Assessing forest growth across southwestern Oregon under a range of current and future global change scenarios using a process model, 3-PG

With improvements in mapping regional distributions of vegetation using satellite-derived information, there is an increasing interest in the assessment of current limitations on forest growth and in making projections of how productivity may be altered in response to changing climatic conditions and management policies. We utilised a simplified physiologically based process model (3-PG) across a 54000 km(2) mountainous region of southwestern Oregon, USA, to evaluate the degree to which maximum periodic mean annual increment (PAI) of forests could be predicted at a set of 448 forest inventory plots. The survey data were pooled into six broad forest types (coastal rain forest, interior coast range forest, mixed conifer, dry-site Douglas-fir, subalpine forest, and pine forest) and compared to the 3-PG predictions at a spatial resolution of 1 km(2). We found good agreement (r(2) = 0.84) between mean PAI values of forest productivity for the six forest types with those obtained from field surveys. With confidence at this broader level of integration, we then ran model simulations to evaluate the constraints imposed by (i) soil fertility under current climatic conditions, (ii) the effect of doubling monthly precipitation across the region, and (iii) a widely used climatic change scenario that involves modifications in monthly mean temperatures and precipitation, as well as a doubling in atmospheric CO2 concentrations. These analyses showed that optimum soil fertility would more than double growth, with the greatest response in the subalpine type and the least increase in the coastal rain forests. Doubling the precipitation increased productivity in the pine type (> 50%) with reduced responses elsewhere. The climate change scenario with doubled atmospheric CO2 increased growth by 50% on average across all forest types, primarily as a result of a projected 33% increase in photosynthetic capacity. This modelling exercise indicates that, at a regional scale, a general relationship exists between simulated maximum leaf area index and maximum aboveground growth, supporting the contention that satellite-derived estimates of leaf area index may be good measures of the potential productivity of temperate evergreen forests.