A multi-scale landscape approach to understand dispersal of the mistletoe by birds in Mediterranean pine forests

Context Seed dispersal by birds plays a crucial role in structuring landscape dynamics. The dispersal process has been extensively studied at local scales; however, landscape scale approaches are scarce. The wide availability of bird species distribution data can be used to explore plant dispersal processes at large spatial scales. Objective We investigated whether and at which scale bird-seed-dispersal indicators (BSDI) based on species distribution models (SDMs), together with climatic and landscape structure metrics related to habitat preferences of birds, aid the explanation of the distribution of Viscum album. Methods We developed a set of BSDI (abundance, species richness, Turdus spp. richness, and specialization) at four different scales (500-3500 m). They served as predictors, along with other environmental variables, in generalized linear models to predict V. album distribution in Catalonia, Spain. Results The indicator Turdus spp. richness (3500 m) explained up to 15% of variance. This contribution was smaller when it was considered alongside other predictors, where climatic conditions and % of olive groves explained ca. 35% and 13% variance, respectively. The abundance or species richness BSDI proved to be poor predictors of V. album distribution. Conclusions We demonstrated the functionality of BSDI based on SDMs to predict the distribution of V. album and the importance to test them at different spatial scales. BSDI showed a clear larger effect with increasing scale, suggesting a strong link between the mobility of birds and the plant dispersal process. The modelling approach could be applied to different localities and species to understand plant seed dispersal at landscape scale.

Automated summary

The study investigated the use of bird-seed-dispersal indicators (BSDI) based on species distribution models (SDMs) in predicting the distribution of Viscum album. Results showed that Turdus spp. richness indicator at 3500 m scale explained up to 15% of variance, while climatic conditions and percentage of olive groves were also significant predictors. The abundance or species richness BSDI were less effective in predicting V. album distribution.