Detecting treatment differences in soil carbon and nitrogen resulting from forest manipulations

Forest management practices may alter soil properties, but experimental evaluation of treatment effects is often difficult because large soil variability yields poor statistical sensitivity. This study was conducted to determine if pretreatment soil sampling enhances our ability to detect treatment differences in pacific Northwest forests. We used statistical simulation and data from 271 pretreatment 30-cm-deep soil cores taken from 70- to 100-yr-oId post-fire Douglas-fir [Pseudotsuga menziesii (Mirb,) Franco] stands at the Siskiyou National Forest in southwestern Oregon. At this site, an unmanipulated control and six treatments of overstory species and woody debris are being established on 6-ha plots and replicated in three blocks. If only post-treatment measurements were made, the minimum detectable difference (MDD) between extreme treatments would be 26 to 57% of current soil C and N concentrations (g kg(-1)) and masses (kg m(-2)), with C generally having smaller relative MDDs than N, In contrast, if the differences between post- and pretreatment were evaluated, relative MDDs for C properties would remain the same, but those for N would decrease by as much as one-half because N has greater pretreatment variability among plots than C. Masses of C and N, which incorporate information about rock volume and either soil mass or bulk density, have relative MDDs as small as one-hair of those associated with concentrations. Evaluation of other sites is required to determine if these observations are consistent for forests in general and, thus, applicable to the design of sampling strategies in large-scale experiments.