Habitat Segregation in Ungulates: Are Males Forced into Suboptimal

Loading...

International Association for Ecology

Habitat Segregation in Ungulates: Are Males Forced into Suboptimal Foraging Habitats through Indirect Competition by Females? Author(s): L. Conradt, T. H. Clutton-Brock and D. Thomson Reviewed work(s): Source: Oecologia, Vol. 119, No. 3 (1999), pp. 367-377 Published by: Springer in cooperation with International Association for Ecology Stable URL: http://www.jstor.org/stable/4222318 . Accessed: 12/12/2012 04:10 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]

.

Springer and International Association for Ecology are collaborating with JSTOR to digitize, preserve and extend access to Oecologia.

http://www.jstor.org

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

Oecologia (1999)119:367-377 119:367377 Oecologia (1999)

1999 © Springer-Verlag Springer-Verlag1999

L. Conradt T.H. Clutton-Brock · D. Thomson

Habitatsegregationin ungulates:are males forced into suboptimal foraginghabitatsthroughindirectcompetitionby females?

Received: 19 August 1998 /Accepted: 13 February 1999

Abstract Sex differencesin habitat use (termed'habitat segregation')are widespreadin sexually dimorphicungulate species. They are a puzzling phenomenon, particularlywhen femalesuse better foraginghabitatsthan males. It has been suggestedthat males, owing to their larger body size and higher forage requirements,are inferiorin indirectcompetitionto femalesand are forced by female grazing pressureinto marginalhabitats ('indirect-competition hypothesis'). This hypothesis has been widely cited and has important theoretical and practical implications. However, evidence for it is inconclusive.The presentpaper presentsthe resultsof the first experimentaltest of the indirect-competitionhypothesis. We manipulatedfemale and male numbersof red deer (CervuselaphusL.) on a large scale on the Isle of Rum, Scotland, and tested the influenceof this manipulationon deer habitat use. We predictedthat where female numbers were reduced, male use of preferred habitatshould increaseand sex differencesin habitatuse should decrease, while a reduction in male numbers should not have such effects. In contrast,we found that the manipulationof female and male numbersdid not affect habitat use, and conclude that the indirect-competition hypothesisdoes not explainhabitat segregation on Rum. Key words Body size dimorphism Grazing behaviour· Herbivoreecology · Jarman-Bellprinciple Sexual segregation

Introduction

1980; Clutton-Brocket al. 1987; Sukamar and Gadgil 1988;Brown et al. 1995). This is termed'habitat segregation'. Habitat segregationis particularlycommon in herbivoresand has attractedwide scientificattention in ungulates (for reviews see Main and Coblentz 1990; Conradt 1997). Since the sexes in ungulatespecies usually share the same physical environment and evolutionary history, sex differencesin habitat preferenceare a puzzling phenomenon,particularlywhen the smaller, subordinatefemales claim the better foraging habitats (e.g. Bowyer 1984; Pellew 1984; Beier 1987; CluttonBrock et al. 1987). Theories of the evolution of habitat segregationare still controversial(for reviews see Main and Coblentz 1990;Conradt 1997;Perez-Barberiaand Gordon 1998). Clutton-Brocket al. (1982, 1987) suggestedas cause of habitat segregationthat males are inferiorto femalesin indirect grazing competition, with the result that females prevail in high-quality foraging habitats and males in marginalhabitats. We term this the 'indirectcompetitionhypothesis'.The backgroundand rationale of this hypothesis are as follows. In relevant species, males are of considerably larger body size than females. Therefore, males have higher forage requirements than females and, thus, need higher forage intake rates than females. Forage intake rates in herbivores dependon bite size and vegetationbiomass (i.e. standing crop; Illius and Gordon 1987, 1992). Since bite size does not increasewith increasingbody size as rapidly as metabolic requirements(Illius and Gordon 1987), males need a higher minimum standing crop than females to achieve efficientforage intake rates. At times of food shortage (e.g. in winter or at high her-

In many sexually dimorphicmammal species, the sexes bivore densities), indirect foraging competition can be differ in habitat use (Gauthier-Hion 1980; Newsome severe in preferred high-quality forage habitats, and standing crop becomes very low in these habitats (e.g. L. Conradt(12) T.H. Clutton-Brock · D. Thomson LARG,Departmentof Zoology, Universityof Cambridge, DowningStreet, Cambridge,CB23EJ,UK

Clutton-Brock et al. 1982; Shank 1982; Gordon 1986). If females graze the vegetation in high-quality forage habitats too low (i.e. down to the minimum forage biomass profitable for females, which is below that for males), males are no longer able to obtain sufficient

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

368

forage intake rates in these habitats and are, thus, forced by indirect female competition into marginal habitats with lower forage quality but higher forage biomass. Clutton-Brocket al.'s (1982, 1987) indirect-competition hypothesisextends the argumentsmade by Jarman (1974) and Bell (1971) for explaining inter-specific competitionbetweendifferent-sizedherbivorespeciesto intra-specific competition between the different-sized sexes of a species. However, evidence in favour of the hypothesisis restrictedto reportsthat in many species, femalesuse more high-qualityforage habitatsand males use more high-forage-biomasshabitats (Watson and Staines 1978;Bowyer 1984;Osborne 1984;Ordwayand Krausman 1986; Beier 1987;Clutton-Brocket al. 1987; Miquelle et al. 1992; Perez-Barberiaet al. 1997), and such sex differencesin habitat use could be explainedin alternativeways (Conradt 1997, 1998a). For example, higher male than female sensitivityto adverse weather conditionscould lead to highermale use of habitatswith high vegetation biomass, which offer good shelter, but are usually of low forage quality (Jackes 1973;Conradt 1997; Conradt and Guinness 1997). Ultimately, only field experimentscan test the validity of the indirectcompetitionhypothesis. In the present paper, we report results from an experimentaltest of the indirect-competitionhypothesis, using red deer (CervuselaphusL.) on the Isle of Rum, Scotland, as experimentalspecies. We manipulatedthe numbersof male and femaledeer differentlyin different parts of the island. We predicted that if the indirectcompetition hypothesis could explain habitat segregation in red deer, and males are constrainedfrom using high-quality foraging habitat by indirect competition from females, then male use of high-quality forage habitat should increasewith decreasingfemale density, and sex differencesin habitat use should be least pronounced when female numbersare low. On the other hand, we expected that male density should have little influence on male and female habitat use and on sex differencesin habitat use. Additionally,we investigated Clutton-Brock et al.'s (1987) prediction that spatial segregation between the sexes is higher within highquality forage habitat subject to intense competition than within habitat types of lower forage quality where competition is less intense. The argumentis as follows. If some areas of Agrostis/Festucagrasslandare grazed more intenselythan other areas,so that standingcrop is lower in these more intensely grazed (and, thus, preferred)areas, then males, becauseof their higherforage requirements,could be forced by female grazing competition into less heavily grazed areas within Agrostis/ Festuca grassland. If Clutton-Brocket al.'s (1987) explanationwerecorrect,one would expecta largerdegree of inter-sexualspatial segregation within Agrostis/Festuca grassland than within other habitat types where grazing competition is relatively low and should not lead to sex differencesin site use (Clutton-Brocket al. 1982).

Materials andmethods Study area and species Data were collected between 1981 and 1996 on the Isle of Rum, Scotland. The island is a nature reserve, owned by the Scottish Natural Heritage, with the explicit aim to contribute to nature conservation, environmentalscience and education. The island is divided into five red deer managementblocks accordingto natural geographicboundariessuch as hill saddles and steep glens, so that deer movements between blocks are restrictedand negligible (see Clutton-Brock et al. 1997). Red deer density is relatively high on Rum (1986: 19.8 deer/km; Clutton-Brockand Albon 1989). In red deer, the sexes are highly sexually dimorphic (males: 113 kg; females: 71 kg; Clutton-Brock et al. 1987) and show a different pattern of habitat use: males use lower-qualityforage habitat than females, particularlyin winter and, as a consequence, the diet of stags has a higherfibreand lower proteincontent than that of hinds (Charleset al. 1977;Staines and Crisp 1978;Yalden 1978;Staines et al. 1982; Osborne 1984; Clutton-Brocket al. 1997). The Isle of Rum and its red deer population have been describedin detail by Clutton-Brockand Ball (1987) and Clutton-Brocket al. (1982).

Differentialculling regime Until 1991, a similar proportion of male and female deer (target: 10-17%) was regularlyculled on Rum in all managementblocks (with the exception of block 4, see below). At the end of 1991, an experimental manipulation of male and female numbers was started (see Clutton-Brocket al. 1997). The design of the experiment was as follows. In one block (block 1) the number of males was greatlyreduced(by 30% from 6.8 males/km2to 4.7 males/km2; respective female densities: 8.8 and 9.6 females/km2;block size: 18.2 km2);in one block (block 2), the previous culling regime of equal numberof males and females was continued (respectivedeer densities: 7.1 and 6.8 males/km2, and 7.2 and 6.6 females/km2; block size: 21.3 km2);in one block (block 3) the numberof females was greatly reduced (by 45% from 13.9 to 7.6 females/km2;respective male densities: 7.6 and 8.8 males/km2;block size: 21.8 km2); one block (block 4) remained unculled from 1974 and was supposed to be close to carrying capacity since about 1985 (respective deer densities: 3.6 and 2.6 males/km2;and 8.8 and 7.1 females/km2;block size: 17.2 km2);and in one block (block 5), the previous culling regime was continued, but deer numberswere so low that this block was excluded from analysis. A large cull was conducted in 1991and had to be continued in 1992,before the final goals of the differentialculling regimes could be reached. Thereafter, the yearly cull was as low as before the onset of the differential culling regimeto keep the populationdensitiesand sex ratios at the desired levels. The differentialcull took place with the least possible disturbance.One experienced,professionalstalkerdid the shooting in all blocks during two brief periods in 1991 and 1992, which was of equal length in the culled blocks and at a time of year when the deer on Rum had, for decades, been traditional culled. The total density of deer in blocks 1, 2, 4 and 5 did not differ significantlybetween 1981-1991 (before the onset of the differential culling regime) and 1992-1996 (since the onset of the differential culling regime), but in block 3, total density dropped considerably after the introduction of the differentialculling regime (by 24%). This drop in deer density in block 3 (and thus in grazing competition) should further augment the predicted effects of the differential cull on male habitat use, if the indirect-competition hypothesisis correct.Note that absolute values of deer/km2is not a useful basis for comparisons between blocks with respect to foraging competition, since blocks differedin the amount of available habitat types. For the purpose of the present experiment,relative changes in deer densitiesare more meaningfulthan absolute values (Clutton-Brock et al. 1997). For testing the indirect-competition hypothesis,block 3 was consideredas the 'experimental'block, and

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

369 blocks 1, 2 and 4 as 'control' blocks (note that in block 2 and block 4 the culling regime has not been changed between 1981-1991 and 1992-1996). Predictionsof the differentialculling regime To test the indirect-competitionhypothesis, we investigated(a) the differencesbetween blocks with respect to male and female use of high-quality forage habitat and with respect to sex differencesin habitat and space use prior to the experimentalmanipulation;(b) the change in male and female use of high-quality forage habitat and the change in sex differences in habitat and space use after manipulation of male and female numbers in the experimental block 3 compared to the three control blocks, and (c) the differences between blocks with respect to male and female use of highquality forage habitat and with respectto sex differencesin habitat and space use consequent to the experimentalmanipulation. We predicted that (a) in block 3, the reduction of female numbers should lead to an increase in use of high-qualityforage habitat by males and to a reduction in degrees of habitat and spatial segregation, while the reduction of male numbers in block 1 and the continuation of the respective previous culling regimes in block 2 and 4 should not influence male habitat selection and degrees of habitat and spatial segregation in the control blocks; and (b) as a result of the experimental manipulation, block 3 (with reduced female numbers)should be the block with the relativelyhighest use of high-quality forage habitat by males and the lowest degrees of habitat and spatial segregation,while block 4 (in which deer density was close to carryingcapacity and grazingcompetition presumably most severe) should show the lowest male selectivity for highquality forage habitat and the highest degreesof habitat and spatial segregation (predictionsare summarizedin Table 1). Habitat definitions The habitat available to deer was classified as (a) Agrostis/Festuca grassland (i.e. 'high-quality forage habitat') and as (b) other vegetation communitiesof considerablylower forage quality (i.e. 'lowquality forage habitat'; these vegetation communities included heather moorland communitiesand poor, oligotrophic grasslands; see Clutton-Brock et al. 1982). Agrostis/Festuca grassland supported vegetation with the highest proportion of easily digestible dry matter (mean + SD: 64.3 ± 4.9%, n = 17; other vegetation communities: 48.6 ± 6.7%, n = 38; source: Conradt 1997), the largest fraction of available protein (mean ± SD: 1.42 + 0.22%,

n = 98; other vegetation communities: 1.05 ± 0.07%, n = 178; sources: Clutton-Brock et al. 1987; Conradt 1997; L. Conradt, unpublished data) and the highest concentration of important minerals (M.E. Ball, personal communication) on Rum, and a relatively low cellulose and lignin content (Gordon 1986, 1989). Agrostis/Festucagrasslandoffered therefore to deer of both sexes, the highest-qualityforage of all vegetation communities(van Soest 1982; Clutton-Brock et al. 1982, 1987; Osborne 1984; Gordon 1986, 1989; Conradt 1997), while its vegetation biomass was relatively low (Clutton-Brock et al. 1982, 1987, 1997; Gordon 1986, 1989;Conradt 1997). Deer discriminatedfor short Agrostis/Festuca grasslandin comparison to other Scottish vegetation communities under natural (Clutton-Brock et al. 1982, 1987, 1997; Gordon 1986, 1989; Conradt 1997) and experimental (Clarke et al. 1995) conditions, and use and availability of short Agrostis/Festuca grassland has been related to reproductivesuccess in hinds (Iason et al. 1986; Langvatn et al. 1996). The analysis of data was repeated defining only short Agrostis/Festuca grassland as 'highquality forage habitat'. Since the results of this additional analysis did not differ from the principal analysis presented here (see Conradt 1997), these additional results are not given in the Results section.

Island censuses From 1981 to 1996, red deer distribution and habitat use were monitored during regular censuses in spring (March-April, four censuses), summer (July-August, four censuses) and autumn (October-November, three censuses). For this purpose, one of the authors (D.T.), or a research assistant, followed on foot in each census six fixed routes on 6 consecutive days, which covered the complete island. From 45% to 100% of the deer population (estimated each year by the red deer commission; see Clutton-Brock et al. 1997) was seen during each census (means: 60% in spring, 70% in summer and 66% in autumn; see Clutton-Brock et al. 1997). In 1994-1996, another of the authors (L.C.) conducted nine additional winter island censuses (three censuses each year in Janurary-February,following the same routes), since most sex differencesin habitat use in red deer have been observed in winter (Charles et al. 1977; Staines and Crisp 1978; Watson and Staines 1978; Staines et al. 1982; Osborne 1984), and late winter/early springis the time of greatestfood shortage on Rum (Clutton-Brock et al. 1982). Therefore, any sex differences in habitat use due to forage competition should be most obvious during these seasons, and we expected the greatest effect of the differential culling experiment in winter and spring. Each winter island census

Table 1 Predictionsmade by the indirect-competitionhypothesis following the differentialculling manipulation(= no change predicted, + increase predicted,(+) small increase predicted,- decrease predicted,x no predictions) Block Culling regime

1 Reduction in male numbers

2 Even cull of both sexes

3 Reduction in female numbers

Status of block

Control

Control

Experimental

4 No cull; population presumedto be close to carryingcapacity Control

= =

+ (+)

=

Highest x Lowest Lowest

Lowest x Highest Highest

Predictedchanges subsequentlyto the differentialculling regime Male use of Agrostis/Festuca grassland Female use of Agrostis/Festuca grassland

= =

Degree of habitat segregation Degree of spatial segregation Predicteddifferencesbetween blocks subsequentlyto the differentialculling regime Male use of Agrostis/Festucagrassland Medium Medium Female use of Agrostis/Festucagrassland x x Medium Medium Degree of habitat segregation Medium Medium Degree of spatial segregation

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

370 took 8 days, since daylight hours were short in winter. On average since it is based on a relativelyobjective criterion (i.e. on spatial 60% of the estimated total deer population were observed per quadrats),and Conradt (1998b) recommendedthe complementary winter census. Censuses could only be made in daylight. Since examination of habitat and spatial segregation. Additionally, we Clutton-Brock et al. (1982) conducted numerous nocturnal measured degree of spatial segregation (based on 100 x 100 m watches on habitat use by deer on Rum and did not find signifi- spatial quadrats)separatelywithin Agrostis/Festucagrasslandand cant differencesin habitat choice between night and daytime, we within other vegetation communities, since Clutton-Brock et al. believe that daytime habitat use was representativefor overall (1987) predicted that according to the indirect-competitionhyhabitat use. pothesis, the highest degree of spatial segregation should be observed within the Agrostis/Festucagrasslandwhich experiencesto intense competition. Definitions of measures For comparisonsof high-qualityAgrostis/Festucagrasslanduse between blocks, the 'selectivityindex' S (Clutton-Brocket al. 1982) The use of high-quality forage habitat (i.e., Agrostis/Festuca was used, since blocks differedin the amount of available AgrostisI grassland)by males (females)was measuredas the proportionof all Festucagrassland. S was defined as: censused adult males (females) seen on that habitat type (minimum S proportion of animals on Agrostis/Festuca grassland in block x sample size 20; males 5 years and older and females 2 years and proportionof area of Agrostis/Festucagrasslandwithin block x older were considered as adults). We also investigated the use of high-quality Agrostis/Festucagrassland by males in the experimental block 3 relative to that of males in each of the control blocks, using RP for that purpose: Data analysis and statistics of males on proportion Agrostis/] Festucagrasslandin block 3 Years were used as independentunits of analysis(note that data for RP each year in each season consisted of three to four individual of males on Agrostis/Festuca grassland (proportion censuses, each of which included a sample of about 700-800 obin control block x) + (proportion of males on Agrostis/ served deer; thus, although sample size in terms of years is small, Festucagrassland in block 3) the estimate of habitat use in each year in each season is based on a The rationale for this was as follows. Use of high-qualityAgrostis/ large sample of deer, and error deviation within years is estimated Festuca grassland can fluctuate with environmental factors other to be small: assuming a conservative(lumped) sample size of 200 than competitor density (e.g. with weather conditions; Clutton- males per block per year in each season, an averageproportionof Brock et al. 1982; Conradt 1997), and such factors are likely to around 0.3 of animals on Agrostis/Festucagrasslandand an overinfluencemale habitat use in all blocks similarly(see Conradt 1997; dispersionfactor of around 3 (as estimatedelsewherein a different the potential differentialinfluence in blocks could have been wind context:Conradt 1997),the binomialerrordeviationis estimatedto direction; however, wind speed x wind direction did not signifi- be around 1%, i.e. ±0.003). The analysis was carried out separesults in summer cantly influence use of high-quality forage habitat by deer). Mea- rately for each season. Since we found that the results in in autumn did not differ from and spring any decisiveway males in the of use the Agrostis/Festuca grassland by suring for reasons of brevity we show only results for experimentalblock 3 relative to that of males in the control blocks (Conradt 1997), controls for such confounding factors. The prediction was that springand winterin the presentpaper. Since sample sizes (in terms males in block 3 would increase Agrostis/Festucagrassland use of years)were relativelysmall, not only weresignificancetest results relative to males in the control blocks, therefore significancetests used for data interpretation,but any obvious trendsin the temporal pattern of the data since 1992 are also mentioned in the Results for these comparisons were one-tailed. To measure the degree of sex differencesin habitat use ('degree section or can be derived from figures. For winter island censuses, of habitat segregation') between high-quality Agrostis/Festuca individual island-widecensuses were treated as independentunits use in individualwintercensuses grassland and low-quality-foragehabitat, and to measure the de- of analysis, since observedhabitat variable within a year than between years (Conradt gree of sex differencesin space use ('degreeof spatial segregation') was not more based on 1000 x 1000 m spatial quadrats, Conradt's (1998b) 'seg- 1997). Non-parametrictests were used throughout the present pathe data distributionsdid not satisfy the conditions of regation coefficient', SC, was used. The minimum sample size for per, since this purpose was 30 animals of each sex, and SC was slightly parametrictests. Tests are two-tailed, unless stated otherwise.Note that in the pairwise comparisons following the Friedman tests in modified as recommendedby Conradt (in press): the Results, the significance threshold had to be lowered to a = 0.0085, since six post hoc comparisons were made and from 1 - ( - P)6 < 0.05 follows P < 0.0085. SC (_IXi Yi where: X: total number of observed males Y: total number of observed females N: total number of observed animals (X + Y) x,: number of males in ith habitat type/spatial quadrate yi: number of females in ith habitat type/spatial quadrate ni: numberof animals in ith habitat type/spatialquadrate(xi + yi) k: number of habitat types/spatial quadrats This measure can range from '0' (no segregation)to '1' (complete segregation)and is stochasticallyindependentof sex ratio and deer density (Conradt 1998b). By modifying the formula, we avoided Conradt's (1998b) binomial approximationand calculated a more precisevalue. Also, since it is known red deer sexes segregate,there was no danger that Conradt's (1998b) segregationcoefficientcould take a negative value, and we could use the geometric mean of the proportion of males and females that segregate rather than the squared value given by Conradt (1998b). Degree of spatial segregation was measured in addition to degree of habitat segregation,

Results Comparisons between blocks prior to the experimental manipulation Prior to the start of the differential culling regime, in spring, males selected for Agrostis/Festuca grassland most strongly in block 4, followed by block 3, block 2 and block 1 (Fig. la: selectivity coefficient for Agrostis/ Festuca grassland in each block). The differences between block 4 and block I and between block 4 and block 2 were significant (Friedman test: Fr = 16.2, n = 7, k = 4, P < 0.01; pairwise comparison: a = 0.0085; block 4-block 1: z = 3.73, P = 0.0002; block 4-block 2: z = 3.11, P = 0.002; Siegel and Castellan 1988). The

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

371 12 11-

10 -

(a)

987654323 10 1

2

3

4

12 11>abL >0

109-

(b)

8-

E~t

65-

0,

3_-

.0" 0 i

-1

I 1

2

3

I

4

'30 0 t

S3 0U L 0 *0

0.9-

0.89-

(c)

0.7 0.60.5 0.4-

*0 0.

Changes in each block subsequent to the experimental manipulation

0 -0.1

I I 1

ison: a = 0.0085; block 4-block 1: z = 3.68, P= 0.00022; block 4-block 2: z = 4.07, P = 0.00006). The selection by females for Agrostis/Festuca grassland in experimental block 3 was not significantly different from that in any of the control blocks (pairwise comparison: a = 0.0085; block 3-block 1: z = 2.13, P > 0.03, n.s.; block 3-block 2: z = 2.52, P > 0.012, n.s.; block 3block 4: z = 1.55, P > 0.12, n.s.). Degree of habitat segregation in spring was lowest in block 1, followed by block 3, then block 4 and block 2 (Fig. Ic). Only the difference between block 1 and block 2 was significant (Friedman test: Fr = 16.2, n = 7, k = 4, P < 0.01; pairwise comparison: a = 0.0085; block 1-block 2: z = 2.69, P = 0.0072). Degree of habitat segregation did not differ significantly in experimental block 3 from that in the control blocks (pairwise comparison: a = 0.0085; block 3-block 1: z = 0.41, P > 0.6, n.s.; block 3-block 2: z = 2.28, P > 0.02, n.s.; block 3-block 4: z = 1.86, P > 0.06, n.s.). Degree of spatial segregation in spring was lowest in block 3, followed by block 1, then block 4 and block 2 (Fig. Id). These differences between blocks were not significant (Friedman test: Fr = 7.6, n = 7, k = 4, P > 0.05).

2

3

4

1.1 10.90.80.70.60.50.40.30.20.1-

Block 1. control block with reduction of male numbers

In spring, neither males nor females in block 1 changed their use of Agrostis/Festuca grassland (i.e. 'high-quality forage habitat'; see Materials and methods) significantly after the experimental reduction in male numbers (Mann-Whitney tests; males: W = 34.5, m = 5, n = 7, P > 0.8, n.s.; females: Wx = 47.5, m = 5, n = 8, P > 0.08, n.s.; Fig. 2), as had been expected. 1 2 3 Also as expected, degree of habitat segregation did not differ significantly between 1981-1991 and 1992-1996 Block (Mann-Whitney test: Wx = 43, m = 5, n = 7, P > Fig. 1 Male (a) and female(b) selectivity(± SD) for Agrostis/Festuca 0.1, n.s.; Fig. 3). However, degree of spatial segregation grasslandand degreeof habitat(c) and spatial(d) segregation(± SD) in each block in spring before (hatchedcolumns)and after (open increased significantly from 1981-1991 to 1992-1996 columns)experimentalmanipulationand in winterafter experimental (Mann-Whitney test: Wx = 46, m = 5, n = 7, P < manipulation (shaded columns) 0.05; Fig. 3). This change was not predicted by the indirect-competition hypothesis. (Table 1). SP

selection by males for Agrostis/Festuca grassland in the experimental block 3 was not significantly different from that in any of the control blocks (pairwise comparison: a = 0.0085; block 3-block 1: z = 1.86, P > 0.06, n.s.; block 3-block 2: z = 1.24, P > 0.2, n.s.; block 3-block 4: z = 1.86, P > 0.06, n.s.). Females selected in spring for Agrostis/Festuca grassland most strongly in block 4, followed by block 3, block 1 and block 2 (Fig. lb: selectivity coefficient for Agrostis/Festuca grassland in each block). Again, the differences between block 4 and block 1 and between block 4 and block 2 were significant (Friedman test: Fr = 21.8, n = 8, k = 4, P < 0.01; pairwise compar-

Block 2. control block with even cull of both sexes In spring, neither males nor females in block 2 changed their use of Agrostis/Festuca grassland significantly in 1992-1996 relative to 1981-1991 (Mann-Whitney tests; males: Wx = 30, m = 5, n = 8, P = 1.0, n.s.; females: Wx = 44, m = 5, n = 9, P > 0.4, n.s.; Fig. 2). In addition, degrees of habitat and spatial segregation did not differ significantly between 1981-1991 and 1992-1996 (Mann-Whitney tests; habitat segregation: Wx = 29, m = 5, n = 8, P = 1.0, n.s.; spatial segregation: Wx = 41.5, m = 5, n = 8, P > 0.4, n.s.; Fig. 3). These

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

372 1

0.

1 0.9 0.8 0.7 0.6 0.50.4 0.3 0.2 0.1 -

Block 1

0.1980.

0. 6-0.50. 4-

0. 30. 0. 1-

X

I .

..-.... ./

. ...

\. I

I

act oj

co

1 0.90.8 0.7 0.60.5 0.40.30.20.1 A

m E c 0

-S

s: o . &-

/

ss

/v~~~~~~~~9;~~~~~~ I

0o.

...............

81 82 83 84 85 86 90 91 91 92 93 94 94 95 95 96 97 86 87 87 88 88 89 81 82 83 84 85 89 90 92 93 96 97

1 0.90.80.7 0.60.50.40.30.20.1-

cu

0 0 I)

VS

Block 3

o 0O (a

I

I

.

I:

\~~~~~~I ,ee-Ie

.

1 0.9 I 0.8 0.7 I 0.6 0.5 0.4 0.3 0.2 I

.....

0.1 I

......

.

.

.

.

1 0.90.8Block 2 0.7Block 2 ' J,-. 0.6 0.50.4 I 0.3 I 0.2 Block 2 0.1 I 0 11111 I 1111I 11 II 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97

81 83 84 87 88 90 91 91 92 93 94 95 96 97 85 86 86 87 88 89 84 85 89 90 94 95 96 97 81 82 82 83 92 93

n

"

82 83 83 84 84 85 87 88 88 89 90 91 92 92 93 93 9495 94 95 96 96 97 97 818182 85 86 86 87 899091

Block 2

.., .../ ..~~~... ~·

N

,.-"

\,/'

II

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 0:

Block 1

Block 3

'·q .

'/ . .i.\ -..

.

.

..

.

.

.

.

.

.

....

·.

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 1 0.90.80.70.60.5 0.40.30.20.1 0(

Block 4

v r ,

I

I

.

·

·

·

86 87 87 88 88 89 85 86 89 90 91 91 92 92 93 93 94 94 95 81818283 82 83 84 85 95 96 96 97 97

year Fig. 2 Proportion of males (dotted line) and females (solid line) observed on Agrostis/Festuca grassland in each block in each year in spring; the vertical lines denote the start of the differential culling regime

results were as expected, since in block 2 the culling regime did not change between 1981-1991 and 1992-1996 (see Materials and methods). Block 3: experimental block with reduction of female numbers Contrary to predictions of the indirect-competition hypothesis, neither males nor females increased their

year Fig. 3 Degree of habitat (solid line) and 1000 x 1000 m spatial (dotted line) segregation in each year in spring in the four blocks; the vertical lines denote the start of the differential culling regime

use of Agrostis/Festuca grassland significantly in spring (Mann-Whitney tests; males: W, = 47, m = 5, n = 8, P > 0.08, n.s.; females: W, = 39.5, m = 5, n = 8, P > 0.5, n.s.; Fig. 2). The proportion of males on Agrostis/Festuca grassland in block 3 did not increase significantly relative to males in block 1 (Mann-Whitney test: Wx = 34, m = 5, n = 7, P > 0.4 one-tailed, n.s.; Fig. 4a: note that there was not even a trend in the data) or in block 2 (Mann-Whitney test: W, = 46, m = 5, n = 8, P > 0.05 one-tailed, n.s.; Fig. 4b);

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

373

Block 4. control block with no cull

I

0.7- -

0.50.4 0.30.2 0.10.2 -

'

^

0.6 -

V (a)

81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 ct ct

1

'4C:

o

0.90.80.70.60.50.40.30.20.1-

i

I,

In spring, males and females in block 4 did not change their use of Agrostis/Festuca grassland significantly in 1992-1996 relative to 1981-1991 (Mann-Whitney tests; males: W, = 31.5, m = 5, n = 8, P = 1.0, n.s.; females: W, = 39.5, m = 5, n = 8, P > 0.5, n.s.; Fig. 2). Degrees of habitat and spatial segregation also did not differ significantly between 1981-1991 and 1992-1996 in block 4 (Mann-Whitney tests; habitat segregation: W, = 36, m = 5, n = 7, P > 0.5, n.s.; spatial segregation: W, = 322, m = 5, n = 7, P > 0.1, n.s.; see Fig. 3). These results were as expected (Table 1), since in block 4, the culling regime had not been changed.

y -.

:

.

Comparisons between blocks after the experimental manipulation

I'

After the start of the differential culling regime, the situation between blocks with respect to male and female rIVI I,. . . . . . . . . . . . . . . selectivity for Agrostis/Festuca grassland did not differ in 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 spring from the situation before the start of the differential culling regime (Fig. la,b). Degree of habitat segregation decreased in control block 2 (equal cull of both sexes) and increased in control block 4 (unculled) rela1 o0. tive to that in other blocks (Fig. Ic). As a result, the 0.90.8degree of habitat segregation no longer differed between .> 0.7block 1 (reduction in male numbers) and block 2 (pair0.6wise comparison: a = 0.0085; block 1-block 2: z= ...... ... ....I \' . . 0.5I I P > 0.06, n.s.), but degree of habitat segregation I a) 1.84, i· 0.4£in block 4 was now significantly higher than in block 1 0.3: : 0.2! (z = 3.06, P = 0.0022) and experimental block 3 (re0.1duction in female numbers) (z = 3.06, P = 0.0022). i . ..... . ) The degree of spatial segregation still did not differ 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 significantly between blocks (Friedman test: Fr = 4.8, n = 5, k = 4, P > 0.1; Fig. Id). We tested whether males selected most strongly for year Agrostis/Festuca grassland in spring in block 3 (where Fig. 4 Proportionof males on Agrostis/Festucagrasslandeach yearin female numbers had been reduced) and least strongly in springin block 3 relativeto that of males in block 1 (a), block 2 (b) block 4 (where deer density was close to carrying and block 4 (c); the verticallines denote the start of the differential capacity). We found that males in block 3 did not select cullingregime significantly more strongly for Agrostis/Festuca grassland than males in the other blocks (pairwise comparisons: a = 0.0085; block 3-block 1: z = 2.20, P > 0.04, however, it did so relative to males in block 4 (Mann- n.s.; block 3-block 2: z = 1.45, P = 0.14, n.s.; block Whitney tests: Wx = 51, m = 5, n = 8, P < 0.01 3-block 4: z = 1.22, P > 0.2, n.s.; Fig. la). Moreover, one-tailed; Fig. 4c). The degree of habitat segregation males in block 4 selected even more strongly for in block 3 did not differ significantly between 1981Agrostis/Festuca grassland than males in the other 1991 and 1992-1996 (Mann-Whitney test: W, = 38.5, blocks (Fig. la), although block 4 was the block in m = 5, n = 7, P > 0.4, n.s.; Fig. 3: note that there which indirect competition by females should have been was not even a trend in the data), although a decrease most severe (see Materials and methods). These differin degree of habitat and spatial segregation was ences between block 4 and the other blocks were signipredicted for block 3 by the indirect-competition ficant for block 1 and block 2 (Friedman test: hypothesis (see Table 1). Moreover, the degree of spa- Fr = 14.0, n = 5, k = 4, P < 0.01; pairwise compartial segregation even increased significantly between ison: a = 0.0085; block 1-block 4: z = 3.43, 1981-1991 and 1992-1996 (Mann-Whitney test: P = 0.0006; block 2-block 4: z = 2.69, P = 0.0072). Wx = 45, m = 5, n = 7, P < 0.05; Fig. 3). The results differ from the predictions of the indirectcompetition hypothesis (Table 1). co'

(b)

-

..

v

-n

II

.

*. .

I

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

374

n = 9, k = 4,

Further, we tested whether degree of habitat and spatial segregationin spring was lowest in block 3 and highest in block 4. We found that degree of habitat segregationin block 4 was significantlyhigher than in

blocks

Fr = 14.6, n = 9, k = 4, P < 0.01; Fig. lb). Degree

Festuca grassland (solid line) and within all other vegetation communities (dotted line) in each block in each year in spring

(Friedman test: Fr = 9.5,

P < 0.05; Fig. Ic: degree of habitat segregation was highestin block 3, followedby block 1, then block4 and block 2). However,the differenceswere not as predicted block 3 (Friedman test: Fr = 13.8, n = 5, k = 4, by the indirect-competition hypothesis(Table 1):degree P < 0.01; pairwise comparison: a = 0.0085; block 1- of habitatsegregationin block 3 was not lowest, but was block 4: z = 3.06, P = 0.0022) and in block 1 even slightlyhigherthan in block 1 and block 4 (Fig. Ic) (z = 3.06, P = 0.0022), but did not differ significantly and significantlyhigher than in block 2 (pairwisecomfrom block 2 (block 4-block 2: z = 1.22, P > 0.2, n.s.). parisons: a = 0.0085; block 3-block 2: z = 2.74, Additionally, degree of habitat segregation was not P = 0.0062), and degreeof habitatsegregationin block significantlylowerin block 3 than in block 1 and block 2 4 was not highest and did not differ significantlyfrom (pairwise comparisons: a = 0.0085; block 3-block 1: degreeof habitat segregationin any of the other blocks z = 0, P = 1.0, n.s.; block 3-block 2: z = 1.84, (pairwise comparisons: a = 0.0085; block 4-block 1: P > 0.06, n.s.; Fig. Ic). Moreover, degree of spatial z = 0.18, P > 0.8, n.s.; block 4-block 2: z = 2.37, segregationdid not differ significantlybetween blocks P = 0.0178, n.s.; block 4-block 3: z = 0.37, P > 0.7, (see above; Fig. Id). Thus, the resultswereonly partlyas n.s.). Furthermore,males in block 3 did not segregate predicted by the indirect-competition hypothesis less from femaleswith respectto spatial use, and males in block 4 did not segregatemore from femalesthan in (Table 1). Finally, for the winter season, we tested whether the other blocks (Fig. Id); the degree of spatial segremales' selection for Agrostis/Festucagrassland was gation did not differ significantly between blocks highest in block 3 and lowest in block 4 and whether (Friedman test: Fr = 0.6, n = 9, k = 4, P > 0.2). degree of habitat and spatial segregationwas lowest in block 3 and highestin block 4. In winter,after the start betweendegreeof spatial of the differentialculling regime, males in block 3 did Comparison within Agrostis/Festucagrassland segregation not select more strongly for Agrostis/Festucagrassland and that within other vegetationcommunities and males in block 4 did not select less strongly for Agrostis/Festucagrasslandthan did males in the other The of spatial segregationbetweenthe sexes was blocks, and the selectivity index for Agrostis/Festuca not degreewithin the high-competitionAgrostis/Festuca higher grasslandby males did not differ significantlybetween grassland than within other habitat types but, rather, blocks (Friedman test: Fr = 7.2, n = 9, k = 4, lower 5; Wilcoxon tests: block 1: sample size too P > 0.05, n.s., Fig. la; note: females selected for small (Fig. for testing; block 2: T= 1, n = 8, P < 0.02, Agrostis/Festucagrassland most strongly in block 4, degree of spatial segregationlower within Agrostis/Fefollowed by block 3 and then blocks 1 and 2, and the differencebetweenblocks was significant:Friedmantest: Fig. 5 Degree of 100 x 100 m spatial segregation within Agrostis/ of habitat segregation differed significantly between

I

1 0o

0.5 -

0.75 -

0.25 -

I

0.5-

0.25 Bloc1 2

Block 1

03 'J;

i

0.75 -

u -- II 82

84

86

88

90

I'/ I 92 94

I,

v

I-96

-M

_.

82

.

...-

I

I

I

I

I

I

I

84

86

88

90

92

94

96

E

8

I

ai

C1

1

0.75 -

0.75 -

0.5 -

0.5 -

0.25,, v

I II

0.25-

Block 3 1

I

I

I

I

I

I

I

82

84

86

88

90

92

94

96

Block 4

o1

82

84

1

I

86

88

90

1

1

1

92

94

96

year

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

375

stuca grasslandthan within other habitattypes;block 3: cull of both sexes and with reductionin male numbers. sample size too small for testing; block 4: T= 11, Furthermore,the degree of spatial segregationbetween n = 10, P = 0.1, n.s.). the sexesdid not differbetweenblocks. Finally,in winter (the season in which we had expected the most pronounced effect of the experimentalmanipulation),none Discussion of the predicted differencesbetween blocks were observed.We concludethat the differencesbetweenblocks After female numbershad been reduced in the experi- with respect to male habitat use and degree of habitat mental block, males did not increase their use of and spatial segregationsubsequentto the experimental Agrostis/Festucagrassland either in absolute terms or manipulationalso did not support the indirect-comperelativeto males in two of the control blocks. The ob- tition hypothesis. served increase in Agrostis/Festucagrassland use by The complete analysis of data was repeateddefining males in the experimentalblock relativeto males in the only short Agrostis/Festucagrassland as high-quality unculledcontrol block was most probablynot a conse- forage habitat, since within Agrostis/Festucagrassland, quence of reduction of female numbers in the experi- the areas of short, young growth are of highest quality mental block, since male use of Agrostis/Festuca and are most intensely grazed (Clutton-Brock et al. grasslandrelativeto male use in the unculledblock in- 1987). The results of this additional analysis did not creasedalso in the block with reductionin male numbers differ from the principal analysis presented here (see and in the block with equal cull of both sexes (Conradt Conradt1997),and thereforeare not givenin the present 1997).Thus, the resultsdo not supportthe predictionof paper but can be found in detail in Conradt(1997). the indirect-competitionhypothesis that a reductionin Clutton-Brocket al. (1987) observed that male and female numbers (and thereforein indirect competition female red deer on Rum not only used Agrostis/Festuca through females) leads to an increase in use of high- grassland to a different extent but, within Agrostis/ qualityforagehabitatby males. Furthermore,the degree Festuca grassland, they used differentsites. They sugof habitat and spatial segregationbetween the sexes in gested that the indirect-competitionhypothesis could the experimentalblock did not (as was predicted)de- explainsex differencesin site use withinthe same habitat creaseafterfemalenumbershad been reduced.Thus, the type, as well as sex differencesin habitat use. If their predictedchanges in habitat use pattern subsequentto explanation were correct, one would expect a larger the experimentalmanipulationwere not observed,and degree of spatial segregation within Agrostis/Festuca the resultsdid not support the indirect-competitionhy- grassland than within other habitat types (CluttonBrock et al. 1987). However, we found that degree of pothesis. After the experimentalmanipulation,blocks should spatial segregation was not larger within Agrostis/ have differedwith respect to male selectivityfor Agros- Festuca grasslandthan within other habitat types. We tis/Festucagrasslandand degreesof habitat and spatial conclude that the indirect-competitionhypothesis also segregation. If female indirect competition had been fails to explain sex differencesin site use within highresponsiblefor sex differencesin habitat use, the block quality-foragehabitat. with reduction in female numbers (in which indirect This was the first experimentaltest of the indirectcompetitionby females was reduced)should have been competitionhypothesis.The resultthat manipulationof the block with the highest male selectivityfor Agrostis/ female density (and thus of degree of indirectforaging Festucagrasslandand the lowest degree of habitat and competition through females) did not influence male spatial segregation,while the unculled block (in which habitat choice and degree of habitat segregationin red deer density was close to carryingcapacity and indirect deer in a mannerpredictedby the indirect-competition competitionmost intense;see Clutton-Brocket al. 1982) hypothesis is supported by an observationalstudy by should have been the block with the lowest male selec- Conradt (1997). In a differentcontext, Conradt (1997) tivity for Agrostis/Festucagrasslandand the largestde- investigatedthe influenceof naturalfluctuationsin stag grees of habitat and spatial segregation.The block with and hind density over the past 20 years on the use of reductionof male numbersand the block with even cull Agrostis/Festucagrasslandby males and females in the of both sexes should have been intermediatewith respect unculled block. She found no negative influenceof feto selectivity and degrees of segregation.These differ- male density on use of Agrostis/Festucagrassland by ences between blocks should have been particularly males or on degree of habitat and spatial segregation. pronouncedin winter, when food for deer is short and This result held when she controlled for environmental competition severe (Watson and Staines 1978). We conditions like weatherand net productionof Agrostis/ found that in spring, degree of habitat segregationwas Festuca grassland.We conclude that the indirect-comhighestin the unculledblock, as predicted.However,so petition hypothesis does not explain sex differencesin was male selectivity for Agrostis/Festuca grassland. habitat use in red deer on the Isle of Rum. Because of Moreover, in the block with reduction in female num- the large scale needed for such an experiment,the numbers, male selectivityfor Agrostis/Festucagrasslandwas ber of treatmentblocks was small (n = 1) and the exnot significantlyhigher and degree of habitat segrega- periment could be criticized on the grounds that the tion not significantlylower than in the blocks with even situationin reddeer on Rum mightnot be representative

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

376

for other populationsor species.However,red deerare a Rum and for support with the cull. The work was supported by typicalsexuallydimorphicgrazingherbivorespecieswith grants from NERC, EU (HCM) and SNH. a large sex differencein body size (Clutton-Brocket al. 1987),and sex differencesin habitatuse on Rum (Charles et al. 1977;Clutton-Brocket al. 1997)are representative References for the rest of Scotland(Stainesand Crisp 1978;Watson and Staines1978;Osborne1984).Moreover,the indirect- Beier P (1987) Sex differencesin quality of white-taileddeer diets. Mammol 68:323-329 competitionhypothesiswas first conceived for red deer BellJRHV (1970) The use of the herb layer by grazing ungulatesin on Rum (Clutton-Brocket al. 1982).The negativeresult the Serengeti.Blackwell,Oxford of the present experiment,therefore,suggests that the Bell RHV (1971) A grazing ecosystem in the Serengeti. Sci Am 225:86-93 indirect-competitionhypothesisis not the correctexplaBleich VC, Bowyer RT, WehausenJD (1997) Sexual segregationin nation for habitat segregationin herbivores. mountain sheep:resourceor predation?Wildl Monogr 134:3-50 This result is surprising. Since sex differences in RT (1984) Sexual segregation in southern mule deer. Bowyer in that are habitat use usually occur herbivores sexually J Mammal 65:410-417 dimorphic,an explanationof habitat segregationbased BrownMR, CorkeronPJ, Hale PT, SchultzKW, BrydenMM (1995) Evidence of sex-segregatedmigration in the humpback whale on the sex differencein body size seemedvery convinc(Megaptera novaeangliae).Proc R Soc Lond B 259:229-234 mathematical models have shown Furthermore, ing. Charles WN, McCowan D, East K (1977) Selection of upland that sex differencesin habitat use due to asymmetryin swards by red deer Cervuselaphus L. on Rhum. J Appl Ecol indirect competition are plausible (Illius and Gordon 14:55-64 1987, 1992). Moreover, the indirect-competitionhy- ClarkeJL, Welch D, Gordon IJ (1995) The influenceof vegetation pattern on the grazing of heather moorland by red deer and pothesis is a modification of the widely accepted Jarsheep. I. The location of animals on grass/heather mosaics. man-Bell principle (Bell 1970; Jarman 1974), which J Appl Ecol 32:166-176 holds that smallerungulatespeciesare more effectivein Clutton-Brock TH, Albon S (1989) Red deer in the Highlands. BSP, Oxford grazingshort, high-qualityswardstherebyforcinglarger species onto lower-quality swards. Thus, it seemed Clutton-BrockTH, Ball ME (1987) Rhum: the natural history of an island. University Press, Edinburgh plausiblethat the same principlecould not only be ap- Clutton-BrockEdinburgh TH, Lonergan ME (1994) Culling regimesand sex sized but also to to differently species, plied differently ratio biases in Highland red deer. J Appl Ecol 31:521-527 sized sexes of the same species. However, the relevant Clutton-BrockTH,Guinness FE, Albon SD (1982) Red deer: behaviour and ecology of two sexes. Universityof Chicago Press, between-speciesdifferencesin body size in Bell's (1970, Chicago 1971) examples were up to eight times higher than the Clutton-Brock TH, Major M, Guinness F (1985) Population regaveragebetween-sexdifferencesin body size within relulation in female and male red deer. J Anim Ecol 54:831-846 evant species (Conradt 1997), and we suggest that sex Clutton-Brock TH, lason GR, Guinness FE (1987) Sexual segredifferencesin body size within species are, in spite of gation and density relatedchanges in habitat use in female and male red deer (CervuselaphusL.). J Zool [Lond] 211:275-289 theoreticalconsiderations(Illius and Gordon 1992),not Clutton-BrockTH, Thomson D, Covey C (1997) Monitoringof red the Jarman-Bell to for principle apply. large enough deer changes on Rum. Prog Rep NCLS 024/97/UPB. Scottish Apart from being of interestto evolutionarytheory, Natural Heritage our result has furtherimplications.Habitat segregation Conradt L (1997) Causes of sex differencesin habitat use in red deer (CervuselaphusL.). PhD thesis, University of Cambridge is often accompaniedby sex differencesin performance L (1998a) Could asynchronyin activity between the sexes and survival (Clutton-Brock et al. 1982, 1985; Prins Conradt cause inter-sexualsocial segregationin ruminants?Proc R Soc 1987), and is, therefore,of relevanceto game manageLond B 265:1359-1363 ment and conservationissues. The indirect-competition Conradt L (1998b) Measuringthe degree of sexual segregationin group living animals. J Anim Ecol 67:217-226 hypothesishas been widely cited as a possible explanation of habitat segregation(Clutton-Brocket al. 1982, Conradt L (in press) Social segregation is not a consequence of habitat segregation in red deer and feral soay sheep. Anim 1987;Illius and Gordon 1987;Main and Coblentz 1990; Behav Illius and Gordon 1992;Miquelleet al. 1992;Koga and Conradt L, Guinness FE (1997) Habitat segregationand weather Ono 1994; Main and Coblentz 1996; Main et al. 1996; sensitivity in red deer (Cervus elaphus L.). Adv Ethol Suppl Ethol 32:172 Bleich et al. 1997) and has the potential to influence A (1980) Seasonalvariationsof diet relatedto species Gautier-Hion managementdecisions (e.g. Clutton-Brockand Lonerand sex in a communityof Cercopithecusmonkeys.J Anim Ecol gan 1994; Clutton-Brocket al. 1997), since it suggests 49:237-269 that high female densities have an unsymmetrically Gordon IJ (1986) The feeding strategiesof ungulateson a Scottish Moorland. PhD thesis, CambridgeUniversity large,deterioratingeffecton male foragehabitatuse and Gordon IJ (1989) The feeding ecology of ungulates on the Isle of nutritionalcondition (Clutton-Brocket al. 1982, 1985, Rum. I. The food supply. J Appl Ecol 26:35-51 1987, 1997). The result of the present study that the lason GR, Duck CD, Clutton-BrockTH (1986) Grazing and reindirect-competitionhypothesis cannot explain habitat productivesuccessof red deer:the effectof local enrichementby gull colonies. J Anim Ecol 55:507-515 segregationis, therefore,important. AcknowledgementsWe thank Prof. T. Roper and Dr. H. Amstrong for comments on an earlierdraft of the manuscriptand the Scottish Natural Heritage (SNH) for the opportunity to work on

Illius AW, Gordon IJ (1987) The allometry of food intake in grazing ruminants.J Anim Ecol 56:989-1000 Illius AW, Gordon IJ (1992) Modelling the nutritionalecology of ungulate herbivores- evolution of body size and competitive interactions.Oecologia 89:428-434

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

377 Jackes AD (1973) The use of wintering grounds by red deer in Ross-shire, Scotland. MPhil, University of Edinburgh JarmanPJ (1974) The social organisationof antilope in relation to their ecology. Behaviour 58:215-267 Koga T, Ono Y (1994) Sexual differencesin foraging behaviour of Sika deer (Cervusnippon).J Mammal 75:129-135 Langvatn R, Albon SD, Burkey T, Clutton-BrockTH (1996) Climate, plant phenology and variationin age of first reproduction in a temperateherbivore.J Anim Ecol 65:653-670 Main MB, Coblentz BE (1990) Sexual segregation among ungulates: a critique. Wildl Soc Bull 18:204-210 Main MB, Coblentz BE (1996) Sexual segregation in Rocky Mountain mule deer. J Wildl Manage 60:497-507 Main MB, Weckerly F, Vernon C (1996) Sexual segregation in ungulates: new directions for research. J Mammal 77:449461 Miquelle DG, Peek JM, Van BallenbergheV (1992) Sexual segregation in Alaskan moose. Wildl Monogr 122:1-57 Newsome AE (1980) Differencesin the diets of male and female red kangeroos in central Australia. Afr J Ecol 18:27-31 Ordway LL, KrausmanPR (1986) Habitat use by desertmule deer. J Wildl Manage 50:677-683 Osborne BC (1984) Habitat use by red deer (CervuselaphusL.) and hill sheep in the West Highlands. J Appl Ecol 21:497-506 Pellew RA (1984) Food consumption and energy budgets of the giraffe. J Appl Ecol 21:141-159 Perez-Barberia FJ, Gordon IJ (1998) The influence of sexual dimorphism in body size and mouth morphology on diet

selection and sexual segregation in cervids. Acta Vet Hung 46:357-367 Perez-BarberiaFJ, Olivan M, Osoro K, Nores C (1997) Sex, seasonal and spatial differencesin the diet of Cantabrianchamois (Rupicaprapyrenaica parva) in Cantabrian Mountains. Acta Theriol 42:37-46 Prins HHT (1987) The buffalo of Manyara. PhD thesis, University of Groningen Shank CC (1982) Age-sex differences in the diets of wintering Rocky Mountain bighorn sheep. Ecology 63:627-633 Siegel S, Castellan NJ (1988) Nonparametricstatistics for the behavioural sciences. McGraw-Hill, New York Soest PJ van (1982) Nutritional ecology of the ruminant. O&B, Corvallis, Ore Staines BW, Crisp JM (1978) Observations on food quality in Scottish red deer (CervuselaphusL.) as determinedby chemical analysis of the rumen contents. J Zool [Lond] 185:253-259 Staines BW, CrispJM, ParishT (1982) Differencesin the quality of food eaten by red deer stags and hinds in winter. J Appl Ecol 19:65-77 Sukamar R, Gadgil M (1988) Male-female differencesin foraging on crops by Asian elephants. Anim Behav 36:1233-1235 Watson A, Staines BW (1978) Differences in the quality of wintering areas used by male and female red deer (Cervuselaphus) in Aberdeenshire.J Zool [Lond] 286:544-550 Yalden DW (1978) Observations on food quality in Scottish red deer (Cervuselaphus)as determinedby chemical analysis of the rumen contents. J Zool [Lond] 185:253-259

This content downloaded by the authorized user from 192.168.52.61 on Wed, 12 Dec 2012 04:10:57 AM All use subject to JSTOR Terms and Conditions

Loading...

Habitat Segregation in Ungulates: Are Males Forced into Suboptimal

International Association for Ecology Habitat Segregation in Ungulates: Are Males Forced into Suboptimal Foraging Habitats through Indirect Competiti...

1MB Sizes 0 Downloads 0 Views

Recommend Documents

Sexual segregation in foraging giraffe
et al., 2006) are based on the Jarman - Bell principle (Bell, 1970;. 1971; Jarman, 1974; Geist, 1974). The principle pre

Segregation of rapidly acetylated histones into a chromatin fraction
Although the function of histone acetylation is unknown, it has been postu- lated that this modification might be necess

Child Trafficking into Forced Labor on Lake Volta, Ghana
Victoria Natsu, Ama Norman, Yaw Ofosu, Andy Olsen,. Andee Cooper Parks, Eric Peasah, Mde. ...... edition_web.pdf” Gros

Boeing Union Workers Forced Into Massive Concessions | Jack Rasmus
Jan 5, 2014 - This past weekend, more than 30000 union workers at Boeing Corp. in Seattle, were forced to accept deep co

Contextualising ethnic residential segregation in Sweden: welfare
Dec 1, 2010 - research questions are: 1) How do welfare, housing, immigration and integration policy goals and ... 5. As

Segregation in like charged polyelectrolyte - Research Explorer
bGroup de Física de Coloides y Polímeros, Departamento de Física de la Materia Condensada,. Universidad de Santiago d

ANAK MALES MAKAN
Pdf bab etika komunikasi pdf elearning mirip etika didefinisikan sebagai usaha yang sistematis dengan menggunakan rasio

MONOLOGUES FOR MALES
[PDF]MONOLOGUES FOR MALEShttp://www.ampa.edu.au/.../AMPA%20Music%20Theatre%20Monologues%20-%20...CachedSimilarCharacter

males – warrsun
Sing my song for you bridgit mendler and shane harper lyrics hold. 2017-11-05. By: admin. On: November 5, 2017. In: Miss

Forced Displacement and Gender Justice in Colombia
Transitional Justice and Displacement Project. From 2010-2012, the International Center for Transitional Justice (ICTJ)