The geographic and climatic distribution of plant height diversity for 19,000 angiosperms in China

Abstract

The geographic distribution of plant form and function has been studied for over a century for purposes ranging from vegetation classification to global vegetation modeling. Despite this attention we have surprisingly few studies that have actually mapped the distribution and diversity of quantitative plant traits on continental scales and quantified the drivers of these spatial patterns. This limitation has been largely due to the inherent patchiness in trait and spatial databases. Here we analyze the distribution and diversity of plant maximum height in relation to climatic gradients for ~ 19,000 Angiosperm species across China. First, we quantify the relationship between the mean maximum height with climatic variables to test the prediction that precipitation and temperature both should restrict the maximum heights possible in a region. Second, we used null model analysis to address the fundamental question of whether gradients in plant species richness coincide with an increased trait range as expected under limiting similarity theory or whether more species are simply packed into the same range of trait values. The results show that the mean maximum height in a plant assemblage is highest in regions with higher temperatures and annual precipitation indicating that increases in precipitation are enough to offset the concomitant increase in temperature, which was expected to limit plant height. The range and packing of height space were found to increase with species richness and in less climatically variable environments. Null modeling results also show that the deviation of the observed results from expected has a distinct spatial signature for herbaceous and woody plants. Our results highlight plant height diversity, including the range and packing of plant height space, are sensitive to environment, and the mechanisms driving the range and packing of height space in the two growth forms may be different.