Home Range, Spatial Overlap, and Burrow Use of the Desert Tortoise in the

West Mojave Desert

Meagan L. Harless1,2, Andrew D. Walde1, David K. Delaney3, Larry L. Pater3, andWilliam K. Hayes2

Understanding the space use patterns of a population may provide crucial information regarding land managementdecisions, such as delineation of protected areas. Herein we provide a comprehensive analysis of factors affecting thespace use of Desert Tortoises in the west Mojave Desert, including physical, social, and environmental variables. Ourobjectives were to determine how spatial overlap, as well as multiple parameters of burrow use, influence home rangesize of this species. Male tortoises were significantly larger than females and exhibited very different patterns of spaceand burrow use; however, body size did not affect these variables. Male home range (100% MCP) and core area (50%

MCP) size averaged 65% and 73% larger than those of females. Burrow use by males and females affected core area size,but not home range area. Females exhibited a strong male bias in overlap of both estimates of space use and burrowsharing, while using a significantly lower number of burrows per year. However, males overlapped and shared burrowswith a similar number of tortoises of either sex. In addition, a high degree of home range overlap between individualssuggests a lack of territoriality in this population, although this warrants further analysis. These results suggest thatsocial factors may not be the primary determinants of space use in this Desert Tortoise population.

BURT (1943:351) introduced the foundation for theconcept of home range as ‘‘that area traversed by theindividual in its normal activities such as food-

gathering, mating, and caring for young.’’ This definitionis heavily used in ecological studies across species to gaininsight into the biological needs of an organism (Kernohanet al., 2001). Through knowledge of the location of resourceswithin a home range, an animal may enhance its fitness byaccessing limiting resources and avoiding or escapingpredation (Peters, 1978). Since Burt proposed the originalconcept of home range, ecologists have been interested indetermining the factors that predict home range size(Powell, 2000).

A number of variables may influence home range size.Sex, age, activity pattern, habitat quality, season, andweather can all influence home range area (Stickel, 1968;Mysterud et al., 2001), as can population density (Alberts,1993). Among reptiles, body size (Perry and Garland, 2002),sex (Goodman et al., 2005; Roth, 2005), and reproductivestatus (Graves and Duvall, 1993; Litzgus and Mousseau,2004; Waldron et al., 2006) are important influences indetermining home range area. Recent studies suggest that,aside from physical and phylogenetic factors, animal spaceuse may be best understood through social factors, namelyinteractions between neighbors (Doncaster, 1990; Sih andMateo, 2001; Stamps and Krishnan, 2001). Home rangeoverlap may be used to assess the degree of interactionamong individuals within a population or to infer territo-riality when behaviors cannot be observed adequately(Powell, 2000). Furthermore, a more complete understand-ing of land use of a species can be gained by investigatingland use at different scales (i.e., hierarchical orders of habitatselection; Johnson, 1980). Analyses should also consider theportion of the area most important to the animal, such asthe core area. Limiting resources are often patchy in a givenenvironment, leading animals to focus land use around coreareas containing precious resources (Kaufmann, 1962;

Samuel et al., 1985; Powell, 2000). Thus, interactionsbetween individuals may be most important within coreareas, and undoubtedly influence the spacing patterns ofindividuals (Borger et al., 2006).

The Desert Tortoise (Gopherus agassizii) represents asuitable model for exploring the spacing patterns ofindividuals. Previously published home range sizes for G.agassizii are wide ranging (Harless, 2007) and suggest moreresearch is needed to understand space use patterns ofDesert Tortoises, especially for application to managementof this threatened species. The home range of G. agassizii isoften described as a network of burrows separated by travelcorridors of varying dimensions (O’Connor et al., 1994;Duda et al., 1999, 2002). Desert Tortoises depend heavily onburrows to meet various life history requirements, includingthermoregulation, predator avoidance, and reproductiveopportunities (Burge, 1977; Bulova, 1994; Duda et al.,1999). Burrows are often the only areas where males andfemales are located together, as males generally seek outfemales that remain within the burrow (Woodbury andHardy, 1948; Berry, 1986). Burrows are known to be alimiting factor in the movements and space use of relatedspecies, the Gopher Tortoise (Gopherus polyphemus: McRae etal., 1981) and the Bolson Tortoise (Gopherus flavomarginatus:Aguirre et al., 1984). Thus, the location of burrows within ahome range likely influences the spatial distribution ofindividual tortoises.

A comprehensive model of the social structure of DesertTortoises remains unclear, although a male-based domi-nance hierarchy is often cited from observations of antag-onistic encounters and territorial displays (Berry, 1986;Duda et al., 2002). In freshwater turtles, social behaviorhas been inferred from observations of home range overlap,agonistic encounters, stability of home ranges betweenyears, and movement patterns (Galbraith et al., 1987;Kaufmann, 1992). Few studies have separately addressedthese aspects of space use in a social context for G. agassizii,

1 ITS Corporation, 8000 San Gregorio Road, Atascadero, California 93422; E-mail: (ADW) awalde@hotmail.com. Send reprint requests toADW.

2 Department of Earth and Biological Sciences, Loma Linda University, Loma Linda, California 92354.3 U.S. Army Construction and Engineering Research Laboratory, P.O. Box 9005, Champaign, Illinois 61826.Submitted: 5 October 2007. Accepted: 30 December 2008. Associate Editor: J. D. Litzgus.F 2009 by the American Society of Ichthyologists and Herpetologists DOI: 10.1643/CE-07-226

Copeia 2009, No. 2, 378–389

leaving a holistic understanding of the social structure ofthis species undefined. To our knowledge, only one previousstudy quantified yearly overlap of home ranges of DesertTortoises (Holt and Rautenstrauch, 1996), although twostudies analyzed overlap in seasonal minimum convexpolygon (MCP) areas (Burge, 1977; O’Connor et al., 1994).The relationship between burrow use and space overlap hasnot been examined. Our objective was to evaluate therelationships between home range area, core area, spatialoverlap, and burrow use of Desert Tortoises in the westMojave Desert.

MATERIALS AND METHODS

Study site.—The study was conducted in the southwestcorner of the National Training Center at Fort Irwin,California (35u149630N, 116u759170W) on land federallydesignated as Desert Tortoise critical habitat (U.S. Fish andWildlife Service, 1994). The study site was irregularlyshaped, approximately 3.74 km2, and comprised typicalMojave Desert vegetation: Creosote–White Bursage scrub(Larrea tridentate–Ambrosia dumosa; Turner, 1982). This areaof the military base is protected from public use and is thusrelatively undisturbed, being approximately 8.5 km fromthe nearest paved road (Walde et al., 2007). Both years of thestudy, 2004 and 2005, experienced record high rainfall forthe west Mojave Desert (National Oceanic and AtmosphericAdministration, 2006).

Radio-tracking.—The study was initiated in 2003, when adultDesert Tortoises were captured during extensive surveyswithin the study-site boundary. Adult age was indicated by amidline carapace length (MCL) greater than 180 mm. Low-duty, cycle-pulsed transmitters (Advanced Telemetry Sys-tems, Isanti, Minnesota; Holohil Systems Ltd., Carp, On-tario, Canada) were mounted directly onto the carapace ofeach tortoise using epoxy, as is commonly done for thisspecies (Boarman et al., 1998). Transmitters weighed lessthan 5% of the tortoise’s mass and had an expected batterylife of 18–36 months. Transmitter mounting varied by sex toensure there was no hindrance to tortoise locomotion,particularly during copulation attempts.

Portable radio receivers (Yaesu VR-500, Vertex StandardInc., Cypress, CA) and three-element hand-held directionalYagi antennas (Wildlife Materials International, Inc., Mur-physboro, IL) were used to locate transmittered tortoises.The geographic location (universal transverse mercator,UTM), the assigned unique burrow number, and theidentities of any nearby conspecifics were recorded duringeach tracking event. Geographic locations were recordedusing a global positioning unit (Garmin 12 PersonalNavigator Unit, Garmin International, Olathe, KS) with aprobable error between 3–6 m. Identical UTMs were used forlocations collected at the same burrow number to limitpossible error in home range and overlap calculations.Geographic locations were imported into ArcView 3.3(Environmental Services Research Institute, Redlands, CA)for home range calculations and spatial analyses.

Transmittered animals were located using a systematicsampling scheme from January 2004–December 2005.During the active season (March–June, August–October),locations were collected twice per week, with locationscollected once per week during less active periods (February,July, and November). During the winter period of decreased

activity (December–January), locations were limited to twiceper month.

Home range, core area, and overlap calculations.—The MCP isthought to be ideal for home range sizes of the DesertTortoise (O’Connor et al., 1994; Rautenstrauch and Holt,1995), despite implications of serial autocorrelation andproblems in sampling methodology (White and Garrott,1990). This method of home range estimation is suited tospecies such as G. agassizii because they are relatively slowmoving, generally do not travel long-distances, and use thesame areas, such as burrows, many times during a season(Row and Blouin-Demers, 2006).

We calculated both 100% (home range) and 50% (corearea) MCP areas for each tortoise during 2004 and 2005using the Animal Movement Extension for ArcView (AME;Hooge et al., 1999). We designated the core area as the 50%

MCP resulting from the algorhythmic removal of half of thelocations furthest from the geographic center (Samuel et al.,1985; White and Garrott, 1990). We used MCP instead ofkernel estimates for home range overlap assessmentsbecause kernel estimates are known to buffer actual datapoints, inherently inflating overlap estimates (Millspaugh etal., 2006). Moreover, multiple centers of activity are oftencombined by the fixed kernel approach, thereby removingarea between the centers of activity and further obscuringareas of overlap (Kernohan et al., 2001).

We calculated overlap in MCP areas using AME inconjunction with the XTools application extension forArcView (DeLaune, 2003). We used two measures tocalculate overlap between neighboring tortoises. First, wecounted the number of overlapping males and females thatshared a portion of an individual’s home range or core area.Individual tortoises were counted as overlapping in homerange where any portion of each MCP covered the samearea. Second, using the polygon shapes of the home rangeand core areas separately, the percentage of MCP overlapbetween two given individuals was calculated. This wasdone with XTools by overlaying the two polygons anddetermining the percentage of an individual’s home rangethat was shared with another tortoise (Kernohan et al., 2001;Bernstein et al., 2006).

We also quantified site fidelity, the percentage of overlapbetween years of an individual’s MCP area, using the AMEand XTools. Each polygon area for the 100% and 50% MCPwas overlaid to determine the shared area of overlapbetween years. We report site fidelity here as the proportionof the 2004 MCP that was contained in the MCP from 2005;in other words, overlap was relative to the 2004 MCP area.

Burrow use.—We examined numerous aspects of burrow useby tortoises to understand its relationship to home rangesize and spatial overlap. We quantified burrow use for eachtortoise at five levels, including: 1) number of burrows usedeach year; 2) burrow fidelity, i.e., the number of burrowsused during both years of study; 3) number of uniqueburrows, i.e., burrows used only by a single individual; 4)number of shared burrows, defined as non-simultaneous useof the same burrow by more than one individual; and 5)frequency of cohabitation, defined as two or more tortoisessimultaneously occupying the same burrow.

Data analyses.—For statistical analyses, we relied primarilyon general linear models, specifically mixed model analyses

Harless et al.—Desert tortoise home range analyses 379

of variance (ANOVA) and analyses of covariance (ANCOVA).Parametric assumptions were met in most cases, althoughsome variables had to be log10- or rank-transformed. Onedependent variable, log10-transformed core area, failed tomeet parametric assumptions; however, we decided toreport the ANCOVA results because supplemental non-parametric tests, with alpha levels adjusted for multipletests, confirmed that results from the general linear modelsinvolving this variable were robust.

For all ANOVA and ANCOVA models, we began with anomnibus model that incorporated all relevant variables,including body size as a covariate. Because body sizeconsistently showed no correspondence with home rangesize, overlap, or burrow use, we omitted it from all finalmodels. We calculated the effect size of independentvariables and interactions using partial g2 values (Mertlerand Vannatta, 2004). When the partial g2 values for maineffects and interactions summed to .1.0, we adjusted theseby dividing each partial g2 by the sum of all partial g2 values.In addition to the general linear models, we used Pearsoncorrelation analyses when parametric assumptions weremet, and either independent t-tests or nonparametricMann-Whitney U-tests for pairwise comparisons. We per-formed all statistical analyses using SPSS v14.0 (2006,Statistical Package for the Social Sciences, Inc., Chicago,IL). All means are reported as mean 6 1 SE.

RESULTS

Radio-tracking.—Initially, we attached radio transmitters to26 male and 15 female adult Desert Tortoises across thestudy site. We removed a total of seven adult tortoises fromthe following analyses for various reasons, includingmortality (n 5 3), long-term transmitter failure resulting ina lack of a sufficient number of locations as per the studyobjectives (n 5 2), and movements followed by home rangere-establishment to an area far outside of the designatedstudy site boundary (n 5 2). Thus, the resultant sample sizefor the following analyses is limited to 20 male and 14female Desert Tortoises (n 5 34). Males were significantlylarger than females based on an independent samples t-test(t 5 3.22, P 5 0.003; Table 1); however, body sizedemonstrated no effect in any of the statistical analyses(see below; Harless, 2007).

We conducted several analyses to evaluate possible biasassociated with the number of locations, year, and sex. A 2 3

2 mixed ANOVA using year (within-subjects) and sex(between-subjects) as independent variables revealed that

the mean number of locations per tortoise in 2004 (2,426total locations; 71.4 6 0.4 per tortoise) was significantly lessthan in 2005 (2,645 locations; 77.8 6 0.0 per tortoise; F1,32 5

231.98, P , 0.001, partial g2 5 0.88; Table 1); in the secondyear, we adhered more closely to the sampling regime. Thenumber of locations did not vary by sex (F1,32 5 0.50, P 5

0.48, partial g2 5 0.02), nor was there a significantinteraction between year and sex (F1,32 5 0.47, P 5 0.50,partial g2 5 0.02). Despite these differences, neither bodysize nor number of observations was significant whenincluded as covariates in ANCOVA models for home rangeand core area size.

Home range and core area size.—To determine possible biasin home range size associated with the number of locationsper year, we used a mixed ANCOVA model with home range(log10-transformed 2004 or 2005) as the dependent variable,with the number of locations (either 2004 or 2005) and sexas the independent variables. Differences between the sexeswere significant in each year (see above; 2004: F1,33 5 22.06,P , 0.001; 2005: F1,33 5 26.08 P , 0.001). The number oflocations did not significantly influence home range size ineither year (2004: F1,33 5 0.01, P 5 0.929; 2005: F1,33 5 0.07P 5 0.785). However, the interaction between sex and thenumber of locations was significant in each year (2004: F2,33

5 11.04, P 5 0.0002; 2005: F2,33 5 13.07 P 5 0.0002), withmales generally having a greater number of locationsbecause the sampling design was more rigorously followedfor them.

We used a 2 3 2 3 2 mixed ANOVA to identify thevariables influencing home range size. This model includedthe MCP size (log10-transformed) as the dependent variableand three independent variables: estimator (100% and 50%

MCP, within-subjects), year (within-subjects), and sex(between-subjects). Each of the three independent variableshad a significant effect on the MCP size (Table 2; Fig. 1).Both MCP areas for all tortoises were on average 20% and24% larger in 2005 than in 2004, respectively (F1,32 5 6.72, P5 0.014, adj. partial g2 5 0.10). Male home ranges and coreareas were on average 65% and 73% larger than those offemales, respectively (F1,32 5 35.67, P , 0.001, adj. partial g2

5 0.29). There were no interactions among these threevariables, indicating that both MCP areas were similarlyaffected by year and sex. Core areas represented a smallfraction of the total home range area. Male core areasrepresented 13.7% of the home range area in both years,while female core areas represented only 9.4% of the homerange in 2004 and 11.2% in 2005. Because of the importance

Table 1. Average Measurements (mean 6 1 SE, range) from Desert Tortoises (Gopherus agassizii; 20 Males, 14 Females) in the West Mojave Desert,Including Midline Carapace Length (MCL, mm), Number of Radio-telemetry Locations, and Burrow Use Statistics for 2004 and 2005.

2004 2005

Male Female All (range) Male Female All (range)

Body size (MCL) 256.1 6 4.3 235.4 6 4.6 247.5 6 3.6 (201–285) — — —Number of locations 71.6 6 0.6 71.0 6 0.5 71.4 6 0.4 (67–76) 77.8 6 0.1 77.8 6 0.2 77.8 6 0.0 (76–78)Number of burrows used

per year15.4 6 0.7 11.4 6 0.9 13.7 6 0.6 (5–22) 16.3 6 0.8 12.4 6 1.0 14.6 6 0.7 (7–24)

Number of burrows usedover both years

— — — 5.7 6 0.6 4.0 6 0.4 5.0 6 0.4 (1–10)

Number of uniqueburrows

11.0 6 0.9 6.8 6 1.0 9.3 6 0.8 (2–19) 11.0 6 0.8 7.6 6 0.7 9.7 6 0.7 (4–19)

380 Copeia 2009, No. 2

of year and sex to analyses of spatial use, we included thesevariables in the following analyses of overlap and burrow use.

Overlap between years (site fidelity).—We examined overlap ofindividual home ranges between 2004 and 2005 as ameasure of site fidelity using a 2 3 2 mixed ANOVA model.We used percent overlap between years as the dependentmeasure, with estimator (rank-transformed overlap of homerange and core areas, a within-subjects factor) and sex(between-subjects factor) as independent variables. Sitefidelity differed significantly for the two estimators (F1,32

5 11.71, P 5 0.002, partial g2 5 0.27; Table 2), with a two-fold or greater percentage of overlap in home rangescompared to core areas. There was no difference betweenthe sexes (F1,32 5 0.15, P 5 0.70, partial g2 5 0.01), nor aninteraction between estimator and sex (F1,32 5 1.75, P 5

0.20, partial g2 5 0.05).

Overlap of individual home ranges.—We used two ANCOVAmodels to examine factors associated with overlap ofindividual homes ranges between years. The first modeltreated number of overlapping individuals as the dependentvariable, whereas the second used percent of home rangeoverlap. Each model included three independent variables:year (within-subjects factor), sex of overlapping tortoise(within-subjects factor), and sex of individual (between-subjects factor). The two-year mean home range (log10-transformed) was used as the covariate. There was asignificant effect of year on the number of overlappingindividuals, with fewer overlapping individuals in 2004than 2005 (F1,31 5 6.70, P 5 0.015, partial g2 5 0.18;Table 2). Males and females overlapped with a similarnumber of individuals (F1,31 5 0.08, P 5 0.784, partial g2

5 0.00), whereas tortoises with larger home ranges over-lapped with significantly more individuals each year (F1,31 5

4.34, P 5 0.046, partial g2 5 0.12). The significantinteraction between year and home range (F1,31 5 11.91, P5 0.002, partial g2 5 0.28) indicated that home range had alarger effect on overlap in 2005. In addition, the significantinteraction between sex of tortoise and sex of overlappingindividuals (F1,31 5 4.40, P 5 0.044, partial g2 5 0.12)resulted from female home ranges overlapping with moremales than females, whereas males overlapped with a similarnumber of tortoises of either sex (Table 2).

The percentage of home range overlap between individ-uals (Fig. 2) was similar for both sexes (F1,31 5 0.66, P 5 0.42,

partial g2 5 0.02) and did not differ between years (F1,31 5

2.10, P 5 0.16, partial g2 5 0.06). However, percent overlapwas significantly affected by sex of overlapping tortoise(F1,31 5 7.07, P 5 0.012, partial g2 5 0.19), with tortoises ofboth sexes overlapping to a greater extent with malescompared to females (Fig. 2). Also, tortoises with largerhome ranges demonstrated a higher percentage of overlapwith adjacent tortoises (F1,31 5 8.69, P 5 0.006, partial g2 5

0.22). There were no significant interactions among thevariables in this model.

Overlap of individual core areas.—We used the same twoANCOVA models described above for home range overlap,except that the dependent variables applied to core areaoverlap and the two-year mean core area (log10-transformed)was the covariate. In the first model for number ofoverlapping individuals (Table 2), a significant interactionwas observed between sex of the individual and sex ofoverlapping tortoise (F1,31 5 8.48, P 5 0.007, partial g2 5

0.22). The core areas for males overlapped with a similarnumber of tortoises of either sex, whereas females had amale bias in core area overlap. No female core areasoverlapped in 2004, whereas the core areas of two femalesoverlapped in 2005. The number of overlapping individualswas similar between years (F1,31 5 0.25, P 5 0.620, partial g2

5 0.01) and independent of core area (F1,31 5 2.75, P 5

0.107, partial g2 5 0.08). No other interactions or maineffects were significant.

For the percentage of core area overlap (Fig. 2), we observeda three-way interaction between sex of the individual, sex ofoverlapping tortoise, and year (F1,31 5 9.13, P 5 0.005, adj.partial g2 5 0.19), resulting in part from three two-wayinteractions. As with the first model, sex of the individual andsex of overlapping tortoise affected core area overlap (F1,31 5

5.38, P 5 0.027, adj. partial g2 5 0.12), with females againdemonstrating a significant bias in percentage of core areaoverlap with males. The interaction between year and sex ofthe individual (F1,31 5 5.91, P 5 0.021, adj. partial g2 5 0.13)confirmed that male overlap was similar for both years,whereas females overlapped with more individuals in 2005compared to 2004. The interaction between year and corearea (F1,31 5 5.67, P 5 0.024, adj. partial g2 5 0.13) resultedfrom core areas being larger in 2005 than in 2004.

Burrow use and fidelity.—Burrow occupation was clearlyimportant to tortoises (Table 1). The majority of tortoises

Table 2. Measurements (Mean 6 1 SE, range) of Site Fidelity and Number of Overlapping Tortoises for Home Range and Core Area MinimumConvex Polygons (MCP; hectares) for 34 Desert Tortoises (Gopherus agassizii; 20 Males, 14 Females) in the West Mojave Desert in 2004 and 2005.

2004 2005

Male Female All (range) Male Female All (range)

Home Range100% MCP size 39.4 6 9.4 13.8 6 1.6 28.9 6 5.9 (6–210) 47.4 6 7.8 16.9 6 2.4 34.8 6 4.5 (5–177)Percent yearly overlap — — — 80.1 6 3.7 76.0 6 3.3 78.4 6 2.6 (30–99)Number of overlapping males 4.9 6 0.5 5.0 6 0.4 4.9 6 0.3 (1–9) 7.1 6 0.8 6.5 6 0.7 6.9 6 0.5 (2–14)Number of overlapping females 3.5 6 0.4 2.6 6 0.4 3.1 6 0.3 (1–6) 4.6 6 0.4 2.4 6 0.4 3.7 6 0.3 (0–7)

Core Area50% MCP size 5.4 6 0.8 1.3 6 0.3 3.7 6 0.6 (0.2–13) 6.5 6 1.0 1.9 6 0.5 4.6 6 0.7 (0.3–15)Percent yearly overlap — — — 39.8 6 8.4 29.2 6 8.6 35.4 6 6.0 (0–100)Number of overlapping males 0.7 6 0.2 0.9 6 0.3 0.8 6 0.2 (0–3) 1.0 6 0.2 0.9 6 0.3 0.9 6 0.2 (0–4)Number of overlapping females 0.7 6 0.2 0 0.4 6 0.1 (0–3) 0.6 6 0.1 0.1 6 0.1 0.4 6 0.1 (0–2)

Harless et al.—Desert tortoise home range analyses 381

Fig. 1. Graphical display of minimum convex polygons (MCP) for male (open polygons) and female (shaded polygons) Desert Tortoises in the westMojave Desert over two years of study. The home range is presented for (A) 2004 and (B) 2005, while core area is presented for (C) 2004 and (D)2005, respectively. Map areas represent the same position on the study site. Scale bars represent one kilometer on the study area.

382 Copeia 2009, No. 2

were found at or within 1 m of a burrow in 2004 (94.6% of n5 2,426 locations) and 2005 (96.2% of n 5 2,645 locations).The number of burrows used by the sample population oftortoises was similar for the two years, with 412 recorded in2004 and 418 in 2005. All tortoises used one or more of thesame burrows in each of the two years (range 5 1–10), with293 burrows used by all tortoises over both years. Roughlyone-third of burrows used by individuals in 2004 were usedagain by the same tortoises in 2005 (Table 1). Relativeburrow fidelity, the number of burrows used in both yearsdivided by the mean number of burrows used in each year,was similar for both sexes (males: 34.3 6 5.5, females: 33.8 6

3.6; Mann-Whitney U 5 95.0, asymptotic P 5 0.115).

Burrow use and home range size.—We used a mixed ANCOVAmodel to examine how the mean number of burrows usedper year varied with year (within-subjects), sex of theindividual (between-subjects), and home range area (log10-transformed mean 100% MCP over both years as thecovariate). The number of burrows used (Table 1) wasindependent of home range size (F1,31 5 0.80, P 5 0.38,partial g2 5 0.03) and was similar for the two years (F1,31 5

0.52, P 5 0.478, partial g2 5 0.02). However, differentialburrow use was observed between the sexes (F1,31 5 5.26, P 5

0.029, partial g2 5 0.15), with males utilizing a significantlygreater number of burrows per year than females. Nosignificant interactions existed among these variables.

To determine whether number of overlapping individualsinfluenced burrow use, we added both number of overlap-ping males and number of overlapping females as covariatesto the preceding model. We again found just one significantrelationship, the difference between the sexes (F1,29 5 6.46,P 5 0.017, partial g2 5 0.18). We obtained similar resultswhen using percent home range overlap of males and offemales instead of number of overlapping individuals ascovariates (for sex: F1,29 5 6.49, P 5 0.016, partial g2 5 0.18).Thus, in models including 100% MCP, the number ofburrows used depended on sex but otherwise was indepen-dent of home range size, year, number of overlappingtortoises, and percent of home range overlap with otherindividuals.

Burrow use and core area size.—We used another mixedANCOVA model to examine how the mean number of

Fig. 2. Mean (+ 1 SE) percentage of overlap of individual minimum convex polygon (MCP) areas with other male and female Desert Tortoises,computed from home ranges (A, B) and core areas (C, D) in 2004 and 2005, respectively. Filled bars represent overlap values for males, with unfilledbars representing overlap values for females.

Harless et al.—Desert tortoise home range analyses 383

burrows used per year varied by year (within-subjects), sex ofthe individual (between-subjects), and core area (log10-transformed mean core area over both years as thecovariate). We obtained very different results when usingcore area, compared with home range as a covariate(preceding section). There was no difference between years(F1,31 5 0.22, P 5 0.646, partial g2 5 0.01) and sexes (F1,32 5

1.63, P 5 0.211, partial g2 5 0.05), but there was asignificant positive association between number of burrowsused and core area (F1,31 5 6.64, P 5 0.015, partial g2 5

0.18). In two additional models, core area was again the onlysignificant main effect when we added either number ofoverlapping males and females as covariates (core area: F1,29

5 4.51, P 5 0.040, partial g2 5 0.14) or the percentage ofoverlap by males and females as covariates (core area: F1,29 5

6.25, P 5 0.018, partial g2 5 0.18). Thus, in contrast tomodels involving home range area, these analyses suggest aclose relationship between burrow use and core area that isindependent of sex, year, and overlap with other tortoises.

Unique burrow use.—More than three-quarters of burrowsused each year were only utilized by a single tortoise (2004:77.6% of 412 total burrows; 2005: 78.5% of 418 totalburrows). A 2 3 2 (year 3 sex) mixed ANOVA for numberof unique burrows used per year per individual (Table 1)revealed that males and females used a similar number ofunique burrows (F1,32 5 3.61, P 5 0.066, partial g2 5 0.10).The proportion of unique burrow use per year relative toburrow use over both years was similar for males (2004: 71.4%

total burrows; 2005: 67.4% total burrows) and females (2004:67.8% total burrows; 2005: 61.2% total burrows). There wasno difference between years (F1,32 5 2.75, P 5 0.107, partialg2

5 0.08) and no interaction between year and sex (F1,32 5 0.13,P 5 0.720, partial g2 5 0.004).

Shared burrow use.—All tortoises used burrows that wereoccupied at some point in time (not simultaneously; seenext section) by another tortoise. To examine selectivity inshared burrow use, we applied a 2 3 2 3 2 (year 3 sex 3 sexof other burrow user) mixed ANOVA to the number of

shared burrows per tortoise. We observed a three-wayinteraction between the variables year, sex, and sex of otherburrow user (F1,32 5 4.63, P 5 0.039, adj. partial g2 5 0.13).This resulted in part from a significant two-way interactionand two significant main effects. The two-way interactionbetween sex and sex of other burrow user (F1,32 5 920.65, P, 0.001, adj. partial g2 5 0.43) indicated that males moreoften shared burrows used by other males, whereas femalesmore often shared burrows used by other females (Fig. 3).The main effect of year (F1,32 5 20.68, P , 0.001, adj. partialg2 5 0.18) indicated a greater frequency of shared burrowuse in 2005 than 2004. The main effect of sex of otherburrow user (F1,32 5 54.95, P , 0.001, partial g2 5 0.28)resulted from a greater number of males sharing burrowsthan females.

Burrow cohabitation.—Tortoises were occasionally foundcohabitating the same burrow. To examine selectivity incohabitation, we examined the number of cohabitedburrows using a 2 3 2 3 2 (year 3 sex 3 sex of cohabitant)mixed ANOVA. Parametric assumptions were not met in thismodel, but the results were highly robust. In additionalmodels, we added one of three covariates: two-year meanhome range, core area, or number of burrows to examinehow cohabitation might be affected by other measures ofspace and burrow use.

The main effect of year was not significant (F1,32 5 0.06, P5 0.811, adj. partial g2 , 0.01); however, there was asignificant interaction between sex and sex of the cohabi-tant (F1,33 5 29.51, P , 0.001, adj. partial g2 5 0.46; Fig. 4).Males were equally likely to cohabit with individuals ofeither sex, whereas females had a significant male bias incohabitation. None of the three covariates were significantwhen added independently to the model, and in each casethe sex 3 sex of cohabitant interaction remained significant.

Factors affecting home range and core areas.—The aboveanalyses suggested that home range area and core area aresubject to different influences. To confirm the relativeimportance of the above factors to each MCP estimator,

Fig. 3. The frequency of burrow sharing (non-simultaneous use; mean + 1 SE) in a population of Desert Tortoises for 2004 (A) and 2005 (B)separated by sex. Filled bars represent the number of males sharing burrows of another individual, and unfilled bars represent the number of sharingfemales.

384 Copeia 2009, No. 2

we used two more ANCOVA models. The first model treatedhome range area (log10-transformed, two-year mean homerange) as a covariate, with core area (log10-transformed, two-year mean core area) as the covariate in the second model.The independent variable in each model was sex, with themean number of overlapping males, mean number ofoverlapping females, and the mean number of burrows usedper year as covariates.

For home range area, sex was the only significant variable,with males having larger home ranges than females (F1,29 5

11.67, P 5 0.002, partial g2 5 0.29). Home range area wasindependent of mean number of burrows used per year (F1,29

5 0.23, P 5 0.633, partial g2 5 0.01) and the mean numberof overlapping individuals of either sex (males: F1,29 5 3.09,P 5 0.089, partial g2 5 0.10; females: F1,29 5 0.14, P 5 0.712,partial g2 5 0.01).

For core area, sex had a similar effect (F1,29 5 10.54, P 5

0.003, partial g2 5 0.27), but there was also a significantlypositive relationship between core area and mean number ofburrows per year (F1,29 5 4.59, P 5 0.041, partial g2 5 0.14).As with the home range (above), the mean number ofoverlapping males and females did not significantly affectcore area size (males: F1,29 5 1.60, P 5 0.22, partial g2 5 0.05;females: F1,29 5 0.05, P 5 0.83, partial g2 5 0.002).

DISCUSSION

Sexual differences in space use.—For both home range andcore area, sexual differences were very important, althoughsocial influences such as the number of overlappingindividuals of either sex, were negligible. We found thatmales had significantly larger home ranges than females.Male Desert Tortoises studied elsewhere have exhibitedlarger home ranges than females (Holt and Rautenstrauch,1996; Duda et al., 1999; Freilich et al., 2000), although insome studies the difference was not statistically significant(Burge, 1977; Barrett, 1990; O’Connor et al., 1994).

Differences between sexes in land use could arise from anumber of physiological or behavioral factors. First, thedisparity could simply be an artifact of body size, as malesare generally larger than females and, thus, need moreresources (Sandell, 1989; Jetz et al., 2004). The effect of body

size on home range area has been shown for otherherpetofauna, including lizards (Turner et al., 1969; Rose,1982; Perry and Garland, 2002), aquatic turtles (Auffenbergand Iverson, 1979; Bury, 1979), and the Texas Tortoise,Gopherus berlandieri (Auffenberg and Weaver, 1969). How-ever, our results indicate home range area, overlap variables,and burrow use were independent of body size within therange of adults examined (201–285 mm; Harless, 2007). Thisagrees with results of O’Connor et al. (1994) who demon-strated no relationship between body size and home rangearea for a separate population of Desert Tortoises. Second,reproductive condition could affect space use. In somereptiles, gravid females undertake long-distance migrationsto nesting areas and exhibit larger home range areas thanmales (Litzgus and Mousseau, 2004; Goodman et al., 2005).Conversely, in other reptiles, gravid females move substan-tially less than non-gravid females and males, whereas non-gravid females and males exhibit similar home range sizesand movements (Viitanen, 1967; Marshall et al., 2006).Environmental conditions were presumably favorable forreproduction during the years of our study (Lovich et al.,1999) as we experienced high rainfall (National Oceanic andAtmospheric Administration, 2006) and had abundantannual vegetation cover. However because we did not x-ray females to either identify enlarged follicles or eggs, orobserve nesting activities, it is unknown if female reproduc-tive condition may explain the smaller home range and coreareas of female tortoises in this population.

Finally, reproductive behaviors could influence space use.Sexual differences in home ranges could arise, for example,from males patrolling larger areas in search of mates, whichhas been proposed for Desert Tortoises (Berry, 1986) andother reptiles (Gibbons, 1986; Duvall et al., 1992; Walker,2000). In Gopher Tortoises, males move longer distancesand use more burrows during the mating season tomaximize the number of potential mates (Diemer, 1992;Boglioli, 1999; Eubanks et al., 2003). The different pattern ofspace use by males and females suggests managementdecisions should consider each sex separately.

There were no differences between the sexes in sitefidelity. The high degree of site fidelity of individual homeranges in this study was similar to other studies on G.

Fig. 4. The mean number of observations of burrow cohabitation between male and female Desert Tortoises in the west Mojave Desert in 2004 (A)and 2005 (B). Values represent the mean number of observations of a tortoise occupying a burrow with another tortoise of either sex. Cohabitatingmales are represented as filled bars, and cohabitating females as unfilled bars for each sex.

Harless et al.—Desert tortoise home range analyses 385

agassizii. In the eastern Mojave, Holt and Rautenstrauch(1996) found that Desert Tortoises demonstrated a highdegree of home range overlap (100% MCPs) betweensuccessive years (mean 5 78%) and across the three years(65%) of their study. In the western Mojave, Freilich et al.(2000) refound over 77% of tortoises within 300 m of theiroriginal capture sites 1–4 years later. Clearly, the causal basisfor sexual differences in space use warrants further study.

Overlap of individual home ranges and core areas.—Sexualdifferences were evident in the analysis of overlap. Forpercentage of area overlap, both sexes exhibited greateroverlap with other males than females. For number ofoverlapping individuals, females exhibited this bias to agreater extent than males. The presence of females within amale’s core area suggests that males actively included one ormore females within their core areas. Indeed, we observedburrow sharing between sexes (see below) and frequentcourting activities (unpubl. obs.). In contrast to home rangeoverlap, overlap in core areas was independent of the size ofthe core area.

The observed male bias in overlap may be the result of anumber of factors. First, males had significantly larger homeranges and core areas than females. As such, a higher degreeof male overlap in both MCP measures would be expected.Second, if female home ranges are relatively evenly spacedacross the landscape (see core area overlap below), then theywould be expected to overlap more with males than otherfemales. Third, this male bias may be the result of a skewedsex ratio in the sample population. We monitored moremales (n 5 20) than females (n 5 14) during the study, as thesex ratio in the area is skewed towards males. We includedall females in the sample area. Thus, the greater overlap inhome ranges of male tortoises may be expected given oursample population.

Burrow use, home ranges, and core areas.—Differences in thenumber of burrows used between sexes in this samplepopulation are similar to previous studies on DesertTortoises. In the Mojave Desert, the number of burrowsused by a tortoise varies by gender, location, season, andyear, ranging anywhere from one to 20 (Burge, 1978;Bulova, 1994; Duda et al., 1999). We found the number ofburrows used was more important in determining core areasthan home ranges. This suggests that core areas are moredependent upon a network of burrows than are homeranges. In contrast to our findings, Duda et al. (1999) founda correlation at each study site between home range areaand number of burrows used during the year by DesertTortoises at a different location in the west Mojave.Differences in study design and home range estimates maycomplicate the comparison between studies, although a realdifference in habitat use may exist. Although we failed todetect a social component in home range and core areas ofDesert Tortoises, as demonstrated in other reptiles (lizards;Perry and Garland, 2002; Haenel et al., 2003; Goodman etal., 2005), we feel that investigators should pay closerattention to social variables.

Sex differences in burrow use were particularly prevalentin analysis of burrow sharing in the sample population.Males used a greater number of burrows that were also usedby other males in the population, presumably due to theoverall use of a greater number of burrows throughout theyear. Females demonstrated a preference for a few choice

burrows that were previously used by other females. FemaleDesert Tortoises nest primarily within burrows (Bjurlin andBissonette, 2004). Favorable nesting sites may be limited inthis sample population, causing the females to compete foroccupation of burrows. Other measures of burrow use hadlittle significance to home range and core area size. Neitherhome range, core area, nor the number of burrows used peryear was dependent on the frequency of burrow cohabita-tion. Home range area was independent of cohabitation byeither sex or measurements of spatial overlap, suggestingthat reproductive opportunities are not a primary determi-nant of home range area in Desert Tortoises (O’Connor etal., 1994). Burrow fidelity was similar between the sexes,with a third of burrows used again in the second year.Although males used a significantly greater number ofburrows in both years, which corresponded to their largercore area, burrow fidelity was similar between the sexes.

Factors affecting home range and core areas.—In our finalstatistical models, home range and core areas were depen-dent on sex and independent of social interactions,measured by the number of overlapping individuals.Accordingly, home ranges of Desert Tortoises at this siteappear to be the result of multiple functions, not solelysocial interactions. If the social structure of the DesertTortoise were a male-based territorial system, we wouldexpect to see certain patterns in home range overlap andburrow use within the population (Brown and Orians, 1970;Powell, 2000). For example, the home ranges of maletortoises would be expected to overlap little with that ofother male tortoises, and more so with female home ranges.Also, we would expect to see a higher degree of site fidelityin male home ranges between years (Brown and Orians,1970; Powell, 2000). In this study, males and females had asimilar degree of site fidelity, but males had larger homerange areas and a greater number of burrows used betweenyears. Male home ranges also included a significant portionof female home ranges, with few male home ranges entirelyencompassing those of one or more females (Fig. 2),common to polygamous species (Stamps, 1977). O’Connoret al. (1994) also noted this lack of home range exclusivity,suggesting a lack of territoriality in Desert Tortoises.However, the lack of overlap among female Desert Tortoisesin this population raises questions as to the presence of afemale-based territory system. Female tortoises may bedetermining the spatial distribution of males, overlappingthe periphery of the female’s home range. In populations ofGopherus polyphemus, similar observations of space use andburrow sharing have led to the designation of a female-based dominance hierarchy (C. Guyer, unpubl.). Moreinsight is needed to clearly identify the social structure ofDesert Tortoises, as well as other members of Gopherus.

Flexibility in territorial behavior is observed in a numberof species, ultimately related to the degree to which theresource is economically defensible (Brown and Orians,1970; Kaufmann, 1983). A high degree of overlap betweenthe home ranges suggests that territories, as far as DesertTortoises are concerned, are not economically defensible(Powell, 2000). However, tortoises may occupy non-exclu-sive home ranges while exhibiting territorial behavior, asdemonstrated in lizards (Haenel et al., 2003; Kerr and Bull,2006). Depending on the nature of the limiting resource(s),the benefit and cost of maintaining a territory may changetemporally and/or spatially (Kerr and Bull, 2006). This may

386 Copeia 2009, No. 2

be particularly true in long-lived species with individualrecognition, such as the Desert Tortoise, where passiveavoidance may be used as a means to refrain from engagingin costly antagonistic encounters (Stamps, 1977).

Conclusions.—Male and female Desert Tortoises demonstratevery different patterns of space use, with male tortoisesoccupying larger areas of land and using more burrows thanfemales. Burrow sharing among members of the populationexhibited a same sex bias, with males cohabitating inburrows more frequently with other male tortoises thanwith females. Space use in the West Mojave populationappears to be influenced by burrow use characteristics, andless so by social variables. Home ranges and core areas ofmale tortoises overlapped to a much greater degree withother males and females, whereas those of female tortoisesalmost never overlapped. These observations suggest analternative social structure is present in our population,possibly flexible territoriality or a female driven socialsystem. Future studies on this and other populations ofDesert Tortoises are needed to adequately address this issue.

ACKNOWLEDGMENTS

Field work would not have been completed without thecoordinated efforts of the following individuals: J. Adams, N.Bieser, L. Bol, B. DeGregario, K. Howard, C. Howey, D. Jackan,M. Kern, C. LaMere, E. Lee, P. Rosenfeld, L. Spenceley, J.Strong, and A. Walde. We thank S. Dunbar, J. Grant, and T.Drummer for providing helpful comments on earlier drafts ofthis manuscript. Research was conducted under United StatesFish and Wildlife Service (USFWS) recovery permit TE066452-1 (to DKD and ADW) and California Department of Fish andGame (CDFG) Memorandum of Understanding for ScientificCollecting Permit #802005-03 (to DKD and ADW). TheConstruction Engineering Research Laboratory (USACERL),which is an element of the U.S. Army Engineer Research andDevelopment Center (ERDC), was the primary fundingagency for this research. ITS Corporation, the contractingagency for USACERL, and the Directorate of Public Works atFort Irwin National Training Center provided additionalfunding as needed.

LITERATURE CITED

Aguirre, G., A. Adest, and D. J. Morafka. 1984. Home rangeand movement patterns of the Bolson tortoise, Gopherusflavomarginatus. Acta Zoologica Mexicana 1:1–28.

Alberts, A. C. 1993. Relationship of space use to populationdensity in an herbivorous lizard. Herpetologica 49:469–479.

Auffenberg, W. A., and J. B. Iverson. 1979. Demography ofTerrestrial Turtles, p. 541–569. In: Turtles: Perspectivesand Research. M. Harless and H. Morlock (eds.). JohnWiley and Sons, New York.

Auffenberg, W. A., and W. G. Weaver, Jr. 1969. Gopherusberlandieri in southeastern Texas. Bulletin of the FloridaState Museum 13:141–203.

Barrett, S. I. 1990. Home range and habitat of the deserttortoise (Xerobates agassizii) in the Picacho mountains ofArizona. Herpetologica 46:202–206.

Bernstein, N. P., R. J. Rightsmeier, R. W. Black, and B. R.Montgomery. 2006. Home range and philopatry in theornate box turtle, Terrapene ornata ornata, in Iowa.American Midland Naturalist 157:162–174.

Berry, K. H. 1986. Desert tortoise (Gopherus agassizii)relocation: implications of social behavior and move-ment. Herpetologica 42:113–125.

Bjurlin, C. D., and J. A. Bissonette. 2004. Desert tortoisenest site selection and embryogenesis in a changingenvironment, p. 7. In: Proceedings of the 2002 and 2003Symposia of the Desert Tortoise Council. D. K. Duncan, G.Stewart, T. B. Egan, M. Tuegel, and D. Pond (eds.). DesertTortoise Council, Inc., Wrightwood, California: Omni-press, Madison, Wisconsin.

Boarman, W. I., T. Goodlett, and P. Hamilton. 1998.Review of radio transmitter attachment techniques forturtle research and recommendations for improvement.Herpetological Review 29:26–33.

Boglioli, M. D. 1999. Burrow dispersion and occupancypatterns as they relate to habitat parameters and socialbehavior in the gopher tortoise, Gopherus polyphemus.Unpubl. M.S. thesis, Auburn University, Auburn, Ala-bama.

Borger, L., N. Franconi, F. Ferretti, F. Meschi, G. DeMichelle, A. Gantz, and T. Coulson. 2006. An integratedapproach to identifying spatiotemporal and individual-level determinants of animal home range size. TheAmerican Naturalist 168:471–485.

Brown, J. L., and G. H. Orians. 1970. Spacing patterns inmobile animals. Annual Review of Ecological Systems1:239–262.

Bulova, S. J. 1994. Patterns of burrow use by desert tortoises:gender differences and seasonal trends. HerpetologicalMonographs 8:133–143.

Burge, B. L. 1977. Daily and seasonal movements and areasutilized by the desert tortoise (Gopherus agassizii) insouthern Nevada, p. 59–94. In: Proceedings of the 1977Symposium of the Desert Tortoise Council. M. Trotter(ed.). Desert Tortoise Council, Inc., San Diego, California.

Burge, B. L. 1978. Physical characteristics and patterns ofutilization of cover sites used by Gopherus agassizii insouthern Nevada, p. 80–111. In: Proceedings of the 1978Symposium of the Desert Tortoise Council. M. Trotter(ed.). Desert Tortoise Council, Inc., San Diego, California.

Burt, W. H. 1943. Territoriality and home range concepts asapplied to mammals. Journal of Mammalogy 24:346–352.

Bury, R. B. 1979. Population ecology of freshwater turtles,p. 571–602. In: Turtles: Perspectives and Research. M.Harless and H. Morlock (eds.). John Wiley and Sons, NewYork.

DeLaune, M. G. 2003. XTools ArcView Extension (Version9/13/2003). Oregon Department of Forestry, Oregon.

Diemer, J. E. 1992. Home range and movements of thetortoise Gopherus polyphemus in northern Florida. Journalof Herpetology 26:158–165.

Doncaster, C. P. 1990. Non-parametric estimates of inter-action from radio-tracking data. Journal of TheoreticalBiology 143:431–443.

Duda, J. J., A. J. Krzysik, and J. E. Freilich. 1999. Effects ofdrought on desert tortoise movement and activity.Journal of Wildlife Management 63:1181–1192.

Duda, J. J., A. J. Krzysik, and J. M. Meloche. 2002. Spatialorganization of desert tortoises and their burrows at alandscape scale. Chelonian Conservation and Biology4:387–397.

Duvall, D., S. J. Arnold, and G. W. Schuett. 1992. Pitvipermating systems: ecological potential, sexual selection, andmicroevolution, p. 321–336. In: Biology of the Pitvipers.

Harless et al.—Desert tortoise home range analyses 387

J. A. Campbell and E. D. Brodie, Jr. (eds.). Selva, Tyler,Texas.

Eubanks, J. O., W. K. Michener, and C. Guyer. 2003.Patterns of movements and burrow use in a population ofgopher tortoises (Gopherus polyphemus). Herpetologica59:311–321.

Freilich, J. E., K. P. Burnham, C. M. Collins, and C. A.Garry. 2000. Factors affecting population assessments ofdesert tortoises. Conservation Biology 14:1479–1489.

Galbraith, D. A., M. W. Chandler, and R. J. Brooks. 1987.The fine structure of home ranges of male Chelydraserpentina: Are snapping turtles territorial? CanadianJournal of Zoology 65:2623–2629.

Gibbons, J. W. 1986. Movement patterns among turtlepopulations: applicability to management of the deserttortoise. Herpetologica 42:104–113.

Goodman, R. M., A. C. Echternacht, and F. J. Burton.2005. Spatial ecology of the endangered iguana, Cylcuralewisi, in a disturbed setting on Grand Cayman. Journal ofHerpetology 39:402–408.

Graves, B. M., and D. Duvall. 1993. Reproduction, rookeryuse, and thermoregulation in free-ranging, pregnantCrotalus v. viridis. Journal of Herpetology 27:33–41.

Haenel, G. J., L. C. Smith, and H. B. John-Alder. 2003.Home-range analysis in Sceloporus undulates (eastern fencelizard): I. Spacing patterns and the context of territorialbehavior. Copeia 2003:99–112.

Harless, M. L. 2007. Spatial ecology of the desert tortoise:sampling frequency and biological influences. Unpubl.M.S. thesis, Loma Linda University, Loma Linda, Califor-nia.

Holt, E. A., and K. R. Rautenstrauch. 1996. Three-yearmovement patterns of adult desert tortoises at YuccaMountain, p. 89–90. In: Proceedings of the 1995 Sympo-sium of the Desert Tortoise Council. V. Dickinson (ed.).Desert Tortoise Council, Inc., San Bernardino, California.

Hooge, P. N., W. Eichenlaub, and E. Solomon. 1999. TheAnimal Movement Program. United States GeologicSurvey, Alaska Biological Science Center.

Jetz, W., C. Carbone, J. Fulford, and J. H. Brown. 2004. Thescaling of animal space use. Science 306:266–268.

Johnson, D. H. 1980. The comparison of usage andavailability measurements for evaluating resource prefer-ence. Ecology 61:65–71.

Kaufmann, J. H. 1962. Ecology and social behavior of theCoati, Nasua nasua, on Barro Colorado Island, Panama.University of California Publications in Zoology 60:95–222.

Kaufmann, J. H. 1983. On the definitions and functions ofdominance and territoriality. Biological Review 58:1–20.

Kaufmann, J. H. 1992. The social behavior of wood turtles,Clemmys insculpta, in central Pennsylvania. HerpetologicalMonographs 6:1–25.

Kernohan, B. J., R. A. Gitzen, and J. J. Millspaugh. 2001.Analysis of animal space use and movements, p. 126–166.In: Radio-tracking and Animal Populations. J. J. Mill-spaugh and J. M. Marzluff (eds.). Academic Press, SanDiego, California.

Kerr, G. D., and C. M. Bull. 2006. Movement patterns in themonogamous sleepy lizard (Tiliqua rugosa): effects ofgender, drought, time of year and time of day. Journalof Zoology 269:137–147.

Litzgus, J. D., and T. A. Mousseau. 2004. Home range andseasonal activity of southern spotted turtles (Clemmys

guttata): implications for management. Copeia 2004:804–817.

Lovich, J. E., P. Medica, H. Avery, K. Meyer, G. Bowser,and A. Brown. 1999. Studies of reproductive output of thedesert tortoise at Joshua Tree National Park, the MojaveNational Preserve, and comparative sites. Park Science19:22–24.

Marshall, J. C., Jr., J. V. Manning, and B. A. Kingsbury.2006. Movement and macrohabitat selection of the EasternMassasauga in a fen habitat. Herpetologica 62:141–150.

McRae, W. A., J. L. Landers, and J. A. Garner. 1981.Movement patterns and home range of the gophertortoise. American Midland Naturalist 106:165–179.

Mertler, C. A., and R. A. Vannatta. 2004. Advanced andMultivariate Statistical Methods: Practical Application andInterpretation. Pyrczak Publishing, Los Angeles.

Millspaugh, J. J., R. M. Nielson, L. McDonald, J. M.Marzluff, R. A. Gitzen, C. D. Rittenhouse, M. W.Hubbard, and S. L. Sheriff. 2006. Analysis of resourceselection using utilization distributions. Journal of Wild-life Management 70:384–395.

Mysterud, A., F. J. Perez-Barberia, and I. J. Gordon. 2001.The effects of season, sex, and feeding style on homerange area versus body mass scaling in temperateruminants. Oecologia 127:30–39.

National Oceanic and Atmospheric Administration.2006. Climatological data. Volume 64–110, 1960–2006.National Climatic Data Center, Asheville, North Caro-lina. http://www.ncdc.noaa.gov/oa/climate/stationlocator.html

O’Connor, M. P., L. C. Zimmerman, D. E. Ruby, S. J.Bulova, and J. R. Spotila. 1994. Home range size andmovements by desert tortoises, Gopherus agassizii, in theeastern Mojave Desert. Herpetological Monographs 8:60–71.

Perry, G., and T. Garland, Jr. 2002. Lizard home rangesrevisited: effects of sex, body size, diet, habitat, andphylogeny. Ecology 83:1870–1885.

Peters, R. 1978. Communication, cognitive mapping, andstrategy in wolves and hominids, p. 95–108. In: Wolf andMan: Evolution in Parallel. R. L. Hall and H. S. Sharp(eds.). Academic Press, New York.

Powell, R. A. 2000. Animal home range and territories andhome range estimators, p. 65–110. In: Research Tech-niques in Animal Ecology: Controversies and Conse-quences. L. Boitani and T. K. Fuller (eds.). ColumbiaUniversity Press, New York.

Rautenstrauch, K. R., and E. A. Holt. 1995. Selecting anappropriate method for calculating desert tortoise homerange size and location, p. 172–173. In: Proceedings of the1994 Symposium of the Desert Tortoise Council. A.Fletcher-Jones (ed.). Desert Tortoise Council, Inc., SanBernardino, California.

Rose, B. 1982. Lizard home ranges: methodology functions.Journal of Herpetology 16:253–269.

Roth, E. D. 2005. Spatial ecology of a cottonmouth(Agkistrodon piscivorous) population in east Texas. Journalof Herpetology 39:312–315.

Row, J. R., and G. Blouin-Demers. 2006. Kernels are notaccurate estimators of home-range size for herpetofauna.Copeia 2006:797–802.

Samuel, M. D., D. J. Pierce, and E. O. Garton. 1985.Identifying areas of increased use within the home-range.Journal of Animal Ecology 54:711–719.

388 Copeia 2009, No. 2

Sandell, M. 1989. The mating tactics and spacing patterns ofsolitary carnivores, p. 164–182. In: Carnivore Behavior,Ecology, and Evolution. J. L. Gittleman (ed.). CornellUniversity Press, Ithaca, New York.

Sih, A., and J. M. Mateo. 2001. Persistence and punishmentpay: a new model of territory establishment and space use.Trends in Ecology and Evolution 16:477–479.

Stamps, J. A. 1977. Social behavior and spacing patterns inlizards, p. 265–334. In: Biology of the Reptilia. C. Gans andD. W. Tinkle (eds.). Academic Press, London.

Stamps, J. A., and V. V. Krishnan. 2001. How territorialanimals compete for divisible space: a learning-basedmodel with unequal competitors. American Naturalist157:154–169.

Stickel, L. F. 1968. Home range and travels, p. 373–411. In:Biology of Peromyscus (Rodentia), Special PublicationNumber 2. J. A. King (ed.). American Society of Mammal-ogists, Stillwater, Oklahoma.

Turner, F. B., R. I. Jennrich, and J. D. Weintraub. 1969.Home range and body size of lizards. Ecology 50:1076–1081.

Turner, R. M. 1982. Mohave Desert scrub. Desert Plants4:157–179.

U.S. Fish Wildlife Service. 1994. Desert tortoise (Mojavepopulation) Recovery Plan. U.S. Fish and Wildlife Service,Portland, Oregon.

Viitanen, P. 1967. Hibernation and seasonal movements ofthe viper, Vipera berus berus (L.), in southern Finland.Annales Zoologici Fennici 4:472–546.

Walde, A. D., M. L. Harless, D. K. Delaney, and L. L. Pater.2007. Anthropogenic threats to the desert tortoise (Go-pherus agassizii): litter in the Mojave Desert. WesternNorth American Naturalist 67:147–149.

Waldron, J. L., J. D. Lanham, and S. H. Bennett. 2006.Using behaviorally-based seasons to investigate canebreakrattlesnake (Crotalus horridus) movement patterns andhabitat selection. Herpetologica 62:389–398.

Walker, Z. J. 2000. The spatial ecology of the timberrattlesnake (Crotalus horridus) in South Central Indiana.Unpubl. M.S. thesis, Purdue University, Fort Wayne,Indiana.

White, G. C., and R. A. Garrott. 1990. Analysis of WildlifeRadio-tracking Data. Academic Press, San Diego, California.

Woodbury, A. M., and R. Hardy. 1948. Studies of the deserttortoise, Gopherus agassizii. Ecological Monographs 18:145–200.

Harless et al.—Desert tortoise home range analyses 389