Correlated Factors Analysis Of Spatial And Temporal Dustribution Of Mycotoxins In Triditional Vinegar Pei And Establishment Of Hyperspectral Prediction Model

Solid-state fermentation is the main production method in the vinegar industry in China.Traditional vinegar is produced from grains such as sticky rice or kaoliang as the main fermentation material through open solid-state fermentation.Due to the contamination of raw materials/substrates and the vulnerability of the open fermentation environment to toxigenic fungal growth,solid-state vinegar is highly susceptible to mycotoxins.In addition,the country and the vinegar industry have not paid sufficient attention to the issue of mycotoxin pollution in vinegar.In the National Standard of the People’s Republic of China（GB 2761-2017）,only a clear limit is set for the content of aflatoxin B1（AFB₁）in vinegar,and there are no limit standards for deoxynivalenol（DON）and zearalenone（ZEN）.Therefore,this study uses a pilot production line of a traditional solid-state fermented vinegar to illuminate the spatial-temporal distribution of the primary mycotoxins during solid-state fermentation process,based on high-performance liquid chromatography technology.Furthermore,the dynamic changes in the physicochemical parameter,bacterial and fungal community compositions in the fermentation basin were revealed.Accordingly,the correlation between the mycotoxins and physicochemical parameter,microorganism diversity during solid-state fermentation process was analyzed to determine the potential effect on the mycotoxins.Finally,the hyperspectral imaging technique and chemometrics methods were used for rapid prediction of the AFB₁,DON and ZEN content of the vinegar Pei.The main research content are as follows.（1）In this chapter,the content of AFB₁,DON,and ZEN in 324 samples of vinegar Pei were conducted by a Immunoaffinity chromatography combined with high performance liquid chromatography,and analyzed the differences in the content of mycotoxins in vinegar Pei at different spatial planes.The results showed that the average DON level exhibited a 17.17%decrease at the end of acetic acid fermentation compared with the first day,while the average AFB₁ and ZEN content increased considerably to 141.67%and 113.64%,respectively.The average DON content fluctuated between 1 d and 7 d.Contrarily,the average AFB₁ and ZEN levels increased significantly at 5-7 d and 7-11 d,respectively.The content of AFB₁,DON and ZEN remained stable at 11-17 d.In addition,the significant variation in the AFB₁ and ZEN content was limited to the cross-sectional sampling planes in the fermentation basin,while DON was heterogeneously distributed on the cross-sectional,horizontal,and vertical sampling planes.（2）In this chapter,we illuminated the dynamic changes in moisture,p H,and bacterial and fungal communities during solid-state fermentation process,and then the correlation between the mycotoxin and physicochemical parameters,microorganism diversity was analyzed to determine the potential effect on the mycotoxins.The results showed that the p H value was negatively correlated with the AFB₁（r=-0.752,P<0.001）and ZEN content（r=-0.558,P<0.001）and positively associated with the DON level（r=0.409,P<0.001）.The moisture content was only positively correlated with AFB₁（r=0.136,P<0.014）.The Spearman correlation coefficients revealed close relationships between ten bacterial,seven fungal species,and three mycotoxins.The AFB₁ presented a negative correlation with Lactobacillus helveticus,Alternaria destruens and Fusarium aethiopicum based on the Spearman correlation coefficients,while it was positively correlated with Acetobacter pasteurianus,Acetilactobacillus jinshanensis,Aspergillus heterocarioticus,and Rasamsonia composicola.DON showed a significantly negative correlation with Acetobacter pasteurianus,while it was positively associated with Lactobacillus helveticus.ZEN displayed significant negative correlations with Lactobacillus helveticus,Lactobacillus acetotolerans,Alternaria destruens,Fusarium aethiopicum and Brettanomyces bruxellensis,while Alternaria destruens（OTU1）and Fusarium aethiopicum（OTU3）were positively correlated with Acetobacter pasteurianus,Acetilactobacillus jinshanensis,Aspergillus heterocarioticus,and Rasamsonia composicola.Among them the Spearman correlation coefficients between ZEN and Lactobacillus helveticus,Acetobacter pasteurianus,Alternaria destruens and Fusarium aethiopicum exceeded 0.7.（3）In this chapter,the external characteristics of the vinegar Pei were extracted from the hyperspectral technology.Comparing the correlation coefficients and RMSE of partial least squares（PLS）,interval partial least square（i PLS）,backward interval partial least squares（bi PLS）and synergy interval partial least square（si PLS）based on full-band spectrum and competitive adaptive reweighted sampling（CARS）spectrum,the optimal algorithm was selected to predict the contents of AFB₁,DON and ZEN.For AFB₁,the good results were obtained by si PLS based on the CARS characteristic band with r_p of 0.83 and RMSEP of 0.09.For DON,the good results were obtained by PLS based on the CARS characteristic band with r_p of 0.84 and RMSEP of 31.20.For ZEN,the good results were obtained by i PLS based on the full band spectrum with r_p of 0.80and RMSEP of 7.29.Overall,these findings verify the potential of hyperspectral imaging for rapid discrimination of the three mycotoxins contamination in vinegar Pei.This study is the first to use a pilot production line of a traditional vinegar,to illuminate the spatial distribution of the primary mycotoxins during the solid-state fermentation.Furthermore,the dynamic changes of the p H,moisture,bacterial and fungal community compositions at different spatial locations in the fermentation basin were revealed.Accordingly,the correlations between the mycotoxin and physicochemical parameters,microorganism diversity were analyzed to determine the potential microbial effect on the mycotoxin distribution during solid-state fermentation.Finally,the prediction models of mycotoxins in vinegar Pei were established based on hyperspectral technique,which provided basic research for preventing and controlling mycotoxin pollution.