Soil fauna modulates the effect of experimental drought on litter decomposition in forests invaded by an exotic pathogen

Litter decomposition is a fundamental process for nutrient cycling and C fluxes between terrestrial ecosystems and the atmosphere. Multiple factors such as environmental conditions, litter quality and decomposer organisms are known to influence this ecological process in direct and indirect ways. However, it remains poorly understood to what extent this entangled bank of complex interactions can be disrupted by several global change drivers such as climate change and invasive species. Here we reported main findings from a litter decomposition experiment conducted in a natural mixed forest invaded by Phytophthora cinnamomi, an exotic oomycete well known for causing tree disease and forest decline. In two Mediterranean forest types (namely woodland and closed forest), we first built a rainfall exclusion infrastructure to reduce natural precipitation by 30%. Then, we followed the spatial variation in C and N leaf litter dynamics as well as the abundance of the exotic oomycete P. cinnamomi for 18 months under both control and rainfall exclusion conditions using a litter bag methodology with different mesh sizes to further evaluate the indirect role of mesofauna abundance and diversity in litter C and N temporal dynamics. Significant reductions in soil moisture produced by the experimental rainfall exclusion increased C litter loss with time directly, and indirectly via increases in overall mesofauna abundance and changes in the ratio between predators and decomposers. N litter dynamics were in contrast modulated mainly by initial litter quality. Surprisingly, P. cinnamomi abundance did not correlate with variation in initial litter quality as previously suggested, but we found that the exotic oomycete was negatively correlated with both decomposer and predator abundances. These results suggest it might participate indirectly in litter decomposition dynamics through the biotic soil component. Synthesis. Contrary to expected, our results show that a moderate but realistic rainfall reduction can accelerate the litter decomposition process. It also points out to litter fauna as a key component modulating the indirect impacts of global change drivers on litter decomposition.