Spatial Partitioning of Predation Risk in a Multiple Predator‐Multiple Prey System

Abstract

Minimizing risk of predation from multiple predators can be difficult, particularly when the risk effects of one predator species may influence vulnerability to a second predator species. We decomposed spatial risk of predation in a 2-predator, 2-prey system into relative risk of encounter and, given an encounter, conditional relative risk of being killed. Then, we generated spatially explicit functions of total risk of predation for each prey species (elk [Cervus elaphus] and mule deer [Odocoileus hemionus]) by combining risks of encounter and kill. For both mule deer and elk, topographic and vegetation type effects, along with resource selection by their primary predator (cougars [Puma concolor] and wolves [Canis lupus], respectively), strongly influenced risk of encounter. Following an encounter, topographic and vegetation type effects altered the risk of predation for both ungulates. For mule deer, risk of direct predation was largely a function of cougar resource selection. However, for elk, risk of direct predation was not only a function of wolf occurrence, but also of habitat attributes that increased elk vulnerability to predation following an encounter. Our analysis of stage-based (i.e., encounter and kill) predation indicates that the risk effect of elk shifting to structurally complex habitat may ameliorate risk of direct predation by wolves but exacerbate risk of direct predation by cougars.

Separating spatial search and efficiency rates as components of predation risk

“…While predator-based studies such as this one have revealed spatial variation in multiple stages of predation such as search and kill rates [26], prey-based studies have also revealed differential avoidance and escape tactics as prey-mediated components of spatial risk [7]. Spatial models of predation risk become increasingly difficult to generalize as one acknowledges the potential for behavioural variation in both predators and prey among species [7,27], among individuals within species [28 -30] and among behavioural states within an individual [31]. Rather than a model of risk for any given prey species or individual, I estimated a multi-species composite model of risk for the average prey individual killed by wolves in my study area.…”

“…Rather than a model of risk for any given prey species or individual, I estimated a multi-species composite model of risk for the average prey individual killed by wolves in my study area. The net spatial pattern of risk with this treatment inherently includes the interacting behavioural mechanisms of predator selection of prey species [32] and the spatial avoidance behaviour and vulnerabilities of prey [27]. This approach tests for links between spatial heterogeneity and predator responses, though such links are undoubtedly mediated by complex and multi-species behavioural games between both predators and prey [7].…”

A “death trap” in the landscape of fear

“…This concept may particularly be true in multipredator systems, especially if there are both stalking and cursorial predators. The risk effect of one predator, which would eventually shift the prey into Bsafer^habitat, should increase the predation risk from another predator, as it was recently evidenced by Atwood et al (2009) for a wolf-cougar-elkmule deer (Odocoileus hemionus) system.…”

Predators choose prey over prey habitats: evidence from a lynx–hare system

“…For example, studies on mule deer (Odocoileus hemionus [Atwood et al 2009]), elk (Cervus elaphus [Hebblewhite et al 2005, Atwood et al 2009, caribou (Rangifer tarandus [Gustine et al 2006]), Eurasian lynx (Lynx lynx [Odden et al 2008]), Stone's sheep (Ovis dalli stonei [Walker et al 2007]), and bottlenosed dolphin (Torsiops aduncus [Heithaus and Dill 2002]) have explicitly assumed a consistent relationship between prey habitat (e.g., resource selection by prey) and resource selection by predators based on food availability, or between predator habitat and resource selection by prey based on avoidance. Other studies, focusing on species such as Atlantic bluefin tuna (Thunnus thynnus [Teo et al 2007]), White-eared and Blood Pheasants (Crossoptilon crossoptilon and Ithaginis cruentus [Jia et al 2005]), Spotted Owls (Strix occidentalis [Irwin et al 2007]), macropods (While and McArthur 2005), Pileated Woodpeckers (Dryocopus pileatus [Lemaıˆtre and Villard 2005]), American marten (Martes americana [Slauson et al 2007]), and dolphins (Cephalorhynchus hectori [Bra¨ger et al 2003]), have relied solely on a ''standard of plausibility'' (Lima and Zollner 1996), using environ- (Charnov 1976) and predation risk predicts that, all things being equal, prey should spend less time in areas that receive greater amounts of habitat use by predators and predators should spend more time in areas that receive greater amounts of habitat use by prey (Brown 1988, Lima andDill 1990).…”