Table Of Content

Atti Soc. it. Sci. nat. Museo civ. Stor. nat. Milano, 148 (I): 135-147, Gennaio 2007 Dario Cesarini Observations on thè Chiroptera of Socotra Island, Republic of Yemen Abstract - Few and limited studies have been conducted previously on thè island of Socotra concerning its bat population. During thè period August-September 2003, more than 20 days/nights of surveys were spent in thè field in order to locate and identify occurring bats. Thanks to these surveys three species of bats have been identified by roost searching and mistnetting: Rhinopoma hardwickii, Aselìia tridens and Rhinolophus clivosus (its first ever recorded sighting in Socotra). Key words: bats, Socotra, Yemen, Rhinopoma hardwickii, Asellia tridens, Rhinolophus clivosus. Riassunto - Osservazioni sulla chirotterofauna dell’Isola di Socotra (Repubblica dello Yemen). Poche e limitate indagini di campo sono state condotte precedentemente sulla chirotterofauna dell’Isola di Socotra. Nel periodo agosto-settembre 2003 sono stati condotti circa 20 giorni/notti di missione sul campo allo scopo di individuare i rifugi e definire la composizione in specie della chirot¬ terofauna dell’isola. Grazie a tali indagini, tre specie sono state identificate mediante ispezione dei rifugi e l’impiego di mist-net'. Rhinopoma hardwickii, Asellia tridens e Rhinolophus clivosus. Per quest’ultima specie si tratta della prima segnalazione per Socotra. Parole chiave: chirotteri, Socotra, Yemen, Rhinopoma hardwickii, Asellia tridens, Rhinolophus clivosus. Introduction The Socotra Archipelago, politically administered as a part of thè Republic of Yemen, is situated in thè north-western section of thè Indian Ocean and comprises four islands: Socotra, Adb-el-kuri, Samha and Darsa. Socotra itself is about 120 by 40 km and covers an area of 3,625 km2. The island is about 240 km from thè near- est mainland (Somalia). Socotra is composed of a basement complex of igneous and metamorphic rocks of Pre-Cambrian age overlaid by sedimentary rocks, mainly limestone and sandstone. According to Wranik (1998) topographically thè island can be divided into three main zones: Projecto “Areas Protegidas”, Delegando do Ministerio do Ambiente e Agricultura, Assomada (Santa Catarina), Repùblica de Cabo Verde, e-mail: [email protected] 136 DARIO CESARINI 1. The costai plains, varying considerably in width. 2. A limestone plateau, extending over most of thè island and averaging 300 to 700 m in altitude. 3. The Hagger Mountains in thè centre, rising up to a height of 1,519 m. The climate is monsoonal. There are no exact figures available on thè annual temperatures and rainfall. As far as is known, thè mean annual temperatures vary between 28°C and 37°C, and thè annual rainfall averages about 130-170 mm. The island is sparsely vegetated and dominated by xeromorphic forms. The vegetation is more luxuriant only in sheltered valleys and higher mountain areas (Mies & Zimmer, 1993; Davis et al., 1994; Mies et al. 1995; Alexander & Miller, 1996). The composition and status of thè Chiroptera in thè Socotra Archipelago is unclear. The information available is limited, and thè data was not collected to a preconceived pian. De Winton et al. (1903) recorded a specimen from Socotra Island for thè first time as Rhinopoma microphyllum, although it was later correct- ly identified as R. hardwickii by Kock (1969). Wranik (1986) reports thè capture of two Asellia tridens near Hadibo, and three male specimens of Rhinopoma hard¬ wickii in thè Diksam plateau. Wranik et al. (1991) report thè morphometrical measurements of two Asellia tridens captured near Hadibo (probably thè same two specimens of thè previously cited work). Guichard (1992) captured two bats: one specimen of Asellia tridens and a subadult of “what was probably” Pipistrelliis bodenheimeri, in thè wadi palm groves of Suq, near Hadibo. The presence of Pipistrellus bodenheimeri in Socotra seems to be confirmed by Menu (1987). The final report of thè zoological multi-disciplinary expedition in 1999 to thè Socotra Archipelago (Evans, 2000) strongly recommended a research effort in order to analyse thè species structure, distribution and status of thè locai Chiroptera. The present study is conceived in order to contribute to this proposed research effort (Cesarini, 2003). We tried for thè first time to realize an extensive monitoring of thè bats present in thè main island of thè archipelago. It is possible to summarise thè objectives of thè present study as follows: • Locate new bat roosts and colonies. • Verify thè species composition and distribution of Socotran bat fauna. • Collect data on population composition, morphological and ecological characteristics of thè recorded species. Materials and methods Preliminarily, in order to obtain useful information for thè field phase, we col¬ lected 35 standardized interviews (28 extension officers of thè Socotra Conservation and Development Program, and 7 community representatives). We visited 23 roosts, caves and other sites, from August 19 to September 17, 2003. Where we found bats, we attempted a sampling capture by mist-net outside thè roost; when this was impossible due to locai conditions, thè capture was done using a hand-net. For each captured individuai we recorded thè following information: - Sex and reproductive status. OBSERVATIONS ON THE CHIROPTERA OF SOCOTRA ISLAND, REPUBLIC OF YEMEN 137 - Age (young, adult). - Morphological measurements (forearm, 5* finger, body length, tail). - Presence of extemal parasites. - Individuai notes. The captured bats were released as soon as possible at thè end of morphologi¬ cal analysis. For species identification we followed Harrison & Bates (1991). For each site we collected thè following information: - Total number of bats counted or estimated (minimum and maximum number). - Temperature and humidity (using a digitai device). - Bat-related signs. - Human-related signs. We also took photographs from most of thè visited sites and captured chiropter- ans; they are available in thè Socotra National Museum’s Photo Library. Results Site surveys From 17 sites at which bats were found, thè sampling capture was successful at 16, where thè number of bats collected ranged from one to nine. Tab. 1 provides information on each site where we found bats, and Fig. 1 shows thè distribution of thè surveyed sites in thè island. Fig. 1 - Bat roosts and sites monitored in Socotra Island (Roost e altri siti cesiti nell’isola di Socotra). ¥ Rhinopoma hardwickii; O Asellia tridens; A Rhinopoma hardwickii and Asellia tridens; f Rhinolophus clivosus. DARIO CESARINI 138 1 1 X P N .2 O5O—ih ObX ) p3 • 3©-O2 CO*—Dh c 0C1/O05 ^2. pS- d^0 °C<5 O2_^ § ^ CP(N-3 *roP Ocdh SbXfk) td • ^2 cd 0 b.2b -d -d ¥d GIOO, 2<^ì mIO, ^4^ 0-d- J<Cpì K d ^05 -0P) —P >s3: CO • b^b <>d a gOh 3 3 3 C_NJ PiOr", C_N PI0O" CdN P'Or"‘ (qNq «»0oP, d 50o iIll ^C/5 <>D * Hh* OoCrt C<D3 1iO3ch X pq X pq X pq X pq X pq o a-> >. V. C/5 122 „,. -h—> 'O *P au £2 J2 ^ cd <D HH S cd C P P C/5 <D c 0 -•5oS! •P —p3 £ £5—1 ^05 5^<o-D1. „£ “ 2 —O =s3s pua»/ X X X cu <D « | s O è o H—' ^d3 <D CD <D -2 e o u Oh T3 V. X) P P X» CO '—2 cNd ^ ^ po ~^ £ X» 'o ^ c Q fi T3 2 -O Oh a» 3 X X X X s* DC/5 C/ì “ .2C/5; <d O 0 3 o P O C/5 ^ Oh ^ -3 o ^ca C/5 -p au d cd C) 4 <D •Kh- m 4—* ri Oh • • hH Uì •g S §2 \0 xp \0 \p \© ^ bb ^ 2 S O -« oO -^è pcr fQfbi c0rN- 030- rc/j~-l £rr-- "cd 00 iXo rr-- Il -0 ~ sc« 2 a <D 2 cd - P bX) § S-H p "« • ^ i • cu H O ^H U**=i; oCd £db CD O£2 cOd Ì*232 • mhH d ^ m 0 d 05 ’d- rf- 2 db c O hF b I b S W mm Mm0 dco md d CXN md rrno ^ o •—1 <-> ^ cd io o Eh w "rt 5—1 c/o O 2 CD P CD -O p ~ Il oU £ B 8 ~d § ^ u 0 ’aOC 4<§2D ^¥§>4 -Ô2h Cdcj oCuD. .b0Cb/5 *_HP<—oD> nP S033/ ¥3- m OPOm1f c_100o35 O<u1—Nn1 1 0’Omd1 - 00X0IO1 , •tb ^ <D 2<D ^r- _2 Oh P 1 4Ì o|p _T p o <D 13 0 b o 0 Si- o ._ O <d o. P\ p db . -H Sp-H Pcd cd =b £ 2 •- (UDh -r>< td £ 6p Hg-H ÒX) (D *22 (D <D tob .-g c £ £ -P bP 0+ Oh Pb P CD £ -C •C 03 O io H—C "P CD • — o ’P3 P a> c^ +-> <d hC-/h5 T_*j rO“ Tn 'P p <D nn Pob ”d- 4P3 cPd h«2 ^<2D OC/5 SP) a5- pbJX c«-1 05 XX) C<D/5 Q.A=- c/5 O p 2 v 8 ». ax 0 g.: £ c 0+ 0+ 0+ 0+ 0+ ’-a H—> S-i O J1 >r, i-H H-H <& cd .—1 -4—> OC/5h ^ O CD .2 ; i2 4= • 0 <0r ' ’d•N ’d d 1 'O ” N P =c i—4 y—4 ro (N P2 O-X ^^ s—^ s—' I ocd CD .c2 -4pP iC2/5 1¥’¥ è ,c (N 0 CN 3" m <D cd TP3 5O-h hC(PD/5 ‘4poP .dâb .2 c Coc/d5 »o5 m r^ì m ro m m m CO c2Pc/) lo T• HC^—DHD> o<D 2>05 .<22ODh .*£rC C£oD ^“£ Q¥Hass/- 00<N 00CONO 00(ONO C00ONO 00CONO 00CONO 00CONO 00<0N0 dC>/5 0oo3 _CP/5 •H H0h—3h. So sc ou0b>) op5- 05 OS dCN <dN iCoN ICON rCN- rCN- bo T3 1, cd (D P P rnn P * HH» ^ ì o .rv5 3 C/5 h^3- OrOh O ?3 CU ? <U CD CD PP cd ._s 5-i h3, C/5 . Sì >3 P <P 3 o*v O "_PcCO/d5 -* soK: POcohd CaUOC>//h55 j<>C<d50 sop <£ppD .2Ncd oPcd cCc/3d5 -u3<C3U/5 X<<D5 3O3 gdcd --~'O0o3o X0<5c-Dd( 4P<Uc<3uDDdh Coo/5 Cp JU(PDh }o_H OCb 0S5 pp- aO r-p Q(U 4c<o3U 3COU stc/d5 H3C(UHU s(U Q<u O o_b• c!•d >0- ¥+£(pp-D» C2So/5 ^¥ . ô •o£ 2-Cpo3P ^(8£\j 4/ *BwW34^ 0c0ì<>d3uD 4Vc(U<^&Ddhì 3<£<<Uuu 0aDtc/Dd" 3CE(OUUJ -X(aC3U)/jl O(c33U-dI <c33du ¥Q S* P o J ^ • bhb cd ^ O cu <-! -a3 (U -3 u cd cd ^ (U 0 C/5 bÓ Q G0 O O Q a a -^ £ P r2 p ip H Ì3 .5 CD H cd -o N "P OBSERVATIONS ON THE CHIROPTERA OF SOCOTRA ISLAND, REPUBLIC OF YEMEN 139 os <n 2! C4 “ oro\ £r OrCS ôc NO isnO g cc ca/3. ..te2: ¥S?+. <OCn c©rloot °°S2O. Slo R CIO p^ OmoòS s£g4 So J iXn ô^\ lîòn g O g cX4 ^m XCN în C_Nq Tin© X m (XN *n <XN ^»n <XN în CXN rm^" CXN in a. o o o O o o o o o Z pq Z pq Z pq Z pq Z pq Z pq Z pq Z pq Z pq s © © = ’S J3 U X = ± o c (j VI in, c - 5 aj © X X SI 5/5 3 2 \p X \C £ £ -d o oc SO r- r- OS in o OS so r- SO so sO OC r- m — d. ^ Os os o OS ON ¥d E ^ © rn oc OS in OS c cn cn CN CN CN (N CN CN fS®' X05 —uZ<=J ©in OOsO oc nCN1 oCinN cn oicn ooo1 CooN co © in © o o CN o (N c g sO E © in C • *0 cc CH- ^4 CN CH- *o m m ^ s— ^^ ^^ CH CH- CH- CH- CH- CH- CN cn CN CN «V rv CN CH- *o ’SO *0 m *o C« _ CN m <N NT '-^ m m n in N" in ro m m m m m cn mo con QJ OO © ©o oo o© oo o o 05 CN CN CN CN CN CN CN CN CN a OC OO O O OS ds Os OS OS r- OS m m m 00 OO un CN <N „-s © Cc/33 03 — X03 Xc/3 4<3u V<D- J <©X © Q 'oc 0000 40=0 ©© 1 Xx03 400303 £g <>D 43 X03 X0CO0 'c© 4_o3p ac GO 43 X5 UcU XC/3 £ C3 03 4S=- 4o43 0CS0 4- ©--3- 4o0 ocn ^Cj © OS 40=3 X) o> So- © 03 Co/3 4(0U C/3 w X 2 £ X O X -3 X C/3 140 DARJO CESARINI We also visited three sites at which we did not find any bats, although they were previously indicated as bat roost: - A small cave on thè border of Wadi Ghelemeneten (translation: Bat’s River) in Hadibo previously hosted a colony of bats, probably Rhinopomatids consider- ing thè characteristics of thè cave. We did not find evidence suggesting thè recent presence of bats. - A cave on thè hill dose to thè port of Howlaf has also hosted a colony in thè past (Ahmed Saeid Suleiman pers. com.), again probably Rhinopomatids con- sidering thè characteristics of thè cave. We found just an old trace of a previous colony on thè ceiling. - In a small cave on thè border of Qaryha Lagoon we found signs of a quite recent bat presence on thè ceiling, and guano on thè floor. It had been quite a large colony and moved from thè site no more than a few months before thè survey. Species account Family Rhinopomatidae Rhinopoma hardwickii Gray, 1831 In total we visited 13 sites used by Rhinopoma hardwickii. At least 8 of them can be considered most likely to be reproductive colonies. The minimum numbers of R. hardwickii for each site varied from 3 to 100 ca. (Gorf Cave). Only once in Gorf Cave (Noged) did we find R. hardwickii using thè same site with another species (Asellia tridens). R. hardwickii demonstrates an impressive adaptability in selecting roost site. We found R. hardwickii in large caves (De Qaseqas), tunnel systems (Wadi Zaerig), small caves, crevices between large rocks, and man-made buildings. Even thè conditions inside thè roosts showed a fair variability. The temperatures ranged from 25.1 to 36.6 °C, and thè range of relative humidity varied from 48 to 77%. We found colonies in almost every monitored area of thè island, and in all thè main different environments. A colony in thè Hagger Mountains was approximate- ly at 1,000 m a.s.l.. However, thè roosts appear more widespread at lower altitudes. Five sites were used by humans as Stores or goat enclosures (Fig. 2), and in three of these people even made fires inside thè site, in order to sterilize thè milk of their goats. In two sites thè guano was collected by people to be used as fertilizer for their vegetable gardens. We captured 46 (23 S, 23 S ) R. hardwickii, at least one for each monitored site. The results of morphometrical measurements are summarised up in Tab. 2. Family Hipposideridae Asellia tridens (E. Geoffroy, 1813) Collectively, we visited 3 sites used by Asellia tridens. All of them can be con¬ sidered most likely to be reproductive colonies. The roost in Gorf Cave is a mixed colony with R. hardwickii. The colonies of Rhahr and De Fareho were composed of at least 500 individuals. The entrance of thè roost hall is always less then 1-1.5 m2, and inside it is com- OBSERVATIONS ON THE CHIROPTERA OF SOCOTRA ISLAND, REPUBLIC OF YEMEN 141 Fig. 2 - The cave of Ghelemeneten (Abel). The cave is used by locai people as goat enclosure. Fig. 2 - La cavità naturale di Ghelemeneten in località Abel. La cavità viene impiegata dalla popola¬ zione locale come ovile. pletely dark with no filtering light. The range of temperatures varies between 29.0 and 33.4 °C, and thè range of relative humidity varied from 67 to 78%. The caves used by A. tridens were not-man made; however, they were not underground wadis. All three colonies found are located below 80 m a.s.l. and placed on hills not far (maximum 600 m) from semi-arid woodland plains. In thè sites of Rhahr and De Fareho thè guano is collected by thè people for their vegetable gardens or for profit. We captured a total of 18 (1S, WS) A. tridens, while sampling every site. One female was captured in Chesa (Noged) by mistnetting in open space; this specimen is not included in Tab. 1, but its morphometrical measurements are summarized in Tab. 2. DARIO CESARINI 142 Family Rhinolophidae Rhinolophus clivosus Cretzschmar, 1828 We recorded this species once, a solitary male, in Hafur Cave, Diksam plateau (Fig. 3). In its noseleaf, thè superior connecting process of thè sella is elevated into a blunt point, when viewed laterally (Fig. 3). It corresponds closely to thè illustra- tion of thè sella of R. clivosus from Yemen in Harrison & Bates (1991, Fig. 59b). We also observed two flying bats in thè huge cave, but thè identification of thè species was impossible. The morphometrical measurements of this specimen are shown in Tab. 2. Fig. 3 - Lateral view of thè sella in Rhinolophus clivosus. Fig. 3 - Profilo della sella di Rhinolophus clivosus. OBSERVATIONS ON THE CHIROPTERA OF SOCOTRA ISLAND, REPUBLIC OF YEMEN 143 G G CS co NO ON o m NO o CO VO m U^ì t/o m) I/O i/o E ^ S c/2 j-i £ - s o co 7- or o CO I/O 7- 3 rS GL 75 E in uo 7f or in or "d- Ti" o <+h —13 NN •> Uh £ a 74 re co in 74 — 7- 04 -G C/2 CO -4 74 —< —1 04 -4 G o Sh ^ O toh ^_y ^_s ^_s ^_. , CD -nf->i in 7ONf oo 74 7f ON in 00 Ô Wfi ) O ^ eII _IcI _ eli CII CII y II 7II NO 7II 7II —r G <—l - ìi m uo O o 7- co «c CXQ ri-ni oynr >inn soor ouor in Oin r-‘ om (D £-T3 t£u£ 4C—d» <D cXs oo 74 CO 74 7- 74 ON co oo Goj rG. - -4p—> G E r- NO 7~ 74 04 04 CO CO G X-? h-< c£ e o\ CO NO t NO On NO ON G *+H ce E >n 7f 74 o <D a oc — or in yri o NO m C/2 ’c C/2 \Ó or in — -4 CO CO CO 4—> o G3 H ^^ GGcGS-/H2 htGCoJDh ’5GGn—i (GD 73 CO-—eIOtI "4 iInI ooIeIoc 0¥'aeII0* ' ' 0_ICI4 77I0I ^ NO7IOOI >04n i>7In—I 7 7^7I—~I -7 CD C/2 5-2 ON oo ON m cn 04 OO CO J* ^ GZCu "rGf-Hi NNOT ri-n’ 7N4O ON NO o04 0044 07~4 <CNO 1 • «n • —G ->o? CD .Hfi â X«3 oc 24 44 or m e O co yn 73 O "g *G CO E 1 1 7~ NO 7- ON oo ON "G g G *3 G 3 T3 < G e OO NO 7f 74 tT t— ON NO NO NO NO t— 7- G G .2 E 4 HGo . T§3 GG oc/2 GN OQ CQ/2 4c>t 7COf oro- N7O4 O7Of OyrOì NCOO u7fo' C/2 <75 > t4=a n—v ^—V G<GD <-G<+Dh ¥GG ’GsGcu QG meII 4eIIt i0cInI0 _ TOeO 7I4I 7I0I 7ii «n o¥7 ¥t—7 •gS3o .-t°soh OGOrv gG> Gos• Oir s o<Nc O>Nn O7N4 IoTT)' N>7O4n CcrOo- O7t ~N Oei—nN to7T~o- CX _o gj 3 5-i ’g cd J-4G£Ig£CGf=-i- ’.1CX—*5<Dc§QD*hIb ¥+oGGCGGGG5--/>42 -T4TOGOGCGG—3/32» XTGGGG§G3I Locality è-4He<g0ooo—D3» eSouth Yeme .^^B <C<£CiO03eDU)- O*'<• 7sCHKC—§33 è4HC<gcooo—1d»0 £4e<£eUooo—duh» 7South Yeme n<DoHecs&cUHddih x5<gc270scU>»udd2h £4<cUo7£oo—duh» '<TcUwestern Sau< dhi and Israel ''C<D-S2U7C7u0S/322h3. and Yemen -O G C/2 K3 73 C/2 C/2 73 C/2 I o^2 C17/23 £G GG NGN Ûh C£^/2 (cUDdh C3c/d2 h4Uo—Ch» Co/2 Z^3£ w G 2 £ « -fi G <D CD W) C/5 in C/5 m r*fi =733 _G ^ S <D 4(—D* 02 02 'vh — cd cd cd hj _ m . ^ C3 23 CQ . ^ OQ OD G -4—» _g o NO (D .s NO Cu CU NO -Gfi ££ <^$ j3= ’ciCQn-3> ooCnI e OONN ^Co OO^Ns ’e(CDd oo74 M7 OONN K7S ^G-4Ss 7O OONN ' Cc0d2 oo04 7O GO^N' 7O OO—NN 7G5 !o—i GG TGO3 < u Ct-3 UCU0/3hh5 o eUdh £cUdh CUUc/dhh5 U mUcdh mcUdh • 5 G G E S se se £ G G <75-51 G "GO 4G-4 _G— «a<5 .523 _o ’g GG T3 5eCi a- _<GD 4G—> "O 4GO—>t GNn t*"“ t’*" Di Oau > -4S DC 523 523 623 G Cj Di SJ 5 *»• a G 0 =*» * «=» >5 >5 V.; Q O o "f> s s © e a G3 73 c^ G fi "5S ¥«a “js -e; ¥4Co—D> 5G-i GO -S5!ì 5! <5 5CJ V«ai *<•>«*3 t•<Haa 5>5J O523j a523 -GG TG3 u .G2 ^H 5««S» « «*¥5H«1!è, I¥Ke» 7"2a u. ?-22» >C5i . -aea. -aaa . C G vi © © ? a^ O in «n ’g o .25 §* !" I- •2 •2 •2 "a oc</D2 AU 4G= 4VG32 COD CQa/2J *> “s»»: °=«Sscwì 0a s 2>u3 oVi o«t/u3 O5/5 >e5c«1 es -caae C/5 H +-» CJ 144 DARIO CESARINI Discussion We monitored three species in thè island: Rhinopoma hardwickii, Asellia tridens and Rhinolophus clivosus. From a global conservationist point of view these species are not considered as endangered and their distribution is widespread. The Socotran populations of R. hardwickii and A. tridens seem to be in good conservation status, both for consistency and distribution. R. clivosus is a typical African bat widespread on thè continent, but also present in thè western part of thè Arabian Peninsula (Harrison & Bates, 1991). There are several records for this species in Yemen (Hayman & Harrison, 1950; Sanborn & Hoogstraal, 1953; Harrison, 1964; Kock & Nader, 1979; Nader, 1982; Kock et al, 2002), but our record seems to be thè first one for this species in Socotra (Al-Jumaily, 1998). Probably this and another record from Oman (Snowden et al., 2000), represent thè eastern limit of its distribution. For this reason thè presence of R. clivosus in thè archipelago could be unstable or weak. Concerning Rhinopoma hardwickii we verified a high adaptability in thè roost site selection. Previous observation of this species suggested this and, in addition, thè inclination to build colonies of smaller size compared with other species (Nader, 1975; Atallah, 1977). Its proximity to humans is particularly interesting. In many areas of thè island people habitually collect guano from thè colonies for per¬ sonal use or it is sold for use as a powerful fertilizer in vegetable gardens. Moreover, people of Socotra frequently use naturai caves or sites for their own activities, located even in thè most inhospitable areas. Such confidence with humans shown by R. hardwickii is probably one of thè main reasons for its success in thè colonization of thè island. The morphometrical analyses of thè Socotran population of R. hardwickii do not show any particular inconsistencies with thè results obtained by Harrison & Bates (1991) for a larger and widespread population in Arabia (Tab. 2.). Otherwise, thè Socotran population of A. tridens seems to have morphological measurements (forearm, body length and tail) on average shorter than thè specimens of thè species mentioned by Harrison & Bates (1991), but nevertheless almost within thè species range. Such a result, confirms that this species shows evi- dent variations between different populations (Gaisler et al., 1972; Wranik et al., 1991). A morphometrical analysis conducted on a limited sample of specimens from Socotra and South Yemen (Wranik et al., 1991) showed that thè two speci¬ mens from Socotra Island were smaller than those from thè mainland. Our study seems to confirm these results. The 18 individuate captured by us show forearm, body length and tail on average smaller than thè Arabian widespread population studied by Harrison & Bates (1991) and thè specimens from southern Yemen (Wranik et al., 1991). The partial isolation of thè bats on thè island could establish thè conditions for a genetic segregation of their populations. Furthermore, interesting evidence was obtained on thè roost site selection in A. tridens. The data clearly indicates an ecological set of criteria adopted by this species. Our findings seem to confirm previous studies involving A. tridens that indicated dry caves were often occupied by hundreds bats as thè most common roost type for this species (DeBlase, 1980; Harrison & Bates, 1991). Very little ecological and morphological hypotheses can be put forward regard- ing Rhinolophus clivosus with our limited sample. However that we found thè bat