802 动 物 interactive effects of coexisting populations of the two species.Most of the above studies were either short-term experimental studies or involved only one habitat. As a part of a 25 year study of population dynamics 0f M.ochrogaster and M.pennsylvanicus(Getz et a1., 1987,2001),we conducted 10—20 year manipulative studies in open populations to test the role of interspecific competition in the dynamics of these two coexisting species of voles.Experimental manipulations involved removal of one species from marginal(bluegrass)and low (tlalrgass)quality food habitats to reduce interspecific competition.Speciifcally,we tested the hypotheses that there all negative interspecific interactions between coexisting populations of M.ochrogaster and M.pennsyl- vanicus that result in depressed densities of the other species and asynchronous population fluctuations. Because of time and space constraints,we were unable to replicate our study sites.Rather,we opted for larger study sites to reduce local site condition bias and for long- term manipulations,to reduce short—term temporal bias. 1 Methods and materials 1.1 Study sites The study sites were located in the University of Illinois Biological Research Area(“Phillips Tract”),6 km NE of Urbana,Illinois,USA(40。15 N,88。28 W). Populations of M.ochrogaster and M.pennsylvanicus well monitored monthly in two habitats:bluegrass Poa pratensis and restored tlalgrass praiire(a mixture of big bluestem, Andropogon gerardii; Indian rgass, Sorghastrum nutans;and switch grsas,Panicum spp.). Tallgrass praiire was the original habitat of both species in Illinois,whereas bluegrass,an introduced species, represents one of the more common habitats in which the two species can be found today in Illinois. lgrass is a very low food habitat,whereas bluegrass provides an intermediate food source(Cole and Batzli, 1979; Lindroth and Batzli,1984;Thompson,1965). hTe restored tallgrass prairie was established in 1968.When the sites well fisrt trapped in September 1 977.praiire vegetation was well established.Lindroth and Batzli(1984)recorded relative abundances of the most prominent plant species: Andropogon gerardii (38%);Chinese lespedeza Lespedeza cuneata(25%); Beard tongue foxglove Penstemon digital ̄(16%);and Sorghastrum nutans (19%). A11 other species represented<1%relative abundance. Bluegrass study sites were estbalished within a former bluegrass pasture that was released from grazing in spring 1971.Relative abundances of plants in the sites well:P.pratensis(70%);dandelion,Taraxacum ofifcinale(14%);wild parsnip,Pastinaca sativa(4%); goat’s beard,Tragopogon pratensis(3%);about 20 other species with relative abundance of≤1%(Getz et a1..1979). 学 报 53卷 A11 study sites well organized on a 10 m interval grid system.One wooden multiple-capture live—trap(Burt, 1940)was placed at each station.Each month,a 2.day prebaiting period was followed by a 3-day trapping session;cracked CO1T1 was used for prebaiting and as bait in traps.We used vegetation or aluminum shields to protect traps from the sun during summer.Wooden traps provided ample insulation in winter,and thus nesting material was not provided in the traps at any time.We estimated trap mortality to be<0.5%.Traps were set in the afternoon and checked at about 08:00 h and 15:00 h on the following three days.All animals were toe-clipped at fisrt capture for individua1 identiifcation(≤2 toes/ foot).All procedures were approved by the University of Illinois Laboratory Animal Care Committee and meet the guidelines recommended by the American Society of Mammalogists(Animal Care Committee,1998). Species, d station,individual identification,sex, reproductive condition(males,testes abdominal or scrotal;females,vagina open or closed,pregnant as determined by palpation,or lactating),and body mass to the nearest 1 g were llcorded at each capture.Animals well grouped by age based on body mass:adult,≥30 g; juvenile,≤29 g(Hasler,1975). 1.3 Manipulations hTe manipulations well carried out in a 2 ha(80× 250 m)bluegrass site(Getz et a1.,2001)surrounded on the north,west,and south sides by cultivated fields or unsuitable vole habitat and on the east by a 2 ha alfalfa ifeld(south 130 m)and an 0.8 ha bluegrass site(north 120 m).Aluminum flashing buried 0.5 m below the surface and extending 0.5 m above the surface divided the bluegrass field into two.1 ha sites. AU M.ochrogaster were removed from the south 1 ha bluegrass site(M.pennsylvanicus alone)from March 1977 through June 1987. M.pennsylvanicus were removed from the north 1 ha bluegrass site (M.ochrogaster alone)from March 1977 through May 1997. e adjacent 0.8 ha bluegrsas site served as the control for the M.pennsylvanicus alone and the M.ochrogaster alone sites from 1977—1987;both species were allowed to inhabit the south 1 ha bluegrass site from 1987—1997,thus serving as the contorl for the M.ochrogaster alone site during these years. Trapping in the 0.8 ha control sited from 1974~ 1976 by Cole and Batzli(1978)suggested this site provided suitable habitat for M.ochrogaster;they removed immigrant M.pennsylvanicus during their study. Both species were allowed to remain in the north 1 ha site from October 1971一June 1977 and in the south 1 ha site from July 1987一May 1997.Demographic data for these periods(Getz et a1.,2001,2005b)indicated both sites afforded suitable habitat for both species. A 2.6 ha(72×360 m)site in Phillips tarct(Getz 维普资讯 http://www.cqvip.com
5期 Lowell L.GETZ et a1.:Interspeciifc competition in vole populations 803 et a1.,2001)was used for the tallgrass manipulative study.The site was surrounded on three sides by cultivated fields and on the east(72 m)side by a site that underwent succession from a forb stage to shrubs and trees during the course of the study. Because M.ochrogaster populations were very low and M.pennsylvanicus very high in tallgrass for most of the study period(Getz et a1.,2001),only effects of M.pennsylvanicus on M.ochrogaster were tested in tallgrass.We removed M.pennsylvanicus from a 0.5 ha (72×75 m)tallgrass site at the west end of the tlalgrass prairie from September 1984 through May 1997;a 0.5 ha site at the east end of the prairie served as the contro1. From July 1 977 through April 1 984 both species were lalowed to remain in the west 0.5 ha site.Population densities of M.pennsylvanicus during this time indicated the site was suitable for the species. Removed animals were released on the opposite side of an Interstate highway,approximately 1 km from the study sites.None returned to the study sites. 1.4 Data analysis We estimated survival as the proportion of animals (totla population)that survived from one month to the next.Animals that weighed≤29 g when first captured were presumed to have been born on the study site since the last monthly trapping session. We calculated persistence of these animals as the time elapsed from fisrt capture to their disappearance from the site;young voles captured in only one month were ven a persistence of one month.Unmarked adult animals caught during a given month were presumed to have been born elsewhere and to have immigrated into the study site since the previous monthly trapping session(Dueser et a1.,198 1; Getz et a1.,2005a;Tamarin,1984).We calculated the number of immigrants per month,adjusted to number of immigrants/ha.We also calculated for each month the proportion of adult females that had been recorded as reproductively active (vagina open,pregnant, or lactating).We compared population density,survival, persistence of young,proportion of immigrants,proportion of reproductively active adult females,and in the site where each species was alone(“experimentla”)with the site where both species were present(“contorl”). We analyzed our data within the capture..mark.. recapture(CMR)framework(Lebreton et a1.,1992; Williams et a1.,2001).We implemented CMR methods using Program MARK(White and Burnham,1999)to estimate demographic parameters,to test speciifc hypotheses regarding these parameters,and to evaluate the effect of removal of the competitor species. Speciifcally,we used Cormack.Jolly.Seber(CJS)models (Cormack,1964;Jolly,1965;Seber,1965)to estimate apparent survival(声)and recapture rates(』D),and to test for variation in survival rates among experimental groups and through time.Because the CMR model did not provide useable estimates of population density and survival at very low densities,we excluded extended trough periods(periods with<20 voles per ha)from our analysis. We applied goodness-of-fit tests using RELEASE TEST 2+3(in program MARK)to test if our global model fits the data.When the dispersion parameter(e) was greater than 1,we corrected for over-dispersion using the calculated e.Next.we used Akaike’s information criterion corrected for small sample size and adiusted for quasi-likelihood(QAIC )for model comparison,and for the identification of the most parsimonious model;the lower the QAIC value,the more parsimonious the mode1. The most parsimonious model in the candidate model set was compared to neighboring models based on the differences in QAIC values,AQAIC .If AQAIC <2, both models are equally supported by the data.If 2<< 7,there iS considerable support for a real difference between the models,and if>7,there iS a strong evidence for a difference between the models(Burnham and Anderson,2002). We performed our analyses separately for three species.habitat combinations: (1)M.ochrogaster in bluegrass habitat,(2)M.ochrogaster in tlalgrass,(3) M.pennsylvanieus in bluegrass habitat(Table 1). Because study duration was up to 20 years in some sites and included extended trough periods,it was not possible to run a single model for each species.habitat combination.We,therefore,divided the data into shorter periods, which resulted in two data sets for M.ochrogaster in bluegrass habitat(data sets#1 and# 2),two data sets for M.ochrogaster in tlalgrass habitat (data sets #3 and #4),and one data set for M.pennsyl vanicus in bluegrass habitat(data set#5; Table 1).Our preliminary analysis showed that survival rates did not differ signiifcantly between sexes in either species;therefore,for each species,we combined data ofr both sexes in all further analyses. Effects of competitor removal on SUrvival and recapture rates were analyzed by examining both combined and additive effects of competitor removal and time (Williams et a1.,2001).For example,a model denoted sa(exp time)includes main effects of competitor removal and time,and also the interaction effect between these two covariates,whereas(exp+time)includes only the main effects of competitor removal and time.The time effect was included as a fixed effect(i.e.,separate estimates for each month). We also tested for correlations between p0pulati0n densitv 0f 0ne species and m0mhly surviva1 mtes, and pmp0rti0n 0f repmductiVely actiVe females 0f the 0tl1er species in the bluegrass c0ntml site fmm March 1977一 June 1 987.These anaJyses fesfed p0femiaJ efrecfs 0f e c0existing species 0n each 0ther within a site and were li ted t0 the time the site served as a c0ntIDl f_0r the 维普资讯 http://www.cqvip.com
804 动 物 学 报 53卷 experimental sites.In order to reduce autocorrelation effects,we used data from every other month for these analyses.1'he CMR model did not provide useable estimates of population density and survival at densities< 20 voles/ha,which was greater than the mean density of food availability during winter, we also tested for correlations between population density of a species and survival and proportion reproductively active females of the other species during December—February. All original capture data and explanatory files from the 25.year study are available to anyone wishing to make use of them at:http://www.1ife.uiuc.edu/getz/and http://hd1.handle.net/2142/172. the two species in bluegrass(Getz et a1.,2001);we, therefore,utilized data from Getz et a1.(2001)for these analysis. Because interspeciic efffects may be most pronounced during periods of low resources,e.g.,low Table 1 Description of each dataset used for capture・mark-recapture analysis:species,habitat,start and end dates,study duration,and number of vgles in control and experimental groups that were used in CMR analysis for each species-habitat eombination Excluding the number of competltor species. 1.5 Statistieal analyses We used paired.sample t.tests to compare within each habitat persistence of young,proportion of 2.2 Demographic variables Goodness-of-ift tests indicated that the global CJS model fit the data with a moderate under-dispersion eprroductively active aduh females, and number of immirants betgween experimentl aand control sites. Because most of the variables did not meet the (<1.0)in one data set(M.pennsylvanicus in bluegrass;Table 2).In the other four data sets,there were slight over-dispersions(1.0<e<3.0),and we used the calculated e values for quasi.1ikelihood requirements for normality(population densities and demographic variables were non normal at the 0.05 level: adjustment of models and their parameter estimates(Table Kolmogorov—Smimov test,Zar,1999),all variblaes were log.transformed.For variables that were“zero’’we added 0.00l prior to transformation.SPSS l0.0.7 for Macintosh 3).Our analyses using CJS models showed that the most parsimonious recapture rate models differed among the five data sets(Table 1).It included no effect in data sets#1 (SPSS,Inc.,2001)was used for these statistical analyses. and#4(Models#2 and#17,Table 3),additive effects of competitor removal and time in data set#2 2 Results 2.1 POpulation density In bluegrass, mean population density of (Model#8)。only a time effect in data set#3(Model #l2).and interactive effects of competitor removal and time in data set#5(Model#21). The most parsimonious survival rate model included only the time effect in data sets#1.#2 and#5.the effect of competitor removal in data set#3.and the additive effects of competitor removal and time in data set M.ochrogaster was lower in the experimental than in the contolr site(MeaR±SE,35.5±4.2 and 41.1±4.9 voles/ha,respectively;t=2.381,af:96,P=0.019; Fig.1).In tlalragss,amplitudes of fluctuation of M.ochrogaster and mean population density were higher #4(Table 3).Survival rates varied substantillay through time in data sets#l,#2。#4,and#5,but not in the experimentl tahan control site(MeaR density。76.6± 9.2 and 37.7±6.2 voles/ha。respectively;t=8.002, dataset#3(Figures l一3).1'he effect of competitor removal was included in the most parsimonious models for df=69,P<0.001;Fig.2).Mean population density of M.pennsylvanicus in bluegrass was lower in the data sets#3 and#4(M.ochrogaster in tlalragss). Comparison of the average survival rates in the experimentl and contraol groups indicated that competitor removal might have had a slight positive effect on M. experimentl tahan in the contolr site(30.6±3.4 and 55.0±4.6 voles/ha,respectively;t=6.170,af=62, P<0.00l;Table 3).Because M.pennsylvanicus densities were high in tllaragss at all times(Getz et a1., ochrogaster in tllarasgs habitat(Fig.4).However, AQAIC between these models and those that excluded the effect of competitor removal was less than 2,indicating that this effect was not substantia1. 2001),it appeared that there was no suppressing efect of M.ochrogaster on M.pennsylvanicus in this habitat. 维普资讯 http://www.cqvip.com
5期 Lowell L.GETZ et a1.:Interspeciifc competition in vole populations 805 ochrogaster in bluegrass (Datasets#1 and≠f2、 暑 专 三 ∽ § ’晶 勺 .宝 盘 j 0 Fig・1 Temporal variation in monthly survival rates of M.ochrogaster in bluegrass habitat(A)and monthly estimates of population densitis(iendiidualvs/hectare)are given for competitor removal and control sites(B) Because competitor removal did not signiifcantly influence survival,survival rates were estimated using model{≠(month)p(.)}for data set #1 and model{ (month)p(exp+month)f for data set#2(Table 3).Grey shaded area indicates 95%co dence intefvals. Table 2 Goodness-0f_fit tests for the global Cormack-Jolly-Seber models for each dataset The values of the variance ilatfnion factor(e)also axe given F0r M.ochrogaster in bluegrass.persistence of (2.3土0.1 and 2.2土0.1 months,respectively:t= young did not differ between the experimental and control 0.247,df=61,P:0.806)and proportion of reproductively active females(0.64土0.07 and 0.66土 sites(1.9±0.1 and 2.0±0.1 months,respectively:t =0.672,df=312,P=0.502).The proportion of 0.05,respectively;t=0.957,df=22,P=0.349) did not difer between the experimental and control sites. he numTber of immigrants was greater in the experimental reproductively active females(0.65±0.03 and 0.76土 0.03,respectively;t=2.102,af=99,P=0.038) was lesser in the experimental than the control site.The than in the contolr site(2.6土0.4 and 2.0±0.4 number of immigrants was lower in the experimental than immirantgs/ha,respectively;t=4.475,df=147,P< 0.001). For M.pennsylvanicus in bluegrass,persistence of young did not difer between experimental and control the control site(3.0±0.4 and 3.6±0.4 immirantgs/ha. respectively;t=3.471,d厂=239,P=0.001). For M.ochrogaster in tallgrass,persistence of young 维普资讯 http://www.cqvip.com 806 动 物 学 报 53卷 ochrogasterin bluegrass (Damsel#3 and蛆) 1 O O O O O O 8 6 4 2 O 至 § ∽ 240 ’ a 160 a 9 j 0 山 80 0 Fig・2 Temporal variation in monthly sur ̄val rates of M.ochrogaster in tallgrass habitat(A)and monthly estimates of population densities(indiVid岫ls,hectare)are given for competitor removal and control sites(B) Because competitor removal did not signiifcantly influence survival,survival rates were estimated using model{ (month)p(month)}for data set#3 and model{ (month)p(.)}for data set#4(Table 3).Grey shaded area indicates 95%conidence ifntervals. sites(2.2±0.1 and 2.1±0.1 months,respectively:t =M.ochrogaster,whether orf lla months(r=0.021,n= 0.425,df=275,P=0.671).Proportion of reproductively active females did not differ between the 49,P=0.885)or for winter alone(r=0.055,n=23, P=0.803),or for lag periods of up to three months. Neither was the pmpo ̄ion of reproductively active female M.pennsylvanicus correlated with population density of experimental and control sites(0.56±0.04 and 0.60-t- 0.04,respectively;t=1.422,df=74,P=0.159), and immigrants were slightly fewer in the experimental than control site(2.8±0.3 and 3.8±0.6 immirantsg/ M.ochrogaster(all months:r=0.033,n=49,P= 0.824;winter:r=0.012,n=23,P=0.958).The proportion of reproductively active female M.pennsylvani— c was positively correlated with population density of ha,respectively;t=2.154,df=Il9,P=0.033). 2.3 Correlation analyses Population densities of M.ochrogaster and M.ochrogaster at a lag period of three months,but not or one or two month lags.f Monthly survival of M.pennsylvanicus were negatively correlated in the bluegrass control site(r=一0.417,n=62,P= 0.001). Monthly survival of M.ochrogaster and M.pennsylvanicus in bluegrass was not correlated, M.ochrogaster in bluegrass was negatively correlated with population density of M.pennsylvanicus for all months(r 0.422,n=49,P=0.003),including lag periods of =whether for lla months(r=0.157,n=36,P=0.361) or only winter months(r=0.143,n=17,P=0.584). Proportion of reproductively active females of the two one to three months,but not orf winter lone(r=0.a322, n=24,P=0.124).The proportion of reproductively active female M.ochrogaster was not correlated with species was correlated for all months(r=0.635.n= population density of M.pennsylvanicus(all months:r= 0.023,n=40,P=0.890;winter:r=0.037,n: 34,P<0.001),but not for the winter months alone(r 0.434.n=15,P=0.160). =23,P=0.867),including lag periods of one to three months. Monthly survival of M.pennsflvanicus in bluegrass was not correlated with population density of 维普资讯 http://www.cqvip.com 5期 Lowell L.GETZ et a1.:Interspeciifc competition in vole populations 807 Diferences in Akaike’s Information Criterion corrected for small sample size and aJjusted for quasi-likelihood(AQAICc),QAICc weights,number of parameters and model deviances are given for each mode1.Symbols are;≠:apparent survival rate,p recapture rate,exp effect of competitor removal,and month:time effect.A plus sign(+)denotes additive effects,and a multiplication sin(。)denotges interaction effect.A perid(.)iondicates constnta value of the parameter (model with intercept only).Only five most parsimonious models are given for each dataset.The most parsimonious model for each dataset is indicated in bold. 3 Discussion Results of experimental removal of Microtus ochrogaster and M.pennsylvanicus from open populations in bluegrass and tallgrass habitats,indicated demography of coexisting populations of the two species Was not negatively affected by interspeciic ifnteractions where food densities,proportion reproductively active females,and number of immigrants were greater in a site where the two species coexisted than where alone.Monthly survival and persistence of young on the site did not differ between the experimental and control sites.Population densities of M.pennsylvanicus in bluegrass sites also were higher where the two species coexisted than where alone. resources were at least moderately high(bluegrass).In fact,population densities of both species in bluegrass were higher where the two species coexisted,e.g.,the control site.Only demography of M.ochrogaster was negatively affected by presence of M.pennsylvanicus Further, survival,persistence of young,proportion eprroductively active females,and number of immigrants did not differ between experimental and control sites. Microtus pennsylvanicus exerted a strong suppressing effect on population densities of M.ochrogaster in lOW food resource tallragss.Krebs(1977)alSO found no suppressive interaction between these two species where food resources were adequate. For M.ochrogaster in bluegrass,opulation ptallgrass.Survival,persistence of young,and proportion reproductively active females did not differ,however, between sites where M.ochrogaster Was alone and where M.pennsylvanicus were present.Only the number of 维普资讯 http://www.cqvip.com 808 动 物 学 报 53卷 pennsylvanicus in bluegrass (Dataset#5) 1.0 31矗_I >_【>.II1力 0.8 0.6 0.4 0.2 0.0 >、 = q 0 . l60 0 0 二 l978 l979 1980 l981 === 1982 Fig.3 Temporal variation in monthly survival rates of M.pennsylvanicus in bluegrass habitat(A)and monthly stiemates of population densiites(indiidualvs/hectare)are given for competitor removal and control sites(B) Because c0mpetitor removal did not signiifcanly itnfluence survival,survival rates were estimated using model{≯(month)p(exp month)} for data set#5(Table 3).Grey shaded area indicates 95%conidence ifntervals. 营 .三 艺 (;; Damset#1 Damset#2 Dataset#3 Dataset≠ Da ̄set#5 Fig.4 Average monthly survival rate estimates for experimental and control groups in each of the five datasets Mean values and 95%c0ntldenee intervals w℃re estimated using the models that included only the removal effect on survival rate i庐 【exp) See Table l forthe description of each dataset, immigrants was slightly greater where M.ochrogaster was lone,iansufficiently SO to affect population densities. control site.but the higher in the experimental than the affect on population density would have been inconsequenti1.Presence of M.ochrogaster did nota Monthly survival of M.ochrogaster may have been slightly 维普资讯 http://www.cqvip.com
5期 Lowell L.GETZ et a1.:Interspeciifc competition in vole populations 809 appear to affect demography of M.pennsylvanicus in opportunity for sufficient interspecific interactions t0 negatively affect demographic variables of the coexisting species.As we observed in our studies.the two species seldom occurred sympatrically at high densities. tallgrass(Getz et a1.,2001). Effects of season.especially periods of low food availability,as occur in winter,could not be tested wi【h the CMR model:data for the trou【gh phase were too few or detailed analfyses of surviva1.Sample sizes also were too small to test for seasonal diierences in tfhe other variables between the experimental and control sites. Getz et a1.(2006 a,b)suggested that population densities of M.ochrogaster and M.pennsylvanicus in our study sites were kept at low densities much of the time bv the net effect of predation from multiple generalist predators.Population densities of generalist predat0r species most likely were controlled by factors in additi0n to vole densities within our study sites.Because of the Correlation analyses of effects of density of one species on the other for coexisting populations in bluegrass indicated, however, no affect of population density on anv demographic variable during winter.In the bluegrass control site,survival and the proportion of reproductivelv active female M.pennsylvanicus were not correlated with population density of M.ochrogaster,whether all seasons or only winter were included in the analyses.Surviva1. but not proportion reproductively active female M.ochrogaster,was negatively correlated with population densities of M.pennsylvanicus when all seasons were included in the analyses. Survival of M.ochrogaster during winter,however, was not correlated with oppulation density of M.pennsylvanicus. The proportion of reproductively active female M.pennsylvanicus and M.ochrogaster were positively correlated in bluegrass,suggesting that what was good for reproduction of M.pennsylvanicus was also good for reproduction of M.ochrogaster.In contrast,survival 0f the two species was not correlated,indicating that the species were subjected to different mortlaity factors(Getz et a1.,2005b).Differentila mortlaity,not reproduction. thus appears to be the primary factor responsible for the negative;correlation in population densities of the coexisting species within a site. Of 25 population fluctuations of M.ochrogaster and 1 4 0f M.pennsylvanicus during the 25 years of the general study,there were only two instances of synchrony of oppulation fluctuations of the two species,both in bluegrass(Getz et a1.,2001).K1att(1986)studied M.ochrogaster and M.pennsylvanicus populations at the same study sites and concluded competitive advantage to the species fisrt entering a site.Tazik and Getz(2007) provided evidence for interspecific territorial behavior between females of the two species,at low to moderate population densities.Getz et a1.(2005a)have shown that the number of immigrants of M.ochrogaster and M.pennsylvanicus into a site is very low most months, thus creating conditions for competitive exclusion of a species through interspeciifc territorial behavior.We suggest that at very low densities,interspecific territorial behavior allowed sufifcient resources for survival and reproduction of the two species.The species with the greatest number of immigrants into a site may have suppressed population growth of the other species through direct effects of interspeciifc territoilal behavior.Thus. over the long term,there would have been reduced independent nature of population fluctuations of such diverse predator species as raptors,Large and small mannals,and snakes,as well as variation in numerical and functional responses of these predators(Pearson, 1985),we sepculate the net effects of predation mav be greater in some years than in others.During years with high predation effects,one or both species may have disappeared from the site.At these times,the species remaining at low densities in the site,or the fisrt one to recolonize the site,may have impeded settlement of immigrants of the other species through interspeciifc territorial behavior. In this manner, interspecific interactions may have been involved in non synchronous population fluctuations of M.ochrogaster and M.penns— ylvanicus within a site.When populations of both species eventually achieved high densities in a site,however, there was no evidence for interspecific interference in the demography of either species where food resources were relatively high(bluegrass). Where food resources (tlalgrass)were limited,presence of M.pennsylvanicus appeared to suppress population density of M.ochro- gaster.The only variable significantly reduced,however, was immigration, not a primary factor involved in population demography of this species(Getz et a1., 2005a).Lin and Batzli(2001)also less immirgation of M.ochrogaster into sites with dense cover when M.pennsylvanicus was present. Our results agreed in general with those of Huitu et a1.(2004),Eccard and Y1fnen(2002,2003a,b), and Johannesen(2003)that interspeciifc interactions between coexisting species of arvicoline rodents are seldom sufifcient to negatively affect population demography of the species.Huitu et a1.(2004)suggested predation effects may keep densities of coex ̄stmg competnig species sufifciently low that there iS no negative interaction effect. Eccard and Y16nen (2002) concluded that direct interference,as in territorial defense,may negatively fafect demography and result in little indirect negative interference,i.e.,on survival and reproduction.That there seldom were concurrent population fluctuations of M.ochrogaster and M.pennsylvanicus,although they coexisted in the same habitats, may reflect similar responses.Typically,populations of the two species did not coexist in sufficient numbers to elicit interspecific 维普资讯 http://www.cqvip.com 维普资讯 http://www.cqvip.com 5期 Lowell L.GETZ et a1.:Interspeciifc competition in vole populations 811 Krebs CJ,Keller BL,Tamarin RH,1969.Microtus population demography: demographic changes in fluctuating populations of Microtus ochrogaster nd M.pennsaylvanicus in southern Indina.Ecolaogy 50:587—607. Seber GAF,1965.A note on multiple—recapture census.Biometrika 52: 249—259. SPSS,Ine,2oo1.SPSS l0.0.7 for Maeintosh.Chicago,IL. Krebs CJ,Myers JH.1974.Population cycles in small manunals.Adv. Eco1.Res 8:267—399 raitt MJ,Krebs CJ,1985.Population dynamics nd cycles.Ian:Tamarin RH ed Biology of New W(1rld Microtus.Spec Publ N0.8.Am.Soc. ManuTl1.535—566.a Krebs CJ,1977.Competition between Microtus pennsylvanicus and M. ochrogaster.Amer.Mid1.Nat.97:42—49. Lebreton JD,Burnham P,Clobert J,Anderson DR,1992.Modeling survival nd taesting biological hypotheses using marked animals——a uniifed approach with easE,一studies.Eco1.Monogr.62:67一ll8. Lin YK.Batzli GO.2oo1.The effect of interspeeiife competition on habitat Tamarin RH,1984.Body nlass as a criterion of dispersal:a critique.J. ManllTll,65:691—692.a Tazik DJ,Getz LL,2007.1nterspeciife territoriliaty in Microtus ochrogaster and M.pennsylvan/cus.Praiire Nat.39:41—48. Thompson DQ,1965.Food preferences of the meadow vole Microtus pennsylvanicus in relation to habitat affinities.Amer.Mid1.Nat.74: selection by voles:an experimentl approach.Caan.J.Zoo1.79: l10—120. Lindroth RL.Batzlj GO.1984.Food habits of the adow vole Microtus pennsylvanicus in bluegrass and praiire habitats.J.Mamma1.65: 60o一606. MarcstrSm V,Hoglund N,Krebs CJ,1990.Peridoic fluctuations in small mammals at Boda.Sweden from 1961—1988.J.An.Eco1.59:753— 761. Morris DW.1996.Coexistence of specilaist and generlaist rdoents via habitat eslection.Ecology 77:2 352—2 364. Norrdabl K,Korpimaki E,2002.Changes in individual quality during a 3一 year population cycle of voles.Oecologia 130:239—249. Pearson OP,1985.Predation.In:Tamarin RH ed.Biology of New World Microtus.Spec.Pub1..Amer.Soc.Mamma1.8:535—566. Pimm SL,Rosenzwig ML,1981.Competitors and habitat use.Oikos 37: l一6. 76—86. Verner L,Getz LL,1985.Signiifcance of dispersal in fluctuating populations of Microtus ochrogaster and M.pennsylvanicus.J.Manlma1.66:338— 347. White GC,Burnham KP,1999.Program MARK:survival estimation from populations of marked animals.Bird Study 46:120一l39. Willi肿s BK,Nichols JD,Conroy MJ,2001.Analysis and management of animal populations.San Diego:Academic Press. Yl ̄nen H.1994.Vole cycles and antipredatory behaviour.Trends Eco1. Evo1.9:426—430. Zar JH.1999.Biostatistical Analysis.4th edn.Upper Saddle River,NJ: Prentice Hal1. Zimmerman EG,1965.A compariosn of habitat and food species of two species of Microtus.J.ManllTl1a.46:6o5—612.
因篇幅问题不能全部显示,请点此查看更多更全内容