TY - JOUR

T1 - Node-based analysis of species distributions

AU - Borregaard, Michael Krabbe

AU - Rahbek, Carsten

AU - Fjeldså, Jon

AU - Parra, Juan L.

AU - Whittaker, Robert James

AU - Graham, Catherine H.

PY - 2014

Y1 - 2014

N2 - The integration of species distributions and evolutionary relationships is one of the most rapidly moving research fields today and has led to considerable advances in our understanding of the processes underlying biogeographical patterns. Here, we develop a set of metrics, the specific overrepresentation score (SOS) and the geographic node divergence (GND) score, which together combine ecological and evolutionary patterns into a single framework and avoids many of the problems that characterize community phylogenetic methods in current use.This approach goes through each node in the phylogeny and compares the distributions of descendant clades to a null model. The method employs a balanced null model, is independent of phylogeny size, and allows an intuitive visualization of the results.We demonstrate how this novel implementation can be used to generate hypotheses for biogeographical patterns with case studies on two groups with well-described biogeographical histories: a local-scale community data set of hummingbirds in the North Andes, and a large-scale data set of the distribution of all species of New World flycatchers. The node-based analysis of these two groups generates a set of intuitively interpretable patterns that are consistent with current biogeographical knowledge.Importantly, the results are statistically tractable, opening many possibilities for their use in analyses of evolutionary, historical and spatial patterns of species diversity. The method is implemented as an upcoming R package nodiv, which makes it accessible and easy to use.

AB - The integration of species distributions and evolutionary relationships is one of the most rapidly moving research fields today and has led to considerable advances in our understanding of the processes underlying biogeographical patterns. Here, we develop a set of metrics, the specific overrepresentation score (SOS) and the geographic node divergence (GND) score, which together combine ecological and evolutionary patterns into a single framework and avoids many of the problems that characterize community phylogenetic methods in current use.This approach goes through each node in the phylogeny and compares the distributions of descendant clades to a null model. The method employs a balanced null model, is independent of phylogeny size, and allows an intuitive visualization of the results.We demonstrate how this novel implementation can be used to generate hypotheses for biogeographical patterns with case studies on two groups with well-described biogeographical histories: a local-scale community data set of hummingbirds in the North Andes, and a large-scale data set of the distribution of all species of New World flycatchers. The node-based analysis of these two groups generates a set of intuitively interpretable patterns that are consistent with current biogeographical knowledge.Importantly, the results are statistically tractable, opening many possibilities for their use in analyses of evolutionary, historical and spatial patterns of species diversity. The method is implemented as an upcoming R package nodiv, which makes it accessible and easy to use.

U2 - 10.1111/2041-210X.12283

DO - 10.1111/2041-210X.12283

M3 - Journal article

VL - 5

SP - 1225

EP - 1235

JO - Methods in Ecology and Evolution

JF - Methods in Ecology and Evolution

SN - 2041-210X

IS - 11

ER -