Barstovian (Miocene) beavers from Stewart Valley, Nevada, and a discussion of the genus Monosaulax based on tooth morphology CLARA STEFEN Stettiner Str. 6, 53757 St. Augustin, Germany; e-mail: cmstefen@web.de The morphomctric variability of beavers from the Stewart Valley, Nevada, is detailed. The best represented Barstovian sites from the Stewart Valley, Tedford Pocket, and Stewart Springs produced ample material of Monosaulax and a single tooth of Anchitheriomys, which form the basis for this report. Most teeth belong to M. pansus and show morphological and metric variability comparable to other beaver populations. A proximal skull fragment of Monosaulax, assigned to M. pansus, is described. Metric analysis of the check teeth differentiated according to wear stages indicates the possible presence of three size groups of Monosaulax and the teeth are tentatively assigned to M. curtis, M. pansus and a larger Monosaulax sp. However, difficulties with the differentiation of beaver species based only on size as generally done are discussed. The morphological analysis of the cheek teeth in different wear stages and comparison to material ol Steneofiber indicate that the distinction between Monosaulax and Steneofiber may be arbitrary. However, the skull fragment of Monosaulax shows differences from Stcneofil/er. Cranial material of both genera has to be studied in detail to reevaluate the generic status of these taxa and their phylogenetic relationship. PaleoBios 21 (1): 1 -14, April 30, 2001 © 2001 University of California Museum of Paleontology INTRODUCTION Stewart Valley is a particularly rich fossiliferous area in southwestern Nevada (Fig. 1A), which revealed one of the most complete terrestrial Miocene ecological associations known for North America (Savage and Russell 1983). An extensive inventory of Stewart Valley, most of which is pub- lic land under the jurisdiction of the Bureau of Land Man- agement, U.S. Department of the Interior, was directed by Scuddcr et al. (1986) primarily to understand the strati graphic framework and to plot the fossil localities. There arc several fossil-bearing horizons predominantly within a lacustrine and fluviatile sequence (Fig. IB). The fossilifer- ous beds have been dated using the K/Ar method to 16- 10.5 Ma (Schorn et al. 1989). The rocks include interbedded stream, lake and volcanic deposits of Hemingfordian, Barstovian and earlier Clarendonian ages. Over 50 sites, most known for mammals, are recorded in Stewart Valley. Most vertebrate remains come from a gray volcaniclastic sandstone or a gray to brown mudstone (Scuddcr et al. 1986), as do the fossils from the principal Barstovian sites Tedford Pocket (UCMP V5915) and Stewart Springs (UCMP 2027). Ini- tially these rocks were assigned to the Esmeralda Formation (Buwalda 1914, Merriam 1916). However, it was recognized thai the rocks of the Stewart Valley probably had a deposi- tional history separate from the Esmeralda Formation (Albers and Stewart 1972). During deposition the ancient Stewart Valley was located north of a structural high today expressed as a range from the Cedar Mountains to the Grabbs Valley Range. In the Miocene this structural high separated the Stewart Valley from the Gilbert and Esmeralda depositional basins indicating independent depositional histories for these basins (Albers and Stewart 1972), and new lithostratigraphic units for the Stewart Valley were introduced (Schorn et al. 1989). Thus, the older, dominantly andesitic volcanics were referred to the Gilbert Andesite; the younger sedimentary rocks and interbedded tuffs are designated to the new Stewart Valley Group with the lower predominantly lacustrine Sav- age Canyon Formation and the upper fluviatile lacustrine Granny Goose Formation (Schorn et al. 1989:162; Fig 1B). The Tedford Pocket and Stewart Springs fossil sites, where the beavers described in this paper were collected, are in the Savage Canyon Formation. The Miocene fauna and flora from the Stewart Valley are diverse and consist of forms living in and around the lake and its drainage basin. Molluscs, arthropods, fish, birds, squamates, turtles, and 50 families of mammals (29 in the Barstovian and 21 in the Clarendonian) comprise the fauna (Schorn et al. 1989). Beavers and a shrew-like inscctivore are the most abundant mammals at Tedford Pocket. Speci mens from the Stewart Valley arc mainly housed at the Uni- versity of California Museum of Paleontology (UCMP) at Berkeley, Los Angeles County Museum (LACM), Univer- sity of Nevada, Reno, and California Academy of Sciences (CAS). The most comprehensive inventory of fossils and work done on them is by Scuddcr et al. (1986). 'The goal of this report is to describe the morphological and metric variability of beavers from Stewart Valley. The abundant material makes it possible to compare them in detail to other beaver populations, and to discuss StirtoiPs (1935) generic diagnosis of Monosaulax. Using material from Stewart Valley for this analysis is most appropriate, because Stirton based his description on material from Stewart Springs (UCMP locality 2027) rather than Cope's original types. The type material for Monosaulax was collected from the Santa Fe marls of New Mexico (Cope 1874, 1877), was lost at the time of Stirton\s (1935) publication, and has only recently been rediscovered at the Yale Pcabody Museum (Korth 2000). As localities UCMP 2027 and V5915 are in the same- strata and only a few meters apart, the beavers from both 2 PAI.KOHIOS, VOL 21, NUMBER /, APRIL 2001 localities are treated here as one paleontological population. An initial analysis with the material from each locality in scperate groups did not reveal morphological or metric differ- Mincral County B O H CO HI o LITHOLOGY fuvium in drainage systems, 1 ight slooe cover Terrace and fan deposits, and heavy slooe cower WHITE 8I0TITS TUFF (KA=4S2, 10.7 »yBr> ICO- at ease - white to gray sandst. becoming 15Cm Dale green higher in section and in turn grading upward to pale pink sandst. BLUE VOLCAMCLAST'.C SANDST. at base ove lain by brown to pale green to gray sandst. Upper naif white-gray sandst. VARIA8LE LITH0L0GIES - lower halr brown to buff poorly sorted siltst., sandst. and pebble congl. A 10m thick gray sandst. in upper half, overlain by gray sandst. and lenticular beds of siltst. with calcareous cement. Brown sandst. at top. 0IATCMACE0US SHALES AND. KUDSTONE - thin bedded shale below becoming massive- bedded in upper 2/3 of unit. MU0ST0NE - white to buff-colored, mudst with buff-colored shales in basal and upper portion. LAMINATED SHALE - laminated to thln- bcdded. brown to dark gray siliceous VARIABLE LITH0L0G1ES - lower 1/3 light- colored siltst., sandst. and congl. 100- Middle portion dark-colored mudst. *nd 150m siltst. Upper 1/3 light-colored siltst sandst. and blue-gray volcaiiclastic sandst. andesite lavas, shales and dacite breccia Fig. 1. A. Schematic drawing of location of Stewart Valley in Nevada, USA. B. Generalized stratigraphy in the Stewart Valley. Tedford Pocket is marked (TP) and radiometric data (e.g., 16.4 Ma) arc noted. Adapted from Schorn (1971, 1985). ences, supporting the treatment of the beavers from both localities as a single group. MATERIALS AND METHODS About 160 teeth, jaw and skull fragments and numerous postcranial elements of beavers from the Stewart Valley housed in the UCMP collections were used in this study. This material includes: Monosttulax teeth from UCMP lo- cality V5915: UCMP 55200, 55202, 55203, 57884, 58919- 58925,61037, 129688, 129689, 173599, 173602-173605, 173606, 173608-173625, 173627, 173628, 173630, 173633, 173634, 173636-173642, 173644-173648, 173650, 173651, 173655, 173656, 173658, 173661- 173666; postcranial material including astragali, calcanci, vertebrae and long bone fragments: UCMP 55153, 55201, 55204-55215, 57890, 58612, 58613; and Monosaulax teeth from UCMP locality 2027: UCMP 19802, 19803, 31426, 31427, 31428, 31429, 31430-31456, 33990-33994, 58158, 58644, 58645; postcranial material: UCMP 35342, 57915; and uncataloged material from both localities; Ancbitheriomys teeth from locality 2027, UCMP 57909 and 55227. The morphological description of beaver teeth follows Stilton (1935:392) and is illustrated schematically in Fig. 2. The wear stages used here are defined as follows (Pig. 2): unworn—no wear can be observed, tooth crown usually not of full height, if these teeth are in the jaw they have not reached occlusion yet; slightly worn—little wear can be observed, the chewing surface is flat and shows the typical pattern of flexids; medium worn—mesoflexus/id is closing or closed; heavily worn—hypoflexus/id is closing or closed. Measurements of the teeth were taken at the chewing sur- face in the plane of the surface and at the base of loose teeth perpendicular to the long axis of the tooth. Abbreviations used are: AMNH—American Museum of Natural I listory. New York; UCMP—University of Califor- nia, Museum of Paleontology, Berkeley, USNM—United States National Museum; Washington; YPM-PU Yale Peabody Museum-Princeton University collection. New 1 la ven, Connecticut. Upper case letters indicate teeth in the upper dentition; lower case letters indicate teeth of the lower dentition. DESCRIPTION OF MATERIAL The beavers at Stewart Valley comprise Ancbitheriomys and Monosaulax. Morphologically, the Monosaulax sample forms a fairly homogeneous group. However, the dental metric data, especially when separated according to wear stages, are not homogeneous and suggest the presence of three possible size groups. As species of beavers have been separated mainly according to size, the teeth studied are provisionally assigned to M. curtis, M. pansus and a larger M. sp., based on observed metric differences. Problems with differentiating beaver species solely on the basis of tooth size are discussed below. STEFEN-MIOCV.NV. HF.AVF.RS FROM NEVADA 3 hypostriid mesoslriid B parafossettid hypoflexid mesoflexid metafossettid parallexid mesoflesid metaflexid hypofossettld D parastiid mesostriid metastriid hypostria hypoflexus parastria mesostriid mesoflexus parafossette mesolossette mesoflexus metafossette hypolossette Fig. 2. Schematic drawing illustrating denial nomenclature for beaver teeth following Stirton (1935). Order: Rodentia Bowdich 1821 Family: Castoriraf. Hemprich 1820 Subfamily: Acjnotocastorinae Korth and Emry 1997 Genus: Anchitheriomts Roger 1898 A ncbitheriomys sp. Anchitheriomys is represented in Tedford Pocket only by one heavily worn p4 (UCMP 57909) and an incisor frag- ment (UCMP 55227). The broad and circular shape of the premolar at the base and the striation of the incisor enamel is characteristic. Subfamily: Castoroidinae Allen 1820 Genus: Monosaui.ax Stirton 1953 Included Species: Monosaulaxpansus (Cope 1874), Monosaulax curtis (Matthew and Cook 1909), Monosaulax sp. Types: M. curtis-AMmi 13871. M. pansus-USNM 1097 labeled as part and USNM 2041; type rediscovered at Yale Peabody Museum: YPM- PU 10575 (Korth 2000). Morphological Description of Teeth The teeth of Monosaulax are subhypsodont. The striae/ striids are usually without cement, but some does occur and occurrence of cement may not even be uniform within one tooth row. The premolars are larger (in particular longer) than the molars (Fig. 3). On the chewing surface the lophs between the flexi/iids or fossettes/iids are generally simple, smooth, and straight, but crenulations occur in unworn teeth and in some M3s. 'Flic teeth generally show a four-flexus/ flexid pattern with hypo- and mesoflexus/id and para- and mctafossette/id. Incisors^Thc enamel of the incisors is smooth and slightly convex. In cross section the incisors are roughly tri- angular and range from 2.9 x 3 mm to 4.2 x 4 mm in si/.e for Monosaulax pansus (Fig. 4). A larger tooth of 4.2 x 4.9 mm is assigned to Monosaulax sp. Lower premolars—The basic outline of P4 is rectangu- lar to "figure-8"-shaped. The outline changes with wear (as already demonstrated by Stirton 1935): unworn teeth are slightly elliptical, with the tip being much smaller than the base, slightly to moderately worn teeth are more rectangu- lar to "8"-shaped. With wear, the teeth generally become more elongated, which is also demonstrated in the length x width data (Fig. 5). The mesoflexid/fossettid is in the middle of the length of the chewing surface, opening to the lingual side, and the hypoflexid in the middle of the length opening to the labial side of the tooth. The mesoflexid is fairly straight and ap- proximately at right angles to the lingual side of the tooth. The hypoflexid is usually straight and at a marked angle to the labial side of the tooth. The hypostriid does not extend to the base, but ends in the lower third of the crown height. The parafossettid extends through about two thirds of the anterior tooth width, and is straight to slightly curved. Some unworn and slightly worn teeth show a laterally open parafossettid that may open into very short parastriid. The metafossettid does not extend further to the side than the distolingual end of the hypoflexid. Lower molars—In outline, lower molars are square to rectangular and generally slightly longer than wide when unworn, and broader than long when worn (Fig. 6). The mesoflexid is approximately perpendicular to the lingual side of the tooth, is straight and in the middle of the length of the chewing surface. The mesoflexid ends approximately in the middle of the width of the chewing surface. The hypoflexid continues diagonally toward the distal end of the PALEOBIOS, VOL. 21, NUMBER 1, APRIL 2001 Fig. 3. Lower jaws of Monosaulax from UCMP loc. V5915. A. UCMP 58919. B. UCMP 173597. C. UCMP 58922. Scale bar for all = 5 mm. tooth. In unworn to slightly worn teeth the hypoflexid does not continue to the midpoint of the chewing surface, end- ing anterior to the mesoflexid. In medium worn teeth the hypoflexid usually ends in the middle and is approximately opposite the mesoflexid, and both flexids may meet in the middle. In strongly worn teeth it continues further lingually and ends posterior to the mesoflexid. The metatbssettid does not extend further than the distolingual end of the hypoflexid in medium worn teeth. The mesofossettid, and to a lesser degree, the parafossettid, become smaller with wear. The metafossettid is laterally open in 12 teeth (about 15%), and the parastriid opens laterally into a very short parastriae in 5 teeth (about 6%), which also have a short metastriid. Addi- tional small fossettids do occur in some teeth anterior to the parafossettid in unworn and slightly worn teeth. Upper premolars -The P4s are rectangular to triangular in outline. The greatest width of the tooth is at the hypo cone. The hypoflexus in the anterior third of the chewing surface, is straight and points slightly towards the anterior. The paraflexus is opposite the hypoflexus and usually "face- to face" to it (Fig. 7). On the chewing surface the paraflexus is larger than the hypoflexus. In strongly worn teeth the parafossette may continue slightly anterior to the hypoflexus. The mesoflexus/fossette is convex and posterior of the hypoflexus on the chewing surface. The metafossette is usu ally small and straight in the most distal part of the tooth. Metrically they are inhomogeneous and are assigned to M. pansus, M. curtisand M. sp. (Fig. 7). Upper molars-The outline of the upper molars is more oval than that of lower molars. When unworn they are slightly rectangular and longer than wide (Fig. 8). With wear, most become broader than long and more oval (Fig. 9). The mcsoflcxus is always convex and posterior of the hypoflexus and may extend to the lingual side of the tooth. The paraflexus/fossette is usually adjacent and "face to face" to the hypoflexus and does not (generally) extend anterior .STfcFiJN-MIOCKNK BEAVERS FROM NEVADA of the hypoflexid. With wear, the paraflexus closes to a parafossette and becomes gradually smaller on the chewing surface; the hypoflexid extends further towards the labial A A A A A A * A AAA ,A\ A A A A » M. pansus ' M. sp? ' M. pansus type Fig. 4. Incisors of Monosniilnx from Stewart Valley. Diagram with length x width data of incisors. Most incisors arc assigned to M. pansus, whereas one (x) is assigned to the larger M. sp. side. Para- and hypoflcxus are in the anterior third of tooth length and are straight as in upper premolars. The metafossette is usually very small posterior to the curved mesoflexus. In unworn and slightly worn upper molars parafossette and metafossette may be open laterally with a short striae; either one of them or both may be open later- ally in one tooth. Three teeth show short mctastriac, and 6 teeth show a laterally open parafossette with one showing additional fossettes between para- and mesofossette that are also open laterally. A single tooth shows a mesofossette with two openings and striae to the lateral side, and one a mesofossette that opens to both sides of the tooth. M3—The M3s are slightly triangular in outline and slightly smaller than the Ml/2s. The M3s are variable in the number of additional fossettes anterior to the mesoflexus and the form of the mesoflexus. In unworn to slightly worn M3s the mesoflcxid may extend on the chewing surface to the distal side of the tooth and may open there into a very tiny striae. Some M3s show a 5.5 5 4.5 4 3.5 5.5 4.5 3.5 A * A A*4f 1 A A Ao % O AA o unw osl A me nst 3.5 4.5 5 length 5.5 6.5 • M. pansus 0 M. curtis X M. sp. 0 *«&' 0 3.5 4.5 5 length 5.5 5.5 5.5 4.5 A 3.5 B 5.5 5 4.5 -g S 3.5 X X 0 o o 3.5 • » • • -• •: • • •M. pansus o M. curtis •• ? M. curtis type x M. sp. 4.5 5 5.5 6 6.5 length ¦ st M. c. X med M. sp ? unw M. c. ? A M. p. type, unw A med M. o. A M. o. type, basal U st M. o. ¦ A .D A AA*4fVB A« a A A AA m D ?*: ox ? A ? A 3.5 4.5 5 length 5.5 6.5 Fig. 5. Diagram showing length (I) x width (w) data for lower premolars of beavers from the Stewart Valley. A. Monmanlax pansiis differentiated according to wear stages, unw—unworn, si—slightly worn, st—strongly worn, med—medium worn. B. Data for me- dium worn teeth at chewing surface, differentiated according to size and species, M. curtis, M. pansus and M. sp.. C. Length x width data at the base of the teeth. Species assignment according to the seize of the chewing surface. See discussion ill the text. D. Data differentiated according to species (at the basis of the chewing surface) and wear stages. Legend as in A. M. c.—M. curtis, M. p.—M. pansus. 6 PALI-OB I OS, VOL. 21, NUMBER 1, APRIL 2001 5 4.5 4 l,5 3 2.5 A° ID A O Ch A A A°£ A A O O A °OAO OOD A AOA A A © O & O 0 unworn o slightly worn a medium worn ost x probl ? M. curlis type A/W. pansus type o 2.5 3.5 length 4.5 Fig. 6. Diagram showing length x width data for lower molars from Stewart Valley differentiated aeeording to wear stage. No distinction between species is made. The data group is inhomo- geneous, especially for strongly worn molars, unw—unworn, si— slightly worn, st—strongly worn, med medium worn, probl -problematic teeth. laterally open metafossettid with a very short metastriid and some show an additional short parastriid. Metric description of teeth Figs. 6, 8, 10 and 11 show the length x width data for the beaver teeth of Stewart Valley at the chewing surface at different wear stages. The variability in size for each tooth category is large at the chewing surface and ranges over 3 mm for the different wear stages of p4 (statistical data, see Appendix 1). The lower premolars vary mostly in length, the molars in width and the upper premolars about equally in both length and width. The size data for the teeth of the different wear stages indicate a dominantly homogeneous beaver population at Stewart Valley. However, there arc a few teeth that ate smaller and others that are larger than the main group. Based on size differences at the chewing sur- face in the teeth of the main group, M. pansus, these teeth arc tentatively assigned to M. curtis (the smaller teeth) and Monosaulax sp. (the larger teeth). Cranial material A proximal skull fragment (UCMP 129688, Kg. 10, Table 1 for measurements) with the frontal, premaxilla with the incisive foramen, maxilla, anterior part of the palatine and part of the right zygomatic arch is present from Stewart Vallev loc. V5915. The infraorbital foramen i.s slender, fairly straight and rectangular (Fig. 10B) and in lateral view cov- ered by the crista facialis or masseteric ridge (Fig. IOC). The infraorbital foramen is distal to the prcmaxilla-maxilla suture, which is fairly straight in dorsal ventral direction. Compared to Castorthe infraorbital foramen is located more dorsally and i.s more elongate. The crista facialis is slightly curved to "C" shaped. It originates close to the zygomatic arch and transverses most of the anterior face of the maxilla, 6.5 6 5.5 5 4.5 -1 3.5 D DD ADD A D °A oo o o unw M. pansus © si M. pansus a me M. pansus ? st M. pansus A M. pansus type * M. curtis 3 3.5 4 4.5 5 5.5 6 6.5 length 6.5 6- 5.5 5' 4.5 ¦ 4- 3.5. B D D DD DDD D D DD D ? O D d M. pansus ¦ M. curtis 3.5 5.5 6.5 length Fig. 7. Length x width data for upper premolars from the Stewart Valley. A. Differentiated according to wear stage and species, unw—unworn, si slightly worn, st—strongly worn, med—medium worn. B. Basal values. Tentative differentiation of species on the basis of size of the chewing surface in different wear stages, unw—unworn, si—slightly worn, st strongly worn, med—medium worn. following a slightly diagonal, curved course trending anterovcntrally. It ends ventrally in a very prominent masse- teric superfacialis process that extends from anterior to pos- terior of the infraorbital foramen (Fig. IOC]). This skull is tentatively assigned to M. pansus, because the teeth fit best the size distribution for that species. The teeth of the skull fragment (UCMP 129688) show that it is juvenile because the 1'4 has not reached occlusion yet. Using data from Castor fiber (Pietschocki and Stiefel 1977, Pietschocki 1986) to estimate age, it can be inferred that the animal was between 10 and 14 months old. Recent beavers show changes in cranial morphology during ontog env, especially in the crista sagittalis and the relation of fron- tal, nasal and parietal. In very young beavers the frontal is nearly straight from the preorbital process to the postor- bital process, and in adult beavers it is constricted at the STEFEN-MIOCENE BKAVFRS FROM NI-VADA 7 A u D ? 11H1 1II 11 a a a a a A II u a A A H A 19 U a a ft O H 11 o unw O O O 0 0 si A med 1 I si 3.5 length 'lil. u 0*1 O O OOI OOOII • M1.2 M. pansus O M3 M. pansus X M1 2 M. sp X M3 M. sp. Fig. 8. Diagram showing length x width data for upper molars from Stewart Valley. A. Data for all upper molars of Monosautax pansus differentiated according to wear stage, unw-unworn, si— slightly worn, st strongly worn, med—medium worn. B. Data for Ml,2 and M3 at the chewing surface for Monosautax pansus (M. p.) and larger teeth assigned to the lager M. sp. However, teeth of the type of M. pansus arc of similar length as M. sp. differentiated here. orbit. Thus, this juvenile skull of Monosautax may not be exactly comparable to adult skulls of Monosautax. In lateral view the crista facialis or masseteric ridge cov- ers the infraorbital foramen. This crest extends to form a very prominent masseteric supcrfacialis process located ven- trally and extending from anterior to posterior of the in- fraorbital foramen. The morphology of the masseteric supcrfacialis process and crista facialis is very similar to that of Dipoidcs{as illustrated in Wagner 1983) and differs from the fairly straight crista facialis in Castor. In Castor, the mas- seteric supcrfacialis process is not visible in lateral view but may be prominent in ventral view. It varies in size between individuals and is generally more prominent in juveniles than in adults (observations on Castor canadensis in the Mu- seum of Vertebrate Zoology, University of California, Berke- ley). Postcranial material Few Miocene beaver postcrania have been described (e.g., ,S'. deperett by Schreuder 1928, S. viciacensisby Filhol 1879). Therefore, species diagnoses are not possible, and the mate- rial reported here can only be ascribed to Monosautax sp. Postcranial material of beavers from Tedford Pocket is com- prised mainly of broken long bones such as humerus, fe- mur, radius, tibia fragments, and vertebrae. The only bones that are usually complete are astragali and calcanci, which differ slightly from Castor (Fig. II). The plantar (towards navicular and cuboid oriented) facet of the calcaneus is tri- angular in Castor, whereas in Monosaulaxh is more rounded. Examined in dorsal view, the sustcnacular and fibular facets in Monosautax are circular with sharp edge to the tuber calcanci. In Castor only the sustcntacular facet is circular, and the fibular facet is more elongate and less projecting than in Monosautax. These morphological differences indi- cate at least a slight difference in the use of the hind feet, suggestive of slightly different life styles or habitat use. How- ever, a discussion of the biomechanics of limbs and lifestyle of Monosautax is beyond the scope of the current work. DISCUSSION Morphologically the Monosautax sample is fairly homo- geneous. Only the metric data indicate some teeth differing from the main group. Based on metric differences at the chewing surface, the teeth were assigned to M. curtis, M. pansus, and M. sp. Most teeth belong to M. pansus, which shows a variability comparable to other beaver populations (e.g., Steneofiber deperett Crusafont er al. 1948, and ,S'. eseri Ste"fcn 1997). The metric variability of beaver teeth renders division into size-based species difficult, as can be illustrated by a consid- eration of p4s in this sample. The unworn and slightly worn teeth are more or less homogeneous in size, but the me- dium worn teeth can be assigned to M. curtis (small), M. pansus (large), and M. sp. (even larger. Fig. 5B). However, in plotting the basal values of the same teeth (Fig. 5C), only one of the teeth assigned to M. curtis on the basis of the chewing surface value differs sufficiently to separate it. The others arc indistinguishable from M. pansus. Only two teeth exhibit a slight divergence from this group and would form a size group intermediate between M. curtis and M. pansus. The teeth group differently depending on whether they are sorted using chewing surface values (small, large, and larger) or basal size (small, intermediate, and large). This difference in interpretation highlights problems with the distinction of beavers on the basis of size, because it has to be determined what is the "right size" or most appropri ate measure to use. In a group that evolves towards hypsodonty an appropriate measure is especially difficult to determine, because the base of the tooth does not exist as a comparably normed "site." The problem of selecting ap- propriate measures and how to deal with teeth that fit into 8 PALF.OBIOS, VOL 21, NUMBER 1, APRIL 2001 W 5 4 P4 w m 4 -• 3 - w m 4 - 3 - X / w M 4 - 3 -¦ Fig. 9. Metric change of Monosaitlax pniisns teeth with wear from the chewing surface at different wear stages to the base of the tooth. 1 length, w—width. A. Lower premolars showing a unidirectional change. B. Upper premolars showing different directions of change. C. Some lower molars showing a unidirectional change. D. Some lower molars which differ in the direction of metric change from the other lower molars. E. Upper molars showing different directions of change. different size groups (and different inferred species) at the chewing; surface and at the base cannot be resolved here. Rather, it requires a substantial review of other populations of Monosaitlax and other species of subhypsodont beavers to evaluate morphological and metric differences as the ba- sis of species differentiation. Here, teeth are separated on the basis of their chewing surface values. Morphological changes with wear have been demonstrated for Monosaitlax (Stirton 1935: figs. 70-73), and for other beavers (e.g.. Castor by Hiinermann 1966, Stcncofibcr cscri by Stefen 1997). Associated metric changes for the Stewart Valley beavers arc documented in Fig. 9. Metric change with wear is clearly unidirectional only for lower premolars. In all other teeth wear-related changes arc variable (Figs. I3B-C). For lower molars two different directions can be distin- guished: most get broader and a little shorter (Fig. 13B) and some get shorter and less wide (Fig. 9C). These het- erogeneous metric changes with wear differ from the largely unidirectional changes seen in S. eseriAnd S. dcpcrcti(Sict'c\-\ 1997, Crusafont et al. 1948, respectively). For other species STEEEN-MIOCENY, BKAVKRS FROM NEVADA 9 Fig. 10. Proximal .skull fragment of Monosaulax pansus from the Stewart Valley, UCMP 129688. Scale bar - 1 cm. A. Ventral view. B. Frontal view. Arrow points to infraorbital foramen. C. Dorsal view. D. Oblique lateral view showing the prominent masseteric super! icialis process. of Stcneofibcr, Monosaulax, and Eucastm; such metric changes are not well documented. The differences in metric changes with wear from the chewing surface to the base could indicate different species. 1 lowcvcr, the teeth diverging in this feature are not distin- guishable on the basis of any other character or by size of the chewing surface. Additional data from different popula- tions and species need to be analyzed to provide an inter- pretive framework for these observations. The direction of metric change with wear could be an important feature to differentiate between metrically similar species. Morphologically the beaver teeth are homogenous, and few teeth show unusual features. A few M3s possess a mesoflextis that is open on the lingual and labial side of the tooth, or on the distolingual side. One upper molar has mesostria and merastria that are about equal in length, and a single I'4 has parastria longer than the mesostria. These teeth do not differ metrically from the average. Generally in Eucastor the parastria is longer than the mesostria in P4 and the presence of this feature may foreshadow the evolu- tionary development towards Eucastor. Another feature that is used to differentiate beaver taxa is the filling of flexids/ii, particularly hypo- and mesoflexids/ii, with cementum. In the material from the Stewart Valley this feature is variable. Taxonomy Stirton (1935) assigned four species to Monosaulax: M. hespcrus (Douglas 1901), M. complexus(Douglas 1901), M. pansus (Cope, 1874), and M. curtis (Matthew and Cook 1909), and noted that a distinction between them may prove invalid. I le distinguished between M. pansus and M curtis primarily in size and between M. complexus and M. pansus 1(1 PALEOBIOS, VOL. 21, NUMBER 1, APRIL 2001 Tabic 1. Cranial measurements in mm and characteristics of Monosaulax pansus, UCMP 129688, from Tedford Pocket. width of rostrum 13.9 height of rostrum 11.9 length of snout (incisor to 27.9 preorbital process) diastema 25 length of incisive foramen 7 foramen in premaxilla anterior yes to incisive foramen form of infraorbital foramen rectangular; not bulged laterally width at preorbital process 21.3 constriction at orbit 13.3 length of palatine in sagittal plane 5.5 anterior end of palatine posterior end of palatine (in sagittal plane) distance P4/P4 distance Ml/Ml approximate height of skull at tooth row within anterior third of Ml within alveolus of M3 5 6.8 22.2 length of premaxilla from incisor to incisive foramen 10 in structure (Stirton 1935:416). Xu (1994) included M. corn- plexus with the European species Stencofiber depereti and the Asian species M. cbanapeinensis and M. tunjjurensis in Stencofiber bespcrus. He assumed that Monosaulax was a synonym of Eucastor (Xu 1994:86). Korth (1996) discussed the type specimens of Monosaulax com plexus And M. bespcrus in detail and synonymized them under a new genus, Neatocastor bespcrus. He distinguishes them mainly on the presence of DP3, which is absent in Monosaulax and most other beavers. The skull from the Stewart Valley lacks DP3 and is therefore distinct from Neatocastor. Another feature that Korth (1996:173) used to distinguish between Neatocastor and species referred to Monosaulax is the complexity of the lophs separating the flexi of the cheek teeth in Neatocastor. However, crenula- tions and complicated minute ridges do occur in fairly un- worn teeth of other beavers as well and can be observed in the material from the Stewart Valley and in the European .S'. escri. M3s seem to have more crcnulated lophs than other teeth. Generally the complexity and crenulation of the lophs is lost during castorid evolution (see also Hugueney 1975). Korth (1996:175) further distinguished between Stencofiber and Neatocastor based on the ratio of the inci- sive foramen to upper diastema length. Unfortunately, it is 7.5 1 7 6.5 H B 5.5 " 5 4.5 "I -I 3.5 3 7 6 ;..s 5 4.5 4 3.5 3 2.5 2 P4 •* \ . >A* ** * * ?W "AAA -*----------1------ • S. eseri S. viciacensis A M. pansus ¦ M. typicus M. prooresus 6 length m ••• .,* %M A /• » 2 2.5 3 3.5 4 length • S. eseri S. viciacensis A M. pansus M. lypicus 1 M. progress 5.5 6 6.5 1 6 5.5 5 4.5 " 4 ¦ 3.5 3 2.5 ' M1/2 D 4 S. viciacensis • S. eseri A M. pansus M. lypicus M. piogicsus 2 2.5 3 3.5 4 4.5 5 5.5 length Fig. 11. Diagram showing length x width data for teeth of Monosaulax pansus from Stewart Valley in comparison to Steneofiber eseri material in SMNS, S. viciacensis measurements of material in Basel and data from Filhol (1879), S. depereti measurements in Basel and M. typicus and M. progresus data from Shotwell (1968). p4—lower premolars, m—owcr molars, M upper molars. not clear from which species of Stencofiber (which differ markedly in size, Stcfcn 1997) he obtains the cited ratio of 0.23-0.25. The ratio for the Monosaulax skull from Stewart Valley is 0.3, which is close to Korth's values for Stencofiber. The diastema is relatively smaller in juveniles than in adults in Castor (Freye 1959) and thus this ratio changes with age. STEEEN-MIOCV.NE BKAVERS FROM NEVADA 11 The ratio of the present juvenile may not be representative for the genus in general. Based on cranial characters, Xu (1994) distinguished be- tween Steneofiber and Eucastor, which he considered syn- onymous with Monosanlax. It is difficult to assess his characters for the material from Stewart Valley due to the state of preservation. Only the masseteric ridge or crista facialis is present in the .specimen from Tedford Pocket, and it is curved to "(""-.shaped, ending in a very prominent masseteric superfieialis process located anterior and ventral to ihe infraorbital foramen (Fig. lOd). The infraorbital fo- ramen is near the middle or slightly ventral to the middle of the snout height and rectangular in frontal view (Fig. 10B); the crista facialis does not bulge laterally but i.s straight in frontal view. In Steneofiber the masseteric ridge is straight to very slightly curved in lateral view (Gervais 1859: pi. 48, Viret 1929: pi. II, Xu 1994, Stefen in press) and ends in a small masseter superfieialis process ventral to posterior of the infraorbital foramen. The infraorbital foramen is in the lower third of the snout height and tear shaped, and the crista facialis bulges laterally around it in frontal view (Stefen in press). Olson (1940) in his pioneering work on cranial foramina of beavers considered only about 27% of the fo- ramina to be of laxonomic significance, including the in- fraorbital foramen. Thus, these observed differences in the infraorbital foramen and crista facialis indicate species dif- ference between Monosaulax pansus and Steneofiber viciacensis\ whether they also indicate generic difference be- tween Monosanlax and Steneofiber has to be evaluated fur- ther as other cranial features can hardly be evaluated with the snout fragment from Stewart Valley. Fractures in the zygomatic arch of the skull from Stewart Valley indicate that the jugal and lacrimal did not meet and that the lacrimal had a small dorsal component only. This morphology and the palatal grooves are similar to Steneofiber (Stefen in press). I lowever, a skull attributed to Monosanlax teiii 1 Korth 1999: fig. 1) exhibits a larger dorsal component of the lacrimal. The skull fragment illustrated by Korth (1999: fig. 1) differs from the skull in Tedford Pocket by the stronger curvature in the premaxilla posterior to the incisors. It i.s not within the scope of this paper to review the European or Asian species that have been assigned to Monosaulax and Steneofiber. The teeth of Monosanlax pansus arc slightly smaller than those of the European S. viciacensis and S. escri from the lower Miocene (Fig. 11). The few teeth in the Stewart Valley sample that are larger than Monosaulax pansus are not assigned to another spe- cies. The assignment of a few beaver teeth to a new species is problematic because a clear metric definition and distinc- tion would be difficult, especially as no consistent morpho- logical feature clearly separates these teeth from the other. Shotwell (1968) described M. processus and states that it is larger than M. pansus, but a comparison with the Stewart Valley material does not show this (Fig. I1C). In 1935 Stirton described the genus Monosanlax and con- fusion prevailed in the literature concerning this name. Monosaulax was introduced as associated with the Mcrycbippus faunas (Stirton 1935:416) and as "evidently confined to the North American late Miocene deposits, but [...] more closely related to 'Monosaulax1 or 'Steneofiber" of Europe than to Palacocastor." Later, Stirton (1935:423) refers European material to "Monosaulax"', specifically "A/." gym (from the lower Miocene) and "A/." minutus from the middle Miocene) and confines the occurrence of the genus Steneofiber to the "Acjuitanian and probably [...] the BurdigaliarT (Stirton 1935:397). The diagnosis and description of Monosaulax is mainly based on material from Stewart Springs (UCMP loe. 2027) (Stirton 1935:416). The material from the Stewart Valley presented here generally supports Stirton's description of Monosaulax, but the analysis of the new material indicates some refinements of the diagnosis. According to Stirton (1935:416) the p4 parafossettid never opens as a flexid as in Eucastor. However, the new material includes unworn and slightly worn lower premolars where the parafossettid opens laterally into very short striids, though not as pronounced as in Eucastor. Stirton (1935) did not comment on the ba- sic outline of the p4s, although it is visible in his drawings that they are slightly triangular (Stirton 1935: fig. 70). These observations and the recent description of mor- phologic and morphometric variability in Steneofiber escri (Stefen 1997) calls into question the distinction between Steneofiber and Monosaulax on the basis of tooth morphol- ogy. The generic differences given by Stirton (1935:397 and 416, respectively) are slight (Table 2). In S. escri (Stefen 1997) the parafossettid of the fairly unworn lower p4 may open laterally into very short striids, not into large ones as in Eucastor. Also, in unworn to slightly worn lower molars para- and metafossettids open laterally. Unworn P4s show a parafossette that opens laterally, and upper molars show ten dencics to form short metastriac. As far as the teeth are concerned there are no marked morphological differences sufficient to justify a generic dis- tinction. Both have subhypsodont cheek teeth. Anterior fossettes of p4 and lower molars may open laterally in un- worn to slightly worn teeth, but are never as pronounced as in Eucastor. "The upper premolar is slightly triangular in outline, and the outline varies with wear. Molars are smaller than premolars; especially lower premolars are longer than lower molars. Mesostriids/striae are shorter than hvpostriids/striae in all teeth, and hypostriid/stria never extend to the base of the tooth. 'The distinction between a convex enamel face in the incisors of Steneofiber and a round one in Monosanlax seems to be arbitrary and not well de- fined. A difference between the Monosaulax skull and Stirton's diagnosis for Steneofiber (1935:397) i.s in the location of the masseter superfieialis process: it is anterior to posterior of the infraorbital foramen in the Monosaulax skull and is 12 PALEOBIOS, VOL 21, NUMBER 1, APRIL 2001 Tabic 2. Differences between Steneqfiber and Monosaulaxaccording to the generic diagnosis ofStirton (1935). Steneofiber Monosaulax incisors with convex enamel lace incisors with round enamel face anterior fossettcs never opening outward as in Eucastor p4 parafosseliid never appearing as flcxids as in Eucastor lower molars with parafosscttids and mctafossettids opening internally in unworn teeth but not as distinct flcxids lower molars smaller and not as elongate as p4 P4 with a short parastria and a long hypostria; mesofosseiie and meiafossette present hypostria of upper teeth, although longer than the mesostria, not extending to base of tooth (comparable to Palcocastor in this respect) I'4 more rectangular than triangular in its basal outline present P4 with a short parastria; mesofossettc and meiafossette dominantiy posterior to the infraorbital foramen in Steneofiber. However, lor a final conclusion on the status of Steneofiber and Monosaulax (and Eucastor, as Eucastor has been as- sumed to be synonymous with Monosaulax, Xu 1994) more- cranial material has to be reviewed and described in detail, because features of the skull are more conclusive for generic differentiation than arc teeth. In particular it has to be evalu- ated how much the masseter ridge and masseteric siipertacialis process vary within one species and genus to evaluate the differences seen in the skulls of Steneofiber and Monosaulax. The skull features are particularly important for beavers as the teeth are very conservative during the evolution of the family. A reevaluation of Monosaulax and Steneofiber and their evolutionary changes would not only clarify their taxo- nomic status but also help to understand the biogeographic development of beavers in Europe and America. SUMMARY The beavers from Stewart Valley include Anchitbcriomys and Monosaulax with Monosaulax dominant. The teeth as- signed to Monosaulax arc morphologically homogeneous and comparable in morphological variability to other beaver populations. However, due to size differences at the chew- ing surface the teeth can be put in three groups and arc- assigned to M. pansus, M. curtis, and a larger form of Monosaulax, M. sp. The difficulties with this assignment are- discussed. Comparison of Monosaulax pansus and Steneofiber cseri indicates that the dental differences described by Stirton (1935) for Steneofiber And Monosaulax Arc variable and t)o not justify the generic distinction. For a final decision on the taxonomic status of both genera, a detailed analysis of cranial material is necessary, as some differences could be observed. ACKNOWLF.DGMF.N IS I thank W.A. Clemens, UCMP, for enabling me to work in his lab, and the Museum of Paleontology and the De- partment of Integrative Biology, University of California, Berkeley, for their hospitality during my studies. My stay in Berkeley was possible through a scholarship from the Deutsche Forschungsgemeinsehaft (grant ST 798/1-1), whose support is gratefully acknowledged. II. K. Schorn was helpful with information on Stewart Valley. A.D. Barnosky, H.M. Wagner, and P.A. Holroyd made useful comments on the manuscript. LITERATURE CITED Buwalda, J.P. 1914. Tertiary mammal beds of Stewart and lone Valleys in west-central Nevada. University of California De- partment of Geological Sciences liiilletin 8:335 363. Cope, E.D. 1874. On a new mastodon and rodent. Proceedings of the Academy of Natural Sciences Philadelphia 1874:221-223. Cope, E.D. 1877. Report upon the extinct Vertebrata obtained in New Mexico by parties of the expedition of 1874. Geographi- cal surveys west of the one hundredth meridian. First Lieut. Geo. M. Wheeler, Corps of Engineers, U.S. Army, in charge. Vol. IV, Palaeontology, pp. 1-370. Crusafont-Pairo, M., F. Villata, and I.R. Bataller. 1948. I.os Castorcs fossiles de Espana. parte 1 et 2. Boletin del Instituto Gcolopico y minero dc Espana 61:319 449. Filhol, M.I I. 1879. Etude des mammifecs fossiles de Saint-Gerand le Puy (Ailler). Annates des Sciencesjjeolojjiqiics 10(3): 1-252. Freye, 11.A. 1959. Descriptive Anatomic des Craniums vom Kibe Biber (Castor fiber albicus Matschie 1907). Wissenschaftlicbe Zeitschrift dcr Universitat Halle, matbematisb- naturwissenschaftliche Rcibc 8:913-962. Gervais, P. 1859. Zoologie et Paleontologie francaise. (Animaux vertcbrcs) Premiere partie: Mammiricrcs. Arthur Bert rand, 2nd Edition, Paris, xx p and pi. 48 fig. 9. Hunermann, K.A. 1966. Der Bau des Biber Praemolarcn und seine Verwendbarkeit fur die Svsrematik der Castoridae (Rodentia, STEFEN-MIOCKNE BEAVERS FROM NEVADA 13 Mammalia), Ncties Jahrbuch fur Geologic und Palaontologie, Abhandlnngcn 125:227-234. Hugueney, M. 1975. Les Castoridae (Mammalia, Rodentia) dans I'Oligoc&ne d'Europe. Colloques Intcrnationaux Centre Na- tional de la Recherche Scicntifique 218:791 -804. Hugueney, M. 1999. Family Castoridae. pp. 281-300 in G. Rossner (ed.). The Miocene Land Mammal Ages in Europe. Pieil Verlag Miinchen. Korth, VV.VV. 1996. A new genus of beaver (Mammalia: Castoridae: Rodentia) from the Arikareean (Oligoccnc) of Montana and its bearing on castorid phylogeny. Annals of Carnegie Museum 65(2): 167-179. Korth, VV.VV. 1999. A new species of beaver (Rodentia, Castoridae) from the earliest Barstovian (Miocene) of Nebraska and the phylogeny of Monosaulax Stirton. Paludicola 2(3):258-264. Korth, VV.VV. 2000. Rediscovery of lost holotype of Monosaulax pansus{Rodentia, Castoridae) Paludicola 1(4):279-281. Merriam, ).C. 1916. Tertiary vertebrates fauna from the Cedar Mountain region of western Nevada. University of California Berkeley Department of Geological Sciences Bulletin 9:161 -198. Piechocki, R., and A. Stiefel. 1977. Zahndurchbruch und /.ahnwcchscl beim Elbebiber, Castor fiber aWicus. Anatomischcr Anzeigcr 142:374 384. Piechocki, R. 1986. Osteologisehe Kriterien zur Altersbcstimmung des Elbebibers Castor fiber a/bieus. Zoologische Abhandlnngcn des Museums fur 'Vicrkundc Dresden 41( 15): 177 183. I'omel, A. 1847. Note sur des animaux fossiles decouverts dans le departement de l'Allier. Bulletin dc Societcgcologique de France IV(2): 378 85. Savage, D.E., and D.E. Russell. 1983. Mammalian Paleofaunas of the World. Addison Wesley, Reading, Pennsylvania, 432 pp. Schreuder, A. 1928. Humerus von Steneofiber depcrcti Mayet. Paldoutologische Zcitschrift 10(2): 125-129. Scudder, H.I., |.R. Firby, T. I.ugaski, J.E. Mawby, D.E. Savage, and U.K. Schorn. 1986. Report of a paleontological inventory of the Stewart Valley fossil area prepared for the Bureau of I .and Management, Nevada Slate Office. U.S Department of the Interior, Reno, Nevada, 171 pp. Schorn, H.E., H.I. Scudder, D.E. Savage, and J.R. Firby. 1989. General stratigraphy and paleontology of the Miocene conti- nental sequence in Stewart Valley, Mineraly County, Nevada, U.S.A. pp. 157-173 in G. Liu, R. Tsuehi, and Q. Lin (eds.). Proceedings of the International Symposium of Pacific Neo- gene Continental and marine Events, National Working Croup of China for IGCP-246. Shotwell, J.A. 1968. Miocene mammals of southeast Oregon. Bulletin of the Museum of Natural History, University of Or- egon 14:1-67. Stefen, C. 1997. Steneofiber eseri(Castoridae, Mammalia) von der Westtangente bei Ulm im Vergleich zu andcrcn Biberpopulationen. Stuttgartcr Bcitrdge zur Naturhundc 255:1-78. Stefen, C. Cranial morphology of Steneofiber (Mammalia: Roden- tia: Castoridae). Gcodivcrsitas. (in press). Stirton, R.A. 1935. A review of Tertiary beavers. University of California Publications in Geological Sciences 23( 13):391 458. Viret, J. 1929. Les faunes de Mammiferes de I'Oligocene superieur de la Limagne Bourbonnaise. Annates dc l.'Univcrsitc de Lyon Nouvclle Scric, I, Sciences Medicine fascicule 47:1 -305. Wagner, H.M. 1983. The cranial morphology of the fossil beaver Dipoides smithii (Rodentia: Mammalia). Contributions in Sci- ence, Natural History Museum Ij>s Angeles 346:1-6. Xu, X.F. 1994. Evolution of Chinese Castoridae. pp. 77-97 inY. Tomida, C.K. Li., and T. Setoguchi (eds.). Rodent and Lago- morph Families of Asian Origins and Diversification. National Science Museum Monographs 8. ADDENDUM Hugueney (1999) states that for beaver material from St. Gerand-Lc Puy, the name Steneofibercastorinus Pomel 1847 has priority over .V. viciacensis Gervais 1848-1852, even though Pomel himself later synonomized S. castorinuswith the German form .V. eseri. Thus the name S. viciacensis used in this paper should read .V. castorinus. However, whereas the type of .V. viciacensis is a fairly complete skull, valuable for taxonomic character- ization of a species and genus, the type of S. castorinus is a lower jaw fragment and a small skull fragment, figured by Pomel (1847) but not located by Hugueney (1999) or my- self as yet. .V. castorinus and .V. viciacensis wax both founded on material from the St. Gerand-Ix-Puy region. APPENDIX 1 Summary data to for the teeth of Monosaulax pansus from Stewart Valley, UCMP localities 2027 and V5915. I—length, w—width, h—height. TOOTH WEAR STAGE NO. OF TEETH X MIN X MAX AVERAGE DEVIATION unworn 1 w 7 7 10 3.35 2.95 4.9 4.4 3.3 6.0 5.3 4.6 3.77 3.12 5.62 4.7 4.0 0.37 0.13 P4 unworn and slightly worn h 0.45 medium worn strongly worn 1 w 20 20 4.0 3.5 0.36 0.25 1 w 10 10 5.1 4.1 6.0 4.9 5.6 4.4 0.28 0.26 14 PALEOBIOS, VOL 21, NUMBER 1, APRIL 2001 APPENDIX 1 (continued) Summary data to for the teeth of Monosaulax pansus from Stewart Valley, UCMP localities 2027 and V5915; 1 —length, w—width, h—height. TOOTH WEAR STAGE NO. OF TEETH XMIN XMAX AVERAGE DEVIATION P4 all basal all chewing surf. 1 w ee ' w 37 37 43 43 5.1 3.8 3.2 2.95 6.8 5.1 6.1 4.3 0.32 0.29 6.0 4.9 4.7 3.9 0.73 0.49 all unworn slightly worn unworn and slightly worn medium worn 1 w 1 w h 1 w 7 7 11 11 14 34 34 3.0 2.6 3.2 2.9 3.4 3.2 3.2 2.9 0.13 0.20 3.8 3.5 3.45 3.2 0.24 0.20 m 3.1 3.0 2.9 4.7 4 0.70 3.9 4.0 3.4 3.5 0.24 0.30 strongly worn all basal 1 w 1 w 20 20 42 46 2.7 3.4 2.2 2.7 4.0 4.8 3.2 4.0 0.34 0.40 3.7 4.8 3.1 3.8 0.26 0.40 unworn 1 w 5 5 3.4 3.2 4.0 4.1 3.8 3.9 0.17 0.10 unworn h 1 - 4.5 - P4 medium worn 1 w 6 6 4.7 4.9 5.0 5.7 4.85 5.2 0.11 0.30 strongly worn 1 w 11 11 4.7 5.0 5.0 6.2 4.95 5.6 0.09 0.40 all basal 1 w 17 17 3.8 5.2 5.0 6.2 4.55 5.8 0.52 0.80 unworn 1 w 5 5 2.9 2.6 3.0 3.5 3.1 3.0 0.0') 0.33 slightly worn 1 w 6 6 2.8 2.7 3.2 3.0 3.0 3.4 0.14 0.32 Ml,2 all unworn and slightly worn medium worn h 1 w 7 20 20 3.2 2.9 3.0 4.0 3.3 4.0 3.4 3.1 3.5 0.70 0.12 0.34 strongly worn 1 w 26 26 3.0 3.2 3.6 4.5 3.2 3.85 0.19 0.29 basal 1 41 2.2 3.8 2.95 0.29 w 41 3.2 4.5 3.9 0.29 M3 medium worn all basal 1 w 1 w 8 7 11 11 2.8 3.0 3.0 3.1 3.5 3.5 4.7 4.1 3.1 3.3 3.1 3.3 0.25 0.23 0.25 0.23 incisors 1 w 50 50 2.2 2.2 4.2 4.3 3.5 3.7 0.41 0.43