Résumé
This report presents the final version of the metacommunity model developed in WP2 (“Predicting biodiversity changes in DRNs”) to investigate biodiversity dynamics in drying river networks. We adapted the individual-based simulation algorithm of metacommunity dynamics in discrete time of Jabot et al. (2020) to drying river networks. Biodiversity dynamics is determined by dispersal-driven regional processes, as well as local demographic processes (birth and death). In each river reach and at each time step, we simulated four processes taking place sequentially: (i) mortality (ii) reproduction (iii) dispersal and (iv) establishment, which accounts for competition for resources. We considered three modes of aquatic and aerial dispersal (drifting, swimming/crawling and flying), reflecting the diversity of dispersal modes among riverine organisms. We explored varying combinations of duration, intensity and location of drying events along river networks. The model consistently reproduced biodiversity patterns observed in river networks: local species richness was higher in downstream reaches or in central reaches for drifting and swimming organisms, respectively. We also found that biodiversity recovery from drying events can be predicted by patch connectivity. Hence, loss of patch connectivity decreased community recovery, regardless of patch location in the river network, dispersal mode or drying intensity. Local communities of flying organisms maintained higher patch connectivity in drying river networks compared to organisms with strictly aquatic dispersal, which explained the higher recovery capacity of flying organisms from drying events. This model of metacommunity dynamics in drying river networks was encapsulated in the R package “cantal” available at https://forgemia.inra.fr/lisc/r-package-cantal (deliverable D2.2). The metacommunity simulator allows to simulate a wide range of drying scenarios, as well as potential interspecific heterogeneity in dispersal abilities and resistance to drying. The simulator can also simulate varying scenarios of drying events based on input parameters coming from empirical data (e.g., river network structure, species dispersal mode and ability, resistance to drying). Part of this work has been submitted for publication and deposited on a public repository: https://www.biorxiv.org/content/10.1101/2022.01.02.474736v1.abstract