No abstract

Current computer simulation models do not treat the physical protection of organic matter in soil mechanistically. A model is presented that describes physical protection explicitly as a function of the capacity of clay particles and aggregates to hold organic matter. The net rate of decomposition of organic matter depends not simply on soil texture but on the degree to which the protective capacity of the soil is already occupied. The rate at which organic matter becomes protected depends on both the amount of free organic matter and the degree to which the protective capacity is filled. The rate at which organic matter is released from protection depends only on the amount of protected organic matter. The model closely followed the buildup and decline of organic matter in 10 soils to which grass residues were added each year for a period of 10 yr and then left without addition for a further 10 yr. An estimate made with the model of the maximum capacity of each of these soils to protect organic matter was closely correlated with the clay fraction in the soils. With this model, we were better able to predict the buildup and decline in amounts of soil organic matter in soils of different textures and initial organic matter contents than with conventional, implicit descriptions of protection. Our model accounted for 80% of the variance in the experimental data compared with 77 and 75% using the other two models. Secondly and importantly, the capacity of soil to protect organic matter was found to be positively and well related with the clay content of the soil in our model, whereas similar relationships were not found between soil texture and the parameters controlling protection in the other models.

No abstract

Abstract. The in‐field calibration of a dielectric probe to measure soil water content is described. The probe uses an access tube analogous to that of the neutron probe. The dielectric constant was measured at soil depths of 10, 20, 30, 40, 60 and 100 cm. Cores of soil were then taken from the face of pits dug 30 cm from the access tube and their soil water contents determined by oven drying. The dielectric constant values measured by the probe were calibrated against water contents from these cores. We found that sensor depth needed to be included to achieve a good calibration model that explained 72% of the variance. It is argued that depth needs to be included because of artefacts introduced during the installation of the access tube.