Data and code for Joint Species Distribution Models unveil co-occurrences between freshwater mussels and their fish hosts

Aim: Freshwater mussels are among the most threatened taxa in the world, partially due to the dependence on fish hosts, also highly threatened, to complete their life cycle. Joint Species Distribution Models (JSDMs) were applied to assemblages of mussels and their fish hosts’ data to assess the role of biotic and abiotic drivers in species occurrences and to test if these models were able to identify mussel-host interactions through patterns of species’ co-occurrences.
Location: Douro River Basin (Iberian Peninsula)
Taxon: Four freshwater mussels and ten fish hosts
Methods: JSDMs were fitted to presence-absence records for mussel and fish assemblages. Variance partitioning among environmental variables and latent variables was conducted to determine the environmental versus biotic drivers of species distributions. Resulting matrices of pairwise species co-occurrences were used to identify co-occurrence patterns.
Results: The distribution of host generalist mussel species was mainly explained by environmental variables related to climate (annual maximum temperature and average precipitation) and topography (elevation and river order). The distribution of the host specialist Margaritifera margaritifera was mainly explained by cropland extent. Strong positive correlations between mussels and the more important fish hosts were consistently captured by JSDMs. Co-occurrence patterns were mainly explained by residual factors, indicating the potential role of biotic interactions.
Main conclusions: While the obligatory interaction between mussels and fish hosts suggests that biotic interactions should play an important role in explaining mussel distributions, the contribution of this factor was only meaningful for the host specialist M. margaritifera. Yet, correlations between mussels and suitable hosts yielded by JSDMs allowed us to infer important fish hosts for freshwater mussels in the Douro River basin from distributional data alone. Furthermore, by finding connections between the ecological requirements of co-occurring species, conservation measures can be oriented towards several species which brings a more holistic perspective to the protection of biodiversity.