The over-arching theme of this work is that soil data affect the performance and realism of vegetation models with particular focus on their ability to predict or explain disturbances such as fire or disease. We tested the sensitivity of the Excel version of the 3-PG model to soil properties and applied this information to understanding bark beetle attacks in drought-stressed forests. We tested the sensitivity of the MC2 model to soil depth with a particular focus on how soils affect the biogeochemistry and fire modules of the Dynamic Global Vegetation Model (DGVM). We found in these sensitivity analyses, soil depth, soil water storage capacity (ASW) and soil texture are among the most important soil factors to map and include in models of forest productivity. Chapter 2 shows how soil available water storage capacity has an effect on where pinyon pines are most likely to be stressed during a drought that lasts longer than one year. Chapter 3 shows how ponderosa pines in western Montana are more vulnerable to disturbance on fine-textured soils during years when precipitation is significantly less than in previous years. Finally, Chapter 4 shows that changes in soil depth affect model simulations of productivity and hydrology across the landscape, but fire simulations are only affected in areas with lower precipitation. The value of this research is the knowledge that soil characteristics such as depth, ASW, soil fertility and soil texture can be used in hindcasts of forest disturbances to determine thresholds where forests may become more vulnerable to disturbance in response to rising temperatures or changes in precipitation.
