Occasional Papers Museum of Texas Tech University Number203 22 December 2000 Two New Genera and Species of Halophytic Desert Mammals from Isolated Salt Flats in Argentina Abstract Two new genera and species of rodents are described from isolated salt pans in far northwestern Argentina. Both are members of the hystricognath family, Octodontidae. The two new genera are both specialized for life in zones of high salinity with halophytic vegetation growing at the periphery of saline flats. The new taxa are most closely related to Tympanoctomys barrerae, the Red Vizcacha Rat, which occurs in salt flats in central Argentina further to the south. One of the new mammals is restricted to an isolated salt flat lying within an enclosed bolson in Catamarca Province. The other is limited to habitat islands that occur within the vast Salinas Grandes (great salt flat) of La Rioja Province in central Argentina. Taxonomic descriptions are included, as is information on relationships within the family Octodontidae. The biogeographic history of the group, and an evolutionary scenario for the evolution of the two new genera, are also discussed. Front cover: The two new genera and their habitats are shown over a satellite composite photo of western Argentina. The isolated Bolson de Pipanaco in the Monte Desert Biome is evident as the type locality for the Golden Vizcacha Rat. Both the rat and its habitat are shown in photos on the left side of the cover. The Chalchalero Vizcacha Rat occurs in the Salinas Grandes, a vast salt flat lying within the Chaco Biome. The position of the type locality for the rat is indicated. A photo of the new genus and species and its habitat are shown on the right side of the cover. Photos: M. A. Marcs. 1 Two New Genera and Species of Halophytic Desert Mammals from Isolated Salt Flats in Argentina Michael A. Mares, Janet K. Braun, Ruben M. Barquez, and M Monica Diaz South Amenca is especially rich in manimals, with more than 1,000 species having been described to date (Patterson, 1994). Although the continental landmass makes up only 12 percent of the world total, 25 percent of the known species of mammals of the world occur in South America (Wilson and Reeder, 1993). In addition to supporting high levels of species richness on the con¬ tinent as a whole. South America also has elevated spe¬ cies richness in particular localities (a diversity). For example, sites in the lowland Amazon rain forest or on the rich moist forested slopes of the Peruvian Andes may support as many as 60 species of bats, along with 79 additional species of mammals (e.g., Patterson et al., 1998; Voss and Emmons, 1996). A pattern of chang¬ ing mixes of species as one moves from the tropics to the temperate regions (p diversity) also contributes to the large numbers of mammal species found in South America (Mares and Ojeda, 1982). There also are portions of the continent that sup¬ port many endemic species, genera, and even families of mammals, and these sites are distributed differen¬ tially in the arid and semiarid parts of South America (Mares, 1992). Endemic and monotypic taxa contrib¬ ute to elevated levels of higher-order genetic diversity that characterize the mammals of the continent’s drier biomes. There has been some controversy about the idea of elevated diversity in arid lands (e.g., Mares, 1992; Voss and Emmons, 1996), but support for the idea of greater numbers of unique taxa or more tlireatened taxa in non-rain forest habitats extends to groups other than mammals (e.g., Barthlott et al., 1996, for flowering plants; Beissinger et al., 1996, for birds; Kalin Arroyo et al., 1988, for flowering plants; Mittemieier et al., 1999, for flowering plants; and Platnick, 1991, for spi¬ ders). In recent years we have been surveying the mam¬ mals of Argentina. Because of a long history of spo¬ radic collecting in the country (e.g., Crespo, 1960; Hershkovitz, 1987; Mares, 1982), and with most col¬ lecting having taken place in populated easternmost Ar¬ gentina, many workers have considered the country to be well studied from the standpoint of its mammal fauna. However, we have encountered undescribed taxa of mammals in all parts of the country with some regular¬ ity. In recent years new taxa have been found in the mesic forests of the Northeast (e.g., Mares and Braun, 2000) and in the mesic Yungas forests of the Northwest (e.g., R. M. Barquez and M. M. Diaz, pers, obs.; M. A. Mares and J.K. Braun, pers. obs.). Most new taxa, however, have been encountered in the deserts, scru¬ blands, and grasslands of western and northwestern Ar¬ gentina (e.g., Salinomys —Braun and Mares, 1995; Andalgalomys —Mares and Braun, 1996; Akodon aliquantulus —Diaz et al., 1999). Northwestern Argentina is interesting for a number of reasons. The entire zone consists of an integradation of Andean and pre-Aidean mountains and lowland desert valleys (Fig. 1). Some of the mountains are isolated from major mountain chains. The lowland desert valleys are also isolated from other similar val¬ leys by the high mountains (e.g., Mares 1975, 1976; Mares etal., 1977; Mares etal., 1997; Morello, 1958; Orians and Solbrig, 1977; Williams and Mares, 1978). The region falls within the Tropical Andes designation as a biodiversity hotspot (Mittenneier et al, 1999), m part because of the verdant Yungas forest that extends along the eastern face of the easternmost mountains as far south as southern Catamarca Province (30® south latitude). The recognition of the region as a biological hotspot is also fitting because recent research suggests that the northern Monte Desert and its sunoundmg mountains are biologically distinct from the southern deserts. The biogeographic regions in far northwestern Argentina were once considered to be the Monte Desert in the lowland habitats (Morello, 1958) and the Puna and Prepuna deserts in the highlands (Cabrera, 1957, 1976). The topographically complex region is now viewed as different enough in its biological features that it has been suggested that it be recognized as a separate zoogeographic region; The Monte Desert of Mountains and Isolated Valleys (Burkart et ah, 1999). Mares etal,, 2000, Two New Genera and Species or HALOPfiYTic Desert Mammals from Isolated Salt Flats in Argentina. Occasional Papers, Museum of Texas Tech University 203a+1 -27, 2 Occasional Papers, Museum of Texas Tech University Presently, we are in the process of naming about a dozen taxa of previously undescribed mammals from northwestern Argentina; most have been collected only within the last few years. In this report, we describe two new genera and species of rodents from isolated salt flats in northern Argentina. These animals are es¬ pecially unusual in that they are specialized for an ex¬ istence in salt flats and forage on salt-adapted plants. Desert mammals specialized on the halophytic vegeta¬ tion of salt flats are extremely rare. Worldwide, only three species of mammals (all rodents) are salt special¬ ists: North American Dipodomys microps —the chisel- toothed kangaroo rat, family Heteromyidae (Kenagy, 1972); North African Psainmomys obesus —the fat- tailed sand rat, family Muridae (Degan et al., 1988); and South American Tympanoctomys barrerae —^tlie red vizcacha rat, family Octodontidae (Mares et al., 1997; Ojeda et al, 1999). Each species forages on halophytic vegetation and inhabits zones of high salt content, low plant diversity, high insolation, high temperatures, and low precipitation. The new genera and species are in the family Octodontidae (tribe Octodontini) from the Monte Desert of Mountains and Isolated Valleys and the Dry Chaco. The family Octodontidae is known from the Late Oli- gocene in South America and the tribe Octodontini is known from the Late Miocene of Argentina (McKenna and Bell, 1997; Vucetich et al, 1999). One of the new taxa occurs in a valley (Fig. 1) that has been isolated from similar habitats by pre-Andean mountain ranges since their late Miocene-early Pliocene uplift (Pascual and Ortiz Jaureguizar, 1990). The other new taxon oc¬ curs in the Salinas Grandes (Great Salt Pan) of central Argentina (Fig. 1). The third known salt-specialist on the continent, T. barrerae, which is also an octodontid, Figure 1. Satellite photo of arid western Argentina showing the isolated Bolson de Pipanaco (enclosed m the square), as well as the Salinas Grandes (lower right in square). The high mountain ranges that demarcate the Monte, Puna, and Patagonian desert valleys are evi¬ dent, with the pre-Andean ranges toward the east and the main Andean Cordillera to the west. occurs in salt flats 600 km to the south (Mares et al, 1997; Ojeda et al, 1999), Herein we describe two new genera and species and report the only known occurrence of tliree species of salt-specialized mammals from the same continent. Moreover, the discovery of two new monotypic genera of desert rnanmials continues to underscore the impor¬ tance of arid South America as a major source of unique genetic diversity at supraspecific levels (Marcs, 1992). Materials and Methods We took 19 external, cranial, dental, and man¬ dibular measurements from specimens of the new taxa and from specimens of the other members of the family Octodontidae: Aconaemys fuscus, Octodon bridgesi, O. degus, O. lunatus, Octodontomys gliroides, Octomys mimax, Spalacopus cyarius, and Tympanoctomys barrerae. External measurements (in millimeters) from the labels were: total length, length from the tip of the snout to the last caudal vertebra; length of head and body, total length less length of tail; length of tail, length of caudal vertebra; length of hind foot, length of pes from heel to tip of longest claw; length of ear, length of pinnae from notch. Temiinology for dentition generally follows Reig (1977). Cranial terminology generally follows Carleton and Musser (1989). Coloration of the holotypes follows those standards proposed by Ridgway (1912). Mares et al.—- Two New Genera of Mammals 3 Cranial measurements (in millimeters) were taken with dial calipers to the nearest 0.1 mm and included: greatest length of skull, least distance from posterior border of supraoccipitals to tip of nasals; basal length, least distance from anterior edge of premaxillae to anteriormost point on lower border of foramen mag¬ num; zygomatic breadth, greatest distance across zy¬ gomatic arches, perpendicular to longitudinal axis of cranium; mastoid breadth, greatest distance across the skull, including the mastoids; least interorbital breadth, least distance across frontal bones; length of nasals, greatest distance between margins of paired nasal bones from posterionnost projection of frontal suture to tip; breadth of rostrum, greatest width of rostrum across in¬ cisive capsules; length of diastema, distance between posterolateral margin of incisive alveolus and anteromedial margin of alveolus of PI; length of maxil¬ lary toothrow, distance from anterior margin of alveo¬ lus of PI and posterior margin of alveolus of M3; length of bulla, greatest length of bulla; width of bulla, great¬ est width of bulla; width of zygomatic plate, width dor¬ sal to insertion of zygomata on skull, parallel to longi¬ tudinal axis of skull; length of mandibular toothrow, distance from anterior margin of alveolus of pi and posterior margin of alveolus of m3; lengtli of mandible, distance between the median margin of alveolus of il and mandibular condyle. Means, standard deviations, and ranges were cal¬ culated for adult specimens of each species. Unpaired comparisons of the means (^-test) for each character for each species were performed using StatView (SAS In¬ stitute Inc., 1998), Means were considered statistically significant at P < 0.05. NTSYSpc (Numerical Tax¬ onomy and Multivariate Analysis System, version 2,0; Rohlf, 1998) was used to generate a principal compo¬ nents analysis to evaluate the relationships among the taxa. Before ordination, characters were standardized (character means = 0, character variance =1). Analy¬ ses were run using mean values for each taxon. A series of 22 discrete-state characters was used for the cladistic analysis of the specimens listed in the Referred specimens and Appendix I. Values for the char¬ acters are given in Table 1. Character descriptions are given in Appendix 11. We used the Phylogenetic Analysis Using Parsi¬ mony program (PAUP 4.0 beta 2; Swofford, 1998) for parsimony analysis. The branch-and-bound algorithm was used, characters were unordered, zero-length branches were collapsed, the maximum number of mini¬ mum-length trees was set at 1000, and the simple addi¬ tion sequence w'as selected. The branch-swapping al¬ gorithm used was tree bisection-reconnection. Ctenomys was used as the outgroup taxon based on studies that indicate that this genus (tribe Ctenomyini) is the sister group of the tribe Octodontini (Kohler et al., 2000; McKenna and Bell, 1997). Results Pipanacoctomys aureus gen. et sp. nov. Type species, —^New species described herein. Included species. —Only the type species. Holotype, —^Adult female; skin, skull, skeleton, and tissues; CML 6137 (Coleccion de Mamiferos Lillo, Universidad Nacional de Tuciiman); collected 6 Octo¬ ber 1998 by M. A. Mares, J. K. Braun, R. M. Barquez, and M. M. Diaz; field number Arg 4915. Referred specimens (Paratypes).—Arg 4916, 4922, 4966, 5045, 5046, 5085. Type locality. —Argentina: Catamarca Province: Departamento Poman: 28 km S, 9.3 km W Andalgala, IT 50’ 03” S, 66° 15’ 50” W; elevation 680 m. Distribution. —Known only from the type local¬ ity in Catamarca Province, Argentina (Figs. 2 and 3). Etymology. — Pipanaco: for the Salarde Pipanaco (Pipanaco Salt Pan) in Catamarca Province, Argentina; octo (Latin): eight, referring to the cheek teeth, which are in the form of the numeral “8”; mys (Greek): mouse; aureus (Latin): gold, for the golden coloration. Com¬ mon name: Golden Vizcacha Rat. 4 Occasional Papers, Museum of Texas Tech University Figure 2. Map of northern Argentina showing locali¬ ties for Pipanacoctomys aureus (closed square) and Salinockmiys loschalchalerosorum (open square). The localities for Tympanoctomys barrerae in Mendoza and La Pampa provinces are shown by closed circles. Diagnosis (Figs, 4-6; Tables 1-4).—A member of the tribe Octodontini, subfamily Octodontinae, fam¬ ily Octodontidae. Size medium to large for the subfam¬ ily; pelage pale golden blond dorsally and white to pale creamy white ventrally; tail long, 76-85% head-body length, with well-developed rufous tuft; pinna about 12% head-body length; hind foot about 22% head-body length; bristle bundles present, but not well developed, and soft. The cranium differs from that of all other octodontids in having the following combination of char¬ acters: distinct gi'oove and flange for infraorbital ner\^es; postorbital process or protuberance located on the fron¬ tal; palate long, extending beyond the postenor border of M2; paraoccipital process wide, flattened, and com¬ pletely adpressed to the bulla; posterolateral border of the incisive foramen raised, which in lateral view ap¬ pears as a protuberance anterior to PI; angle of suture between the jugal and zygomatic process of the maxil¬ lary sharply acute; interpremaxillary foramina small; bulla large and inflated; area anterior to bulla with tw'o foramina, posterior opening of the alisphenoid canal and Figure 3. Saline habitat of Pipanacoctomys aureus showing burrows placed under Heterostachys ritteriana. Sparse smaller shrubs are evident grading across the salty soil to the saline flat in the background. foramen ovale (bony piece separating the two may be lost during specimen preparation); upper incisors light orange with naiTOw white lateral edge; and M3 with metacone and hypocone well developed fomning a tri¬ angular-shaped posterior extension. Description. —Size medium to large (Figs. 4-6; Tables 1-4), total length 298-318 mm; length of head and body 169-178 mm; length of tail 129-145 mm; length of hind foot 37-40 mm; length of ear 20-22 mm. General external characters as for the tribe. Tail long (76-85% head-body length), with pronounced tuft ex¬ tending 35-40 mm beyond tip; tail well haired, scales not visible; pinna medium for subfamily (about 12% head-body length); pinna sparsely haired but outer and inner surfaces covered with fringe of whitish hairs; buc¬ cal bristles present, but not well developed, hairs soft; fore- and hind feet covered with whitish hairs; soles of hind foot naked with six well-developed plantar pads; hind foot with fringes of hairs medially and laterally (well developed); stiff bristle of hairs present above the claws of digits of hind fool. Coloration of the holotype follows terms proposed by Ridgway (1912). Overall dorsal coloration pale blond (see cover photo and Fig. 4). Dorsal hairs (about 21 mm in length) are Neutral Gray basally (10 mm), followed by a band between Chamois and Cream-Buff, and a dark tenninal tip (about 1 mm). Guard hairs (up Mares et al.— Two New Genera of Mammals 5 Figure 4. Photo of Pipanacoctomys aureus. to 26 nim) are dark overall and are about the same color throughout their length. A distinct lateral line is absent, altliough the dorsal coloration grades gradually into the ventral coloration on the sides, which are therefore paler than the dorsum. Overall ventral coloration is whitish or pale cream. Hairs of the chest and throat lack gray bases. Hairs of the belly (about 11 mm) have a basal Pale Neutral Gray band (about 3 mm). Proximal one- third of tail bicolored. Clay above and Light Buff be¬ low'. Distal two-tliirds of tail between Cinnamon-Brown and Ochraceous-Tawny both above and below. General cranial and denial characters as for the tribe. Skull (Fig. 5) medium in size; nasals short, not extending beyond fronto-premaxillary suture; nasal width gradually tapering posteriorly; tips of nasals about equal to gnathic process; interorbital region divergent, ledges and beading present; width of interorbital region greater than width of rostrum; zygomatic arches short and slightly convergent anteriorly; lacrimal small, well developed, and rounded; angle of fronto-parietal suture broadly acute or rounded; protuberance present at fronto-squamosal suture ; protuberance present at squa¬ mosal-parietal suture; mastoid island large, length and width about 5.4 mm; area between mastoid islands broad, width > 8.5 mm; posterior border of interpari¬ etal nearly a straight line; hamular process of squamo¬ sal narrow. In lateral view, the following characters are vis¬ ible, Diastema steeply angled downward from incisor to premolar; posterolateral border of incisive foramen raised, which appears as a protuberance anterior to PI; infraorbital canal with distinct groove and flange for Figure 5. Dorsal, ventral, and lateral view of the skull, lateral view of the mandible, and upper and lower left toothrows of Pipafiacoctomys aureus. the infraorbital ramus of the trigeminal neiwe; angle of suture between jugal and zygomatic process of maxil¬ lary .sharply acute; jugal thick, especially at suture be¬ tween jugal and zygomatic process of maxillary; postglenoid fossa long and narrow. Ventrally (Fig. 5), interpremaxillaiy' foramen an¬ terior to incisive foramina small and round or oval; pala- 6 Occasional Papers, Museum of Texas Tech University tal process of premaxilla short and oblong, the midline raised, and separate from the palatal process of max¬ illa; pair of foramina present at posterior border within incisive foramina; palate long, extending beyond the posterior border of M2; palate with numerous small fo¬ ramina and fragile, tliin plates extending upward from the palate; midline of palate a raised thin plate; pair of posterior palatal foramina between PI and Ml; poste¬ rior margin of hard palate an inverted open “V” shape and without processes; small foramen posterior to M3; presphenoid and basisphenoid broad; pterygoid hamu¬ lus long, thick, and completely adpressed to auditory bulla; pterygoid with distinct groove laterally; auditory bulla large and inflated; foramen ovale and posterior opening of the alisphenoid canal located anterior to the bulla; paraoccipital process wide, flattened, and com¬ pletely adpressed to bulla. Upper incisors light orange with narrow white lateral edges; opisthodont; P1-M2 “8”- shaped, edges rounded; PI slightly narrower in width than Ml-M3; paracone, metacone, protocone, and hypocone of PI about equal in size, metaflexus about equal to hypoflexus; metaflexus of PI and Ml directed slightly posteriorly; paracone and metacone of Ml and M2 slightly smaller than protocone and hypocone, metaflexus shallower than hypoflexus; Ml and M2 about equal in width, size, and shape; paracone and pro¬ tocone of M3 well developed, metacone and hypocone not equal in size to fomier, but with triangular-shaped posterior extension. Lower incisors light orange, thin (width of single incisor, 1.35-1.4 mm); pi slightly narrower than ml and m2, shape of hypoconid and entoconid like that of m 1 and m2, metaconid and protoconid and procingulum fonning a triangle with a rounded apex, hypoflexid di¬ rected anteriorly and slightly larger than mesoflexid, metaconid with slight indentation; ml and m2 “8”- shaped, major conids present, well developed, and about equal in size and shape with rounded edges, hypoflexid directed anteriorly; m3 “T” or “C”-shaped, protoconid and hypoconid present and well developed, entoflexid small or absent, entoconid small or absent, metaconid . present or absent. Angular (lunar) notch deep; coronoid process small; angular process of mandible thin and not flat¬ tened. Thyrohyal bone of hyoid long and narrow; entoglossal process of hyoid small and narrow; atlas with large, thin ventral tubercle 2.5-3.0 mm in length, transverse process large and wing-like with two large foramina on each side, one foramen on each side of the neural arch; axis with neural spine large and extending posteriorly beyond third cervdeal vertebra; 13 ribs, the last very small; 28 caudal vertebrae. Habitat. —The habitat (Fig. 3) consists of perisaline shrublands associated with the Salar de Pipanaco. Low (generally less than 1 m) chenopodiaceous shrubs {Heterostachys sp., A triplex lampa, Suaeda divaricata) dominate the area and the substrate is mainly sand with significant amounts of salt. The animals are restricted to a narrow band of halo- phytic plant habitat that lies between the bare salt Oat and the more Pyq^ical Monte Desert habitats of Prosopis, Larrea, and the other perennial trees and shrubs of the Monte (Mares etal., 1985). Specimens examined (7).—ARGENTINA: Catamarca Province: Departamento Poman: Establecimiento Rio Blanco, 28 km S, 9.3 km W Andalgala, 27°50’ S, 66° 16* W; elevation 680 ± 61 m (holotype CML 6137, paratypes Arg 4916,4922,4966, 5045, 5046, 5085). Salinoctomys loschalchalerosorum gen. et sp, nov. Type species. —New speeies described herein. Included species. —Only the type species. Holotype. —Adult male; skin, skull, skeleton, and tissues; CML 3695 (Coleccion do Mamiferos Lillo, Universidad Nacional deTucuman); collected 25 .July 2000 by M. A. Mares; field number Arg 5111. Referred specimens (Paratype).—Arg 5121. Type locality. —Argentina: La Rioja Province: Departamento Chamical: 26 km SW Quimilo, 30° 02’ 43.4” S, 65°3r 13.4” W; elevation 581 m. Distribution. —Known only from the type local¬ ity in La Rioja Province, Argentina (Figs. 1,2, and 7). Figure 6. Dorsal and ventral views of the skulls and left upper toothrow of: A) Octodon degus\ B) Octodontomys gliroides\ C) Tympanoctomys barrerae, D) Pipanacoctomys aureus, now species (holotype CML 6137); H) Octomys mimax; and F) Salhwctomys loschalchalerosorum, new species (holotype CML 3695). Mares et al.— Two New Gener,\ of Mammals 8 Occasional Papers, Museum of Texas Tech University Etyinology. — Salin: for the Salinas Grandes (Great Salt Pan) in La Rioja Province, Argentina; octo (Latin): eight, referring to the cheek teeth, which are in the form of the numeral ‘‘8”; mys (Greek): mouse; loschalchalerosorum-. named for the great Argentine folklore group, “Los Chalchaleros,” in honor of their 52 years singing the traditional music of western Ar¬ gentina, its habitats, and its histoiy. Common name: Chalchalero Vizcacha Rat. Diagnosis (Figs. 6, 8, and 9; Tables 1-4).—A member of the tribe Octodontini, subfamily Octodontinae, family Octodontidae. Size medium for the subfamily; pelage brownish black dorsally and white ventrally; tail long, 76-77% head-body length, with black tuft; pinnae about 11% head-body length; hind foot about 19% head-body length; bristle bundles present, but not well developed, hairs soft. Tlie cranium differs from that of all other octodontids in having the following combination of characters: distinct groove and small flange for infraorbital nerves present; intcrpremaxillary foramen large; supraoccipital with distinct medial crest; palate short, extending to the middle of M2; paraoccipital process wide, flattened, and completely adpressed to the bulla; angle of suture be¬ tween the jugal and zygomatic process of the maxillary sharply acute; distinct “knob” present posterior to su¬ ture between the jugal and zygomatic process of the maxillary; bulla large and inflated; posterior opening of the alisphenoid canal and foramen ovale present, the latter displaced posteriorly toward the bulla; foramen magnum with a distinct downward orientation; upper incisors light orange with narrow white lateral edge; and M3 with metacone and hypocone well developed laterally. Description, —Size medium (Figs. 6, 8, and 9; Tables 1-4), total length 255, 275 mm; length of head and body 156, 144 mm; length of tail 119, 111 mm; length ofhind foot 30,28 mni; length of ear 17,16 mm. General external characters as for the tribe. Tail long (76-77% head-body length), with tuft extending 20-25 mm beyond tip; tail well haired, scales not visible; pin¬ nae small for subfamily (about 11 % head-body Icngtli); pinna sparsely haired, with inner surfaces covered with fringe of whitish hairs and outer surfaces covered with whitish hairs with brownish tips; buccal bristles present, but not well developed, hairs soft; fore- and hind feet Figure 7. Saline habitat of Sa/inoctomys losch a ich alerosoru m . covered with whitish to pale brownish hairs; soles of hind feet naked with six well-developed plantar pads; hind feet with fringes of hairs medially and laterally (well developed); stiff bristle of hairs present above the claws of digits ofhind feet. Coloration of the holotype follows terms proposed by Ridgway (1912). Overall dorsal coloration between Drab and Hair Brown with Fuscous Black intermixed. Figure 8. Photo of Salinoctomys loschalchalerosorum. Dorsal hairs (about 20 mm in length) are Deep Neutral Gray basally (about 13 mm), followed by a band of Wood Brown (about 5 mm), and with a Fuscous Black tip (about 2 mm), Guard hairs (about 20 mm) arc dark overall and are about the same color throughout their length. A lateral line is absent. Overall ventral colora¬ tion is wliite. Some scattered hairs of the chest and tliroat have Pale Neutral Gray bases. Hairs of the belly (about 10 mm) are white to the base on one specimen and have a basal light gray band in the second specimen. Proxi¬ mal one-third to one-half of the tail is bicolorcd, Wood Brown to Fuscous Black above and whitish below. Dis¬ tal one-half to two-thirds of tail Fuscous Black above and below. Mares et al.— Two New Gener.4 of Mammals 9 General cranial and dental characters as for the tribe. Skull (Fig. 9) small to medium in size; nasals short, not extending beyond fronto-premaxillary suture; nasal width gradually tapering posteriorly; tips of nasals about equal to gnathic process; interorbital region divergent, ledges and beading present; width of interorbital region greater than width of rostrum; zygomatic arches short and only slightly convergent anteriorly; lacrimal small, well developed, and elongate; angle of fronto-panetal suture broadly acute or rounded; protuberance present at fronto-squamosal suture; protuberance present at squamosal-parietal suture; mastoid island large, length and width about 6.3 mm; area between mastoid islands narrow, <8,5 mm; posterior border of interparietal rounded; hamular process of squamosal narrow; su- praoccipital with marked medial crest. In lateral view, the following characters are vis¬ ible. Diastema angled downward from incisor to pre- molar; infraorbital canal with distinct groove and small flange for the infraorbital ramus of the trigeminal nerve; angle of suture between jugal and zygomatic process of maxillary sharply acute; distinct “knob” present poste¬ rior to suture between the jugal and zygomatic process of the maxillary; jugal thick, especially at suture be¬ tween jugal and zygomatic process of maxillary; postglenoid fossa long and narrow. Ventrally (Fig. 9), interpremaxillary foramen an¬ terior to incisive foramina large and oval; palatal pro¬ cess of premaxilla short and oblong, and separate from the palatal process of maxilla; pair of foramina located behind the posterior border of the incisive foramina; palate short, extending to the middle of M2; palate with numerous small foramina and thin plates extending up- w'ard from the palate; midline of palate raised from middle to M2 to posterior border of palate; pair of small posterior palatal foramina located about the middle of M2, either of which may be obsolete; posterior margin of hard palate an inverted open “V” shape and without processes; small foramen posterior to M3; presphenoid and basisphenoid broad; pterygoid hamulus short, thick, and incompletely adpressed to auditory bulla; auditory bulla large and inflated; area anterior to bulla with two foramina, posterior opening of the alisphenoid canal and foramen ovale, the latter displaced posteriorly toward the bulla; pterygoid with distinct groove laterally; paraoccipital process wide, flattened, and completely adpressed to bulla. Figure 9. Dorsal, ventral, and lateral view of the skull, lateral view of the mandible, and upper and low^er left tootlirows oi Salinoctomys loschalchalerosorum. Upper incisors orange wdth narrow white lateral edges; opisthodont; PI-M3 “8”-shapcd, edges rounded; PI slightly narrower in width than M1-M3; paraconc, metacone, protocone, and hypocone of PI about equal in size, metaflexus about equal to hypoflexus; metaflexus of PI directed slightly posteriorly; paracone and metacone of Ml and M2 slightly smaller than pro¬ tocone and h 3 /poconc, metaflexus slightly shallower tlian 10 Occasional Papers, Museum of Texas Tech University Table 1. Values of discrete-state characters for taxa examined. See Appendix U for descriptions and states of characters. Genera are listed alphabetically. Taxa Characters 1 2 3 4 5 67 8 9 1 0 11111 1 2 3 4 5 11112 6 7 8 9 0 2 2 1 2 Aeon aemys fuscus 0000 1 0 0 7 0 1 110 0 0 7 0 0 0 0 1 0 Ctenotnys mendocinus 000 1 1 10010 10000 0 1001 1 0 Octodon bridges i 1 0000 0000 1 1100? 7 ? 1 0 0 1 0 Octodon degus 10010 0000 1 11000 00 100 1 0 Octodon lunatus 1 0 070 0000 1 1 7 0 0? 77 100 1 0 Octodontomys gliroides 2 1000 000 1 1 0 1100 10 10 1 1 0 Octomys mimax 20000 000 1 1 0 0 2 0 0 10 100 11 Pipanacoctomys aureus 2 1112 1112 1 002 1 1 2)110 2 1 Salinoctomys loschalchalerosorum 1110 1 10 12 1 002 1 1 2 1110 2 1 Spalacopus cyanus 00000 0 0 7 0 0 20000 7 0 10 2 0 1 Tympanoctomys barrerae 11201 1 0 2 2 0 002 1 1 2 1110 2 1 hypoflexus; Ml and M2 about equal in width, size, and shape; paracone and protocone of M3 well developed, metacone and hypocone slightly smaller than former. Lower incisors orange and thin (width of single incisor, 1.5 mm); pi slightly narrower than ml and m2, shape of hypoconid and entoconid like that of ml and m2, metaconid and protoconid and procingulum form¬ ing a triangle with a rounded apex, small anteroflexid present, hypoflexid directed anteriorly and about equal to mesotlexid, metaconid with slight indentation; ml and m2 “8”-shaped, major conids present, well devel¬ oped, and about equal in size and shape with rounded edges, hypoflexid directed anteriorly; hypoconid and entoconid of m2 smaller than protoconid and metaconid; m3 “C”-shaped, protoconid and hypoconid present and well developed, metaflexid small or absent, entoconid absent, metaconid present. Angular (lunar) notch deep; coronoid process small; angular process of mandible thin and not flat¬ tened. Thyrohyal bone of hyoid long and narrow; entoglossal process of hyoid a large, thick, well-devel¬ oped knob; atlas with large, thin ventral tubercle 2.1 mm in length, transverse process large and wing-like with two large foramina on each side, two foramina lo¬ cated on each side of the neural arch, height of dorsal process less than that of transverse process; axis with neural spine large, which extends posteriorly to just anterior of the third cervical vertebra; 13 ribs, the last very small; 25 caudal vertebrae. Habitat .—The habitat consists of perisaline shrublands associated with the Salinas Grandes along the edges of thorn scrub islands that rise within the salt flat and provide dense thorn scrub on sandy substrate. Plants associated with the burrows are generally less than 1 m in height, with the exception of tall cacti {Cereus sp.) and occasional trees, and include the chenopodiaceous shrub Heterostachys ritterlana, as well as the chenopodes {Allenrolfea patagonica and Atriplex argentina). Other plants on the site were Montlea aphylla (Scrophulariaceae), Mimroa argentina (Gramineae), Aphyllocladus spartioides (Aster- aceae), Lycium tenuispinosum and Grabowskia boheraviaefolia (Solanaceae), Cortesia cuneifolia (Boraginaceae), Senna aphylla (Fabaceae), Agrostis pyramidalis and Setaria sp. (Gramineae), Alternanthera nodifera (Amaranthaceae), and Prosopis strombulifera (Leguminosae). Salinoctomys is re¬ stricted to a narrow band of halophytic plant habitat that lies between the bare salt flat and the dense thorn scrub habitat of cacti, trees, and shrubs that is typical of the Gran Chaco (Mares et al., 1985). The most abun¬ dant plant in the immediate area of the burrow system is Heterostachys ritteriana. Mares et al.— Two New Genera of Mammals 11 Specimens examined (2).—ARGENTINA: La Rioja Province: Departamento Chamical: 26 km SW Quimilo, 30° 02’ 43.4” S, 65° 31’ 13.4” W; elevation 581 m (holotype CML 3695, paratype Arg 5121). Comparisons A comparison (Tables 1-4; Fig. 6) of Pipanacoctomys and Salinoctomys with other octodontid rodents follows. Both taxa differ from all other octodontid rodents in coloration. Aconaemys and Spalacopus have the following characteristics that dis¬ tinguish them from Pipanacoctomys and Salinoctomys: tail short and untufted; bristle bundles absent; interor¬ bital region narrow, less than width of rostrum; bulla small and iminflated; paraoccipital process small, thick,and fused incompletely to the bulla; angular (lu¬ nar) notch shallow; postglenoid fossa small and round; lacrimal large and triangular-shaped; metacones and hypocones not well developed; postorbital process ab¬ sent (like Salinoctomys y but unlike Pipanacoctomys)'-, paired foramina at the posterior border of the incisive foramina absent; raised lateral edges at the posterior border of the incisive foramina absent (like Salinoctomys, but unlike Pipanacoctomys). Aconaemys is distinguished further by: upper in¬ cisors solid orange; and zygomatic arches parallel-sided. Spalacopus further is distinguished by: dorsal coloration brownish black; ears greatly reduced; upper incisors white and proodont; zygomatic arches ex¬ panded; and distinct groove for infraorbital ramus of the trigeminal nerv'e in the infraorbital canal absent. Octodon has the following characteristics that distinguish it from Pipanacoctomys 2 X\& Salinoctomys: dorsal coloration brownish or grayish (like Salinoctomys, but unlike Pipanacoctomys)', tail color is similai' to that of the dorsum; tail with small tuft (like Salinoctomys, but unlike Pipanacoctomys)\ bristle bundles absent; upper incisors solid orange; zygomatic arches parallel-sided; interorbital region narrow; pos¬ torbital process present (like Pipanacoctomys, but un¬ like Salinoctomys)', shape of suture of jugal and zygo¬ matic process of squamosal squarish; lacrimal large and triangular; paired foramina at the postenor border of the incisive foramina absent; raised lateral edges at the posterior border of tlie incisive foramina absent (like Salinoctomys, but Pipanacoctomys)', bulla small and uninflated; paraoccipital process small, thick, and fused incompletely to the bulla; postglenoid fossa small and round; area anterior to bulla with three foramina, posterior opening of the alisphenoid canal, foramen ovale, and middle lacerate foramina sometimes divided; molariform teeth crescent-shaped; metacones and hypocones not well developed; angular (lunar) notch shallow; angular process flattened. Octodontomys has the following characteristics that distinguish it from Pipanacoctomys and Salinoctomys: dorsal coloration gi'ayish brown; tail with a reddish buffy brush; bristle bundles absent; upper in¬ cisors solid orange; postorbital process absent or re¬ duced (like Salinoctomys, but unlike Pipanacoctomys)] interorbital region narrow, width about equal to the width of the rostrum; shape of suture ofjugal and zygo¬ matic process of squamosal squarish or rounded; lacri¬ mal large and triangular; palate short, an inverted “V” shape, not extending beyond Ml; paired foramina at the posterior border of the incisive foramina absent; raised lateral edges at the posterior border of the inci¬ sive foramina absent (like Salinoctomys, but unlike Pipanacoctomys)', bulla medium and slightly inflated; paraoccipital process small, thick, and incompletely fused to the bulla; postglenoid fossa small and round; hamular process of squamosal well developed; (wo large foramina present, foramen ovale and posterior opening of the alisphenoid canal, middle lacerate foramina tiny; upper incisors orthodont; molariform teeth crescent- shaped; metacones and hypocones not well developed: angular (lunar) notch shallow; angular process shon. slightly flattened. Octomys has the following characicrisiics that distinguish it from Pipanacoctomys and Salinoctomys: dorsal coloration pale tan; tail with well-developed bmsh similar in color to that of the dorsum; bristle bundles absent; upper incisors solid orange; palate short, not extending beyond the posterior border of Ml; paired foramina at the posterior border of the incisix e foramina 12 OccAsroiNAL Papers, Museum of Texas Tech University Table 2, External and cranial measurements (in millimeters) for Pipanacoclomys aureus, and Salinoctomys loschalchalerosorum. Included are mean, standard deviation, sample size, and range. Significant differences in means (X-test\ P < 0.05) are indicated as follows'. Tipanacoctomys v.v. Tympanoctomys; '^Salinoctomys vs. Tympanoctomysi^Pipanacoctomys us. Salinoctomys. Taxa Character Pipanacoctomys aureus Salinoctomys loschalch alerosorum Holotype CML6137 All adults Holotype CML3695 Paratype Arg 5! 21 Total length^’ 315 306+ 12,4 (293-318) 275 255 Length of head and body*-* 170 170 + 5,4 (166- 178) 156 144 Length of taiP 145 135 + 9,4 (127- 145) 119 111 Length of hind foot*"-' 37 37 + 3,4 (32 - 40) 30 28 Length of ear^ '" 22 20 + 1,4 (19-22) 17 16 Greatest length of skull® 43.1 42.6+1.6, 3 (40.8 - 43.9) 39.2 Basal length^ 35.1 37,1 ±2.7, 3 (35.1 -40.2) 36.6 Zygomatic breadth® ' 20.7 20.9 + 0.7,3 (20.3-21.6) 19.5 18,8 Mastoid breadth 23.6 22.3 + 2.6, 3 (19.3-24.1) 20.1 Least interorbilal breadth® 9.0 8.9 + 0.4, 3 (84-9.3) 8.4 7.5 Length of nasals' 13.7 13.8 + 0.1,3 (13.7-13.9) 14.2 14,0 Breadth of rostrum®’' 6.6 67 + 0.1,3 (6.6-6.8) 6.3 6.2 Length of diastema 8.S 8,7 + 0.3, 3 (8.4-9.0) 7.9 8.2 Length of maxillary toothro’w® 6.7 7.1 +0.4, 3 (6.7-7.5) 6.4 5.5 Length of bulla®' 19.2 19,2 + 0.1,3 (19.1 - 19.3) IS.O 17.0 Width of bulla 11.5 11.5 + 0.3,3 (11.2-11.8) 10.7 11.7 Width of zygomatic plate® 1.7 1.6+ 0.1, 3 (1.5-L7) L3 1.3 Length of mandibular toothrow® 6.6 6.8+ 0.2, 3 (6.6- 7.1) 6.6 5.8 Length of mandible 21.2 21.5 + 1.0,3 (20.7 - 22.7) 20,6 19.8 absent; raised lateral edges at the posterior border of the incisive foramina absent (like Salinoctomys, but unlike Pipanacoctomys)\ postorbital process absent (like Salinoctomys, but unlike Pipanacoctomys)\ suture of jugal and zygomatic process of squamosal squarish; paraoccipital process small and incompletely fused to Mares et al.— Two New Genera of Mammals n Table 3, External and cranial measurements (in millimeters) for Tympanoctomys barrerae, Octomys mimax, Octodontomys gliroides, and Aconaemys fuscus. Included are mean, standard deviation, sample size, and range. Significant differences in means (i-test\ P < 0.05) are indicated as follows'. ""Pipanacoctomys V5. Tympanoctomys; '^Salinoctomys vs. Tympanoctomys. Character Taxa Tympanoctomys barrerae Octomys mimax Octodontomys gliroides Aconaemys fuscus Total length* 262 -h 26, 8 369 + 73, 13 324+15. 29 259+ 11,2 (232 - 297) (283 - 508) (301-361) (252 - 267) Length of head and body* 145+ I], 8 212 + 73, 13 176 + 8.5, 29 180+ U, 2 (125 - 160) (140-330) (159- 190) (172 -187) Length of tail 118 ± 18,8 157 + 20, 13 149+11,29 63 + 29, 3 (93 - 147) (105-179) (115-177) (30- 80) Length of hind fooP 34 + 2, 8 34,7+1.3, 13 35.8+ 1.6,30 31+9,4 (30-36) (32.3 - 36.6) (31-4-39.0) (18-37) Length of ear* 16.3+2.2, 8 24.1+2.2, 13 28.2 + 2.3,30 20 + 2, 3 (130-19.4) (21.4-27.4) (21.1-32,0) (18-22) Greatest length of skull* 37.9 + 2.1,3 43.7 + 1.1, 9 453 ± 13,25 43.2 + 0.4,2 (36.1-40.2) (41.8-45.5) (42.4 - 47-7) (42.9 - 43.4) Basal length* 27.8 +O-L 2 37.2 + 1.2,8 39.6+1.3,22 38.6 + 0.4,2 (29.7 -29.8) (35.1- 38.5) (37.3 - 42.4) (383 - 38-9) Zygomatic breadth* 18.8+ 0.7, 3 21.8 + 0.6, 10 233 + 0.7, 26 24.8 + 0.3, 2 (18.1 - 19.3) (20.8 - 22.6) (22.0 - 24.4) (24.6 - 25.0) Mastoid breadth 22.3 + 1.0, 3 21.6 + 0.6, 8 22.1 +0.7,21 183 ±0.4, 2 (21.1 -23.0) (21.0 - 22.7) (21.1 -23.1) (18.0-18.6) Least interorbital breadth* 7.7 + 0.6, 3 10.0 + 0.6, 12 9.41 0.4, 29 8.0+ 0.2, 3 (7.1 -8.2) (9.1 - 11.2) (8-6- 10.2) (7.9 - 8.2) Length of nasals 11.3, 1 15.4 + 0.6, 10 16.6 + 0.9, 26 16,2+ 1.2,4 (LI - 16.4) (13.0”- 17.5) (15.1 - 18.0) Breadth of rostrum* 63 + 0.1,2 7.6 + 03, 11 8.2 + 0.3, 26 9,0+ 0.8, 4 (6.2-6.4) (7.1-8,1) (7,7- 8,9) (7.8”- 9.5) Length of diastema 7.9 + 0.5, 3 9.9 +0.6, 12 11.5 + 0.5, 28 11.0+1.2,5 (7.5-85) (8.7-10.9) (10.7- 12.5) (9.7- 12.5) Length of maxillary toothrow* 5.9+ 0.1,4 8.1 + 0.4, 9 8,1 +0.6, 28 8.5+ 0.7, 5 (5,8-60) (7.4-8.5) (7.4-9.6) (7,9-9.7) Length of bulla* 17.2 + 0.9,3 15.8 + 03, 9 14-6 + 0.7, 23 111 +0.1,2 (16.3- 18.1) (15.3-163) (13.1 - 15.6) (11.0-1 LI) Width of bulla 11.6 + 0.2, 3 11.410,3, 9 n. 3 + 0.3, 23 8.9 + 0.1,2 (1 1.4- 11.7) (10.9-11.8) (10.7-11.9) (8,8-90) Width of zygomatic plate* 1.1 +0,2, 2 L6 + 0.1, 11 1.9 + 0.2, 26 2,2+ 0.3, 5 (0.9- 1.2) (1.5-1.8) (1.3-2.3) (1.8-2.5) Length of mandibular toothrow* 6.0 + 0.4, 3 8.2 1 0.4, 8 8.5 + 0.6, 27 9,5+ 0,7, 5 (5.7 - 6.4) (7,5-8.9) (7,9- 10-1) (8.9- 10.6) Length of mandible 20.0 + 0.6, 4 24.9 + 0.8, 9 27.5 + 0.9, 28 26.7 + 1.7, 3 (19.1-20.6) (23.8 - 25.9) (26.0 - 29.5) (24.8 - 28.1) the bulla; postglenoid fossa small and round; two large foramina present, foramen ovale and posterior opening of the alisphenoid canal, middle lacerate foramina tiny; molariform teeth angular “8”-shaped; metacones and hypocones not well developed; angular (lunar) notch shallow; angular process slightly flattened. 14 Occasional Papers, Museum of Texas Tech University Table 4. External and cranial measurements (in millmeters) for Octodon bridgesi, Octodon degus, Octodon lunatus, and Spalacopus cyanus. Included are mean, standard deviation, sample size, and range. Character Taxa Octodon bridgesi Octodon degus Octodon lunatus Spalacopus cyanus Total length 343+ 24, 2 312 + 23, 9 349 + 23. 9 187+18, 8 (325 - 359) (272-341) (318 - 398) (147 - 203) Length of head and body 191 + 15, 2 181+29, 9 195 ± 15, 9 138± 13, 8 (181 -202) (115-212) (168 - 221) (112- 152) Length of tail 151 ±9,2 130+11,9 153 + 10, 9 50 ±9, 8 (144-157) (120- 157) (145 - 177) (35 - 66) Length of hind foot 38 + 2,4 36 ±2, 9 37± 1,9 29 ± 3, 9 (36-40) (34 - 39) (35 - 39) (25 - 32) Length of ear 25 + 2. 4 28 + 3,7 28 + 2, 8 11 + 1, 9 (23 - 27) (23 - 30) (24 - 29) (9-13) Greatest length of skull 44.7 + 2.6, 4 42.6 + 2.0, 9 46.1 + 1.4, 8 39.5 + 2.7, 7 (40.9 - 46.7) (39.8 -46.5) (44.3 -48.2) (36.9 - 44,2) Basal length 39-2 + 0.9,4 36.4 ± L7, 9 32.3 ±1.5, 8 35.6 + 2.8, 7 (38.3 - 40.2) (33.9 - 39.0) (36.4-41.3) (33.0 - 40.8) Zygomatic breadth 24.2 + 0.6, 3 22.8 ± 0.7, 9 24.5 ± .07, 7 23.4 + 0.8, 6 (23.7 - 24.9) (21.7-23.5) (23.4 - 25.3) (22.8 - 24,9} Mastoid breadth 21.3 ±0.3, 4 17.7 ±3.3, 3 20.8 + 0,2, 8 I7.8± 1.2, 8 (21.0-21.7) (13.9-20.1) (20.5 - 21.1) (16.5-19.4) Least interorbital breadth 9.1 ±0.6, 5 9.6+ 0.4, 9 8.7 ±0.3, 8 8.5 ±0.5, 9 (8.5 - 9.9) (9.1 - 10.2) (8.1 -9.1) (7.9 - 9.5) Length of nasals 17,6 + 0.9,4 15.9 ±0.8, 9 17.9 ± 1.0, 7 13.2 + 0.8, 7 (17.0-18.9) (14.6 - 17.2) (16.5 - 19,3) (12,3 - 14.5) Breadth of rostrum 8.3 ±0.3, 5 7.5 ± 0.4. 9 7.9 + 0.3, 8 7.5 ± 0.7, 6 (8.0 - 8.6) (6.9 - 8.4) (7.6 - 8.5) (6.9 - 8.7) Length of diastema 10.1 ±0,5,5 9.7 ± 0.6, 9 10.3 ±0.8, 8 12,5 + i.O, 8 (9,5 - 10.6) (9,1 - 10.9) (8,7-11.2) (11.7- 14,4) Length of maxillary toothrow 9.7 ± 0.5, 5 8.9 + 0.6, 9 9.7 ±0.4, 8 7.1 ±0.6, 8 (9.0- 10.4) (8.2- 10.0) (9.2 - 10.2) (6.5 - 8,2) Length of bulla 13.3 + 0.4,4 12.9 ±0.9, 9 13.9 + 0.7, 8 10,8 ±0.8, 8 (13.0- 13.7) (11.8- 14.2) (13.1 - 15.4) (9.5 - 11-6) Width of bulla 9.9 + 0.3, 4 9,2 ±0.4, 9 10.1 +0.2, 8 7.6 ± 0.4, 8 (9.5 - 10.1) (8.6-9.7) (9.9 - 10.5) (7.2-8,5) Width of zygomatic plate 1.9+ 0.3, 4 1.6 ±0.3, 9 1.5 ±0.2, 8 1.2+ 0.3, 9 (1.5 -2.2) (1.3 -2.2) (1,0- 1.8) (0.8- 1.6) Length of mandibular toothrow 10.0 ±0.2, 4 9.3 ±0.4, 9 9.9 ± 0.3,8 7.7 ±0.5, 8 (9.8 - 10.3) (8.6- 10.0) (9.4-10.4) (7.1 - 8.3) Length of mandible 28.8 + 0.4,5 26.6 ± 1.2, 8 29.1 ± 1.1, 8 27.0 ± 1.6, 8 (28.3 - 29.3) (24,5 - 27.9) (27.1 - 30.6) (25.6 - 30,3) Tympanoctomys has the following characters that distinguish it from Pipanacoctomys and Salinoctomys: dorsal coloration tan washed with black (like Salinoctomys y but unlike Pipanacoctomys)', tail with a short, black brush (like Salinoctomys, but unlike Pipanacoctomys)', bristle bundles well developed and hairs stiff; hairs of fringe on lateral edge of hind foot long and dense; ratio of length of hind foot to length of head and body > 23% (like Pipanacoctomys, > 21%; unlike Salinoctomys, < 19.5 %); raised lateral edges Mares et al,— Two New Genera of Mammals 15 at the posterior border of the incisive foramina absent (like Salinoctomys, but unlike Pipanacoctomys)\ pos¬ terior border of interparietal nearly a straight line (like Pipanacoctomys, but unlike Salinoctomys); supraoc- cipital w^ith a medial crest (like Pipanacoctomys, but unlike Salinoctomys in which it is marked); foramen magnum with nonnal orientation (\.\k&Pipanacoctomys, unlike Salinoctomys)', distinct groove for infraorbital ramus of the trigeminal nerve in the infraorbital canal absent; postorbital process absent (like Salinoctomys, but unlike Pipanacoctomys)', shape of suture of jugal and zygomatic process of squamosal roundish or squar¬ ish; distinct “knob” absent posterior to suture between the Jugal and zygomatic process of the maxillary (like Pipanacoctomys, unlike Salinoctomys}', width between mastoid islands narrow (like Pipanacoctomys, unlike Salinoctomys)', two large foramina present anterior to the auditory bullae, the foramen ovale and posterior opening of the alisphenoid canal present; pterygoid hamulus short, thin, and incompletely adpressed to bulla (like Pipanacoctomys, but unlike Salinoctomys)', two foramina present on either side of the neural arch (like Salinoctomys, but unlike Pipanacoctomys). Salinoctomys has the following characters that distinguish it from Pipanacoctomys: dorsal coloration tan washed with black; tail with a short black brush; postorbital process absent; lacrimal elongate; infraor¬ bital canal small; interpremaxi 11 ary foramen large and oval; raised lateral edges at the posterior border of the incisive foramina absent; palate short; pterygoid hamu- lar short, not completely adpressed to auditory bulla; Table 5. Principal component loadings based on data for taxa examined. Relatively high loadings are de¬ noted with an asterisk as follows: component I > 0.70; II > 0.70: III > 0.50. Principal Component Character T 11 III Total length 0,60 0,78* 0.05 Length of head and body 0,80* 0.43 0.08 Length of tail 0.32 0.92* 0.07 Length of hind foot 0,54 0.67 0.26 Length of ear 0,78* 0,51 0.10 Greatest length of skull 0.93* 0.28 -0.06 Basal length 0.56* .0.11 -0.76* Zygomatic breadth 0.88* -0.43 0.06 Mastoid breadth -0.12 0.83* -0.10 Least intcrorbital breadth 0.63 0.36 -0.05 Length of nasals 0.93* 0-09 -0.04 Breadth of rostrum 0.86* -0.37 -0.17 Length of diastema 0.56 -0.65 -0.07 Length of maxillary toothrow 0.96* 0.00 0.22 Length of bulla -0.54 0,80* -0 17 Width of bulla -0.32 0.89* -0.24 Width of zygomatic plate 0.76* -0.04 -0.52* Mandibular toothrow length 0,95* -0.18 0.17 Length of mandible 0.90* -0.32 0.20 Eigenvalues 9.90 5.57 1.19 Total variance (87.6%) 52.1% 29.3% 6.2% foramen ovale smaller and displaced posteriorly near bulla; upper tooth row more divergent; two foramina present on each side of the neural arch of atlas. Univariate, Multivariate, and Phylogenetic Analyses Univariate analysis. —^The results of the unpaired comparisons of the means (Mesl) for each character for Pipanacoctomys, Salinoctomys, and Tympanoctomys are presented in Tables 2 and 3. Means were considered to be statistically significant at P < 0.05. For Pipanacoctomys vs. Tympanoctomys the means for the following characters were significant; total length, lengtli of head and body, lengtli of eai’, great¬ est length of skull, basal length, zygomatic breadth, least interorbital breadth, breadth of rostrum, length of max¬ illary toothrow, length of bulla, width of zygomatic plate, and length of mandibular toothrow. For Salinoctomys vs. Tympanoctomys only the mean for length of the hind foot was significant, indicating that there is little size difference between the two taxa. For Pipanacoctomys vs. Salinoctomys the means for the following charac¬ ters were significant; total length, length of head and body, length of tail, length of hind foot, length of ear, zygomatic breadth, length of nasals, breadth of rostrum, and length of bulla. 16 Occasional Papers, Museum of Texas Tech University A B - - Pipanacoctomys aureus Tympanoctomys barrerae Salinoctomys toschalchalerosorum Octomys mimax Octodontomys gliroides Ocfodon bridges! Ocfodon lunatus Octodon degus Aconaemys fuscus Spalacopus cyanus 2.00 1,75 1,50 1.25 1.00 0.7S O.SO !_^_-_J_ \ _ \ _ \ _I Figure 10. Multivariate relationships of species in the tribe Octodontini. A) Projections of scores for principal components I and II were extracted from correlations among 19 external and cranial measurements. Af = Aconaemys fuscus; Ob = Octodon bridgesi; Od = Octodon degus; 01 = Octodon lunatus; Og = Octodontomys gliroides; Om - Octomys mimax; Pa = Pipanacoctomys aureus; Sc = Salinoctomys loschalchalerosorum; Spc = Spalacopus cyanus; Tb = Tympanoctomys barrerae. B) Phenogram (unweighted pair-group method using arithmetic averages) for taxa in the tribe Octodontini based on distance coefficients (cophenetic correlation coefficient = 0.84). Multivariate analysis. —^A principal components analysis of correlations among characters (Table 5) was used to evaluate the relationships among Pipanacoctomys., Salinoctomys^ and other octodontid taxa. Eigenvalues for the first three principal compo¬ nents were 9.90,5.57, and 1.19, respectively, explain¬ ing 87.6% of the total variance (52.1, 29.3, and 6.2%, respectively). About one-half (10 of 19) of the charac¬ ters (Table 5) had high positive loadings on principal component 1; this component may be related to size, however, sample size is perhaps too small to make de¬ finitive conclusions. Component II separates taxa pri¬ marily by total length, length of tail, width of bulla, length of bulla, and mastoid breadth (loadings >0.70). Component III separated taxa primarily by basal length and width of zygomatic plate (loading >0.50). In the bivariate plot of principal components I and II (Fig. 10A), Pipanacoctomys is placed between Tympanoctomys (and Salinoctomys) and Octomys. The Mares et al.— Two New Genera of Mammals 17 51 80 66 100 62 Outgroup Aconaemys fuscus Octodon bridges! Octodon degus Octodon lunatus Octodontomys gliroides Octomys mimax Tympanoctomys barrerae Salinoctomys loschalchalerosorum Pipanacoctomys aureus Spalacopus cyanus Figure 11. Cladogram of parsimony analysis of 22 characters for 10 octodontid taxa. Ctenomys is the outgroup. The tree shown is a bootstrap 50% majority-rule consensus (49 steps, Cl = 0.653, RI = 0.730). Numbers appearing below the lines are bootstrap values. three species of Octodon are clustered closely with Octodontomys. These taxa are placed in the upper half of the plot, Aconaemys and Spalacopus are located in the lower half of the plot. A distance phenogram (Fig. lOB; r= 0.84) shows several distinct clusters. Pipanacoctomys, Salinoctomys, and Tympanoctomys form a cluster dis¬ tinct from the other octodontids, as does Spalacopus. The three species of Octodon are clustered with Octomys and Octodontomys. Aconaemys is distantly associated with this group. Phylogenetic analysis .—A branch-and-bound search using all characters for all taxa resulted in eight most parsimonious trees (46 steps). A bootstrap 50% majority-rule consensus tree (49 steps, CI=0.653, RI=0.730) is shown in Figure 11. The genera Aconaemys and Spalacopus are distinct lineages that are sister to the other members of the tribe. The rela¬ tionship of the three species of Octodon is unresolved. Sister to the Octodon clade is the relatively well supported clade formed by the remaining octodontid taxa {Octodontomys, Octomys, Pipanacoctomys, Tympanoctomys, and Salinoctomys). In this clade, Octodontomys and Octomys are the most basal taxa. The clade formed by Tympanoctomys and Salinoctomys is well supported; Pipanacoctomys is the sister taxon to this clade. Discussion The discovery of two new genera and species of salt-specialized mammals might appear surprising given the rarity of this adaptive type in the deserts of the world, but it is less unexpected when the age and isolation of the region are considered. Argentina, particularly the Northwest, supports numerous extensive salt flats, many lying within isolated valleys (see cover photo; Fig. 1) or separated from similar habitats by large areas of semi- arid thorn scrub or desert vegetation. Pipanacoctomys aureus is not the first new species described from the Bolson dePipanaco (Fig. 12). Tn 1972, Mares discov¬ ered and later described Andalgalomys olrogi (Will- 18 Occasional Papers, Museum of Texas Tech University Figure 12. Aerial photo showing details of the northern Bolson de Pipanaco, including type localities for four taxa in the northern part of the isolated valley and the town of Andalgala. The high mountains fonning the northeastern boundaries of the valley are evident. iams and Mares, 1978), which is a species endemic to the valley. Mares also found an unnamed tuco-tuco (Ctenomys sp.) in 1971 that is also endemic to the val¬ ley. Recently we found a new species of gerbil mouse {Eligmodontia sp.) in the same valley that we are in the process of describing. We are learning that not only is this particular bolson rich in new taxa, but the arid Northwest in general appears to harbor many new spe¬ cies of mammals in its numerous isolated valleys and mountains. Similarly, Salinoctomys loschalchalerosonim is not the only new species we discovered on the habitat island rising from the immense Salinas Grandes (Fig. 13). We also found what appear to be as many as three new species of mammals in known genera. The salt flats of central Argentina are difficult habitats in which to work. For much of the year they are practically inac¬ Figure 13. Aerial photo showing the placement of the Salinas Grandes, including the type locality for Salinoctomys loschalchalerosorum (box). The numer¬ ous habitat islands evident in the figure, even from a satellite, presumably will be found to support additional specimens of this new taxon. cessible, becoming areas of deep, almost bottomless, salt mud that resists vehicular and even animal travel. In the dry, cool winter, when the saline areas are more accessible, the populations of mammals are in decline. We know of few other mammal surveying expeditions that have been made to these habitats. Given the fact that they have not been studied, that they are extensive, and that they are of great age and isolation, it is perhaps to be expected that we would have found some undescribed mammals. The fact that the Northwest contains so many new taxa may result in part from the relatively recent orog¬ eny of the region. In general, the series of Andean and pre-Andean mountain chains that rim the isolated val¬ leys are fairly young, with most of the uplift having occurred since the Miocene. Indeed, significant uplift took place in the Pliocene or even as recently as the Mares et al.— Two New^ Genera of Mammals 19 Pleistocene (Pascual and Ortiz Jaureguizar, 1990). The mountains form barriers as high as 4,000 m that block prevailing winds from the east and lead to orographic rain falling on the eastern slopes of the eastern moun¬ tains, where the complex Yungas forest is found (Mares et ah, 1996). As the uplifted wind currents drop their moisture and flow over the mountaintops, the lowlands to the west of these mountain chains receive warm, dry air. Similarly, the massive Andes, whose peaks ap¬ proach or exceed 7,000 ni, also block any winds that might bring moisture from the west to the isolated val¬ leys situated to the east. Unlike the basin and range to¬ pography of the American West, where mountain chains are generally aligned along a north-south axis, the pre- Andean chains may have long axes that are oriented from east to west or north to south. The result is that in extreme northwestern Argentina a series of isolated desert valleys has been formed. The mountains them¬ selves may be isolated one from the other so that, like the valleys, the mountains become zones for speciation. Since the mountains are largely arid or semiarid, tlie result is that the biogeographic region called The Monte Desert of Mountains and Isolated Valleys is composed of a series of xeric islands at both high and low eleva¬ tions an ideal region for isolation and allopatric dif¬ ferentiation. I'he designation of the new biogeograpliic region thus reflects the unique geographical history of the area, and our data support the singular nature of the zone. In effect, the region is a land-locked, topographi¬ cally rich, terrestrial archipelago of habitat and tectonic islands that are themselves relictual mountains and val¬ leys (Figs. 1 and 12). Mammals that may have colo¬ nized the region before the period of uplift would be expected to have had their geographic ranges broken into numerous populations that would diversify over time, rhe discovery of new monotypic genera in and adjacent to the region supports this hypothesis of bio¬ geography and speciation. We recognize that one of the new genera, Salinoctomys, was found within the great salt flat of central Argentina in the Chaco Domain and outside the strict limits of the proposed biogeographic region. How¬ ever, the type locality is only 125 km distant from the mountains included in the bolson-montane biogeo- Figure 14. Satellite photo of arid western Argentina, The high mountain ranges that demarcate the Monte, Puna, and Patagonian desert valleys are evident, with the pre-Andean ranges toward the east and the main Andean Cordillera to the west. Shown are geographic- ranges of eight endemic genera. 20 Occasional Papers, Museum of Texas Tech University graphic region and the mammals of both areas shared a recent history. Our data suggest that some consideration should be given to including tlie isolated salt flats that rim the proposed new biogeographic region within the region. One must at least recognize the close associa¬ tion between some faunal elements associated with the salt flats and those of the isolated desert valleys and mountains lying slightly to the north and west. Octodontids are of great age (McKenna and Bell, 1997; Vucetich et ah, 1999) and show remarkable cliro- mosomal variation, including a species that has among the highest number of chromosomes known for a mam¬ mal (102 in Tympanoctomys barrerae; Contreras et al. 1990)—the only known tetraploid mammal, Gallardo et ah, 1999). Whether or not either of the new species is a tetraploid has not yet been determined. Phylogenetic analysis of cranial and dental mor¬ phology (Fig. 11) and molecular data (R. J. Baker and F. Hoffmann, pers, comm.) suggest that PipanacoctornySy Salinoctomys^ and Tympanoctomys shared a common ancestor not shared by any other octodontid genera. The range of the putative ancestor was fractured by the uplifting pre-Andean chains re¬ sulting in the eventual formation of two monotypic gen¬ era, one being Pipanacoctomys and the other being the common ancestor of Tympanoctomys and Salinoctomys. Eventually, tliis ancestral population was divided into a northern and southern group as the huge salt flat of the Pliocene was divided into two major salt flats through climate change and habitat disruption. This hypothesis is supported by geological evidence that indicates that prior to the end of the Pliocene a single large basin ex¬ isted in northwestern Argentina (Bailey et al., 1977; Periano, 1957). Two separate inland seas (Pascual and Jaureguizar, 1990; Vuilleumier, 1971) were found in the region: one in the far Northwest (where Salinoctomys and Pipanacoctomys occur today, and one in the south near Mendoza (where Tympa?ioctomys is found today). As these seas retreated due to changing climates and orographic uplift, the modem salt pans were formed. Associated with the salt flats that became in¬ creasingly isolated were the ancestors of today’s salt- specialized octodontids. Tympanoctomys forages on saltbush {Atriplex) in southern Mendoza, although it also consumes Pleterostachys in some parts of its range (Torres-Mura et al., 1989). Each of the three genera possesses the remarkable morphological structures (tooth-like bmshes) that are used by Tympanoctomys to strip the saltbush leaves of their protective salt covering (Mares et ah, 1997), Pipanacoctomys^ which inhabits complex mounds at the edge of the salt pan, forages on Heterostachys ritteriana. The plant has compressed leaves that fomi small, salt-filled balls. The high salt content of the Heterostachys leaves cannot be reduced via mechanical means (as can the leaves Atriplex) and the buccal bmshes of Pipanacoctomys seem to be less well-developed tlian those of Tympanoctomys. In feeding tnals, Pipanacoctomys refused to eat saltbush {Atriplex)mA another species ofchenopodiaceous halo¬ phyte {Suaeda), both of which grow near the mounds, although in much lower frequency than Salinoctomys seems to forage mainly on Heterostachys and grass, both of which were found in abundance in food stores in the burrows (pel's, obs.). Its buccal brushes are more similar to Pipanacoctomys than they are to Tympanoctomys. Unfortunately, we have been unable to capture Salinoctomys alive to detennine how they handle different food items while feeding. Tympanoctomys, Salinoctomys, and Pipanac- octomys show pronounced philopatry to the salt pan habitat (pel's, obs.). Given such unusual habitat require¬ ments, ancesti'al populations would have been unlikely to have dispersed over mountain massifs or across hun¬ dreds of kilometers of non-saline habitat after isolation ensued. Gene exchange would thus have been extremely unlikely and this would explain how the ancestral popu¬ lations diverged sufficiently to be recognized as dis¬ tinct genera after 2 million years. Recently, a monotypic genus of murid rodent {Salinomys delicatns') also was described from wood¬ lands associated with salt pans in the Monte Desert to the south of the Bolson de Pipanaco (Braun and Mares, 1995) and to the southeast of the Salinas Grandes. Salinomys, in a significant range extension, was found to co-occur with Salinoctomys, thus providing a site where two of the rarest mammals in the world occur together. Interestingly, the arid lands of Argentina com¬ prise about 2.1 million km^ (Roig et al., 1992) and sup¬ port eight endemic genera of mammals (Fig. 14; Lestodelphys, Chlamyphorus, Salinomys, Octomys, Pipanacoctomys, Salinoctomys, Tympanoctomys, and Dolichotis', Wilson and Reeder, 1993; this paper). By Mares et al,— Two New Genera of Mammals 21 comparison, the entire contemiinous United States (7.8 million km^) has only six endemic genera of mammals (Microdipodops, Ochrotomys, Podomys^ Arborimus, Neofiber, and Brachylagus; Wilson and Ruff, 1999). These results have relevance to the value of arid lands as important sources of species, genera, and even families of mammals when compared to the better- known Amazonian lowlands (Mares, 1992). The levels of endemism and higher order genetic diversity evi¬ denced by arid land mammals of South America appear to have resulted m part from the extensive geographic isolation present in southern regoins of the continent. Due to the resulting mosaic of relictual mountains, val¬ leys and saline flats—arid habitats requiring extensive levels of specialization at all levels of organization, from cells to populations—series of unique, monotypic gen¬ era developed. Acknowledgments Fieldwork was supported by grants from the Na¬ tional Science Foundation (DEB-9870184) to M. A. Mares and J. K. Braun, and a grant from CONICET to R. M. Barquez. Research also was supported by the Sam Noble Oklahoma Museum of Natural History, University of Oklahoma; PIDBA (Programa de Investigaciones de Biodiversidad Argentina); and the Facultad de Ciencias Naturales e Institute Miguel Lillo, Un 1 vers i dad Nacional deTucum^, Argentina. D. Flores assisted with aspects of the fieldwork. We thank R. J, Baker and F. Hoffmann for making the preliminai 7 re¬ sults of the cytochrome b sequence analyses available to us. We also thank the curators of the many museum collections who permitted us to examine specimens in their care. We acknowledge the support and assistance of the Direccion de Flora and Fauna Silvestres, Buenos Aires, Argentina (V. Lichtstein and S. Goldfeder) and the Direccion General de Medio Ambiente y Desarrollo Sustentable, Catamarca Province, Argentina (E. A. Fra). Literai uRE Cited Bailey, H.P., B.B. Simpson, and F. Vervoorst. 1977. The physical environment: the independent vari¬ able. Pp. 13-49, in Convergent evolution in wanri deserts (G.H. Orians, O.T, Solbrig, eds.). Dowden, Hutchinson,*and Ross, Stroudsburg, Pemisylvania. Barthlott, W., W. Lauer, and A. Placke. 1996. Global distribution of species diversity in vascular plants: towards a world map of phytodiversity. Erdkunde, 50:317-327. Beissinger, S.R., E.C. Steadman, T. Wohlgenanl, G. Blate, and S. Zack. 1996, Null models for as¬ sessing ecosystem conservation priorities: tlireatened birds as titers of tlireatened ecosys¬ tems in South America. Cons. Biol., 10:1343- 1352. Braun, J.K., and M.A. Mares. 1995. A new genus and species of phyllotine rodent (Rodentia: Muridae: Sigmodontme) from South America. J. Mamm., 76:504-521. Burkart, R., N. Barbaro, R. Sanchez, and D, Gomez. 1999. Eco-regiones de la Argentina. Parques Nacionales, Programa Desarrollo Institucionai Ambiental, Buenos Aires. Cabrera, A. L. 1957. La vegetacion de la puna, Argen¬ tina. Revista de Investigaciones Agricolas, Buenos Aires, Argentina, 370 pp. Cabrera, A. L. 1976. Regiones fotogeograficas Argentinas. Enciclopedia Argentino de Agricultura y Jardineria, T. 1, fasc. 1:1-85. Carleton, M.D., and G,G. Musser. 1989. Systematic studies of oryzomyine rodents (Muridae, Sigmodontinae): a synopsis of Microiyzoniys. Bull. Amer. Mus. Nat. Hist., 191:1-83. Contreras, L.C., J.C. Torres-Mura, and A.E, Spotorno. 1990. The highest chromosomal number for a mammal in a South American desert rodent. Experientia, 46:506-508. 22 Occasional Papers, Museum of Texas Tech University Crespo, J.A. 1960. Panorama zoologico argentine: mami'feros. Physis, 22(63):241-247. Degan, A.A., M. Kam, and D. Jurgrau. 1988. Energy requirements of fat sand rats {Psammomys obesus) and their efficiency of utilization of the sdMb\xs)iAtriplexhalimiis for maintenance. J. Zool. Lond., 215:443-452. Diaz, M.M., R.M. Barquez, J.K. Braun, and M.A. Mares. 1999. A new species of Akodon (Muridae: Sigmodontinae) from northwestern Argentina. J, Mamm., 80:786-798. Gallardo, M.H., J.W. Bickham, R.L. Honeycutt, R.A, Ojeda, and N, Kohler. 1999. Discovery of tet- raploidy in a mammal. Nature, 401:341. Hershkovitz, P. 1987. A histoty of the recent mammal¬ ogy of the Neotropical region from 1492 to 1850. Pp, 11-98, in Studies in Neotropical mammalogy: essays in honor of Philip Hershkovitz (B.D. Patterson and R.M. Timm, eds.). Fieldiana, Zool., 39:1-506. Kahn Arroyo, M.T., F.A. Squeo, J.J. Aimesto, and C. Villagran. 1988. Effects of aridity on plant di¬ versity in the northern Chilean Andes: results of a natural experiment. Ann. Missouri Bot. Garden 75:55-78. Kenagy, G.J, 1972. Saltbush leaves: e.xcision of hyper¬ saline tissue by a kangaroo rat. Science, 178:1094-1096. Kohler, N., M.N. Gallardo, L.C. Contreras, and J.C. Torres-Mura. 2000. Allozymic variation and systematic relationships of the Octodontidae and allied taxa (Mainmalia, Rodentia). J. Zool. Lond., 252:243-250. Mares, M.A. 1975. South American mammal zooge¬ ography; evidence from convergent evolution in desert rodents. Proc. Natl. Acad. Sci. USA, 72:1702-1706. Mares, M.A. 1976. Convergent evolution of desert ro¬ dents: multivariate anlaysis and zoogeographic implications. Paleobiology, 2:39-64. Mares, M.A. 1980. Convergent evolution among desert rodents: a global perspective. Bull. Carnegie Mus., 16:1-51. Mares, M.A. 1982. The scope of South American mam¬ malian biology: perspectives on a decade of research. Pp. 1-26, in Mammalian biology in South America (M.A. Mares and H.H. Genoways, eds.). Special Publications Series, Pymatuning Laboratory of Ecology, University of Pittsburgli, 6:1-539. Mares, M.A. 1992. Neotropical mammals and the myth of Amazonian biodiversity. Science, 255:976- 979. Mares, M.A., and J.K. Braun. 1996. A new species of phyllotine rodent, genus Andalgalomys (Muridae: Sigmodontinae), from Argentina. J. Mamm. 77:928-941. Mares, M.A., and J.K. Braun. 2000. Three new species of Brucepattersonius (Rodentia: Sig¬ modontinae) from Misiones Province, Argen¬ tina. Occasional Papers of the Sam Noble Oklahoma Museum of Natural History, 9:1- 13. Mares, M.A., and R.A. Ojeda. 1982. Patterns of diver¬ sity and adaptation in South American hystricognath rodents. Pp. 393-432, in Mam¬ malian biology in South America (M.A. Mares and H.H. Genoways, eds.). Special Publica¬ tions Series, Pymatuning Laboratory of Ecol¬ ogy, University of Pittsburgh, 6:1-539. Mares, M. A., J. Morello, and G. Goldstein. 1985. The Monte Desert and other subtropical semiarid biomes of Argentina, with comments on their relation to North American arid areas. Pp. 203- 237, in Ecosystems of the world (M. Evenari and I. Noy-Meir, eds.). Elsevier, The Nether¬ lands. Mares, M.A., R.M. Barquez, J.K. Braun, and R.A. Ojeda. 1996. Observations on the mammals of Tucuman Province, Argentina. I. Systematics, distribution, and ecology of the Di- delphimorphia, Xenarthra, Chiroptera, Pri¬ mates, Carnivora, Perissodactyla, Artiodactyla, and Lagomorpha. Ann. Carnegie Mus., 65:89- 152. Mares et al.— Two New Genera of Mammals 23 Mares, M.A., R.A, Ojeda, J.K. Braun, and R.M. Barquez. 1997. Observations on tlie mammals of Catamarca Province, Argentina: systemat- ics, distribution, and ecology. Pp. 89-141, in Life among the muses; papers in honor of James S. Findley (T.L. Yates, WX. Gannon, and D.E. Wilson, eds.). Special Publication, Museum of Southwestern Biology, 3:1-290. Mares, M.A., R.A. Ojeda, C.E. Borghi, S.M. Giannoni, G.B. Diaz, and J.K. Braun. 1997. How desert rodents overcome halophytic plant defenses. BioScience, 47:699-704. Mares, M.A., W.F. Blair, F.A. Enders, D. Greegor, J. Hunt, A.C. Hulse, D. Otte, R. Sage, and C. Tomoff. 1977. Tlie .strategies and community patterns of desert animals. Pp. 107-163, in Convergent evolution in wann deserts (G.H. Orians and O.T. Solbrig, eds.). Dowden, Hutchison and Ross, Inc., Stroudsburg, Penn¬ sylvania. McKenna, M.C., and S.K. Bell; with contributions from G.G. Simpson [etal.J. 1997. Classification of mammals above the species level. Columbia University Press, New York. Mittermeier, R.A., N. Myers, P.R. Gil, and C.G. Mittemieier. 1999. Hotspots: earth^s biologi¬ cally richest and most endangered terrestrial ecosystems. CEMEX and Conservation Inter¬ national, Mexico City. Morello, J. 1958. Laprovincia fitogeografica del Monte. Opera Lilloana, 2:1-155. Ojeda, R.A., C.E. Borghi, G.B. Diaz, S.M. Giannoni, M.A. Mares, and J.K. Braun. 1999. Evolution¬ ary convergence of the highly adapted desert rodent Tympanoctomys harrerae (Octo- dontidae). J. Arid. Environ., 41:443-452. Orians, G.H., and O.T. Solbrig (eds.). 1977. Conver¬ gent evolution in warm deserts. Dowden, Hutchinson, and Ross, Stroudsburg, Pennsyl¬ vania. Pascual, R,, and E. Ortiz Jaureguizar. 1990. Evolving clunates and mammal faunas in Cenozoic Soutli America. J. Human EvoL, 19:23-60. Patterson, B.D. 1994. Accumulating knowledge on the dimensions of biodiversity: systematic perspec¬ tives on Neotropical mammals. Biodiversity Letters, 2:79-86. Patterson, B.D., D.F. Stotz, S. Solari, F.W. Fitzpatrick, andV. Pacheco. 1998. Contrasting pattems of elevational zonation for birds and mammals in the Andes of southeastern Peru. J. Biogeogra- phy, 25:593-607. Peirano, A. 1957. Observaciones generalcs sobre la tectonica y los depositos terciarios del cuadrangulo 26° S - 64° 30’0 - 28° 30’S - 67° S en el noreste Argentina. Acta Geol. Lilloana, 1:61-144. Platnick, W. 1991. Pattems of biodiversity: tropical vs. temperate. J. Nat. Hist., 25:1083-1088. Reig, O. A. 1977, A proposed unified nomenclature for the enamelled components of the molar teeth of the Cricetidae (Rodentia). J. Zool. (London), 181:227-241. Ridgway, R. 1912. Color standards and color nomen¬ clature. R. Ridgway and A. Ploen and Com¬ pany, Washington, D. C., 43 pp., 53 plates. Rohlf, F. J. 1998. NTSYSpe. Numerical Taxonomy and Multivariate Analysis System. Version 2.0. Exeter Software, Setauket, New York. Roig, F.A., M.M. Gonzalez Loyarte, E.M. Abraham de Vazquez, E. Martinez Carretero, E. Mendez, and V.G. Roig. 1992. Argentina: desertifica¬ tion hazard mapping of central-western Argen¬ tina. Pp. 50-53, in World atlas of desertifica¬ tion (N.J. Middleton and D.S.G. Thomas, eds.). United Nations Environment Programme, Lon¬ don. SAS Institute Inc. 1998. StatView 5. SAS Institute Inc., Cary, North Carolina. Swofford, D.L. 1998. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Ver¬ sion 4. Sinauer Associates, Sunderland, Mas¬ sachusetts. Torres-Mura, J.C., M.L. Lemus, and L.C. Contreras. 1989. Herbivorous specialization of the South American desert rodent Tympanoctomys barrerae. J, Mammal., 70:646-648. 24 Occasional Papers, Museum of Texas Tech University Voss, R,S,, and L.H. Emmons. 1996. Mammalian di¬ versity in Neotropical lowland rainforests: a preliminary assessment. Bull. Am. Mus. Nat. Hist., 230:1-115. Vucetich, M.G., D.H. Verzi, and J.-L. Harlenberger. 1999. Review and analysis of the radiation of the South American Hystricognathi (Mamma¬ lia, Rodentia). C. R. Acad. Sci. Paris, Sciences de la terre et des planetes, 329:763-769. Vuilleumier, B.S. 1971. Pleistocene changes in the fauna and flora of Soutli America. Science, 173:771 - 780. Williams, D.F., and M.A, Mares. 1978. A new genus and species of phyllotine rodent (Mammalia: Muridae) from northwestern Argentina. Ann. Carnegie Mus., 47:192-221. Wilson, D.E., and D.M. Reeder (eds). 1993. Mammal species of the world: a taxonomic and geo¬ graphic reference. Second ed. Smithsonian In¬ stitution Press, Washington, DC. Wilson, D.E., and S. Ruff (eds.). 1999. The Smitlisonian book of North American mammals. Smithsonian Institution, Washington, DC. Appendix i Specimens examined All specimens that were examined are listed along with localities and catalog numbers. All speci¬ mens are preserved as standard preparations of skin, skull, and skeleton and are housed in the: British Mu¬ seum of Natural History, London (BM); Carnegie Mu¬ seum of Natural History, Pittsburgh, Pennsylvania (CM); Coleccion de Mamiferos Lillo, Universidad Nacional de Tucuman, Tucuman, Argentina (CML); Coleccion Mastozoologica, Institute Argentina de Tnvestigaciones de las Zonas Aridas, Mendoza, Argen¬ tina (lADZA-CM); Michigan State University Museum, East Lansing (MSU); Museum of Vertebrate Zoology, University of California, Berkeley (MVZ); Sam Noble Oklahoma Museum of Natural History, Norman (OMNH); Museum of Texas Tech University, Lubbock (TTU). Field numbers of specimens in the Sam Noble Oklahoma Museum of Natural History and not yet cata¬ loged are designated Arg. Aconaemys fuscus (14).—ARGENTINA: Neuquen: Lago Correntoso (N. end), 1 (159433 MVZ); Lago Correntoso (N. end) (Ruca Malen), 2 (159431, 159434 MVZ). CHILE: Curico: Nr. Peteroa (Valle de las Cuevas, E side of Andes, nr. Vole, of Peteroa, alt, 6,000 R), 1 (55.12.24.195 BM), Malleco: 27 km W Angul in Parque Nacional, 3,800 ft., 2 (6347, 6348 MSU); Parque Nacional, 27 km WNW Angul, 3,650 ft., 1 (6346 MSU); Laguna Malleco, 25 km N Cunacautin, 3,000 ft., 1 (6349 MSU); Osomo: Puesto Montt, 1 (16.11.14.4 BM). Valdivia: vie. Valdivia, 1 (142176 MVZ). Chili, 3 (43.12.30.32, 56.7.1.1, 56.7,1.3 BM). No locality: 1 (no number BM). Ctenomys mendocinus (7).—ARGENTINA: San Juan: 17 km ESE Jose Marti (by road) on road to Chahar Seco, 1 (3352 Arg); Qiiebrada de las Flores, 4 km E and 5 km N Guayamas, 1 (3406 Arg). San Luis: Gran Bajo Salitroso, 1 (3342 Arg) 6 km W Ilualtaran, Parque Provincial Sierra de las Quijadas, 1 (3279 Arg); 8 km W of La Botija, Pampa de las Salinas, 1 (3240 Arg); 15 km E Salinas del Bebedero, 1 (544 Arg); Sali¬ nas del Bebedero, 15 km SE, 1 (3313 Arg). Octodon bridgesi (7).—CHILE: Colchagua: Teno River, 2 (43.7.20.5,55.12.24.196 BM). Malleco: 4 km W Banos Rio Blanco, 3,200 ft., 1 (6345 MSU). Valparaiso: Quilpue, 150 m, 3 (9.6.13.2, 10.7.23.4, 10.7.23.5 BM). Chill, 1 (7.1.1.170 BM). Octodon degus (26).—CHILE: Aconcagua: 3 km N Las Modes, 1 (28634 TTU). Atacama: 40 km E Vallenar, 2 (28636, 28637 TTU). Coquimbo: Fray Jorge, 1,000 ft., 1 (118664 MVZ); Fray Jorge Natl. Park, 5 (28635, 28638, 28639, 28640, 30232 TTU); 15 km S La Serena, 2 (28641,28642 TTU). Santiago; Fundo Santa Laura, 8 km W Tiltil, 1100 m, 1 (150076 MVZ); Fundo Santa Laura, approx. 10 km W Tiltil, 1100 m, 2 (150072,150075 MVZ); Fundo Santa Laura, Cuesta la Doimida, 8 km W of Tiltil, lOOOm, 2 (150078, 150079 MVZ); Puente Alto, 800 m, 1 (4.1.7.13 BM); Mares et al.— Two New Genera of Mammai.s 25 Rinconada de Maipu, 5 (28643-28647 TTU). Valparaiso: Llinllin, Limache, 250 m, 1 (12183 CM); Olmue, 1 (12182 CM); Quilpue, 1 (2107 MSU). Chili, 1 (53.8.29.14 BM). Octodon lunatus (12).—CHILE: Santiago: Fundo Santa Laura, 10 km W of Tiltil, 1100 m, 3 (150090, 150091,150092 MVZ); Fundo Santa Laura, Cuesla la Dormida, 10 km W of Tiltil, 1000 m, 8 (150093, 150098, 150099, 150100, 150101, 150102, 150103 MVZ, 28633 TTU); Fundo Santa Laura, Cuesla la Dormida, 10 km W of Tiltil, 1100 m, 1 (150088 MVZ). Octodontomys gliroides (32).—ARGENTINA: Jujuy: Alfarcito, 2600 m, 3 (21.11.1.74, 21.11.1.76, 21.11.1.78 BM); Casalindo, 4000 m, 2 (19.8.2.26, 19.8.2.27 BM); Cerro Casalindo, 4500 m, 1 (19.8.2.28 BM); 11 km east of Humahuaca, 2 km east of Pucara on road to Cianzo, 2 (23476 OMNH, 4477 lADIZA- CM); Maimara, 2230 m, 2 (12.12.12.42, 12.12.12.43 BM); Maimara, 2300 m, 2 (4917,5007 CM); Maimara, 2600 m, 1 (12.12.12.44 BM); 10 km west of Purmarca on highway 52,1 (3437 CML). Salta: Inca Mayo, 2500 m, 1 (34.11.4.94 BM); M. Sola, 2500 m, 8 (34.11.4.86- 34.1 1.4.93 BM). BOLIVIA: Lipez, 4500 m, 1 (26.6.12.81 BM); Oruro, 3700 m, 2 (2.2.2.5, 2.12.1 BM); Potosi, 4400 m, 1 (2.2.2.2 BM); Est. Yumma, 2200 m, 3 (26.6.12.76, 26.6.12.77, 26.6.12.80 BM); Yuruma, 2200 m, 1 (26.6.12.78 BM); Yuruma, 20 mi. SSE Tupiza, 10,500 R, 1 (120236 MVZ). Octomys mimax (17).—ARGENTNA: Catamarca: Puntilla, Tinogasta, 1000 m, 5 (20.5.11.29- 20.5.11.33 BM), San Juan: Pedemal, 4,040 ft., 1 (23478 OMNH); Pedemal, 1200 m, 3 (21.6,19.11 -21.6.19.13 BM); Parque Provincial Ishigualasto, 3 kmN from head¬ quarters, I (lADIZA-CM 4479); just inside Valle de la Luna, 5 (44068-44072 CM). San Luis: 6 km W Hualtaran, Parque Provincial Sierra de la Qiujadas, 2,800 ft., 2 (Arg 3262, 3271). Spalacopus cyanus (24).—CHILE: Aconcagua: 3 km N La Modes, 1 (28648 TTU). Coquimbo: Fray Jorge Park, 2 (28649, 28650 TTU); Las Plumas, 95 km N Los Vilos, 1 (28651 TTU). Santiago: Cerro Roble, 3 km NW Caleu, 2000 m, 1 (150111 MVZ); Cerro Roble, approx, 2 km S by rd. from summit, 3 km NW of Caleu, 2000 m, 2 (150110, 150112 MVZ); La Parva, 2700 m, 3 (28652-28654 TTU); La Parva, 3200 m, 1 (28655 TTU). Valparaiso: Quilpue^ 1 (2108 MSU); Quilpue, 200 m, 1 (8.4.7.29 BM); Valparaiso, coast hills, 500 m, 8 (0.3.5.2-0.5.4.6, 1.3.21.14,97.5.1.11,98.1.8.5 BM). Chile, 1 (43.7.20.4 BM). S. Chile, 1 (73.12.16.4 BM). No locality, 1 (502.a BM). Tympanoctomys barrerae (8).—ARGEN¬ TINA: Mendoza: 27 kmN Desaguadero, 1,670 ft., 2 (3438 CML, 23477 OMNH); 40 km al N Desaguadero, camino a Arroyito, 1 (2747 lADIZA-CM); 16 km S, 3 km E Nihuil, 5 (3745 CML; 5346 lADIZA-CM; 23902 OMNH; 3789, 3797 Arg). 26 Occasional Papers, Museum of Texas Tech University Appendix ii Descriptions of discrete-character states Descriptions of discrete-character states and char¬ acter state values are listed. Character states are given by numbers in parentheses and are in order (Table 1). External 1 Tip of tail: (0) Not pencilled. (1) Pencil small. (2) Pencil large. 2 Tail color: (0) Similar or somewhat similar to dorsum. (1) Different from dorsum. 3 Bristle bundles: (0) Absent. (1) Present, but not well developed. (2) Present, well developed. Cranial 4 Postorbital processes; (0) Absent. (1) Present, 5 Length of palate: (0) Short, not extending be¬ yond the posterior border of Ml. (1) Short, ex¬ tending to mid-M2. (2) Long, extending at least to the posterior border of M2. 6 Paraoccipital processes; (0) Not adpressed to bullae. (1) Adpressed to bullae. 7 Incisive foramina: (0) Edges not raised. (1) Edges raised. 8 Suture between jugal and zygomatic process of squamosal; (0) Angle squarish. (1) Angle acute, (2) Angle roundish. 9 Foramina anterior to bullae: (0) Posterior open¬ ing of alisphenoid canal, foramen ovale, and middle lacerate foramina present, (1) Poste- nor opening of alisphenoid canal and foramen ovale present; middle lacerate foramina tiny. (2) Posterior opening of alisphenoid canal and foramen ovale present. 10 Infraorbital canal; (0) No distinct groove and flange for infraorbital nerves. (1) Distinct groove and flange present. 11 Zygomatic arches; (0) Convergent anteriorly. (1) Parallel-sided, (2) Expanded. 12 Lacrimals: (0) Small, (1) Large. 13 Bullae: (0) Small and not inflated. (1) Medium and slightly inflated. (2) Large and inflated. 14 Foramina posterior to incisive foramina: (0) Absent. (1) One pair present. 15 Postglenoid fossae: (0) Small and round. (1) Large and narrow. 16 Angular process: (0) Flattened. (1) Slightly flattened. (2) Thin, not flattened. 17 Angular (lunar) notch: (0) Shallow. (1) Deep. 18 Interorbital width: (0) Narrow, less than width of rostrum. (1) Wide, greater than width of rostrum. Dental 19 M3; (0) Metacone and hypocone not well de¬ veloped. (1) Metacone and hypocone moder¬ ately to well developed. 20 Incisor position: (0) Opisthodont. (1) Orthodont. (2) Proodont. 21 Incisor color: (0) White. (1) Orange. (2) Or¬ ange with lateral white edge. 22 Molar shape: (0) Not “8”-shaped. (1) “8”- shaped. Mares et al.— Two New Genera of Mammals 27 Addresses of authors: Michael A. Mares Sam Noble Oklahoma Museum of N atural History and Department of Zoology University of Oklahoma Nomian, OK 73072 e-mail: mamares@ou.edu Janet K. Braun Sam Noble Oklahoma Museum of Natural History University of Oklahoma Nonnan, OK 73072 e-mail: jkbraun@ou.edu RubLn M. Barquez Facultad Ciencias Naturales and Institute Miguel Lillo, UniversidadNacional de Tucuman, and CONICET (Consejo Nacional de Investigaciones Ciendficas y Tecnicas) Tucuman 4.000, Argentina e-mail: mibarquez@xaire.com M. M6nica DIaz Facultad Ciencias Naturales and Institute Miguel Lillo, Universidad Nacional de Tucuman, and CONICET (Consejo Nacional de Investigaciones Cientilicas y Tecnicas) Tucum^i 4.000, Argentina e -mai 1; pidbatuc@i n fovia. c om. ar Publications of the Museum of Texas Tech University It was through the efforts of Horn Professor J Knox Jones, as director of Academic Publications, that Texas Tech University initiated several publications series including the Occasional Papers of the Museum. This and future editions in the series are a memorial to his dedication to excellence in academic publications. Professor Jones enjoyed editing scientific publications and served the scientific community as an editor for the Journal of Mammalogy, Evolution, The Texas Journal of Science, Occasional Papers of the Museum, and Special Publications of the Museum. It is with special fondness that we remember Dr. J Knox Jones. Institutional subscriptions are available through the Museum of Texas Tech University, attn; NSRL Publications Secretary, Box 43191, Lubbock, TX 79409-3191. Individuals may also purchase separate numbers of the Occasional Papers directly from the Museum of Texas Tech University. Layout and Design: Janet K. Braun, R. Richard Monk, and Jacqueline B. Chavez Cover Design: Michael A. Mares and Patrick Fisher Copyright 2000, Museum of Texas Tech University All rights reserved. No portion of this book may be reproduced in any form or by any means, including electronic storage and retrieval systems, except by explicit, prior written permission of the publisher. This book was set in Times New Roman and printed on acid-free paper that meets the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library' Resources. Printed; 22 December 2000 Library of Congress Cataloging-in-Publication Data Occasional Papers, Number 203 Series Editor: Robert J. Baker TWO NEW GENERA AND SPECIES OF HALPHYTIC DESERT MAMMALS FROM ISOLATED SALT FLATS IN ARGENTINA By: Michael A. Mares, Janet K. Braun, Ruben M. Barquez, and M. Mbnica Diaz ISSN 0169-0237 Museum of Texas Tech University Lubbock, TX 79409-3191 USA (806)742-2442