Long-term soil potassium availability from a Kanhapludult to an aggrading loblolly pine ecosystem

A long-term (1962 to 1990) forest biogeochemistry study in the southeastern Piedmont of the USA provided estimates of soil K release in response to fo rest regrowth. We investigated the sources of soil K that buffered the exch angeable K pools during forest growth and we estimated soil K release rates through greenhouse and acid extraction studies for comparison to our field estimate. In these acid Kanhapludults, derived from granitic-gneiss, the disparity be tween measured depletions of soil exchangeable K and estimated forest remov als indicated a buffering of exchangeable K on the order of 0.31 kmol(c) ha (-1) per year. Nonexchangeable K extracted by boiling with 1 M HNO3 exceede d exchangeable K by up to 40-fold. Non-exchangeable K was not depleted duri ng the three decades of stand growth, however, thus was not the long-term s ource of exchangeable K buffering. Total K in these soils ranged from 0.4 t o 3.8% by weight. Mineralogical data indicated a presence of hydroxy-interl ayered vermiculite throughout the upper 4 m of soil for <2 mu m clay fracti on and a presence of micaceous minerals in the 2 to 45 mu m silt fraction. XRD analysis of micaceous flakes extracted from 4 to 8 m in the soil indica ted a presence of muscovite mica. Estimated K releases in the greenhouse and extraction studies were generall y consistent with long-term results. The accumulation of K during two rotat ions of pine seedling growth in the greenhouse exceeded the measured deplet ions in exchangeable and non-exchangeable K over all soil depths tested by 0.007 to 0.026 cmol(c) kg(-1). Potassium removal by sequential extraction/i ncubations with 1 mM HCl and 1 mM oxalic acid continued through 24 extracti ons and K recovered in extract solutions exceeded the sum of depletions in exchangeable and non-exchangeable K pools by 0.001 to 0.028 cmol(c) kg(-1). These excess removals in plant uptake or solution recovery indicate a rele ase of mineral K. Thirty-day extractions with H+-resins in both 1 mM HCl an d 1 mM oxalic acid were well fit by the Elovich equation but were not well correlated with plant K uptake in the greenhouse study. The release rate co efficients ranged from 0.012 to 0.025(cmol(c) kg(-1)) h(-1). Extrapolations to annual releases of K in the greenhouse and sequential ext raction studies were a similar order of magnitude as long-term releases est imated at the long-term Calhoun plots. Surface soil (0 to 15 cm) releases r anged from 0.15 to 0.65 kmol(c) ha(-1) per year while deeper soils ranged u p to 1.54 kmol(c) ha(-1) per year. Results indicate that soils similar to t hose at Calhoun that contain a similar micaceous and HIV component will be able to supply K at rates adequate to keep pare with demands of forest regr owth even under intensive forest management. (C) 2000 Elsevier Science B.V. All rights reserved.