Abstract:[Objectives] Ground-dwelling macroarthropods are important indicators of soil ecosystem. Studying the characteristics of community of ground-dwelling macroarthropods in alpine meadows is helpful to reveal the response of its distribution pattern to the change of alpine meadow ecological factors. [Methods] From May to September 2018, we investigated ground-dwelling macroarthropods community composition and diversity by pitfall traps at four typical alpine meadow habitats (swamp wetland, swamp meadow, meadow and mature meadow). The principal coordinate analysis (PCoA) was used to explain the similarities or differences in ground-dwelling macroarthropods community from different habitats. Permutational multivariate analysis of variance (PerMANOVA) tests the significance of community distinctions among habitats and sampling time. Variance partitioning analysis (VPA) was used to evaluate the soil environmental factors, above ground biomass and sampling time on ground-dwelling maroarthropods community. The influence of specific soil environmental factors on the dynamics of ground-dwelling maroarthropods community structure was analyzed by distance-based redundancy analysis (db-RDA). [Results] A total of 2 545 individual ground-dwelling macroarthropods were collected from May to September 2018, belonged to 30 genera of 29 families, 9 orders, 3 classes. The dominant groups were Lycosa, Camponotus and Formica, and predators and herbivores were dominant in trophic groups (Table 2). The number of ground-dwelling macroarthropod groups in habitat of mature meadow was significantly higher than that of habitats of swamp meadow and meadow (P < 0.05); the Pielou evenness index (Js) of habitat of swamp wetland was significantly higher than that of habitats of meadow and mature meadow (P < 0.05); Shannon-Wiener diversity index (H′) had no significant difference (P > 0.05) among habitats, and the similarity index of ground-dwelling macroarthropod communities between habitats was low (Table 3). Among the Checkerboard score of the matrix (C-score), variance ratio (V-ratio), standardized effect size based on binary data, the Pianka niche overlap, the difference between the observed and simulated values have not regularity, suggesting that the ground-dwelling macroarthropods community assemblage was randomness and competition coexist, and that is structured simultaneously by both random diffusion and ecological filtering (Table 5). PerMANOVA analysis showed that the habitat characteristics of the study area (R2 = 0.636 4, P < 0.001) could more explain the changes of the ground-dwelling macroarthropods community characteristics than seasons (R2 = 0.157 3, P < 0.001). The characteristics of the ground-dwelling macroarthropods community was affected significantly by soil physicochemical properties, of which soil temperature and soil moisture content have the greatest impacts (Fig. 2). The distance based on redundancy analysis (db-RDA) showed that the soil physicochemical properties and aboveground biomass explained 29.88% of the variation of the ground-dwelling macroarthropods community, while the other 70.12% of the variation was not explained, indicating the complexity of the formation of the ground-dwelling macroarthropods community (Fig.3). [Conclusion] The result showed that the ground-dwelling macroarthropods community changed with different habitats and seasons. The effects of habitats on ground-dwelling macroarthropods composition were much greater than those of seasons.