Evolutionary Branching in Complex Landscapes

Abstract

Divergent adaptation to different environments can promote speciation, and it is thus important to consider spatial structure in models of speciation. Earlier theoretical work, however, has been limited to particularly simple types of spatial structure -linear environmental gradients and spatially discrete metapopulations -leaving unaddressed the effects of more realistic patterns of landscape heterogeneity, such as nonlinear gradients and spatially continuous patchiness. To elucidate the consequences of such complex landscapes, we adapt an established spatially explicit individual-based model of evolutionary branching. We show that branching is most probable at intermediate levels of various types of heterogeneity, and that different types of heterogeneity have, to some extent, additive effects in promoting branching. In contrast to such additivity, we find a novel refugium effect in which refugia in hostile environments provide opportunities for colonization, thus increasing the probability of branching in patchy landscapes.Effects of patchiness depend on the scale of patches relative to dispersal. Providing a needed connection to empirical research on biodiversity and conservation policy, we introduce empirically accessible spatial environmental metrics that quantitatively predict a landscape's branching propensity.3

The Biogeography of Adaptive Radiations and the Geographic Overlap of Sister Species

“…Similar to several earlier models of competitive interactions and the evolution of resource specialization (Christiansen and Loeschcke 1980;Brown and Vincent 1987;Dieckmann and Doebeli 1999;Haller et al 2013), we assume that local fitness of a particular phenotype is a function of a single, continuous, evolving trait (e.g., body size) that maps onto a unimodal resource distribution. We use the distribution of trait values to define species and speciation events: clusters of phenotypically similar individuals are viewed as species, and clusters branching into two distinct clusters are registered as speciation events (see details below).…”

“…However, this generality is associated with assumptions that may or may not impact the results. First, similar to, for example, Haller et al (2013), we neglect genetic mechanisms and sexual reproduction. Our results consequently do not encompass mechanisms such as recombination (which may prevent speciation) and nonecological speciation (e.g., Dobzhansky-Mueller incompatibilities).…”

Determining Selection across Heterogeneous Landscapes: A Perturbation-Based Method and Its Application to Modeling Evolution in Space

“…The notion that spatial heterogeneity should promote evolutionary diversification is therefore intuitively appealing. Although this intuition is sometimes foiled (Day 2001;Ajar 2003), several individual-based simulation and population genetics studies do indeed suggest that spatial structure can contribute to evolutionary branching and speciation (Felsenstein 1976;Doebeli and Dieckmann 2003;Servedio and Gavrilets 2004;Gavrilets and Vose 2005;Haller et al 2013). However, understanding the interactions among spatial structure, population dynamics, and evolutionary dynamics has proved challenging.…”

“…A frequently employed approach has been to use spatially explicit, individual-based models that compute traitdependent reproduction and inheritance directly based on various rules (Doebeli and Dieckmann 2003;Gavrilets and Vose 2005;Mágori et al 2005;Birand et al 2012;Haller et al 2013;Kubisch et al 2014). These models exhibit potentially great ecological realism and have broad applicability but suffer from two limitations: they are computationally very demanding, and the results derived from these models are often difficult to analyze and check for robustness (but see the paragraph on moment-closure approximations in "Discussion" for further description and references).…”

Adaptive dynamics on an environmental gradient that changes over a geological time-scale

“…In recent years there have been several research papers on evolution on an environmental gradient (Kirkpatrick and Barton, 1997;Meszena et al, 1997;Case and Taper, 2000;Doebeli and Dieckmann, 2003;Mizera and Meszena, 2003;Leimar et al, 2008;Heinz et al, 2009;Ispolatov and Doebeli, 2009;Debarre and Gandon, 2010;Payne et al, 2011;Haller et al, 2013). So far most papers on evolution on an environmental gradient have assumed the gradient to be fixed in time.…”