PLANT LIFE FORM, STOMATAL DENSITY AND TAXONOMIC RELATEDNESS - A REANALYSIS OF SALISBURY (1927)

1. The results of Salisbury (1927) with regard to interspecific patterns of stomatal density have lead to much theorizing as to the causes for the apparent differences. However, Salisbury treated each species as an independent data point, a practice which may not be valid given that similarity can result from taxonomic relatedness independent of ecological effects. 2. In reanalyses of Salisbury's data for stomatal number, we found that the patterns upon which Salisbury based his conclusion that stomatal density is correlated with the degree of 'exposure' of a species were not upheld when taxonomic relatedness was taken into account. Specifically, we found stomatal density to be greater in shrubs than trees, in trees than herbs and in marginal herbs than understorey herbs, but no significant difference between shrubs and herbs, or woody plants (trees and shrubs pooled) and non-woody plants from the same habitat type. 3. In an additional analysis using data only for one life-form from one habitat type (herbs from the forest margin), we found no difference in stomatal density between amphi- and hypostomatous species.