Temperature functions of the rate coefficients of net N mineralization in sandy arable soils Part I. Derivation from laboratory incubations

Abstract

This study aimed to experimentally determine adequate temperature functions for the rate coefficients of net N mineralization in sandy arable soils from NW Germany. Long-term laboratory incubations were carried out in seven sandy arable soils at 3°C, 10°C, 19°C, 28°C, and 35°C in order to derive the rate coefficients of a simultaneous two-pool first-order kinetic equation. Thereby we differentiated between a small, fast mineralizable N pool, comprising mainly fresh residues, and a larger, slowly mineralizable N pool of old, humified organic matter. The rate coefficients were plotted against temperature, and fits of several different functions were tested: Arrhenius, Q10, and multiple non-mechanistic equations. The two derived rate coefficients showed very different temperature functions. Especially in critical temperature ranges (<5/10°C, >30/35°C) common Q10 functions failed to fit well, and, only below 10°C, the Arrhenius functions were in agreement with mean measured rate coefficients. Over the studied temperature range, only relatively complex, multiple equations could adequately account for the observed patterns. In addition, temperature functions that have been derived earlier from loess soils from NW Germany were found not to be transferable to the sandy arable soils studied. Thus, the results strongly question the use of the same Arrhenius or Q10 function or the same rate modifying factor for different N pools as well as for different soils as is generally done in models. Evaluations with field measurements of net N mineralization in part II of the paper (Heumann and Böttcher, 2004) will show which functions perform best in the field.