Northern populations of red squirrels (Tamiasciurus

Transcription

1 Territrial bequeathal by red squirrel mthers aren Price Behaviural Eclgy Research Grup, Department f Bilgical Sciences, Simn raser University, Burnaby, British Clumbia V5A 1S6, Canada Stan Butin Department f Zlgy, University f Alberta, Edmntn, Alberta T6G 2E1, Canada Received 6 January 1992 Revised May 1992 Accepted June /9/ Internatinal Sciety fr Behaviral Eclgy Red squirrels (Tamiasciurus hudsnicus) defend individual, fd-based territries year rund. These territries are crucial fr verwinter survival, yet sme female red squirrels in nrthern ppulatins bequeath their territries t ffspring and search fr a new ne. Bequeathal invlves active, strategic dispersal by breeding females. Our study dcuments this unusual behavir and investigates its crrelates in tw red squirrel ppulatins in nrthern Canada. Thirty percent f breeding females dispersed. Bequeathal was related t breeding date, with late-breeding females mre ften dispersing, but was nt related detectably t female cnditin r territry quality. As an underlying trade-ff, early-brn juveniles mre likely acquired a territry independently, but early-breeding females lst mass and may have increased their risk f territry lss. We suggest that bequeathal has evlved as a cnsequence f cnditin dependence in breeding date cupled with limited resurces. ey wrds: territry, breeding dispersal, cnditin-dependence, Tamiasciurus hudsnicus, bequeathal. [Behav Ecl :1-150 (199)] Nrthern ppulatins f red squirrels (Tamiasciurus hudsnicus) defend territries yearrund (Smith, 1968). During winter, squirrels rely n cne caches (middens) n each territry fr fd individuals withut a territry at this time generally perish (emp and eith, 1970; Butin S, Alta U, unpublished data). A territry is thus a valuable cmmdity t a red squirrel. Yet, sme females share part f their territry with an ffspring (Butin and Schweiger, 1988), and thers bequeath their entire territry t an ffspring and search fr a new ne. Territry bequeathal is an unusual behavir. In sme species, juveniles may inherit part f their natal territry r may inherit the entire area after a parent's death (e.g., rdents: references in Andersn, 1989; three-ted slths: Mntgmery and Sunquist, 1978; pikas: Smith and Ivins, 198; scrub jays: Wlfenden and itzpatrick, 1978), and in sme nnterritrial species, adult females disperse regularly (e.g., cmmn vles: Byce and Byce, 1988; heath rats: Cckburnetal., 1981; Clumbian grund squirrels: Harris and Murie, 198; reedbucks: Hward, 1986; hairy-nsed wmbats: Jhnsn CN, Tasmania U, persnal cmmunicatin; spider mnkeys: Macarland Symingtn, 1987). Hwever, bequeathal f an entire territry t ffspring fllwed by dispersal in search f a new territry seems very rare (Andersn, 1989; Waser and Jnes, 1991). Investigatin f this seemingly extreme behavir may well shed light n ther cases f female dispersal. Several factrs may influence bequeathal. A female with lw future reprductive ptential shuld invest mre in her current litter (isher, 1958; Stearns, 1976; Williams, 1966a,b) and shuld bequeath mre readily. A female hlding a pr territry may be mre likely t disperse in search f a better ne. Alternatively, a female wning a territry larger than required fr survival might increase her verall reprductive utput by sharing the territry with ne r mre f her ffspring (Butin and Schweiger, 1988). inally, a female breeding late in the -6-week seasn might bequeath mre readily because late-brn juveniles may be at a disadvantage in the cmpetitin fr limited territries. Because lactatin is energetically expensive (Gittleman and Thmpsn, 1988) and because early-breeding females lactate befre summer fd appears, breeding date may depend n bdy cnditin. This paper describes a field study f territrial bequeathal in red squirrels. The study was prmpted by bservatins f vagrant breeding females (i.e., thse n lnger defending territries) cmpeting vigrusly fr vacant spts (Price et al., 1986). We describe bequeathal and investigate its crrelates in an unmanipulated ppulatin. We then examine the underlying trade-ffs f juvenile success and female cnditin and cnclude with a testable hypthesis abut when a female red squirrel shuld bequeath her territry. A related study (Butin et al., 199) investigates bequeathal fllwing experimentally created territrial vacancies. METHODS We cnducted this study n tw 16-ha sites ( and ) in the semi-arid, nrthern breal frest zne f the suthwest Yukn (61 N, 18 W) frm May thrugh August White spruce (Picea glauca) is the dminant tree, and the nly cnifer in the regin. Spruce distributin is clumped n bth sites, with patches f mature spruce interspersed with meadw areas f lw-lying herbs and shrubs (mainly willw, Salix spp.) and immature spruce. We chse sites fr their high squirrel density ( individuals/ha) and ease f visibility. Dwnladed frm at Penn State University (Patern Lib) n ebruary 19, Behaviral Eclgy Vl. N. 2

2 Mnitring f female squirrels We trapped red squirrels in cages that were baited with peanut butter and set n the grund at midden sites. Middens, cmprising cnes cached in a netwrk f tunnels and cne bract debris frm previus years, frm the fcal pint f a squirrel's activity. Target squirrels entered traps n their middens in 5-0 min. We marked all individuals with a numbered metal ear-tag hlding an attached, unique clr cmbinatin f plasticized wire r vinyl. The study sites are well established as part f a lng-term prject, and mst midden sites and squirrels were knwn previusly. Every 10 days ver the summer, we trapped all female squirrels, breeding and nnbreeding, t mnitr their cnditin. We measured mass, reprductive cnditin (small nipples, medium nipples, lactating) and fur cnditin (five-pint scale f fur patchiness and shine). Trappability was 100% while females held territries, but decreased during territrial disputes and after the settlement f juveniles n their mthers' territries. Territry measurements We mapped the territries f all females as well as the brders f adjacent males' territries in early May (befre juvenile emergence frm nests). We fllwed squirrels fr an hur r until they disappeared frm view, recrding squirrel lcatin (relative t a 15-m grid f stakes) and behavir (calling, feeding, sitting, interacting with anther squirrel). We entered lcatins n the territry map whenever the squirrel gave a territrial rattle call r aggressive screech call (Lair, 1990; Smith, 1978), when it fed, and when it sat still fr lnger than 2 min (Price et al., 1986). Squirrels habituated t us rapidly and generally did nt apprach, run away frm, r give alarm calls in respnse t an bserver after tw bservatin perids. We did nt use trapping t map territries because squirrels smetimes mve acrss several territries t reach a baited trap (Price, Butin S, persnal bservatin). We derived territry bundaries using minimum cnvex plygns. In this dense ppulatin, mst territries were cntiguus, and bundaries were usually bvius frm interactins with neighbrs. We assessed territry quality in tw ways: (1) by estimating the cne-prducing capability f the spruce trees n each territry, and (2) by measuring midden area and cntent. We tabulated spruce trees n each territry by 5-cm DBH (diameter at breast height) class and by a subjective measure f tree quality (a five-pint scale rating the crwn length and density f branches; Price, 1990). Spruce DBH is the best predictr f cne crp; crwn length and tree dminance class (related t branch density) are the best predictrs f frequency f cne prductin (Matthews, 196; Waldrn, 1965). We used the number f large (greater than 10-cm DBH), gd-quality (class 1 and 2) trees as a measure f the lng-term cne-prducing capability f a territry. The spruce cne crp failed cmpletely in the regin in We measured midden area (the area cvered with recent cne bracts and/r active tunnels) as an index f territry quality (Rusch and Reeder, 1978) and investigated six tunnel entrances in each midden fr the presence f unpened cnes in August. Selected tunnels were at least 1 m apart, ne in each sextant f the midden. We remved all material frm the penings and investigated cnes fr the presence f whle seed. After a cne failure, nly territries with cnes remaining frm previus years have any stred seed. Juvenile behavir We lcated nests (rugh, grass spheres built 2-15 m high in spruce trees) by bservatin and by raditelemetry f lactating females, and we weighed, eartagged, and sexed the juveniles. We visited nests nce nly unless juveniles were t small t be eartagged n the first visit (7 f 28). emales frequently mved the litter t a new nest fllwing disturbance, but never abandned the litter. We bserved females with litters due t emerge fr 1 h each day. We divided litters int early and late grups, dependent n prjected emergence date (calculated frm the mass f nestlings measured n given dates, see Price et al., 1990). Median emergence dates fr early and late litters were 26 May and 9 June, and the minimum time between litters assigned t the early versus the late grup was 5 days. After their emergence frm the nest, we trapped and weighed juveniles n an pprtunistic basis. Trappability f juveniles was lwer than fr adults, and sme individuals were never caught. We visited each family every 2 r days ver the summer. r the first - weeks after emergence, we bserved juveniles fr a half hur every secnd day. After this time, interactins decreased, and juveniles wandered in search f a territry, eventually attempting t settle. We lked fr each squirrel (mthers and juveniles) fr 15 min every 2nd r rd day and, abut every 10th day, walked transects t lcate wandering juveniles. We als recrded pprtunistic bservatins f juveniles. We recrded dates f emergence, f vagrancy, f first territrial calls, and f first territrial defense fr juveniles, and perids f vagrancy r nndefense fr females. We defined a juvenile as having left hme if it was seen utside its natal territry in tw bservatin perids and was nt seen at hme in the interim. We als nted the lcatin and behavir f any untagged juveniles seen n the study sites (vagrants frm utside the grids). emale behavir The behaviral ptins available t successful mthers (i.e., thse wh did nt either lse all their juveniles r lse their territry t an intruding adult) include keeping the territry, bequeathing the territry t ne r mre juveniles, r giving up part f the territry and keeping part (sharing). We cnsidered a female t have bequeathed nly if she was seen elsewhere subsequently r if her behavir befre disappearance was unusual. Sudden disappearance alne is cnsistent with predatin and hence is insufficient evidence f bequeathal. RESULTS Of an initial sample f 28 breeding females, 25 squirrels bred successfully (i.e., their juveniles Price and Butin Territrial bequeathal in red squirrels 15

3 igure 1 Territry bequeathal strategy f female red squirrels as a functin f litter emergence date. Mean emergence dates are different in females adpting different strategies. Site (pen circles): = 8.25, df = 2,9, p <.05; site (shaded circles): =.72, df = 2,8, p <.05. give all give part #0 c/5 keep them regularly n r near their middens. Accrding t trap data, n males disappeared ver the summer. Adult dispersal in summer is linked with the presence f juveniles. At the time f bequeathal, dispersing females weighed far mre than their inheriting ffspring (females 26 8 g; juveniles g). Juveniles were unsuccessful in cmpetitin with adult females fr vacant territries ver the summer. O Emergence date (days after April 1) Juvenile success Juvenile emergence dates ranged frm 10 May t 12 June, with the earliest juvenile emerging 10 days befre any ther. When juveniles first left the nest, they weighed between 70 and 110 g (linear interplatin between measurements befre and after emergence frm the nest). Early juveniles were being weaned and were already calling (first call days after emergence; end f lactatin days after emergence), while late juveniles were just emerging frm their nests. There was n crrelatin between emergence date and litter size Table 1 Pst-lactatinal mass and fur cnditin (five-pint rank scale) f female red squirrels adpting different bequeathal strategies Strategy Character Site eep Mass (g) ur cnditin n - Mean SE. b Site as blck. ruskal-wallis ANOVA fr each site. 16 Behaviral Eclgy Vl. N. 2 ' f Share Bequeath Test (df) = 0.11 (2,19) b C W = 2. (2) W = 0.62 (2) >.5 >.25 >.5 Dwnladed frm at Penn State University (Patern Lib) n ebruary 19, 2016 entire territry, seven bequeathed their territry t an ffspring, eight shared their territry, and tw lst their territries t intruding adult squirrels. One squirrel lst her territry t a challenging adult befre juvenile emergence. Thirty percent f females with a territry and with juveniles chse t bequeath. ive f the seven dispersing females gave n territrial rattle calls during bservatin perids fr several days befre disappearing ( days; all means are given SE). The remaining tw females suddenly disappeared frm their territry and were nt seen fr 20 days and 56 days, respectively, until they claimed new territries. Tw f the eight females wh eventually shared their territry with an ffspring als stpped calling frm their territry and disappeared fr 2 and 29 days, eventually returning t defend the territry. In the latter case, a juvenile wh had inherited the mther's territry died frm a bite wund inflicted by a grund squirrel befre the female's return. Three f the bequeathing females wn new territries: tw fllwing injuries t the frmer tenant and ne settling interstitially n a small, undefended area. The fur remaining females did nt defend territries n the study sites at the end f August. We saw and trapped three f the fur peridically at distances up t 0.5 km frm their previus territry, especially during territrial disputes. We did nt see the seventh female after her 2-week perid f skulking abut her territry. Seven (f 28) breeding females dispersed. N nnbreeding females (f 1) dispersed, althugh tw were frced frm their territries due t injury. Althugh we did nt bserve males, we trapped Crrelates f bequeathal Territrial bequeathal is related t the emergence date fjuveniles. emales wh bred early were mre likely t keep their territry, while thse wh bred late were mre likely t give up part r all f their territry (igure 1; tw-way ANOVA: = 7.8, df = 1,21; p <.01). Orthgnal cntrasts (Wilkinsn, 1988) shw that significance is attributable t differences between the emergence dates f females keeping their territry as ppsed t thse either giving up part r all f their territry. We culd detect n relatinship between a female's cnditin (lw pst-lactatin mass, pr fur cnditin) and her behavir (Table 1). These data shw n indicatin that females with a lw residual reprductive value gave up their territry mre readily. Neither culd we detect a cnsistent trend between territry quality and female behavir: emales with a small r pr-quality territry did nt bequeath mre frequently, and females with a larger r higher-quality territry did nt share their territry mre frequently (Table 2). emales with smaller middens tended t disperse mre readily than did females with large middens, thugh we culd nt detect a cnsistent trend with the number f tunnels cntaining unpened cnes (Table ).

4 Table 2 Territry size and territry quality f female red squirrels adpting different bequeathal strategies Strategy Character Site eep Share Bequeath P(df) P Territry size (ha) Territry quality* n " (2,18) 0.7(2,18) >.5 >.25 a Site as blck. b Mean SE. Number f class 1 and 2 trees greater than 10 cm diameter at breast height. (site, r =.0, n = 1; site, r =.151, n = 1; mean litter size at emergence = 2.). Early and late juveniles grew at similar rates n the natal territry (regressin f grwth rates frm emergence t 150 g, using mean rate fr each family against emergence date, t = 0.621, df = 16, p >.5; mean grwth rate = 9.6 g/day). Weight rank between families was unchanged ver this time (Spearman rank crrelatin between date f emergence and date juveniles reach 150 g: site, r s =.88, n = 9, p <.01; site, r s =.962, n = 11, p <.01). Thus, grwth rate did nt increase t cmpensate fr late emergence date, and early juveniles were therefre larger than their late neighbrs at any given date. Mst juveniles (51/60) left their natal territry and wandered, smetimes mre than 1 km, returning t the natal territry between explratins. Althugh there is n difference between early and late litters in the frequency fjuveniles leaving hme (early litters, 27/2 left; late litters, 2/28 left; x 2 = 0.02, df = \, p >.25), there is a difference between early and late litters in the prprtin f juveniles wh settled n an independently wn territry, n their mther's territry, r wh were unaccunted fr at the seasn's end (70 days after emergence; x 2 = 9.8, df = 2, p <.01; igure 2). Early juveniles tended t settle n either a small, interstitial territry clse t hme r n a territry left vacant by anther squirrel, whereas late individuals were mre likely t settle n their mther's territry (i.e., late juveniles less ften acquired a territry ther than their mther's). One interesting utcme was that all late litters (12/12) had at least ne juvenile settled, while sme early litters (/1) were cmpletely unsuccessful n the study sites. N untagged juveniles (i.e., frm ff the study sites) acquired a territry n either study site. Althugh we rarely bserved untagged individuals, 11 f the 16 bservatins between 1 July and 17 August ccurred during territrial disputes (after the death r injury f an established squirrel). Juveniles frm a large, surrunding area seem t be aware f, and t cmpete fr, any available territry. emale cnditin Althugh there was n difference between the mass f bequeathing and nnbequeathing females (Table 1), there was a relatinship between female mass and breeding date. emales wh bred early were mre likely t lse mass between the birth and emergence f their juveniles, whereas females wh bred later were mre likely t gain mass (igure ). N nnbreeding females lst mass. Between 20 May and 10 June, breeding female mass either increased r remained cnstant, independent f litter emergence date (site : r =.02, n = 11, p >.1; site : r.5, n = 10,/> >.1; mean mass increase = 16. g). This pattern suggests an external, perhaps seasnal effect. Indeed, the mass increase ccurred at a time crrespnding t the suddenly increased availability f summer fd (buds, pllen cnes, mushrms). Table Area and cntents (number f tunnels with whle cnes) f middens f female red squirrels adpting different bequeathal strategies Strategy Character Site eep Share Bequeath /-(df) P Midden area (m 2 ) Tunnels with whle cnes n b (2,18) 1.85(2,18) >.O7 " Site as blck. b Mean + SE. Price and Butin Territrial bequeathal in red squirrels 17

5 their territries. Challenged females had a prer fur cnditin than did thse nt invlved in fights (females invlved in disputes: cnditin = ; females nt invlved in disputes: cnditin = ; Mann-Whitney U = 12, n, = 15, n2 = 11, p <.05). Territrial disputes may als lead t an increased predatin risk. Squirrels invlved in chases seem t have decreased awareness f their surrundings and smetimes run up t, r even nt, an bserver. Tw f the three gshawk predatin attempts bserved ccurred during extended chasing episdes. The third invlved a vagrant squirrel trespassing n a territry. Tw f the three predatin attempts were successful. CO if) wn mther's unaccunted.a E Late Early DISCUSSION The fur cnditin f breeding females wrsened between juvenile emergence and independence (emergence: ; independence: ; Mann-Whitney U = 70.5, n = 2, p<.001). Nnbreeding females shwed n significant decrease in fur cnditin ver this perid (first measurement: ; secnd measurement: ; Mann-Whitney U = 6.5, n = 8, p =.5). Territry lss Three female squirrels lst their territries t intruding adults n ur study sites (tw n site and ne n site ). These females were all early breeders. They weighed significantly less than did females wh retained their territries (first measurement after birth f the juveniles: females wh lst their territry = 22 6 g; females wh retained their territry = 25 g; t = 2.6, df = 2, p <.05). emales lsing their territry had a lwer mean fur cnditin befre the lss than did ther breeding females ( versus.2 0.1; Mann-Whitney U = 60.0, n, = 2,n2 =,/><.05). One female was nt seen subsequently; her nestlings were emaciated when last seen and very likely died. The remaining tw females became vagrant; f these, ne settled within 2 weeks and the ther disappeared. On bth f these territries, the litters had emerged befre the takever. In ne family, ne juvenile died frm a squirrel-bite wund sustained during the takever; the ther tw juveniles disappeared. In the secnd family, bth juveniles disappeared initially, but ne reappeared 6 weeks later and regained its mther's territry. All three displaced females were invlved in prlnged chasing fr -7 days befre disappearing. Eight ther breeding females were challenged n 0- s B - igure Mass change f breeding female red squirrels between parturitin and juvenile emergence as a functin f litter emergence date. Site (shaded circles): r =.87, n = \\,p <.001; site (pen circles): r =.75, n = 10, p < ' O c in in CO -10" ' ft 0-0- * Emergence date (days after April 1) Behaviral Eclgy Vl. N. 2 Breeding female squirrels actively bequeath territries t their ffspring. It is unlikely that ur bservatins dcument simple searching fr better territries because n nnbreeding females r males dispersed ver the study perid, and we have n reasn t assume that breeding females live n inferir territries. Neither d ur bservatins dcument passive dispersal f females wh are utcmpeted by their juveniles. Bequeathing females were larger than inheriting juveniles, and vagrant females beat juveniles t newly vacated territries (see als Butin et al., 199; Price et al., 1986). Late-breeding females mre ften bequeath their territry t their juveniles than d early breeders. Early-brn juveniles are mre likely t acquire a territry independently, but early-breeding females lse mass and may risk lsing their territry. Such territry lss by a female with unweaned juveniles likely translates int lss f the litter as well as a decreased prbability f survival fr the mther. The chice f breeding date likely depends n a female's cnditin befre breeding. Late-brn juveniles mre ften settled n their mther's territry via bequeathal r sharing. We cannt tell hw well these juveniles wuld survive in the absence f territrial bequeathal withut experimental manipulatin. Data frm ther ppulatins suggest that later litters have prer survival (Gurnell, 198; Rusch and Reeder, 1978), but these studies did nt examine female territrial behavir. A playback experiment (Price et al., 1990) shwed that females defended early litters mre strngly than late litters, suggesting that early-brn juveniles are mre valuable t their mther. We fund n direct evidence that females in pr cnditin were mre likely t bequeath their territries, but there was evidence that they were mre likely t lse their territries t an intruder. The bserved breeding date-dependent change in mass, cupled with the relatinship between breeding date and bequeathal, suggests that female cnditin may be ultimately imprtant. A stchastic, dynamic prgramming mdel investigating bequeathal (Price, 1992) predicts that tw classes f females shuld disperse: thse in pr cnditin late in the seasn (i.e., with pr reprductive ptential), and thse in gd cnditin early in the seasn (i.e., able t win a new territry). This mre cmplex pattern f seasn/cnditin interactins may mask any verall relatinship. Our sample size is t small t distinguish between the seasn and cnditin effects. Dwnladed frm at Penn State University (Patern Lib) n ebruary 19, 2016 igure 2 Number f early and late juvenile red squirrels settling independently, settling n their mther's territry, r unaccunted fr (disappeared r dead).

6 We fund n evidence that squirrels with prerquality territries were mre likely t bequeath, althugh females with smaller middens did tend t give up their territries mre readily. A small midden might prvide insufficient fd fr breeding next year and may decrease the relative cst f dispersing. The lack f any behaviral trend with the number f remaining cnes in a midden (a measure f a territry's immediate wrth, as ppsed t lng-term wrth) des nt supprt this interpretatin. Hwever, particularly lw numbers f cnes in middens, pssibly due t the cne failure, hinder trend detectin. Individuals with large territries did nt share their area mre frequently. In their manpulative study, Butin and Schweiger (1988) shwed that females with expanded territries gave up part t their ffspring. In the present study f an unmanipulated ppulatin, intruder pressure prbably held territry size t lw t bserve this effect. Other crrelates f sharing are yet t be investigated. Our demnstratin f a pssible cst t early breeding wuld be enhanced with knwledge f female cnditin befre breeding. Wauters and Dhndt (1989) fund that Eurpean red squirrels (Sciurus vulgaris) reach a threshld mass befre breeding. Becker (Alta U, unpublished data) fund a similar threshld in red squirrels in Alberta. Neither study examined the time f breeding relative t mass. If a female in pr cnditin delays breeding until summer fd is available, and if, as a cnsequence, she gives up her territry at the seasn's end, it might be better nt t breed that year. A yung female shuld perhaps freg reprductin 1 year and allcate investment instead t fd strage r t grwth. In this way, she wuld increase the chances f breeding early the next year. Alternatively, a female with a pr territry might chse t freg breeding and instead attempt t take ver a better-quality territry. emale red squirrels, particularly in nrthern ppulatins, d freg breeding in sme years. Bequeathal has nt been investigated in suthern ppulatins f red squirrels. With lwer levels f verwinter stress, and hence with well-stcked territries less crucial t survival, juveniles may be mre likely t survive until spring n interstitial territries. uture studies shuld search fr bequeathal. Althugh this study dcuments bequeathal in nly 1 year, a year f cne failure, squirrels als bequeath in years f relative fd abundance (three f seven breeding females dispersed in 198, a mast year with superabundant cnes; Price et al., 1986). The quantitative pattern f bequeathal between individuals likely varies between years. Territrial bequeathal seems a surprising behavir fr tw reasns: first, it is likely t be risky fr the female (the subsequent search fr a new territry may be unsuccessful), and, secnd, it is nt a strategy adpted nly by females in pr cnditin; that is, bequeathal is nt a desperate actin taken by individuals with lw reprductive ptential. We hypthesize that in red squirrels, breeding date depends n female cnditin and that bequeathal has evlved as a cnsequence f this cnditin dependence in the face f limited resurces. If a female delays breeding t take advantage f the increase in available fd and if her juveniles are unsuccessful in cmpeting with their larger, early-brn neighbrs fr territries, we expect that, under certain cnditins, females will bequeath. These cnditins, which can be applied t ther cases f breeding dispersal, include (1) lw verwinter survival f juveniles withut a territry, (2) females being better able t fight fr and acquire a territry than juveniles, and () juveniles being able t defend their natal territry. We thank Rn Ydenberg, Larry Dill, Clive Welham, Sctt rbes, and Dave Daust fr supprt thrughut the prject and fr cmments n the manuscript, Dawn Michalk and Cbey Langill fr assistance in the field, and Christph Rhner, Helene Lair, Clin Clark, and Dusty Becker fr discussin. Helene Lair, Olf Liberg, and Dn ramer did much t strengthen the manuscript. The Natural Sciences and Engineering Research Cuncil f Canada prvided financial assistance thrugh a 1967 Science and Engineering Schlarship t.p. and thrugh perating grants t S.B. and t Rn Ydenberg. REERENCES Andersn P, Dispersal in rdents: a resident fitness hypthesis. Special publicatin n. 9. American Sciety f MammaJgists. Butin S, Schweiger S, Manipulatin f intruder pressure in red squirrels (Tamiasciurus hudsnicus): effects n territry size and acquisitin. Can J Zl 66: Butin S, Tze Z, Price, 199. Pst-breeding dispersal by female red squirrels (Tamiasciurus hudsnicus): the effect f lcal vacancies. Behav Ecl : Byce CC, Byce JL, Ppulatin bilgy f Micrtus arvalis II. Natal and breeding dispersal f females. J Anim Ecl 57: Cckburn A, Braithwaite RW, Lee A, The respnse f the heath rat, Pseudmys shrtridgei, t pyric successin: a temprally dynamic life-histry strategy. J Anim Ecl 50: isher RA, The genetical thery f natural selectin, 2nd ed. New Yrk: Dver. Gittleman JL, Thmpsn SD, Energy allcatin in mammalian reprductin. Am Zl 28: Gurnell J, 198. Squirrel numbers and the abundance f tree seeds. Mammal Rev 1:1-18. Harris MA, MurieJO, 198. Inheritance f nest sites in female Clumbian grund squirrels. Behav Ecl Scibil 15: Hward PC, Spatial rganisatin f cmmn reedbuck with special reference t the rle f juvenile dispersal in ppulatin regulatin. Afr J Ecl 2: emp GA, eith LB, Dynamics and regulatin f red squirrel (Tamiasciurus hudsnicus) ppulatins. Eclgy 51: Lair H, The calls f the red squirrel: a cntextual analysis f functin. Behaviur 115: Macarland Symingtn M, Sex rati and maternal rank in wild spider mnkeys: when daughters disperse. Behav Ecl Scibil 20: Matthews JD, 196. actrs affecting the prductin f seed by frest trees. r Abstr 2:i xiii. Mntgmery GG, Sunquist ME, Habitat selectin and use by tw-ted and three-ted slths. In: The eclgy f arbreal flivres (Mntgmery GG, ed). Washingtn, DC: Smithsnian Institutin Press; Price, Territrial defence and bequeathal by red squirrel (Tamiasciurus hudsnicus) mthers in the nrthern breal frest (MSc dissertatin). Burnaby, British Clumbia: Simn raser University. Price and Butin Territrial bequeathal in red squirrels 19

7 Price, Territrial bequeathal by red squirrel mthers: a dynamic mdel. Bull Math Bil 5:5-5. Price, Butin S, Ydenberg RC, Intensity f territrial defence in red squirrels: an experimental test f the asymmetric war f attritin. Behav Ecl Scibil 27: Price, Brughtn, Butin S, Sinclair ARE, Territry size and wnership in red squirrels: respnse t remvals. Can J Zl 6: Rusch DA, Reeder WG, Ppulatin eclgy f Alberta red squirrels. Eclgy 59: Smith AT, Ivins BL, 198. Clnizatin in a pika ppulatin: dispersal vs philpatry. Behav Ecl Scibil 1: 7-7. Smith CC, The adaptive nature f scial rganizatin in the genus f three [sic] squirrels Tamiasciurus. Ecl Mngr 8:1-6. Smith CC, Structure and functin f the vcalizatins f tree squirrels (Tamiasciurus). J Mammal 59: Stearns SC, Life-histry tactics: a review f the ideas. Q Rev Bil 51:-7. Waldrn RM, Cne prductin and seedfall in a mature white spruce stand. r Chrn 1: Waser PM, Jnes WT, Survival and reprductive effrt in banner-tailed kangar rats. Eclgy 72: Wauters L, Dhndt AA, Bdy weight, lngevity and reprductive success in red squirrels (Sciurus vulgaris). J Anim Ecl 58: Wilkinsn L, SYSTAT: the system fr statistics. Evanstn, Illinis: SYSTAT. Williams GC, 1966a. Adaptatin and natural selectin: a critique f sme current evlutinary thught. Princetn, New Jersey: Princetn University Press. Williams GC, 1966b. Natural selectin, the csts f reprductin, and a refinement f Lack's principle. Am Nat 100: Wlfenden GE, itzpatrick JW, The inheritance f territry in grup-breeding birds. Biscience 28: Behaviral Eclgy Vl. N. 2