Using Multivariate Analyses to Explore Host-Pathogen Coevolution in Complex Trait Space.

Abstract

Lesion severity scores are often used to monitor individual health outcomes following the incursion of certain wildlife diseases. However, collapsing a complex trait such as pathology onto a single axis can mask critical information about host-pathogen interactions. In this study, we apply multivariate techniques (shape and community analyses) to explore potential patterns of coevolution in a well-studied wildlife disease system: House Finches (Haemorhous mexicanus) infected with Mycoplasma gallisepticum, a bacterium that causes conjunctival pathology that is visible and facilitates transmission. We captured hatch-year House Finches from two USA populations that differ in their history of pathogen exposure: a Virginia population that has experienced seasonal epizootics for >25 yr and a Hawaii population that is naïve to the pathogen. We then experimentally infected the birds with one of two isolates that varied in virulence. The Virginia birds showed milder distortions of the eye rim, reflected as shorter distances traveled through disease space, across isolates than did the Hawaii birds. Although birds expressed an overlapping suite of pathologic descriptors, the high-virulence isolate caused Virginia birds to express certain pathologies at different frequencies, leading to depauperate communities, compared with the Hawaii birds in which pathologies were expressed more evenly. Notably, eversion was expressed in nearly half of all Virginia eye-days (number of days an eye was sampled) with pathology in response to the high-virulence isolate despite relatively mild lesion severity scores. This may indicate that pathologies that can enhance host competence without compromising host fitness will be maintained and even selected for during host-pathogen coevolution, especially in wildlife populations trending toward disease tolerance.