Evidence that spatial scale and environment factors explain grassland community assembly in woodland-grassland ecotones.

How communities of living organisms assemble has long been a central question in ecology. The impact of habitat filtering and limiting similarity on plant community structures is well known, as both processes are influenced by individual responses to environmental fluctuations. Yet, the precise identifications and quantifications of the potential abiotic and biotic factors that shape community structures at a fine scale remains a challenge. Here, we applied null model approaches to assess the importance of habitat filtering and limiting similarity at two spatial scales. We used 63 natural vegetation plots, each measuring 5 × 5 m, with three nested subplots measuring 1 × 1 m, from the 2021 field survey, to examine the alpha diversity as well as beta diversity of plots and subplots. Linear mixed-effects models were employed to determine the impact of environmental variables on assembly rules. Our results demonstrate that habitat filtering is the dominant assembly rules at both the plot and subplot levels, although limiting similarity assumes stronger at the subplot level. Plot-level limiting similarity exhibited a positive association with fine-scale partitioning, suggesting that trait divergence originated from a combination of limiting similarity and spatial partitioning. Our findings also reveal that the community assembly varies more strongly with the mean annual temperature gradient than the mean annual precipitation. This investigation provides a pertinent illustration of non-random assembly rules from spatial scale and environmental factors in plant communities in the loess hilly region. It underscores the critical influence of spatial and environmental constraints in understanding the assembly of plant communities.