Quantifying fine-scale variability in soil available water with hot, dry Douglas-fir ecosystems

Dry, grassy sites within Douglas-fir forests are difficult to regenerate following harvesting due to harsh climatic conditions and intense inter-species competition for limited moisture resources. The objective of this study was to sample soil properties of a recently harvested opening (16 ha) to provide information about post-harvest soil conditions and their relationship with water holding capacity within these dry ecosystems. Utilizing soil properties with established soil water characteristic equations, I predicted soil water holding capacity (SWHC) across the site and at four depths. We completed a topographic survey for the site using aerial light detection and ranging (LiDAR) technology to create a high-resolution (~1m) digital elevation model (DEM). We statistically compared multiple topographic variables with water retention properties via multiple linear regression and geographically weighted regression to determine what drives soil moisture distribution on finer scales than previously studied. Coarse fragments (CF) had the highest amount of variability on the site and altered SWHC the greatest compared with other measured soil properties, with a 10% increase in CF corresponding with a 4.7 mm decrease in SWHC. Additionally, geographically weighted regression was found to outperform multiple linear regression for interpolating the measured soil properties using principle component derived topographic predictor variables. However, the models only explained roughly one half or less of the variability in all soil properties (most R2 ≤ 0.50), thereby suggesting that local soil properties be measured to gain accurate representations of any given site prior to conducting logging or site prescription treatments. To maintain sustainable timber resources within dry Douglas-fir forests in a changing climate, an improved understanding of the soil condition prior to regeneration will become increasingly important.