| Nutrient elements are not only important components of soil,but also provide the necessary nutrients for crop growth.But,the content of nitrogen(N),phosphorus(P)and potassium(K)in the soil could not meet the needs of crop growth.The application of fertilizers containing N,P and K could improve the crop yield,and ensure the soil fertility.However,the unreasonable time and method of fertilization not only seriously affect the utilization efficiency of fertilizer,but also cause excessive accumulation of nutrients,causing soil erosion and deterioration of soil structure,inhibition of crop growth and reduction of the yield of agricultural products and other problems.The Straw returning to field technology,as a unique tillage method,can improve the soil structure,and the absorption rate of nutrients in crops.And that can also increase the content of organic matter in soil,avoid sitting obstacles and so on.Not only lowering the cost of crop planted advantageously,but also reducing the probability of environmental pollution.With the development of modern precision agriculture,it is more necessary to carry out the precision fertilization according to the level of fertility in soil and the content of nutrient elements contained in crops.For the precision fertilization,rapidly in-situ and real-time detection of the content of nutrient elements in soil and different physiological parts of crops(leaves,ears,stems,etc.)has become one of the most important methods.Compared with the traditional methods for detecting the content of nutrient elements,Laser induced breakdown spectroscopy technology(LIBS)has the advantages of rapid detection,multi-element on-line detection simultaneously,and the simple sample preparation.Therefore,LIBS is very suitable for the rapid and real time detection of the content and composition of nutrient elements in the agricultural production.LIBS is a technique for rapid analyzing the composition and content of elements in samples by applying laser to the tested samples,reacting with the substances contained in the samples to generate plasma and capturing the spectral energy released in the process of plasma formation and decay.In this study,LIBS technique was used to detect and analyze the content of K in soil,and the nutrient elements(N,K)of the leaf,ear and stem samples of rice planted under different tillage and fertilization methods.During the experiment,the accuracy of the experimental results was influenced due to the inhibition of plasma-forming with the complex structure of soil and rice samples,the emergence of matrix effect,and the low reproducibility of the intensity of the emission spectra of N and K.In order to better realize the application of this technology in agricultural detection,first of all,the experimental parameters of LIBS were optimized to obtain spectral signals with high SNR(signal-to-noise ratio),relatively small standard deviation and high repeatability as much as possible,so as to reduce the error of experimental results.Experimental studies have shown that LIBS has the ability to rapidly detect the content of potassium in soil,and can evaluate the effects of different fertilization and tillage methods on the growth and yield of rice according to the intensity of N and K emission spectral.The main contents of this paper were described as follows:(1)Quantitative analysis of K in soil with five different fertilization methods were conducted by Laser-induced breakdown spectroscopy and flame photometry.The content of K in soils of 5 different fertilization methods ranged from 9 μg·g-1 to 35 μg·g-1.The relative standard deviations of content of K were calculated by flame photometry were 0.075%,0.081%,0.096%,0.134%and 0.47%,respectively.The Lorentz fitting was performed on the emission spectra of KI 766.49nm and KI 769.90nm.And the signal-to-noise ratio of them was calculated and compared.KI 766.49nm was higher than KI 769.90nm.Therefore,KI 766.49nm was selected as characteristic element for quantitative analysis in soil.The spectral data were pretreated with deviation standardization and statistical weight matrix,and the calibration curve was established with the concentration of K in soil.Compared with the original spectral data,the linear correlation of the five calibration curves was increased with 10～20%,which were 0.913,0.917,0.88,0.914 and 0.866,respectively.The experimental results show that this technique provides a theoretical basis for the real-time determination of content of K in soil.(2)LIBS experiments were carried out on the leaves,ears and stems of rice planted with five tillage and fertilization methods at the stage of young ear differentiation.The emission spectral of characteristic elements of rice were 670.88 nm,742.36 nm,744.23 nm,746.83 nm,824.24 nm,856.77 nm,862.92 mn,868.03nm of N and 766.49 nm,769.90 nm of K.By processing the mean and distance normalization methods of the spectral data,the distance between the maximum observed value,the minimum observed value and the occurrence rate of the abnormal point were reduced.In order to reduce the error caused by poor spectral information repeatability,matrix effect and experimental environment,and to eliminate redundant spectral information,principal component analysis(PCA)method was used to analyze the pre-processed spectral data,leaves,ears and stems.The first two principal components of the spectral data of the characteristic elements can explain the total spectral information levels of 91.62%,92.37%and 92.85%,respectively,and as the analysis object.The k-means and Fisher discriminant analysis methods were used to conduct three classifications of fertilization methods and five classifications of tillage and fertilization methods,respectively.When the confidence interval of the error ellipse was 81.96%,the results of the three classifications were superior to that of the five classifications.The results showed that LIBS could provide a potential theoretical basis to evaluate the effects of different tillage and fertilization methods on crop growth and yield. |
PDF Full Text Request