Evaluating soil extraction methods for chemical characterization of ultramafic soils in Kinabalu Park (Malaysia)

Abstract

Soils derived from ultramafic bedrock are known for hosting distinct vegetation types as a consequence of atypical soil chemistries consisting of high trace elements concentrations (Ni, Cr, Co) and exchangeable cation imbalances (high Mg:Ca quotients); the high Mg:Ca ratio causes Ca infertility in plant species not adapted to ultramafic soils. Ecological studies use a range of single-stage extraction methods for chemical characterization of such soils in order to be able to interpret plant response, and ultimately to explain plant community composition. Few studies to date have compared different soil extraction methods in relation to tropical ultramafic soils. This study compares eight commonly used extraction methods on a large number of ultramafic soil samples collected from Kinabalu Park (Malaysia). The tested methods were: for trace elements: NH4Ac, DTPA, CaCl2, Sr(NO3)2 and Mehlich-3, for exchangeable cations: NH4Ac and silverthiorea, and for plant-available phosphorus: Mehlich-3 and Olsen-P. These single-stage extraction methods were compared and evaluated for predictive power for chemically characterizing soils, interrelatedness and ecological application. The methods were also contrasted with a sequential extraction scheme. Finally, several operational parameters including molar ratio (0.01 and 0.1 M CaCl2, Sr(NO3)2) and pH buffering (DTPA-TEA) were also evaluated. The majority of single-stage extraction methods are highly inter-correlated and predictive power could be improved by including independent soil parameters (pH, CEC, pseudo-total element concentration) in the multivariate regression equation. Ecological interpretation remains difficult because of lack of experimental studies in relation to plant uptake response and potential phytotoxicity effects on tropical native plants from ultramafic soils.