Trinity University Digital Commons @ Trinity

Biology Faculty Research Biology Department

4-1997 Social Organization of Neotoma micropus, the Southern Plains Woodrat Sarah A. Conditt Trinity University

David O. Ribble Trinity University, [email protected]

Follow this and additional works at: https://digitalcommons.trinity.edu/bio_faculty Part of the Biology Commons

Repository Citation Conditt, S.A., & Ribble, D.O. (1997). Social organization of Neotoma micropus, the southern plains woodrat. American Midland Naturalist, 137(2), 290-297. doi:10.2307/2426848

This Article is brought to you for free and open access by the Biology Department at Digital Commons @ Trinity. It has been accepted for inclusion in Biology Faculty Research by an authorized administrator of Digital Commons @ Trinity. For more information, please contact [email protected]. Am. Midl. Nat. 137:290-297

Social Organizationof Neotoma micropus,the Southern Plains Woodrat

SARAH A. CONDITT AND DAVID 0. RIBBLE Departmentof Biology, Trinity University, 715 StadiumDrive, San Antonio, 78212

ABSTRACT.-This studydescribed the social organizationof Neotomamicropus, the southern plains woodrat,using radiotelemetry.We studied woodratsfrom September 1994 through April 1995 at the Urban WildernessReserve in S Bexar County,Texas. Sixteen individuals (threeadult males, 11 adult females,and twosubadult males) were radiocollaredand located in the eveningor earlymorning duringJanuary and February.Seventy-four percent of female radiolocationsand 57% of male radiolocationswere fromtheir respective nests. No more than one adult individualwas observed at any nest at the same time.Males had significantly larger(x = 1899 m2) home rangesthan females(x = 220 m2). Male home rangesoverlapped femalehome rangesmore (35.9%) thanfemale ranges overlapped other female home ranges (11.9%). Survivalrates were similarfor males and females. The resultsconfirm previous studies indicatingthat these woodratsare asocial. The spacing of male and female home ranges indicated that these woodratsprobably have a promiscuousmating system.

INTRODUCTION Neotomamicropus is a medium-sizedwoodrat that inhabitsthe southernGreat Plains of the S-centraland southwesternUnited States (Texas, and Oklahoma) and northeasternMexico (Hall, 1981). It is usuallydescribed as solitaryand asocial (Raun, 1966; Wiley,1984; Braun, 1989), meaning thatindividuals show little social behaviorbeyond mat- ing. Home range estimatesfrom live-trapping studies have varied from158 to 1335 m2 for individualsof thisspecies (Johnson,1952; Raun, 1966). However,live-trapping data do not provide a sense of how oftenor how far an travels.Woodrats are normallytrapped near their nests (Cranford,1977) and any evaluationsof home range use and social be- haviorbased onlyon trappingdata are suspect.Radiotelemetry can providea more accurate picture of home range use and social organization(Madison, 1978, 1980). We wished to corroboratethe observationsfrom live-trappingstudies that N. micropusis asocial using radiotelemetry.Specifically, we wanted to ascertainif adults spent any timewith adult con- specificsat theirnest. We also desiredto describethe home rangesand social organizationof individuals,which providesgood evidence for the particularmating system (Madison, 1980). The ensemble of behaviorsand physicaladaptations specific to mating,as well as some of the social con- sequences of these behaviors,describe a population's mating system(Vehrencamp and Bradbury,1984). For mammalianmating systems there is an asymmetryin the interestsof males and femalesbecause female gestationand lactationemancipate males fromcare of the youngwhile imposing additional burdens on the female (Eisenberg,1981; Ostfeld,1985; Barlow, 1988). At one extreme,resources (especially food) may be of high quality and clumped in space or time. In this case, females are predicted to aggregate around the resource to maximize their reproductivesuccess. This distributionsets the stage for mo- nopolization by males; if the aggregationof femalesis defensible,there existsa high po- tentialfor polygyny (Emlen and Oring, 1977). In polygynousspecies, males typicallydefend territoriesduring the breeding season thatcontain home ranges of one to severalfemales (Heske and Ostfeld,1990). Females may,or maynot, show territorialbehavior. A different typeof matingsystem, promiscuity, results when males compete for and convergeon re- 290

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 1997 CONDITT & RIBBLE: WOODRAT SOCIAL ORGANIZATION 291 productivefemales. These adult femalesoccupy exclusivehome ranges duringmost of the breeding season and males have larger,extensively overlapping home ranges. Although subtle,the distinctionbetween polygyny and promiscuityis importantbecause the expected variance in male reproductivesuccess is greaterfor polygynousmating systems (Clutton- Brock, 1988). In this studywe hoped to distinguishbetween polygynyand promiscuitybased on the spatial relationshipsbetween males and females.If our studypopulation is polygynouswe expected female home ranges to be overlapped by only one territorialmale. Males would have home ranges thatdo not overlap other male ranges. In a promiscuousmating system we w'ould expect female home ranges to be nonoverlappingand their ranges would be overlappedby severalmales withlarger home ranges. Males would have home ranges that overlap other male ranges.

METHODS We conducted thisstudy at the Urban WildernessReserve, located at the intersectionof Applewhite andJett roads in south San Antonio, Bexar County, Texas (29?14'45", 98?33'10"). The habitatof the 10-ha studysite was dominatedby honey mesquite-brush(Prosopis glan- dulosa) and pricklypear cactus (Opuntia spp.) on the uplands,with riparian forest lowlands borderingthe Medina River.Neotoma micropus inhabited only the honey mesquiteuplands of the studysite, which were approximately4 ha. No woodrat nests were found in the riparianforest lowlands. The propertywas bordered to the S and W by cultivatedfields, and to the N and E by paved roads. This area of Texas has a subtropicalclimate with hot, humid summers (x high temp. = 29.3 C) and cool winters(x low temp. = 10.4 C) with freezingtemperatures occurring on average 20 days per year.The average annual rainfall is 70.0 cm (Tandy,1987). We trappedNeotoma micropus from late September1994 throughApril 1995. Tomahawk and Sherman trapswere placed in and parallel to woodratruns within3 m of nests.Traps baited withpeanut butterand oats or rat chowwere set in late afternoonand checked after sunrisethe next morning.Generally, one or twotraps were set at each of 10-15 activenests once a week. Activenests were identifiedby the presence of freshscats on runwaysand freshlychewed cactus pads. Periodically,two Sherman trapswould also be placed at nests suspectedto house juveniles. Upon capture,woodrats were transferredto a nylonmesh bag and weighedto the nearestgram with a Pesola springbalance. Sex, breedingcondition and age were noted. Animalswere classifiedin age classes based on weightand breeding con- dition.Females thatwere pregnant,lactating or parous, and males thathad enlarged testes were considered reproductivewoodrats. All reproductiveanimals were classifiedas adults. Animalsweighing less than 150 g were consideredjuvenile. Nonreproductiveanimals weigh- ing more than 150 g were categorizedas subadults.Woodrats were examined for external parasites,wounds and pelage condition,and ear-taggedwith small aluminum numbered tags in both ears. Animalswere released at the point of capture; the directionof retreat was noted. Radiotelemetrywas used to determinehome ranges of adult and subadult Neotomami- cropus.Upon capture,"mouse-style" single stage radiotransmitters (AVM Instruments, Inc.) were fittedto woodratsunder Metofane anesthesia.The collar assemblywas encapsulated withacrylic to hold the package together,waterproof it, prevent the antenna frompulling out, and preventdamage by the studyanimal. This whole transmitterpackage, mounted on plastic cable ties, was fastened snuglyaround the woodrat's neck. Radiotransmitters weighed 3.8 g on average (range 3.5-4.1 g) whichwas, on average,approximately 1.3% of the mass of the woodratto which theywere attached.Woodrats were fittedwith radiotrans-

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 292 THE AMERICAN MIDLAND NATURALIST 137(2) mittersin the field and were released upon recoveryfrom anesthesia (approximately30 min). Afterat least 24 h of acclimation,radiocollared woodrats were located witha Telonics TR-4 receiverand a 3-elementYagi antenna, or an AVM LA12-DSE receiverand 2-element antenna. To accuratelydetermine radiolocations of woodrats,we constructeda map of the study area thatincluded the positionsof all woodratnests, specific landmarks (e.g., fence posts, large trees), and several semipermanentreference points marked with metal spikes. The locationsof woodratnests and otherfeatures were mapped using a Leitz digitaltotal station (combinationrange finderand theodolite).The digitaltotal station converted distance and angular coordinate data into Cartesiancoordinates which were mapped using the Surface III ( Geological Survey) computer program.The locations of some woodrat nests relativeto surveyedpoints were determinedwith a tape measure and magneticcompass. We trackedradiocollared woodrats 2 or 3 nightsper week duringJanuary and February 1995. Woodratswere trackedfor either 3-4 h beginning 1 h aftersunset or 2-3 h before sunrise.We located radiocollaredwoodrats by walkingpre-existing trails included on our map of the studyarea and scanningfor the radio frequenciesof local woodrats.The esti- mated location was noted on the map of the studyarea, along withthe time and activity level (restingor active). Woodratswere considered restingif theirsignal was uniformand constant,or activeif theirradio signal was variablein intensity.We estimatedthe distance to radiotransmitterswithin a 20-mradius based on the strengthof the signal.Our estimates of signalstrength and distancewere testedperiodically with the location ofradiotransmitters placed at known distancesfrom the receivers.Focal samples were collected on selected individualsfor periods of time rangingfrom 1-3 h. All sequential radiolocationswere sep- arated by a minimumof 15 min. The area of a minimumconvex polygon (Mohr, 1947) containingall radiolocationswas calculated for each radiocollared Neotomamicropus. We chose this method because of its widespread use and insensitivityto dependence on successiveobservations (Swihart and Slade, 1987). The area withineach polygonand the areas of overlapwere calculated with the Calhome computerprogram (J. Kie,J. Baldwin and C. Evans,U.S. ForestService, Pacific SouthwestResearch Station,Fresno, California). After collecting data forhome ranges,we continued to monitorthe locations of woodratsonce per week by eitherradiotelemetry or trapping.At the conclusion of the studyin late April 1995, radiocollared Neotomawere presumeddead ifwe could not trapthem and no radio signalwas detected.Assuming these "lost" woodratswere dead, we calculated survivalrates using the softwareand methodsof Heisey and Fuller (1985). Startingdate for each animal was the date when we firsttagged the woodratwith eartags. Ending date for a survivingwoodrat was the day we removedthe radiocollar.The ending date for a woodrat presumed dead was taken as the midpoint betweenthe last day a signalwas detectedand the firstday a signalwas not found and the animal was not subsequentlytrapped. We accepted statisticalsignificance at P ' 0.05. Means are presented? 2 SE throughout, except where indicated.Proportions were arcsin-transformedbefore analyses.

RESULTS During the 6 mo of thisstudy, we set trapsat 31 nests.Thirty-six individuals were trapped and tagged from28 nests;90% of nestswere activeat some time duringthe study.Tagged woodratswere captured 98 timesin 346 trap nights (trap success rate of 28.3%). Ten of the tagged woodratswere captured only once. We trapped an area of 2.4 ha during Sep- temberthrough December. Beginningin January,we expanded the trappingarea 67% to include 3.6 ha and 12 additional activenests.

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 1997 CONDITT & RIBBLE: WOODRAT SOC8L0 ORGANIZATION 293 6

0)

Get Nov Dec Jan Feb Mar Date (1994-1995) FIG. 1.-Populationdensity (/ha) ofwoodrats during the study. The area trappedfrom September throughDecember was 2.4 ha; thearea trappedfrom January through March was 3.6 ha

More adult females than adult males were present in the population at all times.The average ratio of adult males to femaleswas 1:1.6. Population densitywas calculated from the minimumnumber known alive dividedby the area (4 ha) of honeymesquite woodland (Fig. 1). The lowestdensity was 2.0 woodratsper ha in the 1st wk of October.The population densitypeaked at 5.5 woodratsper ha in February.Female breeding activitywas observed throughoutthe study.The proportionof femaleslactating or pregnantpeaked at 50% in October. We found no evidence of lactatingfemales through most of November.Lactating or pregnantfemales were found again sporadicallybetween December and March,but did not exceed 45% of the adult females. We radiocollared 14 adults (three males and 11 females) and twosubadults (both males; Fig. 2). On average,33.5 locations (range 7-51) were collected fromeach female and 36.6 locations (range 26-43) from each male. There was no significantcorrelation between number of radiolocationsand home range area (Spearman's r, = -0.07, P > 0.8). The area of the cultivatedfield was subtractedfrom the home range calculationfor male 2605 since we never observed him crossingthis field.Adult male home ranges (k = 1899.3 ?- 884.0 in2; median = 1696 in2) were significantlylarger than femalehome ranges (k-= 220.9 -+ 98.1 in2 median = 188 in2; Mann-WAhitneyU-test P = 0.01). Subadult male ranges (k 385.8 ?704.5 in2) were not significantlydifferent from females (P = 0.84). Overlap of female home ranges by males was greater (35.9 ? 15.0%) than overlap by otherfemales (11.9 ?- 10.9%; paired t-teston arcsin-transformedpercentages, t = 3.11, d.f. = 10, P < 0.02). The percentof male home ranges overlappedby females (14.6 ? 16.2%)

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 294 THE AMERICAN MIDLAND NATURALIST 137(2)

/ f254254

597 ~~~~N

\ \ SO~~~~~~~~~~~~5meters

f259 __ s\

V Openfield

Boundaryof study area FIG. 2.-Minimumconvex polygons of Neotomamicropus radiotelemetry locations for 14 adultand twosubadult during January and Februaryof 1995.Woodrats are identified by sex (m = male, f = female)and t;4gidentification number. The heavyline marks the boundaries of the honey mesquite uplandsof thestudy area was similarto that overlapped by other males (16.0 ? 17.2%). Mature male home ranges overlapped an average of three female home ranges. The majorityof both male and female radiolocations were made at their respective nests (male: 56.5 ? 25.0%; female: 74.3 ? 10.7%). Individuals did use multiple nests during thisstudy. Four femalesoccupied twodifferent nests. Two radiocollared males had a single nest to which they returned during the day. The male with the largest home range used three differentnests as sheltersites during the study.The two subadult males only visited a single female nest. These radiocollared woodrats had nest sites close (12 and 30 m) to a large female nest. The subadult male closest to the female spent most of his time in the female nest (78.8% of radiolocations). The other subadult visited the female nest only twice. By the conclusion of this studyin late April,seven (two males, fivefemales) of the 16 radiocollaredNeotoma micropus were lost and presumablydead. The 30-daysurvival rates were approximatelythe same for males (0.83, 95% C.I. 0.64-1.00) and females (0.86, 95% C.I. 0.75-0.98). The pooled 30-daysurvival rate forboth males and femaleswas 0.85 (95% C.I. 0.75-0.96) and the pooled yearlysurvival rate was 0.14 (95% C.I. 0.03-0.60).

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 1997 CONDITT & RIBBLE: WOODRAT SOCIAL ORGANIZATION 295

DISCUSSION This is the firststudy to examine the social organizationof Neotomamicropus using ra- diotelemetry.The majorityof our radiolocationsfor both male and femalewoodrats were made while individualswere at their nests and we never observed more than one adult woodrat at a nest at the same time. These observationssupport previousclaims that this species is relativelyasocial Uohnson, 1952; Raun, 1966). We did observetwo subadult males that occupied a nest with a mature female. It is possible that these youngermales were offspringof the adult female. Our resultsalso indicated that male home ranges were sig- nificantlylarger than female home ranges and femalehome rangeswere significantlyover- lapped by male home ranges. These observationson spacing and the female tendencyto have more exclusivehome ranges than males (Fig. 2) led us to conclude thatin thispop- ulation, N. micropustend to have a promiscuousmating system. However, it is likelythat the matingsystem of this species can range frompolygyny (as seen by male 2597; Fig. 2) to promiscuity(males 2605 and 2584; Fig. 2). The distributionand dispersionof females probably dictates the options available to males in these woodrat populations (Ostfeld, 1985). The home range sizes in this studyare larger than those in most previous studies of Neotomamicropus. Raun (1966) determinedhome rangesof N. micropusfrom trapping data in southwesternTexas and concluded thatfemale home rangeswere similarto the area of cactus patches used. The differencebetween male (232 m2) and female (158 m2) home rangeswas not significantlydifferent.Johnson (1952) reportedthat males had smallerhome ranges than females (971 m2 and 1335 m2, respectively).The average adult male home range size calculated in this study(1899 m2) is much largerthan that calculated by either Johnson (1952) or Raun (1966). Much of this differencemay be attributedto different operational definitionsof home range used in each study.Raun (1966) calculated home ranges using minimumconvex polygons,but he excluded points that were far fromthe centerof activityand were assumed to represent"occasional salliesoutside the area, perhaps exploratoryin nature." Raun described the center of activityas the nest and associated cactus patch. This method yieldsacceptable resultswith females because theydo not seem to travelfar fromtheir nests, but is inaccuratefor male home ranges because theirmove- ment is associated not onlywith feeding in cactus patches,but withbreeding activities. We avoided these operationalproblems by defininghome range as all the area traversedby the individualin its normal activitiesof food gathering,mating and caring for young (Burt, 1943). In this study,points thatRaun (1966) considered abnormal,occasional sallieswere shown to occur frequentlywith the use of radiotelemetry. The population densitydetermined during this study(1.6 to 5.8 woodratsper ha) was lower than thatfound in other studies.The largerhome range sizes observedin thisstudy may be due to the lower population density.High den densitieshave been recorded in mesquite-cactushabitat (74.1 dens/ha) and pricklypear-short grass habitat (30.6 dens/ha) (Box, 1959). In one study(Raun, 1966), woodrat densityincreased from14.8 to 31.1/ha in 8 mo, remained stable at 24.7/ha for 10 mo, decreased to 9.1/ha 6 mo later and 0/ha 6 mo later. The differencein the densitybetween this studyand others may be due to census methods or habitatdifferences. Even if everysingle woodrat nest we observedwas occupied, the densitywould not exceed 9/ha. Raun (1966) found a significantpositive correlationbetween the densityof cactus and the population densityof woodrats,although he concluded thatcactus is not absolutelynecessary to supportwoodrats. This observation was supportedat our studysite. No Neotomamicropus were found in the 6 ha of cactus-free

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 296 THE AMERICAN MIDLAND NATURALIST 137(2) riparianwoodland. The lower population densityin thisstudy may be due to lower cactus densityand the lack of appropriateshelter sites. Neotomamicropus is a member of the floridanaspecies group thatalso containsN. flori- dana and N. albigula (Birney,1973). The social organizationof N. micropusis comparable to at least one of the other species in thisgroup. Neotomafloridana is usuallydescribed as solitaryand asocial (Kinsey,1976, 1977). The den is the base of operationsand home ranges are small (Fitchand Rainey,1956). Tate (1970) found thatthe mean home range foradult N. floridanawas 662 M2. Adult males had largerhome ranges than adult females (Goertz, 1970; Tate, 1970). Less is known about N. albigula,although population studies have in- dicated thatfemales survive longer than males (Boggs, 1974; Brownand Zeng, 1989), sug- gestingthat the matingsystem is polygynous.There appears to be variationin social and matingsystems both withinand among these closelyrelated woodrats. These species deserve furtherstudy in order to understandthe effectsof ecology and phylogenyon theirsocial and matingsystems.

Acknowledgments.-WethankJohn Sanders forinvaluable help in the field,Philip Ross forproviding access and use of the Urban Wilderness,Dr. Glenn Kroeger of the TrinityUniversity Geosciences Departmentfor his assistancesurveying the site, Eric M. Andersonfor providing us withthe program to compute mortalityrates, and the TrinityUniversity Department of Biologyfor logistic and monetary support.We also thanktwo anonymousreviewers for theircomments and suggestions.

LITERATURECITED BARLow,G. W. 1988. Monogamyin relationto resources,p. 55-77. In: C. N. Slobodchikoff(ed.). The ecology of social behavior.Academic Press,San Diego. BIRNEY,E. C. 1973. Systematicsof three species of woodrats (Genus Neotoma) in central North America. Univ. Kans. Misc. Publ. Mus. Nat. Hist., 58:1-173. BOGGS,J.R. 1974. Social ecologyof the white-throatedwoodrat (Neotoma albigula) in Arizona.Unpubl. Ph.D. Thesis,Arizona State Univ.Tempe. 118p. Box, T. W. 1959. Densityof plains wood rat dens on fourplant communitiesin south Texas. Ecology, 40:715-716. BRAUN,J. K. 1989. Neotoma micropus.Mamm. Species,330:1-9. BROWN,J. H. ANDZ. ZENG. 1989. Comparativepopulation ecologyof eleven species of rodentsin the Chihuahaun Desert. Ecology,70:1507-1525. BURT,W. H. 1943. Territorialityand home range concepts as applied to .J. Mamm.,24:346- 352. CLuTTON-BROCK,T. H. 1988. Reproductive success, p. 472-485. In: T. H. Clutton-Brock(ed.). Reproductivesuccess. The Univ. Chicago Press,Chicago. CRANFORD,J. A. 1977. Home range and habitat utilizationby Neotomafuscipes as determinedby radiotelemetry.J .,58:165-172. EISENBERG,J. F. 1981. The mammalianradiations. The Univ. of Chicago Press,Chicago. 610 p. EMLEN,S. T. ANDL. W. ORING. 1977. Ecology,sexual selection,and the evolutionof matingsystems. Science,197:215-223. FITCH,H. S. ANDD. G. RAINEY.1956. Ecological observationson the woodrat,Neotomafloridana. Univ. Kans. Publ. Mus. Nat. His., 8:499-533. GOERTZ, J. W. 1970. An ecological studyof Neotomafloridanain Oklahoma.J. Mammal.,51:94-104. HALL,E. R. 1981. The mammalsof NorthAmerica, 2nd ed. John Wiley& Sons, New York. 1181 p. HEISEY,D. M. ANDT. K FULLER. 1985. Evaluationof survivaland cause-specificmortality rates using telemetrydata. J Wildl.Manage., 49:668-674. HESKE,E. J. ANDR. S. OSTFELD. 1990. Sexual dimorphismin size, relativesize of testes,and mating systemsin NorthAmerican voles: J. Mammal.,71:510-519. JOHNSON,C. W. 1952. The ecological life historyof the packrat,Neotoma micropus, in the brushlands of SouthwestTexas. Unpubl. M.S. Thesis, Univ.Texas, Austin.115 p.

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions 1997 CONDITT & RIBBLE: WOODRAT SOCIAL ORGANIZATION 297

KINSEY, K. P. 1976. Social behavior in confined populations of the Alleghenywoodrat, Neotoma floridana magister.Anim. Behav.,24:181-187. 1977. Agonisticbehavior and social organizationin a reproductivepopulation of Allegheny woodrats,Neotoma floridana magister. J. Mammal.,58:417-419. MADISON, D. M. 1978. Movementindicators of reproductiveevents among female meadow voles as revealed by radiotelemetry.J Mammal.,59:835-843. . 1980. Space use and social structurein meadow voles, Microtuspennsylvanicus. Behav. Ecol. Sociobiol.,7:65-71. MOHR, C. 0. 1947. Table of equivalentpopulations of NorthAmerican mammals. Am. Midl. Nat., 37: 223-249. OSTFELD, R. S. 1985. Limitingresources and territorialityin microtinerodents. Am. Nat., 126:1-15. RAUN, G. G. 1966. A population of woodrats (Neotomamicropus) in southernTexas. Tex. Mem. Mus. Bull., 11:1-62. SWIHART, R. K. AND N. A. SLADE. 1987. A test for independence of movementsas shown by live trapping.Am. Midl. Nat., 117:203-207. TANDY, T. L. 1987. ApplewhiteDam and Reservoirand Leon Creek DiversionDam and Lake. Draft EnvironmentalImpact Statement.U.S. ArmyEng., FortWorth, Tex. TATE, W. H. 1970. Movementsof Neotomafloridana attwateriin Brazos County,Texas. Unpubl. M.S. Thesis, Texas A&M Univ.,College Station.39 p. VEHRENCAMP, S. L. AND J. W. BRADBURY. 1984. Matingsystems and ecology,p. 251-278. In: J. R. Krebs and N. B. Davies (eds.). Behavioralecology, 2nd ed. Sinauer AssociatesInc., Sunderland,Mass. WILEY, R. W. 1984. Reproductionin the southernplains woodrat (Neotomamicropus) in westernTexas. Spec. Publ. Mus. Tex. TechUniv., 22:137-164.

SUBMITTED 29 MARCH 1996 ACCEPTED 20 MAY 1996

This content downloaded on Tue, 26 Feb 2013 15:34:44 PM All use subject to JSTOR Terms and Conditions