Aims In biological conservation sciences, implementing nature reserves to protect biological diversity is a hot topic. Specifying plant beta diversity, structural components, and influence factors is the critical basis for creating nature reserves.

Methods In this paper, a national nature reserve named Liangshui, located in the central region of the Xiaoxing’an Mountains, was used to survey 80 quadrats in the size of 20 m × 20 m in the core zone, buffer zone, experimental zone, and outside the reserve. We recorded topographic information (Longitude and latitude, altitude, slope, slope aspect, and slope position), community structure (canopy density, tree age, tree height, tree diameter at breast height (DBH), shrub height, and shrub ground diameter), and species name for each tree, shrub, and herb. We also sampled 0-20 cm soils for soil property determination (soil organic carbon, total nitrogen, pH, electroconductibility, soil water content, and bulk density). Abundance data was used for beta diversity calculation as well as computing species abundance differences and species turnover. The Mantel test, redundancy ordination, and variation partitioning were used to examine the effects of abiotic factors (geography and terrain, protection intensity, soil matrix) and biological factors (community structure) on beta diversity and its partitioning components.

Results (1) Species turnover contributed most to the overall beta diversity (Sørensen dissimilarity) for all layers with a range from 65% (for shrub) to 73% (for herb). (2) Mantel correlation analysis indicated that beta diversity, their two components of abundance differences and species turnover, had close associations with the geography and terrain in the arbor, shrub, and herb layer. In contrast, soil properties expressed more association with beta diversity from the arbor and shrub layer. The slope position, slope degree, tree height, and protection intensity were significantly associated with beta diversity in 3 layers (P< 0.05). (3) Variation partitioning and redundancy analysis ordination indicated that geography and terrain affected the beta diversity of whole forests of the reserve significantly, but differences existed between arbor, shrub and herb layers. The arbor beta diversity was most influenced by biological factors. In contrast, the shrub layer was mainly affected by soil factors, twice more than the geography and terrain and biological factors. In the herb layer, the geography and terrain contributed the most to variation, at a rate of respectively 26-fold and 3-fold higher than that from soil factors and biological factors. Tree DBH had the most significant explanatory effect on the beta diversity variations in the reserve.

Conclusion Our findings highlights that nature reserve settlements in the future should consider combining assemblage of the proper community structure of the arbor, shrub, and herb layers, and soil factors, rather than only emphasis on protection intensity. Such steps will improve the conservation efficiency for plant diversity protection.