A phylogenetic examination of the nearest-living-relative method for reconstructing paleoclimate at mammalian fossil localities [Examen phylogénétique de la méthode du plus proche parent vivant pour la reconstruction paléoclimatique des localités de mammifères fossiles].

Résumé

Overlapping climatic tolerances of extant taxa are commonly used in paleoclimate reconstruction. This method is limited by the stratigraphic ranges of living taxa. Paleoclimate hypotheses have been extended deeper in time through assuming that the nearest living relatives of extinct taxa provide appropriate climatic analogs. These methodologies implicitly assume that the climatic tolerances of taxa do not change through time, and the nearest-living-relative method assumes that there is a phylogenetic control on the climatic tolerances of evolving lineages. The assumptions of the nearest-living-relative methodology can be tested for mammal lineages through a phylogenetic analysis that examines the association of various important climatic parameters with various taxa. Accordingly, this study applied a phylogenetic approach to understand the relationship between climatic features (such as maximum and minimum January and July temperatures, annual precipitation, and seasonality) and evolution of extant species within Marmota and Microtus. Species within these genera have been considered important paleoclimate indicators due to their abundances, distinctive dental characters, and present restriction to certain microhabitats. The association between phylogenetic hypotheses based on molecular data and climatic parameters, both derived from published records, was tested using Felsenstein's independent contrast method. The technique involved mapping the climatic tolerances of various species onto the hypothesized phylogenetic relationships and thereby determining if groups of species were delineated by identifiable climatic boundaries. Preliminary results indicate a correlation between phylogeny and climate tolerances for Microtus species endemic to North America. If additional analysis substantiates these results, the nearest-living-relative method could prove to be an important tool in reconstructing the paleoclimates of some fossil mammal localities.