Estimation of ATP in soil: extraction methods and calculation of extraction efficiency

Abstract

Methodological problems still exist when soil ATP is estimated using a specific extractant in different laboratories or when different extractants are compared in one laboratory. I hypothesised that this may possibly be caused by uncertainties inherent in addition of an ATP internal standard. Therefore, the reliability of this method was tested using 14 agricultural soils. I chose the TCA/HPO42−/paraquat extractant with sonification (Jenkinson and Oades, 1979) to test whether the recovery of an ATP standard added to the extractant represents the true loss of ATP during extraction process. For this purpose soils were repeatedly extracted with fresh extractant, until the amount of ATP reached the limit of detection in the luciferin–luciferase system. This was reached at the latest after six extraction steps. The summed ATP contents obtained in the six extracts were compared with the amount of ATP as estimated by the conventional way of a single extraction and correction for recovery of an added ATP standard. It was found that the latter procedure considerably underestimated the amount of soil ATP, because the recovery of standards underestimated losses. The ATP contents in the extracts 2–6 during the six-fold extraction procedure were by a factor of 2–9 higher than indicated by losses of added ATP standards after a single extraction. The amount of standard added to the extraxtant did not influence its recovery. It was shown that the application of fresh extractant is the decisive step for obtaining higher ATP values by multiple extraction and not a prolonged sonification. Since a repeated extraction of a soil sample is too time- and labour-consuming for practical application, a more convenient and efficient technique was required. Extraction with dimethylsufoxide (DMSO), Na3PO4 (pH 11.7) and NRB® as proposed by Bai et al. (1988) was found to be rapid and efficient. This extractant gave nearly identical ATP contents of soils as did the repeated extraction with the TCA/HPO42−/paraquat extractant. A useful indicator of ATP extraction efficiency is the biomass C–ATP ratio which was 208–217 for all soils and both methods. This good agreement between both methods, and the use of non toxic chemicals encourage me to propose the DMSO based extraction procedure as a suitable technique to estimate soil ATP in agricultural soils.