Abstract
The broad distribution of macroparasites and their thriving populations are matters of health and economic concern. Macroparasites cause damage both directly through their feeding habits, which impact host fitness, and indirectly through the transmission of various infectious diseases of relevance to human and domestic animal health and wildlife conservation. Because the impacts of macroparasites on host health and the risk of disease transmission are directly related to their abundance, understanding the drivers of macroparasite burden is of relevance. Various host traits and environmental factors have been associated with differences in macroparasite abundance. In addition to these variables, spatial scale is increasingly incorporated to understand how these drivers vary across space. However, variation in the relative importance of host traits and environmental factors as predictors of abundance at different scales is not well understood. To further clarify the relationship between scale and drivers of macroparasite abundance, we investigated the effects of host traits and environmental factors on flea abundance in rodents of the Chihuahuan Desert in northwestern Mexico on three levels: within a single site, between sampling sites with different vegetation types, and across the region. This partitioning allowed us to compare drivers at both local and regional scales. Fleas provide a natural model to assess the interplay between host and environmental variables across scales because their life cycles alternate between on‐host and off‐host environments and their hosts have varying ranges of distribution. We sampled 1311 fleas from 674 rodent individuals of 14 different species across 40 sampling plots between 2012 and 2013. Using generalized linear mixed models, we found that flea abundance was associated with different combinations of host traits such as size and sex. The specific combination of predictive variables differed across species, while the effects on flea abundance showed context and scale dependency, although this could only be tested at the full level of analysis on the most abundant species, Dipodomys merriami . Sampling season was the only variable consistently significant across scales, reflecting the far‐reaching effects of large‐scale, interannual environmental fluctuations. These results emphasize that integrating spatial scale can strengthen study design for monitoring macroparasite burden.