Meeting Abstract
The Dilution Effect (DE) hypothesis has played a central role in disease ecology as a way to describe the likelihood of zoonotic disease emergence. The DE states that infectious diseases are less likely to emerge in communities with rich biodiversity because more species diversity can act as a buffer, diluting the number of infected individuals and protecting against emerging infections. The objective of this project was to analyze whether the heterogeneity observed in conclusions in DE hypothesis-based research could be explained by moderators like year of publication and latitude. We performed a meta-analysis using a previously published dataset by Civitello et al. 2015 on vertebrate and invertebrate animal systems and excluded plants. Year of publication could explain some of the variation and graphically reflect changes over time. We were particularly interested in investigating whether there were more studies published supporting the DE hypothesis, since its introduction in the year 2000, suggesting publication bias. A total of 101 animal studies were analyzed using Odds Ratio as the effect size to examine the effect of year of publication estimated 0.0608 (SE=0.025) with marked large heterogeneity (I2=90.30%) using the random effects model. A smaller subset of 25 field studies were used to examine the effect of latitude. We found no significant effect of year of publication or latitude in the smaller data subset. The effect of publication year was significant (p-value = 0.0152) in the full dataset, with an increase in each year more likely to publish research rejecting the DE hypothesis, suggesting potential publication bias in support of the DE hypothesis soon after publication. These results suggest that caution should be taken when implying the DE hypothesis in management decisions based on science conducted more than ten years ago.
