Grazing impacts on the spatial distribution of soil microbial biomass around tussock grasses in a tropical grassland

The role of grass tussocks in supporting soil microbial biomass (SMB) in gr azed ecosystems is not fully understood, nor is the spatial distribution of SMB in response to different grass species. We undertook a study in 1997 t hat examined fine-scale distributions of SMB in grazed experimental paddock s located in the eucalypt woodlands of northern Queensland, Australia. Leve ls of SMB were determined on soil samples collected at seven locations alon g 60 cm transects in the vicinity of three species of tussock grass (Bothri ochloa ewartiana, Chrysopogon fallax, Heteropogon contortus) and bare space s, in replicate paddocks under five different grazing regimes. Data (n = 28 0) were analysed as a split-split plot in a randomised complete block. Slop e position (block) and paddock management were the main plots, micro-patch type (tussocks and bare spaces) the split plot, and location around micro-p atch the split-split plot. Paddock management, micro-patch, and location ef fects were significantly (p < 0.05) different, as were management x locatio n and micro-patch x location interactions. The highest SMB levels were reco rded at tussock centres on ungrazed (control) and lightly grazed paddocks, with lower levels recorded on degrading (15-27% reduction), and degraded/re covering (40-53% reduction) sites. Successively lower levels were noted fro m tussock centres outwards to the most distant locations (+30, -30 cm) with level of paddock degradation. High levels of SMB were noted around tussock s of B. ewartiana and C. fallax, while the lowest were recorded across bare patches (59% of levels for the above species). Heavy grazing reduced input s of organic materials and carbon into the soil, thereby limiting resources available for microbial growth, Fine-scale monitoring of the plant-microbe -soil interface should be combined with large-scale measures of landscape r esponse to properly describe degradation and recovery processes. (C) 1999 E lsevier Science B.V. All rights reserved.