Research ArticlePartitioning of Anthropogenic Watering Sitesby Desert CarnivoresTODD C. ATWOOD,1United States Department of Agriculture, Wildlife Services, National Wildlife Research Center, Fort Collins, CO 80521,USATRICIA L. FRY, United States Department of Agriculture, Wildlife Services, National Wildlife Research Center, Fort Collins, CO 80521, USABRUCE R. LELAND, United States Department of Agriculture, Wildlife Services, San Antonio, TX 78201, USAABSTRACT We investigated the role of water features as focal attractors for gray foxes (Urocyon ciner-eoargenteus), coyotes (Canis latrans), and bobcats (Felis rufus) in west Texas to determine if they were foci forinterspecific interaction. Mixed effects models indicated that species partitioned use of water featuresspatially and temporally. Linear models indicated factors influencing relative activity at water features variedby species. For coyotes and bobcats, the water availability model, containing days since last rainfall andnearest-neighbor distance to water was best supported by the data, with relative activity increasing with timebetween rainfall and distance between waters. For gray foxes, the best approximating model indicated tha trelative activity was inversely correlated to coyote and bobcat activity indices , and positively correlated totopographical complexity. Encounters between carnivore species were low, with most occurring betweencoyotes and gray foxes, followed by coyotes and bobcats, and bobcats and gray foxes. These findings suggest abehavioral-environmental mechanism that may function to modulate resource partitioning by carnivores inthe arid West. ß 2011 The Wildlife Society.KEY WORDS carnivore, Chihuahuan desert, competition, interaction, resource partitioning, water.Interspecific competition is a frequently occurring ph enom-enon (Connell 1983) that can play a determinative role ingoverning community structure where species exploit limitedresources (Schoener 1983, Chase et al. 2002). Amongcarnivores, interspecific interactions typically are asymmet-rical, with one species having a stronger competitive effectthan the other (Atwood and Gese 2008). This asymmetrymost often is influenced by differences in body size (Poliset al. 1989, Crooks and Van Vuren 1995, Davies et al. 2007)with larger species benefiting from a direct size effect byincurring lower costs from interspecific interactions thansmaller species. For subordinate competitors, the risk ofinteraction may motivate spatial and/or temporal segregationof a contested resour ce in an attempt to avoid direct con-frontation (e.g., Mills and Gorman 1997, Valeix et al. 2007)or, in extreme instances, intraguild predation (Palomares andCaro 1999). The latter has been well doc umented amongcanids (e.g., Johnson et al. 1996), and illustrates the risk ofsharing space with fierce competitors. In order to coexist,subordinate competitors must develop behaviors to mitigatethe risk of interspecific strife.Avoidance of a dominant competitor is one means by whicha subordinate competitor can reduce the effects of interspe-cific strife. Generally, avoidance behavior is manifest throughspatial or temporal partitioning of the shared niche(Carothers and Jaksic 1984, Durant 2000). Under thisconstruct, a subordinate competitor may display adaptiveresource selection by altering spatial (Atwood and Gese2008) or temporal (Kotler et al. 1993, Atwood and Gese2010) patterns of resource use based on the presence orabsence of a dominant competitor. The success of suchbehavior is influenced, to a large extent, by heterogeneityin the distribution and quality of resource patches (Palmer2003). For example, Mills and Gorman (1997) concludedthat African wild dogs (Lyca on pictus) minimized the timethey spent in prey-rich areas to avoid interactions with lions(Panthera leo), the primary source of wild dog mortality.Similarly, Thompson and Gese (2007) found that swift foxes(Vulpes velox) were less abundant in structurally complexhabitats where prey and coyotes (Canis latrans), the primarysource of mortality for foxes, were most abundant. Clearly,partitioning a resource is an effective strategy for avoidingstrife when that resource is spatially variable. However, whena resource is spatially fixed, it is likely that aggregation willlead to interference behavior (Vahl et al. 2005) includingaggressive interactions that result in direct contact.The development of water features, such as stock tanks andartificial catchments, has become an integral part of livestockand wildlife management in the arid western United Stateswhere surface water often is absent or only seasonally avail-able (e.g., Hervert and Krausman 1986, Holechek 1992).However, despite the obvious benefits, water features alsohave the potential to become foci of interspecific strife, andas competition for access intensifies so does the probabilityof contact between contestants (Thrash et al. 1995). Inthat way, resource-mediated contact may contribute to theReceived: 8 October 2010; Accepted: 6 March 2011;Published: 3 August 20111E-mail: [email protected] Journal of Wildlife Management 75(7):1609–1615; 2011; DOI: 10.1002/jwmg.225Atwood et al.Partitioning of Water by Carnivores 1609maintenance of infectious disease (Barlow 1996). The mostimportant diseas es of carnivores (e.g., rabies, canine distem-per, canine parvovirus) are caused by generalist pathogenscapable of infecting multiple species and with a high poten-tial for cross-species transmission (Nel et al. 1997, Kauhalaand Holmala 2006, Roemer et al. 2009). Communal use ofscarce resources, and the attendant direct and indirectcontact that is likely to occur when attempting to exploitor defend those resources, may exacerbate the potential forcross-species pathogen transmission. Accordingly, examin-ing the behavioral mechanisms underlying interspecificinteraction at water features will provide a broader under-standing of the connection between resource partitioningand the management of infectious agents in carnivorecommunities.Coyotes, bobcats (Felis rufus), and gray foxes (Urocyoncinereoargenteus) are sympatric over much of their geographicrange (Whitaker and Hamilton 1998), prey mainly on smallmammals (Leopold and Krausman 1986, Neale and Sacks2001a), are habitat generalists (Riley 2006, Neale and Sacks2001b), and share susceptibility to many of the same patho-gens (Riley et al. 2004). As a result, there is great potentialfor