| Mountain broad-leaved forest which has wide distribution, various types and abundant floristic composition is a relatively stable forest community after a long succession. Moreover, it has an important significance on the conservation of regional water resources and protection of the ecological environment. Currently the study of spatial distribution structure of population spatial correlation between population, and population spatial pattern change in the process of succession has become an important research direction of Southwestern Sichuan mountain broad-leaved forest in the study.The purpose of this study is from the perspective of the rational management to natural broad-leaved forest.throughout analysing the natural broad-leaved forest in mountainous area of Xiaoxiangling in different types of forest (evergreen and deciduous broad-leaved mixed forest, natural secondary forest spatial structure characteristic, focusing on the analysis of two kinds of broad-leaved forest tree species, distribution characteristics and spatial structure characteristics,finding the changes of spatial distribution of population in the process of succession of the community space, in order to further improve the methods and measures of the forest spatial structure, to carry out scientific and reasonable protection of natural resources of mountain forest in Xiaoxiangling, and provide technical support on how to develop and manage the mountain broad-leaved forest.Taking mountain broad-leaved forest of Gongyihai nature protection area in Shimian County, Ya’an as study objects, we discuss spatial distribution pattern in the interior of the population and spatial correlation between population in the plots by using Ripley’s K function. And modeling the correlation between the total height, DBH, the space location by using the model of GWR (geographically weighted regression). Results are as follows:(1) The lower layer(<10m)of the trees in the population showed obvious aggregation distribution between similar primitive forest and secondary forest, but the higher layer (≧10m) showed regular distribution or random distribution in most spatial scales. the result obey the rule that the aggregation degree of species decreased with the increase of height about forest layer.(2) Quercus serrata Thunb. showed regular distribution in most spatial scales, Photinia beauverdiana C. K. Schneid. And Viburnum dilatatum Thunb. showed aggregated distribution or significant aggregated distribution in the given range of spatial scales. Aesculus wilsonii Rehd.is present in the secondary forest, distributed regularly in small scale, tending to random distribution with the increasing of scale, and the inducing effect of environmental factor is small. Quercus acutissima Carruth in two plots showed regular distribution mostly in the spatial scale, but still had some differences. In the near primitive forest, Quercus acutissima Carruth mainly showed aggregated distribution in small spatial scales, and showed regular distribution in medium spatial scales. From the secondary forest plot to the near primitive forest plot, it showed less and less aggregated distribution,and the spatial distribution pattern about the main tree species in forest tended to regular distribution and random distribution.(3) The result showed obviously spatial correlation between Quercus acutissima Carruth and Photinia beauverdiana C. K. Schneid.as its associated tree species in two different types of plots. In near primitive forest, Quercus acutissima Carruth and its associated species Quercus serrata showed obviously negative correlation in the 15-30m scales.(4) In the condition of different scales, the phenomenon of non-correlation increased betweenQuercus serrata Thunb.,Quercus acutissima Carruth, Photinia beauverdiana C. K. Schneid from the secondary forest plots to near primitive forest plot.(5) The results of fitting with GWR model in near primitive forest plot and secondary forest plot are ideal, it showed that the relationship between the total tree height, diameter at breast height, crown width, geographic location is positively correlated. In the three factors of affecting forest tree height, the importance factor ofdiameter atbreast height is the strongest, crown width is second, the least is space location. |
PDF Full Text Request