Assessing the effects of quarrying activities on landscape connectivity in two amphibian species with contrasted ecological constraints

Industrial activity is a major driver of fragmentation, but each type of activity has a different impact. Quarrying causes the removal of initial habitat and may hence increase landscape fragmentation (Berhe, 2007). Conversely, it may create particular habitats such as pioneer habitats or wetlands, probably increasing this way the connectivity for species associated with those environments (Tropek & Konvicka, 2008; Santoul et al., 2009). Moreover, the spatial configuration of quarries (i.e. isolated, clustered, or in stepping-stone networks) may modulate those effects. To assess the global impact of quarries on landscape connectivity, we used a landscape genetic approach on two niche-separated anurans: the natterjack toad (Bufo calamita) and the common toad (Bufo bufo). Although these two congeneric anurans look similar in terms of size, life history and general biology, they clearly differ in terms of habitat preferences and dispersal behaviour. The natterjack toad prefers poor-vegetated habitat similar to those encountered in active quarries, while the common toad looks for heavily vegetated and forested habitats. By comparing effective dispersal of both species in landscapes with different spatial configurations of quarries, we tested the hypothesis that quarry density has positive effects on pioneer species connectivity and negative effects on forested species connectivity.

To this end, we sampled respectively 1171 and 1102 tadpoles of common toads and natterjack toads in seven areas with different densities of quarries. Using respectively 10 and 11 polymorphic microsatellite loci for these species, we first used Bayesian and non-Bayesian clustering approaches and multivariate analyses to assess the genetic structure at the site scale and identify potential global barriers or corridors. Then, based on GIS data of soil occupancy and between-pond gene flow and direction, we used an optimization process to assess the relative resistance of quarries and other landscape elements on gene flow in both species. Because genetic patterns reflect both recent and historic landscape structure, GIS data layers were constructed at three time steps, each separated by 10 estimated generations of toads (1946, 1976 and 2006). The optimization process consisted in looking for the resistance values that maximized the correlation between genetic distances and geographic distances accounting for landscape effects. We compared the performance of i) several genetic metrics (e.g. pair-wise genetic distance F_ST) and recent bi-directional migration rates, and ii) several geographic metrics related to increasing spatial information: Euclidian distance for isolation by distance, cost-weighted distance and resistance distance for specie-specific isolation by landscape resistance (McRae, 2006; McRae &  Beier, 2007), and connectivity estimates based on a stochastic movement simulator for specie-specific perceptual range and degree of correlation in movement (Palmer et al., 2011).

Our first results show that resistance distance outperforms classical landscape metrics like Euclidian distance or cost-weighted distance. In the case of natterjack toad they also show a clear corridor-effect due to quarrying activity and a fragmentation effect of intensive agricultural area. Conversely, preliminary results suggest that common toad dispersal is positively influenced by forested habitat, but quarrying impact remains unclear at this time. To dig up and confirm these points, the entire optimization process will be pursued, and the appropriate time scale identified. Last, in light of our results, integrated landscape conservation measures for before, during and after quarrying activity will be proposed.