Effects of land use on the level, variation and spatial structure of soil enzyme activities and bacterial communities

Abstract The aim of this study was to evaluate the long-term influence of contrasting rural land use types on the level, plot-scale variation and horizontal spatial structure of decomposition activities and the bacterial community in soil. Experimental data were collected in the southern boreal zone from topsoil layers of adjacent spruce forest, unmanaged meadow (former field) and organically cultivated field that all shared the same soil origin. The forest soil was sampled separately for the organic and mineral layers. A geostatistical design comprising 50 sampling points per plot area of 10 × 10 m 2 was used. The measured microbiological characteristics included eight different hydrolytic soil enzyme activities involved in C, P and S cycles, bacterial 16S rDNA length heterogeneity profiles (LH-PCR) and total DNA yield as a relative estimate of microbial biomass. Effects of land use were pronounced on both the bacterial community structure and soil enzyme activities. Soil organic matter (SOM) content predicted well the major differences in soil enzyme activities and microbial biomass. Highest enzyme activities were generally found in the forest organic soil whereas the underlying mineral soil showed significantly lower activities with a pattern similar to those of the other mineral soils, especially the cultivated field. Bacterial LH-PCR fingerprints were distinct but at the same time remarkably similar between field and meadow soils whereas the forest organic layer differed clearly from the mineral soils. Within-plot variation of soil microbiological characteristics was best explained by the variation of SOM. Relative standard deviations of soil microbiological characteristics typically decreased in the order: forest organic layer ≈ forest mineral layer > meadow > field. However, bacterial fingerprints showed lowest variation within the meadow. Most of the microbiological variables studied showed no or only weak spatial structure at the scale sampled.