. 1991 for paternal inheritance of mtDNA). Brown hares from the British Isles are convention- ally considered a separate subspecies (ZL. e. occidentalis DE WiINnTon, 1898; cf. CORBET and SOUTHERN 1977, ArnoLD 1993). They might be genetically somewhat distinct from mainland European brown hares. Therefore, we compared the present data with adjusted data sets of continental Eur- opean brown hares published earlier (HArTL et al.1993; SUCHENTRUNK et al. 2000). Material and methods Specimens studied We studied 119 hares from two regions in New Zealand and three in Britain (Fig. 1). In New Zealand, hares were collected in the Wairarapa region (n = 32) of the North Island in September/ October 1993FbyArEICHErux?(CEandeare Re- search”, Lower Hutt), and in the Harper/Avoca catchment (n = 28) on the South Island in Octo- ber 1993 and March 1995 by J. PArKES (“Landcare Research”, Christchurch) and F. SUCHENTRUNK. In Britain, collections were organized by S. TAPPER (The Game Concervancy Trust, Fordingbridge, England) in February 1995 in three regions (Wilt- shire, southern England, n = 20; Loddington, Lei- cestershire, central England, n = 19; Duns, Aber- deenshire, Scotland, n=20). Most hares were dissected by one of the authors (FS). They were sexed by inspection of the internal reproductive organs and aged by checking the lateral epiphy- sial protrusion of the ulna (STROH’s sign), which separates young of the year (4 kb long frag- ments, electrophoresis was continued for two ad- ditional hours and fragment measurements were repeated. Fragment length variants <0.4 kb could not be detected by our procedure. Fragment lengths were determined by comparing with Lambda phage DNA digested with HindIlI and a 100 bp-ladder (1500-500 bp range). We com- pared fragments and cleaveage sites deduced from enzyme-specific fragment patterns with those already found in brown hares from Austria (HARTL et al. 1993; NADLINGER 1994). Because our enzyme set was identical with that used al- ready for Austrian brown hares, we could com- pare our restriction morphs, haplotypes, and in- dices of mtDNA variability directly with brown hare mtDNA data published by these authors (HARTL et al. 1993). We calculated haplotype di- versity (h) and nucleotide diversity (z) (NEI and Lı 1979; Neı 1987) to describe RS-specific gene pool variability. We calculated pairwise net nu- cleotide diversity between RSs based on cleave- age site variations (e.g., AvısE 1994) to obtain es- timates of mtDNA differentiation among RSs. We compared frequencies of the standard haplo- type (i.e., the by far most common type I; HARTL et al. 1993) and other haplotypes (i.e., all others aggregated) between NZ and UK hares by a one-tailed exact Fisher test (hypothesizing lower variability in NZ than UK hares). Finally, we compared RS-specific h- and r-values of NZ and UK hares with the respective values of 18 Aus- trian brown hare samples (HarTL et al. 1993) by MANN-WHITNEY U-tests, basing significance deci- sions on Sequential Bonferroni procedure. Results We found di-allelic variation at six loci (Tab. 1) and a significant excess of homozy- gotes at the Es-1 locus in the NZ-W sample. Except for Acp-1°”° and Mpi”,, all alleles were found earlier in brown hares from central Europe. Allele frequencies did not depend on age category or sex. The indices of genetic variability (H., H., P, A, H’) are listed in table 1. Overall and locus-specific inbreeding coefficients (Fis) are given in ta- ble 2, separately for each RS. Ner’s (1978) unbiased D values, ROGERS’ modified dis- tances, and the fixation coefficients (Fsr) appear in table3, along with the signifi- cances of pairwise comparisons of allele fre- quencies at one or more loci. In table 4, the indices of allozymic variability (based on 49 loci) of the UK and NZ samples are compared to those of 20 Austrian and eight Bulgarian regional samples of brown hares studied earlier (HArTL et al.1993, SUCHEN- TRUNK et al. 2000). Thirtyfive (97.2%) hares had the standard mtDNA haplotype I, that was already 52 F. SUCHENTRUNK et al. Table 1. Allele frequencies at polymorphic loci, sample-specific indices of genetic variability, and mean inbreed- ing coefficients (Fis) in brown hares from New Zealand (NZ) and Britain (UK); for acronyms see figure 1. Allele designations correspond to those given in HARTL et al. (1989, 1993), SUCHENTRUNK (1993), SUCHENTRUNK et al. (1998), and SUCHENTRUNK et al. (2000). n = number of individuals screened for each locus and regional sample. In- dices of gene pool variability (based on 52 loci): H, = observed average heterozygosity, H. = expected average heterozygosity, P(oo%) = rate of polymorphism (99% criterion), A= average number of alleles per locus, H’ = Shannon Weaver information index. Locus Allele P (99%) A H’ Table 2. Locus-specifc (unbiased) heterozygosities in % (upper values) and inbreeding coefficients (lower va- lues) as well as overall inbreeding coefficients (Fıs) for regional samples of brown hares from New Zealand (NZ) and Britain (UK). 31.0 -0.231 Es-1 34.7 +0.0451 Mpi 11.9 -0.067 overall Fis +0.095 Little allozyme and mtDNA variability in brown hares 53 found in the majority of central European brown hares (HaRrTL et al. 1993), and only one hare of the NZ-H sample had a new haplotype. This haplotype deviated from haplotype I by only one additional cleave- age site, produced by Xbal at position 14.8 kb of the restriction map published in HARTL, et al. (1993). Values of haplotype (h) and nucleotide diversity (z) were zero for all RSs except for the NZ-H sample; that had an h-value of 28.57% and a z-value of 0.049%. Haplotype frequencies did not differ significantly between the RSs from Britain and New Zealand. Also, RS-specific h- and x-values of the samples from Britain and New Zealand were not significantly lower (p>0.05, one-tailed Mann-Whitney U-tests) than in brown hares from 20 Aus- trian localities (cf. HArTL et al. 1993). Va- lues of pairwise net nucleotide diversities Table 3. Matrix of pairwise genetic distances and fixation indices (Fsr) among regional samples of brown hares from New Zealand and Britain. Ner's (1978) unbiased D values (first row) and modified RoGers’ distances (WRIGHT 1978) (second row) above the diagonal and Fsr values below. Fsr values were considered differing significantly from zero with significant allele frequencies for at least one locus in pairwise comparisons (significance based on Sequential Bonferroni procedure; nominal & = 0.05, 45 tests). Significance (sig) or no significance (n. s.) is indi- cated below each Fs7 value. For acronyms of regional samples see figure 1. regional samples NZ-W (1) NZ-H (2) UK-W (3) UK-L (4) UK-D (5) Table 4. Comparison of indices of allozymic and mtDNA variability of regional samples of brown hares from New Zealand, Britain, Austria, and Bulgaria. Allozyme data are based on 49 loci and mtDNA RFLP data on 16 restric- tion endonucleases (see Material and methods). Means and range (in parentheses) of observed (H,) and expected (H.) average heterozygosity, Shannon-Weaver diversity index (H’), rate of polymorphism (P - 99% criterion), average number of alleles per locus (A), haplotype (h) and nucleotide diversity (x) are given for each group of local samples. sig = significance as determined by Mann Whitney tests (d.f. = 1) and sequential Bonferroni pro- cedure (@ = 0.05), n. s.= not significant. 2 NZ and 3 UK regional samples (this study) 20 Austrian (HARTL et al. 1993) and 8 Bulgarian (SUCHENTRUNK et al. 2000) regional samples 1.26 (0.9-1.7) H 2.83 (1.7-4.2) H. 122 a0: 1.7) Nr P 3.06 (2.3-4.7) 2.431 (1.73-3.66) 11.15 (8.16-16.33) 1.144 (1.08-1.2) 0.279 (0.172-0.451) 15.9 (0-69.9)' 0.0351 (0-0.184)' 1.029 (0.66-1.29) 4.90 (2.04-6.12) 1.048 (1.02-1.06) 0.314 (0.245-0.490) 5.7 (0-28.57) 0.0098 (0-0.049) (99%) He/P h (%) TU (%) " mtDNA values calculated only from data of 18 Austrian regional samples 54 F. SUCHENTRUNK et al. were zero for the British and NZ-W sam- ples, and amounted to 0.0041% for all pairs involving the NZ-H sample. Respective pairwise values for the presently studied UK and NZ hares and the earlier studied brown hares from 20 Austrian localities (HArTL et al. 1993) ranged between 0.0- 0.093%; this was within the range (0.0- 0.113%) of net nucleotide diversity be- tween local samples of brown hares from Austria (calculated from data produced in our laboratory, see HARTL et al. 1993). Discussion The levels of cross nuclear and mtDNA variability of brown hares from New Zea- land and Britain are similar. This is in con- tradiction to the hypothesis of reduced gene pool variability in hares from New Zealand, owing to an assumed bottleneck during the period of their introduction. Possibly, un- recorded liberations of hares from diverse provenences in Europe in addition to those from the reportedly small number of foun- ders in Australia (LEvER 1985) have in- creased the effective population size during the founder period in New Zealand. But such additional imports from Europe to New Zealand do not seem very lıkely, given the acclimatized hares already available in Australia for the Acclimatization Compa- nies. The essential point to explain the absence of reduced genetic variability in hares from New Zealand as compared to British hares is that brown hares from both New Zealand and Britain exhibit clearly lower allozymic variability than continental European po- pulations. Theoretically, the small number of allozymic alleles found presently in the British brown hares could have been sampled by only few individuals. All com- mon alleles (i. e., those with relative electro- phoretic mobility 100/-100) of the British samples were also common in brown hares from diverse regions of continental Europe (HARTL et al. 1989, 1990, 1992, 1993, 1994; SUCHENTRUNK et al. 2000; SUCHENTRUNK et al. unpubl. data). But only 16% of all var- iant alleles found so far in continental Eur- opean brown hares with the same set of loci (cf. HARTL et al. 1994; SUCHENTRUNK et al. 2000) occurred in the British samples. Quite several with wide distribution in continental Europe (Bed. Hk-2°, Es-1%, Es-r1%, Es-D'*', Mp'”°) are likely absent in British brown hares. And only one (7.1%) of all mtDNA haplotypes found so far in brown hares from central and southeastern Eur- ope (HartL et al. 1993, 1994; NADLINGER 1994; SUCHENTRUNK, unpubl. data) could be detected in British brown hares. But quite a number of regional samples of Aus- trian brown hares did also not display any mtDNA variability (HArTL et al. 1993). British sample sizes are probably too low for such a comparison. Nevertheless, ab- sence of any mtDNA variability of the pre- sently studied brown hares from Britain and the fact that only the standard Eur- opean mtDNA haplotype I was found in Britain, as opposed to its significantly (p<0.0001, exact Fisher’s test) lower fre- quency on the continent, agrees with the in- terpretation of generally reduced genetic varıability in British brown hares. The low genetic variability of British brown hares might result from an “ancient” popu- lation bottleneck or long-term low effective population sizes associated with their colo- nization history. It has been hypothesized that brown hares were introduced in Ro- man times (cf. CoRBET 1986; ARNOLD 1993). Deliberate releases of only few indi- viduals or occasional escapes from farms in Roman times or earlier during the Meso- lithie (?) or Neolithic occupation periods (cf. Jones and KEEn 1993) could have re- sulted in a poor gene pool variability of the pioneer population. In addition, a low sur- vival rate owing to a high predation pres- sure by foxes and other predators, patho- genes, adverse weather conditions etc., could have hampered a quick population growth in the wild. Long-term low effective population size effectively reduces allelic variabililty and heterozygosity (e. g., CHAK- RABORTY and NEIı 1977; HEDRIcK 1985). Con- trary to the hypothesis of a deliberate intro- duction of brown hares to Britain in Roman Little allozyme and mtDNA variability in brown hares 55 times, YALDEN (1982) and RoBERTS (1994) list brown hares along with mountain hares (Leptus timidus) as native to Britain since late-glacial times (see also GrıGson 1983 for the Later Mesolithic excavation site of Cherhill, Jones and KEEN 1993, and cita- tions therein). In this case, brown hares could have lost genetic diversity at low den- sities under adverse late-glacial climate, or later on in small isolated pockets. STUART (1982), however, lists “Lepus sp.” as an ele- ment of the early Flandrian fauna of Star Charr, and considers only L. timidus defi- nitly recorded as fossil in Britain (cf., May- HEW 1975). As to our knowledge, fossil evi- dence of Lepus europaeus is uncertain in north-central and north-western mainland- Europe during the late-glacial period and early Flandrian, before the formation of the Channel (Strait of Dover, c.8000 years BP cf. e.g., Jones and Keen 1993). Brown hares might not have roamed mid-latitude Europe during the early Flandrian and not have managed to arrive there before the formation of the Channel (see also CoRBET 1986). Alternatively, genetic diversity could have been reduced in fragmented popula- tions with long-term low densities in subop- timal habitats after woodland regeneration in the post-Roman period. But it seems that the distribution of wooded and non-wooded land in Britain has not been altered much in the post-Roman period (RoBErTS 1994). The low allozymic diversity of British brown hares unlikely results from recent in- breeding in regional populations. This is in- dicated by quite normal locus-specific he- terozygosities, lack of heterozygote defi- ciencies, and the generally low or even ne- gative inbreeding coefficients. This inter- pretation is also supported by the absence of a significant gene pool substructuring. Allele frequencies of British brown hares do not differ much across regions. Insignifi- cant fixation indices (Fsr) and genetic dis- tance values indicate absence of drift effects among local populations. Obviously, the low amount of genetic variability contained in British brown hares is partitioned among individuals within local populations rather than across larger geographic ranges. The Acp-1”° allele occurred exclusively in the Scottish sample and was not found in a large number of brown hares from many re- gions in continental Europe. It might result from a recent local mutation. Differentiation between the gene pools of British and central European (Austrian) brown hares (cf., HArTL et al. 1993) is negli- gible, and virtually nil when based on the detected mtDNA haplotypes. The corre- sponding pairwise NEr’s (1978) unbiased D values (0.000-0.005) fall within the range encountered among local populations from Austria (e.g., HARTL et al. 1993). Despite their conventional subspecific position (L. e. occidentalis, DE WınTon, 1898), Brit- ish brown hares represent only a genetically depauperate version of brown hares from continental Europe. This is confirmed by the absence of any other mtDNÄA haplotype apart from haplotype I, which has a phylo- genetically central position in brown hares and is most widespread in central Europe (HARTL et al. 1993; SUCHENTRUNK, unpubl. data). As indicated by the low genetic dis- tances, no significant genetic drift has oc- curred between British and continental European brown hare populations. As re- gards mtDNA, we cannot draw any conclu- sions on differentiation among the regions studied in Britain, because of too low sam- ple sizes for two regions. Also, the present data do not permit any conclusion as to ori- gins of the British brown hares. This might be achieved by comparing regional popula- tion samples and continental European samples with a highly resolving molecular marker system (e.g., microsatellites, mito- chondrial d-loop sequences). The lack of distinct gene pool structuring and the poor allozymic diversity of British brown hares suggest that a relatively high proportion of variant alleles was sampled by the few individuals that supposedly con- stituted the founder populations in New Zealand. Theoretically, all currently found alleles of the British brown hares could have been contained in the founders. In New Zealand, hares spread rapidly after their introduction (FLux 1990). Presently, they occupy large ranges including practi- 56 F. SUCHENTRUNK et al. cally all types of habitats on the North and South Islands. Although they commonly do not reach those high densities as in Europe, they even have established successful popu- lations in sub-alpine (PArKES 1984) and rather harsh alpine environments (FLux 1967). Their successful spread in New Zea- land might have been particularly favoured by a low level of parasitic or infectuous dis- eases, a long-term low level of agrochemis- try, and relatively little intensification of farming machinery, little road mortality, low predation pressure on adults etc. (see also FLux 1990). This rapid increase of the founder populations on both main islands, together with repeated imports and subse- quent translocations within and between the islands by diverse naturalization compa- nies (FLux 1990) probably prevented severe loss of genetic variability. Generally, quick population increases following a bottleneck, as was reported for the founder populations of Phillip Island (Rorıs 1969) and New Zealand (Frux 1990), are counteracting re- duction of allelic variability in founder po- pulations (e.g., NEı et al. 1975). As expected, particularly rare alleles (Idh- 2'°° Acp-1°°) of British brown hares are absent in hares from New Zealand, whereas the common British alleles are common in New Zealand too. Interestingly, the Mpi”- allele of both populations from New Zea- land was neither present in the British brown hares nor in any of over 900 brown hares from diverse regions of continental Europe studied so far (HARTL et al. 1994; SUCHENTRUNK et al. 2000; SUCHENTRUNK, unpubl. data). It might stem from a recent mutation in New Zealand or Australia. Al- ternatively, it could be present in regions of Britain or anywhere else in Europe that were not yet sampled, but wherefrom hares were shiped to Australia. The same inter- pretations apply to the exclusive occurrence of mtDNAÄA haplotype V in the Harpa/Avoca catchment on the South Island of New Zealand. This haplotype is phylogenetically closely related to the basal mtDNA haplo- type I of European brown hares, and its evolution can be explained by only one base substitution. The two local populations on both main is- lands of New Zealand have differentiated little since their foundations over one hun- dred years ago. The nuclear and mitochon- drial gene pools of hares from New Zealand are very similar to those of British brown hares. Lack of data of allele frequencies in the source population, unknown effective population sizes, and repeated transloca- tions of hares within and between the North and South Islands, that might have changed allele frequencies, prevented us to compare the presently observed allele frequencies with theoretical frequencies as resulting from drift simulations (see e.g., FiTzSım- MoNS et al. 1997). But both Ner’s (1978) D and Fsrvalues between British and New Zealand populations largely fall within the ranges encountered among regional popula- tions in Europe (D = 0.000-0.019, Fsr = 0.00-0.124, re-calculated from adjusted data sets of 49 locı, cf. HARTL et al. 1993; SUCHEN- TRUNK et al. 2000). In fact, only the Fsr-va- lue for the NZ-H and UK-L samples slightly exceeds the range for regional samples from mainland Europe. This marginally increased level of differentiation between these two populations apparently results from the ab- sence of the Pep-2!°* allele in UK-L and the Es-1° allele in NZ-H. But in general, we cannot see any marked reorganization in the cross gene pool of hares from New Zealand as compared to British brown hares. Rapid population increases in the lo- cal founder populations, repeated releases, and multiple transfers of hares by naturali- zation companies (Frux 1990) likely pre- vented strong genetic drift. However, in spite of apparently well adapted populations in Britain and New Zealand (SUCHENTRUNK et al. 1998), brown hares from these coun- tries harbour less genetic resources for adaptation to future environments than po- pulations from mainland Europe. Acknowledgements We are grateful to Dr J. E. C. FLux (Lower Hutt), Dr J. P. PArkes (Christchurch), Mr Hunter (Mas- terton, New Zealand), Dr S. TArper (Fording- Little allozyme and mtDNA variability in brown hares 57 bridge), Mr A. Purtik (Middle Woodford, Wilt- shire, England), Mr BrockLess (Loddington, Lei- cestershire, England), and to “Bain of Tarves - Ca- tering Butchers, Game and Venison Exporters” (Tarves, Aberdeenshire, Scotland) for organizing shoots and supplying hares. “Manaaki Whenua - Landcare Research” (Lower Hutt and Christ- church, New Zealand) and “The Game Conser- vancy Trust” (Fordingbridge, Hampshire, England) provided facilities. Mrs M.M.Fıux (Belmont, New Zealand), Mr M. CoLEMAN (Christchurch), Zusammenfassung Mrs M TRreBILco, Mr J. Long, MR S. MoRrEBY, and Mr €. STOATE (Fordingbridge) helped with the or- ganisation work or provided otherwise valuable heip. Dr K. NADLinGer (Vienna) helped with com- parisons of mtDNA haplotypes found presently and earlier. Priv-Doz Dr D. Heinrich and Mrs R. Lücht (Institut für Haustierkunde, University of Kiel, Germany) helped with archeozoological literature search, and Mr A. KÖRBER (Research In- stitute of Wildlife Ecology, Vienna Veterinary Uni- versity, Austria) created the figure. Geringe Allozym- und mtDNA-Variabilität bei neuseeländischen und britischen Feldhasen (Lepus europaeus) - eine Folge von Flaschenhälsen? Untersucht wurde die Allozym- und mtDNA-Variabilität bei britischen und neuseeländischen Feldha- sen (Lepus europaeus), um die Hypothese eines Variabilitätsverlustes bei neuseeländischen Hasen infolge eines Flaschenhalsereignisses („Gründer-Effekt“) bei ihrer Einbürgerung im 19. Jahrhundert zu überprüfen. Bei 119 Hasen aus drei britischen und zwei neuseeländischen Stichprobengebieten wurde mittels horizontaler Stärkegelelektrophorese ihre allelische Variabilität an 52 Strukturgenloci ermittelt. Bei 36 Hasen wurde anhand von 16 6-Basenschneidenden Endonukleasen der Restrik- tionsfragmentlängenpolymorphismus (RFLP) in der gesamten mitochondrialen DNA (mtDNA) analy- siert. Alle Indices der Allozym- und mtDNA-Variabilität lagen bei den neuseeländischen Stichproben im Bereich jener der britischen Feldhasen. Somit kann die Hypothese zu Verlusten an genetischer Variabilität bei neuseeländischen Hasen im Vergleich zu den britischen nicht aufrecht erhalten wer- den. Ausgeprägte genetische Drifteffekte zwischen britischen und neuseeländischen Feldhasen konnten ebenfalls nicht festgestellt werden. Jedoch zeigten alle fünf untersuchten regionalen Stichproben signifikant geringere Allozym-Variabilität als Feldhasen vom europäischen Kontinent. Dies läßt die Vermutung eines Verlustes der genetischen Variabilität bei den britischen Feldhasen im Verlaufe ihrer Besiedlungsgeschichte zu. Die nominell als eigene Unterart geführten britischen Feldhasen (L. e. occidentalis, DE WINTON; 1898) erwiesen sich lediglich als genetisch verarmte Var- ianten der kontinental-europäischen Feldhasen. References ARNOLD, H.R. (1993): Atlas of mammals in Brit- CorBET, G. B.; SOUTHERN, H.N. (1977): The ain. ITE research publ. 6. London: HMSO. AVISE, J. C. (1994): Molecular Markers. Natural History and Evolution. New York: Chapman and Hall. AVISE, J. C.; HAMRICK, J. C. (1996): Conservation Genetics. Case Histories from Nature. New York: Chapman and Hall. CHAKRABORTY, R.; NEI, M. (1977): Bottleneck ef- fects on average heterozygosity and genetic distance with the stepwise mutation model. Evolution 31, 347-356. CorBET, G. B. (1986): Relationships and origins of the European lagomorphs. Mammal Rev. 16, 105-110. Handbook of British Mammals. Sec. Ed.; Ox- ford: Blackwell. FITZSIMMonS, N.; BUSKIRK, S. W.; SMITH, M.H. (1997): Genetic changes in reintroduced Rocky Mountain Bighorn Sheep populations. J. Wildl. Managem. 61, 863-872. FrLux, J. E. C. (1965): Incidence of ovarian tumors in hares in New Zealand. J. Wildl. Managem. 29, 622-624. FLux, J. E. C. (1967): Hare numbers and diet in an alpine basin in New Zealand. Proc. New Zeal. Ecol. Soc. 14, 27-33. Frux, J.E.C. (1980): High incidence of missing posterior upper molars in hares (Lepus euro- 58 F. SUCHENTRUNK et al. paeus) in New Zealand. New Zeal. J. Zool. 7, 257-259. FrLux, J. E. C. (1990): Brown hare. In: The Hand- book of New Zealand Mammals. Ed. by C.M. King. Auckland: Oxford University Press. Pp. 162-172. FrLux, J. E. C.; DUTHIE, A. G.; ROBINSON, T. J.; CHAP- MAN, J. A. (1990): Exotic populations. In: Rab- bits, Hares and Pikas. Status Survey and Conser- vation Action Plan. ED. by J. A. Chapman and J. E. C. Flux. Gland: IUCN. Pp. 147-153. FUERST, P. A.; MARUYAMA, T. (1986): Considera- tions on the conservation of alleles and of genic heterozygosity in small managed popu- lations. Zoo Biology 5, 171-179. GILES, RE.; BLAnc, H.; Cann, H.M.; WALLA- CE, D. C. (1980): Maternal inheritance of hu- man mitochondrial DNA. Proc. Natl. Acad. Sci., USA 77, 6715-6719. GRIGSoN, C. (1983): Mesolithic and neolithic ani- mal bones. In: Excavations at Cherhill, North Wiltshire, 1967. Proceed. Prehistoric Soc. 49. Ed. by J. G. Evans, and I. F. SMITH. Pp. 64-72. GRILLITSCH, M.; HARTL, G. B.; SUCHENTRUNK, F.; WirLınG, R. (1992): Allozyme evolution and the molecular clock in the Lagomorpha. Acta Theriol. 37, 1-13. GYLLENSTEN, UÜ.; WHARTON, U. D.; JOSEFSON, A.; Wirson, A.C. (1991): Paternal inheritance of mitochondrial DNA. Nature 352, 255-257. HARRIS, H.; Hopkınson, D. A. (1976): Handbook of Enzyme Electrophoresis in Human Genet- ics. Amsterdam: North-Holland Publ. Co. HARTL, G. B.; Pucek, Z. (1994): Genetic depletion in the European bison (Bison bonasus) and the significance of electrophoretic heterozyg- osity for conservation. Conserv. Biol. 8, 167- 174. HARTL, G. B.; WILLING, R.; NADLINGER, K. (1994): Allozymes in mammalian population genetics and systematics: Indicative function of a mar- ker system reconsidered. In: Molecular Ecol- ogy and Evolution: Approaches and Applica- tions. Ed. by B.SCHIERWATER, B. STREIT, G. P. WAGNER, and R.DESALLE. Basel: Bir- khäuser. P. 299-310. HARTL, G. B.; SUCHENTRUNK, F.; NADLINGER, K.; WirLing, R. (1993): An integrative analysis of genetic differentiation in the brown hare Le- pus europaeus based on morphology, allo- zymes and mitochondrial DNA. Acta Theriol. 38, 33-57. HARTL, G. B.; SUCHENTRUNK,F; WILLING, R.; GRILLITSCH, M. (1989): Biochemisch-genet- ische Variabilität und Differenzierung beim Feldhasen (Lepus europaeus) in Niederöster- reich. Wien. Tierärztl. Mschr. 76, 279-284. HARTL, G. B.; MARKOWSKI, J., KovAcs, G.; GRIL- LITSCH, M.; WirLing, R (1990): Biochemical var- iation and differentiation in the brown hare (Le- pus europaeus) of Central Europe. Z. Säugetierkunde 55, 186-193. HARTL, G. B.; MARKOWSKI, J.; SWIATECKI, A.; JA- NISZEWSKI, T.; WILLING, R. (1992): Genetic di- versity in the Polish brown hare Lepus euro- paeus Pallas, 1778: implications for conservation and management. Acta Theriol. 37, 15-25. HEDRICK, P. W. (1985): Genetics of Populations. Boston: Jones and Bartlett Publ., Inc. Jongs, R. L.; KEEn, D. H. (1993): Pleistocene En- vironments in the British Isles. London: Chap- man and Hall. LANSMAN, R. A.; SHADE, R. O.; SHAPIRA, J. F.; AvI- SE, J. C. (1981): The use of restriction endonu- cleases to measure mitochondrial DNA se- quence relatedness in natural populations. III Techniques and potential applications. J. Mol. Evol. 17, 214-226. LEBERG, P.L. (1992): Effects of population bottle- necks on genetic diversity as measured by al- lozyme electrophoresis. Evolution 46, 477- 494. LEver, C. (1985): Naturalized Mammals of the World. London, New York: Longman. MAHOOoB, 1. T. (1983): Brown hare (Lepus capen- sis). In: The Australian Museum Complete Book of Australian Mammals. Ed. by R. STRA- HAN. London: Angus and Robertson Publ. P. 480. MAYHEW, D. F. (1975): The Quaternary history of some British Rodents and Lagomorphs. Ph. D. Thesis, University of Cambridge. NADLINGER, K.F. (1994): Restriktionsfragmen- tlängenpolymorphismus mitochondrialer DNA in Lepus europaeus, Cervus elaphus und Bison bonasus. PhD Diss., University of Vien- na. NeEı, M. (1978): Estimation of average heterozyg- osity and genetic distance from a small num- ber of individuals. Genetics 89, 583-590. NeEı, M. ( 1987): Molecular Evolutionary Genetics. New York: Columbia Univ. Press. NeEı, M.; Lı, W.H. (1979): Mathematical model for studying genetic variation in terms of re- striction endonucleases. Proc. Nat. Acad. Sci., U.S.A. 76, 5269-5273. NEr, M.; MARUYAMA, T.; CAKRAABORTY, R. (1975): The bottleneck effect and genetic variability in populations. Evolution 29, 1-10. PARKES, J. P. (1984): Home ranges of radio-tele- metred hares (Lepus capensis) in a sub-alpine population in New Zealand: implications for control. Acta Zool. Fenn. 171, 279-281. Little allozyme and mtDNA variability in brown hares 59 PARKES, J. P. (1988): Missing posterior upper mo- lars in Lepus europaeus in New Zealand. La- gomorph Newsl. 7, 12. RıcKwoop, D. (1987): Centrifugation. A Practical Approach. Sec. Ed. Oxford: IRL Press. Rice, W. S. (1989): Analyzing tables of statistical tests. Evolution 43, 223-225. ROBERTS, N. ( 1994): The Holocene. An Environ- mental History. Oxford: Blackwell. Roııs, E.C. (1969): They All Run Wild. The Story of Pests on the Land of Australia. Syd- ney: Angus and Robertson. ROTHE, G. M. (1994): Electrophoresis of Enzymes. Laboratory Methods. Springer LAB Manual. Berlin: Springer. STUART, A.J. (1982): Pleistocene Vertebrates in the British Isles. London, New York: Long- man. SUCHENTRUNK, F. (1993): Variability of minor tooth traits and allozymic diversity in brown hare Lepus europaeus populations. Acta Theriol. 38, Suppl., 2, 59-69. SUCHENTRUNK, F; WirLInG,R.; HaARTL, G. B. (1991): On eye lens weights and other age cri- teria of the Brown hare (Lepus europaeus Pallas, 1778). Z. Säugetierkunde 56, 365-374. SUCHENTRUNK, F.; MARKOWSKI, J.; JANISZEWSKI, T.; HARTL,G.B. (1992): Dental and cranial anomalies in Austrian and Polish brown hares Lepus europaeus populations. Acta Theriol. 37, 241-257. SUCHENTRUNK, F; MICHAILOV, C.; MARKOV, G.; HAIDEN, A. (2000): Population genetics of Bulgarian brown hares Lepus europaeus: allo- zymic diversity at zoogeographical crossroads. Acta Theriol. 45, 1-12. SUCHENTRUNK, F, HARTL,G.B.; FLux,J.E.C.; PARKES, J.; HAIDEN, A.; TAPPER, S. (1998): Al- lozyme heterozygosity and fluctuating asym- metry in brown hares Lepus europaeus intro- duced to New Zealand: Developmental homeostasis in populations with a bottleneck history. Acta Theriol., Suppl. 5, 35-52. SWOFFORD D. L.; SELANDERR.B. (1989): BIOSYS-1. A computer program for the ana- lysis of allelic variation in population genetics and biochemical systematics. Release 1.7. Users manual. Champaign: Illinois Natural History Survey. TIEDEMANN, R.; HAMMER, S.; SUCHENTRUNK, F.; HARTL, G. B. (1997): Allozyme variability in medium-sized and large mammals: determi- nants, estimators, and significance for conser- vation. Biodiv. Letters 3, 81-91. WRIGHT, S. (1978): Evolution and Genetics of Po- pulations. Vol. IV. Variability within and among Natural Populations. Chicago: Univ. Chicago Press. YALDEN, D. W. (1982): When did the mammal fau- na of the British Isles arrıve? Mammal Rev. 12, 1-56. Authors’ addresses: Dr. FRANZ SUCHENTRUNK, Mag. CLAUDIA JASCHKE, ANITA HAIDEN, Research Institute of Wildlife Ecology, Vienna Veterinary University, Savoyen- strasse I, A-1160 Vienna, Austria. E-mail: franz.suchentrunk@vv-wien.ac.at Mamm. biol. 66 (2001) 60-62 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication DS 5. Mammalian Biology CTEREERSS Zeitschrift für Säugetierkunde First record of Rattus rattus in Botswana By K. M. HELGEN Museum of Comparative Zoology, Mammal Department, Harvard University, Cambridge, USA Receipt of Ms. 06. 07. 2000 Acceptance of Ms. 28. 09. 2000 Key words: Rattus rattus, Botswana, bubonic plague, commensalism, sylvatic plague The murid rodents Rattus rattus (Linnaeus, 1758) and Rattus norvegicus (Berkenhout, 1769) are widely distributed in Africa (Kınapon 1974; ROSEVEAR 1969; SKINNER and SMITHERS 1990) and indeed boast nearly cosmopolitan distributions (MUSSER and CARLETON 1993). These ubiquitous commensals are usually associated with hu- man habitation. Several authors have noted the absence of these rodents from Botswa- na (Davıs 1946; DE GRAAFF 1981; SKINNER and SMITHERS 1990), which may be the last African nation for which there is no pub- lished record of either species’ presence. I spent September to December 1999 in Botswana, devoting some time to trapping and preserving rodents, especially in areas inhabited by people, both urban and rural. Four specimens of Rattus rattus were trapped in the eastern sector (Fig. 1). All of them were captured in close association with human habitation: one from under a kitchen stove in Mmathubudukwane (Mu- seum of Comparative Zoology No. 62631), a rural village near the border with South Africa; one in a granary at Francistown (MCZ 62632); and two from a basement in Ramotswa near Gaborone (MCZ 62633- 62634). 1616-5047/01/66/01-060 $ 15.00/0. These specimens of Rattus rattus (Tab. 1) most likely represent an invasion into Bots- wana that occurred within the last thirty years. SMITHERS (1971) noted that it was not recorded in the country despite 5 years of effort during the 1964-1969 Botswana Mammal Survey. He labeled Rattus rattus as a “species not recorded but which may occur, noting that despite the absence of specimens from Botswana, it had been trapped nearby at Kariba Dam in Rhodesia (now Zimbabwe), which lies on a railway line that extends into Botswana. Consider- ing its now long-term presence in neighbor- ing South Africa as well, it is rather surpris- ing that the black rat did not invade Botswana earlier. Other common synan- thropic rodents have been similarly slow to colonize the country; Rattus norvegicus has not been recorded, and SMITHERS (1971) listed only a single specimen of Mus muscu- lus trapped in Botswana. DE GRAAFF (1981) suggested that the aridity of much of the country precludes invasion by exotic rodent commensals like Rattus. Other relevant fac- tors discouraging colonization probably in- clude the extremely low population density of the country, its distance from port cities, and, perhaps, the diversity of native rodents First record of Rattus rattus in Botswana 61 Table 1. Specimens of Rattus rattus collected in Botswana. Measurements are given in millimeters as Total Length - Tail Length - Hindfoot Length - Ear Length. Sex Measurements Date collected Locality adult adult adult 410-220-35-18 381-209-35-25 383-207-36-25 Mmathubudukwane Francistown Ramotswa 28 September 1999 6 December 1999 15 December 1999 juvenile Ramotswa 15 December 1999 Botswana Lesotho South Africa Swaziland Zimbabwe Fig. 1. Localities described in the text; 1, Jao, a small village in the Okavango Delta; 2, vicinity of Toromoja and Xhumo; 3, Francistown, 4, Mmathubudukwane and 5, Ramotswa, sites where Rattus rattus has been recorded. acting as human commensals. Most areas of the country seem to remain free of Rattus rattus, at one such locality, a remote village in the Okavango Delta named Jao, I trapped four native rodents — Aethomys chrysophilus (de Winton, 1897), Mastomys coucha (Smith, 1834), Saccostomus campes- tris Peters, 1846, and Graphiurus murinus (Desmarest, 1822), living in or entering huts and outbuildings. The most common rodent in Botswana is the multimammate mouse Mastomys coucha, which was caught in both wild and settled areas in the Okavango Del- ta, the northern Kalahari region, and the eastern sector (from Francistown to the Ga- borone area). Rattus rattus has not been recorded from any areas but the thickly settled eastern sec- tor, and is probably unlikely to invade the Okavango Delta or the more sparsely settled and arid Kalahari region of Bostwa- na’s interior. Despite reports of Rattus rattus from the Boteti District that were made by rodent surveyors following the bubonic pla- gue epidemic in Xhumo and Toromoja in 1989-90 (MOKGWEETSINYANA pers. comm.), 62 K. M. HELGEN there is no evidence of its existence there. No voucher specimens were preserved, but I have examined survey records at the headquarters of the Ministry of Health in Francistown, and the recorded measure- ments clearly demonstrate that the animals were not Rattus but misidentified Mast- omys. Rattus was thus almost certainly not Acknowledgements I am thankful to S. MOKGWEETSINYANA for access to documents at the Ministry of Health of Bots- wana, to S. MATALAOTE for administrative assis- tance regarding permits and travel, to M. MoT- SHOME for assistance in collecting specimens, and to D. Wiırson for his help in preparing this manu- References Davıs, D. H. S. (1946): A plague survey of Ngami- land, Bechuanaland Protectorate, during the epidemic of 1944-45. S. Afr. Med. J. 20, 462- 467, 511-515. DE GRAAFF, G. (1981): The Rodents of Southern Africa. Durban: Butterworths. KınGoon, J. (1974): East African Mammals: An Atlas of Evolution in Africa. 2B (Hares and Rodents). London, New York: Academic Press. MUSSER, G. G.; CARLETON, M.D. (1993): Family Muridae. In: Mammal Species of the World. 2”@ ed. Ed. by D. E. Wırson and D. M. Ree- DER. Washington, London: Smithson. Inst. Press. Pp. 501-755. involved in this epidemic, which as Davis (1946) noted for a previous epidemic in the region, was a case of sylvatic plague, impli- cating native Tatera as a permanent reser- voir and Mastomys, which travels between Tatera colonies and human dwellings, as a vector for transmission of plague fleas. script. Research costs were defrayed in part with a Barbour Fund grant from the Museum of Com- parative Zoology at Harvard University. Collect- ing was carried out under permit no. WPRES3/ 2 XIV (33) issued by the Department of Wildlife and National Parks, Botswana. ROSEVEAR, D. R. (1969): The Rodents of West Africa. London: Trust. Brit. Mus. Nat. Hist. SKINNER, J. D.; SMITHERS, R.H.N. (1990): The Mammals of the Southern African Subregion. 2"d ed. Pretoria: Univ. Pretoria. SMITHERS, R. H.N. (1971): The Mammals of Bots- wana. Mus. Mem. 4.Salisbury, Rhodesia: Trust. Nat. Mus. Rhodesia. Author’s address: KRISTOFER M. HELGEN, Museum of Comparative Zoology, Mammal Department, Harvard Univer- sity, 26 Oxford Street, Cambridge, MA 02138, USA (e-mail: helgen@fas.harvard.edu). Mamm. biol. 66 (2001) 63-64 3. Mammalian Biology © Urban & Fischer Verlag Ay FR http://www.urbanfischer.de/journals/mammbiol - Zeitschrift für Säugetierkunde Mitteilung der Gesellschaft 75. Jahrestagung der Deutschen Gesellschaft für Säugetierkunde vom 23.-27. September 2001 in Berlin Einladung Die 75. Jahrestagung der Deutschen Gesellschaft für Säugetierkunde wird von Sonntag, dem 23. September bis Donnerstag, dem 27. September 2001 in Museum für Naturkunde und im Zoolo- gischen Garten zu Berlin ausgerichtet. Vorläufiges Programm Anreise ab 16.00 Uhr: Vorstandssitzung der DGS ab 19.00 Uhr: zwangloser Begrüßungsabend 09.00 Uhr: Begrüßungen 09.30 Uhr: Hauptvortrag und Kurzvorträge zum Themenschwerpunkt: „Tiergartenbiologie“ Sonntag, 23. September Montag, 24. September: Dienstag, 25. September: Mittwoch, 26. September: Donnerstag, 27. September: 14.00 Uhr: 15.00 Uhr: 16.00 Uhr: 20.00 Uhr: 09.00 Uhr: Posterdemonstration Kurzvorträge Mitgliederversammlung der DGS Öffentlicher Abendvortrag Hauptvortrag und Kurzvorträge zum Themenschwerpunkt „Ökomorphologie“. 14.00 Uhr: 15.00 Uhr: 16.30 Uhr: 19.00 Uhr: 09.00 Uhr: 14.00 Uhr: 15.00 Uhr: 18.00 Uhr: Posterdemonstration Kurzvorträge Führung durch den Zoologischen Garten Geselliger Abend im Zoologischen Garten Vorträge zu freien Themen Posterdemonstration Kurzvorträge Posterprämierung Wissenschaftliche Exkursion, wahlweise a) Deutsch-Polnischer Nationalpark „Unteres Odertal“ (in Privatwagen oder Kleinbussen) b) Führung durch den Tierpark Berlin-Friedrichsfelde 1616-5047/01/66/01-063 $ 15.00/0. 64 Mitteilung der Gesellschaft Alle Interessenten, Mitglieder und Nicht- mitglieder, sind zu dieser Jahrestagung herzlich nach Berlin eingeladen. Das Pro- gramm mit der Vortragsfolge wird den Mit- gliedern — auf Anforderung auch Nichtmit- gliedern -— rechtzeitig vor der Tagung zugesandt. Sollten Sie eine persönliche Ein- ladung benötigen, so wenden Sie sich bitte an den 1. Vorsitzenden der Deutschen Ge- sellschaft für Säugetierkunde, Prof. Dr. R. SCHRÖPFER, Universität Osnabrück, FB Biologie/Chemie: Ethologie, Barbarastr. 11, D-49069 Osnabrück, (Tel.: +49-(0) 541/ 969-2848, Fax +49-(0) 5 41/9 69-28 62, e-mail: schroepfer@biologie.uni-osnabrueck.de). Wir bitten um die Anmeldung von Kurzvor- trägen (15min+5 min Diskussion) und Posterdemonstration zu den genannten Themenschwerpunkten und zu weiteren Fachgebieten der Säugetierkunde. Bitte melden Sie Ihre Beiträge möglichst frühzeitig, spätestens jedoch bis zum 30. April (Ausschlußfrist) beim Geschäfts- führer der DGS, Prof. Dr. G. B. HARTL, In- stitut für Haustierkunde, Universität Kiel, Olshausenstr. 40-60, D-24118 Kiel an. Der Anmeldung ist eine informative Kurzfas- sung beizufügen. Aus ıhr sollen die Frages- tellung, Methoden, Ergebnisse und die dar- aus gezogenen Schlußfolgerungen hervor- gehen. Alle angenommenen Kurzfassungen werden in einem Sonderheft von „Mamma- lian Biology“ publiziert. Die Kurzfassungen sollen nach folgendem Schema abgefaßt werden: Maximal einseitig bei 1,5-zeiliger Schrift! Deutscher Titel — Leerzeile — englischer Untertitel (Im Titel mit Klein- buchstaben schreiben. Beispiel: Lingual and biting responses ...; bitte ggf. einen „native speaker“ konsultieren!) — Leerzeile — Autorenname/n (Großbuchstaben) mit Adresse (Beispiel: M.MÜLLER und H. WEBER, Institut für Zoologie, Uni- versität....) — Leerzeile — Text. Die Kurzfassung bitte unbedingt ausge- druckt zusenden. Zusätzlich eine Diskette (3,5”, IBM-kompatibler DOS-PC) mit der Kurzfassung in Form eines Word-Doku- ments, Word Perfect- oder ASCII-Eiles bei- fügen. (Bitte diese Datei nicht formatie- ren.) Die Maße für die Poster werden im Rund- brief 2/2000 bekanntgeben. Mit Fragen zum Tagungsort und zur Orga- nisation wenden Sie sich bitte an den Geschäftsführer der DGS, Prof. Dr. G. B. HARTL, Institut für Haustierkunde, Universität Kiel, Olshausenstr. 40-60, D-24118 Kiel, Tel. +49-(0)4 31/880 45 07, Fax +49-(0)4 31/880 13 89, e-mail: ghartl@ifh.uni-kiel.de. Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana- ging editor, Prof. Dr. D. Kruska, Institut für Haustier- kunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni- kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manu- script will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21 x 30 cm) or correspondingly either from a typewriter or legible computer print with a left-hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscript. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscript following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type- written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscript. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings &. Mammalian Biology Ay Zeitschrift für Säugetierkunde depending on the theme, Acknowledgements, Zusam- menfassung (in German), References. A German title must precede the Zusammenfassung in English manu- scripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: Kaıko, E. K. V.; Hanouev, C. O. Jr. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1821), from Panama. Z. Säugetierkunde 59, 257-273. b) From books: HERRE, W.; RöHrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers’ symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscript is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen, Germany (email:peter.langer@anatomie.med. uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). Mammalian Biology Be ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. - 66(2001)1 - 1-64 Contents Reviews Smith, M. H.; Novak, J. M.; Peles, J. M.; Purdue, J. R.: Genetic herterogeneity of white-tailed deer: Management lessons from a long-term study - Genetische Heterogenität beim Weißwedelhirsch: Für die Wildbewirtschaftung relevante Erkenntnisse aus einer langzeitsttudidie 0 0 2 EU Hewison, A. J. M.; Danilkin, A.: Evidence for separate specific status of European (Capreolus capreolus) and Siberian (C. pygargus) roe deer - Beweise für den Artstatus von Europäischem (Capreolus capreolus) und Sibirischem (C. pygargus)Rehtwild —————— zn. 00 DA, Original investigations Goodman, 5. M.; Hutterer, R.; Ngnegueu, P. R.: A report on the community of shrews (Mammalia: Soricidae) occurring in the Mink&b& Forest, northeastern Gabon - Bericht über die Artengemeinschaft von Spitzmäusen (Mammalia: Soricidae) im Minkebe Regenwald, Nord-Ost Gbun _— „an, en; Ulevicius, A.; Sidorovich, V.; Lauzhel, @.: Specificity of non-metric parameters. of American mink (Mustela icon) populations in relation to habitat differences in Belarus - Spezifität nicht-metrischer Parameter von Mink-Populationen (Mustela vison) im Verhältnis zu en inWeißrußland >: fa, Suchentrunk, F.; Jaschke, C.; Haiden, A.: Little allozyme and mtDNA variability i in brown hares (Lebus europaeus) from New Zealand and Britain - A legacy of bottlenecks? - Geringe Allozym- und mtDNA-Variabilität bei neuseeländischen und britischen Feldhasen (Lepus europaeus) - eine Folge von Flaschenhäisen? ___ 43 Short communication Helgen, K. M.: First record of Rattus rattus in Botswana - Erster Nachweis von Rattus rattus in Botswana ____ 60 Mitteilung der Gesellschaft Einladung nach Berlin 00 0. 2.00,, DU IM Il Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record 70 78, Mammalian en Zeitschrift für Säugetierkunde 22001 5, | ISSN 1616-5047 URBAN & FISCHER Mamm. biol. - 66(2001)2 - 65-128 www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Saugetierkunde (German Society for Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel R. Hutterer, Bonn E. Kalko, Ulm P. Lüps, Bern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repennning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.O. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/46 35 70, E-mail: l.still@nature.com Subscribers in the USA can call toll free on ++1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 498.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipient's name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by personal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. If no cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH &) As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANSI/NISO 239.48-1992 (Permanence of Paper). ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Printed in Germany Mamm. biol. 66 (2001) 65-83 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol =: Mammalian Biology Zeitschrift für Säugetierkunde Original investigation Comparison of the social behaviour of captive Sympatric peccary species (genus Tayassu); correlations with their ecological characteristics By G. DuBosT Laboratoire de Conservation des Esp&ces animales, Mus&um national d’Histoire naturelle, Paris, France Receipt of Ms. 03. 01. 2000 Acceptance of Ms. 10. 08. 2000 Abstract Comparison was made of the social behaviour of two congeneric peccary species, the white-lipped (Tayassu pecarı) and the collared peccary (T. tajacu), coexisting in South American rain forests and observed in captivity. In the former species, herd cohesion is strong, and strangers generally are violently attacked. White-lipped peccaries have 2-3 times more contacts with partners of their herd than collared peccaries. In many social behavioural situations, the dominant female is the most active individual in the white-lipped peccary herd, whereas the dominant male is the focal member of the social unit in the collared peccary. Subordinate and subadult males participate in all social behaviour, including sexual, and are very well integrated into the white-lipped peccary herd. In contrast, subordinate collared peccary males are more or less neutral and peripheral indivi- duals. Unlike the latter species, aggressiveness is noticeable in most behaviour of the white-lipped peccary; the dominant male is the main effector of these agonistic contacts, which are frequent and intense. In this species, both sexes belong to only one hierarchic order, with males always superior to females. Conversely, in the collared peccary, there are two distinct monosexual hier- archic orders, and the females dominate the males. These interspecific differences, as well as the total lack of ground marking in the white-lipped peccary, fit well the ecological characteristics of both species: the white-lipped peccary lives in wandering, large multiple-male herds, and the col- lared peccary in small stable and locally resident troops. Key words: Tayassu, social behaviour, peccaries Introduction Unlike species in open habitats, terrestrial or more gregarious forest species on every mammals of tropical forests are generally en- countered alone or in small social units. While there are very few exceptions to this ecological rule, the Bovids (EstEs 1974) and the Suiforms (FrÄDrIcH 1974; BARRETTE 1986; CALDECOTT et al. 1993) each have one 1616-5047/01/66/02-065 $ 15.00/0. continent. There has been no specific study to date to examine the behavioural mecha- nisms underlying the formation of permanent social groups in closed terrestrial habitats. Two peccary species (Tayassuidae), the white-lipped peccary (Tayassu pecari or 66 G. DuBost WLP) and the collared peccary (T. tajacu or CP), coexist in Central and South American rain forests. Both belong to the same genus and share many morphological and physio- logical characteristics; their basic behaviour can therefore be considered broadly similar. Both species live all year long in mixed social groups. This fact probably distinguishes the peccaries from most other Suiforms (BIGOUR- DAN 1948; GUNDLACH 1968; FRÄDRICH 1974; BEUERLE 1975; CALDECOTT et al. 1993). In reality, the two peccary species differ markedly in the size and composition of so- cial units, and in range utilization. White- lipped peccaries are generally encountered in large and apparently nomadic herds, while collared peccaries live in small per- manent troops. The study of their general behaviour permits greater qualitative and quantitative understanding of this diver- gence in terms of life style parameters and social structure. Until now, the collared peccary has mainly been studied from animals originating in the semi-desert southwestern portions of the U.S. A. Nevertheless, there is some in- formation about this species in its forest en- vironment (KırrıE and TERBORGH 1976, 1983; RoBInson and EISENBERG 1985; BoD- MER 1989; HERNANDEZ et al. 1995; PERES 1996). Much of the biology of the white- lipped peccary remains to be described. Given that much of the behaviour of Artio- dactyls, even social, is of an innate nature and appears among captive animals as well as in the wild, captive studies, at least in adequate enclosures, can provide a detailed yet accurate view of the context of each be- haviour in the repertoire and how it is used. Apart from the increase in the knowledge of the biological characteristics of these peccaries, such a study permits a compara- tive analysis of their behaviour, which will help us to understand how these two closely related species could coexist in the same forest habitats. The study also has practical value in that captive peccaries are bred in many parts of the world to replace stocks lost to hunting. Knowledge of their natural behaviour facilitates the husbandry for this endeavour. Material and methods The behavioural data are derived almost exclu- sively from observations of semi-captive animals held at the field station of the Institut National de la Recherche Agronomique (I.N.R.A.), near Kourou, French Guiana. Each social group was maintained in a 1,000-m” enclosure established under an existing forest canopy, surrounded by a 2m-high wire electric fence, and equipped with 4food tubs and 2drinking troughs permitting bathing. Initial social units were comprised of young ani- mals (SCP, 4WLP) captured by local hunters, most often at an age of several days, and hand- reared in groups until weaning. Subsequently, each group was allowed to evolve by itself (in- cluding births and growth of the young), apart from a single daily supply of food, and occasional cleaning. From 1992 to 1994, the CP group in- creased steadily from 7 (2 adult males, 3 adult fe- males, 2 juvenile males) to 16 individuals (4 adult males, 3 adult females, 1 subadult male, 4 suba- dult females, 2 juvenile males, 2 juvenile females). The WLP group almost tripled during the same period, progressing from 4 animals (1 adult male, 3 adult females) to 11 (2 adult males, 3 adult fe- males, 1 subadult male, 1 juvenile male, 4 juvenile females). All individuals were identified by ear- tags or coded ear-notches. Observations were made in both the main rainy (January to June) and dry seasons (July to De- cember). Sampling was undertaken when animals were most active, i.e. in early morning and late afternoon. The observer stood at a distance from which all of the animals could be viewed simulta- neously without any interference with their behaviour. Data were collected by observing each troop member in turn as a focal animal during 30 consecutive minutes. Excluding data on new- born infants, a total of 1,101 individual-hour units of CP observations and 1,136 individual-hour units of WLP were recorded. The different behaviour of each species has been described elsewhere (Dusost 1997). The social behaviour included here is only that involving an approach or physical contact between 2 indivi- duals for peaceful interactions, agonistic or sexual purposes. Given that the whole group was almost always gathered in a small area, it was easy to ob- serve and count all interactions occurring be- tween two individuals. During agonistic contacts, the relative ranking of one individual relative to another was determined, according to the behav- iour displayed by both partners. Cases of equality were defined as occasions where encounters finished with neither individual clearly dominant Comparison of the social behaviour of captive sympatric peccary species 67 over the other. Most social behaviour involved only two individuals at the same time. Neverthe- less, for resting, all individuals composing the group rejoined by an animal were taken into ac- count without any further distinction. Data on be- haviour involving no partner, but still playing a social role, are taken from Dusost (1997). These included long-range communication and olfactory marking. Field data on these species have been collected during many expeditions to French Gui- ana between 1965 and 1996. Comparisons were made using X”, Student t test, or non-parametric Mann-Whitney U test. Corre- lations were examined non-parametrically using Spearman rank coefficient rs. Results Social organization Hierarchical relations within the group Here I consider both direct agonistic interac- tions, which indicate if one anımal is domi- nant or subordinate to another, and situa- tions where one individual avoids an interaction by modifying its initial direction or position to make way for another (e.g. standing up and leaving its resting place). As shown in figure 1, the dominance hierarchies are quite different for the two species. There is a clear dominance hierarchy within males and females in the CP. Dominance rank is inversely proportional to age. How- ever, the dominance hierarchy was less evi- dent in females than males (4 of equal rank out of 39 cases for females vs 0/24 for males). There were no cases where two adult males were apparently equal to each other during encounters. In contrast, young- er males show a high rate of equal rankings (Tab. 1). Between adults of different sexes, cases of equality are frequent (45.6% of en- counters), as they are between a juvenile male and an adult or a juvenile female (41.7-60%), or between two juvenile males or females (44.4-45.5% of cases). On the whole, cases of equality are far more fre- quent between partners of different sexes than within the same sex. Agonistic interactions between females are the most frequent (1.26 times the expected frequency computed using the ratio of fe- subadult ee Se 2 subsdiit 2 Kr U Ss E 24 d' S N young <-> Sound En nn nn nnn men u <> hr WLP au d n . ak <—- sie v yaungiß) inne young/®) Fig. 1. Schema of the hierarchical organization of col- lared (CP) and white-lipped peccaries (WLP), accord- ing to the relative importance of agonistic behaviours performed by one individual towards another (arrows). Dominant and subordinate: >2 years; subadult: >1 year; young: <1 year. male partners to the total number of poten- tial partners for this behaviour). Those be- tween a male and a female are the least common and occur approximately at the expected frequency. Females could be slightly dominant over males: adult females dominated 64.5% of the 31 contacts ob- served with adult males (X? = 0.80; df=1; not significant), and 78.6% of their 14 inter- actions with one juvenile male (x? = 1.80; df = 1; not significant); one juvenile female dominated one adult male in 2/4 cases. In the WLP, in contrast, all individuals of any age or either sex are subject to a single pre- cise hierarchical system, in which dominance 68 G. DuBosT Table 1. Percentage of agonistic encounters during which the two individuals are equal to each other. (CP - col- lared peccary; WLP - white-lipped peccary). Male-male interactions adult male-adult male adult male-juvenile male juvenile male-juvenile male Total Female-female interactions adult female-adult female adult female-juvenile female juvenile female-juvenile female Total Male-female interactions adult male-adult female 45.6% 3.2% adult male-juvenile female 0/4 0/0 juvenile male-juvenile female 60.0% ? adult female-juvenile male 41.7% Total (42/95) 44.2% 0.0% 0.0% 20.0% 0.0% 44.4% 0/0 (9/58) 15.5% (0/101) 0.0% 10.3% 0.0% 20.0% 0.0% 45.5% 1/1 (11/60) 18.3% (1/129) 0.8% 0.0% (5/207) 2.4% interactions are marked and frequent, even between individuals of very different ranks (Fig. 1). Agonistic interactions between two individuals generally reveal a clear domi- nance of one of them. Cases of equalities are non-existent or very rare within each sex (0/101 among males, 1/129 among females), as well as between adults of different sexes 6454 =3.2%% Tabl) There areitherefore no or very few events where the relative sta- tus of each individual is not respected. Thus, contrasts in frequencies of equalities be- tween the two species are always evident in each age or sex category. In contrast to the CP, the most frequent agonistic interactions in WLP occur be- tween males (1.39 times the frequency ex- pected) and the least frequent between fe- males (0.86 times the frequency expected). Adult males dominate females in 72.7% of the 154 cases observed (X? = 15.84; df=1; P< 0.001). As expected, status differences are less evident between juvenile or sub- adult males and adult females, appearing in only 60.4% of the 53 encounters (X? = 0.76; df = 1; not significant). In both peccary species, all 4 subordinate adult females were elevated to the top of the hierarchy immediately after giving birth. They subsequently became equal to the dominant CP female, and to the domi- nant WLP male. Comparative role of different individuals within the group Rates at which naso-body contacts, chin- layings, and mountings are seen in each so- cial category (in average number per ani- mal, per hour of observation and per poten- tial partner) are similar in the two peccary species (respectively, rs = 1.0, 0.95 and 0.90; P<0.01 and <0.05; n=6). These behav- jours are most frequent in adults: generally, rates are highest in the dominant male, next highest in females, third highest in subordi- nate males (Fig. 2). For mounts occurring outside oestrus, the hierarchical order is respected among CP males. The dominant male is implicated in such behaviour 3 times more often than sub- ordinate males: a rate of 0.027 vs 0.009 (U = 0; P = 0.005; nı = 4; n> = 6), and juveniles never. In the WLP, on the contrary, subordinates are involved as often or even more often than the dominant male: respectively 0.077 vs 0.067 (U=1; P<0.05; nn, =3;m=5) and the juveniles also participate in this behav- iour (0.018). Thus, mounting has a lower hierarchical value in WLP than in CP. Comparison of the social behaviour of captive sympatric peccary species 69 dominant male Ballogrooming Brest partners Oıchin laying Emounts Einaso-body contact Elmarking a partner dominant female oO a D oo > oo ke)} L ı N N n WLP Fig. 2. Number of social behaviours or rest partners recorded in the collared peccary (CP) and white-lipped pec- cary (WLP) for each member of the troop per animal, per hour of observation and per potential partner (mean + - standard error). The relative frequencies of other behaviour vary among the different social categories from one species to the other, but without any correlation (rs varying from 0.1 to 0.8; n=6; not significant). Nevertheless, some specific features appear, as shown in fig- urep2% Allogrooming and marking a partner are particularly indicative of the role played by an individual in group cohesion. The major role is held by the dominant male in the CP (respectively, 0.399 vs 0.041-0.238 and 0.558 vs 0.049-0.326 in the other animals: 07 78327 =0.01:n, = 4; 15-20) "and by the adult males and females in the WLP (re- spectively, 0.481-0.760 vs 0.152-0.217 and 0.468-0.802 vs 0.091-0.142 in the other ani- malsa0 6, P= 0.01: n] = 5;n,- 14). CP females show an increased frequency of social play and number of rest partners compared to males: respectively, 0.030- 0.033 vs 0.008-0.029, and 0.220-0.251 vs 0.080202935 U Z077P =0:.00E, Zn: n> = 13). In the WLP, subordinate males and juvenile females are more involved in these behaviours than are other individuals: respectively, 0.076-0.102 vs 0.047-0.084 (UZ22 3REEZ<0 DS In=i2), Fand 0.392-0.414 vs 0.211-0.341 (U=0; B=0. 00. 9; nm2-12): Agonistic behaviour is mainly performed by adults. Adult males and females are nearly equivalent in the CP: 0.073-0.081 vs 0.076 (U = 32; not significant; n, =8; m = 12). In the WLP, however, the dominant male is far more involved in such behaviour than the other adults: 0.288 vs 0.122-0.153 (U=0;P<0.01;n, =3;m = 11). In this spe- cies, the level of aggressiveness of each indi- vidual corresponds to its hierarchical rank. As shown in figure 2, subordinate CP males have a limited social role, often restricted to 70 G. DuBosT play or agonistic contact. They do not seem to have many partners for allogrooming, marking, resting or mounting. They live more or less as satellites to the troop. In contrast, subordinate WLP males are more involved in the various social behaviours than are the Tjuvenile=males (U 3: Pr0:0259n7 = 15-8) Amonezthez Wer males, they show the most intense contacts with the different partners of their social group, especially for allogrooming, play, resting and mounting, and they are very well integrated socially. Performer and receiver of social behaviour In both species, naso-body contacts are made mainly by the dominant male, the dominant female and the subordinate male; the dominant male is also the main receiver (46.1-54.5% of all behaviour performed). Thus, both species appear quite similar to each other (rs = 0.77; P - 0.05; n = 6). Social category and marking of partners also show similar associations in the two species (rs = 0.77; P - 0.05; n=6). This be- haviour is frequently performed by the sec- ond ranking female, the dominant of both sexes being the principal receiver (66.5- 77.4% of all behaviour performed). The initiation of allogrooming shows an op- posite association with social categories in the two species (rs = -0.83; P< 0.05; n=6). In the WLP, such initiations are common in the dominant female and the subordinate male, and the main receivers are the domi- nant male (26.0%) and the adult females (an average of 18.3% each). In the CP, on the contrary, initiations are chiefly per- formed by the dominant animals of both sexes, especially the dominant male, with- out any particular receiver. There are no noticeable differences be- tween CP individuals in rest partners, the dominant individual of each sex being both the principal donor and receiver. On the contrary in the WLP, subordinate adult males and juvenile females often join the other animals, particularly the dominant male (25.5%). Thus, the same individuals do not play the same social role in the two species, as shown by the lack of correlation between them (rs = 0.60; n = 6; not signifi- cant). The dominant male and the second ranking female are the chief performers of chin-lay- ing in the CP, the behaviour being mainly directed to the dominant female. In the WLP, this role is played by the dominant male and female, both also being the main receivers of such behaviour. Nevertheless, there is globally no similarity in the roles played by different individuals in the two species (rs = -0.01; n = 6; not significant). Mounts outside oestrus are usually per- formed by the subordinate male and by the dominant female in the CP, the dominant male being, curiously, a great receiver of such behaviour. In contrast, mounting is the prerogative of any male in the WLP, all the females being receiver animals. Thus, again, individuals behave differently in the two species and without any link between them (rs = 0.04; n = 6; not significant). Taking into account all these behaviours, there are many differences between the two species concerning the role played by different individuals within their social group. This is evident for the juvenile males, whose involvements in the different behav- iours are almost opposite in the two species (rs =-0.81; P - 0.05; n=6). The major ex- ception is the similarity in the social func- tions of subordinate females (rs = 0.83; P<0.05; n= 6). Nevertheless, some other features seem shared between both spe- cies. By calculating the ratio of the number of behaviours performed by an individual and the number received, it is possible to deter- mine if an individual is significantly a donor or receiver of a given behaviour (Fig. 3). On the whole, the dominant CP male is a re- > Fig. 3. Ratio of the number of behaviours performed to the number received by different individuals of the troop (mean + standard error) in the collared peccary (CP) and the white-lipped peccary (WLP). Ratio >1: the animal is chiefly a performer of the behaviour; ra- tio <1: the individual is a receiver. Comparison of the social behaviour of captive sympatric peccary species 71 naso-body contacts allogrooming Eidom. male Dad. female Pisub. male 02 Oly. male My. female social play agonistic behaviours mounts 72 G. DuBosT ceiver of partner marking (24/174: X” = 64.58; df = 1; P< 0.001) and of mount- ing (1/20: X? = 8.64; df=1; P< 0.01), but a donor of allogrooming (267/96: X” = 41.64; df= 1; P< 0.001). Conversely, the dominant WLP male is a receiver of allogrooming (72/267: X? = 59.90; d=1; P< 0.001) and a donor of mounting (31/8: X? = 6.18; df = 1; P<0.02). The dominant CP female is neither receiver nor donor of any behav- iour, whereas its WLP equivalent is a recei- ver of mounting (5/21: X°=4.16; df=1; P< 0.05) and of partner marking (118/233: X” = 19.30; df=1; P< 0.001). Thus, the so- cial role of the dominant animal of both sexes differs greatly between the species. The subordinate males of both species are more often donors than receivers. In the CP, they are donors of naso-body contacts (31/7: X?=7.08;, df=1; P<0.01) and of partner marking (33/10: X”=5.52; df=1; P < 0.02), but receivers of resting partners (8/26: X = 4.04: .df = 150 PJ=0:05) I Insthe WLEP, they are donors of naso-body contacts (66/15: X” = 16.46; df=1; P< 0.001), allo- groomings (252/114: X?=24.18; df=1]; P< 0.001), and mounts (45/5: X? = 17.20; di 1.720.001) Abutsneyermecewversuihe same is true for the juvenile males which are, in the CP, only donors of partner mark- ing (36/11:.X2= 6.06; di= 1; P=0.02), but, in the WLP, only donors of allogrooming (69/28: X°= 8.22; df=1; P<0.01) and of resting partners (53/13: X” = 12.04; df=1; P < 0.001). The subordinate females are as often recei- vers as donors of behaviours in both spe- cies. In the CP, they are receivers of allo- srooming, v (45/140: 1X7=25.02.7 di]; P< 0.001) and donors of partner marking (71/34: X? = 6.02; df=1; P<0.02); in the WLP, they are receivers of mounting (9/41: X? = 10.02; df=1; P< 0.01) and donors of partner marking (239119: X? = 20.00: df=1; P<0.001). The juvenile females are also both receivers and donors. In the CP, they receive allogrooming (31/72: X*=7.68; df=1; P<0.01) but carry out partner marking (35/7: x? =9.06; df=1:; P < 0.01); in the WLP, they receive mounts (3/22: X? = 10.66; df=1; P< 0.01), but are effectors of allogrooming (192/54: X°=40.80; df=1; P<0.001) and resting partners (135/52: "X=1846 di: P < 0.001). Except for mounts in both sexes of the WLP, there are no cases where both the dominant and the subordinates or juveniles of the same sex are both performers or re- ceivers of the same behaviour, and many similarities exist between subordinate and juvenile individuals in both species. In con- trast, the subordinate adult males tend to seek contact with more troop members in the WLP than do their equivalents in CP (U=2;5P=0.057:7n7 = A) Eihessame difference exists in partner choice for social play. Partner choice Twelve different pairings were identified in each species according to sex, age and so- cial status of the individuals. For each be- haviour, the mean number of observations made on each pair was calculated. The most active pairs (number of observations 250% that of the best pair) are particularly distin- guished in table 2. Looking at all pairs, there are detectable differences between the species in choice of partners for playing, allogrooming or resting. The differences are principally due to the fact that the dominant male and the adult females have more frequent contacts with juveniles of both sexes for allogrooming and rest in the CP than in the: WLP (U =: Z0: 29 m 22 2) In contrast, the subordinate WLP adult males tend to seek contact with more troop members than do their CP equiva- lents: (U =2;B=Z0.072:n7- m Asgalhe same difference exists in partner choice for social play. Asarule, young of both sexes are preferred play partners in both species. Nevertheless, the dominant CP male never plays with a juvenile female. Allogrooming is more frequent between adults than with or between juveniles in the-WEP (U = 0; P = 0.002, n7 40920); in contrast to the CP. Comparison of the social behaviour of captive sympatric peccary species 73 Table 2. Mean number of behaviours performed by each pair of partners, according to different social categories. *: number of observations equal to or greater than 50% that of the best pair. CP: collared peccary; WLP: white- lipped peccary. Pairs of partners Behaviour MALE-MALE INTERACTIONS dominant male-subordinate male naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts dominant male-juvenile male naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts subordinate male-juvenile male naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts FEMALE-FEMALE INTERACTIONS adult female-adult female naso-body contacts allogrooming marking a partner rest partners | social play | agonistic behaviours chin-laying mounts adult female-juvenile female naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts juvenile female-juvenile female naso-body contacts allogrooming 74 G. DuBost Table 2. (Continued). Pairs of partners MALE-FEMALE INTERACTIONS dominant male-adult female subordinate male-adult female juvenile male-adult female dominant male-juvenile female subordinate male-juvenile female Behaviour marking a partner rest partners social play agonistic behaviours chin-laying mounts naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts Comparison of the social behaviour of captive sympatric peccary species 75 Table 2. (Continued). Pairs of partners Behaviour juvenile male-juvenile female naso-body contacts allogrooming marking a partner rest partners social play agonistic behaviours chin-laying mounts In the CP, but not in the WLP, resting groups including juveniles tend to be more frequent than those comprising only adults (UF 362 0:055; n1-4;n2= 8). Thus, an adult CP female generally lies with her ju- veniles of different litters. The dominant male may take part in such groups, in con- trast to the WLP, where the dominant male rarely lies near a juvenile male or an adult female. In other behaviours both species show com- parable partner choices. Thus, significant si- milarities exist between CP and WLP for marking, chin-laying and agonistic behav- Ion («Ss 06: B= 0.055 n- 12) as well as for naso-body contact and mounting (es 10838001: n—12). In both species, naso-body contacts are more frequent between adults than with or between juveniles (U = 4; P = 0.024; n, = 4; mo srinsthesep:; ÜU=.0, P= 0.002: 177 =4; n„=8 in the WLP) and are chiefly per- formed by the subordinate males on the dominant, or by the dominant and subordi- nate males on adult females. Contacts be- tween adults are also or tend to be more frequent in both species for agonistic beha- Mouss (U 53 ,P 0.036.014; m -8 in ERBAUT 022000. = 4 m -& ın WLP), and for mounting (U =5; P = 0.036; ne 4:5 8 ın EP: U =6xP = 0.055,n; 4; n»=8in WLP). Marking a partner and chin-laying are also more frequent among adults in the WLP (a7 03B0:0025n7 = 4; nm, = 8), but not'in the PC. Partner marking occurs chiefly be- tween adults of the same sex in the WLP (dominant male with subordinates, adult fe- males with adult females), and between adults of different sexes (dominant male with females) in the CP. Apart from the dominant male-adult female pair in both species, chin-laying often occurs between adult females in the CP, and conversely be- tween adult males (dominant-subordinate) in the WLP. The most frequent agonistic contacts in the CP take place between males, or between a given adult female and another adult except the dominant male. In the WLP, most behaviour results from the dominant male performing agonistic acts towards subordi- nate males or adult females. An important difference between both spe- cies therefore resides in the fact that in the WLP the subordinate male is a favoured partner of the dominant male for many be- haviours, whereas in the CP the same role is held by the adult females. However, these specific preferences are in- verted for mounts occurring outside oestrus, 56.5% of them involving two males in the CP (the subordinate males mounting very often the dominant one), whereas, in con- trast, 87.3% of the mounts in the WLP oc- cur between partners of different sexes (48.1% between a subordinate male and an adult female). Thus, in both species mounts play a social role that differs from the other behaviours. 76 G. DuBost Frequency of interindividual contacts within the group Frequency of social behaviours Taking into account the average number of pooled social behaviour involving one given individual (per hour and per potential part- ner), it appears that each WLP has gener- ally 2 to 3times more social contacts with its partners than an equivalent CP (Tab. 3). On the other hand, both species are very similar in the relative frequencies of indivi- dual social behaviour (rs = 0.94; P = 0.01; n=6), although sexual behaviour is more frequent in the WLP. In both species, mark- ing a partner and allogrooming are the two most frequent social behaviours. Compared to the CP, this greater degree of social contact within a WLP group is also noticeable during rest periods. Each indivi- dual rests on the average with 0.35 part- ners/hour (number of rest partners ob- served/number of potential partners), compared with 0.20 in the CP. Thus, group sizes during resting periods are significantly 1.75 times larger in the WLP than in the CP (t = 6.02; df = 85; P < 0.001). This is the case despite the smaller number of individuals in the WLP groups (4-11 animals in WLP vs 7-16 in CP). Group cohesion The degree of grouping of individuals during activities is indicative of the cohesion level of each group. In this respect, several differences between the two species are apparent. In the CP, one or more individuals fre- quently remain apart from the group in a portion of the enclosure distant from the group’s location. These individuals are neither socially expelled nor actively chased out. At feeding time, they approach the food long after the others. In contrast, all the WLP individuals come together in a very homogeneous unit. During diurnal activities, the grouping of individuals is lower in CP than in WLP: in the former, only 3% of the individuals being separated from the group, compared Table 3. Average number of social behaviours re- corded for each individual, per hour and per potential partner. CP: collared peccary; WLP: white-lipped pec- cary. WLP WLP/CP 0.23 0.17 0.06 0.05 0.03 0.02 0.56 Marking a partner Allogrooming Agonistic behaviours Naso-body contact Social play Sexual behaviours 1.54 2.75 to 45% in the latter (X°=4.7; df=1; P<0.05). In the CP, this trend is more ap- parent during the afternoon than in the morning (1.6% isolated individuals against 4.4%: X” = 14.94; df=1; P< 0.001), in con- trast to the WLP whose percentages are quite similar for both periods (respectively 4.0% and 5.0%: X? = 0.47; df = 1; not signif- icant). This could mean that there are quali- tative changes in the CP behaviour throughout the day: more movements and foraging in the morning, and frequent social interactions in the afternoon. On the whole, CP shows also a weaker allo- mimetic behaviour than WLP: individuals perform a behaviour different from that of the rest of the group in 19.6% of the cases in the former, against 16.0% for the latter species (X” = 64.5; df = 1; P < 0.001). Unlike WLP, CP copy each other less during the morning than in the afternoon: respectively, 23.2% and 15.3% of the animals perform activities different from the rest of the troop (X = 19.39; df = 1; P< 0.001). During resting periods, the dominant male and female of both species are encountered more often alone than the juveniles: respec- tively 28.4-48.6% cases against 3.2-11.1% (X? = 28.6 to 52.5; df = 1; P < 0.001). Subor- dinate adult males usually rest more often alone in the CP than in the WLP (70.8% vs 30.9%: X” = 31.7; df=1; P< 0.001). The same is true for subadult males (20- 22 months old): CPs rest alone 4 times more often than the WLPs (30.3% vs 7.6%: Comparison of the social behaviour of captive sympatric peccary species 77 xX?=11.8; df=1; P< 0.001). On the other hand, subordinate WLP adult females rest alone more often than the CPs (43.9% vs 26.0%: X? = 6.95; df=1; P< 0.01). Integration of new members into the group During tests of sexual receptivity with penned animals, it was always possible to present a female to a new male without any risk in the CP. The opposite often occurs in the WLP: on several occasions under enclo- sure conditions, a male attacked a “new” fe- male so violently that we had to urgently re- move the female and abandon the test. Discussion The behavioural repertoires of these two forest peccary species are broadly similar to each other with few exceptions (DUBoSsT 1997). Contrary to many other suiforms which live either in permanent pairs or in unisex social units outside the reproductive period (BIGOURDAN 1948; GUNDLACH 1968; FRÄDRICH 1974; BEUERLE 1975; KırLrıE and TERBORGH 1983), mixed groups are the basis of the peccary social system. This fact is fa- voured by the persistence of sexual activity throughout the year. There is never any sex- ual segregation, even before or after birth. The two species mainly diverge in the fre- quency or mode of appearance of behav- iours rather than in any real inequality in behavioural repertoire. However, some qualitative or quantitative differences, espe- cially concerning social interactions, are sometimes significant enough to reveal a true divergence in their biology. These fol- low directly from specific etho-ecological characteristics of each species. Behavioural differences between both species In both species, young or subordinate ani- mals of both sexes frequently mark part- ners. This observation, also made by SCHMIDT (1976) for the CP, contradicts the assertion of Sowıs (1974) that dominant anımals are the main effector of such behaviour. Likewise, allogrooming in the WLP is commonly initiated by juveniles of both sexes and subordinate males, and to a lesser degree by dominant females. In the CP, on the contrary, this behaviour is chiefly performed by the dominant animals of both sexes, especially the dominant male. This fact, also noted by ScHmipr (1976), distin- guishes the CP from most other mammals where dominants are generally groomed by subordinates, e.g. Papio hamadryas (KuM- MER 1968), Bos taurus (WALTHER 1979). Contrary to evidence on several ruminants (WALTHER 1979), mounts in peccaries do not have any aggressive significance; they do not play the same role in both species. In male CPs they occur in accordance with hierarchical order, but not so in the WLP. This means that mounting has a different social role and a lower hierarchical value in WLPSs than in CPs. As previously noted by Sowrs (1974) and ScHMIDT (1976), female CPs are slightly dominant to males, but in the WLP, males dominate females, as in several suids (FRÄ- DRICH 1965) as well as in ruminants living in mixed herds - e.g. Taurotragus, Bison, Syncerus (Estes 1974; WALTHER 1979). However, in both peccary species, subordi- nate females are elevated immediately after giving birth to the top of the hierarchy, a fact also observed by SCHWEINSBURG and Sowrs (1972) in the CP. In the CP, but not in the WLP, resting groups often include an adult female with her juveniles of different litters, which led SCHMIDT (1976) to use the term “clan”. Furthermore, according to LOCHMILLER and GRANT (1982), groups of CP are highly bonded units, whose members are intoler- ant of strangers of the same sex. As re- ported by ScHmipt (1976) and ByErs and BEKOFF (1981), individuals which were tem- porarily separated from the group lost their former social status and were never com- pletely reintegrated. However, this exclu- sion does not seem to exist between sexes, 78 G. DuBost because it was always possible to present a CP female to a new male without any risk, contrary to the WLP. We do not know if the intolerance observed in the latter spe- cies is triggered by the odour of the stran- ger and corresponds to a type of group de- fence, but it fits well with the high level of natural agressiveness observed in this spe- cies (DugBost 1997). Social behaviour in relation to group size WLP live in large herds, comprising gener- ally from 30 to 200 animals, and including many adult males, females, and juveniles of all ages (KırrıE and TERBORGH 1976, 1983; Sowıs 1984; MAYER and WETZEL 1987; BE- NIRSCHKE et al. 1989; HERNANDEZ et al. 1995; PEREs 1996; FRAGoso 2000; JUDAS pers. comm.; pers. data). Such social struc- ture appears relatively unique among the artiodactyls, permanent aggregations of so many mature individuals of both sexes being only known seasonally in several mi- gratory, open country ruminants, such as Antilope cervicapra (MunGALL 1978), Con- nochaetes taurinus (ESTEs 1969) and others (WALTHER 1979), but exceptionally also among some forest suids, like Sus barbatus (PFEFFER 1959; FRÄDRICH 1974; CALDECOTT et al. 1993). Conversely, the social unit of the CP is gen- erally composed of a limited number of in- dividuals: 1-2 adult males and 1-3 adult fe- males, wıth several young of different ages. Such small groups are found in French Gui- ana (JUDAsS pers. comm.) as in other forest regions (KırLTIE and TERBORGH 1976, 1983; RosBInson and EISENBERG 1985; BODMER et al. 1988; PEREs 1996). When individual WLPs search for food, they are frequently at a distance from each other, the whole herd being spread out over several tens or hundreds of meters. Because the physical forest environment is largely obstructed at ground level, individuals need mechanisms for intercommunicating efli- ciently at short and medium distances to ensure herd cohesion. In contrast to CP, vi- sual, acoustic or olfactory signals are parti- cularly well developed in WLP, both in ex- pression and in intensity: spectacularly bristled hairs; prolonged yawning as intimi- dation, loud blowing, grunts, teeth snaps or cries of the young (KırrıE and TERBORGH 1976, 1983; SowLs 1984; MAYER and WETZEL 1987; HERNANDEZ et al. 1995; DwuBosT 1997); strong odor (BENIRSCHRE et al. 1989; Dusosrt 1997). The WLP is heavier than the CP: a mean of 37.1kg for the adult males and 35.7 kg for the adult females in French Guiana, versus, respectively, 22.1kg and 19.5kg (Dusost 1997). Unlike the CP, the large body size of individual WLPs belonging to a well- populated and powerful herd provides pro- tection against potential predators and dis- penses with the need for cryptic behaviour. This species has the reputation of attacking jaguars en masse (KırrıE and TERBORGH 1983). Herd cohesion of the WLP must also be fa- cilitated by the behaviours themselves. The results of this study indicate that social be- haviours are from 2.5 to 6.4 times more frequent in this species than in CP. Like- wise, during resting periods, the grouping of individuals at the same spot is almost twice as high in the former than in the lat- ter. Furthermore, in comparison with the CP, the WLP shows a generalization of some behaviours, which are performed by most members of the herd, rather than just by one or several individuals. Such behav- iours include collective fear, body rubbing on the ground in all individuals, play often being contagious, penis trembling and mount by all males. Several other behav- iours occur in a very demonstrative man- ner, including strong reactions to anxiety situations, urinary marking on the standing female, play with objects, marked body dis- plays in threat or submission situations, frequent and strong attacks (DuBost 1997): In the WLP, the involvement of most individuals of the same herd in different so- cial behaviours is also noticeable in sexual behaviour. In this species, all the numerous adult or subadult males are reproductive and may copulate with the females, regard- Comparison of the social behaviour of captive sympatric peccary species 79 less of their relative hierarchical position (pers. obs.), as described in some other mammals (Smuts 1987). On the contrary, only the dominant male is reproductive in the CP. It is the only one to court and copulate with the few reproduc- tive females of the group (BissonETTE 1976; ScHMIDT 1976; pers. obs.). All the other males show much lower sexual hor- mone levels (HELLGREN, in HAnnon et al. 1991; pers. data); they avoid the proximity of the receptive females. Unlike most other artiodactyls and many mammals, they were never observed to be pushed away by the dominant male, as also noted by Sows (1974), ScHhmipt (1976), and ByErs and BEKOFF (1981). Furthermore, they do not leave the social unit to live alone or grouped together in a bachelor herd, like Sus scrofa and Phacochoerus (BIGOURDAN 1948; FRÄDRICH 1974; BEUERLE 1975; BARR- ETTE 1986), the tylopods and many rumi- nants (Kororp 1961; Davıp 1973; EsTEs 1974; FRANKLIN 1974; GosLinG 1974; JOU- BERT 1974; SPINAGE 1974; MunGALL 1978), and several FAquus species (KLinGeL 1974). The fact that these surplus adult males can stay inside the social unit is perhaps due to their sexual inactivity, provided they behave submissively, as in Hippopotamus (KLin- GEL, in ELTRINGHAM 1993) or Hylochoerus (D’HUART 1993). The life within a rather considerable herd does not allow individual WLPs to estab- lish such an elaborate and fine contact with each other as in the CP, whose groups func- tion as very well coordinated and stable units. In the WLP, there are, indeed, fre- quent encounters between animals with lit- tle knowledge of each other, each indivi- dual having to define its own place in relation to its partner. One can understand why, contrary to the CP, most interindivi- dual contacts, even sexual ones, are of a very aggressive nature in the WLP, as noted also by FrRÄDRICH (1986) and BENIRSCHRE et al. (1989). In contrast, there are pro- nounced displays for appeasement and sub- mission purposes (Dugost 1997). Likewise, this species shows a rather strict social or- ganization, where all individuals of both sexes are included in the same linear hier- archy. Finally, since social units need to re- main distinctive within a complex herd, its members are forced to stay close to each other. This would function to strengthen the bonds uniting them and to guarantee their relative isolation within the herd, when necessary, as displayed by the female WLP with her young (Dusost 1997). Thus, the WLP herd can be considered as a multi-male society, whereas the social unit of CP corresponds more to a harem or pseudo-harem. Moreover, some interme- diate situations between these specific so- cial organizations have been observed. In- deed, the formation of a harem in the CP seems directly dependent on the group size. Sowıs (1974), ByErs and BEKOFF (1981), PACKARD et al. (1991) observed, both in cap- tivity and in nature, instances of sexual promiscuity within large groups. Differences of social life between wandering and locally fixed groups Another main difference between both spe- cies lies in the fact that herds of WLP travel more or less constantly throughout a huge area. In the absence of precise biological data, this species was considered nomadic and capable of travelling great distances (KırrıE and TERBORGH 1983; SowLs 1984; HERNANDEZ et al. 1995; PERES 1996). But recently, FRAGOSO (2000) produced data in- dicating that herds of WLP can live for a long time on vast home ranges of 22- 110 km”, where they move over long dis- tances but do not migrate. Quite the con- trary, each group of CP is permanently at- tached to a well-defined terrain, covering an average area of only 190 ha in French Guiana (JuDAs pers. comm.). One finds here the classical link uniting the social be- haviour of a species wıth the characteristics of its environment, as noted in many other artiodactyls. Life in a moving herd does not require to deposite marks on the ground. Thus, the WLP does not show the behaviours that al- low the CP, especially the dominant male, 80 G. DuBosT to assert its presence in the area where the group resides. Indeed, the WLP shows no sign or behaviour indicating any direct bond with the ground occupied, like defe- cation site or glandular marking. The home range of a WLP herd studied by FRAGOSO (2000) was almost completely encompassed by that of another, and the two herds were found together several times at the same site. The WLP sociality is thus nearly exclu- sively dependent on the exchanges existing between the individuals themselves; inter- individual contacts are particularly fre- quent in this species. It ıs also significant that WLP males never urinate on a female lying on the ground, as do CP males, but only on standing or moving females (Du- BOST 1997). Finally, the more or less continuous travel- ling of a WLP herd demands that indivi- duals be well synchronized during their ac- tivities. This is achieved by mass effects and allomimetic behaviour. In counterpart, the great social cohesion, as observed also by FrAGoso (2000) in the wild, engenders a marked intolerance towards strangers. On the other hand in the CP, each home range is generally distinct and belongs to only one social group, even if neighbouring home ranges overlap greatly. Each home range has several rest places (HERNANDEZ et al. 1995; JuDas pers. comm.) and defeca- tion sites (HERNANDEZ et al. 1995; pers. obs.). According to BissonETTE (1976), the home range ıs regularly marked by the lo- cally resident adult male which asserts both his status and presence by many behav- iours, including surveillance, defence of the group in case of danger, display with bristled hairs and ceremonial gait, continu- ous olfactory control of all partners, emis- sion of urine when walking, scraping and glandular marking of the ground and drop- ping of faeces on distinctive places (Du- BosT 1997). The behaviour of the CP ap- pears therefore comparable to that found in many ruminants, and perhaps also in hip- popotamuses and several suids (FRÄDRICH 1974). The harem or pseudo-harem of the CP dif- fers from those of many ruminants by in- cluding several subordinate adult males (Davınp 1973; Estes 1974; GosLinGs 1974; JOUBERT 1974; SPINAGE 1974; MUNGALL 1978). The fact that it stays constantly with the same dominant male within one locally fixed home range differs also from the ru- minant harem, but resembles what is known in Vicugna (KoForD 1961; FRANKLIN 1974). Nevertheless, this similarity is only appar- ent, because the harem of Vicugna lives on two small territories separated from each other by neutral ground. Most behavioural differences between both peccary species thus appear to be the direct result of the way of life adopted by each. In this context, it is reasonable to suppose that the Chacoan peccary, Catago- nus wagneri, shows a social organization si- milar to that of the collared peccary, owing to the great resemblances of these species in both behaviour and group composition (MAyER and BrAnDT 1982; MAYER and WETZEL 1986). Similar variations in social behaviour could exist in Old-World primates. Indeed, between primate species living in multi- male troops and those forming harems, obvious differences exist in the size of the social units, level of sociality, marking, function of the dominant male as the nu- cleus of the group, nature of the male-fe- male relations, and hierarchical system (RowELL 1988). Acknowledgements This study was totally dependent on the facilities provided by the Institut National de la Recherche Agronomique. Many thanks to Dr P. PLANQUETTE for having facilitated this research, to C. DUTER- TRE, P. LAVENTURE, F. PERIACARPIN, and D. TORVIC for having assured the success of this peccary breeding, to J. Jupas for providing unpublished field observationss, and to PDrs J. BRADBURY, R. Day, C.-A. GAUTIER, OÖ. HENRY, R. MAUGET, J.J. MAyER, K.PAcK, M.SCHLEE and an anon- ymous reviewer for their helpful comments and corrections of the manuscript. Comparison of the social behaviour of captive sympatric peccary species 81 Zusammenfassung Vergleiche des Sozialverhaltens zweier sympatrischer Pekariarten (Genus Tayassu) in Menschenobhut; Beziehungen mit ihren ökologischen Merkmalen Die Vergleiche des Sozialverhaltens wurden zwischen dem Weißbart-Pekari (Tayassu pecari) und dem Halsband-Pekari (Tayassu tajacu) in Menschenobhut durchgeführt, die beide im südamerikanischen Regenwald leben. Beim Weißbart-Pekari gibt es einen engen Rudelzusammenhalt und Gruppen- fremde werden im allgemeinen heftig angegriffen. Die Individuen haben zwei- bis dreimal mehr So- zialkontakt mit Gruppenmitgliedern als die Halsband-Pekaris. In vielen sozialen Verhaltensweisen ist beim Weißßbart-Pekari das dominante Weibchen das aktivste Individuum, während beim Hals- band-Pekari das ranghöchste Männchen das meistbeachtete Mitglied der Sozialeinheit ist. Randnie- dere und subadulte Männchen beteiligen sich bei allen sozialen Interaktionen und sind bestens in das Rudel integriert. Beim Halsband-Pekari dagegen leben rangniedere Männchen mehr oder weni- ger als neutrale und periphäre Individuen. Völlig verschieden verhält sich das Weißbart-Pekari mit beträchtlicher Aggressivität in der Mehrzahl der Auseinandersetzungen, wobei das ranghohe Männ- chen der Hauptinitiator der häufigen und intensiven agonistischen Interaktionen ist. Bei dieser Art unterliegen beide Geschlechter einer einzigen, alle Rudelmitglieder einschließenden Rangordnung, in der die Männchen den Weibchen stets überlegen sind. Beim Halsband-Pekari dagegen gibt es zwei getrennte, geschlechtsspezifische Rangordnungen und beim zwischengeschlechtlichen Kontakt dominieren die Weibchen über die Männchen. Diese zwischenartlichen Unterschiede und das völlige Fehlen einer Bodenmarkierung beim Weißbart-Pekari verdeutlichen die ökologischen Merkmale bei- der Arten. Während das Weilbart-Pekari in großen Mehrmännchen-Rudeln umherzieht, lebt das Halsband-Pekari in kleinen stabilen und lokal seßhaften Rotten. References BARRETTE, C. (1986): Fighting behaviour of wild Sus scrofa. J. Mammalogy 67, 177-179. BENIRSCHKE, K.; ByrD, M.L.; Low, R.J. (1989): The Chaco Region of Paraguay. Peccaries and Mennonites. Interdisc. Science Rev. 14, 141- 147. BEUERLE, W. (1975): Freilanduntersuchungen zum Kampf- und Sexualverhalten des euro- päischen Wildschweines (Sus scrofa L.). Z. Tierpsychol. 39, 211-258. BIGOURDAN, J. (1948): Le Phacochere et les Suides dans l’Ouest africain. Bull. Inst. Franc. Afr. Noire 10, 285-360. BIsSONETTE, J. A. (1976): The relationship of re- source quality and availability to social behav- iour and organization in the collared peccary. Diss. thesis, University of Michigan. BODMER, R. E. (1989): Frugivory in Amazon un- gulates. Diss. thesis, University of Cambridge. BODMER, R. E.; Fang, T. G.; MoyA, L. (1988): Es- tudio y manejo de los Pecaries (Tayassu tajacu y T. pecari) en la Amazonia peruana. Notas cientificas/Matero, Ene, 18-25. Byers, J. A.,;, BEKOFF, M. (1981): Social, spacing, and cooperative behaviour of the collared peccary, Tayassu tajacu. J. Mammalogy 62, 767-785. CALDECOTT, J. O., BLOUCH, R. A.; MACDONALD, A.A. (1993): The bearded pig (Sus barbatus). In: Pigs, Peccaries, and Hippos. Status Survey and Conservation Action Plan. Ed. by W.L.R.OLıer. Gland, Switzerland: RHEIN! Davıp, J. H. M. (1973): The behaviour of the Bon- tebok, Damaliscus dorcas dorcas (Pallas 1766), with special reference to territorial be- haviour. Z. Tierpsychol 33, 38-107. D’HuarT, J.P. (1993): The forest hog (Aylo- choerus meinertzhageni). In: Pigs, Peccaries, and Hippos. Status Survey and Conservation Action Plan. Ed. by W.L.R. OLıver. Gland, Switzerland: I. U.C.N. Dusost, G. (1997): Comportements compares du Pecari a levres blanches, Tayassu pecari, et du Pecari ä collier, T: tajacu (Artiodactyles, Tayassuid&s). Mammalia 61, 313-343. ELTRINGHAM, S.K. (1993): The common hippopo- tamus (Hippopotamus amphibius). In: Pigs, Peccaries, and Hippos. Status Survey and Conservation Action Plan. Ed. by W.L.R. OLIVER. Gland, Switzerland: I. U.C.N. Estes, R. D. (1969): Territorial behaviour of the wildebeest (Connochaetes taurinus Burchell, 1823). Z. Tierpsychol. 26, 284-370. 82 G. DuBost Estes, R.D. (1974): Social organization of the African Bovidae. In: The Behaviour of Ungu- lates and its Relation to Management. Ed. by V. GEıst and F. WALTHER. Morges, Switzer- land UÜLEN: FRÄDRICH, H. (1965): Zur Biologie und Ethologie des Warzenschweines (Phacochoerus aethio- picus Pallas) unter Berücksichtigung des Ver- haltens anderer Suiden. Z. Tierpsychol. 22, 328-393. FRÄDRICH, H. (1974): A comparison of behaviour in the Suidae. In: The Behaviour of Ungulates and its Relation to Management. Ed. by V. Geist and F. WALTHER. Morges, Switzer- land: I. U.C.N. FRÄDRICH, H. (1986): Schweine als Zootiere. Zool. Garten 56, 7-19. FRAGOSO, J. M. V. (2000): Home range and move- ment patterns of white-lipped peccary (Tayas- su pecari) herds in the northern Brazilian Amazon. Biotropica (in press). FRANKLIN, W.L. (1974): The social behaviour of the Vicuna. In: The Behaviour of Ungulates and its Relation to Management. Ed. by V. Geist and F. WALTHER. Morges, Switzer- land: I. U.C.N. GosLiInG, L.M. (1974): The social behaviour of Coke’s Hartebeest (Alcelaphus buselaphus cokei). In: The Behaviour of Ungulates and its Relation to Management. Ed. by V. GEIST and F. WALTHER. Morges, Switzerland: I.U.C.N. GUNDLACH, H. (1968): Brutfürsorge, Brutpflege, Verhaltensontogenese und Tagesperiodik beim europäischen Wildschwein (Sus scrofa L.). Z. Tierpsychol. 25, 955-995. HANNoNn, P.G.;, DowDeLL, D. M.; LOCHMILLER, R. L.; GRANT, W. E. (1991): Dorsal-gland ac- tivity in peccaries at various physiological states. J. Mammalogy 72, 825-827. HERNANDEZ, O. E.; BARRETO, G. R.; OJASTI, J. (1995): Observations of behavioural patterns of white- lipped peccaries in the wild. Mammalia 59, 146-148. JOUBERT, S. C. T. (1974): The social organization of the Roan Antelope Hippotragus equinus and its influence on the special distribution of herds in the Kruger National Park. In: The DBehaviour of Ungulates and its Relation to Management. Ed. by V. GEIsT and F. WALTHER. Morges, Switzerland: 1.Ü:@N. Kırrıe, R. A. (1983): Observations on the behav- iour of Rain Forest Peccaries in Peru: Why do White-lipped Peccaries form herds? Z. Tier- psychol. 62, 241-255. Kite, R. A.; TERBORGH, J. (1976): Ecology and behaviour of Rain Forest Peccaries in south- eastern Peru. Nat. Geog. Res. Rep., 873-882. KLinGeL, H. (1974): A comparison of the social behaviour of the Equidae. In: The Behaviour of Ungulates and its Relation to Manage- ment. Ed. by V.Geıst and F. WALTHER. Morges, Switzerland: I. U.C.N. KoFoRD, C. B. (1961): The ecology and manage- ment of the Vicuna in the puna zone of Peru. Terre Vie 108, 342-352. KUMMER, H. (1968): Social organization of Hama- dryas Baboons. A field study. Basel: S. Kar- ger. LOCHMILLER, R. L.; GRANT, W. E. (1982): Intraspe- cific aggression results in death of a Collared Peccary. Zoo Biol. 1, 161-162. MAYER, J. J.; BRANDT, P.N. (1982): Identity, distri- bution, and natural history of the peccaries, Tayassuidae. In: Mammalian Biology in South America. Special Publ. Ser., Pymatuning Lab. Ecol., Univ. Pittsburgh. Ed. by M. A. MARES and H.H.Genoways. Pennsylvania: Pitts- burgh Press. MAYER, J. J.; WETZEL,R.M. (1986): Catagonus wagneri. Mammalian Species 259, 1-5. MAYER, J. J.; WETZEL, R.M. (1987): Tayassu pe- cari. Mammalian Species 293, 1-7. MunGALL,E.C. (1978): The Indian Blackbuck Antelope: A Texas View. Kleberg Studies in Natural Resources. Ed. by CAESAR KLEBERG. PACKARD, J. M.; BABBITT, K.J., FRANCHEK, K. M.; PIERCE, P.M. (1991): Sexual competition in captive collared peccaries (Tayassu tajacu) Appl. Anim. Behav. Sci. 29, 319-326. PERES, C. A. (1996): Population status of white- lipped Tayassu pecari and collared peccaries T. tajacu in hunted and unhunted amazonian forests. Biol. Conserv. 77, 115-123. PFEFFER, P. (1959): Biologie et migrations du Sanglier de Born&o (Sus barbatus Muller 1869). Mammalia 23, 277-303. RoBINSoN, J. G.; EISENBERG, J. F. (1985): Group size and foraging habits of the collared pec- cary Tayassu tajacu. J. Mammalogy 66, 153- 159. RoweELı, T. E. (1988): The social system of gue- nons, compared with baboons, macaques and mangabeys. In: A Primate Radiation. Evolu- tionary Biology of the African Guenons. Ed. by A. GAUTIER-Hion, F. BOURLIERE, J.-P. GAU- TIER and J. Kınapon. Cambridge: Cambridge University Press. SCHMIDT, C. R. (1976) Verhalten einer Zoogruppe von Halsband-Pekaris (Tayassu tajacu). Diss. thesis. Universität Zürich. Comparison of the social behaviour of captive sympatric peccary species 83 SCHWEINSBURG, R. E.; SowLs, L.K. (1972): Ag- gressive behaviour and related phenomena in the Collared Peccary. Z. Tierpsychol. 30, 132- 145. SMuTs, B. B. (1987): Sexual competition and mate choicer In SPrimate, Societies, Ed. by B. B. SmuUTs, D.L. CHENEY, R.M. SEYFARTH, R. W. WRANGHAM, and T. T. STRUHSAKER. Chi- cago: University Chicago Press. Sowrs, L. K. (1974): Social behaviour of the Col- lared Peccary Dicotyles tajacu (L.). In: The Behaviour of Ungulates and its Relation to Management. Ed. by V.Geit and F. WALTHER. Morges, Switzerland: I. U.C.N. Sowıs, L.K. (1984): The Peccaries. Tucson, Ari- zona: University of Arizona Press. SPINAGE, C. A. (1974): Territoriality and popula- tion regulation in the Uganda Defassa Water- buck. In: The Behaviour of Ungulates and its Relation to Management. Ed. by V. GEIST and F WALTHErR. Morges, Switzerland: IUSE>N! WALTHER, F. (1979): Das Verhalten der Hornträ- ger (Bovidae). Hb Zoologie 8 (54), 1-184. Author’s address: G. Dusost, Laboratoire de Conservation des Especes animales, Menagerie du Jardin des Plantes, 57 rue Cuvier, 75005 Paris, France (e-mail: gdubost@mnhn fr). Mamm. biol. 66 (2001) 84-89 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Original investigation Mammalian Biology © Zeitschrift für Säugetierkunde Biochemical identification of three sympatric Apodemus species by protein electrophoresis of blood samples By BRIGITTE A. REuTTER, H. BRÜNNER, and P. VoGEL Institute of Ecology, University of Lausanne, Lausanne, Switzerland Receipt of Ms. 28. 02. 2000 Acceptance of Ms. 20. 09. 2000 Abstract The allelic pattern of seralbumine and general protein 1 of the three sympatric Apodemus species Apodemus sylvaticus, A. flavicollis, and A. alpicola, were studied using electrophoretic analysis of blood samples. This method appears to be a sensitive tool for distinguishing the three Apodemus species in the Alps. Their identification on the basis of external characteristics in the field is some- times extremely difficult, even more so for juvenile specimens. Compared to previously described methods the electrophoretic analysis does not require killing animals and can be used on juveniles. Key words: Apodemus, rodents, protein electrophoresis, biochemical identification Introduction The determination of the two common mouse species in western Europe, Apode- mus sylvaticus (Linnaeus, 1758), and A. fla- vicollis (Melchior, 1834), is not always easy because their morphological characters strongly overlap. Some individuals or popu- lations suggested hybridisation or introgres- sion (ENGEL et al. 1973), but allozymatic studies did not reveal any hybrids (ENGEL et al. 1973; DeBror and MErMmoD 1977; NIETHAMMER and Krapp 1978; GEMMEKE 1980; BENMEHDI et al. 1980; NAscETTI et al. 1980; CsAaıKkL et al. 1980; GEMMEKE and NIETHAMMER 1981; FRAGUEDAKIS-TSOLIS et al. 1983; Nascertı and Fırippuccı 1984; GEBCZYNSKI et al. 1986). Since the recognition of a third sympatric species, A.alpicola Heinrich, 1952, by 1616-5047/01/66/02-084 $ 15.00/0. StorcH and Lürtr (1989) with intermediate morphological traits, the discrimination became even more problematic. A better accuracy of species identification was ob- tained by a discriminant function developed from a limited number of skull measure- ments (REUTTER et al. 1999). Six cranial characters are sufficient to differentiate be- tween the three Apodemus species with a correct classification above 97%. While, this technique is indeed a good tool for re- classifying museum material, it does not overcome the determination problem of young animals and of living individuals dur- ing field studies. A discrimination independent of morphol- ogy should be based on genetic markers, e.g. specific allozyme pattern. The analysis Biochemical identification of three sympatric Apodemus species 85 of allozyme variation by starch gel electro- phoresis has frequently been used in genet- ic and systematic investigations of the genus Apodemus (ENGEL et al. 1973; BEHNMEHDI et al. 1980; CsaıKL et al. 1980; GEMMEKE 1980; FRAGUEDAKIS-TSOLIS et al. 1983; GEM- MEKE 1983; KORPIMÄKI and NORRDAHL 1987; FıLippuccı et al. 1989; FERNANDES et al. 1991; BRITTON-DAVvIDIAN et al. 1991; VAPA et al. 1995). VoGEL et al. (1991) and FıLıp- puccı (1992, 1996) included A. alpicola in their allozyme analysis and confirmed the specific status of this species. DEBROT and MERMmoD (1977) found that the seralbumine pattern obtained by polyacrylamide gel electrophoresis is very distinctive for A. syl- vaticus and A. flavicollis. The aim of this study is to develop a techni- que applicable to all age cohorts of the three sympatric Apodemus species, A. syl- vaticus, A. flavicollis, and A. alpicola, based on blood samples without need to sacrifice anımals, in analogy with the techniques used for sibling species of shrews (HAUSSER and ZUBER 1983; BRÜNNER 1988; NEET 1989; NEET and HAussER 1989, 1990, 1991; BRÜNNER and NEET 1991; TURNI and SCHÖN- HERR 1994). Material and methods Electrophoretic analysis was carried out on 41 in- dividuals of the three species Apodemus sylvati- cus (n= 15), A. flavicollis (n= 18), and A. alpico- la (n=8) from 11localities in Switzerland and neighbouring Italy. Localities and collection num- bers (IZEA: Institut de Zoologie et Ecologie An- imale) of animals investigated are presented in the following list: Apodemus sylvaticus Linnaeus, 1758. Switzerland: Bern: Haslital: (IZEA, 7381, 7382, 7383); Valais: Ayer: (7379), Monnaz: (4887, 4888); Vaud: Aclens (7380); Echichens: (4880, 4 884); Morges: (4886); Renens: (7363, 7366); St. Saphorin: (4883, 4885). Italy: Domodossola: (7 384). Apodemus flavicollis Melchior, 1834. Switzerland: Bern: Haslital: (7389, 7390, 7392, 7393); Valais: Monnaz: (4894, 4895); Vaud: Aclens (7401); Le Brassus: (7395, 7396, 7397, 7398); Echichens: (4890); Morges: (4891); Renens: (7364); St. Sa- phorin: (4881, 4882, 4889, 4892). Apodemus alpicola Heinrich, 1952. Switzerland: Valais: Sanetsch: (7337, 7338, 7339, 7345, 7 346, 7347, 7348, 7361). From every individual a blood sample of about 2 ul was taken from the base of the tail of the ani- mals with heparinized Micro-Hematocrit tubes. The incision of the caudal vein with a razor blade (or another sharp blade) to get blood samples (maximum 0.1 ml) is recommended for mice by the Swiss Federal Office of Veterinary. By being transferred into an Eppendorf tube, the blood could be kept for more than 10 hours at ambient temperature (20°C) or a week at a cool place (4°C), and at least one year and even longer at -20°C. This circumstance allows the use of ani- mals for the electorphoretic analysis that were frozen a few hours after death. Blood samples were diluted in a solution (1:5) of saccharose (40%) and 0.075M Tris/HCl buffer, pH 8.9 with a trace of bromophenol blue. The added amount of the saccharose-buffer solution depended on the volume of the blood sample. For example, to a 2 ul blood sample 8 ul saccha- rose and 50 ul buffer solution were added. The samples were then run in a Polyacrylamid-disc- electrophoresis (resolution gel: 8%, 0.325 M Tris/ HCl, pH 8.9; concentration gel: 3%, 0.056 M Tris/ HCl, pH 6.9; running buffer: 0.05M Tris/HCl, 0.38M glycerine) with a constant power of 4W during migration in the concentration gel and 12W in the resolution gel (gel size: 180x 155x 15 mm; power supplies: Bio Rad 3 000/ 300 and 1000/500; gel support: Zabona AG, Basel). Proteins migrated from cathode to anode during 4-5 h (band of bromophenol blue at 1 cm from lower gel border). Proteins were non-differ- entially stained with Coomassie blue (0.025% Coomassie blue R250, Sigma; 50% methanol; 3.5% glacial acetic acid) for Ih and destained afterwards (50% methanol; 3.5% glacial acetic acid). The method is slightly modified according to HAUSSER and ZUBER (1983). Relative migrating distances of the proteins in re- lation to the bromophenol blue dye front were calculated to identify clearly the different protein bands. All skulls of the examined specimens were pre- pared, measured, and assigned to one Apodemus species by using a discriminant function analysis (REUTTER et al. 1999). To test our results we included 58 unknown Apo- demus specimens from the Bündner Natur-Mu- seum Chur. Allthese animals came from the east- ern part of the Swiss Alps (Graubünden). Blood samples were taken with heparinized Micro-He- matocrit tubes from the defrosted bodies. 86 BRIGITTE A. REUTTER et al. Results The results of the electrophoretic analysis (Fig. 1) are best understood when consider- ing the different allelic pattern of the albu- min and an unknown general protein 1 (GP 1). These two proteins are represented on the gel by three different bands which are labelled from cathode to anode as A, B, and C. Apodemus flavicollis and A. alpi- cola are characterised by the slower migrat- ing albumin allele (band A), whereas the faster migrating albumin allele (band B) is present only in A. sylvaticus. Moreover, A. flavicollis shows an additional protein band C (GP 1), which migrated further than either A and B. Nothing is known about the identity of GP 1, its possible polymorphism and the position of other allelic bands in A..alpicola and A. sylvaticus. However, the presence of the characteristic GP 1 band in A. flavicollis allows the distinction between this species and A. alpicola. When the migration distance of band A is AF AS AA AF AS Table 1. Protein markers and bands (Alb and GP 1) of the three species A. sylvaticus (B), A. flavicollis (A, C) and A. Alpicola (A). marker band A. sylva- ticus A. flavi- collis A. alpicola taken as 100%, the relative distances are for bandB 101.6% (101.3-102.0) and for band C 107% (105.2-108.5) (Tab. 1). In that way 100% of all examined specimens could be determined unequivocally. All 58 testspecimens from Graubünden could be identified according to the protein electrophoresis of blood samples. 52 indivi- duals were assigned to A. sylvaticus, five to A. flavicollis and only one to A. alpicola. Skull measurements using a discriminant function (REUTTER et al. 1999) confirmed this determination. AA AF AS AA AF Albumin A B A GP1 C a» B A A B A A | 6 C Fig. 1. The serum albumin and general protein (GP 1) patterns obtained by protein electrophoresis on polyacryl- amide gels for A. sylvaticus (B), A. flavicollis (A, C) and A. alpicola (A). Symbols: AF = A. flavicollis, AS = A. sylva- ticus, AA = A. alpicola. Biochemical identification of three sympatric Apodemus species 87 Discussion The determination of the three morphologi- cally similar species A. sylvaticus, A. flavi- collis, and A. alpicola remains sometimes difficult. A. sylvaticus and A. flavicollis, are easily distinguishable in northern Europe by morphological characteristics and by the ecological parameters of their habitats. A. flavicollis is larger, with a complete col- lar of yellow-reddish or wide spot on the breast, and inhabits forest. A. sylvaticus is smaller and an eurytopic species with an elongated pectoral spot never forming a collar or without any spot at all. These two species converge morphologically in south- ern Europe, due to clinal variation in body size and pelage colour following opposite trends (Engel et al. 1973). Hence, convergence and overlapping in ex- ternal characters do not always allow a cor- rect specific assignment of specimens, espe- cially in areas where the two sibling species are distributed sympatrically, and when ju- venile individuals are concerned. The recog- nition of A.alpicola further complicated this determination problem in certain re- gions. The alpine mouse resembles the wood mouse in pelage colour while in body size it resembles the yellow-necked mouse. The present results clearly show that the three species A. sylvaticus, A. flavicollis, and A. alpicola can be 100% distinguished biochemically by their albumin and general protein 1 (GP 1) patterns. The electrophore- tic patterns of the albumin and the GP 1 of A. sylvaticus and A. flavicollis in the present study correspond to those of DEBROT and MERMoD (1977), who also analysed animals from Switzerland using the same technique. DARVICHE et al. (1979) reported two specific albumin alleles for A. sylvaticus and A. flavi- collis from France, Corsica, Spain, and Italy, and suggested that these differences are good criteria for differentiation between the two species. Moreover, it has been shown that A. alpicola has an intermediate position between A. sylvaticus and A. flavicollis with regard to allozyme allele frequencies (VoGEL et al. 1991; Fıuiıppuccı 1992). For the albumin locus, the species A. alpicola and A. flavicollis share the same allele. In all analysed individuals of the present study (including the test animals from Graubünden) no heterozygotes were found. These findings support the hypotheses that there is no gene flow between these taxa. The fact that the two alleles do not show a very pronounced difference in gel migration may lead to problems in the case of a monospecific sample of A. sylvaticus and A. alpicola or a mixture between these spe- cies. Therefore, we recommend to load re- ference samples of the two more common species A. sylvaticus and A. flavicollis on every gel. For ecological studies the electrophoretic analysis of blood samples offers not only the advantage that the animal need not be sacrificed as well as the identification of ju- venile individuals, what was impossible with previously described methods. Moreover, the blood samples can be taken from living, freshly killed (even after several hours), as well as frozen animals and can be stored for months at —-20°C. The application of the technique is simple (duration of the whole laboratory procedure about 5 hours). All these advantages allow an application for ecological and long-term studies in the field. Acknowledgements We would like to thank NELLY Dı MARco, who performed most of the electrophoreses. We are also very grateful to JÜRG PAUL MÜLLER from the Bündner Natur-Museum Chur (Switzerland), and to ETIENNE BERTOUILLE who provided us with a lot of wood mice and to TiZIANO MADDALENA and MANUEL RUEDI, who gave us a lot of precious technical advices. Last, but not least, we thank KAREN PARKER who corrected the English manu- script. 88 BRIGITTE A. REUTTER et al. Zusammenfassung Biochemische Bestimmung dreier sympatrisch vorkommender Apodemus-Arten mittels Elektrophorese von Blutproteinen Mittels Elektrophorese von Blutproteinen wurde das Allelbandenmuster von Albumin und eines „General Protein 1” der drei in den Alpen sympatrisch vorkommenden Waldmausarten Apodemus syl- vaticus, A. flavicollis und A. alpicola untersucht. Die Methode erwies sich als ein zuverlässiges Werk- zeug zur Unterscheidung der drei Arten. Ihre Bestimmung anhand äußerer morphologischer Merk- male ist nicht immer einfach, vor allem, wenn es sich um juvenile Tiere handelt. Mit dieser Methode können lebende Individuen aller Altersklassen bestimmt werden, was eine Anwendung in ökologisch ausgerichteten Felduntersuchungen erlaubt. References BENMEHDI, F.; BRITTON-DAVIDIAN, J.; THALER,L. (1980): Premier Apport de la Genetique Bio- chimique des Populations a la Systematique des Mulots de France Continentale et de Corse. Biochem. Genet. Ecol. 8, 309-315. BRITTON-DAVIDIAN, J.; VAHDATI, M.; BENMEHDI, F.; GROoS, P.; NANCE, V.; CROSET, H.; GUERASSI- MOV, S.; TRIANTAPHYLLIDIS, C. (1991): Genetic differentiation in four species of Apodemus from Southern Europe: A. sylvaticus, A. flavi- collis, A. agrarius and A. mystacinus (Muri- dae, Rodentia). Z. Säugetierkunde 56, 25-33. BRÜNNER, H.; NEET,C.R. (1991): A parapatric scenery: the distribution and ecology of Sorex araneus and S. coronatus (Insectivora, Sorici- dae) in southern Germany. Z. Säugetierkunde 56, 1-9. CsAIKL, F.; ENGEL, W.; SCHMIDTKE, J. (1980): On the biochemical systematics of three Apode- mus species. Comp. Biochem. Physiol. 65, 411-414. DARVICHE, D.; BENMEHDI, F; BRrITToNn-DAVI- DIAN, J.; THALER, L. (1979): Donn&es pre&limi- naires sur la syst&matique biochimique des genres Mus et Apodemus en Iran. Mammalia 43, 427-430. DEBROT, S.; MERMOD, C. (1977): Chemiotaxono- mie du genre Apodemus Kaup, 1827 (Roden- tia, Muridae). Rev. Suisse Zool. 84, 521-526. ENGEL, W.; VOoGEL,W.; VoicuLescu,Il.; Ro- PERS,H.H.; ZENZES,M.T.; DBENDER,K. (1973): Cytogenetic and biochemical differ- ences between Apodemus sylvaticus and Apo- demus flavicollis, possibly responsible for the failure to inbreed. Comp. Biochem. Physiol. 44, 1165-1173. FERNANDES, C.; ENGELS, H.; ABADE, A.; COUTIN- Ho, M. (1991): Zur genetischen und morpho- logischen Variabilität der Gattung Apodemus (Muridae) im Westen der Iberischen Halbin- sel. Bonn. zool. Beitr. 42, 261-269. FıLippuccı, M. G.; Sımson,S.; NEVvo,E. (1989): Evolutionary biology of the genus Apodemus Kaup, 1829 in Israel. Allozymic and biometric analyses with description of a new species: Apodemus hermonensis (Rodentia, Muridae). Boll. Zool. 56, 361-376. FıLippuccı, M. G. (1992): Allozyme variation and divergence among European, Middle Eastern, and North African species of the genus Apo- demus (Rodentia, Muridae). Isr. J. Zool. 38, 193-218. FıLippuccı, M. G.; STORCH, G.; MACHOLAN, M. (1996) Taxonomy of the genus Sy/vaemus in western Anatolia — morphological and elec- trophoretic evidence (Mammalia: Rodentia: Muridae). Senck. Biol. 75, 1-14. FRAGUEDAKIS-TSOLIS, S. E.; CHONDROPOU- Los, B. P,; LykAkIs, J. J., ONDRIAS, J. C. (1983): Taxonomic problems of woodmice, Apode- mus ssp, of Greece approaged by electro- phoresic and immunological methods. Mam- malia 47, 333-337. GEBCZYNSKI, M.; NIELSEN, J. T.;, SOMONSEN, V. (1986): An electrophoretic comparison be- tween three sympatric species of rodents from Jutland, Denmark. Hereditas 104, 55-59. GEMMEKE, H. (1980): Proteinvariation und Taxo- nomie in der Gattung Apodemus (Mammalia, Rodentia). Z. Säugetierkunde 45, 348-365. GEMMEKE, H. (1981): Albuminunterschiede bei Wald- und Gelbhalsmäusen (Apodemus syl- vaticus und A. flavicollis, Mammalia, Roden- tia) auch in getrockneten Muskeln und Bäl- gen elektrophoretisch nachweisbar. Z. Säugetierkunde 46, 124-125. Biochemical identification of three sympatric Apodemus species 89 GEMMEKE, H.; NIETHAMMER, J. (1981): Die Wald- mäuse Apodemus sylvaticus und A. flavicollis vom Monte Gargano (Süditalien). Z. Säuge- tierkunde 46, 162-168. GILL, A.; PETRoVv, B.; ZIVKoVvIc, S.; RiIMSDA, D. (1987): Biochemical comparison in Yugosla- vian rodents of the families Arvicolidae and Muridae. Z. Säugetierkunde 52, 247-256. HAUSSER, J.; ZUBER,N. (1983): Determination specific d’individus vivants des deux especes jumelles Sorex araneus et S. coronatus, par deux techniques biochimiques (Insectivora, Sorieidae). Rev. suisse Zool. 90, 857-862. HEINRICH, G. (1952): Apodemus flavicollis alpico- laN. N. J. Mammalogy 33, 260. NASCEITTI, G.; FiLippuccı, M.G. (1984): Genetic variability and divergence in Italian popula- tions of Apodemus sylvaticus and Apodemus flavicollis (Rodentia, Muridae). Ric. Biol. Selv. 9, 75-83. NASCETTI, G.; Tizı, L.; BurLinı, L. (1980): Differ- enziazione biochimica e variabilita genetica in due popolazioni di Apodemus sylvaticus e A. flavicollis (Rodentia, Muridae). Acc. Naz. Lincei, Rendiconti classe Sci. F.M.N. Serie VII, Vol. 9, 131-136. NEET, C. R. (1989): Ecologie comparee et biog£o- graphie Evolutive de deux especes parapatri- ques: Sorex araneus et S. coronatus (Mamma- lıa, Insectivora, Soricidae). Diss. thesis, Universite de Lausanne. NEET, C. R.; HAUSSER, J.(1989): Chromosomal re- arrangements, speciation and reproductive isolation: the example of two karyotipic spe- cies of the genus Sorex. J. Evol. Biol. 2, 373- 378. NEET, C. R.; HAUSSER, J. (1990): Habitat selection in zones of parapatric contact between the common shrew Sorex araneus and Millet’s shrew S. coronatus. J. Anim. Ecol. 59, 235- 250. NEET, C. R.; HAussERr, J. (1991): Biochemical ana- lysis and determination of living individuals of the Alpine karyotypic races and species of the Sorex araneus group. Mem. Soc. Vaud. Sc. Nat. 19, 97-106. NIETHAMMER, J. (1978): Apodemus flavicollis — Gelbhalsmaus, Apodemus sylvaticus — Wald- maus. In: Handbuch der Säugetiere Europas. Vol. 1. Ed. by J. NIETHAMMER and F. KrAPP. Akad. Verlagsges. Wiesbaden. Pp. 325-358. REUTTER, B. A.; HAUSSER, J.; VoGEL,P. (1999): Discriminant analysis of skull morphometric characters in Apodemus sylvaticus, A. flavi- collis, and A. alpicola (Mammalia; Rodentia) from the Alps. Acta Theriol. 44, 299-308. STORCH, G.; LÜTT, ©. (1989): Artstatus der Alpen- waldmaus, Apodemus alpicola Heinrich, 1952. Z. Säugetierkunde 54, 337-346. TUuRNI, H.; SCHÖNHERR, R. (1994): Neue Nach- weise der Schabrackenspitzmaus (Sorex coro- natus) in Baden-Württemberg durch Poly- acrylamidgel-Elektrophorese. Z. Säugetier- kunde 59, 321-325. VoGEL, P; MADDALENA, T.; MABILLE, A.; PA- QUET, G. (1991): Confirmation biochimique du statut specifique du mulot alpestre Apodemus alpicola Heinrich, 1952 (Mammalıa, Roden- tia). Bull. Soc. Vaud. Sc. Nat. 80.4, 471-481. Authors’ addresses: BRIGITTE A. REUTTER and PETER VoGEL, Univer- site de Lausanne, Bätiment de Biologie, Institut d’Ecologie (IE-ZEA), CH-1015 Lausanne-Do- rigny, Switzerland, e-mail: (brigitte.reutter@ie-zea.unil.ch); HARALD BRÜNNER, Wiesenstrasse 6, D-76228 Karls- ruhe, Germany. Mamm. biol. 66 (2001) 90-101 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Diversity of mammals in the Bladen Nature Reserve, Belize, and factors affecting their trapping success By T. M. Caro, RAcHEL Brock, and MARCELLA KELLY Department of Wildlife, Fish and Conservation Biology, University of California, Davis, California, USA Receipt of Ms. 12. 11. 1999 Acceptance of Ms. 20. 04. 2000 Abstract The presence of 33 non-volant mammal species was recorded in the Bladen Nature Reserve, an area of subtropical wet forest in south central Belize, as determined from walking transects and using Sherman and Tomahawk traps to capture mammals. Using trapgrids over 6 075 trapnights, the ef- fects of trap design, bait, moon phase, logging, elevation, and proximity to a river on three mea- sures of trapping success were examined systematically. Open wire mesh traps yielded somewhat higher trapping success than Sherman traps; oats and molasses produced higher trapping success than other kinds of bait; and trapping success was higher in selectively logged than in unlogged forest, and marginally higher at lower elevations and close to a river. Moon phase had no effect on trapping success. These results provide baseline data on mammal diversity in a relatively un- exploited area of central America and, though preliminary, indicate which aspects of trapping tech- nique need to be standardized when comparing species diversity and abundance across neotropical sites. Key words: Mammals, trappability, diversity, Bladen, Belize Introduction Most ecological studies of neotropical 1997; MALcoLM 1997), loss of top predators mammals require data on species diversity and abundance. For example, questions about population dynamics (e.g., O’Con- NELL 1989), population demography (e.g., TORRES-CONTRERA et al. 1997), community structure (e. g., AsQuItH et al. 1997), and re- gions of mammal abundance (e.g., MARES 1992) all require information on the num- ber of mammal species or their relative abundance in an area. Conservation studies sımilarly require data on diversity and abundance in order to understand the ef- fects of habitat fragmentation (LYNAM 1616-5047/01/66/02-090 $ 15.00/0. (WRIGHT et al. 1994; TERBORGH et al. 1997), and mammal exploitation (Dırzo and Mir- ANDA 1991; GLANZ 1991; WRIGHT et al. 2000) on communities of mammals. In the neotropics data on mammal communities are growing (Voss and Emmons 1996) but they still come from only a handful of loca- tions, the most notable of which are La Sel- va in Costa Rica (Timm 1994 a), Los Tuxlas, Mexico (ESTRADA et al. 1994), Barro Color- ado Island, Panama (GLAnz 1990), Cosha Cashu, Peru (Janson and EmMmons 1990) and near Manaus, Brazil (MALcoLm 1990) Diversity of mammals in the Bladen Nature Reserve 91 which limits our ability to make generaliza- tions about how mammalian communities are organized. As studies of mammals in the neotropics in- crease, and comparisons between areas be- come more feasible, researchers must stan- dardize their techniques for estimating mammal abundance, or at least be aware of the biases inherent in different methods of mammal estimation (Wırson et al. 1996). We therefore investigated six differ- ent factors that influence trapping success outside the neotropics and might therefore be of importance elsewhere. Two of these were methodological factors (type of small mammal trap and characteristics of the bait) and four were ecological factors (phase of the moon, selective logging, ele- vation, and proximity to rivers). Trap design can influence trapping success substantially in temperate regions (SEALAN- DER and JAMES 1958) and this problem has been identified in at least one neotropical site (WoopMmAn et al. 1996). We therefore compared measures of trapping success using hand-made wıre mesh small mammal traps which could be seen through with standard Sherman traps which obscure visi- bility. It is also well known that baits can af- fect trapping success in temperate zones (e.8., BUCHALCZYK and OLSZEwsKI 1971). Although far less work has been carried out on this problem in the neotropics, preli- minary evidence suggests that bait is not an important factor in mammal trapping (Woopman et al. 1996). Increased moon- light lowers trapping success in open habi- tats in temperate regions (BROWN et al. 1988) because small mammals often show changes in activity in accordance with a re- duction of predation risk from visually hunting aerial and terrestrial predators (CLARKE 1983). Although moonlight is known to influence activity patterns of some tropical mammals (FENToN et al. 1977, Emmons 1987; ALKoN and SALTZ 1988), it has received little systematic atten- tion in small neotropical mammal trapping studies. Selective logging has many influ- ences on mammal community structure in tropical environments with arboreal mam- mal densities being lower (EISENBERG et al. 1979) and small mammal abundance being higher in logged sites (KASENENE 1984; IsA- BIYRE-BASUTA and KASENENE 1987, MAL- coLM 1995). Increased elevation has been found to change species diversity in some areas of the tropics (RıckART et al. 1991) while proximity to seasonally flooded areas close to rivers may alter species diversity and abundance in complex ways (JANSON and Emmons 1990). Therefore the aim of this study is to investigate the diversity of species at a new site in the neotropics and to analyse factors affecting trapping success in this environment. Material and methods The study was conducted in and adjacent to the Bladen Nature Reserve in the Maya Mountains, Toledo District, Belize (Fig. 1). The reserve en- compasses 350 km? of the watershed of the Bla- den River between latitudes 16°36'18” and 16°24 34’N and longitudes 88°42’16” and 89°0451”W. Elevation ranges from 50 to 1000 m. Rainfall averages around 3000 mm per annum or more; January to April are the driest months with rain starting from June onwards through November. Mean monthly temperatures range from 16° to 33°C in Belize. Much of the re- serve is composed of Coban formation limestone and volcanic rock. The reserve contains subtropi- cal wet forest (HARTSHORN et al. 1984) with the lowest parts, along the main flow of the river where we worked, supporting alluvial soils and tall broad-leaved forest. Most of the Bladen Nat- ure Reserve is unlogged; selective logging has been practiced immediately outside its eastern border, however. The reserve is subject to un- known hunting pressure by local people for game meat. The study was conducted during four periods of fieldwork: June and July 1994, June through August 1995, March 1997 and July 1998. Some rain fell in each of these periods but trapping was suspended under extremely wet conditions because we were concerned about hypothermia of captured individuals. All the work was con- ducted adjacent to the eastern entrance of the re- serve, 1-3 km inside, and 0.5 km outside it. Mammal diversity was assessed in three ways: through night and daytime walks in which obser- vations and calls of mammals were noted, by in- 92 T. M. Caro et al. Belize Belize City ® Belmopan Dongrigae Fig. 1. Belize and the location of Bladen Nature Reserve (in black) in the country. Top right inset shows location of Belize in central America. Bottom right inset shows the boundary of Bladen Nature Reserve (dashed line) and the Bladen Branch River and its tributaries (solid lines). The entrance to the reserve is at Forest Hill shown as a black dot. Three 2 km transects were located (i) just inside the reserve boundary running south west to north east terminating at Forest Hill; (ii) between Forest Hill and the Richardson Creek (fifth tributary upstream from the entrance) confluence south and parallel to the Bladen River; (iii) and outside the reserve from and due north east of Forest Hill. All trapgrids within the reserve were set south of the Bladen River between Forest Hill and the sec- ond tributary upstream from the entrance. Trapgrids outside the reserve were set within 500 m of the reserve en- trance. Black bar denotes 10 km; arrow shows north (adapted from HARTSHoRN et al. 1994). Diversity of mammals in the Bladen Nature Reserve 93 terpreting mammal tracks in the mud, and through live trapping. Abundances of certain mammal species were estimated through repeat trapping using standard Sherman, custom-made small mammal traps, and three sizes of Toma- hawk traps. We used the field guide of EMMONS (1990) to identify mammals in the field. Two pre- vious mammal surveys had been conducted in the Bladen (Brokaw and Lroyp-Evans 1987; THE NATURE CoNSERVANCY 1993). We set up three 2 km transects that were walked at a pace of 1 km/hour. One was located within the reserve and followed an abandoned logging track that serves as the main road into the re- serve; the second was further inside the reserve on the south side and parallel to the Bladen Riv- er; the third was located adjacent to but outside the Reserve in a selectively logged area. Night walks (N=18) were conducted between 18.30 and 21.00 h and animals were spotted using flash- lights; day walks (N =28) took place between 05.30 and 07.30 h. Tracks of mammals were exam- ined opportunistically and in specially prepared flats of mud, at the center at which was placed a commercial carnivore olfactory lure. Small mammals were trapped using standard Sherman traps (23x8x8cm), or traps of the same dimensions custom-made entirely out of galvanized wire mesh (by the late R. SCHWwAB) save for a galvanized aluminum plate door that was part of the floor until the door swung up- wards on closure (see also EmMmons 1984). Traps were usually set in an 8x8 or 7x7 grid with traps 10 m apart; infrequently they were set along a transect precluding calculation of density. Grids of small mammal traps were baited either with a mixture of oats and molasses, with peanut butter, with green or ripe bananas, or with a medley of fruit consisting of apple, banana, and coconut mixed together. Ripe and green bananas were lumped together in bait analyses because green bananas quickly ripened in traps making it diffi- cult to distinguish between the two. Traps were usually set on the flat valley floor of the reserve approximately 20-200 m from the Bladen River bank. In another protocol, trapgrids were set at higher elevations on the steep limestone slope bordering the floodplain approximately 50 and 100 m above the valley floor and 200-250 m from the river. We set two types of Tomahawk traps (40x13x13cm and 40x17x17cm) which we termed middle-sized traps. These were usually set in a 6x8 grid with traps 50 m apart. 'Iraps were baited either with oats and molasses, with green or ripe bananas, raisins, or a fruit found on the forest floor, Warre Cohune palm, Astrocar- ynm mexicanum. These traps were set approxi- mately 50-450 m from the river. In one set of tests these traps were placed at a height of 1-2.5 m in trees and baited with oats and molasses in order to estimate squirrel abundance. We also set large Tomahawk traps (65x 22.5x22.5cm) in a 6x8 grid with traps set 50 m apart. These were baited with either green or ripe bananas, Astrocarynm mexicanum fruit, commercial cat food or fresh fish; these latter two baits were combined in bait analyses since they both contained animal protein and smelled similar. In none of the trapping protocols were different baits run at the same time on either the same or different grids. Traps were usually set for 5 consecutive nights although a small minority was set for fewer or more nights (range 1-7). Traps were set either in unlogged forest with a tall thick canopy and rela- tively open understory inside Bladen Nature Re- serve, or in the area east of the reserve where se- lective logging allowed light to penetrate, producing a thicker understory of vegetation. Traps were opened and baited between 16.00 and 17.00 h and checked next morning between 06.00 and 09.00 h. Captured animals were individually marked by cutting small patches of fur since we were only interested in recaptures over a maxi- mum of 7 days. Quarter of the moon was noted during each sequence of trapping; in some ana- lyses traps set during the first and last quarter spanning the new moon, and then the second and third quarters spanning the full moon were combined. We recorded number of species caught, percen- tage capture success, individual mammals caught/ 100 trapnights, and densities when traps were set in a grid square. Percentage capture success was the number of captures divided by the number of trapnights (i.e., number of traps multiplied by the number of nights on which they were set); in- dividual mammals caught per 100 trapnights was the number of different individuals captured di- vided by the number of trapnights x 100; and den- sities were calculated by dividing the number of individuals captured by the area covered by the grid expressed as number of individuals/km’. We took this area to be the dimensions of the grid plus 5 m either side (i. e., 70 mx 70 m or 4900 m? in a 7x7 grid) because paucity of captures made it difficult to calculate maximum distance be- tween captures of known individuals. We did not use mark-recapture techniques to estimate den- sity because recapture rates were so low. 94 T. M. Caro et al. Data were analysed by comparing trapgrids although the number of trapnights that these re- presented is also presented for clarity. Non-para- metric statistics were used as number of species was an ordinal measure, and captures/trapnight and individual mammals caught/100 trapnights produced too many zeroes (no captures) to justify normalizing the data required for parametric sta- tistics. The use of non-parametric statistics made it difficult to control for confounding variables; in- stead we conducted a series of carefully controlled comparisons among grids by excluding variables that were found to be important in previous ana- lyses even though these resulted in a reduction in sample sizes. & was set at 0.05; nevertheless p va- lues lying between 0.1 and 0.05 are noted and dis- cussed with appropriate caution. Results Captures In this study, twenty eight species of non- volant mammals were identified inside and outside but within 0.5km of the Bladen Nature Reserve, although two of these were equivocal identifications (Tab. 1). In our study all of these species except five, Phi- lander opossum, Urocyon cinereoargenteus, Conepatus semistriatus, Leopardus sp., and Panthera onca were found inside the re- serve; in a previous study conducted by the Rapid Ecological Assessment Team in 1993 a jaguar and a small felid had been identi- fied inside Bladen (THE NATURE CoNSER- vancy 1993). Our results, combined with those of the two earlier surveys (Tab. 1), show that the Bladen area holds a minimum of 33 non-volant species including large predators such as Felis concolor and Panthera onca. Employing small mammal traps, we cap- tured five non-volant species, Heteromys desmarestianus, Ototylomys phyllotis, Tyl- omys nudicaudus, Marmosa robinsoni, Or- yzomys couesi and an unknown species of bat with an average percentage capture suc- cess of 6.5% (sd+5.9), or 5.6 individuals/ 100 trapnights (sd+5.6) (nm = 26 grids; 4236 trapnights). These traps yielded re- spective densities of 6836/km”, 270/km?, 183/km“, 925/km? and 2 127/km? for the five non-volant species (n = 18 grids; 3521 trap- nights). With the middle-sized Tomahawk traps, we caught only two species, Ototyl- omys phyllotis and Tylomys nudicaudas. Trap success was low at 0.7% (sd +1.0), or 0.6 individuals/100 trapnights (sd # 0.8) (n=9grids; 1354 trapnights). Density of these two species was 2/km’ and 10/km’, respectively (n = 5 grids; 1200 trap- nights). With the large Tomahawk traps, we caught Didelphis marsupialis, Didelphis virginianus, Dasypus novemcintus and a Tylomys nudicaudus giving an average per- centage trap success of 4.3% (sd + 4.3), or 3.9 individuals/100 trapnights (sd #3.9) (n=9grids; 1218 trapnights). Density of these species was 28/km?, 8/km’, 2/km“, and 2/km’, respectively (n=5 grids; 1152 trapnights). Excluding bats, we calcu- lated a Shannon-Wiener index of 2.021. Factors affecting trapping success Type of trap: Compared to standard Sher- man traps, custom-made wire mesh traps of the same dimensions caught marginally more terrestrial mammal species (n=6; 20 grids, respectively; 618, 3618 trapnights, Means (Xs)=1.0 (sd+0.9), 2.1 (sd+1.4) species, Mann-Whitney U test, z=-1.763, P = 0.078), demonstrated marginally higher percentage capture success (Xs =2.2% (sd # 3.0), 7.12% (sd #6.) zespeetiyech; z=-1.951, P= 0.051), and caught a margin- ally greater number of individuals per trap- nieht (Xs=2 (sd+3), 7 (sd+6) indivi- duals/100 trapnights respectively, z=- 830, P=Z00 7) Type of bait: For small mammal traps, there were significant differences in the number of species caught (n = 26 grids; 4236 trap- nights, Kruskal-Wallis test, H= 11.444, P=0.01), percentage trap success (H= 8.464, P = 0.037), and individuals captured/ 100 trapnights (H= 7.888, P= 0.048) de- pending upon the type of bait offered. On each measure, oats and molasses were most successful followed by green and ripe bana- nas combined, and then the fruit medley (Tab. 2). There were no significant differ- ences between baits on measures of density, Diversity of mammals in the Bladen Nature Reserve 95 Table 1. List of species of mammals in and immediately adjacent to Bladen Nature Reserve. 1994-1998 this study: Tp: trapped; O: observed; Tr: tracks; H: heard; I/O: inside or outside Bladen Nature Re- serve. 1. refers to species noted by the 1993 Rapid Ecological Assessment Team (THE NATURE ConservacY 1993); 2. refers to species noted by the 1987 Manomet survey (Brokaw and LioYo-Evans 1987). * species may have been Marmosa mexicana; + species may have been Leopardus wiedii (Margay). Scientific name Marsupilia Didelphis marsupialis Didelphis virginianus 2 Philander opossum Chironectes minimus 1 Micoureus alstoni 1 Marmosa robinsoni* Xenarthra Tamandua mexicana Dasypus novemcinctus Chiroptera Balaniopteryx io 1 Noctilio leporinus 5 Primates Alouatta pigra 1 Ateles geoffroyi 1, 2 Carnivora Urocyon cinereoargenteus Nasua narıca 1 Potos flavus Mustela frenata 1 Eira barbara 2 Conepatus semistriatus Lontra longicaudus 2 Leopardus pardalıs + 1 Puma concolor 2 Panthera onca 1 Perissodactyla Tapirus bairdii 1, 2 Artiodactyla Tayassu tajacu 1 Tayassu pecari 1, 2 Mazama americana 1, 2 Odocoileus virginianus 1, 2 Rodentia Sciurus yucatanensis Sciurus deppei 1, 2 Heteromys desmarestianus Oryzomys couesi Tylomys nudicaudus 1 Ototylomys phyllotis Agouti paca 1,2 Dasyprocta punctata 1, 2 Common name Common opossum Virginia opossum Common gray four-eyed opossum Water opossum Alston’s wolly mouse opossum Robinson’s mouse opossum Northern tamandua Nine-banded long-nosed armadillo Least sac-winged bat Greater fishing bat Unidentified species Mexican black howler monkey Central American spider monkey Gray fox White-nosed coati Kinkajou Long-tailed weasel Tayra Striped hog-nosed skunk Neotropical otter Ocelot Puma Jaguar Baird’s tapir Collared peccary White-lipped peccary Red brocked deer White-tailed deer Yucatan squirrel Deppe’s squirrel Forest spiny pocket mouse Coues’ rice rat Naked-tailed climbing rat Big-eared climbing rat Paca Central American agouti 96 T. M. Caro et al. Table 2. Mean (X) and standard deviation (sd) measures of trapping success in all small mammal traps separated by type of bait; round brackets refer to numbers of trapgrids or traplines, square brackets to the number of trapnights. Oats and molasses Number of species Percentage trap success Individuals/100 trapnights Individuals/km° Peanut butter Bananas Fruit medley * traps were set in a line, rather than grid, so estimates of density are unavailable. however. When analyses were restricted only to small mammal wire mesh traps, the number of species captured still differed significantly by type of bait (n = 20 grids; 3,618xtrapniehis HZ 7.279277 0.02) Alt hough this was no longer the case for mea- sures of percentage trap success and indivi- duals/100 trapnights. There were no effects of bait for any measure in the medium-sized or large traps. Moon phase: Considering either small, medium-sized or large traps, there was no effect of moon phase on number of species captured, percentage capture success, num- ber of individuals caught/100 trapnights, or density either when quarters were analyzed separately or when quarters respectively spanning the new and full moons were com- bined. Selective logging: Somewhat more indivi- dual mammals were captured in logged for- est than in unlogged forest using small mammal traps (n = 11,15 grids, respectively; 956, 3280 trapnights, n=8 (sd+7), 4 (sd+4) per 100 trapnights, Mann-Whitney U test, z = 1,664, P = 0.096) but there were no significant differences on the three other measures. Restricting analyses to the cus- tom-made traps that were somewhat more effective in catching small mammals, we found that percentage capture success and individuals/trapnight were significantly greater ın logged forest outside Bladen than in unlogged forest inside (n = 7,13 grids, re- spectively; 741, 2877 trapnights, Xs= 12.2%:; (sd #&5.4), 3.3% (sd 33h) 32 229! P = 0.024; Xs = 12 (sd +4), 4/100 trapnights (sd.#6),. z= 2.539, P =. 0.011). Inzthegiema- hawk traps, a slightly greater number of species was captured in the unlogged than in the logged forest in the medium-sized (n= 7,2 grids, respectively; 1312, 42 trap- niehts, Xs = 0.7 (sd + 0.5), 0 (sd +0) species, z= 1.690, P= 0.091) and in the large traps (n= 7,2 grids, respectively; 1200, 18 trap- nights, Xs = 2.0 (sd + 1.0), 0.5 (sd + 0.7) spe- cies, z- 1.8697 0.092): Elevation: We obtained a marginally greater number of species, percentage capture suc- cess and number of individuals/100 trap- nights on the valley floor than at higher ele- vations on the slope above the Bladen River (n = 24,2 grids, respectively; 3876, 360 trap- nights, Xs = 2.0 (sd+ 1.3), 0.5 (sd +0.7) spe- ciess, Mann-Whitney Utest, z= 1.642, P=0.1;...,Xs=0.7% : . ((d=0 003% (sd’#:0:.4),4. 12=1.832, - PE0. 067x851 (sd’#5.0)5 0.3 (sd#0.4) ZZ RI 0.067). Two of these three results still held after analyses were restricted to custom- Diversity of mammals in the Bladen Nature Reserve 97 made small mammal traps placed in un- logged areas only (n=111low and 2 higher elevation grids; 3258, 360 trapnights, Rs 24 (sd #1.6), 0:5 (sd+#0.7) species re- speetively, 71610, B:0.107: XS 6.3% (sd+5.0), 0.3% (sd+0.4) respectively, 7 287 8108101075 2XS 4.71 (sd#3:5),.0:31 100 trapnights (sd + 0.4) respectively, z= 1.781, P = 0.075). Tomahawk traps were not placed above the valley floor. Proximity to the river: Finally, we compared traps Set within <50 m of the river bank with those placed further away (50-200 m). Results showed that the number of species captured was marginally higher close to the rıver than further away from it (n = 11,11 grids, respectively; PERS), 1349 trapnights, Xs=24 (sd+1.6), 1.2 (sd+0.9) species, Mann-Whitney U test, 24895922 10:073)!but there was'no' effect on percentage trap success or individuals/ 100 trapnights. When analyses were re- stricted to custom-made traps set on the valley floor in unlogged areas, however, proximity to the river resulted in greater numbers of mammals caught on all three measures (n=8,3 grids next to and away from the river, respectively; 2103, 414 trap- nights, Xs = 2.9 (sd + 1.5), 1.0 (sd + 1.0) spe- eiese az 1163, 1 BE=0:0785 XS =7.9% (sd+4.6), 19% (sd +2.8) trap success re- speeiivehawze 20415 BP = 0.0415 Xs 5.7 (sd + 3.2), 1.9 individuals/100 trapnights (sd + 2.8) respectively, z = 1.837, P = 0.066). Discussion Captures Bladen Nature Reserve and land immedi- ately adjacent to it held a minimum of 33 species of non-volant mammals which is comparable to other central American sites. RABINOWITZ and NOTTINGHAM (1989) docu- mented 39 species of non-volant mammals in the Cockscomb basin which is almost ad- jacent to the Bladen; MEDELLIN (1994) re- ported 48species in Selva Lacondona, Chiapas, Mexico but these were compiled over 10 years as opposed to our 6 month to- tal period; Timm (1994b) documented 50 species for La Selva in Costa Rica over 20 years; and GLAanz (1990) reported 39 species on Barro Colorado Island, Pana- ma, which had been studied for 13 years at the time. Small mammal trapping success in Bladen (6.5%) was comparable to that in other neotropical wet forests such as Cockscomb. For example, trapping success in the Gi- gante Peninsula, Panama was 4.2% for the wet and 7.3% for the dry season (MCCLEARN et al. 1994). In contrast to mea- sures of mammal diversity, it is difficult to make many direct comparisons of species’ densities with other sites as data for many of the same species are unavailable. Didel- phis and Dasypus densities appeared low compared to Barro Colorado Island and south American sites (GLAnz 1990), whereas Marmosa densities were higher than either at Barro Colorado or even Gua- topo, Venezuela (EISENBERG et al. 1979). Oryzyomys densities were extremely high in comparison to Barro Colorado, Guatopo, Cosha Cashu and Cabassou, French Guiana (CHARLES-DOMINIQUE et al. 1981; GLANZ 1990) possibly because some traps were set in selectively logged habitats. Factors affecting trapping success There is a substantial literature on the ef- fects of trap type on trapping success in tem- perate regions (e.g., SEALANDER and JAMES 1958; SLADE et al. 1993). The few studies that have been conducted in the neotropics have compared live traps to snap traps and found the latter to catch more species and indivi- duals (Pızzımentı 1979; WooDMAn et al. 1996). The only studies to compare wire mesh and Sherman traps were conducted in temperate climates (HOoLDENREID 1954; O’FARRELL et al. 1994). In both cases a great- er proportion of captures was made using wire mesh traps and heteromyid rodents in particular were captured in mesh traps. O’FARRELL et al. (1994) found that custom- made wire mesh traps captured two to three times more individuals than Sherman traps. Our results from the neotropics replicated 98 T. M. Caro et al. these findings in that they showed margin- ally greater percentage success and indivi- duals caught/100 trapnights in mesh traps. Clearly, comparisons of small mammal den- sities in the neotropics must take into ac- count of whether traps are of mesh or box design. There is also a considerable literature on the effects of baits on trap success but again mostly from temperate regions (e.g., BEER 1964; SLADE et al. 1993; but see LAURANCE 1992) and it is well known that rolled oats and peanut butter is a very effective bait in capturing terrestrial mammals in thıs part of the world. Although very few compari- sons have been reported for the neotropics, WooDMman et al. (1996) found no differences in captures using suet or peanut butter in tropical forest in south-eastern Peru. In contrast, we found that oats and molasses caught more species, generated greater trap success, and captured more individuals/ 100 trapnights than other baits, with peanut butter producing poorest results. In addi- tion, there are many other baits that include anımal protein that we did not use. It there- fore appears premature to suggest that bait has little influence on trapping in the neo- tropics. In contrast to many studies in deserts of North America (e.g., PrıcE et al. 1984), phase of the moon had no effect on mea- sures of trapping in this study. Possibly, the thick canopy obscured the moon to such an extent that little light penetrated to the for- est floor. It is well documented that small mammal abundance is greater in selectively logged habitats in temperate regions (e. g., MONTH- Ey and SOUTHIERE 1995) as well as in the tropics (DELANnY 1971; STRUHSAKER 1997). For example, MALcoLM (1995) showed that terrestrial mammal abundance, richness and diversity were all greater in pasture and young secondary forest than in continu- ous forest north of Manaus. Our results are consistent with his findings in that percen- tage success and number of individuals cap- tured was greater in areas that had been logged outside the reserve than inside it. In addition, we caught a greater variety of spe- cies, principally marsupials, outside the re- serve. There may be many reasons for these associations including a more abundant and predictable insect prey base (MALCOLM 1995), increased seed abundance stemming from increased vertical vegetation density (MoNADJEM 1997), or even reduced threat of predation (DA FonEscA and ROBINSON 1990) but these were not investigated. Studies that have looked into the effects of elevation on small mammals have often found different results. For example, abun- dance increased with elevation in a tropical rainforest in the Philippines but species richness did not change (HEANEY et al. 1989). In contrast, in a temperate rainforest in Chile, number of species, number of in- dividuals, and species diversity all declined with increased elevation (PATTERSoN et al. 1989). We found a marginally reduced number of species, trapping success, and number of individuals at higher elevations although trapping effort was relatively low off the valley floor. In addition, we caught a somewhat greater number of species, had somewhat greater success, and cap- tured somewhat more individuals per trap- night near the river than farther from it. Taking these two results together, there ap- peared to be a gradient of decreasing small mammal diversity and abundance as one progressed away from the Bladen River and up the slope. Whether these findings reflect differences in humidity, soil drai- nage or type, or habitat structure remains unresolved. Our findings are necessarily preliminary be- cause we chose to examine a large number of factors which reduced our sample size. Nevertheless, they highlight the importance of carefully selecting the type of trap and type of bait in trapping studies of neotropi- cal mammals. They also point to the differ- ences that may be expected in estimating mammal abundance and diversity in areas with different logging regimes, elevations and proximity to rivers in neotropical habi- tats. As such, they reinforce the necessity of standardizing techniques when compar- ing species abundance and diversity across neotropical sites. Diversity of mammals in the Bladen Nature Reserve 99 Acknowledgements We thank The Forestry Department and Belize Audubon Society for permission to work in the Bladen Nature Reserve and RAPHAEL MANZA- NERO and JOHN PINELLO for important assistance. We thank Jake and Kerry MARLIN for permission to work in the Belize Foundation for Research and Environmental Education (BFREE). This work was supported by two University of Califor- nia faculty research grants and a President’s un- dergraduate fellowship as well as funding from San Francisco State University. Work was con- ducted under UC Davis Animal Care and Use Protocols. We thank DEIRDRE DoHERTY, DAvE - DROUILLARD, CHRIS GREGORY, BETH JOHNSTONE, ToM MCNAMARA, JEREMY MULL, MEGAN TEMPLE- TON, and SUSAN Woorsy for logistical help in the field; Ron CoL£, JAKE and Kerry MARLIN, and DirK VAN VUREN for loaning and storing equip- ment; STEVEN BREWER, DouG KELT, MARCEL REJ- MANEk, and DIRK VAN VUREN for discussions; and SABINE SCHMID-HoLMESs for translating the sum- mary into German. Zusammenfassung Diversität von Säugern im Bladen Naturreservat, Belize, und Faktoren, die einen Fangerfolg beeinflussen Im Bladen Naturreservat, einem subtropischen Feuchtwaldgebiet, wurde durch Zählungen an Tran- sekten und unter Einsatz von Sherman- und Tomahawk-Lebendfallen das Vorkommen von 33 Säuge- tierarten festgestellt. In insgesamt 6.075 Fallennächten wurden Einfluß von Fallendesign, Köder, Mondphase, Holzeinschlag, Höhenlage und Nähe eines Flusses auf Fangerfolg systematisch unter- sucht. Drahtgitterfallen hatten etwas größeren Fangerfolg als Sherman-Fallen, Haferflocken und Molasse erzielten größeren Fangerfolg als andere Köder, Fangerfolg in Wald mit selektivem Holz- einschlag war größer als in Wald ohne Einschlag und er war etwas größer in höher gelegenen Gebie- ten und näher an einem Fluß. Die Mondphase hatte keinen Einfluß auf Fangerfolg. Die Resultate lie- fern Basisdaten über die Säugetiervielfalt in einem relativ unerforschten Gebiet Zentralamerikas und geben an, wenn auch nur vorläufig, welche Aspekte im Fangdesign standardisiert werden soll- ten, um einen Vergleich der Artenvielfalt zwischen verschiedenen neotropischen Studiengebieten zu ermöglichen. References ALKON, P. U.; Sartz, D. (1988): Influence of sea- son and moonlight on temporal activity pat- terns of Indian crested porcupines (Aystrix in- dica Kerr). J. Mammalogy 69, 71-81. AsauUITH, N. M.: WRIGHT, S. J.; CLauss, M.]. (1997): Does mammal community composition control recruitment in neotropical forests? Evidence from Panama. Ecology 78, 941-946. BEER, J. R. (1964): Bait preferences of some small mammals. J. Mammalogy 45, 632-634. BRoKAw, N. V. L.; LLovpD-Evans, T. L. (1987): The Bladen Branch wilderness. Manomet Bird Observatory, Mass. BROwn, J. S.; KOTLER,B.P; SMITH, R.J.; WiRTZ, II, W. ©. (1988): The effects of owl predation on the foraging behavior of heteromyid ro- dents. Oecologia 76, 408-415. BUCHALCZYK, T.; OLSZEWSKI, J. L. (1971): Behav- ioural response of forest rodents against trap and bait. Acta Theriol. 18, 277-292. CHARLES-DOMINIQUE, P.; AÄATRAMENTOWICZ, M.; CHARLES-DOMINIQUE, M.; GERARD, H.; HLA- DIK, A.; HLADIK, C. M.; PREVoST, M. F. (1981): Les mammiferes frugivores arboricoles noc- turnes d’une foret guyanaise: inter-relations plantes-animaux. Revue Ecologie (Terre et Vie) 35, 341-435. CLARKE, J. A. (1983): Moonlight’s influence on predator/prey interactions between short- eared owls (Asio flammeus) and deermice (Peromyscus maniculatus). Behav. Ecol. and Sociobiol. 13, 205-209. Derany, M.J. (1971): The biology of small ro- dents in Mayanja forest, Uganda. J. Zool. (London) 165, 85-129. Dirzo, R.; MIRANDA, A. (1991): Altered patterns of herbivory and diversity in the forest un- derstory: a case study of the possible conse- quences of contemporary defaunation. In: Plant-Animal Interactions: Evolutionary 100 T. M. Caro etaal. Ecology in Tropical and Temperate Regions. Ed. by DP.W.Prıce, T.M. LEwInsoHnN, G. W. FERNANDEZ, and W. W. BEnson. New York: John Wiley and Sons. Pp. 273-287. EISENBERG, J. F; O’ConNELL, M.; AUuGusrt, P. V. (1979): Density, productivity, and distribution of mammals in two Venezuelan habitats. In: Vertebrate Ecology in the Northern Neotro- pics. Ed. by J. F. EISENBERG. Washington, DC: Smithsonian Institution Press. Pp. 187-207. Emmons, L.H. (1984): Geographic variation in densities and diversities of non-flying mam- mals in Amazonia. Biotropica 16, 210-222. Emmons, L. H. (1987): Comparative feeding ecol- ogy of felids in a neotropical forest. Behav. Ecol. and Sociobiol. 20, 271-283. Emmons, L.H. (1990): Neotropical Rainforest Mammals: a Field Guide. Chicago: Univ. Chi- cago Press. ESTRADA, A.; COoATES-ESTRADA, R.; MERITT JR, D. (1994): Non flyıng mammals and landscape changes in the tropical rain forest region of Los Tuxlas, Mexico. Ecogeography 17, 229-241. FENToN, M. B.; BoyLeE, N. G. H.; HARRISON, T. M.; OxLEy, D. J. (1977): Activity patterns, habitat use, and prey selection by some African insec- tivorous bats. Biotropica 9, 73-85. FoNESCcA, G. A. B. DA; RoBINSON, J. G. (1990): For- est size and structure: competitive and preda- tory effects on small mammal communities. Biol. Conserv. 53, 265-294. FLEMING, T. H. (1973): The number of rodent spe- cies in two Costa Rican forests. J. Mammalogy 54, 518-521. GLANZ, W. E. (1990): Neotropical mammal densi- ties: how unusual is the community on Barro Colorado Island, Panama? In: Four Neotropi- cal Rainforests. Ed. by A. W. GEnTRY. New Haven, Connecticut: Yale Univ. Press. Pp. 287-313. GLANZ, W. E. (1991): Mammalian densities at pro- tected versus hunted sites in central Panama. In: Neotropical Wildlife Use and Conservation. Ed. by J.G. Rosınson and K.H. REDFORD. Chicago: Univ. Chicago Press. Pp. 163-173. HARTSHORN, G. S.; NICOLAIT, L.; HARTSHORN, L.; BEVIER, G.; BRIGHTMAN,R.; CAL,J; CaA- WITCH, A.; DAVIDSoN, W.; Dusois, R.: DyER, C.; GIBSON, J.;, HAwLEy, W.; LEo- NARD, J.; NICOLAIT, R.; WEYER, D.; WHITE, H.; WHITE, C. (1984): Belize: Country Environ- ment Profile. A Field Study. Belize City, Belize: Robert Nicolait and Associates. HEANEY, L. R.; HEIDEMAN, P. D.; RicKART, E. A.; UTZURRUM,R. B.; KLoMPEN, J.S.H. (1989): Elevational zonation of mammals in the cen- tral Philippines. J. Trop. Ecol. 5, 259-280. HOLDENREID, R. (1954): A new live-catch rodent trap and comparison with two other traps. ]. Mammalogy 35, 267-268. ISABIRYE-BASUTA, G.; KASENENE, J.M. (1987): Small rodent populations in selectively felled and mature tracts of Kibale forest, Uganda. Biotropica 19, 260-266. JANSoN, C. H.; Emmons, L.H. (1990): Ecological structure of the nonflying mammal community at Cosha Cashu Biological Station, Manu Na- tional Park, Peru. In: Four Neotropical Rain- forests. Ed. by A. W. GEntky. New Haven, Connecticut: Yale Univ. Press. Pp. 314-338. KASENENE, J. M. (1984): The influence of selective logging on rodent populations and the regen- eration of selected tree species in the Kibale forest, Uganda. Trop. Ecol. 25, 179-195. LAURANCE, W. F. (1992): Abundance estimates of small mammals in Australian tropical rainfor- est: a comparison of four trapping methods. Wildlife Res. 19, 651-655. Lynam, A.)J. (1997): Rapid decline of small mam- mal diversity in monsoon evergreen forest fragments in Thailand. In: Tropical Forest Remnants: Ecology, Management, and Conser- vation of Fragmented Communities. Ed. by W. M.LAuRANcE and R. ©. BIERREGAARD, Jr. Chicago: Univ. Chicago Press. Pp. 222-240. MALCOLM, J. R. (1990): Estimation of mammalian densities in continuous forest north of Man- aus. In: Four Neotropical Rainforests. Ed. by A. W. GENTRY. New Haven, Connecticut: Yale Univ. Press. Pp. 339-357. MALCOLM, J. R. (1995): Forest structure and the abundance and diversity of neotropical small mammals. In: Forest Canopies. Ed. by M.D.LowMmAan and N.M. NADKARNI San Diego: Academic Press. Pp. 179-197. MALCOLM, J. R. (1997): Biomass and diversity of small mammals in Amazonian forest frag- ments. In: Tropical Forest Remnants: Ecology, Management, and Conservation of Fragmen- ted Communities. Ed. by W. M. LAURANCE and R.O. BIERREGAARD, Jr. Chicago: Univ. Chicago Press. Pp. 207-221. MARES, M. A. (1992): Neotropical mammals and the myth of Amazonian biodiversity. Science 255, 976-979. MCCLEARN, D.; KOHLER, J.; McGowan, K.]J.; CE- DENO, E.; CARBONE, L. G.; MILLER, D. (1994): Arboreal and terrestrial mammal trapping on Gigante peninsula, Barro Colorado Nature Monument, Panama. Biotropica 26, 208-213. MEDELLIN, R. A. (1994): Mammal diversity and conservation in the Selva Lacondona, Chia- pas, Mexico. Conserv. Biol. 8, 780-799. Diversity of mammals in the Bladen Nature Reserve 101 MOoNADIJEM, A. (1997): Habitat preferences and biomasses of small mammals in Swaziland. Afr. J. Ecol. 35, 64-72. MOoNnTHEY, R. W.; SOUTHIERE,E.C. (1995): Re- sponses of small mammals to forest harvesting in northern Maine. Canad. Field Nat. 99, 13-18. O’ConNELL, M. A. (1989): Population dynamics of neotropical small mammals in seasonal ha- bitats. J. Mammalogy 70, 532-548. O’FARRELL, M. J.; CLARK, W. A.; EMMERSON, F. H.; JUAREZ,S.M.; Kay, FR.; O’FARRELL, T. M.; GOoODLETT, T. Y. (1994): Use of a mesh live trap for small mammals: are results from Sherman live traps deceptive? J. Mammalogy 75, 692-699. PATTERSON, B. D.; MESERVE,P.L.; LAnG,B.K. (1989): Distribution and abundance of small mammals along an elevational transect in temperate rainforests of Chile. J. Mammalogy 70, 67-78. PIZZIMENTI, J. J. (1979): The relative effectiveness of three types of traps for small mammals in some Peruvian rodent communities. Acta Theriol. 24, 351-361. PRICE, M. V.; WASER, N. M.; Bass, T. A. (1984): Ef- fects of moonlight on microhabitat use by de- sert rodents. J. Mammalogy 65, 353-356. RABINOWITZ, A.; NOTTINGHAM, Jr, B. G. (1989): Mammal species richness and relative abun- dance of small mammals in a subtropical wet forest of Central America. Mammalia 53, 217-226. RicKART, E. A.; HEANEY, L. R.; UTZURRUM, R. C. B. (1991): Distribution and ecology of small mammals along an elevational transect in southeastern Luzon, Phillipines. J. Mammal- ogy 72, 458-469. SEALANDER, J. A.; JAMES, D. (1958): Relative effi- ciency of different small mammal traps. ]. Mammalogy 39, 215-233. SLADE, N. A.; EIFLER, M. A.; GRUENHAGEN, N. M.; DEAvELos, A.L. (1993): Differential effec- tiveness of standard and long Sherman live- traps in capturing small mammals. J. Mam- malogy 74, 156-161. STRUHSAKER, T. T. (1997): Ecology of an African rainforest: logging in Kibale and the conflict between conservation and exploitation. Gai- nesville: Univ. Florida Press. TERBORGH, J.; LOPEZ, L.; TELLO, J.;, Yu, D.; BRU- NI, A.R. (1997): Transitory states in relaxing landbridge islands. In: Tropical Forest Rem- nants: Ecology, Management and Conserva- tion of Fragmented Communities. Ed. by W. F. LAURANCE and R.O. BIERREGAARD, Jr. Chicago: Univ. Chicago Press. Pp. 256-274. THE NATURE ConSERVANCY (1993): Rapid ecologi- cal assessment of the Bladen Nature Reserve, October 1993. Tımm, R.M. (1994 a): The mammal fauna. In: La Selva: Ecology and Natural History of a Neo- tropical Rainforest Ed. by L. A. McDADe, K. S. Bawa, H. A. HESPENHEIDE, and G. S. HARTSHORN. Chicago: Univ. Chicago Press. Pp. 229-237. Tımm, R.M. (1994b): Mammals. In: La Selva: Ecology and Natural History of a Neotropical Rainforest. Ed. by L. A. McDADe, K. S. BAawa, H. A. HEsSPENHEIDE, and G.S. HARTSHORN. Chicago: Univ. Chicago Press. Pp. 394-398. TORRES-CONTRERAS, H.; SILVA-ARANGUIZ, E.; MAR- QUET, P. A.; Camus, P. A.; JaKsıc, F. M. (1997): Spatiotemporal variability of rodent subpopu- lations at a semiarid neotropical locality. J. Mammalogy 78, 505-513. Voss, R. S.; Emmons, L. H. (1996): Mammalian di- versity in neotropical lowland rainforests: a preliminary assessment. Bull. Amer. Mus. Nat. Hist. 230, 1-115. WiLson, D.E.; CoLE,F.R.; NIcHoLs, J.D.; Ru- DRAN, R.; FOSTER, M.S. (1996): Measuring and Monitoring Biological Diversity: Stan- dard Methods for Mammals. Washington DC: Smithsonian Institution Press. WOooDMAN,N.; TIMM, R.M.; SLADE,N. A.; Doo- NAN, T. J. (1996): Comparison of traps and baits for censusing small mammals in neotro- pical lowlands. J. Mammalogy 77, 274-281. WRIGHT, S. J., GOMPPER, M. E.; DELEOoNn, B. (1994): Are large predators keystone species in neo- tropical forests? The evidence from Barro Colorado island. Oikos 71, 279-294. WRIGHT, S. J.; ZEBALLOS, H.; DOMINGUEZ, ].; GAL- LARDO, M. M.; MOoRENO, M.C.; IBANEZ,R. (2000): Poachers alter mammal abundance, seed dispersal, and seed predation in a neo- tropical forest. Conserv. Biol. 14, 227-239. Authors’ address: T.M. CAro, RACHEL BRoOcK, and MARCELLA Key, Department of Wildlife, Fish and Conser- vation Biology, 1 Shields Avenue, University of California, Davis, CA 95616, USA (e-mail: tmcaro@ucdavis.edu) Mamm. biol. 66 (2001) 102-110 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Original investigation Mammalian Biology S} NZ Zeitschrift für Säugetierkunde Welche quantitativen Beziehungen bestehen bei Säugetieren zwischen Schädelkapazität und Hirnvolumen? Von M. RÖHRS und P. EBINGER Institut für Zoologie, Tierärztliche Hochschule Hannover, Hannover Receipt of Ms. 20. 03. 2000 Acceptance of Ms. 10. 09. 2000 Abstract How is cranial capacity related to brain volume in mammals? The measurement of cranial capacity is often used to obtain information about the brain volume in mammals. Brain volume and cranial capacity are of the same size only in small mammals. For 17 mammalian species (451 individuals) we could show that cranial capacity often is larger than brain volume, and the differences in size are not equal in the 17 species. The relation cranial capa- city % brain volume ranges from 102.45% to 116.97%. Therefore, assuming that cranial capacity = brain volume can lead to remarkable errors. The intraspecific allometric relation of cranial capacity to brain volume may be isometric or positive allometric, the interspecific relation is positive allometric. Thus, if the allometric relations are known, it is possible to estimate unknown brain volumes on the basis of known cranial capacity. Key words: Allometry, cranial capacity, brain size, mouse-to-elephant-line Einleitung Ein intra- bzw. interspezifischer Vergleich von Hirngrößen ist ein erster und einfacher Schritt die funktionelle Bedeutung des Ge- hirns bei verschiedenen Arten zu schätzen. Das beste Maß für die Hirngröße ist das Frischhirngewicht (HG) bzw. das Frisch- hirnvolumen (HV = HG/1,036). Die Hirn- größe hängt ab von verschiedenen Fakto- ren: Körpergröße, Evolutionshöhe und Spezialisation. Bevor dieser Einfluß auf die Hirngröße verschiedener Arten bewertet werden kann, muß die Bedeutung der Kör- pergröße für dieses Organ festgestellt wer- 1616-5047/01/66/02-102 $ 15.00/0. den. Ein geeignetes Maß für die Körpergrö- ße ist das Bruttokörpergewicht (BKG). Die Ermittlung des Einflusses des BKG auf das HG ist möglich mit der Allometrieformel: log Organgewicht = log b + a : log Körper- gewicht. Der Wert für a sagt aus, ob der re- lative Anteil der Hirngröße am Körperge- wicht von kleinen zu großen Tieren abnimmt, gleich bleibt oder zunimmt, der Wert b enthält Faktoren, welche zudem das HG beeinflussen. Die Bestimmung von a für den interspezifi- schen Bereich ist möglich bei unterschied- Schädelkapazität und Hirnvolumen bei Säugetieren lich großen Arten naher Verwandtschaft, die sich außer der Körpergröße in anderen Merkmalen nur gering unterscheiden. Ist a bekannt, dann können Unterschiede in b als quantitativer Ausdruck von Verschie- denheiten in der Cephalisationshöhe (Evo- lutionshöhe, Spezialisation) angesehen wer- den (STEPHAN et al. 1986; RöHrs 1986; RÖHRS et al. 1989). Die Beschaffung von Daten über Hirn- und Körpergewicht von möglichst vielen Säuge- tierarten ist schwierig, daher sind zur Er- weiterung der Datenbasis andere Verfahren als die direkte Messung von HG und BKG versucht worden. In den zoologischen und paläontologischen Museen lagern viele Schädel von fossilen und rezenten Säuge- tierarten. Als Ersatzmaß für die tatsächli- che Hirngröße sind bei diesem Material häufig Messungen der Hirnschädelkapazität (HSK) möglich; aber stimmen HSK und HV wirklich überein? Diese Frage soll hier untersucht werden. Auf das Problem der Schätzung des BKG nach Schädel- und Skelettmaßen wird in dieser Arbeit nicht eingegangen. Messungen der HSK zur Gewinnung von Informationen über die tatsächliche Hirn- größe sind eine alte Methode (DArwıINn 1868; KLAtt 1912), und bis heute ist dieses Verfahren noch üblich. Jerıson (1973) be- stimmte die HSK bei fossilen und rezenten Säugetierarten, um Änderungen der Hirn- größe in der Geschichte der Säugetiere zu erforschen. JERISON (1973) setzt dabei vor- aus, daß HSK und HV übereinstimmen; auch Rapınsky (1978), GITTLEMAN (1986), TowE und Mann (199) und MARTIN (1990) nehmen für die von ihnen untersuch- ten Säugetierarten Übereinstimmung von HSK und HV an. JErRIsoN (1973) weist aber darauf hin, daß bei Walen die HSK be- trächtlich größer sein kann als das HV. MARTIN (1990) bemerkt, daß es angebracht sei, die Beziehungen HSK-HV zu erfor- schen, bevor Schlüsse aus der HSK auf die tatsächliche Hirngröße gezogen werden können. Eingehende Untersuchungen über die Be- ziehung von HSK zu HV fehlen bis heute. Nach Daten von RöHrs und EBINGER 103 (1998) beträgt die Hirnschädelkapazität bei 14 Zoo-Przewalskipferden durchschnittlich 116% des Hirnvolumens, wobei die mittlere HSK ein Volumen von 649 cm’ und das HV einen Mittelwert von 558 cm? haben; die al- lometrische Beziehung HSK-HV ist hierbei isometrisch. Bei Nasua rufa (n = 12) beträgt die mittlere HSK 105% des mittleren Hirn- volumens; die intraspezifisch allometrische Beziehung HSK-HV ist positiv (a= 1,19). Der prozentuale Anteil der HSK am HV nimmt von 100,5 beim kleinsten HV bis auf 109,5 beim größten HV zu (RöHRs et al. 1989). Diese Beispiele zeigen, daß HSK und HV beträchtliche Größenunterschiede aufwei- sen können, und daß die Beziehung HSK - HV im intraspezifischen Bereich auch posi- tiv allometrisch sein kann. Wird die HSK ohne Berücksichtigung solcher Tatsachen mit dem HV gleich gesetzt, kann es zu Fehl- beurteilungen der Hirngröße kommen. Wir haben daher an einem umfangreichen Da- tenmaterial untersucht, welche Beziehun- gen im intraspezifischen und interspezifi- schen Bereich zwischen HSK und HV bestehen können, und ob es Möglichkeiten gibt, das HV (damit auch das HG) nach der HSK zu schätzen. Material und Methode Bei 451 Individuen von 17 Säugetierarten erfolgte die Bestimmung des Frischhirngewichts durch Wiegen nachdem das Schädeldach aufgesägt, das Rückenmark durchtrennt und das Gehirn ent- nommen wurde. Durch Division mit dem Fak- tor 1,036 — dem spezifischen Gewicht für Hirn- masse — wurde das Hirnvolumen aus dem HG er- rechnet. Nach der Mazeration des Schädels und Schließen des Schädeldachs konnte die Hirnschä- delkapazität durch Auffüllen des Hirnschädel- hohlraums mit Schrot oder Glaskugeln gemessen werden. Bei allen Individuen wurde auch das Bruttokörpergewicht festgestellt. So kennen wir bei 451 Tieren jeweils HSK, HV, HG und BKG. Das Material stammt von Forschungsreisen (HER- RE und RöHrs, Südamerika 1956/57, 1962; Gala- pagos 1971) sowie aus Sammlungen von HERRE, RÖHRS und EBINGER. Für 12 Arten konnten die intraspezifischen Allo- metriegeraden (Ellipsenhauptachsen) HSK-HV 104 M. RÖHRS u.a. berechnet werden. Weiterhin haben wir für insge- samt 17 Arten jeweils den mittleren prozentualen Anteil der Hirnschädelkapazität am Hirnvolumen bestimmt (HSK%HV). Bei Arten mit positiver Allometrie für die Beziehung HSK zu HV wur- den entsprechend den Allometriegleichungen die Werte HSK%HV für das kleinste und größte HV errechnet. Darüber hinaus haben wir die in- terspezifischen Allometriegeraden HSK-HV für Myrmecophagidae, Mustelinae, Canidae sowie für alle untersuchten 17 Arten ermittelt. Ergebnisse und Diskussion Intraspezifische Allometrien HSK-HV Ist nur die Hirnschädelkapazität bekannt, so können intraspezifisch allometrische Be- ziehungen HSK zu HV geeignet sein, das HV von Individuen einer Art zu schätzen. Dies kann wichtig sein für Vergleiche der Hirngrößen innerhalb von Arten, so z.B. zwischen Unterarten, Geschlechtern sowie Wild- und Haustieren (KruskA 1980; RÖHRS und EBINGER 1983, 1998). In Tabelle 1 sind die intraspezifischen Allo- metriegleichungen HSK-HV für 12 Säuge- tierarten aufgeführt. Bei Mustela nivalis, Mustela, erminea, und Martes foina ist diese Beziehung etwa isometrisch (Abb. 1). Eine solche Isometrie haben wir bei Zoo-Prze- walskipferden bereits 1998 nachgewiesen (RÖHRs und EBINGER 1998). Bei den übri- gen 8 von den 12 Arten ist die Beziehung HSK-HV positiv allometrisch, und die a- Werte reichen von 1,05 bei Dusicyon gym- nocercus bis zu 1,25 bei Hausziegen (Abb. 2). Es gibt also keinen einheitlichen intraspezifischen a-Wert für die Beziehung HSK-HV bei allen Säugetierarten. Bei den von uns untersuchten Spezies ist Tabelle 1. Intraspezifische allometrische Beziehungen Hirnschädelkapazität zu Hirnvolumen (log HSK=log b+a - log HV) bei 12 Säugetierarten. Spezies Oryctolagus cuniculus 1,0191 (n=35, r= 0,9727) Mustela nivalis (n = 78, r= 0,9916) Mustela erminea (n = 29, r= 0,9812) Martes foina (n= 19, r= 0,9763) Dusicyon gymnocercus (n = 56, r= 0,9864) Dusicyon culpaeus (n= 15, r= 0,9840) Canis aureus (n = 32, r= 0,9160) Canis latrans (n = 32, r= 0,9149) Canis lupus (n=65, r= 0,7791) Lama guanacoe (n = 20, r= 0,9663) Hausziegen (n=36, r= 0,9484) Zooprzewalskipferde (n = 14, r= 0,9550) 0,3623 0,7320 1,3512 1,6229 1,7581 1,8423 1,9709 2,1730 2,4427 2,1298 2,8126 -0,0713 1,0812 1,0085 -0,0224 0,3432 0,9916 -0,0230 0,9873 0,7182 0,0720 0,9913 1,3154 -0,0627 1,0505 1,6046 -0,1672 1,1079 1,7377, -0,0578 1,0567 1,7982 -0,1268 1,0833 1,9365 -0,2860 1,1584 2,1226 -0,0735 1,0449 2,4081 -0,4423 1,2456 2,0650 0,0656 1,0000 2,7469 Schädelkapazität und Hirnvolumen bei Säugetieren 105 log (cm? en) :(0 S 3,0 S = 2,5 ©® : = 2,0 Ö (42) = 1,5 <[ 1,0 2,0 3,0 log (cm?) Hirnvolumen Abb. 1. Intraspezifische allometrische Beziehung HSK-HV bei Mustela nivalis, agua = 0,9916 (Standardfehler: x = 0,0093, y = 0,0093). ° (Q — $) 3 ih S\| & | + 132 — 100 — Hirnschädelkapazität 77 113 166 log (cm?) Hirnvolumen Abb. 2. Intraspezifische allometrische Beziehung HSK-HV bei Hausziegen, ara = 1,2456 (Standardfehler: x = 0,0066, y = 0,0082). 106 M. RÖHRS u.a. Tabelle 2. Mittelwerte von HSK (cm*), HV (cm°) und HSK%HV bei 17 Säugetierarten Spezies Oryctolagus cuniculus 10,45 10,20 Mustela nivalıs 2,3056 Mustela erminea 5,3951 Martes foina 22,45 20,67 Fira barbara 50,35 46,39 Dusicyon sechure 33,46 32,63 Dusicyon gymnocercus 41,97 40,23 Cerdocyon thous 50,65 46,57 Dusicyon culpaeus 57,29 54,66 Canis aureus 69,55 62,83 Canis latrans 93,92 86,40 Canis lupus 148,94 132,62 Lama guanacoe 277,14 255,92 Hausziege 134,83 116,14 Tamandua tetradactyla 28,32 26,23 Myrmecophaga trıdactyla 96,58 82,53 649,46 558,39 Zooprzewalskipferd die mittlere HSK immer größer als das HV. Der Größenunterschied ist aber bei den einzelnen Arten nicht gleich. Der Wert HSK%HV reicht von 102% bei Oryctola- gus cuniculus bis zu 117% bei Myrmeco- phaga tridactyla (Tab. 2). Bei Isometrie der intraspezifischen Beziehung HSK-HV ist der Wert HSK%HV auf allen Größenstufen des Gehirns etwa gleich; so bei Mustela ni- valis, Mustela erminea, Martes foina und Zoo-Przewalskipferden (Tab. 2). Bei positi- ver Allometrie steigt der Wert HSK%HV von den kleinen zu den großen Gehirnen an; bei Hausziegen von 110% bis 125% an. Bei diesem Maximum beträgt das HV 156 cm? und die HSK 195 cm’; das ist eine Differenz von 39 cm’. Weitere Beispiele sind in Tabelle 2 aufgeführt. Eine Übereinstimmung von HSK und HV kann besonders bei kleinen Säugetieren vorkommen. Mann et al. (1988) nehmen bei Muridae und Cricetidae die Gleichheit von HSK und HV an. Eine solche Konfor- mität ist aber für Säugetiere nicht allgemein gültig und unsere Ergebnisse machen für den intraspezifischen Bereich deutlich, daß eine Gleichsetzung von HSK und HV zu er- heblichen Überschätzungen der Hirngröße führen kann. 2,2039 5,2264 HSKY%HV Min - Max 102,45 104,62 103,23 108,61 108,35 102,54 = 104,33 103,03-105,57 108,76 - 104,80 103,31-106,69 110,70 109,88-111,71 108,24 107,32-109,76 23 108,53-115,34 103,29 107,70-109,07 116,09 110,32-124,81 107,95 = 116,97 z 116,31 146,314 6,31 101,15-104,00 Isometrie Isometrie Isometrie Interspezifische Beziehungen HSK-HV Mittelwerte der Frischhirnvolumina (oder des Hirngewichts) und der Bruttokörperge- wichte von Arten sind die Basis für inter- spezifische Vergleiche von Hirngrößen. Interspezifische Allometrien HSK-HV könnten die Möglichkeit bieten, Mittelwer- te der Hirnvolumina von Arten nach der Hirnschädelkapazıtät abzuschätzen. Dazu folgen drei Beispiele. Für Tamandua tetradactyla und Myrmeco- phaga tridactyla lautet die interspezifische Allometriegerade HSK-HV: 1,7100 = 0,0778 + 1.0763 16677. 2009819 Er vier Arten der Mustelinae: 1,0367 = 0,0097 + 1,0163 - 1,0108 (r = 0,9999) und für sieben Canidenarten (Abb. 3): 1,7995 = -0,0724 + 1,0583 - 1,7687. € = 0.9993). Nach der interspezifischen Allometrieglei- chung HSK-HV bei Mustelinae wurde für die vier untersuchten Arten das jeweils nach der Hirnschädelkapazität zu erwarten- de Hirnvolumen (,„HV“) errechnet. In Ta- belle 3 sind die Mittelwerte HSK%HV den Mittelwerten „HV“%HV gegenüberge- stellt. Das gleiche Verfahren wurde für sie- ben Canidenarten durchgeführt (Tab. 4). In allen Fällen weicht der Wert „HV“%HV Schädelkapazität und Hirnvolumen bei Säugetieren log (cm?) 170 Be © Cerdocyon thous > Hirnschädelkapazität & © 30 Canis latrans > Canis aureus > < Dusicyon sechure 107 Canis lupus > < Dusicyon culpaeus < Dusicyon gymnocercus 70 170 log (cm?) Hirnvolumen Abb 3. Interspezifische allometrische Beziehung HSK-HV bei 7 Canidenarten, agın = 1,0583 (Standardfehler: x = 0,0783, y = 0,0825). Tabelle 3. Vergleich HSK%HV mit „HV"% HV. „HV” er- rechnet nach der interspezifischen Allometriegeraden HSK-HV bei Mustelinae. HSK%HV „HV"%HV Spezies 100,98 98,29 101,08 99,71 Mustela nivalıs Mustela erminea Martes foina Eira barbara 104,62 103,23 108,61 108,35 Tabelle 4. Vergleich HSK%HV mit „HV"%HV. „HV” er- rechnet nach der interspezifischen Allometriegeraden HSK-HV bei Canidae. HSK%HV Spezies „HV"%HV 102,54 104,33 108,76 104,80 110,70 108,24 za 98,93 99,00 102,36 98,17 103,32 93,68 99,79 Dusicyon sechure Dusicyon gymnocercus Cerdocyon thous Dusicyon culpaeus Canis aureus Canis latrans Canis lupus geringer von 100% ab als der Wert HSK%HV. Je größer das Hirnvolumen, um so mehr Unterschied besteht zwischen Hirnvolumen und Hirnschädelkapazität; so ist beim Wolf die mittlere HSK gegenüber dem HV um 16,3 cm? oder 112% größer, beim erheblich kleineren Dusicyon sechure beträgt die Differenz aber nur 0,83 cm’ oder 103% (Tab. 2, 4). Wir sind der Ansicht, daß auch für weitere Arten der Mustelinae und Canidae mit un- bekanntem Hirnvolumen die genannten in- terspezifischen Gleichungen HSK-HV sgül- tig sind. Damit ist es möglich (bei ausrei- chendem Schädelmaterial), das mittlere HV nach der mittleren HSK abzuschätzen. Für umfangreiche Studien müßten aller- dings zunächst bei weiteren Familien die in- terspezifischen allometrischen Beziehungen HSK zu HV untersucht werden. Es ist aber schwierig, ausreichendes Datenmaterial zu erlangen. Vielleicht ist es machbar, eine Al- 108 M. RÖHRS u.a. lometriegleichung HSK-HV für alle Säuge- tiere einzusetzen, die zwar nur näherungs- weise Schätzungen des HV nach der HSK erlaubt, die aber bessere Werte liefert als die Gleichsetzung von HSK und HV. Bei Säugetieren sind viele Untersuchungen über allometrische Beziehungen zwischen Organgrößen (auch Stoffwechselgrößen) und Körpergrößen durchgeführt worden. Dabei stößt man immer wieder auf das Be- mühen, jeweils einen a-Wert zu finden, der für alle Säugetiere (mouse to elephant line) Gültigkeit hat. Solche Fälle sind aber selten, was sicher mit besonderen Anpassungen und Spezialisationen zu tun hat. Evolutions- höhe und Spezialisationen haben aber wohl kaum eine Auswirkung auf das Verhältnis HSK-HV. Möglich ist jedoch — wie schon gezeigt — der Einfluß der Hirngröße. MAR- TIın (1990) hat für 33 Primatenarten die Be- ziehung HSK-HG berechnet und nennt als Ergebnis: HSK = 0,94: HG'"”. Die Daten- paare stammen allerdings nicht von den sel- ° le! m OD 3 & 91 © oO Canis latrans > Dusicyon SHlDDeRz > Tayra barbara > Dusicyon sechure Martes foina > O1 © Hirnschädelkapazität N oO < Mustela nivalis 1 15 Capra aegagrus > ben Individuen. MARTIN (1990) nennt die Beziehung isometrisch und sagt weiter, daß man bei den untersuchten Primaten HSK und HG gleichsetzen könne, und daß dies auch bei anderen Säugern mit gleichem Größenbereich möglich sei. Ist diese An- nahme richtig? Nehmen wir an, bei einem HV von 1 cm? ist die HSK ebenfalls 1 cm’ und rechnen dann mit dem a-Wert von 1,02 die HSK bei 10cm’ HV aus, so beträgt diese 10,5 cm’. Bei 100 cm’ HV ergibt sich eine HSK von 110 cm’, und bei 1000 cm? HV beläuft sie sich auf 1158 cm”. Ein Pottwal mit 6080 cm’ HV würde nach dieser Gleichung eine HSK von 7237 cm’ haben, damit wären wir bei dem von JErIsoNn (1973) erwähnten großen Unterschied zwischen HSK und HV bei Wa- len. Es ist also nicht angebracht, für die Be- ziehung HSK-HV auch im interspezifischen Bereich für alle Säugetierarten Isometrie an- zunehmen, und es ist hier ebensowenig an- gebracht HSK und HV gleichzusetzen. < Equus przewalskii Lama guanacoe > < Canis lupus < MUSEEN tridactyla < Canis aure < Cerdo < Dusicyon gymnocercus < Tamandua tetradactyla ocyon t Dus < Oryctolagus cuniculus < Mustela erminea 225 650 log (cm?) Hirnvolumen Abb 4. Interspezifische allometrische Beziehung HSK-HV bei 17 Säugetierarten, arena = 1,0222 (Standardfehler: x = 0,1426, y= 0,1458), Schädelkapazität und Hirnvolumen bei Säugetieren Wir haben nun für 17 Säugetierarten (Tab. 2, Abb. 4) aus verschiedenen systema- tischen Einheiten die interspezifische Allo- metriegerade berechnet, sie lautet: log HSK = -0,0015 + 1,02 - log HV (r = 0,0997). Die Beziehung ist positiv allometrisch und stimmt fast mit der von MARTIN (1990) er- mittelten überein. Der Wert HSK%HV nimmt von kleinen bis zu großen Gehirnen zu, und zwar von 102% bis zu 116% (Tab. 2, 5). Das heißt, besonders bei größe- ren Gehirnen kann die Differenz zwischen HSK und HV ein beträchtliches Ausmaß er- reichen. Diese höhere Differenz könnte auch durch Zunahme der Volumina der Subarachnoidalräume und der Dicke der Dura mater bedingt sein. Quantitative Da- ten über die Größen dieser Strukturen lie- gen nicht vor. In Tabelle5S sind die Mittelwerte HSK%HV für die 17 untersuchten Arten aufgeführt und mit denen nach der o.a. in- terspezifischen Allometriegleichung ermit- telten Werte „AV“%HV verglichen. Es ist eindeutig, daß die Schätzung des HV nach der Allometriegleichung HSK-HV weit bessere Werte liefert als die Gleichsetzung von HSK und HV. Wir schlagen daher vor, die von uns errechnete Allometriegerade Zusammenfassung 109 Tabelle 5. Vergleich HSK%HV mit „HV“%HV. „HV” be- rechnet nach der interspezifischen Allometriegeraden HSK-HV für 17 Säugetierarten HSKY%HV Spezies „HV"%HV Oryctolagus cuniculus Mustela nivalis Mustela erminea Martes foina Eira barbara Dusicyon sechure Dusicyon gymnocercus Cerdocyon thous Dusicyon culpaeus Canis aureus Canis latrans Canis lupus Lama guanacoe Hausziege Tamandua tetradactyla Myrmecophaga trndactyla Zooprzewalskipferd 102,45 104,62 103,23 108,61 108,35 102,54 104,33 108,76 104,80 110,70 108,24 I 108,29 116,09 107,95 116,97 116,31 INN 103,02 99,39 101,92 100,09 95,90 96,59 100,28 96,39 101,38 93,50 101,18 96,27 104,82 100,78 106,42 101,52 HSK-HV von 17 Säugetierarten zu verwen- den, wenn für Säuger das Hirnvolumen nach der Hirnschädelkapazität geschätzt werden soll. Wünschenswert sind weitere Datenpaare von HSK und HV, um die Zu- verlässigkeit der Geraden zu verbessern. Zur Ermittlung der Größe des Hirnvolumens von Säugetieren wird häufig die Messung der Hirnschä- delkapazität eingesetzt. Hirnvolumen und Hirnschädelkapazität stimmen aber nur bei kleinen Säu- getieren überein. Für 17 Säugetierarten (451 Individuen) konnten wir nachweisen, daß die Hirnschädelkapazität größer ist als das Hirnvolumen. Die Größenunterschiede sind bei den einzel- nen Arten nicht gleich. Der Mittelwert Hirnschädelkapazität % Hirnvolumen reicht von 102,45% bis 116,97%. Die Bewertung der Größe der Hirnschädelkapazität als Maß für das Hirnvolumen kann zu beträchtlichen Fehleinschätzungen der Hirngröße führen. Die intraspezifische allometrische Beziehung Hirnschädelkapazität - Hirnvolumen kann isometrisch oder positiv allometrisch sein, die interspezifische ist positiv allometrisch. Sind solche Allometrien bekannt, dann kann man für Individuen von Arten mit unbekanntem Hirnvolumen, aber bekannter Hirnschädelka- pazität die Hirngröße abschätzen. Entsprechendes gilt für Arten im interspezifischen Bereich. Literatur Darwin, CH. (1868): Das Variieren der Tiere und Pflanzen im Zustande der Domestikation. Übersetzung von V.Carus. Stuttgart 1906: Schweizerbart’sche Verlagsbuchhandlung. GITTLEMAN, J. L. (1986): Carnivore brain size, be- havioral ecology and phylogeny. J. Mammal- ogy 67, 23-36. 110 M. RÖHRS u.a. JERISON, H. J. (1973): Evolution of the Brain and Intelligence. New York: Acad. Press. KLATT,B. (1912): Über die Veränderung der Schädelkapazität in der Domestikation. Sit- zungsberichte der Gesellschaft naturforschen- der Freunde, Berlin 3, 153-179. KRUSKA,D. (1980): _Domestikationsbedingte Hirngrößenänderungen bei Säugetieren. Z. zool. Syst. Evolut.-forsch. 18, 161-195. MANN, M.D.; GLICKMAN,S.E.; TOWE, A.L. (1988): Brain/body relations among myomorph rodents. Brain Behav. Evol. 31, 111-124. MARTIN, R.D. (1990): Evolution of the Primate Central Nervous System. Weinheim: Chap- man and Hall. RADINSKY, L. (1978): Evolution of brain size in carnivores and ungulates. Amer. Natur. 112, 815-831. RÖHRS, M. (1986): Cephalisation, Telencephalisa- tion und Neocorticalisation bei Mustelidae. Z. zool. Syst. Evolut.-forsch. 24, 157-166. RÖHRS, M.; EBINGER, P. (1978): Die Beurteilung von Hirngrößenunterschieden zwischen Wild- und Haustieren. Z. zool. Syst. Evolut.-forsch. 16, 1-14. RÖHRS, M.; EBINGER, P. (1983): Noch einmal: Wölfe mit unterschiedlichen Cephalisations- stufen? Z. zool. Syst. Evolut.-forsch. 21, 314- 318. RÖHRS, M.; EBINGER, P. (1998): Sind Zooprze- walskipferde Hauspferde? Berl. Münch. Tier- ärztl. Wschr. 111, 273-280. RÖHRS, M.;, EBINGER,P.; WEIDEMANN, W. (1989): Cephalisation bei Viverridae, Hyaeni- dae, Procyconidae und Ursidae. Z. zool. Syst. Evolut.-forsch. 27, 169-180. STEPHAN, H.; BARON, G.; FRAHM, H.; STEPHAN, M. (1986): Größenvergleiche an Gehirnen und Hirnstrukturen von Säugern. Z. mikrosk.- anat. Forsch. 100, 189-212. TOWE, A.L.; MANN, M.D. (1992): Brain size/ body length relations among myomorph ro- dents. Brain Behav. Evol. 39, 17-23. Anschrift der Verfasser: Prof. Dr. MANFRED RÖHRS und Dr. PETER EBINGER (e-mail: Peter.Ebinger@tiho-hannover.de) Insti- tut für Zoologie, Tierärztliche Hochschule Han- nover, Bünteweg 17, D-30559 Hannover Mamm. biol. 66 (2001) 111-115 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication . Mammalian Biology Zeitschrift für Säugetierkunde Reproductive ecology of the endangered monogamous Malagasy giant jumping rat, Hypogeomys antimena By SIMONE SOMMER Zoologisches Institut und Museum, Universität Hamburg, Hamburg Receipt of Ms. 11. 05. 2000 Acceptance of Ms. 06. 10. 2000 Key words: Hypogeomys antimena, reproductive ecology, Madagascar Hypogeomys antimena, the Malagasy giant jumping rat, is the largest extant endemic rodent of Madagascar. Both sexes are ca. 30 cm long and weigh ca. 1.2 kg. It is consid- ered to be one of the most endangered mammalıan species of Madagascar. The dis- tribution of this rodent has greatly dimin- ished during the past two millennia. The last remaining population is restricted to patches of dry deciduous forest wıth a total extension of 20 kmx40 km situated north of the town Morondava, along the western coast of Madagascar. The whole area is sub- ject to slash and burn agriculture and com- mercial logging (GEnINnI 1996; GOODMAN and RAKOTONDRAVONY 1996). Until recently, the information on A. antimena was limited to anecdotal information (PETTER 1972; STARCK 1974) and preliminary data from a nine-week field study by Cook et al. (1991). H. antimena was reported to be strictly noc- turnal, to live in long deep burrows and to move by jumping and running. It was sug- gested that the rodent lives in social units, probably consisting of a pair plus their off- spring. The most surprising information for a rodent species was that it produces only a single offspring per year. Most rodent spe- cies are characterised by large litter sizes, 1616-5047/01/66/02-111 $ 15.00/0. short birth intervals and sexual maturation at an early age (HAsSLER 1975). In order to increase our very limited knowl- edge of the biology, ecology, and behaviour of H. antimena and for conservation pur- poses long-term field studies were initiated in 1992. It turned out that A. antimena has some very unusual lıfe characteristics for a rodent species such as an obligate monoga- mous social and mating system. Pairbonds apparently last until one mate dies. Mates defend an exclusive territory throughout the year (for more details see SOMMER 1996; 1997; 1998; 2000; SOMMER and TicHY 1999). One critical component to under- stand the population dynamics of an endan- gered species ıs its reproductive ecology (for reviews on the behaviour-conservation interface see SUTHERLAND 1998; CARO 1999). The aim of this study therefore was to ınvestigate length of the reproductive period, reproductive rate, and offspring growth of the endangered A. antimena in its natural habitat. Field studies were carried out in the 12500 ha forestry concession of the Centre de Formation Professionnelle Forestiere de Morondava (C.F.P.F.) in the Kirindy Forest (20°03’S 44°39'E) at the research station 112 SIMONE SOMMER of the German Primate Center (DPZ, Göt- tingen, Germany). A detailed description of the area is given in GANZHORN and SORG (1996). Field work took place between Oc- tober 1992 and January 1993, February and April 1994, April and June 1995, Novem- ber and December 1995 and April and June 1996. In a 100 ha study area, all exist- ing burrow systems were known and were regularly monitored and classified as active or inactive. Capture/recapture studies were carried at least once during each field peri- od. Tomahawk live traps (5S1x19x19cm, Tomahawk, Wisconsin) were set in front of the burrow holes before the nocturnal ac- tivity period of the rats started and checked at least once every hour after sunset until the animals entered the traps. Captured an- imals were anaesthetised for 10-15 min the next morning with an intramuscular injec- tion of 0.1 to 0.25 ml ketamine hydrochlor- ide (100 mg/ml), sexed, weighed, and meas- ured. 157 animals from 30 active burrows have been marked individually with a pas- sive integrated transponder (Irovan, Rö- merberg, Germany) since the beginning of the field studies in October 1992. The rats were released during their normal activity period in the evening in front of their bur- rows. The statistical tests were performed with SPSS (1997). The study indicated that the reproduction of H. antimena is seasonal and takes place during the rainy season (Dec-March). The smallest, early born offspring was observed at the beginning of December (Sth Dec) with a body mass of about 200 g and the smallest, late born offspring was observed at end of March (24th March) with a body mass of around 250 g. In contrast to the anecdotal information on the reproductive rate, the capture/recapture studies indicated that not always a male and female couple was accompanied by a single offspring. One single offspring was present in 60 cases out of 78investigated family units but in 11 cases two offspring of the present reproductive period lived together in a burrow system with their assumed par- ents. The sex ratio of offspring was ba- lanced. To answer the question whether this can be explained by the birth of twins or by two consecutive litters per reproductive period, the body mass of offspring which were born during one reproductive period in the same burrow were compared (Table 1). Only data were included in this analysis where all offspring of a pair could be weighed within two days. The mean difference in body mass of offspring trapped in the same burrow was 368+89g (n=6). In one case (Dec 1992, Tab. 1) two offspring of about the same body mass (370 g, 395 g) were trapped at the same time which were as- sumed to be born in the same litter. The data suggest that A. antimena can have two single offspring born consecutively during one reproductive period but also twins might occur in natural populations. The re- productive rate per couple was calculated from trapping results after the reproductive period and was 1.5 offspring ın 1994, 1.5 in 1995 and 1.1 in 1996. The average number of marked offspring per pair and year was 1.4. This might be an underestimation as Table 1. Body mass (g) of offspring which were born during one reproductive period in the same burrow 1. Offspring Trapping date 10.12. 92 24.03. 94 25.03. 94 27.03. 94 27.-28. 03. 94 042=022.00.,95 06. 05. 95 2. Offspring Difference Conclusion twins consecutive litters consecutive litters consecutive litters consecutive litters consecutive litters consecutive litters Reproductive ecology of Hypogeomys antimena 113 Body Mass (g) March April May June a) Month Body Mass (g) March April May June b) Month Fig. 1. Body mass of all offspring trapped between March and June. Early born offspring are symbolised by trian- gles, late born offspring by circles. a: male offspring, b: female offspring. Details on the linear regression lines are given in the text. 114 SIMONE SOMMER offspring spend the first 4-6 weeks of their life in the burrow and regularly leave it and can be trapped after another 4 weeks. The reason that to date Aypogeomys was reported to have only one single offspring per year (PETTER 1972; STARCK 1974; CooK et al. 1991), might be due to a high offspring mortality rate. Radiotracking and capture/ recapture studies revealed a mean offspring mortality of more than 50% (SOMMER 2000). In order to investigate the body mass de- velopment of early and late born offspring of consecutive litters born during one re- productive period and possible sex specific differences, the body mass of all offspring trapped between March and June were analysed (Fig. 1). The body mass develop- ment of early and late born female off- spring and late born male offspring can be described by a significant linear regression (female offspring: early born: R° = 0.46, p = 0.002, late born: R°= 0.74, p = 0.003; male offspring: early born: R° = 0.09, n.s., late born: R? = 0.53, p= 0.01). The present data do not indicate that male and female offspring differ in the development of their body mass (ANOVA: early born: Fı>s = 2.7, n.s., late born: Fi>0 = 0.75, n.s.). The difference of body mass of early and late born offspring during one reproductive per- iod decreases with increasing age in their first year of life. At the end of the dry sea- son (Nov/Dec), female offspring weighed 866+177g (n=11) and male offspring 863+99g (n=5) (t-test: n.s.). Also the analyses of other body measurements (body-, tail-, ear-, hindfoot-, head length, and head width) did not show any age- dependent differences in male and female offspring (SOMMER 1998). Although the study indicated that A. anti- mena can have more offspring per couple and year than suggested previously, the re- productive rate is still very low. The survi- val prospects of thıs endangered species is critical due to changing environmental and ecological conditions as a consequence of the increasing human impact on the re- maining habitat (SOMMER and HOoMMEN 2000). Acknowledgements The studies are supported by the “Commission Tripartite” of the Malagasy Government, the La- boratoire de Primatologie et des Vertebres de l’Universite d’Antananarivo, the Ministere pour la Production Animale et des Eaux et For£ts, the Centre de Formation Professionnelle Forestiere de Morondava, J. GANZHORN, B. RAKOTOSAMIMA- NANA, R. RASOLOARISON, L. RAZAFIMANANTSOA, WWF International, WWF Madagascar, and the German Research Foundation (So, 428/1-1). References Caro, T. (1999): The behaviour-conservation in- terface. Trends Ecol. Evol. 14, 366-369. Cook, J. M.; TREVELYAN, R.; WALLS, S. S.; HATCH- ER, M.; RAKOTONDRAPARANY, F. (1991): The ecology of Hypogeomys antimena, an ende- mic Madagascan rodent. J. Zool. (London) 224, 191-200. GANZHORN, J. U.; SorRG, J-P. (1996): Ecology and Economy of a Iropical Dry Forest in Mada- gascar. Göttingen: Primate Report 46-1. GENINI, M. (1996). Deforestation. In: Ecology and Economy of a Tropical Dry Forest in Madagascar: Ed. by J. U. GAnNZHORN and J.-P. Sorc. Göttingen: Primate Report 46-1, 49-55. GOODMAN, S. M.; RAKOTONDRAVONY, D. (1996): The Holocene distribution of Aypogeomys (Rodentia: Muridae: Nesomyinae) on Mada- gascar: In: Biog£ographie de Madagascar. Ed. by W. R. Lourenco. Paris: Ed. P’ORSTOM. Pp. 283-293. HASLER, J. F. (1975): A review of reproduction and sexual maturation in the Microtine ro- dents. Biologist 57, 52-86. PETTER, F. (1972): The rodents. In: Biogeography and Ecology of Madagascar. Ed. by R. BAr- TIsTINI and G. RICHARD-VINDARD. Den Haag: W. Junk. Pp. 661-665. SOMMER, S. (1996): Ecology and social structure of Hypogeomys antimena, an endemic rodent of the dry deciduous forest in western Madagas- car. In: Biog&ographie de Madagascar. Ed. by W.R.Lourenco. Paris: Ed. P’ORSTOM. Pp. 295-302. SOMMER, $S. (1997): Monogamy in Hypogeomys antimena, an endemic rodent of the deciduous dry forest in western Madagascar. J. Zool. (London) 241, 301-314. Reproductive ecology of Hypogeomys antimena SOMMER, S. (1998): Populationsökologie und -ge- netik von Hypogeomys antimena, einer ende- mischen Nagerart im Trockenwald Westma- dagaskars. Diss. thesis, Universität Tübingen. SOMMER, S. (2000): Sex specific predation rates on a monogamous rat (Hypogeomys antimena, Nesomyinae) by top predators in the tropical dry forest of Madagascar. Animal Behaviour 56, 1087-109. SOMMER, S.; HOMMEN, U. (2000): Modelling the ef- fects of life history traits and changing ecolo- gical conditions on the population dynamics and persistence of the endangered Malagasy giant jumping rat (Hypogeomys antimena). Anim. Conserv. 4, 333-343. SOMMER, S.; TicHy, H. (1999): MHC-Class II poly- morphism and paternity in the monogamous Hypogeomys antimena, the endangered, largest endemic Malagasy rodent. Mol. Ecol. 8, 1259-1272. 115 SPSS (1997): SPSS Base 7.5 for Windows User’s Guide. Version 7.5 for Windows. Chicago: SPSS Inc. STARCK, D. (1974): Die Säugetiere Madagaskars, ihre Lebensräume und ihre Geschichte. Wies- baden: Franz Steiner Verlag GmbH. SUTHERLAND, W. J. (1998): The importance of be- haviour in conservation biology. Anim. Be- hav. 56, 801-810. Author’s address: Dr. SIMONE SOMMER, Zoologisches Institut und Museum, Universität Hamburg, Martin-Luther- King-Platz 3, D-20146 Hamburg. (email: Simone. Sommer@zoologie.uni-hamburg.de). Mamm. biol. 66 (2001) 116-120 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication „ Mammalian Biology Zeitschrift für Säugetierkunde Mating behavior during the estrus cycle in female Mongolian gerbils (Meriones unguiculatus) By R. WeEınany, Syıvıa HOFMANN, and R. GATTERMANN Institute of Zoology, Martin-Luther-University Halle-Wittenberg, Halle, Germany Receipt of Ms. 21. 06. 2000 Acceptance of Ms. 15. 10. 2000 Key words: Mongolian gerbils, estrus cycle, mating behavior, vaginal smears Mongolian gerbils (Meriones unguiculatus) are common socially living rodents in the Steppe and semi-desert regions of Mongolia and Mandchuria (Gromov 1990). In their natural habitat, families, grouped around a founder pair, are strictly territorial (ÄGREN et al. 1989). Male behavior has been shown to be influenced by females (PRoBst and LorENZ 1987). Since the current literature on the female estrus cycle is limited and ambiguous (MaArston and CHAnG 1965; NiısHINo and TOTSUKAWA 1996), a redescrip- tion appears to be necessary. The aim of the present study was to obtain detailed data on the four stages of the estrus cycle in the Mongolian gerbil. Adult Mongolian gerbils (Meriones ungui- culatus) of both sexes from different litters aged 12-28 weeks were selected for this study. They were derived from our own la- boratory stock (Zoh:CRW) and were kept in climatised windowless rooms under a photoperiod of LD=14:10 (lights on at 0500h CET; 200-300 Ix during the light phase, approximately 51x during the dark phase). The room temperature was 23+2°C and the relative humidity varied between 65 and 70%. The animals were housed in plastic cages (55x 33x20 cm) 1616-5047/01/66/02-116 $ 15.00/0. with a wire mesh top. Tap-water and food pellets (Altromin® 7024, Altromin GmbH, Lage) were provided ad libitum. The ani- mal bedding (Allspan®, NL) was renewed every two weeks. Initially, the four different stages of the es- trus cycle were defined in adult females (n= 18) by taking vaginal smears daily be- tween two to four hours after lights on, over a period of two months. The stained smears were microscopically analysed (Leica”, Type DMRBE, x 200). In figure 1 the respective pattern of the four stages of the estrus cycle is depicted. Some females remained in diestrus for up to 14 days, i.e., the cycle became irregular or was arrested for that period of time. How- ever, it was always followed by the prees- trus and the estrus cycle proceeded regu- larly. Mating tests were performed during the four different stages of the estrus cycle of the gerbils. To prevent gravidity, adult but sexually unexperienced males were steri- lized by vasectomy. Two weeks after sur- gery they were taken to perform mating tests. Vaginal smears were taken from all 24 females to evaluate their stage of estrus cycle two hours before the start of the mat- ER * N S => ER SIDE ERE EN Ex = RR & D3 s nr \ RER RE Ex RS N 5 = ® nn & SS PR . %. ; ® S \ » ® no & s‘ = S at SR: N: S S 5 $ R \ \ s® ® SS ! $ 3.2 as : ® s | i \ 8 ® : N } - ES > S SEN j S S S R En \ S & SER ® SE SR S x - \ ; EN Ss ER = k SER EN S S RL S ; ® E i F 5 ER BE S S N | 2 x Ü 8 N SUR SER 4 ER \ Rn i OR = i Sr ER a & BE Fig. 1. Vaginal smears during the estrus cycle. Photo- graphed under microscope (X 200). 1. Preestrus: high number of squamous epithelial cells, absence of leukocytes and almost no cornified epithelial cells; 2. Estrus: low number of squamous epithelial cells, high number of dispersed cornified epithelial cells and no leukocytes; 3. Metestrus: mainly leukocytes, isolated squamous epithelial cells and/or cornified epithelial cells; 4. Diestrus: low num- ber of leukocytes, no or only a few squamous epithelial cells and/or cornified epithelial cells. Mating behavior during the estrus cycle 117 ing tests (20-30 minutes after lights off). The lowest number of females, to which a stage could be unambiguously assigned, counted 11. In the following, always 11 out of 24 females were randomly chosen before every mating test. For each stage the ani- mals were tested in a clean cage with new anımal bedding. Ten minutes before the fe- male was introduced, a vasectomized male was put into the cage. Each test lasted for ten minutes and the frequency of the fol- lowing activities of the females was regis- tered: copulation [c]: female is mounted by the male combined with friction move- ments; copulation trials [ct]: female presses tail to bottom and prevents the male, which tries to mount the female; lordosis [1]: fe- male remains in front of the male with bent hind paws and lifted tail; copulation avoid- ance behavior [cab]: female poses head to- wards the male, vocalizes and/or avoids the male, genitals and tail are directed away. Kruskal-Wallis analysis of variance and sub- sequent two-talled Mann-Whitney U-test (Winstat V 3.1) were used to assess differ- ences in the mating tests. Since multiple tests were run on the same basic dataset, the resulting p-values were corrected by the standard Bonferroni procedure. Differ- ences were accepted as significant at P<=OUOSICEnFEie 2). Figure 2 shows the results of the mating tests. The copulation behavior occurred ex- clusively in estrus (Kruskal-Wallis H-test: H-value = 20.23, n=11, p<0.05; Mann- Whitney U-Test estrus vs. preestrus, metes- trus and diestrus: in all cases U = 27.5, p = 0.0346). The number of copulation trials was highest during the preestrus and lowest at diestrus. This difference was significant (Kruskal-Wallis H-test: H-value = 10.86, n= 11, p< 0.05; Mann-Whitney U-Test pre- estrus vs. estrus: U =31, n.s.; preestrus vs. metestrus: U = 26.5, n.s.; preestrus vs. dies- trus: U=16, p=0.0188; estrus vs. metes- truss U=48 n.s; estrus vs. diestrus: WU -358, ansskznmetestnusehysinediestrus: U=50.5, n.s.). The lordotic behavior was mainly shown in the estrus (Kruskal-Wallis H-test: H-value = 18.37, n=11, p<0.05; Mann-Whitney U-Test preestrus vs. estrus: 118 R. WEInAnDy et al. 75 = 50 = (=) = Fe © Be) = 3 25 PEMD [c] [et] PEMD PEMD PEMD [1] [cab] Fig. 2. Frequency of behavioral parameters during the mating tests. Females (n = 11) were tested during prees- trus (P), estrus (E), metestrus (M) and diestrus (D). [c] = copulation; [ct] = copulation trials; [l] = lordosis; [cab] = copulation avoidance behavior. Median values and the interquartils are shown, differences are significant at p<0.05 and given as asterisks in the graph. U=21.5, p = 0.0369; preestrus vs. metes- trus: U=44, n.s.; preestrus vs. diestrus: Ub57, nlss»vestnussystmetestrusg U 165} p= 0.0048; estrus vs. diestrus: U = 20.5, p = 0.0244; metestrus vs. diestrus: U = 12.5, n.s.). There were no significant differences concerning copulation avoidance behavior towards the males during the estrus cycle (Kruskal-Wallis H-test: H-value = 7.21, n = 11,n.s.). The morning after the females were tested in estrus, 7 of the 24 tested fe- males developed a vaginal plug. In various rodents the uterus and the vagina as targets of ovarian hormones show cycle- dependent proliferation and apoptosis of luminal and glandular epithelium (SArto et al. 1997). The periodical increase and de- crease of squamous epithelial cells, leuko- cytes and cornified epithelial cells in vaginal smears is a consequence of these changes and has already been described for rats (OTHA 1995) or golden hamsters (SANDOwW et al. 1979; GATTERMANN et al. 1985) and re- liably indicates the estrus. In gerbils, the preestrus used to be characterized by an in- creased number of squamous epithelial cells and the absence of leukocytes (NISHINO and ToTsuKAwA 1996). The aggressiveness of the females was low and they displayed only minor copulation avoidance behavior to- wards the males. This belongs to precopula- tory behavior which may have a proceptive function (HoLman et al. 1985). The estrus stage is a period of characteristic behavior including sexual receptivity (lordotic pos- ture) in confrontation with males and the related vaginal smear pattern has already been described (BARFIELD and BEEMANN 1968; Apams and Norris 1973; VIck and Banks 1969). A further indicator for the re- ceptivity in Meriones unguiculatus is a vagi- nal plug (MArsTon and CHAnG 1965; NoR- rıs and Apams 1981). Due to the receptive stage, the interactions initiated by the fe- males were not aggressive during the mat- ing tests. The typical cellular pattern of metestrus was in some cases preceded by clustered cornified cells and isolated leuko- cytes. This has already been described in a previous study and classified as “estrus II” (NısHıno and ToTsuKAwA 1996). Our data do not confirm this suggestion, because our vaginal smear alike was always connected to metestrus behavior. A possible explana- tion for these contradictory results may be found in the diverging procedure, i.e., in Acknowledgements We wish to thank Dr. PETER FRITZSCHE and Dr. KARSTEN NEUMANN for their valuable comments on the manuscript, BIRGIT GEBHARDT for the fine technical assistance and KATE WILLIAMS for re- vising the English. We are grateful to Professor References ADAMS, C. E.; NoRrIs, M. L. (1973): Observations on reproduction in the Mongolian gerbil, Meriones unguiculatus. J. Reprod. Fert. 33, 185-188. ÄGREN, G.; ZHou, Q; ZHong, W. (1989): Ecology and social behaviour of Mongolian Gerbils, Meriones unguiculatus, at Xilinhot, Inner Mongolia, China. Anim. Behav. 37, 11-27. ÄGREN, G.; MEYERSoNn, B.]. (1977): Influence of gonadal hormones on the behavior of pair-liv- ing Mongolian gerbils (Meriones unguicula- tus) towards the cagemate versus non-cage- mate in a social choice test. Behav. Proc. 2, 325-335. BARFIELD, M. A.; BEEMANN,E. A. (1968): The oestrous cycle in the Mongolian gerbil, Meri- ones unguiculatus. J. Reprod. Fert. 17, 247- DS" GATTERMANN, R.; FRITZSCHE, P;, KRAMER,U. S. (1985): Zur Biorhythmik des Goldhamsters. 3. Mitt.: Infradiane Rhythmen. Zool. Jb. Phy- siol. 89, 265-278. GROMov, V. S. (1990): Social dominance and terri- torial behavior of Mongolian gerbil (Meriones Mating behavior during the estrus cycle 119 the cited investigation the animals were in- jected with pregnant mare serum gonado- tropin and human chorionic gonadotropin. The elevated level of gonadotropin might have extended the estrus without affecting the vaginal epithelium. In the present study the females displayed no sexual behavior in that stage. The diestrus is generally defined as a “state of rest” between met- and prees- trus, when the female was not fertilized. As described in an earlier study (ÄGREN and MEYERSoN 1977) the behavior of the fe- males is agonistic and biased towards avoid- ance. Our analysis of the estrus cycle re- vealed characteristic changes in mating behavior of female gerbils. GERALD MorıITZz for supplying the facilities for the photo documentation. This study was sup- ported by the Deutsche Forschungsgemeinschaft (Ga 437/12). unguiculatus). Doklady Akademii SSSR 314, 1268-1271. HOLMAN, S. D.; HUTCHINSON, J. B.; BURLEY, R. A. (1985): The cumulative effects of estrogen on precopulatory behavior in the female Mongolian gerbil. Physiol. Behav 35, 645- 649. MARSTON, J. H.; CHANG, M. C. (1965): The breed- ing, management and reproductive physiology of the Mongolian gerbil (Meriones unguicula- tus). Lab. Anim. Care 15, 34-48. NISHINO, N.; 'TOTSUKAWA, K. (1996): Study on the estrous cycle in the Mongolian gerbil (Meriones unguiculatus). Exp. Anim. 45, 283-288. NoRrRIS, M. L.; ADams, C. E. (1981): Time of mat- ing and associated changes in the vaginal smear of the post-parturient Mongolian gerbil (Meriones unguiculatus). Lab. Anim. 15, 193- 198. OTHA, Y. (1995): Sterility in neonatally androgen- ized female rats and the decidual cell reac- tion. Int. Rev. Cytol. 160, 1-52. PROBST, B.; LORENZ, M. (1987): Increased scent 120 R. WEınanpy et al. marking in male Mongolian gerbils by urinary polypeptides of female conspecifics. J. Chem. Ecol. 13, 851-860. SANDOW, B. A.; WEST, N. B.; Norman, R. L.; BREN- NER, R.M. (1979): Hormonal control of apop- tosis in hamster uterine luminal epithelium. Am. J. Anat. 156, 15-36. SATO, T.; FUKAZAWA, Y.; Koma, H.; ENARI, M.; Isuchi, T.; OTHA, Y. (1997): Apoptotic cell death during the estrous cycle in the rat uterus and vagina. Anat. Rec. 248, 76-83. VicK, L. H.; Banks, E. M. (1969): The estrous cy- cle and related behavior in the Mongolian gerbil (Meriones unguiculatus) Milne-Ed- wards. Commun. Behav. Biol. 3, 117-124. Authors’ address: RENE WEINANDY, SYLVIA HoFMANN, and ROoLF GATTERMANN, Institut für Zoologie, Martin- Luther-Universität Halle-Wittenberg, Dom- platz 4, D-06108 Halle, Germany (e-mail: weinandy@zoologie.uni-halle.de) Mamm. biol. 66 (2001) 121-126 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication GES SEEN DR, FR EN: > 6% ; Mammalian Biology SQ fa! en) Zeitschrift für Säugetierkunde I ERKUNDET The Bolivian bamboo rat, Dactylomys boliviensis (Rodentia: Echimyidae), a new record for chromosome number in amammal By J.L. Dunnum, J. SaLAazar-BRAVo, and T. L. YATEs Museum of Southwestern Biology and Department of Biology, University of New Mexico, Albuquerque, New Mexico, U.S.A. Receipt of Ms. 20. 09. 2000 Acceptance of Ms. 30. 10. 2000 Key words: Dactylomys boliviensis, Rodentia, Bolivia, new record The family Echimyidae, a highly diverse group of rodents, occurs throughout most of Central and South America. There are 16 recent genera and about 70 living spe- cies (Woops 1993), however, new taxa con- tinue to be described (e.g., DA Sınva 1998; PATToN et al. 2000). To date no comprehen- sive phylogenetic analysis is available for the group although great advances have been forthcoming (e.g., LARA et al. 1996; PATTon et al. 2000). The group is highly di- versified ecologically and has had a long evolutionary history in South America (PATTERSON and PAscuaL 1972, WoonDs 1982). Karyologically, less than half of the species have been analyzed but it is known that there is considerable variation in di- ploid (from 2n = 14 to 2n = 96) and funda- mental numbers (from FNa=18 to FNa = 144) (Tab. 1). One of the most spe- cialized groups within the Echimyidae is the subfamily of bamboo rats (Dactylomyi- nae). Woops (1993) placed three genera in the subfamily Dactylomyinae; Dactyloms, Kannabateomys, and Olallamys. The biol- ogy and evolutionary relationships of the Dactylomyinae are poorly known, likely due to their rarity to collectors and subse- quent scarcity in museum collections. What 1616-5047/01/66/02-121 $ 15.00/0. is known is well summarized in PATTon et al. (2000). Until recently (Anıskın 1993) no information on the chromosomal com- plement of any member of this group was available. As part of a long-term survey of the mam- mals of Bolivia, many new and important records for the country were collected (An- DERSON 1997). In July of 1992 and May of 1996, we took a total of five specimens of Dactylomys boliviensis (Bolivian bamboo rat) from a locality in the Yungas of La Paz (SALAZAR et al. 1994). Here we report the karyotype of this species, the highest chro- mosomal number known in amammal. The individuals were located and collected at night in a dense stand of bamboo and secondary growth within the village of La Reserva (Departamento La Paz, Nor Yungas, La Reserva, elev. 840 m, 15°44'S, 67°31'W) by following their distinctive calls and eye shine. The village of La Reser- va lies along Rio La Reserva, a small trıbu- tary of the Caranavi River. The village is at the bottom of a valley in the subtropical montane forest that covers most of the east- ern Andean slopes between 15° and 17°S latitude in the Cordillera Oriental of Boli- via. The foothills at this elevation are cov- 122 J. L. Dunnum et al. SIETETETE TE ET DB R RER REN SE RE AR a Kan 8 zu 9a ua ah Bo ra ad ansarn an DA HA AR 28 on on AR ad aa Ba Ha An RE 37 #= er na 32 sa» an Ar an aA 46 Er | Bi: _E, 5 “an 55 = 58 Fig. 1. Standard karyotype of Dactylomys boliviensis. ered with semi-deciduous vegetation inter- mingled with columnar cacti and bro- meliads. The forest is drier and sparser than at higher elevations. Compared to forests at higher and lower elevations, the trees are smaller, more highly branched, and most grow in open sun. The east facing slope above the river is steep, wıth much vegeta- tion, some secondary growth, and banana and tangerine cultivation. Palms and tree ferns are absent (SALAZAR et al. 1994). Chromosomal preparations were obtained using the technique described in ANDERSON et al. (1987). Metaphase cells were photo- graphed and scored to determine the di- ploid (2n) and fundamental numbers (FNa). One of us (JLD) scored 5 slides per animal and over 20 spreads per slide to de- termine chromosome numbers. The analysis of the morphology of the chromosomes was based on 10 metaphase plates from three Ä -_ 5 Ä % a Pi: ” XIX 3 individuals. Nomenclature for chromosome morphology and fundamental number fol- lows PATTon (1967). Chromosome slides, tissue samples, and cell suspensions are deposited in the Division of Biological Materials, Museum of South- western Biology (MSB). Voucher speci- mens are deposited at MSB (MSB 68547, MSB 85627, NK 40537), the American Mu- seum of Natural History (AMNH 264887, 264884), and the Colecciön Boliviana de Fauna (CBF 2608), in La Paz, Bolivia. The standard karyotype of Dactylomys bo- liviensis is highly asymmetrical, composed of 26 pairs of metacentric or sub-meta- centric autosomes and 32 pairs of acro- centric autosomes. The X chromosome is a large sub-metacentric and the Y chromo- some is a medium sub-metacentric. The re- sulting karyotype has a diploid count of 2n = 118 and FNa of 168 (Fig. 1). Chromo- Dactylomys boliviensis, new record for chromosome number Table 1. Diploid (2n) and fundamental number (FN) for members of the family Echimyidae. Dactylomys boliviensis Dactylomys dactylinus Echimys blainvillei Echimys dasythrix Echimys semivillosus Echimys sp. Echimys sp. Echimys sp. Isothrix bistriata Isothrix bistriata Isothrix pagurus Isothrix sinnamanensis Makalata armata Makalata didelphoides Clyomys laticeps Euryzygomatomys guiara Euryzygomatomys spinosus Hoplomys gymnurus Lonchothrix emiliae Mesomys hispidus Mesomys hispidus Mesomys occultus Proechimys albispinus Proechimys amphicorncus Proechimys brevicauda Proechimys canicollis Proechimys cuvieri Proechimys decumanus Proechimys echinothrix Proechimys gardneri Proechimys goeldii Proechimys guiarae Proechimys gularıs Proechimys guyannensis Proechimys ihering' Proechimys kulinae Proechimys mincae Proechimys oconnelli Proechimys oris Proechimys pattoni Proechimys poliopus Proechimys quadruplicatus Proechimys semispinosus Proechimys simonsi Proechimys steerei Proechimys trinitatus Proechimys unchi Proechimys yonenagae Proechimys sp. Proechimys sp. Proechimys sp. (Balta) Proechimys sp. (Barinas) Thrıcomys aperoides Thrıcomys aperoides Thricomys aperoides Reference this report Anıskın (1993) Reıs (1989) Lima et al. (1998) Acu1LerA et al. (1998) Lima et al. (1998) Anıskın (1993) Reıc (1989) Patron et al. (2000) LimAet al. (1998) Patron and Emmons (1985) VıE et al. (1996) Lima et al. (1998) Lima et al. (1998) Reıc (1989) Anıskın (1993) Reis (1989) Anıskın (1993) AnıskIn (1993) Lima et al. (1998) Patron et al. (2000) Parton et al. (2000) LeAL-MesauIta et al. (1992) Reis (1989) GARDNER and Emmons (1984) GARDNER and Emmons (1984) Mara and LAnGGUTH (1993) GARDNER and Emmons (1984) Patron et al. (2000) Patron et al. (2000) Patron et al. (2000) GARDNER and Emmons (1984) GARDNER and EmMmons (1984) GARDNER and Emmons (1984) Reıc (1989) Patron et al. (2000) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and EmMmons (1984) Patron et al. (2000) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and Emmons (1984) GARDNER and Emmons (1984) LeaL-Mesautma et al. (1992), Rocha (1995) Anıskın (1993) Reıs (1989) Reıs (1989) GARDNER and Emmons (1984) LeAL-Mesautta et al. (1993) Reıc (1989) AnıskIn (1993) 123 J. L. Dunnum et al. 124 180 160 140 - 120 100 80 - Fundamental number (FN) 60 40 20 0 10 20 30 40 50 70 80 90 100 110 120 130 Diploid number (2n) © Dactylomyinae Echimyinae Fig. 2. Karyogram of known echimyid karyotypes. some pair nine exhibits the characteristic satellite chromosome found in other echi- myid rodents. No chromosomal information is available for Kannabateomys, Olallamys, or D. perua- nus. Anıskın (1993) described the karyo- type of D. dactylinus (2n = 94, Fn = 144) from the Loreto Department in Peru. The karyotype of D. boliviensis differs from that of D. dactylinus by the presence of one ad- ditional set of meta or sub-metacentric pairs, and 10 pairs of acrocentric chromo- somes although comparisons are difficult due to the fact that Anıskın (1993) did not identify sex chromosomes. At least 14 Ro- bertsonian rearrangements would be neces- sary to transform the karyotype of one spe- cies into the other. We compiled a list of all species of echimyid rodents for which data were available © Eumysopinae (Tab. 1) and created a karyograph (IAmı and CROZIER 1980) based on chromosomal and fundamental numbers (Fig. 2). A defi- nite pattern of subfamily grouping is clear where two species of Dactylomys assume the highest positions on the plot and the echimyine rodents (Echimys, Makalata, Iso- thrix) are positioned at an intermediate le- vel (with the exception of /. pagurus and I. sinnamariensis). The most speciose and karyologically studied group is the Eumyso- pinae (represented in this sample by Pro- echimys, Clyomys, Euryzygomatomys, Hop- lomys, Lonchothrix, Mesomys, and Thri- chomys). For the most part these fall at the lower end of (Fig. 2). To date, no eumyso- pids have been found with a 2n > 65. Lıma et al. (1998) proposed that Robertso- nian rearrangements were more important in the evolution of the karyotype of arbo- Dactylomys boliviensis, new record for chromosome number real echimyids than other chromosomal re- arrangements because karyotypes of this group appeared to show higher levels of variation in diploid numbers than in funda- mental numbers. Our data do not support Lima et al. (1998). We found statistically significant differences in the levels of var- iation between diploid and fundamental number for the arboreal echimyids (Krus- kal-Wallis; P < 0.004), terrestrial echimyids (One-way ANOVA; a=0.05; P< 0.004), and for the entire echimyid radiation (Kruskal-Wallis; P < 0.001). However, in all cases the fundamental number varied more than the diploid number, suggesting that pericentric inversions may be more common. None the less it is quite likely that several processes may have influenced the evolution of the karyotype in this group. Acknowledgements This work was supported by grants from the Na- tional Science Foundation and a Research Ex- perience for Undergraduates supplement. We would like to thank the 1992 and 1996 members of the joint CBF-MNK-AMNH-MSB field expe- References AÄAGUILERA, M.; SANGINES, N.;, PEREZ-ZAPATA, A. (1998): Echimys semivillosus, a rodent with a very high chromosomal number. Caryologia 51, 181-187. ÄNDERSON, S. (1997). Mammals of Bolivia, Taxon- omy and Distribution. Bull. Am. Mus. Nat. Hist. 231. ÄNDERSON, S.; YATES, T.L.; CooK,J. A. (1987): Notes on Bolivian mammals 4: The genus Cie- nomys (Rodentia, Ctenomyidae) in the eastern lowlands. Am. Museum Novitates 2891, 1-20. ANISKIN, V.M. (1993): Three new karyotypes of prickly chinchillas of the family Echimyidae (Rodentia). Genetika 29, 1500-1507. CONTRERAS, L. C.; TORRES-MURA, J. C.; SPOTOR- no, A. E. (1990): The largest known chromo- some number for amammal, in a South Amer- ican desert rodent. Experientia 46, 506-508. DA SıLva,M.N.F. (1998): Four new species of spiny rats of the genus Proechimys (Rodentia: 125 Prior to our results, the highest chromo- some number reported for a mammal was 2n = 102 in Tympanoctomys barrerae (Con- TRERAS et al 1990). These authors also sug- gested that the family Octodontidae pre- sented the greatest chromosomal diversity. While this remains true for Fundamental number, the Echimyidae now represent the family with the greatest diversity in diploid number (2n = 14 to 2n = 118). Although Tympanoctomys and Dactylomys represent terminal branches in two different families of South American hystricognath rodents with a long history on this continent, they also share another characteristic: both occupy restricted ecological niches and pos- ses highly specialized life history traits. We concur with CoNTRERAS et al. (1990) in sug- gesting that the high chromosomal count ap- pears to be a derived character. ditions for their exceptional work. MIKE BoOGANn, JERRY DRAGOO, and WILLIAM GANNON reviewed earlier drafts. Special thanks to Suzy and SoMIYA. Collecting and exportation permits were facili- tated by the Coleccion Boliviana de Fauna. Echimyidae) from the western Amazon of Bra- zil. Proc. Biol. Soc. Washington 111, 436-471. GARDNER, A.; EMMoNS, L. (1984): Species groups in Proechimys (Rodentia, Echimyidae) as in- dicated by karyology and bullar morphology. J. Mammalogy 65, 10-25. ImAı, H.T.; CROZIER, R.H. (1980): Quantitative analysis of directionality in mammalian kar- yotype evolution. Am. Nat. 116, 537-569. TEARAYM IC ZEPAIONG IE IDRSIEyA,MEIN.E (1996): The simultaneous diversification of South American echimyid rodents (Hystri- cognathi) based on complete cytochrome b sequences. Mol. Phylog. Evol. 5, 403-413. LIMA, J.F. DE, S.; LANGGUTH, A.; DE SousA,L. C. (1998): The karyotype of Makalata_ didel- phoides (Rodentia, Echimyidae). Z. Säuge- tierkunde 63, 315-318. LEAL-MESQUITA, E. R.; FAGUNDES, V.;, YONENA- GA YASSUDA, Y.; RoCHA, P. DA. (1993): Com- 126 J. L. Dunnum et al. parative cytogenetic studies of 2 karyomorphs of Trichomys aperoides (Rodentia, Echimy- idae). Rev. Bras. Genetica 16, 639-651. LEAL-MESQUITA, E. R.; YONENAGA YASSUDA, Y.; CHu TIEn, H.; RocHA, P. da. (1992): Chromo- somal characterization and comparative cyto- genetic analysis of two species of Proechimys (Echimyidae, Rodentia) from the Caatinga domain of the state of Bahia, Brazil. Caryolo- gia 45, 197-212. MAIA, V.; LANGGUTH, A. (1993): Constitutive het- erochromatin polymorphism and NORs in Proechimys cuvieri Petter, 1978 (Rodentia, Echimyidae). Rev. Bras. Genetica 16, 145- 154. PATTERSON, B.; PAscuAL, R. (1972): The fossil mammal fauna of South America. In: Evolu- tion, Mammals and Southern Continents. Ed. by A.Keast, F.Erk and B. Grass. Albany: State University of New York Press. Pp. 247- 309. PATToN, J. L. (1967): Chromosome studies of cer- tain pocket mice, genus Perognathus (Roden- tia: Heteromyidae). J Mammalogy 48, 27-37. PATTON, J. L.; EMMons, L.H. (1985): A review of the genus /sothrix (Rodentia: Echimyidae). Am. Museum Novitates 2817, 1-14. PATToN, J. L.; DA SıLva, M.N.F; MALCOLM, J.R. (2000): Mammals of the Rio Juruä and the evolutionary and ecological diversification of Amazonia. Bull. Am. Mus. Nat. Hist. 244. Reıc, ©. (1989): Karyotypic repatterning as one triggering factor in cases of explosive specia- tion. In: Evolutionary Biology of Unstable Populations. Ed. by A. FONTDEVILA. Berlin, Heidelberg. Springer-Verlag. Pp. 247-289. RocHA, P. L. B. DA (1995): Proechimys yonenagae, a new species of spiny rat (Rodentia, Echi- myidae) from fossil sand dunes in the Brazi- lian Caatinga. Mammalia 59, 537-549. SALAZAR-BRAVO, J.; CAMPBELL, M.; ANDERSON, S.; GARDNER, S.; DUNNUM, J. (1994): New records of Bolivian mammals. Mammalia 58, 125-130. VIE, J., VOLOBOUENV, V.; PATTON,J.;, GRANJON,L. (1996): A new species of /sothrix (Rodentia, Echimyidae) from French Guiana. Mammalia 60, 393. Woops, €. A. (1982): The history and classifica- tion of South American hystricognath ro- dents: Reflections on the far away and long ago. In: Mammalian Biology in South Ameri- ca. Ed. by M. Mares and H. GENowAays. Pitts- burgh, Pennsylvania: Pymatuning Laboratory of Ecology Vol. 6, 377-392. Woops, C. A. (1993): Family Echimyidae (Gray, 1825). In: Mammal Species of the World, A Taxonomic and Geographic Reference. Ed. by D. E. Wırson and D. M. REEDER. Washing- ton, DC: Smithsonian Institution Press. Pp. 789-799. Authors’ address: JONATHAN L. DUNNUM, JORGE SALAZAR-BRAVO, TERRY L. YATEs, Museum of Southwestern Biol- ogy, Department of Biology, University of New Mexico, Albuquerque, New Mexico 87131, U.S.A. Mamm. biol. 66 (2001) 127-128 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Book reviews DENZAU, GERTRUD; DENZAU, H.: Wildesel. Stutt- gart: Jan Thorbecke Verlag 1999. 221 pp., zahl- reiche farbige Abbildungen und Karten. DM 79,—. ISBN 3-7995-9081-1. Eine der frühesten Erinnerungen, welche der Re- zensent aus seinen Studententagen an den frühe- ren Vorsitzenden unserer Gesellschaft, Herrn Professor Dr. Dr. WoLr HERRE, hat, bezieht sich auf dessen Ankündigung, nach der Emeritierung ein Buch über die Domestikation und Kultur- geschichte des Esels schreiben zu wollen. Leider verstarb Prof. HERRE am 12. November 1997 ohne ein entsprechendes Werk hinterlassen zu haben. Es ist aber sehr wahrscheinlich, daß er von dem hier vorzustellenden Band von GERTRUD und HELMUT DENZAU sehr beeindruckt und erfreut gewesen wäre, haben doch die Autoren mit einem klaren und informativen Text die Natur- und auch die Kulturgeschichte der Wildesel dargestellt und ihre Darstellung mit beeindruckenden, häufig so- gar hinreißenden, Farbphotos illustriert. Im vorliegenden Buch wird unterschieden zwischen dem Afrikanischen Wildesel (Equus afri- canus mit drei Unterarten, von denen eine — E. a. atlanticus — ausgestorben ist) und den beiden Asiatischen Wildeseln (Equus hemionus mit sechs Unterarten, davon eine - E. h. hemippus — ausge- storben, und Equus kiang mit drei Unterarten). Zunächst wird in einem kurzen Kapitel die Stam- mesgeschichte der Equiden mit besonderer Be- tonung der Esel besprochen, dann folgen in einem Kapitel mit dem irreführenden Titel „Wis- senschaftsgeschichte“ Darstellungen der taxono- mischen, osteologischen, genetischen und etholo- gischen Charakteristika der Esel, ferner Bemerkungen zu Zoobeobachtungen und zu - weitgehend erfolglosen — Zähmungsversuchen. Ein sehr instruktives, mit anschaulichen, meist far- bigen, Karten illustriertes Kapitel stellt die Le- bensräume von Khur, Kulan, Onager und Dschig- getai (vier Unterarten von E. hemionus), sowie von Kiang (E. kiang) und Somali-Wildesel (E. afri- canus somaliensis) vor. Das folgende ausführliche Kapitel bietet nicht nur Angaben zur Entdeckung, Beschreibung und Erforschung dieser Arten, son- dern auch zum ausgestorbenen Syrischen Halbesel (E. hemionus hemippus). Es enthält auch Auszüge von Beobachtungsprotokollen, welche die Autoren auf ihren Expeditionen anfertigten. Die Kulturgeschichte der wilden Esel wird von der Vor- und Frühgeschichte her aufgerollt und 1616-5047/01/66/02-127 $ 15.00/0. Mammalian Biology Zeitschrift für Säugetierkunde mit Berichten aus verschiedenen Kulturkreisen und Epochen - mit dem Gilgamesch-Epos begin- nend - illustriert. Die Fragen zur Domestikation des Afrikanischen Wildesels (E. africanus), die vermutlich im 5. oder 4. Jahrtausend v. Chr. in Nordafrika oder Westasien stattgefunden hat, so- wie zu Zähmungs- und Kreuzungsversuchen mit Halbeseln, „sind keineswegs endgültig geklärt“. Eine Schilderung der gegenwärtigen Lebensbe- dingungen wilder Esel unter Berücksichtigung von Wiedereinbürgerungsversuchen und der Ein- flüsse des Tourismus auf die Lebensräume der Wildesel, sowie ein Ausblick auf mögliche zu- künftige Entwicklungen schließen den Textteil dieses für den Spezialisten und den Laien höchst reizvollen Buches ab. Es folgt ein Anhang, in dem u.a. ein ausführliches Literaturverzeichnis (Seng bedruckte Seiten), ein Personenregister und ein Index geboten werden. P. LANGER, Gießen FLADE, J. E.: Die Esel, Haus- und Wildesel, Equus asinus. Hohenwarsleben: Westarp Wissenschaften 2000. Die Neue Brehm-Bücherei Bd. 638. 126 pp., 5 Farbtafeln, 41sw Abb., 21 Tab. DM 39,90/ ös 292,-/sFr 39,90. ISBN 3-89432-887-8. Am Ende des vorliegenden Bandes finden sich Informationen der „Interessengemeinschaft für Eselfreunde in Deutschland e.V.“. Dieser An- hang darf nicht zur Vermutung verleiten, daß es sich bei dem Buch nur um eine emotionsgeladene „Liebeserklärung“ für das Grautier handelt; der Text bietet vielmehr Informationen zu Stammes- geschichte und Systematik, sowie zur Verbreitung der Wildesel, auch widmet sich ein Kapitel der Geschichte der Domestikation des Esels, sowie ein weiteres seiner Verbreitung und Kultur- geschichte. Im Text, der durch zahlreiche Tabellen ergänzt wird, bietet der Autor diverse äußere Körpermaße verschiedener Eselrassen und schil- dert das Verhalten und die Sinnesleistungen, so- wie die Haltung von Hauseseln. Ein klares, durch viele Tabellen ergänztes Kapitel beschreibt die verschiedenen Populationen des Hausesels in der Mittelmeerregion, Afrika, Europa, sowie in Mit- tel- und Südamerika. Die Nutzung von Eseln zur Maultierproduktion in vielen Teilen der Welt wird ebenfalls behandelt. Die Fülle der im vorliegenden Band gebotenen Informationen ist bemerkenswert. Es werden ver- gleichende „Seitenblicke“ auf das Hauspferd, ins- besondere auf Ponies, geworfen werden. Der 128 Buchbesprechungen Autor, der seine Sympathie für die behandelte Säugetierart nicht verleugnet, verfaßte einen Band, der sich würdig in die bewährte Reihe der Bände der „Neuen Brehm-Bücherei“ eingliedert. P. LANGER, Gießen FELDHAMER, G. A.; DRICKAMER, L. C.; VESSEY, S. H.; MERRITT, J. F.: Mammalogy. Adaptation, Diversity, and Ecology. New York: McGraw Hill 1999. 576 pp.; numerous figs. and tabs. DM: 160,-. ISBN 0-697-16733-X This new textbook on mammalogy, written by in- ternationally known and experienced authors, covers a broad range of zoological disciplines and briefly summarizes today’s knowledge on this animal group under diverse points of view. It is in- tended to serve in undergraduate and graduate courses at North American universities and is ad- dressed primarily to students with a basic back- ground in vertebrate biology but to teachers as well. Content and length are calculated to cover a one-semester course. The text is arranged in 5 parts with 29 chapters, where of part I (chapters 1-4) deals with mam- malogy as a special science, with its history, with methods and techniques of investigations to reach various aims, and with the emergence of modern mammals in an adaptive evolutionary process from synapsid reptilian roots through mesozoic radiation. Part II (5-9) is devoted to an overview on structure and function. Here, the following features are focused: integument and derivatives, basic skeleton and muscular arrangement, modes of locomotion, intestines, modes of feeding and foraging, sensory organs, central nervous system, endocrine system, biological rhythms, tempera- ture regulations, water balance, reproductive sys- tems, gestation, parturition, and lactation. Part III (10-19) on adaptive radiation and diversity de- scribes morphological and anatomical peculiari- ties and fossil history of the mammalıan orders. The aim of Part IV (20-26) is on overview on be- haviour, ecology, and biogeography with chapters on communication, aggression and spatial rela- tions, sexual selection, parental care, and mating systems, social behaviour, dispersal, habitat selec- tion, migration, and homing, as well as popula- tions and life history, community ecology, and zoogeopraphy. Finally, Part V (27-29) is rather short but attributed to special topics such as para- sites and diseases, domestication and domesti- cated mammals, and conservation. Each of the chapters closes with a brief summary, some spe- cial questions and suggested readings. At the end of the book a glossary is given, a list of references as well as subject and species indices. The text is clearly written and corroborated with many instructive drawings, pictures, other fig- ures, and tables. Thus, the book reads fluently and describes main biological phenomena and special adaptations in convergent and divergent phylogenetic lineages. However, Part V deserves special attention since it covers important themes, usually not dealt with in comparable books. This is in general praiseworthy but unfor- tunately the mode of treatment and presentation are not satisfying. There are several points of se- vere criticism, especially concerning domestica- tion. This chapter is presented rather confusingly and not according to the current state of knowl- edge. The authors do not seem to be well in- formed about this matter and the text does not indicate any general reflection on this special field of zoology. Some of the descriptions and statements are strictly false, contradictory or at least incorrect. Domestication research on mam- mals was not only accomplished to answer ar- chaeozoological questions, to achieve culture- historical results or to describe efforts of ani- mal-husbandry as is mainly mentioned here. A more strongly zoological point of view and ap- propriate evaluations are missing. In this sense, domestication was and still is the longest and greatest experiment of man with anımals gener- ally demonstrating an enormous variability and changeability of the organism at the species level with several convergent adaptations in diverse lineages. Some of these changes are doubtlessly intended through breeding, many more are not. However, the potential for all these must have been present in the ancestral gene pool but re- mained latent under natural selection. Thus, in princeipal, domestication can serve as a model for ideas on phenomena of transspecific evolu- tionary processes in nature although not occur- ring in domestication and leading to other direc- tions through evolution. The origin of new species must have started at the species level from certain preadapted ancestors. There are nu- merous zoological studies since DARWIn, mainly from European countries, that compare indivi- duals of wild ancestral species with domesticated derivatives concerning quantitative anatomy, physiology, ethology and genetics. Accordingly, concluding treatises are available but not men- tioned in this book. Although some of these stu- dies were not published in English, they are not less important. However, altogether this book is an important source and suited to support mam- malogy also at European universities, but correc- tions of the queries seem appropriate in future editions. D. KrRUSskKA, Kiel Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana- ging editor, Prof. Dr. D. Kruska, Institut für Haustier- kunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni- kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manu- script will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21 x 30 cm) or correspondingly either from a typewriter or legible computer print with a left-hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscript. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscnpt following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type- written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscript. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings Mammalian Biology 2 Zeitschrift für Säugetierkunde depending on the theme, Acknowledgements, Zusam- menfassung (in German), References. A German title must precede the Zusammenfassung in English manu- scripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: Kauko, E. K. V.; Hanouey, C. O0. Jr. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1321), from Panama. Z. Säugetierkunde 59, 257-273. b) From books: HERRE, W.; RöHrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers‘ symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscript is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen, Germany (email:peter.langer@anatomie.med. uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). e = URBAN & FISCHER Verla Mammalian Biology Jena Cem - . Be = n ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. : 66(2001)2 - 65-128 Contents Original investigations Dubost, G.: Comparison of the social behaviour of captive sympatric peccary species (genus Tayassu); correlations with their ecological characteristics - Vergleiche des Sozialverhaltens zweier sympatrischer Pekariarten (Genus Tayassu) in Menschenobhut ; Beziehungen mit ihren ökologischen Merkmalen ___ 65 Reutter, B. A.; Brünner, H.; Vogel, P.: Biochemical identification of three sympatric Apodemus species by protein electrophoresis of blood samples - Biochemische Bestimmung dreier sympatrisch vorkommender Apodemus-Arten mittels Elektrophorese von Blutproteinen 0 2 2.00 0m DE Caro, T. M.; Brock, R.; Kelly, M.: Diversity of mammals in the Bladen Nature Reserve, Belize, and factors affecting their trapping success - Diversität von Säugern im Bladen Naturreservat, Belize, und Faktoren, die einen Fangerfolg beeinflussen __ 90 Röhrs, M.; Ebinger, P.: Welche quantitativen Beziehungen bestehen bei Säugetieren zwischen Schädelkapazität und Hirnvolumen? - How is cranial capacity related to brain volume inmammals? an wer 1 Short communications Sommer, 5.: Reproductive ecology of the endangered monogamous Malagasy giant jumping rat, Hypogeomys antımena - Reproduktionsökologie der bedrohten, monogamen madagassischen Springratte, Hypogeomys anttmena____ 0 a0 Weinandy, R.; Hofmann, S.; Gattermann, R.: Mating behavior during the estrus cycle in female Mongolian gerbils (Meriones unguiculatus) - Paarungsverhalten während des Ostruszyklus bei der weiblich Mongolischen Wüstenrennmaus (Meriones unguiculatus) — 202 20 DIEBE Dunnum, J. L.; Salazar-Bravo, J.; Yates, T.L.: The Bolivian bamboo rat, Dactylomys boliviensis (Rodentia: Echimyidae), a new record for chromosome number in a mammal - Die bolivianische Bambusratte, Dactylomys boliviensis (Rodentia: Echimyidae), ein neuer Rekord für die Chromosomenzahl bei einem Säugetierr__ 121 Book reviews OO 2. Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record 2 M lammalian Biology Zeitschrift für Säugetierkunde 32001 — ISSN 1616-5047 URBAN & FISCHER Mamm. biol. - 66(2001)3 - 129-192 _ www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Saugetierkunde (German Society for Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel R. Hutterer, Bonn E. Kalko, Ulm r.LüpsBern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repenning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.O. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6%XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/46 35 70, E-mail: l.still@nature.com Subscribers in the USA can call toll free on +1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 498.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipients name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by personal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. If no cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH &) As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANSI/NISO 239.48-1992 (Permanence of Paper). ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Printed in Germany Mamm. biol. 66 (2001) 129-134 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Small mammal exploitation of upper vegetation strata in non-forest, mixed farmland habitats By K. Norovig, J. REDDERSEN, and T. S. JENSEN Department of Landscape Ecology, National Environmental Research Institute, Kalö and Natural History Museum, Aarhus, Denmark Receipt of Ms. 31. 01. 2000 Acceptance of Ms. 20. 10. 2000 Abstract In September 1998, activity of small mammals in upper herbal and shrub vegetation strata was ex- amined and quantified in a standardized field experiment using a paired upper stratum (height = 0.5 m) vs. ground stratum trapping design. Trapping took place across a range of differ- ent biotope types in a typical Danish mixed farmland within the Kolindsund area, Djursland, Den- mark. Seven small mammal species were encountered among the 409 catches during 776 trapnights. Upper vegetation stratum activity was considerable in Micromys minutus and Apodemus flaviıcollis. Clethrionomys glareolus, Sorex minutus and Sorex araneus also exploited upper vegetation strata but only to a lesser extent. Microtus agrestis and Apodemus sylvaticus were only caught in ground Stratum traps. Our results show that, at low to moderate densities, some small mammal species, such as M. minu- tus, may be greatly underestimated or entirely missed by a traditional ground trapping grid. There- fore, the potential upper vegetation stratum activity in some small mammal species should gener- ally be considered in studies performed in areas or seasons with vigorous herbal or shrub vegetation suitable for climbing. Key words: rodents, shrews, arboreal activity, vertical movement Introduction In general, it is well known that some ro- dent species are capable of climbing trees, shrubs etc. In Europe, this is mainly re- ported for some species of Apodemus, Cle- thrionomys and Micromys and is usually de- scribed as ‘arboreal activity’. However, the issue has only rarely been subjected to sys- tematic and quantitative studies. In unmanaged old forests, OLSZEWSKI (1968) found that Apodemus flavicollis is 1616-5047/01/66/03-129 $ 15.00/0. frequently exploited uprooted trees as ar- boreal runways, while Clethrionomys gla- reolus never did. In spruce-oak woodland, Horısova (1969) demonstrated consider- able arboreal activity in both A. flavicollis (43%) and C. glareolus (17%) and further listed a Sorex minutus specimen caught at a height of 3 m as the first published report of arboreal activity in European shrews. In deciduous woodland, MONTGOMERY (1980) 130 K. NorDdvIG et al. found extensive arboreality in A. flavicollis, A. sylvaticus and C. glareolus, although less so in C. glareolus, while in the same habitat type, in the absence of A. flavicollis, TATTER- SALL and WHITEBREAD (1994) found arbore- ality in both C. glareolus (14%) and A. syl- vaticus (20%). As a part of performing a larger conven- tional small mammal ground trapping pro- gramme across a range of uncultivated farmland biotope types, we wanted to esti- mate species biases when trapping only on the ground. Using a paired ground and upper vegetation stratum trapping design, the present study analyses vertical activity in seven small mammal species across var- ious non-forest biotopes with tall herbal or shrubby vegetation. Material and methods The study was conducted in September 1998 with- in a 1x2 km study area in a typıcal Danish inten- sive mixed farmland landscape situated in the re- claimed former fiord Kolindsund, Djursland, Denmark. Six trap lines were established at one side of each of six existing trapping grids and each comprised from 10 to 28trap points (total no.=97) with each trap point 10 m apart. Trap lines ran through woody, shrubby or tall herbal vegetation, viz. along ditches, fringes of wood fragments, short rotation coppiced (SRC) biomass willow plantations, hedgerow bottoms and grass banks and into uncut perennial set-aside areas. Within and between trap lines, the environment of trap points varied with respect to vegetation type. However, we assumed the trap lines, in total, to be fairly representative of main non-field bio- tope types - in the study area and in Danish farm- land in general. Each trap point consisted of two Usgglan-live traps. One trap was placed on a wooden platform nailed to a stake 0.5 m above ground level, and the other below it on the ground. In all cases, the immediate trap environment was a dense vegeta- tion composed of grasses, reed, mugworts, nettles, bramble of varying height (0.5-2 m), and all plat- form traps were set in a way allowing small mam- mals to gain access to the trap by climbing the ve- getation. Traps were baited with oatmeal and apple slices, supplied with dry bedding and were checked daily in four-day series. Trap lines were run one at a time and only once and hence, the total material comprised 776 trap nights. Captives were sexed, weighed, aged and all rodents individually marked and identified using PIT’s (Passive-Inte- grated-Iransponders) and PIT-scanner, and then released at the capture site. Visiting traps once per day, trap mortality was considerable in shrews, while it was negligible in rodents. For each species, we performed a trap point based pairwise comparison of upper vegetation stratum vs. ground stratum catches across all six trap lines using the non-parametrice “Wilcoxon Matched- Pairs Signed-Ranks Test” (SIEGEL and CASTELLAN, 1988). The null-hypothesis ‚“no differences be- tween levels” was tested two-sided. Maximizing the exploitation of the dataset and the statistical power of the analysis may often conflict with test assumptions of independence between observations in frequency data. Viola- tions of the independence criterion may depend on determent of other species, on attraction to conspecifics in the traps (KALımowskA 1971; MONTGOMERY 1979; ANDRZEJEWSKI et al. 1997) or on trap addiction (TAnton 1965). Accordingly, we produced a hierarchy of progressively reduced datasets and thus progressively conservative tests 1-4 (cf. Tab. 1). 1. Dataset 1 included all captures. 2. Dataset 2 comprised Dataset 1 minus cases with simulta- neous captures in both traps at one trap point to avoid a possible determent effect (second individ- ual more likely to enter trap 2 because trap1 is occupied). 3. Dataset3 comprised Dataset 2 minus cases when more than one individual was caught simultaneously in a trap (excluding deter- ment but also attraction effects). 4. Dataset 4 comprised Dataset 3, but only included the first catch of an individual in a specific trap (avoiding individual trap addiction). Results The total material comprised 409 catches and seven small mammal species corre- sponding to a gross mean of 53 small mam- mal catches per 100 trap nights (including captures, dead and recaptures). Within each species, the total number of catches at the upper and ground stratum, respectively, is presented in table 1. All seven species sig- nificantly responded to trap stratum in Da- taset 1: two species, Micromys minutus and A. flavicollis, occurred more frequently in upper vegetation stratum traps, while the Small mammal exploitation of upper vegetation strata 131 Table 1. Total individual numbers (N :N) in trap catches of small mammals in paired traps, i.e. upper vegetation stratum (0.5 m) vs. ground stratum (0 m) levels, listing also P-values from Wilcoxon Matched-Pairs Signed-Ranks Test (two-sided; null-hypothesis assuming no difference between levels). Species are sorted by proportion of upper stratum activity (descending). For non-significant results and large N, exact P-values are given in (), while, for small N, only ‘NS’ is given as tables only yield critical values. For very small samples, the power of the test is insufficient to detect significance (’nt': no test performed). (E5472370:0004.2 77 =20:01 and 7, P= 0.05) Datasets M. minutus A. flavi- collis Dataset 1 Above ground vs. Ground P Dataset 2 Above ground vs. Ground 2 BHRI22 ? NS (0,17) Dataset 3 Above ground vs. Ground 5:10 2113 P NS (nt) NS (0.23) Dataset 4 Above ground vs. Ground 4:0 1 P NS (nt) NS (0.34) other five species occurred more frequently (or even exclusively) in ground stratum traps. M. minutus showed a consistent and strong tendency to higher catches in upper stratum within all four datasets, but differences were only significant in Dataset 1 and 2. However, non-significance in Datasets 3 and 4 were entirely due to the sample size reduced below the critical size of the test. A. flavicollis exhibited considerably higher catches (c. 2.5x) in upper stratum in Data- set 1. In the progressively reduced Datasets 2-4, the tendency for higher catches in upper stratum remained but seemed less pronounced (c. 1.6x) which, along with the reduction in sample size, resulted in non- significance. This change in proportions of upper and ground level catches was the largest observed between datasets, and was slightly significant (Testing proportions in Dataset 2 against proportions in material discarded from Dataset 1 to produce Data- set 2; Chi-test; X” = 3.92; df=1; P< 0.05). C. glareolus was caught most often on the ground in all four datasets. Still, in all data- sets, more than 20% of catches occurred in C. glareolus 5. minu- S. ara- M. agrestis A. sylva- neus ficus 305159 “Arkr 23:86 Arkr 177353 Akr 12:45 krr upper stratum traps. Sorex minutus and S. araneus, the two shrew species, occurred most often in the traps on the ground (with similar but non-significant pattern in 5. minutus in Datasets 3-4). In both species, however, specimens were occasionally caught at upper stratum. Microtus agrestis and A. sy/vaticus were not caught in upper stratum traps in our material of 19 and 9 catches, respectively. Discussion Our results show that two rodent species, M. minutus and A. flavicollis, may have a high proportion of activity in vegetation strata considerably above the soil surface. In our trapping design within non-forest ve- getation types, we actually observed more catches at the 0.5 m level than at the soil surface. Also, our results showed that most small mammal species may visit upper ve- getation strata, at least occasionally. For that reason, the influence of upper vegeta- tion stratum activity of small mammals should be considered in most kinds of tall 132 K. NorDdvis et al. vegetation when studying population size and density, home range, intra- and inter- specific competition and food energetics. Three-dimensional use of the habitat space might be an integral component of several behaviours such as escape from predators, exploration, fouraging and intra- and inter- specific competition (HoLBRooK 1979). As a consequence, differential changes in food availability between ground and upper ve- getation strata may lead to seasonal changes in arboreality (MONTGOMERY 1980). Our study was performed in September when upper vegetation level has a high cov- erage and offers an abundance of food to frugivorous, granivorous, and insectivorous small mammals. Thus, activity at this stra- tum could be different and possibly lower during other seasons. We trust our results not to be an artefact of our experimental setup, i. e., small mammals gaining access to upper levels by climbing our platform stakes. Using a design similar to ours, MESERVE (1977) demonstrated that such direct stake climbing only accounted for 4% of all upper level records. It is acommon experience that catchability of M. minutus is reduced during the summer months (e. g. TROUT 1978), probably because the species exploits upper vegetation strata in the search for food and nest sites. Analys- ing a large material of nests, FELDMANN (1997) showed that in moist habitats, most nests were placed high (mean H > 0.5 m) in tall grasses, mainly Phalaris arundinacea and Phragmites australis, with some nests also occurring at lower levels (mean H = 0.35 m) in grass tussocks (such as Dac- tylis glomerata and Molinia coerulea). Our results confirm the findings of WARNER and Bart (1976) that catches can be improved by mounting traps on stakes, and demon- strate, that M. minutus utilizes the vertical space intensively. Consequently, the chance of recording M. minutus from a site, using only conventional ground trapping methods, may be rather low, unless densities are high and/or trapping is intensive. Further, we found considerable vertical movements in other small mammal species indicating that such species regularly ex- plore and/or exploit upper vegetation stra- ta, and, in particular, we showed that this applies to non-woody vegetations, e.g., tall herbs and grasses. For this reason, we do not use the phrase ‘arboreal’ and, in gener- al, believe it to be a too narrow and some- what misleading term for describing the biology of the species in question. We found A. flavicollis to have a high pro- portion of upper level catches, 60-71% in Dataset4 and 1, respectively, which is roughly similar to the arboreality reported for this species by Horisova (1969) and MONTGOMERY (1980), viz. 43% and 47- 63%, respectively. In C. glareolus we had more catches at the ground than at the upper level, with the latter comprising 20- 24%. Sımilarly, MONTGOMERY (1980), OL- SZEWSKI (1968), HorısovA (1969), and TAT- TERSALL and WHITEBREAD (1994) all re- ported mean estimates below 50%, viz. 42, 35, 17 and 14%, respectively. The two shrew species, S. araneus and 5. minutus were both most often caught in ground traps, but it is worth noting that, across both species, 7 out of 47 catches (15%) occurred at the upper level. Considering the morphology and feeding biology of M. agrestis, it is not surprising that we caught this species exclusively in ground traps. M. agrestis ıs heavy and has short limbs and tail, which do not facilitate climbing, and its diet is almost exclusively composed of grasses. Climbing does not seem to occur in this species. Regarding climbing in A. sylvaticus, the lit- erature offers different viewpoints. HOFF- MEYER (1973) and CorkE (1974, cited from MONTGOMERY 1980) suggested that coexis- tence of A. flavicollis and A. sylvaticus re- lied on vertical separation, with A. flavicol- lis being more arboreal than A. sylvaticus, whereas MONTGOMERY (1980) and TATTER- SALL and WHITEBREAD (1994) found consid- erable arboreal activity in the species, ca. 50% and 20%, respectively. Our scarce data on A. sylvanicus favours the idea of vertical separation, but must, significant or not, be viewed at with considerable caution. In gen- eral, arboreal activity in this species cannot a priori be excluded. Small mammal exploitation of upper vegetation strata Our study in a unique way tested succes- sively reduced datasets 1-4 in a systematic attempt to avoid various biases which might impair statistical assumptions of indepen- dence. However, with the exception of A. flavicollis (between Dataset1l and 2), our results did not change considerably in this process which may indicate that such analyses are rather robust against well- known potential biases such as trap addic- tion, determent, attraction, etc. At any rate, we suggest that this type of approach may be explored in further detail, analysing the effects of various biases to small mammal trapping data. In conclusion, our results show that, at least temporarily, a few small mammal species utilize the vertical space considerably or Zusammenfassung 133 even entirely and that most species do it at least occassıonally, and that this also applies to other than woody habitats. Especially M. minutus and A. flavicollis catches may, in many situations, be greatly enhanced by trapping at above soil surface levels. Acknowledgements The present work formed a part of a study of the habitat potential for wildlife of SRC-willow bio- mass plantations within mixed farmland land- scapes, which was financed by the Danish Minis- try of Environment and Energy. Supplementary finance from University of Aarhus, Department of Zoology, is likewise gratefully acknowledged. Farmer CLaus WIsTtorT kindly allowed access to his estate. Ausnutzung höherer Vegetationsstrata durch Kleinsäuger in gemischten Ackerland-Habitaten Im September 1998 wurde die Ausnutzung höherer Vegetationsstrata durch Kleinsäuger in Kraut- und Strauchvegetation in einem standardisierten Feldexperiment untersucht und quantifiziert. Das Experiment wurde mit Hilfe von paarweise kombinierten Fallen am Boden und in der Höhe von 0,5 m in einem typischen dänischen Ackerland in Kolindsund, Jütland, durchgeführt. In 776 Fallen- nächten wurden bei 409 Fängen sieben Kleinsäugerarten in den Fallen gefangen. Aktivität in der Höhe von 0,5 m war bei Micromys minutus und Apodemus flavicollis erheblich, während Clethrio- nomys glareolus, Sorex minutus und Sorex araneus die oberen Vegetationsschichten in geringerem Ausmaß nutzten. Microtus agrestis und Apodemus sylvaticus wurden nur in den Bodenfallen gefan- gen. Unsere Ergebnisse zeigen, daß bei niedrigen und mäßigen Dichten das Vorkommen von z.B. M. minutus von einem gewöhnlichen Bodenfallengitter unterschätzt oder vollkommen übersehen werden kann. Daher muß man auch bei Studien der meisten anderen Kleinsäugerarten diese Aktivi- tät in oberen Schichten berücksichtigten. References ANDRZEJEWSKI, R.: BABINSKA-WERKA, J.; LIRO, A.; OWADOWSKA, E.; SZACKIı, J. (1997): The attrac- tiveness of conspecific and interspecific odor for bank voles Clethrionomys glareolus. Acta Theriol. 42, 231-234. CoRKE, D. (1974): The comparative ecology of the two British species of the genus Apodemus (Rodentia: Muridae). Ph. D. Thesis, Univer- sity of London. FELDMANN, R. (1997): Studien zur Autökologie und Fortpflanzungsbiologie der Zwergmaus, Micromys minutus. Abh. Westf. Mus. Na- turkd. 59, 107-115. HOFFMEYER, I. (1973): Interaction and habitat se- lection in the mice Apodemus flavicollis and A. sylvaticus. Oikos 24, 108-116. HOLBROOR, S. J. (1979): Vegetational affinities, ar- boreal activity and coexistence of three spe- cies of rodents. J. Mammalogy 60, 528-542. Horiısova, V. (1969): Vertical movements of some small mammals in a forest. Zool. Listy 18, 121-141. KALINoWSKA, A. (1971): Trapping of Apodemus flavicollis and Clethrionomys glareolus into a Double Trap. Acta Theriol. 16, 73-78. MESERVE, P.L. (1977): Three-dimensionel home 134 K. Norovic et al. ranges of cricetid rodents. J. Mammalogy 58, 549-558. MONTGOMERY, W. I. (1979): An examination of in- terspecific, sexual and individual biases affect- ing rodent captures in Longworth traps. Acta Theriol. 24, 34-45. MONTGOMERY, W. I. (1980): The use of arboreal runways by the woodland rodents, Apodemus sylvaticus (L.), A. flavicollis (Melchior) and Clethrionomys glareolus (Schreber). Mammal Rev. 10, 189-195. OLSZEWSKI, J. L. (1968): Role of uprooted trees in the movements of rodents in forests. Oikos 19, 99-104. SIEGEL, S.; CASTELLAN,N.J. (1988): Nonpara- metric Statistics for the Behavioral Sciences. 2nd ed. New York: McGraw-Hill Intern. Pp. 87-29. TAnton, M.T. (1965): Problems of live-trapping and population estimation for the wood- mouse, Apodemus sylvaticus (L.). J. Anim. Ecol. 34, 1-22. TATTERSALL, F.; WHITEBREAD, S. (1994): A trap- based comparison of the use of arboreal vege- tation by populations of bank vole (Clethrio- nomys glareolus), woodmouse (Apodemus syl- vaticus) and common dormouse (Muscardinus avellenarius). J. Zool. (London) 233, 309-314. TROUT, R. €. (1978): A rewiew of studies on popu- lations of wild harvest mice (Micromys minu- tus (Pallas)). Mammal Rev. 8, 143-158. WARNER, L. J.; BATT, G. T. (1976): Some simple methods for recording wild harvest mouse (Micromys minutus) distribution and activity. J. Zool. (London) 179, 226-229. Authors’ addresses: K. NorpDvic and J. REDDERSEN, National Environ- mental Research Institute, Department of Land- scape Ecology, Grenaavej 12, DK-8410 Ronde, Denmark, (e.mail: JRE@dmu.dk): T.S. JENSEN, Natural History Museum, University Park Build- ing 210, DK-8000 Aarhus C, Denmark. Mamm. biol. 66 (2001) 135-143 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Description, taxonomy, and distribution of Talpa davidiana By B. KRYSTUFEK, FRIEDERIKE SPITZENBERGER, and H. KEFELIOGLU Slovenian Museum of Natural History, Ljubljana, Slovenia; Naturhistorisches Museum Wien, Vienna, Austria; Ondokuz Mayıs University, Samsun, Turkey Receipt of Ms. 27. 10. 2000 Acceptance of Ms. 04. 01. 2001 Abstract The type of Talpa streeti is shown to be cranially very close to and thus probably conspecific with Scaptochirus davidianus, a little known mole described from the border between “Syria and Asia Minor”. The pelvis of 7. davidiana is of caecoidal type, thus disproving any relationship between this species and Scaptochirus moschatus from China. This species is known from two localities in Iran (Kurdistan) and from three regions in Turkey: vicinity of Meydanekbez, Hakkari and Tatvan. Around Tatvan, T. davidiana is sympatric with 7. levantis. Specimens of 7. davidiana from Iran and Hakkari have a complete dental complement of 44 teeth. Due to the absence of lower incisors and the peglike premolars, this number is reduced to 38 in three moles from Tatvan and to 39 (asymmetry) in the type of 5. davidianus. Key words: Scaptochirus davidianus, Talpa streeti, taxonomy, nomenclature, distribution Introduction MILNE-EDWARDS (1884) named and de- scribed Scaptochirus davidianus from a spe- cimen collected in the environs of Akbes at the border between “Syria and Asia Mi- nor”. This taxon remains as one of the least known among mammals of the Western Pa- laearctic fauna. ELLERMAN and MORRISON- Scott (1966) consider it as a synonym of Talpa caeca Saviı, 1822, a statement ac- cepted by DoGramacı (1988, 1989a) and, with reservations, by CorBET (1978). Simi- larly, DEmırsoy (1996) reported on it as T. levantis davidianus, giving Gaziantep as the type locality. In 1965 Lay described a new mole Talpa streeti from Kurdistan province, north-wes- 1616-5047/01/66/03-135 $ 15.00/0. tern Iran. This mole is cranially so well characterised that its specific status was never questioned (CoRBET 1978; CORBET and Hırr 1980, 1986, 1991; Honackt et al. 1982; HUTTERER 1993). Known at first only from its type locality (Lay 1967), T. streeti was later also reported for Turkey (DoGRrA- MACI 1989 a). SPITZENBERGER (in FELTEN et al. 1973) stated that Scaptochirus davidia- nus is a member of the genus Talpa and might be either a species on its own or the oldest name for T. streeti. Assuming that re- ports of T. caeca for Saqgez by MISOoNNE (1959) and for Tatvan by OsBorn (1964) ac- tually refer to T. davidiana this would, to- gether with the type locality of 7: streeti, 136 B. KRYSTUFEK et al. form a plausible distribution area (SPITZEN- BERGER, in FELTEN et al. 1973). GUREEV (1979) and HuTTeErer (1993) gave their opinions about the conspecificity of T. da- vidiana and T! streeti, but did not formally synonymise them. STROGANOV (1948) and GRrULICH (1982) argued, in contrast, that 5. davidianus ıs not a member of the genus Talpa, but is closely related to east Asian 5. moschatus of MILNE-EDWARDS (1867). If this is the case, it would mean that the genus of Scaptochirus is disjunct in distribu- tion, with the east Asian species being sepa- rated from S. davidianus in the Middle East by 5500 km. Further information on the ex- act identities of 7. streeti and S. davidianus is thus of importance also from a Zoogeo- graphical point of view. Therefore, the aim of this study is to elucidate relations within these little-known moles, to define their geographic range and the extent of cranial and dental variation. Material and methods We examined 14 specimens of T. davidiana, in- cluding the types of Scaptochirus davidianus and Talpa streeti. For comparative purposes we also included in the analysis 26 museum specimens of Talpa levantis Thomas, 1906 from Turkey and Iran, three T! romana from Italy and 32 T! stanko- vici from Federal Republic of Yugoslavia and from Macedonia. Specimens were mainly skins and skulls, but in some cases also hip bones. Specimens examined (14). — Iran: Kurdistan, He- zar Darreh (FMNH 96424, type of Talpa streeti; FMNH 96421, 96423, 96425); Kurdistan, 1 mile south of Divandarreh (FMNH 111007). Turkey: Bitlis, Tatvan, Kurtikan (FMNH 82136, 82137); Tatvan (OMU 167); Hakkari, Mergan Zoma in Cilo-Sat-Mts.. (NMW 20326, 20327); Hakkari, Megabuti yaylası (OMU 231); Hakkari, Otluca köyü (OMU 166, 232); Meydanekbez (MNHN 1883-469; type of Scaptochirus davidianus). Material examined from other species. — Talpa le- vantis (26) Iran: Ghilan, 12 km W Chalus (FMNH 96416, 96417, 96418, 96419). Turkey: Trabzon, Co- sandere (BMNH 6.5.1.1, 6.5.1.2, 6.5.1.3, 6.5.1.4); Trabzon, Euthey (BMNH 6.3.6.6); Trabzon, Mer- yemana (BMNH 6.3.6.4; NMNH 327252, 327253); Trabzon, Altindere (BMNH 25.11.1991, the type Talpa caeca levantis); Giresun, Bicik, Ya- vuz-Kemal (NMW 19859); Ordu, Ulubey (NMW 19858); Samsun, Kürtler (PMS 10299); Tamdere, Giresun Dagları, Sehitlor Gegidi (PMS 11372); Bitlis, Tatvan (OMU 233, 234, 236, 237, 238, 239, 240, 241, 242). Talpa romana (3) Italy: Napoli, Roccarainola (PMS 6710, 6711); Quindici, Avelli- no (PMS 6712). Talpa stankovici (32) Macedonia: Mt. Bistra (PMS 2441, 9211, 9532, 7497); Mt. Pe- lister (PMS 9541); Bitola (PMS 7490); Prilep (PMS 7486, 7488, 7521); Resen (PMS 7522, 7505, 7506, 7508); Mt. Sara (PMS 7496); Struga (PMS 7494, 7495); Mt. Galicica (PMS 7492, 7493, 7502). Federal Republic of Yugoslavia: Ulcinj (PMS 3202-3204, 8834-8839, 7500, 7501). Seven linear measurements were taken from each skull with a vernier calliper (accurate to the near- est 0.1 mm). Their abbreviations are: CbL - con- dylobasal length, MxT - maxillary tooth-row length (C - M3), BcB - braincase breadth, BcH -— braincase height (without bullae), RoC - breadth of rostrum over canines, RoM - breadth of rostrum over molars, MdL - length of man- dible. External measurements were deduced from specimen labels: H&B - head and body length, TL - taıl length, HF - hind foot length, W - weight. All measurements are in mm, weight in grams. Overall cranial similarity was assessed by Principal Components Analysis (PCA) of the cor- relation matrix of log transformed measurements. Factor loadings were subjected to Varimax rota- tion. Statistical analyses were performed using STATISTICA analysis system (Release 5.5 ’99). Types of teeth are indicated by letters; capitals in- dicate upper teeth and small letters indicate lower teeth: V/i - incisors, C/c — canines, P/p — premolars, M/m - molars. The number denotes the position of a particular tooth in the tooth row. The following abbreviations were used for collec- tions: FMNH - Field Museum of Natural History, Chicago; MHNP - Museum National d’Histoire Naturelle, Paris; NMW - Naturhistorisches Mu- seum Wien, Vienna; OMU - Zoological collec- tion of the Ondokuz Mayis University, Samsun; BMNH - Natural History Museum London; PMS - Slovenian Museum of Natural History, Ljubljana. Results and discussion Craniometrics T. davidiana is well defined amongst the moles of the Western Palaearctic region by its robust rostrum, a feature best expressed by the breadth across canines (Fig. 1). The Description, taxanomy and distribution of Talpa davidiana type of S. davidianus certainly has nothing in common with the small blind moles of Turkey, traditionally reported on as T. caeca (OsBoRN 1964; DoGramacı 1989 a), but ac- tually representing an independent species, T. levantis (KEFELIOGLU and GENCOGLU 1996). Lay (1965) considered two specimens from Tatvan, Turkey (FMNH 82136 and 82137) to be T. caeca. In spite of their small size (see Tab. 1) and somewhat aberrant dentition, they resemble the type of S. davi- dianus in all other respects. DOGRAMACI (1988) reported Talpa caeca davidianus for Tatvan, however, later on (DoGRAMACI 5,8 D D > EN aRElE 8 m) = 50 DO = Da 8) A ® 46 A 3 A 0o® £ A 9° D 42 88 ® zz [oW:: oXoie) 2 Wonder: 2 3,8 (6) 3,4 27 28 29 30 31 32 33 34 35 Condylobasal length Fig. 1. Plot of rostral breadth over canines against condylobasal length of skull in T. streeti from Iran and Hakkari (diamonds), moles from Tatvan (empty trian- gles), type of 5. davidianus (closed triangle) and 7. le- vantis (circles). Types are indicated by closed symbols. T. levantis from Tatvan is indicated by crosses. 137 1989 a) reported T. caeca for Tatvan and T. streeti for Hakkari, south-eastern Turkey. It is therefore not exactly clear what DoG- RAMACI (1988) understood as T. caeca davi- dianus. As is clear from the available mate- rial (Fig. 1), the Tatvan area is inhabited by two moles: T. levantis and T! davidiana. In the subsequent text, when we refer to Tat- van moles, we have in mind those having the cranial morphology of T. davidiana and not of T. levantis. Moles from the Hakkari region are indistinguishable from the type and topotypes of T. streeti, they also resem- ble the type of S. davidianus, but are of larger size. Projection of specimens onto the first two Principal Components (85.7% of the var- iance explained) clustered groups according to their previous taxonomic assignment and geographic origin (Fig. 2). Principal Com- ponent 1 (PC1) with high character load- ings for CbL, MxT, and MdL, was evidently a factor of general size, which is acommon phenomenon in mammalian morphometrics (LEMEn 1983) — of more interest was the grouping along Principal Component 2 (high loadings for RoC and RoM). There- fore, the close cranıal similarity of $. davi- dianus and T.streeti is beyond doubt. Again, moles from Tatvan were placed within the T. davidiana cluster. Interlocality varlation in size was more strongly ex- Table 1. External and cranial dimensions in 7. davidiana from Turkey and Iran. See list of specimens for their geographic origin. Sex Hu&uBSEnlE SHE, SW. Male Coll. & No. FMNH 96424 FMNH 96421 Male FMNH 96423 Male FMNH 96425 % FMNH 111007 Male FMNH 82136 Female 18 FMNH 82137 Female 18 OMU 166 Male 21 76 OMU 167 Female OMU 231 OMU 232 NMW 20326 NMW 20327 MNHN 1883-469 Female 20 79 Male 18 75 Female Male Male 20 80 17.8261. CbL MxT BcB BcH RoC RoM MdL 138 B. KRYSTUFEK et al. Table 2. Variation in number of teeth in 7. davidiana from Turkey and Iran. First row gives complete dental set in Talpa. (N) - number of specimens examined. See text for explanation of abbreviations. * One specimen with three premolars on the right side. Talpa Iran Hakkari Tatvan Meydanekbes 2 | m] 2 | S 2} SEN SI | DO ODg0 N A ee = | a A [e) | ® QO 2 ® oO E 0) 6) ® ee o [e) (6) (6) ° | SE AOUEN Ss | o® SG AN (6) o El °o [6) | at | [0) ie) ° | -2\ | -3 -2 -1 (6) 1 2 3 Principal Component 1 (75.3%) Fig. 2. Projection of specimens onto the first two principal components (percentage of variance in par- entheses). For symbols see Fig. 1. pressed in T.streeti than in T. levantıs, although the latter originated from a much wider geographical range. Dentition As mentioned by Lay (1965), the type of T. streeti displays the complete dental set of the genus Talpa (i.e. 44 teeth) but this num- ber is reduced to 38 in the three moles from Tatvan (Tab. 2). Reduction affects both the upper and lower premolars, a condition also found in the type of S. davidianus. In the type and topotypes of T. streeti, sub-equal and peglike P2 and P3 are smaller than P1. Moles from Tatvan and the type S. davidia- nus had lost one small premolar; the two remaining premolars between the upper canine and P4 are peglike and of approxi- mately the same size. Between the canini- form pl and large p4, T. streeti has two sub-equal peglike premolars; again, one of them is lost in the moles from Tatvan, and Fig. 3. Variability in the number of lower incisors in T. davidiana. Left - complete set with crowded incisors in a mole from Hakkarı (NMW 20326); right - only two incisors on each side in a mole from Tatvan (FMNH 82137). c - lower canine. Not to same scale. in the type S. davidianus, which result in an increase in the size of the remaining premo- lar. ZIEGLER (1971) stated that in Talpa s.]. the first peglike premolar lost from a row is the anterior one. Consequently, the moles from Tatvan and the type of S. davidianus most likely lack both the p2 and P2. Allthree specimens from Tatvan are unique in having only two lower incisors. The lower incisors are evidently crowded in T! davidi- ana from Hakkari and in the topotypes of T. streeti, a condition which possibly re- sulted in the loss of 13 (Fig. 3). The type spe- cimen S. davidianus has 39 teeth with three incisors in the left mandible and two in the right. Pelvis GRULICH (1971) reported on a mogerid hip bone (with the 5° sacral foramen closed posteriorly by a bony bridge) in $. moscha- tus and presumed such a condition also ex- isted in S. davidianus (GRULICH 1982). Description, taxanomy and distribution of Talpa davidiana However, DoGrAMAcı (1989b) demon- strated that 7. streeti from Hakkari lacks a bony bridge posterior to the 5'" sacral for- amen, having also the 4'® sacral foramen opened posteriorly (caecoidal pelvis). Furthermore, we examined five hip bones from T. davidiana (Iran — 1, Tatvan - 1, Hakkari - 3) and all were of caecoidal mor- phology. Thus, according to the morphology of the pelvis, T. davidiana is close to the genus Talpa from the western Palaearctic, having nothing in common with Scapto- chirus moschatus. Taxonomy MILNE-EDWARDS (1884) based his descrip- tion of S. davidianus on the reduced num- ber of teeth: 3/3, 1/1, 3/3, 3/3 = 40, although a higher number was mentioned in the orig- inal description of 5. moschatus: “Inc. 3/4 can. 1/1 prem. 2/2 mol. 4/4” ı.e. 42 teeth (MILNE-EDWARDS 1867). Fourty teeth have been cited as diagnostic of the genus Scap- tochirus by most subsequent authors (STRO- GANOV 1948; STEIN 1960; GUREEV 1979; GRULICH 1982; NIETHAMMER and KRrAPP 1990). The usefulness of dental formulas in generic diagnostics within the Talpinae, and particularly in the 7alpa group, con- tinues to be a matter of debate. SCHWARZ (1948) considered that the number of gen- era based on dental formulas is grossly ex- aggerated, an opinion shared by ELLERMAN and MorRrIsoNn-Scott (1951) and CorBET (1978). CorBET (1978) also concluded that genera based on dental formulas may cut across other cranial differences. Oligodonties are fairly common within Talpa s.l. (as defined by SCHwARTZ 1948). The fol- lowing teeth are prone to reduction or com- plete loss: i3, p2, p3, P2, and P3 (ZIEGLER 1971). Oligodonties appear to increase in fre- quency across the Western Palaearctic (oli- godonties rare) into the Oriental region (oli- godonties frequent; see ZIEGLER 1971). The reduction or complete loss of i3 is present in Euroscaptor micrura, E.longirostris, and Parascaptor leucura (generic assignments are by HUTTERER 1993). It is evident from STROGANOV (1948) that i3 is also missing in 139 Mogera robusta, GUREEV (1979) however, ex- plained this as a missing canine. An extreme loss of the premolars is seen in Euroscaptor, which lacks all four peglike premolars (1. e. p2, p3, P2, P3; ZIEGLER 1971). Among Talpa s.str. from the Western Pa- laearctic with 44 teeth, i. e., with the conser- vative dental formula of the primitive extant eutherians (ZIEGLER 1971), oligodonties are not as common. In 8184 T. europaea skulls examined by STEIN (1963) only 0.3% lacked at least one peglike premolar in the upper, and 0.6% in the lower jaws. In a smaller sample (N =464) from the Netherlands, NIETHAMMER (1990 a) found 5.6% of the spe- cimens were missing one or more premolars. Loss of premolars was also recorded in T. occidentalis (incidence 3.3%; NIETHAM- MER 1990b), but not in T. caeca (NIETHAM- MER 1990c) and T. stankovici (NIETHAMMER 1990 d; specimens in PMS). A high share of oligodonties was reported ın 7. romana (CA- POLONGO and Panascı 1978): 36.3% of moles lacked between one and four upper premo- lars (N = 255); in three geographic samples, the share of oligodontic moles varied be- tween 14% and 51%. Loss of the two peg- like premolars was common in T! romana, and found in 21.5% of moles on average. ZIEGLER (1971) also recorded a reduction of i3 ın T: romana, in addition to missing of P2. Assuming that electrophoretic diver- gences reflect actual evolutionary relations amongst European moles (FiLippuccı et al. 1987) one would not expect any phyloge- netic background to oligodonties. Moreover, MiLLER (1940) expressed surprise that teeth of such a small size and apparent mechan- ical unimportance as the peglike premolars are so constant in most moles. STEIN (1963) showed that oligodonties in T. europaea are more likely to occur in smaller skulls. In our case, a complete den- tal set was recorded only in the largest moles. Thus, 38 teeth, as observed in moles from Tatvan, are possibly just a case of ex- treme oligodonty and a by-product of size reduction in marginal populations of T. da- vidiana. Although this feature is unique within Talpa s.str., it is supposedly of no taxonomic significance; the asymmetry in 140 B. KRYSTUFEK et al. the number of lower incisors (as seen in the type of S. davidianus) demonstrates that the presence of the small incisor is not stable. In conclusion, the most parsimonious tax- onomy is recognition of a single, although highly variable species T. davidiana, with T. streeti as its junior synonym. Further- more, the caecoidal pelvis suggests their ın- clusion in the genus Talpa not Scaptochirus. The latter is thus a monotypic genus, in dis- tribution restricted to east Asia. Talpa davidiana 1884. Scaptochirus davidianus Milne-Ed- wards, Compt. Rend. Acad. Scıi., Paris, 99: 1141, December 29. Type locality — vicinity of Akbes on the border between Syria and Asıa Mi- Fig. 4. Ventral side of skull in (top left) 7. davidiana (OMU 267), (top right) T. levantis (PMS 10299), (bot- tom left) 7. romana (PMS 6710), and (bottom right) T. stankovici (PMS 2441). nor (“environs d’Akbes, sur les confins de la Syrie et de l’Asie Mi- neure”) = Meydanekbez (also Mey- danı Akbes or Meydan Ikbis), south- west of Gaziantep, Turkey (see SPIT- ZENBERGER in FELTEN et al. 1973). Talpa streeti Lay, Fieldiana Zool., 24: 227, 22 October. Type locality — He- zar Darreh, Kurdistan, Iran. Talpa streetorum Lay, Fieldiana Zool., 54: 131, October 1967. Unjusti- fied emendation of T. streeti (see CorBET 1978). 1965. 1967. Amended diagnosis: A mole with a caecoi- dal pelvis. Rostrum is broader and heavier than in any other Talpa species (T. romana and T. stankovici) with the broadest ros- trum of the European moles (Fig. 4). Ros- trum remains robust also across canines while it is slim in this region in other west Palearctic Talpa. Tail is shorter than in any other west Palaearctic mole (Fig. 5). T. da- vidiana is more inclined towards oligodon- ties than any other species of the genus Tal- pa s.str. Comparison of this species (under Fig. 5. Skins of T. davidiana (left) and T. levantıs (right), both from Tatvan; note the shorter tail in T. davidiana. Description, taxanomy and distribution of Talpa davidiana the name T. streeti) with other moles was provided by Lay (1965). Homonymy: The binomen Scaptochirus da- vidianus was used for the first time by SwInN- HOE (1870) in his list of Chinese mammals. As already pointed out by ELLERMAN and MORRISoN-ScoTT (1951) this is evidently a case of accidental renaming of $5. moschatus. Although SwınHoE (1870 a) used the name S. davidianus, he simultaneousliy refers to the original description of S. moschatus: “(Annales des Sciences Nat. 5e serie, t.7), antea, p.450.” and also credits MILNE- EDwaRDS with authorship. From one of his earlier publications in the same volume of the Proceedings of the Zoological Society (SwINHOE 1870b) it is also clear that SWINHOE was familiar with the description of S. moschatus by MILNE-EDWARDS. The in- troduction of S. davidianus by SWINHOE was simply an error which, however, coincided with the name proposed by MILNE- EDwaArDS (1884) — fourteen years after SWINHOEF’S study was published. If one con- siders 5. davidianus Swinhoe, 1870 and S. da- vidianus Milne-Edwards, 1884 as homo- nyms, then the former is a nomen oblitum as it has never been used as a valid name (INTERNATIONAL COMMISSION OF ZOOLOGICAL NOMENCLATURE 1999, Article 23.9). Distribution: Range is summarised in fig- ure 6. The largest distance between extreme localities is >800 km in a west - east direc- tion and up to 400 km in a north - south di- rection. This range is comparable in size with those of some of the blind Talpa spe- cies with caecoidal pelvis: 7. occidentalis, T. romana, T. stankovici (MITCHELL JONES et al. 1999) and T. caucasica (SOKOLOV and TEMBoToV 1989). All the records are from the southern margin of the Anatolian - Ira- nian high plateau. Specimens were apparently always col- lected at high altitudes (around 2000 m a.s.l.), however, very few facts are avail- able. Habitat is little known; two specimens from Tatvan were collected from burrows in a hayfield, whilst those from Mergan Zoma are from alpine meadows at 2400 m a... (for a photograph of the habitat see SPiIt- ZENBERGER 1976). 141 SSUllE 45°| E Fig. 6. Distributional records of T. davidiana. Turkey: 1 - Meydanekbes; 2 - Tatvan; 3 - Hakkari, Mezralar (1800 m a.s.l.); Megabuti yaylası (3700 m a.s.l.); 4 - Mergan Zoma in Cilo-Sat-Mts. (2400 m a.s.l.); 5 - Hakkari, Otluca köyü (2 000 m a.s.l.); 6 - Yüksekova. Iran: 7 - Lmi south of Divandarreh; 8 - Hezar Darreh. Localities 3 and 5 are based on DoGRAMAcı (1989 a). Meydanekbes seems anomalous from the point of view of both altitude and habitat, and we doubt whether the environs of the town of Meydanekbes are a suitable habitat for the burrowing mole. This specimen might originate from the mountains, either to the north (e.g. Engizek Dagi with the peak of 2814 m a.s.l.; ca. SOkm away) or to the west of Meydanekbes (Bozdag, high- est peak 2240 m; ca. 90 km away). Sympatry: T. davidiana is the southemmost member of its genus, being allopatric throughout the major part of its little known range. The only known incidence of sympatry is from Tatvan. where T. levantis was also collected. Acknowledgements For the access to specimens under their care we thank W.Stantey (FMNH), LiNDA GORDON (NMNH), and PauLA JEnKINS (BMNH). R. HUT- TERER (Bonn) helped with the loan of FMNH spe- cimens to Europe and L. GrAnJon (MNHN) pro- vided photographs of the type of Scaptochirus davidianus. Special thanks go to K. BAUER (Vien- na) for critical comments and help with literature. H. I. GrIFFITHs (Hull) improved the English text and C. MrmAr (Ljubljana) prepared figures 4 and 5. 142 B. KRYSTUFEK et al. Zusammenfassung Beschreibung, Taxonomie und Verbreitung von Talpa davidiana Wegen großer morphologischer Ähnlichkeit der Schädel des Typus von Talpa streeti mit Scaptochirus davidianus wird auf Konspezifität der beiden Taxa geschlossen. Das Becken von Talpa davidiana ist caecoidal, so daß eine Verwandtschaft mit Scaptochirus moschatus aus China ausgeschlossen wer- den kann. 7. davidiana wurde aus dem Grenzgebiet von Syrien und Kleinasien beschrieben und ist bisher von drei Fundorten (Umgebung von Meydanekbez, Hakkari und Tatvan) bekannt. Bei Tatvan lebt die Art sympatrisch mit Talpa levantis. Individuen von T. davidiana aus Iran und Hakkari haben eine Gesamtzahl von 44 Zähnen. Als Folge des Fehlens der unteren Incisivi und der stiftförmiger Prämolaren ist die Zahnzahl bei drei Maulwürfen aus Tatvan auf 38 und beim Typus von 5. davidia- nus auf 39 (Asymmetrie) reduziert. References CAPOLONGO, D.; Panascı, R. (1978): Richerche sulle populazioni di talpe dell’Italia setten- trionale e nuovi dati sulle restanti populazioni Italiane. Annuario Instituto e Museo di Zool- ogia dell’Universita di Napoli 22, 17-59. CoRBET, G. B. (1978): The Mammals of the Pa- laearctic Region: a Taxonomic Review. Lon- don: Brit. Mus. (Nat. Hist.). CoRBET, G. B.; Hırı, J. E. (1980): A World List of Mammalian Species. London: Brit. Mus. (Nat. Hist.). CoRBET, G. B.; Hırı, J. E. (1986): A World List of Mammalian Species. 2” ed. London: Brit. Mus. (Nat. Hist.). CoRBET, G. B.; Hırı, J. E. (1991): A World List of Mammalian Species. 3" ed. Oxford: Oxford University Press. DEMmIrsoy, A. (1996): Türkiye omurgalılan. Tür- kiye omurgali faunasının sistematik ve biyolo- jık özelliklerinin arastirılmsı ve koruma ön- lemlerinin saptanması. Memeliler. Ankara: Meteksan A. S. (In Turkish). DOoGRAMACI, S. (1988): Geographic variation of the Mediterranean mole Talpa caeca, (Mam- malia: Insectivora) in Turkey. Ondokuz Mayıs Universitesi Ziraat Fakültesi Dergisi 3, 207- 222. (In Turkish with English summary). DOoGRAMACı, S. (1989 a): Taxonomy and distribu- tion of moles (genus Talpa: Mammalia) in Turkey. Doga TU Zooloji D. 13, 204-219. (In Turkish with English summary). DOoGRAMACı, S. (1989b): The importance of the pelvis in distinguishing of Turkish Talpa (Mammalia: Insectivori) species. Doga TU Zooloji D. 13, 67-76. (In Turkish with English summary). ELLERMAN, J. R.; MORRISON-SCOTT, T. C. S. (1951): Checklist of Palaearctic and Indian Mammals 1758 to 1946. London: Brit. Mus. (Nat. Hist.). FELTEN, H.:; SPITZENBERGER, F.; STORCH, G. (1973): Zur Kleinsäugerfauna West-Anatoliens. Teil Il. Senckenbergiana biologica 54, 227-2. FıLippuccı, M. G.; NASCETTI, G.; CAPANNA, E.; Burn, L. (1987): Allozyme variation and systematics of European moles of the genus Talpa (Mammalia, Insectivora). J. Mammal- ogy 68, 487-499. GRULICH, I. (1971): Zum Bau des Beckens (Pel- vis), eines systematisch-taxonomischen Merk- males, bei der Unterfamilie Talpinae. Zool. Listy 20, 15-28. GRULICH, I. (1982): Zur Kenntnis der Gattungen Scaptochirus und Parascaptor (Talpini, Mam- malia). Folia Zool. 31, 1-20. GUREEV, A. A. (1979): Fauna SSSR, Mlekopi- tayutschie, tom 4, vyp. 2. Nasekomoyadnye (Mammalia, Insectivora). Leningrad: Nauka. (In Russian). HonAack1, J. H.; KınMAn,K.E.; KoEPPEL, J. W. (1982): Mammal Species of the World. Law- rence: Allen Press Inc. HUTTERER, R. (1993): Order Insectivora. In: Mam- mal Species of the World. A Taxonomic and Geographic Reference. 2”® ed. Ed. by D. E. Wırson and D.-A. M. REEDER. Washing- ton: Smitsonian Institution Press. Pp. 69-130. INTERNATIONAL COMMISSION ON ZOOLOGICAL NO- MENCLATURE (1999): International Code of Zoological Nomenclature. 4" ed. London: The International Trust for Zoological No- menclature. KEFELIOGLU, H.; GENCOGLU, S. (1996): The taxonomy and distribution of Talpa (Mammalia: Insecti- vora) in the Black Sea region. Turkish J. Zool. 20, 57-66. (In Turkish with English summary). Lay, D.M. (1965): A new species of mole (genus Talpa) from Kurdistan province, western Iran. Fieldiana: Zoology 44, 227-230. Description, taxanomy and distribution of Talpa davidiana Lay, D.M. (1967): A study of the mammals of Iran resulting from the Street expedition of 1962-63. Fieldiana: Zoology 54, 1-282. LEMEN, C. L. (1983): The effectiveness of methods of shape analysis. Fieldiana: Zoology 15, 1-17. MILLER, G. S. (1940): Notes on some moles from southeastern Asia. J, Mammalogy 21, 442- 444. MILNE-EDWARDS, A. (1867): Observations sur quelques mammiferes du nord de la Chine. Annales des Sciences Naturelles. Cinquieme serie, Zoologie et Pal&ontologie 7, 375-377. MILNE-EDWARDS, A. (1884): Sur la classification des Taupes de l’ancien continent. Comptes rendus des seances de l’Academie des Sciences 99, 1141-1143. MIsSoNNE, X. (1959): Analyse Zoog&ographique des Mammiferes de I’Iran. Mem. Inst. Roy. Sci. nat. Belgique. Deuxieme serie, Fasc. 59, 1-157. MITCHELL-JONES, A.J.; AMOoRI, G.; BOGDANO- WICZ, W.; KRYSTUFEK, B.; REIUNDERS, P.J.H.: SPITZENBERGER, F.; STUBBE, M.; THISSEN, J. B. M.; VOHRALIK, V.; ZıMA,J. (1999): The Atlas of European Mammals. London: T. and A.D. Poyser. NIETHAMMER, J. (1990a): Talpa europaea Lin- naeus, 1758 — Maulwurf. In: Handbuch der Säugetiere Europas. Vol.3/1. Ed. by J. NIETHAMMER and F.Krapp. Wiesbaden: Aula-Verlag. Pp. 99-133. NIETHAMMER, J. (1990b): Talpa occidentalis Cab- rera, 1907 — Spanischer Maulwurf. In: Hand- buch der Säugetiere Europas. Vol. 3/1. Ed. by J. NIETHAMMER and F.Krappr. Wiesbaden: Aula-Verlag. Pp. 157-161. NIETHAMMER, J. (1990 c): Talpa caeca Savi, 1822 — Blindmaulwurf. In: Handbuch der Säugetiere Europas. Vol. 3/1. Ed. by J. NIETHAMMER and F. Krapp. Wiesbaden: Aula-Verlag. Pp. 145- 156. NIETHAMMER, J. (1990 d): Talpa stankovici V. et E. Martino, 1931 - Balkan-Maulwurf. In: Handbuch der Säugetiere Europas. Vol. 3/1. Ed. by J. NIETHAMMER and F. KrApp. Wiesba- den: Aula-Verlag. Pp. 141-144. 143 NIETHAMMER, J.; KrApp, F. (1990) Handbuch der Säugetiere Europas. Vol. 3/1. Wiesbaden: Aula-Verlag. ÖsSBORN, D. J. (1964): Notes on the moles of Tur- key. J. Mammalogy 45, 127-129. SCHWARZ, E. (1948): Revision of the Old-World moles of the genus Talpa Linnaeus. Proc. Zool. Soc. London 118, 36-48. SOKOLOV, V. E.; TEMBOTOV, A. K. (1989): Mlekopi- tayushchie Kavkaza: Nasekomoyadnye. Moskva: Nauka. (In Russian). SPITZENBERGER, F. (1976): Ein Beitrag zur Kennt- nis der Vogelfauna des Cilo-Sat-Massivs (Vi- layet Hakäri, Türkei). Ann. Naturhist. Mus. Wien 80, 315-323. STEIN, G.H. W. (1960): Schädelallometrien und Systematik bei altweltlichen Maulwürfen (Talpidae). Mitt. Zool. Mus. Berlin 36, 1-48. STEIN, G. H. W. (1963): Anomalien der Zahnzahl und ihre geographische Variabilität bei Insec- tivoren: I. Maulwurf, Talpa europaea L. Mitt. Zool. Mus. Berlin 39, 223-240. STROGANOV, S. U. (1948): Sistematika krotovych (Talpidae). Trudy Zoologicheskogo instituta 8(2). Moskva: Akademija nauk SSSR. (In Russian). SWINHOE, R. (1870 a): Catalogue of the mammals of China (south of the River Yangtsze) and of the island of Formosa. Proc. Scient. Meetings Zool. Soc. London 1870, 615-653. SWINHOE, R. (1870b): Zoological notes of a jour- ney from Canton to Peking and Kalgan. Pro- ceedings of the scientific meetings of the Zo- ological society of London 1870, 427-451. ZIEGLER, A.C. (1971): Dental homologies and possible relationships of recent Talpidae. )J. Mammalogy 32, 50-68. Authors’ addresses: BoRrRIS KRYSTUFEK, Slovenian Museum of Natural History, P.O. Box 290, SI-1001 Ljubljana, Slove- nia (e-mail: boris.krystufek@uni-]j.si); FRIEDERIKE SPITZENBERGER, Naturhistorisches Museum Wien, P.O. Box 417, A-1014 Wien, Austria; HALUK KEFELIOGLU, Ondokuz Mayıs University, 55139 Kurupelit, Samsun, Turkey. Mamm. biol. 66 (2001) 144-154 >. Mammalian Biology © Urban & Fischer Verlag m N http://www.urbanfischer.de/journals/mammbiol © Zeitschrift für Säugetierkunde Original investigation Bats in the Bavarian Alps: species composition and utilization of higher altitudes in summer By JENNNIFER HOLZHAIDER and A. ZAHN Zoologisches Institut, Universität München, München, Germany Receipt of Ms. 10. 03. 2000 Acceptance of Ms. 20. 09. 2000 Abstract Habitat use and species composition of bats in the higher altitudes of the Bavarian Alps were stu- died from May to September 1997. Five hundred buildings at altitudes between 800 and 1800 m above sea level were surveyed; traces of bats or roosting animals were found in 189 of these. Bat occupation of buildings decreased at elevations higher than 1300 m above sea level, and also de- creased with increasing distance of buildings to the surrounding forest. 203 solitary roosting ani- mals and 14 nursery colonies (indicated with*) of the following species were found: Myotis myotis, M. emarginatus, M. mystacinus*, M. brandtı, Pipistrellus pipistrellus*, Pipistrellus nathusii, Eptesicus nılssonii*, Plecotus aurıtus, and Vespertilio murinus. Relative to adjacent lower regions where Myotis myotis is the most abundant species, M. mystacinus and M. brandti together make up an average of 70% of all reliably determined animals at higher altitudes. Within all occurring species, higher re- gions are mainly inhabited by adult males. They might thus avoid competition with nursery colo- nies in the lowlands. Key words: Bats, altitudinal distribution, Bavarian Alps, species composition Introduction The Bavarian Alps are likely to provide good habitats for bats in summer with ex- tensively developed pastures and large areas of natural forests. However, lower temperatures and shorter vegetation peri- ods relative to lowlands (ReısısL and KEL- LER 1987) might be rather disadvantageous for these aerıal insectivores. Studies from neighbouring mountainous countries like Austria and Switzerland show that higher altitudes are in fact inhabited by bats (SPITZENBERGER 1993 a,b: ARLETTAZ et 1616-5047/01/66/03-144 $ 15.00/0. al. 1997; GÜTTINGER 1994). However, cer- tain aspects in the life cycle such as repro- duction seem to be concentrated at lower altitudes (SPITZENBERGER 1993a,b), possi- bly because juvenile growth is favoured by warm conditions in many bat species (TUT- TLE and STEVENSoN 1982; ZAaHn 1999). Also, the composition of species is influenced by altitude (SPITZENBERGER 1993a; ARLETTAZ et al. 1997; GÜTTINGER 1994). This effect may depend both on climatic circumstances and on the availability of suitable roosts. In the “lowland areas” adjacent to the Ba- varıan Alps, the bat fauna is well known (RıcHARZ 1986; RıcHARZ et al. 1989; ZAHN and KrÜGER-BARVELS 1996; ZAHN and MAT- ER 1997). At least 10 of 16 lowland species, which could potentially occur also at higher altitudes, are known to prefer various anthropogenic structures as day roosts (Rı- CHARZ 1986). In this study we therefore fo- cussed on bats roosting in buildings. Build- ings exist in relatively large numbers even at higher altitudes in the Bavarıan Alps and are well suited for a systematic investi- gation of the bat fauna. Moreover, buildings are readily accessible and of better compar- ability than natural roosts such as rock cre- vices or tree holes. Of special interest were the following questions: (1) Which building dwelling species of bats occur over 800 m above sea level in the Bavarian Alps? (ii) Does a species-specific altitudinal distribu- tion exist? (iii) How are roosts in buildings at different altitudes used by bats? (iv) What types of roosts are preferred’? Material and methods The study area ranges from Garmisch-Parten- kirchen (11°06 ©, 47°29 N) in the west to Berch- tesgaden (13°0°O, 47°37 N) in the east and in- cludes altitudes from approximately 800 m to 2963 m above sea level (Zugspitze). About 50% of the area is covered with forest (mostly Fagus sylvatica, mixed with Picea abies, Abies alba and Acer pseudoplatanus). The remaining area con- sists of extensively used mountain pastures and, over about 1800 m above sea level, of unused al- pine meadows, dwarfpines and rocky outcrops. Five hundred buildings (from small wooden mountain cabins to hotel-sized stone houses) at altitudes between 800 m and 1800 m were sur- veyed between May and September 1997. Of these, 451 provided potential roosts for bats. As potential roosts we regarded all kinds of crevices if they were dry, narrow (<5cm wide), dark and without a strong airflow (<15 cm deep). As de- scribed below, these crevices were roughly di- vided into five categories. Shutters, which are fre- quently used by bats (RıcHArRZ 1986) in spite of that fact that they offer less shelter from wind and rain than the other types, were also included. Large accessible attics, which are common roost Bats in the Bavarian Alps 145 sites for bats in the lowlands, were rarely present in the investigated buildings and are therefore not included in the list of potential roosts. Faeces and dead or living animals were taken as evidence for site use. Substantial accumulation of faeces was taken as evidence for colonies. Site use inferred from faeces’ presence is referred to as “indirect proof”, whereas alive animals were taken as “direct proof”. When possible, living ani- mals were caught for species identification. With the exception of Myotis myotis whose droppings reliably could be identified through size and structure, no species identification was attempted using faeces. During the study we did not distin- guish between Pipistrellus pipistrellus and the “new species” Pipistrellus “pygmaeus/mediterra- neus” (HÄUSSLER et al. 2000). For comparison with the lowland bat population, data of the ASK Bavaria (database “Arten- und Biotopschutzkartierung” of the Bavarian State Office of Environmental Protection) were called upon. These data were obtained as well by sys- tematic controls of churches, castles and other buildings that offer suitable attics as by roost con- trols conducted after house owners informed the relevant authorities about bat presence. These roost controls (visits of attics, inspection of cre- vices) were conducted in the same way as in the present study. Results Species composition and abundance of roosting bats At 189 (41.9%) of the 451 potentially suita- ble buildings evidence of site use by bats was obtained. In 100 of these cases, the anı- mals’ presence was indicated by faeces alone. A total of 203 roosting animals (ex- cluding nursery colonies) was found. Of these, 110 were identified to the following nine species: Greater mouse eared bat (Myotis myotis, 4%), Geoffroy’s bat (M. emarginatus, 1%), whiskered bat (M. mystacinus, 50%), Brandt’s bat (M. brandti, 13%), common pipistrelle (Pi- pistrellus pipistrellus, 13%), Nathusius’ pi- pistrelle (Pipistrellus nathusiü, 1%), north- ern bat (Eptesicus nilssoni, 13%), common long eared bat (Plecotus auritus, 5%) and parti-coloured bat (Vespertilio murinus, 2%). With 29 animals, no reliable differen- 146 JENNNIFER HOLZHAIDER and A. ZAHN tiatıon between KM. mystacinus and M. brandti could be made. We assumed that the relative fraction of the two species was the same as that determined reliably, i.e. M. mystacinus about 80%, M. brandti about 20%. In the following, these two species are mostly combined and referred to as M. my./ br. 64 anımals remained unidentified but certainly belonged to smaller species (i.e. larger species like M. myotis, Eptesicus spec., V. murinus or Plecotus spec. could be excluded). Species dwelling in buildings which occur in adjacent areas but were not found in buildings at higher altitudes in- cluded barbastelle (Barbastella barbastel- lus), noctule bat (Nyctalus noctula), sero- tine bat (Eptesicus serotinus) and lesser horseshoe bat (Rhinolophus hipposideros). Altitudinal distribution Bats preferred roosts in buildings at lower altitudes. Figure 1 shows that the percen- tage of buildings with traces of bats - either alive anımals or faeces — decreased signifi- cantly at altitudes higher than 1300 m above sea level (X” = 26.6, p < 0.01). Species composition did not change signifi- cantly with altitude. At all altitudes, M. my./br. were by far the most abundant 60 50 40 30 20 10 Percentage of occupied buildings [%)] <1000 Mindirect proof 1000-1099 1100-1199 1200-1299 1300-1399 1400-1499 species (Fig. 2). Together they made up 70.5% of all reliably identified individuals. Assuming a relation of 20% M. brandti in all M. my./br. individuals, also Brandt’s bat was more abundant than any other species except the whiskered bat. It is also apparent that, within the studied range of altitude, no species showed any preference for particu- lar elevations. The only exception was M. myotis, which was found exclusively be- low 1200 m. However, with M. brandti and E. nilsonii two species were present which are only rarely found in the adjacent low- land areas. The highest site where roosting bats were found was at 1670 m in a group of cabins near Königssee (Berchtesgaden). In five different cabins, a total of 10 animals, in- cluding M. mystacinus, Pl. auritus and V. murinus was found. A total of 14 nursery colonies was found. Nine colonies were indicated through faeces and could thus not be reliably identi- fied on species level. However, concerning pellet size they most likely belonged to Pi- pistrellus spec. or M. my./br. The other co- lonies consisted of M. mystacinus (n=2), P. pipistrellus (n=2) and E.nilssonii (n = 1). While solitary animals reached alti- tudes up to 1670 m, the highest nursery col- ony was recognized at 1400 m. Ten of the E [direct proof l >1500 Altitude[m] Fig. 1. Percentage of occupied buildings at different altitudes. N = number of investigated suitable buildings. MM. mystacinus ME. nilssonii EM. my./br. Indet. Pipistrellus spec. Bats in the Bavarian Alps 147 EM. brandti OPIl. auritus < 1000 1000-1099 1100-1199 1200-1299 1300-1499 >1500 je Altitude Fig. 2. Relative abundance and altitudinal distribution of reliably identified species. N = number of bats. The al- titudes 1300 m-1499 m are drawn together because of only two direct observations between 1300 m and 1399 m. Table 1: Altitudinal distribution of nursery colonies E. nılssonii <1000 m 1000 m-1099 m 1100 m-1199 m 1200 m-1299 m 1300 m-1399 m Total fourteen colonies occurred below 1100 m (Tab. 1). Age structure and composition of sexes In all species, the great majority of solitary animals consisted of adult males. Only 16 fe- males (14.5%) were found outside nursery colonies, the first one on July 23rd. They be- longed to M. mystacinus (n = 9), M. brandti (n = 2), M. myst./brandti indet. (3), E. nilsso- nii (n= 1) and P pipistrellus (n=1). Three mating communities of M. mystaci- P. pipistrellus indicated through Total faeces nus were found, one consisting of three ani- mals (2 females, one male), the others of one male and one female each. Two sub- adults of M. brandti (one male, one female) were found outside colony sites at the be- ginning of September. Preference of roosts and surrounding The occupation of each type of roost by bats deviates significantly from the distribution of existing roosting possibilities (X”-Test, 148 JENNNIFER HOLZHAIDER and A. ZAHN SQ oO Distribution of recordings and potential roosts [%] & p oO oO ee] [e)) oO O1 [@) DD oO —_ oO B2] ER Br Roof oO Roof bar > | other crevices E eu PSEREENE ODistribution of recordings f ak ee M Distribution of potential roosts h Shutters Shingles Fig. 3. Distribution of recordings compared to the distribution of potential roosts. As the number of “other cre- vices” can hardly be estimated, no value is given for this category. X?=17.7, p<0.01, Fig.3), which can be roughly divided into the following cate- gories. The most frequently used type of building roost (174 of 268 recordings, includ- ing direct and indirect evidence) was the crevice behind the board at the ridge of the roof, which was present at almost every sur- veyed building. It was followed in prefer- ence by crevices in the roof itself between tin roof covering and underlying roof beams (used by 47 anımals) which are characteris- tic especially for smaller, traditionally built wooden cabins. In many buildings also spaces behind shingles (7 records), open shutters (7 records) and other crevice-like structures (34 records) were used by bats. Figure 4 shows a typical mountain cabin with the most common types of roosting possibilities. Also, bats significantly fa- voured buildings closer to forests. While 45.6% of the roost sites in buildings less than 100 m away from a forest edge were occupied by bats (n=375), only 24% (n = 76) of all suitable buildings farther than 100 m away were used (X” = 15.4, p < 0.01). Concerning the compass bearings of the roosts, no absolute preference for roosts, openings towards one distinct orientation could be shown. However, when only roosts at cabins built with one distinct ori- entation (1. e. for example all buildings fac- ing in North/South-direction) were consid- ered, southern and south-western exposed roosts were clearly preferred to northern and north-eastern ones (X° =4.3, p< 0.05). Influence of seasons Significantly more buildings were occupied by bats as the summer progressed. The per- centage of buildings with evidence of bats rose from 7.6% (n=13) in May to 40.1% (n = 332) in June/July, and 51.8% (n = 106) in August/September (X? = 0.42, p< 0.05). With the exception of E. nilssonii, which was found only after the end of July, the oc- currence of all species was evenly distribu- ted over the summer. 149 Bats in the Bavarian Alps Ill | II) Fig. 4. Typical Bavarian mountain cabin with various roosting possibilities. rb=crevice behind ridge board, rc =roof crevice (i.e. crevices between tin covering and underlying roof bars), sh = crevice behind shutter, oc = other crevices. Comparison between higher altitudes and lowlands Species composition at higher altıtudes de- monstrates some striking differences rela- tive to the conditions in the contiguous low- land areas. In the latter, the following species were found by RıcHARrZ (1986), Rı- CHARZ et. al (1989), ZAHN and KrÜGER-BAR- vELS (1996) and ZAHn and MAIER (1997) (species where nursery colonies are known are marked with*): Lesser horseshoe bat (Rhinolophus hipposideros)*, greater mouse eared bat*, Bechstein’s bat (M. bech- steini), Daubenton’s bat (M. daubentoni)*, Brandt’s bat*, whiskered bat*, Geoffroy’s bat*, Natterer’s bat (M. nattereri)*, com- mon long eared bat*, barbastelle*, serotine bat, noctule bat (Nyctalus noctula), com- mon pipistrelle*, Nathusius’ bat and parti- coloured bat. Seven of these species were absent at buildings in the study area, viz. R. hipposi- deros, B. barbastellus, M. nattereri, M. dau- bentoni, M. bechsteini, N. noctula and E. serotinus. E. nilssonii was found at mon- tane sites but was virtually absent in the lowland roosts. Apart from these absent species, a compari- son between the database ASK and the pre- sent study shows that a complete shift in the relative dominance of the different species is apparent between the lowlands and high- er altıtudes. In lower regions, Myotis myotis dominates species’ abundance with about 62% of all solitary animals (n = 277). A total of 27 nur- sery colonies is known in the adjacent rural districts. Also Plecotus auritus occurs quite often (17%). On the other hand, at higher altitudes there is a strong dominance of the two “moustached” bat species, Brandt’s bat and whiskered bat. At higher elevations, such roost sites as large church attics were not found. When species that normally roost in these sites in the lowlands are excluded (M. myotis, Pl. auritus, R. hipposideros), the moustached bats represent 54% of the solitary animals in the mountainous study areas, but only 22% in the lowlands (Fig. 5). P. pipistrellus, which reached a frequency of only 14% at higher altitudes, made up 28% of all species in the lowlands. Also M. emarginatus and V. murinus are more abundant in lower regions. Discussion Roosts In general, higher altitudes in the Alps are inhabited by bats, although there are some restrictions. 150 JENNNIFER HOLZHAIDER and A. ZAHN Percentage of species [%)] M> 800 m [L1< 800 m Fig. 5. Comparison of the species composition in solitary roosting bats at high (> 800 m, n=101) and low (<800 m, n=50) altitudes. M. myotis, Pl. auritus and R. hipposideros, which probably find only few suitable buildings in higher regions, are excluded. P. p. = Pipistrellus pipistrellus; P.n. = P. nathusii, M.d. = Myotis dauben- toni, M.n.=M. nattereri, M.e.=M. emarginatus, B.b.= Barbastella barbastellus, V.m.= Vespertilio murinus, E.n. = Eptesicus nilsonii, N. n. = Nyctalus noctula. The number of occupied roosts in buildings decreases significantly over 1300 m above sea level. This is probably due to adverse climatic conditions at higher elevations. The average temperature for July, for example, is 17°C at 600 m but only 13°C at 1200 m (REısısL and KELLER 1987). Furthermore, buildings at these altitudes are often at sites exposed to wind, precipitation and cold temperatures, factors which might thus di- minish their suitability as roosts. The impor- tance of warmer roosts is also supported by the fact that, given the opportunity, bats tend to choose southern and south-western roosting possibilities at the buildings. A re- duced exposure to wind could also explain the observed preference for houses in or near forests. Additionally, forests offer a sheltered flight path between roost and foraging area. Many bat species avoid cross- ing open land without structures such as hedges or tree lines (LimpEns and KAPTEYN 1991). This may additionally lower the va- lue as bat roosts of houses located far from forests. The special climatic conditions in the Alps may also be the reason for the low number of bats roosting behind shutters, which are often used by whiskered bats, pipistrelles and barbastelles in other areas (SPITZENBER- GER 1993 a,b). Probably, their exposure to all sides affects their suitability as a roost in the adverse weather conditions of higher altitudes stronger than it does in lower re- gions. However, the strong preference for crevices behind roof bars that was observed in this study may partially be due to a meth- odological artefact, since this roost type is very easy to investigate. Other potential bat roosts, such as crevices in the roof itself, which offer warm shelters too, are probably occupied more often than is apparent, since bats may roost very hidden at these sites and are easily overlooked. Species composition Some of the species absent in higher areas, such as Natterer’s bat, Bechstein’s bat and Daubenton’s bat, are known to roost prefer- entially in natural structures like tree holes (RICHARZ 1986), so it is not amazing that they were not found in the surveyed build- ings. However, mistnetting at several caves in the Bavarian Alps shows that these spe- cies actually do occur at higher altitudes also in summer: M. bechsteini was caught several times at a cave near Frasdorf (1200 m) (MESCHEDE pers. comm.); M. daubentoni, occurred atacave near Kochel in great num- bers and could regularly be seen foraging at ponds up to 1100. m; M. nattereri even oc- curred at a cave 1800 m high in August and September 1997. Nyctalus noctula, which tends to roost in trees and buildings, was neither found in buildings in the study area nor was it ever seen hunting or identified acoustically in occasional bat-detector surveys. The ab- sence of roosting noctules in study buildings might be due to a lack of suitable roosts, as these bats prefer large crevices in high buildings (Zann et al. 1999). Barbastella barbastellus, on the other hand, is known to live at higher altitudes in Swit- zerland and Austria, and to preferably use roosts of the type that mainly occur in the surveyed buildings: crevices behind ridge- bars and open shutters (SPITZENBERGER 1993 b). It is thus a species almost predes- tined to occur in buildings in the study area. Several animals caught while mistnetting at caves near Kochel, Frasdorf, and Bichlersee in 1997 show that barbastelles actually do occur in the Bavarian Alps (MESCHEDE and RuDoLpH pers. comm.). Nevertheless, whereas it is not very numerous but widely spread in Austria (SPITZENBERGER 1993b), it is one of the rarest species in Germany with only five recordings of nursery colo- nies in southern Bavaria and very few re- cordings of solitary animals (RUDOLPH et al. 2001). We might have failed to find their roosts in buildings simply due to the low population density of this species. It is also possible that they use natural roosts more frequently than is now realised, since radio- tracked individuals favoured roosts behind patches of loose tree bark (STEINHAUSER 2001). Bats in the Bavarian Alps 151 M. myotis, R. hipposideros, and P. auritus, which were rare or absent in the higher re- gions, depend strongly on large attics for roosting which are much more abundant in lower regions. Moreover, M. myotis, which was the dominant species in the lowlands, is known to be strongly thermophilous (SPITZENBERGER 1988; GÜTTINGER 1994; Ru- DOLPH and LıeGr 1990). In Switzerland and Austria, their complete live cycle is concen- trated in relatively low areas. Solitary ani- mals range up to an average of 53lm in Austria, whereas nursery colonies are found only up to 439 m (SPITZENBERGER 1988). Also most colonies in Switzerland occur lower than 600 m (HAFFNER and MOoESCH- LER 1995). More or less the same is true for the lesser horseshoe bat, which mainly oc- curs between 600 m and 900 m in Austria (SPITZENBERGER 1995). It ıs also one of the rarest bat species in Bavarıa with only two known nursery colonies (ZAHN and ScHLAPP 1997). Thus, they could not be ex- pected to occur in the studied buildings. Particularly solitary animals of Plecotus auritus are also known to use tree holes and bird- or bat-boxes as roosts, which were not investigated in this study. They might there- fore be more numerous in the Alps than ıs apparent from this investigation. The whiskered bat is the only species of thıs study for which every aspect of the life cycle (breeding, mating, hunting, hibernating) was observed to take place at higher altı- tudes. The distribution of whiskered bat and Brandt’s bat in other countries show that they are indeed sturdy species in adverse cli- mates: in Scandinavia they reach 64° of lati- tude (SCHOBER and GRIMMBERGER 1987), and in Switzerland nursery colonies of M. mysta- cinus occur up to 1670 m (ZınGG and Bur- KHARD 1995). In a recent investigation of the National Park Hohe Tauern in Austria M. my./br. also make up almost two-thirds of the species composition at higher altitudes (HÜTTMEIR and REITER 1999). P. pipistrellus, on the other hand, has quite similar habitat demands and roost preferences but reaches only 61° of latitude in Scandinavia (SCHOBER and GRIMMBERGER 1987) and seems to prefer the lower altitudes. In Canton Wallis (Swit- 152 JENNNIFER HOLZHAIDER and A. ZAHN zerland), the number of roosts decreases continually over 400 m above sea level (Ar- LETTAZ et al. 1997). It seems that the great flexibility especially of M. mystacinus (TAAKE 1984), its cold- hardiness, and its preference for crevice- like roosts which are so common in the Alps, lead to the strong dominance of this species in the study area. It is still not quite clear whether a true pre- ference for higher altitudes exists for M. brandti, although the virtual absence of this species in lowland areas and multiple records in the study area above 800 m en- courage this suggestion. The low population density in the lowlands could partially be an artefact due to the difficulties in differen- tiating between M.mpystacinus and M. brandti, especially concerning the fe- males. It might well be that some M. brandti colonies in the lowlands have been mista- ken for M. mystacinus. Still, its abundant occurrence at higher altitudes, especially of males, is apparent. Results from Canton Wallis in Switzerland, where M. brandti was found exclusively over 1200 m above sea level (ArLETTAZ et al. 1997), also sup- port that M. brandti ıs indeed well adapted to the conditions in the mountains. Also the northern bat seems to prefer the higher altitudes. While 14 solitary animals plus one nursery colony were found at the higher altitudes, neither solitary roosting animals nor nursery colonies are known from adjacent areas in the last 10 years. Also in Switzerland, the majority of E. nils- sonii was found between 1200m and 2000 m above sea level (ARLETTAZ et al. 1997): However, SKIBA (1995), using a bat detec- tor, recorded up to 80 animals per night in some regions below 800 m in the study area. It is possible that roosting and hunting habi- tats differ for this species, so that foraging of high elevation roosting animals takes place in lower regions. However, hunting activity in the higher regions of the study area was never investigated systematically and there is no apparent reason why this species should not use higher altitudes for foraging as well. Male dominance One of the most apparent features of this study is the strong dominance of solitary males in all occurring species. Whereas in adjacent lower regions almost one-third of all bat roosts recorded in the database ASK are nursery colonies, such colonies only pro- vide 7.4% of all animals in the study area. The reason for this is not a lack of suitable roosts. Especially for two frequently found species, M. mystacinus and P. pipistrellus, most nursery colonies in lower altitudes are known from sites also most abundant in this study: crevices behind roof bars and open shutters. One explanation for the high per- centage of colonies in the lowlands may be the sampling method: While in the moun- tains all buildings were controlled systemati- cally, many lowland collections of data were made after owners had informed bat conser- vationists of bat presence in their houses. Since the presence of a colony may be more obvious than that of a solitary bat, the num- bers of males may be underrepresented in these data. However, the dominance of males in the Alps could be also due to cli- matic factors. Females in nursery colonies depend on relatively high temperatures dur- ing gestation and lactation for optimising foetal growth (AupeEr 1992; RacEy 1969; TUTTLE and STEvVENSON 1982; ZAHN 1999). This may underlie the fact that nursery colo- nies are found more often in lower and therefore warmer regions. Males, which have a lower energy demand and should thus be able to live in harsher conditions (BARCLAY 1991), are able to utilise higher altitudes as day roosts. They may also avoid foraging competition with nursery colonies (Kunz 1974) in the lowland areas by evading into the Alps. Such avoidance of food competi- tion has been inferred for other bat species (Kunz 1974). Moreover, males may even save energy by falling into torpor more often due to lower air temperatures (BARCLAY 1991). However, comparative behavioural and physiological studies of both males and females settling at different altitudes are needed to verify these possible reasons for the prevalence of male bats in the Alps. Climatic chance and altitudinal distribution Climatic conditions appear to be a key fac- tor influencing species composition and po- pulation structure of bats in the Alps. A long term monitoring of populations at dif- ferent altitudes and a comparison of the nursery colonies in respect of roost selec- tion, roost climate, timing of reproduction, growth and mortality of juveniles would of- fer the opportunity to increase our knowl- edge concerning the influence of climate on the population biology of bats. This could allow predictions about possible reac- Zusammenfassung Bats in the Bavarian Alps 153 tions of bat populations to the current de- viations in climatic patterns which may be altering the present scenario of altitudinal distribution. Acknowledgements We thank PD Dr. ROLAND MELZER for critical comments on the manuscript and valuable help with the drawing of the cabin. Thanks to Prof. Ro- BERT DUDLEY and Dr. Les WILLIAMS for improv- ing the English. We are indebted to the Bavarian State Office for Environmental Protection for the permission to analyse the data of the ASK. Fledermäuse in den Bayerischen Alpen: Artenspektrum und Nutzung von höheren Lagen im Sommer Von Mai bis September 1997 wurde das Artenspektrum von Fledermäusen sowie ihre Habitatnut- zung in den höheren Lagen der Bayerischen Alpen untersucht. Von 500 kontrollierten Gebäuden zwischen 800 und 1 800 m über Seehöhe wiesen 189 Spuren von Fledermäusen oder lebendige Tiere auf. Der Anteil an besetzten Gebäuden nahm oberhalb von 1300 m und mit wachsender Distanz zum Wald signifkant ab. Es wurden 203 Einzeltiere und 14 Wochenstuben (mit * gekennzeichnet) folgender Arten nachgewiesen: Myotis myotis, M. emarginatus, M. mystacınus*, M. brandti, Pipr- strellus pipistrellus*, Pıipistrellus nathusii, Eptesicus nilssonii*, Plecotus auritus, Vespertilio murinus. Im Gegensatz zum angrenzenden Flachland, wo Myotis myotis die häufigste Art ist, stellen in höhe- ren Lagen Bartfledermäuse (M. mytacinus und M. brandtı) etwa 70% aller sicher bestimmten Tiere. Bei allen Arten werden die höheren Lagen in erster Linie von adulten Männchen bewohnt. Diese vermeiden dadurch möglicherweise Konkurrenz mit Wochenstubentieren im Tal. References ARLETTAZ, R.; LUGON, A.; SIERRO, A.; DESFAYES, M. (1997): Les chauves-souris du Valais (Suisse): statut, zoog&ographie et Ecologie. Le Rhino- lophe 12, 1-59. AupET, D. (1992): Roost quality, foraging and young production in the mouse-eared bat Myo- tis myotis: a test of the ESS model of group size selection. Diss. thesis, York University, Canada. BARCLAy, R.M.R. (1991): Population structure of temperate zone insectivorous bats in relation to foraging behaviour and energy demand. ]. Animal Ecol. 60, 165-178. GÜTTINGER, R. (1994): Ist in Mitteleuropa das Kli- ma der primär begrenzende Faktor für das Vorkommen von Fortpflanzungskolonien des Großen Mausohrs (Myotis myotis)? Berichte St. Gallischen Naturw. Ges. 87, 87-22. HÄUSSLER, U.; NAGEL, A.; BRAUN, M.; ARNOLD, A. (2000): External characters describing sibling species of European pipistrelles, Pipistrellus pipistrellus (Schreber, 1774) and P. pygmaeus (Leach, 1825). Myotis 37, 27-40. HAFFNER M.; MOESCHLER, P. (1995): Myotis myo- tis. In: Säugetiere der Schweiz-Verbreitung, Biologie, Ökologie. Ed. by Denkschriften- kommission der Schweizerischen Akademie der Naturwissenschaften. Basel, Boston, Ber- lin: Birkhäuserverlag 103, 124-126. HÜTTMEIR, U.; REITER, G. (1999): Vorkommen und Gefährdung gebäudebewohnender Fle- dermäuse (Chiroptera: Rhinolophidae, Ves- pertiliondae) im Salzburger Anteil des Nationalparks Hohe Tauern und in den Nationalparkgemeinden des Pinzgaues. 154 JENNNIFER HOLZHAIDER and A. ZAHN Wiss. Mitt. Nationalpark Hohe Tauern 5, 161-184. Kunz, T.H. (1974): Feeding ecology of a tempe- rate insectivorous bat (Myotis velifer). Ecol- ogy 55, 693-711. LimpEns, H.; KAPTEYN, K. (1991): Bats, their beha- viour and linear landscape elements. Myotis 29, 39-48. RaAcEy, P. A. (1969): Diagnosis of pregnancy and experimental extension of gestation in the pi- pistrelle bat, Pipistrellus pipistrellus. J. Re- prod. Fert. 19, 465-474. ReısıGL, H.: KELLER, R. (1987): Alpenpflanzen im Lebensraum. Stuttgart: Gustav Fischer. RICHARZ, K. (1986): Bedrohung und Schutz der Gebäudefledermäuse. Schriftenreihe Bayer- isches Landesamt für Umweltschutz 73, 15-43, RICHARZ, K.; KruLı, D; ScHhumm, A. (1989): Quar- tieransprüche und Quartierverhalten einer mit- teleuropäischen Wochenstubenkolonie von Myotis emarginatus (Geoffroy, 1806) im Rosen- heimer Becken, Oberbayern, mit Hinweisen zu den derzeit bekannten Wochenstubenquartie- ren dieser Art in der BRD. Myotis 27, 111-130. RUDOLPH, B. U.; LiesL, A. (1990): Sommerverbrei- tung und Siedlungsdichte des Mausohrs Myotis myotis in Nordbayern. Myotis 28, 19-38. RuUDoLPH, B. U.; HAMMER, M.; ZAHN, A. (2001): Die Mopsfledermaus Barbastella barbastellus in Bayern. In: Tagungsband „Zur Situation der Mopsfledermaus in Europa“. Ed. by Ar- beitskreis Fledermäuse Sachsen-Anhalt e.V.. Berlin: IFA-Verlag. (in press) SCHOBER, W.:; GRIMMBERGER, E. (1987): Kosmos Na- turführer: Die Fledermäuse Europas: kennen — bestimmen - schützen. Stuttgart: Kosmos. SKIBA, R. (1995): Zum Vorkommen der Nordfle- dermaus, Eptesicus nilssonii (Keyserling und Blasius, 1839), in Süddeutschland. Nyctalus 5, 593-601. SPITZENBERGER, F. (1988): Großes und Kleines Mausohr, Myotis myotis Borkhausen, 1797, und Myotis blythi Tomes, 1857 (Mammalıa, Chiroptera) in Österreich. Mammalia austria- ca 15. Mitt. Abt. Zool. Landesmus. Joanneum 42, 65-109. SPITZENBERGER, F. (1993 a): Angaben zu Sommer- verbreitung, Bestandsgrößen und Siedlungs- dichten einiger gebäudebewohnender Fleder- mausarten Kärntens. Myotis 31, 69 -109. SPITZENBERGER, F. (1993b): Die Mopsfledermaus (Barbastella barbastella Schreber 1774) in Österreich. Mammalia austriaca 20. Myotis 31, 111-153. SPITZENBERGER, F. (1995): Die Säugetiere Kärn- tens. Teil 1. Carinthia II 185/105, 247-352. SPITZENBERGER, F. (1997): Verbreitung und Be- standsentwicklung der Kleinen Hufeisennase (Rhinolophus hipposideros) in Österreich. Ta- gungsband „Zur Situation der Hufeisennasen in Europa“. Ed. by Arbeitskreis Fledermäuse Sachsen-Anhalt e. V.. Berlin: IFA-Verlag. Pp. 135-142. STEINHAUSER, D. (2001): Untersuchungen zur Ökologie der Mopsfledermaus, Barbastella barbastellus (Schreber 1774), und der Bech- steinfledermaus, Myotis bechsteini (Kuhl 1818). Schriftenreihe für Landschaftspflege und Naturschutz 66 (2). TAAKE, K-H. (1984): Strukturelle Unterschiede zwischen Sommerhabitaten von Kleiner und Großer Bartfledermaus (Myotis mystacinus und M. Brandti) in Westfalen. Nyctalus 2, 16-32. TUTTLE, M. D.; Stevenson, D. (1982). Growth and survival of bats. In: Ecology of bats. Ed. by T.H. Kunz. New York: Plenum Press. Pp. 105-150. ZAHN, A. (1999): Reproductive success, colony size and roost temperature in attic-dwelling Myotis myotis. J. Zool. (London). 247, 275-280. ZAHN, A.; CHRISTOPH, C.; CHRISTOPH, L.; KRED- LER, M.; REITMEIER, A.; REITMEIER, F.; SCHA- CHENMEIER, C.; SCHOTT, T. (1999): Die Nut- zung von Spaltenquartieren an Gebäuden durch Abendsegler (Nyctalus noctula) in Südostbayern. Myotis 37, 61-76. ZAHN, A.; KRÜGER-BARVELS, K. (1996): Wälder als Jagdhabitate von Fledermäusen. Z. Ökol. Naturschutz 5, 77-85. ZAHN, A.; MAIER, S. (1997): Jagdaktivität von Fle- dermäusen an Bächen und Teichen. Z. Säuge- tierkunde 62, 1-11. ZAHN, A.; SCHLAPP, G. (1997): Die Bestandsent- wicklung und aktuelle Situation der Kleinen Hufeisennase (Rhinolophus hipposideros) in Bayern. Tagungsband „Zur Situation der Huf- eisennasen in Europa“. Ed. by Arbeitskreis Fledermäuse Sachsen-Anhalt e.V. Berlin: IFA -Verlag. ZınGg, P. E.; BURKHARD, W.-D. (1995): Myotis my- stacinus. In: Säugetiere der Schweiz — Verbrei- tung, Biologie, Ökologie. Ed. by Denk- schriftenkommission der Schweizerischen Akademie der Naturwissenschaften, Basel, Boston, Berlin: Birkhäuserverlag 103, 104-107. Author’ addresses: Jennifer Holzhaider, Zoologisches Institut, Lu- isenstraße 14, D-80333 München, e-mail: holzhai@zi.biologie.uni-muenchen.de and Dr. Andreas Zahn, Zoologisches Institut, Luisen- straße 14, D-80333 München Mamm. biol. 66 (2001) 155-164 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation The rise of urban fox populations in Switzerland By SANDRA GLOOR, F. BONTADINA, D. HEGGLIN, P. DEPLAZES, and U. BREITENMOSER Zoological Museum and Institute of Parasitology, University of Zürich, and Urban Ecology and Wildlife Research, Zürich, and Swiss Rabies Centre, Institute of Veterinary Virology, University of Bern, Bern, Switzerland Receipt of Ms. 12. 04. 2000 Acceptance of Ms. 15. 10. 2000 Abstract Since 1985 increasingly more foxes have been recorded from cities in Switzerland. The inquiry of town officials showed that foxes are observed in 28 out of the 30 largest Swiss cities today and breeding dens are known in 20 of these cities. Urban foxes are observed more often than one would expect in larger cities than in smaller towns. In Zürich, the largest city in Switzerland, urban foxes were very scarce until the early 19805. According to the hunting statistics, from 1985 onwards, there was a drastic increase in the urban fox population. In the adjacent rural areas, there was also a clear but less extreme increase in the fox population from 1984 onwards due to successful vaccination campaigns against rabies. As an explanation for the presence of foxes in human settlements we sug- gest two alternative hypotheses, which focus either on the population pressure in the rural areas or on the behavioural adaptations of urban foxes. The presence of foxes in urban areas influences behaviour and attitudes of people towards urban wildlife and it has a consequences for the manage- ment of foxes and the treatment of zoonoses such as rabies and the alveolar echinococcosis. Key words: Vulpes vulpes, urban habitat, invasion, adaptation Introduction Since 1985 fox populations have experi- enced a drastic increase in Switzerland (BREITENMOSER et al. 2000). Apart from this development in rural areas, increasingly more foxes have been recorded from large Swiss conurbations and cities such as Zu- rich and Geneva. Game wardens and wild- life biologists have observed foxes in urban areas; people having noticed foxes in their gardens turned to local officials for infor- mation; pictures and articles about foxes in the middle of residential areas have been 1616-5047/01/66/03-155 $ 15.00/0. published. Are these records just occasional observations or do they indicate the coloni- sation of a new habitat by the red fox? Red foxes living in urban areas are known from Great Britain where urban foxes have been observed ın cities such as London since the 1930s (TEAGLE 1967, BEAMES 196951972 ZPAGEZ 95) Inziche21970s and 1980s, fox populations in British cities reached densities of up to five fox family groups per km” (representing 12 adults on average), densities which had never been 156 SANDRA GLOOR et al. observed so far (HArrıs 1981 a; HARRIS and RAYNER 1986a). Similar fox population densities were nowhere recorded in urban areas outside of Great Britain, either on the European continent or in other parts of the distribution areas of the red fox. There- fore, urban foxes were thought to be a Brit- ish phenomenon (Harrıs 1977; MAcDOo- NALD and NEWDICK 1982). In the 1970s and 1980s, the fox population on the European continent experienced a heavy rabies epizootic, which reached Swit- zerland in 1967 (STEcK et al. 1980; MÜLLER et al. 2000). Fox densities decreased drasti- cally, and, as seen from the Swiss hunting record, reached a low in 1984 (BREITENMO- SER et al. 2000). After the success of oral vaccination campaigns against rabies, started in Switzerland in 1978 (WANDELER et al. 1988), the fox population recovered again from 1985 onwards (KAPPELER 1991; BREITENMOSER et al. 2000). At that same time, foxes were increasingly observed in human settlements. Our objectives in this study are to investi- gate the present sıtuation in large Swiss set- tlements, to evaluate the recent develop- ment of the fox population in Zurich, the largest conurbation of Switzerland, and to compare it with the trend in surrounding rural areas. Material and methods Study area Switzerland is a diverse and mountainous country. 24% of its total area of 40 000 km? (excluding lakes), are above 2000 meters in elevation where fox population density is low. The remaining 76% of the country forms heterogeneous and mostly good quality habitat for the red fox. In Switzerland there are 30 cities with more than 20.000 inhabitants, where 19% of the 6.9 million inhabitants live. The largest conurbation of Swit- zerland is the area of Zurich with some 1000 000 inhabitants. However, only 352200 of them live in the actual “city”, the political community of Zurich. The political community of Zurich (92 km?) — which we refer to when we are talking about the “city of Zurich” in the following report — consists of 53% urban area, 24% forest, 17% agricultural areas and 6% water (FEDERAL OFFICE OF STATISTICS 1998). Forest and agricultural areas surround the urban area and are referred to as the rural area of the city in the following. As far as hunting is concerned, the city of Zurich is orga- nised as a game sanctuary. The city of Zurich be- longs to the canton of Zurich, one of the most densely populated cantons of Switzerland (area 1661 km’, 683 inhabitants per km’). The present distribution of urban foxes in Switzerland During a television series about urban foxes in spring 1997, the public was called to report fox sightings in Swiss cities. The sightings were re- corded personally by collaborators of the Inte- grated Fox Project. Only fox sightings within hu- man settlements were recorded. As the call on TV was biased towards the German speaking part of Switzerland, the scanty information from the French and Italian speaking regions of the coun- try were excluded from further analyses. The pro- gram actus (ESTABROOK and ESTABROOK 1989) was used for the statistical test, which performs randomised contingency tables and gives prob- abilities for deviations from expected values. In spring 1999 we carried out a phone inquiry with people or institutions in charge of wildlife management in all 30 Swiss cities (communities) with more than 20000 inhabitants (FEDERAL OF- FICE OF STATISTICS 1998). The experts were asked about occurrence and abundance of urban foxes, evidence of breeding dens in the urban area, the year of the first urban fox sightings and the cur- rent trend in the urban fox population. In cities with official game wardens (18 out of 30), they were interviewed, in all other cities we questioned non-professional hunters and the nature conser- vation officials. In the conurbation of Geneva (three communities with >20000 inhabitants) our contacts were wildlife biologists running an urban fox project; in Zurich we knew the situa- tion from our own project. Development of the urban fox population in the city and the canton of Zurich There are no direct figures on the red fox popula- tion available. Therefore its development has to be shown indirectly through the hunting record and other recorded causes of death. Longtime fig- ures for an urban area are available for the city of Zurich, because it has been a game sanctuary since 1929. All wildlife management tasks in the city are exclusively performed by official game The rise of urban fox populations in Switzerland wardens, therefore the hunting result is recorded and the locations of dead foxes (shot or found dead) are known. For comparison of the data from the canton and the city of Zurich, we used the HIPD (hunting in- dicator of population density; BöcEL et al. 1974). We defined the HIPD as the annual number of foxes hunted per km? excluding lakes and areas above 2000 meters. We did not include data on foxes with other causes of death than hunting be- cause generally these data have only been avail- able since 1968. To compare data from urban and adjacent rural areas within the city of Zurich, we used a total number of foxes shot or found dead (available from 1960 to 1997), and additionally numbers of the two mortality factors “shot” and “found dead” (mostly road casualties; for the whole city available since 1960, for urban and adjacent rural areas separately available since 1984). To analyse the development of the fox population in the city of Zurich we performed simple linear regressions because the fit of regression of the two mortality factors on the years 1984 to 1997 did not improve by exponential or logistic functions. 157 Results The present occurrence of urban foxes in Switzerland After the call for urban fox sightings on Swiss Television in spring 1997, 194 sight- ings from 78 different towns and villages of the German-speaking part of Switzerland were reported. 138 sightings came from towns with more than 10000 inhabitants (Tab. 1). Of those, more sightings than ex- pected concerned cities wıth more than 50000 inhabitants (randomisation test, p<0.01 ), and less sightings than expected from towns with 10000 — 20.000 inhabitants (P=0.05- Tab): According to our inquiry among institutions in charge of wildlife management in 8 out of 9 cities with > 50000 inhabitants, and in 18 of the 19 cities with 20.000 —- 50.000 inha- bitants, foxes were occasıonally found or common (Fig. 1, Tab. 2). Foxes seem not to Table 1. Reported sightings of foxes in urban areas from the German-speaking part of Switzerland (randomisa- tion test). Accumulated number of inhabitants Size of township > 50000 20 000-50 000 10 000-20 000 Total 958 746 335192 897 430 2 191 368 Number of fox reports Expected number of fox Significance reports according to numbers of inhabitants higher (p < 0.01) ns lower (p < 0.05) Table 2. Occurrence and trend of urban fox populations in 30 Swiss cities, according to an inquiry among peo- ple/institutions in charge of wildlife management. The two cities where no urban foxes were observed (Bern, Lu- gano) are excluded. Questions Answers Where are the urban foxes observed? whole of the city outskirts only Are there any urban breeding dens? Since when have urban foxes been present? a) 1985-1999 b) < 1985 c) not known How do you judge the trend of the fox population? a) increasing (b) stable (c) decreasing Cities with few urban foxes (n = 15) Cities with many urban foxes (n = 13) > r 4 1 7 8 7 3 5 5 0 0 158 SANDRA GLOOR et al. \ nr 7 | Fi ® Lugano (@) many urban foxes, breeding dens N ei a y; 2. Ur (®) few urban foxes, breeding dens Ä N ee I er AO few urban foxes, no breeding dens ER ITALY = AfA nourban foxes 25 0 25 50 Kilometers € ES jakes Fig. 1. Distribution of urban foxes in 30 cities with more than 20000 inhabitants according to local wildlife management experts. Circles of adjacent cities are shifted to avoid overlapping. 215 a) 300) b) 150- T Be IT uhr... 28 ua 60 65 70 75 80 85 90 9% 97 Fig. 2. a) Hunting indicator of population density (HIPD) for the city of Zurich (straight line) and the canton of Zurich (dotted line) from 1960 to 1997. b) Rabies cases in the canton of Zurich from 1960 to 1997. The rise of urban fox populations in Switzerland be present in two towns only: in Bern, situ- ated on the Swiss Plateau, and in Lugano, a city in the southern Alps. In all 13 towns where foxes were reported to be common, they were observed throughout the urban area, (including the centre), and they were breeding in the ur- ban area also (Tab. 2). In 4 cities with more than 50000 inhabitants (Zurich, St. Gallen, Luzern, Biel), breeding dens are known even in the very city centre. In most cities (17 out of 28), urban foxes have been per- ceived as a recent phenomenon since 1985. No geographical trend can be recognised as far as the beginning of settlement in dif- ferent cities is concerned. Only in the conurbation of Geneva, with three cities (communities) with > 20.000 in- habitants (Geneva, Lancy, Vernier; Tab. 2) the population is said to decrease because of an outbreak of sarcoptic mange in 1996 (C. FISCHER, pers. comm.). Development of the urban fox population in the city of Zurich The HIPD of the canton of Zurich and the city of Zurich correlate significantly (Spear- man, r = 0.66, p< 0.001; Fig. 2a), the HIPD in the canton always being higher than in the city. Additionally, the HIPD of canton and city are strongly influenced by rabies trends between 1967, the year when rabies reached Switzerland, and 1985, the year with the last cases of rabies found on foxes in the canton of Zurich (Fig. 2a, b). According to the HIPD, the fox population in the city of Zurich and in the whole area of the canton of Zurich seems to have de- veloped in parallel at least since the begin- ning of the 1970s. Both HIPDs are higher after the rabies epizootic than before. The average of the HIPD from 1993 to 1997 compared to the average of the HIPD from 1960 to 1964 is by 1.7 times higher (2.02 vs. 1.19) in the canton and 13.7 times higher (126 vs. 0.09) in the city of Zurich, indicat- ing a stronger population increase in the city than in the canton. The increase of the HIPD started in the canton in 1984 and in the city in 1985, respectively. 159 However, the development of the fox popu- lation in the whole city of Zurich (with ur- ban as well as adjacent rural areas) is not the same as the development of the popula- tion within the urban area. The first peak of the HIPD in 1967 (Fig. 2) only occurred in the records of foxes from the rural part of the city (Fig. 3), whereas in the urban part of the city fox numbers remained low dur- ing the 1960s and 1970s. The trend to an in- creasing urban fox population in fact just started from 1985 onwards. Before 1985, most of the few foxes of the urban area were only recorded at the bor- der of the city, apart from two foxes, one young fox near the city centre in August 1964 and one young fox in the fairly central railway station Enge in June 1967. Rabies cases were recorded in and near the city of Zurich from 1967 to 1983 (Fig. 3). The prophylactic culling of foxes was carried out as intensively as possible from 1965 to 1995. The numbers of foxes found dead and shot, analysed separately for the whole city correlate significantly (Spearman, r = 0.73, p < 0.001). According to these numbers, the population remained low for almost 20 years after the rabies outbreak, and only in 1985, two years after the last rabies cases were re- corded in the area, the fox population started to increase, both in the urban and in the adja- cent rural part of the city. From 1985 to 1997 the number of foxes shot or found dead in the whole city increased by 20 times from 11 to 223. This trend is true for both mortality factors “shot” and “found dead” and exam- ined separately for urban and adjacent rural areas (Tab. 3). Yet the increase in the num- ber of foxes found dead was stronger in the urban than in the rural area (difference of coefficients, t>4 = 4.11,p< 0.001). Discussion Today, urban foxes are recorded in almost all cities of Switzerland. The presence of breeding dens in urban areas up to the city centres indicates that foxes really live in the cities and are not just occasional roam- ers from the vicinity. We ascribe differences 160 SANDRA GLOOR et al. 250 8 urban foxes 7 rural foxes N 200 3 Y/ toxes of the whole city Bi fa) \ foxes with unknown place of death | E= 150 — foxes found dead in the entire city E 7 7 er TA 2 Bl 3 A o 100 | oO Z D, , Ö G Mi Oo G 0 Y0 S G; Bi un G , 772 E 9 0 7 7" 677, GE G % 60 65 70 75 80 85 90 So El year Fig. 3. Fox mortality (animals shot or found dead) in urban and rural areas in the city of Zurich from 1960 to 1997. From the years 1973-1976 and 1982-1983 there are only total numbers of dead city foxes available (widely hatched bars). No precise locations of death are available for some recorded foxes from 1984 onwards (white bars). The years with rabies cases within 5 km of the city centre (Kappeler 1991) are marked with black bars. Table 3. The increase of numbers of recorded dead foxes within the city bordes of Zurich, described by the linear regression of the two mortality factors “shot” and “found dead” (mostly road casualties) from 1984 to 1997. Foxes of urban areas Coefficient 5.215 3.310 Mortality factor Shot Found dead Foxes of adjacent rural areas Coefficient 4.842 0.831 Mortality factor Shot Found dead in fox population densities in Swiss cities of today mainly to the fact that urban foxes have been a recent phenomenon and the development is still going on. Our call for fox sightings on Swiss television revealed that more foxes are recorded from larger towns than from smaller ones, a rela- tion that was also observed by MACDONALD and Newpiıck (1982) in Great Britain. This could be because larger towns may have a higher proportion of suburban habitat, where the highest fox densities are found (HARRIS and RAYNER 1986 b). Although red foxes generally avoid the di- rect presence of humans, some foxes have lived in the neighbourhood of humans set- tlements for a long time, shown, e.g., by the naturalist ScHINZ (in INEICHEN 1997), who noted in 1842, that red foxes had al- ways lived in the moats surrounding the city of Zurich. The hunting statistics of the city of Zurich show that foxes have been pre- sent in the urban area since the early 1960s, but such observations remained iso- lated cases. In 1985 the situation began to change. Due The rise of urban fox populations in Switzerland to successful oral vaccination campaigns against rabies, the fox population in Swit- zerland started to recover (BREITENMOSER et al. 2000), which is recorded in other Eur- opean countries, as well (e.g. Vos 1993; Ar- toıs et al. 1997). It was parallel to this gen- eral trend, when the urban fox population in the city of Zurich and in most other Swiss cities showed a drastic increase. However, hunting statistics have to be inter- preted cautiously, because they do not only correlate with the real fox populations but are also influenced by other factors such as the preferences of the hunters (MACDONALD and Voigt 1985; GoszczyNskI 1989) or out- breaks of zoonoses (KAPPELER and WANDE- LER 2000). A high hunting pressure most probably lasted during the whole period of rabies from 1967 until at least to the end of the 1980s. Therefore, the low HIPD during this period presumedly reflects low densities of fox populations. With the decrease of ra- bies the motivation to hunt foxes probably decreased drastically. The HIPD, on the other hand, was still increasing during the 1990s. We therefore suggest that the real trend of fox populations is underestimated by hunting statistics. The fox population in the canton of Zurich with its high degree of urbanisation must be even more underesti- mated by the HIPD, because foxes are hardly ever shot in most urban areas, where hunting generally is not permitted. The game sanctuary of the city of Zurich is an exception, where a constant hunting re- gime is maintained by official game wardens. The significant correlation of the develop- ment of foxes “shot” and “found dead” within the city confirms, that the increasing numbers of dead foxes are not only the re- sult of an increased shooting effort. A sımilar development of urban foxes as in Switzerland recently took place in other parts of the distribution area of the red fox which is shown by reports, e.g., from Oslo, Norway (CHRISTENSEN 1985), Arhus, Den- mark (MOoLLER NIELSEN 1990), Stuttgart, Germany (T. RoMIG, pers. comm.), Toronto, Canada (Apkıns and Stortr 1998) and Sap- poro, Japan (K. URAGUCHI, pers. comm.). The questions arises why the invasion of ur- 161 ban habitat started and which factors caused this new development. According to Harrıs and RAYNER (1986c), the colonisation of British towns already started in the 1930s. During these years there was a boom of private house con- struction resulting in large districts of mid- dle-class suburbs with low-density housing, and medium-sized gardens. This is the type of habitat which Harrıs and RAYNER (1986b) found to be favoured by foxes. Once established in these residential sub- urbs, foxes moved further into the city and also colonised less favoured habitats. HAR- rıs and RAYNER (1986b) found urban foxes to be less common in areas consisting of council-rented housing, in city centres, and around industrial areas. The colonisation of Swiss cities by foxes re- sulted in a similar phenomenon as known from Great Britain. However, the underly- ing cause for the rise of the urban fox popu- lations seems to be different, because the development of Swiss cities in the past thirty years was unlike British cities in the 1930s. We propose two hypothetical expla- nations for the presence of urban foxes: The population pressure hypothesis (PPH) and, as an alternative, the urban ısland hy- pothesis (UIH). The population pressure hypothesis PPH postulates that urban foxes are simply in- truders from the adjacent rural areas. These foxes invade in human settlements because of a high population density in rural areas. According to the PPH, urban areas would provide suboptimal habitats for foxes, the dynamics of an urban fox population would closely correlate with the trend of the fox population in the adjacent rural areas, and the urban fox population would genetically not be different from the adjacent rural population (Rousser 1999). The alternative urban island hypothesis UIH postulates that urban foxes have adapted to specific urban conditions such as high density of human population. Therefore, urban foxes would be able to use specific urban resources such as sca- venged food items or special hiding places during daytime. The dynamics of such an 162 SANDRA GLOOR et al. urban fox population would be independent from the trend in the adjacent rural areas. The colonisation of urban areas could have been initiated by the behavioural adapta- tions of a few foxes that gave them access to exploit human settlements as a free niche. As only a few individuals founded the new urban population, we would expect it to be genetically isolated from the popu- lation in the rural surroundings. The simultaneous emerging of urban foxes throughout Switzerland along with the in- creasing fox population indicates that the high population pressure has at least in- itiated the immigration of the founder indi- viduals into the cities. MACDONALD and NEwWDICcK (1982) suggested that there was no strict division between rural and urban foxes in Oxford, because they had radio- tracked foxes which regularly commuted between urban and rural areas. Neverthe- less, living in the city requires special adap- tations, and many anecdotal observations reveal that foxes indeed have adapted to this exceptional environment. Further re- search on resource exploitation and genetic structure of the urban fox population will allow to compare the two hypotheses. The presence of foxes in human settlements raises the question of the impact of human behaviour and human attitudes on the urban fox population (BonTADInA et al. 2000). Harrıs (1981b) and DoncAsTER et al. (1990) showed, that food directly or indi- rectly provided by humans can make up a major part of the diet of urban foxes. People feel ambivalent about urban foxes, being either fascinated by this wild carnivore in their neighbourhood or afraid of it because of zoonoses (Bontadina et al. 2000). In fact, foxes in close vicinity to humans and pets could indicate new zoonotic risks (HOFER et al. 2000). The red fox is the main vector of rabies in Europe. Up to now ur- Zusammenfassung ban areas were considered to be barriers to the spread of rabies (STECK et al. 1980), therefore the increase of urban fox popula- tions calls for additional strategies in case of a new outbreak of rabies (MACDONALD and VoıGr 1985; HARRIS et al. 1988). Furthermore, the zoonosis alveolar echino- coccosis (AE), caused by the small fox ta- peworm Echinococcus multilocularis, could become more important through the in- crease of urban fox populations. In Switzer- land, the incidence rate of human AF has not significantly changed over the past 36 years, suggesting a stable epidemiologi- cal situation (ECKERT and DEPLAZES 1999), but regarding the long incubation period of AE of 5-15 years, it would be advisable to study this zoonosis further, especially in ur- ban areas. Results of such studies could have an important impact on the manage- ment of urban fox populations. Acknowledgements The authors thank all collaborators involved in the Integrated Fox Project. We especially thank CLAUDE FISCHER, who conducted the inquiry of game officials in the French part of Switzerland and provided useful information on the urban fox population of Geneva, and CHRISTIAN STAUF- FER for his generous collaboration and for provid- ing access to the data of the game sanctuary of Zurich. We thank ALEXANDER MATHIS, CLAUDE FISCHER, CHRISTIAN STAUFFER, and JEAN-MARC WEBER for comments on an earlier draft. We also thank the game wardens and people in charge of wildlife management in the 30 large cities of Swit- zerland. The study was supported by the Swiss National Science Foundation (grant no. 31- 47031.96), the Fonds zur Förderung des akade- mischen Nachwuchses FAN of the Zürcher Hochschul-Verein and the Swiss Federal Offce for Education and Science (EU FAIR Projekt CT97-3515/BBW Nr. 97.0586). Die Entstehung urbaner Fuchspopulationen in der Schweiz Seit Mitte der 1980er Jahre werden zunehmend Füchse inmitten von Schweizer Städten beobachtet. Die Befragung der zuständigen Behörden ergab, daß heute in 28 der 30 größten Schweizer Städte The rise of urban fox populations in Switzerland 163 Füchse registriert werden. In 20 dieser Städte sind Fuchsbaue mit Jungenaufzucht im Sied- lungsraum bekannt. Dabei werden Stadtfüchse überproportional häufger in größeren Städten als in kleineren Ortschaften beobachtet. In Zürich, der größten Schweizer Stadt, waren gemäß der Jagdstatistik bis zu Beginn der 1980er Jahre Stadtfüchse sehr selten. Erst ab 1985 begann die städtische Fuchspopulation markant anzusteigen. Auch die umliegenden ländlichen Gebiete ver- zeichnen ab 1984 eine deutliche, allerdings weniger starke Zunahme der Fuchsbestände, die u.a. mit der erfolgreichen Tollwutbekämpfung zusammenhängt. Als Erklärung der Präsenz von Füchsen im Siedlungsraum, einem bisher vor allem aus Großbritannien bekannten Phänomen, schlagen wir zwei alternative Hypothesen vor, welche einerseits den Populationsdruck in ländlichen Gebieten, andererseits stadtspezifsche Verhaltensanpassungen der Füchse ins Zentrum stellen. Fuchspopula- tionen im Siedlungsraum beeinflussen das Verhalten und die Einstellung der Bevölkerung gegen- über Wildtieren und haben Konsequenzen für das Fuchsmanagement und den Umgang mit Zoono- sen, wie Tollwut und alveoläre Echinokokkose. References ADKkINS, C. A.; Stort, P. (1998): Home ranges, movements and habitat associations of red foxes Vulpes vulpes in suburban Toronto, On- tario, Canada. J. Zool. (London) 244, 335-346. ARTOIS, M.; LAnGLAISs, M.; Suppo, Ch. (1997): Si- mulation of rabies control within an increas- ing fox population. Ecol. Mod. 97, 23-34. BEAMES, I. R. (1969): Mammals in the London area, 1967. Lond. Nat. 48, 40-47. BEAMES, I. (1972): The spread of the fox in the London area. Ecologist 2, 25-26. BöcGEL, K. A.; ArATA, A.; Moegle, H.; Knorpp, F. (1974): Recovery of reduced fox populations in rabies control. Zbl. Vet. Med. B 21, 401- 412. BONTADINA, F.; CoNTESSE, P; GLooR, S. (2001): Beeinflußt die persönliche Betroffenheit die Einstellung gegenüber Stadtfüchsen? - Aspekte zur Akzeptanz eines Wildtieres im Siedlungsgebiet. Forest Snow and Landscape Research (in press). BREITENMOSER, UÜ.; MÜLLER, U.; KAPPELER, A.; ZA- NONI, R. G. (2000): Die Endphase der Tollwut in der Schweiz. Schweiz. Arch. Tierheilk. 147, 447-453. CHRISTENSEN, H. (1985): Urban fox population in Oslo. Rev. Ecol. (Terre Vie) 40, 185-186. DOoNCASTER, C. P.; DICKMAnN, C. R.; MACDONALD, D. W. (1990): Feeding ecology of red foxes (Vulpes vulpes) in the city of Oxford, Eng- land. J. Mammalogy 71, 188-194. EcKERT, J.; DEPLAZES, P. (1999): Alveolar echino- coccosis in humans: The current situation in central Europe and the need for countermeas- ures. Parasitol. Today 15, 315-319. ESTABROOK, C. B.; ESTABROOK, G. F. (1989): Actus: a solution to the problem of small samples in the analysis of two-way contingency tables. Hist. Meth. 82, 5-8. FEDERAL OFFICE FOR STATISTICS (1998): Statistik der Schweizer Städte. In: Stat. Jahrbuch des Schweizerischen Städteverbandes. Bern: Bun- desamt fiir Statistik. GOSZCZYNSKI, J. (1989): Population dynamics of the red fox in central Poland. Acta Theriol. 34, 141-154. HARRIS, S. ( 1977): Distribution, habitat utilization and age structure of a suburban fox (Vulpes vulpes) population. Mammal Rev. 7, 25-39. HARRIS, S. (1981 a): An estimation of number of foxes (Vulpes vulpes) in the city of Bristol, and some possible factors affecting their dis- tribution. J. Appl. Ecol. 18, 455-465. HARRIS, S. (1981 b): The food of suburban foxes (Vulpes vulpes), with special reference to London. Mammal Rev. 11, 151-168. HARRIS, S.; RAYNER, J. M.V. (1986a): Urban fox (Vulpes vulpes) population estimates and ha- bitat requirements in several British cities. J. Animal Ecol. 55, 575-591. HARRIS, S.; RAYNER, J. M. V. (1986 b): Models for predicting urban fox (Vulpes vulpes) numbers in British cities and their application for ra- bies control. J. Animal Ecol. 55, 593-603. HARRIS, S.; RAYNER, J. M. V. ( 1986c): A discrimi- nant analysis of the current distribution of ur- ban foxes (Vulpes vulpes) in Britain. J. Animal Eeol. 55, 605-611. HARRIS, S.; SMITH, G. C.; TREWHELLA, W. J. (1988): Rabies in urban foxes (Vulpes vulpes), devel- oping a control strategy. State Vet. J. 42, 149- 161. HOFER, S.; GLOOR, S.; MÜLLER, U.; MATHIS, A.; HEGGLIN, D.; DEPLAZES, P. (2000): Urban cy- 164 SANDRA GLOOR et al. cle of Echinococcus multilocularis in the city of Zurich, Switzerland. Parasitology 120, 135- 142. INEICHEN, S. (1997): Die wilden Tiere in der Stadt. Zur Naturgeschichte der Stadt. Frauenfeld: Im Waldgut. KAPPELER, A. (1991): Die orale Immunisierung von Füchsen gegen Tollwut in der Schweiz. Diss. thesis, University of Bern, Switzerland. KAPPELER, A.; WANDELER, A. I. (2000): Entwick- lung der Strategien zur Feldanwendung der oralen Immunisierung von Füchsen gegen Tollwut. Schweiz. Arch. Tierheilk. 147, 439- 446. MACDONALD, D. W.; NEWDICK, M. (1982): The dis- tribution and ecology of foxes, Vulpes vulpes in urban areas. In: Urban Ecology. Ed. by R. BORNKAMM, J. A. Lee, and M. R. SEEWARD. Oxford: Blackwell Scientific Publi. Pp. 123- 135: MACDONALD, D. W.; VoIGT, D. R. (1985): The bio- logical basis of rabies models. In: Population Dynamics of Rabies in Wildlife. Ed. by P.]J. Bacon. London: Academic Press. Pp. 71-103. MOLLER NIELSEN, S. (1990): The food of rural and suburban woodland foxes Vulpes vulpes in Denmark. Natura Jutlandica 23, 25-32. MÜLLER, U.; KAPPELER, A.; ZANONI, R. G.; BREI- TENMOSER, U. (2000): Der Verlauf der Tollwut in der Schweiz — Landschaft prägt die Aus- breitung einer Wildtierepidemie. Schweiz. Arch. Tierheilk. 142, 431-438. PAGE, R.J. C. (1981): Dispersal and population density of the fox (Vulpes vulpes) in an area of London. J. Zool. (London) 194, 485491. RousseTr, F. (1999): Genetic differentiation within and between two habitats. Genetics 151, 397- 407. STECK, F.; WANDELER, A.; NYDEGGER, B.; MANIG- LEY, C.; WeEıss, M. (1980): Die Tollwut in der Schweiz 1967-1978. Schweiz. Arch. Tierheilk. 122, 605-636. TEAGLE, W. G. (1967): The fox in the London sub- urbs. Lond. Nat. 46, 44-68. Vos, A. (1993): Aspekte der Dynamik einer Fuchspopulation nach dem Verschwinden der Tollwut. Diss. thesis, University of Munich, Germany. WANDELER, A.l.; CAPT, S.,;, KAPPELER; A.; HAU- SER, R. (1988): Oral immunization of wildlife against rabies: concept and first field experi- ments. Rev. infect. Diseases 10, 649-653. Authors’ addresses: SANDRA GLOOR, F. BONTADINA, D. HEGGLIN, Ur- ban Ecology and Wildlife Research, Wuhrstrasse 12, CH-8003 Zurich (sandra.gloor@gmx.ch); P. DEPLAZES, Institute of Parasitology, University of Zürich, Winterthurerstrasse 266a, CH-8057 Zürich; U. BREITENMOSER, Swiss Rabies Centre, Institute of Veterinary Virology, University of Bern, Länggassstrasse 122, CH-3012 Bern. Mamm. biol. 66 (2001) 165-173 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Original investigation Mammalian Biology Zeitschrift für Säugetierkunde Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco) By JuANA C. DEL VALLE, MARIA I. LOHFELT, VIVIANA M. COMPARATORE, MARIA S. CID, and CRISTINA BUSCH Departamento de Biologia, Universidad Nacional de Mar del Plata, Mar del Plata Argentina Receipt of Ms. 02. 06. 1999 Acceptance of Ms. 17. 10. 2000 Abstract We tested feeding selectivity and food preference of Ctenomys talarum (tuco-tuco). To test feeding selectivity, above ground and below ground plant biomass from the field was determined and botanical composition of the diet was estimated in stomach contents using microhistological tech- niques. Feeding preferences were studied carrying out laboratory cafeteria experiments. Ctenomys talarum behave as generalist and opportunistic herbivores consuming the greater part of species present in the grassland. The above ground portion was preferred over the subterranean one. Grasses constituted 94% of the above ground vegetative fraction consumed and were generally se- lected. Preference trials also showed that C. talarum prefer above ground parts of grasses to other choices. Key words: Ctenomys talarum diet, feeding selectivity, food preference Introduction Rodents of the genus Ctenomys (tuco-tu- cos) are subterranean herbivores whose populations are distributed in a discontinu- ous pattern throughout Argentina, Para- guay, Bolivia, Uruguay, Perü, Chile, and southern Brazil (Woops 1984). Most herbi- vores inhabit a biotope in which the food plants are more or less continuously distrib- uted in space and time, and whose accessi- bility is restricted by the structural and che- mical properties of the vegetation (ILLIUS and GorDon 1993). They select food items according to their preference, and availabil- ity in the field. Preference is the predilec- tion of a consumer for a particular class of 1616-5047/01/66/03-165 $ 15.00/0. food, and it ıs the result of how well the consumer “likes” this food relative to other ones, when all are equally available (NoRr- BURY 1992). Diet selection in herbivores may be explained by models where the rate of intake ıs maximized with nutrient con- straints, toxins are avoided or their intake is minimized (STEPHENS and KRrEBS 1986). A foraging herbivore maximizes its nutrient intake when greater nutrient intake con- verts directly into greater survival and re- production (nutrient maximization; BELOV- sky and SCHMITZ 1994). Food resources have been implicated as im- portant to both burrow location and burrow 166 JUANA C. DELVALLE et al. system size, suggesting that foraging is a cri- tical component of ecology of subterranean rodents (Busch# et al. 2000). In terms of nu- tritional value, below ground plant tissues may represent a more variable resource than above ground tissue (ANDERSEN 1987). This fact and the high energetic costs of digging may influence food selectivity. HETH et al. (1989) proposed that subterra- nean herbivores cannot afford to be selec- tive feeders since search costs would exceed the benefits of being selective, therefore they should utilize all food that they en- counter. Furthermore, subterranean To- dents are expected to consume a great pro- portion of below ground vegetation (VLeck 1979). Diet selection by herbivores is important in determining their effects on plant commu- nities. Empirical evidence and theoretical models suggest that generalist herbivores may have more wıdespread effects on plant communities than specialist herbivores, since they can greatly reduce, or even elimi- nate, some plant species while persisting on the remaining species. Much remains to be learned regarding the foraging ecology of subterranean rodents. Cafeteria-style test of food preference would help to determine the nature and extent of dietary specializa- tion. In addition, field studies of foraging be- havior would allow to test optimal foraging models (STEPHENS and Kress, 1986) under the condition faced by free-living animals. This study assesses different aspects of the feeding behavior of individuals of a Cie- nomys talarım population inhabiting a coastal dune grassland in the southeastern Buenos Aires province, Argentina. Specifi- cally assessed are: 1) C. talarum feeding se- lectivity in the field, and 2) food prefer- ences of C. talarum in cafeteria test. Material and methods Two studies were conducted. One was carried out in the field to evaluate C. falarum feeding selec- tivity. The other was a cafeteria test developed to determine if food quality (fiber/protein) deter- mines their feeding preference. Feeding selectivity (field data) The study was conducted on coastal dunes at Mar de Cobo (Buenos Aires Province, Argentina), in a natural grassland with the predominance of per- ennial grasses (COMPARATORE et al. 1991). Vegetation and animals were sampled in autumn, winter, spring, and summer. Fifty seven animals were kill-trapped and their stomachs removed. Because above ground foraging occurred near the burrow opening, for each animal captured, four vegetation samples were collected from around the opening (30cm diameter and 30 cm depth). Above ground and below ground samples were separated and dry plant biomass was esti- mated and expressed as percentage of total bio- mass. In addition, the percentage of the above ground fraction of each species was computed. The botanical composition of Ctenomys talarum diet was estimated using microhistological techni- ques. Stomach contents were processed individu- ally according to WırLıams (1969), and the botan- ical composition of the diet was quantified according to SPARKS and MALECHER (1968). The subterranean, above ground and reproductive fractions were quantified. In addition, the species percentages in the vegetative fraction were deter- mined, since it is the only one in which fragments could be differentiated to species level. The seasonal percentages of the components of Ctenomys diets were compared using a Kruskal- Wallis test (P = 0.05). Diet and grassland botanical composition were contrasted, establishing animal selectivity for total above ground and subterra- nean fractions. Reproductive fraction selectivity could not be established because its percentage was not determined in the grassland. In addition, selectivity for the above ground vegetative frac- tion of the species in the diet was computed. The following index (KRUEGER 1972) was used to determine relative species selectivity: SI=% Dixfdi/% Pix fpi, where % Di and fdi are the seasonal mean per- centage and the frequency of component i in the diets, and % Pi and fpi are the seasonal mean per- centage and the frequency of the component i in the grassland. Chi square with 95% confidence was used to determine if the seasonal SI for each component was significantly different from 1. Kulcyznski’s similarity index (HoLECHEK et al. 1984) was used to evaluate the similarity of diets and pasture. The species considered in the analy- sis were those whose seasonal mean percentage by the frequency, in the diets and/or in the grass- land, were over 2%. Results are shown as mean +SD (standard deviation). Feeding selectivity and food preference of Ctenomys talarım (tuco-tuco) Feeding preference (cafeteria test) Animals for the experiments were live-trapped in the coastal dunes of Necochea (Buenos Aires province). Food preference was investigated in the laboratory by the amount of plant matter con- sumed during feeding trials (PHILLıpson et al. 1983). To conduct the trials, animals were set in a feeding apparatus (42x42x6 cm) which consisted of a central nest box with an opening in each of four feeding arenas. The gridded floors of the feeding arenas allowed food remains and faeces to drop into a collecting tray without being mois- tened with urine. Leaves and stems of plant species for the experi- ment were collected at the same site where the animals were trapped. Potatoes and carrots, which have low fiber percentages, were also used as choices. Standard cafeteria trials were conducted performing three different tests. In each one the same wet weight of four different kinds of food was offered simultaneously. Each test lasted for four days with 10 repetitions using different indi- vidual tuco-tucos (5 adult females and 5 adult males). Every day equal wet weight of each plant choice was offered. Intake was calculated on both a fresh-mass and a dry-mass basis. The residual plant material was sorted and weighed on suc- ceeding days and the difference recorded. Then, it was dried at 70-80 °C to invariable weight. Con- versions of fresh mass to dry mass were calculated from samples of plant material that was main- tained in empty cages during the trials. Each choice of food was weighed to the nearest 0.01 g and offered in different compartments of the feeding apparatus. The position of the foods varied at random. Species chosen for the cafeteria test are present in the natural diet of Crenomys ta- larım in Necochea, except for /pomea batatae and Daucus carota. To test if dietary preference correlates with some particular portion of the plant we offered differ- ent parts of two species of grasses that appeared in the field diet: Cynodon dactylon stem, C. dacty- lon leaf, Bromus unioloides stem, B. unioloides leaf. To test if the preference has a relation with the fiber content of the choice C. dactylon stem, B. unioloides stem and /pomea batatae tuber and Dacus carota root (fiber content: 17% and 13%, respectively) were offered. We also tested the preference for forbs or grasses and different parts of them, offering two species from the field diet: Hydrocotile bonariensis (forb) above ground por- tion H. bonariensis below ground portion, Pani- cum racemosum (grass) above ground portion, P. racemosum below ground portion. 167 Protein content was determined by the microbiur- et method (GornALL et al. 1949) and fiber con- tent ‚was determined by the GoERING and VAN SOEST (1970) technique. Results are shown as mean+ S.D. A non-parametric multiple com- parison test (ZAR 1984) was used to ascertain the significance of the preferences observed. Chi square with 95% confidence was used to test whether the proportions of food consumed were equal to expected frequencies, based on the rela- tive dry weight of food offered. Results Feeding selectivity Comparison of botanical composition of grassland and diet Proportion of subterranean biomass was not significantly different from the above ground (P > 0.05; Fig. 1a). Perennial grasses dominate the grassland biomass (79% of to- tal above ground available biomass). The proportion of perennial monocotyledo- neans decreased in spring, while annual monocotyledoneans increased; annual forbs decreased in autumn and perennials in win- ter (Fig. 2a). Monocotyledonean composi- tion was dominated by Panicum racemo- sum, which constituted 25% of the annual biomass. Analysis of the contents of 57 stomachs re- vealed that tuco-tucos exploited at least 16 species of plants annually. The above ground vegetative portion of plants predo- minated (84.5%) in the annual diet, whereas subterranean and reproductive portions constituted only 11% and 4.5%, respectively (Fig. 1b). Grasses comprised the highest proportion of the annual diet. Its average annual occurrence was 94% of the annual above ground vegetative frac- tion (Fig. 2b). Bromus unioloides (46%), Panicum racemosum (16%), and Poa bo- nariensis (10%) were consumed more in- tensively, as they constituted 72% of the dry weight of the annual diet. Seasonal changes in diet Although perennial grasses comprised the highest proportion of the diet year round (Fig. 2b), consumption preference of differ- 168 JUANA C. DELVALLE et al. percentage of biomass spring winter autumn 100% ab 80% 60% 40% percentage of blomass 20% == 5 3 Do o o 0% + [e] Fu 2 autumn je}0} —ı 1% 3 ® 2 =; u: je10} saew 5 3 o 0» ” ® [73 winter sojewa} =) spring ent species varied. Thus, the analysis of the seasonal diet of the tuco-tucos revealed that Bromus unioloides was an important source of nutriment almost year round, but it was consumed less during the autumn season. On the other hand, only during spring Pani- cum racemosum ceased to be an important component of the tuco-tuco’s diet. Poa bo- nariensis was consumed more in autumn, spring, and summer and less in winter. The perennial forb Adesmia incana became an important component in the spring diet (10%). Although there were no significant differences between the average of the be- low ground fraction consumed in each sea- son (P>0.05); the mean consumption (5.3+4) in spring was lower than in the other seasons (Fig. 1b). summer Ol aboveground El belowground summer Fig. 1. Percentages of bio- mass in the grassland and in the diet of tuco-tuco: a) Above ground and subterra- nean percentage of plant biomass in the natural grass- land at Mar de Cobo where (tenomys talarum was trapped; b) Above ground, reproductive and below ground percentage of plant biomass in the stomach of | Ctenomys talarum trapped in | Pbelowground a natural grassland at Mar de Cobo. Different letters indicate significant differences be- tween seasons; * indicates significant differences be- tween sexes in each season. DOlaboveground | reproductive | EJaboveground vegetative 1810} © 3 2 ® (7) Effect of sex on diet Males seem to be more selective than fe- males (male and female diets show a 58% and 66% similarity with the grassland, re- spectively; Kulcyznski’s index). Differences were noticed in winter and in summer when male and female diets exhibited statistical differences in the proportion of Bromus uni- oloides (P < 0.02). Moreover, plant fractions were consumed differentially (Fig.1b). Whereas males consumed the subterranean and vegetative fractions of plants in the same proportions year round, females did not (P< 0.02). Furthermore, in autumn fe- males consumed a higher proportion of the subterranean fraction (P = 0.006) and smal- ler proportion of the above ground vegeta- tive (P = 0.004) than males, and during sum- Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco) 100% - DT a 80% + TIL E Q 2 60% + yo © ® oO) Ss 40% I [= ® 2 2 20% + 0% + autumn winter spring summer 100% - 2 80% - = @ E ‚Q 0 60% - ° D Ss 40% - E © > E 20% - 0% + ’ i autumn winter spring summer 169 EI other forbs Oi perennial forbs annual forbs Elperennnial grasses Eannual grasses Fig. 2. Botanical composi- tion of the grassland and of the diet of tuco-tuco: a) Per- centage of biomass of above ground available vegetation on a natural grassland at Mar de Cobo; b) Percentage of biomass of different types of above ground vegetation in stomachs of Ctenomys ta- larum. DO perennial forbs annual forbs DO perennial grasses EB annual grasses Table 1. Seasonal values of selectivity index (SI) for each above ground vegetative food item and for the subter- ranean fraction. (*) denotes statistically significant difference from 1 (SI # 1) Chi-square test (P = 0.05) AUTUMN N=15 Above ground fraction Annual grasses Perennial grasses Annual forbs Perennial forbs Subterranean fraction mer they consumed a significantly higher proportion of reproductive structures than males (P = 0.028; Fig. 1b). Relative plant selectivity Tuco-tuco ingested perennial monocots and dicots in proportion to their mass (Tab. 1), and thus according to the probability of en- countering them. Nevertheless, tuco-tucos are capable of selective foraging, since the above ground fraction of the plant was not selected by individuals of both sexes in all seasons (P< 0.05). In addition, the analysis WINTER N=15 SPRING N=12 SUMMER N=15 of the stomachs showed that tuco-tucos se- lect some monocots species with preference changing seasonally; the grass Bromus was selected in winter, spring, and summer but was indifferent in autumn, whereas Poa was selected in autumn, spring, and summer but not in winter. Panicum was preferred in autumn and avoided in other seasons (P = 0.05). Furthermore, males and females showed different feeding selectivity for Bromus unioloides, thus, it was selected in winter and summer by males, but not by fe- males (P = 0.05). 170 JUANA C. DELVALLE et al. Feeding preference Tuco-tucos consumed 200457 gind ' d’' of food, and S+3g protein and 26+53 kcal per day. The experiments demostrate that C. talarım is able to discriminate among the plant species tested, and harvested grasses selectively. Although some species and/or part of the plant were consumed more than others (Tab. 2), tuco-tucos con- sumed all plants offered in the test and con- sumption of choices other than the pre- ferred ones make an important contribu- tion to total ingested nutrient (7-44% pro- tein). Results indicate preference for above ground portions of grasses over other choices tested. Furthermore, plant portions with a low fiber/protein ratio were less pre- ferred than those with a high fiber/protein ratio. When offered as above ground sam- ples, significant quantities of all grasses were consumed by the tuco-tucos and no preference for stems or leaves was detected, but as noted above a preference for low quality food was noticed, thus the B. unio- loides leaf, which has the lowest fiber/pro- tein ratio, was eaten to a lesser proportion than the other choices (Tab. 2a). When of- fered monocots (B. unioloides, C. dactylon or P.racemosum) and other choices, grasses represented 70-90% of the total consump- tion and grass stems were preferred to other choices tested (Tab. 2b, c). This pre- ference was independent of the nutritional quality of the other choice, thus the stems of grasses with a higher fiber/protein ratio were preferred to /. batatae, D. carota or to Hydrocotyle bonaroensis above ground pro- portion (Tab. 2b, c). Discussion Feeding selectivity Ctenomys talarım behave as a generalist and opportunistic herbivore since it con- sumes the greatest part of the species pre- sent in the grassland, and changes its diet in relation to food availability. Similar food habits were reported for other Ctenomys species (C. australis, COMPARATORE et al. 1995 and C. mendocinus, MADOERY 1993) and other subterranean rodents such as Table 2. Dry weight consumption and fiber/protein ratio of food items for three different cafeteria tests. (a) First cafeteria test compared the consumption between different portions of two species of grasses; (b) second cafeteria test compared the consumption between items with high and low fiber/protein ratio; (c) third cafeteria test compared between subterranean and above ground fraction of forbs and grasses. Non parametric multiple comparisons test to differentiate among preference fractions (small letters) (P = 0.05). SRECIES QUANTITY EATEN FIBER/PROTEIN g/day + SD RATIO (a) B. unioloides leaf 11.29 44.79 B. unioloides stem C. dactylon leaf C. dactylon stem (b) B. unioloides stem C. dactylon stem I. batatae D. carota (e) H. bonariensis above ground H. bonariensis subterranean P. racemosum subterranean P. racemosum above ground 24.93 +11.61b 23.63 + 12.26 ab 28.87 +12.69b 32.39216.3a Sl. 55=E214.2:1fe 4.65 + 2.18b 2.87.-:2127121b 8.6 + 2.264 15.42 # 12.21 ab 17.23 =4.077be 3IS3EE ZI SIE Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco) Thomomys talpoides (STUEBE and ANDER- sEN 1985), Geomys attwateri (WILLIAMS and CAMERoN 1986), Heterocephalus glaber (BrETT 1991) and Spalax ehrenbergi (NEVo 1979). This behavior would be adaptive for a mammal that supports a high cost of bur- rowing and poor available energy (HETH et al. 1989). In general, the food habits of C. talarum at Mar de Cobo appear to be similar to those reported for C. talarıum at Necochea (COMPARATORE et al. 1995). Indi- viduals of both populations preferred monocotyledoneans, but tuco-tucos con- sumed large amounts of Bromus at Mar de Cobo and of Poa at Necochea (CoMPARA- TORE et al. 1995), suggesting that modifica- tions in the diet may be influenced by changes in food offered. Given the high cost of burrowing (VLEcK 1979) it is not sur- prising that tuco-tucos shift their diet in ac- cordance with habitat availability. C. talarum selected the above ground frac- tion of plants. This may be due to the fact that tuco-tucos live in areas where plant species have different life cycles, therefore the above ground fraction would be avail- able all year round. On the other hand, the lowest consumption of the subterranean fraction during spring is in relation with the active growth of the above ground frac- tion in this season. WILLIAMS and CAMERON (1986) indicated that the difference in the subterranean and above ground proportion of plants in the diets of pocket gophers is related to the different behavior of the ani- mal species. The above ground proportion would be higher in those groups that spend more time out of their burrows. In this sense although tuco-tucos forage within their tunnels, they feed mostly above ground by venturing away for their tunnels for brief periods to gather plant parts from the surface. The vegetation in the vicinity of their holes commonly shows evidence of their feeding activities (ReEıG 1970). At Mar de Cobo where densities were high (65 ind./ha), reproductive structure con- sumption was minimal (4%), whereas at Necochea (13ind./ha) it played an impor- tant role in Cienomys diet (38% of total; COMPARATORE et al. 1995). This suggests that 171 the proportion of high caloric food is higher in animals living in populations of low den- sity. BUJALSKA (1983) reported a similar re- lationship between density and diet quality for Clethrionomys, a forest dwelling micro- tine. The diet of tuco-tucos depends on sex, as females appeared to be less selective than males. The larger consumption of reproduc- tive plant structures by females could re- spond to higher protein requirements for lactation. Differences in preference by re- productive females have been reported for other subterranean mammals like Geomys attwateri (WILLIAMS and CAMERON 1986) and Spalax erhenbergi (Nevo 1991). Feeding preference Choice tests support the fact that Cienomys talarım is a herbivorous generalist with a preference for the above ground fraction of grasses. Thus, although some items were preferred, the diet was supplemented with other choices. In this manner, a varied diet was maintained, even with the abundance of the preferred food resource and without differential foraging costs. Herbivores may select a diet that mixes different types of dietary items to balance the intake of nutri- ents required for proper growth or success- ful reproduction (REZSUTEK and CAMERON 1998). If we accept 200 g fresh weight as an aver- age daily consumption, tuco-tucos intake would amount to 2600 g and 13000 g fresh vegetation per ha consumed each day at Necochea and Mar de Cobo, respectively. This amounts to 996-4680. kg per ha per year, not including vegetation stored un- eaten or used to build nests. Tuco-tuco total energy intake per day was comparable to data reported for the subterranean rodent Thomomys talpoides by STUEBE and An- DERSEN (1985). In our experiments, grasses provided not only most of the daily energy and protein requirements, but also with more than 80% of the daily dietary fiber. As tuco tucos are coprophagous rodents with a large caecum (11% of the gut), they are able to optimize 172 JUANA C. DELVALLE et al. the assimilation of nutrient from this die- tary fiber. Thus, in a study on efficiency of food utilization MARTINo (2000) found that in spite of the high fiber/protein ratio of B. unioloides ıts apparent digestibility (NDF) was high (0.81 + 0.04). Selection of grasses was also reported for the rodent La- gostomus maximus by BRANCH et al. (1994). Assuming food quality is correlated with the amount of annual forbs in the diet, these authors suggested, that grasses may provide essential dietary fiber to maintain caecum motility and the appropriate micro- bial environment in the hindgut. According to optimal-foraging models, plant defenses (e.g., structural, digestive-inhibit- ing chemicals, toxic chemicals, and nutri- Zusammenfassung tional content) may be effective in reducing their intake by mammalian herbivores (BE- LOvskY and ScHMIiTzZ 1994). For example, Hydrocotile bonairensis, which was not pre- ferred by C. talarum, belongs to a genus known to contain relatively high concentra- tions of phytotoxins (JUSCAFRESA 1975). Acknowledgements This research was granted by UNMdP subs. N° 2; Agencia Nacional de Promociones Cientificas y Tecnolögicas subsidio N° 01-00000-01348, CONI- CET-PEI-N° 6429. We thank two anonymous re- viewers for giving constructive comments on ear- lier versions of the manuscript. Auswahl und Bevorzugung von Nahrung bei Ctenomys talarım (Tuco-Tuco) In der Studie wurde die Auswahl und die Bevorzugung von Futterpflanzen bei C. talarum unter- sucht. Um die Auswahl zu schätzen wurde die oberirdische Biomasse der Futterpflanzen bestimmt, und die Zusammensetzung der Nahrung im Mageninhalt mittels mikrohistologischer Technik geschätzt. Die Bevorzugung von Futterpflanzen wurde mit Cafeteria-Experimenten im Laboratorium untersucht. C. talarum verhielt sich als ein Generalist und Opportunist durch Nutzung der meisten Arten, ändert die Nahrungswahl aber gemäß Verfügbarkeit. Oberirdische Pflanzenteile wurden ge- genüber unterirdischen Teilen oder Blüten bevorzugt. Gräser bildeten 94% der oberirdischen Teile. Die Nahrung vom Tuco-Tuco variierte zwischen den Geschlechtern. Männchen verhielten sich selek- tiver als Weibchen. Die Cafeteria-Experimente zeigten auch, daß C. talarum oberirdische Teile der Gräser gegenüber den übrigen bevorzugt. References ANDERSEN, D. C. (1987): Geomys bursarius bur- rowing patterns influence of season and food patch structure. Ecology 68, 1306-1318. BELovskKY, G. E.; SCHMITZ, ©. J. (1994): Plant de- fenses and optimal foraging by mammalian herbivores. J. Mammalogy 75, 816-832. BRANCH, L. C.; VILLARREAL, D.; SBRILLER, A. P; Sosa,R. A. (1994): Diet selection of the plains vizcacha (Lagostomus maximus, family Chinchillidae) in relation to resource abun- dance in semiarid scrub. Can. J. Zool. 72, 2210-2216. BRETT, R. A. (1991): The ecology of naked Mole- Rat colonies: burrowing, food, and limiting factors. In: The Biology of the Naked mole- rat. Ed. by P. W. SHERMAN, J. U. M. JARvIs, and R. D. ALEXANDER. Princeton. New Jersey: Princeton University Press. Pp. 137-184. BUJALSKA, G. (1983): Dynamics and regulation of the population in ecology of the bank vole. Acta Theriol. 28, 148-160. BuschH;, C.; ANTINUCHI, ED DEWVABPREAIRGE KITTLEIN, M. J.; MALIZIA, A. I.; VASALLO, A. ].; ZENUTO, R. R. (2000): Population ecology of subterranean rodents. In: Biology of Subterra- nean Rodents. Ed. by E. A. Lacey, J.M. PAT- Ton, and G.N. CAMERON. Chicago, Illinois: Chicago Press. Pp. 183-226. COMPARATORE, V. M.; CiD, M. S.; Busch, C. (1995): Dietary preferences of two sympatric subter- ranean rodent populations in Argentina. Rev. Chil. Hist. Nat. 68, 197-206. Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco) 173 COMPARATORE, V. M.; MACEIRA,N. O©.; BUSCH, C. (1991): Habitat relations in Ctenomys talarum (Caviomorpha, Octodontidae) in a natural grassland. Z. Säugetierkunde 56, 112-118. GOERING, H. K.; van SOEST, P. J. (1970): Forage fi- ber analyses (apparatus, reagents, procedures, and some applications). Agric. Handbook 379. Washington, D. C: Agriculture Research Service, United States Department of Agri- culture. GORNALL, A.C; BARDWILL, C.J.;, DavıDp, M.M. (1949): Determination of serum protein by means of biuret reaction. J. Biol. Chem. 177, 751-766. HETH, G.; GOLEMBERG, E.M.; NEvo,E. (1989): Foraging strategy in a subterranean rodent, Spalax ehrenbergi. A test case for optimal foraging theoly. Oecologia 79, 617-622. HOLECHEK, J. L; VAVRA, M.; PIEPER, R. D. (1984): Methods for determinating the botanical composition, similary and overlap of range herbivore diets. In: Developing Strategies for Rangelands Managements. Ed. by National Research Council, National Ac. Science. Boulder: Westview Press. Pp. 425-469. ILLıus, A. I.; GoRDoNn, I. J. (1993): Diet selection in mammalian herbivores: constraints and tac- tics. In: Diet Selection. An Interdisciplinary Approach to Foraging Behavior. Ed. by R.N. HucHes. Oxford: Blackwell Scient. Publ. Pp. 157-181. JUSCAFRESA, B. (1975): Enceiclopedia ilustrada Flora medicinal, töxica, aromätica, condimen- ticia. Aedos, Barcelona: Editorial Aedos. KRUEGER, W. C. (1972): Evaluating animal forage preference. J. Range Manage. 25, 471-475. MADOERY, L. A. (1993): Composiciön botänica de la dieta de tuco-tuco (Ctenomys mendocinus) del piedemonte precordillerano. Ecologia Austral 3, 49-55. MARTINO, N. S. (2000). Respuestas nutricionales a corto plazo en Ctenomys talarum (Rodentia, Octodontidae). Tesis de Licenciatura, Univer- sidad Nacional de Mar del Plata, Buenos Aires. Argentina. NEvo, E. (1979): Adaptative convergence and di- vergence of subterranean mammals. Ann. Rev. Ecol. Syst. 10, 269-308. Nevo, E. (1991): Evolutionary theory and process of active speciation and adaptive radiation in subterranean Mole Rats, Spalax ehrenbergi superspecies, in Israel. Evol. Biol. 25, 1-125. NORBURY, G. L.; SANSoN, G. D. (1992): Problems with measuring diet selection of terrestrial, mammalian herbivores. Aust. J. Ecol. 17, 1-7. PHILLIPSON, J.; SARRAZIN COMANS, M.; STOMATO- POULOS, C. (1983): Food consumption by Mi- crotus agrestis and the unsuitability of faecal analysis for the determination of food prefer- ence. Acta Theriol. 28, 397-416. Reıc, ©. A. (1970): Ecological notes on the fossor- ial rodent Spalacopus cyanus (Molina). ). Mammalogy 51, 592-601. RESZUTEK, M. J.; CAMERON, N. G. (1998): Influ- ence of resource removal on demography of attwater’s pocket gopher. J. Mammalogy 79, 538-550. SPARKS, D.; MALECHEK, J. C. (1968): Estimating percentage dry weight in diets using a micro- scopic technique. J. Range Manage. 21, 264- 265. STEPHENS, D. J.; KREBS, J. R. (1986): Foraging the- ory. In: Monographs in Behavior and Ecology. Ed. by J.R. Kress and T. CLUTTON-BROocK. Princeton. New Jersey: Princeton University Press. STUEBE, M. M.; ANDERSEN, D.C. (1985): Nutri- tıonal ecology of a fossorial herbivore: protein N and energy value of either caches made by northern pocket gophers, Thomomys_tal- poides. Can. J. Zool. 63, 1101-1105. VLECcK,D. (1979): The energy cost of burrowing by the pocket gopher Thomomys bottae. Phy- siol. Zool. 52, 122-136. WILLIAMS, L. R; CAMERON, G.N. (1986): Food ha- bits and dietary preferences of Attwater’s pocket gopher Geomys attwateri. J. Mammal- ogy 65, 216-265. WILLIAMS, O. B. (1969): An improved technique for ıdentification of plant fragment in herbi- vore faeces. J. Range Manage. 22, 51-52. Woops, CH. A. (1984): Histricognath rodents. In: Orders and Families of Recent Mammals of the World. Ed. by S. AnDERSON and J. K. Jones. New York: John Wiley. Pp. 389- 446. ZAR,J.H. (1984): Biostatistical Analysis. Inc. Englewood Clıffs. New Jersey: Prentice-Hall. Authors’ addresses: JUANA C. DEL VALLE, MARIA I. LOFHELT, VIVIANA M. COMPARATORE, and CRISTINA BUSCH, Departa- mento de Biologia, Universidad Nacional de Mar del Plata, C.C. 1245, Mar del Plata, (7600), Pro- vincia de Buenos Aires, Argentina, (e-mail: delvalle@mdp.edu.ar); SıLvıa Cıp, Facul- tad de Ciencias Agrarias, Universidad Nacional de Mar del Plata, C.C. 276, (7629) Balcarce, Pro- vincia de Buenos Aires, Argentina. Mamm. biol. 66 (2001) 174-177 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication Bullate stapedes in Mammalian Biology Zeitschrift für Säugetierkunde some phalangeriform marsupials By M. R. SÄNCHEZ-VILLAGRA and SIRPA NUMMELA Zoologisches Institut, Universität Tübingen, Tübingen, Germany and Department of Ecology and Systematics, University of Helsinki, Finland. Receipt of Ms. 06. 10. 2000 Acceptance of Ms. 12. 12. 2000 Key words: Marsupialia, possums, ossicles, stapes, ear Stapes form varies considerably among mammals and has been a disputed topic in morphology-based mammalian systematics (NovAcEK and Wyss 1986; RosE and EMRY 1993; GAuDin et al. 1996). One particular specialization that has received recent atten- tion (WILkins et al. 1999) is the bullate form of the stapes’ footplate. A bullate stapes pos- sesses ‘a highly convex hollow footplate that protrudes into the vestibule of the inner ear’ (Wırkins et al. 1999), instead of being flat or nearly flat like in most mammals. This situa- tion was first reported by Hyrrı (1845) for the common ring-tailed possum Pseudo- cheirus peregrinus (= 'Phalangista cooki’, Pe- tauridae, Marsupialia). Subsequent to this work, other authors have described this ana- tomical specialization in several phylogeneti- cally distant eutherian mammals (DoRAN 1878; SEGALL 1971; BURDA et al. 1992; WIL- Kıns etal. 1999 and references therein). Con- trary to the statements of Wirkıns et al. (1999), P. peregrinus is not the only marsu- pial showing a bullate stapes. SEGALL (1971) reported (but did not illustrate) this for the feathertail glider, Acrobates pygmaeus (Ac- robatidae, Diprotodontia). During the course of our studies on the evo- lution of ear ossicles in marsupials, we ex- amined the stapes in more than 70 speci- mens representing 26species in eight 1616-5047/01/66/03-174 $ 15.00/0. ‘families’. In all cases the stapedial footplate was flat and not bullate, with the following three exceptions (Fig. 1): the brush-tailed opossum, Trichosurus vulpecula (n= 13); the grey cuscus, Phalanger orientalis (n = 2); and the spotted cuscus, Spilocuscus macula- tus (n= 1). Of these three taxa, 7! vulpecula shows this feature most marked, followed by S. maculatus. In T. vulpecula, the depth of the footplate equals that of the crural portion of the stapes, while in the other two taxa the proportion is smaller. Some other marsupial taxa in addition to those mentioned above have a somewhat bullate stapes. SEGALL (1971: 34) reported that in Petaurus norfolcensis ‘the vestibular surface of the plate is only slightly convex. FLEISCHER (1973: 142) noted in his descrip- tion of the stapes of Petaurus breviceps that ‘...seine Basis ist geringfügig ins Vestibu- lum vorgewölbt. The condition in these species of Petaurus approximates that de- scribed here for Phalanger orientalis, as confirmed by examination of a specimen of Petaurus breviceps (SM-64418). Several eu- therians have a convex footplate that ap- proximates the bullate condition, e.g. Sus and Cynocephalus (DoRrAN 1878; Rose and Emry 1993). These cases illustrate well the fact that the definition of a bullate stapes is to some extent a matter of evaluation. Bullate stapedes in marsupials 175 Fig. 1. Stapes of left) Trichosurus vulpecula (WM-pers.coll.) center) Phalanger orientalis (SM-54981) and right) Spilocuscus maculatus (SM-5610). Scale = 0.5 mm. All the marsupial taxa for which a bullate stapes is reported here and elsewhere are phylogenetically close and taxonomically ordered within the Phalangeriformes (KırscH et al. 1997). The stapedes of other members of this group were studied by SE- GALL (1971), including Pseudocheirus her- bertensis, Petauroides volans, and Dactylop- sila trivirgata, and in no case did this author mention any peculiarity in their stapes. Plotting the distribution of bullate stapedes in the phylogenetic tree of Phalan- geriformes based on DNA-hybridization studies by Kirsch et al. (1997), it is obvious that the bullate condition (at least in its marked form) has either evolved independ- ently in several taxa, or has been lost inde- pendently if present in the last common ancestor of Acrobates and the other Phalan- geriformes. In addition to the adult macerated skulls, we examined histologically prepared speci- mens of several South American and Aus- tralasian marsupial taxa. Most species are represented by pouch-youngs, in some cases complete developmental series were exam- ined (for a complete list, see SANCHEZ-VIL- LAGRA 2001). Among the species showing bullate stapes as adults, 7. vulpecula was re- presented by two specimens. An early pouch-young of Trichosurus vulpe- cula shows already a prominently outbulging footplate of the stapes that protrudes into the inner ear (Fig. 2), a condition that persists in the adult. Of all other taxa examined, only an early pouch-young of the eastern quoll, Da- syurus viverrinus also shows this condition. Adults of this species, as well as other adults of the Dasyuromorpha (ArcHER 1976) do not show this feature. For comparison, a pouch- young of Perameles sp. with the plesiomorphic marsupial condition of the stapes’ footplate is shown in figure 2. In the specimens illustrated, the ear ossicles are in a blastemous, pre-carti- laginous stage. Much remodeling and growth takes place in the ear ossicles between these stages and adulthood. The eutherians showing the most pro- nounced bullate stapes are rodents belong- ing to the Heteromyidae and Geomyidae, with highly derived middle ears and special- ized to low-frequency hearing. Of all mar- supials possessing bullate stapedes, only for Trichosurus vulpecula there has been an (electrophysiological) audiogram published (GATEs and Aıtkın 1982). Even though T. vulpecula does not have sımilar hearing abilities to those of the desert rodents men- tioned above, an interesting departure from the few other marsupials (phylogenetically 176 M. R. SÄNCHEZ-VILLAGRA et al. Fig. 2. Cross sections of a portion of the right middle ear of left) Trichosurus vulpecula (ZSH, HL = 7.5 mm) and right) Perameles sp. (ZSH, HL=17.5 mm). m = malleus, i = incus, s = stapes. The arrow indicates the bullate con- dition of the stapes. Not to scale. and ecologically disparate) for which audio- grams are available can be noticed. As pointed out by Aıtkın (1995), T. vulpecula is more sensitive over a wide range to low frequencies than the other marsupials. Based on the distribution of the bullate stapes among mammals, it appears that there is no obvious correlation between the possession of a bullate stapes and any parti- cular habit or ecology. A wide size-range is represented by the marsupial species show- ing a bullate stapes, from the 10-17 g Acro- bates to the much larger Trichosurus reach- ing around 4.5kg (Nowak 1999). They include mostly arboreal species, omnivor- ous-herbivores and predominantly nectar- eaters (HUME 1999). In summary, we report here the presence of a singular specialization of the stapes ın three marsupial taxa. Based on the study of pouch-youngs of one of them, we ob- serve that this feature appears relatively early in ontogeny. A bullate stapes repre- sents either an autapomorphy of Phalangeri- formes lost independently in several mem- bers of this monophylum, or characterizes several clades within this group of diproto- dontian marsupials. Acknowledgements Osteological specimens reported in this study are deposited in the Senckenberg Museum, Frankfurt (SM) or belong to WOLFGANG MAIER’s personal collection (WM). Histological specimens belong to the Lehrstuhl für Spezielle Zoologie, Zoolo- gisches Institut, in Tübingen (ZSH). We thank the following persons and institutions for allowing us to examine collections un- der their care: G. STORCH (SM), A. FORSTEN (Zoological Museum, Helsinki), S.B. MCLAREN and J. R. WıßgLE (Carnegie Mu- seum), J. A. W. KırscH (University of Wis- consin), K.K.SmitH and A. van NIEVELT (Duke University, Durham), and W. MAIER (Tübingen). We thank R. Brıtz for his help with photographic equipment and proce- dures, W. MAIER (Tübingen), T. REUTER and S. HemııÄ (Helsinki) for their support, and W. MAIER and an anonymous reviewer for their helpful criticisms to the manuscript. Ella and Georg Ehrnrooth Foundation, and Oskar Öflund Foundation supported SN. A research travel grant from DFG sup- ported MRSV. References AITKIN, L. (1995): The auditory neurobiology of marsupials: a review. Hear. Res. 82, 257-266. ARCHER, M. (1976): The basicranial region of marsupicarnivores (Marsupialia), interrela- tionships of carnivorous marsupials, and affı- nities of the insectivorous marsupial perame- lids. Zool. J. Linn. Soc. 59, 217-322. BURDA, H.; BRUNS, V.; HICKMAN, G. C. (1992): The ear in subterranean Insectivora and Ro- dentia in comparison with ground-dwelling representatives I. Sound conducting system of the middle ear. J. Morphol. 214, 49-61. DORAN, A.H.G. (1878): Morphology of the mammalian ossicula auditus. Transact. Lin- nean Soc. London 1, 371-497. FLEISCHER, G. (1973): Studien am Skelett des Gehörorgans der Säugetiere, einschließlich des Menschen. Säugetierkdl. Mitt. 21, 131- 239) GATES, G. R.; AITKIN, L. M. (1982): Auditory cor- tex in the marsupial possum Trichosurus vul- pecula. Hear. Res. 7, 1-11. GAUDIN, T. J.; WIBLE, J. R.; HOPSON, J. A.; TURN- BULL, W. D. (1996): Reexamination of the morphological evidence for the cohort Epitheria (Mammalia, Eutheria). J. Mammal. Evol. 3, 31-79. HUME, I. D. (1999): Marsupial Nutrition. Cam- bridge: Cambridge University Press. HYRTL, J. (1845): Vergleichend-anatomische Un- tersuchungen über das innere Gehörorgan des Menschen und der Säugethiere. Prag: Friedrich Ehrlich. KIRSCH, J. A.W.; LAPOINTE, F.-J; SPRINGER, M.S. (1997): DNA-hybridisation studies of marsupials and their implications for me- tatherian classification. Austral. J. Zool. 45, 211-280. Bullate stapedes in marsupials 177 NOVACEK, M.J.; Wyss, A. (1986): Origin and transformation of the mammalian stapes. In: Vertebrates, Phylogeny and Philosophy. Ed. by K. M. FLANAGAN and J. A. LILLEGRAVEN. Con- trib. Geol. Univ. Wyo. Spec. Paper 3, Pp. 35-53. NOWAK, R.M. (1999): Walker’s Mammals of the World. 6 Ed. Baltimore, London: Johns Hop- kins University Press. ROSE, K. D.; EMRY, R. J. (1993): Relationships of Xenarthra, Pholidota, and fossil “edentates’: the morphological evidence. In: Mammal Phylogeny. Ed. by F.S. SZALAY, M. J. NOVA- CEK, and M. C. McKenna. New York: Sprin- ger Verlag. Vol. 2, 81-102. SANCHEZ-VILLAGRA, M.R. (2001): Ontogenetic and phylogenetic transformations of the vo- meronasal complex and nasal floor elements in marsupial mammals Zool. J. Linn. Soc. 131, (in press). SEGALL, W. (1971): The auditory region (ossicles, sinuses) in gliding mammals and selected re- presentatives of non-gliding genera. Fieldiana Zool. 58, 27-59. WIrKINSS KR 1. ZROBERITS, IC 7ROORDA, CS; HAWKINS, J. E. (1999): Morphometrics and functional morphology of middle ears of ex- tant pocket gophers (Rodentia, Geomyidae). J. Mammalogy 80, 180-198. Authors’ addresses: MARCELO R. SANCHEZ-VILLAGRA, Spezielle Zoologie, Zoologisches Institut, Universität Tü- bingen, Auf der Morgenstelle 28, D-72076 Tübin- gen, Germany (e-mail: marcelo.sanchez@uni-tuebingen.de); SIRPA NUMMELA, Department of Ecology and Systematics, University of Helsinki, P.O. Box 17, FIN-00014, Helsinki, Finland. Mamm. biol. 66 (2001) 178-180 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short Communication . Mammalian Biology Zeitschrift für Säugetierkunde Twinning in the big fruit-eating bat Artibeus lituratus (Chiroptera: Phyllostomidae) from eastern Paraguay By R. D. STEVENS Department of Biological Sciences, Texas Tech University, Lubbock, Texas, USA Receipt of Ms. 20. 09. 2000 Acceptance of Ms. 17. 11. 2000 Key words: Artibeus lituratus, twinning, reproduction, Paraguay Although the litter size of bats is variable and ranges from one to five (HAMILTON and STALLING 1972), multiple embryos in Ameri- can leaf-nosed bats are rare and have been reported for only a few species. Twinning in the Phyllostomidae was first reported for Macrotus waterhousii by CoCKRUM (1955) and then by BRADSHAw (1961). BARLOW and Tamsıtr (1968) later reported twinning in three additional species: Glossophaga sorici- na, Erophylla sezekorni, and Artibeus jamai- censis. Herein, I report twinning in A. litura- fus. Artibeus lituratus is widely distributed geo- graphically, ranging from northern Mexico to northern Argentina (KoopMmAn 1993). This species exhibits considerable geo- graphic variation regarding color, morphol- ogy, diet, and reproductive patterns (BAKER et al. 1976, 1977, 1979). Wırson (1979), based on extensive data, suggested that re- productive patterns in this species are geo- graphically varıable, ranging from monoes- try at the northern limit of its range to bimodal polyestry (THuomAs 1972) and acyc- lic breeding (TAamsıtt and VALDIVIESO 1963, 1965; TAamsıtT 1966) in Colombia. Subse- quently, Wırrıs (1985) demonstrated that A. lituratus exhibits seasonal bimodal poly- estry in northeastern Brazil. SazımA (1989) demonstrated that the timing of reproduc- 1616-5047/01/66/03-178 $ 15.00/0. tion is dynamic in this species and depen- dent on weather patterns and primary pro- ductivity. Although patterns of reproduc- tion are well documented, no report of twinning in this species currently exists. Of 864 female A. liruratus collected and ne- cropsied in this investigation, I encountered one gravid female containing two embryos. The female was caught on 29 December 1997 at Yaguarete Forests, located approxi- mately 40 kilometers due east of the town of Santa Rosa de Lima in the department of San Pedro in eastern Paraguay (23° 48.50 'S, 56° 07.68’ W). The twins consisted of one male and one female. Accordingly, they were likely the result of fertilization of two separate ova. They were 11.6 mm and 11.3 mm in length, respectively. Tooth- wear on the mother was relatively slight and she was post lactating, suggesting that she was relatively young in age but had pre- viously produced offspring. Several explanations have been put forth to account for the paucity of instances of twin- ning in the Phyllostomidae. BArRLOw and TAamsıtr (1968) suggested that differences in litter size between vespertilionid and phyllostomid taxa exist because these groups have evolved in or radiated from areas that differ in seasonality and the length of growing seasons. They suggest that temperate vespertilionids should have larger litters because they have a protracted period in which to produce offspring, whereas smaller litters are facilitated in tro- pical phyllostomids by more constant avail- ability of resources. Moreover, phylogenetic as well as mechanical constraints likely maintain the single embryo condition in phyllostomid species. This is supported by the observation that phyllostomid fetuses attain relatively larger size than members from most other families of bats, and multi- ple embryos likely would cause overly great mechanical and physiological strain on the mother (WimsAtT and TrAPıDo 1952). Finally, TAppeEı (1976) suggested that me- chanisms operating during ovulation lim- ited the number of ova released from folli- cles of females of this species. He found that more than one oocyte per ovarian folli- cle (suggestive of the potential for twin- ning) was not uncommon yet none of the individuals examined contained more than a single embryo. These observations com- bine to suggest that twinning is a rare phe- nomenon that results from accidents during References BAKER, R. J.; JONES, Jr, J. K.; CARTER, D. C. (1976): Biology of Bats of the New World Family Phyllostomatidae. Part I. Lubbock: Texas Tech Univ. Press. BAKER, R. J.; JONES, Jr, J. K.; CARTER, D. C. (1977): Biology of Bats of the New World Family Phyllostomatidae. Part II. Lubbock: Texas Tech Univ. Press. BAKER, R. J.; JONES, Jr, J. K.; CARTER, D. C. (1979): Biology of Bats of the New World Family Phyllostomatidae. Part III. Lubbock: Texas Tech Univ. Press. BARLOWw, J. C.; TAMSITT, J. R. (1968): Twinning in American leaf-nosed bats (Chiroptera: Phyllostomatidae). Can. J. Zool. 46, 290- 2% BRADSHAW, G. V.R. (1961): Le cycle de reproduc- tion de Macrotus californicus (Chiroptera: Phyllostomatidae). Mammalia 25, 117-119. CockRuM, E.L. (1955): Reproduction in North American bats. Trans. Kansas Acad. Sci. 58, 487-511. Twinning in Artibeus lituratus 179 ovulation or development. Moreover, twin- ning in the Phyllostomidae likely is a condi- tion that is selected against because of its deleterious effects on the mother. Finally, of the group of Phyllostomid species that exhibit twinning, no phylogenetic or ecolo- gical pattern exists regarding which species should exhibit this condition. This suggests that the phenomenon of twinning, although rare, should be expected from any large col- lection of phyllostomid bats. Acknowledgements L. GiMENEZ and H. STEVENS provided field and la- boratory assistance. Research assistantships were provided by M. Wırrıs and Texas Tech Univer- sity. I. GAMMARA DE Fox and the National Mu- seum of Natural History of Paraguay provided lo- gistical assistance. PAUL MUELLER and YAGUARETE ForESTS allowed field work to be conducted on their property. CELIA LOPEZ-GONZALEZ reviewed a previous version of this manuscript. I am in- debted to each of these people and institutions for their contributions. HAMILTON, R.; STALLING,D.T. (1972): Lasiurus borealis with five young. J. Mammalogy 53, 190. KoopMAn, K.F. (1993): Order Chiroptera. In: Mammal Species of the World: A Taxonomic and Geographic Reference. Ed. by D. WıLson and D. REEDER. Washington: Smithsonian Inst. Press. Pp. 131-241. TAmSITT, J. R. (1966): Altitudinal distribution, ecology, and general life history of bats in the Andes of Colombia. Amer. Phil. Soc. Year- book. Pp. 372-373. TAMSITT, J. R.; VALDIVIESO, D. (1963): Reproduc- tive cycles of the big fruit-eating bat, Artibeus lituratus Olfers. Nature 198, 104. THoMAs, M.E. (1972): Preliminary study of the annual breeding patterns and population fluc- tuations of bats in three ecologically distinct habitats in Southwestern Colombia. Diss. Tu- lane Univ., New Orleans. WirLıg, M.R. (1985): Reproductive patterns of bats from Caatingas and Cerrado biomes in 180 R. D. STEVENS Northeast Brazil. J. Mammalogy 66, 668- 681. Wirson, D. E. (1979): Reproductive patterns. In: Biology of Bats of the New World Family Phyllostomatidae. Part III. Ed. by R. BAkER, J. K. Jones, Jr., and D. CARTER. Lubbock: Texas Tech Univ. Press. Pp. 317-378. Wimsatt, W. A.; Trapido, H. (1952): Reproduction and the female reproductive cycle in the trop- ical America vampire, Desmodus rotundus mu- rinus. Amer. J. Anat. 91, 415-445. Authors adress: RiCHARD D. STEVENS, Department of Biological Sciences, Texas Tech University, Lubbock, Texas, 79409-3131, USA (e-mail: cmrds@ttacs.ttu.edu) Mamm. biol. 66 (2001) 181-184 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication eg SA “ Mammalian Biology DO & 52 TTIERKUND“ Zeitschrift für Säugetierkunde Questionable status of the "Taynguyen civet”, Viverra tainguensis Sokolov, Rozhnov and Pham Trong Anh, 1997 (Mammalia: Carnivora: Viverridae) By J. WaLsToN and GERALDINE VERON Wildlife Conservation Society, Cambodia Program, Phnom Penh, Cambodia and Laboratoire Mammiferes et Oiseaux, Museum National d’Histoire Naturelle, Paris, France Receipt of Ms. 04. 12. 2000 Acceptance of Ms. 12. 01. 2001 Key words: Viverra tainguensis, Vietnam, systematics, morphology CorBET and Hırr (1992) recognized four species of Viverra Linnaeus in the Indoma- layan Region. Two of these are known from Vietnam: the large Indian civet, V. zibetha, and the large-spotted civet, V. megaspila (OscoopD 1932; Dans Huy Huyna et al. 1994). In 1997, SOKOLoVv et al. described the “Taynguyen civet”, Viverra tainguensis, from Vietnam. The description was based on characters of the holotype only, a subadult male. A paratype was designated but its characters were not used in the description. The authors state that they had examined 46 V. zibetha from Vietnam, four V. zibetha from China, two V. megaspila from Vietnam, and eight V. fangalunga from Indonesia and the Philippines. In 1999, RozuHnov and PHAM TRONG AnH assigned an additional five spe- cimens to V. tainguensis and later contribut- ed to another publication detailing addi- tional morphometric parameters of two of the specimens (SoKkoLov et al. 1999). Although the present authors have not ex- amined the holotype, which remains at the Zoological Museum of the Moscow State University (ZMMU), the original descrip- tion of V. fainguensis and both subsequent publications contain a number of factual er- rors and questionable interpretations that cast doubt on the validity of the supposed 1616-5047/01/66/03-181 $ 15.00/0. new species. However, since V. E. SOKOLOV died in early 1998, ıt is not clear to what ex- tent he was involved in the latter publication authored by SoKoLov et al. (1999). Pocock (1939) recognised, as full genera, Viverra, Viverricula and Moschothera. Vi- verra megaspila and V. civettina were placed in Moschothera, which was distinguished from Viverra, in the sense of Pocock, by the absence of sheaths of skin covering the claws of the 3" and 4" digits of the fore- feet. This feature was clearly described and well-figured by PococKk. CorRBET and HırL (1992) placed Moschothera as a synonym of Viverra, but regarded the claw sheathing as an important diagnostic character for dis- tinguishing V. zibetha and V. tangalunga from their congeners. Although the claw sheathing is present in V. tangalunga, this species is restricted to the Sundaic subre- gion and is not known from the Indochinese subregion. In CorgErt and Hırr’s (1992) table 138, the presence or absence of sheathing in each species of Viverra is indicated by a “+” or “o”, respectively, for all species except for V. tangalunga. However, as confirmed by CoRBET (pers. comm.), a typographical er- ror has resulted in the symbols being re- versed. The absence of supporting text or 182 J. Wauston et al. illustrations prevents this error from being easily detected. The remainder of the ta- ble agrees with Pocock’s (1939) find- ings. The most consequential error in the de- scription of V. tainguensis by SOKOLoOV et al. (1997) relates to the confusion over the sheathing of the front claws. Repeating the error of CorBET and Hırr (1992), SOKOLOV et al. d997N)T aklirmedrithez presence of sheathing in V. megaspila, and its absence in V. zibetha, exactly the reverse of the si- tuation found in nature. This error is made repeatedly; the incorrect, reversed, charac- ter are said to have been observed in speci- mens examined; and they are used as the first and most important diagnostic features distinguishing their V. tainguensis from V. zibetha. In order to check the status of the sheathing on V. zibetha, specimens from the Natural History Museum, London (BMNH), were examined. In addition, the 28 specimens at the Institute of Ecology and Biological Resources (IEBR), Hanoi, said to have been examined by SOKOLOV et al. (1997), in the course of their descrip- tion of V. tainguensis, were re-examined (R. J. Timmins, pers. comm.). All specimens conformed to the situation as described by Pocock (1939), rather than that as given by SoKOLoV et al. (1997), with respect to the sheathing. Perhaps the characters as given in table 138 by CorgeEr and Hırr (1992) were simply accepted by SoKoLov et al. (1997), earlier publications were not read carefully, and actual characters of speci- mens of V. zibetha were not ascertained but were merely assumed to be as given by CorBET and Hırır (1992). In any event, if the anımals ascribed to V. tainguensis have sheathed claws, then this trait would be shared between them and specimens prop- erly ıdentified as V. zibetha, rather than being one to suggest a specific distinction between the two. The second supposedly distinguishing fea- ture given by SoKkoLov et al. (1997) for V. tainguensis was body size, which was said to be less than that in V. zibetha. However, the holotype of V. tainguensis is subadult with a head-body length of 600 mm (SoKo- rov et al. 1997) and head-body lengths (of 790 and 780 mm) have been provided for only two additional specimens, both adults (Rozunov and PHAM TRoNG AnH 1999). These measurements are well within the known range of 740-860 mm for V. zibetha (CorBET and Hırr 1992). Although Pocock (1939) was cited by SoKoLov et al. (1997), they made no mention of adult specimens of V. zibetha that Pocock examined from northeastern India, Nepal, and Myanmar, and which had head-body lengths of 742- 863 mm. In addition, THomAs (1927) de- scribed a subspecies of V. zibetha (V. z. sur- daster) from northern Laos and central and southern-central Vietnam; the last locality being less than 50 km from the type locality of V. tainguensis. SOKOLOV et al. (1997) and RoZHNnov and PHAM TRONG AntH (1999) did not mention THomas’s V.z. surdaster, although it was listed by CorBET and HırrL (1992). THoMAs (1927) described V. z. sur- daster as “averaging rather smaller than true Indian zibetha”, and noted further that “among the variable races of the... Indian, civet the form may be distinguished by its comparatively small size and especially by its small bullae”. The condylobasal length of the only existing adult skull of V. tain- guensis is 132.5 mm (SoKoLov et al. 1999), whilst that of the type of V. z. surdaster measures 128 mm, and specimens measured by Pocock (1939) range from 129-135 mm for V. z. pruinosa. The specimens that THo- MAS assigned to V. z. surdaster are clearly important in assessing the valıdıty of V. tainguensis and it appears that fainguen- sis cannot be distinguished from surdaster based on measurements. Certainly the body and skull sizes given for V. tainguensis fall within the range of those known for V. zi- betha, and in no way argue for the specific distincetness of the former. The supposed third distinguishing feature of V. tainguensis given by SOKOLoV et al. (1997) was relative tail length. The tails of the se- ven specimens of V. tainguensis, were Te- ported to average 52% of the head-body length, proportionately smaller than the 55-60% given for V. zibetha. However, the mean tail to head-body length of the adult Questionable status of the “Taynguyen civet” V. zibetha examined by Pocock (1939) was 53.5%, hardly different from that given for the V. tainguensis. The supposed difference becomes even less significant when one considers that the holotype of V. fainguensis is a subadult, that measurements of the adult paratype are not included in the de- scription, and the only two specimens of V. tainguensis with accompanying morpho- metric data had tail to head-body length ra- tios of 53% and 56% (RozHnov and PHAM TrongG Anh 1999). Furthermore, neither publication dealing with V. fainguensis gave relative tail lengths for any V. zibetha speci- mens examined. Both merely quoted the figure from CorsBEr and Hırı, (1992). Clearly, no convincing evidence has been presented to show that relative tail length can be used to distinguish V. zibetha from a second species to be known as V. fainguen- sis. (It should also be noted that SoKOLOV et al. (1997) stated that V. megaspila has a tail 45-55% of the length of its head-body. They appear to mean °‘V. Tangalunga', which, according to CorBET and HirL (1992), does have a tail 45-55% of its head-body length, whereas in V. megaspila it is 30-50%.) A fourth supposed distinguishing feature of V. tainguensis was stated to involve the pe- lage colour pattern. SoKoLov et al. (1997) cited CorBET and Hırr (1992) as stating that V. tangalunga, V. zibetha, and V. mega- spila show little variation in their pelage pattern. Later, Rozunov and PHAM TRONG Ann (1999) cited the same source to sup- port their contention that “Weak variation in external morphology is typical for all species of genus Viverra”. These claims are incorrect. The first claim holds for V. tanga- lunga, but not for V. megaspila and it also involves a misinterpretation of a statement concerning V. zibetha. CorBET and HiırL (1992) mentioned “little regional variation” in V. zibetha. This clearly refers to inter-re- gional, rather than intra-regional, variation. There is clearly a considerable degree of variation, both in pelage colouration and other characteristics in V. zibetha. The de- scription of V. z. surdaster states “colour, as usual, variable, but with less tendency 183 to definite markings on the flanks and hips” (Tmuomas 1927). OsGooD (1932) stated that the species is “variable” and that “doubtless there are several recognizable races.” Pocock (1939), wrote “In V.zi- betha... the coat, colour, and pattern vary considerably... The body-pattern is strongly pronounced in summer, indistinct or even obliterated in winter; and the ground-colour varies individually, even ir- respective of season, from tawny to clear, almost silvery-grey... The differences ... in colour and pattern, now known to be in- dividual and... seasonal, account for the number of names applied to most of the lo- cal races of this civet.” Examination of ex- isting specimens from south-east Asia clearly reveals the variability of the pelage pattern of V. zibetha. Specifically, SOKOLOV et al. (1997) considered there to be three distinctive pelage features of V. tainguensis: the “semi-lunar” spots, the colour pattern of the fore and hind legs, and the light brown stripe running parallel to the crest. All three of these features are present se- parately in V. zibetha specimens at the Mu- seum National d’Histoire Naturelle (MNHN). The “semi-lunar” spots can be observed in combination with both, one or none of the other pelage characteristics claimed for V.tainguensis in specimens from Vietnam, and is also a characteristic of a V.zibetha specimen from China (CG 1962-156 at the MNHN). Another specimen from China (CG 1902-688) dis- plays the colour pattern on its legs but lacks the distinctive spots and lateral stripes, whilst a specimen from Vietnam (CG 1929-390, paratype of V. z. surdaster) has highly distinctive lateral stripes but lacks the colour pattern of the legs and any spots. Semi-lunar spots can also be ob- served on V. zibetha specimens of the BMNH from across the geographic range of the species. The pelage features stated to distinguish V. fainguensis will not sepa- rate this nominal form from all known indi- viduals of V. zibetha. In view of all of the above, insufficient evi- dence has been presented to suggest that V. tainguensis is in any way a distinct spe- 184 J. Wauston et al. cies. However, a proposal to synonymise V. tainguensis with V. zibetha would be pre- mature without an examination of the ho- lotype. Thus, we propose that all records of V. tainguensis, except possibly that of the holotype, be withdrawn and that a re- examination of the holotype be underta- ken. References CORBET, G. B.; HILL, J. E. (1992): The Mammals of the Indomalayan Region: a systematic re- view. London, Oxford: Natural History Mu- seum Publ. and Oxford University Press. DANG HUY HUYNH; DAO VAN TIEN; CAO VAN SUNG; PHAM TRONG ANH; HOANG MINH KHIEN (1994): Checklist of Mammals in Viet- nam. Hanoi: Vietnam National Centre for Natural Science and Technology. LEKAGUL, B.; MCNEELY, J. A. (1977): Mammals of Thailand. Bangkok: Association for the Conservation of Wildlife (as updated 1988). OSGOOD, W. H. (1932): Mammals of the Kelley- Roosevelts and Delacour Asiatic expeditions. Publ. Field Mus. Nat. Hist., Zool. Ser. 18, 193-339. POCOCcK, R. I. (1939): The Fauna of British India, Including Ceylon and Burma. Mammalia, 1. Primates and Carnivora (Felidae and Viver- ridae). London: Taylor and Francis, Ltd. ROZHNOV, V. V.; PHAM TRONG AnNH (1999): A note on the Tainguen civet — a new species of viverrid from Vietnam (Viverra tainguensis Acknowledgements The authors would like to thank Rog Timmins for the examination of the paratype and other speci- mens held at IEBR in Vietnam, and for his obser- vations that initiated this work. To Dr. NGUYEN XUAN Dan we are grateful for his translation of the original description of the “Taynguyen civet’”. We are especially indebted to Dr. Wırr Duck- WORTH and Dr. Ron Pine for their extensive com- ments and guidance on various drafts of the manuscript. Sokolov, Rozhnov and Pham Trong Anh, 1997). Small Carnivore Cons. 20, 11-14. SOKOLOV, V. E.; ROZHNOV, V. V.; PHAM TRONG ANH (1997): New species of viverrids of the genus Viverra (Mammalia, Carnivora) from Vietnam. Zool. Zhurnal 76, 585-589. SOKOLOV, V. E.; ROZHNOV, V. V.;, PHAM TRONG ANH (1999): New data on Viverra tainguensis Sokolov, Rozhnov et Pham Trong Anh, 1997 (Mammalia, Carnivora) from Vietnam. Zool. Zhurnal 78, 759-763. THOMAS, ©. (1927): The Delacour exploration of French Indochina -— Mammals. Proc. Zool Soc. London ‘1927’, 41-58. Authors’address: JoE WaLston, Wildlife Conservation Society, Cambodia program, PO Box 1620, Phnom Penh, Cambodia (e-mail: joe@bigpond.com.kh) and GERALDINE VERON, laboratoire Mammiferes et Oiseaux, Museum National d’Histoire Naturelle, 55, rue Buffon, 75005 Paris, France Mamm. biol. 66 (2001) 185-189 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication GESE! ES, N Sue SL ZN PEN 73 \/ 2 STERKUND“ Mammalian Biology Zeitschrift für Säugetierkunde Response of Apodemus flavicollis to conditions at the altitude limit in the Western Tatra Mountains By NATALIA MARTINKOVA, D. ZIAK, and [/. KOCIAN Department of Zoology, Comenius University, Bratislava, Slovakia Receipt of Ms. 10. 05. 2000 Acceptance of Ms. 09. 11. 2000 Key words: Rodentia, population dynamics, subalpine habitat The occurrence of a species at the distribu- tion border exhibits a dynamic pattern which sensitively reacts to the changing conditions of the environment as well as the state of the population of the species (BE- Gon et al. 1997). The distribution border is set by at least one environmental factor close to ıts limiting value. Regardless of the patchiness of suitable habitat within the dis- tribution range, the quality of habitat changes towards the distribution border from optimal through suboptimal to pessi- mal. Local populations adjacent to the dis- tribution border react to the changes of lo- cal conditions and become extinct or are recolonized (KoZAkıEwIıcz 1993; HANSKI et al. 1996) causing the distribution border to expand and contract (ANGELSTAM et al. 1987). The distribution border of Apodemus flavi- collis (MELCHIOR, 1834) includes a wide range of habitats and diverse climatic con- ditions due to its large distribution range (NIETHAMMER 1978; LurA et al. 1995). The local occurrence and altitude tolerance of the species seems to be affected by the re- quirements of a continental climate (LURA et al. 1995) and food abundance (AnGEL- STAM et al. 1987). The seed abundance ap- pears to be a critical factor affecting the dis- tribution and population dynamics of 1616-5047/01/66/03-185 $ 15.00/0. A. flavicollis. Populations are usually non- cyclic but density outbursts occur in years of high seed crop (GURNELL 1985; ANGEL- STAM et al. 1987; Pucek et al. 1993; ZıaK and Kocıan 1994). Considering habitat, A. flavicollis prefers mature deciduous forests with an open ground layer (GURNELL 1985). Therefore, the altitude range stretches from sea-level (Greece, Italy) to about 2000 m (Alps) (NIETHAMMER 1978; Yoccoz 1992). The aim of thıs study is to investigate the population response and habiıtat selection of A. flavicollis to conditions at an altitude border in an upper subalpine zone. The research was conducted in the Western Tatra Mts., Slovakia. The locality ın the Na- tional Nature Reserve Rohäce Lakes (ele- vation of the trapping grids: 1570-1600 m a.s.l.) represents a transition between a subalpine and an alpine zone due to cli- matic conditions caused by north-western orientation of the mountain range. The ha- bitat is characterized by scattered patches of Pinus mugo cover, the occurrence of wet subalpine meadows, and a talus gra- dient of various rock sizes partially over- grown with vegetation dominated by Juncus trifidus. Small mammal live-trapping was carried out in June, August, and October 1991-99 186 NATALTA MARTINKOVA et al. with the exception of 1992 and 1994 when trapping was carried out twice, in July and October. In the years 1991-95 two trapping grids were established, one 1 ha in size con- tainıng 10x10 live-traps, and the other 0.5 ha containing 6x8 live-traps, both with 10 m spacing. In 1996-99 another 1 ha trap- ping grid was added, and the trap layout was modified in the previous two grids giv- ing in total two I ha grids with 7x7 live- traps at 15 m intervals, one 1.3 ha grid con- taining 10x13 live-traps at 10 m intervals. The traps were baited with rolled oats, op- erated for 3-5 consecutive nights, and were checked twice daily. Animals were marked by toe-clipping, data on species, body weight, sex, reproductive status (scrotal testes, open vagina, gravidity, lactation), and body length were recorded. The habitat characterization was modified from DUESER and SHUGART (1978, 1979), and M’CLoskEy and FIELDWwIcK (1975). The habitat at each trap point was characterized for the years 1996-98. The habitat varıables were recorded in the summer series 1996- 97 ina circle (diameter 10 m, or 15 m) cen- tered on the trap. At each point, the propor- tion of the area covered by rocks, by rocks smaller and greater than 50 cm in diameter was registered. The vegetation structure was estimated by the proportion of the area covered by litter, herbs, shrubs and trees (in- cluding Pinus mugo), and specifically by dominant plant species: Juncus trifidus, Vac- cinium myrtillus, Pinus mugo, and grasses other than J. trifidus. The total number of plant species present at the sample was re- corded. The vegetation density below or above 50 cm was counted as the number of touches on a stick at 20 check-points form- ing a cross 10 + 10 centered on the trap and expressed as percent. The heights of herb, shrub and tree layers were measured at the same intersection depending on the avail- ability of the given layer. Finally, the dis- tance to the nearest patch of Pinus mugo larger than 30 m in diameter was recorded. Discriminant function analysis was used to explore the microhabitat preferences of A. flavicollis. A qualitative model was cho- sen where trap points used by at least one resident individual, defined by the time span between first and last capture being at least two days, were referenced against trap points not used by resident individuals. During the research time span (7200 trap- nights) 12 species of small ground mammals were registered: Sorex araneus, S. minutus, 5. alpinus, Neomys sp., Apodemus flavicollis, Clethrionomys glareolus, Microtus nivalıs, M. agrestis, M. tatricus, Muscardinus avella- narius, Mustela nivaliıs, and M. erminea (ZıaK and Kocıan 1994; N. MARTINKOVA, D. ZIAK, L’. Kocıan unpubl.). Apodemus flavicollis has only been captured in the years 1993, 1996, and 1998 (Fig. 1). A total of 44 indivi- duals was caught, 30 males and 14 females, a significant deviation from an expected sex ratio of 1:1 (x = 5.8, p = 0.016). Except for two individuals, all anımals were captured in one trapping series. The first exception was a sexually inactive female trapped in August and October 1993, and the second was a male trapped in August and October 1996, which was in breeding condition in August, but not so in October. No individuals were registered to stay the entire winter at the locality. Re- captures have also been rare with 60% of all individuals captured only one or two times, the average number of captures per indivi- dual being 2.3 and the maximum was eight (the sexually inactive female present in ser- ies VIIV93 and X/93). Individuals captured five or six times were all trapped in August 1996. No females were found demonstrating perforated vagina, or lactation, but one fe- male may have been pregnant in August 1993. Among males, twenty-two possessed scrotal testes. The discrimination of the preferred habitat of A. flavicollis was significant at p< 0.001. The discrimination model correctly classı- fied 77% of unused trap points and 76% of used trap points (Tab. 1). Habitat variables that possess the highest absolute values of standardized coefficients influenced to a greater extent the position of samples on the discriminant function axes. This means that the variation of these variables best de- scribes the differences between preferred and non-preferred habitat (LEGENDRE and LEGENDRE 1983). Variables associated with A. flavicollis in subalpine habitat 187 [77 Males RERI Females DD [@]} Ce, RS SER ...-.-2122.2 20.000211 202 11202020720 2 01000 0 01 0 u. Tu me nm mn mm un nm nn 0 ee AAMÄ- - = er nun nn nn nn nn nn nn nn nn —n nn ER ER —& © { ‘ ‘ ‘ ‘ ' ' ' ' ' ' ‘ ' ' ' ' \ ' U D ‘ D ‘ ‘ ‘ \ ' ' { (> 1) x ' ' ‘ { \ ' ' ' ‘ ' ı ' ‘ ' ' ' ‘ { ' © ; ı [1 ' ' ’ ı [1 : ' ' ' ' —, CH > < Minimum number alive NIIIÜIIIN: (er) & ER NN 20,4; SS RR IN IQ 17 ea ur nd... AG Fuss ene RR ! de AAz 3 SEELEN Series Fig. 1. Minimum number of Apodemus flavicollis known to be alive each season. Total number of yellow-necked mice being 44 individuals, where one individual was present in series VIII/93 and X/93, and another VIII/96 and X/96. Starting June 1996 methodology was changed so that 228 live-traps on three trapping grids were used in- stead of 148 on two grids. Table 1. Discriminant function analysis coefficients characterizing habitat occupied by resident (time span between first and last capture being at least two days) individuals of A. flavicollis (p < 0.001). Habitat variables are ordered with regard to their importance in discriminating preferred and non-preferred habitat based on the absolute value of standardized coefficients. Habitat variable Standardized coef. of DFA Average Stand. Dev. Vegetation density above 50 cm -0,819 0,31 0,22 Distance from Pinus mugo patch 0,813 8,69 22,84 Pinus mugo cover 0,789 0,35 0,28 Area covered by rocks less than 100 cm in diameter -0,417 0,06 0,14 Juncus tnfidus cover -0,399 0,07 0,17 Herb height -0,393 35,05 15,28 Vaccinium myrtillus cover -0,263 0,36 0,25 Grass cover -0,246 0,43 0,31 Shrub height -0,242 24,13 9,37 Vegetation density below 50 cm -0,240 0,81 0,14 Area covered by rocks more than 100 cm in diameter -0,180 0,06 0,11 Number of plant species -0,125 8,06 2,98 Litter cover 0,071 0,09 0,16 Tree height -0,060 148,40 61,55 Total area covered by rocks 0,031 0,12 0,22 188 NATALIA MARTINKOVA et al. Pinus mugo cover showed a strong indica- tion that Apodemus flavicollis preferred a habitat dominated by P. mugo. However, the raw data show that no used trap was lo- cated within the dwarf pine cover (neither trap point had negative values of distance to the nearest PL mugo patch). This would characterize the habitat of occurrence of A. flavicollis as the edge of P. mugo stands. The typical habitat of Apodemus flavicollis is described as open mature forests, prefer- ably deciduous with open ground level. Its occurrence at ecotones, grasslands or shrub- by habitat is considered atypical (NIETHAM- MER 1978; GURNELL 1985). Yet, the ecotone of dwarf pine and subalpine meadows is the habitat preferred by this species in the subalpine zone in western Tatras. Here, large seeds forming the base of A. flavicollis diet (NIETHAMMER 1978; SMETTAN 1996) are in short supply as well as ın coniferous for- ests in general. These are usually consid- ered suboptimal habitats or serving as corri- dors (AnGELSTAM et al. 1987; KOTZAGEORGIS and Mason 1996; SMAHA 1996). Since the locality does not enable indivi- duals to remain through the winter, but their survival is possible during the vegeta- tion season, it could be considered a subop- timal habitat for this species (GLiwicz 1989, 1993). However, sporadic occurrence of A. flavicollis at the localıty indicates that the species is not a regular seasonal resident to this area. This assumption is supported also by the fact that the sexual ratio is devi- ant from the expected values, which occurs in dispersers (GLiwicz 1988), i.e., most in- dividuals were present at the locality exclu- sively in a single trapping series and by low References ÄNGELSTAM, P.; HAnsson, L.; PEHRSSON, S. (1987): Distribution borders of field mice Apodemus: the importance of seed abundance and land- scape composition. Oikos 50, 123-130. BEGoN, M.; HARPER, J. L.; TOwNSEND, C. R. (1997): Ecology: individuals, populations, and commu- nities. Olomouc: Vydavatelstvi University Pa- lack&ho (Czech translation). number of captures per individual. In a ha- bitat in which reproduction per individual is low, population density has a tendency to decline (GAInES et al. 1994) and the popula- tion shows a high turnover rate (MAZURKIE- wıcz 1991, 1994) in this type of habitat, is referred to as a “sink” habitat. This is the case for the area investigated in the present study. The appearance of yellow-necked mice can be explained by high population densities in altitudes below the research area. Apode- mus flavicollis tends to occur in “nuclei” within occupied forests, which are relatively stable centers of occurrence, and spatially oscillate depending on the population den- sity (GURNELL 1985). If we assume a posi- tive correlation between population density in a given nucleus and the effort that the dispersers make to travel from the nucleus (distance x number of dispersers), then po- pulation density at our study plots indicates the culmination phases at lower altitudes. A crowded habitat would force subdominant individuals to seek vacant space and they will appear at our study plots. If such a sit- uation occurs, the population probably ex- hibits a three year cycle. Acknowledgements The research in the National Nature Reserve was carried out with the permission of the Slovak Ministry of Environment. We are indepted to Mr. HALAK and Mr. HupAcek (Tatra National Park), their support is acknowledged. Financial support of the research provided by VEGA (grant number: 1/4140/97, assigned to L’. KocIan). Dusser, R. D.; Shugart H. H. Jr. (1978): Microha- bitats of a forest-floor small mammal fauna. Ecology 59, 89-98. Dusser, R. D.; SHUGART, H.H. Jr. (1979): Niche pattern in a forest-floor small-mammal fauna. Ecology 60, 108-118. GAINES, M.S.; DIFFENDORFER, J.E.; FOSTER, J.; WRray, F. P; Hort, R. D. (1994): The effects of habitat fragmentation on population of three species of small mammals in Eastern Kansas. Pol. Ecol. Stud. 20, 163-175. GLIwIcZ, J. (1988): Seasonal dispersal in non-cy- clice populations of Clethrionomys glareolus and Apodemus flavicollis. Acta Theriol. 33, 263-272. GLIWICZ, J. (1989): Individuals and populations of the bank vole in optimal, suboptimal and in- sular habitats. J. Anim. Ecol. 58, 237-247. Guiwicz, J. (1993): Dispersal in bank voles: bene- fits to emigrants or residents? Acta Theriol. 38, 31-38. GURNELL, J. (1985): Woodland rodent commu- nities. Symp. Zool. Soc. Lond. 55, 377-411. HaNnSsKI, 1.; MOILANEN, A.; GYLLENBERG, M. (1996): Minimum viable metapopulation size. Am. Nat. 147, 527-541. KOTZAGEORGIS, G. C.; Mason, C. F. (1996): Range use, determined by telemetry, of yellow- necked mice (Apodemus flavicollins) in hedgerows. J. Zool. (London) 240, 773-777. KozaAkIEWIcZ, M. (1993): Habitat isolation and ecological barriers - the effect on small mam- mal populations and communities. Acta Ther- iol. 38, 1-30. LEGENDRE, L.; LEGENDRE, P. (1983): Numerical ecology. Amsterdam, Oxford, New York: Elsevier Scientific Publ. Comp. LURA, H.; LANGHELLE, G.; FREDRIKSEN, T.;, BYRK- JEDAL, I. (1995): Distribution of the field mice Apodemus flavicollis and A. sylvaticus in Nor- way. Fauna Norv. Ser. A 16, 1-10. M’CLoskEy, R. T.; FIELDWICK, B. (1975): Ecological separation of sympatric rodents (Peromyscus and Microtus) :J. Mammalogy 56, 119-128. MAZURKIEWICZ, M. (1991): Population dynamics and demography of the bank vole in different tree stands. Acta Theriol. 36, 207-227. A. flavicollis in subalpine habitat 189 MAZURKIEWICZ, M. (1994): Factors influencing the distribution of the bank vole in forest ha- bitats. Acta Theriol. 39, 113-126. NIETHAMMER, J. (1978): Apodemus flavicollis (Melchior, 1834) — Gelbhalsmaus. In: Hand- buch der Säugetiere Europas. Band 1. Ed. by J. NIETHAMMER and F.Krapp. Wiesbaden: Akad. Verlagsges. Pp. 325-336. PuceK, Z.; JEDRZEJEWSKI, W.; JEDRZEJEWSKA, B.; Pucek, M. (1993): Rodent population dy- namics in a primeval deciduous forest (Bialo- wieza National Park) in relation to weather, seed crop, and predation. Acta Theriol. 38, 199-232. SMAHA, J. (1996): Notes on the mammal fauna of the Krivoklätsko Biosphere Reserve. Lynx (Praha) 27, 37-56 (in Czech with English sum- mary). SMETTAN, H. (1996): Bemerkenswerter Nahrungs- vorrat einer Gelbhalsmaus (Apodemus flavi- collis). Jh. Ges. Naturkde. Württ. 152, 281- 284. Yoccoz, N. G. (1992): Presence de mulot (Apode- mus alpicola ou flavicollis) en milieu alpin. Mammalia 56, 488-491. Zıax, D.; Kocıan, L’. (1994): Population dynamics of small mammals on the morene in West Ta- tras - Rohäce. Vyskum a ochrana cicavcov na Slovensku 1, 49-52 (in Slovak with English summary). Authors’ addresses: NATALIA MARTINKOVA, Department of Zoology, Charles University, Vinicna 7, 12844 Praha 2, Czech Republic (e-mail: natalia@natur.cuni.cz); Divin Zıax, and L’upovır Kocıan, Department of Zoology, Comenius University, Mlynskä dolına Bl, 84545 Bratislava, Slovakia. Mamm. biol. 66 (2001) 190-192 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication Mammalian Biology Zeitschrift für Säugetierkunde Non-invasive PCR sexing of rabbits (Oryctolagus cuniculus) and hares (Lepus europaeus) By BARBARA WALLNER, SUSANNE HUBER, and R. ACHMANN Institut für Tierzucht und Genetik, Veterinärmedizinische Universität Wien, Forschungsinstitut für Wildtierkunde und Okologie, Veterinärmedizinische Universität Wien, Ludwig Boltzmann-Institut für immuno-, zyto- und mole- kulargenetische Forschung, Veterinärmedizinische Universität Wien, Wien Receipt of Ms. 22. 09. 2000 Acceptance of Ms. 27. 12. 2000 Key words: Lepus europaeus, Oryctolagus cuniculus, sex determination, Sry, faeces Genetic sex verification has important im- plications for population studies of free- ranging animals relying on the knowledge of reproductive status and sex ratio of the animals. In the brown hare (Lepus euroaeus) a continuous population decline has been re- ported in many European countries (FLux and ANGERMANN 1990). The reason for the decrease is under debate (McLAREN et al. 1997; PAnEK and KAMIENIARZ 1999; Rey- NoLDs and TAPrER 1995), and population studies of this species are therefore highly needed. The collection of blood causes stress due to trapping and handling of animals (Jessup 1993) that could affect the para- meters under investigation, particularly ın a highly irritable species like the brown hare. Previous studies have demonstrated the po- tential for faeces collected in the field as a suitable source of DNA for genotyping and sexing free-ranging mammals (TABERLET et al. 1997). Since no sex-specific DNA se- quences are known for the brown hare, we initially developed a PCR test for sex deter- mination in rabbits (Oryctolagus cuniculus) and adapted it for sexing hares (Lepus euro- paeus). The assay co-amplifies a part of the Y-chromosomal Sry and the autosomal rab- bit transferrin gene, which is used as an in- ternal amplification control. Because most sex-identification methods are not species specific, some precautions 1616-5047/01/66/03-190 $ 15.00/0. have to be made to be aware of possible contamination with extraneous DNA, espe- cially when animal remains such as hair or faeces are used as source for DNA analysis (TABERLET et al. 1997). In contrast to this universal primer approach, primers de- scribed in this report are placed in rabbit- specific sequence regions. To test the speci- ficity of the assay, we amplified DNA from human, mouse, horse, and sheep, but none of these species amplified even under low stringency conditions (data not shown). We first verified the accuracy of the assay by analysıing genomic DNA from a total of 78 rabbits. Genomic DNA was isolated from 200 ul EDTA-blood from 24 adult males and 27 females of different rabbit breeds (GEMMELL and AkıyAaMA 1996). For 27 new born rabbits, buccal swabs sampled with Q-Tips were used for sexing in order delete to apply a minimal invasive techni- que. The cut cotton-wool end of the Q-tip was placed in a 1.5ml vial containing 600 ul digestion buffer (GEMMELL and ArıyamA 1996) and stored at room tem- perature. Genomic DNA was obtained as described above by digestion of the whole swab with proteinase K (80 ug) for 2 hours at 56 °C. DNA was then extracted from the supernatant (GEMMELL and AkIYAMA 1996). Primers (Tab. 1) amplifying a fragment from the Sry region were designed accord- Non-invasive PCR sexing of rabbits and hares 191 Table 1. Specifications of duplex-PCRs for sex determination in rabbit and hare DNA sequence of rabbits Sry re- gion in SINCLAIR et al. (1990). Rabbit transferrin (Acc. number. X58533). T,: Annealing temperature Primer 5’-primer sequence - 3° Species Fragment length Transferrin GACCTTCTACTATGCTGTGGC Exon 4/5 GTAGCCGAAATACGGCTGAAC Sry AATACAGGAGGAACACGTAAAGTG CAAACTGTCGCTCETTCTGTAGGAT Rabbit Transferrin GCCTTTGTCAAGCAAGAGACC Exon 6/7 CACAGCAGCTCATACTGATCC Sry AATACAGGAGGAACACGTAAAGTG CAAACTGTCGCTCTTCTGTAGGAT Tfexon 4/5 Iikb 3 Tfexon 6/7 0.5kb 3 Sry 0,3kb 3 Primers > | 2 3 + 5 6 7 8 9 Z10 Fig. 1. Sex determination of rabbit and hares using the transferrin (Tf)/Sry duplex-PCR assay. Lane 1 Molecular size marker (Gibco); lane 2 no template control; rabbit: lane 3, 4 blood samples; lane 5, 6 buccal swabs; hare: lane 7, 8 tissue; lane 9, 10 faeces. M = male, F= female ing to a published rabbit-human sequence alignment with their 3’-end being placed in rabbit-specific sequence regions (SINCLAIR et al. 1990). An amplicon from exons 4/5 (Tab. 1) of the rabbit transferrin gene (Tf) was used as amplification control. Duplex- PCR was carried out in a reaction mixture of 15 ul containing 20. ng template DNA, 0.5 U AmpliTag polymerase Gold (PE Bio- systems), 15 mM Tris-HCl pH 8.0, 50 mM KCl, 2mM MsCh, 0.25 mM of each dNTP and primers as indicated in table 1. The cy- cling conditions on a GeneAmp 2400 Cy- cler (PE Biosystems) were: 10 min at 95°C; 30 sec at 95°C, 30 sec at 65°C, and 60 sec at 72°C for 35 cycles. Amplicons were sepa- rated by agarose gel electrophoresis. PCR on male samples amplifies two products (Sry and Tf), whereas from female samples only one product (Tf) is obtained (Fig. 1). The PCR result was consistent with the ani- mals’ phenotypic sex ın the 51 rabbiıt blood samples tested. When using genomic DNA ısolated from 27 buccal swabs we obtained unambiguous results in 24 cases and the as- signed gender was correct. In three cases we could not amplify any fragments be- cause of degraded DNA. The assay was then adopted for sexing brown hares using DNA extracts of tissue samples from 12 individuals. In hares the Sry region could only be amplified under less stringent conditions (Tab. 1). Primers for Tf exons 6/7 were used as internal con- trol. Target specificity was certified by com- paring directly sequenced gel-purified PCR products with published sequences. The am- plified hare Sry sequence has been sub- 192 BARBARA WALLNER etal. mitted to the GenBank database (Acc. number AF230075). Rabbit and hare Sry sequences differed at four nucleotide posi- tions indicating that the amplified region is not completely conserved between these species. Rabbit specific primers may not perfectly match hare target sequences and thus only amplify the respective genomic regions under reduced stringency. DNA extracted from hare faeces can be used for sex determination, but at a higher test dropout rate. We collected fresh faecal sam- ples from 36 individually caged hares with known sex into separate plastic tubes and froze them immediately. We extracted DNA from faeces by a silica-based purification method in order to purify DNA and to break down compounds that inhibit subsequent PCR reactions (after Boom et al. 1990; Con- STABLE et al. 1995). PCR reactions of faeces References Boom, R.; SoL, €. J. A.; SALIMANS, M. M.; JANSEN, C. L.; WERTHEIM-VAN DILLEN, P. M.; VAN DER NOORDAA, J. (1990): Rapid and simple method for purification of nucleic acids. J. Clin. Mi- crobiol. 28, 495-503. CONSTABLE, J. J.; PACKER, C.; COLLINS, D. A.; PUsE, A.E. (1995): Nuclear DNA from primate dung. Nature 373, 292. FrLux, J. E. C.; ANGERMANN, R. (1990): The hares and jackrabbits. In: Rabbits, Hares and Pikas. Ed. by. J.A.CuAapmAan and J.E.C. FLux. Gland: IUCN. Pp. 61-94. GEMMELL, N. J.; AKIYAMA, S. (1996): An efficient method for the extraction of DNA from ver- tebrate tissues. Trends Genet. 12, 338-339. GRIFFITHS, R.; TIwARı, B. (1993): Primers for the differential amplifcation of the sex- determin- ing region Y gene in a range of mammal spe- cies. Mol. Ecol. 2, 405-406. JEssup, D. A. (1993): Remote treatment and moni- toring of wildlife. In: Zoo and Wild Animal Med- icine: Current Therapy. Vol. 3. Ed. by. M. E. Fow- LER. Philadelphia: W.B. Saunders. Pp. 499-504. MCLAREN, G. W.;, HUTCHINGS, M.R.; HARRIS, S. (1997): Why are brown hares (Lepus euro- paeus) rare in pastoral landscapes in Great Britain? Game Wildlife 14, 335-348. PANEK, M.; KAMIENIARZ, R. (1999): Relationships between density of brown hare Lepus euro- paeus and landscape structure in Poland in the years 1981-1995. Acta Theriol. 44, 67-75. samples contained bovine serum albumin (100 ug/ml final concentration) and were in- cubated for 47 cycles. Analysis of 20 out of the 36 DNA samples purified from faeces re- vealed the correct gender. Fifteen samples amplified no PCR product, probably because of poor DNA quality. Only one female was sexed incorrectly, possibly due to male- derived contamination of the cage. Acknowledgements We thank Prof. G. BREM for initiation of the pro- ject, R. SELHOFER and C. Haas for providing sam- ples, B. AIGNER for supplying transferrin primers and U. Bruns for advice on DNA extraction from faeces. We gratefully acknowledge W. ARNOLD and M. MÜLLER for the provision of laboratory fa- cilities. We are grateful to Agrobiogen GmbH and the State government of Lower Austria for fi- nancial support. REYNoLDSs, J. C.; TAPPER, S.C. (1995): Predation by foxes Vulpes vulpes on brown hares Lepus europaeus in central southern England, and its potential impact on annual population growth. Wildlife Biol. 1, 145-158. SINCLAIR, A.H.; BERTA P.; PALMER, M.S.; HAw- KINS, J. R.; GRIFFITHS, B.L.; SMITH, M.).; FOSTER, J. W.; FRISCHAUF, A.M.; LoVELL- BADGE, R.; GOODFELLOW, P.N. (1990): A gene from the human sex-determining region en- codes a protein with homology to a conserved DNA binding motif. Nature 364, 240-244. TABERLET, P; CAMARRA, J; GRIFFIN, S., ÜHRES, E.; HANOTTE, O©.; WAITS, L. P.; DUBoIs-PAGANON, C.; BURKE, T.; BOUVET, J. (1997): Noninvasive ge- netic tracking of the endangered Pyrenean brown bear population. Mol. Ecol. 6, 869-876. Authors’ adresses: BARBARA WALLNER, Institut für Tierzucht und Genetik, Veterinärmedizinische Universität Wien, Veterinärplatz 1, A-1210 Wien, Austria (e-mail: wallner@il22server.vu-wien.ac.at); SU- SANNE HUBER, Forschungsinstitut für Wildtier- kunde und Ökologie, Veterinärmedizinische Uni- versität Wien, Savoyenstraße 1, A-1160 Wien, Austria; R. ACHMANN, Ludwig Boltzmann-Institut für immuno-, zyto- und molekulargenetische For- schung, Veterinärmedizinische Universität Wien, Veterinärplatz I, A-1210 Wien, Austria Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana- ging editor, Prof. Dr. D. Kruska, Institut für Haustier- kunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni- kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manu- script will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21 x 30 cm) or correspondingly either from a typewriter or legible computer print with a left-hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscript. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscript following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type- written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscript. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings - Mammalian Biology Zeitschrift für Säugetierkunde depending on the theme, Acknowledgements, Zusam- menfassung (in German), References. A German title must precede the Zusammenfassung in English manu- scripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: Kaıko, E. K. V.; Hanouev, C. O0. Jr. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1821), from Panama. Z. Säugetierkunde 59, 257-273. b) From books: HERRE, W.; RöhHrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/L, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers’ symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscript is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen, Germany (email:peter.langer@anatomie.med. uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). URBAN & FISCHER Verlag Mammalian Biology Jena/Germany TEE ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. - 66(2001)3 - 129-192 Contents Original investigations Nordvig, K.; Reddersen, J.; Jensen, T. S.: Small mammal exploitation of upper vegeation strata in non-forest, mixed farmland habitats - Ausnutzung höherer Vegetationsstrata durch Kleinsäuger in gemischten Ackerland-Habitaten ___ 129 Krystufek, B.; Spitzenberger, F.; Kefelioglu, H.: Description, taxonomy, and distribution of Talpa davıdıana - Beschreibung, Taxonomie und Verbreitung von fOlpadvidana 2. 2 Du, Holzhaider, J.; Zahn, A.: Bats in the Bavarian Alps: species composition and utilization of higher altitudes in summer - Fledermäuse in den Bayerischen Alpen: Artenspektrum und Nutzung von höheren Lagen im Sommer _ 144 Gloor, S.; Bontadina, F.; Hegglin, D.; Deplazes, P.; Breitenmoser, U.: The rise of urban fox populations in Switzerland - Die Entstehung urbaner Fuchspopulationen in der Schweiz __ 155 Del Valle, J. C.; Lohfelt, M. I.; Comparatore, V. M.; Cid, M. S.; Busch, C.: Feeding selectivity and food preference of Ctenomys talarım (tuco-tuco) - Auswahl und Bevorzugung von Nahrung bei Ctenomys talarım (ucco-tuco) a, Short communications Sänchez-Villagra, M. R.; Nummela, S.: Bullate stapedes in some phalangeniform marsupials - Bullaförmige Stapedes bei phalangeriformen Marsupialia__ 174 Stevens, R.D.: Twinning in the big fruit-eating bat Artibeus Iituratus (Chiroptera: Phyllostomidae) from eastern Paraguay - Zwillingsgeburt bei der großen fruchtfressenden Fledermaus Artibeus Iituratus (Chiroptera: Phyllostomidae) aus dm östichken Poaquay 0 eg Walston, J.; Veron, G.: Questionable status of the "Taynguyen civet", Viverra tainguensis Sokolov, Rozhnov and Pham Trong Anh, 1997 (Mammalia: Carnivora: Viverridae) - Fraglicher Status der "Taynguyen Zibetkatze" Viverra tainguensis Sokolov, Rozhnov und Pham Trong Anh, 1997 (Mammalia: Carnivora: Vivernde) _ 181 Martinkovä, N.; Ziak, D.; Kocian, L'.: Response of Apodemus flavicollis to conditions at the altitude limit in the western Tatra Mountains - Reaktionen von Apodemus flavicollis auf Lebensbedingungen an der Höhengrenze in der westlichen tra __ 185 Wallner, B.; Huber, S.; Achmann, R.: Non-invasive PCR sexing of rabbits (Oryctolagus cuniculus) and hares (Lepus europaeus) - Nicht-invasive Geschlechtsbestimmung mittels PCR bei Kaninchen (Oryctolagus cuniculus) und Feldhasen (Lepus europaeus) __ 190 Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record ” -„„Nammalian Biology Zeitschrift für Säugetierkunde 4-2001 a ISSN 1616-5047 URBAN & FISCHER Mamm. biol. - 66(2001)4 - 193-256 ee. wwwv.urbanfischer de/iournals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Säugetierkunde (German Society of Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen ‚Advisory Board » „5. Halle, Jena Sn eBcdardlkeb SARshabkkerer Bonn. P. Lüps, Bern _ W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repenning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.O. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/46 35 70, E-mail: l.still@nature.com Subscribers in the USA can call toll free on ++1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 498.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipient's name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by personal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. If no cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH &) As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANSI/NISO 239.48-1992 (Permanence of Paper). ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Printed in Germany Mamm. biol. 66 (2001) 193-203 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Supernumerary molars in neotropical opossums (Didelphimorphia, Didelphidae) By D. AsTUA DE MORAES, B. LEMoOS, and R. CERQUEIRA Departamento de Zoologia, Universidade de Säo Paulo, Departamento de Genätica, Universidade Federal do Rio de Janeiro, and Departamento de Ecologia, Universidade Federal do Rio de Janeiro, Brazil msn. Receipt of Ms. 17. 05. 2000 Acceptance of Ms. 28. 09. 2000 Abstract Dental abnormalities, such as the occurrence of extra teeth, are recurrently found in many groups of mammals. Supernumeray molars were found in Didelphis aurıta, D. albiventris, D. marsupialis, Phi- lander andersoni, P. frenata, P. opossum, Chironectes minimus, and Caluromys philander. Frequencies of occurrence of supernumerary teeth in these marsupial species remained within a range similar to that found in other species. Four hypotheses are proposed and discussed to explain the origin of these teeth: appearance of extra teeth due to excessive development in size of the skull, reappear- ance of an atavistic condition, retention of the third deciduous premolar at the eruption of the per- manent premolar, or some sort of ontogenetic disturbance that lead to the duplication of a tooth germ. The first hypothesis is discarded as all individuals have normal sizes for the species. No evi- dence in the marsupial fossil record supports the second. The morphology of the teeth observed does not support the third, as all teeth are apparently permanent (except for one specimen). Finally it is hard to find evidence against or in favour of the fourth, as there is no information available of the development of the museum specimens observed. Key words: Marsupials, abnormalities, dentition, Neotropics Introduction The study of abnormalities can be particularly interesting for those involved in developmen- tal genetics and morphological evolution, providing useful data for medical, evolution- ary, and taxonomic studies. For instance, by focusing on abnormalities one can assess the potentialities for the rise of new variant morphologies. Morphological shifts occur with the rise of such deviant morphologies and its spreading through the taxon. Thus, ab- normalities, as highly deviant morphologies, could initiate such morphological shifts. 1616-5047/01/66/04-193 $ 15.00/0. Mammal dental abnormalities have long been reported, and include teeth specific malformations (e.g. Long and Long 1965; FELDHAMER and STOBER 1993), size reduc- tion or missing of teeth (e.g. MEcH et al. 1970; DREHMER and FERIGOLO 1996), or supernumerary teeth (e.g. KruTzscH 1953; STEELE and PArAMA 1979; Kvam 1985; DREHMER and FERIGOLO 1996). The latter corresponds to the occurrence of more teeth than those expected from the species normal dental formula. Such phenomenon 194 D. AsTUA DE MORAES et al. has been described for eutherian families as diverse as e.g. Mustelidae (LonG and LonG 1965), Otariidae (DREHMER and FERIGOLO 1996), Felidae (Kvam 1985), Canidae (VAN VALEN 1964), Cervidae (FowLE and PAss- MORE 1948; PEKELHARING 1968; STEELE and PARAMA 1979; MEcH et al. 1970), Soricidae (Hoorer 1946), Dipodidae (KRUTZCH 1953), and Sciuridae (Goopwin 1998). Within the marsupial family Didelphidae supernumerary teeth have been reported for the genus Didelphis (ALzen 1901; TAKA- HASHI 1974) only and recently for Philander (HERSHKoVITZ 1997). The aim of this study therefore is to investigate the situation for South American marsupials in more detail. Material and methods The following species were investigated: Calur- omys philander (Linnaeus, 1758), Chironectes minimus (Zimmermann, 1780), Didelphis albiven- tris Lund, 1840, D. aurita Wied-Neuwied, 1826, D. marsupialis Linnaeus, 1758, Philander andersoni (Osgood, 1913), P. frenata (Olfers, 1818), and P. opossum (Linnaeus, 1758), from specimens de- posited in the following collections: Museu Nacio- nal, Rio de Janeiro (MN); Departamento de Zo- ologia, Universidade Federal de Minas Gerais (DZUFMG); Museu de Zoologia, Universidade de Säo Paulo (MZUSP); American Museum of Natural History (AMNH); Field Museum of Nat- ural History (FMNH); National Museum of Nat- ural History (NMNH); and Laboratörio de Ver- tebrados, Universidade Federal do Rio de Janeiro (MC). Collection numbers, sexes and lo- calities for specimens found with supernumerary molars are listed in table 1. Unless otherwise specified tooth morphology no- menclature follows Reıc et al. (1987). Notations in super- or subscript refer to specific (upper or lower, respectively) tooth rows, as traditionally used in tooth formulae nomenclature (e.g. M*), but when no specific row is meant, we chose not to use super- or subscript (e. g. M4). Results Frequency of occurrence The frequency of supernumerary molars oc- currence in each of the investigated species was calculated based on adult specimens with four molars erupted, as the number of individuals presenting any extra tooth di- vided by the total number of specimens ex- amined for that species. The frequency of supernumerary teeth is 0.7% (1/141) in C. minimus,; 12% (1/82) in C. philander; 0.5% (3/655) in D. albiventris, 0.3% (1/337) in D. aurita, 1.0% (9/872) in D. marsupialis; 2.83% (1/36) in P. andersoni; 0.3% (2/244) in P. frenata; and 0.3% (2/767) in P. opos- sum. Tooth morphology and position The location of the supernumerary tooth found is reported in table 1. Supernumerary teeth were found at all molar rows; how- ever, they were more frequent at the upper TOWS. Caluromys philander The only specimen (MZUSP 11591) found with an extra tooth presents it at the end of the right inferior molar series (Fig. 1). It is slightly smaller than the M4. Cuspids are distinguishable with the protoconid exces- sively developed in comparison to the para- conid and metaconid that are reduced. The talonid is slightly reduced, and is divided antero-posteriorly by a crest, not present in the normal teeth, that creates two basins and makes the identification of the talonıd cusps more difficult. This tooth is aligned with the molar series in occlusal view, but its crown is slightly tilted to the lingual side. Chironectes minimus The only occurrence of a supernumerary molar for this species (MZUSP 16545) is an extra molar erupting behind the right M, (Fig. 2). The tooth is not fully erupted (only the protoconid, and the tips of the metaconid and entoconid emerge from the bone), but its crown pattern is clearly visi- ble and identifiable with conids and cristids pattern identical to the My. The tooth is clumped, as there is no space available for it in the mandibular ramus and is erupting Supernumerary molars in neotropical opossums 195 Table 1. Individuals observed with supernumerary teeth, with respective species, museum number, sex, locality of origin and location of supernumerary molars. Legend: M: Male; F: Female; UL: Upper left row; UR: Upper right row; LL: Lower left row; LR: Lower right row. Museum acronyms as in text Sex Locality Museum Number Location of super- numerary molar Species MZUSP 11591 MZUSP 16545 AMNH 63852 DZUFMG 120 MN 22250 AMNH 133034 AMNH 33243 AMNH 93978 Fordländia, Parä, Brasil Cametaä, Parä, Brasil Utcuyacu, Junin, Perü Santa Luzia, Minas Gerais, Brasil Brasilia, Distrito Federal, Brasil Paraiba do Sul, Rio de Janeiro, Brasil UL Esmeraldas, Ecuador Norte do Rio Amazonas, Faro, Para, Brasil Rio Tapajös, Igarap& Amorim, Para, Brasil Rio Tapajos, Inajatuba, Para, Brasil Beni, Bolivia Villanueva, Colombia Rio Tapajös, Caxiricatuba, Para, Brasil Bel&m, Para, Brasil Sierra de M&rida, Merida, Venezuela UL Caluromys philander Chironectes minimus Didelphis albiventris Didelphis albiventris Didelphis albiventris Didelphis aurita Didelphis marsupialıs Didelphis marsupialis u ass 5 mn AMNH 95345 = Didelphis marsupialis AMNH 95361 AMNH 209179 USNM 280966 USNM 51092 USNM 545457 FMNH 22199 Didelphis marsupialis Didelphis marsupialis Didelphis marsupialis Didelphis marsupialis Didelphis marsupialiıs Didelphis marsupialis AMNH 72017 MC 267 MN 5769 AMNH 34373 USNM 337643 Philander andersoni Philander frenata Philander frenata Philander opossum Philander opossum ee nuerszu sn with its crown tilted about 45 degrees lın- gual and anteriorly in comparison to the re- maining molars. Didelphis albiventris Specimen DZUFMG 120 presents a single supernumerary molar, erupted behind the right M* (Fig. 3). It is a small tooth (approx. 1/3 of the size of the normal M*), with an oval crown, with two cusps connected by ceristae. Its reduced size and abnormal crown shape makes it difficult to identify what these cusps and cristae are equivalent to in normal teeth. The tooth is very tight, as there is no place for it on the maxilla, be- hind the M*. The crown of this tooth is ap- proximately in the same occlusal plane as normal molars, but it does not occlude with any inferior tooth. Specimen MN 22250 possesses one super- numerary molar at the end of each series. The extra tooth on the left superior series Rio Curaray, Loreto, Perü UL Marica, Rio de Janeiro, Brasil UL Santa Teresa, Espirito Santo, Brasil UL Bagadö, Chocö, Colombia UL El Recreo, Zelaya, Nicaragua UL is smaller than the M* and similar in shape, yet with much reduced stylar cusp C, and slightly compressed, so that the stylar cusp E (metastyle) is closer to the paracone than in the M*. All remaining cusps are clearly identifiable. It is also slightly rotated coun- ter-clockwise. Its crown lies beneath the oc- clusal plane of the rest of the series but oc- cludes with the extra molar of the inferior series. The extra tooth on the right superior side is identical in shape to the M*, and of about half its size. Unlike the tooth pre- viously described, orientation and cusps are similar to the M*, and apart from the size difference, there is no deformation or other shape difference. Its crown also lies beneath the occlusal plane of the M*, and apparently does not occlude. The left infer- ior extra tooth (Fig. 4) is overall similar to the M,„, especially the trigonid is identical to the M, trigonid, with all cusps identifi- able. In the talonid, however, the hypoco- nulid is closer to the hypoconid, giving the 196 D. ASTUA DE MORAES et al. whole talonid a more elongated or triangu- lar shape. Finally, the supernumerary molar on the right mandible (Fig. 5) also presents a trıgonid ıdentical to the M, trigonid, but with a slightly different talonid, with a hy- poconulid somewhat farther from the trıgo- nıd than in the latter. Prd Fig. 1. Right inferior molar series of Caluromys philan- der (MZUSP 11591 ), with a supernumerary molar (ar- row). Occlusal view. Protoconid (Prd), paraconid (Pad), metaconid (Med). Didelphis aurita Specimen AMNH 133034 possesses one supernumerary molar in each side of the superior row. Both extra teeth have a mo- lariform shape but are slightly reduced. The extra tooth on the right superior series is posteriorly directed, probably due to the lack of space in the row, and is not at the oclusal plane of the remaining teeth. Its cusps are present and distinguishable with a crown pattern resembling a normal M*. The other extra tooth at the left superior series is not at the occlusal plane of the re- maining row and its cusps, although distin- guishable, do not resemble a normal M* pattern. Med Hyde Fig. 2. Right inferior molar series (M2-M4) of Chiro- nectes minimus (MZUSP 16545), with a supernumerary molar (arrow), viewed from the occlusal plane of the extra tooth. Protoconid (Prd), paraconid (Pad), meta- conid (Med), entoconid (End), hypoconid (Hyd). Supernumerary molars in neotropical opossums D. marsupialis Specimen AMNH 93978 possesses one ex- tra tooth on each side of the superior rows. Both are reduced but with a molariform shape. The extra teeth at the right has some cusps visible (but hardly identifiable), and a crista, which is apparently the centrocrista, oriented with the antero-posterior axis of the skull. Philander frenata Specimen MC 267 presents one extra molar on each side, at the end of the upper molar series. On the left side (Fig. 6), the supernu- Fig. 3. Right superior molar series of Didelphis albi- ventris (DZUFMG 120), with a supernumerary molar (ar- row). Occlusal view. 197 merary tooth is smaller than the M*. Some cusps are discernible, as well as what ap- pears to be the centrocrista. Protocone, paracone and metacone are identifiable, but the cusp present on the opposite side End Hyld Fig. 4. Left inferior molar series of Didelphis albiventris (MN 22250), with a supernumerary molar (arrow). Oc- clusal view. Entoconid (End), hypoconid (Hyd), hypo- conulid (Hyld). 198 D. AsTUA DE MoORAES et al. of the centrocrista cannot be identified. Ac- tually, this extra tooth resembles a normal M'* as described by Hersnkovitz (1997) for End Fig. 5. Right inferior molar series of Didelphis albiven- tris (MN 22250), with a supernumerary molar (arrow). Occlusal view. Entoconid (End), hypoconid (Hyd), hy- poconulid (Hyld). Philander, and the M* resembles a normal M°: On the right side the extra tooth is more deformed, somewhat ovally shaped with some cusps visible (but hardly identifi- able), and a partially formed crista (appar- ently the centrocrista, oriented with the antero-posterior axis of the skull). Unlike on the left side, the right tooth shows some degree of wear on the cusps and outer cris- tae. In both cases teeth are partially clut- tered on the M*, and crowns are on a differ- ent occlusal plane than the remaining molars (roots are more developed). Specimen MN 5769 also presents one extra molar on each side, behind the upper left and right M*. On the right side (Fig. 7), the extra molar is partially erupted. It is dislo- cated, due to the lack of space on the den- Fig. 6. Left superior molar series of Philander frenata (MC 267), with a supernumerary molar (arrow). Occlu- sal view. Protocone (Pr), paracone (Pa), metacone (Me), centrocrista (Cc). Supernumerary molars in neotropical opossums Fig. 7. Right superior molar series of Philander frenata (MN 5769), with a supernumerary molar (arrow). Oc- clusal view. Protocone (Pr), paracone (Pa), metacone (Me), centrocrista (Cc), paracrista (Pc). tary behind the M‘“, with its occlusal plane of the tooth posteriorly directed. A centro- crista is also recognisable, as well as proto- cone, paracone and metacone. The paracris- ta is also present, and the tooth lacks stylar cusps C, D and E. On the left side. the extra tooth is more oddly formed, with a crown of ovoid shape. Cusps are perceptible, such as an inner crista, probably the centrocrista. The occlusal plane, however, is aligned with the remaining molars of the series. P. opossum One specimen (AMNH 34373) presented a single supernumerary tooth at the end of each superior row. The extra tooth at the left side is reduced, but otherwise much si- milar to a normal M* in cusp patterns. It is also at the same occlusal plane of the re- maining teeth. The extra tooth at the right superior side is occluding anteriorly or- 199 iented, and seems to be pushing forward the row. It is reduced, with distinguishable cusps. Discussion The occurrence of dental abnormalities such as supernumerary teeth is a rare and unpredictable phenomenon that makes the study of its specific developmental causes and processes very unlikely. Although the alternative explanations to the supernumer- ary teeth phenomenon are not always ex- clusive we provide a critical evaluation of some possible explanations. Altough rare, the frequencies of supernumer- ary molars reported here for the Didelphidae fall within the range of those found in other mammal (placental) groups where the phe- nomenon has been reported (e.g. 1.6% (9 550) in European Iynxes (Lynx lynx (Lin- naeus, 1758)) Kvam 1985; 0.2% (1/580) in red deers. (Cervus elaphus Linnaeus, 1758) and 0.85% (1/130) in wapitis (Cervus canaden- sis Linnaeus, 1758) PEKELHARING 1968; 3.7% (4/109) in mooses (Alces alces (Linnaeus, 1758)), STEELE and PARAMA 1979. First it is important to consider if these supernumerary teeth are a return to a lost primitive condition, thus being an atavistic character. For instance, TAKAHASHI (1974) suggested that the presence of supernumer- ary incisors is an ataviıstic character of some Didelphis specimens. However, BERKOVITZ (1978) states that no more than five incisor tooth germs were ever observed in Didel- phis. Thus, the supernumerary incisors stu- died by TAKAHASsHI (1974) doubtfully can be interpreted as atavistic characters. Furthermore, an atavistic explanation pre- sents some limitations in our case, as the basic marsupial dental formula, exhibited by extant Didelphidae (I3CıP3M4) differs from the basic therian formula at the time of metatherian divergence from eutheria by the lost of premolar teeth on the former (BARBOUR 1977). Besides the proposition of some authors that the third deciduous pre- molars (dP3) could in fact be first molars and the subsequent molars being M2 to 200 D. AsTUA DE MORAES et al. MS (e.g. HERSHKoVITZ 1992), there is no mention of a truly additional fifth molar in marsupials (living or fossil). The only known exception would be the Australian numbat Myrmecobius fasciatus Waterhouse, 1836 (Myrmecobiidae), which can present 5 or 6 molars (THEnIus 1989). In this case however, it is believed that the molar num- ber is a secondary specialisation resulting from jaw elongation and not related to any ancestral condition, or primitive trait. Hence, an atavistic explanation could be advanced for a possible supernumerary pre- molar but not for molars. In fact there is no fossil record relating the occurrence of five molars in marsupials. Another hypothesis relates an excessive size development to the emergence of an extra tooth at the end of the tooth row. However, as all specimens studied here pre- sented standard sizes for their species, such a hypothesis does not seem to be the case. Furthermore, as we reported, in many cases there is actually a lack of space for these teeth to develop, which however did not prevent the teeth from appearing and the amount of space available does not seem to be related to the completeness of the tooth formation. In some reports the occurrence of supernu- merary teeth had been related to develop- mental disorders, such as splitting of the tooth germ (KrUTZcH 1953; LonG and LonG 1965; PEKELHARING 1968; STEELE and PARA- MA 1979). These developmental alterations on the embryonic germ will likely lead to an incomplete development of one or both du- plicated teeth, as observed in cervids, and in some cases resulting in an extra tooth “mir- rored” in relation to the original one (PEKEL- HARING 1968). The supernumerary molars here reported vary from morphological per- fect M*-like teeth to very small vestigial teeth, and never in this “mirrored” situation. In fact, no clear association between lack of space and amount of development of the supernumerary tooth could be found. Furthermore, knowledge on precise tooth development for most of these species is not existant, and all individuals examined are field caught, making an exact determina- tion of the underlying phenomenon hard. Nevertheless, some sort of random develop- mental disorder could be an explanation for many of the cases we studied here. Such an explanation is more difficult to ac- cept in extreme cases such as MN 22250 and AMNH 209179, where four almost fully developed molars are present, requir- ing a simultaneous event of germ duplica- tion in all M4. A genetically based distur- bance with simple mendelian inheritance is also unlikely, as one of the specimens stu- died (MC 267) was field caught but main- tained alive for captive breeding, and none of the offspring presented any similar phe- nomenon. An alternative explanation could be that the premolarıform P3 had erupted without loss of the molariform dP3. If this is the case, the cheek teeth observed would in fact be, in order: Pl, P2, 33,423 715 N PN and MA. The eruption of the P3 would dis- place the whole molar series (at the time of eruption of the P3, only MI and M2 are present), thus forcing the M3 and MA4 to the end of the maxillar bone, which in case of lack of space could explain the deforma- tions eventually observed. According to BERKoVITZ (1978), in Didelphis virginiana the third premolar develops in the embryo- nic dental lamina between the second pre- molar and the deciduous molar. Thus, if the deciduous molar was displaced poster- iorly instead of falling, the sequence of the teeth after eruption would be the one pre- viously stated. However, both upper and lower deciduous premolars are morphologi- cally different from first permanent molars; dP°’s are narrower than M!’s, and dP3’s have narrower trigonids than My,’s, and their talonids are bigger in relation to the trigonid than in the M.. In all animals ex- amined here the first molar in the series has all characteristics of an actual MI, and not of a dP3, which denies this hypothesis. The different shape variations observed in these teeth can be explained by the classic field model of mammalian heterodonty. This model postulates that heterodonty de- rived by the existence of three morphoge- netic fields (incisor, canine, and molar). Supernumerary molars in neotropical opossums These fields determine what the final form of a developing tooth bud will be (BUTLER 1939) but conflicting functional aspects are also important. In the genus Peromyscus Gloger, 1841, for instance, correlations among cheekteeth make M° and M; widths somewhat independent (van VALEN 1962). The characteristics of each type of teeth ap- pear early in tooth ontogeny although initial buds are undifferentiated (ArCHER 1974; BERKoVITZ 1978; BUTLER 1978). At weaning Didelphis has three functional molariform teeth. As it grows the deciduous premolar (which is molariform) is shed at the same time of the eruption of M3 (TRrıgeE 1990). In this sense the deciduous P3’s would actually be molars because when they grow they are located in a molar field (ArcHEr 1974). Al- lometric growth of the bones changes the fields causing the so-called deciduous pre- molar to be shed and substituted by the P3’s now growing in a premolariform field. Different allometric rates among taxa would then explain the differences in onto- geny of dentition among didelphids ob- served by TRıBE (1990) and us (AstÜA DE MOorAES and LEMos, unpubl. data). Apparently it is possible that ancestors of metatherians had two dentitions like the placentals, since some secondary incisive and canine buds appear but are resorbed la- ter in Antechinus flavipes (Waterhouse, 1838) (ArcHER 1974). Thus the dental lami- na of the oral epithelium seems to have the potentiality to develop more teeth buds than teeth that effectively erupt. The super- numerary teeth could then be a congenital accidental anomaly. The resorption of the buds may be determined by the fixation of Zusammenfassung 201 the teeth above them. With available space above buds (lack of teeth), these buds would develop into supernumerary teeth. In fact, supernumerary incisives and molars are more frequent (being the only ones so far described) because there is the available space in the diastema, between incisives and canines and also behind the cheek teeth row. Supernumerary incisives appear on the premaxillary bone only (TAKAHASHI 1974). The additional teeth usually show the char- acteristic morphology of the teeth of the re- gion where they appear, Exceptions are va- lid in cases when where there is no space for the tooth to fully develop into its nor- mal morphology. Thus, varied morphology can be expected to appear. Regardless of an explanation these dental abnormalities do not seem to be especially selectively disadvantageous, as all animals reached adulthood before being captured. Acknowledgements We are grateful to G. A. B. FonsEcA (DZUFMG), A.L. GARDNER (USNM), L.F. B. OLIvEIRA, and L.O. SAarLLes (MN), B. PATTERSON (FMNH), M. DE Vıvo (MZUSP), and R. Voss (AMNH) for access to the collections under their care. To M.A.R. MELLo and LENA GeIsE for translation and corrections in the Zusammenfassung. ARIAN- NA CAMARDELLA gave us precious help sharing her knowledge on didelphid molar morphology, as well as J. A. W. KırscH regarding dental formu- las in Myrmecobius. JAIR DA SıLva did the draw- ings. We thank two anonymous referees for sug- gestions. The work was partially supported by grants from CNPq, FAPERJ, FAPESP, FUJB, PROBIO/MMA and PRONEX. Überzählige Molaren bei neotropischen Beuteltieren (Didelphimorphia, Didelphidae) Abweichungen von der arttypischen Zahnzahl wie das Auftreten zusätzlicher Zähne, werden in vie- len placentalen Säugetiergruppen gefunden. Die Untersuchungen an neotropischen Beuteltieren haben ebenfalls überzählige Molaren ergeben bei: Didelphis aurita, D. albiventris, D. marsupialis, Philander opossum, P. frenata, P. andersoni, Chironectes minimus und Caluromys philander. Die Häu- figkeit des Auftretens überzähliger Zähne bei diesen Arten bleiben innerhalb eines Bereiches, der ähnlich demjenigen bei anderen Arten ist. Folgende Möglichkeiten zur Deutung dieses Phänomens 202 D. ASTUA DE MORAES et al. werden erwogen: 1) Erscheinung der Extrazähne als Folge einer übermäßigen Entwicklung der Schädelgröße; 2) Atavismus; 3) Persistenz des dritten Milchprämolaren bei Erscheinen des Dauer- zahnes; 4) Wachstumsstörungen, die zur Verdoppelung eines Zahnkeimes führen. Die erste Hypo- these wird verworfen, da alle Einzelindividuen eine arttypische Größe haben. Kein fossiler Hinweis unterstützt auch die zweite. Die Morphologie der beobachteten Zähne unterstützt auch die dritte nicht, da es sich um Zähne des Dauergebisses handelt. Schließlich ist es schwierig, Beweise gegen oder zugunsten der vierten Hypothese zu finden, da keine Informationen vorhanden sind, über die Entwicklung der Zähne bei den bearbeiteten Museumsexemplaren. References ALLEN, J. A. (1901): A preliminary study of the north American opossums of the genus Didel- phis. Bull. Am. Mus. Nat. Hist. 14, 149-188. ARCHER, M. (1974): The development of premolar and molar crowns of Antechinus flavipes (Marsupialia, Dasyuridae) and the signifi- cance of cusp ontogeny in mammalian teeth. J. R. Soc. West. Austr. 57, 118-125. BARBOUR, R. A. (1977): Anatomy of marsupials. In: The Biology of Marsupials. Ed. by B. STONEHOUSE and D. GILMORE. London: Macmillan Press. Pp. 237-272. BERKOVITZ, B. K. B. (1978): Tooth ontogeny in Di- delphis virginiana (Marsupialia: Didelphidae). Austr. J. Zool. 26, 61-68. BUTLER, P.M. (1939): Studies of the mammalian dentition. Differentiation of the post-canine dentition. Proc. Zool. Soc. Lond. (B) 109, 1- 36. DREHMER, C. J.; FERIGOLO, J. (1996): Anomalias e patologias dentärias em Arctocephalus G. Saint Hilaire and Cuvier (Pinnipedia, Otarii- dae) da costa do Rio Grande do Sul. Rev. Brasil. Zool. 13, 857-865. FELDHAMER, G. A.;, STOBER, T.L. (1993): Dental anomalies in five species of North American shrews. Mammalia 57, 115-121. FOWLE, C. D.; PASSMORE, R. C. (1948): A supernu- merary incisor in the white-tailed deer. J. Mammalogy 29, 301. Goopwin, H.T. (1998): Supernumerary teeth in Pleistocene, recent, and hybrid individuals of the Spermophilus richardsonii complex (Sciuridae). J. Mammalogy 79, 1161-1169. HERSHKOVITZ, P. (1992): The South American gra- cile mouse opossums, genus Gracilinanus Gardner and Creighton, 1989 (Marmosidae, Marsupialia): a taxonomic review with notes on general morphology and relationships. Fieldiana: Zoology 70, 1-56. HERSHKOVITZ, P. (1997): Composition of the fa- mily Didelphidae Gray, 1821 (Didelphoidea: Marsupialia), with a review of the morphol- ogy and behavior of the included four-eyed pouched opossums of the genus Philander Tiedemann, 1808. Fieldiana: Zoology 86, 1- 103. HooPrEr, E.T. (1946): Extra teeth in shrews, J. Mammalogy 27, 394. Kvam, T. (1985): Supernumerary teeth in the Euro- pean lynx, Lynx lynx Iynx, and their evolution- ary significance. J. Zool. (London) 206, 17-22. KRUTZSCH, P. H. (1953): Supernumerary molars in the jumping mouse (Zapus princeps). J. Mam- malogy 34, 265-266. Long, C. A.; Long, C. F. (1965): Dental abnormal- ities in North American badgers, genus Taxi- dea. Trans. Kansas Acad. Sci. 68, 145-155. MECH,L.D.; FRENZEL, L.D.JR.;, Karns, P. D.; KuEHN, D.W. (1970): Mandibular dental anomalies in white-tailed deer from Minneso- ta. J. Mammalogy 51, 804-806. PEKELHARING, C.)J. (1968): Molar duplication in red deer and wapiti. J. Mammalogy 49, 524- 526 Reıc, ©. A.; KırscH, J. A. W.; MARSHALL, L. G. (1987): Systematic relationships of the living and Cenozoic American “opossum-like” mar- supials (Suborder Didelphimorphia), with comments on the classification of these and of the Cretaceous and Paleogene New World and European metatherians. In: Possums and Opossums: Studies in Evolution. Ed. by M. ARCHER. Sydney, Australia: Surrey Beatty and Sons and The Royal Zoological Society of New ; South Wales. Pp. 1-89. STEELE, D. G.; PARAMA, W. D. (1979): Supernumer- ary teeth in moose and variations in tooth number in north american cervidae. J. Mam- malogy 60, 852-854. TAKAHASHI, F. (1974): Ocorr&ncia de poliodontia em Didelphis L. 1758 e sua possivel correlagäo a uma forma ancestral. An. Acad. Brasil. Cienc. 46, 283-285. THENIUS, E. (1989): Zähne und Gebiß der Säuge- tiere. In: Handbuch der Zoologie. Vol. 8 Supernumerary molars in neotropical opossums Mammalia Part 56. Ed by J. NIETHAMMER, H. SCHLIEMANN, and D. STARCK. Berlin, New York: Walter de Gruyter. TRIBE, €. J. (1990): Dental age classes in Marmosa incana and other didelphoids. J. Mammalogy 71, 566-569. VAN VALEN, L. (1962): Growth fields in the denti- tion of Peromyscus. Evolution 16, 272-277. VAN VALEN, L. (1964): Nature of the supernumer- ary molars of Otocyon. J. Mammalogy 45, 284-286. 203 Authors’ addresses: DiEGoO AsTÜA DE MoRrAES, Departmento de Zool- ogia, Universidade de Säo Paulo. C.P. 11461, 05422-970, Säo Paulo, SP, Brasil, (e-mail: dmoraes@ib.usp.br), BERNARDO LEMOoS and RUI CERQUEIRA, Laboratörio de Vertebrados, Departamento de Ecologia, Universidade Federal do Rio de Janeiro. C.P. 68020. 21941-590, Rio de Janeiro, RJ, Brasil. Mamm. biol. 66 (2001) 204-214 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Original investigation A habitat analysis of badger (Meles meles L.) setts in a semi-natural forest By TATJANA C. GOOD, KARIN HINDENLANG, S. IMFELD, and B. NIEVERGELT Wildlife Research and Conservation Biology, Zoological Institute, University of Zurich, Zurich, Switzerland and Spatial Data Handling Division, Geographical Institute, University of Zurich, Zurich, Switzerland Receipt of Ms. 27. 07. 2000 Acceptance of Ms. 22. 12. 2000 Abstract We studied the size, distribution and habitat characteristics of badger (Meles meles L.) setts in a largely forested area near the city of Zurich, Switzerland. The distribution of the setts was non-ran- dom, as revealed by testing nearest neighbour distances. To evaluate the habitat characteristics that determine sett locations, different parameter categories describing topography, vegetation cover and structure of the forest habitat were analysed with a multiple regression analysis and with a digital terrain model of the forest using a Geographical Information System (GIS). Preferred sett sites were the convex slopes with an inclination of 20-40°. These sites are well drained and offer many opportunities for digging entrances and tunnels, and thus gives the badger the option to leave the sett from several directions. Ideal sett sites were found above 600 metres a.s.l., closer to the forest boundary and adjoining agricultural zones than the random points. These sett sites probably guarantee access to a good food supply year-round and allow badgers to adapt their fora- ging behaviour to seasonal changes in food availability both within the mixed forest stands and in the agricultural fields and meadows outside the forest. Setts were found more than 50 metres from the nearest path and in areas with sparse ground cover. Coniferous stands were avoided. However, single old spruces within deciduous forest stands were frequently used as sett sites for setts con- sisting of one or two entrances only. Spruce trees have shallow roots, which facilitate digging and help prevent the roof of the sett from collapsing. Vegetation cover played an important role in the choice of a sett site. However, just “being out of view” (be it through topographic characteristics or distance from the nearest path) could be a type of cover as well. In this study, the small-scale topography around the setts seemed to play a key role in the choice of sett site. The results pre- sented here suggest that a large, deciduous forest with a Prenounezı topographical variation repre- sents a good badger habitat. Key words: Meles meles, sett distribution, entrance type, sett site, habitat analysis Introduction In general, carnivores not only show great in- 1975). The European badger (Meles meles terspecific diversity in their behavioural ecol- L.) is an example of a species that shows a ogy (BEKOFF et al. 1984; GITTLEMAN 1986) but high degree of plasticity in its behaviour, also marked intraspecific variability (Wırson adapting its social and spatial organisation to 1616-5047/01/66/04-204 $ 15.00/0. A habitat analysis of badger (Meles meles L.) setts different environments and food availability. In high-population-density areas with an abundant and highly predictable food avail- ability throughout the year, badgers usually live in groups, defend small territories and occupy distinctive main setts (CHEESEMAN et al. 1981, 1987, 1988; Kruuk 1978; KRUUK and PArısH 1982, 1987; NoLET and KILLING- LEY 1986; RODRIGUEZ et al. 1996; RoPER et al. 1986; WooDROFFE and MACDOoNALD 1992, 1993). In areas with a seasonally changing, unpredictable food availability, population densities are lower and badgers live in small groups or solitarily, have large overlapping home ranges and use several setts within a range (Bock 1986; CRESSWELL and HARRIS 1988; GRAF et al. 1996; PıGozzı 1989; SKINNER and SKINNER 1988). DONCASTER and WooD- ROFFE (1993) suggested that the spatial orga- nisation of badgers may be influenced by the distribution of suitable sett sites in a given area. Parameters affecting the distribution of badger setts include the type of soil, the amount of cover and the hilliness of the ter- rain (NEAL 1986). Most assessments of suita- ble badger setts have been undertaken in mixed wood- and arable land (e.g. CrEss- WELL et al. 1990). These studies showed an active selection for woodland as sett location. However, open fields and meadows were used as foraging grounds and had an impor- tant effect on sett choice (HoFER 1988; SKIN- NER et al. 1991). The goal of this study was to examine sett density, sett type and the specific habitat parameters affecting the distribution of setts in a highly forested habitat (Sihlwald, Swit- zerland), offering both ideal digging condi- tions as well as good foraging grounds. Furthermore, the correlation between the distribution of the setts in certain parameter categories and the availability of those cate- gories in the study area was explored. Material and methods The study area The Sihlwald forest is situated approximately 10km south of the city of Zurich, Switzerland 205 (47°15’ N, 8°34’ E). It is characterised by a diverse mosaic pattern ot mixed deciduous forest domi- nated by beech (Fagus silvatica), with smaller pro- portions of ash (Fraxinus excelsior), other decid- uous trees, white pine (Abies alba) and the introduced Norway spruce (Picea abies). De- clared a nature reserve in 1994, it covers approxi- mately 1000 ha of a forested hill chain, ranging from 470 metres a.s.l. at the bottom of the valley to over 900 metres a.s.l. on the ridge. It belongs to the Swiss plateau and consists of subalpine mo- lassic sandstone with partly morrainic cover. The dominant soil types contain sandy to silty clay and argillaceous sand. The extreme relief and well-drained soils make Sihlwald an ideal place for digging setts. Methods The study area was searched for setts from March until August 1996. A sett was defined as at least one entrance leading more than two metres un- derground, measured with a two-metre flexible stick. If two entrances were farther than 25 metres apart, they were considered as two se- parate setts. We assumed that practically all setts were found. The setts were classified into small (1 or 2 entrances), middle-sized (3 or 4 entrances) and large (>4 entrances) setts. Five different en- trance types were distinguished: entrances dug di- rectly into the ground, under a boulder/rock, un- der a spruce (Picea abies), under a deciduous tree and under a stump. This study did not differentiate between fox dens and badger setts for the following reasons. Although it is known that fox dens usually have fewer entrances and a different shape and smell than badger setts (STUBBE 1980), the criteria were not as clear-cut in Sihlwald with dens/setts con- sisting of one or two entrances that were sporadi- cally used by one or both species. Badger hairs were found in many setts consisting of one en- trance only, indicating the presence of badgers in single entrance setts as well. Analysis of sett char- acteristics based on sett size revealed no statistical difference between setts. Therefore, all setts were included in the habitat analysis presented here. However, to compare main sett density with those in the literature, all setts with more than two en- trances and definite badger signs were considered “main setts” (Kruuk 1978). In order to test the distribution of the setts for non-randomness, near- est neighbour distances between the setts were compared with a Monte-Carlo-Simulation of nearest neighbor distances (random distribution, 206 D. TATJANA C. GOOD et al. 1000 samples of 123 points) and then tested using a Chi2-test for two independent samples. The habitat parameters chosen for the analysis are summarised in table 1. The parameters were either measured in the field (field data), or ob- tained from a Geographical Information System (GIS) using the software ArcInfo, which also con- tained a digital terrain model of the forest based on 10 metre-contour lines (Tab. 1). The field data were measured within a radius of 25 metres of the approximate centre of the sett. The habitat analysıs was calculated by using a stepwise back- ward logistic regression. The parameters “topo- graphy” and “vegetation unit” were used as cate- gorial variables (equivalent to the traditional group of “dummy variables”) (NIEVERGELT 1981). The parameters derived from the GIS and the field data could not be analysed together, as the parameters derived from the GIS were available for the whole forest, whereas the field data were available only for the sett sites. For the analysis of the parameters derived from the GIS, a good- ness-of-fit test compared the number of observed setts with the number of expected setts for each parameter: area of the parameter number of expected setts = p area of the forest x total number of observed setts To measure the availability of the habitat para- meters measured in the field (i.e. the parameters that could not be derived from the GIS database), the number of setts was compared with the number of random points for each parameter by a y-test for two independent samples. Due to a strong cor- relation between “altitude” and “distance to forest boundary”, the “distance to forest boundary” was analysed separately. The habitat parameters for which the distribution of the setts was non-random were further analysed to see which categories (Tab. 1) best explain the sett distribution. Every parameter category was tested for deviation from the expected value by using a y’-test in conjunction with a Bonferroni z statistic (NEU et al. 1974). Results Sett size and sett type 123 setts were found in the 1000 ha study area of Sihlwald (12.3/100 ha). The setts were classified according to entrance num- ber (one to two, three to four, more than four) and entrance types (five classes, see below). Small setts were most common (71.6%), followed by middle-sized (19.5%) and large (8.9%) setts. Using KRrUUKR’s (1978) definition of “main setts” (setts with >2 entrances), the density of main setts in the forest was 3.5/100 ha. The number of entrances per sett varied from one to eleven with an average number of 2.3 entrances per sett. Of the total 279 entrances, 207 (74.2%) were dug into the ground, while the rest were dug under some type of struc- ture (11.1% under spruces, 2.5% under de- ciduous trees, 7.9% under boulders, 4.3% under tree stumps). Sett spacing The average nearest neighbour distance be- tween two setts was 111 metres, compared to 1558 metres for the average nearest neigh- bour distance between two generated ran- dom points. Thus, the observed distribution of the setts was significantly different from random (y’-test for two independent sam- ples, p < 0.001). When the nearest neighbour distance was calculated for the “main setts” only, the average nearest neighbour distance was 311 metres, compared to the 314 metres for the average nearest neighbour distance be- tween two generated random points. The distribution of “main setts” did not differ from random. Habitat analysis of the setts The results of the stepwise backward logis- tic regression show that sett sites were posi- tively associated with the parameters “con- vex slope”, “inclination” and “distance to nearest path” but negatively associated with “concave slopes”, “flat areas”, “gentle slopes”, as well as “moss-”, “herb-” and “middle layer coverage” (Tab. 2). The re- sults for the different parameter categories are as follows: Forest parameters and vegetation cover: The forest parameters and vegetation cover seemed to play a key role. The results showed that the parameters of the setts dif- fered significantly from the availability of those parameters for “lower layer cover- A habitat analysis of badger (Meles meles L.) setts 207 Table 1. Habitat parameters for setts. The parameters were either measured in the field (field data), or were ob- tained from the Geographical Information System (GIS) Parameter Subscales X-coordinates Y-coordinates Sett location Altitude metres a.s.l. Inclination Aspect Topography* degrees degrees flat - gentle slope concave slope convex slope crest Vegetation cover tree cover (> 1.3 m) shrub cover (0.5-13 m) herb cover (0-0.5 m) moss cover lower-, middle and upper layer coverage Forest parameters”* lower-, middle- and upper coniferous layer coverage vegetation unit stage of development Categories continuous BRAUN-BLANQUET (1964): 0, 1-5%, 6-25%, 26-50%, 51-75%, > 75% Source of data Field data < 600 m; < 700 m; < 800 m; >800 m GIS 0-10°; 11-20°; 21-30°; 31-40°; > 40° N; NE; E; SE; S; SW; W; NW GIS Field data Field data Field data BRAUN-BLANQUET (1964): 0, 1-5%, 6-25%, 26-50%, 51-75%, > 75% BRAUN-BLANQUET (1964): 0, 1-5%, 6-25%, 26-50%, 51-75%, > 75% 1 = young growth; 2 = pole wood; 3 = young timber wood; 4 = middle timber wood; 5 = old timber wood I; 6 = old timber wood II; 7 = old timber wood III Distance to nearest metres path Distance to nearest water metres Distance to forest metres boundary <100IM2S200M2.- < 50 m; < 100 m; < 150 m; < 200 m; >200 m < 50 m; < 100 m; < 150 m; > 150 m .; < 900 m; > 900 m * gentle slope: a slanting surface neither curving inward nor outward concave slope: a slope curving inward convex slope: a slope curving outward, like a segment of a globe. ** data from the forest superintendent’s office (Waldamt der Stadt Zürich). lower layer coverage density: canopy density that reaches at most 1/3 of the dominant tree height. middle layer coverage density: canopy density that reaches 1/3-2/3 of the dominant tree height. upper layer coverage density: canopy density that reaches at least 2/3 of the dominant tree height. ’ age” and “lower-” and “middle coniferous layer coverage” (Tab. 3a). However, more setts than expected were found only in areas lacking a “middle coniferous layer coverage” (Tab. 4; Bonferroni z statistic, p< 0.05). The parameters of the setts dif- fered significantly from the random points for herb- and moss coverage (Tab. 3b; y’- test for two independent samples, p < 0.05 and p<0.0l, respectively), as setts were more frequently found in areas with little “herb-” and no “moss coverage” (Tab. 4; Bonferroni z statistic, p< 0.05 and p< 0.01 respectively). 208 D. TATJANA C. GooD etaal. Table 2. Habitat parameters that best explain the occurrence of the setts. Multiple logistic regression (backward, stepwise), Model-y° = 92.36; df = 10; p < 0.001; R’ = 0.77 Habitat parameter Inclination Topography flat gentle slope concave slope convex slope Distance to nearest path Middle layer coverage Herb coverage Moss coverage Table 3. Comparison of the habitat parameters of the setts derived from the GIS with the availability of those parameters within the study area (a); comparison of the habitat parameters of the setts with those of the random points (b). Only the parameters for which the distribution of the setts is significantly non-random are listed here. Goodness-of-fit test; df = degrees of freedom. For (b): nı = 123; n, = 85 a) Comparison with availability Altitude Aspect Lower layer coverage Lower coniferous layer coverage Middle coniferous layer coverage Distance to forest boundary Distance to nearest path Distance to nearest water b) Comparison with random points Inclination Topography Herb coverage Moss coverage Inclination and topography: Comparison of the habitat parameters of the setts mea- sured in the field with those of the random points (Tab. 3b) showed that setts differed significantly from random points for “incli- nation” and “topography” (both: y’-test for two independent samples, p < 0.001). “Con- vex slopes” with the inclination categories <30° and <40° were sıgnificantly preferred sett sites (Tab. 4; Bonferroni z statistic, p< 0.001). Distances to forest boundary, forest roads and trails, and water: The distribution of the setts was decidedly non-random for NWVow%Pm N N W these three parameters (Tab. 3a). Setts are found significantly closer (<100 m) to the forest boundary than expected (Tab. 4; Bonferroni z statistic, p<0.05). Signifi- cantly more setts than expected were found >50 m but < 100 metres to the nearest road or trail (Tab.4; Bonferroni z_ statistic, p< 0.001). No difference was obtained for any category of the parameter “distance to nearest water”. Altitude and aspect: The distribution of the setts was non-random for “altitude” (Tab. 3a; y-test for two inde- pendent samples, p< 0.001) as well as for “aspect” (p<0.02). More setts were found 209 f badger (Meles meles L.) setts SIS O N A habitat ana T00°0 >d FAR TA L00°0 >d EI’B2T = „X 100°0 >d Ee’82l = ,X 100°0 > d 6L’8T = ,* go'0 > d EIG = X 100°0 >d LIYYT = ,X T00°0 > d g8’Gel=,X T00°0 > d g8'Gel=,* T00°0 > d To’ /2=,X g0'0 >d LIEY=,X T0'0 >d 2607 = ,X LL67'2 =2:50°0=d LL67°2 =2:90°0 = d L1L67°2 =2:50'0 = d LL67'2=2:590'0=d 1/1082 =2:50°0=d:0T =Uu 89/52 =2:90'0 =d 89/19’2=2:90'0 = d 89/92 =2:90'0 = d LL67'2 =2:50°0 = d £gE9’2=2:600=d:9=u LL67'2 =2:500=d:7=u sıaj9weied LUOASJUOG 8crl’0 Leee’0 EYLYAO GL67'0 7562'°0 c029’0 GlLLleO 65750 8006 0 5970 LeL9’O Id ıaddn 86100 ceEELO MLSCHO LeLeO 87600 6/8€°0 7960°0 6STE’0 esel’0 LTYT’O c67Y7'0 "19 J9MO] c000°0 8/60°0 E06L’O 0OT/E’0 1760°0 G6EL’0 8850°0 LY9T’O 0€65°0 6/7210 L8YY70 % UL arewı2s3 0200 Teo’el SONAES TEBSSE 9/5°TI EIT’ZT SIS@ H 65202 er6 cl SEIT E01255 yewLgs] puewsg Äanıgejleny u 008 < u 008 > u00/> u 001 > u 00T > ado]s xaAUO) Aıobae) Spn4]V yyed 07 3Jueysiq Aıepunoq 15840) 0} 39ueIsıq Aydeıbodoj uolyeul]9u] abe1aN0) SSOW abe1an09 QI9H abeıaN09 aA] SN0J34LU0I 3]PPLW 1379 WEILA NensQuL 3U9PLYLU0O9 104 spueJs "L ‘I "uolynqugsıp 7435 Bulaye Alanıyısod saUobaJe9 1ayaweıed Juesyuubıs JJe 104 (4/67 ‘je 22 N3N) JeAUSJUI 39Uuapyuo) Luousjuog *7 aJgeL D. TATJANA C. GooD et al. >50 metres \ to nearest path NT 210 close distance to. jresl boundary } a a altitude: \ > ‚soon meiss: a. S. 2217 / close distance to food gisjde is DE y. > 600 metres a.s.l. (Tab. 4; Bonferroni z sta- tistic, p<0.001). No significant difference was obtained for any category of “aspect” other than North, for which fewer setts than expected were found (Tab. 4; Bonferroni z statistic, p<0.05). Figure 1 illustrates the different parameters positively affecting sett sites in Sihlwald. Discussion Sett size and sett type The number of entrances per sett in Sihl- wald, ranging from one to eleven, was well below the average found in the literature (1 to 21: Kruuk 1978; 1 to 38: Anrys and Lisois 1983; 1-80: RopEr 1992a, b), even compared to that of the other studies in Switzerland (1-28: Grar et al. 1996; 1 to 15: FERRARI 1997; 2-23: MONNIER, unpub- lished data: 1-34: DoLinHSan, unpublished data). Badgers seem to prefer burrowing more setts but with fewer entrances in the forest than in the agriculture zone where sett sıtes are restricted to the little patches of forest between the agricultural fields and meadows (DoLınHSan, unpublished topography: convex slopes / IRRE soil: eutric cambisol, \, low herb- and moss coverage, low coniferous layer coverage inclination: Fig. 1. Parameters posi- tively affecting sett site in Sihlwald. Grey ovals: para- meters measured in this study or taken from avail- able data sets for the study area. White ovals: parameters that are diffi- cult to assess but probably influence the choice of sett sites. Arrows indicate ‚ the interrelations between the parameters. gleyic cambisol / data). We suggest that badgers living in Sihlwald can optimise their foraging effi- ciency by using different setts within their home range according to the proximity of the seasonally most profitable food patches. Future analysıs of the seasonal sett-use to- gether with seasonal variations in foraging behavior in the study area will provide the necessary data to test this hypothesis. In Sihlwald, 28.4% of the setts found showed more than two entrances and could indicate main setts (Kruuk 1978). How- ever, their distribution did not differ from random and therefore did not show a spa- cing-out mechanism indicating territories according to the fixed-territory model pro- posed by DoncAsTER and WWOODROFFE (1993) for a high-density badger popula- tion. With regard to the entrance types, it is sur- prising that 31 entrances (11%) were dug under relatively large spruces. Although sett locations have been analysed in several studies, only Bock (1986) classified differ- ent sett types. However, his study did not mention anything about setts dug under spruces. The spruces in Sihlwald were all ın mixed forests. A possible explanation is that spruce trees are normally shallow rooted A habitat analysis of badger (Meles meles L.) setts (KÖsTLEr et al. 1968; BLANCKMEISTER and Hencst 1971), compared to the dominant beech trees in Sihlwald. Shallow roots facil- itate digging; also the roots keep the roof of the sett from collapsing. It is also of interest to note that 22 entrances (7.9%) were dug under a boulder/rock. To our knowledge, single rocks as a possible habitat parameter for sett location, providing shelter and good thermal insulation has only been mentioned in one other study (VIRGos and CASANOVAS 1999). Cover as key factor The habitat parameters affecting the distri- bution of setts in Sihlwald correspond clo- sely to those identified by NEAL (1986) and THoRNToN (1988): digability, hilliness and (tree-) cover. Cover allows the badgers to leave inconspicuously, and it allows the young cubs to play near the entrance with- out being visible to potential predators. A closer look at the vegetation cover around the preferred sett sites in Sihlwald shows that these sites are areas of sparse ground cover (i.e. low herb- and moss coverage). High herb and moss coverage often is cor- related with humidity and therefore avoided by badgers as sett sites. As ob- served in other studies (NEAL 1986; ZEJDA and NESVADBOVA 1983), coniferous stands providing little vegetation cover and found in rather flat areas were avoided in Sihl- wald. The preference for convex slopes with a high inclination (20-40°) as well as the preference for a minimum distance of 50 m from the next path suggest that the variable “cover” is not necessarily equivalent to ve- getation cover; the small-scaled topography around the sett and the distance to the nearest path (just “being out of view”) can indirectly be a type of cover as well. Topo- graphy, i.e., the physical shape of the area in which a sett is dug, is a parameter that has never been stressed in the literature before and seems to play a key role in the choice of sett site in Sihlwald. Paying at- tention to the small-scale topography (850 metres) around the sett appears to be important. Setts dug in convex slopes 211 have several advantages. The badger can pick up scents from different directions without having to leave the security of the sett and thus have several directions from which to leave a sett. It is also possible that setts on convex slopes are easier to enlarge because the rounded shape of the slope gives the badgers more opportunities for digging entrances and connecting tunnels than an unstructured slope. Inclination (“hilliness” according to THORNTON 1988) is also closely associated with topography. Setts are usually dug in slopes (NEAL 1986; SKINNER et al. 1991). The hilliness of the study area is advantageous to the badger in various ways. Digging in a slope facilitates the removal of the excavated soil, which spills down the slope. A particularly favour- able stratum of soil for digging is more ea- sily found on a slope since it is more likely to be exposed. Sloping land is usually well drained so that the sett is more likely to be warm and dry, and in colder parts a depth below ground is quickly attained which is frost free (NEAL 1986). Sett density and population density The density or setts ın Sihlwald (12.3/ 100 ha) is very high compared to the den- sıty of the nearby agricultural zone (2.7/ 100 ha, DoLınHSan, unpublished data). This implies that suitable sett sites are not a limiting factor in the forest, as suggested by Roper (1993) for British areas. Other re- gions of Switzerland (Canton of Neuchätel: 0.02-0.2/100 ha, MOoNnnIER, unpublished data; Canton of Berne: 4.2/100 ha, GRAF et al. 1996) have also lower sett densities. Still, Sihlwald has a significantly lower sett den- sity than found in Britain (up to 26/100 ha, CRESSWELL et al. 1990). The density of pos- sible main setts in Sihlwald (3.5/100 ha) is comparable to that in the high-badger-den- sity areas in Britain (CLEMENTS et al. 1988, see Kowalczyk et al. 2000, for review). Based on the available earthworm biomass, which is the most important food source for badgers in Sihlwald, the minimum popu- lation size is estimated to be 2.5 to 3 indivi- duals per 100 ha (HINDENLANG, unpub- 212 D. TATJANA C. GooD et al. lished data). Therefore, also the badger density in Sihlwald ıs high compared to the published densities across Continental Eu- rope, but, in contrast to the main sett den- sity, lies much lower than the estimated population densities of the British Isles (KowaLczyk et al. 2000). KowAaLczyK et al. (2000) showed that log densities of badger setts correlate negatively with the percent forest cover in the area. This is certainly not the case in Sihlwald where forest covers approximately 70% of the area used by badgers living in the Sihlwald (pers. obser- vation). In the nearby agricultural zone with a much lower sett density forest covers approximately 17% of the area. We argue that this high sett density and considerably high badger density in Sıhlwald is attained through a combination of ideal sett-site conditions as well as a rich and varied food supply. According to the literature, a mix- ture of woodland and pastures, and wood- land and arable land is among the habitat types preferred by the badger (BRrosETH et al. 1997; HorEr 1988; NEAL 1977, ZEDJA and NESVADBOVA 1983). Also, the setts in Sihlwald are found significantly closer to the forest boundary and adjoining agricul- tural zones than random points. Other stu- dies have noted that badger setts tend to be sıtuated close to habitat edges, i.e. on boundaries between two habitat types (O’CoRRY-CRoOWE et al. 1993; VIRGos and Zusammenfassung CasanovAs 1999). The proximity of setts to the forest boundary and adjoining agricul- tural zones makes access easier to an opti- mal food supply year-round without forfeit- ing optimal sett sites that the forest offers with ıts pronounced topography. The di- verse pattern of mixed deciduous forest stands in Sihlwald itself contains good worm patches even in dry periods (HiNDEN- LANG, unpublished data). Badgers can therefore adapt their foraging to the seaso- nal changes in food availability, both within the mixed forest stands and in the agricul- tural fields and meadows outside the forest. We suggest that the spatial organisation of badgers living in the Sihlwald area is pri- marily determined by the seasonal avail- ability of food resources. Acknowledgements We thank GERULF RIEGER and WERNER SUTER for the helpful comments on the drafts. Furthermore, we would like to thank LoRENnZ GyGAx and WOLF BLANKENHORN for the statistical advice and SU- SANNE REIMANN for helping us draw figure 1. We are also very indebted to all those who helped us find numerous setts. K. HINDENLANG was Sup- ported by the Swiss National Fund (“Adjustment of spatial behaviour of European badger (Meles meles L.) in response to a heterogeneous environ- ment”, Nr. 3100-40846.94/3100-52988.97). Eine Analyse der Habitatcharakteristika von Dachsbauen (Meles meles L.) in einem naturnahen Wald Größe, Verteilung und Habitatcharakteristika von Dachsbauen wurden in einem naturnahen Wald un- tersucht. Die Verteilung der Dachsbaue im Untersuchungsgebiet war nicht zufällig, wobei jeweils die Distanzen zum nächst benachbarten Bau mit den Distanzen zu Zufallspunkten verglichen wurden. Für die Bestimmung der charakteristischen Habitatfaktoren, die die Verteilung der Baue im Untersu- chungsgebiet erklären, wurden verschiedene Kategorien von Habitatparametern für Topographie, Ve- getation und Struktur des Waldhabitats mittels Multipler Regressions-Analyse und mithilfe eines digi- talen Geländemodells in einem Geographischen Informations-Systems (GIS) analysiert. Bevorzugte Standorte waren konvexe Hangrippen mit einer Inklination zwischen 20° und 40°. Sie sind gut ent- wässert und bieten dem Dachs die Möglichkeit, Baueingänge und -röhren zu graben, die ein Verlassen des Baus und das Aufnehmen von Witterung aus verschiedenen Himmelsrichtungen erlauben. Bevor- zugte Standorte für Dachsbaue befanden sich in Höhenlagen über 600 Meter ü.M. sowie näher am A habitat analysis of badger (Meles meles L.) setts 213 Waldrand und damit auch näher an den umliegenden landwirtschaftlich genutzten Flächen als die Zu- fallspunkte. Diese Baue gewährleisten den Dachsen das ganze Jahr über Zugang zu einem optimalen Nahrungsangebot, sei es in den vielfältigen Mischwaldbeständen innerhalb des Waldes oder sei es auf den landwirtschaftlich genutzten Äckern und Wiesen außerhalb des Waldes. Die Dachse können so ihre Nahrungssuche den saisonalen Veränderungen im Nahrungsangebot optimal anpassen. Baue wurden meistens in Flächen mit wenig Bodenbedeckung und weiter als 50 Meter vom nächsten Weg entfernt gefunden. Nadelwaldbestände wurden gemieden. Einzelstehende, alte Fichten (Picea abies) in Laubwaldbeständen jedoch wurden oft als Standorte für Baue mit ein bis zwei Eingängen genutzt. Ihr flaches Wurzelwerk erleichtert das Graben und verhindert das Einstürzen des Baues. Ein hoher Deckungsgrad der Vegetation und damit ein guter Sichtschutz spielte eine wichtige Rolle für die Aus- wahl der Baustandorte. Sichtschutz kann jedoch auch durch andere Faktoren, wie Topographie und Entfernung zum nächsten Weg, gewährleistet werden. Insbesondere die kleinräumig diverse Topogra- phie um die Baue scheint ein bisher wenig beachteter Schlüsselfaktor für die Anlage eines Baues zu sein. Die Ergebnisse dieser Studie sprechen dafür, dass ein großes Laubwaldgebiet mit einer stark ausgeprägten Topographie ein sehr gutes Dachshabitat darstellt. References AnRYS, P.; LiBoIs, R. M. (1983): Choix de l’habitat chez le blaireau europe&en (Meles meles L.) en Belgique. Cah. Ethol. Appl. 3, 15-38. BEKOFF, M.; DANIELS, T. J.; GITTELMAN, J. L. (1984): Life history patterns and the compara- tive social ecology of carnivores. Ann. Rev. Ecol. Syst. 15, 191-232. BLANKMEISTER, J. ; HENGST, E. (1971): Die Fichte im Mittelgebirge. Berlin: Neumann-Neudamm. Bock, W. F. (1986): Kriterien zur Größenbeurtei- lung der Baue des Dachses (Meles meles L.). Säugtierkdl. Mitt. 33, 227-234. BRAUN-BLANQUET, J. (1964): Pflanzensoziologie. Wien, New York: Springer. BROSETH, H.; KNUTSEN, B.; BEVANGER, K. (1997): Spatial organization and habitat utilisation of badgers Meles meles: effects of food patch dis- persion in the boreal forest of central Norway. Z. Säugetierkunde 62, 12-22. CHEESEMAN, C. L.; CRESSWELL, W. J.; HARRIS, S.; MALLINSON, P.J. (1988): Comparison of dis- persal and other movements in two badger (Meles meles L.) populations. Mammal Rev. 18, 51-59. CHEESEMAN, C. L.; JONES, G. W.; GALLAGHER, J.; MALLINSON, P. J. (1981): The population struc- ture, density and prevalence of tuberculosis (Mycobacterium bovis) in badgers (Meles meles L.) from four areas in Southwest Eng- land. J. Appl. Ecol. 18, 795-804. CHEESEMAN, C. L.; WILESMITH, J. E.; Ryan, ]).; MALLINSoN, P.J. (1987): Badger population dynamics in a high-density area. Symp. Zool. Soc. Lond. 58, 279-294. CLEMENTS, E. D.; NEAL, E.; YALDEN, D. W. (1988): The ndational badger sett survey. Mammal Rev. 18, 1-9. CRESSWELL, W. J.; HARRIS, S. (1988): Foraging be- haviour and home range utilisation in a sub- urban badger (Meles meles L.) population. Mammal Rev. 18, 37-49. CRESSWELL, P.; HARRIS, S.; JEFFERIES, D. J. (1990): The history, distribution, status and habitat re- quirements of the badger in Britain. Peterbor- ough: Nature Conservancy Council. DONCASTER, C. P.; WOODROFFE, R. (1993): Den site can determine shape and size of badger terri- tories: Implications for group-living. Oikos 66, 88-93. FERRARI, N. (1997): Eco-Ethologie du blaireau eur- opeen (Meles meles L., 1758) dans le Jura suisse: comparaison de deux populations vi- vant en milieu montagnard at en milieu cul- tive de plaine. PhD thesis. Univ. of Neuchätel, Switzerland. GITTELMAN, J.L. (1986): Carnivore life history patterns: Allometric, phylogenetic and ecolo- gical associations. Am. Nat. 127, 744-771. GRAF, M.; WANDELER, A. 1.; Lürs, P. (1996): Die räumliche Habitatnutzung einer Dachspopu- lation (Meles meles L.) im Schweizerischen Mittelland. Rev. Suisse Zool. 103, 835-850. HoFER, H. (1988): Variation in resource presence, utilisation and reproductive success within a population of badgers (Meles meles). Mam- mal Rev. 18, 25-36. KÖSTLER, J. N.; BRÜCKNER, E.; BIBELRIETHER, H. (1968): Die Wurzeln der Waldbäume. Unter- 214 D. TATJANA C. GOOD et al. suchungen zur Morphologie der Waldbäume in Mitteleuropa. Hamburg, Berlin: Paul Pa- Tey. KOoWwALCZYK, R.; BUNEVICH, A.N.; JEDRZEJEWSKA, B. (2000): Badger density and distribution of setts in Bialowieza Primeval Forest (Poland and Belarus) compared to other Eurasian po- pulations. Acta Theriol. 45, 395-408. KRUUK, H. (1978): Spatial organisation and terri- torial behaviour of the European badger, Meles meles. J. Zool. (London) 184, 1-19. KRUUK, H; PArısH, T. (1982): Factors affecting po- pulation density, group size and territory size of the European badger (Meles meles L.). ). Zool. (London) 196, 31-39. KRrUUK, H.; PARISH, T. (1987): Changes in the size of groups and ranges of the European badger (Meles meles L.) in an area in Scotland. ]. Anim. Ecol. 56, 351-364. NEAL, E. (1977): Badgers. Poole, Dorset: Blanford Press Ltd. NEAL, E. (1986): The Natural History of Badgers. London: Croom Helm. NEU, C. W.; BYERS, C.R.; PEEK, J.M. (1974): A technique for analysis of utilisation-availabil- ity data. J. Wildl. Manage. 38, 541-545. NIEVERGEIT, B. (1981): Ibexes in an African Envir- onment. Ecol. Studies 40. Berlin: Springer. NoLEkrt, B. A.; KiLLinGLEy, C. A. (1986): The ef- fects of a change in food availability on group and territory. Lutra 30, 1-8. O’CORREY-CROWE, G.; EVvEs,J; HAYDEN,T.]. (1993): Sett distribution, territory size and po- pulation density of badgers (Meles meles L.), in East Offaly. In: The Badger. Ed. by T. J. HAyDEn. Dublin: Royal Irish Academy. Pp. 35-56. Pıcozzı, G. (1989): Latrine use and the function of territoriality in the European badger, Meles meles L., in a Mediterranean coastal habitat. Anim. Behav. 39, 1000-1002. RODRIGUEZ, A.; MARTIN, R.; DELIBES, M. (1996): Space use and activity in a Mediterranean po- pulation of badgers, Meles meles L. Acta Theriol. 41, 59-72. ROPER, T. J. (1992 a): The structure and function of badger setts. J. Zool. (London) 227, 691- 694. Roper, T. J. (1992 b): Badger, Meles meles L., setts — architecture, internal environment and func- tion. Mammal Rev. 22, 43-53. ROPER, T. J. (1993): Badget setts as a limiting re- source. In: The Badger. Ed. by T. J. HAYDEn. Dublin: Royal Irish Academy. Pp. 26-34. ROoPER, T.J.;, SHEPHERDSON, D.J.,;, DAvIEs, J.M. (1986): Scent marking with faeces and anal se- cretion in the European badger (Meles meles): Seasonal and spatial characteristics of latrine use in relation to territoriality. Behaviour 97, 94-117. SKINNER, C. A.; SKINNER, P. J. (1988): Food of bad- gers (Meles meles L.) in an arable arca of Es- sex. J. Zool. (London) 215, 360-362. SKINNER, C.; SKINNER, P.; HARRIS, S. (1991): An analysis of some of the factors affecting the current distribution of badger (Meles meles L.) setts in Essex. Mammal Rev. 21, 51-65. STUBBE, M. (1980): Population ecology of the Red Fox — Vulpes vulpes (L. 1758) - in the DDR. In: The Red Fox: Symposium on Behaviour and Ecology. Ed. by E. ZımEn. Biogeographi- ca 18. The Hague: Junk. Pp. 71-96. THORNToN, P.S. (1988): Density and distribution of badgers in south-west England - a predic- tive model. Mammal Rev. 18, 11-23. VIRGÖS, E.; CASANOVASs, J. G. (1999): Badger Meles meles sett site selection in low , density Medi- terranean areas of central Spain. Acta Ther- iol. 44, 173-182. Wiırson, E.O. (1975): Sociobiology: The New Synthesis. Cambridge: Harvard University Press. WOODROFFE, R.; MACDONALD, D. W. (1992): Bad- ger clans: Demographic groups in an antiso- cial species. J. Zool. (London) 227, 696-698. WOODROFFE, R.; MACDONALD, D. W. (1993): Bad- ger sociality — Models of spatial grouping. Symp. Zool. Soc. Lond. 65, 145-169. ZEDJA, J.; NESVADBOVA, J. (1983): Habitat selection and population density of the badger (Meles meles L.) in Bohemia and Moravia. Folia Z00l. 32, 319-333. Authors’ addresses: TATJANA C. GoopD, Department of Ecology and Evolutionary Biology, Princeton University, Prin- ceton, NJ 08544-1003, USA; KARIN HINDENLANG, Swiss Federal Institute for Forest, Snow and Landscape Research (WSL), Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland; STEPHAN IM- FELD, Spatial Data Handling Division, Geogra- phical Institute, University of Zurich-Irchel, Win- terthurerstrasse 190, CH-8057 Zurich, Switzer- land; BERNHARD NIEVERGELT, Wildlife Research and Conservation Biology, Zoological Institute, University of Zurich-Irchel, Winterthurerstrasse 190, CH-8057 Zurich, Switzerland. Mamm. biol. 66 (2001) 215-227 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Age and sex distributions in the catches of belugas, Delphinapterus leucas, in West Greenland and in western Russia By M. P. HEIDE-JORGENSEN and CHRISTINA LOCKYER Greenland Institute of Natural Resources, Nuuk, Greenland and Danish Institute for Fisheries Research, Charlottenlund, Denmark Receipt of Ms. 14. 02. 2000 Acceptance of Ms. 06. 11. 2000 Abstract Age and sex were determined for belugas or white whales, Delphinapterus leucas, harvested in West Greenland in 1985-86 and 1989-1997. There was a clear segregation of whales in the drive fishery conducted during autumn in Qaanaaq and Upernavik. Primarily immature whales of both sexes to- gether with mature females were taken. Age was estimated from Growth Layer Groups (GLGs) in sectioned teeth, assuming the currently accepted criteria of 2 GLGs forming annually. The mean and median ages were increasing slightly in both sexes from Upernavik from 1985 through 1994. Both immature and mature whales were taken on the wintering grounds from Disko Bay and south. Estimation of survival was confounded by the large number of whales where only a minimum age could be assigned because of tooth wear at the crown (i.e. no neonatal line in the dentine). The apparent survival rates for belugas from West Greenland were estimated as 0.81 and 0.79 for fe- males and males, respectively. Correction of these estimates for an observed population decline of 4.17% per year revealed true survival rates of 0.85 and 0.82 for females and males, respectively. The estimates of true survival rates are less than those determined for beluga populations in the White and Kara seas and in Alaska for comparable age truncations. Since the exploitation levels are much lower in these areas the low apparent survival rate from West Greenland strongly supports the evidence of a population decline. Colour change from grey to white occurs at mean ages of 8.5 yr and 9.1 yr and median lengths of 367 cm and 445 cm in females and males, respectively. Key words: Delphinapterus leucas, age structure, Greenland, Russian Arctic Introduction Large numbers of belugas (white whales), Delphinapterus leucas, have been taken in commercial fisheries and Inuit harvests throughout the Arctic during the last 100 years. In Greenland the accumulated catches of belugas in this century amounts to more than 50000 whales (HEIDE-J&ORGEN- 1616-5047/01/66/04-215 $ 15.00/0. sEN 1994). Despite the availability of sam- ples verv little is known about the sex and age frequencies in different harvesting si- tuations; let alone survival rates calculated from catch-at-age data. The belugas that are harvested along West Greenland are believed to be part of the 216 M. P. HEIDE-JORGENSEN et al. stock(s) of belugas that summer in the Cana- dian High Arctic (HEIDE-JORGENSEN 1994). This stock is supposed to winter in West Greenland south of Disko Bay and aerial sur- veys of the population densities of belugas on these wintering grounds have indicated a sub- stantial decline between 1982 and 1994 (HEIDE-JORGENSEN and REEVES 1996). The most likely explanation for this apparent po- pulation reduction is that the large level of exploitation during the 1980s has exceeded the replacement yield of the stock. With this background it was considered important to further evaluate the status of the stock. SERGEANT (1973) provided the first age fre- quencies from catches of belugas in Hudson Bay and while he realised the detrimental impact of tooth wear on age estimations he was able to show that maximum longevity was at least 25 yr for both sexes, under the assumption of deposition of two Growth Layer Groups (GLGs) per annum. BURNS and SEAMAN (1986) provided age frequen- cies from Alaska with a maximum life span of 38 yr but did not present details on the distribution for males and females. DoIDGE (1990) showed from age frequencies ob- tained in Northern Quebec that belugas have a maximum longevity of at least 31 and 33 yr for males and females, respec- tively. By use of a smoothing technique of the age frequencies from teeth without wear DoipGe (1990) estimated annual survi- vorships for both sexes combined ranging from 0.69 to 0.93 for whales between 0 and 9 yr of age and between 0.94 and 0.97 for whales between 10 and 37 yr of age. Survi- vorship estimates for worn and unworn teeth lumped together were slightly lower for all age classes. The survival rates were slightly higher in Northern Quebec com- pared to Alaska especially after age 12 yr, but both studies were evidently violating the assumptions of both equilibrium of the population prior to sampling and represen- tativeness of samples for the population age frequencies. This study examines the changes in the age- and sex selectivity of the harvesting of belu- gas in West Greenland. The survival rates calculated for West Greenland belugas are compared to survival rates from belugas in the White and Kara seas. Material and methods Age frequencies from West Greenland Lower jaws were collected from the catches of be- lugas in West Greenland (Fig. 1). The jaws were registered and stored at freeze houses and kept frozen before and during shipment to the labora- tory in Copenhagen. Teeth were extracted and stored frozen until sections were prepared accord- ing to methods described in HEIDE-J&ORGENSEN, et al. (1994). Two trairfed readers independently counted the number of Growth Layer Groups (GLGs - see PERRIN and Myrıck (1980) for defi- nition of GLG) in the teeth and age was deter- mined as the mean of the two readings. ‘Minimum age’ refers to those teeth where an unknown number of GLGs were missing due to wear of the crown of the tooth. ‘Complete age’ refers to those teeth in which all GLGs, including the neo- natal line, could be counted. Two GLGs were as- sumed to be deposited annually (HEIDE-JORGEN- SEN et al. 1994). The same two persons made all the readings of GLGs except for the samples from 1993-94 where another trained reader replaced one of them. If the discrepancy between the two readers exceeded 3 GLGs the tooth was checked again by one of the readers and if the discrepancy was maintained, the age estimate would either be discarded or another tooth was prepared. Sex was determined by DNA analysis of skin samples extracted from the lower jaws (PALSBOLL et al. 1992). The samples collected were stratified with regard to sex, area, and year. For each area and year, the mean, its standard deviation and the median of ages were calculated for both sexes. This was done separately for samples restricted to whales with minimum age and to whales with complete age. Colouration of skin was classified to three groups; brown or brownish-grey, grey and white, for a subsample of 147 females and 119 male belugas that were examined before being flensed. Classifi- cation of skin colouration from the skin remains on the samples of lower jaws alone proved to be unreliable, whereas all the whales with length measurements had been examined by the first author before they were flensed. For the change in colouration from grey to white colouration mean age and associated variance was deter- mined by use of the method of DEMASTER (1978) after smoothing of a curve fitted to the data. Age and sex distributions in the catches of belugas Age frequencies from the White and Kara seas Age frequency data from the beluga harvest in the White and Kara seas was compared with the age distribution from West Greenland. A sample of 570 whales was collected from the commercial € 2 Upernavik $ ER. r Uummanna 8 Nu Disko Bay (3 55W | | 217 hunt in the White and Kara seas in the 1970s and early 1980s (see OGnETov 1981 for a description of a subset of the sample). The growth of belugas in the White and Kara seas resembles the West Greenland belugas more than any other beluga population (HEIDE-JORGENSEN I . 0 Ss 3 \Ü \ \ RO x som win, 65'N AO'W 60'N | Fig. 1. Map of municipalities and areas in West Greenland mentioned in the text. 218 M. P. HEIDE-J@RGENSEN et al. and TEILMAnN 1994) although their tooth wear begins at a much later age (HEIDE-JORGENSEN et al. 1994). Estimation of survival rates Apparent survival rates (i1.e. survival rates uncor- rected for population changes) were estimated for the age frequencies from West Greenland and the White and Kara seas according to the method of Roßson and CHAPMAN (1961). Because of underrepresentation of the youngest age classes in all samples, all age frequencies were re- coded from the modal age class upwards. Appar- ent survival rates were also calculated by fitting a negative exponential curve where the natural logarithm of the exponent is equivalent to the an- nual survival rate. Statistical analysis Data management and comparisons of means were conducted in Statview for Windows (ver- sion 5.0) and non parametric tests and non linear regressions were carried out in S-Plus (ver- — FEMALES - WEST GREENLAND BEE Complete ages EZZ Minimum ages mu Fitted values FREQUENCY 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 FEMALES - WHITE AND KARA SEAS DD [@)1 DD >) FREQUENCY S Ua [97 0 2 4 6 8 1012 14 16 18 20 22 24 26 28 30 AGE sion 4.5). Unless otherwise stated, significance was always determined at the 5% level. Results Selectivity in the beluga harvest in West Greenland All sex and age classes of belugas are sub- ject to harvesting in West Greenland (Fig. 2). Sampling during ten years between 1985 and 1997 resulted in an overall mean age of 7.7 yr in females and 6.5 yr in males of the harvested population older than l year in all municipalities and minimum and complete ages combined. In the sam- ples more females than males were taken (712 vs. 596), but there was an equal pro- portion of both sexes among calves less than 1 year of age in Greenland (44 fe- males, n=89). The limited number of samples from the municipality of Qaanaaq (formerly called MALES - WEST GREENLAND 9 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 AA... MALES - WHITE AND KARA SEAS 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 AGE Fig. 2. Age-frequency distributions of male and female belugas from West Greenland and the White and Kara seas. Frequencies of complete and minimum ages are shown together with the negative exponential fit (F= a*exp(b*AGE)) from the modal age class. See table 2 for values of parameters from fitted functions. Age and sex distributions in the catches of belugas Avanersuag, see Fig. 1) collected in Sep- tember 1993 showed that the catch primar- ily consisted of young animals (Tab. 1). The mean age of the samples from the munici- pality of Upernavik collected in late Sep- tember and early October 1985 through 1994 or in 1993 alone were not significantly different from those collected in Qaanaaq (ANOVA, p>0.2). The samples of whales taken in Uummannaqg in the autumns of 1993 through 1996 indicated similar age structure as in Qaanaaq and Upernavik. Mainly young whales were taken in all three areas. When contrasting the age and sex frequen- cies obtained from the autumn drive fish- eries in Qaanaag, Upernavik and Uumman- naq, with that from the ice edge and open water hunt in Disko Bay and south, it is ob- vious that older whales are taken in Disko Bay (Tab. 1). This was evident both from comparisons of mean ages from all whales (older than 1 year) and from those with complete ages (t-test). Also, the samples collected from winter and spring catches in Disko Bay showed that less than 30% of the whales could be aged without bias (complete age) in contrast to Upernavik and Qaanaaq where more than 60% had complete age (Tab. 1). Trends in mean age in the catches Linear regressions of mean age on time showed that there was a significant increase in the mean age of harvested female and male belugas for which minimum age had been determined from 1985 through 1994 from Upernavik (ANOVA). The increase was also evident when groups where mini- mum and complete age had been deter- mined were combined but could not be con- firmed for females when complete age was considered alone. For the area from Disko Bay and south no signifcant trends in mean age could be detected between 1990 and 1997 for minimum age or minimum and complete ages combined. However, a signif- icant increase was observed using complete age data for males but not for females (p=0.0588). Since the minimum ages tend 219 to be truncated at the age when the neona- tal. is worn away this category of teeth is less reliable for detecting an increase in the mean age of catches. Thus the category with complete ages should be given more weight, which implies that there is a tendency for older males being harvested in both Uper- navik and Disko Bay since the mid 19808. Comparison with age distribution in the White and Kara seas The observed potential minimum life span of 30 yr in belugas from West Greenland was similar to what was found in the sample from the harvest in the White and Kara seas. This was, however, the only common feature of the two age distributions (see Fig. 2). Whereas immature belugas consti- tuted the largest number of samples in West Greenland, mature whales of more than 7 yr constituted the majority of the samples from the White and Kara seas, where the mean ages also deviate significantly from West Greenland (Tab. 1). Because of the late onset of tooth wear in the belugas from the White and Kara seas less than 16% of the teeth had lost their neonatal line. Survival rates of belugas As tooth wear increases with age indivi- duals with incomplete (minimum) age esti- mates are much more frequent in older age classes. Thus the apparent survival rates were generaily much higher when indivi- duals with minimum ages were included. Moreover, inclusion of minimum age esti- mates still leads to an underestimate of sur- vival rates because the minimum age esti- mations underestimate the true age of the whales. No significant differences were detected for apparent survival rates (age>9, t-test) or frequency distributions of all age samples from north and south of Disko Bay (Kol- mogorov-Smirnov two sample test). Thus the two data sets were pooled. The combined age distribution from West Greenland showed an underrepresentation of age classes O and 1 (Fig. 2), thus for the M. P. HEIDE-JORGENSEN et al 220 sobe 979]d vVezzU109 sajew SEJLINIEN 0861 -0/61 Sv3S vavA ONV 3LIHM sojeWLIOZ L661-9861 HLNOS ANY Ava ONSIA sajeuwug, 9661-661 DVNNVWWNN sabe 979]d Nv wo) sajey sajew1g 7661-5861 AIAVNAIAN sajew NV abe 975] 9/ -W09 YAM % 9T abe xeyw iv N GG UuELPOW 68 qS eg ueoWw sabe 979]0 -W0) sojewuag DVVNVVD abe ‚»23JdWwos, pue ‚wnwıuLW, UM 4393} YJ0Q 04 S13484 „IV, 'SO86T pue sQ/6T ay UL Seas Bley Pue aJLlym au} UL pue /66T Pue G86T UEaMIOq PueJuasıg JsaM UL Seaie JUalayıp ANOJ ul Palsanley AA T ey} Aopjo sebnjaq Jo sabe ueaw °T aJgeL Age and sex distributions in the catches of belugas 221 calculation of survival rates these age classes were excluded. The age distributions from the White and Kara seas were radically different from what is seen in West Greenland. Whereas West Greenland age distribution peaks at 2 yr of age for both sexes the samples of be- lugas from the White and Kara seas were almost constant from age 0 to the peak at 8 yr in females and 10 yr in males (Fig. 2). Estimation of apparent survival rates for the White and Kara seas was thus con- ducted for the age frequencies older than 7 yr and 9 yr for females and males, respec- tively (Tab. 2). The apparent survival estimates for the White and Kara seas were larger, albeit not significantly larger, than the estimates for West Greenland for similar truncations of the age classes (Tab. 2). However, signif- icant difference (t-test) was only detected for the survival rate for male belugas estimated from the negative exponential model. Change in colouration Some beluga calves are born brown and others are grey, but usually the brown col- ouration changes at an earlier age (<5 yrs) and shorter length (<300 cm) than the grey colouration (Figs.3 and 4). There is ob- viously a large overlap in the brown (or brownish-grey) and the grey classification which probably indicates diffculties in dis- tinguishing between these two categories. The colour change from grey to white seems to be completed at a mean age of 8.5 yr for females (SD=0.6) and 9.1 yr for males (SD=0.5) and a median length of 367 cm and 445 cm for females and males, respectively (Figs. 3 and 4). Discussion Reliable calculation of survival rates for be- lugas taken in harvest operations depend on a variety of assumptions: Table 2. Survival estimates for belugas from West Greenland and the White and Kara seas estimated by the meth- od from Roßson and CHAPMAN (1961) and by fitting a negative exponential curve (frequency = a*exp(b*age)). Only age classes older than 1 year of age are included in the estimations for West Greenland. For the White and Kara seas only age classes older than 7 yr for females and 9 yr for males, respectively, are included. 95% Cl are shown in parenthesis FEMALES - WEST GREENLAND Recoded ages from >1 yr (n = 625) Recoded ages from >9 yr (n = 232) MALES - WEST GREENLAND Recoded ages from >1 yr (n = 500) Recoded ages from >9 yr (n = 123) FEMALES - WHITE AND KARA SEAS Recoded ages from >7 yr (n= 152) MALES - WHITE AND KARA SEAS Recoded ages from >9 yr (n= 161 ) RoBSON and CHAPMAN SURVIVAL RATES Negative exponential model 0.86 (0.85-0.88) 0.81 (0.79-0.83) 0.83 (0.82-0.85) 0.79 (0.75-0.82) 0.84 (0 .82-0.87) 0.84 (0.81-0.86) 0.88 (0.85-0.90) 0.84 (0.80-0.87) 0.85 (0.83-0.87) 0.73 (0.69-0.78) 0.87 (0.85-0.90) 0.85 (0.83-0.87) 222 M. P. HEIDE-JORGENSEN et al. NUMBER OF WHALES NUMBER OF WHALES FEMALES EU] BROWN GREY H WHITE 45 6 7 8 9 10111213141516171819 2021 22 23 24 MALES EO] BROWN E GREY HM WHITE 01234567 8 9 101112131415161718 192021222324 AGE (GLGs/2) Fig. 3. Changes in colour phases in relation to age for female and male belugas from West Greenland. . Sımilar or identical methods for age esti- mation must be employed and prefer- ably the techniques need validation from known age animals. The samples collected from the harvest should be representative of the popula- tion and bias - if any - should be clearly discernible. The survival need to be constant over time for the age groups involved in the estimation of survival and the population should be constant over the period when the age material was collected. Age estimation technique Identical methods for age estimation were used for all samples from Greenland and at least one person made readings of all teeth. Thus it seems reasonable to assume that the samples from Greenland are con- sistent with regard to age estimation and reading of GLGs. It is uncertain if the Rus- sian samples are entirely comparable with those from Greenland. For validation of the age estimation method see HEIDE-J@R- GENSEN et al. (1994). Brown Grey White Brown Grey White Age and sex distributions in the catches of belugas 223 * ee + + Md=275cm FEMALES + BZ I z Md=306 cm Md=367 cm or + 150 200 250 300 350 400 450 500 550 LENGTH (cm) Be nz Md=270 cm MALES .. en %+ Md=300 cm Md=445 cm . u 150 200 250 300 350 400 450 500 550 LENGTH (cm) Fig. 4. Changes in colour phases in relation to length for female and male belugas from West Greenland. Md = median of cumulated numbers in each colour classification. Size of diamonds indicate the relative number of animals. Bias in samplings as the selectivity varies with different hunt- ing methods and seasons. The catches in The samples collected for this study do not OQaanaaq and Upernavik are taken in a have an unbiased representation of the age drive fishery during a short period, usually and sex structure of the beluga population the last week of September in Oaanaaq or 224 M. P. HEIDE-JORGENSEN et al. the first week of October in Upernavik. The catches usually consist of entire herds of belugas ranging in size from 20 to 250 whales. The catches can be considered as random samples of the segment of the beluga population that migrates south in a coastal corndor. There may be different composition of the herds encountered and there is probably some selectivity towards larger herds. However, the samples from the catch in Upernavik can at least be ex- pected to reflect the annual composition of that particular fishery better than any of the other samples. According to HEIDE-JÖRGENSEN et al. (1994) mean age at tooth wear, i.e. when neonatal line disappears, for female and male belugas from West Greenland is 7.7 and 6.0 yr of age, respectively. For the samples from Upernavik this implies that less than 30% of the samples were from whales that exceed these ages. It also indicates that females are generally older for the Upernavik samples than the males, because they have a higher mean age but a si- _ milar proportion of ‘minimum age’ classifica- tions. Age at sexual maturity in belugas from West Greenland is around 4-7 yr for females and 6-7 yr for males (HEIDE-JORGENSEN and TEILMAnNn 1994). Thus few mature males were taken in Upernavik, whereas some ma- ture females were included in the catches at least until 1992. In 1993 and 1994 an increas- ing proportion of mature anımals of both sexes were taken. This fact and the apparent increase in mean age of whales taken in Upernavik may be a result of the intensive exploitation of the younger age segments of the population. The whales that were caught in 1993 and 1994 were taken around 1 Octo- ber as in previous years, few whales were available to be driven and nothing suggested any decrease in hunting effort (HEIDE-JoR- GENSEN own observations). The gradual shift to older segments of the population may be explained by a depletion of young whales that exceeded the recruitment to the popula- tion. The hypothesis of a general decline in the beluga population is supported by results from aerial surveys conducted at the winter- ing grounds between 1981 and 1994 (see HEIDE-JORGENSEN and REEVES 1996). The mean age of the whales taken in the autumn catches in Upernavik in 1993 was indistinguishable from the mean age of be- lugas taken in Qaanaag earlier in the same season. Both samples were similar in their age distribution to samples collected at Grise Fjord presumably in the autumn (STEWART 1994). Because there were tem- poral differences in age classes taken in Upernavik, the selectivity in the harvest is likely due to different availability of certain age classes during the autumn migration. The hunt in Disko Bay is conducted from powered boats in open water during No- vember through January or from the ice edge in spring during March through May. Mature whales of both sexes have for sev- eral of the years been found on the winter- ing grounds around Disko Bay. Since the open-water and ice edge hunt in this area is characterised by catches of small herds (<10) of whales or single animals it may be considered to reflect the age structure of the population more randomly than the au- tumn drive fishery for large herds. This is because pods of mature males are often se- parated from mature females with young of both sexes. Changes in population size Estimation of survival rates is confounded by both selectivity in catches and a large proportion of whales with minimum ages which results in underestimates of maxi- mum life span and survival. However, due to differences in tooth wear (HEIDE-JQ@R- GENSEN et al. 1994) the sample from the White and Kara seas have less whales with incomplete age and probably reflect the true age more accurately. Apparent survival rates calculated for West Greenland includ- ing only females >7 yr and males >9 yr in- dicated survival rates that were lower albeit not significantly lower than those estimated for the White and Kara seas. Some of the difference between the two areas can be attributed to differences in tooth wear. The largest catches of belugas in the White and Kara seas took place before 1966 when vessel hunting was stopped (OGNETOV pers. Age and sex distributions in the catches of belugas comm.). During the period when the sam- ples were obtained the beluga population in the White and Kara seas had been sub- ject to some harvesting although exact sta- tistics are not available. The Report of the International Whaling Commission (1982) list catch figures between 135 and 672 belu- gas per annum for 1976-1982 with years without catches in the White, Barents and Kara seas. Other researchers estimate that catches in this area during 1975-1980 re- mained below 1200 whales (OÖGNETOV pers comm.). In the 1980s catches have been dwindling (according to IWC reports from 1985 to 1991) and ceased after 1990 (OGnE- ToV pers. comm.). The reasons for the de- clining catches are more likely related to a reduced economical incentive for selling beluga products than to a reduced availabil- ity of the resource (STANISLAV BELIKOV pers comm., OGNETOV pers. comm.). Recently it has been suggested that the beluga popula- tion in the western Russian Arctic is in a non-declining state and could potentially support a harvest of up to 530 belugas (OGNETOV pers. comm.). During the 1980s catches have remained high in Greenland with an annual reported mean of 740 whales and there are no signs of a decline in hunting effort or in the eco- nomical incentives for the hunt (HEIDE-J@R- GENSEN 1994). But aerial surveys of popula- tion densities on the wintering grounds in West Greenland indicate a linear decline of 4.7% (95% Cl: 2.1-7.2) per year between 1981 and 1994 (HEIDE-JORGENSEN and Reeves 1996). This population decline will severely affect the apparent survival rates calculated from the age material collected from the harvest. The apparent survival rates (q) can be converted into true annual survival rates (p) by p = q/A, for the Robson and Chapman method and by p = exp(Ing + (A -1)) for the exponential model where A, is the observed rate of population change. The true survival rates thereby become 0.85 and 0.83 for females and males, respec- tively, for the Robson and Chapman meth- od applied to the belugas older than 9 yr in West Greenland. Similarly the true survival rates for the exponential model become 225 0.88 and 0.77 for females and males, respec- tively. These estimates are closer to the esti- mates derived from the age samples from the White and Kara seas. In a study of harvested belugas in Alaska, Burns and SEAMAN (1986) estimated age for 528 belugas collected between 1977-79 and 1980-83 and found several whales older than 30 yr and two as old as 38 yr. From the distribution of 332 males and females older than 5 yr they estimated a mean mortality rate of 0.094, equivalent to an annual survi- val of 0.91 for both sexes. This estimate was based on both minimum and complete age classifications and is therefore again an un- derestimate of true survivorship. The onset of tooth wear may occur a few years later in Alaska than in Greenland, but that is probably not the main explanation for the large and significant difference in survival in the two areas. The catches in Alaska have been reported to be between 241 and 345 belugas per year during 1980 to 1983 or an estimated 1.9% to 2.6% of the provi- | sional population estimates and with no in- dications of a population decline over that period (Lowry et al. 1989). Compared to the exploitation status in West Greenland it thus seems reasonable to assume that the apparent survival rates calculated for Alas- kan belugas are less biased than estimates of survival in West Greenland. If the West Greenland age distribution is recalculated from age 5 yr (as for the Alaskan sample), females and males combined and corrected for population change with the negative ex- ponential model, then the true annual survi- val rate becomes 0.9136, which is similar to the estimate from the far less exploited Alaskan population. Changes in colour phase Our estimates of change in colour phases are not entirely comparable to estimates from other studies, because several other researchers (e.g. BURNsS and SEAMAN 1986) have used four colour phases rather than the three as chosen for this study. We found that using three colour phases was less sub- 226 M. P. HEIDE-JARGENSEN et al. jectively based. The age for change from grey to white as reported from other areas is fairly similar to what we have seen in West Greenland (SERGEANT 1973; OGNETOV 1981), but detailed comparisons are not possible because the statistical methods used for deriving the mean age at change of colour are not specified. However, for belugas in East Baffin, BropiE (1971) re- ported that whitening occurs after 6 and 7yr ın females and males respectively, which is also evident from this study. Acknowledgements This study was supported financially by the Fish- eries Directorate of the Greenland Home Rule Government and it was conducted by the Green- land Fisheries Research Institute. The beluga hunters in West Greenland are thanked for pro- viding the samples of their harvest. We thank Royal Greenland and Greenland Trade Depart- ment for help with collection of samples. J. JEN- SEN, J. TEILMAnN, F THoMsEn, and P. BARNER NeEvE did the age determinations and the data entry. G. N. OGNEToV (SEV-Pinro, Arkhangelsk) kindly made his data on white whale age frequen- cies in the White and Kara seas available. P. MoL- LER LUND and U. SIEBERT are gratefully acknowl- edged for making the map and the German translation of the abstract. Zusammenfassung Alters- und Geschlechtsverteilung von Belugafängen, Delphinapterus leucas, in Westgrönland und Westrußland Alter und Geschlecht wurden von Belugas oder Weißwalen Delphinapterus leucas, bestimmt, die von 1985-86 und 1989-97 von Eskimos in Westgrönland erlegt wurden. Der Probenumfang umfaßte 712 Weibchen und 596 Männchen. Es gab eine klare Trennung der Wale in der Jagdfischerei, die während des Herbstes in Qaanaaq (früher als Avenersuag bekannt) und Upernavik, das nördlich des 74 °N Breitengrades liegt, stattfindet. Vor allem nicht geschlechtsreife Wale beider Geschlech- ter wurden zusammen mit geschlechtsreifen Weibchen gefangen. Zähne dienten der Altersbestim- mung. Das Alter wurde an Jahreszuwachsringen (GLGs) im Dentin ermittelt unter der Annahme, daß zwei Zuwachsringe pro Jahr entstehen. Mittelwert und Median für Alter nahmen bei beiden Geschlechtern aus Upernavik von 1985 bis 1994 langsam zu. Sowohl nicht geschlechtsreife als auch geschlechtsreife Wale wurden in den Überwinterungsgebieten der Disko Bucht und südlich des 70 °N Breitengrades entnommen. Die Überlebensrate wurde nach zwei Methoden bestimmt: nach RoBson und CHAPMAN (1961) und über den natürlichen Logarithmus des negativen Exponenten einer an die Altersfrequenz angepaßten Kurve. Die Abschätzung der Überlebensrate wurde erschwert durch eine große Anzahl von Walen, denen bedingt durch eine Abnutzung der Zahnkrone nur ein Minima- lalter zugeordnet werden konnte (d.h. keine Neonatlinie im Dentin). Die offensichtliche Überle- bensrate von Belugas vor Westgrönlands wurde auf 0,81 und 0,79 für Weibchen bzw. Männchen geschätzt. Korrekturen dieser Abschätzung für eine beobachtete Bestandsabnahme von 4,7% pro Jahr ergaben eine tatsächliche Überlebensrate von 0.85 und 0,82 für Weibchen bzw. Männchen. Die Schätzwerte der tatsächlichen Überlebensrate sind geringer als die, welche für die Belugapopu- lation im Weißßen Meer und der Karasee ermittelt wurden, fiir die Altersdaten aus den 70er und frü- hen 80er Jahren zur Verfügung standen, sowie publizierten Raten für Belugas aus Alaska (1977- 83) mit einer vergleichbaren Alterszusammensetzung. Da der Grad der Bejagung in diesen Gebieten wesentlich niedriger ist, bestätigt die geringere Überlebensrate vor Westgrönland deutlich eine Ab- nahme der Population. Der Wechsel in der Hautfärbung von Grau zu Weiß tritt im mittleren Alter von 8,5 und 9,1 Jahren und bei einer mittleren Länge von 367 cm und 445 cm bei Weibchen bzw. Männchen auf. Age and sex distributions in the catches of belugas References BRODIE, P. F. (1971): A reconsideration of aspects of growth, reproduction, and behavior of the white whale (Delphinapterus leucas), with re- ference to the Cumberland Sound, Baffin Is- land, population. J. Fish. Res. Board Can. 28, 1309-1318. BuRNS, J. J.; Seaman, G. A. (1986): Investigations of belukha whales in western and northern Alaska. II. Biology and ecology. Fairbanks, Alaska: Report by Alaska Department of Fish and Game. DoiDgGe, D. W. (1990): Age and stage based analy- sis of the population dynamics of beluga whales, Delphinapterus leucas, with particular reference to the Northern Quebec popula- tion. Ph. D. thesis McGill University, Mon- treal, Quebec. DEMASTER D. P. (1978): Calculation of the aver- age age of sexual maturity in marine mam- mals. J. Fish. Res. Board Can. 35, 912-915. HEIDE-J&ÖRGENSEN, M.P.; REEVES, R.R. (1996): Evidence of a decline in beluga (Delphina- pterus leucas) abundance off West Greenland. ICES J. Marine Science 53; 61-72. HEIDE-JöRGENSEN, M. P. (1994): Distribution, ex- ploitation and population status of white whales (Delphinaterus leucas) and narwhals (Monodon monoceros) in West Greenland. Meddr Grenland, Biosci. 39, 135-149. HEIDE-JORGENSEN, M.P.:; JENSEN, J.;, LARSEN, A.H.; TEILMANN J.; NEUROHR, B. (1994): Age estimation of white whales (Delphinapterus leucas) from Greenland. Meddr Grönland, Biosci. 39, 187-193. HEIDE-JORGENSEN, M.P.; TEILMANN, J. (1994): Growth, reproduction, age structure and feed- ing habits of white whales (Delphinapterus leucas) in West Greenland waters. Meddr Grönland, Biosci. 39, 195-212. 227 Lowry, L. F; BURNS, J. J.; FROST, K. J. (1989): Re- cent harvests of belukha whales, Delphina- pterus leucas, in western and northern Alaska and their potential impact on provisional management stocks. Rep. Int. Whal. Commn 39, 335-339. OGnerov, G.N. (1981): Studies on the ecology and the taxonomy of the white whale (Delphinapterus leucas Pall., 1776) inhabiting the Soviet Arctic. Rep. Int. Whal. Commn 31, 515-520. PALSB@LL, P. J.;, VADER, A.; BAKKE, 1.; EL-GEWEILY, M.R. (1992): Determination of gender in cetaceans by the polymerase chain reaction. Can. J. Zool. 70, 2166-2170. PERRIN, W. F.; MyRrick, A.C. Jr. (1980): Report of the workshop. In: Age determination of toothed whales and sirenians. Ed. by W.F. PERRIN and A.C. MyRrıck Jr. Rep. Int. Whal. Commn, Spec. Issue 3, 1-50. Roson, D.S.; CHAPMAN,D.G. (1961): Catch curves and mortality rates. Am. Fish. Soc., Trans. 90, 181-189. SERGEANT, D. E. (1973): Biology of white whales (Delphinapterus leucas) in Western Hudson Bay. J. Fish. Res. Board Can. 30, 1065-1090. STEWART, R.E. A. (1994): Size-at-age relation- ships as discriminators of white whale (Delphinapterus leucas) stocks in the eastern Canadian Arctic. Meddr Grönland, Biosci. 39, 217-225. Authors’ addresses: M. P. HEIDE-JORGENSEN, National Marine Mam- mal Laboratory, 7600 Sand Point Way NE, Seat- tle, WA, USA (e-mail: madspeter.heide-joergen- sen@noaa.gov) and C. LocKYEr, Danish Institute for Fisheries Research, Charlottenlund Slot, Charlottenlund, Denmark. Mamm. biol. 66 (2001) 228-237 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Original investigation A new species of Aepeomys Thomas, 1898 (Rodentia: Muridae) from the Andes of Venezuela By J. OCHOA G., MARISOL AGUILERA, V. PACHECO, and P. J. SORIANO Asociaciön Venezolana para la Conservaciön de Areas Naturales Caracas, Venezuela; Departamento de Estudios Ambientales, Universidad Simön Bolivar, Caracas, Venezuela; Department of Mammalogy, American Museum of Natural History, New York, USA; Departamento de Biologia, Universidad de Los Andes, M&rida, Venezuela. Receipt of Ms. 25. 04. 2000 Acceptance of Ms. 20. 09. 2000 Abstract A new species of Neotropical rodent of the genus Aepeomys is described based on 24 specimens collected in the Andean region of Venezuela (Lara and Trujillo States). Among the diagnostic char- acters are: large size; first and fifth digits of pes not extending beyond the commissure of digits 2- 3 and the first interphalangeal of digit four, respectively; posterior margin of zygomatic ramus of the maxilla with a distinctive notch; palate extending to the posterior border of M® or behind this molar; and paraflexus of M' and M° divided by an enamel bridge. In addition, the new species shows the following karyological features: 22 chromosomal pairs (2n =44); 46 autosomal arms (FN = 46); a low proportion of two-armed elements; automosal chromosomes with abundant hete- rochromatin around the pericentromeric areas; and short arms of chromosomes X and Y entirely heterochromatic. According to the most recent systematic revision of the species assigned to Aepe- omys, only two forms could be considered as members of this genus: A. lugens (the type species) and the taxon described herein. Both have geographic distributions restricted to highlands from the northern Andes, where the new species inhabits primary cloud forests and päramos located in the northeastern extreme of the Venezuelan Andean Cordillera. Key words: Aepeomys, Thomasomyine, Taxonomy, Andes, Venezuela Introduction Neotropical sigmodontine rodents of the genus Aepeomys are members of the tho- masomyine group, together with six addi- tional genera whose systematic and phylo- genetic relationship; remain unclear: Delomys Thomas, 1917; Phaenomys Tho- mas, 1917; Rhagomys Thomas, 1917; Rhi- pidomys Tschudi, 1844; Thomasomys Coues, 1884; and Wilfredomys Avila-Pires, 1960 (AGUILERA et al. 1994, 2000; GÖMEZ- 1616-5047/01/66/04-228 $ 15.00/0. LAVERDE et al. 1997; MUSSER and CARLETON 1993; REıG 1986; Voss 1993). After the ori- ginal description of Aepeomys by THOMAS (1898), some authors have considered this taxon as a synonym of Thomasomys (e.2., CABRERA 1961; ELLERMAN 1941; HANDLEY 1976). Nevertheless, the results of the most recent systematic revision of these Andean genera (PACHECoO unpubl. data) and several previous publications (e.g., AGUILERA et al. 1994; GARDNER and PATToON 1976; MUSSER and CARLETON 1993; REIG 1986; SORIANO and OcHoA 1997; SoRIANO et al. 1998) are coincident in considering them as differen- tiated taxa. Four nominal species of Aepeomys have been described (CABRERA 1961; MUSSER and CARLETON 1993), although at the pre- sent time only two of them are recognized as valid taxa (both have geographical distri- butions restricted to highlands in the north- ern Andes): Aepeomys lugens (THOMAS, 1896), recorded in several localities from western Venezuela to Andean Ecuador; and A. fuscatus (ALLEN, 1912), known from the western and central Andes of Colom- bia. However, the highly differentiated cra- nial morphology shown by A. fuscatus with respect to A. lugens (the type species of the genus) and other related forms, has been used among the arguments to consider A. fuscatus as representative of a neglected taxon whose evolutionary lineage could be more related with the oryzomyine tribe, re- presenting perhaps an undescribed genus (PAcHEco and Voss unpubl. data). As part of the results of a field study on the small mammal communities inhabiting high- land ecosystems from the Andean region of Venezuela (Lara and Trujillo States), we caught a series of thomasomyine specimens whose general morphology corresponds to Aepeomys (sensu stricto), although their ex- ternal, cranial, and karyological features are not referable to previously known species assigned to this genus. Apparently, they re- present a new species that we describe be- low. Some of these specimens, in addition to others collected in the Venezuelan Andes and cited herein as representatives of the new taxon, were formerly recorded as Aepe- omys lugens or Aepeomys sp. by HANDLEY (1976), SORIANO et al. (1990), and AGUILERA et al. (1994, 2000). Material and methods Specimens examined (all adults) are deposited in the following institutions: American Museum of Natural History (AMNH); the Colecciön de la Es- New species of Aepeomys from Venezuela 229 taciön Biolögica de Rancho Grande (EBRG), Maracay, Venezuela; the Colecciön de Vertebrados de la Universidad de Los Andes (CVULA), Meri- da, Venezuela; and the Colecciön de Vertebrados de la Universidad Simön Bolivar (CVUSB), Cara- cas, Venezuela. Species, individuals and localities corresponding to this material are as follows: Aepe- omys fuscatus (1; holotype). Colombia: Valle del Cauca, San Antonio, near Cali, 2135 m (AMNH- 32230). Aepeomys lugens (21, including two topo- types). Venezuela-Merida State: Päramo Los Co- nejos, 24km W Me£rida, 2928m (AMNH-96169; holotype of A. ottlevi),; 55km E+2km S Tabay (Middle Refugio), 2600 m (EBRG-15569 and 15570); Iikm N+2km W Me£erida (Santa Rosa), 2020 m (EBRG-15571 and 15572); El Morro, 9 km SSW Merida City, 2160 m (EBRG-22009 and 22010; topotypes). Tachira State: Paramo Los Co- lorados (Parque Nacional Päramos Batallön y La Negra), 12km SSE EI Cobre, 3200 m (EBRG- 2151370,215232 EY.UJEAZ5747., 5751, and 5753). Aepeomys reigi (15). Venezuela-Lara State: EI Blanquito, 17km SE Sanare, Parque Nacional Yacambu, 1600 m (CVULA-2738; EBRG-4208, 21735, 21440, and 22580 to 22582); Road El Blan- quito-Sanare, km 6, Parque Nacional Yacambu, 1700 m (EBRG-10621); El Avileno, near El Blan- quito, 9km SE Sanare, Parque Nacional (Yacam- bü, 1600m (CVULA-2710 and 2718). Trujillo State: Macizo de Guaramacal, 9 km ESE Boconö, Parque Nacional Guaramacal, 3100 m (CVULA- 3350); Guaramacal, 5 km E Boconö, Parque Nacio- nal Guaramacal, 2230 m (CVULA-3139); Pica La Toma, 7 km E Boconö, Parque Nacional Guarama- cal, 2300 m (EBRG-22714); 14 t0 15 km E Trujillo, near Hacienda Misisi, 2225 to 2350 m (EBRG- 15567 and 15568). Thomasomys hylophilus (5). Ve- nezuela: 35 km S + 22 km W San Cristobal (Buena Vista), Tächira State, 2395 m (EBRG-15597 to 15601). Thomasomys laniger (5) Venezuela: 4 km S +6.5 kmE Tabay (La Coromoto), M£rida State, 3170 m (EBRG-15227 to 15230); 53km S+ 7 kmE Tabay (near La Coromoto), M£rida State, 3251 m (EBRG-15231). Thomasomys vestitus. (1) Vene- zuela: El Baho, 3km SE Santo Domingo, Merida State, 3010 m (EBRG-32012). Age criteria follow Voss (1991). Cranial measure- ments were taken according to Voss (1988, 1991). Nomenclature of the occlusal components of mo- lar teeth follows Reıc (1977). Karyological ana- lyses were carned out on 13 specimens of A. /u- gens and nine specimens representing the new taxon (five from Lara State and four from Trujillo State), including the sample described by AcGut- LERA et al. (2000). Bone marrow metaphase chro- mosomes were obtained by a modification of Forp and HAMERToN’S (1956) in vivo colchicine 230 OcHoA et al. technique. C- and G-banding patterns were ob- tained as described by BArRos and PATTon (1985) and CHIARELLI et al. (1972), respectively. Chro- mosome nomenclature followed LEVvAan et al. (1964). Fundamental numbers (FN) are autoso- mal arm numbers. Results Aepeomys reigi new species Holotyoe: A female (dry skin, skull, and karyotype analysis; CVUSB-928) with adult 72° 30 729 pelage, fused sphenoccipital suture, and the third molar erupted (age class IV). Col- lected by MARISOL AGUILERA et al. in Au- gust 1986 at El Blanquito, Parque Nacional Yacambuü, 17 km SE Sanare, Lara State, Ve- nezuela, 1600 m (approx. 9°40' N; 69°37’ W; Biel). Paratypes: Seven specimens (6 as dry skins and skulls; one in alcohol) with karyotype analysis: Lara State, Parque Nacional Ya- cambü, EI Blanquito, 17km SE Sanare, 1600 m; 3 males and 1 female collected by 70° 69° 30 Fig. 1. Distribution of A. reigi (triangles) and A. lugens (circles) in Venezuela. White symbols correspond to the type localities. New species of Aepeomys from Venezuela 231 M. AGuUILERA et al. (CVUSB-927, 1365, 1419, and 1420). Trujillo State, Parque Na- cional Guaramacal (approx. 9°15’N; 70°12' W), Pica La Toma, 7km E Boconö, 2300 m; 1 male and 2 females collected by J. OcHoA et al. (EBRG-22715 to 22717). Etymology: The epithet reigi honors the memory of Dr. OswAaLpo ReıG, who de- voted his life to the study of the systematics and evolution of South American rodents, and made important contributions to the education and encouragement of many La- tin American mammalogists. Distribution: Known only in highlands (1600-3230 m) from the northeastern ex- treme of the Venezuelan Andes (Lara and Trujillo States). Diagnosis: Size large for the genus as indi- cated by external and cranial measurements (Tab. 1), in addition to postcranial skeleton development; first and fifth digits of pes not extending beyond the commissure of di- Table 1. Selected external and cranial measurements (in millimeters) of adult specimens of Aepeomys reigi and Aepeomys lugens (age classes 2-4) from Venezuela. Data are: Mean +SD, (range), and sample size. " Sample includes two topotypes and the holotype of Aepeomys (see specimens examined). Measurement Length of head and body Length of tail Length of hind foot Condylo-incisive length Length of diastema Length of molars Length of incisive foramen Breadth of incisive foramen Breadth of rostrum Breadth of palatal bridge Breadth of zygomatic plate Least interorbital breadth Breadth of braincase Zygomatic breadth Depth of incisors Length of orbital fossa 113.6 + 5.88 (104-125)15 127.1+8.29 (116-142)15 27.9+1.30 (25-30)15 27.8 + 0.84 (26.6-29.3)16 8.6+0.31 (8.0-9.0)18 4.5 +0.11 (4.3-4.8)18 5.6+0.22 (5.2-6.0)16 2.4 +0.18 (2.2-2.7)16 5.0 +0.23 (4.6-5.3)14 3.8+0.16 (3.5-4.0)17 1.8 + 0.13 (1.6-2.1)18 6.1+0.17 (5.9-6.4)17 13.3 + 0.22 (12.9-13.7)17 14.8 + 0.37 (14.2-15.6)15 13+0.11 (1.0-1.4)17 9.1+0.24 (8.6-9.6)17 A. lugens' 110.1+7.54 (100-119)7 121.7+4.15 (114-127)7 27.0 +3.96 (20-30)7 26.6 + 0.58 (25.8-27.6)17 8.2 +0.25 (Wzzez)1s 43 +0.12 (4.0-4.4)18 5.5 +0.20 (5.1-5.9)18 2.3 +0.14 (2.0-2.5)18 4.5 +0.27 (4.0-5.1)16 3.5 + 0.25 (3.0-4.0)17 1.8 + 0.14 (1.5-2.0)18 6.0+0.26 (5.6-6.4)18 13.1+0.35 (12.4-13.8)18 14.1 + 0.38 (13.6-14.9)17 1.2+0.11 (1.0-1.4)18 8.4+0.21 (8.0-8.8)17 232 OcHOA et al. gits 2-3 and the first interphalangeal of dig- it four, respectively; posterior margin of zygomatic ramus of the maxilla with a dis- tinct notch; palate extending to the poster- ior border of M° or behind this molar: inter- parietal length (along an antero-posterior axis) near half of parietal length: and para- fiexus of M' and M° divided by an enamel bridge that crosses from the paracone to the base of the anteroloph. Karyotype with 22 chromosomal pairs (2n = 44), 46 autoso- mal arms (FN =46), a low proportion of two-armed elements, the automosal chro- mosomes with abundant heterochromatin around the pericentromeric areas, and the short arms of the chromosomes X and Y entirely heterochromatic. Description: Length of head and body 104- 125 mm. Tail approximately as long as body (Tab. 1), sparsely covered by short dark- brown hairs and unicolored (dark above and below). Legs, heels, and dorsal surface of pes sparsely covered by brown hairs. Body pelage dense and soft (longer in spe- cimens from the highest altitudes). Dorsal coloration ranging from dark gray-brown to reddish gray-brown, with moderately to intensively hoary appearance. Dorsal fur consisting of shorter hairs (approx. 9- 12 mm) with golden tips and scattered long- er hairs (approx. 12-15 mm) with dark brown tips (in a few cases with whitish tips); both having the basal 75% gray. Ventral pelage shorter (approx. 7/mm) and paler than dorsum, ranging from moderately to intensively hoary (hairs with golden tips and the basal 75% gray). Pinnae 18-21 mm long and furred on both sides: inside part yellowish, contrasting in color with the dor- sal fur. Manus cream-colored and paler than hind feet. Pes narrow and long (adapted for terrestrial life; Tab. 1); first and fifth digits not extending beyond the commissure of digits 2-3 and the first inter- phalangeal of digit four, respectively. Incisors narrow and moderately developed (not robust), with sharp tips. Upper incisors with the anterior surface slightly concave. Maxillary and mandibular toothrows relatively short (Tab.1; Fig.2); first molars antero-posteriorly elongated (length averages approximately 50% of their re- spective toothrows). Upper molars with rounded protocone and hypocone, and the paracone and metacone antero-posteriorly sharp. Paraflexus of M' and M° divided by an enamel bridge that crosses from the paracone to the base of the anteroloph, pro- ducing an internal fosseta. M° with triangu- lar shape in dorsal view. M, with a distinc- tive protolophid in most specimens, which reaches the cingulum. Skull with general appearance resembling a typical Aepeomys (see Fig. 2 and 3 for com- parisons with A. /ugens). Rostrum narrow and elongated (approx. 1/3 of the greatest length of skull), with acute profile and only the external capsule of the nasolacrimal foramen exposed in dorsal view; nasal and premaxillary bones extending beyond the anterior surface of incisors and the gnathic process to form a distinct rostral tube. Na- sals laterally concave and flat in dorsal pro- file, forming a continuous surface with the premaxillae: posterior border extending to the level of the zygomatic plate. Interorbital constriction relatively broad, without con- cealing (in dorsal view) the labial ridge of maxillary and the molars. Braincase moder- ately inflated and slightly concave in dorsal profile; the posterior surface concealing the occipital condyles in dorsal view. Interpari- etal length (along an antero-posterior axis) near half of parietal length. Lambdoidal ridges scarcely developed. Zygomatic arches completely ossified, filamentous and fragile. Zygomatic plate relatively narrow (Tab. 1; Fig. 3), with the posterior edge ex- tending to the first molar (Fig. 2). Posterior margin of zygomatic ramus of the maxilla with a distinct notch. Lumen of the infraor- bital foramen compressed laterally and ex- panded dorso-ventrally. Gnathic process scarcely developed. Masseteric tubercle large. Palatal bridge moderately long (Tab. 1; Fig. 2), extending to the posterior border of M° or behind this molar. Poster- ior margin of palate without medial process in most specimens; therefore, the anterior margin of the mesopterygoid fossa has a shallow shape. Incisive foramina extending posteriorly beyond the masseteric tubercle, New species of Aepeomys from Venezuela 233 Fig. 2. Dorsal and ventral views of crania of Aepeomys lugens (topotype; a, b), Aepeomys reigi (holotype; c, d), and Thomasomys laninger (e, f). Approx. X1.9. without reaching the level of the first molar; margins of the anterior half strongly con- vergent anteriorly. Postglenoid foramen compressed dorso-ventrally and expanded antero-posteriorly. Foramen magnum with the inferior border almost reaching the lev- el of the auditory bulla. Auditory bulla moderately inflated. Mandible with the tip of the condylar process behind the angular process. Karyotype with 22chromosomal pairs (2n =44), 46 autosomal arms (FN = 46), and low proportion of two-armed elements (AGUILERA et al. 1994). Automosal chromo- somes with abundant heterochromatin around the pericentromeric areas. Short arms of chromosomes X and Y entirely heterochromatic (AGUILERA et al. 2000). Comparisons: Among the thomasomine group, the genus most closely related to Ae- peomys is believed to be Thomasomys (AGUILERA et al. 2000; GARDNER and PAT- ton 1976), whose cranial morphology is clearly differentiated from A.reigi and A. lugens in the following features (Fig. 2 and 3): shorter rostrum; zygomatic arches more expanded laterally; narrower interor- bital region; braincase less inflated at the le- vel of lambdoidal ridges; broader zygomatic plate; larger incisive foramina (almost reaching the first molars); and shorter pal- ate (posterior border not extending beyond the third molar). In addition, Venezuelan species of Thomasomys are larger (T! aur- eus, T. hylophilus, and T. vestitus) and/or have much paler brownish fur (7. hylophi- OcHoA et al. 234 Fig. 3. Lateral views of crania of Aepeomys lugens (to- potype; a), Aepeomys reigi (holotype; b), and Thomas- omys laniger (c). Approx. X1.8. lus, T. vestitus and T. laninger) than Aepe- omys lugens and A. reigi. With respect to the species previously in- cluded within Aepeomys, A. reigi resembles the external and cranıal morphology of A. lugens, except for the following differ- ences: sıze larger (Tab. 1); fur on head and body shorter and rougher: manus and pes broader; legs, heels, and dorsal surface of pes sparcely haired (densely haired in A. /u- gens); first and fifth digits of pes shorter; posterior margin of zygomatic ramus of the maxilla with a distinct notch, rather than shallow as in A. lugens (as consequence, in A. reigi the orbital fossa is larger, Tab. 1): incisive foramina broader, with margins showing a more convergent position ante- rıorly; interparietal longer in most speci- mens (antero-posterior midline near half of the parietal length, rather than 30-40% as in A. lugens);, palate extending to the pos- terior border of M° or behind this molar (near or before the posterior border of M° in A. lugens); posterior margin of palate without medial process in most specimens (therefore the anterior margin of the me- sopterygoid fossa is shallow rather than in- cipiently biconcave as in A. /ugens); maxil- lary toothrow relatively longer; M° larger and triangular in dorsal view (rounded in A. lugens); M' and M° with paraflexus di- vided by an enamel bridge (continuous in most specimens of A. lugens); coronoid and condylar processes broader and larger, producing deeper sigmoid and angular notches, respectively. M, with a distinctive protolophid in most specimens, which reaches the cingulum (reduced or absent in A. lugens). Some of these features (particu- larly those related with cranial and dental morphology) show the maximun diver- gence in specimens of A. reigi from Lara State. In addition, A. /ugens has a very dif- ferent karyotype, with fewer chromosomal pairs (2n = 28 vs 2n = 44), more autosomal arms (FN=48 vs FN=46), and a lower concentration of heterochromatin (espe- cially conspicuous in the short arms of two autosomal pairs and the Y chromosome; AGUILERA et al. 1994, 2000). These chromo- somal variations were consistent when we compared A. reigi with specimens of A. lu- gens from two localities in Venezuela: the type locality (El Morro, M£rida State) and Päramo Los Colorados, Tächira State (AGUILERA et al. 2000). Regarding A. fuscatus the external and cra- nial features of this species show a high de- gree of differentiation with A. reigi, reveal- ing a morphological pattern that appears to be taxonomically separated from the tho- masomyine group and perhaps corresponds to a taxon whose evolutionary lineage is more related with the oryzomyine tribe (PA- CHEcoO and Voss unpubl. data). Among the most conspicuous characteristics in A. fus- catus supporting this assessment are: darker fur coloration; shorter and broader rostrum (without the acute profile shown by A. reigi and A. /ugens); anterior portion of zygo- matic arches more expanded and broader: broader zygomatic plate; narrower interor- bital breadth; braincase less inflated; short- er incisive foramina and palate; and broad- er mandibular branches. These features, in addition to the extremely high number of chromosomal pairs (2n = 54) and autosomal arms (FN = 62) reported by GARDNER and Patron (1976) for A. fuscatus are clear evi- dences of a differentiated evolutionary pat- tern with respect to Aepeomys. Discussion The morphological variation between A.reigi and A. lugens, in addition to the high degree of differentiation in the num- ber and structure of chromosomes, support the hypothesis of evolutionary divergences in both species, such as it has been pro- posed for other thomasomyine rodents (GARDNER and PArron 1976; GOÖMEZ-LA- VERDE et al. 1997). Despite the higher di- ploid number in A. reigi with respect to A. lugens (44 vs 28), and based on their si- milarities in fundamental numbers (46 vs 48, respectively), we postulate that karyolo- gical differences found in these species could be reached by chromosomal rearran- gements evolving principally robertsonian changes (AGUILERA et al. 2000). An important aspect within the evolution- ary context of Aepeomys species, is the di- rection of chromosomal transformation in A.reigi and A. lugens. According to GARD- NER and PATTon (1976), thomasomyine kar- yotypes are characterized by a generalized condition of diploid number of 42 or 44, ın addition to a predominantly acrocentric autosomal complement. This generalized condition is present in A. reigi and allows to consider it as a primitive form. This fact, together with the great proportion of two- armed elements shown by the karyotype of A. lugens, are arguments to postulate this last species as a derived form (AGUILERA et al. 2000). Some complementary evidences supporting this hypothesis are the differ- ences in quantity and distribution of the constitutive heterochromatin: low and chro- mosomal restricted in A. lugens vs abun- dant and distributed in chromosomes of A. regi; the last pattern has been associated New species of Aepeomys from Venezuela 235 with a primitive condition in eukariotic chromosomal evolution (ImAr 1991). The geographic distribution of A. reigi seems to be allopatric with respect to A. lu- gens, at least in the northeastern extreme of the Venezuelan Andean Cordillera. How- ever, we do not reject the possibility of sym- patric distribution in highlands (> 1500 m) near to the border of Merida and Trujillo States. Future karyological studies, in a more extensive area, are required to pro- vide a further diagnosis on the biogeo- graphic patterns of these taxa. Other non- volant small mammals recorded at Yacam- bu and Guaramacal are: Caluromys philan- der, Didelphis albiventris, Didelphis marsu- pialis, Gracilinanus dryas, Marmosops fuscatus, Micoureus demerarae, Cryptotis meridensis, Mustela frenata, Sciurus grana- tensis, Heteromys anomalus, Akodon urichi, Ichthyomys hydrobates, Microryzomys mi- nutus, Neacomys tenuipes, Oecomys flavi- cans, Oligoryzomys fulvescens, Oryzomys meridensis, Rhipidomys venustus, Rhipid- omys venezuelae and Thomasomys laninger (SORIANO et al. 1990). The known ecological distribution of A. reigi corresponds to prımary cloud forests (humid montane forest according to HUBER and ALARCÖN 1988) and small patches of pära- mos surrounded by continuous masses of cloud forests; these ecosystems, in addition to seasonal forests and evergreen dry forests, have been previously recorded among the ecological conditions used by A. lugens (HANDLEY 1976; SORIANO et al. 1990). A. rei- gi appears to be a relatively uncommon spe- cies along its ecological range. Even though field data for päramos are insufficient, sam- pling efforts of 3 724 trap-nights in cloud for- ests, accumulated during inventories con- ducted by the authors, allowed to catch 27 individuals of A. reigi that represented 10.6% of total non-volant small mammals trapped in this ecosystem (Oryzomys albigu- larıs and Heteromys anomalus were the dominant species). Collected specimens have been found on the ground in densely forested sites (beside logs, at the base of trees, in rocky places, along trails, or near small streams) or in open areas (close to the 236 OCHOA et al. ecotone between päramos and forests) cov- ered by shrubs and herbaceous vegetation (mainly Espeletia schultzii and grasses). We used as bait a mixture of sardine or bacon with oats, peanut butter, and/or kitchen oil. Some specimens maintained in captivity were fed with insects (Orthoptera), domes- tıc fruits, and seeds. Hairs and insect remains (Coleoptera) were found in the stomach content of one specimen from Yacambuü. Three males collected in August and De- cember showed inguinal testis. Acknowledgements We would like to thank the assistance given dur- ing field work by JHONNY MURILLO, RICHARD GAVIDIA, AMELIA DiAZ DE PAscUAL, TINA OLI- Zusammenfassung VEIRA, ROSA MOSCARELLA, ELADIO MÄRQUEZ, ER- NESTO PANIER, MARIAPIA BEVILACQUA, LOURDES SUAREZ, NORBERTO REBOLLEDO, EISAMAR OCHOA, MARIANGELA OCHOA, ELIZANDRA DELGADO, IN- PAROUES, and the Asociaciön Venezolana para el Estudio de los Mamiferos. JAVIER SÄNCHEZ and ROBERT Voss gave important comments to the first draft of the manuscript and allowed revise the specimens housed in the Colecciön de la Esta- ciön Biolögica de Rancho Grande and the Amer- ican Museum of Natural History, respectively. ANTONIO PEREZ ZAPATA (7) and MARCELA GÖMEZ provided valuable information used in this de- scription. Field researches of the senior author are supported by Wildlife Conservation Society. This study is a contribution of the project “Diver- sidad de Pequenos Mamiferos en Ecosistemas de Tierras Altas del Norte de Venezuela” (CONI- CIT-RP-VII-240056). Eine neue Art von Aepeomys Thomas, 1893 (Rodentia: Muridae) aus den Anden von Venezuela Es wird eine neue Art der neotropischen Nagergattung Aepeomys beschrieben. Grundlage bildet eine Serie von 24 Exemplaren, die in Andenregionen (Lara und Trujillo) von Venezuela gesammelt wurde. Diagnostische Merkmale sind unter anderem; große Art; erste und fünfte Zehe reichen nicht über die Kommissur der zweiten und dritten Zehe hinaus; Hinterrand des zygomatischen Astes des Maxillare mit deutlichem Knoten versehen; Palatinum zieht bis zum Hinterrand des M° oder darüber hinaus; Paraflexus am M' und M? durch eine Schmelzbrücke geteilt. Die neue Art ist auch durch ei- nen Karyotyp von 2n =44, NF=46 gekennzeichnet; die perizentromerische Region der Autosomen weist viel Heterochromatin auf; die kurzen Arme der X- und Y-Chromosomen sind vollständig hete- rochromatisch. Die Ergebnisse einer systematischen Revision zeigen, dal nur zwei Arten der Gat- tung Aepeomys angehören: A. lugens (Typusart) und die hier beschriebene neue Art, A. reigi n. sp. Beide sind in ihrer Verbreitung auf die nördlichen Anden beschränkt, wo die neue Art primäre Ne- belwälder und Paramos im äußersten Nordosten Venezuelas bewohnt. References AGUILERA, M.: PEREZ-ZAPATA, A.; OCHOA G. J.; SOR- IANO, P. (2000): Karyology of Aepeomys and two species of Thomasomys (Rodentia: Muri- dae) in Venezuela. J. Mammalogy 81, 52-58. AGUILERA, M.; PEREZ-ZAPATA, A.; MARTINO, A.; BARROS, M. A.: PAaTTon, J. L. (1994): Karyosys- tematic of Aepeomys and Rhipidomys (Ro- dentia: Cricetidae). Acta Cientifica Venezola- na 48, 247-248. ALLEN, J. A. (1912): Mammals from Western Co- lombia. Bull. Am. Mus. Nat. Hist. 31, 71-95. BARROS, M. A.: Patron, J. L. (1985): Genoma evo- lution in pocket gophers (genus Thomomys) III. Fluorochrome-revealed heterochromatin heterogeneity. Chromosoma 92, 337-343. CABRERA, A. (1961): Catälogo de los mamiferos de Ame£rica del Sur. Revista del Museo Ar- gentino de Ciencias Naturales “Bernardino Rivadavia”, Ciencias Zoolögicas 4, 380-513. CHIARELLI, B. A.; SARTI-CHIARELLI, M.; SHAFER, D. A. (1972): Chromosome banding with tryp- sin. Mammal Chrom. Newsl. 13, 64-65. ELLERMAN, J. R. (1941): Family Muridae. In: The families and genera of living rodents. London: British Museum of Natural History. Vol. 2, 1- 69. ForD, E. H.; HAMERTOoN, J. L. (1956): A colchicine, hypotonic citrate, squash sequence for mam- malian chromosomes. Stain Technology 38, 271-274. GARDNER, A.L.; PATTon, J. L. (1976): Karyotypic variation in Oryzomyine rodents (Cricetidae) with comments on chromosomal evolution in the neotropical cricetine complex. Occa. Pa- pers Mus. Zool. Louisiana State Univ. 49, 1-48. GÖMEZ-LAVERDE, M.; MONTENEGRO, D. O.; LoPEZ, A.H.; CADENA, A.; BUENo, M.L. (1997): Kar- yology, morphology, and ecology of Thomas- omys laniger and T. niveipes (Rodentia) in Colombia. J. Mammalogy 78, 1282-1289. HANDLEY, C. O., JR. (1976): Mammals of the Smithsonian Venezuelan Project. Brigham Young Univ. Sci. Bull., Biol. Ser. 20, 1-89. HUBER, O.; ALARCON, C. (1988): Mapa de vegeta- cißon de Venezuela. Caracas, Venezuela: Ministerio del Ambiente y de los Recursos Naturales Renovables. ImAı, H.T. (1991): Mutability of constitutive het- erochomatine (C-bands) during eukaryotic chromosomal evolution and their cytological means. Japonica J. Genetics 66, 365-661. LEVAN, A.; FREDGA, K.; SANDBERG, A. A. (1964): Nomenclature for centromeric position on chromosomes. Hereditas 52, 201-220. MUSSER, G. G.; CARLETON, M.D. (1993): Family Muridae. In: Mammal Species of the World: a Taxonomic and Geographic Reference. Ed. by D. E. Wırson and D. M. REEDER. Washing- ton, London: Smithsonian Institution Press. Pp. 501-755. Reıc, O. A. (1977): A proposed unified nomencla- ture for the enamelled components of the mo- lar teeth of the Cricetidae (Rodentia). J. Zool. (London) 181, 227-241. Reis, O. A. (1986): Diversity patterns and differ- entiation of high Andean rodents. In: High Altitude Tropical Biogeography. Ed. by F. VUILLEUMIER and M. MoNASTERIOS. Oxford: Univ. Press. Pp. 404-439. SORIANO, P. J.; OcHoA, G. J. (1997): Lista actualiza- da de los mamiferos de Venezuela. In: Verte- brados Actuales y Fösiles de Venezuela. Ed. by E. La Marca. ME£rida, Venezuela: Serie New species of Aepeomys from Venezuela 237 Catälogos Zoolögicos de Venezuela, Funda- cite M£rida. Pp. 203-227. SORIANO, P. J.; UTRERA, A.; SosA, M. (1990): Inven- tario preliminar de los mamiferos del Parque Nacional General Cruz Carillo (Guaramacal), Estado Trujillo, Venezuela. Biollania 7, 83-99. SORIANO, P. J.; DiAZ DE PASCUAL, A.; OcHoA G. ].; AGUILERA, M. (1998): Las comunidades de roedores de los Andes venezolanos. In: La Di- versidad Biolögica de Iberoam£rica III. Acta Zool. Mexicana, Nueva Serie. Xalapa, Mexi- co: Instituto de Ecologia, A. C. Pp. 211-226. THoMmAas, O. (1896): On new small mammals from the neotropical region. Ann. Mag. Nat. Hist, series 6, 18, 301-314. THoMmAas, O. (1898): On seven new small mammals from Ecuador and Venezuela. Ann. Mag. Nat. Hist., series 7, 1, 451-457. Voss, R.S. (1988): Systematics and ecology of Ichthyomyine rodents (Muroidea): patterns of morphological evolution in a small adapta- tive radiation. Bull. Amer. Mus. Nat. Hist. 188, 1-493. Voss, R.S. (1991): An introduction to the neo- tropical muroid rodent genus Zygodontomys. Bull. Amer. Mus. Nat. Hist. 210, 1-113. Voss, R.S. (1993): A revision of the Brazilian muroid rodents genus, Delomys with remarks on “Thomasomyine” characters. Amer. Mus. Novitates 3073, 1-44. Authors’ addresses: JosE OcHoA G., Asociaciön Venezolana para la Conservaciön de Areas Naturales (ACOANA) and Wildlife Conservation Souety Apartado 69520, Caracas 1063-A, Venezuela, (e-mail: jochoa@reacciun.ve); MARISOL ÄGUILERA, De- partamento de Estudios Ambientales, Universi- dad Simön Bolivar, Apartado 89000, Caracas 1080-A, Venezuela; VıcTor PACHEcCo, Depart- ment of Mammalogy, American Museum of Nat- ural History, Central Park West at 79th Street, New York, New York 10024, USA; PAscUAL ]. SORIANO, Departamento de Biologia, Universi- dad de Los Andes, M£rida, La Hechicera, Me£ri- da 5101, Estado ME£rida, Venezuela. Mamm. biol. 66 (2001) 238-250 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Developmental stability and protein heterozygosity in a local population of Iberian hares (Lepus granatensis) By P. C. ALves, N. FERRAND, and F. SUCHENTRUNK Centro de Estudos de Ci@ncia Animal, ICETA-UP, Campus Agrärio de Vairäo, Vila do Conde, Departamento de Zoologia-Antropologia, Faculdade de CiEncias, Universidade do Porto, Porto, Portugal, and Research Institute of Wildlife Ecology, Veterinary Medicine University of Vienna, Vienna, Austria Receipt of Ms. 11. 04. 2000 Acceptance of Ms. 13. 09. 2000 Abstract Various studies have revealed a positive effect of heterozygosity on developmental stability in ani- mals of diverse taxa. In homeothermic vertebrates, however, no clear picture has so far emerged in this context. Here, we test the influence of heterozygosity on the developmental stability of adult- sized skulls of 63 Iberian hares (Lepus granatensis) from a local population in Portugal. 44 allo- zyme and blood protein loci were screened by horizontal starch gel electrophoresis, agarose elec- trophoresis, and isoelectric focusing. This yielded eleven polymorphic loci, that were used to calcu- late individual heterozygosity. Levels of fluctuating asymmetry (FA) of three morphological character systems (15 epigenetic dental characters, ten non-metric skull traits, six metric skull vari- ables) were determined and used as indicators of the levels of developmental homeostasis of single hares. Overall individual heterozygosity did not correlate with respective FA levels in any of the three morphological character systems. However, a trend towards a negative relationship between metric FA and heterozygosity suggested that there might be a slight positive influence of hetero- zygosity on developmental stability of the morphometric system, but it could be masked by various seasonal exogenic factors. Key words: Lepus granatensis, heterozygosity, fluctuating asymmetry, developmental homeostasis Introduction Developmental homeostasis, i. e., the ability to buffer against minor random deviations from metabolic pathways during growth, can be reduced by environmental and ge- netic stress (e.g., MOLLER and SWADDLE 1971). The level of fluctuating asymmetry (FA) is commonly used to evaluate the de- gree of developmental stability (e.g., ZA- KHAROV 1981; PALMER and STROBECK 1986; Novak et al. 1993). FA is defined as random 1616-5047/01/66/04-238 $ 15.00/0. departures from the ideal bilateral symme- try of morphological traits with a popula- tion mean around zero and a normal distri- bution (e.g., PALMER 1994). More symmetric animals apparently have higher fitness, as indicated by diverse fitness components, than asymmetric (MoLLer 1997). It has been proposed that heterozygosity stabi- lizes the ontogenetic development, so that genetically determined pathways are more Developmental stability and protein heterozygosity in Lepus granatensis precisely expressed in the phenotype of an organism (e. 8g., HANDFORD 1980; FLEISCHER et al. 1983; Mıtton and GRANT 1984; MIT- Ton 1993 a, b, 1995). In local populations of adult brown hares (Lepus europaeus) from central Europe FA was negatively correlated with population- specific allozyme heterozygosity in non-me- tric but not in metric skull characters. This suggested a differential effect of heterozyg- osity on the developmental stability of non-metrice and metric skull characters (HaRrTL et al. 1995). However, no such rela- tionship was found in brown hares from Britain and New Zealand, that had lower levels of genetic variability than central European brown hares (SUCHENTRUNK et al. 2000), and FA in both character systems was even lower than in brown hares from central Europe (SUCHENTRUNK et al. 1998). Such inconsistencies of the relationship be- tween heterozygosity and FA might result from different evolutionary histories of po- pulation sets. In addition, varying levels of environmental stress may conceal effects of heterozygosity on FA to different degrees (e.8., PANKAKOSKI 1985; PANKAKOSKI et al. 1992; PALMER and STROBECK 1986; BoRISOV et al. 1997; MOLLER and SWADDLE 1997; ZA- KHAROV et al. 1997 a). Here we, examine whether or not hetero- zygosity has a significant effect on develop- mental homeostasis of Iberian hares (Lepus granatensis) from one local population in a relatively homogeneous environment. In particular we examine whether overall indi- vidual heterozygosity is negatively corre- lated with FA in three character systems of the skull; and if so, whether such a relation- ship becomes generally apparent in all three character systems. Material and methods Collection of specimens and samples for genetic analysis Sixty three adult-sized Iberian hares (Lepus grana- tensis) were collected at Pancas (38°48 N/ 8°57 W), approx. 15 km east of Lisbon, Portugal, during regular hunts between October 1997 and 239 October 1999. Pancas is situated in a lowland re- gion (29m a.s.l.) that has a Mediterranean cli- mate, with 500-600 mm average annual rainfall, mainly in October-April, and 16°C mean annual temperature (C.N.A. 1983). The vegetation is characterized by tree stands of OQuercus suber, Pi- nus pinea, as well as marshy areas, pastures, and arable land. All hares were sexed by inspection of their pri- mary reproductive organs. They were classified as “adults”, based on body size and weight, ab- sence of the epiphyseal protrusion of the ulna (cf. SUCHENTRUNK et al. 1991 for European brown hares), the ossification pattern of skull sutures, and the shape and size of the processus supraorbi- tales (e.g. PaALAcıos and LoPEz 1980). Blood samples were taken shortly after the death of the animals by cardiac puncture and collected in ETDA-coated tubes. Red blood cells were se- parated from plasma by centrifugation at 1500 g for 5 min at 4°C and stored at -20 °C. Liver, kid- ney, and spleen tissue samples were taken and frozen at -20°C. Protein heterozygosity Fourty-four locı encoding for allozymes and blood proteins were initially screened for genetic varıability. This set largely included locı that were already studied by Hart et al. (1989, 1990, 1992, 1993, 1994, 1995) for brown hares (L. europaeus), SUCHENTRUNK (1993, 2000), SUCHENTRUNK et al. (1998, 1999, 2000 a, b, c, SUCHENTRUNK, unpubl. data) for brown hares, mountain hares (L. timi- dus), Iberian hares, and several hare species from Mexico, and by Aıves et al. (2000) for Iberian hares. Direct side-by-side comparisons of migrat- ing allozymes on the same gels were made to infer alleles at polymorphic locı from zymograms (cf. HARRIS and Horkınson 1976). Eleven loci re- vealed allelic polymorphism and were considered in the present analyses. They are listed in table 1 along with the respecitve allozyme/blood protein names, E. C. numbers, and methodological speci- fications. Allele frequencies, locus-specific heterozygosities (h,, he), and exact Fisher’s tests for deviation of observed genotypes from Hardy-Weinberg expec- tations were calculated by using the BIOSYS-1 pc package, release 1.7 (SWOFFORD and SELANDER 1989). For the combined analysis with the mor- phological data, genotypes at all polymorphic loci were categorized as “"homozygous” or “heterozy- gous”, irrespective of allele composition. The overall heterozygosity (H) of an individual was calculated as the percentage of heterozygous loci. 240 P. C. Auves et al. Table 1. Enzyme systems/blood proteins and respective loci used in the combined analyses with fluctuating asymmetry of Iberian hares from Pancas. For allele frequencies, see Tab. 4. h. - unbiased locus-specific hetero- zygosity, h, - direct count (observed) locus-specific heterozygosity. Methodological specifications of protein screening (method. specific.) are in the footnote. enzyme systems/blood proteins (name, code, E.C. number) NADH-diaphorase (DIA, 1.6.2.2) Esterases (ES, 3.1.1.1) Acid phosphatase (ACP, 3.1.3.2) Peptidase B (PEPB, 3.4.11) Aminoacylase-1 (ACY-1,3.5.1.14) Mannose phosphate isomerase (MPI, 5.3.1.8) Hemoglobin alpha chain (HBA) Transferrin (TF) Hemopexin (HPX) Vitamin D binding protein (GC) Properdin factor B (BF) method. specific. h. SGE/A - horizontal starch gel electrophoresis and protein staining according to Aıvgs et al. (2000), SGE/G - horizontal starch gel electrophoresis and protein staining according to GRILLITSCH et al. (1992), AGE - agarose gel electrophoresis (TEISBERG 1970), IEF and HIEF - isoelectric focusing in car- rier ampholytes and hybrid pH-gradients according to (Auves et al. 2000). In a few individuals ambiguous allele interpreta- tions at one or more loci resulted in a slightly re- duced number of polymorphic loci for calculation of H. Hence, H-values were used in further ana- lyses only if based on at least nine polymorphic loci. Sex-dependence of locus-specific heterozyg- osity (h) and H was tested by Mann-Whitney U- tests, respectively. Associations of homozygous or heterozygous genotypes among pairs of loci were checked by exact Fisher’s or chi? tests, based on the Sequential Bonferroni procedure (al- pha = 0.05) to account for multiple and partly de- pendent tests (RıcE 1989). The Sequential Bonfer- roni procedure was also applied in all further test series involving non-metric characters, metric variables, as well as in test series of combined morphological and genetic data sets. FA of epigenetic (non-metric) occlusal characters Epigenetic occlusal characters in Leporids con- cern basically presence or absence of enamel folds, notches, grooves or islands, conformation patterns of enamel margins, and presence or ab- sence of cement in folds (e.g., FORSYTH MAJOR 1898; HıBBARD 1963; ANGERMANN 1966; PALACIOS and LorEz 1980). Initially, 40 dichotomized (0/1) occlusal characters were scored for right/left dif- ferences in their respective character states (cf. SUCHENTRUNK 1993; SUCHENTRUNK et al. 1994, 1996, 2000, b). Bilateral asymmetry of a charac- ter was given, if different (0/1, 1/0) character states occurred on the right and left body sides. Only 15 characters were found with clear right/ left-differences. They were used for the FA analy- sis; table 2 details the descriptions of the charac- ters, character states, and character-specific bilat- eral asymmetry levels. The latter were calculated as percentages of individuals with asymmetric characters. A Wilcoxon matched-pairs signed-rank test was run for each character to check for occurrence of FA or directional asymmetry (DA) (PALMER and STROBECK 1986). Since no character showed DA, all were considered indicators of develop- mental homeostasis (PALMER and STROBECK 1986; PALMER 1994; MoLLER and SWADDLE 1997). Associations of FA between pairs of char- acters and sex-dependence of single characters were tested by exact Fisher’s tests, respectively. Individual overall FA of occlusal characters (FAoc) was calculated as the proportion of Developmental stability and protein heterozygosity in Lepus granatensis 241 Table 2. Fluctuating asymmetry (FA) of epigenetic occlusal characters of Iberian hares from Pancas. Current character numbers (CN) and tooth allocation, character description, dichotomized character states (0/1), and level of FA in percent of unequal, i.e., (1)/(0) or (0)/(1) character states are given for each character. Tooth Description of characters dichotomized character states FA Pz Mesial re-entrant fold (filled with cement): present(1)/absent(0) 1.6 Pz Additional mesial re-entrant fold (with cement): present(1)/absent(0) 38 P; Anterior lingual re-entrant fold (with cement): present(1)/absent(0) Pz Posterior external re-entrant fold breaking through the lingual enamel wall and separating trigonid and talonid completely: yes(1)/no(0) Margin of posterior external re-entrant fold forming one extra fold in its most lingual section, extending mesiad and/or distad: yes(1)/no(0) Mesial margin of posterior external re-entrant fold plicate (strong or slight plication) yes(1)/no(0) Distal margin of posterior external re-entrant fold plicate (strong or slight plication): yes(1)/no(0) Distal margin of posterior external re-entrant fold forming one distinct step or extra fold in its lateral part: yes(1)/no(0) Margin of anterior external angle with rather strong plication: yes(1)/no(0) Cement layer of mesial re-entrant fold stretching lingually (covering also anterior linqual re-entrant fold, if present): yes(1)/no(0) Distal margin of lateral fold with extra fold in the buccal section: yes(1)/no(0) Enamel island filled with cement lingually or present (1)/absent(0) buccally of lingual fold (hypostria): central fold with plicated margin: yes(1)/no(0) lingual fold with plicated margin: yes(1)/no(0) labial groove with cement: yes(1)/no(0) Table 3. Non-metric bilateral skull characters used for assessing fluctuating asymmetry (FA). Code, morphologi- cal designation, description of dichotomized (0/1) character states, and levels of FA (%) are given. f. - foramen (foramina). code description and character states (0/1) NM1 Foramen nervi hyperglossi internale: (0) two f. present, (1) > two f. present NM2 Foramen nervi hypoglossi internale accessorium: (0) f. absent, (1) one or more f. present NM3 Foramen condylare: number of f. on both sides: (0) equal, (1) unequal NM4 Foramen alisphenoidale: number of f. on both sides: (0) equal, (1) unequal NM5 Foramen ethmoidale accessorium: (0) absent, (1) present NM6 Foramen palatinus: (0) one f. present, (1) two f. present NM7 _distinct foramen on os maxillare medial of P°-M': (0) no, (1) yes NM3 Foramen frontale mediale: (0) f. absent, (1) f. present NM9 Foramen mandibulare: (0) one f. present, (1) more than one f. present NM10 Foramina along the rostral sulcus of the mandibular ramus (0) equal, (1) unequal number 242 P. C. Auves et al. asymmetric characters of the total set of occlusal characters studied per individual (LEARY et al. 1985). Sex-specific variation of FAoc Was tested by a Mann-Whitney U-test. FA of non-metric skull characters Ten non-metric skull characters (foramıina) were scored on both body sides. They could be easily scored and are largely a subsample of those char- acters that were used in FA analyses in brown hares (Lepus europaeus, see HArTL et al. 1995; SUCHENTRUNK et al. 1998). Character descriptions, dichotomized character states (0/1), and respec- tive asymmetry values appear in table 3. Statisti- cal procedures were analogous to those for dental characters. Calculation of the individual-specific index of overall FA of non-metric skull characters (FAnm) was based on all ten characters. N FA of metric skull variables Six bilateral skull and mandible measurements (Fig. 1) were taken with digital calipers to the nearest 0.01 mm. Measurements were taken ex- clusively by one of the authors (PCA) to elimi- nate the possible inter-observer variability (LEE 1990), and were repeated once to obtain a data basis for evaluating the influence of measurement error on the FA estimation. The effect of mea- surement error on FA values of single variables was calculated by a two-way ANOVA for each variable, based on sides and repeated measure- ments in each individual. The influence of mea- surement error on the asymmetry measurement was considered insignificant, if the sum of mean variance of the side faCtor and the mean variance of the side/individual interaction factor was at least twice as high as the residual mean variance (HARTL et al. 1995; see also PALMER 1994). Fig. 1. Skull measurements used for de- termination of fluctuating asymmetry. CRL = cranium length, LML = lower molar row length, MDL = mandibular diastema length, NL = nasalia length, UML = upper molar row length, UTL = upper tooth row length. Developmental stability and protein heterozygosity in Lepus granatensis Occurrence of DA or antisymmetry (AS) was tested for each variable by a sign test (right minus left measurements), and by a Kolmogorov-Smir- nov test of the frequency distribution of the right-left paired differences, respectively (PALMER and STROBEcK 1986; HARTL et al. 1995). To check for size-dependence of asymmetry, a Spearman correlation between individual values of |right- left | differences and respective arithmetic means was performed in each variable. Despite absence of a correlation, we used an FA index for single metric variables that allowed the calculation of overall individual FA indices, even in cases of missing data for single values in partly damaged skulls. For single variables the following FA index was used: IR-L/[R +L)2], where R and L are the measurements on the right and left sides, respectively (cf. PALMER and STRO- BECK 1986; HARTL et al. 1995; PALMER 1994). Sex- dependence of FA of each variable was checked by a one-way ANOVA. Pairwise correlations of FA of variables were tested by Spearman correla- tions. Overall FA of metric variables (FAy,) was calculated as the arithmetic mean of FA indices of the six (CRL, LML, MDL, NL, UML, UTL), and only in few partly damaged skulls based on five variables. Sex-dependence of FA, was checked by a one-way ANOVA and size-depend- ence by a Pearson correlation between FAy and individual condylobasal length (CBL) (e.g., PAL- MER 1994). Relationships between non-metric and metric FA Relationships between non-metric dental and skull characters were examined by Fisher’s exact tests, and those between non-metric and metric variables by Mann-Whitney U-tests. Relation- ships between FAoc, FAnm, and Fım were exam- ined by Spearman correlations, respectively. FA, skull length, and protein heterozygosity Relationships between either FAoc, FAnm; Fam; CBL, and H were tested by Spearman correla- tions, respectively. For the relationship between FAu and heterozygosity, the following additional statistical approach was carried out: H-values were classified as low (<20%) and high (20-45. 5%); and within each group FA values of single variables were calculated by: var ((R-L)/(R + L)/2], 243 where R and L are the right and left side mea- surements per individual and var is the H-group variance. To maximize the information on FA of all variables and individuals in a comparison of metric FA between the two heterozygosity groups, we performed a two-way ANOVA with variable and group as factors (cf., PALMER 1994). Results The allele frequencies at polymorphic loci are presented in table 4. One significant de- viation of genotype frequencies from Hardy-Weinberg expectations at the Dia-2 locus was found; it was due to a slight het- erozygote deficiencey. Locus-specific ex- pected heterozygosities ranged between 0.016-0.501 and direct count heterozygos- ities between 0.016-0.548 (Tab. 1). There were no sex-specific differences of frequen- cies of homozygote and heterozygote geno- types at any locus. Also, no significant pair- wise associations of homozygous or heterozygous genotypes were found among locı. H values ranged from 0.0% to 45.5% with a mean of 22.52% and a standard de- viation of 12.68%. H values did not vary significantly between the sexes. Levels of asymmetry of non-metric charac- ters appear in tables 2 and 3 and those of metric variables in table 5. Only one metric variable (DIA) showed DA. No character showed AS. Apart from DIA, all metric characters were used as indicators of devel- opmental homeostasis (PALMER and STRO- BECK 1986). In single non-metric characters and metric variables no sex-specific differ- ences of FA were found. Also, no significant associations or correlations of symmetric or asymmetric expressions were detected be- tween pairs of non-metric characters or me- tric variables, respectively. The two-way ANOVA of measurement repeats and side (Tab. 5) did not suggest that FA values of the six metric variables used for FA calcula- tions were confounded by measurement er- TOTS. FA of single metric variables and FA, val- ues were not significantly correlated with CBL. Also, in non-metric characters no sig- nificant differences of CBL values were 244 P. C. Auves etaal. Table 4. Allele frequencies (%) of polymorphic loci in hares from Pancas. Allele designations are not necessarily in alphabetical or numerical order because alleles were assigned in a combined analysis of L. granatensis and L. europaeus and some alleles were not found in the hares from Pancas. Allele designations of the loci Gc, Pep-B, HBA, Bf, and Hpx conform to those in Auves et al. (2000) locus allele frequency locus allele locus allele frequency frequency Tv nor n0 >» Table 5. FA of single metric skull variables selected for estimating overall metric skull FA of hares from Pancas. n= sample size, m.e. = measurement error (see material and methods), (R+L)/2 - mean of the variable size, Ss. e. = standard error of mean. Mean and standard errors of the differences between the sides (R-L) are also given to indicate the absence of DA (cf. PALMER 1994) in conjunction with the not significant sign-test results, based on the Sequential Bonferroni procedure. For variable acronyms and calculation of the FA index, see material and methods. (R+L)/2 (+/-s. e.) (R-L) mean (+/-:.e. FA index mean s.e. 29.70 (0.145) 17.90 (0.08) 20.42 (0.13) 40.11 (0.2) 16.80 (0.07) 43.46 (0.17) 0.046 (0.036 0.044 (0.043 -0.083 (0.045 -0.097 (0.047 -0.06 (0.03) 0.048 (0.035) 0.007 (0.001 0.013 (0.002 0.013 (0.001 0.007 (0.001 0.011 (0.001) 0.005 (< 0.001) found between symmetric and asymmetric Table 6. Relationships between overall individual hete- character states. And there were neither significant correlations between CBL and FAoc Or FAnm values nor significant corre- lations between FAoc, FAnm; FAu values. No significant correlations between FAoc, FAım; FAm, CBL, and H were found. However, there was a slight tendency to- wards lower FA, values in hares with greater H values. The respective correlation coefficients and associated significance le- vels are listed in table 6. The two-way AN- OVA of the variance-based metric FA in- rozygosity (H) and overall fluctuating asymmetry of non-metric occlusal characters (FAoc), non-metric skull characters (FAym), metric skull variables (FAy), condy- lobasal length (CBL). One-tailed Spearman rank correla- tion coefficients (r,), individual numbers (n), and sig- nificance levels (p) are given; n.s. = not significant. 0.0485 -0.0017 -0.2164 0.0824 48 49 49 48 0.289 DES: 0.372 n. S. 0.495 n.S. 0.068 n.S. Developmental stability and protein heterozygosity in Lepus granatensis 245 Table 7. Fluctuating asymmetry (FA) values of single metric skull variables in hares with low (< 20%) and high (> 19%) heterozygosity. Means (M), standard errors (SE), minimum (MIN), maximum (MAX), and sample sizes (n) are given for each variable and heterozygosity group. M, SE, MIN, and MAX values are multiplied by 10°. For acro- nyms of variables, see Fig. 1. FA index |R- L|/[(R + LJ/2] Variable dices revealed a significant effect by vari- ables (p<0.0001) but only a tendency (p = 0.019; Bonferroni criterion for multiple testing: p = 0.01) towards an H group effect. The H group/variable interaction factor was not significant (p = 0.457). The means, stan- dard errors, and extreme values of FA in single metric variables are listed in table 7, separately for each H group. Discussion Heterozygosity is commonly considered to indicate levels of genetic variability within individuals and populations (e.g., MITTON and PıErcE 1980; NEI 1987). Hares with high heterozygosity may harbour less homozygous genotypes with rare recessive alleles that are detrimental to certain meta- bolic processes than hares with low hetero- zygosity. Hence, low heterozygosity could lead to higher developmental instability (“dominance hypothesis”). According to the “overdominance hypothesis” indivi- duals with heterozygote genotypes at many unlinked polymorphic loci should have a better capabiliıty of buffering biochemical processes against various adverse environ- mental effects during ontogenesis (e.g., TURELLI and GiINZzBURG 1983; MITTON 1993b). In Oldfield mice (Peromyscus po- lionotus) the ability of an individual to maintain stable developmental trajectories under fluctuating environmental conditions is related to its genetic variability (TESKA et al. 1991). Higher genetic variability may lead to the production of a higher variabil- ity of biochemical products to buffer di- verse environmental influences. This in turn should lead to a more regular expression of bilateral symmetric morphological traits of an organism (e.g., GINZBURG 1979; MITTON and GRANT 1984; Mırton 1995; M@LLER and SwWADDLE 1997). Bilateral asymmetry of morphological char- acters is only indicative of developmental homeostasis if they show fluctuating asym- metry (FA) (PALMER and STROBECK 1986; PALMER 1994; MSLLER and SWADDLE 1997). Among all traits presently studied, the only case of directional asymmetry (DA) was found in one metric variable (DIA). We do not have a convincing biological explana- 246 P. C. Auves etaal. tion for this significant deviation from FA. It might result from a systematic measure- ment bias due to different positions of the skulls when holding them for taking right and left measurements. Whatever reason, we excluded this variable from all com- bined analyses with heterozygosity. In three other metric variables (MDL, NL, UML) the standard errors of right-left differences (R-L) were quite low compared to the re- spective means (table 5). This might suggest DA in these characters. Nevertheless, we in- cluded these variables in the calculations of metric FA because of nonsignificant sign- tests when based on the Sequential Bonfer- roni procedure. Our results demonstrate that there is FA in all three morphological character systems studied, but no concordance of FA levels among the three morphological systems. And we did not find a significant relation- ship between FA and heterozygosity. This corresponds to the “poikilotherm-home- otherm hypothesis” (HANDFORD 1980; Woo- TEN and SMITH 1986; but see HARTL et al. 1995), according to which a negative rela- tionship between FA and heterozygosity should be more likely in poikilothermic ani- mals because of their supposedly greater sensitivitiy to environmental conditions, whereas homeotherms experience a more stable development during their ontogeny (see also M@LLER and SwApDLE 1997). In some poikilothermic vertebrates, single-lo- cus heterozygosity was found to be nega- tively correlated with FA (Mrrtton 1993a, b, 1995 for overview). However, here we did not check for relationships with single- locus heterozygosity, to avoid too stringent significance levels by the Sequential Bon- ferroni procedure (see PALMER 1994). Nevertheless, in the metric skull characters a tendency towards increased FA in hares with low heterozygosity became apparent. A weak inverse relationship might indeed exist, but could be largely masked by effects of diverse exogenic factors, despite the quite small and homogeneous study area. Muuvey et al. (1994) found a negative rela- tionship between genic variability and FA in the fish Gambusia holbrooki only in cer- tain environmental contexts. Exogenic fac- tors might include diverse weather compo- nents or food stress in different seasons. At Pancas hares are born all year round (lever- ets can be observed in any season). Seaso- nal changes of ground temperature and moisture, wind and rainfall, together with varying food availability for leverets and lactating does could modify levels of FA of hares born in different seasons. Increased dental FA was observed in mice that were born and raised in cold environments (SIE- GEL and DoyLe 1975a). Rats (Rattus norve- gicus) exposed to cold and heat stress in- creased FA of long bones (GestT et al. 1986). FA of humeri of cotton mice (Pero- myscus gossypinus) and Florida mice (P. floridanus) was raised by cold stress (SIEGEL and DoyLe 1975 b). Levels of para- sitic infections may also have varied within the sampling period. FA of antlers of a Nor- wegian reindeer (Rangifer tarandus) herd was enhanced by abomasal nematode infec- tions, and there was a negative relationship between certain immune parameters and FA (LAGESEn and ForstAaD 1998). In addi- tion, seasonal changes of levels of psycho- genic stress due to variable predation pres- sure by foxes, raptors etc. with possible effects on various hormone-based regula- tion systems may influence developmental stability. Social stress had a negative impact on FA of non-metric skull traits in labora- tory rats (Rattus norvegicus) (VALETSKY et al. 1997). PALMER and STROBECK (1986), and PALMER (1994), among others, recommended var- iance-based FA indices and comparisons of FA at the population/group level. In a meta-analysis, VOLLESTAD et al. (cf. MOLLER and SwApDDLE 1997) found a tendency to- wards an inverse relationship between FA and heterozygosity at the population level in poikilotherms but no clear pattern in homeotherms. In fact, in central European brown hares (Lepus europaeus) a signifi- cant negative relationship between non-me- tric FA and heterozygosity was only appar- ent at the population level HarTL et al. 1995). This might result from a better esti- mate of genic variability by groupspecific Developmental stability and protein heterozygosity in Lepus granatensis heterozygosity than by individual hetero- zygosity (PALMER and STROBECK 1986). To increase the power of the comparison of FA between groups of individuals (popula- tions etc.), PALMER (1994) recommended combining the FA information of several traits in a two-way ANOVA with group and trait factors. By this variance-based ap- proach, the initially found (not significant) tendency towards increased FA of metric skull variables in hares with low heterozyg- osity was also evident. However, we empha- size that in this second variance-based test of FA between the two heterozygosity groups, sample sizes per group were natu- rally lower compared to the full set of indi- viduals in the correlation analysis between H and Fım. Moreover, this second ap- proach resulted in a more stringent signifi- cance level to account for multiple (and partly dependent) testing (Rıce 1989). This might have reduced the chance of detecting a significant difference. It has been questioned whether measuring heterozygosity at a comparatively small number of loci provides a good estimate of overall genome heterozygosity. Heterozyg- osity estimates might be valid particularly if: a) a large amount of the genome is struc- tured in blocks, i.e. when there is a large amount of linkeage disequilibrium, b) there is a high degree of inbreeding, and c) there is non-random mating due to small popula- tion size and isolation (M@LLER and SwAD- DLE 1997). The largely absence of concor- dant locus-specific deficiency of hetero- zygous genotypes does not particularly indi- cate a severe level of inbreeding in the hares from Pancas. We also did not find any significant linkage disequilibrium in our sample and we do not have any infor- mation on the mating structure of hares at Pancas. However, there is a substantial lev- el of gene flow between the Pancas popula- tion and other populations in Portugal (Aıves and FERRAND 1999). To increase the predictor quality of heterozygosity, we 247 analysed most of the protein loci that have been found to be polymorphic in the genus Lepus (HARTL et al. 1990, 1992, 1993, 1995 for brown hares; GRILLITSCH et al. 1992; Aıves et al. 2000; SUCHENTRUNK et al. 1998, 1999, 2000, and unpublished data for diverse hare species). Based on eleven polymorphic loci, we did not find a significant positive effect of over- all individual heterozygosity on develop- mental stability. This, however, does not ne- cessarily mean that there is no such effect. We found a tendency for a negative correla- tion between metric skull FA and hetero- zygosity; and such a relationship might in- deed exist, but it could be masked by the combined influence of various unperceived seasonal environmental stressors. Weather parameters or food avaılability, among other stressors, might exert seasonal influ- ences on growing hares. If so, such exogenic stress components would have a clearly higher overall effect on developmental sta- bility than cross genic variability of the hares. This, however, should be tested by a further study based on hares from different birth seasons. Acknowledgements We thank Mr. RAFAEL VINHAIS, the manager of the Pancas hunting estate, the commandant of the Campo de Tiro de Alcochete, and all hunters for their help to obtain the samples. Mrs. AnITA HAIDEN (Vienna), Mrs. HELENA GONCALVES, Mrs. ANA MONTEIRO, Mr. CHRISTIANO LIMA, Mrs. JOANA CASTRO and Mrs. CATARINA FERREIRA (all Vila do Conde) provided help during field and lab work, Mr. RuDoLF WiLLınG (Vienna) provided statisti- cal advise, and Mr. A. KÖrRBER (Vienna) prepared the figure. Financial support for this study was provided by the Instituto de Cooperagäo Cientifi- ca e Tecnolögico International (Ministerio da Ciencia e da Tecnologia (Portugal), the ÖAD (Austrian Academic Exchange Service, Vienna), and by the Direcäo-General de Florestas (Portu- gal). 248 P. C. AuvEs etal. Zusammenfassung Heterozygotie und Entwicklungshomöostase bei Iberischen Hasen (Lepus granatensis) aus einer lokalen Population in Portugal Bisherige Untersuchungen haben vielfach einen positiven Einfluß des Heterozygotiegrades auf die Entwicklungsstabilität von Tieren ergeben. Homeotherme Vertebraten zeigen diesbezüglich aller- dings kein einheitliches Bild. In dieser Arbeit wird der Einfluß des individuellen Heterozygotiegrades von Iberischen Hasen aus einer lokalen Population in Portugal auf ihre Entwicklungshomöostase untersucht. Die Analyse von 44 proteinkodierenden Loci mittels Stärke- und Agarosegelelektropho- rese sowie isoelektrischer Fokusierung ergab bei 63 Hasen elf polymorphe Loci, die zur Berechnung des individuellen Heterozygotiegrades herangezogen wurden. Das Niveau der Entwicklungshomöo- stase einzelner Hasen wurde anhand der fluktuierenden Asymmetrie (FA) in nicht-metrischen und metrischen Merkmalssystemen (Zahnmerkmale, Schädelmeßstrecken und Foramina) ermittelt. Das Ausmaß der individuellen FA war in keinem der drei Merkmalssysteme mit dem Heterozygotiegrad korreliert. Die FA der Schädelmeßstrecken zeigte aber tendenziell einen negativen Zusammenhang mit der Proteinheterozygotie. Solch ein (schwacher) Zusammenhang könnte tatsächlich bestehen, aber durch nicht erfaßte (kaum erfaßbare) exogene Streßßfaktoren weitestgehend verdeckt sein. References AıveEs, P.C.; FERRAND, N. (1999): Genetic varia- bility in Portugese populations of the Iberian hare, Lepus granatensis Folia Zool. 48 (Suppl. 1), 3-10. Auves, P. C.; BRANco, M.; MATIAS, O; FERRAND, N. (2000): New genetic variation in the Eur- opean hares, Lepus granatensis and L. euro- paeus. Biochem. Genet. 38, 87-96. ÄNGERMANN, R. (1966): Beiträge zur Kenntnis der Gattung Lepus (Lagomorpha, Lepori- dae). I. Abgrenzung der Gattung. Mitt. Zool. Mus. Berlin 42, 127-144. Borısov, V.1.; BARANOV, A.S.; VALETZKY, A. V.; ZAKHAROV, V.M. (1997): Developmental sta- bility of the mink Mustela vison under the im- pact of PCB. Acta Theriol., Suppl. 4, 17-26. CoMIssa0o NATIONAL Do AMBIENTE (C.N. A.) (1983): Altas do Ambiente. Direccao General do Ambiente. Lisboa: Ministerio do Am- biente e dos Recursos Naturais. FLEISCHER, R. C.; JOHNSTON,R.F.; KLiTz, W. S. (1983): Allozymic heterozygosity and mor- phological variation in house sparrows. Na- ture 304, 628-630. FoRSYTH MAJOR, C. I. (1898): On fossil and recent Lagomorpha. Trans. Linn. Soc. Lond. 7, 433- 520. GEST, T. R.; SIEGEL, M. I.; AnISTRANSKI, J. (1986): The long bones of neonatal rats stressed by cold, heat, and noise exhibit increased fluctu- ating asymmetry. Growth 50, 385-389. GINZBURG, R. L. (1979): Why are heterozygotes often superior in fitness? Theor. Pop. Biol. 15, 264-267. GRILLITSCH, M.; HARTL, G. B.; SUCHENTRUNK, F.; WirLing, R. (1992): Allozyme evolution and the molecular clock in Lagomorpha. Acta Theriol. 37, 1-13. HANDFORD, P. (1980): Heterozygosity at enzyme loci and morphological variation. Nature 286, 261-262. HARRIS, H.; HoPKInson, D. A. (1976): Handbook of Enzyme Electrophoresis in Human Genet- ics. Amsterdam: North-Holland Publ. Co. HARTL, G. B.; WiLLIinG, R.; NADLINGER, K. (1994): Allozymes in mammalian population genetics and systematics: Indicative function of a mark- er system reconsidered. In: Molecular Ecology and Evolution: Appproaches and Applica- tions. Ed. by DB. SCHIERWATER, B. STREIT, G. P. WAGNER and, R. DESALLE. Basel: Birk- häuser. Pp. 299-310. HARTL, G.B.; SUCHENTRUNK,F., WILLING, R.; GRILLITSCH, M. (1989): Biochemisch-genet- ische Variabilität und Differenzierung beim Feldhasen (Lepus europaeus) in Niederöster- reich. Wien. Tierärztl. Mschr. 76, 279-284. HARTL, G. B.; SUCHENTRUNK,F.; NADLINGER, K.; WirLing, R. (1993): An integrative analysis of genetic differentiation in the brown hare Lepus europaeus based on morphology, allozymes and mitochondrial DNA. Acta Theriol. 38, 33-57. Developmental stability and protein heterozygosity in Lepus granatensis HARTL, G. B.; SUCHENTRUNK,F; WILLING, R.; PETZNEK, R. (1995): Allozyme heterozygosity and fluctuating asymmetry in the brown hare (Lepus europaeus): a test of the developmen- tal homeostasıs. Phil. Trans. Soc. Lond. B, 350, 313-323. HARTL, G. B.; MARKOWSKI, J.; Kovacs, G.; GRIL- LITSCH, M.; WILLInG, R. (1990): Biochemical variation and differentiation in the brown hare (Lepus europaeus) of Central Europe. Z. Säugetierkunde 55, 186-193. HARTL, G. B.; MARKOWSKI, J.; SWIATECKI, A.; JA- NISZEWSKI, T.; WILLING, R. (1992): Genetic di- versity in the Polish brown hare Lepus euro- paeus Pallas, 1778: implications for con- servation and management. Acta Theriol. 37, 15-25. HIBBARD, C. W. (1963): The origin of the P; pat- tern of Sylvilagus, Caprolagus, Oryctolagus and ZLepus. J. Mammalogy 44, 1-15. LAGESEN, K.; FOLSTAD, I. (1998): Antler asymme- try and immunity in reindeer. Behav. Ecol. Sociobiol. 44, 135-142. LEE, J. C. (1990): Sources of extraneous variation in the study of meristic characters: the effect of size and inter-observer variability. Syst. Zool. 39, 47-65. LEARY, R.F.; ALLENDORF, F. W.; KnUDsEn, K.L. (1985): Developmental instability and high meristic counts in interspecific hybrids of sal- monid fishes. Evolution 39, 1318-1326. Mıtton, J.B. (1993a). Enzyme heterozygosity, metabolism, and developmental stability. Ge- netica 89, 47-65. Mıtton, J. B. (1993b): Theory and data pertinent to the relationship between heterozygosity and fitness. In: The Natural History of Inbreeding and Outbreeding. Theoretical and Empirical Perspectives. Ed. by N. W. THORNHILL. Chica- go, London: Univ. Chicago Press. Pp. 1741. Mıtton, J. B. (1995): Enzyme heterozygosity and developmental stability. Acta Theriol., Suppl. 3, 33-54. Mıtton, J. B.; PIERCE, B. A. (1980): The distribu- tion of individual heterozygosity in natural populations. Genetics 95, 1043-1054. Mıtton, J. B.; GRANT, M.C. (1984): Associations among protein heterozygosity, growth rate, and developmental homeostasis. Ann. Rev. Ecol. Syst. 15, 479-499. Möller, A. P. (1997): Developmental stability and fitness: A review. Am. Nat. 149, 916-932. Möller, A. P.; Swaddle, J. P. (1997): Asymmetry, Developmental Stability, and Evolution. Ox- ford Series in Ecology and Evolution. Oxford: Oxford Univ. Press. 249 MULVEY, M.; KELLER, G. P.; MEFFE, G. P. (1994): Sin- gle- and multiple-locus genotypes and life-his- tory responses of Gambusia holbrooki reared at two temperatures. Evolution 48, 1810-1819. Neı, M. (1987). Molecular Evolutionary Genetics. New York: Columbia University Press. NoOVAK, J.M.; RHoEs, O.E.,JR.,;, SMITH, M.H.; CHESSER, R.K. (1993): Morphological asym- metry in mammals: genetics and homeostasis reconsidered. Acta Theriol. 38, Suppl. 2, 7-18. PALAcIOSs, F.; LöPEZ, N. (1980): Morphologia den- taria de las liebres europeas (Lagomorpha, Leporidae). Donana, Acta Vertebr. 7, 61-81. PALMER, R. A. (1994): Fluctuating asymmetry ana- lyses: A primer. In: Developmental Instabil- ity: Its Origins and Evolutionary Implications. Ed. by T. A. MAarkow. Dordrecht, Nether- lands: Kluwer. Pp. 335-364. PALMER, R. A.; STROBECK, C. (1986): Fluctuating asymmetry: Measurement, analysis, patterns. Ann. Rev. Ecol. Syst. 17, 391-421. PANKAKOSKI, E. (1985): Epigenetic asymmetry as an ecological indicator in muskrats. J. Mam- malogy 66, 52-57. PANKAKOSKI,E.; KomwvIsto, I; HyYvaARINEN,H. (1992): Reduced developmental stability as an indicator of heavy metal pollution in the common shrew, Sorex araneus. Acta Zool. Fenn. 191, 137-144. Rıce, W.S. (1989): Analyzing tables of statistical tests. Evolution 43, 223-225. SIEGEL, M.I.; DoyLe, W.J. (1975a): Stress and fluctuating limb asymmetry in various species of rodents. Growth 39, 363-369. SIEGEL, M. 1.; DoyLe, W. J. (1975b): The effects of cold stress on fluctuating asymmetry in the dentition of the mouse. J. exp. Zoology 193, 385-391. SUCHENTRUNK, F. (1993): Variability of minor tooth traits and allozymic diversity in brown hare Lepus eurolapeus populations. Acta Theriol. 38, Suppl. 2, 59-69. SUCHENTRUNK, F. (2000): Epigenetic dental asym- metry of Israeli hares (Zepus sp.): develop- mental stability along an environmental gra- dient. Isr. J. Zoology 46, 103-118. SUCHENTRUNK, F.; FLux, J.E.C. (1996): Minor tooth traits in East African Cape hares and Savanna hares (Lepus capensis and L. victor- iae): A study of intra- and interspecific varia- bility. J. Zool. (London) 238, 495-511. SUCHENTRUNK, F; HARTL,G.B.; WiLLing, R. (1991): On eye lens weight and other age criteria of the Brown hare (Lepus euro- paeus Pallas 1778). Z. Säugetierkunde 56, 365-374. 250 P. C. Auves et al. SUCHENTRUNK,F; WILLING,R.; HARrTL, G. B. (1994): Non-metric polymorphism of the first lower premolar (P3) in Austrian brown hares (Lepus europaeus): a study of regional differ- entiation. J. Zool. (London) 232, 79-91. SUCHENTRUNK, F.; JASCHKE, C.; HAIDEN, A. (2000 a): Little allozyme and mtDNA varia- bility in brown hares (Lepus europaeus) from New Zealand and Britain - A legacy of bot- tlenecks? Mamm. biol. 66, 48-59. SUCHENTRUNK, F.; ALKON, P. U.; WILLING, R.; YoM- Tov, Y. (2000b): Epigenetic dental varıability of Israeli hares (Lepus sp.): ecogenetic or phy- logenetic causation? J. Zool. (London) 232, 503-515. SUCHENTRUNK, F.; MICHAILOV, C.; MARKOV, G.; HAIDEN, A. (2000c): Population genetics of Bulgarian brown hares Lepus europaeus: allo- zymic diversity at zoogeographical crossroads. Acta Theriol. 45, 1-12. SUCHENTRUNKSF.; SIHARTENG. Bi Erux, IE. .C:; PARKES, J.; HAIDEN, A.; TAPPER, S. (1998): Al- lozyme heterozygosity and fluctuating asym- metry in brown hares Lepus europaeus intro- duced to New Zealand: Developmental homeostasis in populations with a bottleneck history. Acta Teriol., Suppl. 5, 35-52. SUCHENTRUNK, F.; POLSTER, K.; GIACOMETTI, M.; RATTI, P.; THULIN, C.-G.; RUHLE, C.; VASILEV, A.G.; SLOTTA-BACHMAYR, L. (1999): Spatial partitioning of allozyme variability in Euro- pean mountain hares (Lepus timidus): gene pool divergence across a disjunct distribu- tional range? Z. Säugetierkunde 64, 1-11. SWOFFORD, D. L.; SELANDER, R. B. (1989): BIOSYS-1. A computer program for the ana- lysis of allelic variation in population genetics and biochemical systematics. Release 1.7. Users manual. Champaign: Illinois Natural History Survey. TEISBERG, P. (1970): High voltage agarose gel elec- trophoresis in the study of C3 polymorphism. Vox sanguinis 19, 47-56. TEsKA, W. R.; SMITH, M. H.; Novak, J.M. (1991): Food quality, heterozygosity, and fitness cor- relates in Peromyscus polionotus. Evolution 44, 1318-1325. TURELLI, M.; GINZBURG, L. (1983): Should indivi- dual fitness increase with heterozygosity? Ge- netics 104, 191-209. VALETZKY, A. V.; DMITRIEVA, I. L.; KRUSHINSKAYA, N. L.; ZAKHAROV, V.M. (1997): Social stress impact on developmental stability of labora- tory rat Rattus norvegicus. Acta Theriol, Suppl. 5, 27-32. WOooTEn, M. C.; SMITH, M.H. (1986): Fluctuating asymmetry and genetic variability in a natural population of Mus musculus. J. Mammalogy 67, 725-732. ZAKHAR0oV, V.M. (1981): Fluctuating asymmetry as an index of developmental homeostasis. Genetika 13, 241-256. ZAKHAROV, V. M.; VALETZKY, A. V.; YABLOKOV, A.V. (1997 a): Dynamics of developmental stability of seals and pollution in the Baltic Sea. Acta Theriol., Suppl. 4, 9-16. Authors’ addresses: PAULO CELIO ALvEs and NUNO FERRAND, Centro de Estudos de Ciencia Animal (CECA), ICETA- UP, Campus Agrärio de Vairäo, 4485-661 Vairäo, Vila do Conde and Departamento de Zoologia- Antropologia, Faculdade de Ci£ncias, Universi- dade do Porto, 4050 Porto, Portugal, FRANZ SUCHENTRUNK, Forschungsinstitut für Wildtier- kunde und Ökologie der Veterinärmedizinischen Universität Wien, Savoyenstrasse 1, A-1160 Wien, Austria, e-mail: franz.suchentrunk@vu-wien.ac.at Mamm. biol. 66 (2001) 251-255 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication . Mammalian Biology Zeitschrift für Säugetierkunde Notes on the ecology of sympatric small carnivores in southeastern China By H. Wang and T. K. FULLER Department of Natural Resources Conservation, University of Massachusetts, Amherst USA Receipt of Ms. 30. 10. 1998 Acceptance of Ms. 20. 02. 2001 Key words: Herpestes urva, Paguma larvata, Vivenicula indica, activity, home range The civets are diverse and prominent ele- ments of old world tropical communities, demonstrating more ecological diversifica- tion in trophic specialization and substrate use than any other family of carnivores (EISENBERG 1981). They are also the least known carnivore group in the world (WEM- MER and WATLInG 1986), especially in Asia, even though many species have ecologıcal and economic significance and have long been harvested for their pelts, meat, and musk. We radio-tracked 5 masked palm ci- vets (Paguma larvata H. Smith, 1827), 2small Indian civets (Viverricula indica Demarest, 1817) and 1 crabcating mon- goose (Herpestes urva Hodgson, 1836) in northern Jiangxi Province, southeastern China during April 1993-November 1994 to understand more about the small carni- vore community there. The study site near Taohong Village is lo- cated in northern Jiangxi Province about 15km south of the Yangtze River and about 490 km WSW of Shanghai. It is in a small, V-shaped vallev about 6 km long at the foot of Mount Taohong and surrounded by a stretch of low and undulating hills (50- 530 m above sea level). The climate is moist monsoon type with typical temperate cli- mate seasonal changes. The average annual temperature is around 16.3°C, and the 1616-5047/01/66/04-251 $ 15.00/0. annual precipitation is 1326 mm, of which over 40% falls as rain during May-July. All the arable lands at the bottom of the valley are under cultivation and many gen- tle hills and slopes also are now farmland. Above the farmland the major vegetation is a combination of tall grasses (Themeda triandra, Imperala cylindirca, and Arundi- nella spp.) and secondary growth of shrub species (Lespecdeza bicolor, L. formosa, Rhus chinensis, and Rhododendron simsi) that is maintained by annual firewood col- lection and frequent fires. Only in some re- mote areas or regions posted by the local forest farms do small patches of deciduous broadleaf and, in rarer frequency, ever- green-deciduous broadleaf forest remain. The northwest part of Taohong Village is included in the Taohongling Sika Deer Reserve that was established to protect a remnant population of the endangered sub- species of sika deer (Cervus nippon kopschi Swinhoe, 1873) in 1981. A general survey on the fauna and flora of the reserve was carried out in and near the reserve during 1988-89; this is the only source of back- ground information for that area (DinG 1990). Most of the study animals were caught by the local trappers with traditional bamboo foot- hold snares (HAn 1960); only one masked 252 H. Wang and T. K. FULLER palm civet was captured in a cage-type live trap. Captured animals were weighed to cal- culate drug dose, then immobilized with Telezol (Tiletamine HCL and Zolazepam HCL) or Ketaset (Ketamine HCI) at a dos- age of lÜ mg/kg body weight. After sex was determined and body measurements re- corded, each animal was fitted with radıocol- lar (with 15-cm whip antennas) weighing <5% of their body weight. All immobilized anımals were held in cages then released after full recovery from the drugs’ effects. Activity of marked animals was determined by listening for changes in radio signal strength during a 60-second period. Conse- cutive readings were taken with an interval of at least 30 minutes. The activity level (% active) was calculated as the number of active readings divided by the total number of readings. Collared animals were gener- ally located a minimum of 3 times a week by walking in on the animals’ resting site during periods of inactivity. The term “daybed” is used to designate those resting sites (RABINoWITZz 1991). The general exter- nal characteristics of daybeds were re- corded to categorize them. Daily move- ments were calculated as the linear distances between 2 consecutive daybeds. Re-use rates of daybeds were calculated as the total number of locations divided by the number of different daybeds (PAro- MARES and DELIBEs 1993, 1994). The resting home ranges of the marked animals were calculated as minimum convex polygons (MonHr 1947) with the RANGES V pro- gram (KENWARD and Hopper 1996), based on the locations of daybeds and capture sites (PALOMARES and DELIBES 1994). Masked palm civets were active >50% of the time between 18.00 and 05.00 hours (Tab. 1). Their activity declined throughout Table 1. Percent of checks indicating activity for radio-marked carnivores studied near Taohong Village, south- eastern China during April 1993-November 1994 Masked palm civet No. of checks Percent active Dr 2 2 2 6 2 0 6 4 Small Indian civet No. of checks Crab-eating mongoose No. of checks Percent active Percent active Notes on the ecology of sympatric small carnivores in southeastern China the morning, with a nadir at 12.00 hours, then remained moderately low until 18.00 hours. Limited data for small Indian civets seemed to mirror that of masked palm civets, but the crab-eating mongoose was clearly very active (56% of 25 readings) when monitored between 10.00 and 18.00 hours (Tab. 1). Masked palm civets have been reported elsewhere to have 2 nocturnal peaks of activity (ZHang et al. 1991), and small Indian civets apparently have either 1 (Wang et al. 1976; RABINO- wırz 1991) or 2 (SHEnG and Xu 1990). Like we found, GAo (1987) indicated that crab- eating mongooses were diurnal. All the daybeds of the masked palm civets were underground burrows, mainly the abandoned dens of porcupines (Aystrix bra- chyura Cuvier, 1822). In contrast, all the daybeds of the small Indian civets we exam- ined were on the ground, usually under dense bushes or among tall grass. In some cases adjacent daybeds were located so close to each other (e.g., 4 daybeds with an area of 2m‘) as to practically form daybed groups. The few daybeds used by the crab- eating mongoose were underground dens. RABınowITz (1991) reported that the small civets (including small Indian civet and masked palm civet in his study) were lo- cated in tree beds 86% of the time; we did not find use of tree beds by the masked palm :civets though there were enough big 253 trees within their habitat. GAo (1987) ob- served that masked palm civets commonly rest in dens in winter and spring and often use the dense bush in the hot summer. Thus, factors other than the availability of proper- sized trees would seem to affect civet daybed selection. All 3 species did not use permanent dens but moved among numerous daybeds. The average daybed reuse rate was 2.5 times for the 5 masked palm civets (range 1.2- 4.2), but increased with the total number of locations obtained (r” = 0.86, 4.d.f., P= 0.005), as might be expected. The reuse rate was 3.6 for the small Indian civet. However, the animals did not use their daybeds ran- domly but showed strong preferences. Over- all, masked palm civets used their daybeds only once (59% of 124), twice (14%) or 3times (11%), but some daybeds were fre- quently used; 7 were used 6-10 times each and 5 were used 10-17 times each. Sımilarly, small Indian civets often used their daybeds only once (43% of 14) or twice (29%), but 4 were used 6-10 times each. The daybeds of the small Indian civet and crab-eating mongoose were located solely in the foothill region adjacent to the farm- land. Some daybeds of masked palm civets were located in the low bushes and tall grass that covered the hilly region border- ing farmland, but many were also farther above in the woods. This difference in habı- Table 2. Altitude (m asl) of daybeds used, distance (m) between consecutive daily locations, and resting home range sizes (ha) of radio-marked sympatric small carnivores studied near Taohong Village, southeastern China during April 1993-November 1994 No. of No. of Sex locations Species daybeds Masked palm civet I2rE23 Sartal7, Naeh ZZ 97 +66 86 + 18 59+14 7,8:1435 Small In- dian civet Crab-eating mongoose different Daybed altitude x+SD° Min. Max.“ x+SD Max. Distance between locations Tracking period 560 + 448 248 + 313 04/93-11/93 06/93-12/93 12/93-07/94 03/94-11/94 10/93-07/94 03/94-06/94 06/94-07/94 02/94-03/94 681 + 414 N7270=1223 613 + 686 254 H. Wang and T. K. FULLER tat use was showed clearly by the altitudes of the locations of the tracked animals (Tab. 2). Although many daybeds of these 3species occurred in close proximity to farmland and trails used by humans, most marked carnivores that were resting were not disturbed by nearby human activity. Elsewhere, masked palm civets prefer dense forest (RABInowITz 1991), and small Indian civets prefer disturbed habitats or forest/agricultural edges (WEMMER and WAT- LING 1986; SHENG and Xu 1990; WANG 1990), similar to our results and indicating an important difference in overall habitat use between those 2 species. The average distance moved between con- secutively used daybeds ranged from 177- 681m (max.=1960 m) for masked palm civets, and 613 m for the small Indian civet (max.=2395, Tab.2). Although masked palm civets and small Indian civets returned to their previous daybeds with similar frequency (32% of 203 movements vs. 38% of 37 movements, respectively), movements to different daybeds were usually shorter for masked palm civets (53% vs. 15% of moves <500m to different daybeds; X” = IST 220.0). The resting home ranges of the 5 masked palm civets located 20-104 times ranged from 182-410 ha (Tab. 2) and did not vary with number of locations (P>0.56). The resting home range for the small Indian civet located 47 times was similar in size (227 ha). The monthly home range for the small Indian civet was 158ha (n=24) in April and 156 ha (n = 20) in May. Although the crab-eating mongoose was only located 7 times, its home range was at least 100 ha. The home ranges of the 2 small Indian civets overlapped, and they and 4 other in- dividuals were caught in the same area. Lo- cal hunters told us that both masked palm civets and crab-eating mongooses were seen in packs of 3 to 4 individuals. Two marked masked palm civets were found sharing the same daybeds more than 20 times. In addi- tion, footprints showed that small Indian civets were usually solitary while crab-eat- ing mongooses often moved in small packs or family groups. In spite of their dietary similarity (WAanG 1999), the resting home ranges of these 3species overlapped extensively. Even though only an unknown portion of all indi- viduals of any 1 species were marked in the study area, daytime home ranges of crab- eating mongooses overlapped 35% and 36%, those of small Indian civets over- lapped 60% and 82%, and those of masked palm civets overlapped 76% and 99% with those of other species. However, none of the radio-marked individuals of any 1 spe- cies used the same daybed simultancously with any other marked individual of an- other species. In the only other telemetry study of masked palm civets, RABINOoWwITz (1991) radio-tracked a single adult female for 12 months. Its total home range was 370 ha, the average daily movement dis- tance was 620m, and the longest daily movement was 1800 m, figures rather close to those we recorded. RABINoWITZz (1991) also reported that an adult male small In- dian civet followed for 6 months had an overall home range of 310 ha, an average daily movement of 500 m, and the longest distance of 2400 m, findings that were again, similar to ours. The home ranges we calculated, however, were certainly mini- mums because they were calculated solely from daybed locations and capture sites, while RABınowITz (1991) used both the daybeds and locations obtained by triangu- lation at night. Acknowledgements We are grateful for the field assistance of TanG Daxong and the logistical assistance of J. Om, S. MEnG, W. Wang, B. WAnG, J. ZHANG, and G. Lı. References Ding, T. (1990): Jiangxi Taohongling Sika Deer Reserve faunal and floral survey. Acta Agric. Univ. Jiangxiensis. Monogr. Nanchang, Jiangxi. (In Chinese) Notes on the ecology of sympatric small carnivores in southeastern China EISENBERG, J.F. (1981): The Mammalian Radia- tions: An Analysis of Trends in Evolution, Adaptation, and Behavior. Chicago: Univ. Chicago Press. GaAo, Y. (1987): Fauna Sinica Mammalia Vol. 8: Carnivora. Beijing: Science Press. (In Chi- nese) Han, Z. (1960): The behavior of and the capture methods for leopards in southern Anhui Pro- vince. Chinese J. Zool. 4, 274-277. (In Chi- nese) KENWARD, R.E.; HoDDEr, K.H. (1990): RANGES V: An analysis system for biological location data. Dorset: Inst. Terrestrial Ecol. MOoHR, C. ©. (1947): Table of equivalent popula- tions of North American small mammals. Am. Midl. Nat. 37, 223-249. PALOMARES, L.; DELIBES, M. (1993): Resting ecol- ogy and behavior of Egyptian mongooses (Herpestes ichneumon) in southwestern Spain. J. Zool. (London) 230, 557-566. PALOMARES, L.; DELIBES, M. (1994): Spatio-tem- poral ecology and behavior of European genets in southwestern Spain. J. Mammalogy 75, 714-724. RABINOWITZ, A.R. (1991): The behavior and movement of sympatric civet species in Huai Kha Khaeng Wildlife Sanctuary, Thailand. J. Zool. (London) 23, 281-298. 255 SHENG, H.; Xu, H. (1990): Range size and activity pattern of small Indian civet in Zhoushan Is- lands, Zhejiang Province by radio-telemetry. J. East China Normal Univ., Mammalian Ecol. Suppl., Pp. 110-112. (In Chinese) Wang, H. (1999): Wildlife conservation in rural southeastern China: wildlife harvest and the ecology of sympatric species. Ph. D. Thesis, University of Massachusetts, Amherst. WAaNnG, P.; SHENG, H.; Lu, H. (1976): The analysis on the food habits of the small Indian civet and its use in captivity breeding. Chinese J. Zool. 20, 39-40. (In Chinese) Wang, Y. (1990). The Viverra zibetha in captivity. Beijing: China Forestry Press. (In Chinese) WEMMER, C. M.; WATLING, D. (1986): Ecology and status of the Sulawası palm civet, Macrogali- dia musschenbroekii Schlegel. Biol. Conserv. 35, 1-17. ZHANG, B.; Su. X.; GAo, C.; ZHANG, W. (1991): Note on the activity and winter dormancy of masked palm civet. Chinese J. Zool. 26, 19- DDR Authors’ address: Haısın Wang and Topp K. FULLER, Department of Natural Resources Conservation, University of Massachusetts. Amherst, MA 01003-4210, USA (e-mail: tkfuller@forwild.umass.edu) Mamm. biol. 66 (2001) 256 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Book reviews MANN, JANET; CONNOR, R. C.; 'TYAckK, P. L.; WHITE- HEAD, H. (eds.): Cetacean societies. Field studies of dolphins and whales. Chicago, London: Univer- sity of Chicago Press 2000. Paperback or cloth, 433 pp., numerous black and white pictures and colour plates. $ US 24.50 or 56.00. ISBN 0-226-50341-0 or 0-226-50340-2. This book gives an account of the state-of-the-art in field studies of dolphin and whale social biol- ogy. Thirteen authors from Canada, USA, and England contribute to this publication. After an introductory chapter on the social lives of Cetacea, the first part of the book, consisting of three chapters, deals with the history of studying cetacean societies as well as with the dynamics of social life and structure. In the second part the so- cial relationships and characteristics of four whale species are dealt with, each of which is presented in a separate chapter, namely, the bottlenose dolphin (Tursiops truncatus), the killer whale (Or- cinus orca), the sperm whale (Physeter macroce- phalus) and the humpback whale (Megaptera no- vaeangliae). In these four chapters detailed information on these species is supplied, not only on behaviour and social relationships, but on bio- logical aspects in general. The social behaviours of these four cetacean species are very different: The bottlenose dolphin lives in a “fission-fusion society”, the killer whale is an active hunter in highly co-ordinated groups. Female sperm whales live a very social life, but males of this same spe- cies are “rovers“ which separate from the female- calve-groups and migrate polewards. The hump- back whale, on the other hand, shows a strong an- nual cycle with seasonal feeding and breeding. The third section of the book discusses results of comparative studies and deals with conservation aspects. For example, reproduction in females and males is discussed in two chapters. The infor- mation on life histories and calf care, as presented by WHITEHEAD and Mann is of general interest and presents information on a large number of odontocete and mysticete species. In a separate chapter functional aspects of cetacean communi- cation — behavioural and acoustic — is presented. Finally, conservation, protection and management of wild cetaceans and an outlook on future behav- ioural studies are discussed in two chapters. The first appendix to the book presents a diagram depicting cetacean phylogeny; in the second ap- pendix a detailed list on cetacean taxonomy 1616-5047/01/66/04-256 $ 15.00/0. Mammalian Biology Zeitschrift für Säugetierkunde (six pages) follows. The list of references is really overwhelming, it fills more than 57 pages, each with more than 30 citations! A citation index re- ferring to authors of original papers mentioned in this book follows and the publication is con- cluded by a subject index of 19 pages. P. LANGER, Giessen CoL£E, T. C. H.: Wörterbuch der Tiernamen; La- tein-Deutsch-Englisch, Deutsch-Latein-Englisch. Heidelberg, Berlin: Spektrum Akademischer Ver- lag 2000. Geb., 970 pp. DM 148,-. ISBN 3-8274- 0589-0. Zur praxisnahen Begutachtung des vorliegenden Wörterbuches wurden stichprobenhaft in der Säu- getierkunde gebräuchlichen Artnamen überprüft. Der Referent nutzte dazu eine von ihm anhand des Werkes von Wırson und REEDER (1993): Mammal species of the world, 2nd ed., zusammen- gestellte Material-Liste von Eutheria, welche von den Insectivora bis zu den Artiodactyla sieben Ordnungen umfaßte. Bei dieser willkürlichen Aus- wahl ergab sich, daß die Mehrzahl der 47 in der Namens-Liste berücksichtigten Arten im Wörter- buch zu finden sind. Drei Vertreter der Goldmulle (Chrysochloridae) werden nicht genannt (Ambly- somus hottentotus, Chrysotalpa duthieae, Ch. scla- teri). Die karibische Seekuh oder Nagel-Manati, Trichechus manatus, wird ebenfalls nicht aufge- führt. Pecari tajacu wird noch als Tayassu tajacu und Moschiola meminna als Tragulus meminna bezeichnet. Im Wörterbuch yon Cole findet sich die Schreibweise Mazama gouazoubira, wo WIL- son und REEDER Mazama gouazoupira schreiben. Das Zwergflußpferd wird an zwei Stellen aufge- führt, unter Hexaprotodon liberiensis und unter Choeropsis liberiansis. Gleiches gilt für das Be- griffspaar Dama dama — Cervus dama. Bei den deutschen Artnamen folgt CoLe der neuen Recht- schreibung: Rupicapra rupicapra ist die Gämse. Im zweiten Teil des Bandes, der die deutsch-latei- nisch-englische Namenstabelle bietet, finden sich folgende Hinweise: „Gämse, Gamswild (Gemse)” und „Gemse, Gamswild (jetzt: Gämse)”. Zoologen, einschließlich der Mammalogen, sind THEODOR C. H. CoLE zu Dank für sein Wörter- buch verpflichtet. Diesem ist ein weiter Benut- zerkreis zu wünschen. Es ist zu hoffen, daß eine zukünftige Auflage unter Hinzuziehung der dann aktuellen Namenslisten bearbeitet wird! P. LANGER, Giessen Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana- ging editor, Prof. Dr. D. Kruska, Institut für Haustier- kunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni- kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manu- script will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21 x 30 cm) or correspondingly either from a typewriter or legible computer print with a left-hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscript. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscnipt following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type- written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscript. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings @:. Mammalian Biology 7° Zeitschrift für Säugetierkunde depending on the theme, Acknowledgements, Zusam- menfassung (in German), References. A German title must precede the Zusammenfassung in English manu- scripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: KaLko, E. K. V.; Hanoıev, C. 0. Ir. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1821), from Panama. Z. Säugetierkunde 59, 257-273. b) From books: HERRE, W.; RÖHRS, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cricetus cnicetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/L, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers“ symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscript is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen, Germany (email:peter.langer@anatomie.med. uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). URBAN & FISCHER Verlag Mammalian Biology Teufen > = Sr > = ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. : 66(2001)4 - 193-256 Contents Original investigations de Moraes, D. A.; Lemos, B.; Cerqueira, R.: - Supernumerary molars in neotropical opossums (Didelphimorphia, Didelphidae) - Uberzählige Molaren bei neotropischen Beuteltieren (Didelphimorphia, Didelphidae) Good, T. C.; Hindenlang, K.; Imfeld, S.; Nievergelt, B.: A habitat analysis of badger (Meles meles L.) setts in a semi-natural forest - Eine Analyse der Habitatcharakteristika von Dachsbauten (Meles meles L.) in einem naturnahen Wald Heide-Jargensen, P. M.; Lockyer, Ch.: Age and sex distribution in the catches of belugas, Delphinapterus leucas, in West Greenland and in western Russia - Alters- und Geschlechtsverteilung von Belugafängen, Delphinapterus leucas, in Westgrönland und Westrußland Ochoa G., J.; Aguilera, M.; Pacheco, V.; Soriano, P. J.: A new species of Aepeomys Thomas, 1898 (Rodentia: Muridae) from the Andes of Venezuela - Eine neue Art von Aepeomys Thomas, 1898 (Rodentia: Muridae) aus den Anden von Venezuela Alves, P. C.; Ferrand, N.; Suchentrunk, F.: Developmental stability and protein heterozygosity in a local population of Iberian hares (Lepus granatensis) - Heterozygotie und Entwicklungshomöostase bei Iberischen Hasen (Lepus granatensis) aus einer lokalen Population in Portugal Short communication Wang, H.; Fuller, T. K.: - Notes on the ecology of sympatric small carnivores in southeastern China - Bemerkungen zur Okologie von sympatrischen kleinen Carnivoren im südlichen China Book reviews Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in 193 204 215 228 238 251 256 Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record Ma immalian Biology Zeitschrift für Säugetierkunde -SMITHSONgHn n > ; 2 6) 0 1 ISSN 1616-5047 URBAN & FISCHER Mamm. biol. - 66(2001)5 - 257-320 www.urbanfischer.de/journalsimammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Säugetierkunde (German Society of Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel . Hutterer, Bonn Kalko, Ulm . Lüps, Bern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storen. Franktunt F. Suchentrunk, Wien F. Trillmich, Bielefeld ‚m m 209 Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repenning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.O. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6%XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/47 61 17, E-mail: (.still@nature.com Subscribers in the USA can call toll free on ++1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 498.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipient's name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by personal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. If no cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH &) As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANST/NISO 239.48-1992 (Permanence of Paper). Printed in Germany ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Mamm. biol. 66 (2001) 257-268 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation The vomeronasal complex in strepsirhine primates and Tarsius By ANGELA WÖHRMANN-REPENNING and M. BERGMANN Zoologie und Vergleichende Anatomie, Universität Kassel, Kassel Receipt of Ms. 11. 09. 2000 Acceptance of Ms. 05. 03. 2001 Abstract The vomeronasal complex (VNC) of several different strepsirhine primates and two Tarsius species was studied with respect to comparative anatomy. All investigated species possess a well devel- oped vomeronasal organ (VNO) in the center of this complex. As Tarsius possesses an extremely small nose, a correspondingly small VNO is present. Its organ is - different to the Strepsirhini and most other mammals - almost completely outlined by an olfactory epithelium. With regard to its histological appearance, however, the VNO should play an important role in all investigated pri- mates concerning sensory ability. It is evident that the VNC in primates follows the progressive de- velopmental line of placental mammals. In this connection the rostral part of the paraseptal carti- lage is an intricate structure and normally forkes into a dorsal and a ventral branch, where the latter usually fuses with the cartilage of the nasopalatine duct. While several Strepsirhini fit into this pattern, some other tend to differ from this scheme, mainly caused by nasal metamorphosis connected with facial reorganisations. All Strepsirhini possess a naked rhinarium split ventrally by a median philtrum, which communicates with the sulcus papillae palatinae. Inside this sulcus, taste buds occur quite frequently at the lateral walls of the palatine papilla. Tarsius a haplorhine primate differs completely by having an unsplit, freely movable, hairy upper lip. Key words: Tarsıus, Strepsirhini, vomeronasal organ, vomeronasal complex Introduction The vomeronasal organ (VNO) is a paired functional importance (BAILEY 1978; WÖHR- accessory olfactory organ present in most MANN-REPENNING 1991) and allows limited mammals and located at the base of the ros- conclusions concerning phylogenetic rela- tral nasal septum. This organ belongs to tionships within the class of mammals an autonomous intranasal system with a (Broom 1898 _WÖHRMANN-REPENNING complicated morphology called the vomero- 1984a, b, 1993b). As a rule, the morphology nasal complex (VNC) (Broom 1898; WöHR- ofthe VNC within a mammalian order fol- MANN-REPENNING 1984a, b). The compara- lows the same type of construction, and tive anatomy of this VNC testifies to its usually the respective species show little in- 1616-5047/01/66/05-257 $ 15.00/0. 258 dividual variations. Those orders that do not fit into this scheme, like insectivores (WÖHR- MANN-REPENNING 1984b) and bats (WiBLE and BHATNAGAR 1996), are of particular in- terest. From this point of view, the order of primates should be of special significance, since it is well known that within the ascen- sion of this order the sense of smell has de- creased from macro- to microsmaty. In- volved in this process are both olfactory systems, the primary nasal, as well as the vo- meronasal olfaction, which in the catarrhine radiation might even have been lost entirely (FrEts 1912; MAIER 1997). Beyond the fact that the facial skull in pri- mates was subject to considerable change during phylogeny (STARCK 1953; BIEGERT 1957; HoFER and Spatz 1963) in this process the eyes oppressed the nasal cavity by mov- ing progressively into a frontal position. In ANGELA WÖHRMANN-REPENNING and M. BERGMANN general, the nose tended to shorten in length almost simultaneously, with the VNC involved. Only few specific publications on the VNC of some species are available (SCHILLING 1970; JoRDAN 1972; HEDEWIG 1980; STARCK 1982; MAIER 1997; BHATNAGAR and MEISAMI 1998). Reference was made to these publi- cations, since they helped to complete the present study whose purpose is to investi- gate the comparative anatomical character- istics of the VNC in a greater number of mainly basale primates with regard to re- corded developmental inquiries. Material and methods For the present study the noses of several, mainly adult primates were available as follows: Fig. 1. Cross section of the VNO of Tarsius bancanus borneanus. 10 um. Azan. The vomeronasal complex in strepsirhine primates and Tarsius 259 Subordo Strepsirhini: Microcebus murinus (Chei- tus, 43mm total length (Galagidae, Lorisoidea) rogaleidae, Lemuroidea), Lemur catta (Lemuri- Galago senegalensis, neonatus (Galagidae, Lori- dae, Lemuroidea), Arctocebus calabarensis (Lori- soidea). sidae, Lorisoidea), Nyticebus coucang (Lorisidae, Subordo Haplorhini: Tarsius bancanus borneanus Lorisoidea), Galago crassicaudatus, adult and fe- (Tarsidae), Tarsius syrichta (Tarsidae). Fig. 2. Schematic and com- parative representation of the merging of the VNO in the in- vestigated species in six Cross sections. Black - bone; stippled - cartilage; 1 - nasal septum; 2 - cartilago parasep- talis; 3 - ductus nasopalati- nus; 4 - ductus vomeronasalis; 5 - vomeronasal organ; 6 - cartilago ductus nasopalatini; 9 - lamina transversalis ante- rior; 10 - outer bar; 11 - un- named cartilage. Bus AR 32 WERKEN TAT AG FIaHE I Ei WIN 2 8 6 4 Nycticebus coucang Tarsius bancanus borneanus 260 The primate material originated to a great extent ANGELA WÖHRMANN-REPENNING and M. BERGMANN from the collection of the late Prof. Dr. H. O. Ho- a light microscope. cross sections FER. A series of differently stained FIT DIT de ee 3 eo Nycticebus coucang Microcebus murinus N ? Sg NW x 1 N 2 2 vi ae Tarsius bancanus borneanus of the rostral noses were studied with the aid of Fig. 3. Schematic and comparative representa- tion of the palatine pa- pilla in the investigated species in Six CrOSS Sec- tions. See Fig.2 for further explanations. The vomeronasal complex in strepsirhine primates and Tarsius Results Vomeronasal organ Apparently, the VNO in all investigated primates is well developed. In Strepsirhini the sensory epithelium covers only the medial side of the organ, while the lateral part is coated by a respiratory epithelium. This coincides with the situation found in most mammals. In Tarsius, however, the or- gan is almost completely covered with olfactory epithelium, and only a few small islands of respiratory epithelium are em- bedded in this area (Fig. 1). This is probably a consequence of the extremely compressed snout of this species, which allows the VNO only to extend up to a length of about 3 mm. In addition, cross sections of the or- gan reveal it to be circularly rounded, while in most mammals the organ is somewhat laterally compressed. Rostrally, the organs of both sides merge into the paired nasopa- latine ducts (Fig. 2). In Lemur, Galago, and Nycticebus this happens in the middle of the ducts deep inside the palate. In Arcroce- bus, as well as in Microcebus, the ducts pe- netrate the gum extremely vertically, while their VNOs merge into the ducts closely neighboured to the nasal floor. Probably therefore the orifices of the VNO are ex- ceptionally oriented towards the oral cavity. Fig. 4. Rostral view of the palate of Galago crassicaudatus. p.p. - palatine papilla; ph. - philtrum. 261 In Tarsius, cross sections show these orifices situated deep inside the nasal floor (Fig. 2). In reality, however, this region proves to be the broad funnel-shaped nasal mouth of the nasopalatine duct. In addition, the openings of the VNO are no small, short ducts like in the other species, but extended longitudinal slits. Rostral palate and palatine papilla All the investigated species possess a well- developed mushroom-shaped palatine pa- pilla, which obviously has the function of a plug, as described in other mammals (WÖHRMANN-REPENNING 1991) (Fig. 3). In Lemur, the papilla is related to its pro- nounced exceptionally large snout and pos- sesses a very broad surface. In Strepsirhini, like in all mammals with a well-developed naked rhinarıum, the sulcus surrounding the lateral sides of the papil communicates directly with the philtrum, which actually is a ventral furrow splitting the rhinarıum (Fig. 4). A cleft exists between their frontal incisors. In accordance their frontal incisors are separated from each other by a median cleft. Tarsius, however, differs from this si- tuation (Fig.5). As in all haplorhine pri- mates (HoreEr 1979), the tip of the nose is a rounded, hairy part connected to uniform 262 ANGELA WÖHRMANN-REPENNING and M. BERGMANN Fig. 5. Rostral view of the palate of Tarsius bancanus borneanus. p.p. - palatine papilla. Fig. 6. Cross section of the palatine papilla of Microce- bus murinus with taste buds (arrows). 10 um Delafield’s hematoxylin and eosin. The vomeronasal complex in strepsirhine primates and Tarsius freely movable upper lips. No median fur- row separates the incisors, instead they are situated closely together. The palatine pa- pilla, not being very prominent, is mainly marked by two small lateral furrows, which are the actual nasal openings of the nasopa- latine ducts. The palatine papilla is a site where taste buds quite often appear in mammals (WÖHR- MANN-REPENNING 1978, 1993 a). The latter studies have shown that they do appear in several strepsirhine primates for instance in Galago, Nycticebus and Microcebus (Fig. 6). Horer (1977) found taste buds in Peridicti- cus potto as well, while no taste buds occur in Lemur, Arctocebus and Tarsius. Cartilages of the VNC In general, the paraseptal cartilage is a cen- tral element of the VNC in mammals, since it is most closely associated to the VNO it- self. It accompanies the organ along its en- tire length, thus forming a gutter-like sup- port. In both Strepsirhini and Tarsius, the paraseptal cartilage reveals a construction typical for most mammals. Its medial edge exceeds the lateral fold significantly in 263 height. An outer bar, which is a cartilagi- nous annular buckle often surrounding the rostral end of the VNO is not commonly distributed in the investigated primates. It is only well-developed in Arctocebus, where it ıs sıtuated caudal to the opening of the VNO, thus forming an extended tubular structure (Fig. 7). Nycticebus tends to have an outer bar of varying form. One indivi- dual showed an outer bar that was devel- oped only at its right side, while in the other anımal both annular bars were developed. These bars, however, are not completely closed, with the lateral fold only showing close contact with the medial side. Here only the perichondria fuse, but not the car- tilages themselves. Obviously exceptionally, there is a distinct outer bar developed in one juvenile Galago crassicaudatus, while this is missing in the adult specimen. With only one exception, i.e. Arctocebus, the paraseptal cartıilage in Strepsirhini has a rostrally forked region. In Microcebus, where the organs open near the back wall of the nasopalatine ducts, this forked region occasionally is situated rostral to the open- ing of the organ. The dorsal branch of this tork ıs always continuous anterior to the la- mina transversalis, while the ventral part en- Nycticebus coucang Arctocebus calabarensis Fig. 7. Schematic representation of the outer bar of the paraseptal cartilage in Nycticebus and Arctocebus. See. Fig. 2 for further explanations. 264 circles the VNO until it opens into the naso- palatine duct. In Tarsius, this part is variable in its construction. Here the forked region seems to be an unstable, delicate element. While such a region was not found in the only investigated individual of T. syrichta, it was present in all three specimens of T. ban- canus, but only in one case, the forked re- gion was distinctly well developed. In both Nycricebus and Tarsius, the ventral branch or part of the paraseptal cartilage rostrally fuses with the cartilage of the naso- palatine duct, which is common in many other mammals as well (WÖHRMANN-REPEN- NING 1984a, b) From this combination a 9 ANGELA WÖHRMANN-REPENNING and M. BERGMANN characteristic sickle-shaped cartilage results which encircles the nasopalatine duct dor- sally (Fig. 8). This seems normal for most mammals, but such a sickle is missing in in- dividuals of Tarsius without a forked region. In those cases there ıs a solid cartilaginous nodal point instead. This sickle formed by two fused cartilages is not existent in both Lemur and Galago. A special situation is present in Microcebus. This species pos- sesses a completely isolated sickle shaped cartilage. SCHILLING (1970) first described and called it “portion en faucille du carti- lage paraseptal”. It encircles the very rostral part of the VNO. The lateral part of this Fa u Fig. 8. A comparative re- presentation of the sickle- \ shaped cartilage in two 8 & strepsirhine primates and 3 2 two non-primate mammals. See Fig. 2 for further expla- Vulpes vulpes Sus scrofa nations. The vomeronasal complex in strepsirhine primates and Tarsius sickle caudally ends abruptly. This special anatomical detail will be discussed later. All investigated species possess a palatine cartilage which serves as a horizontally ori- entated skeletal plug for the incisive win- dow. Except in Microcebus ıt was found to be closely connected with the cartilage of the nasopalatine duct which supports the duct laterally. A papillar palatine cartilage is not com- monly found in the investigated primates. Moreover, this cartilage seems to occur sel- domly, and irregularly. SchiLLinG (1970) did not describe it in Microcebus, while the studied individual has a small cartilage in- side the papil. Correspondingly, it is present and even well developed in only one indivi- dual of an adult Galago crassicaudatus. Finally, another rather odd cartilage in Tar- sius should be mentioned, since it tends to fuse in some cases caudally with the medial edge of the paraseptal cartilage. This ele- ment, already noted and described by STARCK (1982) supports the septal bulge situ- ated dorsal to the VNC (Fig. 2). In T! banca- nus it is either connected with the paraseptal cartilage or nonexistent. In T. syrichta it re- mains in certain distance to the paraseptal cartilage. This skeleton seems to be an un- stable and varying element, primarily meant to stabilize a tuberosity of the nasal septum. Discussion From the present results it might be con- cluded that all strepsirhine primates and all members of the Tarsiiformes have a well developed and functionable VNO. This, however, is not surprising, since all these species possess a distinct marking behavior in which they use urine and/or special glandular secretions (EıBL-EIBESFELD 1953; SEITZ 1969; NıEMITZ 1974; EppLE 1976) and it is suspected that especially pheromones are closely related to a functioning VNO (Estes 1972, MEREDITH et al. 1980; SCHIL- LING et al. 1990; SAasarı et al. 1999). The special histological situation of the VNO in Tarsius might be a consequence of the nar- row intranasal space in this species. 265 In placental mammals the VNC is either based on a primitive or a progressive type of construction. From several comparative studies it can be assumed that the origin of the division into two different lines took place during the early development of the Placentalia. Thus, we find the primitive type in Scandentia, Macroscelidea, Solenodonti- dae as well as in a modified form in Ro- dentia and Lagomorpha (Broom 1898; WÖHRMANN-REPENNING 1980, 1981, 1982, 1984 a, b, 1987). In contrast to this, the ma- jority of placental mammals - also including the majority of insectivores — exhibits a pro- gressively developed VNC. In this case the rostral part of the VNO tends to subside orally, deep into the palate, where it merges directly into the nasopalatine duct, often traversing the palate in an extremely obli- que manner. Due to this situation the ar- rangement of the cartilaginous elements is rather complicated. There is no doubt that primates - in case they have a VNC - with regard to their con- struction, follow the progressively devel- oped line. Here ın this investigation we found that Strepsirhini often demonstrate a pattern which might even be called exemp- lary. Some studied species, however, show slight modifications which can be interpre- tated as first reactions to the tendency ın primates to change and alter the facıal skull. This, for ınstance, leads to some spe- cial features in Microcebus murinus, a spe- cies with a very short nose. The compres- sion of its snout has obviously shifted the VNC. By this means the merging of the VNO is moved to the backside of the naso- palatine duct, and the forked region of the paraseptal cartilage is occasionally situated rostral to this sıte. From this aspect the iso- lated “sickle-shaped” cartılage (SCHILLING 1970) should be the common combination of the paraseptal cartilage and the cartilage of the nasopalatine duct, since this feature is too similar to the situation found in other mammals. Accordingly, the caudal cartilage supporting the lateral wall of the nasopala- tine duct in Microcebus is — taking the shifted VNC into account - the rostral part of the palatine cartilage, which does not 266 fuse in any case with the cartilage of the na- sopalatine duct. The as yet unpublished re- sults of own investigations, show that the two cartilages develop both independently and successively to each other. The results of the investigations conducted on Arctocebus calabarensis should also be discussed, since at the first sight they seem to differ greatly from the normal progres- sive constructional type. This impression arises from the fact that this species has no forked paraseptal cartilage and the fusion of its VNO seems to be sıtuated rather in- side the nasal cavity than in the nasopalatine duct. Cross sections, however, reveal that in this primate the VNC is orientated in an ex- treme vertical position. Due to this position, the VNC of Arctocebus, which originates without doubt from the progressive line, cu- riously regains features of the primitive VNE. Thus, the nasopalatine duct passes through the palate straight down to the oral cavity. Thereby the nasal floor subsides steadily into its tight crater-like opening. Thus in cross sections a missheading impres- sion is gained that the openings of the VNO are still situated inside the nose. Addition- ally, the vertical oriented VNC in Arctoce- bus renders a forked paraseptal cartilage, because the nasopalatine duct passes straight through the incisive foramen. With reference to the primitively constructed VNC, the paraseptal cartilage in Arctocebus has a well-developed outer bar, which, how- ever, is rather an elongated tube than a small bar. In other strepsirhines an outer bar is more or less an exceptional structure. Finally Tarsius should be mentioned. It ıs re- markable that this species, with a nose al- most completely compressed, still has a sur- prisingly well developed VNC (STARCK Zusammenfassung ANGELA WÖHRMANN-REPENNING and M. BERGMANN 1982). In general, it shows several conformi- ties to those features found in strepsirhines. The forked region of the paraseptal carti- lage however seems to have lost its func- tional importance with the very small nose. It is only well-developed in one individual. At the same time, none of the investigated species demonstrates as clearly as Tarsius that the VNC can be altered greatly for ex- ample, when the nasal cavity is forced to transform, and thus forced to reduce struc- tures because of dramatic facial changes. Regarding the external situation of the VNC in strepsirhine primates we find a situation characteristic for many mammals. These re- markable structures are closely connected to the functional mechanism of the VNO (Horer 1977, 1980; WÖHRMANN-REPENNING 1991). The rostral palatal features in Tarsius differ from this in that it belongs to the hap- lorhine primates. In strepsirhine primates taste buds are commonly found at the ven- tro-lateral sides of the papilla palatina. They are mainly situated near the entrance of the nasopalatine duct. Their role in connection with sensory abilities of the VNO in mammals was generally discussed in a previous study (WÖHRMANN-REPENNING 1993a). The pre- sence of taste buds indicates a dual chemo- sensory system combining smell and taste for functions of the VNC. In those cases, where taste buds are missing at the palatine papilla, as it can be seen in Lemur, Arctocebus and in Tarsius, one may assume that lingual taste buds interact with the VN-olfaction. Acknowledgements We would like to thank MARION KÜnn for kindly revising the English text. Der Vomeronasalkomplex bei strepsirhinen Primates und Tarsius Der Vomeronasalkomplex (VNC) verschiedener strepsirhiner Primates und von zwei Tarsius Species wurde unter vergleichend anatomischen Aspekten untersucht. Alle Arten besitzen ein gut entwickel- tes Vomeronasalorgan (VNO). Bei Tarsius ist das VNO auf Grund seiner winzigen Nasenhöhle entspre- chend klein ausgebildet. Offensichtlich um dies zu kompensieren, ist das Organ - anders als bei den The vomeronasal complex in strepsirhine primates and Tarsius 267 Strepsirhini und den meisten anderen Mammalia - in seinem gesamten Lumen fast vollständig von olfaktorischem Epithel ausgekleidet. Das VNO aller untersuchten Primaten läßt auf Grund seiner hi- stologischen Beschaffenheit vermuten, daß ihm eine wichtige Funktion im sensorischen Leben der Tiere zukommt. Die Befunde lassen klar erkennen, daß der VNC der Primates dem progressiv entwik- kelten Typus zuzuordnen ist. Charakteristisch für diese Entwicklungslinie ist ein im rostralen Ab- schnitt diffizil gestalteter Paraseptalknorpel, der sich zudem in diesem Bereich in der Regel in einen dorsalen und einen ventralen Ast gabelt. Beide Teile neigen zum Fusionieren mit anderen Knorpeln. Die Mehrzahl der untersuchten Primaten besitzt einen progressiv entwickelten VNC, dessen Gesamt- struktur dem anderer Mammalia bis in kleinste Details ähnelt. Bei einigen Vertretern jedoch lassen sich Besonderheiten ausmachen, die größtenteils im Zusammenhang mit der für Primates charakteri- stischen Umgestaltung des Fazialschädels zu sehen sind. Alle Strepsirhini besitzen ein nacktes Rhi- narium, dessen Philtrum ventral mit dem Sulcus papillae palatinae kommuniziert. In diesen Sulcus aber münden die Ductus nasopalatini ein, und genau in diesem Bereich befinden sich bei einigen der untersuchten Strepsirhini Geschmacksknospen. Tarsius besitzt kein Rhinarium, sondern als Ange- höriger der haplorhinen Primaten eine ungeteilte, behaarte, frei bewegliche Oberlippe. References BAILEY, K. (1978): Flehmen in the ring-tailed le- mur (Lemur catta). Behaviour 65, 309-319. BIEGERT, J. (1957): Der Formwandel des Prima- tenschädels und seine Beziehungen zur onto- genetischen Entwicklung und den phyloge- netischen Spezialisationen der Kopforgane. Morph. Jb. 98, 77-199. BHATNAGAR, K. P.; MEISAMI, E. (1998): Vomerona- sal organ in bats and primates: Extremes of structural variability and its phylogenetic im- plications. Microsc. Res. Tech. 43, 465-475. Broom, R. (1996): On the comparative anatomy of the organ of Jacobson in marsupials. Proc. Linn. Soc. New South Wales 21, 591-623. BRooM, R. (1998): A contribution to the compara- tive anatomy of the mammalian organ of Ja- cobson. Trans. Roy. Soc. Edinburgh 39, 231- 259) ErPLE, G. (1976): Chemical communication and reproductive processes in nonhuman pri- mates. In: Mammalian Olfaction, Reproduc- tive Processes, and Behavior. Ed. by R.L. Dorty. New York: Academic Press. Pp. 257- 282. EsSTEs, R. D. (1972): The role of the vomeronasal organ in mammalıan reproduction. Mamma- lia 36, 315-341. FRETS, G. P. (1912): Beiträge zur vergleichenden Anatomie und Embryologie der Nase der Pri- maten. Morph. Jb. 44, 409-463. HEDEWIG, R. (1980): Vergleichende anatomische Untersuchungen an den Jacobsonschen Orga- nen von Nycticebus coucang Boddaert, 1785 (Prosimiae, Lorisidae) und Galago crassicau- datus E. Geoffroy, 1812 (Prosimiae, Lorisi- dae). Gegenbaurs morph. Jb. 126, 543-593, 676-722. HoreEr H. ©. (1977): The anatomical relations of the ductus vomeronasalis and the occurrence of taste buds in the papilla palatina of Nycti- cebus coucang (Primates, Prosimiae) with re- marks on strepsirhinism. Gegenbaurs morph. Jb. 123, 836-856. HoreEr, H. ©. (1979): The external nose of Tarsius bancanus borneanus Horsfield, 1821 (Pri- mates, Tarsiiformes). Folia primatol. 32, 180- 192. HorFER, H. ©. (1980): The external anatomy of the oro-nasal region of primates. Z. Morph. Anthrop. 71, 233-249. HOFER, H. O.; SPATZ, W. (1963): Studien zum Pro- blem des Gestaltswandels des Schädels der Säugetiere, insbesondere der Primaten. Z. Morph. Anthrop. 53, 29-52. JORDAN, J. (1972): The vomeronasal organ (of Ja- cobson) in Primates. Folia morphol. (Warsz.) 31, 418-431. Maier, W. (1997): The nasopalatine duct and the nasal floor cartilages in catarrhine primates. Z. Morph. Anthrop. 81, 289-300. MEREDITH, M.; MARQUES, D. M.; O’ConNELL, R. ).; STERN, F.L. (1980): Vomeronasal pump: Sig- nificance for male hamster sexual behavior. Science 207, 1224-1226. NIEMITZ, €. (1974): A contribution to the postna- tal behavioral development of Tarsius banca- nus Horsfield, 1821, studied in two cases. Fo- lia primatol. 21, 250-276. SASAKI K.; OKAMOTO, K; INAMURA, K.; TOKU- MITSU, Y.; KASHIWAYANAGI, M. (1999): Inosi- 268 tol-1,4,5-trıphosphate accumulation induced by urinary pheromones in female rat vomero- nasal epithelium. Brain Res. 823, 161-168. SCHILLING, A. (1970): L’organe de Jacobson du Lemurien malgache Microcebus murinus (Miller 1977) Me&m. Mus. Nat. Hist. Nat. Paris 61, 203-280. SCHILLING, A.; SERVIERE, J.; GENDROT, G.; PER- RET,M. (1990): Vomeronasal activation by urine in the primate Microcebus murinus a 2 DG study. Exp. Brain Res. $1, 609-618. SEITZ, E. (1969): Die Bedeutung geruchlicher Or- ientierung beim Plumplori Nycticebus cou- cang Boddaert 1785 (Prosimiae, Lorisidae) Z. Tierpsych. 26, 73-103. STARCK, D. (1953): Morphologische Untersuchun- gen am Kopf der Säugetiere, besonders der Prosimier, ein Beitrag zum Problem des Formwandels des Säugerschädels. Z. wiss. Zool. 157, 169-219. STARCK, D. (1982): Zur Kenntnis der Nase und des Nasenskelettes von Tarsius (Mammalia, Pri- mates, Tarsioidea) Zool. Garten N.F. Jena 32, 289-304. WIBLE, J. R.; BHATNAGAR, K.P. (1996): Chiropter- an vomeronasal complex and the interfamilial relationship of bats. J, Mamm. Evol. 3, 285- 314. WÖHRMANN-REPENNING, A. (1978): Geschmacks- knospen an der Papilla palatina von Tupaia glis (Diard, 1820), ihr Vorkommen und ihre Beziehungen zum Jacobsonschen Organ. Ge- genbaurs morph. Jahrb. 124, 375-384. WÖHRMANN-REPENNING, A. (1980): The relation- ship between Jacobson’s organ and the oral cavity inarodent. Zool. Anz. 204, 391-399. WÖHRMANN-REPENNING, A. (1981): Die Topogra- phie der Mündungen der Jacobsonschen Or- gane des Kaninchens (Oryctolagus cuniculus) unter funktionellem Aspekt. Z. Säugetier- kunde 46, 273-279. ANGELA WÖHRMANN-REPENNING and M. BERGMANN WÖHRMANN-REPENNING, A. (1982): Vergleichend- anatomische Untersuchungen an Rodentia. Phylogenetische Überlegungen über die Be- ziehungen der Jacobsonschen Organe zu den Ductus nasopalatini. Zool. Anz. 209, 33-46. WÖHRMANN-REPENNING, A. (1984a): Verglei- chend anatomische Untersuchungen am Vo- meronasalkomplex und am rostralen Gaumen verschiedener Mammalia. Gegenbaurs morph. Jahrb. 130, 501-530; 609-637. WÖHRMANN-REPENNING, A. (1984 b): Phylogeneti- sche Aspekte zur Topographie der Jacobson- schen Organe und der Ductus nasopalatini bei Insectivora, Primates, Tupaia und Didel- phis. Anat. Anz. 157, 137-149. WÖHRMANN-REPENNING, A. (1987): Zur Anatomie des Vomeronasalkomplexes von Elephantulus rozeti (Duvernoy, 1830) (Macroscelidea Mammalia). Zool. Anz. 218, 1-8. WÖHRMANN-REPENNING, A. (1991): Functional as- pects of the vomeronasal complex in mam- mals. Zool. Jb. Anat. 121, 71-80. WÖHRMANN-REPENNING, A. (1993 a): The vomero- nasal complex — A dual sensory system for ol- faction and taste. Zool. Jb. Anat. 123, 337-345. WÖHRMANN-REPENNING, A. (1993 b): The anatomy of the vomeronasal complex of the fox (Vulpes vulpes (L.)) under phylogenetic and functional aspects. Zool. Jb. Anat. 123, 353-361. WÖHRMANN-REPENNING, A.; BARTH-MÜLLER, U. (1994): Functional anatomy of the vomeronasal complex in the embryonic development of the pig (Sus scrofa dom.) Acta Theriol. 39, 313-323. Authors’ addresses: Prof. Dr. ANGELA WÖHRMANN-REPENNING, ZO- ology and Comparative Anatomy, University of Kassel, Heinrich-Plett-Str. 40, D-34132 Kassel, Germany (e-mail: woehrep@hrz.uni-kassel.de) Dr. MATTHIAS BERGMANN, Nordring 17, D-33330 Gütersloh, Germany. Mamm. biol. 66 (2001) 269-280 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Review Postnatal brain size decrease, visual performance, learning, and discrimination ability of juvenile and adult American mink (Mustela vison: Carnivora: Mammalia) By KATJA STEFFEN, D. KRUSKA, and R. TIEDEMANN Institut für Haustierkunde, Universität Kiel, Kiel, Germany Receipt of Ms. 24. 07. 2000 Acceptance of Ms. 17. 10. 2000 Abstract The decrease in brain size to an amount of 20% during early life from the juvenile to the adult state in mink and other Mustela species is still poorly understood and unresolved in its general biological and functional relevance. The same holds true tor the decrease in brain cavity and the flattening of the cranial vault. Since the neocortex and other brain parts with higher integrative and associative sensory and motoric functions are especially involved, the question arises as to whether these size changes have any functional consequences, i.e., are the functional capacities reduced concomi- tantly? This was tested for the visual system in 8 juvenile (4, 4) and 8 adult (4, 4) mink (Mustela vison energumenos) of wild descent using twofold-choice discrimination trials. After conditioning and testing for spontaneous side preferences, the individuals had to discriminate black dots of dif- ferent sizes against a white plate from 30 cm distance. Altogether, 16 000 individual data-sets were statistically analysed for differences in visual performance, in learning velocity, and in discrimina- tion ability. No differences occurred between the juvenile and the adult group concerning learning velocity. However, significant differences were found in discrimination ability with regard to age (juvenile mink performed better than adults) and sex (females performed better than males). These results are discussed with regard to the importance of visually guided behaviour of the species, with the behaviour of juveniles and adults in general, and with the ontogenetic decrease in mass of the central nervous tissue. According to this study, there is no indication of any functional im- pact of the ontogenetic reduction in brain size on the capacity of the visual system. Key words: Mustela vison, ontogenesis, vision, brain size Introduction Unlike most other mammals, some species This phenomenon was confirmed by brain of the Mustela genus exhibit a peculiar and skull comparisons in the two domesti- postnatal brain ontogeny, i.e., their brain cated forms of ranch mink (Mustela vison f. significantly decreases in size in both sexes dom.; KruskA 1977, 1979, 1993) and ferret shortly before the adult stage is reached. (Mustela putorius f. furo; APFELBACH and 1616-5047/01/66/05-269 $ 15.00/0. 270 KATJA STEFFEN et al. KruskA 1979; ESPENKÖTTER 1982) as well as in wild populations of feralized ranch mink (Wi 1982, 1985) and the weasel (Mustela nivalis; SCHMIDT 1992). Furthermore, brain size decreases due to domestication were additionally found. This results from com- parisons of adult wıld American mink with adult ranch mink (KruskA 1996, KRUSKA and SCHREIBER 1999) as well as polecat with ferret (ESPENKÖTTER 1982). However, these changes of brain size in the course of do- mestication differ considerably from those in ontogenesis (KRUSKA 1987) since differ- ent brain parts are involved in the two re- spective processes to highly diverse degrees (KruskA 1993, 1996). Although not all spe- cies have been investigated so far, it can be assumed that the ontogenetic decrease in brain size appears in all Mustela species, in- cluding the American wild mink. In comparative brain anatomy, intra- or in- terspecific differences in the mass of brain substance are usually discussed in connec- tion with a functional increase or decrease in general central nervous capacity (KRrUS- KA 1988 a). Because ontogenetic brain size decrease mainly occurs at an individual age from 5 to 7 months in mink, its biological relevance is especially puzzling. This is the time when the maternal family breaks up and the subadults disperse, searching for their own future home ranges (KRUSKA 1988 b). A decrease ın central nervous capa- city at this stage of ontogeny appears unli- kely. Quantitatively, the diencephalon decreases by 9.5% in size independently of body size, but the most extreme decrease was found in the telencephalon, this being 22% smal- ler in size in adults compared with 5- month-old individuals. Here, the total neo- cortex (23%) and especially its grey matter (27%) are most dramatically decreased (KruskA 1993). As these impressive size reductions affect brain parts with especially higher associa- tive, coordinative, and integrative central nervous functions, they should have a bear- ing on general ethology or special behav- ioural performance, if functionally relevant at all. In this framework, the present study aims at comparing the capacities of one ex- emplary sensory system in juvenile vs. adult wild mink (Mustela vison energumenos) namely visual performance, visual learning, and visual discrimination ability. The con- cerned brain parts, especially the occipital neocortex grey matter are especially in- volved in these general visual functions. Material and methods Animals In total 16 (8m, Sf) wild mink (M. vison energu- menos) were used in this study. Eight (4 m, 4 f) ju- venile and 8 (4 m, 4 f) adult individuals were born at the Institut für Haustierkunde of the Univer- sity of Kiel. They were first generation offspring of wild parents that were caught in the vicinity of Whitehorse (Yukon Territory, Canada), trans- ported to Kiel, and kept there in large open air enclosures. The test animals were fed mainly on newly hatched chicks and housed separately in differently sized (22 m’-35 m?) wire mesh sur- rounded open air enclosures on natural ground with vegetation, nest boxes, and water basins. White spots different in size and pattern on the pelage of throat, breast, and belly made them in- dividually recognizable. Additionally they were named using G as initial letter for the juveniles and F for the adults. The juvenile mink were on average 3.6 months (range: 98-117 days) old at the start of the experiments, the adult ones 11.2 months (range: 337-343 days). Experimental arena Twofold-choice discrimination trials were run in a simple experimental arena of wood (1.7m in length, 0.5 m in breadth and 0.2 m in height). This arena (Fig. 1) consisted of one discrimination chamber (A) covered with wire mesh and two ad- jacent reward boxes (B) covered with a wooden plate. Lockable one-way swinging doors (C), 16x16cm in dimension, allowed access to each of the reward boxes if unlocked. Plates with visual discrimination test stimuli could easily be placed on these swinging doors. There was also a parti- tion wall (D) between these swinging doors pro- truding 30 cm into the discrimination chamber, thus forming a left and a right tunnel at its lower end. Entrance to the arena was only possible through a trap door (E) that could not be opened Brain size decrease and visual performance of juvenile and adult Mustela vison 0.2m Fig. 1. Scheme of the experimental arena. 271 A - discrimination chamber; B - reward boxes; C - one-way swinging doors; D - partition wall; E - entrance trap door by the animals. One-way swinging doors at the end of the reward boxes were the exits from the arena. Experimental tests The arena was placed in the diverse rearing enclo- sures of the mink, thus trials were run in accus- tomed environments during normal daylight. The mink were tested to discriminate visually between a totally white, blank plate and diverse white plates with central black circles of different dia- merer (10730725,20515,12597573,and I mm): The experiments were performed in four phases identical for all animals tested. During Phase 1 each of the mink was habituated to the arena and observer (K.S.). At this time all the doors were kept unlocked and pieces of food (chick parts) were randomly placed in one of the reward boxes only. Access was possible through the entrance, so the mink was conditioned to the one-way path through the arena. During Phase 2 individual side preference and its constancy were tested. Equally sized pieces of food were now offered simulta- neously in both reward boxes during each trial. Each animal performed 100 runs over 2 to 3 suc- cessive days. The random right or left side choices 272 KATJA STEFFEN et al. were recorded. During Phase 3 the anımals were tested for spontaneous preference of visual stimu- li. Now, for the first time the discrimination plates were affıxed to the still unlocked swinging doors leading to the reward boxes both loaded with food rewards. In the first trial, the white plate was pre- sented on one side against the white plate with a central black circle 70mm in diameter on the other side. Each anımal had to run 50 trials on one day. During these trıals the position of the plates was changed from left to right and vice ver- sa in accordance with the schedule of GELLERMAN (1933). Hereby it was ensured that (1) the subse- quent side (left/right) was unpredictable to the anımals, (2) no plate was offered more than maxi- mally three times on the same side in consecutive trials to avoid possible side preferences, and (3) both sides (left/right) occurred overall at the same frequency. Phase 4 finally represents the en- tire tests aimed at for visual discrimination and learning. In this, the white plate was designated the incorreet choice and consequently the swing- ing door bearing this plate was locked during all following trials, although the associated reward box always contained equal quantity of meat re- ward. This controlled for choices made on possible olfactory cues. In each trial, the mink had to make their decision at a30cm distance from the swinging doors and choices were evaluated by the observer (K.S.) as positive or negative. These evaluations followed stringent criteria. Choices were counted as posi- tive only in cases where animals ran to the posi- tive sign immediately after entrance into the dis- crimination chamber. In cases where they showed preference for the wrong side or started to move towards the negative stimulus, this was counted as a negative choice even if they seemed to recognize the failure and corrected their choice later. Trials with the same plates were run 50 times on one day by each of the mink followed by a day of relaxation and 50 trials on the subsequent day. A higher number of trials daily led to erratic per- formance in later trials, probably due to satiation after the positive rewards. The time point at which conditioning to the 70 mm eircle had occurred was very clearly recog- nisable through a change in behaviour. From that moment on, mink moved rather rapidly and di- rectly towards the correct sign. The number of runs preceding this conditioning was taken as a measure of individual learning velocity. The en- tire 100 discrimination trials with the 70 mm circle were conducted after conditioning had taken place. Only after completing the 100 trıals on the 70 mm circle during two days the next smaller sign was tested in the same manner, regardless of the animal’s prior performance, and the trials con- tinued in the same way for all smaller circles. The order in which the individual animals were tested and the time of day (08.00 h-12.00 h) the trials were run was constant throughout the experi- ment. Each of the 16 mink therefore completed a total of 1000test trial. Thus, altogether 16000 runs yield the data base for statistical treat- ment. Data analysis Choices during Phase 1 were not analysed statisti- cally. Habituation to the apparatus and condition- ing to find food were continued until the animals performed efficiently. Choices of Phase 2 and 3 were treated statistically. The significance of spontaneous preferences for side (right/left; Phase 2) and visual stimuli (stimulus present/ab- sent; Phase 3) was assessed by a replicated good- ness-of-fit x”-test. During the learning phase, mink were considered as trained to the visual sti- mulus if they correctly chose 63 out of 100 trials. This corresponds to a probability of p = 0.99, i. e. the mean rate of correct choice is significantly larger than expected at random (= the lower limit of the 99% confidence interval for the mean rate of correct choice exceeds 0.5; SOKAL and ROHLF 1995). The total of 16000 individual trials of the entire Phase 4 was jointly analysed using a log-lin- ear model for frequency data (BısnHor et al. 1975; SokAL and RoHLF 1995). By this approach we ex- amined the influences of age, sex, and sign size on the discrimination efficiency as well as the var- ious interactions among these factors. Results General behaviour of the individuals The individual mink differed slightly in gen- eral behaviour prior to as well as during the experiments. The adult mink, e.g., were ac- customed to their enclosures for a longer period and were minimally disturbed by hu- mans. During Phase I, they acted rather hesitantly and shyly. The juveniles, in con- trast, seemed to be more bold and curious. Adult individuals required up to 20 sec. for a run through the experimental arena, while juveniles clearly operated faster, on average only 4-8 sec. per run. Apart from this, some Brain size decrease and visual performance of juvenile and adult Mustela vison individuals of both age groups were espe- cially shy, labile, and less concentrated dur- ing the trials (Gira, Giwa, Feka), while others remained comparably calm and diffi- cult to motivate (Galbus, Fabea, Felia, Fe- dor, Falkano, Fargo). In contrast, still others acted very excited, concentrated, and fast. They often vocalised the typical twitters of excitement (Gilia, Galicia, Gaius, Gagarin, Godu, Fiona, Fibius) during trials. Sex-spe- cific differences were especially marked. In cases of a wrong decision, for example, adult males tended to be aggressive and bit, scratched and tore at the locked en- trance to the reward box but such behav- iour was never observed in females. In com- parable situations, females only corrected their erroneous decision by running to the positive sign with the food reward. Thus, although such a choice was negatively eval- uated for the trial, females showed a some- what greater plasticity in such situations. This is also reported for many other muste- lids.. Females, both adults and juveniles, were also more active in general compared to males, especially during the habituation and conditioning phases. Males were clearly less motivated and comparably disinter- ested. All the animals, however, completed all trials although in individually different ways. Side preference and spontaneous preference for visual stimuli Summarised results of the test for sponta- neous side preference and side constancy of choices are given in table 1. There was no significant overall side preference (repli- cated goodness-fit: sed —_ 023 dt ih p = 0.617), but the side preference was sig- nificantly heterogeneous among individuals (replicated goodness-fit: ee = 40.00, df=15, p< 0.001). This was caused by four individuals exhibiting a sponta- neous side preference, significant at p<0.05. Albeit this preference was only slight (62:38 at maximum) and symmetrical (two individuals preferred left, two right), the potential influence of individual side preference on the visual discrimination effi- 273 Table 1. Individual side preference prior to condition- ing (100 runs per animal) Individual side choice left right Gilia Gira Galicia Giwa Galbus Gaius Godu Gagarin Fiona Felia Feka Fabea Fedor Falkano Fibius Fargo female juvenile adult female ciency was estimated in a separate log line- ar model. Most mink showed no preference for either a blank white plate or a 70 mm black circle. Based on 50 runs per animal and a signifi- cance level of p = 0.05, only two individuals seemed to slightly prefer the circle, while three others did so with the white plate. Learning velocity Regarding individual learning velocity, the juvenile mink clearly showed less variability of run numbers (range: 57 to 95) than the adults (range: 51 to 200) prior to condition- ing (Tab. 2). This was caused by two parti- cular adults (Fibius, Feka) learning very slowly, with 200, respectively, 150 trıals prior to conditioning. However, all the other adults reached values within the var- iance of the juveniles. Over all, there were no significant differences among the four groups, 1. e., juvenile and adult mink of both sexes (Kruskal-Wallis-Test: H = 4.83, df=3, p= 0.185). Thus, learning velocity during conditioning of individual mink can be con- sidered identical and independent of age and sex. 274 KATJA STEFFEN et al. Table 2. Learning velocity during conditioning (runs Table 3. Significant factors determining visual discri- per animal until they were considered conditioned; see mination ability, given as % variation explained (result methods for details). of a log-linear model analysis). Age Sex Individual number of runs % of var- iation ex- juveniie female Gilia plained Gira Galicia Sex <0.001*** ie Age <0.001*** Galbus Gas Sign size <0.001*** Godu Interaction O:00A1S Gagarin sex vs. age adult female Fiona Felia Feka Fabea Fador total explain- Falkano ed variation Interaction <= 0.001222 age vs. sign size Fibius unexplained Fargo variation From 30% unexplained variation, 10% could be at- tributed to individual side preferences (as indicated by a separate log-linear model). 100 (2) © e 90 ® >} oO E ° © y = 0.9979x + 0.1543 = 60 R? = 0.7009 17) D © 40 50 70 90 fitted frequencies (log-linear model) Fig. 2. Correlation between fitted and observed frequences of correct choices in the visual discrimination experi- ment, when sex, age, sign size and the two interactions sex vs. age and age vs. sign size are included in the log- linear model. Brain size decrease and visual performance of juvenile and adult Mustela vison 275 Table 4. Individual visual discrimination efficiencies of the juvenile and the adult mink. Numbers of positive choices are presented out of 100 runs on each of the black circles different in diameter. Grey shaded fields indi- cate performances below the significance level of 63%. Mean values of groups (X) and standard deviations (SD) are also given. Diameter (in mm) of black circle positive sign age/sex 30 25 Gilia Gira Galicia Giwa Galbus Gaius Godu Gagarin juvenile juvenile Fiona Felia Feka Fabea Fedor Falkano Fibius Fargo 795 5.74 712.9 1222 80.3 3.86 7/3081 12.6 females 6.04 Visual discrimination efficiency According to the log-linear model, the key determinant of the visual discrimination efficiency was the sign size (accounting for 35% of the variation in discrimination effi- ciency), followed by age and sex, explain- ing 14% and 4% of the variation found, respectively (Tab. 3). Two two-factor inter- actions were significant, the interactions age vs. sign size and age vs. sex, accounting for 14% and 3% of the found variation in discrimination efficiency, respectively. Al- together, these factors explained 70% of the present variation and yielded a good fit between observed discrimination effi- ciency and values expected by the model (Fig. 2). Of the remaining 30% variation unexplained, about 10% can be attributed 115 12 11.3 8.81 67.3 15.3 7a 8.27 66.1 8.28 68.3 10.9 66.4 8.69 59.3 4.79 56.8 5.80 to individual side preferences (as indicated by a separate log-linear model), while the remaining 20% apparently represents indi- vidual differences and stochastic errors. In- dividual efficiencies to discriminate differ- ently sızed black circles are summarized in table 4. Evidently, the mınk were generally unable to recognize a black dot sign 1 mm in diameter (except for one juvenile male). Additionally, six out of the 16 animals had also difficulties to discriminate the 3mm sign. Thus, to discriminate a motionless sign of this size from a distance of about 30 cm may be problematical for mink in general. In this context, adult males are pe- culiar because of their different efficiencies from sign to sign. This specifically poor performance of adult males — especially with smaller signs — is the apparent expla- 276 KATJA STEFFEN et al. D] juvenile adult 100 - right choices (in %) 70 300 2 20 15 12 9 5 3 1 diameter of sign (in mm) Ofemales EB males 100 right choices (in %) | © diameter of sign (in mm) Fig. 3. Arithmetical means of positive choices for the differently sized signs of juvenile vs. adult and of female vs. male mink. Dotted line indicates the 63% level; vertical lines indicate standard deviations. Brain size decrease and visual performance of juvenile and adult Mustela vison nation for the two significant factor inter- actions (age vs. sign size; age vs. sex). Very obviously their minor concentrations on the tasks are reflected in these results. Nevertheless, except for the smallest signs mink chose on average significantly more frequent correct than wrong, regardless of their sex and age (Fig. 3). Discussion In connection with the aim of this study it seems worth mentioning what is known about the importance of visually guided be- haviour in the biology of the investigated species. Thus mink - like other Mustela spe- cies — are in general basal carnivores with only little anatomical specialisation (EWER 1973). They are predators which regionally and seasonally prey on different vertebrates and invertebrates, hunting and chasing on land but since they are adapted to a semi- aquatic life style larger parts of food are also taken from rivers, lakes, or the sea. They are very efficient hunters, reacting quickly both in air and under water. They are thus dependent on their sensory organs for detection of prey, chasing, and killing but also for recognizing their own preda- tors. Concerning their general orientation, they often are said to be sensory-guided mainly by olfaction followed by hearing and then by vision. However, such ranking in the importance of sensory organs and central nervous circuits for species-specific biological peculiarities seems rather mean- ingless since according to our own observa- tions and those of others as well, mink very efficiently hunt fish in muddy waters even during dark times of the day and also in clear, reflecting waters during bright sun- shine. Thus, it can be suggested that at least during these hunts they are mainly guided by sensory functions of the vibrissal system rather than by olfaction, hearing, or vision. Therefore, it seems more likely to conclude that all sensory systems are of certain im- portance for the biology of mink with dif- ferent priorities at given situations. How- ever, mink have well developed eyes and 277 differentiated central nervous visual struc- tures. The duplex retina is built up by highly differentiated and extremely polarized rods and cones, with the rods being more numer- ous (Dugın and TURNER 1977; BRAEKEVELT 1990; STEFFEN 2000). The lateral geniculate body as the main termination nucleus of vi- sual perception is well differentiated (GuiL- LERY et al. 1979) showing certain laminae and an input of approximately about 75% to 80% crossed and 20% to 25% uncrossed fibres (SANDERSON 1974; SANDERSON et al. 1975). Accordingly, the striate area at the occipital lobe of the neocortex is well orga- nized and histologically clearly distinct from surrounding areas (LEVAy et al. 1987; GUILLERY and ÖBERSDORFER 1977; GUIL- LERY et al. 1979). In relative size this highest sensory field accounts for about 6.83% of the total neocortical grey matter (own preli- minary results). All these anatomical cha- racters can without doubt be indicative for what normally ıs called good vision in con- trast to poor eyesight. This is also con- firmed by ethological and physiological tests and findings (DunstonE 1993; Dun- STONE and SINCLAIR 1978a, b). Acuity of predators for stationary objects (like dots) is less poorly developed than for moving objects. A recognition of black dots differ- ent in sıze might therefore be of limited im- portance for the biology of a species but at least documents its visual acuity and func- tional complexity. However, as our results indicate the anımals acted in response to these stimuli and they were able to discri- minate different sizes. Thus, the reactions of the animals on the experimental tasks may serve as a measure for functional sen- sory and central nervous capakcities. In order to uncover a probable relationship between the size of a brain structure and its complex functional capacity, the quanti- tative structural changes of central nervous tissue from juvenile to adult mink (KRUSKA 1993) can be correlated with quantitative functional differences as resulted from the visual test procedures presented here. Con- cerning such a relation, however, no un- equivocal conclusions can be drawn, as apparent correlation may appear coinciden- 278 KATJA STEFFEN et al. tally. As example, juvenile mink, known to possess significantly larger neocortical and diencephalic brain regions, generally showed significantly better results in reac- tions to visual discrimination tasks than adults both in female and male individuals. Consequently, this could be considered an indication for greater functional capacities of juveniles due to larger central nervous structures. This is, however, contradicted by the finding that learning velocity is inde- pendent of age. On the other hand, differ- ences were observed among sexes concern- ing their discrimination ability, although mass and proportioning of the brain did not differ among sexes, either within the larger brained juveniles or within the adults (KruskA 1993). For these reasons, the post- natal overshooting of brain size has most probably no functional consequences for the biology of mink, at least not for the cap- abilities of the visual system. The fact that juvenile individuals are more playful, cur- ious, easier to be conditioned, open for learning and consequently more efficient in behavioural experiments than adults is commonly known from many other mam- malian species (WÜSTEHUBE 1960; GOETHE 1964; REnscH 1973; TEMBROCK 1982) includ- ing those with normal brain size develop- ment. Thus, the postnatal brain size de- crease most probably has no effects on visual functions and still remains unsolved in its biological relevance. Acknowledgements We greatly thank Prof. Dr. O. AnnEE. Rasa, University of Bonn, for competent criticism on an earlier version as well as linguistic and stylistic advice. We also thank Mrs. ASTRID INGWERSEN, Institut für Haustierkunde, University of Kiel for general technical assistance. Zusammenfassung Postnatale Hirngrößenabnahme, visuelle Leistung, Lernen und Unterscheidungsvermögen von juvenilen und adulten amerikanischen Minken (Mustela vison: Carnivora: Mammalia) Die Abnahme der Hirngröße im Ausmaß von 20% während der Entwicklung vom juvenilen zum adul- ten Stadium beim Mink und anderen Mustela-Arten ist nach wie vor wenig verstanden und in ihrer generellen biologischen und funktionellen Bedeutung ungeklärt. Entsprechendes gilt für die einher- gehende Verkleinerung des Hirncavums, vornehmlich bedingt durch Abflachung des Hirnschädels. Da der Neocortex und andere Hirnteile mit höheren integrativen und assoziativen sensorischen und motorischen Funktionen besonders betroffen sind, erhebt sich die Frage, inwieweit diese Größenänderungen funktionelle Konsequenzen haben; d.h. sind funktionelle Kapazitäten dieser Hirnstrukturen gleichzeitig gemindert? Dieses wurde für das visuelle System bei 8 juvenilen (4,4) und 8 adulten (4,4) Nachkommen von Minken (Mustela vison energumenos) aus freier Wildbahn un- ter Einsatz einer Zweifach-Wahl-Apparatur getestet. Nach Konditionierung und Test auf spontane Seitenstetigkeit mußten die Individuen schwarze Punkte unterschiedlicher Größe von weißen Flä- chen aus 30 cm Entfernung unterscheiden. Insgesamt wurden 16 000 individuelle Datensätze erho- ben und statistisch auf Unterschiede in visueller Leistung, Lerngeschwindigkeit und Unterschei- dungsvermögen analysiert. Es ergaben sich keine Unterschiede in der Lerngeschwindigkeit zwischen juvenilen und adulten Individuen. Allerdings resultierte ein signifikanter Einfluß von Al- ter (juvenile Tiere zeigten bessere Ergebnisse) und Geschlecht (Fähen zeigten bessere Ergebnisse) im visuellen Unterscheidungsvermögen. Diese Ergebnisse werden im Zusammenhang mit der biolo- gischen Bedeutung des Gesichtssinns für diese Tierart, mit generellen Verhaltensunterschieden zwischen juvenilen und adulten Individuen und mit der ontogenetischen Abnahme zentralnervöser Masse diskutiert. Im Hinblick auf diese Untersuchungen gibt es keine Anzeichen für einen funktio- nellen Zusammenhang zwischen der ontogenetischen Hirngrößenabnahme und Leistungsfähigkeit des visuellen Systems. Brain size decrease and visual performance of juvenile and adult Mustela vison References APFELBACH, R.; KRUSKA, D. (1979): Zur postnata- len Hirnentwicklung beim Frettchen Mustela putorius f. furo (Mustelidae; Mammalia). Z. Säugetierkunde 44,127-131. BisHop, Y. M. M.; FIENBERG, S. E.; HALLAND, P.H. (1975): Discrete Multivariate Analysis; Theory and Practice. Cambridge, Mass.: MIT Press. BRAEKEVELT, C.R. (1990): Fine structure of the retinal photoreceptors of the ranch mink Mustela vison. Acta anat. 138, 254-260. Dusin, M. W.; TURNER, L. (1977): Anatomy of the retina of the mink. J. Comp. Neurol. 173, 275- 288. DunsTone, N. (1993): The Mink. London: Poyser Natural History. DUnsTOoNE, N.; SINCLAIR, W. (1978a): Compara- tive aerial and underwater visual acuity of the mink, Mustela vison Schreber, as a function of discrimination distance and stimulus lumi- nance. Anim. Behav. 26, 6-13. DUnSTONE, N.; SINCLAIR, W. (1978b): Orienting behaviour during aerial and underwater visual discrimination by the mink (Mustela vison Schreber). Anim. Behav. 26, 14-21. ESPENKÖTTER, E. (1982): Vergleichende quantita- tive Untersuchungen an Iltissen und Frett- chen. Diss. thesis, Tierärztliche Hochschule Hannover. EwER, R. F. (1973): The Carnivores. London: Wei- denfeld and Nicolson. GELLERMAN, L. W. (1933): Chance orders of alter- nating stimuli in visual discrimination experi- ments. J. Genet. Psychol. 42, 206-208. GOETHE, F. (1964): Das Verhalten der Musteliden. Hdb. Zool. 8, 37. Lief. 1-80. GUILLERY, R. W.; OBERDORFER, M.D. (1977): A study of fine and coarse retino-fugal axons terminating in the geniculate C laminae and in the medial interlaminar nucleus of the mink. J. Comp. Neurol. 176, 515-526. GUILLERY, R. W.; ÜÖBERDORFER, M.D.; MUur- PHY, E.H. (1979): Abnormal retino-geniculate and geniculo-cortical pathways in several ge- netically distinct colour phases of the mink (Mustela vison). J. Comp. Neurol. 185, 623- 656. Kruska, D. (1977): Über die postnatale Hirnent- wicklung beim Farmnerz Mustela vison f. dom. (Mustelidae; Mammalia). Z. Säugetier- kunde 42, 240-255. KruskA, D. (1979): Vergleichende Untersuchun- gen an den Schädeln von subadulten und adulten Farmnerzen, Mustela vison f. dom. 279 (Mustelidae; Carnivora). Z. Säugetierkunde 44, 360-375. KRruska, D. (1987): How fast can total brain size change in mammals? J. Hirnforsch. 28, 59-70. KruskA, D. (1988a): Mammalian domestication and its effect on brain structure and behavior. In: Intelligence and Evolutionary Biology. Ed. by H.J. JErıson and I. JEerıson. Berlin, Hei- delberg: Springer. Pp. 211-250. KruskA, D. (1988b): Marderartige. In: Grzimeks Enzyklopädie Säugetiere Vol.3 Ed. by W. KEIENBURG. Stuttgart: Kindler. Pp. 388- 446. Kruska, D. (1993): Evidence of decrease in brain size in ranch mink, Mustela vison f. dom., dur- ing subadult postnatal ontogenesis. Brain Be- hav. Evol. 41, 303-315. Kruska, D. (1996): The effect of domestication on brain size and composition in the mink (Mus- tela vison). J. Zool. (London) 239, 645-661. KRrUSKA, D.; SCHREIBER, A. (1999): Comparative morphological and biochemical-genetic inves- tigations in wild and ranch mink (Mustela vi- son: Carnivora: Mammalia). Acta Theriol. 44, 377322. LeVay, S.; MCConnteL, S. K.; Luskin, M.B. (1987): Functional organization of primary vi- sual cortex in the mink (Mustela vison), and a comparison with the cat. J., Comp. Neurol. 257, 422-441. RENSCH, B. (1973): Gedächtnis, Begriffsbildung und Planhandlung bei Tieren. Hamburg: Parey. SANDERSON, K. J. (1974): Lamination of the dorsal lateral geniculate nucleus in carnıvores of the weasel (Mustelidae), raccoon (Procyonidae) and fox (Canidae) families. J. Comp. Neurol. 153, 239-266. SANDERSON, K.J; GUILLERY,R. W.; SHAKEL- FORD, R. M. (1975): Congenitally abnormal vi- sual pathways in mink (Mustela vison) with reduced retinal pigment. J. Comp. Neurol. 154, 225-248. SCHMIDT, K. (1992): Skull variability of Mustela ni- valis Linnaeus, 1766 in Poland. Acta Theriol. 37, 141-162. SOoKAL, R. R.; ROHLF F. J. (1995): Biometry. New York: W. H. Freeman. STEFFEN, K. (2000): Vergleichender quantitativer Nachweis sowie topographische Analyse von Ganglienzellen und Zapfen in der Retina von Wildmink (Mustela vison energumenos) und Farmmink (Mustela vison f. dom.) . Diss. the- sis, Universität Kiel 280 KATJA STEFFEN et al. TEMBROCK, G. (1982): Spezielle Verhaltensbiolo- gie der Tiere. Vol. 1. Jena: Fischer Wir, ©. (1982): Bone resorption in the skull of Mustela vison. Acta Theriol. 27, 358-360. Wıc, ©. (1985): Multivariate variation in feral American mink (Mustela vison) from South- ern Norway. J. Zool. (London) 206, 441-452. WÜSTEHUBE, C. (1960): Beiträge zur Kenntnis be- sonders des Spiel- und Beuteverhaltens ein- heimischer Musteliden. Z. Tierpsychol. 17, 579-613. Authors’ address: Dr. KATJA STEFFEN, Prof. Dr. DIETER KRUSKA, Priv.-Doz. Dr. RALPH TIEDEMANN, Institut für Haustierkunde, Universität Kiel, Olshausen- str. 40, D-24118 Kiel, (e-mail: ksteffen@ifh.uni-kiel.de) Mamm. biol. 66 (2001) 281-289 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Original investigation Mammalian Biology a ANZ BEN AS au SM 0 EZ /ERKUND! Zeitschrift für Säugetierkunde Differential predation upon sex and age classes of tuco-tucos (Ctenomys talarum, Rodentia: Octodontidae) by owls By M. J. KITTLEIN, A. I. VASSALLO, and CRISTINA BUSCH Departamento de Biologia, Universidad Nacional de Mar del Plata, Mar del Plata, Argentina Receipt of Ms. 22. 05. 2000 Acceptance of Ms. 05. 04. 2001 Abstract Predation by burrowing (Athene cunicularıa) and short-eared (Asio flammeus) owls upon sexes and age-classes of the subterranean herbivorous rodent Ctenomys talarım were studied by comparing characteristics of field-trapped and preyed-upon individuals. Owls took a greater proportion of males than females in comparison with their respective field densities. This sex-biased pattern of predation was most marked during the breeding season of C. talarum and mainly affected subadult males. A set of ecological features of C. talarum, such as food habits and above ground mobility, that might explain differential vulnerability by sex was analysed and did not support the observed pattern. We suggest that it is determined by higher above ground exposure of subadult males du- ring the breeding season because they interact above ground with adult males and search for settlement sites to establish their own burrows. Key words: Ctenomys talarum, owl predation, prey vulnerability, Argentina Introduction A complex set of behavioural and morpho- logical characteristics of prey and predator species determines the likelihood that an anımal will be captured. Prey activity pat- terns and prey habitat use are behavioural traits that may modify predation risk; while a predator’s attack biomechanics, activity time or habitat use, may likewise cause cer- tain classes of prey to be at higher risk. Non-random predation upon prey species of different size is well known among rap- torial birds (HALLE 1988; MARKS and MARTI 1984; Martı 1974; SCHOENER 1968; SMITH and MurpHYy 1973; STORER 1966; VASSALLO et al. 1994). Even within a single prey tax- 1616-5047/01/66/05-281 $ 15.00/0. on, selection of a given size often occurs, with the young or smaller individuals often preyed upon more frequently than expected (SMITH and MurPpHy 1973; KoTLER 1985; LonGLAND and JENKINS 1987; DONAZAR and CEBALLOS 1989; ZALEWSKI 1996). Predators are thought to be important fac- tors in shaping population traits and habitat use of their mammalian prey (KoTLEr 1984; JaKsıc 1986; PALOMARES and DELIBES 1997; WoLrr 1997; SAITOH et al. 1999). Although evidence that predation limits mammalian abundances is weak, the role of predation as a primary selective force remains undis- puted in population studies. Owls and 282 M. J. Kıttlein etal. hawks are important predators of small mammals (MArTı and HocuE 1979; KoRrPI- MÄKI and NORRDAHL 1989; Jaxsıc et al. 1992; MarTI et al. 1993). Prey selection in these visually oriented predators is con- strained mainly by prey vulnerability, gener- ally an unknown function of prey size (MarTı and HoGuE 1979; ZAMORANO et al. 1984; Bozınovic and MEDEL 1988), prey colour (KAUFMAN 1974a; GÖTMARK et al. 1997), prey activity (KAUFMAn 1974b; LonGLaAnD and Price 1991), and the syner- gistic effects of these factors (HALLE 1993). Generally, the more conspicuous sex (usual- ly males) is more vulnerable to predators in sexually dimorphic prey (SELANDER 1965, 1966; Geist 1971). Moreover, even in appar- ently monomorphic species, body size may exhibit differences in variances between sexes, which can affect vulnerability to pre- dation.. Monomorphic species may also exhibit behavioural differences that make one sex more vulnerable than the other. South American caviomorph rodents of the genus Cienomys (locally known as tuco-tu- cos) are the group with most species of fos- sorial rodents in the world. Of about 125 living species of hypogeic rodents, 55 be- long to this genus (Reıc et al. 1990). Species of Ctenomys are sexually dimorphic, speci- fically male, C. talarım are 33% larger in body mass and 6% in body length than fe- males (ZENUTOo et al. 1999). The aim of this study is to assess the charac- teristics of owl predation upon sexes of C. talarum and to evaluate some behavioural and ecological features of C. talarum that might explain differential vulnerability by sex and age. Material and methods Four censuses of tuco-tucos, Ctenomys talarum, were conducted every three months starting No- vember 1987 and throughout 1988 at Necochea (38°36' S, 58°48' W), Buenos Aires Province, Ar- gentina. Each census took place in a 1.5 ha plot for five consecutive days. Each plot was staked in a grid pattern to establish the spatial location of captured animals. Animals were trapped without injury using plastic livetraps, which were placed at the entrances of all burrows and repeatedly checked and re-set until nearly all animals present within the grid were caught. Removal of indivi- duals from census grids should have had little ef- fect on prey availability, as total hunting area used by owls would have been much larger than the grids. The censuses provided information on: 1) popula- tion density, 2) reproductive status, 3) body weight distribution, 4) the ratio adults: subadults, and 5) sex ratio. Fifty-nine C. falarum were cap- tured, and each individual was killed by ether in- halatıion and autopsied to obtain information on reproductive condition, pregnancy, and relative age. Females were classified as immature (narrow, pale uterine horns and closed vagina) or mature (thick uterine horns, and open or plugged vagina). Males were also classified as immature (lack of spermatozoa in epididymis) or mature (spermato- zoa in epididymis). C. talarum attains sexual ma- turity at an average weight of 95 g. Pregnant fe- males occurred only from July to March (Mariızıa et al. 1991). This well-defined reproduc- tive period in this species allowed comparisons between preyed-upon and trapped individuals during both the breeding and non-breeding sea- Sons. Fresh pellets from burrowing owl, Athene cunicu- laria (n = 149), and short-eared owl, Asio flam- meus (n= 124) were collected simultaneously with our censuses. Total area for pellet collection was 10 km’, which included our census grids. Pel- let collection took place at known roosts, perches, or nests. Because sampling took place along con- sistent routes, sampling intensity did not vary among periods. Thus, number of preyed-upon in- dividuals found in pellets should reflect predation intensity through time. Ctenomys talarım prey remains were catalogued and then identified to species level by reference to dissected and cleaned skeletal elements of lo- cally collected voucher specimens. The minimum number of single or double anatomical elements such as skulls, mandibles, or tooth rows estimated the minimum number of individuals in pellets. Although skulls of C. falarum in owl pellets were frequently found to be partially crushed, a set of skull variables could be measured. Length of maxillary toothrow, rostral width, length of dia- stema, mandibular length, cranial width, and basi- lar length were measured with a hand vernier cal- liper (precision 1/20 mm) and applied in the estimation of body mass and to determine sex of preyed-upon individuals. Estimation of body mass was accomplished through simple linear regres- sion equations. Most regression equations be- Predation upon (tenomys talarum by owls tween skull variables and body mass provided adequate models, thus allowing estimation of body mass with reasonable accuracy. Masses of prey were estimated using that variable which had the smallest sum of prediction error. Deter- mination of sex of prey was undertaken through linear discriminant analysis (CooLEey and LOHNES 1971). Skull measurements were entered as vari- ables for investigating within- and between group variability, testing differences in composition of groups, and finally assigning unknown prey to sex for many preyed-upon individuals. Linear dis- criminant as well as regression equations afforded reliable means for estimating sex and body mass of prey individuals (Tab. 1). Additional information from a neighbour popula- tion of C. talarım at Mar del Cobo (37°52' S, 57°23’ W) was analysed in order to establish pre- sumable correlates with differential sex preda- tion. Data from a capture-recapture study (BUSCH et al. 1989) were used to obtain information on capture frequency distributions and above ground mobility possibly related to dispersal. A second study using microhistological techniques (DEL VALLE et al. 2001) provided information on C. fa- larum’s diet characteristics. The proportion of aer- ial plant parts in the diet and the similarity of diet composition of an individual to that of the vegeta- tion sourrounding each individual’s burrow en- trance were compared in both sexes. Diet-vegetation similarity was expressed using the Morisita’s index (LupwIG and REYnoLDs 1988). Kolmogorov-Smirnov tests were used to compare the distribution of capture frequencies between sexes. Nonparametrice Mann-Whitney U tests (ZAR 1984) were used to compare the proportion of aerial plant parts in the diet, distances moved between successive captures, and diet-vegetation 283 similarity indexes between sexes. Mean prey mass distributions of trapped and preyed-upon indivi- duals were analysed with one-way ANOVA with subsequent Student-Newman-Keuls multiple comparison of means tests. Intraspecific depar- tures from random predation were tested by cal- culating an expected number of individuals of each sex and age category. This was done by mul- tiplying the proportion of each sex or age cate- gory in field censuses by the total number of indi- viduals found in pellets. Differences between observed and expected numbers of prey in pellets were tested using Chi-Square tests (SOKAL and RoHLr 1981). Alpha was initially set to 0.05 in all statistical analyses and adjusted using sequential Bonferroni corrections for the total number of comparisons in each analysis (RıcE 1989). Results Linear discriminant analysis based on skull morphometrics produced significant separa- tion between known males and females of Ctenomys talarım (Wilks’ Lamb- dar 1097-2 19 53rd. 7 3277230: 001): We assıgned skulls of C. talarım from owl pellets to male or female categories if the probability of correct sexual classification was 20.90. Of 112 C. talarum individuals found in pellets 39 skulls were intact and thus provided measurements in those vari- ables necessary for the discriminant analy- sıs. Regressions of body mass on skeletal measurements provided reliable estimates of body mass, allowıng the assıgnment of Table 1. Cranial variables and statistics of linear regressions and discriminant analysis used to estimate body mass and sex, respectively, of preyed-upon individuals of Ctenomys talarum. BL basilar length; CW cranial width; LD length of diastema; LMT length of maxillary toothrow; ML mandibular length; RW rostral width. Number of ob- servations = 39. Body mass Constant Coefficient Constant LD LMT BL RW CW % CORRECT 284 M.J. Kırtıein et al. preyed upon individuals to adult and suba- dult age categories. Of the 39 C. talarum found in pellets we assigned 26 to males (9 adult and 17 subadult), 9 to females (7 adults and 2 subadult), and the remaining 4 were not classified (Fig. 1). Owls showed contrasting consumption of €. talarım. A. cunicularia took most tuco-tu- cos during the breeding season, while A. flammeus preyed most heavily during the non-breeding season. Differences in sex ratios, age category, and mean body masses of C. talarum were ap- parent between trapped and preyed-upon individuals during both breeding and non- breeding seasons (Tab. 2). Male C. talarum were preyed upon significantly more often than expected based on sex ratio of trapped individuals by A. cunicularia during the breeding season (y” = 5.43, P = 0.02). During the non-breeding season no individuals preyed upon by A. cunicularia were re- corded. Sex ratio of C. falarum preyed upon by A. flammeus did not differ with respect to that expected based on trapped indivi- duals either during this period or during the TRAPPED PREYED non-breeding season (x =0.73, P=0.39 and x = 1.09, P = 0.29, respectively). Males preyed-upon by A. cunicularia dur- ing the breeding season were significantly smaller than those field-trapped (SNK test, P = 0.034). The same was observed for males preyed upon by A. flammeus during the non-breeding season (SNK _ test, B=0.072): On the other hand, body mass of females preyed upon by A. cunicularia during the breeding season and by A. flammeus during the non-breeding season did not differ from that of field-trapped females (SNK test, P = 0.290 and P = 0.878, respectively). That males preyed upon by owls were smal- ler than those field-trapped, concurs with a biased adult:subadult ratio as compared to the ratio observed in field-trapped males. Hence, subadult males were significantly overconsumed by A. cunicularia (4° = 30.68, P<0.001) and A. flammeus (x° = 53.78, P< 0.001) during the breeding and non- breeding seasons, respectively. In both peri- ods, the adult:subadult ratio of preyed- upon females did not show significant de- : o.eo 8 0°8 30 0 o FEMALES ® MALES © UNKNOWN ? 2? u) I EL? N Ss ARE 2 0 1 2 BAD DISCRIMINANT SCORE Fig. 1. Sex determination of Ctenomys talarım eaten by owls. Discriminant scores are given for individual cranio- metric measurements from known (trapped) males and females. Individuals of unknown sex in owl pellets had a probability of correct sexual classification > 0.90. Predation upon (tenomys talarum by owls partures from that observed for field- trapped individuals (y’ = 1.34, P= 0.25 and X = 0.07, P = 0.78, respectively). The observed proportion of male C. talar- um in both owls’ pellets for the breeding season was significantly higher than that for the non-breeding season as compared to expected proportions based on field den- sities and sex ratios in both periods (X = 10.05, P = 0.001). On the contrary, the proportion of females in owls’ pellets did not differ between periods (y’=1.41, B2023): A set of attributes that might indicate dif- ferential exposure to owl predation, possi- bly accounting for higher predation upon one sex was analysed. For this purpose we compared data on capture frequencies, dis- tance moved between successive trappings and diet composition between sexes. Comparisons between distributions of cap- ture frequencies (number of individuals that were caught 1, 2, or n times) may de- note differences in site fidelity between sexes, but we did not find significant differ- ences in this regard (males 1.4+0.7, fe- males 1.6 # 1.2; Kolmogorov-Smirnov test, DE02E0.92).: 285 Distances moved between successive cap- tures larger than the average length of the burrow system (14m; ANTINUCHI and Busch 1992) were considered as above ground mobility related to dispersal, taking in account that in this species dispersal oc- curs above ground. Average distance for males (mean +1 SD, 45.36 +# 34.31 m) did not differ from that for females (43.85 +14.55m; Mann-Whitney _ test, UT 02) She Proportionzorsthese transient individuals in the population did not differ between sexes (males = 0.72, fe- males = 0.68: y° = 0.05, P = 0.83). Two attrıbutes of the diet of C. talarum were presumed to be related to above ground ex- posure: the proportion of above ground plant parts in the diet, and the similarity between the botanical composition of the diet and that of the vegetation surrounding each indi- vıdual burrow entrances. A higher propor- tion of above ground plant parts in the diet, and a lower sımilarıty between diet and sur- rounding vegetation for one sex would indi- cate that this sex is more exposed to owl pre- dation. Neither the proportion of above ground plant parts (males = 86.25 + 12.78, fe- males = 81.70 # 17.61) nor diet-vegetation Table 2. Number of male and female individuals trapped in field censuses, body mass (g; mean +# 1SD), and ob- served (expected in parentheses) number of each sex and age category of Ctenomys talarum preyed-upon by owls. Source of specimens and parameter Breeding season Trapped Total 18 23 Adults 16 16 Subadults 2 7 Body mass 129.2 + 31.0 Asıo flammeus No.preyed No. of adults No. of subadults Body mass Athene cunicularia No. preyed No. of adults No. of subadults Body mass 8OSE=31746 Non-breeding season Females 103292: 20.5 0 (1.1) 13. 0:0=51720 286 M.J. KııtLein et al. similarity (males=0.36+0.18, females — 0:36 2.0.2) zediffered 7 /betiwyeengesexes (Mann-Whitney test, U = 337.5, P= 0.29; U985P = 0.9731espectively): Discussion Although tuco-tucos are subterranean ro- dents, they occur above ground when searching for food in the vicinity of bur- rows, as suggested by the high proportion of aerial plant tissues found in their diet (DEL VALLE et al. 2001; CoMPARATORE et al. 1995) and when dispersing (Mauızıa 1994). Unexpected high levels of above ground mobility have been noted in C. talarum (BuschH et al. 1989) as well as in C. australis (E. OUDSHOORN, pers. comm.). These above ground activities differ from what was re- ported for other subterranean rodent spe- cies of Spalacidae and Bathyergidae, where surface exposure is considered incidental (HETH 1991, Jarvıs and BENNETT 1991). The regular above ground activity of tuco-tucos would suggest that predation, principally by visually oriented raptors, is more com- mon than previously assumed in Ctenomys (VassAaLLo et al. 1994; Busch et al. 2000). Our study shows that subadult male Cie- nomys talarum underwent higher predation risk by owls than females during the breed- ing season. As a probable correlate of higher predation risk for males we compared mea- surements related to territory fidelity and above ground mobility between sexes and did not find significant differences. A recent experimental study concerning demographic and reproductive attributes of dispersers in C. talarum, at the study site of Necochea, showed a 1:1 sex ratio of dispersers, transi- ents, and residents. In addition, dispersers did not differ from residents in age composi- tion or body mass (Marızıa et al. 1995). Thus, dispersal does not appear to be the cause of the observed predation pattern. If males and females of a prey species are equally vulnerable to predation, they should occur in owl diets in proportions ap- proximating the sex ratio in the local envir- onment. What characteristics of male C. ta- larum (particularly subadult males) might account for their higher vulnerabilities? The first possibility is that subadults simply lack the experience or sensory skills neces- sary to avoid owl predation. Young indivi- duals of some rodent species became prey soon after leaving maternal care due to inex- perience with their new environment (Lay 1974). Young subterranean rodents leave maternal care rather late and experience the environment outside their natal nests within their mother’s burrow. Size of suba- dult males preyed upon by owls indicate that these individuals had developed sufficient experience with their environment to be al- most equally vulnerable as adult males. It is noteworthy that sizes of preyed upon males (around 85 g) clearly depart from those of natal dispersers, which are approximately 3 months old and weighed around 608g (MaArızıa and Busch 1991; Malizia et al. 1991). Lower body masses of individuals captured at their exclusive burrow were around 60 8, indicating that burrow settle- ment occurs shortly after dispersal from their natal burrow. This suggests that preyed upon males had already established their own burrows when captured by owls. Based on the examined information we do not have evidence to conclude that males are more vagile above ground than females. However, the possibility exists that without being more vagile, males (and particularly subadult individuals during the breeding season) stay longer above ground than fe- males. We ask whether differential preda- tion upon subadult males is due to coinci- dence of either the surface activity patterns and microhabitats used by subadult male tuco-tucos, and whether adult tuco-tucos increase the predation risk of juveniles by forcing them to stay longer above ground in more open areas. We have no evidence concerning differences in microhabitat characteristics, 1.e. differ- ences in plant cover which eventually deter- mine unequal exposure to visual raptors; between adult and subadult males (CoMPAR- ATORE 1990; COMPARATORE et al. 1992). How- ever, data from semi natural enclosures con- cerning social and reproductive behaviour Predation upon (tenomys talarum by owls of C. talarım show that adult males turn from tolerance towards young individuals to higher levels of aggression toward grown-up males (ZENUTO 1999). This obser- vation fits with the high occurrence of scars in field-trapped males (Busc# et al. 1989). When searching for territories — and/or a place within the social hierarchy - a consid- erable number of young males should be in- volved in aggressive interaction, which probably expose them to visually guided raptors. As opposed to females, young males near adulthood are compelled to interact and gain access to mates. Different sources of evidence (ZENUTO et al. 1999) indicate that C. talarum has a poly- gynyc mating system in which male to male competition usually takes the form of ag- gressive interaction leading to dominance ranks (ZENUTO et al. 2001). Taking in ac- count that, compared to other subterranean rodent species, C. talarım makes extensive use of the above ground habitat (Busch et al. 2000), it is conceivable that a substantial portion of social interactions — including but not restricted to inter male aggression - is performed out of their burrows. Tuco- tucos inhabit exclusive burrow systems (CoNTRERAS and REIG 1965; ANTINUCHI Zusammenfassung 287 and BuscH 1992); only during the breeding season one of the sexes must leave its bur- row and enter the other sex’s burrow for mating. If males visit female burrows during the breeding season travelling above ground it is expected that males would suf- fer higher predation risks during this peri- od. However, it is expected that adult males would suffer increased predation under this assumption. Some interaction between both male categories might explain higher expo- sure of subadult males above ground. We suggest that both the higher level of intoler- ance from established adult males, and the active search of a place within the social hierarchy expose subadult males near ma- turity to higher vulnerability to owls. Acknowledgements This research was supported by a grant from the International Foundation for Science, Sweden (C/1706-1), and from the Consejo Nacional de In- vestigaciones Cientificas y Tecnicas (CONICET), PID) N° 3-1049006/88. We wish to express our gratitude to our colleagues at Laboratorio de Ecofisiologia for their constant interest and en- couragement. Alters- und geschlechtsbedingte Unterschiede bei der Prädation von Tuco-Tucos (Ctenomys talarım, Rodentia: Octodontidae) durch Eulen Die Prädation von Kaninchenkauz (Athene cunicularia) und Sumpfohreule (Asro flammeus) auf die Geschlechter und Altersklassen des unterirdisch lebenden, herbivoren Nagers Tuco-Tuco (Ctenomys talarım) wurde untersucht, indem Freilandfänge und Beutespektrum miteinander verglichen wur- den. Die Eulen erbeuteten einen höheren Anteil von Männchen als nach den Dichten im Freiland zu erwarten war. Die Bevorzugung war während der Fortpflanzungsperiode von C. talarum besonders ausgeprägt und betraf insbesondere die Gruppe der subadulten Männchen. Verschiedene ökolo- gische Merkmale von C. talarım wie Nahrungsspektrum und oberirdische Aktivität, die unterschie- dliche Gefährdung der beiden Geschlechter erklären könnten, wurden analysiert, lieferten aber keine Erklärung für die Beobachtung. Wir vermuten daher, dass die vermehrte Prädation subadulter Män- nchen während der Fortpflanzungsperiode Folge einer vermehrten oberirdischen Aktivität ist, die durch oberirdische Interaktionen mit adulten Männchen und die oberirdische Suche nach freien Sie- dlungsplätzen für die Anlage eigener Baue bedingt wird. 288 M. J. Kıttein et al. References ANTINUCHT, C. D.; BuscH, C. (1992): Burrow struc- ture in the subterranean rodent Ctenomys ta- larum. Z. Säugetierkunde 57, 163-168. BozınovIic F.; MEDEL, R. G. (1988): Body size, en- ergetic and foraging mode of raptors in cen- tral Chile: an inference. Oecologia 75, 456- 458. BuschH, C.; MALIZIA, A. I; SCAGLIA, ©. A.; REIG, O. A. (1989): Spatial distribution and attributes of a population Crenomys talarım (Rodentia, Oc- todontidae). J. Mammalogy 70, 204-208. BuschH, C.; AnTINUcHt, C. D.; DEL VALLE, J. C.; KITT- LEIN, M. J.; MALIZIA, A. ].; VASSALLO, A.1.; ZE- NUTO, R. R. (2000): Population ecology of sub- terranean rodents. In: Life Underground: The Biology of Subterranean Rodents. Ed. by J. L. PATTon, E.LAcey, and G.N. CAMERON. Chicago: Chicago University Press. Pp. 183- 226. COMPARATORE, V.M. (1990): Selecciön de häbitat por el tuco-tuco. Diss. thesis. Universidad Na- cional de Mar del Plata, Mar del Plata, Ar- gentina. COMPARATORE, V. M.; AGNUDSDEI, M.; BUScH, C. (1992): Habitat relations in sympatric popula- tions of Ctrenomys australis and Ctenomys ta- larım (Rodentia, Octodontidae) in a natural grassland. Z. Säugetierkunde 57, 47-55. COMPARATORE, V. M.; CiıD, M. S.; BuscH, C. (1995): Dietary preferences of two sympatric subter- ranean rodent populations in Argentina. Rev. Chil. Hist. Nat. 68, 197-206. CONTRERAS, J. R.; REIG, O. A. (1965): Datos sobre la distribuciön del gro. Ctenomys en la zona costera de la pcia. de Buenos Aires compren- dida entre Necochea y Bahia Blanca. Physis (Argentina) 25, 169-186. CooLE&y, W. W.; LoHNES,R.R. (1971): Multivari- ate Data Analysis. New York: John Wiley and Sons. DONAZAR, J. A.; CEBALLOS, ©. (1989): Selective predation by Eagle Owls Bubo bubo on rab- bits Oryctolagus cunniculus: age and sex pre- ferences. Ornis Scandinavica 20, 117-122. GEIST, V. (1971): Mountain Sheep: a Study in Be- havior and Evolution. Chicago: University of Chicago Press. GÖTMARK, F.; POST, P.; OLsson, J.; HIMMELMANN, D. (1997): Natural selection and sexual dimorph- ism: sex-biased sparrowhawk predation fa- vours crypsis in female chaffinches. Oikos 80, 540-548. HALLE, S. (1988): Avian predation upon a mixed community of common voles (Microtus arva- lis) and wood mice (Apodemus sylvaticus). Oecologia 75, 451-455. — (1993): Diel pattern of predation risk in micro- tine rodents. Oikos 68, 510-518. HETH, G. (1991): Evidence of above-ground pre- dation and age determination of the prey in subterranean mole rats (Spalax ehrenbergi) in Israel. Mammalia 55, 529-542. JAKSIC, F. M. (1986): Predation upon small mam- mals in shrublands and grasslands of South America: ecological correlates and presum- able consequences. Rev. Chil. Hist. Nat. 59, 209-221. JAKSIC, F. M.; JIMENEZ, J. E.; CASTRO, S. A.; FEISIN- GER, P. (1992): Numerical and functional re- sponse of predators to a long-term decline in mammalian prey at a semi-arid Neotropical site. Oecologia 89, 90-101. JARVIS, J. U. M.; BENNETT,N. C. (1991): Ecology and behavior of the family Bathyergidae. In: The Biology of the Naked Mole-Rat. Ed. by P. W. SHERMAN, J.U.M.Jarvıs, and R.D. ALEXANDER. Princeton, N.J.: Princeton Uni- versity Press. Pp. 66-90. KAUFMAN, D. W. (1974 a): Adaptive coloration in Peromyscus polionotus: experimental selec- tion by owls. J. Mammalogy 55, 271-283. KAUFMANn, D. W. (1974b): Differential predation on active and inactive prey by owls. Auk 91, 172-173. KoRPIMAKI, E.; NORRDAHL, K. (1989): Predation of Tengmalm’s owls: numerical responses and dampening impact on population fluctuations of microtines. Oikos 54, 154-164. KOTLER, B. P. (1984): Risk of predation on the structure of desert rodent communities. Ecol- ogy 65, 689-701. KoTLeEr B. P. (1985): Owl predation on desert ro- dents which differ in morphology and beha- vior. J. Mammalogy 66, 824-828. Lay, D.M. (1974): Differential predation on ger- bils (Meriones) by the little owl Athene brah- ma. J. Mammalogy 55, 608-614. LONGLAND, W. S.; JENKINS, S.H. (1987): Sex and age affect vulnerability of desert rodents to owl predation. J. Mammalogy 68, 746-754. LONGLAND, W. S.; Price, M. V. (1991): Direct ob- servations of owls and heteromyid rodents: can predation risk explain microhabitat use? Ecology 72, 2261-2273. Luowig, J. A.; REYNoLDSs, J.F. (1988): Statistical Ecology. A Primer on Methods and Comput- ing. New York: John Wiley and Sons. MAarızIA, A. I. (1994): Ecologia poblacional de Ctenomys talarum. Diss. thesis, Universidad de Nacional de Mar del Plata, Argentina. Predation upon (tenomys talarum by owls MALIZIA, A. 1.; Busch, C. (1991): Reproductive parameters and growth in the fossorial rodent Ctenomys talarım (Rodentia: Octodontidae). Mammalia 55, 293-305. MALIZIA, A. 1.; VASSALLO, A. 1.; BuscH, C. (1991): Population and habitat characteris- tics of two sympatric species of Ctenomys (Rodentia, Octodontidae). Acta Theriol. 36, 87-94. MALIZIA, A. 1.; ZENUTO, R. R.; Busch, C. (1995): Demographic and reproductive attributes of dispersers in two populations of the subterra- nean rodent Ctenomys talarım (tuco tuco). Can. J. Zool. 73, 732-738. MARKS, J. S.; MARTI, C. D. (1984): Feeding ecology of sympatric Barn Owls and Long-eared Owls in Idaho. Ornis Scandinavica 15, 135-143. MARTI, C. D. (1974): Feeding ecology of four sym- patric Owls. Condor 76, 45-61. MARTI, C. D.; HoGue, J. G. (1979): Selection of prey by size in Screech Owls. Auk 96, 319-327. MARTI, C.D.; STEENHOF,K.; KocHERT, M.N.; MARKS, J.S. (1993): Community trophic structure: the roles of diet, body size, and ac- tivity time in vertebrate predators. Oikos 67, 6-18. PALOMARES, F.;, DELIBES, M. (1997): Predation upon european rabbits and their use of open and closed patches in Mediterranean habitats. Oikos 80, 407-410. ReıG, ©. A.; BuscH, C.; COoNTRERAS, J. R.; OR- TELLS, M. (1990): An overview of evolution, systematics, population biology, cytogenetics, molecular biology, and speciation in Cre- nomys. In: Evolution of Subterranean Mam- mals at the Organismal and Molecular Levels. Ed. by E.Nevo and O. A. Reıc. New York: Allan Liss. Pp. 71-96. Rice, W. R. (1989): Analyzing tables of statistical tests. Evolution 43, 223-225. SAITOH, T.; BI@RNSTAD, O.N.; STENSETH,N.C. (1999): Density dependence in voles and mice: a comparative study. Ecology 80, 638-650. SCHOENER, T. W. (1968): Sizes of feeding terri- tories among birds. Ecology 49, 123-141. SELANDER, R.K. (1965): On mating systems and sexual selection. Am. Nat. 99, 129-141. SELANDER, R.K. (1966): Sexual dimorphism and differential niche utilization in birds. Condor 68, 113-151. SMITH, D. G.; MURPHY, J. R. (1973): Breeding ecol- ogy of raptors in the eastern Great Basin of Utah. Brigham Young Univ. Sc. Bull., Biol. Series 18, 1-76. 289 SOKAL, R. R.; RoHLF, F. J. (1981): Biometry: the Principles and Practice of Statistics in Biologi- cal Research. (2nd ed.). San Francisco: W. H. Freeman and Co. STORER, R. W. (1966): Sexual dimorphism and food habits in three North American accipi- ters. Auk 83, 423-436. VALLE, J. C. DEL; LOHFELDT, M. I.; COMPARATORE, V. M.; Cıp, M. S.; BuscH, C. (2001): Feeding selectivity and food preference of Ctenomys talarum (tuco-tuco). Mamm. biol. (in press). VASSALLO, A. 1.; KITTLEIN, M. J.; Busch, C. (1994): Owl predation on two sympatric species of tuco-tucos (Rodentia, Octodontidae). J. Mammalogy 75, 725-732. WOLFF, J.O. (1997): Population regulation in mammals: an evolutionary perspective. ]. Anim. Ecol. 66, 1-13. ZALEWSKI, A. (1996): Choice of age classes of bank voles Clethrionomys glareolus by pine marten Martes martes and tawny owl Strix aluco in Biatowieza National Park. Acta Eco- logica 17, 233-244. ZAMORANO, E.; PALOMO, L.; ANTUNEZ, A.; VAR- GAS, J. (1984): Criterios de depredaciön selec- tiva de Bubo bubo y Tyto alba sobre Rattus. Ardeola 33, 3-9. ZAR,J.H. (1984): Biostatistical Analysis. (2nd ed.). Englewood Cliffs, N. J.: Prentice Hall. ZENUTOR.R. (1999): Sistema dc apareamiento en Ctenomys talarım (Rodentia: Octodonti- dae). Diss. thesis, Universidad Nacional de Mar del Plata, Mar del Plata, Argentina. ZENUTO, R.R.; Lacey, E. A.; Busch, CE. (1999): DNA fingerprinting reveals polygyny in the subterranean rodent Cienomys talarum. Mol. Ecol. 8, 1529-1532. ZENUTO, R. R.; MALIZIA, A.1.; Busch, C. (1999): Sexual size dimorphism, testes size and mat- ing system in two populations of Ctenomys ta- larım (Rodentia: Octodontidae). J. Nat. Hist. 33, 305-314. ZENUTO, R. R.; VASSALLO, A. 1.; BuscH, C. (2001): A method for studying social and reproduc- tive behavior of subterranean rodents in cap- tivity, wıth preliminary data from Ctenomys talarum. Acta Theriol. (in press). Authors’ address: MARCELO J. KITTLEIN, ALDO I. VASSALLO, and CRIS- TINA BuscH, Laboratorio de Ecofisiologia, Depar- tamento de Biologia, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de Mar del Plata, C.C. 1245 (7600) Mar del Plata, Argentina. Mamm. biol. 66 (2001) 290-300 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol , Mammalian Biology Zeitschrift für Säugetierkunde Original investigation Kleinsäuger auf forstwirtschaftlich unterschiedlich behandelten Windwurfflächen eines Bergwaldes Von IRENE GLITZNER und H. GOSSOW Institut für Wildbiologie und Jagdwirtschaft, Universität für Bodenkultur, Wien Receipt of Ms. 01. 12. 1999 Acceptance of Ms. 18. 04. 2001 Abstract Small mammal distribution on differently managed storm areas of a mountain forest In 1995, two uprooted forest stands and an old growth area were investigated regarding small mammal distribution. Each stand was 10 ha in size and all were located at the borderline of Styria and Lower Austria. The age of the forest stands at all three study sites was approximately 200 years old. One study area, the windfallen stand was cleared from thrown and broken trees after the dis- turbance, the other one was left untreated. The aim of the study was to document small mammal communities in a mountain forest area in the Northern Limestone Alps and also, to investigate if habitat preferences of yellow-necked mouse (Apodemus flavicollis) and bank vole (Clethrionomys glareolus) were similar in mountain forests compared to previously studied areas at lower eleva- tions. The largest number of Muridae were captured on the untreated storm area. Trapping data showed that the yellow-necked mouse (Apodemus flavicollis) was dominant on all three study sites followed by the bank vole (Clethrionomys glareolus), wood mouse (Apodemus sylvaticus), and field vole (Microtus agrestis). Yellow-necked mice were most often caught at the untreated site and in general were captured in young as well as grass and herb stands. The number of male yellow-necked mice captured was significantly larger on the three study sites as a whole. But taking the respective study sites selectively, the number of males was significantly larger only in the old growth stand. Trapping success suggests that seasonal habitat change occurs with the yellow-necked mouse. Trapping success for the bank vole was highest on the cleared plot and closely associated with young stands, old stands, and forest edge. Trapping success on the three study sites as a whole corresponded with habitat preferences as described in the literature. Key words: Apodemus flavicollis, Clethrionomys glareolus, mountain forest, windfallen stand, forest managment Einleitung Ausgelöst durch ein starkes Sturmtief über Kleinsäuger, speziell Mäuse, reagieren ver- Mitteleuropa, mit Windgeschwindigkeiten stärkt auf bodennahe Standortveränderun- bis zu 160 km/h, wurden im März 1990 in gen, wie etwa plötzlich auftretende struktu- Österreich große Waldbestände entwurzelt. relle Veränderungen durch Totholz und 1616-5047/01/66/05-290 $ 15.00/0. daraus resultierende Deckungs- und Klima- gegebenheiten. Aber sie sind auch in der Lage, sich schnell anzupassen und als Pio- niere neu entstandene Lebensräume rasch zu besiedeln. Untersuchungen zu Kleinsäu- gern und ihrer Habitatwahl in Tieflagen sind zahlreich (z. B. FLOWERDEW 1985), über montane und subalpine Bereiche hingegen sind eher nur spärlich Daten vorhanden. Ziel der Untersuchung war es, Daten über die Kleinsäugerbesiedlung der Bergwald- stufe (im Bereich des nordöstlichen Ausläu- fers der Nördlichen Kalkalpen) zu erhalten. Ebenso war zu untersuchen, ob die aus der Literatur bekannten unterschiedlichen Ha- bitatansprüche von Gelbhalsmäusen (Apo- demus flavicollis) und Rötelmäusen (Cle- thrionomys glareolus) in tieferen Regionen auch auf Sonderstandorte, wie jene forst- wirtschaftlich unterschiedlich behandelten Windwurfflächen (naturnahe und intensive Bewirtschaftung) des montanen Waldes, übertragbar sind. Material und Methoden Untersuchungsgebiet Das Untersuchungsgebiet der Windwurf- und Sukzessionsforschung liegt in den Steirisch-Nie- derösterreichischen Kalkalpen auf rund 1 100 m Seehöhe (derzeitiger Status: JUCN Kategorie Ia „Wildnisgebiet Dürrenstein“, Natura 2000 — Vorschlag) und wird von hohen Bergmassiven großräumig begrenzt. Klimatisch geprägt ist das Gebiet durch ein ozeanisch-subozeanisch über- lagertes Alpenrandklima mit hohen Jahresnie- derschlägen, relativ milden Lufttemperaturen, aber schneereichen Wintern. Die Untersuchun- gen der hier vorliegenden Arbeit konzentrierten sich auf drei je 10 Hektar große und etwa 30 m voneinander entfernt liegende Flächen mit ei- nem Bestandesalter von rund 200 Jahren: Die sogenannte „Edelwiesfläche“ wurde nach dem Windwurf nicht in forstwirtschaftlich üblicher Weise aufgeräumt, sondern unbehandelt belas- sen (U = ungeräumt). Nördlich davon liegt, durch eine Forststraße getrennt, die geräumte Fläche (G = geräumt): das hier angefallene Sturmholz wurde bis auf Wurzelteller und „Frat- ten“ (in Bahnen aufgeschichtetes Astmaterial) beseitigt. Nordwestlich von U befindet sich der Altbestandbereich (A), der in seinem Bestan- Kleinsäuger auf Windwurfflächen 291 desalter dem der Windwurfflächen U und G entspricht, aber nicht vom Sturmereignis be- troffen war. Er diente in der Untersuchung als Kontrollfläche. Kleinsäugerfang Die Kleinsäuger wurden während der Vegeta- tionsperiode vom 18. Juni bis 24. September 1995 mit Lebendfallen vom Fallentyp „Holzkasten- falle“ (Wippbrett-Falltüren-Prinzip; Fa. Deu Fa, Neuburg/Inn) gefangen und mit einem Gemisch aus Erdnußbutter, Haferflocken, Apfelstückchen und Pflanzenöl geködert (vgl. RAppA 1968). Holzwolle und ein über die Falle gestülpter leerer Saftkarton schützten die Tiere vor Kälte und Nässe. Die gefangenen Tiere wurden zur Bestim- mung direkt von der Falle in einen Tiefkühlbeutel entlassen und mit einigen Tropfen Äther betäubt. An den Tieren wurden Art, Geschlecht, Gewicht und Reproduktionszustand erhoben. Eine Mar- kierung war zuerst nicht vorgesehen, da sie für die Fragestellung nicht relevant war (Populations- schätzungen wurden nicht durchgeführt). Ab der 3. Fangperiode wurde angesichts der hohen Fang- ergebnisse der Fangperioden 1 und 2 und dem In- teresse am Umfang der Wiederholungsfänge mit einer Markierung begonnen. Die Tiere wurden mit einem wasserfesten Farbstoff (Xanthin, gelöst in 7/0%igem Alkohol) auf der Ventralseite durch einen Farbpunkt als „gefangen“ markiert. Die Fänge wurden allerdings individuell nicht unter- schieden. Bei jedem Wiederfang wurde der Ven- tralseite dann ein weiterer Farbpunkt hinzugefügt (Markierung im Uhrzeigersinn im Bereich des rechten Vorderfußes beginnend, 4. Punkt: Be- reich des rechten Hinterfußes, 5. Punkt: Bauch- mitte), wodurch die Anzahl der Wiederholungs- fänge pro Individuum registriert werden konnte. Im weiteren erwähnte Wiederfänge beziehen sich immer nur auf die Fangperioden 3-6. Die insge- samt 120 Fallen wurden auf den drei Untersu- chungsflächen zu gleichen Teilen verteilt. Die Fallen wurden, abgesehen von wetterbeding- ten Verschiebungen (Fangperiode 4 auf U entfal- len), im Intervall von zwei Wochen in jeweils dreitägigen Fangperioden fängig gestellt und ei- nen Tag (durchschnittlich 24 Stunden) danach kontrolliert. Insgesamt entsprachen die Fallen der sechs hier berücksichtigten Fangperioden (10 Fangnächte) 1160 Fallennächten (FP1 „Ende Juni“: 18.-20. Juni, FP2 „Mitte Juli“: 14.-10. Juli, FP3 „Ende Juli“: 31. Juli-02. August, FP4 „Mitte August“: 15.-16. August, FP5 „Ende August“: 31. August-0l. September, FP6 „Ende Septem- ber“: 22.-24. September). 292 I. GLITZNER et al. Witterungsbedingt notwendige Fangzahlenkorrektur Für die wetterbedingt entfallene Fangperiode Mitte August (FP4) auf der Fläche U wurden die Fangzahlen für Gelbhals- und Rötelmaus interpo- lierend korrigiert, um dennoch einen direkten Vergleich der Fangzahlen zwischen den verschie- denen Untersuchungsflächen zu ermöglichen: als Bezugsfläche zur Korrektur wurde wegen der Ähnlichkeit der Fangerfolge G herangezogen. Die Differenz der Gesamtfangerfolge (ohne Fang- periode Mitte August, FP4) der einzelnen Arten zwischen U und G wurde berechnet und durch die Anzahl der berücksichtigten Fangnächte (9) dividiert. Der so für jede Kleinsäugerart erhaltene Korrekturfaktor K wurde zum jeweiligen Fanger- folg der Fangperiode 4 auf G hinzugezählt bzw. abgezogen und ergab den theoretischen Fanger- folg der ungeräumten Windwurffläche U für die Fangperiode Mitte August (FP4; K = Differenz der Gesamtfangerfolge zwischen U und G/Anzahl der berücksichtigten Fangnächte). Fallenverteilung Aufgrund extremer Verhältnisse auf den unge- räumten Windwurfflächen konnte eine rasterarti- ge Fallenverteilung mit den zur Verfügung stehen- den Hilfsmitteln nicht realisiert werden. Daher wurden die Kleinsäugerfallen nach der Beliebtheit der unterschiedlichen Mikrohabitate verteilt und nach eingehender Literaturstudie Habitatpräfe- renzen für die im Untersuchungsgebiet in Frage kommenden herbi- und granivoren Kleinsäugerar- ten Gelbhalsmaus (Apodemus flavicollis), Wald- maus (A. sylvaticus), Rötelmaus (Clethrionomys glareolus) und Erdmaus (Microtus agrestis) ab- geleitet (vgl. Gelbhalsmaus: BERGSTEDT 1965; Hansson 1971; ZEIDA 1961; MAZURKIEWICZ und RAJSKA-JURGIEL 1978; NIETHAMMER 1978, MAZUR- KIEWICZ 1984; RAJSKA-JURGIEL 1992; Waldmaus: FELTEN 1952; TELLERfA et al. 1991; JAMmon 1994; TA- TERSALL und WHITBREAD 1994; Rötelmaus: KıKKA- wa 1964; RAppA 1968; Hansson 1971; MILLER und GETZ 1976; BÄUMLER 1981; KırKLAND 1990; Erdmaus: SCHINDLER 1972; CHELKOWSKA et al. 1985; VımALA und HOoFFMEYER 1985; KIRKLAND 1990; NIEMEYER 1993). Diesen Präferenzen ent- sprechend kamen mehr Fallen in von Kleinsäu- gern häufig aufgesuchten Mikrohabitaten zum Einsatz (vgl. ADams und Geis 1983) und verhält- nismäßig weniger an Mikrohabitaten mit geringe- rer Präferenz (s. Tab. 1). Der Abstand zwischen den einzelnen Fallen betrug mindestens 10 m. Auf den Untersuchungsflächen wurden folgende sechs Mikrohabitate als Fallenstandorte gewählt (Habitataufnahme: Quadrat von 2x2 m, Fallen- standort als Zentrum; s. Tab. 1): 1. Jungwuchs: bis 7cm Brusthöhendurchmesser (BHD); 2. Gras: reine Grasstandorte; 3. Kraut/Strauch: krautige Pflanzen, deren unverholzte, oberirdische Stengel im Herbst absterben, und verholzte Pflanzen, die den Winter oberirdisch überdauern mit einer Hö- he bis 2m (z.B. Brennessel, Heidelbeere, Him- beere); 4. Fels/Stein: anstehendes Gestein oder Bereich mit Steinen über 20 cm Durchmesser; 5. Bestand: ab 10 cm BHD, 6. WT;WR*: Wurzelbal- len entwurzelter Bäume. Auf der Kontrollfläche im Altbestand A wurde der Standort Wurzelteller durch den Standort Waldrand ersetzt. Tabelle 1. Fallenverteilung (je Untersuchungsfläche) in Mikrohabitaten, die nach Literaturangaben für verschie- dene Kleinsäugerarten als unterschiedlich attraktiv gelten. Die Punkte in den Feldern entsprechen der nach Litera- turhinweisen vermuteten Nutzungsintensität der Mikrohabitate durch die Kleinsäugerarten (eeee sehr stark ge- nutzt, eee stark genutzt, ee genutzt, e wenig genutzt, - nicht genutzt). Aus der Anzahl der Punkte ergibt sich vertikal die Anzahl der Fallen je Mikrohabitat und horizontal die Anzahl der Fallen, die einer Art auf einer Unter- suchungsfläche zugedacht sind. A auf der Kontrollfläche im Altbestand wurde der Standort Wurzelteller (WT) durch den Standort Waldrand (WR) ersetzt. Mikrohabitate Jung- Gras wuchs Rötelmaus Erdmaus Gelbhalsmaus Waldmaus Fallenzahl je Mikrohab. 9 Kraut/ Strach Bestand WT; WR* Fallenzahl je Kleinsäugerart Fels/ Stein Vegetationsaufnahmen In der Nähe der Fallenstandorte wurden nach BrRAUN-BLANQUET (1964) Mitte Juli (FP2) insge- samt 18 Vegetationsaufnahmen durchgeführt. Teilweise waren es Wiederholungsaufnahmen aus bereits vorhergegangenen Untersuchungen. Auf- grund des Mosaikcharakters der Bodenvegetation wurde für die Aufnahme eine Flächengröße von 2mx2m gewählt. Die Aufnahmen wurden bewußt an Standorten mit unterschiedlichen Au- ßenbedingungen (Exposition, Neigung, Sonnen- einstrahlung, etc.) durchgeführt (je vier Aufnah- meflächen auf G und A, sowie zehn auf U - letztere im Hinblick auf die dort viel größere strukturelle Diversität). Der Abstand der Auf- nahmeflächen betrug mindestens 10 m. Ergebnisse Während des Untersuchungszeitraums wur- den 309 Kleinsäuger der Familien Muridae (Apodemus flavicollis, Apodemus sylvati- cus; zus. = 206) und Cricetidae (Clethriono- mys glareolus, Microtus agrestis;, zus. = 103) gefangen. Die Fangergebnisse der drei Untersuchungs- flächen insgesamt unterschieden sich auf den beiden Windwurfflächen wenig. Der prozen- tual größte Anteil der insgesamt gefangenen Tiere wurde auf den Windwurfflächen U (37%) und G (35%) erreicht, geringer fiel der Anteil auf der Altbestandfläche A mit 28% aus. Deutliche Unterschiede waren hin- gegen bei der Zahl der gefangenen Arten zu bemerken (Tab. 2). Rund 2/3 der gefangenen Kleinsäuger auf Windwurfflächen 293 Tiere auf U waren Gelbhalsmäuse, die restli- chen 36% entfielen auf Rötel-, Wald- und Erdmäuse. Aufgrund der geringen Fangzah- len werden Wald- und Erdmäuse hier in wei- terer Folge nicht mehr berücksichtigt. Gelbhalsmaus Die Gelbhalsmaus war, in Fangzahlen gese- hen (n = 179), auf allen drei Untersuchungs- flächen die dominante Species (Tab. 2). Auch unter Berücksichtigung der Wieder- fänge (9% auf U; 5,4% auf G, 31% auf A) war der Fangerfolg bei Gelbhalsmäusen auf allen drei Untersuchungsflächen am höch- sten. Mitte Juli (FP3) und Ende September (FP6) wurden Gelbhalsmäuse häufig auf der Fläche A gefangen, ab Juni war ein deutlicher Anstieg auf U zu erkennen (Abb. 1). Ende September war insgesamt gesehen die erfolgreichste Fangzeit. Die Anzahl der gefangenen Jungtiere und Subadulten war auf U im Sommer höher als auf A und kehrte sich gegen Ende des Untersuchungszeitraums wieder um: juvenil + subadult A: U: Fangperiode Ende Juni (FP2) 3:4, Fangperiode Mitte Juli (FP3) 3:15, Fangperiode Ende August (FP5) 5:6, Fangperiode Ende September (EBO)A225; Das Geschlechterverhältnis der Fänge fiel auf der Fläche A und auf allen Untersu- chungsflächen insgesamt signifikant zugun- sten der Männchen aus. Auf A unterschied sich die Anzahl der weiblichen und männli- Tabelle 2. Artenspektrum und Fangerfolg (in 100 Fallennächten FN). U: ungeräumte Windwurffläche (insg. 360 FN), G: geräumte Windwurffläche (insg. 400 FN), A: Altbestandfläche (insg. 400 FN). *: theoretischer (korrigierter) Fangerfolg. Um den direkten Vergleich der 3 Untersuchungsflächen zu ermöglichen, wurde die wetterbedingt entfallene Fangperiode interpolierend korrigiert. Korrekturfaktor K wurde zum jeweiligen Fangerfolg der entfallenen Fangperiode auf G hinzu- bzw. weggezählt (K = Differenz der Gesamtfangerfolge zwischen U und G/Anzahl der berücksichtigten Fangnächte). Kleinsäugerarten 23,0% 1,9* 9,4* 0,67 Gelbhalsmaus (Apodemus flavicollis) Waldmaus (Apodemus sylvaticus) Rötelmaus (Clethrionomys glareolus) Erdmaus (Microtus agrestis) Arten gesamt 35,55 Flächen ges. (20.0) 13 13,8 (1,9) 2,8 2,3 (8,9) 11 4,5 (0,6) 0,5 1,25 (31,4) 16,8* 2,35 8,35 0,85 27,3 21,9 el 294 I. GLITZNER et al. chen gefangenen Tiere hoch signifikant voneinander, auf allen Flächen zusammen höchst signifikant (p<0,01; p< 0,001; Tab. 3). In vier der sechs Fangperioden wur- den auf A mehr adulte Männchen gefan- gen, in der Fangperiode Anfang September war der Unterschied signifikant. Einzig auf G konnte in einer Fangnacht Anfang Sep- tember ein höherer Fang an Weibchen ver- zeichnet werden. Ein Großteil der Gelbhalsmäuse wurde in verkrauteten Standorten und Jungwüchsen gefangen (Tab. 4). Rötelmaus Die Rötelmaus war mit 94 gefangenen Tieren die am zweithäufigsten gefangene Art und bevorzugt auf G und U anzutref- fen. Ihre Wiıederfangrate auf diesen Flä- Apodemus flavicollis Z oO IS © N O) = ) LL_ D e=| lo) 42) = 1 2 3 ä 5 6 Fangperioden EA DOG HU Clethrionomys glareolus Ö ac © N Oö) ) U 2 = 9 . 2) OO (0 1 2 8 & 5 6 Fangperioden Abb. 1. Fangergebnis der Untersuchungsflächen für Gelbhalsmaus (Apodemus flavicollis) und Rötelmaus (Clethrio- nomys glareolus). A: Altbestand, G: geräumte Windwurffläche, U: ungeräumte Windwurffläche. Fangperiode 4 auf U: theoretischer Fangerfolg. Um den direkten Vergleich der 3 Untersuchungsflächen zu ermöglichen, wurde die wetterbedingt entfallene Fangperiode interpolierend korrigiert. Korrekturfaktor K wurde zum jeweiligen Fangerfolg der entfallenen Fangperiode 4 auf G hinzu- bzw. weggezählt (K = Differenz der Gesamtfangerfolge zwischen U und G/Anzahl der berücksichtigten Fangnächte). Zeitraum der Fangperioden: FP1: 18.-20. Juni, FP2: 14.-16. Juli, FP3: 31. Juli-02. August, FP4: 15.-16. August, FP5: 31. August-01. September, FP6: 22.-24. September. chen war mit je 17% höher als die der Gelbhalsmaus. Die meisten Fänge konn- ten auch bei dieser Art Ende September erzielt werden. Deutliche Schwankungen in den Fangerfolgen waren auf U und G zu erkennen, wo die Zahl der Fänge Ende Juni (FPl) und Mitte August (FP4) die niedrigsten Werte erreichte (Abb. 1). Auf A war im Verlauf der Untersuchungen ein stufenartiger Anstieg der Fangerfolge zu erkennen, wenn auch die Fangperiode Mitte August (FP4) etwas unter dem er- warteten Wert lag. Auf keiner der Flächen war bei den Fängen ein signifikanter Un- terschied im Geschlechterverhältnis auf- getreten (Tab. 3). Insgesamt wurden auf A und U mehr weibliche Tiere gefangen (53% bzw. 56%), auf G mehr männliche (60%). Rötelmäuse nutzten vor allem Wurzelteller bzw. in A Waldrandbereiche (31% bzw. 69%), Fels/Steinhabitate und Bestandesinseln (= ehemaliger Unterbe- stand; Tab. 4). Kleinsäuger auf Windwurfflächen 295 Vegetation und Witterung Auf den beiden Windwurfflächen U und G waren zum Untersuchungszeitpunkt Kahl- schlaggesellschaften mit Himbeeren (Rubus idaeus) und Berg-Reitgras (Calamagrostis varia) — mit starken bzw. gedämpften Tag- Nacht-Schwankungen der Temperatur - vorherrschend, im Altbestand eine Schnee- rosen-Buchen-Assoziation („Helleboro ni- gri-Fagetum Zukrigl 1973“). Für Gelbhals- und für Rötelmäuse wurden auch bei tieferen Temperaturen (im Mittel 5°C) hohe Fangerfolge erzielt (Abb. 2). In der Untersuchungszeit von Ende August (FP5) bis Ende September (FP6) lag das Temperaturmittel bereits unter zehn Grad Celsius. Obwohl nur drei der insgesamt 10 hier berücksichtigten Fangnächte in diesen Temperaturbereich fielen, wurden 45% der Rötelmäuse in diesem Zeitraum gefangen. Andere Witterungsverhältnisse schienen wenig relevant gewesen zu sein: hohe Fang- Tabelle 3. Geschlechterverhältnis (Weibchen/Männchen w/m) auf den Untersuchungsflächen. U: ungeräumte Windwurffläche, G: geräumte Windwurffläche, A: Altbestandfläche. Untersuchungszeitraum von Juni-September 1995, 6 Fangperioden. Signifikanzniveau (Binomial Test): ***: p< 0,001; **:p< 0,01). Kleinsäuger- arten Gelbhalsmaus Rötelmaus 0,62 68 0,6 1,29 32 0,68 gesamt n w/m 0,38 2 54 1,14 15 0,527 77 1110 0,98 89 Tabelle 4. Angebot und Nutzung. Anzahl der Fallennächte (FN) je Mikrohabitat und Summe der Kleinsäugerfänge während 6 Fangperioden (Juni-September 1995). U: ungeräumte Windwurffläche, G: geräumte Windwurffläche, A: Altbestand. WT: Wurzelteller, WR: Waldrand, *: im Altbestand wurde der Standort Wurzelteller durch den Stand- ort Waldrand ersetzt. Mikrohabitat Fallenzahl Jungwuchs Gras Kraut/Strauch Fels/Stein Bestand WT; WR" Anzahl U,G, A, ges. FN Fangerfolg (in 100 FN): Summe Gelbhalsmaus Rötelmaus 90,8 296 I. GLITZNER et al. ergebnisse wurden sowohl bei Sonnen- schein als auch bei Regenwetter erzielt, die geringsten während bedeckter bzw. bewölk- ter Tage. Diskussion Nach Untersuchungen von GERLACH (1996) scheinen Sturmschadensflächen, unabhän- gig von der Art der Bewirtschaftung, Klein- säugern einen idealen Lebensraum zu bie- ten. Werden sie nach dem Windwurf forstwirtschaftlich geräumt, entsprechen sie Kahlschlägen, deren Strukturierung sich auf zurückgeklappte Wurzelteller und Frat- ten reduziert und deren Vegetation nach ei- niger Zeit von typischen Kahlschlaggesell- schaften mit dichtem Unterwuchs geprägt wird. Freie Flächen dieser Art sind Witterungseinflüssen und kleinklimatischen Schwankungen verstärkt ausgesetzt. Durch das Belassen von Totholz auf Sturmscha- denflächen hingegen entsteht ein reich strukturierter Lebensraum mit mosaikarti- gem Charakter (vgl. Harrıs 1984). Kleinsäuger, wie aus den hier untersuchten Unterfamilien Murinae und Cricetinae, un- terscheiden sich im Hinblick auf Verhalten und Lebensraumwahl. Die sich oberirdisch fortbewegende Waldmausart Apodemus flavicollis ist ein ausgezeichneter Kletterer (vgl. HOFFMEYER 1973; VITALA und Horr- MEYER 1985). Sie nutzt horizontale Struktu- ren in ihrem Habitat zur Fortbewegung und lebt bevorzugt im Waldesinneren. Die Wühlmaus Clethrionomys glareolus hinge- gen lebt vorwiegend unterirdisch in Habita- ten mit dichtem Unterwuchs und Strauch- bestand. Sie kann durch Verbiß von Baumkeimlingen, -trieben, -knospen, -wur- zeln und durch das Schälen von Rinde in 15 2 14,3 = 11,3 5 30 117 ’ = 10° c E = (5 c 20 > Oo Q | | | 210 ) | lab 0 HE 0 1 2 3 4’ 5 6 Fangperioden C. glareolus DIA. flavicollis = T°C (Mittel) Abb. 2. Fangzahlen (in 100 Fallennächten FN) und mittlere Lufttemperatur. “: theoretischer Fangerfolg. Um den direkten Vergleich der 3 Untersuchungsflächen zu ermöglichen, wurde die wetterbedingt entfallene Fangperiode interpolierend korrigiert. Korrekturfaktor K wurde zum jeweiligen Fangerfolg der entfallenen Fangperiode auf G hinzu- bzw. weggezählt (K= Differenz der Gesamtfangerfolge zwischen U und G/Anzahl der berücksichtigten Fangnächte). Zeitraum der Fangperioden von Juni bis September 1995 wie in Abb. 1. der Forstwirtschaft akute Schäden verursa- chen. Gelbhals- und Rötelmäuse waren den Fang- zahlen nach dominante Kleinsäugerarten im untersuchten Bergwald im Altbestand wie auf den Sturmflächen. Die höchsten Fangergebnisse mit Gelbhalsmäusen wur- den auf der ungeräumten Windwurffläche U erzielt. Die morphologisch zum Klettern begünstigten Tiere konnten durch die rei- che Strukturierung der Fläche (liegende Stämme, Astwerk) ihr Territorium in weite- re Ebenen, horizontal und vertikal, ausdeh- nen (Fangerfolg Wurzelteller: 5,3 pro 100 Fallennächte). Auffallend war der saisonale Verlauf der Fangergebnisse auf den Untersuchungsflä- chen. Die Zahl der Fänge im Altbestand A war zu Beginn und am Ende des Untersu- chungszeitraums hoch. Obwohl die Fangpe- riode im Mai in den hier vorliegenden Er- gebnissen nicht berücksichtigt wird, da während dieser FP auf A mit nur 24 statt 40 Fallen gefangen wurde, sei hier erwähnt, daß von den sechs Gelbhalsmäusen vier im Altbestand gefangen wurden. Der Fanger- folg auf U verlief entgegengerichtet: er er- reichte zur Mitte der Fangsaison seinen Hö- hepunkt. Die Fangergebnisse auf G hielten sich in einem mittleren, relativ stabilen Be- reich. Als Erklärung kann für Gelbhalsmäuse auf- grund der vorliegenden Ergebnisse ein sai- sonaler Habitatwechsel in Betracht gezogen werden. Das Nahrungs- wie das Deckungs- angebot für samen- und beerenfressende Mäuse war im Sommer und Herbst auf der Windwurffläche wesentlich höher als im Altbestand. Die Zusammensetzung des Kö- ders blieb über den Untersuchungszeitraum unverändert. Zu Zeiten einer möglichen Nahrungsknappheit sollte die Attraktivität des Fallenköders steigen und bessere Fang- ergebnisse nach sich ziehen. Doch das traf ım Falle der hier untersuchten Altbestand- fläche nicht zu. Saisonaler Habitatwechsel konnte schon früher einige Male bei Apo- demus-Arten beobachtet werden (vgl. z.B. BERGSTEDT 1965; FLOWERDEW 1974, nach VIITALA und HoFFMEYER 1985). Um die hier vorliegenden Ergebnisse statistisch zu bele- Kleinsäuger auf Windwurfflächen 297 gen, sind jedoch Studien über mehrere Jah- re erforderlich. Kraut/Strauchstandorte wurden von Gelb- halsmäusen erstaunlich häufig aufgesucht, wenn man von bekannten bzw. den für das Untersuchungsgebiet abgeleiteten Habitat- präferenzen ausgeht. In der Fangperiode Ende August (FP5) bei Regenwetter und den niedrigsten Außentemperaturen konnte auf Grasstandorten die größte Zahl an Fängen (3/100 FN) verzeichnet werden. Aus Untersuchungen von STOUTJESDIJK und BARKMAN (1992) ist bekannt, daß Grasstel- len (liegende und abgestorbene Gräser) durch ihre kompakte Struktur deutlich hö- here Temperaturwerte aufweisen als die Umgebung. Es wäre denkbar, daß Mäuse bei schlechten Wetterverhältnissen Zu- flucht in Grasbeständen und den darin pla- zierten Fallen suchten. Die Tatsache, daß deutlich mehr Männchen in den Fallen gefunden wurden, steht wahr- scheinlich in Zusammenhang mit der Fal- lenverteilung und dem territorialen Verhal- ten bzw. der Abwanderneigung der jungen Männchen. Die männlichen Tiere der Gelb- halsmäuse besitzen größere Streifgebiete als ihre weiblichen Artgenossen und ver- größern im fortpflanzungsaktiven Zustand ihren Aktionsraum von 0,3-2 ha auf bis zu max. 5 ha, wobei sıe rund 25-40% längere Laufstrecken zurücklegen (vgl. SCHWARZEN- BERGER und KLinGeEL 1995). Durch die unre- gelmäßige Verteilung der Fallen wurde die Untersuchungsfläche nicht systematisch er- faßt. Möglicherweise ergab dies für die mo- bileren Männchen bessere Fangerfolge als für die Weibchen. Auch sind weibliche Säu- getiere generell vorsichtiger als Männchen und aus diesem Grund schwerer zu fangen. Ähnliche geschlechtsspezifische Unter- schiede bei Fängen von Gelbhalsmäusen konnte auch RAISKA-JURGIEL (1992) beob- achten. Rötelmäuse bevorzugten den Fangergebnis- sen nach die geräumte Windwurffläche G. Als Bewohner von Kulturen mit Kräutern und verholzenden Sträuchern (vgl. BÄum- LER 1981) bot ihnen diese Untersuchungs- fläche mit ausgeprägter Schlagvegetation und stellenweise offenen und trockenen Be- 298 IT. GLITZNER et al. reichen einen idealen Lebensraum. Reiches Blockwerk, spaltenreicher Untergrund und tiefgründige Böden in den Mulden haben die Mäuseart, die unterirdische Baue und lange Röhren anlegt, begünstigt, was auch das hohe Fangergebnis von Rötelmäusen in steinigen Bereichen zeigt. Ebenso wurden Rötelmäuse sehr häufig in Bestandesflä- chen und am Waldrand (8,6 Fänge in 100 Fallennächten) gefangen. Erstaunlich bei den Rötelmäusen waren die verhältnis- mäßig schlechten Fangergebnisse in Gras-, Kraut- und Strauchbereichen, die sonst als Präferenzbereiche gelten. Auf A und U stiegen die Fangergebnisse ge- gen Ende des Untersuchungszeitraums an, wobei U allgemein höhere Fangerfolge als A aufwies. Auf G war ein deutlich geringerer Fangerfolg während der 1. und 4. Fangperi- ode zu verzeichnen. Womit diese Schwan- kungen in Zusammenhang zu stellen sind, kann aufgrund des kurzen Untersuchungs- zeitraums nicht geklärt werden. Auf keiner der drei Untersuchungsflächen konnten bei Fängen an Rötelmäusen den Gelbhalsmäu- sen vergleichbare Geschlechtsunterschiede festgestellt werden. Anders als bei den Gelb- halsmäusen wurden hier auf zwei der drei Untersuchungsflächen etwas mehr Weibchen gefangen. Das schon früher beobachtete Ver- halten der erhöhten Mobilität territorialer adulter Weibchen (ANDRZEJEWSKI und OL- ZEWSKI 1963; VIITALA und HOFFMEYER 1985) kann dazu beigetragen haben, daß sich die Fangerfolge männlicher und weiblicher Tiere nicht signifikant unterschieden. Der von RAJSKA-JURGIEL (1992) beobachtete Weib- chenüberschuß während der Fortpflanzungs- zeıt erklärt ebenfalls den guten Fangerfolg bei Weibchen auf U und A. Die Fangergebnisse auf den Untersu- chungsflächen entsprachen insgesamt den aus der Literatur entnommenen Habitat- ansprüchen von Gelbhalsmaus und Rötel- maus. Interessant waren allerdings die Fangergebnisse ın den einzelnen Mikroha- bitaten, die zum Teil von den in der Litera- tur erwähnten Präferenzen abwichen. Für Gelbhalsmäuse gehörten Kraut/Strauch- standorte in der vorliegenden Untersu- chung insgesamt zu den Standorten mit dem höchsten Fangerfolg. Auf Grasstandor- ten konnten in der kältesten und regneri- schen Fangperiode Ende August (FP5) die höchsten Fangzahlen erzielt werden. So- wohl Gras- als auch Kraut/Strauchstandorte gehören laut Literaturangaben nicht zu den bevorzugten Mikrohabitaten der Gelbhals- maus. Die Fangergebnisse der Rötelmäuse wichen ebenfalls im Mikrohabitat Kraut/ Strauch von den erwarteten Ergebnissen ab. Literaturhinweisen zufolge hält sich die Rötelmaus bevorzugt an diesen Standorten auf, die Fangergebnisse waren hier jedoch erstaunlich gering. Mit Hilfe dieser Studie sollten Informatio- nen über die Kleinsäugerbesiedlung eines windwurfbeeinflußten Bergwaldes im Rand- bereich der Nördlichen Kalkalpen gewon- nen werden. Die Dominanzverhältnisse der Arten waren aufgrund der Fangzahlen ein- deutig zuzuordnen. Die aus der Literatur be- kannten Habitatansprüche von Gelbhals- und Rötelmaus aus tieferen Regionen trafen weitgehend auch für diesen montanen Berg- wald zu. Auf den hier vorliegenden Sonder- standorten bevorzugte die Gelbhalsmaus eindeutig die naturnahe Windwurffläche, die durch ihre strukturelle Vielfalt mit einem Bestandesinneren vergleichbar ist. Für sie kann ein saisonaler Habitatwechsel zwischen Altbestand und ungeräumter Windwurfflä- che in Betracht gezogen werden. Die Rötel- maus hingegen nutzte die intensiv bewirt- schaftete kahlschlagähnliche Fläche mit gutem Unterwuchs, Wurzeltellern und steini- gem Untergrund stärker. Da sie zu einem der wichtigsten Forstschädlinge zählt, ist diese Präferenz bei der Wahl der Bewirt- schaftungsmethode auch im montanen Be- reich zu bedenken. Für die Mikrohabitate Gras und Kraut/Strauch deckten sich die Er- gebnisse im untersuchten Bergwald jedoch nicht mit den Literaturangaben. Danksagungen Herzlicher Dank gilt dem Institut für Wild- biologie und Jagdwirtschaft der Universität für Bodenkultur, Wien, für die Bereitstel- lung der Fallen sowie der Rothschild’schen Forstverwaltung, insbesondere Herrn Ing. T. FRITZ, für die Kooperationsbereitschaft. Herzlicher Dank gilt ebenso unseren Kolle- Kleinsäuger auf Windwurfflächen 299 gen wie auch besonders den unbekannten Reviewern für die kritische Durchsicht des Manuskripts. Zusammenfassung 1995 wurden im Steirisch-Niederösterreichischen Grenzgebiet, fünf Jahre nach einer schweren Sturmkatastrophe, eine forstwirtschaftlich geräumte und eine nicht geräumte Windwurffläche sowie eine Altbestandfläche faunistisch und floristisch untersucht. Ziel der Untersuchung war es, Daten über die Kleinsäugerbesiedlung der montanen Waldstufe im Bereich des nordöstlichen Ausläufers der Nördlichen Kalkalpen zu erhalten. Ebenso war zu untersuchen, ob die aus der Literatur bekann- ten unterschiedlichen Habitatansprüche von Gelbhalsmäusen (Apodemus flavicollis) und Rötelmäu- sen (Clethrionomys glareolus) tieferer Regionen auch auf Sonderstandorte, wie jene forstwirtschaftlich differenziert behandelten Windwurfflächen (naturnahe und intensive Be- wirtschaftung) des Bergwaldes, übertragbar sind. Zum Zeitpunkt der Untersuchung herrschten auf beiden Sturmflächen Kahlschlaggesellschaften vor. Auf der nicht geräumten Windwurffläche wurde der größte Fangerfolg an Muriden erzielt. Die Gelb- halsmaus (Apodemus flavicollis) war auf allen drei Flächen die am häufigsten gefangene Art, gefolgt von Rötelmaus (Clethrionomys glareolus), Waldmaus (Apodemus sylvaticus) und Erdmaus (Microtus agrestis). Gelbhalsmäuse wurden vermehrt auf der ungeräumten Windwurffläche in verkrauteten Standorten und Jungwüchsen gefangen. Rötelmäuse gingen hingegen vermehrt auf der geräumten Fläche in Fallen, die auf Fels/Stein-, Bestand- und Waldrandstandorten plaziert waren. Für Gelb- halsmäuse konnte aufgrund der Fangzahlen ein saisonaler Habitatwechsel in Betracht gezogen wer- den. Das Geschlechterverhältnis der Gelbhalsmäuse fiel auf einer Fläche sowie auf allen Untersu- chungsflächen insgesamt signifikant zugunsten der Männchen aus. Die Fangergebnisse der einzelnen Untersuchungsflächen entsprachen weitgehend den aus der Literatur entnommenen Habitatansprüchen. Interessant war allerdings das Fangergebnis in den einzelnen Mikrohabitaten, das zum Teil von den in der Literatur erwähnten Präferenzen abwich. Literatur ADAMS, L. W.; GEIS, A. D. (1983): Effects of roads on small mammals. J. Appl. Ecol. 20, 403-415. ÄNDRZEJEWSKI, R.; OLSZEWSKI, J. (1963): Social behaviour and interspecific relations in Apo- demus flavicollis (Melchior, 1834) and Clethri- onomys glareolus (Schreber, 1780). Acta Theriol. 7, 155-268. BÄUMLER, W. (1981): Zur Verbreitung, Ernährung und Populationsdynamik der Rötelmaus (Cle- thrionomys glareolus) und der Gelbhalsmaus (Apodemus flavicollis) in einem Waldgebiet der Bayerischen Alpen. Anz. Schädlingskde., Pflanzenschutz, Umweltschutz 54, 49-53. BERGSTEDT, B. (1965): Distribution, reproduction, growth and dynamics of the rodent species Clethrionomys glareolus, Apodemus flavicollis and Apodemus sylvaticus in southern Sweden. Oikos 16, 132-160. BRAUN-BLANQUET, J. (1964): Pflanzensoziologie. 3. Aufl. Wien: Springer Verlag. CHELKOWSKA, H.; WALKOWA, W.; ADAMCZYK,K. (1985): Spatial relationship in sympatric populations of the rodents: Clethrionomys glareolus, Microtus agrestis and Apodemus agrarius. Acta Theriol. 30, 51-78. FELTEN, H. (1952): Untersuchungen zur Ökologie und Morphologie der Waldmaus (Apodemus sylvaticus) und der Gelbhalsmaus (Apodemus flavicollis) im Rhein-Main-Gebiet. Bonn. zool. Beitr. 3, 187-206. FLOWERDEW, J. R. (1974). Field and laboratory ex- periments on the social behaviour and popu- lation dynamics of the wood mouse (Apode- mus sylvaticus). J. Anim. Ecol. 43, 499-511. FLOWERDEW, J. R. (1985). The population dy- namics of wood mice and yellow-necked mice. Symp. Zool. Soc. London 55, 315-338. GERLACH, G. (1996): Kleinsäuger auf Sturmflä- chen. Infodienst Wildbiologie und Oekologie in der Schweiz 4, 1-12. 300 I. GLITZNER et al. Hansson, L. (1971): Small rodent food, feeding and population dynamics. Oikos 22, 183-198. HARRIS, L. D. (1984): The Fragmented Forest. Is- land Biogeography Theory and the Preserva- tion of Biotic Diversity. Chicago, London: University of Chicago Press. HOFFMEYER, I. (1973): Interaction and habitat se- lection in the mice Apodemus flavicollis and Apodemus sylvaticus. Oikos 24, 108-116. Jamon, M. (1994): An analysis of trail-following behaviour in the wood mouse, Apodemus syl- vaticus. Anim. Behav. 47, 1127-1134. KIKKAWA, J. (1964): Movement, activity and distri- bution of the small rodents Clethrionomys glareolus and Apodemus sylvaticus in wood- land. J. Anim. Ecol. 33, 259-299. KIrKLAND, G. L. (1990): Patterns of small mam- mal community change after clearcutting of temperate north American forests. Oikos 59, 313-320. MAZURKIEWICZ, M.; RAISKA-JURGIEL, E. (1978): Size and structure of rodent community of various forest stand types. Bull. Acad. pol. Sci. CI II 26, 669-677. MAZURKRIEWICZ, M. (1984): Population density of small rodents as affected by chosen elements of tree stand structure. Bull. Acad. pol. Sci. Cl II 32, 209-217. MILLER, D.H.; GETZ, L.L. (1976): Factors influ- encing local distribution and species diversity of forest small mammals in New England. Can. J. Zool. 55, 806-814. NIEMEYER, H. (1993): Neue Konzeption zur Be- kämpfung der Erdmaus (Microtus agrestis) in forstlichen Verjüngungen Norddeutschlands. Anz. Schädlingskde., Pflanzenschutz, Um- weltschutz 66, 41-46. NIETHAMMER, J. (1978): Apodemus flavicollis (Mel- chior, 1834) — Gelbhalsmaus. In: Handbuch der Säugetierkunde. Bd. 1, Rodentia I. Wies- baden: Akademische Verlagsgesellschaft. RADDA, A. (1968): Populationsstudien an Rötel- mäusen (Clethrionomys glareolus) durch Mar- kierungsfang in Niederösterreich. Oecologia 1, 219-235. RAJSKA-JURGIEL, E. (1992): Demography of wood- land rodents in fragmented habitat. Acta Theriol. 37, 73-90. SCHINDLER, U. (1972): Massenwechsel der Erd- maus, Microtus agrestis, in Süd-Niedersachsen von 1952-1971. Z. angew. Zool. 59, 189-203. SCHWARZENBERGER, T.; KLINGEL, H. (1995): Tele- metrische Untersuchungen zur Raumnutzung und Aktivitätsrythmik freilebender Gelbhals- mäuse Apodemus flavicollis. Z. Säugetier- kunde 60, 20-32. STOUTJESDUK, PH.; BARKMAN, J. J. (1992): Microcli- mate, Vegetation and Fauna. Schweden: Opu- lusPressAB. TATTERSALL, F.; WHITBREAD, S. (1994). A trap- based comparison for the use of arboreal ve- getation by populations of bank vole (Clethrio- nomys glareolus), woodmouse (Apodemus sylvaticus) and common dormouse (Muscardi- nus avellanarius). J. Zool. 233, 309-314. TELLERIA, J. L.; SANTOS, T.; ALCANTARA, M. (1991): Abundance and food-searching intensity of wood mice (Apodemus sylvaticus) in frag- mented forests. J. Mammalogy 72, 183-187. VIITALA, J.; HOFFMEYER, I. (1985): Social organiza- tion in Clethrionomys compared with Micro- tus and Apodemus: Social odours, chemistry and biological effects. Ann. Zool. Fennici 22, 359-371. ZEIDA, J. (1961): Age structure in populations of the bank vole, Clethrionomys glareolus. Zool. Listy 11, 249-264. (Tschechisch, Zusammen- fassung in Englisch). Anschrift der Verfasser: Mag. IRENE GLITZNER und Univ. Prof. Dr. HART- MUT Gossow, Institut für Wildbiologie und Jagd- wirtschaft, Universität für Bodenkultur Wien, Peter Jordanstraße 76, A-1190 Wien, Österreich (e-mail: glitzner@edv1.boku.ac.at). Mamm. biol. 66 (2001) 301-304 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication Mammalian Biology Zeitschrift für Säugetierkunde The occurrence of roof rats (Rattus rattus L., 1758) in Germany during the late 20°" century By S. ENDEPOLS, H. DIETZE, and HEIKE ENDEPOLS Bayer AG Animal Health, Monheim; Landeshygieneinstitut Sachsen-Anhalt, Magdeburg and Zoologisches Institut, Universität zu Köln, Köln, Germany Receipt of Ms. 29. 09. 2000 Acceptance of Ms. 02. 02. 2001 Key words: Rattus rattus, roof rat, black rat, distribution, Germany Roof rats (Rattus rattus Linne, 1758) have been existant in the area covered by pre- sent-day Germany for almost 2000 years. Excavations revealed remains dating back to the second century (LÜTTSCHWAGER 1968), to the 3rd-Sth century (TEICHERT 1985), and to the early middle ages (REICH- STEIN 1974, 1987). Patterns of their distribu- tion in Central Europe appear not to be linked to natural conditions, such as cli- mate and vegetation, but rather to man- made conditions, which tend to change ra- pidly. Roof rats may suddenly be intro- duced by the transport of goods, and they may be extinguished locally when indus- trial sites are abandoned or farms are mod- ernized (BECKER 1978). Pest control opera- tions frequently reduce their populations. The occurrence of roof rats is very dynamic and not characterized by distribution lines used to describe the distribution of endem- ic species. Many sites where roof rats become abun- dant remain unknown to scientists because farmers and pest control operators do not identify the species, and faunistic research projects rarely focus on commensal mam- mals. Although the roof rat has been classi- fied as “extinct or disappeared” during the 1990s in some federal states of Germany, such as Bayern, Nordrhein-Westfalen, 1616-5047/01/66/05-301 $ 15.00/0. Hessen, Baden-Württemberg, and Rhein- land-Pfalz (Nowak et al. 1994), pest con- trol operations targeted this rat in other States. Roof rats are generally less susceptible than Norway rats (Rattus norvegicus) to antıcoagulant rodenticides, which have been predominantly used to control com- mensal rodents during the last few decades. To some degree, they are resistant to war- farın (ENDEPOLS and SCHUSTER 1991). Their control is also ıimpeded by the fact that they use smaller home ranges than Norway rats (TELLE 1966; ENDEPOLS et al. 1989). Incorrect identification of the spe- cies during control operations may have resulted in an underestimated abundance of roof rats. To summarize all occurrences of roof rats ın Germany, we collated those that we identi- fied ourselves and those published by others. In addition, we included analyses of owl pellets, and information obtained from data bases from the Landeshygieneinstitut Sachsen-Anhalt (LHI) in Magdeburg and from Bayer Animal Health in Monheim. The data comprise occurrences of roof rats, which were detected by farmers, millers or pest control operators. Such reports were verified when animals were sighted or car- casses found by the authors, by staff of one 302 S. EnDEPOLS et al. of these institutes or by another hygiene in- stitute. The database of the LHI predomi- nantly contains data from the eastern states of the 1980s. During this time, the LHI ad- vised all rat control programs in the former German Democratic Republic (GDR) and was responsible for the registration of ro- denticides. Most infestations were recorded on farms in Sachsen, Sachsen-Anhalt, and Branden- burg in the 1980s (ERFURT et al. 1986). On large pig farms in these federal states, roof rats established particularly large popula- tions exceeding 10000 individuals (ENDE- PoLS et al. 1989). Poor standards in the ex- tensive animal production in the GDR, such as food spillage, hollow walls and pe- netrable roofs, supported successful repro- duction, even during the cold winters in East Germany (EnDEPoLS 1992). Simultaneously, roof rats were considered locally extinct in some large federal states Fig. 1. Occurrences of roof rats (R. rattus) in Germany, 1980-1999. Each district where roof rats were recorded is marked by a dot. BB= Brandenburg, BW = Baden Württemberg, BY= Bayern (Bavaria), HE= Hessen (Hesse), HH = Hamburg, MV = Mecklenburg-Vorpommem, NI= Niedersachsen (Lower Saxony), NW = Nordrhein-Westfalen (North Rhine-Westphalia), RP = Rheinland-Pfalz (Rhineland-Palatinate), SA = Sachsen (Saxony), SH = Schleswig- Holstein, SL= Saarland, TH = Thüringen (Thuringia). The occurrence of roof rats (Rattus rattus L., 1758) in Germany in western and southern Germany, such as Rheinland-Pfalz (GRÜNWwALD and PREUSS 1983, 1987), Baden-Württemberg (BRAUN 1989), Nordrhein-Westfalen, Bayern, and Hessen (Nowak et al. 1994). Recently, roof rats were re-discovered in some of these areas. Roof rats were detected in all but two small federal states, Schleswig-Holstein and Saar- land, during the last 20 years (Fig. 1). Most populations in western Germany were de- tected in harbour areas along rivers. We found them in grain mills, food mills and si- los along the river Main in the cities of Würzburg (1997) and Hanau (1995), and downstream of the Rhine River in cities, such as Cologne (1999), Düsseldorf (1996), Neuss and Wesel (1995). Roof rats were also detected in southwest Germany (BRÜNNER and TRoJE 1991), at the Boden- see (lake of Constance) (WILHELM, BRAUN, and DIETERLEN, pers. comm.) and in the area of the middle Rhine (DALBEcK 1996). In all regions, we observed the colour varia- tions “rattus”, “alexandrinus”, and “frugi- vorus”. In Hanau on the river Main and in Cologne on the Rhine River, we found pure populations of black rats. However, due to small sample sızes and specimens lacking fur, such as skulls and bones, characteriza- tion of a representative number of popula- tions was not feasible. In general, the three variations of fur colour appear purely or in mixed populations in Germany. Although large populations of roof rats ap- peared in habitats such as pig farms, grain mills and silos, light infestations were also detected on small farms, in restaurants and in small food-producing factories. Even in rural residential buildings a few roof rats were observed. Such infestations were reported solely where large rat popu- lations were established nearby. Although many occurrences of roof rats probably re- mained unnoticed, our data support a pre- vious evaluation that this species is not en- dangered in Germany (Boye et al. 1998). Conservation measures are neither neces- sary nor reasonable because roof rats re- present a pest in the food industry and agriculture. 303 References BECKER, K. (1978): Rattus rattus (Linnaeus, 1758) — Hausratte (HR). In: Handbuch der Säuge- tiere Europas. Vol. 1/I. Ed. by J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Pp. 382-400. BoYyE, P.; HUTTERER, R.; BENKE, H. (1998): Rote Liste der Säugetiere. In: Rote Liste der ge- fährdeten Säugetiere Deutschlands. Bonn: Bundesamt für Naturschutz. Pp. 33-39. BRAUN, M. (1989): Zum Vorkommen der Säuge- tiere in Baden-Württemberg. Veröff. Natur- schutz Landschaftspflege Bad.-Württ. 64/65, 145-201. BRÜNNER, H.; TROJE, N. (1991): Ein Vorkommen der Hausratte (Rattus rattus L.) in Südbaden. Mitt. bad. Landesver. Naturkunde u. Natur- schutz 15, 467-468, DALBECcK, L. (1996): Die Bedeutung von Haus- ratte (Rattus rattus) und Wanderratte (Rattus norvegicus) für dıe Ernährung des Uhus (Bubo bubo) in Eifel und Saarland. Säuge- tierkdl. Inf. 4, 155-162. ENDEPOLS, S. (1992): Untersuchungen zur Popula- tionsökologie der Hausratte (Rattus rattus L.) anhand der Augenlinsen-Trockenmasse. Säuge- tierkd. Inf. 3, 401-407. ENDEPOLS, S.; SCHUSTER, W. (1991): Laborunter- suchungen zur Resistenz der Hausratte (Rat- tus rattus L.) gegen Warfarin in Nordost- deutschland. Z. angew. Zool. 78, 489-495. ENDEPOLS, S.; RÖDER, R.; SCHUSTER, W. (1989): Zur Abundanz und Reproduktion einer Population der Hausratte, Ratftus rattus (L. 1758). Säugetierkdl. Inf. 3, 109-112. ERFURT, J.; RÖDER, R.; SCHUSTER, W. (1986): Zur Verbreitung der Hausratte Rafttus rattus (L. 1758) auf dem Territorium der DDR. Säu- getierkdl. Inf. 2, 303-310. GRÜNWALD, A.; PREUSS, G. (1983): Säugetiere in Rheinland-Pfalz. Verzeichnis der wildleben- den Säugetiere, einschließlich der verscholle- nen und ausgestorbenen Arten (Vertebrata: Mammalıa). Beitr. Landespfl. Rheinland- Pfalz 9, 66-91. GRÜNWALD, A.; PrEUSss, G. (1987): Säugetiere (Mammalia). In: Rote Liste der bestandsge- fährdeten Wirbeltiere in Rheinland-Pfalz. Ed. by Ministerium für Umwelt und Gesundheit Rheinland-Pfalz. Grünstadt: Verlag E. Som- mer. Pp. 13-19. LÜTTSCHWAGER, J. (1968): Hamster- und Hausrat- tenfunde im Mauerwerk eines römischen Brunnens in Ladenburg, Landkreis Mann- heim. Säugetierkdl. Mitt. 16, 37-38. 304 S. ENDEPOLS et al. Nowak, E.; HEIDECKE, D.; BLAB, J. (1994): Rote Liste und Artenverzeichnis der in Deutsch- land vorkommenden Säugetiere (Mammalia). In: Rote Liste der gefährdeten Wirbeltiere in Deutschland. Ed. by E. Nowak, J. BLaB and R. Bess, Schr.-R. f. Landschaftspfl. u. Na- tursch. 42, 27-58. REICHSTEIN, H. (1974): Bemerkungen zur Verbrei- tungsgeschichte der Hausratte (Rarttus rattus, Linne 1758) an Hand jüngerer Knochenfunde aus Haithabu (Ausgrabung 1966-69). Die Heimat, Wachholtz-Verlag Neumünster 81, 113-114. REICHSTEIN, H. (1987): Archäozoologie und die prähistorische Verbreitung von Kleinsäugern. Sber. Ges. Naturfreunde N. F. 27, 9-21. TEICHERT, M. (1985): Beitrag zur Faunenge- schichte der Hausratte, Rattus rattus L. Z. Archäol. 19, 263-269. TELLE, H.-J. (1966): Beitrag zur Kenntnis der Verhaltensweisen von Ratten, vergleichend dargestellt bei Rattus norvegicus und Rattus rattus. Z. angew. Zool. 53, 129-196. Authors’ addresses: STEFAN ENDEPOLS, Bayer AG, Animal Health R & D, Agricultural Centre Monheim, Building 6220, D-51368 Leverkusen, Germany (e-mail: Stefan.Endepols.SE@bayer-ag.de), HoLM DIETZE, Landeshygieneinstitut Sachsen-Anhalt, PF 1736, D-39007 Magdeburg; HEIKE ENDEPOLS, Universi- tät zu Köln, Zoologisches Institut, Weyertal 119, D-50923 Köln Mamm. biol. 66 (2001) 305-307 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication Mammalian Biology j Zeitschrift für Säugetierkunde Caecotrophy in pacas (Agouti paca Linnaeus, 1766) By VERA SaBATINI and M. J. R. PARANHOS DE COSTA Centro de CiEncias e Biotecnologia, Universidade Estadual do Norte Fluminense, Campos, Brasil and ETCO-Grupo de Estudos em Ecologia Animal, Departamento de Zootecnia, Universidade Estadual Paulista, Jaboticabal, Brasil. Receipt of Ms. 04. 10. 2000 Acceptance of Ms. 23. 03. 2001 Key words: Agouti paca, captivity, caecotrophy Caecotrophy, a physiological process which was early documented in rabbits (MoroT 1882), is recognised to occur in mammals of different species (Gorilla gorilla: HAR- COURT and STEWART 1978; Phascolarctos ci- nereus: Osawa et al. 1993; Hydrochaeris hy- drochaeris: BORGES et al. 1996; Myocastor coypus: TAKAHASHI and SAKAGUCHI 1998). Nevertheless, it has been best documented for lagomorphs and some rodents (STIL- LINGS and HACKLER 1966; PICKARD and STE- VENS 1972; BIÖRNHAG and SJIÖBLOM 1977; CRANFORD and JoHNsoN 1989; SoAVE and BrAnD 1991; MAROUNEK et al. 1995). These herbivores ingest differentiated faeces and absorb the protein and carbohydrates synthesized by caecal microorganisms. The paca (Agouti paca) is the second lar- gest neotropical hystricognath rodent with an adult average body weight of S kg. It is distributed from southern Mexico to nor- tern Argentina, in practically all forest ha- bitats up to 2000 m of altitute (Woops 1984). This species has become locally ex- tinct in overhunted areas of Central Ameri- ca (Emmons 1990), and is considered vul- nerable to extinction in some areas of Brazil, because of the reduction of its habi- tats and hunting pressure (AyrES et al. 1991; VIcKErs 1991; BERGALLO et al. 2000). These mammals are mainly frugivorous 1616-5047/01/66/05-305 $ 15.00/0. (Monporrt 1972), but SMYTHE et al. (1983) suggested that pacas could browse on leaves and seedlings during fruit shortage seasons. A study of the behavioural patterns of 11 pacas in captiıvity was conducted at the Universidade Estadual Paulista, in Jaboti- cabal, Brazil between February and March of 1998. Animals were grouped as four mated pairs (three of them with a female offspring) housed separately in 10 m° pens, installed in an open outdoor area. The pens had a 1.7x0.7x0.35m tank full of water; a brickwork den of 1.0x0.75x 1.0 m with a mobile wood cover at the top, and a 0.30x0.30 m entrance near the floor closed by a mobile metal blind. Although living in captivity, these pacas showed nocturnal habits. Every morning around 9.00h, faeces and re- mainding food were removed, drinking water was changed, the water tanks filled, and I kg of hay was placed on the floor, 0.3m from the entrance of the artıfıcıal burrows. Each group was fedwith approxim- atly 240 g of rodent food at 9.00 a.m., and at 5.00 p.m. They received seasonal tropi- cal fruit, green maize, and chopped raw manioc for evening and night consump- tion. To supply their need for gnawing, three to four fresh pieces of eucalyptus 306 V. SABATINI et al. tus branchs, approximately 1.5 m in length and 5.0 cm in diameter, were provided to each pen, and exchanged for new ones after 15 days. The hay that remained out- side the burrows was swept out of the pens every morning. Every ten days, the whole pen floor was washed, even inside the bur- rows after the hay had been removed, and the water tanks were brushed. All mainte- nance of the animals was performed by the same two staff persons, who had already been doing this work for at least one year before the study started. Observations were registered by continuous recording (MARTIN and BATEsoN 1986), with all registry done in a descriptive manner by the same observer. Observations began in February 1998, and were conducted from 7.30 h a.m. to 3.00 h p. m. (daylight) for 40 consecutive days, for a total of 148 hours. Nocturnal observations were made from March to July of the same year, from 5.30 hp.m. to 10.00 hp.m., over scattered days for a total of 31 hours. The night sche- dule was selected, based on a previous study for 72 hours of continuous observa- tion of the activity rhythm. These pacas showed an activity peak from 5.30 hp.m. to 10.00 hp.m. In addition, the animals were observed for two more days, between 6.00 hp.m. to 12.00 ha.m. and 12.00 ha.m. to 6.00 h a.m. In order to acclimate the ani- mals to artificial light, two nights before each nocturnal observation two lamps (with 40 watt each, positioned 5 m equidistant) il- luminated the four pens. Observations in- side the burrows were possible since one corner of the wood cover was lifted 30 cm with a wire. Although defecation occurred mainly at night, caecotrophy was detected only once during the nocturnal observations, and this caecotrophy was of already defecated faeces. Ingestion of faeces directly from the anus was, in contrast, only observed during daytime, always occurring inside the burrows. The paca can rest in the bur- row using three different positions: with the belly upward and the four limbs flexed near the body; with the bodyside and cheek on the floor and the four limbs stretched perpendicular to the body; and, with the sternum on the floor and the limbs close to the body (i.e.: the sternal position). Caecotrophy occurred when the anımals were resting in the sternal posi- tion, by raising the chest off the ground, then putting the snout between the hind legs and repeatedly licking the anus; and fi- nally lıfting the head and chewing for about ten seconds, swallowing soon after. This cycle was repeated up to ten times. All adults and immature pacas over two months old showed this behaviour daily, throughout the diurnal observation period, however, one adult female performed cae- cotrophy during nocturnal observations. Consumption of faeces by captive pacas has previously been reported (MATAMOROS 1982), but no mention was made of caeco- trophy of differentiated faeces. Since pacas have large intestines with a functional caecum (BEntti 1981) and since they are phylogenetically related to hystri- cognaths who perform caecotrophy, the consumption of differentiated faces should be interpreted as related to the feeding ha- bits of pacas and not as an abnormal behav- iour resulting from captivity (GRIER 1984). Although studies concerning the natural feeding habits of this species are lacking, pacas have been considered to be frugi- vores. The occurrence of caecotrophy and their digestive tract anatomy suggests that pacas may be more herbivorous than ex- pected, often browsing on leaves, and not only when fruits are scarce. Acknowledgements We are greatful to Dr. MAurRICIO BARBANTI DUARTE for access to the pacas of the Wild Ani- mal Section, at the Universidade Estadual Paulis- ta-Jaboticabal, and to Dr. CARLOS RuUIZ-MIRANDA for devoted readings, English revisions and sug- gestions for the final manuscript. We also thank the anonymous reviewers for their helpful com- ments on the previous version of this manuscript. Financial support was supplied by the Conselho Nacional de Desenvolvimento Cientifico e Tecno- lögico-CNPgq, a Brasilian governmental institution for scientific and technological development. References AYRES, J. M.; LiMA, D. M.; MARTINS,E.S.; BAR- REIROS, J.L.K. (1991): On the track of the road: Changes in subsistence hunting in a Bra- zilian amazonian village. In: Neotropical Wildlife Use and Conservation. Ed. By J. G. Rogınson and K. H. REDFORD. Chicago: Univ. Press. Pp. 82-92. BENTTI, S. B. (1981): Las lapas, roedores de Amer- ica tropical. Natura 70/71, 40-44. BERGALLO, H. G.; DUARTE DA RocHA,C.F; AL- vES, M. A.S.; Van Suys, M. (2000): A fauna ameacada de extincäo do Estado do Rio de Janeiro. Rio de Janeiro. Editora da Univ. BJÖRNHAG, G.; SJÖBLOM, L. (1977): Demonstration of caecotrophy in some rodents. Swedish J. Agric. Res 7, 105-113. BoRrGESs, P. A.; DOMINGUEZ-BELLO, M. G.; HER- RERA, E. A. (1996): Digestive physiology of wild capybara. J. Comp. Physiol. 166, 55-60. CRANFORD, J. A.; JOHNSON, E. ©. (1989): Effects of coprophagy and diet quality on two microtine rodents (Microtus pennsylvanicus and Micro- tus pinetorum). J. Mammalogy 70, 494-502. Emmons, L.H. (1990): Neotropical Rainforest Mammals - a field guide. Chicago, London: Univ. of Chicago Press. Grier, J. W. (1984): Biology of Animal Behavior. Times Mirror/Mosby College Publ, Missouri: D. Bowen. HARCOURT, A. H.; STEWART, K.J. (1978): Copro- phagy by wild mountain gorillas. E. Af. Wildl. J. 16, 223-225. MAROUNEK, M.; VoOvK, S. J.; SKRIVANOVA, V. (1995): Distribution of activity of hydrolytic enzymes in the digestive tract of rabbits. British J. Nutr. 73, 463-469. MARTIN, P.; BATESon, P. (1986): Measuring Beha- viour — an introductory guide. Cambridge: Cambridge Univ. Press. MATAMOROS, Y. (1982): Notas sobre la biologia del tepezcuinte, Cuniculus paca, Brisson (Roden- tia: Dasyproctidae) en cautiverio. Brenezia 19/20, 1-82. MonpotLrI, E. (1972): La lapa o paca. Def. Nat. 2, 4-16. Caecotrophy in pacas 307 MoroT, M. CH. (1882): Des pelotes stomachales des leporides. Mem Soc. Centr. Med. Vet. 12, 139-239. Osawa, R.; BLANSHARD, W. H.; OCALLAGHAN, P. G. (1993): Microbiological studies of the intest- inal microflora of the koala, Phascolarctos ci- nereus. Pap, a special maternal faeces con- sumed by juvenile koalas. Austr. J. Zool. 41, 611-620. PICKARD, D. W.; STEvENS, C.E. (1972): Digesta flow through the rabbit large intestine. Am. ]. Physiol. 222, 1161-1166. SMYTHE, N.; GLANZ, W. E.; LEIGH JR., E. G. (1983): Population regulation in some terrestrial fru- givores. In: The Ecology of a Tropical Forest: Seasonal Rhythms and Long-Ierm Changes. Ed. By E.G.LeicH JR., A.A. RanD, and D. M. Wınpsor. Washington D.C.: Smithso- nian Inst. Press. Pp. 227-238. SOAVE, O.; BRAND, C.D. (1991): Coprophagy in animals — a review. Cornell Vet. 81, 357-364. STILLINGS, B. R.; HACKLER, L. R. (1966): Effect of coprophagy on protein utilization in the rat. J. Nutr. 90, 19-24. TAKAHASHI, T.; SAKAGUSHI, E. (1998): Behaviors and nutritional importance of coprophagy in captive adult and young nutrias (Myocastor coypus). J. Comp. Physiol. 168, 281-288. VICKERS, W. T. (1991): Hunting yields and game composition over ten years in a Amazon in- dian territory. In: Neotropical Wildlife Use and Conservation. Ed. By J. G. RoBınson and K. H. REDFORD. Chicago: Univ. Press. Pp. 53- 81. Woods, €. A. (1984): Hystricognath rodents. In: Or- ders and Families of Recent Mammals of the World. Ed. By S. AnDERSoNn and J. K. Jones Jr. Canada: John Wiley. Pp. 389-445. Authors’ addresses: VERA SABATINI, Laboratörio de Ci&ncias Ambien- tais/CBB, Universidade Estadual do Norte Flumi- nense/RJ. Av. Alberto Lamego, 2.000 Campos 28015-620, RJ, Brazil, and M.J. R. PARANHOoS - Da Costa, Departamento de Zootecnia, Univer- sıdade Estadual Paulista, Jaboticabal. 14884-900, SP, Brazil. Mamm. biol. 66 (2001) 308-311 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication x. Mammalian Biology Zeitschrift für Säugetierkunde Protein polymorphism in two species of Ctenomys (Rodentia, Ctenomyidae) from Cördoba province, Argentina By A. BORTOLU7ZI, MERCEDES GUTIERREZ, J. BALDO, and CRISTINA N. GARDENAL Museo de Zoologia, Facultad de Ciencias Exactas, Fisicas y Naturales and Cätedra de Quimica Biolögica, Facultad de Ciencias Medicas, Universidad Nacional de Cördoba, Cördoba, Argentina Receipt of Ms. 09. 10. 2000 Acceptance of Ms. 03. 05. 2001 Key words: Ctenomys bergi, Ctenomys rosendopascuali, allozymic polymorphism, speciation Fossorial rodents of the genus Crenomys are widespread in southern South America, from 17° to 54° S (CABRERA 1961; Reıc et al. 1990). The genus comprises 60 recognized species originated by an explosive specia- tion process, promoted mainly by chromo- somal rearrangements (BIDAu et al. 1996). At present, systematic relationships among species of Cienomys are poorly known and/ or controversial. Thomas (1902) cited the species C. bergi for the NW of Cördoba province, Argentina, being Cruz del Eje the type locality. On the basis of geographic criteria, all the po- pulations from the north of that province were included in that species (Bıpau et al. 1996). Chromosomal studies revealed that indivi- duals from the NE of Cördoba have a di- ploid number 2n=52 (FN = 66) but those proceeding from the NW (Salinas Grandes) presented a karyotype of 2n =48 (FN = 90) (REıG et al. 1990). This last form was as- signed to C. bergi and the former was de- scribed as a new species and denominated C. rosendopascuali (CONTRERAS 1995). Several authors have emphasized the im- portance of the application of biochemical and molecular methods in order to confirm and clarify the taxonomic status of different 1616-5047/01/66/05-308 $ 15.00/0. karyotypic forms of Ctenomys (BiDau et al. 1996; MASCHERETTI et al. 2000). The aim of this study is to analyze the allozymic poly- morphism in two populations of Ctenomys from the north of Cördoba, Argentina as- signed to C. bergi and C. rosendopascualı, in order to determine their level of differen- tiation at structural locı. Fourteen specimens of C. bergi from Las Toscas (30°11’S, 64°54° W, near an extense salt mine called Salinas Grandes) and 16 in- dividuals of C. rosendopascualis obtained in the proximity of the mouth of Xanaes river (Mar Chiquita saline lagoon, 30°55S 62°44’ W) were used in this study. Animals were killed by ether anesthesia, lı- ver and kidneys removed immediately and preserved at -30°C until used. Homoge- nates, vertical starch gel electrophoresis and staining procedures were carried out as de- scribed by GARDENAL et al. (1980) and GAR- DENAL and Branco (1985). The following enzymes were analyzed (loci scored and E.C. numbers in parenthesis): liver and kid- ney acid phosphatase (Acpr-1, Acpr-2, Acpx-3, Acpg-4; 3.1.3.2), aspartate amino- transferase (Aat-1, Aat-2; 2.6.1.1), liver so- luble esterases (Es-1} to Es-6,; 3.1.1.1), cat- alase (Cat; 1.11.1.6), phosphoglucomutase (Pgm-1, Pgm-2; 2.7.5.1), leucine aminopepti- dase (Lap-1, Lap-2; 3.4.11.1), malic enzyme (Me; 1.1.1.40), lactate dehydrogenase (Ldh; 1.1.1.27), alcohol dehydrogenase (Adh; 1.1.1.1), glycerophosphate dehydrogenase (Gpdh; 1.1.1.8), malate dehydrogenase (Mdh-1,Mdh-2; 1.1.1.37), isocitrate dehydro- genase (Idh-1, Idh-2; 1.1.1.42), 6-phospho- gluconate dehydrogenase (6-Pgdh; 1.1.1.43) and glucose-6-phosphate dehydrogenase (G-6-pdh; 1.1.1.49). The allele coding for the band migrating fastest to the anode was assıgned the num- ber 100; that controlling the fastest cathodic band, -100. The other alleles were num- bered according to their relative mobility from the origin. Bands with the same mobi- lity were considered homologous. Proportion of polymorphic loci (95% and 99% criteria), mean observed and expected heterozygosities, Rogers’ genetic distance (1972) and Nei’s identity (1975) among po- pulations were calculated using the program Biosys-1 (SWOFFORD and SELANDER 1989). Sixteen out of 27 loci analyzed were poly- morphic at least in one population. Table 1 shows allele frequencies, proportion of polymorphic loci (P), and observed and ex- pected mean heterozygosity per locus (H, and H.) for the two populations analyzed. Locus G-6-pdh was the only one presenting a different allele fixed in each population. Although crossing tests were not per- formed, the genetic control of the electro- phoretice patterns observed was postulated on the basis of similar polymorphisms de- scribed for other rodent species where the Mendelian transmission of variants has been demonstrated (GARDENAL and BLAN- co 1985; GARDENAL et al. 1980; GARcIA and GARDENAL 1989). In all cases, the observed genotypic frequencies did not differ signifi- cantly from the expected ones according to the Hardy-Weinberg equilibrium. Rogers’ genetic distance and similarity be- tween the two species was 0.094 and Nei’s distance and identity were 0.059 and 0.942, respectively. Levels of polymorphism revealed in this study for C. bergi and C. rosendopascuali are particularly high when compared with those reported for other subterranean Protein polymorphism in Ctenomys 309 mammals with low vagility and socially- structured mating system (NEvo et al. 1990). Values of heterozygosity obtained in thıs study are higher than the mean referred for fossorial rodents (H = 0.0311) and for Table 1. Allele frequencies, proportion of polymorphic loci (95% and 99% criteria) and observed and ex- pected heterozygocity in Ctenomys bergi and Ctenomys rosendopascuali from Cördoba province (Argentina). Locus Allele C.bergi C.rosendopascuali Lap-2 100 88 Acp«-1 100 90 81 Adh -100 -50 Gpdh 100 60 Acpı -3 100 78 Acpı -4 100 7 Aat-1 100 72 20 100 93 100 94 100 88 100 89 85 100 89 100 77 Pqm-2 100 82 Me 100 89 G 6pdh 100 87 P (95%) P (99%) Ho (%) He (%) 33.33 40.75 10.1 (s.e. 3) 9.3 40.74 48.15 12.8 (s.e. 3.4) 117 (s.e. 2.8) (s.e. 2.9) 310 A. BortoLuzz1 et al. several species of Ctenomys from Bolivia (Cook and YArtEs 1994), and Chile (GAL- LARDO and PArMA 1992), albeit similar to those obtained in 4 species from southern Brazil (H from 0.11 to 0.17) (MoREIRA et al. 1991). When rapidly evolving loci as es- terases are excluded, H. falls to 0.041 in C. bergi and to 0.067 ın C. rosendopascualı. However, they are still higher than the mean obtained for fossorial rodents, most of them calculated including esterases. SAGE et al. (1986) and ORTELLS and BARRANTES (1994) found lower levels of allozymic poly- morphism in other species of Ctenomys from Argentina. However, estimates were made, in most cases, on the basıs of 1 to 4 individuals, which could explain the results obtained by those authors. Genetic similarity between C. bergi and C. rosendopascuali is within the range re- ported for conspecific populations (KınG 1993). Notwithstanding, in locus G-6-pdh allele ‘100° is fixed in C. bergi and allele ‘87’ has a frequency of 1 in €. rosendopas- cuali, indicating lack of gene exchange be- tween the two forms. Several cases of interspecific homogeneity in allozymic frequencies have been re- ported in Cfenomys. GALLARDO and PALMA (1992) found very low levels of genetic dif- ferentiation among Ctenomys species from Chile, although being very dissimilar in morphological characters and karyotype. MOREIRA et al. (1991) reported an S value of 0.91 between C. minutus and Ctenomys sp. from southern Brazil, inhabiting regions separated by 75km and a wide river. The genus Crenomys is characterized by a large karyotypic heterogeneity, being one example of “explosive” speciation accom- panied by scarce morphological changes (Bıpau et al. 1996; Reıc et al. 1990). Fixa- tion of chromosomal re-arrangements would be favored by the population struc- ture characteristic of all species in the genus: small, semi-isolated groups with low vagıility and continuous extinction, expan- sion, and re-colonization in a variety of en- vironments (Reıc et al. 1990). The low ge- netic distance between C.bergi and C. rosendopascuali would be in agreement with the hypothesis of a rapid speciation by chromosomal re-arrangements, with al- most no differentiation at structural loci, as those coding for proteins. On the basis of morphological, morpho- metric, paleontological, karyological and distributional data, CoNTRERAS and BIDAU (1999) have proposed a hypothesis on the evolution of the complex genus Ctenomys. C. bergi would be closely related to the group designated “mendocinus”, which comprises several species with very similar karyotypes that have originated from a west-south radiatıon. C. rosendopascuali would integrate a separate lineage, the so called “oriental” group, presenting less stable diploid numbers and particular mor- phological features such as sperm asymme- try. However, MASCHERETTI et al. (2000), on the basis of cytochrome b sequences, found a very close relationship between these two species, placing them in the same molecular lineage. Our results would sup- port this last proposal. Acknowledgements We thank Dr. Antonio BLAnco for advice and critical revision of the manuscript. This work was supported by a grant from the Consejo de Investi- gaciones Cientificas y Tecnolögicas de la Provin- cia de Cördoba (CONICOR). C.N.G. is a Career Investigator of the Consejo Nacional de Investi- gaciones Cientificas y Tecnicas (CONICET) of Argentina. References BipAu, C. J.; GIMENEZ, M.D.; COoNTRERAS, J.R. (1996): Especiaciön cromosömica y la conser- vaciön de la variabilidad genetica: el caso del genero Ctenomys (Rodentia, Caviomorpha, Ctenomyidae). Mendeliana 12, 25-37. CABRERA, A. (1961): Catälogo de los mamiferos de America del Sur. Revista del Museo de Ciencias Naturales. Serie Ciencias Zoolögicas, Tomo 4, N° 2, pp. 551, Buenos Aires. CONTRERAS, J. R. (1995): Una nueva especie de tuco-tuco procedente de la llanura cordobesa nor-oriental, Repüblica Argentina (Rodentia, Ctenomyidae). Nötulas Faunisticas 86, 1-6. CONTRERAS, J. R.; BiDAU, €. J. (1999): Lineas gen- erales del panorama evolutivo de los roedores excavadores sudamericanos del genero Cite- nomys (Mammalia, Rodentia, Caviomorpha: Ctenomyidae). Ciencia Siglo XXI, N°1, pp: 1-22. Buenos Aires. Cook, J. A.; YATES, T. L. (1994): Systematic rela- tionships of the bolivian tuco-tucos, genus Ctenomys (Rodentia, Ctenomyidae). J. Mam- malogy 75, 583-598. GALLARDO, M. H.; PALMA, R. E. (1992): Intra-and interspecific genetic variability in Ctenomys (Rodentia: Ctenomyidae). Biochem. Syst. Ecol. 20, 523-534. GARCIA, B. A.; GARDENAL, C.N. (1989): Enzyme polymorphism and inheritance of allozymic variants in Calomys laucha. Com. Biol. 8, 1-10. GARDENAL, C. N.; BLANco, A. (1985): Polimorfis- mo enzimätico en Calomys musculinus: Nue- va estimaciön. Mendeliana 7, 3-12. GARDENAL, C.N.; SABATTINI, M.S.;, BLANCo, A. (1980): Enzyme polymorphism in a popula- tion of Calomys musculinus (Rodentia, Crice- tidae). Biochem. Genet. 18, 563-575. Kıng, M. (1993): Species Evolution, the Role of Chromosome Change. Cambridge: University Press. MASCHERETTI, S.; MIROL, P. M.;, GIMENEZ, M.D.: BIDAU, C. J.; CONTRERAS, J. R.; SEARLE, J. B. (2000): Phylogenetics of the speciose and chromosomally variable rodent genus Cie- nomys (Ctenomyidae, Octodontoidea), based on mitochondrial cytochrome b sequences. Biol. J. Linn. Soc. 70, 361-376. MOREIRA, D. M.; FRANCO, M. H.; FREITAS, T. R. O.; WEIMER, T. A. (1991): Biochemical poli- morphism and phenetic relationships in ro- dents of the genus Ctenomys from Southern Brazil. Biochem. Genet. 29, 601-615. NEvo, E.; FiLipuccı, M. G.; BEILES, A. (1990): Ge- netic diversity and its ecological correlates in nature: Comparisons between subterranean, fossorial and aboveground small mammals. Protein polymorphism in Ctenomys 311 In: Evolution of Subterranean Mammals at the Organismal and Molecular Levels. Ed. By E. Nevo and O. A. Reıc. New York: Alan R. Liss. Pp. 347-366. Neı, M. (1972). Genetic distance between popula- tions. Am. Nat. 106, 283-292. ORTELLS, M. O.; BARRANTES, G. E. (1994): A study of genetic distances and variability in several species of the genus Crenomys (Rodentia, Octodontidae) with special reference to a probable causal role of chromosomes in spe- ciation. Biol. J. Linn. Soc. 53, 189-208. Reıg, O. A.; BuscH, C.; ORTELLS, M. O.; CONTRERAS, J.R. (1990): An overview of evolution, sySs- tematics, population biology, cytogenetics, mMO- lecular biology, and speciation in Ctenomys. In: Evolution of Subterranean Mammals at the Or- ganısmal and Molecular Levels, Ed. By E. NEvo and O.A.Reıc. New York: Alan R. Liss. Pp. 71-96. SAGE, R.; CONTRERAS, J. R.; Roıc, V.; PATTON, ]. (1986): Genetic variation in the South Ameri- can burrowing rodents of the genus Ctenomys (Rodentia, Ctenomyidae). Z. Säugetierkunde 51, 158-172. SWOFFORD, D. L.; SELANDER, R.B. (1981): BIO- SYS-1: aFORTRAN program for the com- prehensive analysis of electrophoretic data in population genetics and systematics. J. Hered. 72, 281-283. Authors’ addresses: ANDRES BORTOLUZZI, MERCEDES GUTIERREZ, and JORGE BALDO, Museo de Zoologia, Facultad de Ciencias Exactas, Fisicas y Naturales, Universidad Nacional de Cördoba, Av. Velez Särsfield 299, 5000 Cördoba, Argentina. CRISTINA N. GARDENAL, Cätedra de Quimica Biolögica, Facultad de Cien- cias Medicas, Universidad Nacional de Cördoba, Casilla de Correo 35, Sucursal 16, 5016 Cördoba, Argentina (e-mail: ngardenal@biomed.fcm.unc.edu.ar). Mamm. biol. 66 (2001) 312-316 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication . Mammalian Biology Zeitschrift für Säugetierkunde Comparative food preference of Microtus brandti and Ochotona daurica in grasslands of Inner Mongolia, China By G. Wang, Q. ZHou, W. ZHong, and D. WAnG Institute of Zoology, Chinese Academy of Sciences, Beijing, China Receipt of Ms. 14. 12. 2000 Acceptance of Ms. 05. 04. 2001 Key words: Microtus brandti, Ochotona daurica, food selection, niche overlap Brandt’s voles (Microtus brandti) and da- hurian pikas (Ochotona daurica) are two common small mammal species in typical steppes of Inner Mongolia, China. The bur- rowing and foraging activities of the voles can impose strong impact on the composi- tion, physiognomy, and productivity of typi- cal steppes (ZHong et al. 1985a). In Inner Mongolian grasslands, the density of burrow entrances of M. brandti reached 5,616/ha in a high density year, and average above ground plant biomass in the area inhabited by voles was only 47% compared to areas without voles (ZHongG et al. 1985b). The voles and pikas are sympatric in this region. Trophic relationships are important to un- derstand the interspecific interactions be- tween these two coexisting species. How- ever, none of previous studies covered seasonal changes in the food habits of M. brandti and O. daurica and their trophic relationships. The objectives of this study were twofold: (1) to report species compoSi- tion of natural diets of M. brandti during spring, summer and autumn in typical steppes of Inner Mongolia, China, and nat- ural diet composition of O. daurica in sum- mer and autumn; (2) to determine the inter- specific trophic relationship between M. brandti and O. daurica. 1616-5047/01/66/05-312 $ 15.00/0. The study was conducted in Hexiten Ban- ner, Inner Mongolia, China. The study site (43°24 N, 116°46’ E) was located in a grass- land of flat topography. The vegetation is characterized as a Stipa krylovii, Artemisia frigida, Aneurolepidium chinense commu- nity. The average annual temperature is about -0.1°C. The average annual rainfall is about 350 mm, and is concentrated in June, July, and August. Snow cover is pre- sent from November to March. Plant growth occurs from April to August (JıanG 1985). We trapped M. brandti and O. daurica ın a 10-ha plot. Snap traps were set at burrow entrances of voles and pikas. Stomachs of captured voles and pikas were removed and preserved in 5% formalin solution. We analysed 22 stomachs of voles in April 1989, 23 in July 1989, and 20 in September 1990. Sample sizes for the pikas were 10 stomachs in July 1989, and 11 in September 1990. All sample sizes were greater than or equal to the minimum sample size for this kind of analysis (BarzLı 1985). Stomach contents were analysed following the proce- dures described by WırLLıams (1962). Refer- ence slides of epidermal layers were made for about 40 plant species of the study site. Epidermal fragments in stomach contents were identified to species whenever possi- Comparative food preference of M. brandti and O0. daurica ble. Percent of each plant species in diet dry weight was estimated following the proce- dure of Sparks and MALECHER (1968). We only listed main food items contributing >1% of diets (BAtzuı 1985). Within plot, above ground biomass was sampled with a 100x100 cm square frame in July, 1989. Ten random frames were cho- sen. Green plants were cut to the ground. Vegetation samples were sorted to species, and were dried in an oven at 60°C for 48 consecutive hours. Dried samples were weighed to the nearest 0.1 g. The total bio- mass of plants in a quadrat (g/m”) and per- cent biomass of each plant species in the to- tal biomass were recorded. We used the proportional similarity index (FEINSINGER et al. 1981), SI=),; min(D,, Dx), to calculate the trophic niche overlap between the two mammal species, where 313 min (D;, D.) is the minimum value between D; and D,, D; the proportion of plant species i in diet dry weight of species j, and D; the proportion of the same plant species in the diet of species k. We computed the jackknife means and varıances (ZAHL 1977) of SI for summer and fall, and then followed the t-test procedure of HUTCHESon (1970) to test for differences in the trophic niche overlaps be- tween the voles and pikas. We also used the proportional similiarity index to determine the sımilarity between the composition of summer diets of the voles and pikas and ve- getation. Similarity between diet and vegeta- tion composition measures diet selectivity of voles and pikas. Trophic niche width was de- termined by Shannon-Wiener diversity in- dex, H=-) D;ln(D;), where D; is the pro- portion of plant species i in a herbivore’s diet. We followed the t-test procedure of Table 1. Natural diets (diet dry weight percent %) of Microtus brandti in spring, summer, and autumn and Ocho- tona daurica in summer and autumn. Blank cells indicate either not used by the voles and pikas or < 1% of diets. Food items Summer Monocotyledons Aneurolepidium chinense Agropyron cristatum Stipa krylovii Carex duriuscula Keolena cristata Cleistogenes squarrosa Other monocotyledons Dicotyledons Astragalus galactites Scutellaria scordifolia Artemisia frigida Potentilla acaulıs Ixeris chinensis Saussurea amara Melissitus ruthenica Potentilla tanacetıfolia Salsola collina Heteropappus altaicus Potentilla bifurca Astragalus adsurgens Other dicotyledons Plant roots Unknown autumn summer autumn 314 G. Wang et al. HUTCHESoN (1970) to detect differences in the niche width between two species as well as between the seasons for the same species. We used preference index (PI = proportion of diet/proportion of forage) to assess if a herbivore responds to availability of a food item, e.g. PI>1 if consistently preferred, PI<1 if consistently avoided (BAtzuı 1985). In spring, the voles consumed seven main plant species, including four species of monocotyledons and three species of dico- tyledons. Monocotyledons made up 91.4% of diet dry weight, and dicotyledons 6.53%. The voles consumed nine main plant spe- cies in summer, four species of monocotyle- dons (49.3%), and five species of dicotyle- dons (47.2%) (Tab. 1). In summer, vole diet composition was different from plant species composition of the vegetation as the similarity index between the diet com- position and vegetation composition was 0.39. Of all available food items, the voles strongly preferred certain dicotyledons in summer (PI > 1.0), such as Potentilla tana- cetifolia, Heteropappus altaicus, and Melis- situs ruthenica (Tab. 2). Although Artemisia frigida contributed 6.5% of the summer diet, the voles did not prefer this plant (PI< 1.0). Autumn diets of voles consisted of 11 main plant species, four species of monocotyledons (33.4%) and seven species of dicotyledons (55.6%). A. chinense was the favorite food of voles in spring, summer, and autumn in terms of percentage. A. cris- fatum was less important during summer (4.7%) and autumn (2.83%) compared with spring (27.6%). However, Melissitus ruthe- nica and A. frigida became more important during summer and autumn. In summer, the pikas selected 11 main plant species, three species of monocotyledons (61%) and eight species of dicotyledons (37.8%). The pikas also showed preference for certain plants in summer, as the similar- ity index between the diet and the vegeta- tion composition was 0.33. The pikas pre- ferred Potentilla tanacetifolia, Astragalus galactites, Heteropappus altaicus, Melissitus ruthenica, and Ixeris chinensis (PI>1, Tab. 2). The pikas selected eight main plant species in autumn, including three species of monocotyledons (26.3%) and five spe- cies of dicotyledons (63.1%, Tab. 1). The diet dry weight percent of A. frigida in- creased from 0.5% in summer to 30% in autumn, while the percent of A. chinense declined from 56.4% in summer to 21% in autumn. Therefore, diets of the pikas had apparent seasonal changes. The overlap index of trophic niche between the voles and pikas was 0.54 in summer, and 0.64 in autumn, but did not differ between Table 2. Percent of main food items of the voles and pikas in the summer above ground biomass of vegetation and preference index (PI). PI > 1.0 indicates consistent preference, PI< 1.0 consistent avoidance. PI of pikas Pl of voles Food items % of vegeta- tion biomass Aneurolepidium 15.32 chinense Agropyron 4.41 cristatum 7.03 0.03 Stipa krylovii Potentilla tanacetiıfolia Heteropappus 0.23 altaicus Astragalus 0.03 galactıtes Pl of voles PI of pikas Food items % of vegeta- tion biomass Melissitus ruthenica Astragalus adsurgens Artemisia frigida Ixeris chinesis Potentilla bifurca Scutellaria scordifolia Comparative food preference of M. brandti and 0. daurica the two seasons (P>0.05). The trophic niche width of voles was 1.32 in spring, 1.69 in summer, and 1.92 in autumn. Like- wise the trophic niche width of pikas in- creased from summer (1.61) to autumn (1.93). The trophic niche width of the voles differed between spring and autumn (P< 0.05), but neither voles nor pikas had significantly different trophice niche widths between summer and autumn (P > 0.05). The voles selected more main food items in autumn than in spring, and the voles con- sumed more monocotyledons in spring (91.4%) than in autumn (37.2%) (Tab. 1). Selection for more food plant species and more even contributions of dicotyledons and monocotyledons in the autumn diet re- sulted in broader trophic niche of the voles in autumn than in spring. Although the standing crop biomass of the steppes of In- ner Mongolia reaches its highest in autumn (Li et al. 1988), the food quality of plants in autumn may be low. Mature plants in grasslands generally have higher fiber con- tent, decreased protein, and increased phe- nolic content that voles tend to avoid (Lin- DROTH et al. 1986; MaArauıs and BATZLI 1989). The reduction of food quality could lower the availability of food plants in au- tumn. Consequently, the voles expanded their trophic niche in autumn to respond to the low availability of food. The trophic niche width was not different between the voles and pikas either in summer or in au- tumn (P > 0.05). BERGMAN and Kress (1993) found that the overlap of the diets of collared lemming (Dicrostonyx kilangmiutak) and tundra voles (Microtus oeconomus) increased when both species foraged in the same ha- bitat. Overlap index of food utilization of voles and pikas under the food selection trial was 0.45 (computed from data of ZHONG et al. 1982 and ZHou et al. 1992), while the overlap of trophic niche in the free-ranging conditions was 0.54 in sum- mer. The voles and pikas had overlapping habitat use on our study site. The greater overlap under the free-ranging conditions might result from the lower availability of preferred food plants of pikas in natural 315 vegetation and higher percent of A. chi- nense in the pika’s natural summer diet. A. chinense made up 6% of daily food con- sumption in the food selection trial (Zmonc et al. 1982), but 56.4% of the nat- ural diet of pikas in summer. The lower availability of preferred P bifurca, A. bi- dentatum, and A. commutata might force the pikas to use the more abundant A. chi- nense, one of the dominant plant species in the plant community on the study site. The limited availability of preferred food items in natural vegetation may cause the voles and pikas to share more common and abundant plant species and may result in greater trophic niche overlap under the free-ranging condition. Acknowledgements We thank GUANGHE Wand for his assistance with the field work. Dr. BıLL McsHEA and anonymous reviewers made helpful comments on our manu- script. This work was financially supported by the Inner Mongolia Grassland Ecosystem Re- search Station, the Chinese Academy of Sciences. References BATZLI, G. ©. (1985): Nutrition. In: The Biology of New World Microtus. Ed. by R. H. TAMAR- In. Washington, DC: Am. Soc. Mammalogists, Spec. Publ. Vol. 8, 779-811. BERGMAN, C. M.; Kregs, C. J. (1993): Diet overlap of collared lemmings and tundra voles at Pearce Point, Northwest Territories. Can. ). Zool. 71, 1703-1709. FEINSINGER, P.; SPEARS, E. U.; PooLE, R. W. (1981): A simple measure of niche breadth. Ecology 62, 27-32. Hutcheson, K. (1970): A test for comparing diver- sities based on the Shannon formula. J. Theor. Biol. 29, 151-154. JıanG, S. (1985): An introduction to the Inner Mongolia Grassland Ecosystem Research Sta- tion, Academia Sinica. In: Research on Grass- land Ecosystem. Vol. 1. Ed. by Inner Mongo- lia Grassland Ecosystem Research Station. Beijing: Science Press. Pp. 1-11. Lı, B.; Yong, S.; Liu, Z. (1988): The vegetation of the Xilin River basin and its utilization. In: Research on Grassland Ecosystem. Vol. 3. 316 G. Wang et al. Ed. by Inner Mongolia Grassland Ecosystem Research Station. Beijing: Science Press. Pp. 84-183. LINDROTH, R.L.; BATZLI, G. O.; SEIGLER, D. S. (1986): Patterns in the phytocheminstry of three prairie plants. Biochem. Syst. Ecol. 14, 597-602. MARauIıs, R. J.; BATZL1, G. ©. (1989): Influence of chemical factors on palatability of forage to voles. J. Mammalogy 70, 503-511. SPARKS, D. R; MALECHEK, J. C. (1968): Estimating percentage dry weight in diets using a micro- scopice technique. J. Range Manage. 21, 264- 265. WILLIAMS, O. (1962): A technique for studying mi- crotine food habits. J, Mammalogy 43, 365- 368. ZAHL, S. (1977): Jackknifing an index of diversity. Ecology 58, 907-913. ZHONG, W.; ZHOU, Q.; Sun, C. (1982): Studies on foods and daily food consumption of Ochoto- na daurica. Acta Ecol. Sın. 3, 269-276. ZHONG, W.; ZHOU, Q.; Sun, C. (1985 a): The basic characteristics of the rodent pests on the pas- ture in Inner Mongolia and its ecological stra- tegies of controlling. Acta Theriol. Sin. 5, 242-249. ZHONG, W.; ZHou, O.; Sun, C. (1985b): Habitat selection of Microtus brandti and its relation- ships with vegetation conditions. In: Research on Grassland Ecosystem. Vol. 1. Ed. by Inner Mongolia Grassland Ecosystem Research Sta- tion. Beijing: Science Press. Pp. 147-153. ZHoU, Q.; ZHONG, W.; Sun, Q. (1992): Food habits and daily consumption of Microtus brandti. In: Research on Grassland Ecosystem. Vol. 4. Ed. by Inner Mongolia Grassland Ecosystem Research Station. Beijing: Science Press. Pp. 61-68. Authors’ addresses: GUIMING WanG, Department of Biological Sciences, Arkansas Tech University, Russelville, AR 72801, U.S. A.; QINGQIANG ZHOU, WENQIN ZHONG, and DEHUA Wang, Institute of Zoology, the Chinese Academy of Sciences, Beijing 100080, China. Mamm. biol. 66 (2001) 317-319 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication ER, Zu: STERKUND“ Mammalian Biology Zeitschrift für Säugetierkunde Records of a few rare mammals from northeastern Peru By CHRISTINE L. HICE Department of Biological Sciences, Texas Tech University, Lubbock, Texas, U.S.A. Receipt of Ms. 06. 12. 2000 Acceptance of Ms. 15. 02. 2001 Key words: Didelphimorphia, Rodentia, Peru, distribution records During 18 months (July 1997 to December 1998) small-mammals were collected near Iquitos, Peru, for certain research projects. Sampling was conducted at the Estaciön Biolögica Allpahuayo (S 3°58; W 73°25'), a 3000 hectare field station operated by the Instituto de Investigaciones de la Amazonia Peruana (IIAP), 25km south of Iquitos, Department of Loreto, in northeastern Peru. The climate is tropical with a mean annual temperature of 26°C; the highest average monthly temperature (31°C) oc- curs in November and the lowest (22°C) in July (Sararı 1985). Average rainfall is 2945 mm per year, with a slightly drier sea- son from June to September (JOHNSON 1976). The elevation of the station ranges from 110 m to 180 m above sea level. Over 1000 mammals were collected and prepared. Concerning our knowledge on their distribution several of these specimens represent substantial range extensions for a few species, whereas others are records of mammals not frequently recorded in mam- malian surveys. The following accounts summarize information about the species of this region not yet recorded as well as what is already known about the species distributions (Emmons 1997; EISENBERG and REDFORD 1999). Any measurements (mm) are given using the standard sequence of to- 1616-5047/01/66/05-317 $ 15.00/0. tal length, taıl length, hind foot length, and ear length. Specimens are deposited at the Museum of Texas Tech University, Lub- bock, Texas, and the Museo de Historia Natural de la Universidad Nacional Mayor de San Marcos, Lima, Peru. Preserved tis- sues (frozen) include heart, kidney, liver, spleen, lung, and muscle and are deposited at the Museum of Texas Tech University. Philander opossum (Linnaeus, 1758) and Philander andersoni (Osgood, 1913) The occurrence of these two species of opossum in the study area represents the first record of sympatry, as well as the first record of P. opossum north of the Amazon River this far east of the Andes (EMMoNS 1997). FLECK and HARDNER (1995) also re- ported sympatry of these species in Jenerro Herrera, south of the Amazon River. How- ever Dre), PEPATIONGof thezUniversity, of California, Berkeley, examined photo- graphs of the animals and determined that the P. andersoni had been misidentified and were in fact P.mcilhennyi (pers. comm.). The species identity of my specimens has been verified by cytochrome b sequence ana- lysis conducted by Dr. Parton. All 39 P. opossum were captured in disturbed habi- tat whereas 12 of 14 P. andersoni were cap- 318 CHRISTINE L. HICE tured in mature forests. P. andersoni was cap- tured throughout the year; nursing females (n=2) were obtained in April and October. The individual captured in April had two pouched young. P.opossum was only cap- tured from July to January, with no captures in the spring. Nursing individuals (n= 11), with an average litter size of 4 (2-5), were obtained in July, August, and November. Gracilinanus kalinowskii Hershkovitz, 1992 This tiny marsupial is extremely rare, known only from seven individuals, three from southern Peru and four from the Guyanan region (HERSHKOVITZ 1992; R.S. Voss, pers. comm.). One young female was captured on 20. 05. 1998 in a Victor rat trap baited with dried, salted fish. The trap was located on the ground in monte alto forest about a 3 h walk north from the road south- west of Iquitos. The animal weighed 4 g and measured 132-76-10-10. The identity of this specimen has been confirmed by R. S. Voss (pers. comm.). This capture represents a substantial range increase and suggests the distribution of this rare species may be quite broad within the Amazon Basin, from southern Peru to the Guyanas. Monodelphis adusta (Thomas, 1897) This species is currently known only from Pa- nama and the eastern slopes of the northern Andes in Peru, Colombia, and Ecuador (Enm- MoNS 1997), with a disjunct population in Madre de Dios in southern Peru (WoODMAN et al. 1995). Six individuals were captured in Allpahuayo. Females, with an average weight of 13 g (12-14 g), were substantially smaller than males, with a mean weight of 29 g (28- 30 g). Average measurements for females were 130-40.5-12.5-11 (130/130-40/41-12/13- 11/11) and for males 164-55-15-12 (157/172- 52/57-15/16-11/12). Five of the animals were captured in pitfall traps with a drift fence, as described by Voss and Emmons (1996). The remaining animal was taken in a Sherman trap baited with a peanut butter/pork fat mixture. They were captured in each of the three types of primary forest present in All- pahuayo (VAsQUEZ MARTINEZ 1997). No re- productive activity was detected for animals captured in the months of November, De- cember, March (both females), May, and Au- gust. This represents the first record of any species of Monodelphis north of the Amazon River in the Iquitos area and a substantial range increase of M. adusta. Scolomys melanops Anthony, 1924 This genus is known from less than 50 indi- viduals and is suggested to be highly loca- lized in its distribution (PATTON and DA SILvA 1995). It is not necessarily unexpected in the Iquitos area, and may be more widespread in the Amazon Basin than is presently re- corded. A large series (24 individuals) was taken during 8 months, but not those of the dry season (June, July, August, and Septem- ber), and in each of the three types of pri- mary forest present in Allpahuayo (VAs- QUEZ MARTINEZ 1997). Eleven individuals were captured in pitfall traps and 9 were taken in Victor traps. Two were taken on fallen logs approximately Im high. A total of 5 females and 19 males was obtained, in- cluding 2 pregnant females with an average litter size of 2.5 (one each in March and April). Males in reproductive condition (average testes size of 3x 6 mm) were cap- tured in March, October, and November. This represents the largest series of $. mela- nops available (PATTon et al. 2000). Galictis vitata (Schreber, 1776) This species is broadly distributed through- out the Amazon Basin, but uncommon in its range (EMMons, 1997). One adult male grison was brought to me by a local hunter on 14. 10. 1997. The animal was shot in up- land monte alto forest. No measurements are available. Acknowledgements I thank many people and institutions for making this research possible. Dr. VICTOR MONTRUEIL with the Instituto de Investiga- Records of a few rare mammals from northeastern Peru ciones de la Amazonia Peruana granted permission to conduct research at the biol- ogical station. Dr. DAvID SCHMIDLY pro- vided comments on an earlier draft that im- proved the manuscript. INRENA supplied necessary collecting permits (#21-98-INRE- NA-DGANPFS-DCFS). This work was supported in part by research grant Al- 39800 from the National Institutes of Health. References EISENBERG, J.F.; REDFORD, K.H. (1999): Mam- mals of the Neotropics. Vol. 3. The central Neotropies: Ecuador, Peru, Bolivia, Brazil. Chicago: Univ. Chicago Press. Emmons, L.H. (1997): Neotropical rainforest mammals, a field guide. 2nd ed. Chicago: Univ. Chicago Press. FLEcK, D. W.; HARDNER, J. D. (1995): Ecology of marsupials in two Amazonian rain forests in northeastern Peru. J., Mammalogy 76, 809- 818. HERSHKOVITZ, P. (1992): The South American gra- cile mouse opossums, genus Gracilinanus Gardner and Creighton, 1989 (Marmosidae, Marsupialia): a taxonomic review with notes on general morphology and relationships. Fieldiana Zool., new series 70, 1-56. JoHNsonN, A.M. (1976): The climate of Peru, Boli- via, and Ecuador. In: Climates of Central and South America. World Survey of Climatology. Vol. 12 Ed. by W. SCHWERDTFEGER. Amster- dam: Elsevier. Pp. 147-210. 319 PATToN, J. L.; DA Sııva, M.N.F. (1995): A review of the spiny mouse genus Scolomys (Roden- tia: Muridae: Sigmodontinae) with the de- scription of a new species from the western Amazon of Brazil. Proc. Biol. Soc. Wash. 108, 319-337. PATTON, J. L.; DA SILVA, M.N.F.; MALCOLM, J.R. (2000): Mammals of the Rio Juruä and the evolutionary and ecological diversification of Amazonia. Bull. Am. Mus. Nat. Hist. 244, 1- 306. SALATI, E. (1985): The climatology and hydrology of Amazonia. In: Amazonia. Key Environ- ments Series. Ed. by G. T. PrAnce and T.E. Lovzsoy. Oxford, UK: Pergamon Press. Pp. 18-48. VASQUEZ MARTINEZ, R. (1997): Flörula de las re- servas biolögicas de Iquitos, Perü: Allphuayo- Mishana, Explornapo Canmp, Explorama Lodge. Mon. Sys. Bot. Missouri Botanical Garden. Vol. 63. St. Louis: The Missouri Bo- tanical Garden Press. Voss, R. S.; Emmons, L. H. (1996): Mammalian di- versity in Neotropical lowland rainforests: a preliminary assessment. Bull. Am. Mus. Nat. Hist. 230, 1-115. WooDMAN, N.; SLADE, N.A.; Timm, R.M,; SCHMIDT, C. A. (1995): Mammalian commu- nity structure in lowland, tropical Peru, as de- termined by removal trapping. Zool. J. Linn. Soc. 113, 1-20. Author’s address: CHRISTINE L. Hıce, Department of Biological Sciences, MS 3131, Texas Tech University, Lub- bock, TX 79409, USA (e-mail: chhice@ttacs.ttu.edu). Mamm. biol. 66 (2001) 320 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Book review HUME, I. D.: Marsupial Nutrition. Cambridge: Cambridge University Press 1999. 434 pp., 93 line- diagrams, 39 half-tones. Paperback. Us$ 54.95. ISBN 0 521 59555 X. The reader will hopefully allow the reviewer to do some calculations first before starting with the re- view, although this might not be a strictly “profes- sional” approach: The present book, as a sequel to “Digestive Physiology and Nutrition of Marsu- pials” by the same author, which was published in 1982, increased considering the number of pages by approximately 70% and the number of papers cited in the references increased by appro- ximately 90%! As the author himself writes in the preface, this remarkable expansion of knowledge during the last two decades is due to intensified research efforts that have gone into marsupials other than kangaroos. Additionally, authors deal- ing with South American marsupials have contri- buted considerably to information on marsupial nutrition, which are meticulously presented here to the impressed reader. In a detailed introductory chapter the author deals with general physiological and nutritional aspects under the title “Metabolic rates and nu- trient requirements”. In this important section the frame for all consecutive chapters is sup- plied: The concept of nutritional niches, signifi- cance of metabolic rates, energy requirements for maintenance, aspects of food intake, torpor and hibernation as it can be found in some mar- supials, as well as requirements of, e.g., water and protein. The following chapter deals with carnivorous mar- supials, such as American caenolestids, some di- delphid species and the only representative of the microbiotheriids, Dromiciops australis, or coloco- los, as well as Australasian Dasyuridae. In the next 1616-5047/01/66/05-320 $ 15.00/0. Mammalian Biology Zeitschrift für Säugetierkunde chapter HuME presents omnivorous marsupials, such as the American Didelphidae and Australa- sian Peramelidae, Peroryctidae, Burramyidae, Pe- tauridae, and Acrobatidae. Two chapters deal with hindgut fermenters; one with wombats and the sec- ond with arboreal folivores, such as tree kanga- roos, the koala (Phascolarctidae) and phalangers as well as pseudocheirids (ring-tails). The follow- ing three chapters present foregut-fermenters, such as the Macropodidae (kangaroos and walla- bies) and Potoroidae (rat kangaroos). All these chapters are clearly organised, well-written, in- structively illustrated (diagrams and half-tones) and full of biological information! Having been informed about the diversity of the digestive tract, its function and about nutritional and physiological aspects in the Marsupialia, the biologically interested reader expects information and comments on how this richness in species and differentiations came into existence. The reader is not disappointed. HUME presents outlines of the Gondwanian origins of American and Australa- sian marsupials and discusses their likely foraging and digestive differentiation. He makes clear that his presentation in this part of the book is specu- lative. After a relatively short “Future directions”, an appendix compiles the classification of marsupi- als. When browsing through this list, serious de- viations from the modern taxonomic reference published by Wırson and REEDER (1993) in asso- ciation with the American Society of Mammal- ogists, could not be detected by the present re- viewer. 51(!) pages of references show the remarkable number of publications, on which the book of I. D. Hume is based. An index of 17 pages makes the information presented in this book readily accessible. P. LANGER, Giessen Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana-ging editor, Prof. Dr. D. Kruska, Institut für Haustierkunde, Christian- Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni-kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manuscript will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Copyright: Submission of a manuscript implies that the submitted work has not been published before (except as part of a thesis or lecture note or report, or in the form of an abstract ); that it is not under consideration for publication elsewhere; that its publication has been approved by all co-authors as well as by the authorities at the institute where the work has been carried out; that written permission of copyright holders was obtained by the authors for material used from other copyrighted sources; that, if and when the manuscript is accepted for publication, the authors hand over the transferable copyrights of the accepted manuscript to the publisher; and that the manuscript or parts thereof will thus not be published elsewhere in any language without the consent of the copyright holder. Copyrights include, without spatial or timely limitation, the mechanical, electronic and visual reproduction and distribution; electronic storage and retrieval; and all other forms of electronic publication or any other types of publication including all subsidiary rights. Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21x 30 cm) or correspondingly either from a typewriter or legible computer print with a left- hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscnpt. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscript following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type-written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed Mammalian Biology Zeitschrift für Säugetierkunde with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscnipt. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings depending on the theme, Acknowledgements, Zusammenfassung (in German), References. A German title must precede the Zusammen- fassung in English manuscripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive Sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: KaLko, E. K. V.; Hanotev, C. O. Jr. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1821), from Panama. Z. Saugetierkunde 59, 257-273. b) From books: HERRE, W.; Röhrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/I, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers” symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscnipt is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen, Germany (email:peter.langer@anatomie.med.uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). URBAN & FISCHER Verlag Mammalian Biology Jena/Germany ER ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. - 66(2001)5 - 257-320 Contents Original investigations Wöhrmann-Repenning, A.; Bergmann, M.: The vomeronasal complex in strepsirhine primates and Tarsius - Der Vomeronasalkomplex bei strepsirhinen Pramates und orsus _ 2 000 Steffen, K.; Kruska, D.; Tiedemann, R.: Postnatal brain size decrease, visual performance, learning, and discrimination ability of juvenile and adult American mink (Mustela vison: Carnivora: Mammalia) - Postnatale Hirngrößenabnahme, visuelle Leistung, Lernen und Unterscheidungsvermögen von juvenilen und adulten amerikanischen Minken (Mustela vison: Carnivora: Mammalia) = 0 8 Kittlein, M. J.; Vassallo, A. I.; Busch, C.: Differential predation upon sex and age classes of tuco-tucos (Ctenomys talarum, Rodentia: Octodontidae) by owls - Alters- und geschlechtsbedingte Unterschiede bei Prädation von Tuco-Tucos (Ctenomys talarıum, Rodentia: Octodontidae) durch udien 0) Glitzner, I.; Glossow, H.: Kleinsäuger auf forstwirtschaftlich unterschiedlich behandelten Windwurfflächen eines Bergwaldes - Small mammal distribution on differently managed storm areas of a mountain forest — _______ 290 Short communications Endepols, S.; Dietze, H.; Endepols, H.: The occurrence of roof rats (Rattus rattus L., 1758) in Germany during the late 20" century - Vorkommen der Hausratte (Rattus rattus L., 1758) in Deutschland am Ende des 20. Jahrhunderts _ 7301 Sabatini, V.; Paranhos de Costa, M. J. R.: Caecotrophy in pacas (Agouti paca Linnaeus, 1766) - Caecotrophie bei Pakas (Agouti paca Linnaeus, 1766) _ 305 Bortoluzzi, A.; Gutierrez, M.; Baldo, J.; Gardenal, C.N.: Protein polymorphism in two species of Ctenomys (Rodentia, Ctenomyidae) from Cördoba province, Argentina - Proteinpolymorphismus bei zwei Ctenomys-Arten (Rodentia, Ctenomyidae) aus der Provinz Cördoba, Äfgentinien 0. 0 00000 u ae Wang, G.; Zhou, Q.; Wang, D.: Comparative food preference of Microtus brandti and Ochotona daurica in grassland of Inner Mongolia, China - Vergleichende Nahrungspräferenz von Microtus brandti und Ochotona daurica im Grasland der Inneren Mongolei, China — > 2 0, en nee Hıce, Ch L.: Records of a few rare mammals from northeastern Peru - Nachweis von einigen seltenen Säugetieren aus dem nordöstlichen Peru 2.2... = 2. 2. ner nn ea Book review — 1) ua nen eu. A DL Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record Mäimmalian Biology Zeitschrift für Säugetierkunde 06-2001 AN 0920 , ISSN 1616-5047 Mamm. biol. - 66(2001)6 - 321-384 www.urbanfischer.de/journals/mammbiol Mammalian Biology Zeitschrift für Säugetierkunde Editor Deutsche Gesellschaft für Saugetierkunde (German Society of Mammalogy) Editorial Office D. Kruska, Kiel - P. Langer, Giessen Advisory Board S. Halle, Jena G. B. Hartl, Kiel R. Hutterer, Bonn E. Kalko, Ulm P. Lüps, Bern W. Maier, Tübingen H. Schliemann, Hamburg S. Steinlechner, Hannover G. Storch, Frankfurt F. Suchentrunk, Wien F. Trillmich, Bielefeld Deutsche Gesellschaft für Säugetierkunde Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) H. Frick, München (1972-1976) M. Röhrs, Hannover (1977-1981) H.-J. Kuhn, Göttingen (1982-1986) E. Kulzer, Tübingen (1987-1991) U. Schmidt, Bonn (1992-1996) H. G. Erkert, Tübingen (1997-2000) Managing Committee President R. Schröpfer, Osnabrück Board Members H. Frädrich, Berlin J. Ganzhorn, Hamburg G.B. Hartl, Kiel D. Kruska, Kiel P. Lüps, Bern A. Wöhrmann-Repenning, Kassel IMPRESSUM Publisher: Urban & Fischer Verlag GmbH & Co. KG, Branch Office Jena, P.O. Box 100537, D-07705 Jena, Germany; Phone: ++49(0)3641/626-3, Fax: ++49(0)3641/62 65 00, E-mail: journals@urbanfischer.de Advertisement bookings: Enquiries should be directed to Urban & Fischer Verlag, Advertising Manager: Sabine Schröter, P.0. Box 100 537, D-07705 Jena, Germany; Phone: ++49(0)3641/62 64 45, Fax: ++49(0)3641/62 64 21. Advertising rates: The price list from January 1, 2001 is effective at present. Distribution and subscription agency: For customers in the German speaking countries: Please turn to your bookshop or to Urban & Fischer Verlag, Aboservice und Vertrieb: Barbara Dressler, Löbdergraben 14a, D-07743 Jena, Germany; Phone: ++49(0)3641/62 64 44, Fax: ++49(0)3641/62 64 43, E-mail: k.ernst@urbanfischer.de For customers in all other countries: Please contact your bookshop or The Nature Publishing Group, Linda Still, Brunel Road, Basingstoke, RG21 6XS, UK; Phone: ++44(0)1256/30 26 29, Fax: ++44(0)1256/47 61 17, E-mail: .still@nature.com Subscribers in the USA can call toll free on ++1 800 747 3187. Terms of delivery (2001): 1 volume consisting of 6 issues. Subscription rate (2001): DM 498.00*/GBP 200.00. Preference price for personal subscribers: DM 150.00*/GBP 60.00. All prices quoted above are exclusive postage. *Suggested list price. Price are subject to change without notice. Personal subscription orders must be addressed directly to the publisher Urban & Fischer in Jena resp. to The Nature Publishing Group. Subscriptions at a personal rate must include the recipients name and mailing address. An institutional or business address is only accepted, if the institute or firm has a subscription as well. Personal subscribers can pay by personal cheque or credit cardmentioned below. US dollar prices are converted at the current exchange rate. We accept the following creditcards: VISA/Eurocard/Master- card/American Express. (Please note card-No. and expiry date.) Banking connections: For customers in the German speaking countries: Postbank Leipzig, Konto-Nr. 149 249 903 (BLZ 860 100 90). Deutsche Bank Jena, Konto-Nr. 390 76 56 (BLZ 82070000); For customers in the other countries: UK Post Office Giro Account No. 519 2455. Full details must accompany the payment. Subscription information: If not terminated the subscription will be effected until recalled. If no cancellation is given until October, 31st the delivery of the journal will be continued. Copyright: The articles published in this journal are protected by copyright. All rights are reserved. No part of the journal may be reproduced in any form without the written permission of the publisher. This includes digitalisation and any further electronic computing, like saving, copying, printing or electronic transmission of digitalised material from this journal (online or offline). Authorization to photocopy items for internal or personal use of specific clients, is granted by Urban & Fischer, for libraries and other users registered with the Copyright Clearance Center (CCC) Transaction Reporting Service, provided that the base fee of $ 15.00 per copy is paid directly to CCC, 222 Rosewood Drive, Danvers, MA 01923, USA, 1616-5047/ 01/66 $ 15.00/0. This consent does not extend to copying for general distribution, for advertising or promotional purposes, for creating new collective works or for resale and other enquiries. In such cases, specific written permission must be obtained from the publisher Urban & Fischer. Type setting and printing, binding: druckhaus köthen GmbH As of vol. 61, number 1 (1996) the paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). ©2001 Urban & Fischer Verlag For detailed journal information see our home page: http://www.urbanfischer.de/journals/mammbiol Printed in Germany Mamm. biol. 66 (2001) 321-331 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol „ Mammalian Biology Zeitschrift für Säugetierkunde Original investigation Recent recovery of the Italian wolf population: a genetic investigation using microsatellites By M. SCANDURA, M. APOLLONIO and L. MATTIOLI Department of Zoology and Anthropology, University of Sassari, Sassari, Italy and Wildlife Management Unit, Provincial Administration of Arezzo, Arezzo, Italy Receipt of Ms. 20. 09. 2000 Acceptance of Ms. 31. 01. 2001 Abstract Genetic differentiation within the Italian wolf population was investigated by microsatellite analy- sis of 38 individuals from 4 distinct sampling sites of the current wolf range throughout the penin- sula. A set of 6 microsatellite loci was used, which showed a high level of polymorphism and a com- bined probability of identity ranging from 10°“ to 10°°. The overall DNA variability detected was considerable and only slightly lower than that found for North American grey wolves. The two largest Italian subpopulations taken into consideration, Tuscan Apennines and central- southern Apennines, proved moderately divergent, and data are consistent with a derivation of the western Alps subpopulation from the former, while the latter showed close similarity to the western coast subpopulation. Gene flow was relatively high across the Italian population and the presence of isolation by distance was supported by our data, as measures of genetic distance were consistent with geographical distribution of sampling sites. High levels of divergence were found between Italian and other European samples. These findings suggest that, despite their absolute mtDNA monomorphism, Italian wolves have preserved a high nuclear DNA heterogeneity and a well-defined genetic identity. A further enlargement of range, which can be expected on the basis of extensive wolf dispersal, might cancel their historical isolation in a few decades, thus favouring a genetic exchange with the east European gene pool. Key words: Canis lupus, microsatellites, variability, population structure, Italy Introduction Since the end of the 19'® century large pre- dator populations have declined in Italy due to progressive habitat disruption and to direct persecution by humans. As a con- sequence, different species approached ex- tinction (i.e., brown bear, wolf, and Iynx). Changes in human activities, wildlife and wood management, and public opinion 1616-5047/01/66/06-321 $ 15.00/0. have led to the restoration of more favour- able environmental conditions and to in- creased protection of these species. Due to such improvements, in the last 30 years an important predator species, the wolf (Canis lupus), has increased in number and en- larged its range (Francıscı and GUBERTI 1993; Bormanı and Cıuccı 1993; MERIGGI 322 M. ScanpurA et al. and LovArı 1996). Wolves in the Italian en- vironment play a key role in wild commu- nities, being the only well-distributed large mammals preying mostly on wild ungulates (Martıoui et al. 1995). During the last two centuries the Italian population became iso- lated from other European populations, due to the extirpation of the species throughout the Alps (CAGnNoLARO et al. 1974). The northern border of its range initially moved southwards, towards the central regions, and the presence was restricted to the less accessible areas of the Apennines and to a wooded area along the Tyrrhenian coast (CAGNOLARO et al. 1974; ZIMEN and BOITANI 1975). In 1973, the number of wolves inha- biting the Italian peninsula was estimated to be approximately 100 individuals (ZIMEN and Bortanı 1975), after which the popula- tion recovered, reaching an estimated size of 400-500 individuals (BoITAnı 1992). The demographic recovery of the species was accompanied by a northward expansıon of its range, which during the last ten years led to the recolonization of the western Alps, and to the consequent appearance of the wolf in France (BREITENMOSER 1998). The effects of population decline and the subsequent range expansion in the genetic diversity of the Italian wolf population were studied by allozyme and mitochondrial DNA (mtDNA) analyses (RanDI et al. 1993; Wayne et al. 1992; Ranoiı et al. 1995; VırA et al. 1999; RanDiı et al. 2000). RANDI and co-workers (1993), using a set of 40 al- lozymes over a sample of 38 wolves, found a level of polymorphism and heterozygosity comparable to that of larger North Ameri- can populations (KEnNEDyY et al. 1991). On the other hand, mtDNA consensus_ se- quences revealed the presence of a single haplotype in all the sampled Italian wolves. This apparent contradiction probably re- sults from the different inheritance systems, based exclusively on female mtDNA trans- mission. In order to investigate the actual effects of the population bottleneck and fragmenta- tion on the genetic structure and variability of wolves, nuclear genetic markers, e.g. mi- crosatellites, are more informative. Differ- ent studies carried out on North American wolf populations demonstrated the effec- tiveness of microsatellite loci as a molecular tool for assessing population structure para- meters (Roy et al. 1994; ForBEs and BoyD 1997), genotyping animals for reintroduc- tion programmes (GARCIA-MORENO et al. 1996), evaluating relatedness among indivi- duals (SmitH et al. 1996), and estimating ge- netic variation following natural coloniza- tions (FORBES and BoyD 1996). The aims of the present study were: 1) to reconstruct the dynamics of the Italian wolf recovery by comparing the genetic pattern of different subpopulations; 2) to evaluate nuclear DNA diversity among and within sub-populations; 3) to estimate the degree of gene flow among different areas. Wolves from historical “stronghold areas” and from recent colonized regions were sampled for comparison. Material and methods Sample description and collection Italian wolf samples were collected from four re- gions (Fig. 1): AR - Tuscan Apennines, which probably repre- sented the northern border of the Italian wolf range along the Apennines for half a century (CAGNOLARO et al. 1974); AB - Central-southern Apennines, the part of Italy where the species has always been present in historic times; VC - Alta Maremma, where the presence of the species was always recorded during the last cen- tury, but the area was characterized by continuous new settlements of breeding packs followed by complete or partial eradication (illegal killing); FR - Alpes Maritimes, France, originating in the early 1990s from individuals moving across from Italy (Ranpı et al. 2000). Its present size is ap- proximately 20-30 units, of which dispersing indi- viduals are colonizing new areas of the western Alps. Seventeen tissue samples came from illegally killed wolves, recovered before 1992 in central and southern Italy and obtained from Prof. G. B. HARTL (University of Kiel, Germany). Thirteen specimens are from the northern Apennines, mostly the province of Arezzo: 12 (3 tissues and 9 hairs), supplied by the Provincial Administra- Recent recovery of the Italian wolf population 323 Fig. 1. Geographical location of sampling sites (solid grey) and present distribution range of wolves (hatching). tion of Arezzo (Tuscany) or by the Corpo Fore- stale dello Stato, derived from animals dying in the period 1991-1999, whereas one hair sample was collected in the field during a survey in the Foreste Casentinesi National Park. Three samples of the Alta Maremma population were obtained from the Veterinary Service of Volterra (Pisa) and one hair sample was collected in a natural preserve near Volterra. For the Alpine popula- tion, 4 tissue samples were provided by Valiere Nathaniel (University of Grenoble, France). In addition, ten wolf samples from Asturias, Spain (SP) and three from Slovakia (SL) were analysed for comparison. DNA isolation and amplification Proteinase K digestion and phenol/chloroform standard protocols were used for genomic DNA isolation from tissues. Nuclear DNA was ex- tracted from hair bulbs either according to HIGU- cHI et al. (1988) or by Chelex isolation (WALSH et al. 1991). Samples were genotyped for five di- nucleotide (AC),„ polymorphic microsatellite loci and one tetranucleotide locus, previously charac- terized in dog (OSTRANDER et al. 1993; FRANCISCO et al. 1996). Amplifications were carried out in 20 ul volume, containing 10 mM Tris-HCl, 50 mM KCl, 2mM MsCh, 0.1% Triton X-100, 200 uM each dNTP, 0.5 units Taq DNA Polymerase (Pro- mega), 5 pmol of each primer and 1-4 ul of DNA solution. An initial denaturation step at 94 °C for 3 min was followed by 35 cycles of amplification each of 45 sec at 92°C, 45 sec at the annealing temperature (55-58°C) and 60 sec at 72°C. The reaction terminated with a polymerization step at PC Tordmmn: In order to verify the successful products of single PCRs, 5 ul of reaction solution were run on a 2% agarose gel (Biorad) containing ethidium bro- mide, and the presence of correctly amplified fragments was detected by comparing their length with a DNA size marker. Single alleles were sized by running denatured PCR products through capillary electrophoresis in an ABI PRISM 310 automatic sequencer (Per- kin-Elmer). Statistical analysis By combining alleles at each locus, individual genotypes were obtained. The GENEPOP soft- 324 M. ScanourA et al. ware program version 3.2a (RAyMonD and Rous- sET 1995) was used to calculate allele frequencies and to test data sets for deviation from Hardy- Weinberg equilibrium (HWE) as well as for geno- typic linkage disequilibrium. Because many rare alleles were present, HWE departures were also tested “with pooling” (HArTL and CLArK 1989), by grouping genotypes into three classes: homo- zygotes for the most common allele, common/rare allele heterozygotes, and all other genotypes. Observed heterozygosity (H,) and unbiased ex- pected heterozygosity (H.) (NEı 1978) were esti- mated for each subpopulation. Probability of gen- otype identity was obtained using the formula Do ) Down, l Ba where p; and p; are the frequencies of the ith and jth alleles at a given locus (PAETKAU and STROBECK 1994). Single locus probabilities were combined to obtain the total probability over all 6 loci, as- suming independence of different loci, as sup- ported by the microsatellite linkage map in the domestic dog (MELLERSH et al. 1997). In order to evaluate the level of genetic variation, the H. estimated for the overall Italian popula- tion (N = 38) over five of the six examined loci (109, 123, 204, 250, and 377) was compared with values recalculated for North American popula- tions over the same loci using published data (Roy et al. 1994; ForBEs and Bovp 1997). Allelic and genotypic differentiations were evalu- ated for each population pair within the Italian range (AR, AB, VC, and FR), and then were pooled and compared with the two other Eu- ropean populations (SP and SL). Two different approaches were used to estimate the level of dif- ferentiation between samples by GENEPOP: a Fisher exact test was performed to test the homo- geneity of allelic distributions across populations (RAymonD and Rousser 1995), whereas a log- likelihood (G) based exact test was used for geno- typic differentiation (GouDET et al. 1996). The significance level was always established using Bonferroni’s criterion for multiple tests. In both cases, an unbiased estimate of the p-value was ob- tained, associated with the null hypothesis of identical distribution across populations. A matrix was created containing the proportions of shared alleles (Pıs), over the six loci, for all pair- wise comparisons of sampled individuals, as de- scribed in Bowcock et al. (1994). In order to obtain a measure of divergence among populations, Pıs values were averaged over population pairs and the pairwise distance value Dıs was calculated as (1- Pas;j) where the second term represents the mean Pıs calculated over all combinations between the ith and the jth subpopulation genotypes. Mean dis- tance values were also computed among indivi- duals of a single sample, in order to evaluate in- tra-group homogeneity. To eliminate the bias originating from different degrees of sample homogeneity, mostly due to different breadths of sampling areas, a new matrix was extrapolated, averaging the differences between inter- and in- tra-population distance values: s (Dasj; Dasi) + (Das; Das;) IASıe mm, 2 Furthermore, Nei’s unbiased genetic distance (Neı 1972) was computed by BIOSYS-2 software (SWOFFORD and SELANDER 1997) between all sub- population pairs. Multilocus F-statistic was calculated by GENE- POP, estimating the Fsr coefficient for each pair of samples and for all Italian samples, according to WEIR and CoCKERHAM (1984). In order to eval- uate gene flow among subpoputations, the same program allowed the effective number of mi- grants per generation (Nm) to be estimated on the basis of the private allele model (SLATKIN 1985; BArToOn and SLATKIN 1986). Thereafter a statistical analysis was carried out using the ISOLDE program, in the GENEPOP package, performing Mantel’s tests (1000 permutations) to highlight the possible presence of isolation by dis- tance in the Italian population. For this purpose, Dis, Nei’s unbiased distance, and Fsr were chosen as measures of genetic divergence and compared with geographic distance. Results A total of 51 wolves was genotyped at six microsatellite loci. All the loci showed poly- morphism in the four Italian subpopulations investigated, except for locus 377 in the Al- pine sample where allele A was fixed. Allele frequencies are given in table 1. Average H. ranged from 0.505 # 0.106 to 0.680 + 0.038, while the probability of identity varied from 1/1700 for the Alpine sample to less than 1/ 100.000 for the central-southern Apennines subpopulation (Tab. 2). In three out of four Italian samples, average H. had a lower val- ue than the observed one (Tab. 3), possibly due to limited outbreeding. Negative F;, val- ues confirm this possibility (mean F;, over 6 Recent recovery of the Italian wolf population 325 Table 1. Allele frequency distributions at 6 microsatellite loci in wolf samples (AR - Tuscan Apennines; AB - Central-southern Apennines; VC - Alta Maremma: FR - Alpes Maritimes: SP - Spain; SL - Slovakia). Locus 109 I (m) an Ina) les) 3 [o) cus 123 zzonmnmmon»> [®) cus 250 A B @ D E G H Lo A B (C D E F G H Locus 377 cus 2158 a Tee Del S Reh te PESTDozrunmD 326 M. Scanoura et al. Table 2. Expected heterozygosity (number of alleles in parentheses) and probability of identity in Italian wolf samples (AR - Tuscan Apennines; AB - Central-southern Apennines; VC - Alta Maremma; FR - Alpes Maritimes). Heterozygosity Probability of Identity 109 0.406 (3) 123 | 0.594 (3) 204 0.594 (3) 250 0.594 (3) 377 0.500 (2) 0. 0.555 2158 0.656 (3) 0. 0.170 All Loci 0.557 3.4x 10“ 01977 0.182 0.557 0.232 0.236 0.125 0.277 0.110 0.128 0.076 8.7%x10°% 0.388 0.248 0.248 0.248 0.375 0.193 AI“ 0.248 0.130 0.392 0.194 1.000 0.248 6.1x10* 0.527 0.680 Table 3. Genetic variation at 6 microsatellite loci and deviation from Hardy-Weinberg equilibrium. N, sample size; A, mean number of alleles per locus; H,, observed heterozygosity (direct count); H., Hardy-Weinberg ex- pected heterozygosity; HWE, significance of chi-square test for Hardy-Weinberg equilibrium without pooling; HWE,, significance of chi-square test for Hardy-Weinberg equilibrium with pooling; SE, standard error; n.s., not significant. Subpopulation/Population AR - Tuscan Apennines 13 AB - Central-southern Apennines 17 VC - Alta Maremma FR - Alpes Maritimes 0.527 (0.061) 0.680 (0.038) 0.557 (0.036) 0.505 (0.106) ) ) ) Total (Italian population) SP - Spain SL - Slovakia 0.644 (0.040 0.693 (0.023 0.435 (0.097 Table 4. Comparison of genetic variation at 5 microsatellite loci among different wolf populations and related canid species (N, sample size; A, mean number of alleles per locus + standard error; H., Hardy-Weinberg expected heterozygosity + standard error). ° recomputed from single locus frequencies data. Species/Population Reference Canis lupus Italy Spain (Asturias) Canada (Northwest Territories) Canada (Alberta) USA (Montana) USA (Yellowstone National Park) Canis simensis Canıs lupus f. familiaris 0.619 # 0.039 0:743==10:013 0.714 # 0.063 0.709 + 0.027 0.659 # 0.055 0.686 +# 0.074 0.167 # 0.011 0.714 # 0.075 this study this study Roy et al. (1994)° Roy et al. (1994)? ForBEs and Boyp (1997)° ForBEs and BoYp (1997)° GoTTeLLl et al. (1994)? GOTTELLI et al. (1994)° locı for the overall Italian population equals -0.039), indicating breeding among non-re- latives. Departures from the Hardy-Wein- berg equilibrium for the different samples proved not significant at all loci. However, as shown in table 3, a recent colonized area (FR) and a strongly fluctuating population (VC) showed a marked excess of heterozy- gotes, although statistically not significant. Linkage disequilibrium for each pair of loci was confirmed by a probability test analysis in GENEPOP. Recent recovery of the Italian wolf population Referring to the restricted analysis (over 5 loci), mean H. for the Italian sample was 0.619+0.039, with a mean number of al- leles per locus (A) of 5.4 + 0.4. A compari- son with other wolf populations and with two canid species (Ethiopian wolf, Canis si- mensis, and domestic dog, Canis lupus f. fa- miliaris) is shown in table 4. Levels of differentiation among subpopula- tions and populations were detected by comparing allelice and genotypic frequency distributions across loci. Significant levels of divergence within the Italian range were obtained only from the comparison be- tween AR and AB samples (allelic data: Fisher exact test, p = 0.00044; genotypic data: G-test p = 0.00042). On the other hand, the whole Italian population showed both a genic and a genotypic statistically significant divergence from the Spanish and the Slovakian samples (allelic data: Fisher exact test, p< 0.001; genotypic data: G-test p< 0.001). Single-locus comparisons allow discrimination of subpopulations due to the presence of private alleles. Both AR and AB subpopulations showed exclusive alleles at different loci. All the alleles but one (allele H at locus 204) present in the VC sample belong to the AB subpopulation also. On the other hand, the FR sample has alleles present in both AR and AB popula- tions, except for three alleles at locus 123, one exclusive to AR and two to AB, respec- tively, and for allele L at locus 2158, absent in other Italian samples. Mean Das (pro- portion of alleles not shared) within the Ita- lian population was 0.466, whereas the mean values of derived Dıs for subpopula- tion pairs ranged between 0.049 and 0.162 (Tab. 5). Nei’s unbiased distances were lower than 0.2 for all the comparisons among Italian samples (Tab. 6a, below diagonal), and higher than 0.6 for all inter-population pair- wise comparisons (Tab. 6b, below diago- nal). In the former case, the minimum va- lues were obtained between AB and VC samples (0.051) and between AR and FR samples (0.054). On the basis of Nei’s un- biased distance, a cophenetic tree may be plotted (Fig. 2). 327 Fsr values accounted for the proportion of total variation due to diversity between samples. Overall, Fsr for the whole Italian population was 0.053, a low value consider- ing that a value of zero expresses the iden- Fig. 2. Upgma phenogram of wolf populations based on Nei’s unbiased genetic distances. Table 5. Pairwise genetic distances by Shared Alleles (D’as) Subpopulation AR - Tuscan Apennines * AB - Central-southern 0.049 Apennines VC - Alta Maremma FR - Alpes Maritimes 0.070 0.062 0.061 ii 0.089 0.162 Table 6. Nei’s unbiased genetic distance (below diag- onal) and pairwise Fs-values (above diagonal) among Italian subpopulations (a) and among European wolf populations (b). 2) Subpopulation AR - Tuscan Apen- 0.058..0:0712.0:.051 nines AB - Central-southern 0.094 Apennines VC - Alta Maremma FR - Alpes Maritimes 0.022 0.042 0.081 0.051 0.054 0.093 0.162 0.121 Population IT - Italy SP - Spain SL - Slovakia 0-19 92652 0.181 0.8512:30.6510 328 M. Scanpura et al. tity of allele frequencies among all subpo- pulations. Fsr values obtained for each sub- population and population pair are shown in table 6a and 6b, respectively (above di- agonal). The multilocus estimate of Nm for the overall Italian population gave a num- ber of about 2.0 migrants per generation, which suggests a relevant amount of gene flow among subpopulations. When calculat- ing Nm between the two most representa- tive subpopulations, AR and AB, a value of 1.7 was obtained. A positive correlation was found, using Mantel’s test, only between D’ıs and geo- graphical distance (Spearman rank correla- tion coefficient, p = 0.038), suggesting the presence of moderate isolation by distance. No significant correlation was found for Nei’s distance and Fs7. Discussion Italian population samples showed a high intra-group diversity, as both H. and P;a were relevant within each subpopulation. Although the standard error was sometimes considerable, due to the small sample size, the A was high for most loci even in small samples. Mean heterozygosity over 5 loci proved very close to the values obtained for North American populations (Roy et al. 1994; Forses and Boyp 1997), and also the mean number of alleles per locus was completely comparable. This agrees with al- lozymic data, whose level of heterozygosity for the Italian population was found to be relatively high (Ranoi et al. 1993). Comparing Italian wolf with a related spe- cies population, Ethiopian wolf (GOTTELLI et al. 1994), which went through prolonged isolation, a vast difference in heterozygosity calculated over the same loci is evident. All samples fitted Hardy-Weinberg expec- tations, whether the %”-test was performed with or without pooling. Excess of hetero- zygotes in strongly fluctuating or recently colonized subpopulations, VC and FR, proved high but not significant. This may be due to random assembling of founder genotypes occupying new territories. High Hardy-Weinberg expected heterozyg- osity, in comparison with the one observed in Italian samples, may arise from limited outbreeding, as confirmed by the negative F;, value. Breeding between unrelated indi- viduals is a common trend in natural wolf populations (SMITH et al. 1996), neverthe- less high levels of induced mortality may enhance the natural turnover of pack mem- bers and favour outbreeding. The most immediate indicator of genetic differentiation is allele frequency distribu- tion. A significant level of divergence among Italian samples was found over all 6 locı only for the AB-AR pair, comparing both allelic and genotypic frequencies. As microsatellites are particularly sensitive to allele frequency differentiation, such differ- ences may not be considered on their own a proof of genetic isolation. Looking at allele frequency distribution, generally, the highest was the frequency across populations, and the widest was spa- tial diffusion. Several rare alleles were pre- sent, with occasional local specificity. Examining the presence of single alleles, AR and AB samples showed a moderate level of diversity due to the presence of pri- vate alleles (5 for AR and 10 for AB). VC and FR samples appear compatible with a possible derivation from the other two sub- populations, but they also have exclusive al- leles. The allele H at locus 204, present in the VC sample, was never found before in wolf individuals, whereas it proved the pre- valent allele in dog samples (data not shown). A possible wolf dog hybridization event among the ancestors of the female in- dividual presenting such an allele cannot be excluded. On the contrary, allele L at locus 2158 in the Alpine sample was found in other European wolves (e.g. in Spain) and might be present also at low frequency in the Italian population, but it was not de- tected in this work as a consequence of the limited sample size. Both Ds and Nei’s unbiased distance val- ues confirm the expected origin of the Al- pine subpopulation from the northern Apennines. The allelic pattern of the VC sample, combined with measures of dis- Recent recovery of the Italian wolf population tance, is compatible with colonization from southern regions (AB). The area represents the most northern tail of the Tyrrhenian coast subpopulation, which is supposed to have maintained links up to the first half of the last century with the central-southern subpopulation and possibly have restored them in the last few decades. Data obtained in the present study seem to confirm this hypothesis. The overall Fsr value was small (0.053), suggesting very limited structuring of the Italian population. Gene flow is relevant as, wıth more than l migrant per generation, differences among subpopulations are reduced, balan- cing the effect of genetic drift (SLATKIN 1987). A value of 1.7, derived from the AR and AB subpopulations, is more reliable with respect to the estimate for the whole Italian population. This is because these subpopulations represent areas where the species was probably never eradicated, and they are close enough to maintain a suffi- cient level of gene flow. These aspects make the assumption of migration-drift equilib- rium (SLATKIN 1993) a plausible one. Evidence for isolation by distance through- out the Italian population was found com- paring D’ıs with geographic distances be- tween different subpopulations. Historical factors and the geographic shape of the re- gion played a key role in establishing a con- tinuous and directional gene flow across the peninsula. Summarizing, we have pointed out that the Italian wolf population constitutes a well- defined and viable natural population, where a high gene flow guarantees a suffi- cient genetic exchange among different areas. The origin of the nuclei settling in Alpes Maritimes should be attributed to movements of dispersing individuals from the northern Apennines, as suggested by the similarity between Alpine samples and specimens from the Tuscan Apennines. The Tyrrhenian subpopulation may have re- stored its continuity with the central Apen- nines, but human-caused mortality continu- ously threatens its stability, favouring high 329 turnover and eventually outbreeding, at least in the peripheral zones of the wolf range. Overall microsatellite diversity is substan- tial and comparable to that in North Amer- ican populations. Inbreeding depression seems to be far from threatening Italian wolves. However demographic factors, diffi- cult to predict, may affect population viabi- lity more than genetic aspects, producing dramatic changes especially in local situa- tions. Therefore it would be advisable to maintain the genetic flow high across the peninsula, in order to balance the effect of local bottlenecks. A long-term differentiation between Alpine and southern subpopulations may be ex- pected as consequence of isolation by dis- tance, while a progressive enlargement of the northern range along the Alps will pro- gressively bring the Italian wolf closer to the Balkan populations. The set of six polymorphic microsatellite locı used in this work represented an effec- tive tool for investigating the actual genetic status of wild wolf populations. The low probability of identity between individuals (in the order of 10°* to 10°°) reveals a high resolution power in resolving pedigrees, i. e., for kinship analysis. Acknowledgements We would like to thank the Provincial Adminis- tration of Arezzo for institutional and financial support in setting up the laboratory and bearing the costs of analysis, and for having provided wolf samples. Additional support for M.S. was pro- vided by a “G. Guelfi” grant from the Accademia Nazionale dei Lincei. We are also grateful to the staff of the Comunita Montana Val di Cecina and the staff of the Veterinary Service U.S.L. 5 in Volterra (Pisa), and to F. BALDASSARRI of the C.T. A. of the Foreste Casentinesi National Park. We also thank S. CAppeLLı and D. DEL CHIARO for help in genetic analysis, P. VARuUZZA and E. AvanZINELLI for technical assistance, G. B. HARTL, V. NATHANIEL, M. NORES-QUESADA, and P. HELLE for having supplied samples, U. BaıscH and J. BURGE for linguistic support. 330 M. ScanouraA et al. Zusammenfassung Neues Anwachsen der italienischen Wolfspopulation: eine genetische Untersuchung mittels Mikrosatelliten. Um die genetische Differenzierung innerhalb der anwachsenden italienischen Wolfspopulation (Ca- nis lupus) zu untersuchen, wurden 38 Individuen von 4 verschiedenen Stellen des derzeitigen Ver- breitungsgebietes mittels Mikrosatelliten analysiert. Die 6 betrachteten Mikrosatellitenloci zeigten einen hohen Grad an Polymorphismus und die kombinierte Identitätswahrscheinlichkeit reichte von 10°* bis 10°°. Insgesamt war die gefundene DNA-Variation beträchtlich und kaum niedriger als jene beim nordamerikanischen Grauwolf. Die beiden ältesten italienischen Teilpopulationen, jene in der Toskana und jene im zentralen bis südlichen Apennin, zeigten nur eine mäßige Diver- genz. Die gefundenen Werte stimmten mit einer Abstammung der ersteren von der Westalpen-Sub- population überein, während die letztere Ähnlichkeit zur Subpopulation der Westküste zeigte. Der Genfluß innerhalb des italienischen Wolfsbestandes war hoch. Die geographische Verteilung der Probengebiete stimmte mit den genetischen Abständen überein, was auf das Vorliegen von ‚Isola- tion durch Distanz’ hindeutet. Hohe Divergenzniveaus wurden zwischen den italienischen Proben und jenen aus anderen europäischen Gebieten gefunden. Insgesamt deuten die Ergebnisse darauf hin, daß, trotz der Einförmigkeit im Bereich der mtDNA, der italienische Wolf noch immer einen ho- hen Grad an Kern-DNA-Variabilität und eine ausgeprägte genetische Identität besitzt. Eine weitere Vergrößerung des Wolfsareals, wie sie aufgrund der hohen Dispersionsrate zu erwarten ist, könnte in einigen Jahrzehnten dazu führen, daß die historische Isolation von einem genetischen Aus- tausch mit dem osteuropäischen Genpool abgelöst wird. References BARToN, N. H.; SLATKIN, M. (1986): A quasi-equi- librium theory of the distribution of rare al- leles in a subdivided population. Heredity 56, 409-415. BoItanı, L. (1992): Wolf research and conserva- tion in Italy. Biol. Conserv. 61, 125-132. BoITAnT, L.; Cruccı, P. (1993): Wolves in Italy: cri- tical issues for their conservation. In: Wolves in Europe, Status and Perspectives. Ed. by C. PROMBERGER and W. SCHRÖDER. Munich: Munich Wildlife Society. Pp. 74-90. Bowcock, A.M.:;: Ru1z-LinAREs, A.; "TOMFOHR- DE, J.; MıncH, E.; KıDD, J. R.; CAVvALLI-SFOR- za,L.L. (1994): High resolution of human evolutionary trees with polymorphic microsa- tellites. Nature 368, 455-457. BREITENMOSER, U. (1998): Large predators in the Alps: fall and rise of man’s competitors. Biol. Conserv. 83, 279-289. CAGNOLARO, L.; Rosso, D.; SPAGNESI, M.; VEN- TURI, B. (1974): Investigation on the wolf (Ca- nis lupus) distribution in Italy and in Canton Tieino and Canton Grigioni (Switzerland). Ric. Biol. Selvaggina 59, 1-75 (in Italian). FoRBES, S. H.: BoyD, D. K. (1996): Genetic variation of naturally colonizing wolves in the Central Rocky Mountains. Cons. Biol. 10, 1082-1090. FORBES, S. H.; BoyD, D.K. (1997): Genetic struc- ture and migration in native and reintroduced Rocky Mountain wolf populations. Cons. Biol. 11, 1226-1234. FRANCISCI, F.; GUBERTI, V. (1993): Recent trends of wolves in Italy as apparent from kill figures and specimens. In: Wolves in Europe, Status and Perspectives. Ed. by C. PROMBERGER and W. SCHRÖDER. Munich: Munich Wildlife So- ciety. Pp. 91-102. FRANCISCo, L. V.; LANGSTON, A. A.; MEL- LERSH, C. S.; NEAL, C.L.; OSTRANDER, FE. A. (1996): A class of highly polymorphic tetranu- cleotide repeats for canine genetic mapping. Mamm. Genome 7, 359-362. GARCIA-MORENO, J. M.; MAToco, M. D.; Roy, M. S.; GEFFEN, E.; WAYNE, R.K. (1996): Relation- ships and genetic purity of the endangered Mexican wolf based on analysis of microsatel- lite loci. Cons. Biol. 10, 376-389. GOTTELLI, D.; SILLERO-ZUBIRI, C.; APPLEBAUM, G.D.; Roy,M.S.; GiIRMAn,D.]J.; GARCIA- MORENGO, J.; ÖSTRANDER, E. A.; WAYNnE,R.K. (1994): Molecular genetics of the most endan- gered canid: the Ethiopian wolf Canis simen- sis. Mol. Ecol. 3, 301-312. GOUDET, J.; RAYMOND, M.: DE MEEÜS, T.; Rous- Recent recovery of the Italian wolf population SET, F. (1996): Testing differentiation in di- ploid populations. Genetics 144, 1933-1940. HARTL, D. L.; CLARK, A.G. (1989): Principles of Population Genetics. 2”® ed. Sunderland, Massachusetts: Sinauer Associates. HiGguchHiı, R.; BEROLDINGEN, C. H. von; MELLERSH, C.S. (1988): DNA typing from single hairs. Nature 332, 543-546. KENNEDY, P.K.; KENNEDY, M. L.; CLARKSOoN, P. L.; Liepins, I.S. (1991): Genetic variability in natural populations of gray wolf, Canis lupus. Can. J. Zool. 69, 1183-1188. MATTIOLI, L.; APOLLONIO, M.;, MAZZARONE, V.; CENTOFANTI, E. (1995): Wolf food habits and wild ungulate availability in the Foreste Casentinesi National Park, Italy. Acta Ther- iol. 40, 337-402. MELLERSH, C.S.;, LANGSTON, A. A.; AÄCLAND, G. M.; FLEMING, M. A.; Ray,K.; WIEGAND, N. A.; FRANCIScCo, L. V.; GiIBBs, M.; AGUIR- RE, G. D.; OSTRANDER, E. A. (1997): A linkage map of the canine genome. Genomics 46, 326-336. MERIGGI, A.; LovaRı, S. (1996): A review of wolf predation in southern Europe: does the wolf prefer wild prey to livestock? J. Appl. Ecol. 33, 1561-1571. Ne, M. (1972): Genetic distance between popula- tions. Am. Nat. 109, 283-292. Neı, M. (1978): Estimation of average heterozyg- osity and genetic distance from a small num- ber of individuals. Genetics 89, 583-590. OSTRANDER, E. A.; SPRAGUE, G. F.; RiNE, J. (1993): Identification and characterization of dinu- cleotide repeat (CA)n markers for genetic mapping in dog. Genomics 16, 207-213. PAETKAU, D.; STROBECK, C. (1994): Microsatellite analysis of genetic variation in black bear po- pulations. Mol. Ecol. 3, 489-495. RAnDI, E.; FRANCISCI, F.; LucchHintı, V. (1995): Mi- tochondrial DNA restriction-fragment-length monomorphism in the Italian wolf (Canis lu- pus) population. J. Zool. Syst. Evol. Res. 33, 97-100. RANDI, E.; LUccHint1, V.; FRAnNcıscı, F. (1993): Al- lozyme variability in the Italian wolf (Canis lupus) population. Heredity 71, 516-522. RANDI, E.; LuccHinı, V.; CHRISTENSEN, M.F.;; Muccı, N.; FUNK, S. M.; DoLr, G.; LOESCHCKE, V. (2000): Mitochondrial DNA variability in Ita- lian and East European wolves: detecting the consequences of small population size and hybridization. Cons. Biol. 14, 464-473. RAYMOND, M.; Roussert, F. (1995): An exact test for population differentiation. Evolution 49, 1280-1283. 331 Roy, M. S.; GEFFEN, E.; SMITH, D.; OSTRANDER, E. A.; WAYNE, R.K. (1994): Patterns of differ- entiation and hybridization in North Ameri- can wolflike canids by analysis of microsatel- lite loci. Mol. Biol. Evol. 11, 553-570. SLATKIN, M. (1985): Rare alleles as indicators of gene flow. Evolution 39, 53-65. SLATKIN, M. (1987): Gene flow and the geographic structure of natural popolations. Science 236, 787-792. SLATKIN, M. (1993): Isolation by distance in equili- brium and non-equilibrium populations. Evo- lution 47, 264-279. SMITH, D.; MEIER, T.;, GEFFEN,E.; MEcH,L.D.; BurcH, J.W; Apams,L.G.; WAaYNE,R.K. (1996): Is incest common in gray wolf packs? Behav. Ecol. 8, 384-391. SOKAL, R. R.; RoHLF, F. J. (1995): Biometry: the principles and practice of statistics in biologi- cal research. 3" Ed., New York: W.H. Free- man and Company. SWOFFORD, D. L.; SELANDER, R. B. (1997): Biosys- 2: acomputer program for the analysis of alle- lic variation in genetics. User manual. VILA, C.; AMORIM, I. R.; LEONARD, J. A.; PosA- DA, D.; CASTROVIEJO, J.; PETRUCCI-FONSECA, F;; CRANDALL, K. A.; ELLEGREN, H.; WAY- NE, R.K. (1999): Mitochondrial DNA phylo- geography and population history of the grey wolf Canis lupus. Mol. Ecol. 8, 2089-2103. WALSH, P. S.; METZGER, D. A.; HıguchHı, R. (1991): Chelex 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material. BioTechniques 10, 506-513. WAYNE, R. K.; LEHMAN, N.; ALLARD, N. W.; HONEY- cUTT, R. L. (1992): Mitochondrial DNA varia- bility of the grey wolf genetic consequences of population decline and habitat fragmenta- tion. Cons. Biol. 6, 559-569. WEIR, B. S.; COCKERHAM, C. C. (1984): Estimating F-statistics for the analysis of population structure. Evolution 38, 1358-1370. ZIMEN, E.; BoITANnI, L. (1975): Number and distri- bution of wolves in Italy. Z. Säugetierkunde 40, 102-112. Authors’ addresses: M. SCANDURA, M. ApoLLonıo, Dipartimento di Zoologia e Antropologia Biologia, Universitä di Sassari, Via Muroni 25, 1-07100 Sassari, Italy (e-mail: marcoapo@discau.unipi.it) L. MATTıoLı, Ufficio Piano Faunistico, Provincia di Arezzo, Piazza della Libertä 3, I-52100 Arezzo, Italy Mamm. biol. 66 (2001) 332-344 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol 5. Mammalian Biology SE \\z Zeitschrift für Säugetierkunde Original investigation The phylogenetic position of southern relictual species of Microtus (Muridae: Rodentia) in North America Von C. J. CONROY, YOLANDA HORTELANO, F. A. CERVANTES and J. A. Cook University of Alaska Museum and Institute of Arctic Biology, University of Alaska Fairbanks, Fairbanks, Alaska, USA and Departamento de Zoologia, Instituto de Biologia, Universidad Nacional Autönoma de M&xico, Coyoacan, Mexico Receipt of Ms. 14. 02. 2001 Acceptance of Ms. 03. 05. 2001 Abstract Climatic fluctuation led to isolation of some populations of temperate species on southern moun- taintops during warming trends. The most southern species of arvicoline rodents (Microtus guate- malensis, M. oaxacensis, M. quasiater, and M. umbrosus) in North America may be relicts, isolated in the mountains of Mexico and Guatemala at the end of the Pleistocene. We used parsimony and likelihood analyses of complete mitochondrial cytochrome b gene sequences of 28 species of Micro- tus, including eight Eurasian species, holarctic M. oeconomus, and all extant North American spe- cies except the island endemic M. breweri. North American species of Microtus were monophyletic under the maximum-likelihood criterion, but polyphyletic under parsimony. Likelihood ratio tests and bootstrapping indicated a rapid basal radiation with short intervals between cladogenic events. However, several sister taxon relationships were robust to bootstrapping or consistent between methods. We found that M. quasiater was sister to M. pinetorum, and these taxa were sister to a clade of M. oaxacensis and M. guatemalensis. The phylogenetic position of M. umbrosus, however, was unclear. Monophyly of the relicts was rejected by a likelihood ratio test, suggesting multiple southern invasions by arvicoline rodents. Phylogenetic data for these and other co-distributed taxa should be used in conservation efforts for these remote areas. Key words: Microtus, Arvicolinae, historical biogeography, conservation, molecular systematics, Mexico Introduction Climatic fluctuations have been linked to the spatial expansion and contraction of species (GRAHAM and GRIMM 1990; GRAHAM et al. 1996). Isolation during contraction phases may have stimulated allopatric di- versification, particularly at the southern edge of ranges. For example, the mountains of Mexico and Guatemala host a highly en- 1616-5047/01/66/06-332 $ 15.00/0. demic flora and fauna (RAMAMOORTHY et al. 1993) that includes a diverse set of or- ganisms associated with mesic environ- ments. These organisms apparently invaded the region during cooler periods and then became isolated at higher elevations as con- ditions at lower elevations became drier and warmer. This invasion and isolation cycle may have occurred numerous times during the Pleistocene and has contributed to the complex biogeographic history of the region (SULLIVAN et al. 1997). The genus Microtus is holarctic and north temperate in distribution and in the New World reaches its southern limit in Central America. At higher latitudes, up to five spe- cies (e.g. M. longicaudus, M. miurus, M. oeconomus, M. pennsylvanicus, and M. xanthognathus in Yukon Territory) may be found in close proximity. However, species tend to be more allopatrically distributed at southern latitudes. For example, M. guate- malensis, M. oaxacensis, M. quasiater, and M. umbrosus are endemic to separate mountains in the cloud and pine forests of Mexico and Guatemala (Fig. 1). M. quasia- ter, the Jalapan vole, is found in the southern Sierra Madre Oriental in Central Mexico. M. oaxacensis, the Tarabundi vole, is iso- lated in the Sierra de Juarez of Oaxaca, and occurs at elevations of 1,600 (SANCHEZ et al. 1996) to 2,499 m (JonEsS and GENOWAYS 1967). M. umbrosus, the Zempoaltepec vole, is restricted to approximately 80 km? at ele- vations ranging from 1,829 to 3,000 m (FREY and CERVANTES 1997) around Mt. Zempoal- tepec in Oaxaca. M. guatemalensis, the most southern species, occurs from the mountains of central Chiapas south to central Guate- mala. These species may be the result of peripheral isolation of ancestors that were more widely distributed during cool periods Mexican Microtus 333 of the early to middle Pleistocene (Horr- MANN and Korppr 1985). In contrast to these four disjunct species, M. mexicanus is widespread in Mexico and oc- curs in limited sympatry or parapatry with M. oaxacensis, M. quasiater, and M. umbro- sus with a discontinuous distribution ex- tending from New Mexico and Arizona to southern Mexico (Hau 1981). Its fossil re- cord is limited to the late Wisconsinan from San Josecito in northeastern Mexico and lo- calities farther north (ZAKRZEWSKI 1985). Phylogenetic relationships between M. mexicanus and other species of Microtus are unclear, but there is little indication that this species shares a common ancestor with other Mesoamerican species. Thus, its pre- sence in Mexico probably reflects an inde- pendent coloniızation. The aim of our study therefore is to investi- gate the phylogenetic relationships of the southern species of Microtus in order to test hypotheses regarding their ancestry in North America and their taxonomic rela- tionships to other species. Material and methods DNA was extracted from ethanol-preserved tis- sues of these four southern Microtus (Tab. 1) with methods described in Conkoy and Cook (1999). The cytochrome b gene (hereafter cyt b) was am- plified in two or three sections and sequenced in Table 1. Specimens of Meso-American species of Microtus examined in this study (in addition to those reported in ConRoY and Cook [2000]) including location from which each specimen was collected, and CNMA (Colecciön Nacional de Mamiferos, Universidad Nacional Autönoma de Mexico) catalog number. Species Collection location | M. umbrosus Mpio. Tlahuitoltepec, 2450 m. M. guatemalensis Mexico: Oaxaca Cerro Zempoaltepetl, 5 km N Sta Ma. Yacochi. Mexico: Chiapas: Cerro Tzontehuitz, 13 km NE San Cristobal CNMA Catalog Number 34890, 34894 35262 de las Casas, Mpio. Chamula, 2880 m. Mexico: Oaxaca: 11 km SW La Esperanza, Mpio. Santiago M. oaxacensis Santiago Comaltepec, 2000 m. M. quasiater Comaltepec, 2000 m.; Oaxaca: 11 km SE La Esperanza, Mpio. Mexico: Veracruz: 5 km W Naolinco, Mpio. Naolinco, 1650 m. 27415, 33815 35282, 35274 334 C. J. Conroy et al. both directions (Perkin-Elmer Prism” dye termi- from 24 species of Microtus and two species of nator kit; Fst-RR, 402119) on an ABI 373a auto- Clethrionomys (Conkoy and Cook 2000). In the mated sequencer. We included cyt b sequences phylogenetic analysis we represented each species OTHER NORTH AMERICAN SPECIES OF MICROTUS M. QUASIATER i. UMBROSUS => Ö M. eine —. M. GUATEMALENSIS Fig. 1. Map of the distribution of species of Microtus in North America and south into Guatemala (redrawn from HOFFMANN and Ko£PPL 1985). with a single individual, but we examined intra- specific variation where possible by including multiple representatives for 24 of the 28 species. We used unweighted parsimony (MP) and maxi- mum likelihood (ML) with the software PAUP*, version 4.0d64 (SwoFFORD 1998). We estimated parameters for likelihood models of increasing complexity (JC: Jukes and CaAntor 1969; HKY85: HaAsEGAwA et al. 1985; and GTR: YAnG 1994 a; HKY85 + T, and GTR + T: Yanc 1994). We tested these with likelihood ratio tests to de- termine significant differences among their likeli- hood scores. We used 1,000 random addition se- quences to locate multiple tree islands in the heuristic MP searches. As tests of node strength, we bootstrapped the parsimony analysis 1,000 times and bootstrapped the final ML analysis 100 times. We tested several alternative phylogenetic topologies against the ML tree by forcing likelıi- hood searches to find the ML topology with parti- cular constraints and then comparing scores (KısHuıno and HAsEGAwA 1989). We tested mono- phyly of the four relictual Mesoamerican species, and the four shortest maximum parsimony trees. Besides examining relationships among these southern latitude taxa, several tests were con- ducted to ascertain the effects of adding these four species to more general hypotheses concern- ing the systematics of Microtus previously con- ducted (Conroy and Cook 2000). We tested the monophyly of subgenus Stenocranius (RAUSCH 1964). This subgenus includes the Asiatic species M. gregalis and the North American M. miurus and M. abbreviatus. It has been hypothesized that the species in this subgenus diverged relatively re- cently. Secondly, we tested the monophyly of a clade of taiga adapted voles (M. richardsoni, M. xanthognathus, M. chrotorrhinus) that was not supported in our previous study (Conroy and Cook 2000). And, third, we tested the.monophyly of all North American taiga voles (HOFFMANN and Korprpr 1985), because previously we were able to reject this hypothesis with the Kishino- Hasegawa test. By including these tests, our inves- tigation explores whether the addition of four southern taxa changes our interpretation of arvi- coline history in North America. Prior to this study, the relationships of southern species to other North American species was unclear. Results Base composition and distribution of vari- able sites of the cyt b gene was similar to other Microtus (Conroy and Cook 2000) as Mexican Microtus 335 well as other mammals (Irwın et al. 1991). Of the 1143 base pairs, 472 were variable across 28 species of Microtus and two species of Clethrionomys. When outgroups were ex- cluded, 460 sites were variable. Of these, 100 were in the first position, 23 in the sec- ond position, and 337 in the third position of codons. Of the 381 amino acids, 76 (20%) were variable across species of Microtus and the replacement pattern was consistent with structural models (e.g. Irwın et al. 1991). There were 351 parsimony informative nu- cleotide sites and g, statistics (gı = -0.325) indicated phylogenetic signal in the data. Maximum parsimony searches recovered four equally parsimonious trees (Fig. 2), each including a basal clade of the North American M. ochrogaster and Asian M. gre- galis. A clade of M. oeconomus, M. midden- dorffi, M. montebelli, M. kikuchii, and M. fortis (hereafter the “Asıan clade”), and the M. pennsylvanicus clade (i.e. M. penn- sylvanicus, M. montanus, M. townsendii, and M. canicaudus) were present in the four trees. M. pinetorum and M. quasiater were sister in all trees, had high bootstrap support in MP and ML analyses (99%) and relatively high decay values (12). M. oaxa- censis and M. guatemalensis were sister taxa in three of four MP trees, but the branch leading to this pair had weak bootstrap sup- port (<50%). Other clades were found in three or four of these shortest trees, but bootstrap support was generally low across basal relationships. M. umbrosus stemmed from the polytomy at the base of the Euro- pean and North American species. Random addition of taxa uncovered these four short- est trees, but simple addition (i.e. alphabe- tical by species name) led to an island of longer length trees. In the ML analysis, the HKY85 + T likeli- hood model (transition/transversion ratio = 3.43, &= 0.213) was chosen since more com- plex models (e.g. GTR + T) were not significantly more likely and produced the same topology (not shown). As in other stu- dies of cytochrome b (e.g. SULLIVAN et al. 1997), the addition of the gamma-distribu- ted rate parameter, allowing among-site rate variation, contributed significantly to 336 C. J. Conroy et al. 12 86/6 54/3 54/2 100/29 N 100/35 98/8 90/3 96/6 A 12 100/43 99/12 SSSSS ® S M. SSSSSSSSSSSSSSISSISESISTETNO . glareolus gapperi gregalis ochrogaster fortis middendorffi montebelli kikuchii oeconomus arvalis rossiaemeridionalis agrestis oregoni longicaudus canicaudus townsendii montanus pennsylvanicus richardsoni californicus mexicanus chrotorrhinus umbrosus xanthognathus . abbreviatus miurus . pinetorum quasiater oaxacensis qguatemalensis Fig. 2. Strict consensus of four trees from the MP search. Each tree had a length of 2038, a consistency index of 0.331, and retention index of 0.360. Values to the left of slash are bootstraps percentages greater than 50% from 1,000 searches with simple addition of taxa; values to the right of slash indicate decay indices calculated with 10 random-addition replicates for each search. the likelihood. This model produded one tree (Fig. 3) in which M. gregalis was basal, followed by the Asian clade. North Ameri- can endemic species formed a clade and the European species formed a sister clade to North American species. The Mesoamer- ican endemics were not basal within the clade of North American species. Three Mexican Microtus 337 Mesoamerican species (Microtus guatema- lensis, M. oaxacensis, and M. quasiater) dis- played the same branching pattern as three of the parsimony trees, while M. umbrosus was sister to M. chrotorrhinus. Other rela- tionships were similar to previous analyses (Conroy and Cook 2000). For example, the M. pennsylvanicus and Asian clades were Clethrionomys glareolus 83 Clethrionomys gapperi M. gregalis M. fortis M. middendorffi 100 M. montebelli M. kikuchii M. oeconomus 100 96 75 97 100 99 57 M. arvalis M. rossiaemeridionalis M. agrestis M. oregoni M. longicaudus M. canicaudus M. townsendii M. montanus M. pennsylvanicus M. ochrogaster M. xanthognathus M. richardsoni M. californicus M. mexicanus M. chrotorrhinus M. umbrosus M. abbreviatus M. miurus M. pinetorum M. quasiater M. oaxacensis ISTHMUS OF ? TEHUANTEPEC M. quatemalensis Fig. 3. Maximum likelihood phylogenetic tree based on the HKY85 + T model (see text for parameter values). Values above branches are bootstrap percentages greater than 50% from 100 bootstraps. 338 C. J. Conrovy et al. supported and M. mexicanus was sister to M. californicus. In Kishino-Hasegawa tests (Tab. 2), few of the alternate topologies could be rejected. Those that were rejected were one of the MP trees and two ML trees constrained to monophyly of the four southern species. Relative depth of divergence was estimated with pairwise likelihood distances between taxa estimated under the same model used for the ML phylogeny. M. abbreviatus and M. miurus, which differ by only 0.015, prob- ably separated at the end of the Pleistocene when rising sea levels isolated M. abbrevia- tus on islands in the Bering Strait. Another late Pleistocene divergence may be M. cani- caudus and M. townsendii which differ by 0.058. Other pairs that were relatively clo- sely related included M. arvalis and M. ros- siaemeridionalis (0.072) and M. montanus and M. pennsylvanicus (0.086). Divergence values between the southern species (M. quasiater and M. pinetorum [0.094], M. gua- temalensis and M. oaxacensis [0.113], and M. chrotorrhinus and M. umbrosus [0.137]) are greater than any of the preceding exam- ples, suggesting older speciation events. These comparisons should be interpreted with caution. Demographic and historical differences among species can alter rates of evolution and species of Microtus probably have evolved at different rates. Also, genet- ic differentiation may not necessarily coin- cide with vicariant events, such as the rise of the Bering Sea. Finally, molecular esti- mates of divergence, particularly for a sin- gle locus, have substantial amounts of sto- chastic variation (AyALA 1999). Discussion Previous investigations into the history of North American species of Microtus sug- Table 2. Kishino-Hasegawa tests of tree topologies. The "Stenocranius” constraint forced M. miurus, M. gregalis, and M. abbreviatus to be monophylyetic. The “Second taiga vole” constraint enforced M. xanthognathus, M. chro- tonrhinus, and M. richardsoni monophyly. The “All taiga voles” constraint enforced the monophyly of the “Second taiga vole” species with a clade of M. pennsylvanicus, M. montanus, M. townsendii, and M. canicaudus. The “Meso-Americans” constraint enforced M. guatemalensis, M. oaxacensis, M. quasiater, and M. umbrosus mono- phyly. Two trees were obtained from the ML search under this last constraint. The parsimony trees were derived from unweighted heuristic searches with ten random-addition replicates. Columns below indicate, from left to right, 1) -log likelihood score of that tree, 2) difference in score between that tree and the best tree, 3) the stan- dard deviation of that comparison, 4) the associated T test statistic, and 4) the associated p value. Tree Constraint 9868.811 9391.258 None Stenocranius Monophyletic Second taiga vole clade 9879.176 All Taiga voles Monophyletic 9897.044 Meso-Americans monophyletic I 9910.346 Meso-Americans monophyletic II 9910.346 Parsimony Tree 1 9903.109 ParsimonyTree 2 9908.743 ParsimonyTree 3 9908.493 ParsimonyTree 4 9919.839 *Significant at P< 0.05 -log likelihood Diff-InL s.d. (diff) T (best) 22.447 106.7,23 0.180 10.365 9.080 0.254 28.233 18.904 0.136 41.535 11.613 0.0004* 41.535 11.613 0.0004* 997298 22.586 0.032 39.932 20.492 0.052 39.682 20.581 0.054 54.027 20.528 0.013* gested an evolutionary history closely tied to fluctuating boreal ecosystems (HoFF- MANN and Korprr 1985). The elevational distribution of species of Microtus in Mexi- co and Guatemala indicates Holocene iso- lation due to the contraction of cool, moist forests. The paucity of synapomorphic mor- phologie characters, but apparent abun- dance of autapomorphic characters has led to their characterization as ancient, highly divergent species; however, their biogeo- graphic history has been enigmatic. This study reconsiders the biogeographic and evolutionary history of the southern relicts in light of molecular characters examined within a wider taxonomic sampling (28 spe- cies) for the genus. Biogeography The Mesoamerican species of Microtus were not basal in the North American clade, nor were they monophyletic. M. quasiater is clo- sely related to M. pinetorum, and M. oaxa- censis and M. guatemalensis are sister to this pair, while the fourth (M. umbrosus) is a weakly-supported sister taxon to M. chrotor- rhinus. Our data do not support the hypoth- esis of HoFFMAnN and Korrrr (1985) that these southern species of Microtus may be relicts of an early colonization of North America prior to the arrival of the ancestors of the other species that are endemic to the higher latitudes of North America. Interspe- cific distances do, however, suggest they be- gan to diverge earlier than suspected late- Pleistocene peripheral isolates such as M. abbreviatus and M. miurus. Assuming an equivalent rate of cyt b evolution across spe- cies of Microtus and Clethrionomys, Mesoa- merican Microtus apparently fragmented into isolated populations prior to the coloni- zation of North America by Clethrionomys, which is thought to have occurred during the early Pleistocene (REPENNING et al. 1990). The split between North American Clethrionomys gapperi and Eurasian C. glar- eolus (Tamura-Nei D = 0.079) occurred about the time their common ancestor in- vaded North America. However, disparate levels of ancestral polymorphism and his- Mexican Microtus 339 tories of population fluctuation in these in- dependent colonizers, among other vari- ables, would impact estimates of ages of differentiation. Two topologies that con- strained ML searches for monophyly of these southern species were significantly less likely than the best ML topology, refuting common ancestry of Mesoamerican species of Microtus. SULLIVAN et al. (1997) and SULLIVAN et al. (2000) found that the Isthmus of Tehuante- pec was a significant geographic barrier for other mesic rodents such as Reithrodon- tomys and the Peromyscus aztecus group in Mexico and Guatemala. They summarized divergence patterns in other co-distributed organisms (see RAMAMOORTHY et al. 1993) and recommended that the significance of the isthmus as a biogeographic barrier be tested with other mesic taxa. M. guatema- lensis occurs east of the isthmus and is sister to M. oaxacensis, found west of the isthmus, suggesting that their common ancestor may have been distributed across the isthmus when conditions were cooler. Western and eastern populations subsequently diverged. Microtus, Reithrodontomys, and Peromys- cus are widely sympatric in this region, and levels of genetic differentiation across the isthmus are surprisingly similar with D = 0.075-0.091 in Peromyscus (SULLIVAN et al. 1997), D = 0.063-0.085 in Reithrodon- tomys (SULLIVAN et al. 2000), and D = 0.100 between M. guatemalensis and M. oaxacensis. If we assume similar rates of molecular evolution, these three clades may have diverged across the Isthmus of Tehuantepec at roughly the same time. High levels of biotic diversity in the region provide ample material to further test hy- potheses regarding the timing and number of mammalian colonizations of the region. The biogeography of other organisms, such as reptiles (e.g. CAMPBELL 1984; CAMPBELL et al. 1989), may also suggest former con- nections across the Isthmus of Tehuantepec. Systematics Microtus quasiater is a member of the sub- genus Pitymys (MUSSER and CARLETON 340 C. J. Conroy et al. 1993) and shares dental morphology with extinet Microtus (Pitymys) meadensis, a widespread species of mid-Pleistocene North America and Mexico (REPENNING 1983). M. pinetorum has often been classi- fied as a member of Pitymys based on shared dental morphology (ZAKRZEWSKI 1985). Thus, it is not surprising that mor- phological characters (MussEr and CARLE- ton 1993) and these DNA sequences place M. quasiater sister to M. pinetorum. M. qua- siater previously was considered sister to M. ochrogaster (MooRE and JANECEK 1990) in an allozyme study, but only nine of 19 North American species and no Palearctic species were examined. The evolutionary relationships of M. guate- malensis and M. oaxacensis have not been addressed in detail, although they have not been suspected to be closely related (Musser and CARLETON 1993). Microtus guatemalensis ıs in the monotypic subgenus Herpetomys, but may have affinities with Pitymys (Martın 1987). Our analyses con- tradict its placement in Herpetomys, or suggestions that it is related to Phena- comys (Hınton 1926). The relationship be- tween M. oaxacensis and other species has also been obscure (MuUssErR and CARLETON 1993), but it also was considered a part of an early pitymyine invasion (HOFFMANN and Korrrpr 1985; MARTIN 1974). A wide- spread ancestor (e.g. M. meadensis) may have given rise to M. pinetorum, M. guate- malensis, M. quasiater, and M. oaxacensis, prior to peripheral isolation in the eastern deciduous forests and southern cloud for- ests (HOFFMANN and KorprL 1985). The mitochondrial phylogeny suggests isolation occurred first between an ancestor of M. pinetorurn-M. quasiater and an ances- tor of M. guatemalensis-M. oaxacensis although the branching order among these clades is weakly supported. The latter pair may have diverged after invasion across the Isthmus of Tehuantepec, while the for- mer pair diverged following an episode of range retraction induced by climatic warm- ing. Significant portions of the temperate flora on mountain tops in Mesoamerica are as old as the late Cenozoic (GRAHAM 1999), however, we do not know of a comparable phylogenetic assessment of vegetation. A sister relationship between M. umbrosus and M. chrotorrhinus was weakly supported in the ML tree and was unexpected because they differ morphologically. Microtus um- brosus is the sole member of the subgenus Orthriomys (MusseEr and CARLETON 1993) and has been considered a relict from an early invasion from Asia during the mid- Pleistocene by the extinct Phaiomys (MAR- rın 1987). Though previously considered closely related to M. xanthognathus (HALL and Keıson 1959), M. chrotorrhinus was la- ter distinguished based on chromosomal complement (RAuscH and Rausch 1974). The lack of similarity between M. chrotor- rhinus and M. umbrosus and the “pity- myine” species suggests an independent in- vasion of the southern latitudes by a common ancestor of M. umbrosus and M. chrotorrhinus. North American monophyly and other phylogenetic hypotheses Our assessment of Microtus includes 28 of approximately 65 species of Microtus (Mus- SER and CARLETON 1993), with the addition of Mesoamerican taxa potentially contri- buting to more accurate estimations of rela- tionships (Hırrıs 1996). This larger study did not significantly alter the ML topology previously obtained based on 24 species of Microtus (Conkoy and Cook 2000). Mono- phyly of all North American species in- cluded in this study and of the M. pennsyl- vanicus clades was supported. The expanded analysis suggested two differ- ences based on Kishino-Hasegawa tests. Without Mesoamerican species, we rejected the topology (p = 0.026; Conroy and Cook 2000) which constrained all North Ameri- can taiga voles as monophyletic (sensu HorFrMAnN and KorprL 1985). However, with the inclusion of Mesoamerican species, this hypothesis was not rejected (p = 0.136). The expanded analysis also rejected a to- pology placing M. chrotorrhinus basal to all species of Microtus sampled and M. gre- galis (Russia) within a clade of North American species (p = 0.013). Morphological material for interspecific comparison is abundant, but the phyloge- netic utility of some morphological charac- ters for arvicolines (e.g. tooth pattern) has been criticized due to high variation within and between species (GUTHRIE 1965; ZAKR- ZEWSKI 1985). Despite the availability of standard karyotypes for many species of Microtus, non-differentially stained chro- mosomes have minimal phylogenetic infor- mation because the rate of chromosomal evolution varies greatly among species (CERVANTES et al. 1997; Monı 1987). Indeed, our phylogeny suggests a complicated series of events are needed to explain chromoso- mal rearrangements in Microtus. Species with low diploid numbers are not sister to each other. For example, the mitochondrial data indicate M. oaxacensis (2N = 30) is sis- ter to M. guatemalensis (2N =52), and M. canicaudus (2N = 24) is sister t0 M. town- sendii (2N =50). These relationships sup- port Cervantes et al.s (1997) contention that the interpretation of chromosomal evo- lution based only on standard karyotypes will be difficult. Conservation Mexico has one of the richest mammalian faunas due partially to an overlap of neo- tropical and nearctic biomes (FA and Mo- RALES 1993). Conservation efforts are com- plicated by this high diversity and a variety of threats (CEBALLOsS and NAVARRO 1991). Species of Microtus in the region are nearc- tic relicts that are dependent on high eleva- tion, mesic habitats. Protection of habitat in the mountains of Oaxaca, a region of high mammalian diversity (ArıtA et al. 1997), could help conserve three endemics, M. oaxacensis, M. umbrosus, and M. quasiater. Whether conservation criteria focus on rar- ity, diversity, or degree of endemism, these southern relicts warrant conservation concern. Molecular systematic studies of other ende- mic taxa should be considered in planning conservation efforts in this region (BAKER Mexican Microtus 341 et al. 1995; Cook et al. 2001). Our analysis suggests that temporal scale may be a cru- cial component to interpreting the signifi- cance of biogeographic barriers. Detection of patterns at different temporal scales could help resolve shared histories of taxa in the region (Avıse 1994) and be used to conserve historical associations of flora and fauna. Species of Microtus in Mexico and Guate- mala are not monophyletic but instead are the result of at least three colonizations of this region during the Pleistocene: one by the ancestors of M. oaxacensis, M. quasia- ter, and M. guatemalensis, a second by the ancestor of M. umbrosus, and a third colo- nization apparently gave rise to M. mexica- nus. A lack of fossils inhibits the dating of cladogenic events among these species; however, depth of divergence relative to other splits within Microtus suggest mid- Pleistocene divergence. Morphological si- milarity between extant species (e.g. M. pinetorum and M. quasiater) and formerly widespread (mid to late Pleistocene) taxa that are now extinct (e.g. M. meadensis) suggest isolation by range retraction is a viable hypothesis. Morphological studies also support the shared history of several of the pitymyine species. The weakly-sup- ported sister relationship between M. um- brosus and M. chrotorrhinus has not been predicted previously and should be tested further. Phylogenetic analysis of other Me- soamerican organisms may help identify regions of shared evolutionary history and the role of significant biogeographic bar- riers in promoting diversification in this biologically rich region. Acknowledgements This work was supported by grants from the UAF graduate school, Hayward Fund (UA Museum), National Science Foundation, U.S. Fish and Wild- life Service, and the Consejo Nacional de Ciencia y Tecnologia, Mexico, to CJC, JAC, and FAC. E. HApıy, J. SULLIVAN, B. RıDDLE, K. WINKER and several anonymous reviewers provided criti- cal comments. M. van TuınEn and S. PyoTT pre- pared the German summary. 342 C. J. Conroy et al. Zusammenfassung Die phylogenetische Stellung südlicher Reliktarten von Microtus (Muridae Rodentia) in Nordamerika. Klimatische Fluktuationen führten in Zeiten der Erwärmung zur Isolation einiger Populationen von Ar- ten aus gemäßigten Zonen auf südlichen Bergen. Die südlichsten Arten arvicoliner Nager (Microtus guatemalensis, M. oaxacensis, M. quasiater und M. umbrosus) in Nordamerika mögen Relikte darstel- len, die am Ende des Pleistozäns in den Bergen von Mexico und Guatemala isoliert wurden. Wir unter- suchten mittels Parsimony- und Likelihood-Analysen die kompletten mitochondrialen Cytochrom b-Gen-Sequenzen von 28 Arten der Gattung Microtus. Darunter befanden sich acht eurasische Arten, holarktische M. oeconomus und alle heute lebenden Arten Nordamerikas außer der inselendemischen M. brewern. Die nordamerikanischen Arten waren in der Maximum-Likelihood-Analyse monophyletisch, in der Maximum-Parsimony-Analyse jedoch polyphyletisch. Likelihood-Ratio-Tests und Bootstrap- Analysen wiesen auf eine rasche adaptive Radiation mit kurzen Intervallen zwischen Kladogenese- Ereignissen hin. Die Analysen von Schwestertaxa zeigten jedoch Konstanz in Bootstrap-Analysen oder unterschiedlichen phylogenetischen Auswertemethoden. Nach unseren Ergebnissen war M. quasiater die Schwestergruppe von M. pinetorum und beide bildeten eine Schwestergruppe zu einer Klade aus M. oaxacensis und M. guatemalensis. Die phylogenetische Stellung von M. umbrosus blieb unklar. Die Monophylie der Relikte erwies sich nach Likelihood-Ratio-Tests als unhaltbar, was auf multiple süd- liche Invasionen arvicoliner Nager hinweist. Die phylogenetischen Daten für 17 diese und andere Taxa sollten in Arbeiten zur Arterhaltung in diesen abgelegenen Gebieten berücksichtigt werden. References ARITA, H.T.; FIGUEROA, F.; FRISCH, A.; RoDRI- GUEZ, P.; SANTOS-DEL-PRADOo, K. (1997): Geo- graphical range size and the conservation of Mexican mammals. Conser. Biol. 11, 92-100. AvISE, J. C. (1994): Molecular Markers, Natural History and Evolution. New York: Chapman and Hall. AyALA, F.J. (1999): Molecular clock mirages. Bioessays 21, 71-75. BAKER, A. J.; DAUGHERTY, C. H.; COLBOURNE, R.; McLeEnNnANn, J.L. (1995): Flightless brown ki- wis of New Zealand possess extremely sub- divided population structure and cryptic spe- cies like small mammals. Proc. Natl. Acad. Sciences USA 92, 8254-8258. CAMPBELL, J. A.; Hırrıs, D. M.; LAMAR, W. W. (1989): A new !izard of the genus Norops (Sauna: Iguanidae) from the cloud forest of Hidalgo, Mexico. Herpetologica 45, 232-242. CAMBBELL, J. A. (1984): A new species of Abronia (Sauna: Anguidae) with comments on the her- petologeography of the highlands of southern Mexico. Herpetologica 40, 373-381. CEBALLOS, G.; NAVARRO, L.D. (1991): Diversity and conservation of Mexican mammals. In: Latin American Mammalogy: History, Biodi- versity, and Conservation. Ed. by M.A. MARES and D.J. ScHhmipry. Norman: Univ. Oklahoma Press. Pp. 167-198. CERVANTES, R. A.; MARTINEZ, J; WARD, O. (1997): The karyotype of the Tarabundi vole (Micro- tus oaxacensis: Rodentia), relict tropical arvi- colid. In: Polaco Homenaje al Profesor Ticul Älvarez. Ed. by J. ARROYO CABRALES. Mexico, D. F.: Instituto Nacional de Antroplogia e Historia. Pp. 87-96. Conkoy, C.J.; Cook, J. A. (1999): MtDNA evi- dence for repeated pulses of diversification in arvicoline and murid rodents. J. Mammal. Evol. 6, 221-245. CoNRoy, C. J.; Cook, J. A. (2000): Molecular sys- tematics of a holarctic rodent (Microtus: Muridae). J. Mammalogy 81, 344-359. CooK, J. A} BIDLACK, ArPETEoNKROFAEeHE DEMBOSKI, J. R.; FLEMING, M. A.; RUNCK, A.M.; STONE, K.D.; MAcDonALD, 8.0. (2001): A phylogeographice perspective on endemism in the Alexander Archipelago of the North Pacific. Biol. Conserv. 97, 215- DIT: FA, J. E.; MORALES, L. M. (1993): Patterns of mam- malian diversity in Mexico. In: Biological Di- versity of Mexico: Origins and Distribution. Ed. by T. P. RAMAMOORTHY, R. ByE, A. LoT, and J. FA. New York: Oxford University Press. Pp. 319-361. FREYy, J. K.; CERVANTES, F. A. (1997): Microtus um- brosus. Mammalian Species 555, 1-3. GRAHAM, A. (1999): The Tertiary history of the northern temperate element in the northern Latin American biota. Am. J. Botany 86, 32- 38. GRAHAM, R. W.; GRIMM, E.C. (1990): Effects of global climate change on the patterns of ter- restrial biological communities. Trends Ecol. Evol. 5, 282-292. GRAHAM, R. W.; LUNDELIUS, E.L., Jr.; GRAHAM, M. A.; SCHROEDER, E.K.; TooMEY, R.S. III; ÄANDERSON, E.; BARNOSKY, A. D.; BURNS, J. A.; CHURCHER, C.S.; GRAYSON, D.K.; GUTHRIE, R. D.; HARINGTON, C.R.; JEFFERSON, G.T.; MARTIN, L. D.; McDonALD, H. G.; MORLAN, R. E.; SEMKEN, H. A., Jr.; WEBB, S. D.; WER- DELIN, L.; Wırson, M.C. (1996): Spatial response of mammals to Late Quaternary en- vironmental fluctuations. Science 272, 1601- 1606. GUTHNIE, R.D. (1965): Variability in characters undergoing rapid evolution, an analysis of Microtus molars. Evolution 19, 214-233. HAıı, E.R. (1981): The Mammals of North America. Sec. ed. New York: John Wiley and Sons. Hau, E. R.; Kelson, K.R. (1959): The Mammals of North America. New York: Ronald Press Co. HASEGAWA, M.; KısHINo, H.; YAno, T. (1985): Dat- ing of the human-ape splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol. 22, 160-174. HırLıs, D. M. (1996): Inferring complex phyloge- nies. Nature 383, 130-131. Hınton, M. A.C. (1926): Monograph of the Voles and Lemmings (Microtinae) Living and Ex- tinct. Vol. 1. London: Trustees of the British Museum of London. HOFFMANN, R.S.; KoEpPpL, J. W. (1985): Zoogeo- graphy. In: Biology of New World Microtus. Ed. by R.H. Tamarın. Spec. Publ., Am. Soc. Mammalogists 8, 84-115. Irwin, D.M.; KocHER, T.D.;, Wırson, A.C. (1991): Evolution of the cytochrome b gene of mammals. J. Mol. Evol. 32, 128-144. JONES, J. K., Jr; GENowAys, H. H. (1967): Notes on the Oaxacan vole, Microtus oaxacensis Good- win, 1966. J. Mammalogy 48, 320-321. Jukes, T. H.; CANTOR, C. R. (1969): Evolution of protein molecules. In: Mammalian Protein Metabolism. Ed. by H.N. Munko. New York: Academic Press. Pp. 21-132. KısHino, H.; HAsEGAwA, M. (1989): Evaluation of the maximum likelihood estimate of the evo- lutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. J. Mol. Evol. 29, 170-179. Mexican Microtus 343 MARTIN, R. A. (1974): Fossil mammals from Cole- man IIA fauna, Sumter County. In: Pleisto- cene Mammals of Florida. Ed. by S. D. WEBR. Gainesville: University Presses, Florida. Pp. 35-99. MARTIN, R. A. (1987): Notes on the classification and evolution of some North American fossil Microtus (Mammalia; Rodentia). J. Vert. Pa- leont. 7, 270-283. Monpı, W.S. (1987): Phylogenetic analyses of the chromosomal banding patterns among the Nearctic Arvicolidae. Syst. Zool. 36, 109-136. MOORE, D. W.; JANECEK, L.L. (1990): Genic rela- tionships among North American Microtus (Mammalia: Rodentia). Ann. Carnegie Mus. 59, 249-259. MUSSER, G. G.; CARLETON, M.D. (1993): Family Muridae. In: Mammal Species of the World: A Taxonomic and Geographic Reference. Ed. by D. E. Wırson and D. M. REEDER. Washing- ton, D. C.: Smithsonian Institution Press. Pp. 501-755. RAMAMOORTHY, T. P.; ByE, R.; LOT, A.; FA, J. (eds.). (1993): Biological Diversity of Mexico: Ori- gins and Distribution. New York: Oxford Uni- versity Press. RaAuscH, R.L. (1964): The specific status of the narrow-skulled vole (subgenus Stenocranius Kashchenko) in North America. Z. Säugetier- kunde 29, 343-358. RAuscH, V.R.; RAuscH, R.L. (1974): The chro- mosome complement of the yellow-cheeked vole, Microtus xanthognathus (Leach). Can. ). Gen. Cytogen. 16, 267-272. REPENNING, C. A. (1983): Pitymys meadensis Hib- bard from the Valley of Mexico and the classi- fication of North American species of Pitymys (Rodentia: Cricetidae). J. Vert. Paleont. 2, 471-422. REPENNING, C.; FEJFAR, O.; HEINRICH, W.-D. (1990): Arvicolid rodent biochronology of the north- ern hemisphere. In: The International Sym- posium on the Evolution, Phylogeny, and Biostratigraphy of Arvicolids (Rodentia, Mammalia). Ed. by O. FEIFAR and W.-D. HEın- RICH. Prague: Pfeil-Verlag. Pp. 385-418. SANCHEZ, H.C.; AIVAREZ, R.C.]J.; ROMERO, A.M.L. (1996): Biological and ecological as- pects of Microtus oaxacensis and Microtus mexicanus. Southwest. Nat. 41, 95-98. SULLIVAN, J.; MARKERT, J. A.; KILPATRICK, C. W. (1997): Phylogeography and molecular sys- tematics of the Peromyscus aztecus species group (Rodentia: Munidae) inferred using parsimony and likelihood. Syst. Biol. 46, 426- 440. 344 C. J. Conroy et al. SULLIVAN, J.; ARELLANO, E.; ROGERS, D. S. (2000): Comparative phylogeography of Mesoameri- can highland rodents: concerted versus inde- pendent response to past climatic fluctuations. Am. Nat. 155, 755-768. SWOFFORD, D.L. (1999): PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Meth- ods). Version 4. Sunderland, Massachusetts: Sinauer Associates. Yang, Z. (1994 a): Estimating the pattern of nu- cleotide substitutions. J. Mol. Evol. 39, 105- 111. Yang, Z. (1994b): Maximum likelihood phyloge- netic estimation from DNA sequences with variable rates over sites: approximate meth- ods. J. Mol. Evol. 39, 306-314. ZAKRZEWSKI, R.J. (1985): The fossil record. In: Biology of New World Microtus. Ed. by R. H. TAMARInN. Spec. Publ., Am. Soc. Mam- malogists 8, 1-51. ZINK, R. M.; BLACKWELL, R. C. (1998): Molecular systematics of the scaled quail complex (genus Callipepla). Auk 115, 394-403. Authors’ addresses: C. J. Conroy, Department of Biological Sciences, Stanford University, 371 Serra Mall, Stanford, California 94305-5020, USA (e-mail: chris.conroy@stanford.edu); YOLANDA HORTELANO and F. A. CERVANTES, Depar- tamento de Zoologia, Instituto de Biologia, Uni- versidad Nacional Autönoma de Mexico. Apdo. Ptal. 70-153, Coyoacan, Mexico, D. F. 04510. J. A. Cook, Department of Biological Sciences, Idaho State University, Box 8007, Pocatello, ID 83209-8007, USA. Mamm. biol. 66 (2001) 345-356 Mammalian Biology © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Zeitschrift für Säugetierkunde Original investigation Assessing competition between forest rodents in a fragmented landscape of midwestern USA By T. E. Nupp and R. K. SWIHART Department of Biological Sciences, Arkansas Tech University, Russellville and Department of Forestry and Natural Resources, Purdue University, W. Lafayette, USA Receipt of Ms. 04. 12. 2000 Acceptance of Ms. 13. 06. 2001 Abstract Forests of the agricultural midwestern United States are highly fragmented, and species of small mammals that rely on the remaining forest fragments exhibit non-random distributions. We tested the extent to which interspecific competition between pairs of five species of granivorous forest rodents has influenced the structure of local assemblages occupying forest patches. We used a re- gression technique and incorporated patch and landscape variables in addition to local habitat variables. After accounting for variation in focal species density explained solely by local habitat variables, significant levels of interspecific competition were implicated for Sciurus niger-Tamias striatus. T. striatus also had a negative effect on densities of Peromyscus leucopus in forest patches > 10 ha. Inclusion of patch and landscape variables increased the explanatory power of regressions for 7. striatus and 5. carolinensis, two species generally regarded as sensitive to agriculturally in- duced fragmentation of forest habitat. Even when allowing for habitat selection at larger spatial scales, our results indicated competitive effects comparable to the analysis incorporating only local habitat variables. One difference was a marginal negative effect of 5. carolinensis on Tamiasciurus hudsonicus after accounting for multi-scale selection. Overall, interspecific competition explained a significant proportion of the variation in densities for only three of the 24 potential interactions. In contrast, habitat and landscape features explained 0.37-0.71 of the variation in densities for all species except 5. niger (0.09-0.20). We discuss the roles of competition and habitat fragmentation in mediating the coexistence of forest granivores. Key words: Rodents, competition, fragmentation, habitat selection Introduction Numerous studies have examined the de- gree to which interspecific competition in- fluences the composition of communities (ABRAMS 1988; BENGTSsoNn 1991; reviewed by ConneLL 1983; DRAKE 1990; HALLETT et al. 1983; McIntosH 1995; MınoT and PER- RINS 1986; NEE and May 1992; ROSENZWEIG 1616-5047/01/66/06-345 $ 15.00/0. et al. 1984). In addition, studies of island biogeography (MACARTHUR and WILSON 1967) have shown that groups of insular faunas, resulting from either geological or anthropological processes, typically exhibit highly ordered (1. e., nested) patterns of spe- cies distribution (CuUTLER 1994; McCoy and 346 T. E. Nupp et al. MUSHINSKY 1994; PATTERSON and ATMAR 1986; PATTERSON 1990; WRIGHT et al. 1998). Nested patterns of distribution often result from differential colonization and/or extinc- tion probabilities among species (ATMAR and PATTERSoN 1993; PATTERSON 1990). Fac- tors influencing these colonization or ex- tinction probabilities may include intrinsic characteristics of species such as minimum- area requirements, vagility, specific habitat affinities, and population stability (PATTER- son 1990; PELTONEN and Hanskı 1991). In the absence of habitat fragmentation, differential habitat affinities of species can ameliorate competitive interactions, thus emphasizing the importance of considering local habitat effects in analyses of interspe- cific competition (ABRAMSKY et al. 1979; CRoWELL and Pımm 1976; HALLETT et al. 1983). Theoretical models predict that habi- tat fragmentation may promote coexistence of competing species by permitting inferior competitors to escape spatially, even in the absence of differences in habitat affinities (ABrAMS 1988; HoLmEs and Wırson 1998; Huxer and HastınGgs 1998; MOILANEN and Hanskı, 1995; NEE and May 1992). More- over, empirical evidence suggests that mammalıan species select habitat at multi- ple spatial scales rather than just at local scales, and patch- and landscape-level selec- tion can have important influences on com- munity composition (LINDENMAYER et al. 1999, 2000; SCHWEIGER et al. 2000). In at least some instances local competition in- teracts with landscape-level habitat selec- tion, influencing community structure (GA- BOR and HELLGREN 2000). Habitat selection by species at the local, patch and land- scape-level may be a major mechanism structuring communities in fragmented landscapes; thus, it would be prudent to ex- amine the importance of habitat structure measured at multiple spatial scales on po- pulations before invoking competitive inter- actions as mechanisms structuring a com- munity. Our aim in the present study is to test for competition after incorporating habitat se- lection at multiple spatial scales rather than only at a single spatial scale. We focus our tests on five species of granivorous forest rodents that occur syntopically in our study area in west-central Indiana, USA: white- footed mice (Peromyscus leucopus), eastern chipmunks (Tamias striatus), red squirrels (Tamiasciurus hudsonicus), gray squirrels (Sciurus carolinensis), and fox squirrels (S. niger). These species exhibit a highly nested distribution among forest patches in agricultural landscapes (Nupp and SWIHART 2000), but they vary considerably in the de- gree to which local, patch, and landscape features influence their density or distribu- tion (Nupp 1997). In addition to our general tests for competition, we tested the hypoth- esis (Nupp and SwIHART 1996, 1998) that white-footed mice mice occupying smaller patches are released from interspecific competition in fragmented landscapes due to the absence of larger granivores. This test was accomplished by conducting separate analyses for mice in forest patches < 10 ha and > 10 ha. Material and methods Study area Our study was conducted on the Indian Pine Nat- ural Resources Area in west-central Indiana. This 259 km” area encompasses two major watersheds in Tippecanoe and Warren counties; 82% of the landscape is subjected to cultivation, principally for production of corn and soybeans. Within this agricultural landscape, woodlands comprise 16% of the area and consist of small, more or less iso- lated farmland woodlots and larger wooded ripar- ian strips (SHEPERD and SWIHART 1995). Thirty- five woodlots (0.1-150 ha) and two sites represen- tative of more extensive wooded areas (-1500 ha) were selected for study based on the criteria of re- latively mature, deciduous woody vegetation. These study sites were 30-870 m from their near- est neighboring forest patch. Determination of density Each study site was sampled at least once during spring of 1992 to 1996 by live-trapping. Sher- man"” live-traps (7.5x9.0x30 cm) were placed at 15-m intervals and Tomahawk"”" live-traps Competition of forest granivores in a fragmented landscape (15x15x60 cm) at 30-m intervals on sampling grids established at each study site. All traps were pre-baited for 2 days and followed by 5 days of trapping. Sherman traps were baited with a mix- ture of rolled oats, sunflower seeds and peanut butter, and Tomahawk traps were baited with English walnuts. Abundance estimates of adult mice (>18g) (CumminGs and Vessey 1994) were calculated using program CAPTURE (Oris et al. 1978), and abundance of other adult small mammals (eastern chipmunks > 80 g, red squirrels > 200 g, gray squirrels > 400 g, and fox squirrels > 600 g) was estimated using minimum number known alive (MNA; Krebs 1966). Density estimates were calculated subsequently using either the entire area of woodlots (when the entire area was trapped) or the area of the trapping grid plus a 7.5-m buffer on all sides (for areas that were too large to cover completely, in which case a grid of -2 ha was used). 347 Quantification of habitat features We used standard line transect and point-count sampling to quantify structural characteristics of the local habitat in each forest patch with 24 vari- ables (Tab. 1). Parallel transects were spaced at 15-m intervals on trapping grids. Diameter at breast height (dbh) was measured for all trees >10cm and <1.5m from a transect line. Trees also were classified as hard-mast (i. e., nut) produ- cers, soft-mast (1. e., samara, fruit) producers, con- ifers, or other. Basal area, average dbh, and fre- quency were computed for all trees, snags, hard- mast producers (further separated into Ouercus, Carya, and Juglans), soft-mast producers, and conifers. Counts of stumps, logs, grapevines, and burrows were obtained along transects and ex- pressed in terms of their frequency per 100 m. At 30-m intervals along each transect, we measured vertical vegetative cover from 0-1 m, 1-2 m, and 2-3m above ground using a modified density Table 1. Habitat, patch, and landscape variables used in principal components analysis for detecting variation in granivorous rodent density as a function of habitat measured at multiple spatial scales. The acronym dbh refers to diameter at breast height. Total and sound mast production, as well as mast production excluding walnuts were used only in the analysis with 37 trapping episodes. Except for fractal dimension, patch and landscape variables were transformed using square roots (core area index) or natural logarithms (all other variables) before analysis. Squared terms, centered on mean values, also were included for patch area, proximity, and nearest neighbor dis- tance. See text for details related to each variable. Local Habitat Variables Basal area of all trees Basal area of hard-mast trees Basal area of oaks (Quercus) Basal area of hickories (Carya) Basal area of walnuts (Juglans) Basal area of soft-mast trees Basal area of conifers Basal area of snags Average dbh of hard-mast trees Average dbh of soft-mast trees Average dbh of snags Number of hard-mast trees Number of soft-mast trees Number of grapevines Number of snags Number of stumps Number of logs Vertical cover, 0-1 m Vertical cover, 1-2 m Vertical cover, 2-3 m Percent canopy cover Total mast production (kg/ha) Sound mast production (kg/ha) Mast production excluding walnuts (kg/ha) Patch and Landscape Variables Area of forest patch Perimeter of forest patch Core area index of patch Fractal dimension of patch Proximity of focal patch to other patches Distance to nearest neighboring patch 348 T. E. Nupp etaal. board (Nupps 1977). Percent canopy closure was measured at 30-m intervals using a spherical den- siometer (LEMMoN 1957). Previous studies have documented the impor- tance of production of hard mast on population dynamics of white-footed mice, eastern chip- munks, and tree squirrels (e.g., MCSHEA 2000; Nixon et al. 1975; WoLrr 1996). Production of hard mast was estimated using seed traps placed at 30-m intervals within the trapping grids. Seed traps were constructed of circular plastic bags (1m? area) elevated off the ground. Traps were placed before mast began to fall, in late August or early September of 1993, 1994, or 1995. Mast was collected from the traps in October, sorted by species and soundness, oven-dried, and weighed. Unfortunately, it was not possible to es- timate production of hard mast in each forest patch in the fall preceding trapping. We collected corresponding data on hard mast production for 37 of 61 spring trapping episodes. To quantify patch and landscape characteristics, forest patches within the Indian Pine landscape were digitized from aerial photographs (1:15000) and the digital map was analyzed. We calculated patch area and perimeter, proximity, nearest-neighbor distance, core area index, and fractal dimension (Tab. 1). Proximity is inversely related to isolation of a forest patch and is the sum of patch area divided by the nearest squared edge-to-edge distances between a neighboring patch and the focal patch, for all neighboring patches within a specified radius of the focal patch. A radius of 1 km was used in our analysis. Core area index is a measure of the ratio of inter- ior to edge habitat, calculated as the percent of the total patch area > 50 m from the patch’s edge. Fractal dimension is a measure of shape complex- ity and is equal to two times the logarithm of patch perimeter divided by the logarithm of patch area. Squared terms for patch area, proximity, and nearest-neighbor distance also were included in the analyses, after centering on mean values to reduce collinearity (NETER et al. 1990). In all ana- lyses, patch and landscape variables were loga- rithmically transformed to stabilize variances. Two exceptions were core area index, which was square-root transformed, and fractal dimension, which required no transformation. Computation of interaction coefficients The technique of determining competition coeffi- cients from census data using regression techni- ques was developed by SCHOENER (1974) and CROWELL and Pımm (1976) and has been the sub- ject of considerable debate (ABRAMSsKY et al. 1986; Hastıngs 1987; Pımm 1985; ROSENZWEIG et al. 1985). Recently, Fox and Luo (1996) addressed a shortcoming of the original technique and used perturbation experiments to demonstrate the va- lidity of a modified Schoener-Pimm analysis. Luo et al. (1998) also have applied the technique to identify seasonal fluxes in the intensity of compe- tition between Rattus luteolus and Pseudomys higginsi in Tasmania. We briefly outline the perti- nent statistical procedures below. The equilibrium population size for species i, N’, can be expressed using the Lotka-Volterra equa- tions for competition, N} = K; + 2@;N}, where K; is the carrying capacity of species i in the absence of competitors, N’ is the equilibrium population size of species j, and a; is the per capita effect of species j on the growth rate of species i (1%). For a community at equilibrium, interaction coef- ficients (@;) can be estimated from population censuses using linear regression. Under conditions of heterogeneous habitat, varia- tion in density could be due to differences in habi- tat selection among species. Incorporating the po- tential effects of M local habitat variables, H,. into a model for predicting the equilibrium den- sity of species i yields (after HALLETT 1982) N’ = ßo + 2a; N} + ZßimHm- Several methods for computing estimates of &; have been proposed for this model. CRowELL and Pımm (1976) used stepwise multiple linear regression performed on principal components for habitat. After habitat components entered the model, densities of other species were allowed to enter the equation. Ro- SENZWEIG et al. (1984) also used a stepwise re- gression performed on habitat components as ex- planatory variables, with density as the response variable. Residuals were saved from each regres- sion, and then a series of regressions were con- ducted on the residuals, with each species taking its turn as the dependent variable. Both the Cro- well-Pimm method and the residual analysis method attempt to estimate interaction coefli- cients between species after accounting for varia- tion in species densities that can be explained by habitat variables. This is a conservative approach, because some segregation due to habitat may be a result of competitive interactions. ROSENZWEIG et al. (1985) proposed a “free” regression ap- proach in which stepwise regression is conducted on habitat variables and species densities simulta- neously, thus permitting species interaction terms to enter the model before habitat components. Dependence of a,; estimates on the variances of species densities is a problem of all of the above methods for estimating interaction coefficients. Fox and Luo (1996) addressed this problem by Competition of forest granivores in a fragmented landscape standardizing density estimates for each species so that mean standardized density equalled O and variance equalled 1. They applied this approach to estimate interaction coefficients for a small mammal assemblage in which field removal ex- periments also were conducted. Interaction coef- ficients estimated using standardized densities matched coefficients computed from removal ex- periments quite well (Fox and Luo 1996). Luo et al. (1998) also used standardized densities to ob- tain reasonable estimates of interaction coeffi- cients for a pair of small mammal species cen- sused across three seasons in Tasmania. Thus, standardized densities should be used instead of unstandardized densities when estimating compe- titive effects. In our analysis, we standardized the density esti- mates so that each species’ density had a mean of 0 and a standard deviation of 1. We then used these standardized densities as response variables in stepwise linear regressions (F>2.0 to enter the model) adhering to the Crowell-Pimm, resi- dual analysis, and free regression methods. A second problem not addressed by published methods for estimating interaction coefficients is their focus on local habitat selection by species. In light of recent evidence indicating the occur- rence of habitat selection at larger spatial scales (LINDENMAYER et al. 1999, 2000; SCHWEIGER et al. 2000), we modified the model of HALLETT (1982) to incorporate the effects of Q patch effects, P,, and R landscape-level effects, L,, on the density of species 1: N’= Bo Ir Za,N;+ en Ir ZyigPa Ar Do: We conducted a separate analysis for the complete set of 61 censuses, and for the subset of 37 cen- suses for which data on hard-mast production also were available. To reduce dimensionality, local ha- bitat, patch, and landscape-level variables were subjected to principal components analysis (PCA). Only scores for principal components with eigenvalues > 1 were used in regression models. In each analysis, we estimated interaction coefficients in the following manner: Two separate regressions were performed for each species of sciurid. In one regression we used as explanatory variables princi- pal components derived only from local habitat varlables. In the other regression we used principal components derived from local habitat variables and principal components derived from pooled patch and landscape variables. Our motivation for partitioning the explanatory variables was to de- termine whether variation in species densities could be explained by variables operating at multi- ple spatial scales, and to assess the degree to which inclusion of patch and landscape metrics affected estimates of interspecific interaction. 349 Results The 61 spring trapping episodes yielded captures of 1669 white-footed mice, with mice captured in all trapping episodes. In addition, we captured 207 fox squirrels in 43 episodes, 264 eastern chipmunks in 47 episodes, 78 gray squirrels in 14 episodes, and 31 red squirrels in 12 episodes. Trap- ping in the spring following estimation of hard-mast production at 37 sites yielded captures of 961 white-footed mice at 37 sites, 148 fox squirrels at 29 sites, 147 eastern chipmunks at 28 sites, 46 gray squirrels at 7 sites, and 18 red squirrels at 6 sites. The number of sites at which we caught gray and red squirrels was marginal for use in subsequent regression analysis. Southern flying squirrels (Glaucomys vo- lans) were excluded from analysis, as they were captured only in the 2 extensive wood- lands and 3 of the largest forest patches. When all 61 episodes were considered, PCA on the local habitat variables yielded seven usable principal components (1. e., ei- genvalues >1), and these components explained 80.2% of the total variation of the original variables. PCA on the patch and landscape-level variables yielded three usable components that together explained 90.3% of the total variation. When the 37 sites with data on mast production were considered separately, PCA on the local ha- bitat variables yielded eight usable compo- nents, and these components explained 83.4% of the total varıation. PCA on the patch and landscape-level variables yielded three usable components that together ex- plained 84.3% of the total variation. For each species, we obtained significant re- gression models relating standardized densi- ties to the principal components derived from local habitat variables. These models explained 9-64% of the variance in standardized density estimates when all trapping episodes were used and 20-60% of the variance when only the 37 episodes with data on hard-mast production were used (Tab. 2). When we constructed regression models using principal components from both local habitat variables and patch and 350 T. E. Nupp et al. Table 2. Coefficients of multiple determination (adusted R°) and P values (in parentheses) for regressions relat- ing standardized density to principal components of either local habitat variables alone or in combination with principal components of patch and landscape variables. Separate analyses were conducted for all spring trapping episodes (n=61) and for those episodes for which hard-mast production was estimated the preceding fall (n = 37). For white-footed mice, too few trapping episodes occurred at sites > 10 ha to permit analysis. Crowell-Pimm and Residual Analysis Methods Local Habitat Only Species All Trapping Episodes White-footed mice, < 10 ha White-footed mice, > 10 ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels 0.41 (0.001) 0.64 (0.004) 0.34 (0.001) 0.09 (0.030) 0.34 (0.001) 0.37 (0.001) Episodes with Mast Data White-footed mice, < 10 ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels 0.42 (0.005) 0.35 (0.002) 0.20 (0.025) 0.54 (0.001) 0.60 (0.001) landscape-level variables, substantial in- creases in adjusted R” values occurred for eastern chipmunks (increases of 0.09-0.12) and gray squirrels (increases of 0.08-0.19), moderate improvements in R” values were noted for red squirrels (increases of 0.03- 0.08), and little or no change in R? values occurred for fox squirrels (0.00-0.03) and white-footed mice (-0.03-0.00) (Tab. 2). Estimates of interaction coefficients gener- ally were similar within a given set of data, irrespective of the regression method used. Generalized community matrices (HALLETT 1982) for interaction coefficients estimated after accounting for components of local habitat indicated significant negative effects of eastern chipmunks on densities of white- footed mice in forest patches >10 ha and reciprocal negative effects between chip- munks and fox squirrels (Tab. 3). When the effects of patch and landscape-level vari- ables were incorporated into the estimation procedure, the negative interactions noted previously with the local habitat variables were retained (Tab. 4). In addition, we noted significant positive effects of mice Free Regression Method Local Habitat Only Local + Patch + Landscape Local + Patch + Landscape 0.41 (0.001 0.61 (0.012 0.46 (0.001 ) 0.41 (0.001) ) ) 0.09 (0.030) ) ) ( 0.64 (0.004) 0.36 (0.001) 0.09 (0.030 0.41 (0.001) 0.61 (0.012) 0.46 (0.001) 0.09 (0.030) 0.45 (0.001) 0.45 (0.001) 0.45 (0.001 0.45 (0.001 ) 0.37 (0.001) 0.37 (0.001) 0.42 (0.003 0.42 (0.005 0.46 (0.001 0.37 (0.001 ) (0.005) 0.42 (0.003 ) (0.001) 0.20 (0.025) 0.27 (0.006) ) (0.001) ) (0.001) (0.003) 0.46 (0.001) 0.30 (0.006) 0.74 (0.001) 0.64 (0.001) 0.71 (0.001 0.55 (0.001 0.63 (0.001 0.60 (0.001 and fox squirrels, and a negative effect of gray squirrels, on density of red squirrels (Tab. 4). Discussion Because of the importance of patch and landscape features to density of some of the species in our assemblage, estimation of competitive interactions should account for variation in density due to habitat fea- tures measured at multiple spatial scales. Of the five species we examined, gray squir- rels and eastern chipmunks are the most sensitive with respect to habitat fragmenta- tion (Nupp and SwIHART 1998, 2000), whereas red squirrels, fox squirrels, and white-footed mice appear to be progres- sively less sensitive (BAYnE and HoBson 2000; Nupp and SWIHART 1998, 2000; SWI- HART and Nupp 1998; GOHEEN and SWIHART unpubl. data). In accord with these differ- ences, models for gray squirrels and eastern chipmunks exhibited the greatest increase in variation explained when patch and land- Competition of forest granivores in a fragmented landscape 351 Table 3. Generalized community matrices for five species of granivorous forest rodents in west-central Indiana, U.S.A. Entries indicate the per capita effect of the column species on the row species, taking into account the effects of local habitat on species density. For a given species pair, interaction coefficients are listed in the fol- lowing order: Crowell-Pimm, residual analysis, free regression. All interaction coefficients in the table exhibited P values < 0.05. White-footed mice All Trapping Episodes White-footed mice <10ha >10ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels Episodes with Mast Data White-footed-mice <10ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels scape components were included in regres- sions.. Smaller increases in explanatory power were evident for red squirrels when patch and landscape components were in- cluded in regression models. For fox squir- rels and mice, inclusion of patch and land- scape components contributed virtually nothing to the models’ explanatory power. Density of white-footed mice actually is in- versely related to patch area (Nupp and SWIHART 1996, 2000), indicating a positive response to habitat fragmentation. Habitat factors measured at multiple spatial scales explained a substantial amount of the total variation in species abundances. In contrast, evidence of strong competitive ef- Eastern chipmunks Fox squirrels Gray squirrels Red squirrels fects among species of small mammals in our study was relatively sparse; only three out of 24 possible species interactions were consistently significant for each type of model constructed (local habitat variables only; habitat, patch and landscape variables combined). Local habitat affinities and lar- ger-scale responses to agriculturally in- duced fragmentation of habitat appear to be the principal determinants of community structure in forest patches, with interspeci- fic interactions relegated to a secondary role. SwIHART and Nurp (1998) drew the same conclusion based on spatially explicit simulation models of gray squirrel, fox squirrel, and red squirrel populations. 352 T. E. Nupp et al. Table 4. Generalized community matrices for five species of granivorous forest rodents in west-central Indiana, U.S.A. Entries indicate the per capita effect of the column species on the row species, taking into account the effects of local habitat, patch, and landscape-level variables on species density. For a given species pair, interac- tion coefficients are listed in the following order: Crowell-Pimm, residual analysis, free regression. All interaction coefficients in the table exhibited P values < 0.05 except red-gray, which was 0.06. White-footed mice Eastern chipmunks Fox squirrels Gray squirrels Red squirrels All Trapping Episodes White-footed mice <10ha >10 ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels Episodes with Mast Data White-footed-mice <10ha Eastern chipmunks Fox squirrels Gray squirrels Red squirrels A possible exception to secondary effects of competition was the mutually negative in- teraction observed between fox squirrels and eastern chipmunks. Both eastern chip- munks and fox squirrels inhabit woodlands throughout the study area (Nupp and Swı- HART 2000). This pattern of co-occurrence places them in potential conflict for a com- mon food source, namely hard mast (Ko- PROWSKI 1994 a; SNYDER 1982). Consistent with this hypothesis, densities of fox squir- rels and eastern chipmunks increase in re- sponse to principal components of habitat variables characterizing basal area of hard mast trees and mast production, respec- tively (NupP 1997). A negative effect of eastern chipmunks also was noted on densities of white-footed mice in large (2 10 ha) forest tracts. Both of these species occur syntopically throughout the study area (Nupp and SWIHART 2000). How- ever, previous studies have demonstrated that eastern chipmunks are sensitive to fragmentation and exhibit lower survival in small forest fragments than in continuous tracts of forest (HEnEIN et al. 1998; NUpP Competition of forest granivores in a fragmented landscape and SWIHART 1998), potentially leading to local extinctions from fragments (HENDER- son et al. 1985). Thus, competitive effects of eastern chipmunks on white-footed mice may be dependent on patch attributes and landscape context. We observed a negative effect of chipmunks on mice in large forest patches but not in small (<10 ha) forest patches, a finding consistent with the hy- pothesis that mice experience release from competition with larger granivores in small forest patches (Nupp and SWIHART 1998, 2000). Competitive release in fragments ap- pears to be common for small mammals with generalist habitat requirements, although social structure may regulate a species’ ability to respond to the absence of competing species (DEBInsKI and Horr 2000). The relatively minor role of interspecific in- teractions in determining current popula- tion densities does not imply that competi- tion was unimportant in the relatively continuous forest that characterized pre- settlement Indiana. Historical influences of- ten are represented in current distributions of species and are difficult to ıdentify using current observational data (ConnELL 1983; DRAKE 1990; Keır et al. 1995). Local com- petition also can influence geographic ranges and the composition of regional bio- tas (Brown et al. 2000). Fragmentation of Indiana’s forest began approximately 150 years ago (PETTY and Jackson 1966). Zusammenfassung 353 Competition could have played an impor- tant role in structuring the distribution and abundance of granivorous rodents in the previously unfragmented forest, but it seems likely that deforestation and conco- mitant reductions in area and increases in isolation of the remaining forest patches have played an increasingly important role in the last century. Our results thus support the notion that observed interactions be- tween two species may be a function of properties intrinsic to the species and, per- haps more importantly, of properties of the landscape in which they co-occur (DANIEL- son 1991; DEBInsKI and Horr 2000). Acknowledgements We thank the numerous landowners in west-cen- tral Indiana who graciously permitted us to use work in their woodlots. Also, we thank $S. AyErs, T. BowmAan, R. CHAPMAN, D. CupDpy, L. CuRTs, G. DAHLE, J. Davıs, T. FAIRCHILD, R.GEHL, L. HUGHEs, J. HuUNT, J. Ivan, D. LoOTTER, R. MiL- LER, J. PARSLOWw, R. RıTcHEY, R. RopDTs, S. SMAL- LIDGE, C. WEBSTER, R. WILLIAMS, and J. WINTERS for field assistance. G. R. PARKER, P. M. WASER, H. P. Weeks, Jr., and two anonymous reviewers provided helpful comments on drafts of the manuscript. Financial support was provided by Purdue University and by National Research Initiative Competitive Grants Program/USDA award 93-37101-8702. This is contribution 16411 to Purdue University Agricultural Research Pro- grams. Abschätzungen zur Konkurrenz von im Wald lebenden Rodentiaspecies in einer fragmentierten Landschaft im mittleren Westen der USA Viele Wälder im landwirtschaftlich intensiv genutzten mittleren Westen der USA sind stark fragmen- tiert, und die Kleinsäugerarten der Habitatinseln zeigen eine nicht zufällige Verteilung. Wir haben den Grad der Konkurrenz zwischen fünf Arten von Samen fressenden Waldnagetieren untersucht, um den Einfluß auf die Struktur des lokalen Vorkommens in Waldinseln abzuschätzen. Dazu wurde die Technik der Regressionskalkulationen erweitert, um Landschaftsvariablen zusätzlich zu den Ha- bitatvariablen einzubeziehen. Nachdem die Varianz der Dichte von interessierenden Arten durch lo- kale Habitatvariablen erklärt wurde, sind signifikante Konkurrenzeffekte für das Artenpaar Sciurus niger-Tamias striatus gefunden worden. 7. striatus hat außerdem eine negative Wirkung auf die Dichte von Peromyscus leucopus in Waldinseln, die größer sind als 10 ha. Die Einbeziehung von Patch-Fläche und Landschaftsvariablen erhöht den Erklärungswert der Regression von T. striatus 354 T. E. Nupp et al. und 5. carolinensis, zwei Arten, die als sensibel gegenüber Fragmentierung angesehen werden. Auch wenn die Habitatwahl in einem größeren Maßstab einbezogen wurde, zeigten unsere Ergebnisse Konkurrenzeffekte, die vergleichbar waren mit der Analyse, die nur lokale Habitatvariablen beinhal- tete. Eine Ausnahme war eine geringe negative Wirkung von S. carolinensis auf Tamiasciurus hudso- nicus nachdem die Habitatwahl auf unterschdiedlicher räumlicher Skala einbezogen wurde. Insge- samt hat die interspezifische Konkurrenz nur für drei von den 24 möglichen Interaktionen einen signifikanten Einfluß auf die Dichte. Im Gegensatz dazu haben Habitat und Landschaftsstruktur 0.37-0.71 der Dichtevarianz für alle Arten außer 5. niger (0.02-0.20) erklärt. Wir diskutieren die Rolle von Konkurrenz und Habitatfragmentierung auf die Koexistenz von Waldsamenfressern. References ABRAMS, P. A. (1988): Resource productivity-con- sumer species diversity: simple models of competition in spatially heterogeneous envi- ronments. Ecology 69, 1418-1433. ABRAMSKY, Z.; DyER,M.I; HARRISON, P.D. (1979): Competition among small mammals in experimentally perturbed areas of short- grass prairie. Ecology 60, 530-536. ABRAMSKY, Z., BOWERS, M. A.; ROSENZWEIG, M.L. (1986): Detecting interspecific competition in the field: testing the regression method. Oikos 47, 199-204. ATMAR, W.; PATTERSoN, B. D. (1993): The measure of order and disorder in the distribution of species in fragmented habitat. Oecologia 96, 373-382. BAYnE, E.; Hoßson, K. (2000): Relative use of contiguous and fragmented boreal forest by red squirrels (Tamiasciurus hudsonicus). Can. J. Zool. 78, 359-365. BENGTSson, J. (1991): Interspecific competition in metapopulations. Biol. J. Linnean Soc. 42, 219-237. Brown, J. H.; Fox, B. J.; Keır, D. A. (2000): As- sembly rules: desert rodent communities are structured at scales from local to continental. Am. Nat. 156, 314-321. CASWELL, H.; COHEN, J. E. (1991): Disturbance, in- terspecific interactions and diversity in meta- populations. Biol. J. Linnean Soc. 42, 193-218. ConNELL, J. H. (1983): On the prevalence and re- lative importance of interspecific competition: evidence from field experiments. Am. Nat. 122, 661-696. CROWELL, K.L.; Pımm, S. L. (1976): Competition and niche shifts of mice introduced onto small ıslands. O1kos 27, 251-258. CUMMINGS, J. R.; VESSEY, S.H. (1994): Agricultur- al influences on movement patterns of white- footed mice (Peromyscus leucopus). Am. Midl. Nat. 132, 209-218. CUTLER, A. H. (1994): Nested biotas and biologi- cal conservation: metrics, mechanisms, and meaning of nestedness. Landscape and Urban Planning 28, 73-82. DANIELSOoN, B. J. (1991): Communities in a land- scape: The influence of habitat heterogeneity on the interactions between species. Am. Nat. 138, 1105-1120. DANIELSON, B. J.; GAINES, M. S. (1987): The influ- ences of conspecific and heterospecific resi- dents on colonization. Ecology 68, 1778-1784. DeBiınskt, D.M.; Hort, R.D. (2000): A survey and overview of habitat fragmentation experi- ments. Conserv. Biol. 14, 342-355. DoLan, P.G.; CARTER, D.C. (1977): Glaucomys volans. Mammalian Species Number 78. Am. Soc. Mammalogy. DRAKE, J. A. (1990). Communities as assembled structures: Do rules govern pattern? Trends in Ecol. and Evol. 5, 159-164. DuEsER, R. D.; HALLETT, J. G. (1980): Competi- tion and habitat selection in a forest-floor small mammal fauna. Oikos 35, 293-297. DyTHam, C. (1995): Competitive coexistence and empty patches in spatially explicity me- tapopulation models. J. Anim. Ecol. 64, 145— 146. Fox, B. J.; Luo, J. (1996): Estimating competition coefficients from census data: a re-examina- tion of the regression technique. Oikos 77, 291-300. GABOR, T. M.; HELLGREN, E. C. (2000): Variation in peccary populations: landscape composi- tion or competition by an invader? Ecology 81, 2509-2524. GRANT,P.R. (1972): Interspecifice competition among rodents. Ann. Rev. Ecol. Syst. 3, 79- 106. HALLETT, J. G. (1982): Habitat selection and the community matrix of a desert small-mammal fauna. Ecology 63, 1400-1410. Competition of forest granivores in a fragmented landscape HALLETT, J. G.; O’ConNELL, M. A.; HonEYCUTT, R.L. (1983): Competition and habitat selection: test of a theory using small mammals. Oikos 40, 175-181. Hasrtincs, A. (1987): Can competition be detected using species co-occurrence data? Ecology 68, 117-123. HENDERSON, M. T.;, MERRIAM, G.; WEGNER, ]. (1985): Patchy environments and species sur- vival: chipmunks in an agricultural mosaic. Biol. Conserv. 31, 95-105. HENEIN, K.; WEGNER, J.,; MERRIAM, G. (1998): Po- pulation effects of landscape model manipula- tion on two behaviourally different woodland small mammals. Oikos 81, 168-186. Hoıımzs, E.E.; Wırson, H.B. (1998). Running from trouble: long-distance dispersal and the competitive coexistence of inferior species. Am. Nat. 151, 578-586. HvuxEL, G. R.; Hastıings, A. (1998): Population size dependence, competitive coexistence and ha- bitat destruction. J. Anım. Ecol. 67, 446-453. KELT, D. A.; TAPER, M. L.; MESERVE, P.L. (1995): Assessing the impact of competition on com- munity assembly: a case study using small mammals. Ecology 76, 1283-1296. Kress, C. J. (1966): Demographic changes in fluc- tuating populations of Microtus californicus. Ecol. Monogr. 36, 239-273. KoPrRowsKk1, J. L. (1994 a): Sciurus niger. Mamma- lian Species Number 479. Am. Soc. Mammal- OBY. KoProwsk1, J.L. (1994b): Sciurus carolinensis. Mammalian Species Number 480. Am. Soc. Mammalogy. LEmmon, P. E. (1957): A new instrument for mea- suring forest overstory density. J. Forestry 55, 667-668. LINDENMAYER, D.B.; CUNNINGHAM,R.B.; PoPE, M.L.; DonneLty, C.F. (1999): The response of arboreal marsupials to landscape context: a large-scale fragmentation study. Ecol. App. 9, 594-611. LINDENMAYER, D. B.; MCCARTHY, M. A.; PARRIS, K. M.; Pope, M.L. (2000): Habitat fragmentation, landscape context, and mammalian assem- blages in southeastern Australia. J Mammal- ogy. 81, 787-797. Luo, J.; Monamy, V.; Fox, B.J. (1998): Competi- tion between two Australian rodent species: a regression analysis. J. Mammalogy. 79, 962- SL MACARTHUR, R.H.; Wırson, E.O. (1967): The Theory of Island Biogeography. Monographs in Population Biology. Princeton, New Jersey: Princeton Univ. Press. 355 McCoy, D. E.; MusHInsKy, H.R. (1994): Effects of fragmentation on the richness of verte- brates in Florida scrub habitat. Ecology 75, 446-457. McInTosH, R. P. (1995): H. A. Gleason’s ‘individ- ualistic concept’ and theory of animal com- munities: a continuing controversy. Biol. Rev. 70, 317-357. MCSHEA, W. J. (2000): The influence of acorn crops on annual variation in rodent and bird populations. Ecology 81, 228-238. MinoT, E. O.; PERRINS, C.M. (1986): Interspecific interference competition-nest sites for blue and great tits. J. Anim. Ecol. 55, 331-350. MOILANEN, A.; HAnskTı, I. (1995): Habitat destruc- tion and coexistence of competitors in a spa- tially realistic metapopulation model. J. Anim. Ecol. 64, 141-144. MUMFORD, R.E.; WHITAKER, J. O., Jr. (1982): Mammals of Indiana. Bloomington: Indiana Univ. Press. NEE, S.; MAy, R.M. (1992): Dynamics of metapo- pulations: habitat destruction and competitive coexistence. J. Anim. Ecol. 61, 37-40. NETER, J.; WASSERMAN, W.; KUTNER, M.H. (1990): Applied Linear Statistical Models, 3" ed. Bos- ton, Massachusetts: Richard D. Irwin. Nixon, C. M.; McCain, M. W.; DONOHUE, R. W. (1975): Effects of hunting and mast crops on a squirrel population. J. Wildl. Manage. 39, 1-25. Nupops, T. D. (1977): Quantifying the vegetative structure of wildlife cover. Wildl. Soc. Bull. 5, 113-117. Nupp, T. E. (1997): Population dynamics and com- munity structure of granivorous forest rodents in a fragmented landscape. Ph. D. diss., Pur- due University, West Lafayette, Indiana. Nupp, T. E.; SWIHART, R.K. (1996): Effect of for- est patch area on population attributes of white-footed mice (Peromyscus leucopus) in fragmented landscapes. Can. J. Zool. 74, 467- 42. Nupp, T. E.; SWIHART, R.K. (1998): Effects of for- est fragmentation on population attributes of white-footed mice and eastern chipmunks. ]. Mammalogy. 79, 1234-1243. Nupp, T. E.; SWIHART, R.K. (2000): Landscape-level correlates of small-mammal assemblages in forest fragments of farmland. J. Mammalogy. 81, 512-526. Oris, D. L.; BURNHAM,K.P; WHITE, G. C.; ÄNDERSON, D. R. (1978): Statistical inference from capture data on closed animal popula- tions. Wildl. Monogr. No. 62. PATTERSoN, B. D. (1990): On the temporal devel- 356 T. E. Nupp et al. opment of nested subset patterns of species composition. Oikos 59, 330-342. PATTERSON, B. D.; ATMAR, W. (1986): Nested sub- sets and the structure of insular mammalıan faunas and archipelagos. Biol. J. Linnean Soc. 29, 65-82. PELTONEN, A.; HAnskı, I. (1991): Patterns of is- land occupancy explained by colonization and extinction rates in shrews. Ecology 72, 1698- 1708. PETTY, R. O.; JAcKson, M. T. (1966): Plant commu- nities. In: Natural Features of Indiana. Ed. By A. A.LinDZEy. Indiana Acad. Sci. Pp. 264- 296. Pımm, S. L. (1985): Estimating competition coeffi- cients from census data. Oecologia 67, 588-590. ROSENZWEIG, M.L.; ABRAMSKY, Z.; BRAND, S. (1984): Estimating species interactions in het- erogeneous environments. Oikos 43, 329-340. ROSENZWEIG, M. L.; ABRAMSKY, Z.; KOTLER, B.; MITCHELL, W. (1985): Can interaction coeffi- cients be determined from census data? Oe- cologia 66, 194-198. SCHOENER, T. W. (1974): Competition and the form of the habitat shift. Theoret. Pop. Biol. 6, 265-307. SCHWEIGER, E. W.; DIFFENDORFER, J. E.; HoLT, R. D.; PIEROTTI, R.; GAINES, M. S. (2000): The interac- tion of habitat fragmentation, plant, and small mammal succession in an old field. Ecol. Monogr. 70, 3-400. SHEPERD, B. F.; SWIHART, R. K. (1995): Spatial dy- namics of fox squirrels (Sciurus niger) in frag- mented landscapes. Can. J. Zool. 73, 2098- 2105. SIMBERLOFF, D.; MARTIN, J. L. (1991): Nestedness of insular avifaunas: simple summary statistics masking complex species patterns. Ornis Fen- nica 68, 178-192. SMITH, A. T.; PEAcocK, M. M. (1990): Conspecific attraction and the determination of metapo- pulation colonization rates. Conserv. Biol. 4, 320-323. SNYDER, D. P. (1982): Tamias striatus. Mammalian Species Number 168. Am. Soc. Mammalogy. SWIHART, R. K.; Nupp, T. E. (1998): Modeling po- pulation responses of North American tree squirrels to agriculturally induced forest frag- mentation. In: Ecology and Evolutionary Biol- ogy of Tree Squirrels. Ed. By M. A. STEELE, J. F. MERRITT, and D. A. ZEGERs. Virginia Mu- seum Nat. History Special Pub. 6, 1-20. WOLFF, J. ©. (1996): Population fluctuations of mast-eating rodents are correlated with pro- duction of acorns. J. Mammalogy. 77, 850-856. WRIGHT, D. H.; PATTERSON, B. D.; MIKKEL- son, G. M.; CUTLER, A.; ATMAR, W. (1998): A comparative analysis of nested subset patterns of species composition. Oecologia 113, 1-20. Authors’ adresses: THoMAsS E. Nupp, Department of Biological Sciences, McEver Hall, Arkansas Tech University, Russellville, AR 72801-2222 USA (e-mail: Tom.Nupp@mail.atu.edu) ROBERT K. SWIHART, Department of Forestry and Natural Resources, Purdue University, W. Lafa- yette, IN 47907-1159 USA Mamm. biol. 66 (2001) 357-364 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Original investigation Mammalian Biology Zeitschrift für Säugetierkunde The daily activity period of the brown hare (Lepus europaeus) By A. J. F. HoLLEY Department of Biological Sciences, University of Durham, Durham, U.K. Receipt of Ms. 17. 10. 2000 Acceptance of Ms. 15. 05. 2001 Abstract The times of 241 entries into and 573 exits by brown hares from their forms on a 65 ha area of the Somerset Levels, South West England, were recorded over a ten year period to reveal the variation during the year of the duration of the daily activity period. In December when nightlength was 16 h, activity was exclusively nocturnal; intervals between entry and sunrise were longer and less consistent than those between sunset and exit. In June with nightlength down to 7.4 h, activity was part diurnal, post-sunrise and pre-sunset, for a total of 6 h. From a peak duration of 14.5 h in December, the activity period declined to 12 h in the third week of March, mirroring night time dura- tion. It then increased to 13.5 h in late midsummer before again reducing in the autumn. It is sug- gested that the proximate cause of this circannual cyclicity is an aversion to daylight activity. Key words: Lepus europaeus, brown hare, activity period Introduction Research on the circadian and circannual ac- tivity of the brown hare (Lepus europaeus Pallas, 1778) has hitherto been effected either by direct observation of hares on their foraging grounds (HoMmoLkA 1986) and tra- velling to and from them (MATUSZWESKI 1981) or by radiotelemetrical monitoring of their movements (PEpın and CARGNELUTTI 1994), the one exception being HoLLEY and GREENWOoD (1984) where, for part of the year, hares were timed exiting their forms in the evening. Such studies all suggest an an- nual cycle in the circadian activity period. The brown hare’s day and that of some other leporids is divided into two distinet periods. One, comprising all or a large part of the daylight hours, is spent crouched in its form 1616-5047/01/66/06-357 $ 15.00/0. or resting place. During much of the other, comprising mostly the night hours, the hare travels, feeds and interacts socially with con- specifics. These activities are, however, inter- spersed with shorter and less formal periods of rest. The combination of the clear-cut dis- tinction between active and inactive periods and the exclusively above-ground lifestyle of the brown hare make it one of the best candidates of the smaller mammals for di- rect observational field studies of its circa- dian rhythm. The objects of this study were, by monitoring the times of entry into and exit from forms throughout the year, to de- fine any annual cycle in the duration of the daily activity period and to consider the rea- son or reasons for it. 358 A. J. F. HoLLey Material and methods The observations reported here were part of a lar- ger study of behaviour of the brown hare (Hor- LEY 1992) on approx. 65 ha, divided between 17 fields in the north western sector of the Somerset Levels, close to Brent Knoll, Somerset, U.K. (51°16 N, 2°58° W). Fields were predominantly permanent pastures with a minority being ploughed in rotation for cereal growing. There were few hedges, fields being divided by drainage channels. Observations were made from Septem- ber 1977 to September 1987 from windows 9m above ground level in the roof space of my house within the study area. Optical equipment em- ployed consisted of 7x50 and 15x80 binoculars, a Zeiss Jena 20/40x 80 binocular telescope and a Celestron 11 catadioptic astronomical telescope, giving magnifications in excess of 100x. The latter could be coupled to a Canon Fl camera as a tele- photo lens with a focal length of 2800 mm. With the equipment it was possible to identify and sex individual animals. Occupied forms were under observation over distances up to 400 m. I recorded the times (Greenwich Mean Time throughout the year) of entry into or departure from forms. All entries and departures, except temporary departures, were included whenever observed although most were obtained during ob- servation periods conducted for the purpose, commencing at least an hour before the earliest or closing at least an hour after the latest ex- pected entry or departure, as the case might be. Usually, hares were noted either entering or de- parting their forms but not both on the same day. These will be referred to as once-a-day (OD) ob- servations. In some cases, which will be referred to as twice-a-day (TD) observations, I recorded both the time a hare entered its form and the time it departed. Preliminary observation having con- firmed the popular notion that hares become more noticeable in March, TD data were col- lected particularly intensively in March of each year. The OD data were collected throughout the study, the TD data being collected during the second five-year period. The population of the study area was thought to be between 5 and 15 adult hares each year. No individual hare was seen in more than four successive years. At some times of the year it was possible to re- cord six or more entries or departures in one ob- servation period. In cases of doubt, when a hare appeared to be entering a new form but might merely be taking a rest before moving on, I only counted an entry if the hare was still in occupa- tion after all other visible hares were also in their forms. For departures, where in doubt, I only in- cluded hares previously observed in occupation of the same form before noon on the day. Tem- porary departures caused, for example, by disturbance from humans, cattle or other hares were disregarded. Apart from such temporary ex- its, hares remained in their forms until the normal departure time, the only exception being when a doe was very close to, or in, oestrus; bucks could then be active throughout daylight. The interval between departing and entering forms is defined as the activity period although it may include within it periods of rest. Likewise, the interval be- tween entering and departing forms is defined as the inactive period. Employing standard optical equipment, it was possible by sweeping the study area to observe departures throughout the year but in December and early January, when the nights were longest, it was not normally possible to detect entries more than 40 mins before sunrise. Because of this problem, I mounted special watch, between 15 December 1985 and 6 January 1986, on one form occupied regularly by the same hare with the principal object of recording entry times. The observed times of entries and exits were dif- ferenced from the day’s sunrise/sunset values, averaged for each day and then averaged over the pertinent week. These weekly differentials in- dicate the time in relation to sunrise/sunset at which activity usually started or stopped. Those results having indicated a cyclic behaviour, har- monic functions were fitted to the a.m. and p.m. weekly mean differentials. The fitted seasonal curves have the form y = A.cos(wx-®) + C Where w=2n/T, T=52, the period in weeks of the function, y, andx = 1,2,3,.... 52, the weekly in- dependent variable. A, C, ©, are constants deter- mined by an Ordinary Least Squares fit to the weekly averages, A being the amplitude, C the long run average of the variable y, and © the phase of the cosine function. The phase can be ex- pressed in terms of the week, P. marking the peak of the cosine function BSOmWDr A test of statistical significance of the computed function, y, can be carried out on the constant A by an approximate F test with 2 and N-3 degrees of freedom, (N = number of weekly observations, somewhat less than 52). Likewise, the significance of the displacement of one cyclic seasonal from the other can be determined by an approximate z test of the difference between the two values of P. The choice of positive and negative for the dif- ferentials was arbitrary. Results Altogether 241 entries into forms and 573 exits from forms were observed; 178 were from TD observations, leaving 152 entries and 484 exits from OD observations. The number and relative frequency of these re- cords spread over the year is shown in ta- ble 1, which also shows in relation to sun- rise and sunset the average weekly differentials of entries and exits of OD and TD observations, with the pattern of re- cords appearing in figure 1. Seasonal varia- tion in the number of observations prima- rily reflects detectability of the forms, except for the peak of TD data in March which reflects application of additional ob- server time. The statistics and tests of signif- icance for the fitted harmonic curves are given in table 2. The divergence between entry and exit curves, which is significant (Tab. 2), justifies Weeks 0 5 10 15 20 25 30 35 40 45 50 4 R 2 5 I -A es. oo -6 Before sunset o ) Fig. 1. Night duration differentials. The constants are respectively sunrise (a) and sunset (b) with the curves of the harmonic functions of entries into and exits from forms throughout the year. Observations of OD hares from which the curves have been constructed (e) and the excluded TD hares (0) are averages for each week. Daily activity period of Lepus europaeus 359 their separate treatment. As nights shorten in the early part of the year, hares begin to leave their forms before sunset on average about two weeks earlier than they begin to enter them after sunrise (Fig 1). There are insufficient data from which to make the same analysis for the equivalent period in the second half of the year. The fit between both sets of data (OD and TD observa- tions) and the curves is clear but there are indications in mid-summer that activity at both ends of the day extends for longer than the curves predict and also that from week 31, the first in August, onwards and until the end of October it is shorter than predicted. In December and early January, when the nights are longest, hares are almost totally nocturnal, entering and leaving their forms in darkness. The time of exit is very consis- tent, being generally no more than 30 mins after sunset. For the reasons given in meth- ods, quantitative data on the timing of entry were not obtained at this time of year. In figure 2 are the results of intensive daily ob- servation of one form occupied by the same hare during December 1985 and January 1986. Exit times averaged 26.8 mins post- sunset (range = 19-36, n= 11 days). Entry time was only observed directly (by moon- light) twice (83 and 56 mins pre-sunrise), while on 6 occasions the hare was already in the form when first observed that day (range of 46-96 mins pre-sunrise). Intervals between entry and sunrise were therefore longer and less consistent than those be- tween sunset and exit. For the majority of entries and exits the identity and sex of the individual was not known. However, in a minority of cases in- dividuals were recognised thereby provid- ing additional information. Over a two year period, I obtained 17 exit times of a known buck, Bolingbroke, who at the time was the alpha male in the dominance hierarchy within the study area (HoLLey 1986). In 13 (76%) instances his exit from the form was before the mean time of exit by the study population. The exceptionally late entry and early exit TD data appearing in figure 1 and table 1 for week 24 refer to a study doe, 360 A. J. F. HoLLey Table 1. The weekly means in hours for OD and TD hares of the differences between entries and sunrise (positive values before, negative values after) and exits and sunset (positive values after, negative values before). Also, nightime duration, sunrise to sunset, and the values of the fitted harmonic functions Week Nightlength Entries am Exits pm Fitted TD hares (n) Fitted TD hares Function Function vosoaunMNwmwme ri DunNH r r r 6 2 6 4 3 6 2 1 1 8 3 9 2 ( 0 4 1 5 5 WUDHRWUWNSJ[WHrRHN vwDrwrhrwpmpd + * Adoe and mateguarding consort Daily activity period of Lepus europaeus 361 Table 2. Statistics of the harmonic function (a) and tests of statistical significance (b). (a) Differentials A(Amplitude) hours Entries Exits (b) Amplitude, A F-statistic 45.77 340.49 Entry differentials Exit differentials Seasonal Displacement Difference of Peaks, P (am) - P (pm) 3.13 (weeks) Bluebell, on the day of her parturition when she occupied her form for only six hours. In mid-December, when night length in the study area exceeded 16h, the activity peri- od, all of it within those hours, occupied about 14.5h (Tab. 1; Fig. 2). In mid-June when night length reduced to a minimum of 7.4h, the activity period occupied about 13.5 h, six of them in daylight (Tab. 1). Is there, during the period of exclusively noc- turnal activity, any relationship between night length and the length of the activity period? That question is addressed in fig- ure 3 which is constructed exclusively from TD data, since these provide an accurate measurement of the activity period of the hare. From a peak of nearly 15 hin week 1, the activity period declined to a minimum of just over 12h in week 12, the third full week in March, mirroring the decline in night duration. That decline was reversed in week 13 when the activity period in- creased to 13.4h and exceeded night dura- tion by two hours. By week 18 it was exceeding night duration by >4h. The re- sults from this figure, supported by data in table 1, demonstrate a close correlation be- tween respectively the lengths of the night hours and of the activity period between the start of the year and the latter part of March. pP Peak-week 1.875 (Jan.) -1.255 (Dec.) Degrees/freedom p-value 2&27 2&41 <.001 <.001 Approximate z-statistic 4.0 Discussion The results of this study suggest that the brown hare is essentially a nocturnal ani- mal. Given sufficient hours of darkness, all activity takes place within them. During the few weeks when there was a marked ex- cess of the dark hours over the activity hours, hares consistently emerged from their forms and commenced activity within 30 mins after sunset, and the excess was re- flected in the time they entered their forms and ceased activity, which was as early as one and a half hours before sunrise. How- ever, it appears that when the night hours are insufficient, the activity period overlaps into daylight at either end of the day. This annual activity pattern conforms clo- sely to that appearing in the observational reports of hare activity on their foraging grounds by Maruszweski (1981) and HoMmoLkA (1986) and also the results of radiotracking analysis of individual activity by Pepın and CARGNELUTTI (1994). Limited evidence suggested, however, that certain individuals, such as pregnant and nursing does and alpha bucks, could be active for periods of up to 16h or more. Such indivi- dual variation is also reported by PEPIN and CARGNELUTTI (1994). The daily activity period follows a seasonal cycle, getting shorter as the hours of day- 362 A. J. F. HoLLey ul; sunrise APAAPAI P Time oO [00] oO oO sunset 1618202224 262830 1 3 5 December January Date Fig. 2. Observed occupation (vertical lines) of one form by the same hare between 15 December 1985 and 6 January 1986. Symbols indicate observed entry or exit (e), earliest sighting of hare already in form (0), hare present in form on the day but entry and exit not observed (P) and hare absent that day (A). Ol Activity period m Night hours Hours u ma a ee ea mal a @O0O0—MDWPOIDM Du26 9 Month Jan Feb 12513 18 Mar] Ar] ] Fig. 3. The mean daily activity period (i.e. length of time out of forms (+SE) based on TD observations) compared with the average night length, sunset to sunrise, during the first 18 weeks of the year. Data only presented for weeks with four or more TD obser- vations (sample size in histogram). light increase. This produces the surprising outcome that it is substantially longer be- fore the breeding season commences in late December (TArper 1991) than it is in the peak reproductive months of February to June. Why should that be? Why does it cy- cle at all? The answer appears to lie in not just a nocturnal preference but an aversion to daylight activity on the part of the brown hare. Preference is shown by choice, aver- sion by withdrawal. Assuming an optimal activity period of over 14h, it would be ex- pected that the transition from a totally nocturnal activity period to a partially diur- nal one, as the night hours drop below 14, would not affect the length of the period. That, however, is not the case, the duration of the period in fact declining from nearly 15 h in the first week of January to a mini- mum of just over 12 h in the third full week of March, mirroring the duration of the night hours. Then, towards the end of March the activity period begins to increase and takes in some of the daylight hours. Why does the duration of the activity peri- od not remain constant instead of reducing substantially January to March and then in- creasing to mid summer? Perhaps hares might need longer hours feeding in winter to meet the higher energetic demand of lower ambient temperature and to allow for the lower calorific value of food. That possibility seems unlikely because there is no sign of any extension of the activity peri- od following the onset at the end of Decem- ber of the reproductive season with its de- pendent energetic demands. The pattern of events does seem to indicate, first, that some inhibitory factor is preventing easy transition from a totally nocturnal to a par- tially diurnal regime and, secondly, that there is a point, at roughly 12 h of duration, beyond which the activity period cannot ea- sily be contracted. This suggests that the in- hibitory factor is daylight itself: hares are shy of daylight activity. Another manifesta- tion of the inhibitory factor is in the differ- ence between the harmonic curves for the times of form entry and departure. In spring, the departure curve crosses the sun- set line into daylight two weeks before the entry curve crosses the sunrise line. Day- light activity extending before sunset pre- cedes daylight activity lasting after sunrise. In the evening, after spending the day in its form, a hungry hare will be less shy of day- light activity than a well fed one returning to its form in the morning. The results from this study show the difference to apply as well before as after commencement of the breeding season and thus cannot be entirely connected to reproductive and agonistic interactions which in both nocturnal and diurnal mammals and birds often peak at activity onset (for review see DAAN and ASCHOFF 1982). Although hares, once they are regularly active diurnally, rapidly ex- tend the duration of daylight activity, none the less the inhibitory factor appears to be still operating to the extent that, in this study area, the duration of activity remains shorter than during the totally nocturnal re- gime. There are indications of a relatively sudden withdrawal from daylight activity at the end of the breeding season in August. Of related species, studies using automatic activity recording of mountain hares (Lepus timidus) in Sweden by CEDERLUND and LEmnELL (1980) and LEMNELL and LINDLOF (1981) showed a close relationship between sunset and onset of activity and sunrise and cessation of activity in winter. Daylight ac- tivity gradually increased up to 50% in summer when the nights were very short. The study of MecnH et al. (1966) on five radio-collared snowshoe hares (Lepus americanus) in Minnesota, USA, also showed that seasonal changes from January to May in both onset and cessation of activ- ity followed the trend of changing sunrise- sunset times, but with the difference that Zusammenfassung Daily activity period of Lepus europaeus 363 contraction of the activity period continued into May when it amounted to less than nine hours. I propose that the proximate cause of cyclicity in the activity period of brown hares is an aversion to daylight activ- ity. It seems that this may be shared by a number of other leporid species. The emergence of the brown hare from darkness on to the daylight arena in March, as demonstrated in this study, adds further clarification to the explanation of the “mad March hare” of literature (CARROL 1865) gi- ven by HoLL£eY and GREENWOooD (1984). Acknowledgements I am greatly indebted to my brother-in-law, Dr. Thomas Hayrton, for valuable statistical analy- sis and advice and to RuTH TEMPLE and Dr. NAn- CY VAUGHAN for discussions and preparation of the figures. My thanks go also to Dr. JoHN CoUL- son and Dr. NEIL METCALFE and to anonymous re- ferees for their very helpful comments on a pre- vious draft and to Kraus HAcKLÄNDER for translation into German. Sunrise and sunset va- lues reproduced, with permission, from data sup- plied by the Science and Engineering Research Council. Die tägliche Aktivitätsperiode des Europäischen Feldhasen (Lepus europaeus) Um die Veränderung der Länge der täglichen Aktivitätspenode im Laufe eines Jahres für Feldhasen zu bestimmen, wurde die Nutzung der Sassen auf einem 65 ha großen Gebiet in Südwestengland aufgezeichnet. Dabei wurde das Aufsuchen (n = 241) und das Verlassen der Sasse (n = 573) proto- kolliert. Im Dezember (Nachtlänge: 16 h) waren die Hasen ausschließlich nachtaktiv. Die Intervalle zwischen dem Aufsuchen einer Sasse und dem Sonnenaufgang waren länger und inkonsistenter als jene zwischen dem Verlassen der Sasse und dem Sonnenuntergang. Im Juni (Nachtlänge: 7,4 h) reichte die Aktivitätsperiode für insgesamt 6 h in den Tag hinein: Sie endete erst nach Sonnenauf- gang und begann bereits vor Sonnenuntergang. Die Länge der Aktivitätsperiode verkürzte sich en- tsprechend der Nachtlänge von 14,5 h im Dezember auf 12 h in der dritten Märzwoche. Darauffol- gend stieg die Aktivitätslänge auf 13,5 h im Spätsommer, um im Herbst wiederum zu sinken. Diese Ergebnisse weisen daraufhin, dass eine proximate Ursache für den circannuellen Rhythmus bei Feldhasen eine Vermeidung von Aktivität bei Tageslicht ist. 364 A. J. F. HoLLey References CARROL, L. (1865): Alice’s Adventures in Wonder- land. London: Macmillan. CEDERLUND, G.; LEMNELL, P. A. (1980): Activity recording of radio-tagged animals. Biotel. Pa- tient Monitg. 7, 206-214. DAANn, S.; ASCHOFF, J. (1982): Circadian contribu- tions to survival. In: Vertebrate Circadian Sys- tems. Ed. by J. AscHorr, S.DAAn, and G. A. Gross. Berlin: Springer. Pp. 305-321. HoLL£y, A.J.F. (1986): A hierarchy of hares: dominance status and access to oestrous does. Mammal Rev. 16, 181-186. HoLLEY, A. J. F. (1992): Studies on the biology of the brown hare (Zepus europaeus) with parti- cular reference to behaviour. Diss. thesis. Uni- versity of Durham, Durham. HoLLEy, A.J.F; GREENWooD, P.J. (1984): The myth of the mad March hare. Nature 309, 549-550. HoMOLKRA, M. (1986): Daily activity pattern of the European hare (Lepus europaeus) Folio zool. 35, 3342. LEMNELL, A.; LinDLorFB. (1981): Diurnal and seasonal activity pattern in the mountain hare. In: Proceedings of the World Lago- morph Conference, Gland. 1981. Ed. by K. Myers and €. D. MAcınnESs. Ontario: Uni- versity of Guelph. Pp. 349-356. MATUSZEWSKI, G. (1981): Circadian activity of European hares in Spring, on the Kampinos Forest border. In: Proceedings of the World Lagomorph Conference, Gland. 1981. Ed. by K. Myers and C.D. MacnInNnEs. Ontario: University of Guelph. Pp. 357-365. MEcH, L. D.; HEEZEN, K.L; Sınıfr, D. B. (1966): Onset and cessation of activity in cottontail rabbits and snowshoe hares in relation to sun- set and sunrise. Anım Behav. 14, 410-413. PEPIn, D.; CARGNELUTTI, B. (1994): Individual var- iations of daily activity patterns in radio- tracked European hares during Winter. Acta Theriol. 39, 399-409. TAPPER, S. C. (1991): Brown hare Lepus europaeus In: The Handbook of British Mammals, 3" ed. Ed. by G. B. CorBET and S. HARRIS. Ox- ford: Blackwell. Pp. 154-161. Author’s address: Dr. A.J.F. HoLLey, Ferndale House, Wick Lane, Brent Knoll, Highbridge, Somerset. TA9 4BU. U.K. Mamm. biol. 66 (2001) 365-370 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication , Mammalian Biology = Zeitschrift für Säugetierkunde Home range size, movements, and habitat utilization of three male European wildcats (Felis silvestris Schreber, 1777) in Saarland and Rheinland-Pfalz (Germany) By H. U. WITTMER Institut für Biogeographie, Universität des Saarlandes, Saarbrücken, Germany Receipt of Ms. 30. 01. 2001 Acceptance of Ms. 20. 04. 2001 Key words: Felis silvestris, home range, habitat utilization, nightly movements Although several radio-telemetry studies of European wildcats have been conducted (e.g. CoRBETT 1979; STAHL et al. 1988; LißE- REK 1999), systematic investigations on Euro- pean wildcats are lacking in Germany. Due to a population size of approximately 1000 individuals (F. RAIMER pers. comm.) and the connection to wildcat populations in neighbouring France and Luxembourg, the population in Rheinland-Pfalz is of great importance for the conservation of wildcats in central and western Europe. Therefore, a radio-telemetry study was conducted in Saarland and adjacent parts of Rheinland- Pfalz, to assess home range size, movements and habitat utilization of European wildcats in this region. The study area encompassed approximately 130 km? in northern Saarland and southern Rheinland-Pfalz, southwest Germany (6°55’ E, 49°36' N). Elevation in the area ranges from 250 to 650 m. Precipitation is greatest at high elevations and ranges an- nually from 700 to 1000 mm and tempera- tures vary from -16.5°C to over +30.0°C (FiscHER 1989). A mean snow depth of 10cm is recorded for 10 to 80 days. Ap- proximately 39% of the study area was forested. The dominate native tree was red beech (Fagus sylvatica), but spruce (Picea 1616-5047/01/66/06-365 $ 15.00/0. abies) and Douglas fir (Pseudotsuga men- ziesii) were commonly planted. At lower elevations agriculture was common and for- est climax consisted of submontane beech- oak-forests (Quercus petrea). Current agri- cultural practices result in many fallow fields and meadows. Over the duration of the study, 3 male wild- cats were caught in box traps, sedated with a Ketamin/Xylazin-mixture (HATLAPA and WIESNER 1982), weighed, measured, and radiocollared (50 g; K. Wagener Cologne, Germany) (Tab. 1). Approximate age was estimated based on tooth succession and condition. Wildcats were primarily located continu- ously during their main active period at night resulting in successful locations ap- proximately once every 50 minutes. Accu- racy of radio locations was evaluated under different habitat conditions, consistent with literature standards (WHITE and GARROTT 1990) and allowed to evaluate a habitat- specific telemetry error of 100 m. For home range analysis, locations were filtered for 2 hour intervals to equally distribute the data. Adaptive Kernel (AK) estimates (Worton 1989) of annual and seasonal home ranges were estimated using the pro- gram Ranges V (KEnwarD 1995). In addi- 366 H. U. WITTMER tion, minimum convex polygons (MCP) (MonHr 1947) were measured for compari- son with previous research. The area enclosed in the 95% AK contour was defined as the home range of an indivi- dual to exclude outliers. Core areas were defined as the area within the 50% contour. Seasonal delineations considered mating behaviour (PıecHocki 1990) and prey avail- ability (SLADER 1973) and included: winter mating season (January-March), spring (April-June), summer (July-September), and autumn (October-December). Over the duration of this study, Ml was moni- tored for 249 days, M2 for 49 days, and M3 for 266 days resulting in a total of 1276 relo- cations with 77% of the locations being re- corded between 18.00 h and 07.00 h. Minimum nightly movements were calcu- lated by summing the distances between re- locations of nights where cats were fol- lowed for >25 hours and both daily resting sites were recorded. The average obser- vation time per night was 9.35 hours. (SD =3.02) resulting in an average of 11 (SD = 4) successful locations per night. Locations were imported into the Geo- graphic Information System (GIS) Map- Grafix (ComGrafix USA). Habitat use was determined using a circular buffer with a 100 m radius around the locations filtered for a l hour interval. Within this 3.14 ha buffer the extent of each habitat was re- corded. When analyzing for seasonal habi- tat use, areas for each habitat category in- cluded in all locations were summed to equal the probability that a location was in a specific habitat category. Analysis of habiı- tat utilization was restricted to Ml and M2 as digitized maps were unavailable for M3. Selection of habitat was assessed by com- paring availability and use of habitat types within the total range (100% AK) of an in- dividual following the method described by NEU et al. (1974). An &= 0.05 was used to determine significance for tests of the null hypotheses. Home range: The largest seasonal home range of 2515 ha was observed in winter 1996 for M2 (Tab. 1). Both MI and M3 had their largest seasonal home ranges during spring. Seasonal home ranges during winter and spring were significantly larger than during summer and autumn (one tailed t- test, d{=6, p= 0.025). Accounting for the total number of relocations, annual home ranges were estimated to be 1407ha for MI and 1916 ha for M3, averaging 1662 ha. Core areas ranged from 92 to 460 ha. Spa- tial overlap between MI and M2 during January and February 1996 was 224ha. Therefore Ml shared 29% and M2 approxi- mately 9% of its range. Core areas were used exclusively. Nightly movements: The seasonal average of distances travelled per night ranged from 2.8km during summer to 11.3km during winter (Tab. 1). Over a period of 14 hours, the longest observed nightly movement was 13.3km recorded for M2 in February 1996. Nightly movements during winter/ spring averaged 5.5km (SD=2.6km) and were significantly longer than movements during summer/autumn averaging 3.0 km (SD=1.2km) (one tailed t-test, df=70, p< 0.001). Habitat use and selection: More than 1 ha- bitat category was found within the 100 m accuracy buffer around 74% of the loca- tions of MI and 65% of the locations of M2. Of these 85 and 79% included forest edge for Ml and M? respectively. Of the lo- cations encompassing only 1 habitat cate- gory, 96% were in forests. For both cats, 3habitat types (forests, fields/meadows, riparian meadows) accounted for 86 to 91% of the relocations. Forests were used most intensively throughout the study period ranging from 50 to 55% of the locations. Fields/Meadows were used by MI more in- tensively during winter and summer, ripar- jan meadows were used most often in spring. Percentages of forested areas in- cluded in seasonal total ranges varied be- tween 40 and 66%. Habitat use was significantly different from expected for M2 during winter, showing preference for fields/meadows and avoid- ance for clearings (7 = 30.41, dr A p < 0.001) (Fig. 1). During the same period M1 used habitats proportionately to their oOccurrence (1 = 4,52, dii=4, pP 054) 367 lvestris 15 51 Home range and habitat utilization of male Fel 152 vs abeiany vvwmm 02 Ge L (u) syybtu 37>]Jdwo) w>y ul sJuawanou ÄQYbıN 8.2 LSY sel BaIE 310) %0S %S6 EEE [Ky23 G60L SILDE eT9T %OOTL jou1ay anıydepy ey ul 9bueı 9WOH GLE je201 18 uwungny I Jawulns 89 Buuds ELL Joyuım = je201 - uwnIny - Jawluns - buuds 66 JoyuLm 105 je201 = uwnINny LI Jawwuns 802 Buuds 9LL JoyuLM u UOSE9S suol} -E90]31 Jo JoquunN syuswanou ÄjJyBıu pue ’sazıs abueı 9Wwoy UMOUJUN 8661 "Ss WI uoljenleIs 9661 '2 '6L umouyun 9661 "6 '6C sn3e3S/aJeq So] b os/S$ b 00/9 b 0087 b 0925 ybıam syyuou VES Anpe 661 80 22 EW syyuou gE< Anpe 966L TO eT ZW syquow Ve< 9661 20 Ye ‘Anpe G66L TO 60 TW aby pa1e]]0) aınyde) "SISUEPEIeYI SPeIPJLM "T aJgeL H. U. WITTMER 368 "papadxa wo Jualayıp Auesyıubıs asn Jeygey = + ‘(7/6T) Je 22 nan BuLpıoa9e sjernajul 39uap -14U0I 3}eILpuL SIeq J01E ‘sn Jeygey aYesıpuı suwn]o9 aıLym ‘sbueı Je}0} Jeuoseas jenpinipui urygım Äyıgtere 3jedıpuı suwnJo9 Adıb :zw pue Tyy UOLP3J3S JeygeH °T "Big Ss Mopeaw Ss MOpeAW sBullegl9 ueiledii Ss Mopeaw/spjal} sqnıys S]S9JO} sbullegld ueisedi4 Ss MOPeU/SPIAN sqnıys S]S9J0) L Br BEN Re a a ee, 0 free m ua N Bao ee ae 0 | LOL rol r rm ‚02 © r02 0 ve: 8 rO0Ee = -0E | ° (e) or? -0r% rOoS Ss r0oS S z + r 09 +09 202 02 J3JUuIM ZN JOyulM IN S MOpeaUI Ss Mopeau sBulleg]l9 ueisedi4 Ss MOPeSW/SPISI} sqnıys S]S910} sBullegjld ueiledı. Ss Mopeaw/spiel, sqnıys S}S3J0} (= — —— tl Tree hm er. 0 "Ger 4. I reo)lE | r DD Bi: £ r0E & - | je} 9 07 2 (7) ros Ss S r 09 02 Jawwns | W Buuds ıW Home range and habitat utilization of male Felis silvestris showed avoidance of fields/meadows in spring (x = 28.63, df = 4,p < 0.001) and used habitats in proportion to their availability during summer (x’ = 1.75, df =4, p = 0.78). The average annual home range size of 1662 ha in this study was larger than ranges reported for radiocollared male wildcats in previous studies (CoRBETT 1979; STAHL et al. 1988). However, MCP seasonal range sizes in this study were smaller than seaso- nal MCP range sizes recorded for male wildcats in Switzerland (LiBEREK 1999), in- dicating that home range sizes in male European wildcats vary under different ecological conditions. Use of home ranges is believed to be exclusive in areas where lagomorphs dominate the diet as opposed to exclusiveness only for animals of the same sex in areas where rodents dominate the diet (STAHL et al. 1988). As the ener- getic requirements of wildcats increase with the number of prey items required to fulfil their energy demand (HEMMER 1993), the observed differences in range sizes may be explained by differences in availability and abundance of prey. First order selection (JoHnson 1980) of areas inhabited by European wildcats have highlighted the importance of large forested areas with clearings interspersed to increase the amount of edge (e.g. Vogt 1985; HEM- MER 1993). First order selection would be strongly influenced by persecution by peo- ple and resulting extirpation from large areas of suitable habitat. Third order selection using radio-telemetry has been studied in east central France (STAHL 1986). Results of this study are con- sistent with STAHL (1986) with seasonal to- tal ranges including 40 to 66% of forested areas showing the importance of cover to wildcats. Because 65 to 74% of the loca- tions encompass more than 1 habitat cate- gory, the importance of habitat boundaries, especially forest edge was indicated. The anımals studied showed individual and sea- sonal variation in habitat selection. The flexibility of wildcat habitat use during this study indicates the ability of wildcats to live in forested landscapes altered by humans and suggests that habitat may not be pre- 369 venting wildcat recovery from low popula- tion numbers. Acknowledgements The presented project was part of a study con- ducted at the Institute of Biogeography at the University of the Saarland, Germany. I thank Dr. P. MÜLLER, Institute of Biogeography, Univer- sity of the Saarland for financial support and added telemetry data and H. CaArıus for telemetry field work. I am thankful to Dr. BRUCE McLer- LAN, Dr. URS BREITENMOSER, and ROGER RAM- CHARITA for comments on earlier drafts of this manuscript. Dr. AnnE OLson and Dr. Dan WEARY provided helpful comments on statistical analyses. References COoRBETT, L. K. (1979): Feeding ecology and social organization of wildcats (Felis silvestris) and domestic cats (Felis catus) in Scotland. Diss. thesis, University of Aberdeen, U.K. FIscHER, H. (1989): Rheinland-Pfalz und Saarland — Eine geographische Landeskunde. Darm- stadt: Wiss. Buchgesellschaft. HATLAPA, H.; WIESENER, H. (1982): Die Praxis der Wildtierimmobilisation. Hamburg: Paul Pa- Tey. HEMMER, H. (1993): Felis silvestris Schreber, 1777 — Wildkatze. In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. KrAPpP. Wiesbaden: Akad. Verlagsges. Vol. 5/II, 1076-1118. JoHnson, D. H. (1980): The comparison of usage and availability measurements for evaluating resource preference. Ecology 61, 65-71. KENWARD, R.E. (1995): Ranges V: An analysis system for biological location data. User man- ual, Wareham, Dorst, U.K.: Institute of Ter- restrial Ecology. LIBEREK, M. (1999): Eco-Ethologie du chat sau- vage Felis s. silvestris Schreber, 1777 dans le Jura Vaudois (Suisse); Influence de la couver- ture neigeuse. Diss. thesis, Universite de Neu- chätel, Switzerland. MOoHR, C. ©. (1947): Table of equivalent popula- tions of North American small mammals. Am. Midl. Nat. 37, 223-249. NEU, C. W.; ByErs, C.R.; PEEK,J.M. (1974): A technique for analysis of utilization-availabil- ity data. J. Wildl. Manage. 38, 541-545. 370 H. U. WITTMER PIECHocKI, R. (1990): Die Wildkatze. Wittenberg: A. Ziemsen Verlag. SLADEKR, J. (1973): Jahreszeitliche und jahresbe- dingte Veränderung der Nahrung der Wild- katze (Felis silvestris Schreber, 1777) in den Westkarpaten. Zool. Listy 22, 127-144. STAHL, P. (1986): Le Chat forestier d’Europe (Fe- lis silvestris Schreber, 1777): exploitations des ressources et organisation spatiale. Diss. the- sis, Universite Nancy I, France. STAHL, P.; ARTOIS, M.; AUBERT,M.F. A. (1988): Organisation spatiale et deplacements des chat forestiers adultes (Felis silvestris Schre- ber, 1777) en Lorraine. Rev. Ecol. 43, 113- 1328 Vogt, D. (1985): Aktuelle Verbreitung und Le- bensstätten der Wildkatze (Felis silvestris sil- vestris Schreber, 1777) in den linksrheinischen Landesteilen von Rheinland-Pfalz und Bei- träge zu ihrer Biologie. Beiträge Landespflege Rheinland-Pfalz 10, 130-165. WHITE, G. C.; GARROTT, R. A. (1990): Analysis of wildlife radio-tracking data. New York: Aca- demic Press. WoRrTOon, B. J. (1989): Kernel methods for estimat- ing the utilization distribution in home-range studies. Ecology 70, 164-168. Author’s address: HEIKO U. WITTMER, Faculty of Agricultural Sciences, University of British Columbia, #270 — 2357 Main Mall, Vancouver, B.C., Canada, V6T 1Z4 (e-mail: wittmer@interchange.ubc.ca) Mamm. biol. 66 (2001) 371-375 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication , Mammalian Biology : Zeitschrift für Säugetierkunde Genetic distinction of roe deer (Capreolus pygargus Pallas) sampled in Korea By H. S. KoH and E. RANDI Department of Biology, Chungbuk University, Korea and Istituto Nazionale per la Fauna Selvatica, Ozzano dell’Emilia, Bologna, Italia Receipt of Ms. 17. 01. 2001 Acceptance of Ms. 30. 05. 2001 Key words: Capreolus pygargus, mitochondrial DNA, molecular taxonomy The roe deer (genus Capreolus, Artiodacty- la, Cervidae) includes two species: the smal- ler European (C. capreolus) and the larger Siberian (C. pygargus) roe deer (GROVES and GruBB 1987; SOoKOoLOV and GROMOV 1990; HEwıson and DANILKIN 2001), which have widespread distributions in central- western Europe, and in Asia and east Eu- rope, respectively. They show a number of diagnostic morphologic, chromosomal and DNA traits (Groves and GruBB 1987; HEPTNER et al. 1989; DoUZERY and RANDI 1997; Ranoı et al. 1998a). Populations of the two species occur in putative contact zones in the Caucasus, where C. capreolus is distributed at the southern slopes and C. pygargus at the northern slopes of the mountain ranges, and along the rivers Volga and Don (HEPTNER et al. 1989; DANILKIN 1996). However, most probably European and Siberian roe deer populations do not hybridize in nature where they overlap in distribution (DAnıLKkın 1996; HEwIsoN and DANILKIN 2001). Populations of the Siberian roe deer show extensive body size and coat colour varia- bility, and up to six subspecies were recog- nized (GruBB 1993). However, recently DA- NILKIN (1996) and HEwıson and DANILKIN (2001) recognized only two subspecies: C. p. pygargus, from western and part of 1616-5047/01/66/06-371 $ 15.00/0. eastern Siberia, and C.p. tianschanicus, from Tian Shan and eastern Asia, including Korea. Other authors referred to roe deer populations from Korea as a distinct sub- species, e.g., C. p. bedfordi (SoKoLov and GroMov 1990), or C. p. ochracea (BARCLAY 1935). The genetic basis and taxonomical significance of morphological variation among Siberian roe deer populations are still unclear. Aim of this study is to assess the genetic distinction and clarify the taxo- nomic status of the C. pygargus population from Cheju Island in Korea. Three specimens of roe deer from Cheju Is- land in Korea were collected and muscle samples were stored in a deep-freezer. Total cellular DNA was extracted from muscle samples digested for 2 hours at 55°C in 500 ul of STE buffer (0.1 M NaCl, 10 mM Tris-HCl, 1mM EDTA, pH 8.0), 25 ul of 10 mg/ml Proteinase K stock solution, and 25 ul of 20% SDS solution. DNA was ex- tracted with equal volumes of phenol- chloroform and chloroform, precipitated with 2 volumes of ethylalcohol, and resus- pended in 50 ul of distilled water. The solu- tion was incubated at 37°C for 30 min with 5 ul of 10 mg/ml RNase stock solution, and DNA was extracted again. The entire mtDNA CR was PCR-amplified using primers L-Pro and H-Phe (DouZzEry 372 H.S. Kon etaal. and RAnpı 1997; Ranpı et al. 1998b) and the following thermal cycle: 94°C for min: 942C7Tior I min, 122C7iozlEnn 54°C for 1 min (32 cycles); 72°C for 5 min. PCR-amplified products were purified using the DNA PrepMateTM kit (Bioneer Co., Cheongwon-gun, Korea), and se- quenced using an automated DNA sequen- cer (Perkin Elmer 377) at the Korea Basic Science Institute (Daejeon). The new mtDNA CR sequences were aligned with homologous complete (Dou- ZERY and Ranpı 1997) and partial (RANDI et al. 1998 b) sequences, using the computer program CLUSTALX (THoMPson et al. 1997). Phylogenetic analyses were performed by maximum parsimony (MP), using PAUP* 4.0b2a (SworrForD 1998), with unordered character states (uninformative nucleotide positions excluded), 10 replicates of ran- dom addition of terminal sequences and TBR branch swapping, and by neighbor- joining procedure (NJ; SAmou and NEI 1987), with TAmurA and NEI (1993) DNA distances (TN93), using PAUP*. Majority- rule (50%) consensus trees were Con- structed when multiple equally parsimo- nious topologies resulted from MP analyses. Robustness of the phylogenies was assessed by bootstrap percentages (BP; FELSENSTEIN 1985), with 1000 random resamplings. Clades can be considered significantly sup- ported when BP values are > 70%. PCR products were clean single bands of the expected molecular weight, and the se- quencing allowed unambiguous nucleotide identifications. Regional organization of the CR from Capreolus was concordant with data from other cervid CRs (Douzery and Ranpı 1997, Ranoiı et al. 1998 b; NAGATA et al. 1998; Cook et al. 1999; Rannı et al. 2001), thus suggesting that they represent true mtDNA sequences and not nuclear co- pies. The CR of Capreolus includes a cen- tral conserved region (CR-II), a left domain (CR-I, on the tRNAs Thr and Pro side), and a right domain (CR-III, on the IRNA Phe side; not shown). A complete alignment 926 nucleotide long was obtained (the new sequences are avail- able from the GenBank with accession nos. AJ311188 and AJ311189). The average TN93 genetic distance between the Cheju samples and European or Siberian roe deer was 4.53%, and 2.5%, respectively. Roe deer from Cheju showed two distinet mtDNA CR haplotypes, which diverged at 0.2% of their sequences. Moreover, we have aligned partial mtDNA CR sequences, 679 nucleo- tide long, obtained from 23 samples of Eu- ropean, Siberian, and Korean roe deer. Equally weighted MP analysis of the CR alignment (with 47 parsimony-informative characters) produced 90 trees (length L = 81; consistency index CI = 0.617; reten- tion index RI = 0.881). The 50% majority- rule consensus tree is equivalent to the NJ phylogenetic tree, which is shown in Hie,1. MP and NJ trees concordantly showed that: (1) European and Siberian roe deer se- quences split into two divergent evolution- ary lineages (BP = 100%); (2) Siberian roe deer sequences split into three significantly different evolutionary lineages (BP = 70%- 100%), including roe deer from east Siberia (Amur Region), Korea, and west Siberia (Kurgan Region); and (3) roe deer se- quences from Korean individuals joined into a monophyletic dade (supported by BP = 91%), which was nested within the Si- berian roe deer clade which is the sister clade of the roe deer sampled in the Kurgan Region (west Siberia). Thus, mtDNA sequences support a genetic distinction between west Siberian and far eastern roe deer, in accordance with both SoKoLov and GRoMoV (1990) and DANILKIN (1996) taxonomical listing.. The new mtDNA sequences analysed in this study strongly suggest that roe deer from Cheju island are more closely related to popula- tions sampled from the geographically dis- tant Kurgan Region in west Siberia (C. p. pygargus) than to those from the geographi- cally close Amur region in south-east Sibe- ria (Fig. 1). The mtDNA sequences from Cheju Island form a significantly distinct clade within C. pygargus, thus indicating that roe deer from Korea belong to distinct populations, and supporting SOKOLOV and GRoMoVv’s (1990) view that roe deer from Systematics of roe deer in Korea 373 the far east including Korea, should be kept gion, and it is bound on the north by north- distinct from the other Siberian roe deer eastern China and far eastern Russia. Cheju populations. The Korean peninsula is apart Island, the largest of the southern Korea is- of eastern Asia within the Palaearctic re- lands, originated by a series of volcanic ac- Sb we Cheju Island (Korea) Korean_1.2 and 1.3 70 Siberian_1.1 Siberian_1.2 E ! 70 West Siberia Capreolus pygargus 23.002 (Kurgan Region) 100 Siberian_1.4 OS East Siberia (Amur Region) Siberian_2.2 European_1.1 European_3.2 European_4.1 50 European_1.2 European_ 5.1 European_2.1 European_3.1 Capreolus capreolus European_7.1 European_ 3.3 European_5.2 European_ 6.1 European_6.5 76 European_ 6.4 European_6.2 European_6.3 —— 0.001 nucleotide substitutions Fig. 1. Neighbor-joining tree (rooted using homologous C. elaphus mtDNA CR sequences as outgroups; not shown) clustering mtDNA CR haplotypes from Capreolus, computed using TAMURA and NEI (1993) DNA distances. At internodes are reported the bootstrap values > 50%. For the identification of C. pygargus (Siberian) and C. ca- preolus (European) haplotypes, see: RannI et al. 1998. 374 H.S. Kon etaal. tivities at the end of the Tertiary (PARK 1985). Cheju Island was connected to the Korean peninsula during the Pleistocene and separated again at the end of the last glaciation, about 10000 years ago. There- fore, roe deer in Cheju could have evolved in ısolation during the last 10000 years. Kon et al. (1997) reported that roe deer from the Korean peninsula showed smaller body size and skull length than west and east Sıberian roe deer subspecies. Thus, mo- lecular and morphological data lend sup- port to the view that roe deer from Korea do not belong neither to C. p. tianschanicus nor C. p. mantschuricus (= bedfordi), and might belong to a distinct subspecies, C. p. ochracea, as described by BArcLAY (1935). Acknowledgements This research was in part supported by the Ko- rean Science Foundation (grant 97-0402-05-01-3 to HunG Sun Kon), and by the Italian Istituto Nazionale per la Fauna Selvatica (I.N.F.S.) and Consiglio Nazionale delle Ricerche (C.N.R.). References BArcLAy,E.N. (1935): The roe deer of Korea. Ann. Mag. Natural History (London) 15, 626- 627. Cook, C. E.; WANG, Y.; SENSABAUGH, G. (1999): A mitochondrial control region and cytochrome b phylogeny of sika deer (Cervus nippon) and report of tandem repeats in the control re- gion. Mol. Phylogenet. Evol. 12 47-56. DAnILkın, A. (1996): Behavioural Ecology of Si- berian and European Roe Deer. London: Chapman and Hall. Douzery, E.; RanDı, E. (1997): The mitochon- drial control region of Cervidae: Evolutionary patterns and phylogenetic content. Mol. Biol. Evol. 14, 1154-1166. ELLERMAN, J. R.; MORRISON-SCOTT, T. C. S. (1951): Check-list of Palaearctic and Indian Mam- mals 1758 to 1946. London: British Museum (Natural History). FELSENSTEIN, J. (1985): Confidence limits on phy- logenies: An approach using the bootstrap. Evolution 39, 783-791. GRoVvES, C. P.; GRUBB, P. (1987): Relationships of living deer. In: Biology and Management of the Cervidae. Ed by C. M. WEMMER. Washing- ton (D.C.): Smithsonian Institution Press. Pp. 21-59. GrRUBB, P. (1993): Order Artiodactyla. In: Mam- mal Species of the World: A Taxonomic and Geographic Reference. Ed. by D. E. WıLson and D. M. REEDER. Washington (D.C.): Smith- sonian Institution Press. Pp. 337-414. HEPTNER, V. G.; NASIMOVICH, A. A.: BANNIKOV A. G. (1989): Mammals of the Soviet Union, Ungu- lates. Vol. I. Leiden: E. J. Brill. HEwIıson, A.J.M.; DaAanILkın, A. (2001): Evi- dence for separate specific status of European (Capreolus capreolus) and Siberian (C. pygar- gus) roe deer. Mamm. biol. 66, 13-21. Kon, H.S.; Yoo, S.K.; LEE,B.K.; (1997): Ana- lyses of external and cranial morphology of roe deer (Capreolus pygargus bedfordi Tho- mas) from Korea. Bull. Nat. Sci. (Chungbuk Univ.) 11, 99-103 (in Korean). NAGATA, J.; MASUDA,R.; Kar, K.; KANEKo, M; YosHIDA, M. C. (1998): Genetic variation and population structure of the Japanese sika deer (Cervus nippon) in Hokkaido Island, based on mitochondrial D-loop sequences. Mol. Ecol. 7, 871-877. PARK, D. W. (1985). Geology and coasts and mountains in Cheju Island. Cheju-Do study (Cheju Province, Korea) 2, 321-322. Ranpı, E.; Muccı, N.; PIERPAOLI, M.; DOUZERY, E. (1998 a): New phylogenetic perspectives on the Cervidae (Artiodactyla) are provided by the mitochondrial cytochrome b gene. Proc. Roy. Soc. Lond., B 265, 793-801. RAnDI, E.; PIERPAOLI, M.; DANILKIN, A. (1998b): Mitochondrial DNA polymorphism in popu- lations of Siberian and European roe deer (Capreolus pygargus and C. capreolus). He- redity 80, 429-437. RANDI, E.; Muccı, N.; CLARO-HERGUETA, F.; BoN- NET, A.; DOUZERY, E. J. P. (2001): A mitochon- drial DNA control region phylogeny of the Cervinae: Speciation in Cervus and implica- tions for conservation. Anımal Conserv. 4, 1-11. SAITOU, N.; NET, M. (1987): The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406- 425. SOKOLOY, V. E.; GROMOV, V. S. (1990): The contem- porary ideas on roe deer (Capreolus Gray, 1821) systematization: Morphological, etholo- gical and hybridological analysis. Mammalia 54, 431-444. SWOFFORD, D. L. (1998): PAUP*: Phylogenetic analysis using parsimony (and other meth- ods), version 4.0b2a. Sunderland, Massachu- setts: Sinauer Associated. TAMURA,K.; NET, M. (1993): Estimation of the number of nucleotide substitutions in the con- trol region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. Evol. 10, 512- 526. TATE, G. H.H. (1947): Mammals of Eastern Asia. New York: MacMillan Co. THoMPSson, J.D.; GiIBSoN, T.J.;; PLEWNIAR, F.; JEANMOUGIN, F.; HiGGIns, D. G. (1997): The CLUSTALX windows interface: Flexible Systematics of roe deer in Korea 375 strategies for multiple sequence alignment aided by quality analysis tools. Nucl. Acids. Res. 24, 4876-4882. Authors’ addresses: Hung Sun Kon, Department of Biology, Chung- buk University, Cheongju 361-763, Korea (e-mail: syskoss@cbuce.chungbuk.ac.kr) ETTORE RANDI, Istituto Nazionale per la Fauna Selvatica, Via Ca Fornacetta, 9, 40064 Ozzano dell’Emilia (BO), Italy Mamm. biol. 66 (2001) 376-378 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication x. Mammalian Biology 1 Zeitschrift für Säugetierkunde A new karyotype of Heliophobius argenteocinereus (Bathyergidae, Rodentia) from Zambia with field notes on the species Von A. SCHARFF, M. MACHOLAN, J. ZIMA and H. BURDA Department of General Zoology, University of Essen, Essen, Germany; Institute of Vertebrate Biology and Insti- tute of Animal Physiology and Genetics, Academy of Sciences, Brno, Czech Republic Receipt of Ms. 05. 03. 2001 Acceptance of Ms. 11. 04. 2001 Key words: Heliophobius argenteocinereus, Bathyergidae, karyotype, field notes The silvery mole-rat (Heliophobius argen- teocinereus) is a little known member of the family Bathyergidae, endemic to east to central Africa eastern of the Great Rift Valley, south of Equator, and north of the Zambezi river, i.e., in Kenya, Tanzania, Zambia, Malawi, and Mozambique (BURDA 2001). The only available information on the biology of Heliophobius is based on a few studies concerning physiology (MCNAB 1966) and burrowing and activity patterns (Jarvıs and SALE 1971; Jarvıs 1973). GEORGE (1979) described the karyotype of this species from Kenya. Since it resembled the karyotype of Heterocephalus glaber (both having 2n = 60), she concluded that the whole family Bathyergidae is chromoso- mally rather conservative. Regarding the fact that recently a large chro- mosomal variation (ranging from 2n = 40 to 2n=78) has been found in Cryptomys, an- other bathyergid mole-rat (SCHARFF 1998; BurDA 2001), the question arises in as much the karyotype established for a Kenyan po- pulation is representative for the whole genus Heliophobius which is distributed across at least 15 latitude degrees. To address this question we have examined karyotypes of AH. argenteocinereus trom Zambia, i.e., close to the southern distributional limit. 1616-5047/01/66/06-376 $ 15.00/0. Three silvery mole-rats (one male, two fe- males), collected in August 1996 in the Lu- balashi Area in the Central Province of Zambia (14°40’S; 29°55’ E) about 160 km east of Lusaka, were examined. Karyotypes were prepared from bone marrow following the splash method ac- cording to Forp and HAMERToN (1956). Chromosomes were differentially stained with the C-banding (SuMNER 1972) and G-banding (SEABRIGHT 1971) methods. Characterisation of chromosomes fol- lowed the nomenclature of Hsu and BENIRSCHKE (1967). The diploid chromosome number in all the examined individuals of Feliophobius ar- genteocinereus from Zambia was 2n = 2. The karyotype consisted of 27 pairs of me- tacentric and submetacentric chromosomes (autosomes) of decreasing size and 3 pairs of small acrocentrics (Nfa = 114). The X- chromosome was the second largest meta- centric, while the Y-chromosome was dot- like (most probably metacentric). The karyotype of Heliophobius argenteoci- nereus from Zambia (2n = 62; Nfa = 114) is very similar to that of silvery mole-rats from Kenya (2n = 60; Nfa = 114; GEORGE 1979). The difference between both karyo- types (one pair of metacentrics vs. two pairs Karyotype and field notes of Heliophobius from Zambia of acrocentrics) is most probably due to a simple Robertsonian fusion or fission. Un- fortunately, the poor quality of G-banding in both GEoRGE’s (1979) and our studies does not allow any detailed comparison and homologization of individual chromo- somes. Although Robertsonian fusions are supposed to be more frequent than fissions in mammals (suggesting thus that the Zam- bian population would be more ancestral) (cf. also NEvo et al. 1986), the opposite pro- cess ot fission cannot be excluded (cf. NEvo 1999). In view of the remarkable chromosome di- versification of Cryptomys in the Zambe- zian region (i.e., in Zimbabwe, Zambia, and Malawi), yet its uniformity in the Southern African subregion (SCHARFF 1998; BurpA 2001), the relative constancy of the karyotype of Heliophobius is of particular interest. It can be assumed that, contrary to mole-rats in most of the Zambezian re- gion, the populations of silvery mole-rats east of the Great Rift Valley have never been fragmented so that also isolation and speciation could not occur. Regarding the paucity of data on silvery mole-rats, it is worth to mention observa- tions which we made on the Zambian sil- very mole-rats. Altogether eleven mole-rats (1 male and 6 female adults, 1 female and 3 male juveniles) were obtained in the Luba- lashi Area, south of the Lunsemfa River in the Luano Valley, in miombo-woodland, mixed with few agricultural spots. The ground of the thin miombo forest was den- sely covered with tall grass which was partly burned by farmers. The adult male weighed 200 g, the average weight of the adult fe- males was 146+20g (range 118-170 g; n=6). Four females reared a single pup each. The male pups weighed 26g, 34 g, and 48 g, whereas the weight of the single female pup was 37 g. At the time of capture (August 1996), the pups were haired and their eyes and ears were open. The high proportion of nursing females in the sample suggests a distinct breeding season, with (small) litters being delivered during the dry season (which lasts from April/May till October/November). 377 Burrow systems (identified by the presence of mounds) of silvery mole-rats were very unevenly distributed in the study area. The mounds measured about 30cm (up to 50cm) in diameter. Burrow systems con- sisted of a main straight tunnel with short side branches and reached about 50 m in length. Most parts of the main tunnels were only 10-20 cm deep but some parts went into depth of more than 150 cm. A few blind ending tunnels or “bolt holes” were found. Diameter of burrows was 8-9 cm on average. One breeding nest was hidden within the system of tree roots, 20-30 cm deep. Heliophobius has been observed in two occasions feeding on (undetermined) grass rhizomes. The grass also served as nesting material. No macroscopic ectoparasites nor intestinal helminths have been found in any of the an- imals. Although silvery mole-rats have been re- ported to be highly aggressive (JARvIS and SALE 1971; Jarvıs 1973), our silvery mole- rats could be kept in pairs and have not en- gaged in serious fighting. Also, presence of juveniles was tolerated. One male offspring lived with its mother in a common cage for more than one year (three other juveniles died within three months of the capture). Furthermore, most of the adult silvery mole-rats were tame immediately after cap- ture and did not try to bite. References BURDA, H. (2001): Determinants of the distribu- tion and radiation of African mole-rats (Bath- yergidae, Rodentia): Ecology or geography? In: Collection Colloques et seminaires. Paris: Editions IRD (in press). ForD, C. E.; HAMERTON, J. L. (1956): A colchicine, hypotonic citrate, squash sequence for mam- malian chromosomes. Stain Technol. 31, 247- 2SlE GEORGE, W. (1979): Conservatism in the karyo- types of two African mole-rats (Rodentia, Bathyergidae). Z. Säugetierkunde 44, 278-285. Hsu, T. C.; BENIRSCHKE, K. (1967): An Atlas of Mammalian Chromosomes. Berlin, Heidel- berg, New York: Springer-Verlag. 378 A. SCHARFF et al. JARVIS, J. U.M. (1973): Activity patterns in the mole-rats Tachyoryctes splendens and Helio- phobius argenteocinerus. Zool. Afr. 8, 101-119. JARVIS, J. U. M.; SALE, J. B. (1971): Burrowing and burrow patterns of East African mole-rats 7a- chyoryctes, Heliophobius and Heterocephalus J. Zool. (London) 163, 451-479. MEeNAB, B. K. (1966): The metabolism of fossorial rodents: a study of convergence. Ecology 47, 712-733. NeEvo, E. (1999): Mosaic Evolution of Subterra- nean Mammals: Regression, Progression, and Global Convergence. Oxford: Oxford Univer- sity Press. NEVO, E.; CAPANNA, E.; CORTI, M.; JARVIS, J. U. M.; HicKMAaNn, G. C. (1986): Karyotype differentia- tion in the endemic subterranean mole-rats of South Africa (Rodentia, Bathyergidae). Z. Säugetierkunde 51, 36-49. SCHARFE, A. (1998): Verhaltensökologie und Sy- stematik sambischer Sandgräber (Bathyergidae, Rodentia). Diss. thesis, University of Essen. SEABRIGHT, M. (1971): A rapid banding technique for human chromosomes. Lancet 2, 971-972. SUMNER, A. T. (1972): A simple technique for de- monstrating centromeric heterochromatin. Exp. Cell Res. 75, 304-306. Authors’ addresses: ANDREAS SCHARFF, Fachklinik Rhein/Ruhr, Neu- rologische Abteilung, Auf der Rötsch 2, D-45219 Essen, Germany HyNEK BURDA, Department of General Zoology, University of Essen, D-45117 Essen, Germany, (e-mail: scharff@frr.de) Mıros MACHOLAN, Laboratory of Genetics and Embryology, Institute of Animal Physiology and Genetics, Academy of Sciences of the Czech Re- public, CZ-60200 Brno, Czech Republic Jan ZıMA, Institute of Vertebrate Biology, Academy of Sciences of the Czech Republic, CZ-60200 Brno, Czech Republic Mamm. biol. 66 (2001) 379-382 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Short communication , Mammalian Biology 2 Zeitschrift für Säugetierkunde New distributional records of small mammals at Beni Biosphere Reserve, Bolivia By JULIETTA VARGAS and J. A. SIMONETTI Colecciön Boliviana de Fauna, Museo Nacional de Historia Natural, La Paz, Bolivia and Departamento de Ciencias Ecolögicas, Universidad de Chile, Santiago, Chile Receipt of Ms. 29. 12. 2000 Acceptance of Ms. 30. 05. 2001 Key words: Bolomys, Marmosops, Mycroryzomys, Oxymycterus, Bolivia The mammalian fauna of Bolivia is among the least known in South America (PınE 1982). Fortunately, the knowledge about the diversity and distribution of this fauna has been increasing in recent years (e.g., ÄANDERSON 1997; EISENBERG and REDFORD 1999). Currently, 316 species are recognized for Bolivia, 71% of which are small mam- mals. The geographic distribution of most species is based on a handful of records from a few sites (ANDERSON 1997). Conse- quently, new distributional records are needed to clarify further the biogeography of Bolivian mammals (e.g., YENSEN et al. 1994; 'TARIFA and ANDERSoN 1997). Much sampling effort has been devoted to the northern highlands and La Paz valley (ERGUETA and SARMIENTO 1992). In the Amazonian region, the Beni Biosphere Re- serve (EBB) has received considerable at- tention in recent years (HERRERA-MACcC- BrYDE et al. 2000), including sampling of bats, marsupials, and rodents (CABBoTr et al. 1986; Wırson and SALAZAR 1989; ANDER- son 1997; YANEZ et al. 1998; BRACE et al. 2000; see also Rumız and HERRERA 2000). The reserve lies in the Llanos de Moxos re- gion, a center of high plant biodiversity. Furthermore, it is regarded as a key area for the conservation of threatened birds in the Neotropics (BRAcE et al. 2000; MOoRAES 1616-5047/01/66/06-379 $ 15.00/0. et al. 2000). Currently, only 11 species of small mammals have been registered, seven rodent and four marsupials (CABor et al. 1986; ANDERSON 1997). However, despite the efforts allocated to inventorying mam- mals at the EBB, ongoing sampling of mammals at both a terra firme forest and forest fragments at EI Porvenir ranch, EBB’headquarters, have revealed four new species for the region. Here we present these noteworthy records. During 1996 a small live-trapping sampling bout was allocated to three forest fragments at El Porvenir (YANEZ al. 1998). Two others have been sampled since 1999. Forest frag- ments sampled during 1999 and 2000 are known as “Taita B” (2.2 ha) and “Airstrip B” (0.3 ha) (14°51’37” S/66°19'68” W 163; BRACE et al. 2000). We also sampled the grassland neighboring a water course and marsh close to the forest fragment named “Porv A”. The sampling site at the terra firme forest, known as “Campo Monos” is located roughly 45 km NW from EI Porvenir (14°39'59" S/ 66°04°60” W and 130 m asl, see MoRAES et al. 2000 for vegetation description). During 1999-2000, sampling consisted of live-trap- ping and collecting for four consecutive nights each time with 200 medium Sherman traps in linear transects, traps being 10m apart. We have also examined prey remains 380 JULIETTA VARGAS et al. in 440 pellets of the barn owl (Tyto alba) col- lected at El Porvenir (VARGAS et al. unpubl.). All specimens collected have been deposited in the Coleccion Boliviana de Fauna (CBF), La Paz. Marsupialia: Didelphidae Marmosops dorothea (Thomas, 1912) is en- demic to Bolivia and regarded as threat- ened by the IUCN (Nowak 1999) This mar- supial has a disjunct distribution with records in the humid Yungas of La Paz (840-2 300 m asl) as well as in the arid low- lands of Santa Cruz (250-620 m asl; ANDER- son and TArIFA 1996). Known from 23 lo- calities and 46 specimens, the two areas of distribution are over 400 km apart (ANDER- son 1997). This broad disjunction led An- DERSON and TARIFA (1996) to suggest that two taxa could be involved. However, we collected it at Campo Monos, a record in the middle of the distribution gap challeng- ing this contention. A single subadult female (CBF 6442; TL 208, T 122, HF 15, E 18; 15 g) was captured (July 1999) in a seasonally flooded forest, close to the Curiraba river, the understory dominated by Heliconia sp, coinciding with known habı- tats of M. dorothea (Emmons 1999). The sin- gle specimen represents 2% of small mam- mals captured in a total of 424 trap/nights. Besides M. dorothea, Oecomys bicolor, Or- yzomys capito, Philander opossum and Proe- chimys sp. were also captured in the same ha- bitat. Rodentia: Muridae Microryzomys minutus (Tomes, 1860): the pigmy rice rat is known from high eleva- tions (2500-3000 m asl) in the Andes of Ecuador, Peru and Bolivia (EISENBERG and REDFORD 1999). It has also been reported for the Monte Zerpa’s cloud forest in Vene- zuela (Diaz 1994). In Bolivia, it is known from 10 localities and 28 specimens of the Yungas from Cochabamba, La Paz and Santa Cruz (ANDERSsoN 1997). Despite being considered a highland species (e. g., Nowak 1999), a subadult female (CBF 7.078; DE31635mm 7E904E24SEIE 2151258) was collected at Campo Monos in Septem- ber 2000. It was captured in a forest tract with an understory dominated by Heliconia sp. In this habitat, Marmosops dorothea, Oecomys bicolor, Oryzomys capito, Philan- der opossum and Proechimys spec. were also captured. Microryzomys minutus re- presents 2% of the 52 individuals captured at Campo Monos, with a trapping success of 0.2% (one out of 424 trap/nights), sug- gesting it might be rare. There were no records of M. minutus at the Department of Beni. This record extends its known distribution roughly 150 km NW of its previously recorded limits. Besides it biogeographical relevance, the record of M. minutus is of medical concern for EBB human populations, as this species might be a reservoir of human cutaneous leishma- niasis (ALEXANDER et al. 1998). Oxymycterus spec. (Waterhouse, 1837): bur- rowing mice inhabit open grassland, marshes, swamps, and grasslands being rare in humid forests (Emmons 1999). Three spe- cies including five subspecies are known from Bolivia, but their biology is unknown (AnDERSoN, 1997). Of these species, Oxy- mycterus inca iris (Thomas, 1901), dwells in the humid forest of the Amazonian low- lands. In Bolivia it is known from 20 local- ities and 63 specimens from La Paz, Santa Cruz and the western portion of Beni. We recorded it as prey of T. alba at El Porvenir, extending its distribution 100 km NE. The single skull recovered represents 0.2% of the prey remains of T. alba over 1998-1999 (VARGAS et al. unpubl.). Bolomys spec. (Thomas, 1916): a single skull of Bolomys spec. was found among the prey of T. alba (VARGASs et al. unpubl.). While the individual undoubtedly belongs to Bolomys, it was not possible to assign it to any of the three species known for Bolivia. Bolomys amoenus (Thomas, 1900) is known from four localities and just 13 specimens, being re- stricted to Cochabamba and Tarija at eleva- tions from 3800 to 4000 m asl (ANDERSON 1997). However, ANDERSoN (1997) includes the southeastern portion of Beni in its distri- bution with no further support. If our speci- men represents B. amoenus, this record will increase its distribution 260 km NW denot- New distributional records of small mammals ing also a notorious change of habitat. An- other species, B.lenguarurm (Thomas, 1898) is widely distributed in the lowlands, including several records from western Beni (AnDERSON 1997; ANDERSON and OLDs 1989). If the skull recovered at EI Porvenir belongs to B. lenguarum, it would represent a further 120 km E expansion of its known distribution. Twenty-two percent of Bolivian mammals are known from one to three localities (An- DERSON 1997). The four species reported here are hence comparatively better known regarding their geographic distribution. Even though and although EBB could be regarded as a relatively well known region (HERRERA-MACBRYDE et al. 2000), these four new records clearly state that much field work needs to be done to assess fully the diversity and distribution of Bolivian mammals. Acknowledgements This work has been supported by Fondecyt 1981050 and 790003 to JAS. C. MıRANDA, EBB Director granted permit for collecting specimens. Support and encouragement of our collegues A. A. GREZ, M. MorAESs, and R. ©. BUSTAMENTE is appreciated. We are indebted to I. HUARECO for field assistance. References ÄGUIRRE, L. F.; DE URIOSTE, R. (1994): Nuevos re- gistros de murci&lagos para Bolivia y los De- partamentos de Beni y Pando. Ecol. Bol. 23, 71-76. ALEXANDER, B.; LOZANO, C.;, BARKER, D. C.; Mc- CAnn, S.H.; ADLER, G. H. (1998): Detection of the Leishmania (Viannia) brasilienzis com- plex in wild animals from Colombian coffee plantations by PCR and DNA hybridization. Acta Tropica 69, 41-50. ÄNDERSON, S. (1997): Mammals of Bolivia, tax- onomy and distribution. Bull. Amer. Mus. Nat. Hist. 231, 1-652. ÄNDERSON, S.; OLDs, N. (1989): Notes of Bolivian mammals. 5. Taxonomy and distribution of Bolomys (Muridae, Rodentia). Am. Mus. No- vitates 2935, 1-22. 381 ÄNDERSON, S.; TARIFA, T. (1996): Mamiferos ende- micos de Bolivia. Ecol. Bol. 28, 45-63. BRACE, R.; HARTLEY, J. C.; BARNARD, C©. ].; Hussr, J. L. (2000): Ecology, biogeography and conservation of forest-island faunas in lowland Bolivia. In: Biodiversidad, conserva- ciön y manejo en la regiön de la Reserva de la Biosfera Estaciön Biolögica del Beni, Bolivia. Ed. by ©. HERRERA-MACBRYDE, F. DALLMEIER, B. MAcBRYDE, J. A. CoMISKEY, and C. MIRANDA. Washington, D.C.: SUYMAB series No. 4, Smithsonian Institution. Pp. 171-202. CABBOT, J.; SERRANO, P.; IBANEZ, C.; BRAZA, F. (1986) Lista preliminar de ares y mamiferos de la reserva “Estaciön Biolögica Beni”. Ecol. Bol. 8, 3744. Diaz, A. (1994): The rodent community of the Venezuelan cloud forest, Merida. Polish Ecol. Studies 20, 155-161. EISENBERG, J. F.;, REDFORD, K.H. (1999): Mam- mals of the Neotropics. Vol.3. The Central Neotropics: Ecuador, Peru, Bolivia, Brazil. Chicago: University of Chicago Press. Emmons, L. H. (1999): Mamiferos de los bosques hümedos de America tropical. Una guia de campo. Santa Cruz de la Sierra: Editorial F.A.N. ERGUETA, P.; SARMIENTO, J. (1992): Fauna silvestre de Bolivia: diversidad y conservaciön. In: Conservaciön de la diversidad biolögica en Bolivia. Ed. by M. Marconi. La Paz: Centro de Datos para la Conservaciön and United States Aid Mission to Bolivia. Pp. 113-164. HERRERA-MACBRYDE, O.; DALLMEIER, F; MaAc- BRYDE, B.; CoMISKEY, J. A.; MIRANDA, C. (eds.) (2000): Biodiversidad, conservaciön y manejo en la regiön de la Reserva de la Biosfera Esta- ciön Biolögica del Beni, Bolivia. Washington, D.C.; SYMAB series No. 4, Smithsonian Insti- tution. MOoRAES, M; RIBERA, M. O; VILLANUEVA, E (2000): La vegetaciön de la Reserva de la Biosfera Estaciön Biolögica del Beni y su importancia para la conservaciön In: Biodiversidad, con- servaciön y manejo en la regiön de la Reserva de la Biosfera Estaciön Biolögica del Beni, Bolivia. Ed. by O. HERRERA-MACBRYDE, F. DALLMEIER, B. MACBRYDE, J. A. CoMISKEY, and C. MırAnDA. Washington, D.C.; SYMAB series No. 4, Smithsonian Institution. Pp. 59- 74. Nowak, R.M. (1999): Walker’s Mammals of the world. Vol. 3. Sixth ed. Baltimore: John Hop- kins Univ. Press. Pine, RH. (1982): Current status of South Ameri- can mammalogy. In: Mammalian Biology in 382 JULIETTA VARGAS et al. South America. Ed. by M. A.MareEs and H. H. GEnoways. Linesville: Spec. Publ. Ser., Pymatuning Laboratory of Ecology, Univer- sity of Pittsburgh. Pp. 27-37. RuMiZ, D., HERRERA, J. C. (2000): Wildlife diversity and selective mahogany logging in Bosque Chimanes, Beni, Bolivia: surveying mammals and other vertebrates by line transects, track quadrants, live-trapping and mist-netting. In: Biodiversidad, conservaciön y manejo en la re- gion de la Reserva de la Biosfera Estaciön Biolögica del Beni, Bolivia. Ed. by O. HER- RERA-MACBRYDE, F. DALLMEIER, B. MACBRYDE, J. A. ComIskEy, and C. MIRANDA. Washington, D.C.; SUMAB series No. 4, Smithsonian Insti- tution. Pp. 235-202. TARIFA, T.; ANDERSON, S. (1997): Two additional records of Glironia venusta Thomas, 1912 (Marsupialia, Didelphidae) for Bolivia. Mam- malia 61, 111-113. WiLson, D. E.; SALAZAR, J. A. (1989): Los murcie- lagos de la Reserva de la Biosfera “Estaciön Biolögica Beni”, Bolivia. Ecol. Bol. 13, 47-56. YANEZ, M. A; VERA, F; SIMONETTI, J. A; GREZ, A. A. (1998): Small mammals of forest islands of the Beni Biological Station, Bolivia. Mast. Neotr. 6, 135-138. YENSEN, E.; TARIFA, T, ANDERSON, S. (1994): New distributional records of some Bolivian mam- mals. Mammalia 58, 405413. Authors’ addresses: JULIETA VARGAS, Colecciön Boliviana de Fauna, Casilla 8706, La Paz, Bolivia (e-mail: jecbf@ceibo.entelnet.bo) JAVIER A. SIMONETTI, Departamento de Ciencias Ecolögicas, Universidad de Chile, Casilla 653, Santiago, Chile. Mamm. biol. 66 (2001) 383-384 © Urban & Fischer Verlag http://www.urbanfischer.de/journals/mammbiol Book reviews Howmes, V.: On the Scent: Conserving Musk Deer - The Uses of Musk and Europe’s Role in its Trade. Brussels: Traffic Europe 1999. 57 pp. ISBN 90-9012795-X As is commonly known the relationship between man and some wild mammalian species is extremely problematic. This holds true especially for those species whose bodily attributes are be- lieved to be essential for human medical or cul- tural needs. Without doubt the musk deer belong to this group. In contrast to older assumptions of only one existing species in total four species of this small and primitive cervid genus are recog- nized in modern taxonomy (Himalayan musk deer -— Moschus chrysogaster, black musk deer — M. fuscus; forest musk deer - M. berezovskii; Si- berian musk deer - M. moschifera). These are dis- tributed throughout a large region in Asia includ- ing certain parts of Afghanistan, Pakistan, Nepal, Bhutan, Myanmar, the Koreas, China, Mongolia, Kyrgystan, Kazakhstan, and Russia (eastern Si- beria, far East, Sakhalin). Males of these species produce musk in an integumental gland and store this substance in the musk sac located between the navel and the sexual organ. Although some other mammalian species (i. e., muskrat - Ondatra zibethica, desman — Desmana moschata, musk oxen — Ovibus moschatus; suni — Neotragus moschatus) or plants (i.e., musk mallow — Hibis- cus abelmoschus,; musk rose — Rosa moschata; musk milfoil — Achillea erba-rotta moschata) are believed to produce similar aromas, these have nothing in common with the chemical substance of musk or with its odour and characteristics. Ori- ginal musk exceeds most other natural constitu- ents in smell intensity, persistence and fixative properties. In Asiatic and Arabic cultures it has been used by man for over 5000 years as a fra- grance, fixative for other fragrances or in medi- 1616-5047/01/66/06-383 $ 15.00/0. Mammalian Biology Zeitschrift für Säugetierkunde cine as a tonic for the heart and mind, for chronic headache and, of course, as an aphrodisiac. In an- cient times musk was introduced to the western world by Arabic doctors and since then increas- ingly has been in demand worldwide. Today, it is one of the most expensive substances derived from any animal. In Europe, for example, the price for musk per g during the 1990s reached three to five times that of gold. Consequently moschus deer males have been extensively pur- sued and hunted by native people with the aim to remove the entire gland (pod) and thus obtain musk. An adult deer male produces approxi- mately 18 to 32 g of musk and by selling 2 musk pods (about 50 g) a Nepalese family in remote mountain regions is reported to have acquired at least a year’s income. This situation has led to a serious decrease in numbers of deer. All musk deer species, therefore, have been included in the Appendices I or II of the Convention on Interna- tional Trade in Endangered Species (CITES) since 1979 and the IUCN Red List of Threatened Animals from 1996 classifies these species as vul- nerable or at least as nearly threatened. In Rus- sia, e.g., between 1989-1993 a total of 90 to 100 thousand individuals was killed and the popu- lation size was estimated to be 53-60 thousand in 1996. However, poaching of deer and illegal trade with musk most probably still exists in vast re- gions of remote areas although musk deer farms have been established since 1958 in China and methods were developed to remove musk from live individuals without injuries. The brochure in hand is a thorough research report on the interna- tional trade and use of musk from musk deer car- ried out by TRAFFIC Europe-Germany between January and July 1998. It contains much informa- tion on the biology of these cervids, on their status as well as on the effects of conservation efforts. D. KruUska, Kiel 384 Buchbesprechungen YALDEN, D.: The history of British mammals. London: T. and A.D. Poyser Ltd. 1999. Hard- cover, 305 pp., numerous black and white pic- tures, numerous maps and tables. £ 29.95. ISBN 0-85661-110-7. In this time of specialisation and over-specialisa- tion, when many authors seem to avoid over- views, it is wonderful to experience a book like that from DEREK YALDEN, the managing editor of the British “Mammal Review”! The author pre- sents a broad approach to the problems related to “The history of British mammals”. His presen- tation is based on own mammological investi- gations, publications dealing with questions of general zoology, as well as on information con- cerning domestic mammals, palaeontology, ar- chaeology, biogeography and history. From these sources Dr. YALDEN has produced an authorative text that offers delightful and informative reading on up-to-date knowledge. Short remarks on Mesozoic and Tertiary mam- mals introduce the reader to the subject at hand. A section on Pleistocene mammals follows; it is il- lustrated with instructive maps and tables. The second chapter deals with “the beginning of his- tory” during the late Pleistocene, but also com- parisons with mammalian faunas of continental Europe can be found. However, the “splendid iso- lation” of Britain since, at least, about 7 000 years before present, forms a biological “laboratory” with scientific problems different from those on the European continent. For the Roman, Saxon, Norse and Norman and later mediaeval periods of British history and for modern times, introduc- tions of mammals, as well as extinctions of other species are described — once again illustrated by diagrams, maps and tables. In these paragraphs detailed information on surviving species is also supplied. The author also takes a look at the twentieth century and beyond. In a special chap- ter problems confronting mammals on large (Ire- land) and small islands, such as Shetland, Orkney, Hebrides, Man, Scilly and Jersey, are addressed. It is impossible in deal here with the multitude of stimulating ideas presented in this book. A very detailed list of references (24 pages) and an index of ten pages concludes this remarkable publica- tion. As the author draws information from a wide range of sources, this is an outstanding refer- ence source also for readers with specialised fields of interest. The present reviewer hopes that the book from DEREK YALDEN will find a wide distri- bution and can thus be appreciated by a wide range of mammalogists and lay persons! P. LANGER, Giessen ELTRINGHAM, S.K.: The Hippos. Natural History and Conservation. London: T. and A.D. Poyser Ltd. (1999). Hardcover, 184 pp., numerous black and white pictures and 8 black and white plates. £ 27.95. ISBN 0-85661-131-X. This book gives detailed information on Hippo- potamus amphibius, the river hippo. The data sup- plied for Hexaprotodon (Choeropsis) liberiensis (pygmy hippo), however, are still limited. For ex- ample, while eleven pages deal with reproduction in Hippopotamus amphibius, less than two pages deal with respective data in the pygmy hippo. Sim- ilar relationships can also be found in the chap- ters on behaviour and on diet and feeding habi- tals. These statements do by no means indicate criticism of KEITH ELTRINGHAM’S accomplish- ments, but they make obvious where the gaps in our knowledge are and where research efforts are necessary. Observations, not to speak of field experiments, of Hexaprotodon liberiensis will be- come increasingly difficult: Most pygmy hippos live in countries — Liberia and Sierra Leone - with practically no means of law enforcement and a wide distribution of firearms because of civil wars and social unrest. For the larger species, the river hippo, one “can conclude...that, over the whole of Africa, there is no immediate threat to the hip- po as a species although some of the constituent populations are certainly at risk” (page 171). The book deals with anatomy and physiology of both species, their palaeontological origins, their social life and reproduction, with diet and feeding habits, ecology, as well as diseases, parasites and commensals and with the relationships between the hippo species and man. Finally, the distribu- tion of the two species is discussed. The origin of the wealth of information that is woven into the text is documented on six pages of references. Special questions can be easily accessed with the help of an index of six pages. The text is clear and direct and sometimes also humorous. It is ob- vious that the author often draws on his own pro- found research experience with both species! P. Langer, Giessen Instructions to authors Submission and acceptance of manuscripts: Manuscripts for publication should be sent to the mana-ging editor, Prof. Dr. D. Kruska, Institut für Haustierkunde, Christian- Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany (e-mail: dkruska@ifh.uni-kiel.de). Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of the manuscript will be confirmed immediately by normal mail or e-mail, and as soon as the peer reviews are received the authors will be informed concerning any decision. All correspondence dealing with details of production should be sent to: Mammalian Biology, Urban & Fischer Verlag, Löbdergraben 14a, D-07743 Jena, Germany (e-mail: e.czauderna@urbanfischer.de Copyright: Submission of a manuscript implies that the submitted work has not been published before (except as part of a thesis or lecture note or report, or in the form of an abstract ); that it is not under consideration for publication elsewhere; that its publication has been approved by all co-authors as well as by the authonties at the institute where the work has been carried out; that written permission of copyright holders was obtained by the authors for material used from other copyrighted sources; that, if and when the manuscript is accepted for publication, the authors hand over the transferable copyrights of the accepted manuscript to the publisher; and that the manuscript or parts thereof will thus not be published elsewhere in any language without the consent of the copyright holder. Copyrights include, without spatial or timely limitation, the mechanical, electronic and visual reproduction and distribution; electronic storage and retrieval; and all other forms of electronic publication or any other types of publication including all subsidiary rights. Form of the manuscript: Each manuscript must be submitted in triplicate, one original and two copies, in English or German in A4 format (21x 30 cm) or correspondingly either from a typewniter or legible computer print with a left- hand margin of 4 cm and true double spacing. Pages must be numbered. Script type should be uniform throughout the manuscript. Use of bold print, italics and spaced-letters must be avoided. Page footnotes and appendices at the end of the text are not allowed. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At the end of the manuscript following References, the exact postal address(es) of the author(s) should be given and the e-mail address of the senior author only. The first page of the manuscript should contain the following information: 1. title (in case of a long title, a running title not exceeding 72 characters must additionally be provided) 2. name(s) of author(s): full first name(s) for all authors (an excessively large number of co-authors should be avoided). 3. department(s), university affiliation(s), city and country Content of the manuscript: Manuscripts can be publi- shed as original investigations, short communications or reviews. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 type-written pages. The text should be divided into: Abstract (in English), Introduction, Material and Methods, Results, Discussion (or together as Results and Discussion), Acknowledgements, Zusammenfassung (in German), References. In English manuscripts a German title must precede the Zusammenfassung; in German manuscripts an English title must precede the Abstract. b) Short communications: Short communications must not exceed 10 typewritten pages and do not require either an Abstract or Summary. They also should not be headed Mammalian Biology Zeitschrift für Säugetierkunde with subtitles into Introduction, Materials and Methods, Results, and Discussion but should be organized according to this form. A German title must be added to an English, an English to a German manuscnipt. c) Reviews: Manuscripts that review and integrate the current state of knowledge in a special field of mammalian biology are also welcome. They must not exceed 50 typewritten pages. The text must provide Abstract (in English), Introduction, special headings depending on the theme, Acknowledgements, Zusammenfassung (in German), References. A German title must precede the Zusammen- fassung in English manuscripts and vice versa an English title must precede the Abstract in German manuscripts. Key words: Up to 5 informative key words, starting with the taxonomic unit(s), must be given following the Abstract or at the beginning of the text in Short communications. Illustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focused and readily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. Labelling on figures should be uniform, printed in a sans-serif font like Arial or Helvetica, and large and clear enough for eventual reduction in size. On the back of each illustration the following information should be given, using a soft pencil: figure number, name of the journal, first author. Captions to figures must be written on a separate sheet of paper. Tables: The number and size of tables must be reduced to a minimum. Each table should appear on a separate page including heading and running number. Comprehensive tables not fitting onto one page should follow on consecutive Sheets. References: Each study cited in the text must appear in alphabetical order at the end of the text. Only internationally available literature and citations from peer reviewed journals must be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples: a) From journals: KaLko, E. K. V.; Hanpuev, C. O. Jr. (1994): Evolution, biogeography, and description of a new species of fruit- eating bats, genus Artibeus Leach (1821), from Panama. Z. Saugetierkunde 59, 257-273. b) From books: HERRE, W.; Röhrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) From reference books: NIETHAMMER, J. (1982): Cncetus cncetus (Linnaeus, 1758) - Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. By J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/T, 7-28 Printing: After acceptance authors may additionally provide the manuscript also as text file on disk together and identical with the final submitted version. The principal author will receive page proofs of the manuscript as well as a set of the illustrations and the original manuscript for proof reading. Only type-setting errors should be corrected. Any major revision undertaken by the author at this stage will result in costs to the authors themselves. The conventional proof-readers° symbols should be used. The corrected page proofs should be signed by the author to indicate that the manuscript is "ready for print" and be sent to Prof. Dr. Peter Langer, Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Aulweg 123, D-35385 Giessen,Germany (email:peter.langer@anatomie.med.uni-giessen.de) Reprints: 50 reprints of each paper are supplied free of charge. When returning the page proofs additional reprints can be ordered at the cost of the author(s). - - URBAN & FISCHER Verla Mammalian Biology Jen /Gernu n - ne = s ISSN 1616-5047 Zeitschrift für Säugetierkunde Mamm. biol. : 66(2001)6 - 321-384 Contents Original investigations Scandura, M.; Apollonio, M.; Mattioli, L.: Recent recovery of the Italian wolf population: a genetic investigation using microsatellites - Neues Anwachsen der italienischen Wolfspopulation: eine genetische Untersuchung mittels Mikrosatelliten _ 321 Conroy, C. J.; Hortelano, Y.; Cervantes, F. A.; Cook, J. A.: The phylogenetic position of southern relictual species of Microtus (Muridae: Rodentia) in North America - Die phylogenetische Stellung südlicher Reliktarten von Microtus (Muridae: Rodentia) in Nordamerika 0 2,5 Nupp, T. E.; Swihart, R. K.: Assessing competition between forest rodents in a fragmented landscape of midwestern USA - Abschätzungen zur Konkurrenz von im Wald lebenden Rodentiaspecies in einer fragmentierten Landschaft im mittleren Westen dr SA 0,5 Holley, A. J. F.: The daily activity period of the brown hare (Lepus europaeus) - Die tägliche Aktivitätsperiode des Europdaischen Feldhasen (lerus uropaeus) Short communications Wittmer, H. U.: Home range size, movements, and habitat utilization of three male European wildcats (Felis silvestris Schreber, 1777) in Saarland and Rheinland-Pfalz (Germany) - Wohngebietsgröße, Ortsveränderung und Habitatnutzung von drei männlichen Wildkatzen (Felis silvestris Schreber, 1777) im Saarland und in Rheinland-Pfalz Deutschiand) 5 Koh, H. S.; Randı, E: Genetic distinction of roe deer (Capreolus pygargus Pallas) sampled in Korea - Genetische Eigenständigkeit von Rehen (Capreolus pygargus Pallas) ausKrea 3, Scharff, A.; Macholan, M.; Zima, J.; Burda, H.: A new karyotype of Heliophobius argenteocinereus (Bathyergidae, Rodentia) from Zambia with field notes on the species - Ein neuer Karyotyp von Heliophobius argenteocinereus (Bathyergidae, Rodentia) aus Sambia und Freilanddaten zur At 2 26ßf, Vargas, J.; Simonetti, J. A.: New distributional records of small mammals at Beni Biosphere Reserve, Bolivia - Neue Nachweise zur Verbreitung von Kleinsäugern aus dem Beni Biosphere Reservat, Bolivien 379 Book reviews ___ 0 200 00 oe en 2 0 Sa Table of Contents now also available via e-mail by free-of-charge ToC Alert Service. Register now: http://www.urbanfischer.de/journals/mammbiol Abstracted/Indexed in Animal Breeding Abstracts/Biological Abstracts/Biological Environmental Sciences/BIOSIS database/Current Advances in Ecological and Environmental Sciences/Current Contents - Agriculture, Biology & Environmental Sciences/Dairy Science Abstracts/Fisheries Review/Helminthological Abstracts/Index Veterinarius/Key Word Index to Wildlife Research/South Pacific Periodicals Index/Veterinary Bulletin/Wild Review (Fort Collins)/Zoological Record | | % Eh a das u N ii 103 Hr Fr ei T | Juneo2 | apa D TENVer nn re Pk RR anna unenan, ni ua Rum Lade Tr, Amen, nn he Saint Im 1, am a ae hr le han x Be “un 3 E fer 2 b weh esi.ı mar s N lee DR > h = { h REN AD 2 Ei er run S . A NE Hrar a ke, en N einge Kun Aare eure B ame Bahnen neke, Bears Vearengjpin, y i i ? ä . £ % DR n S EN aareeee " i E fi € er en N Er 12 . D Anika Ge RER f FEN en 5 weh; Y no ENTER, MAR yARL Br nn Pr ek Arnd se