4. DISCUSS
In this study, the distribution of PWD was predicted based on the MaxEnt model. Environmental factors and actual geographical distribution spaces of species were combined when this model was used to study the potential distribution of species (Yuan et al. , 2020). The actual geographical distribution of species and the environmental factors of species have a stable state when the state parameters obtained through the operation of the system obtain the maximum entropy to determine the potential geographical distribution of the species (Rong et al. , 2019). The MaxEnt model is easy to overfit when simulating the potential distribution of species, resulting in unreliable prediction results. This can seriously affect its application in global change biology and other research fields (Liu et al. , 2018). Therefore, the ENMeval data package was used to adjust the MaxEnt model parameters in the study. The parameters with the lowest complexity were selected to predict the potential distribution areas of PWD disease by analyzing model complexity under various parameter conditions (Muscarella et al. , 2014). The average value of training AUC reached 0.940, indicating that the prediction results of the model had high accuracy and feasibility.
The main climatic factors affecting the potential geographical distribution of PWD were temperature (max temperature of the warmest month [bio5], mean temperature of the driest quarter [bio9]), rainfall (coefficient of the variation of precipitation seasonality [bio14], and precipitation of the wettest quarter [bio16]). Kobayashi et al. (1970) studied precipitation as a crucial factor affecting damage caused by PWD, including the many pine trees killed by PWD during a summer drought. Naoko et al. (2001) studied pine forests in the warm temperature zone of Japan, which PWD had seriously damaged. The maximum temperature in the warmest month affected the diffusion of adultMonochamus alternatus (M.alternatus ), the vector of pine wood nematodes. The population density of B.xylophilus was significantly increased by drought (Zhao et al. , 2003).Currently, researchers agree that the climate conditions of high temperature and drought are conducive to the occurrence of PWD (Kanzaki & Giblin-Davis, 2018).
Under the next six greenhouse gas emission scenarios, the potential distribution area of PWD predicted by MaxEnt will shift to the Northeast, and the potential suitable area will expand significantly under the 2070s SSP5-8.5 scenario. Climate is the key factor affecting the distribution area of species. Climate change will have a far-reaching impact on the distribution of pests, and global warming will be the trend of climate change in the future (Volney & Fleming, 2000; Walther et al. , 2009). Previous studies have reported that the suitable distribution area of PWD in China will expand nearly twice by 2100 with the intensification of climate change, showing a trend of acceleration of the diffusion rate to the North and West (Cheng et al. , 2015). The results of the correlation between temperature and B.xylophilus showed that low temperature may inhibit the spread of B.xylophilus by affecting the reproduction and activity range of M.alternatus (Jikumaru et al., 2008). Global warming is conducive to the activity of the nematode vector B.xylophilus , which significantly enhances the damage caused by PWD (Jikumaru & Togashi, 2000). However, predictions of the potential distribution area of PWD also need to be combined with host plants, topography, soil, and human activities. Abiotic factors, biological factors, and species migration affect the distribution of species during their long-term evolution, and the distribution range of species is different in different historical periods (Soberón, 2007).
If effective prevention and control of PWD is not realized as soon as possible, this major disease may soon spread across a larger area in China, causing hundreds of millions of pine tree deaths every year. This will be an ecological disaster (Lee, 2014). The techniques adopted to control PWD in China mainly include disease quarantine and epidemic situation monitoring, diseased wood removal, and vector insect control. Further, there is an active disease prevention measure, namely, trunk injection. Trunk injection involves injecting effective components into the tree and exerting a drug effect by distributing the chemicals based on the transpiration of the tree. It has advantages of accurate application and high control efficiency, and it is environmentally friendly (Byrne et al. , 2014; Takai et al. , 2000b) (Aćimović et al. , 2014; VanWoerkom et al. , 2014). Emamectin benzoate is a semi-synthetic second-generation avermectin-derived insecticide found to have the strongest nematicide activity against B.xylophilus among different chemical substances (Takai et al. , 2000a). Therefore, it can be considered a strong candidate for use as a preventive trunk injection against pine wilt disease.