USGS2019
The principal motivation for this study is that sagebrush-steppe ecosystems are undergoing significant state changes, and land managers are challenged with optimizing their resources for both short- and long-term use. Yet, limited knowledge is available regarding how the sagebrush-steppe will respond to environmental changes related to precipitation and temperature regimes, and disturbance such as fire. Furthermore, there is a lack of understanding on how fuels reduction and other fuel management activities will impact these ecosystems over the long-term. We addressed these challenges by adapting and testing a vegetation dynamics model, the Ecosystem Demography v2.2 model (EDv2.2), for the sagebrush-steppe. Vegetation dynamics models can provide estimations of ecosystem productivity in their natural and disturbance states, and thus serve as a tool to understand and predict potential changes in various processes and properties of vegetation communities. Yet, there is no vegetation dynamics model that is well-developed for the sagebrush-steppe, and thus significant effort is needed to test EDv2.2 for its application. As part of our efforts to develop the EDv2.2 model into a useful tool for the sagebrush-steppe, we developed a sagebrush plant functional type (PFT) as part of this study, and then performed sensitivity analyses, model calibration, and finally model evaluation. Furthermore, we developed several model scenarios under natural (undisturbed) and disturbed (fire) environments. We compared our model outputs with ground-based data (field and eddy covariance) and remote sensing observations. The results of our project include a sagebrush PFT that can be used in both future EDv2.2 modeling efforts and other vegetation dynamic models. Our results from the model sensitivity analysis indicate that specific leaf area (SLA), stomatal slope (STO_S), cuticular conductance (CUT_C), and carboxylase rate constant (VM0) are sensitive parameters to vegetation productivity in the model (based on gross primary production, GPP), and future modeling efforts will benefit from both lab and field studies of these parameters and sensitivity analyses. Through calibration, we found that the EDv2.2 model estimates of GPP were modeled well at our lowest elevation field site in Reynolds Creek Experimental Watershed (RCEW), which is dominated by Wyoming big sagebrush. On the contrary, we found poorer results at higher elevation site shrub sites. These sites are characterized by either low sagebrush or mountain big sagebrush, and have more forb cover than the low elevation site. In this project we also implemented the fire model in EDv2.2 to explore how shrub and C3 grasses respond to fire by analyzing post-fire GPP. We ran both point and regional model runs with fire introduced. In most fire scenarios, fire substantially reduced shrub GPP and it took several decades for shrub GPP to return to pre-fire conditions. Grass GPP responded more quickly in post-fire conditions. While these processes are representative of what other studies have found, significant efforts to improve the fire processes in EDv2.2 are needed. For example, nuances associated with the fire subroutine in the model (running periodic fire events versus instantaneous fires and fire intensity) will need to be expanded. Another significant contribution to our knowledge gap is that additional PFTs to represent the sagebrush-steppe (e.g. annual grasses such cheatgrass) are needed for EDv2.2. Regardless, this project made significant advances in PFT development and model testing. Moreover, the EDv2.2 provides a useful framework to conceptualize vegetation dynamics, project future conditions, and consider fire as a disturbance. With additional parameterizations, PFTs, and fire routines, EDv2.2 will evolve as a tool for which to better understand future ecosystem dynamics of the sagebrush-steppe.