Spatial Pattern Of An Aged Quercus Forest In Typical Karst Landform Area And Its Environment Interpretation

Understanding the processes underlie spatial patterns of trees is fundamental for understanding the formation and maintenance mechanisms of species coexistence.The Guangxi Yachang Orchid National Nature Reserve,located in the southern-subtropical China,is a typical karst area in which limestone is mainly made up of rock,with highly heterogeneous habitat.Species composition is highly rich,but with an obviously vertical stratification.In this study,a permanent standard fixed plot with an area of 110 m×200 m was established in an aged Quercus forest which composed of Quercus fabri and Quercus variabilis.The pair correlation function,g（r）was used to analyze the spatial pattern and interspecific associations of main populations and different layers before and after excluding the effort of habitat heterogeneity.Meanwhile,marked correlation function,k_mm（r）,was used to analyze the characteristics of tree mark.Moreover,the canonical correspondence analysis（CCA）was used to explore the impact of environmental factors on species distribution and its correlation with species diversity.The main results were listed as follows:（1）A total of 4596 trees and shrubs whose diameter at breast height（DBH）was equal to or larger than 1 cm occurred in the aged Quercus forest.6 common tree species Quercus fabri,Platycarya strobilacea,Quercus variabilis,Lyonia ovalifolia,Keteleeria davidiana,Rhus chinensis,Quercus fabri dominated the forest and had important value of 33.85%,13.63%,12.50%,8.91%,3.97%and 3.24%,respectively.The forest had a typical reversed"J"shape of diameter distribution,and its main populations had a continuous distribution of DBH.（2）When considered the effort of habitat heterogeneity（completely spatial random,CSR）,the dominant tree species were mainly aggregated at r=0-30 m,while were randomly distributed at large scale（r=30-50 m）.In vertical layers,the distribution pattern of the whole community,upstory and understory gradually turned into random from clumped with the increase of scale（r=0-35m）,and had a regular pattern at large scale（r=35-50 m）.The patterns of the main tree species and vertical layers were randomly distributed at most scales and aggregated at small scale（r=0-5 m）when excluded the effort of habitat heterogeneity（heterogeneous Poison process,HP）.（3）The size differentiation of the main populations was not obvious,and different-sized tree occurred together.The size differentiation in the upstory was significant.Small-sized trees tightly gathered at r=0-15 m,while trees of different sizes attracted each other at large scales（r=27-50 m）.Tree size in the understory differentiated slightly,indicating high pressure from the upstory.The concentration degree of tree species decreased with the increase of observation scale,showing a slight negative correlation at large scale（r=27-50 m）,however,the aggregation intensity of the upstory was significantly higher than that of the understory.（4）The spatial correlation of main tree species was complex（CSR）,and positive and negative correlation accounted for the main part,while the proportion of mixed type was low.Most of the time,interspecific exclusion had a higher proportion than that of interspecific attraction.For the vertical layers,the spatial correlation between the upstory and understory gradually changed from positive（r=0-23 m）to random（r=24-29 m）and then to negative（r=30-50 m）with the increase of scale.The main tree species,upstory and understory had no spatial correlation at most scales if the effort of habitat heterogeneity was excluded.（5）The PCA（principal component analysis）analysis well explained the species distribution of the aged Quercus species forest.The area of exposed rock,soil,elevation and slope had extremely significant effects on species distribution（p<0.01）using the analysis of CCA,but convexity has no significant effect on the species distribution.In addition,altitude was extremely significant correlated with the density of standing live trees and Pielou index（p=0.001）,and significantly correlated with species richness（p<0.05）.Slope was only significantly correlated with species richness（p<0.05）.Aspect was significantly correlated with species diversity index and Pielou index（p<0.05）.Concavity was significantly correlated with density and Pielou index（p<0.05）.The interactions among plants and among environmental factors are important factors that affect species distribution patterns.The markedly environmental heterogeneity of karst area has a strong influence on the distribution patterns of the aged Quercus forest.Individuals of different life types and sizes have different adaptability and spatial distribution to different environmental factors.