Măgălie Andreea, Schwartz Daniel A, Lennon Jay T, Weitz Joshua S
School of Biological Sciences, Georgia Institute of Technology, Atlanta, GA, USA; Interdisciplinary Graduate Program in Quantitative Biosciences, Georgia Institute of Technology, Atlanta, GA, USA.
Department of Biology, Indiana University, Bloomington, IN, USA.
J Theor Biol. 2023 Mar 21;561:111413. doi: 10.1016/j.jtbi.2023.111413. Epub 2023 Jan 11.
Organisms have evolved different mechanisms in response to periods of environmental stress, including dormancy - a reversible state of reduced metabolic activity. Transitions to and from dormancy can be random or induced by changes in environmental conditions. Prior theoretical work has shown that stochastic transitioning between active and dormant states at the individual level can maximize fitness at the population level. However, such theories of 'bet-hedging' strategies typically neglect certain physiological features of transitions to dormancy, including time lags to gain protective benefits. Here, we construct and analyze a dynamic model that couples stochastic changes in environmental state with the population dynamics of organisms that can initiate dormancy after an explicit time delay. Stochastic environments are simulated using a multi-state Markov chain through which the mean and variance of environmental residence time can be adjusted. In the absence of time lags (or in the limit of very short lags), we find that bet-hedging strategy transition probabilities scale inversely with the mean environmental residence times, consistent with prior theory. We also find that increasing delays in dormancy decreases optimal transitioning probabilities, an effect that can be influenced by the correlations of environmental noise. When environmental residence times - either good or bad - are uncorrelated, the maximum population level fitness is obtained given low levels of transitioning between active and dormant states. However when environmental residence times are correlated, optimal dormancy initiation and termination probabilities increase insofar as the mean environmental persistent time is longer than the delay to reach dormancy. We also find that bet hedging is no longer advantageous when delays to enter dormancy exceed the mean environmental residence times. Altogether, these results show how physiological limits to dormancy and environmental dynamics shape the evolutionary benefits and even viability of bet hedging strategies at population scales.
生物体已经进化出不同的机制来应对环境压力时期，包括休眠——一种代谢活动降低的可逆状态。进入和离开休眠状态的转变可以是随机的，也可以由环境条件的变化引发。先前的理论研究表明，个体水平上活跃状态和休眠状态之间的随机转变可以在种群水平上使适应性最大化。然而，这种“风险对冲”策略的理论通常忽略了进入休眠状态转变的某些生理特征，包括获得保护益处的时间滞后。在这里，我们构建并分析了一个动态模型，该模型将环境状态的随机变化与生物体的种群动态耦合起来，这些生物体在明确的时间延迟后可以进入休眠状态。随机环境是使用多状态马尔可夫链进行模拟的，通过该链可以调整环境停留时间的均值和方差。在没有时间滞后（或在非常短的滞后极限情况下）时，我们发现风险对冲策略的转变概率与环境平均停留时间成反比，这与先前的理论一致。我们还发现，休眠延迟的增加会降低最优转变概率，这种效应可能会受到环境噪声相关性的影响。当环境停留时间——无论是好是坏——不相关时，在活跃状态和休眠状态之间进行低水平转变时可获得最大种群水平适应性。然而，当环境停留时间相关时，只要环境平均持续时间长于进入休眠状态的延迟时间，最优休眠启动和终止概率就会增加。我们还发现，当进入休眠状态的延迟超过环境平均停留时间时，风险对冲不再有利。总之，这些结果表明了休眠的生理极限和环境动态如何在种群尺度上塑造风险对冲策略的进化益处甚至生存能力。
