Structures and topographical pattern of the tree layer of Fagus engleriana-Cyclobalanopsis oxyodon community in Shennongjia area, Hubei Province, China

Shennongjia represents an area of considerable plant biodiversity, not only for China but for the whole world. The numerous species are distributed along an altitude gradient. The genus Fagus, commonly known as Beeches, constitutes one of the dominant woody species of the humid temperate forests in China. This paper deals with the community structure of a mixed, broadleaved deciduous-evergreen beech forest dominated by the Fagus engleriana and Cyclobalanopsis oxyodon, which occurs along an altitude gradient range in this area. The community consists of 46 woody species, belonging to 22 families and 27 genera. The tree layer can be divided into three sub-strata. The upper layer is composed of deciduous trees, 80.7% of which is E engleriana. The second and third layers are dominated by the evergreen species, such as C. oxyodon and the Rhododendron hypoglaucum. These species increase from 55.9% in the second layer to 80.5% in the third, and regeneration at that rate allows us to assume that the stability of the community may be guaranteed. The micro-site heterogeneity plays an important role in maintaining species diversity in plant communities. In this research, a terrain surface of 0.96 hm<'2> was simulated using the digital elevation model (DEM) in order to analyze the influence of the topography on plant community structures on a small scale. The slope relief aspect obtained with the DEM showed an accentuated heterogeneity. Semivariance analysis was used to measure the heterogeneity scale. The range and nugget variance of the semivariograms for slope were 285.8 and 280.5, respectively; for aspect, were 21.8 and 498, respectively. The elevated value of the nugget variance suggested that even on a small scale, the topographic variations influenced the species distribution. Twenty-four species were selected to estimate the correlation coefficient between the slope, aspect, and orientation. Then based on the correlation results using the Square Euclidean Distance cluster analysis, the 24 species were divided into four groups.