Study On The Influence Factors Of Stable Isotope Composition In Cocoons And Its Origin Traceability

Recently there was a lot of studies in origin assessment of silk due to economic and cultural issues.But there was little information about whether the varieties,colors and seasons of cocoons could affect the ratios of isotopes which in turn affect the accuracy of differentiation of silk from different origins.The objective of this study was to investigate the changes in stable isotope composition caused by cocoons themselves and external factors,and to provide a basis for the construction of origin traceability models for silk fabrics.In this study,we investigated the trends of four stable isotope ratios(δ¹³C,δ¹⁵N,δ²H andδ¹⁸O)in silkworm body and silkworm excrement during the growth of silkworm from larva to silk production,and compared the effects of different varieties on the stable isotope ratios in cocoons.The cocoon samples from four major mulberry production areas,namely Hubei,Shandong,Zhejiang and Sichuan,were selected for the study.The differences in stable isotope composition of cocoon due to its color and feeding season were compared,and the origin traceability models of cocoons,degummed silk and silk protein were established using one-way ANOVA,Fisher Linear Discrimination and Artificial Neural Network.The main findings were as following.1.Silkworm showed different enrichment and depletion effects onδ¹³C,δ¹⁵N,δ²H andδ¹⁸O of ingested mulberry leaves during different developmental stages.In silkworm excrement,the ratios ofδ¹³C,δ¹⁵N,δ²H andδ¹⁸O changed at ranges like those in silkworm body.Stable isotopic composition of silkworm body and silkworm excrement was not different significantly among varieties except few instars under the same feeding conditions,and there was linear correlation between the isotopic ratios ofδ¹³C,δ¹⁵N andδ²H in body and excrement.From the results of the elliptical joint confidence region,it can be seen that the differences in stable isotopic composition in cocoons caused by mulberry silkworm species were not different significantly,and the application of different species of cocoons for origin tracing is valid and feasible.2.Independent-samples T test and elliptical joint confidence region were applied were used to explore the differences in stable isotope composition in cocoons of different colors and feeding seasons from the same geographical origin.The results of the study showed that there were no significant differences between cocoons of different colors from the same production area,except for a few isotopic indicators.The values ofδ²H andδ¹⁸O in cocoons of different seasons were closely related to the feeding of mulberry leaves.Except for theδ¹³C value of cocoon samples from Hubei,there were no significant differences in the ratios of the four stable isotopes in cocoons of different seasons from the same production area.Colors and seasons may cause large variations in a single stable isotope in cocoons,but such variations do not lead to significant differences between cocoon samples.Therefore,the stable isotope technique can be applied to the origin tracing of cocoons of different colors and rearing seasons.3.One-way ANOVA was performed onδ¹³C,δ¹⁵N,δ²H,andδ¹⁸O values in cocoon samples from four provinces,and the results showed that all four stable isotope ratios in samples from different geographical origins were significantly different and were effective traceability indicators.It showed that the origin identification accuracy of single stable isotope index was low,and the discriminant accuracy increased significantly with the increase of index coupling.The origin traceability model based on the cocoonδ¹³C,δ¹⁵N,δ²H andδ¹⁸O indicators was established,and the accuracy of initial discrimination and cross-validation were 83.5%and 79.7%,respectively.The validation accuracy of the training set and test set was 94.9%and 90.0%,respectively,when applying Artificial Neural Network for cocoon origin discrimination.4.The degumming treatment had a small effect on theδ¹³C andδ¹⁸O values and a large effect on theδ¹⁵N andδ²H values.The stable isotope ratios of silk fibers from different origins still differed significantly,and the initial discrimination and cross-validation correct rates of the origin discrimination model established by Fisher Linear Discrimination were 83.5%and 78.5%,respectively.The correct discrimination rate of the training set of Artificial Neural Network reached 89.5%,and the back-verification rate of the training set was 81.8%,with a slight decrease in the discrimination accuracy compared to that of the cocoon samples.5.Usingδ¹³C andδ¹⁵N indicators coupled for origin tracing of silk protein,the initial overall discrimination rate and the correct rate of cross-validation discrimination of the established linear discrimination model were 67.1%and 65.8%,respectively,which were slightly lower than those of cocoon samples.the overall correct origin identification rates of the training sample set and test sample set of BP neural network were 80.0%and 83.3%,respectively.δ¹³C andδ¹⁵N indicators can correctly identify Hubei Shandong and Zhejiang samples,but cannot distinguish Sichuan samples,and need to be combined with other indicators to discriminate.