4.7 Article

Robust Vis-NIRS models for rapid assessment of soil organic carbon and nitrogen in Feralsols Haplic soils from different tillage management practices

Journal

COMPUTERS AND ELECTRONICS IN AGRICULTURE
Volume 153, Issue -, Pages 295-301

Publisher

ELSEVIER SCI LTD
DOI: 10.1016/j.compag.2018.08.036

Keywords

Chemometrics; Partial least squares regression; Tillage systems; No-till; Ferralsols

Funding

  1. National Research Foundation
  2. KwaZulu-Natal Department of Agriculture
  3. Troika Fund

Ask authors/readers for more resources

Application of spectroscopy for assessment of soil nutrition in the field may be affected by the depth at which the radiation spreads to, the analysed nutrient, the nutrient level or management practices such as tillage systems. In this study, the use of visible to near infrared spectroscopy (Vis-NIRS) was explored as a technique to predict soil organic carbon (SOC) and soil organic nitrogen (SON) in different tillage management practices, varying rates of nitrogen and different depth distribution. The tillage treatments were no-till (NT), rotational tillage (RT) and conventional tillage (CT) and nitrogen was applied at a rate of 0, 100 and 200 kg/ha as lime ammonium nitrate. The reflectance spectra of samples from 0 to 10, 10 to 20 and 20 to 30 cm depths were acquired from all tillage treatments using a laboratory bench-top monochromator NIR Systems Model XDS spectrometer. Partial least square regression (PLSR) models were developed using leave-one-out cross validation method. The models were then tested on independent test samples (54) randomly selected from the total 324 samples. Principal component analysis (PCA) was used to differentiate SOC in different tillage treatments and different rates of nitrogen. The best prediction model was observed for SOC with the coefficient of determination (R-2) = 0.993, root mean square error of prediction (RMSEP) = 0.157% and residual predictive deviation (RPD) = 2.55 compared with R-2 = 0.661, RMSEP = 0.019%, RPD = 2.11 for SON. PCA was able to cluster soil samples according to the rates of applied nitrogen but not the tillage systems and depths. This study demonstrated the application of Vis-NIRS for analysis of SOC and SON from soils with varying levels of the nutrients. The robustness of developed models was associated with analysing samples from different depths and combining them during calibrations. Therefore, models developed in this manner were recommended for technicians in the field since they would warrant the assessment of soil in different tillage systems and in the entire rooting zone of crops.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available