Measurement of patterns of soil organic matter in the Mediterranean Montado ecosystem using NIRS

From an environmental point of view, animal production is typically recognized as a major contributor to global greenhouse gas emissions. Nevertheless, it is important to add to this equation the positive impact of the Mediterranean Montado ecosystem on the soil, due to the potential for carbon sequestration through an increase in the soil organic matter content (SOM) resulting from the contribution of trees, pasture and grazing animals. This strategy has recognized interest in mitigating and adapting to climate change. However, monitoring SOM to determine carbon sequestration is time-consuming and expensive, as it requires soil sampling with consecutive laboratory analysis. The concept of precision agriculture (PA) can contribute decisively to resolve this situation. In this work, a combination of near infrared spectroscopy (NIRS) technology and multivariate data analysis (PLS regression) was used to estimate and monitor the evolution of SOM in 8 experimental fields in Portugal. Calibration and validation equations were established between SOM predicted and laboratory reference measurements. The determination coefficients (R2 > 0.80), the RMSE for prediction (RMSE 2.0) demonstrated the use of NIR combined with PLS regression models for efficient estimation of SOM. This results show the practical interest of portable spectrometry, associated with GPS, as expeditious tools for monitoring spatial variability of SOM and, consequently, of soil fertility. Good prospects and opportunities open up for technology-based service providers to develop remote sensing-based computer applications using satellite imagery that enable (i) dynamic management of animal grazing and (ii) definition of homogeneous management zones for the development of site-specific soil fertilization/amendment, particularly important because the vast majority of soils that support the Montado ecosystem in Portugal are poor and acidic