Functional traits and response guilds have experienced a recent resurgence in popularity in the ecological literature. Linking evolutionary traits functionally to environmental resource gradients and stressors is gaining wide use for informing management of river ecosystems. Along rivers, species possessing similar adaptations for water acquisition, resilience and resistance to disturbance, and strategies for dispersing and recolonizing river margins will respond in similar ways to changes in flow regime. These guilds, once identified, can be modeled using species distribution models to explain current distributions of plants along river margins and to enable us to predict changes in response to projections of future human demands for water, climatically driven changes in flow regime, and alternative flow management scenarios. We provide examples of guild distribution models applied to a range of stream types in the southern Rocky Mountains and in the Sonoran desert. Applications of this framework to quantitative evaluation of vegetation shifts in response to alternative possible futures will be provided along with guidance on: 1) developing trait databases, 2) guilding floras, 3) iteratively refining guild membership, 4) conducting species distribution modeling, 5) assessing model fit, and 6) using models to assess alternative future flow scenarios given climate change or management decisions.
Results/Conclusions
Our findings suggest that the best predictive models can be constructed for specialist guilds that are either intolerant of drought conditions or are intolerant of fluvial disturbance. Several guilds in each flora are comprised of ruderal and generalist species whose distributions are not well-predicted within the riverine environment, such species are transient and fugitive, rapidly becoming established between significant flood events. We found that guild modeling provides an elegant alternative to modeling every single species in a local flora or constraining a study to only those species of particular concern. Further, our models developed from guild-based approaches are transferable to other sites in which guilds are shared but species are not. Examples from streams in the Sonoran desert, Colorado Plateau and southern Rocky Mountains will be presented.