Alan R. Townsend1, Gregory P. Asner2, Cory C. Cleveland1, Sasha C. Reed1, and William R. Wieder1. (1) University of Colorado, Boulder, (2) Carnegie Institution
Tropical forests are among the most diverse, productive and threatened biomes on the planet, and as such are critical to understanding a wide range of both basic ecological principles and processes with high social and economic relevance. The past few years have seen considerable discussion about what forces structure highly diverse tropical forest communities, but the ecosystem-level consequences of such structure remain very poorly known. A classic concept in ecosystem ecology is that gradients in soil nutrient availability derived from substrate or geomorphological characteristics create similar variations in plant nutrient content, and evidence for such substrate signals certainly exists in multiple ecosystems, including tropical forests. However, some recent studies have suggested that even local scale variation in soil characteristics is a determinant of community structure in tropical forests. Here, we present evidence that the principal source of chemical variation in diverse tropical forest canopies is a product of biology not geology; detailed surveys of foliar N, P, Ca, Mg and K in multiple tropical forests across several soil types show that species-specific determinants of foliar nutrients create far more variation in the tropical canopy than do even major shifts in soil order. Thus, we argue that local scale variation in soil nutrient content is most likely a consequence, not a cause, of non-random community structure. In addition, we show that local scale differences in foliar chemistry among neighboring tree species on the same soil type lead to distinct patterns in key ecosystem processes such as free living nitrogen fixation and soil carbon turnover.