Geostatistical analysis of the spatial distribution of Dendroctonus valens in Pinus tabuliformis forests with different levels of infestation
Author of the article:GAO Bing-Tao REN Li-Li JIANG Qi LIU Yi-Zhou YU Lin-Feng LUO You-Qing
Author's Workplace:Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China;Beijing Key Laboratory for Forest Pest Control, Beijing Forestry University, Beijing 100083, China
Key Words:Dendroctonus valens; new range; infestation levels; geostatistics; aggregated distribution
Abstract:
[Objectives] To analyze the spatial distribution of Dendroctonus valens LeConte, a major invasive pest that threatens the ecological security of pine forests in northern China. [Methods] The spatial distribution of entrance holes in four infested forests with different levels of infestation was studied using geostatistical methods. [Results] D. valens had an aggregated distribution in different forests, but its spatial distribution differed with level of infestation. The spherical model best approximated the spatial distribution in lightly and moderately infested forest in which spatial dependence ranges were 2.060 m and 2.400 m, respectively. The exponential model best approximated the spatial distribution in severely infested(post-fire)forest, in which the spatial dependence range was 38.610 m. The Gaussian model best approximated the spatial distribution in a severely infested forest, in which the spatial dependence range was 2.633 m. The intensity of local spatial continuity in three forests that had not been burned was greater than 94%, but was less than 90% in forests that had been burnt. Old entrance holes in severely infested forests were also aggregated. The best-fitting model was a Gaussian model with a spatial dependence range of 2.754 m and local spatial continuity of 99.9%. Comparing overlay isoline and vector maps of new and old entrance holes revealed that D. valens spreads from the center to the periphery of branches. [Conclusion] D. valens has a clearly aggregation distribution and spreads around old trees in newly colonized areas.