Problems with DNA-based species delimitation with a Bayesian statistical approach because errors with upstream analyses of assigning individuals to putative species

OLAVE, M.; SOLA, E.; KNOWLES, L.

Lisboa

Congreso; XIV Congress of the European Society for Evolutionary Biology; 2013

Genetic-based delimitation of species typically involves a multistep process in which DNA data is analyzed with a sequence of different programs. Although the performance of the programs associated with each step have been evaluated separately, no analysis has considered how errors in the upstream assignment of individuals to putative species impacts the accuracy of delimited species in downstream analyses, such as those associated with the coalescent-based Bayesian statistical program bpp. Here we show that because of differences in the minimal data requirements for accurate performance in each of the different steps involved in the delimitation process, the reliability of inferences about species delimited from genetic data is compromised. Our results provide important insights into the practice of species delimitation by highlighting how uncertainty in the guide tree, as well as individual-species association (both derived from upstream analyses to bpp) can lead to under and overestimation of biodiversity, even though bpp itself may perform very well.