Soil microbial biomass and its activity estimated by kinetic respiration analysis – Statistical guidelines

The use of kinetic respiration analysis to determine soil microbial biomass its active part and maximum specific growth rate has recently increased. With this method, the increase in soil respiration rate initiated by application of carbon growth substrate, e.g. glucose, and mineral nutrients is used to estimate parameters describing microbial growth in soil. This study refines the method by developing statistical guidelines for the data analysis and processing. The kinetic respiration analysis assumes that microbial growth is not limited by substrate and energy. That is why it is critically important to identify the time period corresponding to the unlimited growth. In this work, we studied how the unlimited growth phase can be identified in less subjective ways by examining 121 datasets of respiration time series of 44 different soil samples taken from field plots. Deflection of the respiratory curve from the exponential pattern indicates growth limitation. Subjective selection of the part of respiratory curve which fits to the exponential pattern resulted in a 30% bias in specific microbial growth rates. We propose rules that are based on inspecting the patterns in a series of plots of residuals of fitted respiration rate. By comparing those rules with a set of statistical criteria we find that the weighted-coefficient of determination (r(2)) can be used to objectively constrain the unlimited growth phase in those cases where double-limitation does not occur. Furthermore, we discuss how the uncertainty of estimated microbial parameters is influenced by a) measurement uncertainty, b) biased measurement at the beginning of the experiment, and c) the number and timing of respiration measurements. We recommend checking plots of fits and residuals as well as reporting uncertainty bounds together with the estimated microbial parameters. A free statistical package is provided to easily deal with those aspects. (C) 2011 Elsevier Ltd. All rights reserved.