Journal of Vertebrate Paleontology 21(3):547-554, September 2001 O 2001 by the Society of Vertebrate Paleontology
ARCHAEODESMAiVA BAETICA, SP. NOV. (MAMMALIA, INSECTIVORA, TALPIDAE) F R ~ M THE MIO-PLIOCENE TRANSITION OF THE GRANADA BASIN, SOUTHERN SPAIN ELVIRA MART~N-SUAREZ', NICOLAS BENDALA', and MATTHIJS FREUDENTHAL'.2 'Departamento de Estratigrafia y Paleontologia, Universidad de Granada, 18071 Granada, Spain; e-mail
[email protected]; 2Nationaal Natuurhistorisch Museum, PO. Box 9517, 2300 RA Leiden, The Netherlands
ABSTRACT-Archaeodesmana baetica, sp. nov. is a new water-mole (Desmaninae, Talpidae) from Purcal4, a locality at the Miocene-Pliocene transition in southern Spain. It is characterized by a very large p2, larger than P2, and a p2/ p3 ratio which is larger than in any known desmanine. Its I1 are bilobed. The premolars are very large in comparison with the molars, or, in other words, the mandibles and maxillae of this new species are greatly enlarged anteriorly. The new species is possibly close to the ancestry of Desmana.
INTRODUCTION The Granada Basin is situated in the central sector of the Betic Cordillera in southern Spain. Its Neogene and Quaternary sediments are of considerable thickness, and cover the contact between the internal and external zones of the Cordillera (Rodriguez-Femhdez, 1982). The latest Miocene and Pliocene continental sediments belong to the Alhambra Formation, a prograding sequence of intensely red alluvial fans, which contain well-rounded, reworked pebbles from the internal zones of the Betic Cordillera, more specifically from Sierra Nevada (Fernkndez et al., 1996). The Purcal (PUR) section is situated in the outer fans of the Alhambra Formation, about 5 krn north of the city of Granada; it is delimited both at its base and at its top by faults, its total
a
formed in a humid climate. An increase of aridity is observed towards the top of the Purcal section. It contains number of superposed fossil mammal localities, the richest of which is PUR-4. Lithologically this locality is a carbonaceous, grayish green, fluvio-lacustrine lutite. The PUR-4 fabnal assemblage includes species typical for the latest Miocene or the earliest Pliocene (Table I), so its age is thought to be close to the Mio-Pliocene transition. This is confirmed by paleomagnetic data, which place the locality at the end of Chron C3r (following Berggren et al., 1995), which correlates with the Mio-Pliocene boundary (Martin Sufirez et al., 1998). Modem desmanines are semi-aquatic and always associated with permanent streams (Grass&,1967). The high relative abundance of Archaeodesmana (18% of all rodents plus insectivores) at PUR-4, along with the abundance of Muscardinus vireti and the presence of two sciurines indicate a wooded biotope. Both the faunal assemblage of PUR-4 and the sedimentological data (Bendala, 1999) indicate humid conditions. t he most abundant European fossil water-moles belong to three genera. We accept the nomenclatural revisions of Hutterer (1995), who recognized Archaeodesmana Topachevskiy and PasNcov, 1983 (Dibolia Riimke, 1985 is a junior synonym), Galemys Kaup, 1829, and Desmana Giildenstaedt, 1777. Topachevskiy and Pashkov (1983) subdivided the genus Desmana into five subgenera, Desmana, Praedesmana, Pliodesmana, Archaeodesmana and Galemodesmana, on the basis of the number of roots in P1 and p l , Pl/C and P2/C indices, relative size of
p2 and p4, and absolute size (Topachevskiy and Pashkov, 1990; Pashkov and Topachevskiy, 1990). Hutterer (1995) elevated Archaeodesmana to full generic rank, and included nine species: Desmana pontica Schreuder, 1940 (type species), Desmana vinea Storch, 1978, Dibolia turolensis Riirnke, 1985, Dibolia adroveri Riimke, 1985, Dibolia luteyni Riimke, 1985, Dibolia major Riimke, 1985, Dibolia dekkersi Riimke, 1985, Dibolia brailloni Riimke, 1985 and Desmana bijida Engesser, 1980. Dibolia Riimke, 1985 and Ruemkelia Rzebik-Kowalska and Pawlowski, 1994 are junior synonyms of Archaeodesmana. For reasons discussed below we also include Desmana verestchagini Topachevskiy, 1963 in Archaeodesmana. In this paper we describe a new species of Archaeodesmana, A. baetica, sp. nov. from PUR-4. It is characterized by having a very large p2, comparable in size to the p2 of the species of Desmana, but the rest of the dentition is similar in size to that of other species in Archaeodesmana.
Dental Terminology and Measuring Methods The nomenclature of crests and cusps follows Riimke (1985). Teeth were measured with a Wild M7S binocular microscope, provided with a Sony Magnescale LM12 digital measuring device. Measurements were only taken on canines, premolars, and molars, because data on incisors are rarely reported in the literature. Orientation is one of the major problems in accurately measuring the teeth of small mammals. No objective orientation criteria for measuring desmanine canines and premolars have previously been published, as far as we know, so we propose the following standards. For each element a reference line is defined along which the specimen should be oriented. Length and width are measured as the smallest circumscribed rectangle with sides parallel and perpendicular to the reference line (Fig. 1). Reference lines include: (1) c, p l , and p3-the reference line follows the anterocristid; (2) p2-the.reference line is the maximum length of the tooth, a line that runs through the apex of the protoconid in unworn specimens; (3) p L t h e reference line is the longitudinal axis of the tooth, through the paraconid, and subparallel to the crests and to the labial border of the tooth; (4) m l , m2, and m3-the reference line runs through the parastylid and entostylid. In the lower molars only length was measured, because width depends strongly on the degree of wear, and seems not to be a reliable parameter for making comparisons; (5) C-the reference line coincides with the maximum
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 21, NO. 3,200I
ml
m2
m3
FIGURE 1. Measunng method for desmanine teeth. The reference lines are indicated by dots.
length, and runs through the small metastyle; (6) PI-the reference line passes through the concavity of the posterior border of the tooth and the lingual end of the anterior cingulum; (7) P2 and P3-the reference line is the axis of maximum length of the tooth. In unworn P3, it runs through the top of the paracone; (8) P L t h e reference line runs through the metastyle and is subparallel to the posterocrista; (9) MI-the reference line runs through parastyle and metastyle; (10) M2-the reference line runs through the parastyle and the mesostyles; and (11) M3-the reference line is the tangent to the anterior border of the tooth. SYSTEMATIC PALEONTOLOGY Family TALPIDAE Fischer, 1817 Mivart, 1871 Subfamily DESMANINAE Genus ARCHAEODESMANA Topachevskiy and Pashkov, 1983 sp. nov. ARCHAEODESMANA BAETICA, (Fig. 2) Holotype--Fragment of a left dentary, with c, p l and p2, PUR-4 968 (Fig. 2C), deposited in the 'Departarnento de Estratigrafia y Paleontologia' of the University of Granada (Spain). Measurements of the holotype: c, 1.14 X 0.97; p l , 1.02 X 0.98; p2, 1.96 X 1.40. TABLE 1. Mammalian faunal list of Purcal 4, Granada Basin, Spain (latest Miocene-earliest Pliocene). Paraethomys meini (Michaux, 1969) Stephanomys dubari Aguilar et al., 1991 Apodemus gorafensis Ruiz Bustos et al., 1984 Occitanomys alcalai Adrover, Mein, and Moissenet, 1991 Apocricetus barrierei (Mein and Michaux, 1970) Blancomys sanzi Adrover, Mein and Moissenet, 1995 Ruscinomys lasallei Adrover, 1969 Muscardinus vireti Hugueney and Mein, 1965 Eliomys truci Mein and Michaux, 1970 Pliopetaurista pliocaenica (Depkret, 1897) Atlantoxerus sp. Sciurini indet. Prolagus michauxi L6pez Martinez, 1977 Galerix iberica Mein and Martin Sukez, 1993 Archaeodesmana baetica sp. nov.
Referred Specimens-6 i l (PUR-4 948-953), 4 i2 (PUR-4 954-957), 6 i3 (PUR-4 958-963), 5 c (PUR-4 981, 982, 984986), 4 p l (PUR-4 964-967), 9 p2 (PUR-4 669-677), 3 p3 (PUR-4 1023-1025), 4 p4 (PUR-4 620-623), 6 m l (PUR-4 626-631), 13 m2 (PUR-4 632-644), 12 m3 (PUR-4 645-656), 8 I1 (PUR-4 987-994), 9 I2 (PUR-4 995-1003), 5 I3 (PUR-4 1004-1008), 4 C (PUR-4 1019-1022), 10 P1 (PUR-4 10091018), 4 P2 (PUR-4 588-591), 4 P3 (PUR-4 978-98,0), 2 P4 (PUR-4 592-593), 6 M 1 (PUR-4 594-599), 6 M2 (PUR-4 600605), 10 M3 (PUR-4 606-615) and 2 humerus fragments (PUR4 1026, 1027). Etymology-After the Latin name for the southern part of Spain. Type Locality-Purcal 4 (Mio-Pliocene transition, Granada Basin, Spain). Co-ordinates UTM 30SVG455218 (Castillo et al., 1990). Diagnosis-Archaeodesmana similar in size to the extant Galemys pyrenaicus, with a short, straight, and bilobed 11; P2 longer than P3; p2 longer than p3; p2 very large, with a p2/p3 ratio of approximately 1.4, larger than in any other known species of Desmaninae. The C is similar in size to the P2 (P2/C = I), whereas in the other species of Archaeodesmana the C is smaller than the P2. Measurements-See Table 2 and Figures 5, 6.
~
Description
1
il-The crown is high, shovel-shaped, and constricted at its base. The labial surface is flat. There is a crest on the lingual face, that becomes more pronounced towards the apex (Fig. 2A). The base of the crown (protruding lingually) and the root (long and straight) have triangular cross-sections. This element corresponds to morphotype "a" of Riimke (1985). i2-The crown is high, subtriangular in cross-section at its base, and compressed at the apex (Fig. 2B). The antekor face is flattened and towards the base there is an enamel-free area. Except for this anterior face, the base of the crown is surrounded by a well-marked cingulum, that forms a small posterior cuspid. The root is straight and long, with a circular crosssection. i3-The crown is low and the tooth is very similar to the canine. The outline is rhomboidal, with rounded anterior and
-
[IIIIIIIIDJIIIIIIID Galemys
A. baotica Archaeodesmana
FIGURE 5.
Ranges for length and width of the lower teeth of Desmana, Galemys, Archaeodesmana baetica, and other Archaeodesmana.
border rounded. The lingual face is practically flat and the labial one is convex (Fig. 21). The anterocristid exists only in the lower part of the crown, and it is curved to the anterolingual border of the tooth, where it meets a very small paraconid, at the base of the protoconid. The posterocristid is faintly marked. The talonid is very narrow, closed by a small entoconid. A posterior cingulum is present. There are two strong, divergent roots, with circular or elliptical cross-sections. m l , m2, and m3-The anterolabial cingulum is well developed, interrupted at the base of the protoconid (Fig. 2J-L). In m2 and m3, the parastylid is large. A small cingulum closes the trigonid lingually. There may be a small cuspid on the cingulum in the re-entrant valley. In m l and m2, the entostylid is very prominent. There are two roots; the larger posterior one has a groove on its internal surface from base to apex. 11-Broad and bifid, with a weak indentation (Fig. 3A). The labial face is convex, the lingual one is concave. On the medial side of the crown there is a wear facet, which proves that the I1 were in contact with each other. The lingual face bears an irregular, very oblique cingulum. The broad and straight root widens towards its end. The tooth corresponds to morphotype AW of Riimke (1985). 12-More or less elliptical in shape, with weakly defined crests (Fig. 3B). The anterolabial face is generally flattened. There may be a poorly developed posterolabial cingulum. The root is straight, and somewhat compressed laterally. 13-The tooth has a subcircular outline and the anterocrista and posterocrista are variably developed (Fig. 3C). Size is slightly smaller than that of 12. The root is straight. C-Very simple tooth with an elliptical outline, narrowest in
their posterior part (Fig. 3D). There are no crests. The anterior cingulum is missing in the majority of the specimens; the posterior cingulum (metastyle) is small and poorly developed. There are two straight divergent roots, which may be connected at their base by a bony flange. PI-Subquadratic crown with slightly convex lingual and labial faces, and flat, or flat-convex anterior and posterior faces (Fig. 3E). There are no crests. The anterolabial and posterolingual borders bear small but prominent cingula. The single root has a deep lingual groove, but no labial groove. P2-Tooth with an elliptical outline, and a flattened posterolingual part (Fig. 3F). The paracone, without crests, lingually bears a vestigial protocone. There is no metastyle, nor an anterior cingulum. The parastyle is very small. There are three roots: one anterior, one posterior, and a third one under the protocone. This third, lingual root may be isolated, or connected, either to the anterior or to the posterior root. P3-Very similar to P2, but its outline is more rounded (Fig. 3G). The paracone, without crests, lingually bears a vestigial protocone, which is smaller than in P2. The protocone issues a poorly developed cingulum, which surrounds the posterolingual part of the tooth. There are two roots (broken in all specimens.) PGRobust tooth with a small concavity in its llabial face (Fig. 3H). The paracone solely bears a poorly-marked posterocrista. The protocone is well developed. The anterior cingulum is short and does not extend to the anterolingual border. A posterolingual cingulum connects the base of the protocone with the metastyle, and continues posterolabially for a short distance. There are three roots: anterior, posterior and a third, smaller one under the protocone.
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 21, NO. 3 , 2 0 0 1
-
Demana Galemys
A. baelica Archaeodemana
FIGURE 6 . Ranges for length and width of the upper teeth of Desrnana, Galemys, Archaeodesmana baetica, and other Archaeodesmana.
MI-Protocone separated from the protoconule by a deep groove; the accessory cusp is small (Fig. 31). The parastyle is large, and formed by two or three very small tubercles; it is connected to a short labial cingulum. The anterior mesostyle is shorter than the posterior one. The metastyle is poorly developed. On the lingual border of the tooth there may be a small cusp between protocone and protoconule and another one between protocone and metaconule. There are three main roots, under the protocone, the paracone, and the metacone, respectively, and an additional smaller root in the center of the tooth. M2-The placement of the main cusps is similar to that in M1 (Fig. 3J). The parastyle is prominent on the labial face of the tooth. The mesostyles are similar in size among themselves. The metastyle is poorly developed. On the lingual border of the tooth there may be a small cusp between protocone and protoconule; if present it is smaller than it is in MI. There are four roots, located as in M1 and a vestigial fifth root under the mesostyles. M3-The anterior border is straight, with a narrow cingulum that connects the small and prominent parastyle with the protoconule (Fig. 3K). The protoconule is small and connected to the protocone. The accessory cusp is reduced or absent. The metaconule is larger than the protoconule, and displaced towards the metacone. There are three roots. Humerus-The humerus (Fig. 4) is characterized by an enormous medial epicondyle, which is as large as the articulation process. Epicondyle and articulation are separated by a deep depression, which is perforated by the entepicondylar foramen. Posteriorly, the olecranon fossa is small and there is no supra-trochlear perforation. The medial epicondyle provides a
large surface for the insertion of the epitrochlear muscles and the internal ligament of the articulation, as may be expected in these aquatic animals (GrassC, 1967). COMPARISONS
Archaeodesrnana baetica is characterized by considerably different dental proportions than those of other desmanines. The p2 and p4 are very large, similar in size to the same elements in the genus Desrnana, and fall outside the range founkl in other species of Archaeodesmana (Fig. 5). The P2 and P4 are also very large; they are much wider than in other Archaeodesmana species, while their lengths fall within the range known for the genus (Fig. 6). The smallest C of A. baetica is of bbout the same size as the largest C of A. major. A diagnostic feature of A. baetica is the length ratio p2Jp3, which attains a value of approximately 1.4. In other Desmaninae, this ratio is smaller. It ranges between 0.94 and 1.27 in other species of Archaeodesmana, between 1.01 and 1.26 in Desmana, and between 0.94 and 1.O5 in Galernys. The size of the other teeth (incisors, c, PI, p l , P3, p3, and milars) fall within the range of other Archaeodesmana, although P3, p3, and m l are relatively large (Figs. 5, 6). Figure 7 shows the length relationships of the teeth of Archaeodesmana baetica, Archaeodesrnana adroveri, and of the extant Desmana rnoschata and Galernys pyrenaicus. Similar diagrams have been calculated for other species of Archaeodesmana (Riimke, 1985:fig. 46). A. baetica is the only desmanine in which the length of p2 is larger than that of P2. In other
MART~N-SUAREZ ET AL.-NEW
A, sdmved (LA)
A, bsslica (Purd)
.
1.54 C
.
. P1
P2
. P3
.
, P4
M1
. M2
I
M3
ARCHAEODESMANA
0.51
.
,
, C
P1
P2
, P3
.
,
I
P4
MI
M2
M3
-
FIGURE 7. Tooth length diagrams of A, Archaeodesmana baetica from PUR-4. B, Archaeodesmana adroveri from Aljezar B. C, extant Desmana moschata. D, Galemys pyrenaicus. Data for B-D after Riimke (1985).
words, p2 is also allometrically enlarged with respect to the other teeth. Direct comparisons were made between the teeth of A. baetica and those of Desmaninae in the Mein collection. Archaeodesmana verestchagini from Spilia-1 (Greece) and A. dekkersi from Kardia and Ptolemais (Greece) both have a p4 with a welldeveloped talonid, much more molarized than in A. baetica. A. adroveri from Aljezar-B (Spain) has P2, p2, P3 and p3 with much better developed cingula than in A. baetica. A. brailloni from S2te and Balaruc-2 (France) has relatively smaller P2, p2, and p4 than those of A. baetica. In A. major from the Spanish type locality Valdecebro-3, all elements are much smaller than in A. baetica. Desrnana inflata from Valdeganga-1 (latest Pliocene, Spain) has a curved I1 with a pointed tip, and molars similar in size to those of A. baetica, but the premolars are much larger in D. inflata. The p4 from both localities are morphologically very similar; in both cases there is a cingulum surrounding the entoconid. In the P4 from Valdeganga, the parastyle and the metastyle are somewhat more developed than those from PUR-4. Desrnana inflata also shares with A. baetica the weak molarization of the premolars. D. inflata may be a descendant of A. baetica or of another species close to A. baetica. If this is true, the anterior regions of the mandibles and maxillae have been enlarged over the course of time. DISCUSSION The three most common genera of European fossil watermoles are Archaeodesmana, Desrnana, and Galemys. One of
FROM SPAIN
I
553
the diagnostic features (at the generic level) used by Riimke (1985) is the shape of the upper incisors: short, s&aight and bilobed I1 in Archaeodesmana (Fig. 3A), and curved with a pointed tip in Galemys and Desrnana; the latter t b o genera differ in size. Another one of Riimke's diagnostic fkatures for these genera is the shape of i l . In Archaeodesmana, J the crown is large relative to the root (Fig. 2A), and the enamel does not overhang the root anteriorly. The i l of both Galemys and Desrnana have a small crown in comparison with the very long root, while the enamel overhangs the root anteriorly. The I1 and i l from PUR-4 show the features of Archaeodesmana. With respect to the shape of 11, Hutterer (1995) considered the bifid tip to be a synapomorphic character of Archaeodesmana. It is possible that the oldest forms had a bilobed 11. But this type of incisor is also found in younger localities, as is the case in Archaeodesrnana brailloni from Gorafe 2 and Escorihuela (Spain, Late Alfambrian, "MN15," Mein, 1989). And curved and pointed I1 are present in Archaeodesmana verestchagini from the older locality of Spilia 1 (Greece, Early Alfambrian, "MN14," Mein, 1989); nevertheless at that locality two types of I1 are found: pointed (Riimke, 1985:fig. 19.2) and bilobed (not figured, personal observations on casts in the Mein collection). Following Topachevskiy and Pashkov (1983), Desrnana verestchagini Topachevskiy, 1963 must be added to the nine species that Hutterer (1995) included in Archaeodesrnana. This species is known from the localities of Spilia-1, Kardia, Ptolemais and Kastellios (Greece); it used to be the only "Desrnana" from the earliest Pliocene. Its allocation to the genus Archaeodesmana implies that the oldest species of ~ e s m a n ais D. kowalskae from Weze ("middle" Pliocene, Poland). According to Topachevskiy and Pashkov (1983), Archaeodesmana is characterized by double-rooted P1 and p l . But in A. verestchagini, the P I and p l have one root, and the p l of A. dekkersi, A. luteyni, and A. major are single-rooted. So, the number of roots is not a good diagnostic criterion for the genus. The desmanine from PUR-4 is allocated to Archaeodesmana because of the shape of its incisors, and cannot be assigned to one of the other genera or subgenera of the ~esmaninae~for the following reasons: it is smaller than the known species of Desrnana, Praedesrnana, and Pliodesmana; its P2 is smaller than its C, whereas in Praedesmana P2 is larger than C; its p2 is smaller than its p4, in contrast with the situation in Pliodesrnana; its P1 is similar in size to the upper canine, so it cannot be included in Galernodesmana (where PI > C); its p2 and p4 are much larger than in Galemys, for molars of similar size; and it is significantly larger than Mygalinia Schreuder, 1940. ACKNOWLEDGMENTS Dr. Pierre Mein (Lyon) put a wealth of reference material at our disposal, provided us with French translations of papers in Russian, and, as always, his comments have been a valuable contribution to this paper. This study was supported by the project DGICYT PB97-0157, the research group RNM0190 of the 'Junta de Andalucia' and the project 'Cambios climiticos en el Sur de Espaiia durante el Ne6geno' of the R. Areces Foundation. The photographs were made using the Zeiss 950 scanning electron microscope of the 'Centro de ~nstrumentaci$nCientifica' of the University of Granada. I LITERATURE CITED Bendala, N. 1999. Micrornamrniferes de la limite ~iockne-~liockne du bassin de Granada (Espagne). Mkmoire DEA, Universit6 Claude Bernard, Lyon 1:1-50. Berggren, W. A,, D. V. Kent, C. C. Swisher 111, and M. P. Aubry. 1995. A revised Cenozoic geochronology and chronostratigraphy. SEPM
554
JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 21, NO. 3 , 2 0 0 1
(Society for Sedimentary Geology) Special Publication 54:129212. Castillo, C., M. Freudenthal, E. Martin SuArez, M. V. Martinez, and I? Rivas. 1990. New localities with fossil micromammals in the Pliocene of the Granada Basin. Scripta Geologica 93:41-46. Femhdez, J., C. Viseras, and J. Soria. 1996. Pliocene-Pleistocene continental infilling of the Granada and Guadix basins (Betic Cordillera, Spain): the influence of allocyclic and autocyclic processes on the resultant stratigraphic organization; pp. 366-371 in P. E Friend and C. J. Dabrio (eds.), Tertiary Basins of Spain. Cambridge University Press, Cambridge. GrassC, P. P. 1967. Trait6 de Zoologie. Vols. XVI(1) and XVII(2). Masson. Paris. Hutterer, R. 1995. Archaeodesmana Topachevskiy and Pashkov, the correct name for Dibolia Riimke, a genus of fossil water moles (Mammalia: Talpidae). Bomer Zoologische Beitrage 45:171-172. Martin-Su*ez, E., 0 . Oms, M. Freudenthal, J. Agusti, and J. M. Paris. 1998. Continental Mio-Pliocene transition in the Granada Basin. Lethaia 31:161-166. Mein, P., 1989. Updating of MN zones; pp. 73-90 in E. H. Lindsay, V. Fahlbusch, and P. Mein (eds.), European Neogene Mammal Chronology. Plenum Press, New York. Pashkov, A. V., and V. A. Topachevskiy. 1990. Novyye predstaviteli roda Desmana (Insectivora, Talpidae) iz pozdnepliotsenovykh
I
otlozheniy Tsentral'noy Yevropy 1 yugo-zapada yevropeyskoy chasti SSSR [New representatives of the genus Desmana (Insectivora, Talpidae) from the Late Pliocene deposits of Central Europe and the southwest of the European part of the USSR]. Vestnik I Zoologii 24(5):25-34. [in Russian] Rodriguez-Femindez, J. 1982. El Mioceno del sector central de las Cordilleras Biticas. Ph.D. dissertation. Universidad de Granada, Spain, 224 pp. Rzebik-Kowalska, B. R., and J. Pawlowski. 1994. Ruemkelia (Mammalia, Insectivora, Talpidae) nom. nov. for Dibolza Riimke, 1985. Acta Zoologica Cracoviensia 35:75-76. Rumke, C. G. 1985. A review of fossil and recent Desmanmae (Talpidae, Insectivora). Utrecht Micropaleontological Bulletins, Special Publication 4: 1-241. Topachevskiy, V. A., and A. V. Pashkov. 1983. Nadvldovaya sistematika vykhukholey rode Desmana (Insectivora, Talpidae). [The supraspecific systematics of the desmans in the genus Desmana (Insectivora, Talpidae)]. Vestnik Zoologii 17:39-45. [in Russian] , and -1990. Novye predstaviteli roda Desmana (Insectivora, Talpidae) iz eopleystotsenovykh otlozhenyi yuga yevropeyskoy chasti SSSR [New representatives of the genus Desmana from the Eopleistocene deposits in the south of the European part of the USSR]. Vestnik Zoologii 24:28-38. [in Russian] Received 15 May 2000; accepted I December 2000
