Box 2: Categorizing enemy-risk effects
The term NCE is frequently used to describe processes at three levels: the enemy-induced trait response (e.g., increased refuge use), the consequences for the individual prey/host (e.g., reduced growth rate), or the consequences at the prey/host population level (e.g., increased emigration). Referring to all three levels as NCEs reduces the important distinctions between them, so we advocate for a more explicit framework (Fig. 1), and clearer terminology (also see Peacor et al. 2020). We will use the terms enemy-risk effect to refer to the overall process, enemy-induced trait response to refer to the mechanism of response, NCE to refer to fitness/population consequences, and trait-mediated indirect effect to refer to effects cascading to trophic levels below the prey/host. A complementary way of conceptualizing enemy-risk effects is to take a more phenomenological approach, focusing on aspects of a pest population: its per-capita impact, abundance, and distribution (box shading in Fig. 1).
Behavioral shifts are a commonly studied trait responses in arthropods, and are generally the most rapid and reversible. Examples include changes in time spent feeding (Thaler & Griffin 2008; Jandricic et al. 2016; Ingerslew & Finke 2017), food source (Schmitz et al. 1997), microhabitat and refuge use (Lucas et al. 2000; Lawson-Balagbo et al. 2007; Penfold et al. 2017), oviposition rate (Deas & Hunter 2013; Hermann & Thaler 2018), oviposition site selection (Angelon & Petranka 2002; Vonesh & Blaustein 2010; Silberbush & Blaustein 2011), short-distance escape (Tamaki et al. 1970; Nelson 2007; Fill et al. 2012), and dispersal (Höller et al. 1994; Henry et al. 2010; Otsuki & Yano 2014b; Welch & Harwood 2014).
Physiological shifts can be direct responses to risk, but they are often consequences of behavioral shifts. For example, a reduction in individual growth rate (physiological) is often a result of reduced foraging effort (behavioral). This can make physiological shifts difficult to categorize within the framework shown in Fig. 1. Examples include changes in growth rate (Kaplan et al. 2014), development time (Bellamy & Alto 2018), and assimilation efficiency (Thaler et al. 2014).
Morphological shifts are generally slower to appear and more difficult to reverse than behavioral or even physiological shifts. They have been less described in terrestrial arthropods, but thoroughly studied in systems such as Daphnia pulex, where predator cues trigger production of carapace protrusions that decrease vulnerability to predation (Havel & Dodson 1984; Tollrian 1995; Rabus & Laforsch 2011). Life history shifts frequently occur over a long timescale and are irreversible for an individual prey/host. They include changes in timing of reproduction or metamorphosis (Ims 1990; Benard 2004; Relyea 2007), quality and quantity of offspring produced (Mappes et al. 1997), and production of winged morphs (Sloggett & Weisser 2002; Kunert & Weisser 2003).
Trait responses carry costs for individuals, and we can categorize NCEs based on these costs. These costs are ultimately tied to individual fitness, including reduced fecundity (Mappes et al. 1997) and reduced survival (Walzer & Schausberger 2009).
Both responses and consequences at the individual level can cascade to affect the entire prey/host population. Finally, community-level impacts include both trait-mediated indirect effects, wherein an NCE reduces the prey population such that they have a smaller effect on a lower trophic level, and interaction modifications, wherein a trait response causes an existing interaction with another species to change. As seen in Fig. 1, these community effects can occur via different pathways that may not be captured equally in all studies.