Evaluation of Relationships Between Oak Site Indices and Properties of Selected Dystrochrepts

Abstract

AbstractSoils derived from argillaceous rocks in the Blue Ridge and Valley and Ridge in Virginia have been mapped and correlated as members of Typic (Berks) and Lithic (Weikert) Dystrochrepts. SCS‐SOI‐5 interpretation records show soils of the Berks series to have higher estimated upland oak site indices than soils of the Weikert series. To explain observed differences in distributions of upland oak site indices, distributions of 34 soil and site properties were evaluated. According to the Shapiro‐Wilk test, the majority of these properties had nonnormal distributions. Therefore, α levels associated with the parametric, two‐sample t‐test would be unreliable. The nonparametric Wilcoxon rank sum test is distribution free and provides reliable α levels for small samples with nonnormal distributions if the samples have similar dispersions. According to the Moses ranklike test, only available water capacity (AH2O) had significantly different (p = 0.05) dispersions for Typic and Lithic Dystrochrepts. Typic Dystrochrepts had significantly larger (p = 0.05) median values for depths to C and R horizons, AH2O, and AL3+ saturation of the cation exchange capacity (CEC). Lithic Dystrochrepts had significantly larger median values for exchangeable Ca2+, total exchangeable bases, CEC, pH‐dependent charge, exchange acidity, ratio of the CEC to clay, and feldspar. Observed median values for upland oak site indices of 56 and 52 for Lithic and Typic Dystrochrepts, respectively, were not significantly different. Significantly larger amounts of feldspar and exchangeable bases, especially Ca2+, and significantly lesser amounts of exchangeable Al3+ saturation of the CEC of Lithic Dystrochrepts are hypothesized as reasons for their slightly higher upland oak site indices. Significantly larger amounts of exchange acidity in Lithic Dystrochrepts are the result of larger amounts of pH‐dependent charge and probably do not adversely influence tree growth. The simultaneous consideration of selected combinations of these properties using discriminant analysis does not produce mutually‐exclusive groups or groups that correspond to limits defined for Lithic and Typic Dystrochrepts. Only the limit of 50 cm depth to bedrock produces classes that correspond to these subgroups.