The quantification of plant biodiversity through time

Abstract

Existing quantitative summaries of the taxonomic diversification of vascular plants are not in numerical agreement, although they permit similar broad evolutionary interpretations of the history of land plants. We examine three aspects of the collection and analysis of such data which demonstrate that quantitative studies are still in their infancy. Comparisons among diversification patterns reflected in two data sets reveal that comparatively minor taxonomic and stratigraphic revisions result in statistically significant quantitative changes as well as qualitative differences in the patterns of diversity change over geological time. Two mathematical models based on three data sets are used to estimate the ‘unsampled’ diversity of fossil plant species. These techniques indicate that considerably more data need to be collected before disparities among vascular plant diversification patterns obtained by different authors can be brought to closure. The question of what constitutes the smallest sampling unit of morphological diversity (i.e. the taxonomic level that yields statistically independent observations) remains unresolved for vascular plants. Nested analysis of the percent distribution of variance for three plant morphological variables (height, tracheid diameter, and xylem cross-sectional area) indicate that the taxonomic level at which major evolutionary innovations become apparent has shifted from the species—genus level to higher taxonomic levels during the early radiation of pteridophytes. Such shifts may have occurred during the taxonomic radiation of other broadly defined plant groups (gymnosperms and angiosperms), and thus may have altered the taxonomic level on which selection pressures have operated. This may influence our interpretation of the mechanisms of diversity response to environmental variation.