The degree of spatial variation relative to temporal variation influences evolution of dispersal

Abstract

In the face of ongoing global climate and land use change, organisms have multiple possibilities to cope with the modification of their environment. The two main possibilities are to either adapt locally or disperse to a more suitable habitat. The evolution of both local adaptation and dispersal interacts and can be influenced by the spatial and temporal variation (of e.g. temperature or precipitation). In an individual based model (IBM), we explore evolution of phenotypes in landscapes with varying degree of spatial relative to global temporal variation in order to examine its influence on the evolution of dispersal, niche optimum and niche width. The relationship between temporal and spatial variation did neither influence the evolution of local adaptation in the niche optimum nor of niche widths. Dispersal probability is highly influenced by the spatio‐temporal relationship: with increasing spatial variation, dispersal probability decreases. Additionally, the shape of the distribution of the trait values over patch attributes switches from hump‐ to U‐shaped. At low spatial variance more individuals emigrate from average habitats, at high spatial variance more from extreme habitats. The comparatively high dispersal probability in extreme patches of landscapes with a high spatial variation can be explained by evolutionary succession of two kinds of adaptive response. Early in the simulations, extreme patches in landscapes with a high spatial variability act as sink habitats, where population persistence depends on highly dispersive individuals with a wide niche. With ongoing evolution, local adaptation of the remaining individuals takes over, but simultaneously a possible bet‐hedging strategy promotes higher dispersal probabilities in those habitats. Here, in generations that experience extreme shifts from the temporal mean of the patch attribute, the expected fitness becomes higher for dispersing individuals than for philopatric individuals. This means that under certain circumstances, both local adaptation and high dispersal probability can be selected for for coping with the projected environmental changes in the future.

Temporal environmental variation may impose differential selection on both genomic and ecological traits

“…Lower body sizes also increase reproductive rates (Savage et al 2004) and thus more offspring dispersal can take place, further assisting species to respond environmental change via dispersal (Berg et al 2010). Additionally, theoretical evidence has suggested that environmental fluctuations may select for intraspecific variation of dispersal ability (Mathias et al 2001, Sieger and Hovestadt 2020), constituting bet‐hedging strategies. Fluctuating environment may also select for increased niche tolerances to cope with varying conditions (Lynch and Gabriel 1987, Holt 1990), although this might be mostly relevant for larger species (due to longer generation times) and poor dispersers (reviewed by Sexton et al 2017).…”

Effects of compositional heterogeneity and spatial autocorrelation on richness and diversity in simulated landscapes

“…This results in a landscape community that is less rich overall but also less globally dominated by a small number of highly abundant species. Due to their larger population sizes and the bet-hedging opportunities afforded by a spatially heterogeneous environment, these populations are likely to survive extreme environmental fluctuations (Sieger & Hovestadt, 2020), resulting in a more stable landscape community composition in the long term. As with all models, our model makes a number of simplifying assumptions.…”

“…Here, we attempt to address this gap using a spatially explicit, using an individual-based metacommunity model implemented in Julia 1.1.1 (Bezanson et al, 2012) based on the model of Sieger and Hovestadt (2020) to systematically explore the effects of landscape structure on patterns of species richness and diversity via their…”

Dealing with software complexity in individual‐based models

“…Overall, our experiences with the language have been almost entirely positive (e.g. Leidinger et al, 2021;Sieger & Hovestadt, 2020;Vedder et al, 2020).…”