Microsite Influences on Variability in Douglas-Fir Seedling Development

Abstract We examined the microsite characteristics of 6,048 Douglas-fir seedlings at three regeneration sites in Washington state. Our objective was to determine the microsite characteristics that were most influential on seedling growth change over time. We analyzed microsite influences both individually and in concert with one another through regression-based techniques. Microsite parameters included soil impedance, topographic, and physical parameter measurements that were recorded at each seedling's location. Akaike's information criterion (AIC) was used to determine combinations of microsite parameters that were most strongly correlated with seedling growth. Multiparameter models explained between 15 and 39 percent of the variance in diameter growth. Prevalent terms from the strongest multiparameter models included soil penetration, log presence, stump presence, skid road presence, and topography. Individual microsite parameters for each regeneration site were also assessed for importance in explaining diameter growth using two additional methods. The first approach was to isolate the parameters that appeared in the strongest multiparameter models and to sum and contrast the AIC weights of all models in which they appeared. The second approach was to regress single parameters against seedling diameter growth. Results varied by site for both methods. AIC weight sums revealed that topographical depression and berms, the presences of logs and stumps, and soil penetration (pounds per square inch) as measured by a penetrometer were most influential, with values ranging from 0.31 (berm) to 0.82 (log). Regression analysis revealed that topographical depression, log presence, and soil penetration were significantly related to diameter growth, explaining between 6 and 29 percent of the variance in diameter growth. Combined results from the three regeneration sites suggest that preferred planting locations are near berms, in the transition zone associated with skid roads, and in soil that is neither too loose nor too compacted. Results from the Randle and Orting sites indicate that planting in topographical depressions should be avoided. Results from Orting indicate that seedlings should not be placed near logs, and Randle findings suggest not planting next to stumps.