MSN d ah, (Suk = Ma WMA yay ’ AS Baa aD, DAA ST 40 Im As ann EIA 4 4 Dan Aas i als Dom! AI ITCR Y 1 0) 4d ACER Te AI hr Y : SMS rd o. Se Yo A a a: ts ” se 2 ) : Pa hay DU A ha TIA 5 2 D as 5 N a sa, : E Fe Tara 4 2 aes vedada A, j ; A i y & un , . . , o 4 oh ai A ; ur va, > ae, “$ Anteater e Phare Caner rey fl ‘os Pre Fr ars anne Set TEO des No tier OA ELLA Nada : AS AS NIG ; Peg sets = t E: A Ira AO “se Mee) Mune ev. E EN Be BONNER ZOOLOGISCHE BEITRAGE Herausgegeben vom Zoologischen Forschungsinstitut und Museum Alexander Koenig Bonn Schriftleitung Rainer Hutterer Band 47, Hefte 1—4, 1997/98 ISSN 0006-7172 Redaktionelle Mitarbeit: H. von Issendorff Vom Band 47, 1997/98, erschienen Hefte 1—2 (p. 1—192) 7. Oktober 1997 Hefte 3—4 (p. 193—448) 30. September 1998 Zoologisches Forschungsinstitut und Museum Alexander Koenig Adenauerallee 160, 53113 Bonn, Germany Druck: JFeCARTHAUS, Bonn ISSN 0006—7172 Inhalt des 47. Bandes Allspach, A.; siehe Taiti Arita, Y.; siehe Gorbunov Bóhme, W.: A note on the gender of the genus Podarcis (Sauria: Lacertidae)..... Bosch, J.; siehe De la Riva Bouton,; siehe Seehausen Catzeflis, F. M.; siehe Jenkins Chen, S.-L.; siehe Matsui Claussen, C.: Die europäischen Arten der Cheilosia alpina-Gruppe (Diptera, Syrphi- TEE) 2 550 0.0 0.0.0.0.0 00 oe Le a ee A Contadini, L.; siehe Galleni Dégallier, N.: Coleoptera Histeridae Hetaeriinae: description de nouveaux taxons, desiemationzdeslectotypeszet MOLES! CaxONOMIGues nn. een. DelaRiva,I., R. Marquez & J. Bosch: Description of the advertisement calls of some South American Hylidae (Amphibia: Anura): taxonomic and methodological SITSSTWENER. sd ll NEN OA AOS Disney, R. H. L.: Post-eclosion heterochrony in the maturation of the adult females oraemitopaloustily (Diptera, PRoridae).. iio. dao oe es Disney, R. H. L.: Recognition of a sibling species of the Australian Eutermiphora abdoninalsaltea (Diptera: Phoridae) Mier. ae a a o na oo Fahr, J.: Die Sandlaufkáfer (Coleoptera: Cicindelidae) des Comoé-Nationalparks, Elfenbeinküste: Faunistik, Zoogeographie und Okologie ..............ooo.o.o.o.. Ferrara, F.; siehe Taiti Frynta, D.; siehe Vohralík Binschrerl2@occinellidae (Coleoptera) aus Rwanda....................22.2...... Galleni, L., R. Stanyon, L. Contadini & A. Tellini: Biogeographical and karyological data of the Microtus savii group (Rodentia, Arvicolidae) in Italy .. Goodman,S. M., P. D. Jenkins & O. Langrand: Exceptional records of Micro- gale species (Insectivora: Tenrecidae) in vertebrate food remalnS ............... Gorbunov, O. G. & Y. Arita: Review of the genus Paradoxecia Hampson, 1919 WepidopteraSesiidas MAI E E a da ile Haacke, W.: Systematics and biogeography of the southern African scincine genus Nana cortas iiReptillas SCIMNCIAL) as o nen en Hershkovitz, P.: Report on some sigmodontine rodents collected in southeastern Brazil with description of a new genus and six new Species ...........o..o.oo... Jenkins, P, M. Ruedi € F. M. Catzeflis: A biochemical and morphological investigation of Suncus dayi (Dobson, 1888) and discussion of relationships in Suneus, Crocidura, and Sylvisorex (Insectivora: Soricidae).................... Jenkins, P. D.; siehe Goodman Jordan, T.: Tersilochus curvator Horstmann und Tersilochus sp. n. (Ichneumonidae, Tersilochinae), neue Parasitoiden der an Birken minierenden Trugmotten (Lepidoptera, ERICA Eh ee RR RA Se EIERN SE Karner, M.: A new species of Notolaemus Lefkovitch from Rwanda (Coleoptera: BTemophlocidae)e Be Bar an LE en ee Kirejtshuk, A. G.: On the evolution of anthophilous Nitidulidae (Coleoptera) in FroprealgandssubiTOpIcalatesionse. an ee aan: Langrand, O.; siehe Goodman Lue, K.Y..; siehe Matsui Márquez, R.; siehe De la Riva Matsui, M., S.-L. Chen € K.Y. Lue: Advertisement call characteristics of a Tai- Naneserorcenttrec iO wm NNACOPNOLUS PIS INAMU nen Melber, A. &H. Pohl: Erster Nachweis einer Strepsipterenparasitierung bei Wanzen ineNirteleuropar(inseeta, Strepsipteraret Heteroptera) a. e Naijt, J.; siehe Weiner Naumann CAVE Zu Gedenkentan Hans Kumerloeve o... oo 187 381 345 175 77 87 321 13 2 135 39 139 193 ZN 411 95 111 165 69 189 Pohl, H.; siehe Melber Ruedi, M.; siehe Jenkins Seehausen, O. & N. Bouton: The community of rock-dwelling cichlids in Lake Vietolla. a ae a A NN Shilenkov, V. G.: New and little known Nebrina (Epinebriola) from the eastern Nepal Himalayas: (Coleoptera; Carabidae) =. sae eee eee Sofianidou, T. S.; siehe Vohralik Spatenka, K.: Neue Glasflügler (Lepidoptera, Sesiidae) aus dem Pamir und dem Hindukusch».. 00 sie cid a Sis RRS BES) OO ee ee eee Spatenka, K.: Neue Glasflügler-Arten und Unterarten aus Europa und der Türkei (Sesiidae, Lepidoptera). 2.0 0... 0 ee eee ee cee Stanyon, R.; siehe Galleni Taiti, S., F Ferrara & A. Allspach: On three species of Periscyphis Gerstaecker, 1873 from Kenya, Sudan and Oman (Crustacea: Isopoda: Oniscidea) .......... Tanasevitch, A. V.: Gorbothorax n. gen., a new linyphiid spider genus from the Nepal Himalayas (Arachnida, Araneae, Linyphiidae)]........................ Tanasevitch, A. V.: New Oedothorax Bertkau, 1883, from Nepal (Arachnida, Araneas,. Linyphiidae)) A ao On eee Tellini, A.; siehe Galleni Tilbury, C.: Two new chameleons (Sauria: Chamaeleonidae) from isolated Afro- montane forests in Sudan and Ethiopia ©...) 2 eee eee Vohralik, V., T.S. Sofianidou & D. Frynta: Reproduction in Mus macedonicus (Mammalia: Rodentia) in the Balkansz. .. .. sac cerca eee eee Weiner, W. & J. Naijt: Collembola Poduromorpha from the Magallanes Province (Chile): o... 0 aa re a 301 313 43 283 Verzeichnis der neuen Taxa im 47. Band Mammalia: Muridae EN OAOREMyStaSENnS SPIHIErShkovitzr ne anne san 220 IBrueenattersomiusne Sg. Hershkovitz sau... ooo eee ee be eee ea 22 Ianlcepattersonius.albinasus.n. sp: Hershkovitz2...2. 222.2... o 55-440" 235 Brucepattersontus eniserufescensn. sp. Hershkoyitz2....2.2.......2.....2...... 233 Bileenattersonius,ieniventris n. sp. Hershkovitz? ooo 232 IBrüecepattersonius. Soricinus'n. sp. HershkovitZ ooo. os 232 Oxaimyeienus:) caparaoem. sp. Hershkovitz o o 244 Sauria: Scincidae ivphlacontias punctatissimus brainein. ssp: Haacke......................... 150 Mynhlaeontiasırüudebeckin. sp. Haack: 20.0 2. oe Bo ee ec cee ee he eee eee 155 Sauria: Chamaeleonidae Chamaeleo (Irioceros) balebicornutus n. sp. Tilbury..:..........2............ 294 ChamacleoWrioceros) controstratum mW. sp. Elbury a mn.n.nn..n... 297 Lepidoptera: Sesiidae BEIMOCClIONA OISIail Ss SPALEMKA G2 65.5 os E ee we es ee 36 Bemocciaenaratecina ms sp: Spatenka ooo eo ns eee ce ee eee tee eens 33 Benibeciaspashtunan> Sp) Spatenka ae hehe ce ee he 31 Bembecia pavicevici dobrovskyi n. ssp. Spatenka.:....................-+-... 48 IBembecianpetersemin. sp. SpatenKa's........ 4052005 - ss. senses A E 46 IBembeeiansyze]oVi Kkappadocican. ssp. Spatenkat a o 50 @namaespheeiankauttı n.ısp. Spatenkan 0. me. es ee o e 53 CGhanaespnecia mudjahida Ww. sp. Spatenka ooo 38 Chamaesphecia taurica n. sp. Spatenka .............. e N AP AS 55 @namaesphecia weidenhofjerin. sp. Spatenka ao oo oe 39 Divenashapecia intermeaia m. sp: Spatenka oca co 52 OT AUOXCEIUMUCICNSIS Me Sp) Gorbunov &-Atita oo 65 Paradoxecia viemnamica n. sp. Gorbunov & Arita...2...-.....--.++-+-+--++-:: 62 Syzansphheeiasherain: Sp. Spatenka 2. .4....c- nose o cece soe: cate ee at 51 Synanthedon andrenaformis tenuicingulata n. ssp. Spatenka .................. 44 Synanthedon stomoxiformis riefenstahli n. ssp. Spatenka.................--.- 43 Diptera: Phoridae ENERO WatSOn! M. Sp: Disney a oes nest ene nee. 90 Diptera: Syrphidae Chenosiagsubpictipennis 1. sp. Claussen.............. te e e ee ee ene eee 405 Coleoptera: Carabidae INebzasschawalleri ns sps Shilenkov. das 314 INebnwantonejelaensisuns sp. Shilenkov ee... neuen. 316 Coleoptera: Coccinellidae Alnidentaammiehleim. sp. Kürsehi ..)5 2 «2% 20. 0 a Ae nee ren na 26 Ianılachnakapiceoeulatan. sp. Fürsch a 24 Naipilachnazcarapacolan. sp. Eürsch. a o dl eel ee so 26 Iipiluchnanconsperzatain. sp. Rürsch. 4. 552-2 5-5 + se ose ee Sees setae es 24 INephUSA(SIGIS) rugulipennis a. sp. Eürsch 250058566 ee os) ees: 18 INCOSE (SIGIS)) Drevipulosus Tl. Sp) Fürsch. 32 9-45-54 2 msn. 18 ORANGES ACIIS Ms O ee 20 PSCUAOSGVIMRUS brunneusın. sp. Rürsch 54.452. 2m .eoae ener eee dees os oes 20 SCOLOSGVINNUS EI QOnIDUOSUS Ne SP: HUISChe semanas sae ees] eee see es. 22 SCOLOSGVININUS MAXIMUS Ml. A ey. een 22. Seumnusı(2ullus)zwandensisin. Sp. Rürsch a... scannen: 16 TEISTARRO SEU a ne ko hc ae ee nen. 20 Coleoptera: Histeridae ESTOS CIN ADE SA NS OR 352 Colonides collesicianensis m= ssp) Déeallicm nen. 357 Neocolonides ng. Désalliera 23. A A A eee 362 MNeoeolonides.hoy.denin. sp: Desi 364 Troglosternus neoeciionis nsps Desallerzzrr. re EL eee 272 Coleoptera: Laemophloeidae Notolaemus wagneri.n. Sp: Karel... ds iota REL Cee 95 Coleoptera: Nitidulidae Brouthina ne os Kirejtshuk. 00 tt E ores OO eC Ee 129 Brounthina aequalisin..sp., Kit eee 129 Claplothorax n.-sg. Kirejtshuk ... . «chs osteo DCL 128 Mandipetes n.: se: Kirejtshuk: 0 3. ug Sr Se 122 Planennipolus.n. se. Kirejtshuk 0... 80 eee Cee 128 Propetes (Propetes) aquılus n. sp Kirejtshuk 2 oe 118 Propetes (Mandipetes) intritus n. sp. Kirejtshuk............. STAs x RN 122 'Propetes (Mandipetes) longipes MAS eee 122 Propetes (Propetes) seychellensis Mp reg eee 120 Uroearpolus.n:.58. Klrejtshuks „ur: o 128 Arachnida: Linyphiidae Gorbothorax 1. 2£.-Tanasevich..... its 0 oe 421 Gorbothorax comatus mn. sp. Tanasevich...2.2. 2 eee 225 Gorbothorax conicusin. sp. Tanasevichi.. 2.12... es 226 Gorbothorax setifer Mn. spy Tanasevich..... 2... re PIDA) Oedothorax angelus ns sp. Tanasevich, 2.2 2 eee 433 Oedothorax clypeelum n. sp. Tanasevich ....... A RE wal ops ay is cad oe 436 Oedothorax coronatus n. sp. Tanasevic 2. 2 ee 432 Oedothorax Jalciferusn. sp. Tanasevich .. 3... 55 ou eee eee 440 Oedothorax malearmatus n. sp. Tanasevich . 222.2 ne 440 Oedothorax modestusn. sp. Tanasevich.. .... 2. oe eee 438 Oedothorax savigniformis n. sp. Tanasevich?. 4..6. 4... ee 439 Oedothorax sexoculorum n. sp. Tanasevich 2... 2.2.2. 2.202 oe eee ee 434 Oedothorax simplicithorax a. sp. Lanas ca 1.0 06 eee 437 Oedothorax tholusus'n. sp. VWanasevich. ... eo 435 Oedothorax unsibbus n. sp. Vanasevich 22 hee 428 Isopoda: Oniscidea Periscyphis albomarginatus n. sp. Taiti, Ferrara & Allspach................... 6 Periscyphis sudanensis n. sp. Taiti, Ferrara & Allspaci oe 1 Collembola: Hypogastruridae Triacanthella vogeli n. sp. Weiner LENIN 102 Collembola: Brachystomellidae Brachystomellides navarinensis n. sp. Weiner & Naijt .............o.o.oooo.o.o.o.. 105 Herausgegeben vom Zoologischen Forschungsinstitut und Museum Alexander Koenig Bonn ae 7 Ri ie ES dl ie IR LIBRA Band 47, Hefte 172, 197 ISSN 00067172 Bonner zoologische Beiträge Herausgeber und Verlag: Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn. Schriftleitung: Dr. Rainer Hutterer, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, D-53113 Bonn. Erscheinungsweise und Bezug: Die Zeitschrift erscheint vierteljährlich zum Preis von 22,50 DM je Heft bzw. 90,— DM je Band einschließlich Versandspesen. Korrespondenz über Abonne- ments, Kauf oder Tausch bitte richten an die Bibliothek, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, D-53113 Bonn. Manuskripte und Besprechungsexemplare: Einsendung an die Schriftleitung erbeten. Manu- skriptrichtlinien sind am Ende dieses Heftes abgedruckt. Druck: JFe CARTHAUS, Bonn. On three species of Periscyphis Gerstaecker, 1873 from Kenya, Sudan and Oman (Crustacea: Isopoda: Oniscidea) Stefano Taiti, Franco Ferrara & Andreas Allspach Abstract. Periscyphis brunneus Budde-Lund, 1912 from Kenya is redescribed on the basis of the type-material and new material examined, and the lectotype is designated. Two new species, P sudanensis from Sudan and P albomarginatus from Oman, are described. All species presently included in Periscyphis are listed with their distributions. Key words. Crustacea, Isopoda, Oniscidea, Eubelidae, Periscyphis, new species, Kenya, Sudan, Oman. The genus Periscyphis Gerstaecker, 1873 belongs to the family Eubelidae which represents the most important fraction of the Oniscidean fauna of tropical Africa and the Arabian Peninsula. The genus is characterized by the pereonite 1 with the posterior corners entire (1.e. without a schisma) and no ventral lobes or teeth, antennal flagellum of two articles, inner branch of the maxillule with two penicils, exopods of pleopods 1 and 2 with monospiracular covered lungs. All species of Periscyphis populate arid or semiarid areas. After Omer-Cooper’s (1926) comprehensive revision, many new species have been ascribed to Periscyphis (Arcangeli 1929; 1934; 1940; Barnard 1940; 1941; Ferrara 1972; 1973; 1974; Schmölzer 1974; Ferrara & Taiti 1982; 1986; 1988; 1996; Taiti & Ferrara 1989; 1991; Erhard € Schmalfuss in press) so that to date the genus has included with certainty 36 species distributed in northeastern Africa, Israel and the Arabian Peninsula. In this contribution one poorly-known species, Periscyphis brunneus Budde-Lund, 1912 from Kenya, is redescribed and two new species from Sudan and Oman are described, bringing the number of species in the genus to 38 (Table 1). Abbreviations: MZUF — Museo Zoologico “La Specola” dell’Universita, Firenze; NRM — Naturhistoriska Riksmuseet, Stockholm; ONHM — Oman Natural History Museum, Muscat; SMF — Senckenberg-Museum, Frankfurt/Main. Periscyphis brunneus Budde-Lund, 1912, Figs 1, 2 Periscyphis brunnea Budde-Lund, 1912 (in Lónnberg & Budde-Lund, 1912): 7, Fig. 5. Periscyphis brunneus; Omer-Cooper, 1926: 398; Paulian de Félice, 1945: 342; Ferrara & Taiti, 1979: 155% Lectotype (here designated): 9, Kenya, Njoro, N of Guaso Nyiri river about 1°N lat., under the sheaths of palm leaf-stalks, leg. E. Lónnberg, 17. II. 1911, NRM 2648. Paralectotype: 1 Q, same data as lectotype. Material examined: 4 70,2 Q 9, Kenya, Kora National Reserve, 00 °02’S—38 °37’E, leg. R. C. Davis, 2. II. 1984, MZUF 1579. Redescription: Maximum dimensions: &, 6x2.5 mm; 9, 6.5x 2.8 mm. Brown with the usual pale muscle spots; uropods pale (red in vivo?). Dorsum smooth with inconspicuous 2 Se Tantızerral D Fig. 1: Periscyphis brunneus Budde-Lund, 1912, or from Kora National Reserve: A, adult spe- cimen, lateral view; B, cephalon, dorsal view; C, cephalon, frontal view; D, right side of pereo- nites 1 and 2, dorsal view; E, pleonite 5, telson and uropods. pointed scale-spines. Eye with about 25 ommatidia. Cephalon with broadly rounded lateral lobes; no frontal margin; interocular line visible only at sides; profrons slightly bulbous in the middle. Pereonite 1 with lateral margin slightly thickened; a shallow sulcus arcuatus only in the anterior third; posterior corners rounded, distinctly protruding backwards. Pereonites 2—7 with posterior corners progressively less protruding backwards. Telson with distal part narrow, triangular with narrowly rounded apex. Antenna short, reaching rear margin of pereonite 2 when pushed back; flagellar articles subequal in length. Uropodal protopod with convex outer margin and posterior margin indented, a small glandular area near postero- lateral corner; small exopod distinctly protruding backwards. Male: Pereopods 1—3 carpus with a brush of pointed spines. Pereopod 7 without special modifications; ischium with straight sternal margin. Pleopod 1 exopod with quadrangular medial part, distal margin with a row of strong spines; endopod with a thickset distal part New species of Periscyphis 3 Fig. 2: Periscyphis brunneus Budde-Lund, 1912, & from Kora National Reserve: A, antenna; B, pereopod 1; C, pereopod 7; D, pleopod 1; E, pleopod 2. and bilobed apex, outer lobe with some small rounded scales, inner lobe triangular with two spines. Pleopod 2 as in Fig. 2E. Remarks: Since no illustrations, except for the apex of the maxilliped, were provided by Budde-Lund (1912) in the description of Periscyphis brunneus, this species is redescribed here and its diagnostic characters are illustrated. It is characterized by the cephalon with no frontal line and distinct interocular line on the vertex, the pereonite 1 with the lateral margin slightly thickened and the sulcus arcuatus present only in the anterior third, and by the male pleopod 1 endopod with a bilobed apex. The colour pattern (brown with pale uropods) is also a useful distinguishing character. 4 S. Taiti et al. Fig. 3: Periscyphis sudanensis n. sp., 2 from Erkowit: A, adult specimen, lateral view. © from Erkowit: B, cephalon, dorsal view; C, cephalon, frontal view; D, left side of pereonites 1 and 2, dorsal view; E, pleonite 5, telson and uropods; F, antenna. Periscyphis sudanensis n. sp., Figs 3, 4 Periscyphis trivialis; Vandel 1964: 735, Fig. 111. Holotype: ©, Sudan, village Erkowit at Mount Erkowit, ca. 50 km SW of Suakin, 1200 m, leg. A. Allspach, 2. III. 1987, SMF 22681. Paratypes:2 00,1 Q, 2 juvs, same data as holotype, SMF 22682; 1 or, 1 9, same data, MZUF 4778. Additional material examined: 3 oo, 4 99, Sudan, Sanganeb Island, N of Port Sudan, at lighthouse, leg. V. Neumann, 31. III. 1991, SMF 22683; 1 ©, 1 Q, same data, MZUF 4779. New species of Periscyphis 5) Fig. 4: Periscyphis sudanensis n. sp., © from Erkowit: A, pereopod 1; B, pereopod 7; C, pleo- pod 1; D, pleopod 2. © from Sanganeb Is.: E, pleopod 1 exopod. Description: Maximum dimensions: ©, 8x 3.2 mm; 9, 9x3.5 mm. Light brown with posterior margin of pereonites, pleon and telson darker; epimera of pereonites and pleonites and uropods pale. Dorsum smooth with inconspicuous pointed scale-spines. Eye with about 21 ommatidia. Cephalon similar to the preceding species, i. e. with broadly rounded lateral lobes, no frontal margin, interocular line visible only at sides and profrons slightly bulbous in the middle. Pereonite 1 with lateral margin slightly thickened; a shallow sulcus arcuatus only in the anterior third; posterior corners rounded, protruding backwards. Pereonites 2—7 with posterior corners slightly protruding backwards. Telson with distal part triangular and rounded apex, reaching tips of uropodal protopods. Antenna reaching middle of pereonite 2 6 S. Taiti et al. when pushed back; first flagellar article almost 3/2 as long as second. Uropodal protopod large, with convex outer margin, a large indentation on posterior margin and a wide transverse elandular area near postero-lateral corner; small exopod distinctly protruding backwards. Male: Pereopods 1—3 carpus with a brush of recurved and pointed spines. Pereopods 2—4 ischium and merus with some small verrucae on sternal margin. Pereopod 7 ischium with a large depression on distal rostral surface, sternal margin almost straight; merus proximally with a small rounded tubercle near tergal margin and a group of four spines near sternal margin. Pleopod 1 exopod subtriangular, with broadly rounded posterior margin equipped with short spines; endopod with distal part pointed and bent outward. In specimens from Sanganeb Island the pleopod 1 exopod is subtrapezoidal (cf. Fig. 4E). Pleopod 2 as in Fig. 4D. Etymology: The name refers to Sudan where the specimens were collected. Remarks: In the structure of the cephalon (without frontal margin) and pereonite 1 (with the sulcus arcuatus present only in the anterior third of the segment) Periscyphis sudanensis is similar to P cavernicolus, P. abyssinicus, P jannonei, P. verhoeffi, P somaliensis and P. brunneus. It is readily distinguished from all these species by the male pereopod 7 merus with a distinct tubercle on the rostral surface and a basal group of spines near the sternal margin, and by the different shape of the telson. Moreover, it differs from P abyssinicus and P jannonei in the absence of a rectangular protrusion on the sternal margin of the male pereopod 7 ischium; from P verhoeffi in the absence of a large lamellar lobe on the tergal margin of the male pereopod 7 merus; from P brunneus, P cavernicolus and P somaliensis in the different shape of the apex of the male pleopod 1 endopod. In the presence of a meral tubercle on the male pereopod 7, PR sudanensis corresponds to P vittatus and P insularis. It is easily distinguished from the former by the presence of the sulcus arcuatus (absent in P vittatus) on pereonite 1 and the telson reaching the posterior margin of the uropodal protopod (it is distinctly shorter in P vittatus) and from the latter essentially by the structure of the telson and uropods (compare Fig. 3E and Fig. 14 in Ferrara & Taiti 1989). In the specimens from Sanganeb Is. the male pleopod 1 exopod is morphologically different from that of the specimens from Erkowit (compare Fig. 4C and Fig. 4E), while all the other characters correspond. With the material at our disposal we are not able to say whether this difference is within the intraspecific variability or, instead, it is indicative of distinct taxa. For the time being, we prefer to include the specimens from Sanganeb Is. in the new species without considering them as paratypes. Vandel (1964: 735) recorded P trivialis from Suakin, Sudan. However, from his illustration of the male pleopod 1 exopod, it is clear that the specimens examined by Vandel do not belong to P trivialis but to P sudanensis, having affinity particularly with the specimens from Sanganeb Is. (compare Fig. 111 in Vandel 1964 and Fig. 4B). Periscyphis albomarginatus n. sp., Figs 5, 6 Holotype: ©, Oman, Masirah Island, near S tip, 20°12’ N—58°40’E, 1100 m, under shrub on mountain side, leg. M. D. Gallagher, 5. VI. 1991, ONHM. Paratype: ©, same data as holotype, MZUF 4780. Description: Maximum dimensions: 14 x 8 mm. Grey-brown with posterior part of cephalon, epimera of pereon and pleon, tip of telson and uropods pale; pereonites with a large pale spot in the middle and two pale paramedian spots in correspondence to the muscle insertions. Dorsum smooth with inconspicuous pointed scale-spines. Eye with 23—24 ommatidia. Cephalon with broadly rounded lateral lobes; frontal margin interrupted in the middle; interocular line reaching middle of the eyes; profrons bulbous in the middle. Pereonite 1 with a conspicuous lateral thickening and a deep narrow sulcus arcuatus in the anterior half; posterior margin slightly concave at sides; posterior corners broadly rounded. Telson with distal part triangular and rounded apex, slightly surpassing tips of uropods. Antenna long with first flagellar article almost twice as long as first. Uropodal protopod subquadrangular with posterior margin indented and small glandular area near exopod insertion; exopods minute. Pereopods 1—3 with a brush of recurved pointed spines on carpus and, progressively New species of Periscyphis 7) Fig. 5: Periscyphis albomarginatus n. sp., ©: A, adult specimen, lateral view; B, cephalon, dorsal view; C, cephalon, frontal view; D, right side of pereonites 1 and 2, dorsal view; E, pleonite 5, telson and uropods; F, antenna. . more reduced, on merus. Pereopods 1—4 ischium and merus with some small verrucae on sternal margin. Pereopod 7 ischium with concave sternal margin and distally with a large depression on rostral surface; carpus flattened, recurved and slightly enlarged, with a distal lobe on caudal surface equipped with three strong spines. Pleopod 1 exopod with rounded distal margin; endopod with pointed apex slightly bent outwards. Pleopod 2 as in Fig. 6D. Etymology: L. albus = white + marginatus = having a margin. The name refers to the characteristic colour pattern (in alcohol) with a large pale stripe all around the body. 8 SeTaltızeralk Fig. 6: Periscyphis albomarginatus n. sp., ©: A, pereopod 1; B, pereopod 7; C, pleopod 1; D, pleopod 2. Remarks: The new species belongs to the granai-group, characterized by the cephalon with large rounded lateral lobes which continue in a frontal margin, entire or interrupted in the middle, large quadrangular uropodal protopod, and male pereopod 7 with sexual modifica- tions on the ischium and carpus. This group includes P granai, P latissimus, P. arabicus, P barnardi, P omanensis, P buettikeri, P insularis and P minor. In the frontal margin inter- rupted in the middle, P albomarginatus corresponds to P buettikeri and P insularis (all the other species in the group have a continuous frontal margin) and it is readily distinguished from both by the telson with rounded, instead of acute, apex surpassing the posterior margin New species of Periscyphis 9 of the uropodal protopod. It also differs from P buettikeri in the sulcus arcuatus anteriorly not bent inwards, the regularly convex lateral margin of pereonite 1 (it has a distinct depression in P buettikeri) and the carpus of the male pereopod 7 with the tergal margin much less convex; and from P insularis in the lack of a tubercle on the merus of the male pereopod 7. Table 1: Periscyphis species* and their distributions. 1) Periscyphis abyssinicus Ferrara, 1972 Ethiopia 2) Periscyphis albescens (Budde-Lund, 1885) Egypt and Sudan 3) Periscyphis albomarginatus n. sp. Oman 4) Periscyphis albus Erhard & Schmalfuss, in press Israel and Egypt 5) Periscyphis arabicus Barnard, 1941 Saudi Arabia and Yemen 6) Periscyphis barnardi Ferrara & Taiti, 1986 Yemen 7) Periscyphis besi Barnard, 1941 Yemen 8) Periscyphis brunneus Budde-Lund, 1912 Kenya 9) Periscyphis buettikeri Ferrara & Taiti, 1986 Saudi Arabia 10) Periscyphis cavernicolus Omer-Cooper, 1926 Ethiopia 11) Periscyphis civilis Budde-Lund, 1908 Somalia and Kenya Periscyphis convexus (Budde-Lund, 1885) Egypt, Sudan, Djibouti and Tanzania (?) 13) Periscyphis felix Taiti & Ferrara, 1989 Saudi Arabia 14) Periscyphis granai Arcangeli, 1929 Eritrea 15) Periscyphis insularis Ferrara & Taiti, 1988 Oman 16) Periscyphis jannonei Arcangeli, 1940 Sudan and Ethiopia 17) Periscyphis lanzai Ferrara, 1973 Somalia 18) Periscyphis latissimus Omer-Cooper, 1926 Eritrea 19) Periscyphis laticarpus Taiti & Ferrara, 1989 Saudi Arabia and Kuwait 20) Periscyphis libycus Arcangeli, 1934 Libya 21) Periscyphis limbatus Omer-Cooper, 1926 Kenya 22) Periscyphis merolobatus Ferrara & Taiti, 1982 Eritrea 23) Periscyphis minor Ferrara & Taiti, 1996 Yemen 24) Periscyphis nigricans Omer-Cooper, 1926 Ethiopia 25) Periscyphis omanensis Taiti & Ferrara, 1991 Oman 26) Periscyphis pulcher Budde-Lund, 1898 Tanzania 27) Periscyphis rubroantennatus Ferrara, 1974 Somalia 28) Periscyphis ruficauda Budde-Lund, 1908 Somalia and Kenya 29) Periscyphis somaliensis Ferrara, 1973 Somalia 30) Periscyphis strouhali Arcangeli, 1929 Eritrea 31) Periscyphis subtransversus Omer-Cooper, 1926 Egypt (?) 32) Periscyphis sudanensis n. sp. Sudan 33) Periscyphis trivialis Gerstaecker, 1873 Ethiopia, Somalia, Kenya and Tanzania 34) Periscyphis tschadensis Schmölzer, 1974 Chad 35) Periscyphis undulatus Omer-Cooper, 1926 Ethiopia 36) Periscyphis vandeli Ferrara, 1973 Somalia 37) Periscyphis verhoeffi Arcangeli, 1929 Ethiopia and Kenya * Four species described in Periscyphis (P kalongensis Arcangeli, 1950 from Zaire, P montanus Schmölzer, 1974, P pallidus Schmölzer, Periscyphis vittatus Omer-Cooper, 1926 Saudi Arabia, United Arab Emirates, Oman, Yemen, Eritrea, Djibouti, Somalia and Mozambique 1974 from Kenya, and P niger Schmölzer, 1974 from Tanzania) certainly do not belong to this genus. 10 Ss: Taitı eral: Acknowledgements We wish to express our sincerest thanks to Dr. L. Sandberg (NRM) for the loan of the type- specimens of Periscyphis brunneus, to Dr. M. D. Gallagher (ONHM) who entrusted us with the specimens of the new species P albomarginatus, and to Dr. V. Neumann (SMF) who collected specimens of the new species P sudanensis during his marine biological stay at Sanganeb Is., Sudan. Zusammenfassung Periscyphis brunneus Budde-Lund, 1912 von Kenia wird auf der Basis des Typenmaterials und neuer Exemplare aus Kenia wiederbeschrieben und der Lectotypus bestimmt. Zwei neue Arten, P sudanensis aus Sudan und P albomarginatus aus Oman, werden beschrieben. Die bis heute bekannten Arten der Gattung Periscyphis werden mit ihrer Verbreitung tabellarisch zusammengestellt. References Arcangeli, A. (1929): Specie nuove o poco note del genere Periscyphis Gerst. ed osser- vazioni sulle cavita incubatorie degli Isopodi terrestri. — Annuar. R. Mus. Zool. R. Univ. Napoli (NZS) 5: 715.20) pig Vil Arcangeli, A. (1934): Due specie ed un genere di Isopodi terrestri nuovi per la Libia. — Boll. Musei Zool. Anat. comp. R. Univ. Torino (3) 44: 213 —220, pls I—V. Arcangeli, A. (1940): Isopodi terrestri dell'Africa Orientale italiana. — Riv. Biol. colon. 3: 381-385. Barnard,K.H. (1940): XXXVI. Entomological Expedition to Abyssinia, 1926-7: Woodlice collected by Mr. J. Omer-Cooper. — Ann. Mag. nat. Hist. (1) 6: 355—366. Barnard, K. H. (1941): 8. Crustacea: Isopoda. — In: British Museum (Natural History). Expedition to South-West Arabia 1937-38, Vol. 1, 57—66. Jarrod & Sons Ltd., Norwich & London. Erhard, F & H. Schmalfuss (in press): The terrestrial isopod genus Periscyphis (Onisci- dea, Eubelidae) in Egypt and Israel. — Stuttgarter Beitr. Naturk., Ser. A. Ferrara, F. (1972): The genus Periscyphis Gerstaecker (Crustacea Oniscoidea Eubelidae) in Ethiopia. — Monitore zool. ital. (N. S.) Suppl. 4: 207—241. Ferrara, F. (1973): Observations on some species of Periscyphis (Terrestrial Isopods) gathered in Somalia. — Monitore zool. ital. (N. S.) Suppl. 5: 61-79. Ferrara, EF. (1974): Researches on the coast of Somalia. The shore and the dune of Sar Uanle. 3. Terrestrial Isopods. — Monitore zool. ital. (N. S.) Suppl. 5: 191—220. Ferrara, F.&S. Taiti (1979): A check-list of terrestrial isopods from Africa (south of the Sahara). — Monitore zool. ital. (N. S.) Suppl. 12: 89—215. Ferrara, F.&S. Taiti (1982): Periscyphis merolobatus, a new species of terrestrial isopod from Ethiopia. — Crustaceana 43: 9—12. Ferrara, F. &S. Taiti (1986): The terrestrial Isopods (Oniscidea) of the Arabian Peninsula. — Fauna Saudi Arabia 7 (1985): 93—121. Ferrara, F & S. Taiti (1988): Terrestrial isopods from Oman (Crustacea). — J. Oman Studies, Special Report 3: 391—396. Ferrara, F &S. Taiti (1996): Terrestrial Isopoda (Crustacea) from Yemen. — Fauna Saudi Arabia 15: 118—136. Lonnberg, E. & G. Budde-Lund (1912): Crustacea collected by the Swedish Zoological Expedition to British East Africa 1911. — Ark. Zool. 7 (No. 26): 1—9. Omer-Cooper, J. (1926): A revision of the Genus Periscyphis Gerst. (Isopoda terrestria). — Proc. zool. Soc. Lond. 24: 349—400. Paulian de Félice, L. (1945): Isopodes terrestres. — In: Mission Scientifique de ’Omo. Vol. 6, fasc. 60. — Mém. Mus. natn. Hist. nat., Paris (N. S.) 19: 331—344. Schmoólzer, K. (1974): Landisopoden aus Zentral- und Ostafrika (Isopoda, Oniscoidea). — Sber. Akad. Wiss. Wien (Math.-nat. Kl., Abt. 1) 182: 147—200. New species of Periscyphis 11 Taiti, S. & E Ferrara (1989): Terrestrial isopods of Saudi Arabia (Part 2). — Fauna Saudi Arabia 10: 78—86. Taiti, S. & E Ferrara (1991): New species and records of terrestrial isopods (Crustacea) from the Arabian Peninsula. — Fauna Saudi Arabia 12: 209—224. Vandel, A. (1964): De l’emploi des appareils respiratoires pour l’etablissement d'une classifi- cation rationnelle des Isopodes terrestres «Oniscoidea». — Bull. Soc. zool. Fr. 89: 730736. Stefano Taiti and Franco Ferrara, Centro di Studio per la Faunistica ed Ecologia Tropicali del Consiglio Nazionale delle Ricerche, Via Romana 17, 50125 Firenze, Italy. — Andreas Allspach, Wetzlarer Straße 50, 35398 Gießen, Germany. rot) tr, Gh tu u. uetzgilainserreil ER erario! | SA Ra MAT A Ñ r \ i oF R '- how Wy es Sige 7 | A ee eee raus er a mie tral” mon Du 30) cy = y ¿e Den ] 4 E ” nn B 1 SE at te q bey » rn ’ ' o ; y a rs, a = « A ae 1 ‘ - -. sil 5 ; o A - a u 7 ‘ i JJ | y er ' ( a ac © 4 ; ” 5 x ’ - i fi ok 3 ~ 5 E 2 1 \ - ig a ry En Ñ CH nn ES t E i Z yo MA 2 : N y Va ad NO O Ele zur N ; | [Det N E A | / J ye Bd. 47 S. 13—29 Bonn, September 1997 Coccinellidae (Coleoptera) aus Rwanda Helmut Fürsch Abstract. Coccinellids from Rwanda, collected by Thomas Wagner by fogging rainforest trees with pyrethrum, are listed and the following new species are described and figured: Afidenta muehlei, Boschalis wagneri, Epilachna carapacola, Epilachna conspergata, Nephus brevipilosus, Nephus rugulipennis, Ortalia gracilis, Pseudoscymnus brunneus, Scotoscymnus glabripilosus, Scotoscymnus maximus, Scymnus rwandensis, Telsimia striata. New synonyms: Scymnus luniferus Sicard is a younger synonym of Scymnus usam- baricus Weise and Scymnus (Nephus) circumcinctus Mader of Nephus (Sidis) burgeoni Mader. Key words. Coccinellidae, Africa, Rwanda, taxonomy, ecology, collecting methods. Einleitung Thomas Wagner erforschte im Oktober 1993 im Rahmen seiner Dissertation über Biodiversität in tropischen Waldökosystemen die Arthropodenfauna Rwandas und des benachbarten Ost-Zaire. Dabei wandte er neben Hand- und Kescherfängen eine neue Methode an: Er benebelte für die entsprechenden Phytozönosen dominierende Bäume mit einem Nebelgerát ’Swingfog SN-50’ und 1%iger Pyrethrumsuspension. Die ausgewählten Bäume waren nicht höher als 8 m, da vom Boden aus gearbeitet wurde und höhere Bäume nicht vollständig benebelt werden konnten. Mit dieser Methode wurden jeweils mehrere Bäume behandelt und die herunterfallenden Ar- thropoden in Trichtern aufgefangen. Damit war es erstmals möglich, den Arthropo- denbestand eines Baumes vollständig zu erfassen. Die Nummern der Einzelbäume sind hinter dem Artnamen der gefundenen Coccinelliden vermerkt. Wagners Vor- gehensweise erbrachte eine einzigartige Zahl neuer Arten, wie das seit den ersten Zeiten der Erforschung der Coccinellidenfauna nicht mehr der Fall war. Zur Abrun- dung wurden hier auch Sammelergebnisse von Hans Mühle aus Nyakabuye mit- bearbeitet. Material und Methoden Das Material von Thomas Wagner, zusammen mit den Aufsammlungen von Hans Mühle (Fürsch 1991) bieten eine Gesamtübersicht der Coccinellidae Rwandas unter Berücksichtigung der Arbeiten Maders (1941, 1950 und 1954) über die Exploration du Parc National Albert (heute Parc National des Virunga in Zaire und Parc National des Volcans in Rwanda) mit Ergänzungen von Fürsch (1991). In diesem Zusammenhang sei auch auf Nummelin & Fürsch - (1992) verwiesen. Das Material befindet sich im Museum Koenig Bonn (ZFMK) sowie in der Sammlung Fürsch (CF in ZSM). Vergleichsmaterial stammt auch aus dem Zoologischen Museum der Humboldt-Universität Berlin (MHB), dem Institut Royal des Sciences Naturelles, Bruxelle (IRSN), dem Zoolog. Museum Helsinki (ZMH), dem British Museum of Natural History, London (BMNH), dem Museum Georg Frey, München (MGF), der Zoologischen Staatssammlung München (ZSM), dem Musée National d’Histoire Naturelle, Paris, sowie dem Musée Royal de |’ Afrique Centrale, Tervuren (MRAC). 14 HPEUES CA 20 km nach Butembo P. N “des eae uhengeri P.N. Akagera RWANDA gai $ mi Karengera * VER ana “ut ep . Foret des Nyungwe Karte der Fundgebiete. Schraffiert: Trockenwaldsavanne; gerastert: Regen- und Nebelwald- gebiete (im Parc National des Virunga (ehemals Parc National Albert) sind grofe Savannen- gebiete enthalten). Gezeichnet nach verschiedenen Vorlagen, vor allem von Th. Wagner. Alle Genitalorgane sind nach Mikropráparaten in Hoyers Gemisch mit Zeichenapparat in vergleichbaren Maßstäben skizziert. Die beigegebene Karte soll einen Überblick über die Lage der Fundorte geben. Ergebnisse Der besondere Wert der Ausbeute Wagners liegt nicht nur in der für heutige Zeiten ungewöhnlichen Zahl neuer Arten, sondern in der Zuordnungsmöglichkeit zu bestimmten Ökotypen. Zunächst werden die Coccinelliden der verschiedenen Wald- typen aufgelistet und dann die neuen Arten beschrieben. Ziffern hinter dem Art- namen bedeuten Nummer des untersuchten Baumes. Wo diese fehlt, bezieht sich die Angabe auf Kescher- oder Handfang in dieser Phytozönose. 1. Trockenwald, südl. des Parc National de Il’ Akagera bei Ibanda Makera. Baumart: Lannea fulva (Anacardiaceae). Boschalis wagneri Fürsch 4 Exochomus troberti concavus Fürsch 1, 2 Nephus castaneicolor Sicard 1 Nephus rugulipennis sp. n. Ortalia argillacea Mulsant 1, 2, 3, 4 Ortalia gracilis sp. n. Ortalia pallens Mulsant 1 Platynaspis capicola Crotch Platynaspis kollari Mulsant Scymnus (Pullus) rwandensis sp. n. 4 Scymnus (Pullus) usambaricus Weise 4 Coccinellidae aus Rwanda 15 Scymnus levaillanti Mulsant Scymnus nummelini Fürsch 4 Scymnus pruinosus Weise 4 2. Galeriewald südl. des Parc National de Il’ Akagera bei Ibanda Makera. Baumart Teclea nobilis (Rutaceae). Chilocorus distigma Klug 10 Declivitata amoenula (Gerstäcker) 1, 3, 8 Nephus castaneicolor Sicard 5 Nephus microglobosus Fürsch 8 Nephus rugulipennis sp. n. Scymnus (Pullus) rwandensis sp. n. 3, 4, 5, 6, 8, 9 Scymnus (Pullus) thiollierei didymus Sicard 3, 10 Scymnus kibonotensis Weise 7 Scymnus levaillanti Mulsant 1, 3, 5, 6, 7, 8, 9 10 Scymnus nummelini Fürsch 6, 7 Scymnus pruinosus Weise 5, 6, 9 Telsimia inornata Casey 6 3. Nebelwald Foret de Nyungwe und Cyamudongo. Baumart Carapa grandiflora (Meliaceae). Aulis korschefskyi Mader 3, 12 Bambusicola centralis (Sicard) 1 (bei Karengera) Boschalis striata sp. n. 17 Boschalis wagneri Fürsch 1, 9, 11, 12, 17 Cheilomenes aurora (Gerstäcker) Chnootriba similis Thunberg 2 (bei Kamiranzovu) Epilachna aestimabilis (Mader) 3 Epilachna carapacola sp. n. 3, 17 Epilachna conspergata sp. n. 1, 2, (bei Kamiranzovu) Epilachna karisimbica (Weise) Epilachna korschefskyi (Mader) Epilachna loveni (Weise) 1 Epilachna novemdecemguttata (Weise) 1 Epilachna paradoxa (Mader) Epilachna tenelloides Fürsch 1 (bei Karengera) Epilachna zuluensis Crotch Exochomus troberti concavus Fürsch 2 (bei Kamiranzovu) Henosepilachna annulata (Kolbe) (bei Karengera) Henosepilachna biplagiata (Kolbe) Henosepilachna humerosa (Weise) 13 Henosepilachna kaesebergi (Weise) Henosepilachna lucifera (Arrow) Lotis neglecta Mulsant 2 (bei Kamiranzovu) Nephus brevipilosus sp. n. 1, 9, 12 Nephus circumcinctus (Mader) 2 (bei Kamiranzovu) Pseudoscymnus brunneus sp. n. 4, 7, 11, 12 Scotoscymnus glabripilosus sp. n. 1, 2, 4, 7, 9, 12, 17 Scotoscymnus maximus sp. n. 3, 4, 9, 11, 12, 17 Scymnus (Pullus) severus Weise 2 Scymnus (Pullus) thiollierei didymus Sicard 11 Scymnus (Pullus) usambaricus Weise 1, 3, 4, 7, 9, 11, 12, 17 Scymnus kibonotensis Weise Scymnus nummelini Fürsch Telsimia striata sp. n. 3, 4 Foret de Nyungwe bei Rwrasenkoko. Baumart Hagenia abyssinica (Rosaceae). Scymnus levaillanti Mulsant 2 16 Sol weiterhin: Chnootriba neglecta Mader auf Alchemilla ellenbeckii Lioadalia sexareata Weise auf Alchemilla ellenbeckii 4. Karisimbi 3000 m, auf Senecio johnstonii. Epilachna gyldenstolpei (Weise) Lioadalia sexareata Weise 5. Oberer Tieflandregenwald, Forschungsstation Irangi in Ost-Zaire. Cheilomenes aurora (Gerstäcker) Declivitata inclusa Mulsant Epilachna aestimabilis (Mader) Epilachna boops Fürsch (bei Kahuzi Tschiranga) Epilachna Kaestneri conjugata Fürsch | Epilachna nigromarginata Fürsch (bei Kahuzi Tschiranga) Epilachna tenelloides Fürsch Epilachna tredecimpunctata (Mader) Henosepilachna mutata fulvicollis Fürsch Scymnus (Pullus) thiollierei didymus Sicard, auf Carapa grandiflora 11, 51 Stethorus weisei Mader, Mt. Ilimu, 1700 m, auf Carapa grandiflora 11 6. Bemerkenswerte und neue Arten Scymnus (Pullus) rwandensis sp. n. (Abb. 1-8, 106) Diagnose: Rundlich, hochgewölbt, gelb, Elytren ringsum geschwärzt, am breitesten an der Basis. Länge: 1.6—1.8 mm; Breite: 1.1—1.3 mm. Beschreibung: Kopf gelb, fein genetzt und mit Punkten von der Größe der Augenfacetten. Die weißen Haare sind nach der Mitte gerichtet. Pronotum etwas dunkler gelb, auf glattem Untergrund Punkte, die etwas kleiner sind als die Augenfacetten. Ziemlich dicht gelbweiß behaart, hinter den Augen wirbelig, sonst halbaufrecht nach hinten (elytrenwärts) gerichtet. Pronotum verbreitert sich nach hinten und ist nur knapp vor der Basis etwas eingeengt. Scutel- lum schwarz. Elytren in der Mitte gelb, sonst schwarz gerahmt, Elytrenhinterrand gelb wie Abb. 106. Hinter der Mitte am breitesten. Oberfläche fein genetzt und grob punktiert. Schul- terbeule deutlich. Die Elytrenwölbung fällt zum Seitenrand sehr steil, fast senkrecht ab. Seiten- randung nur an der Schulter deutlich. Elytrenbehaarung wie auf Pronotum nach hinten gerichtet. Differentialdiagnose: Scymnus alluaudi Sicard zum Verwechseln ähnlich, dieser ist auf den Elytren noch etwas deutlicher punktiert. Auf fast glatter Oberfläche stehen die sehr großen Punkte etwa einen Punktdurchmesser voneinander getrennt. Bei S. rwandensis ist die Punktierung „unruhiger“, rugulöser. Einzig sicheres Unterscheidungsmerkmal: Die Sipho- spitze (Abb. 3,4). Sie ist bei S. rwandensis meist doppelt, während sie bei S. alluaudi in einer dünnen Spitze endet und davor ein ,,Stachelpolster“ aufweist (Abb. 17—20). Dieses sieht man zuweilen auch bei S. rwandensis, doch fehlt hier die fein ausgezogene Spitze. Ähnlich ist auch S. usambaricus Weise, doch ist diese Art meist einfarbig braungelb. Material: Holotypus, männl. Rwanda: Rusumo, Ibanda Makera 10. 1993 auf Teclea nobilis, leg. Th. Wagner. 7 Paratypen mit den gleichen Daten (ZFMK, 2 CF). Weiteres Material: Uganda, W. Prov. Kibale Forest, sweep K 14 5. 1. 1984 leg. N. Nummelin (ZMH, CF). Scymnus (Pullus) usambaricus Weise (Abb. 10-16) Scymnus (Pullus) usambaricus Weise, 1897 Dt. ent. Z.: 301. Fürsch 1966: 176. Scymnus (Pullus) luniferus Sicard, 1931: 229, Syn. nov. Untersuchtes Material: Lectotypus und Paralectotypen von S. usambaricus aus Kwai (MHB), Lectotypus von S. /uniferus, männl. aus den Usambara Mts. bei Amani 20. 2. 1926, “taken by beating coffee foliage” coll. A. H. Ritchie. 2 Paralectotypen mit den gleichen Daten Coccinellidae aus Rwanda 17 Tafel 1: 1-8: Seymnus rwandensis sp. n.: 1—3 Holotypus: 1 Holotypus (b); 2 Tegmen (c); 3 Siphospitze (c); 4 Siphospitze, Kibale Forest (c). 5—9 Paratypen: 8 Siphonalbasis (c); 9 Abdomen (a). — 10—16: Seymnus usambaricus: 10—12 Cyamodungo; 13 Bulengwa; 14, 15 - Paratypen; Lectotypus von S. /uniferus. — 17—20: Scymnus alluaudi, Siphospitzen: 17 Gar- garo Tana; 18 Kamerun, Nkolentangan; 19 Tanzania, Duluti; 20 id. Usa River. — 21—24 Nephus rugulipennis sp. nov. Holotypus: 24 linke Hälfte des 1. Abdominalsternits mit Femo- rallinie (a). —a—c = Maßstäbe wie auf Tafel 4 unter Abb. 77: a = 1 mm; b, c, = 0,1 mm. Alle Detailabbildungen sind in diesen Maßstäben gezeichnet. Die dargestellten Details sind in der Legende nicht aufgeführt, wenn sie vorher schon erläutert worden sind, um die Übersicht- lichkeit zu erhöhen und Wiederholungen zu vermeiden. 18 HakRünsch (Festlegung hiermit) (BMNH). Beide Arten unterscheiden sich weder im Habitus noch am Aedeagus voneinander. Kwai liegt bei Amani. Zahlreichen Belege aus Rwanda leg. Th. Wagner (siehe Liste) (ZFMK), Kenya, Tanzania (MHB, ZSM, CF). Nephus (Sidis) rugulipennis sp. n. (Abb. 21-24, 107) Etymologie: lat. rugula = kleine Runzel; penna = Flügel(decken). Diagnose: Langoval, schwarz, nur Kopf (männl.), Mundwerkzeuge, Beine und ein Fleck in der hinteren Hälfte der Elytren gelbrot (Abb. 107 ). Länge 1.55 mm; Breite: 1.0 mm. Beschreibung: Langoval, fast zylindrisch. Elytrenseiten sehr flach gerundet. Kopf gelbröt- lich, deutlich genetzt, undeutlich punktiert. Mundwerkzeuge und Fühler gelbrötlich. Prono- tum schwarz, seine Vorderwinkel rötlich, auf genetzter Oberfläche fein punktiert (Punkte kleiner als Augenfacetten). Behaarung weiß, ziemlich lang. Elytren schwarz mit braunrotem, verwaschenem Fleck hinter der Mitte wie Abb. 107. Elytrenhinterende schmal rötlich. Die dichte, sehr grobe Punktierung gibt der Oberfläche ein runzeliges Aussehen. Schulterbeule flach, Elytrenwölbung zum Seitenrand sehr steil. Unterseite schwarz mit braunrotem Abdo- men. Aedeagus Abb. 21—23. Differentialdiagnose: In der Form ähnlich den südafrikanischen Arten N. binaevatus Mulsant und posticesignatus Pope sowie der ostafrikanischen N. stigma Weise. Sie unter- scheiden sich alle am männl. Aedeagus, N. stigma ist auf glatter Elytrenoberfläche stärker punktiert, die Schulterbeule ist kräftiger und die Körperform ist breiter. Material: Holotypus: männl. Rwanda: Rusumo, Ibanda Makera 10. 1993 leg. Th. Wagner (ZFMK). Nephus (Sidis) brevipilosus sp. n. (Abb. 25—33, 108) Etymologie: lat. brevis = kurz; pilosus = behaart, wegen der kurzen, halb aufgerichteten Behaarung. Diagnose: einfarbig dunkelrot, hochgewólbt und breit gerundet. Lange: 1,8—2.5 mm; Breite: 1.35—1.80 mm. Beschreibung: Kopf gelb (mánnl.) oder schwarzrot (weibl.), sehr dicht und fein punktiert. Punkte kleiner als Augenfacetten. Sehr dicht gelb behaart. Pronotum rotbraun, auf glatter Oberfläche sehr dicht und fein punktiert, die weißen Haare nach hinten gerichtet. Pronotum- seiten sanft geschwungen, vorne enger als an der Elytrenbasis, fein gerandet. Elytren von der Färbung und Behaarung des Pronotums. Auf fast glatter Oberfläche fein, aber weniger dicht punktiert als Pronotum. Punktzwischenräume größer als ihre Durchmesser. Schulterbeule ziemlich schmal und nicht sehr auffällig. Elytrenwölbung fällt fast senkrecht gegen den fein gekanteten Seitenranad ab. Gleichmäßig gerundet, in der Mitte am breitesten. Differentialdiagnose: In der Größe nur der nächst verwandten Art Nephus burgeoni Mader sehr ähnlich. N. burgeoni ist aber schwarz gerandet, kann allerdings in seltenen Fällen auch ganz braunrot sein wie die neue Art. Die Behaarung von N. burgeoni ist steifer und vor allem hat diese Art einen kaum auffallenden Humeralcallus. Parameren von N. brevipilosus kürzer als bei N. burgeoni. Der Aedeagus der neuen Art weist gewisse Ähnlichkeit zu dem von N. globulus Fürsch auf, diese westafrikanische Art sieht aber völlig anders aus. Eine große äußere Ähnlichkeit besteht zu Pseudoscymnus brunneus sp. n. Material: Holotypus: männl. Rwanda: Nyakabuye, Cyamudongo 10. 1993 auf Carapa gran- diflora leg. Th. Wagner (ZFMK). 3 Paratypen mit den gleichen Daten (1 CF); 2 Paratypen: Rwanda: Karengera 1700 m 10. 1993 auf Carapa grandiflora leg. Th. Wagner (1 CF). Nephus (Sidis) burgeoni (Mader) Scymnus (Nephus) burgeoni Mader, 1950: 64 Scymnus (Nephus) circumcinctus Mader, 1950: 65. syn. nov. Nephus (Bipunctatus) circumcinctus (Mader) — Fürsch 1992: 40. comb. nov. Scymnus (Sidis) longemaculatus Mader, 1950: 59. — Fürsch, 1992: 40. Coccinellidae aus Rwanda 19 Tafel 2: 25—33: Nephus brevipilosus sp. n.: 25—29 Holotypus; 30—33 Paratypen; 33 Sper- matheca (c). — 34—43 Pseudoscymnus brunneus sp. n. Paratypen: 43 Spermatheca (b). 20 He Bwrsich Material: Lectotypus und 5 Paralectotypen von N. burgeoni: Rutshuru (MRAC, MGF, CF); Holotypus (männl.) von S. circumcinctus, Volc. Nyamuragira (IRSN), Lectotypus (mánnl.) von S. longemaculatus: Rwanda, Mt Tamira (Náhe Lac Gando) (MRAC) und viel Material aus Rwanda (leg. Th. Wagner, ZFMK, MRAC, MGF, ZSM, CF) enthält sowohl Tiere mit schwar- zer Elytrenumrandung wie auch solche von der Fárbung des N. brevipilosus, also ganz rot- braun. Die von Mader (1950: 65) angegebenen Differentialmerkmale beziehen sich lediglich auf Größe und Färbung der Unterseite und sind variabel. Mit Nephus oblongosignatus (Mul- sant) hat diese Art, entgegen Maders Angaben, nichts zu tun (vgl. Chazeau, Etienne & Fürsch 1974: 273). Pseudoscymnus brunneus sp. n. (Abb. 34-43, 109) Etymologie: lat. brunneus = braun. Diagnose: Breit, hochgewölbt, dunkel rotbraun, zart rötlich an Länge 1.9—2.5 mm; Breite: 1.4—1.9 mm. Beschreibung: Kopf gelbrot (Holotypus) oder auch rot, dicht und fein punktiert. Punkte kleiner als Augenfacetten. Dicht weiß behaart. Haare im oberen Drittel des Kopfes gegen die Augen gerichtet, davor zur Oberlippe. Pronotum ähnlich dicht punktiert, Oberfläche dazwi- schen fast glatt. Behaarung an den Seiten weiß, in der Mitte rot. Pronotumbreite an der Elytrenbasis am größten, fast gerade nach vorne verengt. Elytrenoberfläche deutlich reticuliert und deutlicher punktiert als Pronotum. Behaarung fein, schütter, rötlich. Schulterbeule breit, nicht besonders deutlich. Zwischen Humeralcallus und Elytrenseitenrand seicht konkav. Elytren hochgewölbt und breit gerundet. Elytrenseitenrand fein gekantet, dunkel. Aedeagus Abb. 34—36, 38, 39, 42. Differentialdiagnose: Die neue Art ist von allen anderen afrikanischen schon in ihrer Größe und der einheitlich rotbraunen Färbung unterschieden. Die anderen Arten haben wenigstens ein etwas helleres Pronotum. Die männl. Genitalorgane unterscheiden sich von den übrigen Arten (vgl. Fürsch 1990). P brunneus hat allerdings große äußerliche Ähnlichkeit mit der im gleichen Gebiet vorkommenden Nephus brevipilosus sp. n. Material: Holotypus, männl.: Rwanda: Nyakabuye, Cyamudongo 10. 1993 leg. Th. Wagner (ZFMK). 11 Paratypen mit den gleichen Daten (3 CF). Ortalia gracilis sp. n. (Abb. 44-47) Etymologie: lat. gracilis = zierlich, wegen der in dieser Gattung geringen Größe. Diagnose: Ober- und Unterseite einheitlich blaßgelb. Länge: 2.9 mm; Breite: 2.2 mm. Beschreibung: Kopf auf glatter Oberfläche dicht punktiert. Punkte etwa von der Größe der Augenfacetten. Mandibelspitzen dunkelbraun. Auch auf den Augen weiß behaart. Prono- tum sehr dicht punktiert, vor der Mitte am breitesten, zur Basis fast gerade sehr leicht verengt. Elytren langoval, hinter der wenig deutlichen Schulterbeule am breitesten, dann fast parallel- seitig. Wie auf Pronotum weiß behaart. Punktierung regelmäßig und dicht. Differentialdiagnose: Sehr ähnlich den Ortalia-Arten, argillacea Mulsant (Abb. 50— 52), ochracea Weise (Abb. 48, 49), sjoestedti Weise und ovulum Weise (Abb. 53, 54). Bei Ortalia ovulum ist der Seitenrand viel breiter, O. ochracea ist größer und stärker gerundet, O. sjoestedti ist länger behaart und auf der Unterseite schwarz. Ein gutes Differentialmerkmal ist die Siphospitze (Abb. 45). Material: Holotypus, männl.: Rwanda: Rusumo, Ibanda Makera 10.1993 leg. Th. Wagner (ZFMK). Telsimia striata sp. n. (Abb. 55-61, 110) Etymologie: lat. striatus -a = gestreift. Diagnose: gerundet, schwarz mit einem rötlichen nierenförmigen Fleck wie Abb. 110 auf jeder Elytra. Zottig weiß behaart. Länge: 1.3—1.4 mm; Breite: 1.1 mm. Coccinellidae aus Rwanda >| x STR, [a] LD a A. S a S 283 SI NES S 61 Tafel 3: 44—47 Ortalia gracilis sp. n. Holotypus: 44 Tegmen ventral (b); 45 Sipho (c); 47 Abdomen (a). — 48, 49: Ortalia ochracea: 48 Siphospitze von dorsal gesehen, 49 von ventral. — 50—52: Ortalia argillacea von Ibanda Makera. — 53, 54: Ortalia ovulum, Kamerun, Joko. — 55—61: Telsimia striata, Paratypen: 55 Aedeagus (c); 56 Hinterbein (b); 57 Hintertarsus (c); 58 Kiefertaster (c); 59 Mandibel (c); 60 weibl. Genitalplatten (b); 61 Abdomen (b). 22 Hakinsıch Beschreibung: Kopf dicht punktiert, Oberfläche dazwischen glatt, weiß behaart. Prono- tum sehr dicht punktiert, Punkte größer als auf dem Kopf. Wirr weiß behaart. Pronotum- seiten gerundet mit deutlicher Randkante. Elytrenpunktierung noch größer als auf Pronotum, Oberfläche dazwischen sehr fein genetzt, glänzend. Behaarung lang, aufrecht. Auf jeder Elytra ein brauner Fleck wie Abb. 110. Elytrenseiten breit gerundet, Schultern vorgezogen. Differentialdiagnose: Durch ihre Zeichnung von allen anderen afrikanischen Arten leicht zu unterscheiden. Material: Holotypus, männl. Rwanda: Cyamudongo, 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (ZFMK). 4 Paratypen mit den gleichen Daten (1 CF); 1 Paratypus: Rwanda: Rusumo, Ibanda Makera 10. 1993 auf Carapa grandiflora leg. Th. Wagner (CF). Boschalis wagneri Fürsch (Abb. 62—66, 111) Boschalis wagneri Fürsch, 1995: 30, Abb. 49-52. Diese von Herrn Thomas Wagner entdeckte Art wurde bereits von Fürsch (1995) ausführlich beschrieben; die Diagnose wird hier durch weitere Abbildungen ergänzt. Material: Holotypus, männl.: Rwanda: Nyakabuye, Cyamudongo, 1700 m 10. 1993 auf Carapa grandiflora leg. Th. Wagner (ZFMK). 7 Paratypen mit den gleichen Daten (2 CF); 1 Paratypus: Rwanda: Rusumo, Ibanda Makera auf Lannea fulva (CF). Scotoscymnus glabripilosus sp. n. (Abb. 67-70, 112) Etymologie: lat. glaber = glatt; pilosus = behaart, wegen der im Vergleich zu anderen Arten relativ glatt nach hinten gerichteten Behaarung. Diagnose: Dunkel rötlichbraun, langoval, Elytrenbehaarung regelmäßig nach hinten gerich- tet. Lange: 1.3—2.0 mm; Breite: 0.9—1.4 mm. Beschreibung: Kopf mit Punkten von der Größe der Augenfacetten ziemlich dicht besetzt, Oberfläche dazwischen glatt. Behaarung dünn, lang, weiß. Pronotumpunktierung noch etwas größer, aber weniger dicht als auf dem Kopf. Pronotum an den Seiten stark gerundet, hinter der Mitte am breitesten. Behaarung weiß, lang, zum großen Teil nach vorne gerichtet. Elytren- punktierung und -behaarung wie auf Pronotum, aber Haare halbaufrecht und gleichmäßig nach hinten gerichtet. Elytrenseitenrand ziemlich breit und deutlich horizontal. Elytrenform länglich, vor der Spitze etwas eingezogen (gutes Merkmal!). 1. und 2. Abdominalsegment ver- einigt, aber Trennlinien doch noch zu erkennen. Differentialdiagnose: Ähnlich S. rotundatus (Weise), bei dem aber die Proa tierung schwácher, dafúr Elytrenpunktierung viel stárker ist. Obwohl die neue Art sehr dunkel rotbraun sein kann, ist sie doch nie völlig schwarz wie S. niger (Weise). In der Behaarung sind die beiden Arten ähnlich. S. niger ist auf den Elytren stärker punktiert. Ein verläßliches Merk- mal ist die Siphospitze (Abb. 70). S. parvulus (Weise) und S. punctipennis (Weise) haben auf- fallendere Seitenränder und andere Aedeagi. S. minutus (Fürsch) ist weniger deutlich punk- tiert als S. glabripilosus und zudem schon an seiner konischen Körperform (kurz hinter den Schultern am breitesten) gut zu erkennen. Material: Holotypus, männl.: Rwanda: Karengera 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (ZFMK). 8 Paratypen mit den gleichen Daten (2 CF); 1 Paratypus: Rwanda: Rusumo, Ibanda Makera 10. 1993 leg. Th. Wagner; 3 Paratypen: Rwanda: Nyungwe, Kamiran- zovu 2000 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner; 14 Paratypen: Rwanda: Cyamu- dongo 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (4 CF). Scotoscymnus maximus sp. n. (Abb. 71-76, 113) Etymologie: lat. maximus = der Größte. Diagnose: Gleichmäßig gerundet, dunkel rotbraun mit auffallend breitem Seitenrand. Eänse 151. 72mm# Breite" 4Smm: Coccinellidae aus Rwanda 23 Tafel 4: 62—66 Boschalis wagneri: 62 Paratypus Tegmen, schräg ventral (b); 63 id. Holo- typus, lateral; 64—66 Paratypen; 66 weibl. Genitalplatte (b). — 67—70: Scotoscymnus glabri- pilosus sp. n. Paratypen. — 71-76: Scotoscymnus maximus sp. n. Paratypen. — 77—83: Epilachna conspergata sp. n.: 77—78 Holotypus; 79—83 Paratypen. Unter 77: Maßstäbe für alle Detailzeichnungen. 24 HeakRinsch Beschreibung: Kopf mit Punkten von der Größe der Augenfacetten locker besetzt, weiß behaart. Pronotum an den Seiten stark gerundet, Punktierung ziemlich fein, ungleich. Prono- tumbehaarung lang, dünn und wirr. Elytrenoberfläche fein genetzt mit viel größeren Punkten als auf dem Pronotum. Elytrenseitenrand besonders unter und hinter der undeutlichen Schul- terbeule breit subhorizontal. Die gleichmäßig gerundeten Elytren sind kurz vor der Mitte am breitesten. Elytrenhaare lang, weißlichgelb, zum größten Teil nach hinten gerichtet. Differentialdiagnose: Viel breiter gerundet als S. glabripilosus und mit breiterem Elytrenseitenrand. Beste Merkmale: Die gekrümmte Dorsalseite des Basallobus, schwache Schulterbeule, stark ausgeprägter Elytrenseitenrand. Material: Holotypus: Rwanda: Nyakabuye, Cyamudongo 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (ZFMK). 27 Paratypen mit den gleichen Daten (9 CF); 1 Paratypus: Rwanda: Karengera 1700 m, sonst gleiche Daten. Weiteres Material: Zaire: Kivu Sud, Irangi 900 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner. Epilachna conspergata sp. n. (Abb. 77-83, 114), Epilachna sahlbergi-Gruppe (vgl. Fürsch 1963) Etymologie: Lat. conspergere = bestreuen, die Oberfläche sieht wie mit Pfeffer bestreut aus. Diagnose: Oval, braun mit wolkig unregelmäßiger schwarzbrauner Zeichnung wie Abb. 114. Lange: 3.8—4.4 mm; Breite: 3.2—3.5 mm. Beschreibung: Kopf gelb mit schwarzbraunem Fleck auf der Stirn, der halb vom Prono- tum verdeckt ist (Holotypus, Paratypus), oder auch ganz schwarzbraun (Paratypus). Fein genetzt und mit feineren Punkten als die Augenfacetten. Pronotum mit Punkten von der Größe der Augenfacetten dicht besetzt. Pronotumbehaarung fein, weißlichgelb. Pronotum an den Seiten breit gerundet. Scutellum gelb, schwarz gerandet (Holotypus) oder ganz gelb. Elytrenpunktierung sehr dicht und größer als auf Pronotum. Behaarung kürzer als auf Prono- tum. Humeralcallus sehr deutlich. Elytrenabdachung gegen den Seitenrand flach. Abdomen dunkelbraun, Beine heller. Siphospitze etwas variabel, wie in der Gruppe um E. gyldenstolpei (Weise) üblich. | Differentialdiagnose: In der Zeichnung sehr ähnlich E. zumpti Fürsch (1963). Bei dieser Art sind die Pronotumseiten gerade, während sie bei der neuen Art auffallend gerundet sind. Wichtigstes Differentialmerkmal: Markanter Höcker an der Basis des Basallobus bei E. zumpti, wogegen die neue Art hier kaum eine flache Erhebung zeigt. Die Elytrenpunktie- rung ist bei E. conspergata etwas stärker als bei E. zumpti. Material: Holotypus, männl.: Rwanda: Nyungwe, Kamiranzovu 2000 m auf. Carapa grandi- flora 10. 1993 leg. Th. Wagner (ZFMK). 1 Paratypus mit den gleichen Daten (CF); 2 Para- typen: Rwanda: Karengera 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (ICF); 1 Paratypus: Rwanda: Nyakabuye 28. 10.—2. 11. 1983 leg. H. Mühle (CF). Epilachna apiceoculata sp. n. (Abb. 84-86, 115), Epilachna sahlbergi-Gruppe Etymologie: lat. apex, apicis = Spitze; oculatus = mit Augenflecken versehen, wegen der auffälligen Zeichnung an der Elytrenspitze. Diagnose: Braun mit schwarzer Zeichnung wie Abb. 115. Basallobus auffallend dick. Länge: 3.3 mm; Breite: 2.6 mm. Beschreibung: Kopf in der Mitte schwarz, neben den Augen braun, dicht punktiert und behaart. Oberlippe, Fühler und Mundwerkzeuge braun. Pronotum braun, gelbweiß und lang behaart. Pronotumseitenrand sanft gerundet, seine Oberfläche kaum sichtbar genetzt, aber dicht punktiert. Punkte von der Größe der Augenfacetten und durch Zwischenräume dieser Größe voneinander getrennt. Scutellum braun. Elytren braun mit schwarzer Zeichnung wie Abb. 115. Ebenso dicht und etwas größer punktiert als Pronotum. Behaarung wie auf Prono- tum. Humeralcallus sehr auffällig. Die Elytrenwölbung fällt steil zum nicht gekanteten Seiten- rand ab. Elytren hinter der Mitte am breitesten. Unterseite und Beine braun. Femorallinie ein gleichmäßiger Bogen, der 2/3 der Länge des ersten Abdominalsternits einschließt und beinahe den Vorderrand erreicht. Coccinellidae aus Rwanda 25 Tafel 5: 84-86 Epilachna apiceoculata sp. n. Holotypus; 87—97 Epilachna gyldenstolpei: 87, 88: Kivu, Shibinda; 89 Butembo; 90 Irangi; 91 Nyakabuye; 92 Parc Nat. des Volcans, Lubero; 94, 95 Ruwenzori; 96 Virunga; 97 Nyakabuye. 26 H. Fúrsch Differentialdiagnose: In der Verwandtschaft um Epilachna gyldenstolpei (Weise) (vel. Fürsch 1963) ist es bei der Variabilität der Zeichnung und der Siphospitzen sehr schwierig, Arten abzugrenzen. Epilachna apiceoculata ist habituell sicher nicht zweifelsfrei zu erkennen, zumal das einzige vorliegende Expl. keine Hinweise auf die Variabilität zuläßt. Das beste Merkmal ist der dicke Basallobus. Das Problem bei diesen Arten ist die Variabilität der Sipho- spitzen. Wagners Ergebnisse erlauben, zusammen mit dem ungeheuren Material des MRAC, eine Darstellung dieser Variabilität (Abb. 87—97). Auch in Form und Färbung des Körpers sind die Arten variabel, ganz besonders E. gyldenstolpei. Bei letzterer ist die Elytrenwólbung flacher, besonders zum Seitenrand hin. E. aulisoides (Weise) ist weniger deutlich punktiert, und die Schulterbeule ist flacher. Material: Holotypus: Rwanda: Prov. Cyangugu, Umgebg. Nyakabuye 13.—15. 5. 1985 leg. H. Mühle. Gen. Präp. Nr. 3153 (ZSM) Epilachna carapacola sp. n. (Abb. 98, 99, 116, 117), Epilachna canina-Gruppe (vgl. Fürsch 1985a) Etymologie: Zusammengezogen aus Carapa = Baumgattung und lat. incola = Bewohner. Diagnose: Ockerbraun mit schwarzbrauner, unvollständiger Netzzeichnung wie Abb. 116, 117. Länge: 5.2—6.1 mm; Breite: 4.6—4.8 mm. Beschreibung: Kopf rotbraun, sehr dicht punktiert. Punkte etwa von der Größe der Augenfacetten, Oberfläche dazwischen kaum mehr sichtbar. Mundwerkzeuge schwarz mit braunen Rändern. Pronotum rotbraun, ebenfalls sehr dicht punktiert. Behaarung wie auf Kopf weiß, ziemlich lang, aber zart. Innerhalb des aufgewulsteten Pronotumseitenrandes eine flache Depression. Elytren heller als Kopf und Pronotum. Elytrenpunktierung sehr dicht, die gröberen Punkte etwa 4mal so groß wie auf dem Pronotum, die kleineren etwa so groß wie auf dem Pronotum. Das schwarze Netz sehr unvollständig wie Abb. 116, 117. Elytrenseiten- rand ganz schmal gewulstet, schwarz, daneben sanft eingemuldet. Femorallinie nahezu voll- ständig. Unterseite und Beine schwarz, Abdomenseiten braun. Differentialdiagnose: Ausgezeichnet durch stark unvollständige Elytrenzeichnung, Siphospitze ohne Serra (Fürsch 1985a: 193) und einen Basallobus, der steiler abfällt als bei einer anderen Art. Nach der Tabelle von Fürsch (1985a) käme man zur südafrikanischen Epilachna dregei Mulsant, die aber eine viel längere Lobusspitze hat, abgesehen von der voll- ständigeren Elytrenzeichnung. Material: Holotypus, männlich: Rwanda: Cyamudongo, 1700 m auf Carapa grandiflora 10. 1993 leg. Th. Wagner (ZFMK). 1 Paratypus weibl. (ZFMK), 1 männl. (CF) mit den gleichen Angaben. Afidenta muehlei sp. n. (Abb. 100-105, 118) Etymologie: Dem Entdecker Hans Mühle, München, gewidmet, der ein Forstprojekt in Rwanda wissenschaftlich betreut hat und dabei eine höchst bemerkenswerte Ausbeute mitbrin- gen konnte. Diagnose: Rot, auf den Elytren mit zusammen 13 kleinen schwarzen, runden Flecken wie Abb. 118. Länge 3.3—3.4 mm; Breite: 2.6—2.7 mm. Beschreibung: Kopf gelb (Holotypus) oder rot mit flachen Punkten, die etwas größer sind als die Augenfacetten, fein genetzter Oberfläche und goldgelber Behaarung. Pronotum rot, in den Vorderwinkeln gelb und auf glatter Oberfläche etwa in gleicher Größe wie auf Kopf punktiert. Elytren rot mit kleinen schwarzen, runden Flecken wie Abb. 118, so stark und ungleichmäßig punktiert, daß sie auch bei stärkerer Vergrößerung rauh erscheinen. Elytren- seitenrand schmal horizontal. Schulterbeule deutlich. Femorallinie Abb. 103. Aedeagus: Spitze des Basallobus leicht paramerenwärts gekrümmt. Differentialdiagnose: Von Afidenta maderi (Korschefsky) kaum zu unterscheiden, letztere Art hat auf der Naht nur zwei gemeinschaftliche Flecken und die neue Art drei. Afidenta malawiensis Fürsch (1996) hat wohl größere Flecken, ist aber im Aedeagus ähnlich. Coccinellidae aus Rwanda 27 - Tafel 6: 98—99: Epilachna carapacola sp. n. Holotypus: 98 Aedeagus (a); 99 Tegmen (b). — 100-105: Afidenta muehlei sp. n.: 100—103 Holotypus; 104, 105 Paratypus. — 106—119: Umrißskizzen: Maßstab für 106—113 neben Abb. 111 = 1 mm; Maßstab für 114—119 neben Abb. 116 = 5 mm. 106 Scymnus rwandensis; 107 Nephus rugulipennis; 108 Nephus brevipilo- sus; 109 Pseudoscymnus brunneus; 110 Telsimia striata; 111 Boschalis wagneri; 112 Scotoscym- nus glabripilosus; 113 Scotoscymnus maximus; 114 Epilachna conspergata; 115 Epilachna apiceoculata; 116, 117 Epilachna carapacola; 118 Afidenta muehlei; 119 Afidenta scitula. 28 Eres In der Tabelle von Fürsch (1986) käme man zur äußerlich gleichen westafrikanischen Afidenta scitula (Weise) (Abb. 119), aber deren Aedeagus ist völlig verschieden, und ihre Pronotum- ränder sind stärker aufgebogen, auch die Punktierung ist etwas gröber als bei der neuen Art. Diese äußerliche Gleichheit zweier Arten ist bei dieser Gattung keineswegs neu (vgl. Fürsch 1986). Material: Holotypus, männl.: Rwanda: Prov. Cyangugu, Umgebg. Nyakabuye 13.—15. 5. 85 leg. H. Mühle Gen. Präp. 3160 (ZSM). 2 Paratypen: gleiche Angaben, aber 10. 1. 1986, 1 Paratypus: 30. 1.—3. 2. 1984 (CF). Weiteres Material: Rwanda: Rangiro 9. 1976 leg. Werner (CF), bei Fürsch (1986: 90) als Afidenta scitula angegeben. Letztere Art bleibt damit auf West- afrika beschránkt. Danksagung Herrn Thomas Wagner sei auch hier nochmals herzlich für die Möglichkeit gedankt, seine Ausbeute bearbeiten zu können. Besonderen Dank schulde ich Herrn Hans Mühle für die Überlassung seiner großen und wertvollen Aufsammlung. Ohne die bereitwillige und verständ- nisvolle Zusammenarbeit mit den großen Museen wären taxonomische Arbeiten heute nicht mehr denkbar, deshalb danke ich den Leitern dieser Institute ganz besonders. Es sind die Damen und Herren Dres. H. André, Tervuren, M. Baehr, München, Nicole Berti, Paris, R. G. Booth, London, M. Cludts Bruxelles, R. L. Davidson, Pittsburgh, H. Silfverberg, Helsinki und M. Uhlig, Berlin. Zusammenfassung Die Coccinellidenausbeute, die Thomas Wagner in Rwanda und Ostzaire vor allem durch Benebeln der fiir bestimmte Phytozónosen charakteristischen Baumarten mit Pyrethrum sam- meln konnte, wurde hier bearbeitet. Dazu wurde weiteres Material von Hans Mühle aus Rwanda ausgewertet. Folgende neue Arten werden beschrieben und abgebildet: Afidenta muehlei, Boschalis wagneri, Epilachna carapacola, Epilachna conspergata, Nephus brevipilo- sus, Nephus rugulipennis, Ortalia gracilis, Pseudoscymnus brunneus, Scotoscymnus glabripi- losus, Scotoscymnus maximus, Scymnus rwandensis, Telsimia striata. Neue Synonyme: Scym- nus luniferus Sicard ist jüngeres Synonym von Scymnus usambaricus Weise und Scymnus (Nephus) circumcinctus Mader ein jüngeres Synonym von Nephus (Sidis) burgeoni Mader. Literaturverzeichnis Chazeau, J., J. Etienne &H. Fürsch (1974): Les Coccinellidae de I’ ile de la Réunion. — Bull. Mus. Nat. Hist. Nat. 3(210) Zool. 140: 265 —297. Fiirsch, H. (1963): Möglichkeiten zur Festlegung niederer systematischer Kategorien, gezeigt an der Epilachna sahlbergi-Gruppe. — Veröff. Zool. Staatssamml. München 7: 161— 287. Fúrsch, H. (1966): Die Scymnus-Arten Westafrikas. — Entomol Arb. Mus. Frey 17: 135—192 Fiirsch, H. (1985): Die afrikanischen Sukunahikonini und Microweiseini mit Diskussion über alle Gattungen. — Dtsch. ent. Z., N.F. 32: 279—295. Fiirsch, H. (1985a): Die Epilachna canina-Gruppe. — Ent. Arb. Mus. Frey 33/34: 189—229. Fiirsch, H. (1986): Die afrikanischen Vertreter der Gattung Afidenta. — Kol. Rdsch. 58: 83—104. Fürsch, H. (1990): Additional notes on African Pseudoscymnius-species. — Coccinella 2: DD Fürsch, H. (1991): Die Epilachnini Afrikas südlich der Sahara. — Entomofauna 12 (18/ 3) a2 i320! Fürsch, H. (1992): Annotated checklist of African Nephus-species South of the Sahara. — Coccinella 4: 35—60. Fiirsch, H. (1995): Revision der Gattung Boschalis Weise, 1897, mit Beschreibung neuer Arten. — Mitt. Münch. Ent. Ges. 85: 21—31. Fürsch, H. (1996): New Coccinellids from Southern Africa. — Coccinella 6: (6—11). Mader, L. (1941): Coccinellidae I in: Exploration du Pare National Albert 34. Coccinellidae aus Rwanda 29 Mader, L. (1950): Coccinellidae II in: idem. Mader, L. (1954): Coccinellidae III in: idem 80. Nummelin, M & H. Fürsch (1992): Coccinellids of the Kibale Forest, Western Uganda: a comparison between virgin and managed sites. — Tropical Zoology 5: 155—166. Sicard, A. (1931): Descriptions d’ especes nouvelles appartenant a la famille des Coccinelli- des. — Ann. Mag. nat. Hist. 10 (8): 228—234. Dr. Helmut Fürsch, Bayerwaldstraße 26, D-94161 Ruderting. ir AS 44 i bl iy ae eet MR att oh a ur LE do Bush Pens Vu Er eth xpos A Y ate ax ena i peda + la i i “4 AE E A e em ; E: ARA 7 . i A A wi & Ú , ay “pt i v Woh i ' ¿De A ey, A i / ' . " N i > "un: | = eae i} i y de Al 3 ie En 1 i N % > E RN 3 i Ee a y Q B ~ Ny y Ye } 4 ma. & ie, a f a 7 E en y de 12 ; E - y Bd. 47 S. 31—41 Bonn, September 1997 Neue Glasfliigler (Lepidoptera, Sesiidae) aus dem Pamir und dem Hindukusch K. Spatenka Abstract. Five new species of clearwing moths from Central Asia are described. Bembecia pashtuna sp. nov. from Pakistan, Baluchistan, is closely related to B. dispar and B. tsvetajevi; the larva lives in Astragalus lacei. B. aloisi sp. nov. originates also from Baluchistan and the larva lives in Onobrychis dealbata. It forms part of the B. ichneu- monisformis-group and resembles B. zebo. The same applies to B. karategina Sp. nov. which was collected by a single pair from Tadzhikistan. Its larval bionomy remains unknown. Chamaesphecia mudjahida sp. nov. is only known from a single specimen from Afghanistan (Nuristan). It is similar in wing shape to Ch. mutilata, but is more loosely related to Ch. masariformis. Its larval bionomy remains also unknown. Ch. weidenhofferi sp. nov. from SW-Uzbekistan ist related to Ch. zimmermanni and Ch. ferganae. The larva lives in Phlomis betonicifolia. Key words. Lepidoptera, Sesiidae, clearwing moths, systematics, Pamir, Hindukush. Zusammen mit A. Pavlicko (Prachatice) und Z. Weidenhoffer (Praha) unternahm ich 1992 eine Sammelreise in einige Gebirge Pakistans, in deren Verlauf es gelang, zwel neue Glasflügler-Arten zu entdecken. Für beide konnte auch die Larvalbionomie geklärt werden. Die Beschreibungen beider Taxa ergánze ich um drei weitere Beschreibungen neuer Arten aus Afghanistan, Tadzhikistan und Uzbekistan. Obwohl von den ersten zwei nur sehr wenig Material zur Verfügung steht, habe ich mich entschlossen, diese Arten zu beschreiben, da sie aus Gebieten stammen, bei denen nur sehr geringe Aussichten bestehen, weiteres Material zu erhalten. Bembecia pashtuna sp. n. (Abb. 1, 2) Material. Holotypus ©, Pakistan, Baluchistan, 7 km O Ziarat, ca 150 km N Quetta, 2500—2700 m, ex p. 29. 6. 1992, coll. Museum Witt, Múnchen'; Paratypen: 10 00,4 99, gleicher Fundort, 19.—24. 6. 1992, davon 1 © in coll. A. Kallies; 1 ©, gleicher Fundort, 15. 6. 1992; 3 o ©, gleicher Fundort, 16. 6. 1992, Anflug an Pheromon Nr. 15 (Dr. Priesner); 1 ©, gleicher Fundort, 16. 6. 1992, Anflug an Pheromon Nr. 16 (Dr. Priesner); 4 © ©, gleicher Fundort, 18. 6. 1992, Anflug an Pheromon Nr. 5 (Dr. Priesner), alle K. Spatenka leg. in coll. K. Spatenka und Museum Witt, München; 1 Q, Afghanistan, Hazarajat, Koh-i-Baba, Shah- tu-Kotal, 4000 m, 20.—21. 6. 1961, G. Ebert leg., in coll. Zoologische Staatsammlung, Mün- chen; 1 9? NO-Afghanistan, Badachschan, Khinsch-e-Andarab, 3500—4000 m, 21. 7. 1957, G. Ebert leg., in coll. Zoologische Staatssammlungen, München (bei diesem Exemplar fehlt das Abdomen). _ Männchen. Spannweite 20,0 mm. Kopf: Labialpalpus dorsal schmutzigweiß, 3. Glied apikal schwarz, innen weißlich, außen schwarz mit langen haarförmigen Schuppen; Stirn schwarz mit einzelnen weißlichen Schuppen vor dem Auge; Scheitel schwarz, dicht mit langen weiß- ! Die umfangreichen Sammlungen des Museums Witt in München werden zu gegebener Zeit an die Zoologische Staatssammlung in München transferiert. 32 K. Spatenka lichen Haarschuppen bedeckt; Antenne schwarz mit sehr langen Cilien, ventral mit einzelnen ockergelben subapikalen Schuppen. Thorax: schwarz; Metathorax mit schmalen, langen, schmutziggelben Haarschuppen; Patagia lateral mit einzelnen weißlichen Schuppen. Vorderbein schwarz, Femur mit langen grauen Haarschuppen, Tibia und Tarsus ventral mit ockergelben Schuppen durchsetzt. Hinter- bein schwarz, Coxa und Femur mit grauen Haarschuppen, Tibia ventral mit einzelnen weiß- lichen Schuppen, Tarsus mit ockergelben Schuppen durchmischt, die 1. Tarsomere ventral weißlich, Sporne schwarz. Vorderflügel: braunschwarz mit schwach entwickelten Glasfeldern. Costalrand, das Apikal- feld, Diskalfleck und Cubitalstamm schwarz, Fransen lang, dunkelbraun, Analrand und die Adern MI, M2 im äußeren Glasfeld (ETA) ockergelb bestäubt. Das hintere Glasfeld (PTA) nur rudimentár, das Keilfeldchen auf etwa die Hälfte verkürzt, das ETA breiter als hoch, durch die Adern in 3 Zellen geteilt, etwa so breit wie der Diskalfleck und das Apikalfeld; Costalrand unterseits weißlich, Analrand ockergelb, Diskalfleck, Apikalfeld und die Adern im ETA schwarz; Fransen dunkelbraun, leicht schmutziggelb durchmischt. Hinterflügel: hyalin, die Adern, die langen Fransen, der breite ea sowie der breite keilförmige Diskalfleck dunkel braunschwarz; unterseits von gleicher Färbung, Adern und der Diskalfleck mit sehr spärlichen ockergelben Schuppen. Abdomen: schwarz, die Hinterränder der Tergite 4, 6 und 7 schmal weißlich beschuppt; ventral schwarz, das vierte Sternit mit schmutziggelbem Hinterrand, der aber medial unter- brochen ist, das siebte Sternit mit einzelnen weißlichen Schuppen am Hinterrand; der Anal- busch schwarz, medial mit einzelnen schmutzig gelbweißen Schuppen, Valven gelbweiß. Weibchen. Körper, Flügel, Beine, Antennen und Labialpalpen dunkel braunschwarz; einzelne schmutziggelbe Schuppen auf der Stirn und dorsolateral an den Hinterrändern der Abdomi- naltergite 4 und 6, schwache schmutzigweiße Bestäubung auch dorsal auf den Palpengliedern 122: Variabilität. Spannweite 18,0—22,5 mm. Die Weibchen variieren sehr wenig, die Männchen stärker, einige haben mehr oder weniger ausgeprägte helle Bestäubung. Das Extrem bildet ein Männchen mit schmutzigweißen Beinen (nur Hintertibia mit schwarzem Ring). Bei diesem Stück finden sich auf der Außenseite des Diskalfleckes einzelne ockergelbe Schuppen, das ETA ist länger und breiter, mit einer rudimentären vierten Zelle, die Analader und die Flügel- wurzel am Analrand des Hinterflügels ockergelb bestáubt, die hellen Ringe auf dem Abdomen breiter und nicht nur weißlich, sondern proximal leicht gelblich, der Analbusch in der Mitte deutlich ockergelb durchmischt. Zwischen der typischen und der Extremform der Männchen sind Übergangsformen zu finden. Bei einem Weibchen ist die braunschwarze Bestäubung des Vorderflügels an der Stelle des Keilfeldchens weniger dicht. Genitalapparat © (Abb. 9). Valve mittelbreit, mit gerundetem Apex, die kahle Fläche dorsal einfach-konkav gewölbt, Crista sacculi gerade, schmal, apikal zugespitzt, nicht den Ventral- rand der Valve erreichend, die Dorsalkante im basalen Drittel mit kräftigen Borsten bestanden, die subapikal quer über die ganze Breite der Crista eine Linie bilden; Gnathos einfach, mit löffelförmigen Lateralkanten, Crista medialis schwach entwickelt; Aedeagus sehr schlank. Genitalapparat Q. Nicht untersucht. Differentialdiagnose. Die neue Art ist mit einer Gruppe deserticoler Arten, wie B. dispar, B. syzcjovi, B. tsvetajevi und B. deserticola, verwandt. Hierher gehören wahrscheinlich auch B. tshimgana und B. vidua. Phänotypisch ähnelt das Weibchen der neuen Art am meisten B. dispar und B. tsvetajevi, teilweise auch B. vidua, das Männchen nur den dunklen Formen von B. tshimgana. Das Männchen von B. tshimgana hat eine gelbe Halskrause (bei B. pash- tuna sp. n. schwarz), gelb begrenzte Tegulae (bei B. pashtuna sp. n. schwarz), der Metathorax (bei B. pashtuna sp. n. schwarz) und auch weitere Körperteile sind gelb statt weiß; dagegen ist der Analrand des Vorderflügels bei B. tshimgana braunschwarz und nicht schmutziggelb. Von B. vidua ist nur der Holotypus bekannt, ein einzelnes Weibchen nicht ganz klarer Her- kunft (Samarkand). Dieses Exemplar ist ebenfalls braunschwarz, hat aber teilweise durchsich- tige Vorderflügel und fast ganz durchsichtige Hinterflügel. Trotzdem kann man nicht aus- Neue Glasflügler aus dem Pamir und Hindukusch 33 schließen, daß B. pashtuna nur eine Subspecies von B. vidua darstellt. Diese taxonomische Frage ist jedoch ohne weiteres Material von B. vidua und ohne Kenntnisse der Bionomie dieser Art nicht lösbar. Das Weibchen von B. dispar aus Nordafrika ist im Durchschnitt größer (27 mm) und hat vollständig braunschwarze Beine und ein ebenso gefärbtes Abdomen (bei B. pashtuna mit heller Beimischung); bei B. dispar ist das dritte Palpenglied schwarz und die Stirn rein weiß, bei B. pashtuna schwarz, und die Stirn hat nur einzelne helle Schuppen. Beim Weibchen von B. dispar kommen sehr oft Stücke mit einer kleinen durchsichtigen Fläche zwischen den Adern M3-Cu2 im Hinterflügel vor, bei B. pashtuna ist hingegen die ganze Flügelfläche braunschwarz. Die 9 Q von B. tsvetajevi treten in zwei Formen auf, einer gelben und einer schwarzen. Die letztgenannte Form ähnelt B. pashtuna. Sie unterscheiden sich aber durch die gelbe Stirn sowie die gelbe Halskrause. Ähnlich wie B. dispar ist auch B. tsvetajevi im Durchschnitt größer (27 mm). Bionomie. Die Raupe ist zweijährig und lebt in den dicken Wurzeln des seltenen, bis 1,5 m hohen, gelbblühenden Astragalus lacei, manchmal zahlreiche Raupen in einer einzigen Wurzel (ähnlich wie B. tshimgana oder B. dispar). Sie baut ein Röhrchen aus Wurzelspánen, das innen fein ausgesponnen ist und das manchmal einige Millimeter aus dem Wurzelkopf herausragt. Die Verpuppung findet in einer kleinen Kammer im Proximalende dieses Röhrchens statt. Die Imagines schlüpfen in der zweiten Junihälfte und vermutlich auch noch im Juli. Sie fliegen vormittags (8.30—11.30) an künstliche Sexual-Pheromone an. Habitat. Lichte Juniperus-Wälder und verlassene Gärten an Abhängen mit nördlicher Exposi- tion, unter dem Ziarat-Paß in Baluchistan in Höhen zwischen 2500—2700 m über NN, wo die Futterpflanze sehr vereinzelt vorkommt. Zusammen mit der neu beschriebenen Art fliegt auf den Lokalitäten auch häufig Chamaesphecia xanthosticta (Hampson, [1893]) und vereinzelt B. aloisi sp. n. Verbreitung. Bisher wenig bekannt, wahrscheinlich im ganzen Hindukusch-System verbreitet. Aus sesiidologischer Sicht stehen die Arten Baluchistans in sehr enger Beziehung zu denen des westlichen Pamir-Alai-Gebietes. So fliegt die bemerkenswerte rotorange Form von Bembecia stiziformis tenebrosa (Püngeler, 1914), die bisher nur aus Aman Kutan südlich von Samarkand bekannt war, ebenfalls am Ziarat-Paß. Weiterhin erwies sich die aus Baluchistan beschriebene Sesia ommatiaeformis (Moore, 1891) als identisch mit Sesia shugnana (Sheljuzhko, 1943) aus Chorog (Westpamir). Bisher ist aus Baluchistan kein Vertreter der orientalischen Glasflügler- fauna bekannt geworden. Derivatio nominis. Die Art ist dem Stamm der Pashtunen gewidmet, deren Siedlungsgebiet großräumig mit dem bisher bekannten Areal von B. pashtuna zusammenfällt. Bembecia karategina sp. n. (Abb. 3, 4) Material. Holotypus ©, Tadzhikistan, Karategin Kette, Romit Tal, 38.40° N, 69.10? O, 1500 m, 12. 7. 1981, J. Vanék leg.; 1 Q Paratypus, wie oben; beide in coll. Museum Witt, München. Der Holotypus & und der Paratypus Q wurden in copula gefunden, so daß die Konspezifitát der beiden Exemplare gesichert ist. Männchen. Spannweite 22 mm. Kopf: Labialpalpus dorsal weißgelb, ventral ockergelb, das erste und zweite Glied außen dicht schwarz beschuppt; Antenne schwarz; Stirn braun, vor dem Auge weißgelb; Scheitel schwarz, mit langen ockergelben Haarschuppen; Halskrause ocker- gelb. Thorax: schwarz; Tegulae am Innenrand schmal gelb gesäumt; Scapularfleck weißgelb. Vor- derbein: Coxa dorsal gelb, ventral schwarz; Femur braunschwarz, dorsal reich gelb beschuppt; Tibia ockergelb mit kleinem dunklem Punkt; Tarsus ockergelb. Hinterbein: Coxa und Femur braunschwarz, Tibia ockergelb, distal mit breitem schwarzem Band; Tarsus ockergelb. Vorderflügel: sehr charakteristisch geformt, der Außenrand nicht wie gewöhnlich konkav gewölbt, sondern fast gerade, dadurch der Apex sehr scharf; Costalrand breit schwarz beschuppt; Diskalfleck breit, schwarz, nur am Außenrand schmal ockergelb begrenzt; der Cubitalstamm, die Adern im Apikalfeld und der schmale Außenrand schwarz; der Analrand und das mittelbreite Apikalfeld ockergelb; Keilfeldchen gut entwickelt; PTA kurz; ETA breiter 34 K. Spatenka als hoch, breiter als Apikalfeld, die Adern im ETA gelb (M1, M2) bzw. schwarz (M3). Sie teilen das ETA in 4 leicht ockergelb bestáubte Zellen. Unterseite áhnlich, der Vorderrand dicht ocker- gelb bestáubt. Hinterflügel: durchsichtig, mit schwarzen Adern, schmalem Außenrand und breitem keil- förmigen Diskalfleck, der den Aderstiel M3-Cul erreicht; die Adern und der Analrand unter- seits dicht ockergelb bestäubt. Abdomen: schwarz; das zweite Tergit mit schmalem, zitronengelbem Ring; viertes Tergit in der Distalhälfte zitronengelb; fünftes medial mit einzelnen Schuppen gleicher Farbe; sechstes Tergit fast und siebtes ganz zitronengelb; Analbusch schwarz, medial mit zitronengelbem Strich. Ventral das zweite Sternit distal sehr schmal gelb begrenzt; die Sternite 4—7 mit breiten zitronengelben Bändern. Weibchen. Spannweite 25 mm, sehr dimorph. Antenne in der Mitte braun; Labialpalpus und Stirn zitronengelb; Scheitel ockergelb; Halskrause ockergelb; Tegulae im Prothorakalbereich ockergelb. | Thorax: lateral ganz zitronengelb; Scapularfleck weißgelb; Metathorax breit ockergelb; Vor- derbein zitronengelb; Hinterbein dorsal zitronengelb, Tibia distal mit undeutlichem schwar- zem Band; ventral ist dieses Band kontrastreich und der Tarsus braun. Vorderflügel: Glasfelder deutlich ausgeprägt; Vorderrand, Außenrand, Cubitalstamm, die Adern im ETA, das Apikalfeld und die innere Hälfte des Diskalfleckes dunkelbraun; Analrand ockergelb; Diskalfleck in der Außenhälfte orangegelb. Der Costalrand ventral leicht ockergelb bestäubt; zwischen den Adern im Apikalfeld kurze ockergelbe Strahlen. Hinterflügel: durchsichtig mit dunkelbraunen Adern, schmalem Außenrand und langen Fransen; der Diskalfleck sehr schmal keilförmig, dunkelbraun, ventral ockergelb bestäubt. Abdomen: schwarz; das zweite Tergit in der Distalhälfte zitronengelb, fünftes Tergit mit schmalem zitronengelbem Ring, sechstes Tergit in den hinteren zwei Dritteln zitronengelb; Analbusch zitronengelb, nur lateral schwarz. Sternite 5 und 6 fast ganz, 2 und 4 ganz zitronen- gelb; das dritte Sternit distal schmal zitronengelb umrandet. Variabilität. Unbekannt. Genitalapparat. Nicht untersucht. Differentialdiagnose. Das Männchen ähnelt ein wenig B. zebo aus der Turkestan-Kette (Tadzhikistan und Uzbekistan). Diese Art zeichnet sich aber durch einen konkaven Außenrand des Vorderflügels, ein großes ETA, einen wesentlich schmaleren Diskalfleck des Vorderflügels und einen sehr schwach entwickelten Diskalfleck des Hinterflügels aus. Besonders aufgrund u un nn nn nn Abb. 1—8: Imagines der neuen Glasflügler. — Abb. 1: Bembecia pashtuna sp. n., Holotypus o. Pakistan, Baluchistan, 7 km O Ziarat, ca 150 km N Quetta, 2500—2700 m, ex pupa 29. 6. 1992, leg. K. Spatenka (coll. Museum Witt, München). — Abb. 2: Bembecia pashtuna sp. n., Paratypus Q. Pakistan, Baluchistan, 7 km O Ziarat, ca. 150 km N Quetta, 2500—2700 m, ex pupa 19.—24. 6. 1992, leg. et coll. K. Spatenka (Praha). — Abb. 3: Bembecia karategina sp. n., Holotypus ©. Tadzhikistan, Karategin-Kette, Romit-Tal, 38.40° N, 61.90° O, 1500 m, 12. 7. 1981, leg. J. Vanék (coll. Museum Witt, München). — Abb. 4: Bembecia karategina sp. n., Paratypus Q. Tadzhikistan, Karategin-Kette, Romit-Tal, 38.40° N, 61.90° O, 1500 m, 12. 7. 1981, leg. J. Vanék (coll. Museum Witt, München). — Abb. 5: Bembecia aloisi sp. n., Holotypus ©. Pakistan, Baluchistan, Ziarat-Paß 9 km O Ziarat, 2700 m, ex pupa 23. 6. 1992, leg. K. Spatenka (coll. Museum Witt, München). — Abb. 6: Chamaesphecia mudjahida sp. n., Holotypus ©. Afghanistan, Nuristan: Bashgal, Menalgal, Dandizenor Mts., 2200—3000 m, 27.—29. 5. 1972, leg. C. M. Naumann, coll. Nr. 1217 (coll. ZFMK). — Abb. 7: Chamaesphecia weidenhofferi sp. n., Holotypus ©. Uzbekistan, 20 km SW Guzar, Gumbulak, 38.30° N, 66.21° O, 750 m, 3. 6. 1995, leg. K. Spatenka (coll. Museum Witt, München). — Abb. 8: Chamaesphecia weidenhofferi sp. n., Paratypus 9. Uzbekistan, 20 km SW Guzar, Gumbulak, 38.30° N, 66.21° ©, 750 m, 3. 6. 1995, leg. K. Spatenka (coll. Museum Witt, München). Neue Glasflügler aus dem Pamir und Hindukusch 35 36 K. Spatenka der Flügelform ähnelt B. karategina keiner der bisher bekannten Bembecia-Arten. Das Weib- chen ist durch das schmale braune Apikalfeld und das zitronengelbe vierte Tergit sowie die Fárbung des Abdomens auf der Ventralseite sehr charakteristisch. Bionomie und Habitat. Die Larvalbionomie ist unbekannt; das vorliegende Paar wurde in einer bunten waldsteppenartigen Vegetation am Nordhang der Karategin-Kette im Kafirnagan- Tal, ca 30 km óstlich von Dushanbe, sitzend gefunden. Anmerkung: Der Typenfundort ist schon seit mehreren Jahren wegen des Krieges in Tadzhiki- stan unzugänglich, so daß es kaum Aussicht auf weiteres Belegmaterial gibt. Deswegen habe ich mich entschieden, die neue Art nach nur einem Pärchen zu beschreiben. Gleiche Gründe führen mich auch zur Beschreibung der folgenden Art. Derivatio nominis. Die Benennung erfolgt nach dem Fundgebiet, der Karategin-Kette in Tadzhikistan. Bembecia aloisi sp. n. (Abb. 5) Material. Holotypus ©, Pakistan, Baluchistan, Ziarat-Paß 9 km O Ziarat, 2700 m, 23. 6. 1992 ex p., coll. Museum Witt, München; Paratypen: 1 © gleiche Daten, 18. 6. 1992, Anflug an „tipuliformis“-Pheromone (Wageningen); 2 o ©, 8km O Ziarat, 2650 m, 18. 6. 1992, unspezi- fisch im Anflug an künstliche Pheromone; 1 ©, 7 km O Ziarat, 2600 m, 18. 6. 1992, unspezi- fisch im Anflug an künstliche Pheromone, alles K. Spatenka leg. et coll. Männchen. Spannweite 18—20 mm. Kopf: Antenne schwarz, subapikal mit einzelnen ocker- gelben Schuppen; Labialpalpus blaßgelb, das dritte Glied schwarz durchmischt, das erste und zweite Glied außen schwarz, mit sehr langen Haarschuppen; Stirn weißgelb; Scheitel schwarz und mit dünnen grauen Haarschuppen bedeckt; Halskrause lateral kürzer und gelb, dorsal lang und grau. Thorax: schwarz mit dichten, langen, grauen Haarschuppen; Scapularfleck hellgelb; Pata- gia lateral hellgelb, an der Basis des Vorderflügels ein gelber Haarpinsel; Metathorax mit sehr langen, gelbgrauen, dünnen Haarschuppen. Vorderbein: Coxa schwarz mit langen grauen Haaren, dorsal und apikal hellgelb; Femur schwarz; Tibia und Tarsus hellgelb. Hintercoxa und -femur schwarz mit grauen Haarschuppen; Tibia gelb, nur an der Basis und am Distalrand schmal schwarz; Sporne und Tarsen gelb. Vorderflügel: Costalrand hellgelb; Costalrand, Außenrand und Cubitalstamm schwarz; Diskalfleck schwarz, schmal, nur an der Basis der Adern M1—M3 mit wenigen hellgelben Schuppen; Analrand dicht hellgelb bestäubt; die Adern MI—M3 im ETA hellgelb bestäubt; das Apikalfeld schmal, halb so breit wie das ETA, hellgelb, die Adern schwach schwärzlich bestäubt; Fransen sehr lang, grau, dreimal länger als die Breite des schwarzen Außenrandes, ETA in 5 Zellen geteilt, konkav gewölbt, die erste Zelle zwischen R3 und dem Aderstiel R4—RS5 extrem verkürzt. Unterseite ähnlich gefärbt, nur der Costalrand dicht weißgelb bestäubt. Hinterflügel: durchsichtig; die Adern und der schmale Rand schwarz; Fransen sehr lang, grau; der Diskalfleck schwarz, sehr schmal, keilförmig, erreicht jedoch den gemeinsamen Stiel der Adern M3—Cul. Abdomen: schwarz; das zweite Tergit distal breit hellgelb begrenzt; das vierte in der Distal- hälfte hellgelb; fünftes Tergit mediodorsal mit gelbem Fleck, sechstes und siebtes Tergit fast ganz hellgelb; der Analbusch rundlich, medial und lateral hellgelb, sonst schwarz. Das zweite Sternit am Distalrand breit gelb beschuppt, das dritte schwarz, Sternite 4—7 mit Ausnahme der schmalen proximalen Ränder hellgelb; Analbusch schwarz, mit gelben Schuppen durch- mischt, medial und lateral an der Basis gelb; Valven hellgelb beschuppt. Weibchen. Unbekannt. Variabilität. Bei den 5 bisher bekannten Exemplaren unbedeutend. Genitalapparat o (Abb. 10). Valve mittelbreit, dorsal gleichmäßig konkav gewólbt, Apex stumpf, die kahle Fläche relativ klein, dorsal einfach konkav gewölbt, Crista sacculi schmal, gerade, an der Dorsalkante bis zur Mitte mit kräftigen, stark sklerotisierten Borsten, subapikal Neue Glasflügler aus dem Pamir und Hindukusch 37 Abb. 9—11: Genitalpráparate. — Abb. 9: Bembecia pashtuna sp. n., Paratypus O. Pakistan, Baluchistan, 7 km O Ziarat, ca 150 km N Quetta, 2500—2700 m, ex. p. 15. 6. 1992, leg. K. Spatenka (Gen.práp. O. Gorbunov 93—03). — Abb. 10: Bembecia aloisi sp. n., Paratypus o”. Pakistan, Baluchistan, Ziarat-Paß 9 km O Ziarat, 2700 m, ex pupa 23. 6. 1992, leg. K. Spatenka (Gen.práp. O. Gorbunov 93—02). — Abb. 11: Chamaesphecia weidenhofferi sp. n., Paratypus &. Uzbekistan, 20 km SW Guzar, Gumbulak, 38.30? N, 66.21 O, 750 m, 25. 5. 1995, leg. K. Spatenka (Gen.präp. O. Gorbunov 95—12). Maßstäbe: 0.5 mm. noch eine kleine Gruppe gleicher Borsten (Seten), Apex der Crista sacculi zugespitzt; Gnathos einfach, mit schmalen bumerangförmigen Lateralkanten; Crista medialis schwach entwickelt. 38 K. Spatenka Differentialdiagnose. Habituell áhnelt diese neue Art mehreren anderen Arten der Gattung, so z.B. B. puella, B. zebo, B. zuvandica, B. parthica, teilweise auch B. sirphiformis und B. flavida. Von diesen unterscheidet sie sich aber in der Genitalmorphologie. B. aloisi sp. n. gehört demzufolge in die „ichneumoniformis“-Gruppe (Gnathos und Crista sacculi einfach), während die übrigen ähnlichen Arten der „uroceriformis“-Gruppe (doppelte Gnathos), „scopigera“-Gruppe (gespaltene Crista sacculi) oder „megillaeformis“-Gruppe (doppelte Crista sacculi) zuzuorden sind. Von den zentralasiatischen Arten steht B. zebo aus Tadzhikistan der neuen Art äußerlich am nächsten. Diese lebt ebenfalls als Raupe in Onobrychis, gehört genitalmorphologisch jedoch in die ,,scopigera“-Gruppe. Aber auch habituell unterscheidet sich B. zebo in einigen Details: die gelb gefärbten Körperteile sind ockergelb statt blaßgelb, die Patagia schwarz (bei B. aloisi sp. n. lateral hellgelb); der Diskalfleck des Vorderflügels breit (bei B. aloisi sp. n. schmal); das PTA im apikalen Drittel ockergelb (bei 2. aloisi sp. n. unbeschuppt). Bionomie. Die Raupe hat offenbar eine einjährige Entwicklung und lebt. in der Wurzel der polsterbildenden Onobrychis dealbata (Fabaceae). Das Auffinden der Raupe in den dicken holzigen Wurzeln der stark dornigen, auf den Lokalitäten häufig vorkommenden Pflanze ist extrem mühsam. Wir konnten trotz großer Mühe in 5 Tagen nur eine Raupe und eine Puppe finden. Die Imagines sind im Freiland wahrscheinlich in der zweiten Junihälfte und im Juli zu finden. Ähnlich wie bei anderen Arten der Gattung Bembecia fällt ihre sexuelle Aktivität in die Vormittagsstunden. Habitat. Trockene Abhänge auf felsigem Grund und Felsen im Ziarat-Tal und auf dem Ziarat- Paß in Höhen von 2600 bis 2700 m, wo die Futterpflanze ein dominierender Vertreter der perennierenden Vegetation ist. Onobrychis dealbata bildet hier, ähnlich wie O. cornuta in Transkaukasien oder wie O. echidna in der Turkestan-Kette, Polsterfluren. Derivatio nominis. Die Art ist nach meinem Freund und Begleiter auf der Pakistan-Expedi- tion, Dipl.-Ing. Alois Pavlicko benannt, der mir beim Suchen nach Glasflüglern in Baluchistan und Kohistan sehr geholfen hat. Chamaesphecia mudjahida sp. n. (Abb. 6) Material. Holotypus ©, Afghanistan, Nuristan: Bashgal, Menalgal, Dandizenor Mts. 2200 — 3000 m, 27.—29. 5. 1972, C. M. Naumann leg., Coll.-Nr. 1217 (coll. ZFMK). Beschreibung. Spannweite 20 mm. Kopf: Labialpalpus dorsal braun, auf dem ersten und zwel- ten Glied mit weißlichen Schuppen, auf der Innen- und Ventral-Seite zitronengelb; Antenne braun, nur ventral am Pedicellus mit einigen gelben Schuppen; Stirn braun; Scheitel schwarz; Halskrause hellgelb. Thorax: schwarzbraun, lateral unter dem Vorderflügel mit gelbem Fleck; mit kleinem weiß- gelbem Scapularfleck; Tegulae an der Flügelbasis mit einer Gruppe haarförmiger gelber Schuppen. Vorderbein braun; Dorsalkante von Coxa, Femur und Tibia ventralwärts und die Distalenden der Tarsen hellgelb; Hinterbein braun; Tibia zitronengelb mit braunem Ring; Sporne hellgelb. Vorderflügel: schmal und lang, braun, mit kleinen Glasfeldern; im Vorderflügelapex zwi- schen den Adern R4—RS5 und R5S—MiI ein kleiner blaßgelber Fleck; die Adern im ETA blaß- gelb; PTA sehr kurz; Keilfeldchen auf ca. die Hälfte verkürzt; ETA oval, halb so breit wie das braune Apikalfeld, durch die Adern in drei Zellen geteilt. Der ventrale Costalrand, der Anal- rand, der Cubitalstamm, die Ränder des Diskalflecks und das Apikalfeld im Apex reich blaß- gelb bestäubt. Hinterflügel: unbeschuppt, mit braunen Adern, schmalem Außenrand, langen braunen Fransen und kurzem, breit keilförmigem, braunem Diskalfleck. Abdomen: braunschwarz; Hinterrand von Tergit 2 schmal zitronengelb bestäubt; Tergit 4 zitronengelb; 6 am Hinterrand und 7 auf der gesamten Fläche zitronengelb bestäubt, wenn auch nicht dicht; Analbusch schmal, hellgelb, nur lateral mit einzelnen braunen Schuppen. Die Sternite 2, 4, 6 und 7 fast ganz hellgelb; das dritte Sternit mit schmalem, das fünfte mit breitem gelbem Distalrand. Neue Glasflügler aus dem Pamir und Hindukusch 39 Variabilität. Unbekannt. Genitalapparat. Nicht untersucht. Differentialdiagnose. Ch. mudjahida sp. n. ist wahrscheinlich mit Ch. minor (Staudinger, 1856), Ch. pechi (Staudinger, 1887) und Ch. masariformis (Ochsenheimer, 1808) nahe ver- wandt, unterscheidet sich aber habituell in vielen Merkmalen. Schon auf den ersten Blick ist bei Ch. mudjahida sp. n. die schmale und relativ lange Form der Flügel auffallend. Das ETA ist klein und sein Innenrand konkav gewölbt. Die oben genannten Arten haben kürzere und breitere Flügel und das ETA ist größer mit geradem Innenrand. Eine ähnliche Flügelform wei- sen die Vertreter der Untergattung Chamaesphecia (Typusart: Sphinx empiformis [Denis & Schiffermüller], 1775) auf, keine dieser Arten (mit Ausnahme der Ch. bibioniformis (Esper, 1800) hat freilich den hellen Scapularfleck. Aus dieser Gruppe ähnelt der neuen Art, gerade durch die Flügelform, die auch aus Afghanistan bekannte Ch. mutilata (Staudinger, 1887). Sie unterscheidet sich aber unter anderem durch die oben genannten Merkmale, durch den ganz schwarzen Kopf und die roten Ringe auf dem Abdomen. Bionomie und Habitat. Unbekannt. Derivatio nominis. Die Art ist den afghanischen Unabhängigkeitskämpfern, den Mudjahedin, gewidmet. Chamaesphecia weidenhofferi sp. n. (Abb. 7, 8) Material. Holotypus ©, Uzbekistan, 20 km SW Guzar, Gumbulak, 38.30° N, 66.21° O, 750 m, 3. 6. 1995, K. Spatenka leg., coll. Museum Witt, München; Paratypen: 1 Q, Uzbekistan, Malguzar Gebirge, Badbad, 39.54° N, 67.15° O, 900 m, 9. 6. 1995, K. Spatenka leg. et coll.; 14 00,5 Q 9 mit gleichen Daten wie der Holotypus, 2 @ Paratypen wie Holotypus, aber 28. 5. 1995, K. Spatenka leg. et coll. (davon 1 9 genitaluntersucht). Paratypen in den Samm- lungen K. Spatenka (Praha), Museum Witt (München) und ©. Gorbunov (Moskau). Männchen. Holotypus ©: Spannweite 21 mm; eine mittelgroße rotgelbe Art. Kopf: Labial- palpus dorsal hell ockergelb, ventrolateral auf dem ersten und zweiten Glied mit langen, ocker- gelben und schwarzen Haarschuppen; Stirn glänzend grau, vor dem Auge mit gelbweißem Saum; Scheitel schwarz, mit dichten, hell ockergelben Haarschuppen; Halskrause hell ocker- gelb, lateral weißlich; Antennen schwarz, dicht hell ockergelb überstäubt. Thorax: grau mit starkem grünlichem Glanz, lateral unter dem Vorderflügel mit mehl- weißem Fleck; Innenrand der Tegulae mehlweiß, Meso- und Metathorax distal mehlweiß beschuppt. Vorderbein: Coxa schwarz, Dorsalkante mehlweiß beschuppt, Femur, Tibia und Tarsus mehlweiß. Hinterbein: Coxa und Femur schwarz mit grün-metallischem Glanz, Femur distal mit mehlweißen Schuppen, Tibia schwarzbraun, bei den schmutzigweißen Spornen mit breiten mehlweißen Bändern, Tarsus mehlweiß, nur am Basalglied mit einzelnen dunkel- braunen Schuppen. Vorderflügel: Costalrand schwarz mit grünem Glanz; Costalkante mehlweiß, zwischen den Costaladern mehlweiße Schuppen; Außenrand schmal, braunschwarz beschuppt: Analrand hell ockergelb; alle Glasfelder dicht beschuppt mit hell ockergelben Schuppen, nur zwischen den Medialadern drei durchscheinende, kurze und schmale, mit weißen Schuppen bedeckte Strahlen; Apikalfeld und Adern hell ockergelb; Fransen dreimal so lang wie die Breite des Außenrandes, braun mit mehlweißen Enden. Hinterflügel: durchsichtig, mit braunschwarzem, schmalem Außenrand, mit schwarzen Adern und schwarzem, breit keilförmigem Diskalfleck, der den Aderstiel M3—Cul erreicht. Abdomen: grauschwarz, mit grünem Glanz, das zweite und fünfte Tergit mit schmalem mehlweißem Hinterrand, das dritte Tergit medial am Hinterrand mit mehlweißem Fleck, die Tergite 4, 6 und 7 fast ganz mehlweiß; Analbusch schwarz, sublateral mit zwei breiten Strahlen mehlweißer Haarschuppen; lateral: auf dem zweiten Tergit mit einigen mehlweißen Schuppen, die Tergite 3—7 mit je einem großen mehlweißen Fleck; Sternite 3—7 grau, mit weißen Schup- pen auf den Hinterrändern, auf dem dritten Sternit nur sehr wenigen. 40 K. Spatenka Weibchen. Spannweite 25 mm, robuster, Labialpalpus mehlweiß; Stirn weißlich; Scheitel rost- gelb; Innenseite der Tegulae breit ockergelb gerandet; Thorax mit unzusammenhängender ockergelber Mediallinie; Vorderflügel ockergelb mit zahlreichen orangegelben Schuppen am Analrand und im Apikalfeld; die Adern im ETA und im Apikalfeld mit verstreuten schwarzen Schuppen. Die hellen Partien des Vorderflügels ventral mehlweiß. Abdomen dick, das erste Tergit am Hinterrand mit einzelnen, hell ockergelben Schuppen, drittes Tergit medial am Hinterrand mit hell ockergelbem Fleck, Hinterránder der Tergite 2, 4—6 mit breiten, hell ockergelben Ringen. Sternite 4—7 mit schmalen mehlweißen Ringen. Variabilität. Wenig ausgeprägt, die Grundfärbung des Vorderflügels ist bei einigen Männchen mehlweiß statt hell ockergelb, bei einigen Weibchen hell ziegelrot statt ockergelb; bei einem Weibchen ist das Abdomen dorsal fast ganz mehlweiß bestäubt. Genitalapparat & (Abb. 11). Ähnlich den naheverwandten, ebenfalls in Phlomis spp. lebenden Arten Ch. ferganae Sheljuzhko, 1924, und Ch. zimmermanni (Lederer, 1872). Valve kurz, ebenso die Crista sacculi, am Ventralrand fehlt die Borstengruppe; Saccus kurz, breit, am Ende abgerundet; Aedeagus kurz, breit und gerade; Scopula gut entwickelt; Gnathos ohne Crista medialis, mit sehr schmalen Lateralkanten. Differentialdiagnose. Ch. weidenhofferi sp. n. ist nahe verwandt mit Ch. zimmermanni und Ch. ferganae und ähnelt beiden Arten. Ch. zimmermanni ist u. a. schlanker, hat schmalere Flügel, schwarze Antennen, gut entwickelte Glasfelder, vollständig ockergelbe Hinterbeine. Das Abdomen ist ventral vollständig hell ockergelb und besitzt dorsal nur auf den Tergiten 2, 4, 6-7 (©) oder 2, 4, 6 (2) gleichfarbige Ringe. Die ebenfalls verwandte Ch. ferganae hat teilweise braunschwarz beschuppte Hinterflügel, und das Abdomen trägt nur einen weißlichen Ring. Eine weitere, nahe verwandte Art Ch. margiana (Püngeler, 1912) hat rote Vorderflügel und ein Abdomen mit sehr schmalen weißen Ringen. Bionomie und Habitat. Die Raupe ist offensichtlich einjährig und lebt in den Wurzeln von Phlomis betonicifolia (Lamiaceae). Sie verpuppt sich in einer Kammer im Wurzelhals. Die Flugzeit der Imagines liegt offensichtlich zwischen Mitte Mai und Mitte Juni. Sie sind in den Vormittagsstunden sexuell aktiv. Sie fliegen nur bei Sonnenschein sehr schnell nahe über dem Boden und setzen sich auf die Blätter der Futterpflanze oder auf deren rote Blüten, an denen sie auch saugen. Die Futterpflanze und wahrscheinlich auch diese neue Glasflügler-Art sind auf Hügeln mit sehr spärlicher Halbwüstenvegetation und auf steinigen, trockenen Weideplätzen in Höhen von 500 bis 1000 m über NN weit verbreitet. Auf solchen Plätzen stellt P betonicifolia die ein- zige höhere Pflanze dar und bildet mehrere Quadratkilometer große Bestände. Solche umfang- reichen Biotope wurden westlich von Sherabad (südwestlichstes Uzbekistan) und in niedrige- ren Lagen des Kugitangtau-Gebirges beobachtet, ferner nördlich von Sairob, westlich vom Akrabat-Paß (Tjubere-Oland-Kette), überall in den Hügeln zwischen Dekhanabad und Guzar und im Malguzar-Gebirge nördlich von Samarkand. Derivatio nominis. Die neue Art trägt den Namen meines Freundes, des Lycaenidologen Dipl.- Ing. Zdének Weidenhoffer (Prag), der mich in den letzten zehn Jahren auf allen Expeditionen begleitet und beim Suchen von Glasflüglern sehr unterstützt hat. Danksagung Ich danke Herrn Prof. Dr. D. Podlech (Botanische Staatssammlung München) sehr für die Bestimmung der pakistanischen Futterpflanzen und Herrn Prof. Dr. C. M. Naumann (ZFMK, Bonn) für die leihweise Überlassung des Exemplars von Ch. mudjahida. Dr. O. Gorbunov (Moskau) fertigte die Illustrationen der Genitalpräparate an, wofür ich ihm zu besonderem Dank verpflichtet bin. Herrn A. Kallies (Greifswald), Herrn Prof. Dr. C. M. Naumann und Herrn Dr. D. Stüning (beide ZFMK, Bonn) danke ich für ihre Unterstützung bei der Vorberei- tung des Manuskriptes. Neue Glasflügler aus dem Pamir und Hindukusch 41 Zusammenfassung Es werden fünf neue Sesiiden-Arten aus Zentralasien beschrieben. Bembecia pashtuna sp. nov. aus Pakistan, Baluchistan, ist mit B. dispar und B. tsvetajevi verwandt; die Raupe lebt in Astragalus lacei. B. aloisi sp. nov., ebenfalls aus Baluchistan, lebt in Onobrychis dealbata. Sie gehört zur B. ichneumoniformis-Gruppe und ähnelt B. zebo. B. karategina sp. nov. wurde in einem Pärchen in Tadzhikistan gefangen und ähnelt ebenfalls B. zebo; ihre Larvalbionomie ist noch unbekannt; Chamaesphecia mudjahida sp. nov. ist bisher nur nach einem Stück aus Afghanistan (Nuristan) bekannt. Sie ähnelt durch die Flügelform Ch. mutilata, ist aber eher mit Ch. masariformis verwandt. Auch ihre Larvalbionomie ist unbekannt. Ch. weidenhofferi sp. nov. wurde in SW-Uzbekistan entdeckt. Sie ist mit Ch. zimmermanni und Ch. ferganae ver- wandt. Die Raupe lebt in Phlomis betonicifolia. Literatur Gorbunov, O. (1989): A new species of the genus Bembecia Hübner, 1819 from the Cauca- sus, USSR. — Atalanta, Würzburg 20: 119—123. Gorbunov, O. (1992): Two new species of the genus Bembecia Hübner, 1819 from Middle Asia. — Atalanta, Würzburg 23: 249—253. Sheljuzhko, L. (1924): Zwei neue palaearktische Aegeriiden-Arten. — Dt. ent. Z. Iris 38: 1832185: Sheljuzhko, L. (1935): Drei neue Aegeriiden-Arten aus Zentralasien. — Z. Oster. ent.-Ver. 2095355. Spatenka,K.&Z.LaStüvka (1988): Typen der Glasflügler aus der Staudinger- und Pün- geler-Sammlung im Zoologischen Museum Berlin (Lepidoptera, Sesiidae). — Dt. ent. Z. OSI 339. Spatenka, K. & O. Gorbunov (1992): Vier neue paläarktische Glasflügler (Sesiidae, Lepidoptera). — Entomofauna 13: 377— 393. Spatenka,K.,Z.LaStüvka, O. Gorbunov, I. ToSevski & Y. Arita (1993): Die Systematik und Synonymie der paläarktischen Glasflügler-Arten (Lepidoptera, Sesiidae). — Nachr. ent. Ver. Apollo (N.F.) 14: 81 — 114. Staudinger, O. (1889): Centralasiatische Lepidopteren. — Stettin. ent. Ztg 50: 22. RNDr. Karel Spatenka, Vyletni 362, CR-142 00 Praha 4 — Pisnice, Tschechische Republik. > ehr irs yy ern ri RL: sl cen dl A y q iT} iste PIO £ ee eae eee de mh are e few alta los ARES de AS iy sere Ph thet volte —_ hi Af i \d | ee ts. ee > A yo i x i bs A N x h Nel { ; A ug 4 p \ Ñ = 1 y Ai Xi I 1) o % ; han er) i 1 ' ‘ - E z - h > / qe 2 N + Ñ Ñ f Ji i af b Sy E, E } Bd. 4 $. 43-57 | Bonn, September 1997 Neue Glasfliigler-Arten und Unterarten aus Europa und der Türkei (Sesiidae, Lepidoptera) K. Spatenka Abstract. Nine new species and subspecies of clearwing moths (Sesiidae) are described from Europe and Turkey. Many specimens of Bembecia peterseni sp. n. have been collected from a number of localities in Anatolia (Turkey). It resembles a number of Bembecia species and is closely related to B. iberica from France, Spain and Morocco. Synanthedon hera sp. n. described from several specimens from the Taurus (Turkey) is closely related to S. manni. Dipchasphecia intermedia sp. n. was collected and bred from specimens originating from the Pontus and Taurus Mts (Turkey). It resembles D. lanipes and D. krocha its larva living in Acantholimon sp. Chamaesphaecia kautti from Nigde (Turkey), closely related to Ch. bibioniformis. Chamaesphecia taurica sp. n. also belongs into the same group. Synanthedon stomoxiformis riefenstahli ssp. n. from southern Spain, Synanthedon andrenaeformis tenuicingulata ssp. n. from north-eastern Turkey and from Armenia, Bembecia pavicevici dobrovskyi ssp. n. from the Taygetos Mts (Greece), and Bembecia syzcjovi kappadocica ssp. n. from Kappadokia (Asia Minor) represent new subspecies of species already known from other parts of the Palaearctic region. Key words. Sesiidae, clearwing moths, systematics, Europe, Turkey. Mit dieser Arbeit wird eine Reihe von Beschreibungen neu entdeckter Glasflügler- Arten fortgesetzt. Besonders in den letzten ca. 10 Jahren vermehrte sich die Anzahl der Anhánger dieser bemerkenswerten Schmetterlingsgruppe in ganz Europa, beson- ders aber in Deutschland. Diese Tatsache ist sicher mit der leichteren Zugänglichkeit synthetischer weiblicher Sexualpheromone verbunden, die auch den Entomologen, die bisher nie im Leben einen Glasflügler gesehen haben, umfangreichere Ausbeuten ermöglicht. Als Folge dieser intensiven Sammelaktivitát wurden viele neue Arten entdeckt. Leider führt dies aber auch dazu, daß dem Studium der Bionomie der Arten weniger Aufmerksamkeit gewidmet wird. Synanthedon stomoxiformis riefenstahli ssp. n. (Abb. 1) Material. Holotypus ©, Spanien, Provinz Alicante, Orcheta, 500 m, 18.—22. 5. 1993, H. Rie- fenstahl leg., coll. Museum Witt, Miinchen!'; Paratypen: 1 © mit gleichen Daten, 4 0,19, Spanien, Provinz Alicante, Relleu, 21.—23. 5. 1993, H. Riefenstahl leg. et coll. Differentialdiagnose. Spannweite 28 mm. Von der Nominatform wie folgt unterschieden: Apikalfeld sehr breit, breiter als das Rundfeldchen (ETA) (sehr reduziert in der Nominat- form), schwarzbraun; Diskalfleck des Hinterflügels wesentlich breiter; die Beine schwarz mit grünviolettem Glanz (Tarsen bei der Nominatform hellbraun); Thorax vollkommen schwarz - (Innenrand der Tegula bei der Nominatform rot). Eine ähnliche Unterart ist aus Kleinasien und Transkaukasien bekannt: S. s. amasina (Stau- dinger, 1856). Bei dieser Unterart ist die Tegula breit rot begrenzt, und nur die Abdominal- sternite 4 und teilweise 5 sind rot, bei S. s. riefenstahli ssp. n. und S. s. stomoxiformis (Hübner, ! Die umfangreichen Sammlungen des Museums Witt in München werden zu gegebener Zeit an die zoologische Staatssammlung in München transferiert. 44 K. Spatenka 1790) sind die Sternite 4—7 rot. Die Raupe der Subspecies amasina lebt in den Wurzeln von Rhamnus palasi (Rhamnaceae). Bionomie und Habitat. Die Raupe ist wahrscheinlich zweijährig und lebt nach den Angaben des Entdeckers in den Wurzeln von Rhamnus cathartica (Rhamnaceae) und Crataegus sp. (Rosaceae). Verbreitung. Bisher nur von der Typenlokalitát bekannt. Derivatio nominis. Die neue Unterart ist ihrem Entdecker gewidmet. Synanthedon andrenaeformis tenuicingulata ssp. n. (Abb. 2) Material. Holotypus ©, Türkei, Avcı Dagları, Girlevik, 39.35° N, 39.44° O, 17. 6. 1993, T. Dobrovsky leg., coll. Museum Witt, München; Paratypen: 1 o, 1 -Q dito, VI. 1993 ex pupa, T. Dobrovsky leg. et coll., 1 © Türkei, Provinz Kars, Sarikamis, 2000 m, 21.—23. 7. 1992, A. Kallies leg. et coll. Männchen. Spannweite 20 mm. Antenne schwarz; Labialpalpus schwarz, unten ockergelb; Kopf und Thorax schwarz; der Vorderflügel mit gut entwickelten Glasfeldern; ETA breiter als das schwarze Apikalfeld, außen konkav gewölbt; der breite Diskalfleck, die Adern und Ränder schwarz. Ventral: Apikalfeld dunkel; der Vorder- und Hinterrand mit hellen, schmutziggelben Schuppen bestäubt. Hinterflügel durchsichtig mit breitem, keilförmigem schwarzem Diskal- fleck und schwarzem Außenrand; Adern, Fransen und Beine schwarz, nur die Tarsen dorsal mit verstreuten ockergelben Schuppen, ventral ockergelb. Abdomen schwarz mit sehr schma- len, hell ockergelben Ringen auf den Tergiten 2 und 4, lateral schwarz mit hell ockergelben Flecken auf den Tergiten 1 und 2, ventral schwarz mit schmutziggelben Sterniten 4—6; Anal- busch schwarz, Valven dunkel ockergelb bestäubt. Weibchen. Ähnelt sehr dem Männchen; ETA schmäler, Diskalfleck des Vorderflügels noch breiter und Beine dunkler; auf dem Abdomen ist nur die Distalhälfte von Sternit 4 schmutzig- weiß. Differentialdiagnose. Die neue Unterart unterscheidet sich in mehreren Merkmalen von der Nominatform: ETA breiter und außen konkav gewölbt (bei der Nominatform sehr schmal und trapezförmig mit geradem oder leicht konvexem Außenrand); Apikalfeld schwarz (bei der Nominatform stets mit mehr oder weniger ausgeprägten hellen Strahlen zwischen den Adern); Diskalflecke auf beiden Flügeln wesentlich breiter als bei der Nominatform; auch die Beine _— _ _a_a_ __ zo e EEE er... Abb. 1—10: Imagines der neuen Glasflügler (Belege in Coll. Museum Witt, München). — Abb. 1: Synanthedon stomoxiformis riefenstahli ssp. n., Holotypus ©. Spanien, Provinz Ali- cante, Orcheta 500 m, 18.—22. 5. 1993, H. Riefenstahl leg. — Abb. 2: Synanthedon andrenae- formis tenuicingulata ssp. n., Holotypus ©. Türkei, Avcı Dagları, Girlevik, 39.35 °N, 39.44 °O, 17. 6. 1993, T. Dobrovsky leg. — Abb. 3: Bembecia peterseni sp. n., Holotypus ©. Türkei, Nigde, Ala-Dag, Demirkazik, 1800 m, 4. 8. 1991, P. Kautt leg. — Abb. 4: Bembecia pavicevici dobrovskyi ssp. n., Holotypus ©. Griechenland, Peloponnes, Taygetos Gebirge, Kardamili, 36.31°N, 23.17%0, 15. 7. 1992, T. Dobrovsky leg. — Abb. 5: Bembecia syzcjovi kappadocica ssp. n., Holotypus ©. Türkei, Kappadokien, Yesilóz, 1300 m, larva 7. 6. 1993, ex. 1. Ende August 1993, K. Spatenka leg. — Abb. 6: Bembecia syzcjovi kappadocica ssp. n., Paratypus 9. Türkei, Kappadokien, Yesilöz, 1300 m, larva 7. 6. 1993, ex. 1. Ende August 1993, K. Spatenka leg. — Abb. 7: Synansphecia hera sp. n., Holotypus O. Türkei, Osttaurus, 1 km S Pürem Paß (N von Kahramanmaras), 37.58°N, 36.350, 1500 m, 30. 6.—1. 7. 1991, K. Spatenka leg. — Abb. 8: Dipchasphecia intermedia sp. n., Holotypus ©. Türkei, Taurus- Gebirge, 10 km N Saimbeyli (Hadjin), 1450 m, 36.07.12°O, 38.03.12 °N, 6. 6. 1993 ex larva ex Acantholimon sp., K. Spatenka leg. — Abb. 9: Chamaesphecia kautti sp. n., Holotypus ©. Türkei, Nigde, Ala Dag, Demirkazik, 25.—26. 7. 1991, 1600 m, P. Kautt leg. — Abb. 10: Cha- maesphecia taurica sp. n., Holotypus oc’. Türkei, Südanatolien, Taurus-Gebirge, Camlıyayla (Namrun), 1200 m, 37.11°N, 34.390, 26.—27. 7. 1991, K. Spatenka leg. 45 Neue Glasflügler aus Europa und der Türkei 46 K. Spatenka und das Abdomen sind dunkler; Analbusch schwarz, ohne die für die Nominatform typischen gelben Schuppen; der gelbe Ring auf Tergit 4 ist vergleichbar schmal wie auf Tergit 2 (bei der Nominatform stets deutlich breiter, besonders beim Weibchen); Sternite 4—6 (bei Weibchen nur 4) sind schmutzigweiß (nicht rein weiß wie bei der Nominatform). Larvalbionomie. Wie bei der Nominatform; die Raupe lebt zweijährig in den Stämmen und Asten von Viburnum lantana (Caprifoliaceae) in lichten Wäldern. Verbreitung. Bisher nur aus Ostanatolien bekannt (Umgebung von Erzincan und Kars) und aus Armenien (Naturschutzgebiet Chosrov bei Vedi). Das Areal umfaßt wahrscheinlich die ganze Osttiirkei und das südliche Transkaukasien. Unklar bleibt die Frage der Existenz einer Kontaktzone zwischen den beiden bisher bekannten Unterarten. Die Südostgrenze der Verbrei- tung der Nominatform ist wenig bekannt, die südöstlichsten Lokalitäten liegen bei Sofia (Bul- garien) und in Südrußland (Kasan, Simbirsk, Ufa, Orenburg, Samara, Saratow, Sarepta, Uralsk). Bartel (1902) führt auch Kleinasien an, ohne daß hierfür Belegmaterial nachweisbar ist. Derivatio nominis. Die Benennung erfolgt aufgrund der besonders schmalen Abdominal- cingulation. Bembecia peterseni sp. n. (Abb. 3) Material. Holotypus ©, Türkei, Nigde, Ala-Dag, Demirkazik, 1800 m, 4. 8. 1991, P. Kautt leg., coll. Museum Witt, München; Paratypen: 7 © co mit gleichen Daten; 8 o © gleiche Angaben, 25.—26. 7. 1991; 3 © ©, Türkei, Kappadokien, Goreme, 7. 8. 1991; 1 ©, Türkei, Adana, Saimbeyli, 1400 m, 10. 8. 1991; 5 oo, Türkei, Bingöl/Mus, Buglan-Paß, 1600 m, alle P. Kautt leg., coll. Museum Witt, München; I © Türkei, Prov. Tunceli, 15 km N Tunceli, 13.—14. 7. 1992, M. Petersen leg., coll. K. Spatenka (GU Nr. 83-393 E. Blum); 1 © mit glei- chen Daten in coll. Kallies; 8 © © Türkei, Kappadokien, Göreme, 24. 6. 1992, H. Riefenstahl leg. et coll.; 1 © Türkei, Provinz Malatya, Kale, 700 m, 12. 7. 1992, A. Kallies leg. et coll., 1 oO Türkei, € Taurus, Sartauul gec. 1650 m, 36:55 ° N, 33:16:25 > E 222 73577531995 K. Spatenka leg. et coll.; 3 or Türkei, Prov. Erzurum, Karagóbek, 1970 m, 40.10.40? N, 41.26? E, 17. 7. 1996, K. Spatenka leg. et coll., 1 © Türkei, Prov. Agri, Arasgünei Del, Tahir, 2250 m, 39.519 N, 42.25.40” E, 14. 7. 1996, K. Spatenka leg. et coll., 13 or Türkei, Prov. Kars, Kar- gapazari Del., Tasoluk, 1700 m, 40.16? N, 41.35.40" E, 14.—15. 7. 1996, K. Spatenka leg. et coll. Männchen. Spannweite 16—22 mm. Eine schlanke Art mit relativ schmalen Vorderflügeln. Labialpalpus weißgelb, lateral außen schwarz; Antenne schwarz- braun, subapikal heller, ven- tral braun; Stirn weißgelb; Scheitel schwarz mit ockergelben Haaren; Halskragen ockergelb. Thorax: schwarz; Patagia lateral ockergelb; Scapularfleck weißgelb; Innenrand der Tegula ockergelb; Metathorax distal zitronengelb. Vorderbein schwarz, -coxa dorsal in ganzer Länge hellgelb, -femur schwarz, ventral hellgelb, -tibia hellgelb, basal und apikal mit schwarzen Punkten, Tarsus hellgelb; Hinterbein: Coxa und Femur schwarz mit langen gelben Haaren, Tibia basal und apikal mit schwarzen Ringen, sonst hellgelb, Tarsus hellgelb mit vereinzelten schwarzen Schuppen, ventral mit schwarzen Dornen. Vorderflügel: ockergelb mit gut entwickelten Glasfeldern; Costalrand braunschwarz mit hell ockergelben Schuppen zwischen den Adern, Außenrand schmal braunschwarz, Fransen lang braunschwarz, Analrand orangegelb beschuppt, Cubitalstamm braunschwarz, Basalhälfte des Diskalflecks dunkelbraun, Apikalhälfte dunkel ockergelb bis orangegelb, halb so breit wie das Rundfeldchen, Apikalfeld etwa so breit wie das Rundfeldchen, hell ockergelb mit schmal, aber deutlich dunkelbraun beschuppten Adern. Das Keilfeldchen bis zur schwarzen Flügelwurzel reichend, das Längsfeldchen apikal durch die ockergelbe Bestäubung auf etwa 2/3 der Lange reduziert, das Rundfeldchen klein, etwa so hoch wie breit, durch die ockergelb beschuppten Adern in 3 Zellen geteilt, die Räume zwischen den Adern R3—R4, R4—R5, M3—Cul und Cul—Cu2 hell ockergelb beschuppt. Bei frischen Exemplaren sind die Glasfelder mit durch- sichtigen, stark perlmuttartig glänzenden Schuppen belegt. Ventral ist auch auch der Costal- und Analrand des Vorderflügels hell ockergelb. Hinterflügel: durchsichtig mit sehr schmalem dunkelbraunem Außenrand, dunkelbraunen Adern und Fransen und kurzem, breit-keilförmigem Diskalfleck, der zwischen den Adern M2 und dem Aderstiel M3—Cul nur aus einer einzigen Schuppenreihe besteht. Neue Glasflügler aus Europa und der Türkei 47 Abdomen: schwarz, mit deutlichen, breit zitronengelben Bándern, Hinterrand von Tergit 1 mit undeutlichem gelbem Ring, Tergit 2 in der Distalhälfte gelb, 3 mit schmalem, aber deut- lichem gelbem Hinterrand, 4 in der Distalhálfte, 5 in den distalen 2/3, 6 fast ganz und 7 ganz gelb. Die gelben Bänder auf den Tergiten 4—7 tragen an den Distalrándern je eine Reihe weiß- gelber Schuppen; Analbusch schwarz, medial und lateral zitronengelb. Abdomen lateral áhn- lich wie dorsal, ventral schwarz, die Sternite 2—7 mit gleichmäßig breiten weißgelben Distal- rändern. Weibchen. Antenne ockergelb mit schwarzer Spitze; Labialpalpen zitronengelb. Patagial- kragen, Innenrand der Tegula und Metathorax breit zitronengelb. Abdomen schwarz, mit sehr breiten zitronengelben Bändern auf den Tergiten 2, 4, 5 und 6, mit breitem, aber kurzem medialen Querband, auch auf dem 3. Tergit, ventral schwarz mit breiten ockergelben Bändern auf den Sterniten 4—6. Vorderrand des Vorderflügels schwarzbraun, lebhaft ockerfarbig bis orange, ebenso wie das Apikalfeld und der Diskalfleck; ETA rundlich, etwa so breit wie das Apikalfeld, in 4 Zellen geteilt, Adern im Apikalfeld schwach braunschwarz beschuppt, Diskal- fleck am Innenrand braunschwarz, alle Glasfelder gut entwickelt. Hinterflügel mit sehr kur- zem und schmalem, nur bis M2 führendem Diskalfleck. Das Weibchen ähnelt den Weibchen von B. fibigeri und B. albanensis, unterscheidet sich aber in vielen Details. B. albanensis hat z. B. kleinere Glasfelder, alle Abdominalránder sind schmaler (12—T6), das PTA beschuppt und die Adern im Apikalfeld sind dicht dunkel bestäubt. B. fibigeri hat schwarze Antennen mit weißlichem subapikalem Fleck, die Tegula sehr schmal gelb begrenzt, die Abdominal- ränder 1 und 3 sehr undeutlich, die Hintertibia fast ohne dunkle Schuppen (bei B. peterseni mit deutlichem schwarzem distalem Fleck), den Costalrand des Vorderflügels ockergelb, braun bestäubt, die dunklen Adern im Apikalfeld sehr deutlich, den Diskalfleck im Hinterflügel breiter und länger (bis M3—Cul). Variabilität. Bei einigen, besonders bei leicht abgeflogenen Exemplaren sind die gelben Abdo- minalbánder schmäler (besonders auf Tergit 3), und der undeutliche gelbe Ring auf Tergit 1 fehlt oft. Bei einigen Exemplaren ist der Diskalfleck des Vorderflügels breiter oder sogar ganz gelb oder orangegelb. Bei Exemplaren mit orangegelbem Diskalfleck ist auch das Apikalfeld orangegelb. Genitalapparat © (Abb. 11). Individuell sehr variabel, im allgemeinen aber ähnlich dem von B. iberica Spatenka, 1992; Valve mittelbreit, mit sehr schmaler bedornter Fläche; Crista sacculi gespalten, ihr dorsaler Processus entweder mächtig wie bei B. iberica oder sehr kurz; Gnathos doppelt, mit ovalen oder langgezogenen Lateralkanten, die cranialen lateralen Cristen ent- weder in einer Höhlung verschmolzen oder proximal stark genähert, jedoch nicht ganz ver- schmolzen. Differentialdiagnose. Die neue Art ist phylogenetisch mit B. iberica verwandt, ähnelt aber auch weiteren Arten der Gattung Bembecia, wie z. B. B. zuvandica Gorbunov, 1987, B. fibigeri Lastüvka € Lastúvka, 1994, B. ichneumoniformis ([Denis & Schiffermüller], 1775), B. alba- nensis (Rebel, 1918) und B scopigera (Scopoli, 1763). B. ichneumoniformis und B. scopigera unterscheiden sich deutlich durch das große Rundfeldchen (5—6 Zellen), schmale gelbe Abdo- minalringe und weitere Merkmale. Die genitalmorphologisch verwandte B. iberica unterschei- det sich habituell durch den teilweise ziegelroten Diskalfleck, das rostbraune Apikalfeld und den rötlichen Analrand des Vorderflügels. Sie hat auch die Räume zwischen den Adern R3—R4 und Cul—Cu2 schwarz. B. albanensis hat bei einigen Exemplaren ähnlich gefärbte Vorderflügel und ähnlich geformte Glasfelder, die Hinterränder der Tergite sind jedoch nur schmal und blaßgelb. B. zuvandica ist noch heller gelb, hat den Analrand des Vorderflügels hell ockergelb beschuppt und die medialen Adern im Apikalfeld gelb und nicht schwarz beschuppt. Sie hat auch ein größeres Rundfeldchen, das breiter als das Apikalfeld ist und in 4 statt 3 Zellen geteilt ist. Sehr ähnlich ist auch die spanische und südfranzösische B. fibigeri, die nur ein wenig robuster ist, der Farbton der Beine und Abdominalringe ist satter, das Api- kalfeld leicht mit schwarzen Schuppen durchmischt und das Rundfeldchen größer (4 Zellen). B. fibigeri unterscheidet sich ferner ganz deutlich im Bau des männlichen Genitalapparates, der eine einfache Gnathos und eine ungespaltene Crista sacculi besitzt. 48 K. Spatenka Bionomie und Habitat. Die Raupe ist einjáhrig und lebt offensichtlich in den Wurzeln einer Ononis sp. (Fabaceae), die der mitteleuropáischen O. spinosa ähnelt. Sie legt ihre Gänge dicht unter der Rhizodermis an. Vor der Verpuppung baut sie aus miteinander versponnenen Spá- nen des Wurzelstockes ein manchmal mehrere cm langes Röhrchen. Die Imago schlüpft im Juli und August. Die Art bewohnt trockene Ruderalflächen, z. B. Weideplätze, Straßenränder, Obstgärten und die Umgebung von Bewässerungsanlagen und Brunnen. Verbreitung. Im mittleren und östlichen Anatolien anscheinend weit verbreitet. Derivatio nominis. Die neue Art ist Herrn M. Petersen (Pfungstadt), einem ausgezeichneten Sesienkenner, gewidmet. Bembecia pavicevici dobrovskyi ssp. n. (Abb. 4) Material. Holotypus ©, Griechenland, Peloponnes, Taygetos Gebirge, Kardamili, 36.31° N, 23.17° O, 15. 7. 1992, T. Dobrovsky leg., coll. Museum Witt, München; Paratypen: 2 o © dito, 13. 7. 1992; 1 © 15. 7. 1992, alles T. Dobrovsky leg. et coll. (Praha); 1 ©, dito, 12. 7. 1992, Krejcik leg. et coll. (Mlada Boleslav); 1 ©, Griechenland, Peloponnes, Taygetos, Tseria, 36.53 ° N, 22.13° O, 14. 7. 1992, T. Dobrovsky leg. et coll.; 3 © ©, Taygetos, Umg. Saidona, 500 m, 9. 7. 1994, A. Lingenhöle leg., coll. K. Spatenka; 10 © © mit gleichen Daten, 17. 7. 1995, A. Lingenhöle leg. et coll.; 10 © ©, Taygetos, 15 km W Mistras, 400 m, A. Lingenhöle leg. et coll.; 8 oo, Taygetos, Schlucht bei Exohori, 450 m, 22. 7. 1995, A. Lingenhöle leg., coll. K. Spatenka; 30 © © mit gleichen Daten in coll. A. Lingenhöle. Männchen. Spannweite 24—27 mm. Das Männchen der neuen Unterart unterscheidet sich in folgenden Merkmalen von B. p. pavicevici ToSevski, 1989: der schmutzigweiße Fleck auf der Antenne nimmt nur !/s der Antennenlänge ein (bei der Nominatform mindestens ein Drit- tel); die Flügel sind deutlich dunkler bestäubt. Vorderflügel: das Längsfeld (PTA) ist durch die dunkelbraune Bestäubung stark reduziert, der Analrand deutlich breiter, der Diskalfleck dunkelbraun, nur mit dunkel ockergelbem Außenrand (bei der Nominatform ist die äußere Hälfte des Diskalfleckes ockergelb), das Rundfeldchen (ETA) ist sehr klein, 3-zellig, oval, halb so breit wie das Apikalfeld, dicht mit ockergelben und perlmuttartigen Schuppen bestáubt (bei der Nominatform 4—5 zellig, nur 1,5 x schmaler als das Apikalfeld); das Apikalfeld ist dunkelbraun, nur mit 3 kurzen dunkel ockergelben Strahlen zwischen den Adern RS—M1, M1—M2 und M2—M3 (bei B. p. pavicevici ist das Apikalfeld vorwiegend ockergelb). Auch der Hinterflügel ist dunkler, mit sehr breitem dunkelbraunem Außenrand, breiter als die Länge der Fransen (bei der Nominatform halb so breit wie die Fransen). Auch der Diskalfleck des A A A A e A, E A 0) __O_--, Abb. 11—17: Genitalpráparate. — Abb. 11: Bembecia peterseni sp. n., Paratypus ©. Türkei, Nigde, Ala-Dag, Demirkazik, 1800 m, 4. 8. 1991, P. Kautt leg. (Gen.práp. O. Gorbunov 95-15). Maßstab: 0.5 mm. — Abb. 12: Synansphecia hera sp. n., Paratypus ©. Türkei, Osttaurus, 1 km S Púrem Paß (N von Kahramanmaras), 37.58 N, 36.350, 1500 m, 30. 6.—1. 7. 1991, K. Spatenka leg. (Gen.präp. O. Gorbunov 93-05). Maßstab: 0.5 mm. — Abb. 13: Dipchasphe- cia intermedia sp. n., Paratypus ©. Türkei, Taurus-Gebirge, 10 km N Saimbeyli (Hadjin), 1450 m, 36.07.12°O, 38.03.12 °N, 6. 6. 1993 ex larva ex Acantholimon sp., K. Spatenka leg. (Gen.práp. I. ToSevski 822). — Abb. 14: Dipchasphecia intermedia sp. n., Paratypus 9. Türkei, Taurus-Gebirge, 10 km N Saimbevli (Hadjin), 1450 m, 36.07.12°O, 38.03.12°N, 6. 6. 1993 ex larva ex Acantholimon sp., K. Spatenka leg. (coll. Gen.präp. O. Gorbunov 96-03). Maßstab: 0.5 mm. — Abb. 15: Chamaesphecia kautti sp. n., Paratypus O. Türkei, Nigde, Ala- Dag, Demirkazik, 25.—26. 7. 1991, 1600 m, P. Kautt leg. (Gen.práp. O. Gorbunov 93-07). Maßstab: 0.5 mm. — Abb. 16: Chamaesphecia kautti sp. n., Paratypus 9. Türkei, Nigde, Ala- Dag, Demirkazik, 20. 7. 1996, K. Spatenka leg. et coll., 1600 m, (Gen.práp. O. Gorbunov 96-06). Maßstab: 0.5 mm. — Abb. 17: Chamaesphecia taurica sp. n., Paratypus ©. Türkei, Südanatolien, Taurus-Gebirge, Camliyayla (Namrun), 1200 m, 37.11 9N, 34.390, 26. 7. 1991, K. Spatenka leg. (Gen.präp. I. ToSevski 762). Neue Glasflügler aus Europa und der Türkei 49 Hinterflügels ist kräftiger, breit keilförmig und ohne gelbe Schuppen (bei der Nominatform schwächer, zwischen den Adern M2—M3 sehr schmal und in der Außenhälfte mit hell ocker- gelben Schuppen). Weibchen. Spannweite 26—28 mm; ebenfalls dunkler und robuster als die Nominatform; der Analrand des Vorderflügels mehr rötlich; die Antenne vorwiegend gelb, nur an der Basis und der Spitze dunkler; mit breiten zitronengelben Bändern auf den Abdominaltergiten 2—6; der Analbusch mit 2 länglichen gelben Strahlen. Es ähnelt dem Weibchen der Individualform „kalavrytana“ von Bembecia albanensis (Rebel, 1918), unterscheidet sich aber in einigen wich- tigen Merkmalen. Der Scheitel ist bei ,,kalavrytana“ rostbraun, bei B. p. dobrovskyi ssp. n. schwarz mit gelben Haaren; die Glasfelder sind bei „kalavrytana“ noch stärker reduziert, ETA und PTA fehlen, das ATA ist sehr schmal und kurz. Der Hinterflügel ist bei „kalavrytana“ - Sh 9 a ANS ER RR ERE ESOS E li, NN N (HZ BE: iy N 50 K. Spatenka im Apex sehr breit schwarz bestäubt, der Diskalfleck rot und endet auf dem Aderstiel M3—Cul, bei B. p. dobrovskyi ssp. n. hingegen auf der Verzweigung von M3—Cul. Kleinere Unterscheidungsmerkmale weisen auch die männlichen Genitalien auf. Die für B. p. pavicevici typische kahle subapikale Partie auf der Crista sacculi fehlt bei der neuen Unterart, der ganze Dorsalrand ist hier dicht mit dicken, stark sklerotisierten Seten besetzt. Diese bedecken basal den Ventralrand des mit Seten bestandenen Teiles der Valve (diese Seten fehlen bei B. p. pavicevici). Crista sacculi subapikal breit durch eine Reihe dicker Borsten mit der bedornten Fläche der Valve verbunden. Diese Verbindung ist bei B. p. pavicevici sehr schwach und wird durch dünnere Borsten gebildet. Variabilität. Vereinzelt wurden in kühlen Schluchten auch ganz schwarze Exemplare gefangen. Bionomie und Habitat. Die Larvalbionomie ist unbekannt. Die Raupe lebt offensichtlich, ähnlich wie die von B. p. pavicevici, in den Wurzeln von Coronilla emerus (Fabaceae). Trockene Waldsteppen und Schluchten in südlichem Taygetos-Gebirge. Verbreitung. Bisher nur von den südlichen Abhängen des Taygetos-Gebirges bekannt. Die Nominatform ist bisher vom Peloponnes nicht bekannt. Ihr bisher bekanntes Verbreitungs- areal zieht sich vom Parnassos-Gebirge in Südgriechenland, über Mazedonien, SW-Bulgarien, die Inseln Korfu und Hvar entlang der dalmatinischen Küste bis nach Istrien und Slowenien. Sie kommt vermutlich auch in Albanien und in NO-Italien vor, von wo Belegmaterial bisher freilich fehlt. Die Art kommt auch auf dem Nord-Peloponnes bei Kalavrita vor (mehrere Exemplare leg. Lingenhöle, 1994). Die Exemplare aus diesem Gebiet unterscheiden sich sowohl von denen aus dem Taygetos als auch von denen vom Festland und von den dalmatini- schen Inseln. Die schwarzen Zeichnungen der Fligel sind etwas breiter als bei der Nominat- form, der helle Fleck auf der Antenne ist von normaler Breite. Auffallend ist die dichte dunkelockergelbe Bestäubung des Apikalfeldes und des Hinterrandes des Vorderflügels. Mit Schuppen gleicher Farbe sind auch alle Glasfelder des Vorderflügels bestäubt. Die taxonomi- sche Stellung dieser Form bleibt bisher unklar. Ähnliche Merkmale (rotorange bestäubte Glas- felder) weist die aus dem gleichem Gebiet von Sheljuzhko beschriebene Form kalavrytana von B. albanensis auf. Derivatio nominis. Die neue Unterart ist nach ihrem Entdecker benannt. Bembecia syzcjovi kappadocica ssp. n. (Abb. 5, 6) Material. Holotypus ©, Türkei, Kappadokien, Yesilöz, 1300 m, larva 7. 6. 1993, ex 1. Ende August 1993, K. Spatenka leg., coll. Museum Witt, München; Paratypen: 4 00,3 Q 9 mit gleichen Angaben, coll. Museum Witt, München; 1 ©, 1 Q mit gleichen Daten in coll. O. Gorbunov; 2 © © Türkei, Kappadokien, Góreme, 24.—31. 8. 1994, 1000 m, A. Lingenhöle leg., coll. K. Spatenka; 10 0 0,5 9 9 gleicher Fundort, 9. 1995, A. Kallies leg. et coll. Männchen. Spannweite 27 mm. Unterscheidungsmerkmale von der Nominatform: Antenne hellbraun mit schwarzer Spitze (in der Nominatform schwarz); Patagia gelb, nur in der Mitte mit einzelnen schwarzen Schuppen (bei der Nominatform schwarz, nur lateral mit einzelnen gelben Schuppen); der Diskalfleck des Hinterflügels erreicht nicht wie bei der Nominatform den Aderstiel M3 —Cul; der schwarze Fleck auf der Hintertibia undeutlich oder fehlend. Auf- fallend stärker gelb bestäubt ist das Abdomen: die Tergite 2—3 sind distal breit gelb gerandet und die Tergite 4—7 ganz gelb (bei der Nominatform sind nur die Tergite 4, 6 und 7 gelb, 2 und 5 nur mit sehr schmalen gelben Ringen und 3 ganz schwarz). Sternit 2 ist breit gelb gerandet (bei der Nominatform schwarz). Weibchen. Sehr vom Männchen verschieden; Antenne, Labialpalpus, Kopf, Thorax und beschuppte Teile der Flügel schwarz; Stirn schmutziggelb; Vordertibia ventral mit braunen Schuppen; Vorderrand des Vorderflügels ventral mit schmutziggelbem Strich; Vorderflügel ähnlich wie beim Männchen mit extrem großen Glasfeldern, Apikalfeld schmal, der Analrand und die Adern freilich schwarz und der Diskalfleck breiter. Hinterflügel unbeschuppt, mit schwarzem Außenrand, Adern und Fransen, der Diskalfleck schwarz, den Aderstiel M3—Cul erreichend. Das Abdomen ist schwarz, die Tergite 4—6 mit Ausnahme der schmalen Vorder- ränder zitronengelb, die hinteren 2/3; der Sternite 4—6 zitronengelb. Neue Glasflügler aus Europa und der Türkei 51 Variabilität. Bei einigen Männchen ist die Antenne ganz schwarz, bei einem Weibchen die Abdominaltergite zitronen- und die Sternite orangegelb, bei einem Stück sogar auch die Ter- gite orangegelb. Bionomie und Habitat. Die Raupe lebt in der Wurzel einer hochwachsenden, gelbblühenden Astragalus sp. (Fabaceae), die bisher noch nicht determiniert werden konnte. Sie ähnelt ande- ren, ebenfalls von Glasflüglern präferierten Astragalus-Arten, wie A. sieversianus, A. tricho- calyx, A. finitimus, A. shahrudensis, A. dipsaceus, A. lacei usw. In Kappadokien kommt diese Art sehr lokal unter A. dipsaceus auf trockenen steppenartigen Abhängen in schmalen tiefen Tälern vor. Derivatio nominis. Die neue Unterart wird nach ihrem Fundgebiet Kappadokien benannt. Synansphecia hera sp. n. (Abb. 7) Material. Holotypus ©, Türkei, Osttaurus, 1 km S Pürem Paß (N von Kahramanmaras), 37.58° N, 36.35 ° O, 1500 m, 30. 6.—1. 7. 1991, K. Spatenka leg., coll. Museum Witt, München; Paratypen: 1 © mit gleichen Daten, coll. K. Spatenka; 1 ©, Türkei, Provinz Adana, Taurus- Gebirge, 10 km N Saimbeyli, 38.06° N, 36.08° O, 1450 m, 29. 6. 1991, K. Spatenka leg. et coll. (Genitalpráparat Gorbunov Nr. 93-05) ; 4 o ©, Türkei, Konya, 50 km S Beysehir, Umgebung Bademli, 1300 m, 8. 7. 1992, A. Lingenhöle leg., coll. K. Spatenka, 10 © © mit gleichen Anga- ben, A. Lingenhöle leg. et coll., 1 © mit gleichen Angaben A. Kallies coll.; 1 © Türkei, Konya, Beysehir-See, 20 km W Beysehir, 7. 7. 1992, A. Lingenhöle leg. und coll. Männchen. Spannweite 23 mm. Im allgemeinen ähnlich S. mannii (Lederer, 1853), im Ver- gleich zu dieser aber auffallend größer und mit größeren Glasfeldern. Antenne schwarz, außen subapikal mit einzelnen ockergelben Schuppen. Labialpalpus ockergelb, das erste und zweite Glied noch mit weißen Schuppen bestäubt, außen schwarz, mit langen haarförmigen Schup- pen; Stirn schneeweiß; Scheitel schwarz mit langen gelbgrauen Haaren; Halskragen ockergelb. Thorax: schwarz, Tegula innen mit blaßgelben Haaren und einzelnen Schuppen begrenzt; Pleura ockergelb. Beine: Vordercoxa schwarz, dicht mit ockergelben haarförmigen Schuppen überstäubt, so daß die Grundfärbung kaum sichtbar ist; Vorderfemur und -tibia ockergelb und schwarz gemischt. Hintercoxa und -femur schwarz, ventral mit langen blaßgelben Haaren; Tibia ockergelb, beim zweiten Spornpaar mit schwarzem Ring; Sporne ockergelb; Tarsus ockergelb, das letzte Tarsomer schwarz. Vorderflügel: Ränder schwärzlich; Diskalfleck schwarz, halb so breit wie das Rundfeldchen (ETA); Apikalfeld schwärzlich mit hell ockergelben Strahlen zwischen den Adern; die Glas- felder gut entwickelt, leicht mit ockergelben und weißen Schuppen überstäubt; ETA höher als breit, mäßig konkav gewölbt, schmaler als das Apikalfeld. Ventral: Costal- und Analrand, Cubitalstamm, Adern im Rundfeldchen und Strahlen im Apikalfeld zitronengelb. Hinterflügel: hyalin mit schwarzbraunem Außenrand, schwarzen Adern und schwarzem keilförmigem Diskalfleck, der den Aderstiel M3—Cul erreicht; Fransen grau, lang. Ventral: Vorderrand und Adern zitronengelb; Außenrand und Diskalfleck mit einzelnen zitronengelben Schuppen. Abdomen: schwarz, die Hinterränder der Tergite 2, 4 und 6 schmal weiß, die Tergite 3—7 noch dicht gelb überstäubt; auf Tergit 2 nur medial spárliche ockergelbe Schuppen; Analbusch schwarz, in der Mitte breit, lateral schmal goldgelb. Lateral über alle Tergite eine durch- laufende goldgelbe Linie. Ventral: schwarz, die Sternite 2 und 3 an den Hinterrändern dünn, 4—7 in fast ganzer Breite dicht goldgelb bestäubt; Analklappen goldgelb. . Differentialdiagnose. S. hera ist nahe verwandt und auch ähnlich mit S. mannii vom Balkan und aus dem Pontus. Diese Art ist im Durchschnitt kleiner (15—17 mm) und dunkler, die gelben Partien sind bei ihr dunkel statt hell ockergelb, was besonders im Apikalfeld auffällt; das erste Palpusglied ist dorsal schwarz, mit gelber Beimischung (bei S. hera gelb mit weißen Schuppen); die Stirn hellgelb bis ockergelb (nicht weiß); die Antenne ventral hellbraun (nicht schwarz); das ETA ist bei S. mannii sehr schmal, schmaler als der Diskalfleck, halb so breit wie das Apikalfeld; das Abdomen ist ventral fast ganz schwarz, nur mit einzelnen ockergelben 52 K. Spatenka Schuppen; der Analbusch schwarz, nur bei einigen Exemplaren lateral mit wenigen ocker- gelben Schuppen, selten auch medial. Genitalapparat © (Abb. 12). Sehr ähnlich dem von S. mannii, die Valve subapikal ein wenig breiter und apikal weniger zugespitzt. Bionomie und Habitat. Die Bionomie ist unbekannt. Nach dem Charakter der Biotope ist zu vermuten, daß die Raupe ähnlich der verwandten S. mannii in einer Geranium sp. (Gerania- ceae) lebt. Waldsteppen mit sehr reicher Vegetation, Lichtungen in Nadel- und Laubwäldern. Verbreitung. Bisher nur von den genannten Lokalitäten bekannt. Derivatio nominis. Die neue Art wird nach der Gattin und Schwester des griechischen Gottes Zeus, der Hüterin der Ehe, benannt. Dipchasphecia intermedia sp. n. (Abb. 8) Material. Holotypus ©, Türkei, Taurus-Gebirge, 10 km N Saimbeyli (Hadjin), 1450 m, 36.07.12° O, 38.03.12° N, 6. 6. 1993 ex larva ex Acantholimon sp., K. Spatenka leg., coll. Museum Witt, München; Paratypen: 5 oo, 4 Q Q mit gleichen Daten; 30 oro Türkei, Taurus-Gebirge, Kan-Paß, 1565 m, 38.18° N, 36.21° O, 29.—30. 6. 1991, K. Spatenka leg. et coll.; 2 © © mit gleichen Daten coll. Z. Lastuvka (Brno, Tschechische Republik); 14 oo, Türkei, Pontus-Gebirge, Gümüs, 1000 m, 40.49° N, 35.10° O, 5. 7. 1991, K. Spatenka leg. et coll; 2 oo mit gleichen Daten, coll. Z. Lastuvka; 2 © ©, Türkei, Kappadokien, Karain, 1200 m, 30.36° N, 35.04 ° O, 3. 7. 1991, K. Spatenka leg. et coll.; 1 ©, gleiche Angaben in coll. A. Kallies; 1 ©, Kappadokien, Goreme, 1100 m, 20.—28. 6. 1992, H. Riefenstahl und M. Petersen leg., coll. A. Kallies, 1 © mit gleichen Daten in coll. H. Riefenstahl. Männchen. Spannweite 13,5—21 mm. Labialpalpus weiß, das dritte Glied schwarz durch- mischt, lateral mit länglichem schwarzem Strich; Antenne schwarz, Pedicellus und Scapus ventral ockergelb; Stirn weiß, medial mit einzelnen braunen Schuppen; Scheitel schwarz mit einzelnen rostbraunen Schuppen; Halskragen rostbraun. Thorax: Schwarz; Scapularfleck sehr undeutlich, nur aus einigen weißlichen Schuppen bestehend oder fehlend; Tegula am Innenrand begrenzt ockergelb; Metathorax mit 2 Gruppen von langen weißen Haaren; Thorax lateral mit großem weißlichem oder ockergelbem Axillar- fleck. Vordercoxa weiß, -femur, - tibia und -tarsus braun; Tibia am Distalrand ockergelb; Tarsomeren an den Distalrändern weiß; Hinterbein braun, -tibia dorsal, besonders im Medial- teil, dicht weiß behaart; Sporne weißlich; Tarsomeren distal weißlich. Vorderflügel: dunkelbraun mit sehr kleinen Glasfeldern; Costalrand vor.dem Apex gelb- weiß, das breite braune Apikalfeld mit kurzen weißgelben Strahlen zwischen den Adern; Diskalfleck braunschwarz, breiter oder ebenso breit wie das ETA; alle Glasfelder sehr klein, bei frischen Exemplaren dicht mit weißlichen Schuppen bedeckt; ETA im typischen Fall punktförmig, höher als breit oder rundlich, durch die Adern in 3 Zellen geteilt. Ventral: Ränder und Apikalfeld dicht schmutzigweiß beschuppt. Hinterflügel: durchsichtig mit schmalem dunkelbraunem Außenrand, braunen Adern, Fran- sen und bei M2 nach unten gewinkeltem Diskalfleck. Abdomen: dunkelbraun, die Tergite 2 und 4 breit schmutzigweiß gerandet, Tergit 6 am Distalrand schmal weißlich begrenzt; Analbusch braunschwarz, lateral an der Basis weißlich. Das ganze Abdomen bei frischen Exemplaren leicht mit schmutzigweißen Schuppen bestäubt. Die breiten Hinterränder der Tergite 2, 4, 6 und 7 lateral weißlich. Ventral: braun, Sternit 2 distal breit schmutzigweiß, die Sternite 3—7 mit mehr oder weniger deutlichen, schmalen schmutzigweißen Hinterrändern. Das ganze Abdomen ventral mit vereinzelten weißen Schuppen. Weibchen. Robuster, mit dickem Abdomen; Stirn schneeweiß; Innenrand der Tegula am Prothorax weißgelb, Meso- und Metathorax breit weiß begrenzt; Prothorax mit deutlicher, ockergelber mediodorsaler Linie. Abdomen dunkelbraun, Tergite 2 und 6 mit schmalem, 4 mit breitem weißlichem Hinterrand; Analbusch mit 2 sublateralen weißlichen Strahlen. Ventral: Sternit 2 braun, dicht weiß überstäubt, Sternite 4—6 mit schmalen weißen Distalrándern, Sternit 3 mit einzelnen weißen Schuppen am Distalrand. Neue Glasflügler aus Europa und der Türkei 33 Variabilität. Im Ausmaß der hellen Beschuppung und in der Größe der Glasfelder relativ stark variabel. In Populationen aus dem Taurus-Gebirge kommen selten Exemplare mit größerem ETA und hellerem Apikalfeld vor; bei einem Exemplar aus Kappadokien sind die Labial- palpen fast ganz weiß, die Tegula breit weiß beschuppt, der Costal- und Analrand des Vorder- flügels dicht weißlich überstäubt und das Apikalfeld heller, mit längeren weißlichen Strahlen zwischen den Adern. Die drei bekannten Exemplare aus Bulgarien sind dunkler, mit größeren Glasfeldern, die Weibchen mit nur zwei schmalen weißlichen Ringen auf den Tergiten 2 und 4, mit sehr dunklem Apikalfeld und weniger ausgeprägten weißlichen Strahlen im Analbusch, Sternit 2 weißlich, sonst nur 4 distal schmal weißlich begrenzt. Die taxonomische Stellung der Exemplare aus Bulgarien ist bisher nicht ganz klar, weshalb diese nicht in die Typenserie auf- genommen wurden. Genitalapparat ~ (Abb. 13). Ähnlich den verwandten Arten. Die ganze Gattung ist genital- morphologisch sehr einheitlich. Genitalapparat Q (Abb. 14). Wie bei den verwandten Arten. Differentialdiagnose. Die neue Art ähnelt Dipchasphecia lanipes und D. krocha und bildet zwischen diesen in vielen Merkmalen einen Übergang. D. lanipes aus Südbulgarien ist bisher leider nur in 4 Exemplaren bekannt, die sich aber konstant und deutlich von der neu beschrie- benen Art unterscheiden. D. lanipes hat ein großes ovales, in 5 Zellen geteiltes Rundfeldchen, ein sehr breites weißgelbes Band auf dem Tergit 4, der Diskalfleck des Vorderflügels ist nur halb so breit oder schmaler als das ETA, der Thorax trägt einen großen weißlichen Scapular- fleck, die helle Begrenzung der Tegula fehlt, das Abdomen ist ventral vollständig weißgelb bestáubt, ohne deutliche Ringe. Die Larvalbionomie von D. /anipes ist unbekannt, die Raupe lebt aber offensichtlich nicht in Acantholimon, sondern eher in einer Limonium-Art (Plum- baginaceae). D. krocha steht der neuen Art noch wesentlich näher. Sie ist aber deutlich kleiner (Spannweite 10—11,5 mm), hat das Apikalfeld konstant sehr hell, das Abdomen dorsal mit sehr schmalen weißlichen Ringen auf den Tergiten 2, 4 und 6 (oder nur auf 2 und 4), ventral mit schmalen weißen Distalrändern auf den Sterniten 2—7. D. krocha lebt ebenfalls in Acantholimon, aber in einer anderen, sehr kleinen Art. Die Futterpflanzen von D. inter- media sp. n. und D. krocha wurden bisher leider nicht genau determiniert, da es sich um eine taxonomisch sehr komplizierte Gattung handelt. Bionomie. Die Raupe ist einjährig und lebt in den holzigen Wurzeln einer kugelförmigen, dornigen, rotblühenden Acantholimon sp., manchmal mehrere Raupen in einer Pflanze. Vor der Verpuppung bauen die Raupen lange, von innen fein ausgesponnene Röhrchen, in denen sie sich auch verpuppen. Die Imagines treten in der Natur im Juni und Anfang Juli auf. Sie fliegen in den späten Nachmittags- und frühen Abendstunden (16—19 Uhr) und sind zu dieser Zeit auch sexuell aktiv. Habitat. Trockene steinige Plätze mit spärlicher dorniger Steppenvegetation mit im allgemei- nen spärlicher Insektenfauna. Verbreitung. Türkei: Pontus (Gümüs), Taurus (Saimbeyli, Kan-Paß, Malatya, Kayseri), Kappa- dokien; Süd- und Südost-Bulgarien: Sliven, Schwarzmeerküste südlich von Burgas. Derivatio nominis. Die Benennung erfolgt in Berücksichtigung der Tatsache, daß die neue Art eine vermittelnde Stellung zwischen den Arten D. lanipes und D. krocha einnimmt. Chamaesphecia kautti sp. n. (Abb. 9) Material. Holotypus ©, Türkei, Nigde, Ala Dag, Demirkazik, 25.—26. 7. 1991, 1600 m, P. Kautt leg., coll. Museum Witt, München; Paratypen: 24 oo 1 Q P. Kautt leg., coll. K. Spatenka; 3 Q 2 © dtto, 20. 7. 1996, K. Spatenka leg. et coll. Männchen. Spannweite 22 mm. Kopf: Labialpalpus dunkelbraun, 1. und 2. Glied dorsal und innen elfenbeinfarben durchmischt, 3. elfenbeinfarben mit braunen Schuppen; Stirn hellbraun mit perlmuttartigem Glanz, vor dem Auge mit weißem Saum; Scheitel schwarz; Antenne 54 K. Spatenka braunschwarz, Pedicellus und Scapus ventral elfenbeinfarben; Halskragen dorsal rostbraun, lateral ockergelb. Thorax: schwarz; Patagia dunkelbraun mit starkem Glanz; Tegula innen schmal ockergelb gerandet; Metathorax am Distalrand mit einzelnen ockergelben Schuppen und weißgelben Haaren und Schuppen; Scapularfleck schmutzigweiß; Axillarfleck ockergelb. Beine: Vorder- coxa braunschwarz, caudal am Rand weiß, Femur, Tibia und Tarsus dunkelbraun. Hinterbein dunkelbraun, Tibia beim ersten Spornpaar mit breitem weißem Ring und am Distalrand mit weißen und ockergelben Schuppen. Vorderflügel: dunkelbraun; Costalkante ockergelb; Cubitalstamm und Analrand mit sehr spärlichen ockergelben Schuppen; im Apikalfeld kurze blaßgelbe Strahlen zwischen den Adern; das Keilfeldchen (ATA) reicht fast bis zur Flügelwurzel; das Längsfeld bis 2/3 der Distanz zum Diskalfleck; das Rundfeldchen (ETA) groß, oval, durch die Adern in 5 Zellen geteilt, die erste Zelle ist vom Costalrand teilweise mit ockergelben Schuppen bestäubt; die Adern im ETA sind ebenfalls teilweise mit ockergelben Schuppen bestáubt; ETA etwa doppelt so breit wie das Apikalfeld und dreimal so breit wie der Diskalfleck. Ventral: Ränder und Adern blaßgelb bestäubt. Hinterflügel: unbeschuppt, mit braunen Adern, schmalem dunkelbraunem Außenrand und braunen Fransen; Diskalfleck dunkelbraun, rechteckig, dünn auf die Ader M2 auslaufend. Vorderrand und Adern ventral weißgelb durchmischt. Abdomen: dunkelbraun, mitteldick, mit schmalem Analbusch; Tergite 4 und 6 am Hinter- rand schmal weiß gerandet; am Hinterrand des Tergites 7 einzelne schmutziggelbe Schuppen; im dunkelbraunen Analbusch lateral an der Basis und medial einige ockergelbe Schuppen. Ventral schwarzbraun, die Distalránder der Sternite 4—7 mit weißen Schuppen. Weibchen. Noch dunkler als das Männchen; Vorderflügel: Längsfeld braunschwarz; Rundfeld- chen klein, vierzellig, die obere Zelle mit braunschwarzen und blaßgelben Schuppen bestäubt, die untere Zelle nur rudimentär ausgeprägt; Apikalfeld etwa so breit wie ETA, braunschwarz, nur mit vereinzelten ockergelben Schuppen; Hinterbein braunschwarz, nur lateral mit weiß- lichem Strich am ersten Spornpaar, dorsal beim ersten Spornpaar und am Distalrand mit eini- gen weißlichen Schuppen; Abdomen schwarzbraun, nur Tergit 4 am Hinterrand weißgelb ~ umrandet. Sternit 4 am Hinterrand mit einer Reihe weißer Schuppen, 6 mit vereinzelten weißen Schuppen. Variabilität. Nur bei den Männchen beurteilt, da nur wenige Weibchen bekannt sind. Bei man- chen Männchen ist die ockergelbe Mediallinie auf dem Abdomen gut sichtbar, die von den Fleckchen auf den Tergiten 3—6 gebildet wird. Bei solchen Exemplaren können auch ganz ver- einzelte weiße Schuppen am Hinterrand des Tergites 2 und eine ganz kurze ockergelbe Medial- linie auf dem Thorax vorhanden sein. Einige Männchen haben fast schwarze Labialpalpen, manche wieder einen nicht rechteckigen, sondern leicht keilförmigen Diskalfleck des Hinter- flügels. Genitalapparat & (Abb. 15). Ähnelt dem von Ch. bibioniformis (Esper, 1800); Valve lang- gezogen, kahle Fläche groß, Crista sacculi breit, einfach, Scopula fehlend, Crista gnathi sehr niedrig und kurz. Genitalapparat © (Abb. 16). Ähnlich wie bei den verwandten Arten. Differentialdiagnose. Ch. kautti sp. n. ist nahe verwandt mit den Arten der Ch. bibioniformis- Gruppe und ähnelt aus dieser besonders Ch. taurica sp. n. Ferner ähnelt sie habituell auch Ch. guriensis (Emich von Emöke, 1872) und Ch. kistenjovi Gorbunov, 1991, aus Transkauka- sien. Ch. taurica unterscheidet sich schon auf den ersten Blick durch die Gestalt: sie ist kleiner, schlanker und hat größere Glasfelder des Vorderflügels. Sie unterscheidet sich unter anderem auch durch weitere Details, so z. B. den schneeweißen Labialpalpus (bei Ch. kautti sp. n. braun), die außen ockergelbe Antenne (bei Ch. kautti sp. n. schwarzbraun), den schneeweißen Scapularfleck (bei Ch. kautti sp. n. ockergelb), die schneeweiße Vordercoxa (bei Ch. kautti sp. n. braunschwarz, nur weiß umrandet), das sehr schwach entwickelte braunschwarze Apikalfeld (bei Ch. kautti sp. n. breit, mit kurzen ockergelben Strahlen), den strichförmigen Diskalfleck im Hinterflügel (bei Ch. kautti sp. n. breit), die weißen Ringe auf den Tergiten 2, 4 und 6 (bei Ch. kautti sp. n. nur auf 4 und 6), das weiße Abdominalsternit 4 (bei Ch. kautti sp. n. braun- Neue Glasflügler aus Europa und der Türkei 55 schwarz). Die beiden obengenannten transkaukasischen Arten gehören in die Ch. empiformis- Gruppe, haben also keinen hellen Scapularfleck, den Thorax lateral dunkel und einige Abdo- minaltergite auf der ganzen Oberfläche diffus weißlich oder weißgelb bestäubt. Bionomie und Habitat. Die Raupe ist wahrscheinlich nur einjährig und lebt in dicken Wurzeln von Euphorbia kotschgiana (Euphorbiaceae). Diese mächtige Pflanze ist in den schotterigen Anschwemmungen im Bachtal und an den kahlen Kalkfelsen oberhalb von Demirkazik am Ala-Dag sehr dominant. Die Falter sind sehr scheu und setzen sich bei Sonnenschein auf die dürren Stengel der Futterpflanze. Derivatio nominis. Die Benennung erfolgt zu Ehren des Entdeckers dieser neuen Art. Chamaesphecia taurica sp. n. (Abb. 10) Material. Holotypus ©, Türkei, Südanatolien, Taurus-Gebirge, Camlıyayla (Namrun), 1200 m, 37.11° N, 34.39° O, 26.—27. 7. 1991, K. Spatenka leg., coll. Museum Witt, München; Para- typen: 7 © © mit gleichen Daten, coll. K. Spatenka; 1 © mit gleichen Daten in coll. I. ToSevski (Beograd, Jugoslawien). Männchen. Spannweite 19 mm. Labialpalpus schneeweiß, das dritte Glied blaßgelb, apikal braunschwarz durchmischt; Stirn dunkelbraun mit perlmuttartigem Glanz, Auge ventral weiß gerandet; Scheitel glänzend schwarz; Antenne dunkelbraun, außen ockergelb, Pedicellus und Scapus ventral ockergelb; Halskragen rostbraun, lateral weiß. Thorax: braunschwarz; Tegula innen schmal ockergelb gerandet; Metathorax mit vereinzel- ten dünnen ockergelben Haaren und mit zwei Gruppen gelbweißer Schuppen am Hinterrand; Patagia lateral ockergelb, unter den Flügeln ein großer schneeweißer Fleck; Scapularfleck schneeweiß. Vordercoxa schneeweiß, Femur, Tibia und Tarsus braun mit weißer Beimischung; Hintercoxa und Femur schwarz, ventral mit schneeweißen Schuppen und Haaren, Tibia dorsal braunschwarz, am Distalrand weiß, lateral beiderseits weiß, distal mit breitem dunkelbraunem Saum, am Distalrand schneeweiß, ventral von der Basis zum ersten Spornpaar schwarz, dann bis zur Mitte zwischen beiden Spornpaaren weiß und das distale Drittel wieder braunschwarz, beide Spornpaare schneeweiß, Tarsus braun, das erste Tarsomer dorsal mit ockergelben Schup- pen bestäubt, an den Distalenden der übrigen Tarsomere immer einzelne ockergelbe Schuppen; ventral braun, an den Distalenden der Tarsomere leicht ockergelb bestäubt. Vorderflügel: mit abgerundetem Apex und sehr gut entwickelten Glasfeldern; Ränder, Adern, Diskalfleck und Fransen braunschwarz, Apikalfeld stark reduziert, braunschwarz; Costalkante bis zum Diskalfleck ockergelb, dann bis zum Vorderapex braunschwarz, subapikal aber mit weißem Fleck; ATA bis zur Flügelwurzel reichend; PTA fast zum Diskalfleck; ETA sehr groß, länger als hoch, durch die Ader in 5 Zellen geteilt; Adern im ETA mit einzelnen ockergelben Schuppen; Apikalfeld sehr schmal, in der Mittelpartie nur ein Viertel so breit wie das ETA, einfarbig braunschwarz. Ventral: Costal-, Analrand und Cubitalstamm dicht ocker- gelb beschuppt. Hinterflügel: unbeschuppt mit braunschwarzen Adern und Fransen, der sehr schmale Außenrand und Diskalfleck ebenfalls braunschwarz; Fransen am Analrand und Adern an der Flügelwurzel weiß; Diskalfleck sehr schmal, strichförmig, den Aderstiel M3—Cul erreichend. Abdomen: dorsal braunschwarz, die Tergite 2, 4 und 6 mit schmalen weißen Hinterrändern, alle Tergite mediodorsal schwach ockergelb bestäubt, diese Bestäubung von Tergit 1 bis 7 abnehmend und eine undeutliche schmale Mediallinie bildend; Analbusch dreieckig, in drei Büschel geteilt, craniolateral und medial mit einzelnen ockergelben Schuppen, sonst braun- schwarz. Ventral dunkelbraun, Sternit 2 schneeweiß, 4 in der hinteren Hälfte schneeweiß, 5—7 - medial mit einzelnen schmutzigweißen Schuppen, weiße Fleckchen auch an den seitlichen Hinterrändern der Tergite 3, 5 und 6; Analklappen dunkelbraun, caudal ockergelb. Weibchen. Unbekannt. Variabilität. Größe 16,5 —21 mm; Scheitel bei ganz frischen Exemplaren mit einzelnen rost- braunen Schuppen, auf dem Prothorax ist eine mediale ockergelbe Linie angedeutet, und die weiße Umrandung der Abdominaltergite 2, 4 und 6 ist leicht ockergelb überstäubt. 56 K. Spatenka Genitalapparat o (Abb. 17). Ähnlich Ch. bibioniformis. Differentialdiagnose. Ähnlich und verwandt mit Ch. bibioniformis, von dieser aber in mehre- ren Merkmalen veschieden. Die polymorphe Ch. bibioniformis bildet in ihrem umfangreichen Verbreitungsgebiet mehrere Ökoformen von unsicherer taxonomischer Stellung. Stücke aus Spanien und Südfrankreich (Futterpflanzen: Euphorbia nicaensis, E. serrata; Euphorbiaceae) sind auffallend groß und robust, von gelbbrauner Grundfarbe und mit sehr großem ETA; in Mitteleuropa, auf dem Balkan, in der Türkei, in Südrußland und in Transkaukasien finden wir Populationen mit breiterem Apikalfeld und grünlicher Grundfarbe vor (Futterpflanze: E. segueriana); in Mazedonien lebt die Art auch in E. myrsinites und E. niciciana, und die Imagines sind auffallend stark gelb bestäubt. In Kappadokien (Türkei) zeichnet sich die Art durch eine sehr reiche weiße Bestäubung aus (Futterpflanze: Euphorbia sp.). Alle diese Formen haben freilich ein breites Apikalfeld mit hellen Strahlen zwischen den Adern und weißer Umrandung des 7. Abdominaltergites. Bei Ch. faurica sp. n. ist das Apikalfeld hin- gegen extrem schmal und einfarbig braunschwarz, und Tergit 7 ist einfarbig braunschwarz. Bionomie. Die Larvalbionomie ist bisher unbekannt, auf der Typenlokalität hat der Autor aber nur Euphorbia kotschyana (Euphorbiaceae) gefunden, die damit die potentielle Futter- pflanze darstellt. Habitat. Die Typenlokalität ist eine sehr pflanzenreiche Waldsteppe in Camlıyayla. Das Gebiet um Camlıyayla stellt eine feuchte Enklave im sonst sehr trockenen südöstlichen Taurus dar. Gemeinsam mit Ch. taurica sp. n. wurden an der Typenlokalität folgende Glasflügler gesam- melt: Tintia myrmosaeformis (Herrich-Schäffer, 1846), Paranthrene insolita insolita Le Cerf, 1914, Synanthedon myopaeformis (Borkhausen, 1789), Bembecia illustris (Rebel, 1901), Pyropteron minianiforme (Freyer, 1843), Chamaesphecia haberhaueri (Staudinger, 1879), Ch. proximata (Staudinger, 1891), Ch. alysoniformis (Herrich-Schäffer, 1846) und von ande- ren Autoren auch Osminia fenusaeformis (Herrich-Schäffer, 1852) und Euhagena palariformis (Lederer, 1858). Derivatio nominis. Die Art wird nach ihrem Fundgebiet, dem Taurus-Gebirge, benannt. Danksagung Für die Überlassung des Materials zur Bearbeitung und Beschreibung sowie für wichtige Informationen bin ich den Herren M. Petersen (Pfungstadt), H. G. Riefenstahl (Hamburg), A. Kallies (Greifswald), P. Kautt (Tübingen), A. Lingenhöle (Biberach), T. Dobrovsky (Praha), Z. LaStüvka (Brno) und O. Gorbunov (Moskau) sehr zu Dank verpflichtet. Dr. O. Gorbunov (Moskau) und Dr. I. Tosevski (Novi Beograd) fertigten freundlicherweise die Zeichnungen der Genitalpräparate an. Dafür danke ich ihnen ganz besonders herzlich. Für die sprachliche Korrektur des Manuskriptes bin ich den Herren A. Kallies (Greifswald) und Prof. Dr. C. M. Naumann (Bonn) zu Dank verpflichtet. Zusammenfassung Der Autor legt Beschreibungen von neun neuen Glasflüglerarten und -unterarten vor. Bembe- cia peterseni sp. n. wurde in größerer Anzahl an mehreren Lokalitäten in Anatolien (Türkei) gefangen; sie ähnelt mehreren Arten der Gattung Bembecia und ist nahe mit B. iberica aus Frankreich, Spanien und Marokko verwandt. Synansphecia hera sp. n. wird in meh- reren Exemplaren aus dem Taurus (Türkei) beschrieben und ist mit S. mannii verwandt. Dipchasphecia intermedia sp. n. wurde im Pontus und im Taurus (Türkei) gefangen und gezüchtet; sie ähnelt D. lanipes und D. krocha; die Raupe lebt in Acantholimon sp. Chamae- sphecia kautti sp. n. aus Nigde (Türkei) ist nahe mit Ch. bibioniformis verwandt. In die gleiche Gruppe gehört auch Chamaesphecia taurica sp. n. Synanthodon stomoxiformis riefenstahli ssp. n. aus Südspanien, Synanthedon andrenaeformis tenuicingulata ssp. n. aus der Nordost- Türkei und aus Armenien, Bembecia pavicevici dobrovskyi ssp. n. vom Taygetos (Griechen- land) und Bembecia syzcjovi kappadocica ssp. n. aus Kappadokien (Türkei) stellen neue geo- graphische Unterarten bereits bekannter Arten dar. Neue Glasflügler aus Europa und der Türkei 7, Literatur Bartel, M. (1912): pp. 375—416. — In Seitz, A.: Die Gross-Schmetterlinge der Erde. I. Die Gross-Schmetterlinge des paläarktischen Faunengebietes. II. Spinner und Schwärmer. Stuttgart. Gorbunov, O. (1987): Novyj vid roda Bembecia (Lepidoptera, Sesiidae) iz Talischa. — Vest. Zool. 1987 (3): 12—18. Gorbunov, O. (1989): A new species of the genus Bembecia Hübner, 1819 from the Cauca- sus, USSR. — Atalanta, Würzburg 20: 119—123. Gorbunov, O. (1991): Review of the genus Dipchasphecia Capuse, 1973. — Atalanta, Würz- bure2222145—167. Lastüvka,Z.& A.LaStüvka (1994): Bembecia fibigeri sp. n. aus Spanien. — Nota lepid. 16: 233—239. Lastüvka,Z.& A.LaStüvka (1995): An illustrated Key to European Sesiidae. — 174 pp. Brno. Spatenka, K. (1992): Weitere neue paláarktische Sesiiden. — Alexanor 17: 427 —446. Spatenka,K., Z. LaStüvka, O. Gorbunov, I. ToSevski & Y. Arita (1993): Die Systematik und Synonymie der paläarktischen Glasflügler-Arten. — Nachr. ent. Ver. Apollo (N. F.). 14: 81 —114. ToSevski, I. (1989): A new species of the genus Bembecia Hübner, 1819 from Macedonia. — Fragm. Balcanica 14: 81 —89. RNDr Karel Spatenka, Vyletni 362, CR-14200 Praha 4-Písnice, Tschechische Republik. del ee mre na ; he A te cpm dl ren 2 tae: me ed pat g Pa Y wi £ 7 Ber thsi dik ; le? ; ur Fa ate — a+ E Ali Mar: ‚u, mb A AA Arial die psi pte rst se fh ate rs sera a ae Er Sr Er N as Da wah = o ? a El N 4 ; 3 = el F a 5 o Du ee Bd. 4 S. 50-68 | Bonn, September 1997 Review of the genus Paradoxecia Hampson, 1919 (Lepidoptera, Sesiidae, Tinthiinae) Oleg G. Gorbunov & Yutaka Arita Abstract. The clearwing moths of the genus Paradoxecia Hampson, 1919 are reviewed. Two new species, P vietnamica sp. n., and P fukiensis sp. n. are described from Vietnam and southeastern China, respectively. Redescriptions of P gravis (Walker, [1865]) and P pieli Lieu, 1935 are also provided. A key to all taxa currently known for the genus is compiled. Key words. Sesiidae, Tinthiinae, Paradoxecia, new species, taxonomy, Vietnam, China. The genus Paradoxecia, erected by Hampson (1919) for Aegeria gravis Walker, [1865], originated from “North China” [=eastern China, Shanghai ?]. In 1935, K.O. Vic- toria Lieu described P pieli after a series of specimens reared from larvae, boring in mulberry-tree twigs in eastern China. Until now, Paradoxecia contained only these two species. In 1933, Gaede placed Paranthrene croconeura Meyrick, 1926, a species from Sikkim, India, in Paradoxecia, as did Heppner & Duckworth (1981). However, we believe croconeura neither belongs to the genus Paradoxecia, nor even to the sub- family Tinthiinae. This species seems to belong to the genus Adixoa Hampson, [1893] or a closely related genus of the Paranthreninae. So, we remove croconeura from the genus Paradoxecia. While investigating the Oriental Sesiidae from the collections of Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn, Germany and Muséum d’Histoire Naturelle, Genéve, Switzerland, we came across a few highly interesting specimens from Vietnam and southeastern China belonging to the genus Para- doxecia and distinct from the two known species of this genus. We describe the two unique species below. So, at present, the genus Paradoxecia includes four species, viz. P gravis (Walker, [1865]), P pieli Lieu, 1935, P vietnamica sp. n., and P fukiensis sp. n. In addition, we redescribe P gravis and P pieli and present a key to all known species of Paradoxecia. Material examined or cited herein has been deposited in the following collections abbreviated in the text as follows: BMNH The Natural History Museum, London, Great Britain; ZFMK Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn, Germany; MHNG Museum d’Histoire Naturelle, Geneve, Switzer- ‘land; MUT Zoological Laboratory, Faculty of Agriculture, Meijo University, Nagoya, Japan; CG collection of O. Gorbunov, Moscow, Russia. 60 ONG, Gorbumnow & Yo Am ta Subfamily Tinthiinae, Tribe Tinthiini Le Cerf Tinthiinae Le Cerf, 1917: 148. Type genus Tinthia Walker, [1865]. Genus Paradoxecia Hampson Paradoxecia Hampson, 1919: 51 (key), 114 (description). Type-species: Aegeria gravis Walker, [1865], by original designation. Dalla Torre & Strand 1925: 180; Gaede 1933: 797 (part.); Naumann 1971: 22, 55; Heppner & Duckworth 1981: 21 (part.); Fletcher & Nye 1982: 118. Medium or large-sized clearwing moths with alar expanse 19—40 mm. Head with antenna filiform, shortly bipectinate and ciliate in male, without hair tuft apically; proboscis well- developed, long, functional; frons and vertex smooth, covered with elongate scales; labial palpus smooth-scaled. Legs with tibia and first tarsomere with short tufts of elongate, apically pointed and rusty coloured setae. Forewing entirely opaque, sometimes with a short and narrow semihyaline stripe in place of posterior transparent area; veins R4 and RS separate, Cu2 short, stalked with Cul (Fig. 7). Hindwing transparent or densely covered with semihya- line with brownish or yellowish tinged scales; veins M1 and M2 nearly reduced basally; vein M2 arising slightly before cross-vein, M3 and Cul arising long before cross-vein; Al well- developed (Fig. 7). Male genitalia (Fig. 10) with uncus well-developed, finger-shaped, covered with simple short setae, well separated from tegumen; tegumen triangular, broad, without gnathos; valva broad, rounded distally, covered with simple hairs and setae; saccus short and broad; aedeagus narrow, long, about twice as long as valva, with long coecum penis and with a strong thorn distally where penis broadens dorsally; vesica with numerous minute spinules. Female genitalia (Figs 8—9) with 8th tergite well-sclerotized, with numerous long setae at distal margin; papillae anales strongly sclerotized with outside curved ventral margin, or nearly membranous, with short setae; apophysis posterior somewhat shorter, equal or longer than apophysis anterior; latter sometimes with distinct ventral appendix; lamella postvaginalis sometimes sclerotized; ostium bursae membranous, on intersegmental membrane; antrum short, broad, slightly or well-sclerotized; ductus bursae short and broad, membranous; corpus _ bursae ovoid, membranous, without or with one or two signa. Diagnosis. Paradoxecia seems to be closely related with Tinthia, Microsphecia Bartel, 1912 (distinct, not congeneric with Tinthia), Paranthrenopsis Le Cerf, 1911, and Zenodoxus Grote & Robinson, 1868. From all of these genera, Paradoxecia differs by the relatively larger size and by the short bipectinate antenna in the male (ciliate in these compared genera). In addi- tion, Paradoxecia can be distinguished from Tinthia and Microsphecia by the shape of the aedeagus in the male genitalia (relatively short and broad with short cecum penis in these compared genera) and by the shape of the female genitalia (8th tergite and both pairs of apophyses relatively longer, antrum long in Tinthia and Microsphecia). Paradoxecia is distinguishable from Zenodoxus by the well-developed uncus and form of the distal part of the aedeagus (uncus small, aedeagus without thorn distally in Zenodoxus). This genus can be separated from Paranthrenopsis by the entirely opaque forewing (with small transparent areas in Paranthrenopsis) and by the shape of tegumen in the male genitalia (small, poorly developed in Paranthrenopsis). Additionally, Paradoxecia can easily be distinguished from the genus Trichocerota Hampson, [1893] by the larger size, entirely opaque forewing and differen- ces in both male and female genitalia. Structure. The genus Paradoxecia currently consists of four species, viz. P gravis (Walker, [1865]), P pieli Lieu, 1935, P vietnamica sp. n., and P fukiensis sp. n. Distribution. Known from the northeastern continental part of the Oriental Region: Vietnam, eastern and southeastern China. Including P gravis in the Palaearctic sesiid fauna (Naumann 1971; Spatenka et al. 1993) is probably incorrect, because P gravis is known only from “North China”, which, in our opinion, is most likely Shanghai, a location somewhat south of the Palaearctic Region (Heppner 1991). Review of genus Paradoxecia 61 Key to species of Paradoxecia based on external characters BENbdomentdorsallyzonly with yellow coloured’ stripes 2.2... nennen. 2 — Abdomen dorsally without yellow coloured stripes or with yellow ACG ORALE SUE ae erties tc ont o ed oe we ed oh 3 2. Alar expanse more than 35.0 mm; labial palpus white ventrally and light brown dorsally; abdomen ventrally entirely dark brown ................. gravis — Alar expanse less than 35.0 mm; labial palpus pale ochreous with a few brown scales both basally and apically; abdomen ventrally WHEN DARRO EMO A GISEIIN a eno cou edo ea a pieli 3. Abdominal tergite 4 with a narrow yellow stripe proximally; tergites 5 and 6 each with a narrow dirty orange stripe distally; ventrally Siemens with a broad yellow stripe proximally ....5............... fukiensis sp. n. — Abdominal tergite 4 with two small orange spots lateroanteriorly; ventrally abdomen entirely dark brown with violet sheen .......... vietnamica sp. n. Paradoxecia gravis (Walker) (Figs 1—2) Aegeria gravis Walker, [1865]: 31. Type locality: “North China” [= eastern China, Shanghai ?]. Holo- type female (BMNH). Hampson 1919: 114 (Paradoxecia); dalla Torre & Strand 1925: 180 (Paradoxecia); Gaede 1933: 797 (Paradoxecia); Heppner & Duckworth 1981: 22 (Paradoxecia); Spatenka et al. 1993: 85 (Paradoxecia). Redescription. Female (holotype) (Fig. 1). Alar expanse 39.5 mm; body length 17.3 mm; forewing 17.7 mm; antennae broken off. Head: antennae broken off; frons rubbed out; labial palpus white ventrally and light brown dorsally; vertex dark brown; pericephalic hairs white. Thorax: patagia dark brown; tegula dark brown to black with a narrow yellow inner margin; meso- and metathorax nearly without scales, dark brown; thorax laterally grey-brown. Legs: fore coxa entirely dark brown; hind tibia dark brown with two tufts of rusty, longitu- dinal and apically pointed scales; inserting both medio-dorsally and apically; spurs dark brown. Abdomen: dark brown; tergite 1 laterally yellow; tergites 2—5 each with a yellow, narrow, distal margin; ventrally entirely dark brown. Forewing: entirely opaque, dark brown with violet sheen; cilia denuded. Hindwing: transparent with brownish tingle; veins and outer margin dark brown; discal spot undeveloped; outer margin narrow, ca. twice as narrow as cilia; cilia dark brown. Female genitalia. Not studied. Male. Unknown. Variability. Unknown. Diagnosis. This species is closest to P pieli Lieu and differs from it by the larger size (alar expanse up to 30.0 mm in pieli) and ventral coloration of the abdomen (sternites 1+2, 3, 4 and 5 each with a very narrow, yellow to pale yellow, distal margin in pieli). It can easily be distinguished from P vietnamica sp. n. and P fukiensis sp. n. by the coloration of the abdo- men (tergite 4 with two small orange spots latero-anteriorly in vietnamica sp. n. and tergite 4 with a narrow yellow stripe proximally, tergites 5 and 6 each with a narrow dirty orange stripe distally in fukiensis sp. n.). These three species are also distinguishable from each other .by the numerous details of the coloration of the body (see descriptions for P vietnamica sp. n. and P fukiensis sp. n.). Bionomics and habitat. Unknown. Distribution. Known only from the type locality: “North China” [eastern China, Shang- hai ?]. Material examined. 1 female (holotype), with labels as in Fig. 2 (BMNH). 62 O. G. Gorbunov & Y. Arita Paradoxecia pieli Lieu (Fig. 3) Paradoxecia pieli Lieu, 1935: 193, figs 1—39. Type locality: China, Hangchow [= eastern China, Zhejiang, Hangzhou]. Holotype male (probably lost, was kept in the author’s collection, Musée Heude, Shanghai). Naumann 1971: 55, figs 18, 57, 172 (as P gravis, misidentification); Heppner & Duckworth 1981: 22. Description. Female (Fig. 3). Alar expanse 30.0 mm; body length 15.0 mm; forewing 13.5 mm; antenna 5.2 mm. Head: antenna dark brown with purplish violet sheen; frons brown; labial palpus pale ochreous with a few brown scales both basally and apically; vertex brown; pericephalic hairs dark yellow dorsally and white laterally. Thorax: patagia brown with a large dark yellow spot lateroposteriorly; tegula brown with a few yellow scales apically; meso- and metathorax brown; thorax laterally brown with a small dark yellow spot at base of forewing. Legs: fore coxa entirely brown to dark brown with violet sheen; hind tibia greyish-brown with violet sheen, internally slightly paler, with two tufts of rusty, longitudinal and apically pointed scales, inserting both medio-dorsally and apically; spurs greyish-brown. Abdomen: brown with violet sheen; tergite 1 yellow laterally; tergites 2, 4 and 5 each with a narrow yellow margin distally; sternites 1+2, 3, 4 and 5 each with a very narrow, yellow to pale yellow, distal margin; anal tuft small, brown. Forewing: entirely opaque, brown with violet sheen, basally somewhat darker; cilia brown with violet sheen. Hindwing: densely covered with semitransparent scales with brownish hue, except on distal half where sparsely covered with brown scales; veins and outer margin brown; discal spot undeveloped; outer margin narrow, ca. twice as narrow as cilia; cilia brown with violet sheen. Female genitalia (Naumann 1971: fig. 172). 8th tergite relatively short but broad, well- sclerotized, with numerous long setae at distal margin; papilla anales relatively small, slightly sclerotized, with numerous short setae; apophysis posterior somewhat longer than apophysis - anterior; latter with distinct, long, ventral appendix; ostium bursae membranous; antrum short, broad, well-sclerotized; ductus bursae short and broad, membranous; corpus bursae ovoid, membranous, with an elongate signum. Male. Somewhat smaller: alar expanse 19.6—22.5 mm; body length 10.6—13.0 mm (Lieu 1935). Variability. Varies in individual size: alar expanse 19.6—30.0 mm; body.length 10.6—15.0 mm. Sometimes abdominal tergite 3 with a narrow yellow margin distally. Diagnosis. This species seems closest to P gravis (Walker, [1865]) and differs from it by the individual size (alar expanse 39.5 mm in gravis) and coloration of the abdomen ventrally (entirely dark brown in P gravis). It is possible that P pieli is only a form of gravis. Unfortu- nately, only a single female (holotype) of gravis is known and its genitalia were not studied. From P vietnamica sp. n. and P fukiensis sp. n., P pieli can easily be distinguished by the key given above. See also “Diagnosis” for comparisons of these species. Bionomics (after Lieu 1935). The host plant is Morus sp. (mulberry) (Moraceae). The larva lives inside the thin twig, occasionally in a thick branch. Its development requires one year; the imago appears in July. Habitat. Mulberry orchards. Distribution. Known only from eastern China, prov. Zhejiang. Material examined. 1 female, China, Hangchow [Hangzhou], 1. VII. 1934 (MUT). Paradoxecia vietnamica sp. n. (Figs 4, 8) Description. Female (holotype) (Fig. 4). Alar expanse 31.6 mm; body length 12.3 mm; forewing 15.5 mm; antenna 5.3 mm. Review of genus Paradoxecia 63 Figs 1—6: Paradoxecia spp. 1. P gravis Walker, [1865], holotype, female (BMNH). Alar expanse 39.5 mm. 2. Ditto, labels. 3. P pieli Lieu, 1935, female, China, Hangchow [Hang- zhou], 1. VII. 1934 (MUT). Alar expanse 30.0 mm. 4. P vietnamica sp. n., holotype, female (MHNG). Alar expanse 31.6 mm. 5. P fukiensis sp. n., holotype, female (ZFMK). Alar expanse 28.0 mm. 6. Ditto, paratype, male (ZFMK). Alar expanse 24.2 mm. Head: antenna dark brown to black with violet sheen; frons dark brown; labial palpus orange with a few pale yellow scales apically; vertex dark brown; pericephalic hairs yellow- orange dorsally and white laterally. Thorax: patagia dark brown to black with bronze-violet sheen, with a small orange spot lateroanteriorly; tegula, meso- and metathorax dark brown with violet sheen; thorax laterally dark grey with violet sheen. Legs: fore coxa dark brown with bronze-violet sheen, with a few orange scales basally; hind tibia dark brown with bronze-violet sheen, with two tufts of rusty, longitudinal and apically pointed scales; both inserting medio-dorsally and apically; spurs dark brown externally and white internally. Abdomen: dorsally dark brown to black with violet sheen; tergite 4 with two small orange spots lateroanteriorly; ventrally entirely dark brown with violet sheen; anal tuft small, dark brown to black with violet sheen. 64 O. G. Gorbunov & Y. Arita Forewing: black basally; costal margin dark brown with a narrow, longitudinal, light brown line; rest of surface covered with dark brown with violet sheen, with a few light brownish scales; transparent areas undeveloped, but covered with slightly lighter scales; cilia dark brown with bronzed sheen. Hindwing: transparent with yellowish hue; veins and outer margin dark brown; discal spot undeveloped; outer margin about twice as narrow as cilia; cilia dark brown with bronzed sheen. Female genitalia. (holotype, genital preparation No. GA—045) (Fig. 8). 8th tergite narrow, well-sclerotized, with numerous long setae at distal margin; papilla anales strongly sclerotized with curved outside ventral margin, with a few short setae; apophysis posterior somewhat shorter than apophysis anterior; latter without ventral appendix; ostium bursae membranous; antrum short, broad, slightly sclerotized; ductus bursae short and broad, membranous; corpus bursae ovoid, membranous, without signum. Male. Unknown. Variability. Unknown. Diagnosis. This new species can be distinguished from P pieli by the orange labial palpus (pale ochreous with a few brown scales in the species compared), coloration of the patagia (with a large dark yellow spot lateroposteriorly in pieli) and abdomen (tergites 2, 4 and 5, sometimes 3, each, with a narrow yellow strip distally in the species compared) and by the transparent hindwing (semitransparent with brownish hue, with distal half sparsely covered with brown scales in pieli). From P gravis, vietnamica sp. n. differs by the smaller size (alar expanse 39.5 mm in gravis) and by the coloration of the abdomen and hindwing. The new species is easily distinguishable from P fukiensis sp. n. by the coloration of the abdomen (tergite 4 with a narrow yellow strip proximally; tergites 5 and 6 each with a narrow dirty orange strip distally; laterally segment 1 yellow; ventrally sternite 4 with a broad yellow strip proximally in the species compared). Also, PR vietnamica sp. n. is clearly separated from P pieli and P fukiensis sp. n. by the female genitalia (compare fig. 9, and Naumann 1971: | fig. 172). Bionomics. The host plant is unknown. The moth has been netted near the end of May. 2 u. == Fig. 7: Wing venation of P fukiensis sp. n. Scale bar: 1.0 mm. Review of genus Paradoxecia 65 Figs 8—9: Female genitalia of Paradoxecia spp. 8. PR vietnamica sp. n., holotype (genital preparation No. GA—045). Scale bar: 0.5 mm. 9. P fukiensis sp. n., paratype (genital prepara- tion No. GA—075). Scale bar: 0.5 mm. Habitat. Unknown. Distribution. Known only from Vietnam. Material examined. 1 female (holotype), Vietnam, Pahia, 24. V. 1950, J. Romieux leg. (MHNG). Paradoxecia fukiensis sp. n. (Figs 5—6, 9—10) Description. Female (holotype) (Fig. 9). Alar expanse 28.0 mm; body length 14.0 mm; forewing 13.0 mm; left antenna broken off, right one with broken tip. 66 O. G. Gorbunov «€ Y. Arita Head: antenna dark brown to black with greenish sheen; frons dark brown with gold-green sheen, with a few dirty orange-yellow scales laterally; vertex dark brown with golden-green sheen; labial palpus pale yellow basally and dirty yellow distally; pericephalic hairs dirty yellow dorsally and pale yellow laterally. Thorax: patagia dark brown with purple-green sheen, with a few dirty yellow scales latero- posteriorly; tegula dark brown with green-violet sheen, with a narrow dirty orange inner margin; meso-, metathorax and thorax laterally dark brown with green-violet sheen. Legs: fore coxa dark brown with golden sheen, mixed with individual dirty orange scales; hind tibia dark brown with purple-violet sheen, with a few yellow-orange scales ventrally, with two tufts of rusty, longitudinal and apically pointed scales; both inserting medio-dorsally and apically; spurs dark brown. Abdomen: segments 1-3 black with purple sheen, all other segments dark brown; dorsally tergite 4 with a narrow yellow stripe proximally; tergites 5 and 6.each with a narrow dirty orange stripe distally; laterally segment 1 yellow; ventrally sternite 4 with a broad yellow stripe proximally; anal tuft small, dark brown tipped with light brown. Forewing: dark brown with bronzed sheen, mixed with individual light brown scales; costal margin narrowly dirty orange; anterior transparent area undeveloped; posterior transparent area very narrow and short, covered with semihyaline brownish scales with blue hue; external transparent area undeveloped, but scales between veins M3 and Cul somewhat lighter and with bluish hue; cilia dark brown. Hindwing: transparent but densely covered with translucent scales with slightly brownish hue; veins dark brown mixed with a few dirty orange scales; outer margin dark brown, narrow, about twice as narrow as cilia; discal spot undeveloped; cilia dark brown. Female genitalia. (paratype, genital preparation No. GA—075) (Fig. 9). 8th tergite relatively short and narrow, well-sclerotized, with numerous long setae at distal margin; papillae anales relatively small, slightly sclerotized, with numerous short setae; apophysis posterior somewhat shorter than apophysis anterior; latter with a short ventral appendix; ostium bursae membranous; antrum short, narrow, well-sclerotized; ductus bursae relatively long and broad, membranous; corpus bursae ovoid, membranous, with two oval signa. Male. Generally less robust than female. Colour pattern as in female (Fig. 6). Male genitalia. (paratype, genital preparation Nos GA—102 and 1570 YA) (Fig. 10). Uncus well-developed, finger-shaped, covered with simple short setae, well separated from tegumen; tegumen triangular, broad, without gnathos (Fig. 10a); valva broad, rounded distally, covered with simple short hairs and long setae (Fig. 10b); saccus short and broad, rounded basally (Fig. 10c); aedeagus narrow, long, about twice longer than valva, with long coecum penis and with a strong thorn distally where penis broadens dorsally; vesica with numerous minute spinules (Fig. 10d). Variability. Vary only in individual size, not in coloration. Size varies as follows. Males: alar expanse 23.0—25.0 mm; body length 11.0—12.0 mm; forewing 10.5—11.5 mm; antenna 4.5—5.0 mm; females: alar expanse 22.5—32.0 mm; body length 10.5—15.0 mm; forewing 9:5—145>smmaantenna4 0—52 mm. Diagnosis. By the coloration of the abdomen (bicoloured background coloration and bicoloured stripes) this species can not be confused with any other congeners. Additionally, by the shape of papilla anales and signum of the corpus bursae of the female genitalia, fukien- sis sp. n. is clearly distinguishable from pieli and vietnamica sp. n. Bionomics. The host plant is unknown. Imagines were collected during June at an altitude of about 2300 m. Habitat. Unknown. Distribution. Southeastern China, prov. Fujian. Known only from the type locality. Material examined. Holotype female, China, Fukien, Kuatun, 2300 m, 2740 N, 117—40 E, 13. VI. 1938, J. Klapperich leg. (ZFMK). Paratypes: 1 male, same locality as holo- type, 8. VI. 1938, J. Klapperich leg. (ZFMK); 2 males, same locality as holotype, 26. VI. 1936, Review of genus Paradoxecia 67 Ey A 5 EEE eS - 7 YE Fig. 10: Male genitalia of P fukiensis sp. n., paratype (genital preparation Nos GA—102 and ' 1570 YA). a. Tegumen-uncus complex. b. Valva. c. Saccus. d. Aedeagus. Scale bar: 0.5 mm. H. Hone leg. (ZFMK); 1 female, same locality as holotype, 8. VI. 1938, J. Klapperich leg. (CG); 1 female, same locality as holotype, 15. VI. 1938, J. Klapperich leg. (ZFMK); 1 male, 3 females, same locality as holotype, 16. VI. 1938, J. Klapperich leg. (male with genital prepa- ration No. 1570 YA, GA—102) (MUT, ZFMK); 1 female, same locality as holotype, 25. VI. 1936, H. Hone leg. (genital preparation No. GA—075) (ZFMK); 1 female, same locality as holotype, 27. VI. 1936, H. Hóne leg. (ZFMK); 1 female, same locality as holotype, 29. VI. 1938, H. Hone leg. (MUT). Acknowledgements We would like to express our cordial thanks to the following colleagues: Prof. Dr. C. M. Nau- mann and Dr. D. Stining (both Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn, Germany), Drs G. S. Robinson, K. Sattler, and M. Shaffer (all The Natural History Museum, London, Great Britain), Dr. D. B. Burckhardt (Muséum d’Histoire Natu- relle, Geneve, Switzerland), Dr. K. Spatenka (Prague, Bohemia) and Dr. S. I. Golovatch (Insti- tute for the Problems of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia) for help during the course of our studies on Oriental Sesiidae. We also have to thank Dr. T. D. Eichlin (California Department of Food and Agriculture, Plant Pest Diagnostics Branch, Sacramento, California, USA) for linguistic help. Zusammenfassung Die Glasfliigler der Gattung Paradoxecia Hampson, 1919 werden revidiert. Neben zwei neuen Arten aus Vietnam und SO-China werden die wenig bekannten Arten P gravis (Walker, [1865]) und P pieli Lieu, 1935 neu definiert. Ein Schlüssel erlaubt die Bestimmung aller heute bekann- ten Arten der Gattung. 68 OE Gorbunova Y Arta References Dalla Torre, K. W. von &E. Strand (1925): Aegeriidae. — In Strand, E., Lepidopterorum Catalogus, 31, 202 pp. W. Junk, Berlin. Fletcher, D. S. & I. W. B. Nye (1982): In Nye, I. W. B. (ed.), The generic names of moths of the world, 4, i-xiv, 192 pp. British Museum (Natural History), London. Gaede, M. (1933): Familie: Aegeriidae. — In Seitz, A., Die Gross-Schmetterlinge der Erde, II. Abteilung: Exotische Fauna, 10 (Die indo-australischen Spinner und Schwármer): 775—802, pl. 94—95. A. Kernen, Stuttgart. Hampson,G.F. (1919): A classification of the Aegeriadae [sic!] of the Oriental and Ethio- pian Regions. — Novit. Zool. 26: 46—119. Heppner, J. B. (1991): Faunal regions and the diversity of Lepidoptera. — Tropic. lepid. 2 (Suppl: Heppner, J. B. & W. D. Duckworth (1981): Classification of the superfamily Sesioidea (Lepidoptera: Ditrysia). — Smithsonian Contrib. Zool. 314: 1—144. Le Cerf, E (1917): Contributions a Pétude des Aegeriidae: Description et iconographie d'especes et de formes nouvelles ou peu connues. —Etud. Lépid. comp. 14: 137—388, pl. 475—481. Lieu, K. O. V. (1935): Study of a new species of chinese mulberry-borer, Paradoxecia piel, n. sp. (Lepidoptera, Aegeriidae). — Notes d’Entomologie Chinoise, Musée Heude (Shang- ha 1852095 5) pl: Naumann, C. M. (1971): Untersuchungen zur Systematik und Phylogenese der holarkti- schen Sesiiden (Insecta, Lepidoptera). — Bonn. zool. Monogr. 1: 1—190. Spatenka,K., Z. Lastuvka, O. Gorbunov, I. Tosevski& Y. Arita (1993): Die Syste- matik und Synonymie der paláarktischen Glasflügler-Arten (Lepidoptera, Sesiidae). — Nachr. entomol. Ver. Apollo, Frankfurt/Main, N.F. 14: 81—114. Walker, F., [1865]. Supplement [Part 1]. — In: List of... Lepidopterous Insects in. . . the British Museum, 31, 321 pp. British Museum, London. Oleg G. Gorbunov, Institute for Problems of Ecology and Evolution, Russian Academy of Sciences, Leninsky prospekt 33, Moscow V-71, 117071 Russia (Visiting Scholar of Zoological Laboratory, Faculty of Agriculture, Meijo University, Tem- paku-ku, Nagoya, 468 Japan). — Yutaka Arita, Zoological Laboratory, Faculty of Agriculture, Meijo University, Tempaku-ku, Nagoya 468 Japan. Bd. 4 S. 69-76 | Bonn, September 1997 Erster Nachweis einer Strepsipterenparasitierung bei Wanzen in Mitteleuropa (Insecta, Strepsiptera et Heteroptera) Albert Melber & Hans Pohl Abstract. Inthe nature reserve Liineburger Heide (Lower Saxony, FR Germany) a new Strepsipteran was discovered in 1993 (Malayaxenos trapezonoti Pohl & Melber, Strepsi- ptera, Corioxenidae), which is parasitizing the two ground bug species Trapezonotus arena- rius (L.) and 7. desertus Seidenst. (Heteroptera, Lygaeidae). This is the first record of para- sitation of Heteroptera by Strepsiptera in Central Europe. First results concerning the univoltine developmental cycle, the biology, and the local distribution of the parasite are presented. Key words. Insecta, Strepsiptera, Corioxenidae, Heteroptera, Lygaeidae, Northwest Germany, parasitation, life history. Einleitung Bei ökofaunistischen Untersuchungen im Bereich des Naturschutzgebietes Lünebur- ger Heide mit dem Ziel, Auswirkungen landschaftspflegerischer Eingriffe auf die Wirbellosenfauna abzuschätzen, wurden neben anderen Tiergruppen auch Wanzen (Heteroptera) bearbeitet. Bei der Präparation von Bodenfallenmaterial fanden sich in den Abdomina einiger Bodenwanzen (Lygaeidae) neben Dipterenlarven (unbe- kannte Tachinidae) verschiedene Entwicklungsstadien von Strepsipteren. Da für Mitteleuropa überhaupt keine Vertreter der Ordnung Strepsiptera mit Hete- ropteren als Wirte gemeldet sind (Kinzelbach 1978), mußte es sich hier um ein neues Taxon oder um einen Erstnachweis für diesen geographischen Raum handeln. Die Untersuchungen des Strepsipterenmaterials ergab, daß hier eine bisher unbekannte Art vorliegt, die von Pohl & Melber (1996) als Malayaxenos trapezonoti beschrieben wurde. Weil die Auswertung des gesamten Bodenfallenmaterials eine Reihe von Aussagen über die Biologie dieses Strepsipters erlaubt, sind im folgenden alle gewonnenen Daten über den neuen Vertreter dieser biologisch noch sehr unzureichend erforschten Gruppe zusammengestellt. Untersuchungsgebiete Die beiden Untersuchungsgebiete „Bockheber“ und „Oberhaverbeck“ liegen im Naturschutz- gebiet Lüneburger Heide (Niedersachsen) 6 km westlich bzw. 10 km nordwestlich von Schne- verdingen und sind nur rund 5 km voneinander entfernt (TK-Code mit Minutenfeldangabe: - 2825.3/13 bzw. 2825.4/01). Bei beiden untersuchten Flächen handelt es sich um trockene Calluna-Heiden auf Sand- untergrund (Genisto-Callunetum-typicum). Im Gebiet „Bockheber“ dominiert überalterte Calluna, was ein recht feuchtes Mikroklima in Bodennähe zur Folge hat, im Gebiet „Ober- haverbeck“ zeigten sich in den beiden Untersuchungsjahren 1993 und 1994 starke Vergrasungs- tendenzen mit Deschampsia flexuosa (L.) Trin. In beiden Fällen war eine ziemlich dicke Rohhumusauflage vorhanden. 70 A. Melber &H. Pohl Material und Methoden Sämtliche untersuchten Wirtstiere stammen aus Bodenfallenfángen. Die Fallen in der Bau- weise nach Melber (1987) waren jeweils ganzjährig exponiert und wurden halbmonatlich geleert. Als Tötungsflüssigkeit diente unverdünntes Ethylenglykol. Die gefangenen Tiere wur- den nach dem Auslesen des Falleninhaltes in 70%igem Ethanol konserviert. In den Jahren 1993 und 1994 waren im Gebiet „Bockheber“ 6 und im Gebiet „Oberhaverbeck“ 31 Boden- fallen aufgestellt. Ergebnisse In Bodenfallen auf den beiden Untersuchungsflächen „Bockheber“ und „Oberhaver- beck“ im Naturschutzgebiet Lüneburger Heide wurden in den Jahren 1993 und 1994 zahlreiche Imagines von Trapezonotus arenarius (Linnaeus) und 7. desertus Seiden- stücker (Heteroptera, Lygaeidae) gefangen, die mit verschiedenen Entwicklungs- stadien des Parasiten Malayaxenos trapezonoti Pohl & Melber (Strepsiptera, Corioxe- nidae) befallen waren. Aufgrund der relativ großen Anzahl gefundener Parasitierungsfälle können neben der Aussage, daß hier erstmals in Mitteleuropa eine Parasitierung von Wanzen durch Strepsipteren nachgewiesen wurde, auch zahlreiche Angaben zur Biologie des Para- siten gemacht werden. Morphologische Charakteristika der Parasitierung Sämtliche Entwicklungsstadien von Malayaxenos trapezonoti wurden im Abdomen der Wirtsimagines gefunden. Abgesehen von 2 Männchen, die im verpuppten Zustand vollständig im Inneren des Wirtsabdomens eingebettet waren, wurde nur 1 Männchen gefunden, dessen Cephalothorax in der Weise, wie es bei dieser Strepsi- pterengruppe üblich ist, intersegmental am Wirtsabdomen ausgebohrt war. Die Aus- trittstelle war in diesem Fall dieselbe wie bei den zahlreich gefundenen weiblichen Cephalothoraces: immer dorsal zwischen dem 2. und dem 3. sichtbaren Abdominal- tergit (morphologisch 3. und 4. Tergit), wo sich eine besonders breite sekundäre Intersegmentalmembran befindet, rechts und links der Abdomenmitte gleich häufig (Abb. 1). Die Cephalothoraces der Weibchen befanden sich, wie die aller bekannten Corio- xeniden, vollkommen unter den Hemielytren des Wirtstieres und sind daher ohne Präparation nicht sichtbar. Eine Begattung durch das Männchen ist wahrscheinlich trotzdem möglich, wobei das Abdomenende des Männchens seitlich um die Flügel- vorderkante des Wirtes herumgeführt wird. Dies hat Kirkpatrick (1937) für Corioxe- nos antestiae Blair gezeigt, eine Art, die Pentatomiden der Gattungen Antestia und Antestiopsis parasitiert, wo ebenfalls die weiblichen Cephalothoraces auf der Abdo- menoberseite vollständig unter den Deckflügeln der Wirtstiere liegen. Entwicklungszyklus von Wirt und Parasit Bei den beiden Wirtsarten Trapezonotus arenarius und T. desertus handelt es sich um univoltine Imaginalüberwinterer. In den Untersuchungsgebieten erfolgt die Eiablage nach der Überwinterung ab März, hauptsächlich im April und Mai, Larven treten ab Mitte Mai auf. Erste Imagines der neuen Generation erscheinen in der ersten Juni- hälfte. Strepsipterenparasitierung bei Wanzen 71 Abb. 1: Rasterelektronenmikroskopische Aufnahme von Malayaxenos trapezonoti, A: Abdo- men von Trapezonotus desertus mit 2 herausgebohrten Malayaxenos, links Weibchen, rechts männliches Puparium; B: männliches Puparium; C: Weibchen. Maßstab: A: 1 mm, B, C: 0,1 mm. Malayaxenos trapezonoti überwintert in den Wirtsimagines als Sekundärlarve. Dieses Entwicklungsstadium war sowohl vor der Winterruhe des Wirtes im August und September als auch danach im März und April in den laufaktiven Wirtsimagines zu finden. Herausgebohrte Parasitenweibchen und -männchen ließen sich dann ab Mitte März bzw. Mitte April nachweisen. Reife Weibchen mit fertig ausgebildeten Primärlarven wurden Mitte Mai bis Mitte Juli festgestellt. Die Primärlarven treffen also bei ihrer Wirtssuche vor allem auf die frisch entwickelten Imagines der neuen Generation, ob sie auch schon in ältere Wirtslarven eindringen, die zu dieser Zeit noch in großer Anzahl vorhanden sind, konnte nicht überprüft werden. Eine graphische Zusammenfassung der bisherigen Erkenntnisse über den jahres- zeitlichen Verlauf der Wirts- und Parasitenentwicklung zeigt die Abb. 2. Spezifität der Parasitierung Die beiden Wirtsarten Trapezonotus arenarius und T. desertus stehen sich taxono- misch sehr nahe; 7! desertus wurde erst 1951 von 7. arenarius abgetrennt. Beide Arten sind zwar in ganz Nordwestdeutschland verbreitet, ihre relative Häufigkeit wird aber 72 A. Melber &H. Pohl deutlich von einem Kontinentalitätsgradienten beeinflußt: In Calluna-Sandheiden tritt im Westen (Emsland) nur 7! desertus auf und erst in Zentralniedersachsen kommt 7: arenarius in ganz geringen Anteilen hinzu, um dann im Osten (z. B. Um- gebung Gifhorn) Anteile um 25 % zu erreichen. Noch weiter im Osten an der Elbe (Wendland) kommt praktisch nur noch 7! arenarius vor. In den beiden Untersuchungsgebieten lag 1993/94 das Verhältnis von 7. arenarius : T. desertus-Imagines in Bodenfallen bei 1:12,2. Das Verhältnis parasitierter Indivi- duen unter den beiden Lygaeiden-Arten war 1:10,0. Da es sich hierbei aber nur um insgesamt 37 Parasitierungsfälle handelt, läßt sich keine signifikante Präferenz für eine der beiden Wirtsarten erkennen. Andere Lygaeidenarten in den Untersuchungsgebieten, die als potentielle Wirts- arten ebenfalls untersucht wurden (Macrodema micropterum Curtis, Pterotmetus staphyliniformis Schilling, Rhyparochromus pini Linnaeus, Scolopostethus decoratus Hahn und Stygnocoris sabulosus Schilling), waren nie parasitiert, so daß zumindest eine strenge Gattungsspezifität der Parasitierung vorzuliegen scheint. Geschlecht des Wirtes und Parasitierung Daß keine Präferenz des Parasiten für ein bestimmtes Geschlecht der Wirtstiere vor- handen ist, konnte ebenfalls gezeigt werden. Das Geschlechterverhältnis beider Trapezonotus-Arten (Männchen zu Weibchen) bei den Bodenfallenfängen insgesamt lag bei 1:1,50, bei den parasitierten Tieren bei 1:1,57. Mehrfachparasitierung Bei 5 der insgesamt 37 untersuchten Parasitierungsfälle lag eine Mehrfachparasitie- rung vor. 2 7. desertus-Männchen und 1 Weibchen waren mit je 2 reifen, dorsal links und rechts herausgebohrten Parasitenweibchen besetzt. In einem 7. desertus-Weib- überwinterte neue Wirtsgeneration Wirtsgeneration ae Laufaktivitatsperiode des Wirtes Monate N nA 2. a Ad u a 2. Monatshälften o Ya 2 1 2 — 2] 1 Larven Sekundärlarven c Primarlarven 3; 32 00 verpuppte, unreife OQ D E QQ mit Eiern go 90 mit Primárlarven $3 ,leere” OO m exe) verpuppte JO leere Puparien @ = tatsächlich nachgewiesene Entwicklungsstadien O = aufgrund der er m vermutete Entwicklungsstaen Abb. 2: Jahreszeitliches Auftreten verschiedener Entwicklungsstadien von Malayaxenos trape- zonoti in Beziehung zum Jahreszyklus der Wirtsarten. Strepsipterenparasitierung bei Wanzen 73 Anzahl Wirtsimagines je Monatshälfte in Bodenfallen Parasitierungsgrad % Monatshälften Monate Abb. 3: Beziehung zwischen der Anzahl der insgesamt in Bodenfallen (1993 + 1994 gefange- nen Imagines von Trapezonotus desertus und T. arenarius beider Untersuchungsgebiete und dem Anteil der durch Malayaxenos trapezonoti parasitierten Individuen. chen fand sich neben einem reifen Malayaxenos-Weibchen, welches normal dorsal ausgebohrt war, ein verpupptes Männchen im Inneren des Wirtsabdomens. In einem fünften Fall wurde ein 7. desertus-Weibchen gefunden, welches neben 2 verpuppten Parasitenweibchen noch eine mittelgroße Sekundärlarve enthielt. Offenbar können 2 Malayaxenos-Individuen in einem Wirtstier zur vollen Entwicklung gelangen, für 3 Parasiten gleichzeitig reicht dann aber die Nahrungsressource nicht mehr aus. Einfluß der Parasitierung auf den Wirt Offensichtlich wird die Laufaktivität des Wirtes, die sich in der Bodenfallenfangrate widerspiegelt, durch die Parasitierung beeinflußt. Das normale jahreszeitliche Muster der Laufaktivität (Abb. 3, oben) zeigt für beide Trapezonotus-Arten einen Aktivitätsbeginn im März, ein Maximum in der ersten Maihälfte und ein Ausklingen der Laufaktivität aufgrund des Absterbens der überwinterten Generation Ende Juli. - Zu dieser Zeit kommen dann schon wieder Imagines der neuen Generation hinzu (in Abb. 3 nicht dargestellt). Der relative Anteil parasitierter Wirtsindividuen in den Bodenfallen wird aber nun mit ansteigender Bodenfallenfangrate immer geringer (Abb. 3, unten), was nur durch eine reduzierte Laufaktivität solcher Tiere in der 2. Aprilhälfte und im Mai erklärbar ist. Wie der Abb. 1 aufgrund des Auftretens her- ausgebohrter Malayaxenos-Weibchen und dem Vorliegen reifer Männchen zu entneh- 74 A. Melber € H. Pohl men ist, dürfte dieser Zeitraum die Periode der Kopulation des Strepsipters sein. Zu dieser Zeit sind also die parasitierten Wirtsimagines deutlich weniger laufaktiv als die unparasitierten. Die Frage, inwieweit die Reproduktionskapazität und Vitalität der Wirtstiere durch die Parasitierung beeinflußt wird, läßt sich nur grob abschätzen. Festzustellen ist, daß in den einfach parasitierten Trapezonotus-Imagines die Gonaden nach Ausreifen der Parasiten immer noch teilweise erhalten und aktiv sind. Es wurden bei den reifen Wirtsweibchen neben den Parasiten in der Regel immer noch reife Eier in Teilen des Ovars gefunden; entsprechend verfügten parasitierte Trapezonotus-Männchen immer noch über Rile reifer Hoden. Ob solchen Tieren aber noch eine erfolgreiche Kopula- tion bzw. Eiablage möglich ist, kann nicht beantwortet werden. Bei den Mehrfach- parasitierungen waren im Wirtsabdomen keine nennenswerten Organreste mehr zu finden, hier erfolgt also eine vollständige parasitäre Kastration. Wie die recht hohen Fangraten parasitierter Wirtstiere gegen Ende der Laufaktivi- tätsperiode (Juni/Juli) zeigen, wird die Lebensdauer der Trapezonotus-Imagines durch die Parasitierung nicht nennenswert verkürzt, im Gegenteil könnte sogar eine Seneszenzverzögerung für das Wiederansteigen des Anteils parasitierter Tiere in den Bodenfallen zu dieser Zeit verantwortlich sein. Regionale Verbreitung von Malayaxenos trapezonoti Um erste Aussagen über die Häufigkeit und Verbreitung von Malayaxenos trapezo- noti zu ermöglichen, wurde älteres Bodenfallenmaterial aus Calluna-Heiden in Niedersachsen, in dem höhere Trapezonotus-Fangzahlen auftraten, auf Stylopisie- rung hin untersucht. Wie die Tab. 1 zeigt, wurden trotz teilweise hoher Trapezonotus- Individuenzahlen keine weiteren Standorte mit Malayaxenos-Parasitierung gefun- den. Sogar in Calluna-Heiden, die in enger Nachbarschaft zu den beiden Unter- suchungsgebieten mit Malayaxenos-Vorkommen liegen, wurde die neue Strepsipte- renart nicht nachgewiesen. Malayaxenos trapezonoti scheint somit also nur in räum- lich sehr engumgrenzten Populationen vorzukommen. Diskussion Die bis jetzt vorliegende Auswertung von 2 Bodenfallenjahrgängen von 2 Standorten mit insgesamt 37 Parasitierungsfällen erlaubt natürlich noch keine vollständige Übersicht zur Biologie von Malayaxenos trapezonoti. Grundlegende Tatsachen aller- dings, wie z.B. das Vorliegen einer univoltinen Entwicklung, können bereits als ge- sichert gelten. Etwas unklar bleibt die Lebensweise der Männchen. Insgesamt wurden neben 30 Strepsipterenweibchen nur 4 Männchen (als Puppen oder leere Puparien) nachgewie- sen. Bei 2 von diesen Männchen-Funden waren die fertig verpuppten und skleroti- sierten Tiere vollständig im Wirtsabdomen eingebettet und der Cephalothorax war nicht herausgebohrt, was bedeutet, daß diese Männchen nicht mehr den Wirt verlas- sen konnten. Vielleicht befindet sich Malayaxenos trapezonoti in einer phylogeneti- schen Entwicklung in Richtung auf eine Parthenogenese, worauf der stark reduzierte Männchen-Anteil in der Population und das Auftreten gewissermaßen „funktions- loser“ Männchen hindeuten könnte. Andererseits zeigen die beiden anderen Männ- Strepsipterenparasitierung bei Wanzen 75 Tabelle 1: Vorkommen von Trapezonotus desertus und T: arenarius-Imagines, die von Malayaxenos trapezonoti parasitiert waren, in 8 verschiedenen Untersuchungsgebieten in Niedersachsen. Anzahl untersuchter gefundene Individuen Parasitierungsfálle Untersuchungsgebiet (mit TK 25- Code) Untersuchungsjahr Tútsberg 1993 + 94 (2825.3) Bockheber 1993 + 94 (2825.3) Oberhaverbeck 1993 + 94 (2825.4) Oberhaverbeck II 1986 (2825.4) Pietzmoor-Heide 1993 (2924.2) Bokeler Heide 1986 (3229.1) NSG Heiliger Hain 1986 + 87 (3329.3) Helstorf-Reiterheide 1989 + 90 (3423.2) chen-Nachweise, daß höchstwahrscheinlich normale Männchen vorhanden sind. Für Corioxenos antestiae konnte Kirkpatrick (1937) zeigen, daß 1 Männchen mindestens 12 Weibchen begatten kann, daher reicht ein geringer Männchen-Anteil für eine nor- male bisexuelle Vermehrung aus. | Da allen hier dargestellten Ergebnissen ausschließlich Bodenfallenmaterial zu- erundeliegt, könnte natürlich ein geringerer Männchen-Anteil bei Malayaxenos auch durch eine reduzierte Laufaktivität von Wirten, die von männlichen Parasiten befallen sind, zustande kommen. Danksagung Herrn Dipl.-Biol. Ludger Schmidt und Mitarbeitern der Norddeutschen Naturschutzakademie in Schneverdingen, insbesondere Herrn Dr. Johannes Prüter, sei für die Überlassung von Material und die Mithilfe herzlichst gedankt. Zusammenfassung An zwei eng benachbarten Calluna-Heidestandorten im Naturschutzgebiet Lüneburger Heide ‘wurde 1993 und 1994 bei Trapezonotus desertus und T: arenarius (Heteroptera, Lygaeidae) erstmals für Heteropteren in Mitteleuropa eine Parasitierung durch Strepsipteren nachgewie- sen. Es handelt sich hierbei um Malayaxenos trapezonoti Pohl & Melber, 1996 (Strepsiptera, Corioxenidae). Die äußerlichen Charakteristika der Parasitierung werden beschrieben und anhand von 37 Parasitierungsfällen der univoltine Entwicklungszyklus des Parasiten rekon- struiert. Eine Präferenz für eine der beiden Wirtsarten liegt offenbar nicht vor, wie auch nicht für eines der Wirtsgeschlechter. Das Auftreten von Doppel- und Dreifachparasitierung wird 76 A. Melber &H. Pohl beschrieben. Durch die Parasitierung wird die Laufaktivitát der Wirtsimagines, gemessen anhand der Bodenfallen-Fangrate im Zeitraum der Kopulationsflüge von Malayaxenos trape- zonoti, drastisch reduziert. Eine weitere Auswirkung der Parasitierung ist eine partielle para- sitäre Kastration der Wirtsimagines. Die erfolglose Suche nach Malayaxenos trapezonoti an anderen Calluna-Heidestandorten in Niedersachsen mit individuenreicheren Vorkommen von Trapezonotus-Arten zeigt, daß diese Strepsipterenart offenbar nur sehr lokal verbreitet ist. Literaturverzeichnis Kinzelbach, R. K. (1978): Fächerflügler (Strepsiptera). — In Dahl, F. (Hrsg.), Die Tierwelt Deutschlands, Bd. 65: 166 pp., Gustav Fischer, Jena. Kirkpatrick, F. W. (1937): Studies on the ecology of coffee plantations in E. Africa. — Trans. R. Ent. Soc. London 86: 281— 341. Melber, A. (1987): Eine verbesserte Bodenfalle. — Abh. Naturw. Ver. Bremen 40: 331 — 332. Pohl, H. & A. Melber (1996): Verzeichnis der mitteleuropäischen Fächerflügler und die Beschreibung einer neuen Art der Gattung Malayaxenos Kifune 1981 (Insecta: Strepsi- ptera). — Senckenbergiana biol. 75: 171—180. Dr. Albert Melber, Lehrgebiet Zoologie-Entomologie, FB Biologie, Universität Han- nover, Herrenhäuser Str. 2, D-30419 Hannover. — Dipl.-Biol. H. Pohl, Institut für Zoologie, Technische Hochschule Darmstadt, Schnittspahnstr. 3, D-64287 Darm- stadt. Bd. 4 S. 71-86 | Bonn, September 1997 Post-eclosion heterochrony in the maturation of the adult females of a termitophilous fly (Diptera, Phoridae) R. H. L. Disney Abstract. Many termitophilous adult insects postpone completion of the expansion of parts, such as the head and femora, until long after eclosion from their pupae. Conse- quently adults of different ages may differ markedly in appearance. Among the Termitoxe- niinae the most extreme post-eclosion changes occur in the females of Clitelloxenia ass- muthi (Wasmann 1902). Documentation of these changes lead to the proposal that five species be synonymised with this species. It has been argued (Assmuth 1913) that the females of this species are likely to ingest haemolymph from the older nymphs of their termite hosts. If so, ingestion of termite-nymph hormones might delay the cessation of expansion and sclerotization of parts of these flies following eclosion. At the least it is concluded that darkening of the cuticle cannot be treated as a reliable indicator of scleroti- zation. Probably darkening and hardening are decoupled in these flies. Key words. Diptera, Phoridae, Termitoxeniinae, synonyms, heterochrony, teneral state, Isoptera, Termitidae. Introduction Heterochrony is the “dissociation, during development, of factors of shape, size, and maturity, so that organisms mature in these respects at earlier or later growth stages” (Nichols 1989). In evolutionary terms it has been defined as “an evolutionary change in the onset or timing of development of a feature relative to the appearance or rate of development of the same feature in the ontogeny of an ancestor” (Lincoln et al. 1982). It has evidently been a widespread mechanism giving rise to novelty during evolution (Matsuda 1987). Among termitophilous insects heterochrony tends to manifest itself as a tendency for parts of certain adults to remain in the teneral state for a prolonged period after eclosion from the pupa. A general review of termitophilous insects (Kistner 1982) has shown that heterochrony occurs among termitophilous Carabidae, Cecidomyiidae, Phoridae, Scarabaeidae and Staphylinidae. In this paper a striking case of an Orien- tal species of Termitoxeniinae (Diptera, Phoridae) is documented and discussed. This study has been made possible by my access to the extensive collections of termitophi- lous Phoridae made by Dr David H. Kistner (California State University, Chico) over many years. In addition I am grateful to the following for the loan of type material in their cares: Dr. H. Ulrich, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn; Dr .F. N. Dingemans-Bakels, Natuurhistorisch Museum, Maastricht; and Dr B. R. Pitkin, the Natural History Museum, London. I am grateful to my colleague Dr Simon Maddrell for useful discussions. My studies are currently funded by a grant from the Leverhulme Trust, to my colleague Dr William Foster for a Research Associate to work on termitophilous Phoridae, and the Isaac Newton Trust (Trinity College, Cambridge). I am grateful to W. M. Lee (Zoology Department, Cambridge University) for his skilled operation of the Scanning Electron Microscope. 78 RSHSESDismnes; Post-eclosion changes in Termitoxeniinae The taxonomy of the aberrant termitophilous Termitoxeniinae, of the famly Phori- dae, is currently undergoing revision. This follows the realisation that the males, the flying-stage females and flightless-stage females had been erroneously assigned to three separate subfamilies by earlier authors (Disney 1989, 1990, 1992, 1994, Disney & Cumming 1992, Dessart 1993). It had long been recognised that the females undergo post-eclosion changes from a stenogastric to a physogastric stage (e.g. Ass- muth 1913, Mergelsberg 1935). It has only recently been realised that the early steno- gastric females shed their wing membranes on entering their host-termite’s nest (Disney & Cumming 1992). Furthermore the procurement of mating pairs of an Afrotropical species has allowed reconstruction of the complete sequence of changes that occur in the adult females after eclosion. This has prompted a closer look at a taxonomically confused complex of six ‘species’ assigned to the Oriental genus Clitelloxenia Kemner, and a re-evaluation of these species in terms of most being different stages in a series of post-eclosion changes within a single species. Clitelloxenia assmuthi (Wasmann 1902) Termitoxenia assmuthi Wasmann 1902: 151. Holotype Q, INDIA: Khandala, near Bombay (Natural History Museum, Maastricht) [presumed cotype material examined] Termitoxenia peradeniyae Wasmann 1913: 20. Holotype Y, SRI LANKA: Peradeniya (Natural History Museum, Maastricht) [examined]. Syn. nov. Termitoxenia clitellaria Schmitz 1915: 36. Holotype Q, SRI LANKA: Maha, Illupalama, Anuradha- pura (Museum Alexander Koenig, Bonn) [examined]. Syn. nov. Termitoxenia longiceps Schmitz 1915: 36. Holotype ©, MALAYSIA: Malacca (Natural History — Museum, Maastricht — only two slide-mounted legs survive) [examined]. Syn. nov. Termitoxenia hemicyclia Schmitz 1931: 176. Holotype Q, JAVA, Buitenzorg (Museum Alexander Koenig, Bonn and also the Natural History Museum, London — a series of presumed cotypes) [series examined]. Syn. nov. Clitelloxenia marshalli Schmitz 1938: 35. Holotype Q , INDIA: Bangalore, Mysore (Museum Alexander Koenig, Bonn — one wing stump; rest in Natural History Museum, London) [examined]. Syn. nov. A mature physogastric female from Java is shown in Figs 1 and 2. It is postulated that the material treated here as a single species exhibits a more marked degree of post-emergence change than most Termitoxeniinae. It is proposed that a failure to appreciate the extent of these changes in the females has misled early workers into describing as new species specimens which merely represented different stages of species already known. In most cases only limited material was available to these early workers. The larger samples now available, mainly because of the collections made by Dr David Kistner, serve to indicate that there are great differences between flies of different ages within a single sample, but litte, if any, difference between flies of the same age from different samples. There is thus now no doubt that C. clitellaria and C. marshalli should be treated as synonyms. The small differences highlighted by Schmitz (1916, 1938) clearly lie within the ranges of variation to be found in a single sample of flies from the same termite nest. The removal of two species by synonymy still leaves us with four “species”, which are based on geographical region — the Indian subcontinent, Sri Lanka, Java and Malaysia. Aggregated samples (of Kistner’s material plus specimens in museums) Maturation heterochrony in a termitophilous fly 79 imm 4@kU 538E1 0672700 SE Figs 1—2: Clitelloxenia assmuthi mature physogastric female. 1, lateral view of head, 2, whole fly from above. 80 RATE Disney, Fig. 3: Clitelloxenia assmuthi females, posterior faces of hind femora. The youngest fly at top and oldest below. (Scale bar = 0.1 mm). from each of these four regions look different from each other. However, this seems to be primarily due to the different age structures of the total sample of specimens available for study from each region. Table 1 presents the length frequencies of the hind femora for females from the four regions. The differences between these aggregated samples for the four regions could be explained by any one of the following hypotheses: (1) An accident of sampling, such that the aggregated samples from each region happen to represent populations of different ages. (2) The four regions are characterised by four different taxa, species or subspecies, with different growth characteristics. Maturation heterochrony in a termitophilous fly 81 Fig. 4: Clitelloxenia assmuthi female, left face of head. s = anterior (lower) supra-antennal bristle; r = rear of head. (Scale bar = 0.1 mm). (3) A single species is subjected to different phenotypic modifications of its growth characteristics in the different regions, possibly due to different species of Odontoter- mes serving as its hosts in the four regions. The relationship between age of fly and length of hind femur is indicated by Fig. 3. This shows the hind femora of females from a single sample from a nest of Odon- totermes grandiceps Holmgren from Bogor, Java (collected by Dr Kistner, 20 June 1977). The fly at the top had two developing eggs only 0.44 mm long, while that at the bottom had a single egg 0.93 mm long. In the topmost specimen the femur is only lightly tinged brown. In the bottom two specimens the femora are dark brown. Furthermore, as indicated by these two specimens, the larger the femur the more likely it is to exhibit distortion of its basic shape. Fig. 4 depicts the head of a female from Sri Lanka. The basiproboscis is seemingly well sclerotized and even more so the occipital region. Fig. 5 depicts the development of the occipital region in a series of females of the same series as for the femora depicted in Fig. 3. In Fig. 6 the ratio of the head length to length of hind femur is plotted for the 82 females of Table 1. The head length was measured from the anterior supra-antennal bristles to the back of the head (Fig. 4). Such a tight scatter of points, along a single straight line, would be highly improbable if the hypothesis of two or more sibling species were to be favoured, to explain the differences in Table 1. Examination of slide-mounts of females made from the material collected by Dr Kistner, and the measurement of the lengths of the dominant developing oocytes in particular, has allowed certain inferences to be drawn. Typically two oocytes start to develop in the young stenogastric female. In a few cases both mature, eventually attaining a length exceeding 0.8 mm. Typically, however, one ceases to grow as the other starts to become more elongate (e.g. Fig. 7). The oocyte whose development is halted frequently proceeds to degenerate, but in some cases it seems to persist in a healthy but static state. Whether it will resume normal development when the first 82 ROTA Diisinieny Fig. 5: Clitelloxenia assmuthi left faces of heads of females of different ages, the youngest at top (from the same series as in Fig. 3). (Scale bar = 0.1 mm). Table 1: Recorded lengths of the hind femora of samples of females of Clitelloxenia assmuthi from different regions. length in mm 0.3—0.4 0.4—0.5 0906 0.6—0.7 0.7—0.8 0.8—0.9 India + Pakistan Sri Lanka Java Malaysia TOTALS 1 28 28 15 Maturation heterochrony in a termitophilous fly 83 Table 2: Recorded lenths of developing eggs in relation to lengths of hind femora of females of Clitelloxenia assmuthi from different regions. length of developing oocyte (egg) in mm Region 0.2-0.4 0.4—0.6 0.6—0.8 0.8—1.0 Pakistan Sri Lanka Malaysia Java TOTALS Pakistan Sri Lanka Malaysia Java TOTALS y = 5.0454e-2 + 0.57937x RA2 = 0.751 0.6 [a] n=82 Head length (mm) 0.3 0.4 0.5 0.6 0.7 0.8 0.9 length of hind femur (mm) Fig. 6: Clitelloxenia assmuthi females, head lengths plotted against lengths of hind femora. egg has been laid is not known. It needs emphasizing that when flies with oocytes and eggs of different sizes are arranged in order, from smallest to largest, then this correlates with the degree of differentiation of the egg. In flies with the shortest hind femora developing oocytes are not discernible in my slide-mounted specimens. In those with intermediate length femora the contents of the developing oocytes or eggs are little, if any, differentiated. In those with the longest femora not only has the 84 Re Eee Disney, Fig. 7: Clitelloxenia assmuthi the two developing oocytes in a single female. The upper oocyte (of the left side) has ceased growing, the lower egg (of the right side) is nearing maturity. (Scale bar = 0.1 mm). development of one egg halted, but in the other egg there is increasing differentiation of the contents, with the formation of an increasingly distinct chorion. The most mature eggs exhibit very little variation in length. Smaller flies do not have smaller mature eggs than larger flies, they have more immature eggs (which are therefore smaller) than are to be found in larger flies. In a few large flies no developing eggs could be observed. In these cases the appearance of the abdomen suggested that the lack of an egg was probably due to it having been recently laid. Measurements of developing eggs found are presented in Table 2. The data in Table 2 provide no support for the hypotheses based on assuming that the females in the four regions have different growth characteristics. The most parsimonious conclusion, therefore, is that hypothesis (1) an accident of sampling, is sufficient to explain the differences between the samples from the four regions. Until defensible differences can be demonstrated, therefore, it is proposed to treat the material from the four regions as belonging to a single species, which continues to lengthen its femora and head right up to the oviposition of its first mature egg at least. Thus C. peradeniyae, C. longiceps and C. hemicyclia are herewith proposed as further synonyms of C. assmuthi. Maturation heterochrony in a termitophilous fly 85 Discussion Clitelloxenia assmuthi exhibits the most extreme post-eclosion changes in the Termi- toxeniinae studied to date. It is perhaps relevant to consider an inference proposed by Assmuth (1913). On the basis of the appearance of the gut contents and the struc- ture of the mouthparts, he suggested that this species probably ingests haemolymph from older nymphs of its termite host. If so, such behaviour would ensure ingestion of hormones prevalent in the blood of these juvenile termites. These termite hormo- nes might then serve to delay cessation of the expansion and sclerotization of the head and femora in these flies. The continuing growth of the occipital region of the head in Clitelloxenia assmuthi after the apparent onset of sclerotization is most surprising. It is certainly evident that the darkening of the femora and occipital region is not correlated with a cessation of growth. At the least this suggests that darkening may be an unreliable indicator of sclerotization. The two familiar indicators of sclerotization, hardening and darkening, may not necessarily be coupled. Indeed it is established that sclerotization sometimes occurs without darkening, as in the case of albino insects (Chapman 1991). It is suggested that in these flies darkening is initiated earlier than hardening. If darkening and hardening are not decoupled in these flies then it would be neces- sary to postulate some reversal of hardening to allow continuing growth. This would seem most improbable. Zusammenfassung Bei manchen termitophilen Insekten wird das Wachstum von Körperteilen, wie Kopf und Femora, erst lange nach der Imaginalhäutung abgeschlossen, mit der Folge, daß Imagines ver- schiedenen Alters recht verschieden aussehen können. Was die Termitoxeniinae betrifft, wur- den die stärksten postmetabolen Veränderungen der äußeren Gestalt bei den Weibchen von Clitelloxenia assmuthi (Wasmann 1902) beobachtet. Auf Grund von Korrelation und Häufig- keitsverteilung der Merkmale werden fünf weitere beschriebene Arten als Altersstadien von C. assmuthi gedeutet und mit dieser synonymisiert. Assmuth (1913) vermutete, daß die Weibchen dieser Art Hämolymphe von älteren Larven der Wirtstermite saugen. Sollte dies zutreffen, so wäre es denkbar, wenn auch unwahrscheinlich, daß der Abbruch des postmetabolen Wachs- tums und die Aushärtung der Cuticula durch Wirkung des aufgenommenen Juvenilhormons verzögert werden. Die Melanisierung der Cuticula setzt lange vor dem Stillstand des Wachs- tums ein und erfolgt demnach offenbar unabhängig von der Sklerotisierung. References Assmuth, J. (1913): Termitoxenia assmuthi Wasm.; Anatomisch-histologische Unter- suchung. — Nova Acta Acad. caesar. leop. carol. 98: 191—316. Chapman, R. F. (1991): General anatomy and function. In: CSIRO. The Insects of Aus- tralia. Second edition. — Melbourne University Press, Victoria. Chapter 2: 33—67. Dessart, P. (1993): Histoire des Termitoxéniides ou les errements de la science. — Natura- listes belg. 74: 61—75. Disney, R. H. L. (1989): The ‘missing’ males of the Alamirinae/Termitoxeniinae complex and the supposed hermaphroditism in Termitoxeniinae (Diptera: Phoridae). — Sociobio- logy 16: 109—120. Disney, R. H. L. (1990): Revision of the Alamirinae (Diptera: Phoridae). — Syst. Ent. 15: 305—320. Disney, R. H. L. (1992): ‘Flightless’ flies and fanciful ‘facts’. — Antenna 16: 150—158. Disney, R. H. L. (1994): From field studies to taxonomy. — Field Studies 8(2): 197—216. 86 RRE DIS Disney, R.H. L. & M.S. Cumming (1992): Abolition of Alamirinae and ultimate rejection of Wasmann’s theory of hermaphroditism in Termitoxeniinae (Diptera, Phoridae). — Bonn. zool. Beitr. 43: 145—154. Kistner, D. H. (1982): The social insect’s bestiary. In: Hermann H. R. editor. Social Insects. Volume III. — Academic Press, New York. Chapter 1: 1—244. Lincoln, R. J.,G. A. Boxshall & P. F. Clark (1982): A dictionary of ecology, evolution and systematics. — Cambridge University Press, Cambridge, 298 pp. Matsuda, R. (1987): Animal Evolution in Changing Environments with Special Reference to Abnormal Metamorphosis. — John Wiley & Sons, Chichester, 355 pp. Mergelsberg, O. (1935): Uber die postimaginale Entwicklung (Physogastrie) und den Hermaphroditismus bei afrikanischen Termitoxenien (Dipt.). — Zool. Jb. Anat. 60: 345 —398. Nichols, S. W. (1989): The Torre-Bueno Glossary of Entomology.-— New York Entomologi- cal Society/American Museum of Natural History, New York, 840 pp. Schmitz, H. (1915): Neue Gattungen und Arten termitophiler Phoriden und Termitoxenii- den. — Soc. ent., Stuttgart 30(7): 35—36. Schmitz, H. (1916): Neue termitophile Dipteren aus den Familien der Termitoxeniiden und Phoriden. — Zool. Jb. Syst. 39: 211—266. Schmitz, H. (1931): Neue termitophile Dipteren von Buitenzorg, Java. — Natuurhist. Maandblad 20: 176. Schmitz, H. (1938): Beitráge zu einer Monographie der Termitoxeniidae (Diptera) I. — Bro- teria 34: 22—40. Wasmann, E. (1902): Termiten, Termitophilen und Myrmekophilen, gesammelt auf Ceylon, mit anderm ostindischen Material bearbeitet. — Zool. Jb. Syst. 17: 99—164. Wasmann,E. (1913): Revision der Termitoxeniinae von Ostindien und Ceylon. — Annls Soc. ent. Bells. 57: 1622. Dr R. H. L. Disney, Field Studies Council Research Fellow, University Department - of Zoology, Downing Street, Cambridge, CB2 3EJ, England. Recognition of a sibling species of the Australian Eutermiphora abdominalis Lea (Diptera: Phoridae) R. H. L. Disney Abstract. Eutermiphora watsoni sp. n. is distinguished from the type species of the genus, E. abdominalis Lea 1911. The females of both species live in the colonies of Nasuti- termes fumigatus (Brauer 1865) in New South Wales. The taxonomic significance of the position of the pre-apical setae of the palps is discussed, with particular reference to the problem of the affinities of the Termitoxeniinae. Key words. Diptera, Phoridae, new species, Isoptera, Termitidae. Introduction The aberrant termitophilous genus Eutermiphora Lea (1911) (Diptera: Phoridae) was established for a single female from a termite’s nest at Sydney, New South Wales, Australia. Unfortunately Lea overlooked both the wing stumps and halteres that, in mature females, are concealed beneath a forward extension of the anterior region of the abdomen that covers much of the thorax. The result was to cause confusion in keys to genera and in attempts to determine the affinities of this genus, apart from its assignment to the Metopininae. Lea's description is very brief and has a single sketchy figure of the whole fly. Beyer’s detailed, but turgid, redescription, based on two females from Victoria, is longer but devoid of figures. Furthermore he also over- looked the halteres and wing stumps. Seven females of Eutermiphora from a colony of Nasutitermes fumigatus (Brauer 1865) at Braidwood, New South Wales, allowed recognition that the mature females retain halteres and wing stumps (it being proba- ble that the young females shed their wing membranes upon entering the termite colony, as with Termitoxeniinae). In the light of this discovery, an illustrated re- appraisal of the genus was provided (Disney 1989). On a recent trip to Australia I collected a single female Eutermiphora from a colony of Nasutitermes fumigatus at Berowra, north of Sydney. On slide mounting this specimen, it proved to be a sibling species of the series from Braidwood. I therefore borrowed the type of E. abdomina- lis. The specimen from Berowra is conspecific. It is concluded that the series from Braidwood constitutes a hitherto unrecognised sibling species. It is named below, its description augmented and its distinction from E. abdominalis specified. The genus Zutermiphora is characterised by Disney (1989) and included in the latest key to the genera of the Phoridae of the world (Disney 1994). It is currently only known from mature, flightless-stage, females. 88 Rea Jal IE. DM my Figs 1—2: Eutermiphora females, gland openings (g) at rear of dorsal faces of fifth abdominal segments. 1, E. watsoni; 2, E. abdominalis. (Scale bars = 0.1 mm). Eutermiphora abdominalis Lea, 1911 Eutermiphora abdominalis Lea, 1911: 77. Holotype 9, AUSTRALIA, New South Wales, Sydney (South Australien Museum, Adelaide) [examined]. Beyer, 1966: 214. 2 9, AUSTRALIA, Victoria, Echuca, 1. 1. 1936. (National Museum of Victoria, Melbourne) (the specimens appear to have been lost). Lea (1911) provided a single figure of the whole fly. His holotype is now remounted on a slide. The principal difference from the newly-recognised species is in the hairing and microtrichia of the abdomen (see below). The microtrichia are exceptionally small and densely crowded (Fig. 5). The measurements in Beyer’s redescription allow recognition that his specimens belong to this species. Female (Figs 2, 5, 8, 10—11). A sibling species of Eutermiphora abdominalis 89 Figs 3—6: Eutermiphora females, abdominal hairs and microtrichia on dorsal face. 3, E. wat- soni, fifth segment; 4, E. watsoni, third segment; 5, E. abdominalis, fifth segment; 6, E. wat- soni, base of bristle on segment 3. (Scale bars = 0.1 mm for 3—5 and 0.01 mm for 6). The host reported for this species by Lea (1911) was “Eutermes fumipennis”, now called Nasu- _ titermes fumipennis (Walker). However, this species is now known not to occur in Australia and Watson & Gay (1980) suggested that most Australian records of this species were probably really Nasutitermes exitiosus (Hill) or, in a few cases, N. walkeri (Hill). However, there is a single soldier termite mounted with Lea's holotype of this fly. I have now identified this termite as N. fumigatus. New material examined: Q, New South Wales, Berowra, 7. VI. 1995 (R. H. L. Disney), in galleries of Nasutitermes fumigatus in log. 90 REE Disney: Eutermiphora watsoni sp. n. Eutermiphora abdominalis Disney 1989: 435. Misidentification Female only known (Figs 1, 3—4, 6—7, 9, 12) The figures supplement those given elsewhere (figs 1—5 in Disney 1989, figs 6.1d, 8.46a—b in Disney 1994). This species is a little larger than E. abdominalis. The third segment of the arista is at least 0.2 mm long, but is nearer 0.1 mm in E. abdominalis. The apical bristle of the palp is longer than the subapical bristle, but is shorter than the subapical in £. abdominalis. The hairs of the dorsal face of the abdomen are less sharply divided into two distinct series, of long bristle-like hairs and short hairs. There is a more continuous gradation from long to short hairs (cf Figs 1 and 2). A clear difference is that the microtrichia of the upper faces of abdominal segments 2—6 are much larger and fewer in number (Figs 3—4), only being at all dense in the vicinity of the gland opening (g in Fig. 1) at the rear of segment 5. Especially on segment 2, these microtrichia tend to be arranged in irregular transverse rows (as in Fig. 1). The metatarsi of the front and hind legs are more elongated than those of E. abdominalis (Figs 7—10). Etymology: The species is named in memory of the late Dr J. A. Watson, who gave me valua- ble information and publications on Australian termites. Holotype 9, AUSTRALIA, New South Wales, about 12 miles west of Braidwood, from nest of Nasutitermes fumigatus, 11. V. 1960 (F. J. Gay) (Australian National Insect Collection, CSIRO, Canberra). 6 Q paratypes, same data as holotype except three deposited in University Museum of Zoology, Cambridge. Discussion The achievement of a satisfactory supra-generic classification of the Phoridae is still a long way off, and is the subject of current controversy. The subfamilies Termi- toxeniinae and Thaumatoxeninae are well circumscribed, but their affinities with the rest of the family remain obscure (Disney & Kistner 1992, 1995). The Aenigmatiinae are probably a polyphyletic assemblage reflecting convergent evolution. The largest subfamily, the Metopininae, embraces the probably monophyletic Megaseliini, the almost certainly polyphyletic Beckerinini and the probably paraphyletic Metopinini. The remaining subfamily, the Phorinae, is likewise probably paraphyletic. Brown’s (1992) proposals designed to rectify some of these defects have served to highlight the latter. However, the proposals themselves have precipitated current controversies (Disney 1993, 1994, Brown 1995, Disney 1995a, 1996). In view of this I highlight an interesting feature of the genus Eutermiphora. The palps carry the pre-apical differentiated setae on the external face (Fig. 12), rather than on the internal face like most Phoridae. The patterns of distribution of these setae in different families and genera is reviewed below. In the Sciadoceridae, the putative sister group of the Phoridae (Cumming et al. 1995), the differentiated setae on the palps are all weak and restricted to the vicinity of the lower margin, but with most being submarginal on the outer face (Fig. 13). In the putative sister group of the Sciadoceridae+ Phoridae, the Ironomyiidae, the setae are larger and more numerous and appear to occur on both faces (Figs 1—2 in McAlpine 1967). This is certainly the case in the supposedly related Lonchopteridae, in which the setae are well developed on both faces of the palp, but are more numerous and stronger on the outer face (Fig. 14). In the presumed basal family of the Cyclorrhapha, the Platypezidae, many genera have no differentiated setae on the A sibling species of Eutermiphora abdominalis 91 Figs 7—10: Eutermiphora females, anterior faces of tarsi. 7, E. watsoni, front leg; 8, E. abdo- minals, front leg; 9, E. watsoni, hind leg; 10, E. abdominalis, hind leg. (Scale bars = 0.1 mm). palps. Others, e. g. Agathomyia Verrall, have strongly differentiated setae on the lower margin and inner face of the palp. The related Opetiidae, have no differen- tiated setae on the palps. In the Empidoidea differentiated setae on the outer face of the palp are common, often with the apical setae further differentiated as bristles. The inner face is typically devoid of differentiated setae. Taken overall, the above outgroup comparisons suggest that the groundplan state in the Cyclorrhapha is possibly with differentiated setae on the both faces of the palp 92 I Jeb Jb IDI Ney Vee 72 Figs 11—14: 11—12: Eutermiphora females: 11, E. abdominalis, right wing stump; 12, E. watson, right palp. 13—14: Left palps: 13, Sciadocera rufomaculata female; 14, Lonchoptera lutea male. (Scale bars = 0.1 mm). but more strongly on the outer face, with one or more apical or subapical setae further differentiated as stronger setae or bristles. Thus Lonchoptera possibly represents the groundplan state. Within the Phoridae every variety of palp setal pattern is encountered, but most frequently the setae are on the inner face with the apical bristles differentiated in varying degrees. The condition reported for Eutermiphora (Fig. 12), therefore, is atypical. However, this pattern is characteristic of all Termitoxeniinae (Disney & Kistner 1995). Furthermore, it occurs in some Metopininae. In most Megaseliini the inner face carries the setae and the apex has differentiated bristles, but in a few species the setae are reduced or lost. In many Metopinini also the pre-apical setae are on the inner face; for example Chonocephalus Wandolleck and Echidnophora Schmitz. In the case of the latter, this observation reinforces the finding that the resemblance between Termitoxeniinae and Echidnophora is due to convergence (Disney 1995b). However, the termitophilous metopinine genera Bolsiusia Schmitz A sibling species of Eutermiphora abdominalis 93 and Penthaplus Schmitz, like Eutermiphora, have the pre-apical setae on the outer face of the palp, as does the type genus of the Metopinini, Metopina Macquart. It is possible, therefore, that this feature is a synapomorphy that unites these genera with the Termitoxeniinae. However, we do not know the extent to which this feature may have evolved independently more than once, and even symplesiomorphy cannot be ruled out. These observations on palp setal pattern suggest the hypothesis that Eutermiphora might be phylogenetically related to the Termitoxeniinae. The female's abdominal physogastry, the shortened costa in relation to the shedding of the wing membranes (presumably at the time of entry to the termite host's nest), the lack of axillary bristles on the wing, the reduction in the size and/or density of the microtrichia on the top of the abdomen and the leg chaetotaxy are additional features of resemblance. The mid-mesopleural ridge and the gland opening at the rear of abdominal segment 5 (g in Fig. 1) distinguish Futermiphora and allow its assignment to the Metopininae. The tapered labella of the proboscis (fig. 2 in Disney 1989) are very similar to those of many Termitoxeniinae. However, this feature is probably not a groundplan feature of the latter subfamily (Disney & Kistner 1995) and so this resemblance most likely represents an example of convergent evolution. The small, subcircular, features clustered around the bases of some abdominal bristles (Fig. 6) invite further investigation. Fresh specimens examined with the scanning electron microscope should establish whether these features are perfora- tions, thus indicating that the associated bristles are probably primitive solutricho- mes that produce a placatory exudate licked up by the host-termite workers (Disney 1995b, Disney & Kistner 1995). So far solutrichome bristles are only known from Termitoxeniinae. The procurement of the unknown males of Eutermiphora would be likely to reveal whether the majority of the similarities between this metopinine genus and Termi- toxeniinae are due to affinity or convergence. Acknowledgements I am gratefulto Dr Leigh Miller (CSIRO, Berrimah, N. T.) for the identification of the termi- tes. Ithank Dr E. G. Matthews (South Australian Museum, Adelaide) for the loan of the holo- type of E. abdominalis and Dr K. Walker (Museum of Victoria, Abbotsford) for his attempts to locate Beyer’s specimens. I thank my cousin John Disney for his hospitality and for intro- ducing me to the locality where I collected the new specimen that prompted this paper. My work on Phoridae is currently funded by a grant from the Leverhulme Trust (London), made to Dr W. A. Foster (Department of Zoology, University of Cambridge) for a Research Associate to work on termitophilous Phoridae, and a grant from the Isaac Newton Trust Trinity College, Cambridge. Zusammenfassung Eutermiphora watsoni sp. n. wird von E. abdominalis Lea 1911, der bisher einzigen Art und Typusart der Gattung, unterschieden. Die Weibchen beider Arten leben in Kolonien von Nasu- titermes fumigatus (Brauer 1865) in New South Wales. Die Anordnung der präapikalen Borsten auf den Maxillarpalpen wird als Merkmal erórtert, besonders im Hinblick auf die ungeklárte systematische Stellung der Termitoxeniinae. 94 RENE Disney References Beyer, E. M. (1966): Neue und wenig bekannte Phoriden, zumeist aus dem Bishop Museum, Honolulu. — Pacific Insects 8: 165—217. Brown, B. V. (1992): Generic revision of Phoridae of the Nearctic Region and phylogenetic classification of Phoridae, Sciadoceridae and Ironomyiidae (Diptera: Phoridea). — Mem. ent. Soc. Canad. 164: 1—144. Brown, B. V. (1995): Response to Disney. — J. nat. Hist. 29: 259—264. Cumming,J. M., B. J. Sinclair £ D. M. Wood (1995): Homology and phylogenetic impli- cations of male genitalia in Diptera — Eremoneura. — Ent. scand. 26: 120—151. Disney, R. H. L. (1989): Re-appraisal of the Australian genus Eutermiphora Lea (Diptera: Phoridae). — Ent. scand. 19: 435—436. Disney, R. H. L. (1993): Mosaic evolution and outgroup comparisons. — J. nat. Hist. 27: MIO IZ ; Disney, R. H. L. (1994): Scuttle Flies: The Phoridae. — Chapman & Hall, London, xii + 467 pp. Disney, R. H. L. (1995a): Reply to Brown. — J. nat. Hist. 29: 1081—1082. Disney, R. H. L. (1995b): Convergent evolution between Echidnophora and Termitoxeniinae (Diptera: Phoridae). — Syst. Ent. 20: 195—206. Disney, R. H. L. (1996/97): Continuing the debate relating to the phylogenetic reconstruc- tion of the Phoridae (Diptera). — G. it. Ent. 7: 103—117 [1994], 338 [1995]. Disney, R.H. L. & D. H. Kistner (1992): Revision of the termitophilous Thaumatoxeninae (Diptera: Phoridae). — J. nat. Hist. 26: 953—991. Disney, R. H. L. & D. H. Kistner (1995): Revision of the Afrotropical Termitoxeniinae (Diptera: Phoridae). — Sociobiology 26: 115—225. Lea, A. M. (1911): On a new Australian genus of Phoridae, associated with termites. — Proc. RESO Wiens), 243 Gaal ie McAlpine, J. F. (1967): A detailed study of Ironomyiidae (Diptera: Phoroidea). — Can. Ent. 997225236: Watson, J. A. L. & F. J. Gay (1980): The identities of Termes australis Walker and Termes: fumipennis Walker (Isoptera). — J. Aust. ent. Soc. 19: 19—25. Dr R. H. L. Disney, Field Studies Council Research Fellow, University Department of Zoology, Downing Street, Cambridge, England. A new species of Votolaemus Lefkovitch from Rwanda (Coleoptera: Laemophloeidae) Michael Karner Abstract. Notolaemus wagneri n. sp. from Rwanda is diagnosed, described and illus- trated. Brief notes on the genus Nofolaemus are given. Key words. Coleoptera, Laemophloeidae, Notolaemus, description, Rwanda. Introduction In his study of European Laemophloeinae, Lefkovitch (1959) described the genus Notolaemus with two species, Notolaemus castaneus (Erichson) and N. unifasciatus (Latreille), and mentioned that the genus is also represented in Asia, America, and Africa. Later, Lefkovitch (1962) provided a key to the African Laemophloeidae and recorded eight species of Notolaemus. Except for the genus Cryptolestes Ganglbauer (which is, because of the economic importance of some species, comparatively well-studied [Thomas 1988], little work has been done on this family since then. It is to be expected, therefore, that the great majority of Laemophloeidae is still undiscovered. A small but very interesting sample of Laemophloeidae and Silvanidae collected near the Akagera National Park, Rwanda, by Thomas Wagner (Bonn), included a new species of Notolaemus which is described here. Notolaemus wagneri n. sp. Types: Male holotype in Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn (ZFMK), with following data: “Rwanda, Rusumo / Ibanda Makera, X. / Th. Wagner leg. 93”. 9 Paratypes with same data, in ZFMK (1 o, 5 Q) and in the author’s collection (1 o, 2 2). Diagnosis Notolaemus wagneri n. sp. is similar to N. liganus Lefkovitch and N. perspicuus (Grouvelle), it can be differentiated from N. liganus by greater body length (see Discussion), longer antennae, and from both species by its more prominent, ir- regularly rounded eyes and by the structure of the male genitalia. Description Length 1.63—2.00 mm, body flattened, moderately shining, testaceous; habitus as in fig. 1. Head transverse, length (from apical margin of clypeus at its middle to imaginary line between hind margins of eyes) 0.30—0.39 mm, width (across eyes) 0.45—0.65 mm, width : length = 1.45—1.70; dorsal surface of head near antennal insertions and along median line shallowly impressed. Eyes prominent, irregularly rounded. Front margin of head with flat antennal and deeper mandibular emarginations, labral emargination evenly rounded. Clypeus emarginate anteriorly, with distinct microreticulation. Disc of head shining, almost impunc- tate medially, laterally punctate, puncture diameter 1.5—2.0 times that of eye facet and punc- tures separated by on an average their diameter, pubescence very fine, setae about as long as 96 M. Karner Fig. 1: Habitus of Notolaemus wagneri n. sp. (scale line 1 mm). Fig. 2: Male genitalia of Notolaemus wagneri n. sp. (holotype, scale line 0.25 mm). New Notolaemus from Rwanda 97 puncture diameter; surface between punctures somewhat uneven, on lateral parts of head with indistinct microreticulation. Antennae long and slender, length 1.00—1.43 mm, reaching almost three quarters of body length in males, in females a little shorter, about 0.6 times as long as body. Ratios of length of antennomeres I—XI as: 1:0.6:0.7:0.6:0.75:0.8:0.85:0.8:0.85:0.9:1.15 (male holotype), or 1:0.6:0.5:0.55:0.5:0.6:0.6:0.55:0.7:0.7:1 (female paratype). Pronotum transverse, length 0.31—0.43 mm, width 0.46—0.63 mm, width : length = 1.35— 1.56; broadest at apical fourth in males, at apical third or somewhat closer to the midpoint in females; posterior angles obtuse, anterior angles acute, rectangular; lateral margin very slightly crenulate. Sublateral lines consisting of internal groove and basally broadened raised ridge. A longitudinal impression located laterad to sublateral line at about midpoint of pronotum. Puncture size as on head, punctures separated by about 1—2 times their diameter; surface between punctures shining, with sparse microsculpture consisting of very fine and short lines (magnification 80 times). Elytraevenly curved laterally, length (measured along suture, including scutellum) 0.95—1.2 mm, width (across maximum joint width) 0.60—0.75 mm, length : wdth = 1.50—1.63. Hume- rus with small, tooth-like projection, sutural stria present only on apical half of elytra, elytral cells I and II absent, cell III impressed basally, only inner stria extending apically, lateral carina complete. Male genitalia as in fig. 2. Derivatio nominis: Dedicated to Thomas Wagner, collector of the new species. Variation: In a single female paratype, two rounded impressions near the pronotal base on both sides of the inner third are present. No further differences to the other specimens at hand could be observed, so this may be a unique deformation. Discussion The study of the holotype and six paratypes of Notolaemus liganus showed that the body length ranges from 1.35 to 1.55 mm (using a calibrated ocular micrometer), differing considerably from the data according to Lefkovitch (1962). Lefkovitch (1962) described the sublateral lines on the pronotum of Notolaemus as represented by raised ridges. Nevertheless, the sublateral lines of both N. liganus and N. wagneri n. Sp. consist of a raised ridge and an internal groove. The transverse sectional view of the pronotum of N. clarus (Grouvelle) given by Lefkovitch (1962) also shows distinct internal grooves. The genus Notolaemus seems to be composed of some species — groups whose systematic positions are far from clear. Lefkovitch (1962) stated that the African species Notolaemus picinus (Grouvelle) belongs to what he called “unifasciatus group”, whereas the other African species are closer to N. castaneus. However, the comparison of e.g. N. liganus with N. castaneus hardly reveals greater similarity than there is between N. castaneus and N. unifasciatus. Any division of the genus Notolae- mus into subgenera or even species groups is obviously premature until much more work has been done especially on tropical species. Acknowledgements The author would like to thank Dr. N. Berti (Museum National d’Histoire Naturelle, Paris) and Dr. E. De Coninck (Musee Royal de Afrique Centrale, Tervuren) for the loan of types, Dr. Th. Wagner (Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn) for the opportunity to study his material, and Christine Ott (Frankfurt am Main), Dr. D. Burck- hardt (Geneva), Bernd Franzen (Cologne), and Dr. M. C. Thomas (Gainesville) for reading and criticizing the manuscript. 98 M. Karner Zusammenfassung Notolaemus wagneri, eine neue afrikanische Laemophloeide, wird diagnostiziert, beschrieben und abgebildet. Die Zusammensetzung der Gattung Notolaemus wird kurz diskutiert. Literature Lefkovitch, L. P. (1959): A revision of European Laemophloeidae (Coleoptera: Cucujidae). — Trans. R. Ent. Soc. London 111: 95—118. Lefkovitch, L. P. (1962): A revision of African Laemophloeinae (Coleoptera: Cucujidae). — Bull. British Mus. Nat. Hist. (Ent.) 12: 167—245. Thomas, M. C. (1988): A revision of the New World Species of Cryptolestes Ganglbauer (Coleoptera: Cucujidae: Laemophloeinae). — Insecta Mundi 2: 43—65. Michael Karner, c/o Dr. D. Kovac, Sektion Entomologie I, Senckenbergisches Natur- forschendes Institut, Senckenberganlage 25, D-60325 Frankfurt. Bd. 47 S. 99—110 Bonn, September 1997 Collembola Poduromorpha from the Magallanes Province (Chile) Wanda M. Weiner & Judith Najt Abstract. Inthis paper we present a study of a collection of Collembola Poduromorpha from the south of Chile. We determined a total of 15 species. Two new species were described; another species already known from Chile was redescribed, and some comple- mentary characters were added for three other species. Key words. Collembola, Poduromorpha, new species, Magallanes Province, Chile. Introduction The Chilean Province Magallanes as well as the Argentinean Tierra del Fuego are very interesting for a scientist because they belong to the Araucarian-Antarctic bio- geographic region (Rapoport 1968, 1971). Collembola were studied there by Schäffer already in 1897. In continuation, other authors published some works on the Collem- bola of that region. Among the most important papers concerning Collembola Podu- romorpha are Wahlgren (1906), Enderlein (1912), Rapoport & Rubio (1968), Izarra (1971), Najt (1973), Rubio (1974). In this paper we present a study of a collection of Collembola collected in Barber’s traps by Dr. M. Vogel (Akademie für Naturschutz und Landschaftspflege, Laufen, Germany) during his field studies in this region. The type material studied in this paper is deposited in the Institute of Systematics and Evolution of Animals, the Polish Academy of Sciences, Kraköw (ISEA) and in the Laboratoire d’Entomologie, Muséum national d’Histoire naturelle, Paris (MNHN). Systematic account Hypogastruridae Hypogastrura purpurescens (Lubbock, 1867) Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphag- num fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 20 specimens on slides and numerous specimens in alcohol; wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum pennamarina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2), 19 specimens on slides and numerous specimens in alcohol; mixed forest of Nothofagus antarctica and Notho- fagus pumilio, about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3), 19 specimens on slides and numerous specimens in alcohol; patch of an old Nothophagus pumilio forest (about 500 years old), large proportion of dead wood, understorey with Berberis ilicifolia, Rubus cheoides, Acaena magellanica and Hierochloe redolens (L.PAR-4), 12 specimens on slides and nume- rous specimens in alcohol. Chabunco, about 15 km east of Puenta Arenas, near main road to Puerto Natales, Patagonian steppe, Barber’s traps, 6—19 February 1985, leg. M. Vogel: 100 W. Weiner & J. Naijt VEO ‘ ‘ 4 ' / O A. N ">. : . . A E 0 N} >” e le CC 7 CY (er oO . a Y ‘ ern cc a r - y SS Cu O) ( e * I s : 5 a \ y 2 2 AO = 0 .. » 0% \ ae) o... . y > eo Ost rans EAS = pe = ’ z ? nn 0° Sa ein, SE DO ÓN, N ID 7 10 OO I. 5 ° Ne O | Gee SoG .".to oe . ; ES DON oy ) e ie =~ So Aime Be j OA PO NADA a, o o . OS BR OLSEN OR ed 5 e 5 o NETA IAS . (2 5 Een — o e OA KOHLER : : ci Oyun e Zoro “ye > a 7 2. Y of Teas nn. hr . Es a) a “ 5 10 te , AO CI 2) ) o O55 20 - O - N ~ - -- 2 o A OS ; - 2 e oe , , y , - x BE o SE = E y ya -- 5 Cr A . ES Zo = - N) on On PRO ASE RGR ony ) a, RB ON [8 So Sr = rim ? > o o o AO o o SS np Be Oe Tes ~ 5 © Sees > o 569 } o A o . A oO 9 Wo . +0 . GeO Ae Ses PLS orn a Senco INF . . 00 Ne BEN ( a = OO “Ea o. eo SD, or : R IE ES x A = . r . . . = -- =. . E Do rs RN By pares = y) . y - SE tty ’ E o og y ON 5 REN ; ie O 0 OF ta “.) > = ns” A: S Sr - ) NN e y SD ES D oc A y ASADO AR OS) IS E y ARE Beg Sc N > ahs o - = . ° > . e) € “> We Ue, 0 o o 4 ¢ en Sid lea =e oe ac Y y \ ot]. SLZLE y . O veer OOO - \ x > EN Se . SD . A 1 Y \ Soo 2 DE Y DN ee en ee 5 c* 9.0 o “. 6 ) EN) OSOS Cis ee a = = an Aa u e = - = == m ) SO 0D o Lee STR > > or > nenn Oe og 0. RO : A ERA = ) MORA ... CO . > y O o or ia . o o y 2 Dis : e See). ( ihe » LER gO O SO) ö » ESTO + 9.00 1 $ Re 7 ips. @ o seis eis Son z SE o \ Voce 9.09 ap © ) Sa er E Se ES ‘ TS ) POLO o y Pa So Nica pastos. mess. CNO ds BO DES 30m Ce > [4 o 9,0 7 CA yo] ~ o e AS. ? SEEN \ er > e. E) 20 SS y Sano Y Si y N 3 , © . — pS Re mm a o 2 20 e OOOO a ) « -- hey ER os (ene N Ehen an San A eee A E O Ss, 0% 4 aa Diam A OTe) ep Cig ie) ig Foun 4 SIT af . . o 9 . . 000 \ ) RO SOLO O o Ro ER 5 DO , 0 > CORR Oke PES Sa : UE gO A Sool eo! IS 707 : 00 a are Jo j 2 ; ee N G Sess A ESA POS Or ( . ) ‘ SEMIS e . a) 7 . ( PR 50 ) ( : 5 02 00 N aoa : ri x Oo OSOS A : - x Cl 0g, 00 20 0000 90 2 RO JS OS So 2% © E a 2d 0 TS S . a { 2 5.0 b . ) x y) E 9 0 AO Ge 25? x t . . » E HER de i ( ”— YA = -- C9567 ( SKE Seno ne ) £ \ ae Ort 4b. 1D 49 a Ly cali ie SOO nO Pip te ‘ . oe o . AY) O 0 . e.) ‘ “o © . yes \ o | / EBD we O 1 q ’ u nO" e SS da Figs 1—2: Schema of dorsal chaetotaxy. Fig. 1 — Triacanthella najtae. Fig. 2 — Triacanthella vogeli sp. n. (Scales in mm). mesic grassland, vegetation consisting of Festuca gracilima, Agropyron spp., Agrostis spp., Deschampsia spp. and Poa spp. (CHA-1), 2 specimens; shrub steppe, peculiarly developed upon former dunes near the coast, vegetation with Berberis buxifolia, Lepidophyllum cupres- siforme, Baccharis magellanica and new growth of Nothofagus antarctica (CHA-2), 3 speci- mens. Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: forest of Nothofagus pumilio, about 20 m a.s.1., understorey sparse, with Acaena ovalifolia and Viola magellanica (NAV-1), 4 specimens on slides and some specimens in alcohol; peat-bog of Sphagnum magellanicum, about 30 m a.s.l., with Marsippospermum grandiflorum, Rostkovia magellanica, Carex magellanica and on the edges dwarfed Nothofagus betuloides (evergreen) (NAV-2), 2 specimens; mixed forest with Nothofagus pumilio (deciduous) and Nothofagus betuloides (evergreen), about 50 m Collembola from the Magallanes Province 101 a.s.1., no understorey, only seedlings (new growth) of Nothofagus pumilio, on the forest edge single Baccharis magellanicus (NAV-3), 18 specimens on slides and some specimens in alcohol. Hypogastrura assimilis (Krausbauer, 1898) Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.1., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphag- num fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 1 specimen. Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: mixed forest with Nothofagus pumilio (deciduous) and Nothofagus betuloides (evergreen), about 50 m a.s.1., no understorey, only seedlings (new growth) of Nothofagus pumilio, on the forest edge single Baccharis magellanicus (NAV-3), 1 specimen. Ceratophysella bengtssoni (Agren, 1904) Material: Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: mixed forest with Nothofagus pumilio (deciduous) and Nothofagus betuloides (evergreen), about 50 m a.s.1., no understorey, only seedlings (new growth) of Nothofagus pumilio, on the forest edge single Baccharis magellani- cus (NAV-3), 3 specimens. Comments: This species, described from Sweden, has very large Holarctic distribution. The data presented here are the first account from the Neotropical region. All morphological characters are similar to those of European specimens, including the variability described by : Thibaud (1967) for French specimens, and also by Christiansen & Bellinger (1980) for North- American specimens. Xenylla subcavernarum Gama, 1969 Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arena, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphag- num fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 1 specimen; wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum penna-marina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2); 1 specimen, mixed forest of Nothofagus antarctica and Nothofa- gus pumilio, about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3), 5 specimens on slides and some specimens in alcohol; patch of an old Nothofagus pumilio forest (about 500 years old), large proportion of dead wood, understorey with Berberis ilicifolia, Rubus cheoides, Acaena magel- lanica and Hierochloe redolens (L.PAR-4), 6 specimens on slides and some specimens in alcohol. Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: forest of Nothofagus pumilio, about 20 m a.s.l.,, understorey sparse, with Acaena ovalifolia and Viola magellanica (NAV-1), 2 specimens on slides and some specimens in alcohol; peat-bog of Sphagnum magellanicum, about 30 m a.s.1., with Marsippospermum grandiflorum, Rostkovia magellanica, Carex magellanica and on the edges dwarfed Nothofagus betuloides (evergreen) (NAV-2), 2 specimens on slides and some specimens in alcohol; mixed forest with Nothofagus pumilio (deciduous) and Nothofagus betuloides (evergreen), about 50 m a.s.l., no understorey, only seedlings (new growth) of Nothofagus pumilio, on the forest edge single Baccharis magellanicus (NAV-3), 4 specimens on slides and some specimens in alcohol; evergreen forest of Nothofagus betuloides, 80 m a.s.l., large fraction of dead wood, no understorey, Blechnum magellanicum in small, wet depressions (NAV-4), 2 specimens on slides and one specimen in alcohol. 102 W. Weiner & J. Naijt Triacanthella najtae Izarra, 1971 Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum penna-marina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2), 4 specimens on slides and one specimen in alcohol. Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: mixed forest with Nothofagus pumilio (deciduous) and Nothofagus betuloides (evergreen), about 50 m a.s.l., no understorey, only seedlings (new growth) of Nothofagus pumilio, on the forest edge single Baccharis magellani- cus (NAV-3), 2 specimens on slides and some specimens in alcohol. Comments: For comparison with 7. vogeli sp. n. we enclose the pattern of the dorsal chaetotaxy (Fig. 1). The dorsal chaetotaxy of T. najtae is composed of smooth mesochaetae and crenaleted short macrochaetae. The formula of macrochaetal chaetotaxy per half tergite and subcoxa or pleurite is the following: 8 / 2+1, 4+2, 4+2, / 2+1, 2+1, 2+1, 3+0, 4. Triacanthella vogeli sp. n. Holotype: female (ISEA), data: Chile, Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.1., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: mixed forest of Nothofagus antarctica and Nothofagus pumilio, about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3). Paratypes: 8 specimens on slides (5 in ISEA, 3 in MNHN), 820 specimens in alcohol (720 in ISEA, 100 in MNHN), the same data as the holotype. Other material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barbar’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphagnum fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 6 specimens on slides and numerous specimens in alcohol; wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum pennamarina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2), 3 specimens on slides and some specimens im alcohol; patch of an old Nothofagus pumilio forest (about 500 years old), large proportion of dead wood, understorey with Berberis ilicifolla, Rubus cheoides, Acaena magellanica and Hierochloe redolens (L.PAR-4), 4 speci- mens on slides and some specimens in alcohol. Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: ever- green forest of Nothofagus betuloides, 80 m a.s.1., large fraction of dead wood, no under- storey, Blechnum magellanicum in small, wet depressions (NAV-4), 2 specimens on slides and some specimens in alcohol. Description: Holotype: female length 2.2 mm, length of paratypes between 1.3 and 2.4 mm. Colour in alcohol: white. Tegumental grain of large size, abdominal segment VI with cuticular ornamentation constituted by rosette-shaped tubercles formed by secondary granules (Fig. 7). Antennal segment I with 10—13 setae, II with 14—17 setae. Sensory organ of antennal segment III consisting of two small hammer-shaped sensillae, two guard sensillae, and one ventro-lateral microsensilla (Figs 3—4). Antennal segment IV with trilobated apical vesicle, small subapical “organite”, one dorso-external microsensilla and 16 subcylindrical sensillae (Fig. 3). Ventral side of antennal segment IV with reduced sensory rasp consisting of about 18 modified setae and one subcylindrical sensilla in ventro-lateral position (Fig. 4). 8 + 8 ocelli, of which two (G and H) are reduced. Postantennal organ with 4 unequal vesicles covered by two tegumentary folds. Buccal cone, maxillae and mandibles typical for the genus. Tibiotarsi I, II, III with 19—20, 19—20, 18 acuminated setae. Claw with two basal inner teeth, one small, basal external tooth and 1 + 1 latero-distal teeth; basis of external lamella with a tridentate structure of pseudonychia-type. Presence of rudimentary empodial append- age (Fig. 6). Collembola from the Magallanes Province 103 Figs 3—7: Triacanthella vogeli sp. n. Fig. 3 — antennal segment III and IV, dorsal side. Fig. 4 — antennal segment III and IV, ventral side. Fig. 6 — leg III and dorsal side of claw. Fig. 7 — abdominal segment VI, dorsal side. Ventral tube with 9 + 9 setae. Tenaculum with 3 + 3 teeth. Dens without apical lobe, adult specimens with 15—17 setae, preadult specimens with 13—16 setae, juvenile specimens with 10—11 setae; all specimens have one long, smooth seta at the base, all other setae are ciliated (Fig. 5). Mucro as in Fig. 5. Dorsal chaetotaxy with mesochaetae of two types: smooth and ciliated, with very long ciliated macrochaetae and sensory setae as in Fig. 2. Macrochaetal chaetotaxy per half tergite and subcoxa or pleurite: 8 / 2+1, 3+1, 3+1, / 2+1, 2+1, 2+1, 3+0, 3+0, 3. Abdomen VI with three anal spines of which posterior one is very short (Fig. 7). Discussion: Triacanthella vogeli sp. n. is very similar to two species from Argentina: T. andina Cassagnau & Rapoport, 1962 from Neuquén (Lago Frias and Nahuel Huapi) and T. najtae Izarra, 1971 from Tierra del Fuego and Isla de Los Estados. These three species share the following characters: presence of rosette-shaped tubercles on abdomen VI, shape of claw and rudimentary empodial appendage, reduction of two ocelli (G and H). The new species differs from the two others by the absence of apical lobe on the dens and very short posterior anal spine. Further differences concern the type of macrochaetae (in 7. andina they are only ciliated in the distal half; in 7. najtae they are only crenelated and short, in 7. vogeli sp. n. they are ciliated at the whole length) and formula of macrochaetal chaetotaxy (in T. andina per half tergite and subcoxa or pleurite = 7 / 2,3,3, / 4,4,4,4,3,4 — after Cassagnau & Rapoport, 1962, in 7. najtae = 8 / 2+1, 4+2, 4+2, / 2+1, 2+1, 2+1, 3+0, 3+0, 4 and in / yogeli sp. n. = 8 / 2+1, 3+1, 3+1 / 2+1, 2+1, 2+1, 3+0, 3+0, 3). Derivatio nominis: The new species is dedicated to our colleague, Dr. M. Vogel. Neanuridae Frieseinae Friesea sp. Material: Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: peat-bog of Sphagnum magellanicum, about 104 W. Weiner & J. Naijt Figs 8—12: Brachystomella ronderosi. Fig. 8 — antennal segment III and IV, dorsal side. Fig. 9 — antennal segment III and IV, ventral side. Fig. 10 — furca. Fig. 11 — dorsal side of claw. Fig. 12 — leg III. 30 ma.s.1., with Marsippospermum grandiflorum, Rostkovia magellanica, Carex magellanica : and on the edges dwarfed Nothofagus betuloides (evergreen) (NAV-2), 3 specimens. Comments: It is a new species but we had only juvenile specimens at our disposal and therefore could not describe it in this paper. This species belongs to the group of Friesea species with 8 + 8 ocelli, furca absent (stage 5), 4 spiniformes setae of abdominal segment VI and some capitated setae on the body. Brachystomellinae Brachystomella ronderosi Najt, 1973 Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphag- num fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 5 specimens on slides and numerous specimens in alcohol; wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum pennamarina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2), 8 specimens on slides and numerous specimens im alcohol. Comments: In 1973 Najt described this species from Isla de Los Estados. We enclose here some important figures: chaetotaxy of body (Fig. 13), dorsal and ventral side of antennal segments III and IV (Figs 8—9), leg III (Figs 11—12) and furca (Fig. 10). Setanodosa fueguensis Najt, 1973 Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum pennamarina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Collembola from the Magallanes Province 105 13 | kl Ss a 15 yea MN EY o Denese SUG Oe” ON aN ; Ne 2 Parr \ | 14 ie N No if dy Cop | Ka \ 2, y N | i) 22 El A 27% > Ss. ES 1° ee PP AA ty, Ml an 4 e De SE | ;? PL ‘ ; a A/A : er A FE ge ine? ! a | i AR 21 DIE BE ? | ? 2 Ly / al 9-1 £ en N A Pe | : A gave Sw | pagans un a ae: | a Ad fp ? ? 7 78 16 A id yee 7 A fr PR! pale eae co A, NM i | ae LG mn RR NS ly Ve en ar ER, 4a ol 2, SN N % Be Br ler | SS nr 2 Ben. al ay; > 7p \ a 5 pr j ? 7 al EN rs, a EN i ” Em 7 $7 a e > en BEN Za zZ, 7 | Fig. 13: Brachystomella ronderosi, dorsal chaetotaxy. Figs 14—17: Setanodosa fuenguensis. Fig. 14 — dorsal chaetotaxy. Fig. 15 — antennal segment III and IV, dorsal side. Fig. 16 — antennal segment III and IV, ventral side. Fig. 17 — leg. III. Trisetum cernum, Arenatherum sp. (L.PAR-2), 12 specimens on slides and numerous speci- mens im alcohol. Comments: This species, described from Bahia Suceso on Tierra del Fuego, seems to be numerous in this region. We compared our specimens from Laguna Parillar with the holo- type. In the present paper some details are given: chaetotaxy (Fig. 14), dorsal and ventral side of antennal segments III and IV (Figs 15—16) and leg III (Fig. 17). Brachystomellides navarinensis sp. n. ‘Holotype: male (ISEA), data: Chile, Navarino Island, Puerto Williams, 55°10’ South, 6930” West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: evergreen forest of Nothofagus betuloides, 80 m a.s.1., large fraction of dead wood, no understorey, Blechnum magellanicum in small, wet depressions (NAV-4). Paratypes: 8 specimens on slides (6 in ISEA, 2 in MNHN), 48 in alcohol (38 in ISEA, 10 in MNHN), the same data as the holotype. Other material: Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: forest of Nothofagus pumilio, about 20 m 106 W. Weiner & J. Naijt 18 20 | aa sh ¡GO | y {7543 Pl € 7 Ny (2 Ua lan 04 N oe A a El (bd ¢ Ge Ye geese fe PR aX Eo RS Igor ¢ SOT AE 9.01 AA et ? ER E ? t \ N 23 & | Cs ee ony 24 | ? cf | | B y — le a j = E le = a oe tre OF O MAR un Figs 18—24: Brachystomellides navarinensis sp. n. Fig. 18 — dorsal chaetotaxy. Fig. 19 — antennal segment III and IV, dorsal side. Fig. 20 — antennal segment III and IV, ventral side. Fig. 21 — maxilla. Fig. 22 — leg III. Fig. 23 — postantennal organ and ocelli. Fig. 24 — furca. a.s.l., understorey sparse, with Acaena ovalifolia and Viola magellanica (NAV-1), 4 speci- mens on slides and some specimens in alcohol. Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.1., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: low peat-bog of Sphagnum fimbratum with Gunnera magellanica, Carex atropicta, Carex capitata, Carex curta and Agrostis flavicola (L.PAR-1), 6 specimens on slides and some specimens in alcohol; mixed forest of Nothofagus antarctica and Nothofagus pumilio; about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3), 5 specimens on slides and some specimens in alcohol. Description: Holotype: male length 0.85 mm, length of paratypes: males and females between 0.79 and 1.0 mm, juvenile specimens 0.5 and 0.58 mm. Colour in alcohol: grey-blue. Tegumental grain middle sized. Antennal segment I with 7 setae, II with 12 setae. Sensory organ of antennal segment III consisting of two small hammer-shaped sensillae, two guard sensillae, of which dorsal one is Collembola from the Magallanes Province 107 longer than ventral one, and one ventro-lateral microsensilla (Figs 19—20). Antennal segment IV with considerable trilobated apical vesicle, subapical “organite”, one dorso-lateral micro- sensilla, only 6 fine sensillae, of which 3 are in dorso-external position and 3 dorso-internal position, and normal setae blunt at the top (Fig. 19). Ventral side of antennal segment IV with sensory rasp consisting of about 30—32 modified sensory setae and some normal blunt setae (Fig. 20). 8 + 8 ocelli. Postantennal organ with 8 fine granulated vesicles (Fig. 23). Buccal cone short. Globular maxillary head with apical tooth and basally with two hyaline lobes (Fig. 21). Mandible absent. Femur ventrally with one very long seta and some short ones. Tibiotarsi I, II, III with 19, 19, 18 acuminated setae. Claw with one inner tooth and a pair of large lateral teeth at the base (Fig. 22). Ventral tube with 3 + 3 setae. Tenaculum with 3 + 3 teeth. Dens dorsally with 6 setae, of which 3 are thick and ciliated (Fig. 24). Mucro as in Fig. 24. Dorsal chaetotaxy as in Fig. 18. Formula of sensory setae per half tergite = 0/022/21111. Discussion: The new species differs from three other species of this genus by the pre- sence of 8 vesicles in postantennal organ (in B. compositus Arlé, 1959 = 4, in B. micropilosus Cassagnau & Rapoport, 1962 = 17—22 and in B. neuquensis Cassagnau & Rapoport, 1962 = 13—17 vesicles). The presence of three thick, ciliated setae on the dens and the maxilla with one apical tooth and two hyaline lobes is characteristic only for the new species. Pseudachorutinae Delamarellina guilleni Rapoport & Rubio, 1963 Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barbar’s traps, 23 January — 6 February 1985, leg. M. Vogel: mixed forest of Nothofa- gus antarctica and Nothofagus pumilio, about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3), 2 specimens: patch of an old Nothophagus pumilio forest (about 500 years old), large propor- tion of dead wood, understorey with Berberis ilicifolia, Rubus cheoides, Acaena magellanica and Hierochloe redolens (L.PAR-4), 1 specimen. Redescription: Female: length 4.5 mm, length of young specimens: 1.6—1.8 mm. Colour in alcohol: dark grey-blue. Body flattened dorso-ventrally with projecting paratergits. Abdomen VI in ventral position. Tegumental grain medium-sized. Plurichaetotic chaetotaxy strongly developed. Antennae III and IV fused dorsally, ventrally distinctly separated. Ratio of antenna IV: antenna III = 3.3 (ventral side). Sensory organ of antennal segment III consisting of two small sensillae bent in the same direction, two long guard sensillae fine and subcylindric, of which dorsal one is 1.2 times longer than ventral one, and one ventral microsensilla. Dorsal guard sensilla and two small sensillae of sensory organ of antennal segment III at the base of apical half of antennal segment HI + IV (Fig. 28). Ventral guard sensilla and ventral microsensilla on ventral part of distinct antennal segment III (Fig. 29). Dorsal side of antennal segment IV with trilobated apical vesicle, distinct bilobated subapical “organite”, small dorso-external microsensilla and 17 short, thick subcylindrical sensillae, all in apical position (Fig. 28). Ventral side with large sensory rasp as in Fig. 29. Ocelli 5 +5. Postantennal organ absent. Buccal cone short. Labium with seta L and 4 labial “organite” (x) as in Fig. 27. Mandible and maxillae as in Figs 30—31. Tibiotarsi I, II, III with 19, 19, 18 acuminated setae. Claw with one inner basal tooth (Fig. 32). Ventral tube with about 30 + 30 setae. Tenaculum with 2 + 2 teeth, without seta on the corpus. Dens reduced, globular, with 11—12 setae, mucro absent (Fig. 33). Dorsal chaetotaxy of head, thorax I, II, abdomen IV and V as in Figs 25—26. Body setae short, only sensory setae long. Comments: This species was described by Rapoport & Rubio (1963) from El Roble Mountains in Chile. In this paper a new redescription is given, including some additional W. Weiner & J. Naijt 108 he | we wus \ | ’ Xt az en N a =. DS | Lo ES A Do! te WSS IN | \ Be ne ES EZ, y E pay, ES SS a sn: ae ELLA LD N N IHR Val LP ; <5. : NA | en a A =—T == S- 30% E 2 AU: a la : O \ se \ eur (5) Den = | NEST cS | NEST, TADA ay) 0, ws ; wet e SEG aN TF OS Lo — = EA 2 hu y ” < 7 ( 2 (CS € I RD AST ERE Be Mae Z i == AA E NES E ES > PA Z ba f era ARA LEER > ‘ == z = RE N RD ES ( PR re un ¢ ZG (preadult specimen). Fig. 26 — dorsal chaetotaxy of abdomen IV and V (preadult specimen). Fig. 27 — labium, right side. Fig. 28 — antennal segment III and IV, dorsal side. Fig. 29 — The adult specimen (female) is bigger than specimens from El Roble, but we antennal segment III and IV, ventral side. Fig. 30 — mandible. Fig. 31 — maxillae. Fig. 32 — leg III. Fig. 33 — furca. consider it asthe same species. In 1962 Cassagnau & Rapoport described Arlesia globulosa as found in Patagonia in Argentina, which is now placed in the same genus Delamarellina Rapoport & Rubio, 1963. D guilleni differs from D. globulosa (Cassagnau & Rapoport, 1962) particularly by the shape of maxillary head, absence of mucro and tenaculum with only 2 + 2 Figs 25—33: Delamarellina guilleni. Fig. 25 — dorsal chaetotaxy of head, thorax I and II teeth (3 + 3 in D. globulosa). characters. Collembola from the Magallanes Province 109 Onychiuridae Tullbergiinae Tullbergia bisetosa Bórner, 1903 Material: Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: forest of Nothofagus pumilio, about 20 m a.s.l., understorey sparse, with Acaena ovalifolia and Viola magellanica (NAV-1), 2 specimens; ever- green forest of Nothofagus betuloides, 80 m a.s.1., large fraction of dead wood, no under- storey, Blechnum magellanicum in small, wet depressions (NAV-4), 1 specimen. Tullbergia trisetosa (Schäffer, 1897) Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: wet bottom of a valley, sloping towards a stream with the most important plants as Gunnera magellanica, Blechnum pennamarina, Acaena magellanica, Carex curta, Carex atropicta, Poa pratensis, Festuca rubra, Trisetum cernum, Arenatherum sp. (L.PAR-2), 2 specimens; patch of an old Nothofagus pumilio forest (about 500 years old), large proportion of dead wood, understorey with Berberis ilicifolia, Rubus cheoides, Acaena magellanica and Hierochloe redolens (L.PAR-4), 1 specimen. Mesaphorura yosii (Rusek, 1967) Material: Navarino Island, Puerto Williams, 55°10’ South, 69°30’ West, Barber’s traps, 26 January — 2 February 1985, leg. M. Vogel: evergreen forest of Nothofagus betuloides, 80 m a.s.l., large fraction of dead wood, no understorey, Blechnum magellanicum in small, wet depressions (NAV-4), 1 specimen. Comments: Cosmopolitan species. Dinaphorura sp. Material: Brunswick Peninsula, Laguna Parillar ca 70 km west of Punta Arenas, 250 m a.s.l., Barber’s traps, 23 January — 6 February 1985, leg. M. Vogel: mixed forest of Notho- fagus antarctica and Nothofagus pumilio, about 80 years old, in the understorey: Gunnera magellanica, Empetrum rubrum, Chilotrichum diffusum, Senecio acantifolius (L.PAR-3), 1 specimen. Comments: It is a new species but we had only one specimen (male) at our disposal and therefore could not describe it in this paper. This species belongs to the group of Dinaphorura species with the pseudocellar formula per half tergite: 11/011/11111 and two anal spines and 5 spiniformes projections. It differs from the other species of this group by its chaetotaxy. Acknowledgements We express our sincere gratitude to Dr. M. Vogel for his kind cooperation and Dr. E. Tabakowska for reading the English version. The work was partly supported by the grant KBN 1828/4/91 from the Polish Committee for Scientific Research given to W. M. Weiner. Resume Dans ce travail nous étudions une collection de Collemboles Poduromorphes du sud de Chili. Nous avons determine 15 especes. Nous décrivons deux espéces nouvelles. Nous redécrivons une autre espece déja connue du Chili et nous ajoutons des caracteres complémentaires pour trois autres especes. 110 W. Weiner & J. Naijt Zusammenfassung Bericht über eine Aufsammlung von Collembolen aus dem Süden von Chile. 15 Arten wurden identifiziert. Zwei Arten werden als neu und eine wenig bekannte Art genauer beschrieben; zusätzliche Merkmale werden für drei weitere Arten benannt. References Cassagnau, P. & E.H. Rapoport (1962): Collemboles d’Amerique du Sud. I. Poduromor- phes. — Biologie de l’Amérique Australe, CNRS, 1: 139— 183. Christiansen, K. & P. E Bellinger (1980): The Collembola of North America north of the Rio Grande. — Grinnell College, Iowa 1322 pp. Enderlein,G. (1912): Die Insekten des Antarkto-Archiplata-Gebietes (Feuerland, Falkland- Inseln, Süd-Georgien). — Kungliga Svenska Vetenskapsakademiesm Handlingar 48 (3): 12170: Izarra, D. C. (1971): Sobre el género Triacanthella Schaeffer con descripción de una nueva especie: 7. najtae (Insecta Collembola). — Physis 30 (81): 345—350. Najt, J. (1973): Sobre algunos Arthropleona de la Isla de Los Estados. I. (Insecta, Collem- bola). — Physis 32 (85): 241—245. Rapoport,E. H. (1968): Algunos problemas biogeográficos del nuevo mundo con especial referencia a la región neotropical. — Biologie de l’Amérique Australe, CNRS 4: 53—110. Rapoport, E. H. (1971): The geographical distribution of Neotropical and Antarctic Collembola. — Pacifics Insects Monograph 25: 99—118. Rapoport, E. H. & I. Rubio (1963): Fauna Colembológica de Chile. — Investigaciones zoolögicas chilenas 9: 95—124. Rapoport,E. H. & I. Rubio (1968): Fauna Colembológica de Chile. II. — Acta Biológica venezueliense 6 (1): 52—67. Rubio, I. (1974): Fauna Colembolögica de la península Muñoz Gamero (Provincia de Magal- lanes, Chile). — Physis 33 (86): 125—131. Schäffer, C. (1897): Apterygoten. — Hamburger Magalhaensische Sammelreise: 1—48. Thibaud, J-M. (1967): Contribution a l'étude du développement postembryonnaire chez les Collemboles Hypogastruridae épigés et cavernicoles. — Annales de Spéléologie 22 (1): 107—198. Wahlgren, E. (1906): Antarktische und Subantarktische Collembolen gesammelt von der schwedischen Südpolexpedition. — Wissenschaftliche Ergebnisse der schwedischen Süd- pol-Expedition 1901—1903 5 (9): 1—22. Dr. Wanda M. Weiner, Institute of Systematics and Evolution of Animals, Polish Academy of Sciences, Slawkowska 17, PL-31016 Kraków, Poland. Dr. Judith Najt, EP 90 du CNRS et Laboratoire d’Entomologie, Museum national d’Histoire naturelle, 45, rue Buffon, F-75005 Paris, France. Bd. 4 S. 11134 | Bonn, September 1997 On the evolution of anthophilous Nitidulidae (Coleoptera) in tropical and subtropical regions A. G. Kirejtshuk Abstract. Independent appearance of anthophagy among different nitidulid groups is considered. Circumstances and regular ways of this process are traced and explained. Simi- lar correlations in transformations of structures, trophics and mode of life among antho- phagous forms from not related groups are shown. Propetes (Propetes) aquilus sp. n., P (P) seychellensis sp. n., P (Mandipetes) intritus subgen. et sp. n., P (M.) longipes sub- gen. et sp. n., Brounthina aequalis gen. et sp. n., Caplothorax subgen. n. and Plapennipo- lus subgen. n. in the genus Carpophilus, Urocarpolus subgen. n. in the genus Nitops stat. n. are described and proposed. Key words. Anthophagy, specialisation, Nitidulidae, phylogeny, taxonomy. Introduction This paper is based on a report read at the International Symposium on Biodiversity and Systematics in Tropical Ecosystems held on the 2—6 May 1994. Because of a necessity to add some taxonomical comments which had increased the text more than twice, the final version of the paper was submitted for publication separately from the proceedings of the symposium. The anthophagy of beetles has drawn the attention of entomologists of many generations. There are a lot of publications on the importance of the anthophagy Of beetles in the evolution of the higher plants, though there are few references where a possibility of reciprocal influence in coevolutionary interconnections between coleopterous groups and plants is regarded. The influence of plant food on the insects does not look so markedly in the results of phylogeny of the order Coleoptera in contrast with that of the angiosperm plants and therefore it is not considered to deserve any special investigation. The anthophagy of coleopterous adults and their larvae is usually studied without any link in the formation of these types of feeding, and a certain regularity in trophic transformation is as yet missed from considera- tion. The Nitidulidae are a comparatively young family including many groups where we can find a considerable resemblance in transformations of structures, trophics and mode of life showing some more or less evident correlations. Such correlations are especially significant among some Nitidulid groups each of which independently became anthophagous (Kirejtshuk 1994a). Moreover, among the anthophilous Nitidulidae distributed in tropical and subtropical regions the regularities here under consideration are more obvious than in subpolar territories. The Mesozoic Cucujoidea (including Nitidulidae) are only known from the Creta- ceous. According to Dmitriev & Zherikhin (1988) the evident Cleroidea, Tenebrionoi- dea and Chrysomeloidea as well as extinct Parandrexidae with unclear position 112 A. G. Kirejtshuk appeared at least no later than the Jurassic. The true Nitidulidae together with other Cucujoid families have been found in different layers of the Cretaceous increasing in number to the end of this period (Ponomarenko 1983; Kirejtshuk & Ponomarenko 1990). If such a distribution in a certain sense reflects dynamics of the appearance of different coleopterous groups, in contrast to a widespread opinion, we are forced to assume that the Cucujoidea had risen no earlier but rather much later than Tenebrionoidea and Chrysomeloidea became distinctly recognisable groups to be identified in fossils. Diversification of the Cucujoidea seemed to arise and be proceeding when the characteristic mesozoic groups of plants became more and yet more rare until they were dislodged by the newly appeared angiosperm plants. Thus the evolution of the Cucujoidea is, perhaps, associated with the development of the Kainophytic flora, even in case this coleopterous group could take its origin some- what earlier. Interconnections between the Cucujoidea and Kainophytic plants were initially mediated through fungi, and at the end of Cretaceous closer and more intimate interconnections seemingly began to establish at first with generative organs of both the gymnosperms and angiosperms and further on with other plant organs. Some years ago a hypothesis on the most possible ways of transitions of beetles to feeding and breeding on living vegetative organs of the higher plants from initial (xylo) mycetophagy was proposed (Kirejtshuk 1989). According to it such transitions had an intermediate stage of feeding and breeding on the pollen and other parts of the generative plant organs: Appearance of complete (i.e. imaginal and larval) anthophagy or carpophagy is a rather important point for understanding the development of many phytophagous beetles which gave by present a wide scope of various trophic types. Primary formation of stable trophic connections with the: generative plant organs opened a perspective for further expansion on other plant organs to the beetles. When anthophagy arises, a special selective regime should be adhered. Essential component of this regime is some coincidence of insect development with the flowering period, i.e. the time of existence of available food and habitat. Not infrequent deviations in the time of maturation cause some advantages for the forms with a more rapid development. Acceleration of development is a more or less neces- sary attribute, which can be interpreted as an adaptation especially important for insects which acquire feeding and breeding in angiosperm flowers with a very short period of anthesis. This acceleration is accompanied by minimization in body size (or so-called miniaturization), and in the sequel with some pedomorphic transforma- tions in imaginal structures as well as with desembryonization of larvae. Inhabitance of larvae inside or near concentrated food resource induces an inactive mode of life facilitating, in turn, desembryonization of larval development. The adaptive tenden- cies of the forms living in flowers and gymnospern cones are the same. It is possible to observe a convergent similization between the representatives of the Jurassic Parandrexidae (Kirejtshuk 1994b) and some extant Propetes from the Epuraeinae here described. That allows to postulate a syndrome of anthophagization at least for the infraorder Cucujiformia (including Chrysomeloidea). A more common type of interactions between beetles and flowers or gymnosperm male cones looks like a simple feeding of insect adults on pollen. Many authors argue the cantharophilous theory considering that such interactions are unilateral or even Evolution of anthophilous Nitidulidae 113 negative for any specimen of flowering plants. However we must admit a participa- tion of beetles in the pollination of plants, including the participation of species whose larvae are not associated with plants. Imagines visited flowers occur among almost all Nitidulid subfamilies (except Calonecrinae, Amphicrossinae and Cyboce- phalinae). But an obliging attendance of imagines in generative plant organs with larval development in other localities is known among some representatives from the Epuraeinae and Nitidulinae. One of the ways of acquiring generative plant organs can be observed in the mode of life of Brachypeplus barronensis Blackburn, 1902 connected with cycad Macroza- mia communis in the temperate and subtropical rainforests of Eastern Australia. Imagines and larvae of this Nitidulid species inhabit the apex of cycad trunks between bases of young leaves and inside both male and female cones of this cycad. Together with the Nitidulids some coccids and pseudococcids live in the same places which yield a lot of sweet exudation which, in turn, give a good resource for growing yeast. The larvae and imagines of Brachypeplus feed on these fungi and the spores lavishly produced during a comparatively long period by male cycad cones. Inhabitance of the coccids in cycad cones is rather usual at present as it was in the past (Tang 1987). Joint life of beetles and Homoptera in cones of the plants could be an initial stage of transition to regular pollinophagy (or spermatophagy). The second example gives the mode of life of Australian Brachypeplus kemblensis Blackburn, 1902 which inhabits inflorescences of Alocasia macrorrhyza from the Araceae (Shaw & Cantrell 1983). The inflorescence of this plant is presented with a long spadix, the basal part of which is covered with a rather big leaf sheet, restricting the free moving of insects in the middle of the spadix inside the inflorescence. Only flat and small beetles (Nitidulid Brachypeplus kemblensis and one Omaliinae species) can easily get in and out. At a certain stage of development of these inflorescences the apical staminate part of the spadix begins to decay assisted by the fungus Fusarium producing a substrate for developing dipteran and coleopterous larvae. Larvae of Omaliinae, perhaps, act mainly as predators of dipteran larvae (Neuro- chaeta), but larvae of Brachypeplus feed on both pollen and fungi in soft decaying spadix. A classic example for the development of anthophagy is a mode of life of the Nitidulid Neopocadius and Oxycorinid Hydnorobius in flowers of the parasitic Neo- tropical plant Prosopanche from the Hydnoraceae (Bruch 1923). According to the opinion of Fegri and van der Peil (1982) the flowers of Hydnoraceae have some resemblance to the bisexual cones of Cycadeoides from the Bennettitales in the hard lignificated outer side as well as in the soft parenchyma of the inner side of these flowers. Moreover, Delevoryas (1968) admitted a considerable similarity also in the character of the holes perforated by insects in both species of plants (extinct and present). After recent studies it is possible to trace independent transition on the angiosperm flowers and gymnosperm cones of some groups from the Aethina- complex of Nitidulinae genera, which includes 9 taxa: Neopocadius, Brounthina gen. n., Idaethina, Anister, Olliffura, subgenus “A”, Aethina, Circopes, Ithyra. Some aspects of relations between these taxa we can see in fig. 1. The only species of Neo- 114 A. G. Kirejtshuk NEOPOCADIUS ee BROUNTHINA gen. n. —— IDAETHINA EI OLLIFFURA Er —— subgenus A AETHINA s. str. Bm CIRCOPES Y en ÓN ITHYRA ANISTER Fig. 1: Composition and tentative relations of the Aethina-complex of genera. pocadius has been mentioned above. Brounthina aequalis gen. et sp. n. from New Zealand is known from imagines labelled without bionomical data. Representatives of Anister from the Afrotropical and Cape regions, and also from Mediterranean and Indochina are exclusively anthophagous in imaginal stage and miners in larvae. Some endemic Australian species of Idaethina feed and breed inside seed capsules of Brachychiton (Sterculiaceae), and others are connected with seeds of plants from the Proteaceae. The rest groups of the Aethina-complex of genera can be divided into two pairs of taxa (Olliffura and subgenus “4”, on the one hand, and Circopes and Ithyra, on the other hand) especially close to Aethina s. str., linking both pairs. The subgenus Olliffura is distributed only in the Indo-Malayan, Australian and Papuan regions. Both larvae and imagines of this subgenus live in flowers of the Malvaceae (mainly Hibiscus). The subgenus “4” includes only 4 species (3 of them remaining undescribed: 1st — from Himalayas and Indochina, 2d — from Sulawesi, 3d — from Australia, 4th — New Hebrides; for now there is no information on their bionomy). The /thyra species occur in the Afro-Madagascarean regions, one of them recorded from Yemen and Sicilia. Species of this group are connected with flowers of Acanta- ceae. The subgenus Circopes is composed of species from different regions of the Eastern Hemisphere (including Australia), though in the Palaearctic region the habi- tat of this group does not extend beyond the Palaearctic or East-Chinese subregion. Imagines of some Circopes species visit both flowers and tree fungi, others have been collected only on blossoming plants, and others show adherence to inflorescences of monocotylodonous plants or cones of cycads where their larvae develop. Finally, the subgenus Aethina s. str. has the widest distribution (except a main part of the Holarc- tic regions, as well as Australian, Papuan, Polynesian and Patagonean regions). Many species of the last group are in all active stages of their living circle connected with substrates decaying with an assistance of fungi, including soft fruits of angiosperms, but some species of it visit blossoming plants, and Aethina tumida can breed in stores of pollen and honey in nests of the domestic honey bee. Evolution of anthophilous Nitidulidae 115 Thus, feeding on living plants has been independently acquired at least by 5 groups of this genera complex: 1. Neopocadius, 2. Idaethina, 3. Anister, 4. Olliffura, 5. Circopes-Ithyra. Anthophagy in the Olliffura species was establishing, perhaps, with the separation of this group from Athina s. str. or later (in case we get to know that the subgenus “4” species are not anthophagous). Anthophagy of Circopes and Ithyra appeared when their common ancestor had been secluded from the Aethina s. str. as some of them preserve an adherence to fungi. Larval mining and imaginal anthophagy of Anister is correspondent with the stage next after complete anthophagy in my scheme of probable ways of trophic transformation. The complete (imaginal and larval) carpophagy of the /daethina species could also be a con- sequence of regular ways of transition mentioned above. In the considered complex of genera it is easy to trace decrease in the body size of imagines and larvae (miniaturization), and also some correlation in pedomorphic transformations of imaginal structures with the profundity of anthophagization. These pedomorphic transformations are expressed in general simplification of struc- tures, shortening of elytra, reduction of surface sculpture and pubescence. As an expression of adaptive tendency to larval desembryonization we can regard a progres- sive reduction of many body appendages and chaetotaxy as well as simplification in mouth parts. Mining larvae of Anister species have obsolete appendages on tergites and lacking urogomphi. Thus, in the considered complex of genera we can observe transformation of larval structures from the silphoid type of differentiation especially among true mycetophagous forms to the bruchoid type in the phyto- phagous Anister. Other anthophagous Nitidulidae: 1. Subfamily Epuraeinae. Complete anthophagy is recorded for some representatives of at least 3 genera of this subfamily (Epuraea, Propetes = Amystrops syn. n. = Platychoropsis syn. n. and Grouvellia), though this feature can be expected among many forms with unstudied bionomy having an appearance which gives reason for such an assumption. Imagines of many others visit flowers being obliged to or regularly. 2. Subfamily Carpophilinae. Many adults visit flowers, and in some cases the pollination of the plants with participation of some species of this subfamily is recorded. The species of genus Nitops stat. n. and in particular Nearctic and (?) Neotropical Urocarpolus subgen. n., aS well as Nearctic and Neotropical subgenera Caplothorax subgen. n. and Plapennipolus subgen. n. of genus Carpophilus are connected with flowers in both active stages of their living circle. 3. Subfamily Meligethinae All adults and larvae of all species of this subfamily subendemic for the Old World with known biology live and feed in flowers of the angiosperm plants (mainly dicoty- lodons). This group is rather monogenous in the structural, ecological and bionomi- cal aspects of all adaptive tendencies of anthophagization. 4. Subfamily Nitidulinae This subfamily includes 5 groups where a trophic transition from primary myceto- phagy to anthophagy and phyllophagy has occurred or is happening now. The recent representatives of the genus Perilopsis known from Chile, Australia and New Guinea 116 A. G. Kirejtshuk seem partly to preserve an ancient connection with the male cones of the gymno- sperm Araucaria and Agathis, though the Chilean species (P flava (Reitter, 1873), non Kirejtshuk, 1986b) lives in Nothofagus inflorescence. Structures of small bodies of the Perilopsis larva and imago are scarcely modified in comparison with those of other groups in the Nitidula-lineage represented by the forms sharing a complete mycetophagy. Anthophagization in the Aethina-complex of genera has been considered above. The third group of the Nitidulinae with an expressed anthophagization is the tribe Cychramini. Regarding almost a world-wide occurrence of the Cychramus species, an early origin of this genus can be admitted. The known larvae of these species breed in fungi, but imagines, at least in the Holarctic, Australian and Neotropical regions, are more or less regular visitors of flowers. Species of other genera of this tribe with known bionomy have a rather small body and are larval miners and imagi- nal feeders on flowers (mainly pollen) of the Brassicaceae (Xenostrongylus, Oxystrongylus, Strongyllodes). The mining larvae of Xenostrongylus have a con- siderable resemblance to the meligethine larvae, but their mandibles with strong teeth along inner edge and an obsolete molar look more similar to those in the Chrysome- loidea and Curculionoidea than to those in other Nitidulidae. The Neotropical tribe Mystropini includes some genera (Mystrops, Cychropiestus, Platychorodes). All species of Mystrops with known bionomy live in palm inflores- cences. Imaginal structures of the above mentioned genera are rather similar and give evidence for supposing at least imaginal anthophagy for these groups. Body size of most representatives of this tribe does not exceed 3.5 mm (only the imaginal body of Cychropiestus corvinus with mandibles reaches to 7.0 mm). Imaginal structures : of head with appendages and abdominal apex have some resemblance to those in the Meligethinae, but the imaginal mandible is quite similar to that in some Chrysome- loidea, in particular among anthophilous forms. On the other hand, the larval legs of the Mystrops species as well as those in the Meligethinae, Xenostrongylus and predaceous Cybocephalinae have a characteristic adoral vesicle at the apex of the last segment of their legs. Emergence of this tribe, perhaps, took place comparatively not so long ago, ie. after separation of South America from other continents of the Southern Hemisphere. As a fifth group in the Nitidulinae, in which a process like anthophagization can be traced, there is the genus Camptodes from the tribe Strongylini, species of which develop in the decaying Opuntia flowers in Central and South America. 5. Subfamily Cillaeinae Many species from this subfamily are known as visitors of flowering plants and some of them live and breed in flowers during the larval period of their living cycle. Interconnections between the Cillaeinae and gymnosperm plants have scarcely a very ancient character. As a rule, the anthophagous forms have been recorded in the groups some species of which are connected with fungi. The members of genera Macrostola and Selis seem to be more or less specialized to inhabit flowers. Adaptive tendencies of anthophagization are not expressed in appearence of these species because of the recent appearance of their anthophagy. Many groups of the Cillaeinae prefer dead and decaying flowers and this preference is an additional evidence of the recent development of this feeding. Little-studied Cillaeinae of Hawaii need a parti- Evolution of anthophilous Nitidulidae 117 cular consideration in aspect of anthophagization and phyllophagization which for now can be supported by the data from labels of museum specimens. In contrast to the regions with temperate and subpolar climate in tropical and sub- tropical ecosystems it is possible to trace some relict interconnections of Nitidulids with gymnosperm cones and flowers. Many anthophagous Nitidulid groups show a tendency to become more similar to other anthophagous groups in comparison with the appearance of their probable closer relatives. Disparity in degree of larval diffe- rentiation and similar simplification of imaginal structures give a comparative evidence to ascertain level and age of anthophagization. Among the considered anthophagous groups a more or less ancient complete anthophagy can be recognized for the species of Perilopsis, Anister and some Cychramini from the Nitidulinae, and also for all representatives of the Meligethinae. The beginning of the anthophagiza- tion of these groups should be supposed near the Rubicon of the Mesozoic and Kainozoic eras. Anthophagization in some Epuraeinae, other genera of the Aethina- complex, the rest Cychramini and Mystropini from the Nitidulinae, perhaps, had a later start, could be in the Palaeogene. And species of the rest anthophagous groups of the Nitidulidae began to acquire this type of trophics even later. On the other hand, in some tropical regions a situation of unususal abundance of unspecialized anthophagous forms (mainly from the Epuraeinae) has been found, particularly expressed on islands of the Polynesian region with comparatively recent _ faunistic complexes. The last peculiarity gives reason to suppose that the anthopha- gization of some Nitidulid groups is continuing till now involving new forms with a mycetophilous mode of life. Taxonomical comments 1. Taxonomical references given in the catalogues by A.H. Grouvelle (1913), are here omitted. 2. On the synonymy of Propetes Reitter, 1875, Amystrops Grouvelle, 1906 and Platy- choropsis Grouvelle, 1912/1913: The recent study of most described and many un- named species regarded in composition of the mentioned taxa makes evident that all of them belong to a group of probably close relatives connected, in known cases, with the Pandanus flowers. As to synonymy Propetes Reitter, 1873 and Amystrops Grouvelle, 1906 (type-species: Amystrops modiglianii Grouvelle, 1906, here design- ated), it was established due to study the type specimen of the first proposed by monotypy (i.e Epuraea nigripennis Redtenbacher, 1867 — Lectotype, male, here designated and 3 paralectotypes — “Fidler, Ceyl, 860”; all specimens in Natural History Museum in Vienna) and thousands of representatives of the second taxon. After a more detailed consideration of a composition of true “Propetes” and “Platy- .choropsis” in as wide scope as possible I could not find any clear character to distinguish these groups. Perhaps, if an indistinct difference in convexity of body between “Propetes” and “Platychoropsis” can be admitted, it is still necessary to revise all members of this group in order to recognise this as a diagnostic feature (perhaps, some of them were described by L. R. Gillogly (1982) among Haptoncus Murray, 1864). Moreover, Propetes seychelensis sp. n. looks more like some species 118 xX GeKurejtshuk from Australia and Polynesia than Indo-Malayan ones, while P aquilus sp. n. has some features which put it apart from all other forms of the group. It can be sup- posed that the Epuraea group here united in Propetes is in a sense analogous with the Neotropical anthophagous Mystrops from the Nitidulinae connected with palm inflorescences. The Afro-Madagascarean Meligethinae also connected with inflores- cences of palms (Endrödy-Younga 1978) can be considered as another anologous group. The genus Propetes consists of a great number of species while other seemingly anthophagous genera and subgenera of the Epuraeinae are monotypic or represented by a few species (Apria Grouvelle, 1919; Parepuraea Jelinek, 1977; Grouvellia Kirejtshuk, 1984; Baloghmena Kirejtshuk, 1987; Polinexa Kirejtshuk, 1989b; Ceratomedia Kirejtshuk, 1990a; Mystronoma Kirejtshuk, 1990a, etc.). Though among the largest groups of the subfamily with a mainly mycetophilous mode of life (Epuraea Erichson, 1843; Haptoncus Murray, 1864; Micrurula Reitter, 1884) some representatives became anthophagous as well. In order to show a scope of structural variability of the considered group 4 aberrant forms are here described, but for two of them a new subgeneric taxon is proposed. The phylogenetic relations of the Propetes proposed for Platychropsis (Kirejtshuk 1986b) remain acceptable for this united group and according to the opinion formerly published the links of this group with some groups of Haptoncus Murray, 1864 can be expected; Tetrisus Murray, 1864 = Trimenus Murray, 1864, Tritesus Heller, 1916 and Baloghmena Kirejtshuk, 1987. The following synonymy can be summarized: Propetes Reitter, 1875 (= Amystrops Grouvelle, 1906, syn. n.; Platychorinus Grouvelle, 1906; Platychoropsis Grouvelle, 1912/1913, syn. n.; Haptoncognathus Gillogly, 1962 - see Grouvelle 1913 and Kirejtshuk 1986b). | 3. Propetes (Propetes) aquilus Kirejtshuk, sp. n. — figs 2—12. Specimens examined — Holotype, male and 53 Paratypes — “E slope Mt. McKinley, Danao Province, MINDANAO, 15: VIII; 46”, “CNHM. Philippine Zool. Exped. (1946—47), H. Hoogstraal & D. Heynemann leg”, “El. 3200—4200”, “in leaf axils on climbing Pandanus” (holotype and most paratypes are deposited in the Field Museum of Natural History in Chicago, some paratypes in the Zoological Institute of the Russian Academy of Sciences and the Zoological Research Institute and Museum Alexander Koenig in Bonn). Male, holotype. Length with mandibles 5.0 (and without 4.3), breadth 2.1, height 1.0 mm. Rather convex dorsally and slightly — ventrally; dorsum dark reddish brown with nearly black mandibles; ventral surface and appendages (except mandibles) reddish, though metepisterna, epipleura and ventrites somewhat darker; dorsum almost dull, but ventral surface almost with a moderate shine; body covered with fine, subrecumbent, moderately conspicuous, greyish golden hairs, length of which about 2.5 times more than the distance between their roots on dorsum; anal sclerite with a brush of long hairs at apex. Head, pronotal and scutellar surface with punctures 1.5—2.0 times as large as eye facets, smaller on head and becoming larger to pronotal base, interspaces between them about a puncture diameter and narrower at pronotal base, nearly alutaceous. Elytral surface somewhat as that on head and pronotum, but with punctures at base larger and denser. Surface of pygidium, preceding tergite and Evolution of anthophilous Nitidulidae 119 Figs 2—18: Propetes (Propetes) spp. — P (P) aquilus sp. n. (2—12): male (holotype: 2 — body from above; 3 — apex of left mandible, fronto-dorsal; 4 — antennal club; 5 — medial part of prosternum with process, ventral; 6 — idem, lateral; 7 — anal sclerite, ventral plate and spicula gastrale, ventral; 8 — tegmen, ventral; 9 — idem, lateral; 10 — penis, dorsal; female: 11 — fore part of head with mandibles, dorsal; 12 — ovipositor, ventral; P (BP) seychellensis sp. n. (13—18): male (paratype): 13 — body from above; 14 — antennal club; 15 — prosternal process, ventral; 16 — tegmen, ventral; 17 — idem, lateral; 18 — penis, dorsal. Scales: A — to figs 2, 13; B — to figs 5, 6; C — to figs 4, 7, 11, 14, 15; D — to figs 8—10, 12, 16—18. ventrites with small punctures about as large as eye facets, interspaces between them less or up to a puncture diameter, finely and distinctly microreticulated. Surface of metasternum with sparse punctures, somewhat larger than eye facets, interspaces 2—3 puncture diameters, smoothedly microreticulated. Prosternal surface with very 120 A. G. Kirejtshuk small and sparse punctures, with very fine and dense undulate microreticulation. Head more than twice shorter than the distance between eyes which consist of rather small facets, with elevated antennal insertions and a wide depression between them. Mandibles with a blunt process before acute apices. Labrum with a deep and wide excision between lobes. Antennae with long and dense hairs, much longer than head broad overreaching pronotal base, their club composing a fourth of total antennal length. Pronotum with a narrow border along perimeter (obsolete in the middle of fore edge), slightly excavated fore and hind edges, and sides gently sloping to extremely narrowly explanate lateral edges. Elytra with arched lateral and oblique apical edges, their sides steeply sloping to extremely narrowly explanate lateral edges. Elytral apices leave pygidium and preceding tergite uncovered. Pygidium with a truncate apex, under which a widely rounded apex of anal sclerite exposed. Terminal segment of maxillar and labial palpi rather long, although of usual structure. Antennal grooves scarcely expressed at arched sides of mentum. Prosternum flattened, its process scarcely medially curved and with a vertically abrupt and rather narrow apex. The distance between mid coxae subequal and that between hind ones about twice more than that between fore coxae. Mesosternum deeply excavated and with a well raised but not sharp medial ridge. Metasternum flattened, with medial suture well expressed nearly along the entire length, angularly excised hind edge between coxae, a well raised intercoxal line between mid coxae and caudal marginal lines behind mid coxae closely following hind edge of coxal cavities. 1st ventrite longest and without a trace of caudal marginal lines behind coxae. Hypopygidium with a nearly truncate or slightly bisinuate apex. Epipleura at base a little narrower than antennal club. Legs moderately raised. Tibiae comparatively short and subequal, nearly as long as prosternum and its process combined, somewhat narrower than antennal club, but a little wider than prosternal process: fore one finely crenulate along outer edge and with a strong subapical tooth curved ventrally; mid and hind ones with long hairs, particularly conspicuous in two rows along their outer edge, and a few subapical stout thorns. Femora with usual outlines of gently convex fore and hind edges: fore and mid ones about 1 and 2/3, but hind ones more than twice as wide as corresponding tibiae. Fore tarsi as wide as fore tibiae, mid and hind ones much narrower, claws moderately long and toothed at base. Tegmen well and penis trunk moderately sclerotized. Female. Externally differs from the male by more or less smaller prothorax and head, flattened surface of head, normally developed mandibles, shorter antennae with moderately raised scapus, less raised medial ridge on mesosternum, moderately rounded pygidial and widely rounded hypopygidial apices. Ovipositor with usual sclerotization. Variations. Minimal length with mandibles 2.7, but maximal one as that of the holotype (with mandibles 5.0 and without 4.3), breadth 1.4—2.5, height 0.8—1.0 mm. Sexual dimorphism is scarcely exhibited in the smallest males which differ from the females mainly by a little longer antennae as well as apices of pygidium and hypopygidium. The same holds true for the medial ridge on mesosternum which is well developed in larger males and rather weakened in smaller ones. A certain variability is expressed in coloration, punctuation, sculpture and pubescence. Evolution of anthophilous Nitidulidae 121 Notes. P (BP) aquilus sp. n. differs from all other members of the genus by comparatively large and robust body, particularly by its unicoloured dark dorsum, a well raised medial ridge of male mesosternum and also by a combination of features in punctuation, sculpture and pubescence of body sclerites. Besides that, this new species shows a resemblance with species of Mandipetes subgen. n. in having of a narrow prosternal process as well as expressed intercoxal line between mid coxae and caudal marginal lines behind mid coxae. Finally, its toothed tarsal claws are also enough diagnostic for this new species among the consubgeners. According to the labels pinned under the studied specimens they have been collected as well as those of P (Mandipetes) longipes subgen. et sp. n. “in leaf axils”, though attendance of these species in inflorescences of Pandanus for feeding and larval development is more probable in comparison with a possibility of their links with leaves or fungi seemingly growing in wet sites on vegetative organs of Pandanus. 4. Propetes (Propetes) seychellensis Kirejtshuk, sp. n. — figs 13—18. Specimens examined — Holotype, male and 1 paratype, male “Seych., Mahe’, Mare aux cochous, 15. 12. 1992, Malicky” (holotype is deposited in Natural History Museum in Vienna and paratype — Zoological Institute of Russian Academy of Sciences). Male, holotype. Length with mandibles 4.0 (and without 3.5), breadth 1.4, height 0.6 mm. Weakly convex dorsally and ventrally; straw coloured with black eyes; body with a faint shine, partly almost dull; thoracic segments and elytra with very fine, subrecumbent, scarcely conspicuous hairs, length of which about the distance between their roots on elytra, but tergites uncovered by elytra and ventrites with denser and more conspicuous yellowish pubescence; anal sclerite with 4, hind edge of hypopygidium with some and other sclerites of abdomen with a few rather long and thick reddish setae. Head surface with distinct punctures a little larger than eye facets, interspaces between them a little more or less than a puncture diameter, nearly alutaceous, especially at base. Surface of pronotum, elytra and metasternum somewhat as that on head, but with punctures somewhat larger, interspaces between them 2—3 puncture diameters, with dense, cellular and partly smoothed microreticulation. Surface of pygidium, preceding tergites and ventrites with punctures about twice smaller than eye facets, interspaces between them approximately as those on head surface, finely and densely microreticulated. Prosternal surface with very small, scarcely visible and sparse punctures, with very fine and dense undulate microreticu- lation. Head more than twice as short as the distance between eyes consisting of comparatively large facets, with rather elevated antennal insertions and a wide and shallow depression between them. Besides it, a medial “endocarina” is traced in basal half. Mandibles rather narrow and long with simple acute apices strongly curved almost turning backwards. Labrum with a deep and wide excision between lobes. Antennae with rather long hairs, reach scutellar apex, their scapus much bigger than their club (nearly as long as the distance between antennal insertions), their club composing about a fifth of total antennal length. Pronotum widest at fore half and almost bisinuate at base, with a narrow border along perimeter, almost convex fore 122 A. G. Kirejtshuk and hind edges, disk flattened and sides comparatively steeply sloping to unexpla- nate lateral edges. Elytra with arcuate lateral and oblique apical edges, their sides steeply sloping and a little curving ventrally at unexplanate lateral edges. Elytral apices leave pygidium and 2 preceding tergites uncovered. Pygidium with a truncate apex, under which a widely rounded apex of anal sclerite is exposed. Terminal segment of maxillar and labial palpi rather long, but of usual structure. Antennal grooves completely untraced at arched sides of mentum which slightly more than twice as wide as long. Prosternum flattened, its process medially somewhat curved and widened before flat and widely rounded apex. The distance between mid coxae twice than that between fore ones and a little less than that between hind coxae. Mesosternum shallowly excavated and without any carina along the middle. Metasternum flattened, with medial suture well expressed nearly along the entire length, shallowly archedly emarginate hind edge between coxae, without a trace of intercoxal line between mid coxae and caudal marginal lines behind mid coxae. Ist ventrite longest and without a trace of caudal marginal lines behind coxae. Hypo- pygidium with a widely rounded hind edge. Legs moderately raised. Tibiae compara- tively short and subequal, nearly as long as prosternum with process, much wider than antennal club, but a little narrower than prosternal process: fore one finely crenulate along outer edge and with 2 strong subapical teeth; mid and hind ones with long hairs, particularly conspicuous in two rows along their outer edge and 1—2 strong subapical spines. Femora with usual outlines of gently convex fore and hind edges, fore and mid ones about 1 and 2/3, but hind ones more than twice as wide as corresponding tibiae. Fore tarsi about 2/3 width of fore tibiae, mid and hind ones significantly narrower, claws moderately long and simple. Tegmen well and penis — trunk moderately sclerotized. Variation. Length with mandibles 3.1 (and without 2.7), breadth 1.2 mm. The second studied specimen (paratype) has smaller head less convex at base, somewhat narrower pronotum, less raised mandibles and antennae. Notes. P seychellensis sp. n. has some more resemblance to the species described by L. R. Gillogly (1962) in the Haptoncognathus as well as to P puberulus (Kirejts- huk 1986b), comb. n. and P subcalvus (Kirejtshuk 1986b), comb. n. (both described as Platychoropsis). This new species is characterized by a rather short head without bulbed temples, but with extremely curved mandibles and rather enlarged antennal scapi, a wide and comparatively short pronotum with straight, almost convex fore edge and gently rounded hind corners and also rather sparse dorsal punctuation. This combination of characters allows to determine it from the species given in the key elaborated by A.G.Kirejtshuk (1986b). Moreover, a traced medial “endocarina” is unknown among other representatitives of the family at all. 5. Mandipetes Kirejtshuk, subgen. n. Type species: Propetes (Mandipetes) longipes subgen. et sp. n. Includes also PR (M.) intritus subgen. et sp. n. As yet recorded from Vietnam and Philippine islands. Notes. This group is quite distinct from Propetes s. str. mainly by the characters of the structure of mouth parts and some others mentioned below. As both species here included in Mandipetes subgen. n. share a more or less stable combination of derived features, a separate taxon seems advisable. Evolution of anthophilous Nitidulidae 123 Propetess.str.: (1) mandibles, if long, with a simple apex or, if short, with a small subapical tooth, although in a few cases (P (P) pacificus Gillogly, 1962, comb. n.) with apices represented by two long teeth; (2) mentum with usual shape and propor- tion, composed of a small or medium part of ventral surface of epicranium; (3) maxillar lobe and palpus comparatively long, but of usual structure; (4) sides of pronotum unnarrowed to its base forming acute apices of hind corners, or gently narrowed with widely rounded apices, and infrequently with a configuration as that in Mandipetes subgen. n., fore edge of pronotum usually with complete or rarely with obsolete carina; (5) legs of medium sizes, tibiae moderately widened apically; Mandipetes subgen. n.: (1) mandibular apex with two long teeth; (2) mentum rather enlarged and subquadrangular, composed of more than half of ventral surface of epicranium; (3) maxillar lobe very narrow and long with setae posteriorly oriented along its inner edge, and palpus very long, with ultimate and penultimate segments dilated at apices and with setae posteriorly oriented along its inner edge; (4) sides of pronotum narrowed as anteriorly as posteriorly and its hind corners with angular blunt apices, fore edge of pronotum with obsolete carina; (5) legs comparatively long and very narrow. Remarks on bionomy. General shape and peculiarities of mouth parts, antennae and legs, including wide lobes of 1—3 tarsomeres give an evidence on anthophagous mode of life for at least imagines of P longipes sp. n. It should be supposed that both species of Mandipetes subgen. n. are associated with Pandanus inflorescences as other species of the genus (subgenus Propetes s.str.). 6. Propetes (Mandipetes) intritus Kirejtshuk, subgen. et sp. n. — figs 19—25. Specimens examined — Holotype, male — “Vietnam, Tam Dao, Son Zuong, hills in valley, 21. II. 1962, O.Kabakov” (in Cyrillic letters) (holotype is deposited in Zoological Institute of Russian Academy of Sciences). Male, holotype. Length with mandibles 2.8 (and without 2.5), breadth 1.5, height 0.7 mm. Moderately convex dorsally and ventrally; bright reddish with somewhat darkened pronotum, thoracic sterna and 1—3 ventrites, and with chestnut brown elytra; body with a particularly bright shine; dorsum with extremely fine, scarcely visible, very short hairs; vertral surface with slightly conspicuous yellowish hairs, length of which somewhat less than the distance between their roots on ventrites. Head, pronotal, scutellar and elytral surface with distinct punctures about 1.5 times as large as eye facets, interspaces between them 2.0—3.5 puncture diameters (a little narrower on elytra), completely smooth. Pygidial surface with not quite distinct punctures a little larger than eye facets, interspaces between them somewhat more than a puncture diameter, densely and conspicuously microreticulated. Surface of metasternum and Ist ventrite with sparse distinct punctures, somewhat smaller than on dorsum, interspaces 1.5—2.5 puncture diameters, smooth, but surface of 2—5 ventrites with smaller and denser punctures and with a trace of microreticula- tion more expressed posteriorly. Prosternal surface with rather small punctures smal- ler than eye facets, interspaces between them with very fine and smoothed microreti- culation. Head about 1.5 times shorter than the distance between eyes which consist of medium facets, weakly convex and with a shallow depression between slightly elevated antennal insertions. Mandibles with a sharp process before acute apices. 124 A. G. Kirejtshuk le 1.0 mm Poss) 100000) ETA AA A A ES A | Figs 19—25: Propetes (Mandipetes) intritus sp. n., male, holotype: 19 — body from above; 20 — antennal club; 21 — mentum with maxillar and labial palpi, ventral; 22 — last segment of maxillar palpi; 23 — anal sclerite, ventral; 24 — tegmen, ventral; 25 — idem, lateral. Scales: A — to fig. 19; B — to figs 20, 21, 23; C — 24, 25. Labrum with a deep and wide excision between lobes. Antennae no longer than head broad, their club very narrow and composing about 2/7 total antennal length. Pronotum with a narrow border along its perimeter (obsolete in the middle of fore edge), excavated fore and slightly emarginate hind edges, and sides gently sloping to unexplanate lateral edges. Elytra with arcuate lateral and oblique apical edges, their sides steeply sloping to bordered but not explanate lateral edges. Elytral apices leave pygidium entirely uncovered. Pygidium with a truncate apex, from which a subangu- lar apex of anal sclerite exposed. Terminal segment of maxillar palpi very long with curved apex and a row of setae along its inner side. Terminal segment of labial palpi rather long and thin, though of usual structure. Antennal grooves scarcely expressed at sides of subquadrangular mentum which twice as wide as long and with emargi- nate fore edge. Prosternum flattened, its process scarcely medially curved and subpa- rallel at sides, with vertically abrupt apex a little narrower than antennal club. The distance between mid coxae 1.5 times and that between hind ones 3.0 times more than that between fore coxae. Mesosternum deeply excavated and with a weak medial carina. Metasternum slightly convex, with a medial suture well expressed along the distal 2/3, archedly emarginate hind edge between coxae, a well raised intercoxal line between mid coxae and caudal marginal lines behind mid coxae closely following hind edge of coxal cavities. Ist ventrite longest and without a trace of caudal margi- nal lines behind coxae. Hypopygidium a little shorter than 1st ventrite with a nearly truncate or slightly bisinuate apex. Legs with a moderate length, but comparatively Evolution of anthophilous Nitidulidae 125 narrow. Tibiae short and subequal, somewhat longer than prosternum and its process combined, nearly as wide as prosternal process: fore one finely crenulate along outer edge and with a medium subapical tooth; mid and hind ones with moderately long hairs, partly disposed in two rows along their outer edge. Femora with usual outlines of gently convex fore and hind edges: fore and mid ones about 1 and 2/3; but hind ones more than twice as wide as corresponding tibiae. Fore tarsi 2/3 as wide as fore tibiae, mid and hind ones much narrower, claws moderately long and toothed at base. Tegmen moderately and penis trunk weakly sclerotized. Notes. This new species differs from another known member of the subgenus, except aedeagal structures, by smaller body, wide base of male mandibles with a setae row along their basal part only, narrower and compact antennal club com- posing ?/7 total antennal length, the distance between mid coxae subequal with width of antennal club, tibiae about as long as prosternum and its process combined, toothed claws. On the other hand, P (M.) intritus subgen. et sp. n. has an appearance somewhat similar to P (P) nigripennis (Redtenbacher, 1867) and some other Indo- Malayan species from Propetes s.str., however, the body of this new species has a particularly bright shine and strongly reduced pubescence, distinctive pronotum archedly narrowed anteriorly as well as posteriorly, comparatively narrow antennal club, structure of male maxillar palpi, longer and narrower labial palpi, large mentum, narrower legs and toothed tarsal claws. The characters of this new species partly correspond with the description of Propetes bicolor (Grouvelle, 1910), comb. . n. (a species still unknown to me), but is formally distinguishable from the latter by 1ts shiny body with reduced pubescence and the shape of the pronotum (pronotum of P bicolorcomb. n. with emarginate fore edge, “angulis rotundatis” fore and back- wards with projected hind corners). 7. Propetes (Mandipetes) longipes Kiretshuk, subgen. et sp. n. — figs 26— 37. Specimens examined — Holotype, male and 1 paratype: “E. slope Mt. McKinley, Davao Prov., MINDANAO, 24 Sept.: 46”, “Elev.: 3300 ft”, “CNHM Philippine Zool. Exped. (1946—47), H. Hoogstraal & D. Heynemann leg”; 6 para- types: ibid. “.. [X:1:46”, “in leaf axils of climbing Pandanus”, “elev. 3000 ft”, “CNHM Philippine Zool. Exped. (1946—47), H. Hoogstraal & D. Heynemann leg” (holotype and 4 paratypes are deposited in the Field Museum of Natural History in Chicago and the remaining paratypes in the Zoological Institute of Russian Academy of Sciences. Male, holotype. Length with mandibles 5.5 (and without 4.6), breadth 2.2, height 1.0 mm. Moderately convex dorsally and ventrally; bright reddish with black elytra; rather shiny; dorsum with sparse, long, subrecumbent and quite conspicuous yello- wish hairs, length of which is nearly 5 times more than the distance between punctu- res; ventral surface with very short and slightly conspicuous pubescence. Head and pronotal surface with punctures as large as eye facets, interspaces between them 2—4 puncture diameters (and less at pronotal sides and base), smooth. Elytral surface with somewhat larger and denser punctures than those on head and pronotum, interspaces between them with a trace of alutation. Surface of tergites uncovered by elytra and prosternum with very small, not quite distinct and modera- tely dense punctures and cellularly microreticulated interspaces between them. Surface of ventrites and metasternum with less distinct punctures than on head, 126 A. G. Kirejtshuk pronotum and elytra which are somewhat smaller than eye facets, interspaces between them 1—2 puncture diameters on ventrites and 3—4 puncture diameters on metasternum, nearly as reticulated as that on pygidium and prosternum. Head 1.5 times shorter than the distance between eyes, composed of moderately small facets, and strongly concave between antennal insertions. Antennae a little longer than head broad, with 3-segmented club. Mandibles with 2 processes along inner edge and densely haired along inner ridge of their ventral surface. Labrum with deeply and widely separate lobes. Maxillar palpi with very long penultimate and ultimate segments widened at apex and densely haired along inner edge. Mentum very large and rather projected anteriorly. Pronotum with slightly and evenly convex surface and obsolete fore edge at neck. Elytra steeply sloping to narrowly bordered lateral edges and transverse apices, remaining pygidium and a part of preceding tergite uncovered. Exposed tergites and anal sclerite well sclerotized and rather convex. Antennal grooves on ventral side of epicranium unexpressed. Prosternum slightly convex with a narrow process slightly medially curved and vertically abrupt at apex. The distance between mid coxae about twice less, but that between hind ones twice more than the distance between fore coxae. Mesosternum deeply excavated and slightly convex at bottom. Metasternum flattened, with a well raised medial suture and a subangular emargination of hind edge between coxae. Hypopygidium almost 1.5 times longer than 1st ventrite and nearly with a transverse apex. Legs narrow, long and rather simplified. Tibiae somewhat narrower than antennal club: fore one finely crenulate, mid and hind — with 2 longitudinal rows of dense, thin and short hairs along their outer edge. Femora with usual outlines of gently convex fore and hind edges, more than twice as wide as tibiae. Fore tarsi somewhat wider, but mid and . hind ones somewhat narrower than the corresponding tibiae, with moderately raised claws slightly toothed at base. Aedeagus well sclerotized. Female. Externally differs from male by less wide fore edge of pronotum and head, different antennal club, suberected mandibles with less projected inner processes, much shorter penultimate and ultimate segments of maxillar and labial palpi, mentum less projected anteriorly, prosternal process more curved medially and with sloping edges, a little shorter legs, widely rounded or nearly subangulate pygidial apex and gently rounded hypopygidial apex. Ovipositor well sclerotized. Variation. Length with mandibles 4.4—5.5 (and without 4.0—4.6) mm. Pronotal disk of some paratypes more or less darkened. The second studied male (paratype) has in contrast with the holotype more arched pronotal sides, somewhat smaller head, different configuration of mandibles and prosternal process nearly as in females. Some variations are expressed in punctuation and sculpture of surface. Notes. P (M.) longipes sp. n. differs from another consubgener here described, except aedeagal structures, by larger body, much longer mandibles (particularly in males) with a narrower base and a row of setae along entire inner edge to base of a long subapical tooth, antennal club with loose articles, composing a fourth of total antennal length, distance between mid coxae considerably less than width of antennal club, tibiae much longer than prosternum and its process combined, comparatively long hypopygidium, almost simple tarsal claws. This new species is easily distinguishable from all the Propetes species by larger head with very wide fore edge and highly specialized characteristic mouth parts, shape of pronotum, convex Evolution of anthophilous Nitidulidae 127 A 20mm DB 10mm Figs 26—37: Propetes (Mandipetes) longipes subgen. et sp. n. Male, holotype: 26 — body from above; 27 — antennal club; 28 — ventral surface of head; 29 — fore leg, dorsal; 30 — mid tibia, dorsal; 31 — anal sclerite, ventral plate and spicula gastrale, ventral; 32 — tegmen, ventral; 33 — idem, lateral; 34 — penis trunk, dorsal; male, paratype: 35 — fore part of head with mandibles, dorsal; female: 36 — ventral surface of head; 37 — ovipositor, ventral. Scales: A — to figs 26; B — to figs 27—31, 35, 36; C — to figs 32—34, 37. and long male anal sclerite, narrow and vertically abrupt prosternal process, long and narrow legs, comparatively long hypopygidium, genitalia of both sexes. 128 A. G. Kirejtshuk 8. Caplothorax Kirejtshuk, subgen. n. Type species: Carpophilus melanopterus Erich- son, 1843. Includes type-species, Carpophilus rufus Murray, 1864, and some still undescribed species from the Neotropical region. Nearctic and Neotropical distribu- tion. Notes. The species of this subgenus are characterized by appearance partly similar to the species of Urocarpolus subgen. n. (perhaps, as a result of convergent evolution of anthophilous mode of life in both groups), but with male anal sclerite and genitalia of both sexes as those in other groups of the genus Carpophilus, but not as in the species of Nitops stat. n. Acute pronotal corners of the Caplothorax subgen. n. are very distinctive from those in other anthophagous forms from the two genera Carpophilus and Nitops stat. n. The new subgenus seems to have some relationship to subgenus Megacarpolus Reitter, 1919 in which some species from the New World should be included, but differs from it by more convex and oval body, peculiarities of sexual dimorphism and aedeagal structures as well as mode of life of its members. On the other hand, species of the Caplothorax subgen. n. seem to be related to those of Plapennipolus subgen. n., but much more convex and dull because of strongly contrasting microreticulation, and with compact and wide antennal club. Finally, the species of the Caplothorax subgen. n. has also some resemblance to the Papuan Carpophilus (Loriarulus) poggii Kirejtshuk, 1987 with unknown bionomy (including acute pronotal corners), but differs from the latter by much shorter last abdominal segment with a simple apex in female and convex dorsum. Perhaps, three subgenera (Loriarulus Kirejtshuk, 1987, Caplothorax subgen. n., Plapennipolus subgen. n.) have their phyletic roots among an ancestor group very close to the Megacarpolus species of which for now maintain a more plesiotypic: appearance and mycetophilous mode of life. The Nearctic Carpophilus longus Fall, 1910 seemingly connected with flowers of Yucca has an unclear position between Caplothorax subgen. n. and Plapennipolus subgen. n., though very different from both by rather slender body and Myothorax-like pronotum with long-fringed sides. 9. Plapennipolus Kirejtshuk, subgen. n. Type species: Colastus yuccae Crotch, 1874. Includes also ? Carpophilus rufiventris Schaeffer, 1911. Endemic Nearctic distribu- tion. Notes. This new subgenus is characterized a particularly flattened dorsally and ventrally body with rather wide elytra and abdomen (external appearance to some of Colopterus species from the Cillaeinae) and more loose and elongate antennal club in contrast with the member of Caplothorax subgen. n. and Megacarpolus (seemingly related to this subgenus). Its position has been regarded above in the consideration of Caplothorax subgen. n. and will be discussed more detailed in one of the further papers. ; 10. Genus Nitops Murray, 1864, stat. n. Type-species: Carpophilus (Nitops) ophthal- micus Murray, 1864. Composition (besides type-species) is referred to in Grouvelle (1913) and Dobson (1972); a group of species included here in Urocarpolus subgen. n. should also be inclued in genus Nitops. The taxa Nitops and Endomerus proposed by Murray (1864) were regarded as synonyms by some authors (Sharp 1887—1905; Grouvelle 1913; Junk’s catalogue and others). In this paper the tradition is tentatively continued until a more detailed revision of this group is available. Evolution of anthophilous Nitidulidae 129 Notes. This genus in contrast with all other groups of the Carpophilinae is characterized by unexcised hypopygidial and abruptly transverse pygidial apices in males, forming together a round foramen for the apically projected anal sclerite. Only one species of the Carpophilinae known from Himalalya and Indochina (Urophorus (Anophorus) prodicus Hinton, 1944) has a similar abdominal structure, i.e. male hypopygidium of this species without the depression characteristic for Anophorus Kirejtshuk, 1990b, but distinctly and widely emarginate at apex. As structure of genital capsule and aedeagus in both mentioned cases is a derived character, a secondary (? reversal) and independent development of the last abdomi- nal segment can be admitted for both species of Nitops stat. et sensu n. and U. (4.) prodicus. The groups of species here regarded as subgenera of Nitops stat. et sensu n. are characterized also by a compact 3-segmented antennal club with the largest ultimate segment (most of the members of Carpophilus have 9th segment partly loose from consolidated 10th and 11th ones). Besides that, the species of Nitops stat. et sensu n. differ from all groups of the subfamily, having strongly convergent antennal grooves and very large eyes composed of comparatively large facets, as well as a highly specialized structure of ovipositor. The species of both subgenera of the considered genus according to the labels attached to museum specimens are usual visitors of flowers of the angiosperms and Connell (1956) had recorded the develop- ment of larvae N. (U.) floralis comb. n. in flowers of Opuntia opuntia. 11. Urocarpolus Kirejtshuk, subgen. n. Type species: Cercus pallipennis Say, 1823. : Other species in this subgenus: Carpophilus floralis Erichson, 1843; Carpophilus mexicanus Reitter, 1873 (? = nigrovittatus Parsons, 1943); Carpophilus longiventris Sharp, 1889; ? Carpophilus obtusicollis Reitter, 1873. Distributed in the Nearctic region and Mexico. Notes. The subgenera of Nitops stat. et sensu n. can be diagnosed after the following features: Nitops s. str.: (1) eyes very large and consisting of large facets, temples not raised; (2) antennal grooves strongly nearly rectilinearly convergent; (3) length of elytra subequal or usually more than their combined width; (4) female with a pygidial apex with rounded or subtruncated hind edge and a beetled process in the middle; (5) Ovipositor wide with unnarrowed and unforked gonocoxites subtruncate at apex. Urocarpolus subgen. n.: (1) eyes moderately large and consisting of moderately small facets, temples exposed behind them; (2) antennal grooves strongly curved and feebly convergent; (3) length of elytra considerably shorter than their combined width; (4) female pygidial apex subacute; (5) ovipositor with slightly modified scleri- tes of gonocoxites narrowed to the forked apex. 12. Brounthina aequalis Kirejtshuk, gen. et sp. n. — figs 38—46. Specimens examined — Holotype, male: “sp”, “Mt. Owen, 26-2-14”, “New Zealand, Broun Coll. 1922—482”; paratypes: 2 — “Glen Hope, 20-2-15”, “New Zea- land, Broun Coll. 1922—482”; 1 — “Pacuratani, 2-1-1915”, “New Zealand, Broun Coll. 1922—482”; 2 — “New Zealand, Gollanz Valley, 1-II. 1924, in fungus, G. V. Hudson”; 1 — “New Zealand, Wallington, — II. 1924, in fungus, G. V. Hudson”; 2 — “NZ. Nelson Lakes N P., 2000” 28—30. V. 1976, O.Kukal”; 1 — “N. ZEAL., BR, Punakaiki, Porarari R., 29. XII—3. I. 1984, 35 m, L. Masner, Nothof., prim. 130 A. G. Kirejtshuk for”; 2 — “N. Zeal.; S. Isl., 30 km W Collingwood, Mangarakau, 50 m, 20. V. 82, S. & J. Peck, mixedforest litter*; 1 — ibid... “13 km NW Takara Washburn Res., 10 m, 19. V. 82, S. & J. Peck, beech log litter”; 2 — ibid... “BR. Nelson Lks. N. P., LK Rotoiti, St. Arnaud Track, 670 m, (14-XII-1984)—(6-I-1985)”, “Nothofagus spp. for., log & leaf litter, A. Newton & M. Thayer”; 2 — ibid... “Mt. Robert Road, 660 m, (26-X11-1984)—(6-1-1985)”, “Leptospermum-Nothof. scrub log & leaf litter, A. New- ton & M. Thayer”; 1 — ibid... “N Slope Mt. Robert, Speargrass Tr., 880 m, (21-XII-1984)—(6-I-1985)”, “Nothofagus spp. for., tree crotch litter, A. Newton & M. Thayer”; 1 — ibid... “SD, Tennyson Inlet, E Side Duncan Bay, 30 m, (15-XII-1984)— (5-1-1985)”, “Nothofagus for. litter u. palm leaf litter, A. Newton & M. Thayer”; 1 — ibid... “SD, Tennyson In-let, W Side Te Mako Bay, 125 m, (15-XII-1984)— (S-1-1985)”, “Nothofagus-podo-pdwd. log & leaf litter, A. Newton & M. Thayer”; 1 — ibid... “NN, Takara R., Cobb Dam., Asbestos For. Walk, 410 m, 2-I-1985”, “Nothofagus-podo-pdwd. log & leaf litter, A. Newton & M. Thayer” (holotype and 4 paratypes are deposited in the Natural History Museum in London; other para- types in Zoological Institute of Russian Academy of Sciences, Field Museum of Natural History in Chicago, Biosystematic Research Institute in Ottawa, Canadian Museum of Nature in Ottawa and Zoological Research Institute and Museum Alex- ander Koenig in Bonn). Male, holotype. Length 4.3, breadth 2.1, heisht 1.0 mm. Moderately convex dorsally and slightly — ventrally; unicoloured reddish with slightly lighter mouth parts and legs; with a feeble shine; dorsum with moderately dense, long, sub- recumbent and contrasting conspicuous yellowish golden hairs, length of which is nearly 3 times more than the distance between their roots; ventral surface with . somewhat short and finer, slightly conspicuous hairs, length of which about twice more than their roots; pronotal and elytral sides moderately ciliate. Head and pronotal surface with punctures about 1.5 times as large as eye facets, interspaces between them nearly a third puncture diameter, densely and cellularly microreticulated. Elytral surface with shallower, smaller, more sparse and less distinct punctures than those on head and pronotum, interspaces between them larger, but as reticulated as those on head and pronotum. Pygidial surface densely and finely punctured, with narrow interspaces densely and cellularly reticulated. Surface of thoracic sterna and 1—4 ventrites with distinct punctures nearly as large as eye facets, interspaces between them 1.5—2.5 puncture diameters, smoothed or smooth on thoracic sterna and reticulated on ventrites. Hypopygidial surface similar to that on 1—4 ventrites, but punctures much larger and with narrower interspaces. Head 1.5 times shorter than the distance between eyes, composed of rather small facets, and concave between antennal insertions. Antennae a little shorter than head broad with 3-segmented and compact club. Mandibles moderately exposed from under frons and labral lobes deeply and widely separated. Maxillar and labial palpi with moderately developed segments, last ones of labial palpi slightly bulbous with oblique apex. Mentum pentangular 2.5 times as wide as long. Pronotum with slightly and evenly convex surface, with a narrow border along base and sides. Elytra steeply sloping to narrowly bordered sides and to acute apices, almost completely covered pygidium. Pygidial apex almost transverse and remaining uncovered angular apex of anal sclerite. Antennal grooves quite distinct and convergent along inner sides. Evolution of anthophilous Nitidulidae 131 Figs 38—46: Brounthina aequalis gen. et sp. n. Male: 38 — body from above; 39 — fore part of head with labral lobes, dorsal; 40 — antennal club; 41 — fore tibia, dorsal; 42 — prosternal process, ventrally; 43 — idem, laterally; 44 — tegmen, ventral; 45 — penis trunk, dorsal; female: 46 — ovipositor, dorsal. Scales: A — to fig. 38; B — to figs 39—43; C — to figs 44—46. Prosternum slightly convex with a process moderately medially curved and almost vertically abrupt at apex. The distance between mid coxae a little more than that between fore ones and that between hind ones, in turn, a little more than the distance between fore coxae. Mesosternum rather excavated and a medial carina at bottom. Metasternum flattened, with well raised medial suture and shallower emargination of hind edge between coxae. Intercoxal line well expressed and disposed far behind fore edge of metasternum almost rectalinearly joining the middles of mid coxal cavities. Caudal marginal lines behind mid coxae well expressed and forming a moderately large triangle at fore corner of metasternum (laterosternite). Caudal marginal line behind hind coxal cavities gently and archedly deviated in medial part of hind edge of coxae and returned to it at the middle of cavity. Hypopygidium 1.5 times longer than 1st ventrite and with distinctly angular apex. Legs rather stout and short. Tibiae triangular: fore one 1 and !/3 wider than antennal club and finely crenulate along outer edge; mid and hind ones a little wider than antennal club and 132 A. G. Kirejtshuk with 2 longitudinal rows of not dense, rather long and moderately thick hairs along their outer edge. Femora with usual outlines of gently convex fore and hind edges: fore one nearly as wide as, mid and hind ones 1.5 times wider than corresponding tibiae. Fore tarsi nearly as wide as antennal club, but mid and hind ones much narrower, with simple and long claws. Aedeagus well sclerotized. Female. Externally differs from male by narrower fore tibiae and tarsi (former as wide as and latter narrower than antennal club), widely rounded pygidial and hypopygidial apices. Ovipositor moderately sclerotized. Variation. Length 2.7—5.0 mm. General coloration varies from light reddish to dark brown, frequently with more or less lightened pronotal sides and appendages, or sometimes scutellar parts of elytra and head also lightened. Large specimens are with comparatively wider pronotal and elytral bases and more narrowed sides of pronotum anteriorly and elytra posteriorly. Elytral surface frequently has a rasp-like view. Pubescence in many specimens recently collected strongly contrast. Some varia- tions are expressed in punctuation and sculpture of surface. Notes. This monotypic genus has an intermediate position between the genera Idaethina Reitter, 1875 from Australia and Neopocadius Grouvelle, 1906 from Argen- tina, but differs from both groups by more deeply excised labrum, carinate mesoster- num, well expressed intercoxal line disposed far behind fore edge of metasternum, more approached one to another hind coxae and characteristic genital structures in both sexes. Besides that, Braunthina aequalis gen. et sp. n. is distinct from the first also by unhaired eye facets, uniform elytral punctuation and weak sexual dimor- phism in tibial shape, and from the second — by character of dorsal punctuation - and pubescence and unwidened apex of prosternal process. Generic name of this new genus is created to be devoted to T. Broun who more successfully investigated the Coleoptera of New Zealand than anyone before him. 13. Subgenus Olliffura Jelinek & Kirejtshuk, 1986 was proposed in Kirejtshuk (1986a). 14. Genus Strongyllodes Kirejtshuk, 1992 is mainly distributed in the Madagas- carean, Indo-Malayan and Australian regions coming to the north up to Korea and Russian Far East. This genus is quite distinctly separated from the Mediterranean Xenostrongylus Wollaston, 1854 by more strongly and evenly convex oval body, less anteriorly projected head and especially elytral epipleura steeply sloped downwards. Acknowledgements I am most grateful to the Organizing Committee of the International Symposium on systematics and biodiversity of tropical ecosystems, in particular to C. M. Naumann, H. Ulrich, W. Böhme, M. Schmitt and F. Krapp whose efforts allowed me to participate in the symposium. I greatly appreciate the assistence of many coleopterists who provided materials for my studies, including M. J. D. Brendell and P. M. Hammond of the London Museum, M. Jách and H. Schónmann of the Vienna Museum, A. F. Newton and M. Thayer of the Chicago Museum, J. Pakaluk and N. Vandenberg of the Washington Museum, A. Smetana, E. C. Becker and B. Gill of the Biosystematic Institute at Ottawa, and B. Anderson and F. Génier of the Ottawa Museum. Evolution of anthophilous Nitidulidae 133 Zusammenfassung Die Arbeit behandelt das mehrfache unabhängige Auftreten von Anthophagie bei verschiede- nen Gruppen von Glanzkäfern der Familie Nitidulidae. Verlauf und Regelhaftigkeiten dieses Evolutionsprozesses werden dargestellt. Ähnliche Korrelationen in den Abwandlungen von Strukturen, Ernährungs- und Lebensweisen von anderen, nicht näher verwandten anthopha- gen Gruppen werden gezeigt. Propetes (Propetes) aquilus sp. n., P (P) seychellensis sp. n., P (Mandipetes) intritus subgen. et sp. n., P (M.) longipes sp. n., Brounthina aequalis gen. et sp. n., Caplothorax subgen. n. und Plapennipolus subgen. n. in der Gattung Carpophilus und Urocarpolus subgen. n. in der Gattung Nitops stat. n. werden als neue Taxa vorgeschlagen und beschrieben. References Bruch, C. (1923): Coleopteros fertilizadores de “Prosopanche Burmeisteri” De Bari. — Phy- sis (Rev. Soc. Argent. Cienc. Natur.) 7: 82—88. Connell, W. A. (1956): Nitidulidae of Delaware. — Agricult. Experemental Sta. Univ. Dela- ware, Newark, 318: 1—67. Delevoryas, T. (1968): Investigations of North American cycadeoids: structures, ontogeny and phylogenetic consideration of cones of Cycadeoidea. — Palaeontographica (A). 121: 02133. Dmitriev, V. Yu. & V. V. Zherikhin (1988): Changes of diversity of insect families accordingto the data of method of accumulation of novelties. — In: Cretaceous biocenoti- cal crisis and evolution of insects, 208—215. Moscow, Nauka (in Russian). Dobson, R. M. (1972): A new species of Carpophilus Stephens (Coleoptera: Nitidulidae from Florida, U.S.A., and a reappraisal of the type material of Carpophilus ophthalmicus Murray. — J. Ent. (B), 41, 2: 145—150. Endrody-Younga, S. (1978): Systematic revision and phylogeny of some Meligethinae genera from the Ethiopian region (Coleoptera, Nitidulidae). — Entomol. Gener. 4: 295—316. Fegri,K. & A. van der Peil (1982): Principles of ecology of pollination. — Moscow, Mir, 1—379 (in Russian). Gillogly, L. R. (1962): Coleoptera: Nitidulidae. — In: Insects of Micronesia (Bernice P. Bishop Museum, Honolulu), 16, 4: 133—188. Gillogly, L. R. (1982): New species and a key to the genus Haptoncus (Coleoptera: Nitiduli- dae). — Pacific insects, 24, 3—4: 281 —291. Grouvelle, A. (1913): Byturidae, Nitidulidae. — In: W. Junk (ed.) Coleopterorum Catalo- gus. — Berlin, 15, 56: 1—223. Grouvelle, A. (1919): Descriptions d’especes nouvelles de Coleopteres de l’Afrique australe. — Mem. Entomol., 2: 47—6l. Heller, K.M. (1916): Die Káfer von Neu-Caledonien und benachbarten Inselgruppen. — Nova Caled. Zool. 2, L. 3: 278—280. Hinton, H. E. (1944): A new nitidulid beetle from Burma, Carpophilus (Urophorus) prodi- cus sp. n. — Entomologist. 77: 172—173. Jelinek, J. (1977): Revision of the genus Epuraea Er. from Africa with remarks to related genera (Col., Nitidulidae). — Acta Entomol. Mus. Nation. Pragae, 39: 345— 397. Kirejtshuk, A. G. (1984): New taxa of Nitidulidae (Coleoptera) from the Indo-Malayan region. — Ann. Hist. — Natur. Mus. Nat. Hung. 76: 169—195. Kirejtshuk, AG. (1986a): Revision of the genus Aethina Er. (Coleoptera, Nitidulidae) of the fauna of the Oriental and Palaearctic regions. — Proc. Zool. Inst. Ac. Sci. USSR. 140: 44-82 (in Russian). KirejtshukA. G. (1986b): New genera and species of the Nitidulid beetles (Coleoptera, Niti- dulidae) from Australian region. 1. — Entomol. Obozr., 65, 3: 559—573 (in Russian). Kirejtshuk, A. G. (1987): New taxa of the Nitidulidae (Coleoptera) of the East Hemisphere (part 1). Omosita nearctica sp. n., vicariant with the palaearctic O. colon (L.). — Proc. Zool. Inst. Ac. Sci. USSR. 164: 63—94 (in Russian). 134 A. G. Kirejtshuk Kirejtshuk, A. G. (1989a): On formation of phyllophagy (phyllophagization) among beetles (Coleoptera). — Proc. Zool. Inst. Ac. Sci. USSR. 202: 147—182 (in Russian). Kirejtshuk, A. G. (1989b): New taxa of the Nitidulidae (Coleoptera) of the East Hemisphaere (part III). — Proc. Zool. Inst. USSR Ac. Sci., 164: 63—94 (in Russian). Kirejtshuk, A. G. (1990a): New genera and species of the Nitidulid beetles (Coleoptera, Nitidulidae) from Australian region. III. — Entomol. Obozr., 69, 4: 857—878 (in Russian). Kirejtshuk, A. G. (1990b): New taxa of the Nitidulidae (Coleoptera) of the East Hemisphere. Part 4. — Proc. Zool. Inst. USSR Ac. Sci. 211: 84—103 (in Russian). Kirejtshuk, A. G. (1992): Fam. Nitidulidae. — in: Identification manual of the insects of the Far East of the USSR. Coleoptera, or beetles. — Leningrad, Nauka. 3, 2: 114—210 (in Russian). Kirejtshuk, A. G. (1994a): System, evolution of mode of life and phylogeny of order Coleoptera (Insecta). Part 1. — Entomol. Obozr., 73, 2: 266—288 (in Russian). Kirejtshuk, A. G. (1994b): Parandrexidae fam. n. — an extinct family of the superfamily Cucujoidea (Insecta, Coleoptera). — Palaeontol. J., N 1: 57—64 + 1 tabl. (in Russian). Kirejtshuk, A. G. £ A. G. Ponomarenko (1990): Fossil beetles of the families Peltidae and Nitidulidae (Coleoptera). — Palaeontol. J., N 2: 78—88 (in Russian). Ponomarenko, A. G. (1983): Historical development of coleopterous insects. — Autho- synopsis of thesis of Doct. Biol. Sciences., Moscow, 1—47 (in Russian). Reitter, E. (1919): Bestimmungs-Tabelle der Coleopterenfamilien: Nitidulidae und Byturidae aus Europa und den angrenzenden Landern. — Verh. nat. Ver. Brunn, 56: 1—105. Shaw, D. E. & B. K. Cantrell (1983): A study of the pollination of Allocasia macrorrhiza (L.) G. Don (Araceae) in southeast Queensland. — Proc. Linn. Soc. N.SW. 106, 4: EDO Tang, W. (1987): Insect pollination in the cycad Zamia pumila (Zamiaceae). — Amer. J. Bot., 31590599: Dr. Alexander G. Kirejtshuk, Zoological Institute of Russian Academy of Sciences, — Sanct-Petersburg, 199034 Russia. Bd. 47 S. 135—138 Bonn, September 1997 Exceptional records of Microgale species (Insectivora: Tenrecidae) in vertebrate food remains Steven M. Goodman, Paulina D. Jenkins & Olivier Langrand Abstract. Records of Microgale species identified from scats and pellet remains of predators are reviewed. Cranial fragments of at least two individuals of Microgale mela- norrhachis and a humerus of M. talazaci were identified from faeces of Cryptoprocta ferox collected at Montagne d'Ambre in northern Madagascar. Remains of M. pusilla were identified from 7yto alba pellets collected on the human-modified Central High Plateau near Antananarivo. The records of M. pusilla suggest that this species is not particularly sensitive to environmental disturbance. Key words. Microgale spp., shrew tenrecs, prey remains, Cryptoprocta ferox, Tyto alba. Introduction Except for the study of Eisenberg & Gould (1970), conducted in the eastern rainforest of Madagascar, remarkably little has been published on the natural history and ecology of the shrew tenrecs (Family Tenrecidae, Subfamily Oryzorictinae) [sensu Hutterer (1993)]. In his recent revision of the genus Microgale, MacPhee (1987) summarized information on the distribution and habitat requirements of this genus based on museum specimens. Further, he identified remains of Microgale species from owl pellets, and in several cases these records represented exceptional range extensions and expanded our knowledge of shrew tenrec habitat requirements. In general, using standard trapping techniques Microgale species are relatively difficult for mammalogists to capture. The presence of Microgale remains in preda- tors’ food remains allow insight into the distribution of shrew tenrecs. In this paper we present some information on Microgale species found in the pellets of owls and the scats of carnivores, and discuss the implications of these data on the distribution and habitat specificity of shrew tenrecs. Results and discussion Scats of Cryptoprocta ferox Bennett, 1833. — On 15 April 1992, Dr. Chris Raxworthy collected four scats in the Montagne d’Ambre National Park (12°28’S, 49 °11’E), 40 km S Antsiranana. The average annual rainfall in the park is approximately 3600 mm (Nicoll & Langrand 1989). The scats were found at 900 m, next to Petit Lac (also known as Matsabory Mahasarika), a crater lake that lies in the heart of the forest, On a large flat rock within 20 m of the forest edge. This site is known to be frequented by Cryptoprocta and its scats are easily distinguished from other native and intro- duced carnivores (Rasolonandrasana 1994). On the basis of size and context, the scats were identified as those of Cryptoprocta. The vegetation surrounding the lake is composed of undisturbed rainforest, with an average canopy height of 30 m, and some emergent trees reaching 35 m. The understory is open and botanically diverse. 136 S. M. Goodman et al. The leaf-litter is thick and permanently moist. Inside the scats, cranial and post-cranial remains of at least two Microgale mela- norrhachis Thomas, 1882. The mandibular rami in which m3 are present show the characteristic shape of the talonid or the toothrows with distinct diastemata, both of which diagnose this species. The distal end of a humerus was also recovered from the scats. On the basis of size, this bone was identified as Microgale talazaci Major, 1896a. In a recent study of the small mammals of the Montagne d'Ambre National Park, Raxworthy & Nussbaum (1994) listed M. melanorrhachis Morrison-Scott, 1948 and M. talazaci, as well as several other species in this genus, as occurring within the local forest. M. melanorrhachis was found between 1125 and 1250 m and falazaci between 660 and 1250 m. MacPhee (1987) treats M. melanorrhachis as a synonym of M. cowani, however Nicoll & Rathbun (1990), Raxworthy & Nussbaum (1994), Stephenson (1995) and Jenkins et al. (1996) consider these taxa as distinct species. Little is known about the food habits of Cryptoprocta. Albignac (1973) noted that in the wild this carnivore preys upon insectivores, particularly Tenrec ecaudatus (Schreber, 1777). The presence of the M. melanorrhachis remains in the scats, an animal weighing on average 12.7 g (range 10.5—15.0 g, n = 14), shows that this carnivore consumes small prey. Barn Owl (Tyto alba) pellets. — Two separate collections of fresh barn owl pellets (n = 18) obtained on the Central High Plateau near the capital city of Antananarivo (1300 m; 18 °55’S, 47°3]’E) contained the bone remains of Microgale pusilla Major, 1896b; a shrew tenrec weighing on average 3.0 g (range 2.6—3.9, n = 12). The first collection, which consisted of a minimum of two M. pusilla, was obtained in February 1993 on the outskirts of Antananarivo in the district of Mahazoarivo © (18 °56’S, 47°33’E) below a roost. The local habitat is mostly human habitation surrounded by a patchwork of small rice paddies, dense eucalyptus and pine planta- tions, and open fields. A collection obtained from this site in January— February 1992 has already been reported on by Goodman & Langrand (1993), but did not contain Microgale remains. A second barn owl pellet collection, including a minimum of three M. pusilla, was obtained in mid-July 1992 near Ilafy (18951'S, 47°34’E), at 1350 m, 13 km NE of Antananarivo, in an area of anthropogenic savanna with rice paddies in valley bottoms and eucalyptus and pine plantations and a few houses on higher ground. The only local natural vegetation consists of a few Dracaena and Ficus trees. The savanna grassland is burnt once or twice a year for grazing purposes. All of these remains were identified as M. pusilla on the basis of size, the shape of the talonid of m3, and tooth socket structure. Nothing is known about the hunting range of Tyto alba on Madagascar or for that matter on the African continent (Fry et al. 1988). In Europe it probably does the majority of hunting within the breeding territory, which varies from 0.4—2.5 km? depending on food availability (Cramp 1985). Thus, it is assumed that the prey found in the remains reported herein were taken in the immediate vicinity of the roost sites. MacPhee (1987) noted that the majority of M pusilla museum specimens came from the eastern rainforest, but there are several “anomalous” records of this animal elsewhere on the island. These include remains found in owl pellets collected along the Mahafaly Plateau (100—200 m), an area in southwestern Madagascar with sub- Records of Microgale in vertebrate food remains 137 arid thorn scrub and no permanent marshes (MacPhee 1986, 1987); at Antsiforakely (1600 m), a locality on the Central High Plateau with little remaining natural forest, but relatively extensive wetlands (MacPhee 1987); and near Antsirabe (1500 m), also on the Central High Plateau, surrounded by little natural habitat except for marsh- land (Major 1897). On the basis of this information, M. pusilla occurs in a wide variety of habitats, including non-forested areas and heavily modified agricultural zones. This same pattern has also been found for other shrew tenrecs, namely M. cowani, M. dobsoni Thomas, 1884, and M. brevicaudata Grandidier, 1899 (Kaudern 1918, MacPhee 1987). In deforested zones of the Central High Plateau several Microgale species, which were previously thought to be forest dependent, persist in areas with extensive anthropogenic habitat modification. Presumably populations of these species remain in marshlands or small vestige patches of forests. Further, M. cowani, M. dobsoni, and M. pusilla are known from some of the remaining forests on the Central High Plateau, which are fragmented (Stephenson et al. 1994). These Micro- gale species cannot be used as “biological indicators” of undisturbed habitat. Other samples of barn owl pellets collected in heavily modified areas of the eastern rain- forest and Central High Plateau have not yielded Microgale remains (Goodman & Langrand 1993, Goodman et al. 1993), and thus, there is either considerable seasonal or individual variation in the food habits of this owl, or remnant populations of Microgale species are exceptionally patchy in distribution. Acknowledgements We are grateful to Dr. C. Raxworthy and Mr. and Mrs. R. Albignac for collecting some of the food remains reported in this note. Zusammenfassung Die Autoren berichten tiber Funde von Tenreciden der Gattung Microgale in Karnivorenkot und in Eulengewóllen. Schádelreste von mindestens zwei Individuen von Microgale melanorr- hachis und ein Humerus von M. talazaci wurden in Kotproben von Cryptoprocta ferox identi- fiziert, die in Nord-Madagaskar in der Montagne d'Ambre gesammelt wurden. Reste von M. pusilla fanden sich in Gewóllen von Tyto alba, die auf einem degradierten zentralen Hoch- plateau nahe Antananarivo gesammelt wurden. Die Funde von M. pusilla deuten darauf hin, daß diese Art nicht besonders empfindlich auf Umweltveränderungen reagiert. Resume La mention de deux especes de Microgale, dont la présence a été mise en Evidence par l’analyse des feces ou des pelotes de rejection de prédateurs, fait ’objet d'une revue. Des fragments de cräne d'au moins deux individus de Microgale melanorrhachis, ainsi qu’un humérus de Micro- gale talazaci ont été identifiés a partir de feces de Cryptoprocta ferox collectes ala Montagne d’Ambre, dans le nord de Madagascar. Des restes de M. pusilla ont été trouves dans des pelotes de Tyto alba collectées dans un milieu fortement anthropique, localisé sur les Hauts Plateaux pres d'Antananarivo. Ces dernieres données suggérent que M. pusilla n'est pas particuliere- ment sensible aux changements environnementaux. References Albignac, R. (1973): Faune de Madagascar. 36. Mammiferes: Carnivores. — ORSTOM/ CNRS, Paris. Bennett, E. T. (1833): Notice of a new genus of viverridous mammalia from Madagascar, presented by C. Telfair, Esq. — Proc. Zool. Soc. London 1833: 46. Cramp, S., ed. (1985): The birds of the western Palearctic. Vol. 4. — Oxford University Press, Oxford. 138 S. M. Goodman et al. Eisenberg, J. F & E. Gould (1970): The tenrecs: A study of mammalian behavior and evolution. — Smiths. Contrib. Zool. 27: 1—137. Fry, C. H.,S. Keith & E. K. Urban (1988): The birds of Africa. Vol. 3. — Academic Press, London. Goodman,S. M. £ O. Langrand (1993): Food habits of the Barn Owl (Tyto alba) and the Madagascar Long-eared Owl (Asio madagascariensis) on Madagascar: Adaptation to a changing environment. — Proc. 8th Pan-African Ornithological Congress, pp. 147—153. Goodman,S.M.,O.Langrand&C. J. Raxworthy (1993): The food habits of the Barn Owl Tyto alba at three sites on Madagascar. — Ostrich 64: 160—171. Grandidier, G. (1899): Description d'une nouvelle espece d’Insectivore provenant de Mada- gascar. — Bull. Mus. Hist. nat. Paris 5: 349. Hutterer, R. (1993): Order Insectivora. — In: Wilson, D. E. & D. M. Reeder, eds.: Mammal species of the world: A taxonomic and geographic reference, 2nd ed., 69—130. Smith- sonian Institution Press, Washington, D.C. Jenkins, P. D., Ss. M. Goodman &C. J. Raxworthy (1996): The shrew tenrecs (Micro- gale) (Insectivora: Tenrecidae) of the Réserve Naturelle Intégrale d'Andringitra, Madagas- car. — Pp. 191—217 in: Goodman, S. M., ed.: A floral and faunal inventory of the eastern slopes of the Réserve Naturelle Intégrale d'Andringitra, Madagascar: with reference to elevational variation. Fieldiana: Zoology, N.S. 85: 1—319. Kaudern, W. (1918): Quartäre Fossilien aus Madagascar. — Zool. Jb. 41: 521—533. MacPhee, R. D. E. (1986): Environment, extinction, and Holocene vertebrate localities in southern Madagascar. — National Geographic Research 2: 441—455. MacPhee, R. D. E. (1987): The shrew tenrecs of Madagascar: Systematic revision and Holo- cene distribution of Microgale (Tenrecidae, Insectivora). — Amer. Mus. Novit. 2889: 145; Major, C. I. Forsyth (1896a): Diagnoses of new mammals from Madagascar. — Ann. Mag. nat-Elist., ser.0,18:318- 325} Major, C. I. Forsyth (1896b): Descriptions of four additional new mammals from Madagas- car. — Ann. Mag. nat. Hist., ser. 6, 18: 461—463. Major, C. I. Forsyth (1897): On the general results of a zoological expedition to Madagascar in 1894-96. — Proc. Zool. Soc. London 1896: 971—981. Morrison-Scott, T. C. S. (1948): The insectivorous genera Microgale and Nesogale (Madagascar). — Proc. Zool. Soc. London 118: 817—822. Nicoll, M. E. & O. Langrand (1989): Madagascar: Revue de la conservation et des aires protégées. — WWE, Gland. Nicoll, M. E. % G. B. Rathbun (1990): African Insectivora and elephant-shrews: An action plan for their conservation. — IUCN, Gland. Rasolonandrasana, B. P. N. (1994): Contribution a l'étude de I’ alimentation de Crypto- procta ferox Bennett (1833) dans son milieu naturel. — D.E.A. en Anthropologie, Service de Paléontologie, Université d'Antananarivo. Raxworthy, C. J.& R. A. Nussbaum (1994): A rainforest survey of amphibians, reptiles and small mammals at Montagne d'Ambre, Madagascar. — Biol. Conserv. 69: 65—73. Stephenson, P. J. (1995): Taxonomy of shrew-tenrecs (Microgale spp.) from eastern and central Madagascar. — J. Zool., London 235: 339—350. Stephenson, P. J, H. Randriamahazo, N. Rakotoarison & P. A. Racey (1994): Conservation of mammalian species diversity in Ambohitantely Special Reserve, Madagas- car. — Biol. Conserv. 69: 213—218. Thomas, M. R. O. (1882): Description of a new genus and two new species of Insectivora from eer — J. Linn. Soc. 16: 319—322, Thomas, R. O. (1884): Description of a new species of Microgale. — Ann. Mag. nat. Hist., ser. oe 14: 337.338. Steven M. Goodman, Field Museum of Natural History, Roosevelt Road at Lake Shore Drive, Chicago, IL 60605, USA & WWE, Aires Protégées, B.P. 738, Antanana- rivo (101), Madagascar. — Paulina D. Jenkins, The Natural History Museum, Crom- well Road, London SW7 5BD, England. — Olivier Langrand, WWE, Aires Proté- gées, B.P. 738, Antananarivo (101), Madagascar. Bd. 4 S. 139-163 | Bonn, September 1997 Systematics and biogeography of the southern African scincine genus Typhlacontias (Reptilia: Scincidae) Wulf D. Haacke Abstract. Typhlacontias is one of several fossorial skink genera of southern Africa, which are limbless or have degenerate limbs. It occurs in parts of the Namib Desert and the northern Kalahari Basin. In this generic revision six species, one with three subspecies, are recognised. The confused description of T. punctatissimus (Bocage, 1895) which has previously been addressed by Andersson (1916), is further discussed. 7. bogerti Laurent is relegated to subspecies level of above species and 7: p. brainei is described as a new and third subspecies. An apparently new species is recognised as one which was described as T. johnsonii in 1916 already. 7. rudebecki is described as a new species and 7. gracilis Roux, T. rohani Angel, as well as T brevipes FitzSimons, are each treated as full species. Key words. Reptilia, Scincidae, Typhlacontias, new species and subspecies, systematics, biogeography, southern Africa, Namib desert, Kalahari. Introduction The genus Typhlacontias was correctly identified as new by Bocage (1873) and only De Witte and Laurent (1943) attempted to split it when they proposed the genus Fitzsimonsia for T. brevipes. As the original syntypes of the type species 7? punctatis- simus were poorly preserved subadults, Bocage (1895) felt that the species description had to be amended when he received an assumed adult topotype in 1884. However, this was an aberrant specimen of a new species, later described as 7. johnsonii (Andersson, 1916), and its supernumerary supranasal scales were considered species specific (Bocage 1895, pl.VII, Fig. 3; De Witte & Laurent 1943, Fig. 53). As a result T. punctatissimus became unidentifiable. Andersson (1916), recognising this confu- sion when he was supplied with seven additional specimens representing both species from the type locality, rediscribed 7. punctatissimus on the basis of a single juvenile and utilised the other six specimens to describe a new species, 7. johnsonii, in a paper which was overlooked ever since its publication. A point of controversy is the presence of leg rudiments in this 7. punctatissimus neonate (HB = 41 mm), in a supposedly limbless species (Bocage 1873; Boulenger 1887). To complicate matters further, Bocage's (1895) three original specimens were destroyed in a fire in the sixties, shortly after Dr D. G. Broadley was able to re-examine the third specimen (= T. johnsonii) and confirm the accuracy of Bocage's (1895) description (Broadley, pers. com.). In 1907 Roux described 7. gracilis from a single specimen from Barotseland in Zambia, the former Northern Rhodesia (vide “Rhodésie” in De Witte & Laurent 1943: 35). The drawings of the head scales (Roux 1907: 84, fig. 3 and 4) are not very accurate, suggesting the presence of only four upper and three lower labials and the parietal area in the dorsal view is inaccurate. In addition the type has a slightly 140 W. Haacke aberrant supraorbital area as the upper anterior temporal scale apparently fused with the third supraorbital (Fig. 4b). Another new species, 7. rohani (Angel 1924), was described from the Kalahari sand of south-eastern Angola, but was overlooked by FitzSimons (1932) when describing 7! ngamiensis. As a consequence the latter has been relegated to a junior synonym (Auerbach 1987). Unfortunately the origin of the type of 7! brevipes was not known when FitzSimons (1939) described this species. The assumption that it came from Angola resulted in the omission of this species from his classic study “The Lizards of South Africa” in 1943. McLachlan (1984: 17) eventually restricted the type locality to Walvis Bay. After the second World War a revival of the interest in faunal studies of southern Africa took place and possibly an improvement in funding provided a new impetus for exploring and collecting. Fresh material was produced from the central and northern Namib when Dr Charles Koch and Dr C. K. Brain of the Transvaal Museum and Dr G. Rudebeck of the Lund University Expedition visited those areas and 7. bogerti was described from south western Angola by Dr R. F. Laurent in 1964. The author of this paper then started collecting during the sixties, intensively sampling the Namib Desert, as well as the Kalahari in north-eastern Namibia and Botswana. Dr D. G. Broadley in Rhodesia (now Zimbabwe) also entered the arena at that time and accumulated long series with the help of enthusiastic amateurs like Richard Japp in western Zambia and the Caprivi Strip of Namibia. Distribution gaps in south-western Zambia and south-eastern Angola still cast doubt at some of the conclusions drawn in this paper. A special problem is provided by the single, mutilated type specimen of 7. rude- becki sp. n., which does not quite fit into the biogeographic picture of Typhlacontias as produced by the other taxa. However it is hoped that this study is a step in the right direction. Genus Typhlacontias Bocage Typhlacontias Bocage, 1873. Jorn. Sci. Lisboa 15: 213 (Type species: Typhlacontias punctatissimus). Boulenger 1887: 429; Bocage 1895: 56; Andersson 1916: 18; De Witte & Laurent 1943: 34; Greer 1970: 14; Welch 1982: 79. A small genus of mostly apodal, fossorial skinks. The snout is flattened, protruding and covered by a large rostral scale. Nostrils pierced in the sides of the rostral, with a suture extending backwards to the edge of the rostral. A small to minute oblong postnasal scale is enclosed in this rostral groove. Three transversally enlarged azygous shields, i.e. internasal, frontonasal and frontal, cover the head between the rostral and the large subtriangular inter- parietal, in which a pineal eye is usually clearly visible. Eyes, small and completely exposed to partly covered, without eyelids, in contact with specific combinations of the second and third upper labials. Three pairs of temporal scales dorsally to the three posterior upper labials. No external ear opening. Body scales smooth, imbricate and in 18 rows around body. Tail length varies according to species from two thirds to one quarter of the body length. No preanal or femoral pores, sexes indistinguishable externally. The skull was discussed by Greer (1970: 14). Pectoral girdle strongly degenerate with no visible remnants of limbs, while the remains of the pelvic girdle may still have rudiments of the femurs. External leg rudiments consistently present in one species and occasionally so in another. Small conical teeth, palate toothless. Tongue squamous, anteriorly notched to slightly bifurcate. Active sand swimmers patrolling leaf litter and decaying vegetable matter for termites, beetle larvae and other small invertebrate prey. Mostly active at night when soil temperatures allow near surface activity. Biology practically unknown but some species are Ovoviviparous. Revision of the scincine genus Typhlacontias 141 Range: The Namib Desert sands from north of Lüderitz into south western Angola and Kalahari sand of north-eastern Namibia, northern Botswana, western Zimbabwe and western Zambia. Although only one record exists from south-eastern Angola it is highly likely that these lizards are of common occurrence there. Remarks: The original generic description was based on two poorly preserved syntypes (Bocage 1873) collected by the well-known naturalist J. O. Anchieta at the Curoca River mouth in the northern Namib Desert in Angola which were donated in 1867. A third, supposedly topotype, was added later and was used to expand the species description (Bocage 1895). It is now clear that the latter was an aberrant individual of 7. johnsonii Andersson, 1916, a then still undescribed species, which caused confusion. All three specimens were destroyed by a fire in the Bocage Museum at the University of Lisbon during the seventies (D. G. Broadley, pers. com.). 7: punctatissimus remains the type species and its original description (Bocage 1873; Boulenger 1887) was reviewed and updated by Andersson (1916: 21). Unfortunately the speci- men, which the latter author considered to represent this species, is also aberrant as it has hindlimb rudiments. Material For this revision 616 specimens have been examined, including the types of 7. brevipes, T. gra- cilis, T. ngamiensis, T. rohani and two syntypes of 7. johnsonii. To save printed space only the specimens of new taxa and lectotypes are individually listed with catalogue number, while for every locality the institution in which a relevant voucher specimen is kept is listed. The following institutional abbreviations are used: CAS = California Academy of Science, San Francisco. EMUS = John Ellerman Museum, University of Stellenbosch. FMNH = Field Museum of Nat. Hist., Chicago. GMS = Gothenburg Museum, Sweden. IRB = Institut Royal des Sciences Naturelles, Brussels. LACM = Los Angeles County Museum, Los Angeles. MCZ = Museum of Comparative Zoology, Boston. MHNP = Museum of Nat. Hist., Paris. NDRS = Namib Desert Research Station, Gobabeb, Namibia. NMZ = National Museum, Bulawayo, Zimbabwe. SAM = South African Museum, Cape Town. SMF = Senckenberg Museum, Frankfurt on Main. TM = Transvaal Museum, Pretoria. ZFMK = Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn. Key to the genus Typhlacontias As tonciy enlareed second lower labial present. .2...2...... nm... 2 id) Ko nlonsedWlowerslabialsersn ee N NER ae 7 Paancumentawyahındlimbs-presentn..% „ann. ee ee le Sek ee OS 3 y NO Maia O N ee et 4 3a) Minute postnasal scale in short contact with 1st upper labial, S © E UE ns ar ae ER Er ER T. brevipes b) Minute postnasal scale enclosed within rostral suture, PRO RUST SARA VEA ASIN eth diy salve Tae Re a (T. punctatissimus ssp.) 4a) Frontal shield subhexagonal to oval shaped, tail about one aer O a nen ee T. johnsonii b) Frontal nearly halfmoon shaped with a strongly curved anterior edge and a nearly straight posterior margin. Original tail about two thirds head/body length ........ (7? punctatissimus sspp.) 5 5a) Five upper labials with second usually reaching eye .................. T. p. brainei b) Six upper labials with second plus third or third only PERCE Var. boo Bice sona o OO IE RLSM cS EA 2 te OR a 6 Ga) Usually only third upper labial reaching eye .................. T. p. punctatissimus b) Usually second and third upper labial reaching eye, SAO A I o N A T. p. bogerti dr upper labialenteridgreye ii ds de ee an T. gracilis PIS ceondeupperslabiallenterinsaeyer 1. A IA A 8 Sa orca lin contact with second upper labial Ms 205 0. oe ee T. rohani D)akorealenot inzeontaet with second upper labial 773. 222222222. T. rudebecki sp. n. 142 W. Haacke Systematic account Typhlacontias johnsonü Andersson (Figs la, 2a, 7) Typhlacontias johnsonii Andersson, 1916. Meddel. Góteborgs Mus. Zool. Afdel. 9: 19. Lectotype Re. ex. 1388a, Lectoparatypes Re. ex. 1388b-f in Gothenburg Museum of Natural History (Type locality: Porto Alexander = Tombua, coll. by D. F. d'Azinhaes, 3 July 1912, don. by H. Skoog). Typhlacontias punctatissimus part. Bocage 1895: 56. Roux 1924: 164. Monard 1937: 86. De Witte & Laurent 1943: 34, fig. 53. Material: 83 specimens examined. Types in Gothenburg Natural History Museum. Two lectoparatypes seen. Type series: The description is based on six syntypes, all recorded under No. 1388 in the Zoological Department of the Gothenburg Museum (now ‘Naturhistoriska Museet, Góte- borg’, Sweden). The largest and first mentioned specimen (HB = 123 mm) of this series, of which the measurements are listed (Andersson 1916: 20) is hereby designated as lectotype (1388a), while the remaining smaller individuals are lectoparatypes (1388b-f). Etymology: Named in honour of Mr. C. O. Johnson, who financed Mr. Skoog’s expedition to Angola. Diagnosis: Tail shorter than in all other species, i.e. about 25 % of SV length. Eye smaller than in other species and partly overlain by edges of loreal and supraocular. Second upper labial about twice as long as tall and in contact with eye. Fig. 1: a) Typhlacontias johnsonii Andersson, TM 63455, Ondondujengo River, Namibia; b) Typhlacontias brevipes FitzSimons, TM 31416, Unjab River, Namibia. Postoculars fused with temporals. (For abbreviations see Fig. 5). Revision of the scincine genus Typhlacontias 143 Fig. 2: a) Typhlacontias johnsonii Andersson, TM 63455. Data as above; b) Typhlacontias punctatissimus bogerti Laurent, NDRS-R 223, Curoca River crossing, Angola. Dorsal view to illustrate the difference in the shape of the frontal shield. Description: Body relatively stout, underside of body flattened. Head wedge-shaped, snout depressed, rostral forming sharp horizontal edge. Relative size of laterally deeply notched rostral less prominent than in other species. Nostrils connected to rear edge of rostral by short suture, enclosing small oblong postnasal scale. Rostral followed by bandlike inter- nasal and frontonasal, a suboval to hexagonal frontal, a large subpentagonal interparietal flanked by pair of striplike parietals which touch behind interparietal (Fig. 2a). Pineal eye in interparietal usually visible. Five upper labials, first largest, second very long and below eye, probably consisting of the fused second and third. Eye exposed but small, with a nearly straight posterior edge, partly covered by second upper labial and the single, large, triangular supraocular, resulting in a subtriangular appearance (Fig. la). Eye also in touch with loreal (= preocular according to Andersson 1916), occasionally a tiny single preocular, the upper anterior temporal which may have fused with the upper postocular and a tiny lower postocu- lar, which may be missing. Frontoparietal in contact with upper anterior temporal. Large mental followed by large postmental, four lower labials of which the second is very large. Head scales show a certain amount of individual variation, s.a. subdivisions of dorsals (eg. fron- tonasal in TM 63344 and supernumerary postnasals in original specimen from Curoca River [Bocage 1895: 56]), minute preoculars may be present or partly obscured by loreal, labials may be split in two or scales may fuse. Row of four preanal scales, subequal in size to the ventrals. Eighteen rows of tight-fitting smooth scales covering body. Cleared specimen TM 63463 (Ondondujengo River) has 72 spinal vertebrae, remnants of pectoral and pelvic girdles present but no limb rudiments. Nentralsakrance 131150, n 735 x — 139.2) SDE 5.1. Subeaud.: Range =35—42, n= SO), =D Sa Ay Size: Max SVL = 117 mm (TM 63461, Ondondujengo River). Max total = 139 (110 + 29) mm (TM 32569 56 km S of Orupembe). Min SVL = 39 mm (TM 63281 Hartmann’s Valley). Min total =51 (41 + 10) mm (TM 63373 Hartmann’s Valley). Relat. tail length: RT 2333870) fh — 53, X= 26.8 90, SD =] 2.4: Colouration: Juveniles are pinkish pale yellow with only small sark spots on frontal and interparietal, an indication of a vertebral double line and a lateral line extends backwards through the eyes. The light dorsal areas on the snout may be light brown to golden yellow, with the tail pale blue. With maturity each scale’s centre spot enlarges, black lines become more prominent and old individuals may become very dark. Apart from the unmarked lower jaw area every body and tail scale has a central dark spot. Size of spots variable and rows of 144 W. Haacke large spots create a dorsal band (2 rows), 1 dorsolateral line, a broad (3 rows) lateral band and often a black ventrum (4 rows). Dorsal band joins dark brown marks on top of head, lateral bands continue through eyes and join on snout. Range: (Fig. 7) Wind blown sand of the Northern Namib Desert from the mouth of the Curoca River on the northern edge of the Curoca sand-sea in south-western Angola, south- wards into the Kunene sand-sea in north-western Namibia. Documented as far south as the Sechomib River valley. In Namibia it avoids the coastal hummocks of the seaward side of the sand-sea where 7: brevipes occurs. Although poorly documented in Angola it apparently occurs closer to the coast there. 7. brevipes, which is of similar size and would therefore probably be in direct competition, has not been recorded from north of the Kunene River. On the inland edge of the Kunene sand-sea, it is usually sympatric with 7. punctatissimus ssp., but only rarely with 7. brevipes when the latter species extends that far inland along drainage valleys. Distribution records: ANGOLA: Porto Alexander (= Tombua) 1511Dd (GMS); Curoca River (nr mouth) 1511Dd (Bocage 1895: 56); Lacrau 1711Bb (TM). NAMIBIA: Okotusu area 1711Bd, E edge of Kunene sand-sea 1712Ac and 1711Db, edge of Hartmann’s Valley 1712Ac, 48 km NW of Ouhandjo 1712Ac (TM); Ondondujengo River at Hartman Valley — Cape Fria road crossing 1712 Cc (SMWN, TM, ZMFK); Sechomib River valley 56 km SSW of Orupembe 1812Da (TM, ZFMK); nr end of Hartmann’s Valley 1712Aa (TM, SMWN); Hartmann’s Valley 1712Cb (TM). Field notes: Active during summer evenings mostly around debris accumulations at the base on the leeward side of dunes. Individuals move just below the sand surface, out of sight, leaving a clear regular undulating track. Occasionally these tracks move across the slipface to the crest of the dune, over the top or down again. When moving, although not actually exposed, the position of the specimen is quite clear under the surface at the blunt end of the track and collecting is easiest at that moment. If disturbed they dive into the sand and become inaccessible. As this material was mostly collected during autumn, no gravid females were found. However, a number of young (about 40 mm HB length) were collected during April 1985. It is assumed that this species is live bearing and the young are born during late summer i.e. February to March. Typhlacontias brevipes FitzSimons (Figs 1b, 6) Typhlacontias brevipes FitzSimons, 1939. Ann. Transv. Mus. 20: 15, figs 14—17. Type S.A.M. 508 in South African Museum, Cape Town. (Type locality: Cape Division. Restricted to Walvisch (= Walvis) Bay, Namibia by G. R. McLachlan 1984: 17). Greer 1970. Welch 1982: 79. Branch 1988: 124. Fitzsimonsia brevipes De Witte & Laurent 1943: 11. Haacke 1964: 8. Material: 141 specimens examined. Type seen. Etymology: Referring to the presence of limb rudiments. Diagnosis: Rostral large and laterally strongly notched with the small enclosed postnasal scale making contact with the first upper labial, third upper labial in contact with eye, second lower labial very large. Minute hindlimb rudiments visible on either side of cloaca. Description: Snout depressed and covered by a large rostral scale which extends beyond the lower jaw and forms a sharp horizontal edge, strongly notched laterally. Edges of nasal suture in rostral scale not touching as small postnasal scale is in contact with the first upper labial. The three bandlike scales i.e. internasal, frontonasal and frontal are subequal in width and breadth, large subpentagonal interparietal is posteriorly edged by a pair of striplike parietals, which slightly overlap. The pineal eye is usually clearly visible in the interparietal, but not so in the four most southern specimens from Groot Anigab (TM 44320-3). Six upper labials, first largest, third touching eye, sixth smallest. Eye small and round, bounded and partly overlain by the third upper labial, loreal, two supraoculars, the upper anterior temporal and a tiny lower postocular. Large mental and postmental, four lower labials with second very large. Body scales smooth, tight-fitting, imbricate in 18 rows. A minute rudimentary hindlimb, present on both sides of the vent. Cleared specimen TM 27450, Swakopmund, has 66 and TM Revision of the scincine genus Typhlacontias 145 56876, Hoanib Oasis, has 68 vertebrae, remnants of the pectoral and pelvic girdles, with femur rudiments. Wentrals Range 1835152) 1m — 95, x= 1403, SD’ == 4:6, Subcaud:: Range = 4556, ne 3% EDAD Size: Max SVL = 133 mm (TM 36886 Walvis Bay). Max total = 158 (125 + 33 dam.) mm (TM 42252 10 km SE of Torra Bay). Min SVL = 39 mm (TM 63349 Ondondujengo River). Min total =54 (39+15) mm (TM 63349 Ondondujengo River). Rel. tail length: Raneer 5) 4400.39) x — 37.6%, SDi= ae 27. Variation: A considerable amount of individual variation occurs amongst the head scales, especially the smaller ones around the eyes, which may be absent or fuse with adjacent scales. Fusions may be symetrical or not. For example, TM 31529 (Munutum River) has both loreals fused to the internasal forming one wide halfmoon-shaped band. Amongst the specimens from just south of the mouth of the Kunene River, the size of the second upper labial, which is usually in touch with the loreal, is reduced to such an extent that the first and third upper labials are in contact above it. The development of the limb rudiments is fairly consistent. The total size, even in big adults, tends to be less than a millimetre in length, although the shape may vary from blunt triangular to slightly elongated. Colouration: Juveniles have well developed dark lateral bands at an early age, a faint dorsal double line which starts on the head, down the yellow-grey back. With increasing size every scale develops a dark blotch to an extent that the specimen is more or less lined but then gets evenly speckled as the dark lines become less distinct. The background colour of back and body may be silver grey to pale straw yellow with the tail pale powder blue. Individuals from Gobabeb on the Kuiseb river are much lighter than the coastal specimens. Most of these specimens have an unmarked light sulphur yellow body with only a few vertebral rows of dark spots. The tails are darker marked with rows of dark specks on a blue-grey background. Range (Fig. 6): Coastal Namib Desert from just north of Lüderitz to the Swakop and then the Omaruru River. No records are known from the Omaruru River to the southern tip of the northern sand-sea, where the road to Khorixas turns inland. Northwards from that point this species is quite common in vegetation islands and hummocks on the seaward side of the large sandbodies as far as the mouth of the Kunene River. It may extend inland along the drainage valleys of the dry rivers. Distributionrecords: NAMIBIA: Cape Fria 1812Ac (TM); Cape Fria or Karlowa’s Hut 1812Aa (SMWN, TM); ‘Dunedin Star’ Wreck 1811Bb (TM); Gobabeb 2315Ca (CAS, MCZ, SMWN, TM, ZFMK); Great Anigab Waterhole 2615Ac (TM); Hoanib River flood plain 1912Bd (SMWN); Hoanib River mouth 1912Bd, Hoanib/Mudorib River confluence 1913Ad, Hoanib River at 1913Ac (TM); Hoarusib River mouth 1912Ba (SMWN); 18 km NW of Hunkab River crossing 1913Ca (TM); Khumib River mouth 1812Cd (SMWN); Kuiseb River dunes at Rooibank 2314Ba (SMF); S of Kunene River 8 km from mouth 1711Bd, Möwe Bay area 1912Bc (SMWN, TM); Munutum River 16 km from mouth 1812Ac (SMWN); Munutum River 40 km from Cape Fria 1812Aa, ‘Oasis’ in Hoanib River 1912 Bc (TM); Omaruru River 16 km from mouth 2114Cd (SMWN, TM); Ondondujengo River 1712Cc (TM); Rocky Point 1812Cd (SMWN); Samanab River 2013Ab, Sandwich Harbour 2314Ad (TM); Sarusas in Khumib River 1812Cd, Sechomib River 1812Ad (SMWN, TM); Swakopmund 2214Da (LACM, TM); Swakop River 2214Da (TM, CAS); Terrace Bay 1913Cc (TM); Torra Bay 2013 Ac (SMWN); 10km and 22km SE of Torra Bay 2013Ad, Unjab River at 2013Aa, Ab (SMWN, TM); 8 km E of Walvis Bay 2214Dc, 10 km S of Walvis Bay 2315Ba (TM). Field notes: This species avoids large bare sand bodies and is usually found in the semistable sand of hummocks and sparsely vegetated dunes, as well as the edges and sandy vegetated banks of dry rivers. In general it does not share its habitat with other fossorial lizards. At Gobabeb, the furthest inland record, 7: brevipes occurs along the banks of the Kuiseb River and although it is sympatric with Typhlosaurus braini there, the latter avoids direct compe- tition by utilising the occasional hummocks on high dunes. However, near Groot Anigab, the currently documented southern range limit of both species, these two species were actually found in the same hummock. South of Torra Bay open valleys are crossed on shifting, barchan dunes which also carry Meroles (Aporosaura) anchietae and Bitis peringueyi. Only when 146 W. Haacke T. brevipes penetrates the northern sandseas along riverbeds towards the eastern side of the dunefields, has it been found to be sympatric with other legless lizards, such as T. johnsonii and 7. punctatissimus. This species, like other congenerics, tends to feed and be active just below the sand surface at night and during the twilight hours when the sand is of a suitable temperature. They leave their undulating tracks all over the windblown sand of dunes and hummocks, feeding on a variety of small insects, their larvae and occasional termites. Although dune slipfaces and debris accumulations are patrolled and frequented, 7! brevipes appears to prefer vegetated hummocks and their leaf litter cushions. Most small juveniles (about 40 mm HB length), which appear to be neonates, were found during March to April. TM 44320, a large female (HB = 110 mm) from Groot Anigab, collected during January contained three well developed foetuses, which suggests that they were probably ready to be born during the latter part of summer, i.e. February to March. Typhlacontias punctatissimus Bocage The original description was based on a pair of poorly preserved subadult syntypes (Bocage 1873). Although referring to them as “limbless”, their state of preservation may have caused the author to overlook some of the finer details. Possibly due to a typographical error the presence of a subocular instead of a supraocular was described (Boulenger 1887). The species description was extended (Bocage 1895: 56) when an apparent adult topotype became avail- able. An important aspect of this specimen was the presence of a pair of small supranasals, now recognized as an individual variation, as it has not been recorded again. This fact was considered as species specific and was used by Monard (1937: 84) and De Witte & Laurent (1943: 34) to differentiate 7. punctatissimus from other species. Due to this compound description 7. punctatissimus became unidentifiable. These three specimens were destroyed by fire in the Museo de Bocage at the University of Lisbon (acc. Prof. Saccarao, D. G. Broadley pers. com.) in the seventies, just after Broadley had the opportunity to re-examine the third specimen and confirm the correctness of its description (Broadley, pers. com.). Although the original syntypes have never been re-examined because of their poor condition, the description - (Bocage 1873) contains a number of points, which are significant for the recognition of this species. The shape of the frontal scale was described as being “large and almost semicircular with the anterior edge rounded and posterior edge straight”. This shape is typical for 7. punca- tissimus (Fig. 2) and differentiates it from the assumed adult topotype which was actually a different species. The aspect providing absolute proof that another species was involved are the body ratios. The relative tail length of the longer of the former syntypes was equal to 51 % of HB length (Bocage 1873: 213), which is typical for 7: punctatissimus (Table 1), whereas in the third or “adult” specimen the tail was only 26 % of the body length, which is diagnostic for T. johnsonii (Bocage 1895: Fig VII, 3+3a. Table 1). Seven new specimens from the type locality, donated by H. Skoog in 1912, were examined by Andersson (1916) and it was obvious to him that two species were involved. According to a single neonate (Re. ex. 1389, HB = 42 mm) he confirmed the original description of 7. punctatissimus (Bocage, 1873) and redescribed Table 1: Variation in size, body proportions and scalation. ER ss max % T/SVL Ventrals Subcaudals Vere SVL T. brevipes 131—152 T. johnsonii 131—150 12. T. p. bogerti 114—135 62 T. p. punctatissimus 116—137 61 T. p. brainei 117—128 — T. gracilis 122-147 E 68 T. rohani 124—146 62,63,66 T. rudebecki 157 — Revision of the scincine genus Typhlacontias 147 it. He added to the controversy by recording the presence of tiny leg rudiments in this specimen, which has now been made available for re-examination and the presence of these limb rudiments is herewith confirmed. Only a single additional specimen with tiny hindlimb rudiments (R 224 Namib Desert Research Station, Gobabeb, from eastern edge of the Curoca Sand Sea, S of Curoca River near road crossing into lona National Park, 1612Ad, Angola) has been traced amongst the 94 specimens of this species that have been examined for this revision. One of two cleared specimens has internal femur rudiments (TM 40727, Kakolo windmill 1612Cd , lona Nat.Park, Angola), while the second (TM 63481) from the Ondondu- jengo River (1712Cc) in Namibia has no indication of any rudiments at all. According to other external morphological characteristics the specimens from the Curoca River mouth and the Ondondujengo River represent the typical subspecies, 7. p. punctatissimus, while the other two specimens mentioned represent 7: p. bogerti. Under normal circumstances leg rudiments carry substantial weight in species recognition of fossorial skinks. At present not enough material is available from the type locality in the vicinity of the Curoca River to evaluate the taxonomic significance in this case and it is for the time being accepted as atavistic intraspecific or individual variation. Unfortunately Andersson’s (1916) paper has been overlooked since then in all studies dealing with Typhlacontias. T. bogerti Laurent (1964) is only subspecifically distinct from T. punctatissimus and trinomens must thus be used. In addition a third subspecies can be defined and is described below. Typhlacontias punctatissimus punctatissimus Bocage (new status) (Fig. 3b) Typhlacontias punctatissimus Bocage, 1873. Journ. Sci. Lisboa 4: 213 (Type locality: Rio Curoca mouth, Mocámedes district, Angola), 1887: 203 and 1895: 56 (Syntypes in ‘Museo de Bocage’, Lisbon, destroyed by fire). Boulenger 1887(3): 429. Andersson 1916: 21. Roux 1924: 164. Monard 1937: 86. De Witte « Laurent 1943: 34. Greer 1970: 14. Welch 1982: 79. Typhlacontias bogerti, Haacke 1965: 22 (part. Namibian specimens). Mertens 1971: 70 (part. Namibian specimens). Material: 39 specimens examined. Syntypes destroyed, not seen. Etymology: Referring to its speckled colour pattern. Diagnosis: The most slender of those species characterised by the enlarged second lower labial and the laterally strongly notched rostral, s. a. 7! brevipes and T: johnsonii. Large frontal, nearly semicircular with straight posterior edge. In this subspecies usually only the third upper labial enters eye. Rudimentary hindlimbs may occur. Description: The original description applies. Snout depressed, large rostral scale with horizontal edge, less strongly notched than in the above mentioned taxa. Nasal groove straight, enclosing minute postnasal scale. Rostral followed by a narrow striplike internasal, a slightly Table 2: Variation in head scalation. : Enlarged | Upper y Supra | Prefront | PF Loreal in ppecies 2nd L.L | labials below oculars | present |”! contact| contact e with FP | 2nd UL T. brevipes T. johnsonii T. p. bogerti T. p. punctatissimus T. p. brainei T. gracilis T. rohani T. rudebecki L ye 3 2) 3 2 3 2 N N O IN ON ION NN NNNN WY 2 and 3 2 148 W. Haacke Fig. 3: a) Typhlacontias punctatissimus bogerti Laurent, topotype TM 40391, 34 km S of Namibé (= Mocämedes), Angola, with supernumerary third upper postocular; b) 7! p. punc- tatissimus Bocage, TM 63215, Munutum River crossing, Namibia; c) 7. p. brainei ssp. n., holo- type TM 68443, Khumib River, Namibia. : wider prefrontal and a fairly large semicircular frontal scale, which has a strongly convex anterior and nearly straight posterior edge (Fig. 2b). Eye round, completely exposed and bordered by one preocular, loreal, the anterior supraocular, two postoculars and the third upper labial, separated from second supraocular by upper postocular. Six, not five upper labials, first largest, second small and curving backwards but separated from eye by tiny preocular. The supposed presence of a subocular reported by Bocage (1873) appears to be due to a typographical error referring to a supraocular. No noticeably enlarged triangular preanal scales could be confirmed. The tiny hindlimb rudiments specifically mentioned by Andersson (1916) for the juvenile topotype (Re. ex. 1389), which was re-examined, are very distinct but could not be traced in any other specimens, except for one from further upstream along the Curoca River. In that case (D.E.R.U. R-224) the limb rudiments are even less well developed. Cleared specimen TM 63481 (Ondondujengo River), has 61 vertebrae, remnants of pectoral and pelvic girdle, but no internal limb remnants. Ventrals: Range = 116137, x= 1227, SD= == 6.99, n= 27. Subeaud 2 Ranse- 50 270% 0102 34 SD. 421222997: | Size: Max SVL = 84 mm (TM 24304 Nangolo Flats). Max total = 115 (77 + 38) mm (TM 63215 Munutum River). Min SVL = 38 mm (TM 63478 Ondondujengo River). Min total = 74 Revision of the scincine genus Typhlacontias 149 (49 + 25) mm (SMWN 8196 Hartmann’s Valley). Rel. tail length: Range = 46—55 Yo, x= SIO UNS SD = ae (MAS Notes: The relative tail length of one of the original syntypes was 51 Y (HB = 55 mm, tail = 28 mm). This matches the mean for this taxon (Table 1), which also applies to the topo- type (SVL = 41 mm, tail = 20 mm, rel. tail length = 49 %) (Andersson 1916: 22). This ratio for the third specimen, presumed to be an adult topotype (Bocage 1895), was “measuring about a quarter of the length of the body,” i.e. 26.1 % (HB = 92 mm, tail = 24 mm) which is typical for 7. johnsonii (Table 1, Range 23—33 Yo, X = 26.8 Wo). As the two syntypes were destroyed by fire 1t is desirable to designate a lectotype. However as the Curoca River is the type locality for this taxon and at the same time appears to be the border between two subspecies, it seems important that any action is avoided which might further contribute to the confusion. It is therefore highly desirable that a topotype should be used for this purpose. At present the only specimen of this taxon from that site and from north of the Kunene River is the above mentioned juvenile with the atypical limb rudiments. Although this is viewed as being an intraspecific apparently atavistic individual variation, it may also represent a situation of taxonomic importance, for which reason this individual is unsuitable as a lectotype. As this problem can only be resolved once additional topotypes become available, it is abstained from designating a lectotype. Colouration: The basic patterns of this taxon are created by the black centre spots of each scale which may be more or less emphasized, thereby creating the impression of lines. The broad (4 scales) dorsal band may be unmarked and of a rich golden brown, but may vary via straw yellow to pinkish grey, in which case two vertebral lines of brownish specks occur. Laterally two well defined black lines start as one near the nostril, pass through the eye and extend to the tail tip. Additional lines of specks may be present ventrally and on the underside of the tail. Scales in cloacal area usually unmarked. On the silvery blue tail the two vertebral dark lines are more prominent than on the body. The top and sides of the head may or may not be marbled with brown, have a few dark spots or an arrow-headed brown mark. Variation: Within the range of 7: p. punctatissimus odd specimens occur with individual variations to the labial formula, such as asymmetrical fusions of labial 1 and 2 (IM 31525) and of 2 and 3 (TM 31525), symmetrical fusions of 1 and 2 (TM 63259) and of 2 and 3 (TM 57556), vertical subdivision of labials 2 and 5 in TM 71475 on the left side created eight upper labials and of these the fourth is in touch with the eye. These variations might produce problems when keying out individual specimens, but similar variations occur in the other two subspecies as well. A relatively significant variation is the rare presence of external rudimentary hindlimbs. Range (Fig. 8): Although the original type locality was the lower Curoca River on the edge of the Curoca Sand Sea, only one further specimen is known from there. At present the typical form is mostly known from south of the Kunene River in Namibia, where it occurs from the inland edge of the Kunene Sand Sea into the Marienfluss Valley and as far south as the Munu- tum River. It is not known from the immediate vicinity of the Namibian coast. Distributionrecords: ANGOLA: Curoca River nr mouth 1511Dd (Bocage 1873) and nr Porto Alexander (= Tumboa) 1511Dd, which is nr mouth of Curoca River (= Type locality). NAMIBIA: Hartmann's Valley, several records from close proximity to one another within the q.d.s. 1712Aa (SMWN, TM); 1712Ac (TM, ZMFK) and 1712Cb (TM); Khumib River (SMWN); Marienfluss 1712Da (SMWN, TM); Munutum River 40 km from Cape Frio 1812Ab (SMWN, TM); Muntum River at Skeleton Coast Park border 1812Aa (TM); Nadas River 32 km from Cape Frio 1812Ac (SMWN); Nangolo Flats in Marienfluss Valley 1712Ad (TM); Ondondujengo River 1712 Cc (TM, ZMFK); Okotusu area 1711Bd (TM); 16 km W of Oru- - pembe 1812Ad (SMWN); Ouhandjo area 1712Ba (SMWN); 48 km NW of Ouhandjo 1712 Ac (TM). Field notes: This species is not normally associated with the major sand seas. Although their regular undulating tracks may be seen in the larger sand accumulations of their range, they are more often found in small patches of sand on stony hillsides and specimens may actually be found under stones in the inland part of their range. In the sandy parts of the Hart- mann's Valley this taxon is sympatric with 7: johnsonii and only capture will allow identifica- 150 W. Haacke tion of the species which produced the track. As most other fossorial skinks, these lizards move just below the sand surface and exposures, as when a stony area is crossed, only rarely happen. It is as yet unclear what predators this species may have. No information about repro- duction is available. The smallest juvenile (TM 63478) of 38 mm HB length, tail broken was collected during April 1985 indicating that parturition takes place during late summer. Typhlacontias punctatissimus bogerti (new status) Laurent (Figs 3a, 8) Typhlacontias bogerti Laurent 1964. Publ. cult. Comp. Diam. Angola 67: 82 (Type locality: 35 km S of Mocámedes (= Namibé)’, Angola. Types in Dundo Museum, Angola). Haacke 1965: 22 (part., Angolan specimens). Greer 1970: 14. Mertens 1971: 70. Welch 1982: 79. Material: 29 specimens examined. Types not seen. Etymology: Named in honour of renowned American herpetologist Dr Charles M. Bogert. Diagnosis: A slender form with enlarged second lower labial scale and general characteristics described for the typical form but second and third upper labial in touch with eye. Description: In general very similar to the typical form apart from the subspecific character, i.e. the second and third upper labial are in touch with the eye. This situation is fairly persistent amongst the series of 14 specimens from Mocämedes (= Namibé), although individ- ual variation occurs, such as the second upper labial being separated from the eye by the preocular as in the typical form (TM 24465, 24469, 24473), or the second upper labial may just be too short to reach (TM 24463, 24469). The Kakolo windmill sample (n = 7) may be seen as of an integrate population, as three are normal 7: p. bogerti, but four have variations which do not allow the second upper labial to reach the eye. Usually two, rarely three postoculars. Cleared specimen TM 40727 ( Kakolo windmill), has 62 spinal vertebrae, remnants of the pectoral and pelvic girdles with tiny remnants of the proximal femur head, while R-224, in the collection of the Namib Desert Research Station, Gobabeb, from the eastern edge of the sand sea south of the Curoca River, has minute limb rudiments adjacent to the cloaca. Ventrals: Range = 114135, x= 122.7, SD= z= 6.99, n =27. Subcaud:. Ranger 59:69 ADA DS AC En JUL Size: Max SVL = 86 mm (TM 40725 Kakolo Windmill). Max total = 123 (83 + 40) mm (TM 40724 Kakolo Windmill). Min SVL = 40 mm (TM 40730 Kakolo Windmill). Min total = 58 (40 + 18) mm (TM 40730 Kakolo Windmill). Relative tail length: Range = 44—62 Yo, SNS) Sa n= O Colouration: Similar to the typical form. Range (Fig. 8): Northern Namib Desert of Angola in the lona Park west of the Curoca Sand Sea and north to Mocámedes (= Namibé). Distribution records: ANGOLA: Curoca River crossing 1612Ad (NDRS); Kakolo Windmill 1612Cd, Mocämedes (= Namibé) 1512Aa (EMUS, IRB, TM, ZMFK); 10 km S of Mocámedes (= Namibé) 1512Ac, 34 km S of Mocámedes (= Namibé) 1512Ca, (TM); 8 km SE of Pico de Azevedo 1512Da (SMWN). Fieldnotes: The types were collected under a stone and under a Welwitschia plant (Laurent 1964), as were those from 10 km S of Mocámedes (= Namibé). In Mocámedes specimens were found next to the trunk and amongst the roots of trees lining the streets (C. Koch, pers. com.). At the Kakolo Windmill the specimens were collected at night while they were active near the surface of red sand accumulations against a stony hillside. The large female paratype collected on 23 November 1949, contains embryos with well developed eyes (Laurent 1964: 84). Typhlacontias punctatissimus brainei ssp. n. (Figs 3c, 8) Etymology: Named after Mr Steve Braine, former ranger of the Skeleton Coast Park, Namibia, for his contribution towards the knowledge of the fauna of that area during his term of office there. Holotype: TM 68443, collected by W. D. Haacke, 11. 4. 1989. Paratypes: 15 specimens. TM 31497-8, Sechomib River Valley 29 km SW of Orupembe Revision of the scincine genus Typhlacontias 151 (1812Ad) coll. W. D. Haacke, 4. 10. *65. TM 32375, same as previous specimens but 7. 5. *66. TM 56863, Hoanib River 2 km E of Amspoort, 19°20’S 13°12’E, coll. S. Braine 18. 9. ’83. TM 68437 —68442, ZFMK 57532, same data as holotype. SMWN 2011, Khumib River 16 km E of Sarusas East (1812Ba), SMWN 2014(3), 29 km SW of Orupembe 1812Ad, coll. W. J. Steyn 4:10:65. Typelocality: Khumib River 25 km from the coast, Skeleton Coast Park, Opuwo District, Namibia, 18°44’S 12°36’E (1812Da). Diagnosis: In general appearance and characteristics similar to the typical form but upper labials 2 and 3 appear to have fused, resulting in five, instead of six upper labials and the large second upper labial is in contact with the eye. Preocular absent or minute and partly covered by loreal. Eye usually in touch with the least number of scales of all taxa, namely second upper labial, loreal, anterior supraocular and two postoculars. Description: Holotype TM 68443, sex undetermined, 112 (70 + 42) mm, tail 60 % of HB length, 18 scale rows round middle, 120 ventrals, 68 subcaudals, five upper labials of which the large second (fused 2 + 3?) is in contact with eye, preocular absent or minute and covered by large loreal, two supraoculars form lateral edge of frontal and larger, anterior also in contact with upper edge of eye, two small postoculars, four lower labials of which the second is very large (Fig. 3c). Variation: Similar to holotype with minor individual variations, s. a. second and third upper labials not fused on both sides (SMWN 2014c) and on right side only (SMWN 2011). Nentrals Range = 117128 n 12, x = 122.3, SDZ = 2:64. Subcaud.: Range = 50—68; me 169x° 161:8, SD = a= 5.63. Size: Max SVL = 70 mm (TM 31498, holotype). Max total = 112 (70 + 42) mm (TM 31489, holotype). Min SVL = 41 mm (TM 56863, near Amspoort, neonate). Min total = 9 (41 + 18) mm (TM 56863, near Amspoort, neonate). Relative tail length: Range = 45—60 Yo, IS 16 SD) = a= 4:86. Range: Recorded from the Sechomib to the Hoanib River valleys, just outside the border of the Skeleton Coast Park, Namibia, and not above an altitude of about 600 m a.s.l. Field notes: Wind-blown sand accumulations on rocky hillsides, flat stone desert, hummocks and vegetated banks along and in drainage valleys. Remarks: The fact that fusion of the 2nd and 3rd upper labial is part of the individual variation within the range of 7. p. punctatissimus, is good supporting evidence that these taxa are only subspecifically distinct. Typhlacontias gracilis Roux (Figs 4b, 7) Typhlacontias gracilis Roux, 1907. Rev. Suisse Zool. 15: 83, Fig. 3—4 (Type locality: ‘Zambezi’. Type in Neuchatel Nat. Hist. Museum). Angel 1920: 617 (Lealui). Pitman 1934: 306. De Witte & Laurent 1943: 35, fig. 54 —55. Greer 1970: 14. Broadley 1971: 58. Welch 1982: 79. Branch 1988: 125. Typhlacontias gracilis gracilis, Broadley 1971: 58. Material: 113 specimens examined. Type seen. Etymology: Referring to its gracile or slender body shape. Diagnosis: Least tendency for head scales to fuse, as a consequence has more than other taxa. Four normal sized lower labials with the fourth slender and horizontal, six upper labials with third entering orbit, prefrontal separated from frontoparietal by third supraorbital (In type antero-upper temporal fused with third supraorbital Fig. 4b). Eye usually in touch with third upper labial, two preoculars, two of three supraoculars and three postoculars. Description: Large rostral not strongly notched laterally. Dorsal head scalation as typical for this genus with frontonasal slightly larger than internasal and frontal. Narrower frontal allows space for a prefrontal wedged between frontonasal and two supraoculars. Frontoparietal fits into notch between frontal and interparietal. Two supraoculars in touch with eye are followed by a third scale, here referred to as third supraocular, separated from eye by upper postocular, but in touch with the frontal thereby separating frontoparietal and prefrontal. All 152 W. Haacke other taxa, except 7. rohani, have two supraoculars of which at least one is in touch with the lateral edge of the frontal. Usually two preoculars and one to three postoculars occur. Type has anterior upper temporal fused with third supraocular and this enlarged scale touches frontal laterally, thereby separating prefrontal from frontoparietal (Fig. 4b). NMZ 7894 from Kalabo has labial one and two fused, with that enlarged scale on the left side showing a slight dorsal notch. Accordingly only five upper labials exist and “second” enters the orbit. Nevertheless, frontoparietal and interparietal are still separated by third supraocular and two preoculars exist, thereby distinguishing it from Kalabo 7. rohani. Opinions of what constitutes a labial scale appear to differ. This reviewer considers 7. gracilis to have six upper labials and usually four lower labials. The fourth lower labial is narrow and horizontal, some- times only visible when the mouth is opened and it extends to below the sixth upper labial. Body covered with 18 tight-fitting scale rows. Anal scales not markedly enlarged. Cleared specimens TM 33866 and TM 33880, both from Kalabo, have 68 and 60 spinal vertebrae, remnants of pectoral and pelvic girdles but no limb rudiments. Ventrals: Range = 122— 147, x = 134.66, n = 43, SD = + 4.74. Subcaud.: Range = 60—65, OS) SIDS ae o. Size: Max SVL = 84 mm (NMZ 7875 Kalabo). Max total = 93 (64 + 29) mm (NMZ 6799 Kalabo). Min SVL = 38 mm (NMZ UM 10060 Kalabo). Min total = 79 (53 + 26) mm (NMSR 3406 Kabompo). Rel. tail length: Range = 40—61 %, x = 49,34 Y, n = 29, SD = + 4.39. Colouration: Buff to straw coloured. Each scale is dark centred, with dark spot varying in size and equal sized spots may form lines, s.a. two faint vertebral lines which start in some dark marks on top of the head. A lateral line, which, when well developed, may be dark brown and include the snout, starts in the eye area extending to the tail tip. The intensity of the lined pattern varies extensively. Range (Fig. 7): Western Zambia, i.e. Barotseland in the upper Zambezi Plains. Distribution records: ZAMBIA: Kabompo 1324Ca (NMZ); Kalabo 1422Dc (NMZ, TM, ZFMK); Lealui 1523Aa (MHNP, Angel 1920); Lyumba 1430Ca (FMNH, Broadley 1971).. Field notes: Found in high rainfall area in Kalahari sand together with other fossorial skinks s.a. Typhlosaurus lineatus jappi, Lygosoma sundevallii as well as several amphis- baenian taxa and their snake predators of the genera Amblyodipsas and Xenocalamus. This taxon occurs in an area with a higher average annual rainfall than any of the other Typhlacontias taxa and as a consequence may have to cope with flooding of the Barotse Plains and associated hazards. Remarks: The re-examination of the long series of specimens from Kalabo provided proof that at that site 7! rohani occurs sympatrically with 7. gracilis and therefore these two taxa are here reinstated as full species. Typhlacontias rohani Angel (Figs 4a, 7) Typhlacontias rohani Angel, 1923. Miss. Rohan-Chabot Angola-Rhodesia (Paris) 4: 162, figs 6—8. (Type locality: Lwankundu, south-east Angola. Type in Paris Nat. Hist. Museum). Angel 1942: 109. De Witte & Laurent 1943: 35. Welch 1982: 79. Typhlacontias ngamiensis FitzSimons, 1932. Ann. Transv. Mus. 16: 37 (Type locality: Mothlatlogo = Bothatogo, Botswana. Type in Transvaal Museum, paratype in Field Museum, Chicago), and 1935: 374, figs 17—19, also 1943: 239. De Witte & Laurent 1943: 35. Mertens 1955: 77 and 1971: 71. Broadley 1962: 805. Haacke 1965: 20. Welch 1982: 79. Griffin 1985: map 51. Typhlacontias gracilis ngamiensis, Broadley & Blake 1979: 7. Branch 1988: 125. Typhlacontias gracilis rohani, Auerbach 1987: 103. Broadley 1988: 376. Material: 233 specimens examined. Type seen. Etymology: Named in honour of M. de Rohan-Chabot who led an expedition in 1914 into the south-eastern part of Angola, an area which is herpetologically still unexplored. Diagnosis: No enlarged lower labial scales, five upper labials with second entering orbit, frontoparietal and prefrontal in contact with one another. Revision of the scincine genus Typhlacontias 153 Fig. 4: a) T! rohani Angel, holotype MHNR 1923.1.1, Lwankundu River, Angola; b) 7! gracilis Roux, holotype, unnumbered. Upper Zambezi River, Zambia (For abbreviations see Fig. 5.) Description: Snout depressed with a horizontal edge. Rostral only as long as the narrow internasal plus the broad fontonasal. Rostral with only slight lateral notches but fairly long nasal sutures enclosing minute postnasals. Five upper labials with the second (2 and 3 fused?) entering the orbit. Usually one or two small preoculars. In the Kalabo series these two scales have fused into a narrow vertically elongate scale separating the loreal from the eye, while in the type the loreal enters orbit above a small preocular (Fig. 4a) and in the type of 7? ngamien- sis no preocular is apparent, as it is often the case in other specimens when the minute scale is covered by the loreal. Frontoparietal and prefrontal in contact, separating the two large supraoculars from contact with frontal, usually two postoculars of which the lower prevents contact between lower anterior temporal and upper labial in contact with eye. Four lower labials of which the last is narrow, horizontally elongate and sometimes invisible unless the mouth is opened. Midbody scale rows 18, four anal scales not significantly different from ventrals. Ventrals: Range = 124—146, n = 193, x = 134.26, SD = + 4.57. Subcaud.: Range = 64—84, 5 exe— 69.4. SD = ae 11. 3: Size: Max SVL = 90.0 mm (TM 30971, Tsodilo Hills). Max total = 125 (83 + 42) mm (TM 38347, 50 km W of Tsumkwe). Min SVL = 33.0 mm (TM 39313, nr Katima Molilo). Min total = 50 (33 + 17) mm (IM 39313, nr Katima Molilo). Rel. tail length: Range = 48—73 %, Br 6957 150-95 Yo, SD E 5:98. Colouration: Similar to 7. gracilis, but in some areas, such as the Kwando River in the Caprivi Strip, Namibia, exceptionally dark specimens occur, as the dark centre blotches of each scale are larger and better developed than usual. In these specimens the body sides are dark brown and the head, especially the snout, is also very dark. Discussion: 7. rohani and T. gracilis are very similar in general characteristics, s.a. size, scalation, colouration, etc. which justifies the idea that they are only subspecifically distinct. However, there are some differences of which the main features are very consistent, nl. the facial scalation. 7. rohani has a fairly wide geographical range in north-eastern Namibia, northern Botswana, western Zimbabwe and probably south-eastern Angola. Throughout this range the available specimens have five upper labials with the second (2 and 3 fused?) in contact with the orbit. Furthermore the frontoparietal and prefrontal are in contact with one another. In contrast T. gracilis has six upper labials and the third enters the orbit while the 154 W. Haacke frontoparietal and prefrontal are separated by a third supraocular scale. As the main range of T. gracilis lies in Barotseland of western Zambia with a collecting gap of about 300 km between Kalabo and the Caprivi Strip of Namibia the subspecies status was probable and difficult to dispute. However, a re-evaluation of a large number of the sample from Kalabo proved that 18 specimens have the characteristics of 7! rohani from further south with no significant variation in the main features. A local variation of that sample is a slender, upright preocular, separating the loreal from the orbit. A single 7. gracilis specimen (NMZ 7894) has only five upper labials as the first and second fused. Of these the left one has a slight dorsal notch, indicating that that large scale is indeed a result of fusion. Two preoculars are present and the supraoculars are divided by the upper of the first temporals. This means, although it suggests a relationship with 7! rohani, it is a T: gracilis with a slight aberration. Greer (1970: 14) compared a skull of 7: gracilis with one of T! rohani (= T. ngamiensis) and found a difference in the palatal rami between the two taxa. To further test the relationship between T. rohani and T. gracilis, especially with reference to Kalabo, some statistical tests were carried out. The ventral scale counts vary significantly and show up differences. Furthermore the subcaudal counts and relative tail lengths of 7? rohani from Kalabo and 7. rohani from elsewhere, were compared but no difference was found, indicating conspecificity. However, when the data for the Kalabo 7! rohani were subjected to a T-test in comparison to the data of 7: gracilis, in both the cases, i.e. subcaudals (sig. level = 5.01784E-4) and tail/body ratio (sig. level = 3.19417E-3) were found to be highly significant, providing additional evidence that two species are involved. As a consequence of proven sympatry of these two taxa at Kalabo, they are accepted as full species of which the ranges overlap in western Zambia. Range (Fig. 7): Kalahari sand with an average annual rainfall above 400 mm p.a. in NE Namibia including the Caprivi Strip, northern Botswana, western Zimbabwe, SE Angola and western Zambia as far as the Zambezi Valley. Distribution records: ANGOLA: Lwankundu River 1620Ad (MHNP); BOTSWANA: Bothatogo 2022Db (TM, FMNH); Dibejama 1821Ad (FMNH); Dodo Crossroads 1825Cb (NMZ); near Gubatsha Hills 1824Ca (TM); Kasane 1725Cc, Maun 1923Cd, Ngwezumba . Bridge 1824Bd (NMZ); Savuti Camp 1823Db, Shakawe 1821Bd (TM); Tamafupi 1926Ac (NMZ); Tsoe 2024Cb (LACM); Tsodilo Hills 1821Db (TM). NAMIBIA: near Grootfontein 1919Bc (TM); Kanovlei 1919Ab (SMW); Katima Mulilo 1724Cb (TM); 15 km and 65 km WSW of Katima Mulilo 1724Ca (NMZ, TM); Kwando River 1723Cd (TM); Kwetche (Mahango Reserve) 1821Ba, Leeupan 1820Bd, 10 km S of Leeupan 1820Db (SMW); Makambu 1718Ad, Mangetti Forest 1818Da (TM); Mpungu 1718Ca (SMW); Odila Dam 1716Bd (TM); Rundu 1719Dd (TM, SMW); Samangeigei 1920Aa (TM); 25 km NW of Tamsu 1820Ba; Tsumkwe 1920Da (SMW); 50 km W of Tsumkwe 1920Ac, 55 km NW of Tsumkwe 1920Aa, Verskyn 1919Aa, Wenela 1723Cb (TM); Western Veterinary Fence 2120Ac (SMW). ZAMBIA: Kalabo 1422Dc (NMZ 6766-7, 6769, 6780, 6782, 6787, 6791, 6793, 6795, 7894, 10048, 10050, 10052-4, TM 33868, 33887, 35580). ZIMBABWE: Bembesi Vlei 1826Dc (NMZ); 10 km NW of St. Mar- tin’s Mission 1927Bb (TM); Dahlia 1827Ca, Dandari 1826Cc, Dett 1826Db, Dina Pan 1826Cd, 6 km S of Dina Pan 1926Ab, Dopi Windmill 1826 Dd, Good Luck Ranch 1827Ca, Gubalala 1826Dc, Kennedy 1827Cc, Lupane 1827Dd (NMZ); Main Camp, Hwange 1826Db (TM); Man- zimbomvu Pan 1826Cc, Maraposa, Nyamandlovu 1928Cc, Shapi Pan 1826Dc, Somavundhla Pan 1927Aa, Teaklands 1927Bd, Tibukai Pan 1825Dd, Verney’s Pan 1826Dd, Victoria Falls 1725Dd, Westwood Ranch 1725Dd (NMZ). Field notes: Active sandswimmers which forage just below the sand surface at night, leaving thin, regularly undulating tracks. As more vegetated environment is utilised these tracks are not as obvious as those occurring in desert sand. May be found by raking in leaf litter under bushes and elsewhere or by turning over logs, as no stones occur. Feeding on small insects s. a. beetle larvae and termites. Biological information is limited. Appears to be ovoviviparous as a specimen from 15 km W of Katima Molilo (NMZ-UM 22889) contains three full term young. However, a specimen from Kalabo (NMZ-UM 6766) is gravid with four large eggs (each about 3.5 x 5.5 mm) which show no sign of development. Unfortunately no collecting dates exist which would allow deductions relating to the season. Other sandswimmers which share the range are Typhlosaurus lineatus lineatus in Namibia and Botswana and 7. /. jappi in western Revision of the scincine genus Typhlacontias 155 b Fig. 5: Typhlacontias rudebecki sp. n., holotype TM 25465, Sáo Nicolau, Angola; a) dorsal, b) ventral, c) lateral view. CS = chin shield, F = frontal, FN = frontonasal, FP = frontoparie- tal, IN = internasal, IP = interparietal, L = loreal, LL = lower labial, M = mental, P = pa- rietal, F = prefrontal, PM = postmental, PO = postocular, PRO = preocular, R = rostral, SO = supraocular, T = temporal, UL = upper labial. Zambia. Several amphisbaenians also share this moist environment s.a. Zygaspis quadrifrons, Monopeltis anchietae, M. leonhardi, M. sphenorhynchus, Dalophia pistillum and D. longi- cauda. Several burrowing snakes are specialised predators of these sandswimmers and will also consume Typhlacontias rohani and T. gracilis. The most probable predators are Amblyodipsas ventrimaculata, Xenocalamus mechowii and X. bicolor. Typhlacontias rudebecki sp. n. (Figs 5a, b, c, 7). Etymology: Named after the collector Dr G. Rudebeck, who was a member of the Swedish Lund University Expedition to South Africa at that time. Holotype: TM 25465, sex undetermined, collected during September 1956 by G. Rudebeck. Type locality: Säo Nicolau, Mocämedes (= Namibé) district, Angola (1412 Ad). Diagnosis: Lower labials not fused, number of ventrals higher than of related 7. gracilis. Rostral not notched laterally, five upper labials with first (= fused 1 and 2?) very large, extending to below anterior edge of eye, loreal separated from second upper labial which is ‘in touch with the eye. Description: TM 25465, Holotype. Tail incomplete and body cut into two parts, 49 mm and 33 mm long with a total of about 152 ventrals and 18 scale rows around the body. Rostral laterally not notched and subequal in length to first three azygus scales on the top of the head i.e. internasal, frontonasal and frontal (Fig. 5a). Nostrils, followed by tiny oblong nasal scale, pierced in sides of rostral with short straight suture connecting them to back edge. Eye round, exposed but with slight overlap by second upper labial and supraoculars. First upper labial 156 W. Haacke NAMIBIA \ | | Ry? 0, en m € >. Fig. 6: Distribution of Typhlacontias brevipes in the Namib Desert. large (1 & 2 fused?), anteriorly extending to above level of eye and posteriorly past anterior edge of the eye. Loreal separated from eye and second upper labial by single preocular. Five upper labials of which second is in broad contact with eye, two small postoculars, two large supraoculars in touch with prefrontal, posterior supraocular in touch with frontal, thereby separating similar-sized prefrontal and frontoparietal. Pineal eye clearly visible in the relatively small subpentagonal interparietal. Frontonasal larger than internasal or frontal. The large mental is followed by a large postmental (Fig. 5b), four lower labials of which the narrow horizontal fourth scale is difficult to see when the mouth is closed. Colour: Dorsum pale with a light brown vertebral stripe two half scales wide, bordered by a dark lateral line three scales wide which extends forward through the eye to the nostril. Ventrally each scale of four rows bears a light brown spot. Discussion: According to the unfused lower labials this taxon is closer to 7. rohani and T. gracilis, than to the Namib species. However, the general proportions differ and the habitat, although not specifically recorded, is coastal Namib Desert. 7: gracilis and 7. rohani are restricted to Kalahari sand at a fairly high altitude with a fairly high average annual rainfall. Revision of the scincine genus Typhlacontias 157 ANGOLA YO U Y O 7. johnsonii © T.rudebecki A T.rohani Y T.gracilis @sSympatry at Kalabo == Edge of Kalahari Fig. 7: Distribution of 7. johnsonii and T. rudebecki sp. n. in the northern Namib Desert and T. rohani and T. gracilis in the northern Kalahari sands. Discussion Character analysis: Greer (1970: 14) gave a description of the general charac- teristics of the genus. He discussed the skull of 7. brevipes, T. gracilis and T. rohani (= T. ngamiensis) and their general external characters. His identification of head- scales differs from that of this author. The supraoculars of Greer, in touch with the sides of the interparietal, are here considered to be frontoparietals as these scales do not reach the eye, but are in touch with frontal and interparietal in all species and usually the parietals as well. The extra scale wedged laterally between frontal and frontonasal and two supraoculars are here referred to as prefrontals. Scales in contact with the dorsal aspect of the eye are referred to as supraoculars, as was done when describing 7. brevipes (FitzSimons 1939: 15). Greer (op. cit.) interprets the presence of a large interparietal and a single transverse internasal (postrostral) as an indica- 158 W. Haacke SE E toon ee eS (A AO _——-4— A AS O Type Locality © T.p.bogerti O T.p.punctatissimus : > A T.p.brainei E Fig. 8: Distribution of the three subspecies of Typhlacontias punctatissimus in the northern Namib Desert. tion of a relationship with Scelotes and he assumes that Typhlacontias evolved in arid areas of southern Africa, with 7! brevipes, due to the presence of external limb rudiments and a relatively large postnasal scale, as the most primitive form. In the light of Broadley’s (1968) evaluation of evolutionary trends in the acontine scincid genus Typhlosaurus, one may have to re-evaluate Greer’s suggestion. A number of morphological characters of the eight taxa are listed in Tables 1 and 2. The presence or absence of the very large second (= fused second and third?) lower labial in combination with the presence of prefrontals separates this genus into two groups. T. brevipes, T. johnsonii and T. punctatissimus, have this large second lower labial, but no prefrontal and are Namib Desert endemics (Figs 6, 7, 8), while 7! gracilis, T. rohani and T. rudebecki sp. n. have no fused lower labials but a prefrontal above the supraoculars. The first two species of the latter group are widespread in Kalahari sand with high rainfall at fairly high altitude above sea level (Fig. 7). In contrast the presence of 7! rudebecki, of which the range is still unknown, on the northern Namib coast at sea level is not yet understood. Broadley (1968: 18) lists three trends in Typhlosaurus Wiegmann (Acontinae), which imply degrees of evolutionary advancement or a greater degree of specialisa- Revision of the scincine genus Typhlacontias 159 tion to a fossorial lifestyle. These might be applicable to Typhlacontias as well, nl. a) a reduction in the number of headshields, b) progressive attenuation of the body, and c) simultaneous increase in the number of ventrals. The latter trend in fossorial lizards is usually correlated with a reduction in tail length. All these changes are obviously related to greater specialisation as sand swimmers. Applying these para- meters, 7. gracilis with the most numerous headshields and long tail has retained with T. rohani a number of primitive characters, although the latter has started to reduce its tail length. 7! rudebecki might rate as the most specialised or advanced in this group due to the amount of head scale reduction and the apparent attenuation of the severed body. Although primitive characters persist, limb loss (one cleared 7. gra- cilis and three 7! rohani checked), an advanced feature, is complete. The Namib Desert endemics with the enlarged second lower labials, i.e. 7. brevipes, T. punctatissimus and T: johnsonii show a similarly involved picture. T. brevipes with its externally visible limb rudiments and enlarged postnasal, which supposedly are retained ancestral characters, is the most primitive species in the eyes of Greer (1970: 15). However, some features appear to be specialised or advanced, such as the high number of ventrals, a relatively short tail and a wedge-shaped head with a reduced number of head scales. 7. punctatissimus has retained a gracile body shape and a long tail similar to 7. gracilis and T. rohani but, although it has a reduced number of head scales, internally 1t may still have minute femur and occasionally even externally visible limb rudiments. Within the three subspecies of this species — small variations in the reduction of head scales exist. T. johnsonii, with its high number of ventral scales, the relatively shortest tail, the most depressed wedge-shaped head and with the greatest reduction of head scales, is according to these criteria the most specialised or advanced taxon of this group and the genus. To test these deliberations a data matrix for 14 characters was generated for the 8 recognised taxa with Sclotes brevipes as an outgroup (Tables 3 & 4). The phylogene- tic systematics computer program Henning86, version 1.5 (Farris 1988), was used to generate a phylogenetic hypothesis, expressed as a cladogram. r 0 Scelotes bipes F 14 = fel T.brevipes 7 E 12 LL 2 T. johnsonii | | r 4 T.p.punctatissimus | Namib L413 =| fF 3_T.p.bogerti | Lıı ats dap brainel = | fe 7 T.rohani =, Lıo =| m6 T.gracilis -- Kalahari Lo=>Lg Tr. rudebecki ]- Namib 160 W. Haacke Table 3: Data list for Typhlacontias taxa and Scelotes bipes as outgroup. le EMAIL, YO 300 = 0, <50 YN S15 S40 Y =2 23000 = 3. 2. Max SVL <100 mm = 0, >100 mm = 1. 3. Ventrals <5 =O, KIS = ll <5 SW, S145 = 3. 4. Vertebrae 6063-108 O—/0 =i, 0-15 = 2% LS = 3: 5. Enl.llab2 Absent = 0, present = 1. 6. Upl./eye ya, AN, YO e fator/upl2, = Contact 05 no cont e Sa UNO SS 0 2= 1 22 oF Hind mb "Ext, 0 extinto — 1, absene 2. 10. Prefrontal Present = 0, absent = 1. 11. Rostr. sut. Shrt & pstn = 0, Ing & pstn = 1, no pstn = 2. 12° >Supocıimt, 83-052 — OZ ZA sola 135° Supocieye 2 0, 1 = 1, 03% 14. Eye lid Present = 0, absent = 1. Table 4: Data matrix for eight Typhlacontias taxa and Scelotes bipes as outgroup. 1 2.3245 (6 TITAN T. brevipes 2 1 27.1 1 0 0 1 20" OS 0 T. johnsonii 3 1 2 he 1 1-07 72.00 DV 1 0 T. p. bogerti OS iOS IKOHRO i PLS) 1 1 1 Ls PRES #30 T. p. punctat. 0. OR EOEI Bi O ON 1 1 1 3 1 0 T. p. brainei 07705407 4,9 al 1 Otal 1 1 1 3 I T. gracilis O QA LO OF 0. 0 207 reel 0: , 20,50 T. rohani U "HL Ve a 1 I or We 2 OO T. rudebecki DO 1077 ¿SAO O 1 OQ OF EEE 1 OO Scel. bipes 3 0 0 -0 0.0. O- (O12 105 OB | A single tree (Length 30 steps, consistency index 0.67, retention index 0.62) was generated, confirming the basic clustering as viewed by the author. T. brevipes, T: johnsonii and T: puntatissimus of the Namib Desert with the enlarged second lower labial, no prefrontal scales and leg rudiments in two taxa form the more ancestral group. 7. gracilis, T. rohani from Kalahari sand and 7! rudebecki, from the Namib coast, with normal-shaped lower labials, prefrontal scales, slender long-tailed bodies and no indication of limb rudiments, form an apparently more advanced group. In this context the status of 7. rudebecki is slightly compromised, as that single tailless specimen could not be cleared, reducing its data content by three. In addition the same data matrix was run using PAUP version 3.1.2d5 (Swofford 1985). PAUP generated two most parsimonious trees, 56 steps in length with a consistency index of 0.46. The measure of phylogenetic content in the data set as measured by the gl-statistic (=0.36) showed a significant phylogenetic signal. A strict consensus analysis of these two trees produced an unresolved trichotomy between T. brevipes, T: johnsonii and T.! rudebecki. This situation lacks credibility and is probably due to the fact that the type specimen contributed three less data points to the matrix. A bootstrap analysis of the two trees indicated a 100 % frequency in all internal branches. It seems essential to increase the data set for a thorough phylogenetic analysis, which is only possible once more material, in particular of 7. rudebecki, becomes Revision of the scincine genus Typhlacontias 161 available. Until such time the Hennig86 must serve as a preliminary hypothesis of phylogenetic relationships. Acknowledgements I am deeply indebted to the Directorate of Resource Management of Namibia and its predeces- sors for the many occasions that I was allowed to collect there, work at the Namib Desert Research Station, Gobabeb, and enter various conservation areas. I am also grateful to the Department of Wildlife, Republic of Botswana and of Angola for being allowed to work there. A special word of thanks in memory of the late Drs Charles Koch and Willem Steyn who first introduced me to the Namib Desert. Many thanks also to the various travel companions, nature conservationists, assistants, guides and colleagues, too many to mention individually, who made those enjoyable trips so successful. I am indebted to Dr E. R. Brygoo, Paris Natural History Museum, Dr J. L. Perret, Geneva, Dr C. Dufour, Museum of Natural History, Neu- chatel and Dr U. Svensson, Gothenburg Natural History Museum, for the loan of type material, as well as to Dr D. G. Broadley, National Museum of Zimbabwe, Mrs E. Griffin, State Museum, Windhoek, Mr M. K. Griffin, Directorate of Wildlife, Namibia, Dr M. K. Seely, Namib Desert Research Station, Gobabeb, for the loan of specimens. Dr M. Kriiger, Mr. M. Whiting and Mr. C. Chimimba assisted with the statistical analysis. Much appreciated assistance was provided by Lomi Brown, Stephanie Ritter, Cecilia Herbst and Lauretta Mahlangu who helped to gather information. Ms Ritter also typed the original manuscript and Michelle Venter assisted in preparing the illustrations. Zusammenfassung Typhlacontias ist eine der zahlreichen, sandwühlenden Skinkgattungen mit degenerierten Gliedmaßen. Die Verbreitung der sechs bekannten Arten ist auf die Namib-Wüste und das zentrale Kalahari-Becken des südlichen Afrikas begrenzt. Die Gattung wurde an Hand von zwei schlecht konservierten Exemplaren von Bocage (1873) beschrieben, so daß der Autor 1895 nach Erhalt eines dritten Exemplares von der Terra typica die Art nachbeschrieb. Letzteres Exemplar repräsentierte jedoch eine unbeschriebene Art, dazu mit abnormaler Beschuppung, wodurch die Art 7. punctatissimus unidentifizierbar wurde. Die drei Exemplare wurden bei einem Brand zerstört. Als Andersson sieben weitere Exemplare vom selben Fundort bekam, erkannte er das Problem, revalidierte 1916 die Originalbeschreibung und beschrieb die Art 7. Johnsonii neu. Das einzige Exemplar in dieser Serie, das zu 7! punctatissimus gestellt werden konnte, hatte sichtbare Beinrudimente, die bisher nur bei einem weiteren Exemplar der Art festgestellt werden konnten, weshalb es nicht als Neotypus geeignet ist. Anderssons Arbeit wurde leider bis jetzt völlig übersehen. Inzwischen wurde sehr viel neues Material gesammelt und neue Arten beschrieben, die aber erst nach dem Wiederauffinden von Anderssons (1916) Arbeit zugeordnet werden konnten. In der vorliegenden Arbeit werden dazu eine neue Art und Unterart beschrieben, die Verbreitung aller acht Taxa dokumentiert und ein Versuch gemacht, die verwandtschaftlichen Beziehungen in dieser Gattung zu verstehen. Bibliography Andersson, L. G. (1916): Notes on the Reptiles and Batrachians in the Zoological Museum at Gothenburg with an Account of some new Species. — Medd. Göteborg. Mus. zool. Afd. 9: 5-41, 3 figs. Angel, F. (1920): Liste de reptiles du Haut-Zambeze et de l'Afrique australe. Description d'une espece nouvelle du genre Monopeltis. — Bull. Mus. Paris 27: 42—44, figs 1—3. Angel, F. (1923): Reptiles. — Pp. 157—169 +1 pl., in: Mission Rohan-Chabot Angola- Rhodésia 1912—1914, 4 (1). Impr. National, Paris. Angel, F. (1942): Sur la synonymie d'une espece de lézard du genre Typhlacontias. — Bull. Mus. Paris, Ser.a, XIV (2): 109. Auerbach,R. D. (1987): The Amphibians and Reptiles of Botswana, 295 pp, 19 pls., illustra- tions, maps. — Mokwepa Consultants, Gaborone, Botswana. 162 W. Haacke Bocage, J. V. Barboza du (1873): Reptiles nouveaux de l’Interieur de Mossamedes. — Jorn. Acad. Sci. Lisboa 4: 247—253. Bocage, J. V. Barboza du (1895): Herpétologie d'Angola et du Congo. — Lisbon. 203 pp., 19 pl. Boulenger, G. A. (1885): Catalogue of lizards in the British Museum (Natural History) 1: 1—436, 32 pls. — London. Branch, W. R. (1988): Field guide to the snakes and other reptiles of southern Africa, 326 pp, 96 pls., maps. — Struik Publishers, Cape Town. Broadley, D. G. (1971): The reptiles and amphibians of Zambia. — The Puku 6: 1—166, 25 pl., 13 maps. Broadley, D. G. (1988): A check list of the reptiles of Zimbabwe with synoptic keys. — Arnoldia, Zimbabwe 9: 369—430. Broadley, DG. & D. H. Blake (1979): A check list of the reptiles of the National Parks and other conservation areas of Zimbabwe Rhodesia. — Arnoldia, Rhodesia 8 (35): 1—15. DeWitte, G.F. £ R. Laurent (1943): Contribution a la systématique des formes dégradées de la famille de Scincidae apparentées au genre Sce/otes Fitzinger. — Mem. Mus. Roy. Hist. Belgique (2) 20: 1—44, 61 figs. Farris, J. S. (1988): Hennig86, version 1.5. Reference. — Private print. FitzSimons, V. F. M. (1932): Preliminary descriptions of new forms of South African Reptilia and Amphibia from the Vernay-Lang Expedition, 1930. — Ann. Transvaal Mus. 1543340. FitzSimons, V. F. M. &D.H. Blake (1935): Scientific results of the Vernay-Lang Kalahari Expedition, March-September 1930. Reptilia £ Amphibia. — Ann. Transvaal Mus. 16: 295397, 30 figs., 2 pls: FitzSimons, V. F. M. & D. H. Blake (1939): Descriptions of some new species and sub- species of lizards from South Africa. — Ann. Transvaal Mus. 20: 15, figs 15—17. FitzSimons, V. F. M. (1943): The lizards of South Africa. — Transvaal Mus. Mem. 1, v-xv, 1—528, 384 figs, 24 pl. Greer, A. E. (1970): The systematics and evolution of the Subsaharan Africa, Seychelles and ' Mauritius scincine scincid lizards. — Bull. Mus. comp. Zool. 140: 1—24. Greer, A. E. (1983): On the adaptive significance of the reptilian spectacle: the evidence from scincid, teiid and lacertid lizards. — Pp. 213—221 in: Rhodan, G. J. & K. Miyata (eds.), Advances in herpetology and evolutionary biology. Mus. comp. Zool., Cambridge Mass. Griffin, M. (1985): A preliminary survey of the occurrence and distribution of amphibians, reptiles and mammals in Bushmanland, Eastern Kavango and Mahango areas with some notes on historical distribution. — Internal Report: Direct. Nat. Cons. Tour., SWA/Nami- bia 32pp, 184 maps. Haacke, W. D. (1964): Description of two new species of lizards and notes on Fitzsimonsia brevipes (FitzSimons) from the central Namib Desert. — Sci. Pap. Namib Desert Res. Station*23 11 Op Haacke, W. D. (1965): Additional notes on the herpetology of South West Africa with descriptions of two new subspecies of geckos. — Cimbebasia 11: 1—40, 5 figs, 7 pls. Laurent, R. F. (1964): Reptiles et amphibiens de l’Angola (Troisieme contribution). — Diamang. Publ. Cult. 67, 165 pp., 40 figs. McLachlan, G. R. (1984): The type locality of Typhlacontias brevipes. — Journ. Herp. Ass. Africa 30: 17. : Mertens, R. (1955): Die Amphibien und Reptilien Stidwestafrikas. — Abh. senckenb. naturf. Ges. 490: 1—172, 24 pl.. Mertens,R. (1971): Die Herpetofauna Südwest-Afrikas. — Abh. senckenb. naturf. Ges. 529: ¡M0! Monard, A. (1937): Contribution a l'herpetologie d’Angola. — Arquivos Museu Bocage 8: 19—154, 3 figs. Pitman,C.R.S. (1934): A checklist of reptilia and amphibia occurring and believed to occur in Northern Rhodesia. — In: Pitman, C. R. S., Report on a Faunal Survey of Northern Rhodesia, Government Printer, Livingstone. Roux, J. (1907): Sur quelques reptiles sud-africains. — Rev. suisse Zool. 15: 77—86. Revision of the scincine genus 7yphlacontias 163 Sclater, W. L. (1891): List of reptiles and batrachians of South Africa, with descriptions of new species. — Ann. S. Afr. Mus. 1: 95—111, 1 pl. Swofford, D. L. (1985): PAUP — Phylogenetic Analysis Using Parsimony. Version 2.4.1. User's manual. — Privately published. Welch, K. R. G. (1982): Herpetology of Africa: a checklist and bibliography of the orders Amphisbaenia, Sauria and Serpentes. 293 pp. — Krieger Publ. Co., Malabar, Florida. Wulf D. Haacke, Curator, Dept. of Herpetology, Transvaal Museum, P.O. Box 413, Pretoria, 0001 Republic of South Africa. hos A eal pla ir ; In mir a Lu Mc i ada | aly jit N | | TE PN 7 ante Zu mir ‘leon DP, in AS ES NE yy 7 Sad Viale | angel wy IE ore u? Ma re) tag. ara a WER UA ae I“ epilh = ar sc MAD Al Tr, ET ur er on fach ont Os ¢ i A N 5 300 ME Mi! A _=- MT Y - b ‘ N a a: t Wie» ag a | 1 A e ES i » y f 1 N y T | ] J 3 J | | J a 5 & z T N | pP [2 7 ' |] = % ” o ® A UT e f i - E Hi e 2 2 + 14 A N y “ LG f ae N ps y E L E si ES ip % | q t bod ; N oan u : Bd. 47 S. 165—174 | Bonn, September 1997 Advertisement call characteristics of a Taiwanese green treefrog Rhacophorus prasinatus Masafumi Matsui, Szu-Lung Chen & Kuang-Yang Lue Abstract. Variations in parameters of three types of notes were analyzed in Taiwanese treefrog, Rhacophorus prasinatus. In contrast to the previous knowledge, the type B notes are shown to be well pulsed. Frequency parameters show variation, and are not correlated to temperature. Note duration is predicted to be shortened, and note repetition rate to increase under higher temperatures. From the advertisement call characteristics, R. prasi- natus is supposed to bridge the gap between R. viridis on the one hand, and R. owstoni on R. moltrechti on the other. Key words. Vocal repertoire, advertisement call, temperature. Introduction Among green treefrogs occurring in Taiwan, Rhacophorus prasinatus (Mou, Risch & Lue 1983) is peculiarly big-growing (mean SVL 56 mm in males and 72 mm in females: Chen 1992) as compared with its relatives, and looks very attractive, being emerald green and golden in the body colour. This species has a complicate nomenclatural history, and is better known by its synonym R. smaragdinus (Lue & Mou 1983; see Zhao € Adler 1993: 158 for a taxonomic review). Its distribution range is restricted to a limited area of northern Taiwan (Lue et al. 1991). Call characteristics of this species recorded at 27—29°C have been reported by Chou (1992). As is well known, however, some of the acoustic parameters of anurans vary with changes of the surrounding temperatures (e. g., Nevo & Schneider 1976, Schneider 1977, Zweifel 1968). Rhacophorus prasinatus is a prolonged breeder and the temperatures surrounding calling males vary greatly from about 13 to 29°C (Chen 1992). Thus, Chou’s (1992) recordings were made at extremely high temperatures, and additional information is needed to better understand the call characteristics of this species. In this article, we will report on the acoustic characteristics of R. prasinatus recorded at temperatures lower than those reported by Chou (1992), and clarify the more detailed structure of some calls. Also, relationships of temperature to call parameters are analyzed in this species. Materials and methods Calls of R. prasinatus were recorded at Mt. Tatung Shan, Feitsui valley, Taipei Hsien on 12, -14, and 24 December 1991. All calls were recorded in the field with cassette tape recorders (Sony TCD5 & TCD5M) with external microphones (Sony ECM-23F & Audio-technica ATM 53) and analyzed using computer programs, SoundEdit Vers. 2 or SoundEdit Pro (Macro- Mind-Paracomp, Inc.) by a Macintosh computer. Air temperatures recorded near the calling males, were 18.2°C, 15.0°C, and 19.0°C, on 12, 14, and 24 December 1991, respectively. In the following description, the term “note” means a pulse group, “note length” the time from the beginning of the first pulse to the end of the last pulse in a note, and “pulse repetition 166 M. Matsui et al. rate” the number of pulses per s. Sample size (n) is indicated as the number of calls or notes analyzed/the number of males recorded, and means are given for the number of calls recorded. For statistic analyses, Wilcoxon rank sum tests were utilized. In order to examine relationships between parameters, analysis-of-covariance (ANCOVA) was performed. The significance level was set at 0.05. Results From the field observation of breeding behavior, advertisement, courtship, en- counter, release, and distress calls (Duellman & Trueb 1986) were differentiated in the vocal repertoire of male R. prasinatus (Chen 1992). Of these, the advertisement call included three types of notes (fig. 1), which assuredly correspond to the types, A, B, and C, of Chou (1992). In the following descriptions, only these three note types are considered. We could not record type B and C notes at 15.0°C. In the type A note, some of the temporal and frequency characteristics differed between calls recorded at dif- ferent temperatures, but differences were insignificant between calls recorded at 18.2°C and 19.0°C. Thus, the following descriptions and comparisons are made mainly on calls recorded at 15.0°C and 18.2°C. 0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 4724 TIME (S) Fig. 1: A sonagram (top) and a sound wave form. (bottom) of an advertisement call of Rhacophorus prasinatus, including one type A, two type B, and four type C notes. Recorded at 18.2°C. Advertisement calls of Taiwanese green treefrog 167 The type A note (fig. 2) was well pulsed and included five to 15 pulses (table 1). The note length increased with increasing number of pulses, and at 15.0°C, the mean length varied from 0.302 s in the note with seven pulses to 0.613 s in the note with 15 pulses. The pulse repetition rate was almost constant, means varying from 23.04 to 24.70. Each pulse, except for the initial and the final one, had clear harmonics, and a clear frequency modulation was seen within a note (fig. 1). The dominant frequency in the mean of the initial pulse was about 1135—1400 Hz, but it increased to about 1500—1600 Hz in the mean of the climax pulse, and rapidly decreased to about 1100—1300 Hz in the final pulse. In the climax pulse, the second dominant frequency was about 4400—4750 Hz, and seven harmonic bands in total were apparent between 0—11 000 Hz (fig. 2). Average harmonic interval, therefore, was about 1550 Hz, and this value corresponded to the fundamental frequency. Thus, the first dominant frequency was the fundamental and the second corresponded to the third harmonic of the spectrogram. Similar trends were found in the calls recorded at 18°C, but the note lengths were significantly shorter than in 15.0°C (table 1; Wilcoxon rank sum tests: p<0.0001 in the eight pulsed note; p<0.0001 in the nine pulsed note; p<0.0003 in the ten pulsed note). The mean note lengths varied from 0.180 s in the note with five pulsed to 0.395 s in the note with 11 pulses. The pulse repetition rates varied from 27.39 to 32.57, and were much larger than in 15.0°C (Wilcoxon rank sum tests; p<0.0001 in the 0 0.1 0.2 0.3 TIME (S) Fig. 2: A sonagram (top) and a sound wave form (bottom) of a type A note of Rhacophorus prasinatus. Recorded at 18.2 *C. M. Matsui et al. Note length (s) 0.302 + 0.020 3/2 0.348 + 0.010 7/4 0.383 + 0.015 13/5 0.425+0.013 8/4 0.446 2/1 0.562 2 0.598 1/1 0.613 A 0.1804 0.004 3/3 0.219+0.007 8/5 0.235+0.028 16/7 0.273+0.031 26/13 0.314+0.043 16/11 0.332+0.086 12/8 0.395+0.040 9/6 Pulse repetition rate 23.244 1.56 3/2 23.04+0.66 7/4 23.43+0.86 11373 23-55 0870 8/4 24.70 2/1 2318 2/1 23.40 17/41 24.49 1/1 27.822058 3/3 27.394 0.94 8/5 30.385 4.96 16/7 29.861 4.57 26/13 SD) 16/11 325710116 12/8 28.11+3.08 9/6 Table 1: Characteristics of type A note of Rhacophorus prasinatus recorded at 15.0 and 18.2°C (Mean#ISD, followed by sample size (number of notes/number of males)). Dominant frequency in climax pulse (Hz) 1500.0+70.7 U 1557.1#34.5 7/4 1584.6+37.6 1375 1556.3+49.6 8/4 1525.0 2/1 1575.0 2/1 1600.0 Al 1500.0 174 1340.04 52.9 3/8 1325.75 64.8 TS 1420.04 154.3 16/7 AS ISE 2318 15225256320 12/11 1486.7+51.9 12/8 1508.9=3-2741 9/6 eight pulsed note; p<0.0001 in the nine pulsed note; p<0.0003 in the ten pulsed note). Some notes had much larger rates than in others, and these rapid notes were emitted by one individual immediately after another’s advertisement call. The mean dominant frequency increased from 938—1027 Hz in the initial pulse to 1325—1522 Hz in the climax pulse, and decreased finally to about 1210—1436 Hz. The dominant frequencies in the climax pulses were significantly lower in calls recorded at 18.2 °C than in 15.0°C (Wilcoxon rank sum tests: p=0.019 in the eight pulsed note; p=0.005 in the nine pulsed note; p=0.020 in the ten pulsed note). Advertisement calls of Taiwanese green treefrog 169 FREQUENCY (KHZ) > o N 0 0.1 0.2 0.3 0.4 TIME (S) Fig. 3: A sonagram (top) and a sound wave form (bottom) of a part of an advertisement call of Rhacophorus prasinatus, showing a type B (eight-pulsed) and a type C notes. Recorded at 18.2 *C. The type B note (recorded at 18.2°C) was also pulsed (fig. 3), and the note length varied from 0.014 s in the note with one pulse to 0.340 s in the note with 14 pulses (table 2). When pairs of notes with the same number of pulses were compared, note lengths in the type A notes were always longer than in the type B (Wilcoxon rank sum tests: p<0.03), in all the combinations except for the notes with ten pulses. The pulse repetition rate varied from 32.83 to 53.48, but in most cases larger than 40. Some notes with a small pulse repetition rate included pulses with irregular intervals. Like in the note length, the type A notes had larger rates than the type B in all the combinations except for the one with ten pulses (Wilcoxon rank sum tests: p<0.03). The dominant frequency varied from about 950 to 1100 Hz, and was much lower than in the type A note (Wilcoxon rank sum tests: p<0.002, in all the combina- tions). In the type C note (recorded at 18.2°C), the note length varied from 0.004 s in the note with one pulse to 1.170 s in the note with ten pulses (table 3). The mean pulse repetition rate decreased from 19.31 to 8.54 with the increment of the pulse number. The dominant frequency was similar to that in the type B (fig. 3), and varied from ‚about 950 to 1200 Hz. As seen in fig. 4, the note duration and the pulse number had positive correlations in all the three note types. Both slopes and positions (table 4) differed among the regression lines (ANCOVA, p<0.0001), and from the multiple comparisons among slopes, it was concluded that the slope decreased in the order of type B, type A (15°C), type A (18.2°C) and type C. In the type A notes, the calculated slope was 170 M. Matsui et al. Table 2: Characteristics of type B note of Rhacophorus prasinatus recorded at 18.2°C (Mean+1SD, followed by sample size (number of notes/number of males)). Note Pulse Dominant N of pulse length repetition frequency (s) rate (Hz) 0.014 — 1050 2/1 — 2/1 0.037 53.48 950 1/1 1/1 17/41 0.069 + 0.006 43.955 4.07 * 1025.0+ 50.0 4/2 4/2 4/2 0.085 47.17 1300 1/1 il Al 0.124+0.012 40.514+3.82 1026.7+40.4 3/3 3/8 3/3 0.1894 0.035 32.832023 1004.3+36.0 7/4 7/4 7/4 0.167+0.024 42.471 4.69 1041.0+71.9 10/5 LOS 10/5 0.202+0.037 40.76+6.42 1060.0+63.2 16/7 16/7 15/7 0.224+0.022 40.47+3.84 10731221038 (393 1578 14/5 0.239+0.019 42.16+3.10 103191222829 12/9 1279 9/9 0.258+0.008 42.65+1.29 1058.31 49.2 8/5 8/5 6/5 0.317+0.019 41.07+2.47 1100.0+180.3 3/1 3/1 Al 0.340+0.028 41.38+3.33 1033 EE oven 3/1 Sal Sal smaller in 18.2°C than 15°C, but the position of the regression line was higher in the former, indicating larger pulse repetition rate. Discussion Acoustic features of R. prasinatus reported here are basically identical to those reported by Chou (1992 as R. smaragdinus), but there is at least one great discord- ance. Chou (1992) reported the note type B to be nonpulsed, but actually it is pulsed as clearly seen in fig. 1. This difference seems to have derived from the difference in the quality of the recording or method of analyses. As shown in the result, call parameters vary in response to variant temperatures, and this becomes more evident when the present data are compared with those given by Chou (1992). Since Chou (1992) did not give relationships between the number of pulse and other parameters, it is impossible to compare his data directly with ours. However, from his data of pulse repetition rate, some parameters in notes with nine and 11 pulses can be estimated. When these data are considered, it is clear that the Advertisement calls of Taiwanese green treefrog 171 N OF PULSE 0 0.2 0.4 0.6 0.8 1 1.2 NOTE DURATION (S) Fig. 4: The relationship between the note duration (in s) and the number of pulse in the three types of notes of Rhacophorus prasinatus. Closed circle=note type A at 15.0°C (Y=23.834X —0.152, r=0.981, n=36 notes/5 males, p<0.0001); open circle=note type A at 18.2°C (Y=17.932X + 3.158, r=0.780, n=89 notes/13 males, p<0.0001); closed triangle=note type B (Y=36.284X +0.846, r=0.922, n=85 notes/9 males, p<0.0001); open triangle=note type C (Y=7.434X +1.249, r=0.984, n=11 notes/8 males, p<0.0001). note decreases in its duration with the increase of the temperature (fig. 5). The relationships of duration (Y) to temperature (X) are expressed as Y=0.0136X +0.577 (r=—0.990, p=0.009) and Y= —0.0160X + 0.688 (r= —0.999, p=0.005) in the notes with nine and 11 pulses, respectively. The regression line for the 11 pulsed note is significantly higher in position than that for the nine pulsed one (ANCOVA: Table 2: Characteristics of type C note of Rhacophorus prasinatus recorded at 18.2°C (Mean=+1SD, followed by sample size (number of notes/number of males)). Note Pulse Dominant N of pulse length repetition frequency (s) rate (Hz) 0.004 — 1200 1/1 — 11/41 0.115+0.042 19.31+7.08 1100.0+204.1 4/2 4/2 4/2 0.247 12418 1000 1/1 1/1 1/1 0.314 14.26 950 2/2 2/2 D2, 0.511 979 1050 2/1 2/1 2/1 1.170 8.54 1100 IVA 1/1 1/1 1072 M. Matsui et al. NOTE DURATION (S) "Aa 16. “48 20. 221124 D6 "DEN AIR TEMPERATURE ( C) Fig. 5: The relationship between the air temperature (°C) and the note duration (in s) in the type A notes of Rhacophorus prasinatus. Data for 27°C and 28°C calculated from Chou (1992). Closed circle=note with 11 pulses (Y= —0.016X +0.688, r= —0.999, n=3, p=0.005); open circle =note with nine pulses (Y=—0.014X+0.577, r=—0.990, n=4, p=0.009). E o gl © > © oo © PULSE REPETITION RATE N w a al RO © POETA aC MSIO AIR TEMPERATURE ( C) Fig. 6: The relationship between the air temperature (°C) and the pulse repetition rate in the type A notes of Rhacophorus prasinatus. Data for 27°C and 28°C calculated from Chou (1992). Closed circle=note with 11 pulses (Y=1.529X+1.116, r=0.997, n=3, p=0.047); open circle=note with nine pulses (Y=1.622X—0.581, r=0.993, n=4, p=0.007). p=0.005), though the slopes did not differ between the two lines (ANCOVA: p> 0.05). Pulse repetition rate also correlates to the temperature, and the number of pulse per unit time is predicted to increase under higher temperature conditions (fig. 6). The relationships of repetition rate (Y) to temperature (X) in the notes with nine and 11 pulses are expressed as Y=1.6216X—0.581 (r=0.993, p=0.007) and Y=1.5294X +1.116 (r=0.997, p=0.047), respectively. Unlike in the note duration, neither the slope nor position differed in the two regression lines (ANCOVA: Advertisement calls of Taiwanese green treefrog 13 p>0.05), and the combined regression line is expressed as Y=1.5831X+0.156 (r=0.994, p<0.0001). In contrast to these two parameters, frequency parameters (initial, climax, and final ones) are insignificantly correlated to the temperature (ANCOVA: p>0.05). Because of the presence of three different note types, Chou (1992) considered the call of R. prasinatus complicated and unique among the rhacophorid frogs of Taiwan, Ryukyu Is., and Japan. However, Chou (1992) himself noted that other Taiwanese species, R. taipeianus and R. moltrechti also had at least two types of calls. In fact, the vocal repertoire of R. prasinatus is not limited to these three types (Chen 1992), and an even more complicated repertoire is found in R. arboreus from Japan (Kasuya et al. 1992). Thus, the ‘complicated calls’ are not unique to R. prasinatus among the rhacophorid frogs of these regions. Chou (1992) emphasized the complexity of the call in R. prasinatus and associated this species with R. appendiculatus and R. bimaculatus (=R. bipunctatus) from Thailand, both of which have been reported to have complicated calls (Heyer 1971). Actually, call structures of the latter two species are quite dissimilar to those of R. prasinatus, so are their morphologies (Matsui unpublished), and acoustic con- vergences or close phylogenetic relationships in these species are unlikely. Instead, R. prasinatusseems acoustically related to species from Taiwan and Japan. On the basis of acoustic characteristics, Kuramoto & Utsunomiya (1981) suggested the possibility of splitting species of Rhacophorus known from Japan and Taiwan : at that time into three groups: 1) R. taipeianus, 2) R. moltrechti and R. owstonii, and 3) R. arboreus, R. schlegelii and R. viridis. The first group is characterized by a very long trill, whereas the second group by clear harmonics and the presence of two phases that differ in pulse repetition rate. The call of the third group is neither very long nor with different phases. The type A note of R. prasinatus is similar to the note of R. viridis in the temporal and frequency characteristics (see the sonogram in Maeada & Matsui 1989). Therefore, the species would be grouped into the third group if only the type A note is considered. However, in R. prasinatus the type B and C notes usually follow the type A note , and this pattern can be well compared to the complex phases seen in frogs of the second group (R. moltrechti and R. owstonii). Thus, R. prasinatus seems to be acoustically intermediate between the second and the third groups. Acknowledgements We thank Dr. Hans Schneider and an anonymous reviewer for helpful comments on an earlier version of this manuscript. We also deeply appreciate Mr. Hidetoshi Ota for his aid in the field survey. Dr. Tsutomu Hikida kindly helped laboratory work and critically read the manuscript. Part of the work was supported by a grant from U. S. National Geographic Society (No. 4505—91) to M. M. Zusammenfassung Fur 3 Ruftypen des Baumfrosches Rhacophorus prasinatus von der Insel Taiwan wurde die Varianz verschiedener Strukturparameter analysiert. Im Gegensatz zu früheren Befunden erwiesen sich Rufe des Typs B als deutlich gepulst. Frequenzparameter variieren, aber nicht mit der Temperatur korreliert. Die Dauer der Rufe sinkt und ihre Wiederholungsrate steigt mit zunehmender Temperatur. Die Strukturmerkmale der Werberufe von R. prasinatus sind inter- mediär zu denen von R. viridis einerseits und R. owstoni andererseits. 174 M. Matsul et al. References Chen, S.L. (1992): The reproductive behavior and ecology of emerald green tree frog (Rhacophorus smaragdinus). — Master’s thesis, Department of Biology, National Taiwan Normal University, Taipei. Chou, W.-H. (1992): Calls of the emerald treefrog, Rhacophorus smaragdinus, in Taiwan. — Bull. Nat. Mus. Nat. Sci. 3: 241 — 246. Duellman, W.E. & L. Trueb (1986): Biology of Amphibians. — McGrow-Hill, New York. Heyer, W. R. (1971): Mating call of some frogs from Thailand. — Fieldiana: Zool. 58: 615982: Kasuya, E, T. Kumaki € T. Saito (1992): Vocal repertoire of the Japanese treefrog, Rhacophorus arboreus (Anura: Rhacophoridae). — Zool. Sci. 9: 469—473. Kuramoto, M. & T. Utsunomiya (1981): Call structures in two frogs of the genus Rhacophorus from Taiwan, with special reference to the relationships of rhacophorids in Taiwan and the Ryukyu Islands. — Jpn. J. Herpetol. 9: 1—6. Lue, K.Y., C-Y. Lin, K-S. Chuang & J.-S. Lai (1991): Review on the current status of amphibians in Taiwan. — Pp. 173—213 in: Lin, Y.-S. & K.-H. Chang (eds.) Proceedings of the First International Symposium of Wildlife Conservation, Republic of China. Council of Agriculture Forest, Taipei. Lue, K.Y. & Y.-P. Mou (1983): Rhacophorus smaragdinus (Anura: Rhacophoridae) a new rhacophorid tree frog from Taiwan. — J. Taiwan Mus. 36: 15—22. Maeda, N. & M. Matsui (1989): Frogs and Toads of Japan. — Bun-ichi Sogo Shuppan, Tokyo. Mou, Y-P., J. P. Risch € K.Y. Lue (1983): Rhacophorus prasinatus, a new tree frog from Taiwan, China (Amphibia, Anura, Rhacophoridae). — Alytes 2: 154—162. Nevo, E. & H. Schneider (1976): Mating call pattern of green toads in Israel and its ecological correlate. — J. Zool., Lond. 178: 133—145. Schneider, H. (1977): Acoustic behavior and physiology of vocalization in the European tree frog, Hyla arborea (L.). — Pp. 295—335 in: Taylor, D. H. &S. I. Guttman (eds.). The Reproductive Biology of Amphibians. Plenum Press, New York. Zhao, E.-M. € K. Adler (1993): Herpetology of China. — Contrib. Herpetol. 10: 1—522. Zweifel, R. G. (1968): Effects of temperature, body size, and hybridization on mating calls of toads, Bufo a. americanus and Bufo woodhousei fowleri. — Copeia 1968: 269—285. Dr. M. Matsui, Graduate School of Human and Environmental Studies, Kyoto University, Sakyo-ku, Kyoto 606 Japan. Mr. S.-L. Chen and Dr. K.-Y. Lue, Department of Biology, National Taiwan Normal University, 88 Roosevelt Road, Section 5, Taipei, 11718 Taiwan, Republic of China. Bd. 47 S. 175—185 Bonn, September 1997 Description of the advertisement calls of some South American Hylidae (Amphibia, Anura): taxonomic and methodological consequences Ignacio De la Riva, Rafael Márquez & Jaime Bosch Abstract. Calling behavior and advertisement call structure of five hylid species from South America are described. Characteristic audiospectrograms and oscillograms are presented, as well as numerical information on spectral and temporal features of the calls of each species. Variation in different call parameters is compared with previously published descriptions of the advertisement calls of these species. Taxonomic and metho- dological issues of the use of bioacoustical data are discussed. The recognition of Ayla walfordi and Scinax parkeri as valid taxa is recommended. Key words. Anura, Hylidae, Hyla albopunctata, H. carnifex, H. granosa, H. walfordi, Scinax quinquefasciata, S. parkeri, Bolivia, Ecuador, advertisement calls. Introduction Anuran vocalizations play a crucial role as mechanisms for species recognition and are essential components of the characterisation of species accounts. In tropical anurans relationships among taxa are particularly complex, and comparisons of advertisement calls represent an alternative way of determining differences and affi- nities based on a quantifiable parameter (e. g. Hödl 1977; Schlüter 1979, 1980, 1981; Duellman & Pyles 1983; Zimmerman 1983; Zimmerman & Hödl 1983; Zimmerman & Bogart 1984). In this paper we contribute to the knowledge of some South American species of hylid frogs by studying the acoustic characteristics of their advertisement calls. This report is a complement to previously published reports of other species of hylids from the Amazonian region (Márquez et al. 1993, De la Riva et al. 1994; 1995) from recordings obtained in Bolivia and Ecuador. The objectives of this paper are to (1) describe the advertisement calls of five species, (2) compare our findings to previous descriptions, (3) discuss taxonomic issues for some of the species studied, and (4) comment about the methodological problems associated with the study of bio- acoustics in anurans. Materials and methods Recordings were obtained in Ecuador and Bolivia. Recording equipment included either a Sony WM D6C or a Sanyo M1120 tape recorder and a Sennheiser Me 80 directional micro- phone. An audiospectrogram and oscillogram were obtained for a 2.5 s recording segment of each species. A longer recording (20—60 s) of a single male was analyzed to obtain numerical information on the spectral and temporal characteristics of the calls. The characteristic call was selected based on quality of the recording and on the subjective criteria of a learned human listener who perceived the call as a “normal” advertisement call of the species (i. e., the call was not emitted in specific situations which could indicate that they would be recipro- cation, release, or distress calls and so on). 176 I. De la Riva et al. Digitalization and editing were completed with a Macintosh-based digital signal analysis system at a sampling frequency of 44.1 kHz and 16 bit resolution with Sound Tools hardware and software. Signalyze software was used to obtain numerical information and to generate audiospectrograms and oscillograms. Information on frequency domain was obtained through fast Fourier transform (FFT) (width 1024 points). Advertisement call terminology follows Heyer et al. (1990). A total of 10 different call characteristics were considered. The variables considered were: number of notes per call, note duration, fundamental frequency, dominant frequency, other frequency with substantial energy, number of pulses per note, pulse rate (pulses/s), pulse ratio (pulse duration/pulse period), note repetition rate (notes per min within the call), and call repetition rate (calls per min). Unless otherwise specified, individual sizes (snout-vent length, SVL) were measured from collected individuals which were deposited in the Museo de Historia Natural “Noel Kempff Mercado”, Santa Cruz de la Sierra, Bolivia. No specimens were collected in Ecuador. Results An audiospectrogram and an oscillogram is presented for a 2.5 sec. section of the call of each species (Figs. 1 to 5). A summary of the numerical information from the sound analyses is shown in Table 1. Hyla albopunctata Spix, 1824 is a medium-sized species [SVL males = 60 mm (Cei 1980] of the A. albopunctata group. It occurs in the cerrado domain of Central Brazil, northeastern Argentina, and eastern Paraguay. Our recordings were obtained from a single male calling from a marshy area at camp Huanchaca I, Noel Kempff Mercado National Park, Santa Cruz Department, Bolivia (13° 54° S / 60° 47 W). This is the first record of the species for Bolivia. The only calls recorded were com- posed of a sequence of three loud, single pulsed notes repeated at regular short inter- Hz. 10000 9000 8000 100 400 700 1000 1300 1600 1900 2200 7000 6000 5000 4000 ’ 3000 Sex 2000 des N 1000 reno eana ss Ye 0 100 400 700 1000 1300 1600 1900 2200 ms. IS; Jl Figs: 1—5: Audiospectrograms and oscillograms of a characteristic advertisement call. Fig. 1 Hyla albopunctata, Fig. 2 Hyla carnifex, Fig. 3 Hyla granosa, Fig. 4 Hyla walfordi, Fig. 5 Scinax quinquefasciata. Advertisement calls of South American Hylidae 10 40 70 100 130 160 190 220 250 100 400 700 1000 1300 1600 1900 2200 PE. b ad a PEA 100 400 700 1000 1300 1600 1900 2200 ms. 100 400 700 1000 1300 1600 1900 2200 ms. 40 50 100 400 700 1000 1300 1600 1900 2200 100 400 700 1000 1300 1600 1900 2200 ms. Ma 178 I. Dela Riva et al. AN GT 140 180 2302603 340 380 Hz. 0000. 0100 2 7 1000 1300 1600 1900 2200 9000 8000 7000 6000 5000 000 cgaestebaes PA 3000 ae mts 2000 IEA II 1000 A PPIIRIIDIA SPRFOPRP ADDED i Peete or NESE Aire rn. EN 100 7 700 1000 1300 1600 1900 2200 ms. FIG. 5 vals (Fig. 1). The structure of the individual pulses was strongly amplitude-modu- lated, showing sub-pulses. The calls showed substantial energy in the 1300—3300 Hz range, with an average dominant frequency of 2692.3 Hz. The dominant frequency increased gradually towards the end of the call (average increase, 289.4 Hz, range 201.9— 424.0 Hz). Hyla carnifex Duellman, 1969 is a relatively small species [mean SVL males = 26.1, Duellman (1969)] in the H. columbiana group that occurs in the Pacific Andean slopes of Colombia and Ecuador. Our recordings were obtained in Las Palmeras, Pichincha Province, Ecuador (00 °17’ S / 78° 45’ W) where males called at night on the vegetation near a river bank, concurrently with males of Centrolene ballux. The advertisement call consisted of two note types: a long note (265 ms duration) fol- lowed by a sequence of three short notes (34 ms duration). The long note is frequency- modulated and tonal in its first half, becoming more pulsed towards the end. The short notes are more pulsed (3 pulses per note), showing some power in the second harmonic (Fig. 2). The dominant frequency of both note types is concentrated around 2400 Hz. Hyla granosa Boulenger, 1882, is a medium-sized species (males, 54 mm SVL) of widespread occurrence in the Upper Amazon Basin and Guianas and belongs to the Hyla punctata group. It has been observed (in Ecuador, Peru and Bolivia), calling from large leaves (as those of the “elephant ear” plants, Heliconia, etc) near or above ponds. In two of the three localities where the species was observed, it was calling in association with Ayla lanciformis. The recording shown in Fig. 3 was obtained near the River Napo, Ecuador (00° 25” S / 77° 47’ W). The call consisted of a series of two to seven short notes of decreasing intensity. The notes were tonal and well- tuned, with a mean dominant frequency of 1327 Hz and substantial power in the second harmonic (2650 Hz). The mean duration of the notes was 46 ms. Hyla walfordi Bokermann, 1962 is a small species (SVL males = 23 mm) of the H. microcephala group. Recordings were obtained in a flooded, open area at Flor de Oro, on the Bolivian bank of the Guaporé river (13° 33’ S / 6° 00° W). This is the Advertisement calls of South American Hylidae 179 first record of the species for the country, although its occurrence in Bolivia was already suspected (Langone & Basso 1987; De la Riva 1990). The only other species heard calling concomitantly was Scinax parkeri. The call is an extremely brief note (average duration, 38 ms) with most of its energy in the 3400—4400 Hz range. An average of five pulses (range 4—7) were discernible in the structure of the note (Fig. 4). Scinax quinquefasciata (Fowler, 1913) is a relatively small species [35 mm SVL in both sexes (Duellman 1971)] of the S. rubra group. It occurs in the Pacific lowlands of Colombia and Ecuador. Recordings were obtained in Same, Esmeraldas Province, Ecuador (00° 50’ N / 79° 55’ W) where males called from a grassy area near a small stream in the mangrove. Trachycephalus jordani and Bufo marinus were the only other anuran species that were found there. The call is composed of a single note, 472.5 ms average duration, with a variable number of pulses (average 16 pulses/note) and a complex spectral structure (range 1400—4000 Hz) with a dominant frequency of 3446.2 Hz (Fig. 5). The call was repeated at a relatively fast rate (average 29.2 calls/minute). Discussion Cardoso (1981) analyzed calls of H. albopunctata from Campinas, Brazil. He distin- guished a nuptial call and a territorial call (both considered, in a broad sense, as “advertisement calls”). He found that the nuptial call has a higher number of pulses and is slightly higher pitched than the territorial call, and that both calls consist of two notes, sometimes followed by a third, different note. Aside from a high number of pulses (118) in the first note of the nuptial call (that we considered equivalent to our subpulses) and a longer note duration (600 ms) in the specimens from Campinas, our results are in general agreement with Cardoso (1981). Haddad et al. (1988) described the calls of A. albopunctata from Serra da Canastra, Minas Gerais, Brazil. They described a call typically consisting of a series of two pulsed notes separated by an interval of 200 ms, the first one having a duration of 500 ms and the second one being longer (900 ms). Occasionally they perceived a third note which was shorter but louder. They determined through playback tests that the third note found to be particularly effective in eliciting calling by other males. Our recordings do not coincide with their description since only one type of note was heard in Bolivia. Heyer et al. (1990) described the call of H. albopunctata from Boracéia (Brazil). Their description does not coincide with ours to the extent that they described two different note types. Their type I note was not heard in the Bolivian specimens. However, our description vaguely resembles their type II note in duration, frequency range, and structure (strong side bands) but differs in that they do not find frequency modulation and we find that the upper emphasized harmonic shows an upwards frequency sweep (2800—3100 Hz). Differences in all these results appear to be due to the different social context in which frogs were at the time of recording. Duellman & Trueb (1983) provided an accurate description of the call of H. carni- Jex from a locality close to ours. Although they found a larger number of calls per minute than us, the rest of the data are coincident. l. 811978609 TPELpEET6EE TILIZ8DEN SLLS—E891 (Er) 78019 THE (971) 9ST (y TI SSE L0-S0 (10) 90 6 sep vé 079 (FO) TSE (KO 811 | (8) TR EOP—P'L] (8:6) TE pjmosofonbumb 'S SE LSIELON GIOPP—PLSE L'E8Sp—S'8E0p L'E8SH—S8E0P 9Y0—yS7 (ED) SEI (LS EISE (EBD TH (EHBI) TE (99) 88 MIE Tor VU€p (70) 78 60—E0 (10) LO Iplofjom “Y (6'L767—8'SOLT) RV ESPI67SEL 8'ESPI675El (TLS) 7198 (y9Z) 697pT IT) ET Sonn SOLD THPSC—TEChe THvst1'EZpT (S'6L9—9'187) (SL9) 6SIP 8 Lp—L8 (200) SIT acid (200) 0 DSOUDIS “HT L'S6—8'S8 70189 Vr —67 Dela Riva et a I 180 num Al) (6p) 116 C1048 (16) £46 NUI /S2JON, 60-70 (10) 90 porad aspnd /uoneinp asjng (8'p1) L'68 v'617—9'607 (69) SAT 87817 (11) 897 puooas /S9S]ng £ 3 (0) LS—98 (LO) 59 {1-7 (90) EI ou /sasind (EB) 19.8» 861-1868] (9:87) €'8161 (L'6£) FORT E tpre—T EChC (EHI) TEE FSOLTTI SENT (L'I1) £2697 (zH) Aouanboyy (zH) Aouanbayy JUEUTUOT (L’6E) Y'06p7 E EPT EH (EHI) TEEN 6 (SEIS TIE (TOT) LTEEI (ZH) Aouanboyy [ejuowepuny "SUOTIEZIIEIOA JY} JO SidjoweIed peorIoumu Jo Alewunms :] 9[qGeL (€°9) EHE 6 1L0—C SSC (811) SEN Frans (E67) #19 (09s) uoneinp 910N [eo sod SOJON pozÁyeue SOON pazAjeue SJenprarpur nous 3uU07 X2/1UIDI "A ppjoundogjv 'H Advertisement calls of South American Hylidae 181 Duellman (1978) and Duellman & Pyles (1983) described the call of H. granosa from Santa Cecilia, Ecuador, a locality close to ours. The three sets of data are consistent. Our audiospectrogram of the characteristic advertisement call of H. granosa (Fig. 3) also closely resembles the description from Surinam published by Hoogmoed (1979). The audiospectrogram in Schlüter (1979) shows great similarity to ours as well, although we do not find the upwards frequency modulation shown in the second harmonic found in Peruvian frogs. Both our audiospectrogram and the numerical data obtained for the sound samples are comparable to the calls described by Zimmerman & Bogart (1984) for frogs recorded in Tapajós (Brazil). Cardoso & Vielliard (1990) also described the call of Brazilian specimens of H. granosa. Their data coincide with ours except for the characteristic three or four notes per call that we found. Their specimens emitted one or two notes per call. The comparison of the calls of H. walfordi and H. nana Boulenger, 1889 can provide some insight in the taxonomic validity of the former species. The taxonomic status of H. walfordi has been controversial. This species was placed in the synonymy of H. nana by Lutz (1973), which was followed by Duellman (1977) and Frost (1985), but not by Heyer (1976; 1977). Langone Basso (1987) resurrected H. walfordi based on comparisons of preserved specimens and supposed differences in the mating call [Cardoso, in Langone Basso 1987)]. However, Duellman (1993) did not include the species in his additions and corrections to Frost's (1985) list. Living specimens of this species can be differenciated from those of H. nana by the shape of the snout and some features of the color pattern (Langone & Basso 1987; De la Riva pers. obs.). Furthermore, if we compare the calls of H. walfordi with those of A. nana obtained with similar analysis and recording equipment (Märquez et al. 1993: 437, 438), substantial differences can be found in call structure, since A. walfordi has a markedly pulsed call structure which is much less evident in H. nana. The call of this species is much more tuned and therefore sounds much more tonal than the calls of H. walfordi. Based on these observations, we support the recognition of H. walfordi as a valid species. Langone é Basso (1987) suggested that the taxonomic status of the small yellow frogs commonly ascribed to A. nana in the central and northern Amazon basin should be reconsidered. Hódl (1977) provided analyses of the call of a so-called Ayla cf. nana frog from Manaus, which supposedly could be actually H. walfordi. Comparisons of his data with ours do not support this hypothesis. The number of calls/minute (159) in H. cf. nana is by far higher than in H. walfordi (x = 8.4). The mean dominant frequency is also a little higher (4835 Hz vs. 4282) and the note duration is much shorter (12 ms vs. 38.3). In fact, the call of H. cf. nana seems to be more similar to that of H. nana, although H. nana shows a lower domi- nant frequency (4526 Hz), a longer note duration (38.3 ms) and a lower number of calls/minute (103) (Márquez et al. 1993). Thus, from these data, no definite conclu- sion seems evident regarding the taxonomic status of A. cf. nana. It seems plausible that A. cf. nana is really A. minima Ahl, 1933, or an undescribed species in the H. microcephala group (see Zimmermann & Rodrigues 1990). Finally, Langone & Basso (1987) stated that the characteristics of the calls of H. minuscula Rivero, 1971 from Venezuela provided by Rivero (1971) are not consistent with those of the same species from Belém, Brasil, recorded by Duellman & Pyles (1983), and Langone « Basso (1987) recommended a reassessment of the taxonomic status of the Brazilian 182 I. De la Riva et al. population. Langone (pers. comm.) suggested that it could be A. walfordi. However, direct comparisons of Venezuelan specimens of A. minuscula in the collection of the Museum of Natural History, The University of Kansas (KU 167131-43) with those from Belém (KU 127646-745) indicate that they are conspecific. Thus, the occurrence of A. walfordi in Belém is not supported by the available data. Duellman (1971) described the calls of specimens of S. quinquefasciata recorded in Santo Domingo de los Colorados, Ecuador, a locality about 150 km southeast (air line) from ours. Although the frequency range and the general structure of the call coincides with our recordings, other characteristics such as the pulse rate [65 pulses / second in Duellman (1971)] remain rather different to our data (35 pulses / second). In addition, the note duration that is represented in the audiospectrogram depicted by Duellman (1971: 216) is about 300 ms in duration, a value that is within the ranges that we find (168—1978 ms) but the range and average duration reported in the text by Duellman (140—150 ms, average 145 ms) is lower than any of the values found by us. In a previous paper (De la Riva et al. 1994) we described the call of Scinax parkeri (Gaige, 1929) from Bolivia, and we stated that no recordings of the species had been published. Actually, we overlooked Duellman & Pyles’s (1983) numerical data from recordings obtained in Belém, although they did not show sonogramas or oscillo- grams. Our results are not consistent with those of Duellman & Pyles. Calls from Bolivian specimens have a higher mean number of notes/minute (151.1 vs. 27.6), are much shorter (185.7 ms vs. 640), have less pulses/s (100.2 vs. 177), and show a lower dominant frequency (2777 Hz vs. 4558). These data seem to belong to different species. Scinax parkeri has been considered by some authors (Lutz 1973; Frost 1985; De la Riva 1990) as a synonym of S. fuscomarginata (A. Lutz, 1925), a species occurring in southeastern Brazil. However, Duellman & Wiens (1992) considered S. parkeri as a valid species, without explanation. There are two other species of the perplexing S. staufferi group described from this part of the Amazon basin, S. goino- rum (Bokermann, 1962) and S. madeirae (Bokermann, 1964) [the later placed also in the synonymy of S. fuscomarginata by Lutz (1973)]. All these species look quite similar in appearance. The call of S. fuscomarginata was described by Cardoso (1981) from specimens recorded in Campinas. He found a single kind of note when indivi- duals were calling alone, and two notes when they were calling in pairs. All the para- meters allowing comparisons from Cardoso (1981) and De la Riva et al. (1994) show strong differences between the calls of S. fuscomarginata and S. parkeri. Because our recordings of S. parkeri were obtained at Buenavista (type locality of the species) while Cardoso's were obtained from the state of Sáo Paulo (type locality of S. fusco- marginata), we conclude that S. parkeri is a valid species. On the other hand, data by Cardoso (1981) are quite similar to those by Duellman « Pyles (1983). A reassessment of the taxonomic status of all the described Amazonian species in the S. staufferi group is needed. As we have seen, the vocalizations of some of the species studied were already described from different localities by other authors. Comparisons with those previous reports sometimes showed concordant results. In other cases, potential inconsistencies were revealed which would suggest lines for further research. The problem is frustrating in some of these cases, because inconsistencies may be due to Advertisement calls of South American Hylidae 183 a vast array of causes. First, slight intraspecific regional variations in the advertise- ment calls may exist. Among other reasons, these variations may be a response to slightly different acoustic niche availability in different communities. The different composition of anuran communities from site to site may drive a particular species to expand or squeeze its acoustic niche, depending on particular acoustic constraints. For example, recordings of Leptodactylus mystaceus from Peru differ markedly from recordings of the same species obtained in Bolivia. This may be due to the presence in Bolivia of a second species with a similar call, L. elenae, or to the fact that two species may actually be involved under the name L. mystaceus (Márquez et al. 1995). While this paper was in press, Heyer et al. (1996, Amphibia-Reptilia 17: 7—31) described Leptodactylus didymus from SE Peru, confirming this hypothesis. Particular conditions at the time of the recording (i.e., temperature), as well as some char- acteristics of the frogs themselves (i.e., size or social context) may also affect the results. In this sense, there are different kinds of calls among those commonly consi- dered as “advertisement calls”. A male can modify some features of the call if there is another male close to it, even though they are not involved in a duet. In such a case, 1t is very common that the “normal” structure of the call undergoes some sort of change. For the researcher not familiar with the vocalizations of a particular species, 1t may be difficult to interpret and distinguish what kind of call is being recorded. Finally, there may be differences in the way in which recordings are obtained, analyzed, and presented, and there are many different kinds of recorders, microphones, vibralizers, hardware, software, and so forth. Thus, it is necessary to be as careful as possible before making taxonomic guesswork derived from these comparisons. Furthermore, we know little about the intraspecific variation of the mating calls in Neotropical anurans. We suggest that unless strong differences are found, no taxonomic implications should be supposed in the absence of another kind of evidence. If material and procedures, as well as the pertinent biological infor- mation, are not thoroughly explained, confusion and erroneous conclusions will result. If we wish to establish useful comparisons between data of recordings by different researchers, a standardization of the methods, variables considered, and terminology employed is needed. Acknowledgements The authors are indebted to the Museo de Historia Natural “Noel Kempff Mercado”, of Santa Cruz de la Sierra, Bolivia, for its support. Field work in South America was possible through a grant from the Asociación de Amigos de Doñana to I. De la Riva. Permits to work in Noel Kempff Mercado National Park were kindly provided by Ing. Esteban Cardona. We thank Louis Jammes for helping in many ways in Huanchaca, and Hermes Justiniano for his hospi- tality in Flor de Oro. Erik Wild made useful comments that improved the manuscript. Sound analyses were funded by project CYCIT PB 92-0091 (PI: P. Alberch) Ministerio de Educación y Ciencia (Spain), and were performed in the facilities of the Estación Bio-Geológica El Ventorrillo (C.S.I.C.). We dedicate this paper to A. Cardoso, for his friendships and note- worthy contributions to Neotropical bioacoustics. Zusammenfassung Das Rufverhalten und die Struktur der Werberufe von 5 südamerikanischen Hyliden-Arten werden anhand von Oszillogrammen und Lautspektrogrammen sowie Messungen ihrer Struk- turparameter darin beschrieben. Die Varianz der einzelnen Parameter wird mit bereits dazu 184 I. De la Riva et al. publizierten Befunden verglichen. Taxonomische und methodische Aspekte der Verwendung bioakustischer Daten werden diskutiert. Ayla walfordi und Scinax parkeri sollten als eigen- ständige Arten anerkannt werden. Bibliography Cardoso, A. J. (1981): Organizacäo espacial e temporal na reproducäo e vida larvaria em uma comunidades de hilideos no sudeste do Brasil (Amphibia, Anura). — Master Thesis. Campinas, Brasil. Cardoso, A. J. & J. Vieillard (1990): Vocalizacöes de anfíbios anuros de um ambiente aberto, em Cruzeiro do Sul, estado do Acre. — Rev. Brasil. Biol. 50: 229—242. De la Riva, I. (1990): Lista preliminar comentada de los anfibios de Bolivia con datos sobre su distribución. — Boll. Mus. reg. Sci. nat. Torino 8: 261—319. De la Riva, I., R. Márquez & J. Bosch (1994): Advertisement calls of Bolivian species of Scinax (Amphibia, Anura, Hylidae). — Bijdragen tot de Dierkunde 64 (2): 75—85. De la Riva, I, R. Márquez & J. Bosch (1995): Advertisement calls of eight Bolivian hylids (Amphibia, Anura). — J. Herpetol. 29: 113—118. Duellman, W. E. (1969): A new species of frog in the Ayla parviceps group from Ecuador. — Herpetologica 25 (4): 241 —247. Duellman, W. E. (1971): The identities of some Ecuadorian hylid frogs. — Herpetologica ENE PARIDAS Duellman, W. E. (1977): Liste der Rezenten Amphibien und Reptilien: Hylidae, Centroleni- dae, Pseudidae. — Das Tierreich 95: 1—225. Duellman, W. E. (1978): The biology of an equatorial herpetofauna in Amazonian Ecua- dor. — Univ. Kansas Mus. Nat. Hist. Misc. Publ. 65: 1—352. Duellman, W. E. (1993): Amphibian species of the world: Additions and corrections. — Univ. Kansas Mus. Nat. Hist. Spec. Publ. 21: 1—372. Duellman, W.E. &R.A. Pyles (1983): Acoustic resource partitioning in anuran communi- ties. — Copeia 1983: 639 — 649. Duellman, W.E. &L. Trueb (1983): Frogs of the Hyla columbiana group: Taxonomy and phylogenetic relationships. — In: Rhodin, A. G. J. & K. Miyata: Advances in herpetology and evolutionary biology, 33—51. Harvard University Press, Cambridge. Duellman, W. E. € J. J. Wiens (199): The status of the hylid frog genus Ololygon and the recognition of Scinax Wagler, 1830. — Occ. Pap. Mus. Nat. Hist. Univ. Kansas 151: 123: Frost, D. R. (1985): Amphibian species of the world: A taxonomic and geographical refe- rence. Allen Press & the ASC, Lawrence, Kansas. 732 pp. Haddad,C.F.B.,G. V.Andrade & A. J. Cardoso (1988): Anfibios Anuros no Parque Nacional da Serra da Canastra, Estado de Minas Gerais. — Brasil Florestal 64: 9—20. Heyer, W. R. (1976): Notes on the frog fauna of the Amazon Basin. — Acta Amazonica 6: 369— 378. Heyer, W. R. (1977): Taxonomic notes on frogs form the Madeira and Purus rivers, Brazil. — Pap. Avuls. Zool., S. Paulo 31 (8): 141—162. Heyer, W.R,A.S. Rand, C. A. Goncalves da Cruz, O: L. Peixoto &€3EasNelsen (1990): Frogs of Boraceiá. — Arq. Zool. S. Paulo 31: 237 —410. Hödl, W. (1977): Call differences and calling site segregation in anuran species from Central Amazonian floating meadows. — Oecologia 28: 351—363. Hoogmoed, M. S. (1979): Resurrection of Hyla ornatissima Noble (Amphibia, Hylidae) and remarks on related species of green tree frogs from the Guiana area. Notes on the herpetofauna of Surinam VI. — Zool. Verhand. 172: 3—46. Langone, J. A. & N. G. Basso (1987): Distribución geográfica y sinonimia de Ayla nana Boulenger 1889 y de Hyla sanborni Schmidt, 1944 (Anura Hylidae) y observaciones sobre formas afines. — Com. Zool. Mus. Hist. Nat. Montevideo 164 (11): 1—17. Lutz, B. (1973): Brazilian species of Ayla. University of Texas Press. Austin & London. 260 pp. Márquez, R., Il. Dela Riva « J. Bosch (1993): Advertisement calls of Bolivian species of Ayla. — Biotropica 25: 426—443. Advertisement calls of South American Hylidae 185 Marquez,R., I. Dela Riva « J. Bosch (1995): Advertisement calls of Bolivian Leptodacty- lidae (Amphibia, Anura). — J. Zool. 237: 313— 336. Rivero, J. (1971): Tres nuevos records y una nueva especie de anfibios de Venezuela. — Carnbb J. Sei, IA 2) 159: Schlüter, A. (1979): Bio-akustische Untersuchungen an Hyliden in einem begrenzten Gebiet des tropischen Regenwaldes von Peru. — Salamandra 15: 211—236. Schlüter, A. (1980): Bio-akustische Untersuchungen an Microhyliden in einem begrenzten Gebiet des tropischen Regenwaldes von Peru. — Salamandra 16: 114—131. Schlüter, A. (1981): Bio-akustische Untersuchungen an Bufoniden in einem begrenzten Gebiet des tropischen Regenwaldes von Peru. — Salamandra 17: 99—105. Zimmerman,B.L. (1983): A comparison of structural features of calls of open and forest habitat frog species in the Central Amazon. — Herpetologica 39: 235 — 246. Zimmerman,B.L.&J.P. Bogart (1984): Vocalizations of primary forest frog species in the Central Amazon. — Acta Amazonica 14: 473—519. Zimmerman, B. & W. Hdd! (1983): Distinction of Phrynohyas resinifictrix (Goeldi, 1907) from Phrynohyas venulosa (Laurenti, 1768) based on acoustical and behavioural para- meters. — Zool. Anz. Jena 211: 341—352. Zimmermann, B. & M. T. Rodrigues (1990): Frogs, snakes, and lizards of the IMPA- WWF reserves near Manaus, Brazil. — In: Gentry, A: Four Neotropical rainforests, 426—454. Yale University Press, New Haven. Ignacio De la Riva, Dept. Biodiversidad y Biologia Evolutiva, Museo Nacional de Ciencias Naturales, 28006 Madrid, Spain. — Rafael Marquez, Jaime Bosch, Dept. Ecologia Evolutiva, Museo Nacional de Ciencias Naturales, 28006 Madrid, Spain. e A note on the gender of the genus Podarcis (Sauria: Lacertidae) Wolfgang Bóhme According to article 34 b of the International Code of Zoological Nomenclature (ICZN) species (and subspecies) names “must agree in gender with the generic name with which it is at any time combined; if the termination is incorrect it must be changed accordingly“ (ICZN 1985). This article has quite often been neglected in the taxonomic/nomenclatural practice. Some popular herpetological examples of recent corrections of a long-term wrong usage are Uromastyx (feminine; Agamidae: Lanza 1983), Adelphicos (neuter; Colubridae: LaDuc 1995) or Scinax (masculine; Hylidae: Köhler & Böhme 1996). Arnold (1973) reinstated the generic name Podarcis Wagler, 1830 as a full genus and referred to Fitzinger (1843) as the first author who fixed a type (cited in the original combination: Seps muralis Laurenti, 1768). The species concerned had mostly been mostly regarded as a sub- genus of Lacerta before (for a detailed discussion see Böhme 1986). In his original description, Wagler (1830) assigned three species to his new genus Podarcis, viz. “Seps muralis Laur? (= Podarcis muralis), “Lacerta velox Pall” (= Eremias velox) and “Lacerta grammica Lichtenst” (= Eremias grammica). It was Fitzinger (1843) who subsequently selected and designated Seps muralis Laurenti (being likewise of masculine gender!) as the type species of Podarcis (see also Loveridge 1957, Arnold 1973). When Arnold (1973), elevated Podarcis to full generic rank again, he did not refer to its gender and consequently left the terminations of the species names (which had been adapted to the feminine gender of their former genus Lacerta before) unchanged. The Greek word “podarkis” is an adjective meaning “swift-footed, nimble”. In his original description, Wagler (1830) used the Latin translation “pedibus celer“ to give the etymology of Podarcis. As he generally did not combine his new generic names with a specific type species, but rather cited the assigned species in their original combinations (e. g. Seps muralis, see above), he never made decisions as to the gender of his new generic names. This is also true for Podarcis, a word derived from a Greek adjective. However, according to article 11 g of the Code (ICZN 1985) “a genus-group name must be or be treated as a noun in the nominative singular”, “podarkis” in his latinized spelling Podarcis has to be treated as a noun of masculine gender for the following reasons: — Wagler was surely aware of “podarkis” being an adjective particularly attributed to the ancient hero Achilleus in the classic Greek (Homerian) literature, i.e. a clearly masculine attribution! The three species assigned by him (1830) to his new genus name were (1) Seps muralis (muralis here being masculine, too, because of the gender of Seps!), (2) Lacerta velox, and (3) Lacerta grammica (velox and grammica being feminine due to their combination with Lacerta). Wagler, therefore, did not make a formal decision as to the gender of Podarcis, but by no means considered this name to be feminine! — Fitzinger (1843) was the first to subsequently and explicitly select a type for Podarcis, viz. Seps muralis, which was the masculine one out of Wagler’s three names. His treatment has to be accepted as the establishment of gender. - The consequence of accepting Podarcis as masculine requires to change the endings of those species of Podarcis that have a specific feminine termination. These are: Podarcis hispanica (to P hispanicus), P peloponnesiaca (to P peloponnesiacus), P sicula (to P siculus), P. taurica (to. P tauricus), and P wagleriana (to P waglerianus). In the case of P tiliguerta, however, the situation is different, as filiguerta is not an adjective but an old-Sardian noun meaning lizard (Corti, pers. comm.). Therefore, this termination is not affected by the gender of the genus name. Of the remaining species of Podarcis, of course all feminine subspecific endings 188 W. Böhme have to be changed accordingly. P (Teira) perspicillata is not affected by this change, as it has recently been transferred to a full genus 7eira by Mayer € Bischoff (1996), thus remaining feminine in gender. The elevation of another subgenus of Lacerta, viz. Timon Tschudi, to full generic rank (Mayer & Bischoff 1996) also causes the change from L. lepida to T. lepidus. I thank Dr. Claudia Corti, Firenze, and Dr. Melitta Haller-Probst, Munich, for valuable comments and help. References Arnold, E. N. (1973): Relationships of the Palaearctic lizards assigned to the genera Lacerta, Podarcis, Algyroides and Psammodromus (Reptilia, Lacertidae). — Bull. Brit. Mus. nat. Hist. (Zool.) 25: 289-366. Bóhme, W. (1986): Handbuch der Reptilien und Amphibien Europas Bd. 2/11 (Echsen 3: Podarcis). — Wiesbaden (Aula), 434 S. Fitzinger, L. (1843): Systema reptilium. — Vindobonae (Braumiiller et Seidel), 106 + ix S. International Commission on Zoological Nomenclature (1985): International Code of Zoological Nomenclature, 3rd ed. — London (Int. Trust. Zool. Nom.). Köhler, J. &W. Böhme (1996): Anuran amphibians from the region of Pre-Cambrian rock outcrops (inselbergs) in northeastern Bolivia, with a note on the gender of Scinax Wagler, 1830 (Hylidae). — Rev. fr. Aquariol. 23: 133 —140. LaDuc, T. J. (1995): The nomenclatural status and gender of Adelphicos. — J. Herpetol. 29: 141-142. Lanza, B. (1983): A list of the Somali amphibians and reptiles. — Monit. zool. Ital. 1983: 1937247: Loveridge, A. (1957): Check list of the reptiles and amphibians of East Africa (Uganda; Kenya; Tanganyika; Zanzibar). — Bull. Mus. Comp. Zool. 117: 153—362 + i-xxxvi. Mayer, W. & W. Bischoff (1996): Beiträge zur taxonomischen Revision der Gattung Lacerta (Reptilia: Lacertidae). Teil 1: Zootoca, Omanosaura, Timon und Teira als eigen-- ständige Gattungen. — Salamandra 32: 163—170. Wasler, J. G. (1830): Natürliches System der Amphibien, mit vorangehender Classification der Säugethiere und Vögel. — München, Stuttgart, Tübingen (Cotta). Prof. Dr. Wolfgang Böhme, Zoologisches Forschungsinstitut und Museum Alexan- der Koenig, Adenauerallee 160, D-53113 Bonn. Bonn. zool. Beitr. S.189—190 Bonn, September 1997 Zum Gedenken an Hans Kumerloeve C. M. Naumann Am 11. 8. 1995 verstarb in München nach dreimonatiger schwerer Krank- heit Museumsdirektor i.R. Dr. Hans Kumerloeve, bis dahin und seit 1970 einziger ehrenamtlicher wissenschaft- licher Mitarbeiter des Museums Koe- nig. Kumerloeve war einer derjenigen biogeographisch und systematisch ar- beitenden Wirbeltierkundler, die be- reits zu Lebzeiten zu einer Legende ihrer selbst geworden sind. Am 5. Sep- tember 1903 in Leipzig geboren, hatte er das große Glück, schon im Eltern- haus viel Verständnis und Unterstüt- zung für seine naturwissenschaftli- chen Neigungen zu finden. So ergab es sich fast von selbst, daß er nach dem Abitur an der Leipziger Universität ein Studium der Naturwissenschaften aufnahm. Dort traf er bald mit einem anderen sächsischen Biologen zusam- men, dessen Herz ebenso wie das von Hans Kumerloeve der Ornithologie gehörte: Günther Niethammer, dem späteren langjährigen Leiter der Orni- thologischen Abteilung des Museums Koenig. Kumerloeve gelang es, auch Niethammer dazu zu bewegen, eine Doktorarbeit bei dem damaligen Inha- ber des Lehrstuhls für Zoologie in Leipzig, Prof. Dr. Johannes Meisen- heimer, zu beginnen. Nach der erfolgreichen Promotion „Vergleichende Untersuchungen über das Gonadensystem weiblicher Vögel“ (3 Bände), Note: summa cum laude, schlug Kummer- löwe (später: Kumerloeve) ein Angebot Meisenheimers, bei ihm als Assistent zu arbeiten, aus und folgte seinen eigenen wissenschaftlichen Interessen mit einer von der Notgemeinschaft der deutschen Wissenschaften finanzierten Studie zur Vogelfauna des Alpenraumes, in deren Ver- lauf er auch längere Zeit an den reichen Sammlungen des Wiener Naturhistorischen Museums arbeiten konnte. Den Grundstein zu seinem wissenschaftlichen Hauptarbeitsgebiet aber legte eine im Jahre 1933 auf einem kleinen Motorrad gemeinsam mit Günther Niethammer durchgeführte orni- thologische Studienreise in die nördliche und westliche Türkei. Noch im gleichen Jahr waren die beiden Zoologen schon wieder in der Türkei, um sich dort Eindrücke vom herbstlichen Vogelzug zu verschaffen. Bereits aus diesen beiden ersten Reisen resultierten mehrere grund- legende faunistisch-zoogeographische Arbeiten. Zunächst allerdings übernahm Kumerloeve im Jahre 1936 die Leitung der Staatlichen Museen für Tier- und Völkerkunde in Dresden, um bereits 1939 nach Wien zu wechseln, wo er zum Ersten Direktor der Wissenschaftlichen Staats- museen ernannt worden war. Seine Beziehungen zum Gründer unseres Instituts sind zumin- 190 C. M. Naumann dest bis in die Zeit vor dem Amtsantritt in Dresden nachweisbar, da Alexander Koenig ihm in einem persönlich gehaltenen Handschreiben, das sicherlich auf älteren Kontakten aufbaut, zu seiner Ernennung in Dresden gratulierte. Nach dem zweiten Weltkrieg sehen wir Hans Kumerloeve als Museumsdirektor i.R. zuerst in Osnabrück und ab 1964 in Gräfeling bei München. Seit Beginn der 50er Jahre fand er mit unermüdlichem Eifer zu seinem eigentlichen Arbeitsgebiet, der Erforschung der Vogel- und Säugetierfauna des Vorderen und Mittleren Orients. Für Reisen erhielt er mehrfach die Unter- stützung der Deutschen Forschungsgemeinschaft, die auch die Drucklegung seiner „Bibliogra- phie Säugetiere und Vögel der Türkei“ (Bonn. zool. Monogr. 21, 1986) mitfinanzierte. Die Auswertung seiner reichen Sammel- und Beobachtungstätigkeit führte Hans Kumerloeve regelmäßig zu kürzeren, in den 50er bis 70er Jahren auch längeren, manchmal mehrwöchigen Aufenthalten an das Museum Koenig, in dessen Obhut auch die umfangreichen ornithologi- schen und theriologischen Sammelausbeuten gelangten. Hier war er Dauergast und konnte 1967 auch einen Mitarbeiter des Museums, Heiner Mittendorf (heute: Klein Windhoek/Nami- bia), als Begleiter und Assistenten seiner Reisen gewinnen. Seine unkonventionelle, ja vielfach burschikose Umgangsweise und seine fundierten geographischen, biologischen und histori- schen Kenntnisse des Vorderen Orients brachten Hans Kumerloeve großen Respekt ein. Hier- bei spielte auch sein besonderes Engagement für den Erhalt der Tier- und Pflanzenwelt dieses Raumes eine besondere Rolle. Seine Bemühungen um den Erhalt der türkischen Populationen des Waldrapps am Euphratufer bei Birecik haben ihn weit über die Grenzen Deutschlands hin- aus bekannt gemacht. Um so bedauerlicher ist es, daß es trotz intensiver Unterstützung durch türkische Politiker und Wissenschaftler nicht gelang, diese letzte große Waldrapp-Population Asiens zu erhalten. Überhaupt war Hans Kumerloeve stets bestrebt, seine Erkenntnisse nicht nur in europäischen Fachzeitschriften zu publizieren, sondern sie auch in der Türkei als seinem ständigen „Gastland“ verfügbar zu machen. Somit pflegte er eine rege Korrespondenz mit tür- kischen Fachkollegen und publizierte auch in türkischer Sprache, so unter anderem in den Mit- teilungen der naturwissenschaftlichen Fakultät der Universität Istanbul (Istanbul Universitesi Fen Fakültesi Mecmuası). Sein wissenschaftliches Gesamtwerk umfaßt rund 400 Publika- tionen. Hans Kumerloeves wissenschaftliche Arbeiten über die Türkei fanden weit über die Grenzen der Bundesrepublik hinaus Beachtung und brachten ihn in Kontakt mit Wissenschaftlern in nahezu allen Erdteilen. Sie legten wichtige Grundsteine für das spätere Engagement vor allem britischer und niederländischer Ornithologen, das noch im Jahre 1995 mit der Publikation der “Songbirds of Turkey — an Atlas of Biodiversity of Turkish Passerine Birds” durch C. S. Rose- laar einen vorläufigen Höhepunkt erreichte. Der Verfasser hat dieses wichtige Grundlagen- werk Dr. Hans Kumerloeve in Anerkennung seiner Leistungen für die Erforschung der türki- schen Vogelwelt und für den Naturschutz im Mittleren Orient gewidmet. Leider hat Hans Kumerloeve das Erscheinen dieses Werkes nicht mehr miterleben dürfen. Es wäre ihm sicher- lich eine tiefe Genugtuung gewesen zu sehen, daß seine Bemühungen nun von anderen weiter- getragen und verfolgt werden. Noch zu Beginn der 90er Jahre hatten wir mehrfach die Ehre und das Vergnügen, Hans Kumerloeve unter uns im Museum Koenig zu sehen. Oft wurde er dabei von seiner Frau Gertraude begleitet, die ihn im übrigen — wie zahlreiche Danksagungen in den Veröffent- lichungen belegen — auf den meisten der Expeditionen begleitet hat, tatkräftig präparierend und die technischen Dinge des Reisealltags in schwierigem Gelände meisternd. Im Zusammen- hang mit dem Ausbau unseres historischen Archivs unterstützte er uns durch die Überlassung von Briefen Alexander Koenigs und anderen Schriftgutes aus der Gründerzeit unseres Insti- tuts. Seinen beschwingten Schritt, seine markante Figur und sein stets humorvolles, der Sache verbundenes Wesen werden wir vermissen. Die Mitarbeiterinnen und Mitarbeiter des Museum Koenig werden Hans Kumerloeve stets ein liebevolles Andenken bewahren. Prof. Dr. C.M. Naumann, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, 53113 Bonn. Bd. 47 S. 191-192 | Bonn, September 1997 Buchbesprechungen Vogelstimmen im Wald. CD Nr. 404.025, AMPLE Edition Musikverlag, Germering. Die Zahl der digitalen Tontráger mit Tierstimmenaufnahmen, die man káuflich erwerben kann, nimmt stándig zu. Die zur Besprechung vorliegende CD gehórt mit weiteren 3 zu einer Serie von 4 CDs mit den Stimmen von insgesamt 100 heimischen Wildvogelarten. Dieselbe Serie ist auch auf MC erhältlich. Ein kleines Beiheft enthält in Stichworten eine sehr kurze Minimalinformation zu jeder der Arten, einen kurzen einführenden Begleittext, bisweilen auch mit Erläuterungen zu der jeweiligen Tonaufnahme, und je ein Photo der 25 Vogelarten, die mit Lautäußerungen auf dieser CD vertreten sind. Einige Arten wie Klappergrasmücke, Grauschnäpper oder Turteltaube, die mit ihren Gesän- gen auch auf dieser CD vertreten sind, kann man aufgrund ihres hauptsächlichen Vorkom- mens sicherlich kaum als ausgesprochene Waldvögel bezeichnen. Die Auswahl der für die ein- zelnen Vogelarten präsentierten Lautformen erscheint bisweilen etwas unausgewogen; so wur- den z. B. für den Habicht nur die Bettellaute von Nestjungen aufgenommen. Wenn diese CD als Einführung für Vogelfreunde ohne viele Vorkenntnisse gedacht ist, so sind die weniger als briefmarkengroßen Photos der einzelnen Vogelarten kaum hilfreich, um diese im Feld sicher zu identifizieren. Die Qualität der einzelnen Tonaufnahmen ist durchweg gut, und sie sind mit meist mehr als 2 Minuten Abspieldauer in der Mehrzahl auch ausreichend lang, man fragt sich - aber, warum der gedruckte Text nochmals auf dem Tonträger in die Vogelstimmenaufnahmen hinein verlesen wird. Das mindert den Eindruck der Vogellaute auf den Hörer. Es wäre sinn- voller gewesen, weitere Vogelarten, die in nahezu allen Aufnahmen im Hintergrund auch zu hören sind — und das ist ja natürlich —, für den Zuhörer zu identifizieren. Vogelstimmen-CDs wie die vorliegende sind sicherlich gut für Personen geeignet, die sich einen ersten Eindruck von den Lautäußerungen unserer heimischen Vögel verschaffen wollen. Wer sich aber eingehender damit befassen will, wird eher auf entsprechende Produkte anderer Hersteller zurückgreifen. G. Peters, Bonn Passarge, H. (1991): Avizönosen in Mitteleuropa. Ber. Bayer. Akad. Natursch. Landsch.- pflege, Beih. 8: 5—128 (diese Serie kürzt sich selbst mit Ber. ANL ab). Avizönosen sind Lebensgemeinschaften von Brutvogelarten; Vorbild sind pflanzensiolo- gische Einheiten. Autor sieht den Sinn der Beschreibung derartiger Gemeinschaften in der Regelhaftigkeit, mit der dieselben Arten unter vergleichbaren Lebensbedingungen stets gemeinsam vorkommen. Die Anwendung sieht er bei der Bioindikation/Umweltmonitoring, beim Natur- und Artenschutz, bei Synökologie/Synchorologie, Synzönologie (Verbund von Avizönosen) und Synethologie (Verhalten bei interspezifischem Zusammenleben). Passarge unterscheidet 6 Kleinvogelzönosen (mit 22 Untergliederungen) und 11 Großvogel- zönosen (mit 25 Untergliederungen). Beispiel: die Greifvogelgemeinschaft der Wälder und Fluren besteht aus 3 Untergliederungen, von denen eine die Bussard-Falken-Gemeinschaften (Buteoni-Falconion) ist. In dieser gibt es die Máusebussard-Turmfalken-Gemeinschaft (Buteoni- Falconetum tinnunculi), die es in verschiedenen Ausbildungen gibt (z. B. B.-F. typicum oder B.-F. circetosum mit Rohrweihe). Der Ansatz erinnert an das Leitarten-Konzept von Flade 1994 (Die Brutvogelgemeinschaf- ten Mittel- und Norddeutschlands, IHW-Verlag), wobei aber Flade wesentlich näher an der Praxis liegt. Passarges Zönosen-Konzept erscheint mir zu typologisch und erstarrt sehr im Formalen. Goetz Rheinwald, Bonn 192 Buchbesprechungen Poot, M., L. M. Rasmussen, M. van Roomen, H.-U. Rósner & P. Südbeck (1966): Migratory Waterbirds in the Wadden Sea 1993/94. Wadden Sea Ecosystem 5, 79 S. ISSN 0946-896X. Zweiter Jahresbericht der drei Anliegerstaaten Dánemark, Deutschland und Niederlande über das Monitoring rastender Vögel im Wattenmeer. Der Band enthält die Mittwinterzáhlun- gen Januar 1994, Ringel- und Weißwangengans im März und Mai sowie die ganzjährig durch- geführten Springtidenzählungen 1993/94. Mit Kormoran, 9 Gänsen und Enten, 19 Limikolen und 4 Möwen sind 33 regelmäßig vorkommende Arten behandelt. Obwohl das gesamte Wattenmeer zwischen Langli (Dänemark) und Texel (Niederlande) behandelt ist, fehlen Anga- ben aus dem Nationalpark Hamburgisches Wattenmeer völlig. Wesentlicher Bestandteil ist die Umsetzung der Zählergebnisse in optisch viel einfacher zu erfassende Karten, die die Mittwinterverbreitung der einzelnen Arten nach Größenklassen zeigen. So können die artlich oft stark voneinander abweichenden wichtigen Rastgebiete auch von Außenstehenden schnell erkannt werden. Die Mittwinterergebnisse sind zusätzlich in Diagrammen mit Zählungen ab 1980 verglichen, was einen guten Einblick in die — allerdings oft von Winterwetter stark beeinflußten — winterlichen Rastbestandsentwicklungen erlaubt. Außergewöhnliche Entwicklungen und Bestände werden kommentiert, die in ausgewählten Gebieten durchgeführte Springtidenzählung wird als Monitoringmethode diskutiert. Stefan Stübing, Schwalmstadt HINWEISE FÜR AUTOREN Inhalt. — Bonner zoologische Beiträge ist eine wissenschaftliche Zeitschrift des Zoologischen Forschungs- instituts und Museums Alexander Koenig, Bonn. In ihr werden Originalarbeiten auf dem Gebiet der syste- matischen und vergleichenden Zoologie veröffentlicht. Manuskripte, die Ergebnisse aus der Arbeit des Instituts “ präsentieren oder auf den Sammlungen des Museums basieren, werden vorrangig ne doch sind andere Beiträge ebenso willkommen. Sprache. — Manuskripte können in deutscher, englischer oder französischer En abgefaßt N Einsendung von Manuskripten. — Manuskripte sind in zweifacher Ausfertigung zu senden an die Schrift- leitung, Bonner zoologische Beiträge, Lose Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, D 53113 Bonn. Auswahlverfahren. — Manuskripte werden einem Gutachterverfahren unterzogen; die Redaktion behält sich das Recht vor, Gutachten von externen Fachleuten einzuholen. Textgestaltung. — Manuskripte sollten inhaltlich und formal gründlich durchgearbeitet sein. Für Beschrei- bungen wird eine kurze und klare Ausdrucksweise empfohlen. Zitierte Literatur wird im Text abgekürzt (Autor, Jahr), die vollständigen Zitate erscheinen im Literaturverzeichnis (siehe unten). Bei Zitaten mit mehr als einem Autor benutze man das Zeichen „&” an Stelle von „und”. Bei der endgültigen Abfassung des Manuskriptes sollten die Autoren eines der letzten Hefte der Bonner zoologischen Beiträge hinzuziehen. Die Beschreibung neuer Taxa muß den Internationalen Regeln für die Zoologische Nomenklatur folgen. Es wird erwartet, daß die Autoren mit diesen Regeln vertraut sind und sie befolgen. Typusexemplare neu aufgestellter Taxa sollten in einem öffentlichen Museum oder Institut hinterlegt werden. Der Aufbau des Manuskriptes sollte sich am folgenden Schema orientieren: Titel, Verfasser, Abstract Key Words, Einleitung, Material und Methoden, Ergebnisse, Diskussion, Danksagung, Zusammenfassung, Literaturverzeichnis, Anschrift des(der) Autors(en), Abbildungslegenden, Tabellen, Abbildungen (jeweils auf der Rückseite mit Nummer und Autorenname[n] versehen). Ein kurzes Ape in englischer Sprache, maximal 10 Key Words und eine deutsche Zusammenfassung sind obligat Manuskripte sind durchgehend weitzeilig mit 3 cm Rand auf weiße DIN A A4 ne zu tippen. Computer- ausdrucke werden nur bei ausreichender Schreibqualitát berücksichtigt. Die Seiten sollten am oberen Rand fortlaufend numeriert sein. Nur wissenschaftliche Namen von Gattungen und niedrigeren Kategorien sol- len unterstrichen werden, alle anderen Auszeichnungen nimmt die Seg de na vor. Die Schreibweise der Literaturzitate sollte streng dem Stil der Zeitschrift folgen; Abkiirzungen fiir Periodika entnehme man der World List of Scientific Periodicals oder einem Pra Werk. Beispiele fiir die Zitierweise folgen: Eisentraut, M. (1963): Die Wirbeltiere des Kamerung Eisentraut, M. & H. Knorr (1957): Les ne AA. de la Guinée francaise. — Mammalia 21: 321335. Musser, G. G., L. R. Heaney & D. S. Rabor (1985): bl rats: a new species of Crateromys from Dinagat nd Am. Mus. Novit. 2821: 1—25. Bei der Planung von Abbildungsvorlagen und von Tabellen sollte oe zu ee stehende Satzspie- gel (126x 190 mm) berücksichtigt werden. Tabellen sollten auf ein Mindestm dungen und Tabellen nicht den gleichen Sachverhalt darstellen. Fußnoten s Korrektur. — Fahnenabzüge des Rohsatzes werden an die Autoren zur Korrektur gesandt. Sonderdrucke. — Von jedem Beitrag werden 50 Sonderdrucke kostenlos geliefert; weitere können gegen Erstattung der Kosten bezogen werden. Bestellformular und Preisliste liegen der Korrekturfahne bei. [English version in the next number] Bonner zoologische Beiträge Band 47, Hefte 1—2, 1997 INHALT On three species of Periscyphis Gerstaecker, -1873 from Kenya, Sudan and Oman (Crustacea: Isopoda: Oniscidea) S. Taiti, E. Ferrara & A: Allspach 772, 2 za ee 1 Coccinellidae (Coleoptera) aus Rwanda H. Fürsch ......:2. 2... 022000, Ss N N ee 13 Neue Glasflügler (Lepidoptera, Sesiidae) aus dem Pamir und dem Hindukusch K. Spatenka .......n 0002 oy woe A o ger 31 Neue Glasfliigler-Arten und Unterarten aus Europa und der Tiirkei (Sesiidae, Lepidoptera) K. Spatenka Lc o 43 Review of the genus Paradoxecia Hampson, 1919 (Lepidoptera, Sesiidae, Tinthiinae) O. G. Gorbunov € Y. Arita 1... 59 Erster Nachweis einer Strepsipterenparasitierung bei Wanzen in Mitteleuropa (Insecta, Strepsiptera et Heteroptera) A. Melber & H. Pohl... ... iia 2... 22 2.2.2 a ee 69 Post-eclosion heterochrony in the maturation of the adult females of a termitophilous fly (Diptera, Phoridae)_ R. H. L. Disney... 2.4 e N HA Recognition of a sibling species of the Australian Eutermiphora abdominalis Lea (Diptera: Phoridae) R. H. L. Disdey ct a At eee 87 A new species of Notolaemus Lefkovitch from Rwanda (Coleoptera: Lae- mophloeidae) Me Karner ...2... ed he whee cs oe) oe os Uae le 95 Collembola Poduromorpha from the Magallanes Province (Chile) R. Weiner € J. Naijt . ...... ...: 0.0.0 oa ik a 99 On the evolution of anthophilous Nitidulidae (Coleoptera) in non 3 and subtropical regions A. Ge Kifejtshuk.... 22... aa. a ete 111 Exceptional records of Microgale species (Insectivora: Tenrecidae) in vertre- brate food remains Ss. M. Goodman, P. D. Jenkins & O. Langrand.... Sy 135 Systematics and biogeography of the southern African scincine genus Typhla- contias (Reptilia: Scincidae) W. Haacke. ..:.e. 2... 22a. 0080 ee 159 Advertisement call characteristics of a Taiwanese green treefrog Rhacophorus prasinatus M. Matsui,“S*L. Chen E KY. Luena da 165 Description of the advertisement calls of some South American Hylidae (Amphibia: Anura): taxonomic and methodological consequences I. De la Riva, R. Marquez de E “Bosch. 173 A note on the gender of the genus Podarcis (Sauria: Lacertidae) W. Böhme........... A SN 187 Zum Gedenken an Hans Kumerloeve C. M. Naumann..\........ 40. de Og a ee 189 Buchbesprechungen ..:.......... a eee 191 This journal is printed on acid-free paper. (FL VA BONNER ZOOLOGISCHE BEITRÄGE Herausgegeben vom Zoologischen Forschungsinstitut und Museum Alexander Koenig Bonn Band 47, Heft 3-4, 1998 ISSN 0006-7172 Bonner zoologische Beiträge Herausgeber und Verlag: Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn. Schriftleitung: Dr. Rainer Hutterer, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, D-53113 Bonn. Erscheinungsweise und Bezug: Die Zeitschrift erscheint vierteljährlich zum Preis von 22,50 DM je Heft bzw. 90,— DM je Band einschließlich Versandspesen. Korrespondenz über Abonne- ments, Kauf oder Tausch bitte richten an die Bibliothek, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 160, D-53113 Bonn. Manuskripte und Besprechungsexemplare: Einsendung an die Schriftleitung erbeten. Manu- skriptrichtlinien sind am Ende dieses Heftes abgedruckt. Druck: JFeCARTHAUS, Bonn. Bonn, September 1998 Report on some sigmodontine rodents collected in southeastern Brazil with descriptions of a new genus and six new species Philip Hershkovitz f Table of Contents NOSE 6 o ae e a ES A A o N 193 TO TUCDOA oo A A UN 193 Gazencemotmcollectings OCalittesws o ee ee ee ew case A 195 Field Museum small mammal surveys in Brazil, 1986-1992 ..................... 196 Review of the Ruschi (1978) and Blair (1989) reports on the mammals of the Parque Nacio (CA Fr ne E ae en 198 Field Museum — Museu Nacional joint survey of small mammals of the Parave Nasal (Ca A A ee oe en ave 204 SJOSCIGS ACUM o o A A Si ns 205 Genus IDO EA AEs ds ATA Ln o E reos Re e LENA 206 GENUS AUSDUDA ÚS A O A ass 214 Ganas TIAS MA 2: 203. O A o lo O. A Ye 221 GENUSWBLUCEHAIIersSONIUSSMEWASENUS: A an 22 CENOS OEA O een han Bene 243 SUCIO. re anne: A TS NUS Cty Ie ee eee Lo 250 SENDEN 0.0.0.0 Beak se salen a A A Orcs, et Jy RRR AA 251 Danal pais. rl de ee ie eel rel eae a INE J) AU uae ANETO 9 253 AAC SITE Pees ca Meteo E A A TNA AUN: AA 253 AS AAC UN A look 254 Licame al. o e a ao & Dene ie No o EL i ER O ere 254 Abstract. The report is based on part of the collections of small mammals made in the Iporanga State Park, Sáo Paulo, the Parque Nacional de Caparaó, Minas Gerais-Espírito Santo, and localities not visited in southeastern Brazil. Accounts are given of five genera (1 new) and 14 species (6 new, 1 undescribed) of sigmodontine rodents. Two earlier reports by other authors on the mammals of the Caparaó National Park are reviewed. Key words. Rodentia, Sigmodontinae, Brazil, field survey, taxonomy, new genus, new species. Introduction Five field surveys of the small southeastern Brazilian mammals were conducted between 1986 and 1992 by the Field Museum of Chicago. Financial support was provided by the Barbara E. Brown Mammal Research Fund. The sponsoring and cooperative Brazilian Institutions are acknowledged individually in the account. This report deals primarily with results of studies of certain sigmodontines mostly collected in the Parque Nacional de Caparaó, Minas Gerais, the Parque Estadual Petar, Iporanga, Sáo Paulo, and comparative material in the collections of the Field Museum. 194 P. Hershkovitz 8 [> ¿RÍO DE JANEIRO > AS 24 j Fig. 1: Map of collecting localities mentioned in text and identified by number in gazetteer. The southeastern South American area shown is a portion of the Southeastern Zoogeographic Province. Localities of specimens mentioned in text and plotted by number in fig. 1: 1 Santa Teresa, Espírito Santo, Brazil; 2 Engenheiro Reeve, Espírito Santo, Brazil; 3 Caparaó, Minas Gerais, Brazil; 4 Lagoa Santa, Minas Gerais, Brazil; 5 Teresópolis, Rio de Sigmodontine rodents from southeastern Brazil 195 Janeiro, Brazil; 6 Centro de Primatológia, Rio de Janeiro, Brazil; 7 Itatiaya, Rio de Janeiro, Brazil; 8 Campos do Jordäo, Sao Paulo, Brazil; 10 Cotia, Sao Paulo, Brazil; 11 Alto da Serra, Sáo Paulo, Brazil; 12 Ribeiráo Fundo, Säo Paulo, Brazil; 13 Iporanga, Säo Paulo, Brazil; 14 Apiai, Säo Paulo, Brazil; 15 Piquete, Säo Paulo, Brazil; 16 Roca Nova, Paraná, Brazil; 17 Taquara, Rio Grande do Sul, Brazil; 18 Sáo Lourenco, Rio Grande do Sul, Brazil; 19 Maldonado, Maldonado, Uruguay; 20 Arazatí, S. Ecilda, San José, Uruguay; 21 Torrecita, Buenos Aires, Argentina; 22 Tobunas, Misiones, Argentina; 23 Caraguatay, Misiones, Argentina; 24 Dos de Mayo, Misiones, Argentina; 25 Puerto Gisela, Misiones, Argentina; 26 Limbani, Puno, Perú; 27 El Choro, Cochabamba, Bolivia; 28 Colomi, Cochabamba, Bolivia. Gazetteer of collecting localities mentioned in text (numbers in parentheses refer to positions on map, fig. 1): ARGENTINA Buenos Aires Torrecita (= Urdampilleta), 3615/6106, 108 m (21) Misiones Caraguatay, Río Paraná 2637/5446, 100 m (23) Dos de Mayo, 2702/5439, 300 m (24) Puerto Gisela, Río Paraná, 2701/5527, 50 m (25) Tobunas, 2628/5354, 200 m (22) BOLIVIA Cochabamba Muelle Muelle, 2700 m, not located Colomi, 1731/6552, 3075 m (28) El Choro, Ayapoc, 1656/6652, 3500 m (27) BRAZIL Espírito Santo Engenheiro Reeve, 2046/4128, 400—600 m (2) Santa Teresa, 1955/4036, 659 m (1) Minas Gerais Caparaó, Parque Nacional de, 2025/4147 (3) Arrozal, 2300 m Cachoeira Bonita, 1750 m Centro de Visitantes, 1300 m Pico da Bandeira, 2700 m; 2890 m Segredo, 2100 m Terreiräo 2370 m Tronqueira, 1970 m Vale Encantado, 1980 m Vale Verde, 1200 m Espirito Santo Pedra Roxa, 100 m Lagoa Santa, 1938/4353, 760 m (4) Parand Roca Nova, Serra do Mar, 2531/4830, 1000 m (16) Rio de Janeiro Centro de Primatologia, 2254/4314, sea level (6) Fazenda Boa Fe (see Teresöpolis) Itatiaya, 2223/4438, 1800 m (7) Teresöpolis, 2226/4259, 902 m (5) 196 P. Hershkovitz Rio Grande do Sul Rio dos Linos [Sinos] (see Taquara) Sao Lourenco (Säo Lourenco do Sul), 3122/5158 (18) Taquara, 2939/5047, 29 m (17) Taquara do Mundo Novo (see Taquara) Sao Paulo Alto da Serra (= Paranapiacaba), 2347/4619, 800 m (11) Apiaí, 2431/4850 200 m (14) Boraceía, Serra do Mar, 2339/4554, 800—900 m (9) Campos do Jordáo, 2244/4535, 2585 m (8) Cotia, 2337/4656, near sea level (10) Iporanga, 2435/4835, 200 m (13) Piquete, Serra da Mantiquiera, 2236/4511, 600—900 m (15) Ribeiráo Fundo, 2415/4745, 30 m (12) PERÚ Puno Limbani, 1408/6942 (26) URUGUAY Maldonado Maldonado, 3454/5457 sea level (19) San José Arazati, S. Ecilda, 3433/5657 sea level (20) Field Museum small mammal surveys in Brazil, 1986 — 1992 Our study of the small mammals of southcentral and southeastern Brazil began with live trapping in the campos cerrados of Brasília, D. F., during July— August 1986. The prime objective at the time was the capture of the fossorial sigmodontine rodent _ Juscelinomys candango Moojen, 1965. The species is known from the original series of 8 specimens, including holotype, taken from a cerrado subterranean system in the Parque Nacional de Brasília. The live animal has never been seen again. Our intensive trapping at ground level, and excavations of burrows proved fruitless. Nevertheless, a collection of about 150 rodents of 15 genera resulted from our search. Most of the material is still under study. The reports published to date include the description of a new species of Thalpomys (Hershkovitz 1990a), a new species of Akodon (Hershkovitz 1990b) and a new genus Microakodontomys and its type species, M. transitionalis (Hershkovitz 1993). Official permits for work in the Parque Nacional de Brasília and cerrado (fig. 2) were granted through the sponsorship of Professor Ulises Caramaschi of the Museu Nacional. The then graduate assistant of the mammal department, Christopher Tribe, was assigned to the project. Others who assisted with field work, material support, hospitality and in many other ways, included Professor Jader Soares Marinho Filho, and Professor Milton Thiago de Mello, both of the Universidade de Brasília, Alberto de Paulo Carlos, of IBAMA (Instituto Brazileiro do Meio Ambiente e de Recursos Naturais Renováveis), Miguel Angelo Marini, a student, and Dr. Braulio F. de Sousa Dias of the Brazilian Geographic Institute. Scott Lindbergh, an American resident in Brasilia D. F., who was studying the behavior of the cerrado howler monkey, Alouatta fusca, volunteered his services without conditions. Scott’s devoted and generous assistance throughout the entire field program accounts for much of the success of our field work here and elsewhere in eastern Brazil. Sigmodontine rodents from southeastern Brazil 197 Fig. 2: Views of Parque Nacional de Brasília, Brazilia, D. F. The Primate Center of Rio de Janeiro about 60 km N of the city, with small remnants of Atlantic rain forest was visited October — November, 1987, for sampling 1ts small mammals. About 120 specimens were collected but the unidentified material has not been released by our then Brazilian sponsor and field associate, Professor Jader Soares Marinho Filho. Field work was facilitated by the Center’s Manager, Lourenco, and by the dedicated assistance of Professor Jader Marinho, his student Marcelo Lima Reis, and Field Museum Associate Barbara E. Brown. I am partic- ularly grateful to the then Director of the Center, Dr. Adelmar Coimbra-Filho for his deep interest in our labors and unstinting cooperation. The Rio de Janeiro State Park at Teresópolis, was worked July— August 1988. Our Brazilian sponsor was again Professor Jader Marinho. The approximately 125 specimens collected still remain in his care and have not been made available for study. Field assistants included Jader Marinho Filho, Marcelo Lima Reis, and our Barbara Brown. The Sáo Paulo State Park Petar, Municipio Iporanga (fig. 3), engaged our survey efforts November — December, 1989. The approximately 220 specimens distributed in about 20 genera were delivered to the Zoology Museum of the University of Sáo Paulo, for processing. The entire collection is now being studied in the Field Museum. Dr. Paulo E. Vanzolini, Director of the Museu de Zoologia, Universidade de Sáo Paulo, was our sponsor. 198 P. Hershkovitz Fig. 3: Four views of Parque Estadual Petar, Iporanga, Sáo Paulo. The Parque Nacional de Caparaó (figs 4, 5), with an area of 16,194 hectares, straddles the comparatively dry part of western Minas Gerais and the humid eastern or Atlantic slope of Espirito Santo. The Park, between 20°19’—20°37’S, 41°43’— 42°53’W (fig. 1) was established May, 1961 by Federal decree, far too late to prevent virtually complete destruction of the original forest cover. The Serra do Caparaó at the northernmost edge of the Serra do Mar is rugged mountain country with elevations from about 800 m to the Pico da Bandeira at 2890 m. The terrain is broken by many steep valleys and waterfalls. The vegetation in 1981 was described as tropical rain forest to about 1800 m then open grass- land above 2400 m. In 1989 the flora was described by IBAMA as almost entirely secondary, considerably degraded, with no resemblance to the original formations. The vegetation had become increasingly poorer at higher altitudes, with dominance of epiphytes, mosses and lichens. The mammalian fauna, according to IBAMA, was reduced to such widespread species as the rabbit, agouti, squirrel, and small predators including eira cat, fox, racoon and coati. Review of the Ruschi (1978) and Blair (1989) reports on the mammals of the Parque Nacional de Caparaó The recent mammals of the Parque Nacional de Caparaó are known from two comprehensive reports. The first, by Ruschi (1978) appears to be a collection of Sigmodontine rodents from southeastern Brazil 199 Fig. 4: Views of Parque Nacional de Caparaó, Minas Gerais. scattered data of mostly personal observations gathered over a period of 40 years. The second, by Blair (1989) is from her unpublished doctoral thesis based on 18 consecutive months of field work. Each report lacks documentation for specimens collected and most of the identifications must be taken at face value. Nonetheless, their reports are the most comprehensive available in terms of the mammals of the region. The Ruschi account The late Augusto Ruschi, a long time resident of Santa Teresa, Espirito Santo, observed the birds and mammals, particularly hummingbirds and bats, of the Caparaó mountains for over 40 years. His (1978) report lists 74 species of mammals, 251 species of birds, and a detailed account of the vegetation and life zones of the Parque Nacional de Caparaó. The account includes names of authors with biblio- graphic references to original descriptions, and scientific and vernacular names of the species. A bat of the genus Tadarida was described. Ruschi (p. 26) emphasized the fact that survival of the woolly spider monkey (Brachyteles arachnoides) in the Park was threatened (none survived) and that the lion tamarin, Leontopithecus rosalia chrysomelas disappeared from Espírito Santo in 1978, most of its habitat having been converted to charcoal. He predicted early extinction of the endemic mouse Abrawayomys from the nearby Forno Grande Biological Reserve. 200 P. Hershkovitz SEGREDO VALE ENCANTADO ARROZAL 7 ES 2 ——-PEDRA ROXA CACHOEIRA BONITA se E - TRONQUE IRA TERREIRÁO PICO DA BANDERA PARQUE NACIONAL DE CAPARAÓ Fig. 5: Map, Parque Nacional de Caparaó, Minas Gerais — Espírito Santo; collecting stations shown are on the Minas Gerais (western) slope, except Pedra Roxa in the Espírito Santo (eastern) watershed. The “Museu de Biologia Prof. Mello Leitao,” issuer of Ruschi's publications, housed no mammals and its Boletim, like the Museum, was the property of Dr. Ruschi. He listed the following Caparaó mammals in 1978. Voucher specimens may not exist but the identifications are presumed to be reliable. Nominal equivalents as used in Mammal Species of the World follow in brackets, the subspecies suppressed by the compilers. The work was edited by Wilson and Reeder (1993). Sigmodontine rodents from southeastern Brazil 201 MARSUPIALIA Didelphidae Didelphis marsupialis Linnaeus Didelphis aurita Wied-Neuwied Metachirus nudicaudatus myosurus Temminck [Metachirus nudicaudatus E. Geoffroy] Philander philander philander [Caluromys philander Linnaeus] Chironectes minimus Zimmermann Monodelphis scalops Thomas Monodelphis iheringii Thomas [Monodelphis iheringi Thomas] Marmosa cinerea cinerea Temminck 1824 [Micoureus demerarae Thomas; cinerea Temminck preoccupied by D. cinerea Goldfuss, 1812] CHIROPTERA Molossidae Molossus ater E. Geoffroy Eumops auripendulus Shaw Tadarida espiritosantensis Ruschi [“probably a synonym of Nyctinomops laticaudatus,‘‘ Koopman, 1993: 240] Phyllostomidae Phyllostomus hastatus Pallas Chrotopterus auritus australis [Chrotopterus auritus Peters] Mimon bennetti Gray Micronycteris megalotis Gray Glossophaga soricina Pallas Artibeus jamaicensis planirostris Spix [Artibeus planirostris Spix] Desmodontidae [Desmodontinae subfamily of Phyllostomidae] Desmodus rotundus É. Geoffroy Diphylla ecaudata Spix Vespertilionidae Myotis nigricans Schinz Lasiurus ega argentinus Thomas [Lasiurus ega Gervais] PRIMATES . Callicebus gigot [Callicebus personatus E. Geoffroy] Alouatta fusca E. Geoffroy Cebus nigritus Goldfuss [Cebus apella Linnaeus] Brachyteles arachnoides E. Geoffroy Callithrix aurita caelestis Miranda Ribiero [Callithrix aurita E. Geoffroy] Callithrix flaviceps Thomas EDENTATA Tamandua tetradactyla Linnaeus Bradypus tridactylus brasiliensis Blainville [Bradypus variegatus Schinz] Euphractus sexcinctus flavimanus Desmarest [Euphractus sexcinctus Linnaeus] Dasypus novemcinctus Linnaeus LAGOMORPHA Sylvilagus brasiliensis minensis Thomas [Sy/vilagus brasiliensis Linnaeus] RODENTIA Sciurus ingrami Thomas [Sciurus aestuans Linnaeus] Thomasomys pyrrhorhinus Wied Neuwied [Wiedomys pyrrhorhinus Wied-Neuwied] Thomasomys dorsalis collinus Thomas [Delomys dorsalis Hensel] Thomasomys sublineatus Thomas [Delomys sublineatus Thomas] Nectomys squamipes olivaceus Hershkovitz [Nectomys squamipes Brants] Rhipidomys masticalis Lund Phaenomys ferrugineus Thomas Akodon arviculoides cursor Winge [4kodon cursor Winge] Akodon subterraneus Hensel [Thaptomys nigrita Lichtenstein] 202 P. Hershkovitz Oxymycterus nasutus Waterhouse Blarinomys breviceps Winge Coendou prehensilis Linnaeus Cavia aperea azarae Lichtenstein [Cavia aperea Erxleben] Dasyprocta aguti Linnaeus Cuniculus paca Linnaeus [Agouti paca Linnaeus] Euryzygomatomys guiara Brandt [Euryzygomatomys spinosus] Echimys medius Thomas [Echimys blainvillei F. Cuvier] CARNIVORA Dusicyon vetulus Lund [Pseudalopex vetulus Lund] Dusicyon thous azarae Wied Neuwied [Cerdocyon thous Linnaeus] Procyon cancrivorus nigripes Mivart [Procyon cancrivorus G. Cuvier] Nasua nasua nasua Linnaeus [Nasua nasua Linnaeus] Nasua nasua solitaria Schinz [Nasua nasua Linnaeus] Potos flavus nocturnus Wied Neuwied [Potos flavus Schreber] Tayra barbara gulina Schinz [Eira barbara Linnaeus] Grison furax Thomas [Galictis cuya Molina] Panthera onca Linnaeus Puma concolor greeni Nelson & Goldman [Puma concolor Linnaeus] Felis pardalis brasiliensis Oken [Leopardus pardalis Linnaeus] Felis wiedii Schinz [Leopardus wiedi] Felis yaguarondi Lacepede [Herpailurus yaguarondi E. Geoffroy] Felis pardinoides Gray [Leopardus tigrinus Schreber] PERISSODACTYLA Tapirus terrestris Linnaeus ARTIODACTYLA Tayassu pecari Link Tayassu tajacu Linnaeus [Pecari tajacu Linnaeus] Mazama americana Erxleben Mazama simplicicornis Illiger [Mazama gouazoupira G. Fischer] Remarks. Most of the species recorded by Ruschi may still be found on the Espírito Santo slope. The list of bats may be longer. Also present are the rabbit, squirrel, and more kinds of sigmodontine rodents. Some of the carnivores may have been extirpated. The once abundant tapir, deer, peccaries and most if not all monkeys may have disappeared. The Blair Report 1979—1980 Blair recorded 53 species captured or observed in the Caparaó National Park from May 1979 through October 1980. The small mammals were live trapped, sexed, marked, measured and released. Trap lines were along transects, hence not selective for species diversity. Bats were caught in mist nets. Carnivores were identified by recovered feces or pellets, burrows, tracks, sight or other external signs. Identifications of captured animals were based on their external measurements compared with those of previously identified specimens preserved in museums or recorded in the literature. Characters other than standard external dimensions taken from the anesthetized captives were not used in the identifications. The identification of feces by size and shape could be problematic. Authorities for generic and specific names were systematically eschewed. None of the animals recorded were preserved for confirmation of their identifica- tions and they are listed as recorded. Sight records are bracketed. Sigmodontine rodents from southeastern Brazil MARSUPIALIA Didelphis marsupialis Metachirops opossum Marmosa cinerea Marmosa murina Marmosa sp. 3 EDENTATA Cabassous tatouay Dasypus novemcinctus Dasypus septemcinctus Euphractus sexcinctus Myrmecophaga tridactyla PRIMATES Cebus apella Callicebus moloch CHIROPTERA Carollia perspicillata Desmodus rotundus Glossophaga soricina Histiotus velatus Phyllostomus hastatus Stenodermatinae sp. indet. (bones from owl pellets) Sturnira sp. Uroderma bilobatum Vampyrops [= Platyrrhinus] lineatus Vampyressa pusilla RODENTIA Akodon arviculoides Akodon cursor Akodon serrensis Blarinomys breviceps Nectomys squamipes Oryzomys bicolor Oryzomys capito Oryzomys concolor Oryzomys eliurus Oryzomys flavescens Oryzomys fornesi Oryzomys intermedius Oryzomys nigripes Oryzomys nitidus Oryzomys ratticeps Oxymycterus hispidus Thaptomys nigrita Sciurus ingrami Thomasomys dorsalis Thomasomys [= Wilfredomys] oenax (bones from owl] pellets) Unknown species: “Red mouse,“ diurnal; in grassland at 2500 m; long-tailed Cavia aperea Coendou prehensilis Dasyprocta agouti Echimys medius CARNIVORA Nasua nasua 203 204 P. Hershkovitz Procyon cancrivorus [Pseudolopex gymnocercus] [Cerdocyon thous] Chrysocyon brachyurus [Lycalopex vetulus] Galictis vittata Eira barbara Felis [Leopardus] wiedi/tigrinus Felis [Herpailurus] yagouaroundi Felis [Leopardus] pardalis Felis [Puma] concolor ARTIODACTYLA Tayassuidae sp. indet. LAGOMORPHA Sylvilagus brasiliensis Field Museum-Museu Nacional joint survey of small mammals of the Parque Nacional de Caparaó, 1992 The Parque Nacional de Caparaó expedition promoted by the Field Museum and sponsored by the Museu Nacional, surveyed the small mammals, particularly the Sigmodontinae, September—October 1992. Participants in field work included Dr. Alfredo Langguth, Curator of Mammals of the Museu Nacional, our official sponsor for the field work, and his two assistants, Stella M. France and Allison A. Sodre. The Field Museum was represented by the author and Associate Barbara E. Brown. Other participants included parasitologist Dr. Pedro Marcos Linardi of the - University of Minas Gerais who focused on the ectoparasites of the captured mammals; Cibele Bonvecino, graduate student of the Universidade Federal de Rio de Janeiro, assisted actively in field work, her primary concern, however, being the cytogenetics of captured rodents. IBAMA representative Alberto de Paulo lent considerable field support. The indispensable Field Museum Associate Scott Morrow Lindbergh assured the success of the expedition with his keenly directed and indefatigable labors. The Park Director José Olimpo Vargas provided housing. The Assistant Director Estevao Marchesini Fonseca helped with our installation, and facilitated our Operations in many ways during the entire stay. The Field Museum-Museu Nacional survey of the mammals of the Parque Nacio- nal de Caparaó resulted in the capture and preservation of approximately 427 small mammals. Represented are 31 species distributed among the 21 genera listed below (number of species in each genus shown in parentheses). | Marmosa Gray (1) Marmosops Matschie (1) Micoureus Lesson (2) Monodelphis Burnett (1) Philander Tiedemann (1) Didelphis Linnaeus (1) Carollia Gray (1) Desmodus Wied (1) Sigmodontine rodents from southeastern Brazil 205 Glossophaga É. Geoffroy (1) Akodon Meyen (3 including 1 new) Thaptomys Thomas (1) Brucepattersonius (new genus; 4 all new) Oxymycterus Waterhouse (2 including 1 new) Calomys Waterhouse (1) Oryzomys Baird (2) Oligoryzomys Bangs (3) Nectomys Peters (1) Delomys Thomas (2) Proechimys J. A. Allen (1) Euryzygomatomys Goeldi (1) Cavia Pallas (1) Results of the study of all the material collected in Caparaó, Iporanga and other localities of the region are being prepared for publication. The taxa included in this first report are the following. Delomys Thomas D. dorsalis Hensel D. sublineatus Thomas Akodon Meyen A. serrensis Thomas A. cursor Winge A. mystax (new species) Thaptomys Thomas T. nigrita Lichtenstein Brucepattersonius (new genus) . SOricinus new species and type species . igniventris (new species) . griserufescens (new species) . albinasus (new species) . Iheringi Thomas . Sp. (from Misiones, Argentina, the “Oxymycterus iheringi” of Massoia [1963], not Thomas) Oxymycterus Waterhouse O. rufus Fischer O. caparaoe (new species) Do by bo bu u io No large mammals other than the opossum (Didelphis) and coati (Nasua) were seen during our short visit. Signs of other predators were absent. Bats were not seen in flight at night but bad weather hindered observation and, except for the most common species, were not located during the day in apparently suitable roosts. The Ruschi and Blair reports present a picture of a rich mammalian fauna. Species accounts -Abbreviations used include: BM = British Museum (Natural History), London; FM = Field Museum, Chicago; MN = Museu Nacional de Brazil, Rio de Janeiro; MZUSP = Museu de Zoologia, Universidade de Säo Paulo, Sáo Paulo; PH = field numbers from 1992 surveys of Caparaó; E = Ear pinna from notch; GSL = Greatest skull length; HB = Length of head and body combined; HF = Hind foot length with claw (cu); without claw (su); T = Tail length. 206 P. Hershkovitz Delomys Thomas Delomys Thomas. Type species Hesperomys dorsalis Hensel, 1872, by original designation. The genus has been revised by Voss (1993) with most of the taxonomic problems resolved. Of the two recognized species 72 specimens of Delomys dorsalis and 46 of D. sublineatus were examined by Voss. The new material studied here, most not seen by Voss, consists of 88 specimens with the registry numbers of the Museu de Zoolo- gia, University of Säo Paulo, the Museu Nacional, Rio de Janeiro, and the Field Museum. Included are 72 specimens of Delomys dorsalis from 12 localities, and 15 specimens of D. sublineatus from 5 localities, all listed beyond under specimens examined, their localities plotted in figure 1 (map). The new material provided a basis for review of parts of Voss's findings. Characters. Delomys is a medium sized sigmodontine; tail from slightly longer (105 %) than combined head and body length (HB) to shorter (75 %); hind foot long, narrow, length with claw 20 % to 25 % HB; ear large, length from notch about 15 % to 16 % HB; condyloba- sal length of skull about 20 % to 25 % HB, or approximately the same proportion as hind foot length with claw; mammae, 2 inguinal, 2 abdominal, 2 postaxial, and occasionally 2 pectoral = 6 or 8. Coloration of upperparts dark brown (dominantly eumelanin) to buffy (dominantly pheo- melanin), underparts dominantly grayish (dilute eumelanin); a more or less defined dark brown middorsal stripe nearly always present. Dorsal contour of skull slightly convex; rostrum elongate; distal part of nasals and premaxillae combined usually project as a tube beyond incisors, the cartilaginous median septum (often lost in preparation) between nasal tips produced slightly beyond in expanded form; interorbital region more or less hourglass shaped, the supraorbital margins posteriorly rounded, square or frequently beaded, the beading continued behind as lateral parietal ridges; zygomatic arches moderately expanded with slight anterior convergence; perpendicular zygomatic plate visible viewed from above; interparietal bone well developed; incisive foramina comparatively short, not extending to level of first molars; palatal bridge short, wide, without conspicuous posterolateral pits, the posterior border not produced behind level of third molars; stapedial and sphenofrontal foramina present, squamosal-alisphenoid groove more or less defined; mandible with capsular tip of incisor root little or not pronounced. Upper incisors small, inclination orthodont to opisthodont; molars brachyodont, pentalo- phodont, tuberculate, m, with well developed ectolophid. Pentalophodonty or fusion of the mesostyle (-id) with the mesoloph (-id), in all molars to form the mesolophostyle (-id), is the one and only consistent and definitive character of thomasomyines (Delomys included) and oryzomyines that distinguishes them from ali other sigmodontines. Muroid molars with a mesostyle (-id) discrete or absent in any tooth, or with a mesoloph (-id) discrete or absent in any tooth, are not pentalophodont. Disjunction between mesostyle (-id) and mesoloph (-id) with reduction and ultimate loss of either or both elements in any tooth of a dental system marks the irreversible transition from pentalophodonty to tetralophodonty (cf. Hershkovitz 1993). Generic Relationships. In 1962, I (p. 21) suggested that Thomasomys, with a complex glans penis and a simple chambered stomach, must be much like the ancestral morphotype which could have given rise to the South American cricetines. The thomasomyine lineage, in my opinion at the time, included Thomasomys, Nyctomys, Otonyctomys, Phaenomys, Rhipidomys, and Wilfredomys. Elsewhere in the account (1962: 84, footnote 3), Aepeomys, Inomys, Delomys and Erioryzomys were referred to the thomasomyine group. It was cautioned, however (1966: 12, footnote), that “some characters shared by members of the group might be phenetic rather than phyletic” Subsequent removal of Nyctomys and Otonyctomys proved this point. With passage of time and increase in knowledge, more modifications were made. The phylogenetic position of Delomys was analyzed by Voss (1993: 21) on the basis of morphological comparisons with Thomasomys and Oryzomys. Thomasomys, with 25 or more Sigmodontine rodents from southeastern Brazil 207 species (Musser & Carleton 1993: 749), was represented in the comparisons by its type species, T. cinereus. Oryzomys, with 36 species (Musser & Carleton 1993: 712), was represented by the type species, Oryzomys palustris, a highly derived if not the most derived species, of the genus. The eighteen characters used by Voss for comparison of Delomys dorsalis and D. sublinea- tus with Thomasomys cinereus and Oryzomys palustris are briefly discussed below. Numbers 19, 20 and 21 are additional sigmodontine characters mentioned separately by Voss (1993: 9, 12). Comparisons of the characters of type species of Delomys and Thomasomys are given in table 1, their nipples in table 2; measurements of Delomys dorsalis are in table 3, and of D. sublineatus and D. dorsalis in table 4. 1. Ungual tufts. Present in all four morphotypes (Delomys dorsalis, D. sublineatus, Thomaso- mys cinereus, Oryzomys palustris). Starting from a hypothetical primitive state of barely covering the claws, the tufts become hypertrophied in Delomys and Thomasomys cinereus. In Oryzomys palustris, they evolved in a negative direction to the point of near disappearance. 2. Plantar pads. The state of the 6 plantar pads (two metatarsal and four interdigital) is regarded as primitive in Delomys and Thomasomys. The system of 5 plantar pads in Oryzomys appears to be derived. 3. Mammae. The primitive number of paired teats or nipples in sigmodontines is 3. The paired formula in mammals developed pari passu from inguinal to abdominal and postaxial as in Thomasomys and nearly all Delomys. A pectoral pair became a fixture in oryzomyines, phyllotines and others, but in Delomys only as an infrequent variable (table 2). Voss’s (1993: 13, 30) interpretation of the comparative morphology appears to be correct. 4. Rostral tube. This highly variable character ranges in Delomys from slightly protruded to extremely so but no more than in certain related species (1. e., Thomasomys cinereiventer). The | rostral tube said to be absent in Thomasomys cinereus is actually more than incipient. 5. Zygomatic notch and zygomatic plate. Definition of the zygomatic notch depends largely on the form of the plate. In some Delomys (MN 31961, 31965) the notch may be as “indistinct” as illustrated by Voss (1993, fig. 5B) for Thomasomys, but in some Thomasomys, the notch may be nearly as “distinct” as illustrated by Voss (1993, fig. SA) for Delomys. Definition of the notch in Thomasomys and Delomys is likewise a continuum beginning with the less deri- ved state in Thomasomys. 6. Interorbital region and temporal crests. The hourglass shape as described and figured by Voss (1993: 15, fig. 6) for Delomys and Thomasomys holds. The posterior half of the supra- orbital margin, however, is described as smoothly rounded, or sometimes gently squared but with sharp edges, beads, or projecting shelves never developed. This is said to be “the plesiomorphic condition.” It does not hold, however, for nearly all specimens at hand of Delomys and Thomasomys cinereus. 7. Palate. The palate of Delomys and Thomasomys is correctly defined as short, that is with posterior margin not produced behind the third molars. It is seen as “probably plesio- morphic.” According to Voss (1993: 15) “the bony palate of Delomys is short because the mesopterygoid fossa extends anteriorly between the third molars.” The mesopterygoid fossa, a feature without definition in itself, does not form the palate or fashion its margin. It is, on the contrary, defined by the bony palate. 8. Mesopterygoid fossa and sphenopalatine vacuities. The sphenopalatine vacuities, slits or fenestrations of the bony roof of the mesopterygoid fossa vary in Delomys from nearly completely ossified, as depicted by Voss (1993, fig. 7 B) for Thomasomys, to 2 and 3 times the size shown for Delomys in his fig. 7 A. The roof of the mesopterygoid fossa in Thomaso- mys cinereus is ossified with few minute exceptions in the 30 specimens examined here. The unfenestrated or completely ossified roof may not be primitive. Ossification is an ongoing developmental process but need not be complete in any one structure. 9. Alisphenoid strut. The strut, absent in Delomys but present in Thomasomys cinereus, may be the most trenchant character for separation of the two genera. 208 P. Hershkovitz Table 1: Comparisons of Delomys and Thomasomys based on their type species Delomys dorsalis Hensel and Thomasomys cinereus Thomas. Thomasomys cinereiventer J. A. Allen included for comparison. Character . Mammae . Tail length . Blackish middorsal stripe . Bevel of posterior surface of lower incisors . Bevel of posterior surface of upper incisors . Zygomatic plate . Nasals with premaxillae . Supraorbital beads or ridges . Temporal ridges . Nasals . Zygomatic arches anteriorly . Frontoparietal suture . Incisive foramina . Ascending mandibular ramus . Coronoid process . Sigmoid notch . Molars . Upper incisors . Lower incisors . Capsule of upper incisor root . Capsular process of lower incisor . Entepicondylar foramen . Articulation of rib 1 . Alisphenoid strut . Ectolophid l weakly developed in old individuals A Delomys . 3 or 4 pairs . From 75 % as long as head and body combined to 5 % longer; average about equal. . Always present, usually well defined . 60—70 % from tip . 50—75 % from tip . Forward projecting, anteorbital foramen (notch) well exposed seen from above . Usually extended as a tube . Partially present . Present but weak . Tapered to obtuse point . Slightly convergent . Widely rounded . Subovate, about as open in front as behind . Wide, nearly square . Short, as spine . Short, rounded, deep . Large, long . Opisthodont to orthodont . Fine, short . Not bulging . Slightly or not indicated . Absent . With thoracic Ist, or cervical 7th and thoracic 1st . Absent . Present B Thomasomys 1. 3 pairs . Always longer . Absent . 40—60 % from tip . Nearly to or quite to alveoli . Hardly or not forward projecting, foramen slightly or not exposed seen from above . Tips of nasals usually on same vertical plane as incisors, sometimes slightly anteriad . Absent . Variable . Slightly expanded to rounded tip . Expanded or rounded . Narrowly rounded . Narrower in front than behind . Narrow, rectangular . Elongate, as process . Long, shallow . Smaller, shorter . Mostly orthodont . Heavier, longer . Bulging . Slightly pronounced . Not available . Not available . Present . Absent e cinereiventer . As in A, but broader, diffuse . Asin B .Asin A . Absent! .Asin A . Asin A . Asin A .Asin A in B in B in B in A .AsinB . Not available . Not available . Present . Absent Sigmodontine rodents from southeastern Brazil 209 10. Carotid circulation. Present as described in Delomys and Thomasomys cinereus. Pattern 1 of Voss (1993: 18, 24) appears to be primitive. 11. Tegmen tympani. This outgrowth of the periotic bone overlies the squamosal bone in Delomys and Thomasomys as stated. It also may merely contact the squamosal bone, and in a few samples is separated from it. Overlapping is said to be primitive but perhaps is an advanced stage of a relationship between bones. 12. Capsular process. A minute point marking the tip of the lower incisor root in the mandible is present in Delomys and most Thomasomys cinereus, but may be slightly swollen in individuals. It appears that the more pronounced the capsular process, the more derived. 13. Incisors. Judgement of the amount of inclination of the upper incisor is largely subjective. Voss (1993: 24, fig. 5, A, B) figures and describes the Delomys incisor as “small and strongly opisthodont that of Thomasomys as “large and weakly opisthodont“. In material at hand, most incisors of Delomys are weakly opisthodont to moderately orthodont. In my view those of Thomasomys cinereus are mostly orthodont with a few weakly opisthodont. Most Delomys incisors are smaller than those of Thomasomys. The weakly opisthodont incisor may be the more primitive. 14. Molar occlusal design. The unworn molars of Delomys and Thomasomys may be described as basically brachyodont, pentalophodont, tuberculate, and crested bilevel. Every stage of wear causes a change in molar crown design. The lophs of the molar crowns of Delomys and Thomasomys cinereus (Voss 1993: figs 9, 10) are the well defined lophs of the pentalophodont pattern. No consistent differences in designs of Delomys and Thomasomys molars exist except for presence of an accessory lophid in Delomys that is absent in Thomaso- mys cinereus (see item 16 below, and Hershkovitz 1962: 69 et. seq.). 15. Anterocone or anteromedian flexus. The feature is present in M! in both Delomys and Thomasomys cinereus as described by Voss (1993: 20, 24). Variability of the conules defined by the flexus is discussed in my description of Delomy dorsalis (p. [43 ms.]). 16. Ectolophid. The well developed loph between protoconid and hypoconid of m, in Delomys originates on the labial side of the mure and fuses with the ectostylid of the cingulum. An ectolophid is sometimes present in m,. The ectolophid is absent in Thomaso- mys cinereus except for an anlage in some specimens. The ectostyle, however, is always present and well developed. It is difficult to judge within the limitations of this comparison whether the ectolophid is primitive or derived. Unlike the mesolophid on the opposite side of the mure, there are no signs of intermediate stages in the genera compared that might point to the direction of change. 17. Number of ribs. Thirteen ribs is the usual number in mammals including Delomys and Thomasomys. In many, if not most individuals, the number may vary from 12 to 14. The mode, therefore, is 13, one rib more or less may be regarded as within normal or primitive limits. 18. Gall bladder. A gall bladder was found by Voss (1993: 21) in all seven specimens examined of Delomy dorsalis but not in “a poorly fixed liver of the only available fluid example of D. sublineatus.” It was present in the only two available specimens of Thomasomys cinereus. 19. Tuberculum of first rib. The articulation is with the transverse process of the seventh cervical vertebra also with first thoracic vertebra as stated by Voss (1993: 9). The number of specimens Voss examined was not given. In a specimen at hand of Delomys sublineatus (FM 149629) and another of D. dorsalis (MN 31935) the articulation is with thoracic only. In remaining 4 available skeletons of D. dorsalis, the articulations are with cervical 7 and thoracic 1. 20. Entepicondylar foramen. The foramen, which provides passage for the median nerve and brachial artery is absent in present material. According to Carleton (1980: 52) “the foramen may be absent in whole groups currently considered monophyletic, e. g., microtines and South American cricetines ... Within the neotomine-peromyscines, the foramen is typically 210 P. Hershkovitz Table 2: Number of nipples in the species of Delomys. Explanation: m6 = 3 pair nipples; m8 = 4 pair nipples (including pectorals). Taxon Localitiy Delomys dorsalis Caparaó Iporanga Teresópolis Boraceia Caraguatay Delomys sublineatus Teresópolis Iporanga present.“ The supracochlear foramen, a perforate of the olecranon fossa, common in South American sigmodontines, may be confused with the entepicondylar foramen where the latter is normally absent. 21. Basihyal (basihyoid) without entoglossal process apud Voss (1993: 9). The few sigmo- dontine hyoids I examined lack an entoglossal process. Habits and habitat. According to information compiled by Voss (1993), both Delomys dorsalis and D. sublineatus inhabit wet tropical and subtropical forest floors. His photographs of habitats suggest a preference for lush vegetation. Sympatric D. dorsalis and D. sublineatus were collected by our group in the humid secondary forests of Iporanga and Apiaí but also in open areas. Only D. dorsalis was taken on the dry slope of the Parque Nacional de Caparaó. Delomys dorsalis Hensel (figs 6, 8, 9, tables 1—4) Hesperomys dorsalis Hensel, 1872: 42, pl. II, figs 16a upper molar row, fig. 16b lower molar row, fig. 26b (m, enlargement from 16b). Hesperomys dorsalis var. obscura Leche, 1886: 696. Brazil: Rio Grande do Sul (type locality, Taquara do Mundo Novo). Akodon dorsalis lechei Trouessart, 1904: 434. New name for Hesperomys dorsalis obscura Leche, wrongly believed preoccupied by Mus obscurus Waterhouse 1834 (a Necromys [= Bolomys]). Delomys dorsalis, Thomas, 1917: 196 — classification. Delomys dorsalis collinus Thomas, 1917: 197 — Brazil: Rio de Janeiro (type locality, Itatiaya, 1800 feet; other specimens from Piquete and Alta da Sierra); holotype, adult male, skin and skull, BM 1422342 Holotype. None specified; 7 paratypes collected by Reinhold Hensel include 5 in spirits, 1 skeleton, and 1 skull; the skeletal material and at least 1 spirit-preserved specimen were used in the original description. The original description of Delomys dorsalis is based on the upper and lower molar rows (both figured), the incisors, and 2 skulls. Measurements of the larger of the skulls are basal length 25.5 mm, nasals 13.3 mm, incisive foramina 6.0 mm, interorbital width 5.3 mm; sagittal diameter of interparietal 4.0 mm, frontal 10.4 mm. Vertebral count is 13 thoracic, 6 lumbar, 32 caudal; 17th vertebra diaphragmatic. External measurements of a large spirit-preserved individual are head and body 135 mm, tail 124 mm, and hind foot 27 mm. The upper and lower molar rows of Hesperomys dorsalis figured and described by Hensel (1872: pl. 2, figs 16a, 16b, 26b) are unambiguous representatives of the species and are hereby designated lectotype. The cotypes or paratypes were originally deposited in the Berlin Natural History Museum. Type locality. State of Rio Grande do Sul, Brazil. Sigmodontine rodents from southeastern Brazil 211 Fig. 6: Delomys dorsalis (MN 31971); skull and molars. GSL, 29.7 mm; molars, 4.8 mm. Distribution. Southeastern Brazil, from Minas Gerais south through the states of Rio de Janeiro, Sáo Paulo, Paraná, Santa Catarina, Rio Grande do Sul, and the province of Misiones, Argentina. External characters. Delomys dorsalis is a medium sized, dark colored, medium to short-tailed, narrow-footed, large eared mouse characterized mainly by a more or less defined blackish middorsal stripe extending from crown or nape to rump or base of tail. Underparts are grayish with thin orange-colored wash, the slaty hair bases usually showing through; division between sides and underparts undefined or marked by a poorly defined, usually broken, orange line from cheeks to tail base. Tail length from about 82 % to 105 % of combined head and body length with average a little under 100 % (table 2); tail thinly pilose, the scales showing through, hair dark brown above, same beneath or with nearly entire ventral surface to one third or one half contrastingly paler; feet long, narrow, whitish above, the short hairs unbanded, the dark skin showing through; mystacial vibrissae usually long, when laid back most vibrissae extend beyond the ears; paired nipples are inguinal, abdominal and postaxial = 6, occasionally a pectoral pair is present to total 8 nipples. Measurements. See table 3. Karyotype. 2n = 82 (Zanchin et al. 1992). Sexual dimorphism. None consistent. Comparisons. The comparisons are between 67 specimens of D. dorsalis of which 49 are from Caparaó, 7 from Iporanga and Apiaí, 7 from Boraceia and 4 from Teresópolis, and 9 of sympatric specimens of D. sublineatus of which 4 are from Iporanga and 4 from Boraceia; and 1 from Teresópolis. Altogether a total of 73 specimens were examined. 212 P. Hershkovitz Table 3: Measurements of Delomys dorsalis from Caparaó, Minas Gerais and Iporanga, Sao Paulo; measurements in mm are of means, extremes, followed by sample size. Caparaó Q | Caparaó © Locality HB! Tail length Hind foot? Ear? CE ZB> IB® BW’ NE$ IF? PL!® UM! D!2 R13 Wwi4 ZP> 132(111—150)25 130(118—146)16 30(26—33)26 20(18—23)26 30.6(28.4—37.8)21 16.5(14.4—17.6)21 5.2(5.0—5.5)21 13.8(12.8—14.4)21 13.8(12.7—15.6)21 7.2(6.6—7.5)21 5.3(4.8—5.8)21 5.0(4.8—5.3)21 9.1(8.5—9.9)21 6.0(5.4—6.5)21 47(25—54)25 3.0(2.8—3.3)21 128(111—138)20 126(118—136)14 30(28—35)24 21(19—24)24 29.5(27.7—31.8)15 16.5(15.6—17.1)15 5.2(4.9—5.5)16 13.9(13.1—14.6)15 13.5(12.4—14.4)16 6.9(6.4—7.5)16 5.2(4.7—5.8)16 4.9(4.6—5.2)16 8.8(8.0—9.5)16 5.8(5.5—6.0)16 43(28—51)23 3.0(2.6—3.5)16 142(133—158)5 139(133 — 150)4 31(29—32)4 7X V+ 23)3 30.8 GOA SES)S 16.7(16.4—17.2)5 5238.2555)5 13.9(13.6— 14.3)5 13.8.281520)3 723 (OS) III) 4.9(4.8—5.0)5 9.0(8.6—9.3)4 6.4(5.9—6.9)4 67(58—75)3 RS NS 144(135—154)6 139(139—141)6 30(29—31)6 22(21—23)6 30.290 S1E9)6 16.6(16.3—16.9)6 3.1(5.052)8 13.9(13.5— 14.2)6 13.9(12.3— 14.2)6 -6.7(6.6—7.0)6 4.9(4.7—5.1)6 4.7(4.6—4.8)6 8.9(8.4—9.3)6 6.1(5.6—6.4)5 58(45 —72)6 ESE DIAS 1 HB = Head and body; ? with claw; ? from notch; *CL = Condylobasal length; > ZB = Zygomatic breadth; $ IB = Interorbital breadth; 7 BW = Braincase width; * NL = Nasal length; ? IF = Incisive foramen; !° PL = Palatal length; 1 UM = Upper molar row; !? D = Diastema; !3 R = Rostral breadth; !* W = Weight, grms; 5 ZP = Zygomatic plate. Pelage of most Delomys sublineatus is shorter, coarser, and paler or ochraceous buff, the pheomelanin subterminal band wider. Pale individuals of the Caparaó series of dorsalis inter- . grade with the sublineatus from Boraceia. The distinctive dark middorsal band is present in all D. dorsalis, albeit ill defined in many. A well-defined middorsal dark band is likewise present in all D. sublineatus from Teresopolis, Iporanga-Apiai and in 4 of 5 from Boraceia. The bright lateral line that marks the color separation of sides of body from underparts is nearly always present and well defined in sublineatus. It is often present in D. dorsalis but usually broken in parts where fur is shaggy. Ears uniformly dark brown in dorsalis, paler or partially unpigmented in sublineatus; tail always shorter in sublineatus (77 % to 92 %) than combined head and body length, and averages shorter than the proportional tail length of dorsalis (82 Y to 105 Yo); mystacial vibrissae appear to be thinner and shorter in sublineatus, when laid back few extend beyond ears; whitish hairs of dorsal surface of hind feet are entirely unbanded in both species; in some, the darkly pigmented skin shows through giving a dusky appearance to the foot; tail dark brown, slightly paler beneath in some, partially bicolor in others more frequently in sublineatus but rarely entirely bicolor; length of outer pedal digits relative to adjacent ones similar in both species. Cranial characters do not distinguish the species although individual and population differences may be noted (Voss 1993: 27) and a slightly greater size of sublineatus compared with sympatric dorsalis (Voss 1993: fig. 13). Among dental traits, the anterolabial conule of m! in D. dorsalis is said to be about the same size as the anterolingual conule. In D. sublineatus, however, it is said to be usually smaller. A tabulation of the comparative sizes of the right and left anteroconules in the Caparaó D. dorsalis revealed the left with 7 larger, 14 smaller, 10 about equal, 13 indeterminate. Whatever the variation of the anterior conules in D. sublineatus, no consistent difference from D. dorsalis can be proven. An enterolophid in m, is consistently present in both species. Specimens examined: Total 73. ARGENTINA: Misiones (Caraguatay, FM 11); BRAZIL: Parana (Roca Nova, 100 m, FM 1); Rio de Janeiro (Teresópolis, FM 2); Sáo Paulo (Boraceia, FM 8; Iporanga, MZUSP 15); Minas Gerais, (Parque Nacional de Caparaó, Arrozal, 2300 m, MN 1; Cachoeira Bonita, 1750 m, MN 17; Pico da Bandeira, 2700 m, MN 2; Segredo, Sigmodontine rodents from southeastern Brazil 213 Fig. 7: Delomys sublineatus (PH 10080); skull and molars. GSL, 31.4 mm; molars, 4.8 mm. 2100 m, MN 6; Tronqueira, 1790 m, MN 1; Terreiráo, 2400 m, MN 15; Vale Encantado, 2100 m, MN 5). Delomys sublineatus Thomas (fig. 7, table 4) Delomys sublineatus Thomas, 1903: 240. Holotype. Old adult male, skin and skull, British Museum (Natural History) no. 3.9.4.58, collected 14 February 1903, by A. Robert, original number 1224. Type locality. Engenheiro Reeve, Espírito Santo, Brazil, elevation between 400—600 m. Distribution. Southeastern Brazil in the States of Minas Gerais, Espírito Santo, Rio de Janeiro, Sáo Paulo, Santa Catarina and Paraná (Voss 1993: 31). Characters. See comparisons with D. dorsalis. Karyotype. 2n = 72, FN = 90 (Yonenaga 1975: 283). Nomenclature. The history of the fossil cranium of a Delomys recovered by Lund from cave deposits near Lagoa Santa, Minas Gerais, and subsequently described as Calomys plebejus by Winge (1887) has been reviewed by Voss (1993: 32). Identification of plebejus as a Delomys, first made by Avila Pires (1960), was accepted without qualification by Voss. Equation of plebejus with sublineatus, however, was rejected. Voss (1993: 34) argued that “synonymizing plebejus Winge 1887 with sublineatus Thomas, 1903 would have the highly undesirable consequence of replacing a type specimen with many characters useful for species discrimination by another with none. Thus, although there is no evidence that the taxon 214 P. Hershkovitz Table 4: Comparative measurements of Delomys sublineatus and D. dorsalis. HB! Tail length Hind foot? Ear? eE2 ZB’ IB® BW’ NL IF? PL! UM!! D!2 R13 W 14 sublineatus Iporanga © 9 137(126—146)4 Gents 29(28—30)4 - 21(20—22)4 30.0(29.0—30.7)4 17.1(16.7—17.4)4 5.2(5.1—5.3)4 13.7(13.3— 14.3)4 13.3(13.0—13.7)4 6.6(6.4—6.8)4 4.7(4.3—5.1)4 4.6(4.4—4.7)4 8.4(8.0—9.0)4 6.1(5.8—6.5)4 60(45—83)4 sublineatus Teresópolis © 9 124, 124, 138 ML, JU, 107 22.30), all 20 28.155303 1 USE, Woes) 4.6, 4.9, 5.0 13.0, 13.4 de WA, aa! 6.0, 6.4, 7.1 Sell, De Do 4.5, 4.7, 4.8 8.2(7.4—8.7)4 6.2(5.6—6.6)4 lo Palo 32 dorsalis Boraceia o 9 118(104—135)7 136(130—145)4 30(28—31)7 19(16— 20)7 29.7(28.2—31.0)7 15.8(14.7—16.5)7 5.4(5.2—5.5)7 13.7(13.4—13.9)7 13:2(12.7—13.8)6 7.3(6.6—8.1)7 5.0(4.6—5.5)7 4.%4.7—5.1)7 8.7(8.3—9.2)7 6.0(5.8—6.2)7 2.8(2.6—3.1)7 dorsalis Boraceia o 9 127(113—135)5 106(100—115)5 27(25—30)5 21(20—22)5 27.9(27.3—28.8)5 15.7(14.8—16.4)5 4.8(4.7—5.0)5 12.8(12.6—13.2)5 12.4(11.5—13.5)5 6.4(6.1—6.9)5 5.0(4.7—5.2)5 4.5(4.3—4.7)5 7.1(7.3—7.9)5 6.0(5.4—6.5)5 48.55 2.8(2.6—3.0)5 DES IT) IHB= Head and body; ? with claw; ?from notch; *CL = Condylobasal length; $ ZB = Zygomatic breadth; 6 IB = Interorbital breadth; 7 BW = Braincase width; $ NL = Nasal length; ? IF = Incisive foramen; PL = Palatal length; ' UM = Upper molar row; !?D = Diastema; 3R = Rostral breadth; !*W = Weight, grms; 15 ZP = Zygomatic plate. represented by the Lagoa Santa fossil is extinct, no biological or nomenclatural purpose is served by synonymizing it with either of the Recent species recognized as valid in this report; plebejus is a nomen dubium that should be used only in reference to Winge's hypodigm.” Reasons given by Voss for rejection of the name Calomys plebejus Lund are personal Opinions with justification. Comparisons. See Delomys dorsalis account. Measurements. See table 4. Origin and dispersal. Phenetic similarity between Delomys sublineatus and D. dorsalis is such that specific distinction between the species is usually predicated on karyotypic differences and sympatry. At one time the two species were regarded conspecific, at another only subspecifically distinct. Zanchin et al. (1992: 168) who studied the karyotypes found “very few elements ... shared by them. This mean[s] that not only centric fusions but pericentrimeric inversions and/or complex rearrangements are responsible for the different karyotypes.” The rearrangements might have been responsible, among other characters of sublineatus, for the short, stiff pelage and selective reduction and partial elimination of the pheomelanin banding of the individual hairs of back and sides. Specimens examined: Total 15. Rio de Janeiro (Fazenda Boa Fe, FM 1; Teresópolis, FM 4); Sáo Paulo (Boraceia, FM 5; Cotia, FM 1; Iporanga, MZUSP 4). Akodon Meyen Akodon Meyen, 1833. Type species Akodon boliviensis Meyen, 1833: 600, pl. 43. The three species of Akodon taken on the western slope of Mt. Caparaó are A. cursor Winge and A. serrensis Thomas, both members of the large-size A. mollis group, and a heretofore undescribed species of the small-size A. boliviensis group. The size Sigmodontine rodents from southeastern Brazil 215 Fig. 8: Delomys dorsalis Hensel, 1872; upper molar row, 4.5 mm; lower, 4.9 mm (figure and measurements from original description of lectotype). categories have been defined by Hershkovitz (1990b). The two large species overlap in most dimensions but otherwise are readily separable. The smaller non-inter- grading third species differs unmistakably from the others and is described below as new. Akodon serrensis Thomas (figs 11, 13, table 6) Akodon serrensis Thomas, 1902: 61. Holotype. Male, skin and skull, British Museum (Natural History) no. 3.7.1.69 collected 15 August 1901, by A. Robert, original number 803. Type locality. Roca Nova, Serra do Mar, Paraná, Brazil, elevation between 930—1150 m. Distribution. Southeastern Brazil, and in the Province of Misiones, Argentina (Justo and Santis, 1977). A. serrensis and A. cursor were taken in collecting stations Vale Verde, Minas Gerais, and Pedro Roxa, Espírito Santo. Characters. Pelage deep, lax, the new pelage of dorsum dark reddish or chestnut with the dark brown or blackish bases showing through increasingly with age; underparts with bright orange wash over otherwise exposed dark bases; lateral line not sharply defined; chin, throat bare except for whitish chin patch; tail dark brown above, slightly paler beneath; pedal claws relatively short, fine, recurved, the digital bristles sparse, scarcely extending beyond tip of claw; manual claws slightly shorter, bristles fewer and shorter. Karyotype. Diploid number 44 (Liascovich & Reig 1989), highest known for the A. mollis group. Comparisons. Distinguished from Akodon cursor by overall smaller size, lax, darker pelage particularly on underside, minute manual claws, larger molars, palatine and maxillary bones more inflated particularly where visible through openings of the palatal foramina. 216 P. Hershkovitz Fig. 9: Skins of Delomys sublineatus showing dark middorsal stripe. All specimens from Sáo Paulo, Brazil. From left to right, dorsal and ventral, MZUSP 26961, FM 26595, FM 141628. Fig. 10: Skins of Delomys dorsalis showing dark middorsal stripe. All specimens from Brazil, Minas Gerais, Parque Nacional de Caparaó. From left to right, dorsal and ventral, MN 31934, PH 10089, PH 10373. Sigmodontine rodents from southeastern Brazil DW Fig. 11: Akodon serrensis (MN 32102) skull and molars. GSL, 27.2 mm; molars, 5.4 mm. Fig. 12: Akodon cursor (MN 32038) skull and molars. GSL, 28.2 mm; molars, 4.5 mm. 218 Table 5: Selected measurements from Winge (1888: pl. Akodon cursor followed by those of his paratypes. Greatest skull length Upper molar row Diastema Length of bulla Palate between m2? Incisive foramen Nasals, length Frontal suture P. Hershkovitz 2957283 4.75 1, fig. 6) of the figured skull of 8.3,..8.3,. o, TAU os a 1407) 4.75 3.0 7.33, MOS ZO ZAS 13:0512.67. 115515912267 10.0 S55) 1G, 142551635 16 Interparietal suture Mandible length Table 6: Sexes compared of Caparaó Akodon serrensis and A. cursor. serrensis O cursor © serrensis Q cursor Q. Head and 105(92—116)28 124(103—157)25 106(95—115)19 111(101—127)19 body Tail Hind foot (c.u.) Weight 84(78—93)21 NAS 2) 25 95(81—106)23 25(23—28)18 85(78—94)10 25(24—28)19 92(81—110)16 25(23—27)20 30(23—45)28 26.0(24.5—26.2)14 50(33—65)27 28.9(26.3—30.4)22 28(18—40)21 25.5(24.4—26.6)14 INMI) 26.7(25.4—27.2)10 Condylobasal length Molar row SOM) ls 4.5(4.0—4.9)21 4.9(4.4—5.4)14 4.5(4.2—4.8)16 Specimens examined: Total 67. Minas Gerais (Parque Nacional de Caparaö, Pico da Bandeira, 2700 m, MN 7; Tronqueira, 1970—2000 m, MN 4; Terreirao, 2400 m, MN 17; Cachoeira Bonita, 1750 m, MN 21; Segredo, 2100 m, MN 9; Vale Encantado, 1980 m, MN ]; locality? MN 1; Arrozal, 2300 m, MN 1; Vale Verde, 1400 m, MN 1; Casa de Julio, MN 1); Espírito Santo (Pedra Roxa, 1100 m, MN 1). Akodon cursor Winge (figs 12, 13, tables 5, 6) Habrothrix cursor Winge, 1887: 25, pl. I, fig. 5 (hindfoot), pl. II, fig. 6 (skull). Lectotype. Winge’s figured skull, basis for the original cranial description, is preserved 1 in the Copenhagen Museum. Syntypes. The original description is derived from a number of skins, skulls and skeletons preserved in the Universitets Zoologiske Museum, Copenhagen. Ximénez et al. (1972) list the following as syntypes: ZMK 221, 222, 224, 237, 6-VIII-1847 9 , 5-VIII-1847 9 , 29-VI-1851, the last three collected by Reinhardt, and an unlisted unnumbered specimen from the Lund collection. Thomas (1902: 60) noted that “one of the co-types . . . [is] in the British Museum.“ Type locality. Lagoa Santa, Minas Gerais, central Brazil. Distribution. Southeastern Brazil from Bahia south, and the central plateau; Uruguay, eastern Paraguay, and Misiones in northeastern Argentina. Sigmodontine rodents from southeastern Brazil 219 Fig. 13: Akodon cursor and A. serrensis, skulls and molars compared. See figures 11, 12 for data. 220 P. Hershkovitz Characters. Pelage soft, deep, adpressed; upper parts individually variable from buffy to dark brown; a poorly defined dark longitudinal band sometimes present middorsally, sides paler, the pheomelanin subterminal band of hairs wider, underparts like sides, lateral line of separation absent but slaty basal portion of hairs showing through weakly; muzzle tip more or less dark brown; tail dark brown, thinly covered with short stiff, blackish hairs, the scales fully exposed; cheiridia brown, digits often whitish; ears short, brown; claws short, weak, hardly one phalanx long, middle digits narrowly webbed. Karyotype. Diploid number, 24 (Liascovich & Reig 1989), lowest known for the A. mollis group. Measurements. Table 6. Specimens examined: Total 53. Minas Gerais (Parque Nacional de Caparaó, Centro Visitantes, 1300 m, MN 18; Vale Verde, 1400 m, MN 22; locality unrecorded, 3); Espírito Santo (Pedra Roxa, 1100 m, MN 10). Akodon mystax, new species (figs 14, 15, 16, 24, table 7) Holotype. Adult female, skin and skull Museu Nacional, Rio de Janeiro, no. 31910, collected 26 October 1992, by Philip Hershkovitz, Scott M. Lindbergh, Alfredo Langguth and Barbara E. Brown; original number, PH 10425. Etymology. The blackish band from tip of rostrum to corner of each eye in many males and fewer females suggests a mustache. Type locality. Arrozal, Pico da Bandeira, western slope Mt. Caparaó, Minas Gerais, Brazil, elevation 2300 m. Distribution. Known from type locality only; captures were at 2300 m, 2400 m, and 2700 m on the western slope of Mt. Caparaó, Minas Gerais, Brazil. Diagnosis. Thin dark rostral band (mustache) usually present, cheiridea unpigmented, tail bicolor; infraorbital foramen and zygomatic plate hardly visible viewed from above; mesopterygoid fossa wide with posterior palatal margin rounded or square; incisive foramina long, the palatal tips rounded, turned outward. Description of holotype. Coloration of dorsum from snout to rump brown, the cover hairs with fine blackish tips, followed by a narrow ochraceous orange subterminal band, the bases dark gray; sides paler, the pheomelanin subterminal band wider, the color dominant on ventral surface but with plumbeous basal portions of hairs exposed; tail brown above, sharply paler beneath; dark rostral band present from snout to side of face; ears brown, cheiridea unpigmented; length of digits normal, claws short, unspecialized. Variation. Pelage deep, fine, almost lax, about 10 mm long on back; dorsum from rostrum to tail base buffy to grayish brown, the subterminal band of hairs buffy to ochraceous becoming paler toward sides of body; underparts paler, with an ill-defined lateral ochraceous buff line of demarcation, the slaty hair bases showing through; tip of rostrum usually with narrow dark brown patch often extending across each side as a mustache to or near corner of eye; ears brown, in some rimmed pale buff; cheiridia pale above and below; first manual digit with nail not extending to base of second digit; digit II with claw extending to base of 2nd phalanx of III; digit IV slightly shorter than III, V with claw extending to base of 3rd phalanx of IV; pedal digit I with claw extending to middle of Ist phalanx of II; length of digits II, III, IV, subequal, the middle slightly longer; interdigital membranes absent; claws short, recurved, the pedal longer than the manual; tail distinctly bicolor, the hairs nearly concealing the scales. The general aspect is of a grayish buffy animal. Sexual dimorphism. Males are larger than females in external dimension and tend to have slightly larger skulls (table 6). Most size differences in this case, however, are inconse- quential; blackish rostral patch or mustache in nearly all males, is less prevalent and extensive in females. Sigmodontine rodents from southeastern Brazil N Fig. 14: Mustached Akodon, Akodon mystax (holotype MN 31910 9); skull and molars; GSL, 24.3 mm; molars, 3.9 mm; Arrozal, Parque Nacional de Caparaó, Minas Gerais. Comparisons. Akodon mystax is a member of the Akodon boliviensis group or size class defined by Hershkovitz (1990b: 3). Males of all species of the boliviensis group are larger than females in external dimensions. Nearest relatives appear to be Akodon sanctipaulensis (Sao Paulo), A. lindberghi (Brasilia, D. F.), and A. azarae (Paraguay, Argentina and Uruguay). Similarities between Andean and Atlantic region akodonts of the boliviensis group may be more than superficial, but the diagnostic characters of mystax are significant; the karyotype is as yet unknown. The posterolophule (pseudomesoloph) often fused with the posterostyle usually absent in A. mystax is not consistently present in other species. Morphometric differences between the species are slight; comparative measurements of most other members of the boliviensis size class are given by Hershkovitz (1990b: 4—5). Measurements. Table 7. Remarks. Several of the specimens were taken in a small stone house in Terreirao occupied only by the mice until Scott Lindbergh and Alfredo Langguth moved in. A mustache like that of Akodon mystax is common among sigmodontines, marsupials and other mammals usually as an individual variable. Specimens examined: Total 21. Minas Gerais (Parque Nacional de Caparaó, Terreirao, 2400 m, MN 13; Arrozal, 2300 m, MN 5; Pico da Bandeira, 2700 m, MN 3). Thaptomys Thomas Thaptomys Thomas, 1916: 339. Type species Hesperomys subterraneus Hensel, 1872 (= Thaptomys nigrita Lichtenstein, 1829). The genus Thaptomys Thomas, 1916, one of the most distinctive of the akodontine assemblage has experienced an equivocal systematic history. Ellerman (1941: 406, 409) characterized Thaptomys as “quite a well differentiated group,” but treated it VEZ Hershkov E 222. sg T—-0'DE'I AS p—T'v)€'p 9(S' p—L'E)6'€ Xr 7—8'1)0°7 os ET ET € 9(19—S"S)8'S 9(6'6—£"8)0'6 AS TIT IDH TI 9(9'p—£'b)v"p AS TI—6 17 TI 9(9°€T—6 ITL 7 9(0°ST—I 7) HZ 9(7Z7T—6I1)IZ 9(LL—¥9)OL 9(LOI —£8)96 $ „sısuaranoq AP TI—-TDEI SUL'v—b'p)s"p 9(0'p—S'E)8 € AET—-L'DOT 8 E-6 DEE s(S'9—8"S)1'9 s(9'01—0'8)r'6 9L TIT IDO TI AL PT PSP 9(6'71—8'1Dv"71 S(0'y7—v" TT EZ S(0'97—S"b7)1 "ST AET07) TZ (78 —79)CL S(001—76)86 nO) „sısuaraıoq 9S I—O DET L(v'v—0'v)7'p 8(7yY—8"8)0'p 8(€°T—6 DIT SlEE-SDT'E 879 —9'S)6°S 8(S'6—7'8)6'8 89° TI—L'ON7 TI sl H—6 Er 8(971—b" 11071 S(T’ET— 806 IT SL’ 97-8 "78" EZ L(S’07—61)07 L(L9—LSIZ9 L(TOL—ZÓ6)L6 ó ¿OD.L DZD ‘Wa (tuequiry) oung ‘Aad ‘sisuatayog uopoyy y Wa (B1N2110L) $241/ souang :WNILNADUV “901020 uopoyy ¿ NW (u 00L7—O0EZ 84 Teuonen OvIedeD) S1D189 soul: “TZ VAS Xvısdw UOPOAY ¿ OISIE NIN “u Q0EZ TEZOLIY Yıeq [BUoneN OvIedeD) simiaH souW “TIZV AA XDISAM UOPOyYY | 6(91—0'DE€'I 6(S' 7-6 Tp 6(Er—L TP AS TU DET 6(9£—0'£)£ € 6(S'9—8"S)1'9 6(s'6—8"L)0'6 6(911—6'0DT'T1 6(£ p—8')0'p 69 7I—-€ TDCI 69 PZ —6 IE 77 6(1'9T—8 ETL HZ 6(S I7—0Z)S 07 6(ZL—LS)S9 6(€lL1—v6)ZOI po) ¿ODADZD LVS 1—6€' D6v'I 8(Lv—T'p)E€'y L(v"9—0'9)€'9 8(ep—9"€)6'€ 8UZ—E DLT 80 E-LO8T 8(L9—p"9)L'9 8(0'6—0'8)9'8 8(711—S'0D6'01 8(9'p—1"b)p'p sr TI—9ID6 TI LT ET-OTIITZ SI TT— I EDE HT L(97—€1)81 8(pI—Z DEI 8(61 —91)9"L1 8(89 —65)L'€9 8(66—L8)b"06 6) XDISsai u s(w9T—0r'DES’I L(0'S—+"p)8'p Z1(8°9—7'9)5°9 TH pS E)6'€ ZHO'Z—9"1)8"1 110°E-9 87 THE L—L'S)8'9 ZHL'OL—E"8)6'8 TIE 11—S 016 01 HQ r—vr' ps UL ZI—SIDUZI OL(0'97—S"ZZI)T EZ Ty ST—0'P7)9 PT ZHST—91)61 IIWI—TDEI II@T—LD61 TI(EL—€9)L9 IT(TOT—S8)S6 O ¿XOJSÁ 6 dAI0J0H ¡XOJSÁ essoJ PIOSA19JdosaN ypIm “wunısoyN JWOISEIA MOI Je[OW 9Je]d INewogAz yJ3u2] ‘SJefed U9WE.IOJ DAISIDUT yY13U9] JeseN YIPIM ISPIUILIS YIPIM [831Qq.10 1930] YIpesiq IewosAz y/13u9] [eseqo¡Apuo) y13ua] [NYS 15978910) MIL M leq 100} PUIH yysug] [eL Apoq pue pesH ‘SISUAIAIIOG “Y PUR 20.020 “y yyım paredulos xvısdu uopoyy JO SJuSwsANsesW :/ 2IgeL Sigmodontine rodents from southeastern Brazil Fig. 15: Skulls of three species of Akodon: A, A. mystax (holotype MN 31910 Q ); GSL, 24.3 mm; Arrozal; B, A. azarae (FM 27616 ©), GSL, 23.9 mm; Uruguay, La Lata; C, A. boliviensis (FM 51293 9); GSL, 24.0 mm; Perú, Puno, Yunguyo. 224 P: Hershkovitz Fig. 16: Molars of three species of Akodon, same shown in figure 15. Upper row, upper molars, A, A. mystax, 3.9 mm; B, A. azarae, 4.0 mm; C, A. boliviensis, 4.0 mm; Lower row, lower molars, A. mystax, 4.0 mm; A. azarae, 4.1 mm; A. boliviensis, 4.0 mm. Fig. 17: Bolivian Akodon boliviensis (FM 51293 ); skull and molars; GSL, 24.0 mm; molars, 4.0 mm; Perú, Puno, Yunguyo. Sigmodontine rodents from southeastern Brazil 225 Fig. 18: Thaptomys nigrita (PH 10112 o); skull and molars; GSL, 24.5 mm; molars, 3.6 mm; Pico da Bandeira, Parque Nacional de Caparaó, Minas Gerais. as a subgenus of Akodon. Cabrera (1961: 453) followed without comment. Reig (1987: 358), with judgement derived from unrevealed sources declared that “separa- tion of Thaptomys from Akodon s. s. is unwarranted.” Nothing, he averred, distinguishes the taxon “beyond the limits of variation with Akodon s. s.” The peculiarity of a single pair of prostate glands in Thaptomys nigrita was not appreciated by Reig, and the highly advanced fossorial adaptations of the mouse (table 12) were dismissed as “alleged,” and “too incipient to deserve any special taxonomic treatment.” It has since been shown (Hershkovitz 1990b: 6), “that the short tail, long manual claws, heavy skull, ridged parietals, short, thick rostrum with long nasal bones, wide interorbital region, squared braincase, relatively small molars, long powerful proodont incisors projecting well beyond the nasals, and the diploid chromosome number 52, are singly or in any combination definitely non-Akodon? Musser & Carleton (1993: 691) however, retained Thaptomys in the synonymy of Akodon. Thaptomys nigrita Lichtenstein (fig. 18) Mus nigrita Lichtenstein, 1829 p. 35, fig. 1 (animal). Thomas 1902: 62 — synonyms: subterraneus Hensel 1872, henseli Leche, 1886, fuliginosus Wagner, 1845, orycter Lund, 1841. Holotype. Male, skin only, Zoological Museum, Berlin. Type locality. Vicinity of Rio de Janeiro, Brazil. Distribution. Southeastern Brazil from the State of Bahia south through Minas Gerais, Sao Paulo, Paraná, Santa Catarina, Rio Grande do Sul into eastern Paraguay, and in Argen- tina, the province of Misiones. 226 P. Hershkovitz Table 8: Thaptomys nigrita: Measurements from three localities. RN Pico de la Bandeira > A 5 Head and body 101(95—107)11 10897105) 9982 UL) Tail 46(40—50)11 47(42—50)9 42(30—50)74 Hind foot OKO Aaya 18 34725)9 19(16—20y7 Ear 10(10—12)11 11.5(10—12)8 1110—12)7, Weight ZAS ZO) 25(17—30)4 26(20—31)4 24.1(23.3 —24.7)5 23.6(22.5—24.3)6 15.20W251520)6 5.0(4.4—5.3)6 IO LO 8.5(8.2—8-7)5 25(22.7—25.4)6 24.2(21.4—25.0)6 13.4(12.1—14.2)6 4.9(4.7—5.0)8 11.7(11.5—12.0)8 8.8(7.5—9.2)8 Greatest skull length Condylobasal length Zygomatic breadth Interorbital breadth Braincase width Nasal length 24.5(23.9—25.3)7 24.0(22.8—25.0)7 13.4(12.8—13.9)6 5.1(4.8—5.3)7 11.7(11.3—12.1)7 9.2(7.5—10.2)6 Incisive foramina ZOO 5.5(4.8—6.0)9 SAS) Palatal length 3.9(3.6—4.7)6 3.7(3.3—4.4)8 SAS, Zygomatic plate NON )S ZO ZO) 2-2) Molar row SMC SO SMS GO SIMS 218) 'MN and FM; ? MZUSP and FM; 3Includes MN: Caira Laguna, Santa Teresa, 3; Castelo, 3; Cachoeira de Hapanurim, 1; 4 Includes some “bobbed-tailed” specimens. External. One of the smallest mammals of the Caparaó region; pelage velvety, coloration brown, ranging from a milk chocolate color to a dark brown on dorsal surface; sides and underparts with orange, the dark slaty hair bases showing through; forwardly directed pelage of neck and throat with less orange; tail half or less length of head and body combined, the short stiff hairs brown above, whitish below, not concealing scales; ears short, eyes small; manual claws long, fine, the middle pair longest, pedal claws shorter. - Eight teats are said to be the norm for Thaptomys (Hensel 1872; Gyldenstolpe 1932; Davis 1947: 6). I count 6 in 2 Caparaó females, 6 in 3 of the FM collection but nipples, especially the pectorals, are difficult to detect in dry skins during the wrong time of the breeding season. Cranial. Rostrum short, nasals not extending to front of incisors; supraoccipital bone inclined forward; zygomatic arches delicate, the perpendicular plate broad, visible from above; interorbital borders smooth, slightly concave and divergent; incisive foramina longer than molar row and extending to plane of metacones of first molars; palate long, extending behind to level of third molars; posterolateral palatal pits present but highly variable in size; spheno- palatine vacuities absent or not indicated. Dental. Upper incisors proodont extending well anteriad to tips of nasal bones; molars moderately high crowned; unworn crowns terraced, anteromedian flexus well defined, the procingulum biconulate; talon and trigon of M'? subequal in width and length, without postcingulum; paralophule (pseudo-mesoloph) usually fused with metalophule; third molar about as large as talon of second, cusps of opposite sides oblique; first lower molar with anteromedian flexid, opposing cusps oblique; protolophid and posterolophid present; posteroflexid of m, absent. Worn upper and lower crowns acquire the typical akodont dished-out eight-shape appear- ance. Measurements. Table 8. Remarks. The later named forms of Thaptomys (fuliginosus Wagner, 1845; subterraneus Hensel, 1872; henseli Leche, 1886) were distinguished by coloration only, and the fossil orycter Lund (1841) by cranial characters. The traits are highly variable, none consistently distinctive of any one described form. The abbreviated rostrum and thick, projecting proodont incisors appear to be adaptations for tunnel burrowing. Sigmodontine rodents from southeastern Brazil DN, Comparisons. Small size, one of smallest of Caparaó mammals, markedly proodont incisors, short thick rostrum and long manual claws among other characters distinguish Thaptomys from all other mammals of the region. Habits and habitat. In his description of Hesperomys subterraneus (= Thaptomys nigrita) from Rio Grande do Sul, Hensel (1872: 45) noted that the species lived in virgin forest and made burrows like that of the European Arvicola arvalis (= A. terrestris). Whether or not it nested in family groups like the microtine was not mentioned. The home range of southeastern Brazilian marsupials and mice was estimated by Davis (1945: 124) by recapturing marked animals in second-growth forest near Teresópolis, Rio de Janeiro. The three common species, Delomys dorsalis, Akodon arviculoides and Thaptomys nigrita were recaptured a total of 152 times. Only 15 individuals were recaptured more than 100 m along the trail between farthest capture sites. The data led Davis to assert that the radius of the home range of each of these species was less than 100 meters. In a second, more detailed account, Davis (1947: 6) described Thaptomys nigrita as an aggressive little mouse, “the second [after Akodon arviculoides| most common mammal in the forests [around Teresópolis, Rio de Janeiro]. It is found under logs and tree roots and even makes tunnels in the leaf litter and soft earth. When released alive from the hand it im- mediately washes its face and cleans its fur and then disappears in a hole. This mouse is very fierce and quickly inflicts a sharp bite when handled. However, several were trained to come out of their holes to look for kernels of grain; they did not eat outside the hole, but grabbed the seed and ran back into the hole. This species is definitely diurnal, for many individuals were captured in the daytime and were regularly seen in the daytime. One mouse was caught 3 times in one afternoon. The sex ratio was about even (80 males and 75 females). There are 4 pairs of mammae. Only 4 pregnant females were examined; 1 had 3 embryos, 2 had 4, and 1 had 5. This little mouse has no conception of climbing, but when placed on a limb runs in any direction and falls to the ground. Recaptures of marked individuals indicate that these mice stay in one place for a long time.“ A test for burrowing was made with an adult I captured live at the mouth of a burrow 30 November 1989 in second-growth forest fringing a banana patch in the Iporanga state park. The mouse was introduced into a 12x 12x 12 wire mesh cage filled nearly to the top with clay. The animal burrowed the instant freed and in seconds dug itself out of sight. The finished tunnel system had three outlets. The first was on top where excavation started. The other two were each at opposing corners of the cage. The mouse was fed corn kernels, peanuts, seeds, and watermelon deposited inside the cage. All was taken into the burrow for eating. The mouse came frequently out of hiding during the day. When frightened it dived into the nearest hole each about 2 inches in diameter but the mouse could negotiate a tunnel a centimeter wide. Specimens examined: Total 24. Minas Gerais (Parque Nacional de Caparaó, Segredo, 2100 m, MN 1; Vale Verde, 1400 m, MN 4; Cachoeira Bonita, 1750 m, MN 3; Terreirao, 2400 m, MN 3; Pico da Bandeira, 2700 m, MN 1). Sáo Paulo (Iporanga, MZUSP 10). Brucepattersonius, new genus (figs 19—26) Type species. Brucepattersonius soricinus (new species). Etymology. The genus is named in honor of Bruce D. Patterson as an expression of my admiration for his mastery of all aspects of mammalogy whether in the field, laboratory or classroom, and in appreciation for his valued friendship. Included species. Oxymycterus iheringi Thomas and the following described as new: B. soricinus (type species), B. igniventris, B. griserufescens, B. albinasus, and an unnamed species (not seen) from Misiones, Argentina. Diagnosis. Head and body combined less than 140 mm; tail from shorter to a fourth longer than head and body combined; eyes minute; rostrum long, tapered; longest manual claw 3 mm or less, pedal 1 or 2 mm longer than manual; digital 228 Sigmodontine rodents from southeastern Brazil N \\ | N N N NX VL Wy N SS N Wy HE SS IE SN ' Wi YY SN y / Fig. 19: Left hands of sigmodontine rodents. Upper row dorsal surface, from left, Chelemys megalonyx, fossorial; Oxymycterus amazonicus, nonfossorial surface digger; Brucepat- tersonius griserufescens, terrestrial semiscansorial. Lower row, palmar surface of same hands in same order as shown above. vibrissae on three middle manual digits extremely sparse or absent on digits I and V; sphenopalatine vacuities absent; nasals produced anteriorly to or beyond plane of incisors, tips rounded or bluntly pointed; rostral width less than that of inter- orbital region; molars tetralophodont, quadritubercular, hypsodont; first molar with anterior median fold; mesoloph(id) present to absent; upper incisors short, narrow, orthodont; cuspids more or less subprismatic. Distribution. Southeastern Brazil from the state of Rio Grande do Sul north into eastern Minas Gerais and probably some part of Espírito Santo; in Argentina known from the province of Misiones. Sigmodontine rodents from southeastern Brazil 229 anterior end of duodenum incisura angularis esophagus fornix ventricularis bordering fold Fig. 20: Unilocular-hemiglandular stomach of Brucepattersonius griserufescens; A, ventral aspect; B, the same bisected, the ventral half removed. 230 P. Hershkovitz Description. Length of head and body combined (HB) between 75—140 mm; tail between 75—120 mm; ear about 14 % —20 Y% HB; hind foot with claw between 23 Y —28 Y HB; adult weight between 18—50 g; general coloration of dorsum dark brown, sides of body usually paler with greater admixture of pheomelanin; underparts and inner sides of limbs dominantly grayish or dominantly ochraceous orange (pheomelanin); rostrum long, tapered, eyeballs small, diameter 1 to 2 mm; hind foot long, narrow, base of middle digits webbed; volar pads 6; pedal claws 3—4 mm; manual claws weak, longest 2—3 mm; manual digit I vestigial, inungulate; gall bladder present; stomach unilocular-hemiglandular with glandular epithelium contained in a pouch-like diverticulum (fig. 20). Nasal bones long, tapered, produced to front of incisors or 1 to 2 mm beyond, tips rounded or bluntly pointed, not squared, flared or trumpet-shaped; premaxillary bones not reaching nasal tips; sphenopalatine vacuities absent; zygomatic arches weak, hardly spread beyond greatest width of braincase; zygomatic plate narrow, markedly reclined, hardly visible viewed from above; interorbital region smooth, wide; braincase smooth, subglobular, interparietal length 1 or 2 mm; incisive foramina long, narrow, terminating slightly behind anterior plane of M!; palate produced to or slightly behind posterior plane of M?. Molars tetralophodont, hypsodont; molar rows parallel-sided or slightly conver- gent posteriorly; first molars with median fold, enamel of anterior margin of m, more or less crenulate; mesoloph(id) present to absent; upper cusps ovate to sub- triangular, cuspids subprismatic, the opposing pairs in echelon. The following description of the molars is based primarily on comparison of the type species of Brucepattersonius with that of the type species of Oxymycterus (O. nasutus). Less worn molars of other than the type species would reveal characters probably undetected in the worn teeth described here. For molar terminology see figs 332% Upper Molars M!. Procingulum with well defined anterior median fold, the anterolabial lophule (5) and anterolingual lophule (c) well defined; anteromedian style (a) absent; other elements absent in B. soricinus but present in Oxymycterus nasutus are anterolophule (d), anteroloph (A), paralophule (m), and enteroloph (w); mesoloph (n) present to absent in Brucepattersonius. M?. Procingulum vestigial, the major and minor enamel folds or flexi shallow but the four principal cusps well defined; mesoloph (n) absent. M?. Very simple with paraflexus (3) and entoflexus (9) reduced to enamel islands. Lower Molars m}. Procingulum with shallow anterior median fold; the anterior enamel margin crenulate; labiolophulid (d) present, usually fused with protostylid (2); ectolophid present; mesolophid (v) present; lophids or lophulids absent in B. soricinus but present in O. nasutus are anterolophid (Ah), anterolophulid (f), mesolophulid (s), metalophulid (x), entolophulid (x); median fossette (b’) and posterior fossette (c”). m,. Short mesolophid (v) fused with metalophulid (u); ectolophid (rn) and entolophid (x) absent; posterior fossette (c’) present. Sigmodontine rodents from southeastern Brazil 231 ms. In this and mp, the superflexid (7) defines what remains of the procingulum; lophids and lophulids absent, enamel folds indicated by indentations of the marginal enamel; m; about 2/3 bulk of m.. Comparisons. Microxus differs by much smaller size; claws longer, as in Oxymyc- terus; palate oryzomyine in length with posterolateral pits; mesoloph complicated or pseudopentalophodont with fused paralophule (12), metalophule (p) and mesostyle (0); mesolophid not certainly distinguishable. Oxymycterus differs by bulbous rostrum; much longer, heavier claws, the manual as long or longer than the pedal; nasals more or less parallel-sided, tips square, often trumpet shaped; cusps (ids) more or less ovate, the inner and outer pairs more nearly opposite. Abrothrix, once linked with Oxymycterus, is unrelated. Among the more obvious differences between it and Brucepattersonius are large or normal-sized eyeballs, simplified baculum, mesopterygoid vacuities present, zygomatic arches expanded, zygomatic plate stout, upright, anterior median folds absent in upper and lower first molars. Relationship. Brucepattersonius is the akodontine apparently most nearly related to Oxymycterus, with which it had been confused. All distinctive characters of Oxymycterus, most notably larger overall size, long front claws, shorter hind claws, small ears and eyes, short tail, long snout, stomach morphology, diet, dental and cranial characters generally and geographic and ecological orientation, can be derived from an ancestral form near Brucepattersonius. Distribution, associations. Brucepattersonius is the soricine sigmodontine supreme with its small size, long, slim body, short limbs, elongate tapered snout, small eyes, short ears and mainly if not entirely insectivorous diet. It is remarkable that the existence of this distinctive faunule, relatively uncommon but widespread over the well studied southern half of the Atlantic rain forest, should not have been appreciated until now. Brucepattersonius with its six currently recognized (including One unnamed) species proves to be one of the more speciose of the akodontine genera. The sylvan associates within its geographic range are Blarinomys breviceps, Akodon serrensis, A. sanctipaulensis, and A. mystax described here. A. cursor is essentially pastoral where trapped in Caparaó but the habitat was formerly sylvan. The monotypic Thaptomys nigrita is the burrowing sylvan form. The pastoral Oxymycterus is represented by the large O. hispidus or O. rufus and a smaller species described here as new. The sylvan oryzomyine-thomasomyine sigmodontines commonly occuring with Brucepattersonius are Delomys (D. dorsalis, D. sublineatus), Oryzomys (O. ratticeps, O. capito, O. intermedius), Oligoryzomys (O. microtis, O. nigripes, O. fornesi, and possibly one or two undescribed forms). A species each of Rhipidomys, Rhagomys(?), Oecomys, and Nectomys are the known remaining sylvan associates. Key to species of Brucepattersonius la. Rostrum white; tail length more than combined head and body length...... albinasus lb. Rostrum dusky; tail longer or shorter than combined head and body length........ % 232 P. Hershkovitz 2a. Tail less than 90 % as long as combined head and body length; underparts grayish with ochraceous: Wash siii an er AS FE soricinus 2b. Tail 90 Y as long or longer than combined head and body length................. 3 32: Underparts dominantlyierayishri. sr er Se Ae eee eee eee griserufescens 3b: Underparts dommantly reddish or orange r er eee eee igniventris Note: The unidentified specimens from Tobunas, Puerto Gisela and Dos de Mayo with measurements given in table 8 represent an unknown species identified as Oxymycterus iheringi Thomas by Massoia (1963a), and Massoia and Fornes (1969). They have not been seen by me. They occur well outside the geographic range of those described here but may be congeneric. Brucepattersonius soricinus, type and new species (fig. 21) — Soricine Brucie Holotype. Adult male, skin and skull FM no. 94480, collected 26 July 1961, by A. M. Olalla, original number 1290. Type locality. Ribeiráo Fundo, Sáo Paulo, Brazil. Distribution. Known from type locality, and nearby localities Primeiro Morro and Morre- tinho, southwestern Sáo Paulo, Brazil. Diagnosis. Size smallest of genus except B. albinasus (new), described below, tail shortest relative to head and body length, general coloration brownish, underparts gray with ochrace- ous wash, skull and rostrum broadest. : Description of holotype. Pelage moderately adpressed, of dorsum about 8 mm long not concealing ears; crown to rump brown, cover hairs with tips minutely tipped blackish, followed by ochraceous orange band, remainder plumbeous; guard hairs entirely blackish; sides Of trunk paler than back, the ochraceous bands wider, basal portions of hairs gray, the whole merging into grayish chest and belly with hairs broadly banded pale ochraceous, bases dark gray; throat, chin dominantly gray, the hairs directed forward; coloration of fore and hind limbs like that of trunk; longest claw of forefoot measured in straight line about 2 mm, of hind foot 4 mm; digital vibrissae sparse; facial vibrissae short, hardly reaching ear base when laid back. Cranial. As described for the genus. Dental. Anterior median fold or flexus present in first upper molar; mesoloph, protoloph, anteroloph and enteroloph absent; opposing cusps slightly oblique, form subprismatic. Lower first molar with anterior flexid; mesolophid absent in first molar, sometimes present in second; labiolophulid present; anterolophid absent; ectolophid present; cusps more or less subprismatic. Measurements. Table 9. Type series. Essentially like holotype but average darker on dorsal suface, more reddish on chest, belly. Comparisons. Smaller than other congenerics except albinasus, underparts less gray than in B. griserufescens (new) described beyond; more nearly like igniventris (new), see below, in size, proportions, coloration except throat, chin more gray, chest less orange; skull and rostrum broader. Specimens examined: Total 6. Sáo Paulo (Ribeiráo Fundo, FM 1; Morretinho, FM 1; Primeiro Morro, FM 4). Brucepattersonius igniventris, new species (fig. 22) — Red-Bellied Brucie Holotype. Adult male, skin and skull. MZUSP no. 27000, collected 1 December 1989, by Philip Hershkovitz, Scott Lindbergh and Barbara E. Brown, original no. 9871. Type locality. Iporanga (Petar) State Park, southwestern Sáo Paulo, Brazil. Sigmodontine rodents from southeastern Brazil 233 Distribution. Known only from type locality in the forested (now second growth) high- lands of southwestern Säo Paulo. Diagnosis. Reddish brown dorsally, reddish orange ventrally; tail shorter than head and body combined; longest manual/pedal claws on digits II, III; muzzle attenuated, nasal tips projected well beyond incisors. Description of holotype. Most reddish brown species on upper parts, most intensely reddish orange on underparts, limbs, sides of head; a broad orange lateral line between reddish underparts and reddish brown upper parts. Description of type series. Pelage of dorsum fine, soft, adpressed, the hairs 7—8 mm long; coloration of snout to rump reddish brown, sides slightly more reddish merging into dominantly reddish orange of cheeks, chin, throat, arms, belly, legs; the chest more uniformly reddish than other parts, the slaty bases of hairs showing through on belly; tail uniformly brownish, thinly clothed, the scales clearly visible, short thin pencil whitish to brownish; ears brown nearly hidden in fur; upper surface of fore and hind feet pale to dark brown; claws little recurved, short, weak; digital vibrissae of three middle toes whitish, vibrissae sparse or absent on outer toes; manual claws small, thin, seemingly ineffectual; facial vibrissae thin, the longest barely reaching ear base when laid back. Cranial. Nasals slender, tips rounded; palate long, extending to posterior plane of m?. Dental. Worn in all 3 available specimens; anterior flexus (id) of m- evident. Measurements. Table 9. Comparisons. Coloration more reddish throughout, particularly of underparts, than in all other known forms; skull smaller than that of griserufescens, and the Argentine (Misiones) species; nasals longer more slender than in B. soricinus. Specimens examined: Total 3. Sáo Paulo (Petar, Iporanga, MZUSP 3). Brucepattersonius griserufescens, new species (fig. 23) — Gray-Bellied Brucie Holotype. Adult female, skin and skull, Museo Nacional, Rio de Janeiro, no. 32016, collected 7 October 1992, by Philip Hershkovitz, Scott M. Lindbergh, Alfredo Langguth and Barbara E. Brown, original number PH 10234. Type locality. Terreiráo, Parque Nacional de Caparaó, Minas Gerais, Brazil, elevation 2400 meters. Distribution. Known only from the western slope of Mt. Caparaó, Minas Gerais; collected from 2100—2400 meters above sea level in remnants of Atlantic forest. Diagnosis. Largest species of the genus; tail nearly as long to longer than head and body combined, underparts dominantly grayish. Description of type series. Dorsum uniformly brownish from rostrum to tail base, pelage silky, hairs about 10 mm long, narrow subterminal band ochraceous orange, gray hair bases entirely concealed beneath long lax fur; sides of trunk and limbs like back; underparts grayish, variably washed pale ochraceous and more or less defined from sides; tail uniformly brown, the dorsal hairs about 1 scale long, ventral hairs, tip and pencil whitish, about 3 scales long, the scales showing through; ears brown, partly hidden in fur; hands and feet pale brown above, the palms pigmented or unpigmented, the soles brown; manual claws 2 mm long, pedal claws 4 mm; tail of holotype bobbed. Remarks. Average tail length about equal to average combined head and body length, but 6 of 9 intact tails are longer, 2 slightly shorter, 1 same, as combined head and body length. Measurements. Table 9. Cranial. Muzzle long, slender, tapered, nasal tips rounded; zygomatic plate little exposed seen from above; interorbital edges rounded, braincase smooth; spread of zygomatic arches 234 P. Hershkovitz Fig. 21: Soricine Brucie, Brucepattersonius soricinus (holotype FM 94480 0); skull and molars; GSL, 27.8 mm; upper molars, 4.2 mm; Ribeiräo Fundo, Säo Paulo. and molars; GSL, 25.5 mm, molars, 4.5 mm; Petar, Iporanga, Säo Paulo. Sigmodontine rodents from southeastern Brazil 235 Fig. 23: Gray-bellied Brucie, Brucepattersonius griserufescens (holotype MN 32016); skull and molars; GSL, 26.4 mm; molars, 4.5 mm; Terreiráo, Parque Nacional de Caparaó, Minas Gerais. about same as greatest width of braincase; interparietal bone 1.5 x 8.0; incisive foramina extending to first enamel fold (supraflexus) of M'; palatal bridge produced to posterior plane of M?; width of mesopterygoid fossa about 2 mm; sphenopalatal vacuity absent; hamular processes of pterygoids destroyed; right auditory bulla, left side braincase damaged; capsular process of lower incisor root weakly pronounced. Dental. Upper incisor thin, short, orthodont; lower incisor shorter than diastema; anterior median fold of M' deep; anterolophule fused with paracone; mesoloph fused with paracone and metacone in M'?; median fossette present in M?; m, with anteromedian fold; mesolophid present; well developed labiolophulid (d) present; ectolophid of m, fused with paralophulid; hypoconulid and ectostylid fused in m,; hypoconulid free in m,. Comparisons. Overall size largest of genus; tail, ears longer; underparts dominantly to entirely grayish contrasting with more reddish underparts of B. igniventris and B. soricinus; underparts as in the much smaller a/binasus (described below); nasals longer, more slender, interorbital region widest of the genus. Specimens examined: Total 15. Minas Gerais (Parque Nacional de Caparaó, Terreiraó, 2400 m, MN 9; Segredo, 2100 m, MN 3; Cachoeira Bonita, 1450 m, MN 1; Pico da Bandeira, 2700 m, MN 1; locality unrecorded, MN 1). Brucepattersonius albinasus, new species (figs 24, 25, 26) — White-Nosed Brucie Holotype. Adult female, skin and skull, carcass in alcohol, Museo Nacional, Rio de Janeiro, no. 32017, collected 7 October 1992, by Philip Hershkovitz, Scott M. Lindbergh, Barbara E. Brown and Alfredo Langguth; original number PH 10246. Etymology. The white triangular rostral patch of this species captures the attention. The white or colorless rostral field is a terminus of the pheomelanic pathway which begins with P. Hershkovitz 236 e(991—S'pI)6'ST L(6'b—b'"b)L'p L8r—E WI r L(y"9—8"S5)0'9 L(9El—8"7D)€'El AS 6T— 18067 E(Sh—OP)EP L(61—91)81 ¿Ly? — 17) ZZ L(06— £8)98 L(TIT—E6)£01 S9UOISIW ‘y'Iz 1eJiseq orf PI “0'pI 8 p ‘sr 99 ‘OL Or “8'p SER cv sr LS '9'S PAE Ar A IA | SS Sia WG 13S 1°87 ‘0°62 81 “81 gET ‘877 €8 ‘S8 LOI “LOT 6P1PSID "Old ‘7 epSIQ OVond :E “OABIA op SO ‘7 “seunqoL (OZE :6961) S9UIO] PUB PIOSSBIA WOT yy “(PEL :£961) BIOSSEJA wol} ‘euNUodIY ‘SIUOISIW py ‘(PEI :£961) BIOSSEJA VOI eunussıy “SOUOISIA "000.7 ASMZIA 2dAJ0]0Yy JO st JUAWIINSPIVL I511J :INOY3NOIYI J9PIO AUIES UL SJUIVISINSEIV [ENPIAIPU] [fey poqqod "Y (Oyump1IoW “1 ‘opuny ovloqry “py ‘OLIOW OOWIIg :ojneg ORS ‘(uomestunumwo) [euosiad) SISAJA AyIYg wo “Ez ‘Teseq ale (ZOE :9681) Seuoy 4q sdAJO[OY jo sjusuamsesul YSU] [eruezo [eurBro "(uolesrunumwos ¡euosiad) sIoAW did voy o16'£l ‘EET US ‘8'P ae Ne vy ‘sv ST ‘ST LV SS = al OE ‘9°71 SERBS ay epee =e GSE = WES 81 “91 v7 “ST 68 “L8 LOT “TIT ¿Seungqo MPO INOYIIM "(80€ :9681) sewoy] WoIJ sdAJO[OY JO SJUIUISINSEIVI [eUIZIIO “¡ng Op apuein ory ‘erenbey ‘A[UO sodAjofoy jo sinyns [e}Id1990-OJoIIed19jUI 0} speseu jo dı] II(6'61—8'91)8"L1 ‘8°91 — *9°97 ZI(8'b—O'P) PP ‘O'P IIW8-T'DS’L “EL AAA zIH6'1—S'D9'1 ‘OT TH6 v—I vr" p ‘6 r Z1(9'9—0'9)h'9 ‘S’9 ZH6"TI—TODS"TL 811 THE EL—VZDLTZIL “671 Z1(9'9—6'S)7'9 ‘19 OLUTEI-FTDE TI ‘LI 119 L7-S’SDIIT 79T II(S'0€ —8°87)h' 67 767 TULT—OT)L EZ “EZ ZH6T—91)S"L1 ‘LI EIIT—FOST “97 OUZIT—L6JEO1 “,S6 ZI(60I—€6)EOI “SOT OLI ‘— LLI| 9(8'91—8°SD)P'9I ‘9°91 =e aes ELIS 10°SZ 7 = 2 67-9 PLP 6'p OL = Gl S(0'L—7'9)9'9 “0'L Sp ‘Ly ‘SY Wr p—T Ey Ep OG = OP 90'7—S DLT ‘SI A Se v—L Ir Tp EIN II 9(1°9—6'S)0'9 “0'9 TU LIT] HIT T'OD6'0L ‘8°01 — ar SAO IIS al 009 ‘— “E9 v(Z:9—8"S)0'9 “0'9 SE FO! TEL vzı ‘p21 TST — SST] EG HT—S' PUL HT “L'v7 0087 ‘— ‘6 LT | = H(B'LZT—S OTE LZ ‘8°LZ SE ‘OE ‘TE 81 “91 “91 SoG SG 665.266 871 “SIT II y13u9] “*S[QIPUeIn ¡100-JUI-OSEN y1pim “unnsoy JWSISEIA MOI IEJOW red INewoskz y13u9] ‘oyefeq U9WEIOF DAISIDUT y13u9] [eseN osesurelg YIPIM Te71q.10.1330] yIpesiıq INewosAz y13u9] [esegojApuoy y33ua] [[NAS IS9I8910) ya Mm Y(LI—SDIT “SI eq 9(97—PZ)SZ “97 1003 pulH E(S8—Z8)€8 ‘PL y1Bua] [TEL 9(O1I—86)EOI “OTT Apoq pue pe>H (08bb6 WA) ySNUIDILOS (000.7 dSNZW) g511JUDA1UB1 (910zE NW) SUIISA{NAISIAB (LIOTE NW) snspu1g]d 8°91°6'68 WA ¿18U119Y1 6 8 Cai Gay SA Vay Vo) [= _ ‘adAJO[OY 94} JO SI UOXB] YOR JO JUSWIOINSeOU ISI ‘SNIUOSJa]JDdaInNIg JO susumads pooja pue so1oads 34] JO SIUIVIIINSBINAN :6 9[ GUL Sigmodontine rodents from southeastern Brazil 237, Fig. 24: Mustached Brucie, Akodon mystax (top), and White-nosed Brucie, Brucepattersonius albinasus (bottom). 238 P. Hershkovitz reddish (e. g., Wiedomys pyrrhorhinus), bleaches to yellowish in others (e. g., Akodon xanthorhinus), and finally white. The rostral patch may also be blackish (eumelanic) as in Akodon mystax described above, and the species of Oxymycterus discussed elsewhere (Hershkovitz, 1994). Type locality. Pico da Bandeira, Parque Nacional de Caparaó, Minas Gerais, Brazil, elevation, 2700 meters. Distribution. Known only from the type locality in the Parque Nacional de Caparaó. B. albinasus and B. griserufescens occur together on the Pico da Bandeira at 2700 meters. Diagnosis. Smallest species of the genus; tail longest, about 114 %, relative to head and body length; general coloration brownish, rostral patch, upper surface of hands, and caudal pencil whitish. Comparisons. Distinguished from all other species by smaller size, tail longer than head and body combined, and white triangular rostral patch. Description. Size, smallest of genus; pelage long, fluffy, hairs of middorsum 10—12 mm long; dorsum from head to rump dark brown, sides similar, well defined underparts grayish with thin ochraceous wash, the slaty hair bases showing through; inner surface of limbs like ventrum; hairs of triangular rostral patch, upper surface of hands (fig. 25) and tail pencil whitish, feet grayish, the pale brownish skin showing through; digital and facial vibrissae whitish, the rostral vibrissae laid back reaching ear bases; tail extremely long, about 114 percent of head and body length, brown above, ventral hairs whitish about 5—7 scales long, the scales not hidden; volar pads six; foreclaws thin, weak, length about 2.5 mm, hind claws sturdy, about 3.5 mm. Cranial. Skull smooth, without crests or ridges; rostrum slender, elongate, length about 40 % percent of skull length; nasal tips rounded, produced slightly less than 1.0 mm beyond incisors, without expansion or formation of trumpet; zygomatic arches weak, hardly expanded beyond sides of braincase; zygomatic plate narrow, sloping back from base, hardly visible viewed from above; interorbital region comparatively wide; braincase subglobular; inter- parietal about 1.0 x 4.7 mm; incisive foramina long, narrow, terminating slightly posteriad to procingulum of m'; palate produced slightly behind posterior plane of m?; greatest width of mesopterygoid fossa 1.7 mm; sphenopalatine vacuities absent; damaged and detached postero- ventral portions of braincase including sphenoidal, occipital, mastoidal, and petrous portions too fragmented for accurate description. Mandible slender, smooth, without defined incisor root capsule. Dental. Molar rows very nearly parallel-sided discounting greater width of first over third molars, opposing cusps oblique. M!: Anteromedian fold present; anteromedian style absent; anterolabial conule larger than anterolingual; minute protostyle (7) present; minute paralophule (m) fused with 7; absent elements compared with Oxymycterus nasutus, include enteroloph (u), enterostyle (x), plesio- style (f), anteroloph (Ah), mesoloph (1), and mesostyle (o); short paraflexus (3) and metaflexus (5) isolated on occlusal surface; median fossette (a’) present; posterior fossette (b’) indicated by a dot. M?: Like M! except procingulum reduced, accessories absent; paraflexus (3) and proto- flexus (8) present; median fossette (a’) coalesced with paraflexus (3); mesoloph (n) absent; mesoflexus (4) absent or combined with metaflexus (5); posterior fossette (b’) coalesced with metaflexus (5); presence of posteroloph (1) indicated by posteroflexus (6). MP: Worn; subtriangular in outline, less than half size M?; paracone (1), protocone (v), hypocone (y) raised, metacone (q) indicated; paraflexus (3) isolated; entoflexus (9) well defined. m,: Anteromedian stylid (a) absent, preflexid (anteromedian fold) (1) absent not to be confused with crenulation of anterior enamel margin; supraflexid (7) present; labiolophulid (d) fused with protoconulid (7); protoflexid (8) poorly defined; included present are hypo- flexid (10), paralophulid (J), ectostylid (rn), metaconid (1), entoconid (y), posterolophid (7), Sigmodontine rodents from southeastern Brazil 239 Fig. 25: White-nosed Brucie, Brucepattersonius albinasus (holotype MN 32017); skull and molars; GSL, 27.8 mm; molars, 4.4 mm; Pico da Bandeira, Parque Nacional de Caparaó, Minas Gerais. i} IM Fig. 26: Left hand of White-nosed Brucie, Brucepattersonius albinasus (holotype MN 32017 9); A, dorsal aspect; B, ventral aspect. 240 P. Hershkovitz mesolophid fused with metalophulid and or metaconid; coronal surface with only short isola- ted mesoflexid (4) and posteroflexid (6). m,: Like m, but with reduced procingulid, distoflexid (17) less worn, better defined. m,: Narrower, shorter than m,; main cusps well defined; mesoflexus (4) isolated, nearly as long as that of m, but size may be increased by fusion with adjacent elements. Measurements. Table 9. Specimens examined: Total 1. Minas Gerais (Parque Nacional de Caparaó, Pico da Bandeira, 2700 m, MN 1 [holotype]). Brucepattersonius iheringi Thomas — Ihering’s Brucie Hesperomys nasutus, Hensel (not Waterhouse), 1873: 43, figs 19a, b, 29a, b (molars) — BRAZIL: Rio Grande do Sul. Hlesperomys]. nasutus, Leche (not Waterhouse), 1886: 700, figs 29—30 (molars) — BRAZIL: Rio Grande do Sul (Taquara do Mundo Novo); cranial and dental characters. Hesperomys (Oxymycterus) nasutus (not Waterhouse), Ihering, 1892, Anuario do Estado do Rio Grande do Sul, para 1893. 9: 109 — BRAZIL: Rio Grande do Sul (Taquara do Mundo Novo). Oxymycterus iheringi Thomas, 1896: 308; description. Thomas 1902: 62 — BRAZIL: Paraná (Roca Nova, Serra do Mar, 1000 m); coloration. Oxymycterus iheringi, Massoia 1963a: 129 (part) — comparisons; taxonomic history of holotype. Massoia & Fornes 1969: 315 (part) — taxonomic history of holotype. Musser & Carleton 1993: 727 — taxonomic comments. Mlicroxus]. iheringi, Thomas 1909: 237 — reclassification. Microxus (?) iheringi, Gyldenstolpe 1932: 134 — characters; taxonomy. [?] Microxus theringii [sic] Vieira 1953: 145 — BRAZIL: Rio Grande do Sul (Sao Lourenco); Sao Paulo (Campos do Jordáo); characters [measurements of male not of O. iheringi — PH]. Massola 1963: 135 — Vieira identifications questioned. Akodon (Microxus) iheringi, Cabrera 1961: 458 — classification; distribution. Holotype. Female, skin and skull, British Museum (Natural History) no. 86.9.16.8; collected by Hermann von Ihering. Type locality (fig. 1). Rio dos Linos (sic = Sinos), Taquara do Mundo Novo, Rio Grande do Sul, Brazil, 29°39’S, 50°47’W; 29 m. Distribution. Forested parts of southeastern Brazil in the states of Rio Grande do Sul, Paraná and perhaps Sáo Paulo. Diagnosis (from literature). Size small, upper parts and sides grayish, underparts undefined from sides; tail nearly as long as head and body combined. Characters. Oxymycterus iheringi Thomas is virtually unknown apart from the original description and some bibliographic references. The specimens from Misiones, Argentina, described by Massoia (1963), and by Massoia and Fornes (1969), as surrogates for true iheringi, do not agree with that species in coloration, size and perhaps other characters. Measurements are reproduced here (table 9), details of color are quoted, and mentioned in the discussion of the Misiones material. The original Thomas description of iheringi and his other contributions are reproduced as follows. “Oxymycterus Iheringi, sp. n” [Thomas, 1896: 308] “Much smaller, more slenderly built, and less Oxymycterine than O. nasutus, rufus, and the other more typical species. Fur soft and thick. General colour uniform grizzled brown, scarcely paler below. Eyes not unusually small. Ears fairly large, thinly haired, brown. Claws much less lengthened than in O. nasutus, but still with the essential fossorial structure characteristic of the group; pollical claw short. Fifth hind toe decidedly longer than the hallux, reaching to the level of the base of the fourth toe. Tail almost as long as the head and body, slender, thinly haired, brown above, rather paler below. Mammae 1-2 = 6. Sigmodontine rodents from southeastern Brazil 241 “Skull not specially elongated anteriorly, although the muzzle shows something of the characteristic Oxymycterus structure. Supraorbital edges smoothly rounded. Interparietal and anterior zygoma-root and other details very much as in O. nasutus, in spite of the great difference between the two in the general proportions of the skull. “Dimensions of the type (an adult female in spirit)” [are reproduced in table 9]. “The two specimens of this species in the Museum are part of the large collection of Taquara rodents worked out by Dr. Leche [footnote, ‘Zool. Jahrb. i. p. 700 (1886).], by whom the present animals were called Oxymycterus nasutus, under which name they have remained in the Museum collection until now. Among other rodents collected by Dr. von Ihering at San Lorenzo, in the same province, there are specimens undoubtedly referable to the true O. nasutus; but these two from Taquara, and no doubt the others seen by Dr. Leche, are so different that there can be no question as to their specific distinction [footnote, “Dr. Leche says that of nineteen skulls examined by him the largest had a basilar length of 22 millim. The basilar length of the true O. nasutus is from 27 to 29 millim?]. Dr. Leche was no doubt led astray by Hensel, in whose classical paper on the mammals of Rio Grande do Sul [footnote, ‘Abh. Ak. Berl. 1872: 43”] the species now described is also referred to O. nasutus. Hensel’s account must therefore in future be assigned to O. Iheringi, to our knowledge of whose structure and habits he makes some valuable contributions. The difference between the two forms is so great that it is difficult at first sight to realize that O. /heringi is an Oxymycterus at all, as it is quite without the extraordinary trumpet-shaped muzzle possessed by O. nasutus and its allies. as yet only known fossil from Lagoa Santa, but has rather a shorter head and shorter palatine foramina; so that I have not been able to assign it to the fossil form, as in the case of the animal next to be described” In a report of a male from Serra do Mar, Paraná, Thomas (1902: 62) noted that “this is the first skin .. . of O. Iheringi which I have seen, the original series all having been in spirit. The _ general colour should rather have been described as grey than brown.” In 1909 (p. 237) Thomas referred iheringi to his newly erected genus Microxus. The specimens of O. iheringi collected by Hensel (1872: 43) in Rio Grande do Sul and recorded as Hesperomys nasutus include 2 skeletons, 2 skulls and 1 or 2 entire in spirits. His measurements of basilar length of the smaller of the two skulls was 20.0 mm, upper molar row 4.1 (M' = 2.0, M? = 1.30, M? = .90). Measurements of the larger skull were basilar length 22.2; nasals 12.5; incisive foramen 5.3; interorbital width, 6.1; interparietal, 1.4 x 7.7; mandibular depth below m, 2.7, below m, 2.3. Vertebral count of the skeleton was 12 thoracic, 7 lumbar, 20 caudal. Measurements of a male identified by Vieira (1953: 145) as “Microxus theringi,“ provenance unspecified, appear to be those of Oxymycterus nasutus. In all likelihood other specimens Vieira recorded at the same time under the same name, one from Säo Lourenco (no. 572), and two females from Campos do Jordáo (nos. 2073-74), are likewise referrable to O. nasutus. Notwithstanding, they are listed in the above synonymy of O. i. iheringi as possible sympa- triots in hypothetical northern and southern extensions of its geographic range. Specimens examined: None. Brucepattersonius sp. (Misiones, Argentina) Oxymycterus iheringi, Massoia (not Thomas) 1963: 129, figs 1—4 (skull) — ARGENTINA: Misiones (Tobunas, Ruta 14, kn*352; Puerto Gisella, Río Paraná); characters; comparisons; taxonomy; habitat. Massoia & Fornes 1969: 315, fig. 1 (animal), fig. 2 (palate), fig. 3 (molars); ARGENTINA: Misiones (Tobunas; Dos de Mayo; Puerto Gisella); taxonomic history; characters; comparisons. The Misiones, Argentina, sigmodontines identified with typical Rio Grande do Sul Oxymycte- rus iheringi Thomas, are apparently congeneric judged by the descriptions and illustrations published by Massoia (1963), and Massoia & Fornes (1969), but not likely conspecific. Massoia's (1963: 133) description of the four Misiones specimens he recorded follows, freely translated from Spanish. “This is a small species of Oxymycterus with comparatively short claws, the length in a straight line between 2.3 —2.6 mm; pelage soft, general coloration of dorsal surface approxi- "MEJO INOYJIM pue MEJO YIM “UNS AIp UT y ‘snpidsiy O Ajaaryequay :(NI) TIzeig “o¡neg oes , (WA) PUNUIBIY “SOUOISIJA y ‘(NA) Izeig ‘opneg oes ; ‚snfnı © Ajaatyequsy “(AA) Tzeig ‘STEIN SRUIA y "(1961) eIOsse WOLF SJUILIIIMSEIA “euNuasıy ‘soIry soUuoNg ; (MI) € ‘eIeT eyung MA) I tiolog “eunus3ry “say souong ; TSE =..EI MEJO IS93UO] JO diy 0) 1007 pury “usar3 JOU y13u9] [re] (ww (ST =.2/16 YISU] [eI0) “uswrmads puosas “uuu ZEl =¿0/1€ Ile] Sa] “2/18 YIU] [BIOL :UI9ISÁS SOW 0) P9J19AUOI SJUDUIIINSEIW YIUIIH PIO ‘TO8I PIBZ Y LO ejep ‘snf{ns sNAaJIAWAXOE jo Ayieso] adá] “eunus3ly ‘sory 9amug | P. Hershkovitz 242 L(9—L"9)6'S (9 S—T ISS 8(6 s—s"s)9"S L(8"S—S"S)L'S AKIS—T'SIE'S ble ST DTS MOY Je[oW L(0’8E— 8 VE)I’IE €(7 8E—6 9E)P'LE HZ SE) — L(S"9€—9"TE)V VE e(9'vE—9"0€)€ TE [eseqo¡Apuoo) LIBE—TEIIE S(8€ —PE)LE 8(9€ —SE)SE LILE—ZE)SE 9(LZ—€7)8°S7Z gh(L7—7Z)S7 gh(TE —6Z)0E GSE }OOJ PUTH E(8p1—TID9EI S(SETI—ETDIEI 8(671 —LIDZTI E(IST—PID6ZI 9(L6—98)€6 Y(96 —V6)56 [s6l SET S(98T—0SDPLI S(6LT—v91)L91 S(ELI—PSDIII LISIT—6EL)ZSI 9(081 —6E TEST r(68I—PvEDESI ¡Sel Apog 27 pe>H /esuelody] gAerenge1e) ¿OLIOJA OAMUNIA „oeıede) ‚eueieg ena ‚eueieg ena ¡4mos „Te ‘pojou se ¿d90xa JOyyne 394) Aq sJuswaınsesur [TY “SOMIBIO[| P3DIYSS WOAIF SOPNIDOY YM paredulo) snfnı snsaJOAWAXE JO SJUSWIIINSBIW [CUISLIO :OT 91ABL Sigmodontine rodents from southeastern Brazil 243 mately dark ochraceous “pardusca ochracea” “(Y YO-6-3*)” gradually passing into a grayish brown (“amarillo pardusco pálido”) “(YYO-17-5°)” on belly, all hair bases being plumbeus gray “(C-6-1°)? Massoia & Fornes (1969: 319) described the dorsal surface of 7 Misiones specimens (including the foregoing) as a mixture of grayish chestnuts and yellows (“castafios, grisaceos y amarillos”) “(0-5-7).“ The color terms used are from the “Atlas de Villolobos Domingues y Villalobos.” The “Atlas” is not available but the color terms are standard. Cranial characters were said by Massoia (1963: 173) to be similar to those of adult Oxymyc- terus nasutus except smaller. The nasals of four specimens (Pto. Gisela, 2; Tobunas, 2) were described as completely fused except for about 1/3 their length, a character not seen before in sigmodontines. The crown surfaces of upper and lower molars of the Tobunas mice were figured by Massoia & Fornes (1969: 318) in comparison with those of Oxymycterus nasutus. The peculiar nasal character was not mentioned. The two Tobunas mice recorded by Massoia were trapped at night in wooded areas, one on the border of a rocky stream, the other in a low brushy area on the edge of a pathway. Measurements. Table 9. Specimens examined: None. Oxymycterus Waterhouse Type species Mus nasutus Waterhouse, 1837: 16, by original designation. Oxymycterus rufus Fischer (fig. 27) Mus rufus Fischer, 1814: 71. Holotype. Not known to exist; name based solely on the description of the rat cinquieme ou rat roux of Azara (1801(2): 94). Type locality. 32°30’S, Rio Paraná = Entre Rios, Argentina (cf. Hershkovitz 1994: 35). The large reddish hocicudo (“long nosed” in Spanish) of Paraguay, Uruguay, the Argentine provinces of Entre Rios and Buenos Aires, and the Brazilian states of Rio Grande do Sul and Santa Catarina, have generally been identified as Oxymycterus rufus Fisher. Other named large hocicudos of the same regions are O. misionalis Sanborn (1931) from Misiones, Argen- tina, O. judex Thomas (1909) from Santa Catarina, and O. quaestor Thomas (1903) from Paraná. All three have since been treated as subspecies of the Bahian O. hispidus (fig. 28) by Cabrera (1961: 467), and as outright synonyms of O. hispidus by Musser € Carleton (1993: 727). Oxymycterus hispidus from Bahia, described by Pictet in 1843, is known from the origi- nal description only. The holotype, if extant, awaits comparison with any of its referred conspecifics, or with a topotype or near topotype of O. rufus, which lacks a type specimen. Coloration of the hocicudo as described by Azara in the Spanish edition (1802: 80) as canela or cinnamon, a hue which could apply to any reddish hocicudo and to most specimens of the localities listed in table 9. Measurements given by Azara in the original description of O. rufus and reproduced in table 9, agree best with those of the mouse from the Delta Paraná, Argentina (table 10) except that hind foot length is much too large. However, total length of the larger of the two hocicu- dos measured by Azara brings other extrapolated dimensions into line with those of the larger individuals listed in the same table. The name Oxymycterus rufus Fisher, therefore, is the earliest available for the species. The Delta Paraná hocicudos with short hind feet may not be Oxymycterus rufus. The large hocicudos from Caparaó, Primeiro Moro, Iporanga, and Caraguatay may not be conspecific but cannot be shown to be either O. rufus (fig. 27) or O. hispidus (fig. 28). They are much larger than the next described hocicudo. Measurements. Table 10. Specimens examined. Total 17. Sáo Paulo (Petar, Iporanga, 8 MZUSP); Espírito Santo (Parque Nacional de Caparaó, Pedra Roxa, 6 MN); Minas Gerais (Parque Nacional de Caparaó, Cachoeira Bonita, 1 MN; Vale Verde, 1 MN); locality unrecorded, 1 MN. 244 P. Hershkovitz Fig. 27: Azara's Rufus Hocicudo Oxymycterus rufus (MN 32002 ©); skull and molars; GSL, 35.8 mm; molars, 5.5 mm; Pedra Roxa, Pico da Bandeira, Parque Nacional de Caparaó, Espírito Santo. Oxymycterus caparaoe, new species (fig. 29, 30) Holotype. Adult female, skin and skull, MN no. 31997, collected 25 October, 1992, by Philip Hershkovitz, Scott M. Lindbergh, Alfredo Langguth and Barbara E. Brown, original no. 10426. Type locality. Arrozal, Parque Nacional de Caparaó, Minas Gerais, Brazil, elevation 2400 m. Distribution. Known only from the western slope of the Pico da Bandeira, Parque Nacio- nal de Caparaó, Minas Gerais. Diagnosis. A small, dark brown hocicudo with orange underparts, and long, slender, trumpet-shaped nasals. Characters. External. Upper surface of body dominantly dark brown modified agouti, pelage thick, long, lax, about 1 cm long on dorsum; individual hairs with tip blackish, the single orange subterminal band minute; basal portion of hairs slate color; sides of body more orange, the subterminal pheomelanin band wider; underparts orange but with slaty basal portion of hairs showing through, the broad ventral midline stripe from throat to anus nearly entirely orange; ears moderately large, pinna dark brown; cheiridia long, narrow blackish above and below each with 6 plantar pads; manual claws longer than pedal claws; manual digit III with claw 7.9 mm, digit IV with claw 5.8 mm, adjacent digits shorter, digit I vestigial, claw, 2.5; digit V, not reaching base of IV, with claw 2.3, reaching phalanx 1; pedal digit I extending to base of II, with claw to base of phalanx 2 of II, digit V about same; claw I, 3.2 mm; Sigmodontine rodents from southeastern Brazil 245 Fig. 28: Hispid Hocicudo, Oxymycterus hispidus (MN 32003). II, 3.8; III, 4.0; IV, 3.8; V, 3.6; interdigital webbing present between second phalanges of digits II, III, IV; tail uniformly dark brown, scutular hairs short, those of underside longer but not concealing scales. Mammae 1-2 = 6. Cranial. Dorsal contour of skull gently sloping; rostrum elongate, nasals with premaxillary bones parallel-sided, their combined tips slightly trumpet-shaped; zygomatic arches slender, zygomatic plate visible when viewed from above; frontal sinuses little inflated; temporal ridges weak; well defined interparietal bone small (3x9.2 mm); incisive foramina (1.7 x 6.9) extending slightly behind level of metacone of M'; posterior palatal border level with posterior border of last molars; mesopterygoid fossa wide (1.8 mm). Dental. Incisors orthodont, combined cutting edges plane; molars tetralophodont, hypsodont, the cusps opposed, worn crowns deeply dished, 8-shaped; M' with anterior median flexus, M? small, the attenuated metacone and hypocone barely distinguishable from each other; lower molar crowns as worn as uppers, opposing cusps oblique; anterior median flexus present; mesolophid, ectolophid and protocunulid defined in m,,; m, nearly twice as large as MP. Measurements. Table 11. | Comparisons. Oxymycterus caparaoe can be distinguished by smaller size alone from larger sympatric Oxymycterus rufus or O. hispidus. It is separated from the similar sized southeastern Brazilian O. nasutus Waterhouse (1837) (fig. 30) by paler coloration (table 12) and more projecting rostrum. Comparative morphometrics are in table 11. 246 Table 11: Oxymycterus caparaoe and Oxymycterus nasutus compared: measurements are P. Hershkovitz means, extremes, sample number. Head and body Tail Hind foot Ear Gr. skull length Condylobasal length Zygomatic breadth Interorbital width Braincase width Nasal length Holotype caparaoe! 126(110—139)28 93(80— 104)22 27(24—29)27 102107 34.4(32.9—35.9)15 SOS(ZSO ) o 13.9(13.6—14.4)9 5.9(5.5—6.4)22 13.8(12.9—14.5)16 12.9(11.4—14.5)16 nasutus2 125(110—137)5 86(79—90)5 27.6(27—28)5 32.2(30.3—33.4)3 29.9(28.8—31.1)5 14.1, 14.2 5.5(5.3—5.7)5 13.0, 13.6 13.0(13.3—14.2)3 nasutus? 129(114—141)9 85(72—91)9 234(22—24)9 17(16—19)9 32.8(31.5—34.5)3 14.0(12.7—15.0)3 5.8(5.5—6.1)3 12.411415 128 Incisive foramina Rostrum width Zygomatic plate Diastema Molar row 6.8(6.5—7.2)19 TACO TESIS = 4.5(4.0—5.0)21 — = 2.1(1.6—2.4)18 ZMZO 25108 UXT) 1s) TS CS) > = 4.7(4.5—5.5)19 4.8(4.7—4.8)3 4.7(4.6—4.9)3 1 HB = Head and body; ? with claw; 3 from notch; *CL = Condylobasal length; $ ZB = Zygomatic breadth; * IB = Interorbital breadth; 7 BW = Braincase width; $ NL = Nasal length; ? IF = Incisive foramen; !? PL = Palatal length; II UM = Upper molar row; 1? D = Diastema; 1% R = Rostral breadth; '* W = Weight, grms; 15 ZP = Zygomatic plate. Table 12: Phenotypes compared of Oxymycterus caparaoe (type series) and O. nasutus topotypes. Character caparaoe nasutus 1. Dorsum Dominantly ochraceous (pheomelanic) modified agouti Dominantly dark brown (eumelanic) modified agouti Dilute orange (pheomelanic) with basal gray showing through . Underparts Saturate orange (pheomelanic) with basal gray showing through Nearly or entirely uniformly orange . Sides of trunk and Dark brown like back or head slightly paler Darker than back . Tail Uniformly blackish . Crown Orange like back Bicolor; pale brown above; distinctly paler beneath Hind, feet Pale above, nearly colorless beneath Blackish above and below . Ears Dark brown Ochraceous Dimensions of Andean Division size groups of hocicudos are roughly comparable to those of the Atlantic Division (cf. Hershkovitz 1994). The small Andean O. hiska and O. hucucha are smaller than any Atlantic Division hocicudo, the large O. inca parallels members of the Sigmodontine rodents from southeastern Brazil 247 Fig. 29: Mt. Caparaó Hocicudo Oxymycterus caparaoe (holotype MN 31997 ©); skull and molars; GSL, 34.3 mm; molars, 4.5 mm; Arrozal, Pico da Bandeira, Parque Nacional de Caparaó, Minas Gerais. O. rufus-hispidus group, and those of the medium-size O. paramensis group equate with the medium and intermediate-size O. nasutus-O. caparaoe group. Morphometrics of O. paramen- sis, O. nasutus, and O. caparaoe are virtually the same but O. caparaoe is intermediate in coloration, O. paramensis being the paler. Most cranial differences between O. caparaoe and O. paramensis vary randomly from population to population. The consistently different traits of O. paramensis include shorter, less protrusive rostrum, the condition reflected in shorter cranial length and incisive foramina. Other distinctions include narrower mesopterygoid fossa, and more inflated frontal sinuses (fig. 30). Remarks. The obvious difference between O. caparaoe and O. nasutus is the dark brown . or dominantly eumelanin outer parts of the first and the dominantly pale reddish or orange pheomelanin outer parts of the second. Expansion of the pheomelanic band of the agouti hairs in O. nasutus with corresponding decrease in width of the eumelanic band are derived conditions. Descriptions of O. nasutus have been provided by Vieira (1953), and Gyldenstolpe (1932). Range of chromatic variation in O. caparaoe is extremely narrow and the probability that the species may be dichromatic or that it intergrades with the paler O. nasutus seems unlikely. P. Hershkovitz Fig. 30: Mt. Caparaó Hocicudo Oxymycterus caparaoe (MN 31997 9); A, skull; B, molars; GSL 34.3 mm; molars, 4.5 mm; Arrozal, Caparaó, Minas Gerais; compared with Oxymycterus nasutus (FM 27652 ©); C, skull; D, molars; GSL, 33.3 mm; molars, 4.7 mm; Uruguay: San José, S. Ecilda. The originally forested habitat of O. caparaoe, is now second growth and scrub. That of O. nasutus is scrub and/or savanna. 249 Sigmodontine rodents from southeastern Brazil + PIQIL| + ua y13u97 IS9JB9J0) SIAJOd snraunH turn eun snoung SLIZE NN D11131U SAwuoldDyL 67001 Hd S1JPDS1O0PP SÁWO]IT 679IvI WA 8c9IvT WA snipouiqns SÁWoa(T gETZE NIN SUBISA{NAIS1AB SNIUOSJA] J DAIINAT 9861£ NIN $861€ NIN v361€ NW 90DIVDADI Y 6956 Hd 114129041 O SO0OTE NIN snfnı snIOJIÁMÁAXO "DOT x SONI pue sJuswaIınseaw [e1919NS :£] OIqQEeL 250 P. Hershkovitz Specimens examined: Total 30. Minas Gerais (Parque Nacional de Caparaó, Segredo 2100 m, MN 1; Vale Encantado, 1980 m, MN 17; Arrozal, 2300—2400 m, MN 7; Terreiráo, 2400 m, MN 9; station unrecorded, MN 5). —__————oo — Ro ———Y Fig. 31: Diagrams of right and left upper molar crown patterns. Explanation for symbols. Note: a—i inclusive = procingulum or loph I; s, t, z = postcingulum or loph V. I—V. lophs of pentalophodont molars anteromedian style (may be fused with b, c, or both) anterolabial lophule anterolingual lophule anterolophule (may be fused with f) anterior fossette plesiostyle (may be fused with d, h, or j) protostyle (may be fused with 1) anteroloph (may be fused with f, j, or both) protoloph (may be fused with g, u, or both) parastyle (may be fused with f, h, k, or a combination) mesolophule (may be fused with h, j, or both) paracone paralophule (may be fused with n, o, or both); element may be multiplied mesoloph (when fused with o = mesolophostyle) mesostyle (may be fused with m, p, or both; when fused with n = mesolophostyle metalophule (may be fused with o, n, or both) metacone posterolophule (may be fused with s) posterostyle (may be fused with r, t, or both) posteroloph (may be fused with s) posteroconule (may be fused with z, usually not differentiated from posteroloph, t) protolophostyle (may be fused with i) protocone enteroloph (may be fused with x) enterostyle (may be fused with w) hypocone distostyle median fossette (may be coalesced with 3, or united with 4) posterior fossette (may be coalesced with 5, or united with 6) protolophule hypolophule mure (border between lingual and labial cusps and lophs) eBPoPs5Truerpemrepanc»e me ande ln) ws SS » Ns SS w O preflexus (anterior median fold) anteroflexus (anterior secondary fold) paraflexus (first primary fold) mesoflexus (first secondary fold) metaflexus (in absence of mesoloph [n] coalesced with first secondary fold [4]) posteroflexus (second secondary fold) supraflexus (anterior lingual fold; in absence of protoloph coalesced with first minor fold [8]) protoflexus (first minor fold) entoflexus (major fold) ; hypoflexus (in absence of enteroloph coalesced with major fold [9]) distoflexus (second minor fold) Foe OO SON UT ale FS) Sigmodontine rodents from southeastern Brazil 2D! Skeleton Bone measurements and ratios of the species of this report were intended to reflect locomotor and functional adaptations. It was found, however, that in most of the few available skeletons, distal ends of the long limb bones had been truncated and left with the foot bones in the prepared study skins (table 13). Convincing inter- pretations of form and function could not be derived from the partial data but the four complete proximal limb bones are instructive. Thaptomys with its extremely long humerus (humerus/femur) is fossorial. In contrast, the comparative length of the Oxymycterus humerus to femur does not indicate adaptation for digging. The vertebral count of the genera examined is the expected 7 cervical, 13 thoracic (rarely 12 or 14), 5 or 6 lumbar, sacrals 2 (Thaptomys), 3—6 (Oxymycterus), 2— 3 (Delomys). The entepicondylar foramen is absent in all sigmodontine species examined. Articulation of the first rib may be with the first thoracic vertebra, the 7th cervical vertebra, or with both. Sympatry All individuals of the same genus living on the western or Minas Gerais slope of Mt. Caparaó from 1100 m upwards to the peak at approximately 2700 m, are regarded as sympatric. The congenerics actually taken in the same trapping stations are Akodon mystax and A. serrensis from Terreiräo and Pico da Bandeira, and Oxymycte- rus rufus and O. caparaoe from the same two stations; Brucepattersonius griserufes- LINGUAL 252 P. Hershkovitz cens and B. albinasus were trapped at Pico da Bandeira. Each of the remaining species of this report may not have a congeneric in the Parque de Caparaó or it remains to be discovered. Tr se Fig. 32: Diagrams of right and left lower molar crown patterns. Explanation for symbols. Note: a—h inclusive = procingulid or lophid I; a’, r, r’ = postcingulid or lophid V. I—V. lophids of pentalophodont molars anteromedian stylid (may be fused with b, c, or both) anterolabial conulid (may be fused with c) anterolingual conulid (may be fused with b) labiolophulid (may be fused with g) anterior fossette anterolophulid (may be fused with h, i, or both) prostylid (may be fused with d) anterolophid (may be fused with f, i, s, or combination) anterostylid (may be fused with f, h, s, or combination) protoconulid (may be fused with g) protoconid paralophulid (may be fused with m, n, or both) . ectostylid (may be fused with 1, n, o, or combination) ectolophid (may be fused with Il, m, o, or combination) hypoconulid (may be fused with m or n) hypoconid posterolophulid (may be fused with r”) posterolophid (may be fused with q) posteroconulid (usually not differentiated from posterolophid, r’) mesolophulid (may be fused with h) metaconid metalophulid (may be fused with v, w, or both) mesolophid (when fused with w = mesolophostylid) mesostylid (when fused with v = mesolophostylid) entolophulid (may be fused with v, w, or both) entoconid distolophulid (may be fused with a’) NM ek EOP HRA OB Bee >. BP op a. posterostylid b. median fossetid c. posterior fossetid d’. protolophulid e. hypolophulid f. murid (zone between lingual and labial lophids and cuspids) preflexid (anterior median fold) anteroflexid (anterior lingual fold; in absence of anterolophid [h] coalesced with 3) metaflexid (first secondary fold) mesoflexid (first primary fold) entoflexid (second secondary fold; in absence of mesolophid [v], coalesced with 4) posteroflexid (second primary fold) supraflexid (anterior labial fold; in absence of labiolophulid [d] coalesced with 8) protoflexid (first minor fold) ectoflexid (major fold) . hypoflexid (in absence of ectolophid [n] coalesced with 9) distoflexid (second minor fold). Bseonsuprunn AS Sigmodontine rodents from southeastern Brazil 253 Dental patterns Patterns of the sigmodontine upper and lower first molar crowns are shown diagram- matically in figures 31 and 32. The terminologies for the dental elements follow. The same diagrams serve for the identification of the enamel elements and folds or flexi (ids) of all muroid molars. Lettering and numbering of dental elements for right and left upper and lower molars are the same but reversed. Because of differences between stationary upper and moving lower molars the symbols are not the same for most apparently homo- logous upper and lower elements. Numbers for the enamel folds or flexi (ids) remain the same for upper and lower molars but reversed for right and left. Acknowledgements The generous help received from many Brazilians during the course of my 5 field trips in southeastern Brazil between 1986—1992 has been acknowledged earlier. Without that assistance this research could not have been possible. My heartfelt thanks to all, especially Professors Ulisses Caramaschi, Paulo E. Vanzolini, and Alfredo Langguth for sponsorship of the field investigations and approval of loans of the collections to the Field Museum for study. On the home front I am indebted to Eunice Hoshizaki for computerization of this manuscript and curatorial assistance, Kathleen Kozol Telfer for technical assistance and scientific illustra- tions, Diane White and Linda S. Dorman of the Field Museum Department of Photography headed, by John Weinstein, for processing and printing the photographic film, and Associate LINGUAL LINGUAL li I TI I II q II IV IV Av Y 254 P. Hershkovitz Barbara Brown for indispensible assistance in every phase of the research. Bruce Patterson performed the heroic task of critically reviewing the manuscript. I am most grateful for the substantial improvements suggested. Financial support for the Southeastern Brazilian project is received from the Barbara E. Brown Mammal Research Fund. Zusammenfassung In diesem Report wird tiber einen Teil der Aufsammlungen von Kleinsáugern im Iporanga State Park, Sáo Paulo, im Parque Nacional de Caparaó, Minas Gerais-Espirito Santo, und an einigen Lokalitäten in Südost-Brasilien, die nicht vom Vefasser selbst besucht wurden, berich- tet. Beschreibungen werden gegeben von fiinf Gattungen (davon eine neu) und 14 Arten (6 neue, 1 noch unbeschrieben) sigmodontiner Nagetiere. Zwei friihere Berichte von anderen Autoren úber Sáugetiere des Caparaó National Park werden kritisch besprochen. Literature cited Avila Pires, E D. (1960): Roedores coleccionados na regiao de Lagoa Santa, Minas Gerais, Brazil. — Arquivos Mus. Nacion. (Rio de Janeiro) 50: 25—45. Azara, F. (1801): Essais sur l’histoire naturelle des quadrupedes de la Province du Paraguay, vol 2. — Paris, 499 pp. Translated from the original Spanish by M. L. E. Moreau- Saint-Méry. Azara, F. (1802): Apuntamientos para la historia natural de los quadrúpedos del Paraguay y Rio de la Plata. — Madrid, 2: i-iii + 328 pp. Blair, K. B. (1989): Islands in the sky: Mammalian taxonomy and community ecology of a neotropical mountain range. — Unpublished doctoral dissertation submitted to the faculty of the Graduate College of the Oklahoma State University, Stillwater, July 1989. Carleton, M. D. (1980): Phylogenetic relationships in Neotomine-Peromyscine rodents (Muroidea) and a reappraisal of the dichotomy within New World Cricetinae. — Miscel. Publ. Mus. Zool., Univ. Michigan 157: vii + 1—146. Cabrera, A. (1961): Catálogo de los mamíferos de America del Sur. — Rev. Mus. Argentino Cienc. Nat. “Bernardino Rivadavia” 4: 309—732. Davis, D. E. (1945): The home range of some Brazilian mammals. — J. Mammal. 26: HO 127 Davis, D. E. (1947): Notes on the life histories of some Brazilian mammals. — Bol. Mus. Nacion. (Rio de Janeiro), no. 76: 1—8. Ellerman, J. R. (1941): The families and genera of living rodents: Volume II. Family Muridae. — British Museum (Natural History), London. Fischer, G. (1814): Zoognosea tabulis synopticis illustrato. Vol. 3. — Moscow. Gyldenstolpe, N. (1932): A manual of Neotropical sigmodont rodents. — Küngl. Svenska Vetensk. Handl. Ser. 3, 11, no. 3: 1—164. Hensel, R. 1872 (1873): Beiträge zur Kentniss der Sáugethiere Siid-Brasiliens. — Abh. Königl. Akad. Wiss. Berlin 1872: 1—130. Hershkovitz, P. (1966): South American swamp and fossorial rats of the scapteromyine group (Cricetinae, Muridae) with comments on the glans penis in murid taxonomy. — Z. Säugetierkunde 31: 81 —149. Hershkovitz, P. (1977): Living New World monkeys (Platyrrhini) with an introduction to primates. Vol. 1. — University of Chicago Press, Chicago. Hershkovitz, P. (1990a): The Brazilian rodent genus Thalpomys (Sigmodontinae, Criceti- dae), with a description of a new species. — J. Nat. Hist. 24: 763— 783. Hershkovitz, P. (1990b): Mice of the Akodon boliviensis size class (Sigmodontinae, Crice- tidae) with the description of two new species from Brazil. — Fieldiana: Zoology 57: I=35: Hershkovitz, P. (1993): A new central Brazilian genus and species of sigmodontine rodent (Sigmodontinae) transitional between akodonts and oryzomyines, with a discussion of muroid molar morphology and evolution. — Fieldiana: Zoology 75: 1—18. Sigmodontine rodents from southeastern Brazil 255 Hershkovitz, P. (1994): The description of a new species of South American hocicudo or long-nose mouse, genus Oxymycterus (Sigmodontinae, Muroidea), with a critical review of the generic content. — Fieldiana: Zoology 79: 1—43. Hoffstetter, R. (1970): Vhistoire biogéographique des marsupiaux et la dichotomie marsu- piaux-placentaires. — Compt. rend. séanc. Acad. Sci. 271: 388—391. Hooper, E. T. & G. G. Musser (1964): The glans penis in neotropical cricetines (Family Muridae) with comments on classification of muroid rodents. — Miscel. Publ. Mus. Zool., Univ. Michigan 123: 1—57. Ihering, H. (1892): Os mamiferos do Rio Grande do Sul. — Annuario Estado do Rio Grande do Sul, 1893, de Graciano A. de Azambuya, pp. 3(96)—30(123) (reprint), Porto Alegre. Justo, E. R. & L. de Santis (1977): Akodon serrensis serrensis Thomas en la Argentina (Rodentia, Cricetidae). — Neotropica 23(69): 47—48. Koopman, K. F. (1993): Order Chiroptera, pp. 137—241. — in Wilson, D. E. & D. A. M. Reeder, eds., Mammal Species of the World: A Taxonomic and Geographic Reference, 2nd edition. Smithsonian Institution, Washington, D.C. Leche, W. (1886): Ueber einige stidbrasilianische Hesperomys-Arten. — Zool. Jb. 1886: 687—703. Liascovich, R. C. & O. A. Reig (1989): Low chromosomal number in Akodon cursor montensis Thomas and karyologic confirmation of Akodon serrensis. — J. Mammal. 70: 391—396. Lichtenstein, H. 1829 (1827—1834): Darstellung neuer oder wenig bekannter Sáugethiere in Abhandlungen und Beschreiben von 65 Arten auf 50 color Steintafeln nach den Origi- nalen. — Zool. Mus. Univ. Berlin, 10 Hefte. Mares, M. A., M.R. Willig, K. E. Streilein, & T. E. Lacher, Jr. (1981): The mammals of northeastern Brazil; A preliminary assessment. — Ann. Carnegie Museum 50: 81 —134. Massoia, E. (1961): Notas sobre los cricétidos de la selva marginal de Punta Lara (Mamma- lia, Rodentia). — Publ. Museo Municipal Ciencias Naturales y Tradicionales, Mar del Plata. Massoia, E. (1963a): Oxymycterus iheringi (Rodentia-Cricetide), nueva especie para la Argentina. — Physis 24: 130—136. Massoia, E. (1963b): Sobre la posición sistemática y distribución geográfica de Akodon (Thaptomys) nigrita (Rodentia-Cricetidae). Physis 24: 73—80. Massoia,E. & A. Fornes (1969): Caracteres comunes y distintivos de Oxymycterus nasutus (Waterhouse) y O. iheringi Thomas (Rodentia, Cricetidae). — Physis 28: 315—321. Meyen, E. J. E. (1833): Beiträge zur Zoologie gesammelt auf einer Reise um die Erde. — Ver- handl. Kaiserl. Leopold.-Carol. Akad. Wiss. 16(2): 551—610, pls. XL-XLVI. Musser, G. G. & M. D. Carleton (1993): Family Muridae, pp. 501—755. — In Wilson, D. E. & D. A. M. Reeder, eds., Mammals of the World: A Taxonomic and Geographic Reference, 2nd edition. Smithsonian Institution, Washington, D.C. Nowak, E. P. € J. L. Paradiso (1983): Mammals of the world. — John Hopkins University Press, Baltimore. Patterson, B. & R. Pascual (1972): The fossil mammal fauna of South America, pp. 247—309. — In Keast, A., F. C. Erk & B. Glass, eds., Evolution, Mammals and Southern Continents. State University of New York Press, Albany. Reig, O. A. (1986): Diversity patterns and differentiation of high Andean rodents, pp. 404—439. — In Vuilleumier, R. & M. Monasterio, eds., High Altitude Tropical Biogeo- graphy. Oxford University Press, Oxford. Reig, O. A. (1987): An assessment of the systematics and evolution of the Akodontini with the description of a new fossil species of Akodon (Cricetidae; Sigmodontinae). — Fiel- diana, Zoology 39: 347— 399. Ruschi, A. (1978): Mamiferos e Aves do Parque Nacional de Caparaó. — Bol. Mus. Biol. Dr. Melo Leitáo, Sér. Zool. 95: 1—28. Sanborn, C. C. (1931): A new Oxymycterus from Misiones, Argentina. — Proc. Biol. Soc. Wash. 44: 1-2. 256 P. Hershkovitz Thomas, O. (1896): On new small mammals from the Neotropical region. — Ann. Mag. Nat. Hist. 6(18): 301—314. Thomas, O. (1902): On mammals from the Serra do Mar, Paraná, collected by M. Alphonse Robert. — Ann. Mag. Nat. Hist. 7(9): 59—64. Thomas, O. (1903): Notes on neotropical mammals of the genera Felis, Hapale, Oryzomys, Akodon and Ctenomys with descriptions of new species. — Ann. Mag. Nat. Hist. 7(12): 234—243. Thomas, O. (1909): Notes on some South-American mammals, with descriptions of new species. — Ann. Mag. Nat. Hist. 8 (4): 230— 242. Thomas, O. (1916): The grouping of South American Muridae commonly referred to Akodon. — Ann. Mag. Nat. Hist. 8 (18): 336—340. Thomas, O. (1917): On the arrangement of the South American rats allied to Oryzomys and Rhipidomys. — Ann. Mag. Nat. Hist. 8 (20): 192—198. Trouessart, E. L. (1904): Catalogus mammalium tam viventium quam fossilium. Quinqui- nal supplementum, fasc. 1, 2, 929 pp. — R. Friedlánder und Sohn, Berlin. Vieira, O.da Cunha (1953): Roedores e lagomorfos do estado do Sao Paulo. — Arq. Zool. Estado Sáo Paulo 8(5): 129—168. Voss,R. S. (1993): A revision of the muroid rodent genus Delomys with remarks on “thoma- somyine” characters. — Amer. Mus. Novit. 3073: 1—44. Voss, R. S. £ M. D. Carleton (1993): A new genus for Hesperomys molitor Winge and Holochilus magnus Hershkovitz (Mammalia, Muridae) with an analysis of its phylogene- tic relationship. — Amer. Mus. Novit. 3085: 1—39. Waterhouse, G. R. (1837): Characters of a new species of the genus Mus from the collection of Mr. Darwin. — Proc. Zool. Soc. London 1837: 15—22, 27—29. Wilson, E. E. £ D. A.M. Reeder, eds. (1993): Mammal species of the world. A taxonomic and geographic reference. 2nd edition. — Smithsonian Institution Press. Washington, D.C., 1206 pp. Winge, O. 1887 (1888): Jordfunde og nulevende Gnavere (Rodentia) fra Lagóa Santa, Minas Geräes, Brasilien. — E Museo Lundii 1(3): 200 pp. Wyss, A.R., J.J. Flynn, M. A. Norell, C.C. Swisher III, R. Charrier, M. J. Nowa- cek & M. C. McKenna (1993): South America’s earliest rodent and recognition of a new interval of mammalian evolution. — Nature 365: 434—437. Ximénez, A.& A. Langguth (1970): Akodon cursor en el Paraguay (Mammalia-Criceti- nae). — Comun. Zool. Mus. Hist. Nat. Montevideo 10(128): 1—7. Ximénez,A., A. Langguth &R. Praderi (1972): Lista sistemática de los mamíferos del Uruguay. — Anal. Mus. Hist. Nat. Montevideo 7: 1—49. Yonenaga, Y. (1975): Karyotypes and chromosome polymorphism in Brazilian rodents. — Caryologia 28: 269—285. Zanchin, N. 1.-T., 1. J. Sbalqueiro, A. Langguth, R. € Boss le EIC zer sErer L. F. B. Oliveira&M. S. Matlevi (1992): Karyotype and species diversity of the genus Delomys. — Acta Theriol. 37: 163—169. Division of Mammals, Field Museum of Natural History, Roosevelt Road at Lake Shore Drive, Chicago, Ill., U.S.A. (deceased; correspondence to B. D. Patterson). Bd. $ 257-276 | Bonn, September 1998 A biochemical and morphological investigation of Suncus dayi (Dobson, 1888) and discussion of relationships in Suncus Hemprich «€ Ehrenberg, 1833, Crocidura Wagler, 1832, and Sylvisorex Thomas, 1904 (Insectivora: Soricidae) Paulina Jenkins, Manuel Ruedi, and Francois M. Catzeflis Abstract. A recent field expedition in South India yielded a series of seven specimens of Suncus dayi (Dobson, 1888), a poorly known crocidurine shrew collected in high- altitude wet evergreen forests in the Nilgiri Hills. The morphology (external, cranial, dental) of this species was investigated in a comparative study with a few taxa of the genera Suncus Hemprich & Ehrenberg, 1833 (e.g. S. stoliczkanus (Anderson, 1877), S. fellowesgordoni Phillips, 1932), Sy/visorex Thomas, 1904 (e.g. S. morio (Gray, 1862); S. granti Thomas, 1907 group) and Crocidura Wagler, 1832 (e.g. C. attenuata Milne- Edwards, 1872). A biochemical survey of isozyme variation at 32 genetic loci allowed the estimation of the genetic differentiation between S. dayi and four other white-toothed shrews: Suncus murinus (Linnaeus, 1766), Crocidura olivieri (Lesson, 1827), C. fuliginosa (Blyth, 1855) and C. russula (Hermann, 1780). In its cranial and dental morphology, Suncus dayi retains many plesiomorphic features common to some members of the African genus Sy/visorex, whereas other species of Suncus instead show many apomorphic states. The electrophoretic survey indicates that S. dayi shares derived characters with Crocidura and Suncus, not with Sylvisorex. In combination, these studies do not exclude the conclusion that the peculiar morphology of Suncus dayi could be the result of convergent evolution related to its possibly scansorial way of life. Key words. Suncus, Insectivora, India, systematics. Introduction Suncus dayi (Dobson, 1888) is a poorly known species from southern India recorded from the holotype deposited in the Natural History Museum, London and one other specimen presumed to be in the collection of the Zoological Survey, Bombay (see Lindsay 1929). Few authors elaborated on its original short description or commented on its status (Blanford 1888; Ellerman & Morrison-Scott 1966) and a century elapsed from the initial discovery, before seven additional specimens were collected from the Nilgiri Hills in southern India by the Catzeflis-Boursot Expedition to India in October 1990. These specimens have enabled us to study variation within the species and to perform more extensive comparisons with other taxa. Although S. dayi has always been associated with species currently assigned to the genus Suncus, it exhibits a suite of characters differing from those found in most other members of this genus, which suggest that its affinities lie elsewhere. Comparisons have therefore been made with other similar and closely related genera belonging to the subfamily Crocidurinae or white-toothed shrews, namely Suncus, 258 P. Jenkins et al. Crocidura and Sylvisorex, and theories of possible relationships have been advanced, as reported below. Besides the morphological examination, a comparative biochemical study of iso- zyme variation was undertaken in Suncus dayi and four other crocidurine shrews: Suncus murinus, Crocidura russula, C. fuliginosa, and C. olivieri. Electrophoresis of homologous proteins has proved an adequate and powerful approach for solving the systematics and evolutionary relationships of species and genera within the family Soricidae, as exemplified by recent studies of Catzeflis et al. (1985), George (1986) and Maddalena (1990b). The primary aim of the genetic survey was to decipher the probable relationships of Suncus dayi with regard to other white-toothed shrews, and to compare the new findings with those published for other Asian, African, and European members of the genus Crocidura sensu lato (Maddalena 1990b, Ruedi et al. 1990, 1993). Material Specimens used in the morphological analysis include the holotype of Suncus dayi BM (NH) 1867.11.14.7 adult of undetermined sex, skin and skull, Trichur, Cochin (10°31’N 77°13’E), and seven individuals from the Nilgiri Hills collected in October 1990 by P. Boursot, A. Orth and F. Catzeflis: V-543 undetermined sex: Ootacamund: wet evergreen forest, 2150 m, 11°24’N 76°42’E; V-557 juvenile male, V-562 subadult male, V-563 juvenile female, V-567 juvenile female: Avallanchi: wet evergreen primary forest, 2100 m, 11°23’N 76°36’E; V-572 subadult male, V-576 juvenile female: Kotagiri: wet evergreen primary and secondary forest, 1500 m, 11°26’N 76°53’E. The following comparative material of other taxa in The Natural History Museum collec- tions was examined: 13 specimens 6 specimens 3 specimens 28 specimens 7 specimens 7 specimens 17 specimens 11 specimens 9 specimens 20 specimens 20 specimens of Suncus stoliczkanus (Anderson, 1877) from Pakistan and India; of Suncus etruscus (Savi, 1822) from India; of Suncus fellowesgordoni Phillips, 1932 from Sri Lanka; of Suncus murinus (Linnaeus, 1766) from the Nilgiri Hills, India; of Suncus lixus (Thomas, 1898) from southern Africa; of Sylvisorex johnstoni (Dobson, 1888) from East and West Africa; of Sylvisorex granti Thomas, 1907 from East and West Africa; of Sylvisorex megalura (Jentink, 1888) from East and West Africa; of Sylvisorex morio (Gray, 1862) from Cameroon, West Africa; of Crocidura attenuata Milne-Edwards, 1872 from Assam, North East India; of Crocidura fuliginosa dracula Thomas, 1912 from North Vietnam. The material used in the electrophoretic study included frozen tissue samples of seven Sun- cus dayi (V-543, V-557, V-562, V-563, V-567, V-572 and V-576) from Nilgiri Hills; eight Suncus murinus (V-546 to V-551, V-554 and V-555) from Mudumalai, Tamil Nadu, South India; three Crocidura olivieri (Lesson, 1827) from Central Africa (Maddalena 1990a, b); five Crocidura russula (Hermann, 1780) from Switzerland (Maddalena 1990a); and three Crocidura fuligi- nosa (Blyth, 1855) from Malaysia (Ruedi et al. 1990). Voucher specimens of all these animals are deposited in the collections of Lausanne (IZEA) and/or London (BM[NH]). Methods Morphological analysis Measurements in millimetres were taken using dial calipers or a microscope measuring stage. The dental nomenclature follows that of Heim de Balsac & Lamotte (1957), Swindler (1976) and Butler & Greenwood (1979). In their bench-mark work on African Sylvisorex, Suncus and Crocidura, Heim de Balsac & Lamotte (1957) determined which external and cranial characters were primitive for this Biochemical and morphological investigation of Suncus dayi 259 group and which were derived. Butler & Greenwood (1979) and Butler et al. (1989) similarly assessed the mandibular characters of fossil and Recent African Soricidae. One of us (PJ) has provided additional characters from the maxillary dentition and external features. Using the studies of these authors as a framework for the current analysis, primitive (plesiomorph) and derived (apomorph) characters have been scored for Asian and African members of the genera Suncus and Sylvisorex, with the objective of obtaining a relative ranking of S. dayi. Electrophoretic study Tissue samples (kidney, liver, heart) were frozen in liquid nitrogen immediately after autopsy in the field, and maintained at -70°C at Lausanne until processing. Electrophoresis of homo- logous proteins was done as described in Ruedi et al. (1993) for studying the variation at the following 32 presumptive nuclear loci: (abbreviations as in Ruedi et al. 1993): Ada, Adh, Ak-1 & -2, Alb, Ck-1 & -2, Est-1, Got-1 & -2, Gpd, G-6-pd, Hk-1, Idh-1 & -2, Lap, Ldh-1 & -2, Mdh-1 & -2, ME, Mpi, Pa, Prot-x, 6-Pgd, Pgi, Pgm, Sod-1, -2 & -3, and Xdh-1 & -2. Alleles were designated by their mobility relative to the most common allele found in C. olivieri (Madda- lena 1990b). Individual genotypes were transformed into allelic frequences by the BIOSYS-1 program (release 1.7; Swofford & Selander 1981) in order to compute Rogers (1972) and Nel (1978) genetic distances between pairs of compared taxa. The Rogers (1972) distances, which are additive, were treated by the Neighbor-Joining (Saitou & Nei 1987) and distance Wagner procedures (in BIOSYS-1) in order to yield dendrograms built without the hypothesis of rate- equality of biochemical changes among lineages. Abbreviations used BM(NH): The Natural History Museum, London, British Isles (formerly British Museum [Natural History]); IZEA: Institut de Zoologie et Ecologie Animale, University of Lausanne, Switzerland; ISEM: Institut des Sciences de l’Evolution de Montpellier, France. c: circa; CBL: condylo-basal length; HB: head and body length; HF: hind foot length; m: metre; n: number; p: page; pers. comm.: personal communication; SD: standard deviation; TL: tail length. Abbreviations for dental nomenclature are given in the text. Results Morphological analysis Diagnosis and description: Suncus dayi is a dark brown, medium sized shrew (HB 70—78, CBL 18.9—20.2), with a long slender tail (TL 83—88), clothed in short hairs but lacking long bristle hairs. The hindfeet are elongated (15.5—16.5; ratio of HF to CBL 79.1—84.7), with elongated cheiridia. The cranium has a relatively short rostrum, with a broad interorbital region and the braincase is rounded and domed (see Figs 1—2 and Table 1). The fourth upper unicuspid (Un?) is slightly smaller than the second (Un?); the cingula on the upper unicuspids are broad and distally flared. The third upper molar (M?°) is long relative to the upper toothrow. The first lower incisor (I,) has two marked denticulations. The last lower molar (M5) has a distinct talonid basin and entoconid with a very short entoconid ridge. Comparison with other taxa: Crocidura is characterised by the possession of three upper unicuspid teeth, so clearly S. dayi does not conform in this diagnostic character; S. dayi also lacks the bristle hairs on the tail characteristic of Suncus and of most species of Crocidura. Crocidura apparently has a restricted distribution in India; only the very small Crocidura horsfieldii (Tomes, 1856) has been recorded from southern India (HB<75, CBL<18.0). Brief comparisons were therefore made P. Jenkins et al. 60 [% I'8—9°9 snunmu * pampa “110ys [dq snosn.ja :S “1 SRULIMU * paxmpal pampa poonpel poonpel “1104s [a snx1 :S “1 snosnjo * MOI. IBU pampa 310 Juasge paxmpal jJuOpoyjsido ‘1apusys “[1e7 [% #8 —9L SNISNAJO * MOLIBU ıhrL> [I Snulinul * V V Vv [y snosn.1jo * d Vv V V d V Vv % IL> [dq snosn.jo * V [9 9S—1S SNXY 'S]) % yS> p>usye]J possoidpe JOMOIIeU “19I10YS quosoid 3uo] juasaJd 3918] s1ydiowody snauns % SL "VL % LIS 1U0p103S9M0/]/9f snoung % S8—8'L % S'L< WyIBU9| MOIYIOO] Joddn 0} YASUE] ¿JA JO oney d dl p>dojA>p [pom “3uo] [y 1uojsuyof *s] d podoyaasp [Jem o[Ajsered [y juojsuyor 'S] al podojaaop [pom 3UO0JOdÁY [y 2uojsuyof *s] d podoyaaop [Jem 9u090}01d d p>doJsAsp [pom “3uo] uoye} d p>dopAsp [Jem ‘peoiq e]n3ulo jen3ur pue jesang 9ZIS UL P29INPII SSI] Jussaıd pidsnorun y}ınoq padoraaap [PM uoJe] juopoolxd ‘InoJs *ıoys dsnd 1oLIsJue :]I [% TL 1uojsuyor 'S] % 88 —9L % SL< yıpesig Arejjıxeu OF yIpesig [eAIOISJuUT JO ONE d I/d peoIq UOI331 JEIIQIOISJUJ % 09—8S % 8S< y13U9] MOIYIOO] Joddn 0} 1Y3r9y ssvoUTeIG JO ONEI P3JBA9]9 IO XIAUOI ISPIUILIJ p>>eds Afppım eIpLIIOUD J9peomg 9 193u0J :J99JPUTH Jussqe SUIUOYOTYI 110US AIOA sırey yuasqe IBSSIIQIA :JIEL 14Dp SNIUNS XOJOSIAJÁS 9781S IIYdIOWOISIIA I9JDBIBYD "9J8IS Jajoereyo o1ydiowode :y !UOISSIIAXI SJEIPIULISJUI JO J9JIBIBYO :] :9]8]S 19J9B18y9 9IYdAJOVIOISIA :q ‘snaung PUB X2LOSIAJÁS 0) IAH] 140p1O3S9MO0]]9/ SHIUNS pue 14VDp SHIUNS JO SIMIBIBYD) :I >]qeL 261 Biochemical and morphological investigation of Suncus dayi poryyduns juasqe 10 JOLI9JUL O] paurjuos MOJIBU % 08 O1 dn “1378913 J9]8913 Juasge juasqe [d snx1] 'S] Vv Vv V [% ET—LI SNXY “S| 3JOUL 10 00 6T—87 01 dn ‘1976913 310U1 10 % 00T o1 dn “131913 yuosqe 1104s JUISIOAIP “19yY3ty SUOU % 07< VI % 78< NA V [q snx1] 'S] V V [% SE—67 SNISNALO “S] % JE< 1918913 1918313 3/88 s1ydiowody V % 78< V snaung d Vv Vv % IZ ‘SI I % 8L ‘CL I 140p103589M0]/9S snauns [% O7—LI Yunis ‘S| % O7—81 % PT> d Id iy SJL % 98> I I Vv d d % ££—67 d % SI—LS d 1dpp SNIUNS XOLOSIAJÁS A SUI|QUIOS91 ‘podojaaop-[[om pruoje] JAJA LLO umnpn3un jensuly IPIM 93P9| pruodojua-3s0d “y *! SS3] IO % L9 y33uo] 01 YIPB9Iq Jo ore “1 Jo q13u9] ST’O 9 “J JO de]1sAo yuasaıd pr[A150301d yuosoid prA150301d SS3] IO % OZ y13u9] W—-!W 0} 14819y |] Jo ones SS3] JO % SL y¿3uo] Iw—-'"W 0} y¿3ua] '] jo one yuasaıd umpn3ur jen3umo1s}sod SUC] ‘ISpıoq jeıusAa 0) Jojfered mo] 93PLI Ten3urjfo1sJue OM} asp JO119JSOd uo SUOIJBA9]O 0% OF UY} SSI] y13u9] TA—'IN 01 yidep snweı [eJUOZIIOY Jo onel % 08 URY SS] JUBISY 07 YIPIM IJÁPUO9 JO one :9[QIPUBIA 978]S MIYAJOULOISI] A panunuo9 :] 91qeL AO) 262 P. Jenkins et al. between S. dayi and the slightly larger Crocidura attenuata found in Assam, north east India and in Southeast Asia (Assam specimens HB 70—87, CBL 19.7—21.6), and also with Crocidura fuliginosa from Southeast Asia, which is considerably larger (Vietnam specimens HB >85, CBL >22.2). The latter was the only Asian species available for the electrophoretic analysis, therefore it was also included in the morphological study, but because it is readily distinguished from Suncus dayi by its much greater size, comparisons given below were restricted to the more comparably sized C. attenuata. Crocidura attenuata differs externally from Suncus dayi in its grey brown pelage colour and shorter, bristle haired tail (ratio of TL to HB 70—95). The skull has a longer, shallower rostrum and the braincase is shallower, less rounded with slightly more angular superior articular facets. The mandible has a deeper, straighter horizontal ramus, while the ramal fossa is smaller and broader. The most obvious dental differences are the absence of the fourth upper unicuspid, and the shape of the first upper incisor, with its elongated anterior cusp (see Fig. 2). Suncus dayi is compared here with other species of Suncus which occur in India: S. murinus, S. stoliczkanus and S. etruscus. Suncus dayi is considerably larger than S. etruscus (HB <55, CBL< 14.0), and considerably smaller than S. murinus (HB > 100, CBL>23.5). Confusion with S. murinus and S. etruscus is therefore unlikely and no further reference is made to these two species. Suncus dayi (HB 70—78, CBL 18.9—20.2) is similar in size to S. stoliczkanus (HB 68—85, CBL 18.6—22.2); however the external appearance of S. dayi and S. stoliczkanus is otherwise quite different: S. dayi is dark brown dorsally and ventrally, the tail is uniformly dark brown, longer than head and body (TL 83—88; - ratio of TL to HB 109.9—123.5), clothed with short hairs but lacking the long bristle hairs which are characteristic of Suncus and most species of Crocidura; the hindfeet are elongated (15.5—16.5, ratio of HF to CBL 79.1— 84.7), the cheiridia are widely spaced and elongated. In contrast, Suncus stoliczkanus ranges in colour from pale grey to grey brown dorsally, paler ventrally; the tail is pale grey to grey brown, shorter than head and body (TL 44—55; ratio of TL to HB 60.8—76.8), with long bristle hairs; the hind feet are medium sized (10.5—15.0; ratio of HF to CBL 56.5—75.2), and the cheridia are adpressed and rounded. Suncus dayi is similar in many features to members of the African genus Sy/visorex, and it is compared here with the slightly smaller Sy/visorex morio, with additional comments on other species of Sy/visorex where S. morio is atypical. In the following external features, S. dayi shows a closer resemblance to members of the genus Sylvisorex than to those of Suncus. The tail is long, slender and, as in Sy/visorex, lacks the long bristle hairs which are characteristic of Suncus and most species of Crocidura; instead the tail has short hairs, which are longer and more numerous than those of S. morio (some Indomalayan species of Crocidura-C. elongata Miller & Hollister, 1921; C. miya Phillips, 1929 and C. paradoxura Dobson, 1887 — also have a long slender tail with few or no long bristle hairs). The cheiridia on the hindfeet of S. dayi and S. morio are elongated, unlike the rounded cheiridia of S. stoliczkanus. In craniodental morphology and proportions, S. dayi is also more similar to Sylvisorex than to Suncus (see Figs 1—6, and Tables 1—2). The skull of S. dayi is more similar to that of S. morio in general shape; the rostrum of both species is Biochemical and morphological investigation of Suncus dayi 263 Fig. 1: Dorsal view of cranium from left to right of Sy/visorex morio (BM [NH] 88.81), Suncus dayi (V-576), Suncus fellowesgordoni (BM[NH] 32.6.11.1), Suncus stoliczkanus (BM[NH] 79.586) and Crocidura attenuata (BM [NH] 27.3.7.2). TTT TT Fig. 2: Lateral view of cranium from left to right of above Sy/visorex morio (BM [NH] 88.81) and Suncus dayi (V-576), below Suncus fellowesgordoni (BM [NH] 32.6.11.1), Suncus stolicz- kanus (BM [NH] 79.586) and Crocidura attenuata (BM [NH] 27.3.7.2). relatively short, while that of S. stoliczkanus is slightly elongated and parallel-sided; the angle between the rostral portion of the maxilla carrying the unicuspid teeth and the premolar and molar portion of the maxilla is much more acute in S. stoliczkanus 264 P. Jenkins et al. Table 2: Selected measurements in specimens of Suncus stoliczkanus, Suncus dayi, Sylviso- rex morio and Crocidura attenuata. Range, mean + standard deviation, sample size. Suncus Suncus Sylvisorex Crocidura stoliczkanus dayi morio attenuata Condylobasal length Upper toothrow length Maxillary breadth at level of M? Interorbital breadth Braincase breadth Braincase height Braincase length Interorbital breadth: maxillary breadth Length of M*: upper toothrow length Braincase breadth: condylobasal length Braincase height: condylobasal length Braincase length: braincase breadth S26 R222 19278221403 11 81105 9.07 0.65 11 ITS) 5.98+0.40 11 os) 3.984 0.14 10 oo) masts 8.73+0.41 10 Ses dW) 4.30+0.32 10 VISIO 8.36+0.30 10 SARA 66.33+0.28 10 6.05720 6.48+0.28 10 42.0—45.3 43.90+1.12 10 19.8232 21.65+1.00 10 91.8—100.0 ZA 2033 10 18.9 2.02 19.66+0.40 7 IAS 8.66+0.13 7 5201620 5.83+0.16 7) 4.3—4.7 4.47+0.13 7 SÓ 9.341 0.22 7) IÜTTISS) Soll Pac), 6 188 8.09+0.21 7 129 88:0 78.10+4.40 7) TESTS 8:0772120.35 7 46.1—48.0 AOS 7 25.127720 25.82+0.58 6 83.9— 89.9 86.57+2.40 Ta INS I7-IEE050 9 7.8—8.4 8.07+0.21 O a 5.29E0.10 9 4.1—4.6 4.31+0.13 9 8.6—9.4 8.97+0.23 9 428522 Sila 0815 8 IR 1.20220-17 8 7559 80288 81.40+3.12 9 SAO 8.964 0.41 5) 47.0—52.6 50.08 + 1.53 9 20.9289 28.06+0.59 8 IO Sl 80.12+2.07 8 19.7239 20.20+0.52 10 0958 8.904 0.47 21 DES 6.004 0.23 2 4.1—4.7 4.37+0.15 14 I DES 9.08+0.32 11 4.4—5.3 4.89+0.25 10 035) 8.26+0.25 10 68.9. 17/8 72.69+2.24 14 7.1820 136=-0925 18 42.0—47.6 44.81+1.42 10 21.9253 24.22+1.04 9 87.9— 96.6 90.95+3.00 10 Biochemical and morphological investigation of Suncus dayi 265 Fig. 3: Buccal view of left upper Incisor and Un! from left to right of above Suncus dayi (V-576) and Suncus fellowesgordoni (BM [NH] 32.6.11.1), middle Sy/visorex granti (BM[NH] 11.4.7.12) and Sylvisorex morio (BM[NH]88.86), right Suncus stoliczkanus (BM[NH] 79.586). Scale 1 mm. Fig. 4: Occlusal view of right Un!—Un* and P* from left to right of Sylvisorex granti (BM [NH] 11.4.7.12), Sy/visorex morio (BM [NH] 88.86), Suncus dayi (V-576), Suncus fellowes- gordoni (BM [NH] 32.6.11.1) and Suncus stoliczkanus (BM [NH] 30.2.11.138). Scale 1 mm. than in S. dayi or S. morio. The interorbital region of S. stoliczkanus is relatively narrower and increases only slightly from anterior to posterior, whereas the posterior portion of the interorbital region is much broader than the anterior in S. dayi and S. morio. Both S. dayi and S. morio have rounded, domed braincases, while that of S. stoliczkanus is shallow, flat and laterally angular. The braincase is short relative to condylobasal length in S. morio, moderate in S. dayi but long in S. stoliczkanus. The parieto-occipital suture is prominently ridged in S. stoliczkanus, scarcely evident in S. dayi and practically absent in S. morio. The shape of the first upper incisor (I!) is more similar in S. dayi and S. morio than in S. stoliczkanus. In S. dayi and 266 P. Jenkins et al. S. morio this tooth is more proodont, the anterior cusp is moderately stout, and less than twice the height of the posterior cusp, while in S. stoliczkanus 1 is opisthodont, the anterior cusp is slender and elongated and more than twice the height of the posterior cusp. The cingula on the upper unicuspid teeth of S. dayi are broad and distally flared as in S. morio but unlike those of S. stoliczkanus, which further differs in the position of the fourth upper unicuspid (Un*), which is partially obscured by the protocone of the premolar (P°) in distal view. In S. dayi the lingual cingula of Un! , Un? and Un‘ and the buccal cingulum of Un? are broader than those of S. morio. In S. dayi and S. morio, P? has a prominent hypocone and well-developed talon, unlike S. stoliczkanus in which the hypocone is low and the talon reduced in size. The third upper molar (M7?) of S. dayi and S. morio is longer relative to the upper toothrow than that of S. stoliczkanus. The first lower incisor (I,) of S. dayi has two marked denticulations and the anterolingual ridge is complete and parallel with the ventral border as in S. morio but unlike S. stoliczkanus which shows no denticulation and the anterolingual ridge is incomplete and divergent. A protostylid is present on the lower premolar (P,) of S. morio and S. granti, but absent in most other species of Sy/visorex as well as in S. dayi and S. stoliczkanus. The talonid of the last lower molar (M;) differs in all three species; in S. morio the talonid basin, entoconid and entoconid ridge are well developed; in S. dayi, the talonid basin and entoconid are distinct but the entoconid ridge is very short, while in S. stoliczkanus the talonid basin is distinct, the entoconid ridge low but the entoconid less distinct. The development of the talonid of M, is variable within Sy/visorex but is usually reduced in Suncus. Historical perspective and morphological relationships: In his original description, Dobson (1888) placed the new species in Pachyura Sélys-Long- champs, 1839, which at that time was considered to be a subgenus of Crocidura but was later recognised as a distinct genus under the senior synonym of Suncus. An illustration of the lateral view of the anterior maxillary toothrow was given by Dobson (1890, Plate 28, Fig. 6). The few subsequent authors (Blanford 1888; Lindsay 1929; Ellerman & Morrison-Scott 1966) who examined this litte known and poorly represented species, only commented on the unusually large fourth upper unicuspid. Ellerman & Morrison-Scott considered that it was probably a member of the Suncus stoliczkanus group. In the course of collaborative work with R. Hutterer, Bonn (pers. comm. 1988), several of the taxa of Asian Suncus were examined, including the holotype of S. dayi and three specimens of S. fellowesgordoni, another poorly known taxon generally regarded as a subspecies of the widespread species Suncus etruscus (see Ellerman & Morrison-Scott 1966; Eisenberg & McKay 1970; Corbet & Hill 1992). These two taxa (S. dayi, S. fellowesgordoni) showed some similarities in several characters not generally found in other species of Suncus (see Table 1). Many of these characters are regarded as primitive, following the views of Heim de Balsac & Lamotte (1957), Butler & Greenwood (1979), and Butler et al. (1989). Butler (1978: p. 62) stated that some species of Suncus in Asia, such as S. fellowesgordoni from Sri Lanka, have primitive characters like Sy/visorex, such as a narrow mandibular condyle and a basined talonid on M,, and they cast doubt on whether the two genera should be Biochemical and morphological investigation of Suncus dayi 267 Fig. 5: Lingual view of right 1, and I, from left to right of above Sylvisorex granti (BM [NH] 71.1828) and Sylvisorex morio (BM [NH] 88.86), middle Suncus dayi (V-576) and Suncus fellowesgordoni (BM [NH] 32.6.11.1), below Suncus stoliczkanus (BM [NH] 79.586). Scale 1 mm. Fig. 6: Lingual view of right M; from left to right of above Sy/visorex morio (BM [NH] 88.86) and Syvisorex granti (BM [NH] 11.4.7.12), below Suncus dayi (V-576), Suncus fellowesgordoni (BM [NH] 32.6.11.1) and Suncus stoliczkanus (BM [NH] 79.586). Scale 1 mm. separated. Butler et al. (1989: p. 402) repeated this view that Suncus includes some primitive forms such as S. fellowesgordoni; they noted that most species of Suncus are from the Indomalayan Region and suggested that this genus may have arisen from a Sylvisorex-like form in Asia, which subsequently extended its range into Africa. 268 P Jenkins et al. Table 1 shows that Sy/visorex is plesiomorphic in the majority of its character states, while the majority of the characters of Suncus are apomorphic. Although not shown here, Crocidura attenuata and C. fuliginosa also show more apomorphic than plesiomorphic characters; it is hoped to discuss the relationships of these species of Crocidura and Suncus in a separate paper. Examination of S. fellowesgordoni confirms the above observations that many of its characters are plesiomorphic (see Table 1) and common to S. dayi and Sylvisorex, but also shows 10 shared derived features (synapomorphies) with Suncus and 3 with the Sylvisorex granti group. Comparison of S. dayi reveals an even larger suite of plesiomorphic characters in common with Sy/visorex, and a few synapomorphic features with Suncus (4), S. fellowesgordoni (3) and 2 with the Sylvisorex granti group. Electrophoretic study Among the 32 assayed presumptive loci, 19 did evidence some intra- and/or inter- specific polymorphism, whereas 13 showed no electrophoretic variation (Ak-1, Ck-1 & -2, Got-2, Gpd, Hk-1, Ldh-1, Mdh-1 & -2, Prot-x, Sod-3, and Xdh-1 & -2). Table 3 indicates the observed allelic frequences at all polymorphic loci. An intra-specific polymorphism for Suncus dayi does exist at the two most variable loci (Ada and Est-1), but only two heterozygotes were observed, namely alleles 65 and 87 at locus Ada for specimens V-543 and V-563. The genetic distances (based on all 32 loci) were computed by using the indices of Rogers (1972) and of Nei (1978) (Table 4), which indicate. a clearcut biochemical (genetical) differentiation between all samples (minimum value of ca. 0.28 for Nei’s : distance and 0.26 for Rogers’ distance). Interestingly, Suncus murinus is roughly equally divergent from S. dayi as from C. fuliginosa, and these are the smallest distances in the matrix of Table 4. According to both kinds of genetic distances, S. dayi is however slightly more related to Suncus murinus than to any other crocidurine shrew so far tested. These relationships are tentatively represented on the dendrograms derived from two clustering methods which do not depend upon the hypothesis of rate-uniformity (Neighbor-Joining and Wagner) neither of which take into account the standard-deviation of branch lengths estimates. Both tree-recon- struction methods indicate strong differences in rates of biochemical change, with for example C. russula having accumulated more changes than the other taxa. More- over, as illustrated by Figure 7, the two branching patterns conflict with regard to the position of the two species of Suncus relative to the Eurasian Crocidura. This is not astonishing, as a casual inspection of the half matrices of Table 4 indicates that there is no strong genetic discontinuity between these crocidurine shrews, as all taxa appear to split off at about the same genetic level of differentiation (at a genetic distance of ca. 0.3). Thus, the branching patterns derived from both clustering procedures suggest the paraphyly of the genus Suncus and/or of the genus Crocidura. Ecology Suncus dayi has recently been collected from tropical evergreen forest (with Rattus satarae Hinton, 1918, Rattus rattus [Linnaeus, 1758], Mus famulus Bonhote, 1898 and Suncus murinus) in the Nilgiri Hills at 1500—2150 m and is also recorded in the Biochemical and morphological investigation of Suncus dayi 269 C. olivieri C. fuliginosa S. murinus C. russula S. dayi [1 C. olivieri C. fuliginosa S. dayi C. russula S. murinus Fig. 7: Phenograms derived from the Neighbor-Joining (above) and Wagner (below) procedures on Rogers (1972) genetic distances. Both networks were arbitrarily rooted at midpoint of longest path. The branch lengths are a good approximation of the observed distances (of Table 4), as indicated by the values of the percent standard deviation (2.87 % and 6.16 %) and of the cophenetic correlation (0.984 and 0.968) for the Neighbor-Joining and Wagner trees, respectively. The bar (scale) is 0.05 units of genetic distance. literature from the Palni Hills (Lindsay 1927). There is doubt about the collection locality of the holotype, as Dobson (1888) stated that the exact locality was unknown but gave the habitat as the Madras Presidency, India. Blanford (1888: 240) surmised that this specimen was probably obtained from the Palni or Travancore Hills in the Madras Presidency; subsequently Blanford (1891: 602) explained that the specimen had been brought to the donor, Dr. F. Day, from Trichur, Cochin. All of these localities are in southern India; the vegetation consists of tropical, high altitude ever- green rain forests. Both Nilgiri and Palni are hill forests, isolated from each other by the surrounding lower lying country which consists of a mixture of secondary forests (most of which are cultivated by man) and cleared, cultivated land (mainly tea and potatoes) and scrub. Heim de Balsac & Lamotte (1957) stated that Sy/visorex and Suncus ecologically differ in Africa, Suncus occurring in semi-arid savanna, while Sy/visorex inhabits wetter forests (although one species, S. megalura — which has clear scansorial adaptations [Vogel 1974, Hutterer 1985] —, has been recorded in grassland). The combination of features such as a long tail, elongated feet with elongated, well-spaced cheiridia, were interpreted as adaptations to a scansorial way of life (Hutterer 1985) and are characteristic of Sy/visorex megalura, Suncus mertensi Kock, 1974 and of some Indomalayan Crocidura such as C. miya, C. paradoxura and C. elongata. The implication is that S. dayi is also scansorial and therefore probably occupies a different ecological niche to the sympatric pigmy, S. etruscus, and the 270 P. Jenkins et al. Table 3: Allelic frequences observed at 19 polymorphic loci among five studied species. N: sample size.. Suncus Suncus Crocidura Crocidura Crocidura Locus/Allele dayi murinus olivieri fuliginosa russula N=7 N=8 N=3 N=3 N=5 Biochemical and morphological investigation of Suncus dayi Din Table 3: continued Suncus Suncus Crocidura Crocidura Crocidura Locus/Allele dayi murinus olivieri fuliginosa russula N=7 N=8 N=3 N=3 NÉS giant, S. murinus, species. Regrettably there are no detailed observations to support these hypotheses. Discussion Species currently assigned to the genus Sylvisorex are confined to Africa. There is some controversy over the number of species recognised as belonging to the genus Suncus, which is found in Asia, Africa and Europe. Following Hutterer (1993) only four species are endemic to Africa, whereas nine occur in Asia and a tenth — S. murinus — is Asian in origin and believed to be a recent introduction to Africa (Heim de Balsac & Lamotte 1957, Hutterer & Tranier 1990). The area of greatest species diversity is generally accepted as the most probable site of origin of the group which, in the case of Suncus, is likely to be Asia. ‘Morphological results Heim de Balsac & Lamotte (1957) suggested that Sy/visorex is ancestral to Suncus, the former possessing a greater number of primitive, and the latter a greater number of derived characters. These authors briefly alluded to the resemblance noted by Ellerman et al. (1953) between the African S. /ixus and the Asian S. stoliczkanus. It was Butler (1978), Butler & Greenwood (1979) and Butler et al. (1989) who produced 272 P. Jenkins et al. Table 4: Genetic distances measured between five species of Crocidurinae. Below diagonal: Rogers (1972) genetic distance; above diagonal: Nei (1978) unbiased genetic distance. S.: Sun- cus; C.: Crocidura. Suncus Suncus Crocidura Crocidura Crocidura dayi murinus olivieri fuliginosa russula S. dayi S. murinus C. olivieri C. fuliginosa C. russula evidence of the possible derivation of African species of Suncus from Sylvisorex and furthermore, suggested that if this did not occur in Africa it might have done so from a Sylvisorex-like ancestral form in Asia which subsequently extended its range into Africa. They suggested that the fossil species of Suncus may have entered Africa rather late and that the main source of the immigrants was the Indomalayan region. They also demonstrated that Suncus lixus is dissimilar to the other African species of Suncus; it shares derived characters with two species of Sy/visorex, the fossil S. olduvaiensis Butler & Greenwood, 1979 and the Recent S. johnstoni (Dobson, 1888b) and could have been derived from a related species of Sy/visorex and, if it did not originate from Sy/visorex in Africa, it must be presumed to be an immigrant - from Asia. Butler et al. (1989) cited the primitive characters, reminiscent of Sylviso- rex found in some species of Asian Suncus, such as S. fellowesgordoni. It is therefore interesting to note that their view is further supported by another Asian species, S. dayi which shows an even higher proportion of primitive characters, so much so that it groups more readily with species of Sylvisorex than with those of Suncus. Butler et al. (1989) observed that Sy/visorex is definable only by primitive characters. These authors noted that Suncus was much more advanced than Sy/visorex and could feasibly be related to the Sylvisorex granti group (defined by these authors to include the species granti, megalura, howelli Jenkins, 1984, johnstoni and olduvaiensis). Table 1 shows that S. dayi has many primitive (plesiomorphic) features in common with Sylvisorex, but fewer with Suncus, which exhibits many more derived (apo- morphic) character states. The results show that Suncus dayi is synapomorphic in only four characters, whereas S. fellowesgordoni is synapomorphic with Suncus in ten characters. This suggests that S. fellowesgordoni has a closer relationship with other species of the genus Suncus than does S. dayi. Suncus dayi, however, retains many plesiomorphic features common to Sy/visorex, shows few synapomorphies with either Sy/visorex or Suncus, and exhibits such a close resemblance to Sy/visorex that it could feasibly be reassigned to this genus. However, as shown in the next section discussing electrophoretic data, the lack of shared derived alleles between S. dayi (this paper) and Sy/visorex (Maddalena et al. 1990a, b) is a strong argument for preferring the conservative decision to continue to assign S. dayi to the genus Suncus. Biochemical and morphological investigation of Suncus dayi 213 Based on the interpretation of characters listed in Table 1, the problem remains as to whether S. dayi is a relict species, differing only in a few apomorphic features from the Sy/visorex-like ancestral form of Suncus, as postulated by Butler et al. (1989); or one of convergent evolution due to adaptation to a similar ecological niche, as suggested by Hutterer (1985) who pointed out that convergent evolution may have occurred several times in scansorial shrews. Additional experiments with genetic markers (DNA sequencing, protein electro- phoresis) are needed to understand the relationships of Asian and African white- toothed shrews of the genera Suncus, Sylvisorex and Crocidura. Furthermore, our comparative morphological analysis should be extended to include Indomalayan scansorial species of Crocidura such as C. miya from Sri Lanka, C. paradoxura from Sumatra and C. elongata from Sulawesi in order to test for morphological conver- gence. Electrophoretic results The endemic South Indian Suncus dayi appears to be genetically well differentiated from the sympatric species Suncus murinus (Fig. 7). However, S. murinus is almost as divergent from S. dayi as from the Indomalayan C. fuliginosa (Nei’s distances of ca. 0.28: Table 4). This observation, as well as the conflicting topologies derived from the two phenetic reconstructions, suggest that all taxa analysed here might have diverged from each other at about the same time. Table 4 also reveals that the genetic distances measured in inter-generic comparisons are quite low by crocidurine standards: for example, Maddalena (1990b) has measured an average Nei’s distance of 0.72 (SD 0.12, range 0.50—0.92) between Suncus etruscus and 20 species of African and European Crocidura. When Indomalayan representatives of Suncus and Crocidura are compared (Figure 11), the apparent paraphyletic placement of Croci- dura is reminiscent of the finding of Maddalena (1990b), and this is in agreement with some traditional morphological studies (Heim de Balsac & Lamotte 1957, Butler et al. 1989). As Sy/visorex could not be included in this electrophoretic study, its genetic relationship with S. dayi cannot be directly assessed. Nevertheless, as Maddalena (1990a, b) used the same standard (Crocidura olivieri) for naming the different alleles, some comparison is possible for a subset of loci. In the cladogram proposed by Maddalena (1990b: 301), Sy/visorex megalura and S. lunaris (which were considered as an outgroup relative to Crocidura) shared two synapomorphic alleles: Ldh-2!”° and Mdh-2°*. Neither of these alleles has been found among our Asian samples, thus suggesting no direct link between S. dayi and these two species of Sylvisorex. On the contrary, allele Idh-1'% in Maddalena (1990a, b) is shared by most other crocidurine shrews, and it is also observed in S. murinus, S. dayi and C. fuliginosa. Several synapomorphies (AK-1'%, Est-1"°, Ldh-2'%, Mdh-2'%, and Pgi'%) link our Asian samples with most species of Crocidura analysed by Madda- lena (1990a, b), and not with Sy/visorex. These few genetic comparisons all indicate that Suncus dayi (and, by extension, the other Asian taxa analysed here) is not closely related to the African Sy/visorex, but instead shares several derived genetic characters with other shrews of the genera Crocidura and Suncus. The genetic affinities of Suncus dayi within a cluster containing Suncus murinus as well as some species of Crocidura indicate polyphyly for one or both genera. In conclusion, the genetic 274 P. Jenkins et al. results, although still preliminary, do not support a close relationship between S. dayi and Sy/visorex, but rather suggest parallelism for the cranial similarities and conver- gence for the external scansorial adaptation. Acknowledgements We thank the federal, state and regional agencies in India (New Dehli, Madras, and Ootaca- mund) for delivering collecting permits in different areas of the Nilgiris Mountains, Tamil Nadu State. The hospitality, kindness, and efficient help of Claude Caratini and his colleagues at the Institut Francais de Pondichery were very much appreciated. Funding of the expedition through NIH-NIAID grant AI29834 attributed to F. Bonhomme, Montpellier. MR thanks Prof. Peter Vogel for laboratory facilities at the Institute of Zoology and Animal Ecology of the University of Lausanne. This is contribution 1996—040 of the Institut des Sciences de PEvolution (UMR 5554) de Montpellier. Zusammenfassung Biochemische und morphologische Untersuchungen von Suncus dayi (Dobson, 1888) und die Diskussion der Verwandtschaft von Suncus, Crocidura und Sylvisorex (Insectivora: Sorici- dae). — Während einer Feldexpedition in den immergrünen Bergregenwald von Südindien konnten im Nilgiri Gebirge 7 Exemplare von Suncus dayi (Dobson, 1888), einer wenig bekann- ten Weißzahn-Spitzmaus, erbeutet werden. Die äußere Morphologie sowie Schädel- und Zahn- merkmale dieser Spezies wurden in einer vergleichenden Studie mit wenigen Arten der Gattun- gen Suncus (z.B. S. stoliczkanus, S. fellowesgordoni), Sylvisorex (z.B. S. morio, S. granti Gruppe) und Crocidura (z. B. C. attenuata) betrachtet. Eine biochemische Untersuchung der Isoenzymvariation an 32 Genloci erlaubt die Abschätzung der genetischen Differenzierung zwischen S. dayi und vier weiteren Weißzahn-Spitzmäusen: Suncus murinus, Crocidura olivieri, C. fuliginosa und C. russula. In der Schädel- und Zahnmorphologie teilt Suncus dayi' viele plesiomorphe Merkmale mit einigen Vertretern der afrikanischen Gattung Sy/visorex, wohingegen andere Arten der Gattung Suncus in diesen Merkmalen apomorphe Muster auf- weisen. Eine elektrophoretische Untersuchung zeigt Übereinstimmung in abgeleiteten Merk- malen zwischen Crocidura und Suncus, aber nicht zwischen Sy/visorex und Suncus. Insgesamt kann diese Studie nicht ausschließen, daß die eigentümliche Merkmalsverteilung bei Suncus dayi ein Ergebnis konvergenter Entwicklung in Verbindung mit kletternder Lebensweise ist. Literature Anderson, J. (1877): Description of some new and little known Asiatic shrews in the Indian Museum, Calcutta. — J. Asiatic Soc. Bengal 46: 261 —283. Blanford, W. T. (1888): The fauna of British India, including Ceylon and Burma. Part 1. — Taylor and Francis, London. Blanford, W. T. (1891): The fauna of British India, including Ceylon and Burma. Part 2. — Taylor and Francis, London. Blyth, E. (1855): Proceedings of the Asiatic Society of Bengal for May 1855. — J. Asiatic Soc. Bengal 24: 354—364. Bonhote, J. L. (1898): Description of a new species of Mus from southern India. — J. Bom- bay Nat. Hist. Soc. 12: 99—100. Butler, P. M. (1978): Insectivores and Chiroptera. — In: Maglio, V. J. & H. B. Cooke: Evo- lution of African mammals., Harvard University Press, Cambridge. Butler, P.M. & M. Greenwood (1979): Soricidae (Mammalia) from the early Pleistocene of Olduvai Gorge, Tanzania. — Zool. J. Linn. Soc. 67: 329—379. Butler, P. M., R.S. Thorpe &M. Greenwood (1989): Interspecific relations of African crocidurine shrews (Mammalia: Soricidae) based on multivariate analysis of mandibular data. — Zool. J. Linn. Soc. 96: 373—412. Biochemical and morphological investigation of Suncus dayi PAYS Catzeflis, E M., T. Maddalena, S. Hellwing & P. Vogel (1985): Unexpected findings on the taxonomic status of East Mediterranean Crocidura russula auct. (Mammalia, Insectivora). — Z. Sáugetierk. 50: 185—201. Corbet, G. B. & J. E. Hill (1992): The mammals of the Indomalayan region: a systematic review. — Oxford University Press, Oxford. Dobson, G. E. (1887): Description of new species of Soricidae in the collection of the Genoa Civic Museum. — Ann. Museo Civico Storia Nat. Giacomo Doria, Genova 4: 564—567. Dobson, G. E. (1888a): Description of two new species of Indian Soricidae. — Ann. Mag. Nat. Hist. 6: 427—429. Dobson, G. E. (1888b): On the genus Myosorex, with descriptions of a new species from the Rio del Rey (Cameroons) district. — Proc. Zool. Soc. London 4: 575—578. Dobson, G. E. (1890): A monograph of the Insectivora, systematic and anatomical. Part 1. — John van Voorst, London. Eisenberg, J.-F £ G. M. McKay (1970): An annotated checklist of the recent mammals of Ceylon with keys to the species. — Ceylon J. Sci., Biol. Sci. 8: 69— 99. Ellerman, J.R.& T.C.S. Morrison-Scott (1966): Checklist of Palaearctic and Indian mammals, 1758 to 1946 — British Museum (Natural History), 2nd edition. — London. Ellerman,J. R., T. C.S. Morrison-Scott & R. W. Hayman (1953): Southern African Mammals 1758 to 1951: A reclassification — British Museum (Natural History). — London. George, S. B. (1986): Evolution and historical biogeography of soricine shrews. — Syst. Zool. 35: 153—162. Gray, J. E. (1862): List of Mammalia from the Camaroon (sic) Mountains, collected by Capt. Burton, H. M. Consul, Fernando Po. — Proc. Zool. Soc. London 2: 180—181. Heim de Balsac, H. & M. Lamotte (1957): Evolution et phylogénie des Soricidés africains. II. La lignée Sy/visorex — Suncus — Crocidura. — Mammalia 21: 15—49. Hemprich, FE. W. &C.G. Ehrenberg (1833): Symbolae physicae seu icones et descriptio- nes mammalium. 2. Berolini. Hermann, J. H. (1780) — In: Zimmermann, E. A. W.: Geographische Geschichte des Menschen, 2. Leipzig. Hinton, M.A. C. (1918): Scientific results from the mammal survey, n° XVIII. — J. Bombay Nat. Hist. Soc. 26: 59—88. Hutterer, R. (1985): Anatomical adaptations of shrews. — Mammal Rev. 15: 43—55. Hutterer, R. (1993): Order Insectivora. — In: Wilson, D. E. & D. M. Reeder: Mammal species of the world. A taxonomic and geographic reference, 69—130. Smithsonian Institution Press, Washington and London. Hutterer, R. & M. Tranier (1990): The immigration of Asian house shrew (Suncus murinus) into Africa and Madagascar. — In: Peters, G. & R. Hutterer: Vertebrates in the tropics, 309—320. Museum Alexander Koenig, Bonn. Jenkins, P. D. (1984): Description of a new species of Sy/visorex (Insectivora: Soricidae) from Tanzania. — Bull. Br. Mus. nat. Hist. (Zool.) 47: 65—76. Jentink, F. A. (1888): Note 1. Zoological researches in Liberia, a list of mammals collected by J. Büttikofer, C. F. Sala und F. X. Stampfli. — Notes Leyden Mus. 10: 1—58. Kock, D. (1974): Eine neue Suncus-Art von Flores, Kleine Sunda-Inseln (Mammalia, Insecti- vora). — Senckenberg. biol. 55: 197—203. Lesson, R. P. (1827): Manuel de mammalogie, ou histoire naturelle des mammiferes. — Paris. Lindsay, H. (1927): Bombay Natural History Society’s mammal survey of India, Burma and Ceylon. Report n° 43. Nelliampathy and Palni Hills. — J. Bombay Nat. Hist. Soc. 31: 591—597. Lindsay, H. (1929): Scientific results from the mammal survey, n° XLVIII. Indian shrews. — J. Bombay Nat. Hist. Soc. 33: 326—340. Linnaeus, C. (1758): Systema naturae. 10th edition. — Holmiae. Linnaeus, C. (1766): Systema naturae per regna tria naturae secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis. 12th edition. — Holmiae. 276 P. Jenkins et al. Maddalena, T. (1990a): Systematics and biogeography of Afrotropical and Palaearctic shrews of the genus Crocidura (Insectivora: Soricidae): An electrophoretic approach. — In: Peters, G. & R. Hutterer: Vertebrates in the tropics, 297—308. Museum Alexander Koenig, Bonn. Maddalena, T. (1990b): Systématique, évolution et biogéographie des musaraignes afro- tropicales et paléarctiques de la sous-famille des Crocidurinae: une approche génétique. — PhD, Université de Lausanne, Suisse. Miller, G. S. & N. Hollister (1921): Twenty new mammals collected by H. C. Raven in Celebes. — Proc. Biol. Soc. Washington 34: 93—104. Milne-Edwards, A. (1872): Recherches pour servir a l’histoire naturelle des Mammiferes. — Paris. Nei, M. (1978): Estimation of average heterozygosity and gemelo distance from a small number of individuals. — Genetics 89: 583—590. Phillips, W. W. A. (1929): New and rare Ceylon shrews. — Spolia Zeylanica 15: 113—118. Phillips, W. W. A. (1932): Additions to the fauna of Ceylon. No 3. A new pigmy shrew from the mountains of central Ceylon. — Ceylon J. Sci. (B) 17: 123—126. Rogers, J. S. (1972): Measures of genetic similarity and genetic distance. — Univ. Texas Stud. Genetics 7: 145—153. Ruedi, M., T. Maddalena, P. Vogel & Y. Obara (1993): Systematic and biogeographic relationship of the Japanese white-toothed shrew (Crocidura dsinezumi). — J. Mamm. 74: 535—543. Ruedi, M., T. Maddalena, H-S. Yong & P. Vogel (1990): The Crocidura fuliginosa species complex (Mammalia: Insectivora) in peninsular Malaysia: biological, caryological and genetical evidence. — Biochem. Syst. Ecol. 18: 573—581. Saitou, N. & M. Nei (1987): The neighbor-joining method: a new method for reconstructing phylogenetic trees. — Mol. Biol. Evol. 4: 406—425. Savi, P. (1822): Osservazioni sopra il mustietto o mustiolo, nuova specie di toporagno Toscano Sorex etruscus. — Nuovo Giorn. Letter., Pisa 1: 60—71. Sélys-Longchamps, E. de (1839): Etudes de micromammalogie. Paris. Swindler, D. R. (1976): Dentition of living primates. — Academic Press, London. Swofford, D. L. & R. B. Selander (1981): BIOSYS-1: a Fortran program for the comprehensive analysis of electrophoretic data in population genetics and systematics. — J. Heredity. 7/2: 23153289: Thomas, [M. R.] O. (1898): On the mammals obtained by Mr A. Whyte in Nyasaland, and presented to the British Museum by Sir H. H. Johnston, K. C. B.; being a fifth contri- bution to the mammalfauna of Nyasaland. — Proc. Zool. Soc. Lond.: 925—939. Thomas, [M. R.] O. (1904): — Abstr. Proc. zool. Soc. Lond. (10): 12. Thomas, [M. R.] O. (1907): On further new mammals na by the Ruwenzori Expedi- tion. — Ann. Mag. Nat. Hist. 7: 118—123. Thomas, [M. R.] O. (1912): New species of Crocidura and Petaurista from Yunnan. — Ann. Mag. Nat. Hist. 8: 686—688. Tomes, R. F. (1856): In Blyth, E. Memoir on the Indian species of shrews. — Ann. Mag. Nate Hist222 0228: Vogel, P. (1974): Note sur le comportement arboricole de Sy/visorex megalura (Soricidae, Insectivora). — Mammalia 38: 171—176. Wagler, [J.] (1832): In Oken, [L.] Isis. 3. Leipzig. Paulina Jenkins, The Natural History Museum, Mammal Group, Cromwell Road, London SW7 5 BD, United Kingdom; Manuel Ruedi, Institut de Zoologie et Ecolo- gie Animale, Université de Lausanne, Batiment de Biologie, CH-1015 Lausanne, Switzerland; Francois M. Catzeflis, Institut des Sciences de l’Evolution, UMR 5554 CNRS, Case 064, Université Montpellier II, 34095 Montpellier, France. Ba. 4 S. 217-282. | Bonn, September 1998 Biogeographical and karyological data of the Microtus savil group (Rodentia, Arvicolidae) in Italy L. Galleni, R. Stanyon, L. Contadini & A. Tellini Abstract. Two species in an early stage of speciation were found in Italy within the Microtus savii complex: M. savii in the described area of the species and M. brachycercus in Calabria. Karyological analyses of specimens from a wider geographic area confirmed the occurrence of the G-banded karyotype of M. savii in north-central Italy and Sicily. The G-banded karyotype of M. brachycercus was found in Calabria and a third chromosome form with an acrocentric X in the south-east of Italy. The C-banded karyotype confirms the homogeneity of autosomal C-bands, apart from a small variant in the third chromo- somes pair, and C-band variations in the sex chromosomes. Key words. Rodents, Microtus savii complex, chromosome banding, polymorphism. Introduction On the basis of slight differences in morphological traits three different subspecies of the Savi pine vole Microtus savii (De Sel.) were described in Italy. Comparisons among banded karyotypes of specimens from Rosarno (Calabria), where the sub- species M. savii brachycercus (von Lehmann, 1961) occurs, and from central Italy (Pisa and Viterbo), where the subspecies M. savii savii (De Selys Longchamps, 1838) is present, revealed identical autosomal complements but marked differences in sex chromosomes (Galleni et al. 1992). Since male hybrid specimens from crosses between the two different karyomorphs were sterile, M. savii brachycercus and M. savii savii were considered different species in an early stage of speciation (Galleni et al. 1994). Krapp & Winking (1976) found a karyotype similar to that of M. savii Savii in a third subspecies, M. savii nebrodensis (Mina-Palumbo, 1868), from Sicily. A different karyological form with an acrocentric X was described in south- eastern Italy (Monte Gargano) (Niethammer 1981) and later classified as M. savii savii on the basis of tooth morphology (Brunet-Lecomte 1988). As no banded karyo- type was reported, nothing can be said about the general level of autosomal similarity with the other karyotypes. For a better knowledge of the level of chromosome polymorphism within the Microtus savii complex (according to Santini 1978), further investigations on C- and G-banded karyotypes of specimens from different Italian populations were carried Out in this study. Material and methods Specimes were trapped at different localities (Fig. 1): Ficuzza (Palermo, Sicily; 1 female (F); Fiume Freddo (Cosenza, Calabria; 1 F); Metaponto (Potenza, Basilicata; 1 F), Monte Cimone (Modena, Emilia; 1 F) and Parco Lambro (Milano, Lombardia, 1 male (M)). 278 L. Galleni et al. e A I > AS N I = ie = 2 \ SL { \ AA ( NA ] IN | ( Di, PAS í ) = SF € \ fs a | Wey IN (ES \ MN) a (ae \ a ‚m RE NR f > A / Si f x R Fig. 1: Collecting sites of specimens reported in the present paper (open symbols) and in literature (filled symbols) (Krapp & Winking 1976; Niethammer 1981; Galleni et al. 1992). Square: Microtus brachycercus; circle: Microtus savii (X a triangle Microtus savii (X acrocentric). Chromosomes were prepared from fibroblast cultures of short terminal tail biopsies (Stanyon & Galleni 1991) and C-, G-banded following previously reported techniques (Galleni et al. 1992). Results The G-banded karyotype of specimens from Ficuzza (PA) Monte Cimone (MO) and Parco Lambro (MI) matched that of Microtus savii from central Italy while the G-banded chromosomes of the female from Fiume Freddo (CS) agreed with those previously found in Microtus brachycercus (see Galleni et al. 1992). The female from Metaponto exhibited the same autosomal pattern of the other specimens but an acrocentric medium size X chromosome which differed from the metacentric X of Microtus savii (Fig. 2). | Autosome C-banding pattern was the same in all the specimens surveyed except a larger amount of pericentromeric heterochromatin in the third chromosomes pairs of specimens with the M. savii X metacentric karyotype (Fig. 3). A centromeric C-band, a band in the middle of the arm and a third band about half way between them were detected in the X chromosome of the female from Metaponto (Fig. 4). All the other specimens showed C-banded karyotypes similar to those described in Galleni et al. (1992). 279 Speciation of the Microtus savii group 2.0 7 Y Were 7 ato wu we e O HAZ AN 5 Pp) 23 a AJA OO Bey “2 Mit yyy Pee Ye o >. 5 Fig. 2: G-banded karyotype of a female M. savii from Metaponto (PZ). Pisa; D = Monte Cimone; = Viterbo; C Milano; B Discussion Fiume Freddo). Rosarno; G Metaponto; F vole of the Italian peninsula actually is a heterogeneous group including two Morphological, karyological data and hybridological tests showed that the Savi pine different species in an early stage of speciation: M. savii in north and central Italy and M. brachycercus in the south-west (Galleni et al. 1994, Galleni 1995). The Fig. 3: C-bands of chromosome 3 (A E 280 L. Galleni et al. nn — | oe Ey ET) : i o 8 ee La es | ; uD |" im A 55 A B C Fig. 4: G-banded (left) and C-banded (right) X chromosomes of M. savii (A: acrocentric, C: metacentric) and M. brachycercus (B). question of the taxonomic state of populations from different localities of southern Italy and Sicily is still open however. The agreement between karyotypes of specimens from northern and central Italy (Parco Lambro and Monte Cimone) and that of the specimen from Sicily supports the classification suggested by Krapp & Winking (1976) who considered these populations as all belonging to the subspecies M. savii savii. It appears, in this case, that allopatric distribution has not been followed by differentiation at specific level at least in respect of morphological and karyological traits. The presence of specimens with the same karyotype as M. brachycercus at Fiume - Freddo (Calabria) confirms the occurrence of this species in the Calabrian peninsula. The X-acrocentric karyotype of the female from Metaponto (Potenza) is similar to that described by Niethammer (1981) for a specimen from Monte Gargano (Foggia). Because of the proximity of the two collecting sites (Fig. 1) 1t is likely that we are dealing with the same form. Hypotheses on the relationships between the metacentric X chromosome of M. savii and the sub-metacentric X of M. brachycercus were discussed previously (Galleni et al. 1992). A similarity of G-banding pattern between the euchromatic part of X chromosome long arm of M. brachycercus and the whole metacentric X of M. savii was also confirmed in this study (Fig. 4). The acrocentric X chromosome is similar to the metacentric X but it is not possible to resolve the bands to ascertain if a pericentric inversion took place in the X of M. savii generating the acrocentric form. According to the taxonomic revision based on tooth morphology (Brunet-Lecomte 1988) the forms of M. savii with different X chromosomes would belong to the same species. However no crosses have been carried out so far to confirm this view. Up to date a clear speciation process has been demonstrated only for karyomorphs (M. savii and M. brachycercus) with different heterochromatin composition of their sex chromosomes, although unequivocal evidence of an active role of this class of chromatin in the formation of reproductive barriers has not been demonstrated yet (Galleni et al. 1994). Finally, heterochromatin polymorphism observed for the third chromosome will be a useful marker for studying phyletic relationships among the different forms of Speciation of the Microtus savii group 281 this group if confirmed on a wider sample. However, this polymorphism does not exceed the level of polymorphism found in many other mammalian species and is not considered to affect fitness. Acknowledgements Thanks are due to M. Sala (University of Modena) who provided the specimen from Monte Cimone and G. Aloise (C.N.R. Roma) who provided the specimens from Ficuzza and Fiume Freddo. This work was supported by grants from Italian MURST (Ministero dell’Universita e della Ricerca Scientifica e Tecnologica). Zusammenfassung Innerhalb des Microtus savıi Komplexes können 2 Arten unterschieden werden: M. savii im beschriebenen Areal und M. brachycercus in Kalabrien. Die beiden Arten befinden sich in einem frühen Stadium der Artbildung. Karyologische Analysen von Individuen eines weiteren geographischen Gebietes bestätigten das Vorkommen von G-gebänderten Karyotypen von M. savii in Nord-Zentral-Italien und auf Sizilien. G-gebänderte Karyotypen von M. brachycer- cus wurden in Kalabrien gefunden und eine dritte Chromosomenform mit einem akrozen- trischen X in Süd-Ost-Italien. Die C-gebänderten Karyotypen bestätigten die Homogenität der autosomalen C-Bänderung, zum Teil von einer kleinen Variante im dritten Chromosomen- paar, und eine Variation der C-Bänderung in den Geschlechtschromosomen. References Brunet-Lecomte, P. (1988): Les campagnols actuels et fossiles d’ Europe occidentale. — These de Doctorat présentée a Université de Bourgogne. Brunet-Lecomte, P. & J. Chaline (1990): Relations phylogénétiques et Evolution des campagnols souterrains d’ Europe (Terricola, Arvicolidae, Rodentia). — C. R. Acad. Sci. Paris, 311, ser. II: 745 —750. Galleni, L. (1995): Speciation in the Savi pine vole, Microtus savii (De Sel. L.) (Rodentia, Arvicolidae) a theoretical biology approach. — Boll. Zool. 62: 45—51. Galleni, L., A. Tellini, R. Stanyon, A. Cicalo &L. Santini (1994): Taxonomy of Microtus savii (Rodentia, Arvicolidae) in Italy: cytogenetic and hybridization data. — J. Mammal. 75: 1040—1044. Galleni, L., R. Stanyon, A. Tellini, G. Giordano &L. Santini (1992): Karyology of the Savi pine vole: Microtus savii (De Sélys Longchamps, 1838) (Rodentia, Arvicolidae): C, G, DA/DAPI and Alu-I bands. — Cytogen. Cell Genet. 59: 290—292. Krapp,F.&H. Winking (1976): Systematik von Microtus (Pitymys) subterraneus (De Sélys Longchamps, 1836) und savii (De Sélys Longchamps, 1838) auf der Appenninen-Halbinsel und benachbarten Regionen. — Sáugetierk. Mitt. 3: 166—179. Niethammer, J. (1981): Über Microtus (Pitymys) savii (De Selys Longchamps, 1838) vom Monte Gargano, Italien. — Säugetierk. Mitt. 29: 45—48. Santini, L. (1978): European field voles of the genus Pitymys McMurtrie and their damage in agriculture, horticulture and forestry. — EPPO Bulletin 7: 243—253. Stanyon, R. & L. Galleni (1991): A rapid fibroblast culture method for mammalian chromosome. — Boll. Zool. 58: 81—83. L. Galleni, A. Tellini, Dipartimento di Coltivazione e Difesa delle Specie legnose Sez. Entomologia agraria, Universita di Pisa. Via San Michele degli Scalzi 2, 1-56124 Pisa, Italy. — R. Stanyon, Istituto di Antropologia fisica, Universita di Genova, Via Balbi 4, I-16124 Genova, Italy . — L. Contadini, Scuola Superiore di Studi Universitari e Perfezionamento “S. Anna”, Via Carducci 40, I-56124 Pisa, Italy. Bd. 47 : 80289292 Bonn, September 1998 Reproduction in Mus macedonicus (Mammalia: Rodentia) in the Balkans Vladimir Vohralik, Theodora S. Sofianidou & Daniel Frynta Abstract. The material of 538 specimens of Mus macedonicus was collected during 1983 —1994 in 64 localities in northern Greece, southern Bulgaria and former Yugoslav Macedonia. Duration of the breeding period, sexual maturation in relation to the body weight, proportion of sexually active individuals in the population, intensity of repro- duction, litter size, and embryonal resorption were evaluated. Reproduction had distinctly seasonal character. Litter size was between 4 and 10, x = 6.73 (n = 45). Our results in Mus macedonicus were compared with the published data on other free living mice species of the genus Mus in the western part of the Palaearctic region. Key words. Mus macedonicus, wild mice, reproduction, litter size. Introduction Mus macedonicus Petrov & Ruzic, 1983, is a strictly free living mouse species inhabiting the south of the Balkan Peninsula (Orsini et al. 1983, Vohralik & Sofianidou 1987, 1992), as well as the large territories of the Middle East extending from Turkey, to Armenia and Azerbaijan in the north, to Iran in the east, and to Israel in the south (Kratochvil 1986, Auffray et al. 1990b, Orlov et al. 1992). Despite its large distribution area it was recognized as a distinct species only in the 1980s, when a considerable research effort was devoted to the taxonomy of Mus musculus sensu lato. Due to this fact, hitherto studies dealing with Mus macedonicus were confined mostly to problems associated with systematics and taxonomy, e. g., bio- chemical systematics (Bonhomme et al. 1984), karyology (Orlov et al. 1992), morphology (Orsini et al. 1983, Kratochvíl 1986, Gerasimov et al. 1990, etc.), and behaviour (Frynta et al. 1992). Despite that its specific status was accepted universally, there is no agreement about its valid name. So far, the following names were suggested successively: Mus abbotti Waterhouse, 1837, Mus tataricus Satunin, 1908, Mus spretoides Thaler, 1983 (nomen dubium, suggested by Bonhomme et al. 1984), Mus macedonicus Petrov & Ruzic, 1983. Recently, the validity of the three former names has been cast into doubt (cf. Marshall 1986, Auffray et al. 1990a) and the later one was suggested by Musser & Carleton (1993) as currently valid name. Although it is much probable that the taxonomic revision of all Mus taxa described from the distribution range of the species will reveal older valid names, we follow the above authors and use the name Mus macedonicus here. There is fairly good information on the reproduction in at least two free living mice species of western Palaearctic. The most extensive data are available about the Mound building mouse, Mus spicilegus. From the 1940s it was studied by various authors in south Ukraine and Moldavia, and the results were summarized in the 284 V. Vohralik et al. book by Sokolov et al. (1990). Valuable information is also contained in the mono- graphy by Mikes (1971) who studied in detail its ecology in the Vojvodina region, Yugoslavia. The other, thoroughly studied species was Mus spretus that received considerable attention only recently, in Spain and France (e. g., Vargas et al. 1984, 1986, 1991, Cassaing & Croset 1985, Durán € Sans-Coma 1986, Durán et al. 1987). Other free living mice populations the taxonomic status of which, unfortunately, remains unclear were studied extensively in Kazakhstan, Central Asia (Borisenko 1977). Also Mus musculus populations collected in fields in former Czechoslovakia by Pelikán (1974) could be included among free living mice. In the south of the Balkans, the only comprehensive study on the biology of free living mice populations (Straka 1966) suffers from the unclear specific status of the used material, that may be composed from three mouse species, at least. According to our knowledge, the only data on the biology of reproduction: in Mus macedonicus are few records on litter size from southern Bulgaria by Markov (1964) who referred this mouse to Mus musculus hispanicus. Therefore, the aim of this study is to provide a basic information about the reproductive biology in Mus macedonicus and to compare it with that in other free living mice species from the western part of the Palaearctic region. Material and methods Altogether, 538 specimens of Mus macedonicus captured during the period 1983—1994 in 64 localities in the south of the Balkan Peninsula, were used in our analysis. Most of the material was collected in northern Greece: 34 localities in Macedonia (275 specimens), 17 localities in Thrace (186 specimens) and 3 localities in Epeirus (9 specimens). For details about localities, © date and number of specimens collected in most of the above material see Vohralík & Sofianidou (1987, 1992) and Sofianidou & Vohralik (1991). Additional 30 specimens were collected in 4 localities of southeastern Bulgaria (Krumovo, district Jambol; Dábovec and Knizovnik, district Chaskovo; Sozopol, district Burgas), 27 specimens in 3 localities of south- western Bulgaria (Sandanski, Strumjani and Baldevo, district Blagoevgrad) and 11 specimens in 3 localities (Star Dojran, Pretor and Vozarci) in former Yugoslav Macedonia. All the material is deposited in collections of the Department of Zoology, Charles University, Prague. Mice were obtained mostly by snap traps. Body weight was recorded to the nearest gram. Next, mice were dissected and the condition of their reproductive organs was ascertained. Embryos were counted macroscopically and their length (in the longitudinal axis) was measured. The presence of placental scars was also recorded. Embryos which were con- spicuously smaller than the remaining ones in the set, were considered to be resorbed. Weight of embryos in the set was estimated according to Zejda (1968). Net body weight of pregnant females (i. e., without weight of embryos) was used in further analysis. The term “mature” females is used for specimens which already have taken part in the reproduction, being either actually pregnant, lactant, or with placental scars. Females not participated in reproduction as yet, were considered to be “immature”. Testis (length and width) were measured with the precision to the nearest 0.1 mm, and length of vesicular glands (from the point of their fusion to their outmost margin) to the nearest millimetre. Area of cross-section of the testes (AC-ST) was used as a criterion of sexual activity. It was computed using formula: AC-ST [mm?] = testis length [mm] * testis width [mm] * 3.1415 * 1/4. We considered males showing values of AC-ST > 25 mm? to be sexually active. This arbitrary criterion was determined on the basis of relationships between AC-ST and the length of vesicular glands in specimens collected in May. In spring and summer there is a full correspondence between the terms sexually mature and sexually active males. However, in autumn and winter samples mature males affected by testes regression are sexually inactive. Reproduction in Mus macedonicus in the Balkans 285 Results Sexual maturation in relation to body weight Intensity of sexual maturation process was expressed as increase in the proportion of mature specimens within subsequent weight categories in the material collected during the period of intensive reproduction (i. e., May to August period in our material). Lightest mature males weight 12 g. Four of nine specimens (44 %) belonging to weight category 11—12 g were mature (Table 1). In the total material of the following three weight categories (13—18 g) only a small proportion of males (7 %, n = 55) remained immature. No immature specimens were found among males exceeding the welght of 18 g (n = 25). Maturation process in females seems to start earlier than that in males (Table 1). The lightest sexually mature female had 10 g. For subsequent weight categories an intensive maturation was typical. As many as 65 % of 20 females between 11 and 16 g were mature. Within weight categories above 16 g, proportion of mature specimens attained the maximal level, with only 8 % (n = 60) remaining immature. However, it is most probable that the real percentage of mature females in our material was slightly underestimated due to following reasons. Among immatures were included also primiparous females during the first five days of their pregnancy, when it is not possible to simply recognize gravidity by means of the macroscopical examination of the uterus (cf. Pelikán 1974). Also, mature females which had participated in reproduction several months before, but had remained sexually inactive in the last months could have been exceptionally scored as immature. Proportion of sexually active males The proportion of sexually active males gradually decreased during the course of the period May to December (Fig.1). The highest percentages of sexually active males were found in May (87 %, n = 37) and June (91 %, n = 11), i.e, in the period in which overwintered, sexually active males highly predominated over males born in Table 1: Proportion of sexually mature males and females within the individual weight categories during the period of intensive reproduction (May to August). Males (n = 100) Females (n = 90) mature mature Body weight | — UW CON WOO hOCO SC 286 V. Vohralik et al. n= 29 n=27 Y PG 2 ; Dec Li Sent Oct ZA Active LJ inactive | < Fig. 1: Variation in the proportion of sexually active males in Mus macedonicus during the period May to December. the current year. As a result of the continual appearance of young, immature individuals, the proportion of sexually active males decreased in July and August to 66 Yo (n = 55) and 73 % (n = 11), respectively. In the following months the decrease continued owing to mass occurrence of young animals, decreased intensity of maturation, and testes regression in mature males. Corresponding values in Septem- ber, October and November were 25 % (n = 56), 9% (n= 55) and 7% (n= 29), respectively. All males captured in December (n = 27) were already inactive. Proportion of sexually mature females Proportion of mature females (Fig. 2) was highest in May 95 % (n = 21). Distinct decrease in the following months was a result of mass occurrence of young animals during the summer (June 64 %, n= 11, July 70%, n= 57, August 75 %, n= 8). Autumnal drop (September 42 %, n = 62, October 40 %, n = 38, November 22 %, n = 23, December 21 %, n = 24) can be attributed in addition to the above factor also to the selective mortality of mature animals and the termination of the maturation process at the end of the breeding season. Duration of breeding season Due to the lack of the material from the January to April period, the only informa- tion available on the beginning of the breeding season can be based on the age estimation in current year specimens collected in May. The best for this purpose seems to be a sample of 48 animals collected during the period 3—11 May 1994 in three localities of Macedonia, Greece. The young, in the current year born individuals were identified among them according to their lower body weight, small body dimensions, fur coloration, and in males also according to considerably smaller length of vesicular glands. Altogether, we found five current year specimens in this sample. The biggest among them were a male (body weight 16 g) and a female (14 g) which was already pregnant (length of embryos 3 mm). According to our knowledge on postnatal growth in laboratory born Mus macedonicus (D. F., un- published data), the birth date of these two specimens can be estimated approxi- Reproduction in Mus macedonicus in the Balkans 287 n=23 n=24 - TZ ZA = Pregnant ZA Other mature L__] immature May June July Aug Sept Oct Fig. 2: Variation in the proportion of mature and pregnant females in Mus macedonicus during the period May to December. mately to the beginning of March and consequently, their conception to mid February. The end of the breeding season was put to October when last pregnant females were collected (Fig.2). They composed small proportion only among mature (20 Yo, n = 15) as well as among all captured females (8 Yo, n = 38). Last pregnant females were from the 10th, 18th and 20th October, and their embryos measured 11, 15 and 16 mm, respectively. These data suggest that parturitions take place till the end of October. Therefore, the duration of the breeding season could be estimated at about 81/2 months (i. e., mid February — late October). However, we ought to keep in mind that the length of the breeding period can vary between years and that our estimation of its beginning is based on data obtained in a single year only. Breeding intensity The reproduction in Balkan populations of Mus macedonicus has distinctly seasonal character. The breeding intensity during the year was expressed as proportion of pregnant among all the females in the sample. As evident from Fig.2, this proportion was fairly high during the spring—summer period (May 38 %, n = 21; June 27 Yo, n= 11; July 35 %, n = 57; August 50 %, n = 8). Marked drop in breeding intensity was recorded in September (13 %, n = 62) and October (8%, n= 38). Pregnant females were not present in November (n = 23) and December (n = 24) samples. Breeding intensity can also be given as percentage of pregnant among mature females. This percentage was also high throughout the entire spring— summer period (May 40 %, n = 20; June 43 %, n = 7; July 50 %, n = 40; August 67 Yo, n = 6). Its rapid decrease was observed from September (31 %, n= 26) till the end of the breeding season (October 20 Yo, n= 15). Litter size In 45 visibly pregnant females, the litter size (i.e., all visible embryos in the set, including the resorbed ones) varied within the range 4—10, x = 6.73 embryos (S. E. = 0.23). The value represents the annual mean of all females, irrespective of 288 V. Vohralik et al. Table 2: Variation in litter size during the period May to October including embryos affected by resorption (mean indicated by asteric was computed for unresorbed embryos only). Month 10 Mean Mean* May June July August September October 6.73 6.49 (0.23) (0.26) NS Stk od Table 3: Variation in resorption rate during the period May to October. Explanations: SLA = size of litters affected (numbers of resorbed embryos are given in parentheses). Percents were computed for two-months periods. sets affected % embryos affected May 7(1) June July 10(2) 5(2) August September 6(2) October 7(1) 6(3) date and the locality of capture. If only unresorbed embryos were counted, the corresponding mean litter size was 6.49 (S. E. = 0.26). Variation in litter size during the year is given in Table 2. Despite very low sample sizes, one-way analysis of variance revealed significant variation of litter size between the two months periods (F = 4.505; P = 0.0169, non-resorbed embryos only). Tukey test confirmed that the mean litter size in May to June (7.73, n = 11) is significantly higher than the corresponding values both in July to August (6.17, n = 24) and in September to October (5.90, n = 10). It is to be remarked here, that the mean net body weight of pregnant females was almost identical in all the above periods (F = 0.198; P = 0.8213) and the effect of season on litter size even increased when the body weight was introduced as covariate into ANOVA model (F = 4.758; P = 0.0149). In the total of 45 visibly pregnant females, 6 embryo sets (13.3 %) were affected by resorption. In two cases, only one embryo was affected, in three cases, two embryos. In one case, even three of six embryos in the set were resorbed. In total, only 11 of 303 embryos examined were affected (3.63 %). Resorption rate showed slightly higher values at the end of the breeding season (Table 3). Reproduction in Mus macedonicus in the Balkans 289 Discussion There is a general agreement between our findings in Mus macedonicus and obser- vations reported by most authors studying free living mice populations (Mus spicile- gus: southern Ukraine, Naumov 1940, Vojvodina, Mikes 1971; Mus sp.: Kazakhstan, Borisenko 1977; Mus musculus: field populations in Czechoslovakia, Pelikán 1974, 1981) that the majority of individuals of both sexes attain their sexual maturity in the size ranging between 10 and 14 grams. Also our estimation concerning the duration of the breeding season in Mus macedonicus (mid February to late October) corresponds fairly with data available on other free living mice species. In various populations of Mus spicilegus in southern Ukraine the following breeding periods were found: mid March to October (Sokolov et al. 1990), early February to October (Pisareva 1948) and mid April to October (Naumov 1940). For the same species from Vojvodina Mikes (1971) reported the period March to November. Similarly, a breeding season lasting from February to October was reported in M. spretus from southern Spain (Vargas et al. 1991). In general, we can conclude that a seasonal reproduction period with an interruption during winter months appears to be a common phenomenon in all free living mice populations of the western Palaearctic, studied so far. Comparing the proportion of pregnant among all the females collected during the period of most intensive reproduction (May to August), it is evident that the value found in our material of Mus macedonicus (36.1 Yo, n = 97) is closest to those in Mus spicilegus from Vojvodina (36.7 Yo, n = 120, Mikes 1971), while the value reported in Mus spicilegus from southern Ukraine (50.0 %, n = 204, Sokolov et al. 1990) is considerably higher. Mean litter size in our total material of Mus macedonicus (6.73) is practically the same as the value computed from the data published by Markov (1964), who had at his disposal a sample collected during 1960—1962 in southern Bulgaria (mean = 6.75, range 3—9, n = 16). Fairly close values were reported in populations of Mus spicilegus from S Ukraine (6.7, n = 189, Sokolov et al. 1990; 6.9, Naumov 1940), as well as in free living mice captured in Bulgaria (6.53, range 3—10, n = 132, Straka 1966). Slightly higher values were reported in Mus spicilegus from Vojvodina (7.30, range 2—15, n= 56, Mikes 1971) and in field populations of Mus musculus from Czechoslovakia (7.85, range 4—12, n = 67, Pelikán 1974). Similar span of the average litter size was found in free living mice populations collected in different parts of Kazakhstan. Corresponding values in the populations from districts Aktjubinsk and Ural (Borisenko 1977), Kustanaj and Turgajsk (Borisenko 1964) and Alma-Ata (Machmutov 1970) were 6.8 (n = 17, range 5—10), 7.7 (n = 68, range 1—14) and 7.0 (n = 65, range 5—11), respectively. On the other hand, mean values of net litter size (1.e., without resorbed embryos) reported in Mus spretus, 5.0 (Durán et al. 1987) and -5.53 (n = 193, range 2— 10; Vargas et al. 1991) are distinctly smaller than average net litter size 6.49 found in our material of Mus macedonicus. Comparing above differences we ought to keep in mind the general tendency in rodents to decrease their litter size in east-west, conceivably also in north-south directions in Europe. For example, in the Wood mouse, Apodemus sylvaticus the differences in mean litter size between populations from the Iberian Peninsula and the Balkans (Frynta & Vohralik 290 V. Vohralík et al. 1992) are nearly the same as those given above in free living mice. However, it has to be mentioned here that Pelikán (1974) found similar differences also between field (7.85, n= 67) and commensal populations (5.58, n= 183) of Mus musculus in Czechoslovakia, which fact suggests that not only latitude and longitude, but also habitat can considerably affect the litter size in murine rodents. It is of interest that the May—June peak in mean litter size found in our material of Mus macedonicus has a parallel in an April —May peak reported by Vargas et al. (1991) in Mus spretus. They attributed this phenomenon to the higher age of over- wintered females prevailing in the population during this period. On the contrary, highest mean litter size values were found during the June—August period in Mus spicilegus of southern Ukraine (Sokolov et al. 1990). Although the percentage of resorbed embryos in our material of Mus macedonicus (3.63 Yo, n = 303) is twice as high as corresponding values 1.57 % (n = 1068) reported by Vargas et al. (1991) in Mus spretus or 1.87 Y (n = 536) reported by Pelikán (1974) in field populations of Mus musculus in Czechoslovakia, the sample size in our material is too small and seasonal variation too high for any further conclusion. Basing on above comparisons we can conclude that, in general, the reproduction pattern in Mus macedonicus resembles that reported in other free living mice species of the western Palaearctic. Acknowledgements We are grateful to our colleagues Dr. J. Flousek (Vrchlabí), Dr. K. Weidinger (Olomouc) and Dr. D. Král (Prague) who provided us with some mice which they collected during their zoological expeditions to the Balkans. We thank Dr. B. Kry3tufek (Ljubljana) for valuable — comments on the earlier version of the manuscript. Zusammenfassung Die Fortpflanzung von Mus macedonicus wurde an 538 Exemplaren dieser Art von 64 Stand- orten in Nordgriechenland, Sidbulgarien und Mazedonien untersucht. Gegenstánde der Untersuchung waren: Lange der Fortpflanzungszeit, Geschlechtsreife im Bezug zum Kórper- gewicht, Anteil geschlechtsreifer Individuen in der Population, Fortpflanzungsintensitat, Wurfgröße, Embryonalresorption. Die Fortpflanzung von Mus macedonicus trägt ausgepräg- ten Saisoncharakter, die Wurfgröße beträgt 4-10, xX= 6.73 (n=45). Die gewonnenen Erkenntnisse wurden mit Literaturangaben über andere freilebende Mäuse der Gattung Mus in Europa und Mittelasien verglichen. Literature Auffray, J-C., J. T. Marshall, L. Thaler & F Bonhomme (1990a): Focus on the nomenclature of European species of Mus. — Mouse Genome 88: 7—8. Auffray, J-C., E. Tchernov, F Bonhomme, G. Heth, S. Simson &E. Nevo (1990b): Presence and ecological distribution of “Mus spretoides” and Mus musculus domesticus in Israel. Circum-Mediterranean vicariance in the genus Mus. — Z. Säugetierk. 55: 1—10. Bonhomme, E, J. Catalan, J. Britton-Davidian, V. M. Chapman, K. Mori- waki, E. Nevo&L. Thaler (1984): Biochemical diversity and evolution in the genus Mus. — Biochem. Genet. 22: 275—303. Borisenko, V. A. (1964): Domovaja my3 Kustanajskoj oblasti v svjazi s raspaskoj celinnych zemel’ [House mouse in Kustanaj region in relation with ploughing of celina grounds]. — Izvestija AN KazSSR, Ser. Biol. 1964(5) (ex Sokolov et al. 1990, in Russian). Reproduction in Mus macedonicus in the Balkans 291 Borisenko, V. A. (1977): [House mouse — Mus musculus Linnaeus (1758)], pp: 396—414. — In: Sludskiy A. A., Bekenov A., Borisenko V. A., Grachev Yu. A., Ismagilov M. I., Kapitonov V. I., Strautman E. I., Fedosenko A. K. & Shubin I. G.: Mlekopitajuscie Kazachstana [Mammals of Kazakhstan]. Vol. I., Part 2. — Alma-Ata: Nauka, 634 pp. (in Russian). Cassaing, J. & H. Croset (1985): Organisation spatiale, competition et dynamique des populations sauvages de souris (Mus spretus Lataste et Mus musculus Rutty). — Z. Sáuge- tierk. 50: 271—284. Duran, A.C. & V. Sans-Coma (1986): Geschlechtsreife bei Mus spretus Lataste, 1883. — Z. Saugetierk. 51: 345—349. Durán, A.C.,F.J. Camprodón, M. Cardo & V. Sans-Coma (1987): La reproducción de Mus spretus Lataste, 1883 en el N-NE de la Península Ibérica. Pp: 111—120. — In: Mamíferos y Helmintos. Un volumen homenaje al Prof. Dr. Hermann Kahmann en su 81 aniversario. Barcelona: Ed. Ketres (ex Vargas et al. 1991). Frynta, D., A. Exnerová, J. Ciháková, P. Kozlová & L. Zelinková (1992): Species dicrimination based on olfaction in males of two related mice species: Mus abbotti and Mus musculus. Pp: 49—56. — In: Horácek I., V. Vohralík (eds.): Prague Studies in Mammalogy. Praha: Karolinum-Charles University Press, 245 pp. Frynta,D.&V. Vohralík (1992): Reproduction in the Wood mouse (Apodemus sylvaticus) in urban habitats of Prague: II. Litter size. — Acta Univ. Carolinae — Biol. 36: 325—340. Gerasimov, S., H. Nikolov, V. Mihailova, J.C. Auffray & F. Bonhomme (1990): Morphometric stepwise discriminant analysis of the five genetically determined European taxa of the genus Mus. — Biol. J. Linn. Soc. 41: 47—64. Kratochvil, J. (1986): Mus abbotti — eine kleinasiatisch-balkanische Art (Muridae-Mam- malia). — Folia Zool. 35: 3—20. Machmutov, S. M. (1970): Sinantropnyje gryzuny goroda Alma-Aty [Synanthrope rodents in the city of Alma-Ata]. — In: Materialy nauénoj konferencii molodych biologov goroda Alma-Aty. Alma-Ata (ex Sokolov et al. 1990, in Russian). Markov, G. (1964): Insektenfressende Sáugetiere und Nagetiere in Thrakien (Súdbulgarien), Pp: 19—53. — In: Fauna na Trakija. Cast 1. (Die Fauna Trakiens. Bd. 1.). — Sofia: Izd. BAN (in Bulgarian, with a summary in German). Marshall, J. T. (1986): Systematics of the genus Mus. — Current Topics in Microbiology and Immunology 127: 12—18. Mikes, M. (1971): Ecologic investigations on Mus musculus hortulanus Nordmann in Vojvo- dina. — Zbornik za prirodne nauke Matice srbske 40: 52—129 (in Serbian, with a sum- mary in English). Musser, G. G. & M. D. Carleton (1993): Family Muridae. Pp: 501—755. — In: Wilson D. F. & D. M. Reeder (eds.): Mammal species of the world: A taxonomic and geographic reference. 224 ed. Washington & London: Smithsonian Inst. Press, 1206 pp. Naumov, N. P. (1940): Ekologija kurgancikovoj mySi Mus musculus hortulanus Nordm. [The ecology of the Mound building mouse, Mus musculus hortulanus Nordm.]. — Trudy Instituta Evoljucionnoj Morfologii AN SSSR 3(1): 33—76 (ex Sokolov et al. 1990, in Russian). Orlov, V. N., R. S. Nadzafova &N. S. Bulatova (1992): Taxonomic isolation of Mus abbotti (Muridae, Rodentia) from Azerbayjan. — Zool. Zurn. 71: 116—121 (in Russian, with a summary in English). Orsini, P, F Bonhomme, J. Britton-Davidian, H. Croset, S. Gerasimov & L. Thaler (1983): Le complexe d’espéces du genre Mus en Europe Centrale et Orientale. II. Criteres d’identification, réparation et charactéristiques écologiques. — Z. Säugetierk. 48: 86—95. Pelikan, J. (1974): On the reproduction of Mus musculus L. in Czechoslovakia. — Acta Sci. Nat. Brno 8 (12): 1-42. Pelikán, J. (1981): Patterns of reproduction in the House mouse. — Symp. Zool. Soc. Lond. 240205229. 292 V. Vohralik et al. Pisareva, M. E. (1948): K ekologii 1 sistematike kurgancikovoj mySi. [On the ecology and systematics of the Mound building mouse]. — Sbornik Rabot Biologiceskogo Fakulteta DGU, Naué. Zap. 32: 68—71 (ex Sokolov et al. 1990, in Russian). Sofianidou, T. S. & V. Vohralik (1991): Notes on the distribution of small mammals (Insectivora, Rodentia) in Epeirus, Greece. — Bonn. zool. Beitr. 42: 125—135. Sokolov, V. E., E. V. Kotenkova & S. I. Ljaljuchina (1990): Biologija domovoj i kurgancikovoj mySej. [The biology of the House mouse and the Mound-building mouse]. — Moskva: Nauka, 207 pp. (in Russian). Straka, F. (1966): Untersuchungen über die Biologie der Hausmaus (Mus musculus L.) unter Freilandbedingungen. — Rast. Nauki., Sofia 3 (7): 87—95 (in Bulgarian with a summary in German). Vargas, J.M., M. Espana, R. Haro & V. Sans-Coma (1984): Estructura poblacional de Mus spretus Lataste, 1883 en cultivos de caña de azúcar del sur de la Península Ibérica. — Misc. Zool. 8: 253—262. Vargas, J. M., L.J. Palomo, M. Espafia, A. C. Duran & V. Sans-Coma (1986): Über die Geschlechtstátigkeit der 9 9 und die Populationsstruktur von Mus spretus Lataste, 1983 in Südspanien. — Z. angew. Zool. 73: 219— 229. Vargas, J. M., L. J. Palomo «€ P. Palmquist P. (1991): Reproduction of the Algerian mouse (Mus spretus Lataste, 1883) in the south of the Iberian Peninsula. — Bonn. zool. Beitr 4222 110! Vohralik, V.& T.S.Sofianidou (1987): Small mammals (Insectivora, Rodentia) of Mace- donia, Greece. — Acta Univ. Carolinae — Biol. 1985: 319 — 354. Vohralik, V. & T. S. Sofianidou (1992): Small mammals (Insectivora, Rodentia) of Thrace, Greece. — Acta Univ. Carolinae — Biol. 36: 341—369. Zejda, J. (1968): A study on embryos and newborns of Clethrionomys glareolus Schreb. — LOMAS 17.215226: Dr. Vladimir Vohralik & Dr. Daniel Frynta, Department of Zoology, Charles Uni- - versity, Vinicna 7, CZ-128 44 Praha 2, Czech Republic. — Prof. Dr. Theodora S. Sofianidou, Department of Zoology, Aristotelian University, Thessaloniki — TK 54 006, Greece. Bd. 47 Ss. 2935299 Bonn, September 1998 Two new chameleons (Sauria: Chamaeleonidae) from isolated Afromontane forests in Sudan and Ethiopia Colin Tilbury Abstract. Two new species of horned chameleon are described from isolated montane forests in eastern Africa. In one of the species, the male has a unique rostral cone unlike any other known chamaeleonid rostral appendage, and in the other, paired annulated rostral horns are present, a feature previously regarded as unique to a group of west African chameleons. Key words. Reptilia, Chamaeleonidae, Sudan, Ethiopia, new species, taxonomy, zoogeo- graphy. Introduction The forests of the Afromontane Archipelago serve as centres of endemism for plants and animals in Africa. It comes as no surprise that isolated and consequently little explored areas of montane forest should still produce new species of reptile. In 1988 the chameleons in the spirit collection of the National Museum of Kenya (Nairobi) were examined. Amongst the specimens a single adult male chameleon was found that originated from the Imatong Mountains in southern Sudan. Its uniqueness was evident by its long non-annulated conical rostro-nasal projection. At the time it was hoped that further specimens of this form could be obtained to allow a more detailed analysis, but due to the civil unrest prevalent in that region, it seems unlikely that further specimens of this chameleon will be collected in the near future and therefore this species Chamaeleo (Trioceros) conirostratum n. sp. is described on the basis of this single adult male holotype. Malcolm Largen (1995) described a new species of chameleon Chamaeleo harennae from the Harenna Forest in Ethiopia on the basis of two specimens — an adult female and an unsexed juvenile LIV 1986.212.245. The latter specimen had been found dead, desicated, severely damaged — presumably by a bird — and not in good condition. However it showed some significant differences from the adult specimen in having more heterogeneous scalation and was also noted to possess two tiny rostral tubercles. Largen felt that the observed differences were likely to be a conse- quence of its immaturity. Since males of a species may possess cranial ornamentation not exhibited by females, the possibility that the juvenile was a male and the adult specimen being a female, would allow that the two could still be conspecific. How- ever On examining the photograph of the adult specimen, well developed hemipenal bulges are evident, suggesting that the hornless holotype is in fact an adult male. Since this would effectively negate the possibility that the juvenile was a horned female, the probability was high that the two specimens in fact represented two distinct taxa. During the course of 1996, a trip was made to the Harenna Forest to test this hypothesis, that resulted in the collection of a good type series of a new 294 ©, Tilbury chameleon Chamaeleo (Trioceros) balebicornutus n. sp. and 4 additional adult speci- mens of Ch. harennae (one male and three females). Chamaeleo (Trioceros) balebicornutus n. sp. (Fig. 1) Diagnosis: A small chameleon with markedly heterogeneous scalation. Each canthal ridge terminates in an annulated horn that is closely aligned to the opposite horn and projects forward over the upper lip. The horns are well developed in the males, and attenuated in females. The casque is flat and has a distinct shallow notch at the occiput. There are no obvious parietal or temporal crests. Gular crest is strong and the dorsal crest is limited to the anterior 2/3 of the keel. Ventral crest is repre- sented by a line of white tubercles. The tail is smooth. Derivatio nominis: The name derives from the fact that this is the only two- horned chameleon inhabiting the Bale Mountains. Holotype: ZFMK 63049, adult male, collected 2nd October 1996 by Colin Tilbury and Stephen Spawls, Katcha clearing, altitude 2400 metres, Harenna Forest, Bale Mountains, central Ethiopia 06 °42’N 39°44’E. Paratypes ZFMK 63050-58 collected by Tilbury and Spawls, 21st September and 2nd October 1996 from roadside vegetation between 1700 m and 2100 m 06°37’N 3944”E and LIV 1986.212.245 from 1980m 06°40’N 39°44’E collected by M. J. Largen on 7th August 1986. Table 1: Morphometric variation in the holotype and paratypes of Chamaeleo (Trioceros) | balebicornutus n. sp. — * Measurements in millimetres. s/v = snout/vent length; c/snt = casque tip to snout length; 1.0.d. = inter-orbital diameter; com/snt = commissure of mouth to snout distance; c/com = casque to commissure distance; g/cr = number of cones in the gular crest; d/cr = number of cones and pointed tubercles in the dorsal crest. — Mus No. = Museum accession number in the ZFMK Bonn, Germany. — MM = adult male, FF = adult female, jm = juvenile male, jf = juvenile female. Holotype cos spa a [es 13 1125 Two new chameleons in Sudan and Ethiopia 295 Fig. 1: Adult male of Chamaeleo (Trioceros) balebicornutus n. sp. from the Harenna forest, Ethiopia. Fig. 2: Sulcal view of the hemipenis of the holotype of Chamaeleo (Trioceros) balebicornutus n. Sp. Description of the holotype: A small chameleon with a total length (TL) of 162 mm. Casque flat. No parietal crest. The lateral crests terminate posteriorly in two large pointed tubercles on each side of the casque giving it a swept up appearance and causing the posterior edge of the casque to be shallowly notched in the midline. There is no temporal crest. The supra-orbital crest has a dentate appearence. The canthal ridges are smooth and each side terminates in an annulated horn which projects forward for 4 mm at an angle of 10° from the horizontal. The 296 ECHT bey two horns are tightly opposed to each other. The nare is sited midway between the anterior orbital rim and the tip of the snout, opens inferiorly and has no indication of a nasal bulge. Gular crest composed of 18 elongated white conical tubercles, largest anteriorly and becoming progressively smaller posteriorly. The gular area is incised by 4 shallow grooves that are most noticeable posteriorly and enclose a few islands of enlarged oval tubercles. Dorsal crest composed of 8 large cones, largest anteriorly and followed by a rapidly diminishing series of pointed tubercles to fade completely by about ?/3 of the way along the keel. There is no crest on the tail. The gular crest is followed posteriorly by a midline ventral row of white tubercles that extends a short way beyond the vent. Scalation heterogeneous with large flattened platelike tubercles 2—3 times the diameter of the surrounding tubercles, profusely scattered over the flanks forming several vague rows. Tail and belly covered with finer sub-homogeneous granules. Eyeball skin coated with fine homogeneous granules. The tail is longer than the snout/vent length, comprising 53 % of the TL. Hemipenis: (Fig. 2). Club-shaped. Truncus is set with shallow simple calyces, the most prominent being on the proximal aspect of the asulcal side. The sulcal lips are almost smooth. The apex is adorned with 2 pairs of similarly sized rotulae with finely dentate free margins. The asulcal pair are orientated almost vertically while the sulcal pair are orientated horizontally towards the sulcal side. Between each of the sulcal and asulcal rotulae is a row of 2—3 large thornlike papillae curved toward the asulcal side. Colour in life: Holotype — Background a light moss green. Flank tubercles divided into many vague rows by dark interstitial skin. A dark patch is present over . the dorsal keel at the nape and pelvis and between these, two broad hourglass shaped dark blotches extend from the keel to the belly overlain by green tubercles. Top of head and temporal area pale off-white. Lower labials pale blue. Limbs blotched with dark brown, as is the tail. Belly off-white to green. Eyeballs with 8 radiating lines. Gular crest white. Anterior dorsal crest spines a rich brown chocolate colour. Female colour: Background pale to bright green. Dorsal crest spines dark orange as are the lateral/supra-orbital crests and the tip of the snout. Gular crest white. No radiating lines on the eyeball. Belly crest indicated only by a faint white line. Variation: (Table I) Little variation of any significance is seen amongst the chameleons. The gular crest varies between 18—23 cones and the number of cones and pointed tubercles in the dorsal crest varies between 14 to 25. Sexual dimorphism is seen in the development of the annulated horns which are much smaller in the female, and in the smaller average number and reduced prominence of the gular cones. The rostral projections are clearly seen in even the smallest specimens. Reproduction: ZFMK 63056 was dissected and found to contain 16 eggs (6 in the left oviduct and 10 in the right). These averaged 8 mm x 6 mm in size. Habitat: Broadleafed undifferentiated Afromontane rain forest between 1500 m — 2400 m. The Harenna forest cloaks the southern slopes of the Bale Mountains. The tops of the tallest trees reach up to about 40—45 metres with the dominant species including Aningeria adolfi-friedericii and Podocarpus latifolius in the lower elevations and Hagenia abyssinica and Schefflera abyssinica at the higher reaches of Two new chameleons in Sudan and Ethiopia 297 this chameleon’s range. The chameleons are found in the lower shrub layer of the forest at perch heights varying between 1.5 metres and 5 metres above the ground. The holotype was collected at a height of approximately 5 metres in a tall specimen of Hypericum revolutum near a rivulet at the edge of a large grassy glade (Katcha clearing). Notes: The Bale mountains form an important centre of endemism in Ethiopia. In addition to Ch. (T.) balebicornutus n. sp., two other species of chameleon are known from these mountains. The recently described Ch. harennae occurs chiefly in the upper reaches of the Harenna forest between 2400 metres and the upper forest line at about 3300 metres. Here it extends into the giant heathers and ericaceous shrubs. The two species are sympatric in the zone around 2400 metres where they both share the same type locality (Katcha clearing). Ch. (T.) balebicornutus n. sp. differs chiefly from Ch. harennae in the presence of paired rostral horns in both sexes, a more heterogeneous scalation and in the complete absence of a parietal crest. The tail is also smooth. The third species Chamaeleo (T.) affinis was not found in the Harenna forest but rather on the northern slopes of the mountains at elevations up to at least 2700 metres. This species has a much finer scalation with no trace of a rostral projection, and either no gular crest or a pair of low gular ridges. Taxonomic affinities: The arrangement of the rostral horns in Ch. (T7.) balebi- cornutus n. sp. is unusual in that all other species with paired annulated horns are strictly west African forms. This, coupled with the absent parietal and temporal crests (also characteristics of the west African Ch. cristatus group (Klaver & Böhme 1992) suggests a west African splinter radiation. This is not however borne out by the affinities of the other herpetofauna of the Ethiopian highlands, which show nothing in common with either the west or central African forests, and are largely unique with a high rate of endemism. In Tanzania, Ch. (T!) deremensis shares some features with balebicornutus n. sp. in that it also lacks a parietal and temporal crest, but has no gular crest and 3 annulated horns. Chamaeleo (Trioceros) conirostratum n. sp. (Fig. 3) Diagnosis: The adult male chameleon is recognised by a unique rostral process comprising of an elongated non-annulated cone 3 mm in length which appears to be attached to a pliable base. The tail is slightly shorter than the snout/vent length. The general habitus, shape of the casque and stucture of the parietal crest ally this taxon to the Ch. bitaeniatus (sensu lato) group. Derivatio nominis: The name conirostratum derives from the unique shape and position of the rostral process. Holotype: NMK L/1949, an adult male collected by Parker and Stubbs in 1982 from Lomoriti, at 3500 feet a.s.l., south west Imatong Mountains, southern Sudan 03°54’N 32°43’E. l Description: A small chameleon with a total length of 130 mm. Casque narrowly elevated and peaked posteriorly. Parietal crest distinct, forks anteriorly with a few small tubercles continuing in a midline row between the orbits. The temporal crests 298 CATE Fig. 3: Detail of the head of the adult male holotype of Chamaeleo (Trioceros) conirostratum n. sp. from Lomoriti, Imatong mountains, southern Sudan. are distinct, comprising 4 enlarged tubercles on each side. The canthal ridges are edged in enlarged, but smooth oval scales (5 on the left and 6 on the right). The canthal ridges merge anteriorly to end in the base of an elongated non-annulated thin cone projecting 3 mm off the front of the upper lip. The base of the cone appears to be pliable (In the preserved specimen this appendage is bent downwards and to the left side over the tip of the snout — possibly an artefactual state caused at the time of preservation). There are no traces of any pre-orbital projections. The nares open facing directly posteriorly, from a nasal bulge located 3/8 of the distance from the anterior orbital rim to the tip of the snout. The gular crest is composed of a continuous series of 24 low cones between the mentum and the angle of the jaw, and continues posteriorly as an even lower series of enlarged midline belly tubercles to the vent. The skin on the sides of the throat are incised by shallow grooves (6 on the left and 5 on the right). The dorsal crest begins at the nape with 3 isolated small cones and is then followed by somewhat triangular laterally flattened tubercles that enlarge successively in groups of 2 to 3 along the dorsal ridge. The crest fades completely by the sacrum and then continues as a low series of angulated tubercles to again fade by the mid tail. Background scalation heterogeneous but not particularly coarse. A single row of slightly enlarged tubercles extends from the shoulder region over the upper third of the flank towards the pelvis. Belly covered with finer almost homogeneous flattened tubercles. The skin of the circular eyeball is covered in roughly homo- geneous granules with a few slightly enlarged rounded tubercles seen scattered near the eye opening. The hemipenes are not extruded. Dimensions: Snout/vent length 67 mm, Tail length 63 mm, Casque tip/snout 21 mm, Casque tip/commissure 13 mm, Commisure/snout 14 mm, Rostral process 3 mm. Two new chameleons in Sudan and Ethiopia 299 Notes: The female of this species is currently undescribed. Taxonomic affinities: Ch. (Trioceros) conirostratum is closely allied to the other chameleons of the “Ch. bitaeniatus group” (Rand 1963) as evidenced by the overall similarity to Ch. (7.) bitaeniatus Fischer, 1884 in body proportion and the shape and structure of its head and body crests. The unusual structure of its rostral process does not exclude it from the subgenus Trioceros where annulated horns are regarded as a synapomorphous character. The structure of its rostral projection does not strictly qualify as a horn and its position and orientation in the living chameleon awaits clarification. Other species of chameleon that are known to occur in the Imatong Mountains include Ch. (Trioceros) bitaeniatus, Ch. (T.) kinetensis Schmidt, 1943 and Ch. (T.) ellioti Günther, 1895 (Böhme & Klaver 1980). None of these species possess horns or rostral projections. Ch. (Trioceros) marsabitensis Tilbury, 1991 from the Marsabit volcano to the south east of the Imatongs, possesses a short single annulated rostral horn, and Ch. (Trioceros) hoehnelii Steindachner, 1891 differs in that its rostral projection is more of a clump of tubercles than a horn. Acknowledgements The author would like to thank Steven and Laura Spawls for providing the opportunity, comradeship and support, both logistical and personal as well as many chocolates during some harrowing yet wonderful field trips to southern Ethiopia. Thanks also to Malcolm Largen and Wolfgang Böhme of the Liverpool Museum and ZFMK Bonn respectively for sharing information and insights into chameleon taxonomy. Also to Damaris Rotich, Anton Espira and Alex Duff Mackay of the Department of Herpetology at the National Museum of Kenya (Nairobi) for allowing me access to the spirit collection of the Museum. Zusammenfassung Es werden zwei neue Arten gehórnter Chamáleons aus isolierten Bergwáldern Ostafrikas beschrieben. Bei einer Art besitzt das Mánnchen einen markanten rostralen Zapfen, der keinem der von Chamáleons bekannten Anhänge ähnlich ist. Die andere Art zeichnet sich durch paarige geringelte Rostralhörner aus, ein Merkmal, welches bisher als charakteristisch für eine Gruppe westafrikanischer Chamäleons angesehen wurde. Literature Böhme, W. & C. Klaver (1980): The systematic status of Chamaeleo kinetensis Schmidt, 1943, from the Imatong mountains, Sudan, with comments on lung and hemipenal morphology within the Chamaeleo bitaeniatus group. — Amphibia-Reptilia 1: 3—17. Klaver, C. & W. Böhme (199): The species of the Chamaeleo cristatus group from Cameroon and adjacent countries, West Africa. — Bonn. zool. Beitr. 43: 433 — 476. Largen, M. J. (1995): A new species of chameleon (Reptilia Sauria Chamaeleonidae) from montane forest in Ethiopia. — Tropical Zoology 8: 333 —339. Rand, A. S. (1963): Notes on the Chamaeleo bitaeniatus complex. — Bull. Mus. Comp. Zool. 13021729. Dr. C. R. Tilbury, Jwaneng Mine Hospital, Private Bag 08, Jwaneng, Botswana. Bd. 4 s. 301-311 | Bonn, September 1998 The community of rock-dwelling cichlids in Lake Victoria Ole Seehausen & Niels Bouton Abstract. New data from southern Lake Victoria show that a large group of probably more than 200 species of stenotopic rock-dwelling cichlids remained unrecognized until recently, thus disproving prior assumptions about the eurytopic character of Lake Victoria cichlids. We discuss the ecology of rock-dwelling cichlids in Lake Victoria based on the community at an isolated rocky island in the Speke Gulf. With emphasis on micro- distribution and feeding ecology we present data on the faunistic and ecological composi- tion of a typical community. Communities are rich in species and ecologically complex. This may make them particularly vulnerable to changes in the environment. Key words. Cichlidae, Lake Victoria, community ecology. Introduction Species rich groups of stenotopic rock-dwelling cichlids made the cichlid species flocks of Lakes Malawi and Tanganyika world famous among fish ecologists, evo- lutionary biologists and aquarists (e.g. Fryer & Iles 1972, Lowe McConnell 1993). It is still widely believed that Lake Victoria cichlids are much less stenotopic and the coexistence of several hundred apparently eurytopic species was considered a violation of the competitive exclusion principle (Greenwood 1981). Though the existence of rock-dwelling cichlids in Lake Victoria was known (van Oijen et al. 1981), it was believed that there are only a handful of such stenotopic species. New research in southern Lake Victoria shows that a huge group of probably more than 200 species remained basically unrecognized. In many respects closely resembling the Mbuna of Lake Malawi, they prove much of the prior assumptions about the euryto- pic character of Lake Victoria cichlids wrong. Most of them are restricted to rocky substrate, have a limited distribution within the lake and geographic variation is well developed. Unfortunately their communities are under severe pressure due to large scale changes in the environment. Several rock-restricted species have already disap- peared (Witte et al. 1992) and, unlike in open waters, Nile perch predation is probably not the direct cause. An understanding of the complex ecological structure of the communities is a first step towards an understanding of these developments. We give here an introduction to the community ecology of rock-dwelling cichlids in Lake Victoria on the example of the community at an isolated rocky island in the Speke Gulf. We present data on the faunistic and ecological composition of a typical community with emphasis on micro-distribution and feeding ecology. We show the ecological complexity of the community and its species richness, qualities which may make it particularly vulnerable to alterations in the environment. 302 O. Seehausen &N. Bouton Speke Gulf Mwanza Gulf Fig. 1: Lake Victoria and the location of Makobe and Ruti Islands. Materials and methods Over a period of two and a half years (from January 1990 through July 1991, March to May 1992, September through December 1993 and February to April 1995) cichlids were sampled at the 0.075 km? (75 000 m?) small rocky Makobe Island which lies 5 km offshore in the Speke Gulf entrance (Fig. 1). Makobe Island is an example of gently sloping islands with small to medium sized rock boulders and moderate geographical isolation. Sampling was done from the surface down to 6m depth with gillnets of mesh sizes between 1 and 2 inches and by angling in rock pools and rocky crevices. The total sampling effort consisted of over 48 000 m? net hours and over 300 hours angling. Additionally underwater censuses were done by snorkeling. Relative abundances were determined with all three techniques. Micro- habitat distribution of the species was scored by restricting samples to areas of one to a few square meters of rather uniform depth, exposure and rock cover. Stomach contents of some individuals of each species were analyzed individually to io food organisms. The contents of 40 individuals per species (each 10 from four seasons) were then centrifuged and quantified. It is necessary to study not only diet composition and spatial distribution but also how the food is obtained. We performed feeding experiments in the laboratory with the very species whose diet was analyzed. We studied the behaviour of fish feeding in experimental aquaria on Aufwuchs (algae and associated fauna) covered rocks from their natural habitat. After collection from the lake the fish were kept in aquaria without food for 40—48 hours and were then offered Aufwuchs-covered stones from the lake. Feeding behaviour was scored over a period of 30 minutes with two or three individuals per The community of rock-dwelling cichlids 303 experiment. Experiments were done for 9 to 15 individuals per species. To get a broader impression of the feeding repertoires that rock frequenting Lake Victoria haplochromines possess, the feeding behaviour of a few individuals of some less abundant species was studied as well. We calculated diet-overlap and behaviour overlap as the sum of the overlap per food item or feeding technique respectively. With three exceptions (Neochromis nigricans, “H.” nyererei, Paralabidochromis chilotes) the species used for this study are not formally described. Preliminary diagnoses and colour photographs were published recently (Seehausen 1996). A number of papers on their ecology, eco-morphology and taxonomy have been published (Seehausen & Bouton 1997; Bouton, See- hausen & van Alphen in press; Seehausen, Bouton, Zwennes & Lippitsch in prep.). In the meantime we use cheironyms for the undescribed species. Species recognition criteria used are those outlined by Seehausen (1996). With regard to the generic taxonomy of Lake Victoria cichlids considerable heterogeneity is prevalent in recent publications with some authors following Greenwood’s last general revision of 1980, others not. Though some of Greenwood’s genera, we believe, need to be rede- fined, we could assign many of the new species. For those that we could not assign, we use the old generic name “Haplochromis” in quotation marks. 45 per cent ss scr bse 1D DDD rw Wh fr Ph eOrhe i hweow i & re? Ff rR 2 Fig. 2: Species abundance distribution at Makobe Island. Means of four sampling periods (n = 5919 fishes): abbreviations as in table 1. n = widely distributed, non-rock-restricted species, w = widely distributed rock-restricted species, r = regional rock-restricted species, e = local endemics, i = introduced species, non-rock-restricted. 304 O. Seehausen &N. Bouton Table 1: List of cichlid species occurring at Makobe Island/Speke Gulf. ++++ widely distributed, non-rock-restricted spp., +++ widely distributed rock-restricted spp., ++ re- gional rock-restricted spp., + local endemics, x introduced species, non-rock-restricted. Algae scrapers Haplochromis “purple yellow” Neochromis nigricans (NIG) Neochromis “giant scraper” (GIA) Neochromis “blue scraper” (BLU) Xystichromis “carp” (CAP) Paralabidochromis “short snout scraper” (SSS) Neochromis “large eye nigricans” (LEN) Xystichromis “jitu” (JIT) Haplochromis “blue obliquidens” (BOB) Algae scrapers / Detritus eaters Oreochromis variabilis (VAR) Xystichromis “copper black” (CBL) Oreochromis niloticus (NIL) Oreochromis leucostictus (LEU) Animal Aufwuchs eaters Psammochromis riponianus (RIP) Paralabidochromis “rockkribensis” (RKR) Paralabidochromis chilotes (CHI) “H.” “zebra nyererei” (ZNY) Paralabidochromis “blue rockpicker” (BRP) Paralabidochromis “yellow pseudorockpicker” (YRP) “FH” “yellow chin pseudonigricans” (GPN) “H>” “pink anal” (SDW) Plankton eaters “FH” nyererei (NYE) Mollusc eaters Astatoreochromis alluaudi (AST) “A “striped crusher” (SCR) Paedophages Lipochromis cf. melanopterus (MEL) food unknown “H.” “offshore” (OFF) SEP bluesseeretzi(BSE) The community of rock-dwelling cichlids 305 Results I. Faunal composition and abundance distribution The community at Makobe Island is composed of at least twenty-seven cichlid species (table 1). They can be assigned to five groups of different gross distribution patterns: 15 Yo are species widely distributed in Lake Victoria and not restricted to rocky substrates. All others are restricted to rocky substrates though they may occur in the immediate vicinity of rocks above other substrata. 30 % are species widely distributed at rocky shores and islands, 30 % are species with regionally restricted distribution (i.e. western Speke Gulf species), 11 Yo are endemics of Makobe Island and 7% are introduced species (Oreochromis niloticus and O. leucostictus). The distribution of two species is not known. Figure 2 gives the mean abundance distribution at Makobe Island obtained over four sampling periods (n = 5919 fishes) and relates relative abundances to the distribution pattern groups. It shows that most of the numerically dominant species are species with a wide distribution, while most regional and all local endemics occur in low densities. This pattern was rather stable over the sampling years. Very similar results were obtained at another island in the Speke Gulf, Ruti Island, which is a steeply sloping island with very large compact rock boulders. This indicates (and is confirmed by samples from many other places), that the species that numerically dominate at one place do so at most of the places where they occur. II. Spatial community structure Figure 3 shows the relative abundance of the four dominant and some other species in different microhabitats and depths. Figure 3a is based on angling and net samples, Figure 3b on underwater censuses. The species show significantly different vertical distributions (Fisher’s exact probability test p < 0.05). Figure 4 gives two examples of horizontal species distribution patterns indicating that horizontal patterns are species specific as well. They reflect not only differences in microhabitat require- ments but likely also different degrees of microhabitat stenotopy. Xystichromis “copper black” lives in a wide range of microhabitats from the surf zone to several meters depth, between rocks of all boulder sizes and at strongly surf exposed places as well as in protected embayments and under floating vegetation. This relatively wide ecological range enables the species to inhabit a large part of the available rocky habitat. Its distribution is not indicated in Figure 4 because it was found in each of the 10x 10 m squares. Widely distributed in the offshore waters but largely absent from the inshore areas is “A? nyererei (Fig. 4) while “H” “zebra nyererei“ is restricted to crevices between medium sized rocks along the shore and at rocky islets. Haplochromis “blue obliquidens” finally, as an extreme example, occurs exclusively in less than 2m depth on the surface of gently sloping rock fields in surf protected embayments (Fig. 4). It appears that more or less specific demands restrict the distribution of species to differently sized fractions of the potentially available rocky habitat. However, it is not yet understood whether interspecific competition plays a role in delimiting the microhabitat distribution of some species. 306 O. Seehausen & N. Bouton II. Trophic community structure Figure 5a shows the result of laboratory feeding experiments for the four most abundant species plus two rare species. Feeding techniques are defined elsewhere (Seehausen, Bouton, van Alphen, Witte submitted). The composition of feeding ethograms of the six species differs quantitatively significantly different (U-test p<0.05). Coefficients for overlap are given in table 2. The species can be assigned to three major feeding behaviour groups according to the predominantly employed feeding technique: pull-scrapers, pickers (both benthic foragers) and snappers (pelagic foragers). Since behavioural differences were recorded in the absence of interspecific competition they are to be considered autecological. INTERSTICES OUTSIDE <1m DEPTH Depth (m) OUTSIDE 1-1.5m DEPTH 1.5-2m DEPTH 2-3m DEPTH CO nig EN gia SS blu WH cb! ZA nye EBzny E bob >3m DEPTH Fig. 3a: Examples of vertical habitat segregation. Relative densities of rock-dwelling haplo- chromine species in different microhabitats at Makobe Island: Interstices between rocks in shallow water and different depth ranges outside of the interstices. Based on net and angling samples. Abbreviations as in table 1. The community of rock-dwelling cichlids 307 Figure 5b shows the result of stomach content analyses for the same species. The four abundant species disperse over three major trophic groups that correspond to the three feeding behaviour groups: scrapers of filamentous algae, loose algae/ animal Aufwuchs eaters, and plankton eaters. A fourth trophic group, not reflected in the feeding behaviour due to absence of its specific prey, are snail eaters. However, their specific feeding technique, pharyngeal crushing, is well known (Greenwood 1974). The interspecific differences in diet are significant (U test p<0.05) with the exception of that between the two algae scrapers N. nigricans and N. “blue scraper”. Coefficients for diet-overlap are given in table 2. Several other trophic groups are rather consistent members of rock cichlid communities: pedophages, piscivores and crab eaters. These larger predators probably always occurred in lower densities than Aufwuchs and plankton feeders, however, after the Nile perch upsurge they completely disappeared at many places. At Makobe Island we frequently caught one pedophagous species (Table 1). OUTSIDE <0.5m DEPTH INTERSTICES Depth (m) 0.5 0.5-0.8m DEPTH 0.8 0.8-1.2m DEPTH O nig gía blu EN cbi IM cap E rkr brp E bse 1.2-2m DEPTH 2.0 Fig. 3b: Same as 3a but based on underwater censuses. 308 O. Seehausen &N. Bouton Legend "H." nyererei "H.""zebra nyererei" "A." "blue obliquidens" main island and rocky islets sampling area 10x10m square ÍA TAN Fig. 4: Examples of horizontal habitat segregation among haplochromine species at the west shore of Makobe Island. Among a pair of sibling species “H” “zebra nyererei” lives more inshore and in the immediate vicinity of rocky outcrops than “A” nyererei. Only the first one occurs frequently in crevices and rockpools within the island in several metres distance from the lake. H. “blue obliquidens” occurs only at small patches of suitable microhabitat. Table 2: Coefficients of trophic niche overlap. Feeding behaviour NIG” BLEW EBEZEZSYZENIE NIG —BLU” "EBE 7ZNSZENME .90 BLU .85 9 37 CBL 47 .56 29 .31 ZNY .36 .45 .12 31172 : : NYE 44 52 .00 .00 3 | AST 22 25 The community of rock-dwelling cichlids 309 Discussion The studied community of rock-dwelling cichlids in Lake Victoria shows a rich ecological structure. It is characterized by few very abundant and many rare species. The abundant species are usually widely distributed and represent the three trophic groups that at many places allow the highest levels of population density: scrapers feeding behaviour diet N. nigricans H. "blue scraper" N. nigricans H. "blue scraper" (n = 43) (n = 43) H. "copper black" H. "zebra nyererei" (n = 8) (n = 6) n= 46) (n = 10) — XA N A. alluaudi H. nyererei (n=1) (n=3) RISSE VIII Ce = snapping zooplankton EH picking insects N combing higher plant material == scraping Bryozoa pull-scraping pulling other material diatoms Nostoc colonies filamentous bluegreen algae filamentous green algae | snails E] snapping 44 picking MMMM scraping s.!. zooplankton other food-items filamentous algae Fig. 5: Left two columns: feeding behaviour ethograms of the four numerically dominant and two other species at Makobe Island. Right two columns: Composition of the diet of the four numerically dominant and two other species at Makobe Island. In the outer circle several categories are lumped to form three principal feeding techniques and the three major food categories that reflect them. 310 O. Seehausen &N. Bouton of filamentous algae, loose algae/animal Aufwuchs eaters and planktivores. Most locally endemic species occur in low densities. This pattern may suggest that the key niches in the communities of rocky habitat islands were filled by a few species early during the formation of the present days lake and that the finer ecological community structure is largely a secondary phenomenon due to ecological differentiation and speciation that happened on regional scales. This is a hypothesis that calls for tests by interdisciplinary approaches. The different species in the rock cichlid community can be well characterized ecologically. Each occupies only a particular portion of the total ecological niche available to rock-dwelling cichlids. The differences, however, are subtle and the ecological packing is very dense. Only the upper five meters are occupied by epilithic Aufwuchs scrapers but ten such species share this narrow depth zone. The peaks of their relative abundance are in different microhabitats and depths but those of neighbouring species are separated by no more than one to two meters. Among them are three of the four numerically dominant species. Trophic morphology of two of them (N. nigricans, N. “blue scraper”) is so similar that it is difficult to identify preserved specimens that lost colouration. Yet they exhibit autecological differences in feeding behaviour which, together with subtle aut- or synecological differences in microhabitat distribution (and possibly synecological competition effects on feeding behaviour), result in significantly reduced diet overlap. There is no evidence that the cichlid species assemblage of Makobe Island violates the competitive exclusion principle but the dense spatial packing of morphologically and ecologically very similar species is likely to make the community highly sus- ceptable to changes in the environment. Tightening and relaxation of interspecific competition may under such circumstances be some immediate response to contrac- tion and expansion of the habitat. Some fluctuation of habitat width is a natural seasonal phenomenon caused by water level fluctuations. However, recently more dramatic changes take place that lead to a more permanent alteration of the habitat: water transparencies are significantly decreasing since at least the late seventies, reducing the habitat width for algae feeders, and upwelling of anoxic waters at steep rocky cliffs affects the communities of planktivores and benthivores in deeper waters. It is possible that such processes add to the species loss in rocky areas that is caused by the impact of turbidity on colour vision and species-assortative mating (Seehausen, van Alphen & Witte 1997) but much more work is needed to understand this. Acknowledgements Many thanks to Jacques van Alphen, Kees Barel and Frans Witte for their critical advice on the manuscript. We are indebted to Mhoja Kayeba, Ruben Enoka, Anna Samwel Terry, Masoud Ilomo, Ali Marwa and Aloys Peter for their expertise in the fieldwork. Radhmina and Gonza Mbilinyi are thanked for much logistic support. The Tanzania Fisheries Research Institute (TAFIRI) and its director Prof. P.O.J. Bwathondi are thanked for the research permit. The director of the Mwanza centre of TAFIRI, Mr. Egid F.B. Katunzi and our colleagues at the centre are thanked for hospitality and support during our stays. The Nyegezi Freshwater Fisheries Research Institute and its Principal Mr. R. R. Mapunda are thanked for logistical help and Martin Brittijn for drawing some of the figures. This study was made possible due to WOTRO grant no. W84-282 and, thanks to the effort of L. Schadhauser from “Wild about animals” magazine, was subsidized by the TETRA Company in 1993. The community of rock-dwelling cichlids Sl References Bouton, N., O. Seehausen & J. J. M. van Alphen (in press): Resource partitioning among rock-dwelling haplochromines (Pisces: Cichlidae) from Lake Victoria. — Ecology of Freshwater Fish. Fryer, G. & T. D. Iles (1972): The cichlid fishes of the great lakes of Africa. — Oliver & Boyd, Edinburgh. 642 pp. Greenwood, P. H. (1974): The cichlid fishes of Lake Victoria, East Africa: the biology and evolution of a species flock. — Bull. Br. Mus. nat. Hist. (Zool.) Suppl. 6: 1—134. Greenwood, P. H. (1980): Towards a phyletic classification of the “genus” Haplochromis (Pisces, Cichlidae) and related taxa. Part II: the species from Lake Victoria, Nabugabo, Edward, George and Kivu. — Bull. Br. Mus. nat. Hist. (Zool.) 39: 1—101. Greenwood, P. H. (1981): Species flocks and explosive evolution. — pp. 61—74 In: P. H. Greenwood « P. L. Forey, eds., Chance, change and challenge — the evolving biosphere. Cambridge University Press & British Museum (Natural History), London. Lowe-McConnell, R. H. (1993): Fish faunas of the African great lakes: origins, diversity, and vulnerability. — Conservation Biology 7: 634—643. Oijen van, M. J. P, F Witte € E. L. M. Witte-Maas (1981): An introduction to ecological and taxonomic investigations on the haplochromine cichlids from the Mwanza Gulf of Lake Victoria. — Netherl. J. Zool. 31: 149—174. Seehausen, O. (1996): Lake Victoria Rock Cichlids, taxonomy, ecology and distribution. — Verduyn Cichlids, Zevenhuizen. 304 pp. Seehausen, O.,J.J.M.vanAlphen&F. Witte (1997): Cichlid Fish Diversity Threatened by Eutrophication That Curbs Sexual Selection. — Science 277: 1808—1811. Seehausen, O. & N. Bouton (1997): Microdistribution and fluctuations in niche overlap in a rocky shore cichlid community in Lake Victoria. — Ecology of Freshwater Fish 6: iS Witte, F, T Goldschmidt, J. Wanink, M. van Oijen, K. Goudswaard,E. Witte- Maas & N. Bouton (1992): The destruction of an endemic species flock: quantitative data on the decline of the haplochromine cichlids of Lake Victoria. — Env. Biol. Fish. 34: 1528: Ole Seehausen & Niels Bouton, Institute of Evolutionary and Ecological Sciences, Sections Morphology and Ecology, Subsection Animal Ecology, Leiden University, P. ©. Box 9516, NL-2300 RA Leiden, The Netherlands. 4 dl pe A 4 » i ti p ' x } y > Ue ae ree. reg Bd. 7 S. 313-320 | Bonn, September 1998 New and little known Nebria (Epinebriola) from the eastern Nepal Himalayas (Coleoptera, Carabidae) * V. G. Shilenkov Abstract. Data on taxonomy and geographical distribution of some species of Nebria, subgenus Epinebriola Daniel, from eastern Nepal are given. Two species are described as new: N. schawalleri n. sp. and N. tangjelaensis n. sp. Key words. Coleoptera, Carabidae, Nebria, taxonomy, new species, Nepal. Introduction The northeastern part of Nepal along the border to Sikkim in the east and to Tibet in the north was hitherto quite poorly investigated under entomological aspects because of difficult accessibility. I was able to study interesting material of Carabi- dae from that region between the Tamur and Arun river systems collected by J. Mar- tens and W. Schawaller during their expeditions in Nepal in 1983 and 1988 (map see fig. 11). The species of Nebria were collected in high altitudes up to 5000 m near snow cover and all belong to a single subgenus, Epinebriola Daniel. Representatives of that subgenus are restricted in distribution to the Hindukush and Himalayas mountain systems and are divided in some species groups. The eastern branch of this subgenus includes most derived species: rasa Andrewes and schawalleri n. sp., both possess a surplus setation on the pronotum, a bisetose penultimate labial palpomere and weakly impressed elytral striae dissappearing laterally. Without doubts these species are closely related to orestias Andrewes and cannot be separated from the subgenus Epinebriola inspite of the unique combination of the above mentioned features. Other material of the genus Nebria from different Himalayan localities was treated earlier by Andrewes (1929, 1932, 1936) and Ledoux (1984, 1985). The descriptions and measurements in the present paper are standardized as in Shilenkov (1982). Abbreviations and material depository: HW, head width; HL, head length; PW, pronotum width; PL, pronotum length; PBW, pronotum base width; EW, elytra width; EL, elytra length; ISU, Irkutsk State University (collection Shilenkov); SMNS, Staatliches Museum für Naturkunde Stuttgart; ZFMK, Museum Alexander Koenig Bonn. *) Results of the Himalaya Expeditions of J. Martens, No. 213. — For No. 212, see Entomol. Basiliensia 20, 1997. — J. M. sponsored by Deutscher Akademischer Austauschdienst and Deutsche Forschungs- gemeinschaft. 314 V. G. Shilenkov Descriptions Nebria (Epinebriola) schawalleri n. sp. Holotype, female: Nepal, Taplejung distr., ascent to Tangje La NW Walungchung Gola, 4400—4600 m, alpine steppe, 23. V. 1988 leg. Martens & Schawaller (SMNS). Description: Size large, standardized body length female 10.8 mm. Pitch-black, shiny, appendages red-brown, femora and antennal scape infuscated. Head (fig. 1) relatively large and wide, with shallow transverse impression behind prominent eyes. Labrum with anterior margin evidently prominent, bearing 6 setae. Clypeus convex, slightly sinuated apically, frontoclypeal suture strong and deep, frontal furrows short and flat, vertex distinctly wrinkled. Surplus setation of head capsule: 2 and 3 supraorbital pores asymmetrically, additional clypeal pore unilaterally and one pore (? artifact) at the middle of the frons. Antennae slender, extending in female almost to the middle of elytra, antennal scape elongate, with 1 seta, second antennomere with 1 seta ventroapically, third antennomere with 6 setae apically. Median tooth of mentum 2.5 times as short as epilobes, sharply bidentate with V-shaped notch. Ligula with fingerlike process bearing 2 long setae apically. Penultimate labial palpomere bisetose, submentum with transverse row of 11 setae. Pronotum (fig. 1) subcordate, basal sinuation of lateral margin moderately long and deep; basal angles short, acute, narrowly rounded apically, projected posteriorly; apical angles widely rounded, weakly prominent anteriorly; basal margin strongly bisinuate, with additional small notches just before basal angles; lateral explanation narrow, slightly broadened at middle; pronotum disc convex, with faint radial wrinkles and a pair of small pitlike impres- sions, basal foveae deep and narrow, basal and apical areas with sparse but rather strong punctuation, transverse and longitudinal impressions strong and deep, anterior transverse impression strongly sinuated basally and deeply impressed; basolateral seta present (at the left basal angle 2 setae), series of 5—6 midlateral setae sinuated in the apical half of pronotum. Proepisternum sparsely and weakly punctate; prosternal process elongate, rounded apically, unmargined. Elytra moderately convex, subovoid, elongate, narrowed basally; humeral sinuation very shallow, subapical sinuation absent; basal margination straight, merged smoothly with lateral margination, humeral carina absent or very slightly developed; apical angles rather sharp, rounded apically; subapical carina evanescent; striae shallow, with traces of punctuation; obliterate at the margins of elytra, stria 8 invisible; scutellar stria long, scutellar seta present; intervals scarcely convex, third interval with 4—6 very small discal setae adjoining stria 3. Metepisterna 1.5 times as long as wide, sparsely and faintly punctate. Hind coxa with 2 (or 1 unilaterally) basal seta(e) and 1 apical seta. Hindwings vestigial, without traces of venation. Third to fifth visible abdominal sterna each with 3 pairs of posterior paramedial setae and with rather deep oblique impression laterally; anal sternite of female with 6 setae on hind margin. Legs long and slender, metatarsomere 4 with long ventrolateral lobe. Microsculpture on head and pronotum isodiametric, consists of very small sculpticells; on elytra strongly reduced, consists of very small and transverse meshes. Apical gonocoxite of female long and slender (fig. 9), bearing 3 lateral setae on inner margin. Proportions: PW/HW=1.29; PW/PBW=1.48; EW/PW=1.61; PW/PL=1.36; EL/EW=1.56. Length 12.5 mm, width 4.7 mm. Diagnosis: The form of the pronotum and the setation are unique and separate this species from all representatives of the subgenus Epinebriola. Geographical distribution: Known only from the type locality in the eastern part of the Nepal Himalayas. Derivatio nominis: With great pleasure I name this species in honour of my collegue Dr. Wolfgang Schawaller from the Staatliches Museum fiir Naturkunde in Stuttgart, who collected this species together with J. Martens and who provided me with this very interesting material. Nebria from the eastern Nepal Himalayas 315 Fig. 1: Nebria schawalleri n. sp., head and pronotum, scale 1 mm. Nebria (Epinebriola) rasa Andrewes, 1936 Material: Nepal, Sankhua Sabha distr., from Thudam to Gabri Khola, 4000—4250 m, dwarf Rhododendron, 27. V. 1988 leg. Martens & Schawaller, 4 males, 3 females. Nepal, Sankhua Sabha distr., Kangla Khola E Thudam, 4100—4200 m, dwarf Rhododendron, rock debris, 24.-25. V. 1988 leg. Martens & Schawaller, 10 males, 1 female (ISU, SMNS, ZFMK). This species was described from Sikkim (Ratong Chu) on a single female collected in an elevation of 11 000 feet. The taxonomical status was unclear, Andrewes (1936) compared it with barbata Andrewes. The redescription is given below. Size moderate, standardized body length male 9.42, female 10.2 mm. Pitch-black, shiny, legs dark piceous with testaceous tarsi and tibiae (in some individuals tibia infuscated), antennae entirely red-brown, or antennal scape slightly infuscated. Head with shallow transverse impression behind prominent eyes; labrum with anterior margin straight or faintly prominent, bearing 6 setae; apical margin of clypeus straight or scarcely concave; frontal furrows hardly discernible, vertex faintly wrinkled; only with 1 pair of supraorbital setae. Antennae long and slender, extending in female to the middle of elytra, in male longer; antennal scape almost straight, with faintly sinuated apical margin, slightly narrowed basally, bearing 1 seta, second antennomere with 1 or 2 seta(e), third antennomere 316 V. G. Shilenkov with 5—6 setae apically. Median tooth of mentum 2.0 times as short as epilobes, bidentate with shallower notch than in the previous species, denticules apically obtuse. Ligula with fingerlike process bearing 2 long setae apically. Penultimate labial palpomere bisetose, submentum with transverse row of 10—12 setae, interrupted medially. Pronotum (fig. 2) convex, subcordate, basal sinuation of lateral margin very short and shallow, situated just before very small and acute basal angles, projected posteriorlaterally; apical angles widely rounded, weakly prominent anteriorly; basal margin bisinuate; lateral explanation narrow, slightly broadened basally; transverse and longitudinal impressions distinct but not so strong and deep as in the previous species; punctuation faint and sparse; basolateral seta present, series of 3—4 midlateral setae situated in the apical half of pronotum. Proepisternum sparsely and weakly punctate; prosternal process elongate, rounded apically, unmargined. Elytra in shape as in the previous species, but apical angles more rounded; striae and inter- vals with the same structure, scutellar seta absent, rarely present unilaterally; discal setae very small, 4—5 setae in third interval adjoining stria 3, 2—4 setae in fifth and 0—4 in seventh interval (in the original description only 6 setae in the third interval are mentioned). Metepisterna 1.5 times as long as wide, smooth. Hind coxae with 2 basal and 1 apical setae. Hindwings verstigial, without traces of venation, shorter than in the previous species. Third to fifth visible abdominal sterna each with 2 or 3 pairs of posterior paramedial setae and with rather deep oblique impression laterally; anal sternite of male with 5—6, of female with 6 setae on hind margin. Legs as in the previous species, in male 3 first tarsomeres of protibia weakly expanded and with pad of adhesive setae ventrally. Microsculpture faintly impressed, on head isodiametric, consists of very small sculpticells, on pronotal disc slightly transverse, almost isodiametric, on elytra consists of very small transverse meshes. Basal part of median lobe (fig. 6) wide, with strongly prominent triangle basal lobes, apex of median lobe long and slender. Apical gonocoxite of female shorter than in the previous species (fig. 8). Proportions: PW/HW = 1.21—1.23; PW/PBW = 1.46—1.54; EW/PW = 1.60—1.70; PW/PL = 1.171.342 ERZAEW 52064: Length 11.0—11.9 mm, width 4.2—4.5 mm. Diagnosis: Differs from schawalleri n. sp. by the form of the pronotum, by the setation on head and pronotum, by the structure of the female apical gonocoxite and by smaller size. Geographical distribution: Eastern parts of the Himalayas in Nepal and Sikkim. Nebria (Epinebriola) tangjelaensis n. sp. Holotype, male: Nepal, Taplejung distr., ascent to Tangje La NW Walungchung Gola, 4800—5000 m, alpine steppe, 23. V. 1988 leg. Martens & Schawaller (SMNS). Paratypes: With the same labels, 5 males, 1 female (ISU, SMNS, ZFMK). Nepal, Taplejung distr., ascent to Tangje La NW Walungchung Gola, 4400-4600 m, alpine steppe, 23. V. 1988 leg. Martens & Schawaller, 1 male (SMNS). Description: Size rather small, standardized body length male 7.74, female 8.10 mm. Dark piceous, shiny, mouthparts, antennae and legs red-brown, rarely femora infuscated. Head with shallow transverse impression behind prominent eyes; labrum with anterior margin straight; apical margin of clypeus straight or scarcely concave; frontal furrows hardly discernible or evanescent, vertex smooth; only 1 pair of supraorbital setae. Antennae long and slender, almost extending to the middle of elytra, antennal scape almost straight, with faintly sinuated apical margin, moderately narrowed basally, bearing 1 seta, second antennomere with 1 seta lateroapically, third antennomere with 5—6 setae apically. Ligula with fingerlike process bearing 2 long setae apically. Penultimate labial palpomere trisetose, submentum with transverse row of 11—12 setae, interrupted medially. Nebria from the eastern Nepal Himalayas 317 Figs 2—10: 2, Nebria rasa Andrewes, pronotum; 3, Nebria tangjelaensis n. sp., pronotum, scale 1 mm; 4, Nebria tangjelaensis n. sp., median lobe from dorsal; 5, Nebria tangjelaensis n. sp., median lobe from left lateral; 6, Nebria rasa, median lobe from left lateral; 7, Nebria rasa, median lobe from dorsal; 8, Nebria rasa, left apical gonocoxite from ventral; 9, Nebria schawalleri n. sp., left apical gonocoxite from ventral; 10, Nebria tangjelaensis n. sp., left apical gonocoxite from ventral. Pronotum (fig. 3) moderately convex, subcordate, uniformly narrowed basally, lateral margins in basal half almost straight, without basal sinuation, or with long and very shallow basal sinuation, basal angles obtuse with small and apical denticules, projected posterior- laterally; apical angles widely rounded, weakly prominent anteriorly; basal margin moderately bisinuate; lateral explanation very narrow, sligthly broadened basally; transverse impressions rather strong, longitudinal impression shallow but distinct, punctuation faint and sparse; basolateral seta present, midlateral seta situated in the apical quarter of pronotum. Proepister- num almost smooth, with sparse and weak punctures; prosternal process elongate, rounded apically, unmargined. Elytra moderately convex, in some individuals longitudinally depressed along suture, subovoid, elongate, faintly narrowed basally; humeral sinuation long and very shallow, sub- apical sinuation almost evanescent; basal margination straight, merged smoothly with lateral 318 V. G. Shilenkov margination, humeral carina absent or very slightly developed; apical angles narrowly rounded; subapical carina short and hardly flattened; striae moderately impressed, distinctly punctate on disc, obliterate at margins and apex of elytra; scutellar stria long, scutellar seta absent; intervals on disc rather convex, flattened laterally and apically; discal setae absent. Thudam a SR Kancnenme Walungchung E, 5804 oy Pomri Lay: 4780 o Au x IT eo/rali” 4500 ¡ O: y .*Hurure er Chichila— as = 3 Rarapan ma i Rhandbari |Khebang Uya m, 2000 „£Worebung Pos "Yektis— wy „Päniporua | / \ X 7 \ “eSheldoti + SEN Deorali YE Dhorpar” 2850 ;\ >. Fig. 11: Map of localities in eastern Nepal. o. orestias, r. rasa, s. schawalleri n. sp., t. tang- Jelaensis n. sp. Nebria from the eastern Nepal Himalayas 319 Metepisterna 1.5 times as long as wide, longitudinally depressed, almost smooth. Hind coxae with 1 basal seta and 1 apical seta, their apical part evidently punctate. Hindwings vestigial, without traces of venation. Third to fifth visible abdominal sterna each with 3 or 4 pairs of posterior paramedial setae and with rather deep oblique impression laterally; anal sternite of male with 4, of female with _ 5—6 setae on hind margin. Legs long and slender, in male 3 first tarsomeres of protibia weakly expanded and with pad of adhesive setae ventrally, metatarsomere 4 with rather long ventrolateral lobe. Microsculpture faintly impressed, on head isodiametric, consists of very small sculpticells, on pronotal disc slightly transverse, almost isodiametric, on elytra consists of very small trans- verse meshes. Median lobe (fig. 5) strongly curved, rather stout, with apex short and acute. Apical gono- coxite of female smaller than in the previous species (fig. 10). Proportions: PW/HW = 1.10—1.21; PW/PBW = 1.52—1.59; EW/PW = 1.57—1.72; PW/PL = lO SSS EL/EW = 1561.65. Length 8.9—9.2 mm, width 3.3—3.4 mm. Diagnosis: Closely related to orestias Andrewes, 1932, but smaller, body paler, pronotum different, especially in the form of basal angles, median lobe of male smaller with shorter apex. Geographical distribution: Known only from the type locality in the eastern Nepal Himalayas. Derivatio nominis: This species is named after the pass Tangje La, the type locality. Nebria (Epinebriola) orestias Andrewes, 1932 Material: Nepal, Taplejung distr., ascent to Tangje La NW Walungchung Gola, 4400— 4600 m, alpine steppe, 23. V. 1988 leg. Martens & Schawaller, 11 specimens. Nepal, Taplejung distr., Anda Deorali between Simbua and Gunsa Khola, 4250—4500 m, alpine zone, 9. IX. 1983 leg. Martens & Daams, 3 specimens. Nepal, Sankhua Sabha distr., from Thudam to Gabri Khola, 4000—4250 m, dwarf Rhododendron, 27. V. 1988 leg. Martens & Schawaller, 2 specimens. Nepal, Sankhua Sabha distr., ascent to Meropapa La from Gabri Khola S Thudam, 4300—4600 m, meadows with dwarf Rhododendron, 26. V. 1988 leg. Martens & Schawaller, 8 specimens. Nepal, Sankhua Sabha distr., descent from Pomri La, S slope, 4550—4450 m, under snow covered stones, 29. V. 1988 leg. Martens & Schawaller, 1 specimen (SU, SMNS, ZFMK). The species was described by Andrewes (1932) from “Sikkim-Tibet, Jelep Pass“. The taxono- mical status of this species was unclear. Recently, Ledoux (1985) included the species in his subgenus Himalayonebria, briefly characterized in the determination key and by figures of the pronotum and male genitalia. Geographical distribution: In the eastern Himalayas in Nepal and Sikkim and probably in the adjacent territory of Tibet. Acknowledgements Ithank Dr. W. Schawaller for the loan of the specimens, for his support and hospitality during my visit in the Staatliches Museum für Naturkunde in Stuttgart and for the help in publishing this paper. Thanks are due also to G. Ledoux for sending types and other important material from the Himalayas. Zusammenfassung Daten zur Taxonomie und geografischen Verbreitung einiger Nebria-Arten, Untergattung Epinebriola Daniel, aus Ost-Nepal werden mitgeteilt. Zwei Arten werden neu beschrieben: N. schawalleri n. sp. und N. tangjelaensis n. sp. 320 V. G. Shilenkov References Andrewes, H. E. (1929): The fauna of British India, including Ceylon and Burma. Coleop- tera. Carabidae. Vol. 1. Carabinae. — Taylor & Francis, London, 431 p. Andrewes, H. E. (1932): Papers on oriental Carabidae. XXVI. — Ann. Mag. nat. Hist. ser. 103927153146: Andrewes, H. E. (1936): Papers on oriental Carabidae. XXX. — Ann. Mag. nat. Hist. ser. 10, 18: 54—65. Ledoux, G. (1984): Contribution a la connaissance des Nebria de la region Himalayenne (Ire note). — Annls Soc. ent. Fr. (NS) 20: 275—282. Ledoux, G. (1985): Contribution a la connaissance des Nebria de la région Himalayenne (2e note). — Entomologiste 41: 279—288. Shilenkov, V. G. (1982): New and little-known ground-beetles of the genus Nebria Latr. (Coleoptera, Carabidae) from Asia. — Insects of Mongolia 8: 241—283 (in Russian). Dr. V. G. Shilenkov, Irkutsk State University, Dept. of Invertebrate Zoology and Hydrobiology, Suche-Bator street 5, 664003 Irkutsk, Russia. Bd. 47 S.321—343 Bonn, September 1998 Die Sandlaufkäfer (Coleoptera: Cicindelidae) des Comoe-Nationalparks, Elfenbeinküste: Faunistik, Zoogeographie und Okologie Jakob Fahr Abstract. The tiger beetles (Coleoptera: Cicindelidae) of the Comoé National Park, Ivory Coast: faunistics, zoogeography and ecology. 23 tiger beetle species are reported from the study area covering about 10 km?. Cylindera lutaria, Habrodera nilotica, Lophyridia fimbriata and Myriochile flavidens are recorded for the first time from the Ivory Coast. Euryarthron planatoflavum is recorded for the first time from Upper Volta. One new species (Dromicoida elegantia) was discovered (Werner 1995). The study site supports a species-rich tiger beetle fauna, which is caused by a highly diverse habitat mosaic. Only very few species are restricted to the Ivory Coast and adjacent countries, many reach as far as Central or East Africa. Possible explanations for this pattern are discussed. Tiger beetles are members of a carnivorous arthropod guild which hunt at very high temperatures. Because of the extreme environmental conditions there are very few competitors. Moreover, tiger beetles seem to compete mostly with sympatric congeners because of very similar life histories. The specific habitat preferences of each species are analysed and the Hutchinsonian ratios of sympatric tiger beetles with respect to mandible length compared. It is shown that species which occur in the same habitat and exhibit a similar behaviour generally follow the predicted character divergence. In contrast, sympatric species with different micro-habitat preferences show only minor divergence. These results are discussed as a response to reduce competition. Key words. Coleoptera, Cicindelidae, tiger beetles, West Africa, Ivory Coast, zoogeo- graphy, distribution, ecology. Einleitung Sandlaufkáfer sind innerhalb der Caraboidea eine der taxonomisch am besten unter- suchten Gruppen. Weltweit sind über 2000 Arten beschrieben. Da die Imagines von Cicindeliden im natürlichen Lebensraum relativ gut zu erfassen und zu beobachten sind, eignen sie sich besonders gut für eine Vielzahl von Fragestellungen. In den letz- ten Jahren sind insbesondere Aspekte der interspezifischen Konkurrenz und Mikro- habitatswahl in den Vordergrund gerückt (Pearson & Mury 1979, Pearson 1980, Pearson & Knisley 1985, Ganeshaiah & Belavadi 1986, Schultz & Hadley 1987, Pearson & Juliano 1991). Dadurch können Aussagen über Standortkonstanz und Nischengröße getroffen werden. In Gebieten, die eine artenreiche Cicindelidenfauna beherbergen, sind zudem charakteristische Vergesellschaftungsmuster festzustellen. Es stellt sich vor allem für tropische Breiten die Frage, ob diese Artenkombinationen durch exogene Faktoren determiniert werden. Dazu ist eine genaue Analyse der spezi- fischen Habitatsansprüche sowie der Phänologie, Larvalbiologie und Verbreitungs- muster notwendig. Sollten diese Vergesellschaftungen einen stabilen Zustand für einen gegebenen Zeitraum darstellen, dann müßten auf Grund des Konkurrenzaus- 322 J. Fahr schlußprinzipes die syntopen Arten eine Nischendifferenzierung aufweisen, die sich als morphologische Differenzierung manifestieren kann. Die meisten Sandlaufkäferarten sind tagaktive, räuberische Caraboiden, die optisch orientiert ihre Beute jagen und spärlich bewachsenen Untergrund mit Roh- bodencharakter bevorzugen (Ausnahmen bilden die in der Regel nachtaktiven Manticorini und die arborealen Vertreter der Collyrinae). Weiterhin zeichnet sich die Familie durch eine hohe morphologische Ähnlichkeit aus. Cicindeliden sind bei sehr hohen Temperaturen aktiv und nur wenige carnivore Arthropoden kommen unter diesen Bedingungen als mögliche Konkurrenten in Frage (z. B. Spinnen und Amei- sen). Im Freiland und in Laborexperimenten wurde nachgewiesen, daß Nahrung als limitierender Faktor sowohl für Imagines als auch für Larven eine entscheidende Rolle spielt (Pearson & Knisley 1985, Knisley, 1987). Eine-mögliche Form der Kon- kurrenzverminderung bei koexistierenden Arten kann durch eine Merkmalsverschie- bung (character divergence bzw. -displacement) erreicht werden, indem ein Merkmal, das zur effektiven Nutzung einer limitierten Ressource entscheidend beiträgt, bei den syntop vorkommenden Arten unterschiedlich ausgeprägt ist (Hutchinson 1959, Schoener 1965). Hutchinson (1959) ermittelte einen empirischen Wert von 1.28 für die mittlere minimale Differenz von Mundwerkzeugen innerhalb syntoper Arten einer Gilde. Bei Sandlaufkäfern ist die Mandibellänge mit der Körpergröße des bevorzugten Beutespektrums positiv korreliert (Pearson & Mury 1979, Pearson & Stemberger 1980). Für bestimmte Habitate wurde gezeigt, daß syntop vorkommende Cicindeliden- arten in Bezug auf Mandibellänge der Hutchinson-ratio folgen (Pearson & Mury 1979, Pearson 1980, Pearson & Juliano 1991). Es muß allerdings beachtet werden, daß eine Nischendifferenzierung durch eine Vielzahl von verschiedenen Variablen ermöglicht werden kann. Es können sich die einzelnen Arten z.B. in ihrer Larval- biologie, bevorzugtem Substrat, Bodenfeuchte, Temperatur, Verhalten gegenüber Prädatoren usw. so unterscheiden, daß Koexistenz ermöglicht wird (Dreisig 1980, 1981, Guppy et al. 1983, Ganeshaiah & Belavadi 1986, Knisley 1987, Mury Meyer 1987, Schultz & Hadley 1987). Ganeshaiah & Belavadi (1986) untersuchten die Mikrohabitatswahl von vier sympatrischen Sandlaufkäferarten an einem Flußufer in Indien und wiesen nach, daß sich ihre Mikrohabitate nicht überlappten und daher auch keine direkte Nahrungskonkurrenz vorhanden war, daß aber Parameter wie vorherrschende Beutegröße, Bodenfeuchte und Deckungsgrad der Vegetation zuver- lässig die Verteilung der einzelnen Arten vorhersagten. Daher ist es notwendig, alle Möglichkeiten der Nischendifferenzierung in Betracht zu ziehen und gegeneinander abzuwägen. In der vorliegenden Arbeit war es auf Grund des beschränkten Untersuchungszeit- raumes nicht möglich, viele der biotischen und abiotischen Parameter ausreichend zu erfassen. Daher stützt sich die im nachhinein durchgeführte Analyse der Mandi- bellängen auf Annahmen, die durch die Feldbeobachtungen der einzelnen Arten gewonnen wurden. Untersuchungsgebiet Die Untersuchung wurde vom 19. 3.—26. 4. 1993 im Comoé-Nationalpark, Elfenbeinküste, Westafrika, durchgeführt (Abb. 1). Der Nationalpark liegt im Nordosten der Elfenbeinküste Sandlaufkäfer des Comoé-Nationalparks, Elfenbeinküste 323 und erstreckt sich über eine Fläche von 11 500 km? (9°6’N—8°5’N; 31'W—44"W). Die ausgeprägte Trockenzeit reicht von November bis März, allerdings variieren Niederschlagsver- teilung und -menge oftmals erheblich. Das eigentliche Untersuchungsgebiet lag in einem Umkreis von ca. 5 km um das „Camp an der Lola“, das vom Zoologischen Lehrstuhl III für Tierökologie und Tropenbiologie der Universität Würzburg unterhalten wird. Das Camp liegt im Südosten (Abb. 2) des Nationalparks und damit im nördlichsten Bereich der Guinea-Zone. % %, oy, Li : red Oe /__Elfenbeinktiste Qs: AE: Cnn: er % Golf io von Guinea 28 Abb. 1: Geographische Lage der Elfenbeinküste in Westafrika (nach Porembski 1991). Ferkessé dougou N 2 2 ER Camp an der Lola Comoe - Dabakala Nationalpark eKatiola 0 50 km Bondoukou oe Abb. 2: Comoé-Nationalpark mit seinen wichtigsten Fliefgewássern sowie Lage des ,Camp an der Lola“ (nach Porembski 1991). 324 I Bahr Durch diese Lage im Übergangsbereich zwischen Guinea- und Sudan-Zone ergibt sich ein mosaikartiges Landschaftsbild, das durch die jeweils kleinräumig vorherrschenden Klima-, Boden- und Wasserverhältnisse bestimmt wird. Dominante Vegetationseinheiten sind die Savannenformationen, Galeriewälder und Inselwälder, als kleinräumig strukturierende Ein- heiten sind Termitenhügel, Felsplateaus, Savannentümpel und vegetationsarme Alluvial- flächen (plaine) zu nennen. Für eine ausführliche Darstellung der Vegetation sei auf Porembski (1991) verwiesen. Die Regenzeit hatte 1993 sehr früh begonnen (erste Niederschläge Ende Februar, Abb. 3), so daß ideale Bedingungen für eine Arbeit an Cicindeliden herrschten, da die Imagines mit Einsetzen der Regenfälle zahlreich erscheinen, um dann im Verlauf der Regenzeit mit zuneh- mender Vegetationshöhe wieder abzunehmen bzw. zu verschwinden. Material und Methoden Im oben beschriebenen Untersuchungsgebiet wurden täglich verschiedene Habitate abgegan- gen, in denen Sandlaufkäfer auftraten. Von jeder zum ersten Mal angetroffenen Art wurden ein bis fünf Exemplare mit einem Handkescher gefangen und mit Ethylacetat abgetötet. Alle wichtigen Parameter des Habitates (Bodenbeschaffenheit und -feuchtigkeit, Laubschicht, Bodenbewuchs, Vegetationsform, Besonnung, Wassernähe etc.) notierte ich auf Fundort- zetteln, die den Käfern beigefügt wurden. Zusätzlich hielt ich fest, welche bereits bekannten Arten mit der jeweiligen Art vergesellschaftet waren oder ob neue Artenkombinationen zu beobachten waren. Außerdem erhielt ich gesammelte Exemplare und Daten zu einigen Arten aus der Zeit vom 20. 5.—11. 6. 1993 von Mark Oliver Rödel. In einem Fall wurden sämtliche Sandlaufkäfer abgefangen, die sich auf einer feuchten Schlickfläche aufhielten, und danach konserviert. Am nächsten Tag wiederholte ich den quantitativen Abfang, um die Wiederbesiedlungsfähigkeit der vorher angetroffenen Arten zu untersuchen. Bei drei Arten (E. walterhorni, E. planatoflavum und R. cinctus) wurden im Bereich des Galeriewaldes einzelne Exemplare mit farbigen Lackfarben individuell markiert, um deren Aktivität im Tagesverlauf und über die Wochen beobachten zu können. Nach Rückkehr wurden sämtliche Käfer (267 Individuen) einheitlich präpariert und anschließend unter einem Stereomikroskop mit Meßokular vermessen (Mandibellänge: Jeweils die linke Mandibel, von der außenliegenden Gelenkstelle bis zur Mandibelspitze; Körperlänge: Vom Kopf (sine labro) bis zum Apex der geschlossenen Elytren). Wenn vorhan- den, wurden pro Art 10 © und 10 Q vermessen. Einige Arten lagen allerdings in geringerer Anzahl oder sogar nur als Einzelexemplare vor. Da bei den in ausreichender Anzahl vorhande- nen Arten die Variabilität vor allem der Mandibellänge äußerst gering war (einzige Ausnahme: C. chrysopyga mit ausgeprägtem Sexualdimorphismus), wurden die in geringerer Anzahl vor- handenen Arten in die Analyse miteinbezogen. Es wurde bei jeder Art der Mittelwert und die Standartabweichung für die Geschlechter separat berechnet. Falls sich die Mandibellängen von © und Q nicht signifikant unterschieden, wurden die Werte gepoolt. Die Bestimmung erfolgte mit Primärliteratur und umfangreichen Vergleichssammlungen (Zoologische Staatssammlung München und Sammlung Museum Frey sowie den Privat- sammlungen von K. Werner, Peiting, und W. Lorenz, Tutzing). Teile des bearbeiteten Materials befinden sich in der Zoologischen Staatssammlung München (ZSM), im Zoologischen Forschungsinstitut und Museum Alexander Koenig, Bonn (ZFMK) und in der Sammlung des Autors. Die taxonomische Einordnung der einzelnen Arten und Großtaxa folgt dem „Verzeich- nis der Sandlaufkäfer der Welt“ von Wiesner (1992). Faunistik und Zoogeographie In einem Gebiet von ca. 10 km? Größe wurden 23 Cicindelidenarten nachgewiesen. Davon waren vier Arten bisher für die Elfenbeinküste unbekannt (L. fimbriata, A. nilotica, C. lutaria und M. flavidens). Eine Art (Dromicoida elegantia) war unbe- Sandlaufkäfer des Comoé-Nationalparks, Elfenbeinküste 325 schrieben (Werner 1995). Nicht im Comoé-Nationalpark nachgewiesen wurden E. deyrollei, R. feisthameli, H. caternaulti, C. flavomaculata ssp. tripunctata, L. sara- liensis und C. decellei, die von anderen Gebieten der Elfenbeinküste bekannt sind (siehe unten). Damit sind insgesamt 29 Sandlaufkáferarten für die Elfenbeinküste nachgewiesen. Das ca. 10 km? große Gebiet innerhalb des Comoé-Nationalparks beherbergt demzufolge 79 Y aller Arten, die bisher für die Elfenbeinküste nach- gewiesen wurden, und zeugt von der hohen Habitatsdiversität des Nationalparks durch dessen Mosaikstruktur (siehe „Untersuchungsgebiet“). Es wäre zu unter- suchen, inwieweit einzelne Arten auf dieses Mosaik angewiesen sind, indem sie im Tages- oder Jahresverlauf bzw. als Larven und Imagines diese unterschiedlichen Bio- Niederschläge [mm] Datum Abb. 3: Niederschlagsverteilung vom 23. 2.—11. 6. 1993 im Bereich des „Camp an der Lola“ zu Beginn der Regenzeit. 13% Unbestimmt : 26% Galerie- & Inselwálder 13% Sandbánke & Flußbett Comoé 17% Schlickufer & Savannengewásser 30% Savanne Abb. 4: Verteilung der 23 nachgewiesenen Arten in den Großbiotopen (unbestimmt: bevorzug- tes Biotop konnte nicht ermittelt werden, z. B. Lichtfänge). 326 J. Bahr tope nutzen. Einige beobachtete Arten scheinen auch Hinweise auf diese Annahme zu liefern (siehe Beschreibung £. walterhorni, E. planatoflavum und L. neglecta). Der Anteil von Savannen- und Galeriewaldarten im Untersuchungsgebiet ist an- nähernd gleich groß und entspricht der Vegetationszusammensetzung im südlichen Teil des Nationalparks (Porembski 1991), der im Vergleich zum nördlichen einen noch relativ hohen Anteil an immergrünen Waldbereichen (Galerie- und Inselwälder) aufweist (Abb. 4). Problematisch bei einer solchen Zuordnung der Arten ist die Tat- sache, daß manche Cicindeliden Übergangsbereiche besiedeln oder im Tages- bzw. Jahresverlauf von einem Großbiotop in ein anderes wechseln (siehe oben). Insgesamt entspricht aber das Verteilungsmuster demjenigen, das man auf Grund der Mosaik- struktur des Untersuchungsgebietes erwartet, und die meisten Arten zeigen eine strenge Habitatsbindung. Betrachtet man die Verbreitungsgebiete aller für die Elfenbeinküste nachgewiese- nen Sandlaufkäferarten, so zeigt sich, daß Dromicoida elegantia „endemisch“ in der Elfenbeinküste und dort im Norden vorkommt. Dies stützt sich aber bisher nur auf die drei im Comoé-Nationalpark gesammelten Exemplare, und man kann daher davon ausgehen, daß sich mit Sammlungen in Nachbarländern die Kenntnis des Verbreitungsgebietes erweitern wird. Es fällt auf, daß nur 3 Arten (ca. 10 %) ein im Bereich der Elfenbeinküste beschränktes Verbreitungsgebiet und ebenfalls nur 4 Arten (ca. 14%) eine rein westafrikanische Verbreitung besitzen (siehe Tab. 1). Tabelle 1: Verbreitungsmuster der 29 in der Elfenbeinküste vorkommenden Arten innerhalb Afrikas. Elfenbeinküste & weitere Verbreitung Afrika (21 Arten) Elfenbeinküste Elfenbeinküste & Elfenbeinküste & „endemisch“ Nachbarländer Westafrika (1 Art) (3 Arten) (4 Arten) D. elegantia E. walterhorni M. bocandei P concinna E. planatoflavum M. denticollis P. leprieurii C. decellei H. caternaulti E. lugubris | M. plurinotata E. deyrollei R. vittatus R. feisthameli R. nysa R. cinctus H. interruptum C. flavomaculata L. fimbriata L. neglecta L. luxerii L. saraliensis HA. nilotica C. lutaria C. octoguttata M. flavidens M. fastidiosa M. melancholica C. chrysopyga Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinkúste 327] Dagegen kommen 21 der vertretenen Arten (ca. 72 %) über große Bereiche Afrikas vor. Davon erreichen 7. nilotica und M. fastidiosa ssp. vicina Nordafrika, P leprieurii den Yemen und M. melancholica dringt über Südeuropa bis nach Indien vor. Der geringe Endemismusgrad und die kleine Anzahl von Arten mit eingeschränktem Ver- breitungsgebiet ist kennzeichnend für die Cicindelidenfauna Westafrikas (siehe auch Pearson & Cassola 1992). Dies steht im scharfen Kontrast zu Ost- und Südafrika, die teilweise einen sehr hohen Prozentsatz endemischer Arten aufweisen (vor allem Län- der wie Äthiopien, Kenia, Tansania, Somalia, Mozambique und Südafrika), aber auch Zentralafrika (Kamerun, Zaire und Angola). Dies könnte darin begründet sein, daß die Sudan/Sahel-Zone in Westafrika ein relativ barrierearmes Areal darstellt (vor allem durch das Fehlen großer Gebirgsstöcke) und paläoklimatisch die weite Ausbrei- tung von Savannen- und Steppenformationen zu den Maxima der Eiszeiten (Living- stone 1975) nur bedingt die Bildung von Endemismen in solchen Habitaten förderte. Verzeichnis der für den Comoé-Nationalpark nachgewiesenen Sandlaufkäferarten Subfam. Cicindelinae Csiki, 1906 Tribus Megacephalini Csiki, 1906 Megacephala bocandei bocandei Guerin, 1848 Megacephala denticollis schultzeorum W. Horn, 1904 Tribus Cicindelini Sloane, 1906 Subtribus Prothymina W. Horn, 1908 (sensu Rivalier, 1971) Prothyma concinna erythrocnema Chaudoir, 1850 Prothyma leprieurii (Dejean, 1831) Euryarthron walterhorni Cassola, 1983 Euryarthron planatoflavum (W. Horn, 1922) Subtribus Cicindelina W. Horn, 1908 Dromicoida elegantia (Werner, 1995) Elliptica lugubris (Dejean, 1825) Ropaloteres vittatus (Fabricius, 1801) Ropaloteres nysa (Guerin, 1849) Ropaloteres cinctus (Olivier, 1790) Hipparidium interruptum (Fabricius, 1775) Lophyridia fimbriata (Dejean, 1831) Lophyra neglecta (Dejean, 1825) Lophyra luxerii (Dejean, 1831) Habrodera nilotica (Dejean, 1825) Cylindera lutaria (Guerin, 1849) Cylindera octoguttata (Fabricius, 1787) Myriochile fastidiosa vicina (Dejean, 1831) Myriochile flavidens (Guerin, 1849) Mpyriochile plurinotata (Audouin & Brulle, 1839) Myriochile melancholica (Fabricius, 1798) Cratohaerea chrysopyga (W. Horn, 1892) Verzeichnis der Cicindelidenarten, die fiir die Elfenbeinkúste, aber nicht den Comoé-Nationalpark nachgewiesen sind Elliptica deyrollei (Guerin, 1849) „Lamto-Savanne“ (Lecordier 1972) Ropaloteres feisthameli (Guerin, 1849) „Korogho á Boundiali; leg. Schmitz“ (Basilewsky 1968) 328 J: Bahr Hipparidium caternaulti (Guerin, 1849) „Bingerville, Koun-Abronso, Akoupé; leg. Decelle“ (Basilewsky 1968) Calochrea flavomaculata tripunctata (Dejean, 1837) „Korogho á Boundiali; leg. Schmitz“ (Basilewsky 1968) Lophyra saraliensis (Guerin, 1849) „Bingerville; leg. Decelle“ (Basilewsky 1968) Cylindera decellei Basilewsky, 1968 „Gagnoa, Korea, Aboisso, Abengourou; leg. Decelle“ (Basilewsky 1968) Bemerkungen zur Biologie der Sandlaufkäferarten Megacephala bocandei. — 5 Exemplare in Eimern einer Anlage zum Fangen von Amphibien, Umgebung Lola-Tümpel, Lola-Plaine (4./10./13./19. 4.). 1 Exemplar im Übergangsbereich Savanne/Galerie (Camp) nachts am Licht. Nachtaktive, flugunfähige Savannenart, die im Comoe-N.P. nur im Bereich der Alluvialflächen gefunden wurde. Verbreitung der Nominat- form: Senegal, Guinea Bissau, Guinea, Elfenbeinküste, Ghana, Togo, Benin, Nigeria. Megacephala denticollis schultzeorum. — 3 Exemplare nachts im Galeriewald des Comoé (Camp), jeweils einen Tag nach Regenfällen (2 Exemplare am 19. 3., am 18. 3.: 7 mm Nieder- schlag; 1 Exemplar am 10. 4., am 9. 4.: 20 mm). Nachtaktive, flugunfähige Waldart, im Gale- riewald entlang des Comoé auf Laubstreu, zum Teil im Licht der Campbeleuchtung jagend, aber nicht vom Licht „festgehalten“. Verbreitung der Unterart: Elfenbeinküste, Obervolta, Togo, Benin, Nigeria, N-Kamerun. Prothyma concinna erythrocnema. — 9 Exemplare, meist im Galeriewald oder Übergangs- bereich Galerie/Savanne, im Halbschatten bis Schatten auf Laubstreu, oft im Schutz des breit- blättrigen Grases Oplismenus hirtellus (L.) jagend, das im Unterwuchs der Galerie häufig anzutreffen ist und lückenhafte Bestände bildet. Diese Art wurde auch zweimal in Inselwäl- dern auf unbewohnten und stark zerfallenen Termitenhügeln angetroffen. Alle beobachteten Exemplare traten einzeln auf, sie hielten sich aber meist in Waldbereichen auf, die auch von E. walterhorni, E. planatoflavum und R. cinctus besiedelt waren. Flugunfähig. (10./11./15./ 21./26. 4.//24. 5.//7. 6.) Verbreitung der Unterart: Elfenbeinküste, Guinea, Äthiopien, Liberia, Ischad, Sierra Leone. Prothyma leprieurii. — 1 Exemplar. Ein Q kam am 15. 6. ans Licht. K. Werner (mündl. Mitteilung) berichtet, daß er diese Art bisher ebenfalls nur an Lichtquellen fangen konnte. Verbreitung der Art: Westafrika, Senegal, Tschad, NO-Zaire, Äthiopien, Jemen. Euryarthron walterhorni. — 19 Exemplare, im Galeriewald oder dessen Randbereichen. Am Vormittag bis ca. 11.00 Uhr wurden Individuen öfters im galerienahen Savannenbereich gefun- den, die sich später im Waldsaum aufhielten. Das gleiche Verhalten zeigte auch E. planato- flavum. Unter Umständen handelt es sich dabei um einen diurnalen Standortwechsel zur Thermoregulation, wobei sich die Tiere in der Savanne vormittags aufwärmen, bis der Galerie- wald geeignete Bodentemperaturen oder Sonnenflecken aufweist, um dann in den inneren Bereich des Waldes zu wechseln. E. walterhorni ist flugunfähig und hält sich im Wald auf lich- ten Stellen auf, die mit Fallaub bedeckt und lückenhaft mit der Poaceae Oplismenus hirtellus bestanden sind, unter die sich die Tiere bei Gefahr flüchten und dort regungslos verharren. Häufig mit E. planatoflavum, teilweise mit P concinna und R. cinctus vergesellschaftet. (4./11./12./15./21. 4.) Verbreitung der Art: Mali (Koulikoro), Elfenbeinküste (Sirakoro); (Cassola 1983). Euryarthron planatoflavum. — 20 Exemplare, im Galeriewald auf offenen Stellen mit spär- lichem Unterwuchs, z. T. auf kleinen Fußwegen oder Hüttenvorplätzen des Camps im Galerie- wald, seltener im Bereich von lehmig-kiesigen Nilpferdfurten. Ähnliches shuttle-Verhalten wie E. walterhorni. Individuell markierte Tiere konnten über Wochen am gleichen Standort beob- achtet werden, zum Teil auf Flächen von nur wenigen Quadratmetern. Diese Art war meistens mit E. walterhorni und R. cinctus vergesellschaftet, teilweise auch mit P concinna und einmal mit A. interruptum. Flugfähig. (2./4./12./15./21. 4.) Verbreitung der Art: Elfenbeinküste (Dimbroko; Horn 1922), Obervolta (Ouagadougou; Coll. Werner, unveröff.). Dromicoida elegantia. — 3 Exemplare. Leider konnten von dieser interessanten Art nur drei Individuen beobachtet und gefangen werden. Ein © jagte auf einem sandigen Fahrweg im Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 329 Bereich der offen Waldsavanne und flog bei Annáherung in angrenzende grasbestandene Flächen. Ein weiteres © hielt sich am Fuß eines stark zerfallenen Termitenhügels auf, wo es kleine sandfarbene Ameisen jagte, möglicherweise die gleiche Art, die auch C. chrysopyga fraß. Dieses Tier zeigte keine nennenswerten Fluchtreaktionen, so daß man sich bis auf wenige Zentimeter annähern konnte. Das dritte Exemplar, ein 9, kam nachts im Camp ans Licht, d. h. im Übergangsbereich Savanne/Galeriewald. Auf Grund der Tatsache, daß die Daten zu dieser Art äußerst dürftig und uneinheitlich sind, können kaum Aussagen über ihre Lebens- weise getroffen werden, außer daß es sich wahrscheinlich um eine Savannenform handelt. (12./19. 4.//11. 6) Verbreitung der Art: Bisher nur vom Comoé-N. P./Elfenbeinküste bekannt. Elliptica lugubris. — 6 Exemplare, große und häufige Art der offenen Savannenwälder und Alluvialflächen. Die meisten beobachteten Tiere jagten zwischen den Horsten der frisch aus- treibenden Savannengräser auf sandig-lehmigen Freiflächen in voller Besonnung. Wenige Exemplare wurden auch auf sandigen Fahrwegen angetroffen, eines kam an eine Lichtfalle, die sich in der Savanne befand. Flugfähig. Zusammen mit L. /uxerii auftretend und im gleichen Großbiotop wie M. plurinotata und D. elegantia (7./10./12./26. 4.) Verbreitung der Art: Senegal, Guinea, Mali, Sierra Leone, Elfenbeinküste, Togo, Nigeria, Kamerun, Äquato- rialguinea, Zentalafr. Republik, Uganda, Kenia, Zaire. Ropaloteres vittatus. — 1 Exemplar, in der Mittagshitze auf sandigem Fahrweg im Bereich der offenen Savannenwälder, im Randbereich der Lola-Plaine (Alluvialfläche). Flugfähig. (6. 4.) Verbreitung der Art: Senegal, Elfenbeinküste (Lamto-Savanne; Lecordier 1972), Sudan, Zaire, Nigeria, Zentralafr. Republik. Ropaloteres nysa. — 5 Exemplare, bis auf ein Exemplar wurden alle Tiere nachts am Licht im Camp gefangen, das heißt im Übergangsbereich von Galeriewald zu einer Alluvialfläche (Lola-Plaine). 1 Exemplar hielt sich auf einem sandigen Streifen auf, der von dichten Gräsern bestanden war und im Saumbereich des Galeriewaldes liegt. Unter Umständen lebt diese Art im Comoé-N.P. im Randbereich von Galerie und Alluvialflächen bzw. Savannenwáldern. Flugfähig. (13./20./21. 4.) Verbreitung der Art: Senegal, Ghana, Togo, Elfenbeinküste, Sierra Leone, Kamerun, Zentralafr. Republik, Zaire, Uganda, Kenia, Sudan, Äthiopien. Ropaloteres cinctus. — 20 Exemplare. Waldart, die häufig im Galeriewald anzutreffen ist, aber auch in Inselwäldern auftritt. Im Bereich des Camps vor allem auf sandig-lehmigen, halbschattigen kleinen Fußwegen und Hüttenvorplätzen. In gewissem Maß als „kultur- folgend“ zu bezeichnen, da die höchsten Individuendichten im Camp auftraten. R. cinctus bevorzugt offene Stellen des Waldbodens zum Jagen, flüchtet aber in dichte Laubstreu oder auf Blätter der angrenzenden Vegetation bis in ca. 50 cm Höhe. Nachts wurden öfters einzelne Tiere angetroffen, die auf bodenfernen Blättern saßen. Diese Übernachtungsplätze werden wahrscheinlich zum Schutz vor nachtaktiven, vagilen Prädatoren, wie z.B. Megacephala denticollis, Skorpionen, Solifugen und Spitzmäusen, aufgesucht (siehe auch Beobachtungen von individuellen und kommunalen Schlafplätzen einiger Sandlaufkäferarten in Peru: Pearson & Anderson 1985). Individuell markierte Tiere zeigten, daß diese Art ebenso wie E. walter- horni und E. planatoflavum sehr standorttreu ist. R. cinctus trat nur innerhalb der Wald- bereiche auf. Ein in Copula befindliches @ wurde dabei beobachtet, wie es eine ca. 1.5 cm große Ponerine, Paltothyreus tarsatus Forell, fraß (M. ©. Rödel, mündl. Mitteilung). U. Braun, ebenfalls mündl. Mitt., beobachtete R. cinctus einige Male beim Fressen dieser wehrhaften Ameise, konnte aber nicht sehen, ob sie von den Cicindeliden aktiv erbeutet worden waren. Werner (1993b) berichtet, daß er R. cinctus mit „einer großen schwarzen Ameise“ in den Mandibeln antraf. Vergesellschaftet ist R. cinctus im Comoé-N. P. regelmäßig mit E. planato- flavum, scheint aber im Gegensatz zu dieser die etwas offeneren Stellen des Bodens im gemeinsam besiedelten Bereich zu nutzen. Außerdem tritt sie mit E. walterhorni und P concinna auf, wobei diese beiden Arten Bereiche dichter Bodenvegetation bevorzugen. Flug- fähig. Die Serie aus dem Comoé-N.P. enthält hauptsächlich rötlichbraune Exemplare mit cremefarbenen Lateralstreifen der Elytren, helle Dorsalmakel sind sehr variabel bzw. fehlen ganz. Einige Tiere besaßen aber auch eine hellgrüne oder schwarze Grundfärbung. (1./2./ 4./5./10./19./21. 4.) Verbreitung der Art: Guinea, Sierra Leone, Togo, Ghana, Liberia, Elfen- beinküste, Nigeria, Kamerun, Äquatorialguinea, Kongo, Zaire, Gabun (Coll. Fahr), Zentral- afr. Republik, Uganda, Sudan, Äthiopien. 330 J. Fahr Hipparidium interruptum. — 13 Exemplare. Ausgesprochene Waldart, die nur im Vollschat- ten der Galeriewálder anzutreffen war. Bevorzugter Lebensraum sind schmale, luftfeuchte, durch Erosion tief eingeschnittene temporäre Zuflüsse des Comoé (,,Nilpferdfurten“), die im Bereich der Sohle ein wechselndes Bodenmosaik aus Sand, Kies und Schlick aufweisen. Außerdem wurde A. interruptum im überschatteten Uferbereich der Kongo- und Iringo-Mün- dung in den Comoé angetroffen, wo sich die Art auf sandigen und nur spärlich mit Laub bedeckten Flächen aufhielt. Insgesamt drei Individuen wurden auf sandigen Flächen inner- halb der Galerie in weiterer Entfernung vom Fluß beobachtet. A. interruptum ist eine scheue, gewandt fliegende Art, die bei Annäherung über verhältnismäßig lange Strecken (bis ca. 30 m) flüchtet. Dabei werden die leuchtend metallisch-roten Tergite sichtbar. In einem Fall trat A. interruptum zusammen mit C. octoguttata (im Uferbereich der Kongo-/Comoé-Miin- dung) auf. (22. 3.//7./15./18./23./26. 4.) Verbreitung der Art: Senegal, Guinea, Sierra Leone, Liberia, Elfenbeinküste, Nigeria, Kamerun, Zentralafr. Republik, Gabun, Zaire, Kongo, Äquatorialguinea, Angola. Lophyridia fimbriata. — 10 Exemplare. Diese Art wurde nur auf einer ausgedehnten Sand- bank gefunden, die sich zungenförmig vom Ufer in den Comoé erstreckte. Im völlig vege- tationsfreien Bereich der Sandbank trat L. fimbriata in hoher Dichte auf, wobei die erhabenen Stellen in der Mitte der Sandbank, wo der Sand trocken war, gegenüber den nassen Bereichen im Spülsaum bevorzugt wurden. L. fimbriata war an dieser Sandbank mit A. nilotica verge- sellschaftet. Dieses gemeinsame Auftreten der beiden Arten ist von vielen Lokalitäten inner- halb des großen Verbreitungsareals von L. fimbriata und A. nilotica bekannt (z. B. Pearson & Juliano 1991, Werner 1993a). Häufig kommt als weitere Art Chaetodera regalis (Dejean) auf solchen gewässernahen, vegetationslosen Sandflächen vor (Pearson & Mury 1979, Pearson & Juliano 1991, Werner 1993a), diese konnte aber im Comoé-N. P. nicht nachgewiesen werden. Erstnachweis von L. fimbriata für die Elfenbeinküste. (23. 3//1./12. 4.) Verbreitung der Nomi- natform: Sudan, Niger, Athiopien, Kenia, Zentralafr. Republik, Zaire, Kamerun, Senegal, Guinea, Benin, Elfenbeinküste. Lophyra neglecta. — 21 Exemplare. Individuen dieser Art wurden ab Anfang April auf Sandfláchen entlang des Comoé beobachtet. Dabei wurden trockene, nahezu vegetationsfreie und stark besonnte Bereiche bevorzugt, die etwas vom Flußufer entfernt lagen (ähnliche Habitatsansprüche von L. neglecta intermedia in Kenia; Werner 1993a). Mit fortschreitendem Ansteigen des Comoé durch die ersten Niederschläge der Regenzeit und der dadurch beding- ten Verkleinerung der Sandflächen im Flußbett nahm die Population in diesem Biotop immer mehr ab. Ab dem 10. 4. erschien L. neglecta auf einer größeren vegetationslosen und halb- schattigen Sandfläche im Randbereich einer Alluvialfläche (Lola-Plaine) und dem Galerie- wald des Comoé. Im angrenzenden Bereich, der lückenhaft mit Gräsern bestanden war, jagten drei Exemplare von Graphipterus voltae Basilewsky. Zwei Tage später drang L. neglecta in den Galeriewald ein, wo sie dann auf halbschattigen und sandigen Hüttenvorplätzen des Camps häufig anzutreffen war. Anscheinend wechselte diese Art mit dem Verschwinden der Sand- flächen durch den Anstieg des Comoé das Habitat, wobei sich interessanterweise das neue in Bezug auf Temperatur, Beschattung und Vegetation stark vom ersten unterschied. L. neglecta trat auf den Sandflächen, auf denen sie entlang des Comoé gefunden wurde, als einzige Sand- laufkäferart auf, im Bereich des Galeriewaldes wurde je einmal R. nysa und C. octoguttata zusammen mit ihr beobachtet. (4./5./6./10./12./21./22. 4.) Verbreitung der Nominatform: Senegal, Sierra Leone, Guinea, Elfenbeinküste, Togo, Tschad, Ghana, Nigeria, Zaire, Angola, Kenia, Äthiopien. Lophyra luxerii. — 1 Exemplar. Auf besonnter, sandiger Fahrspur im Bereich der offenen Waldsavanne. Weitere Exemplare dieser Art wurden ebenfalls in der Savanne angetroffen, wo sie sich meist auf kleinen sandigen Bereichen zwischen lückenhaften Horsten von Gräsern aufhielten. Vergesellschaftet mit E. lugubris und im gleichen Großbiotop wie R. vittatus, M. plurinotata und D. elegantia sowie C. chrysopyga. (22. 4.) Verbreitung der Art: Senegal, Guinea Bissau, Guinea, Sierra Leone, Elfenbeinküste, Ghana, Togo, Nigeria, Kamerun, Zentralafr. Republik, Gabun, Zaire, Uganda, Sudan, Kenia, Äthiopien. Habrodera nilotica. — 7 Exemplare. Gleicher Fundort wie L. fimbriata. Im Unterschied zu dieser Art hielt sich H. nilotica eher am Spülsaum bzw. durchfeuchteten Bereich der Sandbank Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinkúste 331 im Comoé auf. Eine identische Biotopbeschreibung findet sich bei Rensch (1957) für ein Vor- kommen von A. nilotica in Oberágypten. Es war zu beobachten, wie einzelne Tiere bei Flucht und zur Nahrungssuche kurze Strecken ins Wasser liefen. Eine ähnliche Trennung in Mikro- habitate wie bei A. nilotica und L. fimbriata untersuchten Schultz & Hadley (1987) bei Cicindela oregona Leconte und C. franquebarica (Herbst), deren Habitatswahlverhalten an einem sandigen Flußufer in Arizona/USA beobachtet wurde. In diesem Fall bevorzugt C. tranquebarica, analog zu L. fimbriata, die höher gelegenen und trockenen Sandbereiche, während C. oregona, analog zu A. nilotica, die durchfeuchteten ufernahen Bereiche aufsucht. In ökophysiologischen Experimenten zeigten sie, daß C. tranquebarica sowohl eine höhere Letaltemperatur bei 0 % relat. Luftfeuchte (46°—47°C) als auch geringere Wasserverlustraten als C. oregona aufwies, die bei 39°—43 °C in trockener Luft starb. Unter Umständen liegt bei L. fimbriata und A. nilotica ein ähnlicher Fall von physiologisch bedingter Mikrohabitats- trennung wie im oben geschilderten vor. Erstnachweis für die Elfenbeinküste. (23. 3.//1./12. 4.) Verbreitung der Nominatform: Sierra Leone, Elfenbeinküste, Togo, Nigeria, Zentralafr. Republik, Kongo, Zaire, Sudan, Äquatorialguinea, Ägypten, Äthiopien, Kenia, Mozambique, Zimbabwe, Südafrika. Cylindera lutaria. — 18 Exemplare. Dies ist eine von vier dunkel gefärbten, kleinen Cicinde- lidenarten, die im Comoé-N. P. auf Schlickflächen in Gewässernähe oder auf ausgetrockneten Savannentümpeln vorkommen. C. /utaria wurde an drei verschiedenen, jeweils ca. 5 km aus- einanderliegenden Savannentümpeln angetroffen, von denen zwei ausgetrocknet waren, der Schlick aber noch feucht und einer mit Wasser gefüllt war, an dem die Art im schlickigen Uferbereich jagte. C. /utaria wurde nur im Bereich dieser ephemeren, stehenden Savannen- gewässer, nicht aber auf Schlickflächen des Comoé und seiner Zuläufe gefunden. Vergesell- schaftet war diese Art mit C. octoguttata, M. flavidens und M. melancholica. Erstnachweis für die Elfenbeinküste. (18./21./23. 4.) Verbreitung der Art: Guinea Bissau, Elfenbeinküste, Uganda, Sudan, Zaire, Zentralafr. Republik, Angola, Südafrika. Cylindera octoguttata. — 29 Exemplare. Kleinste der nachgewiesenen Arten. C. octoguttata besiedelt wie die vorhergehende Cylindera ebenfalls dunklen, feuchten Schlick als Substrat, zeigt aber eine größere ökologische Plastizität, da sie nicht nur im Bereich von stehenden bzw. ausgetrockneten Savannengewässern vorkommt, sondern auch häufig im schlickigen Spül- saum des Comoé und seiner Zuflüsse zu finden war. Insgesamt drei Exemplare wurden im Untersuchungszeitraum in einem abweichenden Biotop gefunden: Sie hielten sich im halb- schattigen Bereich der sandigen Fußwege und Hüttenvorplätze des Camps innerhalb des Gale- riewaldes auf (7.4.: 9, 18. 4.: Q, 21. 4.: ©). Macfie (1922) berichtet, daß er in Accra (Ghana) häufig C. octoguttata dabei beobachtet hat, wie sie im Uferbereich von Pfützen und Tümpeln nach Anopheles costalis- und Culiciden-Larven „fischte“, indem die Tiere mit den Beinen im Wasser standen und beim Fangen der Beute mit dem Kopf ins Wasser tauchten. C. octoguttata trat zusammen mit C. lutaria, M. flavidens und M. melancholica auf. (27. 3.//3./6./7./8./ 12./18./21./23. 4.) Verbreitung der Art: Senegal, Sierra Leone, Elfenbeinküste, Liberia, Guinea, Ghana, Togo, Nigeria, Kamerun, Zentralafr. Republik, Mali, Benin, Zaire, Gabun, Sudan, Äthiopien, Äquatorialguinea, Kongo, Angola, Namibia, Südafrika, Fernando Poo. Mpyriochile flavidens. — 14 Exemplare. Ebenfalls wie C. /utaria und C. octoguttata eine kleine, dunkel gefärbte Art, die nur an zwei relativ großen Savannentümpeln im Comoé-N. P. gefunden wurde. In einem Fall hielt sie sich im feuchten Schlickbereich des Tümpelufers auf, im anderen auf einer ca. 100 m? großen Schlammfläche, die zu diesem Zeitpunkt kein offenes Wasser mehr besaß. Ähnlich wie C. /utaria war auch M. flavidens nur im Bereich dieser stehenden, ephemeren Savannengewässer zu finden. Die © von M. flavidens besitzen im Unterschied zu den Q auffällig weiß gefärbte Mandibeln, so daß sie sich selbst für den Betrachter deutlich vom dunklen Untergrund abheben. Pearson (1988) beschreibt ähnliche Verhältnisse im Falle von HA. xantophilum W. Horn in Gabun (Makokou). Er beobachtete, wie die © vor potentiellen Paarungspartnern stehenblieben und ihre Mandibeln in schneller Folge öffneten und schlossen. Mit C. lutaria, C. octoguttata und M. melancholica und in einem Fall auch mit M. fastidiosa vicina vergesellschaftet. Erstnachweis für die Elfenbeinküste. (8./21./23. 4.//21. 5.) Verbreitung der Nominatform: Guinea Bissau, Guinea, Sierra Leone, Elfenbeinküste, Obervolta, Togo, Nigeria, Kamerun, Zentralafr. Republik, Zaire, Uganda, Sudan, Äthiopien. 332 TAR Myriochile fastidiosa vicina. — 4 Exemplare. M. O. Ródel sammelte drei Individuen am Licht der Campbeleuchtung im Übergangsbereich Alluvialfláche/Galeriewald. Ein Exemplar wurde zusammen mit M. flavidens auf dunklem Schlick einer vegetationslosen Pfütze gefan- gen. In Äthiopien fand K. Werner (mündl. Mitt.) diese Art auf dünn mit Gras bestandenem, lehmigem Boden. (21. 5.) Verbreitung der Unterart: Mauretanien, Senegal, Guinea, Sierra Leone, Liberia, Elfenbeinküste, Mali, Obervolta, Ghana, Kamerun, Niger, Tschad, Sudan, Äthiopien, Uganda, Tansania, Zaire, Zentralafr. Republik, Kongo, Äquatorialguinea, Angola, Namibia, Malawi, Zimbabwe. Mpyriochile plurinotata. — 9 Exemplare. Als einzige der kleinen, dunkel gefärbten Sandlauf- käferarten wurde M. plurinotata nicht auf feuchtem, meist in Gewässernähe befindlichem Substrat gefunden, sondern trat immer nur in Einzelexemplaren verstreut auf sandigen Flächen, z. B. Fahrwegen, auf. Vier Exemplare wurden nachts am Licht im Randbereich von Galeriewald und Savanne gefangen. Die tagsüber beobachteten Tiere hielten sich vornehmlich im Halbschatten der lückig stehenden Savannenbäume auf. Im gleichen Großbiotop wie E. lugubris, L. luxerii und D. elegantia. (11./20./21. 4.//10./21. 5.) Verbreitung der Art: Maure- tanien, Senegal, Mali, Guinea Bissau, Sierra Leone, Elfenbeinküste, Obervolta, Niger, Nigeria, Zentralafr. Republik. Myriochile melancholica. — 40 Exemplare. Größte Art, die im Comoé-N. P. schlammiges Substrat bewohnt. Überall an Gewässerrändern anzutreffen, wobei anscheinend stehende Savannengewässer bevorzugt werden. Da sich diese erst mit Einsetzen der Regenzeit bilden, war M. melancholica zuerst im Bereich des Comoé und kleiner Zuflüsse zu finden, später trat sie als dominante Art vor allem an etwas größeren Savannentümpeln im schlickigen Ufer- bereich auf. Vergesellschaftet mit C. lutaria, C. octoguttata und M. flavidens, in einem Fall mit M. fastidiosa vicina. Das große Verbreitungsgebiet, das von Südeuropa bis Indien reicht und ganz Afrika einschließt, wobei auch ozeanische Inseln besiedelt sind (z. B. Capverden, Sao Tomé (Coll. Fahr), Seychellen, Comoren, Mauritius etc.), zeugt von der Kolonisations- fähigkeit dieser Art. Insofern scheint sie auch in kleinerem Maßstab dafür prädestiniert zu sein, unvorhersagbare Biotope wie die ephemeren Savannentümpel schnell und erfolgreich zu besiedeln. Solche schlickigen Substrate in Gewässernähe sind besonders zu Beginn der Regen- zeit extrem kurzlebig, da der Boden noch wenig durchfeuchtet ist und andererseits hohe Tem- peraturen und intensive Sonneneinstrahlung vorherrschen, so daß in Abhängigkeit von den Regenfällen geeignete Biotope innerhalb von Tagen bis Stunden entstehen und wieder ver- schwinden. Sucht man nach geeigneten, d. h. feuchten Schlickflächen, so ist M. melancholica fast immer in großer Individuenzahl anzutreffen. (22. 3.//3./8./18./19./23. 4.//21. 5.) Verbrei- tung der Art: Siehe oben. Cratohaerea chrysopyga. — 9 Exemplare. Diese auffällig metallisch blau bis grün gefärbte Art war nur auf den bis zu 5 m hohen Termitenbauten von Macrotermes bellicosus Smeath- man zu finden. Auffälligerweise waren immer nur bestimmte Termitenhügel besetzt, wobei es keinen Unterschied machte, ob die Bauten verlassen oder bewohnt waren. Direkt in der Umgebung befindliche Hügel, die in Bezug auf Besonnung, Vegetation, umgebenden Boden- grund etc. vergleichbar waren, wurden nie besiedelt, während ganz bestimmte Termitenbauten nach dem Abfangen von einzelnen C. chrysopyga meist nach ein bis zwei Tagen erneut besetzt waren. Bis auf zwei Fälle, in denen einmal ein Pärchen in Copula und ein andermal ein getrennt laufendes Pärchen auf demselben Hügel gleichzeitig angetroffen wurde, kam diese Art immer nur in Einzelexemplaren auf dem jeweiligen Termitenbau vor. Ebenso waren keine geschlechtsspezifischen Unterschiede festzustellen, d.h. sowohl © als auch Q bewohnten dieses für Cicindeliden ungewöhnliche Biotop. Nach dem Abfangen von z.B. einem © erschien auf dem gleichen Hügel wenige Tage später ein o& oder Q bzw. vice versa. Somit scheint es sich nicht um temporäre Treffpunkte zur Partnerfindung zu handeln, da ungestörte Individuen über Tage auf demselben Termitenbau verweilten und dort beim Jagen beobachtet wurden. Horn (1921) berichtet, daß sich diese Art in Joko/Kamerun angeblich in der Nähe unterirdischer Termitenbauten aufhält. Weiterhin ist es bemerkenswert, daß C. chrysopyga nicht wie andere Sandlaufkäferarten ständig umherläuft, sondern z.T. bis zu 25 min. regungs- los sitzenbleibt. Bei Annäherung fliegen die Tiere relativ spät auf und landen meist auf der entgegengesetzten Seite des Termitenhügels. Bei größerer Störung fliegen sie entweder vom Hügel fort oder flüchten im Falle von verlassenen Bauten in aufgebrochene Termitengänge. Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 383 Es konnten oft Goldwespen (Chrysididae) beobachtet werden, wie sie langsam die Termiten- bauten fliegend absuchten. Die Größe und Farbe dieser Tiere ähnelt sehr stark der von C. chrysopyga. Ob zwischen diesen beiden Arten irgendeine Form von Wechselbeziehung besteht, konnte nicht festgestellt werden. Wie oben schon erwähnt, verhält sich C. chrysopyga auf den Termitenhügeln sehr ruhig, und nur wenige Male wurde beobachtet, wie die Tiere am Fuß ihres Hügel jagten. Die Beute bestand aus winzigen, 1—1.5 mm langen hellgelben Ameisen. Eine weitere Besonderheit zeichnet diese Art gegenüber anderen Cicindeliden aus: Die Mandibeln der & sind im Durchschnitt um den Faktor 1.3 länger als die der 9, stark sichelfórmig und tragen im Gegensatz zu den Mandibeln der 9 kaum Zähne. Ähnliche Ver- hältnisse liegen bei der Schwesternart C. brunet (Gory) vor, wie an Hand von Museums- exemplaren festgestellt wurde. Unter Umständen dient bei C. chrysopyga dieser für Cicindeli- den relativ ungewöhnliche Sexualdimorphismus dem © zum besseren Festhalten des Q bei der Copulation, da sie sich dabei meist auf den senkrechten Wänden der Termitenbauten befin- den. (12./13./19./26. 4.//20. 5.) Verbreitung der Art: Senegal, Sierra Leone, Guinea, Liberia, Elfenbeinküste, Benin, Togo, Kamerun, Zentalafr. Republik, Zaire, Uganda, Kenia. Artenvergesellschaftungen und Koexistenz Zur Untersuchung, ob sich interspezifische Konkurrenz in Form einer Merkmals- verschiebung auswirkt (siehe Einleitung), wurden sämtliche Arten in Bezug auf Körper- und Mandibellänge vermessen (siehe Material und Methoden). Die Regres- sionsgerade (Abb. 5) zeigt, daß Körper- und Mandibellängen linear und positiv mit- einander korreliert sind. Die einzigen abweichenden Werte sind die von L. fimbriata und den © von C. chrysopyga. Mögliche Erklärungen dafür siehe Beschreibung der Biologie von C. chrysopyga und für L. fimbriata siehe unter „4. Comoé-Sandbánke*. Pearson & Mury (1979) zeigten in Wahlexperimenten, in denen verschiedene Cicin- delidenarten zwischen unterschiedlich großer Beute wählen konnten, daß die Mandi- y = - 0,0618 + 0,2039 x 12 = 0,926 L. fimbriata C. chrysopyga © Mittelwert Mandibellänge [mm] D i) 0 5 10 15 20 25 30 Mittelwert Körperlänge [mm] Abb. 5: Korrelation der Körper- und Mandibellängen von 23 Cicindelidenarten (für C. chry- sopyga sind die Werte nach Geschlecht getrennt aufgetragen). Senkrechte Linien geben die Standardabweichung an. 334 Ja hin: bellángen der jeweiligen Arten direkt mit der mittleren bevorzugten Beutegröße kor- reliert sind. Auf Grund verschiedener Mandibellängen könnten somit syntope Sand- laufkäferarten unterschiedliche Beutespektren nutzen und die interspezifische Nah- rungskonkurrenz vermindern. Dieser Ansatz soll im Folgenden für die jeweiligen Biotope untersucht werden. Wichtige Zusatzinformationen wie Beuteabundanz und Größenspektrum konnten auf Grund des beschränkten Untersuchungszeitraumes leider nicht aufgenommen werden. Diese wären für eine genaue Analyse unabdingbar, und so bleiben einige Aussagen spekulativ oder fußen auf nicht-quantitativen Beobachtungen. 1. Galeriewald Im Bereich des Galeriewaldes traten fünf Arten auf: P concinna, E. walterhorni, E. planatoflavum, H. interruptum und R. cinctus. FaBt man diese zusammen und betrachtet ihre Mandibelverháltnisse (Abb. 6; Tab. 2), so scheinen sie nicht der Vor- aussage einer Merkmalsdivergenz zu folgen. Regelmäßig miteinander vergesellschaf- tet waren allerdings nur E. walterhorni, E. planatoflavum und R. cinctus, während P concinna und A. interruptum meist ohne unmittelbare Begleitarten auftraten (siehe oben, „Bemerkungen zur Biologie“ dieser Arten). E. walterhorni jagte haupt- sächlich im Bereich dichterer Bodenvegetation (vor allem unter den breiten Blättern des Grases O. hirtellus), E. planatoflavum und R. cinctus bevorzugten dagegen offe- nere Bereiche des Waldbodens. In diesem Zusammenhang ist auch zu bemerken, daß die beiden letztgenannten Arten im Gegensatz zu der ersteren flugfähig sind und daher unterschiedliche Strategien der Prädatorenvermeidung besitzen. Betrachtet man die Arten, die die größte Nischenüberlappung aufweisen, nämlich R. cinctus und E. planatoflavum, so zeigt sich, daß sie sich in Bezug auf ihre Mandibellängen deutlich unterscheiden (Verhältnis von 1.32; Tab. 2). Auffällig ist auch die Tatsache, daß A. interruptum und R. cinctus, die an geeigneten Stellen in hoher Dichte auftra- ten und deren Imagines eine sehr ähnliche Lebensweise führen, nie vergesellschaftet waren und andererseits fast identische Mandibellängen besitzen. Tabelle 2: Mandibelverhältnisse für Cicindelidenarten (n = Anzahl der vermessenen Indivi- duen; Mittelwert der > der Galeriewälder. eu concinna (n= 9292238 (n= 19; 2.47) E. planatoflavum (1ME205251) H. interruptun (n=313:93.25) R. cinctus (120332) hi . Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 335 2. Savanne In der offenen Waldsavanne wurden fünf Arten angetroffen: M. plurinotata, D. elegantia, L. luxerii, R. vittatus und E. lugubris. Diese Arten bewohnten den spärlich bewachsenen Boden dieses Großbiotops, während C. chrysopyga ausschließlich auf den Termitenhtigeln von M. bellicosus zu finden war und C. /utaria, C. octoguttata, M. flavidens sowie M. melancholica an Gewässerränder innerhalb der Savannen gebunden waren (siehe „3. Savannengewässer/Schlickufer“) und offensichtlich nicht in Konkurrenz mit den erstgenannten Arten treten konnten. Daher werden nur diese als Artengemeinschaft behandelt. D. elegantia wurde lediglich in drei Exemplaren und in äußerst verschiedenen Kleinbiotopen beobachtet. Auf Grund der Seltenheit Mandibellange [mm] P. concinna E. walterhorni E. planatoflavum H. interruptum R. cinctus Abb. 6: Mittlere Mandibellängen (Punkte) und Standardabweichung (senkrechte Linien) für Cicindelidenarten der Galeriewälder. Mandibellänge [mm] _ D w LS U o D. elegantia L. luxerii R. vittatus E. lugubris 3 3 ° E 3 a = Abb. 7: Mittlere Mandibellängen (Punkte) und Standardabweichung (senkrechte Linien) für Cicindelidenarten der Baumsavanne. 336 J. Fahr dieser Art im Untersuchungszeitraum und der uneinheitlichen Fundumstände, die keine Zuordnung im Sinne einer Artenvergesellschaftung zulassen, wird sie im Folgenden ausgeschlossen. Die verbleibenden vier Arten, die in ihrer Imaginalbiologie starke Nischenüber- lappung zeigen und sich daher im Sinne einer Konkurrenzvermeidung in mindestens einem Parameter unterscheiden sollten, scheinen in Bezug auf die Mandibellängen den Voraussagen einer Hutchinson-ratio zu folgen: M. plurinotata, L. luxerii, R. vittatus und E. lugubris weisen Mandibelverhältnisse von = 1.24 auf (Tab. 3, Abb. 7). Besonders häufig vergesellschaftet waren L. /uxerii und E. lugubris, die einen beson- ders hohen Wert (1.93; Tab. 3) aufweisen. Die z. T. geringen Stichproben vermessener Individuen lassen allerdings keine Statistik zu. 3. Savannengewässer/Schlickufer Für die dunkel gefärbten, schlammigen Ufer der Savannentümpel war die Tatsache charakteristisch, daß fast immer unterschiedliche Artenkombinationen anzutreffen waren. Dies steht in deutlichem Kontrast zu den anderen Habitaten, die durch die Voraussagbarkeit der Artenvergesellschaftungen ausgezeichnet waren. Von allen untersuchten Lebensräumen, die von Sandlaufkäfern besiedelt wurden, ist dieses Habitat durch seine Kurzlebigkeit und Unvorhersagbarkeit gekennzeichnet. Beson- Tabelle 3: Mandibelverhältnisse für Cicindelidenarten (n = Anzahl der vermessenen Indivi- duen; Mittelwert der Mandibellänge) der Baumsavanne. D. elegantia (3371265) L. luxerii (m=—1:71882) R. vittatus (n=11572:83) E. lugubris 0653,52) Tabelle 4: Individuenzahl von drei Cicindelidenarten, die an zwei folgenden Tagen auf einer Schlickfläche quantitativ abgefangen wurden. A uam [7 Bee EITHER ED oma | | Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 337 ders zu Beginn der Regenzeit füllen sich nach Niederschlägen Senken und Mulden mit Wasser, um dann häufig im Verlauf der folgenden Tage durch die starke Sonnen- einstrahlung wieder auszutrocknen. Die von C. /utaria, C. octoguttata, M. flavidens und M. melancholica besiedelten Bereiche, schlammige Gewásserránder oder größere Schlickflächen, verändern sich demzufolge von Tag zu Tag erheblich. Zur Unter- suchung, ob die Artenkombinationen auf Grund der Unvorhersagbarkeit des Habi- tates zufällig zustandekommen, wurden auf einer ca. 20 m? großen Schlickfläche, die nach dem Austrocknen eines Savannentümpels entstanden war, sämtliche Cicindeliden abgefangen. Zu diesem Zeitpunkt war nur noch eine kleine Pfütze vor- handen. Am nächsten Tag wurde der quantitative Abfang wiederholt (Tab. 4). Am dritten Tag war die Oberfläche des Schlicks trocken und keine weiteren Cicindeliden anzutreffen. Durch die Schwierigkeit, eine sowohl vom Zeitpunkt als auch von der Größe (Übersichtlichkeit) geeignete Fläche zu finden, war diese Untersuchung daher auf diese beiden Abfänge beschränkt. An beiden Tagen war M. melancholica annähernd gleich häufig vertreten. C. /uta- ria und C. octoguttata waren dagegen am ersten Tag nur in Einzelexemplaren vorhan- den und fehlten am zweiten Tag vollständig. Dieses Ergebnis befindet sich im Ein- klang mit der Beobachtung, daß M. melancholica an fast allen untersuchten Schlick- flächen auftrat und oft die dominierende Art war. C. octoguttata zeigte ein ähnliches Verbreitungsmuster, wobei sie im Gegensatz zu den anderen schlickbewohnenden Arten auch an den Ufern von Fließgewässern zu finden war. C. lutaria und M. flavi- dens dagegen konnten nur im Bereich stehender oder ausgetrockneter Savannen- tümpel nachgewiesen werden. Außerdem war auffällig, daß die beiden letzteren Arten meist an solchen Biotopen auftraten, die schon eine gewisse Zeit bestanden hatten. Daher stellt sich die Frage, ob es sich bei den jeweils vorgefundenen Arten- kombinationen (es traten sowohl alle vier hier behandelten Arten als auch unter- schiedliche Kombinationen gemeinsam auf) um Sukzessionsstadien zum jeweiligen Zeitpunkt handelt oder ob diese durch zufällige Besiedelungsereignisse jeweils stochastische Gemeinschaften darstellen. Zur Klärung dieser Frage wäre jedoch eine größere Anzahl von Experimenten notwendig. M. melancholica und in geringerem Maße C. octoguttata scheinen in jedem Fall verbreitungsstarke Arten zu sein, was man vor allem an Hand der zoogeographischen Verbreitung von M. melancholica sehen kann (siehe oben, „Bemerkungen zur Biologie“ dieser Art), und daher für eine schnelle und erfolgreiche Besiedelung solcher ephemeren Biotope prädestiniert. Betrachtet man die Mandibelverhältnisse (Abb. 8; Tab. 5), so liegen alle Arten in einem sehr engen Bereich. Dies kann mehrere Gründe haben: Zum einen besteht die Möglichkeit, daß das Größenspektrum der Beutetiere sehr eng ist und dadurch bedingt keine Nischendifferenzierung in diesem Parameter möglich ist. Für diese Annahme gibt es auf Grund von Freilanduntersuchungen an vergleichbaren Gewäs- serrändern in Arizona/USA einige Hinweise (Pearson & Mury 1979, Pearson & Knisley 1985). Dort war sowohl die Beuteabundanz als auch das Größenspektrum im Vergleich zu anderen von Cicindeliden besiedelten Biotopen stark eingeschränkt. Beobachtungen dieser Art konnten auch im Comoé-N.P., allerdings ohne quantita- tive Daten (siehe Einleitung), gemacht werden. Zum anderen könnte es sein, daß Arten, die auf solche ephemeren Biotope spezia- lisiert sind, ihre Beute in sehr kurzer Zeit möglichst effektiv nutzen müssen. Falls 338 J Bahr diese unvorhersagbare Ressource ein geringes Größenspektrum aufweist, wären wie im obigen Fall die Arten auf dieses festgelegt. Interspezifische Konkurrenz könnte in diesem Fall unter Umständen durch die stochastische Besiedelung dieser Biotope herabgesetzt sein. 4. Comoé-Sandbánke Auf den Sandbänken des Comoé wurden drei Arten angetroffen: L. neglecta, H. nilo- tica und L. fimbriata. Interessanterweise wurde L. neglecta nie mit A. nilotica und L. fimbriata zusammen vorgefunden, sondern immer nur alleine: L. neglecta und H. nilotica besitzen fast identische Mandibellángen, und falls im Bereich der Sand- bänke Nahrungskonkurrenz eine starke Rolle spielt, sollte man auch diese Arten- kombination nicht antreffen. Insofern scheinen diese Arten durch unterschiedliche Mikrohabitatswahl (siehe oben, „Bemerkungen zur Biologie“) separiert zu sein. H. nilotica und L. fimbriata sind dagegen eine sehr stabile Artengemeinschaft über ihr großes Verbreitungsgebiet in Afrika hinweg (z. B. Pearson & Mury 1979, Pearson & Juliano 1991, Werner 1993a, b). Betrachtet man Abb. 9, so fällt der große Unter- schied der Mandibellängen zwischen diesen beiden Arten auf. Ihre Hutchinson-ratio (1.98; Tab. 6) würde daher noch eine Art mittelgroßer Mandibellänge „zulassen“. Tatsächlich tritt mit diesen beiden Arten häufig Chaetodera regalis auf, die eine Mandibellänge von 3.1—3.15 mm besitzt, aber interessanterweise eine etwas größere Körperlänge (ca. 16 mm) als L. fimbriata (ca. 14.5 mm) hat (Pearson & Mury 1979). Damit würde C. regalis ziemlich genau auf der Regressionsgeraden von Abb. 5 liegen, während L. fimbriata bei ähnlicher Körperlänge deutlich über dieser liegt. Inwieweit diese Mandibelverhältnisse tatsächlich eine Antwort auf interspezifische Konkurrenz darstellen, bleibt experimentell zu überprüfen. In diesem Zusammenhang sei auf die Anmerkungen zu L. neglecta und A. nilotica verwiesen. Tabelle 5: Mandibelverhältnisse für Cicindelidenarten (n = Anzahl der vermessenen Indivi- duen; Mittelwert der Mandibellänge) der Savannengewässer/Schlickufer. O oem [MA TB mem [econ amo aa | | u | ne [imamate weezy | | ENE Tabelle 6: Mandibelverhältnisse für Cicindelidenarten (n = Anzahl der vermessenen Indivi- duen; Mittelwert der Mandibellánge) der Comoé-Sandbánke. L. neglecta (n=20; 1.85) H. nilotica (n=7; 1.94) L. fimbriata (n= 10; 3.85) Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 339 Schlußfolgerungen Die post hoc durchgeführte Analyse der Artengesellschaften unterschiedlicher Bio- tope an Hand von Mandibellängen der einzelnen Cicindelidenarten ersetzt nicht eine direkte Untersuchung mit vergleichenden oder experimentellen Ansätzen zur Über- prüfung der Frage, ob diese Artengemeinschaften durch interspezifische Konkurrenz strukturiert sind. Trotzdem scheinen die Ergebnisse dieser wie auch anderer Arbeiten (Pearson & Mury 1979, Pearson 1980, Pearson & Juliano 1991, Ganeshaiah & Bela- vadi 1986, Schultz & Hadley 1987) starke Hinweise darauf zu geben, daß es sich im Falle von Sandlaufkäfern um eine Gilde handelt, die stark durch interspezifische Mandibellänge [mm] C. lutaria C. octoguttata M. flavidens M melancholica Abb. 8: Mittlere Mandibellängen (Punkte) und Standardabweichung (senkrechte Linien) für Cicindelidenarten der Savannengewässer/Schlickufer. Mandibellänge [mm] Sg S 3) & el o (= S =) H. nilotica L. fimbriata Abb. 9: Mittlere Mandibellangen (Punkte) und Standardabweichung (senkrechte Linien) fiir Cicindelidenarten der Comoé-Sandbánke. 340 II Konkurrenz geprägt ist. Im Falle der vorliegenden Arbeit konnten nur die Mandibel- verhältnisse quantitativ ausgewertet werden. Viele der wichtigen Parameter, wie z. B. Nahrung, Temperatur, Feuchtigkeitsansprüche, Larval- und Fortpflanzungsbiologie, sowie Prädatoren und Phänologie der jeweiligen Arten konnten nicht oder nur sehr ungenügend erfaßt werden (siehe Einleitung). Da es sich bei Cicindeliden um eine Gruppe handelt, die sowohl im natürlichen Lebensraum relativ leicht zu beobachten als auch experimentell zu manipulieren ist, wäre es wünschenswert, weitere Unter- suchungen zu diesem Fragenkomplex an ihnen durchzuführen. Außerdem sei darauf hingewiesen, daß Sandlaufkäfer im Bereich des Naturschut- zes durch die meist spezialisierten Habitatsansprüche gut als Charakterarten zur Beurteilung von Schutzgebieten verwendet werden könnten. Eine ausführliche Dar- stellung dieses Aspekts findet sich bei Pearson & Cassola (1992). Danksagung Ich danke Jürgen Wiesner (Wolfsburg), Wolfgang Lorenz (Tutzing) und Karl Werner (Peiting) für die Hilfe bei der Bestimmung der Arten sowie für Literaturhinweise. Die Zoologische Staatssammlung München, insbesondere Dr. Scherer, gewährte mir uneingeschränkten Zugang zur Sammlung der ZSM und zur Sammlung Frey. Dank gebührt Mark Oliver Rödel (Tübingen) für die Sammlung einiger Cicindeliden nach meiner Abfahrt sowie Marco Spieler (Würzburg) für die Hilfe bei der Anfertigung von Graphiken und zahlreiche Anregungen. Prof. Dr. K. E. Linsenmair (Würzburg) danke ich für das Ermöglichen dieser Arbeit im Comoé-Nationalpark, Literaturhinweise sowie die Durchsicht des Manuskriptes. Zusammenfassung Im Comoé-Nationalpark wurden 23 Cicindelidenarten auf einer Fläche von ca. 10 km? nach- gewiesen. Cylindera lutaria, Habrodera nilotica, Lophyridia fimbriata und Myriochile flavi- dens waren von der Elfenbeinküste unbekannt. Euryarthron planatoflavum wird zum ersten Mal von Obervolta gemeldet. Eine Art (Dromicoida elegantia) war unbeschrieben (Werner 1995). Das Untersuchungsgebiet zeichnet sich durch einen außergewöhnlichen Artenreichtum aus, der durch das Habitatmosaik des Nationalparks begründet ist. Nur wenige Arten sind zoogeographisch auf die Elfenbeinküste und angrenzende Länder beschränkt, die meisten finden sich auch in Zentral- oder Ostafrika. Mögliche Gründe dafür werden angeführt. Sand- laufkäfer sind Mitglieder einer carnivoren Arthropodengilde, die in Habitaten mit hohen Temperaturen jagen. Durch die extremen Umweltbedingungen kommen nur wenige Organis- men als potentielle Konkurrenten in Frage. Es kann davon ausgegangen werden, daß Cicindeli- den, die syntop vorkommen, auf Grund einer sehr einheitlichen Lebensweise untereinander den stärksten Konkurrenzdruck ausüben. Nach Untersuchung der spezifischen Habitatpräfe- renzen der einzelnen Arten werden die Hutchinson-ratios der Mandibellängen sympatrischer Arten verglichen. Es zeigt sich, daß Arten, die im selben Mikrohabitat auftreten und eine ver- gleichbare Lebensweise haben, im allgemeinen den Voraussagen einer Merkmalsdivergenz folgen, während Arten, die zwar im gleichen Großbiotop leben, aber unterschiedliche Mikro- habitatsansprüche besitzen, häufig überlappende Mandibelverhältnisse aufweisen. Dies wird im Sinne von Strategien zur Konkurrenzverminderung diskutiert. Literaturverzeichnis Basilewsky, P. (1966): Revision des Megacephala d’Afrique (Coleoptera Carabidae Cicinde- linae). — Ann. Mus. Roy. Afr. Centr., Serie IN-08 152: 1—149. Basilewsky, P. (1968): IV. Coleoptera Carabidae. — In: Contributions á la connaisance de la fauna entomologique de la Cóte d'Ivoire (J. Decelle, 1961—1964), I. partie. Ann. Mus. Roy. Afr. Centr., Serie IN-08 165: 27—124. Sandlaufkäfer des Comoé-Nationalparks, Elfenbeinküste 341 Cassola, F. (1983): Studi sui Cicindelidi 35. Una nova specie di Euryarthron Guer., con nota riassuntive sul genere (Coleoptera). — Boll. Soc. Entomol. Ital. 115: 79—85. Dreisig, H. (1980): Daily activity, thermoregulation and water loss in the tiger beetle Cicin- dela hybrida. — Oecologia 44: 376— 389. Dreisig, H. (1981): The rate of predation and its temperature dependence in a tiger beetle, Cicindela hybrida. — Oikos 36: 196 — 202. Eusebi, M. P., L. Favilli & S. Lovari (1989): Some abiotic factors affecting the activity and habitat choice of the tiger beetle Cephalota circumdata leonschaeferi (Cassola) (Cole- optera, Cicindelidae). — Boll. Zool. 56: 143—150. Ganeshaiah,K. N. € V. V. Belavadi (1986): Habitat segregation in four species of adult tiger beetles (Coleoptera: Cicindelidae). — Ecol. Ent. 11: 147—157. Guppy, M., S. Guppy «€ J. Hebrad (1983): Behavior of the riverine beetle Lophyridia dongalensis imperatrix: effect of water availability on thermoregulatory strategy. — Entomol. Exp. Appl. 33: 276—282. Horn, W. (1921): Beitrag zur Faunistik und Lebensweise der Cicindelinae des tropischen Afrika. — Entomol. Bl. Biol. Syst. Káfer 17: 172—184. Hutchinson, G. E. (1959): Homage to Santa Rosalia or Why are there so many kinds of animals? — Am. Nat. 93: 145—159. Knisley, C. B. (1979): Distribution, abundance, and seasonality of tiger beetles (Cicindelidae in the Indiana Dunes region). — Proc. Indiana Acad. Sci. 88: 125—133. Knisley, C. B. (1984): Ecological distribution of tiger beetles (Coleoptera: Cicindelidae) in Colfax County, New Mexico. — Southwestern Nat. 29: 93—104. Knisley, C. B. (1987): Habitats, food resources, and natural enemies of a community of larval Cicindela in southeastern Arizona (Coleoptera: Cicindelidae). — Can. J. Zool. 65: 1191 — 1200. Knisley, C. B. & S. A. Juliano (1988): Survival, development, and size of larval tiger beetles: effects of food and water. — Ecology 69: 1983 —1992. Knisley, C. B.& D. L. Pearson (1984): Biosystematics of larval tiger beetles of the Sulphur Springs Valley, Arizona. Description of species and a review of larval characters for Cicindela (Coleoptera: Cicindelidae). — Trans. Am. Ent. Soc. 110: 465—551. Knisley, C. B., D. L. Reeves & G. T. Stephens (1989): Behavior and development of the wasp Pterobromus rufiventris hyalinatus Krombein (Hymenoptera: Tiphiidae), a parasite of larval tiger beetles (Coleoptera: Cicindelidae). — Proc. Entomol. Soc. Wash. 91: 179—184, Knisley, C. B., T. D. Schultz & T. H. Hasewinkel (1990): Seasonal activity and thermo- regulatory behavior of Cicindela patrulea (Coleoptera: Cicindelidae). — Ann. Entomol. Soc. Am. 83: 911—915. Lecordier, C. (1972): Les Carabiques de la savane de Lamto (Cóte d'Ivoire). — Bull. Inst. Fondam. Afr. Noire Ser. A Sci. Nat. 34: 378—455 Livingstone, D. A. (1975): Late Quaternary climatic change in Africa. — Ann. Rev. Ecol. Syst. 6: 249—280. Macfie, J. W. S. (1922): A note on a beetle which preys on mosquito larvae. — B. Ent. Res. 13: 403. MacNally, R. C. (1988): On the statistical significance of the Hutchinsonian size-ratio para- meter. — Ecology 69: 1974—1982. Morgan, K. R. (1985): Body temperature regulation and terrestrial activity in the ectother- mic beetle Cicindela tranquebarica. — Physiol. Zool. 58: 29—37. Mury Meyer, E. J. (1987): Asymmetric resource use in two syntopic species of larval tiger beetles (Cicindelidae). — Oikos 50: 167—175. Niemelá, J. (1993): Interspecific competition in ground-beetle assemblages (Carabidae): What have we learnt? — Oikos 66: 325—335. Pearson, D. L. (1980): Patterns of limiting similarity in tropical forest tiger beetles (Coleop- tera: Cicindelidae). — Biotropica 12: 195—204. Pearson, D. L. (1984): The function of multiple anti-predator mechanisms in adult tiger beetles (Coleoptera: Cicindelidae). — Ecol. Ent. 10: 65—72. Pearson, D. L. (1988): Biology of tiger beetles. — Annu. Rev. Entomol. 33: 123—147. 342 J. Fahr Pearson, D. L. (1992): Tiger beetles as indicators for biodiversity patterns in Amazonia. — Research & Exploration 8: 116—117. Pearson,D.L.4£J.J. Anderson (1985): Perching heights and nocturnal communal roosts of some tiger beetles (Coleoptera: Cicindelidae) in southeastern Peru. — Biotropica 18: 244—256. Pearson, D. L. & F. Cassola (1992): World-wide species richness patterns of tiger beetles (Coleoptera: Cicindelidae): Indicator taxon for biodiversity and conservation studies. — Conserv. Biol. 6: 376— 381. Pearson, D. L. & K. Ghoparde (1989): Geographical distribution and ecological history of tiger beetles (Coleoptera: Cicindelidae) of the Indian subcontinent. — J. Biogeogr. 16: 333 — 344. Pearson, D. L. & S. A. Juliano (1991): Mandible length ratios as a mechanism for co- occurrence: Evidence from a world-wide comparison of tiger beetle assemblages (Cicinde- lidae). — Oikos 60: 223—233. Pearson, D. L. &S. A. Juliano (1993): Evidence for the influence of historical processes in co-occurrence and diversity of tiger beetle species. — In: Ricklefs, R. E. & D. Schluter, eds.: Species Diversity in Ecological Communities. Historical and Geograhical Perspecti- ves, pp. 194—202. University of Chicago Press, Chicago. Pearson, D. L. & C. B. Knisley (1985): Evidence for food as a limiting resource in the life cycle of tiger beetles (Coleoptera: Cicindelidae). — Oikos 45: 161—168. Pearson, D. L. & R. C. Lederhouse (1987): Thermal ecology and the structure of an assemblage of adult tiger beetle species (Cicindelidae). — Oikos 50: 247—255. Pearson, D. L. & E. J. Mury (1979): Character divergence and convergence among tiger beetles (Coleoptera: Cicindelidae). — Ecology 60: 557—566. Pearson, D. L. & S. L. Stemberger (1980): Competition, body size and the relative energy balance of adult tiger beetles (Coleoptera: Cicindelidae). — Am. Midl. Nat. 104: JUST Porembsky, S. (1991): Beitráge zur Pflanzenwelt des Comoé-Nationalparks (Elfenbein- küste). — Natur u. Museum 121: 61—83. Rensch, B. (1957): Aktivitätsphasen von Cicindela-Arten in klimatisch stark unterschied- | lichen Gebieten. — Zool. Anz. 158: 33—38. Rumpp, N. L. (1977): Tiger beetles of the genus Cicindela in the Sulphur Springs Valley, Arizona, with description of three new subspecies (Cicindelidae-Coleoptera). — Proc. Calif. Acad. Sci. 41: 169—182. Schoener, T. W. (1965): The evolution of bill size differences among sympatric congeneric species of birds. — Evolution 19: 189—213. Schoener, T. W. (1982): The controversy over interspecific oia — Amer. Scientist 7102586995: Schultz, T. D. (1986): Role of structural colours in predator avoidance by tiger beetles of the genus Cicindela (Coleoptera: Cicindelidae). — Bull. Ent. Soc. Amer. 32: 142—146. Schultz, T. D. (1989): Habitat preference and seasonal abundance of eight sympatric species of tiger beetle, genus Cicindela (Coleoptera: Cicindelidae), in Bastrop State Park, Texas. — Southwestern Nat. 34: 468 — 477. Schultz, T. D. & N. F. Hadley (1987): Microhabitat segregation and physiological differen- ces in co-occurring tiger beetle species, Cicindela oregona and Cicindela tranquebarica. — Oecologia 73: 363—370. Schultz, T. D, M. C. Quinlan «€ N. F. Hadley (1992): Preferred body temperature, metabolic physiology, and water balance of adult Cicindela longilabris: a comparison of populations from boreal habitats and climatic refugia. — Physiol. Zool. 65: 226—242. Serrano, A. R. M. (1995): Nouvelle contribution a l'étude des coléoptéres cicindelidés (Coleoptera: Cicindelidae) de Guinée-Bissau. — Bolm. Soc. port. Ent. 156: 69—92. Shelly, T. E. & D. L. Pearson (1978): Size and color discrimination of the robber fly Efferia tricella (Diptera: Asilidae) as a predator on tiger beetles (Coleoptera: Cicindelidae). — Env. Entomol. 7: 790—793. Simberloff, D. (1982): The status of competition theory i in ecology. — Ann. zool. Fennici 19: Al 253) Sandlaufkáfer des Comoé-Nationalparks, Elfenbeinküste 343 Simberloff, D.& W. Boecklen (1981): Santa Rosalia reconsidered: Size ratios and compe- tition. — Evolution 35: 1206—1228. Werner, K. (1993a): Die Sandlaufkáfer Kenias (Coleoptera, Cicindelidae). — Lambillionea 93351716. Werner, K. (1993b): Die Sandlaufkäfer Äthiopiens (Coleoptera, Cicindelidae). — Mitt. Münch. Ent. Ges. 83: 3—38. Werner, K. (1995): Dromicoida Gen. n. from West Africa, with description of a new species (Coleoptera: Cicindelidae). — Koleopterol. Rundsch. 65: 19—22. Wiesner, J. (1992): Verzeichnis der Sandlaufkáfer der Welt. Checklist of the Tiger Beetles of the World. — E. Bauer, Keltern. Wilson, D. S. (1978): Prudent predation: a field study involving three species of tiger beetles. — Oikos 31: 128—136. Wilson, E. O. & D. J. Farish (1973): Predatory behavior in the ant-like wasp Metocha stygia (Say) (Hymenoptera: Tiphiidae). — Anim. Behav. 21: 292—295. J. Fahr, Biozentrum, Zoologie III, Am Hubland, D-97074 Würzburg, Germany. af ir von Ñ er ' \ vil | i 5 ' a E A E el o u y - = = = if j + = N : ha mn a nu y % = = y Mi i N ’ et ade Melee het a ahh, SAE Bd. 47 S. 345— 379 Bonn, September 1998 Coleoptera Histeridae Hetaeriinae: description de nouveaux taxons, designation de lectotypes et notes taxonomiques Nicolas Dégallier Abstract. New data are given for the beetle family Histeridae, subfamily Hetaeriinae. Two collecting methods were fairly productive: a flight interception trap and the flooding of whole antnests. A new cleaning technique was successfully tested for the preparation of genital structures; these are prepared in a basic solution of Genitase proteolytic enzyme. The female genitalia of 12 species of Hetaeriinae are described for the first time. New descriptive, distributional and/or ecological data are provided for 35 species, based on type-specimens, when possible. The following taxa are described as new: Aristonister n. gen.; Colonides collegii guyanensis n. subsp.; Neocolonides howdeni n. gen. n. sp.; Troglo- sternus neoecitonis n. sp. The following new combinations are proposed: Aristomorphus borgmeieri, A. latipes, A. ogloblini, A. perversus, Aristonister sericeus, Hemicolonides parvulus; Latronister breyeri. The following new synonymy is proposed: Oudaimosister Helava in Helava et al. 1985 = Mesynodites Reichardt, 1924. Lectotype and where the case, paralectotypes, are designated for the following species: Aristomorphus borgmeieri, Aristonister sericeus, Cheilister sphaeroides, Clientister henrici, Colonides quadriglumis, Cossyphodister schwarzmaieri, Ecclisister bickhardti, Euxenister caroli, Hemicolonides parvulus, Parasynodites suturacava, Pelatetister pretiosus, Teratosoma amphiphilus, Troglosternus dasypus. The genus Hetaeriarchus Reichensperger is considered as a nomen nudum. Key words. Coleoptera, Histeridae, Hetaeriinae, Neotropics, myrmecophiles, termitophi- les, new synonymies, new combinations, lectotype designations, new genera, new species. Introduction La récente révision générique de la sous-famille des Hetaeriinae, publiée par Helava et al. (1985) n'a malheureusement tenu compte ni des exemplaires-types des taxons révisés (Dégallier 1987), ni des taxons non représentés dans des collections nord- américaines. Ainsi, sur les 103 genres originellement inclus dans cette sous-famille, 26 n’ont pas été révisés. Par ailleurs, ces auteurs ont étudié les genitalia máles de tous les genres révisés a l’exception de 18 d’entre-eux qui n’étaient connus que par des spécimens femelles. Enfin, aucune mention n'a été faite du possible rattachement de deux genres néotropicaux a la sous-famille des Chlamydopsinae, jusque-la composée exclusivement d'éléments indo-australiens (cf. Mazur 1984). Le present travail fournira des compléments aux descriptions (dont l’étude morph- ologique des genitalia femelles de plusieurs genres), désignera des lectotypes (et ‚paralectotypes le cas échéant), proposera de nouvelles synonymies et combinaisons suggérées par l'étude de diverses collections (dont la principale est celle léguée par A. Reichensperger au Laboratoire d’Entomologie du Zoologisches Forschungs- institut & Museum Alexander Koenig a Bonn, Allemagne) et enfin décrira quatre nouveaux taxons. 346 N. Degallier Matériel et Méthodes Préparation des spécimens: Nous avons utilisé tout d'abord la méthode classique du ramollissage dans de l’eau froide ou chaude, suivi de Pextraction des genitalia puis de leur nettoyage apres un séjour dans une solution de potasse a 10 Yo. Une autre methode, publiée recemment par notre ami histeridologue le Dr. P. Kanaar (1990), permet d'obtenir des résultats bien plus satisfaisants, sans dommages possibles, ni pour les pieces nettoyées, ni pour les outils entomologiques, le matériel optique ou les habits, tout en étant beaucoup moins dangereuse pour les yeux et les muqueuses. Les pieces a nettoyer sont placées dans une solution de Genitase (enzyme protéolytique bactérien, distribué par la société Biopraxis: Oegstgeest Pays-Bas) dans de l'eau additionnée de détergent et de carbonate de potassium (ce produit permet d'obtenir un pH optimum compris entre 9,5 et 10). Terminologie: La terminologie morphologique utilisée est celle définie par Wenzel & Dybas (1940) et Helava et al. (1985). Toutefois, les longueurs du pronotum et des parties sternales sont mesurées le long de leur ligne médiane, sans tenir compte des avancées latérales. La longueur de la téte est comprise entre le vertex et le bord antérieur du labre. A notre connaissance, la structure des génitalia femelles n’a encore jamais été décrite en détail chez les Hetaeriinae, ni utilisée pour classer les autres Histeridae. Par analogie avec les pieces génitales máles (cf. Fragoso 1985, pour une discussion sur cette analogie), les sclérites dorsaux et ventraux qui, au repos, forment deux tubes emboftés et téle- scopés a l’interieur du 7eme segment abdominal (= ovipositeur), seront respectivement dénommés tergite et sternite 8, tergite et sternite 9. Le lecteur interesse par la nature exacte et les homologies des sclérites des segments VIII et IX des Coléoptéres pourra consulter la synthese réalisée par Deuve (1988). Chez les femelles de la majorité des sous-familles d’Histeridae (Abraeinae, Chlamydopsinae, Dendrophilinae, Histerinae, Onthophilinae, Saprininae, Tribalinae, etc.), le sternite 8 forme une plaque plus ou moins sclérotisée, sétigere ou non apicalement et prolongée a sa base par des apodémes paires. Le sternite 9 porte distalement des coxites et des styles bien individualisés et souvent setigeres, articulés a leur base sur des apodemes (nommés valviféres par Yélamos 1989) dont le développement est variable (Bickhardt 1916-17: 13—14 et pl. 1 fig. 5; Dégallier 1981; Vomero 1973, 1977; Helava 1978: fig. 9). Entre les bases des coxites est visible un petit sclérite plus ou moins triangulaire et impair qui pourrait correspondre a un reste du segment 10. Chez les Hetaeriinae, ces structures apparaissent modifiées, le plus souvent fortement raccourcies et quelquefois soudées entre elles. Les coxites sont toujours Po de styles sub-apicaux et soudés a leurs apodemes basaux. Bien que seul l’aspect descriptif de ces structures nous intéressera ici, il est utile de signaler que ces modifications des genitalia femelles constituent des synapomorphies complémentaires permettant de définir sans hésitation la sous-famille des Hetaeriinae. Abréviations (en partie selon Arnett, Jr. et al. 1986): AMIC, collection Antonio Marti- nez, Salta, Argentine; BMNH, The Natural History Museum, Londres, UK; CHND, collec- tion de l’auteur, Paris, France; CHYG, collection d’Histérides, Yves Gomy, Fontenay sous Bois, France; CUBO, Carleton University, Department of Biology, Ottawa, Canada; DEIC, Institut fir Pflanzenschutzforschung der Akademie der Landwirtschaftswissen der DDR, Abteilung Taxonomie der Insekten, Bereich Eberswalde Finow, DDR; FIOC, Instituto Oswaldo Cruz, Fundacáo Oswaldo Cruz-FIOCRUZ, Rio de Janeiro, Brésil; FMNH, Field Museum of Natural History, Chicago, USA; MACN, Museo Argentino de Ciencias Naturales “Bernardino Rivadavia”, Buenos Aires, Argentine; ZFMK, Zoologisches Forschungsinstitut & Museum Alexander Koenig, Bonn, RFA; ZMHB, Museum fiir Naturkunde der Humboldt- Universitat, Berlin, RFA; L, longueur du pronotum sur la ligne médiane + longueur des élytres le long de la suture; 1, largeur maximale des élytres pris ensemble; Lm (min.-max.) et Im (min.—max.), moyennes de L et | pour les spécimens mesurés (nombre N), accompagnées des valeurs mini- et maximales; Et. man., imp., etiquette manuscrite, imprimée; ex., exemplaire de sexe inconnu. Nouvelles d’Histeridae 347 Résultats Dans la liste qui suit, les genres sont classés selon l’ordre alphabétique et les especes selon l’ordre chronologique de leurs descriptions originales. Aemulister Reichensperger, 1938 Aemulister: Mazur 1984: 321; Helava et al. 1985: 337. Aemulister borgmeieri Reichensperger, 1938 Aemulister boremeieri: Mazur 1984: 321; Helava et al. 1985: 338. Holotype (par désignation originale): un O étiqueté: “Aemulister borgmeieri Reichensp.” (man. a liséré noir) / “Eciton crassicorne 28 (ou “23” ?).12.35” / Campinas Goyaz, Brasil.” / “Type! Reichensperger” (et. rouge imp. avec liséré noir) / “MUSEUM KOENIG BONN”. Ma 171,8 mm. ZEMK. Descriptioncomplémentaire: Corps (excepte les pattes et antennes) brun rouge, ovale ou oblong, plus ou moins convexe, 0,8 fois aussi large que long. Longueur médiane du prono- tum et des élytres: 2,2 mm. Largeur maximale: 1,8 mm. Tégument (entre la ponctuation) lisse et brillant. Vertex sans tubercule. Front sans stries, sillons, rides ou carénes. Clypeus dans le méme plan ou a peu pres que le front et le labre, ce dernier plus ou moins quadrangulaire. Antennes insérées sous le rebord du front, en avant des yeux. Scape antennaire anguleusement dilaté. Funicule antennaire de sept articles. Massue antennaire en ovale plus ou moins allongé, en partie sclérotisée. Face externe des mandibules non modifiée a la base ou faiblement concave. Pronotum a bord antérieur échancré, normalement convexe, non sculpté ni divisé en aires par des sillons, sans fossettes ni dépressions profondes ou incisions a la base, aux angles antérieurs sans expansion vers l’avant. Disque sans striation supplémentaire ni cótes. Base a surface non modifiée, regulierement convexe, sans impression devant le scutellum. Élytres sans striation supplémentaire entre les stries dorsales. Longueur des sterna (pro-, méso-, méta- et st. abd. 1): 1,0 mm. Prosternum 0,4 fois aussi long que les sterna. Cavités antennaires plus ou moins nettes mais visibles en vue ventrale. Carene prosternale sans processus antérieur. Base du prosternum largement échancrée. Proépisternum avec une fossette ou une aire sétigere. Proépiméron sans fossette ou aire sétigere. Glande propleurale ou latéro-pronotale présente. Lobe prosternal présent (disposition histéromorphe), non subdivisé longitudinalement, a l'apex non modifié ni épaissi. Mésoster- num 0,1 fois aussi long que les sterna, sans fossettes, s'enfoncant en pointe triangulaire dans la base du prosternum. Métasternum 0,35 fois aussi long que les sterna. Méso-métasterna sans bosse lisse. Limite méso-métasternale sans fossette. Premier sternite abdominal 0,15 fois aussi long que les sterna. Tibias sans éperons. Genitalia máles (figs. 33—39): tergite 8 avec une strie antérieure transverse, sans prolonge- ment ventro-postérieur; sternites 8 séparés, avec des disques mais sans soies apicales; tergite 9 avec des apodemes ventraux, des guides internes pour l'édéage et des structures en crochets apicalement; tergite 10 present, membraneux; édéage a base longue et parameres courts. Matériel: Brésil: Campinas Goyaz, 3-1936 (Schwarzmaier), avec Eciton crassicorne, 1 ex. 25mm: |= 1,8 mm. ZEMK Aristomorphus Lewis, 1913 Aristomorphus: Mazur 1984: 303; Helava et al. 1985: 341. Ecitonister: Reichensperger, 1923a: 326, nouvelle synonymie. Ecitonister: Mazur 1984: 310 (partim.). 348 N. Dégallier Parmi tous les genres d’Hetaeriinae connus, seul Ecitonister Reichensperger possede le lobe prosternal excavé antérieurement. Nous avons eu la surprise de retrouver une structure similaire chez l’holotype de Aristomorphus rutilus Lewis, 1913. La combinaison de caracteres qui caractérise ce genre est: — surface dorsale presque plane dans sa moitié antérieure et bombée vers l’arriere; — fronto-clypeus avec trois élévations a ponctuation setigere et une zone médiane concave et lisse; — présence de touffes de trichomes (associées ou non a des fossettes visibles = orifices elandulaires?) pres des angles postérieurs du pronotum, sur la face dorsale, sur les hypomeres, en ligne verticale sur les flancs de la carene prosternale ou sur la limite méso-métasternale; — métasternum extrémement allongé, ler sternite abdominal court. Parmi les cing espéces décrites dans le genre Ecitonister Reichensperger (Mazur, 1984: 310; Helava et al. 1985: 282), quatre sont sans aucun doute congénériques. La cinquiéme, utilisée par Helava et al. (1985: 280) comme représentative du genre, ne peut plus en faire partie, pour les raisons exposées plus loin. L’etude d’exemplaires authentiques des especes en cause permet d'établir le catalogue des especes du genre Aristomorphus de la maniere suivante: — Aristomorphus rutilus Lewis, 1913; — Aristomorphus perversus (Reichensperger, 1923a: 327), nouvelle combinaison; — Aristomorphus borgmeieri (Reichensperger, 1931: 268), nouvelle combinaison; — Aristomorphus ogloblini (Bruch, 1933: 34), nouvelle combinaison; — Aristomorphus latipes (Reichensperger, 1933: 305), nouvelle combinaison; — Ecitonister sericeus Borgmeier, 1948 devient l’espece-type d'un genre nouveau décrit plus loin. Aristomorphus rutilus Lewis, 1913 A. rutilus: Mazur 1984: 303. A. rutilis (err. typogr.): Helava et al. 1985: 342. Holotype (par monotypie): une 9 étiquetée: “Type” (ét. ronde a bordure rouge) / “Corco- vado, Rio de Janeiro G. E. Bryant. 14. V. 1912” (ét. blanche imp. avec une ligne verte au recto) “1913 43” (au verso) / “Aristomorphus rutilus Lewis Type” (et. man.) / “SYNTYPE” (et. ronde a bordure bleue). L = 2,4 mm; | = 1,6 mm. BMNH. Redescription: Forme en ovale allongé, peu convexe (fig. 1); surface lisse et brillante, avec de longues soies éparses, alignées le long de certaines stries ou regroupées en aires. Téte 1,1 fois aussi large que longue, non carénée latéralement; front dépourvu de strie mais montrant, de chaque cóté au-dessus et en avant des yeux, une aire convexe densément ponctuée et longuement setigere; clypeus renflé avec une aire médiane également a ponctua- tion setigere; vertex et front concaves et lisses au milieu; mandibules a pointe tres acérée, non modifiées a leur base; labre quadrangulaire, 2 fois aussi large que long, surface lisse, marge apicale avec une pointe au milieu, une (ou deux) forte(s) soie(s) a chaque angle basalement; massue antennaire ovale, sclérotisée dorsalement, ventralement et sur sa face externe. E O x sl + = 5 11 IR za 0 ZA GG u =» ZA NEN ¡O 350 N. Dégallier Pronotum 1,5 fois aussi large que long; marges latérales presque droites sur leurs 3/4 basaux, convergentes vers l’avant; angles anterieurs tronqués obliquement; échancrure antérieure large et arquée; base en demi-cercle régulier, angles arrondis; surface peu convexe sur le disque et le long des bords latéraux, deux légeres dépressions en arriere des yeux, deux autres plus prononcées en dedans des angles postérieurs, chacune de ces dernieres prolongée d'un sillon vers l’avant et l’exterieur; ponctuation peu dense et extr&mement fine, excepté une bande médiane de quelques points plus gros sur le tiers antérieur; striation comme dans la description originale (strie marginale entiere, strie laterale réduite a sa portion apicale, une strie arquée reliant les fossettes basales); une rangée de soies dressées a l’exterieur de la strie marginale latéralement, cette derniere prolongée le long de la base par des points alignés; de longues soies isolées (ou tout au moins les gros points qui en marquent l’implantation) sont disposées symétriquement, une premiere sur la strie marginale au niveau de l’oeil, une autre environ au milieu de la longueur et une troisieme au niveau du !/se basal, ces trois soies formant un triangle tres étiré dont le sommet, externe, est formé par le deuxieme point sétigere; hypomeére avec deux rangées basales et obliques de soies courtes. Élytres plans ou un peu concaves en avant, convexes le long de la suture et en arriere, déclives et arrondis latéralement; angles antérieurs externes trés aigus et proéminents, bords postérieurs arrondis, tres rétrécis; ponctuation comme sur le pronotum, peu dense et extréme- ment fine, sétigere au moins pres des bords, plus forte et alignée le long des stries, éparse sur le 1/4 postérieur; stries légérement careniformes, intervalles les séparant quasiment plans, de largeurs inégales; 4 stries dorsales entieres (sauf la lere brievement interrompue en arriere), sinuées, les deux leres (externes) recourbées vers la suture a chacune de leurs extrémités; suturale représentée par un court appendice apical et une ligne de points espacés vers l’avant; strie subhumérale (interne?) basale, atteignant le milieu; deux stries épipleurales; épipleurale externe visible du dessus a la base de l'élytre (elle pourrait étre interprétée comme une sub- humérale externe), descendant sur l'épipleure, puis remontant perpendiculairement pour atteindre le '/3 postérieur de lP'élytre; épipleurale interne, formée de points alignés, atteignant a peine le milieu, raccourcie également en avant et représentée par un court appendice de strie a Papex. Propygidium 1,6 fois aussi large que long, peu convexe, la moitié basale et les bords latéraux avec des gros points dont certains portent des soies fortes; deux petites stries en croissants ouverts vers l’exterieur divisent la base approximativement en trois parties (fig. 4). Pygidium 0,8 fois aussi large que long, bombé; ponctuation apicale forte et éparse, une strie ponctuée longe chaque cóté (fig. 4). Sterna de structure tres particuliere, en réalité difficile a décrire avec des mots (v. fig. 2). Prosternum comprimé latéralement; carene dépourvue de stries; base coupée droite, avec une petite incision triangulaire et presque toute la partie située entre les procoxae creusée pour recevoir une large avancée du mésosternum; apex tronqué droit, formant la limite postérieure d'une cavité occupant la partie médiane du lobe prosternal; ce dernier séparé des alae par une suture; stries latérales courtes, tres “ascendantes”, chacune garnie d'une forte sole; stries latérales marginales courtes, longées de soies fortes, se terminant en avant dans une fossette (orifice glandulaire ?); alae ornées d'une vingtaine de tres gros points; bords de la caréne ponctués sur leur moitié apicale, lisses basalement; médialement par rapport a chaque angle postérieur s’ouvre une cavité profonde (orifice glandulaire ?) dont le “toit” est formé d'une rangée de fortes soies contigués. Mésosternum en avancée largement arquée, le rebord en bourrelet; une fossette profonde de chaque cóté, surmontée de 2 touffes de poils; strie marginale entiere, peu nette, arquée; mesocoxae tres écartées, déportées sur les cótés. Métaster- num tres long; stries latérales internes entieres, convergentes vers l’arriere, entre les metacoxae; stries latérales externes formées de gros points alignés, ainsi que les stries postcoxales, ces derniéres legerement récurrentes. Premier sternite abdominal entouré apicalement d'une ligne de points. Longueurs relatives (rapportées a la longueur totale des sterna) du pro-, méso-, metasternum et ler sternite abd. respectivement égales a 0,3; O,16; 0,4; 0,12. Pattes (fig. 3) de longueur modérée, antérieures et médianes avec des soles recourbées, disposées en rangées, postérieures a fémurs presque inermes; ponctuation réguliere, tres dense; tibias aplatis, bord externe non anguleux, garni de denticules. Genitalia máles inconnus; genitalia femelles non étudiés. Nouvelles d’Histeridae 351 Aristomorphus perversus (Reichensperger, 1923a) Ecitonister perversus: Mazur 1984: 311; Helava et al. 1985: 282. Holotype (par monotypie): un ex. étiqueté: “Ecitonister perversus n. Reichensp.” (man. a cadre noir) / “TYPUS” (Et. rouge imp.)/ “St. Catarina Blumenau 22 XII” (imp. et man.) / “Eciton (imp.) caecum (man.)” (et. a cadre noir) / “Coll. Reichensperger” (imp.). ZFMK. L,=21 M0 = 15 mm. Il est fort probable que cette espece ne représente qu’une variété de la précédente (figs. 5—7). Bien que nous n’ayons pu comparer directement les types, les seules differences que nous avons pu relever concernent la striation moins développée et la ponctuation plus discrete chez A. perversus qui ne montre par ailleurs ni striation ni ponctuation particuliere a la base du propygidium. Seul du matériel complémentaire en provenance de chacune des localités- types permettrait d'établir son statut avec certitude. Aristomorphus borgmeieri (Reichensperger, 1931) Ecitonister borgmeieri: Mazur 1984: 310; Helava et al. 1985: 282. Lectotype (présente désignation): une 9 étiquetée: “Ecitonister borgmeieri n. Reichensper- ger” (man. a cadre noir) / “Eciton praedator” (imp. a cadre noir) / “Type ! Reichensperger” (Ersroueerimp)). ZENKE LE = 1,8 mm; 1=-1.3 mm. Cette espéce se distingue des précédentes et de celle qui suit principalement par sa taille infé- rieure, le dessin de la strie pronotale latérale dans les angles antérieurs (fig. 8) et les ponctua- tion et chétotaxie des sterna (fig. 12). Les genitalia máles (figs. 9—11) combinent la présence d'une strie apicale sur le tergite 8, des disques simples et des soies a l’apex de chaque sternite 8, le tergite 9 avec des apodemes ventraux-postérieurs et des guides internes de l’édéage; ce der- nier se compose d'une base longue et de parameres longs; le sternite 9 n'est pas particuliere- ment modifié. Matériel: Brésil: Altamira-Maraba km 18, Para, 10-23/09/85, piege d'interception, 2 ex., 1 ©; Tucurui, Para, 27/10-9/11/85, piege d’interception, 1 ©, CHND. Aristomorphus ogloblini (Bruch, 1933) Ecitonister ogloblini: Mazur 1984: 310; Helava et al. 1985: 282. Holotype (désignation originale): 1 © étiqueté: “Ecitonister Ogloblini Bruch typus (man.) C. BRUCH DETERM. (imp.)” (ét. a cadre noir)/ “con Eciton (L.) coecum” (man.)/ “Typus” (ét. rouge man.)/ “Est. Exp. Loreto (Misiones-Arg.) Dr. A. A. Ogloblin (recto imp.) 21. iii. 1932 (verso man.)”. L = 2,5 mm; |= 1,8 mm. MACN. Les figs. 13 a 20 montrent des structures qui peuvent aider a différencier cette espece des précédentes. Contrairement a A. borgmeieri, on n’observe pas de disques apicaux sur les sternites 8 du mále mais seulement des soies. Matériel. Brésil: Bella Vista, Goiás, 28/12/1934 (Schwarzmaier), avec Eciton caecum, 1 ex., ZFMK. Lm = 2,4 mm; Im = 1,8 mm (N = 2). Aristomorphus latipes (Reichensperger, 1933) Ecitonister latipes: Mazur 1984: 310; Helava et al. 1985: 282. Holotype (par monotypie): un ex. étiqueté: “Ecitonister latipes n. sp. Reichensp” (man. a cadre noir) / “Campo Bello Est. Rio coll. Reichensperger” (imp. a cadre noir) / “Eciton praedator” (imp. a cadre noir) / “Type! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK. De par sa petite taille, absence de strie pronotale basale et de fossettes méso-métasternales (fig. 21—22), la presence d'une large dent sur ses fémurs antérieurs et des parameres longs creusés en gouttiere ventralement (fig. 32), A. latipes pourrait constituer un sous-genre distinct. La strie pronotale laterale est ici tout a fait parallele aux angles antérieurs (fig. 23). tw In ba N. Dégallier En outre, cette espece semble étre la seule dont la forme du pygidium est différente suivant le sexe, plus pointue chez la femelle (figs. 24—25). Matériel: Brésil: Itatiaya, Etat de Rio de Janeiro, avec Eciton praedator (Zikan), 3 ex. dont l o et 1 9, ZFMK; Campo Bello, Etat de Rio de Janeiro, avec Eciton praedator (coll. Reichensperger), 1 ex., MACN. Lm = 1,5 (1,4—1,5) mm; Im = 1,3 (1,2—1,3) mm (N = 4). Aristonister, nouveau genre Ecitonister: Helava et al. 1985: 280. Espece-type: Ecitonister sericeus Borgmeler, 1948 Ce genre se distingue de Aristomorphus par la combinaison de caracteres suivante: face externe des mandibules incisées a la base; deux stries pronotales latérales dont l’interne est tres éloignée du bord; proépisterne et proépimere dépourvus de fossettes ou aires sétigeres; apex du lobe prosternal non excavé; stries métasternales latérales internes trés rapprochées du milieu en avant, sinuées pour rejoindre les angles postérieurs; genitalia máles: sternites 8 non sétigeres a Papex, édéage a piece basale longue et parameres courts. Aristonister sericeus (Borgmeier, 1948), nouvelle combinaison Ecitonister sericeus: Mazur 1984: 311; Helava et al. 1985: 282. Lectotype (présente désignation): un o étiqueté: “Ecitonister sericeus Borgm.” (man. a cadre noir) / “b. E. pseudops Trinidade Bras. 21.9.36” (man. a cadre noir) / “A. Reichensper- ger 1949”‘ (imp. et man.). L = 2,2; 1 = 1,7 mm. ZFMK. Paralectotype (présente désignation): un ex. étiqueté: “Campinas, Goiás Schwarzmaier” (imp.) “16.X11.1938 E. pseudops“ (man.) / “Cotypus” (rouge imp.) / “SEM” (jaune imp.) / “Ecitonister sericeus Borg. Type” (man.) “det. Borgmeier” (imp. a cadre noir). FMNH; les données relatives a ce specimen nous ont été aimablement fournies par le Dr. A. F. Newton, Jr. (FMNH). Bien que ces deux specimens soient sans nul doute des syntypes, il faut cependant noter que les dates portées par les étiquettes et citées ci-dessus ne correspondent pas exactement aux dates publiées par le descripteur: 26/09/1936 et 16/11/1938, respectivement. See Figs 17—39; figs 17—20: Aristomorphus ogloblini — 17, élytre gauche en vue dorsale, 18, méso-métasternum, 19, élytre droit vu de profil (la ligne pointillée délimite l'épipleure), 20, angle antérieur du pronotum; figs. 21-32: Aristomorphus latipes — 21, vue dorsale, 22, méso-métasternum, 23, angle antérieur pronotal, 24, pygidium de la femelle, 25, idem, mále, 26, patte antérieure en vue ventrale, 27, idem, vue dorsale, 28, patte médiane en vue ventrale, 29, patte postérieure en vue ventrale, 30, tergite 9 du mále vu de profil, 31, tergite et sternite 8 du mále en vue dorsale, 32, édéage en vue latéro-ventrale gauche; figs. 33—39: Aemulister borgmeieri — 33, tergite et sternite 8 du mále en vue dorsale, 34, tergite 9 du mále en vue ventrale, 35, idem, profil gauche, 36, édéage, profil droit, 37, idem, paraméres en vue dorsale, 38, idem, vue ventrale, 39, sternite 9 du mále en vue dorsale. Echelles 1 mm: en haut pour les figs. 17—19, 21, au milieu pour les figs. 33—39 et en bas pour les autres figs. 353 Mewvelles d’Histeridae NS z = sd > t 354 N. Dégallier Cheilister Reichensperger, 1924 Cheilister: Mazur 1984: 300; Helava et al. 1985: 319. Cheilister sphaeroides Reichensperger, 1938 C. sphaeroides: Mazur 1984: 300; Helava et al. 1985: 321. Lectotype (présente désignation): un ex. étiqueté: “Cheilister sphaeroides Reichensp‘‘ (man. a cadre noir) / “Hamburg-Farm Costa Rica Nevermann (imp.) 8. XI 8a (man.)” (et. a cadre noir) / “Type! 2 Reichensperger” (ét. rouge imp.) / “E. (Acamatus) pilosum. c. Reichens- perger” (imp. a cadre noir). ZFMK. Paralectotype (présente désignation): un ex. Etiquete: “Cheilister sphaeroides Reichensp.” (man. a cadre noir) / “Hamburg-Farm Costa Rica Never- mann (imp.) 8. XI 8a (man.)” (ét. a cadre noir) / “Paratype! Reichensperger” (ét. rouge imp. á cadre noir) / “E. (Acamatus) pilosum. c. Reichensperger” (imp. a cadre noir) / “MUSEUM KOENIG BONN” (ét. jaune imp.). ZFMK. Clientister Reichensperger, 1935 Clientister: Mazur 1984: 311; Helava et al. 1985: 324. Clientister henrici Reichensperger, 1935 C. henrici: Mazur 1984: 311; Helava et al. 1985: 326. Lectotype (présente désignation): un ex. étiqueté: “Clientister henrici n. g. n. sp. Rei- chensp.” (man. a cadre noir) / “San Jose Costa Rica H. Schmidt” (imp.) / “Eciton burchelli (imp.) 9.1932 (man.)” (ét. a cadre noir) / “Type! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK. Paralectotypes (présentes désignations): un ex. étiqueté: “Clientister henrici Rei- chensp.” (man. a cadre noir) / “San Jose Costa Rica H. Schmidt” (imp.) / “Eciton burchelli (imp.) 9.1934 (man.)” (ét. a cadre noir) / “Paratype! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK; un ex. étiqueté: “Clientister henrici Reichensp.” (man. a cadre noir) / “San Jose Costa Rica H. Schmidt” (imp.) / “Eciton burchelli (imp.) IX.33 (man.)” (et. a cadre noir) / . “Paratype! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK; un ex. étiqueté: “Clientister henrici Reichensp.” (man. a cadre noir) / “San Jose Costa Rica H. Schmidt” (imp.) / “Eciton burchelli imp.) . . . 34. (man.)” (et. a cadre noir) / “Paratype! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK; un ex. étiqueté: “Clientister henrici Reichensp.” (man. a cadre noir) / “San Jose Costa Rica H. Schmidt” (imp.) / “Eciton burchelli (imp.) IV. 33 (man.)” (ét. a cadre noir) / “Co (man.) Type! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK; un ex. étiqueté: “Clientister henrici Reichensp.” (man. a cadre noir) / “Costa Rica F. Nevermann 26.111. 1934” (et. verte imp. et man.) / “Hamburg Farm Reventazon Ebene Limon” (ét. verte imp.) / “Eciton burchelli” (imp. a cadre noir) / “Paratype! Reichensperger” (ét. rouge imp. a cadre noir). ZFMK; un ex. cité par Gaedicke (1984: 460): “Costa Rica, San José, 11. 1933, an Eciton burchelli, leg. H. Schmidt.“ DEIC. A E NAS Figs 40—59; figs 40—42: Colonides collegii guyanensis n. subsp. — 40, moitié droite du pronotum et des élytres, 41, pygidium de la femelle, 42, idem, profil gauche; figs.43—47: Colonides c. collegii — 43, pygidium du male, 44, idem, profil gauche, 45, pygidium de la femelle, 46, idem, profil gauche, 47, idem, détail de l’apex; figs. 48—49: Colonides quadriglu- mis — 48, tergite 9 du male en vue ventrale, 49, genitalia femelles en vue ventrale avec a, sternite 8, b, sternite 9, c, tergite 9, d, tergite 8 et e deux sclérites génitaux; figs. S0—57: Proco- lonides bruchi — 50, sterna (ex. coll. T. Borgmeier, FIOC), 51, tergite et sternite 8 du mále (topotype, BMNH) en vue dorsale, 52, tergite et sternite 9 du mále (ex. topotype, BMNH) en vue dorsale, 53, édéage (ex. coll. T. Borgmeier, FIOC) en vue ventrale, 54, idem, profil droit, 55, tete (ex. coll. T. Borgmeier, FIOC), 56, pygidium (ex. topotype, ZFMK), 57, genitalia femelles en vue ventrale (ex. coll. C. Bruch, MACN); figs. 58—59: Neocolonides howdeni n. gen. n. sp. — 58, pygidia, 59, genitalia femelles en vue ventrale. Echelles 1 mm: a gauche en haut pour la fig. 40; a gauche au milieu pour les figs. 41—46, 58; a droite pour les autres figs. 309 Nouvelles d’Histeridae 356 N. Dégallier Matériel: Costa Rica: San José (H. Schmidt) avec Eciton burchelli, 2 ex., CHND. L = mama: Colonides Schmidt, 1889 Colonides: Mazur 1984: 305; Helava et al. 1985: 198. Xylostega: Mazur 1984: 305; Helava et al. 1985: 200. Helava et al. (1985) ont regroupé dans un méme ensemble, avec cependant quelques réserves, les genres suivants: Colonides (incluant le genre Xylostega Reichensperger qui, pour nous, en est synonyme), Procolonides et Hemicolonides sensu Helava (taxon révisé plus loin). En 1935 (p. 204), Reichensperger mettait son genre Xylostega en synonymie avec Colonides mais sans préciser quelles étaient les “einiger kleiner Unterschiede” [quelques petites différen- ces] subsistant entre ces deux genres. Les caractéres invoqués par Helava et al. (1985) pour differencier Xylostega de Colonides sont assez subjectifs. Ils ne pourraient tout au plus servir qu’a séparer des sous-genres. Ils sont distribués chez les différentes especes de la maniere suivante: — C. drakei ne semble pas posséder de strie élytrale dorsale médiane plus forte que les autres. Par contre, Colonides collegii (Reichensperger) possede cette carene dans un état inter- mediaire entre celui présent chez Colonides hubrichi Bruch et chez C. quadriglumis (Reichens- perger). — les stries prosternales sont tres rapprochées mais la strie mésosternale postérieure (“addi- tionnelle”) n’existe ni chez C. quadriglumis ni chez C. collegii; — le caractere plus ou moins relevé des bords du pronotum et des élytres nous semble difficile a apprécier. En conclusion, s’il fallait séparer ces especes selon deux groupes, nous le ferions donc plutöt de la maniere suivante: gr. 1, C. collegii; gr. 2, C. hubrichi, C. quadriglumis, C. drakei (?). Contrairement a ce qui est indiqué dans le cladogramme et les schémas de genitalia males (Helava et al. 1985: figs. 5, 86 et 90), nous avons observé un 10eme tergite nettement sclérotisé chez toutes les especes de Colonides (v. fig. 48: C. quadriglumis). Chez toutes les femelles des especes de Colonides, les genitalia, une fois dévaginés, forment un tube (= ovipositeur auct.) court avec, dorsoventralement (de bas en haut sur la fig. 49), une plaque impaire inerme (tergite 8: d sur la fig.), l’orifice anal, une seconde plaque impaire inerme (tergite 9: c), une paire de coxites setigeres mais dépourvus de style et leurs apodemes proximaux (sternite 9: b), l’orifice génital flanqué d'une paire de petits sclérites (e; = vestiges du segment abd. 10 ?), une plaque impaire portant distalement et de chaque cóté un prolonge- ment setigere (sternite 8: a). Le dimorphisme sexuel relatif a la disposition des carenes et a la ponctuation pygidiale (fig. 41—47) se retrouve dans tout le genre. Colonides drakei Schmidt, 1889 C. drakei: Mazur 1984: 305; Helava et al. 1985: 200. Holotype: non retrouvé. La description originale de ce taxon s'applique exactement aux spécimens de C. quadriglumis étudiés. Les caracteres principaux en sont: stries élytrales dorsales completes, bordure apicale du propygidium ondulée, méso-métasternum et premier sternite abdominal strigueux. En l’absence de matériel authentique de cette espéce, nous nous bornerons a la rattacher au groupe formé de C. hubrichi et C. quadriglumis. Colonides collegii (Reichensperger, 1923a) C. collegii: Mazur 1984: 305. Xylostega collegii: Reichensperger 1924: 120; Helava et al. 1985: 202. Nouvelles d’Histeridae 31 Holotype (par monotypie: dans la description originale, Reichensperger ne désigne pas d'holotype et ne précise pas le nombre de spécimens en sa possession, contrairement a ce qu'il fait pour d'autres especes décrites dans le méme article; il emploie cependant l’expression “Das Tier erweckt...” qui nous permet de supposer l'existence d'un seul spécimen!): un o étiqueté: “St. Catarina Blumenau” (ét. imp.) / “Eciton burchelli” (ét. manuscrite) / “Xylo- stega n. collegii n. sp. Reichensp.” (ét. man.) / “Xylostega R. = Colonides Schmidt” (ét. man.) / “coll. Reichensperger“ (ét. imp.) / “TYPUS” (et. imp. en rouge). L = 4,2 mm; | = 3,5 mm. ZFMK. Il est notable que la strie élytrale dorsale médiane forme une cóte plus élevée que les autres stries, quoique a un degré moindre que chez C. hubrichi. Cette espece possede une répartition tres étendue, se superposant a celle de son hóte Eciton burchelli (Westwood). On la rencontre du Costa Rica au sud du Brésil, en passant par les Guyanes. Matériel: Costa Rica: Hamburg-Farm Reventazon, Ebene Limon, 1-11-34 (F. Nevermann), 1 9; idem, 27/11/1934,1 &, FMNH, idem, 1 ©, ZFMK. Brésil: Marambo, Rio de Janeiro, 03/1925 (J. F. Zikan), 1 o, 1 Q, 1 ex., ZFMK; Itatiaya, idem, 6/04/25, 1 9, FIOC; Est. Rio de Janeiro, 1 9, MACN. Lm = 4,15 (3,9—4,3) mm; Im = 3,5 (3,4—3,6) mm (N = 5). Colonides collegü guyanensis, nouvelle sous-espece Holotype: une Q: Sail (3 °37,8’N; 53 °11,9’W), Guyane francaise, 3/08/1978, dans colonne migrante d’ Eciton burchelli, 18—20 h. L = 4,8 mm; 1 = 4,0 mm. CHND. Paratypes: deux 9: méme provenance. CHND. Lm = 4,8 (4,8—4,9) mm; Im = 3,8 (3,7—4,0) mm (N = 3). Cette sous-espéce se distingue de l'espece nominale par les caracteres suivants (femelles): — stries élytrales dorsales 1 et 2 (les plus externes) absentes (fig. 40); — 4eme strie dorsale réduite a un court appendice basal et a seulement !/3 de la longueur de l'élytre apicalement (fig. 40). Le raccourcissement de cette strie n'a été observé chez aucune autre espece du genre; — pygidium a ponctuation tres forte et confluente dans sa moitié apicale, rendant indistinc- tes les carenes latérales (figs. 41—42). Labreviation maximale des stries élytrales et la confluence extréme de la ponctuation pygidiale, observées chez ces exemplaires de Guyane francaise, représentent des differences nettement plus prononcées que les variations individuelles observées chez l’espece nominale. Comme le suggere la répartition des différentes sous-especes d’Eciton burchelli présentée par Watkins II (1976), C. c. guyanensis pourrait représenter une différenciation parallele a celle de l’höte E. b. cupiens Santschi qui habite les Guyanes et le versant nord du bassin amazonien tandis que les C. collegii typiques se récoltent avec E. burchelli s. str. du sud brésilien. En Amérique centrale, l'hóte pourrait étre E. b. foreli Mayr. Un C. collegii male provenant du Costa Rica n’exhibait cependant aucune difference notable par rapport a la forme typique. Colonides hubrichi Bruch, 1923 C. hubrichi: Reichensperger 1924: 121; Bruch 1937: 129; Mazur 1984: 305; Helava et al. 1985: 200. Xylostega hubrichi: Bruch 1929: 434. Holotype (par designation originale): une Q étiquetée: “Alta Gracia, Córdoba 4. XII. 921” (ét. man.) / “Typus” (ét.verte imp.) / “Corriendo con Eciton dulcius Jujuyensis For.” (ét. man.). L = 3,5 mm; | = 2,9 mm. MACN. C. hubrichi semble étre spécifiquement associé a Eciton d. dulcius Forel du sud du Brésil et du nord de l'Argentine (Watkins II, 1976). Il serait interessant de rechercher cet histéride chez E. d. crassinode Borgmeier, sous-espece du Costa Rica et du Panama. Matériel: Brésil: Campinas, Goiás, 27/10/33 (T. Borgmeier), 2 0,1 9, ZFMK; idem, 1 ©, FIOC; idem, 1 ©, FMNH, idem (Schwarzmaier), 2 7, 4 9, CUBO; idem, 1 ©, avec Eciton dulcius, AMIC. Argentine: Hersilia, Prov. Santa Fé, 02/1927 (J. Hubrich), 2 0, MACN; Alta Gracia, Córdoba, 01/1925 (C. Bruch), avec Eciton dulcius jujuyensis For., 1 ©, ZFMK. Lm = 3,8 (3,4—4,1) mm; Im = 3,2 (2,8—3,9) mm (N = 12). 358 N. Dégallier Colonides quadriglumis (Reichensperger, 1923) C. quadriglumis: Mazur 1984: 305. Xylostega quadriglumis: Reichensperger 1924: 120; Helava et al. 1985: 203. Lectotype (présente désignation): une Q étiquetée: “Rio Negro Paraná coll. Reichensper- ger“ (et. imp.) / “E. quadriglume” (et. man.) / “Colonides quadriglumis Reichensp.“ (et. man.). L = 4 mm; | = 3,5 mm. ZFMK. Paralectotypes (présente désignation): un © étiqueté: “Xylostega quadriglumis Reichensp.” (et. man. a cadre noir) / “Rio Negro Paraná” (man. a cadre noir) / “E. quadriglume” (man. a cadre noir) / “Paratyp.” (et. rouge man.). ZFMK; une Q étiquetée: “Rio Negro Paraná coll. Reichensperger” (ét. imp.) / “Paratypus” (ét. verte man.) / “Xylostega quadriglumis Reichensp.” (ét. avec liséré rouge man.). L = 4,1 mm; | = 3,7 mm. MACN; un © étiqueté: “Xylostega quadriglumis Reichensp.” (et. man.) / “Eciton (imp.) quadriglume Paraná” (man.) / “A. Reichensperger (imp.) Paratype (man.)” / “Syntypus” (et. rouge imp.). L = 3,9 mm; | = 3,8 mm. DEIC. Lexistence de ce dernier exemplaire a été signalée par Gaedike (1984: 461) et les données relatives a son sexe, son étiquetage et ses dimensions nous ont été tres aimablement communiquées par le Dr. M. Zerche et le Pr. H. J. Müller. Diverses caractéristiques éloignent cette espece de C. collegii et Papparentent a C. hubrichi: stries élytrales non raccourcies, stries prosternales rapprochées, méso-métasternum et ler sternite abdominal a striation longitudinale dense et en réseau, rebord préapical du propygi- dium nettement ondulé, carénes latérales du pygidium prolongées basalement. Par contre, C. quadriglumis est dépourvu de strie mésosternale supplémentaire et ses cótes élytrales sont toutes également élevées. La fourmi-hóte, Eciton quadriglume (Haliday), semble pouvoir étre rencontrée dans tout le bassin amazonien et dans le sud-est du Brésil (Watkins II 1976). Quelques notes sur le comportement de cette espece dans les fourmiliéres et les colonnes de son hóte (et d’E. burchelli!) ont été publiées par Reichensperger (1926: 197). Matériel: Brésil: Rio Negro, Parana (J. F. Zikan), 2 9, FIOC; idem, (P. M. Witte), 2 9, ZFMK; idem, 1 o, 1 9, FMNH; Blumenau, avec Eciton quadriglume, 1 9, ZFMK. Lm = 4,1 3.94) mms Ime—"3.7, 6538) mms Ni) Clé de determination des especes et sous-espece de Colonides Caracteres du genre en partie selon Helava et al. (1985: 133—135): scape antennaire dilate; proépisternum sans fossette ni aire setigere; lobe prosternal non triparti; mandibules sans fovea ni incision basale; élytres et pronotum avec des cótes longitudinales (5 sur le pronotum); marges latérales non incisées en arriere des angles antérieurs; forme du corps ovale et peu convexe, les rebords plus ou moins relevés; suture méso-métasternale sans foveae; seules la cóte pronotale médiane et les deux externes entieres, les autres raccourcies; 8e sternites du male avec des “disques” apicaux et 10e tergiterbien seleuie eg Colonides (= Xylostega) 1 Couleur brune ou marron; ponctuation méso-métasternale forte et confluente mais non longitudinalement strigueuse; rebord préapical du propygidium non nettement ondulé; cótes élytrales dorsales jamais entieres .. 02. 00.0. od. os on SE eee 2 1” Couleur sombre, presque noire; méso-métasternum et ler sternite abdominal longitudinale- ment strigueux; rebord préapical du propygidium nettement ondulé................... 3 2 Une seule cóte élytrale entiere; ponctuation du pygidium de la femelle (male inconnu) confluente, rendant les carenes longitudinales indistinctes ... Colonides collegii guyanensis 2’ Au moins deux cótes élytrales entieres; carenes pygidiales longitudinales bien visibles dans la ponctuation fortermais peu) contiuente a er Colonides c. collegii 3 Cote élytrale dorsale médiane beaucoup plus élevée que les autres; une strie mésosternale supplémentaire en avant de la suture méso-métasternale............. Colonides hubrichi 3’ Toutes les cótes élytrales faiblement mais également élevées; pas de strie mésosternale supplémentaire en avant de la suture méso-métasternale oo eee eee A eee cerco ns oa ae Colonides quadriglumis (? = Colonides drakei) Nouvelles d’Histeridae 359 Cossyphodister Reichensperger, 1936 Cossyphodister: Mazur 1984: 315; Helava et al. 1985: 185. Cossyphodister schwarzmaieri Reichensperger, 1936 C. schwarzmaieri: Mazur 1984: 315; Helava et al. 1985: 186. Lectotype (présente désignation): une Q étiquetée: “Cossyphodister schwarzmaieri Rei- chensp.“ (et. man. a cadre noir) / “Type! Reichensperger (imp.) Q (man.)” (ét. rouge) / “n. g. Sta. Cruz., Goyaz 19.5.35. P. Schwarzm” (man. a cadre noir). ZFMK. La récolte de deux exemplaires en region amazonienne étend sensiblement l’aire de réparti- tion de cette espece. Matériel: Brésil: Campinas, Goias, avec Syntermes, 24/03/1938 (Borgmeier) 1 ex., FIOC; Tucurui, Para, piege d’interception, 1 ex. CHND; Caicarinha, Mirador, Maranhäo, 21—29/05/1993, piege d'interception, 1 ex. CHND. Ecclisister Reichensperger, 1935 Ecclisister: Mazur 1984: 304; Helava et al. 1985: 322. Ecclisister bickhardti (Reichensperger, 1923) E. bickhardti: Mazur 1984: 304; Helava et al. 1985: 324. Lectotype (présente désignation): un ex. étiqueté: “Cyclechinus bickhardti n. sp. Rei- chensp.“ (man. a cadre noir) / “Eciton burchelli” (man. a cadre noir) / “St. Catarina Blumenau” (imp.) / “Coll. Reichensperger” (imp. a cadre noir) / “TYPUS” (ét. rouge imp.) / “Ecclisister Bickhardti Reichensp.“ (man.). ZFMK. Paralectotypes (présentes désignations): un ex. étiqueté: “Cyclechinus bickhardti n. sp. Reichensp.” (man. a cadre noir)/ “Eciton burchelli” (man. a cadre noir) / “St. Catarina Blumenau” (imp.) / “Coll. Reichensperger” (imp. a cadre noir) / “Cotype” (et. rouge imp.) / “Ecclisister Bickhardti Reichensp.“ (man.). ZFMK; un ex. étiqueté: “Cyclechinus bickhardti n. sp. Reichensp.“ (man. a cadre noir) / “Eciton burchelli” (man. á cadre noir) / “St. Catarina Blumenau” (imp.) / “Coll. Reichens- perger” (imp. a cadre noir) / “Cotype” (et. rouge imp.) / “Ecclisister Bickhardti Reichensp.” (man.). ZFMK. Matériel: Brésil: Utinga (IPEAN), Belém, Para, 18/08/1984, dans nid d’Eciton dans tronc creux au sol, 1 ex.; Altamira-Maraba Km 18, Para, 14/04/1986, dans bivouac d'Eciton burchelli, 1 ex., CHND. Guiana: Kartabo, 15/04/1919, avec Eciton hamatum, lex. CHYG. Euxenister Reichensperger, 1923a Euxenister: Mazur 1984: 321; Helava et al. 1985: 203. Euxenister caroli Reichensperger, 1923a E. caroli: Mazur 1984: 321; Helava et al. 1985: 205. Lectotype (présente désignation): un ex. étiqueté: “Euxenister n. caroli n. Reichensperger” (man. a cadre noir) / “Eciton burchelli” (man. a cadre noir) / “St. Catarina Blumenau” (imp.) / “Coll. Reichensperger” (imp.) / “TYPUS I” (ét. rouge imp. et man.). ZFMK. Paralectotype (présente désignation): un ex. étiqueté: “Euxenister caroli n. Reichensp.“ (man. a cadre noir) / “Eciton burchelli” (man. a cadre noir) / “St. Catarina Blumenau” (imp.) / “Reichensperger” (imp.) / “Cotype” (ét. rouge imp.). ZFMK. Matériel: Guyane francaise: Satil, 3/08/1978, dans colonne migratrice d’Eciton burchelli (dE E: Overal), 29; 10, CHND: 360 N. Dégallier Hemicolonides Reichensperger, 1939 Hemicolonides: Mazur 1984: 305. Oudaimosister sensu Helava in Helava et al. 1985: 261; figs 187—190. Oudaimosister Helava, in Helava et al. 1985 = Mesynodites Reichardt, 1924, nouvelle synonymie. Lillustration et la description fournies pour définir le genre Oudaimosister Helava s'appli- quent a peu de détails pres a l’holotype d’Hemicolonides plaumanni (fig. 64). Par contre, la description originale de Mesynodites verruculosus Reichensperger, espece-type de Oudaimosister, est en contradiction avec la diagnose de ce dernier genre en ce qui concerne les caracteres suivants: ponctuation sétigere, pronotum regulierement convexe a partir des cótés, stries subhumérales des élytres non carénées, pygidia simples, lisses, a ponctuation peu dense, strie mésosternale antérieure indistincte, ponctuation sternale tres espacée, effacée par places, pas de suture visible entre le méso- et le métasternum, pas de stries postcoxales sur le ler sternite abdominal, etc. En conséquence, nous considérons le genre Oudaimosister synonyme du genre Mesynodites Reichardt. Hemicolonides plaumanni Reichensperger, 1939 H. plaumanni: Mazur 1984: 305. Holotype (par monotypie): un ex. étiqueté: “E. Plaumann Nova Teutonia Brasilien” (et. imp.) / “Eciton praedator (imp.) 12.11.37 (man.)” / “Hemicolonides Plaumanni Reichensp.” (et. man.) / “Geno- (man.) TYPUS (imp.)” (et. rouge). L = 1,6 mm; 1 = 1,4 mm. ZFMK. Laspect de cette espece en vue dorsale est montré sur la fig. 64. Elle se distingue de l’espece suivante par sa forme moins allongée (rapport L/l = 1,14 contre L/l = 1,30 pour A. parvulus), par l’absence de strie transversale antérieure sur le pronotum et par la presence de 8 cótes dorsales (au lieu de 6) sur chaque élytre. Hemicolonides parvulus (Lewis, 1891), nouvelle combinaison Colonides parvulus Lewis 1891: 404; Mazur 1984: 305; Helava et al. 1985: 200. Lectotype (présente désignation): un O étiqueté: “Mexico A. G“ (et. man.) / “Colonides parvulus Lewis Type.” (ét. man.) / “Type” (ét. ronde a bordure rouge) / “Georges Lewis Coll. B. M. 1926—369” (et. imp.) / “SYNTYPE” (et. ronde a bordure bleue) / “Hemicolonides parvulus (Lewis) n. comb. lectotype N. DEGALLIER” (ét.man.). L = 1,4 mm; | = 1,0 mm. BMNH. Letude des types d’Hemicolonides plaumanni Reichensperger et de Colonides parvulus Lewis nous a convaincu de l’appartenance de ces deux especes au méme genre, par ailleurs tres distinct du genre Colonides sensu Schmidt. En outre, le specimen ayant servi de base a la description de Oudaimosister verruculosus sensu Helava est extrémement ressemblant sinon identique (nous ne l’avons pas eu en notre possession) au lectotype d’Hemicolonides parvulus (Lewis). Cependant, chez ce dernier, les 8e sternites (figs. 68 —69) ne sont pas “séparés” mais soudés en une plaque impaire portant de chaque cóté des “soies apicales”, les 9e tergites (figs. 70—71) sont munis d’apodemes ventraux et les “guides internes” de l’édéage, assez globuleux, ne sont pas transversaux mais paralleles a Paxe longitudinal. Nous completerons les données relatives a ce taxon par les figs 65 a 67. Matériel: Brésil: route Altamira-Maraba, km 18 (52°O3’W —3°09’S), Pará, 3—21/04/1986 piege d’interception, 1 ©. L = 1,4 mm; 1= 1,0 mm. CHND. Hetaeriarchus Reichensperger, 1936a, nomen nudum Hetaeriarchus Reichensperger, 1936a: 190; Borgmeier, 1948: 377. La description de ce genre, tres succincte (“ .. sie gibt ganz den gewöhnlichen Histeridentyp wieder, besitzt aber die verwachsene Fühlerkeule und die Fühlergruben der Hetaeriomorphi- nen.”), a été appliquée a un (ou plusieurs ?) Histéride(s) récolté(s) par “ .. Herrn Plaumann in Südkrasilien (sic.!) bei Eciton praedator ...”. Lauteur n’a malheureusement pas désigné d'espece-type, rendant ce nom indisponible (Code, 1985: Art. 13(b)). Nouvelles d’Histeridae 361 Figs 60—71; figs 60—63: Neocolonides howdeni n. gen. n. sp. — 60, patte antérieure an vue ventrale, 61, patte médiane en vue ventrale, 62, patte postérieure en vue ventrale, 63, sterna et pleures, vue latéro-ventrale droite; fig. 64: Hemicolonides plaumanni, habitus en vue dorsale; figs. 65—71: Hemicolonides parvulus — 65, patte antérieure en vue dorsale, 66, patte médiane en vue ventrale, 67, patte postérieure en vue ventrale, 68, tergite et sternite 8 du mále vus de profil gauche, 69, idem en vue ventrale, 70, tergite et sternite 9 du mále en vue dorsale, 71, idem en vue ventrale. Echelles 1 mm: en bas pour les figs. 60—63; en haut pour les autres figs. 362 N. Dégallier Une recherche effectuée dans l'ensemble de la collection Reichensperger ne nous a pas permis de trouver d'exemplaire correspondant a cette publication. Par contre, dans le fichier bibliographique manuscrit de A. Reichensperger, mis a notre disposition par le Dr. H. Roer (ZFMK, Bonn), nous avons pu lire la mention: “Hetaeriarchus nom. nud. = Phelister sculptu- ratus u. praedatoris.”, taxon publié comme espece en 1939, Latronister Reichensperger, 1932 Latronister: Mazur 1984: 320; Helava et al. 1985: 298. Latronister breyeri (Bruch, 1931), nouvelle combinaison Euxenister breyeri: Mazur 1984: 321; Helava et al. 1985: 205. Holotype (par monotypie): une Q étiquetée: “Euxenister Breyeri Bruch (man.) C. BRUCH DETERM. (imp.)” / “con Ectatomma quadridens F. Sm.” (man.) / “Typus” (rouge man.) / “1031.” (man.) / “REP. ARGENTINA PATQUIA” (imp.). MACN. Les figs. 72 a 74 montrent quelques particularités de cette espece. D'apres les descriptions (Pholotype de l’espece-type du genre, L. rugosus Reichensperger n’a pu étre retrouvé) et le seul exemplaire femelle étudié, cette espece ne peut étre maintenue dans le genre Euxenister. Du matériel complémentaire permettra, par létude des genitalia males, de confirmer ou d'infirmer sa réunion au genre Latronister. Neocolonides, nouveau genre Espece-type: N. howdeni, nouvelle espece Comme nous l'avons déja noté (Dégallier 1987: 220), les descriptions et illustrations fournies par Helava et al. (1985: 195—197) pour caractériser le genre Hemicolonides Reichensperger ne correspondent aucunement a celles publiées par l’auteur de ce genre. De plus, un exemplaire probablement conspécifique de Hemicolonides n. sp. 1 sensu Helava et al. (1985) s'est révélé étre extrémement different de ’holotype de l’espece-type A. plaumanni Reichensperger. Le genre Neocolonides est donc caractérisé par la description (“Diagnosis”) et les figs. 82 a 85 données par Helava et al. (1985: 195—197) ainsi que nos figs. 58 a 63. Nous compléterons cette diagnose par les notes suivantes: ponctuation des téguments forte et confluente, formant une réticulation; labre non setigere; pygidium avec, en plus de la carene médiane, une caréne longeant les cótés et l'apex (fig. 58); carenes prosternales divergeant en arriére et convergeant sans se réunir en avant; stries prosternales latérales externes (“lateral marginal striae” de Helava et al. 1985) se terminant dans une fossette de chaque cóté a la base du lobe pronotal (fig. 63); genitalia femelles comme dans le genre Colonides mais avec les sclérites ventraux (st. 8) dépourvus de soies, moins allongés et les coxites (st. 9) a extrémité triangulaire (arrondie chez Colonides) (fig. 59). II AE ES E E E mer, Figs 72—93; figs 72—74: Latronister breyeri — 72, prosternum vue du profil droit, 72, pygidia, vus du profil droit, 74, genitalia femelles en vue ventrale; figs. 75—81: Panoplitellus comes — 75, tergite et sternite 9 du mále en vue dorsale, 76, tergite 9 vu du profil gauche, 77, tergite et sternite 8 du mále en vue dorsale, 78, idem, profil gauche, 79, édéage, profil gauche, 80, idem en vue dorsale, 81, genitalia femelles en vue dorsale; figs. 82—88: Para- synodites suturacava — 82, sterna, 83, tergite et sternite 8 du mále en vue ventrale, 84, tergite 9 du mále en vue dorsale, 85, édéage en vue dorsale, 86, idem, profil gauche, 87, sternite 9 du male, profil gauche, 88, genitalia femelles en vue dorsale; figs. 89 —93: Pelatetister pretio- sus — 89, pronotum et téte en vue antérieure, 90, tergite et sternite 8 du mále en vue ventrale, 91, tergite 9 du mále en vue dorsale, 92, édéage en vue ventrale, 93, idem, profil gauche. Echelles 1 mm: a gauche en haut pour les figs. 72—73, 77—80; a gauche au milieu pour les figs. 74—76, 81; a droite pour les autres figs. 72 ‘2 Nouvelles d’Histeridae 363 364 N. Dégallier Neocolonides howdeni, nouvelle espece Holotype: 1 Q étiquetée: “PANAMA-Canal Zone Barro Colorado Island VI-4: 63” (imp.) / “Host: Eciton hamatum Colony E-340” (imp.) / “R. D. Akre #100 preserved VI-4-63” (man.) / “Xylostega n. sp. 771” (man.). L = 2,6 mm; | = 2,3 mm. FMNH. Paratype: 1 or, illustré par la fig. 82 (et peut-étre les figs. 83 a 85: voir ci-dessous) du travail de Helava et al. (1985: 196) et cité de “PANAMA. Canal Zone: Barro Colorado Island. HOST. Neivamyrmex pilosus.” Remarque sur le choix de l’holotype: Nous n’avons malheureusement pas pu examiner l’exemplaire mále figure par Helava et al. (1985) car sa localisation nous est restée inconnue, ceci malgré les interventions bienveillantes du Pr. H. F. Howden aupres de J. V. T. Helava. Vus, d'une part, l’incertitude qui subsiste dans l’association entre le mäle et la femelle de cette espece et d'autre part, le fait que les fourmis hótes de ces deux exemplaires appartenaient a des genres différents, il nous a semblé plus “prudent” de designer comme holotype la femelle (voir a ce propos la Recommandation 73B du Code International de Nomenclature Zoologique). Description: Tous les caracteres, exceptés ceux concernant les genitalia máles, énumérés par Helava et al. (1985) dans la diagnose du genre Hemicolonides sensu Helava, s'appliquent a Pholotype désigné. Nous les compléterons par les suivants: corps entierement brun-rouge avec des zones ou táches noires irrégulieres; élytres avec quatre cótes entieres, les deux premieres (externes) faiblement marquées, les deux suivantes plus relevées, une cóte suturale de hauteur intermédiaire; pattes comme sur les figs. 60 a 62; longueurs des pro-, méso-, métasternum et ler sternite abd. respectivement égales a 0,5 mm, 0,1 mm, 0,3 mm et 0,2 mm. Nomadister Borgmeier, 1948 Nomadister: Mazur 1984: 317; Helava et al. 1985: 355. Nomadister papillatus Borgmeier, 1948 N. papillatus: Mazur 1984: 317; Helava et al. 1985: 356. Holotype (par désignation originale): une Q étiquetée: “Nomadister papillatus Borg. Holo- typus det. Borgmeier” (man. avec bordure noire) / “typus” (rouge imp.) / “Goiás, Trindade Schwarzmaier 21. 9. 36 pseudops“ (imp. et man.) / “N. pseudops” (man.) / “Nomadister papillatus Borgm.“ (man. a bordure noire) / “Neivam. pseudops” (man. a bordure noire) / “MUSEUM KOENIG BONN” (jaune imp.). L = 1,5 mm; 1= 1,3 mm. ZFMK. La structure des genitalia femelles (fig. 114) confirme l’appartenance de ce genre a la sous- famille des Hetaeriinae. Panoplitellus Hedicke, 1923 Panhoplites: Reichensperger 1924: 129 et suiv. Panoplitellus: Mazur 1984: 299; Helava et al. 1985: 287. Description des genitalia: o: tergite 8 (figs. 77—78) avec une strie transverse anté- rieure, sans expansions ventrales sur le bord postérieur; sternites 8 séparés, avec des disques, sans soies a l’apex; tergite 9 (figs. 75—76) avec des apodémes ventraux et des guides internes pour l'édéage, sans prolongements apicaux en crochet; sternite 9 épaissi dans sa moitié basale; tergite 10 présent, membraneux; édéage (figs. 79—80) a piece basale longue et parameres courts. 2 (fig. 81): sternite 8 a apodemes tres courts, avec des lobes setigeres larges a la base et pointus a Vapex; tergite 9 sans prolongement médian proximal; sternite 9 avec des coxites setigeres, sans style et soudés a leurs apodemes qui sont tres courts. Les caractéres ci-dessus permettent de confirmer le placement de ce genre dans le groupe E, sous-groupe E3 de Helava et al. (1985). Nouvelles d’Histeridae 365 Panoplitellus comes (Reichensperger, 1923a) P comes: Mazur 1984: 299; Helava et al. 1985: 289. Deux exemplaires récoltés en Amazonie correspondent parfaitement aux descriptions et illustrations publiées pour ce taxon. Matériel: Brésil: Altamira-Maraba km 18 (52°0O3’W—3°09’S), Pará, 14/04/1986, dans bivouac de Eciton burchelli, 1 0,1 9, CHND; P. N. Itatiaia 800 m, Rio de Janeiro, 02/1957 (Vulcano & Martinez), avec Eciton quadriglume (Dégallier dét.), 1 ex., AMIC. Lm = 4,1 (3,9—4,4) mm; Im = 3,3 (3,2-— 3,4) mm (N = 3). Parasynodites Bruch, 1930 Parasynodites: Mazur 1984: 311; Helava et al. 1985: 357. Description complémentaire: Corps 1,2 fois aussi long que large, garni de longues soles éparses, plus denses le long des cótés du pronotum. Téte aussi large que longue, faiblement carenée sur les marges allant du vertex au clypeus; vertex et front élargis, le premier sans stries divergentes; scape antennaire anguleusement dilaté; massue presque entierement sclérotisée; labre a apex coupé droit, deux fois aussi large que long, soudé au clypeus mais avec une suture visible les séparant; mandibules avec une fossette pres de la base sur leur surface externe. Pronotum deux fois aussi large que long, dépourvu de striation en réseau ou de fossettes. Elytres sans 4eme strie dorsale et sans cótes. Pro-, méso-, métasternum et ler sternite abdominal (fig. 82) de longueurs respectivement égales a 0,3; 0,1; 0,3 et 0,3 fois la longueur totale des sterna (0,8 mm); lobe prosternal non divisé, incliné presque a angle droit par rapport au prosternum; carene avec quelques soies alignées entre les stries carenales et les stries latérales; carene sans processus antérieur; base prosternale largement incisée; proépisternum et proépimere sans touffes de soies; une fossette ronde creuse les angles postérieurs de chaque cóté. Mésosternum sans fossettes profondes mais creusé d'une large excavation qui se continue sur le métasternum. Hanches médianes et postérieures tres éloignées, la distance entre les premieres égale a 1,3 fois la longueur du méso- métasternum. Métasternum avec une seule strie latérale. Tibias sans éperons. Genitalia males (figs. 83—87): tergite 8 avec une bande transversale apicale translucide sinon une strie apicale, sans expansion ventro-postérieure; sternites 8 presque fusionnés, avec des disques, sans soles apicales; tergites 9 sans apodemes ventraux, avec des guides internes de l’édéage, la marge apicale sans prolongement en crochet mais portant une lamelle dorsale triangulaire et translucide; tergite 10 présent; édéage a piece basale longue et parameres longs, recourbés presque a angle droit, tres largement et profondément fendus (sur !/3 de leur longueur) apicalement. Genitalia femelles (fig. 88): sternite 8 a apodemes proximaux tres courts, des lobes setigeres apicaux; tergite 9 sans prolongement médian proximal; sternite 9 a coxites a apex sétigere, sans style, soudés a leur apodemes et médialement. Remarque: Par Paspect de ses téguments, la conformation des sterna et la morphologie des genitalia máles, ce genre rappelle le genre Lissosternus Lewis. Il s’en distingue cependant par Paspect des mandibules, des bords latéraux du pronotum et la forme des pattes. Parasynodites suturacava Bruch, 1930 P suturacava: Mazur 1984: 312; Helava et al. 1985: 357. Lectotype (présente désignation): un © étiqueté: “Parasynodites suturacava Bruch (man.) C. BRUCH DETERM. (imp.)” (ét. blanche a cadre noir) / “Typus” (ét. verte man.) / “700.” (man.) / “ROSAS-F. C. Sud Provincia de Buenos Aires JUAN B. DAGUERRE” (et. blanche imp.). L=1,0 mm; 1=0,9 mm. MACN. Paralectotypes (présente désignation): une Q étiquetée: “Parasynodites suturacava Bruch (man.) C. BRUCH DETERM. (imp.)” (et. 366 N. Dégallier blanche a cadre noir) / “Typus” (et. verte man.) / “700.“ (man.) / “ROSAS-F. C. Sud Provin- cia de Buenos Aires JUAN B. DAGUERRE” (et. blanche imp.). L = 1,0 mm; 1 = 0,9 mm. MACN. Un o étiqueté: “ROSAS-F. C. Sud Provincia de Buenos Aires JUAN B. DAGUERRE?” (et. blanche imp.) / “Typus” (ét. verte man.). L = 1,0 mm; 1 = 0,9 mm. MACN. Un © étiqueté: “700.” (man.) / “ROSAS-F. C. Sud Provincia de Buenos Aires JUAN B. DAGUERRE” (et. blanche imp.). L = 1,0 mm; 1 = 0,9 mm. MACN. Un ex. étiqueté: “Para- synodites suturacava Bruch (et. man.)” / “cotypus” (man.) / “cotypus“ (ét. rouge imp.) / “ROSAS-F.C. Sud Provincia de Buenos Aires JUAN B. DAGUERRE“ (et. blanche imp.). ZFMK. Un ex. étiqueté: “Parasynodites suturacava Bruch (Et. man. a cadre noir)” / “Para- typus (imp.) Co (man. par dessus)” (Et. rouge)/ “Cotypus” (ét. verte man.) / “ROSAS-F. C. Sud Provincia de Buenos Aires JUAN B. DAGUERRE” (imp.). ZFMK. Pelatetister Reichensperger, 1939 Pelatetister: Mazur 1984: 112; Helava et al. 1985: 357. Pelatetister pretiosus Reichensperger, 1939 P pretiosus: Mazur 1984: 112; Helava et al. 1985: 359. Lectotype (présente désignation): un © étiqueté: “Pelatetister pretiosus Reichensp.” (Et. man. a liséré noir) / “Hamburg-Farm Costa Rica Nevermann” (imp.) / “E. (Acamatus) pilo- sum. c. Reichensperger” (imp.) / “Type! Reichensperger” (et. rouge). L = 1,0 mm; 1 = 0,9 mm. ZFMK. Paralectotype (présente désignation): un O étiqueté: “Pelatetister pretiosus Rei- chensp.“ (et. man. a lisere noir) / “Hamburg-Farm Costa Rica Nevermann (imp.) 8 35 XI (man.)” (et. a cadre noir) / “E. (Acamatus) pilosum. c. Reichensperger” (imp.) / “Paratype! Reichensperger” (ét. rouge). L = 1,0 mm; 1 = 0,9 mm. ZFMK. Dans le travail de Mazur (1984), ce taxon est classé dans la sous-famille des Chlamydopsinae, endémique de la région indo-australienne. Mise a part une ressemblance superficielle avec les représentants de ce groupe (Reichensperger 1939), cette interprétation ne peut étre maintenue car Pelatetister est bien un Hetaeriinae. Il ne possede aucune des synapomorphies mises en evidence chez les Chlamydopsinae: pincements ou fentes (“épaulettes”) au niveau des épaules élytrales, massues antennaires au moins trois fois plus longues que larges et non sclérifiées sur une partie de leur surface, scape antennaire recouvrant la massue au repos etc. Pelatetister possede au contraire au moins deux des synapomorphies présentes chez les Hetaeriinae, qui sont la massue antennaire ovale et en partie sclérifiée, fermant les cavités antennaires au repos (fig. 89) et les genitalia máles “raccourcis” (figs. 90—93), assez semblables a ceux figurés par Helava et al. (1985) pour le genre Prerotister. Les genitalia femelles sont malheureusement inconnus. Le caractere “saprinomorphe” (Wenzel 1944) ou raccourcissement extreme du lobe prosternal, partagé avec les Chlamydopsinae, ne peut étre considéré comme une synapo- morphie chez ces derniers car on le rencontre aussi chez d'autres groupes d’Histeridae. Matériel: Costa Rica: Revertazon, Hamburg-Farm (F. Nevermann), avec Eciton (Acamatus) pilosum, 1 ©, ZFMK. L = 1,0 mm; 1 = 0,9 mm. Procolonides Reichensperger, 1935 Procolonides: Mazur 1984: 305; Helava et al. 1985: 193. La description et lillustration des genitalia males données par Helava et al. (1985) indiquent Pabsence de soies apicales et de “disques” sur le 8e sternite et absence de 10e tergite, alors que nous avons observé ces structures (figs. 51—52) chez tous les spécimens males étudiés! Procolonides bruchi Reichensperger, 1935 P. bruchi: Mazur 1984: 305; Helava et al. 1985: 193. Holotype (par désignation originale): non retrouvé. Nouvelles d’Histeridae 367 La collection A. Reichensperger, a Bonn, ne contient qu’un seul exemplaire de cette espece, accompagné d'une étiquette rouge manuscrite: “Type!” mais il ne peut s'agir que d'un topo- type car la date de récolte (1951) est postérieure a celle indiquée apres la diagnose originale! Ce spécimen, dont nous figurons quelques détails morphologiques jusqu’ici non décrits (fig. 56), correspond cependant exactement a la description de l’espece. Des differences par _ rapport a ce topotype et a la description donnée par Helava et al. (1985) sont a mentionner pour deux spécimens, l’un cité d'Argentine par Bruch (1937: 129) et l'autre present dans la collection T. Borgmeier (FIOC). Ces differences ont trait principalement aux caracteres suivants: carene médiane du front se divisant en avant en formant un V inversé (fig. 55); fossettes pronotales non limitées par un bourrelet relevé sur leur rebord interne, un tel bourrelet n'est développé que le long de leur bord externe; stries et carenes élytrales dorsales apicalement indistinctes sur une longueur variant de !/s a 2/3 de la longueur des élytres, la strie externe étant la plus courte; propygidium a concavité dorsale reguliere jusqu’au rebord apical, qui n'est pas bombé comme chez la forme typique; ponctuation pygidiale a points plus ou moins alignés dans la moitié apicale, plus forte basalement; pas d’impression basolatérale circulaire ponctuée de gros points ronds (fig. 56) (ni Reichensperger 1935, ni Helava et al. 1985 ne signalent ce caractere, net chez la forme typique); stries prosternales formant en avant un angle moins effilé que chez la forme typique (fig. 50); carene prosternale, vue de profil, sans pointe nette mais seulement un peu ondulée; stries métasternales internes obsolétes ou nulles, aucunement ramifiées ou arborescentes (fig. 50) (“baumartig” in Reichensperger 1935: 206); extrémités antérieures des stries longitudinales internes du premier sternite abdominal plus proches des coxae que du milieu de la suture métasterno-abdominale; édéage (spécimen de la FIOC) a bords paralleles (fig. 53). Il est cependant difficile de conclure a l'existence d'un taxon distinct en l'absence de matériel complémentaire en provenance d'au moins une de ces localités. P bruchi semble étre habituellement associé a des fourmis du genre Solenopsis (S. saevis- sima: Reichensperger 1935: 207; S. s. tricuspis: Bruch 1937: 129) mais il a aussi été récolté au Bresil en compagnie de Eciton diversinode (det. T. Borgmeier, in litt., confirmée par W. L. Overal)! Matériel: Brésil: Nova Teutonia, Sta. Catharina, 300—500 ft., 9-1951 (F. Plaumann), 1 ©, ZFMK; idem, 15/08/1950, 1 ex.; idem, 9/09/1950, 1 o, BMNH; idem, 1 ©; idem, 08/1972, 1 ©; idem, 15/08/1950, 1 ©, FMNH; Campinas, Goiás, 1963, 1 o, FIOC. Argentine: Loreto, Misiones 14/08/36 (A. Oglobin), 1 9, MACN. Lm = 2,5 (2,0—2,8) mm; Im = 1,6 (1,3—1,8) mm (N = 8). Pterotister Reichensperger, 1939 Pterotister: Mazur 1984: 112; Helava et al. 1985: 214. Pterotister schwarzmaieri Borgmeier, 1948 P schwarzmaieri: Mazur 1984: 112; Helava et al. 1985: 216. Matériel: Brésil: 10—27/09/1992, piege d'interception, Monte Alegre- Oriximina, Para, 1 ©. CHND. Ce genre, plac& par Mazur (1984) dans la sous-famille des Chlamydopsinae, est bien un Hetaeriinae, pour les mémes raisons données au sujet de Pelatetister. Synoditinus Reichensperger, 1929 Synoditinus: Mazur 1984: 310; Helava et al. 1985: 360. Synoditinus herteli Reichensperger, 1929 S. herteli: Bruch 1930: 7, figs 1—3; Mazur 1984: 310; Helava et al. 1985: 361. Bien que la localité-type ait été citée comme “Porto Alegre Para”, nous pensons qu'il s'agit 368 NiDsg aller de la capitale de l'état du Rio Grande du Sud, frontalier avec l’Argentine, et non d'une localité située dans l’etat du Para, situé au nord du Brésil. L’holotype de cette espece, unique de son genre (!), n'a pu étre retrouvé. Cependant, l'iden- tification faite par C. Bruch du seul spécimen que nous ayons pu avoir a notre disposition semble avoir été confirmée par Reichensperger (Bruch, 1930) et sera considérée comme fiable. La caracteristique principale de cet insecte est la forte courbure vers le dos (anatomiquement parlant) du tiers apical de sa caréne prosternale (fig. 109). Le lobe prosternal est divisé en trois parties. Le tergite 8 du mále (fig. 110) ne possede pas d’extensions ventrales mais montre une strie apicale; les sternites 8 possedent des disques apicaux. Les tergites 9 (fig. 111) ont des expansions postéro-ventrales et des guides de l'édéage mais le tergite 10 semble absent. L’edeage (fig. 112) est formé d'une piece basale longue et de parameres longs, divergeant a Papex. Matériel: Argentine: Rosas F. C. Sud Buenos Aires, (J. B. Das) lo, MAG — mos — 1k Ol mm: Teratosoma Lewis, 1885 Teratosoma: Mazur 1984: 321; Helava et al. 1985: 213. Teratosoma longipes Lewis, 1885, sensu Helava et al. (1985) T. longipes: Mazur 1984: 321; Helava et al. 1985: 213. Genitalia máles: tergite 8 avec une strie antérieure transverse, sans prolongements ventro- postérieurs; sternites 8 soudés médialement, sans soies apicales mais avec des disques sétigeres (fig. 101); tergite 9 avec des apodemes ventraux, des guides internes pour l'édéage et ’apex non prolongé en forme de crochets; sternite 9 en Y, a base non élargie (fig. 102—103); tergite 10 present; édéage a base et parameres longs (figs. 104—105). Les fémurs sont plus sinués (en forme de S) que chez l’espece suivante (figs. 106—108). Matériel: Brésil: Passa Quatro, avec Eciton legionis, 1 ©, FMNH. Colonia Alpina Rio de Janeiro, avec Eciton legionis, 10/1894 (A. Góldi),1 ex., ZFMK. Teratosoma amphiphilus (Bruch, 1926) Neolister amphiphilus: Mazur 1984: 329. T. amphiphilus: Helava et al. 1985: 213. xo ____JJJ_—_—___—_—__———_—_— —___——_—— > Figs 94—114; figs 94—100: Teratosoma amphiphilus — 94, tergite et sternite 8 du mále en vue ventrale, 95, idem, profil gauche, 96, tergite et sternite 9 en vue ventrale, 97, idem, profil gauche, 98, édéage, profil gauche, 99, tergite et sternite 8 de la femelle en vue ventrale, 100, tergite et sternite 9 de la femelle en vue dorsale; figs. 101—108: Teratosoma longipes — 101, tergite et sternite 8 du mále en vue dorsale, 102, tergite et sternite 9 du mále en vue dorsale, 103, tergite 9, profil gauche, 104, édéage, profil gauche, 105, idem en vue ventrale, 106, patte postérieure, vue dorsale, 107, patte médiane en vue dorsale, 108, patte antérieure en vue dorsale; figs. 109—112: Synoditinus herteli — 109, prothorax en vue latéro-ventrale (coté droit), 110, tergite et sternite 8 du mále en vue ventrale, 111, tergite et sternite 9 du mále en vue dorsale, 112, édéage en vue ventrale; fig. 113: Xenister schwarzmaieri, genitalia femelles en vue ventrale; fig. 114: Nomadister papillatus, idem. Echelles 1 mm: a gauche pour les figs. 106—108; a droite en haut pour les figs. 94—105, 109, 113 —114; a droite en bas pour les autres figs. 369 Nouvelles d’Histeridae , A ASS 370 N. Dégallier Lectotype (présente désignation): un exemplaire étiqueté: “Neolister amphiphilus Bruch (man. par Bruch) C. BRUCH DETERM. (imp.)” (ét. blanche) / “Typus” (ét. verte imp.) / “Foto” (ét. verte imp.) / “Alta Gracia La Granja Sierras de Córdoba C. Bruch leg.” (ét. blanche imp.). L = 1,8 mm; 1= 1,2 mm. MACN. Paralectotype (présente désignation): un exemplaire étiqueté: “Typus” (ét. verte imp.) / “Alta Gracia La Granja Sierras de Córdoba. C. Bruch leg.” (ét. blanche imp.). MACN. Genitalia máles un peu different de ceux de l’espece précédente: les disques des sternites 8 sont glabres, la partie médiane de la strie apicale du tergite 8 est arrondie a Papex (figs. 94—95), le tergite 9 est plus allongé, avec les processus antérieurs développés en forme de cuillere a concavité tournée vers le haut (figs. 96—97); tergite 10 present (fig. 96); édéage a base et parameres longs (fig. 98). Genitalia femelles: sternite 8 a apodemes tres réduits, les lobes distaux sétigeres (fig. 99); tergite 9 sans prolongement proximal médian; sternite 9 avec les coxites soudés a leurs apode- mes, sans soies apicales et sans style (fig. 100). Matériel: Argentine: Alta Gracia, La Granja, Sierras de Córdoba, 10/01/1927 (C. Bruch), 1 o; idem, 15/01/1927, 1 ©, 1 9, MACN. 1 ex. (coll. Reichensperger), ZFMK. Lm = 1,9 1320) Mm AZ NS Troglosternus Bickhardt, 1917 Troglosternus: Mazur 1984: 316; Helava et al. 1985: 259. Troglosternus dasypus Bickhardt, 1917 T. dasypus: Mazur 1984: 316; Helava et al. 1985: 261. Lectotype (présente désignation): un o étiqueté: “Rio Grande do Sul” (et. blanche imp.) / “Type” (ét. rouge imp.) / “dasypus Bickh.” (ét. man.) / “Troglosternus Bickh.” (ét. man.) / Zool. Mus. Berlin” (et. imp. en bleu). L = 3,3 mm; | = 2,9 mm. ZMHB. : Pronotum 1,9 fois aussi large que long; striation transversale du disque pronotal atteignant a peine le milieu (fig. 115). Pro-, méso-, métasternum et ler sternite abdominal respectivement égaux a 0,3; 0,1; 0,3 et 0,2 fois la longueur totale des sterna (2,3 mm) (fig. 116); stries de la carene prosternale éloignées entre les hanches antérieures d'environ 0,4 fois la distance séparant la strie basale du lobe de la base du prosternum; métasternum densément strié ne emani. Pattes comme sur les figs. 123 a 125. Les genitalia mäles de cette espece different plus de ceux de 7. /isaevedouae que de ceux de 7: ecitonis: les sternites 8 sont soudés médialement et sont garnis d'une rangée de petites soles droites a l’apex (fig. 119); les guides internes de l'édéage (tergite 9) sont prolongés vers Pavant, armés d'un crochet sur leur face interne (figs. 120—122); les paraméres (figs. 117—118) sont pointus, non aplatis en forme de spatule a l’apex. Troglosternus ecitonis Mann, 1925 T. ecitonis: Mazur 1984: 316; Helava et al. 1985: 261. Matériel: Panama: Canal Zone, Barro Colorado I., 25/01/59, bivouac abandonné de Eciton burchelli, 1 ex. comparé avec un paratype (R. L. Wenzel in litt.); idem, 3/03/1963 (R. D. Akre), avec Eciton hamatum, 1 ex., FMNH. Costa Rica: Punta S. Vito, Las Cruces, 1200 m, 22/02—3/03/1983 (B. Gill),:1-9, CHND! Em = 2,5 (2,3520) mm: Im 21022) (N = 3). Brésil: Cruzeiro do Sul, Acre, Janvier-Février 1988, piege d’interception, 1 ©, CHND. 2S mins le 4 ora T. ecitonis se distingue des autres especes du genre par la ponctuation non uniforme des pygidia, caractére signalé dans la description originale mais sans doute non considéré par Nouvelles d’Histeridae 371 Figs 115—125: Troglosternus dasypus, 115, pronotum et élytre gauche en vue dorsale, 116, sterna, 117, extrémité de l'édéage en vue dorsale, 118, édéage, profil droit, 119, tergite et sternite 8 du mále en vue ventrale, 120, tergite 9 du mále, profil droit, 121, apex du guide interne droit de lédéage en vue ventrale, 122, apex du tergite 9 (moitié gauche) en vue dorsale, 123, patte antérieure en vue dorsale, 124, patte médiane en vue ventrale, 125, patte postérieure en vue ventrale. Echelles 1 mm: a gauche pour les figs. 115—116, 123—125; a droite pour les autres figs. 372. N. Dégallier Reichensperger (cf. infra). Alors que la moitié ou les 3/4 basaux de ces structures sont densément ponctués ridés, leur partie apicale est lisse ou tres éparsément ponctuée. La position de la dent du bord externe du tibia postérieur par rapport a l’apex (fig. 130) semble étre un bon caractere pour séparer 7. ecitonis d'une espéce qui a été confondue avec lui et qui est décrite ci-apres. Les figs 126 et 127 montrent le dessin des stries dorsales et ventrales qui caractérisent cette espece. Pexemplaire récolté en Amazonie brésilienne (Etat de l'Acre) présente une striation du metasternum assez dense, comparable a celle observée chez T. dasypus. Letude de matériel complémentaire reste cependant nécessaire pour classer définitivement cet insecte. Troglosternus lisaevedouae Reichensperger, 1938 T. lisaevedouae: Mazur 1984: 316; Helava et al. 1985: 261. Matériel: Costa Rica: San José (H. Schmidt), 2 ex., FIOC; 3 o et 4 9, ZFMK. Lm = 2,8 PASS 20 (2532) mara UNS) Cette espece est facilement reconnaissable par la pilosité tres abondante et la ponctuation rugueuse et dense recouvrant presque toute la face dorsale du corps et notamment les pygidia. Par ailleurs, les metatibia sont plus étroits et paralléles que ceux des autres especes, avec la dent externe située la plus proche de l’apex (fig. 132). Les mesures de ce caractere et de l’allon- gement relatif du métatibia par rapport au corps montrent néanmoins un faible recouvrement entre 7: lisaevedouae et l’espece décrite ci-apres. Les genitalia males montrent plusieurs diffé- rences: expansions apicales des guides internes du tergite 9 (fig. 136) moins proéminentes que chez 7: dasypus; parameres élargis préapicalement-et aplatis en spatules a l’apex (fig. 137); genitalia femelles (fig. 139) paraissant caractéristiques du genre, le sternite 8 trilobé apicale- ment, les deux lobes externes setigeres; coxites (sternite 9) larges et setigeres apicalement. Troglosternus neoecitonis, nouvelle espece T. ecitonis: Reichensperger 1938: 79. Holotype: un © étiqueté: “Costa Rica San Jose H. Schmidt” (et. imp.). L = 3,7 mm; ] = 3,2 mm. ZFMK. Allotype: une Q étiquetée: “Costa Rica San Jose H. Schmidt” (ét. imp.). L =3,3 mm; 1=2,9 mm. ZFMK. Paratypes: tous de la méme provenance que les deux specimens précédents: un ex., FIOC; une Q et un O, ZFMK et CHND. Lm = 3,5 (3,3—3,7) mms) ln 3140932) mo (NS) Les spécimens-types de cette nouvelle espece ont été originellement rapportés a 7. ecitonis par Reichensperger (1938: 79) et Borgmeier a sa suite. Ils s’en distinguent cependant par le pygidium éparsément et uniformément ponctué et la partie basale du bord externe des métatibias (avant la dent) convexe (fig. 150). La dent n’est donc pas comprise entre deux concavités comme chez les autres especes. Lemplacement de cette dent est nettement plus proche de Papex que chez T! ecitonis. T. neoecitonis sp. nov. ne possede pas de stries A A Figs 126—139; figs 126—130: Troglosternus ecitonis — 126, pronotum et élytre gauche, 127, sterna, 128, patte antérieure en vue ventrale, 129, patte médiane en vue ventrale, 130, patte postérieure en vue ventrale; figs. 131—139: Troglosternus lisaevedouae — 131, sterna, 132, patte postérieure en vue ventrale, 133, patte médiane en vue ventrale, 134, patte antérieure en vue ventrale, 135, métatibia gauche en vue postérieure, 136, tergites 8 et 9 du mále en vue ventrale, 137, apex de l'édéage en vue dorsale, 138, édéage, profil droit, 139, genitalia femelles en vue ventrale (tergite 9 non représenté). Echelles 1 mm: a gauche pour les figs. 126—135; a droite pour les autres figs. 373 Nouvelles d’Histeridae U) Ma N iy a ...” \ \ \ e A \ \ fa N Ty, 5 EN 7 A, te he SSA AAA SS e] < ANTAS 374 N. Dégallier supplémentaires ramifiées sur le disque du métasternum (figs. 141 vs. 127) et ses stries élytrales dorsales sont plus nettes que chez 7: ecitonis (figs. 140 vs. 126). Compare a T. dasypus, T. neoecitonis n. sp. n’a pas les pro-, méso- et métasternum striés longitudinalement (figs. 116 vs. 141) et il possede des stries transversales sur le tiers postérieur du pronotum (fig. 140). Bien qu'il soit proche de 7. lisaevedouae par la structure des genitalia femelles (apex des lobes du sternite 8 large: fig. 142) et males (apex des parame£res dilatés: figs. 145—146), il s’en distingue par Paspect beaucoup moins pileux, l’absence d'un strie métasternale laterale interne (figs. 141 vs. 131) et la forme des tibias postérieurs (figs. 150 vs. 132). La fig. 152 montre la relation entre la position relative de la dent externe des tibias posterieurs et la longueur du corps pour chacune des quatre especes précédentes. Xenister Borgmeier, 1929 Xenister: Mazur 1984: 311; Helava et al. 1985: 362. Description complémentaire: Téte non ou peu carénée latéralement; vertex et front non élargis, le premier avec des stries divergentes; scape antennaire élargi; massue sclérotisée par endroits; labre triangulaire soudé au clypeus; mandibules non modifiées a la base. Pronotum sans stries en réseau, avec des fossettes. Elytres a 4eme strie dorsale droite, sans cótes. Prosternum a lobe non divise, sans processus antérieur; carene sans soies entre les stries carenales et latérales; largement incisé a la base; proépisternum et proépiméron sans touffes de soies. Mésosternum sans fossette. Métasternum avec une strie latérale. Tibias sans éperons. Genitalia femelles: sternite 8 avec les apodemes proximaux plus courts que la partie distale, avec des lobes sétigeres; tergite 9 sans prolongement proximal médian; sternites 9 avec des coxites setigeres, séparés mais soudés a leurs apodemes, dépourvus de styles. Genitalia máles inconnus. Xenister schwarzmaieri Borgmeier, 1929 X. schwarzmaieri: Mazur 1984: 311; Helava et al. 1985: 365. Holotype (par désignation originale): une Q étiquetée: “Xenister schwarzmaieri Borg. Holotype” (ét. man. a liséré noir) / “Eciton schlechtendali” (imp. a liséré noir): “Campinas Goiaz 15. V. 29. Schwarzmaier” (imp. a liséré noir) / “TYPUS” (imp. rouge) / “MUSEUM KOENIG BONN” (et. jaune imp.). L = 3,3 mm; 1=2,6 mm. ZFMK. | ES AA AS A Figs 140—151: Troglosternus neoecitonis n. sp., 140, pronotum et élytre droit, 141, sterna, 142, genitalia femelles en vue dorsale (tergite 9 non représenté), 143, tergite et sternite 8 du mále en vue ventrale, 144, tergite 9 du mále en vue ventrale, 145, apex de l’edeage en vue ventrale, 146, édéage, profil droit, 147, sternite 9 du mále en vue dorsale, 148, patte antérieure en vue dorsale, 149, patte médiane en vue ventrale, 150, patte posterieure en vue ventrale, 151, métatibia gauche en vue postérieure. Echelles 1 mm: a gauche en haut pour la fig. 140; a gauche au milieu pour les figs. 142—146; a droite pour les autres figs. Nouvelles d’Histeridae 375 376 N. Dégallier Discussion et conclusions La présente étude n’est qu’un préalable a un travail plus ample qui abordera les problemes posés (a) par l’identification des Hetaeriinae au niveau générique et (b) par l’etablissement d'une classification plus “naturelle”, basée sur la méthode phylo- génétique. Ces objectifs ne pourront en effet pas étre atteints sans les études morphologiques et taxonomiques (détection des synonymies, fixation des types etc.) de la majeure partie des taxons décrits. Certaines especes, dont les descriptions originales sont insuffisantes et les types inexistants, resteront cependant mal définies. Le seul moyen de lever ces incertitudes consiste en la récolte de matériel dans les localités-types lorsque celle-ci sont connues avec précision et non encore détruites par la civilisation! La récolte des insectes commensaux d’insectes sociaux n’a jamais été réputée aisée; cependant, nous pouvons des lors recommander deux techniques nous ayant donné des résultats inespérés en Amazonie brésilienne et qui seront décrites en détail dans un travail postérieur. Il s’agit de la récolte de nids entiers de fourmis légionnaires (Eciton spp.) (Dégallier & Gomy 1983) et de Putilisation d'un piege d’interception de vol [“window flight intercept trap” des auteurs anglo-saxons] (Peck & Davies 1980; Paulian 1985). 0,68 AT. dasypus 0,63 0,58 0,53 AT. ecitonis T. ecitonis 0,48 = (Acre) AT. neoecitonis Position de la dent externe du tibia postérieur AT. lisaevedouae 0,43 2,4 2,6 2,8 3 3,2 3,4 3,6 3,8 Longueur du corps (mm) Fig. 152: Position relative de la dent externe des tibias postérieurs (rapport de la distance entre la dent et la base sur la longueur du tibia) en fonction de la longueur du corps pour 4 especes du genre Troglosternus. Nouvelles d’Histeridae 7) Remerciements Nos plus sincéres remerciements s’adressent a tous ceux qui, d'une maniere ou d'une autre, nous ont aidé au long des années d'élaboration de ce travail. Conscient du risque de ne pas étre exhaustif, nous citerons néanmoins (par ordre alphabétique): R. J. W. Aldridge, G. Arriagada, M. E. Bacchus, A. O. Bachmann, N. Berti, O. V. Ferreira, B. D. Gill, Y. Gomy, F. Hieke, H. F. Howden, P. Kanaar, D. H. Kistner, A. F. Newton, Jr., W. L. Overal, H. Roer, F. Schwartz, J. J. Shaw, R. I. Storey, M. Uhlig et R. L. Wenzel. Zusammenfassung In diesem Beitrag werden neue Ergebnisse über Stutzkáfer (Histeridae) dargestellt. Zwei Sammelmethoden waren gleichsam ertragreich: eine Flug- bzw. Fensterfalle und das Schwem- men ganzer Ameisennester. Eine neue Reinigungstechnik nach P. Kanaar wurde erfolgreich für die Präparation von Genitalstrukturen eingesetzt; dafür wurde eine basische Lösung eines proteolytischen Enzyms benutzt. Die weiblichen Genitalien von 12 Spezies von Hetaeriinae wurden zum ersten Mal beschrieben. Neue Beschreibungen, Verbreitungsangaben und ókolo- gische Daten für 35 Spezies werden präsentiert, weitgehend basierend auf dem Studium der Typusexemplare. Die folgenden Taxa werden als neu beschrieben: Aristonister n. gen.; Colonides collegii guyanensis n. subsp.; Neocolonides howdeni n. gen. n. sp.; Troglosternus neoecitonis n. sp. Die folgenden neuen Kombinationen werden vorgeschlagen: Aristomorphus borgmeieri, A. latipes, A. ogloblini, A. perversus, Aristonister sericeus, Hemicolonides parvu- lus; Latronister breyeri. Das folgende neue Synonym wird vorgeschlagen: Oudaimosister Helava, in Helava et al. 1985 = Mesynodites Reichardt, 1924. Lectotypen und/oder Paralecto- typen wurden für die folgenden Spezies designiert: Aristomorphus borgmeieri, Aristonister sericeus, Cheilister sphaeroides, Clientister henrici, Colonides quadriglumis, Cossyphodister schwarzmaieri, Ecclisister bickhardti, Euxenister caroli, Hemicolonides parvulus, Parasynodi- tes suturacava, Pelatetister pretiosus, Teratosoma amphiphilus, Troglosternus dasypus. Die Gattung Hetaeriarchus Reichensperger wird als nomen nudum angesehen. References Arnett, Jr, R. H., G. A. Samuelson, J. B. Heppner, G. M. Nishida, J.C. Watt & R. E. Woodruff (1986): The insect and spider collections of the world. — E. J. Brill/Flora & Fauna Publications, Gainesville, FL, USA, 220 p. Bickhardt, H. (1916—1917): Histeridae, 1—112, 113—302. — In: Genera Insectorum, fasc. 166 a—b, éd. P. Wytsman, La Haye, 302 p. Borgmeier, T. (1929): Um novo histerideo mirmecophilo. — Bol. biol. (16): 85—091. Borgmeier, T. (1948): Zur Kenntnis der bei Eciton lebenden myrmekophilen Histeriden (Col.). — Rev. Ent. 19: 377—400. Bruch, C. (1923): Estudios mirmecológicos con la descripción de nuevas especies de dipteros (“Phoridae”) por los RR. PP. H. Schmitz y Th. Borgmeier y de una araña (“Gonylepti- dae”) por el Doctor Mello-Leitäo. — Rev. Mus. La Plata 27: 172—220. Bruch, C. (1926): Nuevos Histerideos ecitofilos (Col.). — Rev. Mus. La Plata 29: 17—33. Bruch, C. (1929): Neue myrmekophile Histeriden und Verzeichnis der aus Argentinien bekannten Ameisengáste. — Zool. Anz. 82: 421—437. Bruch, C. (1930): Histéridos huéspedes de Pheidole. — Rev. Soc. ent. Arg. 3 (1): 1—12. Bruch, C. (1931): Algunos mirmecofilos y termitofilos nuevos y poco conocidos de la Argen- tina. — Rev. Ent. 1: 387—395. Bruch, C. (1933): Coleópteros mirmecófilos de Misiones (Staph. Pselaph. Hister.). Rev. Ent. ll): 1237. Bruch, C. (1937): Coleópteros: mirmecófilos y termitófilos (Col. Hist. Staph. ). — An. Mus. Arg. Cien. nat., Entomologia, Buenos Aires 39 (publ. no. 153): 125—133. Code International de Nomenclature Zoologique, adopté par la XXe Assemblée générale de Union internationale des Sciences biologiques (1985): International Trust for Zoologi- cal Nomenclature & British Museum (Natural History), London, XX + 338 p. 378 N. Dégallier Dégallier, N. (1981): Etude des Euspilotus du groupe azureus [Coleoptera, Histeridae, Saprininae]. — Rev. fr. Ent. (N. S.) 3: 59—67. Dégallier, N. (1987): Analyse d'ouvrage. — Helava (J. V. T.), Howden (H. F.) & Ritchie (A. J.), 1985. A review of the New World genera of the myrmecophilous and termitophi- lous subfamily Hetaeriinae (Coleoptera: Histeridae). — Nouv. Revue Ent. (N. S.) 4: 220. Dégallier, N. & Y. Gomy (1983): Caracteres généraux et techniques de récolte des Coléop- teres Histeridae. — LEntomologiste 39: 9—17. Deuve, T. (1988): Les sternites VIII et IX de ’abdomen sont-ils visibles chez les imagos des Coléopteres et des autres Insectes Holométaboles? — Nouv. Revue Ent. (N. S.) 5: 21—34. Fragoso,S. A. (1985): The terminalia as a basis for the classification of Cerambycidae sub- families (Part I, terminology and genital morphology of Cerambyx cerdo). — Rev. bras. Ent. 29: 125—134. Gaedicke, H. (1984): Katalog der in den Sammlungen der Abteilung Taxonomie der Insek- ten des Institutes fir Pflanzenschutzforschung, Bereich Eberswalde (ehemals Deutsches Entomologisches Institut), aufbewahrten Typen — XXI (Coleoptera: Pselaphidae, Histeri- dae). — Beitr. Ent. 34: 441 — 462. Hedicke, H. (1923): Nomina nova. IV. — Deutsche ent. Z.: 431. Helava, J. V. T. (1978): A revision of the nearctic species of the genus Onthophilus Leach (Coleoptera: Histeridae). — Contr. Amer. ent. Inst. 15(5): 1—43. Helava, J. V. T, H. F Howden &A. J. Ritchie (1985): A review of the new world genera of the myrmecophilous and termitophilous subfamily Hetaeriinae (Coleoptera: Histeri- dae). — Sociobiology 10: 127—382. Kanaar, P. (1990): The use of a proteolytic enzyme in clearing genital preparations. — Ent. Ber. 50: 141—142. Lewis, G. (1885): New species of Histeridae, with synonymical notes. — Ann. Mag. nat. Hist. Sa 152 45652473) Lewis, G. (1891): On new species of Histeridae. — Ann. Mag. nat. Hist. ser. 6, 8: 380—40S. Lewis, G. (1913): On new species of Histeridae and Notices of others. — Ann. Mag. nat. Hist. ser 8, 2.281872 Mann, W. M. (1925): Guests of Eciton hamatum (Fab.) collected by Prof. W. M. Wheeler.. — Psyche 32: 166-177. Mazur, S. (1984): A world catalogue of Histeridae. Swiatowy katalog gnilików (Histeridae ). — Polskie Pismo ent./Bull. ent. Pol. 54 (3—4): 1—376. Paulian, R. (1985): Un piege a Col&opteres Crépusculaires. — LEntomologiste 41: 35—36. Peck,S. B. £ A. E. Davies (1980): Collecting small beetles with large-area “window” traps. — Coleopt. Bull. 34: 237—239. Reichardt, A. N. (1924): Zwei neue Histeriden aus Turkestan. — Ent.. Bl. 20: 162—166. Reichensperger, A. (1923): Neue südamerikanische Histeriden als Gäste von Wander- ameisen und Termiten. — Z. wiss. Ins.-Biol. 18: 243—252. Reichensperger, A. (1923a): Neue siidamerikanische Histeriden als Gáste von Wander- ameisen und Termiten. I. Systematischer Teil. — Mitt. schweiz. ent. Gesch. (Bull. Soc. ent. Suisse) 13: 313—336. Reichensperger, A. (1924): Neue siidamerikanische Histeriden als Gáste von Wander- ameisen und Termiten. II Teil. — Rev. suisse Zool. 31: 117—152. Reichensperger, A. (1926): Neue Beitráge zur Artenkenntnis und zur Lebensweise myrmekophiler Histeriden. — Verh. III int. Ent. Kongr. 2: 184552082 Reichensperger, A. (1929): Neue Ameisengáste und ein neuer Termitengast (Paus. Elist. Staph. ). — Ent. Bl. 25: 132-137. Reichensperger, A. (1931): Die Wirte der Mesynodites-Gruppe nebst Beschreibung neuer ecitophiler und termitophiler Histeridenarten (Coleopt. Hist.). — Zool. Jb. Abt. Syst. 61: 263 —284. Reichensperger, A. (1933): Neue ecitophile Histeriden aus Mittel- und Súdamerika (Cole- opt.). — Zool. Anz. 101: 299—309. Reichensperger, A. (1935): Beitrag zur Kenntnis der Myrmekophilen- und Termitophilen- fauna Brasiliens und Costa Ricas III (Col. Staphyl. Hist.). — Arb. morphol. taxon. Ent. Berlin-Dahlem 2 (3): 188—218. Nouvelles d’Histeridae 379 Reichensperger, A. (1936): Beitrag zur Kenntnis der Myrmekophilen- und Termitophilen- fauna Brasiliens und Costa Ricas. IV. (Col. Hist. Staphyl. Pselaph.). — Rev. Ent. 6: DDD 2A, Reichensperger, A. (1936a): Ergebnisse neuerer Forschungen an Ameisen- und Termiten- gásten. — Arb. phys. angew. Ent. 3: 186—192. Reichensperger, A. (1938): Beitráge zur Kenntnis der Myrmekophilen- und Termitophilen- fauna Brasiliens und Costa Ricas. V. (Col. Hist. Staph.). — Rev. Ent. 9: 74—97. Reichensperger, A. (1939): Beitráge zur Kenntnis der Myrmekophilenfauna Costa Ricas und Brasiliens VII, nebst Beschreibung der Königin von Eciton (Acamatus) pilosum. — Zool. Jb. Abt. Syst. 73: 261—300. Schmidt, J. (1889): Neue Histeriden aus Paraguay. — Berlin. ent. Zeit. 33: 317—324. Vomero, V. (1973): Troglobacanius n. gen. with four new species, a line of cave-adapted Mexican Histeridae (Coleoptera), 325—361. — In: Accademia nazionale dei Lincei, anno 370, quaderno N. 171 “Subterranean fauna of Mexico” -part II- Further results of the Italian zoological missions to Mexico, sponsored by the National Academy of Lincei (1969 and 1971), Roma. Vomero, V. (1977): Anapleus wenzeli, una nova specie di Dendrophilinae (Col. Histeridae) proveniente da una grotta del Messico meridionale, 341—348. — In: Accademia nazionale dei Lincei, anno 374, quaderno N. 171 “Subterranean fauna of Mexico” - part III- Further results of the Italian zoological missions to Mexico, sponsored by the National Academy of Lincei (1973 and 1975), Roma. Watkins Il, J. F. (1976): The identification and distribution of New World army ants (Dorylinae: Formicidae). — The Markham press Fund of Baylor University Press, Waco, Texas, 102 pp. Wenzel, R. L. (1944): On the classification of the histerid beetles. — Fieldiana, Zool. 28: SIS Wenzel, R. L. &H. Dybas (1941): New and little known neotropical Histeridae (Coleop- tera). — Fieldiana Zool. 22: 433 —472. Yelamos, T. (1989): Der Geschlechtsapparat von Hypocaccus brasiliensis (Paykull, 1811) (Coleoptera, Histeridae). — Ent. Nachr. Ber. 33: 135—136. Nicolas Dégallier, rés. Le Paradis esc. C, 14 imp. des Capucines 34000 Montpellier France. ) | | “a Bein u DT We 2 cio) bee m N 1 Y y 1) F) 007 N u Ye { : I 7 | A \ y 3 Ñ 4 i Y } 1 ' A 1 4 { j + i R ~ _— i E u de | + 1 + Fi - i iS) y = j { ' be ; id x ; = x fe af a ur Is N Bd. 47 S. 381—410 Bonn, September 1998 Die europäischen Arten der Cheilosia alpina-Gruppe (Diptera, Syrphidae) Claus Claussen Abstract. The European species of the Cheilosia alpina-group (Barkalov 1983a) (Diptera, Syrphidae) are revised. The following European species are recognized and described: Cheilosia alpina (Zetterstedt, 1838), Cheilosia chrysocoma (Meigen, 1822), Cheilosia kuznetzovae Skufjin, 1977, Cheilosia montana Egger, 1860, Cheilosia pictipennis Egger, 1860, and Cheilosia subpictipennis sp. n. Four new synonyms are identified: Cheilosia montana Egger, 1860 = Cheilosia braueri (Becker, 1894) syn. n. = Cheilosia thalhammeri (Szilädy, 1938) syn. n.; Cheilosia pictipennis Egger, 1860 = Cheilosia bureschi (Delkes- kamp, 1942) syn. n. = Cheilosia zmilampis Violovitsh, 1975, syn. n. Lectotypes are designated for Cheilosia alpina (Zetterstedt, 1838), Cheilosia phantoma (Zetterstedt, 1838), Cheilosia chrysocomoides (Strobl, 1910), Cheilosia montana Egger, 1860, and Cheilosia pictipennis Egger, 1860. A key for the European species of the Cheilosia alpina-group, new records, and a summary of their known distribution are presented. Key words. Diptera, Syrphidae, Cheilosia alpina-group, Europe, new species, new syno- nyms, lectotypes, distribution, key. Einleitung Für die sibirischen und fernöstlichen Arten der Gattung Cheilosia Meigen, 1822 (Eristalinae: Rhingiini) errichtete Barkalov (1983a) nach Merkmalen des Hypo- pygiums 8 Artengruppen: nigripes-, longula-, illustrata-, alpina-, velutina-, scanica-, sachtlebeni- und formosana-Gruppe. Die bekannten europäischen Cheilosia-Arten lassen sich den ersten 6 Gruppen zuordnen. Die beiden letztgenannten Gruppen sind hingegen ostpaläarktisch beziehungsweise orientalisch verbreitet. Das Auffinden einer unbeschriebenen Art aus der alpina-Gruppe wird zum Anlaß genommen, deren europäische Vertreter näher zu untersuchen. Ziel der vorliegenden Arbeit ist es, die alpina-Gruppe zu beschreiben, durch Revision der verfügbaren Typen die gültigen Taxa für den europäischen Raum festzustellen, die Synonymie zu klären und die bislang unzureichend bekannten Arten durch Wiederbeschreibungen neu abzugrenzen. Die verwandtschaftlichen Beziehungen der behandelten Arten untereinander und die Stellung der alpina-Gruppe innerhalb der Gattung sind nicht geklärt. Es scheint jedoch ein Schwestergruppenverhältnis zwischen der alpina-Gruppe und der illu- strata-Gruppe (sensu Barkalov 1983a) vorzuliegen. Die wenigen Beobachtungen zur Bionomie der besprochenen Arten lassen ver- muten, daß sich die Larven als Minierer in Stengeln und/oder Wurzeln von Apiaceae . oder Asteraceae entwickeln. Material und Methode Das untersuchte Material entstammt den folgenden Institutionen und privaten Sammlungen (vorangestellt, das im Text verwendete Kürzel): MZHF, Zoological Museum, University of 382 C. Claussen Helsinki (G. Stähls); MZLS, Musée zoologique de Lausanne (Prof. Dr. P. Goeldlin de Tiefe- nau); MZLU, Lunds Universitet, Zoologiska Museet, Lund (Dr. R: Danielsson); NHMW, Naturhistorisches Museum Wien (Dr. R. Contreras-Lichtenberg); NMBA, Naturhistorisches Museum des Benediktinerstiftes Admont (Prior P. B. Hubl); SUEL, Bakonyi Természettudo- mányi Muzeum, Zirc (Dr. S. Toth); ZFMK, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Bonn (Dr. H. Ulrich); ZMAN, Zoölogisch Museum Amsterdam (Dr. V. S. van der Goot & Dr. B. Brugge); ZMHB, Museum fiir Naturkunde der Humboldt-Universitat, Berlin (Dr. H. Schumann); ZMUH, Zoologisches Museum Universitat Hamburg (Prof. Dr. R. Abraham); ZSMC, Zoologische Staatssammlung Miinchen (W. Schacht); AB, Coll. Dr. A. V. Barkalov, Zoological Museum, Academy of Sciences, Novosibirsk; AM, Coll. Dr. A. Marcos- Garcia, Universidad de Alicante; AV, Coll. Dr. A. Vujic, University of Novi Sad; CC, Coll. C. Claußen, Flensburg; CK, Coll. C. F. Kassebeer, Kiel; DD, Coll. D. Doczkal, Malsch; ET, Coll. E. Torp, Jelling; FG, Coll. F. Geller-Grimm, Frankfurt/M.; FM, Coll. Dr. F. Malec, Kassel; GS, Coll. G. Schwendinger, Dornbirn; HB, Coll. H. Bartsch, Járfálla; JL, Coll. Dr. J. A. W. Lucas, Rotterdam; JS, Coll. J-H. Stuke, Freiburg; MD, Coll. Dr. M. Daccordi, Verona; MH, Coll. M. Hauser, Darmstadt; MS, Coll. Dr. M. C. D. Speight, Dublin; PL, Coll. P.-W. Löhr, Mücke-Merlau; PR, Coll. Dr. P.-F. Róseler, Würzburg; RB, Coll. R. Borcherding, Kleinschmalkalden; RT, Coll. R. Treiber, Freudenstadt; ST, Coll. Dr. S. Töth, Zirc; TN, Coll. Dr. T. R. Nielsen, Sandnes; UB, Coll. U. Buchholz, Freiburg; US, Coll. U. Schmid, Stuttgart; VB, Coll. V. Brädescu, Bukarest. Die Abbildungen der Genitalstrukturen wurden nach mazeriertem Material in Glyzerin mit Hilfe einer Mikroprojektion angefertigt. Die morphologischen Termini für die Strukturen des Hypopygiums folgen weitgehend McAlpine (1981), teilweise Gaunitz (1960), abweichend davon werden jedoch die gelenkig mit dem Hypandrium verbundenen Anhänge als Gonostyli gedeutet. Abkürzungen morphologischer Termini: f = Femur, p = Bein, S = Sternit(e), T = Tergit(e), t = Tibia; sonstige Abkürzungen nach McAlpine (1981). Abb. 1: Cheilosia illustrata (Harris, 1780), Hypopygium lateral. Abkürzungen: bph = Basiphal- lus, ce = Cercus, dphs = Sklerit des Distiphallus, eja = Ejakulatorapodem, ep = Epandrium, godl = dorsaler Lobus des Gonostylus, gon = Gonostylus, govl = ventraler Lobus des Gono- stylus, hp = Hypandrium, klm = Kammlamelle des Surstylus, pha = Phallapodem, spk = Spermakanal, spp = Spermapumpe, sst = Surstylus. Die europäischen Arten der Cheilosia alpina-Gruppe 383 Diagnose der alpina-Gruppe o 0: MittelgroBe bis große Arten (9,5—13 mm) mit breitem Abdomen und dichter, aufrechter, meist langer Körperbehaarung. Seiten des Gesichts ohne längere Haare. Augen lang behaart. Antennengruben durch einen medialen Fortsatz der Lunula getrennt. Thorax und Abdominaltergite überwiegend glänzend, schwarz bis schwarz- braun; Seiten des Scutums vor der Quernaht, Pleuren sowie das 1. Abdominaltergit höchstens mit zarter Bestäubung (nicht dicht weißgrau bestáubt, wie in der ¿llustrata- Gruppe). Hinterrand des Scutellums und meist auch Postalarcalli ohne stärkere Borsten. Katepisternum (Sternopleuron) durchgehend lang behaart. Beine hell und dunkel gefärbt, zumindest die Spitzen der Femora und der basale Abschnitt der Tibien hell. Flügel zumeist mit mehr oder weniger deutlichem braunen Mittelfleck zwischen Subcostalzelle und Basis der Diskalmedialzelle, und zwar in beiden Geschlechtern meist deutlich bei pictipennis und subpictipennis, bei den übrigen Arten der Gruppe oft nur bei den Q 9 mehr oder weniger ausgeprägt. Sternite glán- zend oder von der Hinterleibsspitze zur Basis des Abdomens hin zunehmend grau bestäubt. ©: Hypopygium: Surstyli sehr unterschiedlich: meist kurz, mit ausgeprägter Kamm- lamelle; zuweilen sehr stark verlängert, und dann die Kammlamelle fast vollständig fehlend (chrysocoma, pictipennis); intermediäre Ausprägungen des Surstylus finden sich bei einigen ostpaläarktischen Arten (gorodkovi, nudiseta, subalbipila). Ventraler Lobus des Gonostylus aufrecht, distal gelegentlich leicht nach innen gebogen (aber nicht im basalen Drittel auffällig dorsad gekrümmt wie in Abb. 16h). Die lateralen Sklerite des Distiphallus umschließen diesen kapselartig und laufen ventral in 2 basale und 2 apikale Zipfel aus. Q: Das 3. Fühlerglied kreis-rundlich bis kurzoval, gelegentlich mit einer Oberecke (chrysocoma) und/oder auffällig vergrößert (chrysocoma, pictipennis, subpictipen- nis). Die Seitenfurchen der Stirn breit, nach innen bogenförmig begrenzt, zusammen etwa die Hälfte bis 2/3 der Stirnbreite einnehmend. Die 6 folgenden Arten der alpina-Gruppe sind aus Europa bekannt: Cheilosia alpina (Zetterstedt, 1838) = C. akela Violovitsh, 1973 Cheilosia chrysocoma (Meigen, 1822) = C. phantoma (Zetterstedt, 1838) = C. chrysocomoides (Strobl, 1910) Cheilosia kuznetzovae Skufjin, 1977 Cheilosia montana Egger, 1860 = C. trisulcata (Becker, 1894) =. braueri (Becker, 1894). Syn. n. =. thalhammeri (Szilady, 1938). Syn. n. Cheilosia pictipennis Egger, 1860 = C. innominata (Becker in Bezzi & Stein, 1907; als Varietát von pictipennis) =. bureschi (Delkeskamp, 1942). Syn. n. =C. zmilampis Violovitsh, 1975. Syn. n. Cheilosia subpictipennis sp. n. =. pictipennis Barkalov in Violovitsh, 1983, nec Egger 384 E. Claussen Nach den Beschreibungen und Abbildungen in Stackelberg (1963), Peck (1971), Peck in Stackelberg & Peck (1979) und Barkalov in Violovitsh (1983) müssen auch die folgenden ostpaläarktischen Arten zur alpina-Gruppe gerechnet werden: C. asio- montana Peck, 1971, C. balu Violovitsh, 1966, C. gorodkovi Stackelberg, 1963, C. heptapotamica Stackelberg, 1963, C. kaszabi Peck in Stackelberg & Peck, 1979, C. lutea Barkalov, 1979, C. nudiseta (Becker, 1894), C. subalbipila Violovitsh, 1956 und C. tibetana Stackelberg, 1963. Abb. 2—8: Cheilosia, Surstyli lateral. — 2 C. chrysocoma, 3 C. pictipennis, 4 C. alpina, 5 C. montana (Lectotypus), 6 C. montana (Österreich, Alpen), 7 C. kuznetzovae, 8. C. sub- pictipennis sp. n. Die europáischen Arten der Cheilosia alpina-Gruppe 385 Verwandtschaft der alpina-Gruppe Die Arten der alpina-Gruppe stehen habituell und nach dem Bau des Hypopygiums den Arten der i//ustrata-Gruppe (sensu Barkalov 1981, 1983a) sehr nahe. Beide Grup- pen besitzen kapselartig den Distiphallus umschließende, ventral in 2 basale und 2 apikale Zipfel auslaufende Distiphallussklerite, ein vermutlich synapomorphes Merkmal, das ein Schwestergruppenverhältnis (alpina + illustrata-Gruppe) wahr- scheinlich erscheinen läßt. Weitere gemeinsame Grundplanmerkmale sind vermut- lich: a) ein dunkler Mittelfleck zwischen Flügelstigma und Basis der Discoidalzelle, b) sehr breite, nach innen bogig verlaufende Seitenfurchen auf der Stirn der © Q und c) ein auf den Seiten behaartes Gesicht. Als Autapomorphie der illustrata-Gruppe ist möglicherweise der nicht endständige dorsale Lobus des Gonostylus anzusehen. In der alpina-Gruppe werden die ventral ausgerichteten Zipfel der Sklerite des Disti- phallus und der Verlust der Gesichtsbehaarung als apomorphe Gruppenmerkmale betrachtet. Bestimmungsschlüssel für die europäischen Arten und Unterarten der Cheilosia illustrata- und -alpina-Gruppe 1. TI lateral und Basis von T2 lateral dicht grauweiß bestáubt (bei 9 Q deutlicher als bei © ©). Flügel mit braunem Mittelfleck (bei frischen Tieren gelegentlich undeutlich). Eee er ee 2 — Ti lateral und Basis von T2 lateral ohne dichte grauweiße Bestäubung, höchstens T1 und T2 medial schwach grau bestäubt. Flügel mit oder ohne Mittelfleck ............... 3 2. Gesichtsseiten gewöhnlich lang behaart, selten mit nur vereinzelten langen Haaren oder ganz kahl. 3. Fühlerglied schwarz bis schwarzbraun, gelegentlich mit rötlicher Basis. Kör- perbehaarung auffällig dreifarbig, überwiegend fahlgelb, aber Scutum hinter der Quer- naht und T3 mit mehr oder weniger vollständiger schwarzer Haarbinde, T4 und T5 mit ici GIA a A AS ess ude nee oe Raa ee illustrata illustrata (Harris) — Gesichtsseiten unbehaart. 3. Fühlerglied rötlichbraun mit dunkler Spitze. Kórperbehaa- rung farblich sehr variabel (Kaukasus).......... illustrata portschinskiana Stackelberg 2. ADICTA “CHC AR O crn eta rire > a ER 4 = JPR CUSC WO SBR A A A A A E A ER 9 4. Hypopygium: Surstylus mehr als 3mal so lang wie breit (Abb. 2—3, 9—10). ....... 5) — Hypopygium: Surstylus etwa 2mal so lang wie breit (Abb. 4—8, 11—15). .......... 6 5. Basalhälfte von S2 auch medial mit lang abstehenden Haaren. Hypopygium: Surstylus (Lateralansicht) sich apikal verjiingend (Abb. 2). .............. chrysocoma (Meigen) — Basalhälfte von S2 medial ohne lang abstehende Haare oder die Haare hier kürzer und spárlicher als auf den Seiten. Hypopygium: Surstylus (Lateralansicht) apikal breit abge- PEE ADOS A A lA A NE A pictipennis Egger 6. Hypopygium: Beide Loben des Gonostylus schlank und etwa gleich lang (Abb. 16€). O O a o e RE Ree subpictipennis sp. n. — Hypopygium: Ventraler Lobus des Gonostylus länger und schlanker als der dorsale CA LEDC A E A N es ee 7 7. 3. Fühlerglied fast stets schwarz oder schwarzbraun (bei unausgefärbten Exemplaren auch rotbraun), etwa so lang wie breit, distal gerundet (Abb. 18a—b). Hypopygium: Gonosty- SLE me in ANS) NA ee Jen ee ee ee SE 8 — 3. Fühlerglied rotbraun, etwas länger als breit (Abb. 18c). Hypopygium: Gonostylus wie Ad Zentraltubland). ooo a ia ta ee kuznetzovae Skufjin 8. Katepisternum ventral, p3-Coxen, f3 anterodorsal und S8 (Prágenitalsegment) über- wiegend schwarz behaart (gilt für Tiere aus den Alpen. — Bei einzelnen Stücken aus den Karpaten und dem vorliegenden Exemplar aus Spanien ist die Behaarung deutlich heller). Flügel: cual schmaler oder gleich breit wie die basale Hälfte der dm (Abb. 20). Gesicht 386 10. E-Claussen (Abb. 42) meist etwas stárker herabgezogen (variabel). Hypopygium: Kammlamelle des Surstylus (Ventralansicht) flach (Abb. 12—13). — (Alpen, Karpaten, Tatra-Gebirge, Kantabrisches Gebirge; oberhalb der Waldgrenze). .................. montana Egger Katepisternum ventral, p3-Coxen, f3 anterodorsal und S8 überwiegend hell behaart, aber distales !/4 von f3 anterodorsal mit längeren schwarzen Haaren. Flügel: cual etwas brei- ter als die basale Hälfte der dm (Abb. 22). Gesicht (Abb. 41) nicht so stark herabgezogen. Hypopygium: Kammlamelle des Surstylus (Ventralansicht) meist etwas kürzer und stärker nach außen gewölbt (Abb. 11). — (Nordskandinavien, Karelien). ................... DA ERS SRH). SITES E ESE RE JU alpina (Zetterstedt) 3. Fühlerglied höchstens 2mal so breit wie das 2. Fühlerglied, fast stets schwarz bis schwarzbraun (bei unausgefärbten Exemplaren gelegentlich heller) (Abb. 19a—b). Flügel: MI spitzwinklig zu R4+5 verlaufend (Abb; 29, 2.025.225 55-65 oe eee 10 3. Fühlerglied 2.5—3mal so breit wie das 2. Fühlerglied, für gewöhnlich hell bis düster orange (selten schwarzbraun) (Abb. 19c—f). Flügel: M1 steiler zu R4+5 verlaufend (Abb. 25,26)... sl a nette ahnt Sl. A We 11 Katepisternum ventral, Scutum lateral in Höhe der Flügelwurzeln, Scutellum und Hinter- coxen (anterolateral) meist mit schwarzen, selten mit ganz hellen Haaren. Distales !/3 der f3 anterodorsal mit schräg abstehenden längeren schwarzen, selten hellen Haaren (Abb. 27). Flügel: cual im basalen Abschnitt schmal, nicht deutlich breiter als Basis der dm (Abb: 21) ash ars ehe a che ok Cn eee montana Egger 10 Abb. 9—10: Cheilosia, Epandrium mit Surstyli und Cerci dorsal. — 9 C. chrysocoma, 10. C. pictipennis. Die europäischen Arten der Cheilosia alpina-Gruppe 387 — Katepisternum ventral, Scutellum und Hintercoxen hell behaart. Scutum lateral nur mit vereinzelten schwarzen Haaren. Distales !/3 der f3 anterodorsal mit feinen, hellen Haa- ren, diese stärker aufgerichtet (Abb. 28). Flügel: cual im basalen Abschnitt meist breiter als Basis der dm (Abb. 23). (Abgrenzung von montana-Q 9 gelegentlich schwierig)... o A RO alpina (Zetterstedt) 11. Hinterrand von T5 (Dorsalansicht) gerundet (Abb. 34). 3. Fühlerglied hellorange, oft mit deutlicher Oberecke (Abb. 19e), gelegentlich (nordspanische Population) distal gebráunt und gerundet (Abb. 19e). Kórperbehaarung meist auffállig dicht und fuchsrot (lichter und blasser in Populationen aus Nordskandinavien und Nordspanien)................... ee RER RE N NR ee Tezhee WO chrysocoma (Meigen) 15 Abb. 11—15: Cheilosia, Epandrium mit Surstyli und Cerci dorsal. — 11 C. alpina, 12 C. mon- tana (Lectotypus), 13 C. montana (Österreich, Alpen), 14 C. kuznetzovae, 15 C. subpictipennis sp. n. 388 C. Claussen — Hinterrand von T5 (Dorsalansicht) gerade abgestutzt (Abb. 36), mit den Seitenrändern nicht zu einem Bogen verbunden. 3. Fühlerglied rotorange bis braunrot (selten dunkler), mit oder ohne Oberecke (Abb. 19c, d, f). Kórperbehaarung orange- bis blaßgelb, auf den hinteren Abdominaltergiten gelegentlich auch schwarz. . ............ 2.22. 12 12. Relativ kürzer behaarte Art: Die abstehende Behaarung medial auf T3 deutlich kürzer als der Durchmesser des p2-Basitarsus; Behaarung auf dem Scutellum wie geschoren, kürzer als Durchmesser der f3. Basale Hälfte der f3 anteroventral mit vereinzelten längeren, abstehenden Haaren, die kürzer sind als der Durchmesser von f3 (Abb. 31). Apikales !/3 der f3 anteroventral hell behaart, gelegentlich mit vereinzelten kurzen schwarzen Börst- Chen eee odo ocio kuznetzovae Skufjin — Relativ länger behaarte Arten: Die abstehende Behaarung medial auf T3 deutlich länger als Durchmesser des p2-Basitarsus; Behaarung auf dem Scutellum struppig abstehend und länger als der Durchmesser der f3. Basale Hälfte der f3 anteroventral und apikale Hälfte der f3 anterodorsal mit langen, abstehenden Haaren, diese deutlich länger als Durchmesser von f3 (Abb. 29—30). Apikales !/3 der f3 anteroventral mit kurzen hellen und schwarzen Haaren (variabel): .......2...2..0.. 22... So teins 13 13. Behaarung der basalen Hälfte von S2 medial schmal unterbrochen oder hier kürzer als an den Seiten; Haare auf der basalen Hälfte von S3 medial meist niedergedrückt. Die kurze Behaarung auf der Ventralseite von f3 annähernd gleich lang und durchschnittlich kurzer als */3 des Dürchmessers von 13 (Abb: 30) Er re eee pictipennis Egger — Behaarung der basalen Hälfte von S2 medial nicht unterbrochen und hier meist deutlich länger als an den Seiten. Haare auf der basalen Hälfte von S3 auch in der Mitte aufrecht (oft aber nur in einem schmalen Saum). Die kurzen Haare auf der Ventralseite von f3 ungleich lang und zumeist länger als !/3 des Durchmessers von f3 (Abb. 29). (Die Abgrenzung von pictipennis-Q 9 gelegentlich schwierig). ....... subpictipennis sp. n. Beschreibung und Diskussion der Arten Cheilosia alpina (Zetterstedt, 1838) Eristalis alpina Zetterstedt, 1838; Ins. Lapp.: 611; loc. typ.: Schwedisch Lappland: Jockeltind. Cheilosia akela Violovitsh, 1973; Trudy Biol. Inst., Sib. Otd. Akad. Nauk SSSR: 145; loc. typ.: West- sibirien: Altai (Synonymie durch Barkalov 1983b: 633). e f g h Abb. 16 a—h: Cheilosia, Gonostyli lateral. — a C. illustrata, b C. alpina, c C. montana (Lecto- typus), d C. kuznetzovae, e C. subpictipennis sp. n., f C. chrysocoma, g C. pictipennis, h C. barbata. Die europäischen Arten der Cheilosia alpina-Gruppe 389 Typen: Eristalis alpina Zetterstedt: Lectotypus ©, hiermit festgelegt. Etikettierung: „E. alpina O. Jockelt.“, in Dipt. Scand. Coll. MZLU. Erhaltungszustand: Vollständig und gut erhalten. — Paralectotypen: 1 © ,,E. alpina ©. Jockeltind“ in Ins. Lapp. Coll. MZLU; 1 9 „E. alpina 2. Jockelt.“, in Dipt. Scand. Coll. MZLU; 1 © ohne Daten, in Dipt. Scand. Coll. MZLU. In der Sammlung Zetterstedt befinden sich die vier oben aufgeführten Exemplare, die als Syntypen aufgefaßt werden. Lectotypus und Paralectotypen wurden entsprechend etiket- tiert. Weiteres Material: Schweden: 1 ©, L. Brundin leg., Torne Lappmark, Abisko, 1928 (ET). 1 ©, J. Struve leg., Torne Lappmark, Gállivare, Kakerjaure RN1610/7502, 29. 6. 1982 (HB). 2 OC, S. Gaunitz leg., Lappland, Sorsele (weitere Angaben unleserlich), 30. 7. 1929 (MZLU). 1 9, S. Gaunitz leg., Ly Lpm. (= Lycksele Lappmark), Ammarnäs, 26. 7. 1971 (MZLU). Nor- wegen: 1 ©, T. R. Nielsen leg., STJ, Oppdal, Kongsvoll EIS 79, 24. 6. 1988 (TN); 1 © gleicher Fundort, 26. 6. 1988 (TN); 2 o © gleicher Fundort, 27. 6. 1988 (TN/CC); 1 ©, Arne Nielsen leg., Kongsvoll STI, Oppdal, 18. 7. 1966 (TN); 1 9, T. R. Nielsen leg., Spiterstulen On: Lom, 1. 7. 1974 (CC). Russische Föderation: 14 © © 3 9 9, Mikkola, Hippa & Jalava leg., SW- Altai, 15 km S Katanda, Bert-Kum, 2000—2500 m, 10.—14. 7. 1983 (MZHF). Verbreitung: Norwegen, Schweden, Finnland (Frey 1941), Russische Fóderation: Karelien, Nordwest-Sibirien (Stackelberg 1970), óstlich bis zum Baikalsee, Altai-Gebirge. Mongolei (Peck 1977). — Zu dem Nachweis aus Nordspanien (Marcos-Garcia 1989) siehe unter C. mon- tana! Abb. 17 a—g: Cheilosia, Aedoeagus mit Phallapodem, lateral, e —f mit Gonostylus (schraf- fiert). — a C. illustrata, b C. alpina, c C. montana, d C. kuznetzovae, e C. subpictipennis sp. n., f C. chrysocoma, g C. pictipennis. 390 e3elanssen . 0,5 mm d e Abb. 18 a—f: Cheilosia © ©, drittes Glied des linken Fühlers mit Arista von außen. — a C. alpina, b C. montana, c C. kuznetzovae, d C. subpictipennis sp. n., e C. chrysocoma (Mitteleuropa), e” C. chrysocoma (Nordspanien), f C. pictipennis. Abb. 19 a—f: Cheilosia 9 Q, drittes Glied des linken Fühlers mit Arista von außen. — Zuordnung der Arten wie in Abb. 18. | Die europäischen Arten der Cheilosia alpina-Gruppe 391 Beschreibung Diagnose: Morphologisch sehr áhnlich montana; Gesicht (Abb. 41) meist etwas stárker vorspringend und weniger weit herabgezogen; Flügel: cual im basalen Abschnitt meist etwas breiter als der entsprechende Abschnitt der dm. Körperbehaarung insgesamt heller als bei montana, Katepisternum ventral ohne schwarze Haare. o": Körper schwarz bis dunkel-olivbraun, fein punktiert. Kopf (Abb. 41): Stirn mit deutlicher Mittelfurche, nur an den Augenrändern schmal grau bestäubt, lang schwarz behaart, selten mit einzelnen hellen Haaren vermischt. Gesicht ausgedehnt graugelb bestäubt, seitlich über dem Mundrand schwarz glänzend. Wangen (Genae) wenig breiter als das 2. Fühlerglied, oben wie das Gesicht bestäubt und hell behaart, nach unten hin stärker glänzend und zunehmend auch mit schwarzen Härchen. Scheitel lang schwarz oder schwarz und hell behaart. Fühler (Abb. 18a) schwarz, bei unausgefärbten Stücken rötlichbraun, 3. Fühlerglied im Umriß etwas varia- bel: meist distal gerundet, aber gelegentlich auch mit undeutlicher Oberecke; Arista fast nackt. Lunula gelblich bis bräunlich. Augen lang hell behaart. Thorax: Behaarung des Scutums ungleich lang, abstehend graugelb bis fuchsrot, zwischen den Flügelwurzeln eine mehr oder weniger ausgeprägte schwarze Haarbinde, die sich seitlich craniad in die helle Behaarung hineinziehen kann. Scutellum ganz oder überwiegend hell behaart, median gelegentlich mit kürzeren schwarzen Haaren. Behaarung der Pleuren über- wiegend hell, meist etwas blasser als auf dem Scutum, aber Anepisternum, Anepimeron und dorsale Hälfte des Katepisternums in unterschiedlicher Ausdehnung auch mit schwarzen Haaren; ventraler Abschnitt des Katepisternums hell behaart (Unterschied zu typischen montana). — Flügel (Abb. 22) ohne deutliche dunkle Querbinde, aber an den Queradern r-m cual 20 21 22 23 | 1mm Abb. 20—23: Cheilosia, Flügelausschnitt. — 20 C. montana ©, 21. C. montana Q, 22 C. alpina ©, 23 C. alpina 9 (dm = Discalmedialzelle, cual = vordere Cubitalzelle). 392 C. Claussen und m-cu oft auffállig gebráunt, cual etwas breiter als basale Hálfte von dm (relativ kon- stanter Unterschied zwischen skandinavischen alpina- und alpinen montana-Populationen). Squamulae weiflich, Saum gelblich; Halteren gelbbraun mit dunklem Kópfchen. — Beine: Femora schwarz, apikal sehr schmal gelblich; Tibien im basalen !/3 und apikal schmal gelb- braun, mit zwischenliegendem schwarzem Ring (gelegentlich undeutlich); Tarsen, bis auf die beiden basalen Glieder der p2, dorsal geschwárzt, ventral oft heller. Behaarung der Beine variabel, gelblich und schwarz (insgesamt aber deutlich heller als bei montana): f3 überwie- gend hell behaart, nur ventral mit schwarzen Borsten und in der apikalen Hälfte anterolateral sowie im apikalen !/3 auch anterodorsal mit schwarzen Haaren. p3-Coxa überwiegend hell behaart, nur apikaler Zipfel mehr oder weniger schwarzhaarig. 25 26 Abb. 24—26: Cheilosia, Flügelausschnitt. — 24 C. montana 9, 25 C. subpictipennis sp. n. Q, 26. C. kuznetzovae Q (Winkel in Abb. 24 und 25 vgl. Text). Die europäischen Arten der Cheilosia alpina-Gruppe 393 Abdomen: Kurz und breit, überwiegend glänzend, T2 (und T3) medial matt: T1—4 lang abstehend rötlich bis blaßgelb behaart, Hinterecken von T2—4 gelegentlich mit einzelnen schwarzen Haaren. S1—2 sowie Seiten von S3—4 lang abstehend hell behaart; S3 medial mit kurzer, schräg abstehender heller oder heller und schwarzer Behaarung, auf S4 sind diese kurzen Haare überwiegend schwarz. Hinterrand von S4 auch mit längeren schwarzen Haaren. S8 (Prägenitalsegment) meist überwiegend hell behaart. — Hypopygium (Abb. 4, 11, 16b, 17b): Sehr ähnlich wie bei montana, aber Kammlamelle des Surstylus (Abb. 11) zumeist etwas kürzer und stärker nach außen gewölbt. Maße: Körper 9,7—11,7 mm; Flügel 8,5—10,0 mm. O: Bis auf die üblichen Geschlechtsunterschiede dem & sehr ähnlich, aber die schwarze Körperbehaarung und die dunkle Färbung der Beine reduziert. Kopf (Abb. 40) hell behaart, nur am unteren Augenrand mit ganz vereinzelten schwarzen Haaren. Seitenfurchen der Stirn grob punktiert und mit leicht bogenförmiger Runzelung, an ihrer breitesten Stelle zusammen 1/2—2/3 der Stirnbreite einnehmend. Fühler (Abb. 19a) schwarz, die beiden basalen Glieder und Basis von Glied 3 gelegentlich rötlichbraun; 3. Glied apikal gerundet, gelegentlich mit Andeutung einer Oberecke. Thorax: Hell behaart, gelegentlich Seiten des Scutums vor der Quernaht und neben den Flügelwurzeln mit vereinzelten schwarzen Haaren; Pleuren lang und fein hell behaart, ohne schwarze Haare. — Flügel (Abb. 23): Basalhälfte von cual meist deutlich breiter als dm. — Beine überwiegend hell behaart, nur Tarsen in unterschiedlicher Ausdehnung mit kurzen schwarzen Härchen und Borsten sowie fl und f2 im apikalen !/3 posteroventral und f3 im apikalen !/3 anteroventral mit schwarzer und heller Behaarung. Die anterodorsale Behaarung auf f3 (Abb. 28) hell (bei montana zumindest teilweise schwarz). Hintercoxen hell behaart (bei montana zumindest anterolateral meist mit schwarzen Haaren). Abdomen: Tergite glänzend, nur Basis von T2 medial matt, ohne schwarze Haare. Maße: Körper 9,5—10,5 mm; Flügel 8,7—9,0 mm. Variabilität: Nach dem vorliegenden Material variiert die Art in Nordeuropa kaum. Bei den Tieren aus dem Altai-Gebirge ist im Flügel die Weite der cual im Verhältnis zur Weite von dm deutlich variabler als in europäischem Material, so daß sich hier bezüglich dieses Merk- mals die Unterschiede zu montana verwischen. Präimaginalstadien: nicht beschrieben. Cheilosia chrysocoma (Meigen, 1822) Syrphus chrysocomus Meigen, 1822; Syst. Beschr. 3: 280; loc. typ.: ? Stolberg bei Aachen. Eristalis phantoma Zetterstedt, 1838; Ins. Lapp.: 611; loc. typ.: Norwegen: Raschstind, an der Mündung des Malangerfjordes. Chilosia chrysocomoides Strobl, 1910; Mitt. naturw. Ver. Steierm. 46: 101; loc. typ.: Österreich: Steier- mark, Admont. Typen: Syrphus chrysocomus Meigen: Typen nicht untersucht. Von den ursprünglich vier Syntypen befindet sich noch 1 © in Coll. Meigen (Paris), das zweifelsfrei zu der vorliegenden Art gehört (M. C. D. Speight briefl., siehe auch Becker 1902). Eristalis phantoma Zetterstedt: Lectotypus ©, hiermit festgelegt. Etikettierung: „E. Phan- toma ©. Raschstind.“, in Dipt. Scand. Coll. MZLU. Erhaltungszustand: Arista des rechten Fühlers fehlt. — Paralectotypen: 1 © „E. Phantoma O Raschstind“, in Ins. Lapp. Coll. MZLU; 1 o „Dovre“, darüber zwei kleine, rechteckige Zettel ohne Angaben, in Dipt. Scand. Coll. MZLU. Chilosia chrysocomoides Strobl: Lectotypus ©, hiermit festgelegt. Etikettierung: „Admont Steierm. Strobl.“, in Coll. Strobl NMBA. Erhaltungszustand: Sehr gut. — Paralectotypus 1 Q mit denselben Angaben wie der Lectotypus, in Coll. Strobl NMBA. — Der Lectotypus ist ein frisches Exemplar von chrysocoma. Der Paralectotypus ist ein unausgefärbtes Exemplar 394 Gy Cliauissien YOU VE WIE SINT 28 INN N AWS N \ \\ NN V Y YA) N Abb. 27—31: Cheilosia Q 2, f3 lateral von vorn . — 27 C. montana, 28 C. alpina, 29 C. sub- pictipennis sp. n., 30 C. pictipennis, 31 C. kuznetzovae. Die europáischen Arten der Cheilosia alpina-Gruppe 395 von pictipennis, mit fehlender Flügelbinde und langer schwarzer Behaarung am Hinterrand von T2 und auf T3—5. Lecto- und Paralectotypen wurden als solche etikettiert. Weiteres Material: Schweden: 1 9, J. A. W. Lucas leg., Torne Lappmark, Abisko, 485 m, 5. 7. 1980 (JL); 1 9, Jeekel en Piet leg., Torne Lappmark, Abisko, 10. 7. 1955 (ANZM); Il Qe JB) Gaunitz leg., Lappmark, Sorsele, 13. 7. 31 (MZLU). Norwegen: 1 9, T. R. Nielsen leg., Astereit, HOy: Asane, 30. 5. 1971 (CC); 1 ©, A. Fjelberg leg., HOy: Fana, 1. 6. 1968 (CC); 2 oo, T. R. Nielsen leg., Isdalen, HOy: Bergen, 1. 6. 1963 (ZMAN/CC); 3 oo, J. A. W. Lucas leg., STI: Uppdal, Kongsvall, 4. 7. 1977 (JL); 1 ©, J. A. W. Lucas leg., By: Hoi, Hangastol, 1150 m, 6. 7. 1976 (JL). Irland: 1 ©, M. C. D. Speight leg., N-4060 Ballynafid L. Co. Westmeath, 6. 5. 1985 (CC); 1 9, M. C. D. Speight leg., NR. Slane, Louth, 7. 6. 1971 (@@) Niedenlande: 1.072.900, B. v. Aartsen leg., Limbricht, 23. 4. 1972 (JL); 1 o, Y. $. van der Goot leg., Echt De Dort, 5. 5. 1970 (ZMAN); 1 9, v. Doesburg leg., Baarn Utr., 6. 5. 1947 (ZMAN); 1 9, J. A. W. Lucas leg., Ulvenhout, 12. 5. 1969 (JL); 1 Q, V. S. van der Goot leg., Bunder Bos, 16. 5. 1970 (ZMAN); 1 ©, V. S. van der Goot leg., Rijckholt, Savelsbos, 25. 5. 1970 (ZMAN). Belgien: 1 o Hohes Venn bei Sourbrodt, 25. 4.—3. 5. 1957 (ZFMK); 1 o, Br. Arnoud leg., Rekum, 12. 4. 1950 (ZMAN). Polen: 1 o (Schlesien) ,Wolfels- fall 17/5 26513“ „Original v. Becker“ (ZMHB). Deutschland: 1 © Hint. Sächsische Schweiz, Gr. Winterberg, S-Kuppe, 420—490 m, 5. 5. 1968 (ZMAN); 1 © Bayr. Alpen, Leitzachtal, 14. 4. 1946 (ZSMC); 1 ©, K. Kormann leg., Grötzingen, 21. 4. 1975 (CC); 1 © Straussdorf bei Ebersberg, 26. 4. 1969 (JL); 1 Q, F. Stócklein leg., Starnberger-Seegebiet, Haarkirchen- Mautal, 30. 4. 1941 (ZSMC); 1 ©, Stöck. leg., Starnberg a. See, Würmtal, Petersbrunn, 9. 5. 1975 (JL); 1 Q, Dr. v. Rosen leg., Wolfratshausen, 11. 5. 19. (ZSMC); 1 9, F. Malec leg., 800 m nó Silberborn NC33, 30. 6. 1987 (FM); 1 ©, Weiffenbach leg., Mittelhessen, Lahntal zwischen Gießen und Marburg, 20. 4. 1983 (ZSMC); 1 ©, A. Haas leg., Münster Sentmaring Park, 23. 5. 1953 (ZSMC); 1 9, H. Ulrich leg., Südwestdeutschland, Ottmarsheim, 22. 4. 1957 (ZFMK); 1 9, R. Rombach leg., Eifel, Genfbachtal bei Nettersheim, 18. 6. 1987 (ZFMK). Osterreich: 1 © „Austria“, „Egger 1866“ (NHMW). Schweiz: 1 ©, A. Aptroot leg., Berner Oberland, Grindelwald, 1000 m, 22. 5. 1982 (ZMAN). Italien: 1 9, J-H. Stuke leg., Südtirol, Vinschgau: Planeiltal, 1800—2000 m, 27. 6. 1992 (JS). Ungarn: 1 o, S. Tóth leg., Tihany: Csucshegy, 17. 4. 1983 (CC); 1 ©, S. Toth leg., Tihany: Kiserdó tetó M-cs., 26. 4. 1983 (SUEL); lo 1 9,S. Toth leg., Hárskut: Esztergáli-vólgy, 13. 5. 1983 (SUEL). Jugoslawien/Serbien: 1 ©, A. Vujic leg., Fruska Gora, Vrdnik, 22. 4. 1988 (CC). Spanien: Kantabrisches Gebirge: 400 3 0Q0,V. S. van der Goot & J. A. W. Lucas leg., Oviedo or Leon, Puerto de Pajares, 1350—1700 m, 11.—18. 7. 1972 (JL). Verbreitung: Weite Teile Europas und Sibiriens: Norwegen und Schwedisch Lappland bis Nordspanien, Norditalien (Südtirol), Serbien, Rumänien (Brádescu 1991); von Irland über Mitteleuropa, die Taiga- und Waldgürtel des europäischen Rußland (Stackelberg 1970) und Westsibirien bis zum Baikalsee (Violovitsh 1983). Beschreibung (mitteleuropäische Exemplare) Diagnose: Sehr dicht und auffällig fuchsrot behaart (siehe aber unten: ,,Variabilitat“); 3. Fühlerglied (Abb. 18e, 19e) leuchtend blaßorange. ©: Surstylus stark verlängert und apikal verjüngt (Abb.2, 9). 9: Hinterrand von T5 gerundet (Abb. 34). ©: Kopf (Abb. 35): Stirn lateral grau bestäubt mit deutlicher Mittelfurche, abstehend blaßgelb behaart. Gesicht schwarz, zart grau bestäubt; Mittelhöcker und ein Dreieck über dem seit- lichen Mundrand glänzend. Wangen (Genae) etwas breiter als das 2. Fühlerglied und wie Augen, Scheitel und Occiput fein hell behaart. Fühler (Abb. 18e): Gestalt etwas variabel, oft (aber nicht konstant) mit deutlicher Oberecke, Arista zart, dunkel, fast nackt. Lunula und 2 Grundglieder der Fühler hell. Thorax: Scutum und Scutellum fuchsrot, sehr dicht und mäßig lang abstehend behaart, selten vereinzelte schwarze, borstenförmige Haare neben der Flügelwurzel. Pleuren (Epister- num und Epimeron) zart bestäubt und lang blaßgelb behaart. — Flügel: Adern der basalen Flügelhälfte gelblich, distal dunkler; r-m und m-cu meist etwas stärker gebräunt, gelegentlich 396 ©, Claussen 37 Abb. 32—39: Cheilosia. — 32 C. pictipennis Q , Kopf lateral, 33 C. subpictipennis sp. n. 9, Kopf lateral, 34—35 C. chrysocoma: 34 9, Tergite 4—5 dorsal, 35 or, Kopf lateral (ohne Behaarung und Bestáubung); 36 C. subpictipennis sp. n. Q, Tergite 4—5 dorsal, 37—39 C. kuznetzovae: 37 Q, Stirn dorsal, 38 ©, Kopf lateral (ohne Behaarung), 39 ©, Kopf lateral. Die europáischen Arten der Cheilosia alpina-Gruppe 397 auch die angrenzende Flügelfläche bráunlich. Squamulae weißlich, Saum rötlich; Halteren orange, bisweilen Köpfchen etwas dunkler. — Beine: Femora schwarz mit heller Spitze; Tibien hellorange mit undeutlicher, blaßbrauner Binde, meist überwiegend hell behaart, aber auch mit einzelnen schwarzen Haaren vermischt; Tarsen mit Ausnahme der Endglieder hell, aber pl-Iarsen und p3-Basitarsus dorsal geschwärzt. | Abdomen: T2—4 dicht und lang abstehend fuchsrot behaart, selten Hinterecken von T3 und T4 mit vereinzelten schwarzen Haaren. Sl grau bestäubt; S2—4 glänzend; Behaarung auf S1—2 sowie lateral auf S3—4 abstehend blaßgelb; S3—4 medial mit kurzen, anliegenden Haa- ren, die caudad zunehmend mit schwarzen Härchen vermischt sein können. — Hypopygium (Abb. 2, 9, 16f, 17f): Surstylus sehr stark verlängert und apikal verjüngt. Maße: Körper 9,3—11,6 mm; Flügel 8,3—10,0 mm. Abb. 40—43: Cheilosia, Köpfe. — 40 C. alpina 9 dorsal, 41 C. alpina © lateral, 42 C. mon- tana © lateral (Alpen), 43 C. montana © lateral (Kantabrisches Gebirge, ohne Behaarung und Bestäubung). 398 C. Claussen 9: Dem or sehr ähnlich, aber etwas kürzer und dichter behaart. Kopf: Gesicht etwas weniger bestäubt; Stirn mit schwacher Mittelfurche und den für die Gruppe typischen, sehr breiten Seitenfurchen. Fühler (Abb. 19e) hellorange, das 3. Glied oft mit deutlicher Oberecke, gele- gentlich auch distal gerundet. Thorax: Scutum und Scutellum sehr dicht und gleichmäßig fuchsrot behaart, neben den Flügelwurzeln und gelegentlich auch auf den Postalarcalli einzelne schwarze Borsten. — Flügel: M1 steiler als beim © in R4+5 mündend. — Beine: Tibien heller als beim o, oft ohne dunkle Binden. Abdomen: Kurzoval, der Hinterrand von TS gerundet (Abb. 34); T2—3 lang abstehend fuchsrot behaart; auf T4—S wird die Behaarung caudad allmählich länger und blasser. Maße: Körper 10,2—12,0 mm; Flügel 9,1—10,3 mm. Variabilität: Die Art variiert in Mitteleuropa farblich kaum. In Mittelnorwegen und Mittelschweden treten zunehmend Exemplare auf, die sich durch blassere (graugelbe) und etwas weniger dichte Behaarung, dunkle Tibienbinden und ausgeprägten Flügelfleck von mitteleuropäischen Vertretern unterscheiden; sie entsprechen dem Konzept von phantoma (Zetterstedt, 1838). Die vorliegenden Exemplare aus Irland sind kupferrot behaart, und damit etwas dunkler als kontinentale Vertreter. Bei dem Q zeigt sich außerdem eine scharfe Grenze zwischen der kupferroten, gleichlangen Behaarung auf T2—3 und der blaßgelben, deutlich längeren und struppigeren Behaarung auf T4—5. In diesen Merkmalen scheinen sich nach Verrall (1901: 239) auch die englischen von kontinentalen Exemplaren zu unterscheiden. Aus Nordspanien (Kantabrisches Gebirge) liegt eine Serie vor, die durch ihre blasse und lockere Körperbehaarung den nordskandinavischen Populationen ähnelt. Die spanischen Exemplare unterscheiden sich außerdem durch eine etwas kürzere Körperbehaarung, die Reduktion der schwarzen Borsten an den Seiten des Scutums und durch das etwas schlankere, dorsal geschwärzte und distal gerundete 3. Fühlerglied (Abb. 18€, 196) (siehe auch Marcos-Garcia 1989). Präimaginalstadien: Die Larve von C. chrysocoma soll nach Batra et al. (1981) in den Blattrosetten („crowns“) und Stengeln von Carduus nutans (Asteraceae) minieren. Ei- ablagen konnten in Süddeutschland, Baden-Württemberg, am 11. Mai, an Angelica sylvestris (Apiaceae) beobachtet werden, wobei die Eier einzeln oder in großem Gelege unter eine Blatt- scheide an der Stengelbasis plaziert wurden (Doczkal 1996). Cheilosia kuznetzovae Skufjin, 1977 Cheilosia kuznetzovae Skufjin, 1977; Novye i maloizvestnye vidy nasekomykh Evropeiskoi chasti SSSR, Leningrad: 57—60; loc. typ.: Rußland: Lipetsker Region, Galichya Gora. Typen: Designierte Typen wurden nicht untersucht. Die Daten des vorliegenden Materials sind jedoch mit denen der Typus-Serie identisch. Untersuchtes Material: Rußland: 1 ©, Kuznetzova leg., Galichya Gora, 15. 5. 1974 (AB); 1 9, Kuznetzova leg., Galichya Gora, 6. 5. 1974 (AB); 1 9, Kuznetzova leg., Galichya Gora, 6.—15. 5. 1974 (ZMHB). Verbreitung: Bisher nur vom Typus-Fundort bekannt. Beschreibung Diagnose: Die Q 9 sind durch die kürzere und auf den Abdominaltergiten weniger dichte Körperbehaarung, die © © durch die Form des Gonostylus (Abb. 16d) von den übrigen euro- päischen Arten der alpina-Gruppe zu trennen. ©: Kopf (Abb. 39): Gesicht an Mittelhöcker und Mundrand wenig vorspringend, schwarz glänzend, unterhalb der Fühler und an den Seiten zart gelbgrau bestäubt. Augen lang blaßgelb behaart. Stirn mit deutlicher Mittelfurche, entlang der Augenränder schmal silbergrau bestäubt; Behaarung von Stirn, Scheiteldreieck und Occiput blaßgelb, gelegentlich vereinzelte Die europäischen Arten der Cheilosia alpina-Gruppe 399 dunkle Haare an der Spitze des Scheiteldreiecks. Fühler (Abb. 18c) gelb bis dunkel-orange, 1. Glied dunkel. Genae etwa so breit wie der Durchmesser der tl, glänzend schwarz, blaßgelb behaart. Thorax: Scutum und Scutellum glänzend schwarz, abstehend, mäßig dicht und mäßig lang gelbgrau behaart, lateral und vor dem Scutellum gelegentlich vereinzelte dunkle Haare bei- gemischt. Pleuren zart grau bestäubt mit langen, sehr blassen graugelben Haaren. — Beine: Apices der Femora relativ scharf abgesetzt gelb; Tibien mit undeutlichem dunklen Ring (oft kaum wahrnehmbar); pl-Iarsen dorsal, p3-Basitarsus dorsal und die beiden Endglieder aller Tarsen geschwärzt. — Flügel schwach gelblich mit Andeutung eines bräunlichen Mittelflecks (variabel); Adern der Flügelbasis gelblich. Halteren gelbbraun mit dunklerem Köpfchen. Abdomen: Tergite glänzend schwarz, mäßig lang abstehend behaart (T1—3 rötlichgelb, T4—S heller). S2—5 schwarz glänzend, S1 grau bestäubt; S2 durchgehend und lang abstehend hell behaart. — Hypopygium (Abb. 7, 14, 16d, 17d) ähnlich alpina und montana, aber dorsaler Lobus des Gonostylus länger und kräftiger. Maße: Körper 11,0—12,0 mm; Flügel 8,5 mm. Q: Farblich dem © sehr ähnlich. — Kopf (Abb. 38): Gesicht dichter und ausgedehnter bestäubt. Fühler (Abb. 19c): 3. Glied größer und heller. Stirn (Abb. 37) und Scheitel abstehend blaßgelb behaart. Thorax: Scutum fein punktiert, mit relativ kurzen (kürzer als der Durchmesser von f3), dichtstehenden rötlichen Haaren von annähernd gleicher Länge. Tibien ohne deutliche dunkle Zeichnung, wie auch die Tarsen heller als beim ©. Die von f3 ventral und dorsal abstehenden Haare kürzer als der Durchmesser von f3 (Abb. 31). Abdomen: T1—2 lateral mit längeren rötlichen Haaren (auf T2 mehr als doppelt so lang wie auf dem Scutum); T1—2 medial sowie T3—S sehr locker und kurz abstehend blaßgelb behaart (auf T3 medial kürzer als der Durchmesser des p2-Basitarsus). Sternite wie beim ©, aber kürzer behaart. Maße: Körper 9,5—12 mm; Flügel 8,4 mm. Präimaginalstadien: nicht beschrieben. Cheilosia montana Egger, 1860 Cheilosia montana Egger, 1860; Verh. zool.-bot. Ges. Wien 10: 350; loc. typ.: „Austria Alpes“. Chilosia trisulcata Becker, 1894; N. Acta Acad. Leop. 62(3): 426; loc. typ.: Norditalien, Turin. Chilosia braueri Becker, 1894; N. Acta Acad. Leop. 62(3): 435; loc. typ.: Zentralalpen, Ortler, Stilfser Joch. Syn. n. Chilosia thalhammeri Szilády, 1938; Annls hist.-nat. hung., Pars Zoologica 31: 142; loc. typ.: Ost- und Nordkarpaten: Kórósmezó und Tátraszéplak. Syn. n. Typen: Cheilosia montana Egger: Lectotypus ©, hiermit festgelegt. Etikettierung: , montana det. Egger.“ „montana Egg. det. Becker.“ , ¡montana Egg“, in Coll. Schiner & Egger NHMW. Erhaltungszustand: Gut; das Genital wurde mazeriert und befindet sich in einem Glasróhr- chen an der Nadel des Lectotypus. — Paralectotypus: 1 weiteres © auf derselben Nadel wie der Lectotypus. In der Schiner-Eggerschen Sammlung stecken 3 von Egger determinierte Exemplare: Die beiden oben aufgeführten o © entsprechen der Originalbeschreibung und werden als Syntypen aufgefaßt. Das obere der zusammen genadelten Exemplare ist der Lecto- typus. Das 3. Exemplar, ein Q von montana, trägt die gleichen Etiketten wie der Lectotypus. Es zeigt jedoch auf den Seiten des Scutums eine deutliche Beimischung schwarzer Haare und steht damit im Widerspruch zur Originalbeschreibung. Dieses Exemplar wird deshalb nicht als Syntypus betrachtet. Lectotypus und Paralectotypus wurden als solche etikettiert. Chilosia trisulcata Becker: Holotypus Q. Etikettierung: „frisulcata Beck.“, „234/32 Turin, Italie, coll. Ferrero“, „3863. canicularis“, „234/12“, „710“, in Coll. Museum d’Histoire natu- relle, Genf (vgl. Claußen 1988). 400 C: Claussen Chilosia braueri Becker: Holotypus oO". Etikettierung: ,Brauer-Handl. Stilfserjoch 90“, „Braueri det. Becker“, „Braueri Beck. Typ.“, in Coll. NHMW. Erhaltungszustand: Tarsen 2—5 an p2 links fehlen, sonst vollständig. — Das Exemplar entspricht der Originalbeschreibung. Es ist ein stark abgeflogenes Exemplar von montana mit fast völlig abgeriebener Gesichts- bestäubung. Das Hypopygium wurde mit dem des Lectotypus von montana direkt verglichen: Bei braueri ist die Kammlamelle des Surstylus geringfügig flacher als bei dem Vergleichstier, ansonsten stimmen die Genitalia beider Exemplare völlig überein. C. braueri Becker, 1894 wird daher als Synonym zu C. montana Egger, 1860 gestellt, Syn. n. Chilosia thalhammeri Szilady: Syntypen (mehrere © 0’) zerstört. Nach der Original- beschreibung „mit Ch. montana verwandt, aber durch tiefschwarze Grundfarbe und ganz schwarze, struppige Behaarung verschieden.. .“ (Szilády 1938: 142). Alle in der Original- beschreibung zur Trennung von montana angeführten Merkmale liegen innerhalb der Varia- tionsbreite von montana, so daß C. thalhammeri Szilády, 1938 als Synonym zu C. montana Egger, 1860 gestellt wird, Syn. n. Weiteres Material: Österreich: 1 ©, E. Torp leg., Groß-Glockner, Edelweißspitze, 2570 m, 27. 6. 1973 (ET); 1 o 1 9, CC leg., West-Tirol, Paznauntal N Ischgl, Madleinsee, 2400 m, 13. 7. 1986 (CC); 1 o 2 9 2, CC leg., West-Tirol, Bezirk Paznaun, S Galtür, oberes Jamtal, 1800 — 2200 m, 15. 7. 1986 (CC); 5 © 1 Q 9, CC leg., West-Tirol, Bezirk Paznaun, S Mathon, oberes Laraintal, 2200—2400 m, 18.—25. 7. 1986 (CC); 1 Q, CC leg., West-Tirol, Bezirk Paznaun, SE Kappl, oberes Grübletal, 2000 m, 21. 7. 1986 (CC); 1 Q, CC leg., Montafon, oberes Kromertal, Saarbrücker Hütte, 2500 m, 22. 7. 1986 (CC); 1 9, Mhs leg., Tirol, Naitener Lizum u. Mölztal, 2000 m, 24. 7. 1953 (ZFMK). Deutschland: 1 Q, Doczkal leg., D Alpen Oberstdorf, Nebelhorn, Koblat, 1920— 2220 m, 4. 7. 1994, 8528SW (DD); 1 ©, Doczkal leg., D Alpen Oberstdorf, Fellhorn Umg. Schlappoltsee, 1719—2038 m, 5.7.1994 (DD). Schweiz: 1 ©, M. Daccordi leg., Valais, dint. Col Bretolet, 1800—2000 m, 17. 7. 1989 (MD); 1 ©, J. A. W. Lucas leg., Graubünden, Davos, Ischalp, 1931—2074 m, 23. 7. 1979 (CC); 1 ©, W. Barkemeyer leg., Davos, Flüelapaß, 2380 m, 7. 8. 1979 (CC). Italien: 2 vo 2 Q 9, CC leg., Südtirol, Vinschgau, Planeiltal, 2000—2400 m, 7. 7. 1988 (CC); 1 © 1 9, D. Doczkal leg., gleicher Fundort, 2150 —2400 m, 28. 6. 1992 (DD); 1 o 1 2, J.-H. Stuke leg., Südtirol, Vinschgau, Langtaufertal, 1900—2100 m, 30. 6. 1992 (JS); 4 vo 6 Q 9, CC leg., Südtirol, Vinschgau, Val di Mázia, 2100 m, 9. 7. 1988 (CC); 2 99, M. Daccordi leg., Gran San Bernardo AO, 2400 m, 18. 7. 1990 (MD/CC); 1 Q, W. Barkemeyer leg., Aosta, Gr. S. Bernardo, 2300 m, 29. 7. 1979 (CC); 5 oro 5 2 9, M. Daccordi leg., Passo Stelvio SO, 1600—2000 m, 22. 7. 1988 (MD/CC); 6 ©o 3 9 9, M. Daccordi leg., gleicher Fundort, 2600 m, 30. 7. 1991 (MD/CC); 2 9 9, M. Daccordi leg., Trentino, Rabbi Mge. Stablat, 1600 m, 27. 7. 1986 (MD/CC); 4 © ©, M. Daccordi leg., Trentino, Passo Gavia, 27. 7. 1990 (MD). Rumänien: 2 © ©, V. Brädescu leg., Südkarpaten, Mt. Fágáras, V. Simbetei, 1400 m, 8.—9. 7. 1988 (CC); 1 o 1 9, V. Brádescu leg., Karpaten, Mt. Retezat, Taul Stirbu, 21. 7. 1982 (CC); 1 ©, V. Brädescu leg., Karpaten, Mt. Retezat, Taul Negru, 31. 7. 1976 (VB); 1 o, V. Brádescu leg., Karpaten, Mt. Retezat, Gemenele, 4. 8. 1976 (VB). Spanien: 1 o, M. A. Mar- cos-Garcia leg., Kantabrisches Gebirge, Leon: Leitariegos, Laguna de Arbás, 1700 m, 13. 6. 1986 (AM) (vgl. Marcos-Garcia 1989, sub nom. Cheilosia alpina). Verbreitung: In den Hochgebirgen Mitteleuropas; oberhalb der Waldgrenze: Kantabrisches Gebirge (Nordspanien); Alpen (Schweiz, Deutschland, Osterreich, Italien); Tatra-Gebirge (Kempny 1956: Slovakei; Malski 1959: Polen); Karpaten (Rumänien). [Nachweise aus Bulgarien (Bankowska 1967) erscheinen fraglich: 1 9 „Bulgaria Sofija Vitosa 21. 5. 1959 leg. R. Bankowska — Inst. Zool. PAN Warszawa 59/59 — Cheilosia montana Egg. det. R. Bankowska“ und 1 © „Bulgaria Pirin 1500 m npm 1. 6. 1964 leg. R. Bankowska — Inst. Zool. PAN Warszawa 97/64 — Cheilosia montana Egg. det. R. Bankowska“ in Coll. E. Torp, Jelling, Dänemark, wurden überprüft. Beide Exemplare gehören zu Cheilosia melanura Becker, 1894. — Nach Vuji¢ (1996) beruhen auch die Angaben für die Vojvodina (Serbien) (Glumac 1972: 53) und für Macedonia (Glumac 1968: 865, sub nom. trisulcata) auf Fehlbestimmungen. — Barkalov (briefl. 1991) kennt keine Exemplare aus Sibirien und der Mongolei, so daß sich vermutlich auch die Angabe für Turkestan (Szilädy 1938: 142) nicht auf montana, sondern auf eine der asiatischen Arten der alpina-Gruppe bezieht]. Die europäischen Arten der Cheilosia alpina-Gruppe 401 Beschreibung Diagnose: Sehr áhnlich alpina; Kórper tiefschwarz, glánzend und fein punktiert; meist durch das etwas stärker herabgezogene und im Verhältnis zur Kopfhöhe weniger weit vorsprin- gende Gesicht (Abb. 42), die stärkere Ausdehnung der schwarzen Behaarung sowie die im Ver- hältnis zu dm schmalere cual (Abb. 20—21) zu unterscheiden (siehe aber unter „Variabilität“!). or: — mit den folgenden Unterschieden zu alpina: Kopf (Abb. 42, 43): Genae ausgedehnter, gelegentlich ganz schwarz behaart. Fühler (Abb. 18b) schwarz, 3. Glied bei frischen Tieren bräunlich aufgehellt. Lunula dunkelbraun bis schwarz, nur wenig mit der Stirn kontrastierend; Oberrand der Lunula in der Mitte meist deutlich V-förmig eingekerbt (meist durchlaufend bei alpina). Thorax: Scutum lateral und vor dem Scutellum ausgedehnter schwarz behaart (oft über- wiegend schwarz), auf der Vorderhälfte medial helle Haare beigemischt (variabel). Scutellum überwiegend schwarzhaarig. Behaarung der Pleuren oft überwiegend schwarz, insbesondere Katepisternum ventral fast stets mit kräftigen schwarzen Haaren. — Flügel (Abb. 20): meist ohne Mittelfleck, aber r-m und m-cu gebräunt; cual schmal, nicht breiter als basale Hälfte von dm. Flügeladern auch in der basalen Hälfte dunkel. — Beine schwarz, äußerste Spitzen der Femora sowie Basis (1/3) und Apices der Tibien gelblich bis düster-orange, oft tl und t3 auch distal dunkel. Tarsen bis auf die beiden basalen Glieder von p2 zumindest dorsal geschwärzt. Behaarung der Beine schwarz und gelblich (sehr variabel): Die langen Haare auf der Hinterseite von fl und f2 ganz schwarz oder von der Basis der Femora her mit hellen Haaren vermischt. Distale Hälfte von f3 anterodorsal meist mit lang abstehenden schwarzen Haaren (helle Haare bei einzelnen Stücken aus den Karpaten). Abdomen: schwarz glänzend; Tergite lang abstehend blaß-orange behaart oder in unter- schiedlicher Ausdehnung schwarzhaarig: oft Vorder- und Hinterecken von T2—4 mit schwar- zen Haaren, seltener T4 auch am Hinterrand schwarzhaarig oder T2—4 fast vollständig schwarz behaart (1 ©, Osterreich, Groß-Glockner). Sternite ähnlich behaart wie bei alpina, aber die lang abstehenden Haare auf S2—4 oft schwarz. S8 (Prägenitalsegment) ganz oder überwiegend schwarz behaart. — Hypopygium (Abb. 5, 6, 12, 13, 16c, 17c) ohne konstante Unterschiede zu alpina; meist Kammlamelle des Surstylus geringfügig länger und flacher. Maße: Körper 9,5—12,2 mm; Flügel 7,8—10,5 mm. 9: Dem © sehr ähnlich, aber insgesamt etwas kürzer und heller behaart. — Kopf: Seiten- furchen der Stirn ähnlich breit wie bei alpina. Stirn, Scheitel und Occiput meist teilweise- schwarzhaarig. Fühler (Abb. 19b) schwarz, 3. Glied gelegentlich bräunlich, nicht mehr als dop- pelt so hoch wie das 2. Fühlerglied. Dorsaler Rand der Lunula meist V-förmig eingeschnitten. Thorax: Scutum und Scutellum lang abstehend rötlich-gelb behaart, lateral sowie vor und auf dem Scutellum meist mit eingesprengten schwarzen Haaren. Pleuren lang abstehend gelb- lich und schwarz behaart, zumindest dorsale Hälfte des Katepisternums immer auch mit schwarzen Haaren. — Flügel ohne deutlichen Mittelfleck, aber r-m und m-cu gebräunt; MI spitzwinklig in R4+5 mündend (Abb. 24); cual schmaler oder gleich breit wie basale Hälfte der dm (Abb. 21). — Beine: Tibien und Tarsen oft etwas ausgedehnter hell. Distales 1/3 von f3 (Abb. 27) anterodorsal fast stets mit lang abstehenden schwarzen (sehr selten ganz hellen) Haaren, diese stärker nach außen geneigt als bei alpina (Abb. 28), in Seitenansicht daher weni- ger weit aufgerichtet. Hintercoxen anterolateral mit schwarzer, selten ganz heller Behaarung. Abdomen: Breitoval, schwarz glänzend, fein punktiert. Tergite abstehend fahlgelb bis röt- lich-gelb behaart. Behaarung der Sternite hell. Maße: Körper 9,0—12,5 mm; Flügel 8,2—10,5 mm. Variabilität: Gesichtsprofil in beiden Geschlechtern variabel, so daß dieses Merkmal nicht immer zur Abgrenzung von alpina herangezogen werden kann. Bei den © © können die Abdominaltergite 2—4 ganz hell bis fast ganz schwarz behaart sein. 1 © (Italien: Südtirol, Vinschgau, Planeiltal, 2150—2400 m, 28. 6. 1992) trägt auf beiden Gesichtsseiten 4—5 lange, schwarze Borstenhaare. Die vorliegenden © © aus den Karpaten sind auf T2—4 intensiver rötlich behaart als Tiere aus den Alpen; schwarze Haare fehlen auf T2—4. Das 3. Fühlerglied 402 C. Claussen ist bráunlich aufgehellt, obwohl die Stúcke stark abgeflogen und folglich ausgefárbt sind. — Das einzige bisher aus Spanien bekannte Exemplar (O) trägt auf T2—4, den Pleuren, den Coxen der p3 und dorsal auf f3 keine schwarzen Haare. Es wurde von mir früher zu alpina gestellt (vgl. Marcos-Garcia 1989). Aufgrund des Fligelgeáders (cual schmal) wird diese Bewertung jetzt revidiert. Der Status der nordspanischen Population muß jedoch auf einer breiteren Materialgrundlage geklärt werden. Präimaginalstadien: nicht beschrieben. Cheilosia pictipennis Egger, 1860 Cheilosia pictipennis Egger, 1860; Verh. zool.-bot. Ges. Wien 10: 352; loc. typ.: Österreich, Alpen: „Schneeberg“ (bei Wien). Chilosia „innominata“ (Becker in Bezzi & Stein, 1907); Kat. pal. Dipt., 3: 36 (als Varietát von pictipennis); loc. typ.: , Silesia”. Chilosia bureschi Delkeskamp, 1942; Mitt. dt. ent. Ges. 11: 11; loc. typ.: Bulgarien, Pirin-Gebirge, Potschiren-Dom bei Bansko. Syn. n. Cheilosia zmilampis Violovitsh, 1975; Taksonomiya i ekologiya zhivotnykh Sibiri (Novye i maloizvestnye vidy fauny Sibiri, 9): 82 (Cheilosia); loc. typ.: Altai, Karasuk. Syn. n. Typen: Cheilosia pictipennis Egger: Lectotypus ©, hiermit festgelegt. Etikettierung: „Schneebg. [unleserlich: ,Brauer’ ?]“, „signatipenis“, „pictipennis Typ. Egger“, „pictipenis det. Egger“; in Coll. Schiner & Egger NHMW. Erhaltungszustand: gut. — Der Lectotypus wurde als solcher etikettiert, er entspricht der Originalbeschreibung. Aus dem Naturhistorischen Museum Wien wurden 5 weitere, als fragliche Syntypen von pictipennis ausgewiesene Exemplare untersucht. Aufgrund der Daten wird hier keines dieser Exemplare als Syntypus betrachtet: 1 9 „Schiner 1869“, „chrysocoma Egger“, „chrysocoma Egg. det. Beck.“ ist pictipennis; 1 © „Schiner 1869“, „pictipenis det. Schiner“ ist subpictipennis; 1 9 „Gumpoldskrn., N.O. 28. 4. 12, Czerny“ ist subpictipennis; 1 Q „Austria 5.18.56“, „pictipennis Egg. det. Michl“ ist subpictipennis; 1 Q „Mödling“, „asilica Fall. Alte Sammlung“ ist subpictipennis). Chilosia „innominata“ (Becker in Bezzi & Stein): In Coll. ZMHB wurde nur ein Exemplar gefunden, das als Varietät von pictipennis gekennzeichnet ist: „Chilosia pictipennis var m 2885“, auf der Rückseite des Etikettes: „Landov 8/7 84“. Da die Tergite 1—4 bei diesem Tier hell behaart sind, kann es sich nicht um die von Becker (1894: 409) beschriebene Varietät handeln. — Die Synonymie von „innominata“ erscheint nach der Beschreibung jedoch nicht zweifelhaft. Chilosia bureschi Delkeskamp: Holotypus 9. Etikettiert: „Iypus“, „Bulgarien, Pirin- Gebirge 1200 m, Potschiven-Dom bei Bansko, leg. 1—10. 7. 38 Ziegenhagen“, „Chilosia bureschi Delk. det. Delkeskamp 1941“, „F 163“, „Cheilosia pictipennis Egger det. Claußen 1987“, in Coll. ZMHB. Erhaltungszustand: gut. — Ein Exemplar mit schwach ausgeprägten Flügelflecken; in allen weiteren Merkmalen mit der schwarzhaarigen Form von pictipennis identisch. C. bureschi Delkeskamp, 1942 wird daher als Synonym zu C. pictipennis Egger, 1860 gestellt. Syn. n. Cheilosia zmilampis Violovitsh: Typen nicht untersucht. Das vorliegende Material aus Innerasien (Altai, Sajan) sowie die Abbildungen des Genitale in Violovitsh (1983: 192) lassen jedoch keinen Zweifel an der Synonymie zu. C. zmilampis Violovitsh, 1975 wird daher als Synonym zu C. pictipennis Egger, 1860 gestellt. Syn. n. Weiteres Material: Osterreich: 1 Q, G. Strobl leg., bei Admont a. d. Ems, Ende April (Paratypus von chrysocomoides) (NMBA); 1 Q, CC leg., Hohe Tauern, Virgental: Maurertal, 1600 m, 11. 7. 1991 (CC); 1 ©, CC leg., West-Tirol, Paznauntal, ca. 2 km NE Mathon, 1400—1500 m, 14. 7. 1986 (CC). Deutschland: 1 o 3 Q 9, D. Doczkal leg., D Alpen, Oberst- dorf Mádelegabel Vorderer Bockkar, 1880—2084 m, 3. 7. 1994 (DD); 1 ©, D. Doczkal leg., D Alpen, Oberstdorf Nebelhorn Koblat, 1920—2220 m, 8528SW, 4. 7. 1994 (DD); 2 Q 9, D. Doczkal leg., D Alpen, Oberstdorf Fellhorn Umg. Schlappoltsee, 1719—2038 m, 5. 7. 1994 (DD). Schweiz: 20 © 1 9,J. A. W. Lucas leg., Uri, Andermatt, Gürschenalp, 2000—2200 m, 9. 7. 1985 (JL); 12 Q 9, J. A. W. Lucas leg., Graubünden, Bergún, Latsch, 1588 m, 6. 7. 1979 (JL); 1 9, J. A. W. Lucas leg., Graubünden, Bergün, Val de Taors, 1400—1950 m, 7. 7. 1979 Die europáischen Arten der Cheilosia alpina-Gruppe 403 (JE); 1 O, J. A. W. Lucas leg., Graubünden, Davos, Sertigtal: Witi, 1600—1700 m, 16. 7. 1979 (JL); 1 O, J. A. W. Lucas leg., Graubünden, Davos, „ZO zijde Davoser See“, 1500—1600 m, DATO TO (IE); 2 oo, J. A, W: Lucas leg., gleicher Fundort, 24. 7. 1979 (JL); 2 9 9, V. S. van der Goot leg. Graubünden, Clavadel, 1600—1700 m, 24. 7. 1974 (ZMAN); 2 oc, P. Goeldlin leg., Gr. Val Chamuera, 25. 7. 1980 (MZLS); 1 ©, Steck, Wallis, Useigne, 14. 5. 1934 (MZLU); 1 ©, A. Maibach leg., Valais Lac Tanay, 16. 6. 1986 (MZLS); 1 ©, A. Maibach leg., Valais Moosalp, Chalte Brunne, Moos 2042 m, 26. 6. 1986 (MZLS); 1 ©, Valais, Cols de Cou et de Bretolet, 1900 m, 7. 7. 1968 (MZLS); 3 © Q, J. Aubert leg., Valais, Col de Bretolet, 7.—13. 6. 1976 (MZLS); 1 Q, gleicher Sammler u. Fundort, 9. 7. 1972 (MZLS): 1 Q, Bretolet, 14. 7. 1965 (MZLS); 1 ©, Bretolet, 15. 7. 1964 (MZLS); 1 Q, L. Rezbanyai leg., Rigi-Kulm SZ, 1600—1700 m, 14. 7. 1978, Tf (DD); 1 2, P. Goeldlin leg., Vaud Les Dévens, Bex, 18. 4. 1967 (MZLS); 1 Q, P. Goeldlin leg., Vaud Les Pléiades, 14. 6. 1968 (MZLS). Italien: 1 o 1 9, D. Doczkal leg., Südtirol, Trafoi Tartscher Wald, 1500—2100 m, 31. 5. 1993 (DD); 2 © ©, J.-H. Stuke leg., Vinschgau, Tartscher Wald, 31. 5. 1993 (JS); 4 9 Q, D. Doczkal leg., Südtirol Schlinigerbach, 1750—1850 m, 1. 6. 1993 (DD/CC); 1 o 1 Q, D. Doczkal leg., Südtirol Schlinig, Niedermoor zw. St. Sebastian u. äuß. Alm, ca. 1850 m, 1. 6. 1993 (DD); 2 oo 3 Q 9, J.-H. Stuke leg., Vinschgau, Schliniger Tal, 1. 6. 1993 (JS); 1 ©, Südtirol, Planailtal, 2. 6. 1982 (JL); 1 o, E. Brockmann leg., NW Alpes, Valle de Aosta nórdl. Nus (ONO Aosta) Valle di S Barthelemy bei Arlod, oberhalb Abfahrt Trois Villes, 1500 m, Fels-Matten, 1. 6. 1992 (PL); 1 Q, Borsato-Brutti leg. Verona, Boscochiesanuova Rifugio S. Giorgio, 1800 m, 14. 6. 1987 (MD). Jugoslawien/Serbien: 1 ©, A. Vujic leg., Kopaonik, 24. 5. 1987 (AV); 1 ©, A. Vujié leg., 199 F. Durmitor Skrcko Zdrijelo, 8. 7. 1991 (CC). Bulgarien: 3 Q Q, S. Tóth leg., Rodopehegység Batak, 24. 5. 1982 (ST). Rumänien: 1 o 1 9, V. Brádescu leg., Mt. Retezat Toul Negru, 31. 7. 1976 (VB). Russische Föderation: 1 ©, A. Barkalov leg., Westlicher Sajan, Umgebung von Abasa, 25. 5. 1981 (CC); 1 Q, A. Barkalov leg., Westlicher Sajan, 17. 6. 1981 (CC); 2 © ©, Exp. Mikkola, Hippa & Jalava leg., USSR, SW Altai 15 km S Katanda Bert-Kum, 2000—2500 m, 10.—14. 7. 1983 (MZHP). Verbreitung: Europa, in der Nadelwaldzone der Hochgebirge, oft im Bereich alter Fichten- oder Lárchen-Fichtenwálder (Piceetum, Larici-Piceetum). — Alpen (Schweiz, Italien, Deutschland, Osterreich); Balkangebirge: Kopaonik (Serbien), Berg Durmitor (Monte Negro); Pirin-Gebirge, Rodopen (Bulgarien); Südkarpaten (Rumänien); Kaukasus (Stackelberg & Richter 1968: 247; Barkalov briefl.). — Innerasien: Altai, Westlicher Sajan. — Weitere Litera- turangaben sind wegen der bisherigen Verwechslung mit subpictipennis, die zusammen mit pictipennis auftreten kann, nicht zu verwenden. Beschreibung Diagnose: Sehr ähnlich subpictipennis und in einzelnen Exemplaren ähnlich chrysocoma. S2 medial in einem schmalen Langsstreifen nackt oder die Behaarung hier kürzer als lateral und niedergedrückt. or: Surstylus sehr lang und distal breit gerundet (Abb. 3). @: die kurze Behaarung auf der Ventralseite von f3 sehr gleichmäßig (Abb. 30); Hinterrand von TS gerade abgestutzt, Gesichtsprofil (Abb. 32) zwischen Fühlerwurzel und Mittelhöcker etwas weniger stark nach unten verlaufend als bei subpictipennis (Abb. 33). o: Kopf: Gesicht schwarz glänzend, unter den Fühlern dicht gelblich-weiß bestáubt, Bestáu- bung lateral oft bis zum vorderen Mundrand herabreichend, so daß bei frischen Stücken nur der Mittelhöcker sowie ein Dreieck über dem seitlichen Mundrand unbestäubt bleiben. Genae etwa so breit wie der distale Abschnitt der tl, mäßig glänzend mit längeren gelblichen Haaren. Stirn schwarz, glänzend, zart grau bestäubt, an den Augenrändern mit etwas dichterer Bestäu- bung, Mittelfurche deutlich; Lunula gelblich-braun; Behaarung lang abstehend, gelblichweiß, seitlich der Fühlerwurzeln meist vereinzelte dunkle Haare. Scheitel lang abstehend hell behaart. Augen dicht und lang hell behaart, Occiput ganz hellhaarig. Fühler (Abb. 18f): 1. Glied meist dunkel, 2. und 3. Glied meist rötlich aufgehellt, 3. Glied etwas variabel: meist etwas länger als hoch, gelegentlich mit Andeutung einer Oberecke, dunkel-orange (selten schwarzbraun) mit fast kahler, schwarzer Arista. 404 CrClaussen Thorax: Scutum glänzend schwarzbraun, mäßig dicht und mäßig fein punktiert, dicht und lang abstehend blaß-orange bis fuchsrot behaart; zwischen den Flügelwurzeln meist eine oft unvollständige Binde aus kürzeren schwarzen Haaren und vor der Quernaht lateral gelegent- lich vereinzelte längere dunkle Haare beigemischt. Behaarung der Pleuren lang, weißlichgelb. — Flügel: Fläche zart bräunlich, Adern im basalen Drittel gelblich, distal dunkler; zwischen Subcostalzelle und m-cu meist ein diffuser bräunlicher Fleck. Halteren gelbbraun mit dunklem Köpfchen. Squamulae blaß-orange, Saum geringfügig dunkler. — Beine: Coxae und Trochan- teren schwarzbraun; alle Femora schwarz, apikal schmal gelb; Tibien dunkel-orange mit undeutlicher brauner Binde etwas unterhalb der Mitte (Breite der Binden auf tl und t2 etwa 1/3, auf t3 !/2 der Tibienlánge); Tarsen dorsal geschwárzt, aber die beiden basalen Glieder der p2 sowie Basitarsen der pl und p3 distal meist gelb, oft auch weitere basale Tarsenglieder der p3 dorsal heller. Basitarsus der p3 (im Vergleich zu subpictipennis) schlank, etwa 3,5mal so lang wie an der Basis breit. Behaarung der Beine etwas variabel: an fl vorn kurz, hell; hinten lang, an der Basis überwiegend hell, zur Spitze hin überwiegend schwarz; an f2 im distalen 1/3 auch vorn kurz und schwarz; an f3 vorn lang hell und kurz schwarz, ventral in oft ganzer Länge mit schräg-abstehenden schwarzen Börstchen und einzelnen langen hellen oder hellen und schwarzen Haaren; Tibien und Tarsen in unterschiedlicher Ausdehnung kurz anliegend hell und schwarz. Abdomen: Behaarung der T1-4 lang abstehend, farblich sehr variabel: TI und der größte Teil von T2 meist blaß-orange (wie das Scutum); T3—4, oft auch der äußerste Hinterrand und die Hinterecken von T2 schwarz; aber auch T1—4 ganz hell oder Übergänge zwischen diesen Extremen. Oberfläche der Tergite auf den schwarz behaarten Abschnitten stark glänzend, tief schwarz, auf den hellen Partien mit bronzenem Schimmer, Segmente medial oft etwas matt. — SI graugelb bestäubt, mäßig lang abstehend behaart; übrige Sternite glänzend schwarz, jedoch Vorder- und Hinterrand von S2 bestäubt; Behaarung von S2—4 an den Seiten lang abstehend hell, S2 medial in einem schmalen Streifen nackt oder kurz anliegend behaart, S3—4 medial kurz anliegend und überwiegend schwarz behaart. — Hypopygium (Abb. 3, 10, 16g, 17g,) durch die langen, distal breit gerundeten Surstyli gekennzeichnet. Maße: Körper 12,0—13,8 mm: Flügel 10,8—12,0 mm. 9: Bis auf die üblichen Geschlechtsunterschiede dem © ähnlich. — Kopf (Abb. 32): Gesicht schwarz glänzend, wie beim © in unterschiedlicher Ausdehnung bestäubt. Stirn mit den für die Gruppe charakteristischen breiten Seitenfurchen, deutlich punktiert und abstehend hell behaart. Fühler (Abb. 19f) einschließlich der beiden Grundglieder orange, mit zunehmendem Alter dunkler werdend, bis fast schwarz, das 3. Glied in der Form etwas variabel. Thorax: Insgesamt etwas kürzer behaart als beim ©; Scutum und Scutellum hell behaart, höchstens mit sehr vereinzelten schwarzen Haaren. — Flügel: MI relativ steil in R4+5 mün- dend, Flügelfleck meist deutlich. Halteren gelblich. — Beine: meist etwas heller als beim ©; die kurze Behaarung auf der Ventralseite von f3 sehr gleichmäßig und im Unterschied zu sub- pictipennis durchschnittlich kürzer als !/3 des Durchmessers von f3 (Abb. 30). Abdomen: T1—5 mit Ausnahme der Seiten von T2 kürzer behaart als das Scutum; Haare abstehend, meist hell, aber Hinterränder von T2—5 gelegentlich auch schwarzhaarig. — Sternite etwas kürzer behaart, sonst wie beim ©’; S2 medial in einem schmalen Lángsstreifen nackt oder mit kürzeren niedergedrückten Haaren; auch am Vorderrand von S3 fehlen medial kurze aufgerichtete Haare (relativ konstanter Unterschied zu subpictipennis). Maße: Körper 12,0—14,0 mm; Flügel 10,4—11,8 mm. Variabilität: Die helle Behaarung variiert in beiden Geschlechtern von blaßgelb bis fuchs- rot. Die vorliegenden © © aus Jugoslawien, Rumänien und Bulgarien gehören sämtlich zu der Form mit überwiegend schwarz behaarten Abdominaltergiten. Die untersuchten Exemplare aus Innerasien sind auf TI—4 (bzw. 1—5) gänzlich hell behaart. In den Alpen treten beide Farbvarietäten nebeneinander auf. Präimaginalstadien: nicht beschrieben. Die europáischen Arten der Cheilosia alpina-Gruppe 405 Cheilosia subpictipennis sp. n. Cheilosia pictipennis Barkalov in Violovitsh, 1983, nec Egger, 1860 Holotypus ©. Etikettierung: „Schiner 1869“, „pictipenis det. Schiner“, in Coll. Schiner & Egger (NHMW); loc. typ.: Vermutlich Österreich, Alpen. — Paratypen: Frankreich: 1 ©, C.-F. Kassebeer leg., Pyr. [Pyrénees] Central Porte Puymoreus, 1900 m, 16. 6. 1991 (CK); 1 9, C.-F. Kassebeer leg., Mas. Central Cheires Hautes, 900 m, 2. 6. 1991 (CK); 1 9, C.-F. Kassebeer leg., Haute Provence Montagne de Lure, Jas de Bailles, 1200—1400 m, 22. 5. 1993 (CK); 49 Q,C. J. Zwakhals leg., Savoi, Valmeinier, 1500 m, 13. 7. 1984 (JL); 1 9, M. C. D. Speight leg.; Malvau, c. 1800 ft., Vosges, N. E. France, Lambert grid 531.050, 12. 6. 1979 — “large clearing by stream, edge Picea abies plantation. Flying low among picea stumps” (MS); 1 9, M. C. D. Speight leg., Ballon d’Alsace, Vosges, N. E. France Lambert grid 531.050, ca. 3000 ft., 21. 6. 1979 — “Picea abies woods in glade on Flwrs. Sorbus aucuparia” (MS). Deutschland: 1 9, F. Malec leg., Hessen, Weißenborn, Werra-Meißner-Kreis, Graburg, Schäferburg, 15. 5. 1980 (FM); Baden-Württemberg, Schwarzwald: 1 9, R. Treiber leg., Bad Rippoldsau, 27. 5. 1987 (RT); 1 Q, R. Treiber leg., Nordschwarzwald, Herrenwies, 7. 6. 1988 (RT); 2 9 9, D. Doczkal leg., Nordschwarzwald, Lautenbach-Sohlberg, Sumpfwiesen n Steighof, 640— 670 m, 7414SE, 30. 4. 1994 (DD); 1 9, D. Doczkal leg., Nordschwarzwald, Ottenhöfer-Unterwasser, Kolbenloch, 550-650 m, TK7515SW, 30. 4. 1994 (DD); 1 Q, R. Treiber leg., Freudenstadt, Kniebis, 11. 5. 1990 (RT); 2 o o 2 9 Q, D. Doczkal leg. Nordschwarzwald, Kniebis, Zollstock- hütte, 935 m, 7515NE, 16. 5. 1994 (DD/CC); 1 9, M. Hauser leg., Freudenstadt, Kniebis, 15. 6. 1988 (MH); 1 ©, R. Treiber leg., gleicher Fundort, 30. 6. 1987 (RT); 1 Q, R. Treiber leg., gleicher Fundort, 1000 m, 10. 9. [!] 1986 (RT); 1 o 1 Q, M. Hauser leg., Bernau-Dorf, Scheibenfelsen, 15. 6. 1988 (MH); 1 9, R. Treiber leg., Bernau, 1300 m, 16. 6. 1988 (RT); lo 10, J-H. Stuke leg., Feldberg, Toter Mann, 22. 4. 1991 (JS); 1 o 1 9, J.-H. Stuke leg., gleicher Fundort, 14. 5. 1992 (JS/CC); 1 © 1 9, C.-F. Kassebeer leg., gleicher Fundort, 1000 m, 22. 5. 1991 — an Salix (CK); 1 ©, J-H. Stuke leg., Feldberg 1 SE, Todtnauer-H., 22. 5. 1992 (JS); 1 @, Röseler leg., Feldberg, 30. 5. 1964 (PR); 10 9 Q, U. Buchholz leg., Feld- berg, Zastler Loch, 1280 m, 13.—16. 6. 1988 (UB); 1 Q, J.-H. Stuke leg., Feldberg, Zastler Hütte, 16. 7. 1992 (JS); 1 ©, Röseler leg., Wutach, W.-Grabenmel., 8. 5. 1960 (PR); 1 Q, Röse- ler leg., Wutach, Waldácker, 14. 5. 1960 (PR); 1 ©, D. Doczkal leg., Vogelskopf, Grinden, 1020—1056 m, TK7415NW, 16. 5. 1994 (DD); 1 9, C.-F. Kassebeer leg., Zastler, Rappenfelsen, 22. 5. 1991 (CK); 1 o 1 Q, E. Krüger leg., Neustadt (Schwarzwald), 18. —25. 5. 1944 (CC); 1 co 1 2, J.-H. Stuke leg., Gr. Kappler Tal, 24. 5. 1992 (JS/CC); 1 Q, J-H. Stuke leg., Waldau/St. Márg., 29. 5. 1993 (JS); 2 Q Q, W. Maassen leg., Schluchsee, 3. 6. 1982 (JL); Baden-Württemberg: 1 ©, D. Doczkal leg., Geisingen (Donau), NSG Hórnekopf UTM MUT], 750—780 m, 7. 5. 1990 (DD); 1 Q9, D. Doczkal leg., Mössingen, NSG Filsenberg UTM NU06, 750—805 m, 18. 5. 1990 (DD); 2 Q Q, D. Doczkal leg., Kandel w Kandelhof, 1200—1230 m, 7914NW, 1. 6. 1994 (DD); 2 oc, R. Treiber leg., Schwäbische Alb, Tieringen Lochenstein, 9. 6. 1980 (RT/CC), Thüringen, Thúringerwald: 1 ©, R. Borcherding leg., Kleinschmalkalden, 3. 5. 1994 (RB); 1 9, 6. 5. 1994(RB);3 oo 1 9, 11. 5. 1994 (RB/CC); 1 9, 5. 6. 1993 (RB); 1 9, Hartmann leg., Vessertal, Breitenbach, 14. 5. 1988 (PL); 1 9, Creutzburg leg., Friedrichs- höhe (Gelbschale), 27. 5. 1988 (PL); Bayern: 1 ©, F. Stöcklein leg., Starnberger-Seengebiet, Maisingerschlucht, 22. 5. 43 [1943] (ZSMC); 1 Q, gleicher Sammler und Fundort, 8. 5. 47 [1947] (JL). Schweiz: 1 ©, P. Goeldlin leg., Vaud Agi Hes, 1. 6. 1968 (MZLS); 1 9, F. Geller- Grimm leg., GB Lantsch/Lenz, 900—1200 m, 31. 5. 1991 (FG); 1 9, J. A. W. Lucas leg., Grau- bünden, Bergün, 1367 m, 5. 7. 1979 (JL); 2 oo 1 Q, J. A. W. Lucas leg., Graubünden, Latsch, 1588 m, 6. 7. 1979 (JL): 1 Q, V. S. van der Goot leg., Graubünden, Bergün, 1400—1600 m, 7. 7. 1974 (ZMAN); 1 ©, P. Goeldlin leg., VS Vers chez les Reuses, 19. 5. 1975 (MZLS); 1 9, M. Hauser leg., Wallis, Sion, 20. 5. 1991 (MH); 2 oo 2 Q Q, P. Goeldlin leg., Valais, Lac Champex, 5. 6. 1968 (MZLS/JL); 1 ©, P. Goeldlin leg., Valais, Van den Haut, 22. 6. 1968 (MZLS); 1 ©, J. Aubert leg., Col de Bretolet, 6. 7. 1972 (MZLS); 1 9, M. Hauser leg., Niedergampe, 15. 5. 1991 (MH); 2 oo, M. Hauser leg., Klausen, 3. 6. 1987 (MH). Italien, Alpen: 1 ©, L. Verlinden leg., Val Venosta, Malles, Mazia 1800 m, 28. 5. 1982 (ZMAN); 3 oo 5 Q, D. Doczkal leg., Vinschgau, Planeiltal, 1750—1850 m, 30. 5. 1993 (DD/CC); 13 o o 16 Q 9, J.-H. Stuke leg., Fundort und Datum wie vorher (JS/CC); 2 9 9, 406 C. Claussen D. Doczkal leg., Südtirol, Trafoi, Köpfel, 2200—2400 m, subalp. Weide, 31. 5. 1993 (DD); 1 9, D. Doczkal leg., Südtirol, Trafoi, Tartscher Wald, 1500— 2100 m, 31. 5. 1993 (DD); 1 ©, J.-H. Stuke leg., Fundort und Datum wie vorher (JS); 1 9, D. Doczkal leg., Südtirol, Schlinig, Niedermoor zw. St. Sebastian u. äuß. Alm, ca. 1850 m, 1. 6. 1993 (DD); 1 9, D. Doczkal leg., Südtirol, Schlinigerbach, 1750—1850 m, 1. 6. 1993 (DD); 1 9, J.-H. Stuke leg., Vinschgau, Schliniger Tal, 1. 6. 1993 (JS); 1 9, L. Verlinden leg., Val Venosta, Malles, Planol, 1700 m, 2. 6. 1982 (ZMAN); 1 9, J.-H. Stuke leg., Vinschgäu, Planeiltal, 1800—2000 m, 27. 6. 1992 (JS); 1 9, Miksch leg., Gardasee, Molina di Ledro, Val di Pur, 600—1200 m, 28. 5. 1992 (US); 1 ©, Borsato-Brutti leg., Verona, Boscochiesanuova, Rifugio S. Giorgio, 1800 m, 14. 6. 1987 (MD); 1 ©, Borsato-Brutti leg., gleicher Fundort, 1500 m, 20. 6. 1987 (MD). Österreich: 1 9, „Austria 18. 5. 56“, „pictipennis Egg. det. Mich!“ (NHMW); 1 9, Czerny, Gumpoldskrn, N. O., 28. 4. 12 [vermutl. Gumpoldskirchen, Nieder-Österreich, 1912] (NHMW); 1 Q, „Mödling“, „asilica Fall. Alte Sammlung“ (NHMW); 1 9, M. Hauser leg., Neusiedler See, Winden a. S., Bärenhöhle (Leithargebirge), 30. 4. 1993 (MH). Ungarn: 1 ©, S. Töth leg., Plattensee, Bakony- wald, Keszthelyi-hg. Petó-hegy, 22. 4. 1977 (ST). Slowenien: 1 9, A, Vujic leg., Vrsic-Kranj. Gora, 1200 m, 23. 5. 1989 (AV). Jugoslawien/Makedonija: 2 9 9, Konjsko-Kuzuv, 995 m, 13. 6. 1975 (AV); Griechenland: 1 9, F. Borchmann leg., Olymp-Prioni, 1000 m, 3.—13. 6. 1975 (ZFMK). Verbreitung: Nadelwaldzone: Zentraleuropäischer Mittelgebirgsraum (Zentralmassiv, Vogesen, Nordhessische Mittelgebirge, Schwäbische Alb, Thüringerwald), Schwarzwald, Pyrenäen, Alpen und Leithargebirge, Bakonywald, Balkangebirge (Makedonija, Olympos- Gebirge). — Westsibirien (Barkalov in Violovitsh, 1983: 80, sub nom. „pictipennis“). Beschreibung Diagnose: Sehr ähnlich pictipennis, mit folgenden Unterschieden: Gesichtsprofil (Abb. 33) zwischen Fühlerbasis und Mittelhöcker etwas schräger nach unten verlaufend, S2 median lang abstehend behaart. o: Surstylus (Abb. 8) kurz; dorsaler und ventraler Lobus des Gonostylus (Abb. 16e) gleich lang. 9: Die kurze Behaarung auf der Ventralseite von f3 (Abb. 29) neo und ungleichmäßiger als bei pictipennis (Abb. 30). ©: Kopf: Gesicht zart graugelb bestäubt, Mittelhócker und Seiten oberhalb des Mundrandes schwarz glänzend (Bestäubung bei älteren Tieren oft bis auf einen Streifen unterhalb der Fühlerwurzeln abgerieben). Gesichtsprofil zwischen Fühlerbasis und Mittelhöcker etwas schräger nach unten verlaufend als bei pictipennis, der Mittelhöcker dadurch etwas tiefer liegend (vgl. Abb. 32 und 33). Genae etwa so breit wie tl distal, lang hell, unten aber meist auch schwarz behaart. Stirn schwarz, mäßig fein punktiert mit deutlicher Mittelfurche, an den Augenrändern schmal graugelb bestäubt, lang abstehend hell bis überwiegend schwarz behaart. Lunula meist wie die Stirn gefärbt, selten heller. Scheiteldreieck und die seitlich anschließenden Orbiten lang abstehend hell oder hell und schwarz behaart. Augen vollständig und lang behaart, Augenhaare farblich sehr variabel, von weißlich bis schwarzbraun. Fühler (Abb. 18d): Grundglieder meist geschwärzt, 3. Glied in der Form etwas variabel, dunkel- orange (selten heller), Arista gleichmäßig verjüngt, schwarz. Thorax: Scutum glänzend schwarzbraun, dicht und mäßig fein punktiert, vor dem Scutel- lum kaum merklich grau bestäubt; Behaarung lang abstehend, überwiegend blaßgelb bis röt- lichgelb, zwischen den Flügelwurzeln oft eine Binde aus kürzeren und längeren schwarzen Haaren, die an den Seiten craniad in die helle Behaarung hineinziehen können; gelegentlich auch das Scutellum medial mit kürzeren schwarzen Haaren. Postalarcalli selten mit vereinzel- ten längeren schwarzen Haaren. Pleuren schwach grau bestäubt, lang abstehend, überwiegend hell behaart, Anepisternum und Anepimeron selten mit vereinzelten schwarzen Haaren. Flügel bei ausgefärbten Tieren schwach bräunlich-grau getönt, zwischen Subcostalzelle und m-cu ein diffuser bräunlicher Fleck (bei frischen Tieren undeutlich); Adern der basalen Flügelhälfte gelblich, distal geschwärzt. Halteren blaßbraun, Köpfchen dunkel. Squamulae weißlich mit etwas dunklerem Saum. — Beine: Femora schwarz, apikal schmal gelb; Tibien gelblich, distale 3/5 undeutlich gebräunt; Tarsen überwiegend gelblich, aber die beiden Tarsenendglieder aller Beine oft dunkel sowie der pl- und p3-Basitarsus dorsal geschwärzt Die europáischen Arten der Cheilosia alpina-Gruppe 407 (variabel); Behaarung der Beine sehr variabel: an fl vorn kurz hell, hinten lang hell (basal, in sehr unterschiedlicher Ausdehnung) und schwarz (distal); an f2 ähnlich wie an fl, vorn jedoch etwas länger und in der distalen Hälfte auch schwarz; an f3 vorn in unterschiedlicher Ausdehnung lang hell (basal) und schwarz (distal), ventral meist in ganzer Länge mit schräg abstehenden, kurzen schwarzen Borsten und längeren hellen und schwarzen Haaren; Tibien und Tarsen kurz anliegend hell und schwarz behaart (sehr variabel). Abdomen: Tergite schwarzbraun, teilweise mit bráunlich-violetten Reflexen, T1 und Vorder- ecken von T2 sehr schwach grau bestäubt; T2, gelegentlich auch T3, medial matt, die übrigen Partien der T2—4 einschließlich Hinterrand von T2 stark glänzend. Behaarung von T1—4 lang abstehend, dicht, farblich sehr variabel: hell (weißlich bis orange) und schwarz, die hell- sten Exemplare sind vollständig hell behaart, die dunkelsten bis auf Tl, die Vorderhälfte von T2 und die Vorderecken von T3—4 schwarzhaarig. — Sl und ein schmaler Vorderrand von S2, gelegentlich auch die Hinterränder von S2—4 graugelb bestáubt, die übrigen Partien dieser Sternite schwarz glänzend. Behaarung der Sternite: SI—2 (auch medial) lang abstehend hell, S3—4 seitlich lang abstehend hell oder hell und schwarz, medial überwiegend kurz und anliegend, zumindest teilweise schwarz behaart, oft auch vor dem Hinterrand von S2 medial ein Feld kürzerer schwarzer Haare zwischen den langen hellen. — Hypopygium (Abb. 8, 15, 16e, 17e): Surstylus kurz, dorsaler und ventraler Lobus des Gonostylus gleich lang. Maße: Körper 9,9—13,5 mm; Flügel 9,3—12,0 mm. 9: dem © bis auf die üblichen Geschlechtsunterschiede sehr ähnlich, aber Behaarung etwas kürzer und weniger schwarz, Beine heller. - Kopf (Abb. 33): Genae oft auch im unteren Abschnitt ohne schwarze Haare; Stirn, Scheitel und Orbiten des Hinterkopfes abstehend hell behaart, die dicht und kräftig punktierten Seitenfurchen der Stirn 2/3 der Stirnbreite ein- nehmend; Augen etwas heller und kürzer behaart als beim oc’. Fühler (Abb. 19d): oft auch die beiden Grundglieder rötlich, 3. Glied groß, etwas „fleischig“, in Form, Größe und Farbe variabel (hell- bis dunkel-orange, gelegentlich schwärzlich), Arista zart, schwarz, sehr kurz pubeszent. Thorax: etwas kürzer behaart als beim ©, blaßgelb bis fuchsrot, für gewöhnlich ohne Beimischung schwarzer Haare (nur eines der untersuchten @ Q mit deutlicher schwarzer Binde zwischen den Flügelwurzeln). — Flügel (Abb. 25): der braune Mittelfleck meist deut- licher und in der vorderen Flügelhälfte gelegentlich bis zur Flügelspitze reichend; M1 sehr steil in R4+5 mündend. Köpfchen der Halteren hell. — Beine: Tibien und Tarsen meist etwas heller als beim & und weniger schwarz behaart; die kurze Behaarung auf der Ventralseite von f3 (Abb. 29) etwas länger als 1/3 des Durchmessers von f3 und ungleichmäßiger als bei picti- pennis. Abdomen: oft nur TI mehr oder weniger deutlich bestáubt; Behaarung der T1—5 aufrecht, in der Länge variabel, deutlich kürzer als beim ©, hell (blaßgelb bis hellorange) ohne Bei- mischung schwarzer Haare. — Sternite ähnlich bestäubt und behaart wie beim ©, aber ins- gesamt etwas kürzer und heller, nur selten sind einzelne der kurz anliegenden Härchen medial auf S3—5 schwarz; S2 auch medial mit lang abstehenden Haaren und S3 im basalen !/3 mit einem durchlaufenden Saum aufgerichteter, kurzer Härchen (bei pictipennis zumindest medial niedergedrückt). Maße: Körper 10,5—13,5 mm; Flügel 9,5 —10,3 mm. Variabilität: Nach dem vorliegenden Material variiert die Art geographisch kaum. Präimaginalstadien: D. Doczkal stellte großzügig die folgenden Beobachtungen zur Eiablage von C. subpictipennis im Nord-Schwarzwald (Baden-Württemberg) zur Verfügung: „1) Unterwasser bei Oberkirch, Kolbenloch, 570 m, frische Magerwiese, 30. 4. 1994, 2 9 9 mit Eiablageverhalten an Meum athamanticum Jacq. (Apiaceae). 2) Lautenbach-Sohlberg bei Oberkirch, 650 m, frische Magerwiese, 30. 4. 1994, 4 Q Q mit Eiablageverhalten an Meum athamanticum. 3) Gleicher Fundort, 8. 5. 1995, mehrere Q Q mit Eiablageverhalten, davon 1 Q bei der Ablage von 5 Eiern. Die Eier wurden unter die Blattscheide der Stengelbasis, wenige cm über dem Erdboden abgelegt, dabei waren die Eier nicht an die Wirtspflanze ange- 408 C. Claussen klebt, wie bei anderen Cheilosia-Arten, sondern nur sehr lose angeheftet. Meum war bisher als Wirtspflanze für Cheilosia-Larven nicht bekannt. Da außerhalb des Schwarzwaldes Meum athamanticum nicht an allen Fundorten von C. subpictipennis vorkommt, müssen weitere Pflanzenarten zum Wirtsspektrum von C. subpictipennis gehören. Im Nord-Schwarzwald wurde C. subpictipennis ausnahmslos an den wenigen Standorten von Meum athamanticum festgestellt. Wegen des starken Bestandsrückganges von Meum athamanticum in diesem Gebiet muß C. subpictipennis zumindest hier als stark gefährdete Art betrachtet werden.“ Danksagung Den im Abschnitt „Material und Methode“ aufgeführten Personen und Institutionen sei für den Zugang zu ihren Sammlungen und/oder die Ausleihe von Material herzlich gedankt. Mein besonderer Dank gilt außerdem Herrn Dr. A. V. Barkalov (Novosibirsk) für die Diskussion einzelner Arten, Herrn D. Doczkal (Malsch) für Beobachtungen zum Eiablageverhalten von C. subpictipennis, Frau I. Duty (Rostock) für Übersetzungen aus dem Russischen, Herrn Dr. A. C. Pont (Goring-on-Thames, Reading) für Korrekturen des „Abstract“, Herrn Prof. Dr. K. W. Skufjin (Woronesch) für die Überlassung schwer erreichbarer Literatur, Herrn Dr. Martin C. D. Speight (Dublin) für Informationen über den Typus von C. chrysocoma, sowie Herrn Dr. H. Ulrich (Bonn) für eine eingehende Diskussion der Hypopygium-Strukturen. Schließlich danke ich herzlich meinem Sohn Harm Claußen (Flensburg) für die Übersetzung einiger Texte aus dem Lateinischen und seine geduldige Hilfe bei allen aufgetretenen Schwierigkeiten am Computer sowie Herrn Werner Barkemeyer Gene für die kritische Durchsicht einer früheren Fassung des Manuskriptes. Zusammenfassung Die europäischen Arten der Cheilosia alpina-Gruppe (Barkalov, 1983a) werden revidiert. Die folgenden Arten werden festgestellt und beschrieben: Cheilosia alpina (Zetterstedt, 1838), Cheilosia chrysocoma (Meigen, 1822), Cheilosia kuznetzovae Skufjin, 1977, Cheilosia montana Egger, 1860, Cheilosia pictipennis Egger, 1860 und Cheilosia subpictipennis sp. n. Vier Arten werden in die Synonymie versetzt: Cheilosia montana Egger, 1860 = Cheilosia braueri (Becker, 1894), syn. n. = Cheilosia thalhammeri (Szilády, 1938), syn. n.; Cheilosia pictipennis Egger, 1860 = Cheilosia bureschi (Delkeskamp, 1942), syn. n., = Cheilosia zmilampis Violovitsh, 1975, syn. n. — Lectotypen werden für die folgenden Arten festgelegt: Cheilosia alpina (Zetterstedt, 1838), Cheilosia phantoma (Zetterstedt, 1838), Cheilosia chrysocomoides (Strobl, 1910), Cheilosia montana Egger, 1860 und Cheilosia pictipennis Egger, 1860. Die Arbeit enthält neue Nachweise, Verbreitungsangaben, Angaben zu Präimagi- nalstadien und einen Bestimmungsschlüssel für die behandelten Arten. Literaturverzeichnis Bankowska, R. (1967): Matériaux pour Pétude des Syrphides (Diptera) de Bulgarie. — Fragm. faun. 13: 345—389. Barkalov, A. V. (1981): Taxonomy of species closely allied to Cheilosia illustrata Harris (Diptera, Syrphidae). — Izv. Sib. Otdel. Akad. Nauk SSSR, ser. biol. nauk 1: 112—116. [Russisch]. Barkalov, A. V. (1983a): The role of structure of hypopygium in the systematics of the genus Cheilosia Meigen, 1822 (Diptera, Syrphidae). — Diptera (Insecta), their systematics, geographic distribution and ecology. Leningrad: 3—7. [Russisch]. Barkalow, A. V. (1983b): New data on synonymy and distribution of the hover flies of the genus Cheilosia Mg., 1822 (Diptera, Syrphidae). — Ent. Obozr. 62(3): 633—634. [Rus- sisch]. Batra, S. W. T., J. R. Coulson, P. H. Dunn & P. E. Boldt (1981): Insects and fungi associated with Carduus thistles (Compositae). — Techn. Bull. U.S. Dep. Agric. 1616: I— III + 1—100. Washington. Die europäischen Arten der Cheilosia alpina-Gruppe 409 Becker, T. (1894): Revision der Gattung Chilosia Meigen. — N. Acta Acad. Leop. 62(3): 199 — 524, Tafeln XI— XXIII. Bezzi, M. & P. Stein (1907): Cyclorrapha Aschiza. — In: T. Becker, M. Bezzi, K. Kertész & P. Stein (Hrsg.): Katalog der paläarktischen Dipteren, Bd. 3, S. 1—189. Budapest. Brädescu, V. (1991): Les Syrphides de Roumanie (Diptera, Syrphidae). Clés de détermina- tion et répartition. — Trav. Mus. Hist. nat. „Grigore Antipa“ Vol. 31: 7—83. Claussen, C. (1988): Neue Synonyme in der Gattung Cheilosia (Diptera: Syrphidae). — ENTEZA9S: 205205: Delkeskamp, K. (1942): Eine neue Art der Gattung Chilosia Meigen aus Bulgarien (Dipt. Syrph.). — Mitt. dt. ent. Ges. 11: 11—13. Doczkal, D. (1996): Observations on host plants and behavior of egg-laying femals of Cheilosia Meigen (Diptera, Syrphidae) in Central Europe. — Volucella 2: 77—-85. Stutt- gart. Egger, J. (1860): Dipterologische Beiträge. — Verh. zool.-bot. Ges. Wien 10: 339—358. Frey, R. (1941): Enumeratio insectorum Fenniae. VI. — [i] + 63 S. Helsingfors. Gaunitz, S. (1960): Syrphidenstudien III. (Dipt). — Ent. Tidskr. 81: 35—44. Glumac, S. (1968): Sirfide (Syrphoidea, Diptera) u Makedoniji. — Godisnjak filoz. Fak. Univ. Novi Sad 11: 845—880. Glumac, S. (1972): Catalogus faunae Jugoslaviae Syrphoidea. — Acad. Scient. Art. Slov. 3(6): 1—71. Ljubljana. Kempny, L. (1956): Beitrag zur Kenntnis der Cheilosien (Cheilosia Meigen, 1822, Dipt. Syrphidae) in der CSR. — Pfir. sbor. Ostrav. kraje 17: 542—562. Malski,K. (1959): The Syrphidae of the Polish Tatra Mts. (Diptera). — Acta zool. cracov. 4: 447 —510. Marcos-Garcia, M. A. (1989): Nuevas especies de Cheilosia (Meigen, 1822) para la entomofauna Iberica (Diptera, Syrphidae). — Fragm. entomol. Roma 21: 145—151. | McAlpine, J. E. (1981): Morphology and terminology — adults. S. 9—63. — In: Manual of Nearctic Diptera. Vol. 1. — Agric. Can. Monogr. 27: I—VI, 1—674. Ottawa. Meigen, J. W. (1822): Systematische Beschreibung der bekannten europäischen zweiflügeli- gen Insekten. Bd. 3. — X + 416 S., Taf. 22—32. Schulz-Wundermann, Hamm. Peck, L. V. (1971): A description of new and little-known flies of the genera Chellosia, Eristalis and Eumerus (Diptera, Syrphidae) from Kirghizia. — Ent. Obozr. 50(3): 695—705. [Russisch]. Peck, L. V. (1977): New data on the hover-fly fauna of Mongolia (Diptera, Syrphidae). — Insects of Mongolia 5: 698 — 710. Skufjin, K. V. (1977): New species and subspecies of the genus Cheilosia Mg. (Diptera, Syrphidae) from Galichya gora (Lipetsk Region). — Novye i maloizvestnye vidy nasekomykh Evropeiskoi chasti SSSR, Leningrad: 57—60. [Russisch]. Stackelberg, A. A. (1963): Neue paläarktische Cheilosia-Arten (Diptera: Syrphidae). — Beltre Ent. 13: 513-522. Stackelberg, A. A. (1970): [Fam. Syrphidae — Schwebfliegen]. — In: Bei-Bienko, G.Ya.: [Bestimmungsschliissel für die Insekten des europäischen Teils der UdSSR] 5(2): 11—96. Nauka, Leningrad. [Russisch]. Stackelberg, A. A. &L. V. Peck (1979): Syrphiden von der Mongolei (Diptera, Syrphi- dae). — Folia ent. hung. 32: 129— 147. Stackelberg, A. A. & V. A. Richter (1968): Hover-Flies (Diptera, Syrphidae) of the Caucasus. — Trudy vszesz. ent. Ob. 52: 224—274. [Russisch]. Strobl, G. (1910): Die Dipteren der Steiermark. II. Nachtrag. — Mitt. naturw. Ver. Steier- mark 46(1909): 45 — 293. Szilády, Z. (1938): Über paláarktische Syrphiden. II. — Ann. hist.-nat. Mus. nat. hung. 31: 573143: Verrall, G. H. (1901): British flies, 8. Platypezidae, Pipunculidae, and Syrphidae of Great Britain. — 691 S. London. Violovitsh, N. A. (1973): New Palaearctic species of Syrphidae (Diptera) from the Altai. — Tr. biol. Inst. sib. Otdel. Akad. Nauk SSSR 16: 145—149. [Russisch]. 410 C. Claussen Violovitsh, N. A. (1975): Some new species of hover-flies (Diptera, Syrphidae) from the fauna of the USSR. Taksonomiya i ekologiya zhivotnykh Sibiri. — Novi i maloizvestnye vidy fauny Sibiri, 9. Novosibirsk: 73—89. [Russisch]. Violovitsh, N. A. (1983): Siberian Syrphidae (Diptera). — 242 S. Nauka, Novosibirsk. [Russisch]. Vujic, A. (1996): Genus Cheilosia Meigen and related genera (Diptera: Syrphidae) on the Balkan Peninsula. — 196 S. (+ 4 S. Inhaltsverzeichnis), Matica srpska, Department of natural sciences, Novi Sad. Zetterstedt, J. W. (1838 —1840): Insecta Lapponica. VIS. + 1140 Spalten (2 je Seite). Voss, Lipsiae [= Leipzig]. (Diptera auf Sp. 477—868, erschienen 1838.) Claus Claußen, Twedter Holz 12, D-24944 Flensburg, Deutschland. Bd. 47 S. 411—419 Bonn, September 1998 Tersilochus curvator Horstmann und Tersilochus sp. n. (Ichneumonidae, Tersilochinae), neue Parasitoiden der an Birken minierenden Trugmotten (Lepidoptera, Eriocraniidae) Thorsten Jordan Abstract. Tersilochus curvator Horstmann and Tersilochus sp. n. (Ichneumonidae, Tersilochinae), new parasitoids of birch leaf mining Eriocrania species (Lepidoptera, Eriocraniidae). — Tersilochus species (Ichneumonidae, Tersilochinae) are known to be koinobiont endoparasites of concealed living beetle larvae (Coleoptera, mainly Curculioni- dae and Nitidulidae). A new rearing record from birch leaf mining Eriocraniidae, a primitive group of Lepidoptera, is presented. Tersilochus curvator was reared from Eriocrania cicatricella, and a new species, Tersilochus sp. n., from Eriocrania unimaculella. The parasitoids attack their host larvae in spring and develop inside the host cocoons in the soil. Adults remain in the cocoons from autumn to the following spring. Tersilochus curvator attacks young larvae of E. cicatricella while Tersilochus sp. n. attacks older larvae of E. unimaculella by attaching eggs with an anchor-like mechanism on the inner side of the hosts skin. Attacked larvae were recorded throughout Central Europe, from northern and southern Germany, the Alsace, the Swiss Rhone Valley and from northern Austria. Key words: Tersilochinae, leaf miner parasitoids, Eriocraniidae, Tersilochus-eggs, new parasitoid-host relationship. Einleitung Die meisten Vertreter der artenarmen Unterfamilie der Tersilochinae (Ichneumoni- dae) entwickeln sich als solitáre Endoparasiten vorwiegend an Rüsselkäfern (Col., Curculionidae) und Glanzkäfern (Col., Nitidulidae) (Gauld & Bolton 1988). Von den 33 bekannten Arten aus der Gattung Tersilochus Holmgren, 1858, wurden allerdings erst 7 Arten einem Wirt zugeordnet, von denen sich 5 Arten an Rüsselkäferlarven und jeweils eine an einer Glanzkäfer- und einer Blattkäferart (Col. Chrysomelidae) entwickeln (Horstmann 1971, 1981; Klingenberg & Ulber 1994). Im Rahmen einer Studie über Birkenminierer schlüpften Tersilochus-Imagines aus Kokons von Klein- schmetterlingen der Gattung Zriocrania Zeller (Lepidoptera, Eriocraniidae). Dieser ungewöhnliche Zuchtbefund war Ausgangspunkt einer genaueren Analyse des Para- sitoidenkomplexes der mit Birken assoziierten Trugmotten, die 1987 in Schleswig- Holstein begonnen und von 1990—1993 auf verschiedene Regionen Mitteleuropas ausgeweitet wurde. Im Verlauf der Arbeit wurden Tersilochus curvator Horstmann, 1981 (= T. saltator F., 1781), aus E. cicatricella sowie eine bislang noch unbeschrie- bene Art, Tersilochus sp. n. (Horstmann, pers. Mitt.), aus E. unimaculella gezüchtet. An den anderen gleichzeitig an Birken minierenden Eriocrania-Arten, E. sangii und E. semipurpurella, wurden bisher keine Tersilochus-Eier gefunden, auch schlüpften keine Tersilochus-Imagines aus den entsprechenden Zuchten. Da die Biologie und Wirtsbindung der beiden Tersilochus-Arten bisher unbekannt waren (Horstmann 1981), schien es lohnend, Untersuchungen zur Wirts- und Stadienspezifität der Para- 412 T. Jordan sitoiden sowie zu ihrer Entwicklungsbiologie durchzuführen. Um den Einfluß der Parasitoiden auf die Populationsentwicklung ihrer jeweiligen Wirtsarten abzuschát- zen, wurde der Parasitierungserfolg in verschiedenen Regionen Mitteleuropas erfaßt und für ausgewählte Standorte dargestellt. Material und Methoden Trugmotten (Eriocraniidae) stellen eine der ursprünglichsten Familien der Lepidoptera dar (Common 1975), deren paläarktische Arten alle in der Gattung Eriocrania Zeller, 1851, zusam- mengefaßt werden (Birket-Smith & Kristensen 1974). Von den 9 europäischen Eriocrania- Arten (Burmann 1958; Heath 1976) entwickeln sich 6 Arten im Frühjahr und Sommer an Birken (Betula spp.). Eriocrania cicatricella (Zetterstedt, 1839) (= E. haworthi Bradley, 1966; Karsholt et al. 1994), E. sangii (Wood, 1891), E. semipurpurella (Stephens, 1835), und E. unimaculella (Zetterstedt, 1840) minieren im Frühjahr und sind in Mittel- und Nordeuropa weit verbreitet (Davis 1978; Heath 1976; Koponen 1974). Die Weibchen legen ihre Eier in die aufbrechenden Knospen der Wirtspflanzen. Die Larven (Lı—L.) fressen etwa 4—5 Wochen in den Minen, bevor sie Mitte bis Ende Mai aus den Minen abwandern und in den oberen Bodenschichten einen Kokon spinnen, in welchem sie, zunächst als Larve, ab August im Puppenstadium, bis zum nächsten Frühjahr ruhen (Hering 1957; Davis 1978). Trugmottenarten wurden in verschiedenen Regionen Schleswig-Holsteins (n = 11 Stand- orte), in Hessen (2), Bayern (3), Baden-Württemberg (3), im französischen Elsaß (3), Schweizer Wallis und im Waldviertel (5) in Nordösterreich untersucht. Blattminen mit reifen Larven wurden Ende April bis Mitte Mai im Freiland gesammelt, in höheren Lagen bis Anfang Juni. Für Zuchten wurden abgewanderte Wirtslarven nach Arten getrennt auf sterilisiertes Sand / Blumenerdegemisch gelegt und die Zuchtgefäße (1,3 1-Plastikdosen) frostfrei in einer Erd- kammer überwintert. Parasitierungsanalysen wurden an minierenden Wirten oder an abwan- dernden Altlarven durchgeführt. Zur Untersuchung der Parasitoidenentwicklung wurden reife Larven in feingesiebte Zuchterde überführt und ihre Kokons später isoliert. Die Eriocrania- Larven wurden nach Hering (1957) und die Imagines nach Heath (1976) bestimmt. Die Deter- mination der Tersilochus-Imagines übernahm freundlicherweise K. Horstmann, Würzburg. Ergebnisse Imagines von Tersilochus curvator sind aus Nord- und Mittelschweden, Südfinnland, Südengland und Norddeutschland bekannt (Horstmann 1981). Die während dieser Untersuchung gezüchteten Imagines wurden ausschließlich aus E. cicatricella-Larven erhalten und stammen aus Walliser Larvenproben, dem Elsaß und in wenigen Exem- plaren aus Norddeutschland. In den übrigen Regionen fehlte 7. curvator im Para- sitoidenkomplex von E. cicatricella, wie an Hand von Zuchten und Untersuchungen an minierenden Larven festgestellt wurde. Die Parasitierungsraten von jüngeren und erwachsenen Larvenstadien waren im Wallis ähnlich (Parasitierungsrate L3 = 3,4 %, L4=3,7%, y2=0,13 korrigiert nach Yates: nicht signifikant unterschiedlich, n = 320), wobei diese jeweils nur mit einem Parasitoidenei belegt waren. Lediglich in Junglarven (Li und L2) wurden gelegentlich 2, 3 oder 4 7! curvator-Eier pro Larve vorgefunden (Wallis 1993, in 9 % der Wirtslarven, n = 55). Die Eier (Abb. la) sind gurkenförmig und werden frei im Wirt abgelegt. Die Parasitoiden schlüpfen noch in den minierenden Wirtslarven aus ihren Eiern (etwa Ende April bis Mitte Mai). In den Wirtskokons wurden bis Anfang Juli Para- sitoiden-Junglarven vorgefunden (Li—L2; n = 86 Parasitoide untersucht), die sich auf Grund der charakteristischen Larvengestalt (Abb. 1b) eindeutig 7! curvator zuordnen lassen. Die weitere Entwicklung der Parasitoidenlarven (L3—Ls) erfolgte Neue Parasitoiden minierender Trugmotten 413 bis Anfang August (n = 81). Die Altlarven (Kopfkapsel siehe Abb. 1c) spinnen ab etwa Mitte August einen eigenen Kokon innerhalb des Wirtskokons, in dem sie sich verpuppen. Ab Ende August waren die ersten, ab Mitte September etwa die Hälfte der Imagines (48 %, n = 62) in den Kokons geschlüpft, in denen sie bis zum nächsten Frühjahr verblieben. Imagines aus Walliser Proben verließen die Kokons bei angenä- herten Außentemperaturen (8—15°C) zwischen Mitte und Ende März. Der Schlupf- verlauf war gleichmäßig (50 % Schlupferfolg der Männchen und Weibchen jeweils am 25. 3. 1993), das Geschlechterverhältnis ausgeglichen (205 Männchen und 203 Weibchen geschlüpft). Einige Individuen verblieben längere Zeit in Diapause und schlüpften erst ein Jahr später (n = 3, Schlupf 1994). Abb. 1: Jugendstadien von 7: curvator; Wirt: E. cicatricella. a) Eiform, b) Junglarve (Li), c) Kopfsklerite der Altlarve. Maßstab jeweils 0,1 mm. 414 TE Jordan Die Parasitierungsraten waren im allgemeinen gering, wie im Elsaß (4,4 Yo + 1,9 S.E.) und im Wallis, wo der Parasitierungserfolg von 1990 mit 2,8 % +0,8 S.E. auf 1,2 % +0,5 S.E. abnahm (1991) und dann 1992 auf 3,8% +1,1 S.E. anstieg. 1993 waren im Wallis 9,9 % + 1,1 S.E. der E. cicatricella-Larven durch 7! curvator para- sitiert. Eine bisher unbeschriebene Art, Tersilochus sp. n. (det. Horstmann), wurde aus- schließlich aus E. unimaculella-Larven gezüchtet. Eier dieses Endoparasitoiden wurden in allen 5 untersuchten Regionen, an 15 von 22 Sammelplátzen mit Befall von E. unimaculella, nachgewiesen, wobei die Parasitierungsraten zwischen 0,6 Yo und 35,1 % schwankten. In den 5 größten Wirtspopulationen (Probengröße n = 210—820 E. unimaculella-Larven, gesamt n = 2210) waren im Mittel 18,9 % # 7,8 S.E. der Wirtslarven (La) parasitiert. Befallene Wirtslarven wurden ebenfalls bei niedrigen Dichten (Anzahl gefundener E. unimaculella Minen < 80) nachgewiesen, wobei die Parasitoiden 3 %—35 % der Wirtslarven mit Eiern belegten (Tab. 1). An einem Standort in der Nähe von Kiel (Norddeutschland) war Tersilochus sp. n. ein dominanter Parasitoid von E. unimaculella. Die Parasitierungsraten betrugen 1989 25,5% +5,1 S.E., 1990 24,5 Y +4,3 S.E. und 1991 19,2 % +4,5 S.E. Die Minen- dichten von E. unimaculella sanken in diesem Zeitraum von 7,9 % befallener Birken- blátter (1989, n = 3110 Blätter untersucht) auf 4,4 % (1990; n = 1750), bzw. 3,7 % ab (1991, n = 1440). Tersilochus sp. n.-Weibchen legen ihre Eier wenige Tage vor dem Abwandern der Wirtslarven ab, wie Untersuchungen in Kiel ergaben. So waren am 15. Mai in Blatt- proben noch keine E. unimaculella-Larven parasitiert (n = 107, 60 % Ls), jedoch die 3 Tage später aufgesammelten Wirtslarven zu 37 Yo durch Tersilochus sp. n. befal- len, wobei Dritt- und Viertlarven gleichmäßig belegt wurden (x? = 0,02, korrigiert nach Yates; n = 161). Superparasitierungen mit 2 oder 3 Tersilochus sp. n.-Eiern pro Wirtslarve traten in der Regel erst bei Befallsraten von über 10 % auf (Tab. 1). Das Ausmaß der Mehrfachbelegungen stieg in den untersuchten Wirtspopulationen signifikant mit der Zunahme der Parasitierungsraten an (Spearman Rank Korrela- tionskoeffizient r, = 0,983; Test zweiseitig, a = 0,05, n = 17). Genauere Untersuchungen der Eiablage an E. unimaculella ergaben einen unge- wöhnlichen Befund: Tersilochus sp. n. befestigt die Eier mit einem Ankerapparat in der Wirtslarvenhaut. Die Eier tragen einen zentralen, scheibenförmigen Ankerappa- Tabelle 1: Anzahl Eier pro Wirtslarve und Parasitierungsraten (in %) von Tersilochus sp. n. an E. unimaculella in Norddeutschland (1), in Hessen (2), im Elsaß (3) und im Wallis (4). Anzahl Eier pro Wirtslarve Parasitierungs- Standort 2 3 rate (Y) : Dieburg (2) Kiel (1) Vellescot (3) Sierre (4) Hammoor (1) Bordelum (1) Hammoor (1) Neue Parasitoiden minierender Trugmotten 415 außen Wirtslarvenhaut Abb. 2: Jugendstadien von Tersilochus sp. n.; Wirt: E. unimaculella. a) Eiform, b) Erstlarve in der Eihülle, c) Junglarve (L>), d) Kopfsklerite der Altlarve. Maßstab jeweils 0.1 mm. 416 1) Jordan Tabelle 2: Verteilung abgelegter Tersilochus sp. n.-Eier an E. unimaculella-Larven. Thorax- Abdomensegment Il 7782970212229 Ablageort dorsal lateral ventral Summe rat, mit dem sie innen in der Larvenhaut befestigt werden, wobei der Anker die Ei- ablagestelle verschließt (Abb. 2a). Die hellbraun gefärbten, am cephalen Pol farb- losen Eier waren bevorzugt im Thorakalbereich der Wirtslarven abgelegt (93 % aller Eiablagen, Tab. 2) und durch die Wirtslarvenhaut hindurch sichtbar. Die Parasitoi- denlarven verbleiben einige Zeit in der Eischale (Abb. 2b) und fressen im Haemocoel des Wirtes. Die Entwicklung der Parasitoidenlarven in den Wirten wurde nur an wenigen Individuen verfolgt. Neun Tage nach dem Abwandern der Wirtslarven hatte sich 1 von 3 Tersilochus sp. n.-Erstlarven zur L2 (Abb. 2c) entwickelt. Drei Wochen nach dem Abwandern der Wirte wurden 2 Li und 10 L2 in ihren Wirtslar- ven aufgefunden. Mitte August, etwa 12 Wochen nach dem Abwandern der Wirte, spannen sich die Tersilochus sp. n.-Altlarven (Kopfkapsel siehe Abb. 2d) in ihren Kokon ein. Die Imagines schlüpfen im Herbst und überwintern im Kokon. Im Okto- ber vorzeitig befreite Individuen liefen träge umher, waren aber offensichtlich noch nicht flugfähig. Imagines aus norddeutschen Proben schlüpften bei angenäherten Außentemperaturen zwischen dem 27. März und dem 2. April 1988 (n = 12). Die Lebensspanne umfaßte 16 bis 36 Tage, im Mittel 27 + 4 S.E. Tage (bei natürlichen Lichtverhältnissen und Temperaturen, mit Rosinen und Wasser gefüttert). Im allge- meinen schlüpften die Imagines im darauffolgenden Jahr nach der Probennahme (Zuchten 1988— 1994, n = 70). Lediglich 2 Individuen schlüpften 3 und 4 Jahre nach der Aufsammlung aus den Larvenproben, waren also in eine verlängerte Diapause eingetreten. Diskussion Die beiden untersuchten Tersilochus-Arten lassen sich als strikt monophage Larven- parasiten von E. cicatricella und E. unimaculella charakterisieren. Die vollständige Kongruenz von Wirtswahl und Wirtsbelegung mit den unterschiedlichen Eiformen bestátigt die hohe Spezifitát der Parasitoiden. Sie haben zudem eine unterschiedliche Stadienspezifitát. Tersilochus curvator parasitiert bereits Junglarven (Li und Lz) von E. cicatricella. Daim Wallis keine Zunahme der Parasitierung durch JT. curvator bei álteren Larvenstadien im Vergleich zu júngeren E. cicatricella nachgewiesen wurde, scheint die Eiablage in einem kurzen Zeitraum zu Beginn der Wirtslarvenent- wicklung zu erfolgen. Tersilochus sp. n. hingegen befällt E. unimaculella erst am Ende der Larvenentwicklung, kurz vor dem Abwandern der Wirte aus ihren Minen, wie in Kiel festgestellt wurde. Neben der Wirtswahl, der unterschiedlichen Stadien- spezifität und divergierenden Abundanz der beiden Tersilochus-Arten unterstützen IeTorden 417 Merkmale der Parasitoiden-Jugendstadien, wie unterschiedliche Eiformen und Kopfsklerite der Altlarven, zusätzlich die Zuordnung der gezüchteten Parasitoiden- Imagines zu unterschiedlichen Arten. Die erfolgreiche Parasitierung bei niedrigen Wirtsdichten weist auf ein hohes Wirtsfindevermögen der Tersilochus-Weibchen hin, die außerdem in der Lage sind, zwischen den einzelnen Eriocrania-Wirtsarten zu unterscheiden. Eine Diskrimination bereits parasitierter Wirtslarven zur Verminderung innerartlicher Konkurrenz, wie sie bei vielen echten Schlupfwespen (Ichneumonidae) verbreitet ist, manifestiert sich im allgemeinen durch niedrige Superparasitierungsraten sowie eine geringe Anzahl Eier pro Wirt bei mehrfachen Belegungen (Pschorn-Walcher 1985). Für 7. curvator wurden fast ausschließlich Solitärparasitierungen und für Tersilochus sp. n. gelegent- liche Superparasitierungen mit höchstens drei Eiern pro Wirtslarve festgestellt. Jedoch entsprach die Verteilung der Parasitoideneier in den jeweiligen Wirtslarven- populationen einer Zufallsverteilung (Verteilung nicht signifikant unterschiedlich zu einer Poisson-Verteilung). Somit erscheint für die beiden Tersilochus-Arten ein ent- sprechendes Diskriminationsvermögen fraglich. Die Entwicklungsbiologie von 7. curvator und Tersilochus sp. n. entspricht weit- gehend jener anderer Tersilochus-Arten, die sich an versteckt lebenden Wirten ent- wickeln (Gauld & Bolton 1988). Beide Endoparasiten verzehren den Wirt im Kokon im Boden, spinnen einen eigenen Kokon und verpuppen sich noch im Sommer. Die Imagines schlüpfen vor der Überwinterung, verbleiben aber bis zum Ausschlupf im Wirtskokon. Das Verbreitungsareal der untersuchten Eriocrania-Wirtsarten erstreckt sich über ganz Europa. Davis (1978) gibt das Vorkommen der an Birken minierenden Zriocra- nia-Arten von der Baumgrenze am Polarkreis bis zu den Alpen und bis nach Asien hin an. Imagines von 7. curvator sind im nördlichen Verbreitungsgebiet ihres Wirtes bekannt (Horstmann 1981) und wurden hier zusätzlich für das Elsaß und die Schweiz nachgewiesen. In höheren Lagen des Schwarzwaldes und im Waldviertel war die Wirtsart E. cicatricella nicht oder nur in geringen Dichten vorhanden (Jordan 1992), so daß nicht sicher beurteilt werden kann, ob 7! curvator dort vorkommt. Tersilochus sp. n. war dagegen regelmäßig in allen untersuchten Regionen mit £. unimaculella- Befall vertreten. Die Verankerung der Tersilochus sp. n.-Eier in den Wirtslarven scheint eine Beson- derheit innerhalb der Gattung zu sein. Allerdings sind erst von wenigen Tersilochus- Arten die Wirtsbeziehungen bekannt (Horstmann 1971, 1981), so daß hier keine endgültigen Aussagen gemacht werden können. Einige Arten der nahe verwandten Gattung Diasparsis (Ichneumonidae, Tersilochinae) besitzen knopfförmige Anker oder sehr kleine Anheftestellen an den Eiern und werden ebenfalls auf der Innenseite der Wirtslarvenhaut befestigt (Dysart et al. 1973). Ausschlüpfende Diasparsis-Larven und zum Teil auch ältere Larvenstadien fressen aus der fixierten Eischale heraus im Haemocoel des Wirtes, was von Parker & Berry (1950) als Verteidigung gegen aggres- sive endoparasitische Konkurrenten gedeutet wird. Daneben gibt es Diasparsis-Arten, deren Eier keine Befestigungsstrukturen aufweisen und die frei im Haemocoel der Wirtslarven flottieren (Dysart et al. 1973; Montgomery & DeWitt 1975). Unter den Tersilochus-Arten weisen z. B. die gut untersuchten Arten 7. conotracheli Riley und T. melanogaster Thoms. Eier ohne Befestigungsstrukturen auf (Cushman 1916, Jourdheuil 1960), wie sie auch bei 7? curvator vorgefunden wurden. 418 To jordan Tersilochinae sind im allgemeinen spezialisierte Endoparasiten phytophager Käfer- larven, die sich im Boden verpuppen, oder von phylogenetisch „altertümlichen“ Blattwespen wie den Xyelidae (Gauld & Bolton 1988). Davis (1978) nennt eine (frag- liche) Tersilochus-Art als Parasitoid der nearktischen Trugmotte Dyseriocrania griseocapitella (Wals.) (Eriocraniidae). Inwiefern eine Adaptation an die phylogene- tisch ursprünglichen Zriocrania-Arten erfolgte und die Frage, warum an den gleich- zeitig an Birken minierenden Arten E. sangii und E. semipurpurella keine Tersilochus- Arten leben, muß zunächst offen bleiben. Der bisher spärliche Kenntnisstand über die übrigen europäischen Trugmottenarten sowie deren assozierte Parasitoiden lassen in Zukunft die Entdeckung neuer Parasitoidenarten erhoffen und damit fundiertere Aussagen über das Ausmaß der Wirtsbeziehungen der Tersilochus-Arten zu den phylogenetisch ursprünglichen Zriocrania-Arten erwarten. Danksagung Mein Dank gilt Prof. Dr. H. Pschorn-Walcher, Neulengbach, Österreich, für die Anregung zu dieser Arbeit und die Durchsicht des Manuskriptes. Weiterhin danke ich Prof. Dr. K. Horst- mann, Würzburg, für die Bestimmung der Tersilochus-Arten und Dr. E. Altenhofer, Groß Gerungs, Österreich, für die Unterstützung bei der Probennahme im Waldviertel. Ein Teil dieser Arbeit wurde von der Deutschen Forschungsgemeinschaft finanziert (P/S 7-3/1, 1990). Zusammenfassung Tersilochus-Arten (Ichneumonidae, Tersilochinae) waren bisher als koinobionte Endoparasiten versteckt lebender Käferlarven (Coleoptera, hauptsächlich Curculionidae und Nitidulidae) bekannt. Hier werden erstmals Wirtsbeziehungen zu in Birkenblättern minierenden Trug- motten (Lepidoptera, Eriocraniidae), einer phylogenetisch ursprünglichen Kleinschmetter- lingsgruppe, vorgestellt. Der Parasitoid Tersilochus curvator wurde aus Eriocrania cicatricella und eine noch unbeschriebene Art, Tersilochus sp. n., aus Eriocrania unimaculella gezüchtet. Beide Parasitoiden-Arten befallen die Wirtslarven im Frühjahr in den Blattminen und ent- wickeln sich im Wirtskokon im Boden weiter. Die Imagines verbleiben vom Herbst bis zum Ausschlupf im darauffolgenden Frühjahr im Kokon. Tersilochus curvator parasitiert bereits Junglarven von E. cicatricella, während Tersilochus sp. n. ältere Larvenstadien von E. uni- maculella belegt, wobei die mit einem Ankerapparat versehenen Eier unter der Larvenhaut befestigt werden. Befallene Wirtslarven wurden in Nord- und Süddeutschland, Österreich, dem Elsaß und im Wallis nachgewiesen. Literatur Birket-Smith, J. & N. P. Kristensen (1974): The skeleto-muscular anatomy of the genital segments of male Eriocrania (Insecta, Lepidoptera). — Z. Morphol. Tiere 77: (1571748 Burmann, K. (1958): Eriocrania alpinella nov. spec. (Lepidoptera, Eriocraniidae). — Z. Wiener Entomol. Ges. 43: 269—271. Common, I. F. B. (1975): Evolution and classification of the Lepidoptera. — Ann. Rev. Entomol. 2071832203: Cushman, R. A. (1916): Thersilochus conotracheli, a parasite of the plum curculio. — J. Agr. Res. 6: 847 — 854. Davis,D.R. (1978): A revision of the North American moths of the superfamily Eriocranioi- dea with the proposal of a new family Acanthopteroctetidae (Lepidoptera). — Smith- sonian Contrib. Zool. 251. Dysart,R.J., H. L. Maltby &M. H. Brunson (1973): Larval parasites of Oulema mela- nopus in Europe and their colonization in the United States. — Entomophaga 18: 153167 Neue Parasitoiden minierender Trugmotten 419 Gauld, I.D. &B. Bolton (1988): The Hymenoptera, British Museum. — New York: Oxford Univ. Press. Heath, J. (1976): The moths and butterflies of Great Britain and Ireland, volume I. Microp- terigidae-Heliozelidae. — Blackwell Scientific Publications, Oxford & Curven Press, London. Hering, E. M. (1957): Bestimmungstabellen der Blattminen von Europa einschließlich des Mittelmeerbeckens und der Kanarischen Inseln, Band I-III. — Uitgeverij Dr. W. Junk, s'Gravenhage, Den Haag. Horstmann, K. (1971): Revision der europáischen Tersilochinen (Hymenoptera, Ichneumo- nidae) Teil I. — Veröff. Zool. Staatssamml. München 15: 45—138. Horstmann, K. (1981): Revision der europäischen Tersilochinae II (Hymenoptera, Ichneu- monidae). — Spixiana Suppl. 4. Jordan, T. (1992): Biologie und Parasitoide der im Frühjahr an Birken minierenden Trug- motten (Lep., Eriocraniidae). — Dissertation Universität Kiel, Deutschland. Jourdheuil, P. (1960): Influence de quelques facteurs écologiques sur les fluctuations de population d'une biocenose parasitaire. — Ann. Epiphyties 11: 472—479. Karsholt, O., M. V. Kozlov & N. P. Kristensen (1994): Eriocrania cicatricella (Zetter- stedt, 1839), the correct name of the moth currently known as Eriocrania haworthi Bradley, 1966 (Lepidoptera, Eriocraniidae). — Entomol. medd. 62: 91—93. Koponen, S. (1974): On the occurrence and ecology of Eriocrania spp. (Eriocraniidae) and other mining insects of the birch in northernmost Fennoscandia in 1973. — Rep. Kevo Subarctic Res. Stat. 11: 52—64. Klingenberg, A. von B. Ulber (1994): Untersuchungen zum Auftreten der Tersilochinae (Hym., Ichneumonidae) als Larvalparasitoide einiger Rapsschädlinge im Raum Göttingen 1990 und 1991 und zu deren Schlupfabundanz nach unterschiedlicher Bodenbearbeitung. — J. appl. Entomol. 117: 287 —299. Montgomery, V. E. £ P. R. De Witt (1975): Morphological differences among immature © stages of three genera of exotic larval parasitoids attacking the cereal leaf beetle in the United States. — Ann. Ent. Soc. Am. 68: 574—578. Parker, H. L. & P. A. Berry (1950): Vegetable weevils and their natural enemies in Argen- tina and Uruguay. — Tech. Bull. USDA Washington: 1016. Pschorn-Walcher, H. (1985): Konkurrenz und Mannigfaltigkeit bei Parasitoiden. — Z. zool. Syst. EvolForschung 23: 268—298. Dr. Thorsten Jordan, Lohgerberstr. 1A, D-25840 Friedrichstadt. mint | soaks ede A sem ru de dE Ve: i a y a ¡NIDO ni Dr nfa y pp AN in u Gi Emm E 1 tl ih ta do eel mn oe a ; j Au Jl u, b Je A Ñ O ay x 179 el f ne I y i #7: i h ; Piss =" x t a ' An 4 Y j ' ; 1; Ar J = Ai THA ee ; ee Al 1] cP j > x ii 7 a - ih shy q I p 4 © 7 M fe” Dr} ri | R 2 Sl Bd. 47 S. 421—428 Bonn, September 1998 Gorbothorax n. gen., a new linyphiid spider genus from the Nepal Himalayas (Arachnida, Araneae, Linyphiidae)” Andrei V. Tanasevitch Abstract. A new spider genus, Gorbothorax n. gen., and four new species, comatus n. sp. (the type species), conicus n. sp., setifer n. sp., and ungibbus n. sp., are described from the Nepal Himalayas. One more species is transferred to Gorbothorax, i.e. maculatus (Wunderlich, 1974), comb. n. ex Oedothorax Bertkau, 1883. Key words. Arachnida, Linyphiidae, taxonomy, Nepal. This work continues my studies on the linyphiid fauna of Nepal, this time devoted to the description of a new genus and a few new species. Holo- and the majority of paratypes have been deposited in the Senckenberg Museum, Frankfurt a. M. (SMF), some paratypes have become part of the collections of the Zoological Museum of the Moscow State University, Moscow (ZMMU) and the Zoologisches Forschungs- institut und Museum A. Koenig, Bonn (ZFMK). The following abbreviations are used in the text and the figures: Ti— tibia, Mt — metatarsus, Tm I— position of the metatarsal trichobothrium, ED — embolic division, SA — suprategular apophysis. Chaetotaxy is given in the following formula: 2.2.1.1 refers to the number of dorsal spines on Ti I—IV. The sequence of leg segments in the measurement data is as follows: femur + patella + tibia + metatarsus + tarsus. All measurements are given hereinafter in mm. Scale — 0.1 mm, if not otherwise indicated. Gorbothorax n. gen. Type species: Gorbothorax comatus n. sp. Etymology: The generic name refers to the specific shape of the male carapace, both deriving from gorb (in Russian meaning “hump”), and Latin fhorax. Diagnosis: The genus is characterized by the combination of the following features: chaetotaxy 2.2.1.1, presence of a trichobothrium on Ti IV and its distal position on Ti I, large paracymbium, large, complete and well sclerotized suprategu- lar apophysis, the embolic division divided into two separate sclerites, embolus long and thin. Description: Medium-sized erigonines, total length 1.80—2.60, with a pale brown coloration and a distinct abdominal pattern. Male carapace modified (except in ungibbus n. sp.). Chelicerae proportionate in size with cephalothorax, unmodified, D Results of the Himalaya Expeditions of J. Martens, No 216. — For No. 215, see Senckenbergiana biol. 77, 1997. — J. M. sponsored by Deutscher Akademischer Austauschdienst and Deutsche Forschungsgemeinschaft. 422 A. V. Tanasevitch Figs 1—3: Male carapace of Gorbothorax spp. — 1) comatus n. sp., paratype from Mai Pokhari; 2) conicus n. sp., holotype; 3) setifer n. sp., holotype. without frontal tooth. Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I: 0.70—0.85. Palpal tibia modified, with two outgrowths: a thick and rounded dorso-retrolateral as a rule bearing a few very short, thick, terminal spines, and a dorso-prolateral one. Paracymbium relatively large, black, carrying several spines in distal part. Suprategular apophysis well-developed and sclerotized, complex in shape. Embolic division divided into two separate sclerites connected with short, poorly visible membrane. The first sclerite — embolus — long, thin and curved. The second Gorbothorax, a new linyphiid spider genus 423 Figs 4—7: Gorbothorax comatus n. sp., male paratype from Mai Pokhari. — 4,5) right palp; 6) palpal tibia (dorsal view); 7) abdomen (dorsal view). one is relatively large, well sclerotized and elongated. The latter was previously called by Merrett (1963) “lamella”, and as this term was wittingly invalid, the author and followers have used it in quotation marks. For this separated sclerite of the embolic division, which connects with the embolus by the membrane, I propose a new term, convector, which in Latin means “companion”. This term may apply at least to the closely related genera of the Gongilidium genus group of Millidge (1977), Oedothorax Bertkau, 1883, Gongylidium Menge, 1868, Gongylidioides Oi, 1960 and Gorbothorax n. gen. In these genera this sclerite is obviously homologous. Taxomonic remarks: This genus is a member of the Gongylidium genus group of Millidge (1977) and closely related to the Far-Eastern Gongylidioides. Gorbotho- 424 AN Tama seviten rax n. gen. differs by the shape of the convector (elongate and without outgrowth) and a curved embolus. From Oedothorax, another closely related genus, the new genus differs by a well-developed and sclerotized suprategular apophysis and a long embolus. Species included: The new genus currently comprises five species, four of them being new to science. These are comatus n. sp., conicus n. sp., setifer n. sp., and ungibbus n. sp., as well as maculatus (Wunderlich, 1974), comb. n. ex Oedothorax. Distribution. All species of Gorbothorax are distributed in the Nepal Himalayas. The close relation to Gongylidioides demonstrates important relationships between Himalayan and East Asian faunas. Figs 8-12: Gorbothorax conicus n. sp., male paratype from Omje Kharka. — 8,9) right palp; 10) palpal tibia (dorsal view); 11) embolic division; 12) abdomen (dorsal view). Neue Parasitoiden minierender Trugmotten 425 Figs 13—15: Gorbothorax setifer n. sp., male holotype. — 13,14) right palp; 15) palpal tibia (dorsal view). Gorbothorax comatus n. sp. (Figs 1, 4—7) Material: Holotype male (SMF), Nepal, Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller. — Paratypes: 4 males (SMF), 1 male (ZFMK), 1 male (ZMMU), same locality, together with holotype; 2 males (SMF), Ilam Distr., Mai Pokhari, 2100— 2200 m, forest, 25.—27. III. 1980, leg. J. Martens & A. Ausobsky; male (SMF), Panchthar Distr., between Paniporua and Hinwa Khola Valley, 1850—2300, cultivated land, tree-rich, 20. IV. 1988, leg. J. Martens & W. Schawaller; 2 males (SMF), Taplejung Distr., Yamputhin, cultivated land, open forest, 1650—1800, 26. IV.—1. V. 1988, leg. J. Martens & W. Schawaller. Etymology: The specific name can be translated from Latin as “shaggy”. Diagnosis: The new species is well recognizable by the specific shape of the male carapace, as well as by the shape of the frontal part of the convector, which carries numerous small, pointed teeth. Description: Male (female unknown). Total length 2.33. Carapace 1.13 long, 0.80 wide, pale reddish-brown. Cephalic part of carapace behind eye area with a hump-shaped elevation carrying numerous long, curved spines (Fig. 1). Chelicerae 0.40 long. Legs pale brown. Leg I 4.01 long (1.10 + 0.30 + 1.03 + 0.95 + 0.63), IV—-4.10 long (1.15 + 0.30 + 1.05 + 1.05 + 0.55). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I—0.70. Palp (Figs 4—6): Dorso-retrolateral outgrowth of tibia conical, with several very shot spines terminally, dorso- prolateral outgrowth blunt. Paracymbium large, dark, carrying a few spines in distal part. Frontal part of convector black, with numerous small teeth. Abdomen 1.23 long, 0.85 wide, pattern as in Fig. 7. 426 A. V. Tanasevitch Taxomomic remarks: This species is closely related to conicus n. sp., being well distin- guished by the shape of the cephalic elevation, frontal process of the embolic division (narrow and toothless in conicus n. sp., thick and serrate in comatus n. sp.), and longer embolus. Gorbothorax conicus n. sp. (Figs 2, 8—12) Material: Holotype male (SMF), Nepal, Taplejung Distr., ridge Lasse Dhara und pasture Lassetham, 3000—3300 m, 6—7. IX. 1983, leg. J. Martens & B. Daams. Paratype: male (ZFMK), Taplejung Distr., Omje Kharka NW Yamputhin, natural mixed broadleaved forest, 2300—2500 m, 1.—6. V. 1988, leg. J. Martens & W. Schawaller. Etymology: The specific name describes the shape of the male cephalic elevation. Diagnosis: The new species is well recognizable by the specific shape of the male carapace, as well as by the shape of the palpal tibia and long, narrow and pointed distal part of the suprategular apophysis. Description: Male (female unknown). Total length 2.43. Carapace 0.93 long, 0.75 wide, pale reddish-brown. Cephalic part of carapace partly depigmented, with a conical “hump” behind eye area carrying long, strongly curved setae (Fig. 2). Chelicerae 0.45 long. Legs pale brown. Leg I 3.39 long (0.93 + 0.28 + 0.83 + 0.80 + 0.55), IV—-3.46 long (0.95 + 0.25 + 0.85 + 0.88 + 0.53). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I —0.70. Palp (Figs 8—11): Dorso-retrolateral outgrowth of tibia conical, with several very shot spines terminally, dorso-prolateral outgrowth pointed. Paracymbium dark, carrying a few spines in Figs 16—19: Gorbothorax ungibbus n. sp., male holotype. — 16,17) right palp; 18) palpal tibia (dorsal view); 19) abdomen (dorsal view). Gorbothorax, a new linyphiid spider genus 427 GORBOTHORAX Gebiete zwischen 3000 und 6000m Ka = Kathmandu 50 100 200km Fig. 20: Collecting sites of Gorbothorax species in Nepal (Martens collection). — 1) between ridge Lassedara and pasture Lassetham, also Omje Kharka (G. conicus n. sp.); 2) Yamputhin (G. comatus n. sp.); 3) Tinjura Dara (G. setifer n. sp.); 4) Paniporua and descent to Hinwa Khola (G. comatus n. sp.); 5) Mai Pokhari (G. comatus n. sp.); 6) Sanishare (G. ungibbus n. sp.). — According to species: G. comatus n. sp.: 2, 4, 5; G. conicus n. sp.: 1; G. setifer n. sp.: 3; G. ungibbus n. sp.: 6. distal part. Suprategular apophysis large, distally elongate to slightly curved and pointed. Frontal part of convector narrow, toothless. Abdomen 1.53 long, 0.95 wide, dorsal pattern as in Fig. 12. Taxomomic remarks: This species is closely related to comatus n. sp. (see above). Gorbothorax setifer n. sp. (Figs 3, 13—15) Material: Holotype male (SMP), Nepal, Terhathum Distr., Tinjura Dara, 2450—2850 m, species-rich broadleaved forest, Berlese funnels, 17. IX. 1983, leg. J. Martens & B. Daams. Etymology: The specific name is derived from both seta meaning a “strong spine”, and ferens meaning “bearing”. Diagnosis: The new species is well distinguishable from other congeners by the very peculiar shape of the carapace, structure of the palpal tibia, as well as by the shape of the suprategular apophysis. Description: Male (female unknown). Total length 2.55. Carapace 1.20 long, 0.85 wide, greyish-brown, modified as in Fig. 3. Chelicerae 0.50 long. Legs pale brown. Leg I 4.22 long (1.18 + 0.33 + 1.05 + 1.03 + 0.63), IV—4.27 long (1.23 + 0.30 + 1.03 + 1.13 + 0.58). Chaeto- taxy 2.2.1.1, leg spines very short. Each metatarsus with a trichobothrium. Tm I— 0.77. Palp (Figs 13—15): dorso-retrolateral outgrowth of tibia rounded, with numerous, very short, barb- like spines terminally, dorso-prolateral outgrowth small. Distal part of paracymbium narrow, dark, carrying a few spines. Suprategular apophysis large, complex in shape. Frontal part of convector narrow, dark. Abdomen 1.38 long, 0.88 wide, dark grey, dorsal pattern very vague. 428 A. V. Tanasevitch Oedothorax ungibbus n. sp. (Figs 16—19) Material: Holotype male (SMF), Nepal, Ilam Distr., Sanishare, 5 km N, feet of Siwalik Mts., 270—300 m, mixed Shorea forest, 3.—5. IV. 1988, leg. J. Martens & W. Schawaller. Etymology: The specific name describes the shape of the male carapace referring to the absence of a “hump”. Diagnosis: The new species is diagnosed by the unmodified male carapace, relatively small and spineless paracymbium, very long and curved embolus, as well as by the shape of the frontal part of the convector. Description: Male (female unknown). Total length 1.83. Carapace 0.85 long, wide, 0.68, greyish-yellow, unmodified. Chelicerae 0.38 long. Legs pale yellow. Leg I—-3.23 long (0.88 + 0.24 + 0.85 + 0.78 + 0.48), IV — 3.13 long (0.85 + 0.20 + 0.80 + 0.83 + 0.45). Chaeto- taxy 2.2.1.1. Tm] —0.84. Palp (Figs 16—18): Dorso-retrolateral outgrowth of tibia with several very short spines terminally. Paracymbium relatively small, spineless. Suprategular apophysis complex in shape, distally pointed. Embolus very long and curved. Frontal part of the convec- tor with several large black teeth. Abdomen 1.03 long, 0.65 wide, dorsal pattern as in Fig. 19. Gorbothorax wunderlichi (Brignoli, 1983) comb. n. Oedothorax maculatus Wunderlich, 1974. Senckenbergiana biol. 55: 185, figs 51—58. Oedothorax wunderlichi Brignoli, 1983. A Catalogue of the Araneae: 324, nom. nov. pro Oedothorax maculatus Wunderlich, 1974, praeocc. This species was described from East Nepal by Wunderlich (1974) as Oedothorax maculatus Wunderlich, 1974. According to the above author's figures, this is a good member of Gor- bothorax, comb. n. ex Oedothorax. The collecting sites of all species mentioned are shown in Fig. 20. Acknowledgements I am very grateful to Prof. J. Martens (Mainz) and Dr. W. Schawaller (Stuttgart), whose collec- tions served as the basis for the present paper. In addition, Dr. S. I. Golovatch (Moscow) kindly checked the English of the final draft. This work was partly supported by the Soros Foundation (“Biodiversity”). Zusammenfassung Auf der Basis von in Nepal gesammelten Materials wird die Spinnengattung Gorbothorax n. gen. mit den vier Arten comatus n. sp. (Typusart), conicus n. sp., setifer n. sp., und ungibbus n. sp. beschrieben. Eine weitere Art wird von Oedothorax Bertkau, 1883 nach Gorbothorax transferiert. References Merrett, P. (1963): The palpus of male spiders of the family Linyphiidae. — Proc. Zool. Soc. London 140: 347—467. Millidge, A. FE. (1977): The conformation of the male palpal organs of linyphiid spiders, and its application to the taxonomic and phylogenetic analysis of the family (Araneae: Linyphiidae). — Bull. Brit. arachnol. Soc. 4: 1—60. Wunderlich, J. (1974). Linyphiidae aus Nepal. II. Die Gattung Oedothorax Bertkau 1833 (Arachnida, Araneae). — Senckenbergiana biol. 55: 169—188. Dr. Andrei V. Tanasevitch, All-Russian Institute on Nature Conservation and Reser- ves, 113628, O.P. Vilar, Moscow, Russia. Bd. 47 S. 429—441 Bonn, September 1998 New Oedothorax Bertkau, 1883, from Nepal (Arachnida, Araneae, Linyphiidae)” Andrei V. Tanasevitch Abstract. Eleven new species of the spider genus Oedothorax Bertkau, 1883, from the Nepal Himalayas are described. Key words. Arachnida, Linyphiidae, new species, Nepal. The spider genus Oedothorax Bertkau, 1883, seems to be one of the most speciose among Himalayan erigonines. From various parts of the Himalayas, no fewer than 12 species of this genus have hitherto been described: dubius Caporiacco, 1935 (from a juvenile specimen) from Karakorum (Caporiacco 1935); annulatus Wunderlich, 1974, asocialis Wunderlich, 1974, dismodicoides Wunderlich, 1974, elongatus Wun- derlich, 1974, hirsutus Wunderlich, 1974, lineatus Wunderlich, 1974, lucidus Wunder- lich, 1974, sexoculatus Wunderlich, 1974, unicolor Wunderlich, 1974, and wunder- lichi Brignoli, 1983, all from Nepal (Wunderlich 1974, Brignoli 1983), and globiceps Thaler, 1987, from Kashmir (Thaler 1987). This paper augments the Nepalese fauna : by another 12 new Oedothorax. Although many females are present in the material, the majority of the descriptions are based on the male sex only. The reason for that lies in the difficulty to match the male with the proper female when several congener candidates are plausible, all deriving from the same sample. Contrary to numerous Far Eastern or African Oedothorax, which seem to actually belong to other genera, the Himalayan fauna appears to comprise a swarm of species clearly congeneric with the type-species, Oedothorax gibbosus (Blackwall, 1841). However, wunderlichi Brignoli, 1983 (nomen novum pro Oedothorax maculatus Wunderlich, 1974, praeocc.) proves to be a member of another genus. In 1974 J. Martens stated (in Wunderlich 1974: 171) that the eleven species of Oedothorax described in that paper originated from a comparatively narrow vertical area belt of the southern macroslope of the Nepal Himalayas, the localities being situated between 1700 and 2900 m. They comprise the subtropical evergreen moun- tain forest and the evergreen cloud forest communities. J. Martens collected, how- ever, during his long-lasting stays in Nepal also at altitudes even above 5500 m (Martens 1987), but no Oedothorax species showed up above 3000 m, namely the Palaearctic part of the Nepal Himalayas. All Oedothorax localities are heavily influenced by monsoon precipitations (except lake Titi, which is on the borderline to drier habitats). In dry rain shadow areas north of the Himalayan main chain, no representative of the genus was discovered. Results of the Himalaya Expeditions of J. Martens, No 217. — For No. 216, see Bonn. zool. Beitr. 47, 1998. — J.M. sponsored by Deutscher Akademischer Austauschdienst and Deutsche Forschungs- gemeinschaft. 430 A. V. Tanasevitch The new records of eleven further Oedothorax species, which are presented here, fall, too, exactly into this distribution scheme. Again, no record is higher than 3000 m, the lowest, as already indicated by Wunderlich (1974), no lower than 1700 m, but these belong well to the subtropical belt. In general, Oedothorax in Nepal shows no clear-cut affinities to Palaearctic climate and biotopes, and it is to be questioned how close the affinities to the true Palaearctic species set of the genus really are. Holo- and the majority of paratypes have been deposited in the Senckenberg Museum, Frankfurt a. M. (SMF), some duplicate paratypes have become part of the collections of the Zoological Museum of the Moscow State University, Moscow (ZMMU) and Zoologisches Forschungsinstitut und Museum A. Koenig, Bonn (ZFMK). | ; The following abbreviations have been accepted in the text and figures: Ti — tibia, Mt — metatarsus, Tm I — position of the metatarsal trichobothrium, AME — anter- ior, and PME — posterior median eyes, respectively. Chaetotaxy is given in the following formula: 2.2.1.1 refers to the number of dorsal spines on Ti I—IV. The sequence of leg segments in the measurement data is as follows: femur + patella + tibia + metatarsus + tarsus. All measurements are given hereinafter in mm. Scale — 0.1 mm, if not otherwise indicated. Figs 1—5: Oedothorax assuetus n. sp., male and female paratypes. — 1) male carapace (lateral view); 2) right palp; 3) palpal tibia (dorsal view); 4) male abdomen (dorsal view); 5) epigynum. New Oedothorax Bertkau, 1883, from Nepal 431 Oedothorax assuetus n. sp., Figs 1—5. Material: Holotype male (SMF), Nepal, Kathmandu, Godawari, foot of Phulchoki Mt., 1700 m, 19. III. 1980, leg. J. Martens & A. Ausobsky. — Paratype: 1 male (SMF), same locality, together with holotype, leg. J. Martens & A. Ausobsky. Diagnosis: The new species is diagnosed by the shape of the male carapace's cephalic elevation and structure of the palpal tibia. The epigyne is of ordinary shape and close to many other congeners. Etymology: The specific name assuetus is a Latin adjective meaning “ordinary”, referring to the common shape of the epigyne. Description: Male. Total length 2.38. Carapace (Fig. 1) 1.08 long, 0.75 wide, pale brown, with a wide dark ring (except for cephalic division). Chelicerae 0.48 long. Legs pale brown. Leg 1— 3.62 long (0.95 + 0.88 + 0.28 + 0.88 + 0.63), IV — 3.68 long (1.00 + 0.95 + 0.28 + 0.95 + 0.50). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.57. Palp as in Figs 2—3. Abdomen 1.53 long, 0.90 wide, dorsal pattern as in Fig. 4. Female. Total length 2.45. Carapace 1.03 long, 0.70 wide, unmodified. Chelicerae 0.45 long. Leg 1— 3.39 ¡ong (0.93 + 0.85 + 0.30 + 0.78 + 0.53), IV — 3.54 long (0.98 + 0.88 + 0.28 + 0.90 + 0.50). TmI — 0.53. Abdomen 1.70 long, 1.20 wide. Epigynum as in Fig. 5. Body and legs coloration, chaetotaxy as in male. Oedothorax coronatus n. sp., Figs 6—9, 10—13. Material: Holotype male (SMF), Nepal, Ilam Distr., Mai Pokhari, 2100—2200 m, forest, 25.—27. IM. 1980, leg. J. Martens & A. Ausobsky. — Paratypes: 5 males (SMF), 2 males (ZMMU), same locality, together with holotype, 25.—27. III. 1980, leg. J. Martens & Figs 6—9: Male carapace of Oedothorax coronatus n. sp., 6—7) form a, paratype from Ilam Distr.; 8 —9) form b, paratype from Panchthar Distr. — 6 & 8) lateral view; 7 & 9) dorsal view. 432 A V Tanaseviteh A. Ausobsky; 2 males (SMF), 2 males (ZFMK), Ilam Distr., Mai Pokhari, 2100, forest, 31. III.—1. IV. 1980, leg. J. Martens & A. Ausobsky; 1 male (SMF), Ilam Distr., Gitang Khola, 1900—2100 m, cultivated land, 31. III. 1980, leg. J. Martens & A. Ausobsky; 1 male (SMF), 1 male (ZMMU), Ilam Distr., Mai Pokhari, 2100—2200 m, Castanopsis forest remains, 9.—10. IV. 1988, leg. J. Martens & W. Schawaller; 4 males [carapace form b] (SMF), Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller; 4 males [carapace form b] (SMF), 1 male [carapace form b] (ZMMU), Taplejung Distr., Worebung Pass, degraded broadleaved forest, 2000, 21. IV. 1988, leg. J. Martens & W. Schawaller. Diagnosis: The new species is characterized by the shape of both male carapace and palpal tibia. Etymology: The specific name coronatus is a Latin adjective meaning “crowned”, referring to the male cephalic structure. Description: Male (female unknown). Total length 2.48. Carapace 1.18 long, 0.90 wide, medially with a large conical “hump”. Cephalic division elevated, with four specific setae. Two different shapes of carapace: carapace form a — Figs 6—7 (specimen from Ilam Distr.), and carapace form b — Figs 8—9 (specimen from Panchthar and Taplejung distr.). No differences whatever in palpal structure regardless of carapace shape. Chelicerae 0.48 long. Legs pale brown. Leg I—4.46 long (1.25 + 1.10 + 0.30 + 1.13 + 0.68), IV —4.78 long (1.38 + 1.25 + 0.30 + 1.30 + 0.55). Chaetotaxy: 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.80. Palp as in Figs 10—12. Abdomen 1.38 long, 0.90 wide, dorsal pattern as in Fig. 13. Taxonomicremarks: The new species is very closely related to angelus n. sp., being well distinguished by the shape of the cephalic setae, presence of a carapace “hump”, as well as by the shape of the palpal tibia. Figs 10—13: Oedothorax coronatus n. sp., male paratype from Mai Pokhari. — 10) right palp; 11) palpal tibia (dorsal view); 12) embolic division; 13) abdomen (dorsal view). New Oedothorax Bertkau, 1883, from Nepal 433 Oedothorax angelus n. sp., Figs 14—18 Material: Holotype male (SMF), Nepal, Panchthar Distr., Dhorpar Kharka, Rhododen- dron & Lithocarpus forest, 2700 m, 13.—16. IV. 1988, leg. J. Martens & W. Schawaller. — Paratypes: 4 males (SMF), 1 male (ZMMU), same locality, together with holotype, leg. J. Martens & W. Schawaller; 4 males (SMF), 2 males (ZFMK), 1 male (ZMMU), Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller. Diagnosis: The new species is characterized by the shape of both male carapace and palpal tibia. Etymology: The specific name from Latin is translated as an “angel” and refers to the shape of the male cephalic setae looking like the angel wings. Description: Male (female unknown). Total length 2.00. Carapace 0.95 long, 0.73 wide, cephalic part carrying two large and two small characteristic setae situated in central part of eye area (Figs 14—15). Chelicerae 0.28 long. Legs pale brown. Leg 1— 3.29 long (0.88 + 0.25 + 0.90 + 0.78 + 0.48), IV—3.18 long (0.90 + 0.20 + 0.85 + 0.83 + 0.40). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I— 0.81. Palp (Figs 16—18). Abdomen 1.05 long, 0.68 wide, dorsal pattern close to that of coronatus n. sp. Taxonomic remarks: The new species is close to coronatus n. sp. (see above). Figs 14—18: Oedothorax angelus n. sp., male paratype from Dharpar Kharka. — 14—15) carapace (lateral and dorsal views, respectively); 16—17) right palp; 18) palpal tibia (dorsal view). 434 A. V. Tanasevitch Oedothorax sexoculorum n. sp., Figs 19—23. Material: Holotype male (SMF), Nepal, Terhathum Distr., Tinjura Dara, 2450—2850 m, species-rich mixed broadleaved forest, Berlese funnels, 17. IX. 1983, leg. J. Martens & B. Daams. Diagnosis: The new species is diagnosed by the shape of the male carapace, presence of six Operating eyes: PME covered by the cephalic elevation (maybe this is a teratology: one specimen is known only!), elongated and distally pointed suprategular apophysis, elongated parts of the embolic division, as well as by the absence of a scapuliform apophysis on the embolic division (s. Fig. 12). Etymology: The specific name is a Latin adjective translated as “possessing six eyes”. Description: Male (female unknown). Total length 2.00. Carapace modified (Figs 19—20), 0.98 long, 0.68 wide, pale brown, with a wide dark ring. Posterior median eyes covered by cephalic elevation, thus male with only six operating eyes. Chelicerae 0.40 long. Legs pale brown. Leg 1— 2.99 long (0.80 + 0.25 + 0.73 + 0.73 + 0.48), IV —3.06 long (0.83 + 0.23 + 0.75 + 0.80 + 0.45). Chaetotaxy 2.2.1.1. Tm I—0.76. Palp as in Figs 21—22. Abdomen 1.08 long, 0.65 wide, dorsal pattern as in Fig. 23. Taxonomic remarks: There is another Himalayan congener which has six eyes, namely O. sexoculatus Wunderlich, 1974, described from Jiri, East Nepal (Wunderlich 1974). How- ever, both species in question are well distinguishable by the shape of the palpal tibia. 0.5 21-22 23 19-20 0.5 Figs 19—23: Oedothorax sexoculorum n. sp., male holotype. — 19—20) carapace (lateral and frontal views, respectively); 21) right palp; 22) palpal tibia; 23) abdomen (dorsal view). New Oedothorax Bertkau, 1883, from Nepal 435 Oedothorax tholusus n. sp., Figs 24—29. Material: Holotype male (SMF), Nepal, Kaski Distr., above Dhumpus, broadleaved forest, 2100 m, 8.—10. V. 1980, leg. J. Martens & A. Ausobsky. — Paratype: 1 female (SMF), same locality, together with holotype, leg. J. Martens & A. Ausobsky. Diagnosis: This species is easily distinguished by the shape of the male carapace, small suprategular apophysis, absence of a scapuliform apophysis on the embolic division (s. Fig. 12), as well as by the shape of the palpal tibia. Etymology: The specific name is a Latin adjective meaning “cupola of temple”, this referring to the shape of the male cephalic elevation. Description: Male. Total length 2.70. Carapace 1.25 long, 0.95. wide, reddish-brown, cephalic elevated as in Figs 24—25. Chelicerae 0.55 long. Legs reddish-brown. Leg I 4.01 long (1.13 + 0.30 + 0.93 + 1.00 + 0.65), IV—4.18 long (1.15 + 0.30 + 1.05 + 1.10 + 0.58). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.60. Palp as in Figs 26—27. Abdomen 1.50 long, 0.93 wide, pattern as in Fig. 28. Se | 26-27, 29 28 24-25 Figs 24—29: Oedothorax tholusus n. sp., male holotype and female paratype. — 24—25) male carapace (lateral and frontal views, respectively); 26) right palp; 27) palpal tibia; 28 male abdomen (dorsal view); 29) epigynum. 436 A. V. Tanasevitch Female. Total length 2.18. Carapace 0.75 long, 0.63 wide. Chelicerae 0.30 long. Leg I — 2.06 long (0.60 + 0.23 + 0.48 + 0.45 + 0.30), IV — 2.26 long (0.65 + 0.20 + 0.55 + 0.53 + 0.33). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.72. Abdomen 1.55 long, 1.08 wide. Epigynum as in Fig. 29. Body and legs coloration, chaetotaxy as in male. Taxonomic remarks: This species seems to be close to globipes Thaler, 1987, from Kashmir, being well distinguishable by the structure of the male palp. Oedothorax clypeellum n. sp., Figs 30—33. Material: Holotype male (SMF), Nepal, Kathmandu, Phulchoki Mt., pitfall traps, 2600 m, 21. III.—14. V. 1980, leg. J. Martens & A. Ausobsky. Diagnosis: This new species is easy recognizable by the shape of both palpal tibia and clypeus, as well as by the presence of a tooth on the frontal surface of the chelicerae. Etymology: The specific name refers to the characteristic shape of the male clypeus. Description: Male (female unknown). Total length 2.53. Carapace 1.13 long, 0.85 wide, brown-grey, with a pale margin; cephalic division not elevated, clypeus conical (Fig. 30). Chelicerae 0.48 long, basally with a large conical tooth (Fig. 30). Legs pale brown. Leg I — 3.16 long (0.83 + 0.30 + 0.80 + 0.78 + 0.45), IV — 3.48 long (0.88 + 0.30 + 0.90 + 0.95 + 0.45). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I— 0.88. Palp as in Figs 31—32. Abdomen 1.38 long, 0.90 wide, dorsal pattern as in Fig. 33. 0.3 Ar 1 31-32 33 30 Figs 30—33: Oedothorax clypeellum n. sp., male holotype. — 30) carapace (frontal view); 31) right palp; 32) palpal tibia; 33) abdomen (dorsal view). New Oedothorax Bertkau, 1883, from Nepal 437 Oedothorax simplicithorax n. sp., Figs 34—35. Material: Holotype male (SMF), Ilam Distr., Gitang Khola Valley, 4/nus forest along river, 1750 m, 11.—13. IV. 1988, leg. J. Martens & W. Schawaller. — Paratype: 1 male (SMF), same locality, together with holotype, leg. J. Martens & W. Schawaller. Diagnosis: The new species is diagnosed by the structure of the palpal tibia, as well as by the small both suprategular apophysis and “embolus”. Etymology: The specific name refers to the unmodified shape of the male carapace. Description: Male (female unknown). Total length 1.93. Carapace 0.88 long, 0.73 wide, pale brown, unmodified. Chelicerae 0.33 long. Legs pale brown. Leg I — 3.23 long (0.85 + 0.25 + 0.83 + 0.75 + 0.55), IV—3.41 long (0.95 + 0.28 + 0.85 + 0.83 + 0.50). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.60. Palp as in Figs 34—35. Abdomen 1.13 long, 0.70 wide, dorsally grey, with a pale median stripe. 0.5 0.5 == 34-35, 37-38 39 36 Figs 34—39: Oedothorax simplicithorax n. sp. (34, 35), male holotype, and O. modestus n. sp. (36—39), male holotype. — 34 & 37) right palp; 35 & 38) palpal tibia (dorsal view); 36) carapace (lateral view); 39) abdomen (dorsal view). 438 A. V. Tanasevitch Oedothorax modestus n. sp., Figs 36—39. Material: Holotype male (SMF), Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller. — Paratypes: 5 males (SMF), 1 male (ZMMU), same locality, together with holotype leg. J. Martens & W. Schawaller. Diagnosis: The new species is characterized by the shape of both carapace and palpal tibia. Etymology: The specific name is a Latin adjective meaning “moderate”, “temperate”, this referring to the poorly modified male carapace. Description: Male (female unknown). Total length 2.05. Carapace 1.00 long, 0.75 wide, pale brown with a pale area situated behind cephalic division (Fig. 36). Chelicerae 0.33 long. Legs pale brown. Leg I — 3.78 long (1.00 + 0.30 + 0.93 + 0.90 + 0.65), IV — 3.98 long (1.10 + 0.27 + 0.98 + 1.00 + 0.63). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm T— 0.56. Palp as in Figs 37—38. Abdomen 1.15 long, 0.75 wide, dorsal pattern as in Fig. 39. 0.5 0.3 Ws 42-44 45 40-41 Figs 40—45: Oedothorax savigniformis n. sp., (40—44) male holotype, and (45) male para- type. — 40—41) carapace (lateral and frontal views, respectively); 42—43) right palp; 44) palpal tibia (dorsal view); 45) abdomen (dorsal view). A. V. Tanasevitch 439 Taxonomic remarks: This species is closely related to falciferus n. sp., the structure of the embolic division being almost identical. Both species compared are well distinguishable by the shape of the cephalic elevation and the structure of the palpal tibia. Oedothorax savigniformis n. sp., Figs 40—45. Material: Holotype male (SMF), Taplejung Distr., Yamputhin, ascent to pass Deorali, 2600 m, cultivated land, 16. V. 1988, leg. J. Martens & W. Schawaller. — Paratype: male (SMF) Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller. Diagnosis: This species is well distinguishable by the shape of the cephalic division. 085 ee Lao > 47-48, 51-52 49 46,50 Figs 46—52: Oedothorax falciferus n. sp. (46-49), male holotype, and O. malearmatus n. sp. (50—52), male holotype. — 46 & 50) carapace (lateral view); 47 & 51) right palp; 48 & 52) palpal tibia; 49) abdomen (dorsal view). 440 A. V. Tanasevitch Etymology: The specific name refers to the male cephalic shape similar to that of Savignia species. Description: Male. Total length 2.05. Carapace 1.05 long, 0.70 wide, modified (Figs 40—41), dirty-grey, with radial stripes, with a partly depigmented cephalic area. Chelicerae 0.33 long. Legs pale brown. Leg I — 3.01 long (0.83 + 0.25 + 0.70 + 0.70 + 0.53), IV — 3.14 long (0.88 + 0.25 + 0.78 + 0.80 + 0.43). Chaetotaxy 2.2.1.1. Each metatarsus with a tricho- bothrium. Tm I 0.64. Palp as in Figs 42—44. Abdomen 1.05 long, 0.73 wide, dorsal pattern of paratype as in Fig. 45. Abdomen of holotype dark grey, dorsal pattern absent. Taxonomic remarks: This species is closely related to both modestus n. sp. and falciferus n. sp., being well different by the shape of the cephalic part of the carapace. Oedothorax falciferus n. sp., Figs 46—49. Material: Holotype male (SMF), Ilam Distr., Worebung Pass, degraded broadleaved forest, 2000 m, 21. IV. 1988, leg. J. Martens & W. Schawaller. Diagnosis: The new species is diagnosed by the shape of both carapace and palpal tibia, as well as by the relatively large suprategular apophysis. Etymology: The specific name is a Latin adjective meaning “carrying a sickle”, thus referring to the shape of the male palpal tibia. Description: Male (female unknown). Total length 2.33. Carapace 1.13 long, 0.85 wide, pale brown, cephalic part with an elevation (Fig. 46). Chelicerae 0.48 long. Legs pale brown. Leg I—4.17 long (1.13 + 0.30 + 1.01 + 1.00 + 0.73), IV —4.36 long (1.20 + 0.25 + 1.13 + 1.13 + 0.65). Chaetotaxy 2.2.1.1. Each metatarsus with a trichobothrium. Tm I — 0.60. Palp as in Figs 47—48. Abdomen 1.20 long, 0.88 wide, dorsal pattern as in Fig. 49. Taxonomic remarks: This species is closely related to modestus n. sp. (see above). Oedothorax malearmatus n. sp., Figs 50—52. Material: Holotype male (SMF), Panchthar Distr., Paniporua, 2300 m, mixed broadleaved forest, 16.—20. IV. 1988, leg. J. Martens & W. Schawaller. Diagnosis: The new species is characterized by the shape of both carapace and palpal tibia. Etymology: The specific name consists of two Latin words: malus meaning “poorly, badly”, and armatus “armed”, referring the structure of the palpal tibia. Description: Male (female unknown). Total length 2.05. Carapace with a small cephalic elevation behind PME and with a very small ridge-shaped tubercle situated between PME and AME (Fig. 50), 0.83 long, 0.65 wide, pale brown, with a wide dark ring. Chelicerae 0.35 long. Legs pale brown. Leg I — 3.41 long (0.85 + 0.28 + 0.80 + 0.83 + 0.65), IV — 3.51 long (0.95 + 0.25 + 0.90 + 0.88 + 0.53). Chaetotaxy 2.2.1.1. Each metatarsus with a tricho- bothrium. Tm I— 0.44. Palp as in Figs 51—52. Abdomen 1.13 long, 0.73 wide, dark grey, dorsally with a pale median stripe at fore half of abdomen. Acknowledgements I am grateful to Prof. J. Martens (Mainz), Dr. W. Schawaller (Stuttgart), A. Ausobsky (Bischofshofen) and B. Daams, now Mrs. Martens (Mainz), whose collections have served as the basis for the present paper. In addition, Dr. S. I. Golovatch (Moscow) kindly checked the English of the final draft. This work has partly been supported by the Soros Foundation. Zusammenfassung In diesem Beitrag werden elf neue Arten der Spinnengattung Oedothorax Bertkau, 1883, aus dem Nepal-Himalaya beschrieben. New Oedothorax Bertkau, 1883, from Nepal 441 OEDOTHORAX Gebiete über 6000m NR ER, / N : N Na N Gebiete zwischen 3000 und 6000m 1 + ; + Ka - Kathmandu te === = 50 100 200km Fig. 53: Collecting sites of Odothorax species in Nepal (Martens collection, species of both papers, Wunderlich [1974] and the present study). — 1) Lete (O. dismodicoides Wunderlich); 2) lake Titi (O. dismodicoides, O. lineatus Wunderlich); 3) Gorapani Pass and between Gora- pani Pass and Ulleri (O. dismodicoides, O. lineatus); 4) above Dhumpus (O. tholusus n. sp.); 5) between Ramche and Dhunche (O. hirsutus Wunderlich); 6) Godawari and Phulchoki Mt. (O. unicolor Wunderlich, O. assuetus n. sp., O. clypeellum n. sp.); 7) Jiri and Mt. Chordung (O. annulatus Wunderlich, O. asocialis Wunderlich, O. sexoculatus Wunderlich, O. lucidus Wunderlich); 8) Tinjura Dara (O. sexoculorum n. sp.); 9) between Yamputhin and Pass Deorali (O. savigniformis n. sp.); 10) Worebung (O. coronatus n. sp., O. falciferus n. sp.); 11) Paniporua (O. coronatus n. sp., O. angelus n. sp., O. melearmatus n. sp., O. modestus n. sp., O. savigniformis n. sp.); 12) Dhorpar Kharka (O. angelus n. sp.); 13) Gitang Khola (O. coronatus n. sp., O. simplicithorax n. sp.); 14) Mai Pokhari (O. coronatus n. sp.). — According to species: O. angelus: 11, 12; O. annulatus: 7; O. asocialis: 7; O. assuetus: 6; O. clypeellum: 6; O. coronatus: 10, 11, 13, 14; O. dismodicoides: 1, 2, 3; O. falciferus: 10; O. hirsutus: 5; O. lineatus: 2; O. lucidus: 7; O. malearmatus: 11; O. modestus: 11; O. savigni- formis: 9, 11; O. sexoculatus: 7; O. sexoculorum: 8; O. simplicithorax: 13; O. tholusus: 4; O. unicolor: 6. References Caporiacco, L. (1935): Aracnidi dell’Himalaia e del Karakoram, raccolti dalla Missione italiana al Karakoram (1929 — VII). — Mem. Soc. ent. ital. (1935): 113-263. Martens, J. (1987): Remarks on my Himalayan Expeditions. — Courier Forsch.-Inst. Senckenberg 93: 7—31. Thaler, K. (1987): Über einige Linyphiidae aus Kashmir (Arachnida: Araneae). — Courier Forsch.-Inst. Senckenberg 93: 33—42. Wunderlich, J. (1974): Linyphiidae aus Nepal. II. Die Gattung Oedothorax Bertkau 1883 (Arachnida, Araneae). — Senckenbergiana biol. 55: 169—188. Dr. Andrei V. Tanasevitch, All-Russian Institute on Nature Conservation and Reserves, 113628, O. P. Vilar, Moscow, Russia. Ri Me m G an nl ü Md | yl ran. AR Fea A ch IA (in f in i ‘ ly } 7 > 3 } | er yl y ? ria inti - dí Ñ A i ña INFORMATION FOR CONTRIBUTORS Content. — Bonner zoologische Beitraege is a publication series of the Alexander Koenig Institute and Museum of Zoology, Bonn. It contains papers bearing on systematic and comparative zoology and related topics. Results of research carried out at the Institute or on the collections of the Museum may be given priority but other contributions are welcome. Language. — Manuscripts may be written in German, English or French. Submission of manuscripts. — Two copies of the manuscript should be sent to the Editor, Bonner zoolc sche Beitraege, Zoologisches Forschungsinstitut und Museum Alexander Koenig, Adenauerallee 150— 164, D 53113 Bonn, West Germany. Review. — Manuscripts will be reviewed by a board of associate editors and appropriate referees. Presentation. — Manuscripts should be clear and concise in style. Telegraphic style is recommended for descriptions. Literature should be cited in abbreviated form in the text and fully in the References section. Use “&” instead of “and” when citing co-authors (see below). Before preparing the final manuscript a recent issue of Bonner zoologische Beitraege should be consulted. The establishment of new taxa must conform with the requirements of the International Code of Zoolo- gical Nomenclature. Authors are expected to be familiar with the code and to comply with it. Type speci- mens of new taxa should be deposited in an institutional collection. The sequence of material should be: Title, Author(s), Abstract, Key Words, Introduc Methods, Results, Discussion, Acknowledgements, German Summary, References, Re iddress( Figure legends, Tables, Figures (each numbered and identified). Manuscripts should be typed, double-spaced throughout (including tables and legends) « on one side of DIN A4 paper sheets, with a 3 cm margin all round. Computer prints will only be conside is acceptable. Pages must be numbered on top. Only scientific names of genera and ce categories be underlined; leave other indications to the editor. References should strictly follow the style of the journal. Abbreviations of journals should follow the: World List of Scientific Periodicals or a similar work. a fk the ion of OS are: Dyte, C. E. (1959): Some interesting habitats of larval Dolichopodidae (Dipter: 95: 139—143. Dyte, C. E. (1975): Paty Pu — In: arado M. D. 8 D. E. en Schuchmann, K.-H., K. Krüger & R. Prinzinger (1983): Leer in hummingbirc 34.273277. For the design of figures and tables the format of the journal (126x 190 mm) should be considered. Ta should be kept to a minimum. Footnotes should be avoided. Proofs. — Galley proofs will be sent to authors for correction. Reprints. — Fifty reprints will be supplied free; additional reprints may be ordered with returned proof. 4 [Deutsche Fassung im nächsten Bonner zoologische Beiträge Band 47, Hefte 3—4, 1998 INHALT Report on some sigmodontine rodents collected in southeastern Brasil with descriptions of a new genus and six species P. Hershkovitz ...:..2%... es ee A biochemical and morphological investigation of Suncus dayi (Dobson, 1888) and discussion of relationship in Suncus Hemprich & Ehrenberg, 1833, Crocidura Wagler, 1832, and Sylvisorex Thomas, 1904 (Insectivora: Soricidae P Jenkins, M. Ruedi &M. Catzeflis.:. 2... (22. 2 2 pe Biogeographical and karyological data of the Microtus savii group (Rodentia, Arvicolidae) in Italy L. Galleni, R. Stanyon, L. Contadini & A. Tellini 2 7 Reproduction in Mus macedonicus (Mammalia: Rodentia) in the Balkans V. Vohralik, T. S. Sofianidou & D: Frynta ...:.... Vs Two new chameleons (Sauria: Chamaeleonidae) from isolated Afromontane forests in Sudan and Ethiopia : Co Tilbury 2 2 ol ee N The community of rock-dwelling cichlids in Lake Victoria O. Seehausen & N. Bouten...........2 2.2 22.202.222 202 New and little known Nebria (Epinebriola) from the eastern Nepal Himalayas (Coleoptera, Carabidae) V. G. Shilenkov........0 Gc. 022%. ze a aa 2 Die Sandlaufkäfer (Coleoptera: Cicindelida) des Comoé-Nationalparks, Elfen- beinküste: Faunistik, Zoogeographie und Okologie J. Fahr cis. eis as tues o beta nee Coleoptera Histeridae Hetaeriinae: description de nouveaux taxons, désignation de lectotypes et notes taxonomiques N. Degallier .....%2...... 8.020.220. 2 bana Aa Die europäischen Arten der Cheilosia alpina-Gruppe (Diptera, Syrphidae) C. Claussen. 206 sia. We ed Tersilochus curvator Horstmann und Tersilochus sp. n. (Ichneumonidae, Tersilochinae), neue Parasitoiden der an Birken minierenden Trugmotten (Lepidoptera, Eriocraniidae) T. Jordan... 2 3. a ee Gorbothorax n. gen., a new linyphiid spider genus from the Nepal Himalayas (Arachnida, Araneae, Linyphiidae) A. V. Tanasevitch... „ou... ee See er New Oedothorax Bertkau, 1883, from Nepal (Arachnida, un Linyphiidae) A. V. Tanasevitch. .. alt sea hn eres cee or eiene e This journal is printed on acid-free paper. 193 257 27% 283 293 301 313 321 345 381 411 421 4 UN 2 Pa DUNA A Ad HECKMAN BINDERY, TE NIC? Bound-To-Please” MAY 00 | N. MANCHESTER, INDIANA 46962 SMITHSONIAN INSTITUTION LIBRARIES >in i ne 3 9088 01257 3648_ | ¿ab bet . > . , ’ ... . hos. . .. e 4 ‘ ‘ ‘ee . ’ a . a \ ñ ys a “oo mp A . Fire ala 4 : : I ' ‘ / ‘ . FARBE" et ‘ ‘ . Ace sis y 1 dE ] Al ha , A e os Tee . . 5 . u tt . , os "os ee e . S a . e , . E A u. ‘ nte, . oven ; > ... . ; 7 x » R ‘ .. tte Pet e . Ret o AA 5 here eee y oe 5 a , Ñ = rear e " Ne. « a : D D 5 . , ag iv? : , : E da e . er. = a, . ” A : Er © „fie Fi 4 af . ‘ x mae EEE >} i tates ahs ys 5 er een | A ; 5 y A pt : y : N E P uo. . A UN ei le y Ñ * ss daha > : . 5 : Pete x Lo.” y E . 1 . = . car 5 » A ee a A Sec 2 len gen & ‘ ey ; ir Bei ti Han) ea: BR ne Cry S FE > A y DEE OE cies NN i N $ ate r 248 ure se . Sees CODA > 7 . = h PR 3 i MN cus : F 5 a at po ps , nie, ; pS “peas: . ean é ‘ ; : . Pi vs S Se Er qe ci: ‘ a 5 y e e y N ae trad ie - ; A : P : ier a hi . Gus ee 5 = . TEST Een z : : e Ñ rs y 3 ‘ E ¿ O A a ic | y rn ae 5 : * . . . > E . € A A : ‘ 7 . $ , a or . ag z = a E LEN 2 ie e ART be cr ae ‘ . . . a . . Erster, : dl A F dien, ta : ‘ ‘ aie > Pre Er ER 2 . Thor ies d : Peed teat a] É ce 2 andren treed. E 4 Satara ge . a 4 . e La “. “se = 1 sl Sur .. u 1 Aa ale . Br 2 : ‘ “erst Sota . R . : 5 A E en E EOS . E £ 5 by af EL! H wis, : = ‘ : - OS 4 > 77 . Sis E Lia