Do model choice and sample ratios separately or simultaneously influence soil organic matter prediction?

This study was performed to examine the separate and simultaneous influence of predictive models’choice alongside sample ratios selection in soil organic matter(SOM).The research was carried out in northern Morocco,characterized by relatively cold weather and diverse geological conditions.The dataset herein used accounted for 1591 soil samples,which were randomly split into the following ratios:10%(∼150 sample ratio),20%(∼250 sample ratio),35%(∼450 sample ratio),50%(∼600 sample ratio)and 95%(∼1200 sample ratio).Models herein involved were ordinary kriging(OK),regression kriging(RK),multiple linear regression(MLR),random forest(RF),quantile regression forest(QRF),Gaussian process regression(GPR)and an ensemble model.The findings in the study showed that the accuracy of SOM prediction is sensitive to both predictive models and sample ratios.OK combined with 95%sample ratio performed equally to RF in conjunction with all the sample ratios,as the latter did not show much sensitivity to sample ratios.ANOVA results revealed that RF with a∼10%sample ratio could also be optimum for predicting SOM in the study area.In conclusion,the findings herein reported could be instrumental for producing cost-effective detailed and accurate spatial estimation of SOM in other sites.Furthermore,they could serve as a baseline study for future research in the region or elsewhere.Therefore,we recommend conducting series of simulation of all possible combinations between various predictive models and sample ratios as a preliminary step in soil organic matter prediction.

Mapping soil nutrients via different covariates combinations:theory and an example from Morocco

Background:Mapping of soil nutrients using different covariates was carried out in northern Morocco.This study was undertaken in response to the region’s urgent requirement for an updated soil map.It aimed to test various covariates combinations for predicting the variability in soil properties using ordinary kriging and kriging with external drift.Methods:A total of 1819 soil samples were collected at a depth of 0–40 cm using the 1-km grid sampling method.Samples were screened for their pH,soil organic matter(SOM),potassium(K_(2)O),and phosphorus(P_(2)O_(5))using standard laboratory protocols.Terrain attributes(T)computed using a 30-m resolution digital elevation model,bioclimatic data(C),and vegetation indices(V)were used as covariates in the study.Each targeted soil property was modeled using covariates separately and then combined(e.g.,pH~T,pH~C,pH~V,and pH~T+C+V).k=tenfold cross-validation was applied to examine the performance of each employed model.The statistical parameter RMSE was used to determine the accuracy of different models.Results:The pH of the area is slightly above the neutral level with a corresponding 7.82%of SOM,290.34 ppm of K_(2)O,and 100.86 ppm of P_(2)O_(5).This was used for all the selected targeted soil properties.As a result,the studied soil properties showed a linear relationship with the selected covariates.pH,SOM,and K 2O presented a moderate spatial autocorrelation,while P2O5 revealed a strong autocorrelation.The cross-validation result revealed that soil pH(RMSE=0.281)and SOM(RMSE=9.505%)were best predicted by climatic variables.P_(2)O_(5)(RMSE=106.511 ppm)produced the best maps with climate,while K_(2)O(RMSE=209.764 ppm)yielded the best map with terrain attributes.Conclusions:The findings suggest that a combination of too many environmental covariates might not provide the actual variability of a targeted soil property.This demonstrates that specific covariates with close relationships with certain soil properties might perform better than the compilation of different environmental covariates,introducing errors due to randomness.In brief,the approach of the present study is new and can be inspiring to decision-makers in the region and other world areas as well.

Soil quality estimation using environmental covariates and predictive models:an example from tropical soils of Nigeria

Background:Information addressing soil quality in developing countries often depends on results from small experimental plots,which are later extrapolated to vast areas of agricultural land.This approach often results in misin-formation to end-users of land for sustainable soil nutrient management.The objective of this study was to estimate the spatial variability of soil quality index(SQI)at regional scale with predictive models using soil–environmental covariates.Methods:A total of 110 composite soil samples(0–30 cm depth)were collected by stratified random sampling schemes at 2–5 km intervals across the Cross River State,Nigeria,and selected soil physical and chemical properties were determined.We employed environmental covariates derived from a digital elevation model(DEM)and Senti-nel-2 imageries for our modelling regime.We measured soil quality using two approaches[total data set(TDS)and minimum data set(MDS)].Two scoring functions were also applied,linear(L)and non-linear(NL),yielding four indices(MDS_L,MDS_NL,TDS_L,and TDS_NL).Eleven soil quality indicators were used as TDS and were further screened for MDS using principal component analysis(PCA).Random forest(RF),support vector regression(SVR),regression kriging(RK),Cubist regression,and geographically weighted regression(GWR)were applied to predict SQI in unsampled locations.Results:The computed SQI via MDS_L was classified into five classes:≤0.38,0.38–0.48,0.48–0.58,0.58–0.68,and≥0.68,representing very low(classⅤ),low(classⅣ),moderate(classⅢ),high(classⅡ)and very high(classⅠ)soil quality,respectively.GWR model was robust in predicting soil quality(R^(2)=0.21,CCC=0.39,RMSE=0.15),while RF was a model with inferior performance(R^(2)=0.02,CCC=0.32,RMSE=0.15).Soil quality was high in the southern region and low in the northern region.High soil quality class(>49%)and moderate soil quality class(>14%)dominate the study area in all predicted models used.Conclusions:Structural stability index,sand content,soil oganic carbon content,and mean weight diameter of aggregates were the parameters used in establishing regional soil quality indices,while land surface water index,Sentinel-2 near-infrared band,plane curvature,and clay index were the most important variables affecting soil quality variability.The MDS_L and GWR are effective and useful models to identify the key soil properties for assessing soil quality,which can provide guidance for site-specific management of soils developed on diverse parent materials.