L... -..J.' V OF ILLh.oiw LIBRARY AT UR3ANA-CHAV.PAIGN N/mr*''^ ------- -M-.,^ / ^FIELDIANA Zoology Published by Field Museum of Natural History Volume 65, No. 8 October 23, 1975 Notes On Rodents of the Genus Gerbillus (Mammalia: Muridae: Gerbillinae) from Morocco Douglas M. Lay Department of anatomy University of North Carouna at Chapel Hill The current systematics of rodents of the genus Gerbillus from Morocco and adjacent eireas of Algeria are based upon the works of Cabrera (1932), Ellerman and Morrison -Scott (1951), and St. Girons and Petter (1965). Much remains to be learned of the distribution, ecology, evolution, and systematics of these rodents in North Africa. Africa. Species of Gerbillus have the plantar surface of the hindfeet either covered with hair or naked and classification reflects this dichotomy at either the generic or subgeneric levels (e.g., Allen, 1939; Ellerman and Morrison -Scott, 1951; Petter, 1959; Wassif et al., 1969; Harrison, 1972; Schlitter and Setzer, 1972). Generally, hairy -footed Gerbillus inhabit areas of sand or light sandy soils and naked -footed Gerbillus occur on compact substrates containing much rock. This study concerns only hairy -footed species of Gerbillus. During 1971 and 1972, I obtained specimens of Gerbillus from several localities in Morocco for cytogenetic study. I visited Morocco in 1973 and collected additional specimens, karyotypes, and data on the environmental and physiographic characteristics of specimen localities. Analysis of this material suggests that at least three medium -sized species of hairy -footed Gerbillus occur in Morocco, two of which are described as new. MATERIALS AND METHODS Chromosomes were analyzed from cell suspensions of femoral bone marrow (Nadler and Lay, 1967). Karyotypes were arranged in Library of Congress Catalog Card No.: 75- 18214 US ISSN 0015-0754 lATHAL KTWT 9M Publication 1213 89 iiFC 2 21975 UBRABir 90 FIELDIANA: ZOOLOGY, VOLUME 65 pairs according to size and centromere position (metacentric, submetacentric, acrocentric). The fundamental number of chromosome arms (FN) is computed by counting biarmed autosomes as two arms and acrocentric autosomes as one arm. Voucher specimens of karyotyped animals and other specimens were prepared (usually as skins and skeletons) and deposited in the collections of the University of Michigan, Museum of Zoology (UMZ) and Field Museum of Natural History (FMNH). Other specimens and cytogenetic preparations remain in my personal collection (DML), but will ultimately be deposited in the FMNH collection. Specimens in the collections of the British Museum (Natural History) (BM) were used in compeuisons. The following cranial dimensions, useful in distinguishing the taxa treated herein, were taken with dial calipers and rounded to the nearest 0.1 mm.: greatest breadth across the posterior mastoid bullae; greatest breadth across squamous portion of temporal bones; the width of the anterior process of the bsisioccipital at its articulation with the sphenoid was recorded indirectly as the least width between the medial walls of the tympjmic bullae at the level of the above articulation; narrowest point of the interorbital constriction; height from base of tympanic bulla to superiormost point of suprameatal triangle (measured with ventral parts of tympanic bullae and incisors placed on glass slide and then subtracting slide thickness); greatest length of skull; length of nasals from distal tip to frontal articulation; greatest breadth between the lateral tips of the zygomatic process of the temporal bones. The anatomical terminology of middle ear structures follows that of Lay (1972). Color references are based on the color guide of Palmer (1962). Gerbillus hoogstraali sp. nov. Hohtype. — Adult female, in fluid, skull removed, FMNH, no. 114647, from 7 km. S. Taroudannt, Morocco, elevation about 116 m.; taken on July 16, 1973 by D. M. Lay. This locahty lies about 1 km. south of the Sous river. Description. — A middle-sized species, mean and extremes of standard external measurements of seven specimens are (in mm.): total length, 208.1 (190-221); tail length, 113.8 (103-123); hind foot length, 27.3 (26-28); pinna length, 16.6 (16-17). Dorsal hair color is intermediate to cinnamon and buffy brown and is clearly delimited from the white ventral fur. A whitish postauricular spot exists, but its prominence is reduced due to a buffy suffusion. The posterior borders of the nasal bones at the articulations with frontal bones are either squared off or concave o c a «D iH rH 73 •a s I 3 1 CO Q OS «0 « CO 2 rH 0) CO 00 ^ CO (N O d d i^; ^ o O 00 CO lO O CD O ; o 00 CO i-H o CO d CO CO 13 ^ O •= c a, ecies differ significantly. G. hoogstraali has a diploid complement (2N) of 72, and FN = 80. The karyotype contains six submetacentric and 64 acrocentric autosomes. The sex chromosomes are a large submetacentric and a medium -sized metacentric (fig. 3). G. occiduus shows a 2N = 40 and FN = 76. The karyotype shows eight metacentric and 30 submetacentric autosomes. The sex chromosomes are medium -sized metacentrics (fig. 4). G. hesperinus Cabrera possesses 2N = 58 and FN = 76. The karyotype reveals 20 submetacentric and 36 acrocentric autosomes. The sex chromosomes are large and medium -sized metacentrics (fig. 5). The diagnostic cranial features of G. hesperinus that dis- tinguish it from both G. hoogstraali and G. occiduus are: less voluminous auditory bullae with smaller tympjuiic and posterior mastoid chambers; short and constricted posterior palatine fora- minae; narrow lateral pterygoid processes, which produce smaller fossae for the medial pterygoid muscles; small posterior inferior mastoid chamber (about one-half as voluminous as in either of the other two species) (figs. 1, 2). The crania of G. hoogstraali and G. occiduus can always be identified by the anterior process of the basioccipital that is broad in the latter and narrow in the former and the least interorbital breadth which varies in the same manner (fig. 1, table 1). This difference in basioccipital width is also constant between subadult It ft* •■ it II II II il II II A» II II II II 1% II II II II 11 41 II II II II II II II »A i« A4 It .. .. J, X Y Fig. 3. Karyotype of G. hoogstraali with 2N »■ 72. It is not possible to distinguish between the X and Y chromosomes until female specimens are studied. li U X^ ly IK XX XV lA IK hh %A "ht X Y XX X X PlC. 4. Karyotype of G. occiduus with 2N - 40. 96 LAY: RODENTS OF GENUS GERBILLUS 97 u u u 1^ u n U %% C^ h< «« Jin if\ HA n hh 4A M III M A/) a U M 19 Af Ad Ift K X Y Fig. 5. Karyotype of G. hesperinus with 2N = 58. It is not possible to distinguish between the X and Y chromosomes until female specimens are studied. specimens of both species. The nasal bones (at the articulation with the frontal bones) are truncated into a posteriorly directed wedge in G. occiduus but are broader and squared off or are concave anteriorly in G. hoogstraali (fig. 1). Beading of the supraoribital ridge is usually more prominent in G. occiduus particularly along the posterior third of the ridge (figs. 1, 2). The external orifices of both hypoglossal canals are readily visible when the inferior surface of the skull is viewed from the anterior in G. occiduus, but the laterally broadened occipital condyles obscure these canals in G. hoogstraali. The portion of the tympanic bullae anterior to the external auditory meatus is more inflated in all dimensions in G. occiduus than in G. hoogstraali (figs. 1, 2). Further, these two species differ significantly in at le£ist five cranial dimensions (table 1). These species differ in a number of pelage features. A narrow but distinct dark eye ring characterizes G. occiduus but is less prominent than that in G. hoogstraali, in which a distinct ring of M FIELDIANA: ZOOLOGY, VOLUME 65 dark hairs surrounds the eye. The eye ring of G. hesperinus is very similar to that of G. occiduus, but lacks the clear white band of hair between the eye and base of pinna that characterizes the latter. The pinnae of G. occiduus show a slight, dark pigmentation around the distal free edge while the distal one-half to two-thirds are darkly pigmented in G. hoogstraali and G. hesperinus. IHnna length of G. occiduus averages shorter than in the other two species. Mean pinna length, range, and sample sizes are: G. hoogstraali (16.3, 14- 17, n « 9); G. hesperinus (16.1, 14-19, n = 12); G. occiduus (13.8, 12-16, n = 8). Prominent black terminal rings on the fresh dorsal pelage of G. hoogstraali and hesperinus give this region a much darker appearance than in G. occiduus. Further, the subterminal rings are more cinnamon colored in the former two species; the latter tends toward buffy brown. Other Specimens Examined. — G. hesperinus, total of 14 from Morocco: 5 km. NE-5 km S Essouira ( = Mogador), UM 122755-57, 22.5.30.24-31, (BM) 7.6.17.3, DML 4582; Cape Tefelney, (BM) 66.6067. Discussion. — The ranges of G. hesperinus, G. hoogstraali, and G. occiduus are allopatric, each is apparently limited to a relatively small area of sand, and all lie near or along the Atlantic coast west of the High Atlas and Antiatlas mountains. The High Atlas extend to the ocean north of Agadir and separate the sandy lowlands inhabited by G. hesperinus and G. hoogstraali. Similarly, an extension of the Antiatlas reaches the sea south of Tiznit forming a barrier between the ranges of G. hoogstraali and G. occiduus. To my knowledge, hairy-footed Gerbillus are known to occur only in and around places of abundant, loose sand. Thus, the availability of loose sand seemingly constitutes an ecological constraint on the distribution of these gerbils and particularly the three above-mentioned species. The intervening mountain ranges are formidable barriers. Not only are they almost exclusively rock, but the higher regions support(ed) extensive forests, which could also form a significant barrier to dispersal. If the distributions of these three species are to be interpreted on the basis of the present geomorphology and ecology of southwestern Morocco, one could conclude that G. hesperinus and G. hoogstraali dispersed across two and one mountain range, LAY: RODENTS OF GENUS GERBILLUS 99 respectively, to reach their present ranges. Another explanation seems more likely. The crustal orogeny that produced the present elevations of the High Atlas and Antiatlas mountains began in the late Miocene and continued into the middle Pliocene (Choubert, 1945). The Sous valley, a broad syncline, was formed during the initial uplift of the High and Antiatlas ranges. Pliocene and Early Pleistocene eustatic sea levels were high (200-300 m. above present MSL), but fluctuated, and most or all of the areas now inhabited by G. hesperinus, G. hoogstraali, and G. occidims were inundated (Choubert, 1950). Sea levels for the period between 1,000,000 and 400,000 BP are insufficiently known, but were considerably higher than at present during most of this time (Lecointre, 1952); but, levels during the period from the latter date to the present are known with considerable accuracy (Fairbridge, 1961; Kassler, 1973). Mean sea level varied between 20 and 100 m. above the present level from 400,000 until about 120,000 BP, the beginning of the Riss-Illinoian glacial. The Riss and Wurm glacial, interstadial and interglacial sea level changes are well documented and averaged much lower than present levels (Fairbridge, 1961; Kassler, 1973). The Riss marine regression exposed a corridor along the present continental shelf that at maximum was 9 - 23 km. wide between Aoreora and Essouira. The corridor was open for about 20,000 years and closed during the Monastrian (Sangamon) interglacial. Bottom profiles and contour maps reveal that the continental shelf off Morocco slopes gently seaward from the coast to the 100 m. contour (Heezen et al., 1959). The Wurm regression reopened this corridor for a period of more than 40,000 years, and maximum exposure and width (12-30 km.) occurred about 20-17,000 BP. By 6000 BP, the Flandrian transgression reached present sea levels (Fairbridge, 1961; Kassler, 1973). On the basis of the known geological record it is unlikely that G. hesperinus or G. hoogstraali could have spread into south- western Morocco during the Pliocene or early Pleisotcene, assuming that either/both existed then. There is a distinct possibiUty that hairy-footed Gerbillus dispersed northward from the Spanish Sahara region during either or both the Riss and Wurm low stands or possibly at some as yet unknown earlier Pleistocene low stand(s). The discovery by Tchemov (1968) of fossils of the extant G. dasyrus in deposits dated as ± 125,0(X) BP suggests that other 100 FIELDIANA: ZOOLOGY. VOLUME 66 modem species such as G. hesperinus and G. hoogstraali could have existed then. In any event, the Flandrian transgression onto the continental shelf seems likely to have isolated the latter two species or their ancestors about 6000 BP. Additional study of these rodents is necessary to clarify the details of their origin and evolution. Fortunately, what is known of their distribution and affinities offers encouraging prospects for estimating rates of differentiation and speciation within Gerbillus. ACKNOWLEDGEMENTS I thank E. T. Hooper, H. Setzer, and L. Robbins for specimens. Alan Feduccia kindly criticized the manuscript. Kirk Agerson rendered valuable technical assistance. Funds for part of this research were provided by William and Janice Street and the Medical Foundation of North Carolina. REFERENCES Allen, G.M. 1939. A checklist of African mammals. Bull. Mus. Comp. Zool., 83, 763 pp. Cabrera, A. 1932. Los mamiferos de Marruecos. Trab. Mus. Nac. Cien. Nat., Ser. Zool. no. 57, 361pp. Choubert, G. 1945. Note preliminaire sur le Pontien au Maroc (Essai de synthese orogenique du Moroc Atlasique). Bull. Soc. Geol. France, ser. 5, 16, pp. 677-764. 1950. La limite du Pliocene et du Quatemaire au Maroc. Proc. Int. Geol. Cong., Part IX, section H, pp. 11-18. Ellerman, J.R. and T.C.S. Morrison-Scott 1951. Checklist of Palaearctic and Indian mammals. Brit. Mus. Nat. Hist., London, 810 pp. Fairbridge, R.W. 1961. Eustatic changes in sea level, pp. 99-185 in Ahrens, L.H., F. Press, K. Rankama, and S.K. Runcorn, eds.. Physics and Chemistry of the Earth, vol. 4, New York. Harrison, D.L. 1972. The mammals of Arabia, vol. 3. E. Benn Ltd. London, pp. 385-670. LAY: RODENTS OF GENUS GERBILLUS 101 Heezen, B.C., M. Tharp, and M. Ewing 1959. The floors of the oceans. Geol. Soc. America, Spec. Paper, 66, 122 pp. Kassler, p. 1973. The structural and geomorphic evolution of the Persian Gulf, pp. 11-32. in Purser, B.H., ed., The Persian Gulf, New York. Lay, D.M. 1972. The anatomy, physiology, functional significance and evolution of specialized hearing organs of Gerbilline rodents. Jour. MorphoL, no. 138, pp. 41-120. Lecointre, G. 1952. Recherches sur le Neogene et le Quatemaire marins de la cote Atlantique du Maroc, Part 1 Stratigraphy. Notes Memoires Service Geol. Maroc, no. 99, 198 pp., 8 pi. Nadler, C.F. and D.M. Lay 1967. Chromosomes of some species of Meriones. Z. Saugetierk, no. 32, pp. 285-291. Palmer, Ralph S. 1962. Handbook of North American Birds, Vol. 1, Yale Univ. Press, New Haven, 567 pp. Petter, F. 1959. Evolution du dessin de la surface d'usure des molaires des gerbillides. Mammalia, no, 23, pp. 304-315. Saint Girons, M. and F. Petter 1965. Les rongeurs du Maroc. Trav. Inst. Sci. Cherifien., Zool. Ser., no. 31, 58 pp. Schlitter, D.A. and H.W. Setzer 1972. A new species of short-tailed gerbil (Dipodillus) from Morocco (Mammalia: Cricetidae: Gerbillinae). Proc. Biol. Soc. Washington, no. 84, pp. 385-392. TCHERNOV, E. 1968. Succession of rodent faunas during the upper Pleistocene of Israel. Berlin. Wassif, K., R. G. Lutfy, and S. Wassif 1969. Morphological, cytological and taxonomical studies of the rodent genera Gerbilbis and Dipodillus from Egypt. Proc. Egypt Acad. Sci., no. 22, pp. 77-97.