Effects of habitat destruction on coevolving metacommunities

Abstract

Habitat destruction is a growing threat to biodiversity and ecosystem services. The ecological consequences of habitat loss and fragmentation involve reductions in species abundance and even the extinction of species and their interactions. However, we do not yet understand how habitat loss alters the coevolutionary trajectories of the remaining species or how coevolution, in turn, affects their response to habitat loss. To investigate this, we develop a spatially explicit model which couples metacommunity and coevolutionary dynamics. We show that, by changing the size, composition and structure of local networks, habitat destruction increases the diversity of coevolutionary trajectories of mutualists across the landscape. Conversely, in antagonistic communities, some species increase while others reduce their spatial trait heterogeneity. Furthermore, we show that while coevolution dampens the negative effects of habitat destruction in mutualistic networks, its effects on the persistence of antagonistic communities tend to be smaller and less predictable.

Habitat restoration and the recovery of metacommunities

“…For restoration efficiency of regional interaction abundance, we used linear mixed models in the lme4 package in R. We chose metanetwork structure, metanetwork size, interaction extinction threshold (fraction of habitat loss at which that interaction becomes globally extinct), and interaction degree (sum of degrees of the two species involved in the interaction, Gawecka et al, 2022) as the fixed effects, and metanetwork ID as the random effect. We fitted these models for each restoration simulation (i.e., restoration starting at different fractions of habitat loss), scaling the explanatory variables such that their slope parameter estimates are comparable for a given restoration simulation (Schielzeth, 2010).…”

Habitat restoration and the recovery of metacommunities

“…We then investigated which properties of the metanetwork or the interaction affect the restoration efficiency, R. For the network level restoration efficiency measures (local interaction richness and network similarity), we used linear models with metanetwork structure (PC1) and metanetwork size (total number of species in the metanetwork) as the explanatory variables. For restoration efficiency of regional interaction abundance, we used linear mixed models in the lme4 package in R. We chose metanetwork structure, metanetwork size, interaction extinction threshold (fraction of habitat loss at which that interaction becomes globally extinct) and interaction degree (sum of degrees of the two species involved in the interaction, Gawecka et al, 2022) as the fixed effects, and metanetwork ID as the random effect. We fitted these models for each restoration simulation (i.e.…”

Multi-scale Effects of Habitat Loss and the Role of Trait Variation

“…5, row 2). Evolutionary responses to HL and fragmentation have been studied theoretically, for example for evolution of dispersal distance (North et al, 2011) and coevolution of interacting species (Gawecka et al, 2022), as well as observed empirically. For example, selection of better colonization ability was observed in a large butterfly metapopulation in Finland (Fountain et al, 2016).…”

“…Furthermore, when the trait variation is heritable, it can also help mitigate the effects of environmental change through "evolutionary rescue" (Bell & Gonzalez, 2011;Boeye et al, 2013;Gonzalez et al, 2013), or conversely aggravate the negative effects through "evolutionary trapping" (Ferriere & Legendre, 2013). These reciprocal effects can lead to eco-evolutionary feedbacks, between the population dynamic consequences of HL and trait variation, which are increasingly being recognised (Legrand et al, 2017;Faillace et al, 2021;Gawecka et al, 2022).…”