The ‘Landscape of Fear’ concept
Mammalian predators often do not to use their home range or territory uniformly, and thus areas of high activity (core areas) would be expected to be associated with a high density of kairomones and other predatory chemical cues (Atkins et al. 2019). Prey individuals, therefore, should avoid such areas and may be able to assess relative predation risks in different areas (Atkins et al. 2019). Conversely, the removal or reduction of predation risk may enable the dispersal of prey populations; declines in Persian leopards (Panthera pardus ) and African hunting dogs (Lycaon pictus ) allowed the recovery and distribution of Cape bushbuck (Tragelaphus sylvaticus ) in Mozambique (Atkins et al.2019). Relationships such as these, however, have formed across multiple generations of predator and prey species and, as such, alterations may not result in observable ecological changes immediately; a 50 to 130 year absence of predators may result in the inability of prey to identify recolonizing predators as a threat (Berger et al. 2001). Such phenomena have given rise to the concept of a “landscape of fear” (LOF), where prey may reduce their temporal and spatial exposure to core predator areas in relation to perceived predation risk, although a variety of other biological and evolutionary factors besides fear also influence habitat use (Laundré et al. 2001; Laundré et al.2010; Bleicher 2017).
As a prerequisite for prey to actively avoid predation threats, it must be possible to identify and associate predatory cues with predation risk (Griffith 1920; Schaller 1974; Mech 1970: Table 1). In some cases, the ability to identify scents specific to a predator species may be advantageous; foraging beavers (Castor fiber ) show significantly fewer defensive behaviours in response to the odours of dogs (Canis familiaris ) compared to those of wolves (Canis lupus ) (Rosell & Czech 2000); grey kangaroos (Marcopus fuliginosis ) can discriminate the urine of coyotes (Canis latrans ) from dingoes (Canis lupus dingo ) (Parsons et al.2007); mule deer (Odocoileus hemionus ) respond to the faecal remnants of coyotes and mountain lion (Felis concolor ), but not those of African lions (Panthera leo ), leopards, or tigers (Panthera tigris ) (Mulller-Schwarze 1972). The type of kairomone left by a predator is often species-specific; weasel (Mustela nivalis ) tracks are rich in sulphur compounds, while those of beech martens (Martes foina ) contain no sulphur-rich components (Schildknecht & Birkner 1983; Apfelbach et al. 2015). The ability to detect kairomones also differs between prey species; the abundance of trace amine-associated receptors (TAARs), a group of olfactory receptors used to detect kairomone presence, differs between prey species (Borowsky et al. 2001; Ferrero et al. 2011). Rodents may possess up to seventeen TAARs, while primates only six (Borowsky et al. 2001; Ferrero et al. 2011). The ability to differentiate between predatory threats and benign, albeit from closely related predators, chemical cues may reduce the energetic expenditure spent on vigilance and evasion.
Furthermore, the extent of predation risk likely varies, and prey species should possess the ability to distinguish between the least threatening and more ominous indicators. These abilities are often demonstrated in the occurrence of “giving up distributions” (GUDs), with the emphasis on whether prey animals sacrifice foraging behaviour in relation to the perceived risk of predation (Brown 1988; Brownet al. 1994; Laundré et al. 2001). For instance, moose (Alces alces ) quit more than half of experimental areas following repeated exposure to urine from wolf and grizzly bears (Ursus arctos ) (Pyare & Berger 2003). More specifically, a forager should terminate feeding in a food patch when the value of its harvest rate (H) no longer exceeds the sum of the energetic cost of foraging (C), predation risk (P) and missed opportunity cost (MOC): H=C+P+MOC (Brown 1988), and has been shown to apply in a range of taxa under varying ecological conditions (Stokes et al. 2004; Merwe & Brown 2008; Shrader et al. 2008; Valeix et al. 2009; Iribarren & Kotler 2012; Clinchy et al. 2013; Gallagher et al. 2017).