Quantifying grass productivity using remotely sensed data: an assessment of grassland restoration benefits

This study sought to evaluate the utility of remotely sensed data in estimating the impact of wattle invasion and clearance on native grass species productivity using Sentinel-2 multispectral instrument (MSI) imaging and the partial least squares regression (PLSR) algorithm. Therefore this study assessed grass above ground biomass (AGB) at various levels of wattle invasion In assessing the impacts of wattle invasion on grass AGB the study found that, wattle invasion significantly reduces grass AGB when, compared with uninvaded and cleared plots. Mean grass AGB was 89.64 g m−2, 43.87 g m−2 and 83.36 g m−2 for the cleared, moderately invaded and uninvaded, respectively. The study further found no significant differences between cleared and uninvaded plots (p = 0.826). However, moderately invaded plots were significantly lower than the cleared (p < 0.0001) and uninvaded plots (p = 0.001). In assessing the applicability of remotely sensed data, the findings of this study showed that the most influential variables in estimating biomass were red-edge-based VIs. Specifically, the simple ratio VI (band5/band2) was the most optimal variable for predicting grass AGB across various levels of wattle invasion yielding high accuracies (root mean square error of prediction [RMSEP] = 19.11 g m−2 and R2 = 0.83). Additionally, PLSR results showed that the moderately invaded treatment was most optimally predicted with RMSEP of 13.06 g m−2. Overall, the results underscore the utility of remotely sensed data in monitoring grassland degradation and restoration. Keywords: exotic plants, invasive plants, rehabilitation, remote sensin