Study On Habitat Suitability Of Cydia Pomonella (L.) In China

Cydia pomonella is a kind of invasive species which hazards apple, pear and other fruit. It has now spreaded to the Shandan counties of Zhangye City from the west, and seriously threats to the main apple producing areas in the south of Gansu province, Loess Plateau, the fruit industry and production of the eastern region of China and will have a severe impact on the fruit export of China. In this study, based on climate, topography, vegetation ecogeographical variables, we predict the potential distribution of Cydia pomonella in China using ENFA, GAPR, MAXENT, DOMAIN and BIOCLIM five Ecological Niche Models (ENMs) and we use the method of Receiver Operating Characteristic (ROC) curve analysis in assessing the accuracy of different ENMs. Habitat suitability study of codling moth in China under the natural state will have important guiding significance for controlling and monitoring the spread of insect and international trade of the apple. The application of these ENMs for studying habitat preferences has great potential benefits for invasive species and pest risk analysis.Main results are as follows:1) The suitable area of codling moth are not only correlative with climatic conditions, but also decided by the host plant spatial distribution pattern.The model constructed using all 17 variables is more rational and logical than the one constructed using 15 variables. The host plant variables have a greater impact on the distribution for ENFA, GARP and MAXENT whose optimal area reduces, the total suitable area of the both ENFA and MAXENT models reduces, suitable area of both GARP models increases; DOMAIN and BIOCLIM are affected little by the host plant, optimal area of both DOMAIN models include the host plant distribution, and optimal area of both BIOCLIM models are irrelevant with the host plant.2) The result of the overlay of five kinds of models predictes that Xinjiang, Gansu, Inner Mongolia, Ningxia, Shaanxi, Shanxi, Hebei, Beijing, Tianjin, Shandong, Liaoning province has a broad distribution of suitable areas, high-altitude areas of Yunnan, Sichuan, Guizhou province and the middle of Taiwan in the south of China apprear singly or at widely scattered suitability.3) ENFA showed that: codling moth prefers dry and hot climate. The core areas are more concentrated in dry and hot areas where are low rainfall, low relative humidity, High solar radiation and long sunshine hours. The results also indicate that the main environments which limited the distribution of the pest are corresponding with their ecosystem characteristics mutually, and the factors that influenced the range niche of the pest are annual accumulated temperature, sunshine hours, low winter temperature and an average annual temperature which determine the spread boundary of the insect.4) The comparison of five ENMs'prediction distribution shows that: among these ENMS, the biggest value of AUC was assigned to ENFA, indicating that the result gained by ENFA should be better than the other four models. According to the results of analysis of variance (ANOVA), there is not a remarkable difference in AUC between each model except for ENFA17 and BIOCLIM17. The comparison results also show that the optimal, suitable and non-suitable areas of the five ENMs are basically the same. Optimal area includes the main producing areas of apples, pears and other host plants, the suitable areas are mainly in Xinjiang, Gansu, Inner Mongolia, Shaanxi, Shanxi, Hebei, Liaoning province, and the habit suitability of vast areas of south of the Qinling Mountains are low or none; MAXENT and BIOCLIM view that the HSI of Taklimakan Desert, Kumutage desert, the Gurbantunggut Desert in Xinjiang, the Gobi in Hexi Corridor in Gansu, Badain Jaran Desert, Tengger Desert, the Mu Us Sandy Land in Inner Mongolia are naturally very low or zero, and the remaining three models show these areas are suitable in various degrees.5) ENFA and MAXENT models are more superior than the other tree GARP, DOMAIN, BIOCLIM models in terms of the accuracy and interpretation of the results and the use and application of model.