| Rice is a major food crop,and nearly half of the world’s population feed on it.Rice plants have strong absorption characteristics of cadmium.Low level cadmium pollution will cause the cadmium accumulation in rice roots,stems,leaves and seeds,which can cause poison to plants,destroy cell structure,inhibit enzyme activity,affect metabolic activity,change the phenotypic information of rice and eventually form cadmium-rice.Rice roots and stems are important organs for the absorption,transfer and accumulation of heavy metals in rice.Traditional heavy metal laboratory analysis methods are cumbersome,costly,and time-consuming.There is an urgent need for a method and means for rapid detection of heavy metals to realize the monitoring and prevention of heavy metal pollution.The aim of this study is to quickly obtain the phenotypic information of rice roots and stems.This study investigated the morphological changes,physiological responses,mineral elements and cadmium content of rice roots and stems under heavy metal stress and achieved accurate and rapid detection of crop phenotypes with multiple information,which is of great significance for crop stress and heavy metal detection and repair.The main research contents and conclusions are as follows:(1)The response law of rice roots and stems morphological characteristics were revealed,and quantitative relationships between root micro-morphology,root length and surface area,root tip number and branch number were established,which provided theoretical support for the research of heavy metal absorption and transport in crop roots.Based on the root analysis system Win RHIZO,the morphological parameters of rice roots and stems were obtained,and the significance and variation of root and stem morphological characteristics between groups under different stress conditions were explored;the morphological characteristics of the microstructure and subcellular structure of rice root and stem under different cadmium stress conditions were revealed based on scanning electron microscopy and transmission electron microscopy techniques.The results showed that:(1)Cd stress had a "low-promoting and high-inhibition" effect on root and stem morphological characteristics.The best growth condition of rice root and stem morphological characteristics occurred under low Cd concentration(CK/5/25μM)for a long time(15/20 days),and the worst growth condition occurred under high Cd concentration(100 μM)for a long time(20 days).(2)Root length had a strong correlation with surface area,tips and forks,with correlation coefficients of 0.9052,0.8987 and 0.9412,respectively;surface area had a strong correlation with total volume and forks,with correlation coefficients of 0.9340 and 0.8841,respectively;tips had a strong correlation with forks,with correlation coefficient of 0.8369.(2)The physiological response pattern and mechanism of rice roots and stems under Cd stress were investigated,and a rapid detection model for the physiological information(soluble protein and malondialdehyderic)of roots and stems in adversity was established,which provided technical support for the rapid detection of physiological information of crop roots and stems in adversity.The physiological information(soluble protein and malondialdehyde)of rice roots and stems under different Cd stress concentrations(CK,5 μM,25μM,50 μM and 100 μM)and different stress times(5,10,15 and 20 days)were detected.And the physiological response pattern and mechanism of rice roots and stems under Cd stress were analyzed.A hyperspectral model for rapid detection of physiological information of rice roots and stems in adversity based on full spectrum and characteristic bands was established.The results showed that:(1)Low-concentration cadmium stress promotes protein synthesis and metabolism in cells;high-concentration cadmium stress promotes the activity of proteolytic enzymes and strengthens the decomposition of soluble proteins.Low concentration of cadmium stress can inhibit the synthesis of malondialdehyde in roots;high concentration of cadmium stress can cause metabolic disorder of antioxidant enzyme system,stimulate tissue to produce and accumulate a large amount of ROS,and promote the synthesis of malondialdehyde in root.(2)SPA,RF and CARS methods were used to select characteristic wavelengths related to the physiological information of adversity,based on the regression methods of PLSR,SVM and ELM to establish a rapid detection model of adversity physiology,and the correlation coefficient(R)and root mean square error(RMSE)were used to evaluate the effect of the model.For soluble proteins in rice roots,the optimal model was CARS-ELM(Rp=0.9161,RMSEP=0.0687)and the input variables are reduced by 89.25%;for malondialdehyde in rice roots,the optimal model was CARS-ELM(Rp=0.9021,RMSEP=0.2507)and the input variables are reduced by 85.25%;for soluble protein in rice stems,the optimal model was RF-ELM(Rp=0.7718,RMSEP=0.0233)and the input variables are reduced by 90.50%;for malondialdehyde in rice stems,the optimal model was CARS-ELM(Rp=0.7994,RMSEP=0.4824)and the input variables are reduced by 90.50%.(3)The accumulation pattern of cadmium elements in rice roots and stems were revealed,a fusion analysis method of full spectrum,characteristic spectrum and univariate and multivariate were proposed,a rapid diagnosis model of the degree of cadmium stress in roots and stems was established,and a rapid detection method of cadmium elements in roots and stems was studied,which provided method support for the rapid detection of the degree of heavy metal stress and elemental content in crops.The accumulation pattern and resistance mechanism of cadmium elements in rice roots and stems were investigated based on detection of the accumulation of cadmium elements in rice roots and stems by using the standard chemical method;the optimal collection parameters of the LIBS system were determined,and the cadmium stress degree and the quantitative detection method of cadmium element were established.PCA was used to realize the cluster analysis of rice root and stem with different cadmium stress and the rapid diagnosis of samples was realized based on PLS-DA,SVM and ELM models;univariate analysis models were established using Cd II 214.44 nm,Cd II 226.50 nm and Cd I228.80 nm as the characteristic spectral lines;multivariate analysis models were established based on PLSR,LSSVM,ELM and RBFNN models using the combination of characteristic spectral lines,230-884 nm band and 211-232 nm band as inputs;multivariate analysis models were established based on PLSR,LSSVM,ELM and RBFNN models using i PLS,SPA and RF feature screening methods.The results showed that:(1)The range of cadmium content in roots was 0-1033.77 mg/kg,and the range of cadmium content in stems was 0-302.00 mg/kg.Although in the same stage of stress,the Cd content in the stems was much less than the Cd accumulation in the roots during the same period.(2)For rice roots and stems,the optimal delay time was 2.5 μs and2 μs,respectively,and the optimal energy was 85 m J and 90 m J,respectively.(3)For rice roots and stems with a single stress time and all stress times,excellent discrimination results can be obtained in the 230-884 nm band and the 211-232 nm band,and the prediction set accuracy of the optimal model exceeded 88%,respectively.(4)For the rapid detection of cadmium in rice roots,the optimal model was the full-spectrum model based on the 211-232 nm short band(Rp=0.9958,RMSEP=29.58 mg/kg),which reduces the number of variables by 95.35% compared with the230-884 nm band.(5)For the rapid detection of cadmium in rice stems,the optimal model was the SPA-ELM model(Rp =0.9965,RMSEP=7.46 mg/kg)based on the 211-232 nm band.Compared with the 211-232 nm band,the numbe of variables has been reduced by 99.51%,which has important reference value for the development of portable instruments.(4)The accumulation pattern of heavy metal Cd elements and mineral elements in rice roots and stems were investigated,and a rapid detection model of mineral elements in rice roots and stems was established based on LIBS technology,which realized the simultaneous,accurate and rapid detection of multiple mineral elements in crops and provided model support for the development of portable instruments.The accumulation pattern of Cd elements and mineral elements(K,Ca,Na,Mg,P,Fe,Mn,Zn,Cu)in rice roots and stems were analyzed based on standard chemical methods;preprocessing methods such as area normalization,abnormal spectrum elimination,wavelet transform and spectrum averaging were used to improve the stability of detection signals and reduce the interference of point-to-point fluctuations and background noise.The effects of quantitative detection based on full spectrum and feature bands were compared and the accuracy and reliability of LIBS for the detection of mineral elements were investigated based on the PLSR,LSSVM,ELM and RBFNN models and the i PLS,SPA and RF feature screening methods.The results showed that:(1)The contents of K,Na,P,Ca,Mg,Mn and Zn in the roots were lower than those in the stems,while the contents of Cd,Fe and Cu in the roots were higher than those in the stems.(2)The correlation coefficients of the optimal prediction model for mineral elements in roots,K,Mg,Na and Mn are greater than 0.98,Ca,P and Fe are greater than 0.95,Zn and Cu are greater than 0.90;the correlation coefficients of the optimal prediction model for mineral elements in stems,Na,Fe and Mn are greater than 0.98,Ca is greater than 0.95,K,Mg,P,Zn and Cu are greater than 0.90.LIBS technology has achieved good predictive effects on the mineral elements in roots and stems,and provided support for rapid detection and precise regulation of rice growth process information. |
PDF Full Text Request