Changes in capture availability due to infection can lead to correctable biases in population-level infectious disease parameters

Abstract

Correctly identifying the strength of selection parasites impose on hosts is key to predicting epidemiological and evolutionary outcomes. However, behavioral changes due to infection can alter the capture probability of infected hosts and thereby make selection difficult to estimate by standard sampling techniques. Mark-recapture approaches, which allow researchers to determine if some groups in a population are less likely to be captured than others, can mitigate this concern. We use an individual-based simulation platform to test whether changes in capture rate due to infection can alter estimates of three key outcomes: 1) reduction in offspring numbers of infected parents, 2) the relative risk of infection for susceptible genotypes compared to resistant genotypes, and 3) change in allele frequencies between generations. We find that calculating capture probabilities using mark-recapture statistics can correctly identify biased relative risk calculations. For detecting fitness impact, the bounded nature of the distribution possible offspring numbers results in consistent underestimation of the impact of parasites on reproductive success. Researchers can mitigate many of the potential biases associated with behavioral changes due to infection by using mark-recapture techniques to calculate capture probabilities and by accounting for the shapes of the distributions they are attempting to measure.