Short-term insurance versus long-term bet-hedging strategies as adaptations to variable environments

Abstract

11Understanding how organisms adapt to environmental variation is a key challenge of biology. Central to 12 this are bet-hedging strategies that maximize geometric mean fitness across generations, either by 13 being conservative or diversifying phenotypes. Theoretical models of bet-hedging and the multiplicative 14 fitness effects of environmental variation across generations have traditionally assumed that 15 environmental conditions are constant within lifetimes. However, behavioral ecology has revealed 16 adaptive responses to additive fitness effects of environmental variation within lifetimes, either through 17 insurance or risk-sensitive strategies. Here we explore whether the effects of adaptive insurance interact 18 with the evolution of bet-hedging by varying the position and skew of fitness functions within and 19 between lifetimes. When insurance causes the optimal phenotype to shift from the peak to down the 20 less steeply decreasing side of the fitness function, then conservative bet-hedging does not generally 21 evolve on top of this, even if diversifying bet-hedging can. Canalization to reduce phenotypic variation 22 within a lifetime is almost always favored, except when the tails of the fitness function are steeply 23 convex and produce a novel risk-sensitive increase in phenotypic variance akin to diversifying bet-24 hedging. Importantly, using skewed fitness functions, we provide the first example of how conservative 25 and diversifying bet-hedging strategies might coexist. 26 Keywords 27Fluctuating selection; environmental stochasticity; variance-sensitivity; geometric mean fitness; cliff-28 edge effect; phenotypic canalization.

Generalists versus specialists in fluctuating environments: a bet‐hedging perspective

“…Environmental fluctuations across different time scales pose a challenge for the evolution of organisms subjected to them, as well as for the researchers studying them (Botero et al , Tufto ). Notably, different fields of research have taken very different approaches to studying adaptations to stochastic environments, resulting in somewhat disparate bodies of literature dealing with quite similar topics (Haaland et al ).…”

“…A related CBH strategy is ‘playing it safe’ in the face of an asymmetric fitness function, which provides an adaptive explanation for some apparently suboptimal phenotypes seen in nature (Simons and Johnston , Haaland et al ). In their review, Starrfelt and Kokko () state that CBH and DBH are not mutually exclusive, but rather represent two ends of a continuum, and that both strategies can operate on the same trait.…”

“…However, in the popular illustrative model of an annual organism experiencing yearly variation in rainfall considered in theirs and other bet‐hedging papers (Seger and Brockmann , Crowley et al ), it is not entirely clear how such a continuum should work. Haaland et al () have recently provided a solution to this issue via these ‘playing it safe’ CBH strategies in response to asymmetric fitness functions, which under certain conditions allow CBH and DBH to co‐exist in the same trait as suggested by Starrfelt and Kokko (). However, it is currently unclear how the original ‘compromise’ or generalist CBH strategy could coexist with DBH, as was originally suggested, at least under discrete ‘two‐patch’ scenarios of environmental fluctuation.…”

“…Nevertheless, many topics in behavioral ecology echo the arguments of CBH and DBH while seeking to maximize only individual arithmetic mean fitness, and it is therefore hard to know how and when these predictions fail or succeed in capturing actual evolutionary outcomes. For example, insurance strategies in the face of an asymmetric fitness function due to starvation or predation risk (Brodin , Dall ) have much in common with ‘playing it safe’ CBH strategies (Haaland et al ), and generalist versus specialist foraging strategies as adaptations to within‐lifetime fluctuations in prey availability (Dall and Cuthill , Tinker et al ) likewise focus on individual‐level strategies and benefits (Fig. B) despite apparently mirroring the logic of DBH (Fig.…”

“…arithmetic mean fitness) may fail to predict evolutionary outcomes. As Haaland et al () showed for insurance and ‘playing it safe’ CBH (i.e. with skewed fitness functions), traits maximizing arithmetic mean fitness in response to within‐generation stochasticity may well reduce the scope for bet‐hedging traits increasing geometric mean fitness in response to between‐generation stochasticity.…”