An application of terrain and environmental modelling in a large-scale forestry experiment

Within-site heterogeneity, particularly of soil and climatic variables is a difficulty facing forestry researchers in the design of field trials. Fore st soils are notoriously variable and small changes in soil properties can have a large impact on tree productivity. Likewise, within-site variation i n microclimate is known to influence growth. Within a particular climatic e nvelope, this variation is controlled to a large extent by topography. In t his study, a range of terrain-derived attributes and indices was calculated for each of 40 growth plots from a digital elevation model (DEM) in an att empt to quantify the effects of spatial heterogeneity and establish whether they improved the interpretation of data from a field experiment. In addit ion, selected soil profile measurements were analyzed for their impact on t ree growth and correlated with calculated terrain attributes. The presence of gleying (evidence of waterlogging) in the profile accounted for a signif icant proportion of the residual variation in growth response after treatme nt effects were removed. A combination of predicted erosion index (ei), rel ative available soil water (raswc), dynamic wetness index (dynwet) and tang ential curvature (tancurve) contributed to significant improvements in mode l predictions of growth. A physiologically based model of tree growth (PROM OD) was applied to the site to model the effect of these variables on tree growth. We found that by using a terrain-derived waterlogging or gleying va riable as input, PROMOD could successfully account for variation in tree gr owth due to waterlogging (gleying) and some additional between-plot variati on in growth due to soil depth. (C) 1999 Elsevier Science B.V. All rights r eserved.