Estimating relative per capita predation rates from molecular gut content analysis

Abstract

The problem of estimating predation rates from molecular gut content data has been challenging. Previous work showed how per capita predation rates could be estimated from quantitative molecular gut content data using the average prey quantity in the predator, the decay rate of the prey in the predator and a conversion constant to convert measured prey quantity into prey numbers or biomass. Based on this previous work, we developed and illustrated a method to estimate relative per capita predation rates for a single prey species consumed by one predator species. This method does not require estimation of either the decay rate of the prey in the predator or the conversion constant. We describe how gut content data from qPCR, quantitative ELISA, metabarcoding and unassembled shotgun reads (Lazaro) can be used to estimate relative per capita predation rates. The method was used to estimate the relative per capita predation rate in a laboratory feeding trial to evaluate the precision and accuracy of the method using Lazaro data. Ten independent estimates were statistically similar, but precision was related to the number of observed prey reads. We estimated the relative per capita predation rate by the ant Pheidole flavens on another ant Pheidole tristis in a field experiment and by the ladybeetle Hippodamia convergens on the aphid Lipaphis pseudobrassicae on organic production farms. We found that higher P. flavens activity-density was associated with lower relative per capita predation rates, therefore indicating lower predation rates on P. tristis. The absence of variation among farms in relative per capita predation by H. convergens suggested that the farms were biological replicates. Using relative per capita predation rates can provide a rapid way to assess how a predator–prey interaction changes over space and time and may help identify factors that limit or enhance biological control of pests.