The Study On Spatial Distributions And Adaptive Cluster Sampling For Dongzhaigang Mangrove In Hainan

The goal of forestry in China is changing from the pattern of production for traditional wood primarily to giving priority to ecological construction. Forest resources investigation target is turning from the traditional forest resources survey to the more forest resource investigation. Traditional sampling method was easy to cause the survey costs sharply or estimated result significant deviation. In 1990, Thompson presented adaptive cluster sampling technique which was got application and attention in rare and gathered distribution.It is urgent to address the problem how to apply adaptive cluster sampling method to survey rare vegetation overall resources effectively. In addition, sampling survey method technology choice and sampling results are directly influenced by the plant distribution pattern. Therefore, the space structure of the forest resources need to be studied in investigation.Based on the population data of two species, Sonneratia caseolaris and S. Apetala Buch-Ham at the Dongzhaigang Mangrove Natural Reserve located at Hainan Province northeast., this paper has carried out sampling case research. Adopts point mode analysis were carried out to analyze space distribution patterns of two species. Based on population data of these two species, adaptive cluster sampling, systematic adaptive cluster sampling and two-stage sequential adaptive cluster sampling have been carried out sampling simulating experiment under various kinds of designs. Four criterions, which are density variance, mean of density variance, density variance of replications and relative efficiency, have been used to evaluation sampling efficiency. Research conclusions are mainly as follows:1. Firstly used adaptive cluster sampling in coastal Mangroves vegetation surveys at home.2. It was proved that both S. caseolaris and S. Apetala Buch-Ham were aggregation distribution. Three kinds of adaptive cluster sampling designs were all more effective than traditional sampling designs, so adaptive cluster sampling was applicable to be used in rare and clustering vegetation.3. From the standpoint of the influence factors to the ACS sampling design effect and the rule. It was proved that the systematic sampling design was more effective than the simple random sampling design. Two-stage sampling design was more effective than system sampling design. To the optimal samples unit sizes of S. caseolaris and S. Apetala Buch-Ham, the suitable unit area was 25m (5×5) for system adaptive cluster sampling, and 20m2(2×10) for the two-stage sequential sampling.