Trade-off 4: Larger-scale spatial or temporal escape from stressors
As an alternative to coping with the suite of stressors an organism faces locally and immediately (the direct and indirect effects which are discussed in the preceding sections on trade-offs 1-3), some organisms can escape environmental stressors in space or time through long-distance dispersal or some form of substantial, relatively long-term reduction in metabolic demands. Escape in space (EIS) involves an organism, temporarily or permanently, relocating to a new environment. Seasonal migration, exhibited by various mammals, birds, and insects, is a common, cyclical form of temporary EIS, often tracking predictable large-scale variations in weather patterns and food availability. EIS can also be triggered by anthropogenic environmental stressors (Berg et al. 2010). For example, the onset of human hunting, rather than the onset of severe weather (e.g., snowfall), was a primary driver of autumn migration by red deer (Rivrud et al.2016). Similarly, individual mule deer that undertake long migrations have been shown to dramatically reduce their risk of being harvested by human hunters relative to short- and moderate-distance migrants in the same population (Sawyer et al. 2016). Furthermore, non-migratory butterfly species have moved over vast areas of habitat made unsuitable by anthropogenic climate change to occupy new locations in Europe (Parmesan et al. 1999). Escape in time (EIT) involves reducing exposure and avoiding the costs of tolerance or dispersal by instead entering into torpor (Humphries et al. 2003; Geiser 2004), dormancy (including hibernation or estivation; (Danks 2000)), diapause (a special case of dormancy based on suspended development; (Chapman 1998)), or resting stages (Smirnov 2014). EIT is commonly used by animals to address extreme temperatures, drying conditions, and a limited food supply (Thomas et al. 1990; Danks 2000; Gotoet al. 2001; Sarmaja-Korjonen 2003; Hairston Jr. & Fox 2009).
Theory on the evolution of adaptive dispersal and/or dormancy provides insights regarding factors that influence when organisms should attempt to escape in space or time (Levin et al. 1984; Snyder 2006; Bonte & de la Pena 2009; Bonte & Dahirel 2017). Whether dispersal or dormancy is adaptive depends on the expected net fitness payoff, which depends on how organisms handle trade-offs 1-3 (Fig. 1) in both the current environment, and in alternative environments, as well as costs of escape in space or time (including mortality and the need for a substantial front-end investment in energy stores that increases escape success). Both dispersal and dormancy can involve substantial uncertainty about expected fitness. For long-distance dispersal, there is uncertainty about transit costs (that depend on both cost per unit of distance or time, and distance relative to mobility) and often great uncertainty over likely payoffs in prospective new environments. This might be especially true now, following the unprecedented human-induced rapid environmental change that is shaping natural habitats globally (Crowley et al. 2019; Van de Waal & Litchman 2020).
Importantly, when organisms disperse to a new habitat, they might face a different set of stressors that require a different set of behavioral and physiological responses. Thus, the suitability of a new environment could hinge on the organism’s plasticity in behaviour and physiology. The degree of dissimilarity between the suite of dominant stressors in an organism’s former environment compared to its new environment can come with distinct costs; e.g., new stressors could require greater energetic investment in establishing appropriate physiological or behavioral responses. Furthermore, if, as is often the case, there are behavioural or physiological carryovers (i.e., earlier experiences with stressors influence later responses), then EIS can expand the scope of multiple stressors to include stressors that do not co-occur in space or time (see Box 4).
The benefit of escaping in time or space to a new environment is proportional to not just the increase in quality in the new environment but also to how long the new environment will remain of higher quality (i.e., the degree of temporal stability). If stressor levels fluctuate frequently or intensely over time, this can dilute the benefits of escape. In short, the key for adaptive dispersal or dormancy is not the spatiotemporal pattern of stressors per se; it is, instead, the spatiotemporal pattern of fitness adjusted for costs of dispersal or dormancy. In addition, because escaping to a new environment might also result in greater competition or predation risk (e.g., if competitors or predators make the same escape decision, or population demography yields this result), there is a game-theoretic aspect to this dynamic that can further complicate expectations.
Despite the various sources of complexity that can arise when considering whether an organism should stay and cope with stressors or, instead, attempt to escape them, simplified scenarios offer qualitative insights. Generally, we expect that the probability that an organism will attempt to escape stressors would scale with the potential for that escape to be possible: if stressors are highly localized in space or time, we would expect escape to be much more likely, relative to when stressors are widespread over space or time (i.e., chronic), and, thus, difficult or impossible to escape. If stressors are widespread in only one-dimension (i.e., time or space), we expect a threshold to exist for the other dimension, such that increasing the stressor’s (or suite of stressors’) presence over this dimension causes the organism to eventually shift from an optimal strategy of escape to one of tolerance (Figure 3A). In other words, when stressors occur at large enough scales in space and time, they become infeasible to avoid and tolerance becomes the sole strategy.
If we expand to consider animal behavior in in the context of multiple stressors, the horizon of possibilities quickly becomes much more complex. For example, considering only two partially correlated stressors, expressed in time and space relative to the scales of these dimensions experienced by a focal organism, reveals that ten qualitatively distinct sets of possible behavioural responses emerge (Figure 3B). Generally, whenever one or more stressors is escapable in space or time, formulating quantitative predictions about when organisms will choose this strategy will hinge on the magnitude of effects of the escapable stressor(s) and the cost-benefit ratio of choosing to escape relative to choosing to tolerate the stressor(s).