Integrating host biological and ecological variables to predict probability of haemosporidian infection in raptors

Variations in host traits that influence their exposure and susceptibility may impact probability of vector-transmitted diseases. Therefore, identifying the predictors of infection probability is necessary to understand the risk of disease outbreaks during expanding environmental perturbation. Here, we conducted a large survey based on microscopic examination and molecular analysis of haemosporidian parasite infection in raptors rescued at the Beijing Raptor Rescue Centre. Combining these data with biological and ecological variables of the raptors, we determined predictors that affect the probability of haemosporidian infection using generalized linear mixed models and multimodel inference. Our results showed that infection probability exhibited considerable variation across host species in raptors, and body mass, sex, and evolutionary history played relatively weaker roles in driving infection probability. Instead, activity pattern, age, geographic range size, migration distance, and nest type were important predictors of the probability of haemosporidian infection, and the role of each predictor differed in the three main haemosporidian genera (Plasmodium, Haemoproteus, and Leucocytozoon). This macro-ecological analysis will add to our understanding of host traits that influence the probability of avian haemosporidian infection and will help inform risk of emerging diseases.