A Preliminary Study On The Main Influencing Factors And Prediction Models Of Licorice Yield And Qualit

Glycyrrhiza uralensis Fisch is currently the main species introduced and cultivated in the country.Glycyrrhiza uralensis Fisch is the most widely distributed in the country,and its distribution area traverses 13 provinces,municipalities and autonomous regions dominated by the Three Northern Regions,extending 10 latitudes from north to south and east to west.Across 52 longitudes.In such a vast region,the climate,soil and other ecological and environmental conditions vary greatly from region to region.Based on the above reasons,this paper has mainly carried out systematic research in four aspects:1 National cultivation of Glycyrrhiza uralensis Fisch cultivation techniques and yield and quality status;2 Glycyrrhiza uralensis Fisch production and quality The main factors affecting the analysis;3 environmental conditions on the yield and quality of Glycyrrhiza uralensis Fisch;4 irrigation and fertilization treatment on the yield and quality of Glycyrrhiza uralensis Fisch.Statistical analysis A mathematical model for predicting the yield and quality of Glycyrrhiza uralensis Fisch during the test period was established,and a stand-alone test software for the Glycyrrhiza uralensis Fisch production and quality prediction system based on the mathematical model was established.The results obtained are as follows:1.The Glycyrrhiza uralensis Fisch yield and active ingredient content in different planting patterns,ages,and place of production showed a certain change pattern.The study found that the yield of Glycyrrhiza uralensis Fisch was nearly twice that of transplanting in the live broadcast method,but the individual plant weights of the two were relatively close.The Glycyrrhiza uralensis Fisch production(g/m2)is greatly affected by age.The Glycyrrhiza uralensis Fisch production from 1 to 3 years increased with age,the production from 5 to 8 years was higher than 4 years,slightly lower than that from 2 and 3 years old Glycyrrhiza uralensis Fisch;The root weight increased continuously with the increase of the number of years,and the weight increase from 1 to 4 years was more even.When the birth weight was greater than 4 years,it showed a sharp increase.The content of glycyrrhizin and glycyrrhizic acid increased with age.The increase showed a certain degree of difference.The content of glycyrrhizin increased to a certain extent with the increase of longitude.Glycyrrhizic acid gradually decreased,indicating that the active ingredient content of Glycyrrhiza uralensis Fisch may have a threshold in a fixed longitude.2.Inconsistent main factors affecting the yield and quality of Glycyrrhiza uralensis Fisch in different producing areas or longitude range.Correlation analysis of all Glycyrrhiza uralensis Fisch yield samples showed that the correlations between yield and density,fertilization amount,and planting pattern were relatively large(the absolute value of the correlation coefficient r was from high to low),and the Glycyrrhiza uralensis Fisch yield of each producing area was analyzed separately.The main factors affecting the yield of Glycyrrhiza uralensis Fisch herbs are density,total number of irrigations,amount of fertilization and planting patterns(relevance in descending order);the main factors influencing the yield of Glycyrrhiza uralensis Fisch herbs in Ningxia are density,planting patterns,and fertilization amount(correlation coefficient).The absolute value of r is high to low.There was a significant correlation between Glycyrrhiza uralensis Fisch production and planting methods and fertilizer levels in Inner Mongolia.There was a significant correlation between Glycyrrhiza uralensis Fisch production and density in the Northeast region.Analysis of all Glycyrrhiza uralensis Fisch quality samples showed that there was a significant weak correlation between glycyrrhizin content and age,average annual rainfall,the warmest month’s highest temperature,longitude,soil available water level,average annual temperature,July average temperature,and soil cation exchange capacity.The correlation was from strong to weak;the content of glycyrrhizic acid was significantly correlated with age,the highest temperature of the warmest months and altitude,and the correlation was from strong to weak.At the same time,the content of glycyrrhizic acid and the content of glycyrrhizin showed a high positive correlation,indicating that under the cultivation conditions,the accumulation of these two active ingredients in Glycyrrhiza uralensis Fisch will influence each other.The influencing factors on the quality of each longitude segment alone analysis results:x2∈75.00～94.99,the content of glycyrrhizin in Fisch has a significant positive correlation with the average temperature in July and the highest temperature in the warmest month(in order of strength of correlation),and the correlation coefficient r Both were up to 0.8,and there was no significant correlation between glycyrrhizic acid content and the influencing factors.The contents of glycyrrhizin and glycyrrhizic acid in x2 ∈ 95.00～104.99 were significantly positively correlated with altitude,and two precipitation factors in geographic factors.There are significant negative correlations between temperature factors and positive correlations with age,but the correlation coefficients of the two are different.This indicates that the content of active ingredients in Glycyrrhiza uralensis Fisch is greatly influenced by environmental factors;x2 E 105.00～114.99 is only related to age.Significant correlation;x2∈115.00～124.99 The content of glycyrrhizin was significantly correlated with the average temperature in July,altitude,and the warmest month(in order of strength of correlation),glycyrrhizic acid content and July average temperature,and most There is a significant correlation between the warmest month’s maximum temperature and altitude(in order of relevance).Correlation analysis of the four longitude segments showed a downward trend in the correlation coefficient between temperature factors(x4～x6)and precipitation factors(x7～x8)in the 1st to 4th longitude,and the average annual precipitation in the precipitation factor(x7)and soil factors(x9～x 12)have gradually increased,indicating that the degree of interaction between climate and soil factors is not uniform across different longitudes.Only the correlation analysis is used in this chapter.Although the individual indicators are classified for further analysis,it is possible to attempt multiple verifications using the grey correlation degree and regression analysis.3.The yield and active ingredient content of Glycyrrhiza uralensis Fisch with the same germplasm were different under different environmental conditions.When the same batch of Glycyrrhiza uralensis Fisch germplasm material was transplanted under different environmental conditions,the yield and active ingredient content of Glycyrrhiza uralensis Fisch varied greatly.The yield indicators showed that the Glycyrrhiza uralensis Fisch values of Jingtai and Aksu were higher,and that of Jinta and Chifeng were the lowest;the content of glycyrrhizic Glycyrrhiza uralensis Fisch was highest in the Ganzhong experimental site in Lanzhou,and the highest in glycyrrhizic acid in the Glycyrrhiza uralensis Fisch experiment site in Chifeng,both Aksu and Baicheng.Glycyrrhizin and glycyrrhizic acid of Glycyrrhiza uralensis Fisch were the lowest in the test sites,and there were no significant differences in the other 5 test sites.Based on the principal component analysis of the environment and soil factors,five eigenvalues with a cumulative contribution rate of 80%were selected,which are integrated factors,geographic factors,soil charge factors,soil moisture factors,and soil nutrient factors,indicating the main factors affecting the distribution of Glycyrrhiza uralensis Fisch cultivation fields.It is a soil condition.The multivariate linear regression of the five factors after reducing the dimensions of the climate and soil factors using principal component analysis and the yield and active ingredient contents(total flavonoid content,total saponin content,glycyrrhizin content,and glycyrrhizic acid content)was tested,and the test was nonlinear.There is no significant correlation between single-independent variables and multivariate independent variables.On the contrary,the collinearity of multivariate independent variables is significant.The similarity of the control finger prints of the herbal medicines from different regions was 0.835-0.987(270nm)and 0.691-0.971(365nm).and the lower similarities were S7(Jilin Baicheng)0.835 and S8(Shaanxi Shunde).4.The mathematical relationship between irrigation,fertilization and licorice yield,or the mathematical relationship between irrigation,fertilization and active ingredient contents differ greatly;Irrigation and fertilization treatment options for the best active ingredient content of licorice from different origins are different.Under the same independent variables,the dependent variables are different,and the return is obtained.The equations differ greatly(Tab.5-33).In the Jingtai area,the Glycyrrhiza uralensis Fisch irrigation was performed 4 times,and the application rate of base fertilizer was 131.6 g/m2.The content of glycyrrhizin and glycyrrhizic acid were the highest.With the increase of the number of irrigation,the optimum amount of fertilizer for Glycyrrhiza uralensis Fisch needed to be properly reduced;In the Chifeng area,when the irrigation was performed 4 times,and the application rate of basal fertilizer was 32.9 g/m2,the yield was the highest.When the rrigation was performed once,and the amount of basal fertilizer was 98.7 g/m2,the content of Glycyrrhizin and glycyrrhizic acid was highest.5.Combination of Random Forest and Ridge Regression is a Suitable Method for Establishing Mathematical Model for Forecasting Glycyrrhiza Production and QualityLinear regression was used on all variables,R2 was at 0.65,then AIC was used for bivariate variable screening on this model,R2 was at 0.83,the screening variables were placed into the support vector regression model,and the random forest model and Ridge regression model,R2 were 0.71,0.92,0.89,respectively,using tri-fold cross validation,test linear regression model,support vector regression model,random forest model,and ridge regression model to verify The set performance R2 is0.18,0.11,0.07,and 0.10.