Font Size: a A A

Spatial Ecology Of Citrus Huanglongbing Disease Based On Niche And Multi-agent Models

Posted on:2021-05-29Degree:MasterType:Thesis
Country:ChinaCandidate:M ZhangFull Text:PDF
GTID:2513306038486064Subject:Ecology
Abstract/Summary:PDF Full Text Request
Citrus Huanglongbing(HLB)is one of the most harmful diseases in the citrus industry in China.As of now,11 provinces of 19 citrus-cultivated provinces in China have suffered from Citrus Huanglongbing,and the affected area is nearly 80%of the total area,causing huge economic losses.Guangxi Zhuang Autonomous Region is one of the major citrus planting areas in China.In 2018,the citrus planting area has reached 7 million-acre.Therefore,understanding the spatial nature of Citrus Huanglongbing is critical for the citrus industry.However,previous studies on citrus Huanglongbing focused mainly on the pathology of Citrus Huanglongbing and the efficiency of pesticides.There were few studies on the spatial ecology of Citrus Huanglongbing except for some at the global or national scale,which provides limited guidance for the control of citrus Huanglongbing at the provincial scale.To support the prediction of citrus Huanglongbing on multiple spatial-temporal scales,this study firstly investigated the performances of ten commonly used niche models for predicting the distribution of citrus Huanglongbing in Guangxi Zhuang Autonomous Region.Then the spatial distribution,susceptible risk areas and critical environmental factors of citrus Huanglongbing were carried out via a model-assembling approach.Lastly,an agent-based model of citrus Huanglongbing was built on the NetLogo platform for Mengshan County of Guangxi,and the patterns of spatial spread under different management scenarios were evaluated through the agent-based model simulations.The main results are as follows:1)Ten Ecology niche models of citrus Huanglongbing at the provincial scale were evaluated via three measures:ROC,TSS,and Kappa.The results showed that whether only 19 bioclimatic variables or 38 environmental variables(19 bioclimatic variables,18 ENVIREM environmental variables,elevation)were employed for modeling and evaluation,the random forest models(RF)were the best in all tested ecological niche models.The AUC,TSS and Kappa values of the RF model based on 19 bioclimatic variables were 0.919,0.730,and 0.713,respectively;while the AUC,TSS and Kappa values of RF model based on 38 bioclimatic variables were 0.932,0.768,and 0.769,respectively.2)The simulation results of a multi-model combination show that the high-risk area of citrus huanglongbing predicted by the model with 38 environmental variables was 50242 square kilometers,which is higher than the area predicted by the model with 19 biological climate variables.When only 19 bioclimatic variables were employed to model,the first three environmental variables with the highest contribution were temperature seasonality(0.349),mean temperature of the coldest quarter(0.304),and annual precipitation(0.239),respectively.When modeling with all 38 environmental variables,the first three environmental variables with the highest contribution values were terrain roughness index(0.876),elevation(0.471),and annual potential evapotranspiration(0.378).3)The results of multi-model combination simulation based on 38 environmental variables show that the areas with high-risk citrus Huanglongbing infection in Guangxi Zhuang Autonomous accounted for 20.96%of the total area of Guangxi Zhuang Autonomous,including Liucheng,Liuzhou,Luzhai,Xiangzhou,and Rongan in central Guangxi Zhuang Autonomous;Hepu,Pubei,Lingshan,Bobai,Qinzhou,Luchuan,Beiliu,Rong county,Yulin in the south of Guangxi Zhuang Autonomous;Fusui,Chongzuo,Longzhou,Ningming in the southwest of Guangxi Zhuang Autonomous;Fuchuan,Zhongshan,Hezhou,Pingle,Yangshuo,Lipu in the east of Guangxi Zhuang Autonomous.Areas where citrus huanglongbing has not been reported,such as Baise,Tianyang,Tiandong,Yizhou,and Huanjiang maonan autonomous county,are also potentially high-risk areas.4)The simulation of the NetLogo agent-based model of Citrus Huanglongbing(starting from the spatial distribution pattern of Citrus Huanglongbing in October 2018)showed that the area with low-risk infection increased by 249.22%in a month with the absence of any control measures.The area of moderate-risk infection increased by 57.14%and the high-risk area increased by 0.90%.After three months,as compared with the infection area after a month,the low-risk area went up by 117.67%.Meanwhile,the moderate-risk area and the high-risk area almost remained unchanged.Six months later,compared with the simulation model after three months,the low-risk area and the moderate-risk area infected by citrus Huanglongbing increased by 73.71%and 11.35%,respectively while the high-risk area showed no change.After one year of simulation,the low-risk area and the moderate-risk area infected by citrus Huanglongbing went up by 121.40%and 65.30%,respectively,however,almost no change of the high-risk area was observed in contrast to the simulation after six months.If the pesticides with a fatality rate of 90%of Citrus psyllids were applied to citrus each week,the spread of citrus Huanglongbing was thus basically hindered in Mengshan County after two months.In addition,the area of the low-risk areas was 63.48%less than that without using pesticides,The area of moderate and high-risk areas of Citrus Huanglongbing is basically zero.5)According to the results of the comprehensive analysis model,the potential spread of Citrus Huanglongbing in Mengshan County is relatively large,but low-risk areas of Citrus Huanglongbing are mostly formed,while the areas of moderate and high-risk areas are small,and they are mostly distributed at the edges of orange forests.Considering that Citrus Huanglongbing is a systemic disease,there is no way to cure it after infection,and only the whole plant is excavated.Therefore,this study suggests that the citrus planting area should strengthen the inspection of the border of the orange grove,detect and control early,and remove diseased plants in time.In addition,it is necessary to strengthen the popularization of knowledge about the identification of Citrus Huanglongbing and Citrus psyllids,strengthen unified control and rule,standardize the use of high-efficiency pesticides to kill Citrus psyllids,promptly clean up the lost citrus forest,and stop the spread of Citrus Huanglongbing.
Keywords/Search Tags:Citrus Huanglongbing, Diaphorina citri Kuwayama, Spatial distribution, Ecological niche model, Agent-based model, Guangxi Zhuang Autonomous Region
PDF Full Text Request
Related items