Small-scale spatial variability of selected soil biological properties

Abstract

A strategy for sampling soil from intact monolith lysimeters was established based on measurements of spatial heterogeneity within the lysimeter area. This was part of an ongoing study to determine relationships between soil microbial diversity and nutrient loss by leaching. The sampling protocol had to allow collection of soil on a regular basis (as opposed to destructive sampling) and ensure high spatial independence of subsamples. On each of the two sites (one developed under organic crop management and the other under conventional crop management), ten 15 cm soil cores (sampling points) were taken from three areas (replicates) of 50 cm diameter (lysimeter surface area) and separately analysed for biotic (microbial biomass carbon and nitrogen; arginine deaminase activity) and abiotic (total carbon and nitrogen) soil properties. The data were tested for variability, expressed as coefficient of variance (biotic and abiotic), and spatial heterogeneity using geostatistics (biotic properties). The biotic soil properties showed significant differences among sampling points, whereas the abiotic parameters were useful in differentiating on a larger scale, i.e. between sites. For all soil properties tested, the differences among the replicates were smaller than those between the sites or among points indicating that, in the main experiment, all treatments can be sampled following the same pattern. Geostatistical analysis and fitting of an exponential model showed that a spatial structure exists in the biotic soil properties and that the samples are independent beyond separation distances of 25–30 cm. A revised sampling pattern consisting of 11 samples per lysimeter is described.