The escalating awareness of non-forested landscapes and realization that more emphasis is needed for an all lands approach to management increasingly requires timely information to improve management effectiveness. The Forest Vegetation Simulator (FVS) has been used in a large number of studies to project future vegetation conditions and is often used in conjunction with the richly populated Forest Inventory and Analysis (FIA) database. The FVS lacks the routines and algorithms for dealing with non-forest landscapes and is therefore only applicable on a limited number of landscapes. This obviates the need for simulation capabilities in non-forested environments, in addition to improved characterization of understory conditions (herbaceous and shrub species) in the FVS. In response to this need, we developed the Rangeland Vegetation Simulator (RVS) and new models for estimating understory conditions in forested landscapes. The RVS is calibrated on 112 unique sites and enables simulation of ecological dynamics, production and fuels in either a spatially explicit manner or as a processor of inventory data much like the FVS. Validation of the RVS, in this inaugural development, suggests significant promise for its use to describe vegetation and fuel data when the structure and composition are given, but its ability to describe succession is limited and in some cases unrealistic. The premier outputs of the vegetation simulator are: 1) Standing biomass, carbon, and annual production of herbs and shrubs (including standing dead herbaceous material). 2) Vegetation structure, composition, and seral stage 3) Fuelbed properties (1, 10, 100, 1000 hr fuel) and fire behavior fuel models 4) Response of these attributes to herbivory and fire.
The Ecological dynamics underlying the RVS are LANDFIRE’s Biophysical Settings (BPS) models. These models do not account for invasive species and other contemporary ecological theory. As a result, in this project, a prototype application using Ecological Sites and their associated state-transition models were merged with the RVS fuel, growth and biomass algorithms to test the efficacy for describing fuel bed properties. This was funded as a separate project but is really part of this larger effort to improve non-forest simulation and management Results indicate significant promise for Ecological Sites replacing the BPS models of succession for a national strategy involving rangeland ecological simulation in support of rangeland management and administration.
A final component of this project was development of equations to predict understory response to overstory and site factors in forested environments, especially focused on herbaceous and shrub structure. The understory equations were developed for 4 major forest types covering approximately 44 million of acres in the U.S. including: lodgepole pine, Douglas fir, grand fir, and ponderosa pine. McFadden statistics range from 0.22 for herb height to 0.87 for tally tree cover of the Pacific Northwest FIA Region. Predictions of shrub cover and height are relatively more accurate than those of herbaceous species.