| National Nature Reserve of Longqishan (NNRL) is an ideal environment for the forest eco-tourism and it is famous for the diversity of species, the richness of rare animals and plants and the typical sub-tropical forest vegetation. However, the weakness of Reserve economic base and the limitation of government financial resources have no way to throw into sufficient funds to build and perfect the infrastructure construction and scenic spot touring. All leads to hardly realize the coordinated growth of ecological, social and economic benefits.According to the weakness of studies on environment and landscape pattern of NNRL, this article focused on analyzing of Landscape pattern of NNRL, the distribution law of patch, and the relationship of outside interference pattern on the landscape impact of NNRL. Then, it also evaluated the eco-quality of forest landscape using AHP, BRF, PPE, and SVM. Last, the paper brought up and predicted the capacity of eco-tourism, and gave some advices. The results of the paper can develop a systematic and targeted regional sustainable development strategy for Reserve managers. Thus, it is useful for coordinated development of ecological, social and economic benefits.The results listed below.1. It analyzed of forest landscape pattern of NNRL with landscape indexes, which were computed under patch-type-level, class-type-level, and landscape-type-level using FRAGSTATS software with grid picture converted under ARCGIS environment.1) The results of analysis of forest landscape pattern indicate that there are nine type forest landscape, and their areas, perimeters, and the number of patch distributes very unevenly. The results further present that broadleaf forest, pine forest, and bamboo forest are mainly landscape styles, and that the sum of their area is 95.3078 percent of the NNRL, while 85.6365 percent of the perimeter of NNRL. But, they also suggest that the numerical area and perimeter varies sharply discrimination. Besides, because the nature reserve has its unique protection character, the number of the bigger patches is not big enough.2) The results of analysis of shape features landscape elements show that the patches shape of the broadleaf forest landscape is the most complicated, and that landscape connectivity is good. They further indicate the degree of fragmentation and patches separation degree are fairly low, while the degree of fragmentation in non-wood landscape and economic landscape are high which accords with the theory that landscapes with human interference suffer from high degree of fragmentation. But, as a reserve of national-level forest ecosystem, NNRL enjoys a high predominance, uneven landscape distribution and a low variety on the whole.2. Normal, Lognormal, Weibull, Gamma and Beta distribution respectively are used to analyze the distribution law in the size, the area, and the fractal indices .The analysis results of Landscape patches characteristic distribution law proves a level relativity exists among landscapes patches. The results also show normal distribution fitness is the best, and that Gama distribution has a better effect on the area distribution and fractal distribution fitness, while the Weibull distribution fitness has the worst of all. They further present that the size, the area, and the fractal indices submit to normal distribution mostly, while the parties of indices submit to gamma distribution.3. The article uses nonlinear modeling of modern statistical to simulate and analyze the effects of landscape pattern by Natural disturbance and human disturbance.1). It chooses 10 indices as a natural disturbance factors such as size, volume, age, soil thickness, shrub cover, herbaceous cover, average altitude, slope, aspect and slope position to construct of artificial neural network model, multi-model adaptive regression splines, Support vector machine model and project pursuit regression model to analog analyze of fractal index separately. That the value of fractal index is changing when the value of one index changes, proves and riches the conclusion on interference affecting on landscape pattern. Through leave-one-out cross-validation for the simulation of natural disturbance on the impact of landscape pattern, multiple adaptive regression splines model has the most ability of revealing the relationship between interference and impression of landscape pattern.2). Human beings cause changes in landscape pattern by changing the number of landscape types and it is one of the performances of human Interference activities on the landscape pattern affect. It chooses the number of landscape types as human interference factors to construct artificial neural network model, multi-model adaptive regression splines, support vector machine model and projection pursuit regression model to respectively analog analysis human disturbance on the impact of the landscape diversity, dominance and evenness. The results show that the change of the number of landscape types because of human beings interference is remarkable to diversity, but weak to dominance and evenness. It also proves and riches the conclusion on interference affecting on landscape pattern. By comparing the results of leave-one-out cross-validation for the simulation analysis of the number of landscape patches on the landscape diversity, the analytical capacity of the above-mentioned models are stronger than the generalization ability of each model. Among them, the projection pursuit regression model is the optimal performance. While according to the analysis used to simulate the number of landscape patches dominance on the landscape and the impact of uniformity, the above-mentioned models have not been able to obtain better performance.4. From the landscape ecological protection and the development and utilization level,it chooses the characteristics of Nature, value, diversity, representativeness, suitability, stability, interference, and usability, and ornamental, research, and regional and cultural values as indicators to construct RBF networks, support vector machines, Projection Pursuit Evaluation Model and AHP Evaluation Model to evaluate the eco-quality of forest landscape.1). The result of AHP shows the index of eco-quality of forest landscape of NNRL is 0.79, which means it lies in a better state of eco-quality.2). The results of RBF model, SVM model, and PPE model also show the eco-quality of forest landscape of NNRL lies in a better state of eco-quality.3). Within the scope of all levels, it randomly generated 10 samples data and used these methods to evaluate. In the case of 0.99 Credibility and highly concerning Support Vector Machine, RBF network and analytic hierarchy process, the results of evaluation are consistent. On the contrary, Projection Pursuit evaluation model is an independent method, so the results of evaluation are different.4)Combined with the total numbers and theoretical analysis, it is used in evaluating the eco-quality of landscape. The methods which ranged from the best to the poor are Support Vector Machine, RBF network, the level analysis and projection pursuit evaluation model in turns.5. We plan for ecotourism according to different function areas, estimate and predict the capacity of environment and the capacity of tourism based-on bionomics, eco economics, and the tourism economics theory driven by some planning aims. The results of estimation of the capacity of environment show that it is about 280 days for tourism in one year. They also present that the daily capacity is 883 persons each time per day for near future and is 1663 persons each time per day for long-date. They further indicate that the annual capacity is 247240 persons each time per year for near future and is 465640 persons each time per year for long-date. The results of estimation of the capacity of eco-tourism show that the daily capacity is 622 persons each time per day for near future and is 1159 persons each time per day for long-date. They also indicate that the annual capacity is 174160 persons each time per year for near future and is 324520 persons each time per year for long-date. The results of prediction on the size of eco-tourism show that the daily size is 360 persons each time per day for near future and is 726 persons each time per day for long-date. They also indicate that the annual size is 101.1 thousand persons each time per year for near future and is 203.4 thousand persons each time per year for long-date. While in the process of carrying on ecotourism, we should properly deal with the relationship between development and protection, strengthen and lead extensive propaganda, work together and intensify contents management to update the service level. |
PDF Full Text Request