Quantitative Evaluation And R Programming Of Forest Spatial Structure Based On The Relationship Of Neighborhood Trees

Quantitative descriptions and analyses of forest structure have recently become common tools in the modern management of forests. Indices of forest spatial structure based on the relationship of neighborhood trees, such as the mingling, dominance and uniform angle index, are widely used in both Chinese and foreign studies of the analysis of forest spatial structure, calculations of competition and dominance, measurements of species diversity, reestablishment of forest structure and optimization of management. Analyzing forest spatial structure based on the relationship of neighborhood trees has laid a foundation for the study of the characteristics of spatial structure, but the analysis of forest spatial structure is becoming increasingly common, so the theory and practice need further supplementation and development. The establishment of the indices, exploration of methods of quantitative evaluation and development of analytical tools are the major themes in the study of forest spatial structure.Ten typical secondary forests at the Huoditang region in the mid-altitudes of Qinling Mountains are the research subjects in the present study. We identified the mechanisms of community assemblages of secondary forests in the region based on species composition and the distribution of species abundance, established an index system of forest spatial structure and a method of quantitative evaluation and developed the forestSAS R package for analytical systems of forest spatial structure.(1) We presented an index system for the evaluation of forest health at the stand level based on vigor, organization and resilience of the VOR model for evaluating ecosystemic health, including biomass distribution, stand density, tree diversity, diameter distribution, tree height distribution, successional characteristics, tree quality and the distributional pattern. The index system provided a theoretical framework for identifying the indices of forest spatial structure.(2) We developed two new indices of spatial structure, “Differentiation” and “Ideal state”, based on the relationship of neighborhood trees, which can characterize the variation and social status of tree attributes, respectively.(3) We established an index system for total forest spatial structure, which included species mingling, storey differentiation, succession ideal state, tree quality ideal state, crowding, uniform angle index, diameter dominance and biomass dominance.(4) We proposed the concept and calculation method of “Preference value”, which can identify the trend from actual to ideal forest spatial structure. The index average value evaluation(IAVE) and index preference value evaluation(IPVE), two evaluation models of the heterogeneity of forest spatial structure, were developed based on the preference value for comprehensively evaluating forest spatial structure and providing a scientific basis for forest management plans and measures.(5) We presented a new systematic and effective survey method for individual trees based on the conditions of crowns, stems and roots combined with the survey method for crown condition of the Forest Health Monitoring Program for North America and established a uniform system for evaluating tree quality. These methods can contribute to the surveying and evaluation of tree quality in China.(6) A total of 50 tree species belonging to 30 genera in 16 families were identified among 5686 trees(DBH ? 5 cm) in 50 plots(totaling 4.5 hm2) in the 10 typical secondary forest stands in two forest belts in the mid-altitudinal zone of the Qinling Mountains. Twentyfive plots in the birch belt contained 2934 trees of 43 species(27 genera, 16 families). Twenty-five plots of the pine-oak belt contained 2752 trees of 41 species(28 genera, 14 families). Four species, Quercus aliena var. acutiserrata, Pinus armandii Franch., Toxicodendron vernicifluum(Stokes) F. A. Barkl. and Carpinus turczaninowii Hance, had the broadest distributions, irrespective of forest type. The dominant species in the birch belt were Betula albo-sinensis Burk., P. armandii, Acer davidii Franch. and T. vernicifluum. The dominant species in the pine-oak belt were Q. aliena var. acutiserrata, P. tabuliformis, P. armandii and T. vernicifluum. Both types of belts had rich species compositions and similar floristic components but clearly different community structures.(7) The values of IAVE were between 0.467 and 0.611, with an average of 0.549, and the values of IPVE were between 0.431 and 0.592, with an average of 0.529, both of which indicated that the most of the heterogeneity of spatial structure of the secondary forests was in the average class, and only a few of the forests were in the good class. These results indicated that the heterogeneity of spatial structure was very low, and the effect of the ecological restoration was not ideal.(8) We developed the forestSAS R package, which includes 22 R functions that can help users to calculate the indices and to evaluate the heterogeneity of forest spatial structure. The main functions of the package include edge correction, nearest-neighbor analysis, calculation of spatial-structure indices, evaluation of spatial-structure heterogeneity, cutting simulation and vegetation establishment. The package has the advantages of being open source, simple, effective and easy to use and maintain.