The distribution of soil microorganisms is generally believed to be patchy and to reflect habitat heterogeneity. Despite this general rule, the amount of existing data oil species distribution patterns is scarce. Testate amoebae (Protozoa; Rhizopoda) are an important component of soil microbial communities and are increasingly used in ecological and paleoecological studies of Sphagnum-dominated peatlands, but data on the spatial structure of communities are completely lacking. This is an important aspect since quantitative models used for paleoecological reconstruction and monitoring are based on species assemblages. We explored the distribution patterns of testate amoebae distribution in a macroscopically homogeneous Sphagnum carpet, down to a scale of several centimeters. Distributions maps of the species and spatially constrained sample groups were produced. Multivariate and individual spatial autocorrelations were calculated. The importance of spatial structure was quantified by canonical correspondence analysis. Our ultimate goal is to find the finest resolution of environmental monitoring using testate amoebae. The distribution patterns differed among species, resulting in a complex spatial structure of the species assemblage in a whole. Spatial structure accounted for 36% of the total variation of species abundance in a canonical correspondence analysis constrained by spatial variables. This structure was partly correlated to altitude (microtopography) at a very fine scale. These results confirmed the existence of significant broad-and fine-scale spatial structures within restate amoebae communities that could in parr be interpreted as effects of ecological gradients. This shows that, on a surface area of 0.25 m(2), ecological conditions which look uniform from a macroscopic point of view are not perceived as such by Sphagnum-inhabiting organisms. Therefore, restate amoebae could prove very useful to monitor fine-scale ecological processes or disturbances. Studies of the species' spatial distribution patterns in combination with autoecological studies are needed and should be included in the toolbox of biomonitoring itself.