Movement of faecal indicator organisms in earthworm channels under a loamy arable and grassland soil

Abstract

It is a critical issue for the application of animal slurry in water protection areas whether macropores such as earthworm burrows can contribute to the contamination of groundwater with faecal organisms. The aim of the current field experiment was to investigate the effects of the factors: slurry addition, land use (grassland/arable land), sampling date, depth (15, 45 and 75 cm), and drilosphere (soil surrounding an earthworm burrow) on the movement of faecal indicator organisms (total coliforms, faecal coliforms and faecal streptococci) into the subsoil. The experiment was carried out under worst case conditions, i.e., a soil at field capacity, irrigated directly after application of fresh cattle slurry with high rates of water. Under these conditions, we showed that significant numbers of faecal coliforms were transferred after slurry addition down earthworm burrows to the subsoil. We also showed that the numbers transferred in the grassland soil exceeded those of the arable soil. However, only about 0.16% (=1.3×10 3 m −2) of the applied faecal coliforms were found under grassland around the earthworm burrows at a depth of between 40 and 80 cm. Total coliforms or faecal streptoccoci were not affected by slurry addition at any depth. In the drilosphere, the numbers of all faecal indicator organisms were always significantly increased in comparison to the bulk soil and also the contents of microbial biomass which were similar to those of the surface horizons. This means that the competition with the autochthonous soil microflora and thus, the decay rate of the introduced faecal bacteria, were also similar to the surface soil. Total and faecal coliforms were significantly related to each other and to the total microbial biomass, indicating that a certain percentage of these two groups belongs to the autochthonous soil bacterial microflora. The number of faecal streptococci varied enormously between replicate plots and the variation appeared to be independent of the factors studied.