Does home range of the black-tailed tree rat(Thallomys nigricauda Thomas 1882) change

with season along an aridity gradient?Joy C. Coleman* & Colleen T. Downs*

School of Biological and Conservation Sciences, Private Bag X01, Scottsville,

Pietermaritzburg, 3209 South Africa

Received 9 November 2009. Accepted 6 April 2010

Home-range size may reflect the physiological state and energetic requirements of an individual.In the black-tailed tree rat (Thallomys nigricauda) we expected differences in the home-rangesize of populations in habitats of differing productivity, as measured along an aridity gradient,and between seasons. We studied the variations in home-range size of black-tailed tree rats byradiotracking in winter 2006, 2007 and summer 2007 at three study sites along an ariditygradient in southern Africa. Home ranges were estimated using 100% and 95% minimumconvex polygons (MCP) and 95% and 50% fixed kernels. Home ranges varied widely, from 166to 80 199 m2 for males and from 46 to 8810 m2 for females. Males had larger home ranges thanfemales, which supports a promiscuous mating system reported for the species. Althoughrange size was reduced in both sexes in winter, this was not significant. We found no significantdifference in home-range size along the aridity gradient. It is suggested that a combination ofhabitat productivity, season and breeding system influences the size of home range and thatthis species displays phenotypic flexibility in terms of its behavioural responses to thesefactors.

Key words: home range, rodent, cavity-dweller, radio-telemetry.

INTRODUCTIONHome range, defined as ‘that area traversed bythe individual in its normal activities of foodgathering, mating, and caring for young’ (Burt1943: 351), is the standard measure of space use bysmall mammals (Bowers et al. 1996). Interspecificvariation in home-range size has been attributedto numerous factors, including body size (McNab1963), energetic requirements (McNab 1963;Harestad & Bunnell 1979; Lindstedt et al. 1986,Swihart et al. 1988), mating system (Ostfeld 1990),habitat productivity (Swihart et al. 1988), popula-tion density (Mares et al. 1976), and predation(Swihart et al. 1988). This variation may differbetween seasons, between sexes and also due toreproductive status (Clutton-Brock et al. 1982).Despite the multitude of studies on factors affect-ing home-range size, there is no agreement as towhich factors are primarily responsible for observedvariation in home-range size, which may in partbe the result of the studies being conducted ondifferent spatial and temporal scales (McLoughlin& Ferguson 2000).

Intraspecific analyses are needed to determinewhether specific ecological arguments hold at

different hierarchical levels (Gompper & Gittle-man 1991) and the extent to which ecologicalfactors affect social organization can best be mea-sured by examining a single variable species in arange of environments (Lott 1991). Yet, relativelyfew studies have focused on intraspecific variationin home-range size, more particularly in popula-tions from different environments (but seeSchoener & Schoener 1982; Ruby & Dunham 1987;Corp et al. 1999).

Habitat productivity and/or rainfall are theprimary determinants of food availability interrestrial systems (Noy-Meir 1973) and may havea constraining effect on the energy budgets ofmammals both at the species level (Lovegrove2000; Mueller & Diamond 2001; Lovegrove 2003)and at the intraspecific-intrapopulational level(Bozinovic et al. 2007). More specifically, habitatproductivity has been shown to influence energybudgets, as measured by basal metabolic rate,between populations from habitats differing inproductivity (Bozinovic et al. 2009). The primarydeterminant of home-range size at the populationlevel is habitat productivity (Tufto et al. 1996;McLoughlin & Ferguson 2000) through the limitsit may impose on energy acquisition (Corp et al.1999). Harestad & Bunnell (1979) proposed the

African Zoology 45(2): 177–188 (October 2010)

*Authors for correspondence.E-mail: downs@ukzn.ac.za / colemanj@ukzn.ac.za

habitat-productivity hypothesis, suggesting thathome-range size in habitats of greater productiv-ity should generate smaller home ranges. In addi-tion, if primary production is directly proportionalto precipitation (Leith 1975), size of home rangefor a species should be larger in areas of relativelylower precipitation (Mares et al. 1976; Harestad &Bunnell 1979), which has been supported bystudies (e.g. Harris & Leitner 2004). A more recentmodel (resource, habitat and predation model)was developed by Brashares & Arcese (2002), link-ing precipitation with habitat structure, femaledispersion and male behaviour. Following thismodel, female home-range size should varyinversely with food abundance and quality, andmales should travel greater distances when femaleshad larger home ranges.

In seasonally varying habitats, habitat produc-tivity varies during the year (Forman 1995). Sinceseasonal studies have demonstrated the influenceof habitat productivity on the energy budgets ofrodents (Bozinovic et al. 2007), one might expectchanges in energetic requirements to be reflectedin the movements of small mammals (Schradin &Pillay 2006) since home-range size is related to theenergetic requirements and reserves of the animal(Hawes 1977).

When winter coincides with harsh environmen-tal conditions, animals may, for example, reducemovements to save energy (McDevitt & Speakman1994). Thus studies of seasonal change in home-range size at the population level (McLoughlin &Ferguson 2000) may lend support to the habitat-productivity hypothesis (Relyea et al. 2000).

Owing to the association between season andreproductive activity, any estimate of seasonaleffects on home-range size should include thepossible impact of reproductive activity of thespecies (Slade & Swihart 1983). Thus, while moststudies have shown changes in home-range sizewith season, it is generally not possible to removethe associated effects of breeding behaviour (Don1983).

Numerous studies of space-use patterns of smallmammals have reported differences in home-rangesize between sexes, male home ranges being typi-cally larger than female home ranges. A commoninterpretation of this difference is that males covera large range to gain access to receptive females,whereas the range of a female mainly reflects thesize needed to obtain resources to raise offspring(Madison 1980; Ostfeld 1985; Gaulin & Fitzgerald1988). In polygynous and promiscuous species,

males usually have larger home ranges thanfemales (Clutton-Brock 1989; Fisher & Lara 1999).

The black-tailed tree rat (Thallomys nigricaudaThomas 1882) is an arboreal, nocturnal, muridrodent, occurring in mesic to xeric regions of SouthAfrica, Botswana, Swaziland, Zimbabwe andNamibia (Skinner & Chimimba 2005). They arefolivores, feeding on young leaves and fine twigsand the outer green coating of seed pods, but maysupplement their diets with insects and gum (deGraaff 1981; Skinner & Chimimba 2005). This is incontrast with the majority of xeric rodents, whichare predominantly granivorous or opportunisti-cally insectivorous. It was suggested that the wateryield is sufficient to permit black-tailed tree rats tosurvive without drinking water (Bradfield inShortridge 1934). While the vast majority of rodentspecies evade the harsh conditions by burrowing,black-tailed tree rats are one of a few species in aridenvironments to have adopted an arboreal life-style with nests composed of sticks and leaves inforks and hollows of trees (Skinner & Chimimba2005; Coleman & Downs 2009).

Information on the degree to which behaviouralflexibility affects species range in southern Africais lacking, which is surprising considering itsimportance with regard to climate change. Whenthe environments within the range of a speciesdiffer, phenotypic variability could provideincreased environmental tolerance (Via et al. 1995).The family Muridae is characterized by phenotypicplasticity in social organization (Ostfeld 1990).Since these populations of black-tailed tree ratsoccur over an aridity gradient, inter-populationcomparisons have the potential to reveal thevariability in home-range size due to differences inhabitat productivity associated with this gradient.Furthermore, one would expect phenotypic flexi-bility in its behaviour in response to the diverseenvironmental conditions to which it is exposed toplay a role in the success and range of the species.

The aim of this study was to determine whetherthere were differences in the home-range size ofpopulations in habitats of differing productivity,as measured along an aridity gradient, and betweenseasons. More specifically, in accordance with thehabitat-productivity hypothesis, we expected thehome ranges in the most arid site, Molopo, to havethe largest size and the most mesic site, Weenen, tohave the smallest size. In addition, we expectedthat black-tailed tree rats would enlarge theirhome ranges in winter and that home ranges ofmales would be larger than those of females.

178 African Zoology Vol. 45, No. 2, October 2010

METHODS

Study sitesThe study was conducted at three sites along an

east–west aridity gradient (Rutherford & Westfall1994; Schulze et al. 1997; Coleman & Downs 2009)(Fig. 1). The mesic locality, Weenen Game Reserve(KwaZulu-Natal Province, South Africa, 28°50’43”S,29°59’12”E), had a mean annual rainfall rangingfrom 682 mm to 752 mm (Coleman & Downs 2009)and mean annual temperature ranging from16.5°C to 17.9°C (Mucina & Rutherford 2006). Thesemi-xeric site, Haina Game Farm (Botswana20°56’56”S, 23°40’40”E), on the northern boundaryof the Central Kalahari Desert, had a mean annualrainfall of approximately 400 mm (Coleman &Downs 2009) and a mean annual temperature ofbetween 20 and 22°C (Schulze & McGee 1978).The xeric site, Molopo Nature Reserve (southernKalahari savanna, North West Province, SouthAfrica 25°50’52”S, 22°55’37”E), had a mean annualrainfall of 333 mm (Coleman & Downs 2009) and amean annual temperature of 19.1°C (Mucina &Rutherford 2006). Aridity index (Emberger 1955)data for the three sites are detailed elsewhere(Coleman 2008; Coleman & Downs 2009). Degreelocations were plotted using a Magellan GPS 4000XL.

Telemetry proceduresWe had ethical clearance from the University

of KwaZulu-Natal. Black-tailed tree rats werelive-trapped at each study site during winter(May–July) in 2006 and 2007, and in summer(January–February 2007) using Elliot traps (ElliotScientific Equipment, Victoria, Australia) (detailsin Coleman & Downs 2009).

Black-tailed tree rats were sexed, weighed, ear-clipped and reproductive status was noted(Coleman & Downs 2009). Black-tailed tree rats(>40 g) were anaethetised using isoflorthane (2%)or Anakel – V (100 mg/ml) and Chanazine(20 mg/ml) (Coleman & Downs 2009). A total of60 black-tailed tree rats, 28 in winter and 32 insummer, were fitted with radio-transmitters(C. Dearden, Pietermaritzburg) and released backinto their natural environment at the point ofcapture after a 4–12 h recovery period. Transmit-ters were two-stage radio-transmitters encased inepoxy resin and attached to a cable-tie collar with a15 cm vertical whip aerial.

Tracking was conducted on foot with a handheld three element aerial and an Alinco DJX10(Osaka, Japan) receiver. Fixes (the location of ananimal at one point in time) were recorded by theirgrid coordinates using a handheld GPS (MagellanGPS 4000 XL). Locations of black-tailed tree ratswere obtained from direct sightings/homing in(White & Garrott 1990) and activity, where possi-ble, was recorded with each fix. Use of homingtechniques had no effect on diurnal locations. Inaddition, there was no evidence that nocturnal

Coleman & Downs: Home range of the black-tailed tree rat along an aridity gradient 179

Fig. 1. Location of the three study sites of Thallomys nigricauda in southern Africa.

locations were affected by homing techniquessince our close proximity to black-tailed tree ratsrarely appeared to affect their behaviour or move-ment patterns (J.C.C., pers. obs.). At the end of thestudy, all animals were recaptured and radio-transmitters removed.

Winter nights and days were divided into 14-and 10-hourly intervals, respectively, while sum-mer nights and days were divided into 11- and 13-hourly intervals, respectively.

Home-range estimation/data analysesWe estimated home-range sizes for black-tailed

tree rats using the Home Range Extension (HRE;Rodgers & Carr 1998) for ArcView 3x (ESRI,Redlands, California, U.S.A.). We selected animalswith >34 fixes for home-range calculations aspreviously determined by Eccard et al. (2004).

Data with independent sampling intervals havebeen found to underestimate home-range sizeand produced less accurate estimates thanautocorrelated fixes (Reynolds & Laundre 1990; deSolla et al. 1999). In addition, current methods usedto detect autocorrelation are not necessarily suit-able for all organisms (Andersen & Rongstad1989). Therefore, using a sampling methodologythat captures biologically independent data,which is species specific, is perhaps more impor-tant than the elimination of autocorrelation whenusing utilization distribution estimation tech-niques.

We used all fixes for each animal to calculate thesize of the home range since kernel estimators arerobust to violations of independence (Swihart &Slade 1997) and estimates should incorporate theentire activity cycle of the animal (Harris et al.1990). Prior to the study, we evaluated independ-ence biologically for this species guided by thedefinition in Lair (1987). We ensured that eachblack-tailed tree rat was radio-tracked every hourof the day during the study period and to confirmnesting sites. At night we allowed at least onehour to elapse between successive locations onindividual mice which was sufficient to allow anindividual to move across its home range, therebyensuring independence of data points, avoidingproblems of autocorrelation (Kenward 1987;Swihardt & Slade 1985), while optimizing thenumber of locations.

We conducted minimum convex polygon (MCP)analyses using both 100% of the locations (100%MCP) and 95% of the locations (95% MCP) with afloating mean approach (Rodgers & Carr 1998)

for each black-tailed tree rat. Minimum convexpolygons were calculated because it is the mostcommonly reported method in the literature(Harris et al. 1990) and thus allows for comparisonwith other studies. All (100%) of fixes were includedto determine the maximal area used and 95%MCPs were calculated to exclude the occasionalexcursion.

We also conducted 95% fixed-kernel analyses(Worton 1989). All were conducted with a 95%volume and default resolution (70 × 70-cell grid)options in HRE (Rodgers & Carr 1998). We alsoused the fixed kernel method as it is less sensitiveto autocorrelated data (Millspaugh & Marzluff2001), has been shown to produce a less biasedhome-range size calculation and better surface fitthan the adaptive kernel method (Worton 1989;Seaman & Powell 1996; Seaman et. al. 1999) provid-ing a more reliable estimate of the outer contoursand centres of activity (Kernohan et al. 2001).Kernel estimates are less likely to be biased thanMCP estimates, given the problems concerningautocorrelated data (Swihart & Slade 1997).

We used the smoothing parameter calculated byArcView (href) except in the cases where a reducedh value (e.g. 0.9 href, 0.8 href, 0.7 href … 0.1 href) resultedin a smaller, non-fragmented home range. In suchcases, we chose the smallest increment of href thatresulted in a contiguous rather than disjoint 95%kernel home-range polygon, referred to as huser.This was used in 8 of 28 analyses. Probability con-tours of 50% were calculated which representedan individual’s core area of greatest use.

We compared male and female home-range sizeto that expected on the basis of body sizes. Weestimated home-range size (A) from body masswith the formula for herbivorous mammals(Swihart et al. 1988):

A = 4.90M1.56 (1)

where M was body mass (kg).Statistical analyses were performed using the

Statistica 7 package (Statsoft Inc., USA). A generallinear model was used to determine differences inmass between sexes. We evaluated the combinedeffects of site, season and sex on mass usingfactorial analysis of variance (ANOVA). Home-range data were not distributed normally, thusnon-parametric analyses were used. Differencesbetween home-range estimates were tested usingFriedman one-way nonparametric ANOVA.Differences between sites, seasons and sex wereassessed using Kruskal-Wallis one-way nonpara-

180 African Zoology Vol. 45, No. 2, October 2010

metric ANOVA. All mean values are presented asmean ± S.E. Significance was assessed at an alphavalue of 0.05.

RESULTSMean mass of female and male tree rats was63.41 ± 4.41 g (n = 16) and 71.47 ± 4.46 g (n = 12),respectively. There was no significant difference inbody mass when the combined effects of site,season and sex was taken into account (ANOVA,F(2,16) = 1.31, P = 0.30), or between sexes (ANOVA,F(1,26) = 1.16, P = 0.22).

Radiotracking

Several black-tailed tree rats died during thestudy, mainly due to predation. Consequently,telemetry data for 13 and 15 tree rats were usedfor home-range analysis in winter and summer,respectively. A total of 1274 fixes was recorded, ofwhich 65.1% were nocturnal fixes. A mean of 45.5(±1.53) fixes was recorded for each black-tailedtree rat. Totals of 406, 493 and 375 fixes wererecorded in Weenen, Haina and Molopo, respec-tively. A mean of 53.1 fixes was recorded for eachhourly interval. Mean number of night fixes foreach animal was 29.6 whereas mean number ofday fixes was 15.9. A mean of 2.1 night fixes wasrecorded for each animal for each hourly interval,whereas a mean of 1.6 day fixes was recorded foreach animal for each hourly interval. Locations of20 animals were autocorrelated according toSchoener’s index (Rodgers & Carr 1988).

Home-range analysis

We found no significant difference in any home-range estimate between sites (Table 1, Fig. 2). Thenon-significant results were due to variability inhome-range sizes (Fig. 2). On average, cores repre-sented 18.2% in Weenen, 17.5% in Molopo and21.1% in Haina of the 95% kernel estimation.

Season did not explain range variability since wefound no significant difference in any home-rangeestimate between seasons (Table 1, Fig. 3). Onaverage, cores represented 18.9% in summer and18.0% in winter of the 95% kernel estimation.

Female mean 95% kernel (2030.1 ± 620.5 m2) waslarger than mean 100% MCP (1341.5 ± 405.8 m2)and mean 95% MCP (743.4 ± 262.8), and femalesused a mean core area (50% kernel) of 301.0 ±100.46 m2 (Fig. 4). On average, core areas repre-sented 14.8% of the 95% kernel estimation.Home-range size of females varied from 4.67 to5248.9 m2 (100% MCP), from 97.0 to 3834.3 m2 (95%

MCP) and from 46.2 to 8806.8 m2 (95% kernel).Female 50% core size ranged from 6.0 to 1499.1 m2.Females had significantly smaller home ranges forall home-range estimates than males (Table 1).

Male mean 95% kernel (20 976.6 ± 7227.49 m2)was larger than mean 100% MCP (8739.7 ±3016.31 m2) and mean 95% MCP (6860.6 ±2261.27), and males used a mean core area (50%kernel) of 3959.9 ± 1448.71 m2 (Fig. 4). On average,core areas represented 18.9% of the 95% kernelestimation. Home-range size of males varied from181.2 to 28721.4 m2 (100% MCP), from 197.4 to23 211.0 m2 (95% MCP) and from 165.7 to80 199.4 m2 (95% kernel). Male 50% core sizeranged from 22.9 to 16365.9 m2.

Black-tailed tree rats showed wide variation intheir home ranges based on the different methods(Figs 2–4). There was a significant differencebetween 100% MCP and 95% MCP estimates(Friedman ANOVA(1,24), P < 0.001). Similarly, therewas a significant difference between 95% MCPand 95% kernel estimates (Friedman ANOVA(1,24),P = 0.00), and again between 100% MCP and95% kernel estimates (Friedman ANOVA(1,27), P <0.001), as well as between 95% MCP and 50%kernel estimates (Friedman ANOVA(1,27), P < 0.001).

Predicted home-range size for males based onbody size was 0.08 ha. Predicted area for femaleswas 0.07 ha. In males, observed 95% MCP and 95%kernel estimates were 8.58 and 26.25 times larger,respectively, than the predicted values. In females,observed 95% MCP and 95% kernel estimateswere 1.12 and 3.06 times greater, respectively, thanpredicted values.

Coleman & Downs: Home range of the black-tailed tree rat along an aridity gradient 181

Table 1. Summary of H and P values of statisticalanalyses (Kruskal-Wallis) of home-range size estimatesof male and female Thallomys nigricauda at the threestudy sites in winter and summer. An asterisk (*) indi-cates significance.

Estimate H-value P-value

100% MCP Site 6.02 0.05Season 1.27 0.26Sex* 5.61 0.02

95% MCP Site 1.89 0.39Season 1.55 0.21Sex* 9.63 0.002

95% kernel Site 1.9 0.39Season 2.26 0.13Sex* 5.04 0.02

50% kernel Site 1.43 0.49Season 2.56 0.11Sex* 4.82 0.03

In winter, all black-tailed tree rats had one corearea except for one male which had three coreareas. In summer, all black-tailed tree rats hadbetween one and two core areas in their range.

DISCUSSION

Radiotracking

We expected our data to be autocorrelated asstudy animals sometimes did not move betweenrelocations and often returned to previously usedportions of their home ranges (Hansteen et al.1997), resulting in a nonrandom use of theirhome range. In addition to returning to commonfeeding or resting sites, a common pattern amongterrestrial mammals is the use of well-known‘runways’ to travel to and from familiar feedingspots or nesting sites. This further contradicts theassumption of random temporal use of home-range area and thus statistical independence oflocations (Samuel et al. 1985; Samuel & Garton1987). As expected, due to the fact that they make

use of cavities, some black-tailed tree rats hadnonstationary home ranges. This resulted fromshifts in centers of activity and marked expansionof home range, which violates the assumption of astationary home range in tests of independence(Swihart & Slade 1997).

Site

Home-range size should increase with de-creased habitat productivity according to the habi-tat- productivity hypothesis (Harestad & Bunnell1979). We would then expect the home ranges inMolopo, the most arid site, to have the largest sizeand Weenen, the mesic site, to have the smallestsize. Unlike predicted, home-range size of black-tailed tree rats did not differ as a function of site.

One suggestion is that home range of black-tailed tree rats may vary due to ecological differ-ences, as measured along an aridity gradient,between study sites, but such differences may bemasked by the variation in home-range size ateach site. Estimates of individual home-range size

182 African Zoology Vol. 45, No. 2, October 2010

Fig. 2. Box-plots of mean home-range size (m2) of Thallomys nigricauda at the three study sites using estimates(a) 100% MCP, (b) 95% MCP, (c) 95% kernels and (d) 50% cores, irrespective of season or sex (� = mean, � = mean ±S.E., I = mean ± 1.96 × S.E., see text for details of abbreviations).

exhibited a trend of high intraspecific variation.Intraspecific variation in home-range size indifferent areas has been reported (Nevo et al. 1982)and can be extensive (Gompper & Gittleman1991). In field studies, it can be difficult to identifycauses of differences between individuals due tothe extreme variation shown by individualswithin a local population and differences due toany one factor would need to be of sufficientmagnitude to supersede the great individualvariation (Mares et al. 1982). The variation withineach population in the present study needs to beconsidered concerning the lack of significantdifferences between sites.

Patterns can be evaluated by relating home-range size to variables linked with productivity,such as latitude and precipitation. Fisher & Owens(2000) used rainfall as an index of environmentalproductivity in macropod marsupials and foundthat annual rainfall explained most variation inhome-range size, suggesting a strong link between

climate and social organization. In the presentstudy, an aridity index was used as a proxy forhabitat productivity. The aridity index is calculatedover a number of years and thus does not take intoconsideration rainfall and primary productivity ina particular year. This might mask the effects of ahigher than average summer rainfall on winterhabitat productivity in xeric and semi-xeric sites inone particular year on home-range size.

Sex

Home-range size of black-tailed tree rats variedintraspecifically in relation to sex as predicted. Forall methods of estimating home-range size, maleblack-tailed tree rats occupied significantly largerhome ranges than females. One reason might bethat males are larger than females and thus requirea larger range for increased energy demands(McNab 1963). In this study there was no signifi-cant difference in body mass between males andfemales and thus differences in body mass alone

Coleman & Downs: Home range of the black-tailed tree rat along an aridity gradient 183

Fig. 3. Box-plots of mean home-range size (m2) of Thallomys nigricauda in winter and summer using estimates(a) 100% MCP, (b) 95% MCP, (c) 95% kernels and (d) 50% cores, irrespective of sex or site. (� = mean, � = mean ±S.E., I = mean ± 1.96 × S.E., see text for details of abbreviations).

could not explain the significantly larger homeranges of males.

Since home-range size can often be determinedby the mating system of a species (Ostfeld 1990),the observed differences in home-range size ofmales and females may reflect differences inparental investment of the sexes (Safi et al. 2007).Eccard et al. (2004) found that male black-tailedtree rats in Molopo had significantly larger homeranges than females during the breeding season,suggesting a promiscuous mating system. Thelarger home range of male black-tailed tree ratsmay result from increased ranging by malesduring the breeding season, although no signifi-cant difference was found in summer and winterwhen sex was taken into account. Male and femaleblack-tailed tree rats 100% MCP home-range sizein the breeding season (summer) in Molopo fallwithin the range reported in a previous study in

the same reserve (Eccard et al. 2004). However,sizes of the 100% MCP estimate in both male andfemale black-tailed tree rats in the non-breedingseason (winter) were smaller than the range givenin the previous study. This would be expectedresponse, according to the habitat productivityhypothesis, to a winter of increased habitatproductivity, following a summer of unusuallyhigh rainfall.

Generally, black-tailed tree rats have a summerbreeding season (Skinner & Chimimba 2005).However, it appears that black-tailed tree ratsexhibit reproductive flexibility and reproduceopportunistically as a lactating female with new-born offspring was found in July 2006 in MolopoNature Reserve (J.C.C., pers. obs.). Opportunisticreproduction has been reported in other rodentsand is probably an adaptation to the unpredict-ability of the climate, allowing reproduction at

184 African Zoology Vol. 45, No. 2, October 2010

Fig. 4. Box-plots of mean home-range size (m2) of male and female T. nigricauda using estimates (a) 100% MCP,(b) 95% MCP, (c) 95% kernels and (d) 50% cores, irrespective of season or site (�= mean, �= mean ± S.E., I = mean ±1.96 × S.E., see text for details of abbreviations)

varying times of the year, for example duringwinter, if conditions are suitable (Jackson &Bernard 2005). This would have implications forseasonal differences in home range size.

Season

In seasonally varying habitats, habitat produc-tivity changes during the year (Forman 1995), andhence seasonal studies are able to test the habitatproductivity hypothesis (Harestad & Bunnell1979). Unlike that predicted, home-range size ofblack-tailed tree rats did not differ as a function ofseason.

Variations in individual movements of black-tailed tree rats may represent dynamic responsesto fluctuating habitat productivity, through itsinfluence on food resources. Home range is likelyto be influenced primarily by food availabilitywithin populations (McLoughlin & Ferguson2000). When food resources are a determinant ofspatial distributions of individuals within a popu-lation, one alternative would be to possess adynamic home-range, movement patterns beingrapidly adjusted to short term fluctuations in foodresources (Mares et al. 1982). Species that wereopportunistic and adaptable in their food andshelter requirements would be better able toexploit available resources than habitat specialists(Haythornwaite & Dickman 2006). In an unpre-dictable environment, such as the semi-xeric andxeric sites in this study, it would be beneficial formale black-tailed tree rats to expand their homerange in winter after high summer rainfall andthus high primary productivity, in support of thehabitat productivity hypothesis (Harestad &Bunnell 1979). A winter with high primaryproductivity would provide metabolic require-ments needed for enlarged home ranges for malesin search of receptive females and for gestationand lactation in females in rodents that displayopportunistic breeding. Reproductive flexibility,as demonstrated in Molopo in winter (see above),could explain the lack of significant differences inhome-range size between seasons. Male black-tailed tree rats breeding opportunistically out ofthe expected breeding season would be expectedto have similar home-range sizes to those in thebreeding season. Male black-tailed tree rats appearto be flexible in this regard, exhibiting seasonallylabile movements. Confounding effects evidentlyexist between season and reproductive activity inthe present study. Moreover, the utilization of dryleaves of A. mellifera and A. luederitzii in winter

(Meyer et al. 2008) suggests that an increase inhome-range size may not be required.

Body size

Home ranges of both male and female black-tailed tree rats were larger than predicted usingthe model of Swihart et al. (1988). In a previousstudy, home-range sizes of males were larger andthose of females smaller than predicted (Eccardet al. 2004). Home-range size of other mammalsexceeded the prediction of the model (Hanski et al.2000; Shibata et al. 2004), suggesting that otherfactors, including the sex of an animal and themating system may be more important than bodysize in determining home-range size.

CONCLUSIONSIt is suggested that home-range size in black-tailedtree rats is determined by a complex interactionbetween habitat productivity, season and breed-ing system, and thus it is difficult to separate theireffects in the field. Any changes in home-rangesize in response to variation in habitat productiv-ity were masked by high individual variation andbehavioural interactions between individuals.

The fact that neither aridity site nor season influ-enced the home-range size of black-tailed tree ratssuggests that they have sufficient behaviouralflexibility allowing opportunistic responses toprevailing conditions in an unpredictable climateas evidenced by intraspecific variability in home-range size. During a winter of unusually highprimary productivity, females could be expectedto breed opportunistically. In response, maleblack-tailed tree rats would expand home-rangesize, resulting in similar home-range sizes betweenaridity sites and seasons. This interaction supportsthe ‘resource, habitat and predation model’(Brashares & Arcese 2002). Knowledge of spaceuse in these subpopulations of black-tailed treerats may help in understanding how smallmammals are able to cope with environmentalchange. Adapting to current climate extremesshould be consistent with adapting to futureclimate change (Hulme 2005). Studies which payattention to geographical variation in responses ofpopulations in different environments enable oneto establish the extent to which variation is due tophenotypic flexibility (Chown & Storey 2006). Thishas consequences for the conservation of smallmammal species in the light of climate change insouthern Africa.

Coleman & Downs: Home range of the black-tailed tree rat along an aridity gradient 185

ACKNOWLEDGEMENTS

We thank our research assistants for their help inthe field, and the management and staff ofWeenen Game Reserve, Molopo Nature Reserveand Haina Game Farm for the use of their proper-ties. We are grateful to Cliff Dearden for the con-struction of radio-transmitters. Veterinarians inEstcourt (Estcourt Veterinary Clinic), Vryberg(Vryberg Animal Hospital) and Maun (VeterinaryAnimal and Agricultural Consultants) are thankedfor performing veterinary procedures. We aregrateful to the National Research Foundation(NRF) for funding (Grant 65723) and to MazdaWildlife for vehicle support.

REFERENCES

ANDERSEN, D.E. & RONGSTAD, O.J. 1989. Home-range estimates of red-tailed hawks based on randomand systematic relocations. Journal of Wildlife Manage-ment 53: 802–807.

BOWERS, M.A., GREGARIO, K., BRAME, C.J., MATTER,S.F. & DOOLEY, J.L. 1996. Use of space and habitatsby meadow voles at the home range, patch and land-scape scales. Oecologia 105: 107–115.

BOZINOVIC, F., MUNOZ, J.L.P. & CRUZ-NETO, A.P.2007. Intraspecific variability in the basal metabolicrate: testing the food habits hypothesis. Physiologicaland Biochemical Zoology 80: 452–460.

BOZINOVIC, F., ROJAS, J.M., BROITMAN, B.R. &VASQUEZ, R.A. 2009. Basal metabolism is correlatedwith habitat productivity among populations ofdegus (Octodon degus). Comparative Biochemistry andPhysiology A 152: 560–564.

BRASHARES, J.S. & ARCESE., P. 2002. Role of forage,habitat and predation in the behavioural plasticity ofa small African antelope. Journal of Animal Ecology 71:626–638.

BURT, W.H. 1943. Territoriality and home range as ap-plied to mammals. Journal of Mammalogy 24: 346–352.

CHOWN, S.L. & STOREY, K.B. 2006. Linking molecularphysiology to ecological realities. Physiological andBiochemical Zoology 79: 314–323.

CLUTTON-BROCK, T.H. 1989. Mammalian matingsystems – Review lecture. Proceedings of the RoyalSociety of London B 236: 339–372.

CLUTTON-BROCK, T.H., GUINESS, F.E. & ALBON, S.D.1982. Red Deer: the Behaviour and Ecology of Two Sexes.Edinburgh University Press, Edinburgh.

COLEMAN, J.C. 2008. Integrated physiology andbehaviour of Thallomys nigricauda along an ariditygradient. Ph.D. thesis, University of KwaZulu-Natal,South Africa.

COLEMAN, J.C. & DOWNS, C.T. 2009. Variation in urineconcentrating ability and water balance of the Black-tailed tree rat Thallomys nigricauda, along an ariditygradient. Comparative Biochemistry and Physiology A154: 508–513.

CORP, N., GORMAN, M.L. & SPEAKMAN, J.R. 1999.Daily energy expenditure of free-living male woodmice in different habitats and seasons. FunctionalEcology 13: 585–593.

DE GRAAFF, G. 1981. The Rodents of Southern Africa.Butterworths, Durban.

DE SOLLA, S.R., BONDURIANSKY, R. & BROOKS, R.J.1999. Eliminating autocorrelation reduces biologicalrelevance of home range estimates. Journal of AnimalEcology 68: 221–224.

DON, B.A.C. 1983. Home range characteristics andcorrelates in tree squirrels. Mammal Review 13:123–132.

ECCARD, J.A., MEYER, J. & SUNDELL, J. 2004. Spaceuse, circadian activity pattern, and mating system ofthe nocturnal tree rat Thallomys nigricauda. Journal ofMammalogy 83: 440–445.

EMBERGER, L. 1955. Afrique du Nord-Ouest. In: PlantEcology: reviews of research, (ed.) UNESCO. UNESCO,Paris.

FISHER, D.O. & LARA, M.C. 1999. Effects of body sizeand home range on access to mates and paternity inmale bridled nailtail walabies. Animal Behaviour 58:121–130.

FISHER, D.O. & OWENS, I.P.F. 2000. Female home rangesize and the evolution of social organisation inmacropod marsupials. Journal of Animal Ecology 69:1083–1098.

FORMAN, R.T.T. 1995. Land Mosaics: the Ecology of Land-scapes and Regions. Cambridge University Press,Cambridge.

GAULIN, S.J.C. & FITZGERALD, R.W. 1988. Home-range size as a predictor of mating systems in Micro-tus. Journal of Mammalogy 69: 311–319.

GOMPPER, M.E. & GITTLEMAN, J.L. 1991. Homerange scaling: intraspecific and comparative trends.Oecologia 87: 343–348.

HANSKI, I.K., STEVENS, P.C., IHALEMPIA, P. & S.E.LONENEL, V. 2000. Home-range size, movements,and nest-site use in the Siberian flying squirrel,Pteromys volans. Journal of Mammalogy 81: 798–809.

HANSTEEN, T.L., ANDREASSEN, H.P. & IMS, R.A. 1997.Effects of spatiotemporal scale on autocorrelation andhome range estimators. Journal of Wildlife Management61: 280–290.

HARESTAD, A.S. & BUNNELL, F.L. 1979. Home rangeand body weight – a reevaluation. Ecology 60: 389–402.

HARRIS, J.H. & LEITNER, P. 2004. Home-range size anduse of space by adult mohave ground squirrels,Spermophilus mohavensis. Journal of Mammalogy 85:517–523.

HARRIS, S., CRESSWELl, W.J., FORDE, P.G., TREW-HELLA, W.J., WOOLLARD, T. & WRAY, S. 1990.Home-range analysis using radio-tracking data – areview of problems and techniques particularlyas applied to the study of mammals. Mammal Review20: 97–123.

HAWES, M.L. 1977. Home range, territoriality, andecological separation in sympatric shrews, Sorexvagrans and Sorex obscurus. Journal of Mammalogy 58:354–367.

HAYTHORNWAITE, A.S. & DICKMAN, C.R. 2006.Long-distance movements by a small carnivorousmarsupial: how Sminthopsis youngsoni (Marsupialia:Dasyuridae) uses habitat in an Australian sandridgedesert. Journal of Zoology 270: 543–549.

186 African Zoology Vol. 45, No. 2, October 2010

HULME, P.E. 2005. Adapting to climate change: is therescope for ecological management in the face of aglobal threat? Journal of Applied Ecology 42: 784–794.

JACKSON, C. & BERNARD, R.T.F. 2005. Effects ofsupplementary food on the winter inhibition ofreproduction in male and female four-striped fieldmice (Rhabdomys pumilio). Reproduction, Fertility andDevelopment 17: 393–400.

KENWARD, R.E. 1987. Wildlife Radio Tagging. AcademicPress, London.

KERNOHAN, B.J., GITZEN, R.A. & MILLSPAUGH, J.J.2001. Analysis of animal space use and movements.In: Radio Tracking and Animal Populations, (eds) J.J.Millspaugh & J.M. Marzluff. Academic Press, SanDiego.

LAIR, H. 1987. Estimating the location of the focal centerin red squirrel home ranges. Ecology 68: 1092–1101.

LEITH, H. 1975. Primary production of the major vegeta-tion units of the world. In: Primary Productivity of theBiosphere, (eds) H. Leith & R.H. Whittaker. Springer-Verlag, New York.

LINDSTEDT, S.L., MILLER, B.J. & BUSKIRK, S.W. 1986.Home range, time, and body size in mammals. Ecol-ogy 67: 413–418.

LOTT, D.F. 1991. Intraspecific Variation in the Social Systemsof Wild Vertebrates. Cambridge University Press,Cambridge.

LOVEGROVE, B.G. 2000. The zoogeography of mamma-lian basal metabolic rate. The American Naturalist 156:201–219.

LOVEGROVE, B.G. 2003. The influence of climate on thebasal metabolic rate of small mammals: a slow–fastmetabolic continuum. Journal of Comparative Physiol-ogy B 173: 87–112.

MADISON, D.M. 1980. Space use and social structure inMeadow voles, Microtus pennsylvanicus. BehavioralEcology and Sociobiology 7: 65–71.

MARES, M.A., WATSON, M.D. & LACHER Jr, T.E. 1976.Home range pertubations in Tamias striatus. Oecologia25: 1–12.

MARES, M.A., LACHER Jr, T.E., WILLIG, M.R., BITAR,N.A., ADAMS, R., KLINGER, A. & TAZIK, D. 1982. Anexperimental analysis of social spacing in Tamiasstriatus. Ecology 63: 267–273.

McDEVITT, R.M. & SPEAKMAN, J.R. 1994. Limits tosustainable metabolic rate during transient exposureto low temperatures in short-tailed field voles (Micr-otus agrestis). Physiological Zoology 67: 1103–1116.

McLOUGHLIN, P.D. & FERGUSON, S.H. 2000. Ahierarchical pattern of limiting factors helps explainvariation in home range size. Ecoscience 7: 123–130.

McNAB, B.K. 1963. Bioenergetics and the determinationof home range size. The American Naturalist 97:133–140.

MEYER, J., RAUDNITSCHKA, D., STEINHAUSER, J.,JELTSCH, F. & BRANDL, R. 2008. Biology and ecologyof Thallomys nigricauda (Rodentia, Muridae) in theThornveld savannah of South Africa. MammalianBiology 73: 111–118.

MILLSPAUGH, J.J. & MARZLUFF, J.M. 2001. RadioTracking and Animal Populations. Academic Press, SanDiego.

MUCINA, L. & RUTHERFORD, M.C. 2006. The Vegetationof South Africa, Lesotho and Swaziland. Strelitzia 19.

South African National Biodiversity Institute,Pretoria.

MUELLER, P. & DIAMOND, J. 2001. Metabolic rate andenvironmental productivity: well-provisioned animalsevolved to run and idle fast. Proceedings of the NationalAcademy of Sciences 98: 12550–12554.

NEVO, E., GUTTMAN, R., HABER, M. & EREZ, E. 1982.Activity patterns of evolving mole rats. Journal ofMammalogy 63: 453–463.

NOY-MEIR, I. 1973. Desert ecosystems: environmentand producers. Annual Review of Ecology and Systemat-ics 4: 25–51.

OSTFELD, R.S. 1985. Limiting resources and terri-toriality in microtine rodents. The American Naturalist126: 1–15.

OSTFELD, R.S. 1990. The ecology of territoriality in smallmammals. Trends in Ecology and Evolution 5: 411–415.

RELYEA, R.A., LAWRENCE, R.K. & DEMARAIS, S. 2000.Home range of desert mule deer: testing the body-size and habitat-productivity hypothesis. Journal ofWildlife Management 64: 146–153.

REYNOLDS, T.D. & LAUNDRE, J.W. 1990. Time intervalsfor estimating pronghorn and coyote home rangesand daily movements. Journal of Wildlife Management54: 316–322.

RODGERS, A.R. & CARR, A.P. 1998. HRE: The HomeRange Extension for ArcView. Ontario Ministry ofNatural Resources, Centre for Northern ForestEcosystem Research, Canada.

RUBY, D.E. & DUNHAM, A.E. 1987. Variation in homerange size along an elevational gradient in theiguanid lizard Sceloporus merriami. Oecologia 71:473–480.

RUTHERFORD, M.C. & WESTFALL, R.H. 1994. Biomesof southern Africa: an objective categorization. Memoirsof the Botanical Survey of South Africa 63: 1–94.

SAFI, K., KONIG, B. & KERTH, G. 2007. Sex differences inpopulation genetics, home range size and habitat useof the parti-colored bat (Vespertilio murinus, Linnaeus1758) in Switzerland and their consequences forconservation. Biological Conservation 137: 28–36.

SAMUEL, M.D. & GARTON, E.O. 1987. Incorporatingactivity time in harmonic home range analysis. Jour-nal of Wildlife Management 51: 254–257.

SAMUEL, M.D., PIERCE, D.J. & GARTON, E.O. 1985.Identifyng areas of concentrated use within thehome range. Journal of Animal Ecology 54: 711–719.

SCHOENER, T.W. & SCHOENER, A. 1982. Intraspecificvariation in home-range size in some Anolis lizards.Ecology 63: 809–823.

SCHRADIN, C. & PILLAY, N. 2006. Female striped mice(Rhabdomys pumilio) change their home rangesin response to seasonal variation in food availability.Behavioral Ecology 17: 452–458.

SCHULZE, R., MAHARAJ, M., LYNCH, S.D., HOWE, B.J.& MELVIL-THOMSON, B. 1997. South African Atlasof Agrohydrology and Climatology. Water ResearchCommission, Pretoria.

SCHULZE, R.E. & McGEE, O.S. 1978. Climatic indicesand classification in relation to the biogeography ofsouthern Africa. In: Biogeography and Ecology ofsouthern Africa, (eds) M.J.A. Werger & A.C. vanBruggen. Dr. W. Junk, The Hague.

SEAMAN, D.E. & POWELL, R.A. 1996. An evaluation of

Coleman & Downs: Home range of the black-tailed tree rat along an aridity gradient 187

the accuracy of Kernel density estimators for homerange analysis. Ecology 77: 2075–2085.

SEAMAN, D.E., MILLSPAUGH, J.J., KERNOHAN, B.J.,BRUDIGE, G.C., RAEDEKE, K.J. & GITZEN, R.A.1999. Effects of sample size on kernel home rangeestimates. Journal of Wildlife Management 63: 739–747.

SHIBATA, F., KAWAMICHI, T. & NISHIBAYASHI, K.2004. Daily rest-site selection and use by the Japanesedormouse. Journal of Mammalogy 85: 30–37.

SHORTRIDGE, G.C. 1934. Mammals of South West Africa.Heinemann, London.

SKINNER, J.D. & CHIMIMBA, C.T. 2005. The Mammals ofthe southern African Subregion. Cambridge UniversityPress, Cambridge.

SLADE, N.A. & SWIHART, R.K. 1983. Home rangeindices for the hispid cotton rat (Sigmodon hispidus)in northeastern Kansas. Journal of Mammalogy 64:580–590.

SWIHART, R.K. & SLADE, N.A. 1985. Influence ofsampling interval on estimates of home-range size.Journal of Wildlife Management 49: 1019–1025.

SWIHART, R.K. & SLADE, N.A. 1997. On testing forindependence of animal movements. Journal ofAgricultural, Biological and Environmental Statistics 2:48–63.

SWIHART, R.K., SLADE, N.A. & BERGSTROM, B.J. 1988.Relating body size to the rate of home range use inmammals. Ecology 69: 393–399.

TUFTO, J., ANDERSEN, R. & LINNELL, J. 1996. Habitatuse and ecological correlates of home range size in asmall cervid: the roe deer. Journal of Animal Ecology 65:715–724.

VIA, S., GOMULKLEWICZ, R., SCHEINER, S.M.,SCHLICHTING, C.D. & VAN TIENDEREN, P.H.1995. Adaptive phenotypic plasticity: consensus andcontroversy. Trends in Ecology and Evolution 10:212–217.

WHITE, G.C. & GARROTT, R.A. 1990. Analysis of WildlifeRadio-tracking Data. Academic Press, San Diego.

WORTON, B.J. 1989. Kernel methods for estimatingthe utilisation distribution in home-range studies.Ecology 70: 164–168.

Responsible Editor: G.N. Bronner

188 African Zoology Vol. 45, No. 2, October 2010