Biogeographically distinct controls on C 3 and C 4 grass distributions: merging community and physiological ecology

Abstract

Aim C 4 photosynthesis is an adaptation that maintains efficient carbon assimilation in high‐light, high‐temperature conditions. Due to the importance of C 4 grasses for carbon and surface energy fluxes, numerous models have been proposed to describe their spatial distribution and forecast their responses to climate change. These models often rely on broad climatic predictors (e.g. temperature and precipitation) but fail to integrate other ecologically relevant factors like disturbance and competition, which may modify realized C 3 / C 4 grass distributions. Here, we present a combined evaluation of the contribution of ecological factors and climatic predictors to realized C 3 / C 4 grass distributions. We consider multiple biogeographic regions of N orth A merica using a multisource database of over 40,000 vegetation plots. Location The conterminous U nited S tates of A merica ( USA ). Methods We identified a comprehensive pool of climatic models in the literature and used information theoretic criteria to select a primary climatic predictor of C 3 and C 4 grass distributions. Subsequently, the best model was combined with ecological predictors (e.g. fire, tree cover) using a multiple regression framework and tested within eight regions. Results Surprisingly, grass‐dominated communities across the USA exist largely in C 3 ‐ or C 4 ‐dominated states. Transitions between C 3 / C 4 dominance were best explained by models that integrated temperature and precipitation with ecological factors that varied according to region. For some regions, like E astern T emperate F orests, local ecological factors were comparable in strength to broad‐scale climatic predictors of C 3 / C 4 abundance. Main conclusion Local ecological factors modify C 3 / C 4 grass responses to broad‐scale climatic drivers in ways that manifest at regional scales. In E astern T emperate F orests, for example, C 4 grass abundances are maintained below climatic expectations where tree cover creates light limitation but above expectations where frequent fires reduce tree cover. Thus, local ecological factors, which vary among biogeographic regions, contribute to large‐scale climate disequilibrium.

Bibliographic Information