Effects of forest structure on the interaction between avian hosts, dipteran vectors and haemosporidian parasites.

backgroundForest habitats are important biodiversity refuges for a wide variety of bird species. Parasitism may modulate host species presence and abundance, and parasite effects can change according to forest management practices. Such processes are not well studied in vector-borne avian haemosporidians. We analyzed the effects of forest management on bird-dipteran-haemosporidian interactions, using seven common bird species in managed and unmanaged beech forest habitats in northeastern Germany. We assumed that forest structural heterogeneity affects parasite population parameters in avian hosts (i.e., prevalence and parasitemia), through its effect on the condition of the avian host but also through varying vector abundances.

resultsParasite prevalence was high (about 80%) and homogeneous across different beech forest categories (i.e., young, old, unmanaged) and for all bird species, except Erithacus rubecula (35%). Parasitemia varied across bird species but not across forest categories within each avian species (lowest parasitemia were found in E. rubecula, Turdus merula, and Turdus philomelos). In our study system, we found that vector abundance was not the main driver of parasite dynamics. We found that forest structure affects parasite infection probability directly and potentially host condition via available resources that have to be used either to combat infections (i.e., high parasitemia) or to maintain a good body condition.

conclusionsThe effects of each of the predictors were bird species-specific, and we found that Diptera vectors were not the foremost influence in our host-vector-parasite system. Effects of forest habitat variables indicated that for most bird species in this study, habitat regulation of infection probability was more likely (i.e., E. rubecula, Fringilla coelebs, Sylvia atricapilla), whereas for Parus major habitat characteristics impacted first individuals' body condition and subsequently the probability of infection. Our findings emphasize the need of species-specific analyses and to use continuous forest structural parameters (e.g., the proportion of gap, south facing aspect) to better understand habitat and land use effects on host-vector-parasite dynamics.