The use of DNA for inferring species boundaries offers great promise. However, limitations with the current methodologies pose serious obstacles. The accuracy of such inferences can be questionable because determining a species identity often depends on the type of data used. This project will apply recent theoretical and methodological advances developed in population genetics to determine species limits. The Caribbean cricket genus Amphiacusta will be used to explore ways in which population genetic models and alternative sampling designs can improve the accuracy of inferred species boundaries when different types of data and different rates of divergence exist among populations.
The methods developed by the project will enable improved sampling design for a diversity of approaches used in species delimitation, including those applied in conservation biology and DNA-barcoding initiatives. The results of these analyses will also produce recommendations about how genetic data, including data from multiple loci, can be combined with other data types to produce a unified approach for recognizing species. These tools will be of use for a wide range of taxa, and are expected to impact a large cross-section of disciplines given the central role that species, and their delimitation, play in biology and society at large.
The project identified how combining different data types can provide not only the resolution for identifying species boundaries that would otherwise go undetected by relying on DNA exclusively, but also additional insights about how species diverge. For example, with the empirical application in the Caribbean cricket genus Amphiacusta, studying divergence across genetic, ecological and morphological characters identified how such factors are associated with the different stages of species divergence. Such information has important implications for deciphering how the sampling design and approach used to recognize species will affect the accuracy of species delimitation. In addition to recommendations about how genetic data, and specifically data from multiple loci, can be combined with other non-genetic data, collaborative efforts produced a Bayesian statistical framework for such analyses, as well as valuable biodiversity data, including the addition of specimens to museum collections and deposition of DNA sequences in the public database Genbank. Presentation of the work in seminars and workshops has also helped stimulate additional research that includes both empirical applications (including to groups organisms of conservation concern) and methodological developments that will complement future efforts, such as the expansion of DNA-barcoding initiatives to multi-locus data. The project provided broad training to both students and postdoctoral fellows. This included molecular skills with next-generation sequencing technologies and familiarity with the population genetic underpinnings of analytical procedures. This training can be applied in many other scientific endeavors unrelated to the goal of species delimitation. For example, such pursuits from students and postdocs that received training from the funded project include studies on the impact of climate-induced distributional shifts, tests of the role of habitat heterogeneity in shaping biodiversity patterns, and study of the consequences of conservation management decisions on population genetic structure.
