Research On Early Detection Of Anthracnose Of Camellia Oleifera Based On LIBS Technology

Camellia oleifera is a kind of tea tree that can produce edible oil.Its roots,leaves,seeds and shells all have good ecological and economic benefits.However,with the continuous increase of the planting area of Camellia oleifera,more and more diseases are harming the Camellia oleifera,among which anthracnose is a highly destructive disease commonly occurring in Camellia oleifera industry,and there is no effective prevention and control agent developed at present,which seriously hindered the healthy development of the Camellia oleifera industry.Therefore,timely and accurate detection of anthracnose is of great significance to the healthy and sustainable development of Camellia oleifera industry.In this paper,the laser induced breakdown spectroscopy（LIBS）technology was proposed to study the spectral response characteristics of nutrient elements in Camellia oleifera leaves.Through the change of the nutritional status of Camellia oleifera leaves,the early detection of Camellia oleifera anthracnose was realized quickly,effectively and nondestructive.The main research contents of this paper were as follows:（1）The variation rule of LIBS spectral characteristics of nutrient elements in Camellia oleifera leaves was studied,and the discriminant analysis model based on different degrees of anthracnose damage was established.The LIBS spectral data of healthy,mild,moderate and severe anthracnose Camellia oleifera leaves were collected,and the relationship between the characteristic peak intensity of Fe,Mn,Cu,Zn,Ca and Mg in four kinds of samples and anthracnose disease was analyzed.PLS-DA,ELM and SVM discriminant analysis models based on full spectrum and characteristic variables were established for leaves of anthracnose.The results showed that:（1）The LIBS spectral curves of the four kinds of samples were basically similar,only the spectral line intensity was different.（2）The damage degree of anthracnose damage was correlated with the characteristic peak intensities of the six nutrient elements.With the aggravation of anthracnose damage,the characteristic peak intensities of the six nutrient elements were gradually reduced.（3）The recognition rate of SVM model based on the polynomial kernel function of characteristic variables was the best,and the correct recognition rates of healthy,mild,moderate and severe leaves of Camellia oleifera were 97.22%,94.44%,72.22%,72.22%,respectively.（2）The content of nutrient elements in Camellia oleifera leaves was analyzed by flame atomic absorption spectrometry.The true contents of Fe,Mn,Cu,Zn,Ca and Mg in different type of samples were detected by flame atomic absorption spectrometry,and the range and variation rules of the contents of nutrients in each type of samples were compared and analyzed.The results showed that:（1）With the increase of infection degree of anthracnose,the maximum contents of each nutrient element decreased gradually.（2）The order of contents of six nutrient elements in leaves of Camellia oleifera was Ca>Mn>Mg>Fe>Zn>Cu.（3）Using single factor variance and multiple Tamhane analysis in SPSS 25.0 software,the significant difference of the contents of each nutrient element in different degrees of anthrax damage was obtained.The quantitative analysis model of nutrient elements in Camellia oleifera leaves was established.Based on external standard method,PLS and BP artificial neural network was established respectively.The results showed that the BP artificial neural network model has better quantitative analysis performance than external standard method and PLS method,and could accurately predict the concentration of nutrient elements.The R²,RMSEP and RSD of Fe in the leaves were 0.97,61.23μg·g^-1 and 21.96%,respectively.The R²,RMSEP and RSD of Mn in the leaves were 0.93,191.74μg·g^-1 and 10.44%,respectively.The R²,RMSEP and RSD of Cu in the leaves were 0.97,1.19μg·g^-1 and 19.45%,respectively.The R²,RMSEP and RSD of Zn in the leaves were 0.85,53.57μg·g^-1and 48.69%,respectively.The R²,RMSEP and RSD of Ca in the leaves were 0.95,348.97μg·g^-1 and 15.46%,respectively.The R²,RMSEP and RSD of Mg in the leaves were 0.81,206.88μg·g^-1 and 11.39%,respectively.The above research results showed that the LIBS technology could realize the nondestructive detection and damage classification of Camellia anthracnose,and could realize the quantitative detection of a variety of nutritional elements,which provided technical support for the early detection and prevention of Camellia anthracnose.