Research On The Evolution Of Paddy Soil Fertility Quality In Cixi Area Based On Imaging Spectroscopy Technolog

It is essential for future security and soil sustainability to understand how soil fertility evolves under long-term anthropogenic influences.Spectroscopic techniques can compensate for the shortcomings of traditional indoor methods used to measure soil properties.The main objective of this study is using rice soil formed by human cultivation in Cixi City,Zhejiang Province over the past millennium as the research object to investigate the feasibility of using laboratory imaging spectroscopy(IS)technology to predict the fertility properties(SOM,TN,TP,TK,p H)of soil profiles at different ages of rice planting.Three methods,support vector machine(SVM),partial least square regression(PLSR)and back propagation neural network(BPNN),were used to calibrate the models.The prediction performances of full-spectrum multivariate models and variable-selected(i.e.,the bootstrapping soft shrinkage(BOSS)algorithm)optimization models were compared.The best models were used to map the fertility properties for every profile and the synthetic index for soil fertility of profiles were computed and used to quantify the soil fertility.Then the differences in soil properties with depth and time between laboratory IS technology and traditional sampling methods were compared,and the evolution of soil fertility properties and quality during the process of rice soil formation in the study area was analyzed.Additionally,the potential of this technology for revealing the soil evolution in research was evaluated.The results showed that:(1)Among the six different model combinations,the soil fertility estimation model established by combining the BOSS feature extraction algorithm with support vector machines had the highest accuracy.According to the independent validation results,SOM,TN were predicted excellent(R_P~2≥0.90,RPIQ_P≥4.05),p H was predicted well(0.82≤R_P~2<0.90,3.37≤RPIQ_P<4.05),and the prediction performance for TP and TK was acceptable(0.50≤R_P~2<0.66,2.02≤RPIQ_P<2.70).(2)SOM,TN,and TP contents of the topsoil(0～40 cm)are relatively high and decrease rapidly with depth.In the deeper layers(below 40 cm),the contents are lower but tend to be stable.The vertical variation range of TK content is relatively small,and the regularity is not obvious.Some soil profiles show an increase in TK content with increasing depth.The soil p H gradually increases with depth.By comparing the differences in the characteristics of the changes with depth between the predicted results and the measured values,it was found that the continuous depth variation trends of SOM,TN,TK,and p H extracted from the profiles were basically consistent with the measured values,with more detailed changes within each depth.The continuous depth variation curve of TP showed poor regularity and had no obvious relationship with the measured results.(3)The differences between estimated and measured values of soil fertility properties over time were small.The storage of SOM and TN showed a linear increase on a millennium timescale.The storage of TP increased rapidly during the initial hundreds of years,and then gradually decreased with the increasing age of rice planting.The storage of TK in the surface layer(0～20 cm)showed no correlation with the age of rice planting,while that in the deeper layer(below 20 cm)gradually increased with the age of planting.Soil p H showed a linear decrease with the increasing age of rice planting,indicating that rice cultivation led to the development of soil towards acidity.(4)The continuous depth variation characteristics of soil fertility quality are generally similar to the measured values,with the highest fertility quality in the surface layer(0～30 cm)gradually decreasing with increasing soil depth,and lower fertility quality in deeper layers(30～70 cm)but tending to be stable.The linear function fit the changes in soil fertility quality with planting age better than the power and logarithmic function,indicating that the soil fertility quality in this area may have a high capacity for improvement.The current research results showed that IS technology could be applied to high-resolution mapping of fertility properties across the whole soil profile to achieve rapid monitoring of soil fertility(without the limitation of high precision conditions).For the future application of IS technology,it is necessary to improve the prediction accuracy of phosphorus and potassium.These results can offer new insights on soil fertility evolution and future soil sustainability under long-term anthropogenic influences.