Inference of biogeographic history by formally integrating distinct lines of evidence: genetic, environmental niche, and fossil

4 April 2019

Hoban, Sean; Dawson, Andria; Robinson, John; Smith, Adam; Strand, Allan

A primary focus of historical biogeography is to understand changes in species ranges, abundance, genetic connectivity, and changes in community composition. Traditionally, biogeographic inference has relied on distinct lines of evidence, including DNA sequences, fossils, and hindcasted ecological niche models. In this review we propose that the development of integrative modeling approaches that leverage multiple data types from diverse disciplines has the potential to revolutionize the field of biogeography. Although each data type contains information on a distinct aspect of species’ biogeographic histories, few studies formally integrate multiple types in analysis. For example, post hoc congruence among analyses based on different data types (e.g., fossils and genetics) is commonly assumed to reflect likely biogeographic histories. Unfortunately, analyses of different data often reach discordant conclusions. Thus, fundamental and unresolved debates continue regarding speed and timing of postglacial migration, location and size of glacial refugia, and degree of long distance dispersal. Formal statistical integration can help address these issues. Specifically, formal integration can leverage all available evidence, account for inherent biases associated with different data types, and quantify data and process uncertainty. Novel, quantitative approaches to integration of data and models across fields are increasingly enabled by recent advances in cyberinfrastructure, spatial modeling, online and aggregated ecological databases, data processing, and quantitative methods. Our purpose is to make the case for, and give examples of, rigorous integration of genetic, fossil, and environmental/ occurrence data for inferring biogeographic history. In particular, we (1) review the need for such a framework; (2) explain common data types and approaches used to infer biogeographic history (and the challenges with each); (3) review state-of-the-art examples of data integration in biogeography; (4) outline a series of novel improvements on current methods; and (5) provide an outlook on technical feasibility and future opportunities.