OCCASIONAL PAPERS

Museum of Texas Tech University Number 235 5 May 2004

HISTORICAL ZOOGEOGRAPHY AND TAXONOMIC STATUS OF THE PRAIRIE ( OCHROGASTER) FROM THE SOUTHERN PLAINS OF TEXAS AND OKLAHOMA

FREDERICK B. STANGL, JR., JIM R. GOETZE, AND KIMBERLY D. SPRADLING

The (Microtus ochrogaster) today is with distinct cranial characters andpale pelage colora­ considered a species ofthe Great Plains, whose range tion, they described M. o. taylori. extends from the interior grasslands ofsouthern Canada to central Oklahoma and the northernTexas Panhandle, Choate and Williams (1978) perfonned a mor­ and is generally bordered by the Rocky Mountains to phometric analysis ofthe prairie vole, which included the west and the eastern deciduous woodlands (Bee et material reported onby Hibbard and Rinker (1943) as al. 1981; Hall 1981; Hoffmann andKoeppl 1985; Caire M. o. taylori. Each of their external body measure­ et al. 1990; Stalling 1990). Duringthe late Pleistocene, ments and four of the seven cranial characters (skull the species occurred throughout much ofTexas, leav­ length, zygomatic breadth, mastoidal breadth, and least ing fossil evidence as far south as the Edwards Pla­ interorbital breadth) exhibited significant geographic teau and parts ofthe GulfCoast (Dalquest and Schultz variation. While no geographic patterns were evident 1992; Graham 1987). The species retreated north­ to suggest subspecific levels of differentiation, popu­ ward with the advent ofincreasingly arid climate dur­ lations from the Dakotas to Kansas demonstrated a ing the Holocene. At least one isolated population, M. slight north-to-south elinal increase in size of cranial o. ludovicianus, was stranded during this time, and is characters. Among their conclusions were that the known to have survived in extreme southeasternTexas population referred to as M. o. taylori by Hibbard and and adjacent Louisiana until the earliest 1900s (Bailey Rinker (1943) no longer exists. They hypothesized 1905). that either the type locality and vicinity was colonized by adjacent populations following extirpation of the Hibbard and Rinker (1943) reported a series of original taylori stock by drought, or that genetic M. ochrogaster taken in 1942, which extended the swamping enveloped the few remaining drought sur­ known range ofthe prairie vole southward into Meade vivors. Consequently, in their view, all M. ochrogaster County of southwestern Kansas. The were from the Central Plains (which would include all prai­ found to be restricted to the few bogs and meadows, rie of Nebraska, Kansas, Colorado, and Okla­ having survived increasingpressures ofagriculture and homa) should be assigned to the subspecies M. o. a recent drought. On the basis of these specimens, haydenii. considered by them to represent a relictual population 2 OCCASIONAL PAPER", MUSEUM OF TEXAS TECH UNIVERSITY

Since 1986, populations ofthe prairie vole have ofthe late Pleistocene or early Holocene. He referred been reported from south-central Oklahoma (Choate his specimens to M. o. haydeni ongeographic grounds. 1989; Smith 1992a), and the panhandles ofOklahoma Even at a glance, the pelage coloration of these ani­ (Dalquest et al. 1990a; Reed and Choate 1988) and mals is dark--comparable to reference series from Texas (Choate and Killebrew 1991; Jones et al. 1988; Kansas, Nebraska, and Missouri--and in striking con­ Manning and Jones 1988). Each ofthese populations trast to the pale pelage of specimens from southern represents marginal extensions to the south ofthe pre­ Colorado and the panhandle regions ofOklahoma and viously mapped ranges ofthe species. Reed and Choate Texas. (1988) suggested that their Oklahoma panhandle speci­ mens represented an invasion byM. o. haydenii stock, This studycharacterizes the morphometric varia­ but others (Dalquest et al. 1990a; Manning and Jones tion ofsouthern peripheral populations ofprairie voles, 1998) have tentatively referred specimens from the and provides the ftrst critical review ofthe species to Oklahoma andTexas panhandle regions to M o. taylorl. include specimens from Texas and Oklahoma. We examine the historical zoogeography and subspeciftc Presence of the prairie vole in southern Okla­ status of the species from those states, with special homa first was documented by Choate (1989), who attention to the question ofM. o. taylori as an extant considered the south-central Oklahoma prairie vole taxon and its validity as a subspecies. population from the Wichita Mountains to be a relict

MATERIALS AND MEmoDs

Specimens of Microtus ochrogaster from the lary toothrow) were taken with digital calipers and re­ collections ofMidwestern State University, Texas Tech corded to the nearest 0.01 mm. We depart from their University, and Oklahoma State University were sur­ protocol by excluding standard body measurements veyed from southern localities of the species' range. on the basis ofinescapable variations in methodology Of these, 129 specimens representing adult individu­ and skills of different preparators. als (two oldest of four age classes, sensu Choate and Williams 1978) taken from three populations at the Sexes were pooled for all subsequent analyses southern periphery ofthe species' range in Oklahoma after two-way (sex, age) multiple analysis ofvariance and Texas (Figure 1), and reference series from Colo­ (MANOVAs) for eachmeasurement revealed that sexual rado (western peripheral population) and Kansas (in­ dimorphism and interactions between age and sex were terior sample) were selected for ftnal study. Each of not signiftcant. the three nominal subspecies (sensu Hall 1981; Hoffmann and Koepp11985; Stalling 1990) are repre­ Age variation between the two adult age classes sented in the analyses: M o. taylorl from the Texas in our study was signiftcant for each character except and Oklahoma panhandles; M o. haydenii from south­ least interorbital breadth and prelambdoidal breadth ern Colorado and southern Oklahoma; and M o. (one-way MANOVAs; P = 0.2987 and P = 0.1779, ochrogaster from eastern Kansas. However, we note respectively), and only the least interorbital breadth that Choate and Williams (1978) have synonymized (one-way MANOVA, P:s 0.0001) among young adult M o. ochrogaster with M o. haydenii, and we follow specimens exhibited signiftcant geographic variation. their revision. Based on these fmdings, further geographic analyses were restricted to four samples comprised of speci­ Seven cranial measurements described by Choate mens representing the oldest of the four age catego­ and Williams (1978; length ofskull, zygomatic breadth, ries. Only for the Oklahomaffexas panhandles sample mastoidal breadth, least interorbital breadth, was it necessary to pool individuals from more than prelambdoidal breadth, length of rostrum, and maxil- one collecting locality. STANGL ET AL-PRAnm: VOLE (MICROTUS OCHROGASTER) 3

--- H ;~ --- I 36 , \ 3t- I H f \

0 300 : + KM N I I 102 96

Figure 1. Southern distribution and superimposed locality records for Microtus ochrogaster. Holocene records from Dalquest et al. (1990b) are indicated by "H". Site represented by skulls examined from pellets is indicated by "X". Nominal subspecific taxa (sensu Hall 1981 ) repre­ sented are: a) M. o. haydenii; b) M. o. taylori; and c) M. o. ochrogaster.

Principal components analysis ofthe covariance (MWSU 14219, 14221, 14222, 14224, 14226, 14431, matrix was usedto assess patterns ofgeographic varia­ 14463, 14474, 14476, 14569, 14632, 14633, 14741, tion between populations, and discriminant functions 14776, 14777, 14854-14856, 14858-14860, 14862, analyses were applied to assess uniformity ofdistinct­ 14872, 14873, 15018, 15019, 15068, 15069, 15594, ness ofthose samples. Duncan's Multiple Range tests 15741); Mo. ochrogaster-KANSAS: Franklin Co., were used to examine morphometric associations be­ 1-5 mi. S Ottawa,19 (MWSU 5367, 5368,7176-7180, tween populations. All computations were accom­ 7191, 7483-7485, 9186, 9188, 9192-9197, 15960); plished with NCSS, Version 5.3 (Hintze 1990). M o. taylori--OKLAHOMA: Beaver, Co., 3.8 mi. W Bryan's Comer, 1 (MWSU 16508); Texas Co., 11.1 Specimens examined.--Following is the list of mi. E Guymon, 1 (MWSU 15674); 4 mi. NE Tyrone, "old adult" individuals (sensu Choate and Williams, 2 (MWSU 15715, 15735); TEXAS: Hansford Co.: 8 1978), listed by nominal subspecific designation, that mi. S Speannan, 4 (ITU52211,52449,52450,52423). comprise the basis for this study: M. o. haydenii-­ Specimens from Texas are deposited in The Museum COLORADO: Huerfano Co., 4 mi. W Walsenburg, 6 of Texas Tech University (TTU), and all others are (MWSU 18156-18159, 18161,18162); OKLAHOMA: from the Collection of Recent at Midwest­ Comanche Co., Fort Sill, Wichita Mountains, 30 ern State University (MWSU). 4 OCCASIONAL PAPERS, MUSEUM OF TEXAS TECH UNIVERSITY

REsULTS

Each ofthe cranial characters, except zygomatic tion for five ofthe seven measured characters. These breadth in old adult Microtus ochrogaster, exhibited differences are significant (Duncan's Test at P ~ 0.05 highly significant geographic variation between the level) for length ofskull, rostral length, and length of studiedpopulations (Table 1). The interiorsample from maxillary toothrow. Colorado is represented by an as­ Kansas averaged smaller than any peripheral popula- semblage ofparticularly robust specimens, averaging

Table 1. Comparison ofcranial measurements (in mm) of63 "old adult" specimens ofthe prairie vole (Microtus ochrogaster) from four southern populations. Descriptive statistics are sample size (N), arithmetic mean, standard deviation (SD), extreme measurements (minimum-maximum), confidence intervals (CI), and coefficients ofvariation (CV). Texas/Oklahoma Wichita Mountains, Huerfano Co., Franklin Co., panhandles southern Oklahoma southern Colorado eastern Kansas (N= 8) (N = 30) (N=6) (N = 19) Character Mean±SD Mean±SD Mean±SD Mean±SD (Min. -max.) (Min. -max.) (Min. -max.) (Min. -max.) 95%CI 95%CI 95%CI 95%CI 01 01 01 01 Skllll 28.60:1:0.94 28.38±O.70 28.52±0.83 27.33±O.76 length (27.29-30.13) (26.52-29.64) (28.77-30.80) (25.71-28.56) •• 27.82-29.39 28.12-28.65 28.52-30.25 27.00-27.67 3.29 2.47 2.82 2.56

Zygomatic 16.09±O.74 15.99±O.51 16.12±O.28 15.59:1:0.43 breadth (15.18-17.15) (l4.92-17.03) (15.80-16.53) (14.70-16.32) N.S. 15.48-16.70 15.79-16.18 15.83-16.41 15.38-15.79 4.58 3.16 1.72 2.75

Mastoid 12.27±0.45 12.63±O.31 12.36±O.23 12.16±0.35 breadth (11.84-13.87) (12.05-13.27) (12.14-12.74) (11.58-12.74) •• 1I.89-12.64 12.51-12.75 12.11-12.60 11.99-12.33 3.70 2.47 1.88 2.90

Interorbital 3.86±0.31 4.31±O.19 4.16±O.16 4. 17±0.22 breadth (3.16-4.15) (3.90-4.71) (4.06-4.47) (3.69-4.69) •• 3.61-4.12 12.51-12.75 12.11-12.60 11.99-12.33 3.70 2.47 1.88 2.90

Prelamdoidal 9.89±0.46 10.52±O.27 10.33±O.38 10.14±0.41 breadth (9.13-10.50) (7.65-9.36) (8.93-9.84) (7.42-8.62) •• 9.50-10.27 10.42-10.63 9.94-10.73 9.94-10.33 4.69 2.54 3.66 4.00

Rostral 8.77±O.35 8.62±0.38 9.24±0.39 8.00:1:0.36 length (8.23-9.19) (7.65-9.36) (8.93-9.84) (7.42-8.62) •• 8.48-9.06 8.48-8.76 8.82-9.65 7.83-8.18 3.94 4.34 4.25 4.52

Maxillary 6.4I±O.31 6.42±0.22 6.68±O.07 6. 16±O.27 toothrow (6.02-7.01) (6.07-6.94) 6.56-6.79) (5.52-6.70) length 6.15-6.66 6.34-6.51 6.60-6.76 6.02-6.29 •• 4.81 3.45 1.13 4.38 One-wayMANOVA: ··=P s;O.OOI STANGL ET AL-PRArmE VOLE (MICROTUS OCHROGASTER) 5 significantlylargerthan all otherpopulations for zygo­ highly correlated with the first factor, which contrib­ matic breadth, rostral length, and maxillary toothrow. uted more than half (51.1%) ofthe reported variabil­ The southern Oklahomapopulationis significantlylarg­ ity. The second factor accounted for 19.9% of the est for mastoid, interorbital, and prelambdoidal mea­ variation, and was most influenced byprelambdoidal, surements. The pooled sample from the Texas and mastoidal, interorbital, and zygomatic measurements. Oklahoma panhandles is most variable with highest The third factor accounted for 10.6% ofthe variation. coefficients ofvariation for all measurements but ros­ trallength, and is characterized by a significantly nar­ Discriminant functions analysis ofthe seven cra­ row least interorbital breadth. nial measurements from all four samples served to accurately predict the geographic origin for 81% (50 The first three principal components accounted of 62) ofthe individuals. When this analysis was re­ for 81.6% ofthe total variation between populations. stricted to the three peripheral populations (exclusion Ofthe two-dimensional scatterplots examined, the first of the Kansas sample), 88% were correctly assigned two components provided clearest separation ofspeci­ to locality. Misclassifications occurred within each mens by region-of-origin (Figure 2). All characters sample and appeared to be random in effect. butprelambdoidal and interorbital measurements were

2

o N U a..

•- southern OK • - OKITX panhandles A- southern CO 0- eastern KS

-4 -1------.------...------.------,...------1 -2 o 3 PC 1 Figure 2. Scatter plot offirst two principal components for four populations ofMicrotus ochrogaster. 6 OCCASIONAL PAPERS, MUSEUM OF TEXAS TECH UNIVERSITY

DISCUSSION

Historical Climatology ing the past century (Figure 3), with the exceptions of warming in Lubbock, cooling in Wichita Falls, and in­ The southern Great Plains ofthe late Pleistocene creasing moisture in Oklahoma City. was a cooler, moister, and more equable time than the prevailing conditions oftoday (Graham 1987). Pollen Historical zoogeography of southern records suggest that expanses of sandsage (Artemi­ peripheral populations sia) grasslands dominated the western High Plains, and comparativelymesic prairies cloaked the RollingPlains Southern Oklahoma.-The southernmost popu­ ofTexas and Oklahoma to the east. These vegetation lation ofMicrotus ochrogaster may well have occu­ associations likely had no modern analog (Hall and piedthe WichitaMountains ofsouth-central Oklahoma Valastro 1995), but conditions clearly favored such continuously since the Pleistocene. The fossil record mesic grasslands forms as Microtus ochrogaster; its in the immediate vicinity is problematic, for the loose remains occur in Pleistocene sediments throughout the teeth that usually represent this species in fossiliferous region strata cannot be distinguished from those ofthe wood­ land vole (M pinetorum), a locally sympatric species Emergence from the Pleistocene into a warmer that also occurs in the Pleistocene. Holocene deposits and drier Holocene has been anything but gradual and from two locations within 30 miles ofthese hills, and regionally uniform. Studies ofsedimentarydeposition dated at 2,000 YBP, contain microtine remains that and erosion across numerous sites of the High Plains might be either ofthese species (Goetze 1989; Smith of the Texas Panhandle suggest a prolonged drought 1992b). However, each of these two sites also con­ between 6,500 and 4,500 YBP (Holliday 1989). Evi­ tained the remains of prairie indicator species (thir­ dence from buried trees and water table fluctuations teen-lined ground squirrel, Spermophilus in the canyons ofthe southern OklahomaRollingPlains tridecemlineatus; hispid pocket mouse, Chaetodipus (Hall and Lintz 1984) indicate a local shift to a moister hispidus; hispid cotton rat, Sigmodon hispidus), sug­ environment about 2,600YBP, which coincides with a gesting that at least some ofthese microtine teeth might mesic interlude in the desertification of Trans-Pecos represent M ochrogaster. Texas (Bryant 1977). These conditions apparently lasted until about 1,000YBP, which marked the advent The rugged granite hills, a mesic refugium marked ofthe modern trend toward increasing semiarid con­ by wooded slopes and rocky soils, were never as at­ ditions (Hall and Lintz 1984). tractive as the surrounding level plains to homestead­ ers ofthe 1880s, thereby protecting comparatively lush Historical deviations from the current progres­ grassy parks from the plow and extensive overgraz­ sion towards a progressively warmer and drier envi­ ing. ronment are sometimes of considerable duration and magnitude. The "Little Ice Age" of 1450-1890 A.D. The prairie vole only rarely is encountered else­ was a period ofincreased precipitation and coolertem­ where today in southern Oklahomabeyondthe Wichita peratures that was apparently global in effect (Crowley Mountains (Smith 1992a; Stangl et al. 1992), but it is and North 1991). locally common within the protective confines ofthese hills, from where it was first reported by Choate Modem weather patterns ofthe Southern Plains (1989). Absence of any earlier documentation seems are subject to dramatic fluctuations, and can vary re­ only a matter ofthe lack ofsystematic collecting, as is gionally. Sustained temperature extremes over weeks evident from an earlier compilation ofsmall or months, and droughts ofup to several years in du­ records from the Wichita Mountains by Glass and ration, are not uncommon, and each can have a locally Halloran (1961). Much of this isolated range of this devastating effect on resident species. However, few vole falls within the boundaries ofthe Wichita Moun­ significant long-term changes in annual mean climatic tains National Wildlife Refuge (first established as a variables have occurred over the Southern Plains dur- national forest in 1901) and Fort Sill (an army installa- STANGL ET AL-P1w:mE VOLE (MICROTUS OCIIROGASTER) 7

Annual Mean Temperature (OF) Annual Total Precipitation (in) '0.0,-- -, 60."1------__-,

6s.l1 45.0 AMARILLO

,= +0.002 P=N.S. 5Oi~:t----:-'m~---,aSO~--'9~1S---2OOO-I 0;~8OO!:---::,9::':2S:----'-9SO~--'9~1S---2-JOOO

'0.0,------_-, 6O.O,- ~ '* -0.063 P=N.S. 65.0 45.0 LUBBOCK 50. 30.0

55. '5.0 ,= +0.282 P = < 0.01 501~:t---,~"'"'---,9S0~--'9~1-S--2000-1 °i~:t---I92~S---'9S0~--,9~1S---2000-I

10.0,------, 60.0,.------,

6S "'.0 WICHITA

FALLS 60.0 30.0

55.0 15.0 ,= -0.528 ,=+0.187 P < 0.001 P=N.S. SO.O+---~--~--~--_l 00+---~--~--~--_l 1900 1925 2000 1900 1025 laso 1815 2000

'00,------, 60.0>,------,

65.0 OKLAHOMA CITY 6O·~'jS~R.,'?1

55.0

r = -0 078 r = +0.238 P =N.S. P = <0.05 SO.C-I---~--~---~----1 O.O+---~--~--~--_./ ,aoo '9'" '950 2000 1800 HI2S '950 2000

10.0.,.------, 00.0.,.------,

6'0 045.0

DALLAS 600 30.0

SSO ,=+0.191 ,=+0.047 P= N.S. P=N.S. SO.O+----~--~--~--__:__I OO+---~--~--~--__:__I 1900 1925 1950 1915 2000 19oo UI2S 1950 '015 2000

Figure 3. One-hundred year weather trends from five cities in Texas and Oklahoma that bracket the southern periphery of the range of Microtus ochrogaster. Data of annual means for temperature and precipitation are from National Atmospheric and Oceanic Administration. 8 OCCASIONAL PAPERS, MUSEUM OF TEXAS TECH UNIVERSITY tion of 98,000 acres, established in 1869), which to­ Oklahoma panhandles, represent a reclamation ofear­ day provide security for this relict population. lier abandoned range, or simply the discovery of the species in an area that has only in the past two de­ Texas/Oklahoma panhandles.-The semiarid cades been systematically collected. Earlier and more short-grass prairies of the High Plains provide mar­ occasional collecting efforts might have easily over­ ginal habitat today for the prairie vole. The fact that looked this locally uncommon and sporadically dis­ recent range extension reports in the region (Manning tributed species. and Jones 1988; Reed and Choate 1988; Dalquest et al. 1990a; Choate and Killebrew 1991) are from along Taxonomic status of Southern Plains transportation routes may be significant. Fencerows, prairie voles pavement runoff, and water collection along railroad andhighwayrights-of-waypromote comparativelylush The morphometrically distinct ofeach of vegetation. These conditions offerbothadequate con­ the four populations examined is consistent with the ditions and avenues ofdispersal, as well as protection findings of Choate and Williams (1978) from across from livestockovergrazingonadjacentrangelands that the northern range ofMicrotus ochrogaster, and repre­ are, in any event, privately owned and not usually ac­ sents a continuum of the patterns of distribution ob­ cessible to collectors. served by those authors. We attribute the magnitude ofdifferentiation ofspecimens in our study area to the The semiaridshort-grass prairie ofthe High Plains degree ofisolation expected from a less-than-uniform currently occupied by western peripheral populations distribution atthe periphery, where conditions are gen­ ofthe prairie vole provide only locally marginal habi­ erally marginal at best. Areas ofoccupation are prob­ tat. The history ofthe prairie vole in the Texas/Okla­ ably best portrayed as scattered isolates or tendril-like homapanhandles and adjoiningparts ofColorado, New extensions into areas ofsuitable habitat--with little or Mexico, and Kansas extends well back into the Pleis­ no lateral gene flow. With the exception of Texas/ tocene (Hibbard 1949; Graham 1987; Dalquest and Oklahoma panhandle populations, we follow Choate Stangl 1989; Dalquest and Schultz 1992). Several ex­ andWilliams (1978) inreferring prairie voles from this tralimital carbon-dated Holocene deposits (Dalquest et study to M. o. haydenii, because these populations al. 1990b) from the Oklahoma Panhandle and north­ seem only to represent the extreme ofa cline ofmor­ eastern New Mexico (Figure 1) document the pres­ phometric differentiation and increase in size. How­ ence ofM ochrogaster at various times during an in­ ever, the distinctive population referred to as M o. terval ofat least 3,860-530 YBP. This does not neces­ taylon by Hibbard and Rinker (1943) remains an ex­ sarily suggest a continuous regional occurrence ofthe tant entity. prairie vole, but more likely, a period of time during which the species locally has experienced a long ebb­ In their original diagnosis of M. o. taylori, and-flow history ofperipheral range fluctuations. Hibbard and Rinker (1943) placed considerable em­ phasis on the distinctive interorbital ridge. Choate and One-hundred year weather data for the region Williams (1978) correctly noted that the development (Figure 3) demonstrate few significant weather trends ofthis feature is a function ofthe distinctively narrow over that period, although pronounced episodes such interorbital breadth (lOB) of M. o. taylori. This is as the Dust Bowl era and droughts of the late 1990s clearly evident with the illustrations providedbyHibbard have occurred in the interim. The discovery of M and Rinker (1943:257, Figure 1), and yet they failed to ochrogaster by Hibbard and Rinker (1943) in south­ refer specifically to the comparatively narrow lOB, or western Kansas came after the 1930's Dust Bowl era, to include that character in their diagnosis. and extended the known range into extreme southern Kansas at a time when local populations were appar­ The extreme constriction of the interorbital ently restricted to lush growth fringing the few re­ breadth is a defining and readily visible feature ofM maining isolated bogs ofthe region. o. taylori, and one which exemplifies specimens taken in recent years from the Texas/Oklahoma panhandles It is difficult to discern if the more recent dis­ (Table 1). Fortunately, this component ofthe skull is coveries since the late 1980s to the south, in the Texas/ durable, and often remains intact after disintegration STANGL ET AL.-PRAm.m. VOLE (MICROTUS OCHROGASTER) 9 and loss of other cranial components (Stangl et a1. tions is not unusual among voles. This phenomenon 1993), and is a character that does not usually differ has been demonstrated in peripheral populations for by age or sex in the prairie vole (Choate and Williams an array of microtine taxa, suggesting that increased 1978). This permits us to assign the fragmented re­ size confers some locally adaptive advantage for cop­ mains from owl pellets (N = 10, mean = 3.94) re­ ing with the stresses of existing in marginal habitat ported by Dalquest and Baskin (1991) from the Okla­ that could logically pertain to the periphery of a spe­ homa Panhandle to M o. taylorl. Similarly, we can cies' range. date this subspecies back to at least 530 ± 60 YBP from Trampiros Creek of Union Co., New Mexico In support ofhis proposed model ofcentrifugal (Dalquest et al. 1990b), on the basis ofthe fortuitous speciation, Brown (1957) noted that peripheral popu­ preservation and recovery ofthe braincase ofa speci­ lations ofEuropean microtines tendedto be more vari­ men (lOB = 3.82) from ancient pond sediments. able and larger in size than their interior counterparts. Chitty (1958) initiated a series of studies (later sum­ Ofparticular interest is the status of southwest­ marized by Krebs [1978]) which detailed the attain­ ern Kansas populations. The utility ofthe lOB mea­ ment oflarge size ofa suite ofmorphological charac­ surement as a diagnostic tool transcends both spatial ters among microtine populations as an adaptive re­ and temporal bounds, does not vary by sex or age in sponse to environmental stress. More recently, ex­ the prairie vole (Choate andWilliams 1978) and canbe ceptionally large size was noted for individuals among applied to all but the most damaged and fragmentary populations of M califomicus recovering from the skull. It should prove useful to test the hypothesis of rigors of peak densities, and usually associated with Choate and Williams (1978)--thatM o. haydenii is the poor habitat (Lidicker and Ostfeld 1991). current prairie vole of southwestern Kansas, having occupied the historical range ofM o. taylorl. The large size ofSouthern Plains prairie voles is certainly consistent with the clinal increase in size re­ Peripheral populations and large size ported by Choate and Williams (1978) for the species, in microtine but are the stresses of"life on the frontier" contribu­ tory? Examination of peripheral populations of this Finally, we note that the especially large and ro­ species to the north, east, and west could prove infor­ bust nature ofperipheral Microtus ochrogaster popula- mative.

ACKNOWLEDGMENTS

For arranging the loan ofcomparative materials J. Baker made available the Texas Tech specimens for that were not included in the final analyses, we thank our study. Catherine Parker and Clyde Jones offered Karen McBee of Oklahoma State University. Robert comments on an earlier draft ofthis manuscript.

LITERATIJRE CITED

Bailey, V. 1905. Biological survey ofTexas. North Ameri­ Bryant, V. B., Jr. 1977. Late Quaternary pollen records can Fauna, 25:1-222. from the east-central periphery of the Chihuahuan Desert. pp. 3-21, in Transactions of Bee, J. W., G E. Glass, R. S. Hoffmann, and R. R. Patterson. the symposium of the biological resources of the 1981. Mammals in Kansas. University of Kan­ Chihuahuan Desert region, United States and sas Museum of Natural History, Public Educa­ Mexico (R. H. Wauer and D. H. Riskind, eds.). tion Series, 7:ix + 1-300. National Park Service, Proceedings and Trans­ Brown, W. L., Jr. 1957. Centrifugal speciation. Quarterly actions Series, 3:xxii + 658 pp. Review in Biology, 32:247-277. 10 OCCASIONAL PAPERS, MUSEUM OF TEXAS TECH UNIVERSITY

Caire, W., J. D. lYler, B. P. Glass, and M. A. Mares. 1990. Goetze, J. R. 1989. Mammalian faunas of a late Pleis­ Mammals ofOklahoma. University ofOklahoma tocene-Holocene terrace ofthe Red River, Tillman Press, Norman, xiii + 567. County, Oklahoma. Texas Journal of Science 41:205-209. ' Chitty, D. 1958. Self-regulation of numbers through changes in viability. Cold Spring Harbor Sympo­ Graham, R. W. 1987. Late Quaternary mammalian faunas sium in Quantitative Biology, 22:277-280. and paleoenvironments of the southwestern Choate, J. R., and S. L. Williams. 1978. Biogeographic plains of the United States. pp. 24-86, in Late interpretations of variation within and among Quaternary mammalian biogeography and envi­ populations of the prairie vole, Microtus ronments of the Great Plains and prairies (R. W. ochrogaster. Occasional Papers, Museum, Texas Graham, H. A. Semken, Jr., and M. A. Graham, Tech University, 49:1-25. eds.). Illinois State Museum Science Papers, 22:xiv+ 1-491. Choate, L. L. 1989. Natural history ofa relictual popula­ tion of the prairie vole, Microtus ochrogaster, in Hall, E. R. 1981. The mammals of . John southwestern Oklahoma. Occasional Papers, Wiley & Sons, New York, 2nd ed., 2:vi + 1-600 + Museum, Texas Tech University, 129:1-20. 91. Choate, L. L., and F. C. Killebrew. 1991. Distributional Hall, S. A., and C. Lintz. 1984. Buried trees, water table records of the California myotis and the prairie fluctuations, and 3000 years of changing climate vole in the Texas Panhandle. Texas Journal of in west-central Oklahoma. Quaternary Research, Science, 43:214-215. 22:129-133. Crowley, T. 1., and G R. North. 1991. Paleoclimatology. Hall, S. A. ,and S. Valastro, Jr. 1995. Grassland vegetation Oxford Monographs in Geology and Geophys­ in the southem Great Plains during the last gla­ ics, Oxford University Press, New York, 16:1-339. cial maximum. QuaternaryResearch,44:237-245. Dalquest, W. w., and J. A. Baskin. 1991. Local abundance Hibbard, C. W. 1949. Pleistocene stratigraphy and pale­ of prairie voles in Beaver County, Oklahoma. ontology of Meade County, Kansas. Contribu­ Texas Journal ofScience, 43:104-105. tions, Museum of Paleontology, University of Michigan, 7:63-90. Dalquest, W. W., and G E. Schultz. 1992. Ice Age mam­ mals of northwestern Texas. Midwestern State Hibbard, C. W., and G C. Rinker. 1943. A new meadow University Press, Wichita Falls, Texas, 309 pp. mouse (Microtus ochrogaster tay/ori) from Meade County, Kansas. University of Kansas, Dalquest, W. w., and F. B. Stangl, Jr. 1989. Late Pleis­ Science Bulletin, 29:255-268. tocene Mammals from the northwestern comer of the Oklahoma panhandle. Texas Journal of Hintze, J. L. 1990. Number cruncher statistical systems. Science, 41:35-47. Version 5.3. Pacific Ease Co., Santa Monica, Cali­ fornia, 442 pp. Dalquest, W. W., F. B. Stangl, Jr., and 1. K. Jones, Jr. I990a. Mammalian zoogeography ofa Rocky Mountain­ Hoffmann, R. S., and J. W. Koepp!. 1985. Zoogeography. Great Plains interface in New Mexico Oklahoma Pp. 84-115, in Biology of the New World Micro­ and Texas. Special Publications, Mu~eum, Tex~ tus (R. H. Tamarin, ed.). Special Publication, Tech University, 34:1-78. American Society of Mammalogists, 8:xiii + 1­ 893. Dalquest, W. w., F. B. Stangl, Jr., and M. 1. Kocurko. I990b. Zoogeographic implications of Holocene mam­ Jones, 1. K., Jr., R. W. Manning, C. Jones, and R. R. Hol­ mal remains from ancient beaver ponds in Okla­ lander. 1988. Mammals of the northern Texas homa and New Mexico. Southwestern Natural­ Panhandle. Occasional Papers, The Museum, ist,35:105-110. Texas Tech University, 126:1-54. Glass, B. P., and A. F. Halloran. 1961. The small mammals Holliday, V. T. 1989. Middle Holocene drought on the of the Wichita Mountains Wildlife Refuge, Okla­ southern High Plains. Quaternary Research, homa. Journal ofMammalogy, 42:234-239. 31:74-82. Jones, J. K., Jr., R. W. Manning, C. Jones, and R. R. Hol­ lander. 1988. Mammals of the northern Texas Panhandle. Occasional Papers, Museum, Texas Tech University, 126:1-54. STANGL ET AL,-PRAnuE VOLE (MICROTUS OCHROGASTER) 11

Krebs, C. J. 1978. A review of the Chitty Hypothesis of Smith, K.. S. 1992a. The prairie vole, Microtus ochrogaster, population regulation. Canadian Journal of Zo­ in Caddo County, Oklahoma. Texas Journal of ology, 56:2463-2480. Science, 44:116-117. Lidicker, W. Z., Jr., and R. S. Ostfield. 1991. Extra-large Smith, K.. S. 1992b.AHolocene mammalian fauna from Box body size in California voles: causes and fitness Elder Creek, Caddo County, Oklahoma. Proceed­ consequences. Oikos, 61:108-121. ings of Oklahoma AcademyofScience, 72:39-44. Manning, R. W., and J. K.. Jones, Jr. 1988. A specimen of Stalling, D. T. 1990. Microtus ochrogaster. Mammalian the prairie vole, Microtus ochrogaster, from the Species, 355:1-9. northern Texas panhandle. Texas Journal of Sci­ Stangl, F. B., Jr., W. W. Dalquest, and R. J. Baker. 1992. ence, 40:463-464. Mammals of southwestern Oklahoma. Occa­ Manning, R. W., and C. Jones. 1998. Annotated checklist sional Papers, Museum, Texas Tech University, of Recent land mammals ofTexas, 1998. Occa­ 151:1-47. sional Papers, Museum, Texas Tech University, Stangl, F. B., Jr., J. R. Goetze, and C. B. Carr. 1993. Value 182:1-20. of the interorbital breadth in the discrimination Reed, K. M., and J. R. Choate. 1988. Noteworthy south­ of some problematic species ofPeromyscus and western records of the prairie vole. Southwest­ Reithrodontomys. Texas Journal of Science, ern Naturalist, 33:495-496. 45:186-187.

Addresses ofauthors:

FREDERICK B. STANGL, JR. KIMBERLY D. SPRADLING

Department ofBiology Department ofBiological Sciences Midwestern State University University ofNorth Texas Wichita Falls, Texas 76308 Denton, Texas 76203 email: [email protected] Email: kdS0052@untedu

JIM R. GoETZE

Science Department Laredo Community College Laredo, Texas 78040 email: [email protected] PUBLICATIONS OF THE MUSEUM OF TEXAS TECH UNIVERSITY

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 num­ bers of the Occasional Papers directly from the Museum ofTexas Tech University.

ISSN 0149-175X

Museum ofTexas Tech University, Lubbock, TX 79409-3191