Ecological Distribution of Small Mammals in the De Long Mountains of Northwestern Alaska

Transcription

1 ARCTIC VOL. 37. NO 2 (JUNE 1984) P Ecological Distribution of Small Mammals in th D Long Mountains of Northwstrn Alaska RICHARD J. DOUGLASS ABSTRACT. Th cological distribution of small mammals (< 200 g) was studid in th foothill tundra of th D Long Mountains in northwstrn Alaska. Thr spcis of shrws and fiv spcis of microtin rodnts wr trappd on 15 liv-trapping grids during 1978 and Emphasis was placd on th thr most abundant microtin spcis (Clrhrionomys rurilus, Microtus oconomus and M. grgalis). During lat summr up to six spcis of small mammals wr capturd pr habitat typ which rangd from wt madow through msic shrubland to dry ridgs. Following snowmlt most habitats containd only a singl spcis and som containd non. Only four habitat typs wr continuously occupid by small mammals during both summrs. Spcis divrsity was variabl among habitats. Most spcis of small mammals wr capturd on ight or fwr of th 15 trapping grids. Only Clrhrionomys rurilus was capturd on all grids. Th numbr of habitats occupid by Clrhrionomys rurilus incrasd from 4 to 14 as population dnsitis incrasd. Th numbr of habitats occupid by th othr spcis smd to b indpndnt of population dnsity. Avrag population dnsitis for th microtins wr low (< 15.ha- I) and, for ach spcis, varid according to habitat typ. Only Clrhrionomys rurilus populations dmonstratd markd intra-annual fluctuations (3.ha- I to 37,ha- I). Ky words: rodnts, habitat, Alaska, populations, Clrhrionomys ncrilus, Microtus oconomus, Microrus grgulis RESUME. La distribution tcologiqu d ptits mammifrs (200 g) fut Btudi dans la toundra ds contraforts ds montagns D Lung uans l nordoust d I Alaska. Trois spbcs d musaraigns t cinq sp&cs d rongurs microtins furnt capturs dans ds grills A pikgs a rcipint n 1978 t L tud porta surtout sur ls trois spkcs praominants d microtin6s (Clthrionomys rurilus, Microtus oconomus t M. grgalis). Vrs la fin d I Ctt, jusqu a 6 sptcs d ptits mammifrs furnt capturs par habitat, c drnir variant ntr ds prs font ds nigs, la plupart ds habitats n contnaint qu un sul spbcs t crtains n n contnaint aucun. Sulmnt quatr gnrs d habitats taint continullmnt occup& au cours ds dux 6th. La divrsit ds spbcs variait parmi ls habitats. La plupart ds spbcs d ptits mammifrs furnt captur% sur 8 ou moins ds 15 grills d pris au pikg. Sul l Clrhrionomys rurilus fut captur dans touts ls grills. L nombr d habitats occuph par l Clthrionomys rurilus augmnta d 4 a 14 a msur qu ls dnsits d population augmntaint. L nombr d habitats occup6s par ls autrs spkcs smblait &r inddpndant d la dnsit d population. La dnsit6 moynn d population pour ls microtin6s Ctait pu ClvC (15.ha- I) t variait pour chaqu spkc slon l gnr d habitat. Suls ls populations d Clrhrionomys rurilus dtmontraint ds fluctuations importants au cows d un m6m ann (d 3.ha-1 a 37.ha-I). Mots lks: rongurs, habitat, Alaska, populations, Clrhrionomys rurilus, Microtus oconomus, M. grgulis Traduit pour l journal par Mauric Guibord. INTRODUCTION Substantial information xists concrning population dynamics and cology of arctic rodnts (Batzli, 1975; Batzli r al., 1980). Most data, howvr, ar from th arctic coastal plain (Thompson, 1955; Pitlka, 1973; Martll and Parson, 1978; Martll and Fullr, 1979). Littl is known about th cology and population dynamics of rodnts in foothill or mountainous arctic tundra. B and Hall (1956) and Pruitt (1966) prsntd data on th cology of rodnts in foothill tundra but placd mphasis on rodnts of th coastal plain. In this papr I discuss cological distribution and short-trm (two summrs) population dynamics of rodnts living in foothill tundra in northwstrn Alaska. Bcaus of hilly and mountainous topography, th vgtation of th foothill tundra is mor patchy than coastal tundra. This patchinss provids an opportunity to xamin th ffcts of habitat on th structur of rodnt communitis and th rspons of populations of individual spcis to diffrnt habitat typs, Bsids dscribing cological distribution, I assssd diffrncs in svral aspcts (.g. divrsity, biomass, nich ovrlap) of small mammal communitis among habitats. I also xamind th spcific hypothsis that populations of microtin rodnts fluctuat in ssntially th sam mannr in all habitats in which thy occur (Krbs and Myrs, 1974). METHODS Th study was conductd in th D Long Mountains 160 km north of Kotzbu, Alaska. Th D Long Mountains (68 07 N, W) ar locatd approximatly 58 km inland from th Chukchi Sa and ris to btwn 610 m and 1190 m abov man sa lvl (asl). Most paks, howvr, ar lss than 900 m asl. Most lowlands ar dominatd by cottongrass (Eriophorurn sp.) and othr sdgs (Carx spp.), dwarf birchs (Btula spp.), and willows (Mix spp.). Tall stands of riparian willows occur along most drainags. Th highr lvations in th wstrn D Long Mountains ar dominatd by Dryas, lichns, dwarf willows, and bar rock. Trapping habitats wr slctd subjctivly on th basis of dominant plant spcis and microtopography, and rprsntd most of th major plant communitis in th study ara. On liv-trapping grid was placd in ach habitat typ. I documntd th vgtation by stimating th prcnt covr of mosss, lichns, grasss, sdgs, shrubs, and total shrubs (ignoring ovrlap of spcis) on 15 trapping grids (Daubnmir, 1959). Nin quadrats wr sampld on ach column of ach grid (a total of 36 on ach of 14 grids, and 72 quadrats on on grid). Tabl 1 dscribs som of th physical and vgtativ charactristics of th mammal sampling locations. Fourtn grids, ach consisting of 100 stations of on trap ach, wr trappd during 1978and Grids wr four Montana Collg of Minral Scinc and Tchnology, Butt, Montana 59701, U.S.A

2 SMALL MAMMAL DISTRIBUTION, NW ALASKA I49 TABLE 1. Habitat charactristics of 15 liv-trapping grids in th D Long Mountains, Alaska n -c( Elvation (m) dpth Man to prmafrost 49 (cm) maximum Man hight of vg. (cm) X Covrag Bar Littr Lichn Moss sdg I 5 7 Grass Hrb 5 I All shrubs Btula nana I salix sp Ldum gronlandicum Shprdia canadnsis Emptrum Vaccinium uliginosum V Andromda polifolia columns wid by 25 rows long and ncompassd aras of 0.76 ha. During 1979, I addd a grid containing 200 stations in ight columns and 25 rows (1.76 ha) in an upland shrub habitat. Columns on all grids wr 15 m apart and rows wr 7 m apart. Grids wr sampld for thr conscutiv days ach during Jun, July, and August 1978 and Shrman liv-traps (23 x 8 x 9 cm) wr covrd with plywood, baitd with rolld oats and panut buttr, and chckd onc daily. Animals wr individually numbrd by toclipping and rlasd at th point of captur. Th numbr, location of captur, sx, brding condition, and wight wr rcordd for ach animal ach tim it was capturd. Population dnsitis ar xprssd by th numbr of individuals capturd pr trapping grid and th numbr of individuals pr hctar (numbr capturd/ara of, grid). No attmpt was mad to stimat population dnsity by capturrcaptur tchniqus bcaus of problms of nonrandom sampling (Krbs, 1977). Biomass dnsitis ar rprsntd by th sum of wights of individuals dividd by th ara occupid by ach grid. Spcis divrsity was masurd by th Shannon-Wavr formula (Pilou, 1966); nich bradth was calculatd according to Lvins (1%8); spcis association was calculatd according to Dic (1945) (for comparisons with Pruitt's (1966) data); and nich ovrlap was calculatd using Pianka's (1973) formula. Spcis Distribution RESULTS AND DISCUSSION Numbrs of small mammals trappd in various habitat typs rangd from thr (bar ridg) to 96 (riparian grass-willow) (Tabl 2). Two of th thr tussock grids in thdlong Mountains had rlativly high numbrs of rodnts but th third had vry fw animals. A shrub-covrd location had rlativly high numbrs of rodnts and th mat cushion typ had low numbrs of rodnts. In contrast, Pruitt (1966), working on th coastal plain nar th D Long Mountains, found that tussock habitats containd th largst numbr of small mammals whras madows and mat cushion aras containd intrmdiat numbrs, and shrublands containd th smallst numbr. Th numbr of spcis in th various habitat typs rangd from on (bar ridg) to six (riparian grass-willow and birchtussock) with most habitats supporting four to fiv spcis of small mammals. Th divrsity indics wr lowst (0) on th bar ridg and highst (1.54) in th riparian grass-willow. Pruitt (1966) found tussock aras to hav th highst divrsity and shrublands th lowst. In th D Long Mountains, tussock aras had lowr small mammal divrsitis than shrublands. Th avrag monthly biomass was dirctly rlatdtoth numbr of individuals, with highst (340.3 g-ha") occurring in th birch flat and th lowst (11.3 g-ha") occurring on th

3 I50 R.J. DOUGLASS TABLE 2. Numbr of small mammals liv-trappd on 15 grids in th D Long Mountains, Alaska, during 1978 and 1979' Priodically Floodd Northrn rd-backd vols Tundra vols I I 69 IO 4.23 mmings Brown I I lmmings Collard I s Singing I Maskd shrws Arctic shrws (Sorx arcticus) 0 0 I 0 0 I 7 I Dusky shrws I Numbr of spcis indx Divrsity Avrag monthly biomass (g) Avrag chang in biomass (g) Jun-August Chang in biomass (g) August 1978-Jun NA Monthsthatanimals JJA JJA 1978 JJA NA JJA JJA -JA JJA JJA J-A -JA -JA -JA -JA -JA 1979 prsnt wr JJA JJA -JA JJA --- JJA JJA -JA J-- -JA JJA JJA -JA --A -JA 1 'Divrsity = E Pi log Pi;nichbradth = I/EPi* n 'Grid largr than othr grids. Numbrs = x th actual numbr capturd. bar ridg habitat. Th two-yar avrag for all aras was g.ha" which was lowr than th thr-yar ( ) avrag in th Ogotoruk drainag wst of th D Long Mountains (51 = gsha-l, rang g.ha"; Pruitt, 1966). Six habitats containd small mammals during all six sampling priods, and all containd animals during at last two of thsix priods. Of thos without continuous occupation by small mammals, all but on wr' vacant during Jun in on or both yars. This may indicat that fw, if any, animals wr abl to surviv wintr or spring flooding in thos habitats. Tundra vols (Microtus oconomus) apparntly survivd th wintr in th sdg tussock habitat and wr prsnt in Jun 1979 but disappard by August aftr a priod of intns rain. That habitat was occupid in August by northrn rd-backd vols (Clthrionomys rutilus). Th numbr of habitats occupid by a spcis was not ncssarily dirctly rlatd to its abundanc. Som of th last abundant spcis (.g. maskd shrw [Sorx cinrus]) wr capturd in mor habitats than mor abundant spcis (.g. tundra vol). Northrn rd-backd vols wr ubiquitous, and maskd shrws and tundra vols occurrd in most habitats. Although all spcis wr prsnt in svral habitats, non wr prsnt in qual dnsitis in all habitats. Som spcis (.g. dusky shrw [Sorx vagrans]) occurrd in high dnsitis in only on or two habitats with fw individuals occurring in othr habitats. Th singing vol (Microtus grgalis), which occurrd in th sam numbr of habitats as th dusky shrw, was prsnt in mor uniform numbrs among various habitats than th dusky shrw. Th nich bradth indx accounts for both th numbr of habitats occupid and th vnnss of occupation among habitats. In th D Long Mountains, northrn rd-backd vols had th broadst nich (Tabl 2). Pruitt (1966) also found northrn rd-backd vols to hav broad habitat tolranc whil tundra vols and brown lmmings had rstrictd habitat tolrancs. Batzli and Jung (1980) did not find northrn rd-backd vols north of th Brooks Rang, but found tundra vols in fwr habitats than brown or collard lmmings. Howvr, th discrpancy in nich bradths of lmmings and tundra vols btwn Batzli and Jung's (1980) study ara and th D Long Mountains cannot b fully assssd with th limitd data (on captur ach of brown and collard lmmings) from th D Long Mountains. Whitny (1976) compard tundra vols and northrn rd-backd vols in th boral forst and also found that tundra vols occupid a mor rstrictd nich than northrn rd-backd vols. Tmporal Rlationships Populations of microtin rodnts wr variabl in th D Long Mountains; most populations rmaindblow1o.ha-l whras som rachd dnsitis ovr 30.ha-' (Fig. 1). Th numbrs and fluctuations wr within th rangs rcordd for ths spcis in othr studis. Most population data for northrn rd-backd vols show that thy undrgo a rgular sasonal population fluctuation, starting with low numbrs in Jun and raching high dnsitis in August or Sptmbr (Fullr, 1969; Whitny, 1976; Martll and Fullr, 1979). Dnsitis in various studis rangdfrom

4 SMALL MAMMAL DISTRIBUTION, NW ALASKA 151 MINIMUM NUMBER PER HECTARE FIG. I Population dnsitis of rodnts on 15 liv-trappinggridsduring1978 and 1979 inth D Long Mountains, Alaska. Lgnd: I. Birch ridg 6. Crk bottom I 1 I. Madow 2. Willow bar 7. Sdg-willow madow 12. Madow shrub 3. flat Birch 8. Tussock Tussock 2 4. Tussock 1 9. Sdg-willow ic scar 14. Riparian-willow 5. Bar ridg 10. Crk bottom Shrubland Jun populations of 2.ha-I to 32.ha- (Fullr, 1969) to fall populations of 5ha-1 to 62-ha- (Whitny, 1976). In most aras whr this spcis has bn studid, it also undrgos priodic fluctuations (Fullr, 1969); howvr, in som aras northrn rd-backd vol populations hav bn found to b rlativly stabl intrannually (Whitny, 1976; Martll and Fullr, 1979). In th D Long Mountains northrn rd-backd vols undrgo th typical sasonal fluctuation but it is not yt possibl to dtrmin th natur of long-trm fluctuations. Populations of tundra vols in th D Long Mountains wr rlativly low, as low as thos dscribd in th first thr yars of Batzli and Jung s (1980) study nar Mad Rivr. Batzli and Jung (1980) found that tundra vol population dnsitis nar Mad Rivr rmaind rlativly low (2-6.ha-I) ovr a thryar priod but incrasd th fourth yar to 35.ha-. Whitny (1976) also found tundra vol populations to b quit variabl. In Whitny s study ara nar Fairbanks, tundra vols rachd population dnsitis of ha- during th scond yar of study, rapidly dclind th sam yar, and wr no longr trappabl by th third yar. During two summrs, populations of brown lmmings in th D Long Mountains wr vry low. In othr aras populations of this spcis fluctuat gratly or rmain fairly stabl. Brown lmming populations nar Barrow charactristically hav thr- to six-yar cycls with lows of approximatly 0.02.ha-I and highs of up to 255.ha-I (Thompson, 1955; Pitlka, 1973; Batzli t al., 1980). Howvr, nar Mad Rivr, about 100 km south of Barrow, brown lmming populations rmaind rlativly low (2-6.ha-I) and stabl ovr a four-yar priod (Batzli and Jung, 1980). Dnsitis of singing volsrmaind low in th D Long Mountains but wr slightly highr during 1979 than during Pruitt (1966) found singing vols to b th scond most common rodnt occurring in th rgion wst of th D Long Mountains. Thy wr found in only on habitat (snow-bd community). Numbrs of singing vols capturd in qual numbrs of traps dcrasd from 5 1 animals in 1959 to on in This suggsts that singing vols can undrgo larg fluctuations in numbrs. In summary, populations of thr microtin rodnts in th D Long Mountains during 1978 and 1979 occurrd in dnsitis similar to thos found in othr subarctic and arctic rgions of North Amrica. Northrn rd-backd vols undrwnt sasonal incrass which ar apparntly typical for this spcis. Tundra vols and singing vols rmaind low throughout both summrs. According to various studis, thos spcis may or may not undrgo cycls in population dnsitis. Dnsitis of brown lmmings and collard lmmings rmaind xtrmly low through both summrs. 14- I Il- : 10-, 9- + a 8- z u- I -A A Tundra vol Singing vol Northrn rd-backd vol I I I I I I I I Numbr of animals capturd on 14 livtrapping grids FIG.?. Numbr of habitats occupid by rodnts in rlation to population dnsity in th D Long Mountains, Alaska. Sasonal Changs In Habitat Associations Som ffcts of population incrass on habitat associations ar dmonstratd in Figurs 1 and 2. As northrn rd-backd vol populations incrasd ach summr, thy occupid mor habitats (r = 0.948, d.f. = 4, 0.01 < PcO.05). Howvr, within th small rangs of population dnsitis of tundra and singing vols, tundra vols (r = , d.f. = 4, P>0.50) and singing vols (r = 0.725, d.f. = 4, O.O5<P<O. 10) dmonstratd no consistnt rlationship btwn population dnsity and th numbr of habitats occupid. Populations of rodnts in th D Long Mountains vary both spatially and tmporally (Fig. I). I assumd that habitats occupid in Jun wr thos whr animals survivd th wintr and wr, thrfor, most important to th spcis. Rodnts dmonstratd distinct sgrgations in Jun with ovrlap in habitat occupation occurring only in 1978 in th riparian grasswillow. As th summr progrssd and northrn rd-backd vol populations incrasd, habitat sgrgation bcam lss distinct. By August, all habitats wr occupid by ithr northrn rd-backd vols or by variously mixd populations of northrn rd-backd vols, tundra vols, and singing vols.

5 R.J. DOUGLASS d rlativly low, most of th ovrlap in occupation of habitats was th rsult of northrn rd-backd vols invading nw habitats as thir numbrs incrasd. Contrary to my rsults, Krbs and Wingat (1976) found that nich bradth did not incras as northrn rd-backd vol populations incrasd. Littl isknownabout invasions ofnw habitats by small microtin rodnts. KrbsandMyrs (1974), in a rviw of microtinpopulationdynamics,dscribd a gnralpopulationsyndrom in whichmigrationoccurrd during thincras phas of population cycls. Th fat of migrating animals was not known and th influnc of habitat on migrating animals was not discussd. Howvr, populations of northrn rd-backd vols in th D Long Mountains sm to occupy additional marginal habitats as thir populations incras, in a mannr similar to muskrats (Ondurru zibrhicus) (Errington, 1967)and arctic ground squirrls (Sprtnophiius purryii) (Carl, I97 1 ). TABLE 3. Comparison of ovrwintring habitats of thr spcis of rodnts in th D Long Mountains, Alaska, Low High F Valu Rlativ Habitat Valus Prmafrost (cm) 1.54 RV,~ - Tv Rv, Sv Maximum hight of vg. (cm) Rv, Sv w, " Littr (% covr) _. Tv Rv, Sv - Rv, Shrubs > 50 cm 9.66 Rv, Sv Tv Rv, Lichn "- Rv,Rv, Tv Sv Moss I.06 N.S. Sdg Sv "_ Tv Rv, RJ, Grass Sv Rv, Rv, Tv Hrbs Sv Rv, Rv, Tv Shrubs <OS m I.90 N.S. Birch (Brula nana) Rv, "- Tv Rv, Sv Willow (Salk sp.) Rv, Tv Rv, Sv Labrador ta (Ldum gronlandicum) Rv, Rv, Tv Sv Crowbrry (Emprrum nigrum) 0.02 N.S. Linonbrrv - - (Vaccinium. viris-idaa) 4.89 Rv, -~ Tv Sv Rv, Blubrry (V. uliginosum) Rv, Rv, Tv Sv 'Th Nwman-Kuls tst was usd to compard among variabls. A significanc lvl of P<0.05 was usd. *Habitats: Rv, = Northrnrd-backd vol, birchridg Rv, = Northrnrd-backd vol, birchflat Tv = Tundra vol, madow (shrub) Sv = Singing vol, riparian grass-willow Eightn habitat variabls (Tabls I and 3) wr stimatd and compard (Nwman-Kuls Tst [Zar, 19741) to dtrmin if habitats continuously or frquntly occupid by ach of th mostcommonlycapturdrodnt spcis wr distinct from thos occupid by th othr spcis. Northrn rd-backd vols continuously occupid two habitats (birch ridg and birch flat) and tundra vols continuously occupid on habitat (madow shrub). Singing vols did not continuously occupy anyof th habitats, butwrtrappd in th riparian grasswillow during all priods xcpt July Th two habitats occupid by northrn rd-backd vols wr dissimilar (Tabl 3). Thr wr, howvr, consistnt diffrncs btwn ach of th two northrn rd-backd vol habitats and thos continuously occupid by tundra vols or singing vols. Habitats continuously occupid by northrn rdbackd vols consistntly diffrd from thos of tundra vols in having a lowr prcnt covr of littr and willows and a highr prcnt covr of lichns, birch, and Labrador ta. Tundra vol habitat diffrd from singing vol habitat in having shortr maximum vgtation hight, lowr prcnt covr of grass, hrbs, and crowbrry, and highr prcnt covr of littr, lichn, birch, sdg, and low willow. Consistnt and significant diffrncs among habitats occupid by ach of th thr spcis occurrd only in th prcnt covr of littr, lichn, and birch. Both of th northrn rdbackd vol habitats also containd significantly lss sdg covr than habitats occupid by tundra and singing vols. Tundra vols continuously occupid habitats with th highst prcnt covr of littr and sdgs, as suggstd by Gtz (1961) and Grant (1971) for othr Microtus spp., and singing vols occupid habitats with an intrmdiat but substantial accumulation of littr. Northrn rd-backd vols occupid habitats with spars littr covr, and aras with high covr of birchs and lichns. This is consistnt with th occupation by northrn rd-backd vols of aras with high covr of birch and lichns in th Northwst Trritoris (Douglass, 1977) and by othr Clthrionomys spp. lswhr (Pollard and Rlton, 1973). Prcnt covr of birch was intrmdiat in th tundra vol habitat and absnt from th singing vol habitat. Habitats continuously or most frquntly occupid by northrn rd-backd vols, tundra vols, and singing vols apparntly wr sgrgatd along a moistur gradint, with northrn rd-backd vols occurring in th drist habitats and tundra vols occurring in th wttst habitats. Littr, sdgs, and willows tnd to hav highr covr valus on poorly draind sits whras lichns and birch gnrally hav highr covr valus on wll-draind sits. Additional rsarch, including an assssmnt of th rol of rhizomatous vs. nonrhizomatous monocots, would b usful in clarifying habitat sgrgation among ths rodnts. Th prsnc of on spcis of small mammal may dtrminhowothr spcis us a givnhabitat (Morris, 1969; Grant, 1972; Douglass, 1976), may affct population dynamics of othr spcis (Rdfild t ai., 1977), or may hav no ffct at all (Grant, 1972; Krbs, 1977). To obtain som insight into intractions among spcis of rodnts in th D Long Mountains, I calculatd (basd on data from all grids) cofficints of association (CA) and nich ovrlap (03) for th most abundant spcis (Tabl 4). A CA of lss than 1.0 indicats that th two spcis occurrd togthr lss oftn thanwouldbxpctd by chanc. A CAvalu gratr than 1.0 indicats that th two spcis occurrd togthr mor frquntly than xpctd. In th D Long Mountains during 1978 and 1979, northrn rd-backd vols occurrd with tundra vols at th bginning of th summr lss frquntly than xpctd by chanc, but not at th nd of th summr (Tabl 4). This could indicat that

6 ~~~~ SMALL 153 TABLE 4. Cofficints of association (CA) and nich ovrlap (Qj) btwn pairs of spcis of vols in th D Long Mounains (1978 and 1979), and CA in th Ogotoruk rgion (1959-!%1)1 rd-backd Northrn vol Singing vol CA Month Yar Oil CA Oii Month Yar ~ Tundra vol 1978 Jun Jun July July 0.70 August August Jun OOO 1979 Jun 0.00 o.oo0 July July August August Man Ogotoruk Rgion I Northrn vol 1978 Jun July August Jun 0.00 o.oo0 July August Man Ogotoruk Rgion 'Coflcinr of association = CA = hn/ab, whr h = numbr of sampls in which both spcis a and b occur, n = numbr ofsampls, a = numbr of sampls in which spcis u occurs, and b = numbr of sampls in which spcis b occurs. Data for ar from hitt (1966). Nich OVrhP = Oij = E(xik Xij)/dEX'ik E*X,k. whr Xik = proportion of spcis i in habitat k, and X,k = proportion of j in habitat k. som mild comptitiv xclusion was acting to sparat th spcis in arly summr, or thatthyhad diffrnt habitat prfrncs. Singing vols did not show a consistnt pattrn of association with ithr of th othr vols. Pruitt (1966). howvr, found that northrn rd-backd vols and tundra vols occurrd togthr wst of th D Long Mountains mor frquntly than xpctd, and singing vols wr not capturd with th othr two spcis. Existing data ar inadquat to xplain th discrpancy in rodnt spcis associations btwn th D Long Mountains and th coastal plain to th wst. Th nich ovrlap calculations suggstd thath thr spcis of rodnts occupid discrt nichs (Tabl 4). In this study th nich ovrlap indx masurs th similaritis of hhbitat nichs; 0.0 indicats no similarity and 1.0 indicats similarity. Tundra vols and singing vols had th most similar nichs (Oij = 0.923) but only during July Evn though northrn rd-backd vols occupid mor habitats as thir numbrs incrasd, thir habitat nich rmaind distinct from th othr spcis (Oijc0.5). This is consistnt with th analysis of spcific habitat variabls discussd abov. CONCLUSIONS In this study I was intrstd in dtrmining whthr small mammal communitis rspondd to habitat patchinss rlatd to topography in foothill tundra. Svral associations wr vidnt. All habitats containd small mammals including rodnts and insctivors, but most habitats wr dominatd by on to thr spcis of microtin rodnts. Th most numrous spcis dpndd on typ of habitat and th month in which th habitat was trappd. In arly summr just aftr snowmlt, many habitats wr not occupid and othrs wr occupid only by singl spcis. During arly summr northrn rd-backd vols, tundra vols, and singing vols occurrd in sparat habitats, apparntly along a moistur gradint. As th summr progrssd, sgrgation was not as obvious bcaus northrn rd-backd vols vntually occupid all habitats. 1 conclud that patchinss in foothill tundra was usdby smallmammals for habitat sgrgation andthis sgrgation was most pronouncd during th critical spring priod (Fullr, 1969) whn many habitat typs wr inundatd by snowmlt or runoff. During this priod th topography of th foothill tundra not only providd aras with diffrnt vgtation but also providd a varity of aras whr rodnts could scap th ffcts of mlting snow. 1 also was intrstd in dtrmining if populations of microtin rodnts fluctuatd similarly in all habitats as suggstd by Krbs and Myrs (1974). At th dnsitis that occurrd in th D Long Mountains during 1978 and 1979, microtin rodnt populations appard not to fluctuat similarly among all habitats in which thy occurrd. This was particularly vidnt in th spring xtinctions in many habitats dscribd abov. Most of th microtin rodnts discussd by Krbs and Myrs (1974) rachd vry high dnsitis (>50.ha"). If rodnt populations in th D Long Mountains incras to such dnsitis in th futur, thy may do so similarly in all habitats. If futur rsarch is prformd on cological distribution of small mammals in foothill tundra, I would rcommnd invstigating thr aspcts: I. What spcifically happns whn northrn rd-backd vol populations incras and nich bradth broadns? This could b important to gnral populationthory by providing insight into th fat of migrating animals. 2. Do rodnts rspond dirctly to spcific habitat variabls as

7 I54 R.J. DOUGLASS suggstd abov or ar ths variabls and th rodnts rsponding to somthing ls? This could b xamind through mor sophisticatd dscriptiv rsarch or by using an xprimntal approach. 3. What ar th long-trm rlationships btwn populations of rodnts in foothill tundra vs. thos prviously dscribd in coastal tundra? This could b discovrd only with longtrm (- IO-yar) population monitoring. ACKNOWLEDGEMENTS Data for this papr wr collctd whil fild studis wr bing conductd for GCO Minrals Co. I would lik to xprss my apprciation to D. Knndy, J. Britton, and J. Frntt for making this study possibl. 1 also would lik to thank D. Hollingsdal for assistanc with drafting and D. Yokl for assistanc in th fild. Finally, I xtnd my apprciation to G.O. Batzli and A.M. Martll for thir hlpful suggstions on an arlir draft. REFERENCES BATZLI, G.O Th rol of small mammals in arctic cosystms. In: Golly, F.B., Ptruswicz, K. and Ryzkowski, L. (ds.). Small Mammals: Thir Productivity and Population Dynamics. Cambridg, England. Cambridg Univrsity Prss and JUNG, H.J.G Nutritional cology of microtin rodnts: rsourc utilization nar Atkasook, Alaska. Arctic and Alpin Rsarch 12: BATZLI, G.O., WHITE, R.G., MacLEAN, S.F., Jr., PITELKA, F.A. and COLLIER, B.C Th hrbivor-basd trophic systm. In: Brown, J., Millr, P.C., Tiszn, L. and Bunnll, F. (ds.). An Arctic Ecosystm: Th Coastal Tundra at Barrow, Alaska. Intrnational Biological Programm 12. Cambridg: Cambridg Univrsity Prss. 571 p. BEE, J.W. and HALL, E.R Mammals of Northrn Alaska. Univrsity of Kansas Musum of Natural History, Miscllanous Publication No. 8. Lawrnc, Kansas: Alln Prss. 309 p. CARL, E.A Population control in arctic ground squirrls. Ecology 52: DAUBENMIRE, R.F A canopy covrag mthod of vgtation analysis. Northwst Scinc 33: DICE, L.R Masurs of th amount of cological associations btwn spcis. Ecology 26: DOUGLASS, R.J Spatial intractions and micro-habitat slction of two sympatric vols. Microrus monfanus and M. pnnsylvanicus. Ecology Effcts of wintr roads on small mammals. Journal of Applid Ecology ERRINGTON, P.L Of Prdation and Lif. Ams, lowa: lowa Stat Univrsity Prss. 277 p. FULLER, W.A Changs in numbrs of thr spcis of small rodnts nar Grat Slav Lak, N.W.T., Canada, , and thir significanc for gnral population thory. Annalr Zoologica Fnnica 6: GETZ. L.L Factors influncing th local distribution of Microrus and Synaptomys in southrn Michigan. Ecology 42: GRANT, P.R Th habitat prfrnc of Microrus pnnsylvanicus and its rlvanc to th distribution of this spcis on an island. Journal of Mammalogy 52: I Intrspcific comptition among rodnts. AnnualRviw of Ecology and Systmatics 3: KREBS, C.J Comptition btwn Microrus pnnsylvanicus and Microtus ochrogastr. Amrican Midland Naturalist 97:4249. and MYERS, J.H Population cycls in small mammals. Advancs in Ecological Rsarch 8: KREBS, C.J. and WINGATE, I Small mammal communitis of th Kluan rgion, Yukon Trritory. Canadian Fild-Naturalist 90: LEVINS, R Evolution in changing nvironmnts: Som thortical xplorations. Princton, NJ: Princton Univrsity Prss. 120 p. MARTELL, A.M. and FULLER, W.A Comparativ dmography of Cithrionomys rutilus in taiga and tundra in th low Arctic. Canadian Journal of Zoology 57: MARTELL, A.M. and PEARSON, A.M Th small mammals of th Macknzi Dlta Rgion, Northwst Trritoris, Canada. Arctic 31: MORRIS, R.D Comptitiv xclusion btwn Microtus and Clrhrionomys in th aspn parkland of Saskatchwan. Journal of Mammalogy 50: PIANKA, E.R Th structur of lizard communitis. Annual Rviw of Ecology and Systmatics 4: PIELOU, E.E Th masurmnt of divrsity in diffrnt typs of biological collctions. Journal of Thortical Biology PITELKA, F.A Cyclic pattrn in lmming populations nar Barrow, Alaska. In: Gunn, W.W.H. (4.). NARL Twnty-fifth Annivrsary Symposium. Washington, D.C.: Th Arctic Institut of North Amrica POLLARD, E. and RELTON, J.H A study ofsmall mammals in hdgs and cultivatd filds. Journal of Applid Ecology PRUITT, W.O., Jr Ecology of trrstrial mammals. In: Wilimovsky, J.H. and Wolf, J.N. (ds.). Environmnt of th Cap Thompson Rgion, Alaska. Washington, D.C.: U.S. Atomic Enrgy Commission REDFIELD, J.A., KREBS, C.J. and TAIT, M.J Comptition btwn Promyscus manicuiatus and Microrus rownsndii in grasslands of coastal British Columbia. Journal of Animal Ecology 46: THOMPSON, D.Q Th rol of food and covr in population fluctuations of th brown lmming at Point Barrow, Alaska. Transactions of th North Amrican Wildlif Confrnc 20: WHITNEY, P Population cology of two sympatric spcis of subarctic microtin rodnts. Ecological Monographs 46: ZAR, J.H Biostatistical Analysis. Englwood Cliffs. Nw Jrsy: Prntic-Hall. 620 p.