CAJ | 학술논문

Eleven soil types, which can be identified and delineated using conventional soil survey procedures, were characterized for loblolly pine (Pinus taeda L.) productivity. Four 4-hectare study sites, each containing four measurement plots, were established for every soil type studied. In a stepwise multiple regression, both soil parent material (i.e. a combination of subsoil texture and geology) (p<0.001), and drainage class (p=0.006) were significant predictors of site index (tree age 25), and the overall linear regression model had an R2 value of 0.55. The extremes of soil parent material differed by 3.9 m site index (loamy subsoil on the Wicomico-Penholoway surfaces versus clayey subsoil on the Pamlico-Princess Anne surfaces). Each increment of drainage class differed by 0.7 m site index. For example, a poorly drained soil had 0.7 m lower site index than a somewhat poorly drained soil. For seven of the eleven soil types studied, there is greater than 80% probability that estimated mean site index is within ±0.8 m of the actual soil type mean site index. The other four soil types (labeled G, I, C and K) need to be either redefined or sampled more intensively. Two of these need to be subdivided in order to adequately characterize site quality, one based on geology (Soil type G) and one based on soil drainage class (Soil type I). Variation in soil drainage class and varying amounts of topsoil displaced into windrows were both factors influencing site quality variation of a third soil type (Soil type C). The wide variation in site index data for a fourth soil type (Soil type K) appeared to be due, in part, to sampling study locations and individual measurement plots with less than optimum bedding and/or artificial drainage. Soil parent material (subsoil texture and geology) along with drainage class were found to be important factors influencing site quality on the South Carolina Lower Coastal Plain.