Testing a novel pasture quality index using remote sensing tools in semiarid and Mediterranean grasslands

Estimating the nutritional value of herbage is a core aspect of livestock pasture management. Quantitative information related to significant findings is essential for assessing nutritional provision and preventing pasture degradation. Here, we developed a new Pasture Quality Index (PQI) that indicates the nutritional quality of pastures based on their nutritional value and the composition of plant functional groups of the vegetation. We applied the model on a regional scale using VENμS satellite data, collected more than 450 plant samples, and applied pasture quality parameters analysis. Pasture samples were collected from paddocks under two different grazing intensities in Israel in a semiarid and a Mediterranean grassland. Pasture quality was estimated five times over two years at mid- and peak vegetation growth stages. The PQI included: protein content, fiber content, dry matter digestibility and the proportion of unpalatable spiny thistles. A support vector machine regression model was used to train statistical models that relate ground truth PQI data to the reflectance values of the VENμS satellite. Large-scale forage quality maps at 5 m spatial resolution were created. Fine-resolution drone orthomosaics were used to enhance the satellite-based predictions accuracy, resulting in a model accuracy of R2 = 0.86. We observed that nutritional quality showed strong dependence on seasonality and the grazing regime, with higher quality observed during mid-growth compared to peak growth. Higher quality was also observed under grazing compared to ungrazed paddocks. The occurrence of unpalatable thistles in grazed paddocks significantly reduced the pasture quality at the Mediterranean grassland. To summarize, the PQI enables the integration of several quality indicators into an index that helps in detecting the effects of grazing management on rangelands. In addition, high accuracy was achieved in relating the overall PQI and high spatiotemporal resolution satellite imagery data. Our developed methodology enables site-specific, spatially explicit, frequent, and area-extensive pasture quality assessment that can aid in optimizing livestock management in various ecosystems.