The effect of cattle activity on pastures can be subdivided into three categories of disturbances: herbage removal, dunging and trampling. The objective of this study was to assess separately or in combination the effect of these factors on the potential activities of soil microbial communities and to compare these effects with those of soil properties and plant composition or biomass. Controlled treatments simulating the three factors were applied in a fenced area including a light gradient (sunny and shady situation): (i) repeated mowing; (ii) trampling; (iii) fertilizing with a liquid mixture of dung and urine. In the third year of the experiment, community level physiological profiles (CLPP) (Biolog Ecoplates™) were measured for each plots. Furthermore soil chemical properties (pH, total organic carbon, total nitrogen and total phosphorus), plant species composition and plant biomass were also assessed. Despite differences in plant communities and soil properties, the metabolic potential of the microbial community in the sunny and in the shady situations were similar. Effects of treatments on microbial communities were more pronounced in the sunny than in the shady situation. In both cases, repeated mowing was the first factor retained for explaining functional variations. In contrast, fertilizing was not a significant factor. The vegetation explained a high proportion of variation of the microbial community descriptors in the sunny situation, while no significant variation appeared under shady condition. The three components of cattle activities influenced differently the soil microbial communities and this depended on the light conditions within the wooded pasture. Cattle activities may also change spatially at a fine scale and short-term and induce changes in the microbial community structure. Thus, the shifting mosaic that has been described for the vegetation of pastures may also apply for below-ground microbial communities.