The large observed functional diversity in grasses and their importance in carbon cycling and interactions with climate motivate accurate process representations in land models. Current strategies to disaggregate grass diversity in models likely over-simplifies functional diversity with potentially large effects on predictive capability. In this talk we review the current modeling approaches to represent carbon, nutrient, water, and energy cycling in grasses and motivate the need for a novel, integrative framework that reorganizes grass vegetation types around phylogeny-driven functional diversity. We will highlight processes and parameterizations that differ between the traditional plant functional type (PFT) approach and a new representation organized around evolutionary relationships, the lineage functional type (LFT). We describe modeled processes that affect vegetation structure and competition (between other plants, microbes, and abiotic processes), because these processes are critical to prediction of current and 21st century grass functioning and carbon cycling and can be parameterized based on existing trait observations.
Results/Conclusions
We find that modeled grass species should be disaggregated into at least into two LFTs containing most of the dominant C4 grasses (Andropogoneae (warm, wet, and seasonally burned) and Chloridoideae (warm and semi-arid), and one dominated by C3 grasses (Pooideae (cool and dry)). Using two models of differing complexity, we describe how differences between traits associated with structure, phenology, reproduction, and disturbance recovery affect growth and carbon cycle dynamics in several representative regions.