Patterns of life history variation in the Funariales, Polytrichales and Pottiales

Abstract

A combination of Principal Component and Cluster Analyses is used to test for recurring suites of life history variation among 357 species of the Funariales, Polytrichales, and Pottiales. About 85% of among-species covariation in life expectancy, spore size and number, and sexuality is summarized by three Principal Components. Cluster Analyses, using species scores on these three derived variables, identify six groups of species based on similar grades of covariation in life history traits. These groups also differ significantly in water relations, gametophyte size, modes of spore dispersal, and prevalence of asexual reproductive propagules, and apparently are ecologically coherent, at least with regard to spatio-temporal scales of habitat availability and heterogeneity.

A number of differences emerge in comparisons of Principal Component Analyses performed separately for each of seven families. In particular, the distribution of overall variance among successive components varies greatly, suggesting that the number of biologically independent axes of life history specialization differs among families. Likewise, differences in loadings of the life history variables on Principal Components indicate that their patterns of covariation change as a function of familial membership.

In the Pottiales, the likelihood of producing sporophytes decreases with increasing life expectancy, and also shows a negative association with the production of asexual propagules. The likelihood of producing asexual propagules is higher among dioicous species and those not known with gametangia than among monoicous species, and is also positively associated with life expectancy. Alternative historical and ecological explanations are offered to account for these patterns of association.

These results support the idea that recurring suites of life history traits occur among moss species, and the patterns of life history variation appear to be associated with particular sets of environmental circumstances. The influence of phylogenetic history on life history variation is also strong. Until explicitly phylogenetic approaches can be employed, the extent to which observed patterns are adaptive rather than historically contingent must remain unresolved.