Technological innovation combined with greater networking and rising interest in tracking how biodiversity responds to global environmental change has led to a great increase in the availability of large ecological datasets. For many species we have multiple sources of information; for example, traditional survey data may be augmented with GPS tracking, species traits, environmental DNA, or phylogenetic data. Moreover, data for single species can be viewed in the context of matching data for other co-occurring species. However, most methods do a poor job of integrating these different sources of information. This presentation will review recently developed methods for including multiple information sources in biodiversity models targeting both single- and multiple-species outputs.
Results/Conclusions
Two marine case studies show how information sources with very different coverage and calibration (in these cases, GPS trakcing combined with either aerial or shore-based counts) can be combined with meta-information (such as researcher knowledge about survey effort) to improve abundance and uncertainty estimates. In an additional case study, we show that conditioning joint species distribution models with species richness models reduces uncertainty in the predictions of both models. Finally, we present a hypothetical case study where combining occurrences with dynamic time series data fit ina theoretical model can be used to better predict species responses to environmental change.