Phylogenetic evidence for a rapid radiation of pleurocarpous mosses (Bryopsida)

We tested the hypothesis that the Hypnales, the Hookeriales, or both, underwent a rapid radiation early in their phylogenetic history.

Diversification rates (speciation minus extinction) were investigated using lineage-through-time (LTT) plots and alternative models of diversification were tested using the statistical approach of Paradis. We used sequence data from two chloroplast loci (rps4 and trnL-trnF) from 241 taxa in the Ptychomniales, Hookeriales, and Hypnales. One of the MP trees was arbitrarily selected for the analysis of biodiversity patterns in the pleurocarps (Fig. 1).

As in previous analyses, the shape of the tree suggests a rapid radiation of hypnalian mosses because of relatively short branches within the order. The three pleurocarpous orders differ in species richness; Ptychomniales: 100 species, Hookeriales: 743 species, and Hypnales: 4418 species. The data set included 71 Hookeriales and 149 Hypnales, yet the Hookeriales included more phylogenetic diversity (PD; Faith 1992) than the Hypnales (Table 1). We estimated hypnalian PD from 100 subsamples of 71 taxa for a more direct comparison with our hookerialian sample and found that, on average, the Hookeriales contained twice the PD as did the Hypnales (48% vs 23%).

The number of segregating nucleotide sites (P), two estimates of theta (theta-S & theta-pi?), and nucleotide diversity (theta-pi / # sites) suggest similar patterns: Hookeriales encompass more molecular diversity than Hypnales despite the higher species diversity of Hypnales (Table 1). Indeed, even the Ptychomniales, which contain only 100 species, and which were represented by only 21 samples, contained higher nucleotide diversity than did the Hypnales.

Tests of different diversification models assume that sampling within clades is exhaustive at the species level. To test the effect of sampling on the shape of LTT plots and the statistical tests of DIVERSI we implemented simulations. Our phylogenetic sample of Hypnales consisted of 141 taxa, or approximately 4% of the total hypnalian species. We generated 100 replicate phylogenies under a constant diversification rate with 4418 extant taxa (the number of species recognized in the order), then randomly pruned all but 141 taxa from the resulting trees and constructed plots from these subsamples to generate a mean LTT curve with 95% confidence intervals (Fig. 2). The analyses were intended to simulate a constant diversification rate phylogeny for the Hypnales, with subsequent random sampling of 4% of terminals for diversification analyses. A similar procedure was conducted for the Hookeriales, but with 743 extant taxa and all but 71 randomly pruned from the trees to generate LTT plots simulating our roughly 10% sampling of hookerialian taxa (Fig. 3). We evaluated the empirical LTT plots by comparison with the simulated curves and their 95% confidence intervals. In addition, we used DIVERSI (E. Paradis) to fit the best diversification model to each of the 100 simulated phylogenies in order to assess the robustness of the statistical algorithms to incomplete sampling. We were able to reject a constant diversification rate for Hypnales (despite incomplete sampling), but not for the Hookeriales.

We concluded that the Hypnales underwent an initial period of rapid diversification that lasted approximately 20% of their history. Moreover, a model of diversification characterized by two discretely different rates provided a better fit to the data than either a constant rate or a gradually changing rate, based on likelihood ratio tests and the Akaike Information Criterion (AIC) implemented by the program DIVERSI.

Table 1. Phylogenetic Diversity (PD) and molecular diversity in pleurocarpous mosses

Reference:

Shaw, A.J., C. J. Cox, S. B. Boles, & B. Goffinet. 2003. Phylogenetic evidence for a rapid radiation of pleurocarpous mosses (Bryopsida). Evolution 57: 2226-2241. reprint

<--- TOP

 Shaw Laboratory
 139 Biological Sciences Bldg.
 Box 90338
 Department of Biology
 Duke University
 Durham
 North Carolina 27708
 U.S.A.

 Phone: (+1) 919 660-7345
 Fax: (+1) 919 660-7293