Landscape features impact connectivity between soil populations: a comparative study of gene flow in earthworms

Abstract

Landscape features are known to alter the spatial genetic variation of aboveground organisms. Here, we tested the hypothesis that the genetic structure of belowground organisms also responds to landscape structure. Microsatellite markers were used to carry out a landscape genetic study of two endogeic earthworm species, Allolobophora chlorotica (N = 440, eight microsatellites) and Aporrectodea icterica (N = 519, seven microsatellites), in an agricultural landscape in the North of France, where landscape features were characterized with high accuracy. We found that habitat fragmentation impacted genetic variation of earthworm populations at the local scale. A significant relationship was observed between genetic diversity (H , A ) and several landscape features in A. icterica populations and A. chlorotica. Moreover, a strong genetic differentiation between sites was observed in both species, with a low degree of genetic admixture and high F values. The landscape connectivity analysis at the regional scale, including isolation by distance, least-cost path and cost-weighted distance approaches, showed that genetic distances were linked to landscape connectivity in A. chlorotica. This indicates that the fragmentation of natural habitats has shaped their dispersal patterns and local effective population sizes. Landscape connectivity analysis confirmed that a priori favourable habitats such as grasslands may constitute dispersal corridors for these species.

Functional connectivity in replicated urban landscapes in the land snail (Cornu aspersum)

“…Moreover, landscape‐specific effects were analysed by linear models computed for each landscape (as classically done in nonreplicated landscape studies—e.g., Dupont et al., ). This one‐by‐one landscape approach was applied to population and pairwise F ST analyses.…”

“…We first described the overall genetic structure of our data set by means of a hierarchical analysis of molecular variance using the R package HIERFSTAT (Goudet & Jombart, 2015). Moreover, landscape-specific effects were analysed by linear models computed for each landscape (as classically done in nonreplicated landscape studies-e.g., Dupont et al, 2017). This one-by-one landscape approach was applied to population and pairwise F ST analyses.…”

“…The classically used resistancebased models are least-cost path (LCP, Adriaensen et al, 2003) and circuit-theory (CT-)-based analyses (McRae, Dickson, Keitt, & Shah, 2008). The different scenarios based on these two models can vary according to the resistance value sets, habitat types and environmental conditions, which support assessing of their relative performances (e.g., Braaker et al, 2017;Dupont, Torres-Leguizamon, Ren e-Corail, & Mathieu, 2017;Fenderson et al, 2014).…”

Methods for studying earthworm dispersal

“…It suggested that soil properties influence dispersal in the field, and estimated that gene flow occurs at approximately 7 m.year-1. At the landscape level, it showed the role of landscape features on the dispersal of two earthworm species, A. chlorotica and Aporrectodea icterica (Dupont et al, 2017).…”

“…This process can be realized with different scenario of costs. By comparing different scenarios of costs, this technique allows identifying which elements are most likely to play a role on dispersal, either by facilitating it (corridors) or by impeding it (barriers) (Dupont et al, 2017). In order to match the distance matrix with genetic features, the genetic differentiation between all sampling sites must also be estimated.…”