Files
Abstract
This research investigates soil variables relevant to growth in loblolly pine (Pinustaeda L.) plantations in the Piedmont of Georgia, USA. Under conditions of completecompeting vegetation control, soil-site productivity relationships were investigated to testwhether nitrogen (N) availability indexes and descriptive soil-site variables were correlated tostand growth. In addition, relationships between soil N availability and the soil organic matterlight fraction (LF) were investigated.I hypothesized that: 1) under simplified forest conditions due to complete competitioncontrol, indices of soil available N will be strong predictors of pine growth and 2) indices of soilavailable N will be positively correlated to the amount of LF in the soil.In plots on Udults, and Udifluvents that received herbicide (H) and herbicide plusfertilization (HF) treatment resin extractable N, and potentially mineralizable N in long-termlaboratory incubations were estimated. Resin extractable and laboratory estimates of Nmineralization were correlated to stand growth and to the quantity of LF. In addition, plots(N=66) under experimental treatment combinations of Control (C), Fertilization (F), H, and HFwere evaluated for changes in soil nutrients and carbon.2Resin extractable N was found well correlated (R=0.7) to mean annual increment(MAI) for 13 out of 17 sampling dates. Based on four sampling dates, resin core estimates of N2availability were also significantly correlated (R=0.5) to the corresponding laboratory estimatesof soil potentially mineralizable N. Another significant relationship was found between the LFpool and resin core estimates of N mineralization (r=0.63, n=25). Finally, the factorial analysisindicates that H, F, and HF treatments can alter the quantity of nutrients other than N with a cleareffect of H treatments on soil carbon and exchange capacity.On a regional scale, when controlling for understory competing vegetation on Nlimited soils, it is possible to improve on soil-site productivity relationships. The significantcorrelation between soil potentially mineralizable nitrogen and resin core estimates of Navailability suggests that both techniques are valuable tools for the prediction of growth on Nlimited soils and that the LF is an important variable affecting N availability.