Trait diversity in plant communities maintained by competition for water and light

Ecological communities frequently exhibit remarkable taxonomic and trait diversity, and this diversity is consistently shown to regulate ecosystem function and resilience. However, ecologists lack a synthetic theory for how this diversity is maintained when species compete for limited resources, hampering our ability to project the future of biodiversity under climate change. Water-limited plant communities are an ideal system in which to study these questions given (1) the diversity of hydraulic traits they exhibit, (2) the importance of this diversity for ecosystem productivity and drought resilience, and (3) forecast changes to precipitation and evapotranspiration under climate change. We developed an analytically tractable model of water and light competition in age-structured perennial plant communities and demonstrated that high diversity is maintained through phenological division of the time between storms. We modeled a system where water arrives in the form of intermittent storms, between which plants consume the limited pool of soil water until it becomes dry enough that they must physiologically shut down to avoid embolism. Competition occurs because individuals, by consuming the shared water pool, cause their competitors to shut down earlier, harming their long-term growth and reproduction. When total precipitation is low, plants in the model compete only for water. However, increases in precipitation can cause the canopy to close and individuals to begin competing for light. Variation among species in the minimum soil water content at which they can sustain growth without embolizing leads to emergent phenological variation, as species will shut down at varying points between storm events. When this variation is paired with a trade-off such that species that shut down early are compensated by faster biomass accumulation, higher fecundity, or lower mortality, there is no limit to the number that can coexist. These results are robust to variation in both total precipitation and the time between storms. The model therefore offers a plausible explanation for how hydraulic trait diversity is maintained in a wide array of natural systems. More broadly, this work illustrates how the phenological division of an apparently singular resource can emerge because of common trade-offs and ultimately foster high taxonomic and trait diversity.