|Abstract or Summary
- Many studies have shown that net primary production in old-growth Douglas-fir/western hemlock forests is lower than in younger forests in similar sites, although the cause is still not clear. One possibility is that overall carbon assimilation, or GPP, is lower in older forests. However, it is difficult to measure GPP on an ecosystem scale, particularly in complex terrain where approaches such as eddy covariance are not possible. One approach is to use process models, although care must be taken to ensure that the model adequately represents "reality". I used a detailed soil-plant-atmosphere (SPA) continuum model, which explicitly links CO2 and H2O fluxes through stomatal conductance. The SPA model has mLlltiple canopy and soil layers where the hydraulic system is modeled as analogue to an electrical circuit. The model accounts for many structural and physiological differences in the stands. I carefully calibrated the model to accurately represent transpiration in a young (ca 25 years old) and an old-growth (ca 450 years old) Douglas-fir/western hemlock stands at Wind River, WA, based on detailed measurements of soil water depletion and sap flow and then used the model to predict variation in GPP among the sites. The modeling approach of this study allowed me to perform sensitivity tests to examine the likely influence of individual structural and physiological parameters on transpiration and GPP. The variables tested were height, above ground leaf specific conductivity, capacitance, leaf area and root vertical distribution, minimum mid-day water potential, canopy water storage, rain interception, root resistivity, and photosynthetic capacity. Within the natural range of variation of these variables, the most influential of these on GPP and Transpiration was height, followed by above ground leaf specific conductivity, and minimum mid-day water potential. By identifying the sensitivity of water and carbon fluxes to the range of natural variation found in these variables, it was possible to examine the likelihood of various hypotheses proposed to explain the age-related decline in productivity. Finally, I modified the model software in order to improve its versatility and make it available to other researchers for use in additional types of forests.