Predation is a strong evolutionary force acting as a selective pressure on various traits within a prey species including, but not limited to, size, morphology, coloration, development of sexual maturity, and patterns of behaviour (Werner & Anholt 1993; Maynard-Smith & Harper 2003; Scott-Phillips 2008). Ultimately, the distribution, abundance, community composition, size, diversity, breeding success and ecosystem functioning of one or more species are impacted (Hall et al. 1976; Berger et al. 2001; Kneitel & Chase 2004; Beauchamp et al. 2007; Hawlena & Schmitz 2010; Letnic et al. 2011). Interactions between predators and prey can over time make communities vulnerable to biotic and/or abiotic perturbations (Mooney et al. 2010), including the invasion of non-native species as a consequence of human activity (Snyder et al. 2004). At the most general level, predation results in the reduced survival of prey (consumptive effects), but the (real or perceived) risk of predation can also affect prey animals in a variety of ways (Zanetteet al. 2011), including stinted growth, poor reproductive output and reduced foraging efficiency (Sih 1980; Lima & Dill 1990; Peckarskyet al. 1993; Abrams et al. 1996; Peacor & Werner 1997). Individuals most capable of evading predators have an increased potential for producing offspring, and therefore should achieve higher reproductive success than less capable conspecifics; female cheetahs (Acinonyx jubatus ) with higher fecundity avoided habitats with indications of elevated spotted hyaena (Crocuta crocuta ) or lion (Panthera leo ) densities more than females with lower fecundity (Durant 2000).
In response to the severity of the risks posed by predators, prey species have evolved a wide range of anatomical, physiological, neurobiological and behavioural mechanisms to increase their chances of survival. Adaptations related to morphology and physiology are particularly energetically expensive, and thus prey may display different responses to a threat depending on how severe they perceive it to be, known as the threat-sensitive predator avoidance hypothesis (Chivers et al. 2001; Monclús et al. 2009).
Some adaptations reduce the chance of initial detection by a predator including: camouflage (Cott 1940; Ruxton et al. 2004; Stevens & Merilaita 2008) which includes crypsis where an individual resembles its surroundings and masquerading, when an individual resembles a single inanimate object (Ruxton et al. 2004; Quicke 2017); and Batesian mimicry in non-toxic species that have evolved to imitate the appearance of toxic species (Lindström et al. 1997).
Others do not decrease initial detection, and in some cases may increase the chance of detection, but instead reduce the chance of the predator successfully carrying out the job and serve as a means of protection from a predator after detection including: defensive structures such as spines, shells or scales (Ruxton et al. 2004; Yadun & Halpern 2008); autotomy (loss of a body part) and regeneration (Maginnis 2006); warning colouration and associated toxins (Fisher 1930; Harvey et al. 1982), both of which are also associated with mimetic species (Huheey 1961; Brower & Brower 1963; Brower et al. 1963; Poughet al. 1973); venoms (Holding et al. 2016); construction of appropriate refugia e.g. bolt holes (Baugh & Deacon 1988); and plastic growth and development (Edmunds 1974; Lima & Dill 1990; Tollrian & Harvell 1999; Sokolowska et al. 2000). For example, the wide variation in shell structure in marine invertebrates is believed to be a response against shell-destroying predators (Vermeij 1997). Similarly, the freshwater three-spine stickleback (Gasterosteus aculeatus ) exhibits a complex of distinct morphological forms, varying in the degree of spinal and lateral plates as a consequence of predation risk (Colosimo et al. 2005).
Group aggregations are another example of an adaption in response to predator threat, leading to benefits including increased vigilance (Treves et al. 2001), selfish-herd effects (King et al.2012), and active defence e.g. musk ox Ovibos moschatus (Dixon 1998). Furthermore, such group behaviours have given rise to the development of group warning/alarm calls in some species (Smith 1986), and sometimes leads to sentinel behaviour in which one individual scans for predators whilst the other group members forage, observed in vervet monkeys (Chlorocebus pygerythrus ) (Horrocks & Hunte 1986). In Gunnison’s prairie dogs (Cynomys gunnisoni) this has led to species-specific warning calls, allowing the group to react in accordance to the threat (Slobodchikoff et al. 1991; Kiriazis & Slobodchikoff 2006; Slobodchikoff et al. 2009).
In addition to the mechanisms outlined above, prey animals use basic sensory modalities (visual, aural, olfactory) to detect the presence of predators in the same way that predators use both direct and indirect cues to identify and locate the position of the prey (Dwernychuk & Boag 1972; Sugden & Beyersbergen 1986; Nams 1997; Santisteban et al.2002; Carthey et al. 2011). In these contexts, visual, aural and olfactory cues each offer advantages and disadvantages to predator and prey in terms of being detected and/or avoiding detection. Vision is a directional cue and relies upon direct “line of sight” between two individuals, so can be used by prey to indicate the exact position and identity of the predator, and vice versa (Hemmi 2005). Predators may reduce the risk of detection while residing amongst vegetation, for instance. Often, prey may only observe their immediate environment at the expense of feeding, resulting in a trade-off between foraging and vigilance. Though social species may deploy sentinels to detect approaching predators, such individuals in turn experience a foraging cost by foregoing feeding to protect other group members (Doolan & Macdonald 1996; Whittingham et al. 2004). The slender-tailed meerkat (Suricata suricatta ), vervet monkey (Chlorocebus pygerythrus ), and black-tailed prairie dog (Cynomys ludovicianus ) exhibit such complexes within their respective social groups (Magle et al. 2005). Furthermore, visual cues are of limited use for nocturnal species and those living in dense vegetation.
Auditory cues are also directional, although generally to a lesser extent than visual cues due to sound resonating in a non-linear fashion from a point source. Predators, therefore, may inadvertently reveal their proximity when stalking prey, but such signals may not disclose exact locations and are less valuable to prey than direct sight. Furthermore, auditory signals are attenuated and reflected by vegetation and other surfaces, reducing the distance they travel and obscuring their origin (Goerlitz et al. 2008). Similar effects may arise in poor weather conditions, where sounds are refracted in strong winds (Clinchy et al. 2013). Consequently, predators that stalk prey must approach conspicuously and, preferably, upwind. Sound, however, is not likely a significant concern for ambush (sit-and-wait) predators (Boonstra et al. 2013).