| Nutrient elements in plant organs are important indicators of plant growth and development.The determination of carbon(C),nitrogen(N),phosphorus(P)and phosphorus fractions(orthophosphate,monophosphate,diphosphate,pyrophosphate)in plants plays an important role in studying of the growth and development of plants.Classical and traditional methods such as chemical analysis are generally used for the determination of different nutrient contents in plants.However,traditional chemical analysis has a large amount of work,cumbersome process,time-consuming and labor-intensive waste of reagent.Near Infrared Spectroscopy(NIRS)analysis technology,with the advantages of fast detection,high efficiency,cost-effiective and no pollution,has been widely used in agricultural and medical industry and other industries.NIRS through the known nutrient content of plant sample and spectral data fitting modeling can realize the rapid determination and analysis of similar bulk samples.In order to verify and establish a near-infrared model of carbon,nitrogen and phosphorus contents and phosphorus components in various organs of subtropical woody plants,this study choose 18 kinds of typical subtropical tree species of Dashanchong national forest park in hunan province,collected branches,leaves and roots.Roots and branches can be divided into primary,secondary and third level,leaves divided into new leaves and fallen leaves.Through traditional chemical methods measuring organic carbon(OC),total nitrogen(TN),total phosphorus(TP),and 31P-NMR were used to determine the phosphorus component concentration of orthophosphate,monophosphate,diphosphate and pyrophosphate.Then,Fourier transform infrared spectroscopy(FTIR)was used to obtain the near infrared spectral data of each plant sample,and partial least squares(PLS)method was used to establish the C、N、P of the roots,branches and leaves of subtropical woody plants.The aim of this study was to establish and verify the rapid determination method by near infrared spectroscopy which can alternative of traditional chemical analysis of organic carbon,total nitrogen,total phosphorus and phosphorus classification in leaves,branches and roots of subtropical woody plants.The rapid determination method of nutrients in plants by near infrared spectroscopy can provide a rapid and accurate detection for the future research of subtropical woody plants and powerful data and technical support for more scientific experimental research.The application of near infrared spectroscopy technology in the determination of plant nutrients is discussed in this studyThe main results are as follows:(1)The NIR spectral information of each plant sample was concentrated in the range of 3750-11750nm.After preprocessing with 7 methods,including linear compensation subtraction,linear subtraction,vector normalization,minimum-maximum normalization,multiple scattering correction,first derivative and second derivative method,most of the spectral information was concentrated in the range of 4400-9400nm.The obvious absorption peak was between 3750-5250nm,and the spectrum with obvious characteristics was concentrated between 3750-5250nm.The pretreated spectrum basically retained the absorption characteristics of the original spectrum,and the outline of the obvious absorption peak became clearer,which can better judged and analyzed its spectral data.(2)Partial least squares(PLS)method was used to construct near-infrared spectral prediction models of organic carbon,total nitrogen,total phosphorus and phosphorus fractions in plant organs.The optimal spectral processing results were selected by seven pretreatment methods,and the data were analyzed by OPUS software to obtain the model.The optimal model was selected and evaluated according to RMSECV,R2 and RPD.The fitting effect of prediction model of total nitrogen content in whole organs was the best,with correlation coefficient R2 was 0.95 and RPD was 4.56.The correlation coefficients R2 of total phosphorus and organic carbon models were 0.60 and 0.46,respectively,and RPD was 1.59 and 1.53,respectively.The fitting degree of organic carbon model was the lowest,indicating that the fitting degree of each element in the near-infrared prediction model of subtropical woody plants from high to low was TN>TP>OC.It indicated that NIR spectrum had the strongest and most accurate prediction ability for nitrogen.(3)Compared the fitting results of NIRS model.The fitting effect of leaves was the best in organs.The correlation coefficients R2 and RPD of organic carbon,total nitrogen and total phosphorus were 0.75 0.89 0.87 and 3.02 2.02 2.73,respectively.The fitting correlation coefficients R2 of organic carbon and total nitrogen of branches were 0.86 and 0.85,and RPD was 2.72 and 2.55,respectively.The fitting value of total phosphorus was 0.63 and RPD was 1.65.The correlation coefficient R2 of total nitrogen and total phosphorus organic carbon in the root model was 0.91 0.72 0.32,and the RPD was 3.371.89 1.21,respectively,indicating that the fitting degree of NIRS prediction model for each organ of subtropical woody plants from high to low was leaf>branch>root,which may be due to the highest homogeneity in leaves.However,the absorption of NIR spectra of branches and roots may be affected by the differences of the lignification degree and the mycorrhizal infection degree of roots,thus affecting the prediction effect of the model.(4)The fitting degree of the prediction model of plant organic carbon,total nitrogen and total phosphorus contents was higher than phosphorus fractions.The prediction value of the former model was closer to the true value,and the coefficient of the former prediction model was between 0.32(root OC)and 0.95(whole organ TN),and the range of RPD was between 1.21 and 4.56.The contents of phosphorus fractions in different organs of plants were orthophosphate>monophosphate>diphosphate>pyrophosphate,and the R2 range of NIR model fitting was from 0.25(leaf pyrophosphate)to 0.62(leaf diphosphate),and the RPD range was 1.16-1.63.It is suggested that further discussion and research are needed,because NIR spectroscopy has a higher requirement on the sample of predictiing phosphorus fractions in plants. |
PDF Full Text Request