Spatial patterns are an important consideration in soil animal community ecology, as they form the base for pattern-process research and the mechanism of biodiversity maintenance in the soil animal community. Spatial patterns of soil animals have been studied over several years, but remain unclear at small scales. Spatial explicit research based on geostatistics helps to reveal the spatial patterns of soil animals at different spatial and temporal scales. This study was conducted to investigate the dynamics in the spatial patterns of adult Coleoptera communities in a temperate deciduous broad-leaved forest at Maoershan Forest Ecological Station, Heilongjiang Province. Three replicates (5m×5m plots for Sample A, Sample B, and Sample C) were set up. Each sample plot was divided into 25 squares with 1m×1m subplots. Geostatistical methods were used to reveal the spatial patterns dynamic of Coleoptera communities. Pitfall traps were used to collect the ground-adult Coleoptera beetles. This study was conducted in August and October in 2014 and in June and October in 2015. Global Moran's I index was used to explain the spatial autocorrelation of the abundance of Coleoptera communities and groups. Spatial semi-variogram was used to determine the spatial heterogeneity of the abundance of Coleoptera communities and groups. Furthermore, a combination of a cross-variogram was performed to identify the spatial relationships between different Coleoptera groups, while the ordinary kriging method was used to map the spatial pattern of the abundances of Coleoptera communities and species. The main results are as follows:the survey captured 1021 individuals belonging to 11 families and 29 groups. Coleoptera adult communities showed significant spatial and temporal variations at different survey times. The coefficient of variation indicated that the Coleoptera communities had strong spatial variability at all survey times. Moran's I coefficient results showed that the abundances of adult Coleoptera communities and a relatively small proportion of the Coleoptera groups at each survey time had significantly positive spatial autocorrelations. The spatial model could be fitted by Spherical, Exponential, Gaussian, and Linear models. According to the C/C₀+C values, spatial heterogeneities of the communities and groups were mainly influenced by random factors or a combination of structured and random factors. The structure proportions of some groups showed a pure nugget effect resulting from sampling errors or spatial variability occurring within the minimum distance interval (1m). For each community, the results of cross-variogram showed that the spatial relationships between different adult Coleoptera groups were complex at multiple scales, and non-existent, positive, or mostly negative spatial relationships were all observed. The C/C₀+C values demonstrated that the spatial differentiation of adult Coleoptera groups were mainly influenced by structured factors or random factors. Maps of ordinary kriging interpolation showed that the spatial patterns of Coleoptera communities were significantly different at different survey times. These results showed that adult Coleoptera communities obviously aggregated in spatial distribution at a small scale (5m). Spatial relationships between different groups were complex and unstable. In summary, the above-ground Coleoptera adult community has obvious spatial heterogeneity at the small scale, and the dynamic spatial heterogeneity is obvious. However, this study was conducted at a fine scale of 5m×5m, and studies at both larger and much finer spatial scales are needed.