Experimental examination of behavioural interactions between free-ranging wild and domestic canids

Transcription

1 Behv Ecol Sociobiol (2009) 64: DOI /s z ORIGINAL PAPER Experimentl exmintion of behviourl interctions between free-rnging wild nd domestic cnids Abi Tmim Vnk & Mri Thker & Mtthew E. Gompper Received: 5 July 2009 /Revised: 5 July 2009 /Accepted: 11 August 2009 /Published online: 16 September 2009 # Springer-Verlg 2009 Abstrct The structure of mmmlin crnivore communities is strongly influenced by both intrguild competition nd predtion. However, intrguild interctions involving the world s most common crnivore, the domestic dog (Cnis fmiliris), hve rrely been investigted. We experimentlly exmined the behviourl responses of smll cnid, the Indin fox (Vulpes benglensis), to the presence of dogs nd dog odours. Resource competition between dogs nd Indin foxes is low, so it is uncler whether foxes perceive dogs s interference competitors. To test this, we exposed foxes to neutrl, live dog, nd niml odour stimuli t food trys, nd recorded the time spent t food trys, the mount of food eten, nd vigilnce nd non-vigilnce behviours. When dogs were visible, foxes continued to visit the food trys, but reduced the mount of time spent nd food eten t those trys. Foxes were more vigilnt during dog trils thn during neutrl nd odour trils nd lso exhibited lower levels of non-vigilnce behviour (resting nd plying). In contrst, dog odours did not ffect fox forging nd ctivity. These results show tht Communicted by E. Korpimäki A. T. Vnk : M. E. Gompper Deprtment of Fisheries nd Wildlife Sciences, University of Missouri, Columbi, Missouri 65211, USA M. Thker Deprtment of Biology, Indin Stte University, Terre Hute, Indin 47809, USA A. T. Vnk (*) : M. Thker School of Biologicl nd Conservtion Sciences, University of KwZulu-Ntl, Westville Cmpus, Pvt. Bg 54001, Durbn 4000, South Afric e-mil: bivnk@gmil.com vigilnce/forging trde-offs due to interference competition cn occur between ntive nd domestic crnivores despite low dietry overlp. These negtive effects of dogs on smller member of the crnivore guild rise conservtion concerns, especilly for endngered crnivores. In mny prts of the world, free-rnging dog densities re high due to humn subsidies, nd these subsidized predtors hve the potentil to excerbte the indirect effects of humn presence. Keywords Intrguild interction. Free-rnging dogs. Resource voidnce. Forging-vigilnce trde-off. Crnivore conservtion Introduction Mmmlin crnivores cn drmticlly influence the distribution nd density of co-predtors through direct nd indirect competition (Cro nd Stoner 2003; Plomres nd Cro 1999). Theoreticl nd empiricl studies of intrguild competition nd the ssocited occurrence of intrguild predtion suggest tht this dynmic is unidirectionl, with lrger crnivores negtively influencing smller crnivores (Linnell nd Strnd 2000). Interference competition results in direct interction which in turn results in sptil exclusion, hrssment, or t n extreme, mortlity by intrguild predtion (Holt nd Polis 1997). In the ltter cse, one species kills, but often does not consume the other (Dondio nd Buskirk 2006; Plomres nd Cro 1999). Interference competition hs been well-documented mong mmmlin Crnivor nd especilly in the fmily Cnide (Johnson et l. 1996). In most cses of interference competition, the subordinte competitor uses one of two behviourl strtegies to

2 280 Behv Ecol Sociobiol (2009) 64: reduce encounters with dominnt competitor(s), either by voiding rnge overlp, or by modifying home-rnge use to reduce interctions while still llowing for home rnge overlp. For exmple, red foxes, Vulpes vulpes, void coyotes, Cnis ltrns, (Fedrini et l. 2000; Gosselink et l. 2003) nd Arctic foxes, Alopex lgopus, void red foxes (Elmhgen et l. 2002; Tnnerfeldt et l. 2002). In these cses, interference competition results in non-overlpping territories in fvour of the dominnt competitor. In contrst, cnids such s kit foxes, Vulpes mcrotis, hve home rnges tht overlp with those of coyotes despite the fct tht coyotes re the min cuse of mortlity for kit foxes (Cypher nd Spencer 1998; Rlls nd White 1995; White nd Grrott 1997). Here, kit foxes use tctics such s multiple diurnl den use nd hbitt prtitioning to reduce interference interctions with coyotes (Nelson et l. 2007). At the popultion level, the costs of intrguild interference interctions re unmbiguous for the subordinte competitor: incresed mortlity rtes nd decresed popultion sizes (Creel et l. 2001; Linnell nd Strnd 2000). At the level of the individul, however, the behviourl effects of these interctions re less well understood nd re likely more subtle, involving the need to temporlly nd sptilly void prticulr res or lter ctivity budgets (Nelson et l. 2007; St-Pierre et l. 2006). For exmple, Africn hunting dogs, Lycon pictus, incur cost in hunting efficiency by voiding prey-rich hbitt due to the presence of lions, Pnther leo (Creel et l. 2001). After the reintroduction of wolves, Cnis lupus, into Yellowstone, coyotes reduced forging nd incresed vigilnce when in wolf territories or when pproching ungulte crcsses (Atwood nd Gese 2008; Switlski 2003). Similr forging-vigilnce trde-offs were lso experimentlly demonstrted by Scheinin et l. (2006) who exposed red foxes to lrger competitor (golden jckl, Cnis ureus) t food trys. Foxes decresed food consumption t food trys when jckl ws present, but did not show similr rection to jckl odours (urine). Such exmples demonstrte tht the mere presence of dominnt competitor results in risk version tht is mnifested by reduced willingness to use n re nd the ssocited loss of ccess to high qulity resources. Given these strong intrguild interctions between crnivores, it is striking tht the effects of one of the most common crnivores, the domestic dog, Cnis fmiliris, on symptric ntive crnivores hve received little ttention Vnk nd Gompper (2009b). Dogs hve been introduced wherever humns hve settled nd re probbly the most numerous crnivore in the world tody (Dniels nd Bekoff 1989; Wndeler et l. 1993; WHO/WSPA 1990). Throughout rurl res of much of the world dogs re free-rnging; tht is, lthough they my be owned by individuls or ffilited with specific humn hbittions, much of their dily ctivity budget involves free-rnging behviour (Butler et l. 2004; Fiorello et l. 2006). During the course of this rnging, these nimls hve numerous opportunities to interct with ntive crnivore species t multiple levels, for exmple s predtors, prey, nd disese reservoirs (Butler nd du Toit 2002; Butler et l. 2004; Edgonkr nd Chellm 2002; Fiorello et l. 2006). There is little informtion on the competitive dynmics of dogs nd ntive cnids. Such interctions deserve close exmintion becuse the direct nd indirect effects of dogs on other crnivores my be fr-reching. For exmple, in Austrli, dogs (including dingoes, wild dogs, nd their hybrids) fulfill the role of lrger predtor in their interctions with non-ntive meso-predtors such s introduced red foxes nd ferl cts, Felis ctus (Glen et l. 2007; Johnson et l. 2007; Mitchell nd Bnks 2005). However, Austrlin dogs re truly ferl (sensu Boitni et l. 1995; Green nd Gipson 1994) nd there is high overlp in diet between these species (Glen et l. 2006; Mitchell nd Bnks 2005). Contrry to populr belief, few dog popultions re truly ferl with regrds to food cquisition (but see Cmpos et l. 2007; Kruuk nd Snell 1981) s they re typiclly subsidized by humn-derived food sources to lrge extent (Butler nd du Toit 2002; Dniels nd Bekoff 1989). Presumbly, competition for resources is primry driver of interference interctions mong crnivores (Atickem 2003; Cro nd Stoner 2003; Cse nd Gilpin 1974; Linnell nd Strnd 2000). However, even where dogs do not compete with wild crnivores for food (Butler nd du Toit 2002; Vnk nd Gompper 2009), they my still influence the behviour of smller predtors if they ct or re perceived s predtors or interference competitors (Vnk nd Gompper 2009b). Especilly in res where the crnivore community is disrticulted, dogs could fulfill the role of medium-sized crnivore through interference competition nd intrguild predtion (Glen nd Dickmn 2005; Johnson et l. 2007; Vnk nd Gompper 2009b). Dogs in rurl Indi re typicl of free-rnging dogs elsewhere in the world s they re hevily dependent on humn subsidies nd show low dietry niche overlp with wild cnids such s Indin foxes, Vulpes benglensis (Vnk nd Gompper 2009). Indin foxes re typicl smll cnid nd re common throughout the Indin subcontinent (Gompper nd Vnk 2006). We designed field experiment to determine whether dogs, like wild cnids, cn induce behviourl chnges indictive of interference competition in the smller Indin fox merely by their presence. We predicted tht foxes will rect to the presence of dogs becuse dogs re mid-sized cnid similr to golden jckls, C. ureus. Specificlly, the presence of live dog will cuse foxes to reduce visittion rtes to high vlue food sources nd s consequence, consume less food. We lso expect foxes to show greter levels of

3 Behv Ecol Sociobiol (2009) 64: vigilnce in the presence of live dog. Mny vertebrtes including crnivores use olfctory cues to identify potentil predtors nd competitors (Gormn nd Trowbridge 1989; Kts nd Dill 1998; Muller-Schwrze 2006). However, we predict tht dog odours lone will be insufficient to induce chnge in the forging nd ctivity of foxes becuse dogs re widespred nd occur t high densities in the study re, nd thus, we expect tht foxes will be hbituted to the presence of dog odours. Mterils nd methods Study re nd species This study ws conducted in grsslnd ecosystem in nd round the Gret Indin Bustrd Snctury t Nnnj, Mhrshtr in centrl Indi. The regionl lndscpe consists of mtrix of sugrcne fields, sesonl crops, communl grzing lnds, protected grsslnds, nd forestry plnttions. The study re of pproximtely 135 km 2 comprises the protected grsslnds of Nnnj which is bordered by severl villges with combined humn popultion of pproximtely 50,000. This region is semirid nd experiences wet seson from July to October during which 95% of the precipittion occurs (temperture rnge=16 32 C, men nnul precipittion=600 mm), cool dry seson from November to Februry (temperture rnge=6 37 C) nd hot dry seson from Mrch to June (temperture rnge=18 47 C). The locl economy is bsed on gro-pstorlism, nd thus, domestic dogs in this region (dult weight rnge kg) cn be roughly ctegorized into three types: (1) herding dogs tht ccompny grzing livestock in grsslnds (fox hbitt) during the dy, (2) villge dogs tht occsionlly trverse fox hbitt during the dy or t night, nd (3) frm dogs tht re continuous presence in peripherl fox hbitt. Frm dogs occur t densities of 28 dogs per squre kilometer (±3.2 SE) in the study region nd therefore, re common in the lndscpe. Dog diet consists primrily of humn-derived food mterils (grbge nd crops) nd scvenging from livestock crcsses. A smll component of their diet consists of wild cught food, minly rodents (Rttus spp.) nd blck-nped hre (Lepus negricollis) (Vnk nd Gompper 2009). The primry source of dog mortlity ppers to be enzootic virl diseses, intrguild predtion by wolves (the home rnge of one wolf pck overlpped the study re; Hbib 2007), nd trgeted killing of rbid or ggressive individuls by humns (A. T. Vnk nd M. E. Gompper, unpublished dt). The Indin fox is smll fox (dult weight rnge kg) endemic to the Indin subcontinent. It is crepusculr nd nocturnl nd is n opportunistic omnivore tht is common in open short grsslnds (Gompper nd Vnk 2006; Vnk nd Gompper 2007). It is the most common wild crnivore in the study re where it inhbits protected grsslnds, s well s, grzing lnds nd griculturl fields (Vnk nd Gompper, in press). It does not compete directly with dogs for food, s rodents, insects, fruits, nd reptiles mke up lrge prt of the fox s diet (Vnk nd Gompper 2009). The min cuses of mortlity for foxes in this region re poching by humns, disese, nd predtion by lrger crnivores, including dogs, which kill but do not consume foxes (Vnk 2008). Indin foxes breed during the winter, t which time foxes use den tht is mintined by both individuls of mted pir. The breeding seson coincides with the beginning of the dry seson (from Jnury to My) when wter nd food resources become incresingly scrce, nd grsslnd height decreses drmticlly. Experimentl design Trils were conducted from Februry to April 2007, during which time foxes were loclized t breeding dens nd hd predictble dily ctivity ptterns. Behviourl trils were centred on fox dens which were inhbited by resident breeding pir (except for one den, which hd only one dult femle), nd two to four pups ged 3 4 months old. Behviours were recorded for ll foxes t ech den during these trils but den ws considered the smpling unit. To determine if nd how the presence of dogs influences fox ctivity nd forging behviour, foxes t ech den site were exposed to three trils: (1) neutrl (no stimulus), (2) live dog, nd (3) niml odours. The neutrl tril consisted of two un-mnipulted forging trys for the foxes. The dog nd niml odour trils consisted of two forging trys with either n djcent control condition ( non-thretening stimulus) or tretment condition with potentilly thretening stimulus. The difference in fox responses between these pired forging trys nd cross the three trils indictes the level of interference tht dogs hve on fox forging nd ctivity. Ech set of trils ws preceded by 2 3 dys of hbitution of foxes to the food source nd the observers. Two plstic trys ( cm deep) were plced 200 m prt on opposite sides of the min den entrnce. The plstic trys were lined with soil obtined from the site to reduce unfmilir smells. Two boiled chicken eggs divided into eight equl pieces nd 200 g of loclly grown grpes were plced on ech try. We ssumed tht this type of food would be highly ttrctive forging option becuse of the presence of pups t the den sites t time of sesonl food scrcity. The mount of food in the trys ws lso pproprite s foxes would often et ll the food in both trys during the hbitution period, thus, indicting tht

4 282 Behv Ecol Sociobiol (2009) 64: they did not stite fter forging t single try. Ech pir of trys ws plced 100 m from the den site 1 hour prior to sunset, nd observtions were recorded continuously from the time the first fox (either dult or pup) ws sighted within 100 m of either try until 1 hour fter sunset. Behviours of foxes were recorded by two people from the bck of pickup truck 150 m from the den site using spotting scope nd two Sony minidv cmcorders. Videos were then scored by ATV. Foxes were considered hbituted when they visited nd te from both food trys during the 2 hr observtion period. During the hbitution period, ll foxes showed preference for one try over nother, nd therefore, we ssigned the tretments for the dog nd niml odour trils non-rndomly to control for initil try bis (Fig. 1). A preferred try ws identified s the try where foxes consumed more food (>50%) or spent more time (>50%) during the hbitution period nd ws often the first try discovered (in five out of seven trils). Initil preference for try ws mostly reduced by the time the neutrl tril ws conducted 2 3 dys lter (see results). Following this initil hbitution period, we exposed ech set of foxes to the neutrl, dog, nd niml odour trils on consecutive dys in tht order. The order of the trils ws explicitly set to test predictions. Odour cues re likely to be reinforced by the ctul presence of dog during the live dog tril. Therefore, lck of difference between tretment nd control trys during the odour tril would be stronger support for our Control or Tretment 20 m Food try m 100 m Fox Den Food try 2 20 m Control or Tretment Experimentl setup No stimulus (preferred food try) Dy 1: Neutrl No stimulus (non-preferred food try) Live dog in cge (Tretment) Dy 2: Live dog Empty cge (Control) Blckbuck dung nd wter (Control) Food try 1 Food try 1 Food try 1 Food try 2 Order of trils Dy 3: Animl odour Food try 2 Food try 2 Dog sct nd urine (Tretment) Fig. 1 Experimentl design nd order of neutrl, live dog, nd niml odour trils. Note the lternting ssignment of tretments nd controls t trys prediction tht dog odours do not ffect fox forging nd ctivity. For the neutrl tril, no stimulus ws plced t either try, thus, providing bseline mesure of undisturbed ctivity nd forging behviour. For the live dog tril, trined dog ws plced in cge ( cm) in plin sight 20 m from one try (tretment) nd similr-sized empty cge ws plced t the other try (control). The live dog tretment ws intentionlly plced t the preferred try (Fig. 1). The stimulus dog ws fully vccinted, locl mixed-breed niml rised by ATV, nd ws hbituted to the cge prior to these trils. Thus, the dog st clmly in the cge during the trils nd ws not gitted except, in one instnce when he growled t n pproching fox. Although the cge ws designed to collect ny excrement from the dog to reduce confounding odours t the site, the dog did not urinte or defecte during ny of the trils. We did not use got or other domestic niml s control for the live dog tretment becuse their ssocition with humns (who use dogs for herding nd hunting) would be confounding fctor. Further, we did not use n innimte model (s in Scheinin et l. 2006) becuse the novelty of the object could lso potentilly confound the experiment. Thus n inherent ssumption of the live dog tril is tht chnges in fox behviour re result of the dog being recognized s thret (interference competitor/predtor) by the fox, rther thn response to the novelty of the dog. For the niml odour tril, we reversed the position of the tretment nd control t the food trys (Fig. 1) to ensure tht specific try ws not ssocited with dog stimuli. The odour tretment ws fresh (<24 hr old) dog sct nd cotton bll soked with fresh (<2 hr old) dog urine, nd the odour control ws fresh (<24 hr old) dung from blckbuck Antilope cervicpr nd wter-soked cotton bll. Blckbuck re common in the study re nd re presumbly non-thretening ntive herbivore. Odours were plced directly on the ground 20 m from the try. During ech 2-hour tril period, we recorded the presence of pups t the den site nd t trys nd ll occurrences nd durtion of the following behviours by dult foxes: time of first ppernce of t lest one fox t the site, number of visits to ech try, nd time spent t ech try irrespective of the number of foxes tht visited the try. We lso recorded the durtion of vigilnce behviours by ny dult fox nd durtion of non-vigilnt behviours by ll foxes. Vigilnce behviours were defined s lert postures, lrm brks, nd gitted movements, while non-vigilnce included grooming, sleeping, resting, plying, nd suckling. For sttisticl nlyses, we grouped these behviours in the two ctegories becuse not ll foxes exhibited ll behviours (for exmple, mles did not suckle nd pups did not lrm cll). We combined the durtions of ech behviour exhibited by ll foxes t ech den site s mesure of behviourl occurrences per den site.

5 Behv Ecol Sociobiol (2009) 64: For exmple, if more thn one fox ws present t try, we only recorded tht s one event nd did not tlly the time spent by ll individul foxes present concurrently t the try. We ended trils 1 hour fter sunset nd mesured the remining food present in ech try. Ech set of trils took between 5 6 dys to complete, nd we tested the behviourl responses of 13 dult foxes nd their pups t seven den sites. Dt nlysis All behviourl mesures met the ssumptions of normlity s determined by Kolmogorov Smirnov tests, nd ssumptions of sphericity were met ccording to Muchly s sphericity test. For ech tril (neutrl, live dog, nd niml odour), we used pired t test to determine whether the mount of food eten nd number of visits by foxes differed between control nd tretment trys. There ws no significnt difference in the mount of eggs nd grpes consumed during ech tril so these food types were combined for nlysis. We clculted the difference between the number of visits to nd the mount of food eten from the control try nd tretment try for ech tril (neutrl, live dog, nd niml odour) nd used Repeted Mesures Anlysis of Vrince (RM ANOVA) to test for the effect of tril on these differences. We lso tested whether foxes differed in the time they spent t control nd tretment trys during ech tril using pired t tests. We quntified the durtion of vigilnce nd non-vigilnce behviours nd totl time spent t both trys, s proportions of totl tril durtion to correct for the differences in observtion periods between trils. Using RM ANOVA, we lso determined the effect of tril (neutrl, live dog, nd niml odour) on lloction of time spent t trys nd in vigilnce nd non-vigilnce behviours. All sttisticl nlyses were conducted using SPSS 15.0 for Windows. Results The number of foxes tht visited the trys did not differ mong trils (F 2, 12 =1.5, P=0.26; men ± SE=2.62± 0.016). Pups were present nd ctive t ll den sites during the trils nd mostly visited the trys with dults, except for two cses when pups independently visited try nd two cses when they did not visit the trys t ll. During the neutrl nd odour trils, there ws no difference in which try ws first visited. Foxes did investigte both the odour stimuli t four of the seven sites by sniffing or licking the urine-soked cotton bll (tretment) nd wter-soked cotton bll (control). In contrst to the neutrl nd odour trils, the first try visited by ny fox during the dog tril ws lwys the control try with the empty cge. During the dog trils, the ltency to rrive t ny try by ny fox (time first rrivl to time try set up) ws longer (men=87.43±6.08 SE) thn the ltencies during the neutrl (men=59.14±4.31 SE) nd odour trils (men=57.71±9.74 SE; F 2, 12 =7.78, P=0.007). There ws no difference in the proportion of food eten between the two trys during the neutrl tril (t= 1.903, P=0.106; Fig. 2), thus, indicting tht initil forging preference for try ws lost fter the hbitution period. Proportion food eten t trys Number of visits to food trys Time spent t food trys (min) b c Neutrl Live dog Odours Neutrl Live dog Odours * Neutrl Live dog Odours Fig. 2 Proportion of food consumed by foxes t control (open brs) nd tretment (shded brs) trys during the neutrl, live dog, nd niml odour trils. b Number of visits by foxes to control (open brs) nd tretment (shded brs) trys during the neutrl, live dog, nd odour trils. c Amount of time spent by foxes t control (open brs) nd tretment (shded brs) trys during the neutrl, live dog, nd odour trils. For ll figures, the sterisk mrk (*) denotes significnt difference between trys t p<0.05 (men±se, n=7 dens) * * *

6 284 Behv Ecol Sociobiol (2009) 64: There ws lso no difference in the proportion of food eten between the two trys during the odour tril (t=0.801, P=0.454; Fig. 2). In contrst, foxes significntly reduced food consumption t the live dog try compred to the empty cge try (t=9.033, P=0.0001; Fig. 2). The men difference in proportion of food consumed between control nd tretment trys ws significntly greter during the live dog tril thn during the neutrl nd niml odour trils (F 2, 12 =21.89, P<0.0001). Foxes did not completely void the live dog try, but rther, they mde significntly fewer visits compred to the empty cge try (t=2.856, P=0.029; Fig. 2b). There ws no difference in the number of visits between trys during the neutrl (t= 2.295, P=0.062) nd niml odour (t=1.291, P=0.244) trils (Fig. 2b). Consequently, the men difference in the number of visits between control nd tretment trys ws significntly greter during the live dog tril thn during the neutrl nd niml odour trils (F 2, 12 =8.3, P=0.006). Foxes did not compenste for the decrese in the number of visits nd proportion of food consumed t the live dog try by incresing these behviours t the control try during the live dog tril s their responses t control trys were similr for ll three trils. During the neutrl tril, foxes spent mrginlly less time t the non-preferred try compred to the preferred try (t= 2.408, P=0.053; Fig. 2c), nd thus, we ssigned the preferred try to the tretment condition during the following dy s dog tril. Foxes spent significntly less time t the try within sight of the cged dog compred to the empty cge try (t=3.258, P=0.017; Fig. 2c).There ws no significnt difference in time spent t the two trys Proportion of time spent Neutrl Live dog Odours b Non-vigilnce Vigilnce At food trys Fig. 3 Proportion of time spent by foxes t the food trys nd in vigilnce nd non-vigilnce behviours during the neutrl, live dog, nd niml odour trils. Letters denote significnt difference (Fisher s protected lest squres difference p<0.05) mong trils for ech behviour (mens±se, n=7 dens) b b,c,c during the following dy s odour tril (t=0.754, P=0.48; Fig. 2c). As expected, the men difference in the time spent between control nd tretment trys ws significntly greter during the live dog tril thn during the neutrl nd niml odour trils (F 2, 12 =7.7, P=0.007). For ll three trils, time spent on vigilnce behviours by foxes ws negtively correlted with the mount of food eten t the food trys (Person s r 2 = 0.59, P=0.005, n=21). There were significnt differences mong the trils in the totl proportion of time spent t both food trys (F 2, 12 =11.13, P=0.0018) nd in non-vigilnce (F 2, 12 = , P=0.0003) nd vigilnce (F 2, 12 =55.24, P<0.0001) behviours by foxes (Fig. 3). Foxes were most vigilnt during the dog tril (men±se=30.35±4.87 min) compred to the neutrl (men±se=2.49±0.6 min) nd niml odour (men±se=5.1±1.26 min) trils (Fig. 3). Consequently, foxes significntly reduced time spent on nonvigilnt behviours during the dog tril (men±se=5.84± 1.27 min) compred to the neutrl (men±se=14.37± 3.22 min) nd niml odour (men±se=13.9±2.98 min) trils (Fig. 3). Similrly, foxes significntly reduced time t the food trys during the dog tril (men±se=9.6± 2.81 min) when compred to the neutrl tril (men±se= 17.17±3.3 min; Fig. 3), but not to the niml odour tril (men±se=14.41±3.22 min; Fig. 3). For ll three responses, there ws no significnt difference in how time ws llocted between the neutrl nd niml odour trils (Fig. 3). Discussion This experiment demonstrtes the profound effect dogs cn hve on the ctivity nd forging behviour of Indin foxes. When dog ws visible, foxes reduced their visittion rtes nd spent less time t the food trys. Not only did the presence of live dog result in less food consumed t the tretment try, but foxes consumed less food overll compred to the neutrl nd niml odour trils. Foxes were more vigilnt when dog ws visible, exhibiting lert postures, gitted movements, nd lrm clling. Furthermore, foxes rrived much lter t food trys, spent less time forging t the trys nd reduced other non-vigilnce ctivities such s resting, plying, nd grooming. The presence of dog odours (fresh urine nd sct) hd little effect on fox forging nd ctivity. These results re similr to those reported by Scheinin et l. (2006) where the presence of live jckl, but not jckl odours, elicited chnge in red fox forging nd behviour. Although unlikely, it is possible tht foxes could not distinguish between the odours of dogs nd blckbucks, s their behviourl responses t both food trys were similr. However, dogs re present t very high

7 Behv Ecol Sociobiol (2009) 64: densities in the study re, nd foxes re likely to often encounter dog sct nd urine. Therefore, the presence of dog odours is unlikely to be novel or prticulrly thretening stimulus. Since olfction is n importnt component of cnid communiction (Gormn nd Trowbridge 1989), it is possible tht dog nd blckbuck odours my provide some informtion which would explin why foxes investigted the odours during the trils. In ny cse, the reduction of forging time during the odour trils compred to the neutrl trils ws mrginl (men<1 minute). Overll, foxes re not ffected by the indirect indictors of dog presence (urine nd sct), but insted rect to the potentilly more direct thret posed by ctul dog presence. This response is expressed s reduction of food intke nd n increse in vigilnce. Vigilnce hs been observed in other crnivores, including coyote (Switlski 2003), cheeth, Acinonyx jubtus (Hunter et l. 2007), dwrf mongoose, Helogle undulte, (Rs 1989) nd estern quoll, Dsyurys viverrinus (Jones 1998). In ech of these cses, there ws lso lrge trde-off between vigilnce behviour nd forging. It is likely tht the energetic costs in terms of loss of forging opportunity during the dog trils might be high for the foxes, especilly, when rising pups t time of generl food scrcity. The combined effects of reducing forging, incresing vigilnce, nd lowering other nonvigilnt ctivity (such s ply behviour, grooming, nd suckling) when under thret my hve importnt fitness nd survivl consequences for these individuls (Lim nd Dill 1990). The responses of Indin foxes to dogs were similr to the rections of red foxes to golden jckls described by Scheinin et l. (2006). In our study re, golden jckls occur t low densities but they likely compete for the sme food resources, such s rodents nd Zizyphus berries, s foxes (Home 2005; Mukherjee et l. 2004). Such dietry overlp could be mjor driver of interference competition between those two species. However, dogs nd foxes do not compete for food resources t this study re (Vnk nd Gompper 2009), but dogs pper to dominte foxes vi interference interctions. Yet, it is importnt to note tht foxes did not completely void the food trys when dogs were visible. Rther, they modified their behviour while still pproching the dog nd the food. Given tht in other systems, subordinte competitors use diverse strtegies to reduce encounters with dominnt competitors (Atwood nd Gese 2008; Creel et l. 2001; Nelson et l. 2007), we believe our dt reflects similr competitor voidnce scenrio: foxes trde-off forging for vigilnce t rich resources ptches when dogs re present. These results suggest tht foxes view dogs s lrger member of the cnid guild nd thus, s n interference competitor. Dogs my, therefore, fulfill the role of mid-sized cnid in this community despite their close ssocition nd dependence on humns. The broder effects of interference interctions between dogs nd other crnivores hve been observed in t lest one system. In Austrli, non-ntive red foxes voided res tht were recently visited by wild dogs (dingoes, ferl dogs, nd their hybrids), lthough, they re symptric with dogs t the lndscpe level (Mitchell nd Bnks 2005). There, in the bsence of lrge mmmlin crnivores, dingoes nd ferl dogs hve ssumed the role of mid-sized pex predtor in their interctions with both introduced eutherin nd ntive mrsupil crnivores (Glen et l. 2007; Johnson et l. 2007). Domestic dogs re clssic exmple of subsidized predtor (Bormn 2003; Kys nd DeWn 2004). They re insulted from food scrcity becuse of humn-derived food subsidies nd re generlly buffered from intrguild feedbck mechnisms (e.g., predtion by lrger crnivores such s wolves nd leoprds) by the presence of sfe refuges in or ner humn hbittions. Such scenrio is problemtic for conservtion prctitioners since in some regions the enhnced dog popultion likely influence species of conservtion. Although Indin foxes re not species of conservtion concern in Indi (Sillero-Zubiri et l. 2004), the results from this study cn be extrpolted to endngered smll crnivores since the underlying mechnisms nd consequences of intrguild competition likely remin the sme. In such cses, the interference competition from dogs could result in lower popultion sizes of symptric crnivores which in turn might increse the likelihood of locl extirption (Cypher et l. 2001; Linnell nd Strnd 2000). Acknowledgements This study ws funded through grnts from the University of Missouri Reserch Bord (M.E. Gompper) nd the Wildlife Conservtion Society Reserch Fellowship Progrm (A.T. Vnk). This study ws pproved by the Institutionl Animl Cre nd Use Committee of the University of Missouri (reference number 4265) nd complies with lws governing reserch in Indi. We thnk Mr. B. Mjumdr (wildlife Chief Wrden) nd the Mhrshtr Stte Forest Deprtment for providing necessry logistic help nd permits (number 22(8)/Reserch/2563/ ). We re especilly thnkful to A. Kulkrni, S. Mhmne, nd V. Shitole for their help in crrying out the reserch. We lso thnk three nonymous reviewers for helpful comments on n erlier version of this mnuscript. References Atickem A (2003) The ecology of domestic dog (Cnis fmiliris) nd its effect on the survivl of the Ethopin wolf (Cnis simensis) in the Web Vlley of the Ble Mountins Ntionl Prk, Ethopi. MS Thesis, Addis Abb University Atwood TC, Gese EM (2008) Coyotes nd recolonizing wolves: socil rnk medites risk-conditionl behviour t ungulte crcsses. Anim Behv 75:

8 286 Behv Ecol Sociobiol (2009) 64: Bormn WI (2003) Mnging subsidized predtor popultion: reducing common rven predtion on desert tortoises. Environ Mng 32: Boitni L, Frncisci F, Ciucci P, Andreoli G (1995) Popultion biology nd ecology of ferl dogs in centrl Itly. In: Serpell J (ed) The domestic dog. Its evolution, behviour, nd interctions with people. Cmbridge University Press, Cmbridge, UK, pp Butler JRA, du Toit JT (2002) Diet of free-rnging domestic dogs (Cnis fmiliris) in rurl Zimbbwe: implictions for wild scvengers on the periphery of wildlife reserves. Anim Conserv 5:29 37 Butler JRA, du Toit JT, Binghm J (2004) Free-rnging domestic dogs (Cnis fmiliris) s predtors nd prey in rurl Zimbbwe: threts of competition nd disese to lrge wild crnivores. Biol Conserv 115: Cmpos CB, Esteves CF, Ferrz K, Crwshw PG, Verdde LM (2007) Diet of free-rnging cts nd dogs in suburbn nd rurl environment, southestern Brzil. J Zool 273:14 20 Cro TM, Stoner C (2003) The potentil for interspecific competition mong Africn crnivores. Biol Conserv 110:67 75 Cse TJ, Gilpin ME (1974) Interference competition nd niche theory. Proc Ntl Acd Sci USA 71: Creel S, Spong G, Creel N (2001) Interspecific competition nd the popultion biology of extinction prone crnivores. In: Gittlemn JL, Funk SM, Mcdonld DW, Wyne RK (eds) Crnivore conservtion. Cmbridge University Press, Cmbridge, UK, pp Cypher BL, Spencer KA (1998) Competitive interctions between coyotes nd Sn Joquin kit foxes. J Mmml 79: Cypher BL, Clrk HO Jr, Kelly PA, Job CVH, Wrrick GW, Willims DF (2001) Interspecific interctions mong wild cnids: implictions for the conservtion of endngered Sn Joquin kit foxes. Endnger Species Updte 18: Dniels TJ, Bekoff M (1989) Popultion nd socil biology of free rnging dogs, Cnis fmiliris. J Mmml 70: Dondio E, Buskirk SW (2006) Diet, morphology, nd interspecific killing in Crnivor. Am Nt 167: Edgonkr A, Chellm R (2002) Food hbit of the leoprd, Pnther prdus, in the Snjy Gndhi Ntionl Prk, Mhrshtr, Indi. Mmmli 66: Elmhgen B, Tnnerfeldt M, Angerbjorn A (2002) Food-niche overlp between rctic nd red foxes. Cn J Zool 80: Fedrini JM, Fuller TK, Suvjot RM, York EC (2000) Competition nd intrguild predtion mong three symptric crnivores. Oecologi 125: Fiorello CV, Noss AJ, Deem SL (2006) Demogrphy, hunting ecology, nd pthogen exposure of domestic dogs in the Isoso of Bolivi. Conserv Biol 20: Glen AS, Dickmn CR (2005) Complex interctions mong mmmlin crnivores in Austrli, nd their implictions for wildlife mngement. Biol Rev 80: Glen AS, Fy AR, Dickmn CR (2006) Diets of symptric red foxes Vulpes vulpes nd wild dogs Cnis lupus in the Northern Rivers Region, New South Wles. Aust Mmml 28: Glen AS, Dickmn CR, Soulé ME, Mckey BG (2007) Evluting the role of the dingo s trophic regultor in Austrlin ecosystems. Austrl Ecol 32: Gompper ME, Vnk AT (2006) Vulpes benglensis. Mmm Species 795:1 5 Gormn ML, Trowbridge BJ (1989) The role of odor in the socil lives of crnivores. In: Gittlemn JL (ed) Crnivore behvior, ecology, nd evolution. Cornell University Press, NewYork, pp Gosselink TE, Vn Deelen TR, Joselyn MG, Wrner RE (2003) Temporl hbitt prtitioning nd sptil use of coyotes nd red foxes in est-centrl Illinois. J Wildl Mng 67: Green JS, Gipson PS (1994) Ferl dogs. In: Hygnstrom SE, Timm RM, Lrson GE (eds) The hndbook: prevention nd control of wildlife dmge. University of Nebrsk, Lincoln, pp 1 7 Hbib B (2007) Ecology of Indin wolf (Cnis lupus pllipes Sykes, 1831), nd modeling its potentil hbitt in the Gret Indin Bustrd Snctury, Mhrshtr, Indi. Ph.D. Thesis, Aligrh Muslim University Holt RD, Polis GA (1997) A theoreticl frmework for intrguild predtion. Am Nt 149: Home C (2005) Resource utiliztion by the Indin fox, Vulpes benglensis in Kutch, Gujrt. M.S. Thesis, Surshtr University Hunter JS, Durnt SM, Cro TM (2007) To flee or not to flee: predtor voidnce by cheeths t kills. Behv Ecol Sociobiol 61: Johnson WE, Fuller TK, Frnklin WL (1996) Symptry in cnids: A review nd ssessment. In: Gittlemn JL (ed) Crnivore behvior, ecology, nd evolution, vol 2. Cornell University, Ithc, New York, pp Johnson CN, Isc JL, Fisher DO (2007) Rrity of top predtor triggers continent-wide collpse of mmml prey: dingoes nd mrsupils in Austrli. Proc R Soc Lond B Biol Sci 274: Jones ME (1998) The function of vigilnce in symptric mrsupil crnivores: the estern quoll nd the Tsmnin devil. Anim Behv 56: Kts L, Dill L (1998) The scent of deth: chemosensory ssessment of predtion risk by prey nimls. Ecoscience 5: Kys RW, DeWn AA (2004) Ecologicl impct of inside/outside house cts round suburbn nture preserve. Anim Cons 7: Kruuk H, Snell H (1981) Prey selection by ferl dogs from popultion of mrine iguns (Amblyrhynchus cristtus). J Appl Ecol 18: Lim SL, Dill LM (1990) Behviourl decisions mde under the risk of predtion: review nd prospectus. Cn J Zool 68: Linnell JDC, Strnd O (2000) Interference interctions, co-existence nd conservtion of mmmlin crnivores. Divers Distrib 6: Mitchell BD, Bnks PB (2005) Do wild dogs exclude foxes? Evidence for competition from dietry nd sptil overlps. Austrl Ecol 30: Mukherjee S, Goyl SP, Johnsingh AJT, Pitmn MRPL (2004) The importnce of rodents in the diet of jungle ct (Felis chus), crcl (Crcl crcl) nd golden jckl (Cnis ureus) in Srisk Tiger Reserve, Rjsthn, Indi. J Zool 262: Muller-Schwrze D (2006) Chemicl ecology of vertebrtes. Cmbridge University Press, Cmbridge, UK Nelson JL, Cypher BL, Bjurlin CD, Creel S (2007) Effects of hbitt on competition between kit foxes nd coyotes. J Wildl Mng 71: Plomres F, Cro TM (1999) Interspecific killing mong mmmlin crnivores. Am Nt 153: Rlls K, White PJ (1995) Predtion on Sn Joquin kit foxes by lrger cnids. J Mmml 76: Rs OAE (1989) The costs nd effectiveness of vigilnce behviour in the dwrf mongoose: implictions for fitness nd optiml group size. Ethol Ecol Evol 1: Scheinin S, Yom-Tov Y, Motro U, Geffen E (2006) Behviourl responses of red foxes to n increse in the presence of golden jckls: field experiment. Anim Behv 71: Sillero-Zubiri C, Hoffmnn M, Mcdonld DW (2004) Cnids: foxes, wolves, jckls nd dogs. Sttus survey nd conservtion ction pln. IUCN/SSC Cnid Specilist Group, Glnd, Switzerlnd nd Cmbridge, UK

9 Behv Ecol Sociobiol (2009) 64: St-Pierre C, Ouellet JP, Crete M (2006) Do competitive intrguild interctions ffect spce nd hbitt use by smll crnivores in forested lndscpe? Ecogrphy 29: Switlski TA (2003) Coyote forging ecology nd vigilnce in response to gry wolf reintroduction in Yellowstone Ntionl Prk. Cn J Zool 81: Tnnerfeldt M, Elmhgen B, Angerbjorn A (2002) Exclusion by interference competition? The reltionship between red nd rctic foxes. Oecologi 132: Vnk AT (2008) Intrguild interctions between ntive nd domestic crnivores in centrl Indi. Ph.D. Disserttion, University of Missouri Vnk AT, Gompper ME (2007) Effectiveness of non-invsive techniques for surveying ctivity nd hbitt use of the Indin fox Vulpes benglensis in southern Indi. Wildl Biol 13: Vnk AT, Gompper ME (2009) Dietry niche seprtion between symptric free-rnging domestic dogs nd Indin foxes in centrl Indi. J Mmml 90 (in press) Vnk AT, Gompper ME (2009b) Dogs Cnis fmiliris s crnivores their role nd function in intrguild competition. MmmlRev39(inpress) Vnk AT, Gompper ME (in press) Multi-scle resource selection nd sptil ecology of the Indin fox in humn-dominted dry grsslnd ecosystem. J Zool Lond Wndeler AI, Mtter H-C, Kppeler A, Budde A (1993) The ecology of dogs nd cnine rbies: selective review. Revue Scientifique et Technique, Office Interntionle des Epizootics 12:51 71 White PJ, Grrott RA (1997) Fctors regulting kit fox popultions. Cn J Zool 75: WHO/WSPA (1990) Guidelines for dog popultion mngement. World Helth Orgniztion, Genev