Assessing the Effects of Climate on Host-Parasite Interactions: A Comparative Study of European Birds and Their Parasites

BACKGROUND: Climate change potentially has important effects on distribution, abundance, transmission and virulence of parasites in wild populations of animals. METHODOLOGY/PRINCIPAL FINDING: Here we analyzed paired information on 89 parasite populations for 24 species of bird hosts some years ago and again in 2010 with an average interval of 10 years. The parasite taxa included protozoa, feather parasites, diptera, ticks, mites and fleas. We investigated whether change in abundance and prevalence of parasites was related to change in body condition, reproduction and population size of hosts. We conducted analyses based on the entire dataset, but also on a restricted dataset with intervals between study years being 5–15 years. Parasite abundance increased over time when restricting the analyses to datasets with an interval of 5–15 years, with no significant effect of changes in temperature at the time of breeding among study sites. Changes in host body condition and clutch size were related to change in temperature between first and second study year. In addition, changes in clutch size, brood size and body condition of hosts were correlated with change in abundance of parasites. Finally, changes in population size of hosts were not significantly related to changes in abundance of parasites or their prevalence. CONCLUSIONS/SIGNIFICANCE: Climate change is associated with a general increase in parasite abundance. Variation in laying date depended on locality and was associated with latitude while body condition of hosts was associated with a change in temperature. Because clutch size, brood size and body condition were associated with change in parasitism, these results suggest that parasites, perhaps mediated through the indirect effects of temperature, may affect fecundity and condition of their hosts. The conclusions were particularly in accordance with predictions when the restricted dataset with intervals of 5–15 years was used, suggesting that short intervals may bias findings.