Premium
Simplified models of anoxia and denitrification in aggregated and simple‐structured soils
Author(s) -
ARAH J.R.M.,
VINTEN A.J.A.
Publication year - 1995
Publication title -
european journal of soil science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.244
H-Index - 111
eISSN - 1365-2389
pISSN - 1351-0754
DOI - 10.1111/j.1365-2389.1995.tb01347.x
Subject(s) - denitrification , soil water , anoxic waters , chemistry , soil science , nitrate , saturation (graph theory) , porosity , aggregate (composite) , biological system , environmental chemistry , nitrogen , environmental science , materials science , mathematics , organic chemistry , combinatorics , composite material , biology
Summary Two complementary approaches to modelling soil anoxia and denitrification are compared. The first postulates a structurally and biologically heterogeneous soil matrix (physical aggregates) and Michaelis—Menten kinetics; the second a random distribution of cylindrical air‐filled pores, a uniform metabolic activity and zero‐order reaction kinetics. Simple functional approximations to both models are developed, allowing their differences to be explored. At water contents corresponding to aggregate saturation the anoxic fractions and denitrification rates predicted by the aggregate model exceed those predicted by the simple‐structure model. In drier soils the predictions of the latter model typically exceed those of the former by between one and three orders of magnitude, the discrepancy lessening as the oxygen reaction potential and the mean radius of the aggregates increase. The aggregate model is much more sensitive to air‐filled porosity, and shows a decreased denitrification efficiency when nitrate concentrations are low. It is likely to predict sharper rainfall‐induced denitrification events, and a smaller background activity, than the distri‐buted‐pore treatment. Whichever of these or other process‐based treatments of denitrification is adopted, simple approximations like those presented here greatly facilitate inclusion in larger‐volume systems models.