COS 67-6 - Evolutionary lineage, not plant functional type, explains the most trait variation in grasses

Background/Question/Methods

There is an increasing appreciation for the role of evolutionary history in driving patterns of trait variation across plants. Consequently, there are concerns regarding the utility of plant functional types (PFTs) as a means to group grass species in grassland ecosystem modelling and synthesis analyses. Instead, grouping grasses by evolutionary lineage in lineage-based functional types (LFTs), may produce more accurate models and better represent grass functional diversity. To explore the role of lineage in shaping suites of traits, we measured 11 structural and physiological traits of 75 species of grasses occurring at the Konza Prairie Biological Station (KPBS; Manhattan, KS, USA). We tested whether grouping species based on their photosynthetic pathway (as done in PFTs), life history (annual/perennial), or lineage (tribe or C4 lineage) explained the most trait variation across a diverse suite of species in natural habitats experiencing similar ambient climate histories.

Results/Conclusions

Trait variation among the 75 species of grasses varied significantly by photosynthetic pathway, life history, and evolutionary lineage. Photosynthetic pathway primarily explained physiological traits (δ13C, osmotic potential, C:N, and Vcmax; p < 0.05) and life history primarily explained structural traits (SLA, LDMC, maximum flowering height, maximum vegetative height, C:N, leaf thickness, and osmotic potential; p < 0.05). Critically, we found that grouping grass species by their lineage explained the most variation in all types of traits, both physiological and structural (δ13C, Vcmax, SLA, LDMC, maximum flowering height, maximum vegetative height, C:N, leaf thickness, and osmotic potential; p < 0.05), as these traits may be conserved evolutionarily. Thus, there is limited value of models that group species by photosynthetic type if they include structural traits, such as SLA and LDMC, in their dataset. Our results support a move away from using PFTs and towards a lineage-based approach for grassland ecosystem modelling. This strategy will hopefully improve how grass diversity is represented in these models and increase accuracy in making predictions of how grassland ecosystems will change in the future.