Baited traps as flawed proxies for carcass colonization

In both fundamental and applied sciences, the use of surrogates to measure phenomena that are challenging to study directly is a common practice. However, this requires validating the appropriateness of the surrogates. This study examines if traps, used to measure flight activity of necrophagous flies, can serve as effective surrogates for predicting oviposition on whole carcasses, a topic still under debate in forensic science. We used three sets, a calibration and validation subsets comprising monitoring data of the flight activity of four necrophagous blow fly species, and a test set comprising the oviposition activity of these species on carcasses. Each set also included measurements of abiotic parameters. Using Random Forest for each species, we quantitatively and qualitatively modeled flight activity as a function of abiotic parameters and validated these models. However, when we examined the extent to which flight activity predicted oviposition on carcasses, the models performed poorly, only explaining a fraction of the variance. As the first study making use of small baited trap data to model oviposition on animal carcasses, this study presents mixed results that suggests that traps, despite their utility in addressing various forensic entomology questions, currently appear to be unreliable proxies for predicting carcass colonization.