The early Pliocene (Blancan) Always Welcome Inn Ophiomys and the migration of Ophiomys into the Pacific Northwest

Misty Bork, John Rinehart, and Jay Van Tassell

Science Department, Badgley Hall, Eastern Oregon University, La Grande, OR 97850

ABSTRACT

Measurements of the lengths of the occlusal surfaces of the lower first molars found at the Always Welcome Inn fall into the same range as those of Ophiomys taylori, but this may reflect a high percentage of juveniles or the small number of specimens.  The mean width of the teeth is closest in size to Ophiomys cf. mcknighti.  Cladistics analysis suggests that the Always Welcome Inn vole is closely related to Ophiomys mcknighti from the ~4.3 m.y.-old White Bluffs fauna of Washington and more distantly related to “Mimomys” sawrockensis from the ~4.8 m.y.-old Upper Alturas fauna of California.  These results suggest that Ophiomys migrated up the Sacramento River at ~ 4.8 Ma; traveled along the shores of the Glenns Ferry lake during the Pliocene stage of Lake Idaho to the Always Welcome Inn area between ~4.8-4.3 Ma; and then crossed the drainage divide and migrated through the Grande Ronde Valley and down the Grande Ronde River into the Columbia River system at ~4.3 Ma. 

INTRODUCTION

Our knowledge of the early migration of the arvicoline Ophiomys through the western United States is based on information from only a few localities (Fig. 1).  These include:  1) the earliest Blancan (~4.8 Ma) Alturas Formation of northern California, where  Ophiomys has been found in planar-bedded, diatomaceous, silty lake margin deposits with reed and rush remains (Repenning, 1987, 2003; Repenning and others, 1995); 2) the ~4.3 Ma White Bluffs, the ~3.9 Ma Blufftop , and ~3.0-2.8 Ma Taunton Faunas of the Ringold Formation in the Columbia River Basin near Pasco, Washington (Gustafson, 1978; 3) the ~3.7 -3.1 Ma Hagerman and ~2.2 Ma Grandview faunas of the western Snake River Plain in Idaho (Wilson, 1933; Hibbard, 1959; Hibbard and Zakrzewski, 1967; Zakrzewski, 1969; Shotwell, 1970; Repenning and others, 1995; and 4) the ~2.4-1.8 Ma Shooting Iron Formation fauna from the Jackson Hole area of Wyoming (Barnosky, 1985).  Mou (1995) named a new species, Mimomys panacaensis, from the ~4.8 Ma Panaca Formation of southeast Nevada that represents an evolutionary stage similar to that of the early Blancan Ophiomys found in the White Bluffs Formation of Washington.

The information from these sites led to the theory that Ophiomys immigrated to North America from Eurasia 4.8 million years ago and followed the Pliocene Pacific coastline to the Columbia River, where it moved inland up the Columbia River, which was not yet connected to the Snake River, to the Pasco basin area by 4.3 Ma (Fig. 2).  Ophiomys also migrated along the Pliocene coastline as far south as the San Francisco area, where it moved up the Sacramento River to Alturas by 4.8 Ma and reached the Glenns Ferry lake area (Pliocene Lake Idaho) at Hagerman by 3.7 Ma (Repenning and others, 1995; Repenning, 2003).  Ophiomys then migrated eastward through Yellowstone Pass during the late Pliocene (Barnosky, 1985; Repenning, 2003).

              This paper describes Ophiomys fossils from the early Pliocene (Blancan) Always Welcome Inn fossil locality in Baker City, Oregon, and what they suggest about the early migration patterns of Ophiomys in the Pacific Northwest.  Based on beaver (Dipoides), swallow (Hirundo aprica), and the vole fossils, the age of the Always Welcome Inn sequence is ~4.8-4.3 Ma, the same age as the Fox Canyon local fauna of Kansas.

Figure 1.  California, Idaho, Oregon and Washington Pliocene vole localities and representative first molar occlusal patterns.

Figure 2.  Possible vole migration routes during the Pliocene, based in part on Repenning and others (1995).  A: Alturas; AWI: Always Welcome Inn; H: Hagerman; P: Panaca; V: Verde; and WB: White Bluffs.

The sedimentary sequence at the Always Welcome Inn fossil locality in Baker City, Oregon (Fig. 3) has been described in detail by Van Tassell and others (2007).  The 10-meter thick sequence records a change from shallow-water lacustrine deposits at the base to meandering stream and floodplain deposits at the top which are separated from late Pliocene-early Pleistocene chert-rich gravels by an angular unconformity.  Ophiomys fossils are most abundant in a trough cross-bedded silty sand layer at the base of a fining-upward cycle in the middle of the stream deposits, but have also been found in the silty sand layer that marks the base of the stream deposits and in a sandy silt layer in the upper part of the lacustrine sequence.

Figure 3.  The sedimentary sequence at the Always Welcome Inn and photos of the modern equivalents of some of the plant and animal fossils found at the site.  The dashed line separates the lower lacustrine part of the sequence and the upper meandering stream deposits.  Dia: diatomite; Slt: silt; sSlt: sandy silt; sltS: sandy silty sand; and Gr: gravel.  The gravels at the top of the sequence are late Pliocene to Pleistocene in age.

NOMENCLATURE

The arvicoline genus Ophiomys was first described by Hibbard and Zakrzewski (1969) based on studies of specimens from the Pliocene Lake Idaho deposits (Glenns Ferry Formation) of Idaho.  The Ophiomys genotype is a specimen originally described as Mimomys? parvus by Wilson (1934).  Repenning (1987) redefined Ophiomys as a subspecies of the genus Mimomys.  This nomenclature was extensively revised by Repenning (2003), who reassigned the species Mimomys (Ophiomys) parvus to the genus Ophiomys.  Martin (2003) concluded that only two species of Mimomys, the late Blancan or early Irvingtonian M. dakotaensis Martin, 1989 from the Java local fauna of South Dakota, and M. virginianus from the Irvingtonian Cheetah Room local fauna of Hamilton and Porcupine Caves in Colorado can be identified in North America based on extensive study of their dental characteristics and suggested that Mimomys came to North America during the earliest Pleistocene, not at the beginning of the Pliocene.  For this reason, we have chosen to use the nomenclature of the papers in which species were originally defined (Hibbard and Zakrzewski, 1967, Gustafson, 1978, Mou; 1997) rather than the new species names proposed by Repenning (1987, 2003).  Equivalent names are shown in Table 1.

Table 1.  Nomenclature

                                                          Hibbard and Zakrzewski (1967)         Repenning  (1987)                                      Repenning (2003)

Blancan                                            Gustafson (1978)

V        Mimomys? parvus                   Ophiomys parvus                       Mimomys (Ophiomys) parvus               Ophiomys parvus

                  (Wilson, 1934)

                  Grandview fauna,

                  Jackass Butte, ID (2.2 Ma)

IV              ----------                                 Ophiomys magilli                     Mimomys (Ophiomys)                             Mimomys meadensis

                                                                 (Hibbard and Zakrzewski,           taylori-parvus                                            (including Ophiomys

                                                                 1969)                                                                                                                 taylori from Taunton fauna,

                                                                 Sand Draw fauna, NB                                                                                       WA (3.0-2.8 Ma)

                                                                  (2.6-2.5 Ma

-----------------------------------------------------------------------------------                

              Pliophenacomys                    Ophiomys meadensis        

                  meadensis

                  (Hibbard, 1956)

                  Sanders and Dixon

                  faunas, KS (2.6-2.5 Ma)

III           Pliophenacomys                     Ophiomys taylori                        Mimomys (Ophiomys)                            Mimomys taylori

              idahoensis/                                                                                      taylori

              Nebraskomys?

              taylori

              (Hibbard, 1959)

               Hagerman fauna, ID

               (3.7-3.1 Ma)

II                ----------                                   Ophiomys cf. mcknighti              Mimomys (Ophiomys)                          Mimomys sawrockensis-

                                                                   (Gustafson, 1978)                           mcknighti-taylori                                  taylori

                                                                   Blufftop fauna, WA (3.9 Ma)

I                 ----------                                   Ophiomys mcknighti                    Mimomys (Ophiomys)                         Mimomys sawrockensis

                                                                   (Gustafson, 1978)                            mcknighti

                                                                   White Bluffs fauna, WA (4.3 Ma)

                                                                   Alturas fauna, CA (4.8 Ma)

                                                                                                                            Mimomys panacaensis

                                                                                                                            (Mou, 1995)

                                                                                                                            Panaca fauna, NV (4.8 Ma)

METHODS

Samples have been collected at 0.25 m intervals throughout the Always Welcome Inn sequence.  The samples were dry-sieved on-site through screen with a square 1.5 mm mesh.  The Ophiomys teeth and jaws were found in the intervals 2.75-3, 5.25-5.5, and 6.75-7 meters above the base of the sequence.

              The characteristics of the vole teeth were carefully observed using a Leitz binocular microscope.  Measurements of the mesial-distal length (also called the “anterior-posterior length”) of each of the Always Welcome Inn vole first molars and the maximum width perpendicular to the mesial-distal length were made according to the techniques described by Barnosky (1985).  Mean values and standard deviations were calculated from these measurements for the Always Welcome Inn vole and compared with those of voles from other late Miocene and Pliocene faunas.

              Next we determined the characteristics of the Always Welcome Inn according to the criteria used in Mou's (1997) cladistic analysis of Mimomys panacaensis and other species.  These characteristics include:

• The presence or absence of a Mimomys kante.
• The development of a dentine tract on the labial side of the anteroconid complex of     the lower first molars,         

              3)  The presence of a posterior enamel islet on the upper third molar,

              4)  The number of roots on the upper third molars,

              5)  The number of roots on the upper second molars,

              6)  The presence of an enamel islet on the lower first molar, and

              7)  The length of the lower first molar.

              Finally, we inputted Mou's (1997) data into the the cladistic program Phylip 3.67 (Pars subprogram; Wagner parsimony), which is similar to the Paup program used by Mou (1997) and McClade.  The program was set to choose the most parsimonious tree.  After the program produced results identical to those of Mou (1997), we inputted the data from the Always Welcome Inn and other late Miocene and Pliocene vole species found in the western United States.     The sources of the data for the vole teeth used in this study include:

Promimomys mimus, 7-6 Ma                      Mou (1997)

“Mimomys” panacaensis, 4.8 Ma   Mou (1997)

“Mimomys” sawrockensis, 4.8 Ma              Repenning (2003)

Ophiomys mcknighti, 4.3 Ma                      Gustafson (1978), Mou (1997)

Ophiomys cf. mcknighti, 3.9 Ma                  Repenning (2003)

Ophiomys taylori, 3.7-3.1 Ma                       Hibbard and Zakrkzewski (1967), Zakrzewski (1969)

Ophiomys magilli, 3.3 Ma                           Mou (1997)

Ophiomys meadensis, 2.4-1.8 Ma             Hibbard (1956), Barnosky (1985)

Ophiomys parvus, 2.2 Ma                           Hibbard and Zakrzewski (1967)

Cosomys primus, 3.7-3.1 Ma                      Hibbard and Zakrzewski (1967), Zakrzewski (1969), Lich                                                                           (1990), Mou (1997)

Ogmodontomys sawrockensis                     Mou (1997)

Ogmodontomys cf. poaphagus                    Czaplewski (1990)

THE ALWAYS WELCOME INN OPHIOMYS

              The identity of the primitive vole fossils found in the early Pliocene sediments behind the Always Welcome Inn in Baker City, Oregon, has been puzzling ever since the first vole teeth were found in the sequence in May 2005.  These teeth included two incisors and a broken lower third molar from a juvenile vole.  Leithner and Van Tassell (2005) noted that the shape of the molar was closer to the shapes of Ophiomys taylori than it was to the shapes of other Pliocene voles from the area of Hagerman, Idaho, but were not able to positively identify the Always Welcome Inn vole.  Additional Ophiomys teeth found at the site convinced Van Tassell and others (2007) that the Always Welcome Inn vole is most closely related to Ophiomys mcknighti from the White Bluffs locality of Washington, but they noted that one of the juvenile lower first molars has a narrow anteroconid, in part due to deep penetration by lingual reentrant 4, which is almost never seen in Ophiomys specimens that have atolls, especially early forms such as "Mimomys" panacaensis or Ophiomys mcknighti.  This is a feature, along with the bulbous form of the anterior part of the anteroconid above lingual reentrant 4, that is seen only in Cosomys molars from the Glenns Ferry Formation in the Hagerman, Idaho, area (Van Tassell and others, 2007).  Ophiomys mcknighti also has features similar to those of Cosomys primus (Gustafson, 1978).

SYSTEMATIC DESCRIPTION

Family ARVICOLIDAE Gray, 1821

Subfamily Ondatrinae Repenning, 1982

Tribe Ogmodontomyini Martin, 2007

Genus Ophiomys Hibbard and Zakrzewski, 1967

Description of the Teeth

              The Ophiomys fossils collected at the Always Welcome Inn site as of December 2008 include 2 lower jaws (including one with a first and second molar and one with a first molar), 22 individual molars, and several incisors.  Including the teeth in the jaws, 7 lower first molars (m1), 5 lower second molars (m2), 1 lower third molar (m3), 6 upper first molars (M1), 5 upper second molars (M2), and 1 upper third molar (M3) have been found (Fig. 4). 

              The seven lower first molars include 6 left teeth and 1 right tooth.  These second molars consist of a posterior loop, three alternating triangles, and a complex anterior loop.  The anterconid complex (ACC) of one of the teeth is broken.  Five of the remaining 6 lower first molars (83%) have an enamel islet on the AC.  A Mimomys kante is present on 4 of the teeth and the dentine tract height on the labial side of the ACC is low.  Three of the six teeth that have an ACC have an enamel pit, prism fold, and crenulations; one has an enamel pit and a prism fold, but no crenulations; and two have a prism fold or an enamel pit, but not both.  Four of the teeth show moderate wear and have fully developed roots and three show heavy wear, with a crown that is less than 1 mm in height and an enamel pattern that is not interrupted.  The anterolabial margin of the anterior loop is rounded and is crenulated in the less worn adult teeth.  The apex of lingual reentrant 1 is straight in 5 out of 6 teeth that show it.  Lingual reentrants 2 and 3 are constricted and turn anterad.  The fourth lingual reentrant present on the unbroken teeth is 1/3 to 1/2 as deep as the posterior lingual reentrant.

              The five lower second molars include 4 left teeth and 1 right tooth.  The lower second molars consist of a posterior loop, 2 alternating triangles, and an anterior loop.  One of the lower second molars is broken.  The teeth show moderate to heavy wear and the roots are fully developed.  The juvenile second molar described by Leithner and Van Tassell (2005) has no roots.  The other second molars have two roots.  The posterior root is two times the width of the anterior loop.  Two out of five lower second molars have a pronounced prism fold in the anterior loop.  The lingual reentrants do not turn anterad at the tips.

              The one lower third molar that has been found is a left molar with two roots.  It consists of a posterior loop, two alternating triangles and an anterior loop.  The tooth shows moderate wear and the roots are fully developed.  The lingual reentrants do not turn anterad at the tips.

Figure 4.  Representative vole teeth from the Always Welcome Inn.  Scale bars are in millimeters.

              The six upper first molars include 3 left and 3 right teeth.  The teeth consist of an anterior loop, three alternating triangles, and a posterior loop.  The teeth show moderate to heavy wear and have fully developed roots.  On one of the teeth the anterior root is the largest.  The nature of the reentrants is variable.  On two teeth, all reentrants except the anterolabial turn posterad at the tip; on another only the lingual reentrant turns posterad at the tip; on another all reentrants except the first lingual turns posterad at the tip, and on another all reentrants turn posterad at the tip.

              The five upper second molars include 4 left teeth and 1 right tooth.  The teeth consist of an anterior loop, two alternating triangles, and a posterior loop.  The teeth have three roots, one posterior and two anterior.  The wear varies from slight to heavy.  All reentrants except the anterolabial turn posterad at the tips, except in the tooth which is slightly worn, where none of the reentrants turn posterad at the tips.  The dentine connecting the triangles becomes narrower with wear.

              The single upper third molar is a left molar with two roots.  It consists of a posterior loop, two alternating triangles, and an anterior loopo.  The lingual reentrants do not turn posterad or anterad.

              Overall, the characteristics of the Always Welcome Inn Ophiomys teeth closely resemble those of Ophiomys mcknighti.  Differences include:  1) There is no suggestion of fusion of two roots to form the posterior root in the Always Welcome Inn Ophiomys lower second molars, 2) the lingual reentrant on the lower third molar does not curve slightly anterad at the tip as it does in Ophiomys mcknighti, 3) the curvature of the reentrants on the upper first molar is much more variable than reported for Ophiomys mcknighti, and 4) the lingual reentrants do not turn posterad at the tips like those of Ophiomys mcknighti.

Sizes of the Lower First Molars

              Two of the seven Always Welcome Inn lower first molars are broken.  The remaining five have a mean length of 2.43 ± 0.06 mm.  The lengths range from 2.38 to 2.52 mm.  The widths of all seven lower first molars have a mean value of 1.19 ± 0.07 mm and range from 1.10 to 1.30 mm (Table 2).  These measurements are slightly different from those of Van Tassell and others (2007), who reported lower first molar lengths ranging from 2.22-2.36 mm and widths ranging from 1.07-1.09 mm.

              The lengths of the Always Welcome Inn vole lower first molars fall within the same range as those reported for Ophiomys taylori lower first molars found in Hagerman, Idaho, and overlap with the range of sizes reported for "Mimomys" panacaensis

(Table 3).

              The widths of the Always Welcome Inn lower first molars are most similar to those of Ophiomys cf. mcknighti from Blufftop, Washington.  The range of widths of the lower first molars at the Always Welcome Inn overlaps with the range of widths of lower first molars of Ophiomys mcknighti and “Mimomys” panacaensis (Table 4).

Table 2.  Always Welcome Inn vole lower first molar lengths and widths.

Table 3. Lengths of vole lower first molars from other Pliocene faunas.

Table 4. Widths of vole lower first molars from other Pliocene faunas.

CLADISTIC ANALYSIS

Mou (1997) introduced the use of a cladistic program (PAUP) for unraveling the evolutionary history of primitive voles.  She used four characteristics of the lower first molar, one feature of the upper second molar, and two characteristics of the upper third molar that had been identified as important in defining species by earlier workers.  Using a similar cladistic program (Phylip 3.67) and its subprogram (PARS) that allows ordering of each characteristic and adding the characteristics of the Always Welcome Inn vole to the data (Table 5) yielded a result that is identical to what Mou (1997) found with the addition of the Always Welcome Inn vole splitting off of “Mimomys” panacaensis at the same time as Ophiomys mcknighti (Figure 5).  Adding four additional species (“Mimomys” sawrockensis from the Upper Alturas fauna of California, Ophiomys cf. mcknighti from the Blufftop fauna of Washington; Ophiomys taylori from the Hagerman fauna of Idaho, and Ophiomys parvus from the Grandview fauna of Idaho produced a similar pattern but with the significant difference that the younger faunas split off from “Mimomys” sawrockensis” rather than “Mimomys” panacaensis.  There is also a clear division of the Always Welcome Inn and faunas from Washington from the younger faunas from Idaho and the Midwestern United States (Figure 6).

Table 5.  Data used for cladistic analysis

Figure 5.  Results of cladistic analysis adding the Always Welcome Inn Ophiomys to the list of species analyzed by Mou (1997).

Figure 6.  Results of cladistic analysis using data from Mou (1997), plus data for the Always Welcome Inn vole and other Pliocene voles that were not included in Mou’s (1997) analysis.  The cladistics analysis suggests that “Mimomys” sawrockensis from the Upper Alturas fauna may be an ancestor of both the Always Welcome vole and the Ophiomys species of Washington.

MIGRATION OF OPHIOMYS INTO THE PACIFIC NORTHWEST

Our observations and measurements support Van Tassell and others’ (2007) suggestion that the Always Welcome Inn is most like Ophiomys mcknighti from the Columbia River Basin.  The lengths of the Always Welcome Inn Ophiomys first molars are slightly smaller than those reported by Gustafson (1978) for Ophiomys mcknighti and some of the features of the occlusal surfaces of the teeth differ, but 6 out of 7 of the characteristics chosen by Mou (1995) for cladistic analysis are identical and the one difference, the percentage of lower first molars with enamel islets is very slight.  The results of our cladistic analysis suggest that both the Always Welcome Inn Ophiomys and Ophiomys mcknighti may be derived from “Mimomys” sawrockensis from the Alturas Basin of California.

              These results reinforce Repenning’s (2003; also see footnote in Gustafson, 1978) conclusions that the ~4.8 m.y.-old “Mimomys” sawrockensis of the Sacramento River drainage basin and the ~4.3 m.y.-old Ophiomys mcknighti of the Columbia River Basin are closely related.  The Always Welcome Inn fauna includes a beaver (Dipoides) that is more like the Dipoides in the Lake Idaho area than the species in the Columbia River drainage (Van Tassell and others, 2007) and a sucker (Catostomus) which, according to Smith and others (2000) does not occur in the Columbia River drainage basin until 3.9 Ma.  This suggests an eastward connection to Lake Idaho rather than a westward connection to the Columbia River drainage.  This evidence indicates that Ophiomys may have bypassed the Columbia River during its migration at 4.8 Ma and continued its southward migration along the Pacific coast, moved up the Sacramento River system to Lake Idaho, and then migrated westward along the shores of Lake Idaho and up the present-day Burnt River system to the Always Welcome Inn area (Figure 7).  Ophiomys then migrated westward into the Grande Ronde River drainage and the Columbia River system at ~4.3 Ma as climate grew wetter and the level of the Glenns Ferry stage of Lake Idaho rose towards its peak at ~4 Ma (Figure 8), when Lake Idaho began to drain through the Grande Ronde Valley into the Columbia River system (Livingston, 1928; Van Tassell and others, 2001).

Figure 7.  Possible Ophiomys  migration routes through the western U.S. based on the results of this study.

Figure 8.  Lake Idaho levels (based on Wood and Clemens, 2002) and paleoenvironments in the Grande Ronde Valley (based on Van Tassell and others, 2001).  Ophiomys migrated from the Powder Valley to the Columbia River drainage during a period of rising lake levels and wet climate at ~4.3 Ma.

              Muskrats migrated along a similar route up the Sacramento River system at ~4.1 Ma (Dolomys, Etchegoin Formation, Kettleman Hills, California) and reached the Glenns Ferry lake at Hagerman (Pliopotamys minor) by ~3.7 Ma and the Columbia River basin at ~2.9 Ma, just before Lake Idaho began to drain out through Hells Canyon (Repenning, 1987).  The exact path that Ophiomys and the muskrats followed up the Sacramento River to Lake Idaho during the Pliocene is speculative because the evidence for the late Miocene and Pliocene drainage of Lake Idaho prior to its connection with the Columbia River at ~4 Ma has largely been eliminated by Basin and Range faulting and lava flows.  Similarities in fossil and living freshwater fish (Miller, 1965) and mollusks (Taylor, 1966) suggest an earlier drainage during the late Miocene Chalk Hills phase of Lake Idaho out of southern Idaho along the Owyhee River and through northern California and into the Sacramento River.  It is possible that Ophiomys followed this route and crossed the drainage divide into the Owyhee River and the Lake Idaho system, but it is also possible that Ophiomys migrated up along the drainage route for the Pliocene Glenns Ferry phase of Lake Idaho proposed by Repenning and others (1995):  1) up the Pit river of California to the Alturas Basin, 2) eastward to the Black Rock Desert and Winnemucca in Nevada, and 3) north to the Humboldt River, into the eastern part of northern Nevada, and finally reaching the shore of the Glenns Ferry Lake at the present position of the Snake River near Twins Falls, Idaho.

              Why didn’t Ophiomys migrate up the Columbia River at 4.8 Ma and why did the muskrat Dolomys also bypass the Columbia River and continue its migration south along the Pliocene coastline at ~4.1 Ma?  One possibility is that the ancestral Columbia River was clogged with vitric volcanic glass (hyaloclastic) sands (Mark Ferns and Jason McClaughry, personal communication, 2008).  These vitric sands were derived from the eastern Cascade Range Simcoe volcanics flowing into the ancestral Columbia River, quenching explosively, and then being redeposited downstream as sand, forming the early Pliocene upper Troutdale Formation (Trimble, 1967; Tolan and Beeson, 1984; Swanson, 1986).

DISCUSSION

In the late 1930’s, Claude Hibbard of the University of Michigan and Miklos Kretzoi in Hungary, along with their students, started trying to sort out the complex history of microtine rodents, including voles.  By 1965, Kretzoi recognized that new and more advanced species that appeared abruptly in the European fossil record evolved in the steppe environments of Asia and migrated from there to central Europe.  Kretzoi saw a strong similarity between European and North American fossil arvicolines, but Hibbard was convinced that the North American arvicoline fauna was not the result of invasions from Asia.  Charles Repenning (1987, 2003), a strong supporter of the theory that arvicolines emigrated to North America from Asia, has proposed that the primitive microtine Promimomys immigrated to North America at ~6.7±0.5 Ma and Ophiomys and Cosomys emigrated from Asia to North America at ~4.8±0.2 Ma and evolved into a distinct fauna in the United States by ~4.25±03 Ma.

              Underlying the debate about whether or not these arvicolines migrated from Asia to North America or evolved on the North American continent is considerable disagreement about the classification of these groups of arvicoline rodents.  For example, Cosomys primus was first described in the ~3.0 m.y.-old sequences in the Coso Mountains of southern California, by Wilson (1932), who reasoned that the few differences and the great geographic separation from Europe suggested that assigning this genus to Mimomys was not appropriate.  This conclusion was disputed by Hinton (1932), who felt that Wilson’s Cosomys mimus was a species of Mimomys  When Wilson (1934) discovered Cosomys primus in the fossil beds at Hagerman, Idaho, he described it as Mimomys (Cosomys) primus, agreeing with Hinton’s opinion, but retaining the name Cosomys as a subgenus.  Hibbard (1949, 1952) described a microtine from the Saw Rock Canyon fauna of Kansas as Cosomys primus, stating that the Saw Rock microtine was not Mimomys.  Hibbard (1957) later reassigned the Saw Rock microtine to a different genus and renamed it Ogmodontomys sawrockensis.  Based on lesser hypsodonty of the teeth of Ogmodontomys sawrockensis, Hibbard (1964) suggested it was the ancestor of Cosomys primus, which must have originated from the Ogmodontomys in the Great Plains and migrated westward.  In contrast, Repenning (1987) returned the species Ogmodontomys sawrockensis to the genus Cosmomys, where Hibbard (1941, 1952) had put it (as a genus) and to the genus Mimomys following Hinton (1934) and Wilson (1934) and concluded that Cosomys fossils found in the ~4.8 m.y.-old Maxum fauna of the San Francisco area of California and other faunas in the western United States provided evidence that Cosomys emigrated to North America from Asia.

              As stated by Robert Martin (2006, written communication):  “The problem is not easy, as I’m sure you can tell from the literature.  Most of the archaic voles with rooted molars from the western United States that I have seen are referable to Ophiomys.  The late Charles Repenning and a few others lump all these ancient voles with a three triangle lower first molar into Mimomys, but I think the New World experienced a parallel, endemic radiation to what went on in Eurasia, and our Ophiomys and Ogmodontomys represent a New World clade with its own evolutionary tendencies.  Cosomys is another vole in this group.  At some point, of course, all these forms “regress” to a common ancestor that would look very much like the more primitive Ogmodontomys and Ophiomys.  We do not know if this ancestor gave rise to both Ophiomys and Ogmodontomys in eastern Asia before they dispersed here, or if the common ancestor, perhaps something like Promimomys antiquus, came over here and then diversified."  It is clear that the Always Welcome Inn vole is closely related to the taxa that represent the early radiation of the clade including Ophiomys, Cosomys, and probably Ondatra, but, as stated by Martin (2007, p. 487), "the features that distinguish these genera become obscured as we move back in time, into the "metaregion" (defined by Martin and Tesakov (1998) for the group."

              Our study of the Always Welcome Inn Ophiomys adds a small piece to this puzzle since it suggests that the ancestors of the Always Welcome Inn vole migrated eastward from the Pliocene Pacific Coast rather than westward from the Great Plains and because, like that of Mou (1997), it suggests that the Always Welcome Inn Ophiomys and Ophiomys mcknighti of the Columbia River Basin share a common ancestor with Cosomys primus.  Like Mou (1997), we were not able to differentiate the relationships between many of the vole species that we included in our analysis and we have discovered that more diagnostic criteria will be needed in order to make cladistic analysis a sensitive enough tool to unravel the evolutionary history of primitive voles in North America.  Adding the Always Welcome Inn vole and other species to the cladistic analysis has provided new insights to those of Mou (1997) but it also suggests how important it is to view our results as preliminary because the conclusions are likely to change as other early Pliocene fossil vole sites, like the one at the Always Welcome Inn, are found.

CONCLUSIONS

The lengths and widths of the lower first molars of the Always Welcome Inn vole fall into the same range as those of Ophiomys taylori and Ophiomys cf. mcknighti respectively, but this may reflect a high percentage of juveniles in the population or the small number of teeth that have been found so far.  The characteristics of the Always Welcome Inn vole teeth are most closely related to Ophiomys mcknighti from the ~4.3 m.y.-old White Bluffs fauna of Washington as previously suggested by Van Tassell and others (1997).  A cladistic analysis that incorporates data for 12 species, 5 more than used by Mou (1997), suggests that the Always Welcome Inn Ophiomys and Ophiomys mcknighti are closely related to “Mimomys” sawrockensis from the ~4.8 m.y.-old Upper Alturas fauna of California.  Combined with evidence suggesting connections between the Always Welcome Inn area and Lake Idaho to the east rather than with the Columbia River drainage basin to the west, the analysis suggests that Ophiomys migrated south along the Pliocene Pacific coast at ~4.8 Ma, bypassed the Columbia River, migrated up the Sacramento River to Lake Idaho, moved westward into the Always Welcome Inn area, and then migrated down the Grande Ronde River into the Columbia River system at ~4.3 Ma. 

ACKNOWLEDGMENTS

We are very grateful to the Langrells, the owners of the Always Welcome Inn, for giving us permission to dig at their site and for their enthusiastic support throughout the project.  Leslie Davis and Mark Bork helped sieve vole teeth and also drove Misty back and forth to La Grande to do her research during her pregnancy.  Special thanks go to Nicole Virtue for babysitting Misty and Mark’s daughter Athena and helping with the vole teeth measurements.  Mark Ferns and Jason McClaughry helped solve the puzzle of why the voles may not have migrated up the Columbia River during the early Pliocene.  Special thanks go to Greg McDonald and Bob Martin for help identifying the vole teeth and for suggesting that a cladistics analysis would be a useful tool to try.  Thanks to everyone who has helped find vole fossils at the Always Welcome Inn!

REFERENCES CITED

Barnosky, A.D., 1985, Late Blancan (Pliocene) microtine rodents from Jackson Hole, Wyoming: 

     Biostratigraphy and biogeography: Journal of Vertebrate Paleontology, v. 5, no. 3, p. 255-271.

Czaplewski, N.J., 1990, The Verde local fauna: Small vertebrate fossils from the Verde

     Formation, Arizona: San Bernadino County Museum Quarterly, v. 37, no. 3, p. 1-39.

Gustafson, E.P., 1978, The vertebrate faunas of the Pliocene Ringold Formation, south-central

     Washington: Bulletin No. 23, Museum of Natural History, University of Oregon, Eugene,

     Oregon, 62 p.

Hibbard, C.W., 1941, Mammal from the Rexroad fauna of the upper Pliocene of southwestern

     Kansas: Kansas Academy of Science Transactions 44, p. 265-313.

Hibbard, C.W., 1949, Pliocene Saw Rock Canyon fauna in Kansas:  University of Michigan

     Museum Contributions 7, p. 91-105.

Hibbard, C.W., 1952, Vertebrate fossils from late Cenozoic deposits of central Kansas:  University

     of Kansas Paleontologic Contributions, Vertebrata 2, p. 1-14.

Hibbard, C.W., 1956, Vertebrate fossils from Meade Formation of southwestern Kansas: 

     American Journal of Science, v. 256, p. 51-59.

Hibbard, C.W., 1957, Two new Cenozoic microtine rodents: Journal of Mammalogy, v. 38, p. 39-

     44.

Hibbard, C.W., 1959, Late Cenozoic microtine rodents from Wyoming and Idaho:  Papers of the 

     Michigan Academy of Science, Arts, and Letters, v. 44, p. 3-40.

Hibbard, C.W., 1964, A contribution to the Saw Rock Canyon local fauna of Kansas: Papers of

     the Michigan Academy of Science, Arts, and Letters 49, p. 115-127.

Hibbard, C.W., and Zakrzewski, R.J.,1967, Phyletic trends in the Late Cenozoic microtine,

     Ophiomys gen. nov., from Idaho: Contributions from the Museum of Paleontology, University

     of Michigan, v. 21, no. 12, p. 255-271.

Hinton, M.A.C., 1932, Not on “Cosomys” Wilson, from the Pliocene of California: Journal of

     Mammalogy, v. 13, p. 280-281.

Leithner, A., and Van Tassell, J., 2005, Pliocene vole fossils, Always Welcome Inn, Oregon:

     Eastern Oregon Geology, v. 3, http://www.eou.edu/~geology/Leithner.html

Lich, D.K., 1990, Cosomys primus, a case for stasis: Paleobiology, v. 16, no. 3, p. 384-395.

Livingston, D.C., 1928, Certain topographic features of northeastern Oregon and their relation to

     faulting:  Journal of Geology, v. 36, no. 8, p. 694-708.

Martin, R.A., and Tesakov, A., 1998, Introductory remarks. Does Allophaiomys exist? in Martin,

     R.A., and Tesakov, A., ed., The early evolution of Microtus, Paludicola, v. 2, p. 1-7.

Martin, R.A., 2007, Arvicolidae, in Janis, C.M., Gunnell, G.F., and Uhen, M.D., eds.,Evolution of

     Tertiary Mammals of North America, v. 2, Cambridge, Cambridge University Press, p. 480-

     497.

Mou, Y., 1997, A new arvicoline species (Rodentia, Cricetidae) from the Pliocene Panaca

     Formation, southeast Nevada: Journal of Vertebrate Paleontology, v. 17, no. 2, p. 376-383.

Repenning, C.A., 1987, Biochronology of the microtine rodents of the United States, in

     Woodburne, M.O., ed., Cenozoic mammals of North America: Geochronology and

     biostratigraphy, Berkeley, University of California Press, p. 236-268.

Repenning, C.A., 2003, Chap. 17, Mimomys in North America: Bulletin of American Museum of

     Natural History, no. 279, p. 469-512.

Repenning, C.A., Weasma, T.R., and Scott, G.R., 1995, The early Pleistocene (latest Blancan-

     earliest Irvingtonian) Fromans Ferry fauna and history of the Glenns Ferry Formation,

     southwestern Idaho:  U.S. Geological Survey Bulletin 2105, p. 1-86.

Smith, G.R., Morgan, N., and Gustafson, E., 2000, Fishes of the Pliocene Ringold Formation of

     Washington and history of the Columbia River drainage: University of Michigan Museum of

     Paleontology Papers on Paleontology, v. 32, 42 p.

Shotwell, J.A., 1970, Pliocene mammals of southeast Oregon and adjacent Idaho:  Museum of

     Natural History, University of Oregon, Bulletin no. 17, 103 p.

Swanson, R.D, 1986, A stratigraphic-geochemical study of the Troutdale Formation and Sandy

     River mudstone in the Portland Basin and Lower Columbia River Gorge: M.S. thesis,

     Portland State University, 103 p.

Taylor, D.W., 1966, Evolution of freshwater drainages and molluscs in western North America, in

     Smiley, C.J., Leviton, A.E., and Berson, M., eds., Late Cenozoic history of the Pacific

     Northwest: Interdisciplinary studies on the Clarkia fossil beds of northern Idaho: San

     Francisco, American Association for the Advancement of Science, Pacific Division, p. 265-

     321.

Tolan, T.L., and Beeson, M.H., 1984, Intracanyon flows of the Columbia River Basalt Group in the

     lower Columbia River Gorge and their relationship to the Troutdale Formation: 

     Geological Society of America Bulletin, v. 95, p. 463-477.

Trimble, D.E, 1967, Geology of Portland, Oregon and adjacent areas: U.S Geological Survey

     Bulletin 1119, 119 p.

Van Tassell, J., Ferns, M., McConnell, V., and Smith, G.R., 2001, The mid-Pliocene Imbler fish

     fossils, Grande Ronde Valley, Union County, Oregon, and the connection between Lake Idaho

     and the Columbia River:  Oregon Geology, v. 63, no. 3, p. 77-84, 89-96.

Van Tassell, J.; Bergey, E.; Davis, C.; Davis, M.; Grimshaw; B., Kisselburg; J., Ledgerwood, R.;

     Miller, S.; Morris, C.; Steele; J., Wehymiller, C.; Ferns, M.L.; Smith, G.R.; McDonald, H.G.;

     Mead, J.I.; and Martin, R.A., 2007, Early Pliocene (Blancan) Always Welcome Inn local fauna,

     Baker City, Oregon: Oregon Geology, v. 68, no. 1, p. 3-23.

Wilson, R.W., 1932, Cosomys, a new genus of vole from the Pliocene of California:  Journal of

     Mammalogy, v. 13, p. 150-154.

Wilson, R.W., 1934, A rodent fauna from later Cenozoic beds of southwestern Idaho:

     Contributions to Paleontology, Carnegie Institute of Washington, v. 473, p. 15-34.

Wood, S.H., and Clemens, D.M., 2002, Geologic and tectonic history of the western Snake River

     Plain, Oregon and Idaho, in Bonnichsen, B., White, C.M.,  and McCurry, M., eds., Tectonic

     and magmatic evolution of the Snake River Plain volcanic province:  Idaho Geological Survey Bulletin 30, p. 69-103.

Zakrzewski, R.J., 1969, The rodents from the Hagerman local fauna: Contributions from the

     Museum of Paleontology, The University of Michigan, v. 23, no. 1, p. 1-36.