Temporal And Spatial Differentiation Of Forest Landscape In Relation To Naturalness

This study was conducted in the Wuliangshan National Nature Reserve and its adjacent area in Yunnan to fill up the study gap on temporal and spatial differentiation of forest landscape in relation to naturalness, and to provide close-to-nature forest management and sustainable forest management with theoretical and technical basis at landscape level, with its coverage on the current status of forest landscape and characteristics of temporal and spatial differentiation of forest landscape in terms of naturalness. The methods and theories applied in the study included landscape ecology, plant community theory, quantitative ecology, forest management, gray system theory, and remote sensing technology, geographical information system technology, global positioning system technology, plant community survey method, analytic hierarchy process (AHP), gray related degree analysis, non-quantitative landscape classification, point sampling technology, canonical correspondence analysis (CCA), detrended canonical correspondence analysis (DCCA), landscape structure analysis, Markov Model etc.Findings from the study are the following:(1) The study area maintains relatively natural status. Forest naturalness is broadleaved evergreen forests > deciduous broadleaved forests > coniferous forests in terms of species group gradient, nature forests > artificial forests in terms of their origin gradient, and close to mature and mature forests > young and middle aged forests in terms of age group gradient. It is basically in line with the actual forest naturalness, and indicates that the fact indicated that the selection of index for calculating naturalness was reasonable.(2) Criteria for the 2 levels classification of forest naturalness were established by using a method combining qualitative and quantitative approaches with more persuasion than by using onlyqualitative or quantitative criteria.(3) Criteria for the 3 levels classification of forest landscape are instituted based on naturalness, the 3 levels classification of forest landscape are conducted and quantitatively measuring the area of 3 levels forest landscape are realized.(4) It is made clear the distribution patterns of the forest landscape of 3 classes which is relation to 5 forest ecological factors: altitude, slope orient, slope gradient, slope position and soil depth, 2D and 3D visualizations of forest landscape based on naturalness are realized. The vivid effect of 3D visualization of forest landscape suggests that 3D visualization is the best approach to describe spatial structure and distribution pattern of forest landscape.(5) Altitude, soil depth and slope gradient are the main factors affecting landscape distribution of forested land, slope position and slope orient are less dominating.(6) Within 20 years forested land and newly planted land tends to increase while brush land, scatter forest, grass land and crop land tends to decrease in the study area.(7) Suggestions on close-to-nature forest management in the study area based on naturalness are provided from 3 aspects of management type classification, management zoning and management practices.(8) Suggestions on sustainable forest management scale, forest sustainable management zoning, sustainable forest management monitoring system are provided at landscape level based on naturalness.The above study results indicate that this study has not only successfully combined forest naturalness with forest landscape but also provided with beneficial experiences on selecting index for calculating naturalness, temporal and spatial differentiation of forest landscape in relation to naturalness, close-to-nature forest management and sustainable forest management. The results from this study are in accordance with the actual situation of the study area, hence they were scientific and reasonable and would be able to provide the study area with important theoretic and technical support for its scientific management, further more they are of great reference value for relevant studies.