Study On The Impact Of Meterological Factors On The Production Of Deoxynivalenol In Wheat And Establishment Of Predictive Model For Deoxynivalenol In Jiangsu Province

Ultra performance liquid chromatography coupled with tandem mass spectrometry(UPLC-MS/MS)was optimized for simultaneously detection of multiple mycotoxins in wheat grain.And the contents of mycotoxins in wheat harvested in Jiangsu Province from the year 2018 to 2020 were examed.Then the effects of temperature,water content of wheat,fermentation time and wheat varieties on the contents of deoxynivalenol(DON)and its derivatives caused by Fusarium graminearum Schwabe infection in wheat were studied in the laboratory.Finally,the main meteorological element variables that affect the content of DON were screened by using gene expression programming algorithm(GEP).The coupling model of DON and meteorological elements in wheat in Jiangsu Province was constructed.The specific work and results are as follows:(1)Solvent extraction combined with UPLC-MS/MS was established for simultaneous detection of DON,3-acetylated deoxynivalenol(3-Ac DON),15-acetylated deoxynivalenol(15-Ac DON),ZEN and aflatoxins(AFB₁,AFB₂,AFG₁,AFG₂).The results show that the optimized extraction solvent for 8 mycotoxins was acetonitrile-water-acetic acid(80:19:1,v/v/v).Based on the method,the standard curve coefficients for all 8 mycotoxins were greater than 0.997,and the relative standard deviation were less than 15%.The detection limits were from 0.02 to 4.54(?)g/kg,the recovery range was 100.50%?119.37%,which suggests that the method can be used for wheat sample detection.Mycotoxins analysis results for wheat sample harvested from 2018 to 2020 in Jiangsu Province indicated that DON contamination was serious in 2018.The wheat producted in Northern Jiangsu,including Suqian,Huai'an and Liyang,showed high DON level polluted.However,the mycotoxins level in wheat producted in the year of 2019 and 2020 was very low.The correlation analysis among the 8 mycotoxins showed that the DON,3-Ac DON and 15-Ac DON had significantly positive correlation(r>0.8,p<0.05).The correlation coefficients for the other 5 toxins were low.(2)The influence of water content,temperature,inoculation amount and inoculation time on the production of DON and its derivatives in wheat by Fusarium graminearum PH-1 had been simulated studied.The influence of different factors followed with the order of the interaction of moisture content and time>moisture content>time>temperature>time quadratic.The inoculation conditions for the DON yield of 4000?6000(?)g/kg were deployed for different varieties of wheat based on the response surface studied.There were significant differences in the content of DON and its derivatives among different varieties of wheat.The correlation coefficients between the content of DON with disease index,15-Ac DON and3-Ac DON were 0.4080,0.6120 and 0.7380(p<0.05)respectively.There is also a significant positive correlation between the disease index and the falling index with the correlation coefficient value of 0.4160(p<0.05).The correlation coefficient value between 15-Ac DON and 3-Ac DON was 0.9730(p<0.01).(3)Correlation of DON content and meteorological factors in wheat harvest in Jiangsu Province in 2018 were analyzed by GEP method,and the result showed that the average temperature,cumulative precipitation,precipitation days,continuous precipitation days,and average humidity have a coupling relationship with the classification of DON.The average correlations were 0.197,0.124,0.101,0.201,and0.058,respectively.The DON predictive model was established by using above meteorological factors and the number of days with temperature greater than 25?,relative humidity greater than 80%,and days with daily precipitation greater than 2mm as explanatory variables,and different natural months,blooming period and blooming-harvesting period as the time window.It was found that the model with blooming period as the time window(GEP?PSO model)is the best with an accuracy rate of 77.15%.The DON contents of wheat from Jiangsu Province in 2018 were used to verify the GEP?PSO model,and the result showed the accuracy for the training group is 76.45%,and the accuracy for the test group is 63.61%.The probability of‘false positive'is 59%,and the probability of‘false positive'is 28%.The GEP?PSO model is more accurate compared with multiple regression(MA)model,partial least squares(PLS)model and support vector machine(SVM)model.This study aims to clarify the impact of meterological factors on the production of DON,and build a coupling model between DON of wheat and meteorological elements through screening the meteorological variables to achieve efficient early warning of the DON contaimination.