Species Limits, Phylogeny and Niche Evolution: a Case Study in Carex (Cyperaceae)

Smith, Tyler

ABSTRACT: Two conflicting processes are believed to influence the relationship between taxonomic and ecological diversity: adaptive radiation and niche conservatism. No consensus has yet emerged regarding the conditions that favour conservatism of traits due to common ancestry over adaptive radiation and the resulting ecological divergence. Carex is an ideal group for studying these phenomena, as it is incredibly diverse taxonomically and ecologically, and includes many recently formed lineages. I chose a group of eight closely related Carex taxa from eastern North America as the focus of my study. Multivariate analysis of 30–33 morphological variables measured on 456 plants from 110 populations confirmed the existence of eight different entities. Analysis of 141 AFLP fragments from 435 plants from the same populations revealed the same eight groups. I recognise each of these taxa as distinct species. These include C. roanensis, a globally rare species whose taxonomic status had been questioned, and C. complanata and C. hirsutella, a pair of species that some authors had considered a single variable taxon. Six previously unknown sterile or nearly sterile hybrids are documented with a combination of morphology, restriction fragment, and AFLP data. Parsimony and distance analyses of the AFLP data failed to fully resolve the phylogeny of this group, but did establish consistent estimates of the relative evolutionary distances among species. Morphology does not reflect the phylogenetic relationships in this group. There were significant differences among species for niche measures calculated from climate variables, ecological niche models, and local habitat variables. Niche divergence was significantly correlated with phylogenetic distance, indicating niche conservatism is an important process governing ecological evolution in this group. Furthermore, the shape of the relationship suggested that phylogenetic distance limits the maximum possible ecological divergence, but within this limit other unmeasured factors also contribute to the observed pattern. Comparing my results to other recent work, I propose a framework for identifying the factors that influence the strength of niche conservatism.