Spatial Configuration And Habitat Association Of Plants In A Subtropical Evergreen Broad-Leaved Forest

Climate change,environmental pollution,deforestation and habitat loss caused by human activities have put great pressure on forest sustainable development and biodiversity conservation.In this context,the research on spatial configuration of forest plants,forest habitat and their correlation can provide data support and theoretical guidance for forest management and protection.By analyzing the community characteristics of spatial configuration of plant community,this paper revealed the driving effect of forest habitat on spatial configuration of plant community,and discussed the primary and secondary roles of habitat filtering,dispersal limitation,intra-and inter-specific competition,density dependence and other mechanisms in the process of spatial configuration of plant community.In addition,by comparing the differences of community characteristics after two surveys and the changes of the correlation between community characteristics and habitat factors,the key community ecological processes such as community succession stage,succession direction and succession mechanism were discussed.The biological factors and habitat factors of the community were investigated in typical subtropical evergreen broad-leaved forests based on 10-hm~2permanent plot in Kanghe Nature Reserve of Guangdong province.The results were as follows:(1)The DBH distribution of the community and the dominant species showed inverted"J"shape,which suggestedregeneration was good.Castanopsis carlesii,Schima superba and Itea chinensiswere mainly distributed in the ridge and the northeast of the plot,Castanopsis fargesiiwas relatively concentrated in the middle(mountainside),and Ardisia quinquegonawas relatively concentrated in the foothills and valleys.The spatial distribution pattern of dominant species were complementary to each other,reflecting the effect of habitat heterogeneity and dispersal limitation on species distribution.The dominant species were mainly clustered distribution,and the aggregation become weaker with the increase of scale,and tended to be random and uniform distribution.(2)Based on the variation trend of Sum of squares within groups of k-means clustering,the forest layer was divided into understory,sub-canopy,lower canopy,middle canopy and upper canopy.The number of individuals,average DBH,average tree height,total basal area and biomass of varied significantly among forest layers.The results of non-parametric correlation analysis indicated that the total basal area and biomass in the middle canopy and upper canopy were the main influencing factors of the total basal area and biomass of the community and the understory was the main influencing factor of the number of individuals of the community.The closer the forest layer was,the stronger the correlation of structural factors was.The SAR had better effect on the fitting of stand structure factors than the OLS.Topographic factors had the strongest correlation with stand structure factors,followed by soil chemical factors.Stand structure factors were obviously influenced by habitat heterogeneity,and responses of different forest layers and different structure factors to habitat factors were different,which might be caused by differences in biological characteristics of plants at different growth stages.(3)As the forest layerincreased,species richness,Simpson index and Shannon-Wiener index decreased,and Berger-Parker dominance increased first and then decreased.There were significant differences in species diversity index and species distribution between different forest layers.The closer the forest layer was,the closer the composition and distribution of species were,otherwise,the difference was greater.In addition,the species composition and distribution of community were similar to those of understory,indicating that the species and distribution of community are mainly composed of understory.The indicator species were mainly distributed in understory,indicating that the number and abundance of species in understory exceeded and covered the species in other forest layers.In most cases,the species nestedness component obtained byβdiversity index decomposition was larger than the species turnover component,indicating that the species composition difference between communities was mainly caused by species nestedness variation.The results of CCA showed that 10 habitat factors explained 21.8%of the variance of the community species distribution.The explanation effect of elevation was the strongest,with the contribution rate of explanation variance reaching 42.4%.With the increase of forest layer,the explanatory power of habitat factors on species composition and distribution decreased.(4)The results of v/m method and negative binomial parameter method showed that the spatial distribution pattern of community in different forest layers was aggregation distribution,and the intensity and scale of aggregation decreased with the increase of forest layer.The dominant species mainly showed the aggregation distribution pattern.As the spatial scale increased and the forest layer increased,the aggregation of the dominant species decreased and tended to random distribution and uniform distribution.When the dominant species were in low forest layer and small scale,the spatial association was mainly positive.With the increase of spatial scale and the increase of forest layer,the spatial association decreased continuously,and the positive association tended to be no significant association and negative association.In four types of habitats:foothills and gully,less-steep slope,steep slope and ridge,Castanopsis carlesii showed a significant negative correlation in the foothills and gully,and a significant positive correlation at the ridge.A large number of Ardisia quinquegona trees grew in the foothills and gully,forming distinct niche complementary with the seedlings and saplings of Castanopsis carlesii.(5)In the process of succession,the dominantposition of Castanopsis carlesii and Castanopsis fargesiideclined,while the dominant position of Schima superba,Ardisia quinquegona,and Itea chinensisrose.The results of MRPP showed that the species composition and distribution of communities,understory,and upper canopy had significant differences during forest succession,and there were no significant differences in other forest layers.Spatial autocorrelation analysis and Moran’s I coefficient results showed that the community characteristics have obvious spatial properties.In RDA analysis,soil chemical factors and topographic factors were the key influencing factors of community characteristic distribution.In general,seedlings and saplings in the lower forest layer tend to die off as a result of environmental stress and stress from other individuals.The large trees in the high forest were relatively stable and their biomass increased significantly over time.The large Castanopsis carlesii trees will maintain its dominant position for a long time.However,in the seedlings and saplings,heliophytic tree species such as Castanopsis carlesii and Castanopsis fargesii were gradually replaced by mesophytic tree species such as Ardisia quinquegona,and Aidiapycnantha,and existing communities are at the early and middle stage of succession and tends to the middle and late stage of succession.