- » . + pr ? PR “ ” r 4 Bee je RR N & : 18 > y Aust Nie Er af DON N i il u : ; Ru vente : HERE g INNERE: ? > £ Ren ? 5 ne 5 > B = 3 Ken ERDE TEL: - TUR ER TIERE. Bor ee 5 NEW : \ v N R N a - } e ’ 2 ler 5" n Sunny nareialk ira Nr SR ; en ER ERRCR BL \ ; i \ TEHNEeTUEE Er Sue vr n x e H £ PN = Ri ty TR A La f BR r Dat I rent NTRE 2 5 A i 5 N haPa alter ne 2 K } EB SER Eren : # i EN L N » x ran er par NE Frans UM NRES ART ERS Bun. Malle BOHH NE > Ar DIESEN ei Baaduge vr ” R: Y ua R % a “ er j a ET ER SRN: SITE DE NZ ei at I via ER Er eher FR ERTS PEN SEN RATEN Na nenan “ E E SIEHE N x R nannte ae wär BESEETMENNE SR a Selena er herein Dt KENNEN: weraiuren Ar; “er Ann ; 5 An r Ban Pi ne Eh N ur Ainar rss ERS i ; Pr Ey Brain 2 e { y NS ERLS RR TINER EI 20 KU BET f Keane ha 2, & R v. SE VERUg ee van rin bit) a 2 ER YTEIE u Ein ES st Sat ; ; KERT ne RRSITEHE ERNES 5 A Da r R eat B E Ve War DE Def messe = Ä c RE EEE SER iur r 5 ERZLENNBÖRK hi x ; Wk RATEN E Rn ER ” Y N Sa 2 ten de a RERLN 5 Vorl Ks enter 2 x r hr er DR TERN Ee . URL wo 4 N Dr TRINKT) N ae x ie EN ERONLEREIER s : 3 BER H x ; ; 7 uw en 5 e =; a > 2 > “ IND: S h tina EEE Dahn a a A LEE 4 PRRERCRTT Fast i r 3 aß Ivy MYRRNEN : i kalt ? es ns i B: PR BES a s i 2 N 2 re wände ehr ent Dee E 3 ERRAR $ x Saspalte hmeNen ee Eryens u ö $ ER EURE, : L : S vr ns ea Sarah GAR Dit Ta en STE DE TOW ER TERNE er B E Era ! R Ereste " ger WIR ee Frese BEINE tert DER EN REITS x ; * ; ; > 2 , ER - Eu nr rt En x en BEEPTERFENER ya r $ 5 : B 5 Y ul uNa a PELEuRIETE agree Van tar Bye Ne : Eee 2 Honor ; = FRE EFNN IHRE in Bulıe wear: IRELESEIEN ? X 3 R N g 5 E i N Na & rTer ä > ER RRS AUT ER t Se: Ser Tara rr Fa ER RRNe f une Senna aber ehhmeitn 5 , “ PLPRE ES a YuR ran er 5 5 x umwelt ; ß Bert = RE “ Speer \ 3 EV. ® z x 2 Bi 2 r z 3 Rn g S n x “ Ash bt sen BE Ne ir x 2 7 “ i “ R ardmasje ii Mm Y I“ Az L: » En \ 7 g n ei d ie e B , BERNER LTE ac“ g ts RN : : i SONNE u Bon BAER s RER BE hd yet ne eh wi BSR erhal (RE RER ENESTE ee ee 5 x Nrertart eher Bit enn. Sr vn NN x Ne te en ie Uns ante gen = 3 er Beer ERLINNE TR TE En en Ba i ’ u von Bd Flle apart Vorige eite Hu ZU us Rare te ae BEE Da FRE , N Baer f De fr 2 RR PT i Yu ur re REP R Wellen een ne x . a " a an S ’ aa In $ BUSCENSTEEESE a a ak REN PEN TED. re are Ber 5 S : : } ’ TB EN R E 5 MED Re en Mia Se Fre Anni Eee" \ Sa s h u rn SEHETe tu ai u RE N a eat. Were are Wut re RT IE IE TERN Nena un Mt vn Ze ee) Er Zu er EIS REURIN BRIGERETERRE ’ Sera Fer hen uam ren Per re rund EN NT tr eg RATEN ish re St: are 2 Par Nie Ar plain m ea een nee s deng “aradın nahe KA! ABER LATET we rh Keane: ns Anti a ie md “ En NE ar a yew NE ERP er As : 1a. BER En Bruce IR N VURRELSE A ale 3 ’ BR ER LE Ana en : i en ht d. Y.ı R 7 ; a u ; ana. INA FAR | HU ER AR U iR FINE N | N ı a. IF . N) 2 ER va fi No N ı { g ru ” . L I ie ri ) j HR H FOL N IB La ah tms rn auf Au f Hi, f r ne 2; f . Kt I n UrsE { | . , 1 DR UR ge N AN LT ’ Bi - AR f, P v r Sa 9) PRTELUN | 1 Ye ie er ar ZEITSCHRIFT * SÄUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Organ der Deutschen Gesellschaft für Säugetierkunde Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kiel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel - R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel - M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt - P. Vogel, Lausanne GSUSTAV i FISCHER ms JENA » STUTTGART. NEW YORK 1 9 96 ın m E< o m 2 N u BONIS ARTIBUS II Inhaltsverzeichnis von Band 61 Wissenschaftliche Originalarbeiten Apollonio, M.; Vailati, G.: Functional morphology of metatarsal glands in Fallow Deer (Cervus dama). — Funktionsmorphologie der Metatarsaldrüsen beim Damhirsch (Cervus dama) ............................. Banaszek, Agata; Ratkiewicz, M.: Fedyk,S.; Szalaj, K. A.; Chetnicki, W.: The chromosomes and isoen- zymes in marginal populations of the Common shrew (Sorex araneus) in the Vistula delta. - Die Chro- mosomen und Isoenzyme in Randpopulationen der Waldspitzmaus (Sorex araneus) im Weichsel- Delta... 2202202 se ea ee ee RE Barbosa, A.; Benzal, J.: Diversity and abundance of small mammals in Iberia: peninsular effect or habitat suitability? — Diversität und Abundanz bei Kleinsäugern in Spanien: Halbinsel-Effekt oder Lebens- raumanpassung? #..„snaeee Beolchini, Francesca; Dupr£, E.; Loy, Anna: Territorial behavior of Talpa romana in two different habi- tats: food resources and reproductive needs as potential causes of variation. — Revierverhalten von Talpa romana in zwei verschiedenen Habitaten: Nahrung und Reproduktionsbedürfnisse als poten- tielle,Gründe,der Variation .....22.: 2:9 7-22 re Corti, M.; Civitelli, Maria Vittoria; Castiglia, R.; Bekele, Afework; Capanna, E.: Cytogenetics of the genus Arvicanthis (Rodentia, Muridae). 2. The chromosomes of three species from Ethiopia: A. abys- sinicus, A. dembeensis and A. blicki. — Cytogenetik der Gattung Arvicanthis (Rodentia, Muridae). 2. Die Chromosomen von drei Arten aus Äthiopien: A. abyssinicus, A. dembeensis und A. blicki ....... Fischer, K.: Der Pubertätsverlauf bei männlichem Damwild (Cervus dama). — The course of puberty in male fallow:deer (Cervus dama)........ .......2...2....2.0... 020020 anne Ganem, Guila; Alibert, P.; Searle, J. B.: An ecological comparison between standard and chromosomally divergent House mice in Northern Scotland. — Ein ökologischer Vergleich zwischen Standard- und chromosomal abweichenden Hausmäusen in Nordschottland.......... 2. ......2.2.2.. 22er Giagia-Athanasopoulou, Eva B.:; Markakis, G.: Multivariate analysis of morphometric characters in the Eastern hedgehog Erinaceus concolor from Greece and adjacent areas. — Multivariate Analyse mor- phometrischer Merkmale beim Ostigel Erinaceus concolor in Griechenland und benachbarten Gebie- z Groves, C. P.: The taxonomy of the Asıan Wild Buffalo from the Asian mainland. — Taxonomie des Was- serbüffels (Bubalus arnee) auf dem asiatischen Eestland........ 2... 2.22.2222. 22 Hartl, G. B.; Kurt, F.; Tiedemann, R.; Gmeiner, Christine; Nadlinger, K.; U Mar, Khyne; Rübel, A.: Popu- lation genetics and systematics of Asian elephant (Elephas maximus): A study based on sequence var- iation at the Cyt b gene of PCR-amplified mitochondrial DNA from hair bulbs. — Populationsgenetik und Systematik des Asiatischen Elefanten (Elephas maximus): Eine Studie auf der Grundlage von Se- quenzvariation am Cyt-b-Gen von PCR-amplifizierter Mitochondrien-DNA aus Haarwurzeln.......... Hudson, Robyn: Bilkö, Agnes: Altbäcker, V.: Nursing, weaning and the development of independent feeding in the rabbit (Oryctolagus cuniculus). - Säugen, Entwöhnung und die Entwicklung unabhängi- gen Freßverhaltens/beim Kaninchen (Onyctolagus’ceuniculus) 2.2 Ivanitskaya, Elena; Shenbrot, G.; Nevo, E.: Crocidura ramona sp. nov. (Insectivora, Soricidae): a new spe- cies of shrew from the central Negev desert, Israel. - Crocidura ramona sp. nov. (Insectivora, Sorici- dae): eine neue Spitzmausart aus dem Zentralteil der Negev Wüste, Israel ....................ueeeseeeo.. Janossy, D.: Kleinsäuger aus dem Pleistozän und Holozän Südost-Europas und des Mittleren Ostens. — Small mammals from the Pleistocene and Holocene of Southeast-Europe and the Middle East.......... Kauhala, Kaarina: Habitat use of Raccoon dogs, Nyctereutes procyonoides, in southern Finland. — Habi- tatnutzung beim Marderhund, Nyctereutes procyonoides, ım südlichen Finnland........................... Korz, V.; Hendrichs, H.; Militzer, K.: Behavioural and anatomical correlates of sympathetic arousal and stress in male Central American Agoutis (Dasyprocta punctata). — Verhaltens- und anatomische Kor- relate zur Sympathikusaktivität bei männlichen Mittelamerikanischen Agutis (Dasyprocta punctata) .. Krystufek, B.; Hrabe£, V.: Variation in the baculum of the European souslik, Spermophilus citellus. — Var- iation des Baculum zwischen Populationen des Europäischen Ziesels Spermophilus citellus ............. Kurt, F; U Mar, Khyne: Neonate mortality in captive Asian elephants (Zlephas maximus). — Neonaten- sterblichkeit bei asiatischen Elefanten ın Menschenobhutf.......... 2.2.2....2.. me u. ..2 22 reger 65 236 193 5339 165 176 129 327 285 39 93 242 269 112, 228 Inhaltsverzeichnis von Band 61 Langer, P.: Comparative anatomy of the stomach of the Cetacea. Ontogenetic changes involving gastric proportions — mesenteries — arteries. —- Vergleichend-anatomische Untersuchungen am Magen der Ce- tacea. Ontogenetische Proportionsveränderungen —- Mesenterien — Arterien ...........2222222202eeeeessen: Lorenzini, Rita; Mattioli, L.; Rustioni, M.; Patalano, Marianna: Allozyme and craniometric variability in the Roe Deer (Capreolus capreolus L.) from Central Italy. - Allozym- und kraniometrische Variabili- tasbesrehent(@apreolusteapreolus)),aus)Mittelitallen. rg. Macholän, M: Multivariate morphometric analysis of European species of the genus Mus (Mammalia, Muridae). - Morphometrische multivariate Analyse europäischer Arten der Gattung Mus (Mamma- ie, Mierilda®)} o 4.0.8000 Km ea EEE Matson, J. O.; Christian, D. P.: Patterns of variation in cranial size and shape in two coexisting gerbilline rodents. — Variationsmuster von Schädelgröße und Schädelform bei zwei syntopen Nagetierarten (Gerbillirae, |Rodenii@) s»s.%2 80000 does ee EN N Meia, J.-S.; Weber, J.-M.: Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains. — Sozialstruktur des Rotfuchses (Vulpes vulpes) im Schweizer Jura ................22222222eeeeeeseseneneeeeen Nogales, M.; Medina, F. M.: A review of the diet of feral domestic cats (Felis silvestris f. catus) on the Can- ary Islands, with new data from the laurel forest of La Gomera. - Ein Überblick über die Nahrung von verwilderten Hauskatzen (Felis silvestris f. catus) auf den Kanarischen Inseln unter besonderer Benue&sichveungden Borbeerwalder aut PaGomeran.... nen oneannoaannenenesnennoeenner san: Pandolfi, M.; De Marinis, Anna Maria; Petrov, l.: Fruit as a winter feeding resource in the diet of Stone marten (Martes foina) in east-central Italy. — Früchte als Winternahrung des Steinmarders Martes foina m Minel-Ost-[allen ..000sssssansae ee ER RE OERNELERFIERERER Perez-Barberia, F. J.; Machordom, Annie; Fernandez, J.; Nores, C.: Genetic variability in Cantabrian cha- mois (Rupicapra pyrenaica parva Cabrera, 1910). — Genetische Variabilität bei der Kantabrischen GensehKupIeaprayDyTenaicapanvar@abrera, 1910) nenne eeennenen sea nnnnnesemenene nenne ns Rieger, I.: Wie nutzen Wasserfledermäuse, Myotis daubentonii (Kuhl, 1817), ihre Tagesquartiere? — How desDaubentons;sipats Myotis/7daubentonii, use their day 100SstS?...............0.20222nseneoeennesenonnenoe: Rosi, Maria I.; Cona, Mönica I., Puig, Silvia; Videla, F.; Roig, V. G.: Size and structure of burrow systems of the fossorial rodent Ctenomys mendocinus in the piedmont of Mendoza province, Argentina. — Maße und Struktur der unterirdischen Gänge von Nagetieren der Art Ctenomys mendocinus auf dem BiednonsmfdedEroynz7 Mendoza, Argentinien... ass snnsnensnnne nenne nen enneinne Schwarz-Weig, Esther; Sachser, N.: Social behaviour, mating system and testes size in Cuis (Galea muste- loides). — Sozialverhalten, Paarungssystem und Hodengewichte bei Wieselmeerschweinchen (Galea munstelondles)) ©. 0...00% 0800 A ee Tiedemann, R.; Harder, J.,; Gmeiner, Christine; Haase, E.: Mitochondrial DNA sequence patterns of Har- bour porpoises (Phocoena phocoena) from the North and the Baltic Sea. — Untersuchungen an der mitochondrialen DNA von Schweinswalen (Phocoena phocoena) aus Nord- und OÖstsee................. Torre, I.; Tella, J. L.; Arrizabalaga, A.: Environmental and geographic factors affecting the distribution of small mammals in an isolated Mediterranean mountain. — Verbreitungsbedingte Umweltfaktoren für Kiemsauremunemenasolierien mediterranen Gebirge... een seen sangen nnnenenene een Turni, H.; Müller, E. F.: Unterscheidung der Spitzmausarten Sorex araneus L., 1758 und Sorex coronatus Millet, 1828 mit Hilfe einer neuen Diskriminanzfunktion. — Discrimination of the shrew species Sorex araneus L., 1758 and Sorex coronatus Millet, 1828 by help of a new discriminance function .............. Virgös, E.; Casanovas, J. G.; Blazquez, T.: Genet (Genetta genetta L., 1758) diet shift in mountains of cen- tral Spain. - Nahrungswechsel bei Ginsterkatzen (Genetta genetta L., 1758) in den Gebirgen von Zen- TASDDMIEN, s 000 00845 0000 R Bo ee NE Wissenschaftliche Kurzmitteilungen Gattermann, R.: Interindividuelle Zyklusdesynchronisation bei Goldhamsterweibchen, Mesocricetus aur- atus. - Interinduvidual estrous desynchrony in female Golden hamsters Mesocricetus auratus ..........- ' Hingst, Oona; Blottner, S.; Meyer, H.H. D.: Role of apoptosis in seasonal involution and recrudescence of testis. - Die Rolle der Apoptose bei saisonaler Involution und Reaktivierung des Hodens............ Hirakawa, H.: Hard faeces reingestion in the Mountain hare Lepus timidus. - Wiederaufnahme von har- (EmaK@laHeimESchnechasenellen stud ee see dene: III 140 304 295 215 276 202 352 25 104 365 18 221 54 I IV Inhaltsverzeichnis von Band 61 Kock, D.; Ebenau, C.: The desert hedgehog, Paraechinus aethiopicus (Ehrenberg, 1833), new to the fauna of Syria. Der Wüstenigel Paraechinus aethiopicus (Ehrenberg, 1833), neu für die Fauna Syriens......... 189 dos Reis, S. F.; Pombal, Jr., S. F.; Pombai, J. P.; Nessimian, J. L.; Pessöa, Leila Maria: Altitudianal distribu- tion and feeding habits of Blarinomys breviceps (Winge, 1888) (Rodentia: Muridae). - Höhenverbrei- tung und Nahrungsgewohnheiten von Blarinomys breviceps (Winge, 1888) (Rodentia: Muridae) ....... 253 Robinson, T. J.; Bothma, J. du P;; Fairall, N.; Harrison, W. R.; Elder, F. F. B.: Chromosomal conservatism in southern African Klipspringer antelope (Oreotragus oreotragus): a habitat specialist with disjunct distribution. -— Chromosomaler Konservatismus bei südafrikanischen Klippspringern (Oreotragus or- eotragus), einem Habitatspezialisten mit disjunkter Verbreitung”... 49 Topping, M.; Kruuk, H.: Size selection of prey by otters, ZLutra lutra L.: An experimental approach. — Größenselektion der Beute von Fischottern Lutra lutra L.: Ein experimenteller Ansatz.................. 376 Mitteilungen der Gesellschaft .......... ...... 1.22. Er e e 62, 192, 382 Buchbesprechungen!!!..2......2..2... 0222.02 Re EEE E N PEEREREEREIEEER 126, 256, 320, 384 Indexed in Current Contents ® Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database Printed in Germany © Gustav Fischer Verlag Jena 1996 Teo © 487 MAMM ZEITSCHRIFT SÄUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Nogales, M.; Medina, F. M.: A review of the diet of feral domestic cats (Felis silvestris f. catus) on the Canary Islands, with new data from the laurel forest of La Gomera. - Ein Überblick über die Nahrung von verwilderten Hauskat- zen (Felis silvestris f. catus) auf den Kanarischen Inseln unter besonderer Berücksichtigung der Lorbeerwälder auf nennen inunsannehonssaunsnnnasnesnnnnnsnenssssnnn one nunnesedunn 1 Lorenzini, Rita; Mattioli, L.; Rustioni, M.; Patalano, Marianna: Allozyme and craniometric variability in the Roe Deer (Capreolus capreolus L.) from Central Italy. - Allozym- und kraniometrische Variabilität bei Rehen (Ca- Eee neanesanasnnnuennunnannssnenennnnnnneenaenenennhenne 7 Schwarz-Weig, Esther; Sachser, N.: Social behaviour, mating system and testes size in Cuis (Galea musteloides). — Sozialverhalten, Paarungssystem und Hodengewichte bei Wieselmeerschweinchen (Galea musteloides) ............ 25 Hudson, Robyn; Bilkö, Agnes; Altbäcker, V.: Nursing, weaning and the development of independent feeding in the rabbit (Oryctolagus cuniculus). — Säugen, Entwöhnung und die Entwicklung unabhängigen Freßverhaltens beim a nannaenusonnenunsnennnssnasnnnnans ss nn esunansnonnehin nennen 39 Wissenschaftliche Kurzmitteilungen Robinson, T. J.; Bothma, J. du P.; Fairall, N.; Harrison, W. R.; Elder, F. F. B.: Chromosomal conservatism in southern African Klipspringer antelope (Oreotragus oreotragus): a habitat specialist with disjunct distribution. - Chromoso- maler Konservatismus bei südafrikanischen Klippspringern (Oreotragus oreotragus), einem Habitatspezialisten ER 49 Gattermann, R.: Interindividuelle Zyklusdesynchronisation bei Goldhamsterweibchen, Mesocricetus auratus. — Inter- induvidual estrous desynchrony in female Golden hamsters Mesocricetus auratus...........z22444@s4@4eeennneennennen 54 Hingst, Oona; Blottner, S.; Meyer, H. H. D.: Role of apoptosis in seasonal involution and recrudescence of testis. — Die Rolle der Apoptose bei saisonaler Involution und Reaktivierung des Hodens.................222s2222sssssssen 00 59 a nnaanannen nenn sun nennen ensure namen nennen 62 Indexed in Current Contents ® Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database ICSUSTAV ıssh 0044-3468 B Z. Säugetierkunde in FISCHER Vol. 1-64 1996 Bet oO OÖ) JENA »STUTTGART»NEW YORK Februar 1996 ZEITSCHRIFT FÜR &: SÄUGETIERKUNDE Sk INTERNATIONAL JOURNAL % Ir Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kiel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G..B. Hartl, Kiel - R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel — M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt — P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) - W. Herre, Kiel (1962-1966) - H. Frick, München (1972-1976) -— M. Röhrs, Hannover (1977-1981) - H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amntierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H. G. Erkert, Tübingen - W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel — Marialuise Kühnrich, Hamburg Z. Säugetierkunde 61 (1996) 1-6 ZEITSCH RIFT® w "FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY A review of the diet of feral domestic cats (Felis silvestris f. catus) on the Canary Islands, with new data from the laurel forest of La Gomera By M. NoGALe&s and F. M. MEDINA Departamento de Biologia Animal (Zoologia), Universidad de La Laguna, Tenerife, Canary Islands, Spain Receipt of Ms. 28. 08. 1995 Acceptance of Ms. 16. 11. 1995 E APR 16 196 | LIBRARIES In this study we present ar on the diet of the feral cat on the Canary archipelago, providing the first data from the relict laurel forest (Garajonay National Park). Among the 403 prey items identified in this habitat, rats were most numerous followed by reptiles. A total of 1,047 scat groups has been stud- ied in the last eight years from the main habitats of the Canaries. Most of the 2,963 prey items identi- fied represent introduced mammals (rabbit, Oryctolagus cuniculus, mouse, Mus cf. musculus and rat, Rattus sp.). These prey were commonly captured in most habitats with the exception of high mountain scrub. In this habitat, reptiles were taken instead introduced mammals. They were also more commonly included in the diet from open ground and open forest than from the denser forest. Birds were more fre- quently consumed in the forest than in open areas. Arthropods appeared in significant proportions in the scats from habitats where these items reached high numbers. The results indicate that the diet of fer- al cats clearly varies according to the different habitats. Abstract Introduction The domestic cat (Felis silvestris f. catus) has long been associated with humans and has travelled with them to most parts of the world, including many remote and uninhabited is- lands (FITZGERALD 1988). The Canary archipelago is no exception and today feral cats are widely distributed over all the major islands, being a top predator in the food chain among the vertebrates (NoGALES et al. 1992). Most islands contain a high diversity of habitats, and their distribution depends mainly on latitude and altitude. Nevertheless, the majority of the studies concerned with the diet of feral cats carried out on other oceanic islands has been made in single habitats (MAR- SHALL 1961; van AARDE 1980; KARL and Best 1982; FITZGERALD and VEITCH 1985; RAU- zon 1985). Based on these studies, some general tendencies in the frequency of different prey items — according to latitude - have been revealed (see review by FITZGERALD 1988). Nevertheless, this variation can be very confuse and changeable, depending on the types of habitat studied. Despite the enormous difference in size, latitude and altitude of the islands, no studies have revealed the variation in the spatial patterns of the diets among the different main habitats. In this respect, the Canary Islands provide a suitable framework for such studies, due to the habitats being strongly distributed as a basic function of altitude and orienta- tion. In this study we demonstrate the spatial variation pattern of the diet of feral cats among the main habitats on the Canary Islands. 2 M. NoGALESs and F. M. MEDINA Although five publications have described the diet of feral cats on the Canaries, none of them has examined statistically the prey items found in the different habitats studied. To date, no information has been published on the diet of feral cats in the relict laurel for- est of the North Atlantic islands. Material and methods The Canary Islands are situated in the Atlantic Ocean at the closest point some 100 km from the Afri- can continent, Iying between 27°37—29°25’ N and 13°20’-29°25’ W. Garajonay National Park is situated in the upper areas of La Gomera Island (highest point: 1,487 m a.s.l.) and its area extends to over 3,974 ha. The dense vegetation is composed of native laurel forest, a relict woodland which formerly covered the margins of the Tethys Sea during the Tertiary. This wood- land is composed of some twenty species of trees, such as Laurus azorica, Ilex canariensis, Myrica faya, Erica arborea. 135 scot groups from Garajonay N.P. were analysed and a total of 1,047 scat groups of feral cats (including those from Garajonay) have been studied in the last eight years from the main different habi- tats of the Canary Islands (Tab. 1). Table 1. Studies carried out in the different habitats of the Canary Islands. Habitats = Altitude n° scats Island Authors (ma.s.l.) analysed Xerophytic scrub (islet) 0-300 110 Alegranza NOGALES et al. (1992) Xerophytic scrub (island) 0-300 200 Tenerife MEDInA and NoGALzs (1993) Juniper forest 300-600 248 El Hierro NOGALES et al. (1988) Laurel forest 600-800 135 La Gomera present study Pine forest 800-1800 185 Gran Canaria SANTANA et al. (1986) High mountain scrub 1800-3700 221 Tenerife NOGALESs et al. (1990) The scat analysis was carried out at a magnificatian of 16x after previously saturation in water. For more information on the analysis of the scat, see SANTANA et al. (1986) and NoGALss et al. (1988). Due to the fact that scats remain unaltered over a long period before their eventual disintegration, (usually more than one year), the material studied in each habitat corresponded to all seasons. For this reason it is only possible to compare the general patterns of the diet in the different habitats represented in the archipelago. Although the diet of feral cats can change according to the seasons (LiBERG 1984; HUBBs 1951; Jones and CoMAN 1982; FITZGERALD and Karı 1979), this aspect was not considered ın the present study. Statistical analyses (Chi-square test) were performed by utilizing frequencies of occurrence in scat groups for each food item. Significant differences were considered at p < 0.05. Results The diet in Garajonay National Park The analyses of scats yielded a total of 403 prey items. Introduced mammals, mainly rats (Rattus sp.), appeared most frequently and reptiles ranked second (Fig. 1). Birds and ar- thropods were less frequent in the diet. From the point of view of biomass introduced mammals were clearly the most important prey. Diet of feral domestic cats in island habitats 3 ® Mammals Birds 120 16 a 14 - 100 + — 12 + —— a 10 + 60 - 40 - 20 - 0 XSSI XSI JE LF PF HM XSSI XSI JF LF Habitats Habitats [1% Oceurrence % Biomass _ [__]% Occurrence % Biomass Reptiles Arthropods 100 80 - ae 40 - 60 - Er 30 + 40 - 20 - a 20 ,- 10 - li Um MM : | ERRRSSr XSSI xSI JF LF PF HM XSSI xSI JF LF PF HM Habitats Habitats [_]% Occurrence VUN % Biomass [__)& Occurrence VD % Biomass Fig. 1. Results obtained from the scat analyses of feral cats (Felis silvestris f. catus) in the different habi- tats of the Canary Islands. XSSI - Xerophytic scrub (small islet), XSI - Xerophytic scrub (island), JF - Juniper forest, LF - Laurel forest, PF - Pine forest, HM - High mountain scrub. Review of the diet on the Canary Islands A total of 2,963 prey items has been identified in the 1,047 scat groups analysed. Introduced mammals appeared less frequently in the diet of the high mountain areas compared with the remainder of the habitats (X? = 32.8; df= 1; P & 0.001). Rabbits (Or- 4 M. NoGALEs and F. M. MEDINA yctolagus cuniculus) showed a lower frequency of occurrence in the laurel forest than in the other habitats I = 91.5; df= 1; P < 0.001), but rats were most frequently consumed in the laurel forest (X* = 256.1; df=1; P & 0.001). Mice (Mus cf. musculus) were mostly consumed in those habitats situated below 900 m a.s.1. (x? = UP IR ee << NL). Birds were more frequently consumed in the forests than in open territorial habitats (X? =41.8; df=1;P < 0.001). In open territories and open forest areas, reptiles are more frequently included in the diet than in those forest habitats that have a dense vegetation cover (X? = 92.3; df =2; P <& 0.001). Arthropods appeared in significant proportions of the scats in those habitats where the prey items were of large sıze and relatively abundant (e.g. Juniper forest: Coleoptera, Pine forest: Chilopoda, and high mountain scrub: Orthoptera) (X? = 136.9; df=1]; P <& 0.001). Discussion Where introduced mammals are available they usually form the major camponent of the diet of cats on islands (Jones 1977, Dırks 1979; Cook and YALDEN 1980; KARL and BEST 1982). On the Canary Islands, introduced mammals are the main prey items in most habi- tats, except in the high mountain scrub. The low numbers of rabbits in this habitat are compensated by lizards of medium size (about 25 g). Taking into account the fact that fer- al cats are versatile generalist predators (ANDERSSON and ErLINGE 1977, PascAL 1980; VeitcH 1985), and endemic lizards (eg. Gallotia galloti) are very common in the high mountain scrub, and are active virtually throughout the year, it is not difficult to under- stand this alternative diet. A further example of the generalist predator behaviour of the feral cats can be ob- served with the substitution of rabbits for rats in the laurel forest. Rats are very abundant in this habitat, being particulary dependent upon seeds from some arboreal species which produce fleshy fruits. On other oceanic islands, rats are heavily preyed upon where rab- bits and mice are absent, such as on Stewart Island (New Zealand) (KArL and Best 1982). Although mice were relatively abundant among the diet of cats in habitats of the Ca- nary Islands situated at altitudes between 0 and 900 m a.s.1., the xerophytic scrub of the islet of Alegranza where this rodent is extremely abundant, showed the highest predation rate. Mice are most frequent in the diet in studies from latitudes between 28° and 50° North. Although mice are present on several tropical islands where cats have been studied (see review by FITZGERALD 1988), they were either not recorded in the guts or scats, or were recorded infrequently. On the Canary Islands, birds are commonly included in the diet of cats in forest areas, but elsewhere, seabirds comprise the largest proportion of the birds eaten, especially on small oceanic islands (DERENNE 1976; VAN AARDE 1980; FITZGERALD and VEITcH 1985; RAuzon 1985; KIRKPATRICK and RAuzon 1986). Land birds predominate as the major com- ponent of the diet of rats only on the New Zealand mainland and offshore Stewart Island, and on the North Atlantic islands of the Canary archipelago, Heisker (Scotland) and Hel- goland (Germany). Reptiles were frequently consumed in open habitats, even in the juniper forest, which may probably be due to their higher vulnerability in these areas while thermoregulating. The frequency of reptiles in the diet of cats on islands shows a similar pattern to that on the continent, with high frequencies of occurrence at low latitudes (see Konecny 1983; LAurIE 1983; FITZGERALD 1988). This interpretation should be considered carefully be- cause the highest frequency of occurrence in the world has now been recorded in the high Diet of feral domestic cats in island habitats I mountain scrub of the Tenerife Island (Canaries), which is situated in a subtropical region (NoGALss et al. 1990). For this reason, it is necessary to take into account the fact that the diet of the feral cat can change according to the different habitats, which are significantly influenced not only by latitude, but also by altitude and orientation. Arthropods appeared at significant ratios where they were of large size. The predation pattern on arthropods is not clear, but some cats in European countries apparently catch many invertebrates when they are young (Howe 1982). Insects normally vary greatly dur- ing the different seasons (Hußes 1951). Feral cats in many parts of their wide distribution are opportunistic predators (An- DERSSON and ERLINGE 1977). Our results could provide a general notion of the abundance of some prey on the Canary Islands habitats. Acknowledgements The authors would like to express their gratitude to Ornistudio S.L. specially to K. EMMERSON, M. OrAMAS and E.C. HERNANDEZ for donating some of the material and data. A. MARTIN and J. C. RAnDo identified some prey items, and the ICONA’s personal from La Gomera Island gave logis- tic support. All the studies carried out on the diet of feral cats on the Canaries were done without any government economic support. BERNIE ZONFRILLO revised the manuscript and made helpful comments. Zusammenfassung Ein Überblick über die Nahrung von verwilderten Hauskatzen (Felis silvestris f. catus) auf den Kana- rischen Inseln unter besonderer Berücksichtigung der Lorbeerwälder auf La Gomera In dieser Studie wird eine Übersicht über die Nahrung verwilderter Hauskatzen (Felis silvestris f. catus) auf den Kanarischen Inseln gegeben. Erstmals werden Daten aus den Restbeständen der Lorbeer- wälder im Garajonay Nationalpark geboten. Unter den 403 Beutebestandteilen, welche in diesem Le- bensraum bestimmt werden konnten, waren Ratten besonders zahlreich. Reptilien stellten die zweithäufigste Beute dar. In acht Jahren sind 1047 Kotproben untersucht worden, welche aus verschiedenen Lebensräumen der Kanaren stammten. Den Hauptanteil der 2963 identifizierten Beutereste stellten auf die Kanaren eingeführte Säuger, wie Kaninchen (Oryctolagus cuniculus), Hausmaus (Mus musculus) und Ratten (Rattus sp.) dar. Mit Ausnahme der Buschzonen auf höheren Bergen wurde diese Beute in allen Lebens- räumen der Katzen geschlagen. Auf den Bergen traten Reptilien an die Stelle der eingeführten Säuge- tiere. Auch im offenen Gelände und in lichten Wäldern spielten Reptilien eine größere Rolle als in dichten Wäldern. Vögel bildeten in Wäldern eine häufigere Beute als in offenen Landschaften. Arthro- poden waren in signifikanten Anteilen aus jenen Habitaten nachweisbar, in denen sie in größerer An- zahl auftraten. Diese Studie zeigt, daß sich die Nahrung der Katzen in verschiedenen Lebensräumen unterscheidet. References ANDERSSON, M.; ERLINGE, S. (1977): Influence of predation on rodent populations. Oikos 29, 591-597. Cook, L. M.; YALDEN, D. W. (1980): A note on the diet of feral cats on Deserta Grande. Bocagiana 352, 1-4. DERENNE, P. (1976): Notes sur la biologie du chat haret de Kerguelen. Mammalia 40, 531-595. Dırks, P. J. (1979): Observation on the food of feral cats on Campbell Island. New Zealand J. Ecology 2, 64-66. FITZGERALD, B. M. (1988): Diet of domestic cats and their impact on prey populations. In: The domestic cat: the biology of its behaviour. Ed. by D. G. Turner and P. BATEson. Cambridge: Univ. Press. Pp. 123-144. 6 M. NoGALES and F. M. MEDINA FITZGERALD, B. M.; KARL, B. J. (1979): Food of feral house cats (Felis catus L.) in forest of the Orongo- rongo Valley, Wellington. New Zealand J. Zoology 6, 107-126. FITZGERALD, B. M.; VEITCH, C.R. (1985): The cats of Herekopare Island, New Zealand; their history, ecology and effects on birdlife. New Zealand J. Zoology 12, 319-330. Howe, C. (1982): Yorkshire kittens rule O.K.!. In: The domestic cat: the biology of its behaviour. Ed. by D. C. TURNER and P. BATEson. Cambridge: Univ. Press. Pp. 132. Husss, E. L. (1951): Foods habits of feral house cat in the Sacramento Valley. Calif. Fish and Game 37, 177-189. Jones, E. (1977): Ecology of the feral cat, Felis catus (L.) (Carnivora: Felidae) on Macquarie Island. Australian Wildl. Res. 4, 249-262. JonEs, E.; CoMan, B. J. (1982): Ecology of the feral cat, Felis catus (L.) in South Eastern Australia. III. Home ranges and population ecology in semi-arid North West Victoria. Australian Wildl. Res. 9, 409-420. Kart, B. J.; Best, H. A. (1982): Feral cats on Stewart Island; their foods, and their effects on Kakapo. New Zealand J. Zoology 9, 287-294. KIRKPATRICK, R. D.; RAUZon, M. J. (1986): Food of feral cats Felis catus on Jarvis and Howland Island, Central Pacific Ocean. Biotropica 18, 72-75. Koneceny, M. J. (1983): Behavioural ecology of feral house cats on the Galapagos Islands, Ecuador. Diss. thesis Univ. Florida, Gainsville. LAURIE, A. (1983): Marine iguanas in Galapagos. Oryx 17, 18-25. LißErG, ©. (1984): Home range and territoriality in free ranging house cats. Acta Zool. Fennica 171, 283-285. MARSHALL, W. H. (1961): A note on the food habits of feral cats on Little Barrier Island, New Zealand. New Zealand J. Science 4, 822-824. MEDINA, F. M.;, NoGALESs, M. (1993): Dieta del gato cimarrön (Felis catus L.) en el Piso Basal del macizo de Teno (Noroeste de Tenerife). Donana Acta Vertebrata 20, 193-199. NOGALES, M.; MARTIN, A.; DELGADO, G.; EMMERSON, K. (1988): Food spectrum of the feral cat (Felis ca- tus L., 1758) in the juniper woodland on El Hierro (Canary Island). Bonner Zool. Beitr. 39, 1-6. NOGALES, M.; ABDOLA, M.; ALONSo, C.; Quiuis, V. (1990): Premieres donn&es sur l’alimentation du chat haret (Felis catus L., 1758) du Parc National du Teide, Tenerife (Iles Canaries). Mammalia 54, 190- 196. NOGALES, M.; RODRIGUEZ, J. L.; DELGADO, G.; QuiLis, V.;, TrUJILLO, O. (1992): The diet of feral cats (Fe- lis catus) on Alegranza Island (North of Lanzarote, Canary Islands). Folia Zool. 41, 209-212. PAscAL, M. (1980): Structure et dynamique de la population de chats harets de l’archipiel des Kergue- len. Mammalia 44, 161-182. RAUZzon, M. J. (1985): Feral cats on Jarvis Island: Their effects and their erradication. Atoll Res. Bull. 282, 1-30. SANTANA, F.; MARTIN, A.; NOGALES, M. (1986): Datos sobre la alimentaciön del gato cimarrön (Felis ca- tus Linnaeus, 1758) en los montes de Pajonales, Ojeda e Inagua (Gran Canaria). Vieraea 16, 113- 11972 VAN AARDE,R.J. (1980): The diet and feeding behaviour of feral cats, Felis catus at Marion Island. South African J. Wildl. Res. 10, 123-128. VEITCH, C.R. (1985): Methods of eradicating feral cats from offshore island in New Zealand. In: Con- servation of Island Birds. Ed. by P. J. Moors. ICBP Techn. Publ. 3, 125-141. Authors’ address: Dr. MANUEL NOGALES and FELIX M. MEDINA, Departamento de Biologia Animal (Zoologia), Universidad de La Laguna, E-38206 Tenerife, Canary Islands, Spain Z. Säugetierkunde 61 (1996) 7-24 ZEITSCHRI FT® < & u FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Allozyme and craniometric variability in the Roe Deer (Capreolus capreolus L.) from Central Italy By RırtA LoRENZINI, L. MATTIOLI, M. RUSTIONI, and MARIANNA PATALANO Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale”, Teramo, Italy; D.R.E.AM. Settore Fauna, Poppi, Arezzo, Italy; Dipartimento di Scienze della Terra, Universita di Firenze, Italy. Receipt of Ms. 16. 05. 1995 Acceptance of Ms. 29. 09. 1995 Abstract Roe deer from the provinces of Siena and Arezzo, central Italy, were investigated for genetic and mor- phometric differentiation. A total of 162 individuals from seven sampling sites (4 in Siena, 3 in Arezzo) were analysed at 40 presumptive gene loci by means of vertical polyacrylamide gel electrophoresis. Se- ven loci were polymorphic for at least two alleles. The mean proportion of polymorphic loci (P) was 16.2% in Siena, and 15.3% in Arezzo. Mean expected heterozygosity (H.) was 5.3% and 4.0% in Siena and Arezzo, respectively. Both the values of relative (Fsr=8%) and absolute (mean Ner’s 1972 D = 0.007, mean modified Rogers D = 0.084) genetic distance suggest a significant differentiation be- tween the subpopulations of Siena and Arezzo, possibly due to the different demographic origin of roe deer from the two provinces. Morphometric variation in 28 cranial and mandibular traits, measured in 48 individuals, was subjected to principal component and discriminant function analyses. Both methods allowed a quite clear separation of specimens from the two provinces, revealing significantly greater skull dimensions in the roe deer from Arezzo. Although genetic factors might be involved, a higher den- sity of deer within the wooded areas of Siena was hypothesized as the main environmental cause of morphometric differentiation. Introduction The roe deer, Capreolus capreolus (Linnaeus, 1758) is recognized as one of the most eco- logically adaptable species of deer (NEUHAUS and ScHAICH 1985; LEHMANN and SÄGESSER 1986; Kurr 1991). Its great biological plasticity enabled this species to colonize different habitats successfully, e.g. wooded areas or open cultivated landscapes. Field and forest dwelling roe deer have been identified as different ecotypes, showing remarkable pheno- typic variability and differences in breeding biology and social organization (PIELOWSKI 1977; FRUZINSKI et al. 1982; Fruzınskı and LABUDZKI 1982; MAJEWSKA et al. 1982; PıE- LOwskI and BRESINSKI 1982; Kurt 1991). The intraspecific morpho-ecological variation observed between different forms could be generated by comparatively high genetic differentiation, justifying the existence of dis- tinct subspecies or geographical ecotypes. On the other hand, the reaction norm of the genotype of roe deer could be responsible for a wide range of phenotypic variation in- duced by gene-environment interactions. Extensive studies on the morphometry, ecology, and biochemical-genetic variability of the European roe deer have been conducted to date. These investigations attributed the wide intraspecific variation of this deer to the dy- 8 RıTA LorRENZINI etal. namic rearrangement of genetic diversity within populations and to different habitat con- ditions (MARKOWSKI and MARKOWSKA 1988; ZEJDA and KouBEK 1988; ZımA 1989; ZımA et al. 1989; Fanpos and ReıG 1993; MARKOWSKI 1993), to human activities such as hunting and/or reintroductions (HARTL et al. 1991; LorRENZINI et al. 1993), to the breeding behav- iour (APpoLLonIo and HARTL 1993; Kurt et al. 1993), and to population histories (LOREN- zını et al. 1993). These results seem to indicate that there are no emerging subspecies of Capreolus capreolus, rather they suggest a high phenotypic plasticity of this species. This view is reinforced by analyses of non-metric variation. FANDos and ORUETA (1991) de- tected a high variability in epigenetic skull traits of roe deer from the western Cordillera Cantabrica and suggested an adaptive value for most of those characters. Similar investi- gations revealed a high level of genetic variability also in populations from Poland (MAR- KOWsKI and MARKOWSKA 1988). However, no significant differences between animals living in field and forest habitats exist either in cranial non-metric traits, or in fluctuating asymmetry (MARKOWSKI 1993). We studied forest dwelling roe deer for biochemical-genetic and morphological varia- tion. The aim was to investigate the level of differentiation between some populations from central Italy in relation to their different demographic backgrounds and environ- mental conditions. Material and methods A total of 162 roe deer from the provinces of Siena and Arezzo, Tuscany, Italy, were investigated. Speci- mens were collected from four sampling sites (called subpopulations here) in Siena and three in Arezzo (Fig. 1). The Arezzo population inhabits areas situated at 210-870 m above sea level, with amean wood cov- er 0f 57.4%. According to censuses carried out in 1989 and 1990, the mean density was 10.3 deer/100 ha of the total area (Lovarı et al. 1991). This population, derived from an original stock which lived in the Foreste Casentinesi, North-Eastern Apennines, has been supposed to be autochthonous (LORENZINI et al. 1993). Apart from a post war release of four individuals from the Alps, no other introduction of roe deer occurred from other stocks (CRUDELE 1988; MATTıoLı 1994). The Siena population lives at lower al- titudes (150-500 m above sea level) with amean wood cover of 26.3%. In 1989/90 the mean density was 9.6 deer/100 ha of the total area (Lovarı et al. 1991). It was formed by animals supposed to come from the Latium Maremma, and with allochthonous individuals released into the province probably from Great Britain and the former Czechoslovakia at the beginning of the century (MAZZoNI DELLA STELLA 1991). The distribution areas of the two populations met along a narrow strip only very recently, due to the spreading of roe deer from their centers of origin (Fig. 1). Heart and liver samples were collected from regular hunting during the seasons from 1989 to 1991, and stored at -20 °C until electrophoresis. Preparation of tissue extracts, electrophoretic conditions and staining procedures followed standard laboratory protocols (LoRENZINI et al. 1993) adapted from SHAW and PrAsAap (1970), Harrıs and Hopkınson (1976), and Murphy et al. (1990). Thirty enzyme and pro- tein systems encoded by 40 presumptive gene loci were analysed for polymorphisms. Loci were the fol- lowing (E. C. numbers are given in parentheses): Sod-1, -2 (1.15.1.1), Aat-2 (2.6.1.1.), Est-1, -2 (3.1.1.1), Gpd (1.1.1.49), Pep-A, -B (3.4.11), Fum (4.2.1.2), Adh-1, -2 (1.1.1.1), Ldh-1, -2 (1.1.1.27), Mod-2 (1.1.1.40), Idh-2 (1.1.1.42), Pgm (2.7.5.1), Mpi (5.3.1.8), Ck (2.7.3.2), B-Gus (3.2.1.31), Ada (3.5.4.4), Pgd (1.1.1.44), Pox-2 (1.11.1.7), Acp-1.@.1.32), Mor-1, -2 (1.1.1.37),.Dia-17.(.622), 6Ep2 & 39) a@dh (1.1.1.47), Xdh (1.2.3.2), Hk-2 (2.7.1.1), Ak-1, -2 (2.7.4.3), Hbdh (1.1.1.30), Alb, Trf, Hb-1, -2, Heart-Gp- 1, -2, -3. Genotype frequencies were obtained directly by scoring the gels. Observed single locus hetero- zygosity (h,) was tested against the expected heterozygosity (h.) based on the Hardy-Weinberg law by chi-square goodness-of-fit analysis (SoKAL and RoHLF 1981). For each subpopulation estimates of ob- served (h,) and expected (h.) average heterozygosity, and the proportion of polymorphic loci (P) were computed. F-statistics (WRIGHT 1978; Neı 1977) were calculated for each locus to quantify the relative genetic differentiation among subpopulations (Fsr), and the level of inbreeding in each subpopulation (Fis) or in the population as a whole (Fır). The inbreeding coefficient F (WrıcHT 1951) was also com- puted. Statistical significance of the Fsr and Fıs fixation indices was tested with a chi-square test follow- Genetic and morphometric variation in the Roe deer 9 e Sampling sites ı Distribution of roe deer Province of Arezzo 'S| Province of Siena Tus 50 Km Fig. 1. Sampling sites of Capreolus capreolus in the provinces of Siena (CN = Crete Nord, CS = Crete Sud, BAG = Bagnaia, VAS = Valdasso), and Arezzo (CAS = Casentino, CF = Castiglion Fibocchi, VT = Val Tiberina). ing the methods of WoRKMAN and NISWANDER (1970), and NEı (1977). The significance of Fir was tested with a t-test according to GouLson (1993). UPGMA (SnEATH and SokAL 1973) and Fitch-Margoliash (FELSENSTEIN 1979) dendrograms were constructed using NEr’s (1972) standard genetic distance. A dis- tance Wagner tree (Farrıs 1972) was computed with modified Rogers D (WRIGHT 1978) and rooted at midpoint. Electrophoretic data were evaluated statistically by the programs BIOSYS-1 of SWOFFORD and SELANDER (release 1.7, 1989), and PHYLIP of FELSENSTEIN (1989). A total of 28 cranial and mandibular measurements were made on skulls of 48 males. Since the skulls belonged to pre-existing collections, we were not able to trace the respective sampling sites of specimens to infer some microgeographic differentiations within provinces. Therefore, the samples un- der study could only be ascribed as belonging to the two main populations of Siena and Arezzo. Roe deer is supposed to be completely developed in skull and mandible growth at about three years of age, although full agreement is not achieved by all authors (GoTTscHLicH 1963; HAnus and FiISEr 1979). Be- cause of hunting management, the adult male population was shot randomly. Thus, two-year-old ani- mals were sampled as well. This was not a source of bias since the difference in the proportion of the two-year-old specimens in both Siena and Arezzo samples was not significant (chi-square = 0.43, “dr 2,9 >0.75). The age was estimated by tooth wear (AITkEn 1975). Measurements, recorded with a dial caliper to the nearest 0.01 mm, were as follows: condylobasal length (CBL), basal length (BASL), short lateral fa- cial length (LFL), prosthion-P2 length (PML), greatest prosthion length (PRL), length of the nasals (NASL), total skull length (TSL), greatest width across the nasals (NASW), greatest frontal width (FRW), least width between the orbits (ORW), neurocranium width (NEUW), width of the occipital 10 RITA LoRENZINI etal. condyles (COW), zygomatic width (ZYW), width of the right pedicle (RPW), width of the left pedicle (LPW), length from prosthion to M3 (M3L), minimum width at the crest of the maxilla (CMW), length of the premolar row (PMAL), length of molar row (MTL), length of the upper third molar (M3LE), length of the mandible from infradentale to condyle (COLM), height of the mandible (HEM), length of the diastema (DL), length of the cheektooth row measured along the alveoli on the buccal side (TRL), length from the angle of the mandible (GOLM), length of the premolar row of the mandible (PRML), length of the molar row of the mandible (MML), length of the third molar of the mandible (M3MA). Variables were named and measured according to VON DEN DRIESCH (1976). Mean, standard deviation, and coefficient of variation were computed for all the absolute values of dimensions. Interpopulation difference in mean values for each measurement was evaluated using the Student’s t-test. Ratios of mean values of the two samples were represented as logarithms in a ratio dia- gram. Morphometric relationships in multivariate space were assessed using two methods: principal component analysis, PCA, and discriminant function analysis, DFA, (AnDErson 1971; MorRISoN 1975). PCA allows the projection of multidimensional clouds of points into a space of two or three dimensions which provides a graphical representation of the reciprocal location in space of different groups. More- over, examination of variable loadings on each axis gives an indication of the relative weight of any sin- gle measurement to variation along that axis. The principal components were extracted from a correlation matrix and subjected t0 VARIMAX rotation to obtain simplified orthogonal axes. DFA was performed on the components extracted by PCA, so that the number of original measurements was lim- ited to a few common factors. The choice of a subset of variables which best summarizes the total varia- tion is sometimes recommended, since characters with little discriminating power can have destabilizing effects on correct classifications (WILLIAMS et al. 1990). Occasionally skulls were damaged and hence not all measures could be recorded for each sample. Records with missing values were deleted. To satis- fy completely the assumptions of multivariate normal distribution of the data, a transformation to na- tural logarithms of the absolute values of all measurements was performed prior to the computations. However, the use of log. instead of raw data did not yield different results. In order to remove partly the effect of size variation between groups, multivariate analyses were also conducted on the relative values of measurements, expressed as percentage of the total skull length. All the statistical analysis were done using the SPSS/PC+ package release 4.0 by Nie et al. (1975). Results Genetic variation Biochemical-genetic variants were observed at seven out of the 40 loci clearly resolved in all subpopulations (Tab. 1). No fixed allelic differences were found among samples, although one of the two low frequency alleles at the Hk-2 locus (Hk-2 “a”) was present in only the subpopulations of the province of Siena, whilst the other (Hk-2 “b”) was found in those of the province of Arezzo (“private alleles”, cf. SLATKIN 1985). Samples revealed marked differences in allele frequencies at five loci (Mpi, Mod-2, Pgm, Ak-1, Hk-2), this being rather due to heterogeneity between the two provinces than to differences among subpopulations within the provinces. Differences among subpopulations within the prov- inces in terms of WRIGHT’s (1978) fixation index Fsr were not significant (p>0.1, not shown). The differentiation among all subpopulations gave Fsr values (Tab. 2) ranging from 0.038 at Hk-2 to 0.195 at Mod-2. Single locus contingency chi-square tests yıelded highly significant Fsr values for Mod-2, Pgm and Ak-1 (p < 0.001). Mpi and Hk-2 showed chi-square values significant at p < 0.05, whereas no significant differences were found for the Acp-l and the Pep-B allele frequencies. Single locus contingency chi-square tests for departures from Hardy-Weinberg equilibrium revealed a significant heterozygote defi- ciency at Pgm (chi-square = 10.0, d.f.= 1, p<0.01) in one of the subpopulations of Siena (VAS). Mpi showed a significant heterozygote deficiency as well (chi-square = 13.2, d.f.= 1, p< 0.001) in one of the subpopulations of Arezzo (CF). In these subpopulations a consistent excess of homozygotes was demonstrated also by the positive values of WRicHT’s (1951) inbreeding coefficient (F = 1.000 and 0.774, respectively, for Pgm and Genetic and morphometric variation in the Roe deer ul Table 1. Allele frequencies at seven polymorphic loci in seven roe deer subpopulations from the provinces of Siena and Arezzo, central Italy. CN = Crete Nord, CS = Crete Sud, BAG = Bagnaia, VAS = Valdasso, CAS = Casentino, CF = Castiglion Fibocchi, VT = Val Tiberina. Sample sizes in parentheses. Allele Arezzo cs BAG (29) (22) Table 2. Summary of F-statistics by locus for the seven roe deer subpopulations studied. EP P<0.0R *** n<(0.001, NS = not significant. Pep-B Mod-2 Pgm Mpi Acp-l Ak-1 Hk-2 Mean -0.022 NS 0.128 NS 0.116 NS 0362 -0.142 NS 0.007 NS 0.063 NS 0.023 NS -0.011 NS 09987 NIS 0.410*** -0.113 NS 0.122 NS 0.099 NS 0.100 NS 0.011 NS 0495 VS 0.076* 0.025 NS ONINZ = 0.038* 0.079*** Mpi loci). At Mpi the deviation from equi- librium within subpopulations and relative to the whole population was also indicated by the significant positive values of Fıs (p<0.01, Tab. 2) and Fır (p< 0.001). The average Fsr value of 0.079 was significantly different from zero (p<0.001, Tab.2), pointing out that approximately 8% of the total amount of genetic variation was due to differentiation among subpopulations. The Fsr value yielded an estimate mNe (WRIGHT 1931) of 3.0. Expected heterozygosities in the subpo- pulations of Siena (Tab.3) ranged from 0.046 in VAS to 0.060 in CN, with a mean value of 0.053 (SD 0.007). Arezzo showed an expected average heterozygosity of 0.040 (SD 0.002) with values for subpopulations ranging from 0.039 in both CAS and VT to 0.043 in CF. The proportion of polymorphic ‚loci ranged from 0.125 to 0.175, with mean values of 0.162 (SD 0.014) and 0.158 (SD 0.029) in Siena and Arezzo, respectively. Genetic distances — NEr’s (1972) standard D and modified Rogers D (WrıcHT 1978) are given in table 4. Mean values of genetic distances averaged by province were Nei’s D 0.007 (SD 0.002) and modified Rogers D 0.084 (SD 12 RıTA LorRENZINI etal. 0.011). A rooted tree (UPGMA) and unrooted dendrograms (Fitch-Margoliash tree, Wagner network) displaying genetic relationships among subpopulations are shown in figure 2. Table 3. Indices of genetic variability in seven roe deer subpopulations from central Italy. h. (h.) = observed (expected) heterozygostity (calculated over 40 loci), P = proportion of polymorphic loci (0.99 criterion), n = mean number of alleles per locus. Arezzo Mean/SD CF Mean/SD 0.056 0.040 0.053/0.009 0.038 0.038/0.000 0.052 0.046 0.053/0.007 0.039 0.040/0.002 0.150 0.150 0.162/0.014 0.125 0.158/0.029 115 1.15 1.15 Table 4. Ner’s (1972) standard genetic distance (above diagonal) and modified Rogers distance (WRiGHT 1978, below diagonal) among the seven roe deer subpopulations studied. (A) c.c = 0.894 | mas pm Joe Soma Komme Lone Om NEI's D (x 100) 0.0 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Fig. 2(A) Genetic and morphometric variation in the Roe deer 13 (B) e.c=0.992 PP — ge dm 1 Distance from the root 0.00 0.01 0.02 0.03 0.04 0.05 0.06 Fig. 2 (B) (©) 0.001 —I Fig. 2(C) Fig. 2. Genetic relationships among the seven roe deer subpopulations studied. (A) Ner’s (1972) D/ rooted UPGMA. (B) Modified Rogers D (WRIGHT 1978)/distance Wagner tree with midpoint rooting (FArrıs 1972). (C) Ner’s (1972) D/unrooted Fitch-Margoliash tree (negative branch lengths not al- lowed). c.c. = cophenetic correlation. Morphologie measurements Univariate t-test revealed significant differences in 15 out of 28 morphometric characters between skull samples (Tab. 5). Appreciable differences involved all the cranial dimen- ‚sions, the measurements being generally larger in the skulls from the Arezzo than those from the Siena population. All the length dimensions (CBL, BASL, LFL, PML, PRL, TSL, COLM, PRML;), except length of the upper third molar (M3LE), showed greater mean values in Arezzo (Fig. 3), whereas most of the dimensions defining the width of the 14 RıTA LoRENZINI et al. Table 5. Mean (x), standard deviation (SD), coefficient of variation (CV), and univariate t-test of 28 variables measured on skulls of roe deer from central Italy. Charac- Arezzo n 24: (SD) == <0.05, = =p< 0.0, 727 —_p< 0.00 5ns Znogsieniieant * Allmeasurements are in mm > See text for explanation of variable acronyms skull were not significantly different between samples (FRW, ORW, ZYW, CMW) or attained higher mean values in Siena (LPW, RPW, NEUW; Fig. 3). The only measure related to the heigth of the skull (HEM) was significantly greater in Siena as well. Lower variability of morphometric characters was recorded in the latter population, as suggested by its generally smaller coefficients of variation. Values of CV range from 2.2 to 7.9 in Siena and from 3.4 to 9.9 in Arezzo, breadth across the nasals (NASW), width of pedicles (RPW, LPW) and width at the crest of the maxilla (CMW) showing higher variation than other skull measurements (Wilcoxon signed-ranks test = 3.7, p< 0.001). Principal component analysis of log.-transformed data extracted six components with eigenvalues greater than 1. The first component (PC-I) explained 40% of the total amount of phenotypice variability (Fig. 4). The other five components altogether ac- counted for an additional 45% of variation. The first component was of the size-type, with all coefficients positive, which suggests an influence of overall skull size on group se- paration (in Fig. 4 only coefficients equal to or greater than 0.5 are shown). The second (PC-II) and the third (PC-III) factor explained 14% of the variance each, mandibular Genetic and morphometric variation in the Roe deer 16 log ratio (x 10%) Er SEN SER Er TE BE lo 7 II IT] | CBL BASL LFL PML PRL TSL NASW COW NEUW RPW LPW M3LE HEM COLM PRML Variables Fig. 3. Diagram showing variables significantly greater in the population from Arezzo (squares in the upper part of the graph) and in that from Siena (squares in the lower part of the graph), expressed as logarithms of the ratios of mean values. (TRL, PRML, MML, M3MA) and maxillar (PMAL, MTL) length dimensions showing highly positive loadıngs on PC-I, whereas zygomatic width (ZYW), width of the right (RPW) and the left pedicle (LPW) loading positively on PC-III. Minimum width at the crest of the maxilla (CMW) had a positive loading on PC-IV (not shown). Three charac- ters loaded positively on the fifth component PC-V (M3LE, NEUW, HEM). It explained only 4% of the variation. However, since most of the meaaurements which were signifi- cantly greater in Siena (Tab. 5) loaded on this axis, it was the only one that, projected on PC-I, allowed that samples overlapped only slightly, thus forming two quite distinct clus- ters of points (Fig. 4). The structure of variable loadings suggests that roe deer with rela- tively shorter and broader skulls tended to have negative values on PC-I and positive values on PC-V. Out of the six components extracted by PCA, only the ones that minimized the Wilks’ lambda (PC-I to PC-VI, PC-IV excluded) were entered in the discriminant analysis. Be- cause of the maximization of distances, DFA was more efficient in separating the groups, which appeared as clearly separated clusters (Fig. 5). A new classification of the data based on the results of this analysis prevented misidentifications, since 100% of cases were correctly classified. When the relative values of measurements were used to calculate PCA, the remaind- ing phenotypic variation between groups was reduced (Fig. 6). The percentage of variance accounted for by the first principal components was smaller than that obtained using the absolute data. Eight components had to be extracted to exceed 80% of total variation ex- plained. Clusters formed by PC-I projected on PC-I showed a wider overlap as a conse- 16 RıTA LORENZINI etal. PC-I PC-V ln 1 0.8 ] = 0.8 0.6 J u ———u 0.6 Fe 0.4 0.4 - 0.2 | 0.2 0 0 02 - FEN Sg x EEE -0.2 LEE 04- 0.4 0.6 J -0.6 ] 0.8 - -0.8 | 1- a PC-V (4%) 3 D Siena = Arezzo 2 0 -3 -2 -1 0 1 2 3 4 9 6 1 PC-I (40%) Fig. 4. Principal component analysis of the log.-transformed values of 28 cranial measurements in two roe deer populations of central Italy. The loadings of variables relative to axes I and V are shown in the upper part of the graph (only loadings equal to or greater than 0.5 are shown). The percentage variation explained by each axis is given in parentheses. quence of the size effect of the skull which had been partly removed. Similar pictures were obtained by projecting further principal components (not shown). The first and the second components accounted for 27% and 16% of variation, respectively (Fig. 6). Most of the tooth length dimensions related to the mandible (MML, TRL, PRML, M3MA) and the maxilla (MTL, M3L, PMAL, M3LE) had all high positive loadings on PC-I, which can be interpreted again as a general size axis. As happened with the absolute data, variables failed to provide a good separation of groups along this axis. The values of vari- Genetic and morphometric variation in the Roe deer 1 count ! | |Arezzo 6 | Siena 5 4 — 3 0 u 2 ı -4.0-3.5-3.0-2.5-2.0-1.5-1.0-0.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Discriminant score Fig.5. Histogram of discriminant scores of the log.-transformed values. able loadings on PC-IH differed in magnitude and sign, implying a contrast between width (RPW, LPW, ZYW, FRW, NEUW) and length (LFL) of the skull. Roe deer with long and narrow skulls tended to have relatively large negative values on this axis. Other factors were not important since the percent of variation they accounted for was very low. Thus, the separation of the two samples was mostly attributable to differences in both the size and shape of the skulls. DFA was once again more efficient than PCA in identifying Siena and Arezzo speci- mens as different groups. The plot of discriminant function scores shows slightly overlap- ping clusters (Fig. 7). Only two out of 48 skulls analysed were misindentified (95% of cases were correctly classified). Discussion Recent investigations revealed that, in contrast to other cervids such as the fallow deer, Dama dama (PEMBERTON and SMITH 1985; HARTL et al. 1986; RanDı and APOLLONIO 1988), the moose, Alces alces (Ryman et al. 1980; Baccus et al. 1983), the red deer, Cer- vus elaphus (GYLLENSTEN et al. 1983; HARTL et al. 1990) and the sika deer, Cervus nippon (HERZOG 1988; MARKoV et al. 1992), roe deer exhibit high levels of protein variation (HARTL et al. 1991; WEHNER et al. 1991; LorEnZINI et al. 1993; Hewıson 1995). Values of expected average heterozygosity and polymorphism obtained in this study suggest that the roe deer is among the deer species with highest levels of genetic variability within po- pulations (cf. SHEFFIELD et al. 1985; RED 1985; HartL et al. 1993; MÖRSCH and LEIBEN- GuUTH 1994). In previous studies, based on comparable sets of biochemical markers screened, polymorphisms were observed nearly at the same loci in different European po- pulations (HARTL et al. 1991, 1993; LorRENZINI et al. 1993). Not surprisingly, the same vari- 18 RıTA LorRENZINI etal. PC-I PC-II D Siena m AÄrezzo -3 -2 =] 6) 1 2 3 4 5 6 7% PC-I (27%) Fig. 6. Principal component analysis of the relative (divided by total skull length) values of 28 cranial measurementis. able loci were found in all the Italian roe deer subpopulations under study, except those in which the small sample size might have influenced the finding of the less frequent al- leles. Patterns of allele frequency distribution at polymorphic loci showed considerable genetic homogeneity among the subpopulations of the same province, revealing the ab- sence of a relevant microgeographic differentiation. Subpopulations formed two major clusters in both the rooted and unrooted trees, as expected from patterns of allelic diver- sity. Within both the “Siena” and the “Arezzo” groups, the clustering of subpopulations remained basically unchanged, regardless of their low genetic distance and unequal sam- ple sizes. In each province, samples from different sites revealed no significant genetic di- Genetic and morphometric variation in the Roe deer 19 | || Arezzo 5 _ Siena 5 | m | _ | 0 m j j 1 | -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Discriminant score Fig. 7. Histogram of discriminant scores of the relative values. versity, leaving the main differentiation to the whole populations of Siena and Arezzo. In the subpopulations of Siena H- and P- values were among the highest as yet detected in this species (BAccus et al. 1983; HArTL et al. 1991, 1993, WEHNER et al. 1991). The relative genetic differentiation of 8% was well in the range of values obtained in other investigations (6.4%, 8.5%, and 12.6% Hartı et al. 1993, 1988, 1991). Ner’s (1972) absolute genetic distance between the two main populations was similar to or slightly higher than the values obtained previously in this species and generally for local popula- tions in deer (Ryman et al. 1980; PEMBERTON and SMITH 1985; ROED 1986; HArTL et al. 1990). The effects of random drift could account for the differences in allele frequencies and distribution of polymorphisms in the roe deer from Siena and Arezzo. By contrast, they are not pronounced enough to be explained by different selective pressures. The human influence and the eco-geographical conditions of their habitats are, on average, very simi- lar indeed, as revealed also by the lack of microgeographic structuring. Therefore, during the relative short time the populations have inhabited those regions, environmental con- straints and selection should have acted in a similar way, thus preventing any noticeable genotype differentiation. Moreover, they have lived as neighbouring populations for the last few years, so that some extent of gene flow exists, which might have prevented a sub- stantial genetic isolation (SLATKIN 1987, MAYNARD SMITH 1991). There must be a demo- graphic explanation (LorENZINI et al. 1993). The contribution of different gene pools to the population of Siena possibly produced a high heterozygosity. However, we cannot ex- clude that a higher level of genetic variability was proper to the original stock, native of the Latium Maremma. The Arezzo population did not experience any remarkable intro- duction of allochthonous genotypes from outside, or suffered any drastic demographic event which could reduce its variability. Thus, the heterozygosity of the original nucleus was retained. 20 RıtA LorRENZINI et al. As far as the morphometrical analysis of skulls and mandibles is concerned, size varia- tion was the main source of difference between samples, skulls of the Arezzo population being generally larger than those of the Siena population. Nevertheless, some differences in shape were also apparent, mostly when the ratios of measurements were used. Due to a different approach in gathering the morphometric measurements, the set of characters used in this investigation is not fully comparable with that of previous studies. In spite of this, some rough comparisons could be made with the data from other European popula- tions. Such comparisons revealed that the Arezzo roe deer have greater skull dimensions than those studied so far, whereas roe deer from Siena are more similar to the ones de- scribed in the literature (MEUNIER 1981; ZEIDA and KouBEK 1988; ZımA et al. 1989; FAnN- Dos and Reıc 1993, Fanpos 1994). Moreover, the level of cranial differentiation obtained in this study is higher than the ones reported in other investigations, where no evidence of shape differences was found among local populations (ZEJDA and KouBEK 1988; FAN- pos and Reıc 1993). Some divergence between samples was first suggested by the univariate comparisons of means, and became more evident from the discriminant analysis. This method, better than principal component analysis, allowed full separation of the two sets of skulls, when log. transformations of absolute data were used. Even when discriminant analysis was per- formed with the relative values, the residual phenotypic variation allowed a quite clear se- paration of the two populations. Only one skull per sample was ıidentified as belonging to the other a-priori subdivided group on the basis of the discriminant scores. The pattern of allele frequencies revealed that a genetic factor is involved in the diver- sification of the Siena and Arezzo populations due to their different historical back- grounds. However, when dealing with metrical traits, the environmental component is also of great importance. Beyond the complexity of the gene-environment interactions, it remains problematic to understand and to quantify which environmental factor affects the observed phenotypic variation. In our case, overall food resources and environmental sta- bility of the habitats are very similar for both the roe deer populations. On the other hand, the population densities within the wooded areas are considerably different in the two provinces. Density is one of the main factors in the dynamics and the ecology of po- pulations (RıcKLers 1979). Density proved to be a significant factor for the social organi- zation of deer populations, affecting behaviour and reproduction (AroLLonio et al. 1991; VıncEnt and BipeAU 1992). The emergence of the new ecological form of field roe deer has been interpreted as the result of the overcrowding of forest populations due to persis- tent habitat degradation, which was followed by movements of the animals towards agri- cultural landscapes (PıELowskı 1977, PıELowsKI and BrEsınskI 1982). High densities increase the interactions among individuals for food supplies, and be- tween males during the rutting period (KLEin and STRANDGAARD 1972; Kurt 1991). In the province of Siena the high density in forested areas could cause a stronger individual competition and, on average, poorer feeding conditions. The latter may affect growth, weight, body size and other morpho-physiological traits. In the Arezzo population this si- tuation may have occurred less intensively, due to the lower density in the wooded areas. There, roe deer have larger body sizes and longer skulls. Lovarı et al. (1991) reported that also lower jaw and metatarsal dimensions are greater in roe deer from Arezzo than in those from Siena. Moreover, MEUNIER (1981) showed that the growth of some cranial di- mensions can be influenced by changing conveniently the dietary regimes, lengths increas- ing more than widths. Significant differences in cranıal, mandibular and antler dimensions have been docu- mented between populations from two geographically separated regions in the former Czechoslovakia (ZEIDA and KoußEk 1988). However, a precise cause-effect relationship between variation of morphological traits and environmental effects could not be found. In contrast to our results, the Czechoslovakian roe deer which lives at high density had Genetic and morphometric variation in the Roe deer | DA larger body sizes and greater cranial dimensions than those living at lower densities. How- ever, the latter are supported by poorer quality food, so that density is not the principal limiting factor. Ackowledgements We wish to thank R. MAZzonI DELLA STELLA for his contribution in providing the samples of Siena and in tracing the history of the same population. We are also grateful to the local hunters and to the D.R.E.AM. team (C. LovaRı, V. MAZZARONE, P. PEDONE, and N. SIEMoNI) for their assistance in the field. Special thanks are addressed to G. HARTL, S. LOVARI, J. MARKOWSKI, F. SUCHENTRUNK, and S. MATTIOLI for their comments on an earlier version of the manuscript. The graphical help of S. SANTARELLI is grate- fully ackowledged. The German summary was graciously provided by CLAUDIA WEIss. Zusammenfassung Allozym- und kraniometrische Variabilität bei Rehen (Capreolus capreolus) aus Mittelitalien Rehe aus den Provinzen Siena und Arezzo, Mittelitalien, wurden auf genetische und morphologische Differenzierung untersucht. Bei insgesamt 162 Individuen aus sieben Stichprobengebieten (4 in Siena, 3 in Arezzo) wurde mittels vertikaler Polyacrylamidgel-Elektrophorese die Variabilität an 40 hypothe- tischen Strukturgenloci erfaßt. Sieben Loci waren für mindestens zwei Allele polymorph. Die durch- schnittliche Polymorphierate (P) betrug in Siena 16,2% (s= 1,4%) und in Arezzo 15,8% (s = 2,9%). Der durchschnittliche erwartete Heterozygotiegrad (H.) betrug in Siena 5,3% (s = 0,7%) und in Arezzo 4,0% (s= 0,2%). Sowohl die relative (Fsr=8%) als auch die absolute (D nach Neı 1972 = 0,007, s = 0,002; D nach Rogers = 0,084, s = 0,011) genetische Distanz zeigen eine deutliche Separierung der Stichproben aus Siena von jenen aus Arezzo, was auf die unterschiedliche Herkunft der Rehbestände zurückzuführen sein dürfte. Die bei insgesamt 48 Individuen gemessene morphometrische Variabilität in 28 Schädelmaßen wurde einer Hauptkomponentenanalyse und einer Diskriminanzanalyse unterwor- fen. Beide Methoden zeigten eine klare morphometrische Differenzierung zwischen den Individuen aus den jeweiligen Provinzen. Die Rehe aus Arezzo wiesen signifikant größere Schädelmaße auf. Obwohl auch genetische Faktoren eine Rolle spielen könnten, betrachten wir die höhere Individuendichte der Rehe in den Waldgebieten von Siena als Hauptursache für diesen Befund. References AITKEN, R. J. (1975): Cementum layers and tooth wear as criteria for ageing roe deer (Capreolus capreo- lus). J. Zool. (London) 175, 15-28. ANDERSON, T. W. (1971): An introduction to multivariate analysis. New York: J. Wiley and Son. APOLLONIO, M.; STRIGLIONI, F.; SCHENONE, L.; SPANO’, S. (1991): Different mating systems in fallow deer, possible consequences of high density on reproduction costs. Ongul&s/Ungulates 91. Toulouse, Ab- stracts, 5. APOLLONIO, M.; HARTL, G. B. (1993): Are biochemical-genetic variation and mating systems related ın large mammals? In: Ecological Genetics in Mammals. Ed. by G. B. HArTL and J. MARKOWSKI. Acta theriol. 38, Suppl. 2, 175-185. Baccus, R.; Ryman, N.; SMITH, M. H.; REUTERWALL, C.; CAMERON, D. (1983): Genetic variability and dif- ferentiation of large grazing mammals. J. Mammalogy 64, 109-120. CRUDELE, G. (1988): La fauna. In: Le foreste di Campigna-Lama nell’Appennino Tosco Romagnolo. Ed. by M. PADuLA and G. CRUDELE. Regione Emilia Romagna, Bologna, 325-401. DRIESCH, A. voN DEN (1976): A guide to the measurement of animal bones from archeological sites. Peabody Mus. Bull., Univ. Harvard Press, 35-57. FAnDos, P. (1994): Skull biometry of Spanish roe deer (Capreolus capreolus). Folia Zool. 43, 11-20. FANDos, P.; ORUETA, J. F. (1991): Morphological variations of the roe deer skull (Capreolus capreolus L.). Donana, Acta Vertebrata 18, 159-171. DD RıTA LoRENZINI et al. FANDos, P.; Reıc, S. (1993): Craniometric variability in two populations of roe deer (Capreolus capreo- lus) from Spain. J. Zool. (London) 231, 39-49. FARRIS, J. S. (1972): Estimating phylogenetic trees from distance matrices. Amer. Nat. 106, 645-668. FELSENSTEIN, J. (1979): Alternative methods of phylogenetic inference and their interrelationship. Syst. Zool. 29, 49-62. FELSENSTEIN, J. (1989): PHYLIP. Phylogenetic inference package. Version 3.2. University of Washing- ton, Seattle, Washington. FRUZINSKI, B.; LABUDZKI, L. (1982): Demographic processes in the forest roe deer population. Acta theriol. 27, 377-384. FRUZINSKI, B.; KALUZINSKT, J.; BAKSALARY, J. (1982): Weight and body measurements of forest and field roe deer. Acta theriol. 27, 479-488. GOTTSCHLICH, H. J. (1982): Vergleichende Untersuchungen über das Schädelwachstum bei Rot- und Rehwild. Beitr. Jagd- und Wildforschung 3, 81-92. Gourson, D. (1994): Allozyme variation in the butterfly, Maniola jurtina (Lepidoptera: Satyrinae L.): evidence for selection. Heredity 71, 386-393. GYLLENSTEIN, U.; RYMAN, N.; REUTERWALL, C.; DRATCH, P. (1983): Genetic differentiation in four Eu- ropean subspecies of red deer (Cervus elaphus L.). Heredity 51, 561-580. HANUS, V.; FISER, Z. (1979): Craniological measurements in the roebuck (Capreolus c. capreolus L.). Fo- lia Venat. 9, 15-32. HARRIS, H.;, Hopkinson, D. A. (1976): Handbook of enzyme electrophoresis in human genetics. Amster- dam: North Holland. HARTL, G. B.; REIMOSER, F. (1988): Biochemical variation in roe deer (Capreolus capreolus L.): are T- strategists among deer generally less variable than k-strategists? Heredity 60, 221-227. HARTL, G. B.; SCHLEGER, M.; SLOWAK, M. (1986): Genetic variability in fallow deer, Dama dama L. Anim. Genet. 17, 335-341. HARTL, G. B.; REIMOSER, F.; WILLING, R.; KÖLLER, J. (1991): Genetic variability and differentiation in roe deer (Capreolus capreolus L.) of Central Europe. Genet. Sel. Evol. 23, 281-299. HARTL, G. B.; WiLLIngG, R.; LANG, G.; KLEIN, F.; KÖLLER, J. (1990): Genetic variability and differentiation in red deer (Cervus elaphus L.) of Central Europe. Genet. Sel. Evol. 22, 289-306. HARTL, G. B.; MARKOV, G.; RUBIN, A.; FINDO, S.; LAnG, G.; WILLING, R. (1993): Allozyme diversity within and among populations of three ungulate species (Cervus elaphus, Capreolus capreolus, Sus scrofa) of Southeastern and Central Europe. Z. Säugetierkunde 58, 352-361. HERZOog, S. (1988): Polymorphism and genetic control of erythrocyte 6-phosphogluconate dehydrogen- ase in the genus Cervus. Anim. Genet. 19, 291-294. Hewiıson, A.J. M. (1995): Isozyme variation in roe deer in relation to their population history in Bri- tain. J. Zool. (London) 235, 279-288. KLEIN, D. R.; STRANDGAARD, H. (1972): Factors affecting growth and body size of roe deer. J. Wild. Man- age. 36, 64-79. Kurt, F. (1991): Das Reh in der Kulturlandschaft - Sozialverhalten und Ökologie eines Anpassers. Hamburg, Berlin: Paul Parey. Kurt, F.; HARTL, G. B.; VöLK, F. (1993): Breeding strategies and genetic variation in European roe deer Capreolus capreolus populations. In: Ecological Genetics in Mammals. Ed. by G. B. HARTL and J. MARKOWSKI. Acta theriol. 38, Suppl. 2, 187-194. LEHMANN, E. VON; SÄGESSER, H. (1986): Capreolus capreolus Linnaeus, 1758. In: Handbuch der Säuge- tiere Europas. Vol. 2/II. Ed. by J. NIETHAMMER and F. KrApp. Wiesbaden, Aula-Verlag. Pp. 233-268. LORENZINI, R.; PATALANO, M.; APOLLONIO, M.; MAZZARONE, V. (1993): Genetic variability of roe deer Capreolus capreolus in Italy: electrophoretic survey on populations of different origin. In: Ecologi- cal Genetics in Mammals. Ed. by G. B. HArTL and J. MARKOWSKI. Acta theriol. 38, Suppl. 2, 141- 11 LovaRı, C.; MATTIOLI, L.; MAZZARONE, V.; SIEMONI, N.; PEDONE, P. (1991): Biometric statistics of two po- pulations of roe deer (Capreolus capreolus L.) in Tuscany, Italy. In: Ongul&s/Ungulates 91. Ed. by F. Spitz, G. JANEAU, G. GONZALEZ, and S. AULAGNIER. Toulouse: S.F.E.P.M. - I.R.G.M., 133-136. MAJEWSKA, B.; PIELOwSsKI, Z.; LABUDZKI, L. (1982): The level of some energy metabolism indices in for- est and field populations of roe deer. Acta theriol. 27, 471-477. MARKOV, G.; DANILKIN, A.; HARTL, G. B. (1992): Lack of biochemical-genetic variation in native sika deer (Cervus nippon hortularum) from the far east of the Asian continent. Z. Säugetierkunde 57, 118-119. MARKOWwsSKI, J. (1993): Fluctuating asymmetry as an indicator for differentiation among roe deer Ca- Genetic and morphometric variation in the Roe deer 23 preolus capreolus populations. In: Ecological Genetics in Mammals. Ed. by G.B. HArTL and J. MARKOWSKI. Acta theriol. 38, Suppl. 2, 19-31. MARKOWSKI, J.; MARKOWSKA, M. (1988): Non-metrical variation in three populations of roe deer. Acta theriol. 33, 519-536. MAYNARD SMITH, J. (1989): Evolutionary Genetics. Oxford: Univ. Oxford Press. MATTIOL1, S. (1994): Le reintroduzioni di cervidi in Italia, con particolare riferimento all’Appennino. In: I Convegno Barrasso: La fauna appenninica e la sua conservazione. Ed. by M. InGtisA. Pp. 19-25. MAZZONI DELLA STELLA, R. (1991): Indagine sulla presenza del capriolo e del daino nella provincia di Siena. Amministrazione Provinciale di Siena. Pp. 1-65. MEUNIER, K. (1981): Analyse der Größen- und Proportionsverhältnisse. In: Über Rehe in einem stei- rischen Gebirgsrevier. 3. Aufl. Ed. by A. V. BAYERN and J. V. BAyErn. Muenchen: Bayerischer Land- wirtschaftsverlag. Pp. 141-146. MORRISON, D. F. (1976): Multivariate statistical methods. London: McGraw-Hill. MÖRSCH, G.; LEIBENGUTH, F. (1994): DNA fingerprinting in roe deer using the digoxigenate probe (GTG) 3. Anim. Genet. 25, 25-30. MURPHY, R. W.; Sıtes, J. W.; DonALD, G. B.; HAUFLER, C. H. (1990): Protein I: Isozyme Electrophoresis. In: Molecular Systematics. Ed. by D. M. Hırrıs and C. Morıtz. Sunderland, Massachusetts, USA: Sinauer Ass. Neı, M. (1972): Genetic distance between populations. Am. Nat. 106, 283-292. Neı, M. (1977): F-statistics and analysis of gene diversity in subdivided populations. Ann. Human Gen- et. 41, 225-233. NEUHAUS, A. H.; SCHAICH, K. (1985): Das Rehwild. Hamburg, Berlin: Paul Parey. Nie, N. H.; Hurı, C. H.; JENKINS, J. G.; STEINBRENNER, K.; BENT, D. H. (1975): Statistical package for the social sciences. New York: Mc Graw-Hill. PEMBERTON, J. M.; SMITH, R.H. (1985): Lack of biochemical polymorphism in British fallow deer. Her- edity 55, 199-207. PıIELowsK1, Z. (1977): Das Feldreh. Wild der Zukunft in der Agrarlandschaft. Beitr. Jagd- und Wild- forsch. 10, 193-200. PIELOWSKI, Z.; BRESINSKI, W. (1982): Population characteristics of roe deer inhabiting a small forest. Acta theriol. 27, 409-425. RANDI, E.; APoLLoNIo, M. (1988): Low biochemical variability in European fallow deer (Dama dama L.): natural bottlenecks and the effect of domestication. Heredity 61, 405410. RIcCKLEFS, R. E. (1979): Ecology. 2nd ed. Middlesex, UK: Nelson. Rep, K.H. (1985): Genetic variability in Norwegian semi-domestic reindeer (Rangifer tarandus L.). Hereditas 102, 177-184. Reep, K.H. (1986): Genetic variability in Norwegian wild reindeer (Rangifer tarandus L.). Hereditas 104, 293-320. Ryman, N.; REUTERWALL, C.; NIGREN, K.; NIGREN, T. (1980): Genetic variation and differentiation in Scandinavian moose (Alces alces): are large mammals monomorphic? Evolution 34, 1037-1050. SHAw, C. R.; PrAsAD, R. (1970): Starch gel electrophoresis of enzymes. A compilation of recipes. Bioch. Genet. 4, 297-320. SHEFFIELD, S. R.; MORGAN, R. P. II; FELDHAMER, G. A.; HARMAN, D.M. (1985): Genetic variation in white-tailed deer (Odocoileus virginianus) populations in Western Maryland. J. Mammalogy 66, 243-255. 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 populations. Science 236, 787- VOR, SNEATH, P. H.; SoKAL, R. R. (1973): Numerical taxonomy. San Francisco: W. H. Freeman and Co. SOKAL, R. R.; RoHLEF, F. J. (1981): Biometry. San Francisco: W. H. Freeman and Co. SWOFFORD, D. L.; SELANDER, R. K. (1989): Biosys-1. A computer program for the analysis of allelic var- iation in population genetics and biochemical systematics. Release 1.7. Illinois Nat. Hist. Surv., Champaign. VINCENT, J. P.; BıpEAU, P. (1992): Influence of density on spatial and social organization of forest roe deer (Capreolus capreolus L.) In: Ongul&s/Ungulates 91. Ed. by F. Spitz, G. JANEAU, G. GONZALEZ, and S. AULAGNIER. Toulouse: S.F.E.P.M. - I.R.G.M., 267-269. WEHNER, J.; MÜLLER, H. P.; KIERDORF, H. (1991): Untersuchungen zur genetischen Situation ausgewähl- ter rheinischer Rehwild-Populationen unter besonderer Berücksichtigung isolationensbedingter Veränderungen. Z. Jagdwiss. 37, 40-48. 24 RıTA LorRENZINI et al. WILLIAMS, B. K.; Titus, K.; Hınes, J. E. (1990): Stability and bias of classification rates in biological ap- plications of discriminant analysis. J. Wild. Manag. 54, 331-341. WORKMAN, P. L.; NISWANDER, J. D. (1970): Population studies on southwestern Indian tribes. II. Local ge- netic differentiation in the Papago. Amer. J. Human Genet. 22, 24-49. WRIGHT, S. (1931): Evolution in Mendelian populations. Genetics 16. 97-159. WRIGHT, S. (1951): The genetical structure of populations. Ann. Eugenet. 15, 323-354. WRIGHT, S. (1978): Evolution and the genetics of populations. Vol. 4. Variability within and among nat- ural populations. Chicago: Univ. Chicago Press. ZEJDA, J., KOUBEK, P. (1988): On the geographical variability of roebucks (Capreolus capreolus). Folia Zool. 37, 219-229. ZIMA, J.; LIBOSVARSKY, J.; BAUEROVA, Z., KOUBEK, P;; ZEIDA, J. (1989): Comparison of metric and non-me- tric morphological distances between four populations of roe-deer (Capreolus capreolus). Folia Zool. 38, 45-58. Zıma, J. (1989): Non-metrical variability in the skull of roe deer (Capreolus capreolus). Folia Zool. 38, 119-137. Authors’ addresses: RıTA LORENZINI, MARIANNA PATALANO, Istituto Zooprofilattico Sperimentale dell’Abruzzo e del Molise “G. Caporale”, Campo Boario, I-64100 Teramo, Italy; LucA MATTIOL, D.R.E.AM. Settore Fauna, Via Roma 172, 1-52014 Poppi, Arezzo, Italy; Marco Rusrionı, Dipartimento di Scienze della Terra, Universitä di Firenze, Via La Pira 4, I-50121 Firenze, Italy. Z. Säugetierkunde 61 (1996) 25-38 DS FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Social behaviour, mating system and testes size in Cuis (Galea musteloides) By ESTHER SCHWARZ-WEIG and N. SACHSER Lehrstuhl für Tierphysiologie, Universität Bayreuth, Bayreuth, Deutschland Receipt of Ms. 02. 06. 1995 Acceptance of Ms. 02. 11. 1995 Abstract Social behaviour, mating system and testes size were investigated in cuis (Galea musteloides) in order to elucidate the social mechanisms in an obviously promiscuous rodent. Therefore, quantitative behaviour- al measures were recorded with a total of more than 200 h observation time in 4 groups of animals (1—4 adult males; 3-5 adult females in 12-25 m? enclosures). Testes and body masses were determined in 21 males. The most conspicuous feature in all groups of cuis was the daily rhythm of huddling. Short-last- ing phases of high activity alternated several times daily with phases of inactivity during which all or nearly all members huddled together with close bodily contact. The mating system was promiscuous. In none of the 8 mating chases did the alpha male succeed in preventing the lower ranking males from co- pulation. Promiscuous mating was achieved by the females’ behaviour, which attracted the attention of all the males and made it impossible for a single male to monopolize her. The males’ relative testes size was extremely high, which thus strongly supports the supposition that a promiscuous mating system is typical for cuis. Surprising for such a mating system were the high levels of intermale aggression which were even found when no female was receptive, and although clear dominance relationships existed be- tween the males. The females were also organized in linear dominance hierarchies. In contrast to the males, however, they frequently directed socio-positive behaviour towards conspecifics of the same sex. Introduction As a general guideline, mating systems can be divided into four different categories: monogamy, polygyny, polyandry and promiscuity (Kress and Davıs 1993). In mammals polygynous systems predominate, whereas monogamy and above all polyandry are rare (Wırson 1975; EMLEN and OrınG 1977; CLuTTon-Brock 1989; Arcock 1993; KREBS and Davis 1993). These three forms, however, have been studied extensively (see for example KLEIMAn 1977; HENDRICHS 1978; GOLDIZEN 1987; CLUTTON-BRocK 1989; VAN SCHAIK and DunBaAr 1990) and there is good evidence that variations in mating systems can be attrib- uted to variation in resource distribution and paternal care (EMLEN and OrınG 1977; CLUT- ToN-BrRock 1989; Arcock 1993; Kress and Davıs 1993). Also the social mechanisms bringing forth these differences in mating behaviour — for example, territory formation, attaining and respecting dominance, building up social bondings, mate guarding — are well understood (cf. Wırson 1975; KuMMER 1984). Relatively little is known, however, about promiscuous systems, although they are widespread among mammalian species (KENAGY and TROMBULAK 1986). Especially in small mammals the social mechanisms inducing both male and female to mate several times with different individuals are largely unknown. Variation in mating systems is frequently related to variation in relative testes size (HARCoURT et al. 1981; Harvey and HArRcoURT 1984; KEnAGY and TROMBULAK 1986; 26 ESTHER SCHWARZ-WeEIG and N. SACHSER MO@LLER 1988). In species in which only one male mates (polygynous and monogamous species, respectively) low testes weights in relation to body weights are found, whereas males of promiscuous species are characterized by high relative testes weights. Supposing that testes size is positively correlated with volume of ejaculate, sperm counts and sperm motility (MoLLeEr 1988; HArvEY and May 1989), it is reasonable to assume that in promis- cuous species males do not compete for access to females via agonistic encounters, but that the competition is yielded to their semen, that is, sperm competition should arise. In- deed, in the promiscuous chimpanzee, for example, high relative testes weights and no ag- gressive competition for mating is found among males (Van LAwIcKk-GooDALL 1968; SHorT 1979). But while the knowledge about the relationship between behaviour and testes size is advanced in primates, relatively few detailed data are available in small mammals of other groups. The yellow-toothed cavy or cuis (Galea musteloides) is a small South-American rodent which occupies the open habitats of savannas and bushland. In man-made habitats, such as rock piles and brush fences, considerably high population densities are attained. Field observations on behaviour and the mating system are not available. The study of 5-8 males and 8-12 females in a 25 m” enclosure suggests, however, that a promiscuous mat- ing system exists: Although the alpha male frequently guards the female during estrous, he obviously is not successful in preventing other males from mating (Roop 1972). To extend our investigations in guinea pigs (e.g. SACHSER 1986, 1994; SACHSER et al. 1994) to a closely related non-domestic species we chose Galea musteloides, a member of the same subfamily Caviinae, which has already been studied under specific behavioural aspects (WEır 1971, 1973; Roop 1972; König 1985; KÜNKELE and HoEk 1989). When we compared the abdomens of intact males of both species, testes size appeared to be much greater in cuis than in guinea pigs, although they had a significantly lower body weight. These differences in morphological traits are obviously related to variation in behaviour: In colonies of guinea pigs a polygynous mating system is found (SACHSER 1986), whereas Roop’s (1972) observations point to a promiscuous mating system in cuis. We therefore investigated the social behaviour and mating system in groups of cuis and determined testes size quantitatively in order to elucidate social mechanisms in a small promiscuous non-primate species. Material and methods Subjects The cuis (Galea musteloides) used were either descendants of 4 pairs (4 males, 4 females) obtained from the Zürich Zoo (Switzerland) more than 10 years ago, or they were provided by Prof. Dr. H. HENDRICHS (Department of Behavioural Physiology, University of Bielefeld, Germany) in 1989 (1 male, 4 females) and 1990 (6 males, 11 females), respectively. All animals were marked individually using commercial fur dye (“Nyanzol D”, Belmar Inc., North Andover, USA). Animal care and procedures Four groups of cuis (group 1, 2, 3, 4) were studied; these groups had been established 2-12 months be- fore the investigations began. All individuals were sexually mature, i. e., the males had a minimum age of 3 months and the females had a minimum age of 4weeks (Roop and Weır 1970; Horr 1977). Groups 2 and 3 were kept in outdoor enclosures of about 15 m”; group 4 was kept in an indoor enclo- sure (size: 12 m; 12:12 L:D cycle, photoperiod 07:00 to 19:00 h; temperature: 20+2°C). Group 1 was housed in an outdoor enclosure (size: 15 m?) during the first 3 months of the study. The space available was subsequently enlarged by opening the door to an indoor enclosure (10 m). The indoor enclosures had stone flooring, which was covered with woodshavings and hay. The floor of the outside enclosures Social behaviour, mating system and testes size in Galea musteloides ZT was covered with earth and planted with small trees. Part of each outdoor enclosure was roofed, under which the floor was covered with straw and hay. All enclosures were structured with branches and root- stocks. Three to five nestboxes were placed to every group. During the cold season the outdoor enclo- sures were equipped with radiators. Commercial guinea pig diet, flaked oats, maize and water were available ad libitum. This diet was supplemented regularly with fruits. The enclosures were cleaned at intervall of 3 to 6 weeks. In the 4 groups studied, 1-4 adult males and 3-5 adult females were present. Social behaviours were recorded quantitatively over periods of 1-20 consecutive weeks with a total of 232 h observation time (group 1: 177 h; group 2: 16 h; group 3: 29 h; group 4: 10 h; focal group sampling, continuous recording). The identity of each interaction partner was noted as well as who performed and who received the be- haviour displayed. The period of observation was generally 2 h, either between 08:30 h and 12:00h or between 15:00h and 19:00 h. In addition, in group 1 the rhythm of “huddling” was determined during the 16 h-light-phase of a summer day: On 4 consecutive days we recorded for 4h at minute intervals the number of animals who huddled together with bodily contact. Data recording was conducted by hand, using prepared checksheets. All cuis were weighed regularly at 14 day intervals and marked if necessary. Behavioural measures The categorization and nomenclature of behavioural patterns followed Roop (1972) and SCHWARZ (1991). The behavioural patterns considered were: huddling, sitting in bodily contact, approach; agonis- tic behaviours: retreat, head-thrust, bite, attack-lunge, chase, pursue, curved body posture, turn-around; courtship and sexual behaviours: naso-anal licking, chin-rump following, mounting, intromission, ejacu- lation, taking the lordosis position; vocalization: churr. A description of these behavioural measures is given below where required. Testes and body masses Testes and body masses were determined in 21 healthy, mature males aged from 5 to 27 months which had lived in groups of 1-4 males and 1-2 females. Body masses were taken immediately before the males were sacrificed. The testes were subsequently rapidly cleaned from fat and adherent tissue, and for each individual the mass of both testes (excluding epididymides) was determined to the nearest 0.01 g. Relative testes size was calculated according to KEnAGy and TROMBULAK (1986), that is: relative es size = observed mass of testes / predicted mass of testes; predicted mass of testes = 0.035* body mass”. Statistics Testes and body masses are given as means +SD. Behavioural data are presented as medians. Results Huddling and rhythms of activity In all four groups the cuis showed distinct ultradian rhythms of activity: Short-lasting phases of high activity, which were often characterized by frequent agonistic interactions alternated several times daily with phases of inactivity during which all or nearly all mem- bers of a group huddled together with close bodily contact (Figs. 1, 2). Even two cuis which had just been engaged in an escalated fight could be seen in this contactual resting posture only a few minutes later. A quantitative analysis of 89 huddling events in group 1 revealed that it took on average 13 min until all members of the group came together to huddle. Huddling itself lasted for a mean period of 46 min and it was ended on average within 10 min. Surprisingly, huddling was displayed at a wide range of temperatures, from lower than 0°C in winter to about 30 °C in summer. 28 ESTHER SCHWARZ-WeEIG and N. SACHSER Fig. 1. Group of huddling cuis. number of animals huddling time Fig. 2. Diurnal rhythm of huddling in a group of 8 adult cuis (group 1: 4 males, 4 females) living in an outdoor enclosure. The data were recorded in July during the whole phase of daylight. Agonistic behaviours and dominance relationships A complete range of agonistic behaviours of differing intensities was found. At the lowest intensity one individual approached another, which in turn immediately retreated. Threat displays and fights were hereby avoided. During more escalated forms of agonistic en- counters overt aggressive behaviours were performed, that is head-thrusting, biting, at- tack-lunges, chasing and pursuing occurred. A still higher escalation was the display of Social behaviour, mating system and testes size in Galea musteloides 29 curved body posture by one or two individuals simultaneously which might change into an escalated fight. At this highest agonistic intensity both individuals jumped at each other, bit and turned around. Dominance relationships between the individuals could be derived from the direction of the aggressive behaviours displayed as well as from the out- come of agonistic encounters. Although the lower ranking animals generally retreated im- mediately when a more dominant approached, the dominant frequently chased, pursued and bit them. As a consequence, most subordinate males and some subordinate females had scratches on their backs (cf. Roop 1972). Dominance relationships between males In all four groups the males established clear dominance relationships. High ranking posi- tions were maintained by frequently directing overt aggressive behaviours towards sub- ordinates. Changes in dominance positions occurred, as is illustrated exemplarily for 5 > g—d-—-0-—0 9 9a 8 a 7 Fig. 3. Distribution of offensive aggressive behaviours (sum of head-thrusting, biting, attack-lunges, chasing, pursuing) among the adult males in group 1. The arrows point to recipients; the width of the arrows is proportional to frequencies. The 20-week observation time (177 h) is divided into 5 phases: A: 07.08.-11.08. (10 h); B: 13.08.-15.08. (6 h); C: 16.08.-17.08. (3 h); D: 20.08.-14.09. (38 h); E: 17.09.-21.12. (120.h). 30 ESTHER SCHWARZ-WeEIG and N. SACHSER group 1. At the beginning of the observation the males had organized themselves into a linear dominance hierarchy (Fig. 3A): M 11 was the highest, M 12 the 2nd, M 14 the 3rd and M 13 the lowest ranking male. (From previous observations it was known, that this rank order had been stable for at least 4 weeks before the start of the detailed observa- tion.) M 11 held the highest ranking position throughout the whole 20 weeks of observa- tion time. During the 2nd week changes in lower ranking dominance positions occurred. First - after an escalated fight - M 14 and M 13 reversed their positions (Fig. 3B). There- after M 13 and M 14 attained dominance over M 12 (Fig. 3C) and frequently chased and bit this male. Although no woundings were detected, M 12 died a few days later. How- ever, it could not be decided whether death was the consequence of the changed domi- nance relationships, or whether M 12 was defeated because he fell ill and died from an undiagnosed disease. During the following 4 weeks the remaining 3 males organized themselves into a linear dominance hierarchy (Fig. 3D). After the death of the lowest ranking male (M 14) a clear and stable dominance relationship persisted between the alpha and beta males during the following 14 weeks (Fig. 3E). In two groups a change in the alpha position occurred. Surprisingly, in both cases the subordinates had lower body masses than the dominants when they defeated their oppo- nents (407 vs. 546 g; 317 vs. 364 g). In general the alpha positions, however, were filled by the heaviest males. Dominance relationships between females In three out of four groups (group 2, 3, 4) the females organized themselves in a similar way as the males: They formed linear dominance hierarchies including all adult females. This finding is shown exemplarily for group 3 (Fig. 4). F31 took the alpha position and her daughters F 33, F32 and F 34 the 2nd, 3rd and 4th respective hierarchal positions. In group 1, female F 15 and her three daughters were also organized in a linear hierarchy, with F15 at the alpha position throughout the whole of the 20 week observation time. Three older unrelated females, however, which took the lowest ranking positions during 15 out of 20 weeks, hardly ever directed aggressive behaviours towards each other. So in this group it was not possible to arrange all females in a linear order. The adult females’ rank positions were largely independent of their body weights and ages. A juvenile female in group 1, for example, attained dominance over two fully grown females when she was not older than 6 weeks and had a distinct lower body weight than her opponents (170 g vs. 325 g and 264 g, respectively). Aggression between the sexes In all groups offensive aggressive behaviours were not only directed towards individuals of the same sex but also towards cuis of the opposite sex: Males and females could be in- volved both in agonistic interactions of low intensities as well as in threat displays and es- calated fights. The frequencies of aggressive behaviours, however, were much lower between the sexes than within each sex. Furthermore, aggressive behaviours among males occurred on average more than twice as often as among females (Fig. 5). In contrast to the intrasexual aggression, aggressive behaviours between the sexes were often bidirectional: Those males or females, respectively, which frequently displayed aggressive behaviours towards certain cuis of the opposite sex, were frequently attacked themselves by these same individuals. Moreover some males and females were never seen to direct aggressive behaviours towards each other. Thus it would not seem to be reason- able to calculate dominance relationships between the sexes. Social behaviour, mating system and testes size in Galea musteloides 31 Fig. 4. Distribution of offensive aggressive behaviours (sum of head-thrusting, biting, attack-lunges, chasing, pursuing) among the adult females in group 3 during a 4 week observation time (29 h). The ar- rows point to recipients; the width of the arrows is proportional to frequencies. Socio-positive behaviours The cuis were not only involved in agonistic interactions, but socio-positive behaviours also occurred. Frequently one individual approached another and sat down in close bodily contact. Such neighbouring pairs often synchronized their feeding. In contrast to hud- dling, sitting in bodily contact occurred during phases of activity. It was frequently re- corded as occurring between the females as well as between the sexes. However, it rarely occurred among the males (Fig. 5). Mating system During the present study an estrous could be observed 8 times (in 3 different groups, in 7 different females; cf. Tab. 1). In all cases mating behaviour was promiscuous. Not only the dominant male mounted, intromitted and ejaculated but also the second and third-rank- ing males. It was true that the dominant male tried to guard the receptive female, but he never succeeded. Promiscuous mating was achieved by the behaviour displayed by the fe- 32 ESTHER SCHWARZ-WeEIG and N. SACHSER frequencies of aggressive behaviour/h among dd 9% R SCH X 7 $, frequencies of socio-positive behaviour/h IX T7? (2 [7 25 0$ & 0 NN] among dd 22 Fig. 5. Frequencies of aggressive behaviours (A) and socio-positive behaviours (B) (sitting in bodily contact during non-huddling times) among adult cuis (group 1) during a 19-week observation time (167 h). Circles and triangles give the frequencies of these behaviours for each male-male (N = 6), male- female (N = 24) and female-female dyad (N = 18); columns represent medians. A statistical evaluation is not feasible because the samples consist of dependent and independent data. male: She unexpectedly started to race around, stopped abruptly and started racing again. Thereby she often changed direction. By this rather conspicuous behaviour she attracted the attention of all the males, and these at once engaged in a mating chase, i.e., they fol- Table 1. Mating chases number of males males present group female DD WN DD WW % x PPRDNDNWN DD W 1 1 1 1 3 2 1 1 DD m * Estrus was not observed from the beginning of the mating chase number of copulating lowed the running female. This procedure was accompanied by extremely high lev- els of aggression between the males. Especially the alpha male directed ag- gressive behaviour towards subordinate males. The males omitted a typical vocali- zation (the “churr”, cf. Roop 1972) which only occurred around estrous. From time to time the female stopped abruptly, took the lordosis position, and copulated with that male which was just directly behind her. In none of the 6 cases which were re- corded from the beginning was the mat- ing period terminated before the female had copulated with all males present (Tab. 1), irrespective of whether it was a spontaneous or a post-partum estrus. Social behaviour, mating system and testes size in Galea musteloides 33 Testes mass The mass of both testes of the males (excluding epididymides) amounted to 728 g+1.5lg (n=21). Their body weights were 402 g+58g (n= 21) on average. The re- lative testes size, calculated according to the formula of KEnAGy and TROMBULAK (1986), amounted to 2.77 +0.44. This value can be directly compared with the relative testes sizes of other mammals. Figure 6 shows that the relative testes size of Galea musteloides was much higher than that which is found in species in which only a single male copulates with each female. In contrast, these values were among the highest ever recorded in a ter- restrial mammalian species with a known promiscuous mating system. A B C © SR EREEEREE > xx \7 II relative testes size ET EEE, 6% RR single male multi male Galea musteloides species Fig. 6. Relative testes masses (according to KEnAaGY and TROMBULAK 1986; for details cf. methods) of mammals. A: species with single-male mating systems (monogamous and polygynous species); B: species with multi-male mating systems (promiscuous species); C: cuis (Galea musteloides). Data in A and B refer to KenAGy and TROMBULAK (1986). Species listed in their publication were only consid- ered here when their mating systems were known and sample size for testes mass was >1. In A and B each circle and triangle, respectively, represents the mean relative testes mass of a given species, col- umns give mean values for single- (N = 18) and multi-male species (N = 21), respectively. In C triangles represent relative testes masses of individual male cuis (n = 21); the column represents the mean value for this species. 34 ESTHER SCHWARZ-WEIG and N. SACHSER Discussion The most conspicuous feature in all groups of cuis was the daily rhythm of huddling (cf. also Roop 1972). Short-lasting phases of high activity, which were often characterized by frequent agonistic interactions, alternated several times daily with phases of inactivity dur- ing which all or nearly all members of a group huddled together with close bodily contact. Huddling was the behavioural pattern shown most of the time by the cuis (SCHWARZ 1991), its behavioural significance, however, is not yet clear. There are three likely func- tions: (1) It might have thermoregulatory benefits, as demonstrated in muskrats and degus (CAnars et al. 1989; Bazın and MACARTHUR 1992). Surprisingly, cuis perform huddling not only at cold temperatures, but also at temperatures of up to 30°C, which points against thermoregulation (CONTRERAS 1984). (2) A further function of huddling might be to synchronize the behavioural activities of the group members. The risk of predation might be hereby decreased. Indeed, the cuis were frequently seen feeding directly before and after huddling, and in the case of the closely related Cavia aperea it is well estab- lished that they synchronize their feeding in the wild (Roop 1972; Cassını 1991). (3) A role in the development of an odour typical of the group might also be possible. Subman- dibular glands are present in both sexes (see photograph in WEır 1971). The secretions are released on the surface of the skin. During huddling the odour can be spread through- out the anımals of a group. Hereby group cohesion might be promoted. This mechanism would also explain the extremely aggressive response of group residents to unfamiliar in- truders (Roop 1972). The behavioural data presented here point to a promiscuous mating system in Galea musteloides. In none of the 8 mating chases observed did the alpha male succeed in pre- venting the lower ranking males from copulating. Thus, throughout the whole observation time of this investigation both males and females mated several times with different indi- viduals. From this aspect our data confirm Roop’s (1972) observations. However, we do not agree that during the mating chase the female is driven around by the males. Accord- ing to our data the female, in contrast, played the active role: She always initiated the mating chase; it was her conspicuous behaviour that attracted all the males’ attention and made it impossible for a single male to monopolize her. We therefore state that the pro- miscuous mating was primarily due to the females’ behaviour. What might be the adaptive significance of this phenomenon? A high benefit for the female should be expected, since her behaviour results in an optically and acoustically rather conspicuous mating chase, which might attract not only further male conspecifics but also predators. It is generally assumed that sexual selection favours mate choice in fe- males. The more selective the female’s behaviour is, the better her chance of finding the best partner (TRIvErS 1985; Arcock 1993; KrEBs and Davıs 1993). In our and Roop’s (1972) study, however, the receptive females did not behave in a selective manner at all. In contrast, they attracted the attention of all males and showed no signs of refusing to mate with specific males. Apparently a selection pressure has acted on the females of this species to copulate with as many males as possible. Thus, the number and quality of sperm competing to fertilize the ovae might be increased. To our knowledge it has not yet been determined in mammals whether promiscuous mating will increase a female’s repro- ductive success. Adders, however, which copulate with more than a single male, produce more live offspring than females with single matings (MaDsen et al. 1992). At present we are investigating whether these findings are also true in the case of Galea musteloides. From laboratory studies it cannot be reliably derived which mating system a given spe- cies will display in the wild. Our data on testes size, however, strongly support that a pro- miscuous mating system is typical for cuis. The relative testes size was not only much higher than in species with a polygynous mating system, but it was one of the highest values ever recorded for a terrestrial mammal with a promiscuous mating system (cf. Social behaviour, mating system and testes size in Galea musteloides 35 KeEnaAGy and TROMBULAK 1986). The comparison with the closely related guinea pig (which is amember of the same sub-family of the Caviinae) further supports that a functional re- lationship exists between testes size and mating system in cuis. Guinea pigs living in colo- nies do not establish a promiscuous mating system but a polygynous one (JacoBs 1976; SACHSER 1986): The highest ranking males are able to monopolize “their” receptive fe- males because female guinea pigs behave more passively during estrous than female cuis. Furthermore, the highest ranking males respect each other’s ownership. Body weights are much higher in guinea pigs than in cuis (about 1000 g vs. 400 g). The mass of both testes in the males is, however, significantly higher in cuis (7.28 g vs. 4.50 g; cf. SACHSER 1984). In addition Horr (1977) reported in two individual male cuis 737x 10° and 970 x 10° sperm/ ml ejaculate, respectively, whereas in guinea pigs a distinctly lower concentration was found (42 x 10° sperm/ml ejaculate; FrEunD 1969). Thus, the Caviinae obviously provide a low-level taxon with diverging mating systems and corresponding functional variations in testes size. In those mammalian species studied so far with promiscuous mating systems a high compatibility between the males was described even during estrous, i.e., levels of aggres- sion were low. Indeed, males should not waste time and energy in agonistic encounters or even risk the danger of being injured when reproductive success is decided by “sperm competition” (SHORT 1979). Surprisingly, in cuis we recorded extremely high levels of aggression between the males around estrous. Especially the alpha male directed aggressive behaviour towards subordinate males. These observations agree with Roop’s (1972) findings. One might ar- gue that these high levels of aggression resulted from artificial housing. An alternative ex- planation, however, would seem to be more likely. In the literature there is not much information about which male will be the father of the offspring in a promiscuous mating system. There is, however, good evidence that go- nadal activity might be rank-dependent in dominance structured social systems, especially when high ranking positions are established and maintained by overt aggression (e.g. CHRISTIAN 1980; KEVERNE 1983; SACHSER and PRÖVE 1986; IzArp 1990). Thus, chronic so- cial subordination is frequently associated with low testosterone titers and hereby the amount and quality of sperm might be impaired (see, for example, von Horst 1969). In this case a male would benefit by suppressing male conspecifics through aggressive behav- iour, even if he is not able to prevent them from mating. This would explain the high fre- quencies of overt aggressive behaviour even when no female was receptive. Furthermore, this makes it comprehensible why dominants frequently chased, pursued and bit lower- ranking males, although clear dominance relationships were already established. In accordance with Roop (1972) we found linear dominance hierarchies among the females of Galea musteloides, a trait that seems to be common in the Caviinae (Galea spixüi, Kerodon ruprestris (LACHER 1981), Cavia aperea (Roop 1972), Cavia aperea f. por- cellus (THyEn and HENnDRrIcHS 1990). Compared to the females of Kerodon rupestris (LACHER 1981), female cuis were generally competitive. It is, however, not clear what the benefit of a high hierarchical position might be. LACHER (1981) supposed that in Kerodon reproductive success is correlated with dominance rank. Quantitative data concerning this aspect are, however, missing in all wild species of the Caviinae. In colonies of the domes- tic guinea pig, high ranking females do indeed have higher reproductive success than the lower-ranking ones (SACHSER 1984). Surprisingly, there were not only aggressive encoun- ters found among the females of Galea, but they also directed socio-positive behaviour towards each other. Furthermore, communal nursing usually occurs in this species (KÜnkELE and Hork 1989). It remains to be studied which factors trigger the permanent change between conflict and cooperation in female cuis. 36 ESTHER SCHWARZ-WEIG and N. SACHSER Acknowledgements The research was supported by a grant from the Deutsche Forschungsgemeinschaft. We wish to thank Prof. Dr. DIETRICH von Horst for stimulating discussions throughout the study and critical comments on the manuscript. We should also like to thank HAZEL SCHMIDT for correcting the English. Zusammenfassung Sozialverhalten, Paarungssystem und Hodengewichte bei Wieselmeerschweinchen (Galea musteloides) In mehr als 200 Beobachtungsstunden wurde das Verhalten von Wieselmeerschweinchen, die in vier Gruppen lebten, quantitativ erfaßt (Gruppenzusammensetzungen: 1-4 Männchen, 3-5 Weibchen; Ge- hegegrößen: 12-25 m’). Zusätzlich wurden von 21 Männchen die Körper- und Hodengewichte be- stimmt. In allen Gruppen zeigten die Tiere einen auffälligen ultradianen Aktivitätsrhythmus. Phasen hoher Aktivität wechselten sich täglich mehrfach mit ausgeprägten Ruhezeiten ab, in denen alle Indi- viduen einer Gruppe in engem Körperkontakt bei- und übereinander lagen. Das Paarungssystem war promiskuitiv. Während keiner der acht beobachteten Paarungsjagden gelang es dem dominanten Männchen, ihm unterlegene Artgenossen an der Kopulation zu hindern. Hervorgerufen wurde die Promiskuität durch ein auffälliges Verhalten der Weibchen: Während des Östrus lenkten sie die Auf- merksamkeit aller Männchen auf sich und machten es unmöglich, von einem einzigen Männchen monopolisiert zu werden. Das relative Hodengewicht der Wieselmeerschweinchen war im Vergleich zu anderen Säugetierarten extrem hoch. Auch dieser Befund unterstützt die Annahme eines promis- kuitiven Paarungssystems bei dieser Spezies. Überraschend für ein solches Paarungssystem war je- doch das hohe Maß an aggressivem Verhalten zwischen den Männchen, das auch auftrat, wenn kein Weibchen östrisch war, und obwohl klare Dominanzbeziehungen zwischen den Männchen existierten. Die Weibchen organisierten sich ebenfalls in einer linearen Dominanzhierarchie. Im Gegensatz zu den Männchen richteten sie jedoch häufig sozio-positives Verhalten gegen gleichgeschlechtliche Art- genossen. References ALCOock, J. (1993): Animal Behavior. 5th Ed. Sunderland, Massachusetts: Sinaur Associates. BAZin, R. C.; MACARTHUR, R. A. (1992): Thermal benefits of huddling in the muskrat (Ondatra zibethi- cus). J. Mammalogy 73, 559-564. CANALS, M.; ROSENMANN, M.; BozinovIic, F. (1989): Energetics and geometry of huddling in small mam- mals. J. Theor. Biol. 141, 181-189. Cassını, M.H. (1991): Foraging under predation risk in the wild guinea pig Cavia aperea. Oikos 62, 20- 24. CHRISTIAN, J. J. (1980): Endocrine factors in population regulation. In: Biosocial mechanisms of popula- tion regulation. Ed. by M.N. CoHEn, R.S. MaArpass, and M. G. Krein. New Haven, London: Yale Univ. Press. Pp. 55-115. CLUTTON-Bkock, T. H. (1989): Mammalian mating systems. Proc. Roy. Soc. Lond. (B) 236, 339-372. CONTRERAS, L. C. (1984): Bioenergetics of huddling: Test of a psycho-physiological hypothesis. J. Mam- malogy 65, 256-262. EMLEN, S. T.; OrınG, L. W. (1977): Ecology, sexual selection, and the evolution of mating systems. Science 197, 215-223. FREUND, M. (1969): Interrelationship among the characteristics of guinea pig semen collected by elec- tro-ejaculation. J. Reprod. Fert. 19, 393-403. GOLDIZEN, A. W. (1987): Facultative polyandry and the role of infant-carrying in wild saddle-back ta- marins (Sanguinus fuscicollis). Behav. Ecol. Sociobiol. 20, 99-109. HARCOURT, A. H.; HARVvEY, P. H.; Larson, S. G.; SHORT, R. V. (1981): Testis weight, body weight and breeding systems in primates. Nature 293, 55-57. HARVEYy, P.H.; HARCOURT, A.H. (1984): Sperm competition, testes size, and breeding systems in pri- mates. In: Sperm competition and the evolution of animal mating systems. Ed. by R. L. SMITH. Or- lando: Academic Press. Pp. 589-600. | | | Social behaviour, mating system and testes size in Galea musteloides Sl HARVEY, P. H.; MAy, R.M. (1989): Out for the sperm count. Nature 337, 508-509. HENDRICHS, H. (1978): Die soziale Organisation von Säugetierpopulationen. Säugetierkdl. Mitt. 26, 81- 116. Hour, W. V. (1977): The postnatal development of the testis in the cuis (Galea musteloides). Lab. Anim. 11, 87-91. IzARD, M. K. (1990): Social influences on the reproductive success and reproductive endocrinology of prosimian primates. In: Socioendocrinology of primate reproduction. Ed. by T. E. ZiIEGLER and F. B. BERCoVITcCH. New York: Wiley-Liss. Pp. 159-186. JAacogs, W. W. (1976): Male-female associations in the domestic guinea pig. Anim. Learn. Behav. 4, 77- 83 KEnAGY, G. J.; TROMBULAK, S. C. (1986): Size and function of mammalian testes in relation to body size. J. Mammalogy 67, 1-22. KEVERNE, E. B. (1983): Endocrine determinants and constraints on sexual behaviour in monkeys. In: Mate choice. Ed. by P. BATEson. Cambridge: Cambridge Univ. Press. Pp. 407-420. KLEIMAN, D. G. (1977): Monogamy in mammals. Quart. Rev. Biol. 52, 39-69. König, B. (1985): Maternal activity budget during lactation in two species of Caviidae (Cavia porcellus and Galea musteloides). Z. Tierpsychol. 68, 215-230. Kress, J. R.; Davis, N. B. (1993): An introduction to behavioural ecology. Oxford: Blackwell. KÜNKELE, J.; HoEcK, H.N. (1989): Age-dependent discrimination of unfamiliar pups in Galea muste- loides (Mammalia, Caviidae). Ethology 83, 316-319. KUMMER, H. (1984): From laboratory to desert and back: a social system of hamadryas baboons. Anim. Behav. 32, 965-971. LACHER, T. E. (1981): The comparative social behaviour of Kerodon rupestris and Galea spixii and the evolution of behaviour in the Caviidae. Bull. Carnegie Mus. Nat. Hist. 17, 1-71. MADSEN, T.; SHINE, R.; LOMAN, J.; HAKANSSon, T. (1992): Why do female adders copulate so frequently? Nature 355, 440-441. MSLLER, A.P. (1988): Ejaculate quality, testis size and sperm competition in primates. J. Hum. Evolut. 17, 479-488. Roop, J. P. (1972): Ecological and behavioural comparison of three genera of Argentine cavies. Anim. Behav. Monogr. 5, 1-83. Roop, J. P.; WEır, B. J. (1970): Reproduction in female wild guinea pigs. J. Reprod. Fert. 23, 393-409. SACHSER, N. (1984): Auswirkungen unterschiedlicher sozialer Situationen auf Verhalten und Physiologie von Hausmeerschweinchen. Diss. Unversität Bielefeld. SACHSER, N. (1986): Different forms of social organization at high and low population densities in guinea pigs. Behaviour 97, 253-272. SACHSER, N. (1994): Sozialphysiologische Untersuchungen an Hausmeerschweinchen. Hamburg und Berlin: Parey. SACHSER, N.; LICK, C.; STANZEL, K. (1994): The environment, hormones, and aggressive behaviour: A 5- year-study in guinea pigs. Psychoneuroendocrinol. 19, 697-707. SACHSER, N.; PRÖVE, E. (1986): Social status and plasma testosterone-titers ın male guinea pigs (Cavia aperea f. procellus). Ethology 71, 103-114. SCHWARZ, E. (1991): Verhalten und soziale Organisation der Wieselmeerschweinchen (Galea muste- loides Meyen, 1833). Diplomarbeit, Universität Bayreuth. SHORT, R. V. (1979): Sexual selection and its component parts, somatic and genital selection, as illus- trated by man and the great apes. Adv. Stud. Behav. 9, 131-158. TAyvEn, Y.; HENDRICHS, H. (1990): Differences in behaviour and social organization of female guinea pigs as a function of the presence of a male. Ethology 85, 25-34. TRIVERS, R. (1985): Social evolution. Menlo Park, California: Benjamin Cummings Publ. VAN LAWICK-GOODALL, J. (1968): The behaviour of freeliving chimpanzees in the Gombe Stream Re- serve. Anim. Behav. Monogr. 1, 161-311. VAN SCHAIK, C. P; DuNBAR, R. I.M. (1990): The evolution of monogamy in large primates: A new hy- pothesis and some crucial tests. Behaviour 115, 30-62. Von Horst, D. (1969): Sozialer Stress bei Tupajas (Tupaia belangeri). Die Auswirkungen des sym- pathischen Nervensystems und ihre Beziehung zu hormonell ausgelösten ethologischen und physio- logischen Veränderungen. Z. Vergl. Physiol. 63, 1-58. Weir, B. J. (1971): The evocation of oestrus in the cuis, Galea musteloides. J. Reprod. Fert. 26, 405-408. 38 ESTHER SCHWARZ-WEIG and N. SACHSER WEIR, B. J. (1973): The role of the male in the evocation of oestrus in the cuis, Galea musteloides. J. Re- prod. Fert. Suppl. 19, 421-432. Wirson, E. ©. (1975): Sociobiology. The new synthesis. Cambridge (Mass.): Belknap Press of Harvard Univ. Press. Authors’ address: Dipl.-Biol. ESTHER SCHWARZ-WEIG und Prof. Dr. NORBERT SACHSER, Institut für Neuro- und Verhaltensbiologie, Abt. Verhaltensbiologie, Westfälische Wilhelms- Universität Münster, Badestr. 9, D-48149 Münster, BRD | | I | j | | Z. Säugetierkunde 61 (1996) 39-48 ZEITSCHRI FT NZ FÜR © 1996 Gustav Fischer, Jena SÄUG ETI E RKÜ N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Nursing, weaning and the development of independent feeding in the rabbit (Oryctolagus cuniculus) By Rogyn HuDson, AGnEs BıLkö, and V. ALTBÄCKER Institut für Medizinische Psychologie, Ludwig-Maximilian-Universität, München, Deutschland and Department of Ethology, Eötvös Lorand University, Budapest, Hungary Receipt of Ms. 28.09. 1995 Acceptance of Ms. 02. 11. 1995 Abstract Maternal care in the European rabbit is limited to one brief nursing visit a day. To investigate the na- ture of this unusual mother-young relationship, four domestic does and their litters were kept separately except for the once-daily nursing, and the following parameters were recorded; from post-natal days 1 to 30, the duration of nursing bouts, daily milk yield, deposition of faecal pellets in the nest by does, dai- ly weight gain of pups, eating of faecal pellets and nest material by pups, their water intake, and from post-weaning days 31 to 44, their weight gain. Does were mated immediately after giving birth, and the measures for the first litters raised when does were pregnant were compared with the results for the sec- ond litters raised when does were not pregnant. Four control does and their litters were treated in the same way but without separating mothers and young. Pups progressed from drinking milk alone, to nib- bling faecal pellets, to ingesting nest material, drinking water and finally to eating lab food. However, growth rates and the pattern of weaning depended on does’ reproductive state. The first litters, raised by pregnant does, were significantly lighter and were weaned earlier than the second litters raised by the same does when not pregnant. The rabbit thus provides a particularly good opportunity to investi- gate the processes underlying the transition to independent feeding in amammalian species. Introduction Mammals are defined as a taxonomic class by the ability of the mother to feed the new- born young on milk from the mammary glands. This is a physiologically and behaviou- rally complex process for both parties. For the young, the transition from the suckling period during which this milk represents the sole or primary source of nourishment to adult patterns of ingestion is certainly one of the major challenges of early life (GALEF 1981; HArı 1990). For the mother, on the other hand, the provision of milk represents a major investment, the costs of which presumably must be weighed against the conse- quences for her future reproductive fitness (TRIvVErs 1974; PARTRIDGE et al. 1982; CLUT- ToNn-Brock et al. 1989). This apparent conflict of interests raises many questions concerning the nature of the weaning process and the relative contribution of the mother and young to it (GALEF 1981; BATEson 1994). The European rabbit (Oryctolagus cuniculus) provides an unusually good opportunity to investigate the weaning process in a familiar laboratory species. Not only is the repro- ductive physiology and natural ecology of the rabbit relatively well understood (THoMP- son and Kına 1994) but its remarkably limited pattern of maternal care (Hupson and DistEL 1982, 1989) and renowned reproductive efficiency make it particularly suitable for the study of mother-young interactions. 40 Roßyn HuDson, AGnes BıILKö and V. ALTBÄCKER When given the opportunity both wild and domestic does dig a nursery burrow in which they build a nest of grass, and fur pulled from their chest and belly. After giving birth the doe leaves the pups, blocks the burrow entrance and only returns to reopen it and nurse approximately every 24 hours for about 3-4 minutes. On entering the nest the doe simply positions herself over the litter, remaining almost motionless during nursing and not giving the pups any direct behavioural assistance to suckle. Towards the end of nursing she deposits a few faecal pellets in the nest and then jumps abruptly away, leaving the pups alone until the following day (Deutsch 1957; ZARROW et al. 1965; BROEKHUIZEN and MULDER 1983; Hupson and Dister 1982, 1989). Despite such limited care the young develop rapidly, and by weaning at about postna- tal day 27, have increased their birth weight approximately twelve-fold. Eyes and ears open around day nine, and during the third week both domestic and wild pups start to leave the nest (ZARROWw et al. 1965; MyKyrowycz and DupzinskI 1972; KrAFT 1979; HuD- son and DisteL 1982). This rapid development is important as the doe normally comes into oestrus immediately after parturition and is particularly receptive at this time (BRAMBELL 1944). While this enables her to raise several litters in a season, the rapid weaning of the young associated with preparation for the birth of the next litter means that pups have to make the transition to independent feeding largely alone (Hupson and ALTBÄCKER 1994). Given this unusual pattern of maternal care it was the aim of this study to provide a first description of the changes in ingestive behaviour of young rabbits from birth, through weaning, to fully independent feeding, and of corresponding changes in the nurs- ing behaviour of the doe. The findings demonstrate that the development of ingestive be- haviour constitutes a well-defined sequence of stages but that the timing of weaning depends on the reproductive condition of the doe, and in particular, whether she is preg- nant with a second litter. Material and methods Animals A total of eight domestic chinchilla-breed does (Chbb, Thomae, Biberach, Germany) and their litters was used. Does were kept singly in standard metal rabbit cages (75x45x35 cm) at 20°+2°C and on a 16L/8D light-dark cycle (lights on at 06.00 hours), with lab chow (Altromin”) and water available con- tinuously. Several days before parturition they were also provided with nest boxes (40x 30x30 cm) and hay. Test procedures To approximate the natural breeding situation in which does are frequently both pregnant and lac- tating, the development of the first litters from the eight does was studied after the mothers had been mated post partum and so were pregnant during lactation, and this was compared with the de- velopment of the subsequent eight litters raised without remating the mothers. On the day of birth (day 0) the nest boxes were removed, the pups were weighed and individually colour-marked in the ears, and the litters reduced to six pups each. After weighing the nest material, counting the faecal pellets deposited by the doe in the nest, and placing a foam rubber mat in the box to absorb pups’ urine, the nest material, pellets and pups were returned to the nest boxes. Does were mated within a few hours of giving birth, and all responded with immediate lordosis and gave birth to large lit- ters 31-32 days later. These second litters were raised and tested using the same procedure as for the first litters. After weighing and returning newborn pups to the nest boxes, does and their litters were randomly allocated to one of two experimental conditions for the duration of the study. Development of independent feeding in rabbits 41 Condition I: Pups separated from mother This condition was designed to allow accurate recording of nursing behaviour and changes in pups’ in- gestive behaviour without does eating the nest material or causing other disturbance. On the day of birth the nest boxes were removed to a separate room and each morning between 09.00-10.00 h the pups and nest material were separately weighed, faecal pellets were counted, and the pups were re- turned to their box without nest material and brought to the doe for nursing. After nursing, pups were again weighed, fresh faecal pellets were counted, and the pups and fresh pellets, together with the nest material and old pellets were returned to the box until the following day. On post-natal day 14 the pups, nest material and pellets were transferred to large acrylic rodent cages (60 cmx38 cmx20 cm). On day 16 the nest material was replaced with 100 g of fresh hay which was renewed each day, and on day 18 the cages were equipped with a calibrated water bottle. Pups were brought to the doe each day until she refused to nurse, or until day 30, at which time they were transferred in pairs to standard metal rabbit cages, provided with lab chow and water ad libitum, and weighed daily until day 44. Condition II: Pups wıth mother In this condition the nest boxes containing the pups were left permanently with the mothers, enabling does to nurse at any time. The pups were only removed for weighing at 09.00 and 18.00 h each day to monitor their growth and to determine the approximate time of nursing. On post-natal day 26, pups of the first litters were separated from their mother (second litters on day 30), and were transferred in pairs to standard metal rabbit cages and treated as for Condition I. Data analysis Data were subjected to analysis of variance (ANOVA) for repeated measures followed by the Duncan post hoc test in the case of significant F values. For this analysis the nursing period was defined as days 1-25, and the post-weaning period as days 31-44. The alpha level was set at 0.05. 1000 First litters (does pregnant) Weaning | — separated from mother | o with mother | 22 800 Second litters (does not pregnant) -- separated from mother W h e with mother e S a 600 DS wu S O 400 | S | 200 (0) 4 8 2 10, 7.20, 224 Z802732 36 Days post partum Fig. 1. Average daily cumulative weight gain of pups nursed by the eight does from Conditions I and Il when pregnant (“first litters’, N = 8x6 pups), and of pups nursed by the same does when not pregnant (‘second litters’, N= 8x6 pups). Means and SDs are given and the shaded zone represents the termina- tion of nursing as determined by the pregnant does or by the experimenter on day 30. 42 Roßyn Hupson, AGnes BiıLKö and V. ALTBÄCKER Results Two to three days before parturition all does built substantial nests of hay weighing be- tween 98 and 155 g, which they lined with fur. In most cases some faecal pellets were also deposited in the nest at this time although does never urinated there. Using the measures described above for does and litters raised separately (Condition I), the period of mater- nal dependence was found to divide naturally into three main stages: the first week dur- ing which the pups’ sole source of nourishment was the mother’s milk; weeks 2-3 during which they started nibbling at the faecal pellets and hay and started drinking water; and week 4 during which they ate substantial amounts of solid food and at which time, if the mother was pregnant, nursing was abruptly terminated. Litters from Condition II, left per- manently with their mother, only showed an increase in weight when weighed in the morning, suggesting that they also were only nursed once every 24h and most probably during the night. Week 1: Suckling Figure 1 shows the average increase in the weight of pups from all eight first litters raised by pregnant does, and of the eight second litters raised by the same does when not preg- nant. Using a 3-way ANOVA (state of pregnancyxconditionxage) with repeated mea- sures it was found that whereas the second litters from non-pregnant does showed a significantly greater weight gain by day 25 than the first litters from pregnant does [F (1,86) = 21.5, p<0.01], there was no significant difference during the first week when most pups doubled their body weight. However, by day25 the effect of age [F (24,2064) = 3074.8, p< 0.01] and condition [F (1,86) = 9.74, p< 0.01], as well as the inter- action between state of pregnancy and age were significant [F (24,2064) = 19.4, p<0.01], as was a threefold interaction among the factors [F (24,2064) = 2.9, p < 0.01]. The uniform growth of pups was reflected in the similar amounts of milk obtained by each of the separated litters during this period, which ranged from an average of 10 g/pup on day 1 to about 22 g/pup on day 7 (Fig. 2). As can be seen in figure 3, pups generally —0O— First litters 60 (does pregnant) —— Second litters (does not pregnant) 50 40 30 20 Milk intake [g/pup] 10 = - z > 0 4 8 12 16 20 24 28 Days post partum Fig. 2. Average daily milk intake of pups from Condition I nursed by does when pregnant (“first litters’, N = 4x6 pups), and of pups nursed by the same does when not pregnant (‘second litters’, N = 4x6 pups). Data are from the same litters raised separately as in figure 1. Means and SDs are given. Development of independent feeding in rabbits 43 5 — ® I 4 .S S 3 < .Q S = 2 2 1 —O-—- First litters RZ (does pregnant) >= —#— Second litters 0 (does not pregnant) 4 8 12 16 20 24 28 Days post partum Fig. 3. Average time spent nursing each day by the four does from Condition I when pregnant (first lit- ters’) and when not pregnant (‘second litters’). These data correspond to the milk intake values in fig- ure 2. Means and SDs are given. had to obtain these amounts of milk in less than 4 min/day. While the four does from which these measures were taken showed small individual differences in the length of the nursing visits, the duration of these visits remained very stable and did not change mark- edly with the advance of lactation. Weeks 2-3: Transition to solid food During the second and third weeks milk intake increased steadily from about 25 g/pup on day 8 to about 40 g/pup on day 19 at the peak of lactation (Fig. 2). By this time pups weighed between 350-400 g althoush litters from non-pregnant does tended to be heavier (Fig. 1). Differences were also recorded in the average milk intake [F(1,34) = 51.5, —o-— First litters Defecation period (does pregnant) EHER 50 —#— Second litters (does not pregnant) 5 40 ) S EN 2 30 2 a 3 20 \$) Q u 10 0) 0 A 8 12 16 20 24 28 Days post partum Fig. 4. Mean cumulative number of faecal pellets deposited in the nest by the four does from Condi- tion I when pregnant (‘first litters’) and when not pregnant (‘second litters’) until the end of the defeca- tion period (shaded bar), and the cumulative number disappearing from the nest, presumably eaten by the pups. Means and SDs are given. 44 Roßyn HuDson, AGnes BıLKö and V. ALTBÄCKER p<0.01], with litters from pregnant does obtaining significantly less milk by day 23 (Fig. 2) than the litters from non-pregnant does. As shown in figure 4, early in lactation does usually deposited several faecal pellets in the nest at each nursing visit. Although there were considerable individual differences be- tween does in the total number of pellets added to the nest, this behaviour was most con- sistent; pellets were deposited on almost every visit up to day 10, after which time this behaviour abruptly stopped. From about day 11 pellets showed clear signs of having been nibbled, and by the third week they started to disappear at the rate of about one a day. As pups were frequently seen chewing on fragments, the pellets appeared to have been actually eaten and not to have just disintegrated. Towards the end of the second week pups also started nibbling the nest material. At first they appeared simply to bite through the long stalks, reducing the hay to a kind of rough chaff. Despite difficulties in accurately weighing the hay when it became damp from pups’ urine, by about day 18 clearly measurable amounts were being eaten as de- duced from the decrease in the net weight of the remaining material. Week 4: Termination of nursing Weaning appeared to start between days 18-20 with a decline in milk yield which was considerably steeper when the does were pregnant (Fig. 2). At this time pups also started to drink from the bottle, with the litters from pregnant does drinking considerably more than litters from these same does when not pregnant (Fig. 5). Between days 26 and 28 the pregnant mothers completely refused to nurse. Whereas on the preceding day they re- mained in the nest box for the usual time (Fig. 3) and showed apparently normal nursing behaviour, the following day they refused to enter the box and vigorously struck at, bit and cuffed away any pup trying to suckle. In fact, the litters had to be quickly removed to prevent them coming to harm. This contrasted with the behaviour of the non-pregnant does which were never seen responding aggressively to their pups. The transition to lab food resulted in an immediate acceleration in the growth rate of all litters (Fig. 1). This was particularly clear from the difference in weight gain between litters from non-pregnant does from Condition II left with the mother and thus with free 30 —o- First litters (does pregnant) 25 —#— Second litters r (does not pregnant) I S220 & x so 75 .S S 10 W S 5 fo) E 18 22 26 30 Days post partum Fig. 5. Average daily water intake by litters from Condition I nursed by does when pregnant (‘first lit- ters’, N = 4), and of litters nursed by the same does when not pregnant (‘second litters’, N = 4). Means and SDs are given. Development of independent feeding in rabbits 45 access to lab chow during the period of declining milk intake, and litters from non-preg- nant does from Condition I kept separately from the mother and so without access to lab food until they were fully weaned on day 30 (Fig. 1). Using a 3-way ANOVA (state of pregnancyxconditionxage), it was found that during the post-weaning period from days 31-44, litters from pregnant does remained significantly lighter than litters from non- pregnant does [F(1,85) = 21, p<0.01] although the difference between pups from Condi- tions I and II was no longer significant [F(1,85) = 0.8, ns]. Discussion This study provides the first systematic description of the changes in ingestive behaviour of young rabbits from birth through to fully independent feeding, and of corresponding changes in the nursing behaviour of the doe. At first sight the pattern of behaviour of both mothers and young appears stereotyped and largely lacking in the flexibility we nor- mally associate with mammalian mother-infant relationships. Thus, having no other con- tact with their mother than three to four minutes of suckling once a day, pups progressed from drinking milk alone, to nibbling faecal pellets, to ingesting nest material, drinking water and finally to eating lab food, and at very similar ages. Nursing behaviour was also stereotyped, with the duration of does’ visits remaining very constant across litters and throughout lactation. However, flexibility was clearly demonstrated by the marked differences in the pat- tern of weaning depending on does’ reproductive state. In pregnant does, milk yield not only declined earlier and more steeply, but the willingness of does to continue nursing late in lactation ended more abruptly, with pups being rejected and even attacked from one day to the next. Furthermore, neither the limited care nor the change in does’ nurs- ing behaviour appeared to present difficulties for the pups, as evidenced by their steady growth during the period in the nest, clear acceleration in weight gain following the tran- sition to solid food, and good survival rate. This unusually efficient pattern of maternal care cannot simply be attrıbuted to the ar- tificial conditions of this study in which young were raised separated from their mothers. The behaviour and development of the separated litters was very similar to that of the control pups left with their mothers, and is in good agreement with the previous report of similar growth rates for young whether left with the doe or raised separately (ZARROW et al. 1965). The present findings are also consistent with previous reports of constant nurs- ing times (LincoLn 1974), peak milk yields at about day 20 (CowıE 1969; LincoLn 1974; McNitt and Moopy 1990), the abrupt refusal of pregnant does to nurse four to five days before the birth of the next litter (LimcoLn 1974), and with an acceleration in pups’ weight gain following the transition to solid food (VEnGE 1963; McNitt and Moopy 1990). Such precise regulation of the mother-young relationship obviously raises important questions concerning the underlying mechanisms. With regard to the doe these concern the fine timing of the neural and endocrine processes motivating her to nurse at the same time and for a specific period each day (Hupson et al. 1995), and then if pregnant, to steeply reduce milk yield after day 20, and end nursing four to five days later. Almost without pause she must then repeat the cycle; build a new nest, and following parturition, produce milk with a composition appropriate for the newborn young (CowıE 1969) and resume the same tightly timed pattern of nursing. With regard to the development of ingestive behaviour, the fact that pups started eat- ing faecal pellets and hay at similar ages and at a time when milk intake was still increas- ing, suggests that the maturation of endogenous mechanisms associated with the hunger system probably play a significant role. This is consistent with the report of age-depend- ent changes in the motivation to attach to nipples and suckle in rat pups (Harr et al. 46 RoBYN HuDson, AGnes BILKö and V. ALTBÄCKER 1977), and in rabbits by day 13 at about the time they start eating nest material (DistEL and Hupson 1984). On the other hand, as all litters gained weight more rapidly after the transition to solid food, it also seems that pups might be pushed quite early to make this transition by the inability of the mother to provide enough milk to meet their increasing demands (THIELS and ALBERTS 1985), and by the steep decline in lactose, and therefore in the caloric content of milk after day 20 (CowıE 1969). Whether, as has been reported for the rat (GALEF 1981), this transition is accompanied by a decline in lactase, the enzyme permitting the young to absorb and utilize lactose, has yet to be investigated. A further question raised by this study concerns the significance of the faecal pellets deposited in the nest. As they only disappeared from the nest at the rate of about one/ day, they can hardly have constituted a significant food source. However, they may pro- vide pups with gut flora aiding in the digestion of plant food (SmıtH 1965; HÖRNICKE and BJÖRNHANG 1980), a possibility it would be simple to test by monitoring the weight gain and later survival of pups raised without pellets. In this respect it might be relevant that pellet eating was always observed to precede eating of hay by several days. Furthermore, previous findings suggest that does might transmit information concerning their diet via the faecal pellets, thus indirectly influencing pups’ later food choice (Hupson and ALT- BÄCKER 1994; BıLKö et al. 1994; ALTBÄCKER et al. 1995). Sımilarly, ıt is not yet clear to what extent the provision of grass by the doe contri- butes to the development of ingestive behaviour. Again there are several possibilities; that the nest material is of direct nutritional value, that it is of little nutritional value but helps prepare pups for the digestion of plant food, that pups acquire information from the plants selected by the doe about what to eat at weaning, and possibly even from which plants to construct their own nests later. Whereas the first two possibilities could be read- ily investigated by raising pups in nests without plant material, the second two possibili- ties are currently being tested in the laboratory in Hungary, with first findings suggesting that when given a choice of grasses, pregnant does do indeed preferentially select those species from which their own natal nest was constructed. In conclusion, the findings of this study demonstrate that the development of ingestive behaviour in the rabbit constitutes a well-defined sequence of stages but that the timing of weaning depends on the reproductive condition of the doe, and in particular, whether she is pregnant with a second litter. Given the costs to the mother of nursing one litter while pregnant with the next, as suggested by the slower growth rate of first litters in the present study, it should now be interesting to investigate under what nutritional or other conditions does invest in first or subsequent litters, and to what extent they vary litter size or sex ratio (HAmMmonD 1965; Boyp 1985). Particularly interesting in this regard is the abil- ity of rabbits to resorb foetuses and modify or reverse a reproductive decision as late as day 20 in the 31-day gestation (BRAMBELL 1942, 1944). However, to assess the real costs and benefits of the different patterns rearing requires monitoring the later survival and subsequent reproductive performance of both mothers and young, and preferably under natural conditions (CLUTTON-BRocK et al., 1989). Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (Hu 426/1), by the Hungarian Acad- emy of Sciences (F5254/1 991), and by the MHB Foundation for Hungarian Science (No 90/92/IH). We particularly thank Apam MiıkKrösı for assistance with the statistics. Development of independent feeding in rabbits 47 Zusammenfassung. Säugen, Entwöhnung und die Entwicklung unabhängigen Freßverhaltens beim Kaninchen (Oryctola- gus cuniculus). Bei Kaninchen beschränkt sich die Jungenfürsorge auf einen einzigen kurzen Säugebesuch pro Tag. Um diese bei Säugetieren ungewöhnliche Mutter-Kind Beziehung näher zu untersuchen, wurden vier Würfe von ihren Müttern getrennt und für das tägliche Säugen zu ihnen gegeben. Folgende Meßdaten wurden erhoben: von Tag 1 bis Tag 30 die Dauer des Säugens, die Milchaufnahme und die Gewichtszunahme der Jungen, das tägliche Absetzen von Kot von der Mutter in das Nest, der Beginn des Fressens von Kotkugeln und Nistmaterial durch die Jungen, der Beginn und die Menge der selbständigen Wasserauf- nahme, und nach dem Absetzen, von Tag 31 bis Tag 44 die Gewichtszunahme der Jungen durch selb- ständiges Fressen von Laborfutter. Die Mütter wurden unmittelbar nach der Geburt erneut gedeckt, und die Werte ihrer ersten Würfe — aufgezogen während sie trächtig waren — mit den Werten ihrer zwei- ten Würfe — aufgezogen während sie nicht trächtig waren — verglichen. Entsprechende Meßwerte wur- den bei vier Kontrollhäsinnen erhoben, jedoch hatten diese Mütter jederzeit Zugang zu ihren Jungen. Das Freßverhalten der Jungen entwickelt sich von Milchaufnahme alleine hin zu zusätzlichem Knab- bern der Kotkugeln, weiterhin zum Fressen von Nistmaterial, und schließlich zum selbständigen Fressen von festem Laborfutter. Gewichtszunahme und Zeitpunkt des Absetzens hängt jedoch vom Fortpflan- zungszustand der Mütter ab. Diejenigen Würfe, die von gleichzeitig trächtigen Häsinnen aufgezogen wurden, waren signifikant leichter und wurden früher abgesetzt als diejenigen Würfe, die von nicht trächtigen Häsinnen aufgezogen wurden. Aufgrund dieser Untersuchungen scheint das Kaninchen für Untersuchungen über Entwöhnung und Entwicklung von Freßverhalten bei Säugetieren besonders geeignet zu sein. References ALTBÄCKER, V.; Hupson, R.; BıLkö, A. (1995): Rabbit mothers’ diet influences pups’ later food choice. Ethology 99, 107-116. BATESson, P. (1994): The dynamics of parent-offspring relationships in mammals. Trends Ecol. Evol. 9, 399-402. BiıLkö, A.; ALTBÄCKER, V.; HupDson, R. (1994): Transmission of food preference from mother to young in the rabbit: the means of information transfer. Physiol. Behav. 56, 907-912. Boyp, I. L. (1985): Investment in growth by pregnant wild rabbits in relation to litter size and sex of the offspring. J. Anim. Ecol. 54, 137-147. BRAMBELL, F. W. R. (1942): Intra-uterine mortality of the wild rabbit, Oryctolagus cuniculus (L.). Proc. Roy. Soc. London B 130, 462-479. — (1944): The reproduction of the wild rabbit Oryctolagus cuniculus (L.). Proc. Zool. Soc. London 114, 1-45 BROEKUIZEN, $.; MULDER, J. L. (1983): Differences and similarities in nursing behavior of hares and rab- bits. Acta Zool. Fennica 174, 61-63. CLUTTON-BROocK, T. H.; ALBoN, S. D.; Guinness, F. E. (1989): Fitness costs of gestation and lactation in wild mammals. Nature 337, 260-262. CowiE, A.T. (1969): Variation in the yield and composition of the milk during lactation in the rabbit and the galactopoietic effect of prolactin. J. Endocrinol. 44, 437-450. Deutsch, J. A. (1957): Nest building behaviour of domestic rabbits under seminatural conditions. Brit. J. Anim. Behav. 5, 53-54. Dister, H.; Hupson, R. (1984): Nipple-search performance by rabbit pups: changes with age and time of day. Anim. Behav. 32, 501-507. GALEFE, D. G. Jr. (1981): The ecology of weaning. In: Parental care in mammals. Ed. by D. J. GUBERNICK and P. H. KLoprer. New York: Plenum Press. Pp. 211-241. HALL, W. G. (1990): The ontogeny of ingestive behavior. In: Handbook of behavioral neurobiology. Vol.10: Neurobiology of food and fluid intake. Ed. by E. M. STrRICKEr. New York: Plenum Press. Pp. 77-123. =; CRAMMER, C. P; Brass, E. M. (1977): Ontogeny of suckling in rats: transitions toward adult ingestion. J. Comp. Physiol. Psychol. 91, 1141-1155. 48 Roßyn HuDson, AGnes BILKÖ and V. ALTBÄCKER HAMMOND, J. Jr. (1965): The effects of high and low planes of nutrition on reproduction in rabbits. New Zeal. J. Agric. Res. 8, 708-717. HÖRNICKE, H.; BJÖRNHAG, J. G. (1980): Coprophagy and related strategies for digesta utilization. In: Di- gestive physiology and metabolism in ruminants. Ed. by Y. RUCKEBUSCH and P. THIVEND. Lancaster: MTP Press. Pp. 707-730. Hupson, R.; ALTBÄCKER, V. (1994): Development of feeding and food preference in the European rab- bit: environmental and maturational determinants. In: Behavioral aspects of feeding. Ed. by B. G. GALEF, Jr.; M. MAINARDI and P. VALSEccHI. Chur: Harwood Academic Publishers. Pp. 125-145. -; DisteEL, H. (1982): The pattern of behaviour of rabbit pups in the nest. Behaviour 79, 255-271. -;— (1989): Temporal pattern of suckling in rabbit pups: a model of circadian synchrony between mother and young. In: Development of circadian rhythmicity and photoperiodism in mammals. Ed. by S.M. REpPperT. New York: Perinatology Press. Pp. 83-102. -; MÜLLER, A.; KEnNEDy, G. (1995): Parturition in the rabbit is compromised by daytime nursing: The role of oxytocin. Biol. Reprod. 53, 519-524. KRAFT, R. (1979): Vergleichende Verhaltensstudien an Wild- und Hauskaninchen. I. Das Verhaltensin- ventar von Wild- und Hauskaninchen. Z. Tierzücht. Züchtungsbiol. 95, 140-162. Lincorn, D. W. (1974): Suckling: a time-constant in the nursing behaviour of the rabbit. Physiol. Behav. 13, 711-714. MEeNITT, J. 1.; Moopy, G. L. Jr. (1990): Daily milk intake by rabbit kits. J. Appl. Rabbit Res. 13, 176-178. MykyTowycz, R.;, Dupzinsk1, M.L. (1972): Aggressive and protective behaviour of adult rabbits Oryc- tolagus cuniculus (L.) towards juveniles. Behaviour 43, 97-120. PARTRIDGE, G. G.; FULLER, M. F.; PULLAR, J. D. (1982): Studies on the protein and energy requirements of the lactating rabbit. Anim. Prod. 34, 384. SMITH, H. W. (1965): The development of the flora of the alimentary tract in young anımals. J. Path. Bact. 90, 495-513. THIELS, E.; ALBERTS, J. R. (1985): Milk availability modulates weaning in the Norway rat (Rattus norve- gicus). J. Comp. Psychol. 99, 447-456. THompson, H. V.; Kıng, C. M. (Eds), (1994): The European rabbit: the history of a successful colonizer. Oxford: Oxford Univ. Press. TRIVERS, R. L. (1974): Parent-offspring conflict. Amer. Zool. 14, 249-264. VENGE, ©. (1963): The influence of nursing behaviour and milk production on early growth in rabbits. Anim. Behav. 11, 500-506. ZARROW, M. X.; DENENBERG, V. H.; ANDERSON, C. O. (1965): Rabbit: frequency of suckling in the pup. Science 150, 1835-1836. Authors’ addresses: Dr. Rosyn Hupson, Institut für Medizinische Psychology, Ludwig-Maximilian Universität, Goethestr. 31, D-80336, München, Germany; A. BıLkö and V. ALT- BÄCKER, Department of Ethology, Eötvös Loränd University, H-2131 Göd, Javorka S. u. 14, Hungary. u Z. Säugetierkunde 61 (1996) 49-53 ZEITSCHRIFT DE "FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY WISSENSCHAFTLICHE KURZMITTEILUNGEN Chromosomal conservatism in southern African Klipspringer antelope (Oreotragus oreotragus): a habitat specialist with disjunct distribution. By T. J. RoBınson, J. DU P. BoOTHMA, N. FAIRALL, W. R. HARRISON, and F. F. B. ELDER Department of Zoology and Entomology, University of Pretoria, Pretoria, Centre for Wildlife Manage- ment, University of Pretoria, Pretoria, Western Cape Division of Nature Conservation, Jonkershoek Research Station, Stellenbosch, South Africa, and Department of Pathology and Laboratory Medicine, The University of Texas Medical School - Houston, Health Science Center, Houston, Texas, U.S.A. Receipt of Ms. 17. 10. 1995 Acceptance of Ms. 13. 11. 1995 Klipspringer are small antelope, standing approximately 60 cm at the shoulder, which in- habit mountainous, rocky habitats to which they are closely confined (SMITHERS 1983). Although widely distributed from the Western Cape Province of South Africa through to east Africa (Fig. 1), their occurrence within this range is patchy and discontinuous. This disjunct distribution raises the possibility that the reported phenotypic differences be- tween the 11 recognized subspecies (AnseLL 1972 but see GrußB 1993) reflect underlying genetic partitioning promoted, in part, by limited gene flow between demes. Should these genetic differences extend to fixed chromosomal rearrangements between geographically defined populations, the translocation of this antelope from one area to another as part of game ranching practices, and ecotourism considerations, could have significant long-term conservation implications for the species. We have argued elsewhere (Rogınson and ELDER 1993) that matings between speci- mens characterized by different cytotypes can result in perinatal mortality or, at a later stage, in reduced fertility of offspring heterozygous for chromosomal rearrangements. The current investigation was prompted by concerns relating to the relocation of klipspringer from the northern reaches of South Africa into suitable habitat in the extreme southwes- tern portion of the Western Cape Province, and the fear that the translocation could nega- tively impact on the viability of resident populations in the area. Our aim was therefore to ascertain whether geographically discrete populations of the klipspringer differed de- tectably in chromosome structure. In so doing, we would provide wildlife managers with genetic guidelines (albeit only at the level of gross genetic incompatibility) for the devel- opment of a translocation policy for this species within southern Africa. Additionally, as far as we have been able to ascertain, these cytogenetic data are the first recorded for O. oreotragus, an endemic African antelope of the subfamily Neotraginae. Ear clippings were taken in the feld from adult klipspringer specimens, transported to the laboratory, and used to establish fibroblast cell cultures following routine procedures. The country of origin, collection localities and number of specimens analyzed are: Zim- babwe: Bulawayo (20°07’S 28°35’ E); one male and one female; South Africa: Grabouw (34°09' S 19°01’E); one male; Kruger National Park (23°50’S 31 °30'’ E); one male and one female. Air-dried slides were G-banded using ENZAR-T trypsin (Intergen Cat. number 7000- 65) and C-banded using a barium hydroxide 2x SSC treatment (SuUMNER 1972). G-banded 50 T. J. ROBINSON, J. DU P. BOTHMA, N. FAIRALL, W. R. HARRISon, and F. F. B. ELDER Er: Fig. 1. Pan-African distribution of the klipspringer, Oreotragus oreotragus (from SMITHERS 1983). White dots indicate approximate collection localities for specimens used in the present study. karytoypes from each specimen were arranged and numbered according to the standard cattle GTG-banded karyotype (IscnpA 1989). However, the contracted state of the klip- springer chromosomes, especially the smaller autosomes, confounded comparisons with the cattle standard and in some instances their placement as cattle homologs is equivocal. The klipspringer analyzed herein, irrespective of their geographic origin, are charac- terized by an invariant karyotype with 2n = 60; all chromosomes, including the X and Y, are acrocentric in morphology (Fig. 2). This striet chromosomal conservatism, reflected by both G-banding and diploid number, extends also to the species’ C-bands. The klipsprin- ger constitutive heterochromatin is centromeric. No marked heteromorphisms were evi- | Klipspringer chromosomes Jul Fig. 2. Typical G-banded karyotype of the klipspringer (Oreotragus oreotragus). dent within or between specimens, and no interstitial C-bands have been detected in the specimens examined. The Y chromosome appears entirely heterochromatic following C- banding, but is not as darkly staining as the pericentromeric material present in all auto- somes and the X. This presumably reflects some difference in the repeated sequences comprising the heterochromatin at these sites. The chromosome number within the family Bovidae varies from 2n =30 to 2n = 60 but the number of chromosomal arms has remained relatively constant at 56-58 for most karyotyped bovids (GALLAGHER and WoMAcK 1992). It is generally regarded that bovid chromosomal evolution has progressed from a primitive karyotype with 2n = 60 compris- ing 58 acrocentric autosomes, a configuration currently retained in representatives of many of the bovid subfamilies (WURSTER and BENIRSCHKE 1968; GALLAGHER and WOMACK 1992; GALLAGHER et al. 1994; Roßınson and HARLEY 1995). Therefore, in spite of its pro- pensity to population fragmentation due to habitat constraints, the klipspringer, like so 52 T. J. RoBINsSon, J. DU P. BOTHMA, N. FAIRALL, W. R. HARRISoN, and F. F. B. ELDER Fig. 3. Typical C-banded metaphase chromosomes of the klipspringer (Oreotragus oreotragus); inset shows the heteropycenotic nature of the Y chromosome. many of the Bovidae, appears to have retained the ancestral autosomal condition. Of course it is not possible to isolate a single factor which may have contributed to the main- tenance of this primitive state, but, clearly, it does raise the question of how efficient po- pulation subdivision and the formation of small demes in this species is. Small effective population size (N.) is generally considered a prerequisite for the fixation of chromoso- mal rearrangements (Wiırson et al. 1975; BusH et al. 1977; Kıng 1993). Occasional ex- changes between apparently isolated populations will decrease stochastic genetic loss (caused through the effects of inbreeding) and increase the effective population size (CHESSER 1983) and may therefore act to constrain chromosomal evolution. In this re- spect it is noteworthy that in spite of their saxicolous habit and the accompanying ex- treme specialization in hoof morphology (klipspringer walk on the tips of their hooves — an adaptation to their life among rocks), there are reports documenting relatively long range movements of animals between rocky areas which are separated by unsuitable habi- tat (NIETHAMMER 1942; Wırson and CHıLD 1965; SMITHERS 1971). Such observations sug- gest that the klipspringer will show weak geographic structure when assessed by other genetic markers. This prediction receives preliminary support from mitochondrial DNA data which show no meaningful differences between populations in east and southern Africa. While our investigation removes one of the more obvious concerns governing the movement of klipspringer over large geographic distances, it must be emphasized that the apparent lack of regional cytotypes does not imply the absence of genetic differences be- Klipspringer chromosomes 33 tween distant populations. Quite clearly, additional studies providing data at a finer level of resolution and a more extensive geographic representation are prerequisites for the de- velopment of successful translocation policies. Until such data become available, propo- sals to relocate these antelope over extensive geographic distance should be considered cautiously. Acknowledgements Financial support from the Centre for Wildlife Management and the South African Foundation for Re- search Development is gratefully acknowledged. References ANSELL, W. F.H. (1972): Order Artiodactyla. In: The mammals of Africa: an identification manual 15: Ed. by J. MEESTER and H. W. SETZER. Washington, D. C.: Smithsonian Institution Press. Pp. 1-84. BusH, G. L.; Case, S. M.; WıLson, A. C.; PATToN, J. L. (1977): Rapid speciation and chromosomal evolu- tion in mammals. Proc. Nat. Acad. Sci. U.S.A 74, 3942-3946. CHESSER, R. K. (1983): Isolation by distance: relationship to the management of genetic resources. In: Genetics and Conservation. Ed. by C. M. SCHONEWALD-Cox, S. M. CHAMBERS, B. MACcBRYDE and L. Thomas. London: Benjamin/Cummings. Pp. 66-77. GALLAGHER, D. S.; WoMAcK, J. E. (1992): Chromosome conservation in the Bovidae. J. Hered. 83, 287-298. GALLAGHER, D. S.; DERR, J. N.; WomMmAack, J. E. (1994): Chromosome conservation among the advanced pecorans and determination of the primitive bovid karyotype. J. Hered. 85, 204-210. GRUBB, P. (1993): Order Artiodactyla. In: Mammal species of the world: A taxonomic and geographic reference. Sec. Ed. Ed. by D. E. Wırson and D. M. REEDER. Washington, D. C.: Smithsonian Institu- tion Press. Pp. 377—414. IscnDA (1990): (International System for Cytogenetic Nomenclature of Domestic Animals). The second in- ternational conference on standardization of domestic animal karyotypes, Jouy-en-Josas, France May 22-26, 1989. DIBERARDINO, D.; HAYES, H.; FRIES, R.; Long, S.) Cytogenet. Cell Genet. 53, 65-79. Kıng, M. (1993): Species evolution the role of chromosome change. Cambridge: Univ. Press. NIETHAMMER, G. (1942): Über den Klippspringer Deutsch-Südwest-Afrikas. Zool. Gart. Leipzig 14, 139- 149. ROBINSON, T. J.; ELDER, F. F. B. (1993): Cytogenetics: its role in wildlife management and the genetic conservation of mammals. Biol. Conserv. 63, 47-51. ROBINSON, T. J.; HARLEY, E. H. (1995): Absence of geographic chromosomal variation in roan and sable antelope and the cytogenetics of a naturally occurring hybrid. Cytogenet. Cell Genet. 71, 363-369. SMITHERS, R. H.N. (1971): The mammals of Botswana. Mus. mem. Natl. Mus. Monum. Rhod. 4, 1-340. SMITHERS, R.H.N. (1983): The Mammals of the Southern African Subregion. Pretoria: University of Pretoria. SUMNER, A.T. (1972): A simple technique for demonstrating centromeric heterochromatin. Expl. Cell. Res. 75, 304-306. WiLson, A. C.; BusH, G. L.; CAs£, S. M.; Kınc, M. (1975): Social structuring of mammalian populations and rate of chromosomal evolution. Proc. Nat. Acad. Sci. U.S.A. 72, 5061-5065. WiLson, V. J.; CHILD, G. T. F. (1965): Notes on the klipspringer from tsetse fly control areas in eastern Zambia. Arnoldia Rhod. 1, 1-9. WURSTER, D. H.: BENIRSCHKE, K. (1968): Chromosome studies in the superfamily Bovoidea. Chromoso- ma 25, 152-171. Authors’ addresses: T. J. Rogınson, Department of Zoology and Entomology, University of Pretoria, Pretoria, South Africa, J. Du P. BOTHMA, Centre for Wildlife Management, Univer- sity of Pretoria, Pretoria, South Africa, N. FAIRALL, Western Cape Division of Na- ture Conservation, Jonkershoek Research Station, Stellenbosch, South Africa, and W.R.Harrıson and F.F.B. ELper, Department of Pathology and Laboratory Medicine, The University of Texas Medical School-Houston, Health Science Cen- ter, Houston, Texas, U.S.A. Z. Säugetierkunde 61 (1996) 54-58 © 1996 Gustav Fischer, Jena Interindividuelle Zyklusdesynchronisation bei Goldhamsterweibchen, Mesocricetus auratus Von R. GATTERMANN Institut für Zoologie, Martin-Luther-Universität Halle-Wittenberg, Halle Eingang des Ms. 20. 05. 1995 Annahme des Ms. 24. 07. 1995 Zyklussynchronisation beschreibt die interindividuelle zeitliche Abstimmung der weibli- chen Sexualzyklen hinsichtlich Physiologie und Verhalten. Sie ist häufig Bestandteil von saisonalen, lunaren und tidalen Fortpflanzungszyklen, die primär durch geophysikalische Zeitgeber synchronisiert werden. Diese biorhythmische Synchronisation kann durch eine inter- oder intrasexuelle ethologische Synchronisation, das heißt durch spezifische Verhal- tensinteraktionen zwischen Männchen-Weibchen und Weibchen-Weibchen, verstärkt und beschleunigt werden. Intersexuelle, hauptsächlich zyklusinduzierende und damit auch syn- chronisierende Effekte der Männchen, sind für zahlreiche Arten belegt (HANDELMANN et al. 1980; McCrmtock 1981). Intrasexuelle Zyklussynchronisation kommt bei Ratten- (McCuimrtock 1978), Wildschwein- (DELcRoOIX et al. 1990) und Affenweibchen (WALLIıs et al. 1986; FRENCH und STRIBLEY 1987) vor, sofern sie stabilen Sozietäten angehören und physische Kontaktmöglichkeiten haben. Ebenso kann bei Frauen, die in enger Wohnge- meinschaft leben, eine Synchronisation der Menstruationszyklen nachgewiesen werden (McCımrtock 1971). Alle diese Arten haben eine soziale Lebensweise und ihre artspezi- fischen Zykluslängen weisen eine mehr oder weniger umfangreiche intra- und interindivi- duelle Variabilität auf. Im Gegensatz dazu, ist der Sexualzyklus der Goldhamsterweibchen außergewöhnlich stabil. Er beträgt unter standardisierten Licht-Dunkel-Bedingungen (L:D=12:12, 14:10) exakt 4 Tage und läßt sich in die Phasen Proöstrus (28h), Östrus (16h), Met- östrus (24h) und Diöstrus (28h) unterteilen (GATTERMANN et al. 1985; FRITZSCHE 1987). Abweichende Zykluslängen von 3 oder 5 Tagen kommen nur beim Übergang von zyklisch zu azyklisch (anöstrisch) und vice versa während der Frühjahrs- und Herbstmonate, unter künstlichen Dauerlicht- (LL) oder Dauerdunkelbedingungen (DD) sowie bei größeren Stressbelastungen vor. Unsere mehrjährigen Beobachtungen unter weitgehend natürli- chen Bedingungen zeigen, daß beide Geschlechter solitär leben und territorial sind. Weib- chen sind streng territorial gegenüber Artgenossinnen, und sie räubern und zerstören jeden neu angelegten Bau eines fremden Weibchens. Männchen und deren Baue dulden sie in ihrem Territorium nur während der Fortpflanzungszeit. Die Territorien der Männ- chen sind wahrscheinlich größer als die der Weibchen. Die Goldhamstermännchen durch- streifen regelmäßig mehrere Weibchenterritorien und prüfen die Duftmarken am Baueingang. Treffen sie auf ein östrisches Weibchen, so folgen sie in den Bau. Ist das östrische Weibchen noch im Bau, so wird er geöffnet, und das Männchen dringt ein. Die Kopulationen finden im Bau des Weibchens statt, ziehen sich über fast die gesamte nächt- liche Aktivitätszeit hin und werden nur von einem Männchen vollzogen. Eine Zyklussyn- chronisation der Territoriumsnachbarinnen würde zusätzliche Konkurrenz schaffen und auch den Fortpflanzungserfolg der Männchen reduzieren, die pro Nacht nur einen Bau Interindividuelle Zyklusdesynchronisation bei Goldhamsterweibchen 55 aufsuchen. Deshalb sollten benachbarte Weibchen kein phasensynchrones Paarungsver- halten aufweisen. Vergleichend untersucht wurden über 120 Tage insgesamt 48 Goldhamsterweibchen (Mesocricetus auratus Waterhouse, 1839) des Zuchtstocks Zoh: GOHA im Alter von 12- 14 Wochen zu Beginn der Untersuchungen. Sie wurden einzeln gehalten (a) ohne phy- sische Kontaktmöglichkeiten in den üblichen Käfigen (Plastikschale 60x40x20 cm mit Drahtdeckel) in einem Regal nebeneinander stehend und (b) in Anlehnung an die Frei- landsituation in sechs 80x 60x40 cm großen Drahtkäfigen mit 1cm Maschenweite und physischen Kontaktmöglichkeiten zu benachbarten Weibchen und einem Männchen (Abb. 1). Alle untersuchten Tiere befanden sich gemeinsam mit anderen Männchen, Weibchen und Jungtieren in einem fensterlosen Raum mit L:D=12:12 (300:5 x), Licht-an um 7.00 Uhr MEZ, Temperatur 24 +1°C, relative Luftfeuchte 65-75%. Pelle- tiertes Standardfutter und kommunales Leitungswasser standen ad libitum zur Verfügung. Die Untersuchungsgruppen von je 6 Weibchen wurden 100% zyklussynchron (92d, 94a, 94b, 94c, 94d), 75% zyklyssynchron (92b) und zyklusdesynchron (92a, 92e) zusammen- gestellt. Das Zyklusstadium wurde anhand der Scheiden-Abstrichbilder bestimmt, die in regelmäßig oder zufällig gewählten Abständen von allen Weibchen gleichzeitig angefer- tigt wurden. Die Berechnung des Synchronisationsgrades einer Untersuchungsgruppe er- folgte nach: V (np = Mm). (Ma — mE n Synchronisationsgrad = n = Anzahl der Individuen im P = Proöstrus, Ö = Östrus, M = Metöstrus und D = Diöstrus In diese Formel gehen die Phasenverschiebungen der Zyklen als Vektorlängen ein. So- mit wird beispielsweise die Phasendifferenz zwischen zwei Weibchen im Proöstrus und Metöstrus stärker bewertet als zwischen zwei im Proöstrus und Östrus. Die untersuchte Gruppe gilt als „desynchron“ bis zu einem Synchronisationsgrad <0,70, da in einer 2er- Gruppe eine Zyklusdifferenz von nur einer Phase (z.B. P und Ö) als noch „synchron“ be- wertet wird. Einzeln gehaltene Goldhamsterweibchen in üblichen Käfigen ohne physische Kontakt- möglichkeiten behalten ihre Sexualzyklen über die Untersuchungszeit von 120 Tage un- verändert bei, denn der Synchronisationsgrad der Untersuchungsgruppen bleibt konstant (Abb. 2 oben). Nur im Fall der Gruppe 94c war bei einem Weibchen spontan der Zyklus um einen Tag verschoben. Können einzeln gehaltene Goldhamsterweibchen physischen Kontakt zu ihren Nachbarinnen und zu einem Männchen aufnehmen (Abb. 2 unten), dann beginnt nach etwa vier Wochen eine interindividuelle Zyklusdesynchronisation, so daß im Mittel nach sechs Wochen die 70%-Schwelle unter- schritten wird und Desynchronität der Zyklen besteht. Dieser Zustand stabilisiert sich und in keinem Fall kann eine Resynchronisation beobachtet werden, auch nicht bei der von Beginn an zyklusdesynchronen, zusätzlichen Kontrollgruppe 92e. Prüft man, bei welchen der Weib- chen die Zyklusänderungen beginnen, so waren es 5mal die von drei Seiten mit zyklussynchronen Nachbarinnen umgebenen Weibchen (W3 und W4, Abb. 1) und nur 2mal die übrigen. Der Unterschied (5:2) ist gemessen am Erwartungswert (6:12) signifikant (Chi? = 4,57; p = 0,03), was ebenfalls für eine aktive Zyklusdesyn- Abb. 1. Anordnung der mit je einem Weibchen (W1-W6) besetz- ten Drahtkäfige, die von einem Männchen (M) aufgesucht werden FD, 5 f konnten. chronisation spricht. Die dafür notwendigen Signale 56 R. GATTERMANN scheinen olfaktorischer Art zu sein und nur im direkten Nahfeld zu wirken. Ob direkte physische Kontakte, ein individuelles Kennen oder Pheromone der Flanken- oder akzes- sorischen Geschlechtsdrüsen bzw. andere Substanzen im Kot oder Urin der Nachbarinnen oder des Männchens beteiligt sind, ist bisher unbekannt. Eine Übertragung mit dem Luft- strom über Strecken >10 cm oder eine akustische Kommunikation sind auszuschließen, denn die normale Käfignachbarschaft zu Weibchen und auch zu Männchen oder die An- wesenheit zyklussynchroner Tiere in einem Raum bewirken keine Desynchronisation. Analoges wird für die zyklussynchronisierenden Mechanismen beschrieben. Zyklus- synchronisation (McCLintock 1978; FRENCH und STRIBLEY 1987; DELCROIx et al. 1990) und auch die Stabilisierung irregulärer Zyklen (WarLıs et al. 1986) ist nur bei engeren Kon- taktmöglichkeiten oder über Urin- und Kotmarken (PELLISSIER-ScoTT 1986) gegeben. Im r- 90,4 c e 0,3 an 0,2 0,1 - 92a 0,0 + zul: T T T T LTE T az Vz len] 1 7 ı3 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103 109 115 Tage MW 0,2 92e 92d 0,1 0,0 T T im amt T T ST a A a RE ee! | BEE) Peaa y | 1 2 1319725731 37 43 49 55 61 67 73 197857 IE I7E03SHO Is Tage Abb. 2. Zyklusdesynchronisation zwischen einzeln gehaltenen Goldhamsterweibchen in üblichen Käfi- gen (oben) und in Drahtkäfigen mit physischen Kontaktmöglichkeiten (unten). Jeweils 6 Tiere pro Un- tersuchungsgruppe; Symbole kennzeichnen die unterschiedlichen Zeiten der Abstrichentnahme; MW = geglättete Mittelwertskurve (ohne Kontrollgruppe 92 e). Interindividuelle Zyklusdesynchronisation bei Goldhamsterweibchen I Gegensatz dazu stehen Untersuchungen von HANDELMANN et al. (1980), die abweichend von anderen (LAUBSCHER und MAGALHAES 1962; Gross 1977; GATTERMANN et al. 1992), für Goldhamsterweibchen eine Zyklussynchronisation bereits nach 1-4 Tagen, d.h., inner- halb eines Zyklus ergaben. Eine erste Zyklussynchronisation beginnt bei Ratten nach 15- 20 Tagen (3-5 Zyklen) und ist stabil nach 30 Tagen (7-8 Zyklen). Bei Frauen sind es 3-4 und 7 Monate bzw. Zyklen (McCuirtock 1971). Beim Goldhamster (Abb. 2 unten) sind erste Zyklusdesynchronisationen nach etwa 28 Tagen (7 Zyklen) erkennbar, eine Stabilität wird im Mittel zwischen 40-60 Tagen (10- 25 Zyklen) erreicht, auch das bestätigt wiederum die außergewöhnliche Zyklusstabilität beim Goldhamsterweibchen. Zum anderen wird in Gruppen von 2-5 Weibchen sehr schnell eine despotische Rangordnung etabliert, und in Abhängigkeit von der sozialen Belastung (Soziostreß) kommt es bei den subordinierten Weibchen zu Zyklusirregularitä- ten und Anöstri, jedoch nicht zur Zyklussynchronisation (GATTERMANN et al. 1992). Spermienkonkurrenz und „female choice“ sind für Goldhamster unter Laborbedin- gungen nachgewiesen (Huck und Liısk 1985; Huck et al. 1986), scheinen jedoch aufgrund der Territorialität der Männchen sowie der langen Aufenthaltsdauer immer nur eines Männchens im Bau bis zum Abschluß der Ovulation (1-3 h nach Mitternacht) für Frei- landbedingungen von untergeordneter Bedeutung. Zusammenfassend kann festgestellt werden, daß im Gegensatz zu sozial lebenden Ar- ten, die solitären und streng territorialen Goldhamsterweibchen über keine Möglichkei- ten der Zyklussynchronisation verfügen. Sie können aber ihre sich zufällig in Phase befindlichen Sexualzyklen wechselseitig desynchronisieren, wenn sie über ein eigenes Ter- ritorium verfügen und enge Kontaktmöglichkeiten zu den benachbarten Weibchen haben. Der Einfluß reproduktiver Männchen ist noch nicht abgeklärt, denn durch die Zyklusde- synchronisation wird das operationale Geschlechterverhältnis zugunsten der Männchen verschoben. Danksagungen Für die sorgfältige technische Assistenz danke ich Frau K. WAEGNER und Frau B. GEBHARDT. Herrn Dipl.-Physiker Dr. G. TscHhucH möchte ich für Aufstellung der Formel zur Berechnung des Synchronisa- tionsgrades danken. Mit Unterstützung durch die Deutsche Forschungsgemeinschaft (Ga 434/1-2). Literatur DELCROIX, I.; MAUGET, R.; SIGNORET, J. P. (1990): Existence of synchronization of reproduction at the level of the social group of the European wild boar Sus scrofa. J. Reprod. Fertil. 89, 613-618. FRENCH, J. A.; STRIBLEY, J. A. (1987): Synchronization of ovarian cycles within and between social groups in golden lion tamarins (ZLeontopithecus rosalia). Am. J. Primat. 12, 469-478. FRITZSCHE, P. (1987): Zur Infradianrhythmik (Circaquadrianrhythmik) des Goldhamsters (Mesocricetus auratus). Zool. Jb. Physiol. 91, 403-418. GATTERMANN, R.; FRITZSCHE, P.; KRAMER, S. (1985): Zur Biorhythmik des Goldhamsters. 3. Infradiane Rhythmen. Zool. Jb. Physiol. 89, 279-285. GATTERMANN, R.; NEUHÄUSER, P.; MILER, H. (1992): Zur sozialen Synchronisation biologischer Rhyth- men bei weiblichen Goldhamstern. In: Verh. Deutschen Zool. Ges. Stuttgart, Jena, New York: Fischer Verlag. Pp.111. Gross, G. H. (1977): A technique for sustained synchronization of hamster estrous cycles by hormonal means. Horm. Behav. 9, 23-31. HANDELMANN, G.; RAvIzZA,R.; Ray, W.J. (1980): Social dominance determines estrous entrainment among females hamsters. Horm. Behav. 14, 107-115. Huck, U. W.; Lisk, R. D. (1985): Determinants of mating success in the golden hamster. IV. Sperm com- petition. Behav. Ecol. Sociobiol. 17, 239-252. 58 R. GATTERMANN Huck, U. W.; Lisk, R. D.; ALLIson, J. C.; VAN DoNGeEN, €. G. (1986): Determinants of mating success in the golden hamster: Social dominance and mating tactics under seminatural conditions. Anim. Be- hav. 34, 971-989. LAUBSCHER, J. A.; MAGALHAES, W. (1962): The estrous cycle in golden hamster. Am. Zool. 2, 423. MCCLINTock, M.K. (1971): Menstrual synchrony and suppression. Nature 229, 244. MCCLinTock, M.K. (1978): Estrous synchrony and its mediation by airborne chemical communication (Rattus norvegicus). Horm. Behav. 10, 264-276. MCCLinTock, M.K. (1981): Social control of the ovarian cycle and the function of estrous synchrony. Amer. Zool. 21, 243-246. PELLISSIER-SCOTT, M. (1986): The timing and synchrony of seasonal breeding in the marsupial Antechi- nus stuartii interaction of environmental and social cues. J. Mammalogy 67, 551-560. WALLIS, J.; Kıng, B. J.; ROTH-MEYER, CH. (1986): The effect of female proximity and social interaction on the menstrual cycle of crab-eating monkeys (Macaca fascicularis). Primates 27, 83-94. Anschrift des Verf.: Prof. Dr. habil. ROLF GATTERMANN, Institut für Zoologie, Martin-Luther-Universi- tät Halle-Wittenberg, Domplatz 4, D-06108 Halle/Saale Z. Säugetierkunde 61 (1996) 59-61 © 1996 Gustav Fischer, Jena Role of apoptosis in seasonal involution and recrudescence of testis By OonA Hiınsrt, S. BLOTTNER, and H. H. D. MEYER Institute for Zoo Biology and Wildlife Research, Berlin Receipt of Ms. 09. 10. 1995 Acceptance of Ms. 29. 11. 1995 Transitions between total arrest and recrudescence of spermatogenesis occur in several mammals with seasonal cycles of testicular activity. The involution of testicular parenchy- ma could result from a cessation of proliferation in seminiferous tubules. However, the significant decrease in testicular weight and function —- an efficient mechanism to adjust physiology and behaviour of the animals to the needs of the non-breeding season — can not be explained exclusively by such a process. A second mechanism of regulated, pro- grammed reduction of tissue such as apoptosis (programmed cell death) must be in- volved. Apoptosis is an active, genetically governed, signal-induced process of selective cell elimination (SCHWARTZMAN and CiDLowskI 1993; TENNISWOOD et al. 1992). Apoptosis seems to be activated and inactivated at different times in the annual cycle as an antago- nist of seasonal testicular proliferation. Thus, changes in levels of apoptosis could contri- bute to involution and reactivation of the testis and to fluctuations in sperm production. Therefore, quantitative measurements of both the proliferation-specific antigen TPS and apoptotically produced nucleosomes have been compared with sperm and testosterone production during the transition from breeding to non-breeding seasons in testes from 80 adult roe deer, Capreolus capreolus (May 1994-October 1995) and from 50 brown hares, Lepus europaeus (June 1993-February 1994). Testes of the animals investigated, were removed immediately after hunting. Testis weights were measured. The number of testicular spermatozoa was counted in prepared homogeneous suspensions of spermatogenic cells with a hemocytometer and calculated as sperm/g testis parenchyma and spermitestis. The value of proliferation-specific TPS in testicular parenchyma was measured using a monoclonal antibody by an immunora- diometric assay (TPS-IRMA, Beki, Stockholm) as described by BLOTTNER et al. (1994) and was given in units/g testis (U/g). Apoptosis of testicular cells was measured in the same homogenised tissue samples according to the method of Hıncst et al. (1995) using a cell death detection ELISA (Boehringer-Mannheim) and expressed in units/mg tissue (U/mg). Testicular testosterone was measured by EIA as described by MEYER and Horr- MANN (1987). In roe deer the highest testicular and epididymal weights were found in the rutting period in late July and early August (27.2+8.6g and 3.3+0.8g, respectively). Gonadal size corresponded with the numbers of testicular spermatozoa/g parenchyma and sperma- tozoa per testis, respectively (Fig. 1). The most intensive apoptosis (60.4 + 34.2 U/mg tes- tis) was found during the period of testicular involution. The amount of testicular apoptosis was negatively correlated to gonadal weight (r = -0.5268, p < 0.05). The number of testicular spermatozoa/g parenchyma was highly correlated to testicular proliferation. Changes in apoptosis and proliferation showed inverse correlation. The highest testi- cular proliferation (108.7 +58.9 U/g testis) was found during the rutting period in early 60 Oona Hinssr, S. BLOTTNER, and H. H. D. MEYER g testis ; 100 Mio spermitestis (sS3} Buw/n) sısoydody V VI Vı vn IX X +testis weight +spermitestis $apoptosis Fig. 1. Testis weight, total testicular spermatozoa and the level of testicular apoptosis in roe deer during the transition from pre-rutting to post-rutting period (May-October). August coinciding with the lowest amount of apoptotic nucleosomes (13.7 # 1.2 U/mg tes- tis). In contrast, the activated apoptosis during the post-breeding period was associated with a low rate of proliferation (32.3 # 29.7 U/g testis). Differences in apoptosis and prolif- eration were significant between breeding and non-breeding periods (p<0.001 and p< 0.01, respectively). Testosterone production increased during the pre-rutting period and showed a peak in the first half of August (1.2 +0.6 ug/g testis). Its concentration dropped to 0.3 + 0.3 ug/g in September. Also in brown hare the testicular weight showed a cyclic involution and recrudes- cence. The regression of the testes in August/September is associated with the most active apoptosis and a diminished sperm production per gram of parenchyma and per testis (Bign2): The maxima and minima of testicular apoptosis and proliferation during involution and reactivation of testes were inversely related. The highest apoptotic level was found 15 12.5 (s4S3} Buuyn) SISOoYdody g testis ; 100 Mio spermitestis VERY RE a x +-testis weight +spermitestis $ apoptosis Fig. 2. Testis weight, total testicular spermatozoa and the level of testicular apoptosis in European brown hare during the transition from the breeding to non-breeding season and then again to the breed- ing season (June-February). Role of apoptosis in seasonal involution and recrudescence of testis 61 when proliferation was low (0.06 # 0.06 U/g testis; n = 10). Proliferation was newly acti- vated in November/December (0.85 + 0.34 U/g testis; n = 10), preceding the increase in testicular size. At this time, apoptosis showed low levels. The differences were highly sig- nificant between the compared stages for both parameters (p< 0.001). Testosterone level in July/August was 12.0 + 10.4 ng/g testis and 67.5 + 86.4. ng/g testis in November/Decem- ber (p< 0.05). The results as presented from two seasonal breeders with different periods of repro- duction within the annual cycle, indicate that apoptosis as an antagonist to proliferation plays an important role in seasonal regulation of testicular activity. The highest testoster- one concentrations occurred simultaneously with highest proliferation and lowest apopto- tic levels. This result is in accordance with the assumption that gonadotropins and androgens play an essential role in prevention of apoptosis in the testis (TAPANAINEN et al. 1993; TENNISWooD et al. 1992). Recent report from a tropical bat demonstrated also that the sexual dormancy is characterized by a marked apoptosis (OnyancGo et al. 1995). The counteraction of proliferation and apoptosis seems to be a general phenomenon in sea- sonal breeders. The characterisation of proliferation and apoptosis by simultaneous quan- tification could be a valuable approach to study seasonal changes on the gonadal level and its dependence on different conditions. Acknowledgement Parts of the presented study were supported by a grant of Deutsche Forschungsgemeinschaft. References BLOTTNER, S.; BLOTTNER, A.; HiINGsST, O.; SCHADOWw, D. (1994): Use of TPS as a proliferation marker in studies of mammalian spermatogenesis. In: Current tumor diagnosis: Applications, Research, Trends. Ed. by R. KLAappor München, Bern, Wien, New York: Zuckschwerdt-Verlag, Pp. 285-288. Hinssrt, O.; BLOTTNER, S. (1995): Quantification of apoptosis (programmed cell death) in mammalıan testes by DNA-Fragmentation ELISA. Theriogenology 44, 313-319. MEYER, H. H. D.; Horrmann, B. (1987): Development of a sensitive microtiterplate enzyme immunoas- say for the anabolic steroid trenbolone. Fd Add Contam 4, 149-160. OnyanGo, D. W.; GACHOKA, J. M.; OTIANGA-OWITI, G. E.; HENDRICKX, A. G. (1995): Seasonally depend- ent testicular apoptosis in the tropical Long-fingered bat (Miniopterus inflatus). Z. Säugetierkunde 60, 206-214. SCHWARTZMAN, R. A.; CIDLOWSKI, J. A. (1993): Apoptosis — the biochemistry and molecular biology of programmed cell death. Endocr. Rev. 14, 133-151. TAPANAINEN, J. S.;, TiLry, J. L.; ViHKo, K. K.; HsueH, A. W. J. (1993): Hormonal control of apoptotic cell death in the testis: gonadotropins and androgens as testicular survival factors. Mol. Endocrinol. 7, 643-650. TENNISWOOD, M. P.; GUENETTE, R. S.; LAKINS, J.;, MOOIBROEK, M.; Wong, P; WELSH, J. E. (1992): Active cell death in hormone-dependent tissue. Cancer Metastasis Rev. 11, 197-220. Authors’ address: OonA Hincst, Dr. STEFFEN BLOTTNER, Dr. H.H.D. Meyer, Institut für Zoo- und Wildtierforschung, PF 1103, D-10252 Berlin, Germany Z. Säugetierkunde 61 (1996) 62-64 ZEITSCH ER 3, © 1996 Gustav Fischer, Jena > — ” on SÄUGETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY MITTEILUNGEN DER GESELLSCHAFT Einladung Auf Einladung von Herrn Prof. Dr. D. KruskA und Herrn Prof. Dr. G. B. HARTL, Kiel, fin- det die 70. Jahrestagung der Deutschen Gesellschaft für Säugetierkunde e.V. von Sonn- tag, dem 22. September, bis Donnerstag, dem 26. September 1996, an der Christian- Albrechts-Universität zu Kiel statt. Sonntag, 22. September: Montag, 23. September: Dienstag, 24. September: Mittwoch, 25. September: Donnerstag, 26. September: Vorläufiges Programm Anreise 16.30 Uhr: Vorstandssitzung ab 19.00 Uhr: Zwangloser Begrüßungsabend im Hotel Kieler Yacht Club 09.00 Uhr: Grußworte und Eröffnung der Tagung durch den 1. Vorsitzenden 09.30 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt: „Evolution und Domestikation“ 13.30 Uhr: Posterdemonstration 14.30 Uhr: Kurzvorträge 16.30 Uhr: Mitgliederversammlung 20.00 Uhr: Empfang im Rathaus der Stadt Kiel 09.00 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt: „Populationsökologie“ 13.30 Uhr: Posterdemonstration 14.30 Uhr: Kurzvorträge 19.00 Uhr: Geselliger Abend 09.00 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt: „Säugetierschutz“ 13.30 Uhr: Posterdemonstration 14.30 Uhr: Kurzvorträge 16.30 Uhr: Posterprämierung 19.00 Uhr: Besuch des Zoologischen Museums der Universi- tät 09.00-17.00 Uhr: Exkursion ins Wikingermuseum Haithabu und nach Schleswig 09.00-12.00 Uhr: Symposium der „Koordinationsgruppe Fle- dermausschutz“/Workshops anderer Arbeitsgruppen Mitteilungen 63 Die Vorträge und Posterdemonstrationen finden statt in den Räumen des Biologie- Zentrums der Christian-Albrechts-Universität, Am Botanischen Garten 9, D-24118 Kiel. Alle Interessenten, Mitglieder und Nichtmitglieder, sind zu dieser Jahrestagung 1996 der Deutschen Gesellschaft für Säugetierkunde in Kiel herzlich eingeladen. Falls eine per- sönliche Einladung gewünscht wird, wenden Sie sich bitte an den 1. Vorsitzenden der Gesellschaft, Prof. Dr. U. SCHMIDT, Zoologisches Institut, Poppelsdorfer Schloß, D-53115 Bonn (Tel.: 0228/7354 68; Fax: 02 28/735458). Das Programm mit der Vortragsfolge wird den Mitgliedern - auf Anforderung auch Nichtmitgliedern - rechtzeitig vor der Tagung zugesandt. Wir bitten um Anmeldung von Tagungsbeiträgen. Außer Beiträgen zu den genannten Themenschwerpunkten werden auch dieses mal wieder Kurzvorträge und Posterpräsenta- tionen zu anderen Fachgebieten der Säugetierkunde angemessen berücksichtigt. Bitte melden Sie Kurzvorträge (15 Min. + 5 Min. Diskussion) sowie Posterpräsentatio- nen möglichst frühzeitig, spätestens jedoch bis zum 30. April (Ausschlußfrist) beim Geschäftsführer der DGS, Prof. Dr. H. ERKERT, Zoologisches Institut, Auf der Morgen- stelle 28, D-72076 Tübingen (Tel.: 07071/292958; Fax: 07071/294634) an. Der Anmel- dung ist eine maximal einseitige, informative Kurzfassung (1,5zeilig) beizufügen. Aus ihr sollen die Fragestellung, Methoden, Ergebnisse und die daraus gezogenen Schlußfolge- rungen hervorgehen. Alle Kurzfassungen werden wieder in einem Sonderheft der Zeit- schrift für Säugetierkunde publiziert. Sie sind nach folgendem, schon im letztjährigen Ab- stractheft eingeführten Schema abzufassen: Deutscher Titel, Leerzeile, englische Titelübersetzung (kleine Anfangsbuchstaben im Text; bitte ggf. einen „native speaker“ konsultieren), Leerzeile, Initialen und Familienname(n) des/der Autors(in) bzw. der Auto- ren(innen) in Großbuchstaben, Adresse, Leerzeile, Text (nicht formatiert!). Aus arbeits- ökonomischen Gründen bitten wir dringend darum, zusätzlich zu diesem ausgedruckten Abstract möglichst noch eine fehlerfreie Fassung auf Diskette (5,25 oder 3,5”, IBM kom- patibler PC, DOS oder Windows) in Form eines Word- (5,0 oder 5,5) oder ASCH-Files mitzuschicken. Bitte verwenden Sie als Filebezeichnung den eigenen Namen (Initialen und Familienname, z.B. KFISCHER.TXT/DOC). Die Maße für Poster werden im Juni- Rundschreiben der DGS bekanntgegeben. Mit Fragen zum Tagungsort und zur Organisation wenden Sie sich bitte an Herrn Prof. Dr. D. KruskA oder Herrn Prof. Dr. G. B. HARTL, Institut für Haustierkunde, Univer- sität Kiel, Olshausenstr. 40, D-24118 Kiel (Tel.: 0431/8804513 oder 0431/88045 07; Fax: 04 31/880 13 89). Mitteilung der „Koordinationsgruppe Fledermausschutz in Deutschland“ Die „Koordinationsgruppe Fledermausschutz in Deutschland“ beabsichtigt, im Rahmen des Schwerpunktthemas „Säugetierschutz‘“ auf der 70. Jahrestagung der DGS in Kiel wie- der einen „Fledermaustag“ zu gestalten. In Form von Kurzvorträgen und Postern sollen dabei aktuelle Ergebnisse der Fledermausforschung und des Fledermausschutzes darge- stellt werden. Für die Anmeldung von Beiträgen gelten die in der Einladung zur Jahres- hauptversammlung der DGS (s. 0.) genannten Bedingungen und Fristen. 64 Mitteilungen Einladung Auf Einladung von Herrn Prof. Dr. R. R. HormAnn, Berlin, findet vom 18. bis zum 21. September 1996 in Berlin das „1° International Symposium on Physiology and Etho- logy of Wild and Zoo Animals“ statt. Das Symposium wird vom Institut für Zoo- und Wildtierforschung (IZW), Berlin, in Kooperation mit der Deutschen Gesellschaft für Säu- getierkunde (DGS) und der European Association of Zoos and Aquaria (EAZA) ausge- richtet. Es soll sich schwerpunktmäßig mit den „größeren Säugetieren“ befassen. Vorläufiges Programm Mittwoch, 18. September: Anreise, Begrüßungsabend im Hotel „Berlin City Apart- ments, BCA“, Rhinstr. 159, D-10315 Berlin Donnerstag, 19. September: 09.00 Uhr: Eröffnung 09.30 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt „Reproduktionsbiologie“ 14.00 Uhr: Kurzvorträge und Posterdemonstration 1 Freitag, 20. September: 09.00 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt „Ethologie“ 14.00 Uhr: Kurzvorträge und Posterdemonstration 2 18.00 Uhr: Besichtigung des IZW 19.30 Uhr: Kongreß-Dinner (FH) Samstag, 21. September: 09.00 Uhr: Hauptvortrag und Kurzvorträge zum Themen- schwerpunkt „Ernährungsphysiologie“ 13.00 Uhr: Abschlußveranstaltung 14.30 Uhr: Workshops (,Noninvasive Monitoring“) Führung durch den Tierpark Die Vorträge, Workshops und Posterdemonstrationen finden im Gebäude der Verwal- tungsakademie, Alt-Friedrichsfelde 60, sowie im IZW, Alfred-Kowalke-Str. 17 in Berlin statt. Informationen über das „1“ International Symposium on Physiology and Ethology of Wild and Zoo Animals“: Dr. Franz KıımA, Organizing Committee, Institute for Zoo Biology and Wildlife Research (IZW), Alfred-Kowalke-Str. 17, D-10315 Berlin, Germany. Tel. 0(+49) 0 30-5 16 80; Fax: (+49) 0 30-5 12 61 04. Redaktionelle Hinweise Einsendung und Annahme von Manuskripten: Ms. zur Veröffentlichung sind einzusenden an den 1. Schriftleiter: Herrn Prof. Dr. Dieter Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel. Über die Annahme von Ms. zur Veröffentlichung wird gemäß Geschäftsordnung der Deutschen Gesellschaft für Säugetierkunde entschieden. Der Eingang von Ms. wird sofort bestätigt, und nach Erhalt der Gutachten wer- den die Autoren über die Entscheidung umgehend informiert. Schriftwechsel, der die technische Drucklegung betrifft, ist zu richten an: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena. Außere Form der Manuskripte: Die Ms. sind 3fach und vollständig als Original mit 2 Kopien in deutscher oder englischer Sprache einzureichen. Sie müssen mit Schreibmaschine oder als deutlich lesbare Computerausdrucke auf DIN A4-Blättern (21x 29,7 cm), mit einem linken Heftrand von 4 cm und durchgehend mit einem doppelten Zeilenabstand geschrieben sein. Jedes Ms. soll auf der 1. Seite folgende Angaben enthalten: Titel der Studie; Name(n) der Autoren (männliche Vornamen nur in Initialen, weibliche ausgeschrieben); Institution(en), an der (denen) die Studie durchgeführt wurde. Ferner soll bei längeren Titeln ein kurzer Kolumnentitel (Seitenüber- schrift) von maximal 72 Anschlägen angegeben werden. Die Schrifttypen müssen einheitlich sein, ohne halbfette, gesperrte, kursive etc. Hervorhebungen. Fußnoten auf den Seiten oder Appendices am Ende sind nicht zugelassen. Autorennamen im Zusammenhang mit Literatur- zitaten sind im Text mit Bleistift zu unterstreichen. Wissenschaftliche zoologische und botanische Gattungs-, Art- und Unterartnamen sind mit einer Schlangenlinie zu unterstreichen. Der Name von Erstbeschreibern bei wis- senschaftlichen Namen wird nicht unterstrichen. Im Ms. ist am Rand mit Bleistift zu vermerken, an welcher Stelle die einzelnen Abbildungen und Tabellen ein- gefügt werden sollen. Am Ende der Studie, im Anschluß an das Literaturverzeichnis, sind die genauen postalischen Adressen des (der) Autoren anzugeben. Umfang und Aufbau der Manuskripte: Manuskripte können als Originalarbeiten oder als wissenschaftliche Kurz- mitteilungen veröffentlicht werden. a) Originalarbeiten: Originalarbeiten sollen einschließlich Abbildungen, Tabellen und Literaturverzeichnis einen Gesamtumfang von 30 Ms.-Seiten nicht überschreiten. Der Text soll eine klare Gliederung aufweisen in: Abstract (in englischer Sprache), Einleitung, Material und Methode, Ergebnisse, Diskussion (oder zusammengefaßt als Er- gebnisse und Diskussion), Danksagungen, Zusammenfassung und Literatur. Bei deutschsprachigen Studien ist über dem englischen Abstract der Titel in englischer Sprache anzugeben. Bei englischsprachigen Studien ist entsprechend über der deutschen Zusammenfassung der Titel in deutscher Sprache anzugeben. b) Wissenschaftliche Kurzmitteilungen: Wissenschaftliche Kurzmitteilungen sollen einen Gesamtumfang von 5 Ms.-Seiten nicht überschreiten. Sie enthalten kein Abstract und keine Zusammenfassung. Bei deutschsprachigen Ms. ist jedoch auf der 1. Seite zusätzlich eine englische Übersetzung des Titels anzugeben, bei englischsprachigen entsprechend eine deutsche Übersetzung. Wissenschaftliche Kurzmitteilungen sind äußerlich nicht in die Kapitel Einleitung, Material und Methode, Ergebnisse und Diskussion gegliedert, müssen aber in der textlichen Darstel- lung diese Strukturierung erkennen lassen. Abbildungen: Anzahl und Größe der Abbildungen sind unbedingt auf das notwendige Mindestmaß zu beschrän- ken. Es können nur kontrastreiche, reproduktionsfähige Originale verwendet werden, farbige Abbildungen wer- den nicht angenommen. Beschriftungen müssen einheitlich und wegen späterer Verkleinerung groß und kräftig sein. Auf der Rückseite sind laufende Numerierung, Zeitschrift und Autorennamen anzugeben. Alle Bildunter- schriften sollen auf einem gesonderten Blatt gelistet werden. Tabellen: Tabellen müssen auf ein Mindestmaß beschränkt werden. Jede Tabelle muß auf einem gesonderten DIN A4-Blatt angelegt sowie mit Überschrift und laufender Numerierung versehen sein. Umfangreichere Tabel- len müssen auf Folgeblättern als Fortsetzung erscheinen. Literaturverzeichnis: Jede zitierte Arbeit ist in alphabetischer Reihenfolge aufzuführen. Es soll nur international erreichbare Literatur zitiert werden. Die Zeitschriften werden mit ihrem Kurztitel angegeben. Beispiele: a) Zeitschriftenbeiträge: KALko, E.K. V; HANDLEY, C. O. Jr. (1994): Evolution, biogeography, and description af a new species of Fruit-eat- ing bat, genus Artibeus Leach (1821), from Panamä. Z. Säugetierkunde 59, 257-273. b) Bücher: HERRE, W.; Rönrs, M. (1990): Haustiere — zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) Handbuchbeiträge: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) -— Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. by J. NIETHAMMER und F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28. Drucklegung: Der jeweilige Hauptautor erhält sowohl die Andrucke der Bilder zur Prüfung wie auch den komplet- ten Umbruch seines Artikels zusammen mit dem Ms. Es dürfen nur tatsächliche Satzfehler korrigiert werden. Vom Ms. abweichende Änderungen werden dem Autor in Rechnung gestellt. Es sind die üblichen Korrekturzeichen zu verwenden. Ein als druckfertig bezeichneter Abzug ist zu senden an: Herrn Prof. Dr. Peter Langer, Institut für Anatomie und Zytobiologie, Justus-Liebig-Universität, Aulweg 123, D-35385 Giessen. Sonderdrucke: Verfasser von Originalbeiträgen und Wissenschaftlichen Kurzmitteilungen erhalten 50 Sonder- drucke gratis. Mehrbedarf kann gegen Berechnung bei Rücksendung des Umbruchs bestellt werden. Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective, or for resale. For copying from back volumes of this journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 100537, D-07705 Jena; Telefon (03641) 6263, Telefax (03641) 62 65 00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A. Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zeitschriftenvertrieb, Frau B. Dressler, Post- fach 10 05 37, D-07705 Jena; Telefon (03641) 62 64 44, Telefax (03641) 62 65.00. ‘* For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 42882 01. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2945,-/382,50 (plus postage). Single issue price (DM/ÖSI/SFr): 80,-/592,-/77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 86010090; Deutsche Bank, Konto- Nr. 3907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400 00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 u " "FÜR SAUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Banaszek, Agata; Ratkiewicz, M.; Fedyk, S.; Szataj, K. A.; Chetnicki, W.: The chromosomes and isoenzymes in margin- al populations of the Common shrew (Sorex araneus) in the Vistula delta. - Die Chromosomen und Isoenzyme in Randpopulationen der Waldspitzmaus (Sorex araneus) im Weichsel-Delta ..................ussssesssesssenseeenennenen 65 Turni, H.; Müller, E. F.: Unterscheidung der Spitzmausarten Sorex araneus L., 1758 und Sorex coronatus Millet, 1828 mit Hilfe einer neuen Diskriminanzfunktion. — Discrimination of the shrew species Sorex araneus L., 1758 and Be eor0natas Niillet, 1828 by help of a new discriminance function .......--.-:...-zuunus0sr20200 000000 mansnane nee 73 Ivanitskaya, Elena; Shenbrot, G.; Nevo, E.: Crocidura ramona sp. nov. (Insectivora, Soricidae): a new species of shrew from the central Negev desert, Israel. - Crocidura ramona sp. nov. (Insectivora, Soricidae): eine neue Spitz- By eeatalieu der Negev Wüste, Istael 2.2.2 2n...nssnendnensen se nasnansensenenönnarnnnennn hend 3 Tiedemann, R.; Harder, J.; Gmeiner, Christine; Haase, E.: Mitochondrial DNA sequence patterns of Harbour por- poises (Phocoena phocoena) from the North and the Baltic Sea. - Untersuchungen an der mitochondrialen DNA von Schweinswalen (Phocoena phocoena) aus Nord- und Ostsee ............uu..-eseeseensnesennnnnnnnennnnnnnnenennnnnn 104 Korz, V.; Hendrichs, H.; Militzer, K.: Behavioural and anatomical correlates of sympathetic arousal and stress in male Central American Agoutis (Dasyprocta punctata). — Verhaltens- und anatomische Korrelate zur Sympathikusakti- vität bei männlichen Mittelamerikanischen Agutis (Dasyprocta punctata) ...........ususeressssneeseeseenneneen en 12 LE De Er BRÜSSEL BEPEERLLPEIBERREL FELL RELRELSETE TER EL rer 126 Indexed in Current Contents “Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database CSUSTAV ss 0044-3468 FISCHER 1.215" £ BONIS ARTIBUS SEMPER JENA «-STUTTGART-NEW YORK April 1996 _ co OÖ) AT NT Dt LIN 23 : SÄUGETIERKUNDE OF MAMMALIAN BIOLOGY ER, Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kıel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel — R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel — M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt — P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) - W. Herre, Kiel (1962-1966) — H. Frick, München (1972-1976) -— M. Röhrs, Hannover (1977-1981) — H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amtierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H. G. Erkert, Tübingen —- W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel -— Marialuise Kühnrich, Hamburg Z. Säugetierkunde 61 (1996) 65-72 © 1996 Gustav Fischer, Jena INTERNATIONAL JOU On OF MAMMALIAN ap Sy 17 1906 \ The chromosomes and isoenzymes in marginal populati ofthe Common shrew (Sorex araneus) in the Vistula delta By AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. Szaras, and W. CHErTNIckI Institute of Biology, Warsaw University Bialystok Branch, Biatystok, Poland Receipt of Ms. 21. 11. 1995 Acceptance of Ms. 28. 11. 1995 Abstract The chromosomes and isoenzymes of the common shrew populations in the Vistula delta were studied. Twenty nıne out of 33 shrews were characterised by the karyotype XX/XYıY>, af, bc, yl, hi, g, k, m, n, o, PD, 9, r, tu. We suggest that they be described as the Nogat race. The occasional occurrence of ko and gm metacentrics is the result of introgression from the neighbouring races inhabiting adjacent areas. Of 22 loci scored, six appeared to be polymorphic. The studied area was homogenous with respect to protein variation. The hypothesis of the presence of a gentle cline of Pgm-3° allele frequency in the hybrid zone Druzno/Legucki Miyn was not supported. The frequencies of esterase (Est-1) and manno- sephosphate isomerase (Mpi) fit closely to the pattern of variation for Poland. Introduction The karyotype of the common shrew (Sorex araneus L., 1758) shows intra- and interpopu- lation variation based mainly on centric fusions. The species is subdivided into a number of chromosome races, which differ with respect to their metacentric composition or the number of metacentrics (WöjJcık 1993; BRÜNNER 1991). The eastern border of West European Karyological Group (WEKG), according to HAuSSER et al. (1994), runs through Poland and within its range a few local chromosome races differing only in number of metacentrics were described. Central Poland is inhab- ited by monomorphic populations of Stobnica race with metacentrics jl, hi, ko, gm, np (Wörcık 1993; FEDyK et al. 1993). This monomorphic centre is surrounded by an area of polymorphism for ko, gm, np metacentrics and this variation is of clinal nature. The individual metacentric clines do not coincide. The frequencies of np metacentric are the first to decay and the cline of np metacentric frequency is the narrowest north- wards from the monomorphic centre. The populations characterized by the four remain- ing metacentrics jl, hi, ko, gm were described as Laska race (I) (Wöscık 1986; FEDyK and LEnıec 1987, Szaraı et al. 1995). Further north, along the eastern border of WEKG range, the gm metacentric frequencies decay gradually and the shrews with metacentrics jl, hi, ko were described as Druzno race (WöjJcık and FEpykK 1985). East of this area, the frequencies of metacentrics hi, ko of Druzno race form narrow clines in the hybrid zone between Druzno and Legucki Miyn race (BAnAszEk 1994). The Legucki Miyn race is characterised by metacentrics jl, hk, io, gr, mn (FEpyk and Lenıec 1987; Wöscık 1993). In north-western Poland the cline of gm metacentric is wider than that of ko arm com- bination. In consequence, the Polish sea coast is inhabited by populations of the Ulm race with the metacentrics jl, hi, gm (Wöscık 1986, 1993). 66 AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. SzAra7, and W. CHETNICKI The distance between the closest known sites of Druzno and Ulm races is about 130 km. It is not known whether the range of ko and gm metacentrics overlap in the area between those points. We have studied chromosomally part of Vistula delta with the aim to describe the local geographic distribution of these metacentrics (Fig. 1). The studies of karyotype variation were supplemented by the electrophoretical analysis. It is known that genic differences between chromosome races of the common shrew are small (FryKMman et al. 1983; SEARLE 1985; CATZEFLIS 1984). Clines of allele frequencies between the races were detected only for Mpi locus in northern Sweden (FRYKMAN et al. 1983). RATKIEWICZ et al. (1994) suggested also the presence of a gentle cline of the Pgm- 3B allele frequency in the Druzno/Legucki Mlyn hybrid zone in northern Poland. How- ever the transect studied through the zone was rather short, 18.6 km only. The sample from the Vistula delta can be treated as the westward extension of the transect (Fig. 1) and the hypothesis of the presence of an allele frequency cline can be tested. Material and methods Common shrews were collected during 1993 and 1994 from five sites in Zulawy Wielkie. The area is part of Vistula River delta, situated between the Nogat and Vistula Rivers (Fig. 1). The delta was inhab- ited by shrews rather late, probably in historical times. The karyotypes of thirty three common shrews were determined (Tab. 1). Chromosome prepara- tions were made in the field by the standard method from spleen (FEpyK 1980). The slides were tryp- sinized and stained with Giemsa for G-bands (SEABRIGHT 1971). Chromosome arms were labelled according to the nomenclature of SEARLE et al. (1991). For electrophoretic studies we used 50 individuals from four chromosomally studied sites. The karyotyped shrews were all included in the sample, except for one individual from Dabrowa. Horizontal Eu u Fig. 1. The distribution of the studied sites; L I - Lubstowo I, L IH - Lubstowo II, R - Rybina, K - Kepki, D - Dabrowa, closed triangle - Krzewsk population, stripped area - the hybrid zone Druzno/Legucki Mtyn, closed square - Wezina population, thick line - the geographical border of the Vistula delta. Chromosomes and isoenzymes of Sorex araneus 67 starch electrophoresis of kidney homogenates was carried out as described in HArrıs and HoPKINSON (1976). Buffers and staining procedures were done according to SELANDER et al. (1971), HARRIS and Hopkinson (1976) and Quavi and Kır (1980). Fourteen protein systems coded by 22 presumptive loci were screened. A locus was considered poly- morphic when more than one allele was detected in at least one individual. Alleles were designated with letters of Latin alphabet according to the relative mobility of corresponding bands on the gel (RATKIEWICZ et al. 1994). Results Karyotypes Individuals with the autosome number 2Na = 26 homozygous with respect to arm combi- nations jl, hi were the most common in all studied populations (Tab. 1, Fig. 2). Three jl heterozygotes with 2Na = 27 were also found in Lubstowo I (Tab. 1). The polymorphism of jl pair was restricted to that one population only. Table 1. Material used for chromosome studies. LI - Lubstowo I, L II - Lubstowo II, R- Rybina, K - Kepki, D - Dabrowa. Karyotype Sample sıze jl jl, hihi, iV-, hihi, il jl, hihi, il jl, hihi, il jl, hihi, Fig. 2. A karyotype of shrew from the Nogat race. 68 AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. SzArAJ, and W. CHETNICKI Metacentric ko occurred in three individuals. We found two heterozygotes, one in Lubstowo II and one in Rybina, and one metacentric homozygote in Lubstowo I (Tab. 1). The frequency of metacentric ko was 0.08 ın Lubstowo I and II and 0.07 in Rybina. One metacentric gm was found in Lubstowo II (Tab. 1), which gives the frequency 0.08 of this metacentric in the population. The other chromosome arms of the variable part of the karyotype occurred as acrocentrics in all studied shrews. Electrophoresis At the 16 following locı only single alleles were detected: malate dehydrogenase (Mdh-1, Mdh-2) 6-phosphogluconate dehydrogenase (6-Pgd), a-glycerophosphate dehydrogenase (a-Gpd-1l, a-Gpd-2), lactate dehydrogenase (Ldh-1, Ldh-2), carbonate dehydrogenase (Est-D), glucosephosphate isomerase (Gpi), superoxide dismutase (Sod-1, Sod-2, Sod-3), glutamate oxalate transaminase (Got-1, Got-2), malic enzyme (Me-2) and isocitrate dehy- drogenase (Idh-1). The sıx remaining loci appeared to be polymorphic (27.3%): isocitrate dehydrogenase (Idh-2), phosphoglucomutase (Pgm-1, Pgm-3), mannosephosphate isomerase (Mpi), ami- noacylase (Acy) and esterase (Est-1). Allelic frequencies at the polymorphic loci are listed in table 2. There were no significant deviations from the Hardy-Weinberg equili- brium in the populations (P > 0.05). At the Idh-2 locus in three of the analyzed sites we found two alleles but only the Idh- 2° allele was detected in Lubstowo II. On the other hand the frequency of the Idh-2* Table 2. Allele frequencies at the polymorphic locı in the populations studied and in the pooled sam- ple; in parentheses: number of specimens scored for each locus. Pooled sample Population EIN R 0.00(14) 0.96 0.04 0.75(14) 0.25 0.00(14) 0.61 0.39 0.00 0.14(14) 0.75 0.11 0.73(13) 0.27 0.00(7) 0.93 0.07 0.71(7) 0.29 0.07(7) 0.79 0.14 0.00 0.14(7) 0.72 0.14 0.57(7) 0.43 0.07(14) 0.89 0.04 0.75(12) 0.25 0.07(14) 0.68 0.21 0.04 0.11(14) 0.85 0.04 0.61(14) 0.39 0.00(15) 1.00 0.00 0.89(14) 0.11 0.00(15) 0.73 0.27 0.00 0.14(14) 0.82 0.04 0.69(16) 0.31 0.03(50) 0.94 0.03 0.79(48) 0.21 0.03(50) 0.69 0.27 0.01 0.13(49) 0.80 0.07 0.66(50) 0.34 * the allele Mpi“ corresponds with the allele Mpi“ in Wöscık and Wörcık (1994), allele Mpi” with Mpi“, Mpi“ with Mpi®. We do not known if the rare allele Mpi” corresponds with the allele Mpi” in Wöscık and Wöjcık (1994). ** the alleles notations are compatible with those in Wörcık and Wörcık (1994). Chromosomes and isoenzymes of Sorex araneus 69 allele in the neighbouring population Lubstowo I reached 0.32. Lubstowo II population differs significantly with respect to allelic frequencies at Idh-2 locus from any other popu- lation (Fisher’s Exact Probability Test P< 0.05). Three alleles were found at the Pgm-1 locus in the studied area. Only one Pgm-1® al- lele was Seelen in Kepki population. It was the common allel in all samples studied. The rare Pgm-1“ allel was also found in all samples except Kepki, while the rare allel Pgm-1” was detected only in Rybina. Those ‚a0 rare alleles were only found in a hetero- zygous state Pgm-1”/Pgm-1”, Pgm-1"/Pgm-1* The other four loci showed polymorphism in all Benulebons studied. Four alleles were recorded at the Mpi locus. The alleles Mpi" and Mpi” were detected in all samples and the Mpi” allel was the commonest. The rare allel Mpi* was found in two populations Ry- bina and Lubstowo II, while the Mpi’ was found in Rybina only. Overall five genotypes were recorded: Mpi"/Mpi", Mpi"/Mpi“, Mpi“/Mpi“, Mpi /Mpi“, Mpi"/Mpi”. We found three alleles at the Acy locus. The Acy” allele was the commonest in all samples. Two locı Est-] and Pgm-3 were clearly polymorphic in all populations. Two alleles were detected at both locıi. Discussion Out of 33 shrews karyotyped, 29 had in their karyotypes jl and hi metacentrics only. The frequencies of the metacentrics ko and gm were low and did not exceed the value 0.08 in either population. The shrews with the karyotype XX/XYıY>, af bc, VI, hi, g, k, m, n, 0, p, g, r, tu were thus classıfied as a distinct chromosome race, which we named Nogat race after the river running close to the sampling sites. The type locality of the race is in Kepki (19°19’ E; 54°12’ N). The occeurrence of ko and gm metacentrics in the area inhabited by Nogat race is the result of introgression. Metacentrics ko were probably derived from the east from the Druzno race. The frequencies of ko in Krzewsk population, which is the type locality of Druzno race, reach the value 0.55 (WöJcık and FEDYK 1985), and in Lubstowo I and II si- tuated west of Nogat - 0.08. We do not know the frequencies of these metacentrics on the eastern banks of Nogat. It is probable that the river forms the barrier to chromosome flow and sets the eastern border of Nogat race range. On the other hand the frequency cline of ko may be very sharp and the metacentric frequency may be low on both banks of the river. An abrupt decrease in ko frequencies was found on the eastern side of Druzno Lake, where the ko frequency was 0.33 in the Wezina population (BANASZEK 1994) while in the Krzewsk population on the western side of Druzno Lake it reached the value 0.55 (WöjJcık and FEDYK 1985). Metacentric gm was found in only one population in the eastern part of the study area and its frequency reached 0.08. This metacentric might have been introgressed not from the west but rather from the southern populations of Laska race along the Vistula River. We do not know the distribution of this metacentric west of Vistula River. We suppose that Vistula forms a strong barrier similar to that of the Nogat River and that this limits the range of Nogat race to Zulawy Wielkie. However, the arm combination gm is poly- morphic in the populations of Laska and Ulm races in north-western Poland (Wöjcık 1986, 1993). It is possible (although less likely) that the cline of gm frequencies is narrow in northern Poland and the metacentric does not occur in some areas west of Vistula River. If this holds the occurrence of the shrews of Nogat race would be also expected there. The populations of Nogat race are situated at the north-eastern edge of the continu- ous range of WEKG in continental Europe. The number of autosomes in this race (2Na = 26 - 27) is the highest yet reported in Poland. These data support the hypothesis 70 AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. SzAraAJ, and W. CHETNICKI of prevalence of acrocentric chromosomes in marginal populations (Zıma et al. 1994). The karyotype of Nogat race is probably ancestral for the WEKG. It is interesting that shrews with ancestral karyotypes are found in the areas of pre- sumptive pleistocene refugia and also in the areas most distant from them (Zıma et al. 1994). It seems plausible that the rate of chromosome evolution was the quickest in the centre of the migration wave. In the refugia, however, as it was suggested by Zıma et al. (1994), and also in the forefront of the migration wave, the chromosomal evolution was slow or completely absent. Such hypothesis was also put forward by Wöhcık (1993) in his model of chromosomal evolution in the shrews in Central Europe. He argued that centric fusions were formed somewhere within the species range during post-glacial expansion and later extended their ranges. The electrophoretic study has revealed that Zulawy Wielkie area is rather homoge- nous with respect to protein variation. The only statistically significant difference was the absence of the Idh-2® allele in Lubstowo II. However, we treat this difference with great caution as ıt can be probably attrıbuted to sampling error given the small sample size and the probability of close relations between individuals, since six of them were caught in the same {rap. At the eastern border of Vistula River delta the hybrid zone between Druzno and Legucki Mliyn race was localized (BAnASZEK 1994). Six populations from this hybrid zone were studied electrophoretically (RATKIEwIcZ et al. 1994). The loci Pgm-1 and Idh-2, poly- morphic in the present material, were described as monomorphic in this hybrid zone (RATKIEwIczZ et al. 1994). The description was erroneous, caused by the difficulties in reading zymograms. After the reinterpretation of the data, we found rare heterozygotes at the Pgm-1 locus similarly to this in the present material (RATKIEWICZ unpubl. data). We detected also two alleles at the Idh-2 locus in the hybrid zone. The differences between the samples from Zulawy and the hybrid zone in allele frequencies at this locus were not significant (RATKIEWICZ, unpubl. data). We compared the allelic frequencies at the Pgm-3 and Est-1 locı, which were clearly polymorphic in Zulawy and Druzno/Legucki Mlyn hybrid zone area. We found two al- leles at the Pgm-3 locus in the Nogat race sample. Two rare alleles at this locus, which were detected in the shrews of Legucki Mlyn race, were neither found in Druzno race in the hybrid zone (RATKIEWIcCZ et al. 1994), nor in the Nogat race. Moreover, RATKIEWICZ et al. (1994) suggested the presence of the gentle cline for the Pgm-3” allele frequency through the hybrid zone. The frequency of the Pgm-3" allele decreased westwards from 0.81 to 0.62. But in the westernmost population the tendency was reversed and the fre- quency of this allele slightly increased to 0.73 (RAtkIEwicz et al. 1994). The extension of the transect westwards to Zulawy, where the Pgm-3 allele frequency was 0.79, confirmed the reversal of the tendency. The hypothesis of the presence of the gentle cline of Pgm-3" allele frequency in the hybrid zone Druzno/Legucki Mlyn is therefore not supported by the present data. The slight decrease in the Pgm-3° allele frequency and corresponding in- crease in the Pgm-3“ allele frequency can be rather observed in the centre of the hybrid zone. At the Est-1 locus we found two alleles. Two alleles were also detected in the Druzno/ Legucki Mliyn hybrid zone (RArkızwicz et al. 1994). The rare Est-1© allele was detected by Wöscık and Wörcık (1994) in the samples of Bialowieza and Drnholec races. But this allele did not occur in the sample of Stobnica race, which is phylogenetically close to the Druzno and Nogat races. | Wöscık and Wöscık (1994) noted that the allele Est-1” frequency increases westwards in Poland. The frequency of the Est-1” allele was 0.66 in Nogat race sample. It therefore seems that the frequency of this allele increases in the south-west direction from the low- est values in Bialowieza and Nogat race to the maximum in Drnholec race. Slightly high- er frequencies of the Est-1” allele were found in the centre of the Druzno/kLegucki Miyn Chromosomes and isoenzymes of Sorex araneus zu hybrid zone. But it is possible that higher frequencies of the Est-1” allele were accumu- lated in the hybrid zone, whereas in the border populations of the transect the frequency was 0.5 on Druzno race side and 0.54 on Legucki Mlyn race side (RATKIEWICZ et al. 1994). The frequencies of the Mpi” and Mpi“ alleles form clines ranging from Sweden in the north to Poland in the south (Frykman et al. 1983; Wöscık and Wörcık 1994). The fre- quencies of Mpi” -0.69 and Mpi“ -0.27 in the total sample from Zulawy fit closely to that pattern of variation. Those frequencies are very similar to the frequencies found in the geographically closest populations of Stobnica race -0.67 and 0.24, respectively (Wöscık and Wörcık 1994). Acknowledgements We wish to thank Dr. M. KonArzEewskı and Dr.J. B.SEARLE for improving our English and M. Sc. A. MiısHtA for her help in the field work. This study was supported by the Committee of Scienti- fic Research within the 6 P204. 03405 grant. Zusammenfassung Die Chromosomen und Isoenzyme in Randpopulationen der Waldspitzmaus (Sorex araneus) im Weichsel-Delta Die Chromosomen und Isoenzyme von Waldspitzmauspopulationen in einem Teil des Vistula-Deltas wurden untersucht. Von insgesamt 33 Spitzmäusen waren 29 durch den Karyotyp XX/XY}Y>, af, bc, y, hi, g, k, m, n, 0, p, q, r, tu charakterisiert. Wir schlagen vor, dessen Träger als „Nogat-Rasse‘“ zu bezeich- nen. Das gelegentliche Vorkommen von metazentrischen ko- und gm-Chromosomen ist das Ergebnis einer Introgression von Rassen aus benachbarten Lebensräumen. Von 22 untersuchten Genloci zeigten sechs einen Polymorphismus. Im Hinblick auf die Proteinvariation erwies sich das Untersuchungsgebiet als homogen. Die Hypothese einer geringfügig klinalen Verteilung der Frequenz des Allels Pgm-3” in der Hybridzone Druzno/Legucki Miyn wird durch unsere Daten nicht gestützt. Die Allelfrequenzen an den Loci Est-1 und Mpi stimmen gut mit der generellen Verteilung dieser Allele über Polen überein. References BANASZEK, A. (1994): The chromosomal structure of the hybrid zone between Druzno and Legucki Miyn races in the common shrew (Sorex araneus) in north-eastern Poland — preliminary results. Fo- lia Zool. 43 (Suppl. 1), 11-20. BRÜNNER, H. (1991): The karyology of the common shrew, S. araneus L., 1758 (Insectivora, Soricidae) in Southwestern Germany. Z. Zool. Syst. Evolut.-forsch. 29, 73-81. CATZEELIS, F. (1984): Systematique biochimique, taxonomie et phylogenie des musaraignes d’Europe (Soricidae, Mammalia). Universit& de Lausanne. PhD thesis. FEDYK, S. (1980): Chromosome polymorphism in a population of Sorex araneus L. at Bialowieza. Folıa Biol. 28, 83-120. FEDYK, S.; LENIEc, H. (1987): Genetic differentiation of Polish populations of Sorex araneus L. I. Varia- bility of autosome arm combinations. Folia Biol. 35, 57-68. FEDYK, S.; ZAJKOWSKA, M.; CHETNICKI, W. (1993): Study of a contact between two chromosomally mono- morphic races of Sorex araneus L. (common shrew). Heredity 71, 221-226. FRYKMAN, 1.; SIMONSEN, V.; BENGTSSON, B. O. (1983): Genetic differentiation in Sorex. I. Electrophoretic analysis of the karyotypic races of Sorex araneus in Sweden. Hereditas 99, 2792922. HARRIS, H.; Hopkınson, D. A. (1976): Handbook of enzyme electrophoresis in human genetics. Oxford: North-Holland. HAUSSER, J.; FEDYK, S.; FREDGA, K.; SEARLE, J. B.; VOLOBOUEY, V.; WÖSCIK, J. M.; ZıMaA, J. (1994): Defini- tion and nomenclature of the chromosome races of Sorex araneus. Folia Zool. 43 (Suppl. 1), 1-9. 12 AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. SzAra), and W. CHETNICKI Quavı, H.; Kır, S. (1980): Electrophoretic patterns of aminoacylase-1 (Acy-1) activity. Biochem. Genet- ics 18, 669-679. RATKEWICZ, M. A.; BANASZEK, A.; LOBODZINSKA, J. (1994): Isoenzyme variation in the common shrew (Sorex araneus) from the hybrid zone between the chromosomal races Druzno and Legucki Miyn: preliminary results. Folia Zool. 43 (Suppl. 1), 21-28. SEABRIGHT, M. (1971): A rapid banding technique for human chromosomes. Lancet 2, 971-972. SEARLE, J. B. (1985): Isoenzyme variation in the common shrew (Sorex araneus) in Britain, in relation to karyotype. Heredity 55, 175-180. SEARLE, J. B.; FEDYK, S.; FREDGA, K.; HAUSSER, J.; VOLOBOUEY, V. T. (1991): Nomenclature for the chro- mosomes of the common shrew (Sorex araneus). Mem. Soc. vaud. Sci. nat. 19, 13-22. SELANDER, R.K.; SMITH, M. H.; Yang, S. Y.; JOHNSON, W. E.; GENTRY, J. B: (1971): Biochemical poly- morphism and systematics in the genus Peromyscus. I. Variation in the old-field mouse (Peromyscus polionotus). Studies in Genetics VI, Univ. Texas Publ. 7103, 49-90. SzArAJ, K. A.; FEDYK, S.; BANASZEK, A.; CHETNICKI, W. (1995): Maximization of the frequency of recom- binants in the hybrid zone of Sorex araneus in northern Poland. Acta theriol. 40, 225-236. WöjJcık, J. M. (1986): Karyotypic races of the common shrew (Sorex araneus L.) from northern Poland. Experientia 42, 960-962. WöjcıkK, J. M. (1993): Chromosome races of the common shrew Sorex araneus in Poland: a model of karyotypic evolution. Acta theriol. 38, 315-338. Wöl1cıkK, J.; FEDYK, S. (1985): A new chromosome race of Sorex araneus L. from Northern Poland. Ex- perientia 41, 750-752. Wöj1cık, J. M.; Wöscık, A.M. (1994): Protein variation in the common shrew (Sorex araneus L.) in Po- land, in relation to karyotype. Folia Zool. 43 (Suppl. 1), 53-61. ZIMA, J.; MACHOLAN, M.; FiLipuccı, M. G.; REITER, A.; ANDREAS, M.; LiPA, M.; KRYSTUFEK, B. (1994): Karyotypic and biochemical status of certain marginal populations of Sorex araneus. Folia Zool. 43 (Suppl. 1), 43-51. Authors’ address: AGATA BANASZEK, M. RATKIEWICZ, S. FEDYK, K. A. SZALAJ, W. CHETNICKI, Institute of Biology, Warsaw University Bialystok Branch, Swierkowa 20B, 15-950 Bialys- tok, Poland Z. Säugetierkunde 61 (1996) 73-92 ZEITSCHRI FTSE "FÜR © 1996 Gustav Fischer, Jena SÄUGETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Unterscheidung der Spitzmausarten Sorex araneus L., 1758 und Sorex coronatus Millet, 1828 mit Hilfe einer neuen Diskriminanzfunktion Von H. Turnı und E. F. MÜLLER Zoologisches Institut, Abt. Physiologische Ökologie, Universität Tübingen, Tübingen Eingang des Ms. 01. 12. 1995 Annahme des Ms. 12. 12. 1995 Abstract Discrimination of the shrew species Sorex araneus L., 1758 and Sorex coronatus Millet, 1828 by help of a new discriminance function. Described is a new function to discriminate Sorex araneus from $. coronatus on the basis of skull charac- teristics. The new function is based upon the measurement of four easily obtainable distances, which hardly correlate with each other and usually are present in material from owl pellets. Among 114 bio- chemically (PAGE) determined specimens the new discriminant function assigned 100% correctly, com- pared to 92,1% with the function developed by HAusser and Jammor (1974) and 89,81% with the index x5/x6 (HANDWERK 1987). In order to be able to discriminate between uncompletely conserved skulls in- dices were determined for skull and mandibles. Using the new discriminant function as well as additional indices and characteristics material from collections and owl pellets from almost all over Baden-Würt- temberg was investigated (ca. 4000 Sorex araneus/coronatus). A comparative study based upon suitable distances showed that there exist geoclimatic variations: e.g. individuals from populations in the Alpine foreland are bigger on average than those from other climatic regions. Among different climatic regions interspecific variations are usually larger than intraspecific variations. However, with regard to the con- stancy of the new discriminant function the variations are neglectable because overlapping of single dis- tances remains small. In sympatric populations a character displacement could not be found. Einleitung Eine sichere Unterscheidung der sehr nahe verwandten Arten Sorex araneus und Sorex coronatus ist gegenwärtig nur anhand ihrer Karyotypen oder mit biochemischen Metho- den möglich, z.B. mit Hilfe der Polyacrylamidgel-Elektrophorese (PAGE) des Serumpro- teins Albumin (HAusser und ZUBER 1983) — mit Ausnahme der südwestalpinen Chromo- somenrasse S. araneus „Valais“ (HAusser et al. 1991) — sowie der Elektrophorese des Urin-Pepsins (NEET und HAusser 1991). Für eine flächendeckende Erfassung der Verbreitung beider Arten können diese Methoden jedoch kaum herangezogen werden, da mit ihnen ein erheblicher Zeit- und Materialaufwand verbunden ist. Es wurden daher zur Arttrennung geeignete morpholo- gische Merkmale gesucht. Befunde in Nordrhein-Westfalen, wonach sich beide Arten nach der Fellfärbung gut unterscheiden lassen (von LEHMANN 1955; OLERT 1973, HUTTERER und VIERHAUS 1984), konnte NEET (1992) für die Schweiz nicht bestätigen. HAUSSER und Jammor (1974) entwickelten eine Diskriminanzfunktion für 4 Mandibelmeßstrecken, die 95,3% des überwiegend aus der Schweiz stammenden Materials richtig zuordnete. Bei der Überprüfung dieser Trennfunktion an europäischem Material (LocH 1977; HAusser 1984; 74 H. Turnı und E. F MÜLLER Mys et al. 1985) wurde jedoch festgestellt, daß die Mandibelmeßstrecken erheblich geo- graphisch variieren. HAussEr et al. (1990) schlugen aus diesem Grund vor, die Konstante der Trennfunktion jeweils geographisch anzupassen. Ein von HANDWERK (1987) ermittel- ter Trennindex sowie Merkmale anderer Autoren (u.a. PIEPER 1978; von BüLow 1989) bedürfen noch der Überprüfung an umfangreichem, cytologisch oder biochemisch gesi- chertem Material. Im Rahmen der vorliegenden Untersuchung sollte zunächst die Zuverlässigkeit der Methoden und Merkmale zur morphologischen Unterscheidung von S. araneus und 5. coronatus überprüft werden. Dazu wurden biochemisch sicher bestimmte Tiere aus ganz Baden-Württemberg herangezogen. Darauf aufbauend sollte schließlich eine verbes- serte Trennmethode entwickelt werden, mit welcher umfangreiches Material aus Gewöl- len und Sammlungen in Baden-Württemberg determiniert werden kann. Material und Methoden Für die vorliegende Untersuchung standen 114 biochemisch (PAGE) sicher identifzierte Tiere aus ganz Baden-Württemberg zur Verfügung. 23 S. araneus und 5 S. coronatus stammten aus dem Süd- und Nordschwarzwald (BRÜNNER 1988, 1990), 10 $. araneus und 7 S. coronatus aus dem Naturpark Schön- buch bei Tübingen (KuLZER et al. 1993) und 43 $. araneus und 26 S. coronatus aus den übrigen Landes- teilen Baden-Württembergs (TURNI und SCHÖNHERR 1994). Die Schädel wurden mit Hilfe von Speckkäferlarven gesäubert. An den 114 PAGE-Tieren und an weiteren 4000 Schädeln aus ganz Baden-Württemberg (Samm- lung Rosensteinmuseum Stuttgart, Gewöllfunde aus dem Projekt „Wildlebende Säugetiere in Baden- Württemberg“ sowie eigene Gewöll- und Totfunde) erfolgte die Überprüfung aller in der Literatur vor- geschlagener Trennmethoden und Merkmale. Hierzu wurde eine Präzisions-Schieblehre mit digitaler Anzeige (Meßgenauigkeit: 0,01 mm) und eine Stereolupe mit Meßokular benutzt. Gerätebedingte und beim Messen der Objekte unvermeidlich auftretende Meßfehler wurden dadurch vermindert, daß für jede Meßstrecke nur der Mittelwert aus mehreren Meßvorgängen verwendet wurde. Folgende Meßstrecken, Merkmale und Trennmethoden wurden auf ihre Zuverlässigkeit überprüft: — Die Diskriminanzfunktion für die 4 Mandibelmeßstrecken a, f, y und ö von HAUSSER und JAMMOT (1974). Beide Autoren führten noch folgende Hilfsmerkmale an: — Die Fossa temporalis interna ist bei $. araneus geschlossen dreieckig, bei $. coronatus offen eiför- mig (Abb. 1). — Der Processus coronoideus verläuft bei S. araneus gerade oder leicht nach hinten geneigt, bei S. coronatus hingegen eher nach vorn (Abb. 2). — Das Foramen mandibularis tritt oft doppelt bei S$. araneus auf, meist bis zur Mitte des Unter- randes der Fossa temporalis interna reichend; bei S. coronatus einfach und im hinteren Teil unter- halb der Fossa temporalis interna. — Das Foramen mentale liegt bei $. araneus unter der vorderen Hälfte des M,, bei S. coronatus un- ter der Mitte desM.. -— Der untere Incisivus verläuft bei $. araneus als horizontale Verlängerung der Mandibel, bei 5. coronatus steigt er leicht an. — von BüLow (1989) erwähnt zwei buccale Knochenwülste am Kopf des Processus coronoideus, die bei $. araneus weiter auseinander liegen als bei $. coronatus (Abb. 2). — Auf die Caudalfläche des Processus articularis machte erstmals PıEPEr (1978) aufmerksam. Dem- nach weist S. araneus einen relativ niedrigeren und an der Basis breiteren Condylus auf als 5. coronatus (Abb. 3). HANDWERK (1987) nahm am Condylus 4 Meßstrecken (X18, X19, X20, X21) und trennte mit dem Index X21/X19 93,4% seines Materials. Er benutzte für seine Messungen u.a. eine Schieblehre, die auf 0,1 mm Dicke zurückgeschliffen wurde. Für seine Meßstrecke X21 ist eine korrekte Messung anders kaum möglich; vermutlich wurde sie deshalb in der Praxis (vgl. BERGER et al. 1992) durch die buccale Artikularhöhe ersetzt. Unterscheidung der Spitzmausarten S. araneus und S. coronatus 19 In der vorliegenden Arbeit wurde jedoch auch die buccale Artikularhöhe verworfen, da sich eine einheitliche Meßweise nicht ergab und die Abweichung aus jeweils mehreren Messungen zu groß war. Deshalb wurde die Qualität der gut meßbaren Strecke X18 geprüft. Für X18 und X19 werden im folgen- den die Synonyme AH und AB (Abb. 3) verwendet. — Zusätzlich gemessen wurden die Coronoidhöhe (Corh) und die Breite der knöchernen Verbindung zwischen der Incisura sigmoidea und der Vertiefung oberhalb des Foramen mandibularis (IV) (Abb. 4). — Handwerk (1987) konnte auch Unterschiede am Oberkiefer herausarbeiten: Abb. 1. Form der Fossa temporalis interna bei S. coronatus (oben) und $. araneus (unten) Das Foramen magnum ist bei S. araneus eng und spitz, bei S$. coronatus weit und rund. Der Processus zygomaticus ist bei $. araneus relativ kurz, bei S. coronatus relativ lang. Der Abstand der Foramina vas- cularıa war an seinem Material bei S. araneus durchschnittlich enger als bei S. coronatus. Aus dem Abstand zwischen den Postglenoidalfortsätzen (X5) und der Zygomaticus-Breite (X6) er- stellte HANDWERK (1987) den In- dex X5/X6. Bei Index-Werten über 1,0 könne von S. araneus, bei Werten unter 1,0 von S. coronatus ausgegangen Wer- den. Mit diesem Index trennte er 90,1% seiner Tiere. Auch diese Merkmale entbehrten der Bestätigung durch anderes, sicher determiniertes Material und wurden deshalb in der vorliegenden Arbeit untersucht. Die berücksichtigten Oberkiefer- Meßstrecken (Abb. 5) waren: BR... Abb. 2. Neigung des Processus coronoideus sowie der Ab- stand der buccalen Knochenwülste bei S. coronatus (oben) und S. araneus (unten) Abb. 3. Meßstrecken AH und AB 76 H. Turnı und E. F MÜLLER Cbl (Condylobasallänge), FV (Abstand der Foramina vascularia, Lochmitten), PB (Abstand der Post- glenoidalfortsätze = X5 bei HANDWERK 1987) sowie Zyg (Zygomaticusbreite = X6 bei HANDWERK 1987). Corh Abb. 4. Meßstrecken Corh und IV Cbl Abb. 5. Meßstrecken Cbl, PB, FV und Zyg (Schädel von $. coronatus) Statistische Methoden Der Umfang biochemisch (PAGE) eindeutig identifizierter Individuen (n = 114) blieb etwas zu gering, um eine aussagekräftige biometrische Analyse durchzuführen. Aus diesem Grund erfolgte zunächst eine vorläufige Artzuordnung an 347 Individuen aus den zentralen Landesteilen mit Hilfe eines Kom- plexes ausgewählter Trennmethoden und Merkmale, für welche It. Literatur zumindest regional, z.T. aber auch überregional eine hohe Zuverlässigkeit angenommen wurde: — Diskriminanzfunktion von HAUSSER und JAMMoT (1974) (unter dem Vorbehalt, daß die Konstante C regional angepaßt werden muß) Neigung des Processus coronoideus (HAUSSER und JAMMoT 1974) Form der Fossa temporalis interna (HAusser und JAmMmoT 1974) Caudalfläche des Processus articularis (PIEPER 1978) Index X5/X6 (HANDWERK 1987) relativer Abstand der Foramina vascularia (HANDWERK 1987) Dh Nach dieser Zuordnung lagen 197 Sorex araneus und 150 Sorex coronatus vor. An ihnen erfolgten die ersten Messungen und statistischen Auswertungen. Die sehr umfangreiche Diskriminanzanalyse und der Kolmogorov-Smirnov-Iest wurden mit Hilfe des Programmes SPSS/PC* durchgeführt. Unterscheidung der Spitzmausarten S. araneus und S. coronatus en LiNDER und BERCHTHOLD (1982) geben zu bedenken, daß die Daten in einer Diskriminanzanalyse doppelt verwendet werden: Zunächst wird mit den Daten eine Diskriminanzfunktion erstellt, um dann am gleichen Datenmaterial diese Trennfunktion zu überprüfen. Die Folge ist oft ein zu optimistisches Resultat. Deshalb wurde vorgeschlagen, die Überprüfung der ermittelten Trennfunktion an anderem Material vorzunehmen. In der vorliegenden Untersuchung wurde so verfahren, daß für die Diskrimi- nanzanalyse die 347 Individuen aus den zentralen Landesteilen herangezogen wurden. Die erste Überprüfung der berechneten Diskriminanzfunktion erfolgte am gleichen Material, eine zweite an den 114 PAGE-Tieren. Ergebnisse Die Lage- und Streuungswerte beider Verteilungen waren für jede Meßstrecke im Erwar- tungsbereich der Literaturangaben. Tabelle 1. Die ermittelten Lage- und Streuungszahlen einzelner Schädelmaße. Meßstrecke Sorex araneus Sorex coronatus n = 197 ne #190 min — max min - max 18,795 + 0,354 5,538 + 0,149 1,269 + 0,086 2,181 + 0,085 1,345 + 0,064 1,037 # 0,069 6,982 + 0,199 15199=20:095 122 E01 0,314 # 0,105 17,49 - 19,60 5,19 - 5,93 0,98-1,53 1,98 - 2,38 1,17 -1,49 0,86 - 1,21 6,35 - 7,51 1,00 - 1,49 1,54 - 2,45 0,00 - 0,58 18,370 + 0,287 5,304 + 0,133 1,487 # 0,088 2,306 # 0,082 1,206 # 0,058 1,100 # 0,072 6,652 + 0,166 PPSIESN0N 2,124=204155 0,132 # 0,086 17,51-19,10 4,96 - 5,61 1,24 - 1,70 2110222153 1,07 - 1,38 0,96 - 1,27 6,14 - 7,01 1,08 - 1,58 1,78 - 2,49 -0,06 - 0,33 „Kolmogorov-Smirnov-Test (K-S-Test) für die Güte der Anpassung“ und t-Test Mit dem K-S-Test für die Güte der Anpassung (SacHs 1992) wurde die Wahrscheinlich- keit p für eine Normalverteilung jeder einzelnen Meßstrecke berechnet. Daraus ging her- vor, daß die Annahme einer Normalverteilung für die Meßstrecken IV bei S. coronatus und für £ bei S. araneus am vorliegenden Material unwahrscheinlich ist und verworfen werden muß. Normalverteilung, nicht zu kleine Stichprobenumfänge und gleiche Stan- dardabweichungen waren Voraussetzung für den t-Test (LoRENnZ 1988). Mit Ausnahme von IV und £ erfüllten sämtliche Meßstrecken die Voraussetzungen für den t-Iest und zeigten dort jeweils hochsignifikante Unterschiede. Andere Meßstrecken Zuweilen war der Processus zygomaticus abgebrochen und eine korrekte Meßweise von Zyg nicht mehr möglich. Zyg wurde im folgenden nur noch für die Überprüfung des „HANDWERK-Indexes“ X5/X6 herangezogen. Die Coronoidhöhe (Corh) zeigte zwar im t- Test einen signifikanten Unterschied, der Überschneidungsbereich beider Verteilungen war jedoch so groß, daß diese Meßstrecke für weitere Berechnungen zur Trennung beider Arten keine Verwendung fand. Andere Meßstrecken wurden verworfen, weil selbst bei größeren Stichproben nur geringe Unterschiede auftraten, oder weil eine einheitliche Meßmethode nicht möglich war. 78 H. Turnı und E. F MÜLLER Lineare Abhängigkeit der einzelnen Meßstrecken untereinander Als Maß für die lineare Abhängigkeit zwischen zwei Parametern gilt der Maßkorrela- tionskoeffizient r (Lorenz 1988). In den Tabellen 2 und 3 sind die Koeffzienten für jedes Meßstreckenpaar angegeben. Tabelle 2. Maßkorrelationskoeffizient r bei Sorex araneus 0,3478 — 0,2579 0,0572 0,4905 0,0256 0,5534 _ 0,1453 0,0448 0,3314 0,2812 — 0,2549 0,0308 0,2719 0,1901 0,2649 _ 0,1323 0,0309 0,4201 0,1461 -0,014 -0,021 0,2245 0,0699 0,0819 0,1319 0,0642 0,3574 -0,346 0,0701 -0,048 -0,033 -0,017 -0,072 0,0323 0,2449 -0,216 Tabelle 3. Maßkorrelationskoeffizient r bei Sorex coronatus 0,1921 0,1775 0,5683 0,0991 0,4418 0,1401 - 0,0127 720 330, 02 PZN 259 — 0,0993 7.0.3341 70.251727 .021457. 0.018 -0,078 0,0472 -0,058 -0,073 0,1788 -734 - -0,107 -0,099 -0,062 -0,099 0,0367 0,1077 -0,172 Mys et al. (1985) übernehmen für die Meßstreckenpaare Thorpe’s Limit |r| > 0,7, d.h. alle Paare, die diesen Wert übersteigen, korrelieren zu stark, weshalb dann jeweils eine der beiden Meßstrecken verworfen werden sollte. Eine leichte Korrelation bestand zwischen den Paaren Cbl/a und AH/AB, jedoch er- reichten auch sie die kritische Grenze nicht, womit kein Anlaß zur Ablehnung vorlag. Diskriminanzanalyse Zunächst wurde eine Diskriminanzfunktion (DF) für 9 Meßstrecken (PB, FV, AH, AB, IV, a, ß, y und ö) ermittelt. Weitere Analysen und Überprüfungen erlaubten das Weglas- sen der Meßstrecke IV. Keine Einbußen in der Trennwirkung hatte schließlich der Ver- zicht auf die Mandibelmeßstrecken a, ß, y und ö. Mit diesem Schritt erreichte die neue DF ihre Unabhängigkeit von jener, welche HAusser und JAMmMoT (1974) entwickelten. | Die neue Diskriminanzfunktion lautet: Z = 2,961 (PB) + 7,040 (FV) + 9,163 (AH) - 11,854 (AB) + C; C = 0,9085 Z<0 — Sorex araneus, Z>O > Sorex coronatus Unterscheidung der Spitzmausarten $. araneus und S. coronatus 79 Bei C = 0,9085 (die Konstante wurde für den Scheidewert Null angeglichen) hatten die Mittelwerte den Betrag +2,5148; s (S. araneus) = 1,045 und s (5. coronatus) = 0,938. Der Unterschied war hochsignifikant (t-Test). Die erste Überprüfung dieser neuen DF am gleichen Material ergab eine „richtige“ Zuordnung von 99,42% aller Individuen (345/347). Alle weiteren Kombinationen mit zu- sätzlichen Meßstrecken brachten keine Verbesserung der Trennwirkung. Anzumerken ist, daß die Artvorgabe für alle Individuen durch den Komplex von Merkmalen und Trennmethoden (s. o.) bereits mit einer Restunsicherheit erfolgte. Für die beiden „falsch“ zugeordneten Individuen berechnete das Programm eine hohe Wahr- scheinlichkeit für eine andere Artzugehörigkeit. Am selben Material zeigt die Zuordnung durch die Diskriminanzfunktion von HaAus- SER und JAMMoT (1974) mit den Mandibelmeßstrecken a, ß, y und ö einen größeren Überschneidungsbereich. Zunächst fand die in HAusser et al. (1990) angegebene Kon- stante mit dem Wert C = 8,1598 Verwendung. Die unterschiedlichen Beträge der Mittel- werte beider Verteilungen erforderten für Baden-Württemberg jedoch eine Anpassung der Konstante C. Die Optimierung der Trennung wurde durch eine Nullpunktverschie- bung erreicht, wodurch sich mit C = 8,3098 eine maximale „richtige“ Zuordnung von 317/ 347 Individuen = 91,35% ergab. Der Betrag der Mittelwerte beider Verteilungen war: 22570. Artdiagnose an unvollständigen Schädeln Schädel aus Eulengewöllen oder von auf andere Weise beschädigten Tieren sind zuweilen so unvollständig, daß entweder nur die Mandibel oder nur Oberschädelteile erhalten ge- blieben sind. Damit auch solche Stücke bestimmt werden können, wurden entsprechende Indices ermittelt. Am Oberschädel wurden die Meßstrecken PB und FV gegeneinander abgetragen. Aus der Punktwolke (Abb. 6) geht bereits eine deutliche Trennung hervor. Für beide Vertei- lungen wurden Konfidenzintervalle errechnet (Sachs 1992) (Tab. 4). S.coronatus FV A 4 S.araneus 1,00 + —— — ———— — 4,90 5,00 5,10 5,20 5,30 5,40 5,50 5,60 5,70 5,80 5,90 6,00 6,10 PB Abb. 6. Punktwolken-Diagramm für die Meßstrecken PB und FV 80 H. Turnı und E. F. MÜLLER Tabelle 4. Werte für den Index PB/FV 99%-Schranken Sorex araneus 197 3,75 - 5,07 4,38 09275 3,66 - 5,09 Sorex coronatus 150 3,15 - 4,08 ID 0,199 3,05 — 4,09 Tabelle 5. Werte für den Index AH/AB 99%-Schranken Sorex araneus 197 15471582 1,62 0,068 AS 880 Sorex coronatus 150 1.742218 1,91 0,075 1,72-2,11 1,60 1,50 1,40 S.coronatus AB 1,30 4 S.araneus 1,00 — 4 4 43 1,90 2,00 2,10 2,20 2,30 2,40 2,50 2,60 AH Abb. 7. Punktwolken-Diagramm für die Meßstrecken AH und AB Aus der Überschneidung beider Intervalle konnte folgende Trennung festgelegt werden: Index PB/FV > 4,09 > Sorex araneus Index PB/FV < 3,66 —> Sorex coronatus Auf die gleiche Weise wurde mit den Mandibel-Meßstrecken AH und AB verfahren. Auch hier ergab sich eine deutliche Trennung beider Verteilungen (Tab. 5, Abb. I): Aus der Überschneidung beider Intervalle ergab sich folgende Trennung: Index AH/AB < 1,72 > Sorex araneus Index AH/AB > 1,80 > Sorex coronatus Endgültige Überprüfung der Trennmethoden Aus der biometrischen Analyse ging eine neue Diskriminanzfunktion für die Meßstrecken PB, FV, AH und AB hervor. Eine Zuordnung durch diese Trennfunktion am gleichen Ma- Unterscheidung der Spitzmausarten S$. araneus und S. coronatus 81 terial (347 Tieren aus den zentralen Landesteilen) deutete auf eine sehr große Zuverläs- sigkeit (99,42%) hin. Dieses erfreuliche Resultat bedurfte aber erst der Bestätigung an anderem Material, wofür die 114 durch PAGE identifizierten Individuen (76 S. araneus und 38 $. coronatus) herangezogen wurden. Diese 114 PAGE-Tiere erschienen auf Grund ihres Umfanges und ihrer Verteilung über 51 Meßtischblätter Baden-Württembergs hin- reichend geeignet, die Ergebnisse der biometrischen Analyse sowie die wesentlichsten Trennmethoden zu überprüfen. Diskriminanzfunktionen Die neu ermittelte Diskriminanzfunktion trennte 113/113 (= 100%) PAGE-Individuen richtig (Tab. 6). Auf Grund der mit der biometrischen Analyse weitgehend übereinstim- menden Werte und Verteilungen bestand keine Notwendigkeit, diese Diskriminanzfunk- tion zu verändern. Die Verteilung der beiden Arten durch die neue DF wurde in einem Säulendiagramm dargestellt (Abb. 8). Tabelle 6. Überprüfung der neuen Diskriminanzfunktion durch PAGE-Tiere (Meßstrecken PB, FV, AH und AB) Sorex araneus Sorex coronatus gesamt 1 38 —2,494 2,361 S 0,983 0,837 min — max 4,997 - 0,318 0,561 - 4,197 richtig** 755 38 in % 100 100 * Ein Individuum konnte nicht berücksichtigt werden, da der Oberkiefer stark zerstört war. ** auch Werte <|0,1| galten als nicht zuverlässig getrennt Anzahl der Individuen E S.coronatus MI S.araneus | Abb. 8. Verteilung der beiden Arten durch die neu ermittelte Diskriminanzfunktion (für die Meß- strecken PB, FV, AH und AB) 82 H. Turnı und E. F. MÜLLER Die DF von HAusser und JAmMoT (1974) mit der angepaßten Konstanten C = 8,3098 ordnete 105/114 (= 92,10%) richtig zu (Tab. 7), allerdings lagen noch weitere Werte mit knapp >|0,1| in der Nähe des Nullpunktes. Insgesamt besitzt auch diese DF in Baden- Württemberg eine hohe Zuverlässigkeit (Abb. 9). Tabelle 7. Überprüfung der DF von HaAusser und JamMmor (1974) durch PAGE-Tiere (Meßstrecken ©, ß, yund ö) mit Konstante C = 8,3098. Sorex araneus Sorex coronatus gesamt 76 38 1,429 -1,413 S 1,011 1,069 min - max 0,365 - 4,453 -3,601 - 0,692 richtig** 71 34 in % 93,42 89,47 ** auch Werte < 0,1| galten als nicht zuverlässig getrennt Anzahl der Individuen en N en & =) DD > UIELIERTTITTIITENTEN IT = I E SE -S685 = 0 = EN =i =TIEB>TER—TER—TEB— Da en Sl en Ze en PET a ES Fe el | E S.araneus S.coronatus | Abb. 9. Verteilung der beiden Arten durch die Diskriminanzfunktion von HAUSSER und JAMMOoT (1974) (Meßstrecken a, ß, y und ö) Konstante C = 8,3098 Der Handwerk-Index X5/X6 Es wurde bereits darauf hingewiesen, daß der Processus zygomaticus durch die Präpara- tion oder in Eulengewöllen zuweilen abgebrochen vorliegt, und dann eine korrekte Mes- sung der Meßstrecke Zyg (= X6 bei HANDWERK 1987) kaum möglich ist. So konnten auch nur 108 von 114 PAGE-Tieren für eine Zuordnung mit dem „HANDWERK-Index“ X5/X6 herangezogen werden. Mit 97/108 (= 89,81%) richtig bestimmten Individuen erwies sich dieser Index als recht zuverlässig. Unterscheidung der Spitzmausarten S. araneus und S. coronatus 83 Hilfsmerkmale Zur Überprüfung der Hilfsmerkmale wurden zusätzlich zu den 114 PAGE-Tieren etwa 4000 Schädel aus ganz Baden-Württemberg herangezogen. Da diese Hilfsmerkmale me- trisch kaum faßbar sind, unterliegen sie stets der subjektiven Beurteilung. Unter diesem Vorbehalt ließen sich nur die folgenden Hilfsmerkmale als geeignet bestätigen: a) die Neigung des Processus coronoideus (Abb. 2), b) der Abstand der buccalen Wülste am Kopf des Processus coronoideus (Abb. 2) und c) die Form der Fossa temporalis interna (Abb. 1) Balgmaterial Aus dem Vergleich des vorhandenen Balgmaterials mit der jeweiligen Artdiagnose durch PAGE bzw. nach Schädelmerkmalen ging hervor, daß sich zwar Tendenzen, jedoch keine zuverlässigen Unterschiede in der Fellfärbung beobachten lassen. Die für $. araneus ty- pische dreistufige Färbung mit breiter Schabracke konnte — auch bei adulten Tieren - nicht so häufig beobachtet werden wie erwartet. Es traten zweifarbige und einfarbige Tiere sowie Individuen mit sehr schmaler Schabracke auf. Bei $. coronatus stellte sich zwar öfter als bei S. araneus die sehr schmale Schabracke ein, jedoch konnten auch Indivi- duen mit breiter Schabracke und sogar einheitlich braun gefärbte Tiere beobachtet wer- den. Neben den typisch zweifarbigen traten auch dreifarbige $. coronatus auf. Tendenziell waren Schabrackenspitzmäuse etwas heller gefärbt, doch variiert die gesamte Färbung mit dem Lebensraum. Geographische Schwankungen einzelner Meßstrecken in Baden-Württemberg HaAusser (1984) ermittelte mit Hilfe statistischer Methoden die individuelle, geoklima- tische und genetische Varianz dieser Arten und konnte zeigen, daß die interspezifische morphologische Varianz wesentlich stärker auf geoklimatische Einflüsse als auf gene- tische Ursachen zurückzuführen ist. Die neu ermittelte Diskriminanzfunktion basiert nun auf Daten, die an Schädeln aus den zentralen Teilen Baden-Württembergs erhoben wurden. An den Meßstrecken PB, FV, AH, AB und Cbl sollte deshalb geklärt werden, ob a) sich die Schädelproportionen in Baden-Württemberg verändern, b) geoklimatisch bedingte intra- und interspezifische Unterschiede erkennbar sind und c) die Qualität der neuen Irennfunktion durch möglicherweise zu große Schwankungen der Meßstrecken beeinträchtigt wird. Für den Vergleich wurden Populationen beider Arten aus größeren, klimatisch homo- genen Regionen untersucht. Hierzu wurde Baden-Württemberg in 11 Klimabezirke unter- teilt, als Grundlage dienten die Klimabezirke aus dem Klima-Atlas von Baden-Württem- berg (1954) (Abb. 10). Für die einzelnen Klimabezirke wurden folgende Abkürzungen verwendet: TÜB - Zentrale Landesteile des Oberen Neckarlandes (Zentrum: Tübingen) BRACK = Kraichgau und Neckarbecken (Zentrum: Brackenheim) MURR = Bauland und Schwäbische Waldberge (Zentrum: Murrhardt) ALB = Schwäbische Alb SCHWARZ = Schwarzwald so = Südliches Oberrhein-Tiefland BODEN = Bodensee-Hügelland und Schwäbisches Alpenvorland DONAU = Donau-lller-Lech-Platten NO = Nördliches Oberrhein-Tiefland ODENW = Odenwald MAINFR = Mainfranken und Mittelfranken 84 H. Turnı und E. F MÜLLER Klime-Atlas von Beden-Württemberg 8 o Höhenstufen 4) III 8 N I) SEIN A SS u DAN o1 D 4 IF VENS I KR NN « v D HIRTEN of SEK RR ZIEL, ER Ss SS NN ER S N ER IN NA IE "= ug, N ! Se < X DER Entwurf: Dr. M. Manig Druck: Wir Größchen KG Dortmund Abb. 10. Klimabezirke in Baden-Württemberg Es gingen nur solche Individuen in die Berechnung ein, die aus möglichst zentralen Bereichen der jeweiligen Klimabezirke stammten. Zur Klärung der intraspezifischen Un- terschiede wurde für jede Meßstrecke jeweils die Abweichung [in %] des Mittelwertes eines Klimabezirkes (X;) von jenem der zentralen Region (Xrvg) ermittelt. Als Ausdruck des interspezifischen Verhältnisses galt der Quotient X;saraneus/Xis.coronams (= %)- Auch hier wurde jeweils die Abweichung [in %] des Quotienten a; vom Quotienten der Zentra- len Population (Opus) errechnet und in Tab. 12 zusammengestellt. Zur graphischen Ver- anschaulichung dieser Unterschiede wurden „Box-and-Whisker-Plots“ (LoRENZ 1988) Unterscheidung der Spitzmausarten S. araneus und $. coronatus 85 erstellt. Die „Box“ enthält die mittleren 50% aller Werte um den Median, die „Whiskers“ entstehen durch die Extremwerte. Condylobasallänge (Cbl) Für eine vergleichende Untersuchung der Meßstrecke Cbl war die teilweise geringe An- zahl vollständig erhaltener Schädel begrenzender Faktor. Einstweilen können folgende Tendenzen festgehalten werden: S. araneus aus der Südlichen Oberrheinebene (SO) und die Populationen beider Arten aus dem Alpenvorland sind auffällig größer als in anderen Regionen Baden-Württembergs. Beide Arten zeigten kein „Süd-Nord-Gefälle“ für diese Meßstrecke, sondern geoklimatische Unterschiede. Meßstrecken PB, FV, AH und AB Die verschiedenen Box-and-Whisker-Plots (Abb. 11 bis 14) veranschaulichen, daß die ein- zelnen Meßstrecken zwischen den Klimabezirken variieren. Zunächst mußte nun unter- sucht werden, wie stark die Meßstrecken in den einzelnen Klimaregionen im Verhältnis zur zentralen Region (TÜB) intraspezifisch abweichen (Tab. 8 bis 11). Tabelle 8. Abweichung der X; von Xrüg der Meßstrecke PB [in %]. BRACK MURR ALB SCHWARZ SO BODEN DONAU NO ODENW MAINFR 2 Sorex 0,03 0,45 0,86 0,02 2,0 araneus Sorex 0,39 0,54 0,02 0,43 { 0,45 0,09 coronatus 2,83 1,04 0,77 0,95 0,52 Tabelle 9. Abweichung der X; von Xrüg der Meßstrecke FV [in %]. BRACK MURR ALIB SCHWARZ SO BODEN DONAU NO ODENW MAINFR Sorex —2,12 2,52 -133 -0,86 2,44 267 -1,81 24 -165 -1,97 araneus Sorex 0,06 1,54 -0,06 -1,48 —2,69 -0,33 0,13 coronatus Tabelle 10. Abweichung der X; von Xrüg der Meßstrecke AH [in %]. BRACK MURR ALB SCHWARZ SO BODEN DONAU NO ODENW MAINFR 3 Sorex -1,79 -0,46 1,42 -1,23 1,8 0,32 -0,55 0,23 0,04 -0,36 araneus Sorex —2,29 -0,73 0,39 0,21 - 1,82 - - -1,47 0,56 coronatus Tabelle 11. Abweichung der X; von Xrüg der Meßstrecke AB [in %]. BRACK MURR ALB SCHWARZ SO BODEN DONAU NO ODENW MAINFR Sorex —2,08 -1,11 0,74 3,34 0,67 -0,07 -1,04 -0,81 0,15 -1,26 araneus Sorex —2,41 1247 23516 0,99 — 3.31 = = 0,41 0,33 coronalus 86 H. Turnı und E. F MÜLLER Abweichungen der interspezifischen Unterschiede in den verschiedenen Klimaregionen Da nur Individuen aus möglichst zentralen Bereichen jedes Klimabezirkes herangezogen wurden und aus manchen Regionen ohnehin wenige Nachweise von S. coronatus vorla- gen, war der Stichprobenumfang in einigen Klimabezirken relativ gering, teilweise nicht ausreichend. Unter diesem Vorbehalt müssen die folgenden Ergebnisse eher als Tendenz betrachtet werden. Tabelle 12. Abweichung der Quotienten a; vom Quotienten drug [in %]. BRACK MURR ALIB SCHWARZ SO BODEN DONAU NO ODENWMAINFR -0,09 0,86 0,48 0,48 0,38 3,94 -1,21 0,68 -1,26 -1,78 0,32 1,07 -1,41 1,58 0,24 —2,31 -0,39 4,27 0,24 -1,56 246 222 63 100 39 75 132 225 121 66 75 (n) Sorex araneus 210 50 34 121 61 (-) 24 (-) (-) 30 17 (n) Sorex coronatus 6,1 6,0 5,9 5,8 5,7 5,6 SS) 5,4 313 3,72 SH 5,0 4,9 TÜB BRACK MURR ALB SCHWARZ so BODEN DONAU NO ODENW MAINFR Abb. 11. Box-and-Whisker-Plots für PB in den einzelnen Klimaregionen Diskussion Biometrische Analyse Im vorliegenden Fall mußte zunächst die Behauptung, daß durch einen Komplex von Trennmethoden und Merkmalen eine Arttrennung vorgenommen werden kann, überprüft werden. Zu diesem Zweck wurden 347 Individuen untersucht und es können folgende Aussagen festgehalten werden: Unterscheidung der Spitzmausarten S. araneus und S. coronatus 87 210 54 34 119 61 (-) 22 (-) (-) 30 17 (n) Sorex coronatus 240 219 61 93 36 74 132 221 118 64 73 (n) Sorex araneus TÜB BRACK _MURR ALB SCHWARZ SO BODEN DONAU NO ODENW MAINFR Abb. 12. Box-and-Whisker-Plots für FV in den einzelnen Klimaregionen 150 57 3] 27 7 (-) 26 (-) (-) 33 20 (n) Sorex coronatus 197 288 60 114 44 93 140 247 129 71 83 _(n) Sorex araneus TÜB BRACK MURR ALB SCHWARZ so BODEN DONAU NO ODENW MAINFR Abb. 13. Box-and-Whisker-Plots für AH in den einzelnen Klimaregionen 88 H. Turnı und E. F MÜLLER 197 288 61 116 42 93 140 246 129 71 82 (n) Sorex araneus 150 57 32 126 70 (-) 26 (-) (-) 33 20 (m) Sorex coronatus TÜB BRACK MURR ALB SCHWARZ so BODEN DONAU NO ODENW MAINFR Abb. 14. Box-and-Whisker-Plots für AB in den einzelnen Klimaregionen Die Lage- und Streuungswerte beider Verteilungen waren für jede Meßstrecke im Er- wartungsbereich (Vergleich mit Literaturangaben). Die einzelnen Meßstrecken waren jeweils hochsignifikant unterschiedlich. Bei der Diskriminanzanalyse gingen von jedem einzelnen Objekt mehrere Parameter gleichzeitig in die Berechnung ein. Mit der berechneten Diskriminanzfunktion wurden die Artvorgaben überprüft und 345/347 Individuen bestätigt. Auch diese DF war hochsignifi- kant unterschiedlich. Durch die biometrische Analyse konnte also gezeigt werden, daß sich der Komplex dieser Trennmethoden und Merkmale mit großer Wahrscheinlichkeit zur Arttrennung eignet. Darüber hinaus konnte eine neue Diskriminanzfunktion für die Meßstrecken PB, FV, AH und AB entwickelt werden. Diese Meßstrecken sind untereinander nur gering korreliert, linear unabhängig von den Meßstrecken a, ß, y und ö (und somit auch unabhän- gig von der durch HAusser und Jammor (1974) entwickelten Diskriminanzfunktion) und fast immer an Gewöllematerial vorhanden. Ein weiterer Vorteil der neuen DF besteht darin, daß die Meßstrecken PB, FV, AH und AB auch ohne aufwendige Meßvorrichtung problemlos meßbar sind. Bei der von HAussErR und JAaMmMoT (1974) entwickelten DF muß dagegen für die Mandibel- meßstrecken a, ß, y und ö eine Stereolupe mit Meßokular (bzw. Zeichenspiegel oder Bildprojektor) verwendet werden. Die nicht immer einheitliche Bildebene bereitet je- doch meßmethodische Schwierigkeiten. Bereits Mys et al. (1985) machten auf teilweise beachtliche Schwankungen der Meßwerte für diese Mandibelmeßstrecken aufmerksam. Da sich auch in der vorliegenden Untersuchung diese Schwankungen trotz sorgfältiger Meßweise nicht vermeiden ließen, galten Funktionswerte <|0,1| als nicht sicher zugeord- net. Unterscheidung der Spitzmausarten S. araneus und S. coronatus 89 Endgültige Überprüfung aller Trennmethoden und Merkmale Die neue DF hat von allen bekannten morphologischen Trennmethoden mit annähernd 100% (345/347 bzw. 113/113 Individuen) die größte Trennwirkung. Auch die für Baden- Württemberg optimierte Form (Konstante C = 8,3098) der Trennfunktion von HAUSSER und JamMoT (1974) erreichte höchstens eine richtige Zuordnung von 92,10%. Bemerkens- wert erscheint, daß der Abstand der Mittelwerte bei der neuen DF im Vergleich zur DF von HaAusser und Jammor (1974) wesentlich größer und folglich der Überschneidungs- bereich kleiner ist. Für eine Artdiagnose an unvollständigen Schädeln wurden die Indices PB/FV für den Oberschädel und AH/AB für den Unterkiefer ermittelt. Der Index PB/FV bietet eine gute Ergänzung zu dem recht zuverlässigen (89,81%) „HANDWERK-Index“ X5/X6 für den Oberschädel. Auf Grund der leichteren bzw. korrekteren Handhabung ist der Index AH/AB ohne Qualitätseinbußen meßmethodisch den Indices X21/X19 (HANDWERK 1987) bzw. dem „Höhen/Breiten-Index“ (BERGER et al. 1992) vorzuziehen. Überprüfungen an vorliegendem Material ergaben, daß eine zuverlässige Unterschei- dung von S. araneus und S. coronatus nach der Fellfärbung in Baden-Württemberg nicht möglich ist. Zu dem gleichen Resultat führte eine Untersuchung an Schweizer Material (NEET 1992). Diese Befunde stehen jenen aus Nordrhein-Westfalen entgegen, wonach sich dort beide Arten - sofern nicht juvenile Individuen vorliegen — gut nach der Fellfärbung trennen lassen sollen (von LEHMANN 1955; OLERT 1973; HUTTERER und VIERHAUS 1984; HANDWERK 1987). HANDWERK (1987) weist allerdings auch darauf hin, daß eine Un- terscheidung in manchen Gebieten nicht möglich war. Geographische Schwankungen einzelner Meßstrecken in Baden-Württemberg Für Sorex araneus kann festgehalten werden, daß Tiere aus dem Alpenvorland und der Südlichen Oberrheinebene durchschnittlich größer sind als solche aus den nördlicheren Landesteilen. Dennoch besteht kein Süd-Nord-Gefälle hinsichtlich der Schädelgröße, viel- mehr dürfte die morphologische Varianz durch klimatische Unterschiede bedingt sein, wofür die erheblichen Unterschiede zwischen den Populationen der Südlichen Oberrhein- ebene und des Schwarzwaldes, der Schwäbischen Alb und des Donaugebietes, des Oden- waldes und des Kraichgaus sprechen. Für Sorex coronatus scheinen ähnliche Verhältnisse vorzuliegen wie für Sorex araneus: Die Tiere aus der Voralpenregion sind relativ größer als in allen anderen Klimabezirken. Auch hier müssen geoklimatische Ursachen für die morphologische Varianz in Betracht gezogen werden - es ist kein Süd-Nord-Gefälle erkennbar. Zur Klärung der interspezifischen Verhältnisse zwischen den einzelnen Klimabezirken wurde die Tab. 12 erstellt. Durchschnittlich ergaben sich -— sofern ein Vergleich möglich war — größere interspezifische als intraspezifische Abweichungen zwischen den Klimabe- zirken. Auch die verschiedenen Box-and-Whisker-Plots veranschaulichen, daß die Meß- strecken nicht überall gleichsinnig variieren. Ob die geoklimatischen Einflüsse einer größeren, klimatisch homogenen Region eine unterschiedliche Wirkung auf jede der bei- den Arten hinsichtlich ihrer Morphologie haben, oder ob diese Unterschiede auf die teil- weise geringen Stichprobenumfänge zurückzuführen sind, kann noch nicht abschließend geklärt werden. Welche Folgen ergeben sich für die neue Trennfunktion in Baden-Württemberg, wenn die Meßstrecken geoklimatische Schwankungen zeigen und eventuell nicht überall gleich- sinnig variieren? An den Box-and-Whisker-Plots wird deutlich, daß hinsichtlich der Ab- stände der Mediane und der Boxen die Überschneidungsbereiche der Meßstrecken in allen Klimabezirken gering bleiben. Überprüfungen ergaben, daß die Größenordnung der 90 H. TurnI und E. F MÜLLER einzelnen Schwankungen (Abweichungen in %) in Hinblick auf die Beständigkeit der neuen Trennfunktion vernachlässigbar ist. Ist ein „character displacement‘ erkennbar? Geschwisterarten wie S. araneus und S. coronatus weisen nur geringe morphologische Un- terschiede auf und zeigen sehr ähnliche ökologische Ansprüche. Deshalb dürften sie nur selten gemeinsam vorkommen und es müßte dann ein „character displacement“ stattfin- den. Das hieße, daß bestehende geringe Unterschiede in Lebensweise und Morphologie bei sympatrischer Verbreitung verstärkt würden. Die Schädelmeßstrecken von S. araneus sollten also in Gebieten, wo S. araneus und S. coronatus sympatrisch sind, deutlichere Un- terschiede aufweisen (etwa deutlich größere Condylobasallängen) als in Gebieten, wo ausschließlich S. araneus vorkommt. Gegen ein „character displacement“ sprechen die durchschnittlich alle anderen an Größe übertreffenden Schädel der südlichen Oberrheintiefebene von S$. araneus, die dort fast ausschließlich allein vorkommt. Auch $. araneus aus der nördlichen Oberrhein-Tief- ebene oder aus dem Donautal — jeweils klar dominierend - sind größer als etwa S. araneus aus dem Schwarzwald oder dem Kraichgau. Ähnliches beobachtete NEET (1989, zit. in HAusser 1994) an einer Schweizer Population: Die morphologischen Un- terschiede von S. araneus und S. coronatus waren zwischen allopatrischen Populationen größer als in den Kontaktzonen, wo beide Arten gemeinsam vorkommen. Gültigkeit der neuen Diskriminanzfunktion außerhalb von Baden-Württemberg Die Trennwirkung der neuen Diskriminanzfunktion wird durch den relativ großen Ab- stand der Mittelwerte beider Verteilungen gekennzeichnet. Erste Hinweise und eigene Untersuchungen an Stichproben aus Bayern, Hessen, Nordrhein-Westfalen, Thüringen, Sachsen-Anhalt, Brandenburg und Sachsen lassen die Vermutung zu, daß die neue TIrenn- funktion auch außerhalb Baden-Württembergs große Zuverlässigkeit besitzt. In Ost-Thüringen, Sachsen-Anhalt, Brandenburg und Sachsen konnten einige auffal- lend kleine Schädel von $. araneus beobachtet werden. Manche zierlichen Unterkiefer wiesen relativ schlanke, „coronatus‘ — ähnliche Condyli auf. Von 531 untersuchten Unter- kiefern lagen nach der Zuordnung durch den Index AH/AB 12 (= 2,26%) knapp im „coronatus“-Bereich. An der Zugehörigkeit dieser fraglichen Stücke zur Art S. araneus bestand jedoch nach Überprüfung weiterer Merkmale bzw. mit Hilfe der neuen Diskrimi- nanzfunktion kein Zweifel. Zusammenfassung Eine neue Trennfunktion zur Unterscheidung von Sorex araneus und S. coronatus anhand von Schädel- merkmalen wird beschrieben. Die neue Funktion basiert auf der Messung von vier einfach zu ermitteln- den Meßstrecken, die untereinander kaum korrelieren und an Gewöllematerial fast immer vorhanden sind. Die neue Diskriminanzfunktion ordnete von 114 biochemisch (PAGE) bestimmten Individuen 100% richtig zu, im Vergleich zu 92,1% bei der Diskriminanzfunktion von HAUSSER und JAmMoT (1974) und 89,81% beim Index x5/x6 von HANDWERK (1987). Um auch unvollständig erhaltene Schädel bestim- men zu können, wurden für den Oberschädel und die Mandibel jeweils Indices ermittelt. Mit Hilfe der neuen TIrennfunktion und gegebenenfalls anhand der Indices und Hilfsmerkmale wurde Sammlungs- und Gewöllematerial fast flächendeckend aus ganz Baden-Württemberg determiniert (ca. 4000 Sorex araneus/coronatus). Eine vergleichende Untersuchung an geeigneten Meßstrecken zeigte dabei geo- klimatisch bedingte Schwankungen: z.B. sind Individuen aus Populationen im Alpenvorland durch- schnittlich größer als solche aus den anderen Landesteilen. Zwischen verschiedenen Klimabezirken sind die interspezifischen Unterschiede meist größer als die intraspezifischen. Hinsichtlich der Bestän- Unterscheidung der Spitzmausarten S$. araneus und S. coronatus 91 digkeit der neuen Trennfunktion ist die Größenordnung der Schwankungen vernachlässigbar, da in allen Klimabezirken die UÜberschneidungsbereiche der einzelnen Meßstrecken gering bleiben. Bei sympatrischer Verbreitung der beiden Arten konnte kein „character displacement“ festgestellt werden. Danksagung Ein Teil des biochemisch (PAGE) sicher bestimmten Materials geht auf die Untersuchungen von H. BRÜNNER (1988, 1990) zurück. Für die freundliche Überlassung dieser eingefrorenen Tiere zur Prä- paration der Schädel, aber auch für wertvolle Hinweise sei ihm an dieser Stelle vielmals gedankt. Für weiteres Material möchten wir uns außerdem bei der Leitung des Projektes „Wildlebende Säugetiere in Baden-Württemberg“ Frau M. Braun, Dr. F. DIETERLEN und Dr. R. FLösser sowie für die Überlassung des „Schönbuchmaterials“ bei Prof. Dr. E. KuLZEr bedanken. Ganz besonderer Dank gilt Dr. G. Arm- BRUSTER der die Durchführung der Diskriminanzanalyse mit dem SPSS/PC" ermöglichte. Literatur BERGER, M.; FELDMANN, R.; REHAGE, N. O.; SKIBA, R. (1992): Kleinsäugetier-Zönosen bachbegleitender Feuchtgebiete des südwestfälischen Berglandes. Abh. Westf. Mus. Naturkd. Münster 54, 1-47. BRÜNNER, H. (1988): Untersuchung zur Verbreitung, Ökologie und Karyologie der Waldspitzmaus (Sorex araneus Linne, 1758) und der Schabrackenspitzmaus (Sorex coronatus Millet, 1828) im Frei- burger Raum mit Bemerkungen zu einigen anderen Spitzmausarten. Diplomarbeit Univ. Freiburg. BRÜNNER, H. (1990): Zur Verbreitung von Waldspitzmaus (Sorex araneus Linne, 1758) und Schabracken- spitzmaus (Sorex coronatus Millet, 1828) in Oberrheinebene und Schwarzwald. Unveröff. Bericht im Auftrag des Forschungsvorhabens „Wildlebende Säugetiere in Baden-Württemberg“, Karlsruhe. BürLow, B. von (1989): Beitrag zur Verbreitung der Kleinsäuger im westlichen Münsterland. Natur und Heimat 49, 17-21. HANDWERK, J. (1987): Neue Daten zur Morphologie, Verbreitung und Ökologie der Spitzmäuse Sorex araneus und S. coronatus im Rheinland. Bonn. zool. Beitr. 38, 273-297. HAUSSER, J. (1984): Genetic drift and selection: Their respective weights in the morphological and ge- netic differentiation of four species of shrews in Southern Europe (Insectivora, Soricidae). Z. zool. Syst. Evolut.-forsch. 22, 302-320. HAUSSER, J. (1990): Sorex coronatus Millet, 1828 — Schabrackenspitzmaus. In: Handbuch der Säugetiere Europas. Hrsg. von: J. NIETHAMMER und F. KraApp, Wiesbaden: Aula-Verlag. Pp. 279-286. HAUSSER, J. (1994): The Sorex of the araneus-arcticus group (Mammalia: Soricidae): Do they actually speciate? - Special publication of Carnegie Museum of Natural History 18, 295-306. HAUSSER, J.; JAMMOT, D. (1974): Etude biome£trique des machoires chez les Sorex du groupe araneus en Europe continentale (Mammalia, Insectivora). Mammalia 38, 324-343. HAUSSER, J.; ZUBER,N. (1983): Determination specifique d’individus vivants des deux especes jumelles Sorex araneus et Sorex coronatus, par deux techniques biochimiques (Insectivora, Soricidae). Revue suisse Zool. 90, 857-862. HAUSSER, J.; HUTTERER, R.; VOGEL, P. (1990): Sorex araneus Linnaeus, 1758 — Waldspitzmaus. In: Hand- buch der Säugetiere Europas. Hrsg. von: J. NIETHAMMER und F. KrApp. Wiesbaden: Aula-Verlag. Pp. 237-278. HAUSSER, J.; BOSSHARD, F.; TABERLET, P.; WoJcıKk, J. (1991): Relationships between chromosome races and species of Sorex of the araneus group in the western Alps. M&moires de la Societe vaudoise des Sciences naturelles 19, 79-95. HUTTERER, R.; VIERHAUS, H. (1984): Waldspitzmaus — Sorex araneus Linnaeus, 1758; Schabrackenspitz- maus — Sorex coronatus Millet, 1828. In: Die Säugetiere Westfalens Hrsg. von R. SCHRÖPFER, R. FELDMANN und H. VIERHAUS, Abh. Westf. Mus. Naturkde. 46, 54-60. KULZER, E.; LINDEINER, A. VON; WOLTERS, 1.-M. (1993): Säugetiere im Naturpark Schönbuch. Beih. Ver- öff. Naturschutz Landschaftspflege Bad.-Württ. 71, 1-212. LEHMANN, E. von (1955): Über die Untergrundmaus und Waldspitzmaus in NW-Europa. Bonn. zool. Beitr. 6, 8-27. LINDER, A.; BERCHTHOLD, W. (1982): Statistische Methoden III: Multivariate Verfahren. Birkhäuser Verlag. 92 H. Turnı und E. F. MÜLLER LocH, R. (1977): A biometrical study of karyotypes A and B of Sorex araneus Linnaeus, 1758, in the Netherlands (Mammalia, Insectivora). Lutra 19, 21-35. LORENZ, R. J. (1988): Grundbegriffe der Biometrie (2. Auflage). Jena: Gustav Fischer Verlag. Mys, B.; VAN DER STRAETEN, E.; VERHEYEN, W. (1985): The biometrical and morphological identification and the distribution of Sorex araneus L., 1758 and Sorex coronatus Millet, 1828 in Belgium (Insecti- vora, Soricidae). Lutra 28, 55-70. NEET, €. R. (1992): The use of fur colour characters to distinguish the sibling species Sorex araneus and Sorex coronatus (Insectivora, Soricidae): a field test in a zone of parapatric contact. Z. Säugetier- kunde 57, 176-178. NEET, C. R.; HAUSSER, J. (1991): Biochemical analysis and determination of living individuals of the Al- pine karyotypic races and species of the S$. araneus group. Memoires de la Societe vaudoise des Sciences naturelles 19, 97-106. OLERT, J. (1973): Schädelmessungen an rheinischen Wald- und Schabrackenspitzmäusen. Bonn. zool. Beitr. 24, 366-373. PIEPER, H. (1978): Zur Kenntnis der Spitzmäuse (Mammalia, Soricidae) in der Hohen Rhön. Beitr. Nat- urkdl. Osthessen 13/14, 101-106. SacHs, L. (1992): Angewandte Statistik (7. Auflage). Heidelberg: Springer-Verlag. TURNI, H.; SCHÖNHERR, R. (1994): Neue Nachweise der Schabrackenspitzmaus (Sorex coronatus) in Ba- den-Württemberg durch Polyacrylamidgel-Elektrophorese. Z. Säugetierkunde 59, 321-3235. Anschrift der Verf.: HENDRIK TUrNI und Prof. Dr. EwALD F. MÜLLER, Zoologisches Institut, Abt. Phy- siol. Ökologie, Universität Tübingen, Auf der Morgenstelle 28, D-72076 Tübingen. FÜR Z. Säugetierkunde 61 (1996) 93-103 ZEITSCHRIFT © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Crocidura ramona sp. nov. (Insectivora, Soricidae): a new species of shrew from the central Negev desert, Israel By ELENA IVANITSKAYA, G. SHENBROT and E. NEvo Institute of Evolution, University of Haifa, Haifa, Israel and Ramon Science Center, Ben-Gurion Univer- sity ofthe Negev, Mizpe Ramon, Israel Receipt of Ms. 29. 05. 1995 Acceptance of Ms. 22. 12. 1995 Abstract A new species of Crocidura (C. ramona) from Central Negev desert, Israel is described. The new spe- cies has small size, flattened skull, and light silver-gray dorsal pelage coloration. The diploid chromo- some number is 28, NFa = 42, X-chromosome - acrocentric. The species belongs to the Palearctic group of the genus Crocidura. Within this large group, the nearest relative of C. ramona is C. pergrisea. Introduction For decades, only three taxa of the genus Crocidura have been described in the fauna of Israel: C. leucodon judaica Thomas, 1919; C. russula monacha Thomas, 1906; and C. suaveolens portali Thomas, 1920 (Harrıson 1964; ATALLAH 1977; MENDELSSOHN and Yom-Tov 1987). The last taxonomic revision based on chromosomal, biochemical, and morphometrice data (CATZErLIıs et al. 1985) presented evidence of conspecifity of C. russula monacha and C. suaveolens. This point of view was accepted, and now only two species of Crocidura (C. leucodon and C. suaveolens) are included in the list of Israeli mammals (HArrıson and BATEs 1991; SHALMoN 1993). During the course of regular trapping sessions of rodents, lizards, and insects in the Central Negev, between 1993 and 1995, we found seven specimens of Crocidura clearly different from known species, both morphologically and karyologically. Investigation of zoological collections revealed four additional specimens of the same taxon. The new spe- cies, Crocidura ramona from the Negev Desert, is described. Material and methods Two males, one female and one embryo (also female) of Crocidura sp. were used for karyotype analysis. The chromosomes of adult animals were prepared by the standard method from bone marrow and spleen cells. Chromosomes from embryos were obtained from liver tissue by use of a short-term cell cul- ture method: the liver cells were incubated at 37°C for two hours in Medium 199 and calf serum (Sig- ma) at a proportion of 6:1. Then two drops of 0.004% colchicine solution were added to 6 ml of the suspension, and after 20 minutes the conventional chromosomal technique was employed. Chromo- some preparations were stained with 4% Giemsa. For G- and C- band staining, the trypsin (SEABRIGHT 1971) and BSG (Sumner 1972) methods were applied. As comparative material we used G-banded chromosomes of one male of C. suaveolens captured on Mt. Carmel, Israel. 94 ELENA IVANITSKAYA, G. SHENBROT and E. NEvVo Skull morphology was studied on eight specimens of Crocidura sp. trapped in the Makhtesh Ramon and ın the vicinity of Mizpe Ramon, Central Negev desert. These were then compared with 47 speci- mens of Crocidura suaveolens and 12 specimens of Crocidura leucodon from zoological collections at Tel-Aviv (TAU) and Hebrew (HU) Universities. Four cranial measurements were analysed: condylo-in- cisive length (Lei), zygomatic breadth (Bz), preorbital breadth (Bpo), and skull height (H). All mea- surements were performed according to KAHMANN and VESMANIS (1974) and ZAITsEV (1991). Results and discussion Skull morphometry Measurements of the examined specimens are presented in table 1. All of the Crocidura specimens from the vicinity of Mizpe Ramon (seven of ours and one from the HU collec- tion) are clearly different from both C. leucodon and C. suaveolens in skull size and pro- portions. The specimens of the Central Negev are much smaller than C. leucodon, but similar in size (or slightly smaller on average) to C. suaveolens. Likewise, skulls of the spe- cimens from the Central Negev are flattened; this character is shared with C. leucodon, but different from C. suaveolens (Fig. 1). Three additional specimens, indistinguishable from the Mizpe Ramon sample, were found in the TAU collections: two from Sede Boger (Central Negev, 30 km north of Mizpe Ramon) and one from Sartaba (northern edge of Judean Desert). Table 1. Skull measurements (upper line - min-max, lower line - mean +SD) of the three Crocidura species of Israel. Skull measurements: condylo-ineisive length (Lei), zygomatic breadth (Bz), preorbital breadth (Bpo), skull height (H), and relative skull height index (H/Bpo) Species C. suaveolens (n = 47) 17.0 - 20.0 5.2 - 6.3 3.0 - 3.6 3.1-3.8 0.97 -1.12 181692. 0/025255779250923 3335-4014 3.46 #0.15 1.041 # 0.035 C. ramona sp.n.(n=11) 17.7 -18.4 5.1-5.6 3.2 - 3.6 3.1-3.3 0.86 - 0.97 LS05-A02 62553723045 3.44 #0.13 3.152. 0.072091 9E-0:051 C. leucodon (n = 12) 19.7 - 21.7 6.2 - 6.8 3.4 4.0 3.4 - 3.8 0.91 - 1.06 20 18EZ0,D0,,0693E3020 327323020 3.612 04275.0970530:071 Karyotype analysis All karyotyped specimens of Crocidura sp. have 28 chromosomes in the diploid comple- ment (Fig. 2a). Three groups of chromosomes can be distinguished based on their mor- phology: the first group consists of three pairs of submeta- subtelocentric chromosomes; the second group includes five pairs of subtelocentrics; and the third group involves five acrocentric pairs (NFa = 42). The X-chromosome is acrocentric, comparable in size with the first chromosome pair of the second group. The Y-chromosome is acrocentric, equal in size with the last acrocentric autosomes. The karyotype of Crocidura suaveolens from Mt. Carmel has 40 chromosomes: four pairs of subtelo- submetacentric autosomes, 15 pairs of acrocentric autosomes and submetacentric X-chromosome and acrocentric Y- chromosome (NFa = 46). The karyotype of C. suaveolens from Mt. Carmel does not differ from the chromosome complements of C. suaveolens that have been studied from the vi- cinity of Tel-Aviv (CATZEFLIS et al. 1985) and from other localities (CATZEFLIS et al. 1985; REUMER and MEYLAN 1986; IvanıTsKAYA 1989; GRAFODATSKY et al. 1991). Crocidura ramona sp. nov. (Insectivora, Soricidae) 95 1.14 mr Er er Zr Pi = i Tey 11 | ° ° oO, ° 1.06 2 s So ° oe n OA [0) oO 000 o0oo © e ) A 1.02 Ä a 0 0 000000 or A JE: [ 0.98 } & 6) Ai A A \ r, A A 0.94 L Br ) 0) — & i A 4 | ;® o e 0.9 [ “ [ o .. 0.86 en ZZ 6,5 1725 18.5 19.5 20.5 24r5 22.5 Loi Fig. 1. Scatter-plot of cranial measurements of the three Crocidura species of Israel. Lci - condylo-inci- sive length, H/Bpo - index of relative skull height (skull height/preorbital breadth). Solid circles — C. ramona sp. nov., open circles - C. suaveolens, triangles — C. leucodon. Dotted ellipses confine the 95% confidence areas. The C-band method revealed a very small amount of autosomal heterochromatin in Crocidura sp., localized in the pericentromere regions of 9 chromosome pairs with differ- ent degrees of expression. The short arms of the first pair of subtelocentric autosomes are almost entirely C-positive. The X-chromosome does not contain the heterochromatin ma- terial, and the Y-chromosome completely consists of heterochromatin (Fig. 2b). All spe- cies of Crocidura studied with the C-banding method have a small amount of heterochromatin in the autosome complement and C-positive Y-chromosomes (HARADA et al. 1985; TApA and OBARA 1986; GRAFODATSKY et al. 1988; IvAnITSKAYA 1989). At the present time the karyotypes of 50 species of the genus Crocidura are known. Most of these species are differentiated karyologically. Intraspecific chromosomal varia- bility has been revealed in six species (MADDALENA and Ruepı 1994; RUEDI and VOGEL 1995). C. suaveolens is the widest geographically distributed species of the genus; of more than 30 Karyologically studied populations of this species, only four possess chromosomal variability. Three populations from Switzerland are polymorphic for the number of chro- mosomes, due to the presence in some karyotypes of B-chromosomes (MEyLAnN and HAus- SER 1974). In addition, one population from Japan is characterized by 2n =39-40, as a result of a Robertsonian translocation (TsucHiyA 1987). The other chromosomally poly- morphic populations of Crocidura occur only in Southeastern Asia and in the Afrotropi- cal region. The diploid numbers in Crocidura species vary from 2n = 22 in C. pergrisea to 2n =60 in €. cf. bicolor and NF - from 34 to 86. The high numbers of 2n and NF are typi- cal for species of Afrotropical origin; species with low chromosomal numbers are distribu- ted in the Palearctic and Oriental regions (MADDALENA and Rurpı 1994). Thus, Crocidurasp. belongs to the second group by virtue of its chromosomal formula. C. leucodon also has the same number of chromosomes (?2n =28) as Crocidura sp., but 96 ELENA IVANITSKAYA, G. SHENBROT and E. NEVo AN AK an Na Ni AA An nA hr aa aa An an Di a GUY RUE M 2 = > er Fig. 2. Conventionally stained (a) and C-banded (b) chromosomes of C. ramona sp. nov. (male, no. M-9342 (TAU), holotype). these two species vary distinctly in their number of chromosomal arms: NFa = 42 in C. ramona and NFa = 52 in C. leucodon. In addition, C. leucodon, like most Crocidura spe- cies, has a biarmed X-chromosome. To date, an acrocentric X-chromosome has been re- vealed only in two species: one Palearctic, C. pergrisea (GRAFODATSKY et al. 1988) and one Oriental, C. lepidura (RuEpı and VoGEL 1995). Comparative analysis of G-banded chromosomes of Crocidura sp. and C. suaveolens revealed similar G-band patterns in some chromosomes (Fig. 3). All acrocentric chromo- somes, the Y-chromosome and two pairs (1 and 3) of the first group of Crocidura sp. have a banding pattern that is practically identical to the corresponding chromosomes of C. suaveolens. Subtelocentric chromosomes of Crocidura sp. match only partially with C. suaveolens chromosomes. The acrocentric X-chromosome of Crocidura sp. has inverted G-band patterns in comparison to the biarmed X-chromosomes of C. suaveolens. The same type of rearrangement in X-chromosome was recorded for C. pergrisea (GRAFODATS- KY et al. 1988). Apparently, the X-chromosomes of Crocidura sp. and C. pergrisea are clo- sely related. Crocidura ramona sp. nov. (Insectivora, Soricidae) 97 Fig. 3. Comparison of G-banded chromosomes haploid sets of C. ramona sp. nov. and C. suaveolens. In each pair the left chromosome represents C. ramona and the right one is of C. suaveolens. Unpaired chromosomes in the lower row represent chromosomes of C. suaveolens that have no analogs in the karyotype of C. ramona. Triangles mark the centromere positions. It is not possible to derive the karyotype of Crocidura sp. directly from a karyotype of any recent species. Comparative analysis of G-banded chromosomes allowed MADDALENA and Ruepı (1994) to reconstruct the ancestral Crocidura karyotype with 2n=38 and NF=54-58. The chromosomal complement of C. suaveolens is close to this Karyotype. The origin of C. leucodon (2n=28) and C. pergrisea (2n = 22) karyotypes can be ex- plained mainly by Robertsonian type of rearrangements (GRAFODATSKY et al. 1988). Five pericentric inversions (or centromere shifts) and five centromere-telomere fusions are ne- cessary for evolving from the hypothetical ancestral karyotype to the karyotype of Croci- durasp. If we accept the inverted structure of X-chromosomes as an apomorphous character, C. pergrisea and Crocidura sp. can be considered as sister taxa within the Pa- learctic group of Crocidura. Thus, both morphological and karyological analyses lead to the conclusion that Croci- dura sp. represents a new species. Crocidura ramona, sp. nov. Holotype: 3 adult, No. M-9342 deposited in TAU collection, NE region of Makhtesh Ramon, Negev (30° 40’ N 34° 56 E), 1. 04. 1995, skull and skin. Paratypes: 3 adult, No. M-9343, TAU collection, central region of Makhtesh Ramon, Negev (30° 35’ N 34° 51’ E), 13. 03. 1993, skull; 2, adult, No. M-9344, TAU collection, SW 98 ELENA IVANITSKAYA, G. SHENBROT and E. NEvo region of Makhtesh Ramon, Negev (30° 30’ N 34° 40’ E), 4. 08. 1994, skull; © adult, No. M-9345, TAU collection, northern rım of Makhtesh Ramon, Negev (30° 35’ N 34° 42’ E), 21. 01. 1995, skull; $ adult, No. M-9346, TAU collection, NE region of Makhtesh Ramon, Negev (30° 40° N 34° 56 E), 1. 04. 1995, skull, skin; 3 adult, No. M-9347, TAU collection, central region of Makhtesh Ramon, Negev (30° 35’ N 34° 50’ E), 27. 04. 1995, skull, skin; g adult, No. M-9348, TAU collection, central region of Makhtesh Ramon, Negev (30° 34’ N 34° 44’ E), 30. 04. 1995, skull, skin. Additional material: No. M-3347 HU collection, Mizpe Ramon; No. M-8675, M-8676 TAU collection, Sede Boger; No. M-8703 TAU collection, Sartaba. Diagnosis: Shrews of relatively small sıze: condylo-incisive length of skull 17.7- 18.4 mm. Skull is flattened: the ratio of skull height to preorbital breadth less than 1.00. Third upper molar ıs moderately reduced. Dorsal pelage coloration is light silver-gray. Chromosome set: 2n = 28, NFa = 42, X-chromosome: acrocentric. Fig. 4. Skulls of C. ramona sp. n. (a) and C. suaveolens (b) in lateral view. Crocidura ramona sp. nov. (Insectivora, Soricidae) 99 Description and comparison: Size is relatively small: external measurements (based on four specimens) are as follows: head and body length: 58-63 mm; tail length: 42-43 mm; hindfoot length: 10-11 mm; and ear length: 7.5-8.0 mm; body weight: 4.0-5.0 g. In compar- ison with other North African, Arabian, and eastern Mediterranean species (SPITZENBER- GER 1971; OsBorn and HELMY 1980; HUTTERER and HARRISON 1988; HARRISON and BATES 1991; ZArtsev 1991; HUTTERER 1992, 1994), C. ramona is much smaller than C. leucodon, C. lasia, C. pergrisea, C. somalica, and C. olivieri. Body length of C. ramona is smaller on average than C. suaveolens and C. floweri, equal to that of C. religiosa, but greater than C. arabica, however, it is still within the limits of variation for the first two species. Tail length of C. ramona is within the limits of variation for C. suaveolens, shorter than C. floweri and longer than C. arabica and C. religiosa. Hindfoot length of C. ramona is less than C. suaveolens and C. floweri and equal to C. arabica and C. religiosa. The pelage coloration is light. The hairs are bicoloured throughout with slaty gray bases. The tips are sılvery white on the ventral surface and silvery gray on the dorsal sur- face. The line of demarcation along the flanks ıs clear, but not very sharp. The tail is indis- Fig. 5. Rostral part of skulls of C. ramona sp. n. (a) and C. suaveolens (b) in ventral view. 100 ELENA IVANITSKAYA, G. SHENBROT and E. NEvo 30°40 30°30' 35°00' To Eilat 34°50' To Beer Sheva C. suaveolens, squares - C. leucodon. Mizpe Ramon collection data. Asterisks — C. ramona sp. n., circles — Fig. 6. Distribution map of Crocidura species in Israel by N 34°40 $ ee S = S Q ® =] H tinctly bicoloured, with its dorsal surface lighter than its back. In coloration, C. ramona is distinct from most Crocidura species, with the exception of C. pergrisea. The skull (Fig.4) is smaller than that of C.leucodon, C.lasia, C. somalia, and C. olivieri, it is larger than that of C. religiosa, and is approximately similar in size to C. suaveolens, C. pergrisea, C. arabica, and C. floweri. The skull is markedly flattened in Crocidura ramona sp. nov. (Insectivora, Soricidae) 101 dorsal profile. The extent of this flatness in C. ramona is much greater than in most other Crocidura species, including C. leucodon, but less than in C. pergrisea. The upper incisor (i') is powerful, as that of C. leucodon, and noticeably heavier than in other species. In the maxillary dentition, unicuspid' is the largest; the crown area of unicuspid“ ıs smaller than that of unicuspid’, the character shared with C. leucodon, but not with other species. The internal cusp (hypoconus) of the upper premolar is very small, noticeably smaller than that of C. suaveolens, C. pergrisea, and C. arabica, but not comple- tely reduced as in C. leucodon. The third upper molar is moderately reduced posteriorly (Fig. 5). It is smaller than that of C. suaveolens and C. pergrisea, but larger than that of C. leucodon, C. arabica, and African species. Etymology: Named for the Ramon erosion cirque (Hebrew name -— Makhtesh Ra- mon), Central Negev, Israel, where the new species was found. Distribution: The species has been found in three regions: the vicinities of Mizpe Ra- mon, Sede Boger, and Sartaba (Fig. 6). The first two regions are located on the Negev Highlands, and the third is on the north edge of the Judean Desert. All three regions are located in rocky desert, from 400 to 950 m a.s.l. We did not find any other Crocidura spe- cies co-occurring with C. ramona. All localities of C. suaveolens in Israel are confined to regions with a Mediterranean climate or to lowland deserts (i.e., Mediterranean sandy coastal plain, Rift Valley). The only shrew species found together with C. ramona was Suncus etruscus. Habitats: Of seven specimens from the vicinity of Mizpe Ramon, five were caught in the dry riverbeds with Retama and Tamarix vegetation among gravel plain with rock out- crops. The last two specimens were caught near dry riverbeds with dense Arriplex vegeta- tion among rocky hills, partially covered with loess. Remarks: Based on its morphological and karyological characters, C. ramona undoubt- edly belongs to the Palaearctic, but not to the Afrotropical, group of the genus Crocidura (in the sense of MADDALENA and Ruepı 1994). Within this large group, it is more or less equally distant from C. suaveolens, C. leucodon, and C. pergrisea. However, based on col- oration, some skull characters, the structure of X-chromosome, and habitat distribution, it is possible to suggest that the nearest (but not closely related) relative of C. ramona sp. nov. is C. pergrisea. This being true, an ancestor of C. ramona penetrated into the territory of Israel from Asia Minor. This penetration presumably occurred in the Early Pleistocene, as the first records of Crocidura from Jordan Valley are dated to that time (TCHERNOV 1988). Acknowledgements We thank Dr. Borıs KrAasnov (Ramon Science Center, Ben-Gurion University of the Negev) for field assistance. We also thank Prof. YorRAMm Yom-Tov, Dr. TsıLA SHARIV (Tel-Aviv University) and Prof. EıTAN TCHERNoV (Hebrew University) for providing possibility to study collection material. This is publication no. 34 of Ramon Science Center. Zusammenfassung Crocidura ramona sp. nov. (Insectivora, Soricidae): eine neue Spitzmausart aus dem Zentralteil der Negev Wüste, Israel. Eine neue Art von Crocidura aus dem Zentralteil der Negev Wüste in Israel wird beschrieben. Diese neue Art ist durch kleine Körpergröße, flachen Schädel und silbergraue Färbung des Rückenfells ge- kennzeichnet. Die Art gehört zu der Palaearktischen Gruppe der Gattung Crocidura. Nächster Ver- wandter von C. ramona innerhalb dieser großen Gruppe ist C. pergrisea. 102 ELENA IVANITSKAYA, G. SHENBROT and E. NEvo Literature ATALLAH, S. I. (1977): Mammals of the Eastern Mediterranean Region; their ecology, systematics and zoogeographical relationships. Säugetierkundl. Mitt. 25, 241-320. CATZEFLIS, F.; MADDALENA, T.; HELLWING, S.; VOGEL, P. (1985): Unexpected findings on the taxonomic status of East Mediterranean Crocidura russula auct. (Mammalia, Insectivora). Z. Säugetierkunde 50, 185-201. GRAFODATSKY, A. S.; RADZHABLI, S. I.; SHARSHOV, A. V.; ZAITSEV, M. V. (1988): Karyotypes of five Croci- dura species of the USSR fauna. Tsitologiya 30, 1247-1250 (In Russian). GRAFODATSKY, A.S.; RADZHABLI, S. 1.; ZAITSEV, M. V.; SHARSHOV, A. V. (1991): The levels of chromo- some conservatism in the different groups of insectivores (Mammalia, Insectivora). Proc. Zool. Inst. USSR Acad. Sci. 243, 47-57 (In Russian). HARADA, M.; YosıDaA, T. H.; HATTORI, S.; TAKADA, S. (1985): Cytogenetical studies on Insectivora: IIl. Karyotype comparison of two Crocidura species in Japan. Proc. Japan. Acad. Ser. B 61, 371-374. HARRISON, D. L. (1964): The mammals of Arabia. Vol. 1. London: E. Benn Ltd. HARRISON, D. L.; BATEs, P. (1991): The mammals of Arabia. 2nd ed. Sevenoaks: Harrison Zool. Mus. Publ. HUTTERER, R. (1992): Order Insectivora. In: Mammal Species of the World: A Taxonomic and Geo- graphic Reference. Ed. by D. E. Wırson and D. M. REEDER. 2nd ed. Washington: Smithsonian Insti- tution Press. Pp. 69-130. HUTTERER, R. (1994): Shrews of ancient Egypt: Biogeographical interpretation of a new species. In: Ad- vances in the Biology of Shrews. Ed. by J. F. MERRITT, G. L. KIRKLAND, and R.K. Rose. Pittsburgh: The Carnegie Mus. Natur. Hist. Pp. 407413. HUTTERER, R.; HARRISON, D. L. (1988): A new look at the shrews (Soricidae) of Arabia. Bonn. zool. Beitr. 39, 59-72. IvanıTsKAYA, E. (1989): Constitutive heterochromatin and nucleolar organizer regions in karyotypes of some shrews (Soricidae, Insectivora). Genetika 25, 1188-1198 (In Russian). KAHMAnN, H.; VESMANISs, I. (1974): Morphometrische Untersuchungen an Wimperspitzmäusen (Croci- dura). 1. Die Gartenspitzmaus Crocidura suaveolens (Pallas, 1811) auf Menorca. Säugetierkundl. Mitt. 22, 313-323. MADDALENA, T.; RUEDI, M. (1994): Chromosomal evolution in the genus Crocidura (Insectivora: Sorici- dae). In: Advances in the Biology of Shrews. Ed. by J. F. MERRITT, G. L. KıRKLAND, and R.K. Rose. Pittsburgh: The Carnegie Mus. Natur. Hist. Pp. 335-344. MENDELSSOHN, H.; Yom-Tov, Y. (1987): Plants and animals of the land of Israel. Vol. 7. Mammals. Tel- Aviv: Ministry of Defence/the Publishing House (In Hebrew). MEYLAN, A.; HAUSSER, J. (1974): Position cytotaxonomique de quelques musaraignes du genre Croci- dura au Tessin (Mammalia, Insectivora). Rev. suisse Zool. 81, 701-710. OsBoRrNn, D. J.; HELMY, I. (1980): The contemporary land mammals of Egypt (including Sinai). Fieldiana Zool. 5, 1-579. REUMER, J. W. F; MEYLAN, A. (1986): New developments in vertebrate cytotaxonomy IX, Chromosome numbers in the order Insectivora (Mammalia). Genetica 70, 119-151. Rueoı, M.; VoGEL,P. (1995): Chromosomal evolution and zoogeographic origin of southeast Asian shrews (Genus Crocidura). Experientia 51, 174-178. SEABRIGHT, M. (1971): A rapid banding technique for human chromosomes. Lancet 2, 971-972. SHALMON, B. (1993): A field guide to the land mammals of Israel, their tracks and signs. Jerusalem: Ke- ter Publ. House Ltd. (In Hebrew). SPITZENBERGER, F. (1971): Eine neue, tiergeographisch bemerkenswerte Crocidura (Insectivora, Mam- malia) aus der Türkei. Ann. Naturhistor. Mus. Wien 75, 539-552. SUMNER, A.T. (1972): A simple technique for demonstrating centromeric heterochromatin. Exp. Cell Res. 75, 304-306. TADA, T.; OBARA, Y. (1986): Karyological relationship between the Japanese house shrew, Suncus muri- nus riukiuanus and the Japanese white-toothed shrew, Crocidura dsinezumi chisai. Proc. Japan. Acad. Ser. B 62, 125-128. TCHERNOoV, E. (1988): The biogeographical history of the southern Levant. In: The Zoogeography of Is- rael: Distribution and Abundance at Zoogeographical Crossroad. Ed. by Y. Yom-Iov and E. TCHERNov. Dordrecht, Netherland: Dr. W. Junk Publishers. Pp. 159-250. Crocidura ramona sp. nov. (Insectivora, Soricidae) 103 TsucHIyA, K. (1987): Cytological and biochemical studies of Insectivora in Tsushima Island. In: Nature of Tsushima, Tsushima Natural Resource Research Report. Nagasaki Prefecture. Pp. 111-124 (In Japanese, English summary). ZAITSEV, M. V. (1991): Species composition and questions of systematics of white-toothed shrews (Mammalia, Insectivora) of the fauna of USSR. Proc. Zool. Inst. USSR Acad. Sci. 243, 3-46 (In Russian). Authors’ addresses: Dr. ELENA IvANITSKAYA, Prof. EVIATAR NEVo, Institute of Evolution, University of Haifa, Mt. Carmel, Haifa 31905, Israel; Dr. GEORGY SHENBROT, Ramon Science Center, Ben-Gurion University of the Negev, Mizpe Ramon 80600, Israel. Z. Säugetierkunde 61 (1996) 104-111 ZEITSCH RIFT® an, = FÜR © 1996 Gustav Fischer, Jena SÄUG ETI ERKÜ N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Mitochondrial DNA sequence patterns of Harbour porpoises (Phocoena phocoena) from the North and the Baltic Sea By R. TIEDEMANN, J. HARDER, CHRISTINE GMEINER, and E. HAASE Institut für Haustierkunde, University of Kiel, Kiel, Germany Receipt of Ms. 29. 11.1995 Acceptance of Ms. 22. 01. 1996 Abstract To investigate genetic differentiation between harbour porpoises (Phocoena phocoena) of the North and the Baltic Sea, a total of 39 individuals were screened for sequence polymorphisms at a highly poly- morphic part of the mitochondrial DNA control region. DNA was extracted from liver or skin samples of stranded animals. After PCR amplification and direct sequencing, 420 bp were scored. Nine haplo- types were found, differing from one another by one to four transitions. Haplotypes separated out into two clusters A and B by a specific nucleotide substitution. All Baltic harbour porpoises showed type A haplotypes, which was found only in 45% of the North Sea specimens. Genetic variation in terms of nu- cleotide and haplotype diversities was much lower in the Baltic Sea than in the North Sea population. Haplotype composition and nucleotide divergence suggest a colonization of the Baltic Sea several thou- sand years ago and a limited genetic exchange since then. The genetic differentiation between Baltic and North Sea populations of harbour porpoises is corroborated by published data both on skull charac- ter differences and enzyme polymorphisms. Introduction The harbour porpoise (Phocoena phocoena) is a small cetacean species inhabiting coastal waters in the Northern hemisphere (Nowak 1991; KınzeE 1994). On a global scale, the po- pulations of the North Atlantic, the North Pacific, and the Black Sea differ significantly in morphology, especially in skull characteristics, and have been described as three different subspecies Phocoena phocoena phocoena, P.p. vomerina, and P. p. relicta (TomıLın 1967; Yurıck and GaskIn 1987; Amano and MiyAzakrı 1992; KınzE 1994). The subspecies status is corroborated by mitochondrial control region differentiation (RoseL et al. 1995). A further subdivision into local populations has been proposed due to limited migration and the incoherence of suitable habitat (GAskın 1984; KınzeE 1994). The population definition is controversial for the Eastern part of the North Atlantic: while some authors define three to four local populations (GAsKIn 1984), others assume the entire area to be inhab- ited by a coherent population (AnDERSEN 1972). However, there is some support for a se- parate Baltic population from skull character analyses (KınzE 1985) as well as from enzyme electrophoretic data (ANDERSEN 1993). From a population genetic point of view, the alternatives of population segregation vs. panmixia may be settled by a quantitative approach. Our comparison of the harbour por- poises from the North Sea and the Baltic Sea aims at estimating the amount and the di- rection of gene flow between those areas. In this context, mitochondrial DNA sequence patterns have been proven to be a suitable measure of genetic diversity within and among populations (cf. Avıse 1994). MtDNA patterns in Harbour porpoises Phocoena phocoena 105 Material and methods A total of 39 harbour porpoises stranded on the coast of the North Sea (n = 19) and the Baltic Sea (n = 20) were analysed (Fig. 1). Liver or skin samples were stored for 0-24 months at -80 °C until analy- sis. Fig. 1. Sampling localities (place of stranding) of the analysed harbour porpoise specimens. Numbers give sample size for each site. Total DNA was exctracted from 100 mg of tissue using the SuperQuikGene DNA isolation kit (Analytical Genetic Testing Center, Denver, USA) according to manufacturer’s instructions. The DNA was dissolved in a final volume of 100 ul Tris (pH 8.5). 5 ul DNA solution were used for an enzymatic amplification of a part of the mitochondrial control region via polymerase chain reaction (PCR), using the primers S-CACCACCAACACCCAAAGCT-3 and S-CCTGAAGTAAGAACCAGATG-3. These primers produce a product of 471 basepairs (bp), containing 45 bp of the Prolin-tRNA and 426 bp of the control region. The PCR-product contains the most variable part of the control region in Cetacea (Arnason et al. 1993). The following amplifcation profile was applied: After an initial denaturation step at 95°C for 5 min, 40 cycles were carried out with a denaturation at 94°C for 1:30 min, annealing at 51.9°C for 1:15 min, and extension at 72°C for 1:30 min, followed by a final extension at 72°C for 2:30 min. The PCR-products were cycle-sequenced directly using the SequiTherm Cycle Sequencing Kit (Epicentre, Madison, USA) and the Digoxigenin-labelled oligonucleotide 5’-DIG-CACCAA- CACCCAAAGCT-3 for 30 cycles, each with a denaturation at 95°C for 30 s, an annealing at 53.2 °C for 30 s, and an extension at 70 °C for 1 min. Samples were run on a direct blot sequencing device (RICHTER- ıcH et al. 1989), and sequences were detected using an anti-Digoxigenin/alkaline phosphatase conjugate (Boehringer, Mannheim, Germany) and the chemiluminescent substrate CDP-Star (Tropix, Bedford, USA), following manufacturer’s instructions. Mitochondrial haplotypes were defined on the basis of 420 bp scored sequence of the control region and compared in terms of pairwise nucleotide divergence (NEı and Jın 1989), using the Neighbor-Join- ing-method (SAıtou and NEı 1987) and the PHYLIP 3.5 c computer package (FELSENSTEIN 1993). A published sequence of a Black Sea harbour porpoise was used as an outgroup (RoseL et al. 1995). A median graph of the relationships between mitochondrial haplotypes was determined according to BAn- DELT (1992). As measures of genetic variation within and among populations, mean nucleotide diversi- ties within and between the two geographic areas were estimated (cf. Quinn and WHITE 1987), and haplotype diversities within populations were calculated (NEı and TAyımA 1981). 106 R. TIEDEMANN, J. HARDER, CHRISTINE GMEINER, and E. HAASE Results The harbour porpoises analysed comprised nine different mitochondrial haplotypes as de- fined by sequence patterns of the Control Region. These were defined arbitrarily as Pho I to Pho IX, differing from one another by one to four transitions (Fig. 2, Tab. 1). An analy- sis of haplotype sequence divergence by the Neighbor-Joining procedure suggested two major clusters A and B, containing Pho I, Pho III, Pho VII, Pho VIII (cluster A), and Pho II, Pho IV, Pho V, Pho VI (cluster B) (Fig. 3). The haplotype Pho IX was sorted to none of these clusters and showed the highest sımilarity to a specimen from the Black Sea (BS), which served as an outgroup. As shown in a median graph, each haplotype could be derived from at least one other haplotype by a single point mutation, and the clusters A and B could be defined by a C and T, respectively, at position 198 (Fig. 4). IIOOOOIUIOOCOIILLCOLIOOCOOOLLTIILIIITTI11172222222223338838 388, EEE 001122333445556666788889990112222344446912445677900111225556888 697837179130281247401456899020126901890897158707413789020153567 GTGTTGTATCTTAGTATTCACCTTCTESCT-CCACACATCCACCCGETCATAATCCTEETEET Gr: EIG, FRE Le >. 0.18C, oc Le SHE: E® Le BE J [ I (ep) rg @) HJ } m .Iie m . 1. oe Ian Seele ae +3 Fig. 2. Sequence polymorphisms in the different haplotypes of harbour porpoise found in this study, as compared to specimens from the North Pacific (NP) and the Black Sea (BS) (np 1 and bs 1 from Rose et al. 1995). Position numbers are indicated by three vertical digits. Table 1. Number of different nucleotides (above diagonal) and pairwise nucleotide divergence in % (below diagonal) between mitochondrial haplotypes of harbour porpoise. 0.71 0.48 0.24 0.71 0.48 MtDNA patterns in Harbour porpoises Phocoena phocoena 107 Pho IX Pho| Pho VII Pho VIII Pho III Pho Il Pho IV Pho VI Pho V BS Fig. 3. Sequence divergence among haplotypes of harbour porpoises from North and Baltic Sea (Neigh- bor-Joining-tree, based on nucleo- tide divergence according to NEı and Jın 1989). A specimen from the Black Sea population (BS; cf. Fig. 2) served as an outgroup. The geographic distribution of the haplotypes is given in table 2. The most common haplotypes Pho I and Pho VII were found in 62% of the animals. Over- all mean nucleotide diversity was 0.37%, overall hap- lotype diversity was 0.80%. The genetic variation of the Baltic Sea population in terms of within popula- tion nucleotide diversity was less than 50% of the var- iation in the North Sea population, and also the haplotype diversity was considerably lower in the Bal- tic Sea (Tab. 2). The difference in the haplotype com- positions of these two populations was highly significant (p = 0.0001; Fisher’s exact test): all Baltic specimens showed cluster A haplotypes (95%-confi- dence limits: 83%-100%). These had only a frequency of 45% in the North Sea (95%-confidence limits: 26%-69%). The frequency of the most common hap- lotype Pho I had an increasing tendency from 26% in the North Sea through 40% in the Western Baltic Sea to 60% in the Eastern Baltic area (the last value was based on a very small sample size, however). As a measure of the divergence between the North Sea and the Baltic Sea population, net nucleotide diversity be- tween them was estimated to be 0.13%. Net nucleo- tide diversity between the Western and Eastern subpopulation of the Baltic Sea was negligible (0.03%). Fig. 4. Median graph of the relationships among haplotypes in terms of nucleotide substitutions. Note, that the transition at position 17 is included twice, i.e. as two different vectors, to allow a three-dimen- sial realisation of the graph. 108 R. TIEDEMANN, J. HARDER, CHRISTINE GMEINER, and E. HAASE Table 2. Geographic distribution of mitochondrial haplotypes, nucleotide diversity (z) and haplotype diversity (ö) within populations of harbour porpoise. Haplotype Total North Sea Baltic Sea Western Eastern Baltic Sea Baltic Sea Pho I Pho II Pho III Pho IV Pho V Pho VI Pho VII Pho VII Pho IX Discussion For the total distribution range of the subspecies P. p. phocoena, i.e. the whole North Atlantic, mean nucleotide diversity in the control region was estimated to be 0.48%, and the West Atlantic population was found to be more diverse (0.58%) than the population of the East Atlantic (0.40%; all values recalculated from Roser et al. 1995, using the for- mulae of Quinn and WHITE 1987). Our value for North and Baltic Seas together (0.37%) was in good agreement with the value for the East Atlantic. Since all haplotypes except Pho I were only found either in the West or in the East Atlantic and net nucleotide diver- sity between these areas was high (0.28%; calculated for the combined data set of RoSEL et al. 1995 and this study), gene flow between these areas appears very limited or even ab- sent. The predominance of the two ubiquitous haplotypes Pho I and Pho VII, which dif- fered only by one nucleotide, suggests these types to be ancestral. This is corroborated by the results of a comparable study on eight harbour porpoises from Norway and the Dan- ish Skagerrak, containing 4 times Pho I, once Pho VII, and three additional haplotypes, of which two could be derived from Pho I by a single transition (Rose et al. 1995). More- over, PhoI is the only haplotype that has been found both at the East and the West Atlantic coasts (Roseı et al. 1995). On the contrary, the characteristic transition at posi- tion 198, defining the cluster B haplotypes, was found neither on the West coast of the Atlantic nor in the Black Sea (Rose et al. 1995) and may thus have arisen locally. Considering the status of the Baltic harbour porpoise population, three alternative hy- potheses concerning initial colonization and current extent of gene flow are to be dis- cussed: 1. North Sea and Baltic Sea are completely panmictic (cf. ANDERSEN 1972). We would expect the same haplotypes, occuring at similar frequencies, and similar nucleotide and MtDNA patterns in Harbour porpoises Phocoena phocoena 109 haplotype diversities within both populations. Net nucleotide diversity between the areas should be close to zero. 2. The Baltic Sea has been colonized once and then remained genetically isolated from the North Sea. Then, the Baltic Sea population could be expected to be genetically less diverse due to a persisting founder effect. The haplotype composition could be differ- ent as an effect of random genetic drift. 3. North Sea and Baltic Sea are inhabited by separated populations, but gene flow oc- curs occasionally. The Baltic Sea population might again be genetically less diverse, but its haplotype composition could represent a subset of the North Sea haplotypes. The significant differences found in haplotype composition and nucleotide diversity prompt us to reject the first hypothesis of total panmixia: Cluster B haplotypes are absent in the Baltic Sea, and the nucleotide diversity of the Baltic Sea population is only half that of the North Sea population. When considering the second hypothesis of complete isolation between the two populations, we may invoke the concept of the molecular clock (cf. Avıse 1994): Assuming that the part of the mitochondrial genome studied here may have diverged at a rate of about 15% per million years in an intraspecific comparison (RoseL 1992), the net nucleotide diversity suggests a divergence between Baltic and North Sea population about 8500 years ago. This coincides quite well with the age of the Baltic Sea as a brackish sea (Litorina-period; cf. LIEDTKE 1981). However, these values must be taken with caution, since both the divergence rates and the nucleotide diversities may contain stochastic errors. Nevertheless, these values provide some support for the second hypothesis. The haplotype composition of the two respective populations shows that all haplotypes found in the Baltic Sea were also present in the North Sea, except for Pho VIII, which was found only in the Eastern part of the Baltic Sea. This pattern could be explained by a colonization of the Baltic Sea by harbour porpoises with ubiquitous haplotypes (Pho I, Pho III, Pho VII). The low nucleotide and haplotype diversities in the Baltic Sea indicate a persisting founder effect. However, the Baltic population might be sufficiently large that it is not driven to haplotype uniformity by random genetic drift. Re- cent census data estimated a population size of about 50,000 specimens in the Baltic Sea, which is almost twice as high as in the Eastern part of the North Sea (Germany and Den- mark combined; Kınze 1994; HAmMonD et al. 1995). It should be noted, however, that the majority of the Baltic Sea population is located in the Kattegat area (HAamMmonD et al. 1995), which was not sampled in this study. Considering the third hypothesis, the significant differences in haplotype composition between the two populations do on a first glimpse not support the interpretation of con- siderable gene flow between them. If we apply a model of the relation between the amount of gene flow and the frequency of exclusive genetic characters (1. e. exclusive hap- lotypes), we get a rough estimate of the number (Nm) of individuals migrating between separated populations per generation (SLATKIN 1985). In this study, 55% of the North Sea and 10% of the Baltic Sea harbour porpoises have exclusive haplotypes, which gives an estimate for Nm = 0.01, i.e. one migrating specimen per 100 generations. Taking into ac- count that Harbour porpoises are not fertile until the age of three to four years (SORENSEN and KınzE 1993), this would correspond to only one migration event every several hundred years. This value apparently underestimates gene flow, since it would pro- pose only about 10 to 20 migration events since the beginning of the Litorina-period of the Baltic Sea. The Nm value is possibly biased towards an underestimation for mainly two reasons: 1. There is some evidence that there is a seasonal pattern in the population structure within the North Sea (AnDERSEN 1993; KınzE 1994). If there would be a season- ality both in the mitochondrial haplotype compositon in the North Sea and in the time of migration between North and Baltic Sea, the number of excluxive North Sea haplotypes may be overestimated, which consequently leads to an underestimate of Nm. 2. Mitochon- drial DNA is maternally inherited. Thus, our Nm estimate does not include immigration 110 R. TIEDEMANN, J. HARDER, CHRISTINE GMEINER, and E. HAASE of males. Despite these cautions on the Nm estimate, it is an indication for only limited genetic exchange between the Baltic and the North Sea population of harbour porpoise. The apparent genetic isolation of the Baltic Sea is surprising considering the known winter migration of Baltic harbour porpoises out of the Baltic (DupDok vAn HEEL 1962), where they are likely to meet North Sea individuals (Kınze 1994). However, fertilization occurs in July and August (MaHr-HAnsen 1954), when porpoises are in the Baltic Sea (SCHULZE 1987). Thus, the observed genetic segregation in the maternally inherited mito- chondrial DNA between North and Baltic Sea could be explained by philopatry, assuming that at least females return to their area of birth for fertilization. Differences between these areas are also suggested from investigations on isoenzymes, which are encoded in the nuclear genome and thus susceptible for gene flow in both sexes (ANDERSEN 1993). Moreover, morphological characters indicate population differentiation (KınzE 1985). Thus, we conclude that the Baltic population of harbour porpoise has been genetically isolated after a postglacial colonization and gene flow into this population is a rare event. Acknowledgements We thank Dr. HARALD BENKE, Dr. ROLAND Lick, Prof. Dr. JANUSZ MARKOWSKI, Dipl.-Biol. GERHARD SCHULZE, and Dr. URSULA SIEBERT for providing samples. We also thank Prof. Dr. DiETER KRUSKA for support and fruitful discussions during the progress of the study. Prof. Dr. Jan DuInkEr and Dipl.- Biol. HELGA DRÖFN HÖGNADOTTIR provided valuable comments on the manuscript. Financial support is acknowledged from the German Ministry of Research and Technology. Zusammenfassung Untersuchungen an der mitochondrialen DNA von Schweinswalen (Phocoena phocoena) aus Nord- und Ostsee Um das Ausmaß genetischer Differenzierung zwischen der Nord- und der Ostsee-Population des Schweinswals (Phocoena phocoena) zu analysieren, wurden 39 Individuen auf DNA-Sequenzun- terschiede in einem hochpolymorphen Abschnitt der mitochondrialen Kontrollregion untersucht. DNA wurde aus Leber- und Hautproben gestrandeter Tiere isoliert. Nach einer Amplifikation mit der Poly- merase-Kettenreaktion (PCR) und direkter Sequenzierung der Amplifikate wurden 420 Basenpaare ausgewertet. 9 mitochondriale Haplotypen wurden gefunden, die sich jeweils an ein bis vier Nukleotid- positionen durch Transitionen unterschieden. Auf der Grundlage einer spezifischen Substitution wur- den Haplotypen in zwei Gruppen (A und B) eingeteilt. Alle untersuchten Schweinswale aus der Ostsee zeigten Haplotypen der Gruppe A, die nur bei 45% der Nordseetiere gefunden wurden. Die genetische Variation, gemessen als Nukleotid- und Haplotypendiversität, war in der Ostseepopulation deutlich ge- ringer als in der Nordseepopulation. Aufgrund der Frequenzen gefundener Haplotypen sowie der Nu- kleotiddiversität zwischen den Populationen ist davon auszugehen, daß die Ostsee vor einigen tausend Jahren durch Schweinswale besiedelt wurde. Nach dieser Besiedelung war der genetische Austausch mit der Nordseepopulation wahrscheinlich sehr gering. Die in dieser Untersuchung gefundene genetische Differenzierung zwischen Schweinswalen aus Nord- und Ostsee steht im Einklang mit Literaturanga- ben zu Unterschieden in Schädelmerkmalen und Enzympolymorphismen. Literature AMANO, M.; MiYAZakrt, N. (1992): Geographic variation in skulls of the harbour porpoise, Phocoena pho- coena. Mammalia 56, 133-144. ÄNDERSON, S. (1972): Status over den danske hvalbestand. In: Status over den danske dyreverden. Ed. by B. BEnzons. Copenhagen: Zoologisk Museum, Pp. 239-242. MtDNA patterns in Harbour porpoises Phocoena phocoena ala ANDERSEN, L. W. (1993): The population structure of harbour porpoise, Phocoena phocoena, in Danish waters and part of the North Atlantic. Mar. Biol. 116, 1-7. ÄRNASOoN, Ü.; GULLBERG, A.; WIDEGREN, B. (1993): Cetacean mitochondrial DNA control region: se- quences of all extant Baleen whales and two Sperm whale species. Mol. Biol. Evol. 10, 960-970. Avıse, J. (1994): Molecular markers, natural history and evolution. New York, London: Chapman and Hall. BANDEILT, H. J. (1992): Generating median graphs from Boolean matrices. In: L1-Statistical analysis. Ed. by Y. Dopece. North Holland: Elsevier. Pp. 305-309. Dupok VAN HEEL, W. H. (1962): Sound and Cetacea. Neth. J. Sea Res. 1, 407-507. FELSENSTEIN, J. (1993): PHYLIP (Phylogeny Inference Package) version 3.5 c. Seattle: Department of Genetics, University of Washington. GAsKIn, D. E. (1984): The harbour porpoise Phocoena phocoena (L.): Regional populations, status, and information on direct and indirect catches. Rept. Int. Whal. Commn. 34, 569-586. HAMMOND, P.; HEIMLICH-BoORAN, S.; BENKE, H.; BERGGREN, P.; COLLET, A.; HEIDE-JORGENSEN, M. P.; LEO- POLD, M.; @IEN, N. (1995): The distribution and abundance of harbour porpoises and other small ce- taceans in the North Sea and adjacent waters. Cambridge: Final Report of the European Union project LIFE 92-2/UK-U27. Kınze, C. C. (1985): Intraspecific variation in Baltic and North Sea harbour porpoises (Phocoena pho- coena (L., 1758)). Vidensk. Medd. dansk naturh. Foren. 146, 63-74. Kınze, C. C. (1994): Phocoena phocoena (Linneaus, 1758) — Schweinswal. In: Handbuch der Säugetiere Europas. Ed. by D. RoBINEAU, R. Dusuy, and M. Krıma. Wiesbaden.: Aula-Verlag. Pp. 242-264. LIEDTKE, H. (1981): Die nordischen Vereisungen in Mitteleuropa, 2. ed. Trier: Zentralausschuß für deutsche Landeskunde. MoHL-Hansen, U. (1954): Investigations on reproduction and growth of the Porpoise (Phocaena pho- caena (L.)) from the Baltic. Vidensk. Medd. dansk naturh. Foren. 116, 369-396. Neı, M.; Jın, L. (1989): Variances of the average numbers of nucleotide substitutions within and be- tween populations. Mol. Biol. Evol. 6, 290-300. NEı, M.; TAyımA, F. (1981): DNA polymorphism detectable by restriction endonucleases. Genetics 97, 145-163. NowAK, R. M. (1991): Walker’s mammals of the world Vol. II, 5th ed. Baltimore, London: John Hopkins Univ. Press. Quinn, T. W.; WHITE, B.N. (1987): Analysis of DNA sequence variation. In: Avian genetics. Ed. by F. Cooke and P. A. BuckL£y. London: Academic Press. Pp. 163-198. RICHTERICH, P.; HELLER, C.; WURST, H.; PoHı, F.M. (1989): DNA sequencing with direct blotting elec- trophoresis and colorimetric detection. Bio Techniques 7, 52-58. Rose, P.E. (1992): Genetic population structure and systematic relationships of some small cetaceans inferred from mitochondrial DNA sequence variation. PhD thesis, Univ. California, San Diego. Rose, P.E.; Dızon, A. E.; HAyGoop, M. @. (1995): Variability of the mitochondrial control region in populations of the harbour porpoise, Phocoena phocoena, on interoceanic and regional scales. Can. J. Fish. Aquat. Sci. 52, 1210-1219. SAITOU, N.; NEı, M. (1987): The Neighbor-Joining method: A new method for reconstructing phyloge- netic trees. Mol. Biol. Evol. 4, 406-425. SCHULZE, G. (1987): Die Schweinswale. Wittenberg Lutherstadt: A. Ziemsen Verlag. SLATKIN, M. (1985): Rare alleles as indicators of gene flow. Evolution 39, 53-65. S@RENSEN, T. B.; Kinze, C. C. (1993): Reproduction and reproductive seasonality in Danish harbour porpoises Phocoena phocoena (L.). Ophelia 39, 159-176. Tomirin, A. G. (1967): Mammals of the U.S.S. R. and adjacent countries Vol. IX. Cetacea. Jerusalem: Is- rael Program for Scientific Translations. YURICK, D. B.; GAskIn, D. E. (1987): Morphometric and meristic comparisons of skulls of harbour por- poise Phocoena phocoena (L.) from the North Atlantic and North Pacific. OPHELIA 27, 53-75. Authors’ address: Dr. RALPH TIEDEMANN, JÜRGEN HARDER, Mag. CHRISTINE GMEINER, and Prof. Dr. EBERHARD HAASE, Institut für Haustierkunde, Christian-Albrechts-Uni- versität zu Kiel, Olshausenstraße 40, D-24118 Kiel, Germany Z. Säugetierkunde 61 (1996) 112-125 ZEITSCH RIFT © 1996 Gustav Fischer, Jena SÄUG ETI ERKÜI N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Behavioural and anatomical correlates of sympathetic arousal and stress in male Central American Agoutis (Dasyprocta punctata) By V. Korz, H. HENDRICHSs, and K. MILITZER Lehrstuhl für Verhaltensforschung, Fakultät für Biologie, Universität Bielefeld and Zentrales Tierlabora- torium, Universitätsklinikum Essen, Germany Receipt of Ms. 19. 04. 1995 Acceptance of Ms. 18. 12. 1995 Abstract Studied were distinctions and relationships between behavioural indicators of sympathetic arousal in twelve male Central American agoutis (six living in large groups and six in male-female pairs) in two so- cial situations (pair-bonded and non pair-bonded). Indicators were scaled by correspondence analysis; scrapemark, scentmark (in pair living males), bury and scrape indicating high, scratch, nibble-of-body- sides, lick and tremble low sympathetic activity. Pair-bonded males displayed low and non pair-bonded males showed high amounts of behavioural elements indicating high sympathetic arousal. Morphologic- anatomical data were used for parallel analysis of long-term stress reactions. Pair-bonded males showed low and non pair-bonded males high relative portions of adrenal medulla at the central section area of adrenal glands. This can be regarded as adaptation to long lasting high sympathetic activity in the non pair-bonded males along with considerable stress reactions (enlarged hearts and diminished testes and anal glands). Thus the classification of males by their amounts of behavioural indicators of sympathetic arousal corresponds to their classification by physiological indicators of continuous sympathetic activity. The differentiating recording and analysis of appropriate behavioural elements provide a non invasive method for detecting arousal states and stress in individual animals, even in their “everyday life”. Introduction In agoutis the social unit is a bonded pair in a defended area (SmyTHE 1978; DuBost 1988). In captivity the realization of social bonds is of central importance for the internal dy- namics of the individual (Korz and HENDRIcHs 1989; Korz 1991). One important variable of these dynamics is arousal, meaning both the produced levels of general arousal and the locally produced states of specific arousal, both of which the animal is able to regulate de- pending on the quality of its social relationships (Korz and HEnDRIcHs 1989). Social bonds (the integration of behavioural systems into a specific internal organization, BOWL- BY 1969) provide security (an internal state of an animal in correlation to the quality of in- ternal organization) and hence prevent escalated arousal (WicKLER 1976; HENDRICHS 1988), which stresses the organism. Lack of social attachments can lead to considerable stress reactions (HENRY and STEPHENS 1977). For assessments of individual arousal in the agouti, frequencies of specific behavioural indicators have been used (Korz and HENDRICHS 1989; Korz 1991). Findings in other ro- dent species indicate that these indicator elements are mainly correlated with sympathetic arousal. Application of adrenaline and noradrenaline produced higher frequencies of burying, scraping, scratching, and self grooming in rats (MAzEs 1979; HAwKıns and AVERY 1983; DE BoER et al. 1990). In male agoutis variations of amounts of burying, scraping and Behavioural and anatomical correlates of sympathetic arousal and stress 113 scentmarking displayed were correlated with variations in rhythmic and phasic compo- nents of body temperature (Korz and HEnDkricHs 1995). Males displaying decreased amounts of these elements when socially challenged, show decreased diurnal oscillations and increased mean body temperature. Hightened amounts were correlated with a more regular oscillation and decreased body temperature. In the generation of circadian oscilla- tions of several physiological parameters in mammals the sympathicus is involved (REuss 1993). The sympathoadrenomedullary system is mainly activated in animals which cope actively with social challenges while the pituitary adrenocortical system is more active in anımals with inhibited behavioural activity (Weıss et al. 1970; HEnRY and STEPHENS 1977; Houst et al. 1983; SACHSER and Lick 1989). The present study is based on data of twelve males, six living in larger groups and six in male-female pairs. The main aims were to elucidate relationships and distinctions be- tween the different indicators of arousal, and to classıfy the males using the amounts of indicators they displayed. Anatomical and histological data were included for parallel analysis of long-term stress reactions. Material and methods Animals Observed were three larger groups of agoutis, descendants of animals captured in Guatemala and bred at the Department of Ethology since 1974 (Tab. 1). For the first group (Group 1, consisting of 12 ani- mals: the male M1, four adult and two subadult females and five young animals) data were collected from June 1985 until February 1986 (120 hrs of observation), for the second group (Group 2, living since 1988 in the same enclosure and consisting of four males: M 23 (grown up in group 1), M 81, M 102, M 312, in that sequence of superiority; and six females) from November 1988 to May 1989 (56 hrs of ob- servation). The third group (Group 3, consisting of one male: M 5, introduced into the group in July 1989, and five females) was observed from March to April 1991 (35 hrs of observation). Pair groups, liv- ing together for 6.5 months to 8.5 months, were observed from August 1990 until November 1992. Housing conditions The large groups were kept in roofed outdoor enclosures of 6.75 m to 9.00 mx 4.65 m with three adjoin- ing wooden sleeping cabins each of 1.75 mx1.10 m which were heated in winter. Pair groups were kept in adjacent roofed outdoor enclosures of 4.65 mx 2.25 m, each with an adjoining video-monitored Table 1. Label, age, body weight and group membership of the observed males. Age [years. months] Body weight [g] Group M 1271 M 458 M 658 M 2218 M 21118 M 1571 GL: group living, PL: pair living 114 V. Korz, H. HEnDRICHs and K. MILITZER sleeping cabin (size as above). The cabins were illuminated by an electric bulb (75 watt) from one hour before observation to 15 min after the end of session. The enclosures contained branches and wooden or stony refuges where the animals were visible. The ground consisted of gravel covered with sand. The animals were exposed to natural light-dark-cycle. The ambient temperature varied from -10°C to +30 °C for the large groups and from +6 °C to +26°C for pair groups. The animals were fed on corn, rolled oats, fresh vegetables, fruits and leaves once a day (dry food and water ad libitum). Behavioural data The behaviour of animals was recorded by written entries in one-minute intervals as scan samples at 6- 12 a.m. (large groups) and 17-21 p.m. (large and pair groups) when all animals were active. For each pair group a first quantitative observation phase (prephase) of five consecutive days in their home cages was followed by a testphase after the animals were set in a similar, but strange enclosure. After five days the pairs were set back into their home cages. For each phase ten hours of behavioural protocols were analyzed. The following behavioural elements were recorded: Bite-attempt: A biting movement in the direction of another animal while sitting, standing or quickly approaching. Bury: The animal takes a piece of food in its mouth, digs a hole into the ground with its forefeet, places the piece of food in the hole and covers it up, pressing the substrate with its forefeet. Chase: The animal runs after another one for at least two meters. Chin-raise: A male approaches a female with raised chin. Drive: A male follows a female with raised chin for at least two meters. Follow: A male follows a female in a distance of up to two meters. Head-raise: Standing in front of or parallel to another animal at a distance of up to one meter the ani- mal having raised its head stares at the other. Head-thrust: The anımal thrusts its raised head at another. Joust: Standing opposite on their hind legs two anımals drum on each other with their forelegs. Lick: The animal raises a forefoot and licks it. Nibble: The animal sits at the side of another nibbling the other’s head region from ear to corner of the mouth. Nibble-of bodysides: The animal nibbles at its flanks. Nuzzle: The animal nuzzles the corner of another’s mouth (mostly when the latter one has a piece of food in its mouth). Scentmark: The animal drags its anal glands over the ground leaving a secretion. Scrape: The animal scrapes the ground with its forefeet (can change into burying and vice versa). Scrapemark: After scentmarking the animal scrapes the ground with its hindlegs. Scratch: The animal scratches its flanks with its hind leg, turning its body. Sniff: The animal sniffs at parts of another’s body. Tremble: Brief body-trembling with hair erected. Urine-spray: A male approaches a female, raises its body and sprays a jet of urine in the direction of the female. Anatomical data All the males of Group 2 died within 30 days (May to June 1989): M 1 died in February 1986. The bodies, covered in plastic, were frozen at -20°C within a few hours after death. In October 1989 the bodies were dissected after 18 hrs of thawing and organs were histologically and microbiologically ex- amined. Body weights did not differ before freezing and after thawing. Hearts, kidneys, adrenal glands, testes and anal glands were weighed and fixed in 4% neutral formalin. For each body side, adrenal glands were divided in the largest sectional area parallel to the longitudinal axis. After imbedding in Paraplast", several sections were produced and stained with hematoxylin-eosin. The size of adrenal cor- tex and medulla for each body side was measured at a magnification 600x by planimetry, and the tissue area was determined in mm’. In November/December 1991 four of the pair group living males were killed and dissected within an hour in the same manner as described above. The legal requirements for this procedure have been adhered to. Behavioural and anatomical correlates of sympathetic arousal and stress 1415 Statistics The behavioural data were analyzed by correspondence analysis (CA), the morphological data by prin- cipal component analysis (PCA). These multivariate techniques are eigenvector methods of ordination developed to describe a set of intercorrelated variables in terms of a smaller number of factors and de- termines components of association between variables and populations. CA extracts components of an x°-“distance”-measure of association for frequency tables (Hırı 1973), whereas PCA does this in a simi- lar way for Euclidean distances and continued data. Eigenvalues separate the total information into r factors. For each factor a set of row and column coordinates is computed. Distances of dots from the ori- gin indicate the distance from total profile of row and column vectors. Associated dots can be deter- mined by neighbourhood along the axes (vertically or horizontally, not diagonally). CA has the advantage of simultaneous plotting of individual males and behavioural variables (amounts of be- havioural elements). The analyses were carried out using the PCA and P4M paragraphes of the statisti- cal computer program package BMDP (version 1990). Differences in group means of behavioural data were evaluated by use of the Mann-Whitney-U-test, of anatomical data by use of the Welch’s t-test (each two-tailed). The criterion of significance was set atp < 0.05. Results Male-female relationships Each group living male showed a preference for a specific female (defined as directing courtship behaviour more than twice as much towards a specific female than on average towards every female; Korz 1991). The assessment of a social relationship, especially of a pair-bond, is more difficult in pair living than in group living males, because there is no possibility to separate the qual- ity of relationship between pair mates towards other conspecifics. Exclusivity of behav- iour between two animals not restricted to times of estrus indicates a pair-bond (LAMPRECHT 1973; WICKLER 1976). Mutual and exclusive nibble was shown by the group living males M 1 and M 23 and specific females. Therefore, only these males (of the group living ones) are taken to be pair-bonded. The behavioural profiles of contact (nibble, nuz- zle, follow, and sniff) and agonistic behaviour (chase, joust, head-thrust, head-raise, and bite-attempt) the pair living males displayed towards their females were compared to those of the group living males directed towards their preferred females. Behavioural amounts were analyzed separately for pair and group living males using the CA. The co- ordinates of all males for the first two axes were graphically displayed in figure 1. Spatial neighbourhood of pair living males to the pair-bonded group living males M 1 and M 23 indicates a similarity of behavioural profile. It is assumed that behavioural similarity indi- cates a similarity of relationships towards the females. Therefore, the pair living males M2218 and M658 are taken to be pair-bonded. A clear coordination of M 1571 and M 21118 to the division pair-bonded/non pair-bonded is not possible with the present data. All other males are regarded to be non pair-bonded. Arousal-indicating behaviour In a first step for evaluation of relations and differences of the arousal-indicating behav- ioural elements, the intercorrelations between variables were computed (Tab. 2). There are high correlation coefficients between bury and scrape on the one hand and scrape and the other variables on the other hand. Because of statistical dependency of data more sig- nificant correlation coefficients are to be expected than for the comprehension of the es- sential variance contributing variables are necessary. For reducing the variables to a single factor explaining most of the variability of data the profiles of these behavioural elements for each male were analyzed by CA, and the coordinates of the single elements were pro- =\ N ON V. Korz, H. HEnDRICHSs and K. MILITZER Axis 2 [Score] -1.6 -1.2 -0.8 -0.4 0 0.4 Axis 1 [Score] Fig. 1. Scores (y°-distances to the centroid) of group (N = 6, solid frames) and pair (N = 6, broken frames) living males for the correspondence analyses of amounts of contact and agonistic behavioural elements they displayed towards their respective females. NIGM: Nibble occuring in groups, NIPM: Nibble occurring in pairs, *: pair-bonded males, +: non pair- bonded males, ?: no coordination possible. jected on the first axis. The relative distances of elements from the origin of all elements and between coordinates of single elements along the axis are regarded as a scaling along a common dimension. Following the findings mentioned in the introduction, that occur- rence and the frequencies of these behavioural elements are correlated to the sympathetic activity, the underlying dımension is seen as the sympathetic tone. In figure 2 three situations are shown: an analysis for all males (group living males and pair living males in the prephase) without the elements scentmark and scrapemark and the analyses for pre- and testphase of the pair-living males considering these behav- ioural elements. Besides the fact that scrapemark was not displayed by group living males there is another reason for this differentiation: Scentmark is attached to territorial beha- viour. It is shown by dominant males regularly while patrolling through their enclosure. Separating this kind of marking from spontaneously and arousal-indicating scentmarking is possible in the pair living males with comparable social status (“territory-holder with subordinated female”). Thus scentmark did not serve as arousal-indicating behaviour for group living males. The order of behavioural elements along the axis is following objec- tive criteria, whereas the direction of axis depends on assumptions by the observer. The most simple assumption is that for a given sympathetic tone the probability of occurrence of a specific behavioural element presupposing a lower sympathetic tone is higher than Table 2. Spearman rank-correlation coefficients for behavioural indicators of arousal. 0.85* LI 0.38 0.60* Os 07705 0.81* 0.78* 0.83* BU: Bury, SP: Scrape, SC: Scratch, NB: Nibble-of-bodysides, LI: Lick, TR: Tremble. *: Critical values (df = 10, p < 0.05, two tailed). Behavioural and anatomical correlates of sympathetic arousal and stress 117 for an element presupposing a higher sympathetic activity. The frequencies of behavioural elements and their corresponding axis scores should be negatively correlated. In the case of positive correlations the direction of axis was inversed. This procedure only affects the sign of the scores (BonD 1992). The three analyses are significantly indicating high prop- ability of dependence between behavioural profiles along the specific axes (for all males x: 1328.5, df: 55, p< 0.001, two axes explaining 82.9% of the variability of data, axis 1: 62.4%, axis 2: 20.6%; for pair living males in the prephase x”: 1374.5, df: 35, p < 0.001, two axes explaining 93.9% of the variability of data, axis 1: 60.8%, axis 2: 33.1%; and in the testphase y”: 1810.1, df: 35, p < 0.001, two axes explaining 93.7% of the variability of data, axis 1: 50.3%, axis 2: 42.9%). The quasi-independence model allows the scaling of behavioural data of all males independently of their social situation. For more detailed in- formation cf. Korz et al. (1995). For an assessment of the sympathetic tone the behaviour- al elements bury, scrape and nibble-of-bodysides closest correlated with the first axis (Tab. 3), and the first two ones distinctly separated from the other elements are of special importance. Nibble-of-bodysides indicates a low, bury and scrape a high sympathetic tone. For an animal displaying bury frequently the sympathetic tone is assessed to be higher than for an animal showing bury less frequently. In the prephase, under usual conditions, the element scrapemark points to a higher sympathetic tone than bury (hardly separated, not very evident); scentmark supplies no further information besides scrape. In addition, scentmark and scrapemark are not well represented on the first axis (Tab. 3). In the analy- sis of the testphase, under social conditions that stresses the animals, the marking be- haviours are clearly separated from bury and scrape, indicating a very high and possibly escalated sympathetic activity. Along with the increasing importance of the marking be- havioural elements they are now fit to the first axis (Tab. 3). The meaning of scrape and its correlation with the first axis decrease. High correlation coefficients indeed only imply how good the respective elements can substitute the combination of all elements. In order to get an appropriate interpretation of the axis the linear combination of all variables has to be taken into account (Harrıs 1985; JAMEs and McCurrocH 1990). For pair living males displaying high amounts of scrapemark and bury a higher sympathetic tone is as- sessed than for males showing these elements less frequent, or, in the case of scrapemark, not at all. It was not an aim to construct a numerical scale, but to arrange the behavioural elements with respect to relative importance for an assessment of sympathetic activity. NBSSTR SCe Lie SP sBl a) NERSSTR SCeLlle SMesP EU „PM b) LieeeNB/SCETR SPe BU SMe® PM®c) Axis 1 [Score] Fig. 2. Scaling of behavioural indicators of sympathetic arousal by correspondence analysis. a): analysis for all males (N = 12) group living males and pair living males in prephase, b): analysis for pair living males (N = 6) in prephase, c): analysis for pair living males (N = 6) in testphase (ten hours each). Given are the y”-distances to the centroid. BU: Bury, SP: Scrape, SC: Scratch, NB: Nibble-of-bodysides, LI: Lick, TR: Tremble, SM: Scentmark, PM: Scrapemark. 118 V. Korz, H. HEnDRICHs and K. MILITZER Table 3. Correlation of profiles of behavioural indicators of sympathetic arousal and behavioural pro- files of individual males with the first two axes (Al, A2) of correspondence analysis. Group living Pair living males all males males FTSE BISSCHEN] KESRLIE IT TÜREN FETT TEE Prephase Testphase SP SC NB JE TR SM PM M 23 M 81 M 102 M 312 MI MS5 M 1271 M 458 M 658 M 2218 M 21118 M 1571 Correlation: ratio between squared length of the projection on the axis from the origin and the squared distance from the origin in all dimensions (the signs result from those of coordinates). For abbreviations see Fig. 2. In a second step the association of individual males to specific behavioural elements was analyzed. Figure 3 shows the results of the analyses for group (Fig.3a) and pair (Fig. 3b) living males (group living males x: 350.8, df: 25, p < 0.001, two axes explaining 89.6% of the variability of data, axis 1: 67.7%, axis 2: 21.9%; for pair living males see above). The classification of group living males is not explainable by the differences in age (Tab. 1) but by their different social situation (Fig. 1). The non pair-bonded males are associated with elements which point to a high sympathetic arousal like bury and scrape (group living males) as well as scrapemark, bury and scentmark (pair living males), whereas the pair-bonded males are associated with low sympathetic arousal indicating be- havioural elements like lick, nibble-of-bodysides and tremble. M 2218 is arranged be- tween the low sympathetic tone indicating element lick and the higher sympathetic tone indicating scrape (Fig. 3b). | To summarize the analysis of behavioural indicators of sympathetic arousal gives the result that pair-bonded males, whether living in large groups or in pair groups, show on the average a lower level of sympathetic arousal than non pair-bonded males. Comparing the frequencies of behaviour indicating high sympathetic arousal (bury, scrape and scentmark) between group and pair living males, there is an unexpected result (Fig. 4). Pair living males show significantly higher frequencies of these behaviours than group living males. One should assume that the life in large groups with high rates of ago- nistic encounters is more stressful than life in pair groups with a subordinated female. Behavioural and anatomical correlates of sympathetic arousal and stress 119 o 025 ° © m, N 0 n x < -0.25 -0.5 -0.5 -0.25 0 0825 055 a) Axis 1 [Score] [3 8: \B% = M658* TR N (7) = -0.7 < 14 -2.1 -0.7 -0.35 0 085 b) Axis 1 [Score] Fig. 3. Results of correspondence analyses of amounts of behavioural indicators of sympathetic arousal for group living (a) and for pair living (b) males. Labels for males are framed. Given are the x-dis- tances to the centroid. *: pair-bonded males, +: non pair-bonded males, ?: no coordination possible. For abbreviations see Big=2. This result points out that because of the very stressfull life in group living males not mainly the sympathicus adrenomedullary system but also the pituitary adrenocortical sys- tem is highly activated effecting the behavioural output. Additional data presented in the next paragraph contribute to this explanation. Morphologic-anatomical indicators of long-term stress Group living males showed significantly higher relative weights of adrenal glands (0.29 +/-0.10, Fig.5) and lower body masses (2776.0 +/- 175.0g) than the pair living males (arithmetic average of relative weights of adrenal glands: 0.14 +/- 0.02, t= 3.08, p = 0.03, df=4.5; average of body masses: 3 268.7 +/- 251.5 g, t=-3.33, p = 0.02, df=5.2; Welchs’ t-test). The absolute weights of adrenal glands are also significantly higher in group living (0.79 +/- 0.26 g) than in pair living (0.47 +/- 0.09 g, t = 2.53, p = 0.05, df = 5.2) males, while testes weights are higher in pair living (9.42 +/- 0.80 g) than in group living 120 V. Korz, H. HENDRICHs and K. MILITZER N/h BZ Group living DO Pair living Bury Scrape Scentmark Fig. 4. Frequencies of bury, scrape and scentmark of group living (N = 6) and pair living (N = 6, pre- phase) males. Median with interquartile range. N/h: average amount of behavioural element per hour. Statistics: Mann-Whitney-U-test, two tailed, *: indicates significant differences: U = 0, p < 0.01; n. s.: not signifi- cant. A n.S. 94 D zE 817 OT © s 5 ZB Group living Den n.S. |O Pair living BES DD MD 1 0 N DZZZPPFZZZE N. Heart Kidney Adrenal Testes Anal glands glands Fig. 5. Relative weights of organs of group (N =5) and pair (N = 4) living males. Arithmetic average and standard deviation. Statistics: Welchs’ t-test, two tailed, *: p = 0.03; n. s.: not sig- nificant. Relative weight in °/y0 of body weight. (6.21 +/- 2.22 g, t=-2.99, p = 0.03, df=5.2) males (given are the arithmetic average and standard deviation). The absolute weights of hearts, kidneys and anal glands did not differ significantly between group and pair living males. For individual profiles of organic parameters the data were analyzed by principal com- ponent analysis (Fig. 6). Factors are orthogonally rotated by the Varimax-method. Two factors explain 65.3% of data variance (first factor: 36.4%, second factor 28.9%). The first factor accounts for the difference between males in respect of the relative portions of adrenal cortex and medulla at the central section area of adrenals, indicating that these parameters mainly correlate together, not with a third one. The second factor is highly loaded by weights of hearts (positively) and anal glands (negatively) showing that males with high heart weights exhibit small anal glands. It depends on the condition of M 102 that heart weigths and relative portions of adre- nal medulla did not highly load on the first factor with same sign. This is the only non pair-bonded male exhibiting a small portion of adrenal medulla at the central section area of adrenals. At the same time he is the most stressed male indicated by the fact that he shows the highest relative weight of adrenals (0.38) of all males and by additional findings of autopsy: several stomach ulcerations, an enlarged spleen and, along with the highest weight of heart of all males, a myocarditis. MıLITZER and REINHARD (1982) observed an Behavioural and anatomical correlates of sympathetic arousal and stress 121 Heart# M102 “ Adrenal glands ® Mm312] 4 di Factor 2 Ad.medulla# * Ad.cortex# Factor 1 Fig. 6. Results of principal component analysis of relative weights of organs ["/oo of body weight], cen- tral section areas of adrenals [mm?] and relative portions (%) of adrenal medulla and adrenal cortex at the central section area of adrenals of group (solid frames) and pair (broken frames) living males. Factor loadings of parameters and factor scores (x°-distances to the centroid) for individual males. Per- centages are arcussinus-transformed. Ad. medulla: adrenal medulla, Ad. cortex: adrenal cortex, *: pair- bonded males, +: non pair-bonded males, #: indicates significant loadings, critical value for df = 7, p < 0.05: 0.666, two tailed. increased adrenal cortex size along with a reduced adrenal medulla size in stressed rates. In the other males there were found: fat depots in abdominal cavity (M 81), spermatogo- nial giant cells in the testes of M 81 and M 102, pointing to degenerative processes prob- ably effective in enhanced fertility (MıLıTzEr and Rasım 1990). The separate analysis for the pair living males gives high positive loadings of heart weights and portions of adrenal medulla on the first factor. The arrangement of individual scores of the males shows that pair-bonded males associate with weights of testes and anal glands, suggesting a specific “male status’ and low stress. Proximity to the relative portion of adrenal cortex results from their corresponding low relative portion of adrenal medulla. M 1 and M 23 show the lowest absolute portions of adrenal medulla of all group living males. There were no symptoms of an increased activation of the pituitary adrenocortical system in pair-bonded males. The arrangement of non pair-bonded males points to effects of chronic stress, the males associated to relative portion of adrenal medulla at the central section area of adre- nals (besides M 102) indicate continuously increased sympathetic activity with increased cardiovascular activity (M 458, M 81, M 312, M 102). Discussion The main question was aimed at the possible use of the relationships and distinctions of different behavioural indicators of arousal for a classification of the animals with regard to their sympathetic activation. The results were: a) behavioural elements indicated differ- ent levels of sympathetic arousal, b) frequencies of behavioural elements displayed by in- dıvidual males differed according to their social situation (pair-bonded males displaying lower frequencies of high sympathetic arousal indicating behavioural elements than non pair-bonded males), c) classification of the males by behavioural amounts corresponded to their classification by anatomical and histological parameters (for the pair-bonded males indicating low and for the non pair-bonded males indicating high sympathoadreno- medullary activity). 902 V. Korz, H. HENDRICHS and K. MILITZER The usual procedure is to combine different behavioural elements to more complex behavioural reactions, e.g. attack or wıthdrawal, for which a common motivation is as- sumed. The single elements may reflect different intensities of the underlying motivation, e.g. different states of escalated aggressiveness (BArLow and BArLin 1976; Bonn 1992). By analyzing the behavioural reaction of an animal its motivational state can be charac- terized (BAERENDS 1976). Similarly the different behavioural elements, the single ele- ments reflecting different levels of sympathetic arousal, can be combined to characterize the arousal state of an animal. Behavioural reactions can result from a mixture of differ- ent motivations. Display behaviour for example can be produced by sexual and by agonis- tic motivations, each accompanied by specific arousal processes. The elements were chosen for the following reasons: a) indicators should occur frequently enough to exclude chance influences, b) elements should not occur in immediate reactions towards conspeci- fics (to avoid direct social motivations), c) they should be comparable to elements found in other rodent species. The combination of elements is also significant because of the high intercorrelations. A direct connection between sympathetic activity and the occurrence of this behav- iour complex is not supposed, only that the complex hormonal and neural mechanisms leading to this behavioural output correlate with sympathetic arousal. Moreover, the as- sessment of arousal and stress using behavioural indicators is possible in a social field of regularly reacting animals, not in (pathologically) inactive animals. The increased, but not disorganized or even stereotypic, specific motoric activity of non pair-bonded males can be seen as a coping response to elevated arousal. The possibi- lity of displaying organized patterns of behaviour can serve as a behavioural release me- chanism that can reduce damaging physiological effects (ConnEr et al. 1971). Such mechanisms are described e.g. for body temperature in the agouti (Korz and HENDRICHS 1995), for the cardiovascular activity in rats (FOKKEMA and KooLHAASs 1985). LEVINE and WIENER (1988) found lower plasma cortisol titers during high frequencies of vocalization in young squirrel monkeys (Saimiri sciureus), and VANDERWOLF and VANDERWART (1970) and Erriott (1975) registered lower rates of heartbeat during increased frequencies of self-grooming in the rat. The pair living non pair-bonded males are in a position to cope actively with the social challenge of living and breeding with a female without a pair- bond. The group living males are even able to live closely with several conspecifics of both sexes. This capability is also found in group living females to such an extent that stress could be held low and reproduction remained possible (Korz 1991). Possibly the regulation of arousal challenges the non pair-bonded animals in a specific way. Compen- sating coping responses lead to behavioural and physiological hypertrophies and in the long-term to increased stress. The higher masses of adrenal glands and the lower body weights in all group living males in relation to the pair living males point to higher stress of the first independent of dominance status and of whether they are pair-bonded or not. Possibly this can be ex- plained by the particular challenges of group life with frequent social perceptions and ac- tivations. Among the group living males, however, the subordinated and non pair-bonded males are more stressed than the pair-bonded M 1 and M 23. The results in figure 6 and the additional findings mentioned point to an increased activity of the pituitary adreno- corticalsystem (M 102 and M 312), an inhibition of the pituitary gonadal system for all non pair-bonded males (leading to infertility in M 102 and M 81) and a considerable acti- vation of the sympathoadrenomedullary system. The strong stress of the non pair-bonded, group living males can be seen as a result of subordination and the frequent agonistic be- haviour of dominant animals together with the lack of a pair-bond. The comparable posi- tions of non pair-bonded group and pair living males in figure 6 indicate that mainly the absence of a pair-bond leads to the described coping reactions. With regard to the finding that pair living males in spite of lower stress show significantly higher frequencies of high Behavioural and anatomical correlates of sympathetic arousal and stress 123 sympathetic activity indicating behaviour than group living males, the following explana- tion is suggested: Along with an increased activity of the antagonistic vagus as an adapta- tion to chronice stress in group living males, the increased activation of pituitary adrenocortical system, indicated by the higher weights of adrenal glands, with high titers of cortisol possibly reduces the basal adrenomedullary activity. SZEMEREDI et al. (1988) found this effect during chronical glucocorticoid treatment in rats. Reflexive sympatho- adrenomedullary activation was not affected. As mentioned above, high cortisol titers mainly are found in animals which hardly activate in behaviour, whereas a sympathoadre- nomedullary activation is found in actively coping animals. HEnrY and STEPHENS (1977) report an increased activation of both systems during chronic stress. An adjusting balance between these antagonistic activities may explain the lower frequencies of bury, scrape and scentmark in group living males. In pair living males the sympathetic tone accom- plishes the production of higher frequencies of the behavioural elements. Thus the assess- ment of arousal and stress by behavioural indicators of an individual animal have to consider the living conditions. The physical and social environment are relevant for under- standing physiological reactions during stress, as pointed out by Horst (1977). The differentiating recording and analysis of appropriate behavioural elements pro- vide a non invasive method for detecting arousal states and stress in individual animals. Data on different physiological parameters (e.g. heart rate) may help to evaluate further which patterns of behaviour can reflect specific conditions of the anımals. Zusammenfassung Verhaltens- und anatomische Korrelate zur Sympathikusaktivität und zu Belastungen bei männlichen Mittelamerikanischen Agutis (Dasyprocta punctata) Es wurden Daten von sechs männlichen Mittelamerikanischen Agutis mit Weibchen in drei größeren Gruppen und sechs Männchen in 3/?-Paaren in zwei unterschiedlichen sozialen Situationen (paarge- bunden und nicht-paargebunden) analysiert. Geeignete Verhaltensindikatoren für die Aktivität des sympathischen Nervensystems (Futtervergraben, Scharren, Kratzen der Körperseiten, Knabbern der Körperseiten, Lecken der Vorderpfoten, Schütteln, Markieren und Scharrmarkieren) wurden mittels Korrespondenzanalyse hinsichtlich der Höhe des ihnen jeweils unterliegenden Sympathikotonus ska- liert. Es zeigte sich, daß Futtervergraben und Scharren einen hohen und unter akut belastenden Bedin- gungen Markieren und Scharrmarkieren einen sehr hohen Sympathikotonus anzeigen. Paargebundene Männchen zeigten geringe und nicht-paargebundene Männchen hohe Mengen der einen hohen Sym- pathikotonus anzeigenden Verhaltenselemente. An fünf gruppenlebenden und vier paarlebenden Männchen konnten morphologisch-anatomische Daten für die Analyse langfristiger Belastungsreaktio- nen erhoben werden. Paargebundene Männchen zeigten niedrige und nicht-paargebundene hohe rela- tive Anteile der Nebennierenmedulla an der zentralen Anschnittfläche der Nebennieren. Dieser Befund wird als Anpassung an langfristige hohe Sympathikusaktivität gesehen. Einhergehend mit die- ser Aktivität, zeigten sich deutliche Stressreaktionen bei nicht-paargebundenen Männchen: erhöhte Herzmassen sowie verringerte Hoden- und Analdrüsenmassen. Die Klassifikation der Männchen an- hand der Verhaltensparameter korreliert mit derjenigen anhand der physiologischen Parameter für langfristig erhöhte Sympathikusaktivität. Die differenzierte Erfassung und Analyse geeigneter Verhal- tenselemente versprechen einen nicht-invasiven Zugang zur Erfassung von Erregtheitszuständen und Belastungen bei Tieren auch unter „Alltagsbedingungen“. References BAERENDS, G. P. (1976): The functional organization of behaviour. Anim. Behav. 24, 726-738. BARLOWw, G. W.; BALLin, P. J. (1976): Predicting and assessing dominance from size and coloration in the polychromatic midas cichlid. Anim. Behav. 24, 793-813. 124 V. Korz, H. HENDRICHs and K. MILITZER BOER DE, S. F.; SLANGEN, J. L.; VAN DER GUGTEN, J. (1990): Plasma catecholamine and corticosterone levels during active and passive shock-prod avoidance behavior in rats: Effects of Chlordiazepoxide. Physiol. Behav. 47, 1089-1098. BowLsy, J. (1969): Attachment and loss. Vol. 1. Attachment. New York: Basic books, Inc., Publishers. Bon, A.B. (1992): Aggressive motivation in the midas cichlid: Evidence for behavioral efference. Be- haviour 122, 135-152. Dusosr, G. (1988): Ecology and social life of the red acouchy, Myoprocta exilis,; comparison with the or- ange-rumped agouti, Dasyprocta leporina. J. Zool. (London) 214, 107-123. COoNNER, R. L.; VERNIKOS-DANELLIS, J.; LEVINE, S. (1971): Stress, fighting and neuroendocrine function. Nature 234, 564-566. ELLIOT, R. (1975): Heart rate, activity, and activation in rats. Psychophys. 12, 298-304. FOKKEMA, D. S.; KoOLHAAS, J.M. (1985): Acute and conditioned blood pressure changes in relation to social and psychosocial stimuli in rats. Physiol. Behav. 34, 33-38. HARRIS, R. J. (1985): A primer of multivariate statistics. New York: Academic Press. HAwKINS, M. F.; AvEry, D. D. (1983): Effects of centrally-administered bombesin and adrenalectomy on behavioral thermoregulation and locomotor activity. Neuropharm. 22, 1249-1255. HENDRICHS, H. (1988): Lebensprozesse und wissenschaftliches Denken. Freiburg, München: Alber. HENRY, J. P.; STEPHENS, P. M. (1977): Stress, health, and the social environment. A sociobiologic ap- proach to medicine. New York, Heidelberg, Berlin: Springer. Hırr, M. ©. (1973). Reciprocal averaging: An eigenvector method of ordination. J. Ecol. 61, 237-249. Horst von, D. (1977): Social stress in tree-shrews: problems, results, and goals. J. Comp. Physiol. 120, 71-86. Horst von, D.; FucHs, E.; STÖHR, W. (1983): Physiological changes in male Tupaia belangeri under dif- ferent types of social stress. In: Biobehavioral bases of coronary heart disease. Ed. by T. M. DEMBROWwSKI, T. H. SCHMIDT, and G. BLÜMCHEN. Basel: Karger biobehavioral medicine series. Vol. 2, 332-390. JAMES, F. C.; McCULLocH, C. E. (1990): Multivariate analysis in ecology and systematics: Panacea or Pandora’s box? Annu. Rev. Ecol. Syst. 21, 129-166. Korz, V. (1991): Social relations and individual coping reactions in a captive group of Central American Agoutis (Dasyprocta punctata). Z. Säugetierkunde 56, 207-218. Korz, V.; HENDRICHS, H. (1989): Einflüsse der sozialen und räumlichen Eingebundenheit auf Aktivie- rungsfähigkeiten bei mittelamerikanischen Agutis, Dasyprocta punctata (GRAY, 1842). Zool. Beitr. NSES321 51172. Korz, V.; HENDRICHS, H. (1995): Spontaneous behavior and body temperature in male Central Ameri- can Agoutis (Dasyprocta punctata) under different social conditions. Physiol. Behav. 58, 761-768. KoRrZz, V.; SCHADE, U.; LAUBENSTEIN, U.; HENDRICHS, H. (1995): Artificial neural networks in the analysis of behavioural topology. Naturwiss. 82, 479-481. LAMPRECHT, J. (1973): Mechanismen des Paarzusammenhalts beim Cichliden Tilapia mariae Boulenger. Z. Tierpsychol. 32, 10-61. LEVINE, S.; WIENER, S. G. (1988): Psychoendocrine aspects of mother-infant relationships in nonhuman primates. Psychoneuroendocrinology 13, 143-154. Mas, H. (1979): Time course and transition analysis of the behavioural effects of microinjection of pen- tobarbital and noradrenaline into the ventromedial hypothalamus of the rat. Behav. Process. 4, 341— 358. MILITZER, K.; RAsım, R. (1990): Riesenzellen und Atrophie des Tubulusepithels im Hoden von Gold- hamster, Ratte und Aguti. Berliner u. Münchener Tierärztl. Wochenschrift 103, 432. MILITZER, K.; REINHARD, H.-J. (1982): Rank positions in rats and their relations to tissue parameters. Physiol. Psychol. 10, 251-260. Reuss, S. (1993): Das Werk der inneren Uhr. Zur Neuroanatomie des circadianen Systems der Säuger. Naturwiss. 80, 501-510. SACHSER, N.; LicK, C. (1989): Social stress in guinea pigs. Physiol. Behav. 46, 137-144. SMYTHE,N. (1978): The natural history of the Central American Agouti (Dasyprocta punctata). Wa- shington: Smithsonian contributions to Zoology 257. Smithsonian Institution Press. SZEMEREDI, K.; BAGDy, G.; STULL, R.; CALOGERO, A. E.; KoPin, 1. J.; GOLDSTEIN, D. S. (1988): Sympatho- adrenomedullary inhibition by chronic glucocorticoid treatment in conscious rats. Endocrinol. 123, 2585-25. Weiss, J. M.; STONE, E. A.; HARELL, N. (1970): Coping behavior and brain norepinephrine level in rats. J. Comp. Physiol. Psychol. 72, 153-160. Behavioural and anatomical correlates of sympathetic arousal and stress 125 VANDERWOLEF, C. H.; VANDERWART, M.L. (1970): Relations of heart rate to motor activity and arousal in the rat. Canad. J. Psychol./Rev. Canad. Psychol. 24, 434-441. WICKLER, W. (1976): The ethological analysis of attachment. Sociometric, motivational and sociophysio- logical aspects. Z. Tierpsychol. 42, 12-28. Authors’ addresses: Dr. VOLKER Korz, Institut für Zoologie, Martin-Luther-Universität Halle-Witten- berg, Domplatz 4, D-06099 Halle/S., Prof. Dr. Dr. HUBERT HENDRICHS, Lehrstuhl für Verhaltensforschung, Fakultät für Biologie, Universität Bielefeld, Post- fach 100131, D-33501 Bielefeld, PD Dr. KLAus MILITZER, Zentrales Tierlabora- torium, Universitätsklinikum Essen, Hufelandstr. 55, D-45122 Essen, Germany Z. Säugetierkunde 61 (1996) 126-128 FÜR © 1996 Gustav Fischer, Jena SÄUG ETI ERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Buchbesprechungen NYHART, L. K.: Biology takes form. Animal morphology and the German universities 1800-1900. Chica- go und London: University of Chicago Press 1995. 414 S., 7 Abb., 9 Porträts und 3 tab. Anhänge. Paper- back: 19,85 £/24,95 $, ISBN 0-226-61088-8; clothbound: 59,95 £/75,- $, ISBN 0-226-61086-1. Die Autorin des vorliegenden Werkes, eine amerikanische Wissenschaftshistorikerin, welche an renom- mierten amerikanischen Universitäten und in Heidelberg tätig war, bietet eine bewunderungswürdige Materialfülle und unterzieht diese einer gründlichen Analyse. Sie wertet dabei vor allem die in deutschen Landes-, Universitäts- und Fakultätsarchiven verfügbaren Informationen aus, doch stützt sie sich auch auf persönliche Erinnerungen bedeutender Morphologen, sowie auf Nachrufe, auf hinterlas- sene Schriften und Korrespondenzen. Im ersten Kapitel des Buches (,„Situating Morphology“) wird in einer Art „Vorspann“ der Leser zu- nächst verblüfft durch eine auf den ersten Blick steril erscheinende Gruppierung der deutschen Mor- phologen des 19. Jahrhunderts in sechs „Kohorten“, die nicht nur in sechs unterschiedlichen und voneinander gut abgesetzten Perioden geboren wurden, sondern auch in verschiedenen Zeitspannen ihre wissenschaftliche Hauptwirkung an den Universitäten der deutschen Territorien entfalteten. Es stellt sich bei der Lektüre des Werkes schnell heraus, daß die erwähnte Zuordnung der Morphologen zu Alterskohorten den Weg für eine übersichtliche Darstellung der Entwicklungen Öffnet. Die Geschichte der Morphologie ist im Deutschland des 19. Jahrhunderts eng mit der Geschichte der Universitäten verbunden, ganz besonders in der zweiten Hälfte des Jahrhunderts, mit der sich das Buch schwerpunktmäßig befaßt. Nach der oben erwähnten Charakterisierung der Morphologie werden im ersten Hauptabschnitt des Buches (,‚Morphology and Physiology“) in zwei Kapiteln die Beziehun- gen des Faches zur Physiologie und zur allgemeinen Entwicklung der biologischen Wissenschaften geschildert. Der zweite Hauptteil („Evolutionary Morphology, 1860-1880“) beschäftigt sich in vier Ka- piteln mit der Periode nach dem Erscheinen der „Entstehung der Arten“ von CHARLES DARWIN. Die Aktivitäten der Evolutionsmorphologen ERNST HAECKEL und CARL GEGENBAUR werden besonders aus- führlich beschrieben. HAEcKEL war zwar ein bedeutender Verbreiter der Vorstellungen DARwıns, durch seine Polemik sowie wegen seiner Reserviertheit gegenüber experimenteller, insbesondere auch histolo- gischer Arbeitsweise, konnte er seine Vorstellungen langfristig nicht durchsetzen. Er bot seinen Studen- ten nicht die Möglichkeit zum Erwerb praktischer und technischer Fähigkeiten. CARL GEGENBAUR wirkte als Anatom in medizinischen Fakultäten, zunächst, wie HAEcKEL, in Jena und ab 1873 in Heidel- berg. Er schuf zwar eine Schule junger Anatomen und widmete sich intensiv dem anatomischen Unter- richt, doch wegen seiner engen fachlichen und menschlichen Verbindung mit Ernst HAEcKEL erfuhr auch GEGENBAUR heftigen Widerspruch. Ihm wurde durch JAKOB HENLE und WILHELM WALDEYER VOTge- worfen, er betreibe keine deskriptive, sondern eine evolutionsbiologisch orientierte Morphologie. Es wurde bezweifelt, daß er für die Ausbildung junger Mediziner, die sich für ihre spätere Tätigkeit möglichst ein transparentes Bild des menschlichen Körpers erarbeiten sollten, der geeignete Lehrer sei. Da befürchtet wurde, daß die Ausrichtung der Forschung sich auch auf die Schwerpunkte des Unter- richts auswirken könne, sollte die evolutionsmorphologisch orientierte vergleichende Anatomie mö- glichst an die philosophischen Fakultäten delegiert werden. GEGENBAUR konnte nur einen beschränkten Erfolg erzielen, da die in medizinischen Fakultäten arbeitenden Anatomen sich vom Problem der Evo- lution ab- und sich Fragestellungen zuwandten, denen sie für die Ausbildung junger Mediziner größere Bedeutung zumaßen. Im abschließenden dritten Hauptabschnitt des Buches (,„Morphology and Biology, 1880-1900“) werden in vier Kapiteln die beiden letzten Jahrzehnte des 19. Jahrhunderts mit Ausblicken bis zum I. Weltkrieg behandelt. Die Auseinandersetzungen der Phase von 1860 bis 1880 im Zusammenhang mit der Durchsetzung des Darwinismus traten in den Hintergrund. Die Erweiterung des Forschungsspek- trums und die Gunst der politischen Situation während der Gründerzeit hatten eine große Zahl von An- stellungsmöglichkeiten nicht nur für in medizinischen Fakultäten arbeitenden Anatomen, sondern auch für Morphologen in zoologischen Instituten geschaffen. Ab etwa 1880 trat die experimentelle For- schung, beispielsweise in der Embryologie, sowie kausalanalytisch-morphologische Aspekte in den Vor- Buchbesprechungen 127] dergrund. In den letzten beiden Dekaden des Jahrhunderts wurde der Stellenmarkt für junge Morpho- logen so stark eingeschränkt, daß Wissenschaftler die Universitäten verlassen mußten und u.a. ins Lehramt an Schulen auswichen. Die Autorin kommt dank ihrer detaillierten Analysen zu dem interes- santen Schluß, daß die beinahe mystische Verklärung der deutschen Universität als Urquell („foun- tainhead‘) der wissenschaftlichen Forschung in dieser Form nie begründet gewesen sei. Der Unterricht stellte immer die zentrale Aufgabe („fundamental mission“) der deutschen Universitäten dar. Die For- schung verdrängte die Wissenschaft nie, sie wurde vielmehr auf diese „aufgepfropft“. Mit Schmunzeln stellt der Leser fest, daß sich die Grundsituation der Morphologie an den Hochschulen in einhundert Jahren nur wenig verändert hat! Der vorliegende Band wird ergänzt durch Porträts bedeutender Morphologen (SIEBOLD, GEGEN- BAUR, HAECKEL, SEMPER, LEUCKART, WEISMANN und EHLERS), sowie durch drei interessante tabellarische Anhänge. Eine Liste der Zoologen und (humanmedizinischen) Anatomen nennt die Lehrstühle, welche von 1810 bis 1918 an deutschen Universitäten besetzt waren und gliedert deren Inhaber in die o.g. „Kohorten“ ein. In den beiden folgenden Tabellen wird für die Zeit von 1850 bis 1918 die Besetzung der Lehrstühle für Zoologie und für Anatomie in 20 Universitäten zusammengestellt. Ein ausführliches Literaturverzeichnis sowie ein Register schließen den Band ab. Diesem Buch ist eine weite Verbreitung zu wünschen! Es wäre auch höchst erfreulich, wenn eine deutsche Übersetzung dieses interessanten, ja spannenden, Werkes erscheinen könnte. Nicht nur an Form und Funktion interessierte Biologen und Mediziner werden sich durch diesen Band angesprochen fühlen, auch an der Geschichte der Naturwissenschaften, sowie an Mentalitätsgeschichte interessierten Lesern gewährt Lynn K. NyHART Einsichten in die Entwicklung der biologischen Gedankenwelt, sowie in die äußere und innere Entwicklung der deutschen Hochschulen im 19. Jahrhundert. P. LANGER, Gießen KRAFT, R.: Xenarthra. Handbuch der Zoologie. Bd. VII, Mammalia, Teilbd. 59. Berlin, New York: Wal- ter de Gruyter u. Co. 1995. 80 S., 60 Abb., 260,- DM. ISBN 3-11-014428-X Im vorliegenden Beitrag werden die fossilen und rezenten Vertreter der Ordnung Xenarthra in ihrer Biologie und taxonomischen Zuordnung gekennzeichnet. Nach einer kurzen, einleitenden Diagnose werden zunächst gemeinsame morphologische und anatomische Merkmalskomplexe behandelt, ins- besondere wird dabei auf Haut und Hautorgane, Schädel, Gebiß, postcraniales Skelett, Gehirn, Verdauungstrakt, Urogenitalsystem und Embryonalentwicklung eingegangen. Der anschließende systematische Teil beginnt mit einer Zuordnung der Familien zu den 3 Unterordnungen Cingulata, Tar- digrada und Vermilingua, Abstammung und Verbreitungsgeschichte werden diskutiert unter Einbezie- , hung von Eurotamandua joresi aus dem Eozän der Grube Messel, dem bislang einzigen gesicherten fossilen Ameisenbären außerhalb von Amerika. Es folgen ausführlichere Schilderungen über die Be- sonderheiten der Arten. Die Beschreibungen sind durch zahlreiche Abbildungen und Zeichnungen von guter Qualität ergänzt. Die Verbreitung der rezenten Arten ist in eindrucksvollen Verbreitungskarten dokumentiert. Der Beitrag ist entsprechend der systematischen Gliederung konsequent konzipiert und klar, knapp und kritisch geschrieben. Der momentane Kenntnisstand über diese Säugetierordnung ist erschöpfend belegt, so daß eine beeindruckende Dokumentation vorliegt. Dieser Teilband kann jedem Interessenten ‚ empfohlen werden. Der etwas hohe Preis wird durch die sehr gute Ausstattung verständlich. | D. Kruska, Kiel I 128 Buchbesprechungen Buchbesprechung ENGELHARDT, W. VON; LEONHARD-MAREK, S.; BREVES, G.; GIESECKE, D. (Eps): Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction. Proceedings of the eighth Symposium on Ruminant Physiology. Stuttgart: Ferdinand Enke Verlag, 1995. 626 pp., numerous illustrations and tables. 248,— DM. ISBN 3-432-26851-3. This book publishes the papers presented during an international symposium held in Willingen, Ger- many, in the autumn of 1994. Following an introductory chapter, 29 articles are grouped into nine “parts” and one “special lecture”. The first section deals with the regulation of digestion (three chap- ters), in the subsequent four chapters of the second part data on absorptive and secretory mechanisms are presented. The current state of knowledge concerning the control of feed and water intake is dealt with in the following four chapters. The functionally important advances in ruminant microbiology are presented in two articles. Environmental, ecological and nutritional aspects can be found in the fifth sec- tion. In the latter one of the 1995 winners of the Nobel Prize in chemistry, P. J. CRUTZEN, discusses the role of methane in atmospheric chemistry and climate. Considering global aspects, CRUTZEN demon- strates that about 30% of the total anthropogenic CH, per year is released by domestic animals; about 75% of this gas produced by domestic stock comes from cattle alone. In the sixth section of this book ruminant intermediary metabolism is dealt with (three chapters). The following two parts are closely related to animal production: Four chapters deal with the regulation and improvement of reproduction, lactation and growth of ruminants. In the following three articles growth of fetal and young ruminants is discussed. Under the headline “Comparative Physiology” (part 9) information on the sites and modes of action of melatonin in sheep is presented and thermore- gulation and water balance in the dromedary is compared with the same parameters in pecoran rumi- nants. In a “special lecture” P. J. Van SOEST, E. S. DIERENFELD, and N. J. ConkLın deal with the digestive strategies and limitations of ruminants. In a final section the reader is presented short reports from five workshops, all dealing with special aspects of ruminant physiology. The book is concluded by a detailed index of eight pages. A wealth of information on the physiology — especially on nutritional and digestive physiology - is presented and thoroughly discussed in this volume. For the reader with mammological interests, how- ever, it is disappointing that the book starts with an unbalanced article on the incorporation and role of domestic ruminants in early rural societies, authored by AnGELA VON DEN DRIESCH. A readership inter- ested in the problems of applied animal breeding and animal production as well as in a wide field of physiological and biochemical problems, needs as prerequisite a presentation of the comparative as- pects of the problem of domestication, including a biological approach. For example, the following types of questions might be asked: What are, generally speaking, the essential differences between wild, domestic and feral animals? What are the morphological or ethological consequences of domestication? How is it possible that a domestic animal can be relatively easily handled, as compared to its wild rela- tives? Why does a domestic animal that is able to “run wild”, not become a “wild animal” again, but a feral one? In her chapter A. von DEN DRIEScH deals with differentiations of body size and emphasizes results of archaeological investigations on domestic ruminants. These findings are mainly taken from her own studies and those of her mentor, Prof. J. BoEsSNEcK. She almost completely avoids those biolo- gical questions that make domestic animals such fascinating subjects of research! P. LANGER, Gießen Instructions to Authors Submission and Acceptance of Manuscripts: Manuscripts for publication should be sent to the managing editor, Prof. Dr. D. Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany. Acceptance of the manuscript follows the bylaws of the German Society for Mammalogy (Deutsche Gesellschaft für Säugetierkunde). Receipt of tne manuscript will be confirmed immediately by mail, and as soon as the peer reviews are received the authors will be informed concerning the decision for acceptanc?. All correspondence dealing with details of production should be sent to: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena, Germany. Form of the Manuscript: Each manuscript should be submitted in triplicate, one original and two copies, in English or German in DIN A4 format either from a typewriter or legible computer print with a left-hand margin of 4cm and true double spacing. The first page of the manuscript should contain the following information: title name(s) of author(s): full first name for female authors, first initials only for male authors department and university affiliation In case of long titles, a running title not exceeding 72 characters should be provided. 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. Names of authors cited in the text are to be underlined in pencil. Botanical and zoological generic and species names should be underlined in pencil with a wavy line, whereby the name of the scientist who first described the species is not underlined. Authors should indicate in the margins the approximate location in the text for illustrations and tables. At tthe end of the manuscript following References, the exact postal address(es) of the author(s) should be given. Content of the Manuscript: Manuscripts can be published as original investigations or short communications. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 typewritten, double-spaced 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. If written in German, an English title should precede the English abstract; if written in English, a German title should precede a German summary. b) Short communications: Short communications must not exceed 5 typewritten, double-spaced pages and do not require either an Abstract or Summary. If written in German, an English title should be given on page 1; if written in English, please give a German title on page 1. Short communications need not be headed with subtitles into In- troduction, Materials and methods, Results, and Discussion but should be organized according to this form. Ilustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focussed and read- ily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. La- belling on figures should be uniform, 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 should 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 DIN A4 format sheets of paper. References: Each paper cited in the text must appear in alphabetical order at the end of the text. Only internation- ally available literature should 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; HANDLEY, C. O. Jr. (1994): Evolution, biogeography, and description of a new species of Fruit-eat- ing bat, 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 und F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/I, 7-28. Printing: 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 under- taken 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 paper 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. 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). Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective, or for resale. For copying from back volumes of this journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 100537, D-07705 Jena; Telefon (03641) 6263, Telefax (03641) 62 65.00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A. Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zeitschriftenvertrieb, Frau B. Dressler, Post- fach 10 05 37, D-07705 Jena; Telefon (03641) 62 64 44, Telefax (0 3641) 62 65.00. For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 4288201. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2945,-/382,50 (plus postage). Single issue price (DM/ÖS/SFr): 80,-/592,-/77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 86010090; Deutsche Bank, Konto- Nr. 3907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400.00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Giagia-Athanasopoulou, Eva B.; Markakis, G.: Multivariate analysıs of morphometric characters in the Eastern hedgehog Erinaceus concolor from Greece and adjacent areas. — Multivarıate Analyse morphometrischer Merk- male beim Ostigel Erinaceus concolor in Griechenland und benachbarten Gebieten rer ID Langer, P.: Comparative anatomy of the stomach of the Cetacea. Ontogenetic changes ınvolving gastric proportions — mesenteries — arteries. — Vergleichend-anatomische Untersuchungen am Magen der Cetacea. Ontogenetische Proportionsveränderungen — Mesenterien — Arterien 140 Kurt, F.; U Mar, Khyne: Neonate mortality in captive Asian elephants (Elephas maximus). — Neonatensterblichkeit bei asiatischen Elefanten in Menschenobhut 4155 Fischer, K.: Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) [he course of puberty ın male fallow deer (Cervus dama) 165 Ganem, Guila; Alibert, P.; Searle, J. B.: An ecological comparison between standard and chromosomally divergent House mice ın Northern Scotland Ein ökologischer Vergleich zwischen Standard- und chromosomal abweı chenden Hausmäusen in Nordschottland 176 Wissenschaftliche Kurzmitteilung Kock, D.; Ebenau, C.: The desert hedgehog, Paraechinus aethiopicus (Ehrenberg, 1833), new to the fauna of Syria Der Wüstenigel Paraechinus aethiopicus (Ehrenberg, 1833), neu für dıe Fauna Syrıens | 189 Mitteilungen der Gesellschaft , 192 Indexed in Current Contents Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database > U SIAV ISSN 0044-3468 FISCHER vs. 12:2 BONIS ARTIBUS SEMPER JENA »STUTTGART»NEW YORK Juni 1996 1 996 ZEITSCHRIFT FÜR 4 a je INTERNATIONAL JOURNAL &° OF MAMMALIAN BIOLOGY Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kıel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel - R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - ©. AnneEE. Rasa, Bonn -H. Reichstein, Kiel — M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt - P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) - W. Herre, Kiel (1962-1966) - H. Frick, München (1972-1976) -— M. Röhrs, Hannover (1977-1981) — H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amntierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H. G. Erkert, Tübingen - W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel - Marialuise Kühnrich, Hamburg Z. Säugetierkunde 61 (1996) 129-139 u FÜ R © 1996 Gustav Fischer, Jena SÄUG ETI ERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Multivariate analysis of morphometric characters in the Eastern hedgehog Erinaceus concolor from Greece and adjacent au SUN] IAN H \ JUL 30 1996 ) LIBRARIES Receipt of Ms. 19. 06. 1995 Acceptance of Ms. 06. 12. 1995 By EvA B. GIAGIA-ATHANASOPOULOU and G. MARKAKIS Section of Animal Biology, Department of Biology, University of Patras, Greece and Department of Biology, University of Crete, Iraklio, Greece Abstract A number of subspecies of the eastern hedgehog Erinaceus concolor has been described in the southern Balkan peninsula. Despite extensive morphological and karyological descriptions, their taxonomic sta- tus is still unclear. In the present study we have attempted to identify this status in different populations from mainland and insular Greece and to clarify their relationships to subspecies from adjacent areas. Multivariate analysis of 21 somatic and cranial characters was performed on 191 specimens from Greece, and also on published data (76 from Bulgaria, 21 from Yugoslavia and 9 from Asia Minor). Our analysis showed that throughout the mainland of the southern Balkan peninsula, one large- sized subspecies, E. c. bolkayi, is distributed. Two insular subspecies of smaller body size are found on the Greek islands: FE. c. nesiotes on Crete, Kyklades islands and Ionian sea islands and F. c. rhodius in the eastern Aegean sea islands Rhodos, Samos, and Chios. The subspecies E. c. transcaucasicus also ap- pears to occur on the islands of Lesvos and Kos. Introduction The eastern hedgehog Erinaceus concolor has an extensive distribution in eastern Europe and west Asia: Poland, Czech Republic, Slovakia, Austria, north-eastern Italy, Balkan peninsula, Turkey, southern Russia, west side of the Caspian Sea and western Siberia as well as in Israel, northern Iraq and northern Iran (CorgEr 1988; HoLz and NIETHAMMER 1990). In this vast area, the hedgehog displays such a degree of variation, particularly in colour and size, that a considerable number of subspecies has been described. A melanic form Erinaceus concolor concolor (or E. c. ponticus) is distributed on the eastern shore of the Black Sea and a pale form E. c. palidus in west Siberia (OGNEV 1928; CorBET 1988). Using a multivariate analysis of skull measurements, Horz (1978) showed that apart from the western and eastern hedgehog, a third type exists in south-eastern Europe (Ro- mania-Bulgaria) with a smaller cranial capacity and a longer cranium. It is not certain whether this type belongs to the species E. concolor, to the subspecies E. c. transcaucasicus (Satunin, 1905), or to a third species not yet described. KRrATocHVIL (1980) states that the hedgehogs inhabiting lowland steppes of the Balkan peninsula have a smaller mean capacity of the skull cavity and belong to the form de- scribed as E. c. drozdovskii (MARTINO 1933). Thus, according to KrATOCHVIL (1980) the European continent is populated by two separate subspecies of Erinaceus concolor differ- ing in somatic and craniological characters i.e. roumanicus and drozdovski. The form drozdovskii has been described as a “morpha” of the subspecies Erinaceus roumanicus 130 EvA B. GIAGIA-ATHANASOPOULOU and G. MARKAKIS bolkayi (MARTINO 1930), being lighter in colour and larger in size and distributed in the mountaneous regions of Montenegro, Yugoslavia (MARTIno 1933; PETROV 1940). Some authors considered bolkayi and drozdovskii two separate subspecies of the eastern hedge- hog, larger in body sıze than the subspecies FE. c. roumanicus (MARKoV 1957; MARKoV and DoBrIJANoV 1974; MALEC and STORCH 1963; Durıc and Murıc 1967; OnDkrıas 1966). Ac-ı cording to WETTSTEIN (1966) the hedgehogs from Asia Minor belong to E. c. transcaucasi- cus (Satunin, 1905). He considered it as a subspecies synonymous with drozdovskii, which : belongs to Anatolian populations. This subspecies has also been found on Kos (DE BEAUX 1929): YUGOSLAVIA >» BULGARIA Fig. 1. Distribution map of the subspecies of Erinaceus concolor in Greece. e=E. c. bolkayi, A =E. c.nesiotes, 8 = E. c. rhodius, * = E. c. transcaucasicus. Continental Greece: a = Attica (aı = Koropi, a, = Parnitha), b = Boiotia (Thebes), e = Evoia (e = Aliveri, &z = Edipsos), | = Fthiotida (Lamia), f = Fokida (f, = Lidoriki, f, = Tichio), g = Aetoloakarnania (Agrinio), i = Epiros (loannina), s = Thessalia (sı = Karditsa, s> = Trikala, s; = Kalambaka, s, = Elassona), m = Makedonia (m; = Florina, m» = Alexandria, m; = Kilkis, my = Serrae), t = Thrace (tı = Alexandroupolis, t = Soufli). Peloponni- sos: 1 = Kastritsi (Patra), 2 = Pyrgos, 3 = Kiparissia, 4 = Astros (Kinouria), 5 = Korinthos. Crete: 1 = Chania, 2 = Rethymno, 3 = Iraklio, 4 = Agios Nikolaos, 5 = Ierapetra. Multivariate analysis of morphometric characters in Erinaceus concolor 131] An extensive karyological investigation of hedgehogs in Greece has shown that three subspecies of the eastern hedgehog E. concolor are distributed throughout the country (GIaGIA and OnDRIAS 1980). The large-sized hedgehog of the Balkan peninsula was first de- scribed in the mainland as E.c. drozdovskü. However, a recent morphological study showed that it should be denominated E. c. bolkayi (Giacıa and OnDrias 1995). An insular subspecies, E. c. nesiotes (Bate, 1905), is found on the island of Crete and is karyotypically identical to the mainland form, while a second insular subspecies E. c. rhodius (Festa, 1925) is found on the island of Rhodos and differs karyotypically from the other two subspecies (GıaGIA and OnDkrıas 1980). Some hedgehogs from the island of Vis in Yugoslavia have been described under the name E. c. nesiotes (DuLıc and Tvrrkovic 1979). In the present study, we attempted to give a global view of the morphometric relation- ships among the populations of E. concolor throughout Greece and to clarify their rela- tionships with the neighbouring subspecies that have been previously described. Material and methods The present analysis was performed on 191 specimens constituting eleven populations collected from the mainland of Greece and its numerous islands (Fig. 1). Only adults were used to minimize the effect Fig. 2. Measurements taken on skulls of Erinaceus concolor (GSL = greatest skull length, CBL = condylobasal length, BL = basal length, ZB = zygomatic breadth, MB = mastoid breadth, . IC = interorbital constriction, NL = nasal length, RB = rostrum breadth, PB = postorbital breadth, , PW = palate width, UT = upper toothrow, ML = mandible length, LT = lower toothrow, HS = height of | skull, PL = palate length, HRM = height of ramus mandibulae, RA = rostral angle). 192 EvA B. GIAGIA-ATHANASOPOULOU and G. MARKAKIS of allometric variation associated wıth growth. Twentyone measurements, four bodily (HB = head and body length, TL = tail length, EL = ear length, and HFL = hind foot length) and seventeen cranial mea- surements (Fig. 2), were selected for the present analysis. In addition, two indices that have been used for the study of the hedgehog (WETTSTEIN 1942; 1966; RUPRECHT 1972) were calculated: maxillary index (Mi=a-b/a-c) and length index (Li= CBL/ZB). Additional data from Bulgaria (MArKov 1957), Yugoslavia (DuLıc and TvrTkovıc 1979; MARTINO 1933; PETROV 1940) and Asia Minor (WETTSTEIN 1966) were used for comparative purposes. A list of all populations including the measurements used is given in table 1. Cranial measurements were taken as defined by HrABE (1976) with a precision of 0.1 mm or 0.5 degree in the case of the ros- tral angle (RuPRECHT 1972). Data were subjected to canonical (discriminant) analysis (CA) and principal components analysis (PCA) after their logarithmic transformation. MANOVA was used to detect differences between popu- lations (Morrison 1976; REYMENT et al. 1984) and Mahalanobis distances between population centroids in order to construct a UPGMA phenogram (SnEATH and SoKAL 1973). The statistical packages SYSTAT (Wiırkınson 1987) and NT-SYS (RoHLF 1992) were used throughout this study. Table 1. Populations, sample sizes and variables (in common) studied. Description Code Sample size Variables used Continental Greece HB, TL, HFL, EL, GSL, CBL, BL, ZB, MB, IC NEI/RBIBBSBLNAUE ML, LT, HS, PL, HRM, RA Peloponnisos % Crete (E. c. nesiotes) Rhodos (E. c. rhodius) Lesvos Kyklades islands Kerkyra Kephalonia Kos Samos Chios Bulgaria (E. c. drozdovskii) HB, TL, HFL, EL, GSL, CBL, ZB, MB, IC, NL, RB, PB, UT, ML, LT 2 3 4 3 6 7 8 9 10 Bulgaris (E. c. bolkayi) Bulgaria (FE. c. roumanicus) Yugoslavia (E. c. drozdovskii) HB, TL, HEEJEE,GBEFZBSU/F IMIIE BIS Yugoslavia (FE. c. bolkayi) HB, TL, HFL, EL, CBL, ZB, MB, NL, RB, UT, ML, LT ” Yugoslavia (E. c. nesiotes) Asia Minor (E. c. franscaucasicus) CBL, ZB Results Preliminary analysis A preliminary analysis included the pooling of sexes and the pooling of specimens from all mainland Greek localities. Sexual dimorphism was tested in the largest sample from Peloponnisos (34 males and 49 females) and was not significant (MANOVA, p > 0.05). As the number of specimens collected from different localities of the mainland is rather low, they were pooled and tested by the inspection of the principal component plots (the first two, accounting for 53% of the total variability, are shown in Fig. 3). No Multivariate analysis of morphometric characters in Erinaceus concolor 1133 pattern of variation was found among the specimens from these localities. Similar results were obtained for other component plots (not shown). Thus, the pooled specimens from the Greek mainland will be referred to as the Continental Greek population. (CG in Tab. 1). 0.15 0.09 0.03 2nd PC -0.03 -0.09 -0. 15 -0. 15 -0.10 -0.05 0.00 0.05 0. 10 Ist PC Fig. 3. First and second principal component (PC) plot of the continental Greek population (CG). Bor- ders of the largest subpopulations (Macedonia and Thessalia) are also plotted for comparative pur- poses. Abbreviations as in figure 1. Canonical analysis (CA) CA of Greek populations From the eleven Greek populations studied (Tab. 1) the first five were considered as main populations because their samples comprise a sufficient number of individuals and repre- sent the known subspecies of Frinaceus concolor distributed in Greece (except for those from the island of Lesvos). The six further populations that come from differenf islands of Greece all contain a small number of specimens and one cannot be certain to which sub- species they belong. The main Greek populations were subjected t0o CA. MANOVA was significant (Wilk’s A = 0.0352, F = 8.44, p< 0.0001) as well as all pairwise contrasts be- tween them (p<0.0001). Canonical scores are plotted in figure 4. It is clear that the first canonical variate (CV) discriminates the insular from the mainland populations while the second one discriminates the Cretan population (E.c. nesiotes) from that of Rhodos (E. c. rhodius). The population from Lesvos of an uncertain taxonomic status, lies inbe- tween of all (Fig. 4). From the canonical coefficients in table 2 it can be concluded that CBL, ZB, BL, PL, and RA (Ist CV) are the measurements playing an important role in the discrimination of mainland from insular populations, while UT, LI, CBL, GSL and PB are those that mostly discriminate Rhodos and Cretan populations (2nd CV). 134 EvA B. GIAGIA-ATHANASOPOULOU and G. MARKAKIS 2nd CV Ist CV Fig. 4. First and second canonical variate (CV) plot of the five main Greek populations (nos. 1-5). Ab- breviations as in table 1. Table 2. Standardized canonical coefficients for canonical analysis of the five main Greek populations (1-5). First CV accounts for 65.6% and second for 17.3% of the total between group variability. lst CV 2nd CV 3rd CV. 4th CV. -0.153 -0.306 0.352 0.440 0.321 0.034 -0.112 0.102 0.192 0.063 -0.314 0.261 0.540 0.664 -0.050 0.630 0.352 0.615 0.344 -0.056 0.180 0.161 0.470 0.349 0.378 0.618 -0.054 -0.096 -0.139 -0.215 0.019 0.334 -0.160 0.206 0.778 0.017 0.504 -0.333 0.136 -0.025 0.347 0.075 1.243 0.348 0.113 0.179 -0.818 -1.221 0.533 0.177 1.144 0.042 -0.306 0.468 -0.397 0.529 0.455 0.442 -0.119 0.065 0.073 -0.276 0.111 Multivariate analysis of morphometric characters in Erinaceus concolor 135 The insular populations (no. 6-11 in Tab. 1) were subsequently classified into the main classes and the results are shown in table 3. It is clear that the population from Kyclades (KY as well as those from the Ionian islands, i.e. Kephalonia (KP) and Kerkyra (KE), are all classified into the Cretan (CR) population. The population from Samos (SA) is mostly classified in the population from Rhodos (RH). The populations from Chios (CH) and Kos (KO) have uncertain positions. Note that the specimens from Kos are karyologically closely related to Rhodos population (MAnDAHL 1978). Table 3. Subsequent classification of the small Greek insular populations (6-11) into the main cate- gories (1-5). Abbreviations as in table 1. This classification was based on the Mahalanobis distances from each specimen to the centroids of the main populations. CA of the Bulgarian populations Canonical variate plot of the three Bulgarian subspecies (Fig. 5) described by MARKOV (1957) reveals that £. c. roumanicus and E. c. drozdovskü are clearly different subspecies (MANOVA, p < 0.0001), while E. c. bolkayi is an intermediate form significantly different from E. c. roumanicus (p = 0.005) but not from E. c. drozdovski (p = 0.022). 2nd CV -5 -3 -1 1 3 5 Ist CV Fig. 5. First and second canonical variate (CV) plot of the three Bulgarian populations (nos. 12-14). Abbreviations as in table 1. 136 EvA B. GIAGIA-ATHANASOPOULOU and G. MARKAKIS CA of Greek and Bulgarian populations Using the 15 measurements (common to all Greek and Bulgarian material, see Tab. 1) an 8-group CA including both main Greek and Bulgarian populations was made. The UPGMA tree (Fig. 6) based on Mahalanobis distances revealed that the Greek mainland populations (CG and PE) are different from all the Bulgarian populations (MANOVA, p<0.001). Insular Greek populations clearly form another cluster, with the exception of the island of Lesvos which is closer to the continental Greek populations. 32 24 16 8 0 Te Sf a En Fee ee a a a u u Zr Fig. 6. UPGMA phenogram of the five main Greek (nos. 1-5) and the three Bulgarian (nos. 12-14) po- pulations, based on the Mahalanobis distances between centroids. Abbreviations as in table 1. Comparison of neighbouring populations from southern Balkan Peninsula and Asia Minor The Yugoslavian population of hedgehogs from Vis island, described as E. c. nesiotes (DuLic and TvrrKkovıc 1979), was compared to the Greek insular populations from the is- lands of Crete, Rhodos and Lesvos. The Mahalanobis distances calculated confirm that the Yugoslavian population is closer to the Cretan form than to the other Greek insular populations, but remains significantly different from it (p< 0.001). The specimens from the Ionian islands, Kerkyra and Kephalonia, were pooled in one population and compared to Yugoslavian insular population from Vis, because these is- lands are geographically closer to Vis than any other Greek island. We found a significant difference between them (p < 0.001). The mainland Yugoslavian populations, representing E. c. bolkayi (DuLıc and TVRTKO- vıc 1979) and E. c. drozdovskii (MARTINO 1933; PETROV 1940), due to the small size of samples cannot be analysed with confidence. The Mahalanobis distances between cen- troids indicated that: a) Yugoslavian bolkayi is closer to Bulgarian drozdovskü than to its Bulgarian counterpart. b) Yugoslavian drozdovskii is closer to its Bulgarian counterpart than to any other population. c) Yugoslavian bolkayi is very close to the Greek mainland populations from continental Greece and Peloponnisos d) Yugoslavian drozdovskii and Yugoslavian bolkayi are very closely related. Finally, the specimens from Asia Minor, described as E. c. transcaucasicus by WETT- STEIN (1966), were compared to all other populations. With only two measurements avail- able (CBL and ZB), they were found to be different from all populations, except those from Lesvos (MANOVA, p = 0.20). Multivariate analysis of morphometric characters in Frinaceus concolor 1377 Indices The basic statistics for Mi and Li in all populations are given in table 4. Length index is generally constant among the studied populations of Erinaceus concolor, except those from the islands of Rhodos and Samos which differ significantly from the others (ANO- VA, p< 0.001). Maxillary index varies amongst the five main Greek populations and the insular Yugo- slavian population (no. 1-5 and no. 17 in Tab. 1) while the ANOVA (Tukey’s test), groups together the populations from continental Greece, Peloponnisos, Crete and Vis island, on one hand, and the populations from Rhodos and Lesvos, on the other. These results are in accordance with our karyological data which differentiate the specimens of the Greek mainland and Crete from those of the islands of Rhodos and Lesvos (GlAGIA and On- DRIAS 1980). Table 4. Basic statistics (means and standard deviations = sd) of the two indices in the 18 populations studied. Population Length index Maxil. index Discussion The results of the above morphometric analysis of the Greek populations support the pre- viously expressed view that one large-sized subspecies of eastern hedgehog is distributed throughout mainland Greece (GlAGIA and OnDkrIAs 1995). The analysis of Bulgarian hedgehogs, based on data given by Markov (1957), showed that two distinct subspecies occur throughout the country; the small-sized E. c. roumani- cus and one large-sized subspecies including bolkayi and drozdovskii, which were pre- viously described as being different. The comparison of the neighbouring populations from Yugoslavia, Bulgaria and Greece showed that previously there was confusion concerning the name of the large- sized hedgehog described by Marrıno (1930, 1933) and reported by Horz (1978) and KraArocHviL (1980) as occurring in the southern Balkan peninsula (MALEC and STORCH 1963; PETROV 1940; WETTSTEIN 1942). 138 EvA B. GIAGIA- ATHANASOPOULOU and G. MARKAKIS Although the populations from the mainland of Greece and Bulgaria are quite clearly different, they do overlap to some extent and are characterized by morphometric continu- ity. It is suggested that one large-sized subspecies E. c. bolkayi is distributed throughout the southern Balkan peninsula. The CA of the Greek insular populations showed that two small-sized subspecies are distributed throughout Greek islands. E. c. nesiotes (Bate, 1905) from Crete appears to oc- cur also on other islands of the Aegean sea such as Syros, Tinos, Naxos in the Kyclades, as well as on the Ionıan sea islands, Kerkyra and Kephalonia. According to DuLic and Tvrrkovic (1979) the distribution of E. c. nesiotes extends up the Adriatic sea to the is- land of Vis. The analysis showed that the Yugoslavian insular population is rather differ- ent and must be reexamined. The subspecies E. c. rhodius (Festa, 1929) from the island of Rhodos is also distributed on Samos and probably Chios, as the karyological data confirm (GıaGIA and OnDrIAs 1980). The hedgehogs from the island of Kos although karyologi- cally identical to those from Rhodos (MAnDAHL 1978) hold an uncertain position. DE BEAux (1929) described some hedgehogs from the island of Kos as belonging to E. c. transcaucasicus which is found in Asia Minor. Finally, the position of hedgehogs from Lesvos island in CA and their similarity to the specimens from Asıa Minor, support the view that E. c. transcaucasicus is probably dis- tributed on Lesvos, although the set of measurements is not sufficient to draw definite conclusions. This may be supported by the fact that many species distributed in Asia Min- or are also found amongst Lesvos fauna (GIAGIA et al. 1982; OnDrIAS 1966). An extensive study of hedgehogs from Asıa Minor as well as the study of more specimens from the East Aegean sea islands might clarify the relationships among the insular populations (E. c. rhodius) and those from Asia Minor (E. c. transcaucasicus). Ackowledgements We are grateful to Dr. STELLA FRANGUEDAKIS-TSOLIS for her advice and help in taking the cranial mea- surements and Prof. Dr. JoHNn ONDRIAS for his helpful remarks and discussion. Zusammenfassung Multivariate Analyse morphometrischer Merkmale beim Ostigel Erinaceus concolor in Griechenland und benachbarten Gebieten Von der südlichen Balkanhalbinsel sind eine Reihe von Unterarten des Ostigels Erinaceus concolor be- schrieben worden. Trotz umfangreicher morphologischer und karyologischer Beschreibungen ist ihr ta- xonomischer Status immer noch unklar. In der vorliegenden Studie wird versucht, den Status der verschiedenen griechischen Festland- und Inselpopulationen zu identifizieren und ihre Beziehungen zu Subspezies in Nachbarregionen zu klären. Dazu wurde eine multivariate Analyse von 21 somatischen und kranialen Merkmalen an 191 griechischen Exemplaren sowie anhand veröffentlichter Daten (76 aus Bulgarien, 21 aus Jugoslawien und 9 von Kleinasien) durchgeführt. Die Analyse zeigt, daß auf dem gesamten Festland der südlichen Balkanhalbinsel die großwüchsige Subspezies E. c. bolkayi verbreitet ist. Zwei Insel-Unterarten mit geringerer Körpergröße kommen auf griechischen Inseln vor: E. c. nesiotes auf Kreta, den Kykladen und den lonischen Inseln, E. c. rhodius auf den ostägäischen Inseln Rhodos, Samos und Chios. Die Subspezies E. c. transaucasicus kommt auf den Inseln Lesbos und Kos vor. References CORBET, G. B. (1988): The family Erinaceidae: a synthesis of its taxonomy, phylogeny, ecology and zoo- geography. Mammal Rev. 18, 117-172. Multivariate analysis of morphometric characters in Erinaceus concolor 139 De BEAux, O. (1929): Richersche faunistice nelle isole Italiane dell Egeo, Mammiferi. Arch. Zool. Ital. Napoli 13, 5-24. Duuic, B.; Murıc, D. (1967): Catalogus Fannae Jugoslaviae. IV/4 Mammalia. Ljubljana: Cons. Acad. SCHHR.SKEN. Duuic, B.; Tvrrkovic, N. (1979): On some mammals from the central Adriatic and south Adriatic is- lands. Acta Biol. (Zagreb) 43, 15-35. GiIAGIA, E. B.; ONDRIAS, J. C. (1980): Karyological analysis of eastern hedgehog Erinaceus concolor in Greece. Mammalia 44, 59-72. GIAGIA, E. B.; Savıc, 1.; SOLDATOVIc, B. (1982): Chromosomal analysis of the mole rat Microspalax from Greece and Turkey. Z. Säugetierkunde 47, 231-236. GIAGIA-ATHANASOPOULOU, E. B.; ONDRIAS, I. C. (1995): The Distribution of the Eastern Hedgehog Eri- naceus concolor bolkayi in continental Greece. Säugetierkundl. Mitt. 36, 153-173. Hoız, H. (1978): Studien an europäischen Igeln. Z. zool. Syst. Evolut.-forsch. 16, 148-165. Horz, H.; NIETHAMMER, J. (1990): Erinaceus concolor Martin, 1838 — Weißbrustigel, Ostigel. In: Hand- buch der Säugetiere Europas. Hrsg. von J. NIETHAMMER and F. KrApp. Wiesbaden: Aula Verlag. Vol. 3/1, 50-64. HRABE, V. (1976): Variation in cranial measurements of Erinaceus concolor roumanicus (Insectivora, Mammalia). Zool. Listy 25, 303-314. KRATOCHVIL, J. (1980): Hirnmasse der mitteleuropäischen Arten der Gattung Erinaceus (Insectivora, Mamm.). Folia Zoologica 29, 1-20. MALECc, F.; STORCH, G. (1963): Kleinsäuger (Mammalia) aus Makedonien, Jugoslavien. Senck. Biol. 44, 155-173. MANDAHL, N. (1978): Variation in c-strained chromosome regions in European hedgehog. Hereditas 89, 107-128. MARKoVv, G. (1957): Die Insekten fressenden Säugetiere in Bulgarien. Sofia: Zool. Inst. Bulgar. Acad. Wiss. Sofia. Fauna Bulgariens Ill. MARKOV, G.; DOBRIJANOV, D. (1974): Karyologische Analyse der Weißbrust- oder Ostigel (Erinaceus rou- manicus Barr.-Ham.). in Bulgarien. Zool. Anz. 193, 181-188. MARTINO, V. E. (1930): Notes on the ecology of some mammals from Yugoslavia. Zap. Russ. Naush. Inst. Belgrade 2, 53-65. MARTINO, V. E. (1933): Novi jer iz vardarske bavonine. Prirod Razpr. Ljubljana 2, 56-57. MOoRRISoN, D. F. (1976): Multivariate statistical methods. (2nd ed.). New York: Mc Graw-Hill. OGney, S. I. (1928): Mammals of Eastern Europe and Northern Asia. Vol. 1 Insectivora and Chiroptera. Moscow. (English translation 1962, Jerusalem: Israel Programme of Scientific Translations). ONDRIAS, J. C. (1966). The taxonomy and geographical distribution of the rodents of Greece. Säugetier- kundl. Mitt. 14, 1-136. PETRov, B. M. (1940): Bemerkungen über die Systematik und Ökologie der Säugetiere Süd-Serbiens. d.h. Mazedoniens. Zap. Russ. Naush. Inst. Belgrade 16, 57-64. REYMENT, R. A.; BLACKITH, R. E.; CAMPBELL, N. A. (1984): Multivariate morphometrics. (2nd ed.). Lon- don: Academic Press. RoHLF F. J. (1992): NT-SYS. Numerical Taxonomy System. Exeter: Software. RUPRECHT, A.L. (1972): Correlation structure of skull dimensions in European Hedgehogs. Acta Ther- iol. 17, 419-442. SNEATH, P. H.; SOKAL, R. R. (1973): Numerical Taxonomy. San Franscisco: W. H. Freeman. WETTSTEIN, O. (1942): Die Säugetierwelt der Ägäis, nebst einer Revision des Rassenkreises von Erina- ceus europaeus. Ann. Nat.-Hist. Mus. Wien 52, 245-278. WETTSTEIN, ©. (1966): Bemerkungen über einige Säuger des griechisch-kleinasiatischen Raumens. Sit- zungsberichte österr. Acad. Wiss. mathem. naturwiss. Kl, 175, 357-302. WILKINSON, L. (1987): SYSTAT, The System for Statistics. Evanson, IL: Systat Inc. Authors’ addresses: Dr. EvA B. GIAGIA-ATHANASOPOULOU, Section of Animal Biology, Department of Biology, University of Patras, GR-26001 Patras, Greece and Dr. G. MARKAKIS, Department of Biology, University of Crete, GR-71400 Iraklio, Greece. Z. Säugetierkunde 61 (1996) 140-154 u 2 FÜR © 1996 Gustav Fischer, Jena SÄUGETI RRÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Comparative anatomy of the stomach of the Cetacea. Ontogenetic changes involving gastric proportions — mesenteries — arteries By P. LANGER Institut für Anatomie und Zellbiologie, Justus-Liebig-Universität Giessen, Giessen, Germany Receipt of Ms. 18. 12. 1995 Acceptance of Ms. 15. 01. 1996 Abstract This study discusses three aspects of the comparative anatomy of the stomach in seven species of odon- tocete cetaceans: ontogenetic changes in the proportions of four gastric regions, the lines of fixation of the mesogastria to the stomach and the mode of branching of arteries that supply the organ. An interspecific comparison revealed a considerable postnatal increase in the relative volume of the forestomach, however, in the harbour porpoise (Phocoena phocoena) there is a constant decrease in the size of the forestomach during the time period when the total length of the embryo lies between 7% and 12% of the neonatal length. The forestomach, which is not differentiated in species of the cetacean family Ziphiidae (Mesoplodon densirostris was investigated here), is completely differentiated on the ontogenetically left side of the stomach anlage, a situation that has only been found once - in the Brady- podidae or tree sloths - in terrestrial mammals. The branching mode of the arterial supply of the stomach showed only one synapomorphy between Stenella longirostris and Tursiops aduncus, where the A. gastrica sinistra is a branch of the A. hepatica communis. In all other comparisons only symplesiomorphies could be discerned; between Mesoplodon densirostris and Grampus griseus in as many as three cases. This relatively high number of symplesio- morphies indicates that the gastric arterial system in both species did not change significantly from the basic situation at the beginning of cetacean evolution. Introduction Most of the published studies on the anatomy of the stomach of the Cetacea deal with only one species. The number of comparative investigations is limited (WEBER 1888; PERN- KOPF and LEHNER 1937, SLIJPER 1962). Due to this deficit, the present author attempted to consider material from different cetacean species. This study deals with three main aspects of comparative anatomy of the cetacean stomach. 1. The changes in relative size of the four gastric compartments are dealt with in an interspecific comparison. According to findings in terrestrial mammals (LANGER 1988) it is assumed that changes in volume allow an extrapolation to the functional importance of the gastric chambers. 2. The lines of fixation of the Mesogastrium dorsale and the Mesogastrium ventrale to the stomach are discussed. With the help of these fixation-lines an idea can be obtained about the gastric regions that increased in size to form separate compartments. 3. The arterial supply of the gastric regions and their modes of branching were investi- gated. Interspecific comparisons allow differentiation between plesiomorphic (“basal”, “primitive”, or “ancestral”) and apomorphic (“derived”) branching modes. In a second step differentiation was made between symplesiomorphic and synapomorphic characters of arterial branching. Synapomorphies are characters that differentiated in the same direc- tion in different species (MAyr 1975, 1976, HENNIG 1982). Comparative anatomy of the stomach of the Cetacea 141 Material and methods Specimens of cetacean species from the Indian Ocean were made available through the Centre for Dol- phin Studies, Port Elizabeth Museum (PEM), South Africa (taxonomy according to Ross 1984 and Evans 1987). Species marked with an asterisk (*) are depicted in figure 1: MESOPLODON DENSIROSTRIS (PEM 1451) (®) ; SEcm baum N ps.» EXIT) SER END: Superfam.: ZIPHOIDEA Fam.: ZIPHIIDAE *) Mesoplodon densirostris (Blainville, 1817) — Blainville’s beaked whale, PEM 1022, PEM 1451, Superfam.: DELPHINOIDEA Fam.: DELPHINIDAE Subfam.: Globicephalinae *) Grampus griseus (Cuvier, 1812) — Risso’s dol- phin, PEM 1658, PEM 1858, Subfam.: Delphininae Delphinus delphis Linneus, 1758 —-— Common dol- phin, PEM 1848, PEM 1850, PEM 1855, PEM 1856, *) Tursiops aduncus (Ehrenberg, 1832) — Indian Ocean bottlenosed dolphin, PEM 1178, PEM 1849, PEM 1854, PEM 1857, PEM 1959, Stenella longirostris (Gray, 1828) — Spinner dol- phin, PEM 1030, PEM 1319, Subfam.: Steninae Sousa plumbea (Cuvier, 1829) — Humpback dol- phin, PEM 963, PEM 1586. The above-mentioned material originated either from specimens stranded on the Cape and Natal Coasts of the Republic of South Africa, or from individuals that were accidentally caught and drowned in shark-nets off swimming beaches of the coast of Natal. The material was either col- lected directly by the Port Elizabeth Museum or by the Natal Sharks Board, Umhlanga Rocks, Na- Fig. 1. Stomachs of three odontocete species. In Mesoplodon densirostris (Ziphiidae) a forestomach is not differentiated and in the two delphinid spe- cies Grampus griseus and Tursiops aduncus it is of different relative size and position. In Grampus gri- seus the connecting stomach cannot be distin- guished from outside. In all three species the Ductus hepatopancreaticus opens into the proximal duodenum, which in Gram- pus griseus and Tursiops aduncus is dilated to an ampulla duodeni. Abbreviations: O = oesophagus, FS = forestomach, MS = main stomach, CS = connecting stomach, PS = pyloric stomach, D = duodenum. 142 P. LANGER tal, and then transferred to the collections of the Port Elizabeth Museum. The above-mentioned ceta- cean specimens where stored in toto in formalin solution, all of them with opened abdominal and thor- acic cavities. The conservation of the in situ position of the organs thus remained undisturbed. The studies were started by fenestrating the body wall either from the left or the right side to docu- ment photographically the position of the internal organs. After stepwise removal of the ribs, liver, and diaphragm, the digestive tract was totally removed and immersed in 5% formalin solution. The material was sealed in strong plastic bags and sent via Air Cargo to Giessen, Germany, where it arrived in good condition and was stored in cooled containers filled with 5% formalin solution. After rinsing the speci- mens and their arterial vascularisation were studied macroscopically. The material from the Port Elizabeth Museum (PEM) consisted not only of specimens belonging to different species, but also to different developmental stages. The changes of relative volumes of the four gastric compartments, comparing all available material, were determined. The “baseline” of reference was the total length of the neonatal animal (Ln) as well as the total volume of the stomach (foresto- mach [FS] + main stomach [MS] + connecting stomach [CS] + pyloric stomach [PS]). The present author is fully aware that results of such interspecific comparisons can be biased because of consider- able differences in gastric proportions, even in very young animals. The following procedure to deter- mine volumes was applied: After different gastric portions (forestomach, main, connecting, and pyloric stomachs) were opened, cleaned of the contents and after the Tunica mucosa was removed (!) they were tightly filled (but not over-filled!) with moist cellulose pulp, which was pliable into any possible shape. The volumes of pulp were determined by packing them tightly (but not compressed!) into volu- metric cylinders. The volumes of the forestomach (FS), main stomach (MS), connecting stomach (CS), and pyloric stomach (PS) were expressed as percentage of total gastric volume. The total body length (Lt) of each specimen had already been determined when the animal was sal- vaged by the museum crew. From Ross (1984) the length of a newborn (Ln) of the species was known and the relative body length of each specimen could be expressed as a relative value (100 Lt/Ln). To obtain an idea of the length of the weaning period, body length at weaning (Lw) and at birth (Ln) (raw data from TomıLın 1967 and Ross 1984) were compiled and body length at weaning relative to birth length (100 Lw/Ln) could be calculated; its value ranged from 113% in Grampus griseus to 194% in Stenella longirostris. Additionally, serial sections (thickness 10 um) of harbour porpoises embryos (Phocoena phocoena, collection M. KLımA [MK]) were supplied by Professor Dr. HELMUT A. OELSCHLÄGER, Frankfurt/M. The total length of the five specimens were as follows: MK 70: 42 mm, MK 67: 5l mm, MK 64: 60 mm, MK 62: 70 mm, MK 61: 95,5 mm. From these sections models of the external shape of the stomach were reconstructed from plates of “Styropor” (average thickness 2 mm). Secondly, the perimeters of the in- ternal gastric Jlumina were traced with a “Videoplan” (Kontron Elektronik, München) to reconstruct three-dimensional shapes from serial sections. From the data of the luminal perimeters the “topological skeleton” of the six stomachs, i.e., the connecting lines of the “centres of gravity” of the respective cross-sections were processed with “3D-Top”, a topological object processing program developed by T. SCHwEBEL and Dr. A. KRIETE, Giessen, and distributed by IMA-TEC, Bad Tölz, Germany. The length of the forestomach (Lf) as well as the length of the main stomach (Lm) from the end of the oesophagus to its tip were compared and it was possible to determine the relative length of the forestomach (Lf) as compared to the combined lengths of the forestomach plus main stomach (Lf + Lm). Results and discussion Ontogenetic development of stomach compartments in Cetacea Relative volumes of the four gastric compartments in five odontocete species were plotted against relative body length and the following relationships became visible (Fig. 2): After birth the forestomach volume increases strikingly and the pyloric stomach volume decreases. The volume of the main stomach decreases but later shows a tendency to increase. These changes take place during the weaning period. The bar on top of the diagram in figure 2 represents the range of the weaning period. Increase in the relative volume of the main stomach is advantageous during the later part of the weaning period Comparative anatomy of the stomach of the Cetacea 143 70 BERNER ENG ET TE ET | weaning period S.l. relative volumes (%) relative total length (%) a ES SEES ESI—eZPS Fig. 2. Relationships between the relative total length, expressed as percentage of the length of the newborn animal (from Ross 1984), versus the relative volume of gastric compartments, expressed as percent of the total gastric volume. Material from six specimens, belonging to five odontocete species was used. An interspecific comparison is made possible, but results of such a study can be biased be- cause of considerable differences in gastric proportions of different species. Four gastric compartments were considered: FS = forestomach, MS = main stomach, CS = connecting stomach, PS = pyloric stomach. Abbreviations: T. a.= Tursiops aduncus, G.g. = Grampus griseus, S.p. = Sousa plumbea, S.1. = Stenella longirostris, D.d. = Delphinus delphis. because during this time the adult digestive process begins to develop. Autoenzymatic di- gestion (LANGER 1988) of protein, i.e., digestion with the mammal’s own proteolytic en- zymes, becomes increasingly important. Milk, which in cetaceans is extremely rich in fat (FLinpt 1985; OFtEDAL 1993) is later replaced by adult food, consisting of cephalopods, fish and, to some small extent, crustaceans, which are all rich in protein. The forestomach is the region where the massive Tunica muscularis of the gastric wall is able to compress the ingested food into a semi-liquid mass, which passes into the main stomach (Cock- CROFT pers. comm.). The relative length of the forestomach of the harbour porpoise (Phocoena phocoena), expressed as a percentage of the combined lengths of both forestomach plus main sto- mach, increases, but later decreases. This prenatal “overshoot” might document different growing intensities in the compartments of the stomach. In cattle (Bos primigenius) AUERNHEIMER (1909) and BEckER et al. (1951), as well as in red deer (Cervus elaphus) DRESCHER-KADEN (1981) showed that the ruminoreticulum is relatively larger during the first half of the gestation period than during the second half. For the gastric gland of Tri- chechus manatus (West Indian manatee) LANGER (1988) showed a similar “overshoot” in relative size. 144 P. LANGER It is interesting to compare the data of the considered gastric sections in the above- mentioned odontocets with data from the same regions of a mysticete species, the minke whale (Balaenoptera acutorostrata). The three animals investigated by OLsen et al. (1994) had a body length of about 200 to more than 400% of their neonatal length of 2.7 m (NoRMAN et al. 1963) and 2.1 m to 2.74 m according to ToMmıLin (1967). The forestomach volume of the minke whale ranges from 44 to about 70% of the total gastric volume, the main stomach (“fundic chamber” according to OLsen et al. 1994) from about 22 to 44%, and the pyloric stomach from 8.5 to 12%. Only the values for the main stomach lie well above those depicted for Delphinus delphis (PEM 1848) in figure 2. It can be speculated that there is further increase in the relative volume of this gastric compartment during the post-weaning period of all cetacean species. The above-mentioned discussion refers to born cetaceans. In the case of material from the harbour porpoise (Phocoena phocoena) some information on prenatal changes of gas- tric proportions can be gained. The length of the forestomach (Lf) was expressed as a per- centage of the sum of both forestomach and main stomach (Lm) lengths (100 Lf/Lf + Lm) and plotted against the total length (Lt) of the five specimens, expressed as percentage of neonatal length (Ln) of the animals (100 Lt/Ln) (Fig. 3). 60 a [9,7 relative forestomach length (%) a [e]} 45 5 6 7 8 9 10 11 12 relative total length (%) Fig. 3. Relationship between the total length, expressed as percentage of the length of the newborn ani- mal of the habour porpoise, Phocoena phocoena, versus the relative forestomach length, expressed as the sum of the length of the forestomach plus the length of the main stomach. The relative length of the forestomach increases up to a total length (Lt) of 5l mm, which is 6.4% of neonatal length and decreases in longer fetuses. This “overshooting” of the forestomach size could also be observed in some mammals with a voluminous gastric section that is “set off” from the direct connection between oesophagus and duodenum, such as in the red deer (Cervus elaphus, Ruminantia) (DRESCHER-KADEN 1981) and the Caribbean manatee (Trichechus manatus, Sirenia) (LANGER 1988). Comparative anatomy of the stomach of the Cetacea 145 Lines of fixation of the dorsal and ventral mesogastrium Two of the investigated species were used to depict the lines of fixation of the Mesogas- trium dorsale et ventrale to the stomach: Grampus griseus (PEM 1858) (Fig. 4) and Del- phinus delphis (PEM 1856) (Fig.5). All other species showed a principally similar situation. GRAMPUS GRISEUS (PEM 1858) Lines of mesenterial fixation Oesophagus Mesogastrium dorsale Ductus hepatopancreaticus | Mesogastrium ventrale Ampulla duodeni Pe] Se : L) EEE Magscubuun Pylorus Mesogastrium dorsale Lien Forestomach Main stomach Fig. 4. Line of fixation (bold broken line) of the dorsal and ventral mesogastria to the stomach of Grampus griseus. The line of fixation of the dorsal mesogastrium is partly hidden behind the pyloric sto- mach, where it is represented by a thin broken line. The stippled area at the base of the oesophagus is the zone of tight fixation of this region to the diaphragm. DELPHINUS DELPHIS (PEM 1856) Lines of mesenterical fiıxation Duodenum Ampulla duodenı Oesophagus Pylorus Ductus hepato- rien pancreaticus Be & Mesogastrium Forestomach #-— 4 ee, AR 15 ne dorsale 5 cm Mesogastrium ventrale Maın stomach Pylorıc stomach Fig. 5. Line of fixation of the mesogastria to the stomach of Delphinus delphis. For further details see legend to Fig. 4. 146 P. LANGER The Mesogastrıum ventrale is attached at the gastric wall close to the oesophageal opening where the stomach is tightly connected with the diaphragm (Figs. 4 and 5). It runs over the oral part of the main stomach, follows the lesser curvature of the pyloric stomach and the pylorus and ends at the opening of the Ductus hepatopancreaticus into the Ampulla duodeni. The course of the line of fixation of the Mesogastrium dorsale is much more compli- cated because of the formation of a forestomach and the increase in volume in the region of the main stomach. Once again, the mesentery starts at the zone of fixation between the cardiac region and the diaphragm. From here it runs on the topographically left side of the main stomach. The round (Grampus griseus) or slightly triangular (Delphinus delphis) spleen is fastened to the main stomach via the dorsal mesogastrium. This passes over the greater curvature and follows the fold between the main and pyloric stomach regions. With a sharp bend the mesenterial fixation line crosses over to the pyloric stomach, which it follows towards the greater curvature, then passes over the pylorus and the duodenal ampulla to the tube-like part of the duodenum. In all investigated cetaceans the small greater omentum was small; in Delphinus delphis (PEM 1856), for example, this omentum just connects the main stomach with the oral end of the pyloric stomach and does not form an omental pouch. In previous studies (LANGER 1973, 1988) it was shown that the plurilocular stomachs of mammalian herbivores lies on the ontogenetically left side of the Fornix (= Fundus) ven- triculi. It is highly probable that a hypothetical plane through the two lines of fixation of the mesogastria separates the ontogenetically right from the left part of the stomach. The forestomach in Cetacea belongs completely to the ontogenetically left side of the Fornix ventriculi, whereas the main stomach is partly derived from it. In terrestrial mammals it was possible to show that the different forestomach regions are generally partly derived from the ontogenetically left and right sides of the organ (LANGER 1988). Only in tree sloths (Bradypodidae) the position of the total forestomach pouch on the ontogenetically left side of the stomach could be demonstrated. This simi- larity does, of course, not indicate a close relationship between Bradypodidae and Ceta- cea. Arterial branching in the vicinity of the stomach The arterial supply to the stomach and its surrounding organs was dissected in five spe- cies: Mesoplodon densirostris (PEM 1451) (Fig. 6a), Grampus griseus (PEM 1858) (Fig. 6b), Delphinus delphis (PEM 1856) (Fig.6c), Tursiops aduncus (PEM 1178) (Fig. 6d), and Stenella longirostris (PEM 1319) (Fig. 6e). The arterial branching was not only documented in drawings, but also in schematic diagrams that generally list the Aorta abdominalis plus ten branches Tab. la, b. A compilation of the information obtained by the schematic graphs of the branching mode gave the following information (Tab. 2): The Aa. hepatica communis, lienalis, gastrica sinistra et dextra, gastroepiploica sinsistra et dextra, and duodenalis can be branches of the Truncus coeliacus. On the other hand, the A. hepatica communis can give off the following branches: Aa. hepatica propria, gastrica dextra et sinistra, gastro- epiploica dextra, duodenalis, mesenterica cranialis, and lienalis. Four of these above- mentioned vessels, the Aa. gastrica sinistra et dextra, lienalis and gastroepiploica sinis- tra represent either branches of the Truncus coeliacus, the A. hepatica communis, or the A. lienalis. In Mesoplodon densirostris, Grampus griseus, and Delphinus delphis, the Aa. lienalis and gastrica sinistra are branches of the Truncus coeliacus. This is the usual situation in unilocular stomachs such as in man (LippErT and Past 1985), horse and dog (Koch and BerG 1985). Only in Delphinus delphis the A. gastrica dextra branches from the Truncus coeliacus, in the four other species the right gastric artery Comparative anatomy of the stomach of the Cetacea 147 a) MESOPLODON DENSIROSTRIS (PEM 1451) A.lıenalis Tr. coeliacus A.mesenterica ee: IN: gastroepiploica sin. A.gastroepiploica dext-; A.hepatica communis A.gastrica dextra A.gastrıca sinistra A. hepatica propria b) GRAMPUS GRISEUS (PEM 1858) Tr. coelıacus ii A.hepatica comm. 2 I AS A.gastrica sın.g A.mesenterica cranıalıs A. gastrıca dex A.hepatica pro. A.lienalıs Fig. 6a, b c) DELPHINUS DELFHIS (PEM 1856) Truncus coliacus A. gastroepıiploica sınıst A. hepatica communis A. lienalıs Aa. hepaticae propriae A. gastrıca sınıstra A. gastroepiploica dextra A.gastrica dextra A. duodenalıs Ductus hepatopancreaticus Fig. 6c is a branch of the A. hepatica communis. This latter situation can also be found in land mammals, such as man (Lippert and PABst 1985), horse, pig, and dog (all with a unilocular stomach), as well as in three ruminant species, cattle, sheep, and goat (KocH and BErG 1985). In the three species of the subfamily Delphininae the A. gastroepi- ploica sinistra is a branch of the A. lienalis, just as in the unilocular stomachs of man, horse, pig, and dog. 148 P. LANGER d) TURSIOPS ADUNCUS (PEM 1178) Tr. coeliacus A. mesenterica cranialis A. lienalis A.duodenalis A. gastroepiploica sinistra AR A.hepatica communi A.gastrica sinistra P ommunis A. gastrica dextra (a hepatica propria e) STENELLA LONGIROSTRIS (PEM 1319) A.hepatica communis A.lienalis Tr. coeliacus A.mesenterica cranialis f A A.gastroepiploica sinistra A. duodenalıs ne A.gastrica sinistra A. gastroepiploica A.gastrica dextra dextra 3 am Züe A.hepatica propria Fig. 6d, e Fig. 6. Arterial supply of the stomach of a) Mesoplodon densirostris (stomach seen from the right), b) Grampus griseus (stomach seen from dorsally and from the left), c) Delphinus delphis (stomach seen from dorsally and from the right, stippled area at base of oesophagus indicates the zone that is tightly fixed to the diphragm), d) Tursiops aduncus (stomach seen from left), and e) Stenella longirostris (sto- mach seen from the right). In all five cases the oesophagus is marked with an arrow. A branching mode that can be found in the majority of the five investigated cetacean species, as well as in man, horse, and dog, is a plesiomorphic differentiation. The follow- ing four plesiomorphic modes can be documented: 1. A. lienalis is a branch of the Truncus coeliacus, 2. A. gastrica sinistra is a branch of the Truncus coeliacus, 3. A. gastrica dextra is a branch of the A. hepatica communis, 4. A. gastroepiploica sinistra ıs a branch of the A. lienalıs. The following four apomorphic branching modes can be listed: I. A. lienalis is a branch of the A. hepatica communis, II. A. gastrica sinstra is a branch of the A. hepatica communis, III. A. gastrica dextra is a branch of the Truncus coeliacus, IV. A. gastroepiploica sinstra is a branch of the Truncus coeliacus. With the exception of Grampus griseus (PEM 1858) where technical reasons only al- lowed the comparison of three branching modes, it can be seen that the four remaining species each show at least one apomorphy: Mesoplodon densirostris: apomorphy No. IV, Delphinus delphis: apomorphy No.Ill, Tursiops aduncus: apomorphy No. III, Stenella longirostris: apomorphies Nos. I and Il. Comparative anatomy of the stomach of the Cetacea 149 Table 1. Schematic illustrations for five cetacean species of the branching mode of arteries supplying the stomach. Arterial names with asterisks are either classified as plesiomorphic or apomorphic charac- Table 1a Superfam.: ZIPHIOIDEA Fam.: ZIPHIIDAE Mesoplodon densirostris (PEM 1451) Aorta abdominalis Truncus coeliacus A. gastroepiploica sinistra* A. lienalis*) A. hepatica communis*) A. hepatica propria A. gastrica dextra A. gastrica sinistra*) Ds gastroepiploica dextra A. mesenterica cranialis*) OOo PrPumNmDH N Superfam.: DELPHINOIDEA Fam.: DELPHINIDAE Subfam.: GLOBICEPHALINAE Grampus griseus (PEM 1858) Aorta abdominalis Truncus coeliacus A. lienalis*) A. hepatica communis*) A. hepatica propria A. gastrica dextra A. gastrica sinistra*) oO OO oOoNOaANWMPWND m N A. mesenterica cranialis*) In total, Mesoplodon densirostris and Grampus griseus have three symplesio- morphic branching modes, which is the highest number of symplesiomorphies seen in this study (Tab. 3). The lowest number of symplesiomorphies, namely one, can be found when comparing Ste- nella longirostris with Mesoplodon densi- rostris, Grampus griseus, and Delphinus delphis. It should be kept in mind that symplesiomorphies do not indicate that bearers of these characters are closely re- lated to one another, thus forming a ters. Table 1b Superfam.: DELPHINOIDEA Fam.: DELPHINIDAE Subfam.: DELPHININAE Delphinus delphis (PEM 1856) jaREN oO OooSıoataPpUuwmrNmD HH Bau abdominalis Truncus coeliacus r A. gastroepiploica sinistra*) A. lienalis*) A.hepatica communis*) L Aa. hepaticae propriae A. gastrica dextra A. gastrica sinistra*) A. duodenalis A. gastroepiploica dextra Superfam.: DELPHINOIDEA Fam.: DELPHINIDAE Subfam.: DELPHININAE Tursiops aduncus (PEM 1178) N 1 2 3 A 5 6 7 8 9 0 Rn abdominalis Truncus coeliacus rA. gastroepiploica sinistra*) A.lienalis*) A.hepatica communis*) A. hepatica propria EA gastrica dextra A. gastrica sinistra*) A. duodenalis A. mesenterica cranialis*) Superfam.: DELPHINOIDEA Fam.: DELPHINIDAE Subfam.: DELPHININAE Stenella longirostris (PEM 1319) jeN DD m oO OD SONO WR W wa abdominalıs Truncus coeliacus A. gastroepiploica [ sinistra*) rA. lienalis*) A. hepatica communis*) A. hepatica propria En gastrica dextra A. gastrica sinistra*) A. duodenalis A. gastroepiploica dextra A. mesenterica cranialis*) 150 P. LANGER Table 2. Symplesiomorphies (P), i. e., Joint possession of ancestral characters, and synapomorphies (A), i. e., Joint possession of derived characters (MAyr 1976), in the branching mode of arterial vessels are indicated in this compilation by multiple x’s in a horizontal line. Five cetacean species are considered. Md: Mesoplodon densirostris,; Gg: Grampus griseus; Dd: Delphinus delphis; Ta: Tursiops aduncus; Sl: Stenella longirostris. Branches of the Aorta abdominalis and origin of arteries supplying the stomach in five species of the Cetacea Md Ge Dd Ta Sl Aorta abdominalis Truncus coeliacus A. hepatica communis A. hepatica propria A. gastrica dextra (P) A. gastrica sinistra (A) A. gastroepiploica dextra x A. duodenalis x A. mesenterica cranialis A. lienalis (A) x A. lienalis (P) x x L A. gastroepiploica sin. (P) A. gastrica sinistra (P) x x A. gastrica dextra (A) A. gastroepiploica sinistra (A) x A. gastroepiploica dextra x A. duodenalis x x A. mesenterica cranialis KESERE DES 8 28 x I 28 28 28 28 RR EEE X x x Table 3. The number of synapomorphies are depicted in the upper part of this table, symplesiomor- phies can be found at the bottom. Number of symplesiomorphies and synapomorphies in the arterial branching of gastric vessels in ceta- ceans Synapomorphies Mesoplodon Grampus Delphinus Tursiops Stenella Mesoplodon Grampus Delphinus Tursiops Stenella Symplesiomorphies monophyletic group. However, when two species show a common synapomorphic charac- ter, they are of monophyletic origin (HENNIG 1982). Such a synapomorphy is found in Ste- nella longirostris and Tursiops aduncus, where the A. gastrica sinistra is a branch of the A. hepatica communis. Comparative anatomy of the stomach of the Cetacea 151 Table 4. Differentiations of the four gastric regions in 24 cetacean species are compiled from the litera- ture; FS = forestomach, MS = main stomach, CS = connecting stomach, PS = pyloric stomach. Informations from the following publications have been considered: 1. AmasakI et al. (1989), 2. BERZIN (1972), 3. BOULART and PILLiET (1884), 4. CATTANEO (1898), 5. CAvE (1982), 6. GAsKIn (1978), 7. GREEN (1972), 8. HARRISON et al. (1970), 9. Home (1807), 10. JunGKLAUSs (1898), 11. KaryA et al. (1982), 12. OHE (1951), 13. PERNKOPF and LEHNER (1937), 14. Pıng (1926), 15. Rıce and WoLmANn (1990), 16. SLIJPER (1962), 17. SmitH (1972), 18. TARPLEY et al. (1987), 19. TURNER (1867), 20. TURNER (1886), 21. TURNER (1889), 22. VROLIK (1848), 23. WEBER (1886), 24. WEBER (1888), 25. YAMASAKI and KAMIYA (1981), 26. YAMASARI and TAKAHASHI (1971). Gastric regions in 24 cetacean species Genera and species FS MS CS PS References 6, 25, 26 125 25 25, 26 own own 8 374, 9=13, 217o0wn 7,13, 16, own 13, 24 own 13219920824 5,6, 10, 17, 24, own 14 15 2 2UFZE own 6 29223 1 6, 16 128) 18 Platanistidae Pontoporia blainvillei Inia geoffrensis Lipotes vexillifer Platanista gangetica Delphinidae Sousa plumbea Stenella longirostris Stenella roseiventris Delphinus delphis Tursiops truncatus Lagenorthynchus albirostris Grampus griseus Globicephala melaena Phocoenidae Phocoena phocoena Neophocaena phocaenoides Physeteridae Kogia breviceps Physeter catodon Ziphiidae Mesoplodon bidens Mesoplodon densirostris Ziphius cavirostris Hyperoodon spec. Balaenopteridae Balaenoptera acutorostrata Balaenoptera physalus Balaenoptera spec. Balaenidae Balaena mysticetus Ba a a a Ba Ba Ba Be a a ee BE ee ae Eee ee TS HH Hr H HH HH HH HH HH HH Hr ++ + SE ++ rn HH HH HHr HH HH HH HH HH HH Hr + HE + + a en re ee Very probably, the interspecific differences of the branching types of arteries that sup- ply the stomach and organs in its vicinity, represent a complex of morphological differen- tiations. In spite of the remarkable diversity of branching modes, it is possible to differentiate between “primitive” or plesiomorphic and “derived” or apomorphic charac- ters. According to modern studies on ribosomal DNA sequenc and some morphological characteristics (MILINKOWITCH et al. 1993; MEYER 1993; Kııma and OELSCHLÄGER 1994) the Ziphiidae in general are widely removed from the odontoces, and the genus Mesoplo- don is perhaps the most primitive modern ziphiid (DuFFIELD KuLu 1972). On the other hand, the branching of the A. gastroepiploica sinistra from the Truncus coeliacus has to be categorized as an apomorphy (No. IV). The genus Grampus belongs to the delphinid . subfamily Globicephalinae (Evans 1987). DUFFIELD Kuru (1972) concluded that Gram- ‚ pus belongs to the more primitive of the higher odontocetes. A differentiation within the cetacean family Delphinidae becomes possible: The arter- _ial supply of the stomach of Grampus griseus (Subfamily Globicephalinae) is more similar to that of Mesoplodon than to the three representatives of the subfamily Delphininae (Delphinus delphis, Tursiops aduncus, Stenella longirostris). 113% P. LANGER According to the literature all four gastric regions, the forestomach, the main sto- mach, the connecting stomach and the pyloric stomach can be identified in the Delphini- dae, the Phocoenidae, the Physeteridae, the Balaenopteridae and the Balaenidae (Tab. 4)., i.e., they can be found in Odontoceti as well as in Mysticeti. The situation in the Platanis- tidae is ambiguous with Pontoporia blainvillei lacking a forestomach (GAskIn 1978; YA- MASAKI and KamIyA 1981; YamAsAKı and TAKAHASHI 1971), but two other species (Inia geoffrensis and Platanista gangetica) having a forestomach. In the Ziphiidae a foresto- mach is not differentiated. A recent illustration of a Cuvier’s beaked whale, Ziphius cavi- rostris, published by PFEIFFER (1993), depicts a forestomach in this species, but it cannot be decided from this drawing whether the slight external groove is related with an inter- nal separation between forestomach and main stomach. The following conclusions can be drawn from the present observations: Although the external form of the stomach is variable, a forestomach is differentiated in most Cetacea. It increases in size after birth. However, there are some groups (Zipiidae and some Plata- nistidae) where a forestomach is not differentiated at all. It is assumed that this is an apo- morphic character because all other Cetacea — Mysticetes included(!) — differentiate a forestomach. This section of the stomach developed on the ontogenetically left side of the stomach anlage, i.e., it lies to the left of the twisted area between the lines of fixation of the two mesenteries (mesogastria) of the stomach. The arterial supply of the stomach of different cetacean species shows symplesiomorphies between species with a forestomach and Mesoplodon densirostris, which does not have this differentiation. Acknowledgements The author is sincerely grateful for a travel grant funded by the Deutsche Forschungsgemeinschaft (La 325/10-1 ). Dr. Vic G. CockckoFTs, curator of mammals and head of the Centre for Dolphin Studies, Port Eliza- beth Museum, Humewood, South Africa, made material from his collection of cetaceans from the In- dian Ocean available. He and Mr. W. M. MAHOLA of the same institution supplied logistic help. Prof. Dr. HELMUT A. OELSCHLÄGER and Prof. Dr. M. Krıma, both of the Zentrum der Morphologie, Universität Frankfurt/M., supplied material of fetuses of harbour porpoises. TIM SCHWEBEL and Dr. A. KRrIETE of the Institut für Anatomie und Zellbiologie developed and made available a topologi- cal object processing computer-program that allowed measurments on the material of harbour por- poises. Frau BıancaA Hess, also from the Institut für Anatomie und Zellbiologie produced the photos of the illustrations and Dr. R. L. SnıpEs from the same institution gave linguistic advice. The help of all these colleagues is greatly appreciated! Zusammenfassung Vergleichend-anatomische Untersuchungen am Magen der Cetacea. Ontogenetische Proportionsverän- derungen - Mesenterien - Arterien. In dieser Studie werden drei vergleichend-anatomische Aspekte des Magens von sieben Odontoceten- Arten behandelt: Die Veränderungen der Proportionen der vier Magenabschnitte im Laufe der Ontoge- nese, die Befestigungslinien der beiden Magenmesenterien an diesem Organ und der Verzweigungsmo- dus der den Magen versorgenden Arterien. Bei einem interspezifischen Vergleich zeigt sich, daß das relative Volumen des Vormagens (bezogen auf das Volumen des Gesamtmagens) nach der Geburt beträchtlich zunimmt, daß aber beim Schweinswal (Phocoena phocoena) die Größe des Vormagens vor der Geburt abnimmt, und zwar in der Periode, in der die Gesamtlänge der Embryos zwischen 7% und 12% der Länge des Neugeborenen beträgt. Der Vorma- gen, der bei Arten der Familie Ziphiidae nicht ausgebildet ist (Mesoplodon densirostris wurde hier unter- sucht) entsteht in seiner ganzen Ausdehnung auf der ontogenetisch linken Seite der Magenanlage. Diese Situation wurde bei Landsäugetieren bislang nur ein Mal, und zwar bei den Bradypodidae, gefunden. Comparative anatomy of the stomach of the Cetacea 153 Nur bei Stenella longirostris und Tursiops aduncus ist die A. gastrica sinistra ein Ast der A. hepatica communis, ein Verzweigungsmodus der den Magen versorgenden Arterien, der als synapomorpher Charakter zu bezeichnen ist. Bei allen anderen Vergleichen konnten nur Symplesiomorphien nachge- wiesen werden; zwischen Mesoplodon densirostris und Grampus griseus war dies sogar dreifach mög- lich. Diese relativ hohe Zahl von Symplesiomorphien läßt vermuten, daß sich das arterielle Versorgunssystem des Magens bei beiden Arten nicht weit von der Ausgangssituation, wie sie sehr wahrscheinlich zu Beginn der Cetaceen-Evolution herrschte, entfernt haben dürfte. References AMASAKI, H.; DAIGo, M.; TAGUchHı, J.; NISHIYAMA, S. (1989): Morphogenesis of the digestive tract in the fetuses of the Southern Minke Whale, Balaenoptera acutorostrata. Anat. Anz. 169, 161-168. AUERNHEIMER, ©. (1909): Größen- und Formveränderung der Baucheingeweide der Wiederkäuer nach der Geburt bis zum erwachsenen Zustand. Diss. med. vet., Zürich. BECKER, R. B.; Dix ARNOLD, P. T.;, MARSHALL, S. P. (1951): Development of the bovine stomach during fetal life. J. Dairy Sci. 34, 329-332. BERZIN, A. A. (1972): The sperm whale. Jerusalem: Israel Program for Scientific Translations. BOULARD, R.; PILLIET, A. (1884): Sur l’estomac du dauphin. J. Anat. Physiol. (Paris) 20, 432-440. CATTANEOo, G. (1898): Ancora sullo stomaco dei delfini. Atti Soc. Ligustica Scienze Natur. Geograf. (Genova) 9, 452-465. CAvE, A. J. E. (1982): The Vaterian ampulla of certain cetaceans. In: Investigations in Cetacea. Vol. 14. Ed. by G. Pırrerı. Pp. 131-147. DRESCHER-KADEN, U. (1981): Vergleichende Untersuchungen zur Nahrungswahl von Gams- und Rehwild unter besonderer Berücksichtigung der mikrobiellen Besiedlung und der Verdauungsvor- gänge im Pansen. Fortschr. Tierphysiol. Tierernähr. Beiheft 12. DUurFIELD Kuruv, D. (1972): Evolution and cytogenetics. In: Mammals of the sea, Biology and Medicine. Vol. 3. Ed. by S. H. Rıpaway. Springfield, III.: C. C. Thomas. Pp. 503-527. Evans, P. H.H. (1987): Classification. In: The natural history of whales and dolphins. Ed. by P. G. H. Evans. London: Christopher Helm. Pp. 32-75. FLinDT, R. (1985): Biologie in Zahlen. Stuttgart, New York: Gustav Fischer Verlag. GAskIn, D. E. (1978): Form and function in the digestive tract and associated organs in Cetacea, with a consideration of metabolic rates and specific energy budgets. Oceanogr. Mar. Biol. Ann. Rev. 16, 313-345. GREEN, R. F. (1972): Observations on the anatomy of some Cetaceans and Pinnipeds. In: Mammals of the sea, Biology and Medicine. Ed. by S. H. Rıpaway. Springfield, III.: C. C. Thomas. HARRISON, R. J.; JOHNSON, F. R.; Young, B. A. (1970): The oesophagus and stomach of dolphins (Tur- siops, Delphinus, Stenella). J. Zool. (London) 160, 377-390. HENNIG, W. (1982): Phylogentische Systematik. Berlin, Hamburg: Verlag Paul Parey. Home, E. (1807): Observations on the structure of the different cavities, which constitute the stomach of the whale, compared with those of ruminating animals, with a view to ascertain the situation of the digestive organ. Phil. Trans. Roy. Soc. London 97, 93-102. JUNGKLAUS, F. (1898): Der Magen der Cetaceen. Jenaische Z. Naturwiss. 32, 1-94. KaAıyA, Z.; YUEMIN, L.; PILLERI, G. (1982): The digestive tract of Inia boliviensis (d’Orbigny, 1834). In: In- vestigations in Cetacea. Vol. 13. Ed. by G. PıLrerı. Pp. 125-137. KLIMA, M.; OELSCHLÄGER, H. A. (1994): Phylogenie und Systematik der Cetacea. In: Handbuch der Säu- getiere Europas. Vol. 6/IA. Ed. by J. NIETHAMMER; F. KrApp; D. ROBINEAU; R. Ducuy; M. KLıma, Wiesbaden: Aula-Verlag. Pp. 31-48. Koch, T.; BERG, R. (1985): Lehrbuch der Veterinär-Anatomie. Bd. 3. 4. Aufl. Jena: VEB Gustav Fischer Verlag. LANGER, P. (1973): Vergleichend-anatomische Untersuchungen am Magen der Artiodactyla (Owen, 1848). Gegenbaurs morph. Jahrb. 119, 514-561 and 633-695. LANGER, P. (1988): The mammalian herbivore stomach. Comparative anatomy, function and evolution. Stuttgart, New York: Gustav Fischer Verlag. Lippert, H.; PABst, R. (1985): Arterial Variations in Man. Classification and frequency. München: J. F. Bergmann Verlag. Mayr, E. (1975): Grundlagen der zoologischen Systematik. Hamburg, Berlin: Verlag Paul Parey. 154 P. LANGER MAayR, E. (1976): Evolution and the diversity of life. Selected essays. Cambridge, Mass., London, Eng- land: Belknap Press of Harvard University Press. MEYER, A. (1993): Molecular approaches to the phylogenetic study of vertebrates. Verh. Dtsch. Zool. Ges. 86, 131-149. MILINKOVITCH, M. C.; ORTI, G.; MEYER, A. (1993): Revised phylogeny of whales suggested by mitochon- drial ribosomal DNA sequences. Nature 361, 346-348. NORMAN, J. R.; FRASER, F. C.; KREFT, G.; SCHUBERT, K. (1963): Riesenfische, Wale und Delphine. Ham- burg, Berlin: Verlag Paul Parey. ÖFTEDAL, O.T. (1993): The adaptation of milk secretion to the constraints of fasting in bears, seals and baleen whales. J. Dairy Sci. 76, 3234-3246. OH, T. (1951): Iconography on the abdominal cavity and viscera of the Balaenoptera, with special re- marks upon the peritoneal coverings. Sci. Rep. Whales Res. Inst. 5, 17-40. OLSEN, M. A.; NoRD@Y, E. S.; BLIX, A. S.; MATHIESEN, S. D. (1994): Functional anatomy of the gastroin- testinal system of Northeastern Atlantic minke whales (Balaenoptera acutorostrata). J. Zool. (Lon- don) 234, 55-74. PERNKOPF, E.; LEHNER, J. (1937): Vorderdarm. Vergleichende Beschreibung des Vorderdarmes bei den einzelnen Klassen der Kranioten. In: Handbuch der vergleichenden Anatomie der Wirbeltiere, Bd. 3. Ed. by L. Bork; E. GöPPERT; E. KArLıus; W. LuBoschH. Berlin, Wien: Urban und Schwarzen- berg. Pp. 349-476. PFEIFFER, C. J. (1993): Neural and muscular control functions of the gut in odontocetes: Morphologie evidence in beaked whales and beluga whales. J. Physiol. (Paris) 87, 349-354. Pıng, C. (1926): On some parts of the visceral anatomy of the porpoise, Neomeris phocoenoides. Anat. Rec. 33, 13-28. Rice, D. W.; WoLMAN, A. A. (1990): The stomach of Kogia breviceps. J. Mammalogy 71, 237-242. Ross, G. J. B. (1984): The smaller cetaceans of the south east coast of southern Africa. Ann. Cape Pro- vincial Museums, Natural History 15, 173-410. SLIJPER, E. J. (1962): Whales. London: Hutchinson. SMITH, G. J. D. (1972): The stomach of the harbor porpoise Phocaena phocaena (L.). Can. J. Zool. 50, 1611-1616. TARPLEY, R. J.; Sıs, R. F.; ALBERT, T. F.; DALTon, L. M; GEORGE, J. C. (1987): Observations on the anato- my of the stomach and duodenum of the Bowhead Whale, Balaena mysticetus. Am. J. Anat. 180, 295-322. TomıLın, A.G. (1967): Cetacea. In: Mammals of the U.S.S.R. and adjacent countries. Ed. by V. G. HEPTNER, translated from Russian by O. Ronen. Jerusalem: Israel Program for Scientific Translations. TURNER, W. (1867): A contribution to the anatomy of the pilot whale (Globiocephalus svineval, Lace- pede). J. Anat. Physiol. 2, 66-79. TURNER, W. (1886): The anatomy of a second specimen of Sowerby’s whale (Mesoplodon bidens) from Shetland. J. Anat. Physiol. 2, 144-188. TURNER, W. (1889): Additional observations on the stomach in the ziphoid and delphinoid whales. J. Anat. Physiol. (London) 23, 466-492. VROLIK, W. (1848): Natuur- en Ontleedkundige Beschouwing van den Hyperoodon. Haarlem: De Erven Loosjes. WEBER, M. (1886): Ueber den Magen und Darmkanal. In: Studien über Säugethiere. Ein Beitrag zur Frage nach dem Ursprung der Cetaceen. Ed. by M. WEBER. Jena: Gustav Fischer Verlag. WEBER, M. (1888): Anatomisches über Cetaceen. Gegenbaurs morph. Jahrb. 13, 618-653. YAMASAKI, F.; KAMIYA, T. (1981): The stomach of the Boutu, /nia geoffrensis: Comparison with those of other platanistids. Sci. Rep. Whales Res. Inst. 33, 69-81. YAMASAKI, F.; TAKAHASHI, K. (1971): Digestive tract of Ganges Dolphin, Platanista gangetica. I. Oeso- phagus and stomach. Okajimas Folia Anatomica Japonica 48, 271-293. Author’s address: Prof. Dr. PETER LANGER, Institut für Anatomie und Zellbiologie, Justus-Liebig-Uni- versität Giessen, Aulweg 123, D-35385 Giessen, Germany. g Z. Säugetierkunde 61 (1996) 155-164 ZEITSCHRIFT ® FÜR © 1996 Gustav Fischer, Jena SÄUGETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Neonate mortality in captive Asian elephants (Elephas maximus) By F. Kurt and Knyne U MAR Institut für Haustierkunde, Christian-Albrechts-Universität Kiel, Kiel, Germany and Myanmar Timber Enterprise, Yangon, Myanmar (Burma) Receipt of Ms. 08. 11. 1995 Acceptance of Ms. 24. 01. 1996 Abstract One third of Asian elephants born in European zoos and circusses are stillborn (16.0%) or killed or re- fused by their mothers (15.7%). Stillbirths and infanticides are rare in extensively kept and wild-living elephants. Infanticide could be related to life history of the mothers: Females which had grown up in the company of an older, motherly female adopted their offsprings without complications. Those having lacked such affection, tended to kill or at least not to adopt their neonates. Stillborn calves show higher neonate weights (124.6 +20.8kg) than surviving calves (92.0 427.6 kg). Positive correlations were found between gestation period and neonate weight as well as between neonate weight and relative weight (body weight/shoulder height) of the mother. As female elephants in modern Zoos and circusses are relatively heavier than those living in Asian camps, they produce calves after longer gestation peri- ods (644.4 + 19.5 days) with larger neonate weights (105.6 + 26.6 kg) than extensively kept females in Asia (598.1 +51.6 days; 74.0 + 21.6 kg). Chances to survive parturition are negatively correlated with lenght of gestation and neonate weight. Introduction The reproduction rate of Asian elephants kept in western Zoos and circusses is very low ‚, as compared to that in extensively kept working elephants in South Asian camps due to ‚ low availability of reproducing bulls, relative short reproduction periods in females and high mortality of neonates (Kurt 1995 b). Between 1902 and 1992, 44 out of 121 Asian elephants born in Europe were stillborn or died before reaching the age of one year. This infant mortality of 36.4% is more than five times higher than the infant mortality (7%) in wild-living populations of Sri Lanka (Kurt 1974) and South India (SUKUMAR 1992) or in extensively kept working elephants in Myanmar (Burma) (ToKE GALE 1974; Kurt 1995 a). High neonate mortality in zoo and circus elephants is due to infanticide by the mothers and stillbirths, respectively. Aggression of mothers towards neonates has been rarely reported in wild-living and extensively kept Asian elephants. On the contrary, nu- merous reports mentioned adoption of neonates and even suckling by females other than the mother (Kock 1994; Naır et al. 1992; ToKE GALE 1974). Not all these females adopt- . Ing neonates had previously experienced the successful raising of its own calf, however, not all captive females show adopting behaviour (e.g. Kock 1994; WEILENMANN and IsSEnN- . BÜGEL 1992). Hence, it is reasonable to assume that presupposition for maternal behav- iour is established under certain conditions already at a neonate or juvenile age. Therefore, the first hypothesis of this study reads: Females having experienced motherly affections as neonates and juveniles more readily accept their own offsprings than females lacking such affection, when they come in a zoo or circus already as young animals. 156 F. Kurt and KHyNnE U MAR Certain females are known to establish “special relationship” independent of rank and ages leading to the well-known phenomenon of allo-mothers or “aunts” (GArAI 1992), and it is known that “aunts” are paramount before, during and after parturition to calm the parturating female and secure the life of the neonate. It can be assumed that such spe- cial relationships are established only after a certain timespan and that females shifted from one Zoo to another are less engaged in special relationships than females which al- ways lived in only one establishment with relevant female partners available. Therefore, the second hypothesis reads: Shifted females parturate less and raise their offsprings less successfully than females which always lived within one and the same group. Premature stillbirths of underdeveloped calves are rarely reported from wild-living and captive Asian elephants, and occur either in connection with capturing operations (Dittrich 1967; Kurt 1992) or twin-births (ToKE GALE 1974), but never in connection with social distress. Hence, the third hypothesis states that stillbirths are not related to processes of socialization but to physical characteristics of the mother. It ıs generally agreed upon that captive elephants are heavier than wild-living ones of the same shoulder height, and that females kept in zoos and circuses are relatively heavier than female working elephants in Asian camps (BEnEDıcT 1936; KurT and NETTA- SINGHE 1968; SUKUMAR et al. 1988; Kurr 1995 a). The fourth hypothesis of this study therefore reads: Neonate mortality is positively correlated with the relative weight of fe- males (body weight/shoulder height). Material and methods Data were collected on 48 neonate elephants in modern zoos and circuses in Europe, the U.S.A. and Ja- pan (Anni 1962; Dittrich 1967, 1977, Knie pers. com.; Kock 1994; Kock pers. com.; LAnG and EGGEN- BERG 1991; Prıns 1992; REUTHER 1969; RIETKERK et al. 1993; TURKINGTON 1991; WEILENMANN and IseEnßBÜGEL 1992) and on 21 calves born in South Asia (DERANIYAGALA 1995; KRISHNE GoWDA 1969, 1971; RATNASOORIYA et al. 1991; SANDERSON 1907; TOKE GALE 1974). Data from the Zoos of Myosore (south India) and Dehiwala (Sri Lanka) have been included in those from Asian camps, since these ele- phants live under similar conditions as traditionally kept working elephants (Tab. 1). Gestation periods are given in days and are considered to last from the last successful mating up to parturition. It is assumed that mating was successful when insertion of the penis to the vagina, evidence of exudation of seminal fluid from the vagina following dismount and resting of the male after dis- mounting were observed (see PooL£ 1989). Neonate weights are given in kg. Shoulder height represents the height at the withers in cm. Fore some pregnant females relative body weight was calculated as the body weight (kg) divided by shoulder height (cm). For these females shoulder height was measured with a device when they were standing on all four legs. The results were rounded off to the nearest 10cm. They were weighed on heavy-duty weight-bridges, and the results rounded off to the nearst 100 kg. In Carl Hagenbeck’s Tierpark (Ham- burg), the Swiss National Circus Knie and the Zoo in Zürich, elephants are regulary weighed and mea- sured. It was therefore possible to obtain accurate values for the time of pregnancy and parturition, if necessary by interpolations. Results Of 121 Asian elephants born in European zoos, twelve (9.9%) had been killed by their mothers shortly after births and seven (5.83%) had to be raised artificially, as their mothers refused them (HAUFELLNER et al. 1993). In several cases, females showing high aggression toward their neonates learned later with assistance of keepers and “aunts” to care for their offsprings (Kock 1994). Of 426 elephants born in camps from Myanmar (Burma) and Kerala only two (0.5%) were refused by their mothers and raised by “aunts” (NAır et al. 1992; ToKE GALE 1974). Neonate mortality in captive Asian elephants (Elephas maximus) 197 Table 1. Gestation periods (days), birth weights (kg) and shoulder height (cm) of captive born Asian elephants. The numbers given for calves born in Europe correspond with numbers in the list given by HAUFELLNER et al. (1993). Successful births are marked with *, stillbirths with SB. Animals born in the zoos of Dehiwala and Mysore are marked with ' and twins with '. Sources of information: AnchHi (1962); DERANIYAGALA (1955); DITTRICH (1967, 1977); KnIE (pers. com.); Kock (1994); Kock (pers. com.); KRISHNE GoWDA (1969, 1971); Lang and EGGENBERG (1991); Prıns (1992); RATNASOORIYA et al. (1991); REUTHER (1969); RIETKERK et al. (1993); SANDERSON (1907); TOKE GALE (1974); WEILENMANN and Isenbügel (1992). Country Number Gest. Birth Shoulder period weight height Modern zoo and circus: N. America F F 19 M M 12 F M M Bi F F F F BE M B F M M M E E E M RB IE M [BE 5 M M RB EB M M M M M E M M je) 158 F. Kurt and KHYNnE U MAR Table 1. (continued) N. America South Asıan camps: SB * * 1 2 3 4 I 6 fl 8 9 a a a a a m mo Sri Lanka x“ *% x 3x ı ZZzmZu Zumımiı un 0 * * SAU RUND» Of 28 reproducing females and their neonates in European Zoos and circuses life his- tories are roughly known (Tab. 2). Females cared for by an older female when young, la- ter accepted their offsprings significiantly more often than females lacking such affection (chi 18.52; P< 0.001). Such a significant difference could not be found between fe- males shifted on to four times between several herds and those never shifted. Females well cared for by another female at young ages parturated more often than other females (Mann-Whitney U Test, P< 0.01). Females which had never been shifted, parturated more often than females which had been shifted one to four times from one to another zoo (Mann-Whitney U Test; P< 0.01). Of 121 Asian elephants born in European zoos and circuses 20 (16.5%) were stillborn (HAUFELLNER et al. 1993), but of 296 captive-bred elephants in Myanmar only 20 (4.0%) stillbirths were recorded (Toke Gale 1974). The difference between the rates of stillbirths of these different management systems is significiant (chi, = 18.94; P < 0.001). Neonate weights vary between 49 and 159 kg (mean value: 100. kg. Gestation periods vary between 510 and 678 days (mean value: 618 days). There are no significant differ- ences between male and female calves. In Asian elephants kept in zoos and circuses the mean gestation period (644 days) is significantly longer and the mean birth weight (106 kg) significantly heavier than in south Asian neonates with a mean gestation period of 598 days and a mean birth weight of 74kg. The mean weight of stillborn calves (125 kg) is significantly higher than the mean weight of surviving neonates (92 kg). Still- Neonate mortality in captive Asian elephants (Elephas maximus) 159 Table 2. Number and survial chances of the offsprings of 28 female Asian elephants in European zoos and the Swiss National Circus Knie. Females are distinguished according to their life histories: 1: Fe- males which while growing up, had close contact to an older female (MT-K'). 2: Females lacking this opportunity (MT-K). 3: Females which always lived in the same establishment (U 0). 4: Females which had been shifted for one to four times between zoos (U 1-4). MT-K* MT-K" UO U1l-4 1. Number of females: 2. parturitions: 3. neonate accepted: 4. stillbirths: 5. neonate killed or not accepted: 6. losses (4. + 5.): 7. number of neonates per female: 6: J: D: DE je birh are produced after a significantly longer mean gestation period of 644 days than sur- viving neonates with a mean gestation period of 616 days (Tab. 3). There is no correlation between shoulder height of neonates and duration of gestation, but neonate weights are positively correlated with the duration of gestation (Tab. 1; Spearman rank correlation goethieient..r,— 0.67; P.- 0.001). Since neonate weights and stillbirths correlate positively with gestation periods (Tab. 3), the chances to survive parturition can be estimated: After a gestation period of some 600 days calves are born with a mean weight of approximately 60 kg and their chance to survive parturition is practically 100%. Calves born after 650 days of gestation show mean birth weights of approximately 110 kg; their chance to survive parturition is still 70%; but calves born after a long gestation period of 680 days have a mean weight of 130 kg and a low survivial chances of approximately 20%. According to the few available data, lenght of gestation periods and neonate weights, respectively, do not seem to correlate with ages or weights of mothers but with the rela- tive body weights of the mother. In a small sample of seven neonate-mother pairs from Carl Hagenbeck’s Tierpark in Hamburg, the swiss National Circus Knie and the Zürich Zoo, neonate weights are known as well as weights and shoulder heights of mothers. For this admittably small sample a significant negative correlation was found between relative weight of mother at the beginning of the second year of pregnancy and neonate weight. In captive Sri Lankan elephants relative weights of females in reproducing age and neo- nate weights are smaller (Tab. 4). The stillbirths of male twins at Port Lymphe Zoo Park, Kent, in 1990 (Nr. 110 and 111; Tab. 1) had neonate weights of 135 kg and 116 kg, respec- tively, and were delivered after gestation periods of 682 days and 718 days, respectively. The neonate weights and the gestation periods of these twins are above mean values of zoo-born calves. At the beginning of pregnancy their mother had a weight of 4000 kg and short before parturition of 4500 kg (TurkınGron 1991). Although she was described to be very tall (shoulder height assumed 280 cm) she had a relative weight of 14.3 to 16.0 kg/ cm. As birth weight correlates positively with the frequency of stillbirths as well as the re- lative weight of mothers it can be concluded that high relative weight of the mother in- creases the probability of being stillborn. 160 F. Kurt and KHynE U MAR Table 3. Weights and gestation periods of Asian elephants under different conditions of management. 1. all records: 2. males: 3. females: 4. stillbirths: 5. norm. births: 6. zoo and circus: 7. South Asia: 8. all records: 9. males: 10. females: 11. stillbirths: 12. norm. births: 13. zoo and circus: 14. South Asia: Extremes Average Mann-Withney i U test max. min. Neonate weight (kg) 150 49 148 49 159 54 P > 1.00 150 93 159 49 P< 0.001 159 50 136 49 P< 0.001 Gestation period (days) 678 510 668 580 678 520 P » 1.00 668 619 678 510 2015 678 614 666 510 P< =0.001 Table 4. Comparison of neonate weight with the relative weight of the mother. Relative weight (body weight in kg/shoulder height in cm) was calculated for the beginning of the second year of pregnancy. Numbers given for calves born in Zoos and circuses correspond with the numbers in the list of HAUFELL- NER et al. (1993). Data on 13 Sri Lankan females at reproducing age stem from KURT and NETTASINGHE (1968) and those on five neonates from RATNASOORIYA et al. (1991). For the seven mother-neonate pairs of European zoos and circuses a significant correlation between relative weight of the mother and neo- nate weight was found (Spearman rank correlation coefficient, r, = 0.94; P< 0.01). Nr. of neonates birth Mother weight weight (kg) (kg) Zoo and circus Mother Mother shoulder rel. weight heigt (cm) A:B 154 3400 230 14.8 145 3500 280 12,5 142 2 300 190 au 139 2 600 230 1.3 125 2 800 260 10.8 100 2500 230 10.9 7] 2.600 240 10.3 126.0 + 28 2814 + 460 236 + 26 11.922315 Sri Lanka 61.6+19 2146 + 350 (m5)) (m) 2ER (n—=15)) a Neonate mortality in captive Asian elephants (Zlephas maximus) 161 Discussion The first hypothesis of this study ıs fully and and the second partly confirmed by the pre- sent results. Obviously, shifted females do not find adequate social contacts in the new en- vironment and reproduce accordingly less. They may or may not adopt their newborns according to their own experience at young ages. Concerning the number of stillbirths, there are no significant differences between the four groups of females dstinguished here. Hence, the third hypothesis is confirmed: Stillbirths are not correlated with the socializa- tion processes, however, they correlate positively with relative weight of the mother. Hence, the fourth hypothesis of this study is confirmed. In wild-living population allo-mothers are inevitably close relatives of parturating fe- males (Kurt 1992), but in captive ones special relationships rely mainly on behavioural char- acteristics of the females concerned (GARAT 1992). In intensively kept elephants in south Asia, where captive propagation is economically unimportant, hence neglected and hardly successful, “aunts” are not consired paramount social partners for parturating females (AZEEZ et al. 1992). But in extensively kept working elephants in south India and Myanmar the phenomenon of “aunts” is well known. Here “aunts” are allowed to follow their highly pregnant female partners to assist during parturition and for the first months or year in rais- ing and protecting the neonate (ToKE GALE 1974; KRISHNAMURTHY 1992). In Zoo and circus elephants “aunts” can play an important part in successful calf-raising (Kock 1994). The present study shows that close special relationships between certain females cor- relate positively with the number of parturitions per female. Shifting single females from one Zoo to another is a rather questionable practice for captive propagation of a highly social species with extreme matrilinear reproduction pattern as shown by the following sample (for details, see HAUFELLNER et al. 1993): Four bulls in the Zoos of Copenhagen, ' Paris, Rotterdam and Zürich mated with a total of eight females of their own group. All of these females gave birth to at least one offspring. Furthermore, the four bulls mated with at least nine females, shifted temporarily to their groups from other zoos, but only two of them gave birth to at least on offspring. The high frequency of infanticide in western Zoos and circuses is correlated with the lack of the sozialization processes. The experiences a certain female makes during neonate and juvenile ages, when looked after by an older one, later have a positive effect on its ‚ quantitative and qualitative reproductive success. In several European Zoos (HAUFELLNER | et al. 1993) as well as in intensively kept populations in Kerala (Azerz et al. 1992) calves ‚ are weaned before reaching one year of age. In extensively kept working elephants such ' early weaning is considered to increase calf mortality (KRISHNAMURTHY 1992). In Tamil ‚ Nadu and Thailand weaning takes place between the first and the second year (KRISHNA- MURTHY 1992; GERBET 1994), but in Myanmar, where a relative large proportion of captive- born females reoproduce, weaning takes place only after the fourth year (KuynE U MAR ‚ 1992; TOKE GALE 1974). These data from different south Asian elephant-keeping establish- ments advocate for the results of the present study that motherly affection received by ele- phants when they are young is positively correlated with later reproductive success. Data presented in this study show that elephants in modern Zoos show longer gesta- ‚ tion periods and higher neonate weights than those born in south Asian camps. Accurate ‚ data are missing for wild-living elephants, but of five neonates born shortly after capture ‚in the Mysore Khedda in 1967 and in one dead neonate found in 1970 in the Kakankote forests, their weights were estimated by the first author to range between 50 and 70 kg. DirrricH (1967) mentioed a 48-kg neonate of a wild south Indian female. Modern studies on population dynamics and ecology of wild-living elephants (e.g. EISENBERG und Lock- HART 1972) assume average gestation periods of 22 months, but south Asian experts indi- ‚ cate shorter gestation periods between 17 and 22 months (overview in TOKE GALE 1974; ‚ RATNASO0RIYA et al. 1991). From captive elephants in Asian camps as well as zoos, mating 162 F. Kurt and KuynE U MAR has been reported during pregnancy (e.g. KRISHNE GowDA 1971; TOKE GALE 1974; RATNA- SOoRIYA et al. 1991), hence actual lenght of the gestation period could have been underes- timated, when the hormonal changes were not monitored (e.g. MAınKkA und LOTHROP 1990; TAyLor 1995). In Sri Lanka six reproducing females underwent continuous ethologi- cal and physiological observation, and the results indicated a shorter mean gestation peri- od (627.5 +9.5 days; RATNASOORIYA et al. 1991) than in zoos and circuses (644.4 + 19.5 days). Neonate weight is positively correlated with gestation period and relative weight of the mother. Parturating females in modern zoos and circuses can reach relative weights of 10.8 to 16.0 kg/cm, while tame Sri Lankan females in reproductive ages show a mean rela- tive weight of 9.6+1.1kg/cm. In Tamil Nadu the mean relative weight of extensively kept females at reproductive ages is 9.2 +#1.2 kg/cm, and it can be assumed that in wild-living ones it is even smaller (SUKUMAR et al. 1988). Stillborn calves are heavier than surviving neonates and parturated after compara- tively long gestation periods. One can only speculate on the gynecological reasons for stillbirths: Placentae are relatively large due to overweight of pregnant females and births take place only after the fetus can no longer be sufficiently supplied with oxygen (NAAKT- GEBOREN and SLIJPER 1970). In modern Zoos and circuses stillbirths and infanticides are re- sponsible for a neonate mortality of 26.5%. A high neonate mortality is also known in extensively kept elephants in Tamul Nadu: Obviously, due to premature weaning and ne- glect 22.5% of calves died in Tamil Nadu within the first year of live (SUKUMAR 1991; KRISHNAMURTHY 1992). In wild-living populations neonate mortality is low: In a two-year study several births were observed at close range; neither stillbirth nor infanticide oc- cured (Kurr 1992). It is reasonable to estimate that within the first five years of live in wild-living elephants mortality rates are five percent per annum in females and nine per- cent per annum in males (Kurr 1974, 1995 a; SUKUMAR 1992). These values correspond with a loss of 35% of all calves born after five years. In modern zoos and circuses about as many young elephants die within a short period before, during and after parturition. The findings of this study suggest the following management practices in Zoos: (1) Young females should grow up in the company of at least one motherly older female. (2) Females should be kept in such numbers that special relationships can be established. (3) Shifting single females for breeding reasons between Zoos destroys established special relationships and reduces the reproductive potential of a female. A transfered female should always be accompanied by another female, with which it has already established a special relationship. (4) As stillbirths correlate positively with overweight of mothers, the causes of overweight should be ruled out by changing the feeding practices and allowing captive elephants more movement. Acknowledgements Data on birth weights and relative weights were provided by KArı Kock of Hagenbeck’s Tierpark, Hamburg, the Swiss National Circus Knie (FrRAnco KNIE, Louis KNIE, and CHRIS KRENGER) and Zoo Zürich (Dr. A. RüßEL and R.E. HoneEGGEr). Prof. G. B. Hartı (Kiel) and L. AEcKERLE (Aichach) made valuable comments on the manuscript. Zusammenfassung Neonatensterblichkeit bei asiatischen Elefanten in Menschenobhut Ein Drittel aller asiatischen Elefanten, die in europäischen Zoos und Zirkussen zur Welt kamen, wur- den tot geboren bzw. von ihrer Mutter kurz nach der Geburt getötet oder zumindest nicht aufgenom- Neonate mortality in captive Asian elephants (Elephas maximus) 163 men. Bei wildlebenden oder in Südostasien extensiv gehaltenen Arbeitselefanten sind Totgeburten und Kindstötungen selten. Kindstötung hängt zusammen mit der Lebensgeschichte der Mutter: Weibchen, die in Gesellschaft eines älteren mütterlichen Weibchens aufwuchsen, nahmen ihre Jungen in der Regel problemlos an. Jene, die nicht entsprechend umsorgt wurden, zeigten die Tendenz zur Kindstötung oder wenigstens zur Verweigerung des Neugeborenen. Totgeborene Kälber haben höhere Gewichte (124,6 +20,8 kg) als überlebende Kälber (92,0 +#27,6kg). Positive Korrelationen wurden gefunden zwischen Tragzeit und Neonatengewicht und zwischen Neonatengewicht und relativem Gewicht (Kör- pergewicht/Schulterhöhe) der Mutter. Da Elefantinnen in modernen Zoos und Zirkussen relativ schwerer sind als in südasiatischen Elefantencamps, bringen sie ihre Kälber nach längerer Tragzeit (644,4 + 19,5 Tage) und mit höheren Geburtsgewichten (105,6 + 26,6 kg) zur Welt, als die extensiv ge- haltenen Elefantinnen in Südasien (598,1 + 51,6 Tage; bzw. 74,0 # 21,6 kg). Die Überlebenschancen bei der Geburt sind negativ korreliert mit der Tragzeitlänge und dem Neonatengewicht. References ANGHI, C. G. v. (1962): Breeding Indian elephants, Flephas maximus, at the Budapest Zoo. Int. Zoo Yb. 4, 83-86. AZEEZ, M. A.; UNNIKRISHNAN, M. D.; RAGHAVAN, I. K.; MENon, T. G. (1992): Insurance coverage to ele- phant. In: The Asian elephant; ecology, biology, diseases and management. Ed. by E. G. SıLas, M. KrısHnA NAIR, and G. NIRMALAN. Trichur: Lumiere Printing Work. Pp. 127-129. BENEDICT, F. G. (1936): The physiology of the elephant. Washington: Carnegie Institution. DERANIYAGALA, P. E. P. (1955): Some extinct elephants, their relatives and the two living species. Colom- bo: Government Press. Dittrich, L. (1967): Beiträge zur Fortpflanzung und Jugendentwicklung des Indischen Elefanten, Ple- phas maximus, in Gefangenschaft mit einer Übersicht über die Elefantengeburten in europäischen Zoos und Zirkussen. Zool. Garten (NF) 34, 56-92. Dittrich, L. (1977): Über die Nachzucht des Asiatischen Elefanten (Elephas maximus) in europäischen Tiergärten. Zool. Garten (NF) 47, 296-302. EISENBERG, J. F.; LOCKHART, M. (1972): An ecological reconnaissance of Wilpattu National Park, Cey- lon. Smithonian Contribution to zoology 101, 1-118. GERBET, S. (1994): L’Elephant de travail en Thailand. Nantes: Ecole nationale veterinaire. GARAI, M. E. (1992): Special relationships between female Asian elephants, Elephas maximus, in zoolo- gical gardens. Ethology 90, 197-205. HAUFELLNER, A.; KURT, F.; SCHILFARTH, J.; SCHWEIGER, G. (1993): Elefanten in Zoo und Zirkus, Doku- mentation Teil 1. München: Karl Wenschow. Khyne U Mar (1992): Development of artificial insemination in Myanmar elephants (Elephas maxi- mus). Forestry Science Research Paper. Yangon: Ministry of Forestry. Kock, K. (1994): Elefanten - mein Leben. Hamburg, Zürich: Rasch und Röhring. KRISHNAMURTHY, V. (1992): Care and management of elephant calves in captivity. In: The Asian ele- phant; ecology, biology, diseases and management. Ed. by E. G. Sı.as, M. KrısHNA NAIR, and G. NIRMALAN. Trichur: Lumiere Printing Work. Pp. 82-85. KRISHNE GoOWDA, C. D. (1969): A brief note on breeding Indian elephants, Flephas maximus, at Mysore Zoo. Int. Zoo Yb. 9, 99. KRISHNE GowDA, C.D. (1971): Breeding Indian elephants, Elephas maximus, at Mysore Zoo. Int. Zoo Yıb. 12, 121122. Kurt, F. (1974): Remarks on the social structure and ecology of the Ceylon elephant in the Yala Na- tional Park. In: The behaviour of ungulates and its relation to management. Ed. by V. GEIST and F. WALTER. Morges: IUCN Publ. new ser. 24, 618-634. Kurr, F. (1992): Das Elefantenbuch. Hamburg, Zürich: Rasch und Röhring. Kurt, F. (1995 a): Asian elephants (Elephas maximus) in captivity and the role of captive propagation for maintenance of the species. In: Proceedings of the Eighth U. K. Elephant Workshop. Ed. by N. G. SPOONER and J. A. WHITEAR. Southhampton: North of England Zool. Soc. Pp. 69-96. Kurt, F. (1995 b): The preservation of Asian elephants in human care - acomparison between different keeping systems in South Asia and Europe. Animal Research and Development 41, 38-60. KURT, F.; NETTASINGHE, A. P. W (1968): Estimation of body weight of the Ceylon elephant (Zlephas maximus). Ceylon vet. J. 16, 24-26. 164 F. Kurt and KHuynE U MAR Lang, E. M.; EGGENBERG, U. (1991): Protokoll einer Elefantengeburt. Zool. Garten (NF) 61, 5-7. MAINKA, S. A.; LOTHROP, C. D. (1990): Reproductive and hormonal chances during the oestrus cycle and pregnancy in Asian elephants (Zlephas maximus). Zoo Biology 9, 411-419. NAAKTGEBOREN, C.; SLIJPER, E. J. (1970): Biologie der Geburt. Hamburg, Berlin: Paul Parey. NAIR, G. P.; RADHAKRISHNAN, K.; CHANDRASEKHARAN, K. (1992): Mating behaviour of the Asian ele- phant in captivity. In: The Asian elephant; ecology, biology, diseases and management. Ed. byE.G. SILAS, M. KRISHNA NAIR, and G. NIRMALAN. Trichur: Lumiere Printing Work. Pp. 78-81. POOLE, J. H. (1989): Mate guarding, reproductive success and female choice in African elephants. Ani- mal Behav. 37, 842-849. Prins, P. J. (1992): Op goed geluk - de geboorte van een zondagskind. Zoo Informatie Emmen 1, 1-32. RATNASOORIYA, W. D.; FERNANDO, S. B. U.; MANATUNGA, A.M.V.R. (1991): Pregnancy duration of Sri Lankan elephant (Elephas maximus maximus) in captivity. Med. Sci. Res. 19, 623-624. REUTHER, R. T. (1969): Growth and diet of young elephants in captivity. Int. Zoo Yb. 9, 168-179. RIETKERK, F. E.; Hıppıng, H.; Van Disk, S. (1993): Hand-rearing an Asian elephant, Elephas maximus, at the Noorder Zoo, Emmen. Int. Zoo Yb. 32, 244-252. SANDERSON, G. P. (1907): Thirteen years among the wild beasts of India. Edingburgh: John Grant. SUKUMAR, R. (1991): The breeding and management of captive Asian elephants. In: Proceedings of the fifth U. K. elephant workshop. Ed. by N. G. SpoonEr. Windsor. Pp. 40-49. SUKUMAR, R. (1992): The Asian elephant - Ecology and management. Cambridge Univ. Press. SUKUMAR, R.; JosHI, N. V.; KRISHNAMURTHY, V. (1988): Growth in Asian elephant. Indian Acad. Sci. (Animal Sci.) 97, 561-571. TAYLOR, V. (1995): The social behaviour and breeding physiology of a group of captive Asian elephants. In: Proceedings of the Eighth U. K. Elephant Workshop. Ed. by N. G. SPOONER and J. A. WHITEAR. Southhampton: North of England Zool. Soc. Pp. 58-68. ToKE GALE (1974): Burmese timber elephants. Rangoon: Trade cooperation. TURKINGTON, J. (1991): Asian elephant pregnancy and birth at Porth Lymphe. In: Proceedings of the fifth U. K. elephant workshop. Ed. by N. G. SPOONER. Windsor. Pp. 37-39. WEILENMANN, P.; ISENBÜGEL, E. (1992): Keeping and breeding the Asian elephant at Zurich Zoo. In: The Asian elephant; ecology, biology, diseases and management. Ed. by E. G. SıLas, M. KRISHNA NAIR, and G. NIRMALAN. Trichur: Lumiere Printing Work. Pp. 78-81. Authors’ addresses: Dr. FRED KURT, Pfarrer Steinacker-Str. 12, D-86551 Aichach, Germany and DAaw KHyNE U Mar, Manager Research, Myanmar Timber Enterprise, Ahlone, P. ©. B., Yangon, Myanmar (Burma) Z. Säugetierkunde 61 (1996) 165-175 ZEITSCHRIFT® © I FÜR © 1996 Gustav Fischer, Jena SÄUGETI ERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) Von K. FISCHER I. Zoologisches Institut, Universität Göttingen, Göttingen, Deutschland Eingang des Ms 17. 07. 1995 Annahme des Ms 07. 02. 1996 Abstract The course of puberty in male fallow deer (Cervus dama) Investigated was the process of puberty from the beginning onwards (onset of coronet growing at about 6 1/2 months of age) up to the end (4th/5th year of life). The end of puberty, that is the time of ontogen- esis when the general growth is completed, and all the primary and secondary sex characters have reached the level of prime age males in amplitude and frequency. The process of puberty is divided into successive steps by annual periodicit, e.g. during the winter months no general growth occurs (“winter- rest-weight”). The different sex characters do not appear at the same time. Maximal sperm production already occurs in the 3rd. maximum of testicular volume and the end of general growth in the 4th year of life. Neck girth, fat-season-syndrom and blood-testosterone on the other side reach their final maxi- ma during the Sth year of life. At the beginning of puberty the seasonal phases of secondary sex charac- ters diverge markedly from their final position and shift to their endpoints in the course of puberty (shifting-phenomenon). These instabilities are a typical criterion of puberty. On the other side, seasonal stability over years concerning onset, maximum, minimum and end of the events are characteristic of adult male fallow deer. Einleitung Untersuchungen an männlichen Hirschartigen (Cervidae) zum Pubertätsbeginn und teil- weise auch zu ihrem Verlauf liegen vor aus der Gruppe der Telemetacarpalia (GROVES und GruBB 1987) für Maultierhirsch (Odocoileus hemionus) (Wooo et al. 1962; Banpy et al. 1970), Ren (Rangifer tarandus) (McEwan 1968; WHITEHEAD and McEwaAn 1973; LEA- DER-WILLIAMS 1979), Elch (Alces alces) (Franzman et al. 1978) und Reh (Capreolus ca- preolus) (SEMPERE und Lacroıx 1982); und aus der Gruppe der Plesiometacarpalia, Unterfamilie Echthirsche (Cervinae) für Rot- (Cervus elaphus) (LincoLn 1971 a; SUTTIE and Kay 1982) und für Damhirsch (Cervus dama) (CHarLin und WHITE 1972; CHAPMAN und CHAPMAN 1980; CHAPMAN et al. 1981; FISCHER et al. 1985). In der vorliegenden Arbeit werden Befunde aus einer Langzeitstudie an männlichem Damwild vorgestellt zu Beginn, Verlauf und vor allem zum Ende der Pubertät, sowie über die Interaktionen zwischen Pubertätsverlauf und der Jahresperiodik. In unserer Herde können für adultes Damwild (>5Jahre) folgende saisonale Ereignisse mit großer zeitlicher Genauigkeit vorausgesagt werden: Abwurf: 20. April- erste Maitage; Fegen: 20. August-erste Septembertage (der 1. Geweihzyklus weicht von diesem Schema ab). Brunst: 15./20. Oktober-10./15. November. Satzzeit: 15. Juni-erste Julitage. 166 K. FISCHER Material und Methode Am 3./10. Januar 1987 wurde die Studie mit 5 ca. 6 1/2 Monate alten Damhirschkälbern begonnen und im Mai 1991, etwa 6 Wochen vor Vollendung des 5. Lebensjahres (LJ), mit noch 3 Individuen beendet. Zwei Tiere wurden im August 1987 durch ein drittes ersetzt. Aus Gründen der Gruppenanpassung gin- gen seine Werte erst ab November 1987 in die Berechnungen mit ein. Im Oktober 1988 starb ein Hirsch aus unbekannten Gründen. Die Tiere lebten in Göttingen (51°32’N, 9° 56’ E) unter Freilandbedingungen in einem etwa 220 m? großen Gatter ohne die Konkurrenz älterer Männchen und ohne Weibchen. Sie wurden optimal gefüt- tert mit Grünfutter während der Vegetationsperiode und Heu während der anderen Jahreszeiten. Zu- sätzlich dazu erhielten sie zu allen Jahreszeiten Wasser und ein käufliches Kraftfutter ad lib. Immobilisation: Im Abstand von etwa 4 Wochen wurden die Hirsche mit 1-3 ml einer Mischung aus 500 mg Rompun® (Bayer AG) gelöst in 5 ml 10% Ketavet” (Parke Davis Corp.) anästhetisiert. Appli- ziert wurde das Narkosemittel intramuskulär über ein Luftdruck-Betäubungsgewehr. Hodenvolumen (HV): Mit einer Schublehre wurden Länge und Breite beider Testikel ohne Epidi- dymis gemessen und das Volumen nach der Formel zur Berechnung eines Rotationsellipsoids be- stimmt: V = an - (ab?) (a = longitudinaler, b = transversaler Radius). Die Volumina beider Hoden wurden addiert. Halsumfang (HU): Der Halsumfang wurde über dem Kehlkopf und etwa 10 cm caudad gemessen. Beide Meßwerte wurden gemittelt. Um einen konstanten Zug bei den Messungen zu erreichen, war am Meßband eine auf 8 Newton eingestellte Federwaage befestigt. Lebendgewicht (LG): Das LG wurde auf 500 g genau ermittelt. Statistik: Die Befunde wurden nach dem Student’s T-Test verglichen. Die Wachstumsverläufe wer- den dargestellt durch die Steigungsgeraden. Die annualen Kurvenverläufe der HV wurden auf Ähnlichkeit über den Pearsonschen Korrelationskoeffizienten (r) mit entsprechenden Signifikanzen ge- prüft. Für die vergleichende Betrachtung korrespondierender Punkte bei gleichsinnigen Kurvenverläu- fen wurden im Rahmen des T-Tests folgende Irrtumswahrscheinlichkeiten festgelegt: p: <0,05: nicht signifikant; p: <0,01 (signifikant); p: < 0,005 (hochsignifikant). Wegen der geringen Individuenzahl wer- den nur hochsignifikante Unterschiede durch Zahlen belegt. Bei signifikanten Differenzen wird dies er- wähnt. Alle Zahlenangaben im Text geben Mittelwerte wieder. Ergebnisse Hodenvolumina (HV) Wie aus Abbildung 1a ersichtlich, setzt eine deutliche HV-Zunahme bei den jungen Dam- hirschen am Anfang des 2. LJ im Juli kurz vor dem Fegen des 1. Geweihs (1.-20. August) ein. Das erste Maximum der HV liegt im September/Oktober bei 58 ml. Die Hirsche sind jetzt 15/16 Monate alt. Noch während und gleich nach der Brunst fällt im November das HV steil ab auf ca. 40 ml. Danach wird eine deutlich verlangsamte Abnahme bis zum Minimum von 13 ml im Mai/Juni kurz vor dem Ende des 2.LJ sichtbar. Beginn und Dauer des Minimums fallen zusammen mit dem Abwurf des 1. Geweihs zwischen Mitte Mai und den ersten 10 Tagen im Juni und der Wachstumsperiode des zweiten am Beginn des 3. LJ. Im August, kurz vor bzw. während des Fegens des 2. Geweihs, steigt das HV dann wieder deutlich an. Dieser annuale Rhythmus im HV erscheint in den folgenden Jahren immer wieder in der gleichen Form. Die 4 Jahreskurven sind positiv und hochsigni- fikant miteinander korreliert (r = 0,95). Nur die Maximalwerte im September/Oktober steigen noch an im 3. LJ auf 71 ml und im 4. auf 104 ml. Vom 4. zum 5. LJ ist kein wei- terer Zuwachs mehr erkennbar. Die Minimalwerte der HV liegen in allen 4 Zyklen zwischen 10 und 20/24 ml. Ein signifikanter altersabhängiger Anstieg ist nicht zu erken- nen. Abbildung 1b zeigt die Jahreskurve der HV-Veränderungen bei adulten Damhirschen. Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) 167 3. | | 5.LJ Abb. 1. Der Pubertätsverlauf (Beginn bis Ende) bei männlichem Damwild. Die Entwicklung des Hodenvolumens in ml (x + SD): a) bei Jungtieren; b) bei adulten Männchen. Symbole: Abszisse: Monate; 1.-5. LJ; V u. A: Abwurf- und Fegetermine; [7]: Brunst. Halsumfänge (HU) Wie aus Abbildung 2a ersichtlich, steigt der HU vom ersten Meßtermin im Januar des 1. LJ von 24,7 cm kontinuierlich an bis zu einem Maximum von 37,1cm im Januar/Fe- A: . 60 u | 5 . EN | 22 | ,.| ..| 5.LJ V A A VEN V A V na "ATBIERENeNavaraTa nun MM) ar MMISW) MM Abb. 2. Der Pubertätsverlauf (Beginn bis Ende) bei männlichem Damwild. Die Entwicklung der Hals- umfänge (HU) incm (x+SD): a) bei Jungtieren; b) bei adulten Männchen. Symbole: siehe Abb. 1. 168 K. FiscHER bruar des 2. LJ. Es folgt ein steiler Abfall auf Minimalwerte von 32,7 cm zwischen März und Juli am Ende des 2./Beginn des 3. LJ. Im August des 3. LJ zum Fegetermin des 2. Geweihs hin wird der nächste Anstieg im HU sichtbar. Das nächste Maximum liegt zwischen Mitte November und Mitte Januar bei 44,5 cm. Das nächste Minimum wird zwischen Mai und Juli mit 38,4 cm am Ende des 3./Anfang des 4. LJ erkennbar. Im 4. LJ steigen die Maximalwerte der HU zwischen September/November auf 54,1cm an. Nach der Brunst nimmt der HU bis Dezember steil auf 47,8 cm ab, danach langsam bis zum nächsten Minimum auf 41,6 cm zwischen Mai und Juli am Ende des 4./Anfang des 5. LJ. Im 5. LJ erreicht der HU das Maximum im September und Oktober mit 59,7 cm. Bereits Mitte November, also noch während und gleich nach der Brunst, wird die Abnahme der HU deutlich. Im April/Mai kurz vor Beendigung des 5. LJ liegt das nächste Minimum bei 41,1 cm. Abbildung 2b zeigt die Jahreskurve der HU-Veränderungen bei adulten Damhirschen. In Abbildung 3 sind die Steigerungsgeraden durch die durchschnittlichen annualen Minima (a) und Maxima (b) im HU in den aufeinanderfolgenden Jahren eingezeichnet. Um zu einer einheitlichen Berechnungsgrundlage zu kommen, wurden die Minima festge- legt auf die Monate April und Mai. Das ist ab dem 2. Geweih der Zeitraum des Abwer- fens und des Beginns der Neubildung des nächsten Geweihs. Die Minima der HU 6 3 b) 60 IA +56cm ss 7 I +17,9cm 50 nen R be % iz 7,%&cm +8,75 cm S j isn Bel 40 17 + T 7 a) se A lke2sen +6,1cm Tr ZE 30 er 27 1. Dior went] a ala] un IN ASM JE SZASMINSI SE AZMEISEN ZA SSS OA Abb. 3. Entwicklung der HU (x + SD) im Pubertätsverlauf (Beginn bis Ende) bei männlichem Dam- wild. Die Steigungsgeraden: a) durch die Minima (April/Mai) in den annualen Zuwächsen; b) durch die saisonalen, androgenbedingten Maxima; Monatsangaben auf der Abszisse: reale Lage der Maxima und Minima im 1. -5. LJaus Abb. 2a: Ordinate: cm; |: Lot von den Maxima auf die die zugehörigen Minima verbindenden Steigungsgeraden. Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) 169 nehmen bis zum 4. LJ um ca. 13,6 cm zu. Zum 5. LJ hin ist kein weiterer Anstieg mehr zu sehen. | Die Maxima (b) sind in der Abbildung stets in die Mitte zwischen die vorangehenden und die nachfolgenden Minima gelegt. Die Monatsangaben auf der Abszisse geben ihre reale Lage in den Jahreskurven (Abb. 2a) wieder. Im 2. LJ beträgt der mittlere Abstand zwischen dem Maximum als Lot gefällt auf die die zugehörigen Minima verbindende Stei- gungsgerade 6,25 cm; im nächsten Zyklus sind es 8,75 cm; im 4. LJ beträgt die Differenz 13,42 und im 5. 17,9 cm. Die Gesamtzunahme im maximalen HU beträgt zwischen 2. und SIE NDPAGTEMm! Lebendgewichte (LG) Wie aus Abbildung 4a hervorgeht, steigt das LG der Damhirsche ab Meßbeginn im Ja- nuar von 28,1 kg bis zum September im 2. LJ steil an auf 58 kg. Danach flacht der Ge- wichtsanstieg bis Mai deutlich ab. Die Kurve zeigt eher eine fallende Tendenz. Der nächste signifikante Anstieg im LG um 5 kg (p: <0,01) wird erkennbar von Mai nach Juni am Ende des 2. LJ. Bis zum Dezemer im 3. LJ steigen die LG auf 71 kg. Bis zum April ist dann kein weiterer Anstieg im LG zu erkennen. Der nächste signifikante Anstieg des LG um 11,1 kg liegt zwischen April und Mai am Ende des 3. LJ, das nächste Maximum von 95,8 kg im September des 4. LJ. Über Oktober bis Mitte November verlieren die Hirsche 16,2 kg. Bis zum März des 4. LJ liegt das LG bei 80,8 kg. Der nächste Anstieg im LG um 7,7 kg (p: <0,01) wird deutlich zwischen März und April im 4. LJ. Das nächste Maximum liegt im September des 5.LJ bei 109 kg. Bis Mitte November verlieren die Hirsche 25,5 kg. Anschließend bis zum März des 5. LJ liegt dann das LG wieder bei 80. kg. Im Vergleich zum Vorjahr ist im Winter keine weitere Gewichtszunahme mehr erkennbar. Abbildung 4b zeigt die Jahreskurve der LG-Veränderungen bei reifen Damhirschen. In Abbildung 5a sind die Steigerungsgeraden durch die gemittelten LG zwischen Ja- nuar/März in den aufeinanderfolgenden LJ miteinander verbunden. Auf diesem Niveau nimmt das LG bis zum 4. LJ um 49,6 kg zu. Zum 5. LJ hin ist kein weiterer Anstieg mehr erkennbar. a a Be EISEN NBEM SERN ar MRMEEIDSSTEN EIER MEAN 2 SEON DRMeM | | | Bl 5 Abb. 4. Der Pubertätsverlauf (Beginn bis Ende) bei männlichem Damwild: Die Entwicklung des Le- bendgewichts (LG) inkg (x+SD): a) bei Jungtieren; b) bei adulten Hirschen. Symbole: - - - -: WRG; —: FS; andere: siehe Abb. 1. 170 K. FISCHER 120 b) 110 +13,2kg 100 PA: +25,5kg +24,8kg 80- TI FF —— a) N,1kg 70- y 50 MM +26, kg 2 ! % | 2. | 3. | [A | 5.1 | N J-M VE EN Abb. 5. Entwicklung der LG in kg (x + SD) im Pubertätsverlauf (Beginn bis Ende) bei männlichem Damwild. Die Steigungsgeraden: a) durch WRG (Januar-März): annualer Schub im allgemeinen Wachstum; b) durch die FS-bedingten Maxima. Monatsangaben auf der Abszisse: reale Lage der Maxi- ma und Minima aus Abb. 4a im 1.-5. LJ. Andere Symbole: siehe Abb. 3. In Abbildung5b verbinden die Steigungsgeraden die gemittelten annualen Maxima der LG. Aus Gründen der Übersichtlichkeit sind die Maxima in die Mitte zwischen die vorangehenden und die nachfolgenden Minima zwischen Januar/März gezeichnet. Die Monatsangaben auf der Abszisse geben ihre reale Lage in den Jahreskurven (Abb. 4a) wieder. Im 3. LJ beträgt die Differenz 8,1 kg; sie ist nicht signifikant, daher kann nur von einer steigenden Tendenz gesprochen werden. Im 4. LJ sind es 21,2 und im 5. 25,5 kg. Die Kurve zeigt zusätzlich die annualen Zuwächse zwischen den erreichten Maxima. Vom 3. zum 4. LJ beträgt er 24,8 kg und vom 4. zum 5. LJ nochmals 13,2 kg. Danach ist kein wei- terer Anstieg im Vergleich zu adulten Hirschen zu sehen (Abb. 4b). Diskussion Mit der vorliegenden Langzeitstudie zur Pubertät bei männlichem Damwild wurde in der 1. Januarhälfte des Geburtsjahres begonnen, weil in dieser Zeit bei den Hirschkälbern aus unserer Herde in der Regel das Rosenstockwachstum einsetzt (FISCHER et al. 1985). Die Rosenstöcke sind bei den Echthirschen ein sekundäres Geschlechtsmerkmal; ihr Wachs- tumsbeginn signalisiert den Beginn der Androgensekretion und damit das Einsetzen der Pubertät (LincorLn 1971a). Der Zeitpunkt hängt vom allgemeinen körperlichen Entwick- lungszustand ab (LincoLn 1971a; SuTTIE und Kay 1982). Bei den von CHAPMAN und CHAPMAN (1980) untersuchten Damhirschkälbern setzt das Rosenstockwachstum erst im Alter von etwa 9 Monaten im März ein. Ich führe diese zeitliche Differenz auf den opti- malen Entwicklungsstand unserer Tiere zurück. Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) Al Das Rosenstockwachstum ist bei sehr kräftigen Individuen schon gegen Mitte Februar abgeschlossen. Der äußerlich meßbare Wachstumsbeginn des 1. Geweihs setzt dann zwischen Mitte/Ende März und Anfang April ein (FiscHEr et al. 1985). Während des Wachstums der ersten Geweihgeneration kommt der erste Aktivitätsschub im reprodukti- ven System zum Stillstand und setzt erst kurz vor dem Fegetermin wieder ein (CHAPMAN und CHAPMAN 1980; CHAPMAN et al. 1981). Das wird verständlich aus der Tatsache, daß bei den borealen Echthirschen Geweihwachstum nur möglich ist, wenn die Blutandrogen- spiegel auf basalem Niveau liegen (BUBENIK 1982; RoLr und FiscHEr 1990; KoLLe et al. 1993). Da in dieser Zeit zudem die Spermabildung bei praktisch allen Individuen zum Er- liegen kommt (ASsHER et al. 1987; GoscH und FiscHEr 1989), kann diese Periode auch als „gonadale Ruhephase“ bezeichnet werden. Die gängigen Definitionen über die Pubertät sagen in der Regel nur etwas aus über ihren Beginn, nichts jedoch über ihren Verlauf und vor allem nichts über ihr Ende (Do- NOVAN und VAN DER WERF TEN BoscH 1965). So hat sich z.B. gezeigt, daß junge Rot- und Damhirsche (Spießerstadium) ohne die Konkurrenz älterer Männchen während der Brunst die Weibchen durchaus erfolgreich beschlagen können (LincoLn 1971 b; CHAPMAN und CHAPMAN 1975; FISCHER 1983). Niemand wird jedoch solch jungen Hirschen den Er- wachsenenstatus zugestehen. Das Problem, daß die reproduktive Potenz zwar vorhanden ist, das junge Männchen sich aber in der Sozietät noch nicht durchsetzen kann, ist älteren Autoren durchaus bewußt gewesen. Der Pubertätsbegriff wurde daher schon früh um den Begriff der sexuel- len Reife erweitert (AspELL 1946). Diesen Zustand erreicht ein männlicher Hirsch, wenn er sich eine effektive Rolle im Reproduktionsgeschehen der Population erobern kann. Auch diese Definition ist problematisch, denn ein Hirsch kann sehr wohl die sexuelle Reife erreicht haben; ob er sich aber während der Brunst gegen seine Rivalen durchset- zen kann, ist eine andere Frage. Um solche sekundären Einflüsse auszuschließen, defi- niert FISCHER auf der Basis morphologischer, physiologischer und endokriner Parameter das Ende der Pubertät wie folgt: „Es ist die Zeit im Leben eines Individuums, in der das allgemeine Körperwachstum faktisch abgeschlossen ist und alle primären und sekundären Geschlechtsmerkmale während der Fortpflanzungszeit ein Aktivitätsniveau aufweisen, wie es für adulte Hirsche charakteristisch ist“. Wie alt muß nun ein Damhirsch werden, um diesen Zustand zu erreichen? Um allgemeine und androgenbedingte Wachstumsprozesse im Pubertätsverlauf von- einander trennen zu können, benutzten Woopn et al. (1962) zur Beschreibung der LG-Ent- wicklung bei Maultierhirschen 4 Kurven: die erste für den Verlauf der Entwicklung von der Geburt bis zum Einsetzen der Pubertät. Diesen Teil haben wir nicht gemessen. Die 2. Kurve beschreibt die annualen Veränderungen im LG (und bei den primären und den sekundären Geschlechtsmerkmalen) bei adulten Individuen. Es sind die Bezugsgrößen, ge- gen welche die in der Ontogenese ablaufenden Veränderungen gemessen werden müssen. a) HV: Minimale HV finden sich bei den adulten Damhirschen in der gonadalen Ru- hephase. Vier bis 5 Wochen vor den Fegeterminen steigen sie wieder an und erreichen kurz vor und während der Brunst das Maximum. Gleich nach der Brunst gehen sie zu- nächst rasch und danach langsam bis zum Minimum am Beginn der nächsten gonadalen Ruhephase zurück (ASHEr et al. 1987; GoscH und FiscHErR 1989). Diese annuale Rhyth- mik im HV ist bei reifen Damhirschen über Jahre hinweg konstant, sofern keine Verän- derungen im sozialen Bezugssystem und im Gesundheitszustand eintreten. b) LG: Das LG des reifen Damhirsches zeigt einen ausgeprägten jahresperiodischen Verlauf, der sich in 3 deutlich getrennte Phasen unterteilen läßt: 1. Eine Periode hoher Gewichtszunahme um 25 bis 30% zwischen Mitte März und letzter Septemberwoche. 2. Eine sich direkt anschließende Vor- und Hauptbrunst umfassende Fastenperiode, in der der Hirsch die Nahrungsaufnahme fast vollständig einstellt und das zuvor zugelegte Gewicht wieder vollständig verliert. Der saisonale Gewichtsanstieg und die sich an- 172 K. FISCHER schließende etwa 6 Wochen dauernde Fastenperiode sind ein sekundäres Geschlechts- merkmal, das „Feistzeit-Syndrom (FS) (Fat-season-syndrom)“ (FiscHErR 1991). Die 3. Periode umfaßt die Wintermonate, in denen auch bei bestem Futterangebot eine Ge- wichtszunahme nicht zu erreichen ist. Die Tiere nehmen aber auch nicht ab, wenn nur Heu geboten wird. Wir sprechen daher vom „Winter-Ruhe-Gewicht“ (WRG). Wegen sei- ner Stabilität eignet sich das WRG besonders gut zur Beschreibung des allgemeinen kör- perlichen Entwicklungszustandes. Beginn, Verlauf und Wechsel zwischen FS und zwischen WRG sind beim reifen Damhirsch Jahr für Jahr voraussagbar. c) HU: In vergleichbarer Weise lassen sich beim adulten Damhirsch die saisonalen Veränderungen im HU beschreiben (FiscHEr 1991). Minimale Werte werden stets gefun- den in der Zeit zwischen dem Abwurf des alten Geweihs und dem Ende der Wachstums- periode des neuen. Mit dem Fegen steigt der HU wieder steil an und erreicht das Maximum kurz vor und während der Brunst. Danach geht der HU zunächst bis Dezem- ber steil und dann langsam zurück bis zum Erreichen des Minimums während der gonada- len Ruhephase. Nach Lincorn (1971b) sind diese saisonalen Veränderungen im HU ein sekundäres Geschlechtsmerkmal; es entfällt nach Kastration. Die 3. Kurve beschreibt nach dem Vorgehen von Wooo et al. (1962) die annualen Ver- änderungen der Minimalgewichte (und der Minima der primären und sekundären Ge- schlechtsmerkmale) im Laufe des Pubertätsgeschehens. Die Steigungsgeraden durch die WRG verdeutlichen die basale LG-Zunahme bis zum 4. LJ. Zum 5. L)J steigt die Kurve nicht weiter an. Bei diesem Parameter haben die Damhirsche somit im Winter ihres 4. LJ das WRG adulter Hirsche erreicht. Die Steigungsgeraden durch die annualen Zuwachsraten der Minima im HU spiegeln ebenfalls das allgemeine Körperwachstum wieder. Hier ist wie beim LG im 4. LJ der Er- wachsenenstatus erreicht. Beim Vergleich der annualen minimalen HV während der gonadalen Ruhephase ist zwischen 1. und 5. LJ ein statistisch faßbarer Anstieg nicht erkennbar. Die 4. von Woon et al. (1962) erstellte Kurve, beschreibt den annualen Zuwachs im Maximum der LG (und in den Maxima der primären und der sekundären Geschlechts- merkmale) im Verlauf der Pubertät. Die Kurven der Steigungsgeraden, welche die WRG einerseits, und die Maxima der LG während des Höhepunktes der Fettdeponierungspha- sen andererseits miteinander verbinden, driften ab dem 4. LJ deutlich auseinander. Das Auseinanderdriften der beiden Kurven ab dem 4. L)J markiert das erste signifikante Auf- treten des FS in der Ontogenese. Der annuale Zuwachs im maximalen LG erreicht aber erst im 5. LJ im September das Niveau reifer Damhirsche. Das FS ist somit erst ab dem 5. LJ voll ausgeprägt. Die saisonalen androgenbedingten Veränderungen im HU sind dagegen bereits im 2.LJ deutlich erkennbar und treten in den folgenden Jahren immer stärker in Erschei- nung. Die Endstufe wird im 5. LJ erreicht. Hierzu paßt sehr gut die in der Ontogenese ab- laufende Zunahme der annualen maximalen Plasma-Testosteron-Werte kurz vor und während der Brunst (RoLr und FiscHEr 1990). Auch bei diesem Kriterium wird erst im 5.1LJ der Erwachsenenstatus erreicht. Eine vergleichbare altersabhängige Zunahme im maximalen Testosteron-Spiegel finden auch BUBENIK und ScHAms (1986) beim Weißwe- delhirsch. Bei dem primären Geschlechtsmerkmal HV erreichen die Maxima im 4. LJ die End- stufe. Zieht man in diese Betrachtung den annualen Spermatogenesezyklus mit ein, so zeigt sich, daß beim Damhirschkälbchen während der ersten Brunstperiode in seinem Le- ben noch keine Spermatozoen im Ejakulat zu finden sind. Aber schon ab dem 3.LJ gibt es dann keinen Unterschied mehr zu den reifen Hirschen (GoscH 1992). Bei alleiniger Wertung dieses Kriteriums wäre ein Damhirsch bereits im 3. LJ erwachsen. Die hier vorgestellten Befunde zeigen, daß beim Damhirsch der Status des Adultus in unterschiedlichen LJ erreicht wird je nachdem, welche Kriterien verwendet werden. Be- Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) 173 trachtet man die Merkmale jedoch in ihrer Gesamtheit, so ist nach FiscHErs Definition beim Damhirsch das Ende der Pubertät im 4.LJ fast und im 5. voll erreicht. Ein Rothirsch erreicht diesen Zustand etwa ab seinem 7./8. LJ (LincoLn 1971 a). Anders lie- gen die Dinge bei Vertretern aus der Gruppe der Telemetacarpalia. Beim Schwarzwedel- hirsch z.B. tritt der annuale LG-Zyklus bereits im 2. LJ deutlich in Erscheinung (Wooo et al. 1962). Für den Rehbock finden SEMPERE und LaAcroix (1982) bereits zur Brunst im 2. LJ maximale HV und Testosteron-Spiegel. Pubertätsbeginn und -verlauf sind bei diesen Arten deutlich akzeleriert. Woop et al. (1962) berichten von einer erfolgreichen Verpaa- rung zweier Schwarzwedelkälber und beim Reh werden im Winter immer wieder träch- tige Kitze gefunden. Betrachtet man bei den einzelnen Parametern die zeitlichen Zuordnungen in den auf- einander folgenden Jahren genauer, so fällt auf, daß bei den sekundären Geschlechts- merkmalen Aktivitäts-Beginn und -Ende am Anfang der Pubertät beträchtlich von der zeitlichen Lage, wie sie bei adulten Hirschen gefunden wird, abweichen. Ihre endgültige Position in den Jahreskurven erreichen sie erst nach mehreren Jahren. FiscHEr spricht vom „Einschwing-Phänomen“. Nach FiIscHEr et al. (1985) liegen im ersten Geweihzyklus der Wachtumsbeginn der Spieße 3-4 und der Fegetermin gegen 2-3 Wochen früher als bei den Folgegeweihen. Das Abwerfen der Spieße dagegen erfolgt 3-4 Wochen verspätet. Erst mit dem Fegen des 2. Geweihs im 3. LJ wird die für ältere Damhirsche so typische und präzise zeitliche Fest- legung der Fege- und der Abwurftermine erreicht (FiscHeEr 1985). Deutlich kommt das Einschwing-Phänomen auch bei der Lage der HU-Maxima im Laufe der Ontogenese zur Geltung. Zwischen 2. und 5. LJ kommt es zu einer schrittwei- sen Vorverlegung um fast 3 Monate. Lincorn (1971a) beschreibt bei jungen Rothirschen eine Vorverlegung um 1 1/2 bis 2 Monate. Das Einschwing-Phänomen zeigt sich auch beim annualen Beginn des FS; er erfährt im Laufe der Pubertät eine Vorverlegung um ca. 8 Wochen. Dagegen ist bei den Parametern HV, Spermaproduktion und Testosteronspiegel der annuale Rhythmus bereits ab dem 2. LJ weitgehend, ab dem 3. dann völlig stabil. Nur die Maxima der HV und der Testosteronspiegel erhöhen sich nochmals im 4. bzw. im 5. LJ. Die Synopsis aller vorgestellten Parameter zeigt, daß der Damhirsch im 4. LJ weitge- hend, im 5. voll erwachsen ist. Von jetzt an sind mit großer zeitlicher Genauigkeit die jahresperiodisch wiederkehrenden Aktivitäts-Beginne, -Höhepunkte und -Enden bei pri- mären und sekundären Geschlechtsmerkmalen voraussagbar. Die Ergebnisse zeigen zudem, daß der Pubertätsverlauf kein Kontinuum ist; er wird vielmehr durch die dominante photoperiodisch gesteuerte Jahresperiodik in einzelne, auf- einander aufbauende Schübe zerlegt (GoscH 1992; RoLF und FIscHER 1996). Nach Lincorn (1971 bb) ähnelt bei reifen Hirschen der annuale Aktivitäts-Ruhe-Wech- sel im gesamten reproduktiven System einer jährlich wiederkehrenden Pubertät. Diese Vorstellung kann nach den hier gezeigten Befunden nicht aufrecht gehalten werden. Denn für die Entwicklungsvorgänge im reproduktiven System sind im Verlauf der Puber- tät neben der annualen Amplitudenerhöhung der sich über mehrere Jahre erstreckende Einschwing-Prozeß charakteristisch. Für den reifen Hirsch dagegen ist die erstaunlich prä- zise Jahresperiodik aller von uns untersuchten Parameter Wesensmerkmal. Zusammenfassung An Damhirschen wurden vom Alter von ca. 61/2 Monaten an (Beginn des Rosenstock- wachstums) bis kurz vor Vollendung des 5. LJ Pubertäts-Beginn, -Verlauf und -Ende an den Veränderungen im HV, LG, FS, WRG und HU verfolgt. Der Pubertätsverlauf wird durch die Jahresperiodik in aufeinander aufbauende Schübe zerlegt. Die einzelnen Para- 174 K. FISCHER meter erscheinen im Laufe der Ontogenese nicht gleichzeitig und zum Teil weichen ihre jahreszeitlichen Positionierungen anfänglich noch deutlich von der Lage ab, die für die adulten Hirsche so charakteristisch ist (Einschwing-Phänomen). Das Ende der Pubertät hat ein Individuum erreicht, wenn Beginn, Dauer, Ende, Maximum und Minimum der Aktivitäten bei allen Parametern in Amplitude und Frequenz das Niveau reifer Dam- hirsche erreicht haben. Dieser Zustand wird erst im 4. besser im 5. LJ erlangt. Die Stabili- tät im saisonalen Auftreten der Ereignisse bei reifen Damhirschen ist ein grundlegender Unterschied zu deren Instabilität während des Pubertätsgeschehens. Daher kann der an- nuale Rhythmus ım Fortpflanzungssystem der adulten Damhirsche nicht als eine saisonal wiederkehrende Pubertät bezeichnet werden. Danksagung Mein Dank gilt dem Minister für Ernährung, Landwirtschaft und Forsten des Landes Niedersachsen für die über Jahre hinweg gewährte finanzielle Unterstützung unserer Arbeiten am Damwild (Grant Nr. 401 F 04 032/1-1145/91 ff.). Literatur ASDELL, S. A. (1946): Patterns of mammalian reproduction. New York: Comstock, Ithaca. ÄSHER, G. W.; Day, A.M.; BARRELL, G.K. (1987): Annual cycle of liveweight and reproductive changes of farmed male fallow deer (Dama dama) and the effect of daily oral administration of melatonin in summer on the attainment of seasonal fertility. J. Reprod. Fert. 79, 353-362. Banpy, P.J.; Cowan, I. McT.; Woop, A.J. (1970): Comparative growth in four races of blacktailed deer (Odocoileus hemionus). Part I. Growth in body weight. Can. J. Zool. 48, 1401-1410. BUBENIK, G. A. (1982): The endocrine regulation of the antler cycle. In: Antler Devolopment in Cervi- dae. Ed. by R.D. Brown Kingsville, Texas: Caesar Kleberg Wildlife Research Institute. Pp. 73-107. BUBENIK, G. A.; SCHAMS, D. (1986): Relationship of age to seasonal levels of LH, FSH, Prolactin and tes- tosterone in male whitetailed deer. Comp. Biochem. Physiol. A, 83, 179-183. CHAPLIN, R. E.; WHITE, R. W. G. (1972): The influence of age and season on the activity of the testes and epididymides of the fallow deer, Dama dama, J. Reprod. Fert. 30, 361-369. CHAPMAN, D. J.; CHAMPMAN, N. G. (1975): Fallow deer. Lavenham: Terence Dalton Ltd. CHAPMAN, D.J.; CHAPMAN, N.G. (1980): Morphology of the male accessory organs of reproduction of immature Fallow deer (Dama dama L.) with particular reference to puberty and antler develop- ment. Acta Anat. 108, 51-59. CHAPMAN, D. J.; CHAPMAN, N. G.; KENNAUGH, J.H. (1981): Development of the preputial gland of imma- ture Fallow deer (Dama dama Linnaeus) with particular reference to puberty. Z. Säugetierkunde 46, 322-330. Donovan, B. T.; Van DER WERF TEN Bosch, J.J. (1965): Physiology of puberty. London: Edward Arnold. FISCHER, K. (1983): Untersuchungen zur Fortpflanzungsfähigkeit von jungem weiblichem und männli- chem Damwild (Dama dama L.). Z. Jagdwiss. 29, 137-142. FIscHER, K. (1985): Jahresperiodische Schwankungen physiologischer Parameter beim Damhirsch (Dama dama L.). In: Transact. 17th Congr. IUGB. Ed. by S. A. DE CROMBRUGGHE. Brüssel. Pp. 183- 190. FISCHER, K. (1991): Seasonal changes in neck girth and the fat season in male fallow deer. 2 secondary sex characteristics and their ontogenesis. In: Transact. 20th Congr. IUGB. Ed. by S. Csanyı and J. ERNHAFT. Gödöllö, Hungary. Pp. 547-551. FISCHER, K.; GOSCH, B.; MENNERICH, B. (1985): Individueller Pubertätsverlauf bei im Juni und verspätet im September gesetzten Damhirschen (Dama dama L.). Z. Jagdwiss. 31, 211-221. FRANZMANN, A. W.; LE RESCHE, R. E.; RAUSCH, R. A.; OLDEMEYER, J.L. (1978): Alaskan moose measure- ment in weight relationships. Can. J. Zool. 56, 298-306. GoscH, B.; FISCHER, K. (1989): Seasonal changes of testis volume and sperm quality in adult fallow deer (Dama dama) and their relationship to the antler cycle. J. Reprod. Fert. 85, 7-17. GoscH, G. (1992): Untersuchungen zur Jahresperiodik der testikulären Aktivität beim Damhirsch (Dama dama L.). Diss. Math.-Nat. Fachbereiche, Universität Göttingen. Der Pubertätsverlauf bei männlichem Damwild (Cervus dama) 178 GROVES, C. P.; GRUBB, P. (1987): Relationships of living deer. In: Biology and management of the Cervi- dae. Ed. by C.M. WEMMER. Washington, D. C.: Smithonian Institution Press. Pp. 21-59. KoLLE, R.; KIERDORF, U.; FISCHER, K. (1993): Effects of an antiandrogen treatment on morphological characters and physiological functions of male fallow deer (Dama dama L.). J. Exp. Zool. 267, 288- 298. LEADER-WILLIAMS, N. (1979): Age related changes in the testicular and antler cycles of reindeer, Rangi- fer tarandus. J. Reprod. Fert. 57, 117-126. Lincorn, G. A. (1971a): Puberty in a seasonally breeding male, the red deer stag (Cervus elaphus L.). J. Reprod. Fert. 25, 41-54. LincorLn, G. A. (1971b): The seasonal reproductive changes in the adult red deer stag (Cervus elaphus L.). J. Zool. (London) 163, 105-123. McEwan, E.H. (1968): Growth and development of the barrenground caribou. II. Postnatal growth rates. Can. J. Zool. 46, 1023-1029. RoLr, H. J.; FISCHER, K. (1990): Serum testosterone (T) and 5-a-dihydrotestosterone (DHT) in male fal- low deer (Dama dama): seasonality and age dependence. Comp. Biochem. Pysiol. A. 95, 445-452. RoLr, H.J; FISCHER, K. (1996): Serum testosterone, 5-a-dihydrotestosterone and different sex characters in male fallow deer (Cervus dama): a long term experiment with accelerated photoperiods. Comp. Biochem. Physiol. (in press). SEMPERE, A.J.; LACROIX, A. (1982): Temporal and seasonal relationships between LH, testosterone and antlers in fawn and adult male roe deer (Capreolus capreolus L.). Acta Endocrinol. 99, 295-301. SUTTIE, J.M.; Kay, R.N.B. (1982): The influence of nutrition and photoperiod on the growth of antlers of young red deer. In: Antler development in Cervidae. Ed. by R.D. Brown. Kingsville, Texas: Cae- sar Kleberg Wildlife Research Institute. Pp. 61-71. WHITEHEAD, P.E.; McEwan, E.H. (1973): Seasonal variation in the plasma testosterone concentration of reindeer and caribou. Can. J. Zool. 51, 651-658. WooD, A. J.; Cowan, I. M.; NoRDAN, H.C. (1962): Periodicity of growth in ungulates as shown by deer of the genus Odocoileus. Can. J. Zool. 40, 593-603. Anschr. des Verf.: Prof. Dr. Kraus FIscHER, I. Zoologisches Institut der Universität Göttingen, Berli- ner Straße 28, D-37073 Göttingen Z. Säugetierkunde 61 (1996) 176-188 FÜR An SÄUG ei INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY An ecological comparison between standard and chromosomally divergent House mice in Northern Scotland By GUILA GANEM, P. ALIBERT, and J. B. SEARLE Department of Zoology, University of Oxford, Oxford, U.K. Receipt of Ms. 10. 08. 1995 Acceptance of Ms. 08. 02. 1996 Abstract In western Europe there are populations of the house mouse that diverge in karyotype from the stan- dard 40-chromosome complement as a result of Robertsonian fusions between pairs of chromosomes. Very little is known about the ecology of such Robertsonian populations of the house mouse. The pre- sent study focuses on the Robertsonian system of north east Scotland. We distinguish seven “chromo- somal zones” along north-south and east-west transects (2?n=32 in the extreme north east, 2n=34S(outh), 2n = 34 W(est), 2n=368S,2n=36W,2n=40S, 2n=40 W). We describe and com- pare the different chromosomal zones in relation to their habitat and population characteristics. The 2n= 32 chromosomal race is characterised by the lowest density as well as the lowest frequency of preg- nant and lactating females. The highest densities are found in both the 2n = 40 zones. Intermediate chromosomal zones appear to be separated from each other by habitat where the house mouse does not occur. Thus, passive transport may represent the only means by which mice could move between zones. Within the 32, 34 and 36S zones movements of materials occur, while farmers from these areas have few, if any, interactions with farmers in the other zones. This pattern of agricultural contact would tend to facilitate chromosomal flow between the 32, 34 and 36S chromosomal zones and isolate them from the three other zones. Chromosomal flow between the 32 zone and adjacent areas may threaten the low-density, low-fertility 32-chromosome race with extinction. Introduction Several populations of the house mouse (Mus musculus domesticus) diverge chromoso- mally from the standard type (2n=40chromosomes, all with terminal centromeres), through the fusion of pairs of chromosomes at their centromeres (Robertsonian fusions). These mutations have been found in natural populations of mice and involve most auto- somes. Robertsonian populations occur in Europe and northern Africa (BAucHAU 1990), and based on archaeozoological and molecular data, it is largely accepted that this diver- gence took place less than 10,000 years ago (AUFFRAY et al. 1990 b, NACHManN et al. 1994). Very little is known about the ecology of Robertsonian populations of the house- mouse (AUFFRAY et al. 1990 a, GAnEM 1991, 1993; Saıp and Brıtton-DAviDIan 1991; BER- rY et al. 1992; HAUFFE 1993). This is due to the fact that most studies concerning these po- pulations have been aimed solely at a description of karyotypes. However, estimates of population size, rates of turnover and extinction, and patterns of social and geographical structure of nearby Robertsonian and standard populations could help to: a) understand past events which promoted the fixation of Robertsonian fusions and their subsequent spread (assuming that we can relate present-day ecology to past ecology), b) analyse pre- sent dynamics of hybrid zones between standard and Robertsonian populations, and c) make predictions about future spread or extinction of such populations. An ecological comparison between House mice 1777. This study focuses on Robertsonian populations occurring in the extreme north of mainland Scotland, where there is a race characterised by 32 chromosomes (homozygous for 4 centric fusions: 4.10, 9.12, 6.13, 11.14; ScRIvEN and BROOKER 1990; SEARLE 1991), which forms a hybrid zone with the more widespread standard type. The hybrid zone ex- tends along both the northern and eastern coast of the counties of Caithness and Suther- land. It has a staggered structure, such that there is a2n = 36 race (homozygote for two centric fusions: 4.10, 9.12) in the middle of the hybrid zone (SEARLE et al. 1993). This stag- gered structure allow to distinguish different “chromosomal zones” corresponding to zones where a particular diploid number predominate. This situation provides a particu- larly interesting model to address the above considerations. For the ecological study presented here we sampled seven “chromosomal zones” (Fig. 1), the 2n = 32 zone in the vicinity of John 0’Groats, areas immediately south and west of this zone where mice had karyotypes close to 2n=34 (34S and 34W respec- tively), and even more southerly and westerly zones characterised by 2n=36 and 2n=40 (368, 36 W, 40S, 40 W). This study will address the following questions, i) do the chromosomal zones present different habitat characteristics, related or not to agricultural practices? ii) are there differences in fertility between Standard and Robertsonian popula- tions? iii) are habitat and population characteristics relevant to the chromosomal evolu- tion of house mice in northern Scotland’? 2n=36W zone a Melvich Densities 3% 5 = >30 mice/100 night traps 10-30 mice/100 night traps © =] 0-10 mice/100 night traps Helmsdale 3 Altitudes 2n=40S EI <1i0om EEE 200-400m zone 100-200m EE >400m Fig. 1. Geographic location of the different chromosomal zones described in this study with special re- ference to the densities of mice in these zones. The area shown constitutes the counties of Caithness (to the north-east) and Sutherland (to the south-west} with the boundary occurring in a line approximately between Melvich and Hemsdale. 178 GUILA GANEM, P. ALIBERT, and J. B. SEARLE Material and methods Forty one different farms were visited during two field sessions (March, 1992; September, 1992), as well as three outdoor localities (completely independent of human activities). In March 1992, mice were found in 18 out of 26 farms sampled. In September 1992 out of 24 farms, 16 produced mice. Six farms were successfully sampled during both trapping sessions, these will allow to detect seasonal differences. A total of 280 mice were trapped during 1900 trapnights (Tab. 1). Despite extensive trapping in the out- door localities (300 trapnights) no house mice were trapped while wood mice (Apodemus sylvaticus) were numerous in one of the sites and common shrews (‚Sorex araneus) were present in another. Longworth traps were baited with peanut butter and were set in several contiguous and non-contig- uous rooms in each farm; the number of traps set depended on room sizes and varied between 2 to 20; the trapping period varied between 3 to 7 nights depending on trapping success. Trapping sites were noted precisely for each mouse, and when the same or very close traps (less than 1 metre apart) repeat- edly caught mice, this was recorded as evidence of spatial association of mice. From the data collected, population sizes for each farm are expressed as number of mice/100 trap nights, and sex ratio is described in terms of total number of males/total number of females. Trapping visits which resulted in nil returns were excluded from the analysis. For each chromosomal zone the mean body weight of mice caught, and the frequency distribution of body weights were also analysed. The reproductive condition of mice was assessed in live individuals in the field. For live individuals we calculated the ratio of the number of pregnant and lactating females/total number of females, and the ratio of number of males with descended testes/total number of males. Testis and seminal vesicle weights were measured on post-mortem specimens (Mettler AE163 balance) which were sacrificed after being kept for 3 months in male-female pairs under laboratory conditions. Results Habitat characteristics (Tab. 1, Fig. 1) Many small and some medium sized farms occur in the 32 zone. They are bordered by the sea, agricultural fields or heather moorland. Most of the farms do not have large stocks of grain and straw. The 34S zone is separated from the 32 zone by hilly ground covered in heather moor- land and peat bogs, where many wood mice occur (outdoor site n? 1). The 34S zone is connected to the 34W zone by a series of small and large farms. The 34W zone is separated from the 32 zone by large sand dunes. Some medium and large farms in the 34 W zone show intensive agricultural activity, very favourable for mice, and could constitute a mouse reservoir for the zone. The 36 W and 36S zones are separated from each other by geographical features such as hills, lochs, and woods, in which there are few farms. The 36S zone extends from Wick to Lybster, but is limited to a strip of land along the east coast. It is characterised by a se- ries of closely-located active small farms surrounded by cultivated fields. The 36 W zone occurs south west of Thurso, and south of Melvich Bay. It is limited to several small and medium sized farms occurring along the Halladale river. This agricultur- al strip is surrounded by natural landscape, mostly uninhabited by humans. It is separated from the 34W zone by natural geographical features. The 40W and 40S zones are approximately 75 km distant from each other, and coin- cide with the western and southern borders of Caithness. They are isolated from the clo- sest 36 zones by hills and mountains where farms are rare. They mainly consist of large to very large farms, which contain abundant resources for mice. The 40 W zone seems to be mainly limited to the area of Tongue which is surrounded by the sea, mountains and loch, uninhabited zones where farms are rare. In the different zones farmers tend to regulate mice by the use of cats; poison is not used extensively. Chromosomal zone An ecological comparison between House mice 179 Table 1. Description of localities where mice were caught Sampling session 1: March 1992. Sampling session 2: September 1992 L: large sized farm, > 400 m? M: medium sized farm, 200 < x < 400 m? S: small sized farm, < 200 m? Farm grid reference Sampling Number of Size Presence of Presence of other (Ordinnance Survey) session mice predators small mammals (cat, dog). Use of poison 3388/9709 Apodemus sylvaticus (Biel of Duncasby) 3378/9732 Apodemus sylvaticus (John 0’Groats 5) 3376/9719 (John 0’Groats 6) 3372/9734 (Mill) 3368/9734 (John 0’Groats 7) 3360/9736 (Pier) 3355/9724 Apodemus sylvaticus, (Canisbay) Microtus agrestis 3354/9729 Microtus agrestis (Seater) 3341/9729 Sorex araneus (Kirkstyle) 3357/9627 Apodemus sylvaticus (Keiss 1) 332719512 cat (Haster) 3344/9447 poison/cat (Thrumster) 3346/9437 cat (Sarclet) 3235/9380 (Smerlie 1) 3238/9379 cat (Smerlie 2) 3239/9379 poison Sorex araneus (Smerlie 3) 3070/9201 cat (Ousdale) 2957/9107 poison when 0 mice Apode- (Crakaig) mus sylvaticus; Sorex araneus 180 GUILA GANEM, P. ALIBERT, and J. B. SEARLE Table 1. (Continued) 2925/9078 (Brora 1) 2919/9078 Apodemus sylvaticus (Brora 2) 3234/9734 (Brough) 3211/9713 cat (Dunnet 2) 2895/9521 dog, cat Microtus agrestis (Bunahoun) 2894/9582 (Achiemore) 2889/9574 Apodemus sylvaticus (Upper Bighouse) 2888/9566 cat (Laidham) 2584/9539 poison (Ribigill) Population characteristics Six farms from the three main zones of the north-south transect (32, 36S, 40S) were suc- cessfully sampled during both trapping sessions. The analysis shows that densities within each farm, and thus within each zone, did not differ between sessions (Kruskal-Wallis test, KW =0.4, p= 0.5), however, densities were different between zones (KW = 9.4, p = 0.009; Fig. 2a). The same difference between zones was observed when the entire set of data was ana- lysed (KW = 12.75, p = 0.05; Fig. 2b). The Robertsonian 32 zone is characterised by the lowest density of mice, and the 40S and 40 W zones by the highest densities. Sex-ratios were not significantly different between zones. The average sex-ratio was 1.2 #0.2; this does not differ significantly from 1.0. Only in the 32 zone males showed a tendency to be more numerous than females during both sessions (respectively 69.2 and 62.0% of males). Taking all the different sampled farms, we found that various numbers (2-8) and combinations of mice (males, females or males and females, in different proportions of adults, more than 16 g, and immatures, less than 14 g) may occur in close spatial proxi- mity. The weight structure of populations was assessed by subdividing mice into six cate- gories (< 10 g, 10-12 g, 12-14 g, 14-16 g, 16-18 g, > 18 g). Males and females showed simi- lar distributions and there were no significant differences between chromosomal zones or between trapping sessions (Kolmogorov-Smirnov two samples tests p > 0.05). However, two patterns of weight frequency distribution occur (Fig. 3). One is the “adult dominant” type (Fig. 3a) found during both trapping sessions in the 36S zone, and during the first session in the 40S, 32 and 40W zones. The alternative “all age classes” type (Fig. 3b) characterised the 40S and 32 zones during the second session, and the 36 W zone during the first session. An ecological comparison between House mice 181 a) D session 1 (March, 1992) BE] session 2 (September, 1992) Mean number of mice/100 trap-nights 32 32 36S 368 365 40S Chromosomal zones b) Mean number of mice/100 trap-nights 40W 36W 34W 32 348 365 405 Ww NE S Fig. 2. a) Variation in number of mice (mean + S. E.) at six different farms along a 80 km north-south transect through the Scottish hybrid zone sampled during both trapping sessions. (Farms from left to right: John 0’Groats 5, Seater, Smerlie 1,2 and 3, Brora 1). b) Population size variation (mean + S.E.) along the two transects of the Scottish Robertsonian system, based on all successful trapping visits to farms. Reproductive conditions assessed on live individuals in the field Frequencies of pregnant and lactating females were not statistically different in the differ- ent zones (KW=38.73, p=0.19) and during the two trapping sessions (KW = 1.91, p=0.17). The mean frequency was 13.0%. However, population size and frequency of pregnant and lactating females were significantly correlated (Spearman rank correlation n = 33, rs = 0.67, p<0.001). The fewest pregnant and lactating females were found in the 32 zone (3%) where there was the lowest density of mice (3-10 mice/100 trap nights). 182 GUILA GANEM, P. ALIBERT, and J. B. SEARLE (a) Sarclet/males Thrumster/males Smerlie 1/males Smerlie 3/ males Smerlie 2/ males Sarclet/females Thrumster/ females Smerlie 1/ females Smerlie 3/ females Smerlie 2/ females S ENENDBNE St Number of mice —_ oO 10 002 oa nase eels weight (g) (b) Brora 1/males Brora 2/ males Brora1/ females Brora 2/ females HSEN Number of mice <10 10-12 12-14 14-16 16-18 >18 weight (g) Fig. 3. Examples of weight structures observed in different mouse populations. (a) The “adult domi- nated” distribution observed in the 36 S chromosomal zone during both trapping sessions. (b) The “all age classes” distribution of weights frequencies observed in the 40 S chromosomal zone during the sec- ond trapping session. The frequency of males with descended testes was significantly lower during the second session (18% of males) than the first (72% of males; KW = 7.84, p = 0.005), indicating a greater sexual activity in the spring. There were no significant differences between chro- mosomal zones but the 34W and 32 zones tended to have the lowest frequencies of such males. An ecological comparison between House mice 183 Reproductive conditions as assessed on post-mortem specimens after 3 months in the laboratory Males of the 40W, 40S, 36 W, 36S, and 32 chromosomal zones were compared for their testis and seminal vesicle size. Mice collected during session 1 were sacrificed in July 1992, those collected during session 2 in December 1992, we thus assumed that they were all adult when these organs were weighted, given that they were all at least 3 months old. At the time the mice were sacrificed, there was no significant variation in body weight, neither between zones nor between time of sampling (Tab. 2). Table 2. Body, testes and seminal vesicles weights (mean +SE) of mice from five different races of the Scottish hybrid zone sampled during the two trapping sessions. Races 2n=40W 2n=36W 2m 32 Trapping session N 10 11 7 11 19 (*18) 14 Bodyaweishz3163+0.9 716.5+1.6 18.1+1.5 15.1#0.9 17.6#0.6 16.912. 16.9+0.77 14.90.83 at capture (g) Body weight 19.5+0.6 17.4+0.5 20.9+1.3 20.2+#0.7 20.6+0.4 19.5+0.5 20.0+#0.5 19.0+0.7 after 3 months in lab (g) Testes weight 163.7+6.1 158.0+5.0 161.1+7.8 167.8+10.3 159.8+4.8 157.6+2.7 *116.6+4.5 138.2+4.2 (mg) Seminal 143.8+22.8 64.3+5.8 87.3+12.0108.9+12.7101.2+11.5107.8+13.9 58.1+6.8 68.5+8.6 vesicles weight (mg) Seminal vesicles/Body weight (x 10°) Body weight and seminal vesicle weight were found to correlate positively both for mice collected during the first and second sessions (rs=0.43, p=0.001, n=57 and rs = 0.72, p = 0.0001, n = 32, respectively). Therefore, it was considered appropriate to ex- press seminal vesicle weights per unit body weight. These “relative seminal vesicle weights” differed significantly between chromosomal zones for both the first session (Tab. 2; Anova, F4.49 = 8.83, p= 0.0001) and the second (Tab. 2; F>>9 = 5.47, p = 0.01). Mice from the 40S zone had the lowest relative seminal vesicle weight and those from the 40 W zone, the highest. Testis weights were not correlated with body weights and so raw values were com- pared (Tab. 2). Mice differed significantly among the different chromosomal zones during both trapping sessions (session 1: F44s = 17.45, p < 0.001; session 2: F5 29 = 8.13, p = 0.002). Mice from the 40S zone were characterised by the lowest testis weights for both sessions (p < 0.05). 184 GUILA GANEM, P. ALIBERT, and J. B. SEARLE Discussion Habitat and population characteristics House mice appear not to occur or to be rare in the outdoor (not agricultural fields) of Caithness and Sutherland. One can propose that either environmental conditions are not favorable for the house mouse to establish populations, still the house mouse is known to occur in various conditions when on islands (BERRY and Bronson 1992), or the presence of competitors such as Apodemus sylvaticus would limit house mouse to commensal habi- tats (BERRY 1986; NAVAJAS Y NAVARRO et al. 1989). In the latter case, considering the geo- graphical features of the area and the distribution of commensal habitats in Caithness and Sutherland, the chromosomal zones appear to be separated from each other by habitat unsuitable to mice, with the exception of the 34W and 34S and 36S zones. Passive trans- port might represent the only means by which mice could move between most of the chromosomal zones. Commensal habitats are considered to favour high densities of mice and to allow con- tinuous reproduction (BRonson 1979). Moreover, human habitat is thought to induce spa- tial substructure in mouse populations such that they would function more or less like metapopulations (BAKER 1990). Our results show that even though all mice studied were from commensal habitats, density and fertility could be variable. Low fertility was found where densities were low (the 32 zone). This result is surprising since one would expect that social influence would reduce the numbers of pregnant females in high density better than in low density populations (BRonson 1989). As far as the populations studied here are concerned, density appears not to be high enough to provoke reproduction reduction. Lower fertility in the 32 zone may be related to unfavourable environmental conditions, or to genetic peculiarities of these mice. Considering the variation in population structure across the hybrid zone there appears to be the following pattern (refer to Fig. 1). The 32 zone occupies a small area and con- sists of several small farms. The mice occur at low density and were the least fertile of the animals examined (in terms of number of pregnant and lactating females). The 34 and 36 zones occupy larger areas and consist of larger farms than in the 32 zone and the mice occur at high densities. Large farms and very high densities also occur in the 40 zones, but the 40W zone, at least, is very restricted in area. The particularly low density of mice in farms from the 32 zone was confirmed during both trapping sessions in the present study and during other field trips by JBS (pers. obs.: 1987 and 1989) and CHRISTIANNE PALMER (pers. comm.: 1994). Features of farming in Caithness and Sutherland Three characteristics are important to consider in relation to the dynamics of the Scottish Robertsonian system: the evolution of agricultural practices, human trade between chro- mosomal zones and the effects of poisoning. In the agricultural triangle of NE Caithness (32, 34 W, 34S chromosomal zones), tradi- tional stacks for the long-term storage of unthreshed oats, used to be very common. These stacks provided food and shelter for mice. The replacement of threshing by com- bine harvesting over the last half-century has probably involved a dramatic change in mouse populations (BERRY 1981). The use of stacks certainly maintained very high density populations of mice in this area, with very few extinction events since poisons were not used in the stacks. These days there are almost no traditional stacks. In 1992 we only observed them in two very localised areas of Caithness: in the village of Thrumster (36S zone), and at one farm in John 0’Groats. In 1989 there was also a stack in John o’Groats; trapping studies An ecological comparison between House mice 185 have confirmed the extremly high density of mice that occur in such structures (GG and JBS pers. obs.) At the time when stacks were commonly used in Caithness, mice might have been considerably more abundant in the 32, 34S and 34W zones than they are nowadays. In other parts of the studied area, agricultural practices probably have not changed so dra- matically, because stacks were not used and farming consists mainly of stock breeding. Another aspect of farms of considerable importance to our understanding of the Scot- tish hybrid zone is the degree of human contact between the different areas. Discussion with farmers from the different chromosomal zones provided us with interesting informa- tion regarding possible passive transport of the house mouse in this region. Within the 32, 34, and 36S zones there seems to be extensive movements of materials directly among farms or via cooperatives based in Wick. Farmers in the vicinity of John 0’Groats also ex- change agricultural goods with farmers in Orkney (archipelago north of Caithness). So, passive transport of mice could occur between these zones. However, the farmers in the 32, 34 and 36S zones have few, if any, interactions with the 36 W, 40W, and 40S farmers. In the 40S zone, farmers exchange materials among themselves and with farmers further south. A final observation on farming practice relates to the effects of poisoning on popula- tions of mice. Rodenticide is not commonly used in the area, but when applied can dra- matically reduce mouse populations. For example, the farm at Crakaig had a very large population in March 1992 but no animals were caught in September 1992, apparently as a result of poisoning. However, the farm had a substantial number of mice again in autumn 1994 (C. P. pers. obs.). Such “boom-bust” population changes, whether the result of poison or other fac- tors, may have profound impact on the population genetics of the house mouse. If the mice actually go extinct during the poisoning phase then recolonisation of the farm in- volves migration of mice from elsewhere, causing gene flow and (in the context of the Scottish hybrid zone) chromosomal flow. If the mice do not go extinct, then the poisoning phase may represent a population bottleneck causing enhanced genetic drift and de- creased genetic heterozygosity. These constitute important factors which could influence the rates of fixation of Robertsonian fusions (MICHALAKIS and OLIVIERI 1993). Difference in testis weight between 40 S and other mice The specimens that were maintained in captivity for three months were very informative. Although individuals from the different chromosomal zones did not differ in body weight, the 40S males had smaller testes than mice from the other chromosomal zones. Testis weights have been examined for a second chromosomal hybrid zone in the house mouse. In Upper Valtellina (northern Italy), HAUFFE (1993) found that standard mice had signifi- cantly larger testes than Robertsonian mice, contrary to our finding that certain standard race mice (40S) had small testes. Given this discrepancy and the difference in testis weights between the 40S (small testes) and 40 W (large testes) mice in Scotland, there is clearly no consistent relationship between testis weight and karyotype. Testis size may reflect the mating system of the individuals examined. Relatively large testes would be expected in species or populations where there is substantial sperm compe- tition, i.e. apromiscuous mating system (KEnAGY and TROMBULAK 1986). In shrews (Sorex), there is evidence that sperm competition is more likely when individuals are at high density (STOCKLEY and SEARLE, in prep.). Thus, on the basis of present population sizes, we may have expected 40 S mice to have testis as large as 40 W mice do, which is not the case. Inves- tigations on inbred strains of the house mouse have shown considerable variation in testis size (HAYwARD and SHIRE 1974), and that at the different stages of the ontogenesis and growth of a mouse, testis size is under the control of the X and Y chromosomes as well as 186 GUILA GANEM, P. ALIBERT, and J. B. SEARLE autosomal genes (Hunt and MırttwocH 1987, CHußg 1992). Thus small testes in mice from the 40S zone may reflect genetic variabılity between populations of mice. Evolution of the Scottish hybrid zone It is difficult to make any sensible deductions about the origin of the 32-chromosome race in Caithness from our ecological studies. It could have arisen at any stage during the 2,000 years the house mouse are known to have existed in Britain (BROTHwELL 198]; NACHMANN et al. 1994). It is not clear, for instance, whether the race arose when the den- sity of mice was low or high, although further archaeological studies and genetical analy- sis may give some insight into this point. One interesting feature of the hybrid zone between the 32-chromosome John 0’Groats race and 40-chromosome standard race is that it has a staggered structure. Thus, instead of the 40-chromosome race directly abutting the 32-chromosome race, there is an inter- mediate 36-chromosome form (and mice with 34 chromosomes are found at the contact of the 32- and 36-chromosome races). This staggered structure could have arisen at the time of formation of the hybrid zone or it could have been a more recent event (SEARLE 1991; SEARLE et al. 1993). While our ecological study does not help us distinguish between the various models to explain the origin of the staggered hybrid zone, it shows that hu- man behaviour could be very important to take into account in order to understand popu- lation features of the commensal house mouse. Changes in population density over the last 50 years may have influenced the degree of staggering within the hybrid zone. To illustrate this we present one possible scenario. We propose that the agricultural triangle in northeastern Caithness, presently incorporating both the 32 and 34 chromosomal zones, may have formerly been occupied almost exclu- sıvely by 32-chromosome mice, at high density because of the presence of many oat stacks. This 32 zone could have been in equilibrium with neighbouring 34 zones, with a balance of dispersal in and out of the agricultural triangle. However, with the disappearance of the oat stacks and decrease in mouse density within the agricultural triangle, there may, in recent times, have been a greater tendency for mice to migrate into the triangle than out. In this way, the acrocentrics 6, 11, 13 and 14 (found in the 34 zones) may have penetrated the 32 zone (characterised by metacentrics 6.13 and 11.14), such that the 34 zones moved into the agricultural triangle. A 34 zone represents polymorphism, such that chromosomes 6, 11, 13, 14, 6.13 and 11.14 are all found in the same area. Thus the separation (or “stagger”) of the John 0’Groats and standard races would have been enhanced. Clearly, on this scenario, the acrocentrics 6, 11, 13 and 14 may be continuing to spread at the expense of metacentrics 6.13 and 11.14. If so, and unless behavioural traits inhibit the spread of “foreign” mice into the 32 zone GANEM and SEARLE in press., ultimately the 32- chromosome race may go extinct such that the agricultural triangle will become occupied by the 36-chromosome race. Whether or not the hybrid zone has been evolving in this way, the John 0’Groats race does appear to be endangered. It occurs in a very limited area at low density and may be susceptible to extinction by disease or further adverse changes in agricultural practice. Acknowledgements G.G. and J.B. S. were funded by the French Foreign Ministry (Bourse Lavoisier) and the Royal Society of London, respectively. We are very grateful to the Percy Sladen Fund (Linnean Society of London) for financial help with field work expences. Tim JonEs and JEFF Rırz provided invaluable assistance in the field, Mr. Rırz also drew the map, and R. SCHWEITZER provided a german translation of the ab- stract. G.G. thanks the Laboratoire de Genetique et Environnement for facilities during the writing of this study. This study would not have been possible without the generosity of farmers in Caithness and Sutherland. An ecological comparison between House mice 187 Zusammenfassung Ein ökologischer Vergleich zwischen Standard- und chromosomal abweichenden Hausmäusen in Nordschottland Im Westen Europas kommen Populationen der Hausmaus vor, die als Folge Robertson’scher Fusionen vom Standard-Chromosomensatz (?n =40) abweichen. Über die Ökologie solcher Robertson’scher Hausmauspopulationen ist sehr wenig bekannt. Die vorliegende Arbeit behandelt ein karyotypisch he- terogenes Gebiet im Nordosten Schottlands. Wir unterscheiden sieben „chromosonale Zonen“ entlang zweier Nord-Süd- bzw. Ost-West-Transsekte (2n=32 im äußersten Nordosten, 2n=34 S(üden), 2n=34 W(esten), 2n=36S,2n =36W,2n=40S,2n=40W). Die verschiedenen chromosomalen Zo- nen werden bezüglich ihrer Habitat- und Populationseigenschaften beschrieben und verglichen. Die chromosomale Rasse 2 n = 32 ist durch die niedrigste Dichte sowie die geringste Häufigkeit von trächti- gen und säugenden Weibchen charakterisiert. Die höchsten Dichten wurden in beiden 2n =40 Zonen gefunden. Intermediäre chromosomale Zonen scheinen durch Habitat voneinander getrennt zu sein, in dem die Hausmaus nicht vorkommt. Folglich wäre passiver Transport das einzige Mittel, durch das sich Mäuse zwischen den Zonen bewegen können. In den Zonen 32, 34 und 36S finden landwirtschaftliche Transporte statt, während die Bauern dieser Gebiete, wenn überhaupt, nur wenig Kontakt zu Bauern der anderen Zonen haben. Dieses Muster landwirtschaftlicher Beziehung sollte chromosomalen Fluß zwischen den Zonen 32, 34 und 36S erleichtern und diese von den anderen drei Zonen isolieren. Durch chromosomalen Fluß zwischen der Zone 32 und den angrenzenden Gebieten könnte die durch niedrige Dichte und geringe Fruchtbarkeit gekennzeichnete 2 n = 32-Rasse vom Aussterben bedroht werden. References ÄUFFRAY, J.-C.; BELKHIR, K.; CASSAING, J.; BRITTON-DAVIDIAN, J.; CRosET, H. (1990 a): Outdoor occur- rence in Robertsonian and standard populations of the house mouse. Vie et Milieu 40, 111-118. ÄUFFRAY, J.-C.; VANLERBERGHE, F.; BRITTON-DAVIDIAN, J. (1990 b): The house mouse progression in Eura- sia: a paleontological and archaeozoological approach. Biol. J. Lin. Soc. 41, 13-25. BAKER, A.E.M. (1990): Does group selection occur in commensal house mice (Mus domesticus)? In: Living in a patchy environment. Ed. by B. SHORRocKS and I. R. SWINGLAND Oxford: Oxford Science Publ. Pp. 197-218. BAucHAu, V. (1990): Phylogenetic analysis of the distribution of chromosomal races of Mus musculus domesticus Rutty in Europe. Biol. J. Linn. Soc. 41, 171-192. BERRY, R.J. (1981): Town mouse, country mouse: adaptation and adaptability in Mus domesticus (M. musculus domesticus). Mammal Rev. 11, 91-136. BERRY, R. J. (1986): Genetical processes in wild mouse populations. Past myth and present knowledge in current Topics in Microbiology and Immunology, Vol. 127, 86-94. BERRY, R.J.; BERRY, A.J.; ANDERSON, T.J.C.; SCRIVEn, P. (1992): The house mice of Faray, Orkney. J. Zool. London 228, 233-246. BERRY, R.J.; Bronson, F.H. (1992): Life history and bioeconomy of the house mouse. Biol. Rev. 67, 519-550. Bronson, F.H. (1979): The reproductive ecology of the house mouse. Quart. rev. Biol. 54, 265-299. Bronson, F.H. (1989): Mammalian reproductive biology. Chicago, London: Univ. of Chicago Press. BROTHWELL, D. (1981): The Pleistocene and Holocene archaeology of the house mouse and related spe- cies. Symp. Zool. Soc. London 47, 1-13. ChuBB, C. (1992): Genes regulating testis size. Biol. Reprod. 47, 29-36. GANEM, G. (1991): Commensalisme, fonction cortisosurrenalienne et Evolution chromosomique chez la souris domestique. These de Doctorat, Universite Montpellier I. GANEM, G. (1993): Ecological characteristics of the house mouse Robertsonian populations. Is their ha- bitat relevant to their evolution? Mammalia 57, 349-357. GANEM, G.; SEARLE, J. B. (in press): Behavioural discrimination among chromosomal races of the house mouse (Mus musculus domesticus) in J. Evol. Biol. HAUFFE, H.C. (1993): Robertsonian fusions and speciation in a house mouse hybrid zone. DPh thesis, University of Oxford. HAYWARD, P.; SHIRE, J.G.M. (1974): Y chromosome effect on adult testis size. Nature 250, 499-450. 188 GUILA GANEM, P. ALIBERT, and J. B. SEARLE HunT, S. E.; MırrwocH, U. (1987): Y-chromosomal and other factors in the development of testis size in mice. Genet Res. Camb. 50, 205-211. KENAGY, G. J.; TROMBULAK, S.C. (1986): Size and function of mammalian testes in relation to body size. ]. Mammalogy 67, 1-22. MICHALAKIS, Y.; OLIVIERT, I. (1993): The influence of local extinctions on the probability and time to fixation of chromosomal rearrangements. J. Evol. Biol. 6, 153-171. NACHMAN, M. W.; BoYER, S.N.; SEARLE, J. B.; AQUADRO, C. F. (1994): Mitochondrial DNA variation and the evolution of Robertsonian chromosomal races of house mice, Mus domesticus. Genetics 136, 1105-1120. NAVAJAS Y NAVARRO, M.; CASSAING, J.; CROSET, H. (1989): Demographie et dispersion d’une population sauvage insulaire de Mus musculus domesticus: comparaison avec une population continentale. Z. Säugetierkunde 54, 286-295. SaiD, K.; BRITTON-DAVIDIAN, J. (1991): Genetic differentiation and habitat partition of Robertsonian house mouse populations (Mus musculus domesticus) of Tunisia. J. Evol. Biol. 4, 409-427. SCRIVEN, P.N.; BROOKER, P.C. (1990): Caithness revisited: Robertsonian chromosome polymorphism in Caithness house mice. Heredity 64, 25-27. SEARLE, J.B. (1991): A hybrid zone comprising staggered chromosomal clines in the house mouse (Mus musculus domesticus). Proc. R. Soc. London B 246, 47-52. SEARLE, J. B.; NARAIN NAVARRO, Y.; GANEM, G. (1993): Further studies of a staggered hybrid zone in Mus musculus domesticus (the house mouse). Heredity 71, 523-531. Authors’ addresses: GUILA GANEM and P. ALIBERT, Laboratoire de Gen£tique et Environnement, Insti- tut des Sciences de l’Evolution, Universite Montepellier II, C.C. 065, Pl. E. Batail- lon, F-34095 Montpellier cedex 05, France and J. B. SEARLE, Department of Biology, University of York, PO Box 373, York YO15YW, U.K. > N Z. Säugetierkunde 61 (1996) 189-191 ZEITSCHRIFT SF © 1996 Gustav Fischer, Jena SÄUG ETl ERRÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY WISSENSCHAFTLICHE KURZMITTEILUNG The desert hedgehog, Paraechinus aethiopicus (Ehrenberg, 1833), new to the fauna of Syria By D. Kock and C. EBENAU Forschungsinstitut Senckenberg, Frankfurt a. M. Receipt of Ms. 05. 01. 1996 Acceptance of Ms. 29. 02. 1996 The mammal fauna of desert regions of Syria is still imperfectly known both with regard to species number and to distribution. Recent field work in the Euphrates valley (by C.E.) resulted in the addition of a third species of hedgehog, Paraechinus aethiopicus (Ehrenberg, 1833), to the Syrian fauna. Previously only Erinaceus concolor Martin, 1838, and Hemiechinus auritus (Gmelin, 1770) were known from this country (HARRISON and BATEs 1991). Besides Syria, only Israel, Palestine and Jordan host such a high species di- versity of spiny hedgehogs (Erinaceinae) due to the geographic position at the crossing of three faunal realms, the Palaearctic, Oriental and Aethiopian region. Material: Cater Magara Cave at Hussein al-Achmad village, 35.53.N - 39.01.E, ca. 2.5km W of Ratla, S-bank Euphrates, II. 1993. unsexed subad. (Skull and mandibles) Senckenberg-Museum Frankfurt SMF 79445, C. EBENAU leg. Comparative material: Saudi Arabia: Riyadh, 1958, unsexed (skull, skin) SMF 19919. Near Abgeig (= Abgaig), 25.26.N - 49.40.E, unsexed (skull) American Museum Nat. His- tory, New York, AMNH 166942 (labeled “hypomelas”, emend. IX.1977, not measured). P.ae. pectoralis (Heuglin, 1861): Jordan: Azraq area, 24. III. 1977, male (skull, skin) SMF 54967, R. KINZELBACH leg. Measurements, in this sequence: Ratla, Riyadh, Azraq: Occipito-premaxillary length 46.2, -, 50.8; condylobasal length 46.55, 47.6, 50.8; zygomatic width 28.55, 27.5, 29.5; inter- orbital constriction 12.9, 12.5, 13.65; postorbital constriction 11.6, 10.5, 11.5; mastoid width 26.1, -, 27.9; upper toothrow I1/-M3/, alveolar 21.9, 21.75, 24.5; mandible length from con- dylus 32.2, 34.7, 36.6; mandibular angular length 34.6, —, 37.5; PM4/ width 3.58, 4.43, 4.42. The broadly flared zygomatic arches, high brain case, large tympanic bullae, and ante- ro-dorsally elongated condylar process of the mandible identify the present Syrian speci- men as P.aethiopicus (cf. Frost et al. 1991) and differentiate it from other Arabian hedgehog species. In skull characters it agrees with a large sized specimen of P. ae. pecto- ralis (loc. typ. Petra, Jordan) from Azraq by inflation of bullae and pterygoids; it has, how- ever, the upper canine less strong and the last upper premolar narrower. The taxonomy of the Saharo-Arabian desert hedgehogs related to P. aethiopicus varies from recognition of several species to a reduced number of subspecies; for reviews, discus- sions and divergent taxonomy see CorBET (1988), Frost et al. (1991), HARRISON and BATEsS (1991), and NADER and AL-Sarapı (1993). For the populations of P.aethiopicus in the Middle East, Harrıson and BArtEs (1991) consider local Arabian subspecies other than P. ae. pectoralis of doubtful validity, P. ae. Judlowi Thomas, 1919 possibly excepted be- cause of ıts lighter coloration. Subspecies definitions for P. aethiopicus in Arabia are thus not established. Coloration as a character is not available in the Syrian specimen. In skull 190 D. Kock and C. EBENAU measurements, condylobasal length excluded because of subadult age, it does not differ appreciably from the pooled measurements given by HARRISON and BATEs (1991) for P.aethiopicus throughout Arabia and by NADER and AL-SAraDı (1993) from SW Arabia. The reduced size of PM4/ has not been used as a criterion for subspecies definition. De- spite of originating from the region where P. ae. ludlowi is expected to occur (loc. typ. Hit, south bank of the Euphrates, Iraq), the length of the upper toothrow (I1/-M3/, per- manent dentition) differs obviously (21.9 versus 25.2 in /udlowi; THoMAs 1919). P. aethiopicus occurs widely in Arabia (Harrıson and BArtEs 1991; additionally PALF- ERY 1988: Summan Plateau; NADER and AL-SAFADI 1993: numerous localities in SW Saudi Arabia and nothern Yemen; Kock and NADER 1996). The present specimen documents the first record from Syria and extends the species range by ca. 350 km NW from Haditha (Hart 1959), the nearest Iraqi locality downstream in the Euphrates valley. The presently known northwestern limit of the species range in the Middle East is marked by the following records from SW to NE (Fig. 1): Fig. 1. Paraechinus aethiopicus: New record in Syria (triangle) and northwestern records known (dots) in the Near East (after HArrıson and BATESs 1991 and this study). In Israel: Beer Sheba, Negev (Harrıson and Bates 1991); in Palestine: Adamah Bridge 40 km N Dead Sea (Yom-Tov 1988); in Jordan: Azraq Shishan (Boyp 1966); Dib- been, SW of Jerash; Qasr Amra; Shaumari Wildlife Reserve (CLARKE 1977); Wadi Dhu- layl; Mafrag; Qasr al-Halabat (Amr and Dısı 1988); in Iraq: 7 miles S Oasr al-Helqum, 33.48.N - 40.35.E, Syrian Desert (HArrıson 1964); 40 km E of Rutbah, 33.03.N - 40.18.E, Syrian desert (NADACHowskI et al. 1990); Hadithah, 34.07.N - 42.32.E (Hart 1959), Hit (Thomas 1919), and Hamam al-Alil 25 km SE of Mossul, 36.10.N - 43.16.E (BHATNAGAR and EL-Azawı 1978). Paraechinus aethiopicus new to the fauna of Syrias 191 This new find of P. aethiopicus enlarges the area of sympatry with Hemiechinus auritus (see CoRBET 1988; HArrRIısoNn and BArtEs 1991). As far as presently recorded, A. auritus ranges into the Syrian Desert west to a line running north-east from Damascus — Oarya- tayn — Palmyra (Harrıson and Bares 1991) — Rasafah (= Ressafe, Sergiopolis; SMF 60370) and along the south bank of the Euphrates from Halabiye (SMF 74067) southeast to Duro Europos (= As-Salahiya; SMF 80624) and into Iraq. P. aethiopicus and H. auritus have as yet not been found to occur syntopically in the Syrian desert, but may do so along the Euphrates valley. The present specimen, found as a carcass at the cave en- trance, may have originated from nearby desert habitat and could have been transported to this site. References AMR, Z. S., Disı, A. M. (1988): Jordanian mammals acquired by the Jordan University Natural History Museum. Amman: University of Jordan. BHATNAGAR, A. N.; EL-AZAwI, T. F. (1978): A karyotype study of chromosomes of two species of hedge- hogs, Hemiechinus auritus and Paraechinus aethiopicus (Insectivora: Mammalia). Cytologia 43, 53- 5% Bovp, J. M. (1966): International Jordan Expedition 1966. Nature 212, 663-665. CLARKE, J. E. (1977): A preliminary list of Jordan’s mammals. Amman: Royal Society for Conservation of Nature. CoRBET, G. B. (1988): The family Erinaceidae: A synthesis of its taxonomy, phylogeny, ecology and Zoo0- geography. Mamm. Rev. 18, 117-172. FRoST, D. R.; WOZENCRAFT, W. CH.; HOFFMANN, R. (1991): Phylogenetic relationships of hedgehogs and gymnures (Mammalia: Insectivora: Erinaceidae). Smiths. Contrib. Zool. 518, 1-69. HARRISON, D. L. (1964): The mamals of Arabia, I. Introduction, Insectivora, Chiroptera, Primates. Lon- don: E. Benn. HARRISON, D. L.; BATES, P. J. J. (1991): The mammals of Arabia. 2nd ed. Sevenoaks: Harrison Zool. Mus. HATT, R.T. (1959): The mammals of Iraq. Miscellan. Publ. Mus. Zoll. Univ. Michigan 106, 1-113. Kock, D.; NADER, I. A. (1996): The mammal fauna of the Jubail Marine Wildlife Sancturary. In: A ma- rine Wildlife Sanctuary for the Arabian Gulf. Environmental Research and Conservation following the 1991 Gulf War Oil Spill. Ed. by F. Krupp, A. H. ABUZINADA, and I. A. NADER, Frankfurt a.M., Riyadh: Forschungsinst. Senckenberg, Natn. Commission Wildl. Conserv. Development (in press). NADACHOWSKI, A.; SMIELOWSKI, J.; RZEBIK-KOWALSKA, B; DAouD, A. (1990): Mammals from the Near East in Polish collections. Acta zool. cracov. 33, 91-120. NADER, 1. A.; AL-SAFADI, M. M. (1993): The Ethiopian hedgehog Paraechinus aethiopicus (Ehrenberg, 1833) and Brandt’s hedgehog Paraechinus hypomelas (Brandt, 1836) (Mammalia: Insectivora: Eri- naceidae) from Northern Yemen. Fauna Saudi Arabia 13, 397-400. PALFERY, J. (1988): Observations on the behaviour of the white-crowned black wheatear in eastern Ara- bia. Sandgrouse 10, 1-25. Thomas, O. (1919): Some new mammals from Mesopotamia. J. Bombay nat. Hist. Soc. 26, 745-749. Yom-Tov, Y. (1988): The zoogeography of the birds and mammals of Israel. In: The zoogeography of Is- rael. The distribution and abundance at a zoogeographical crossroad. Ed. by Y. Yom-Tov and E. TCHERNov. Dordrecht, Boston, Lancaster: Dr. W. Junk Publisher. Pp. 389-409. Authors’ addresses: Dr. DIETER Kock, Forschungsinstitut Senckenberg, Senckenberg-Anlage 25, D-60325 Frankfurt a.M. and CARSTEN EBENAU, Barchemhöhe 27, D-45357 Essen, Germany. Z. Säugetierkunde 61 (1996) 192 # FÜR ee, ZEITSCHRIFT ee ale INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY MITTEILUNGEN DER GESELLSCHAFT Announcement Euro-American Mammal Congress: Challenges in Holarctic Mammalogy Santiago de Compostela, Spain, 1998 A joint meeting between the American Society of Mammalogy (ASM) and the Societas Europaea Mammalogica (SEM), will be held in Santiago de Compostela, Spain in 1998. Technical sessions are tentatively scheduled for 20-24 July. The meeting will be hosted by the Societad Espanola para el Estudio y Conservacion de los Mamiferos (SECEM) and the Universidad de Santiago de Compostela under the sponsorship of the Xunta de Galicia (Autonomous Community Government). The purpose of the meeting is to pro- mote Holarctic Mammalogy and cooperation among North American and European mammalogists; and to foster cooperation among, and the development of Societies of Mammbology. Located in the Northwestern region of Spain known as Galicia, Santiago de Compos- tela is the final stop on the medieval pilgrimage to the tomb of Saint James. The city is the focal point of a large number of medieval crossroads; these roads have been and re- main places of cultural exchange, of encounters and communication among all types of people and ideas across the centuries. It is this intellectual fervour from which was born the University of Santiago de Compostela, which celebrated its 500" anniversary in 1995. This cultural diversity is likewise surrounded by and linked to a very rich biodiversity formed by Marine ecosystems, Atlantic Mountains, and areas with Mediterranean influ- ence; all of these areas are in a good state of conservation. The first circular describing the meeting in more detail will be mailed in May 1996. To suggest a symposium topic or to receive more information prior to receive the first circu- lar contact Steering Committee co-chairs: Dr. FERNANDO PALAcCIOS, Museo Nacional de Ciencias Naturales, CSIC, Jose Gutierrez Abascal, 2, Madrid 28006 SPAIN; e-mail: Fer- nan@pinarl.csic.es; FAX (34) 1 564-5078; or Dr. Luıs RuUEDAS, Museum of Southwestern Biology, Department of Biology, University of New Mexico Albuquerque, NM 87 131- 1091 USA, e-mail: Lruedas@sevilleta.unm.edu; FAX (1) 505 277-0304. Redaktionelle Hinweise Einsendung und Annahme von Manuskripten: Ms. zur Veröffentlichung sind einzusenden an den 1. Schriftleiter: Herrn Prof. Dr. Dieter Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel. Über die Annahme von Ms. zur Veröffentlichung wird gemäß Geschäftsordnung der Deutschen Gesellschaft für Säugetierkunde entschieden. Der Eingang von Ms. wird sofort bestätigt, und nach Erhalt der Gutachten wer- den die Autoren über die Entscheidung umgehend informiert. Schriftwechsel, der die technische Drucklegung betrifft, ist zu richten an: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena. Äußere Form der Manuskripte: Die Ms. sind 3fach und vollständig als Original mit 2 Kopien in deutscher oder englischer Sprache einzureichen. Sie müssen mit Schreibmaschine oder als deutlich lesbare Computerausdrucke auf DIN A4-Blättern (21x 29,7 cm), mit einem linken Heftrand von 4cm und durchgehend mit einem doppelten Zeilenabstand geschrieben sein. Jedes Ms. soll auf der 1. Seite folgende Angaben enthalten: Titel der Studie; Name(n) der Autoren (männliche Vornamen nur in Initialen, weibliche ausgeschrieben); Institution(en), an der (denen) die Studie durchgeführt wurde. Ferner soll bei längeren Titeln ein kurzer Kolumnentitel (Seitenüber- schrift) von maximal 72 Anschlägen angegeben werden. Die Schrifttypen müssen einheitlich sein, ohne halbfette, gesperrte, kursive etc. Hervorhebungen. Fußnoten auf den Seiten oder Appendices am Ende sind nicht zugelassen. Autorennamen im Zusammenhang mit Literatur- zitaten sind im Text mit Bleistift zu unterstreichen. Wissenschaftliche zoologische und botanische Gattungs-, Art- und Unterartnamen sind mit einer Schlangenlinie zu unterstreichen. Der Name von Erstbeschreibern bei wis- senschaftlichen Namen wird nicht unterstrichen. Im Ms. ist am Rand mit Bleistift zu vermerken, an welcher Stelle die einzelnen Abbildungen und Tabellen ein- gefügt werden sollen. Am Ende der Studie, im Anschluß an das Literaturverzeichnis, sind die genauen postalischen Adressen des (der) Autoren anzugeben. Umfang und Aufbau der Manuskripte: Manuskripte können als Originalarbeiten oder als wissenschaftliche Kurz- mitteilungen veröffentlicht werden. a) Originalarbeiten: Originalarbeiten sollen einschließlich Abbildungen, Tabellen und Literaturverzeichnis einen Gesamtumfang von 30 Ms.-Seiten nicht überschreiten. Der Text soll eine klare Gliederung aufweisen in: Abstract (in englischer Sprache), Einleitung, Material und Methode, Ergebnisse, Diskussion (oder zusammengefaßt als Er- gebnisse und Diskussion), Danksagungen, Zusammenfassung und Literatur. Bei deutschsprachigen Studien ist über dem englischen Abstract der Titel in englischer Sprache anzugeben. Bei englischsprachigen Studien ist entsprechend über der deutschen Zusammenfassung der Titel in deutscher Sprache anzugeben. b) Wissenschaftliche Kurzmitteilungen: Wissenschaftliche Kurzmitteilungen sollen einen Gesamtumfang von 5 Ms.-Seiten nicht überschreiten. Sie enthalten kein Abstract und keine Zusammenfassung. Bei deutschsprachigen Ms. ist jedoch auf der 1. Seite zusätzlich eine englische Übersetzung des Titels anzugeben, bei englischsprachigen entsprechend eine deutsche Übersetzung. Wissenschaftliche Kurzmitteilungen sind äußerlich nicht in die Kapitel Einleitung, Material und Methode, Ergebnisse und Diskussion gegliedert, müssen aber in der textlichen Darstel- lung diese Strukturierung erkennen lassen. Abbildungen: Anzahl und Größe der Abbildungen sind unbedingt auf das notwendige Mindestmaß zu beschrän- ken. Es können nur kontrastreiche, reproduktionsfähige Originale verwendet werden, farbige Abbildungen wer- den nicht angenommen. Beschriftungen müssen einheitlich und wegen späterer Verkleinerung groß und kräftig sein. Auf der Rückseite sind laufende Numerierung, Zeitschrift und Autorennamen anzugeben. Alle Bildunter- schriften sollen auf einem gesonderten Blatt gelistet werden. Tabellen: Tabellen müssen auf ein Mindestmaß beschränkt werden. Jede Tabelle muß auf einem gesonderten DIN A4-Blatt angelegt sowie mit Überschrift und laufender Numerierung versehen sein. Umfangreichere Tabel- len müssen auf Folgeblättern als Fortsetzung erscheinen. Literaturverzeichnis: Jede zitierte Arbeit ist in alphabetischer Reihenfolge aufzuführen. Es soll nur international erreichbare Literatur zitiert werden. Die Zeitschriften werden mit ihrem Kurztitel angegeben. Beispiele: a) Zeitschriftenbeiträge: KArko, E. K. V; HANDLEY, €. O. Jr. (1994): Evolution, biogeography, and description af a new species of Fruit-eat- ing bat, genus Artibeus Leach (1821), from Panamä. Z. Säugetierkunde 59, 257-273. b) Bücher: HERRE, W.; RöhHrs, M. (1990): Haustiere - zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) Handbuchbeiträge: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) -— Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. by J. NIETHAMMER und F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/I, 7-28. Drucklegung: Der jeweilige Hauptautor erhält sowohl die Andrucke der Bilder zur Prüfung wie auch den komplet- ten Umbruch seines Artikels zusammen mit dem Ms. Es dürfen nur tatsächliche Satzfehler korrigiert werden. Vom Ms. abweichende Änderungen werden dem Autor in Rechnung gestellt. Es sind die üblichen Korrekturzeichen zu verwenden. Ein als druckfertig bezeichneter Abzug ist zu senden an: Herrn Prof. Dr. Peter Langer, Institut für Anatomie und Zytobiologie, Justus-Liebig-Universität, Aulweg 123, D-35385 Giessen. Sonderdrucke: Verfasser von Originalbeiträgen und Wissenschaftlichen Kurzmitteilungen erhalten 50 Sonder- drucke gratis. Mehrbedarf kann gegen Berechnung bei Rücksendung des Umbruchs bestellt werden. Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective, or for resale. For copying from back volumes of this journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 100537, D-07705 Jena; Telefon (03641) 6263, Telefax (03641) 6265.00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A. Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zeitschriftenvertrieb, Frau B. Dressler, Post- fach 1005 37, D-07705 Jena; Telefon (03641) 62 64 44, Telefax (03641) 62 65.00. For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 4288201. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2945,-/382,50 (plus postage). Single issue price (DM/ÖSI/SFr): 80,-/592,-/77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 86010090; Deutsche Bank, Konto- Nr. 3907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400.00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY , Organ der Deutschen Gesellschaft für Säugetierkunde _ Beolchini, Francesca; Dupre£, E.; Loy, Anna: Territorial behavior of Talpa romana in two different habitats: food re- sources and reproductive needs as potential causes of variation. — Revierverhalten von Talpa romana in zwei verschiedenen Habitaten: Nahrung und Reproduktionsbedürfnisse als potentielle Gründe der Variation ........... 193 Rieger, I.: Wie nutzen Wasserfledermäuse, Myotis daubentonii (Kuhl, 1817), ihre Tagesquartiere? - How do Dauben- En ebentonit, use tbeit day TOOSiIS?...creneennnnenneenaeenesnnnnanen nenn anna en een nenn sanieren 202 Pandolfi, M.; De Marinis, Anna Maria; Petrov, I.: Fruit as a winter feeding resource in the diet of Stone marten (Martes foina) in east-central Italy. - Früchte als Winternahrung des Steinmarders Martes foina in Mittel-Ost-Italien........ 215 Virgös, E.; Casanovas, J. G.; Bläzquez, T.: Genet (Genetta genetta L., 1758) diet shift in mountains of central Spain. — Nahrungswechsel bei Ginsterkatzen (Genetta genetta L., 1758) in den Gebirgen von Zentralspanien ...............- 221 Krystufek, B.; Hrab£, V.: Variation in the baculum of the European souslik, Spermophilus citellus. - Variation des Ba- | culum zwischen Populationen des Europäischen Ziesels Spermophilus citellus ...........2ss@sssseeeeeenee nennen nenn 228 Barbosa, A.; Benzal, J.: Diversity and abundance of small mammals in Iberia: peninsular effect or habitat suitability? — Diversität und Abundanz bei Kleinsäugern in Spanien: Halbinsel-Effekt oder Lebensraumanpassung? ............- 236 Jänossy, D.: Kleinsäuger aus dem Pleistozän und Holozän Südost-Europas und des Mittleren Ostens. — Small mam- mals from the Pleistocene and Holocene of Southeast-Europe and the Middle East ..............zuussssss2sseeeeennnnn 242 Wissenschaftliche Kurzmitteilung dos Reis, S. F.; Pombal, Jr., S. F.; Pombai, J. P.; Nessimian, J. L.,; Pessöa, Leila Maria: Altitudianal distribution and feeding habits of Blarinomys breviceps (Winge, 1888) (Rodentia: Muridae). - Höhenverbreitung und Nahrungsge- wohnheiten von Blarinomys breviceps (Winge, 1888) (Rodentia: Muridae).............uus2022222sssHsnen nn nenensen nenn 253 LTR NIE Re N RRE te + © VS6 PT ORRERE 10) Pie AR REED SE Ey DUBTERURSE LERE 192 Indexed in Current Contents “Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database GUSTAV ISSN RTRORER.NN a ee: ” y; SEMPER FISCHER Vol, ie N ansuee JENA «STUTTGART» NEW YORK August 1996 BONIS ARTIBUS ASS U © ZEITSCHRIFT FÜR A: SÄUGETIERKUNDE INTERNATIONAL JOURNAL SAW OF MAMMALIAN BIOLOGY < TERKUNDE Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kiel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel - R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel - M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt — P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) -— W. Herre, Kiel (1962-1966) - H. Frick, München (1972-1976) - M. Röhrs, Hannover (1977-1981) — H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amntierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H.G. Erkert, Tübingen - W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel — Marialuise Kühnrich, Hamburg n, =: Z. Säugetierkunde 61 (1996) 193-201 ZEITSCHRIFT & = FÜR © 1996 Gustav Fischer, Jena SÄUG ETIE RKÜU N DE INTERNATIONAL JOURNAL %7 MAMMALIAN BIOLOGY LıB Territorial behaviour of Talpa ro RARIES two different habitats: food resources and reproductive needs as potential causes of variation By FRANCESCA BEOLCHINI, E. DUPRE, and AnnA Loy Department of Animal and Human Biology, University of Rome ‘La Sapienza’, Rome, Italy Receipt of Ms. 02. 10. 1995 Acceptance of Ms. 15. 03. 1996 Abstract Differences observed in the territorial behaviour of two populations of the Roman mole, Talpa romana, have been explored through quantitative and qualitative analyses of food resources available in the soil. Samples of soil were collected and analysed to detect the composition and the biomass of the soil fauna. Data are related to spatial and temporal differences observed in the home ranges of the two popula- tions. Results indicate a general inverse relationship between the availability of food and the size of ter- rıtories, which is particularly evident for females. Differences in the pattern of food dispersion in the two areas are associated with the spatial distribution of female moles, which in turn relates to the terri- torial behaviour of males during the reproductive season. Introduction The Roman mole (Talpa romana) is a fossorial insectivore endemic to south-central Italy. Like the better known T. europaea, it is a solitary and territorial species, spending most of its active periods in the defence of home range boundaries and in search of food, repre- sented by the invertebrate fauna inhabiting the soil (Loy et al. 1992, 1994 a, b). In the last four years we have carried out radio-telemetric studies in two different habitats of central Italy, an olive grove and a pasture, in order to collect data on the territorial behaviour of the species. The results revealed some differences in the spatial distribution pattern of the two populations, which mainly involved seasonal variation in the size and the relative po- sition of home ranges (Loy et al. 1994 a, b). In an attempt to explain these differences, we focused our attention on the amount, composition and dispersion pattern of food re- sources available in the soil. This choice was based on the consideration that food repre- sents the most important variable in a substantially homogeneous habitat like the subsoil (Nevo 1979), and that it is considered to be a primary force in competition among indivi- duals (OstrELp 1990). Moreover, according to the resource dispersion hypothesis (RDH) of MAcDonALD (1983), spatial and temporal variation of resources are thought to be main factors affecting territory size and social behaviour. Material and methods The two radiotelemetric studies were carried out at Formello, near Rome, from February 1991 to Janu- ary 1992, and at Beviglie, near Assisi, from October 1992 to May 1993. The same techniques of data re- cording and analysis were employed in both studies to facilitate comparisons. Results and details about techniques have already been published for the two areas (Loy et al. 1992, 1994 a, b). E, and AnnA Loy E. Dupr b) FRANCESCA BELOCHINI 194 oe] elldie) s]laue)ı ofewoy ofewo] XIdoN HIVAa XS AJAOAD HATIO HUOD day-uoN AOIAHd („u) AH XIHoN ofeusF ejldue)) Sfeu1sJ SfeweJ Sfeu1s] ofews ofeus] SfewsJ ofeuo] ofewu ofeuu ofewo] ofeusF ofeup] fewsJ ofewe] A1VA xHS JANLSVd HAOD day-uoN AOIAHAd ‘(q 'ep661 Te 19 AOT WOAJ P9J10d91 Je eyep) (XII „N) SPXL Jo 1oquınu 0} Jurp.IOHse poyJou UOFAJOA X9AUOT wnunum >yJ Aq Psınduo9 spouiod (doy-uoN) sAansnpo1ds1-uou ayy pue (day) sAnanpo1daı sy Jurımp sPeILgey Om} Hy) ur Poy9e.1 speunue Jo („w UI YH) SOdue.ı 9woy uo eJeq] "T OIgeL Territorial behaviour of Talpa romana 195 The first area consisted of a degraded pasture land at an altitude of 138 m where the soil was typi- cally volcanic and characterised by stratified tufa. The other area was an olive grove in a hilly area at an altitude of about 400 m, where the soil was rich in carbonates and clay. Fourteen moles (7 males, 7 females) were successfully radio-tracked in the pasture and 9 moles (4 males, 5 females) in the olive grove. Since some animals were racaptured and monitored more than once, the data refer to a total of 31 tracking sessions for the two areas. Details of the animals tracked and the estimated sizes of their home ranges are drawn in table 1. The soil fauna for both areas was analysed by sampling both sites at the same time of the year, in December 1993 and April 1994, in order to account for possible changes in the invertebrate fauna du- ring the non-reproductive and reproductive periods of moles. Ten random samples of soil were col- lected in each habitat, each one consisting of 64 litres of earth from a cubic pit with a side of approximatively 40 cm, corresponding to the usual depth of the mole tunnels observed in the two areas. The earth was immediately sieved, and all visible invertebrates collected and stored in 70% alcohol. In- vertebrates were then sorted, weighed and classified to ordinal level within a few days, in order to avoid weight loss caused by alcohol dehydration. To verify the correspondence between the diet of the moles and the quantitative and qualitative composition of soil fauna, a sample of 18 moles, 6 males and 12 females, was kill-trapped in the olive grove in April and November 1992 and in April and November 1993, and their stomach content ana- lysed to assess the composition of preys. All stomachs were preserved in 5% formalin, according to the method outlined by SKOCZEN (1966), and their contents weighed and then sorted and identified using a binocular microscope. The total and relative biomasses of soil fauna found in the two areas were correlated with the aver- age size of home ranges of males and females through multiple correlations and analyses of variance. The use of the average values was due to the lack of data on the soil fauna present inside each territory, which was in turn related to the fact that collection of soil invertebrates was performed at a different time with respect to the radio-tracking sessions. Data were analysed separately for the reproductive and the non-reproductive season. To compare the spatial distribution of soil fauna in the two areas, the Kol- mogorov-Smirnov test for homogeneous distribution of the biomasses was conducted. All statistical analyses were computed using the packages SAS (version 6.0, 1985) and STATISTICA (version 4.3, 13) Results The composition of the soil fauna in the two areas is shown in table 2. In terms of bio- mass, the fauna of both areas was dominated by Oligochaeta (percentages vary from 60% to 90% of total biomass, with the lowest values found in the olive grove). Other impor- tant groups were Mollusca, Coleoptera (larvae and adults), Dermaptera, Lepidoptera (larvae and pupae) and Hymenoptera. These other invertebrates are consistently present in the olive grove, while in the pasture they represent less than 20% of the total biomass. Thus, the soil fauna of the pasture appears less diverse with respect to the olive grove, even if almost the same number of invertebrate classes is present in both areas. In the olive grove little variation was observed in the quantity of earthworms during different seasons, while Mollusca almost doubled in biomass during autumn-winter; the larvae of Coleoptera in the pasture land also showed a significant increase during spring (Tab. 2). A comparison of the soil fauna composition in the olive grove and the stomach con- tents of moles captured in that habitat reveals that the two parameters are related (Fig. 1), with values for Pearson correlation coefficient between relative biomasses ran- ging from 0.96 to 0.99 (P < 0.0001). This evidence indicates that Talpa romana is an Oppor- tunistic feeder, as was already described for T. europaea (SKOCZEN 1966; OPPERMANN 1968; FunMILAYo 1979) and suggests that the abundance and distribution of soil invertebrates can be used as a general indicator of food availability for moles. Total biomass of soil invertebrates collected in the two areas in winter and spring is shown in table 2. Soil fauna in both periods was less abundant in the olive grove than in 196 FRANCESCA BELOCHINI, E. DUPRE, and AnnA Loy Table 2. Relative weights and percentages of the different classes of soil invertebrates collected in the two habitats. OLIVE.GROVE, 7 PASTURE77 7 O0LIVE-GROVEZ ZPASIURE spring spring winter winter (8) % Mollusca Oligochaeta Oligochaeta, cocoons Arachnida Chilopoda Diplopoda Isopoda Diplura Protura Embioptera Dermaptera Heteroptera Lepidoptera, pupae Lepidoptera, larvae Diptera, pupae Diptera, larvae Coleoptera, adults Coleoptera, pupae Coleoptera, larvae Hymenoptera, adults Hymenoptera, pupae Hymenoptera, larvae Totals the pasture, an evidence confirmed by the GAmEs and Howerı (1976) test for equality of means run on the two sets of sampled biomasses in pasture and olive grove, which is sig- nificant at a level of P< 0.05. The total biomass in the olive grove did not change over the two periods, while in the pasture a marked seasonal variation was observed, with a great- er abundance in autumn-winter. The relation between the abundance of soil fauna and the size of territories in the two habitats was analysed by computing the correlation between the average size of home ranges of males and females during the reproductive and the non-reproductive periods, and the total biomass of soil invertebrates during the two periods. An inverse relation be- tween territory size and the biomass of preys in the soil is shown in figure 2: a reduction in the size of territories is observed when food availability increases. The smaller terri- tories are those of male and female moles of the pasture land during the non-reproduc- tive season (males x = 1665 m’, S.D. = 181; females x = 1761.8 m”, S.D. = 1 047.9), which is also the period of maximum abundance of soil invertebrates (54 g). The only contradic- tory value to the observed trend is represented by the home range size of males in the pasture during the reproductive period (coded MRP in Fig. 2). This increase of male terri- tories during the reproductive season land has been related by Loy et al. (1994 a) to the reproductive behaviour, i.e. the active searching of sexually receptive females, rather than to trophic needs. In fact, if we exclude this contradictory value, the inverse relation be- tween biomasses and territory sizes becomes significant (r = -0.383; P< 0.05). Results of multiple correlations between the variation of home range sizes of males and females and Territorial behaviour of Talpa romana 197] 100.00 7 80.00 + 3 soil 2 Ol stomach contents 8 60.00 + = = ® 2 40.00 + o 23 20.00, 0.00 lm 8- a oO [®) = & 2: em IN < ee: ın je!) © oa Een 22 8 aa a 8 je 77) > Q SEO» 5 3 o ”o © & 2 oo = ° = non-reproductive season 100.00 7 80.00 - B soil 2 Dlstomach contents s 60.00 + = o = 40.00 + m 20.00 - [DDOT OD and ae De ee ee oO o be! Ss En Our: m < Si 1 un ® SHARGL, (een E EV TEIOEEOgELa en 8 Re ee — reproductive season Fig. 1. Comparison between the percentages of various invertebrates found in the stomach contents of 18 moles captured in the olive grove, and in the soil fauna of the same area, analysed separately for the non-reproductive (top) and reproductive (bottom) periods (arac = Arachnida; coc = Oligochaeta co- coons; chil = Chilopoda; col.a = Coleoptera, adults; col.| = Coleoptera, larvae; derm = Dermaptera; dip = Diplopoda; dit.l = Diptera, larvae; heter = Heteroptera; hymen = Hymenoptera; isop = Isopoda; lep.| = Lepidoptera, larvae; olig = Oligochaeta). the variation of biomasses of different classes of invertebrates, pooled over the two areas and periods, are shown in table 3. A significant correlation is found between the size of female home ranges and the amount of earthworms, slugs, arachnida and isopods, while the size of male home ranges is related only to the minor occurrences present in the pasture solely during spring (Pro- tura, larvae of Coleoptera and Hymenoptera, pupae of Diptera and Hymenoptera). The only inverse significant relation is found between the size of female territories and the amount of earthworms, that always occurs in the soil of the two habitats and constitutes the greatest percentage to the total soil fauna biomass (Tab. 2). Finally, the patterns of food distribution in the two habitats (Fig. 3) were compared through the Kolmogorov-Smirnov test for homogeneous distribution, run on the two sets of data. The test results indicate a higher d-value for the pasture than for the olive grove (pasture land: d = 0.421 P< 0.01; olive grove: d = 0.283 P< 0.1), showing that the soil fau- na is more homogeneously distributed in the olive grove than in the pasture land. 198 FRANcCESCA BELOocHINI, E. DuPprR£, and AnnA Loy 14000 12000 10000 8000 6000 4000 MCP mean home range estimate (square meters) Total soil fauna biomass (g) Fig. 2. Scatter plot showing the relation between total soil fauna biomass and the mean home range size of males and females in the two areas for the two periods. The first coded letter refers to the sex (M = male; F = female), the second to the period (P = pre-reproductive, R = reproductive), and the third to the habitat (O = olive grove; P = pasture). The regression line (r = -0.383, P< 0.05) has been computed excluding the contradictory value MRP, which refers to mean territory size of males in the pasture during the reproductive season (see text for details). Table 3. Correlation between biomasses of diverse soil invertebrates and home range size of males and fe- males, pooled over the study areas and the different seasons. Levels of significance: *=P<0.1;**=P< 0.05. home range of males home range of females Mollusca Oligochaeta Oligochaeta, cocoons Arachnida Chilopoda Diplopoda Isopoda Diplura Protura Embioptera Dermaptera Heteroptera Lepidoptera, pupae Lepidoptera, larvae Diptera, pupae Diptera, larvae Coleoptera, adults Coleoptera, pupae Coleoptera, larvae Hymenoptera, adults Hymenoptera, pupae Hymenoptera, larvae Territorial behaviour of Talpa romana 199 25000 pasture BEE olive grove 20000 en) E a 15000 & = = #= = 5 = 10000 ® 5000 samples Fig. 3. Distribution of biomasses of soil fauna in each sample in the two areas. Discussion The relation between stomach contents and soil fauna composition confirms that Talpa romana is an opportunistic feeder, whose diet reflects the most abundant classes of inver- tebrates occurring in the soil. This feeding behaviour was already described by one of the authors in a wider sample of Roman moles (Loy 1992), and it is consistent with that de- scribed for Talpa europaea (SKOCZEN 1966; OPPERMANN 1968; FUNMILAYO 1979). The territorial behaviour of male and female moles appears to be influenced differ- ently by food and reproductive needs. According to our results, the size of female terri- tories in both habitats is related to the availability of food, i.e. each female mole tends to defend an area, the size of which is adequate to guarantee the food supply in the long term. This consideration is also supported by other data on the territorial behaviour of the species, and particularly those regarding the use and turnover of the core areas (Loy et al. 1994 a). While this relationship between food supply and territory size is always con- firmed for females, in both habitats and periods, it does not adequately explain the avail- able data on the distribution pattern of males. The seasonal variation of the male territories observed in the pasture land suggests that this is more influenced by reproduc- tive needs than by the availability of trophic resources. Specifically, the size of male terri- tories appears to be influenced by the spatial distribution of females, which we suppose to be related to the kind of dispersion of food resources. The dis-homogeneous distribution of soil fauna observed in the pasture land is likely to imply a more scattered distribution of females; as a consequence males should widely enlarge their home ranges in order to include the territories of more than one female. In this kind of habitat increased male mobility during the breeding season may become a primarily selected trait, as has been observed for other solitary and territorial species of small mammals (SCHWAGMEYER 1988; OstFELD 1990). By contrast, a different mating be- haviour should characterize the male moles in the olive grove, where a more homoge- neous distribution of food resources would induce females to live in closer adjacent 200 FRANCESCA BELOCHINI, E. DUPRE, and Anna Loy territories. Therefore males do not extend their home ranges much to include those of several females. The existence of different factors affecting the territorial behaviour of males and fe- males is supported by the hypothesis that the mating system commonly adopted by these territorial species is a form of polygyny in which the location of sexually receptive fe- males represents the male’s main competitive hurdle for reproductive success (GORMAN and Stone 1990; Loy et al. 1994 a). Although our data are not sufficient to represent an indisputable proof for our hypothesis, they suggest that the reproductive behaviour of moles is in accordance with the general statement that among polygamous rodents and insectivores the reproductive success of males depends on access to females, while for females it depends strongly on acquisition of food (TRıvErs 1972; EMLEN and Orıng 1977; Tew and MAcDonALD 1994). Acknowledgements We would like to thank GIUSEPPE CARPANETO for his valuable help in the diagnostic work on soil inverte- brates, and Prof. Luıcı BoITAnI for revising the manuscript. Prof. ERNESTO CAPANnNA warmly encouraged this work by offering both financial support and valuable suggestions. The project was supported partly by Consiglio Nazionale delle Ricerche, C. N. R. (project N. 93.04294.04), and partly by the Italian Minis- tero della Ricerca Scientifica e Tecnologica, MURST 40%. Zusammenfassung Revierverhalten von Talpa romana in zwei verschiedenen Habitaten: Nahrung und Reproduktionsbe- dürfnisse als potentielle Gründe der Variation Die Unterschiede im Revierverhalten von zwei Talpa romana-Populationen wurden durch qualitative und quantitative Analysen der verfügbaren Nahrungsdichte im Boden erforscht. Es wurden Bodenpro- ben gesammelt und untersucht, um die Zusammensetzung und Biomasse der Bodenfauna zu schätzen. Die Ergebnisse wurden in Beziehung zu den beobachteten Unterschieden in der Reviergröße der zwei Populationen gestellt; es ergab sich eine umgekehrte Relation zwischen der verfügbaren Nahrungs- dichte und der Reviergröße, die bei den Weibchen besonders deutlich war. Die unterschiedliche Nah- rungsverteilung (homogen im ersten, inhomogen im zweiten Gebiet) wurde in Beziehung zu der unterschiedlichen Verteilung der Weibchen gebracht und diese mit dem Revierverhalten der Männchen während der Paarungszeit. References EMLEN, S. T.; OrInG, L. W. (1977): Ecology, sexual selection, and the evolution of mating systems. Science 197, 215-223. FUNMILAYO, ©. (1979): Food consumption, preferences and storage in the mole. Acta Theriol. 24, 379- 389. GAMES, P. A.; HowELL, J. F. (1976): Pairwise multiple comparison procedures with unequal N’s and/or variances: A Monte Carlo study. J. Educ. Stat. 1, 113-125. GORMAN, M. L.; STONE, R. D. (1990): The natural history of the mole, London: Cristopher Helm. Loy, A.; DuprE, E.; STONE, R. D. (1992): Biology of Talpa romana Thomas (Mammalia, Insectivora: Tal- pidae). 1. Home range and activity patterns: preliminary results from a radiotelemetric study. Rend. Lincei, Sc. Fis. Mat. Nat. 9, 173-182. Lov, A.; DUPRE, E.; CAPANNA, E. (1994 a): Territorial behaviour in Talpa romana, a fossorial insectivore from southcentral Italy. J) Mammalogy 75, 529-535. Lov, A.; BEOLCHINI, F.; MARTULLO, S.; CAPANNA, E. (1994 b): Territorial behaviour of Talpa romana in an olive grove habitat in Central Italy. Boll. Zool. 61, 207-211. Territorial behaviour of Talpa romana 201 MACDONALD, D. W. (1983): The ecology of carnivore social behaviour. Nature (London) 301, 379-384. OPPERMANN, J. (1968). Die Nahrung des Maulwurfs (Talpa europaea L. 1758) in unterschiedlichen Le- bensräumen. Pedobiologia 8, 59-74. OSTFELD, R. S. (1990): The ecology of territoriality in small mammals. Trends Ecol. Evol 5, 411-415. NEvo, E. (1979). Adaptive convergence and diverence of subterranean mammals. Ann. Rev. Ecol. Syst. 10, 269-368. SAS INSTITUTE, Inc. (1985): SAS user’s guide: statistics, version 6.04 edition. Cary, North Carolina: SAS Institute, Inc. SCHWAGMEYER, P. L. (1988): Scramble-competition polygyny in an asocial mammal: male mobility and mating success. Am. Nat. 131, 885-892. SKOCZEN, S. (1966): Stomach contents of the mole, Talpa europaea Linnaeus, 1758 from Southern Po- land. Acta Theriol. 11, 551-575. SOKAL, R. R.; RoHLF, F. J. (1981): Biometry. 2nd Ed.. New York: Freeman an Co. STATISTICA FOR WınDows (1993): Version 4.3 for Personal Computer. Stat. Soft. Inc. Tew, T. E.; MACDONALD, D. W. (1994): Dynamics of space use and male vigour amongst wood mice, Apodemus sylvaticus, in the cereal ecosystem. Behav. Ecol. Sociobiol. 34, 337-345. TRIVERS, R. L. (1972): Parental investment and sexual selection. In: Sexual selection and the descent of man. Ed. by B. CampgELr. Chicago: Aldine Press. Pp. 136-179. Authors’ address: FRANCESCA BEOLCHINI, EUGENIO DUPRE and AnnA Loy, Dipartimento di Biologia Animale e dell’Uomo, Unversita di Roma, via Borelli 50, I-00161 Roma, Italy. Z. Säugetierkunde 61 (1996) 202-214 “FÜR © 1996 Gustav Fischer, Jena SÄUG ETI ERKU N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Wie nutzen Wasserfledermäuse, Myotis daubentonii (Kuhl, 1817), ihre Tagesquartiere? Von I. RIEGER Dachsen, Schweiz Eingang des Ms. 24. 07. 1995 Annahme des Ms. 13. 03. 1996 Abstract How do Daubenton’s bats, Myotis daubentonii, use their day roosts? Daubenton’s bats, Myotis daubentonii, in the area of the Rhinefalls (Switzerland) use up to seven day roosts. On an average, they change roosts every second day. Some individuals visited two or more day roosts within 24 hours, others used exclusively the same day roost for more than two weeks. Males chan- ge roosts twice as often as females. A newly used day roost is some 600 m away from the old roost. All day roosts in a forest are connected with one another by individual roost changes. Such day roost net- works are restricted to one forest. We never found roost networks with day roosts in different forests. Following rainy days, Daubenton’s bats have a higher tendency to change roosts than after dry days. Our knowledge is too small for a functional description of day roosts of Daubenton’s bats. We dis- tinguish the following day roost types: Type A day-roosts have high animal numbers without much chan- ge from day to day. These day roosts are favourite, based on the number of changes at these roosts. Their entrances are two to five m above the ground. Bat groups in Type A day-roosts vary during the hunting season. Type B day-roosts have entrances ten m or more above the ground. 30 to 100 individuals stay to- gether in type Bl roosts. Group size changes daily. Only single animals or a few animals stay in Type B2 day-roosts, but they use these roosts exclusively for two or more weeks. Type C day-roosts have entrances two to five m above the ground. The number of bats staying together is small, usually less than 15, quite often less than 6 animals. They stay longer in a Type C roost than in a Type A roost. Einleitung Seit mehreren Jahren werden Verhalten und Ökologie von Wasserfledermäusen, Myotis daubentonii, in der Region Rheinfall studiert (RıEGER et al. 1990, 1992; RiEGER und WALZ- THÖNY 1993; RIEGER und ALDER 1993, 1994; RıEGER 1994, 1996). Diese Fledermäuse, die sich tagsüber in hohlen Bäumen im Wald aufhalten, sind schwierig zu beobachten (Cer- VENY und BÜRGER 1989; FUHRMANN und GODMANN 1991, 1994; GODMANN 1994; HELMER 1983; LABEs et al. 1989; SLUITER und VAN HEERDT 1966; STUTZ und HAFFNER 1985; STRAT- MANN 1978; WoLz 1986). Es scheint, als würden Begriffe wie Zwischenquartier, Wochen- stubenquartier, Männchenquartier die Quartiere von den waldlebenden Fledermäusen nicht ebenso gut beschreiben wie bei synanthropen Fledermaus-Arten. In der Region Rheinfall halten sich Wasserfledermäuse tagsüber hauptsächlich in vitalen, hohlen Rotbu- chen und Eichen auf. Die meisten Rotbuchen mit Wasserfledermaus-Quartieren sind forstwirtschaftlich wenig wertvoll, sie eignen sich höchstens als Brennholz. Quartierbäu- me haben auf Brusthöhe einen Umfang von 30 bis 300 cm, die Quartieröffnungen mün- den zwischen 50 cm und über 15 m, die meisten zwischen 1 und 5 m über Boden (einige Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 203 Quartierbäume haben Öffnungen, die so hoch liegen, daß man sie vom Boden aus nicht sehen kann). Die höchsten Punkte der Baumhöhlen sind 1 bis 170 cm über der Quartier- öffnung. Mehr als 40% aller Quartierbäume sind höchstens 30 m vom Waldrand entfernt. Sie stehen bevorzugt an Süd- bis Westhängen. Vermutlich sind geeignete Quartiere ein li- mitierender Faktor für die Verbreitung der Wasserfledermäuse, auch wenn in der Region Rheinfall der Anteil hohler Buchen am Gesamt-Baumbestand relativ hoch ist (RIEGER 1996). Von einer Tierart, die ein so großes Verbreitungsgebiet hat wie die Wasserfledermaus (BopGAanowıcz 1994), darf man annehmen, daß sie ganz unterschiedlichen Umweltbedin- gungen begegnet und ihr Verhalten diesen Bedingungen optimal anpaßt. Verhaltenswei- sen und ökologische Besonderheiten einer Teilpopulation müssen nicht zwingend in andern Teilpopulationen auftreten. Die vorliegende Übersicht faßt erste Resultate zusam- men, die zeigen, wie Wasserfledermäuse in der Region Rheinfall ihre Baumhöhlen-Quar- tiere nutzen. Material und Methode Untersuchungsgebiet: Das Haupt-Untersuchungsgebiet ist eine rund 64km” große Region um den Rheinfall (8°40' östliche Länge, 47°40' nördliche Breite; Kilometer-Koordinaten des Bundesamts für Landestopographie 687/278 bis 695/286). Der Rhein-Wasserspiegel liegt oberhalb des Kraftwerks Schaffhausen (KWS) 392 müM, zwischen KWS und Rheinfall 385 müM, unterhalb des Rheinfalls 359 müM. Dies ist der tiefste Bereich im Untersuchungsgebiet. Die höchsten Punkte im Untersuchungs- gebiet liegen 481 müM (Ryhirt), 551 müM (Engiwald), 567 müM (Buechbüel), 570 müM (Cholfirst). Rund ein Drittel dieses Gebiets ist bewaldet, knapp 4% sind Gewässer, und ein Fünftel ist Siedlungs- und Industriegebiet. Die Region ist teils dicht besiedelt (Schaffhausen, Neuhausen), teils Landwirt- schaftsgebiet (Zürcher Weinland). Wald: Im Untersuchungsgebiet liegen große Waldparzellen auf den Hügelkuppen. Sie haben keinen direkten Rhein-Anstoß. Einzig im Osten und im Südwesten fließt der Rhein entlang von größeren Waldparzellen (Schaaren-Wald vis-a-vis von Büsingen; Schwaben-Wald auf der deutschen Rheinseite bei Dachsen und Rheinau). Die Waldgesellschaften gehören vorwiegend zum Typ Waldmeister-Buchen- wald (SCHMIDER et al. 1993). Rotbuchen, Fagus sylvatica, und Eichen, Quercus robur, sind die häufig- sten Baumarten. Die Buche ist mit einem Drittel am gesamten Holzvorrat die häufigste Holzart. Eichen werden seltener, ihr Anteil sank von einem Viertel in den 20er Jahren auf ein Sechstel in den 80er Jahren. Fichten, Picea abies, steigerten ihren Anteil von rund einem Fünftel auf rund ein Viertel. Der Anteil an Weißtannen, Abies alba, sank von einem Fünftel auf ein Achtel. In den letzten 6 Jahr- zehnten verdoppelte sich das Holzvolumen im Cholfirst-Wald. Die meisten Quartier-Untersuchungen wurden in den Wäldern „Ryhirt“ (2 km?), „Schaffhauser Wald“ (0,5 km?) und „Cholfirst“ (11 km?) ge- macht. Tierfang: Wasserfledermäuse wurden mit dem Japannetz (etwa 1,5 cm Maschenweite, 0,2 mm Fa- dendurchmesser) gefangen. Fangorte waren Flugstraßen und die Ausflugbahn vor Quartieröffnungen. Vor Quartieröffnungen wurden die Tiere auch mit einem Plastikschlauch, einem Plastiktrichter oder der TUTTLE-Trap gefangen. Telemetrie: Es wurden Radiotelemetriesender mit Thermistor von Biotrack, UK, eingesetzt (0,85 g schwer, 9 mm breit, 13 mm lang und 5 mm hoch, mit einer 15 cm langen Stahldrahtantenne). Die Sen- der wurden mit einem Cellophan-Halsband und einem Tropfen eines giftfreien Kontaktklebers im Nak- ken der Wasserfledermaus befestigt. Zum Empfang dienten 2- bis 4teilige Richtantennen auf dem Autodach, auf dem Velo, dem Moped und in der Hand. Bei idealen Bedingungen, d.h. trockener Luft und Sichtverbindung, hatten die Sender und Antennen eine Reichweite von 700 m, meist betrug die Reichweite aber nur 300 bis 500 m. Triangulieren erübrigte sich zumeist, da man mit dem Telemetrie- Empfänger so nahe beim Tier war, daß man aus der Richtung, der Signalintensität (dem geschätzten Abstand), der Geländestruktur und den Habitaten recht genau auf den Aufenthaltsort des Tiers schlie- ßen konnte. Zur Kennzeichnung wurden Armklammern des Naturhistorischen Museums Genf verwendet. In der Regel wurden nur Tiere in einem Gebiet markiert, in dem auch öfter gefangen wurde, d.h. es wur- den Tiere mit einer gewissen Wiederfang-Wahrscheinlichkeit gekennzeichnet. 204 I. RIEGER Für Quartier-Ausflugbeobachtungen wurden infrarot-empfindliche Bildwandler eingesetzt (je nach spektraler Empfindlichkeit Wärmebildgerät = Optronisches Beobachtungsgerät OBG oder Infrarot- Fernsehkamera IR-TV). Das OBG nutzt Wellenlängen zwischen 8 und 12 um und wandelt die Wärme- abstrahlung eines Körpers in ein Bild um. Die IR-TV-Kamera nutzt Wellenlängen zwischen 4 und 10 nm. Die Infrarot-Lichtquelle ist ein Scheinwerfer mit vorgesetztem IR-Filter. Neben der Registrierung mit technischen Ausrüstungen wurde der Quartierausflug optisch beob- achtet. Dabei saß der Beobachter beim Quartierbaum (nicht in der Ausflugbahn der Fledermäuse) und beleuchtete die Ausflugbahn, nicht aber die Quartieröffnung, mit einer Taschenlampe. Zudem war ein Bat Detektor mit maximaler Empfindlichkeit bei rund 40 kHz eingeschaltet. Definitionen: Das Tagesquartier (Tab. 1) ist ein Ort, wo sich die Wasserfledermäuse tagsüber auf- halten und ruhen. Als Quartierwechsel gilt, wenn eine Wasserfledermaus nach einer Aktivitätsphase außerhalb des Tagesquartiers sich in ein anderes Tagesquartier zurückzieht. Ergebnisse und Diskussion Tagesquartiere von radiomarkierten Wasserfledermäusen Es wurden 25 Wasserfledermäuse im Mittel während je 12 (1 bis 30) Tagen telemetriert. 24 Tiere hielten sich zeitweise in Tagesquartieren auf. Eine Wasserfledermaus benutzte ein bis sieben verschiedene Tagesquartiere. Aufenthaltsdauer im Tagesquartier Wasserfledermäuse benutzen ohne Unterbrechung das gleiche Tagesquartier von null bis 23 Tagen. Null Tage im gleichen Tagesquartier bedeutet, daß sich eine Wasserfledermaus innerhalb von 24 Stunden in mindestens zwei verschiedenen Tagesquartieren aufhielt. Dies trifft zu vor allem im April, aber auch im Juli und Oktober waren radıomarkierte Wasserfledermäuse (2? und 34) innerhalb einer Nacht in mehr als einem Baumhöhlen- Quartier. Im Durchschnitt sind Wasserfledermäuse während zwei aufeinanderfolgenden Tagen im gleichen Quartier und wechseln dann in ein anderes. Die durchschnittliche individuelle Aufenthaltsdauer in einem Tagesquartier variiert von Individuum zu Individuum. Weibchen bleiben durchschnittlich doppelt so lange in einem Tagesquartier wie Männchen (Tab. 2). Distanzen zu neu genutzten Tagesquartieren Es wurden rund 100 Wechsel einer radiomarkierten Wasserfledermaus von einem Tages- quartier in ein anderes registriert (Abb. 2). Während der gesamten Telemetriedauer wech- selten die Tiere durchschnittlich über Distanzen von etwa 600 m; der größte Wechsel ging über 4km, der kleinste über 15 m. Alle Wechsel über Distanzen größer als 1900 m fan- den im April statt. Die Quartierwechsel über sehr große Distanzen waren nicht mit dem Fang und Ausrüsten des Tiers mit einem Radiosender korreliert. Große Distanzen bei Quartierwechsel fanden in den ersten 24 Stunden statt, aber auch Tage nachdem eine Wasserfledermaus radiomarkiert wurde. Höhe der Quartieröffnung über dem Boden der neu genutzten Tagesquartiere Bei den meisten Quartierwechseln beziehen Wasserfledermäuse Quartiere mit Öffnungen bis 5m über Boden. Weniger häufig wechseln die Tiere in Quartiere mit Öffnungen zwi- schen 5 und 10 m über Boden. Noch seltener sind Wechsel in Quartiere mit Öffnungen, die mehr als 10 m über Boden liegen. Es gibt keine bevorzugten Höhen von Quartieröff- nungen zu verschiedenen Zeiten der Telemetriephase (Tab. 3). Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 205 Tabelle 1. Tagesquartiere von Wasserfledermäusen in der Region Rheinfall (siehe auch RıEGER 1996). Quartier- Baum Quartieröffnung Standort Code Rot- Eiche Hage- an- Stamm- Spalt ausge- an- Höhe Wald- Höhe buche buche derer umfang faulter dere über rand- über auf Ast Boden abstand Meer Brusthöhe [cm] [m] [müM] Schaffhauser-Wald 08701 80 680 5 459 QS702 105 162 5 472 QS703 47 z 146 5 455 QS704 76 _ 340 5 470 QS705 78 160 5 472 QS706 97 390 10 448 QS707 70 _ 473 1 458 QS709 a 300 17 468 QS710 6 WE 87 10 469 08711 | 5 460 08712 u | 59 100 470 QS713 155 en 490 1 462 QS716 202 200 460 08717 u 543 14 470 QS718 ze 350 10 465 08719 a 345 15 470 QS720 98 180 465 08721 62 s= 460 200 460 QS722 82 80 458 QS723 zu 30 10 470 QS724 350 ze 20 470 QS725 51 790 8 470 QS726 Dan 237 470 QS727 282 2 440 QS728 ı u 528 22 465 206 I. RIEGER Tabeile 1. (Fortsetzung) Quartier- Baum Quartieröffnung Standort Code Rot- Eiche Hage- an- Stamm- Spalt ausge- an- Höhe Wald- Höhe buche buche derer umfang faulter dere über rand- über auf Ast Boden abstand Meer Brusthöhe [cm] [m] [müM] [em] Ryhirt-Wald Cholfirst-Wald Attraktive Tagesquartiere Aufgrund der Anzahl radiomarkierter Wasserfledermäuse ist der Ryhirt-Wald das teleme- trisch bestuntersuchte Wasserfledermaus-Waldhabitat. In diesem Wald sind rund 50 Baumhöhlen-Quartiere von Wasserfledermäusen bekannt. Es wurden hier auch mehr als 80 Quartierwechsel von radiomarkierten Tieren registriert. Nur ein halbes Dutzend Quartiere erhielten mehr als 4% dieser Quartierwechsel, Aufgrund der Quartierwechsel ist das Quartier QS709 das attraktivste (mehr als 15% aller Quartierwechsel), gefolgt von QS719 (11%) und QS713 (10%). Auf die übrigen 28 Quartiere verteilten sich weniger als zwei Drittel aller Quartierwechsel. Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 207 gg Don 3 | 0 ——4 | 0% ——5 | 0% ——6 | 0% son ww 0o9 jo 2 ar aa Fin Abb. 1. Histogramm der Aufenthaltsdauer [in ganzen Tagen] von radiomarkierten Wasserfledermäu- sen im gleichen Tagesquartier. 09% —— 1|01% ——2 |09%”—— 3 | 0% ——4 | 0% ——5 | 0% ——6 | 0% 99 600.35 15.00 3922.00 718.63 Distanzen [m] bei Quartierwechsel 3 |0%——4 | 0% ——5 |) ——6 | 0% Abb. 2. Histogramm der Distanzen [in Metern] bei Quartierwechseln von radiomarkierten Wasserfle- dermäusen. Individuelle Tagesquartier-Netze Durch individuelle Wechsel von einem Quartier in ein anderes sind alle Baumhöhlen- Quartiere im Ryhirt-Wald miteinander vernetzt. Abgesehen von wenigen Ausnahmen (Tab. 4), ist kein Tagesquartier-Netz mit Quartieren in verschiedenen Wäldern bekannt. Wasserfledermäuse, die Tagesquartiere in hohlen Bäumen im kleinen Schaffhauser-Wald (nördlich von Büsingen) nutzten, hielten sich nie in hohlen Bäumen im nahen Ryhirt- Wald auf. 208 I. RIEGER Tabelle 2. Aufenthaltsdauer [d], während der eine radiomarkierte Wasserfledermaus das gleiche Tages- quartier benutzt Individuum Tage im gleichen Tagesquartier Sex max SD 900508 W 1.00 1 1 2 920 702 W 1.00 1 jl 2 920 705 W 3.15 Il 8 3.10 7a - 920 766 W 2.10 1 6 1.60 2.29 10 921 006 W 133 0 3 1.03 0.89 6 930 404 W 5.00 1 10 4.58 14.0 5 930407 W 9.50 il 18 12.0 NER 2 940 404 W 1.86 0 6 27, 3.84 7 940 406 W 2.50 0 6 2.26 4.25 6 940 505 W 1.10 0 6 179 2.89 10 940 508 W 1.50 1 2 0.71 025 2 940 802 W 25 2 23 14.8 110 2 940 803 W 9.25 1 14 32977 26.7 4 mean SZ 920405 M 2.67 2 3 0.58 0.22 3 920 406 M 0.75 0 1 0.50 0.19 4 920721 M 133 1 2 0.58 0.22 3 930403 M 3.00 1 8 2.68 6.00 6 930405 M 0.93 0 3 1.07 1207 14 934408 M 2.00 1 3 1.41 1.00 2 mean 1.78 Tabelle 3. Höhe der Quartieröffnung über Boden [cm] eines neu benutzten Tagesquartiers von radio- markierten Wasserfledermäusen individuelle mittlere kleinste größte Standard- Anzahl Telemetrie-Tage Höhe Höhe Höhe (*) abweichung Meßwerten [d] [cm] [cm] [cm] SD (*) bei Baumhöhlen mit nicht sichtbaren Quartieröffnungen wurden Werte von 1500 bis 1552 cm einge- setzt. Quartierwechsel und Regen Es regnete an 45% der Beobachtungstage (n = 69), an denen eine radiomarkierte Wasser- fledermaus ihr Tagesquartier wechselte, teilweise ergiebig mit mehr als 100 mm Nieder- schlag. Es regnete an 38% der Beobachtungstage (n = 97), an denen eine radiomarkierte Wasserfledermaus das gleiche Tagesquartier benutzte wie am Vortag. Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 209 Tabelle 4. Bemerkungen zu Quartierwechsel über große Distanzen Tier-Nr. Telemetrietag neues Quartier Sex des Q-Wechsels 920404 2 OP201, 10 : 10 m große Region bestimmt. Telemetriesender fällt aus. 920406 8 QS050, Dehnungsfuge in Brücke, eine zweite Wasserfleder- maus im Quartier. 920 766 9 QS051, Dach Gaswerk. Vermutlich allein in diesem Quartier. 940 402% OQOB403, abgestorbene Eiche. 9405087 QOB801, Föhre. Tier offensichtlich gestreßt (Radiosender schlecht montiert). Tags darauf in Jagdpausen-Quartier auf Weide am Rheinufer von Hand gefangen (Tier anschließend von Sender befreit, aufgefüttert und wieder freigelassen). 940 505 2 3/4/5/5/11 Wechselt mehrmals zwischen Quartieren im Ryhirt-Wald (QS705, QS719, QS745, QS713) und nahe am Rhein (QS052, QS101) hin und her. Zusammensetzung der Gruppen in Tagesquartieren Aufgrund der Zusammensetzung der Gruppen, kann man von „großen“ (mit 15 oder mehr Tieren) und von „kleinen“ Quartieren (mit weniger als 15 Tieren) sprechen. In be- stimmten Wäldern (z.B. Schaffhauser-Wald) scheinen vorwiegend 33 zu leben, in andern (z.B. Ryhirt) überwiegen die 9. Die Gruppenzusammensetzungen scheinen sich im Lauf der Jagdsaison zu verändern (z.B. QS713). Gruppengröße und Aufenthaltsdauer im Tagesquartier Wasserfledermaus-Tagesquartiere mit hohen Beständen (aufgrund von Ausflugszahlen und Fangerfolgen) sind nicht identisch mit Tagesquartieren, in denen sich ein Wasserfle- dermaus-Individuum über längere Zeit aufhält (Daten von radiomarkierten Tieren). In Quartieren mit kleinen Beständen bleibt ein Individuum oft während einer langen Zeit (in einem Fall mehr als 3 Wochen). Nicht nur Wasserfledermäuse, auch andere einheimische Fledermäuse mit Baumhöh- len-Tagesquartieren, wechseln mehr oder weniger oft ihr Tagesquartier (CERVENY und BÜRGER 1989; CHAPUISAT et al. 1988; von HELVERSEN 1989; Henkel et al. 1989; WoLz 1986). Nach der Einteilung von Lewis (1995) gehören Wasserfledermäuse zu den quar- tierlabilen Fledermaus-Arten, d.h. sie wechseln das Tagesquartier meist häufiger als ein- mal innerhalb von 10 Tagen. Lewis (1995) nennt folgende Gründe, weshalb Fledermäuse ein Quartier verlassen: (1) Energiehaushalt: Eine Fledermaus, deren Tagesquartier nahe beim Jagdgebiet liegt, spart Energie, weil der Flug in das Jagdgebiet kürzer ist, als wenn sie tagsüber entfernt vom Jagdgebiet ein Quartier nutzt. (2) Störungen am Quartier: z.B. Fangaktionen an Tagesquartieren, ausgeführt zum einen von Raubfeinden, zum andern von Forschern mit Telemetrieabsichten. Meist wechselten die Fledermäuse nach solchen Aktionen das Quartier. 210 I. RIEGER (3) Änderungen im Mikroklima. (4) Änderungen in der Quartiergeometrie. (5) Parasiten im Quartier. Die Wasserfledermäuse in der Region Rheinfall wechselten ihre Tagesquartiere z.T. aus den von Lewis (1995) aufgezählten Gründen: Energiehaushalt: Oft benutzten Wasserfledermäuse nach Quartierwechseln über „große“ Distanzen Tagesquartiere nahe beim Jagdhabitat Rhein. Verschiedene Wasserfle- dermäuse wechselten ihr Aufenthaltsgebiet großräumig, benutzten aber keine Tagesquar- tiere (72 940 402), sondern hielten sich tagsüber in einem „Jagdpausen-Quartier“ auf, oder aber ihr(e) Tagesquartier(e) wurde nicht gefunden (3 920 702). Störungen am Quartier/„Montageschock“: Wenn eine Wasserfledermaus an einem Ta- gesquartier abgefangen und mit einem Radiosender ausgerüstet wird und das Tier an- schließend das Quartier wechselt, dann sprechen wir von einem „Montageschock“. Am ersten Telemetrietag wurden viele Wechsel in Quartiere mit Öffnungen über 10 m über Boden registriert. Dennoch liegt die durchschnittliche Höhe von Quartieröffnungen nach einem Quartierwechsel im Lauf der Telemetriedauer immer etwa bei 9m. Es scheint, als wäre die Konkurrenz um hohle Bäume groß. Fledermäuse sind vermut- lich allen Konkurrenten (Specht, Star, Siebenschläfer, Hornisse) unterlegen. Für Fleder- mäuse ist die Konkurrenz in Baumhöhlen mit großen (>50 cm) Aufrißöffnungen am kleinsten, denn die meisten Säugetiere und Vögel nutzen den Dom der Baumhöhle nicht; bei großen Öffnungen stört ein allfällig vorhandener Konkurrent im unteren Höhlenteil die ein- und ausfliegenden Fledermäuse nicht. Dies dürfte mit ein Grund dafür sein, daß Wasserfledermäuse im Untersuchungsgebiet ihre Tagesquartiere in den meisten Fällen in Buchen mit Spaltöffnungen haben (RıEGER 1996). Mikroklima: Im Optronischen Beobachtungsgerät erscheinen Buchen heller abgebil- det als andere Waldbäume, d.h. Buchen sind wärmer als andere Bäume. Dieser Unter- schied besteht nicht nur in der Abenddämmerung, sondern während der ganzen Nacht. Der Wärmeunterschied zwischen verschiedenen Waldbäumen ist daher nicht auf unter- schiedliche Abstrahlungswärme von tagsüber akkumulierter Wärme durch Sonnenbe- strahlung zurückzuführen, sondern scheint die Folge von Stoffwechselvorgängen (z.B. Fäulnisprozesse) im Baum zu sein. Die meisten Tagesquartiere liegen im Stammbereich, meist 2 bis 5m über Boden. In einem vertikalen Temperaturgradienten vom Waldboden zu den Baumwipfeln ist dieser Bereich der wärmste (MiTscHERLICH 1988). Wasserfledermaus-Tagesquartiere liegen somit in relativ warmen Zonen und in warmen Bäumen. Nach ALper (1994) schwankte die Temperatur in einer Wasserfledermaus-Baumhöh- le (mittelgroße Buche) im Juni zwischen 4°C über dem Minimum der Außentempera- tur und 2°C unter dem Maximum der Außentemperatur (QS709, Umfang 70cm). In einer andern Höhle (kleinere Buche) schwankte sie zwischen 1 bis 2°C über dem Mini- mum der Außentemperatur und 2°C unter dem Maximum (QS710, Umfang 46 cm). Lange Meßreihen von Parametern des Baumhöhlen-Mikroklimas sind bis jetzt nicht zu- gänglich. | Man darf annehmen, daß sich das Mikroklima in wenigstens einigen Baumhöhlen- Quartieren verändert, wenn es regnet, denn bei einzelnen Bäumen sieht man nach dem Regen eine viel deutlichere Feuchtspur unterhalb der Quartieröffnung als während einer Trockenperiode. Zudem gibt es Hinweise, daß Wasserfledermäuse dazu neigen, nach Re- gentagen eher das Quartier zu wechseln als darin zu bleiben. Quartiergeometrie: Messungen an Baumhöhlen-Quartieren in Jahres- oder Zweijah- resabständen zeigen, daß sich die Höhlen-Dimensionen während eines Jahres in der Grö- ßenordnung von Zentimetern bis Dezimetern verändern können. Fäulnisprozesse lassen zum einen den Höhlenraum größer werden. Auf der andern Seite können aber herabfal- Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 214 lende Holzteile sich weiter unten in der Höhle verkeilen und das den Fledermäusen zu- gängliche Volumen verkleinern. Aus verschiedenen Gründen wird bei einigen Quartieren die Öffnung nach und nach kleiner, so daß Wasserfledermäuse nicht mehr hindurchschlüpfen können: Spaltförmige Öffnungen bei Rotbuchen verändern sich. In einem Fall (OT001) über- wallte der Baum die Öffnung so stark, daß Wasserfledermäuse nicht mehr in das Baum- höhlen-Quartier hineinschlüpfen können (die kritische Weite liegt bei 10 mm). Bei einem andern Quartier überwächst Efeu die Öffnung, so daß die Fledermäuse das Quartier nicht mehr benutzen. Fledermaus-Kot verstopft die Quartieröffnung. Nach Entfernung des Kotes sind oft schon am nächsten Tag wieder Fledermäuse im Quartier. Einige %2 nutzten ununterbro- chen und sehr lange (10 bis 23 Tage) dasselbe Tagesquartier. Auch DIETZ (unveröff.) be- richtet von Individuen, die während mehr als zweieinhalb Wochen das gleichen Quartier nutzten. Im Sinne von Lewis (1995) verhalten sich diese Individuen quartiertreu. Die herkömmlichen Fledermaus-Quartierbezeichnungen wie „Wochenstube, Zwi- schenquartier, Männchenquartier, Ausweichquartier etc.“ befriedigen nicht, wenn ver- sucht wird, Quartiere von Wasserfledermäusen zu beschreiben. Das Quartier QS713 scheint zeitweise eine Wochenstube zu sein (z.B. 9. August 1994), zu andern Zeiten ist es ein reines $g-Quartier (14. Oktober 1994) oder ein „Haremsquartier“ (29. April 1993: da- mals wurden 409% gefangen gleichzeitig hielt sich das radiomarkierte $ 930403 im Quar- tier auf, ließ sich aber nicht fangen). QB203 wäre aufgrund der Fangdaten ein SJ- Quartier, bei einer Ausflugskontrolle am 30. Mai 1992 flogen knapp hundert Tiere aus. In diesem Fall ist die Bezeichnung SS-Quartier kaum angezeigt. Wasserfledermaus-Tagesquartiere lassen sich erst mit funktionellen Begriffen beschrei- ben, wenn Beobachtungen während mindestens einer Jagdsaison an mindestens je einem der unten beschriebenen Quartiertypen durchgeführt wurden. Aufgrund der bisher vorlie- genden Resultate können Wasserfledermaus-Tagesquartiere in drei bis vier Typen einge- teilt werden. Wasserfledermäuse sind ein, zwei Tage mit vielen Artgenossen in einem Typ-A-Quar- tier zusammen, dann längere Zeit in einem Typ-C-Quartier mehr oder weniger allein. Quartiere vom Typ A scheinen sozial äußerst wichtige Quartiere zu sein: Sie enthalten große Bestände, die möglicherweise von Tag zu Tag anders zusammengesetzt sind. Einzig an einem Typ-A-Quartier (QS713) wurden Einflüge kurz nach der normalen Ausflugzeit (im Juni, noch keine Säuglinge zu diesem Zeitpunkt) beobachtet. Typ-B-Quartiere liegen in hohen Bäumen. Die Quartieröffnungen befinden sich auf 10 oder mehr Meter über Boden. Diese Quartiere haben zwei verschiedene Nutzungsmu- ster. Im ersten Fall (Typ Bl) sind 30 bis 100 Tiere zusammen. Die Bestände wechseln von Tag zu Tag. Beispielsweise flogen am 30. Mai 1992 aus dem Quartier QB203 96 Tiere aus, am darauffolgenden Morgen flogen nur 49 Tiere ins Quartier ein. Im zweiten Fall (Typ B2) scheint ein einzelnes Tier während langer Zeit ein solches „hohes“ Quartier zu nutzen. Möglicherweise spielt bei der Nutzung der Typ-B-Quartiere das Mikroklima, vor al- lem die Temperatur im Quartier, eine wichtige Rolle. An sich ist die Lage des Quartiers für den Wärmehaushalt ungünstig: 10 und mehr Meter über dem Waldboden ist es kühler als 2 bis 5m über dem Boden. Eichen, ein häufiger Typ-B-Quartierbaum, sind nach Beob- achtungen mit dem Optronischen Beobachtungsgerät weniger warm als Buchen. Dieses vermutete Wärmedefizit gleichen die Wasserfledermäuse aus, indem sie große Gruppen bilden (Typ Bl). Wie aber einzelne Tiere in einem Typ-B2-Quartier den Wärmehaushalt optimieren, ist nicht bekannt (das radiomarkierte ? 920 766, mit einem Telemetriesender mit Thermistor, übermittelte Nackenfelltemperaturen aus Typ-B2-Quartieren (QS711, QS051), die er- staunlich hoch waren). 212 I. RIEGER Wasserfledermäuse waren mit Großen Abendseglern (Nyctalus noctula) in Quartieren vom Typ A (z.B. 0S713), Typ Bl (OK001) und Typ € (z.B. QS704) zusammen. Im Typ- A-Quartier OS 709 fanden sich einmal, neben Wasserfledermäusen, auch zwei Mausohr- 92 (Myotis myotis). Im Oktober war eine Langohrfledermaus (Plecotus sp.) im Typ-C- Quartier QS712. Wasserfledermäuse in der Region Rheinfall wechseln von Quartier zu Quartier über Distanzen von durchschnittlich 600 m. Die Quartier-Netze der radiomarkierten Wasserfledermäuse in der Region Rheinfall sind vermutlich gleichzusetzen mit den „Kolonien“, die GEIGER (unveröff.) mit Hilfe beringter Wasserfledermäuse in Mittel- franken fand. Quartierwechsel über Distanzen von mehr als 1 km waren selten. Beim jetzigen Stand des Wissens lassen sich die großräumigen Wechsel wie folgt interpretieren: Die 2? 920404 und 940 404 reagierten mit „Montageschock“ (Wechsel am ersten Tele- metrietag). Das % 940 508 verhielt sich atypisch wegen des schlecht montierten Senders. Die großräumigen Wechsel von 3 920406 und 9 920766 lassen sich als Wechsel in Quartiere nahe beim Gewässerjagdhabitat kennzeichnen. $ 940505 hielt sich mehrere Ta- ge in Quartieren am Rhein (OS101, QS052) auf, dann wieder in Quartieren im Ryhirt- Wald. Diese Wechsel machte es mehrere Male. Die meisten Quartierwechsel machten Wasserfledermäuse zwischen den Tagesquartie- ren des gleichen Waldes. Keine Wasserfledermaus wechselte von Quartieren in Bäumen im Schaffhauser-Wald in Quartiere im benachbarten Ryhirt-Wald, obwohl Tiere aus Ry- hirt- und Schaffhauser- Wald-Quartieren auf gleichen Flußabschnitten jagten. Danksagung Viele Daten über Verhalten und Ökologie von Wasserfledermäusen trugen die Kolleginnen und Kolle- gen, die an Projekten der Fledermaus-Gruppe Rheinfall mitarbeiten, bei. Mein Dank gilt HAnsUELI AL- DER, FABIO BONTADINA, CLAUDIA und ROLAND BÖSIGER, JÜRG CAMBENSY, SANDRA GLOOR, MARTIN GRAF, ERIKA und JÜRG MERKI, ANDREAS MÜLLER, THOMAS MÜLLER, RUEDI SCHNEIDER, DORIS WALZTHÖNY; Mitglieder des Vereins für Vogel- und Naturschutz Schaffhausen TURDUS und des Naturschutzvereins Dachsen NVD, und Teilnehmer an einem Kurs der Lehrerfortbildung des Kantons Schaffhausen. NVD und die Naturforschende Gesellschaft Schaffhausen unterstützen mit Geldbeträgen die Projekte der FMGR. Pnır W. RıcHARDSson überarbeitete die englische Zusammenfassung. Zusammenfassung Wasserfledermäuse in der Region Rheinfall nutzen bis zu sieben verschiedene Tagesquartiere. Durch- schnittlich wechseln sie alle zwei Tage das Quartier. Einzelne Individuen findet man innerhalb von 24 Stunden in zwei oder mehr Tagesquartieren. Andere Tiere nutzen während mehr als zwei Wochen ununterbrochen dasselbe Tagesquartier, SS wechseln das Tagesquartier doppelt so häufig wie ?9. Das neue Quartier ist durchschnittlich 600 Meter vom alten entfernt. Die Quartiere in einem Wald sind durch individuelle Quartierwechsel untereinander vernetzt. Solche Tagesquartier-Netze sind immer auf einen Wald beschränkt, es gibt keine Netze mit Quartieren in verschiedenen Wäldern. Nach Regenta- gen neigen Wasserfledermäuse eher zu einem Quartierwechsel als nach trockenen Tagen. Für eine funktionelle Beschreibung der Wasserfledermaus-Tagesquartiere ist die Datenbasis zu klein. Folgende Quartiertypen werden unterschieden: Typ-A-Quartiere haben große Tierbestände, die von Tag zu Tag wenig variieren. Die Quartiere sind attraktiv, gemessen am Anteil von Quartierwech- seln, die zu diesen Quartieren führten. Die Quartieröffnungen liegen im Stammbereich zwischen zwei und fünf Metern über Boden. Die Tiergruppen in Typ-A-Quartieren sind je nach Saison verschieden zu- sammengesetzt. Typ-B-Quartiere haben ihre Öffnungen über 10 m über Boden. In Typ-B1-Quartieren sind 30 bis 100 Tiere zusammen, die Bestände variieren von Tag zu Tag sehr stark. In Typ-B2-Quartie- Wie nutzen Wasserfledermäuse ihre Tagesquartiere? 213 ren sind nur einzelne Tiere, die oft zwei und mehr Wochen bleiben. Typ-C-Quartiere haben Öffnungen zwei bis fünf Meter über Boden. Es finden sich meist weniger als 15 Tiere, meist sogar weniger als 6 Tiere in solchen Quartieren. Die Tiere bleiben aber meist länger als in Typ-A-Quartieren. Literatur ALDER, H. (1994): Erste Erfahrungen mit dem Data Logger: Ereigniszählung vor Baumhöhlenquartie- ren von Wasserfledermäusen, Myotis daubentoni, bei gleichzeitiger Messung mikroklimatischer Werte. Mitt. natf. Ges. Schaffhausen 39, 117-131. CERVENY, J.; BÜRGER, P. (1989): Bechstein’s Bat, Myotis bechsteini (Kuhl, 1818 in the Sumava Region. In:European Bat Research 1987. Ed. by V. HANAKR, I. Horäcek, and J. GAISLER. Praha: Charles Univ. Press. Pp. 591-598. CHAPUISAT, M.; DELACRETAS, P; REYMOND, A.; RUEDI, M.; ZUCHUAT, O. (1988): Biologie du Murin de Daubenton (Myotis daubentonii) en periode de reproduction. Le Rhinolophe 5, 1-8. FUHRMANN, M.; GODMANN, ©. (1991): Natürliche Quartiere der Waldfledermäuse schützen! Konsequen- zen aus einer Baumhöhlenuntersuchung im Rheingau. Allg. Forstzeitschrift 46, 982-983. FUHRMANN, M.; GODMANN, ©. (1994): Baumhöhlenquartiere vom Braunen Langohr und von der Bech- steinfledermaus: Ergebnisse einer telemetrischen Untersuchung. In: Die Fledermäuse Hessens. Geschichte, Vorkommen, Bestand und Schutz. Verlag Manfred Hennecke. Pp. 181-186. GODMANN, O. (1994): Methoden der Fledermauserfassung und ihre Effektivität bezüglich der verschie- denen Arten und deren Schutz. In: Die Fledermäuse Hessens. Geschichte, Vorkommen, Bestand und Schutz. Verlag Manfred Hennecke. Pp. 103-106. HELMER, W. (1983): Boombewonende Watervleermuizen Myotis daubentonii (Kuhl, 1817) in het rijk van nijmegen. Lutra 26, 1-11. HELVERSEN, O. von (1989): Schutzrelevante Aspekte der Ökologie einheimischer Fledermäuse. Schrif- tenreihe Bayer. Landesamt für Umweltschutz 92, 7-17. HENKEL, F.; BORNKESSEL, G.; TRESS, C.; FISCHER, J. A.; TREss, J. (1989): Beobachtungen an Wochenstu- ben der Nordfledermaus (Eptesicus nilssoni) und über den Witterungseinfluß. In: Populationsökolo- gie von Fledermausarten Ed. by D. HEIDECKE und M. StusgBeE. Martin-Luther-Universität Halle- Wittenberg. Wissenschaftliche Beiträge 20, 177-188. LABES, R.; LABES, S.; SAWALLISCH, D. (1989): Erstnachweis des Kleinen Abendseglers (Nyctalus leisleri) in einer Wasserfledermausgesellschaft (Myotis daubentoni). In: Populationsökologie von Fleder- mausarten. Ed. by D. HEiDEcKE und M. STUBBE. Martin-Luther-Universität Halle-Wittenberg. Wis- senschaftliche Beiträge 20, 111-112. Lewis, S. E. (1995): Roost fidelity of bats: A review. J. Mammalogy 76, 481-496. MITSCHERLICH, G. (1988): Wald, Wachstum und Umwelt — Eine Einführung in die ökologischen Grundlagen des Waldwachstums. Zweiter Band. Waldklima und Wasserhaushalt. 2. überarb. und erw. Aufl. Frankfurt am Main: J. D. Sauerländer’s Verlag. RIEGER, 1.; WALZTHÖNY, D.; ALDER, H. (1990): Wasserfledermäuse, Myotis daubentoni, benutzen Flug- straßen. Mitt. naturf. Ges. Schaffhausen 35, 37-68. RIEGER, 1.; ALDER, H.; WALZTHÖNY, D. (1992): Wasserfledermäuse, Myotis daubentoni, im Jagdhabitat über dem Rhein. Mitt. naturf. Ges. Schaffhausen 37, 1-34. RiEGER, I.; WALZTHÖNY, D. (1993): Fixstreifen-Taxation: Ein Vorschlag für eine neue Schätzmethode von Wasserfledermäusen, Myotis daubentoni, im Jagdgebiet. Z. Säugetierkunde 58, 1-12. RIEGER, I.; ALDER, H. (1993): Weitere Beobachtungen an Wasserfledermäusen, Myotis daubentoni, auf Flugstraßen. Mitt. natf. Ges. Schaffhausen 38, 1-34. RIEGER, I. (1994): Wasserfledermäuse, Myotis daubentoni, in einem großen offenen Unterstand. Mitt. natf. Ges. Schaffhausen 39, 61-91. RIEGER, I. (1996): Quartiere von Wasserfledermäusen, Myotis daubentoni, in hohlen Bäumen. Schweiz. Z. Forstwes. 147, 1-20. RIEGER, 1.; ALDER, H. (1994): Verhalten von Wasserfledermäusen, Myotis daubentoni, am Baumhöhlen- Quartier. Mitt. natf. Ges. Schaffhausen 39, 93-118. SCHMIDER, P., KÜPER, M., TSCHARNER, B., KÄsEr, B. (1993): Die Waldstandorte im Kanton Zürich -— Waldgesellschaften, Waldbau, Naturkunde. Zürich: Verlag der Fachvereine an den schweizerischen Hochschulen und Techniken AG. SLUITER, J. W.; VAN HEERDT, P. F. (1966): Seasonal Habits of the Noctule Bat (Nyctalus noctula). Ar- chives N&erlandaises Zool. 16, 423-439. 214 I. RIEGER STRATMANN, B. (1978): Faunistisch-ökologische Beobachtungen an einer Population von Nyctalus noctu- la im Revier Ecktannen des StFB Waren (Müritz). Nyctalus (N.F.), 1, 2-22. STUTZ, H. P.; HAFFNER, M. (1985): Baumhöhlenbewohnende Fledermausarten der Schweiz. Schweiz. Z. Forstwes. 136, 957-963. Woız, I. (1986): Wochenstuben-Quartierwechsel bei der Bechsteinfledermaus. Z. Säugetierkunde 51, 65-74. Anschr. des Verf.: Dr. InGo RiEGER, Chratzhöfli 4, CH-8447 Dachsen Z. Säugetierkunde 61 (1996) 215-220 ZEITSCHRIFT && INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Fruit as a winter feeding resource in the diet of Stone marten (Martes foina) in east-centrai Italy By M. PANDOLFI, ANNA MARIA DE MARINIS, and I. PETROV Istituto di Scienze Morfologiche, Universita di Urbino, Urbino, Italy and Institute of Forestry, Bulgarian Academy of Sciences, Sofia, Bulgaria Receipt of Ms. 24. 10. 1995 Acceptance of Ms. 05. 02. 1996 Abstract The present study was carried out in Le Cesane (east-central Italy) during the winter of 1992. The envir- onment is a mosaic of woodland (mostly conifer plantations), sown and uncultivated lands. A total of 89 scats of stone marten Martes foina was collected. Standard techniques were applied for analysing the content of scats. Diet composition is given as percentage of occurrence and percentage of estimated vol- ume. The winter diet of the stone marten is almost completely frugivorous. The fruit represents 89% of the total volume in the diet with a percentage of occurrence of 55%. Wild fruits are the most common food items. Berries from juniperus (Juniperus sp.) and from sloe (Prunus spinosa) are the most impor- tant feeding resource (42.2% and 29.7%, respectively). A significant relationship has been found be- tween latitude and winter consumption of fruit in Europe. A trend of increasing consumption of winter fruit has been detected from northern to southern Europe. Introduction Numerous studies have been carried out in Europe on the diet of the stone marten Martes foina Erxleben, 1777 in rural environments (e.g. WAECHTER 1975; DELIBES 1978; AMORES 1980; MARCHESI et al. 1989; RAsmussen and MADsEnN 1985; SKIRNISSON 1986; TES- TER 1986; CHEYLAN and BAYLE 1988; ROMANoWwSKI and LesinskI 1991; RuIz-OLMmo and PA- LAZON 1993; LoDE 1994). The feeding habits of this mustelid in the Italian peninsula are still poorly known (Pozıo and GrADoNnI 1981; BERTOLINO and DorE 1991; SERAFINI and Lovarı 1993). According to these authors the stone marten is omnivorous and feeds on mammals, birds, reptiles, amphibians, insects and fruit. Fruit shows a high incidence in the diet of this predator mainly in Mediterranean and temperate habitats (WAECHTER 1975; SKIRNISSON 1986; BERTOLINO and DorE 1991; Rurz-OLmo and PALAZoN 1993; SERAFINI and Lovarı 1993; LoDE 1994). The geographic variation in the consumption of fruit might be related to geographical factors, among which latitude (which clearly influences the local availability of fruit throughout the year in different habitats) plays a major role. This con- dition of a higher fruit consumption at lower latitudes may correlate closely with the no- tion of the stone marten as an opportunistic predator, whose diet is determined by local availability of feeding resources (CLEVENGER 1994). The aim of this study is twofold: 1) to report data on the winter food of the stone mar- ten in east-central Italy, focussing on an analysis of the consumption of wild and culti- vated fruit; 2) to test the hypothesis that geographic variation in winter consumption of fruit in Europe may be related to latitudinal changes. 216 M. PANDOLFI, ANNA MARIA DE MARINIS, and I. PETROV Material and methods The study was carried out in the Cesane State Forest, located in the Marche region of east-central Italy (43°43’N 0° 13’E), between 150 and 648 m altitude. The climate is mediterranean. Average monthly temperature varies between 4°C in January and 25°C in July; the mean annual rainfall is 900 mm. Pine plantations cover most of the study area. Pinus nigra, P. pinea, P. halepensis, P. maritima are the main species used in reforestation. Cedars (Cedrus atlantica and C. deodara) and cypresses (Cupressus sem- pervirens and C. arizonica) are scattered in the pine plantations. The remaining part of the study area is covered by mixed deciduous formations of Quercus pubescens, Ostrya carpinifolia, Fraxinus ornus, Acer opalus and Corylus avellana. Sown and uncultivated fields are interspersed within the woodland and cover only a few hectars. The occurrence of the pine marten Martes martes has not been recorded in the present study area, thus avoiding the possibility of collecting scats of pine marten by mistake instead of those of stone mar- ten. A total of 89 scats was collected every fortnight in January and February, 1992. The sample was ana- lysed according to conventional methods (KorscHGEn 1980). The different prey items were determined by macroscopical and microscopical analyses and compared with our reference collections. The results are given as relative percentage of occurrence (number of each food item/total number of food items). The percentage of estimated volume (Kruuk and PARISsH 1981) of each food item in the diet was calcu- lated on 70 scats. In order to assess geographical variation in winter consumption of fruit in Europe, data were col- lected from 8 studies relative to the diet of the stone marten: CLEMENT and SAINT GIRONS 1982; SKIRNIS- son 1986; TESTER 1986; MARCHESI et al. 1989; BERTOLINO and DorE 1991; Ruız-OLMo and PALAZON 1993; SERAFINI and LovaArı 1993; LoDE 1994. The number of studies reviewed is not high, since only stu- dies carried out in winter (January, February, and March) and in rural habitats have been considered. The arcsin-transformed percentage of occurrence for different values of n (FREEMAN and TukEy 1950) was regressed against latitude of the different study areas. Table 1. Percentage of occurrence, number of each food item/total number of food items, in 89 scats of stone marten. N = number of occurrence. Food category Fruit Wild fruit/seeds Juniperus sp. Prunus spinosa Sorbus domestica Rosa canina Clematis vitalba Crataegus monogyna Cornus mas Cultivated fruit Vitis vinifera Ficus carica Pyrus communis Malus sylvestris Unidentified fruit/seeds Mammals Birds Insects Unidentified material Total items N MHMHoOomADWNWHMR HD RB ES m WW m SI \0 Fruit as a winter feeding resource in the diet of Stone marten 2A, Results The categories of prey found in stone marten faeces included fruit, mammals, birds, and insects (Tab. 1). Fruit was the most common item occurring in 55% of the samples, with ESTIMATED VOLUME (%) 100 VvoL 80 IN TOTAL DIET 60 50 % 40 20 % 20 10 % 5% ) 1% 20 40 60 80 OCCURRENCE (%) Fig. 1. Estimated volume of food categories versus their per- centage of occurrence in 70 scats of stone marten. 1: fruit; 2: mammals; 3: insects; 4: birds; 5: unidentified material. 89% volume (Fig. 1). Eleven spe- cies were identified, four of them being cultivated. The berries from juniper (Juniperus commu- nis vel oxycedrus) play an impor- tant role in the winter food of stone martens and represent the most frequently consumed fruit. They occur in 42.2% of the sam- ples and are followed in impor- tance by sloe, Prunus spinosa (29.7%) (Tab. 1). The remaining 9 species of fruit occur in 21.5% of the samples. Clematis vitalba has nonsucculent seeds and thus cannot be considered palatable for the stone marten. The seeds may have been swallowed by chance during hunting activities of this mustelid. From a total number of 121 only 13 (10.8%) corresponded to cultivated fruits: ARCSIN-TRANSFORMED % OCCURRENCE 40 42 44 46 48 LATITUDE 50 52 54 56 Fig. 2. Relationship between winter consumption of fruit and latitude in the diet of the stone marten in Europe. Arcsin-transformed percentage of occurrence for different values ofn (FREEMAN and TUKEY 1950). NW Italy (BERTOLINo and Dore 1991); CW Italy (SERAFINI and Lovarı 1993); Spain (RuIZ-OLMO and PaALAzon 1993); NW France (CLEMENT and SAINT GIRoNs 1982); NW France (Lope 1994); C Swit- zerland (MArcHESsı et al. 1989); N Switzerland (TESTER 1986); N Germany (SKIRNISSON 1986). 218 M. PANDOLFI, ANNA MARIA DE MARINIS, and I. PETROV apples, pears, figs and grapes. Mammals and insects represented less used food resources and constituted 18.6% and 16.8% of the sample, respectively. These feeding categories have a very slight incidence in the total volume of the diet (Fig. 1). Birds are only occa- sıonally taken. Figure 2 depicts the regression of winter consumption of fruit in Europe against lati- tude. A significant relationship has been found (r = 0.795; F = 12.03; P = 0.01) suggesting that stone marten feeds more on fruits in southern Europe than in the north. The figure shows NW France separated from the other areas. The study area in NW France is lo- cated in marshy habitat (Lake Grand-Lieu close to Nantes), where the fruit availability may be lower than the expected value. Discussion The stone marten reveals the most frugivorous diet among the European species of mus- telids (WEBER 1989; CLEVENGER 1994). This predator exploits fruits found on the ground but also may have access to those still on trees (WAECHTER 1975). The climbing ability of this carnivore may have determined this food selection (TESTER 1986; SERAFINI and LOVARI 1993). Fruits are rich in carbohydrates and do not require any special adaptation to be di- gested and can represent an alternative feeding resource in the cold months when the en- ergy requirements are higher. Wild fruit such as berries from juniper and sloe are the staple winter food of the stone marten in the study area. Juniper usually represents the most frequent food item in Medi- terranean habitats (Lıgois 1991) and its berries are in general replaced by Rosaceae fruit in temperate habitats (Lisoıs 1991). In east-central Italy cultivated fruit represents a feed- ing resource of secondary importance. Elsewhere in Europe cultivated fruit, such as grapes or cherries, are often consumed (WAECHTER 1975; RuIz-OLMmo and PALAZoNn 1993). The high incidence of fruit in the winter diet of this mustelid in east-central Italy may be related to habitat characteristics. Succulent fruit are very abundant and with a clumped distribution in the study area. The low incidence of small mammals recorded in the diet of stone marten in this study could be due to the relative scarcity of this food item in the study area. Monocultures of exotic conifers in the Cesane state forest are probably associated with lower diversity and availability of alternative prey (e.g. non arboreal small mammals). A combination of reasons (the availability of fruit and seeds and the scarsity of small mam- mals) may explain why the stone marten diet can be based so extensively on vegetable mat- ter in east-central Italy. These results concern, however only two months in one single winter. Differences in the stone marten diet could be recorded from year to year in relation to differences in the availability of feeding resources (KALPERS 1980). Therefore, these re- sults could not be considered conclusive. Nevertheless, they seem to confirm the general trend detected in Europe in winter consumption of fruit. The high geographic variability observed in the studies considered seems to be related to latitude. A significant trend in the form of an increase in winter consumption of fruit has been found from northern to south- ern Europe. The results suggest that where the habitat and climate allow for readily accessi- ble fruit, this food item represents a very common dietary resource (e.g. southern Europe). Also, it is evident that the stone marten can switch to different feeding resources in relation to local and temporal food availability (Lope 1994), thus confirming the view of an oppor- tunistic feeding behaviour of this species. Acknowledgements We wish to thank G. Mocsaı for his valuable assistance in the identification of fruit remains; S. LOVARI and M. Masserı for critical reading of the manuscript; J. MoGGı CEcchi for the correction of the Eng- lish text and C. CorTi for the translation of the German summary. Fruit as a winter feeding resource in the diet of Stone marten DA Zusammenfassung Früchte als Winternahrung des Steinmarders Martes foina in Mittel-Ost-Italien Diese Studie basiert auf einer Analyse von Kotproben aus dem Winter 1992 in Le Cesane (Mittel-Ost- Italien). Das Habitat besteht aus einem Gemisch von Waldflächen (hauptsächlich aus aufgeforsteten Koniferen), Acker- und Brachland. 89 Kotproben von Martes foina wurden gesammelt und nach Stan- dardmethode analysiert. Die Zusammensetzung der Nahrungskomponenten ist prozentual auf den zah- lenmäßigen Anteil und das abgeschätzte Volumen kalkuliert worden. Die Winterdiät des Steinmarders besteht hauptsächlich aus Früchten. Diese stellen zahlenmäßig einen Anteil von 55% und volumen- mäßig einen Anteil von 89% der Gesamtmasse dar. Wildfrüchte sind somit die häufigste Nahrungska- tegorie. Juniperus sp. und Prunus spinosa Beeren sind die wichtigste Nahrungsquelle (mit 42,2% bzw. 29,7%). Breitengrad und Winterfruchtkonsum sind in Europa signifikant korreliert. Die steigende Ten- denz von Früchtekonsum in der Winternahrung verläuft in Europa von Norden nach Süden. References AMORES, F. (1980): Feeding habits of the Stone marten, Martes foina (Erxleben, 1777), in south western Spain. Säugetierkdl. Mitt. 28, 316-322. BERTOLINO, S.; DoRE, B. (1991): Dati preliminari sulla dieta della faina (Martes foina) nel Parco Regio- nale “La Mandria” (Torino). Suppl. Ric. Biol. Selv. 19, 643-646. CHEYLAN, G.; BAYLE, P. (1988): Le regime alimentaire de quatre especes de mustelides en Provence: la fouine Martes foina, le blaireau Meles meles, la belette Mustela nivalis et le putois Putorius putorius. Faune de Provence 9, 14-26. CLEMENT, R.; SAINT GIRONS, M.C. (1982): Notes sur les mammiferes de France. 18. Le regime de la fouine, Martes foina (Erxleben, 1777), dans l’agglomeration nantaise et en milieu rural. Mammalia 46, 550-553. CLEVENGER, A. P. (1994) : Feeding ecology of Eurasian Pine Martens and Stone Martens in Europe. In: Martens, Sables, and Fishers. Biology and Conservation. Ed. by S. W. BusKirKk, A.S. HARESTAD, M. G. RAPHAEL, and R. A. Powerr. London: Cornell University Press. Pp. 326-340. DELIBEs, M. (1978): Feeding habits of the Stone marten, Martes foina (Erxleben, 1777), in northern Bur- gos, Spain. Z. Säugetierkunde 43, 282-288. FREEMAN, M. F.; TUKEY, J. W. (1950): Transformations related to the angular and the square root. Ann. Math. Statist. 21, 607-611. KALPERS, J. (1980): Contribution a l’Etude Ecoethologique de la fouine (Martes foina): strategie d’utilisa- tion du domaine vital et des ressources alimentaires. I. Introduction generale et analyse du regime alimentaire. Cahiers Ethol. appl. 3, 145-163. KoRSCHGEN, L. J. (1980): Procedures for food habit analyses. In: Wildlife management techniques man- ual. Ed. by S. D. SCHEMNITZ. Washington, D. C.: The Wildlife Society. Pp 113-127. KRrUUK, H.; PARISH, T. (1981): Feeding specialization of the European badger (Meles meles) in Scotland. J. Anım. Ecol. 50, 773-788. Lisois, R. (1991): La Fouine (Martes foina Erxleben, 1777). In: Encyclopedie des Carnivores de France. Ed. by M. Arroıs and P. DELATTRE. Nort s/Erdre: Soc. Frangaise pour l’Etude et la protection des mammiferes. Pp 18-25. Lope, T. (1994): Feeding habits of the Stone marten Martes foina and environmental factors in western France. Z. Säugetierkunde 59, 189-191. MARCHESI, P.; LACHAT, N.; LIENHARD, R.; DEBIEVE, P.; MERMOD, C. (1989): Comparaison des r&gimes ali- mentaires de la fouine (Martes foina Erx.) et de la martre (Martes martes L.) dans une region du Jura Suisse. Revue Suisse Zool. 96, 281-296. Poz1o, E.; GRADONI, L. (1981): Spettro trofico della Volpe (Vulpes vulpes L.) e della faina (Martes foina Erxleben) in provincia di Grosseto. Natura, Milano 72, 185-196. RASMUSSEN, A. M.; MADSEN, A.B. (1985): The diet of the Stone marten Martes foina in Denmark. Na- tura Jütl. 21, 141-144. ROMANOWSKI, J.; LESINSKI, G. (1991): A note on the diet of stone marten in southeastern Romania. Acta Theriol. 36, 201-204. RUIZ-OLMO, J.; PALAZON, S. (1993): Diet of the Stone marten (Martes foina Erxleben, 1777) in the north- eastern Spain. Donana, Acta Vert. 20, 59-67. 220 M. PANDOLFI, ANNA MARIA DE MARINIS, and I. PETROV SERAFINI, P.; LovaRrı, S. (1993): Food habits and trophic niche overlap of the red fox and the stone mar- ten in a Mediterranean rural area. Acta Theriol. 38, 233-244. SKIRNISSON, K. (1986): Untersuchungen zum Raum-Zeit-System freilebender Steinmarder (Martes foi- na Erxleben, 1777). Beitr. Wildbiol. 6,1-200. TESTER, U. (1986): Vergleichende Nahrungsuntersuchung beim Steinmarder Martes foina (Erxleben, 1777) in großstädtischem und ländlichem Habitat. Säugetierkdl. Mitt. 33, 37-52. WAECHTER, A. (1975): Ecologie de la fouine en Alsace. Rev. Ecol. (Terre Vie) 29, 399-457. WEBER, D. (1989): The diet of polecats (Mustela putorius L.) in Switzerland. Z. Säugetierkunde 54, 157- IZak Authors’ addresses: Dr. MAssımo PANDOLFI, Dr. ANNA MARIA DE MARINIS, Istituto di Scienze Morfolo- giche, Universita di Urbino, Via M. Oddi 23, I-62100 Urbino, Italy; Dr. Ivan PETRov, Institute of Forestry, Bulgarian Academy of Science, 132 St. Kl. Ohridski, BG-1756 Sofia, Bulgaria. Z. Säugetierkunde 61 (1996) 221-227 x - 2 "FÜR © 1996 Gustav Fischer, Jena SÄUG ETI ERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Genet (Genetta genetta L., 1758) diet shift in mountains of central Spain ByE. VıRrGös, J. G. Casanovas, and T. BLAZQUEZ Departamento de Biologıa Animal I (Vertebrados), Facultad de Biologia, Universidad Complutense de Madrid, Madrid (Spain) Receipt of Ms. 29. 08. 1995 Acceptance of Ms. 06. 03. 1996 Abstract This study focuses on the seasonal shift of the main prey and patterns of prey size consumed by a popula- tion genet (Genetta genetta L.) in central Spain as determined by scat analysis. Four different “middens” (186 scats) were studied. Data were presented as frequency of occurrence and biomass of each prey group considered. Seasonal variation for the occurrence of each prey group in the diet was tested with the Chi-square goodness of fit test. Mean prey size index was used to analyse the distribution pattern of prey weights consumed. The mean prey size obtained was compared with others from different areas. As in earlier studies, woodmouse (Apodemus sylvaticus) was the predominant prey group. How- ever, in winter the diet was based on birds. In addition, this group was the most important throughout the year with respect to frequency of occurrence data. This shift is discussed from an energetic point of view. The role of dung beatles was also significant and their inclusion in mean prey size calculations was recommended. The remaining groups only showed unimportant seasonal fluctuations. Mean prey size obtained is lower than in other populations. The relative role of woodmouse, arthro- pods and rabbits (Oryctolagus cuniculus) explains these differences. Introduction Studies on the diet of genet (Genetta genetta L.) in Spain and other countries are rela- tively numerous (DELIBES 1974; MAGALHAES 1974; ALcovER 1984; CALviNo et al. 1984; CuGnasseE and Rıors 1984; LoDE et al. 1991; PALOMARES and DELIBES 1991; HAMDINE et al. 1993) and have shown the generalist habit of this species. However, the species clearly prefers small mammals, which were always the main prey item in the diet, despite their seasonal or local availability (Lop&£ et al. 1991; HAMmDine et al. 1993). The remaining prey items experienced spatio-temporal variations according to their potential availability (Lı- vET and ROEDER 1987). Many studies have pointed out the important intraspecific plasti- city of carnivores in their diet selection (see BEKOFF et al. 1984) and the functional response to changes in abundance of their main mammalian prey (AnDERSoN and ER- LINGE 1977; JaKsıc 1989). Recently, studies on the generalist or specialist condition in spe- cies like the badger, Meles meles (Roper 1994; MARTIN et al. 1995) showed the importance of focusing future research on the population level, according to earlier sug- gestions (PARTRIDGE and GREEN 1985). With regard to the genet, this species has been noted to have a clear preference for the woodmouse (Apodemus sylvaticus) despite some evidence showing changes towards other prey items in certain individuals or populations (ALcoveEr 1982, 1984; DELIBES et al. 1989). On the other hand, the influence of other aspects of trophic ecology such as the size DDR E. VIRGöS, J. G. Casanovas, and T. BLAZQUEZ distribution of the prey consumed have, so far, been largely ignored. ERLINGE (1987) and Kınc (1991) demonstrated the importance of these aspects in the ecology of small muste- lıds. Knowledge of the size pattern could also permit future studies concerned with opti- mal diets and how these can change, e.g., within and between populations or individuals. These baseline data will aid research into other related ecological subjects. In this study, the seasonal changes of trophic preferences are described in a local po- pulation of genets. Material and methods The field work was conducted in the Alto Manzanares Regional Park, located on the southern slope of the Sierra of Guadarrama (central Spain). All bioclimatic and vegetation stages of central Spain moun- tains are represented in this area, which varies in altitude from 700-1200 m (Rıvas-MARTINEZ et al. 1987). To study the genet diet we selected four middens (heaps of scats) located in different but nearby areas. The first one was located in a dehesa (open wood with pastures) where holm oak (Ouercus ilex) and ash (Fraxinus angustifolia) were the dominant tree species. The second and third ones were located in a holm oak closed forest. The fourth midden was located in a pyrenaica oak forest (Ouercus pyrenai- ca). All middens were located within 850-1 000 m range. Laboratory analyses were carried out following the procedures described by CorBETT (1989) and REYnoLDs and AEBIscHER (1991). The remains were determined using diagnostic features (teeth, feath- ers, flakes, seeds, etc.) supported by keys and reference collections. The prey items considered can be consulted in table 1. Results were presented as frequency of occurrence (number of scats with prey item presencel/total number of scats x 100) and biomass for each prey item, following the methods and recommendations from CorBETT (1989). Transformation factors suggested by PALOMARES and DELIBES (1991) were applied to calculate consumed biomass, except for fruits where factors were based on Branco (1988) from a study on the red fox (Vulpes vulpes). B-diversity index of Levins (1968) was used for the frequency occurrence data in each season. Asso- ciation between diversity and occurrence data was analysed by non-parametric Spearman rank correla- tions (SIEGEL and CASTELLAN 1988). To analyse seasonal variation a 2x4 contingency table with the frequency of occurrence data was used (REYNoLDs and AEBISCHER 1991). Prey items that did not satisfy test conditions were excluded from the analysis (SoKAL and RoHLF 1981). Bonferroni sequential corrections were applied to avoid type I error (Rıce 1989). In order to analyse the prey sizes consumed, the mean prey-size index suggested by ERLINGE (1987) was used and the prey distribution was classified into three categories: <15 g; 16-50 g and >50 g follow- ing CLEVENGER (1994). DELIBES (1974) and RODRIGUEZ (1993) mean weight data were used in calcula- tions: Microtus (30 g),;, Apodemus (25 g); Mus (20 g); Rattus (150 g); Oryctolagus (150 g); small bird (30 g); medium-size bird (100 g); lacertid (3 g); snake (100 g); amphibian (15 g); arthropod (1 g). Results and discussion Frequency of occurrence and biomass results for each category are shown in table 1. The minimum number of individuals of each taxon according to category is shown in table 2. The largest group with respect to contributed biomass was small mammals, whereas birds represented the group most frequently found in the scats throughout the year. Gen- ets also consume a large number of insects and in certain seasons amphibians (winter), reptiles (summer), and fruit (autumn). Significant differences in frequency of occurrence of the five most outstanding prey items (small mammals, birds, reptiles, amphibians and insects) were found (Tab. 3), espe- cially insects and reptiles. The former had very low winter frequencies and the reptiles were consumed more in the summer. Genet diet shift in mountains of central Spain 223 Table 1. Frequency of occurrence (f) and biomass (b) of the prey groups considered; scat number ana- lysed (n) and B-index of Levıns (1968) Mammals Birds Reptilies Amphibians Insects Myriapods Eggs Carrion Fruits B-index The most important item throughout the year, was the group small-mammals except in winter when birds clearly predominated in the diet, in biomass as well as in frequency of occurrence. Diversity reached its maximum in winter, and minimum in spring coinciding, respec- tively, with the highest and the lowest frequencies of occurrence of small mammals in the scats. Frequency of mammals was negatively correlated with trophic diversity (Tab. 3). In- sects also correlated negatively with trophic diversity, but this difference was not signifi- cant. The remaining groups showed no significant positive correlations with diversity. These data reveal a shift in the main prey-item from one group to another, in accord- ance with earlier studies on the diet of genet (ALcovEr 1982, 1984; DELIBES et al. 1989). Birds composed the bulk of prey consumed in winter and they were the most significant prey-item in frequency of occurrence for the whole year, although the woodmouse re- mained the main prey item in biomass and correlated negatively with diversity. Contrary to investigations of ALCoVER (1982, 1984) and DELIBES et al. (1989), the main prey is re- placed by an equivalent group from an energetic point of view. The high winter predation suggests a specialization in roost of small passerines. They constitute abundant (TELLERIA 1987) and predictable resources in this area, both characteristics that determine high en- counter rates and make them easy to capture (high success rate). High encounter and suc- cess rate determine high prey vulnerability to predation (GREENE 1986). Moreover, they make an important contribution to biomass. As a result, the high profitability of this prey item could cause the shift in preferences. However, because of the lack of availability esti- mates we cannot assess whether this shift is a consequence of selective choice or an indi- cation of changes in main prey availability (e.g. decline). 224 E. VIRGöSs, J. G. CAsanovAs, and T. BLAZQUEZ Table 2. Prey species in each prey group consid- ered and minimun number of individuals (N) found. Mammals Oryctolagus cuniculus Crocidura russula Talpa spec. Pitymys duodecimcostatus Apodemus sylvaticus Mus spec. Muridae Small-mammals Birds Turdus merula Fringilla coelebs Passerines Birds undetermined Reptiles Psammodromus algirus Ophidia Reptiles undetermined Amphibians Anura Insects Typhaeus typhoeus Geotrupidae Escarabeidae Carabidae Cetonidae Melolonthidae Elateridae Blaps spec. Coleoptera Orthoptera Gryllotalpa spec. Formicidae Insects undetermined Eggs Turdus merula Undetermined Fruits Juniperus oxycedrus Rubus spec. Ficus carica Others Julus terrestris Buthus occitanus *: seed number found Table 3. Chi-square test for 2x 4 contingency tables on seasonal variation in presence/absence data and Spearman correlation coefficient (r,) between trophic diversity (B-index) and frequency of occur- rence data. Mammals 3432-2 Birds DIA EEE Reptiles 40.01*** Amphibians 1952 Insects AS DEES 772010088 Insects also play an important role with respect to frequency of occurrence. Other studies have already indicated the signifi- cance that this group can acquire in certain periods of the year (DELIBEsS 1974; DELIBES et al. 1989). The explanation for this high consumption in a not particularly thermic area can be found in the nature of the in- sects captured and their ecology. Most of them are dung beetles with nocturnal activ- ity, very abundant in the habitat studied due to high availability of cow faeces (PALMER et al. 1989). In the case of reptiles the Psammodro- mus algirus stands out, a plentiful species in these habitats (Diaz and CARRASCcAL 1991). Amphibians were well represented in winter. Coinciding with the last few days of this sea- son, some species were concentrated in re- productive pools (RODRIGUEZ-JIMENEZ 1983; LöPEZ-JURADO 1988) becoming an easy and predictable food resource. Fruits were eaten exclusively in autumn, when their availabil- ity reached a maximum. Only two species were significantly consumed, blackberries (Rubus spp.) and figs (Ficus carica). How- ever, these are not very important in the overall diet, especially compared to other Mediterranean generalist carnivores (CIAM- PALINI and LovArı 1985; BLAnco 1988; ALE- GRE et al. 1991). The prey size found (17 g) is very small compared with that recorded for other genet populations: 27g (Galicia, CALvIiNo et al. Genet diet shift in mountains of central Spain 22 1984); 24 g (Mallorca, ALcovEr 1984); 40 g (Donana, PALOMARES and DELIBES 1991); 12 g (Algerie, HAmDine et al. 1993), and is primarily influenced by the high rate of insects and other arthropods in the diet. In Algerian genets the prey size value was lower still and they also ingested a high proportion of insects (HAMDINE et al. 1993). Data from Mallorca and Galician populations show values within the range of woodmouse weights, in accordance with a higher and lower ingestion of woodmouse and arthropods, respectively, in these areas (ALCoVER 1984; CALvINo et al. 1984). The value obtained from Donana (PALOMARES and DELIBEs 1991) revealed the importance of rabbits (Oryctolagus cuniculus) and rats (Rattus spp.) in the diet of this population. Nevertheless, it is impossible to determine whether these large differences indicate a higher level of trophic stress (HEGGEBERT and Moseıp 1994) in populations with an important arthropod contribution in their diet with- out data on the physical condition of individuals. On the other hand, smaller prey size could be an optimal adaptation on many occasions (JUAnES 1994) despite early predictions from classical optimal foraging theory (STEPHENS and Kres 1986). GITTLEMAN (1985) and FisHErR and DicKmAn (1993) pointed out a tendency for smaller prey sizes in insectivorous carnivores. This interspecific pattern could be extrapolated to the intraspecific level. Moreover, KınG (1991) discusses whether arthropods should be included when apply- ing to the index proposed by ErLinGeE (1987) and considers that this should depend on whether this group is incidentally predated. Regarding the ecological characteristics of the majority of insects found in the diet, we consider that their location in predictable and abundant patches (cow faeces) render them especially rewarding prey items (high prey vulnerability). In addition, their rapid and easy ingestion results in a minimum of time lost before they can begin searching for more profitable prey (see FisHER and DICKMAN 1993). Likewise, it is uncertain whether other prey groups (amphibians and reptiles) should be included in calculation of the index in view of their probable incidental preda- tion. However, their seasonal and local importance preclude their exclusion. In any case a greater consensus is required about which groups to include in the application of these in- dices for generalist carnivores. Extrapolation of the criteria used for small specialist mus- telids (ErLinGE 1987; KınG 1991) is not reasonable. The absence of larger prey is noteworthy and is probably due to the scarcity of rabbits and rats in the study area. As in other subjects related to trophic ecology, it would be ne- cessary to know the range of available sizes in the environment in order to determine whether observed patterns reflect selective prey choice by genets. All the above data manifest the generalist character of genets in this area (despite quantitative estimates of food availability). The shift of the main prey item towards birds verifies the plastic behaviour of this species. Acknowledgements We are very grateful to Y. CorTEs, F. J. SAMBLAS, D. GARcCIA, M. A. GARcIA, and R. MARTINEZ, to G. P. FARINÖS, J. C. ATIENZA, and J. BARREIRO for helping us with prey determination; to G. P. FARINÖS and G. G. NıcoLA for helping us with the translation, to J. L. TELLERfA, J. A. ALCOVER, and J. C. BLANcO for providing suggestions to earlier drafts of the manuscript and to F. PALOMARES for some useful refer- ences. This work has been written while E. V. was under a grant from DGICYT PB92-0238 project. Zusammenfassung Nahrungswechsel bei Ginsterkatzen (Genetta genetta L., 1758) in den Gebirgen von Zentralspanien Diese Arbeit beschäftigt sich mit dem aus Kotproben ermittelten jahreszeitlichen Wechsel der Nah- rungsbestandteile und der Verteilung der Größe der Beute in einer Ginsterkatzen-Population in Zen- 226 E. VIRGöS, J. G. Casanovas, and T. BLAZQUEZ tralspanien. 186 Kotproben von vier Absetzstellen wurden untersucht und die Ergebnisse als Frequen- zen der Häufigkeit und als Biomasse pro Beutegruppe dargestellt. Die jahreszeitliche Variabilität in der Verteilung auf Beutegruppen wurde mit dem Chi-Quadrat-Test ermittelt. Der mittlere Index der Beutegröße wurde zur Analyse der Verteilung der Gewichte der Beute benutzt. Die mittlere Beutegröße wurde mit solchen aus anderen Regionen verglichen. Wie aus früheren Untersuchungen bekannt stellte die Waldmaus (Apodemus sylvaticus) die wich- tigste Beute dar. Im Winter bestand die Beute jedoch weitgehend aus Vögeln. In bezug auf die Häufig- keit im Auftreten in der Nahrung stellten Vögel durch das ganze Jahr hindurch den bedeutendsten Futterbestandteil dar. Dieser Wechsel wird vom energetischen Standpunkt her besprochen. Auch war der Anteil an Mistkäfern bemerkenswert. Es wird empfohlen, Mistkäfer bei der Ermittlung der mittle- ren Beutemengen zu berücksichtigen. Die übrigen Nahrungsgruppen weisen nur geringe jahreszeitliche Schwankungen auf. Die mittlere Beutegröße ist geringer als bei anderen Populationen. Die relative Bedeutung von Waldmäusen, Arthropoden und Kaninchen (Oryctolagus cuniculus) erklärt diese Unterschiede. References ÄLCOVER, J. A. (1982): On the diet of Carnivora in islands: a method for analyzing it and a particular case. Doniana, Acta Vertebrata 9, 321-339. ALCOVER, J. A. (1984): Über die Nahrung der Ginsterkatze, Genetta genetta, auf den Inseln Mallorca, Ibiza und Cabrera. Säugetierkdl. Mitt. 31, 189-195. ÄLEGRE, J.; HERNANDEZ, A.; PURROY, F.; SALGADO, J.; FUERTES, B. (1991): Dieta otono-invernal de la gar- duna, Martes foina, en un häbitat rural de Leön (Espana). Ecologia 5, 265-274. ANDERSSON, M.; ERLINGE, S. (1977): Influence of predation upon rodent populations. Oikos 29, 591- 5972 BEKOFF, M.; DANIELS, T. J.; GITTLEMAN, J. L. (1984): Life history patterns and the comparative social ecology of carnivores. Ann. Rev. Ecol. Syst. 15, 191-232. Branco, J. C. (1988): Estudio ecolögico del zorro (Vulpes vulpes) en la Sierra de Guadarrama. Tesis Doctoral; Univ. Oviedo CALVINO, F.; DE CASTRO, A.; CANALS, J. L. S.; GUITIAN, J.; Bas, S. (1984): Regimen alimenticio de la gine- ta, Genetta genetta, en Galicia, noroeste de la Peninsula Ib£rica. Bol. Estaciön Central de Ecol. 13, 2941. CIAMPALINI, B.; LovaRrı, S. (1985): Food habits and trophic niche overlap of the badger (Meles meles) and the red fox; (Vulpes vulpes) in a Mediterranean coastal area. Z. Säugetierkunde 50, 226-234. CLEVENGER, A. P. (1994): Comparative feeding ecology of the Eurasian pine marten (Martes martes) and stone marten (Martes foina) in Europe. In: Martens, sables and fishers: biology and conserva- tion. Ed by S. W. Buskikk, A. S. HARESTAD, M. G. RAPHAEL, and R. A. Powerr. Ithaca, NY: Cornell Univ. Press. Pp. 326-340. CORBETT, L.K. (1989): Assessing the diet of dingoes from faeces: a comparison of three methods. ]. Wildl. Manage. 53, 343-346. CUGNASSE, J.-M.; Rıors, C. (1984): Contribution a la conaissance de l’ecologie de la genette (Genetta genetta) dans quelques departements du sud de la France. Gibier Faune Sauvage 1, 25-55. DELIBES, M. (1974): Sobre alimentaciön y biologia de la gineta (Genetta genetta) en Espana. Donana, Acta Vertebrata 1, 143-199. DELIBES, M.; RODRIGUEZ, A.: PARRENO, F. (1989): Food of the common genet (Genetta genetta) in north- ern Africa. J. Zool. (London) 218, 321-326. | Diaz, J. A.; CARRASCAL, L. M. (1991): Regional distribution of a Mediterranean lizard: influence of habi- tat cues and prey abundance. J. Biogeogr. 18, 291-297. ERLINGE, S. (1987): Why do European stoats (Mustela erminea) not follow Bergmann’s rule?. Holarct. Ecol. 10, 33-39. FISHER, D. O.; DicKMan, C.R. (1993): Body size-prey size relationships in insectivorous marsupials: tests of three hypotheses. Ecology 74, 1871-1883. GITTLEMAN, J. L. (1985): Carnivore body size: ecological and taxonomic correlates. Oecologia 67, 540- 554. GREENE, C.H. (1986): Patterns of prey selection: implications of predator foraging tactics. Am. Nat. 128, 824-839. Genet diet shift in mountains of central Spain 227 HAMDINE, W.; THEVENOT, M.; SELLAMI, M.; SMETK. DE (1993): Regime alimentaire de la genette (Gen- etta genetta) dans le Parc National du Djurjura, Algerie. Mammalia 57, 9-18. HEGGEBERT, T. M.; Moseip, K.-E. (1994): Prey selection in coastal Eurasian otters, Lutra lutra. Ecogra- phy 17, 331-338. JAKsıc, F. M. (1989): Opportunism vs selectivity among carnivorous predators that eat mammalian prey: a statistical test of hypotheses. O1kos 56, 427-430. JuanEs, F. (1994): What determines prey size selectivity in piscivorous fishes?. In: Theory and applica- tion in fish feeding ecology. Ed. by D. J. STOUDER, K.L. FrEsH and R. J. FELLER. SC: Univ. South Carolina Press. Pp. 79-100. Kıng, C. (1991): Body size-prey size relationships in European stoats, Mustela erminea: a test case. Holarct. Ecol. 14, 173-185. Levins, R. (1968): Evolution in changing environments. Princeton: Princeton Univ. Press. LivErt, F.; ROEDER, J. J. (1987): La genette (Genetta genetta): In: Encyclopedie des carnivores de France. Ed. by M. Arroıs and P. DELATTRE. S.F.E.P.M. 16, 1-33. LoDe, T.; LECHAT, I.; LE JAcQuzs, D. (1991): Le regime alimentaire de la genette en limite nord-ouest de son aire de r£partition. Rev. Ecol. (Terre Vie) 46, 339-348. LoPEZ-JuURADO, L. F. (1983): Estudios sobre el sapo corredor (Bufo calamita) en el sur de Espana. III. Reproducciön. Donana, Acta Vertebrata 10, 19-39. MAGALHAES, C. M. P. (1974): Habitos alimentares da raposa (Vulpes vulpes silacea) e da genetta (Gen- etta genetta) na Tapada de Mafra. Secretaria de Estado da Agricultura. Dir Gnal. dos Servicios Forestais e Agricolas. MARTIN, R.; RODRIGUEZ, A.; DELIBES, M. (1995): Local feeding specialization by badgers (Meles meles) in a Mediterranean environment. Oecologia 101, 45-50. PALMER, M.; GARCIA-PLE, C.; MoREy, M. (1989): Explotaciön del recurso por los escarabajos copröfagos del genero Aphodius. Bases para un modelo. Elytron 3, 115-124. PALOMARES, F.; DELIBES, M. (1991): Alimentaciön del meloncillo (Herpestes ichneumon) y de la gineta (Genetta genetta) en la Reserva Biolögica de Donana, S. ©. de la Peninsula Iberica. Donana, Acta Vertebrata 18, 5-20. PARTRIDGE, L.; GREEN, P. (1985): Intraspecific feeding specializations and population dynamics. In: Be- havioural ecology: ecological consequences of adaptive behaviour. Ed. by R.M.Sısry and R. H. SmitH. Oxford: Blackwell Press. Pp. 207-226. REYNOLDS, J. C.; AEBISCHER, J. (1991): Comparison and quantification of carnivore diet by faecal analy- sis: a critique, with recommendations, based on a study of the fox, Vulpes vulpes. Mammal Rev. 21, 97-122. Rice, W. R. (1989): Analyzing tables of statistical tests. Evolution 43, 223-225. RıvAs-MARTINEZ, S.; FERNANDEZ-GONZALEZ, F.; SANCHEZ-MATA, D. (1987): El Sistema Central: de la Sierra de Ayllön a Serra da Estrela. In: La vegetaciön de Espana. Ed. by M. PEınADo and S. Rıvas- MARTINEZ. Madrid: Serv. Publ. Univ. Alcalä. Pp. 419-451. RODRIGUEZ, J. L. (1993): Guia de campo de los mamiferos terrestres de Espana. Barcelona: Omega. RODRIGUEZ-JIMENEZ, A.J. (1988): Fenologia de una comunidad de anfibios asociada a cursos fluviales temporales. Donana, Acta Vertebrata 15, 29-43. ROPER, T. J. (1994): The European badger (Meles meles): food specialist or generalist?. J. Zool. (Lon- don) 234, 437-452. SIEGEL, S.; CASTELLAN, N. J. (1988): Nonparametric statistics for the behavioral sciences. 2nd ed. New York: McGraw-Hill. SOKAL, R. R.; ROHLF, F. J. (1981): Biometry. 2nd ed. San Francisco: Freeman. STEPHENS, D.; KREBS, J. R. (1986): Foraging theory. Princeton, NJ: Princeton Univ. Press. TELLERIA, J. L. (1987): Biogeografia de la avifauna nidificante en Espana central. Ardeola 34, 145-166. Authors’ addresses: EMILIO VIRGös, Dept. Biologia Animal I (Vertebrados), Universidad Complutense de Madrid, E-28040 Madrid; JORGE G. CAsanovAs, Grupo de Ecologia Aplicada. C/Bernardina Garcia 27, E-28047 Madrid; Tomäs Bläzquez, C/Camino de Vina- teros 135, E-28030 Madrid, Spain. Z. Säugetierkunde 61 (1996) 228-235 © 1996 Gustav Fischer, Jena SÄUG ETl ERKÜN R DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Variation in the baculum of the European souslik, Spermophilus citellus By B. KryStureEk and V. HRABE Slovene Museum of Natural History, Ljubljana, Slovenia and Institute of Landscape Ecology, Brno, Czech Republic Receipt of Ms. 15. 01. 1996 Acceptance of Ms. 15. 03. 1996 Abstract We analysed 35 bacula from six geographical subsamples of the European souslik, Spermophilus citellus (Linnaeus, 1766) from the Pannonian Plain and the Balkan Peninsula. Highly significant between-sam- ple differences were found in all four scored parameters: baculum length, basal breadth, breadth of the spoon-shaped distal spatula, and the number of tooth-like projections (denticles) along the spatula’s border. The greatest interpopulation differences found were in the spatula breadth and in the number of denticles. Interpopulation differentiation shown by the bacular characters follows the pattern of var- iation previously observed in non-metric cranial traits, and is not affected by simple geographic dis- tance. In S. citellus the baculum contains more phyletic information than do conventional taxonomic characters such as the size and proportions of the skull. Introduction The baculum or os penis is a heterotopic bone in most mammals. Its morphology varyies considerably between species but it generally remains constant within species. Its function is to support the penis during copulation and it may also play a role in “internal court- ship” (sensu EBERHARD 1985). Since the work of THomas (1915) it has been known that the analysis of bacular characters is often superior to the use of cranial traits for phyloge- netic reconstructions in groups such as squirrels (Sciuridae). Accordingly, bacular mor- phology has often been used to elucidate taxonomic problems in Nearctic sciurids (e.g. WHITE 1953; BurT 1960), although only rarely applied to the study of Palaearctic squirrels (e.g. RESHETNIK and BALAKHNIN 1967). The European souslik, Spermophilus citellus (Linnaeus, 1766), is the most western Eurasian ground squirrel, and is characterised by a high degree of interpopulation varia- tion. Untill now, no less than 9 subspecies have been described, usually on the basis of conventional taxonomic characters such as colour, size and skull proportions (RuZıc 1978; KrySTUFEk 1995). Although these characters do not allow the recognition of clearly de- fined races (KRYSTUFEK 1995), variation in nonmetric cranial traits demonstrates signifi- cant interpopulation differences, at least in one segment of the species’ range (KRYSTUFEK 1990). This is particularly evident in the Balkans, where differences between samples have been ascribed to allopatric divergence, whilst the relative homogeneity seen in Pannonian sousliks presumably reflects their comparatively recent colonisation of the Pannonian Plain (see KrySTtUFEk 1990, 1995). The baculum of the European souslik was first described and drawn by Dipier (1952), although he did not indicate the origin of the single specimen available to him. RESCHET- Variation in the baculum of Spremophilus citellus 229 NIK and BALAKHNIN (1967) drew the baculum of S. citellus, and compared it with the same bone in S. suslicus and S$. pygmaeus. MECZYNSKI (1971) studied the male genitalia includ- ing the bacula of five European sousliks from Opole voivodship in Poland, and RuZıc (1978) drew a specimen from southern Banat, Serbia. KAyA and SımseEk (1986) studied ba- cular variability in a representative sample of S. citellus thracius and compared it with $. xanthoprymnus; unfortunately their figures are too small to show finer details of bacu- lar structure. Despite several attempts to describe the baculum of the European souslik, the nature and range of variation of this structure remains obscure. The aim of the present study is to describe baculum interpopulation variability in six geographic samples of the European souslik, and to evaluate the phyletic weight of bacular morphology in the study of evolu- tionary divergence within this species. Material and methods We examined 35 bacula, originating from six geographically distinct regions (Fig. 1). Samples, their des- ignation numbers, and sample sizes (in brackets) are as follows: Sample 1 - Czech Republic: Trebic, Stu- denec (n = 2). Sample 2 - Serbia, Srem: Indjija; Fruska gora Mt. (n = 9). Sample 3 - Serbia, Deliblatska Fig. 1. Site localities for the six samples of $. citellus, connected by the Minimum Spanning Tree with the length of the edges. See text for designation numbers of samples. Approximate distribution (shaded) is modified from RuZıc (1978). 230 B. KrySturek and V. HRABE peScara (south): Banatska Palanka; Samos (n = 5). Sample 4 - Serbia, Deliblatska peScara (north): Orlo- vat (n=4). Sample 5 — Macedonia: lowlands along the Vardar River: Lake Dojran; Gevgelija (n = 6). Sample 6 — Macedonia: Jakupica Mt.: Gorno Begovo; Solunsko pole (n = 9). Sample 3 is topotypical with S. c. laskarevi, sample 5 with S. c. gradojevici, and sample 6 with S. c. karamani. Only adult males (those having attained at least their second calendar year) were selected for analysis. Sousliks were aged by tooth wear and date of collection (RuZıc 1966). One year old specimens strongly predominated in our material with 77%. The rest (1.e. 17%) were two years old animals, while older sousliks (2 in all) were exceptional. Consequently we presume that age did not influence our results. Glans penis and adjoining part of the corpus were removed from freshly killed animals and pre- served in 70% ethanol until processed. We adopted the method described by AnDErson (1960), i.e. ma- ceration of the glans penis for 2 to 4 days in 2-3% potassium hydroxide with a few drops of a saturated alcoholic solution of Alızarin Red S. The stained baculum was placed in glycerine, with a small quantity ofthymol added as a disinfectant. The material is stored, together with standard museum specimens, in the Natural History Museum of Slovenia at Ljubljana. Four linear parameters were scored for each baculum using a dissecting microscope fitted with an eyepiece graticule: LB - length of baculum, BB - basal breadth, BS - breadth of spatula, ND — number of denticles. Differences between and within groups were analysed by One way Analysis of Variance (ANOVA), Principle Components Analysis (PCA), and Clustering (UPGMA) performed on the Average Taxo- nomic Distance (ATD) matrix obtained from z-standardized data. Procedures were performed using the NTSYS-pc routine (RoHLF 1989) and the Statgraphics (version 5) statistical programme. A Mini- mum Spanning Tree (MST) based on sample means was calculated from the ATD matrix. In an attempt to detect local distortions the MST was superimposed on the first three principal components axes and on the projection of samples on a geographic map. Results The baculum of $. citellus consists of a corpus, a thickened base, and a spoon-like ex- panded distal spatula which bears ventrally oriented, sharp, tooth-like projections along its margin (Meczynskı 1971). The baculum is asymmetrical, this being seen most clearly in the shape of the spatula. This asymmetry is directional, the right side of the spatula usually being more well-developed. Marked inter-sample differences in shape are evident in all parts of the baculum (Fig. 2). The corpus is most robust in sample 5, and slimmest in sample 6. The basal bor- der is concave in sample 1, but it is straight in all the others. A sagittal, knob-like, ante- rior projection on the dorsal side of the spatula is robust in the majority of samples, particularly so in sample 5, but inconspicuous in sample 6. Not surprisingly, all four bacular parameters showed highly significant interlocality variation (F>8.0, p<0.0001; Tab. 1). On the other hand, intrapopulation variation was high (see coefficients of variation values; Tab. 1) in three out of four parameters (BB, BS, ND). For example, the conylobasal length of the skull (CbL) as a single measurement which represents the size of the animal better than any other cranial trait (KRYSTUFEK 1995) is less variable than bacular parameters, with the coefficient of variation in samples 2 to 6 being between 1.62 and 2.35. Intercorrelations between bacular parameters were low, and only two of six pairwise comparisons correlated significantly (Tab. 2). Poor fit between BS and ND suggests that the number of denticles depends on the breadth of the spatula. Only BB was correlated significantly with the condylobasal length of the skull as a measure of general body size. Other bacular parameters were independent of size as long as all the material was pooled together. Nevertheless, a significant positive partial correlation existed between LB and CbL when sample 5 was removed from the analysis (r = 0.66, p < 0.0005). Thus, the length of the baculum cannot be predicted from the general size of the animal. The largest Eu- ropean sousliks (sample 5) have relatively shortest bacula, measuring between 5.9 and Variation in the baculum of Spremophilus citellus DS 9° Aug Fig. 2. Bacula of six S. citellus populations in ventral (lower row) and lateral view (upper row). (Sce Fig. 1 for the locality designation numbers.) Table 1. Means (x; in mm), Standard deviations (SD), Coefficients of variation (CV) and F-test values of four bacular parameters among six geographic samples of $. citellus. See text for acronyms of bacular parameters and sample designation numbers. 22 B. KrySturek and V. HRABE Table 2. Correlation of four bacular parameters 6.3% of the CbL (average 6.1 + 0.16). Re- and with the Condylobasal length ofskull(CbL)as Jative length was between 6.1 and 7.8% an index of averall size in $. citellus. See text for of the CbL in sousliks from the remain- ing samples. The baculum was relatively longest in sample 2 where it was, on aver- age, 7.3.=.03070l2thex@hle Sousliks from sample 5 also had the broadest bacular base, being on average acronyms of bacular parameters. Asterisks indicate correlation coefficients (r) significantly different fOoMZer04 2.037,92 00 BB =. 41.1+5.98% of LB. This value amounted En S Be to 29.2% in sample 6, which had the nar- er = ne rowest base. It ıs notable that the two geographically most closely samples dis- played the extremes of variation of this quotient. The spatula is broadest relative to the base of baculum in sousliks from sample 2 (average 161 + 16.3% of BB). The lowest values for this quotient were those from sample 5 (113+21.5%), where the spatula was as broad as the bacular base in extreme speci- mens. The number of teeth on the spatula was lower, between 6 and 12, in sousliks from the Balkan Peninsula (samples 5, 6) than in specimens from the remaining samples that had 10 to 16 teeth. These two groups formed two clusters in the One way ANOVA (p < 0.05). PCA was based on a correlation matrix of z-standardised bacular data. The projection of sample centroids onto the first three principal components is shown in figure 3. The first principal component (46.9% of variance explained) had high positive loadings for BS and ND and separated the two Balkan populations with narrow spatulae and low numbers of teeth from the remaining samples. The second and third components (ex- plaining 28.4 and 24.2% of variance, respectively) had high negative loadings for both LB and BB. Results obtained from UPGMA clustering of the ATD matrix of z-standardised data showed the same intergroup relations as suggested by ordination (Fig. 4). However, the projection of the MST onto a geographic map of the localities (Fig. 1) suggests that differ- ences exist between geographic distance and phenetic distance, expressed as ATD. Two samples from closely-placed sites in the Balkans (5 and 6) showed no mutual relation- ships, but both were linked to sample 3. There was no significant correlation between geo- Fig. 3. Projection of group centroids of six S. citellus populations onto the first three Principal compo- nents axes. Centroids are connected by the Minimum Spanning Tree. (See Fig. 1 for locality designation numbers.) Variation in the baculum of Spremophilus citellus 2353 ATD 2.0 1.6 1.2 0.8 0.4 Fig. 4. Cluster analysis of six S. citellus populations, based on UPGMA dendrogram. Cophenetic corre- lation r = 0.895. (See Fig. 1 for locality designation numbers.) graphic distance and the ATD matrices (r = 0.38, Mantel t-test = 1.052, not significant). The divergence between samples revealed by bacular morphology is therefore not af- fected by geography. Discussion The baculum of the European souslik lies in a superficial position (sensu PATTERSON 1983) and the female choice hypothesis predicts that such a baculum should function as an ‘in- ternal courtship’ device (EBERHARD 1985). If so, competition between males for mates could be expected to result in rapid evolutionary divergence in bacular morphology, whilst the remainder of the phenotype remained unaffected. This seems to be consistent with the observations on the European souslik. The characters which are under direct en- vironmental selection (size, colour) diverged between S. citellus populations to a lesser de- gree (KrYSTUFEK 1995) than bacular characters. Interpopulation relations between five souslik samples from Serbia and Macedonia sug- gested by the clustering of bacular parameters are the same as those obtained from non- metric cranial traits (KrySTUFEK 1992). Accepting the assumption that nonmetric skeletal traits reflect genetic differentiation between populations (e. g. HARTMAN 1980), the struc- ture of the baculum is likely to contain a considerable ammount of phyletic information, making it an appropriate structure for the study of evolutionary divergence in S. citellus. For example, bacular morphology suggests $. c. karamani (sample 6) to be the most dis- tinct, which is consistent with analyses of cranial traits, both nonmetric (KrYSTUFEK 1992) and metric ones (KrYSTUFEK 1995). The basic difference between the two sets of data is ca- tegorical interpopulation variation of baculum, while cranial data overlap to a large extent between samples. It is thus possible to ascribe each baculum of S$. citellus karamani to the actual group, which is not the case with the skull. Anyhow, we should underline that we have data on bacular variation from only one part of the species distribution range. Any in- crease in samples studied could diminish the diagnostic value of the baculum shape as is the case with cranial morphology (compare KrySTUFEK 1993, 1995). Differences in bacular morphology in some Nearctic ground squirrels (BUrRT 1960) are of approximately the same magnitude as those demonstrated in this study between geo- graphic samples of S. citellus, and a similar impression is obtained from studying drawings of the bacula of three closely-related western Palaearctic taxa: S. citellus, S. pygmaeus and 234 B. KrySturek and V. HRABE S. suslicus (RESHETNIK and BALAKHNIN 1967). However, before reaching any definite con- clusions about evolutionary and taxonomic significance of the bacular variability in the European souslik, data is needed on genetic divergence between European souslik popu- lations, data on interpopulation variability in bacular morphology on a larger geographic scale, and for other species of ground squirrel. Acknowledgements We thank Dr. R. HUTTERER (Bonn) for providing much stimulating discussion, Dr. H. GrIFFITHS (Hull) for improving the English text and style and Dr. A. Hırre (Bonn) for translating the abstract into Ger- man. This research was partly supported by the Ministry of Science and Technology of Slovenia (grant PI-5079-0614 to BK). Zusammenfassung Variabilition des Baculum zwischen Populationen des Europäischen Ziesels Spermophilus citellus Untersucht wurden 35 Bacula des Europäischen Ziesels Spermophilus citellus (Linnaeus, 1766) in 6 geo- graphischen Stichproben von der Pannonischen Ebene und der Balkanhalbinsel. Alle 4 untersuchten Merkmalsstrukturen — Länge des Baculum, Breite an der Basis, Breite der löffelförmigen distalen Spa- tula und Zahl der Zähnchen (Dentikel) am Außenrand der Spatula — waren hochsignifikant zwischen den Stichproben verschieden. Die größten Unterschiede entfielen auf die Spatulabreite und die Zahl der Dentikel. Die Populationsunterschiede in den Baculumausprägungen assoziieren mit bereits festgestellter Merkmalsvariation in qualitativen Schädelmerkmalen und zeigen sich stabil gegenüber der geographischen Interdependenz der Stichproben. Verschiedene Baculumstrukturen von S. citellus enthalten daher mehr phyletische Information als konventionelle taxonomische Merkmale wie etwa Größe oder Schädelproportionen. References ÄNDERSON, S. (1960): The baculum in Microtinae rodents. Univ. Kansas Publ. Mus. Nat. Hist. 12, 181- 216: Burrt, W.H. (1960): Bacula of North American Mammals. Misc. Publ. Mus. Zool. Univ. Michigan 113, 1-76. Dipier, R. (1952): Etude syst&matique de l’os penien des mammiferes. Mammalia 16, 7-23. EBERHARD, W. G. (1985): Sexual selection and animal genitalia: Cambridge, Mass.: Harvard Univ. Press. HARTMAN, S. E. (1980): Geographic variation analysis of Dipodomys ordii using nonmetric cranial traits. J. Mammalogy 61, 436-448. Kaya, M. A.; SımseK, N. (1986): The importance of the baculum in distinguishing the subspecies of ground-squirrel, Spermophilus citellus (L. 1766), (Mammalia: Rodentia) in Turkey. Doga TU Bio. D 10, 385-390. (In Turkish with summary in Eng.). KRYSTUFEK, B. (1990): Nonmetric cranial variation and divergence of European sousliks (Citellus citel- lus) from Yugoslavia (Rodentia: Sciuridae). Boll. Zool. 57, 351-355. KRYSTUFEK, B. (1993): European sousliks (Spermophilus citellus; Rodentia, Mammalia) of Macedonia. Scopolia 30, 1-39. KRYSTUFEK, B. (1995): Phenetic variation in the European souslik, Spermophilus citellus (Mammalıa: Rodentia): Bonn. zool. Beitr. 1-2 (in press). MECZYNSKT, S. (1971): Morphohistological analysis of the male genital organs of the genus Citellus. Acta Theriol. 16, 371-386. PATTERSON, B. D. (1983): Baculum-body size relations as evidence for a selective continuum on bacular morphology. J. Mammalogy 64, 496-499. PATTERSON, B. D.; THAELER, C. S. Jr. (1982): The mammalian baculum: hypotheses on the nature of bacu- lar variability. J. Mammalogy 63, 1-15. Variation in the baculum of Spremophilus citellus 235 RESHETNIK, E. G.; BALAKHNIN, I. A. (1967): Certain differences in three species of the genus Citellus in the fauna of the Ukrainian SSR. Dopovidi Akademii nauk Ukrainskoi RCR 5, 465-468. (In Ukrai- nian with summaries in Russian and Engl.). RoHLfr, F. J. (1989): NTSYS-pc. Numerical Taxonomy and Multivariate Analysis System. Version 1.50. New York: Exeter Publ., LTD. Ruzic, A. (1966) Odredjivanje uzrasnih kategorija u populaciji tekunice Citellus citellus L. Arhiv bio- loskih nauka 18, 65-70. (In Serbian with summary in Russian). Ruzic, A. (1978): Citellus citellus (Linnaeus, 1766) -— Der oder das Europäische Ziesel. In: Handbuch der Säugetiere Europas. Vol. 1/I, Ed. by J. NIETHAMMER und F. KrApp. Wiesbaden: Akad. Verlagsges. Pp. 123-144. THomas, O. (1915): The os penis or “baculum”, as a guide to the classification of certain squirrels. An- nals and Magazine of Natural History 15, 383-387. WHITE, J. A. (1953): Taxonomy of the chipmunks, Eutamias quadrivittatus and Eutamias umbrinus. Univ. Kansas Publ., Museum Nat. Hist. 5, 563-582. Authors’ addresses: BoRIS KRYSTUFEK, Slovene Museum of Natural History, PO Box 290, 1001 Ljubl- jana, Slovenia; Vır HRABE, Institute of Landscape Ecology, AS CR, Kvetna 8, 603 65 Brno, Czech Republic. Z. Säugetierkunde 61 (1996) 236-241 S 2 FÜR © 1996 Gustav Fischer, Jena SÄUG ETI E AKO N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Diversity and abundance of small mammals in Iberia: peninsular effect or habitat suitability? By A. BArBosA and J. BENZAL Museo Nacional de Ciencias Naturales, Departamento de Ecologla Evolutiva, CSIC, Madrid, Spain Receipt of Ms. 26. 04. 1994 Acceptance of Ms. 29. 02. 1996 Abstract The influence of geographical and climatic factors (latitude, longitude, altitude, mean annual precipita- tion and mean annual temperature) were investigated on the diversity and abundance of Iberian small mammal fauna. We analysed 92 983 small mammal preys from barn owl pellets. Data were taken both from the literature and by ourselves in 183 localities throughout the Iberian Peninsula. No relationships were found between altitude and mean annual temperature. In contrast, a positive relationship between diversity, abundance, latitude and mean annual precipitation was found. Such a trend is discussed in re- lation to the “peninsular effect” hypothesis. Our results do not support this hypothesis, since we did not find differences in small mammal abundance between southern Spain and northern Morocco. We con- clude that the latitudinal pattern observed in the diversity and abundance is most likely the result of the poor conditions of the Mediterranean environment for small mammals. Introduction The diversity of small mammals is affected by both geographical and climatic factors (PıanKkA 1966; BonD et al. 1980; DELIBES 1985; ALCANTARA 1989; BRÜNNER and NEET 1991; MorRENno and BARBOSA 1992). The Iberian Peninsula shows a great variation from wet and temperate conditions of the Eurosiberian region in the north, to xeric and warm conditions of the Mediterranean region in the south. Altitude also shows great variability along such geographical gradi- ents. In spite of this stated variability, there are few studies available dealing with the var- iation of species in respect to abundance and diversity among small mammals, which consider both main geographical variables (latitude, longitude and altitude) and main cli- matic variables (precipitation and temperature) together. On the other hand, many authors have debated the presence or absence of the “pen- insular effect” (species abundance decreases from base to tip of the peninsula) (SIMPSON 1964) in several animal groups and several peninsulas (MCARTHUR and Wırson 1967; KıE- STER 1971; TAyLor and REGAL 1978; SEı8 1980; BusAck and HEDGE 1984; MEANS and SIM- BERLOFF 1987). In western Europe, abundance of small mammal species decreases from north th south (HERRERA 1974). The aim of the present study was: 1) to test whether the pattern found by HERRERA (1974) at the continental scale is consistent at the regional scale, and 2) to test the “Peninsular effect” on small mammals in Iberia. Material and methods We used data drawn from barn owl diet as the sample method. In spite of the limitations of this sam- pling procedure (SAINT-GIRoNS and Spitz 1966), it has been shown to be valid for biogeographical stu- dies of small mammals (HERRERA 1974; BRUNET-LECOMTE and DELIBES 1984; CLARcK and Bunck 1991; Diversity and abundance of small mammals in Iberia 237, Moreno and BARBoSA 1992). Data were collected from literature and by ourselves (BEnZAL 1982; BAR- BOSA et al. 1992). We analysed about 92 983 small mammal preys from 183 localities (Fig. 1). Localities with less than 100 prey items were not considered. Fig. 1. Area of samples on the Iberian Peninsula (stripped area). Eurosiberian region (above the dotted line). Mediterranean region (below the dotted line). Small mammal diversity was calculated using the Shannon-Weaver Index and abundance was taken as the number of species at each locality. Climatic data were obtained from Erıas and Ruız (1977) and from the Instituto Nacional de Mete- rologia Data Bank for the points nearest to sampling sites. We selected mean annual precipitation and mean annual temperature as the two most independent and comprehensive indicators of climate. Partial correlation analyses were used to test for the effect of latitude, longitude, altitude, tempera- ture and precipitation on the small mammal diversity and abundance (NETER and WASSERMAN 1974; ZAR 1984). Stepwise multiple regression analyses (SoKAL and RoHLF 1981) were performed when more than one variable was significant in partial correlation analyses (see MoREno and BArBosA 1992 for a similar approach). To test for the peninsular effect hypothesis, abundance was compared among northern and southern Spain and northern Morocco (data from CoRBET and OVENDEN 1980; AULAGNIER and THEVENOT 1986). If a peninsular effect exists, the number of species should increase on the African side of the Strait of Gibraltar. Results Simple correlation analyses show a significant positive relationship for diversity and abun- dance with altitude, latitude, mean annual precipitation and a negative relationship with mean annual temperature (Tab. 1). These results also show a high correlation between abundance and diversity (r = 0.51 p< 0.0001). No relationships were found between diver- 238 A. BArRBosA and J. BENZAL Table 1. Simple and partial correlations between abundance, diversity and climatic and geographical variables. MAT = Mean annual temperature, MAP = Mean annual precipitation. ** P< 0.01, #=* P < 0.001, n.s. = not significant Altitude Latitude Longitude MAT MAP Simple correlations Abundance 031875 0.42*** Ss: 2080 = 0.35*** Diversity 0397 0582 sS _ (),49*** 0.37%** Abundance as: 1 Diversity 0502 sity or abundance and longitude. The importance of altitude and mean annual tempera- ture disappears when the relationship among variables is analysed using of partial correla- tion. Only latitude correlates with both diversity and abundance. Mean annual precipitation correlates only with abundance (Tab. 1). Latitude was the first variable se- lected by the stepwise multiple regression analysis performed on the relationship between abundance, latitude and mean annual precipitation (r=0.19p< 0.01). Abundance shows a strong decrease from the north to the south of Spain (27 species and 15 species respectively; CoRBET and OVENDEN 1980); data from AULAGNIER and THE- VENOT (1986) showed an abundance of 14 species in north Morocco. Discussion Geography (latitude) appears to be the main factor explaining the variation in abundance and diversity of small mammals on the Iberian Peninsula. Also, mean annual precipita- tion seems to be more important than temperature. We earlier calculated that abundance of most small mammalian species in Iberia depends on variations in precipitation more than on variation in temperature. The absence of a relationship between both abundance and diversity, and temperature are contradictory with results obtained for several other organisms. TURNER et al. (1987) found temperature and sunshine to be closely related with species diversity of butterflies. CurRIE (1991) concluded that one of the three stron- gest correlates of species abundance among birds, mammals, amphibians and reptiles, was mean annual temperature. However, in a study of breeding birds on the Iberian Peninsu- la, TELLERIA et al. (1992) did not assure any relationship between abundance and mean annual temperature, similar to the results obtained here. Potential causal factors could be related to scale and geographic situation of the selected area in the theoretical gradient of small mammal abundance distribution in the Palearctic region (TELLERIA et al. 1992). TiL- MAN (1982) pointed out that abundance is distributed along productivity gradients accord- ing to curves characterized by ascendant, peak and descendant phases along a range of low to high levels of resource availability. TELLERIA et al. (1992) stated according to TIL- MAN’s (1982) model that Iberian climatic gradient is too small to produce significant changes in abundance of forest birds, making it necessary to examine the abundance pat- tern on a broader, or more contrasting, geographical scale. This does not seem to be the case for small mammals, since results at broader scale (HERRERA 1974) or more constrain- ing geographical scale (MorEno and BarBosA 1992; BARBoSsA et al. 1992) show the same pattern as this peninsular study (but see RosEnzwEiG 1992 for north American rodents). Diversity and abundance of small mammals in Iberia 239 This reinforces the importance of latitudinal changes other than temperature for explain- ing changes in diversity and abundance (Ronpe 1992). Latitudinal gradients in abundance and diversity have been known for over a century (WALLACE 1878), with species numbers decreasing as latitude increases (STEHLI 1968; McCoy and Connor 1980; CLARK and Bunck 1991). They have been shown to occur in many different organisms and habitats (RoHnDE 1992). Our results, however, show an in- crease in small mammal abundance, as well as in diversity, as latitude increases (HERRERA 1974; MorENo and BARBoSA 1992). Two reasons could explain this apparent contradiction to the general pattern found on a larger scale (from the Equator to the Poles); 1) a very scanty small mammal fauna is usually associated with Mediterranean environments (HER- RERA and HırALDO 1976; WARBURG et al. 1978; Copy et al. 1983); 2) the peninsular effect (Sımpson 1964; MCARTHUR and Wırson 1967; SEıB 1980; GıLpım 1981) influences the spe- cies distribution. Comparison of species abundance shows a strong decrease from north to south in Spain, but the abundance is almost the same in both south Spain and north Morocco. If a peninsular effect exists, then the number of species should decrease from north to south in Spain, however it should increase on the African side of the Strait of Gibraltar. The present results did not support this hypothesis. This suggests that the change in the small mammal fauna from northern to southern Spain is not related to the peninsular effect. Other authors (TAyLor and REGAL 1978; WAMER 1978; SEIB 1980; LAawLor 1983; BUSACK and HEDGEs 1984; MEANS and SIMBERLOFF 1987) did not find a peninsular effect in other peninsulas (Florida and Baja California) which have a Mediterranean environment. They conclude that southern depauperation is most likely caused by habitat insufficiency in the south. On the other hand, several authors (Harrıs 1952; BATzLı 1968; DELANY 1981) have pointed out that differences in the diversity of a small mammal community are primarily due to habitat structure complexity. On the Iberian Peninsula there is an increase in habi- tat structure complexity from south to north (PEINADo and Rıvas-MARTINEZ 1987). These statements reinforce the hypothesis that Mediterranean environments per se are less sui- table to the small mammalian community considered as a whole. Acknowledgements We thank Data Bank of the Instituto Nacional de Meteorologia for supplying climatic data. E. MORENO, M. ALCANTARA and an anonymous referee made suggestions that improved the manuscript greatly. M. GOMENDIO corrected the English. Zusammenfassung Diversität und Abundanz bei Kleinsäugern in Spanien: Halbinsel-Effekt oder Lebensraumanpassung? Untersucht wurde der Einfluß geographischer und klimatischer Faktoren (geographische Breite und Länge, Höhe, jährlicher Durchschnitt der Niederschläge und Jahresmitteltemperatur) auf die Diversität und Abundanz der iberischen Kleinsäugerfauna. Hierzu wurden 92 983 Kleinsäugerfunde aus Schleier- eulen-Gewöllen analysiert. Die Daten wurden sowohl aus der Literatur, wie auch von den Untersu- chern an 183 iberischen Orten gewonnen. Irgendeine Beziehung zwischen der Höhe über dem Meer und der Jahresdurchschnitts-Temperatur konnte nicht ermittelt werden. Eine positive Korrelation be- steht zwischen der Diversität, der Abundanz, der geographischen Breite und dem mittleren jährlichen Niederschlag. Dieser Trend wird im Zusammenhang mit der Hypothese des „Halbinsel-Effektes“ be- handelt. Die Ergebnisse stützen diese Hypothese nicht. Unterschiede zwischen Südspanien und Nord- marokko in der Abundanz an Kleinsäugern konnten nicht gefunden werden. Vermutlich ist die Verteilung sowohl der Artenhäufigkeit, wie auch der Artenmannigfaltigkeit, parallel zu den Breiten- kreisen das Resultat der ungünstigen Bedingungen für Kleinsäuger in mediterranen Lebensräumen. 240 A. BAaRBosA and J. BENZAL References ALCANTARA, M. (1989): Anälisis de la distribuciön altitudinal de la fauna de micromamiferos de la Sier- ra de Guadarrama. Acta Biol. Mont. 9, 85-92. ÄULAGNIER, S.; THEVENOT, M. (1986): Catalogue des mammiferes sauvages du Maroc. Travaux de l’insti- tut scientifique. Serie Zoologique 4. BARBOSA, A.; LOPEZ-SANCHEZ, M. J.; NIEvA, A. (1992): The importance of geographical variation in the diet of Tyto alba Scopoli in central Spain. Global Ecology and Biogeography Letters 2, 75-81. BAtzuı, G. ©. (1968): Dispersion patterns of mice in California annual grassland. J. Mammology 49, 239-250. BENZAL, J. (1982): Contribuciön al conocimiento de los micromamiferos (Insectivoros y Roedores) del norte de Burgos. Sistemätica, taxonomia y zoogeografia. Msc. Thesis. Universidad Complutense de Madrid. BOND, W.; FERGUSON, M.; FORSYTH, G. (1980): Small mammals and habitat structure along altitudinal gradients in the Southern Cape Mountains. S. Afr. J. Zool. 15, 34-43. BRUNET-LECOMTE, P.; DELIBES, M. (1984): Alimentaciön de la Lechuza Comün Tyto alba en la cuenca del Duero. Donana Acta Vertebrata 11, 213-229. BRÜNNER, H.; NEET, C. R. (1991): A parapatric scenery: the distribution and ecology of Sorex araneus and Sorex coronatus (Insectivora, Soricidae) in southwestern Germany. Z. Säugetierkunde 56, 1-9. BUSAcK, S. D.; HEDGES, S. B. (1984): Is the peninsular effect a red herring? Am. Nat. 123, 266-275. CLARCK, D. R.; Bunck, C. M. (1991): Trends in North American small mammals found in common barn- owl Tyro alba dietary studies. Can. J. Zool. 69, 3093-3102. Copy, M.L.; BREYTENBACK, G.)J.; Fox, B.; NEWSOME, A.E.; Quinn, R.D.; SIEGFRIED, W.R. (1983): Mineral nutrients in Mediterranean ecosystems. S. Afr. Nat. Scient. Pogr. Report. 71, 91-114. CoRBET, G.; OVENDEN, D. (1980): The mammals of Britain and Europe. London: Collins. CURRIE, D. J. (1991): Energy and large-scale patterns of animal and plant species richness. Am. Nat. 137, 27-49. DELIBES, J. (1985): Distribution and abundance of small mammals in a gradient of altitude. Acta Zool. Fennica 173, 53-56. DELANY, M. J. (1981): Ecologia de los micromamiferos. Barcelona: Omega. Euias, F.; Ruız, L. (1977): Agroclimatologia de Espana. Madrid. Publ. del Ministero de Agricultura. GiLpIn, M. E. (1981): Peninsular diversity patterns. Am. Nat. 118, 291-296. HaRrRıs, V. T. (1952): An experimental study of habitat selection by prairie and forest races of the deer mouse Peromyscus maniculatus. Contrib. Lab. Vert. Biol. Univ. Michigan 56, 1-53. HERRERA, C. M. (1974): Trophic diversity of the Barn owl Tyto alba in continental Europe. Ornis Scandi- navica 5, 181-191. HERRERA, C. M.; HIRALDo, F. (1976): Food-niche and trophic relationships among European owls. Ornis Scandinavica 7, 29-41. KiESTER, A. R. (1971): Species density of North American amphibians and reptiles. Syst. Zool. 20, 127-137. LAwıor, T.E. (1983): The peninsular effect in mammalian species diversity in Baja California. Am. Nat. 121, 432-439. MCARTHUR, R. H.; WırLson, E. O. (1967): The theory of Island Biogeography. Princeton Univ. Press. McCoy, E. D.; Connor, E.F. (1980): Latitudinal gradients in the species diversity of North American mammals. Evolution 34, 193-203. MEANS, D. B.; SIMBERLOFF, D. (1987): The peninsula effect: habitat-correlated species decline in Flori- da’s herpetofauna. J. Biogeography 14, 551-558. MORENO, E.; BARBOSA, A. (1992): Distribution patterns of small mammal fauna along gradients of lati- tude and altitude in Northern Spain. Z. Säugetierkunde 57, 169-175. NETER, J.; WASSERMAN, W. (1974): Applied linear statistical models. Homewood, IL: Richard D. Irwin, Inc. PEINADO, M.; Rıvas-MARTINEZ, S. (1987): La vegetaciön de Espaha. Madrid. Publ. Univ. Alcalä de He- nares. PIANKA, E. R. (1966): Latitudinal gradients in species diversity: A review of concepts. Am. Nat. 100, 33- 46. ROHDE, K. (1992): Latitudinal gradients in species diversity: the search for the primary cause. Oikos 65, 514-527. ROSENZWEIG, M.L. (1992): Species diversity gradients: we know more and less than we thought. J. Mammalogy 73, 715-730. Diversity and abundance of small mammals in Iberia 241 SAINT-GIRONS, M. C.; Spitz, E. C. (1966): A propos de l’etude des micromammiferes par l’analyse des pelotes des rapaces: Interet et limitations de la methode. Terre et Vie 1, 3-18. SEIB, R. L. (1980): Baja California: a peninsula for rodents but not for reptiles. Am. Nat. 115, 613-620. Sımpson, G. G. (1964): Species density of North American recent mammals. Syst. Zool. 12, 57-73. SoKAL, R. R.; ROHLF, F. J. (1981): Biometry. San Francisco: Freeman. STEHLI, F. G. (1968): Taxonomie diversity gradients in pole location: the recent model. In: Evolution and Environment. Ed. by E. T. DRAKE. New Haven: Yale University Press. Pp. 163-227. TAYLOR, R.J.; REGAL, P. J. (1978): The peninsular effect on species diversity and the biogeography of Baja California. Am. Nat. 112, 583-593. TELLERIA, J. L.; SANTOS, T.; SANCHEZ, A.; GALARZA, A. (1992): Habitat structure predicts bird diversity distribution in Iberian forest better than climate. Bird Study 39, 63-68. TILMAN, D. (1982): Resource competition and community structure. Princeton NJ, Univ. Press. TURNER, J. R.; GATEHOUSE, C. M.; CoREY, C. A. (1987): Does solar energy control organic diversity? But- terflies, moths and British climate. Oikos 48, 195-205. WALLACE, A.R. (1878): Tropical nature and other essays. London. MacMillan. WAMER, N. ©. (1978): Avian diversity and habitat in Florida: an analysis of a peninsular diversity gradi- ent. M. S. Thesis, Thallahassee: Florida State University. WARBURG, M.R.; BEN-HorIn, A.; RANKEVICH, D. (1978): Rodent species diversity in mesic and xeric habitats in the Mediterranean region of North Israel. J. Arid Environment 1, 63-69. ZAR, J. H. (1984): Biostatistical analysis. 2nd ed. Englewood Cliffs, NJ: Prentice Hall. Authors’ address: ANDRES BARBosA and JESUS BENZAL, Departamento de Ecologia Evolutiva, Museo Nacional de Ciencias Naturales, CSIC, C/Jose Gutierrez Abascal, 2, E-28006 Madrid, Spain. Z. Säugetierkunde 61 (1996) 242-252 ZEIESECH! RIFTL © 1996 Gustav Fischer, Jena SÄUG ETl ERKÜR N DE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Kleinsäuger aus dem Pleistozän und Holozän Südost-Europas und des Mittleren Ostens Von D. JANOSSY Paläontologische Abteilung, Ungarisches Naturhistorisches Museum, Budapest Eingang des Ms. 25. 01. 1995 Annahme des Ms. 29. 02. 1996 Abstract Small mammals from the Pleistocene and Holocene of Southeast-Europe and the Middle East According to new data of the last decades, the Pleistocene and Holocene evolution and the migration of most small mammal genera of adjacent territories in Europe are compiled. Most new data on the faunal successions refer to the Russian-Ukrainian Plain and the Levant, while new informations about the events in the Balkans and in Asia Minor are scarce. The above mentioned data are elaborated for about 40 genera of small mammals in our territory. Einleitung Vor mehr als 35 Jahren habe ich unsere damaligen Kenntnisse über die Entwicklung der Kleinsäugerfauna Europas zusammengestellt (JAnossy 1961), und diese Daten wur- den in der Literatur weitgehend als grundlegend betrachtet (WoLDsTEeprt 1969; KAHLKE 1981). Erhebliche Erweiterung des Wissensstandes auf diesem Gebiet der Kleinsäuger- forschung erfolgte jedoch in den letzten Jahrzehnten und hauptsächlich in den letzten Jahren. Viele dieser neueren Daten liegen aus Europa vor. Von diesen werden in dieser Stu- die in erster Linie die südwestlichen Teile der russisch-ukrainischen Ebene und die Krim- Halbinsel berücksichtigt. Daraus resultiert, daß das heutige Europa eigentlich eine Halb- insel des asiatischen Kontinentes ist („Subkontinent“). In der Zusammenstellung der vor- liegenden Studie habe ich aufgrund der Walterschen Vegetationskarte als Gebiete in erster Linie den westlichen Teil des pontischen geographischen Elementes (mit heute 88% europäischen Formen), den Westen des iranischen Gebietes (75% europäisch) und des sindischen Gebietes (eigentlich nur levantisches Gebiet, mit 60% europäischen For- men) hervorgehoben. Kleinasien, die Levante und der Balkan sind noch sehr sporadisch dokumentiert. Die Iberische Halbinsel habe ich wegen der außerordentlich hohen Zahl von Endemismen bei den Vergleichen nicht berücksichtigt. In den vergleichenden Tabellen habe ich unter großem Vorbehalt die üblichen Faunenetappen der russischen Ebene mit jenen des mittleren und westlichen Europas bio- chronologisch (also zeitlich) zu korrelieren versucht: Jungpleistozän — Werchnepaleolititscheskj Komplex Mittelpleistozän — Syngilskij und Chosarskij Komplex Altpleistozän — Tiraspoljskij und Tamanskoj Komplex. Kleinsäuger aus dem Pleistozän und Holozän 243 Ich habe versucht, die mitteleuropäischen Daten mit jenen der russisch-ukrainischen Tief- ebene in tabellarischer Form zu vergleichen. Dabei habe ich mich hauptsächlich auf die Daten von W. I. Gromov (1948), I. M. Gromov (1981), TCHERNOV (1988), RABEDER (1981) und ZAZHIGIn (1982), sowie auf viele andere Literaturangaben gestützt. Kollege REKO- vETS (Kiew) hat die Tabellen in liebenswürdiger Weise gründlich revidiert, wofür ich auch an dieser Stelle meinen besten Dank aussprechen möchte. Ergebnisse und Diskussion Die hier zur Diskussion stehenden etwa 40 Gattungen wurden in systematischer Reihen- folge zusammengestellt. Es sollen zuerst die Wandlungen bei Insektenfressern erörtert werden. Maulwürfe waren keineswegs so allgemein im Pleistozän der osteuropäischen Tiefebene verbreitet wie im mittleren und westlichen Europa. Talpa fehlt in den meisten Faunenlisten des ukrainisch-russischen Pleistozäns. Einige Funde (z.B. von Chadzibej, Ilowai, Tiligul; To- PACHEVSKIJ und SKORIK 1977) sprechen aber dafür, daß Maulwürfe wegen klimatischer und Bodenbedingungen sporadisch vorhanden waren oder sich auf die europäische Halbinsel zurückgezogen haben. Aus Kleinasien haben wir praktisch auch keine Belege. Die aus- schließlich wasserbewohnenden Desmane waren durchgehend im Pleistozän vorhanden, mit vergrößerten Köperdimensionen, wie auch im übrigen Europa und in Kleinasien (To- BIEN 1981) Galemys kormosi, Desmana nehringi-thermalis-moschata (WANGENGEIM 1982). Unter den kleinwüchsigeren Spitzmäusen soll in erster Linie die Gattung Crocidura, mit tropisch-subtropischem Verbreitungsschwerpunkt erwähnt werden. Sie erscheint viel- leicht zuerst im Pliozän (AGADJANIAN und Kowarskı 1978), aber eindeutig im älteren Pleistozän (z.B. in Taman, Krim). Sie ist aber keineswegs ein so beständiges Element in „interglazialen“ Faunen dieses Gebietes wie im mittleren und westlichen Europa oder gar im Mittleren Osten, wo sie vom älteren Pleistozän an (Ubeydiya) ständig vorhanden ist. Die neueren Untersuchungen bestätigen viel ältere Wurzeln der Gattung Sorex in Europa als früher angenommen wurde (ab dem unteren Pliozän häufig, z.B. REUMER 1983). Ei- nige Relikt-Formen (z.B. Beremendia, Petenyia) starben dabei nach unseren heutigen Kenntnissen in der russischen Tiefebene etwas früher aus als im übrigen Europa. Sie feh- len bis jetzt in den pleistozänen Faunen Kleinasiens und des Nahen Osten. Übergehend zu den Nagern soll zuerst folgendes erwähnt werden: Eine der größten Überraschungen der vergangenen Jahrzehnte ist das pleistozäne Auftreten eines Mit- gliedes der bis jetzt ausschließlich auf das Tertiär beschränkten Familie Eomyidae: Estra- momys simplex (JAnossy 1969). Die Art ist bis jetzt aus vier ungarischen und sechs südukrainisch-russischen (etwa 20° 30° — 40° 30 östliche Breite, 45°-47° nördlicher Länge: Kotlavina, Odessa und Umgebung; Kryzanovka, Zhevakhova Gora; Tiligul-Liman, Obu- khovka 1-2; unterer Don; brieflich A. TEsaKov) etwa ältest- bis altpleistozänen Fundstel- len bekannt, was für eine weite geographische Verbreitung spricht. Der älteste Fundort ist das altpliozäne Podlesice in Süd-Polen (19° 32’ östl. Br., 50° 34’ nördl. L.). Unseren der- zeitigen Kenntnissen gemäß war Estramomys ein östliches Element in Europa. Zu den Hörnchen kann folgendes gesagt werden. Funde von Eichhörnchen (Sciurus spp.) gelten als sporadisch im Pleistozän von England bis zur Ukraine. Ein Fund verdient allerdings unser besonderes Interesse. Das ist der gut stratifizierte Überrest aus einer alt- pleistozänen Flußterrasse des Ilowai (etwa 40° östl. Br. und 53° nördl. L., im Hangenden von Kryoturbierten Schichten), zusammen mit Mimomys savini gefunden (AGADJANIAN 1981). Er beweist, daß durch Fauna definierbare „interglaziale“ Ablagerungen auch auf der russischen Ebene existierten, die meisten wurden aber durch die Tätigkeit späterer Moränen vernichtet. Im gemäßigten Europa bedeutet das endgültige Erscheinen des Eichhörnchens (Sciurus vulgaris) den Beginn „unseres“ Interglazials, des Holozäns (Na- 244 D. JAnossy dachowski 1989). Die Eichhörnchenart Sciurus anomalus war vom mittleren Pleistozän bis zum Neolithikum in der Levante, von Griechenland und vom älteren Pleistozän bis heute in Kleinasien vorhanden (TCHERNOV 1988; STORCH 1988). Unsere Kenntnisse über die alt- bis jungpleistozänen Funde von Marmota vermehrten sich in den vergangenen Jahrzehnten beträchtlich. Wir haben in immer größerem Gebiet über die einstige Verbreitung bis zu den Pyrenäen, in ganz Mittel- und Osteuropa (das ganze russisch-ukrainische Tiefland) und im nördlichen Balkan neue Kenntnisse. Dabei scheint die eingehende Analyse der Bezahnung dafür zu sprechen, daß die zwei heutigen Arten, M.bobak und M. marmota, sich im mittleren Pleistozän voneinander auf unserem „subkontinent“ getrennt haben (CHALINE 1970). I. M. Gromov et al. (1965) haben versucht, die komplizierten Evolutionswege der Zie- sel kritisch zu deuten. Laut deren Hypothese lebte eine kleine Form (C. citelloides-citel- lus-xanthoprymnus) vom mittleren Pleistozän des östlichen und südlichen Europa bis nach Kleinasien bis heute. Die Zwillingsart dieses Taxons wäre C. severskensis des Jung- pleistozäns der russischen Ebene. Die altpleistozäne Art, C. (cf. Urocitellus) primigenius, — wahrscheinlich der Ahne von C. (U.) nogaici, — wurde auf russischem Gebiet nicht gefun- den. Citellus dietrichi dürfte der Vorfahr jener großen Form gewesen sein, die zweimal (?) im Mittel- und im Jungpleistozän England und S-Frankreich erreichte und unter dem Namen C. superciliosus und C. major beschrieben wurde. Ein interessantes Negativum soll an dieser Stelle erwähnt werden: Nach heutigem Kenntnisstand erreichten die Ziesel die Levante nicht. Die paläontologischen Daten über die Schläfer hat StorcH (1978) in Einzelheiten erörtert, so daß an dieser Stelle nur einige ergänzende Bemerkungen gemacht werden sol- len. Die Schläfer sind vorwiegend europäischen Ursprungs und einige Arten haben den Mittleren Osten nicht erreicht. Die Mausschläfer (Myomimus) sind nicht nur ständige Elemente der bis jetzt bekannt gewordenen pleistozänen Faunen Kleinasiens und des Na- hen Ostens, sondern finden sich sogar in den heute eher als Pliozän angesprochenen Schichten der russischen Tiefebene (AGADJANIAN 1977). Die Gliriden waren keineswegs so beständige Elemente der pleistozänen Faunen wie im mittleren Europa, sie fehlen me- ist völlig in den Faunenlisten über die pleistozänen Faunen der russisch-ukrainischen Ebene (TorAcEvskn 1965; REKOVETS 1985; MAur 1990). Hier soll nur noch der Glis-Fund von Ilowai erwähnt werden (Einzelheiten über die Fundstelle s. unter Sciurus), der mehr als tausend Kilometer östlich von den bis jetzt bekannten pleistozänen Funden des Siebenschläfers liegt. In den vergangenen Jahren sind wieder einmal (nicht publizierte) Funde des Gartenschläfers aus gut definierten Ablagerungen des Eneolithikums in Un- garn zutage gekommen (Füzesabony). Rätselhaft ist das fossile Vorkommen in Nord-Afri- ka und rezent in der Levante. Die neuere Revision der fossilen Blindmäuse (Spalacidae) (hauptsächlich KorDos 1988; TopacEvsKıJ 1969) hat erwiesen, daß die Verwandtschaftsbeziehungen nicht so ein- fach sind, wie früher gedacht wurde: Spalax stammt nicht unmittelbar von Prospalax ab, vielleicht haben sie gemeinsame Ahnen. Fossile und subfossile Funde aus Kleinasien be- weisen neuerdings, daß Blindmäuse schon vom älteren Pleistozän an bis heute in diesem Gebiet vorhanden waren (STorcH 1975, 1978, 1988). | An dieser Stelle sollen einige Gedanken über die asiatischen Springmäuse mitgeteilt werden. In dem mittelasiatischen, seit dem klassischen Miozän (Sarmat) mehr oder weni- ger gesperrten Turanischen Tiefland war die Situation ähnlich wie auf einer Insel im Ozean. Etwa 25 Arten von rezenten Dipodiden wurden beschrieben und wenigstens eben- soviele fossile und rezente andersrangige Taxa (Genera, Subspecies) wurden aufgestellt. Die Dipodiden erreichten schon im Miozän das westliche Marokko (Protalactaga moghre- biensis JAEGER 1975) und später, während des Pleistozäns wiederum dasselbe Gebiet (Ja- culus cf. jaculus). Verschiedene Gattungen mit unterschiedlichen ökologischen An- sprüchen haben sich in den südlichen Teilen der russischen Ebene, meist nur bis zum Kleinsäuger aus dem Pleistozän und Holozän IN SISELE | VORA REOZD EINER DAE un! > E ö & [@) ® > leg} 58 55 |: So) =: = ASTTTERSELINN ASSSTSSTLTTTTATSEAN = SSTRÄTTELTETTTTIN [NNRRRRNTRÄNURAAANS NSTSSSESETSERESTN JUNG- PLEISTOZÄN MITTEL- PLEISTOZÄN RÄLTTRSRSTTTERSEESESERISSSSESSESSETSSREESEERTERTSSESS SE SS SSSSSSESSESRUSSSRNNSENTSRURUNRRUNNIS ALT- PLEISTOZÄN 1. Talpa 7. Petenyia 13. Marmota 2. Desmana nehringi Gr 8. Estramomys 14. Muscardinus 3. D. moschata Gr 9. Sciurus 15. Glis 4. Crocidura 10. Citellus primigenius - nogaicı 16. Dryomys 53. Sorex 11. C. superciliosus-major 17. Eliomys 6. Beremendia 12. C. citelloides-citellus 18. Spalax IE SPALACIDAE- DIPODIDAE \S ALITLTETTTTTETASLTERLTERSRTETESTETESETSETSNSE | | 5) 2 9 a 2 2 TTTETLSANAÄSTT RAUS \SÄILTTTTTTII ESSTTTTTITTIN 19. Allactaga 20. Allacıagulus 21. Dipus 22. Sicista 245 Abb. 1. Skizze der zeitlichen und räumlichen Verbreitung der Kleinsäuger-Gattungen im Pleistozän und Holozän des Mittleren und des Östlichen Europas. I. Dicke Linie: Vorhandensein in Mitteleuropa bestätigt; Schräg schraffiert: im Östlichen Europa bestätigt; Strichellinie: keine Funde aber wahrschein- lich in der angegebenen Zeitspanne im Gebiete vorhanden oder Vorkommen unsicher. Gr. = Gruppe. RODER IN A ee CASTORIDAE |, | u HOLOZAN | 8 [ JUNG- | PLEISTOZÄN MITTEL- PLEISTOZÄN IWURUAULUAAULLALUNAUANANUMDNAANAAN! CSSSTSTSSTSTTTTTTSSSTSSTTSSTSSSTSSSSTTSSSESESTET T STSSSTSSSTSELSTEN ALT- - PLEISTOZAN | E E . Castor LAGO- MORPHA KLITTTISTILSSLTTSSTSESTTSSESETTSTSTTESETTSTTESTSSELSTSESTTTTTSTTTETTS TEN 1 13. Mimomys 19. Pitymys 25. Ochotona 2. Trogontherium 8. Cricerus 14. Pliomys 20. Prolagurus 26. Hypolagus 3. Hystrix major Gr 9. Allocricetus 15. Clethrionomys 21. Lagurus 27. Lepus 4. Hystrix vinogradovi Gr. 10. Cricetulus 16. Arvicola 22. Eolagurus 5. Apodemus 11. Meriones 17. Allophaiomys 23. Lemmus 6. Mus 12. Ellobius 18. Mierotus 24. Dicrostonyx Abb. 2. Skizze der zeitlichen und räumlichen Verbreitung der Kleinsäuger-Gattungen im Pleistozän und Holozän des Mittleren und des Östlichen Europas. II. 246 D. JAnossy Dnjepr, eingebürgert (Alactagulus, Dipus, Scirtopoda usw.). Der Pferdespringer (Allactaga jaculus Pallas nec Linne) hat zweimal im Jungpleistozän nach Westen vorrückend den Rhein erreicht, und er wurde neuerlich bis zum mittleren Balkan in jungpleistozänen Schichten vorgefunden (etwa 25 °30’ östl. Br., 43° nördl. L., von den nördlichen Ausläu- fern des Balkan-Gebirges, Bacho Kiro-Höhle, KowArskı und NADACHowSsKI 1982). Im Na- hen Osten und in Kleinasien kommt vom älteren Pleistozän an bis heute eine endemische Form (Allactaga euphratica) vor. Neue Daten bereichern unsere Kenntnisse über die zeitliche und räumliche Verbrei- tung der Streifenmäuse (Sicista). Im Süden der russischen Ebene und im Süden von West- Sibirien kommen sie regelmäßig vor (TopacEvskıs 1965; TOPACEVSKIJ und SKORIK 1977), wobei sie aber im Jungpleistozän desselben Gebietes praktisch völlig fehlen (REKOVETS 1985). Der südlichste, zeitlich unsicher korrelierbare Fund (Mittelpleistozän?) stammt vom westlichsten Kleinasien (Chios, STORCH 1975). Über die ausgestorbene Bibergattung Trogontherium wurden in den vergangenen Jah- ren intensive morphometrische und geographische Studien publiziert (MAı 1979; MAYHEW 1978). Es wurde eine allmähliche Vergrößerung einiger Zähne statistisch analysiert. Die- ser früher als ausschließlich nördlich-eurasiatisch angesehene Säuger wurden in weiteren Gebieten gefunden (etwa 200 km S von Peking). Unter der Bezeichnung Trogontherium minus wurde das dominante Vorkommen in altpleistozänen Schichten im westlichen Ana- tolien (etwa 38° 39’ nördl. L. und etwa 30° östl. Br., ToßıEn 1981) festgestellt. Ein außerordentlich glücklicher Fund - der übrigens so seltenen — Stachelschweine (Hystrix) im Altpleistozän Nord-Ungarns hat bewiesen, daß die vom Jungpleistozän be- schriebene kleine Form, Hystrix vinogradovi, schon damals eine selbständige Evolutions- linie vertrat und mit der großen Hystrix etrusca-major Gruppe zusammenlebte (JAnossY 1972): Für die Mäuse (Muridae) im engeren Sinne kann folgendes festgestellt werden: In den heute schon durch eine große Zahl von Fundstellen bekannt gewordenen Faunen der rus- sischen Ebene sind Apodemus und Mus im Mittel-Pleistozän der südlicheren Teile vorhan- den. Aus den kaltkontinentalen Schichten des Jungpleistozäns fehlen sie vollständig. Nach unseren bisherigen Kenntnissen lebte Mus von älteren Pleistozän des Nahen Ostens (Übey- diya) und in Kleinasien (Kalymnos, STorcH 1978) bis heute in diesem Gebiet. In den vergan- genen Jahrzehnten stellte sich heraus, daß Mus im jüngeren Mittelpleistozän ein ständiges Element der Faunen Ungarns war (JAnossy 1986). Im Westen Marokkos wurden auch pleis- tozäne Mus-Reste gefunden (JAEGER 1969). Eine offene Frage ist noch heute die Herkunft der bis jetzt als in historischer Zeit nach Europa eingeschleppt betrachteten Ratten (Rattus). Im alten bis jüngeren Pleistozän Asiens schon längst bekannt, kommt Rattus vom oberen Jungpleistozän (Kebaran + 17.000 Jahre alt, TCHERNOV 1988) der Levante vor. Stratigra- phisch eigenartig und alleinstehend sind die Rattus-Funde aus dem Mittelpleistozän von Chios und von der Dobrudscha (RADULEScCo und Samson 1973; STORCH 1975; TERZEA 1973). Auf die wegen der großen Homogenität in der Morphologie der Bezahnung so schwer überschaubaren Hamster wirft eine große Menge von Funden der letzten Jahrzehnte ein ganz neues Licht. Schon FAHLBUSCH (1969) hat versucht, die phylogenetischen Zusam- menhänge mit rein paläontologischer Interpretation darzustellen. Nach dieser Studie stammen die altweltlichen Hamster aus den Cricetodontiden des älteren Miozän, und Cri- cetus (siehe auch FREUDENTHAL und KorDos 1989) sowie ALLOCRICETUS dürften euro- päischen Ursprungs sein. Nach dieser Zusammenstellung stammen die kleinen Formen (Cricetulus, Phodopus) aus Asien. Das Pleistozän betreffend kann festgestellt werden, daß eine sehr große Menge von Daten über die europäische Verbreitung der Hamster vorliegt (neuerlich auch von Süd-Polen, NADAcHowskı 1985, 1989). Demgegenüber ken- nen wir bis jetzt aus einer „alt-ruscinischen“ Fundstelle von Kleinasien (Maritsa, Rhodos) einen Endemiten: Cricetus lophidens Bruijn, Dawson und Mein 1970, und aus dem älter- en Pleistozän von Ubeydiya wird ein kleiner Cricetus cricetus (nanus?) angegeben Kleinsäuger aus dem Pleistozän und Holozän 247 (TCHERNoV 1986, 1988). Der „große“ Hamster fehlt übrigens nicht nur im Pleistozän von ganz Kleinasien, sondern auch im Nahen Osten und im nördlichen Afrika. Ohne sich in systematisch-nomenklatorische Fragen zu vertiefen, soll an dieser Stelle vermerkt werden, daß eine mittelgroße Hamsterform im Alt- bis Jungpleistozän aus West-, Süd- bis Mitte- leuropa als Allocricetus bursae beschrieben wurde. Neben dieser, nur für Spezialisten in der Zahnmorphologie unterscheidbaren Form, wurden Hamster unter den Namen Meso- cricetus newtoni, oder brandti, oder auratus von der Balkan-Halbinsel, aus Kleinasien und von der Levante bekannt. Zu diesen Formen gesellt sich der noch kleinere Cricetulus mi- gratorius, der heute in der Levante die einzige Art unter den Zwerghamstern ist (TCHER- Nov 1988). Den speziellen ökologischen Ansprüchen gemäß, hat die zu den Rennmäusen ge- stellte Meriones nur die südlicheren Teile unseres Gebietes erobert. Nach neueren Daten (Kuss und STorcH 1978) im westlichen Anatolien schon im älteren Pleistozän vorhanden und sporadisch bis heute verbreitet (FELTEn et al. 1971), war Meriones auch schon im äl- teren Pleistozän im nördlichen Afrika vorhanden (JAEGER 1969). Auch im älteren Pleisto- zän der Levante erscheinen die Rennmäuse, und sie leben dort noch heute (TCHERNOV 1986). In der russischen Ebene fehlen diese Mäuse aus klimatisch-ökologischen Gründen vollständig, der einzige bis jetzt bekannte Fund stammt vom Kaukasus-Gebiet (vom mitt- leren Pleistozän der Umgebung von Baku, Binagady, V. I. Gromov 1948). Das dürfte mit der Tatsache gedeutet werden, daß die Rennmäuse keinen Winterschlaf halten und an aride klimatische Verhältnisse angepaßt sind. Zuletzt sollen die neueren Daten über fossile Wühlmäuse in Betracht gezogen wer- den. Die große Mannigfaltigkeit verschiedener Taxa, wo ein meist unübersichtliches Ge- flecht phylogenetischer Reihen und Konvergenzen wahrzunehmen ist, macht die Orien- tierung sehr schwierig. Trotz dieser Probleme wurde versucht, die verwandtschaftlichen Beziehungen wenigstens skizzenhaft herauszufinden (z.B. KrErtzoı 1969; I. M. GrOMoV und PoLJakov 1977). GroMmov und PorJaXov (1977) haben sogar einen „klassischen“ Stammbaum aufgestellt, bei dem die Arvicoliden in sieben Tribus unterteilt werden. Es stellt sich heraus, daß bei einigen primitiven Formen nicht zu entscheiden ist, ob sie Crice- tiden odr Arvicoliden sind (Baranomys, Pannonicola) und neben „superarvicoliden“ Gat- tungen (z.B. /Ischymomys) noch absolut „cricetine“ Genera auftraten (Microtoscoptes). Die frühere Hypothese, die ältesten Vorfahren der Wühlmäuse im Oberen Miozän (Turolium) des mittleren Asien (Mongolei) und im ähnlichalten Nordamerika (Hemphil- lian) zu suchen, wurde durch neue Funde nur verstärkt. Die geologisch nahestehenden europäischen Funde (Microtocricetus FAHLBUSCH und Mayr 1975) können nur als eine Sackgasse der Evolution betrachtet werden. Die Übersicht soll nun mit Ellobius weitergeführt werden. Die Moll- oder Mull-Lem- minge sind heutzutage echte mittelasiatische Faunenelemente. Wenn sie in immer neuen Fundstellen der russischen Ebene vom älteren Pleistozän an — wenn auch immer spora- disch — nicht die heutige Westgrenze von Ellobius talpinus überschritten haben, so ist die übrige bis jetzt bekannte Verbreitung doch recht eigenartig. Sie erreichten schon im älte- ren Pleistozän NW-Afrika (Marokko, Algerien, JAEGER 1969, 1970), und wir kennen Reste von Ellobius nur vom Jungpleistozän der Levante. Früher wurde angenommen, daß Mull- Lemminge nie in die mittleren Teile Europas eingewandert wären, seitdem aber GROMOV und BARANoVoJ (1981) Ellobius mit Ungaromys, Germanomys und Betfiomys synonymi- siert haben, dürfte dieses Problem wieder revidiert werden. Endlich betrachten wir die typischen Micromammalier des Pleistozäns, die Wühl- mäuse in engerem Sinne (Arvicolidae). Mit den wurzelzähnigen Arvicoliden beginnend, soll Promimomys als eine weit verbreitete ursprüngliche Form hervorgehoben werden. Die aus den Bergen des westlichen Balkans bekannten „Bergmäuse“ figurieren in man- chen Bestimmungsbüchern noch immer als Dolomys milleri und nicht als Dinaromys bog- danovi (MARTINo, 1921), was den paläontologischen Daten nicht entspricht. Neue Funde 248 D. JAnossy aus Nord-Italien (D. allegranzii SaLA 1994) und aus der westlichen Ukraine (Tiligul, D. to- patschevskii Nesın und SKORIK 1989), sprechen für die weite Verbreitung dieser Form we- nigstens im älteren Pleistozän. Die explosionsartige Evolution der Gattung Mimomys (bis jetzt mehr als 40 Arten beschrieben!) macht dieses Genus - verbreitet im älteren Pleistozän der ganzen Holarktis (außer Nord-Afrika und dem Nahen Osten, aber bis zum mittelasiatischen Gebiet) — für stratigraphische Erwägungen außrordentlich geeignet. Die diesbezügliche Literatur ist heute kaum noch zu überschauen, und selbst die hauptsächlich in der russischen Ebene so weit verbreitete letzte zu Arvicola führende Art, Mimomys savini, charakterisieren mehr als hundert alt- bis mittelpleistozäne Vertebratenfaunen von England bis zum S-Ural (und wahrscheinlich noch weiter östlich). Ein glücklicher Fundkomplex aus Ungarn, wo sich ganze Schädel mit Mandibeln erhalten haben, machte endlich einige Evolutionslinien klarer (JAnossy und MEULEN 1975). Die Wurzeln von Clethrionomys gehen sehr tief in die Vergangenheit, ebenso wıe jene von Lemmus. Die aus einem Evolutionszweig der Mimomys-Gruppe stammenden Graulemminge oder Steppenlemminge (Lagurini) zeigen nach neueren statistischen Untersuchungen eine vielleicht noch schnellere Evolution als bei Mimomys, mit der Evolutionsreihe (überaus vereinfacht!): L pannonicus — L. transiens — L. lagurus. Sie sind im Pleistozän der rus- sischen Ebene - bei oft massenhaftem Vorkommen - weit verbreitet. Der große, heute als typisch mittelasiatische Form zu betrachtende, ZL. /uteus hat dabei westlich kaum die Linie des Prut bzw. des Dnjesters überschritten. Die kleineren Formen haben nach neueren Literatur-Daten wenigstens zwei Vorstöße nach Westen gehabt. Das erste Mal bis zum nordöstlichen Teil Spaniens (Bagur, Lagurodon pannonicus, LOPEZ-MARTINEZ et al. 1976), das zweite Mal im jüngeren Pleistozän westlich bis England, nördlich bis S-Polen und dem mittleren Deutschland, südlich bis Frankreich, bis zum Balkan und sogar bis zum westlichen Kleinasien (Insel Chios). Lagurus lagurus ist vom mittleren Pleistozän an ver- einzelt sogar bis zum mittleren Jungpleistozän vorzufinden (NADACHOWSKI hat unlängst ein Exemplar, „Mittel-Würm“, in alten Sammlungen von Ungarn, Höhle Iställoskö, gefun- den! SUTCLIFFE und Kowarskı 1976; STORCH 1975; CHALINE 1970). Endemische Formen im älteren Pleistozän des Mittleren Ostens hatten sich in solchem Grade spezialisiert, daß sie als selbständige Gattungen aufgefaßt werden (Jordanomys, Kalymnomys, KOENIGSWALD et al. 1992). Früher schien Allophaiomys, der bis heute für einen gemeinsamen Vorfahren von den wenigstens 50 Microtus (Pitymys)-Arten gehalten wird, auf das mittlere und südlichere Europa beschränkt zu sein. In den vergangenen Jahren wurde unser Wissen durch mehr als hundert Fundstellen vermehrt, die Verbreitung dehnte sich damit von W-Spanien durch das gesamte gemäßigte und mediterrane Europa, Mittel-Asien (TYuTkovA und KAr- pova 1994, SO-Kazachstan) bis Nordamerika aus (MARTIN, S-Dakota, mündl. Mittl.). Diese, in absoluter Chronologie 1,2-1,5 Millionen Jahre alte Periode wurde früher als Mit- telpleistozän, heute eher als Altpleistozän bezeichnet. Meiner Meinung nach ist die nach- folgende artenreiche holarktische Microtus-Fauna in dem gemäßigten Teil dieser Zone, auch in der russischen Ebene, polyphyletisch entstanden (Pitymys-Gruppe, M. arvalıs, agrestis, gregalis, oeconomus). Die Wühlmäuse von Kleinasien und der Levante sind dabei eintöniger, hier kommen seit dem Mittelpleistozän meist nur Microtus guentheri, socialis und irani, und als neue Einwanderer Microtus arvalis und Microtus nivalis in den Bergen vor. Das von SPITZENBERGER (1971) belegte Vorkommen der „Kaukasischen Wühlmaus“, Microtus gud (später teilweise als M.nivalis spitzenbergerae beschrieben), spricht dafür, daß diese letztgenannte Form einst weit verbreitet war (STORCH 1988). Die große morphologische Homogenität der Zähne der Lemminge (hauptsächlich die Genera Lemmus und Synaptomys) macht es verständlich, daß aus dem älteren Pleistozän ausgestorbene Taxa nur in den 70er Jahren beschrieben wurden (zusammenfassend Ko- wauskI 1977). Lemminge erreichten im älteren Pleistozän vereinzelt das mediterrane Ge- Kleinsäuger aus dem Pleistozän und Holozän 249 biet (die Pyrenäen und das süd-ungarische Villänyer Gebirge, 43° bzw. 46° nördl. Br., CHALINE 1976; JAnossy 1983). Ihr Vorkommen in größerer Zahl wird durch verschiedene Faunen des westlichen Europas und der russischen Ebene vom älteren Pleistozän an „zo- nal“ nördlich angedeutet, im osteuropäischen Gebiet meist zwischen den südlichen Gren- zen des Inlandeises der Riß- und des Würm-Glazials gelegen (AGADJANIAN 1977; JANOSSY 1986). Die Dokumentation der Wurzeln der Dicrostonychini verschiebt sich immer mehr nach Osten (Beringia). Noch einiges soll an dieser Stelle über die Schermaus bzw. Wasserwühlmaus (Genus Arvicola) gesagt werden. FEJFAR sowie KOENIGSWALD bemerkten zum ersten Mal, daß sich ähnlich wie bei Mimomys auch bei Arvicola die Evolution in den Änderungen der Schmelzdicke und -struktur ausprägt (KoENIGSWALD 1980). Es schien früher eine lineare Änderung zu sein, die auch für eine geologische Zeitbestimmung behilflich sein könnte (JAnossy 1976). Der Vorgang der Evolution ist aber in diesem Falle auch buschartig und lokal und kann nicht für das ganze Verbreitungsgebiet als gültig anerkannt werden. Man muß sogar vom älteren Pleistozän an mit Endemiten rechnen. So sollte z.B. „Arvicola jor- danica“ aus der Liste von UÜbeydiya gestrichen werden und aufgrund der Schmelzstruktur zu der großwüchsigen Microtus [Tibericola] jordanica (Haas, 1966) gerechnet werden. Die Pfeifhasen (Gattungen Ochotona, Proochotona, Prolagus, Ochotonoides) werden durch immer neuere Funde als östliche Elemente belegt, die im unteren Pliozän gerade dominant erscheinen (z.B. oberer Don, AGADJANIAN und KowALskI 1978). Sie waren stän- dige Elemente der russischen Ebene bis zum Jungpleistozän. Während wenigstens zweier pleistozäner Vorstöße nach Westen erreichten sie nicht nur Frankreich, sondern auch den mittleren Balkan (Arnissa, Macedonien, MAyHEw 1977) und die Dobrudscha (RADULESCO und TERZEA 1963). In Kleinasien wurden Pfeifhasen außer den erwähnten pliozänen Fun- den weder rezent noch fossil gefunden. Ein interessantes Fragezeichen der Tiergeogra- phie ist das geologisch kurze Erscheinen von Ochotona ım jüngeren Pleistozän der Levante (Acheulien, Oumm-Qatafa, TCHERNOV 1988). Die Verbreitung der westlichsten rezenten Ochotona-Arten erreicht dabei kaum die Wolga. Die in Mittel-Europa so regelmäßig registrierbare Ablösung von Aypolagus durch Le- pus ist in der russischen Ebene nicht so eindeutig erkennbar. Diese kurzgefaßte Übersicht über die Kleinsäuger Ost-Europas und des Mittleren Os- tens im Pleistozän macht uns auf folgende Tatsachen aufmerksam: 1. Die in geologischem Sinne schnelle Evolution einiger Kleinsäuger-Gruppen im pa- läarktischen Raum (Wühlmäuse) zeigt auch in diesem Gebiet, außer einigen Endemismen (Lagurini), ein ähnliches Bild wie im gemäßigten Europa, natürlich mit dem entsprechen- den tiergeographischen Kolorit. 2. Wir können also, wenigstens in großen Zügen, mit ähnlichen Faunenwellen rechnen (z.B. „Promimomys“ von der russischen Ebene bis W-Anatolien) wie in Europa. Nach Süden fortschreitend wird die faunistische Grenze zwischen Holozän und Jungpleistozän immer mehr verwischt. 3. Es wäre wünschenswert, in der Zukunft bei der Beurteilung der Faunenentwick- lung dieses Gebietes die Rolle geographischer Barrieren (Gebirge und aride Zonen) und der Dardanellen- und Bosporus-Brücke stärker in Betracht zu ziehen. 4. In der russischen Ebene dürften neben der Terrassenbildung und der klimatisch-zo- nalen Verteilung des Faunenbildes (z.B. Microtus, gegenüber Lemmus-Faunen) die seku- lären Hebungen der russischen Tafel während des Pleistozäns in höherem Maße in Betracht gezogen werden. Zusammenfassung Eine Kompilation der pleistozänen und holozänen Evolution der meisten Gattungen der Kleinsäuger der angrenzenden Gebiete Europas von Daten aus den letzten Jahrzehnten wurde erstellt. Die meisten 250 D. JAnossy Angaben stammen von der ukrainischen-russischen Ebene und aus dem Nahen Osten. Wenig Informa- tionen haben wir von diesen Geschehnissen auf der Balkanhalbinsel und in Kleinasien. Die erwähnten Angaben wurden für etwa 40 Gattungen zusammengestellt. Literatur AGADJANIAN, A.K. (1977): Quartäre Kleinsäuger aus der Russischen Ebene. Quartär 27/28, 111-145. — (1981): Mestonahozdenie melkich mlekopitauscie rannevo Pleistocena na reki Ilowai (Frühpleisto- zäne Fundstelle von Kleinsäugern am Fluß Ilowai). Novije dannie po stratigrafii i paleogeografii verchnego pliocena ı pleistocena centralnich rajonov Evropejskoj Casti SSSR. Inqua, Moskau. 1982, 32-51 (Russisch) -; KowAuskı, K. (1978): Prosomys insuliferus (Kowalski, 1958) (Rodentia, Mammalia) from the Plio- cene of Poland and of the European Part of the USSR. Acta Zool. Cracov. 23, 29-54. Brunn, H.; Dawson, M.; MEın, P. (1970): Upper Pliocene Rodentia, Lagomorpha and Insectivora (Mammalia) from the Isle of Rhodes (Greece). I. II. III. Proc. Koninkl. Nederl. Akad. Wetensch. B. 73, 535-584. CHALINE, J. (1970): Les rongeurs du pleistocene moyen et superieur de France. These de Doct. es Sci. Nat. 1-3, 596. Dijon. - (1976): In: CLoT, A. et al.: Montousse 5 (Hautes-Pyrenees), un nouveau remplissage de fissure. Une faune de vertebres du pleistocene inferieur. Ge£obios (Lyon) 9, 511-514. FAHLBUSCH, V. (1969): Pliozäne und pleistozäne Cricetinae (Rodentia, Mammalia) aus Polen. Acta Zool. Cracov. 14, 99-138. — Mayr, H. (1975): Microtoide Cricetiden (Mammalia, Rodentia) aus der Oberen Süßwasser-Molasse Bayerns. Paläont. Z. 49, 78-93. FELTEN, H.; SPITZENBERGER, F.; STORCH, G. (1971, 1973). Zur Kleinsäugerfauna West-Anatoliens. I-II. Senckenb. biol. 52, 393-424; 54, 227-290. FREUDENTHAL, M.; Korbos, L. (1989): Cricetus polgardiensis sp. nov. and Cricetus kormosi Schaub, 1930 from the Late Miocene Polgärdi localities (Hungary). Scripta Geol. 89, 71-100. GROM©v, 1.M. et al. (1965): Nazemnye belici (Marmotinae). (Erdhörnchen (Marmotinae). In: Fauna SSSR, Mlekopitauscie, IIV2, (Russisch) -; POLJAKOV, 1. J. (1977): Mlekopitauscie. Poljevki (Microtinae), (Wühlmäuse). In: Fauna SSSR. IIV/8. Le- ningrad: Akad. Nauk SSSR. (Russisch). -; BARANOVOJ, G. I. (1981): Katalog Mlekopitajuscich SSSR. (Katalog der Säugetiere der Sowjetunion). Nauka Leningrad. (Russisch). GROMov, V. I. (1948): Paleontologiceskoje i archeologiceskoje obosnovanje stratigrafii kontinentalnich otlozenij cetverticnovo perioda na territorri SSSR. (Mlekopitajuscie, Paleolit) (Paläontologische und archäologische Begründungen zur Stratigraphie der Kontinental-Ablagerungen des Quartärs in der UdSSR). Trudy inst. geolog. A. N. SSSR. 64, 17, (Russisch) JAEGER, J.J. (1969): Les rongeurs du pleistocene moyen de Ternifine (Algerie). C. R. Acad. Sci. 269, 1492-1495. - (1970): Decouverte au Jebel Irhoud des premieres faunes de rongeurs du Pleistocene inferieur et moyen du Maroc. C.R. Acad. Sci. 270, 920-923. — (1975): Les rongeurs du miocene moyen et superieur du Maghreb. These de Doct. Sci. Nat., Montpellier. -; MIcHAUX, J.; DAvıp, B. (1973): Biochronologie du Miocene moyen et sup£rieur continental du Magh- reb. C.R. Acad. Sci. 277, 2477-2478. JAnossy, D. (1961): Die Entwicklung der Kleinsäugerfauna Europas im Pleistozän (Insectivora, Roden- tia, Lagomorpha). Z. Säugetierkunde 26, 1-11. — (1969): Uj Eomyida (Rodentia, Mammalia) a bödvaszilasi Osztramosi köfejtö 3. lelethelyenek also- pleisztocen faunajabol. (A new species of Eomyidae (Rodentia, Mammalia) from the Lower Pleis- tocene fauna of loc. No. 3. of the Osztramos quarry of Bödvaszilas). Öslenytani Vitäk 13, 5-40. — (1972): Ein kleiner Hystrix aus dem Altpleistozän der Fundstelle Osztramos 8. (Nord-Ungarn). Ver- tebr. Hung. 13, 163-182. - (1976): Die Revision jungmittelpleistozäner Vertebratenfaunen in Ungarn. Fragm.Min. et Pal. 7, 29-54. — (1983): Lemming-remain from the Older Pleistocene of Southern Hungary (Villany, Somssich-hill 2). Fragm.Mimn. et Pal. 11, 55-60. Kleinsäuger aus dem Pleistozän und Holozän DS — (1986): Pleistocene Vertebrate faunas of Hungary. Budapest: Akademiai Kiadö; Amsterdam: Elsevier. -; MEULEN, A.J.v.d. (1975): On Mimomys (Rodentia) from Osztramos 3, North Hungary. Proc. Ko- nikl. Nederl. Akad. Wetensch. B. 78, 381-391. KAHLKE, H. D. (1981): Das Eiszeitalter. Leipzig: Urania Verlag. KOENIGSWALD, W. v. (1980): Schmelzstruktur und Morphologie in den Molaren der Arvicolidae (Ro- dentia). Abh. der Senckenb. Naturforsch. Ges. 539, 129. -: FEJFAR, O.; TCHERNOV, E. (1982): Revision einiger alt- und mittelpleistozäner Arvicoliden (Rodentia, Mammalia) aus dem östlichen Mittelmeergebiet (Ubeidiya, Jerusalem und Kalymnos-Xi). N. Jb. Geol. Pal. 184, 1-23. Korpos, L. (1988): A Spalax nemzetseg (Rodentia) euröpai megjelenese Es a plio-pleisztocen hatärkerdes. (The appearance in Europe of the genus Spalax (Rodentia) and the problem of the Plio-Pleistocene boundary). M. äll. Földtani Int. Evi Jelentese as 1986 &vröl, 469-491. Kowarskı, K. (1977): Fossil Lemmings (Mammalia, Rodentia) from the Pliocene and Early Pleistocene of Poland. Acta Zool. Cracov. 22, 298-317. KRETZOI, M. (1969): Skizze einer Arvicoliden-Phylogenie, Stand 1969. Vertebr. Hung. 11, 155-193. Kuss, S. E.; STORCH, G. (1978): Eine Säugetierfauna (Mammalia: Artiodactyla, Rodentia) des älteren Pleistozäns von der Insel Kalymnos (Dodekanes, Griechenland). N. Jb. Geol. Pal. Monatsh. 4, 206- DON: LoPEZ MARTINEZ, N.; MICHAUX, J.; VILLALTA, J. F. (1976): Rongeurs et Lagomorphes de Bagur-2 (Pro- vince de GeErone, Espagne). Nouveau remplissage de fissure de debut du Pleistocene Moyen. Acta Geol. Hisp. 11, 46-54. Maı, H. (1979): Die Biberart Trogontherium aus den Mosbacher Sanden bei Wiesbaden. Mainzer Na- turw. Archiv 17, 41-64. Maur, L. (1990): Überblick über die unterpleistozänen Kleinsäugerfaunen Europas. Quartärpaläont. $, 153-191. MAYHEWw, D. F. (1977): Late Pleistocene small mammals from Arnissa (Macedonia, Greece). Palaeont. Proc. Konikl. Nederl. Akadem. Wetensch. B. 81, 302-321. — (1978): Reinterpretation of the extinct beaver Trogontherium (Mammalia, Rodentia). Phil. Transact. Royal Soc. 281, 983, 407-438. NADACHOWSKI, A. (1985): Biharian voles (Arvicolidae, Rodentia, Mammalia) from Kozi Grzbiet (Cen- tral Poland). Acta Zool. Cracov. 29, 13-28. — (1989): Origin and History of the Present Rodent Fauna in Poland Based on Fossil Evidence. Acta Theriol. 34, 37-53. -; Kowarskı, K. (1982): Rodentia. In: Excavation in the Bacho Kiro cave (Bulgaria). Panstwowe Wy- dawnictwo Naukowe. Ed by J. Kozlowski. Warszawa. Pp. 45-51. NEsin, V. A.; SKORIK, A. F. (1989): Pervaja Nachodka Polevki Roda Dinaromys (Microtinae, Rodentia) v SSSR. (First Find of the Dinaromys Vole (Rodentia, Microtinae) in the USSR. Vestnik Zool. 5, 14-17. (Russisch) RABEDER, G. (1981): Die Arvicoliden (Rodentia, Mammalia) aus dem Pliozän und dem älteren Pleisto- zän von Niederösterreich. Beitr. Paläont. Österreich 8, 343. RADULESCO, C.; SAMSON, P. (1973): Sur la d&couverte d’une espe£ce nouvelle de Murid& (Rodentia, Mam- malia) dans le Pleistocene moyen de Roumanie. Rev. Roumaine Biol. Ser. Zool. 18, 185-187. -; TERZEA, E. (1963): In: Pestera „La Adam“ statione pleistocena (La grotte „La Adam“ station pleisto- cene). Edby M. Dumitkescu et al. Lucralile Inst. Speol. „E. Rakovita“, Bucarest, 1-2, 229-284. REKOoVvETS, L. I. (1985): Mikroteriofauna desnjansko-podnieprowskowo pozdnewo paleolita (Kleinsäu- gerfauna des späten Paläolits entlang der Desna und des angrenzenden Dniepers). Naukowa Dum- ka Akad. Nauk. 166. (Russisch) REUMER, J. W. F. (1983): Ruscinian and Early Pleistocene Soricidae (Insectivora, Mammalia) from the Tegelen (the Netherlands) and Hungary. Diss. Univ. Utrecht. SALA, B. (1994): Dinaromys allegranzii n. sp. (Mammalia, Rodentia) from Rivoli Veronese (N. Italy), in a villanyian association. Abstr. in: Neogene and Quarternary Mammals of the Palaearctic. Conf. Krakow May 1994, 60-61. SPITZENBERGER, F. (1971): Zur Systematik und Tiergeographie von Microtus (Chionomys) nivalis und Microtus (Chionomys) gud (Microtinae, Mamm.) in S-Anatolien. Z. Säugetierkunde 36, 370-380. STORCH, G. (1975): Eine mittelpleistozäne Nager-Fauna von der Insel Chios, Ägäis (Mammalia: Roden- tia). Senckenb. biol. 56, 165-189. — (1978): Familie Gliridae Thomas, 1897 — Schläfer. In: Handbuch der Säugetiere Europas Bd. 1, Hrsg. von J. NIETHAMMER und F. KrApp. Wiesbaden: Akad. Verlagsges. Pp. 201-280. 2SD, D. JAnossy — (1988): Eine jungpleistozäne/altholozäne Nager-Abfolge von Antalya, SW-Anatolien (Mammalia, Rodentia). Z. Säugetierkunde 53, 76-82. SUTCLIFFE, A.J.; Kowauskı, K. (1976): Pleistocene Rodents of the British Isles. Brit. Mus. Nat. Hist. Bull. Geol. 27, 31-147. TCHERNOV, E. (1986): Les mammiferes du pleistocene inferieur de la Valle&e du Jourdain a Oubeidiyen. Mem. et Trav. du Centre de Rech. Frangais de Jerusalem. 5 — (1988): The biogeographical history of the southern Levant. In: The zoogeography of Israel. Vol. 8. Dordrecht: W. Publishers. Pp. 159-248. TERZEA, E. (1973): Les genres Rattus et Lemmus (Rodentia, Mammalia) dans le Pleistocene de Rouma- nie. Liv. cinqu. Inst. Speol. „Emile Racovitza“. Pp. 427-436. TIUTKoVvA, L.; KAıpovAa, G. (1994): Late Pliocene and early Quarternary microtheriofaunas of South- Eastern Kazakhstan. Abstr. in: Neogene and Quaternary Mammals of the Palaearctic. Conf. Kra- kow. May 1994, 73-74. ToBIEN, H. (1981): Micromammals from the Pliocene/Pleistocene Boundary in Turkey. Manuskript 2. ToPACEVSKNW, V. A. (1965): Nasekomoadnye i gryzuny nogajskoj pozdne-pliocenovoj fauny (Insekten- fresser und Nager der Spätpliozänen Fauna von Nogajsk). Kiev. (Russisch) — (1969): Fauna SSSR Mlekopitajuscie 3.3 (Spalacidae). Akad. Nauk SSSR. Leningrad: „Nauka“. (Russisch) -; SKORIK, A. F. (1977): Gryzuny rannetamanskoj fauny tiligul’ skogo razreza (Die Nagetiere der früh- tamanischen Fauna des Aufschlusses von Tiligul). Kiev. (Russisch) WANGENGEIM, E. A. (1982): Paleozoogeographiceskie osobennosti rasprostranenija mlekopitajuscich (Paläozoogeographische Übersicht der Verbreitung der Säuger). Stratigraphia SSSR Vol. 1 Cetver- ticnaja Sistema, „Nedra“ Moskwa. Pp. 279-283. (Russisch) WoLDSTEDT, P. (1969): Quartär. Handbuch der Stratigraphischen Geologie, Bd. 2. Stuttgart: F. Enke Verlag. ZAZHIGIN, V.S. (1982): Rodentia. Stratigraphia SSSR 11, Cetverticnaja Sistema, „Nedra“ Moskwa. Pp. 294-305. (Russisch) Anschrift des Verfassers: Dr. DEnES JAnossy, Geologische-Paläontologische Abteilung, Ungarisches Naturhistorisches Museum, Pf. 330, H-1370 Budapest, Ungarn. Z. Säugetierkunde 61 (1996) 253-255 ZEITSCHRI FTSE FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Altitudinal distribution and feeding habits of Blarinomys breviceps (Winge, 1888) (Rodentia: Muridae) By S. F. Dos Reıs, J. P. POMBAL, Jr., J. L. NEssıMIAN, and LEILA MARIA PESSÖA Departamento de Parasitologia, Universidade Estadual de Campinas, Säo Paulo, Departamento de Vertebrados, Museu Nacional, Rio de Janeiro and Departamento de Zoologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil Receipt of Ms.. 17. 11. 1995 Acceptance of Ms. 22. 02. 1996 Blarinomys breviceps (WINGE, 1888) is an Akodontini rodent in the subfamily Sigmodon- tinae, with morphological features associated with fossorial habit characteristics, including a short and conical head, noticeably reduced eyes, short ears and tail, and prominent claws (Nowak and PArADIso 1983). This ensemble of features renders B. breviceps very si- milar to shrews in overall morphology, and clearly distinguishes this remarkable species from all other neotropical murids (ABrAavaya and Marson 1975). Since its description from a skull fragment from the Pleistocene of Lagoa Santa in Minas Gerais in eastern Brazil, the few reports available have described details of the capture of specimens and general aspects of natural history (THomAs 1896; GoELDıI 1902; Davıs 1944, 1945; Nowak and PArADIso 1983). The most complete account is that of ABravaya and Marson (1975), where they reported on morphological and ecological data from 31 specimens collected in Espirito Santo in eastern Brazil during the period of August, 1970 through April, 1974. These authors reported data on individual, sexual, and geographical variation, but empha- sized two aspects of the biology of B. breviceps. Firstly, ABravaya and Marson (1975) re- marked that all known specimens, for which precise geographical records are available, have been collected from localities above 750 m, suggesting that 2. breviceps is restricted zoogeographically to the southeastern Brazilian highlands. Secondly, they argued that the observed low coefficients of variation for cranial measurements might be associated with a presumptive insectivorous habit, although according to ABrAvaya and Marson (1975) “There is nothing known of the feeding habits of B. breviceps.” The last account on B. breviceps dates back nearly 20 years ago (MArtson and ABRA- vayA 1977), a fact that underscores the rather sporadic appearance of this species in biolo- gical surveys. In this note we report on two specimens of B. breviceps from a new locality, Aracruz, in the state of Espirito Santo, Brazil. We believe that the capture of the speci- mens from Aracruz is important, not only because it adds two more specimens to the rela- tively small sample of reported 40 specimens in museums worldwide (MArtson and ABrAVAYA 1977), but also because the data provide novel and conclusive information which is relevant for the definition of the altitudinal range of distribution and the feeding niche of B. breviceps. The two individuals of B. breviceps reported herein were collected in pitfall traps, and this represents a first account of this type of capture since the other specimens obtained so far have been caught either by hand (Davıs 1944; Apravaya and Marson 1975) or with conventional snap traps (MATson and ABrAVAYA 1977). The pitfalls were set originally as 254 S. F. Dos Reis, J. P. POMBAL, JR., J. L. NEssIMmIAn, and LEILA MARIA PESSÖA part of a long-term project started in January of 1994, intended to sample and monitor the herpetofauna of a second growth forest surrounded by eucalyptus plantations in the locality of Aracruz (19° 49’S, 40° 16° W), state of Espirito Santo, eastern Brazil. Buckets of approximately 50 liters were used as pitfalls with drift fences arranged in linear transects. The two specimens of B. breviceps, one male and one female, were col- lected on consecutive nights (3-4 July, 1995), in the same pitfall trap. The stomachs of the two specimens were removed and preserved in 75% ethanol and the specimens were pre- pared as study skins and skulls and deposited in the mammal collection of the Museu Na- cional, Rio de Janeiro (MN 37029 and MN 37030; male and female, respectively). Stomach contents were examined under a binocular microscope to determine the nature of food items. For further details on the methodology to study stomach contents, see MiL- Ton and NEssımIan (1984). The food items are deposited as vouchers in the Museu Nacio- nal. Blarinomys breviceps has been recorded from several localities, primarily montane for- ests, in the states of Bahia, Minas Gerais, Espirito Santo, and Rio de Janeiro in eastern Brazil, and in all instances the elevation for all localities was above 750 m (Martson and ABRAVAYA 1977). It is interesting to note that MAtson and ABrAVvAYA (1977) speculated that records from Ilheus in the state of Bahia (Moosen 1952; Avıra-PırEs 1960) might in- dicate that B. breviceps did occur at sea level. However, in the view of the fact that the re- gion of Ilheus also includes montane areas, MATson and ABrAvAYA (1977) suggested that “B.breviceps occurs chiefly in montane areas.” The locality of Aracruz in the state of Espirito Santo, where the specimens reported here were collected, is located 60 m above sea level, and this record definitely establishes that 2. breviceps is not restricted in distri- bution to montane forests in the highlands of eastern and southeastern Brazil as pre- viously thought. The diet of B. breviceps is so far unknown, although authors have suggested that this species is an insectivore (ABRAvAYA and MATson 1975; MATson and ABrRAVAYA 1977). De- tailed analysis of the stomach contents in the two specimens of B. breviceps showed ar- thropods of six different taxonomic orders, representing at least five families (Tab. 1). We present the data separately because the arthropod taxonomic groups represented in the stomach contents differed in the male and female B. breviceps. Approximately 90% of the Table 1. Insects and arachnids recovered from the stomach contents of a male and female Blarinomys breviceps Order Specifications Minimum number of individuals Coleoptera Curculionidae (one species) Scarabaeidae (one species) Hymenoptera Formicidae (two species) Araneae Ctenidae (Ctenus sp.) Coleoptera Scarabaeidae (one species) 1 Hymenoptera Formicidae (four species) 6 Isoptera one species 13 workers, 2 soldiers Blattariae one species remains of ootheca Araneae Trechaleidae (one species) 1 Amblypygi one species 1 Altitudinal distribution and feeding habits of Blarinomys breviceps 255 diet is composed of insects, and the remaining items are represented by arachnids. The two individuals share food items in one beetle family, Scarabaeidae, one family of Hymen- optera, Formicidae, and the order Araneae, although different families are present in the male and female. Specimens representing Isoptera, Blattariae, and Amblypygi were found only in the famale stomach. The data obtained from the two specimens described here in- dicate that B. breviceps feeds primarily on insects, and thus confirm the conjecture of ABRAVAYA and Marson (1975). Acknowledgements R.P.Bastos and J. Gasparını helped with field work. We are indebted to R.L.C. BaprıstA for identification of the arachnids. Earlier versions of this manuscript were improved by comments and suggestions from C. J. Von ZuBEn and P. R. MANZANI. Research supported by ARACRUZ, S. A. Work by SFR, JPP, and LMP is supported by re- search fellowships from Conselho Nacional de Desenvolvimento Cientifico e Tecnolögico. Literature ABRAVAYA, J. P.; Marson, J. ©. (1975): Notes on a Brazilian mouse, Blarinomys breviceps (Winge). Con- trib. Sci. Nat. Hist. Mus., Los Angeles Co. 270, 1-18. AVILA-PIRES, F. D. (1960): Roedores colecionados na regiäo de Lagoa Santa, Minas Gerais, Brasil. Arg. Mus. Nac., Rio de Janeiro 50, 25-43. Davıs, D. E. (1944): The capture of the Brazilian mouse Blarinomys breviceps. J. Mammalogy 25, 367- 369. Davis, D. E. (1945): The annual cycle of plants, mosquitos, birds, and mammals in two Brazilian forestes. Ecol. Monog. 15, 243-295. GoELDI, E. A. (1902): Dois roedores notäveis da familia dos ratos do Brazil. Bol. Mus. Paraense Hist. Nat. e Ethnog. 3, 166-180. MATSON, J. O.; ABRAVAYA, J. P. (1977): Blarinomys breviceps. Mamm. Spec. 74, 1-3. MiıLTton, K.; NEsSIMIAN, J. L. (1984): Evidence for insectivory in two primate species (Callicebus torqua- tus lugens and Lagothrix lagothricha lagothrica) from northwestern Amazonia. Am. J. Anthrop. 6, 367-371. MOooJen, J. (1952): Os roedores do Brasil. Rio de Janeiro, Brazil: Instituto Nacional do Livro. Nowak, R. M.; PARADISO, J. P. L. (1983): Walker’s Mammals of the World. 4th ed. Baltimore: The Johns Hopkins Press. THomas, O. (1896): On new mammals from the Neotropical region. Ann. Mag. Nat. Hist. 6, 301-314. WinGe, H. (1888): Jordfundne og nulevende Gnavere (Rodentia) fra Lagoa Santa, Minas Gerais, Brasi- lien, E. Museo Lundii 1, 1-200. Authors’ adresses: SERGIO F. Dos Reıs, Departamento de Parasitologia, Caixa Postal 6109, Universi- dade Estadual de Campinas, 13083-970 Campinas, Säo Paulo, Brazil; J. P. POMBAL, Jr., Departamento de Vertebrados, Museu Nacional, Quinta da Boa Vista, 20940- 040 Rio de Janeiro, Rio de Janeiro, Brazil; J. L. Nessımıan and LEILA MARIA PES- söA, Departamento de Zoologia, Universidade Federal do Rio de Janeiro, 21941- 590 Rio de Janeiro, Rio de Janeiro, Brazil Z. Säugetierkunde 61 (1996) 256 FÜR © 1996 Gustav Fischer, Jena lose. INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Buchbesprechung KAHLKE, R.-D.: Die Entstehungs-, Entwicklungs- und Verbreitungsgeschichte des oberpleistozänen Mammuthus-Coelodonta-Faunenkomplexes in Eurasien (Großsäuger). -— Abh. Senckenberg. Natur- forsch. Ges. 546: 1-164, 23 Karten, 4 Tab. Frankfurt: Verlag Waldemar Kramer 1994. DM 55,00. ISBN 3-7829-1137-7. In der letzten pleistozänen Kaltzeit, dem Weichsel-Glazial, bestand in den gemäßigten und nördlichen Breiten der Palaearktis des Ökosystem der Steppentundra oder Mammutsteppe. Es reichte vom Atlan- tik zum Fernen Osten und nach Beringia und war von einer charakteristischen Assoziation extrem kälteadaptierter Großsäuger bewohnt, welche der Verfasser nach den Leitarten Mammut-Fellnashorn- Faunenkomplex bezeichnet. Es gibt keine rezente Analogie zu diesem Ökosystem, und das dramatische Scenario des raschen Zerfalls dieser Säugergemeinschaften am Ende des Pleistozäns, verbunden mit dem Aussterben imposanter Formen, hat zu vielfältigen Betrachtungen Anlaß gegeben (z.B. „overkill“- Hypothese). Der Verfasser wendet sich jedoch einem eher vernachlässigten Thema zu, nämlich der Entwicklung des Mammuthus-Coelodonta-Komplexes in Zeit und Raum. Es ist keine Literatur-Kompilation. Die Arbeit gewinnt sehr dadurch, daß der Verfasser die relevan- ten Sammlungen und viele Fundstellen in Europa, China, der Mongolei und GUS aufsuchen und aus er- ster Hand und zum ersten Mal in dieser Form eine Synthese der Verhältnisse in Europa und dem riesigen asiatischen Raum vornehmen konnte. Er legt dar, daß die Entwicklung kälteadaptierter Großsäugerfaunen bis in das Jungtertiär zurückzuverfolgen ist. Die Stammformen waren vorwiegend Bewohner offener und trockener Lebensräume, vor allem Steppen (z.B. Mammut, Fellnashorn, Yak, Saiga), des weiteren auch von Tundren (Ren, Moschusochse) und Waldsteppen (Fleckenhyäne, Tiger). Aus zoogeographischer Sicht sind die Vorläufer überwiegend palaearktisch-asiatischer Herkunft, kom- plementiert durch nearktische (Equus, Canis, Rangifer) und äthiopische Elemente (Crocuta, Mam- muthus). Im Verlauf des Pleistozäns erfolgte vor dem Hintergrund immer rascherer klimatischer Oszillationen und einschneidender werdender Klimadepressionen die Ausbildung charakteristischer Kaltzeitfaunen. Sie ging hauptsächlich von östlichen, insbesondere nordöstlichen palaearktischen Ge- bieten aus und erreichte Europa und Nordamerika. Zunächst blieben die Faunen der Tundren Berin- gias und Steppen Innerasiens getrennt. Jeweils während der Kälteperioden des Elster-, Saale- und Weichsel-Glazials hatten sich Tundren- und Steppenfaunen unter Reduktion des Taiga-Gürtels zum Mammuthus-Coelodonta-Komplex überlagert und bildeten Teil eines neuartigen Ökosystems, der Step- pentundra. Der Verfasser dokumentiert diese dynamischen Prozesse, erörtert die vielfältigen Anpassungsstra- tegien verschiedener Säugergruppen an Kälte und Trockenheit und legt detaillierte Verbreitungskarten für das Weichsel-Glazial der Palaearktis vor. Die behandelten Taxa sind Canis lupus, Alopex lagopus, Ursus spelaeus und maritimus, Gulo gulo, Crocuta crocuta, Lynx Iynx, Panthera leo und tigris, Mam- muthus primigenius, Equus spp., Coelodonta antiquitatis, Megaloceros spp., Alces alces, Rangifer taran- dus, Bos grunniens und baikalensis, Bison priscus, Spirocerus kiakhtensis, Saiga tatarica und Ovibos moschatus. Der Referent empfiehlt dieses Opus magnum von R.-D. KAHLkE nicht nur Vertretern von Paläo-Disziplinen. Es trägt viel zum Verständnis auch heutiger systematischer, ökologischer und zoo- geographischer Verhältnisse bei. Der Faunenwechsel zu unserer heutigen gemäßigten Waldfauna liegt noch nicht lange zurück! Zwei kleine Kritikpunkte: Der Referent vermißt stratigraphische/biochronolo- gische Übersichtstabellen zur Orientierung über das zeitliche Auftreten von Leitformen und die Alters- stellung von bedeutenden Referenzlokalitäten. Auch kann er sich nicht so recht mit dem Gebrauch der biostratigraphischen Termini Villafranchium und Epivillafranchium anfreunden. G. STORCH, Frankfurt/M. Instructions to Authors Submission and Acceptance of Manuscripts: Manuscripts for publication should be sent to the managing editor, Prof. Dr. D. Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany. 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 mail, and as soon as the peer reviews are received the authors will be informed concerning the decision for acceptance. All correspondence dealing with details of production should be sent to: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena, Germany. Form of the Manuscript: Each manuscript should be submitted in triplicate, one original and two copies, in English or German in DIN A4 format either from a typewriter or legible computer print with a left-hand margin of 4cm and true double spacing. The first page of the manuscript should contain the following information: title name(s) of author(s): full first name for female authors, first initials only for male authors department and university affiliation In case of long titles, a running title not exceeding 72 characters should be provided. 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. Names of authors cited in the text are to be underlined in pencil. Botanical and zoological generic and species names should be underlined in pencil with a wavy line, whereby the name of the scientist who first described the species is not underlined. 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. Content of the Manuscript: Manuscripts can be published as original investigations or short communications. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 typewritten, double-spaced 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. If written in German, an English title should precede the English abstract; if written in English, a German title should precede a German summary. b) Short communications: Short communications must not exceed 5 typewritten, double-spaced pages and do not require either an Abstract or Summary. If written in German, an English title should be given on page 1; if written in English, please give a German title on page 1. Short communications need not be headed with subtitles into In- troduction, Materials and methods, Results, and Discussion but should be organized according to this form. Hlustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focussed and read- ily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. La- belling on figures should be uniform, 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 should 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 DIN A4 format sheets of paper. References: Each paper cited in the text must appear in alphabetical order at the end of the text. Only internation- ally available literature should 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; HANDLEY, ©. O. Jr. (1994): Evolution, biogeography, and description of a new species of Fruit-eat- ing bat, 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 und F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28. Printing: 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 under- taken 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 paper 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. 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). Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promionozal purposes, for creating new collective, or for resale. For copying from nz erumes of this journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 100537, D-07705 Jena; Telefon (03641) 6263, Telefax (03641) 62 65.00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A.Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zetsche Frau B. Dressler, Post- fach 10 05 37, D-07705 Jena; Telefon (03641) 62 64.44, Telefax (0 3641) 62 65 00. For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 4288201. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2945,-/382,50 (plus postage). Single issue price (DM/ÖSI/SFr): 80,-/592,-/77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please 'specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 860 10090; Deutsche Bank, Konto- Nr. 3907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400 00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 SÄUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Meia, J.-S.; Weber, J.-M.: Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains. — Sozialstruk- tur des Rotfuchses (Vulpes vulpes) im Schweizer Jura..... EN Kauhala, Kaarina: Habitat use of Raccoon dogs, Nyctereutes procyonoides, in southern Finland. — Habitatnutzung beim Marderhund, Nyctereutes procyonoides, im südlichen Finnland te ... 269 Perez-Barberia, F. J.;, Machordom, Annie; Fernändez, J., Nores, C.: Genetic variability in Cantabrian chamois (Rupi capra pyrenaica parva Cabrera, 1910). — Genetische Variabilität bei der Kantabrischen Gemse (Rupicapra pyre- naica parva, Cabrera, 1910)... Rr 276 Hartl, G.B.; Kurt, F; Tiedemann, R.; Gmeiner, Christine; Nadlinger, K.; U Mar, Khyne; Rübel, A.: Population genetics and systematics of Asian elephant (Elephas maximus): A study based on sequence variation at the Cyt b gene of PCR-amplified mitochondrial DNA from hair bulbs. — Populationsgenetik und Systematik des Asıa tischen Elefanten (Elephas maximus): Eine Studie auf der Grundlage von Sequenzvariation am Cyt-b-Gen von PCR-amplifizierter Mitochondrien-DNA aus Haarwurzeln | RE 285 Matson, J. O.; Christian, D. P.: Patterns of variation in cranial size and shape in two coexisting gerbilline rodents. Variationsmuster von Schädelgröße und Schädelform bei zwei syntopen Nagetierarten (Gerbillinae, Rodentia) 295 Macholan, M: Multivariate morphometric analysis of European species of the genus Mus (Mammalıia, Muridae) Morphometrische multivariate Analyse europäischer Arten der Gattung Mus (Mammalıa, Muridae) DA Buchbesprechung............ in ur 320 Indexed in Current Contents “ Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database SA en FISCHER Vol. a 257-320 E40 BONIS ARTIBUS SEMPER JENA +» STUTTGART»-NEW YORK Oktober 1996 1 996 REGESEN, (5 ZEITSCHRIFT FÜR &&: SÄUGETIERKUNDE INTERNATIONAL JOURNAL San OF MAMMALIAN BIOLOGY & TIERKUNDE Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kiel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel — R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel — M. Röhrs, Hannover -H. Schliemann, Hamburg - G. Storch, Frankfurt — P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) - W. Herre, Kiel (1962-1966) — H. Frick, München (1972-1976) - M. Röhrs, Hannover (1977-1981) — H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amntierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H. G. Erkert, Tübingen - W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel — Marialuise Kühnrich, Hamburg Z. Säugetierkunde 61 (1996) 257-268 ZEITSCHRIFT ® FL © 1996 Gustav Fischer, Jena SÄUG ET] ann INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Social organization of Red foxes (Vulpes vulpes) in the | Mountains By J.-S. MEıaA and J.-M. WEBER Institute of Zoology, University of Neuchätel, Neuchätel, Switze Receipt of Ms. T#. 75 Acceptance of Ms. 20. 05. 1996 Abstract We studied the social organization of red foxes Vulpes vulpes for four years (September 1989-August 1993) in a 30-km? area in northwest Switzerland. We caught and tagged 64 foxes, fitted 25 of them with radio-collars, and established a total of 18641 locations. We made 1470 direct observations of foxes, in- cluding 234 intra-specific encounters. There were five main findings: (1) Foxes formed spatial groups in mountainous areas; (2) in addition to the alpha pair and the helpers, the groups also included at least one additional non-helping female (old female that has lost its alpha rank); (3) although we usually ob- served only one litter in each group, mating was not limited to the dominant pair; (4) there were very few encounters between individuals. When conspecific foxes meet, they ignore each other in more than half the cases. Even at the resting sites, which were shared, foxes usually did not meet; (5) the social or- ganization of foxes appeared more stable than expected. We observed no change in social organization, despite a large decrease in the density of the fox’s main prey, the water vole Arvicola terrestris scher- man. The Resource Dispersion Hypothesis (R.D.H.) is the best existing model on group formation in solitary carnivores to explain the observed situation. Introduction Until recent years red foxes (Vulpes vulpes L.) were considered to be typical solıtary car- nivores. Then the radio-tracking studies of NıEwoLD (1976, 1980), MAcponALD (1979, 1980, 1981) HARRIS (1980), LiNDSTRÖM (1980) and von SCHANTZ (1981) showed that foxes can live in “spatial groups” (MAcDonALD 1983) in different habitats (e.g. urban areas). In such situations, more than one pair of adults share the same area. Although individuals of a group show overlapping home ranges, encounters are rare (MACDoNALD 1981; WHITE and HaArrıs 1994). The social organization of foxes in certain areas and/or habitats re- mains unclear, and relations between members of a group need further investigation (White and Harrıs 1994). In central Europe, the social organization of foxes has not been studied, except by PouLe£ et al. (1994). It has been suggested that foxes could be more solitary there because of particular conditions such as hunting and rabies (ARrTOoIS 1989; Arrtoıs et al. 1990). Several hypotheses have been proposed to explain why solitary foragers, such as foxes or European badgers (Meles meles L.), share their home range with conspecifics. They were recently summarized and discussed by WooDROFFE and MAcDoNALD (1993). Accord- ing to these hypotheses, group formation is governed by food quantity and/or distribution. Most of these hypotheses were based on observations in one type of habitat and need to be tested and discussed in relation to other conditions. 258 J.-S. MEıA and J.-M. WEBER In this context, we studied a fox population in the Swiss Jura Mountains in an area characterized by a fluctuating prey, the water vole Arvicola terrestris scherman Shaw (WEBER and Ausry 1993). Three questions were addressed: 1. What was the social organi- zation of foxes in this area”; 2. Was the social structure affected by changes in food avail- ability, ı.e. the fluctuations of water voles?; and 3. Which hypotheses could best explain the observed social structure? Material and methods Study area The 30-km? study area is in the northwestern part of Switzerland, 20 km north of Neuchätel (47°09’ N 6°56’E). The altitude ranges from 900 to 1290 m. Average annual rainfall is 1329 mm and average annual temperature 6.3°C. Snow cover is present sporadically from November to March (up to I m). Pastures and meadows cover 55% of the surface, pastures 25%, and spruce, Picea abies dominated forests to 20%. Most of the soil cover is thin and therefore the number of dens is low (see MEIA and WEBER 1992). Cattle breeding is the main agricultural activity. About 80 farms are evenly distributed over the area linked by a network of small roads. Fox hunting is traditional and is still carried out in some win- ters, e.g. the number of families in the study area was reduced from eight to two in winter 1887-88 (A. PARATTE, pers. comm.). Throughout northern Switzerland, including the study area, foxes have been vaccinated against rabies (Steck et al. 1982; WANDELER et al.1988). Only one case of a rabid fox was re- ported in the area during our study (A. HENNET, pers. comm.) In the study area, foxes are mainly nocturnal (WEBER et al. 1994) and rest by day either in a den or above ground according to the amount of cover available (MEıA and WEBER 1993). Home range sizes are small (seasonal estimates: 0.48-3.06 km?) and they are not affected by changes in food availability (MEIA and WEBER 1995). Social organization Foxes were caught using stopped-neck snares, foot snares, or by placing a net around a den into which we introduced a terrier dog. They were then tranquillized and individually tagged with one or two col- oured eartags (Dalton Supplies Ltd, Henley-on-Thames, UK). We fitted adult-sized foxes with activity- monitoring transmitters (Wildlife Materials Inc., Carbondale Ill, USA), and subadults with expanding radio-collars (AVM Instrument Co., Livermore Ca, USA). The collared foxes were monitored weekly during continuous tracking sessions of 24 hours, or of six hours during the night, and located during the diurnal period each day. A portable spotlight was used to confirm the radiolocation and to observe the behaviour of the individual. The presence of another fox (< 100 m, PoULLe et al. 1994) and possible in- teractions were also recorded. We also observed foxes: 1. during night lighting sessions, along a standard circuit (see below), and 2. during den-watching. All observations were made with 10x40 binoculars. Foxes were considered juvenile up to 31 August, subadult between 1 September and 31 December of their first year, and adult after 31 December of their first year. We calculated range overlap between Minimum Convex Polygons (MCP, MoHRr 1947) including all locations established in one season, excepting some scarce excursions (ZIMEN 1984). The overlap be- tween two home ranges is presented as the mean of their percentage overlaps (GEFFEN and MACDONALD 1992). We also recorded the number of unmarked foxes seen during tracking sessions or night lighting in the home ranges of the collared foxes. The sex of the individuals could be sometimes determined by the posture when marking (LABHARDT 1990) or the size and shape of the individual. When two foxes met, we considered that an interaction occurred when the behaviour of one individual influenced in a perceptible way the behaviour of the second one (McFARLAnD 1990). The attitudes during encounters were interpreted using the indications of MEcH (1970), Artoıs (1989), and LABHARDT (1990). Density, reproduction, mortality We estimated the density of foxes in two ways. First, we counted the individuals observed during night lighting sessions once a month on three successive nights along a 30-km representative circuit. Since the Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains 259 number of foxes seen fluctuated (see WEBER et al. 1991; MEıA et al. 1993 for previous discussions), we have chosen to use seasonal data. The possible changes in the density were calculated using seasonal means, whereas the real density was estimated using the higher number of foxes observed per session in each season: we divided this number by the surface of the study area to obtain a minimal estimate, and by the surface lit up with a spotlight to obtain a maximal estimate. Second, we counted the number of families, i.e. the number of breeding dens used each year (see MEıA and WEBER 1992 for more details). To estimate annual reproduction, we counted the number of cubs at each breeding den. We evalu- ated mortality by collecting dead foxes. We also used the body weight of the cubs caught as an indica- tion of their life expectancy. Food Fox scats were collected every month to analyse the diet composition (for detailed methods, see WEBER and Augry 1993; FERRARI and WEBER 1995). The density of water voles was estimated in spring, summer and autumn (PAscAr and MEYLAN 1986; WEBER and Ausry 1993). No trapping occurred in winter be- cause of frozen ground and snow cover. Other food resources (small mammals, domestic stock, earth- worms, wild fruits, and exploitable scraps) were measured and their abundance estimated in fixed plots (FERRARI and WEBER 1995). Results Number of captures, number of observations Between September 1989 and August 1993, we made 79captures, corresponding to 64 foxes. We caught 35 females and 29 males; the difference was not significant (Fisher test, p > 0.05). All subadults/adults (15 individuals: four males and 11 females) and 10 ju- veniles were fitted with a radio-collar; they were monitored between one and 45 weeks. Altogether, we established 18641 locations during 143 24-hour tracking sessions, 84 6-hour tracking sessions, and 2670 daily diurnal locations. We have described 1470 direct observations of foxes including 234 intra-specific encounters. Social organization a) Overlap between home ranges An adult female (F 11) showing a nomadic range use (ZımEn 1984; MULDER 1985; MEIA and WEBER 1995) has been separated from resident foxes; its range overlapped widely with those of two resident females (Tab. 1). Overlap in resident foxes varıied from 0.1 to 58.2% (Tab. 1). The overlap was large (x =45.3%, Tab. 1) between foxes with home ranges containing the same breeding den, and small (x = 3.9%, Tab. 1) between foxes with ranges containing two breeding dens. There was no significant difference (Mann-Whitney U-test, p > 0.05). Table 1. Overlap between seasonal home ranges (MCP) of subadult/adult foxes. Overlap between home ranges [%] a) Resident-nomadic b) Resident-resident same breeding den different breeding den * Number of observations 260 J.-S. MEıA and J.-M. WEBER Tracking and night lighting sessions confirmed that several individuals shared the same areas. Results from autumn and winter have not been used to estimate the number of foxes sharing the same areas because these seasons correspond to the dispersal period, which included movements of subadults that could not be morphologically distinguished from adults. We observed up to six additional adults in the home range of the collared foxes, but more frequently between one and three (Tab. 2). No trend of increase/decrease was observed during the study period. We noticed that the same foxes were not seen in two different ranges, unless a favourable feeding patch was present near the boundary (e.g. F25 or M21 groups in spring 1993, when an innkeeper provided food in the small overlapping part of both ranges, Tab. 2). Table. 2. Adult individuals (f, females; m, males; i, not Known) observed in the home ranges of the radio-tracked foxes (spring and summer). individuals N* individuals f,m f,m spring 92 summer 9 spring 91 F10, Fil,ii summer 91 ER ı EISEN me ED) JENL, an, st PR u DD m,f,i mel summer 91 spring 92 summer 93 F10, m, f M21, m “short tail”, m, f, i “one-eyed” i “one-eyed”, i,i spring 93 summer 93 mais 1, spring 93 summer 93 E25 NINANIMDARe F15,m,f spring 93 summer 93 DD ON DD W ww ın (99) DD W * Number of individuals b) Encounters Although several subadult/adult foxes shared the same area, they were rarely seen to- gether. Collared individuals were observed with another fox in less than 15% of the ob- servations (Table 3). There was no difference in this proportion at den sites (23.1%) and other sites (13.5%) (Fisher test, p > 0.05, Tab. 3). Few differences were noticed between seasons (Fisher test, p> 0.05, Tab. 3). Encounters between more than two individuals were scarce (two foxes: 88.4%, three: 9.3%, four: 2.3%). Resting sites (especially den sites) were often shared by several foxes which usually went in and out at different times and did not meet. Of the 1282 observations of foxes away from the den site, only 11.83% concerned individuals seen together, and then the foxes ignored each other in more than half the cases (Fig. 1). In those cases, the foxes, up to four, were mainly observed foraging in the same area (47 cases/82) or resting close to each other (12/82). These observations without interaction were more usual in autumn (62.2%), when several subadults still shared the parental range; they were less frequent in winter (30.0%) (Fisher test, p < 0.05). Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains 261 Table 3. Number of observations of the radio-tracked subadult/adult foxes, alone or with conspecific(s). Season Elsewhere with conspecific(s) with conspecific(s) Winter 6 (85.7%) 1 (14.3%) 182 (84.3% ) 34 (15.7%) Spring 14 (66.7%) 7(33.3%) 314 (87.2%) 46 (12.8%) Summer 0 0 276 (88.5% ) 36 (11.5%) Autumn 20 (83.3% ) 4 (16.7%) 404 (85.8% ) 67 (14.2%) Total 40 (76.9% ) 1210331%%)) 1176 (86.5% ) 1831 13570)) 100 7 N = 1131 80 y 8 I - S.S. synchronized interactions activities Fig. 1. Observations of foxes (without the observations at a den and the observations with juvenile foxes). Percentages of foxes observed with conspecific(s), percentages of interactions, and percentages of sensu stricto interactions (see text). 262 J.-S. MEıA and J.-M. WEBER There was an interaction between individuals in only 69 cases, between two foxes. As shown in figure 1, we defined two categories. First, we distinguished the “synchronized ac- tivities” that concerned situations with two foxes doing the same activity at the same time, often at a very short distance (<5 m) from each other. In all these cases, the behaviour of one individual seemed clearly to influence the behaviour of the other one: one fox followed the other (11 cases/28), both were foraging side by side (11/28), or fled together (6/28). Sec- ond, we recognized several interactions with a succession of attitudes taken alternatively by the two foxes; they were called “sensu stricto (s.s.) interactions” and were aggressive and non-aggressive encounters (24/41), play (9/41), or mating (2/41). The encounters were usually brief and, in most of the cases (20/24), we did not notice any aggressive attitude. In four cases, the encounters indicated the social status of tagged animals. c) Cooperation We observed adult foxes taking care of the young at 15 breeding dens. In most cases (11/ 15), only one adult (usually a female) was observed rearing the pups. In one case, they were two adults (one female and one indeterminate). In two cases, they were three adults (two females and one male) and in one case five adults (three females, one male, and one inde- terminate). Our observations at dens with adults and juveniles (N = 30) showed that one or two adults could rest in the den with the pups (N = 9), but, more frequently, the adults were observed reaching the den site in late afternoon (N = 17). Then they suckled, provided food or guarded the area while the cubs were playing. Only one adult fox was usually ob- served with cubs (N = 30), even if several adults were seen at the den during the rearing period; in eight cases, we noted that the adults came one after the other. We very rarely ob- served several adults rearing the cubs at the same time (N = 2). The mother dominated the other vixens at the den (N = 5), other vixens showed typical subordinate attitudes; in three cases they were fightened away by the mother. In three cases, the subordinate vixens were young animals of previous years; they took care of the young, and one of them was ob- served suckling. No evidence was collected on the relative status of the mother and the father of the pups. The males that we observed at dens played with the cubs, provided food or guarded the area (N =5). In two cases, radio-tracking showed that an old vixen that had reared young in the past could live in the area of the breeding den without taking part ın the rearing process. These females appeared to dominate the other foxes of the area, even though they had lost their breeding status. One of these old vixens was seen mating with the dog of the area, although she did not rear cubs the following spring. Other observations suggested that mating was not limited to one pair of foxes in each group: for example, we saw one male trying to mate with a subordinate female, and we radio-tracked young fe- males that had cubs but lost them in the first weeks after the birth. Food abundance and characteristics of the fox population a) Food availability Water voles were very abundant at the beginning of the study but decreased strongly be- tween 1989 and 1993 (Fig. 2a). Scat analysis showed that the abundance of water voles was correlated with their consumption by foxes (WEBER and AusryY 1993; FERRARI and WEBER 1995). The decrease in importance of water voles in the fox’s dıet was accompa- nied by an increase in the frequency of occurrence of all other food resources, but espe- cially wild fruits and scavengeable items (FERRARI and WEBER 1995). b) Density of foxes The seasonal means of the number of foxes counted during the night lighting sessions var- ied between four (winter 91-92) and 19 (autumn 92). We observed some seasonal differ- ences: there were fewer observations in winter than in other seasons due to snow cover in Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains 263 winter limiting movements and foraging. The arrival of cubs in spring, moving increas- ingly until autumn, led to more observations. Overall, no significant changes in density oc- curred (r,=0.27, p>0.05, Fig. 2b). This stability appeared also when counting the number of families each year (1990: 11 families, 1991: 10, 1992: 13, 1993: 11; x” goodness- of-fit, p > 0.05). The estimates of density (based on night lighting sessions and number of families) lay between 0.4 (minimum estimate) and 3.2 (maximum estimate) individuals per km“. c) Number of young Reproduction was stable. The total number of cubs in each year was, respectively, 46, 30, 43, and 43cubs (Tab. 4). We counted between one and seven cubs/family (x = 3.98, sd = 1.54); there was no difference between years (Kruskal-Wallis one-way ANOVA, p > 0.05, Tab. 4). Table 4. Number of cubs per family. Number of cubs per family * Total number of cubs ** Total number of families d) Mortality Mortality was difficult to estimate. A few foxes were found killed by traffic or mowing, especially juveniles when the breeding den was close to a road or meadow. The mortality of subadult/adult foxes remained uniform during the whole study period, because hunting pressure remained low. On the other hand, juvenile mortality seemed to increase: body weight of the juveniles caught decreased from year to year, although cubs were captured in the same period (r, =-0.4, p< 0.01, N=51, Fig. 2c). Our observations showed that the adults had difficulties in providing food for the young when there were fewer water voles. In 1993 we noticed at least two cubs dying of starvation. Dead cubs were usually eaten by siblings. Discussion Social organization In the groups of foxes observed during our study, the home ranges of the individuals did not overlap totally as described by NırwoLp (1980), HERSTEINSSON and MACDONALD (1982), or MuLDeEr (1985) but corresponded more closely to the situations observed by HARRIS (1980), von SCHANTZ (1981) and PouLLe et al. (1994) who noted that subordinate individuals use only a small part of the group range. There was no (or only a very small) overlap between the ranges of adjacent groups, as most frequently described (SARGEANT 1972; NıEwoLD 1980; HERSTEINSSON and MACDONALD 1982; MULDER 1985; PHILLIPS and CAtLinG 1991), but it appeared that foxes entered the range of adjacent groups to reach a rich feeding patch (e.g. human food supply), as observed in other areas (HaArrıs 1980). Our observations of the relations between group members agree with the descriptions of 264 J.-S. MEıA and J.-M. WEBER a) 1000 a 800 Z Be 600 8 9 400 © SS 200 2 0 Au Wi Sp Su Au Wi Sp Su Au Wi Sp Su Au Wi Sp Su 1989 1990 1991 1992 1993 7 b) 5 E 2 c 2 ® & e S : Z Au Wi Sp Su Au Wi Sp Su Au Wi Sp Su Au Wi Sp Su 1989 1990 1991 1992 1993 4000 la = - 0.4 C) N S Weight (g) Sp 9 Sp 91 Sp 92 Sp 3 Fig. 2. Food abundance and characteristics of the fox population during the study period. a) density of water voles (no trapping occurred during winter); b) number of foxes estimated from night lighting ses- sions; c) weight of the juvenile foxes. MACcDoNALD (1979, 1980, 1981) who noted, as PoULLE et al. (1994) and WHITE and HARRIS (1994), that individuals met only rarely. In our area, the relations between group mem- bers were less intense than those observed by MAcponALD (1979, 1980, 1981), because the members of a group usually did not meet in the feeding patches, but used them serially. Such behaviour seems advantageous if food resources are homogeneously distributed and quickly renewed (GEFFEn and MACDoNALD 1993). This was probably the most frequent si- tuation in our area (voles when abundant, earthworms during favourable nights). It was also indicated by patterns of movement and core areas (MEıA and WEBER 1995). In the few observations of foxes foraging together, we noted that the resources were clumped (group of berry trees, meadow after mowing), as they are in urban areas where several foxes visit a garden at the same time (Harrıs 1980; MAcpDonALD 1981). Even when rear- ing the young, or in the resting sites which corresponded to important meeting points (PouLLE et al. 1994), foxes did not meet. PouLLE et al. (1994) supposed that sharing of resting sites provided the opportunity of direct contacts between group members and thus Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains 265 played a role in the maintenance of social cohesion within the group. This was probably not so in our area. Resting sites and feeding patches were used in a similar way: foxes shared them if necessary or profitable, even with individuals of adjacent groups, but showed usually no interactions. Mating was not limited to the dominant pair. This differs noticeably from the descrip- tions of spatial groups (MAcponALD 1979, 1980, 1981) but agrees with observations on other species, like the European badger (CREssweELL et al. 1992). Admitting that the sim- ple presence of the dominant leads to “stress” and a decrease of the mating ability of the subordinates (MEcH 1970; MAcDoNALD 1980), explains why all foxes mated when encoun- ters were scarce in the study area. Red fox polygyny was already observed by ZABEL and TAGGART (1989) and suggested by other authors, e.g. von SCHANTZ (1981). Loss after mat- ing (von SCHANTZ 1981) probably explains why only one litter per group was seen. An- other reason could be a low ovulation rate in subordinate females when food is scarce (CArR and MAcDoNALD 1986; LiINDSTRÖM and LinDsSTRöM 1991). Subordinate females ob- tain less of the food in the group range than dominant individuals. Earlier studies of red foxes in mountainous areas (JONES and THEBERGE 1982; BOITANI et al. 1984; Branco 1986) did not examine social behaviour in detail, although the situa- tion appears similar to the present study. The social structure of foxes in mountainous habitats has probably remained unclear because of the difficulty in studying it. Stability of the social structure: explaining models There was no change in the social structure of the fox population, despite a large de- crease in the water vole population, although it has been shown that the number of fa- milies, the number of juveniles, and/or the number of individuals in groups depend on the quantity of available food (EnGgrunn 1980a; Korg and HEwson 1980; vVoN SCHANTZ 1984 b; LinpsTröM and LinDsSTRöM 1991). The Resource Dispersion Hypothesis R.D.H. (MAcponALD 1983; CArR and MAcDo- NALD 1986) predicts that the distribution of food patches determines the size of home ranges, whereas the richness of patches determines the number of individuals in the groups. According to this model, there is enough food to have more than one pair of indi- viduals in each range even when water voles are scarce. Thus, Artoıs (1989) assumes that the habitats of central Europe are very favourable for foxes and contain enough food even when rodents are scarce. Assuming the distribution of food patches in the study area did not change between 1989 and 1993, and that the total quantity of food did not de- crease substantially (since excluding some group members was not necessary), this model could explain group formation in our area. The Constant Territory Size Hypothesis C.T.S.H. (Linpström 1980; voN SCHANTZ 1984 a) was proposed for areas with fluctuating resources. In such habitats, the animals chose their territory when resources were scarce and maintained it, accepting conspeci- fics, when resources became abundant. This model supposes that the fluctuation period of prey ıs shorter than the life of the predator. As the cycles of water voles are plurian- nual (4-8 years), whereas most of the foxes probably do not reach an age of four years, this hypothesis does not work well in our study area. Moreover, there was no change in group sıze with the decreasing quantity of food. However, there is a delay between prey decrease and effects on predators (EnGLunD 1980 b; LinDsTRöM et al. 1994) and the ef- fect of food shortage in the study area may not yet have become apparent. We consider, like Artoıs (1989), this model inappropriate for central Europe, because it focusses on one food resource only, without taking in account the many other resources. The Territory Inheritance Hypothesis T.I.H. (Linpström 1986) predicts that groups form when subadults have difficulty in dispersing because all territories are occupied, and then remain in their parent range without breeding initially, but eventually acquiring re- 266 J.-S. MEIA and J.-M. WEBER productive status. This model could be applied in our study, since we observed subadult females that did not disperse and were helpers in the following years. Later observations (S. MEYER, pers. comm.) showed that these females could finally reach a high rank in the hierarchy and rear their own cubs. Nevertheless, we believe, like Artoıs (1989) and WooDROFFE and MACDONALD (1993), that this model explains how groups form and are maintained, rather than why groups form or not. The Prey Renewal Hypothesis P.R.H. (WAser 1981) suggests that when prey is re- newed quickly, being territorial is not necessary, and then home ranges can be shared with conspecifics. Our observations are consistent with this hypothesis: we observed that foxes formed groups and we consider that the main prey (water voles, earthworms and exploita- ble scraps) renews quickly. However, this hypothesis supposes that other criteria (“food patches” or dens) determine the territory, as already emphasized by WooDROFFE and MAc- DONALD (1993), and in this way it should be used in association with R.D.H. There are two problems in explaining fox social organization. First, social life in foxes is inconspicuous. Second, social life in foxes is very flexible and the hypotheses to explain why foxes can form groups appear to be too limited. Even if the R.D.H. seems to be the best model, the dispersion of resources is probably not the only influencing factor. Environ- mental and behavioural factors also influence the social life (SAUNDERS et al. 1993). Our work, like other recent studies in carnivore social life (e.g. HoFEr and East 1993; 'THURBER and PETERSoN 1993), reveals a situation that partially differs from classical descriptions. Acknowledgements This publications presents the main results of the PhD thesis of J. S. MEıA which was awarded the För- derpreis 1995 der Deutschen Gesellschaft für Säugetierkunde. We wish to thank particularly S. AuBRY for his considerable technical and field assistance. We thank also N. FERRARI, S. MEYER, A. BADSTUBER, N. LACHAT FELLER, E. Rıon, M. FELLRATH, and J. JOoSPIN-FRAGNIERE for field assistance, A. CoLLAUD and J.-P. JEANNERET for technical assistance, and J. MoRrET for statistical assistance, as well as C. MERMOD for his support. We are also grateful to M. Arroıs, L. BoITanI, H.-J. BLANKENHORN, M.-L. PoOULLE for com- ments on earlier drafts. Especially helpful have been the comments from R. Hewson and C. Boss. The study was funded mainly by the Swiss National Science Foundation (grant 31-27766.89). Zusammenfassung Sozialstruktur des Rotfuchses (Vulpes vulpes) im Schweizer Jura Von September 1989 bis August 1993 wurde im Schweizer Jura eine Rotfuchspopulation untersucht. Es wurden 64 Füchse markiert und 25 davon mit Radiosendern ausgerüstet. Insgesamt wurden 18500 Peilungen durchgeführt und 1470 Beobachtungen, mit 234 intraspezifischen Begegnungen, be- schrieben. Wir haben die folgenden Resultate erhalten: (1) In gebirgigen Habitaten können Füchse in Gruppen leben; (2) Die Gruppen bestehen aus einem Alpha-Paar und einigen Helfern und einer Fähe, die nicht als Helferin auftritt (alte Fähe, die nicht mehr als Mutterfähe wirkt); (3) Obwohl sich die Paa- rungen nicht nur auf das Alpha-Paar beschränken, wird immer nur ein Wurf pro Gruppe beobachtet; (4) Füchse begegnen anderen Füchsen selten. Wenn sie sich trotzdem treffen, nehmen sie keinerlei No- tiz voneinander. Kontakte werden vermieden; (5) Die Sozialstruktur ist viel stabiler als wir angenom- men haben. Wir können trotz einer starken Verminderung der Schermaus Arvicola terrestris scherman, die die Hauptbeute des Fuchses in diesem Gebiet bildet, keine Veränderung beobachten. Die “Re- source Dispersion Hypothesis (R.D.H.)” ist das beste Modell, um unsere Beobachtungen zu erklären. References Arroıs, M. (1989): Le Renard roux (Vulpes vulpes Linnaeus, 1758). In: Encyclopedie des carnivores de France. Vol. 3. Ed. by M. Artoıs and P. DELATTRE. Paris, Societ& frangaise pour l’Etude et la protec- tion des mammiferes. Pp. 1-90. Social organization of Red foxes (Vulpes vulpes) in the Swiss Jura Mountains 267 ARTOIS, M.; AUBERT, M.; GERARD, Y. (1982): Reproduction du renard roux (Vulpes vulpes) en France. Rythme saisonnier et fecondite des femelles. Acta Oecol., Oecol. Applic. 3, 205-216. ARTOIS, M.; AUBERT, M.; STAHL, P. (1990): Organisation spatiale du renard roux (Vulpes vulpes L., 1758) en zone d’enzootie de rage en Lorraine. Rev. Ecol. (Terre Vie) 45, 113-134. BLaAnco, J. C. (1986): On the diet, size, and use of home range and activity patterns of a red fox in cen- tral Spain. Acta Theriol. 31, 547-556. BOITANT, L.; BARASSO, P.; GRIMOD, Y. (1984): Ranging behaviour of the red fox in the Gran Paradiso Na- tional Park (Italy). Boll. Zool. 51, 275-284. CARR, G. M.; MACDONALD, D. W. (1986): The sociability of solitary foragers: a model based on resource dispersion. Anim. Behav. 34, 1540-1549. CRESSWELL, W. J.; HARRIS, S.; CHEESMAN, C. L.; MALLINSON, P. J. (1992): To breed or not to breed: An analysis of the social and density-dependent constraints on the fecundity of female badgers (Meles meles). Phil. Trans. R. Soc. London B338, 393-407. EnGLUnD, J. (1980 a): Population dynamics of the red fox (Vulpes vulpes) in Sweden. In: The red fox. Biogeographica. Vol. 18. Ed. by E. Zımen. The Hague: Dr W. Junk. Pp. 107-122. ENGLUND, J. (1980 b): Yearly variations of recovery and dispersal rates of fox cubs tagged in Swedish coniferous forests. In: The red fox. Biogeographica, Vol. 18. Ed. by E. Zımen. The Hague: Dr W. Junk. Pp. 195-207. FERRARI, N.; WEBER, J. M. (1995): Influence of the abundance of food resources on the feeding habits of the red fox, Vulpes vulpes, in western Switzerland. J. Zool. (London) 236, 117-129. GEFFEN, E.; MACDONALD, D. W. (1992): Small size and monogamy: Spatial organization of Blanford’s foxes, Vulpes cana. Anim. Behav. 44, 1123-1130. GEFFEN, E.; MACDONALD, D. W. (1993): Activity and movement patterns of Blanford’s foxes. J. Mam- malogy 74, 455-463. HARRIS, S. (1980): Home range and patterns of distribution of foxes (Vulpes vulpes) in an urban area, as revealed by radio tracking. In: A Handbook on biotelemetry and radio tracking. Ed. by C. J. AMLANER and D. W. MAcDoNALD. Oxford: Pergamon Press. Pp. 685-689. HERSTEINSSON, P.; MACDONALD, D. W. (1982): Some comparisons between Red and Arctic foxes, Vulpes vulpes and Alopex lagopus, as revealed by radio tracking. Symp. zool. Soc. London 49, 259-289. HoFER, H.; EAsT, M.L. (1993): The commuting system of Serengeti spotted hyaenas: how a predator copes with migratory prey. I. Social organization. Anim. Behav. 46, 547-557. Jones, D. M.; THEBERGE, J. B. (1982): Summer home range and habitat of the red fox (Vulpes vulpes) in atundra habitat, northwest British Columbia. Can. J. Zool. 60, 807-812. Kors, H. H.; Hewson, R. (1980): A study of fox populations in Scotland from 1971 to 1976. J. Appl. Ecol. 17, 7-19. LABHARDT, F. (1990): Der Rotfuchs. Hamburg, Berlin: Paul Parey. LiNDSTRÖM, E. (1980): The red fox in a small game community of the south taiga region in Sweden. In: The red fox. Biogeographica Vol. 18. Ed. by E. Zımen. The Hague: Dr W. Junk. Pp. 177-184. LINDSTRÖM, E. (1986): Territory inheritance and the evolution of group-living in carnivores. Anim. Be- hav. 34, 1825-1835. LINDSTRÖM, E.; LINDSTRÖM, C. (1991): Monitoring red fox Vulpes vulpes reproduction. In: Global trends in wildlife management. Ed. by B. BoBEK, K. PERZANOWwsKI, and W. REGELIN. Krakow: XVIIIth IUGB Congress. Pp. 393-397. LINDSTRÖM, E.; AANDREN, H.; ANGELSTAM, P.; CEDERLUND, P.; HOERNFELDT, B.; JAEDERBERG, L.; LEM- NELL, P. A.; MARTINSSON, B.; SKOELD, K.; SWENSON, J. E. (1994): Disease reveals the predator: sar- coptic mange, red fox predation, and prey populations. Ecology 75, 1042-1049. MACDONALD, D. W. (1979): ‘Helpers’ in fox society. Nature 282, 69-71. MACDONALD, D. W. (1980): Social factors affecting reproduction amongst red foxes, Vulpes vulpes. In: The red fox. Biogeographica. Vol. 18. Ed. by E. Zımen. The Hague: Dr W. Junk. Pp. 123-176. MACDONALD, D. W.: (1981): Resource dispersion and social organization of the red fox (Vulpes vulpes). In: Worldwide Furbearer Conference Proceedings. Ed. by J. A. CHapman and D. P. PursLEy. Frost- burg, Maryland: World Furbearer Conference 1980. Pp. 918-949. MACDONALD, D. W. (1983): The ecology of carnivore social behaviour. Nature 301, 379-384. MCFARLAND, D. (1990): Dictionnaire du comportement animal. Paris: Robert Laffont. MecnH, L.D. (1970). The wolf: the ecology and behavior of an endangered species. New York: Natural History Press. MEIA, J. S.; WEBER, J. M. (1992). Characteristics and distribution of breeding dens of the Red fox (Vulpes vulpes) in a mountainous habitat. Z. Säugetierkunde 57, 137-143. 268 J.-S. MeEıA and J.-M. WEBER MEIA, J. S.; WEBER, J. M. (1993): Choice of resting sites by female Foxes Vulpes vulpes in a mountainous habitat. Acta theriol. 38, 81-91. MEIA, J. S.; WEBER, J. M. (1995): Home ranges and movements of red foxes in central Europe: stability despite environmental changes. Can. J. Zool. 73, 1960-1966. MEIA, J. S.; AUBRY, S.; FERRARI, N.; LACHAT, N.; MERMOD, C.; WEBER, J.M. (1993): Observations noc- turnes au phare dans le Jura bernois: Septembre 1988-aoüt 1991. Mitt. Naturforsch. Ges. Bern 50, 193-202. MOoHR, €. ©. (1947): Table of equivalent populations of North American small mammals. Am. Nat. 37, 223-249. MULDER, J. L. (1985): Spatial organization, movements and dispersal in a Dutch red fox (Vulpes vulpes) population: some preliminary results. Rev. Ecol. (Terre Vie) 40, 133-138. NIEWOLD, F. J. J. (1976): Aspecten van het sociale leven van de vos. Natura 73, 234-241. NIEWOLD, F. J. J. (1980): Aspects of the social structure of Red fox populations: a summary. In: The red fox. Biogeographica. Vol. 18. Ed. by E. Zımen. The Hague: Dr W. Junk. Pp. 185-194. PAscAL, M.; MEYLAN, A. (1986): L’echantillonnage lin&aire des populations de la forme fouisseuse du campagnol terrestre (Arvicola terrestris scherman (Shaw)): description des techniques. Def. veg. 237, 3-12. PHiLLıps, M.; CATLing, P. C. (1991): Home range and activity patterns of red foxes in Nadgee Nature Re- serve. Wildl. Res. 18, 677-686. POULLE, M. L.; ArToIs, M.; ROEDER, J. J. (1994): Dynamics of spatial relationships among members of a fox group (Vulpes vulpes: Mammalia: Carnivora). J. Zool. (London) 233, 93-106. SARGEANT, A.B. (1972): Red fox spatial characteristics in relation to waterfowl predation. J. Wildl. Mgmt. 36, 225-236. SAUNDERS, G.; WHITE, P. C. L.; HARRIS, S.; RAYNER, J. M. V. (1993): Urban foxes (Vulpes vulpes): food ac- quisition, time and energy budgeting of a generalized predator. Symp. zool. Soc. London 65, 215- 234. SCHANTZ, T. von (1981): Female cooperation, male competition, and dispersal in red fox Vulpes vulpes. Oikos 37, 63-68. SCHANTZ, T. von (1984 a): Spacing strategies, kin selection, and population regulation in altrician verte- brates. Oikos 42, 48-58. SCHANTZ, T. von (1984b): “Non-breeders” in the red fox Vulpes vulpes: a case of ressource surplus. Oikos 42, 59-65. STECK, F.; WANDELER, A.; BICHSEL, P.; CAPT, S.; SCHNEIDER, L. (1982): Oral immunisation of foxes against rabies. Zbl. Vet. Med. B 29, 372-396. THURBER, J. M.; PETERSON, R. O. (1993): Effects of population density and pack size on the foraging ecology of gray wolves. J. Mammalogy 74, 879-889. WANDELER, A.; CAPT, S.; KAPPELER, A.; HAUSER, R. (1988): Oral Immunization of wildlife against rabies: concept and first field experiments. Rev. Infec. Dis. 10, 649-653. WASER, PM. (1981): Sociality or territorial defense? The influence of resource renewal. Behav. Ecol. Sociobiol. 8, 231-237. WEBER, J. M.; Augry, S. (1993): Predation by foxes, Vulpes vulpes, on the fossorial form of the water vole, Arvicola terrestris scherman, in western Switzerland. J. Zool. (London) 229, 553-559. WEBER, J. M.; AUBRY, S.; LACHAT, N.; MEIA, J. S.; MERMOD, C.; PARATTE, A. (1991): Fluctuations and be- haviour of foxes determined by nightlighting. Preliminary results. Acta theriol. 36, 285-291. WEBER, J. M.: MEIA, J. S.; AuBry, S. (1994): Activity of foxes, Vulpes vulpes, in the Swiss Jura Mountains. Z. Säugetierkunde 59, 9-13. WHITE, P. C. L.; Harrıs, S. (1994): Encounters between red foxes (Vulpes vulpes): implications for terri- tory maintenance, social cohesion and dispersal. J. Anim. Ecol. 63, 314-327. WOODROFFE, R.; MACDONALD, D. W. (1993): Badger sociality - modes of spatial grouping. Symp. zool. Soc. London 65, 145-169. ZABEL, C. J.; TAGGART, S. J. (1989): Shift in red fox, Vulpes vulpes, mating system associated with El Nins in the Bering Sea. Anim. Behav. 38, 830-838. ZIMEN, E. (1984): Long range movements of the red fox, Vulpes vulpes L. Acta Zool. Fennica 171, 267-270. Authors’ addresses: Dr. J.-S. MEıA, Rue de la Tour4, CH-2520 La Neuveville, Switzerland and Dr. J.-M. WEBER, Institut de Zoologie, Universit& de Neuchätel, Emile-Argand 11, CH-2007 Neuchätel, Switzerland. Z. Säugetierkunde 61 (1996) 269-275 ZEITSCHRIFT ® = © 1weGuafichenim _— ___________SÄUGETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Habitat use of Raccoon dogs, Nyctereutes procyonoides, in southern Finland By KAARINA KAUHALA Finnish Game and Fisheries Research Institute, Helsinki, Finland Receipt of Ms. 23. 01. 1996 Acceptance of Ms. 10. 05. 1996 Abstract Habitat use of raccoon dogs (Nyctereutes procyonoides) was studied in southern Finland during the snow-free seasons of 1990-1992 using radio tracking. Habitat selection within the study area and habitat use within the home range were examined. Raccoon dogs favoured shore areas especially during early summer. Shore areas with dense undergrowth provide food (e.g. frogs) and shelter, and raccoon dogs often escape into water when attacked. During autumn, raccoon dogs favoured moist heaths with abun- dant berries, which serve as an important food source before entering winter dormancy. The habitat use of raccoon dogs is thus affected by the availability of food, shelter and suitable den sites. Two features are common to dogs in all areas: 1) the are very often found near water and 2) during autumn they are more or less dependent on fruits and berries, which affects their habitat selection. Introduction In Europe the raccoon dog (Nyctereutes procyonoides Gray, 1834) is an introduced canid which has rapidly colonized Finland since the mid-1950s (HELLE and KAuHALA 1991). To- day its population density is rather high in southern and central Finland (KAUHALA and HELLE 1995), but very little is known of its effects on Finnish native fauna. Knowledge of the diet, home ranges and habitat use would provide background information on possible predation on or competition with native fauna. Home ranges of the raccoon dog in Fin- land have been studied previously (KAUHALA et al. 1993 a), and a preliminary study of its diet in this country has also been published (KAuHALA et al. 1993). The aim here was to examine habitat selection within the overall study area, and habi- tat use within its home range in southern Finland. Habitat use between seasons, mainly early summer and autumn, was also compared. Material and methods Habitat types The study area (about 30 km?) is located in the boreal zone (25°10’E, 61°14'N) around Evo Game Re- search Station in southern Finland, and consists mainly of industrial forests, small pıne swamps and many small lakes and streams. The forests were classified according to soil type as moist or barren heaths, and also by age: clear cuts (0-2 years), plantations (3-9 years), young forests (10-30 years) and old forests (>30 years). The basis for this classification lies in the undergrowth of each forest type. Moist heaths (including spruce swamps) are more productive than barren heaths (including pine 270 KAARINA KAUHALA swamps). Berries (mainly bilberries Vaccinium myrtillus and lingonberries V. vitis-idaea) grow abun- dantly in moist heaths, especially in those older than 30 years. Ferns grow abundantly in old spruce swamps, providing shelter for raccoon dogs, while grasses and some raspberries (Rubus idaeus) grow in plantations of moist heaths. Few lingonberries grow in barren heaths, where undergrowth consists mainly of lichens and heather. In addition to forests, the study area includes shore areas, farmyards, a fisheries research station and small fields. Shore areas included all habitats within 50 m of water. I included seven habitat classes in most analyses and figures (for habitat classes, see Fig. 1). Habitat use Habitat use of 12 adult (>1-year-old) raccoon dogs was investigated during early summer, midsummer and autumn of 1990-1992, using radio tracking. These animals lived as three pairs (Pair 1: female 33 and male 26, Pair 2: female 30 and male 23, Pair 3: female 140 and male 139) and six single individuals (males 99, 116, 150 and 141 and females 19 and 51). These single raccoon dogs probably also had mates, but these could not be caught. Raccoon dogs were caught with live traps or hounds and fitted with radio collars. They were located several times per week at different times of the day and night during snow- free periods (for details of radio tracking, see KAUHALA et al. 1993 a). The habitat of each fix (n= 961) was determined and its distance to the nearest lake, pond or stream was measured, while locations within 200 m of a known den were excluded. Harmonic mean home ranges (Dıxon and CHAPMAN 1980) were also calculated and the proportion of each habitat type in the 60% core area (mean size 1.3 km”) determined. Home ranges were calculated only for data sets with at least 35 independent locations (for testing independence, see KAUHALA et al. 1993 a). Thirty-five locations are sufficient for determining home range size (KAUHALA et al. 1993 a). Although the habitat use of individuals may differ, I first combined the data for all individuals to obtain an overall picture of habitat use in different seasons. I compared the frequency distributions of observed locations (number of locations in each habitat) with that of 532 random fixes from the study area (G-test; SOKAL and RoHLF 1981). Random fixes were obtained by determining the habitat at cross- points of a 200-m grid map overlay. I calculated a selection index for each habitat type to determine which habitats were especially favoured. The selection index is a ratio of the percentage of locations (L) to the percentage of habitat available (H): log L/H (see STorcH et al. 1990). When I compared the pro- portion of locations in a certain habitat with the proportion of random fixes in that habitat (i.e. tested the significance of the selection index), or compared the proportion of locations in each habitat during different seasons, I used the t-test (comparison of percentages). I studied the habitat use of the three pairs and the two single females in more detail (those cases with at least 35 independent locations per season). I combined the data of the male and female shar- ing the same home range (the pairs) because their habitat use did not differ (tested from frequency distributions of locations in each habitat, G-test). I compared the frequency distributions of locations of individuals/pairs with that of random fixes (G-test) and likewise the data for different seasons. I also compared the proportion of each habitat type in the 60% core area with the proportion of loca- tions in each habitat in the home range. The snow-free period was divided into three seasons: early summer or pup-rearing season (May-June), midsummer when pups begin to forage with their parents (July) and autumn when the young disperse and raccoon dogs prepare themselves for winter (August- October). Results Combined data for all raccoon dogs studied Habitat use of raccoon dogs differed between early summer and autumn (G =52.5, df=6, P<0.001); raccoon dogs used more old moist heaths and less shore areas and young barren heaths during autumn than in early summer (Fig. 1). The frequency distri- bution of locations differed in all seasons from that of random fixes (early summer G = 225.7, df = 6, P < 0.001, midsummer G = 64.3, df=6, P<0.001 and autumn G = 108.4, di=6, P= 0.001). Habitat use of raccoon dogs 271 EM earlysummer MM midsummer [| autumn || random fixes (N = 369) (204) (388) (532) 50 40 - — (dp) Ar e © 305 S (®) 7 © | & 20. 7 — 10 0 Mi HU = | | | clearcut young old clearcut young old shores Moist heath Barren heath Fig. 1. Habitat use of raccoon dogs around Evo Game Research Station, southern Finland, during early summer (May-June), midsummer (July) and autumn (August-October) of 1990-1992. The statistical differences between early summer and autumn are indicated by asterisks (t-test; * P< 0.05, ** P< 0.01, *** P<0.001). Random fixes were also obtained from the Evo research area. Clear cuts include 3-9-year-old plantations. Young forests are <30 and old forests >30 years old. 'e | — BE summer I] autumn >= clearcut Ss | (N = 369) (388) = young forest - ME __ uk 7) | [®) (***) > old forest - BE c clearcut - EEE T | & young forest - (——n 5 (***) | | old forest - (***) sn & au A%kx% T I ! T -0.6 -0.4 202 ) 0.2 0.4 0.6 0.8 Selection index Fig. 2. Habitat use of raccoon dogs in Evo research area, southern Finland, during early summer and autumn of 1990-1992. The selection index is calculated according to STORcH et al. ( 1990). A positive selection index means that the habitat is favoured, negative values indicate that the habitat is used less frequently than expected. Habitat use is compared with the availability of each habitat which is esti- mated on the basis of the distribution of 532 random fixes from the Evo area. The statistical significance is indicated by asterisks (t-test; * P< 0.05, ** P< 0.01, *** P< 0.001). If the habitat was used less fre- quently than expected the significance is in parentheses (see text). DR. KAARINA KAUHALA Shore areas were favoured in early summer and autumn, while young moist heaths were favoured in autumn (Fig. 2). Shores were favoured also in midsummer (P < 0.001). Information on which habitats were favoured is here considered crucial, because the pro- portions of locations in each habitat are not independent: if the animal spends much time in a certain habitat, ıt inevitably spends less time in other habitats (AEBISCHER et al. 1993). The mean distance of the locations from water was 191 m (SD = 170.2, n = 961). During early summer the distance was 156m (SD = 146.6, n = 369), in July 193m (SD = 151.9, n = 204), and during autumn 223m (SD = 192.6, n = 388). Random fixes were located a mean of 222m (SD = 161.0, n = 100) from the water. The mean distance from water was significantly smaller during early summer than in the other seasons. During early summer the distance was smaller than that of random fixes (P< 0.001, ANOVA). Habitat selection of different individuals/pairs Habitat selection on the basis of frequency distributions of locations (number of locations in each habitat) during early summer and autumn differed from random in all cases stud- ied (Tab. 1). Habitat selection also differed between seasons in the three cases in which it could be tested (Tab. 1) Table 1. Results of the G-tests for habitat selection/use of different raccoon dog individuals/pairs du- ring early summer (May-June) and autumn (August-October) in southern Finland during 1990-1992. Habitat use was compared with the frequency distribution of 532 random fixes from the study area. Pair/ind. Comparison to random Comparison between seasons Pair 1 Summer G=-583.0%- 670.001 G=590F4R 6 Autumn G,- 511.07 dt = O9EZ0:008 P < 0.001 Pair 2 Summer G= 8216. dt 67B2<0.001 G AS BUN 6 Autumn G = 48.3, df=6, P< 0.001 P < 0.05 Pain Summer &- 647,086, 20001 Female 19 Summer G 84.8201 - 08 Fig. 1. Geographic origin of Asian elephants included in the present study. Population genetics and systematics of Asian elephant 287 Upper Chidwan on the Myanmar side; 4. Southern Myanmar and northern Thailand, since a consider- able movement of elephants occurs between those areas; 5. Eastern Thailand and Vietnam, with a con- tinuous population until recent times (cf. SUKUMAR 1992). Amplification and sequencing of the mitochondrial cytochrome b gene An about 480 bp part of the mitochondrial cytochrome b gene was amplified via the Polymerase Chain Reaction (PCR) using the primers L 14724 and H 15149 (Irwın et al. 1991). Assay conditions were 2.5 U Tag-polymerase (Perkin Elmer), 0.2 uM of each primer, and 50 uM of each nucleotide ATP, CTP, GTP, and TTP in a reaction volume of 75 ul. The bulb end of a single hair was set directly into the reac- tion solution, without any prior treatment. After an initial denaturation step at 97°C for 15 min, 39 cy- cles were carried out with denaturation at 93°C for 1:30 min, annealing at 53°C for 1:15 min, and extension at 72°C for 1:30 min, followed by a final extension at 72°C for 3 min. The PCR-products were purified using a commercial kit (Oiagen) and cycle-sequenced using the Se- quilherm Cycle Sequeneing Kit (Epicentre) and the Digoxigenin-labelled oligonucleotide 5’-DIG- GCTITGATATGAAAAACCATCGTTG-3 for 30 cycles, each wıth a denaturation at 95°C for 30 s, an annealing at 52.7 °C for 30 s, and an extension at 70°C for 1 min. Samples were run on a direct blot se- quencing device (RICHTERICH et al. 1989). After blotting onto a nylon membrane, sequences were de- tected using an anti-Digoxigenin/Alkaline phosphatase conjugate (Boehringer) and the chemi- luminescent substrate CDP-Star (Tropix), following manufacturer’s instructions. Data analysis Based on 335 bp scored sequence of the cytochrome b gene, mitochondrial haplotypes were defined and compared in terms of pairwise nucleotide divergence (NEı and Jın 1989), using the Neighbor-Joining method (SAaıtou and Ne 1987) and the PHYLIP 3.5c computer package (FELSENSTEIN 1993). Published sequences of the African elephant (Loxodonta africana, Irwın et al. 1991), and of the Siberian woolly mammoth (Mammuthus primigenius, HAGELBERG et al. 1994) served as outgroups. A median graph of the relationships between mitochondrial haplotypes was determined according to BAnpeır (1992). Mean nucleotide diversities within and between geographic subsets of the analysed samples were calcu- lated (cf. Quinn and WHITE 1987). Standard errors of these values were estimated by bootstrap resam- pling of nucleotides for 1000 times. Haplotype diversities at different localities were calculated according to NEı and TayımA (1981). Results Compared with the corresponding Cyt b sequence in the Afrıcan elephant, the sequence of the Asian elephant differed by 18 to 25 nucleotide substitutions. Excluding the first 95 nucleotides not scored by HAGELBERG et al. (1994), Asıan elephant differed from both the African elephant and the Siberian woolly mammoth by 8 to 15 nucleotide substitu- tions. All but one substitutions were transitions. The only transversion (A &C at position 129) found separated the African elephant from the other two species (Fig. 2). In the 53 Asıan elephants analysed, a total of eight mitochondrial haplotypes were found, differ- ing from one another by one to six nucleotides at altogether seven polymorphic sites (Fig. 3). All seven mutations were transitions at the third codon position, without an ef- fect on the amino acid composition of the protein. Using the Neighbor-Joining procedure, haplotypes separated out by nucleotide diver- gence into two major clusters A and B, consisting of MAX I, II, III, V, VIII, and of MAX IV, VI, VII, respectively (Fig. 4). Each cluster contained one of the two most common mi- tochondrial haplotypes, MAX V (found in 18 specimens) and MAX VI (found in 12 speci- mens), respectively. Within each cluster, all other haplotypes can be derived by one to two transitions from these frequent haplotypes (Fig. 5). While the frequent haplotypes MAX V and MAX VI were distributed over the whole study area, some of the other hap- 288 G. B. Hart etal. lotypes (e.g. MAX I, MAX IV, MAX VIII) were found only in particular study units (Tab. 1). LO) 20 30 40 50 60 ATGACCGACATTCGAAAATCTCACCCCTTACTTAAAATCATCAATAAATCCTTCATTGAT 153,9 140 150 160 18740 180 CTAATTACACAAATCCTAACAGGATTATTCCTAGCCATACATTATACACCCGACACAATA 200 210) 220 230 240 ACTGCATTTTCATCTATATCCCATATTTGCCGAGATGTAAACTACGGCTGAATTATTCGA 25:0 260 270 280 290 300 CAACTACACTCAAACGGAGCATCCATTTTCTTCCTCTGCECTATACACACACATTGGACGA 310 320 330 AACATCTACTATGGGTCCTACCTATACTCGGAAACTTGAA Fig. 2. Nucleotide sequences of part of the Cyt b gene in Asian elephants (E. m.) in comparison to pub- lished sequences from one African elephant (1. a., Irwın et al. 1991), and two Siberian woolly mam- moths (M. p.1, M. p.2, HAGELBERG et al. 1994). Additional capital letters indicate polymorpkhic sites in Asıan elephant. Population genetics and systematics of Asian elephant 289 1.1.11 2223 0194671 238208 HEGIRE CH: SERIE CA.ATCC EA.A.CC Fig. 3. Sequence polymorphisms in the different haplotypes of Asian elephants. Position numbers are indicated by three vertical digits. MAX IV MAX VI B MAX VIl MAX II MAX VIII MAX III MAXI MAX V Pr Loxodonta Fig. 4. Sequence divergence among haplotypes of Asian ele- phant (Neighbor-joining tree, based on nucleotide divergence according to Neı and Jın 1989). The African elephant (Loxodonta) was used as an outgroup (according to our results the transition/transversion ratio was estimated to be 20/1). A and B refer to clusters of similar haplotypes as defined in Figs. 2 and 3. Fig. 5. Median graph according to BAnDELT (1992), showing relationships among haplotypes in terms of nucleotide substitutions (small bars). Supposed transient haplotypes not found in this study are indi- cated by circles (see Figs. 2 and 3 for definition of haplotypes). Nucleotide diversity and haplotype diversity, characterizing genetic variability within study units, are given in table 1. Haplotype diversity was highest in Sri Lanka and in northeastern India/northern Myanmar. The latter area also showed the highest nucleotide diversity. Genetic differentiation among study units in terms of nucleotide diversity is given in table 2. 290 G. B. Hart etal. Table 1. Geographic distribution of cytochrome b haplotypes (MAX I-MAX VIII) found in Asian ele- phant. Letters in parentheses refer to major groups of haplotypes as defined in Fig. 4. Total T = total number of haplotypes found in the respective samples. S SL = southern Sri Lanka, NE SL = northern and eastern Sri Lanka, S In (NE In) = southern (northeastern) India, N My (S My) = northern (south- ern) Myanmar, N Tha (S Tha) = northern (southern) Thailand, Viet = Vietnam. Numbers in parenth- eses refer to study units created by pooling of single samples. h = haplotype diversity, z = nucleotide diversity (in per cent) for the respective study units (standard error in parentheses). NE SL S In NEIn NMy S My NIhassEaTha Viet (1) (2) (3) (3) (4) (5) (5) nn 3 M=9d no M=S I (A) II (A) III (A) IV (B) V(A) VI(B) VIL(B) VIII (A) 2 total T: 3 h 0.84 0.67 0.75 - 0.91 0.77 0.75 (0.36) (0.33) (0.33) Table 2. Pairwise nucleotide diversity (in per cent, above diagonal) and net nucleotide diversity (in per cent, below diagonal) between study units (1-5, see Table 1) of Asian elephant (standard error in pa- rentheses). 0.90 (0.37) 1.18 (0.47) 0.88 (0.37) 1.19 (0.45) 0.06 (0.03) 1.21 (0.45) 0.90 (0.40) 1.06 (0.43) 0.13 (0.06) 0.22 (0.09) ie 1.17 (0.47) 1.38 (0.55) 0.06 (0.03) 0.18 (0.09) 0.22 (0.11) = 1.26 (0.53) 0.28 (0.10) 0.21 (0.09) 0.31 (0.12) 0.41 (0.17) 4 Discussion Cytochrome b variation in the Asian elephant Data available on sequence variation of the mitochondrial cytochrome b gene have proved to be inconsistent with respect to genetic variation within mammalian populations (see e.g. SHIELDS and KocHER 1991; Rannı et al. 1994 for bears; PErRY et al. 1995 for seals; CArr and Huches 1993 for deer; Wayne and JENKS 1991; GIRMAN et al. 1993 for wild canids). Population genetics and systematics of Asian elephant 291 Using restriction enzymes for cutting a 2,450 bp fragment of mtDNA, GEORGIADIS et al. (1994) detected considerable genetic variation within and among populations of the African elephant (Loxodonta africana). This result was interpreted to indicate a large long-term effective population size and high gene flow between subpopulations. Exhibit- ing a total of 7 variable sites and considerable haplotype diversity within each of the local populations studied, also the Asian elephant, assumed to originate from Africa and to have invaded southern and southeastern Asia during the Pliocene (cf. HAynes 1991), can be considered at least moderately polymorphic. Despite a remarkable decline and increas- ing isolation of local Asian elephant populations in historical times due to human exploi- tation (Kurt et al. 1995), the observed variation may indicate the ancient effective population size to have been large. Relationships among haplotypes in comparison with their geographic distribution Haplotype MAX VI was previously described in six Asıan elephants without mentioning their geographic origin (HAGELBERG et al. 1994). This haplotype is the one most closely re- lated to those of both the African elephant and the Siberian woolly mammoth, and taking also into account its widespread geographic distribution it must be considered a very bas- ic one. According to their geographic distribution the same holds for haplotype MAX V and, although being absent or rare in the central and easternmost part of our study area, respectively, for haplotypes MAX III and MAX II. Haplotype MAX I, apparently con- fined to Sri Lanka may have locally arisen through a new mutation, which is in agreement with its marginal position in the median graph. However, due to their more central posi- tion in the median graph, the other haplotypes found only in single study units (MAX IV, MAX VII) or in a group of adjacent study units (MAX VII) may rather represent relics of a formerly more widespread polymorphism. Altogether, the disjunct distribution of many haplotypes suggests that they had already been present at the time when elephant colonized southern and southeastern Asia during the Pliocene (c. f. HAynes 1991). This ar- gument is supported by the presence of several well differentiated haplotypes at Cytb, a generally rather conservative gene with respect to varıation at the population level (cf. Avıse 1994). The alternative hypothesis, a segregation of haplotypes MAX IV and MAX VIII in the Indian subcontinent, and of haplotype MAX VII in Southeastern Asia, respectively, from a basic set consisting of haplotypes MAX VI, MAX V, MAX II, and MAX III in the original founder population would imply a considerable extent of homo- plasy, which is unlikely to have arisen at the population level. Genetic diversity within populations The assumption of local losses of Cyt b haplotypes in Asian elephant is supported by lev- els of nucleotide and haplotype diversity found in extant populations. However, it is inter- esting to note that, in a comparison of populations studied, estimates of nucleotide diversity do not correspond with the respective levels of haplotype diversity. This can be explained by both the number of haplotypes and the extent of nucleotide divergence among haplotypes being involved in the calculation of nucleotide diversity. The largest number of haplotypes is found in Sri Lanka, but with the exception of two animals with haplotype MAX VlI all of them belong to group A. Hence, the comparatively high nucleo- tide diversity in that area mainly reflects a large number of haplotypes present. In north- ern India/northern Myanmar the number of haplotypes found is smaller than in Sri Lanka, but the proportions of haplotypes belonging to group A and B, respectively, the two being separated by two mutation steps, are more similar than in the previous case. In eastern Thailand/Vietnam the high nucleotide diversity is almost exclusively due to a large nucleotide divergence among only a few haplotypes. As a methological consequence 292 G. B. Hart etal. we suggest that both haplotype diversity and nucleotide diversity are essential for a prop- er characterization of genetic variability in populations. The most striking result with respect to levels of genetic variability is the high haplo- type diversity present on the island of Sri Lanka as compared to the respective values found in most areas of the Indian subcontinent and in southeastern Asia. During the last five centuries, nowhere in the range of the Asıan elephant has capturing and killing by Mogul and colonian forces been carried out as extensively as in the area between south- ern and northeastern India. There numbers of elephants have declined from at least 375,000 at 330 BC to less than 20,000 at present (i.e. to about 5 per cent of the original population). In southern Myanmar and in Thailand large numbers of elephants had been caught for timber extraction between the late 18th and the early 20th century (Evans 1910; STRACEY 1963). In addition it must be stressed that, due to habitat destruction by man, on the Indian subcontinent the remaining elephant population is subdivided into a number of mostly small isolates, which are exposed to genetic drift (Kurt 1992, KURT et al. 1995). A similar situation prevails in the region from southern Myanmar to Vietnam (SANTIAPILLAI and JACKSoN 1990). By contrast, in Sri Lanka still some 20 per cent of the original elephant population of about 12,000 individuals are present, and human encroachment hardly prevented migra- tion across the whole island until the middle of this century (Kurr 1992). Also in north- eastern India and northern Myanmar, harbouring the second highest haplotype diversity, human impact on population size of elephant had been comparatively low. Altogether, the different levels of genetic variation found in populations of Asian elephant are in good agreement with records on population histories. Genetic diversity among populations Although there seems to be a tendency of haplotypes belonging to group A to decrease from the western to the eastern part of our study area, haplotype frequencies and esti- mates of net nucleotide diversity among samples do not correspond very well with the geographic distribution of the local populations studied. A certain bias due to limited sample sizes of individuals cannot be excluded, but in general this result seems to reflect a considerable extent of genetic drift caused by human impact on population size and structure. Implications for systematics and conservation of Asian elephant Estimates of net nucleotide diversity among populations do not reveal a major separation of the Sri Lanka population from all other study units. Thus, our data do not provide sup- port for subspecies distinction between elephants on Sri Lanka and on the Asian main- land, which is in accordance with the electrophoretic data of HartL et al. (1995). In the light of our present results the apparently fixed difference at the Sod-2 locus reported by NozawA and SHOTAKE (1990) can be explained by short-term drift processes in local popu- lations rather than by a long-term divergence between subspecies. Regarding breeding of Asian elephants in zoos differentiation among wild-living popu- lations, especially between those of Sri Lanka and southern India, formerly believed to re- present different subspecies, seems to be low enough to justify cross-breeding of parents of geographically different origin. Future management of wild-living and captive popula- tions in Asia should follow the traditional Sri Lankan practice, where the off-take from the wild-living population for maintaining considerable captive stock has been as high as necessary but low enough for preserving the highest level of genetic variability as yet known (see also Kurr et al. 1995). Population genetics and systematics of Asian elephant 293 Acknowledgements Collection of samples was made possible by H. E. the Minister of Forestry Lt. Gen. CHırt SwE (Yangon); Lyn DE Aıwıs, Chairman of the Asıan Elephant Specialist Group of the IUCN (Colombo and Singapur); Dr. (Mrs.) JAYANTHI ALAHAKOON (Colombo); Dr. VısIt ARSAITHAMKUL (Bangkok); Prof. J. V. CHEERAN (Trichur); Dr. C. HAGENBEcK and K. Kock (Hamburg); Dr. S. K. KARUNARATHE (Colombo); Dr. ALONG- KORN MAHANNOP (Bangkok); U Myar Win (Yangon); Dr. G. RUEMPLER (Münster); Prof. K. C. PANICKER (Trichur); Dr. Wısıp WıcHasiLpa (Bangkok). This research was supported by the R. and R. SCHLAGETER Foundation (Zoo Zürich), and by A. HAUFELLNER (European Elephant Group, Munich). Zusammenfassung Populationsgenetik und Systematik des Asiatischen Elefanten (Elephas maximus): Eine Studie auf der Grundlage von Sequenzvariation am Cyt-b-Gen von PCR-amplifizierter Mitochondrien-DNA aus Haarwurzeln Um die genetische Variabilität und Differenzierung bei Asiatischen Elefanten (Elephas maximus) zu charakterisieren, wurde bei insgesamt 53 Individuen aus Sri Lanka, dem Süden und Nordosten Indiens, dem Norden und Süden von Myanmar (Burma), dem Norden und Osten von Thailand sowie aus Viet- nam die Sequenzvariation am mitochondrialen Cytochrom-b-Gen untersucht. Ein ca. 480 bp langer Ab- schnitt des Gens wurde mittels PCR aus Haarwurzeln amplifiziert und im Thermocycler direkt sequenziert. Nach Auftrennung in einer Direkt-Blot-Sequenzieranlage wurden 335 bp analysiert. Sie- ben Sequenzpositionen waren polymorph, und acht Haplotypen wurden definiert, die sich durch eine bis sechs Transitionen an der dritten Codon-Position unterschieden. Die Haplotypen wichen im Ver- gleich zu publizierten Sequenzen an mindestens 18 Positionen vom Afrikanischen Elefanten (Loxo- donta africana) und an mindestens 8 Positionen vom Mammut (Mammuthus primigenius) ab. Entsprechend ihrer Nukleotiddivergenz ließen sich die Haplotypen zwei Gruppen zuordnen, die sich in ihrer geographischen Verbreitung unterschieden. Die Divergenz zwischen Elefanten von Sri Lanka und dem asiatischen Festland, die bisher als eigene Unterarten beschrieben worden waren, war zum Teil kleiner als zwischen Elefantenbeständen auf dem asiatischen Festland. Die untersuchten Bestände zeig- ten deutliche Unterschiede hinsichtlich der Haplotypen- und Nukleotiddiversität. Die phylogeogra- phischen Muster der Haplotypen legen nahe, daß die Asiatischen Elefanten zur Zeit ihrer pliozänen Einwanderung nach Süd- und Südostasien eine zusammenhängende Population gebildet haben. Unter Berücksichtigung der jeweiligen Bestandesgeschichte läßt sich die gegenwärtige geographische Differ- enzierung als Folge einer unterschiedlich starken menschlichen Einflußnahme auf die Bestandesgrößen und -strukturen interpretieren. References AVISE, J. C. (1994): Molecular Markers, Natural History and Evolution. New York, London: Chapman and Hall. BANDEILT, H.J. (1992): Generating median graphs from Boolean matrices. In: L1-Statistical Analysıs. Ed. by Y. Donce. Elsevier: North Holland. Pp. 305-309. DRYSDALE, T. A.; FLORKIEWICZ, R. F. (1989): Electrophoretic variation within and between the two ex- tant elephant species (Mammalia: Proboscidea). J. Mammalogy 70, 381-383. CARR, S. M.; HucHes, G. A. (1993): Direction of introgressive hybridization between species of North American deer (Odocoileus) as inferred from mitochondrial-cytochrome-b sequences. J. Mamma- logy 74, 331-342. Evans, G. H. (1910): Elephants and their Diseases. Rangoon: Government Printing. FELSENSTEIN, J. (1993): PHYLIP (Phylogeny Inference Package) version 3.5c. Seattle: Dep. Genetics, Univ. Washington. GEORGIADIS, N.; BISCHOF, L.; TEMPLETON, A.; PATTON, J.; KARESH, W.; WESTERN, D. (1994): Structure and history of African elephant populations: I. Eastern and Southern Africa. J. Heredity 85, 100-104. GIRMAN, D. J.; KAT, P. W.; MırLs, M. G. L.; GINSBERG, J. R.; BORNER, M.; WILson, V.; FANSHAWE, J. H.; FITZGIBBoN, C.; LAU, L. M.; WAYNE, R.K. (1993): Molecular genetic and morphological analyses of the African wild dog (Lycaon pictus). J. Heredity 84, 450-459. 294 G. B. Harrtı etal. HAGELBERG, E.; THOMAS, M. G.; Cook, C. E. Jr.; SHER, A. V.; BARYSHNIKOV, G. F.; LiSTER, A.M. (1994): DNA from ancient mammoth bones. Nature 370, 333-334. HARTL, G. B.; KURT, F.; HEMMER, W.; NADLINGER, K. (1995): Electrophoretic and chromosomal variation in captive Asian elephants (Elephas maximus). Zoo Biology 14, 87-95. HAYyNnes, G. (1991): Mammoths, Mastodonts, and Elephants — Biology, Behavior, and the Fossil Record. Cambridge: Univ. Press. Irwin, D. M.; KocHERr, T. D.; Wırson, A. C. (1991): Evolution of the Cytochrome b gene of mammals. ]. Mol. Evol. 32, 128-144. Kurr, F. (1992): Das Elefantenbuch. Hamburg: Rasch und Röhring. Kurt, F.; HARTL, G. B.; TIEDEMANN, R. (1995): Tuskless bulls in Asian elephant Elephas maximus. His- tory and population genetics of a man-made phenomenon. In: Ecological genetics in mammals II. Ed. by G. B. HArTL and J. MARKOWSKI. Acta Theriol. Suppl. 3, 125-143. McKay, G.M. (1973): Behaviour and ecology of the Asian elephant in Southeastern Ceylon. Smithso- nian Contr. Zool. 125, 1-111. NeEı, M.; TAsımA, F. (1981): DNA polymorphism detectable by restriction endonucleases. Genetics 97, 145-163. Neı, M.; Jın, L. (1989): Variances of the average numbers of nucleotide substitutions within and be- tween populations. Mol. Biol. Evol. 6, 290-300. NOZAWA, K.; SHOTAKE, K. (1990): Genetic differentiation among local populations of Asian elephant. Z. zool. Syst. Evolut.-forsch. 28, 40-47. PERRY, E. A.; CARR, S. M.; BARTLETT, S. E.; DAvıpson, W. S. (1995): A phylogenetic perspective on the evolution of reproductive behavior in pagophilic seals of the Northwest Atlantic as indicated by mi- tochondrial DNA sequences. J. Mammalogy 76, 22-31. Quinn, T. W.; WHITE, B. N. (1987): Analysis of DNA sequence variation. In: Avian Genetics. Ed. by F. Cooke and P. A. Buckre£y. London: Academic Press. Pp. 163-198. RANDI, E.; GENTILE, L.; BoscAGLI, G.; HUBER, D.; ROTH, H. U. (1994): Mitochondrial DNA sequence di- vergence among some west European brown bear (Ursus arctos L.) populations. Lessons for conser- vatıon. Heredity 73, 480-489. RICHTERICH, P.; HELLER, C.; WURST, H.; Pont F.M. (1989): DNA sequenceing with direct blotting electro- phoresis and colorimetric detection. BioTechniques 7, 52-58. SAITOU, N.; NEI, M. (1987): The neighbour joining method: A new method for reconstructing phyloge- netic trees. Mol. Biol. Evol. 4, 406-425. SANTIAPILLAI, C.; JACKSON, P. (1990): The Asian Elephant. An Action Plan for its Conservation. Gland: IUCN, WWE. SHIELDS, G. F.; KOCHER, T. D. (1991): Phylogenetic relationships of North American ursids based on analysis of mitochondrial DNA. Evolution 45, 218-221. STRACEY, P. D. (1963): Elephant Cold. London: Weidenfeld and Nicholson. SUKUMAR, R. (1990): Ecology of the Asian elephant in southern India. II. Feeding habits and crop raid- ing patterns. J. Tropical Ecol. 6, 33-53. SUKUMAR, R. (1992): The Asian Elephant: Ecology and Management. Cambridge: Univ. Press. WAYNE, R. K.; JENKS, S. M. (1991): Mitochondrial DNA analysis implying extensive hybridization of the endangered red wolf Canis rufus. Nature 351, 565-568. Authors’ addresses: Prof. Dr. G. B. HARTL, Dr. F. Kurt, Dr. R. TIEDEMANN, Mag. CHRISTINE GMEINER, and Dr. K. NADLINGER, Institut für Haustierkunde der Christian-Albrechts-Univer- sıtät zu Kiel, Olshausenstraße 40, D-24118 Kiel, Germany; Dr. KHyNE U MAR, Myanmar Timber Enterprise, Ahlone, Myanmar; Dr. A. RüßeL, Zoo Zürich, Zürichbergstraße 221, CH-8044 Zürich, Switzerland. an BA W; RR INTER Z. Säugetierkunde 61 (1996) 295-303 ZEITSCH RIFTS Fr FÜR © 1996 Gustav Fischer, Jena SÄUG ETI E RKÜ N D E INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Patterns of variation in cranial size and shape in two coexisting gerbilline rodents By J. ©. Marson and D. P. CHRISTIAN Department of Biological Sciences, San Jose State University, San Jose, California and Department of Biology, University of Minnesota, Duluth, Minnesota, USA Abstract Patterns of cranial variation within and between two coexisting gerbilline rodents from the Namib De- sert were assessed. Size of individual cranial characters is more variable in Desmodillus auricularis than in Gerbillurus paeba, but G. paeba is more variable in aspects of morphological shape. This contrasting pattern of variability in size and in shape has implications for ecomorphological studies. Without infor- mation on the relative functional importance of these two aspects of morphology, there appears to be little basis for using data on morphological variation to derive ecological generalizations. It is suggested that the strikingly different degree of correlation among cranial measurements in these two species is related to differences in growth rates and demographic seasonality. Introduction A variety of hypotheses have suggested relationships between intrapopulation morpholo- gical varıability and ecological patterns. One of the most popular and widely-tested is the “niche variation” hypothesis (VAn VALEN 1965 a) which argues that populations occupying wider niches should be morphologically more variable than those with narrower niches (see also HAamırron and JoHNnsSToNn 1978; PATTERSON 1981, 1983; ROTHSTEIN 1973; SOULE and STEWART 1970; VAn VALEN and GRANT 1970; Wırrson 1969; Wırıson et al. 1975). Other examples of hypotheses relating morphological variation and particular aspects of ecology were summarized in HESPENHEIDE (1973), WEINs and ROTHENBERRY (1980), WER- NER and GILLIAM (1984) and FuruyMmA and MoRrENoO (1988). Tests of these hypotheses have yielded varying outcomes and a number of authors (PATTERSoN 1881, 1983; ROTH- STEIN 1973; SOULE and STEwART 1970, WEINsS and ROTENBERRY 1980; WıLıson et al. 1975) have discussed problems with the formulation of these hypotheses or the way they com- monly have been tested. Recently, SCHMITT et al. (1995) suggested the “niche-width” (= niche variation) hypothesis to explain the association of genetic and ecological variabil- ity in fruit bats in the Indonesian Archipelago. However, their results indicated no corre- lation with morphological variability. A basic premise of these ideas is that morphologically different individuals within a population differ in patterns of resource use. The degree of difference among individuals (1.e., variation) should thus parallel the variety of a population’s ecological patterns (e.g., geographical range, range of prey sizes, or some more general measures of niche breadth). The expectation that morphological variation should translate into functional varlation is entirely reasonable. However, most tests of these hypotheses have rested on the additional assumption that functional variation is most importantly influenced by var- lation in size (sensu stricto) of morphological structures. Constraints of body size and size of particular structures, of course, are of unquestionable functional importance. Several 296 J. ©. Marson and D. P. CHRISTIAN authors have demonstrated that differential resource use (either between species or among individuals of a single species) may be related to variation in linear dimensions of anatomical traits (most frequently examined have been trophic structures; ABBOTT et al. 1975, GRANT 1981; HERRERA 1978; PULLIAM and EnDERS 1971; Wiırrson 1971). In other cases, however, relationships between variation in size and variation in function have been assumed but not directly tested (see PATTERSON 1983 for a review). The shortcoming of this approach is illustrated by the existence of structures in which size per se may not be of clearly greater functional importance than aspects of morphology like shape or rela- tive proportions of component parts (for mammals see BROWNFIELD and WUNDER 1976; HERRING and HERRING 1974; NELSON and SHUMP 1978; FREEMAN 1979; 'THORINGTON and HEANEY 1981; and, SmıtH 1993; RusseL and THoMAson 1993 for reviews). In this study we present data on contrasting patterns of variation in size and shape or proportion of cranıal characters in two coexisting gerbilline rodents from the Namib De- sert of Namibia. The results demonstrate the virtual necessity of understanding the rela- tive functional importance of sıze/shape variation in eco-morphological studies. Additionally, we discuss possible developmental and demographic correlates of mor- phological differences between these two species. Material and methods Specimens Examined The two species of gerbilline rodents considered were Gerbillurus paeba (Smith 1836) and Desmodillus auricularis (Smith 1834). Specimens used in assessing morphological variability were collected at Gorra- sis (25°18°S, 15°55’E) on the eastern edge of the Namib Desert in Namibia. Specimens were collected in a single habitat (the bush plain habitat of Christian 1979 a, 1980) more or less evenly throughout the period between March 1974 and April 1975. Our sample is, thus, temporally and spatially very consis- tent. Cranial sutures were well ossified in all specimens and there was at least slight wear on the molar teeth. Additionally, body weights of all specimens were well above minimum weights of reproductively active animals observed in the field. Therefore we feel that these specimens were both morphological and potentially reproductive adults. Precipitation essentially failed completed on the study site a year prior to the start of the field work during which the animals were collected. Rodent populations con- sisted of only widely scattered individuals when field work began in 1974 (Christian 1979a, 1980). Thus, the sample consisted almost entirely of animals born during 1974-1975. Morphological Characters Several of the morphological characters we selected almost certainly have a direct relationship to feed- ing (incisor width, length of maxillary and mandibular toothrows; SmitH 1993) and predator detection (length and width of auditory bulla; LAy 1972). While the functional significance of the other characters chosen (skull length, zygomatic breadth, and diastemal length) is less clear, these structures are likely related to feeding (sensu lato; see SmıtH 1993 for a review of morphological studies of the mammalian feeding apparatus). The eight cranial characters were measured with dial calipers to the nearest 0.1 mm. To insure accu- racy of the measurements, each specimen was measured at least twice, one week apart. If the second measurement deviated by 0.1 mm or more, it was measured a third time. In our second series, less than 10 per cent of the original measurements had to be repeated. | Analyses The mean body weight of specimens was 24 g in G. paeba and 45 g in D. auricularis. Because of this size difference, we log-transformed the raw data before computing variances (LEwonTIN 1966; for ratio- nale, see MorIaArıTY 1977). The log-transformations scale the raw data so that variances of measure- ments of different magnitude are comparable while preserving the relative variability. Using log- transformed data we computed variances for each character in each species/sex class. These variances Patterns of variation in cranial size and shape in two coexisting gerbilline rodents 297 were compared between sexes within each species and within sexes between species by a series of F- tests. One technique used to examine questions about variation in shape or proportion was to compute Pearson product-moment correlation coefficients (r) between each pair of characters within each sex and species. We assume that a high degree of correlation among cranial characters is indicative of rela- tively low variation among individuals in cranial shape. Differences between species and sexes in the distribution of significant F-tests and correlation coefficients were tested for independence by Chi- square. Another method to examine relationships among cranial characters was to compute the proportion of variation unique to each character. This was accomplished by performing a Principal Factor Analysis (PFA) in which we forced the procedure to extract eight new multivariate factors (NIE, et al. 1975). The procedure calculated an estimate of the communality (h”) for each character. The communality ex- presses the proportion of the total variance of each character that is due to intercorrelations with the other characters. From this value, we calculated a uniqueness index (1-h?) for each character (i. e., the proportion of the total variance of that character that cannot be attributed to correlations with the other characters). Results The distribution of measurements for both species gave no indication of bi- or poly-mod- ality, indicating that the size variation discussed here ıs continuous. Univariate compari- sons showed no significant difference in variability between sexes within each species. Univariate comparisons by sex between species indicated that all variances were larger in D. auricularis than in G. paeba (Tab. 1). F-tests on these variances indicated that D. auricularis was significantly more variable than G. paeba in six of the eight characters for females and in seven of the eight characters for males. A Chi-square test of independ- ence on the distribution of significant F-test results was significant for either probability of the F-tests (X° > 4.00, P < 0.05, df = 1). These results indicate that the number of signifi- cantly larger variances in D. auricularis was greater than would be expected due to chance alone. Thus, in terms of size of cranıal characters, D. auricularis is clearly more variable than G. paeba. Table 1. Variances of log-transformed measurements of cranial characters in male and female Desmo- dillus auricularis and Gerbillurus paeba. Probability (P) based on F computed as S? for D. auricularis divided by S° for G. paeba. Males Females D. auricularius G. paeba D. auricularius G. paeba CHARACTER N SS N BIN ENKORES "OMAN IERLS? Skull Length (SL) 17 0.002504 0.000738 0.002319 0.000559 0.3). The number of significant correlations was greater in female D. auricularis than in G. paeba (X” 24.1, P< 0.05 for either level of significance of r, df= 1). In males, there were more significant r’s in D. auricularis than in G. paeba (X? = 3.27, 0.05 length (LM2i); M3 length (LM3i); M3 width (LaM3i); length of the lower toothrow (LM13i); length of the first upper molar (LM1s); M? length (LM2s); length of the upper toothrow (LM13s). Only adult individuals were measured. In order to estimate age of the animals under study, the con- dition of their reproductive organs, their weight (LAUrRIE 1946; PELIKAN 1981), and the level of abrassion of their molars (KELLER 1974) were assessed; where possible, a combination of the three approaches 306 M. MACHOLAN was taken into account (see MACHOLAN 1996, for details). All individuals of doubtful age were exluded from subsequent investigations. Multiple group principal component analysis (MGPCA, THorPE 1983 a), based on a pooled within- group varlance-covariance matrix, was used in order to assess the contribution of within-group compo- nents to the between-group discrimination. The ‘size’ vector was searched in order to be eventually ex- tracted from the data. This ‘size’ component is generally the eigenvector corresponding to the first principal component; however, three conditions have to be met for such an assumption: an eigenvector expressing general size should have coefficients of the same sign (1) and ‘similar’ magnitude (2); and the first principal components within localities should have the same orientation (3). The latter condi- tion can be tested by comparing the first eigenvectors across localities. Since substantial between-character differences due to different scales in individual variables was expected, the contribution of each character to a component was compared by computing the pooled within-population correlation between the character and the component score according to the formula (THorPE 1983 a): a where r;; is the pooled within-group (within-population) correlation between the ith character and the jth eigenvector, a;; is the coefficient for ith character for the normalized jth eigenvector, 4; is the latent root (eigenvalue) of the jth eigenvector and s; is the pooled within-group standard deviation of the ith character. Before computing the correlations, the eigenvectors were normalized so that each compo- nent coefficient was divided by Dar, where a;; is as defined above and k is the number of characters. Correlation coefficients were then compared and their significance was tested (SoKAL and ROHLF 1981; THoORPE and LEAMY 1983). Since for p close to #1.0, the distribution of sample values of r is mark- edly asymmetrical, we have to transform r to a function z; standard normal deviate value t, is then de- 1 1 i;E = N ander wm En I r is the correlation coefficient as defined 2(n; —3) an above, k is the number of populations, and n; is the sample size of the ith population. Since z is approxi- mately normally distributed and we are using a parametric standard deviation, t, is compared with t.7-] (where a = 0.01). Two techniques were employed in order to extract the size vector: one produces new data as princi- pal component scores with the first eigenvector removed (THoRPE 1983 b), whereas the other is based on BURNABY’S (1966) adjustment as suggested in ROHLF and BooKSTEIN (1988). Population interrelationships were assessed by subjecting the component scores to canonical variate analysis (CVA, FısHEr 1936). This multivariate ordination method separates groups so that between- group variation is maximized while within-group variation is minimized (CAMPBELL and ATCHLEY 1981). As multiple-group PCA uses pooled within-group covariances, CVA performed on all of the MGPCA component scores (‘size-in’ analysis) gives the same results as CVA on the original data (THoRrPE et al. 1982; THoORPE 1983 a). CVA computed on MGPCA component scores with the ‘size’ vector extracted (‘size-out’ analysis) revealed the same results as CVA performed on BurNABY-adjusted data. Matrices of Mahalanobis generalized distances D’, computed as a part of canonical variate analysis, were employed both in the Mantel test comparing the results of ‘size-in’ and ‘size-out’ multivariate ana- lyses, and subjected to cluster analysis. The System for Statistics (SYSTAT, Release 5.02, Wırkınson 1990) and Numerical Taxonomy Sys- tem (NTSYS-pe, Version 1.60, RoHLF 1990) packages were used for all the statistical analyses. fined as Z/o,, where o, = Results As stated above there are three assumptions for the first principal component, extracted from the pooled within-population covariance matrix, to be treated as the ‘size’ vector. The first principal component within localities appeared to be of the same orientation as substantiated by checking the signs of the first eigenvectors for each locality. Further- more, the coefficients corresponding to the first principal axis were all of the same sign Multivariate morphometric analysis of European species of the genus Mus 307 (see the first column in Tab. 1). However, differences in their magnitude were strikingly high. Since the (pooled) variance-covariance matrix was used, the relative magnitude of the coefficients was dependent on the variances of the original data, i.e., on the scale of the respective characters. Therefore, two transformation techniques were used in order to decrease the differences in the variances: firstly, the variates were converted to loga- rithms, and secondly, the data were normalized by subtracting the mean and dividing by the standard deviation as provided by the standard SYSTAT routine. Either transformation of data is only feasible under the expectation of a general im- provement in linearity. Since non-linear relations between variables would result in a low- er inter-character correlation, an improvement in linearity should generally be apparent by a larger first eigenvalue in the correlation matrix. As shown in table 1, both the trans- formations yielded a slight (although insignificant) decrease in curvilinearity between the variables. If we compare the total variance explained by the first principal component for the three data sets we can see a steady decrease in the percentage from the original to the normalized data sets. Although logging the variates reduced the differences among individual component coefficients of the first PC their magnitude still remained highly heterogeneous. More- over, whereas six characters showed insignificant character-component correlations in the raw data, this number was increased to as many as nine in log transformed characters (all of them being the tooth measures). Hence, it is obvious that log-transforming data may Table 1. Principal component coefficients of the first normalızed eigenvector of MGPCA (left col- umns) and character-component correlation coefficients (right columns) for raw (RAW), log-trans- formed (LOG) and normalized (NORM) data. Nonsignificant correlations are in parentheses. Below, the percentage of the total variance explained by the first principal component, the variance explained by the first three components (all the eigenvalues being extracted from the covariance matrix), and the proportion of the first eigenvalue computed from the correlation matrix, respectively, are given. Character Coefficients/Correlations LOG A 0.033 0.47 0.774 0.89 0.194 0.50 B 0.038 0.50 0.392 0.66 0.169 0.54 LCb 0.652 1.00 0.136 0.60 0.397 0.90 LB 0.632 1.00 0.150 0.62 0.411 0.90 LaR 0.097 0.59 0.134 0.48 0.312 0.70 LaC 0.128 0.57 0.062 0.39 0.313 0.70 LaZ 0.296 0.80 0.130 0.56 0.327 0.80 hC 0.084 0.38 0.056 0.25 0.224 0.52 LD 0.229 0.88 0.161 0.50 0.341 0.78 LMIı 0.004 (0.10) 0.013 (0.07) 0.119 032 LaMlı 0.007 0.23 0.025 (0.11) 0.149 0.42 LM2i 0.005 (0.13) 0.016 (0.05) 0.137 0.35 LM3i 0.002 (0.06) 0.013 (0.03) 0.082 0.25 LaM3i 0.002 (0.06) 0.013 (0.04) 0.128 0.30 LM13i 0.013 0.16 0.014 (0.07) 0432 0.43 LMIs 0.000 (0.01) 0.012 (0.04) 0.084 0.22 LM2s 0.004 (0.09) 0.027 (0.08) 0.117 0.27 LM13s 0.017 0.18 0.021 (0.11) 0.151 0.43 Ist V eigenvalue 81.09% 42.24% 33.80 % 3 V eigenvalues 91.45% 68.02% 61.21% lst C eigenvalue 31.32% 53221820 31.61% 308 M. MACHOLAN not adequately standardize the varıance of the characters. In addition, it is apparent from table 1 that this transformation substantially changed the relative contribution of indivi- dual characters to the total variation explained by the component (cf. the coefficients and correlations in the RAW and LOG columns of the Tab. 1). On the contrary, MGPCA computed on the normalized data revealed all the correla- tions to be signifcant although the magnitude of the coefficients was neither the same nor ‘sımilar’. These results suggest that although the differences in the magnitude of the coef- ficients of the first principal component were partly due to differences in character var- iances, there was still some portion of persistent variance which could not be associated with the ‘size’ component: this especially concerned the dental measures. Therefore, the size-adjustment must be used with caution as these results indicate that the first principal component is also very likely to contain some ‘shape’ information which would be lost on 4 1 T T UKR 2 - | Hass AUT mu | | MH | | DA ak M Ga > 0’E PT yMM = 6) | | I Mc | SPR *°® DCH | | | m MSP 2 en E | Rn —4 | | i IL -6 4 -2 0 2 4 a CV1 Fig. 1a 4 a UKR | Ju > AUT | MS DA MU | MM | Q | | ee | 5 on - | ; MC 2 | | | | SPR ! DCH | | | f* Ms Ar | en GR 4 I -6 —4 => 0 2 4 b CV1 Multivariate morphometric analysis of European species of the genus Mus 309 Gransee Hin area T au AUT a MS UKR en Fesiu (N N MH San Gi N DA MM 5 SPR \ DCH N m MSP u / —5) 8 -6 | 2 GR Dun 4 | je L | 5 A -2 0 2 4 c CV1 Fig. 1c 4 T a AUT Ik %% x UKR MS \ DR m N \ MH > -0-}+ 5 \ \ MM \ DCH # —* Msp SPR N | za a GR 4 | se | | | -6 4 =2 0 2 4 d CV1 Fig. 1d Fig. 1. Plots of the first two canonical variate scores for group centroids; (a), (b)-'size-in’ analysis; (c), (d) — ‘size-out’ analysis; plots (a) and (c) show results of the log-transformed data, whereas (b) and (d) concern the normalized data. Minimum spanning trees were superimposed on the plots. extraction of the ‘size’ vector; on the other hand, we may expect a small part of residual size information to be contained in following (intentionally ‘size-free’) components. Thus results of both the 'size-in’ and ‘size-out’ analyses should be taken into account (accord- ingly, these terms are rather loose and henceforth will only be used for convenience). Figure 1 shows plots of scores for the first two canonical variates. For log-transformed and normalized data (Fig. 1a, c vs. b, d), similar results were revealed. The first variate (CV1) apparently separated the two major lineages, i.e. aboriginal and commensal spe- cies groups, while the second one (CV2) identified individual species (or groupings of po- pulations) within the lineages. According to relative values of discriminant coefficients, 310 M. MACHOLAN the contrast of A and B (the so-called zygomatic index), and, to a lesser degree, also LM13i and LaZ were the variables contributing the most to the first discriminant function. In con- strast, in the second canonical variate, the relative contribution was not so clear, with the highest coeffcients being those for LB (contrasted by LCb), LMIs, LD and B. As displayed by the minimum spanning tree, the two M. spicilegus populations formed the most remote group within the aboriginal lineage in the two-dimensional discriminant space, whilst M. spretus was closer to M. macedonicus. When 'sıze-in’ and 'size-out’ ana- lyses were compared, the patterns were similar except for the changed relative position of the Ukrainian and Austrian spicilegus samples, and the Albanıan M. domesticus popula- tion which tended to be closer to M. musculus populations in both the ‘size-in’ analyses contrary to the ‘size-out’ ones. GR [a9) Oo MH ul SPR UKR AUT MS -1 / Fig. 2. A three-dimensional plot of the first three canonical variate scores for the 'size-out’ CVA on the normalized data. Group centroids are connected by the minimum spanning tree. A three-dimensional plot of the first three canonical variates is presented in figure 2 for the ‘size-out’ CVA on normalized data. The third canonical axis, based mostly on the relative rostral width (LaR as compared to LCb), placed the Albanian mice into the do- mesticus cluster and separated Bohemian and Moravian M. musculus populations from the Slovakian, Hungarian, and Ukrainian ones. Within the aboriginal group the species were not distantly separated by CV3 with, again, M. spretus being between the two east- ern short-tailed indoor species. There were no substantial differences between the 'size-in’ and ‘size-out’ and between logged and normalized data analyses on the third canonical axis. Multivariate morphometric analysis of European species of the genus Mus Sal The correctness of canonical discriminations were assessed by a posteriori classifica- tion tests. In table 2, actual memberships (rows) were tabulated against predicted ones (columns) from ‘size-out’ CVA on log-transformed data, where populations were pooled within species except the hybrid (MSP) and DA samples. In this analysis, 84.12% cases were classified correctly. When the MSP sample was excluded, the correctness increased to 91.51-92.25% (Cohen’s Kappa ranging between 0.884 and 0.895) depending on the type of the analysis used (see Tab. 3). Table 2. A posteriori classification testing the correctness of the assignment of each individual to a par- ticular group for the ‘size-in’ CVA performed on log-transformed data. Here, actual group membership (rows) is tabulated against predicted (columns). The populations are pooled within species except for DA and MSP samples. MAC SPI SPR DA DOM MUS MSP 7% 25 43 24 10 44 124 26 T 25 43 22 19 50 98 39 296 Table 3. A comparison of the correctness of canonical discriminations on different data sets; LOGIN, ‘size-in’ CVA on logged data; LOGOUT, ‘size-out’ CVA on logged data; NORMIN, 'size-in’ CVA on normalized data; NORMOUT, 'size-out’ CVA on normalized data. In columns, percentages of erro- neous assignment for each species (plus DA sample, MSP excluded), total classification error (in %), and Cohen’s Kappa are given, respectively. Cohen’s Kappa is an association measure testing if counts along the diagonal in Table 2 are significantly greater than those expected by chance alone; values great- er than 0.75 are usually said to indicate strong agreement (WirLKınson 1990). LOGIN LOGOUT NORMIN NORMOUT DA 0.00 0.00 0.00 0.00 DOM 11.36 19236 9.09 13.64 MUS 9.68 10.48 11.20 10.40 Total 7.18 7.18 773 8.49 Kappa The 'size-in’ and ‘size-out’ analyses were compared so that matrices were plotted against each other using the Mantel procedure (NTSYS-pc) separately for each data set (Fig. 3). In both cases, CVA revealed similar results when distances were close to the diag- onal (product-moment correlations r = 0.998 and r = 0.991, respectively). However, there were some differences between size-adjusted and non-adjusted data, especially in the nor- malized variables, mainly due to M. macedonicus which tended to show higher distances in the ‘size-in’ CVA compared to the ‘size-out’ analyses. Results of a UPGMA cluster analysis based on the Mahalanobis distances are shown in figure 4. Because of the hybrid nature of the MSP population this sample was excluded from the clustering. Interestingly, the ‘size-in’ and ‘size-out’ procedures gave the same trees, while there was a difference between log-transformed and normalized data sets: the Swiss M. domesticus (DCH) appeared in the M. musculus MC + MM cluster with the logged data (Fig. 4a), whereas both the species made separate clusters 312 M. MACHOLAN 1 0 T T — = & © - A N ©) e7 ar je) 6-n 2 or 3° 2 | E | 02 6) 4 N | 1 | fi 4 6 8 10 a size-out Fig. 3a 10 T T T L 4 A 2 8 8 A = ie A 2° I L 22 | > 7 Dos © 6 - = — ® 4 n | f | ı 4 6 8 10 b size-out Fig. 3b Fig. 3. Mantel plots of ‘size-in’ (ordinate) and ‘size-out’ (abscissa) Mahalanobis distances betweeen group centroids; (a) log-transformed data, r = 0.998; (b) normalized data, r = 0.991. In both the cases, M. macedonicus is marked by triangles, while all other populations are indicated by circles. Multivariate morphometric analysis of European species of the genus Mus 318 8.0 ZARO 6.0 5.0 4.0 D B [9 > = Fig. 4a I) SPR Fig. 4b Fig. 4. UPGMA dendrograms based on the Mahalanobis distances; (a) log-transformed data; (b) nor- malized data. Both the ‘size-in’ and ‘size-out’ analyses gave the same results for each type of the data transformation; MSP hybrid sample was excluded from the clustering. 314 M. MACHOLAN ({IMC + MM) + [[MS + MU] + MH]} and {DA + [DCH + DWM]}, respectively) with the normalized data (Fig. 4b). Albanian mice, although being quite distantly related, clus- tered with M. domesticus in both cases. What should be noted in the dendrograms is the small distance between M. macedonicus and M. spretus in comparison to the distance be- tween both the M. spicilegus populations and among populations within the domesticus and musculus clusters. The mound-building mouse (M. spicilegus) thus appeared to be morphologically the most distant species within aboriginal mice, yet interspecifice morpho- logıical distances within this group were, in general, strikingly small. Discussion Although size can provide significant information on morphometric differences among taxa, it is sometimes desirable to avoid size variation as it may cause a substantial bias in an assessment of group interrelationships due to the growth allometry (RöHrs 1961; THorPpE 1976, 1983 a). This especially concerns organisms with indeterminate growth; how- ever, nutritional, seasonal, sexual, ecological and other factors are also likely to affect morphological characters (LEAmY 1981) and thus the size-adjusting of data may be neces- sary. Several techniques have been employed to remove the size component from this ana- lysis. In morphometrics, the most familiar methods are those using ratios; however, it has been argued repeatedly that because of some undesirable statistical properties and con- ceptual difficulties ratios should be avoided (ATcHLEY et al. 1976; Corruccinı 1977; ATCH- LEY 1978; ATCHLEY and AnDERSoN 1978). Nor does taking logarithms of the ratios (BLACKITH and REYMENT 1971; Dopson 1978; Hırıs 1978) entirely remove size from data as stated by HumPpHRies et al. (1981) and evidenced by Reıst (1985). Another possibility is a univarlate regression analysis of variables on a standard size measurement such as snout-vent length in reptiles and amphibians, standard length in fish, wing length ın birds, or condylobasal length in mammal skulls (THorPE 1975; Corruccını 1977; KuHry and Marcus 1977). However, since size is designated a single variable in these techniques, only one particular variable is partialed out. As pointed out by HumPpHRies et al. (1981) and THoRrPE and LEAMmY (1983) size is not equal to any single measurement and using a multivariate ordination method for comparing size and shape differences among groups is more appropriate. Multiple group principal component analysis (MGPCA) is now widely used in var- ious types of studies (THoRPE 1983b; Corrı and THORPE 1989; ALLEGRUCCI et al. 1992; BEKELE et al. 1993) as a method of evaluating the relative contribution of size and shape to the between-group variation and to extract the size component from data. Nevertheless, some criticisms have appeared concerning the biological meaning of the general size aspect of the first principal component (SHEA 1985), and/or pointing out that the first component may contain shape information and remaining vectors may re- tain size information (reviewed in HUMPHRIES et al. 1981; Reıst 1985). While the first criticism is irrelevant to this study, the second may pose a problem, because even though the coefficients related to the first principal component appeared all to be of the same sign, their magnitude was very different even after transformation. This may indi- cate some residual size component could be resident in the second and following axes, whereas a proportion of shape information is likely to be lost with an extraction of the first component. Comparison of two transformation techniques employed for standardization of the measurements with different scales (log-transformation and normalization) proved to be of interest. While both approaches resulted in a slight improvement in between-variable linearity, logging the data (probably the most widely used method in morphological stu- Multivariate morphometric analysis of European species of the genus Mus SS dies) tended to change the relative contribution of variables to the variation explained by principal components. Moreover, the influence of non-equality of the variances of the ori- ginal characters, although being lowered relative to the raw data, was not entirely re- strained, and in turn the number of variables with insignificant character-component correlations was even increased by 50% in the log-transformed vs. original data. Normali- zation, on the other hand, resulted in better results both in the standardization of the var- iances and in the relative contribution of individual variables. The results of CVA computed on the MGPCA scores revealed similar results for both the normalized and log-transformed variates and only slight differences between the ‘size- in’ and ‘size-out’ analyses. In all the cases, the two separate evolutionary and ecological lineages were clearly discriminated. Within the aboriginal lineage, the most distant species were M. macedonicus and M. spicilegus with M. spretus being morphologically intermedi- ate between them. This result is rather surprising given the close genetic relationships be- tween the former two species (BONHOMME et al. 1983, 1984). Likewise, GERASIMoOV et al. (1990), using the stepwise discriminant analysis, found these forms to be morphologically very similar. On the contrary, in a previous study focused mostly on uni- and bivariate analyses (MAcHoLAN 1996) M. spicilegus appeared to be closer to M. spretus when origi- nal untransformed data were analysed, whereas the latter showed greater similarity with M. macedonicus when the variables were size-adjusted using THorPE’s (1975) allometric formula. A comparison of the univariate (MACcHoLAN 1996) and multivariate (this study) study carried out on the same material shows the former to be more affected by the growth/size influence than the latter. It is not clear, however, to what extent the close si- milarity between M. macedonicus and M. spretus reflects the cırcum-Mediterranean ecolo- gical vicariance of the two species (AUFFRAY et al. 1990). Within the commensal lineage, a somewhat peculiar position was displayed by the M. domesticus sample from Albania (considered as M.d. brevirostris, REICHSTEIN 1978; MARSHALL 1981, but see the discussion about valıdity of subspecific categories in M. domesticus ın FERRIS et al. 1983; SAGE et al. 1986; MAcHoLAN 1996) which was rather distinet both from other M. domesticus populations and from M. musculus, mainly due to its relatively narrow rostrum (MACHOLAN 1996). However, since 10 Albanian individuals were only studied, more animals should be investigated and perhaps other measurements should be included before the systematic relationships of Albanian and other commensal house mice can be established. In his multivariate morphometric analyses of house mice from eastern Europe and central Asia, LAVRENCHENKO (1990, 1994) found the variation within M. musculus to be categorical rather than clinal and this led him to distinguish three subspecies: musculus from the European part of the former USSR, southern Siberia and eastwards to the Far East; wagneri from lowlands north of the Caspian Sea, Kazakhstan, and ex-Soviet Central Asia; and raddei from eastern Kazakhstan, Altai, most of Mongolia, and eastern Transbai- kalıa. In this study, the investigation of M. musculus populations from western parts of its range showed quite different patterns and the variation changed rather continually in the east-west direction. This conclusion is corroborated by the results of the univariate analysis (MACHOLAn 1996) where some measurements were shown to be similar to M. domesticus in western localities, especially when the raw variates were taken. This suggests a possibility of introgressing domesticus alleles into the musculus range across the hybrid zone in western Bohemia and south-eastern Germany (SAGE et al. 1986; Tucker et al. 1992; MACHoLAn and ZımA 1994). The introgression of polygenic traits is similar to that of biochemical markers (MAcHoLAn and ZımA 1994 and unpubl. results) _ but the gene-flow distance might be much higher as indicated by the results of the pre- sent study. 316 M. MACHOLÄN Acknowledgements I wish to express my gratitude to the following persons for kindly providing material (listed in alphabeti- cal order): Dr. M. AnpErA (National Museum, Prague), Dr. J.-C. AurrrAy (University of Montpellier), Dr. K. BAUER (Museum of Natural History, Vienna), Mrs. S. BERANKOVA (Institute of Landscape Ecol- ogy, Brno), Prof. J. HAUSSER (University of Lausanne), Dr. S. V. MEZHZHERIN (Institute of Zoology, Kiev), Dr. F. SPITZENBERGER (Museum of Natural History, Vienna), Dr. V. VOHRALIK (Charles Univer- sity, Prague), Dr. I. V. ZAGORODNYUK (Institute of Zoology, Kiev). I am indebted to Dr. HEıpı HAUFFE for helpful comments and a language revision of the manu- script. This work was partly supported by the Ministry of Environment of the Czech Republic (GA/846/ 93), and by the Grant Agency of the Czech Republic (204/93/0531 and 206/95/1 596). Zusammenfassung Morphometrische multivariate Analyse europäischer Arten der Gattung Mus (Mammalia, Muridae) Die systematischen Beziehungen zwischen 12 Populationsgruppen von fünf europäischen Hausmausar- ten wurden mit Hilfe verschiedener Methoden zur morphometrischen Multivarianzanalyse untersucht. Die ‚Multiple group principal component analysis (MGPCA) wurde dazu genutzt, den Beitrag der Größe zur Gesamtvariation zu beurteilen. Es wurde gezeigt, daß ein Teil der ‚Form‘-Information in der ersten Hauptkomponente enthalten sein könnte, und weitere Komponenten ähnlich dazu eine residuale ‚Größen‘-Information beinhalten könnten. Deshalb sollte das Eliminieren der ‚Größen‘-Information mit Vorsicht und erst nach einer angemessenen Überprüfung der Daten vorgenommen werden. Die ka- nonische Diskriminanzanalyse brachte ähnliche Resultate wie die ‚size-in‘ und ‚size-out‘ MGPCA-Un- tersuchungen. Die erste Kanonische Achse grenzte die Freiland- und die kommensalen Artgruppen der Mäuse voneinander ab, währenddessen die zweite Achse Artgruppen identifizierte. Die dritte kano- nische Achse teilte Populationsgruppen innerhalb der kommensalen Arten ab. Sowohl die CVA als auch die Clusteranalyse zeigten, daß (1) M. macedonicus und M. spretus morphologisch miteinander mehr Ähnlichkeit haben als eine der beiden Arten mit M. spicilegus; daß (2) die Distanz zwischen M. musculus und M. domesticus ähnlich groß ist wie die Distanz zwischen den Freilandarten; und daß (3) die Distanz zwischen den Populationen verglichen mit den zwischenartlichen Beziehungen relativ hoch ist. Reference ALLEGRUCCI, G., BALDARI, F.; CESARONI, D.; THORPE, R. S.; SBORDONI, V. (1992): Morphometric analysis of interspecific and microgeographic variation of crayfish from a Mexican cave. Biol. J. Linn. Soc. 47, 455-468. ATCHLEY, W. R. (1978): Ratios, regression intercepts, and the scaling of data. Syst. Zool. 27, 78-83. ATCHLEY, W. R.; ANDERSON, D. (1978): Ratios and the statistical analysis of biological data. Syst. Zool. 27, 71-78. ATCHLEY, W. R.; GAskiIns, C. T.; ANDERSON, D. (1976): Statistical properties of ratios. I. Empirical re- sults. Syst. Zool. 25, 137-148. AUFFRAY, J.-C.; MARSHALL, J. T.,; THALER, L.; BONHOMME, F. (1990a): Focus on the nomenclature of European species of Mus. Mouse Genome 88, 7-8. AUFFRAY, J.-C.; TCHERNOV, E.; BONHOMME, F.; HETH, G.; SIMSoN, S.; NEvo, E. (1990 b): Presence and eco- logical distribution of Mus “spretoides” and Mus musculus domesticus in Israel. Circum-Mediterra- nean vicariance in the genus Mus. Z. Säugetierkunde 55, 1-10. BEKELE, A.; CAPANNA, E.; CoRTI, M.; MARCUS, L. F.; SCHLITTER, D. A. (1993): Systematics and geo- graphic variation of Fthiopian Arvicanthis (Rodentia, Muridae), J. Zool. (London) 230, 117-134. BLACKITH, R. E.; REYMENT, R. A. (1971): Multivariate morphometrics. London, New York: Acad. Press. BONHOMME, F.; CATALAN, J.; BRITTON-DAVIDIAN, J.; CHAPMAN, V. M.; MORIWAKRI, K.; NEVo, E.; THALER, L. (1984): Biochemical diversity and evolution in the genus Mus. Biochem. Genet. 22, 275-303. Multivariate morphometric analysis of European species of the genus Mus 317 BONHOMME, F.; CATALAN, J.; GERASIMOV, S.; ORSINI, P.; 'THALER, L. (1983): Le complexe d’especes du genre Mus en Europe Centrale et Orientale. Z. Säugetierkunde 48, 78-85. BONHOMME, F.; GUENET, J.-L. (1989): The wild house mouse and its relatives. In: Genetical Variants and Strains of the Laboratory Mouse. Ed. by M. F. Lyon and A. G. SEARLE. Oxford: Oxford Univ. Press. Pp. 649-662. BOußsoT, P; AUFFRAY, J.-C.; BRITTON-DAVIDIAN, J.;, BONHOMME, F. (1993): The evolution of house mice. Ann. Rev. Ecol. Syst. 24, 119-152. BURNABY, T. P. (1966): Growth-invariant discriminant functions and generalized distance. Biometrics 22, 96-110. CAMPBELL, N. A.; ATCHLEY, W.R. (1981): The geometry of canonical variate analysis. Syst. Zool. 30, 268-280. CoRBET, G. B. (1978): The mammals of the Palaearctic region: a taxonomic review. London, Ithaca: Cor- nell Univ. Press. CORRUCCINI, R. S. (1977): Correlation properties of morphometric ratios. Syst. Zool. 26, 211-214. CorTI, M.; THORPE, R. S. (1989): Morphological clines across a karyotypic zone of house mice in Central Italy. J. evol. Biol. 2, 253-264. DARVICHE, D.; Orsını, P. (1982): Criteres de differentiation morphologique et biometrique de deux especes de souris sympatriques: Mus spretus et Mus musculus domesticus. Mammalia 46, 205- TE Davis, S. J. M. (1983): Morphometric variation of populations of house mice Mus domesticus in Britain and Faroe. J. Zool. (London) 199, 521-534. Dopson, P. (1978): On the use of ratios in growth studies. Syst. Zool. 27, 62-67. ELLERMAN, J. R.; MORRISON-SCOTT, T. C. S. (1951): Checklist of Palaearctic and Indian mammals, 1758 to 1946. London: Trustees of the British Museum. EnGers, H. (1980): Zur Biometrie und Taxonomie von Hausmäusen (Genus Mus L.) aus dem Mittel- meergebiet. Z. Säugetierkunde 45, 366-375. EnGeLs, H. (1983): Zur Phylogenie und Ausbreitungsgeschichte mediterraner Hausmäuse (Genus Mus L.) mit Hilfe von „Compatibility Analysis“. Z. Säugetierkunde 48, 9-19. FERRIS, S. D.; SAGE, R. D.; PRAGER, E. M.; RıTTE, U.; Wırson, A. C. (1983): Mitochondrial DNA evolu- tion in mice. Genetics 105, 681-721. FISHER, R. A. (1936): The use of multiple measurements in taxonomic problems. Ann. Eugen. 7, 179- 188. GERASIMOV, S.; NIKOLOV, H.; MIHAILOVA, V.; AUFFRAY, J.-C., BONHOMME, F. (1990): Morphometric step- wise discriminant analysis of the five genetically determined European taxa of the genus Mus. Biol. J. Linn. Soc. 41, 47-64. Hırıs, M. (1978): On ratios - A response to Atchley, Gaskins, and Anderson. Syst. Zool. 27, 61-62. HUMPHRIES, J. M.; BooKSTEIN, F. L.; CHERNOFF, B.; SMITH, G. R.; ELDER, R. L.; Poss, S. G. (1981): Multi- variate discrimination by shape in relation to size. Syst. Zool. 30, 291-308. KELLER, A. (1974): Determination de l’äge de Mus musculus Linne par l’usure de la dentition. Rev. suisse Zool. 81, 839-844. KRAFT, R. (1985): Merkmale und Verbreitung der Hausmäuse Mus musculus musculus L., 1758, und Mus musculus domesticus Rutty, 1772 (Rodentia, Muridae) in Bayern. Säugetierkdl. Mitt. 32, 1- 12, KRATOcCHVIL, J. (1986 a): Die intraspezifische Evolution der Art Mus domesticus. Acta Sc. Nat. Brno 20, 1-49. KRATOcHVIL, J. (1986 b): Mus abbotti — Eine Kleinasiatisch-Balkanische Art (Muridae - Mammalia). Fo- lia Zool. 35, 3-20. KuHry, B.; Marcus, L. F. (1977): Bivariate linear models in biometry. Syst. Zool. 26, 201-209. LAURIE, E. M. ©. (1946): The reproduction of the house mouse (Mus musculus) living in different envi- ronments. Proc. R. Soc. Lond. B 133, 248-281. LAVRENCHENKO, L. A. (1990): Systematic analysis of the superspecies complex Mus musculus s. lat. PhD Thesis, IEMEA AS USSR, Moscow. [In Russian] LAVRENCHENKO, L. A. (1994): Analysis of craniometric characters of house mice Mus musculus sensu lato (Rodentia, Muridae): A multivariate approach. Zool. zh. 73, 169-178. [In Russian with Engl. summary] ' LeAamy, L. (1981): Heritability of morphometric ratios in random bred house mice. In: Mammalian Po- pulation Genetics. Ed. by M.H. SmitH and J. JouLe. Athens, Georgia: Univ. Georgia Press. Pp. 254- ZI 318 M. MACHOLAN LYALYUKHINA, S.; KOTENKOVA, E.; WALKOVA, W.; ADAMCZYK, K. (1991): Comparison of craniological para- meters in Mus musculus musculus Linnaeus, 1758 and Mus hortulanus Nordmann, 1840. Acta ther- iol. 36, 95-107. MACHOLAN, M. (1996): Morphometric analysis of European house mice. Acta theriol. 41 (in press). MACHOLAN, M.; ZIMA, J. (1994): Mus domesticus in western Bohemia: anew mammal for the Czech Re- public. Folia Zool. 43, 39-41. MARSHALL, J. T. (1981): Taxonomy. In: The Mouse in Biomedical Research, Vol. 1. Ed. by L. FosTEr, J. D. SMALL, and J. G. Fox. New York: Academic Press. Pp. 17-26. MARSHALL, J. T.; SAGE, R. D. (1981): Taxonomy of the house mouse. Symp. zool. Soc. Lond. 47, 15- 25% ORsint, P.; BONHOMME, F.; BRITTON-DAVIDIAN, J.; CROSET, H.; GERASIMOV, S.; THALER, L. (1983): Le com- plexe d’especes du genre Mus en Europe Centrale et Orientale. II. Criteres d’identification, r&parti- tion et caracteristiques Ecologiques. Z. Säugetierkunde 48, 86-95. PALOMO, L. J.; ESPANA, M.; LÖPEZ-FUSTER, M.-J.; GOSÄLBEZ, J.; SANS-CoMA, V. (1983): Sobre la variabili- dad fenetica y morfometrica de Mus spretus Lataste, 1883 en la Peninsula Iberica. Misc. Zool. 7, 171-192. PELIKAN, J. (1981): Patterns of reproduction in the house mouse. Symp. zool. Soc. Lond. 47, 205-229. REICHSTEIN, H. (1978): Mus musculus Linnaeus, 1758. -— Hausmaus. In: Handbuch der Säugetiere Euro- pas. Bd. 1. Ed. by J. NIETHAMMER and F. Krapp. Wiesbaden: Akad. Verlagsges. Pp. 421-451. Reiısr, J. D. (1985): An empirical evaluation of several univariate methods that adjust for size variation in morphometric data. Can. J. Zool. 63, 1429-1439. RoHLF, F. J. (1990): NTSYS-pc. Numerical Texonomy and Multivariate Analysis System, Version 1.60. Setauket, New York: Exeter Publ. ROHLE, F. J.; BOOKSTEIN, F.L. (1988): A comment on shearing as a method for “size correction”. Syst. Zool. 36, 356-367. Rönrs, M. (1961): Allometrie und Systematik. Z. Säugetierkunde 26, 130-137. SAGE, R. D. (1981): Wild mice. In: The Mouse in Biomedical Research, Vol. 1. Ed. by L. FoSTEkr, J. D. SMALL, and J. G. Fox. New York: Acad. Press. Pp. 39-90. SAGE, R. D.; WHITNEY, J. B. III.; Wırson, A. C. (1986): Genetic analysis of a hybrid zone between domes- ticus and musculus mice (Mus musculus complex): Hemoglobin polymorphisms. Curr. Top. Micro- biol. Immunol. 127, 75-85. SAGE, R. D.; ATCHLEY, W. R.; CAPANNA, E. (1993): House mice as models in systematic biology. Syst. Biol. 42, 523-561. SANS-CoMA, V.; LÖPEZ-FUSTER, M.-J.; GOSÄLBEZ, J. (1979): Über die Hausmaus, Mus musculus Linng, 1758, auf der Insel Meda Crossa, Katalonien, Spanien. Säugetierkdl. Mitt. 27, 107-113. SCHWARZ, E.; SCHWARZ, H. K. (1943): The wild and commensal stocks of the house mouse, Mus muscu- lus Linnaeus. J. Mammalogy 25, 59-72. SCRIVEN, P. N.; BAUCHAU, V. (1992): The effect of hybridization on mandible morphology in an island po- pulation of the house mouse. J. Zool. (London) 226, 573-583. SERAFINSKI, W. (1965): The subspecific differentiation of the central European house mouse (Mus mus- culus L.) in the light of their ecology and morphology. Ecol. Polska, Ser. A 13, 305-348. SHEA, B. T. (1985): Bivariate and multivariate growth allometry: statistical and biological considera- tions. J. Zool. (London) 206, 367-390. SOKAL, R. R.; ROHLF, F. J. (1981): Biometry. 2nd ed. New York: Freeman and Co. THALER, L.; BONHOMME, F.; BRITTON-DAVIDIAN, J. (1981): Processes of speciation and semi-speciation in the house mouse. Symp. Zzool. Soc. Lond. 47, 27-41. THORPE, R. S. (1975): Quantitative handling of characters useful in snake systematics with a particular reference to intraspecific variation in the Ringed snake Natrix natrix (L.). Biol. J. Linn. Soc. 7, 27- 43. THORPE, R. S. (1976): Biometric analysis of geographic variation and racial affinities. Biol. Rev. 51, 407- 452. THORPE, R. S. (1983 a): A biometric study of the effects of growth on the analysis of geographic varıa- tion: Tooth number in green geckos (Reptilia: Phelsuma). J. Zool. (London) 201, 13-26. | THORPE, R. S. (1983 b): A review of the numerical methods for recognising and analysing racial differen- tiation. In: Numerical Taxonomy. Ed. by J. FELSENSTEIN. Berlin, Heidelberg, New York, Tokyo: Springer-Verlag. Pp. 404-423. THORPE, R. S.; CorTI, M.; CAPANNA, E. (1982): Morphometric divergence of Robertsonian populations/ species of Mus: a multivariate analysis of size and shape. Experientia 38, 920-923. Multivariate morphometric analysis of European species of the genus Mus 319 THORBE, R. S.; LEAMY, L. (1983): Morphometric studies in inbred and hybrid House mice (Mus sp.): Mul- tivariate analysis of size and shape. J. Zool. (London) 199, 421-432. TUCKER, P. K.; SAGE, R. D.; WARNER, J. H.; WıLson, A. C.; EICHER, E.M. (1992): Abrupt cline for sex chromosomes in a hybrid zone between two species of mice. Evolution 46, 1146-1163. WILKINSOoN, L. (1990): SYSTAT: The System for Statistics, Release 5.02. Evanson, IL: SYSTAT, Inc. Author’s address: Dr. MıLoS MACHOLAN, Laboratory of Genetics and Embryology, Institute of Ani- mal Physiology and Genetics, Academy of Sciences of the Czech Republic, Ve- ver 97, CZ-60200 Brno, Czech Republic. = NE Z. Säugetierkunde 61 (1996) 320 ZEITSCHRIFT& > "FÜR © 1996 Gustav Fischer, Jena SÄUGETI ERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Buchbesprechung STARCK, D.: Lehrbuch der Speziellen Zoologie. Bd. II: Wirbeltiere, Teil 5/1-2: Säugetiere. Jena: Gustav Fischer Verlag 1995. 1241 S. 504 Abb., 62 Tab., DM 198,-. ISBN 3-334-60453-5. Wenn man als älterer Säugetiertaxonom ein neues Handbuch über Säugetiere in die Hand bekommt, beginnt man das Werk meist mit einer gewissen Skepsis zu lesen. Man hat schon soviele Bücher dieser Art gelesen, daß man den Eindruck hat, sie alle seien Variationen eines bekannten Themas. Man ist blasiert geworden. In seltenen Fällen erlebt man beim Lesen des neuen Werkes, daß das Thema neu und aufregend ist, und daß man sich bezüglich seines eigenen Spezialgebietes wieder jung und enthu- siastisch fühlt, daß man wiederholt das Buch aufschlägt, um zu lesen, was der Autor über einen bes- timmten Sachverhalt geschrieben hat oder was über eine bestimmte Säugetierordnung berichtet wird. Dem Rezensenten wiederfuhr diese zweite Reaktion nach dem Studium der beiden Teilbände über Säugetiere von D. STARCK. Der Inhalt der Bände ist ein erfrischendes Gemisch aus deutscher Gründlich- keit und einer amerikanischen, sachlichen Annäherung, die durch hervorragende Illustrationen und eine umfassende Literaturliste ergänzt werden. Und weil alle Kapital der beiden Teilbände vom glei- chen Autor stammen, sind sie ausgewogen - nicht zu kurz und nicht zu detailliert. Kurzum: so, wie ein gutes Handbuch sein sollte. Nach einem knappen Vorwort, dem Inhaltsverzeichnis und einer kurzen Einleitung folgen Kapitel über Homoiothermie, Wärmehaushalt, Temperaturregulation, Lethargie, Winterschlaf und Karyologie. Den zentralen Teil des Werkes bildet das Kapitel über Systematik, Phylogenese und Verbreitung, in dem alle Säugetierordnungen und -familien und ihre typischen Eigenschaften eingehend besprochen werden. Hierbei sind auch Tiere ausgestorbener Ordnungen mit berücksichtigt. Die Bände schließen mit ausführlichen Literaturlisten, einem Verzeichnis der wissenschaftlichen und trivialen Tiernamen und einem Sachregister. Diese Teilbände sind nicht billig, aber in Anbetracht des Inhaltes, der Ausführung und des zukünfti- gen Gebrauchswertes sicher eine ausgezeichnete Geldanlage. Seit Jahrzehnten ist das Buch „Die Säuge- tiere“ von MAx WEBER das Handbuch schlechthin gewesen, aber jetzt gibt es einen würdigen, modernen Ersatz in Gestalt des hier vorliegenden Werkes. P. J. H. van BREE, Amsterdam Redaktionelle Hinweise Einsendung und Annahme von Manuskripten: Ms. zur Veröffentlichung sind einzusenden an den 1. Schriftleiter: Herrn Prof. Dr. Dieter Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel. Über die Annahme von Ms. zur Veröffentlichung wird gemäß Geschäftsordnung der Deutschen Gesellschaft für Säugetierkunde entschieden. Der Eingang von Ms. wird sofort bestätigt, und nach Erhalt der Gutachten wer- den die Autoren über die Entscheidung umgehend informiert. Schriftwechsel, der die technische Drucklegung betrifft, ist zu richten an: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena. Äußere Form der Manuskripte: Die Ms. sind 3fach und vollständig als Original mit 2 Kopien in deutscher oder englischer Sprache einzureichen. Sie müssen mit Schreibmaschine oder als deutlich lesbare Computerausdrucke auf DIN A4-Blättern (21x 29,7 cm), mit einem linken Heftrand von 4cm und durchgehend mit einem doppelten Zeilenabstand geschrieben sein. Jedes Ms. soll auf der 1. Seite folgende Angaben enthalten: Titel der Studie; Name(n) der Autoren (männliche Vornamen nur in Initialen, weibliche ausgeschrieben); Institution(en), an der (denen) die Studie durchgeführt wurde. Ferner soll bei längeren Titeln ein kurzer Kolumnentitel (Seitenüber- schrift) von maximal 72 Anschlägen angegeben werden. Die Schrifttypen müssen einheitlich sein, ohne halbfette, gesperrte, kursive etc. Hervorhebungen. Fußnoten auf den Seiten oder Appendices am Ende sind nicht zugelassen. Autorennamen im Zusammenhang mit Literatur- zitaten sind im Text mit Bleistift zu unterstreichen. Wissenschaftliche zoologische und botanische Gattungs-, Art- und Unterartnamen sind mit einer Schlangenlinie zu unterstreichen. Der Name von Erstbeschreibern bei wis- senschaftlichen Namen wird nicht unterstrichen. Im Ms. ist am Rand mit Bleistift zu vermerken, an welcher Stelle die einzelnen Abbildungen und Tabellen ein- gefügt werden sollen. Am Ende der Studie, im Anschluß an das Literaturverzeichnis, sind die genauen postalischen Adressen des (der) Autoren anzugeben. Umfang und Aufbau der Manuskripte: Manuskripte können als Originalarbeiten oder als wissenschaftliche Kurz- mitteilungen veröffentlicht werden. a) Originalarbeiten: Originalarbeiten sollen einschließlich Abbildungen, Tabellen und Literaturverzeichnis einen Gesamtumfang von 30 Ms.-Seiten nicht überschreiten. Der Text soll eine klare Gliederung aufweisen in: Abstract (in englischer Sprache), Einleitung, Material und Methode, Ergebnisse, Diskussion (oder zusammengefaßt als Er- gebnisse und Diskussion), Danksagungen, Zusammenfassung und Literatur. Bei deutschsprachigen Studien ist über dem englischen Abstract der Titel in englischer Sprache anzugeben. Bei englischsprachigen Studien ist entsprechend über der deutschen Zusammenfassung der Titel in deutscher Sprache anzugeben. b) Wissenschaftliche Kurzmitteilungen: Wissenschaftliche Kurzmitteilungen sollen einen Gesamtumfang von 5 Ms.-Seiten nicht überschreiten. Sie enthalten kein Abstract und keine Zusammenfassung. Bei deutschsprachigen Ms. ist jedoch auf der 1. Seite zusätzlich eine englische Übersetzung des Titels anzugeben, bei englischsprachigen entsprechend eine deutsche Übersetzung. Wissenschaftliche Kurzmitteilungen sind äußerlich nicht in die Kapitel Einleitung, Material und Methode, Ergebnisse und Diskussion gegliedert, müssen aber in der textlichen Darstel- lung diese Strukturierung erkennen lassen. Abbildungen: Anzahl und Größe der Abbildungen sind unbedingt auf das notwendige Mindestmaß zu beschrän- ken. Es können nur kontrastreiche, reproduktionsfähige Originale verwendet werden, farbige Abbildungen wer- den nicht angenommen. Beschriftungen müssen einheitlich und wegen späterer Verkleinerung groß und kräftig sein. Auf der Rückseite sind laufende Numerierung, Zeitschrift und Autorennamen anzugeben. Alle Bildunter- schriften sollen auf einem gesonderten Blatt gelistet werden. Tabellen: Tabellen müssen auf ein Mindestmaß beschränkt werden. Jede Tabelle muß auf einem gesonderten DIN A4-Blatt angelegt sowie mit Überschrift und laufender Numerierung versehen sein. Umfangreichere Tabel- len müssen auf Folgeblättern als Fortsetzung erscheinen. Literaturverzeichnis: Jede zitierte Arbeit ist in alphabetischer Reihenfolge aufzuführen. Es soll nur international erreichbare Literatur zitiert werden. Die Zeitschriften werden mit ihrem Kurztitel angegeben. Beispiele: a) Zeitschriftenbeiträge: KALko, E. K. V; HAnDLEY, C. O. Jr. (1994): Evolution, biogeography, and description af a new species of Fruit-eat- ing bat, genus Artibeus Leach (1821), from Panama. Z. Säugetierkunde 59, 257-273. b) Bücher: HERRE, W.; RÖöHrs, M. (1990): Haustiere — zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. c) Handbuchbeiträge: NIETHAMMER, J. (1982): Cricetus cricetus (Linnaeus, 1758) -— Hamster (Feldhamster). In: Handbuch der Säugetiere Europas. Ed. by J. NIETHAMMER und F. Krapp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28. Drucklegung: Der jeweilige Hauptautor erhält sowohl die Andrucke der Bilder zur Prüfung wie auch den komplet- ten Umbruch seines Artikels zusammen mit dem Ms. Es dürfen nur tatsächliche Satzfehler korrigiert werden. Vom Ms. abweichende Änderungen werden dem Autor in Rechnung gestellt. Es sind die üblichen Korrekturzeichen zu verwenden. Ein als druckfertig bezeichneter Abzug ist zu senden an: Herrn Prof. Dr. Peter Langer, Institut für Anatomie und Zytobiologie, Justus-Liebig-Universität, Aulweg 123, D-35385 Giessen. Sonderdrucke: Verfasser von Originalbeiträgen und Wissenschaftlichen Kurzmitteilungen erhalten 50 Sonder- drucke gratis. Mehrbedarf kann gegen Berechnung bei Rücksendung des Umbruchs bestellt werden. Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective, or for resale. For copying from back volumes of this journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 1005 37, D-07705 Jena; Telefon (03641) 62 6436, Telefax (03641) 62 65 00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen | Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A. Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zeitschriftenvertrieb, Frau B. Dressler, Post- fach 1005 37, D-07705 Jena; Telefon (03641) 62 64.44, Telefax (0 3641) 62 65 00. For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 4288201. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2945,-/382,50 (plus postage). Single issue price (DM/ÖSI/SFr): 80,-/ 592,-/ 77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 86010090; Deutsche Bank, Konto- Nr. 3907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400 00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 zZ L) SÄAUGETIER Apollonio, M.; Vailati, G.: Functional morphology of metatarsal glands in Fallow Deer (Cervus dama). — Funktions- morphologie der Metatarsaldrüsen beim Damhıirsch (Cervus dama) Groves, C. P: The taxonomy of the Asıan Wild Buffalo from the Asıan mainland Taxonomie des Wasserbüffels (Bubalus arnee) auf dem asiatischen Festland Corti, M.; Civitelli, Maria Vittoria; Castiglia, R.; Bekele, Afework; Capanna, E.: Cytogenetics of the genus Arvi- canthis (Rodentia, Muridae). 2. The chromosomes of three species from Ethiopia: A. abyssinicus, A. dembeensis and A. blicki. - Cytogenetik der Gattung Arvicanthis (Rodentia, Muridae). 2. Die Chromosomen von drei Arten aus Äthiopien: A. abyssinicus, A. dembeensis und A. blicki Rosi, Maria I.; Cona, Mönica I., Puig, Silvia; Videla, F.; Roig, V. G.: Size and structure of burrow systems of the fos- sorial rodent Ctenomys mendocinus in the piedmont of Mendoza province, Argentina. -— Maße und Struktur deı unterirdischen Gänge von Nagetieren der Art Ctenomys mendocinus auf dem Pıedmont ın der Provinz Mendoza, Argentinien........ "Torre, I.; Tella, J. L.; Arrizabalaga, A.: Environmental and geographic factors affecting the distribution of small mam mals in an isolated Mediterranean mountain. — Verbreitungsbedingte Umweltfaktoren für Kleinsäuger ın einem isolierten mediterranen Gebirge......... ey Wissenschaftliche Kurzmitteilungen Topping, M.; Kruuk, H.: Size selection of prey by otters, Lutra lutra L.: An experimental approach. — Größenselek- tion der Beute von Fischottern ZLurra lutra L.: Ein experimenteller Ansatz RR sea: x ae, Hirakawa, H.: Hard faeces reingestion in the Mountain hare Lepus timidus. -— Wiederaufnahme von hartem Kot beim Schneehasen Lepus timidus .... Sr EIER a Mitteilungen der Gesellschaft ......... a re er EA CEO er: er ee he Indexed in Current Contents “ Agriculture, Biology & Environmental Sciences; Biological Abstracts; BIOSIS database ws ITSCHRIFTT FÜR KUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY DI 339 SIE rt FISCHER Vol. 61 e 321-384 JENA »-STUTTGART»NEW YORK Dezember 1996 ONIS ARTIBUS ZEITSCHRIFT FÜR ‘=. SÄUGETIERKUNDE INTERNATIONAL JOURNAL A OF MAMMALIAN BIOLOGY Kr L Herausgeber/Editor Deutsche Gesellschaft für Säugetierkunde Schriftleitung/Editorial Office D. Kruska, Kiel - P. Langer, Giessen Wissenschaftlicher Beirat/Advisory Board P. J. H. van Bree, Amsterdam - W. Fiedler, Wien - G. B. Hartl, Kiel - R. Hutterer, Bonn - H.-G. Klös, Berlin - E. Kulzer, Tübingen - P. Lüps, Bern - W. Maier, Tübingen - O. Anne E. Rasa, Bonn -H. Reichstein, Kiel — M. Röhrs, Hannover - H. Schliemann, Hamburg - G. Storch, Frankfurt — P. Vogel, Lausanne Deutsche Gesellschaft für Säugetierkunde Altvorsitzende/Living Past Presidents D. Starck, Frankfurt (1957-1961, 1967-1971) - W. Herre, Kiel (1962-1966) — H. Frick, München (1972-1976) -— M. Röhrs, Hannover (1977-1981) - H.-J. Kuhn, Göttingen (1982-1986) - E. Kulzer, Tübingen (1987-1991) Amtierender Vorstand/Managing Committee Vorsitzender/President: U. Schmidt, Bonn Mitglieder/Board Members: H. G. Erkert, Tübingen - W. Fiedler, Wien — H. Frädrich, Berlin - R. Hutterer, Bonn - D. Kruska, Kiel -— Marialuise Kühnrich, Hamburg Z. Säugetierkunde 61 (1996) 321-326 © 1996 Gustav Fischer, Jena INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY (Cervus dama) By M. ArorLLonIo and G. VAILATI Animal Behaviour and Anthropology Department, University of Pisa and Biology Department, University of Milan, Italy Receipt of Ms. 10. 11. 1995 Acceptance of Ms. 30. 07 1996 Abstract Samples of metatarsal glands were studied from fallow deer (Cervus dama), of different age and sex classes collected in September, October, November, and March. The amount of sebaceous and sudori- ferous gland layers, the diameter of the apocrine tubuli, and their relative number were measured. In each sex and age class two different developmental situations were found, the sudoriferous component being mainly responsible for both. Such differences corresponded to a more or less pronounced glandu- lar activity, evidenced by the thickness of the sudoriferous glandular dermal zone (sud-GDZ) and the density of the apocrine tubuli. Well-developed glands were significantly more prevalent in adults than in subadults and fawns. No seasonal variation was found in the rut and the post-rut period for adult fe- males, while it was evident in adult males. In the latter, glands are underdeveloped during the rut, sug- gesting no relationship with the sexual hormone cycle, but become well-developed in spring, i.e. when males regroup after the rut. The obtained results suggest that the metatarsal gland in fallow deer could play a similar functional role as in black-tailed deer (i.e. to discharge alarm pheromons, being more de- veloped in the sex and age classes involved in social groups with a leadership role). Introduction The metatarsal gland is one of the specialized dermal glands involved in chemical commu- nication, and consists of enlarged holocrine and apocrine secretory units. Seasonal or sex and age variations in such structures seem to be related with behavioural changes and could provide information concerning their function; all these data are thus of interest from an ethological point of view. In cervids they are present both in Cervinae and Odo- coileinae (GosLinG 1985) and have been especially well studied in the latter (HERSHKO- vırz 1958; QuAy 1959; QuAay and MÜLLER-SCHWARZE 1970; MÜLLER-SCHWARZE 1971). In black-tailed deer (Odocoileus hemionus columbianus), MÜLLER-SCHWARZE 1971 demon- strated their function ın releasing scent in “fear-inducing” situations. For fallow deer Po- cock (1910) and ALvarEZ et al. (1975), gave only an external description, while the histology of this gland in Cervus dama was described by ZIETZSCHMANN (1903), BRINKER- MANN (1912) and SCHAFFER (1940) although no functional explanation was given. Here we describe the histology of the metatarsal gland in fallow deer and we attempt to use histological data concerning different sex and age classes and different periods in order to assess: i) any differences in sex and age classes possibly linked with a functional explanation ii) any seasonal fluctuations of metatarsal gland activity likely to be dependent on sex- ual hormone cycle. 322 M. AroLLonIo and G. VAILATI Materials and methods Metatarsal glands were collected from 198 fallow deer culled from September to December during the years 1985, 1986, 1987, and 1992 in the San Rossore Preserve near Pisa, Italy. Samples were fixed in Bouin’s fluid. Skin samples were dissected into easy to handle pieces by removing a strip from the center of the metatarsal gland. Sections 10 um in thickness were stained with Mayer’s acıd alum hematoxylin and : eosin. The thickness of the glandular dermal zone (GDZ, subdivided into sebaceous glandular dermal , zone, seb-GDZ, and sudoriferous glandular dermal zone, sud-GDZ), the outer diameter of the apocrine tubuli and their density were measured with the aid of an ocular micrometer. The thickness of GDZ refers to the dermal portion actually occupied by glandular layers, discounting strips showing only connective tis- sue. It was measured where glandular tissue was most developed. The density of sudoriferous tubuli was determined referring to a reticulum of 1.562 square millimeter placed at the middle of the measurement Y axis. Their mean outer diameter was obtained from the values of 10 tubuli randomly chosen inside the reti- culum (cf. MossınGg and KÄrrouiıst 1981). Variations in the amount of secretory tissue have been used to indicate the extent of secretory activity; therefore, we distinguished between developed and non-devel- oped glands considering both sudoriferous and sebaceous components and we tested the measurements described above for all sex and age differences. The age classes, determined by dentition and antlers di- mensions and shape, were as follows: male and female fawns between 4 and 10 months old; male and fe- male yearlings between 15 and 21 months; adult males and females older than 27 months. Statistical analyses included Kruskall-Wallis Anova, Mann-Whitney U-test, y’ and Fisher exact test. Results and discussion The metatarsal gland in fallow deer shows two glandular components: one displaying superficial aciniform sebaceous secretory units and a second with deep tubular sudorifer- ous bodies (Fig. 1). The overlying skin is intensely pigmented and a circumglandular hair Fig. 1. Developed (left) and non-developed (right) metatarsal glands of adult fallow deer females. Scale bar is 0.5 mm, * = hair follicle, D = dermis, E = epidermis, SE = sebaceous glands, SU = sudoriferous glands. Functional morphology of metatarsal glands in Fallow Deer (Cervus dama) 323 tuft is present. In the centre of this area the keratinized ridge present in Odocoileus (QuAy 1959; Quay and MÜLLER-SCHWARZE 1970) is absent. Overall thickness is 3-4 mm, depending on the development of glandular components. The epidermis covering the glandular area is about 50 um thick and consists of a pluristratified keratinized epithelium with 4-6 cellular rims and a thin stratum corneum. The underlying corium is about 0.5 mm thin. The sebaceous component consists of irregularly shaped bulbs about 200 um wide, usually connected with smooth musculature (erector pili) of about 150 um. The ex- cretory duct of these bulbs opens into the higher tract of the pilosebaceous follicle near the open surface. The sudoriferous component shows many glomerular apocrine glands whose tubuli have an outer diameter of 70-100 um and are surrounded by a monostrati- fied cubic epithelium of 7-15 um. Their excretory ducts are very thin and open inside a pi- liferous follicle above the level at which the sebaceous glands open. Unlike in Odocoileus (QuAy 1959; QuAay and MÜLLER-SCHWARZE 1970), sudoriferous units do not form a com- pact glandular layer, but are separated by slender connective tissue. Close to the sudori- ferous tubuli myoepithelial cells are numerous and observable. Differences in metatarsal gland development concern sebaceous and sudoriferous components; higher values for thickness being 1.3 mm in sebaceous glands and 1.25 in sudoriferous glands, and the lower values about 185-187 um in both of them. Age and sex classes showed differences in the thickness of the sudoriferous component (p < 0.00001, K-W Anova) but not in the thick- ness of the sebaceous component. Histological description of this gland is comparable to the findings Of ZIETZSCHMANN (1903) and SCHAFFER (1940). However, we observed developmental differences in age and sex classes in the sudoriferous but not in the sebaceous component; this does not corre- late with the results of SCHAFFER (1940). Analysing seven individuals only, this author found such differences in the sebaceous but not in the sudoriferous component. The su- doriferous component showed remarkable differences in age and sex classes in spite to what was observed by QuAy and MÜLLER-SCHWARZE (1970) ın black-tailed deer. It should be taken into account that these authors considered six males and one female only. To look for differences among age and sex classes in the proportion of developed and non-developed glands, we qualitatively discriminated the two categories of development in relation to the thickness of seb-GDZ and sud-GDZ. This evaluation was tested with Mann-Whitney U-test for each sex and age class for the sudoriferous component by the use of the sud-GDZ thickness, tubuli density (TD) and tubuli diameter (TDI) in a sample of ten (or fewer when not available) developed and non-developed glands. Sud-GDZ and TD proved to fit well with the evaluation, as significant differences were observed bet- ween developed and non-developed glands in all cases but one (subadult females, sud- GDZ, P = 0.07). Some comparisons were then performed with x” test. Differences were found between age classes, both with cumulated and segregated sex classes, but not be- Table 1. Mean thickness value #SE expressed in um of sudoriferous and sebaceous glandular dermal zones (sud-GDZ and seb-GDZ,) of metatarsal glands in different age and sex classess in fallow deer. Sex and age class No. of samples Mean thickness of Mean thickness of sudoriferous glandular sebaceous glandular dermal zone dermal zone Female fawns 412 #145 SONS Yearling females 496 + 145 826 + 263 Adult females 660 + 245 874 + 244 Male fawns 431 #110 766 + 230 Yearling males 584 + 209 763 + 216 Adult males 756 + 263 814 #185 324 M. AroLLonIo and G. VAILATI Table 2. Comparisons made with y2 of proportion of developed and non-developed metatarsal glands with respect to the sudoriferous component in different age and sex classes. From above a) comparison between age classes, b) comparison between age classes in males and c) females, d) comparison be- tween sexes within the same age class. Yearlings Adults x = 13.29 p < 0.001 Yearlings Fawns Adults Yearlings Adults _ x=026 p>05 N.s. Yearlings = Fawns Adults Yearlings Adults — x =9.13 p < 0.005 Yearlings — Fawns Adult females Yearling females Female fawns Adult males X=047 p>025 N.S. Yearling males Y=154 p>01 N.S. Male fawns x = 0.0008 p>0.975 N.S. tween sex classes of the same age (Tab. 2). Differences between developed and non-devel- oped glands within each category for the sebaceous component was based only on the thickness of the seb-GDZ. This evaluation was tested as described above with a compari- son of ten developed and non-developed glands for each sex and age class, and proved to fit well in all cases. However, no differences were found in the proportion of developed and non-developed glands with respect to the sebaceous component in the various age or sex classes. Functional morphology of metatarsal glands in Fallow Deer (Cervus dama) 325 An interpretation of the functional role of metatarsal glands lies in the different devel- opment of the sudoriferous and the sebaceous components. Only the former showed de- velopmental differences between age classes and this component in some cervids could develop volatile pheromones, favouring rapid olfactory communication. The metatarsal gland of black-tailed deer, for example, produces a scent that acts as an alarm phero- mone. This scent is attributable to the sudoriferous secretion, and has a low concentration and is highly volatile (MÜLLER-SCHWARZE 1971). Considering the close relationship of the two species, it would not be surprising ıf metatarsal glands in fallow deer serve a similar function. The higher developmental level reached by the sudoriferous component in adult fallow deer could depend on the higher social responsibility of this age class: adult fe- males, particularly, maintain a permanent link with their fawns (BrAzAa 1975; SCHAAL 1982) and often play a leader role in herds (HEIDEMAnN 1973; BRAZA 1975). A further ele- ment to be considered is the higher vigilance activity of adult females in comparison to other age and sex classes reported by SCHAAL and RoPpPrArRTZ (1985). Seasonal changes between rut and post-rut periods were documented by statistical analysis within the various age and sex classes only for adult males (Fisher exact test p = 0.013) who showed higher proportions of developed metatarsal glands in the post- rut period. This result suggests a possible connection with the dissolution of male herds dur- ing the breeding season due to the arousal of interindividual aggression. Since, at this time, males do not maintain other social relationships than intense competition for fe- males, the scent alarm communication would be quite useless. In our opinion this seaso- nal variation therefore should not be interpreted as depending on sexual hormones as in the cases of preputial glands of adult fallow deer males (KENNAUGH et al. 1977) or subor- bital glands in reindeer (MossınG and KÄrrouıst 1981). Metatarsal gland activity in adult fallow deer males seems in fact to be depressed by breeding activities and the consequent increase of testosterone. Acknowledgements We are most grateful to the staff of the San Rossore Presidential Estate for their help in collecting sam- ples. We gratefully acknowledge the help of ROBERTO CoLoMBO who greatly improved this manuscript both in concepts and style. Zusammenfassung Funktionsmorphologie der Metatarsaldrüsen beim Damhirsch (Cervus dama) Es wurden Metatarsaldrüsen von Damhirschen (Cervus dama) unterschiedlichen Alters und Ge- schlechtes untersucht. Die erforderlichen Proben wurden in den Monaten September, Oktober, Novem- ber und März gesammelt. Die Ausbildung des sebazeischen und sudoriferen Stratum sowie der Durchmesser und die relative Anzahl der apokrinen Tubuli wurden gemessen. In jeder Alters- und Ge- schlechtsklasse wurden zwei unterschiedliche Entwicklungszustände gefunden. Die Unterschiede entsprachen einer mehr oder weniger ausgeprägten Drüsenaktivität, wie sie durch die Dicke des sudori- feren Stratum und die Dichte der apokrinen Tubuli nachgewiesen wurde. Gut entwickelte Drüsen tra- ten signifikant häufiger bei Adulten als bei Subadulten und Kälbern auf. Bei den weiblichen Tieren zeigten sich zwischen der Brunftperiode und der Zeit danach keine Unterschiede in der Drüsenreife. Bei den Hirschen waren die Drüsen während der Brunft unterentwickelt, was auf einen fehlenden Zu- sammenhang mit dem Zyklus der Geschlechtshormone hindeutet. Beim Wiedereintreten der Hirsche in Männchengruppen nach der Brunft waren die Drüsen jedoch gut entwickelt. Nach diesen Ergebnissen scheinen die Metatarsaldrüsen beim Damhirsch dieselbe Funktion zu haben wie beim Schwarzwedel- hirsch, nämlich die Abgabe von Warnpheromonen. Diese Funktion ist in jenen Alters- und Ges- chlechtsklassen stärker ausgeprägt, in denen soziale Gruppen und die Führerrolle von Einzeltieren von Bedeutung sind. 326 M. ApoLLonIo and G. VAILATI References ALVAREZ, F.; BRAZA, F.; NORZAGARAY, A. (1975): Estructura social del gamo (Dama dama, Mammalia, Cervidae) en Donana. Ardeola 21 (Especial), 1129-1142. BRAZA, F. (1975): Censo de gamo (Dama dama) en Donana. Naturalia Hispanica, 3 ICONA, 1-27. BRINKERMANN, A. (1912): Die Hautdrüsen der Säugetiere (Bau und Sekretionsverhältnisse). Erg. Anat. Entw. 20, 1173-1231. GosLing, L. M. (1985): The even toed ungulates: order Artiodactyla. Sources, behavioural context, and function of chemical signals. In: Social odours in mammals, Ed. by R.E. Brown and D. W. MACDONALD. Oxford: University Press. Pp. 550-618. HEIDEMANN, G. (1973): Zur Biologie des Damwildes (Cervus dama L., 1758). Hamburg, Berlin: Verlag Paul Parey. HERSHKOVITZ, P. (1958): The metatarsal glands in white-tailed deer and related forms of the Neotropical Region. Mammalıa 22, 537-546. KENNAUGH, J. H.; CHAPMAN, D. I.; CHAPMAN, N. G. (1977): Seasonal changes in the prepuce of adult fal- low deer (Dama dama) and its probable function as a scent organ. J. Zool. (London) 183, 301-310. Mossıng, T.; KÄLLouist, L. (1981): Variations in cutaneous glandular structures in reindeer (Rangifer tarandus). J. Mammalogy 62, 606-612. MÜLLER-SCHWARZE, D. (1971): Pheromones in black-tailed deer (Odocoileus hemionus columbianus). Anim. Behav. 19, 141-152. Pocock, R. I. (1910): On the specialized cutaneous glands of ruminants. Proc. Zool. Soc. London. Pp. 840-986. Qua, W. B. (1959): Microscopic structure and variation in the cutaneous glands of the deer, Odocoileus virginianus. J. Mammalogy 40, 114-128. Quay, W. B.; MÜLLER-SCHWARZE, D. (1970): Functional histology of integumentary glandular regions in black-tailed deer (Odocoileus emionus columbianus). J. Mammalogy 51, 675-694. SCHAAL, A. (1982): Influence de l’environment sur le composantes du groupe social chez le daim cervus (Dama dama L.) Rev. Ecol. Terre Vie 36, 161-174. SCHAAL, A.; ROPPARTZ, P. (1985): Le comportement de surveillance chez le Daim (Dama dama): effet de variables liees a l’individu, au groupe social et a l’habitat. C.R. Acad. Sc. Paris, t. 301, Serie III, 16, 731-736. SCHAFFER, J. (1940): Die Hautdrüsenorgane der Säugetiere mit besonderer Berücksichtigung ihres histo- logischen Aufbaues und Bemerkungen über die Proktodäaldrüsen. Berlin, Wien: Urban und Schwarzenberg. ZIETZSCHMANN, E. H. (1903): Beiträge zur Morphologie und Histologie einiger Hautorgane der Cervi- den. Z. wiss. Zool. 74, 1-63. Authors’ addresses: MARCO APoLLoNIo, Dipartimento di Scienze del Comportamento Animale e dell’ Uomo, Universita di Pisa, Via Volta 6, I-56126 Pisa and GIAnnI VAILATI, Diparti- mento di Biologia, Universita di Milano, Via Celoria 26, I-20133 Milano, Italy. Z. Säupetierkunde 61 (1996) 327-338 ZEITSCHRI FT® ur S19mGuvficheniem _ _ _ ________ SÄUGETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY The taxonomy of the Asian Wild Buffalo from the Asian mainland By C. P. GrovES Department of Archaeology and Anthropology, Australian National University, Canberra, Australia Receipt of Ms. 01. 11.1995 Acceptance of Ms. 23. 01. 1996 Abstract Examined were skulls and horns of wild buffalo (Bubalus arnee) from India and Southeast Asia. Three subspecies are clearly distinguishable: one from Assam, one from Nepal and Central India, and one from Thailand and Cambodia. Several names are available; their type specimens had been lost or mis- identified, but were discovered in this study, and allocated to one or other of the two Indo-Nepalese taxa. A new subspecies is described for the Thailand and Cambodian population. Introduction Wild buffalo (Bubalus) in Asia are known from India and Nepal, Sri Lanka, Burma, Thai- land, Cambodia, Vietnam, Malaysıa (Sarawak), Indonesia (Sumatra, Java, Sulawesi), and the Philippines (Mindoro). Those from Sulawesi are assigned a separate subgenus, Anoa, and those from Mindoro, while referable to the nominotypical subgenus, are regarded as belonging to a unique species, Bubalus mindorensis (see GRoVEsS 1969). All others belong to the same species as the domestic buffalo, Bubalus bubalis (Linnaeus, 1758). THomas (1911) found that LinnAeus’s (1758) description of Bos bubalis refers to the domestic buffalo, and fixed the type locality as “Italy (Rome)”. LinnaeEus’s (1758) brief ‚descriptive notes make it clear that he had domestic buffaloes in mind. Whether wild species should be referred to by names first given to domestic forms is a controversial matter, and has been most recently been discussed by GrovEs (1995 a), who refers to domestic forms as “parataxa”, rather than true taxa; as such, it can be main- tained that they have no place in biological nomenclature. Pending a new consensus on the question, in this study, names given to domestic buffalo will be ignored; the prior available name for a wild buffalo is Bubalus arnee (Kerr, 1792). The former and present distribution of Asian wild buffalo is carefully reviewed by Hepses (1995). The species is today in danger of extinction as a wild animal. On the Asıan mainland it still exists in the following pockets of population (A. CHOUDHURY, H.K. DivEKAR, J. HEINEN, T. PRAYURASUDHI, SUN HEAN, LE Vu KHoı, pers. comm.): As- sam and adjacent states, and Bhutan (several reserves); Bastar and Raipur districts, Mad- 'hya Pradesh (four reserves); Nepal (Kosi Tappu National Forest); Thailand (Huai Kha ‚Khaeng Wildlife Sanctuary); scattered populations on the Cambodia-Vietnam border. Of ‚these the only reasonably substantial population, numbering some 3300-3500, is that centred on Assam (CHOUDHURY 1994). In 1979, an undoubted wild buffalo turned up in the Panna district, far from its known present-day distribution (Hasan 1980). As later ‚pointed out (Arun SınGH 1980), this distance, about 225 km northwest of the species’ nearest present occurrence in Raipur district, is well within the known seasonal ranging distance of buffaloes in the last century. 328 C. P. GroveEs Feralized buffalo (domestic stock run wild) have established themselves in many places. In most cases there should be little problem in distinguishing them from true wild buffalo; this is especially the case in peninsular India and Pakistan, where the domestic buffalo are “River” breeds, quite different in appearance from wild buffalo. River buffa- loes have from time to time been exported to Southeast Asia, and even run wild; the type specimen of so-called Bubalus moellendorffi Nehring, 1894 (ZMB 14803), from Busuanga in the Philippines, ıs a feral river buffalo. In the case of “Swamp” Buffalo, the much less highly bred domestic buffaloes of Southeast Asia, there would be more of a problem; a specimen from the Upper Chindwin (BM 17.7.8.3) is very small and short-horned com- pared to specimens from Assam, Chittagong and Pegu, and assorts in all analyses with do- mestic swamp buffalo, but no such criteria can be applied in the case of the insular populations (see above), and it is still an open question whether any are genuinely wild, rather than feral (HEDGEs 1995). For the moment, it should be noted that the insular buf- faloes are much smaller, and have very much shorter horns, than those from the main- land. The cranıial capacity criterion (HERRE and Rönrs 1990) could presumably be applied; the true status of the wild-living buffalo of Sri Lanka and Southeast Asia will be the subject of future investigations. There is also the problem of interhreeding between wild buffalo and local domesti- cates. For the moment, let it be said that with one exception, which will be discussed be- low, all specimens of mainland wild buffalo which I have been able to study (mostly collected before the middle of this century) are sharply distinct from domestic buffalo of whatever breed. Finally, the interrelationships of the mainland groups of wild buffalo are in need of re- vision: are there subspecies; if so, how many? This problem, a rather crucial one for con- servation, is the subject of the present study. Materials and methods Skulls and horns of wild buffaloes were studied in the following collections (with their abbreviations): Natural History Museum, London (BM); Powell-Cotton Museum, Birchington, Kent (PC); Royal Scot- tish Museum, Edinburgh (RSM); National Museum of Natural History, Paris (MNHNP); Natural His- tory Museum, Marseille (MAR): Zoological Museum, Hamburg (ZMH):; Zoological Museum, Berlin (ZMB):; University Zoological Museum, Copenhagen (CPH): Natural History Museum, Florence (FIR); Royal Natural History Museum, Stockholm (RMS); Indian Museum (IM) and Zoological Sur- vey of India (ZSIC), Calcutta; Zoological Survey of India, Madras (ZSIM); Indian Forestry College, Dehra Dun (IFC); National Zoological Reference Collection, Bangkok (NRC); private collection of Van InGEn and Van InGEn Taxidermy, Mysore (VI): private collection of Maharana of Wankaner, Gu- jarat (WAN); private collection of Maharaja of Jaipur (JAI); and private collection of CHOKCHAI BULU- KUL, Bangkok (CHOK) (Tab. 2). The following skull and horn measurements (in mm) were taken on each specimen, if complete en- ough (some were frontlets and horns or stuffed heads and horns only), and if fully accessible for mea- surement (some were hung on walls and could not be moved): (1) GSL: Greatest Skull Length (ends of premaxillae to back of occiput): (2) BB: Biorbital Breadth (across lower margins of orbits); (3) PB: Postorbital Breadth (least width of forehead behind orbits); (4) OBG: Occipital Breadth Greatest (in mastoid region): (5) OBL: Occipital Breadth Least (at constriction between horn bases and mastoids); (6) BHB: Breadth Between Horn Bases (least distance across forehead between horn cores); (7) PBN: Posterior Breadth of Nasals (across nasals, at widest part at suture with frontal); (8) ABN: Anterior Breadth of Nasals (across nasals, at widest part at premaxillary sutures); (9) NL: Nasal Length (greatest length along internasal suture); (10) BL: Basal Length (ends of premaxillae to basion); (11) MT: Maxil- lary toothrow length; (12) TT: Tip-to-Tip (distance between tips of horns); (13) BT: Base-to-Tip (dis- tance between a horn tip and nearest point of base on same side); (14) SP: Span (greatest distance across lateral surfaces of horns); (15) BD: Basal Diameter (greatest diameter of one horn at base); (16) BO: Base-to-Outer (from a horn base to most distant point of curve on same side). Note that all horn measurements were taken on the sheath, not the core. The taxonomy of the Asian Wild Buffalo 329 The specimens were divided according to region: Assam, “Upper India” (meaning Nepal; from the Hopcson collection of the Natural History Museum, London), “Central India”, Thailand, and Cambo- dia. There were also individual specimens from the Chittagong Hills; the Sunderbans; Upper Chindwin; Pegu; and Song R., Lagna, S. Vietnam, which were considered separately, as was an unlocalised skull from the RSM which was identified as the type specimen of Bos arnee Kerr (see below). The different regions, and individuals, were compared by univariate (t-tests and coefficient of difference) and multi- variate methods, the latter using SPSS Discriminant on the Durras Mainframe at the Australian Na- tional University. Three Discriminant Analyses were run: (1) all cranial measurements (nos. 1-10); (2) all horn measurments (nos. 12-15); and (3) a combination of skull (nos. 1-3, 5, 7-9) and horn (12-14, 16) measurements, designed to maximise the number of available specimens. After initial runs, “Upper India” and “Central India” were combined as “C/U.India”, as was Thailand with Cambodia: despite the small sample sizes, no clear discrimination emerged between them. The analyses were thereafter based on three samples (Assam; C/U India, and Thai/Cambodia), with Chittagong, Sunderbans, Upper Chind- win, Pegu, Vietnam and the presumed type of arnee entered as unknowns. Results Sexing Rather few skulls (25 in all) are sexed; those that are show a good average difference in the diameter of their horns at the base. In the Assam and C/U.India samples, Basal Dia- meter is 138-185 (mean 166.11) mm in 9 males, 114-158 (mean 129.38) mm in 8 females; in Thai/Cambodia, 136-163 (mean 148.33) in 3 males, 100-133 (mean 118.67) in 3 females. Table 1. Comperative skull and horn variables for mainland wild buffalo (in mm) (SD = standard devia- tion; n = number) Assam C/U.India Thai/Cambodia Geatest skull length (GSL;) 561.5 SD 192 n 11 Biorbital breadth (BB) 239.4 SD 10.5 n u Upper toothrow length (MT) 160.4 SD 73 n ja MTas % GSL 28.2 SD : ci n 11 Horn span (SP) 1141.5 SD 220.8 n 15 SP as % GSL 192.6 SD 26.3 n 11 TTas % SP 56.5 s.d. 18.1 n 1 330 C. P. GrovEs Table 2. List of specimens and localities (for abbreviations of collections, see text) Mishmi Hills Sadıya Dhubri Bansbaree (= Palnsbari) Tezpur Mikir Hills Gola Ghat Guwahati Kazıranga Kuch Behar Faridpur Chittagong Hills “North from Bengal” Ganjam Raipur “Upper India” (= Nepal) Nepal “Central India” Junga Gomrapodor Torenga Indgaon Purneah Sunderbans Kukri-Mukri Pegu 28.00 N 96.00 E 27.49 N 95.38 E 26.01 N 90.00 E 26.07 N 91.30 E 26.38 N 92.49 E 26.30 N 93.00 E 26.30 N 93.59 E 26.10 N 91.45 E 26.50 N 93.30 E 26.18 N 89.32 E 232INIIAIIE 22-23 N 92-92.30 E BM 41.73, 84.1.22.4, 1938.7.1.1, Acton unreg.; KOL; CPH 288; IFC; VI; ZSIC unreg. (3); NRS unreg.; WAN; ZMB 37377 BM 91.8.7.215 (type fulvus) Bentham 10 (?paratype fulvus) BM 20.5.14.1, 12.10.31.84, 91.8.7.213 FIR 8070, —1 BM 55.1.10.1 ZMH 384 BM 44.129, -130 BM 12.10.31.83 FRI BMS72I 1 ZSIC unreg. BM 30.10.5.1 Central and “Upper” India 19.28 N 85.05 E 21.16 N 81.42 E 20.08 N 82.17 E 19.57 N 32.23E 20.11 N 82.20 E 20.05 N 82.26 E DSATNESTE2SIE 22-23 N 88-91 E ?southwestern Sun- derbans ?RSM (?type arnee) FRIM unreg. BM unreg. BM 45.1.8.142, -3 BM 59.471 BM 41.172 PEEIRG Bei BEER BEEPSFT7 BM 91.8.7.214 PC Sund.5 type septentrionalis (not seen) Thailand, Cambodia, Burma, Vietnam 177 A83NF90-31SE) Kokan, E. Siam (? = Khon Kaen, 16.25 N 102.40 E) Mae Wong, Nakhon Sawan 15.41 N 100.07 Thailand “Siam” Cambodia Langna, Song River 11.00 N 107.20 E BM 26.9.5.2 CPH unreg. NRG unreg. CHOK (7) CEPE715512624963 NRC unreg.; MAR 122, 138 MNHNP 1932 Between males and females, for India t=4.559 at 15 d.f, p<0.001; for S.E. Asia, t = 2.363 atA4.d.f.,p<0.1> 0.05. I attempted to sex other skulls on this criterion, and pub- lished mean skull lengths for the two sexes, allocated by this method (GrovsEs 1995). Considering the great overlap, however, this procedure may not be safe: and the sexual dimorphism did not amount to much in any case. While admitting that there may be some size difference, it seems most realistic to combine all specimens by sex, and reexamine the question later when more material may become available. The taxonomy of the Asian Wild Buffalo Söll Univariate and bivariate analyses The results of the univariate analyses and bivariate ratios are given in table 1. Skulls from Assam are very much larger than those from elsewhere; those from Southeast Asia (Thailand, Cambodia, Pegu) are slightly smaller than C/U.India in their skull measure- ments, but their toothrows are very much shorter. SP, however, is greatest in Thai/Cambodia, followed by Assam, followed by C/U.India. In 9 out of 26 Assam specimens and 6 out of 11 C/U.India ones, SP is less than twice GSL, but in only a single Thai/Cambodia specimen; on the other hand in 2 from Assam and 3 from Thailand SP is more than 3 times GSL. Thai/Cambodia also have much less inturned horn tips (though not as much as in domestic swamp buffaloes, in which TT is often more than 90% of SP). There is no difference between the three samples in the degree to which the horns reach posteriorly as opposed to laterally: BT as percent of SP averages 61.4 (SD 4.20) in Assam, 59.6 (SD 7.86) in C/U.India and 57.8 (SD 7.37) in Thai/Cambodia. The Royal Scottish Museum specimen, identified here as the probable type of Bos ar- nee Kerr, falls into the C/U.Indıa range, though at the upper end of it for size: GSL = 581, BB = 243, MT=158 272% of GSL), SP=1130 (194.5% of GSL), TT=51.9% of SP. None of these values are below the lower limit for the Assam sample (BB nearly is), but on the whole they are much closer to C/U.India. A specimen from Lagna, southern Vietnam, is difficult to place. GSL is 557, BB = 224, TT=57.8% of SP, all well within the Thai/Cambodia range, and different from any do- mestic buffaloes; but MT = 169 (30.3% of GSL), within the range of domestic and Sara- wak feral buffalo; SP is only 877 (157.5% of GSL), below any wild buffalo but somewhat large compared to a domestic swamp buffalo. This is the only mainland specimen that might be best interpreted as a wild/domestic hybrid. Multivariate analyses The first Discriminant Analysis, using all skull measurements alone (Fig. 1), separates the three samples completely. The Chittagong skull, classed as an unknown for the purposes -6 -4 -2 0 2 = 6 Fig. 1. Discriminant Analysis of 10 skull measurements. First Discriminant Function accounts for 80% of total variance; second, for remaining 17%. 1 = Assam, 2 = C/U.India, 3 = Thai/Cambodia, solid cir- cle = Chittagong, solid rectangle = Lagna, Vietnam. * Group Centroid. 332 C. P. Groves of the analysis, fits well inside the Assam range. The Lagna (Vietnam) skull is closest to the Thai/Cambodia dispersion, but on the edge of it. Inspection of the Discriminant Func- tion coefficients, and the correlations between the Discriminant Functions and original measurements, shows that Function 1 (horizontal) emphasises large size, contrasted with relatively short nasals and narrow posterior part of skull; Function 2 (vertical) contrasts broad skull and long nasals with narrow occiput and width between horn bases. Fig. 2. Discriminant Analysis of 5 horn measurements. First Discriminant Function accounts for 64% of total variance; second, for remaining 36%. 1 = Assam, 2 = C/U.India, 3 = Thai/Cambodia, solid cir- cle = Pegu, solid rectangle = Lagna. Vietnam. * Group Centroid. Fig. 3. Discriminant Analysis of 11 skull and horn measurements. First Discriminant Function accounts for 82% of variance; second, for remaining 18%. 1 = Assam, 2 = C/U.India, 3 = Thai/Cambodia, solid circle = presumed type of Bos arnee Kerr, solid rectangle = Pegu. * Group Centroid. The taxonomy of the Asian Wild Buffalo 388 The second analysis (Fig. 2), based on all horn measurements, separates the three sam- ples much less clearly; they differ on average only. Function 1 contrasts wide basal dia- meter and base-outer distance with lower tip-to-tip; Function 2 contrasts wide base-to-tip distance with narrow span and base-to-outer. The third analysis (Fig. 3) separates Assam well from the others, but there is a slight overlap between C/U.India and Tha//Cambodia. Function 1 contrasts overall skull size with narrower horn span; Function 2 contrasts wide horn measurements (especially tip-to- tip) with short nasals and narrow postorbital region. Discussion Nomenclature The discriminant analysis based on cranial measurements separates the Assam, C/U.India and Thai/Cambodia samples completely; on this criterion, and on their nearly complete separation in the other two discriminant analyses and their good separation on univariate and bivariate comparisons, they amply qualify as distinet subspecies. Assam buffaloes are distinguished by their very large size; the C/U.Indian form is smaller; and the Thai/Cam- bodian buffalo, small like the C/U.Indian, has small teeth and exceedingly long horns that turn in less at the tips. A skull from Chittagong fits without difficulty in the Assam sam- ple, and one from Pegu (presumably meaning the Irrawaddy delta) fits in the Thai/Cam- bodia range. A specimen from Lagna, Vietnam, may be a hybrid between the Thai/ Cambodia subspecies and domestic swamp buffaloes. The first Asian wild buffalo was described by Kerr (1792) as Bos arnee from a skull in the Edinburgh College Museum and a frontlet and horns from Weir’s Museum. He de- scribed it as having “long erected semilunar horns... Inhabits India north from Bengal... is of vast size”. His figure (No. 746) shows a typical wild buffalo skull. The fate of Weir’s Museum is not recorded, but the Royal Scottish Museum, Edinburgh, acquired all the ma- terial of the former Edinburgh College Museum (pers. comm. from the late Mr. Ian Lis- TER), but these specimens are not labelled. I examined several wild buffalo skulls from the old collections, none of known history, in the RSM, and only one of them corresponds to Kerr’s (1792) figure; its correspondance is in fact very good, both in general appearance _ despite evident mistakes in the engraving and in its measurements, if Kerr’s (1792) scale is at all accurate. It is not unlikely that this really is KERR’S specimen (Fig. 5). Its measure- ments, as discussed above, fit best with the Central/Upper India sample, although is verges towards the Assam sample somewhat, and may represent an intergrade popula- tion. The first variant of this same name that I can trace was of BLUMENBACH (1807), who called it “der Riesenbüffel”, Bos arni, and ascribed it to “den gebirgichten Ge- genden von Nord-Hindostan”. Given the lack of any more precise indication, this name is most conveniently placed in the synonymy of arnee. SMITH (1827) listed Bos arni as “the wild buffalo of the Central districts of Bengal” (Shaugur Island and the road to Patna are mentioned), and compared it with the generally domestic, short- horned form, but at the same time distinguishing it from “the Gigantic, or Taurele- ' phant Arnee” from “the upper eastern provinces and forests at the foot of the Hima- ‚ laya”, nearly seven feet high, black, with the tail barely reaching the hocks, and with ‚ horns (in a N.Bengal specimen) requiring “the outstretched arms of a man to hold the ' points”; the Common Arnee he describes as nearly a foot lower at the shoulders, with the tail reaching to near the heels, and the hide more scantily covered with hair. This “Taurelephant Arnee”, given the huge size and the distribution, may be the first indi- cation of Assam buffalo. 334 C. P. GrovEs Fig. 4. Skull of buffalo from Mae Wong, Thailand, NRC unregistered, type of Bubalus arnee theerapati new subspecies. vw u a i .. s = ee __6 6 _ E u Fig. 5. Skullin Royal Scottish Museum, probably type of Bos arnee Kerr. Hopsson (1841) listed the wild buffalo among the mammals of Nepal, with yet an- other variant of the name: “Bubalus arna, foem.Arnee, two varieties. Macrocerus, et Speirocerus, Nob.H.T. (Habitat Terai)”. Hopsson (1841) appears never to have described his two “varieties”, but Gray (1852) in the course of cataloguing British Museum skulls, under Bubalus buffelus, refers to HopGson’s forms: — Two skulls with horns, wild variety. India. Presented by B.H.Hopcson, Esq. B. B. macrocerus, Hodgson, 1. c., 912. -— Skull, with horns, tame variety. India. Presented by B.H.Hopcson, Esg. B. B. spirocerus |sic], Hodgson, 1. c., 912. The taxonomy of the Asian Wild Buffalo 335 (The page numbers refer to Hopcson’s 1841 study). The museum (now the Natural His- tory Museum) actually has three adult skulls and one immature, as well as two frontlets with horns, presented by Hopcson in 1845. The three adult skulls are 604 k, I and, j (old registrations) = 45.1.8.142, -3 and -4 (current registrations); the first two, male and female respectively, are listed as “wild” by Gray (1873), the third as “domestic variety”. Inferen- tially, then, the first two would be syntypes of Hopsson’s (1841) macrocerus, the third the holotype of his sp/eJirocerus which, being domestic, is of no further concern to us here. They are presently labelled as being from “Upper India”. Gray (1852) quotes Hopsson directly, giving no source (perhaps a letter) or date: “In the wilderness, as in the cow-house, there is a marked distinction between the long- (Macrocerus) and curve-horned (Spirocerus) buffaloes”. These characterisation of macro- cerus, rudimentary though it is, would serve to make the name available; the author and date of the name is thus GrAy, 1852. Today’s Nepalese domestic buffalo is of mixed breed, more Swamp than River type (CockrıLL 1974) though some at least have rather well-curved horns more like a River buffalo (figs. 165, 166); this may be what is meant by the “curve-horned (Spirocerus)” variety, although its evident occurrence in the wild, and that of the “long-horned (Macro- cerus)” variety in a domestic state, is curious. A wild buffalo, captured in Kosi Tappu, in Katmandu Zoo, has wıde-spread horns, and Dr JoEL HEINEN (pers. comm.) states that this is usual in the Kosi Tappu buffaloes although there is some variation. BLANFORD (1891: 492) mentioned Hopcson’s (1841) two forms as if they were both wild, describing macrocerus as having the horns “almost straight till near the end, where they turn more rapıdly upward”, and spirocerus as “with horns approaching a circle”; and in addition “a very distinct race of a dun colour that inhabits Upper Assam”, which he de- scribed as Bos bubalus, var. fulvus. He mentioned “two heads of bulls”, in the British Mu- seum and the Indian Museum - from the context, he evidently meant skulls, as he described a convex forehead and short rostrum and nasals. The Indian Museum specimen must be the one from Sadıya (which is ın “Upper” Assam, i.e. along the upper reaches of the Brahmaputra ın Indian terntory), no. 10 in BENTHAM'’s (1908) list. This is probably an unregistered skull now in the Zoological Survey of India, which has the requisite charac- ters: ıts measurements correspond approximately to those given by BENTHAM (1908), and it has short nasal bones only (223 mm, nearly all others being above 240 mm). The British Museum (= Natural History Museum, London) skull, however, is readily identifiable: BM 91.8.7.215, a complete skull, was collected in the Mishmi Hills (as far up the Brahma- ‚ putra as one can get and still be in India) by Sir G. CAMPBELL and presented to the mu- seum by A. ©. Hume in 1891. Stick-on labels give it the Hume Catalogue number 44.A.l, and ıt has a noticeably convex forehead and short rostrum, including nasals which are 230 mm long. Of these two skulls the more certain identification ıs the London one, which may be fixed as Lectotype of fulvus. Finally MATscHıE (1912) described a new subspecies, Bubalus bubalis [sie] septentrio- nalis, from “Kuckri-Muckri, N. W. Vorderindien”, from a specimen collected by the Crown Prince of the German Reich and of Prussia; Kukri-Mukri (sic) is a somewhat iso- lated island in the Bay of Bengal, off the Sunderbans, the Ganges Delta islands, (Fr. SIGRID RITTHALER, Archivist of the Hohenzollern family, pers. comm.). All these names, with the exception of fulvus Blanford, refer to buffaloes from regions . where the “C/U.India” form is known; indeed, in some cases the type specimens them- ' selves have been studied by me, and fall within this form. No names are available for the ‚ mainland Southeast Asian wild buffalo, which below is described as new. Grovss (1994) included a sample of supposedly wild buffaloes from “Bihar” in a preli- minary analysis. In retrospect, these are more likely to be domestic (swamp) buffaloes. Even discounting this sample, the hypothesis of derivation of domestic buffaloes from something nearest the present-day Central Indian stock seems quite plausible. 336 C. P. GrovEs Taxonomy Three subspecies of the wild buffalo of the Asian mainland may be distinguished, as fol- lows: Bubalus arnee arnee Kerr, 1792 Synonyms: arni Blumbenbach, 1807; macrocerus Gray, 1852; septentrionalis Matschie, O1, Distribution: formerly, from the Sunderbans (Ganges delta) southwest into Madhya Pradesh and Andra Pradesh, and northwest into Nepal. Still occurs in the Raipur and Bas- tar districts of eastern Madhya Pradesh, and the Kosi Tappu reserve, southeastern Nepal. Diagnosis: a small subspecies, with somewhat larger teeth than others, and horn span less. Greatest skull length generally less than 570 mm, horn span usually less than 1 200 mm., tip-to-tip distance nearly always less than 80% of span. Toothrow length more than 27% of skull length. Notes: the Panna buffalo (HAsan 1980) was black in colour with contrastingly white lower limbs below the knees and hocks, and a white muzzle. The tail reached about to the hocks. Photos of buffalo from Kosi Tappu, Nepal, in the Katmandu Zoo are very similar. This subspecies is reduced to tiny remnant stocks, about 100 animals in Kosi Tappu, Nepal (JoEL HEINEN, pers. comm.); and, in 1988, 25 in Uddanti Reserve and 27 spread be- tween Indravati, Bhairamgarh and Pamed Reserves, Madhya Pradesh (H.K. DivEKAR pers. comm.). Bubalus arnee fulvus Blanford, 1891 Distribution: Brahmaputra valley, formerly from Kuch Behar to the Mishmi Hills; south to the Chittagong Hills. Still occurs in the Brahmaputra valley from about 92.30 to 96.00 E, and in the Manas district on both sides of the India/Bhutan border (CHOoUDHURY 1994). Diagnosis: a subspecies of very large size, with widely spreading horns whose tips are well turned in. Greatest skull length usually over 570 mm, horn span more than 1 100 mm in most specimens, tip-to-tip distance (as in nominotypical arnee) less than 80% of span in almost all specimens. Toothrow length approximately 26-28% of skull length. Notes: photos of buffalo in Kaziranga and Manas seem to indicate a lighter grey or brownish grey, less blackish, animal than Kosi Tappu specimens or the Panna buffalo, and in particular less contrastingly white on the limbs. Interestingly, given the remarks of SMITH (1827) quoted above, the tail always falls well short of the hocks. Further observa- tions are needed to demonstrate whether these apparent differences hold good over lar- ger series. A. CHOUDHURY (pers. comm.) estimated the following population numbers for this subspecies in 1992: Manas (Assam and Bhutan), 1200; Kaziranga and adjacent areas (Assam), 1100; Laokhowa and adjacent areas (Assam), 200; Laikhimpur (Assam), 100- 150; Dibru-Saikhowa and adjacent areas (Assam and Arunchal), 500; Balkharam (Me- ghalaya), 200; scattered populations in Assam, 29-32; total, 3 300-3 500. Bubalus arnee theerapati new subspecies Distribution: formerly from the Irrawaddy delta through Thailand to Cambodia, and probably Vietnam. Now known mainly from Huai Kha Khaeng Wildlife Sanctuary, Thai- land; reported from Cambodia. Type specimen: NRC, Bangkok, unregistered skull, from Mae Wong, Nakhon Sawan, Thailand. (Fig. 4). Judging from its rather slender horns, the specimen may be a female. Hypodigm: “Koken, E. Siam”, 1; Thailand (including “Siam”), 9; Cambodia, 3; Pegu (probably Irrawaddy Delta), 1. Skulls (with horns) only. The taxonomy of the Asian Wild Buffalo 33 Diagnosis: the smallest mainland subspecies on average, with small teeth; horns excep- tionally widely spreading, with tips less inturned. Greatest skull length below 570 mm as in nominotypical arnee; horn span usually over 1200 mm, tip-to-tip distance more than 80% of span. Toothrow length 24-27% of skull length. Notes: the only photo I have seen of this subspecies, a distance shot taken by Mr T. PRAYURASUDHI, shows a gray anımal, but apparently rather contrastingly marked with white. Photos in LEKAGUL and McNEEIY (1977) were taken in Kaziranga, and do not de- pict the present subspecıies. The number in Huai Kha Khaeng does not exceed 50 (T. PRAYURASUDHI, pers. comm.). A few still exist in far eastern Cambodia (Sun HEANn, pers. comm.). A pair of wild buffalo horns was found in Yok Don, Vietnam, on the Cambodian border, in 1987 (LE Vu KnHoı, pers. comm.). Etymology: the name honours Mr THEERAPAT PRAYURASUDHI (Royal Forestry Depart- ment, Thailand), whose continuing fieldwork in Huai Kha Khaeng has added notably to our understanding of the ecology of gaur, banteng and the 50-100 remaining wild buffalo, laying a sound basis for their conservation. Acknowledgements Grateful thanks are due to the following curators for allowing access to material in their care, and for as- sistance in studying it: P. JENKINS, J. CLUTTON-BROCK, L. BARTON, the late JAN LISTER, M. TRANIER, P. BAYLE, H. SCHLIEMANN, R. ANGERMANN, F. W. BRAESTRUP, L. AZZAROLI, L. WERDELIN, the late B. BıswAs, S. CHAKRABORTY and the staff of the Indian Museum, G. U. Kurup, staff of the Indian For- estry College, DEHRA Dun, staff of the National Zoological Reference Collection, Bangkok, ]. and R. van InGEn, the Maharana of Wankaner, the staff of the Maharaja of Jaipur, and the late Dr. BoonsonG LEKAGUL1. Finally I thank my friends and colleagues at the CAMP workshop, Chonburi, July, 1995, at whose behest this study is written; M. K. RanjJItsınH for sending me several photos of the Panna buffalo and Ms. Shoshana Greenburg for computing assistance. Zusammenfassung Taxonomie des Wasserbüffels (Bubalus arnee) auf dem asiatischen Festland Schädel und Hörner des Wasserbüffels (Bubalus arnee) aus Indien und Südostasien wurden morphome- trisch untersucht. Trotz einer hohen individuellen Variationsbreite konnten drei Unterarten unterschie- den werden. Eine davon, die Population in Thailand und Kambodscha umfassend, wurde in der vorliegenden Arbeit neu beschrieben. Aus dem Verlust oder der Fehlzuordnung von Typusexemplaren resultierende nomenklatorische Probleme wurden geklärt. References ARUN SINGH, T. (1980): Feedback: “The enigmatic buffalo”. Hornbill 1980, 4. BENTHAM, T. (1908): An illustrated Catalogue of the Asiatic Horns and Antlers in the Collection of the Indian Museum. Calcutta: Indian Museum. BLANFORD, W. T. (1891): Fauna of British India. London: Taylor and Francis. BLUMENBACH, J. F. (1807): Handbuch der Naturgeschichte. 8th ed. Göttingen: Heinrich Dieterich. CHOUDHURY, A. (1994): The decline of the wild water buffalo in north-east India. Oryx 28, 70-73. CockRILL, W. R. (1974): The Husbandry and Health of the Domestic Buffalo. Rome: Food and Agricul- ture Organization of the United Nations. Gray, J. E. (1852): Catalogue of Mammalia. Vol. 3, Ungulata. London: British Museum Trustees. GRAY, J. E. (1873): Handlist of the Edentata, Thick-skinned and Ruminant Mammals in the British Mu- seum. London: British Museum Trustees. 338 C. P. GroveEs GROvES, C. P. (1969): Systematics of the Anoa (Mammalia, Bovidae). Beaufortia (Amsterdam) 17, 1-12. GRrovESs, C. P. (1995 a): On the nomenclature of domestic animals. Bull. Zool. Nomendcl. 52, 137-141. GROovESs, C. P. (1995 b): Domesticated and commensal mammals of Austronesia and their histories. In: The Austronesians: Historical and Comparative Perspectives. Ed. by P. BELLwooD, J. J. Fox, and D. Tryon. Canberra: Research School of Pacific and Asian Studies, Australian National University. Pp. 152-163. Hasan, S. M. (1980): The enigmatic buffalo. Hornbill 1980, 26-27. HEDGESs, S. (1995): Asian Wild Cattle and Buffaloes: Status Report and Conservation Action Plan. 2 vols. WWF Project 3950/Asia. HERRE, W.; RÖHRs, M. (1990): Haustiere — zoologisch gesehen. 2. Aufl. Stuttgart, New York: Gustav Fischer. Hopsson, B.H. (1841): Classified catalogue of Mammals of Nepal (corrected to end of 1841, first printed in 1832). J. Asiat. Soc. Bengal 10, 907-916. KeERR, R. (1792): The Animal Kingdom of the celebrated Sir Charles Linnaeus. Class I. Mammalia. Lon- don: J. Murray. LEKAGUL, B.; McNEETY, J. A. (1977): Mammals of Thailand. Bangkok: Assoc. Conservation of Wildlife. LinNAEUSs, C. (1758): Systema Naturae per Regna Tria Naturae, secundum Classes, Ordines, Genera, Species, cum Characteribus, Differentiis, Synonymis, Locis. I: Regnum Animale. 10th ed. Holm: Laurentius Salvius. MATSCHIE, P. (1912): Beutestücke aus fernen Ländern. Dt. Jäger-Zeitung 59, 97-103. SMITH, C. H. (1827): Vol. 4 of GRIFFITH.: E. The Animal Kingdom, arranged in conformity with its orga- nisation by the Baron Cuvier. London: Geo. B. Whittaker. THomas, ©. (1911): The Mammals of the Tenth Edition of Linnaeus; an Attempt to fix the Types of the Genera and the exact Bases and Localities of the Species. Proc. Zool. Soc. Lond 1911, 120-158. Author’s address: Corın P. Grovss, Department of Archaeology and Anthropology, Australian Na- tional University, Canberra, A. C. T. 0200, Australia. Z. Säugetierkunde 61 (1996) 339-351 us © 1996 Gustav Fischer, Jena S ÄUG ETIE = N DIE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Cytogenetics of the genus Arvicanthis (Rodentia, Muridae). 2. The chromosomes of three species from Ethiopia: A. abyssinicus, A. dembeensis and A. blicki. By M. Corri, MARIA VITTORIA CIVITELLI, R. CASTIGLIA, AFEWORK BEKELE, and E. CAPANNA Dipartimento di Biologia Animale e dell’Uomo, Universita di Roma ‘La Sapienza’, Roma, Italia and Biology Department, University of Addis Ababa, Addis Ababa, Ethiopia. Receipt of Ms. 15. 02. 1996 Acceptance of Ms. 05. 05. 1996 Abstract The karyotypes of three Ethiopian species of Arvicanthis, the lowland A. dembeensis, the highland A. abyssinicus, and the endemic mountain species A. blicki are described. Their diploid numbers and autosomal Fundamental Numbers (aFN) are 2n=62 (aFN = 62), 2n=62 (aFN =64), and 2n =48 (aFN = 64), respectively. Except for a large submetacentric (no. 4) in A. abyssinicus and a small meta- centric in A. dembeensis and A. abyssinicus, all their chromosomes are acrocentric, while the karyotype of A. blicki is characterized by having seven pairs of Rb metacentric chromosomes. C- and G-banding patterns and Ag-NORs positions were used to identify chromosomal rearrangements (Rb transloca- tions, pericentric inversions and heterochromatin additions). Phylogenetic hypotheses are suggested based upon the comparisons among karyotypes investigated and with the previously described karyo- type of ‘A. niloticus’ from West Africa (Benin). Introduction Arvicanthis, the unstriped grass rat, is a rodent very common to the South and East of the Sahara and North of the Zambezi (Kınapbon 1974). The genus comprises of few named species, with A. niloticus (Desmarest, 1822) being the most common in the Nile basin and in central and western sub-Saharan regions, and a group of species present in the Horn of Africa and East Africa, i.e. A. abyssinicus Rüppel, 1842, A. dembeensis Rüppel, 1842, A. somalicus Thomas, 1902, and A. blicki Frick, 1914, according to CorBET and Hırı (1991), or A. abyssinicus, A. blicki, A. somalicus, and A. nairobae Allen, 1909, according to MuSSER and CARLEToN (1993). However, cytogenetic findings (CAPANNA and CIVITELLI 1988; CivitELLı et al. 1995; GRANJoN et al. 1992; MATTHEY 1965; VIEGAS-PEQUIGNOT et al. 1983; VoLOBOUEV et al. 1987, 1988) suggested that A. niloticus represents rather a species complex with different karyomorphs. A. abyssinicus is adapted to the highlands between 2000 and 3000 m above sea level and it reaches 3700 m in Simien. A. blicki has a restricted range occurring at the higher altitudes of the Bale mountains between 2750 and 4100 m (YALDEN and LARGEN 1992). Finally, the range of A. dembeensis extends from sea level up to 2000 m (YALDEN et al. 1976). A. abyssinicus and A. blicki are genetically related (they diverged during recent Pleis- tocene), as shown by Nei’s genetic distances, and have an older divergence from A. dembeensis (CaPuLa et al. 1996). 340 M. Corrietal. A multivariate investigation of morphometric traits (AFEWORK BEKELE et al. 1993) has shown that A. dembeensis and A. abyssinicus are distinct and that there is a clear pattern in morphometry related to altitudinal variation suggesting trends in adaptation to these very dif- ferent environments. Nonetheless, the taxonomy of A. dembeensis ıs still puzzling: CoRBET and Hırr (1991) and YALDEN et al. (1976) consider it as a separate species, while MusseEr and CARLETON (1993) include it within A. niloticus. There is an apparent correspondence between the karyotypes of the two (at least for the Egyptian populations of A. niloticus, VIEGAS-PE- QUIGNOT et al. 1983), so that A. dembeensis could be a geographic variant of A. niloticus. How- ever, to solve its taxonomic status, a clarification of the entire ‘niloticus’ complex is needed. ‚In the present study we describe the karyotypes of three species reported by DE WINTON (1900) and YALDEN and LARGEN (1992) in their study of the Ethiopian fauna, i.e. A. abyssinicus, A. dembeensis, and A. blicki. Moreover, based upon chromosomal rear- rangements, we suggest their phylogenetic relationship. Material and methods Animals were collected from the following localities: A. abyssinicus: Sululta (9° 15’ N-38° 43’ E, 2700 m a.s.l.), cropland, 13 and 4%; Managhesha (9° 00’N-38° 35’ E, 2200-2 300 m a.s.l.), savanna and forest limits, 53. A. dembeensis: Koka (08° 24 N-39 01’E, 1650 m a.s.l.), open savanna and cropland, 25 and 16 12 34 38 42 46 Fig. 1. Map of Ethiopia demonstrating distribution of A. dembeensis (open circles), A. abyssinicus (closed circles) and A. blicki (open triangles) (redrawn from YALDEN et al. 1976), with the location of trapping sites. 1 = Koka; 2 = Sululta; 3= Managhesha; 4 = Bale. Isohypses represent 1 600 m asl. Ethio- pia and Benin are shown in dark in the inset. Chromosomes and phylogeny of Arvicanthis 341 12. A. blicki: Sannetti Plateau, Bale Mts. (6° 52’N-39° 52’ E, 4000 m a.s.l.), alpine moorland, 1%. The karyotypes of samples of A. niloticus from Benin (CiVITELLI et al. 1995) were used for outgroup compar- ison and are from Lokossa (1° 37’ E-6°43’N), open savanna, 45 and 5%; Attogon (2° 10’ E-6°43’N), cropland; 1 and 2%, and Toffo (2° 6° E-6° 49’ N), forest, 1% (all at sea level) (Fig. 1). Metaphases were obtained from bone marrow preparations (Hsu and PArron 1969) at Addis Ababa University. Slides and cell suspensions in fixative solution (3:1 methanol/acetic acid) were then trans- ported at 4°C to the Dipartimento di Biologia Animale e dell’Uomo, Rome, for differential staining. Metaphases standard staining was performed by Giemsa 4% in phosphate buffer pH7. G-bands were obtained following SEABRIGHT (1971) and C-bands according to BıcKHAM (1979). Nucleolus Orga- nizer Regions (NORSs) were enhanced by means of silver reaction following Howeıı and BLAck (1980). To identify chromosomes carrying Ag-NORs, the silver stained metaphases were also treated with tryp- sin to highlight the G-banding. To construct phylogenetic relationships between species, homologies in chromosomal elements were identified and then coded into binary form (additive binary coding, SNEATH and SokAL 1973). To find the shortest tree (Wagner parsimony) a global swapping procedure was used and characacter state trees were polarized using, as an outgroup, A. niloticus from Benin. Besides geographic considerations, the use of this species as an outgroup is supported by our find- ings from multi-locus protein electrophoresys (CAPuLA et al. 1996) which indicates a very high Nei’s ge- netic distance (D = 0.847) between the Benin and the Ethiopian species and a lower Nei’s genetic distance (D = 0.186) among the latter. Furthermore, there is also the advantage that this is the only po- pulation for which C- and G-banding and Ag NOR staining are available (Civiternı et al. 1995), and this allows proper character state tree polarization. Chromosomes carrying Ag-NORs were used cautiously (see SCHUBERT and Wopus 1985) and then only when a peculiar localization of ribosomal cistrons was found and/or if shared by two or more branches. Results Karyotype descriptions A. dembeensis. The diploid number is 2n=62, the autosomal Fundamental Number (aFN) is 62. All autosomes are acrocentrics of decreasing size, except a metacentric pair of small size (Fig. 2). The X-chromosome is a large submetacentric (the largest of the complement) and the Y-chromosome is a metacentric of smaller size (Fig. 2). All auto- somes are characterized by a large centromeric heterochromatic region (Fig. 3). The Y- chromosome is entirely C-positive. The short arm of the X-chromosome is completely het- erochromatic and the long arm shows a C-positive region near the centromere and a smal- ler one before the distal end (Figs. 3 and 4). A total of 7 chromosomes carrying NORs was found. These NORs have a telomeric position in two pairs of large chromosomes (nos. 2 and 5) and are paracentromeric in two other pairs of medium-size chromosomes (probably nos. 18 and 19) (Fig. 3). Chromosomes were numbered according to their size (no. 1 being the largest). The karyotype of A. dembeensis was chosen as the reference point for comparison of other species, since its diploid and fundamental numbers are most similar to those of A.niloticus from terra typica (VIEGAS-PEQUIGNoT et al. 1983). The Karyotypes of A. abyssinicus and A. blicki were consequently numbered based upon the homologies identified with the chromosome elements of A. dembeensis. The notation used by CivITEL- rı et al. (1995) for the chromosomes of A. niloticus from Benin will be shown in brackets when the karyotypes are compared (Fig. 5). A. abyssinicus. The diploid number is 2n = 62 and the aFN is 64. All autosomes are acrocentrics which decrease in size, except a submetacentric pair of medium size and a metacentric pair of small size (Fig. 1). All autosomes are characterized by a large centro- meric heterochromatic region (Fig. 3). A variability in shape and size of the X-chromo- some was observed: two females out of four have shown a heteromorphic condition for 342 NnAN aonmn nndnnnanann On oa mann on AN NA Dem MA Hr Am Are AM AM aa Ar Ri un er De = A x < Arvicanthis dembeensis 2n=62 aFN=62 AAN ANAaAAn AANO MA HR AN OA HR N Bun nn N MAN AA AR AG AA IEOCAum Mm AA Ma am Arvicanthis abyssinicus 2n =62 aFN = 64 x Arvicanthis blicki 2n = 48 aFN = 64 Ku X Fig. 2. Karyotypes of A. dembeensis, A. abyssinicus and A. blicki. The karyotype of A. dembeensis has been arranged according to the chromosomes decrease in size. The lower diploid number of A. blicki (2n = 48) is determined by seven pairs of Rb metacentrics. Note the fourth pair which is acrocentric in A. dembeensis but submetacentric in A. abyssinicus and A. blicki. Note also the smallest metacentric (the first in each karyotype third row) which is common to all three species. Chromosomes and phylogeny of Arvicanthis 343 Abb. 3a Fig. 3. Part A: Metaphase plates of the three Ethiopian species with, on the left, C-bands (arrows indi- cate sex chromosomes), and, on the right, Ag-NORSs (arrows indicate chromosomes with Ag-NORs). Part B: Chromosomes carrying Ag-NORs in the three Ethiopian species. 344 M. Corrietal. ie] gg WW, Arvicanthis dembeensis Arvicanthis blicki Abb. 3b this chromosome that can be either subtelocentric or submetacentric, with the latter con- figuration similar to that found in A. dembeensis (Fig. 4). The large submetacentric var- iant of the X-chromosome possesses large heterochromatic blocks in the short arm not present in the subtelocentric form (Figs. 3 and 4). All the six males analysed presented a submetacentrice X-chromosome. The Y-chromosome is a metacentric of medium size (Fig. 2) and entirely C-positive (Figs. 3 and 4). A total of nine chromosomes with NORs were detected. Three acrocentric pairs (chromosomes nos. 1, 7 and 12) have NORs at the centromere, one pair (no. 2) has both telomeric and centromeric NORs, and another pair (no. 5) has telomeric NORSs only (Fig. 3). A. blicki. The species is characterized by 2n = 48 and by aFN = 64. The autosomal set is composed by 14 pairs of acrocentrics and 9 pairs of biarmed chromosomes (meta- centrics and submetacentrics) (Fig. 2). The X-chromosome is submetacentric (Fig. 2), and its G- and C-banding patterns are identical to those found in A. dembeensis and to the equivalent form described in A. abyssinicus (Fig. 4). As no males were analyzed, there is no information relative to the Y-chromosome. All biarmed chromosomes are character- ized by C-positive centromeric regions; acrocentric chromosomes have centromeric het- erochromatic spots except for a few pairs of medium size (Fig. 3). C-bands in A. blicki are less evident than in the two species described above. A total of six chromosomes have shown NOR:s. The NORs occupy a centromeric position on the largest chromosome pair (no. 1), a telomeric position on the smallest (no. 5) and have both a telomeric and centro- meric position on chromosome no. 2 (Fig. 3). Chromosomes and phylogeny of Arvicanthis 345 Xm X sm ex Y A. blicki A. demb. A. abyss. Pi BG 3 BE A. nilot. Fig. 4. The sex chromosomes of A. blicki, A. dembeensis, A. abyssinicus, and A. niloticus. Figure 4 re- presents the metacentric (X m), submetacentric (X sm) and subtelocentric (X st) shapes found for the X chromosome. For each condition the G- (left) and C-banding (right) patterns are shown, as well as for the Y chromosome. Arrows indicate the plausible rearrangement sequence for variants of the X chromosome; 1 indicates pericentric inversion, 2 and 3 indicate heterochromatin addition. Y chromo- some has not been described for A. blicki. Karyotype comparisons The complete comparison of homologies and chromosomal rearrangements as detected by G-banding is shown in figure 5. The identification of homologies has been possible without ambiguity for most of the chromosomes. There is a minor group of chromosomes 346 M. Corrietal. d = Arvicanthis dembeensis 4 Ah a “ #- a= A. abyssinicus x nun ’ a hd b = A. blicki 5 “% beue n = A. niloticus mi = “Mr Chromosomes and phylogeny of Arvicanthis 347 (at the bottom of Fig. 5) for which the small size and the limited number of G-bands do not allow the identification of homologies. Within this group, there is a medium sized me- tacentric of A. blicki whose arms were not identified. Character state trees (cst) were constructed to describe changes in each chromosome in order to allow the identification of phylogenetic relationships between the species (a=A. abyssinicus, b=A.blickiı, d=A. dembeensis and n=A.niloticus). The form in which character state trees are represented follows the standard representation as in For£y et al. (1992). In case of doubt, the choice was made assuming the most common re- arrangement as the most primitive. Comparisons are as follows: Chromosome 1 — the chromosome has the same acro- centric morphology and banding pattern in the three Ethiopian species, while A. niloticus shows a pericentric inversion; NORs are centromeric in A. abyssinicus, A. blicki and A.niloticus and absent in A. dembeensis (cst=n- ab-d). Chromosome 2 - the chromo- some has the same acrocentric morphology and banding pattern in the three Ethiopian spe- cies, while it is a subtelocentric in A. niloticus; NORs have a telomeric and centromeric position in A. abyssinicus and A. blicki, while they are telomeric in A. dembeensis and A. niloticus (cst =n - d- ab). Chromosome 3 - the chromosome has the same morphology and banding pattern in the three Ethiopian species, while in A. niloticus it is submeta- centric (cst=n- abd). Chromosome 4 - the chromosome is acrocentric in A. dembeensis, submetacentric in A. abyssinicus and A. blicki, and acrocentric with a deletion in A. niloticus (cst = n - d- ab). Chromosome 5 — the chromosome is an acrocentric with the same morphology, banding pattern and telomeric NORs in all four species. Chromo- somes 6 and 8 - the chromosomes have the same acrocentric morphology and banding pat- terns in all four species. Chromosome 7 - the chromosome is acrocentric in all four species; two distinct banding patterns characterize the centromeric region in A. dembeensis and A. abyssinicus on one side, and A. blicki and A. niloticus on the other; A. abyssinicus pos- sesses Ag-NORSs in a centromeric location (cst = bn - d- a). Chromosome 9 - the chromo- some is a large acrocentric in A. niloticus; it is amedium sized acrocentric in the Ethiopian species; A. dembeensis and A. abyssinicus show an additional centromeric band (cst=n- ad-b). Chromosomes 10 and 14 - chromosome 10 is a medium size acrocentric in A. dembeensis, A. abyssinicus and A. niloticus, with a dark centromeric band in the former two; chromosome 14 is an acrocentric shared by the same three species with the same band- ing; the two chromosomes form the 10/14 Robertsonian (Rb) metacentric in A. blicki (cst=n-ad-b). Chromosome 11 — chromosome 11 is a medium size acrocentric in A. dembeensis, A. abyssinicus and A. niloticus; this chromosome forms the 11/? Rb meta- centric in A. blicki through fusion with an unidentified element (?) in A. dembeensis and A. abyssinicus but which, on the contrary, was identified by CiviteLıı et al. (1995) in A.niloticus and numbered as 21 (cst=ad-n-b). Chromosomes 12 and 13 -— chromo- some 12 is medium sized acrocentric in A. dembeensis, A. abyssinicus and A. niloticus; A. abyssinicus has centromeric NORs; chromosome 13 is an acrocentric shared by the same Fig. 5. G-banding comparison of the autosomes of the three Ethiopian species (A. dembeensis, A. abyssinicus, A. blicki, respectively in first, second and third position) and A. niloticus from Benin (in fourth position). The chromosomes of A. dembeensis have been numbered according to their decrease in size. This species constitutes the reference point for the chromosome numbering of the others. Rober- stonian metacentrics are shown twice near the corresponding acrocentric elements. Question marks {?) indicate arms that were not identified. Chromosome numbering assigned by CivITELLI et al. (1995) to A. niloticus is preserved and enclosed into brackets. Elements from number 20 onwards (chromosome 25 excluded) for which identification was difficult or impossible are shown in the four rows at the bot- tom of the figure. 348 M. Corrtietal. three species with the same banding pattern; these two chromosomes are fused in A. blicki into the 12/13 Rb metacentric (cst=nd-a-b). Chromosomes 15 and 16 - chromo- somes 15 and 16 are the same medium sized acrocentric in A. dembeensis, A. abyssinicus and A. niloticus; chromosome 16 shows ın A. niloticus an addition at the pericentromeric region; the two chromosomes form the 15/16 Rb metacentric in A. blicki (cst=n-.ad-.b). Chromosome 17 — chromosome 17 ıs one of the smallest acrocentrics of intermediate size with the same morphology and banding described for A. dembeensis and A. abyssinicus; this chromosome corresponds to the small acrocentric numbered as 23 in A. niloticus by Cı- n N <. & — 8 .a8 & Ss Ar 6) o o © % more. 5 Ser Sole Eee (3) © © 8 = SS 83% sSsS=s < BR HHHHHHHH © (&) KR) (S ST SEN I im Fig. 6. Phylogeny of Ethiopian Arvicanthis, with the most parsimonious trees found. Homoplasies in the tree in part A concern chromosome 7 and the two Rb metacentrics 11/? and 17/?; homoplasies in the tree in part B concern chromosomes 2, 4 and the Rb metacentric 12/13. Cladograms have been out- group rooted on A. niloticus. Letter a indicates the most primitive chromosome condition, b the inter- mediate and c the most derived. Chromosomes and phylogeny of Arvicanthis 349 VITELLI et al. (1995); this chromosome forms the 17/2? Rb submetacentric in A. blicki. The unidentified (?) arm corresponds to an acrocentric chromosome of A. niloticus numbered as 16 by Civiterni et al. (1995), but not yet identified in A. dembeensis and A. abyssinicus (cst=ad-n-b). Chromosomes 18 and 19 - these chromosomes are acrocentrics in A. dembeensis, A. abyssinicus and A. niloticus; the former two carry centromeric AgNORSs; the two chromosomes form the 18/19 Rb metacentric in A. blicki (cst = d- na - b). Chro- mosomes 20-30 - all these chromosomes are acrocentrics in A. dembeensis and A. abyssinicus, except one small metacentric (no. 25) which is common to the three Ethio- pian species. Also this chromosome possibly corresponds to no. 30 in the karyotype of A. niloticus from Benin. Since it has not been possible to identify homologies in banding patterns from chromosome 20 onwards, these have been excluded from a phylogenetic as- signment; for their general morphology, see the previous section. A most parsimonious solution was computed using the character state trees of chro- mosomes 1, 2, 4, 7, 9, 10/14, 11, 12/13, 15/16, and 17. Two trees of equal length were pro- duced (Fig. 6). The most primitive chromosome condition is indicated in the figure as a, the most derived as c and the intermediate as b. The two trees differ in the phylogenetic relationships of A. abyssinicus which, in one case, is connected to A. blicki (part A) and in the other to A. dembeensis (part B). The length of the trees is 25 and the consistency in- dex is 0.88. Three homoplasies occur in the trees (indicated as filled rectangles); these concern chromosomes 7, 11/? and 17/? in the first tree, and chromosomes 2, 4 and 12/13 in the second tree (Fig. 6). Discussion A. niloticus from Benin belongs to the group of the three or more species identified by VOLOBOUEV et al. (1988), GRANJoN et al. (1992) and Civiternı et al. (1995) in Central and West Africa. There is a series of chromosomal rearrangements characterizing this lineage that are different from the karyotype of A. niloticus from terra typica (VOLOBOUEV et al. 1988). The Benin population is therefore the sister group of the Ethiopian species and is the proper outgroup for character polarızation. Furthermore, its divergence from the Ethiopian group has been placed back at the Early Pliocene (about 4.2 Myr), as estimated from Nei’s genetic distances (CAPANNA et al. 1996; see also YALDEN and LARGEN 1992). The three Ethiopian species have karyotypes which are more similar to each other and constitute a monophyletic group. A. dembeensis and A. abyssinicus differ only for one pericentrie inversion and for some of the chromosomes carrying Ag-NOR:. A. blicki has the karyotype showing the greatest differences, with a diploid number reduced by Robert- sonian fusions, as evidenced by the nFÄA (64) which is equal to that of A. abyssinicus. The isolation of A. blicki and its range restricted to the Afroalpine habitat of the Bale Moun- tains account for its peculiar diversity (Corti et al. 1995). It is still an open question which one of the two alternative equivalent trees may re- present the true phylogeny of Ethiopian Arvicanthis, i.e. whether A. abyssinicus is the closest relative to A. dembeensis or to A. blicki. The similar adaptation to the highlands in A. abyssinicus and A. blicki, together with Ag-NORs position, would favour the sec- ond hypothesis. Indeed, the two pairs of chromosomes carrying Ag-NORs in A. blicki (nos. 1 and 2) are the same as in A. abyssinicus. In particular, the peculiar position of Ag- NORS in the second pair of these two species, i.e. the nucleolus organizer regions which are both telomeric and centromeric, constitutes a synapomorphy (i.e. it derived from their direct common ancestor) rather than a parallelism. Furthermore, the pattern of ge- netic relationships as deduced from Nei’s genetic distances (CAPpuLA et al. 1996) points in favour of the hypothesis of a close phylogenetic relationship between A. abyssinicus and A. blicki. 350 M. Corrietal. The sex chromosomes provide some insight into the systematics of these species. They show very similar hetero- and euchromatic patterns in the three Ethiopian species, and this reinforces the monophyletic hypothesis for the Ethiopian group. A high intra-popula- tional variability has been described by Civiterti et al. (1995) in the Benin population with three different variants, 1.e. subtelocentric, submetacentric, and metacentric. Rear- rangements involved are, respectively, a pericentric inversion and addition of heterochro- matin. The same subtelocentric form occurs in A. abyssinicus with the same G- and C- banding pattern. However, there is a second variant of the X chromosome in A. abyssinicus which ıs submetacentric and differs due to an addition of heterocromatin in the short arm. This condition is shared also by A. dembeensis and A. blicki. Due to the low number of individuals examined, the occurrence of a polymorphism in the X chromo- some cannot be excluded a priori for A. dembeensis and A. blicki. Only the comparison with the ‘true’ A. niloticus from terra typica, i.e. the Nile delta, would solve the systematics of these species (see MussEer and CARLETON 1993). If A. dembeensis was a geographic variant or a subspecies of A. niloticus, then the genus would be represented by an Egyptian-Ethiopian radiation (A. niloticus, ‘A. dembeensis’, A. abyssinicus, and A. blicki) and by a Central-Western African one (VOLOBOUEV et al. 1987, 1988), including the karyotypes described by CiviteLni et al. (1995), GRANJoN et al. (1992) and by VoLOBOUEV et al. (1988) as ‘A. centralis’ and ‘A. solatus’. Acknowledgements Field trips in Ethiopia during 1994-1995 were supported by grants from the Accademia Nazionale dei Lincei, C.N.R. (no. 93.4294.04), and Universita di Roma ‘La Sapienza’, project ‘Modelli dı evoluzione in ambiente tropicale’. The authorities of the Ethiopian Wildlife Organization and of the Bale Mountain National Park are deeply acknowledged. MARIA LuIsA PUCCETTI AZZAROLI (Firenze) and GoIToM RED- DA (Addis Ababa) provided valuable help during field trips and laboratory work at Addis Ababa Uni- versity. MAURIZIO ZuccotTI (London) provided an early draft of this manuscript and contributed with criticisms and linguistic revision. Zusammenfassung Cytogenetik der Gattung Arvicanthis (Rodentia, Muridae). 2. Die Chromosomen von drei Arten aus Äthiopien: A. abyssinicus, A. dembeensis und A. blicki Es werden die Karyotypen von drei äthiopischen Arvicanthis-Arten beschrieben, A. dembeensis aus dem Flachland, A. abyssinicus aus dem Hochland sowie der endemischen Gebirgsart A. blicki. Ihre di- ploiden und autosomalen FN-Zahlen (aFN) betragen 2n=62 (aFN = 62), 2n=62 (aFN =64) und 2n=48 (aFN=64). Mit Ausnahme eines großen submetacentrischen (no. 4) Chromosoms von A. abyssinicus und eines kleinen metacentrischen Chromosoms, sind alle Chromosomen von A. dembeensis und A. abyssinicus akrocentrisch, wohingegen der Karyotyp von A. blicki durch sieben Paare Rb metacentrischer Chromosomen gekennzeichnet ist. C- und G-Bandenmuster und Ag-NORs- Lagen werden zur Identifikation von chromosomalen Umordnungen herangezogen (Rb-Translokatio- nen, pericentrische Inversionen). Auf der Grundlage von Vergleichen zwischen diesen Karyotypen so- wie mit dem kürzlich beschriebenen Karyotyp von “A. niloticus” aus West-Afrika (Benin) werden Hypothesen zur Phylogenie entwickelt. References AFEWORK BEKELE; CAPANNA, E.; CoRTI, M.; MARCUS, L. F.: SCHLITTER, D. A. (1993): Systematics and geo- graphic variation of Ethiopian Arvicanthis (Rodentia, Muridae). J. Zool. (London) 230, 117-134. Chromosomes and phylogeny of Arvicanthis Sa BIcKHAM, J. W. (1979): Banded karyotypes of 11 species of American bats (genus Myotis). Cytologia 44, 789-797. CAPANNA, E.; AFEWORK BEKELE; CAPULA, M.; CASTIGLIA, R.; CIVITELLI, M. V.; CoDJA, J-CL.; Corrı, M.; FADDA, C. (1996): A multidisciplinary approach to the systematics of the genus Arvicanthis Lesson, 1842 (Rodentia, Murinae). Mammalia (in press). CAPANNA, E.; CIVITELLI, M. V. (1988): A cytotaxonomic approach of the systematics of Arvicanthis nilo- ticus (Desmarest, 1822) (Mammalia Rodentia). Trop. Zool. 1, 29-37. CAPULA, M.; CIVITELLI, M. V.; CorTI, M.; AFEWORK BEKELE; CAPANNA, E. (1996): Genetic divergence in the genus Arvicanthis (Rodentia, Muridae). Biochem. Syst. Ecol., in press. CIVITELLI, M. V.; CASTIGLIA, R.; CoDJIA, J.-CL.; CAPANNA,E. (1995) Cytogenetics of the genus Arvi- canthis (Rodentia, Muridae). 1. Arvicanthis niloticus from Republic of Benin (West Africa). Z. Säu- getierkunde 60, 215-225. CoRBETT, G. B.; HıLı, J. E. (1991): A world list of mammalian species. Oxford: Oxford Univ. Press. CorTI, M.; CIVITELLI, M. V.; AFEWORK BEKELE; CASTIGLIA, R.; CAPANNA, E. (1995): The chromosomes of three endemic rodents of the Bale mountains, South Ethiopia. Rend. Fis. Acc. Lincei s. 9, 6, 157- 164. De Wiınton, W. E. (1900): On the mammals obtained in Southern Abyssinia by Lord Lovat during an expedition from Berbera to the Blue Nile. Proc. Zool. Soc. London 1900, 79-84. FOoREY, P. L.; HUMPHRIES, C. J., KiTtcHing, I. L.; SCOTLAND, R. W.; SIEBERT, D. J.; WILLIAMS, M. D. (1992): Cladistics. A practical course in systematics. Oxford: Oxford Univ. Press. GRANIJON, L.; DUPLANTIER, J. L.; BRITTON DAviDIan, F. (1992): Karyotypic data on Rodents from Sene- gal. Isr. J. Zool. 38, 263-276. HoweLL, W. M.; BLAck, D. A. (1980): Controlled silver staining of nucleolus organizer regions with a protective colloidal developer: a 1-step method. Experientia 36, 1014-1015. Hsu, T. C.; PATTon, J. L. (1969): Bone marrow preparation for chromosome studies. In: Comparative mammalian cytogenetics. Ed. by K. BENIRSCHKE. Berlin: Springer-Verlag. Pp. 454-460. KınGoon, J. (1974): East African Mammals: an atlas of evolution in Africa. Vol. II, part B: hares and ro- dents. London: Academic Press. MATTHEY, R. (1965): Etudes de cytogenetique sur les Murinae africains appartenant aux genres Arvi- canthis, Praomys, Acomys et Mastomys (Rodentia). Mammalıa 29, 228-249. MUusSER, G. G.; CARLETON, M. D. (1993): Family Muridae. In: Mammal Species of the World. A taxo- nomic and geographic reference. Ed. by D. E. Wırson and D. A. W. REEDER. Washington, London: Smithsonian Institution Press. Pp. 576-578. ORLOoVv, V. N.; BASCKEVICH, M. I.; BULATOVA, N. SH. (1992): Chromosomal sets of rats of the genus Arvi- canthis from Ethiopia. Zool. Zh. 71, 103-112. SCHUBERT, I.; WoßBus, U. (1985): In situ hybridization confirms jumping nucleolus organizing regions in Allium. Chromosoma 92, 143-148. SEABRIGHT, M. A. (1971): A rapid banding technique for human chromosomes. Lancet 2, 971-972. SNEATH, P. H. A.; SOKAL, R. R. (1973): Numerical Taxonomy. The principles and practice of numerical | classification. San Francisco: W. H. Freeman and Co. VIEGAS-PEQUIGNOT, E.; DUTRILLAUX, B.; PROD’HOMME, M.; PETTER, F. (1983): Chromosomal phylogeny of Muridae: a study of 10 genera. Cytogenet. Cell. Genet. 35, 269-278. VOLOBOUEY, V.; VIEGAS-PEQUIGNOT, E.; PETTER, F.; DUTRILLAUX, B. (1987): Karyotypic diversity and taxo- nomic problems in the genus Arvicanthis (Rodentia, Muridae). Genetica 72, 147-150. VOLOBOUEN, V.; VIEGAS-PEQUIGNOT, E.; LOMBARD, M.; PETTER, F.; DUPLANTIER, J. M.; DUTRILLAUX, B. (1988): Chromosomal evidence for a polytypic structure of Arvicanthis niloticus (Rodentia, Muri- dae). Z. Zool. Syst. Evolut.-Forsch. 26, 276-285. YALDEN, D. W.; LARGEN, M. J.; Kock, D. (1976): Catalogue of mammals of Ethiopia. 2. Insectivora and Rodentia. Mon. Zool. It. (NS), suppl. 8, 1-118. YALDEN, D. W.; LARGEN, M. J. (1992): The endemic mammals of Ethiopia. Mammal Rev. 22, 115-150. Authors’ addresses: Prof. E. Capanna, Dr. R. CASTIGLIA, Prof. M. V. CivitELLı, Dr. M. Corri, Diparti- mento di Biologia Animale e dell’Uomo, Universita di Roma “La Sapienza”, via A. Borelli 50, 00161 Roma, Italy and Dr. AFEWORK BEKELE, Biology Department, University of Addis Ababa, P.O. Box 1176, Addis Ababa, Ethiopia. Z. Säugetierkunde 61 (1996) 352-364 ZEITSCHRIFT“ © 1996 Gustav Fischer, Jena SÄUG ETl ERKÜ NDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Size and structure of burrow systems of the fossorial rodent Ctenomys mendocinus in the piedmont of Mendoza province, Argentina By MARIA 1. Rosı, MöNICA I. CoNA, SILVIA PuiG, F. VIDELA, and V. G. RoıG Instituto Argentino de Investigaciones de Zonas Aridas, Unidad de Zoologta y Ecologla Animal, Consejo Nacional de Investigaciones Cientificas y Tecnolögicas, Mendoza, Argentina Receipt of Ms. 05. 07. 1995 Acceptance of Ms. 01. 06. 1996 Abstract The structure and size of burrow systems of Ctenomys mendocinus were analysed in the present study. These burrows showed a linear pattern, with a main axis from which branches and laterals forked of Burrow systems of males were larger than those of females, with a longer main tunnel and greater num- ber of branches. Lengthening of the main tunnel was achieved by adding new segments rather than by excavating longer segments. Both male and female burrows showed the same geometric configuration. There were no differences in the percentages estimated for main tunnel, branches and laterals. The an- gular variables (directional angle, branch angle and angle of ascent of laterals) did not show any differences between both sexes either. Burrows showed a constant heading along their path with mean directional angles close to 0°, even though in most systems left- and right-ward deviations from the main tunnel followed a random sequence. Branches originated at right angles to the main tunnel. C. mendocinus appears to minimize the energy cost of burrowing by increasing the angle of ascent of the lateral instead of its length as the main tunnel grows deeper. Home range, perimeter, and linearity were significantly higher in males. Introduction Fossorial mammals, just as subterranean mammals, are largely confined to underground life, but they venture a few centimeters outside their burrows for foraging (CONTRERAS and McNaß 1990). Both groups of herbivorous mammals construct complex tunnel sys- tems, which ensure them a relatively constant microclimate, protection from predators and access to food (NEvo 1979; REICHMAN and SMITH 1990). The subterranean environment is largely constant, and selection has led to the devel- opment of highly convergent structural features in the burrow systems of diverse mamma- lian taxa (NEevo 1979; HıckMAn 1990). Knowledge of burrow structure and its adaptıve features is a major aspect in understanding the unique evolution and adaptive biology in subterranean mammals (HıcKkMmAn 1990). The architecture of burrow systems of the species of Cienomys represents an almost unknown aspect of the ecology of this South American endemic fossorial rodent. Informa- tion available is either fragmentary or based on the study of only a few animals. Linear systems composed of a main axis and lateral tunnels have been reported for C. opimus, C. peruanus (PEARSON 1959) and C. pearsoni (AıtunA 1983). A similar pattern was de- scribed for C. mendocinus in an environment of low productivity (Puic et al. 1992). Line- ar systems associated with low resource availability have also been reported for Thomomys bottae (REICHMAN et al. 1982), Tachyoryctes splendens (Jarvıs and SALE 1971) and Spalax ehrenbergi (HETH 1989). Size and structure of burrows of Ctenomys mendocinus SD) This study aims at elucidating the most relevant structural features of burrow systems of C. mendocinus, and their possible differences between sexes. Material and methods Burrow systems of C. mendocinus were studied at Las Higueras (Mendoza, Argentina, 32°30°S and 68°55 W), located in the foothills of the Andes (1125 m elevation). The zone presents low hillocks and depressions furrowed by wadis. Soils are made up of lithosols in a matrix of strongly compacted clay open tunnels ---- home range +4+- plugged tunnels & food storage @ feeding ond resting chomber & mounds ——9 blind loterol © burrow openings —> nest chamber Fig. 1. Schematic representations of excavated burrow systems of Ctenomys mendocinus from the pied- mont of Mendoza (Argentina). Burrows of adult (A) male and (B) female. Tu = main tunnel, Br = branch, La = lateral. 354 Rosı, MARIA I., etal. and silt. Shrub communities of Larrea divaricata and L. cuneifolia prevail, alternating with dense ripar- ian vegetation (Roıc 1976). Complete burrow systems of 12 adults (6 males, 6 females) were excavated and mapped in Septem- ber and October 1991. After occupants were captured, Oneida Victor kill-traps were kept for 24 hours in the burrow openings, in order to record the number of occupants per system. Sex, weight, reproduc- tive condition and relative age were determined, based on criteria of Rosı et al. (1992). Length, depth, diameter and directionality of tunnel sections and chambers were measured for every system. Burrow openings and surface mounds were mapped. Burrow structural features were described with the same terminology used for other subterranean mammals (VLEcK 1981; REICHMAN et al. 1982; ANDERSEN 1988; HıckMANn 1990). The main tunnel was defined as the longest axis of the system (Fig. 1). When a fork was found, the longest tunnel was selected to follow the path of the main tunnel (REIcH- MMAN et al. 1982). Most selected tunnels showed the smallest angular deviation from the main axis (An- DERSEN 1988). Branching tunnels forking off of the main tunnel were classified into: a) laterals: straight tunnels that never branch, and end in a surface opening with or without a soil mound. Laterals also in- cluded all those tunnels ending in a “cul-de-sac”, hereafter referred to as blind laterals, b) branches: tun- nels made up of at least one segment and one lateral. Such tunnels sometimes have secondary branches. In burrow systems described for other fossorial rodents the term “segment” is used to define tunnel sections that extend between laterals (VLEcK 1981; REICHMAN et al. 1982; ANDERSEN 1988). In C. mendocinus, these tunnel sections were comparatively longer and showed great variability, owing to the small number of laterals recorded. For this reason, every fragment of such tunnel sections deter- mined by a change in heading was called segment, therefore our segment is not equivalent to the one defined by the above authors. One of the ends of the main tunnel was arbitrarily chosen as the initial segment of the system, since we did not know the sequence of segment construction. Deviation angles (directional angles) between consecutive segments of the main tunnel were measured from that initial segment, considering their left-ward (negative) or right-ward (positive) deviation. The mean of directional angles permitted us to know system directionality, according t0 ANDERSEN (1988). The branch angle was measured as the smal- lest angle between the initial segment of every branch and the main tunnel. The angle of ascent (VLECK 1981) was estimated only for those laterals that reached the surface (blind laterals were excluded). The area and perimeter in every system were measured with Tecktronix 4958 digitizer using the IN- CYTH-CRA'’s programme for calculating areas. For this purpose, an irregular polygon was drawn, by joining the ends of all laterals in a clockwise direction. For every angle between two segments from which no lateral came off, the bisector of the convex angle was drawn. The length of an estimated lat- eral was projected on this bisector. This lateral was obtained by averaging the perpendicular distances between the end of every actual lateral and the originating tunnel segment. Linearity of a burrow sys- tem was determined based on criteria of REICHMAN etal. (1982). Angular variables were analysed by ZAr’s (1984) circular statistics, using traditional statistics for the remaining variables. All variables were first tested for homogeneity of variance using the variance ratio test (Zar 1984). According to results, the Student’s t-test or Mann-Whitney U test were used to analyse for differences between sexes. An approximate testing procedure for differences between two proportions was used as applied by ZAR (1984). Linear regressions were performed between different metric variables estimating data ad- justment by using Pearson’s product-moment correlation coefficient r. Results Size and structure of subterranean systems All excavated systems (n= 12) corresponded to sexually active adult animals. No signifi- cant differences were recorded (t=1.69; P<0.10) in dry lens weight between males (n=5; X -322ms; SD=2.9) and females (n 6; X 29.0 ms SPS) AnEcontrase males (n=5; X = 219g; SD = 52.7) proved to be significantly heavier (U = 27; P = 0.025) than females (n= 6; X = 140.5 g; SD = 10.8). Wide ranges of body weigth were obtained within each sex (140 to 260 g for males, and 126 to 155 g for females). Every burrow was inhabited by only one specimen, except for a female’s burrow with five youngs in the nest. Size and structure of burrows of Ctenomys mendocinus 399 Total length of tunnels in a system ranged from 15 to 31 m for females, and from 23 to 77m for males. Significant differences were found (U =34; P=0.005) between male (n = 6, X = 50.5 m; SD = 21.0) and female burrows (n = 6; X = 22.4 m; SD = 6.8) concern- ing the average of the total length of tunnels. The diameter of feeding tunnels was largely constant, with mean values of 8.03 cm (n = 17; SD = 1.72) for females, and 8.5 cm (n=33; SD = 1.51) for males. Slight enlargements were only observed in the branching points of the main tunnel. Main tunnel: The main tunnel represented the greatest proportion of the total length of the system (males: 63.9%; females: 60.3%), there being no differences between sexes (Z = 0.68; P< 0.50). Total length and number of segments of the main tunnel were signifi- cantly greater in males (Tab. 1). Segment length varied within a wider range for males (0.20 to 3.20 m) than for females (0.20 to 1.60 m). Mean segment lengths were not signifi- cantly different either within or between sexes (Tab. 1). The main tunnel length showed a significant correlation with the number of segments (r = 0.92; df= 10; P<0.001), with their mean length (r=0.78; df=10; P<0.005), and with the occupant’s body weight (r = 0.69; df = 10; P< 0.02). In spite of the width of weight ranges within each sex, correla- tions with the main tunnel length were not significant. The main tunnel depth showed little variation both within and between burrows; moreover, no differences were found related to sex (Tab. 1). Greatest depths recorded did not exceed 0.40 m. The directional angles of consecutive segments of the main tunnel ranged from -128° to 132° (n = 350). Figure 2 shows the bimodal and symmetrical distribution of directional Table 1. Comparison between sexes of metric variables (X + SD) in burrow systems of €. mendocinus, where t and U are the statistics of the Student’s test and Mann-Whitney test, respectively; n is the num- ber of values and (k) the number of means used to estimate second-order means. Variable Value ofthe Significance Eee Teen REF ZERRETTE STALISUIE level MAIN TUNNEL Total length (m) 6 13.53 # 4.42 32-311 -3314.05 Number of segments 6 26.50 + 9.01 41.60 # 11.91 Mean segment length (m) 159 0.52 + 0.26 0.68 + 0.46 Mean depth (m) 112 0.30 + 0.09 0.27 # 0.06 BRANCHES Number of branches 20: 1022 4.00 #1.51 Mean length (m) jRSAEESIR 07, 2OSE30:03 Mean number of seg- ments 2.96 + 1.46 3.06 # 0.63 Mean segment length (m) 0.47 # 0.26 0.72 +0.61 LATERALS Mean number of laterals ISYS3E=ADD 22,83==110:26 Mean length (m) 0.34 + 0.16 0.40 + 0.24 Mean depth (m) 0.26 # 0.08 0.26 # 0.06 HOME RANGE Area (m’) RIES 13a 5300 p = 0.01 Perimeter (m) 1.07 EN) NENA93=13585 33:00 p = 0.005 Linearity 2.60 + 0.36 Il END Zah p<0.05 356 Rosı, MARIA I., etal. angles > and <0° in all twelve systems. Mean values for each system varied between -11° and 8° (n = 12) (Tab. 2), showing no significant differences either within or between sexes (Tab. 3). Moreover, when considering separately the positive and negative directional an- 60 40 20 / ‚af N A -110 -70 70 110 Fig. 2. Frequency distribution of directional angles (n = 350) in main tunnels of burrow systems of C. mendocinus. Table 2. Mean values (degree) and angular deviations (S) of the directional angle for main tunnel and branches. Animal Mean directional Directional angles of main tunnel (*) Mean directional angles of main 777 een ezianeiesjoisbemneh tunnel (+ and —) Positive Negative segment (**) FO +0 40 ) +O -K N mn m 7 4 1 4 4 7 2 6 4 Os Os Os Os Os Os +40 (*) According to left-ward (negative) and right-ward (positive) deviations from the main tunnel. (**) Estimated only for systems with branches composed of more than two segments. Size and structure of burrows of Ctenomys mendocinus 357 gles in every system, differences in number and in mean values were found to be minor (Tab. 2). By use of the one-sample test for the mean angles (Zar 1984) it was verified for every burrow that the mean directional angle did not deviate significantly from 0°. No significant deviation was recorded when analysing the mean values obtained for each sex (Tab. 3) at the 99% confidence level (length of mean vector r = 0.62 in males and 0.70 in females). Table 3. Comparisons within each sex (using Chi-square contingency test) and between sexes (using Watson’s U? test) of angular variables (X = mean, S = angular deviations) of burrow systems of C. mendocinus, where n is the number of values, (k) the number of means used to estimate second- order means and P level of significance. Directional angle Branch angle Angle of ascent of main tunnel of lateral Females .n(k) 152(6) 13 31(6) REIS OAl447, 88.1+18.1 40.4 + 18.3 Chi-square (df) 45.6(35) 23.4(18) 44.3(35) P=3025 p<0.25 P)=025 Males .n(k) 229(6) il 395) X+S 0.4 +49.8 85.8 + 19.9 Ile Chi-square (df) 46.4(45) 30.0(20) 25.6(24) p<0.50 p<0.10 p< 0.50 Comparison between sexes U?’(df) 0.095(6, 6) 0.068(13, 21) 0.103(6, 5) p < 0.20 p <0.50 p < 0.50 To test whether the sequence of construction of right-ward and left-ward segments was random, the two-tailed runs test (ZAr 1984) was used. In only three of the twelve sys- tems considered was the null hypothesis for randomness rejected, which indicates that in most systems segment deviations towards either side of the main tunnel did not alternate. Branches: in every system, one to six branches forked off from the main tunnel. The mean number of branches per system was significantly greater in males than in females (Tab. 1), whereas the mean number of branches per meter of main tunnel (males = 0.13; SD = 0.03; females = 0.20; SD = 0.14) did not differ significantly between sexes (U = 25; P>0.10). Branch mean length, as well as length and mean number of branch segments, showed no variations related to sex (Tab. 1). The proportion of branching tunnels in the total system length did not differ significantly (Z= 0.32; P>0.5) between male (X = 17.5%) and female burrows (X = 16.2%). Branch number, but not mean length, was significantly correlated with the main tunnel length (r = 0.79; df = 11; P< 0.002). The highest frequencies of branch angles ranged between 80° and 100° (Fig. 3). Means of these angles were similar for both sexes (Tab. 3) and did not depart significantly from 90° (length of mean vector r = 0.95 in females and 0.94 in males, one-sample test for mean angles, at 99% confidence level, ZAr 1984). Most branches composed of more than two segments showed a mean directional angle close to 0° (Tab. 2). Branch depth (X = 0.26 m; SD =0.06) did not differ significantly (t = 0.75; df = 22; P< 0.25) from the main tunnel depth (X = 0.28 m; SD = 0.07). 358 Rosı, MARIA l., etal. Frequency Du III SS 7 Un Fig. 3. Histogram of branch angles (n = 34) measured in all excavated systems of C. mendocinus. mr AAN, 85 Branch Angle Taegrses Laterals: 8 to 37 laterals per system were found. Their mean number was significantly lower than the mean number of segments composing the system, both in males (t=3.8; df= 10; P<0.001) and in females (U = 32; df = 6.6; P< 0.025). No significant differences between sexes were recorded in the mean number of laterals (Tab. 1) or in the proportion of laterals relative to total segments (females: 47%; males = 45%; Z = 0.45; P > 0.50). The length of main tunnel sections, comprised between two laterals reaching the surface, showed a wide range of variation (n= 71; X=3.7m; SD=3.9). This high variability was also observed when blind laterals were included (n = 151; X=1.8m; SD = 1.8). Lateral length varied considerably both within and between burrows (0.15 to 1.50 m), although few laterals exceeded 0.90 m in male (6.6%) and female (1%) systems. Lateral depth at the originating point was not significantly correlated with lateral length. Nevertheless, depth and angle of ascent of laterals were correlated (r = 0.46; df = 69; P< 0.001). These three variables showed no significant differences between sexes (Tabs. 1 and 3). Angles of ascent varied between 21° and 58° with their highest frequen- cies between 15° and 45°, and only 7% were higher than 60° (Fig. 4). Total length of laterals in proportion to total length of system tunnels was similar for both sexes (males: 19%; females: 23%; Z=1.10; P< 0.50). The proportion of laterals reaching the surface (54.7%, n = 127) was similar to that of blind laterals (45.3%). The latter were usually as deep as the main tunnel; 11.4% of them (n = 105) were plugged with shredded plant material mixed with loose soil. About 25% of laterals (30.4% of their overall length) were totally or partially plugged with slightly compacted soil. Chambers: eight systems showed oval chambers (1 to 3) that outsized the diameter of feeding tunnels. Considering their contents and location, they were classified into nest chambers, and feeding-resting chambers. No defecation chambers were found. Size and structure of burrows of Ctenomys mendocinus 859 18 15 u D Frequency 35 55 Angle of Ascent (degree) Fig. 4. Angles of ascent of laterals (n = 70) measured in eleven burrow systems of C. mendocinus. The chamber located at a greater depth than the remainder of the system with dry plant material covering its walls was regarded as the nest. Nests occupied an eccentric po- sition in most systems, never being situated in the distal ends of main tunnels or branches. No significant differences were recorded (t = 0.64; P>0.25) in nest size between males (n = 4; X = 0.32 m; SD = 0.005) and females (n = 4; X = 0.29 m; SD = 0.003). Depth of fe- male nests (n = 4; X = 0.64 m; SD = 0.04) was significantly greater (t= 3.16; P<0.01) than that of male nests (n = 4; X = 0.34 m; SD = 0.001). Feeding-resting chambers were smaller than nests and were located in the main tun- nel, at the same depth. They were usually empty, although fresh plant material was found in some of them. Storage of plant material was found into little blind laterals in most bur- TOWS. Mounds: the mean number of mounds recorded per burrow was 6.2 (n= 10; SD = 3.08). Most of them were roughly circular, with a mean diameter of 0.52 cm (n=21; SD = 0.14). Connection of these mounds with the burrow could not be found in 44% of them, in spite of their adjacent position to the main tunnel or blind laterals. Nearly all the mounds (95%) were old, given their flattened shape and high degree of soil compactness. Home range Area, perimeter and linearity were significantly higher in male than in female burrows (Tab. 1). Home range size varied between 7 and 16 m” in females, and between 14 and 99 m” in males. In both sexes, the perimeter of home ranges (n=12; X =53.1 m; SD = 32.5) departed significantly (U = 135; df = 12.12; P< 0.005) from the value expected for a circular system having the same area (n = 12; X = 17.0 m; SD = 7.8). 360 Rosı, MARIA I., etal. Area (r = 0.9; df= 11; P< 0.001), perimeter (r = 0.98; df = 11; P< 0.001), and Imeanity, (r = 0.69; df= 11; P nn _ a ? S £ =; © I « Barcelona EN V / Fa INS x ——— BE 600- 1712m BEE 1400 - 1600m 1200 - 1400m 1000 - 1200m 800 - 1000m 600 - 800m 400 - 600m 200 - 400m 0 - 200m = z Fig. 1. Topography and situation of the study area and the 25 localities sampled. Localities 1-4: Central Plateau (NW); Localities 5-7: Mediterranean and mid-European highlands (N); Localities 8-11: East- ern humid lowlands (NE); Localities 12-25: Mediterranean lowlands (S-SW). Small mammals in an isolated Mediterranean mountain 367 this method is useful for studying patterns of distribution of small mammal fauna and allows the com- parison with other studies based on the same methodology (ALEGRe et al. 1989; MORENO and BARBOSA 1992). Although the barn owl’s diet shows annual and seasonal changes (WEBSTER 1973; MaArTIı 1973), these fluctuations are not significant since samples were taken over many years and different seasons (MorENo and BarBosA 1992). We gathered information from studies published in the area about the feeding habits of the barn owl (see TORRE et al. 1996), and also included our own data. We collected in- formation from 25 different localities and 16,916 small mammals in the four zones described. Percen- tages of occurrence of the species were homogenized equalizing sample sizes for all localities, and we then obtained percentages respect to the total number of individuals of each species found in all of the localities. This new data set of percentages relates closely to the data set before transformation (r, values between 0.84 and 0.99, p < 0.0001). The geographical variables of the sampled localities (altitude, latitude, and longitude) were taken from 1:50,000 cartography, and the climatic data (temperature and rainfall) were obtained from maps of the Parc Natural del Montseny as well as from the Centre Meteorolögic de Barcelona. The geo- graphical and climatological information allowed us to arrange the localities into four groups (see study area) which were characterized by their small mammal fauna. Possible differences between them were tested with ANOVA. Before statistical analysis, percentages of species were arcsine transformed to equalize variances amongst them (CLark and Bunck 1991), and geoclimatic variables were log(X + 1) transformed (Zar 1984). A factorial analysis was performed with the geographic and climatic variables in order to obtain in- dependent orthogonal factors which could explain the observed variability. These factors, rotated by the varımax procedure, are interpreted as geographical and climatic gradients on which the distribution re- quirements of small mammals can be placed. The origin of the factorial space represents the average of the localities sampled in this study. The average situation (centroid) of every species was obtained fol- lowing CARRASCAL and TELLERIA (1990). The average and standard deviation of the localities in the four geoclimatic zones were also calculated. Diversity was calculated by Shannon-Wiener index (H’) and distribution breadth as e'". A matrix of simple correlations was obtained among all the variables. Stepwise regression models were performed to determine which environmental or geographical variable primarily affected the pre- sence of each species. Table 1. Number of small mammal individuals found in barn owl pellets corresponding to the four en- vironmental zones described (n = number of localities sampled in each zone). SPECIES Mediterranean Humid Central Highlands lowlands (n = 14) lowlands plateau (m) (n = 4) (n=4) Talpa europaea Sorex minutus Sorex araneus Neomys anomalus Suncus etruscus Crocidura russula Eliomys quercinus Apodemus sylvaticus Rattus rattus Rattus norvegicus Mus musculus Mus spretus Mus sp. Clethrionomys glareolus Microtus duodecimcostatus Microtus agrestis Arvicola sapidus m [0,0] nn \o &) (05) nn WW DD ES RVVDOUNOcoDOoo No prND 1 —I WW QHooouN0oO MH 0 Oo © D m 368 I. TORRE, J. L. TELLA and A. ARRIZABALAGA Results Information on 16 small-mammal species (sıx Insectivora and ten Rodentia) was recorded (Tab. 1). Table 2. Results of the factorial analysis performed with the geographic and climatic variables from 25 localities of the Montseny area and level of significance of the correlations between variables and fac- tors(C—P< 0.057790 EZ 0 N FH NN FACTOR 1 BACHORD VARIABLE Altitude 09057555 0.281 Rainfall Or 059055 Temperature -0.914**** 0.017 Latitude OST 0.149 0.260 0,950 Longitude Eigenvalue % Variance Ac. % Varlance FACTOR 2122 2°2) 2 7154 05005 ı 18 "NEE FACTOR 1 (59.1%) Fig. 2. Average situation of the species on the factorial space defined by factors one and two, and mean ands.d. of the four groups of localities sampled. Species: Te — Talpa europea; Sm — Sorex minutus; Sa — Sorex araneus; Na - Neomys anomalus; Se - Suncus etruscus, Cr - Crocidura russula; Eq - Eliomys quercinus; Ap- Apodemus sylvaticus; Rr — Rattus rattus; Rn — Rattus norvegicus; Ms — Mus spretus and Mus musculus; Cg - Clethrionomys glareolus,; Ma - Microtus agrestis; Md — Microtus duodecimcostatus; Ar - Arvicola sapidus Small mammals in an isolated Mediterranean mountain 369 The factorial analysis yielded a model with only two factors with eigenvalues greater than one, explaining together the 86.3% of variance (Tab. 2). The first factor correlated positively with altitude, rainfall and latitude, and negatively with temperature. The second factor was correlated with longitude and rainfall. The average situation of the species and the average situation and standard deviation of the four zones are represented in figure 2. These zones differed in the composition of the small mammal fauna (Tab. 3). A matrix of correlations between species enabled us to determine possible associations (Tab. 4). Spe- cies having significant correlations with factors showed clear patterns of distribution, their relative abundances increasing or decreasing along these multivariate gradients (Tab. 5). Species positively correlated with factor 1 (Sorex minutus, p< 0.01; Sorex araneus, p< 0.01; Apodemus sylvaticus, p < 0.05; Microtus agrestis, p < 0.001) showed an increase of their pre- sence towards colder, higher, northern, and rainy localities, whereas species with negative correlations (Crocidura russula, p<0.01; Rattus rattus, p<0.05; genus Mus, p< 0.001) Table 3. Mean and s.d. values of geography, climatology, diversity, richness, and small-mammal compo- sition (rare species excluded) of the four groups of localities considered. Differences between them tested with ANOVA and level of significance (see Tab. 2) VARIABLE GROUPI1 GROUP 2 GROUP 3 GROUP4 ANOVA (n= 14) (n=4) (n=4)) (m) FandP Altitude 240 609.5 9723 (85.7) (84.2) (132.7) Rainfall 710 1055 1024.6 (73.6) (40) (198.8) Temperature 14.9 1122 10.3 (0.5) (0.86) (0) Latitude 41°39 41°48' 41°47 Longitude 2°24' 31 ul DEE Sorex 0.14 15.78 ä 10.83 minutus (0.54) (27) ä (4.21) Suncus 4.13 3.95 4. 1.36 etruscus (2.85) 4. (2.36) Crocidura 4.45 i 1.9 russula (d132)) 5 (1.76) Apodemus 3419 8.56 sylvaticus (0.87) ! (3.36) Rattus 5.70 e : 0 rattus (5.65) h 5 (0) Rattus Ill ä 0.53 norvegicus (8.14) e (0.82) Mus spp. 5.24 3 (1.59) . («15)) Clethrionomys 3.86 2.96 glareolus (5.42) 3 (0. (4.14) Miecrotus DAS 1 2 duodecimcost. (8:97) h j (SA) Microtus 0 ; 2 21.4 agrestis (0) (12.54) DIVERSITY 1.94 E 1.43 (0.15) (0.89) RICHNESS 8.71 6 (1.54) (1.63) (diey3)) 370 I. TORRE, J. L. TELLA and A. ARRIZABALAGA Table 4. Correlation matrix among the small mammal species (+ positive correlation, — negative corre- lation) and level of significance (see Tab. 2); mnemonics as in Fig. 2. SPECIES RR RN MS MM MT Apodemus sylvaticus Rattus rattus Rattus norvegicus Mus spretus Mus musculus 2% a ka BR rk Mus total Kar Ka okakakak 1 akakakak Clethrionomys glareolus Microtus agrestis io M. duodecimcostatus showed an increase towards warmer, lower, southern and drier localities. Species positively correlated with factor 2 (Sorex minutus, p< 0.01; Apodemus sylvaticus, p< 0.01; Clethrio- nomys glareolus, p< 0.001) showed an increase in their abundances towards eastern and rainy localities, whereas species with negative correlations (Rattus rattus, p < 0.01; Microtus duodecimcostatus, p< 0.001) showed an increase towards western and drier localities. The average situation (centroids) of the species on the factorial space also revealed trends of distribution (Fig. 2). The species placed close to the origin of the factorial space can be considered as widespread. This conclusion arose from the relationship between dis- tribution breadth and the euclidean distance of each species to the origin of the space (E-059, pP 00yn= 5)! Four species displayed the greatest values of distribution breadth (>2): C. russula, A. sylvaticus, genus Mus and S. etruscus. The minimum values (<1.6) were recorded for rare species: S. araneus, E. quercinus, T. europaea and N. anomalus (Tab. 6). The stepwise regression analysis revealed the primary variable affecting the distribu- tion of each species and the species’ diversity (Tab. 6). The climatic variables (tempera- Table 5. Average situation of the small mammal species on both factors, correlation coefficients be- tween the abundances and factors, and level of significance of such correlations (see Tab. 2). SPECIES Talpa europaea Sorex minutus Sorex araneus Neomys anomalus Suncus etruscus Crocidura russula Eliomys quercinus Apodemus sylvaticus Rattus rattus Rattus norvegicus Mus spp. Clethrionomys glareolus Microtus agrestis M. duodecimcostatus Arvicola sapidus Insectivora Rodentia Diversity Small mammals in an isolated Mediterranean mountain 371 Table 6. First variable selected, t-value, level of significance (see Tab. 2) and percentage of variance ex- plained by the stepwise regression models performed with the small mammal species abundances as de- pendent variables and the geographic and climatic variables as independent ones (n = 25 except for M. spretus, M. musculus and diversity with n = 13). Distribution breadth (e'’) for the species is also shown. SPECIES VARIABLE t-VALUE % VAR BREADTH Talpa europaea Sorex minutus Rainfall 5,05+*+** Sorex araneus Rainfall 3.44** Neomys anomalus Suncus etruscus Temperature SD Crocidura russula Temperature 30. Eliomys quercinus Apodemus sylvaticus Rainfall A.18*** Rattus rattus Rainfall —4.26*** Rattus norvegicus Mus spretus Altitude EIGAE Mus musculus Temperature +4.94*** Mus total Temperature + ASEE* Clethrionomys glareolus Longitude 4.06 * Microtus agrestis Temperature _3,86*** M. duodecimcostatus Longitude 089 Arvicola sapidus INSECTIVORA Latitude Sa RODENTIA Temperature 2.46* DIVERSITY Rainfall —5,51*** ture and rainfall) explained the distribution of most of the species (eight species), whereas the geographic ones only explained the distribution of three species. Five species were independent of environmental variables. The number of species of Insectivora increased with latitude, while the number of spe- cies of Rodentia increased with temperature. Diversity was negatively affected by rainfall (Tab. 6) and altitude (Tab. 5). The position of the species in the space defined by factorial analysis and their correla- tions with factors, together with the stepwise regression analysis and the relationships among the species, allowed us to define three groups of species: 1. Species inhabiting preferentially Mediterranean zones: S. etruscus, C. russula, E. quercinus, R.rattus, R.norvegicus, M. musculus, M.spretus, A.sapidus, and M. duodecimcostatus. 2. Species inhabiting preferentially mid-European zones: S. minutus, S. araneus, M. agrestis, T. europaea, A. sylvaticus and C. glareolus. The first three species were more striet in distribution requirements, whereas the others displayed significant penetration into the Mediterranean region (Tab. 3). 3. Species with no clear preferences: N. anomalus. Discussion In general, Mediterranean communities have lower values of diversity and richness than mid-European ones (HERRERA 1974; IzA et al. 1985; GoNZALEZ and RoMAN 1988). How- ever, in the Montseny diversity was lower in the mid-European characterized localities. 372 I. TORRE, J. L. TELLA and A. ARRIZABALAGA This apparent contradiction can be explained by the higher elevation and the relative small extensions of the mid-European habitats: number of species and diversity decrease with altitude (Age 1982; DELIBES 1985; PATTERsoN et al. 1989) due to the island effect of mountains (MCÄARTHUR 1972; LomoLino et al. 1989; GoRrMAN 1991). These results contrast with the distribution patterns found in the Pyrenees, where diversity increases with alti- tude (MorENoO and BARBosA 1992). Climatic and geographic factors could be interacting with taxa under consideration, leading to opposing results (Owen 1990). The five mon- tane species found in the Pyrenees (Neomys fodiens, Sorex coronatus, Microtus nivalis, M. pyrenaicus, and M. arvalis, GıL et al. 1986; MorREno and BArBosA 1992) are absent in the nearby Montseny, probably due to its current isolation and few suitable areas to inhabit, which could lead to extinctions following the Pleistocene and the lack of further recolonizations (BROWN 1971; LomoLino et al. 1989). This fact could explain the higher di- versity ın highland areas of the Pyrenees. Other five mid-European species (Talpa euro- paea, Sorex minutus, S. araneus, Clethrionomys glareolus, and Microtus agrestis) are also present in the Montseny, where they are at the southern limit of their distribution in NE Iberia. The limited number of montane species in the Montseny could also explain why alti- tude only affects the distribution of one species, in spite of the important altitudinal range of the localities sampled (more than 1000 m), while altitude affected many species in higher Spanish mountains (DELIBES 1985; ALCANTARA 1989; MoRENoO and BARBosA 1992). In spite of the restricted latitudinal rank, the number of Insectivora species increased with latitude, following the general pattern found in larger mountain ranges (MORENO and BArBosA 1992). However, latitude did not affect the distribution of any particular species, while this variable affected the distribution of some species in the Pyrenees (MoRENO and BARBOoSA 1992). Longitude conditioned the distribution of M. duodecimcostatus and C. glareolus in op- posing ways, according to the Mediterranean requirements of the former (Lisois et al. 1983: GosALBEZ 1987; GONZALEZ and RoMAn 1988) and the Eurosiberian distribution of the latter (IzA et al. 1985; GosALBEZ 1987; GONZALEZ and RomMAn 1988). However, C. glareolus showed important penetrations into Mediterranean habitats in this area. The clımatic variables seemed to explain the distribution of the small mammals better than the geographic ones. Rainfall negatively conditioned the distribution of R. rattus and favoured S. minutus, S. araneus, and A. sylvaticus. The increase of $. minutus with rainfall has already been explained (IzA et al. 1985; GoSALBEZ 1987; GONZALEZ and ROMAN 1988; MorEno and BarBosA 1992), and reached the highest mean values of occurrence in the humid eastern region. S. araneus was primarily conditioned by rainfall (Spitz and SAINT- Girons 1969; GosSALBEZ 1987) but altitude also seemed to be important, preferring cold and rainy highlands. This pattern differs from that of the Pyrenees (GosALBEZ 1987), probably because suitable habitats for the species are restricted to high altitudes in the Montseny. Finally, A. sy/vaticus showed an increase of abundance from dry and temperate localities to rainy and cold ones. Our results agree with MorEno and BArBosA (1992) and Iza et al. (1985), but opposing patterns were also found (DELIBES 1985; GONZALEZ and RoMAn 1988). The generalist character of this species (ALCANTARA 1989; GoSALBEZ 1987) could explain these differences. Temperature positively affected the number of Rodentia species found in the area, as well as the distribution of four particular species (M. agrestis, S. estruscus, C. russula, and M. musculus). The former was affected negatively, thus being restricted to higher eleva- tions (coinciding with other mountains, DUENAS and Perıs, 1985; DELIBES 1985; MORENO and BArBosA 1992), while in some regions of northeastern Iberia it is even found at sea level (GosÄLBEZ 1987). S. etruscus, C. russula and the genus Mus tended to occupy low- land areas because of their elevated temperatures, agreeing with their supposed Mediter- ranean requirements (Spırz and SAINT-GIRONS 1969; LıBois et al. 1983; GosALBEZ 1987). Small mammals in an isolated Mediterranean mountain 378 Five species (T. europaea, N. anomalus, E. quercinus, R. norvegicus, and A. sapidus) did not show any relation with either geographic or with climatic variables. Although the presence of T. europaea was greater in the humid eastern region suggesting thus mid- European trends, the soil characteristics could greatly determine its distribution (GoSAL- BEZ 1987). In spite of the small number of records for E. quercinus, R. norvegicus, and A. sapidus, they showed a weak Mediterranean distribution. However, no geographic or climatic requirements were previously found (GosSALBEZ 1987; GONZALEZ and ROMAN 1988; MorENo and BaArBosA 1992). Finally, our results confirm that N. anomalus should be conditioned by other factors, such as the hydrographic network characteristics, rather than geographic or climatic ones (ToRRE and TELLA 1994). Acknowledgements P. J. CoRDERO provided information on two localities, A. VILLARROYA helped with the statistical proce- dures, C. GoORTAZAR translated the abstract into German, E. and N. Lawson and T. LEVERIDGE improved the English. Zusammenfassung Verbreitungsbedingende Umweltfaktoren für Kleinsäuger in einem isolierten mediterranen Gebirge Es wurde die Verbreitung der Kleinsäuger im Montseny-Gebirge (Catalonia, Nordwestspanien) unter- sucht. Es handelt sich um ein isoliertes mediterranes Gebirge, welches die südliche Grenze der mittel- europäischenFauna und Flora darstellt. Das Untersuchungsmaterial stammt aus 25 Stellen (von 119 bis 1140 m über dem Meeresspiegel) und umfaßt 16 916 Kleinsäuger (sechs Insectivora- und zehn Roden- tia-Arten), die aus der Analyse von Schleiereulengewöllen stammen. Das Vorkommen (relative Abun- danz) der verschiedenen Arten wurde mit Umweltfaktoren (Höhe, Breiten- und Längengrad, Regenfall und Temperatur) verglichen. Regenfall und Temperatur beeinflussen die Verbreitung der meisten Ar- ten, wobei Regenfall negativ mit der Diversität verbunden ist, welche in den tieferen Regionen höhere Werte zeigt. Diese Tendenz ist anders als in den Pyrenäen, der nahesten großen Gebirgskette. Vier ty- pische Arten aus höheren Lagen der Pyrenäen sind nicht im Montseny anzutreffen, möglicherweise auf- grund der isolierten Lage. Der Montseny bildet die südliche Grenze der Verbreitung von fünf Säugetierarten. Durch eine stufenweise Regression, factorielle Analyse und einer Korrelationsmatrix konnten drei Artengruppen unterschieden werden: Diejenigen die mediterrane Biotope bevorzugen (S. etruscus, C. russula, E. quercinus, R. rattus, R. norvegicus, M. musculus, M. spretus, M. duodecimco- status und A. sapidus), solche, die mitteleuropäische Biotope vorziehen (T. europaea, S. minutus, S. araneus, A. sylvaticus, C. glareolus und M. agrestis) und zuletzt eine einzige Art (Neomys anomalus), die keine klaren Vorzüge zeigt. References ABE,H. (1982): Ecological distribution and faunal structure of small mammals in central Nepal. Mam- malia 46, 477-503. ALCANTARA, M. (1989): Analisis de la distribuciön altitudinal de la fauna de micromamiferos de la Sier- ra de Guadarrama. Acta biol. mont. 9, 85-92. ÄLEGRE, J.; HERNÄNDEZ, A.; PURROY, F.; SANCHEZ, A. J. (1989): Distribuciön altitudinal y patrones de afi- nidad tröfica geogräfica de la lechuza comün Tyto alba en Leön. Ardeola 36, 41-54. AMORI, G.; CRISTALDI, M.; ConToLı, L. (1984): Sui rodentori (Gliridae, Arvicolidae, Muridae) dell’Italia peninsulare ed insulare in rapporto all’ambiente bioclimatico mediterraneo. Animalia 11, 217-269. BonD, W.; FERGUSON, M.; FORSYTH, G. (1980): Small mammals and habitat structure along altitudinal gradients in the Southern Cape Mountains. S. Afr. J. Zool. 15, 34-43. 374 I. TORRE, J. L. TELLA and A. ARRIZABALAGA Brown, J. H. (1971): Mammals on mountaintops: non-equilibrium insular biogeography. Am. Nat. 105, 467-478. CARRASCAL, L. M.; TELLERIA, J. L. (1990): Impacto de las repoblaciones de Pinus radiata sobre la avifau- na forestal del Norte de Espana. Ardeola 37, 247-266. CLARK, D. R.; Bunck, C. M. (1991): Trends in North American small mammals found in common barn- owl (Tyto alba) dietary studies. Can. J. Zool. 69, 3093-3102. De Borös, ©. (1983): Vegetaciö del Montseny. Diputaciö de Barcelona. DELIBES, J. (1985): Distribution and abundance of small mammals in a gradient of altitude. Acta Zool. Fennica 173, 53-56. DUuENAs, M. E.; Perıs, S.J. (1985): Anälisis de egagröpilas de 7yto alba en la Sierra de Gata (W de Espana). Alytes 3, 109-144. Dusser, R. D.; SHUGART, H. H. (1979): Niche pattern in a forest-floor small-mammal fauna. Ecology 60, 108-118. GiL, J.;, GONZALEZ, F.; Puig, D. (1986): Alimentaciö de l’öliba (7yro alba): Distribuciö dels mamifers In- sectivors Rosegadors al Ripolles. Butll. Centre d’Estudis del Ripolles 10, 22-33. GONZALEZ, J.; ROMAN, J. (1988): Atlas de los micromamiferos de la provincia de Burgos. Burgos: Ed. J. GONZALEZ and J. ROMAN. GORMAN, M.L. (1991). Ecologia insular. Barcelona: Ed. Vedra. GOSALBEZ, J. (1987): Insectivors i rosegadors de Catalunya. Barcelona: Ketres S.A. HERRERA, C. M. (1974): Trophic diversity of the Barn Owl Tyto alba in continental Europe. Ornis Scand. 5, 181-191. IzA, J. B.; CASTIEN, E.; MENDIOLA, 1.; PEMAn, E. (1985): Algunos aspectos de la ecologia de los microma- miferos del Pais Vasco. Munibe 37, 101-110. Lisois, R. M.; Fons, R.; SAINT-GIRONS, M.C. (1983): Le regime alimentaire de la Chouette Effraie Tyto alba dans les Pyrenees-Orientales. Etude des variations Ecogeographiques. Terre Vie 37, 187- DIE LoMOLINO, M. V.; BROwNn, J. H.; Davis, R. (1989): Island biogeography of montane forest mammals in the American southwest. Ecology 70, 180-194. MARTI, C. D. (1973): Ten years of Barn Owl prey data from a Colorado nest site. Wilson Bull. 85, 85- 86. MCARTHUR, R.H. (1972): Geographical ecology: Pattern in the distribution of species. New York: Har- per and Row. McCovy, E. D.; Connor, E. F. (1980): Latitudinal gradients in the species diversity of North American mammals. Evolution 34, 193-203. MESERVE, P.L. (1976): Habitat and resource utilization by rodents of a California coastal sage scrub community. J. Anim. Ecol. 45, 647-666. MIRALLES, J.; TERRADAS, J. (1986): El patrimoni biolögic del Montseny. catälegs de flora i fauna, 1. Ser- vei de Parcs Naturals. Barcelona: Diputaciö de Barcelona. MORENO, E.; BARBOSA, A. (1992): Distribution patterns of small mammals along gradients of latitude and altitude in Northern Spain. Z. Säugetierkunde 57, 169-175. Owen, J. G. (1990): Patterns of mamalian species richness in relation to temperature, productivity, and variance in elevation. J. Mammalogy 71, 1-13. PASCUAL, X.; MoNToRrı, A. (1983): Contribuciön al estudio de Rana temporaria L. (Amphibia, Ranidae) en Sta. FE del Montseny (Barcelona). I. Descripciön de la zona y estima de la poblaciön. Misc. Zool. 7, 109-115. 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. Rıvas-MARTINEZ, S. (1983): Pisos bioclimäticos de Espana. Lazaroa 5, 33-43. ROCAMORA, G. (1987): Biogeographie et ecologie de l’avifaune nicheuse des Massifs peri-mediterra- neens d’Europe occidental. Thesis. Ecole Nationale Superiere Agronomique de Montpellier. Rocovin, K. A.; SHENBROT, G. 1.; SUROV, A. V.; Iprıs, M. (1994): Spatial organization of a rodent commu- nity in the Western Rajasthan desert (India). Mammalıia 58, 243-260. SAINT-GIRONS, M. C.; SPITZ, F. (1966): A propos de l’etude des micromammiferes par l’analyse des pe- lotes des rapaces. Interet et limites de la methode. Terre Vie 1, 3-18. SPITZ, F.; SAINT-GIRoNS, M. C. (1969): Etude de la repartition en France de quelques Soricidae et Micro- tinae par l’analyse des pelotes de rejection de Tyto alba. Terre Vie 3, 246-268. TORRE, 1.; TELLA, J. L. (1994): Distribution of the Cabrera water shrew (Neomys anomalus) in Northeast- ern Spain. Z. Säugetierkunde 59, 282-288. Small mammals in an isolated Mediterranean mountain 375 TORRE, 1.; ARRIZABALAGA, A.; TELLA, J. L. (1996): Influencias geogräficas y climäticas en la distribuciön de los micromamiferos del Montseny. Actas de la III Trobada d’Estudiosos del Montseny. Barcelo- na: Ed. Diputaciö de Barcelona (in press). WEBSTER, J. A. (1973): Seasonal variations in mammal contents of Barn owl castings. Bird Study 20, 185-196. ZAR, J. H. (1984): Biostatistical analysis. Prentice Hall, Englewood Cliffs, New Jersey. Authors’ addresses: IGnNAcıo TORRE and ANTONI ARRIZABALAGA, Museu de Granollers-Ciencies Natu- rals, Avda. Francesc Macia 51, E-08400 Granollers (Barcelona) and Jos£ L. TELLA, Estaciön Biolögica de Donana (CSIC), Avda. M® Luisa s/n, Pabellön del Perü, E-41013 Sevilla, Spain 2 Z. Säugetierkunde 61 (1996) 376-378 ZEITSCH RIFTS FÜR © 1996 Gustav Fischer, Jena SÄUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY WISSENSCHAFTLICHE KURZMITTEILUNGEN Size selection of prey by otters, Lutra lutra L.: An experimental approach By M. TorrınG and H. KruuK Ecology and Evolution Group, Department of Zoology, University of Western Ontario, London, Ontario, Canada and Institute of Terrestrial Ecology, Banchory, Scotland, U.K. Receipt of Ms. 05. 02. 1996 Acceptance of Ms. 27. 07. 1996 Detailed observations of otter (Luftra lutra L.) foraging in freshwater habitats (stream, riv- ers and lochs) are extremely difficult. In these habitats, analysis of prey remains in the ot- ters’ spraints can provide some information on the sizes and species of fish taken (WEBB 1975; Conroy et al. 1993). Spraint analysis has revealed that the most common freshwater prey ıtems of otters in northern Britain and Ireland are eels (Anguilla anguilla) and sal- monid fish (Salmo sp.) (JENKINS et al. 1979; MurPHY and FAIRLEY 1985; ConRoy and JEN- KINS 1986; Kyne et al. 1989). In northeastern Scotland, where otters forage in both lochs and streams, spraint analysis indicates that salmonids are taken in relation to abundance, and no preference is shown for any particular size class (Kruuk et al. 1993). In contrast, larger eels are preferentially predated (CArss and Eıston 1996). As information on the foraging behaviour of otters is unavailable, it is still unclear whether the conclusions from spraint analysis reflect size preference by otters, or differences in the vulnerability to pre- dation by different size classes of fish (Kruuk 1995). The purpose of this study was to investigate the size selection of otters using the two most common freshwater prey species (eels and salmonids) in an experimental situation. By offering captive otters a choice between two fish of different size, we investigated the size selection of prey independent of the availability of sıze classes. Prey selection trıals were performed with 4 (2 male; 2 female) captive otters housed at the Institute of Terrestrial Ecology, Banchory, Scotland. Their normal diet of chicken and haddock was supplemented or replaced by experimental trials, using either brown trout (Salmo trutta) or eels (Anguilla anguilla). All animals were exposed to both eel and trout prior to the experiments. All fish were obtained by electrofishing at nearby lochs and riv- ers. Fork length (trout) or total length (eels) was measured to the nearest mm. Fish were then frozen at -20 °C. In preparation for experimental trials, the fish were removed from the freezer, and placed in a cold room (2-5 °C) to defrost for 24 hours. All trials were performed by offering a choice of two prey items of the same species. The fish were placed on a tray, such at the fish were visible and could be taken only from the front of the tray. We randomly selected the side of the tray where the larger fish was placed. Only one otter was present in the compound while trials were taking place. The tray was lowered into the compound, and the otter was allowed to choose a prey item from the tray. Binomial tests were used to determine whether the probability of first choosing the larger fish was significantly different from 50%. Significance was accepted at p = 0.05. Size selection of prey by otters, Lutra lutra L.: An experimental approach Sl Data were collected from 85 trials using eels and 221 trials using trout as the prey spe- cies. The difference in the number of trals involving eels and trout reflected the availabil- ity of each species of fish. Eels used for selection trials ranged in length from 152- 622 mm, with a range of length differences during trials of 4—304 mm. Lengths of trout ranged from 51-196 mm, and the length difference between trout ranged from 1-109 mm. These ranges corresponded with the sizes of prey taken by otters in the wild (trout, 19- 250 mm; eels, 130-620 mm; JENKINS et al. 1979; JENKINS AND HARPER 1980; KrUUK et al. 1933): The results of length selection trials involving eel are shown in table 1. The probability of the larger eel being taken was always greater than 50 % and statistically significant for 6 of the 8 length difference categories, resulting in a significant tendency to select the lar- ger eel. Overall, otters appear to show selection for the larger eel (larger eel taken in 72 of 85 trials; sign test significant at p< 0.001), even when the length difference between the two eels is small. The results of the trout length selection trials are shown in table 2. The larger of the two trout was selected more than 50% of the time in only 6 of the 10 categories, and this was significant in only one category. Thus, we conclude that there was no size selection for trout when examining absolute length difference. Overall, it appears that otters show no size selection for trout (larger trout taken in 117 of 221 trials; sign test non significant at D= ID): Table 1. Results of selection trials examining length differences between eels. Asterixes indicate length difference categories where the probability of taking the larger fish was significantly greater than 50% (Binomial test). Length Difference Number of Trials Larger Fish % Larger Category Taken Fish Taken 0-39 mm 40-79 mm 80-119 mm 120-159 mm 160-199 mm 200-239 mm 240-279 mm 280-319 mm Table 2. Results of selection trials examining length differences between trout. Asterixes indicate length difference categories where the probability of taking the larger fish was significantly greater than 50% (Binomial Test). Length Difference Number of Trials Larger Fish & Larger Category Taken Fish Taken 1-9 mm 10-19 mm 20-29 mm 30-39 mm 40-49 mm 50-59 mm 60-69 mm 70-79 mm 80-89 mm 90+mm 378 M. ToppinG and H. KruuUK The conclusion that in a captive, experimental situation, otters show a preference for lar- ger eels, but no sıze selection for trout is in agreement with observations in the field in northeast Scotland (Kruuk et al. 1993; Carss and ELston 1996). The pattern of prey selection observed may be at least partly explained by the varia- tion in lipid content of different sized fish, in that a proportionately greater lipid reward may be achieved by selecting the larger eel, even when the length difference is small (Decaniı et al. 1986). In contrast, even though a larger trout offers a larger meal, the lipid content does not vary greatly between different sizes of trout (ELLıoT 1976). This suggests that the nutritional value of prey may play a role in prey selection in captivity. However, the relative importance of the nutritional value of prey in the wild may be offset by other factors such as density of prey and competition. Acknowledgements We thank Dr. D. Carrs and M. McCann for collecting the fish used in these experiments. For advice and assistance during various stages of this study, we thank Prof. P. RAcev, Dr. J. Conroy, Dr. P. TAYLoRr, L. BALHARRY, and D. LITTLEJOHn. Thanks also to Dr. J. MıLLar, Dr. M. CHAseE, and the Small Mammal Discussion Group (U.W.O.) for critical review of this manuscript. References Casss, D. N.:; ELsTon, D. (1996): Errors associated with otter, Lutra lutra faecal analysis. II. Estimating prey sıze distribution from bones recovered in spraints. J. Zool. (London) 238, 319-332. CoNRoY, J.W. H.: JENKINS, D. (1986): Ecology of otters in Northern Scotland. VI. Diving times and hunt- ing success of otters (Lutra lutra) at Dinnet lochs, Aberdeenshire and in Yell Sound, Shetland. ]. Zool. (London) 209, 341-346. CoNROoY, J. W. H.; WATT, J.; WEBB, J. B.; Jones, A. (1993): A guide to the identification of prey remains in otter spraint. London: The Mammal Society, Occ. Publ. 16. DEGANI, G.; HAHAMU, H.; LEvANOoNn, D. (1986): The relationship of eel body size, lipid, protein, glucose, ash, moisture composition and enzyme activity. Comp. Biochem. Physiol. 84A, 739-745. ELLIOT, J. M. (1976): Body composition of Brown trout in relation to temperature and ration size. ]J. Anim. Ecol. 45, 273-289. JENKINS, D.; HARPER, R. J. (1980): Ecology of otters in Northern Scotland. II. Analyses af otter (Zutra lutra) and mink (Mustela vison) faeces from Deeside, N.E. Scotland in 1977-78. J. Anim. Ecol. 49, 737-754. JENKINS, D.; WALKER, J. G. K.; McCowan, D. (1979): Analysis of otter (Zutra lutra) faeces from Deeside, N.E. Scotland. J. Zool. (London) 187, 235-244. Kruuk, H. (1995): Wild otters: predation and populations, Oxford: Oxford University Press. KRUUK, H.; Carss, D. N.; ConRoy, J. W. H.; Duregin, L. (1993): Otter (Zutra lutra L.) numbers and fish productivity in rivers in N.E. Scotland. Symp. Zool. Soc. London 65, 171-191. Kyne, M. J.; SMAL, C. M.: FAIRLEY, J. S. (1989): The food of otters (Lutra lutra) in the Irish midlands and a comparison with that of mink (Mustela vison) in the same region. Proc. R. Ir. Acad. 89, 33-46. MURPHY, K. P; FAIRLEY, J. S. (1985): Food of otters (Zufrra lutra) on the South shore of Galway bay. Proc. R. Ir. Acad. 85B, 47-55. WEBB, J. B. (1975): Food of the otter (Lutra lutra) on the Somerset levels. J. Zool. (London) 177, 486-491. Authors’ addresses: Dr. MicHAEL G. ToppıinG, Ecology and Evolution Group, Department of Zoology, University of Western Ontario, London, Ontario, N6A 5B7, Canada: Dr. HAns KrUUk, Institute of Terrestrial Ecology, Hill of Brathens, Banchory, Aberdeenshire, Scotland, AB31 4BY, UK. Z. Säugetierkunde 61 (1996) 379-381 ZEITSCHRI FTSEF Omen — ______SÄUGETIERKUNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY Hard faeces reingestion in the Mountain hare Lepus timidus By H. HIRAKAWA Forestry and Forest Products Research Institute, Toyohira, Sapporo, Japan Receipt of Ms. 08. 03. 1996 Acceptance of Ms. 09. 07. 1996 Leporids are known to reingest their own faeces. This fact first became widely accepted thanks to the rediscovery and confirmation of Moror’s (1882) study by MADsen (1939) and TAyLor (1939). They showed that the domestic rabbit produces hard and soft types of faeces and reingests all soft faeces by taking them directly from the anus. The soft faeces, which had not been known until that time, was subsequently identified as being derived from fermented digesta in the caecum rich in vitamins and proteins (EDEn 1940; HUANG et al. 1954; Kurwıch et al. 1953; THACKER and BrANnDT 1955). Later, soft faeces were found in the stomachs or colons of many other leporid species (Hamırrton 1955; HEWSON 1962; LAynE 1958; LECHLEITNER 1957; SPENCER 1955; Watson and TAyLor 1955). As a re- sult, it became understood that caecotrophy (reingestion of soft faeces) was a normal phy- siological digestive process widely practiced in the leporids. Hard faeces are produced when the separation process at the proximal colon selectively diverts fluid and fine particles of digesta to the caecum (BJÖörnHAG 1972; CHEEKE 1987; EHR- LEIN et al. 1983; HÖRNICKE et al. 1984; PıcKARD and STEVENS 1972; RUCKEBUSCH and Hör- NICKE 1977). Hard faeces are hence composed mostly of poorly digestible large particles and had not been regarded as being normally reingested. However, the frequent reinges- tion of hard faeces was recently reported in the domestic rabbit (EBıno et al. 1993). More- over, the Japanese hare Lepus brachyurus was observed to reingest all the hard and soft faeces excreted during daytime rest (HıraKAwA 1994). Based on the various pieces of evi- dence, HırAKAwA (1994) suggested that the reingestion of all daytime hard faeces should be a normal practice in leporids. In this communication, the reingestion of hard faeces in the mountain hare L[. fimidus is reported as the evidence to support the prediction. The reingestion activities during the daytime rest were observed during the first two weeks of September 1995 in two mountain hares: a male kept in a large outdoor enclo- sure (32 mx28m) and a female in a small roofed pen (3mx4 m). Both were caught in February 1995 in Sapporo and had been fed with commercial pellets. Natural vegetation was also available to the male. They were accustomed to rest at some fixed sites on the ground (forms), which were monitored by video recording from dawn until dusk. The male hare entered the form at around sunrise (4:50), and left it at around sunset (18:00). The reingestion bouts for hard faeces (repeated faeces-taking actions each fol- lowed by mastication) occurred 3 times (n = 7, range: 2-5) within an hour after the hare en- tered the form (Fig. 1). Then, soft faeces reingestion (one faeces-taking action at a time without mastication) started at 5:40 (n=13, range: 5:21-5:57) and occurred 14 times (n = 10, range: 13-18) until early afternoon. Reingestion bouts for hard faeces began again at 13:38 (n = 12, range: 12:46-14:19) and occurred 4.6 times (n = 11, range: 3-7) until the hare left the form. Thus, the hare reingested all hard as well as soft faeces while resting in the form. Supplemental observation showed a similar reingestion pattern in the female. 380 H. HIRAKAWA J je) DOOOO00 oO SIOHSTEOHES 3:09 6:00 9:00 12:00 15:00 18:00 21:00 Time of day Fig. 1. Faeces reingestion activities of amale mountain hare during one day. Each circle represents a single faeces-taking action; solid circles are for hard-faeces-taking action followed by mastication; open circles for soft-faeces-taking action with no mastication. The thick horizontal bar indicates the time du- ring which the form was observed. The thin horizontal bar indicates the period when the hare was in the form. Circle piling indicates that faeces-taking actions occurred in a bout. The results of observation showed essentially the same pattern of reingestion of day- time hard and soft faeces in the Japanese hare /[. brachyurus (HıraAKAwA 1994). The only notable difference was that the morning reingestion bout for hard faeces occurred several times in the mountain hare whereas it occurred usually once in the Japanese hare. How- ever, both of these observations were conducted for a limited number of individuals, hence it is not known whether the difference is interspecific. The excretion of soft faeces in the mountain hare is reported to occur from the morn- ing to the early afternoon as in many other leporid species (FLux 1970; HEewson 1962; PEHRSoN 1983; PSHENNIKOV et al. 1988). This matches with the period in which the behav- iour of soft faeces reingestion was observed in this study. Hence, the generality of the pre- sent observation is supported. This is the second leporid species in which reingestion of all daytime hard faeces was observed. The temporal and behavioural patterns of reingestion in the mountain hare was also essentially the same with those in the Japanese hare. This strongly suggests that the reingestion of all hard faeces during daytime rest is a normal habit in leporids. However, to assure the conclusion, the study on the species other than Lepus is further required. References BIÖRNHAG, G. (1972): Separation and delay of contents in the rabbit colon. Swed. J. agric. Res. 2, 125- 136. CHEEKE, P.R. (1987): Rabbit feeding and nutrition. Orland: Academic Press. EBıno, K. Y.; SHUTOH, Y.; TAKAHASHI, W. (1993): Coprophagy in rabbits: autoingestion of hard feces. Exp. Anim. 42, 611-613. EDEN, A. (1940): Coprophagy in the rabbit: Origin of “night” faeces. Nature 145, 628-629. EHRLEIN, H.-J.; REICH, H.; SCHWINGER, M. (1983): Colonic motility and transit of digesta during hard and soft faeces formation in rabbits. J. Physiol. 338, 75-86. Frux, J. E.C. (1970): Life history of the Mountain hare (Lepus timidus scoticus) in north-east Scotland. J. Zool. (London) 161, 75-123. HaMmILToN, W. J. (1955): Coprophagy in the swamp rabbit. J. Mammalogy 36, 303-303. Hewson, R. (1962): Food and feeding habits of the mountain hare ZLepus timidus scoticus Hilzheimer. Proc. Zool. Soc. (London) 139, 515-526. HırakAawa, H. (1994): Coprophagy in the Japanese hare (Lepus brachyurus): reingestion of all the hard and soft feces during the daytime stay in the form. J. Zool. (London) 232, 447-456. HÖRNICKE, H.; RuOFF, G.; VoGT, B.; CLAuss, W.; EHRLEIN, H.-J. (1984): Phase relationship of the circa- dian rhythms of feed intake, caecal motility and production of soft and hard faeces in domestic rab- bits. Lab. Anim. 18, 169-172. Hard faeces reingestion in the Mountain hare ZLepus timidus 381 Huang, T. C.; ULRICH, H. E.; McCaAy, C. M. (1954): Antibiotics, growth, food utilization and the use of chromic oxide in studies with rabbits. J. Nutr. 54, 621-630. KuıwicH, R.; STRUGLIA, L.; PEARSON, P. B. (1953): The effect of coprophagy on the excretion of B vita- mins by the rabbit. J. Nutr. 49, 639-645. LAYNE, J. N. (1958): Notes on mammals of southern Illinois. Am. Midl. Nat. 60, 219-254. LECHLEITNER, R. R. (1957): Reingestion in the black-tailed jack rabbit. J. Mammalogy 38, 481-485. MADSEn, H. (1939): Does the rabbit chew the cud?? Nature 143, 981-982. MOoRroT, C. (1882): Des pelotes stomacales des leporides. Mem. Soc. Centr. Med. Vet. 12, 137-137, (not seen). PEHRson, Ä. (1983): Caecotrophy in caged mountain hares (Lepus timidus). J. Zool. (London) 199, 563- 574. PICKARD, D. W.; STEVENS, C. E. (1972): Digesta flow through the rabbit large intestine. Am. J. Physiol. 222, 1161-1166. PSHENNIKOV, A. E.; BoRISOV, Z. Z.; VASILIEV, 1. S. (1988): Coprophagy and its rhythms in mountain hare (Lepus timidus) of Central Yakutia. Zool. J. 67, 1357-1362. RUCKEBUSCH, Y.; HÖRNICKE, H. (1977): Motility of the rabbit’s colon and caecotrophy. Physiol. Behav. 18, 871-878. SPENCER, J. L. (1955): Reingestion in three American species of Lagomorph. Lloydia 18, 197-199. TAYLor, E. L. (1939): Does the rabbit chew the cud? [Appendix to Madsen’s note]. Nature 143, 982-983. THACKER, E. J.; BRANDT, C. S. (1955): Coprophagy in the rabbit. J. Nutr. 55, 375-387. WATSON, J. S.; TAYLoR, R. H. (1955): Reingestion in the hare Lepus europaeus Pal. Science 121, 314-314. Author’s address: Dr. HırOFUMI HıRAKAWA, Forestry and Forest Products Research Institute, Hitsuji- gaoka 7, Toyohira, Sapporo 062, Japan. Z. Säugetierkunde 61 (1996) 382-384 ZEITSCHRIFT < er 2 FÜR © 1996 Gustav Fischer, Jena SÄUG ETIERKÜNDE INTERNATIONAL JOURNAL OF MAMMALIAN BIOLOGY MITTEILUNGEN DER GESELLSCHAFT Protokoll über die Mitgliederversammlung der Deutschen Gesellschaft für Säu- getierkunde e. V. am 22. September 1996 im Hörsaal E 62 des Biologie-Zentrums der Universität Kiel. Der 1. Vorsitzende, Herr SCHMIDT, eröffnet die Versammlung um 16.30 Uhr. 1. Die Tagesordnung wird angenommen und nach Abstimmung bei fünf Enthaltungen um den Tagesordnungspunkt 8a (Austritt aus der Union deutscher biologischer Ge- sellschaften) erweitert. 2. Der Geschäftsführer, Herr ERKERT, verliest den Bericht über das Jahr 1995. Die 69. Jahrestagung der Gesellschaft fand auf Einladung von Herrn Kollegen Kraus FiscHER vom 24. bis 28. September 1995 in Göttingen statt. Schwerpunktthemen waren „Säugetiere in der Kulturlandschaft“, „Chronobiologie und Aktivitätsrhyth- men“ und „Fortpflanzungsbiologie der Säugetiere“. 48 Vorträge und 33 Posterprä- sentationen boten den über 200 Teilnehmern ein interessantes wissenschaftliches Programm auf hohem Niveau. Zum Gelingen beigetragen haben ferner die Führung durch das Deutsche Primatenzentrum durch Herrn Kollegen H.-J. KuHnn und die Ex- kursion in den Nationalpark Harz, die von Forstdirektor Dr. W. E. BARTH sachkundig geführt wurde. Herr ERKERT dankt dem Veranstalter, seinen Mitarbeitern und allen Organisatoren für die Ausrichtung der gelungenen Tagung. Herr van BREE, Amsterdam, wurde im Rahmen der Eröffnung der Tagung 1996 zum Ehrenmitglied ernannt. Der Frıtz-Frank-Förderpreis ging zu gleichen Teilen nach Österreich und in die Schweiz. Herrn Dr. WERNER HABERL aus Wien wurde er für seine Dissertation mit dem Thema „Zur Ökologie einheimischer Spitzmäuse (Soricoidae, Insectivora) und ihres Lebensraumes am Beispiel eines Waldviertler Feuchtbiotops und experimen- telle Bearbeitung ausgewählter ökologischer Fragestellungen“ verliehen. Herr Dr. JEAN-STEVE MEIA aus Neuchätel erhielt ihn für seine Dissertation mit dem Titel „Organisation sociale d’une population de renard (Vulpes vulpes) en milieu montag- nard“. Die Preise des Poster-Wettbewerbs wurden vergeben an: Frau J. Fıerz und Herrn W. SCHLUND, Tübingen, die Arbeitsgruppe von Herrn H. KLinGer, Braunschweig, so- wie an Frau H. FREYTAG-GRUNERT und Herrn R. SCHRÖPFER, Osnabrück. Die Zeitschrift für Säugetierkunde erschien in neuem Gewand in sechs Heften mit insgesamt 384 Seiten. Sie enthielt 34 größere Originalarbeiten, elf wissenschaftliche Kurzmitteilungen sowie einige Mitteilungen der Gesellschaft und Buchbesprechun- gen. Den beiden Schriftleitern und den aktiven Herausgebern wird ebenso gedankt wie Frau Dr. SCHLÜTER vom Fischer Verlag Jena, durch deren persönlichen Einsatz der Verlagswechsel völlig reibungslos verlaufen ist. Mitteilungen der Gesellschaft 383 Die Mitgliederzahl hatte sich bis Ende 1995 auf 602 verringert. Durch den Tod verlor die Gesellschaft folgende Mitglieder: Dr. PAuL BÜHLER, Gschwend Prof. Dr. RICHARD DEHM, München und Dozent Dr. WALTER PODUSCHKA, Wıen/Wultendorf. . Herr ERKERT erläutert den von Frau KÜHnRrıchH abgefaßten detaillierten Kassenbericht und dankt der Schatzmeisterin für ihre sorgfältige und umfangreiche Arbeit. . Die Herren BoHLKEN und SCHLIEMANN haben die Kontounterlagen der Gesellschaft in Hamburg geprüft und für korrekt befunden. Sie weisen in ihrem Bericht darauf hin, daß sparsamste Haushaltsführung auch weiterhin vonnöten sei. . Die Anträge auf Entlastung der Schatzmeisterin und des Vorstandes werden bei zwei Gegenstimmen und 15 Enthaltungen angenommen. . Die Herren BOHLKEN und SCHLIEMANN werden bei einer Gegenstimme und drei Ent- haltungen als Kassenprüfer für das Geschäftsjahr 1996 gewählt. Beide sind mit der Wahl einverstanden. . Der Vorstand schlägt vor, die Mitgliedsbeiträge für 1997 unverändert zu lassen. Dies wird bei einer Gegenstimme angenommen. . Die Mitgliederversammlung nimmt die Einladung der Herren FiscHEr und HALLE an, die 71. Jahrestagung vom 21.-25. September 1997 ın Jena abzuhalten. Als Schwer- punktthemen sind vorgesehen „Motorik bei Säugetieren“, „Methoden der Freiland- forschung“ sowie „Biologie der Musteliden“. Per Akklamation wird angenommen, daß Herr HUTTERER für 1998 nach Bonn eingela- den hat. 8a. Bei sechs Gegenstimmen und drei Enthaltungen faßt dıe Mitgliederversammlung an- 10. gesichts notwendiger Sparmaßnahmen den Beschluß, daß die Deutsche Gesellschaft für Säugetierkunde aus der „Union deutscher biologischer Gesellschaften“ austritt. Die Tagesordnungspunkte 9 und 11 werden aus Zeitmangel zugunsten der Neuwahl des Vorstandes zurückgestellt. Für den neuen Vorstand, der vom 1.1.1997 bis 31. 12. 2000 amtieren soll, schlägt Herr SchmiDT im Namen des alten Vorstandes Frau KÜHnrIcH und Frau WÖHRMANN-REPENNING sowie die Herren ERKERT, FRÄDRICH, HUTTERER und SCHRÖPFER vor. Schriftlich, einzeln und geheim werden gewählt als 1. Vorsitzender: Herr ERKERT 2. Vorsitzender: Herr HUTTERER 3. Vorsitzende: Frau WÖHRMANN-REPENNING Geschäftsführer: Herr SCHRÖPFER Schriftführer: Herr FRÄDRICH Schatzmeisterin: Frau KÜHNRICH Die Gewählten nehmen die Wahl an. Wegen der ausstehenden Tagesordnungspunkte wird die Mitgliederversammlung auf den 25. September 1996 vertagt. Der erste Teil endet um 19.15 Uhr . Herr ScHMipT eröffnet den 2. Teil am 25. September 1996 um 13.30 Uhr und stellt das von der Artenschutzkommission der DGS unter der Federführung von Herrn SCHRÖP- FER erarbeitete Positionspapier zur Diskussion. Die Mitglieder erhalten Gelegenheit, zu jedem einzelnen Absatz redaktionelle und inhaltliche Änderungswünsche vorzutra- gen. Die von Herrn SCHRÖPFER zu redigierende Endfassung ist dazu gedacht, dem ein- zelnen Mitglied der Gesellschaft ein Hilfsmittel in die Hand zu geben, das ihm im Hinblick auf Anträge bei Behörden usw. von Nutzen sein kann. 384 Mitteilungen der Gesellschaft 11. Es wird angeregt, bei künftigen Jahrestagungen darauf zu achten, den Begrüßungs- abend und den geselligen Abend für Studenten erschwinglich zu gestalten. Die Sitzung endet um 15.15 Uhr. Prof. Dr. U. SCHMIDT Prof. Dr. H. ERKERT Dr. HAns FRÄDRICH 1. Vorsitzender Geschäftsführer Schriftführer Buchbesprechung WUKETITS, F. M.: Die Entdeckung des Verhaltens. Eine Geschichte der Verhaltensforschung. Darm- stadt: Wissenschaftliche Buchgesellschaft 1995. 187 S.,; 23 Abb. DM 39,80. ISBN 3-534-12268-2 Im vorliegenden Werk möchte der Autor, gemäß Vorwort, eine Geschichte der Gedankenwelt der Ver- haltensforschung vorlegen. In einem einleitenden Abschnitt werden Definitionen der Begriffe „Ver- halten“ und „Verhaltensforschung“ Kritisch diskutiert und die Konzeption des Buches begründet. Im ersten Kapitel wird die „vorwissenschaftliche“ Phase in der Beobachtung des Tierverhaltens abgehan- delt. Insbesondere werden die Hintergründe für eine anthropomorphe Betrachtung von Tieren und de- ren Lebensäußerungen sowie der Unterschied zwischen einer beschreibenden Naturgeschichte und einer theoretisch begründeten Naturwissenschaft erörtert. Im zweiten Kapitel legt der Autor die Bedeu- tung des Evolutionsdenkens für die Verhaltensforschung dar. Die gedanklichen Wurzeln der Verglei- chenden Verhaltensforschung, wie etwa das Streben nach der Gewinnung neuer Merkmalssysteme zur Rekonstruktion von Stammbäumen, werden erläutert. Außerdem werden die Rolle der Gestaltwahr- nehmung als Quelle wissenschaftlicher Erkenntnis, Konzepte der Verhaltensphysiologie und die Frage nach einer Sonderstellung des Menschen beleuchtet. Das dritte Kapitel ist den „großen Kontroversen“, wie etwa der Auseinandersetzung zwischen Vertretern der Vererbungstheorie und der Milieutheorie, gewidmet. Die heuristische Bedeutung und die Problematik der Betrachtung von Tieren als „Reflex- maschinen“ (PawLow) sowie die Grundannahmen und Probleme des Behaviourismus (WATSON, SKIN- NER) und der Zweckpsychologie (UEXxKÜLL, McDouGALL) werden kritisch diskutiert. Danach beschreibt der Autor die Entstehung der Ethologie als Synthese aus dem Spannungsverhältnis zwischen einer jeweils zu einseitigen Betonung des Instinkts oder des Lernens durch verschiedene Schulen der Verhaltensforschung bzw. der Psychologie. Den Abschluß des Kapitels bilden Gedanken über die Evo- lution und die Modifikation des Verhaltens sowie Ausführungen zum Thema Aggression. Das vierte und letzte Kapitel beschließt das Buch mit einer Vorstellung von interdisziplinären Konzepten sowie neueren Forschungsansätzen, aber auch neuen Kontroversen im Bereich der Ethologie. Die Human- ethologie (EIBL-EIBESFELDT, HASSENSTEIN) und die Soziobiologie (WILSON, MAYNARD SMITH, WICKLER) werden in ihren Grundzügen charakterisiert, die Verhaltensökologie und -genetik kurz gestreift. Etho- logie wird hinsichtlich ihres Bezuges zu Erkenntnistheorie und Ethik diskutiert. Das vorliegende Buch gibt dem Leser einen gut gegliederten und interessant geschriebenen Abriß sowohl der Geschichte als auch vieler wichtiger Konzepte und Kontroversen der Verhaltensforschung. Etwas bedauerlich erscheint, daß die Darstellung des Fachgebietes in seiner Gesamtheit doch fast ausschließlich der im deutschsprachigen Raum, und hier vor allem der in der LorEnz’schen Schule verbreiteten Tradition verhaftet bleibt. Trotz der Betonung von Synthesen werden verschiedene neuere, für den evolutions- und populationsbiologisch interessierten Leser relevante Bereiche der Verhaltensbiologie eher ausge- klammert denn integriert. So tragen die sparsamen Ausführungen zur Verhaltensökologie und Verhal- tensgenetik den seit mehreren Jahren sehr umfangreichen Entwicklungen in diesen Disziplinen und deren Bezug zu anderen Sparten der Ethologie nur sehr ungenügend Rechnung. G. B. HaArTL, Kiel Instructions to Authors Submission and Acceptance of Manuscripts: Manuscripts for publication should be sent to the managing editor, Prof. Dr. D. Kruska, Institut für Haustierkunde, Christian-Albrechts-Universität, Olshausenstr. 40-60, D-24118 Kiel, Germany. 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 mail, and as soon as the peer reviews are received the authors will be informed concerning the decision for acceptance. All correspondence dealing with details of production should be sent to: Zeitschrift für Säugetierkunde, Gustav Fischer Verlag, Villengang 2, D-07745 Jena, Germany. Form of the Manuscript: Each manuscript should be submitted in triplicate, one original and two copies, in English or German in DIN A4 format either from a typewriter or legible computer print with a left-hand margin of 4 cm and true double spacing. The first page of the manuscript should contain the following information: title name(s) of author(s): full first name for female authors, first initials only for male authors department and university affiliation In case of long titles, a running title not exceeding 72 characters should be provided. 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. Names of authors cited in the text are to be underlined in pencil. Botanical and zoological generic and species names should be underlined in pencil with a wavy line, whereby the name of the scientist who first described the species is not underlined. 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. Content of the Manuscript: Manuscripts can be published as original investigations or short communications. a) Original investigations: In addition to the text, original investigations should include illustrations, tables and references, and must not exceed 30 typewritten, double-spaced 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. If written in German, an English title should precede the English abstract; if written in English, a German title should precede a German summary. b) Short communications: Short communications must not exceed 5 typewritten, double-spaced pages and do not require either an Abstract or Summary. If written in German, an English title should be given on page 1; if written in English, please give a German title on page 1. Short communications need not be headed with subtitles into In- troduction, Materials and methods, Results, and Discussion but should be organized according to this form. Ilustrations: Number and size of illustrations must be kept to an absolute minimum. Only well focussed and read- ily reproducible original prints showing optimal contrast can be accepted; colour plates cannot be processed. La- belling on figures should be uniform, 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 should 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 DIN A4 format sheets of paper. References: Each paper cited in the text must appear in alphabetical order at the end of the text. Only internation- ally available literature should be quoted. The names of the journals should be given in abbreviated form in conformity to international use. Examples a) From journals: KArxo, E. K. V; HANDLEY, C. ©. Jr. (1994): Evolution, biogeography, and description of a new species of Fruit-eat- ing bat, genus Artibeus Leach (1821), from Panamä. 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 und F. KrApp. Wiesbaden: Akad. Verlagsges. Vol. 2/1, 7-28. Printing: 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 under- taken 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 paper 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. 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). Vorbehalt aller Rechte Die in dieser Zeitschrift veröffentlichten Beiträge sind urheberrechtlich geschützt. Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von Abbildungen und Tabellen, der Funk- und Fernsehsendung, der Mikroverfilmung oder der Vervielfältigung auf anderen Wegen, blei- ben, auch bei nur auszugsweiser Verwertung vorbehalten. Das Vervielfältigen dieser Zeitschrift ist auch im Einzel- fall grundsätzlich verboten. Die Herstellung einer Kopie eines einzelnen Beitrages oder von Teilen eines Beitrages ist auch im Einzelfall nur in den Grenzen der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundes- republik Deutschland vom 9. September 1965 in der Fassung vom 24. Juni 1985 zulässig. Sie ist grundsätzlich ver- gütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des Urheberrechtsgesetzes. Gesetzlich zulässige Vervielfältigungen sind mit einem Vermerk über die Quelle und den Vervielfältiger zu kennzeichnen. Copyright-masthead-statement (valid for users in the USA): The appearance of the code at the bottom of the first page of an article in this journal indicates the copyright owner’s consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condi- tion however, that the copier pay the stated percopy fee through the Copyright Clearance Center, Inc., 21 Con- gress Street, Salem. MA 01970, USA, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective, or for resale. For copying from back volumes of this Journal see “Permissions to Photo-Copy: Publisher’s Fee List” of the CCC. Zeitschrift für Säugetierkunde (International Journal of Mammalian Biology) Publisher: Gustav Fischer Verlag Jena GmbH, Postfach 1005 37, D-07705 Jena; Telefon (03641) 62 6436, Telefax (03641) 62 65 00. Type setting, printing, binding: Druckhaus Köthen GmbH, Friedrichstr. 11/12, D-06366 Köthen Advertising sales: Gustav Fischer Verlag Jena GmbH, Anzeigenverwaltung, Frau A. Schütz, Postfach 100537, D-07705 Jena; Telefon (03641) 626430, Telefax (03641) 626500. The price list from October Ist, 1994 is effec- tive at present. Distribution and subscription: Gustav Fischer Verlag Jena GmbH, Zeitschriftenvertrieb, Frau B. Dressler, Post- fach 1005 37, D-07705 Jena; Telefon (03641) 62 64 44, Telefax (0 3641) 62 65.00. For the USA and Canada only: VCH Publishers, Inc., Distribution Center, 303 N.W. 12th Avenue, Deerfield Beach, FL 33442-1788; Telefon (305) 42855 66, Telefax (305) 4288201. Terms of delivery (1996): 1 volume consisting of 6 issues. Subscription rate (1996): Price per volume (DM/ÖS/SFr): 398,-/2 945,-/ 382,50 (plus postage). Single issue price (DM/ÖSI/SFr): 80,- /592,-/77,- (plus postage). We accept the following credit cards: Visa / Eurocard / Mastercard / American Express (Please specify Card No. and Expiry Date) Subscription information: Please, send your order to your scientific bookshop, to your hitherto dealer or directly to our publishing house. If not terminated the subscription will be effective until recalled. If no discontinuance is given until October, 31 the delivery of the journal will be continued. Banking connections: Postgiroamt Leipzig, Konto-Nr. 149249-903, BLZ 86010090; Deutsche Bank, Konto- Nr. 3 907 656, BLZ 820 700 00; Commerzbank AG, Filiale Jena, Konto-Nr. 2581 122, BLZ 820 400 00. 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. Printed in Germany © Gustav Fischer Verlag Jena GmbH 1996 Y Re, | | mung: Er un nee nn en we re er a ee De me my are. were TE a N ETEr 2 & wegeir% BT EEE ren mmnenen anenyny- rei ur x ap ann rem ae RD a ne TE RESTE TE Ne - im EEE LE GBR STE ar a I Eng ine Ka nern RENT RR ehren rare 5 CH PERF er nen ums en hir By, Bern ern ver RT FAGT, m warez Bere menge RR. 2 mangeenn amrasr Dunsan BLCLTaRIE Ba Bu BES ES EEE EEE rung. Zar un Pre Petra Ken en R NER PH FEE? Ferner ann eg ar ya rare warm wir A N nr at I a een map am me lan Dev Aue ae un amann ne RENTEN RER ne hr a ANZ En nt, RR ang a are ran de ankrun nen ESP e HELFER RRÄRNENE: ” Er ee Dr SE LE et enge ARSTER) ir & Keane m De ann u ale PS ER KL WET ET RENS SB FRE REIT ARE a RER er % 2223 \ EERRIR A n heak ELPITEFER TEEN ET, REN AA “ NE EEE ER EEE f war hegen ar LADE EE Ä nern N e ae EEE iy, LEE ara g Art meh ara Een kaykıı VE LE Er RÄT Kies; Veisanar I ee RER, Kaleanını Was annshegz EHE TEIN AT RE We ANARNRN RE want, x Ag AP FURTEVA FERIEN TATUERN e : N N ran i ae ha nenngzt a ER a BE RE EN EIER, ANA NEN RL? \ LERVEr ER REF, ANSAGE ihr De : ’ De RER ö. TER ET ea rang hr re [1207 VO OT ZEN Pal Bed Rn a nee Buer ae engine era het NEN m Bone, nenn ! ih, DR EL a isn PB Se Fee er Bu nem len En R aaayaae ARE ER ae rn, u en BEN ER are) ee rar ns) nr en. ws dermna nen ea 7 ALTE ENTER DEEP EN ETT: Fr nihadg aa Dr 2