Very-Fine Resolution Multispectral Imagery for Rangeland Forage Quality Assessment

ABSTRACT Understanding forage nutritional dynamics across rangeland landscapes is essential for sustainable grazing management and wildlife habitat conservation. Traditional field sampling methods are constrained by limited spatial coverage, time, and labor-intensive protocols. The goal of this research was to assess the spectral and temporal relationships of native grasses’ nutritional content in a semi-arid rangeland. To this end, monthly drone flights were conducted from November 2023 to May 2024 over monoculture plots of pink pappusgrass ( Pappophorum bicolor E. Fourn), seacoast bluestem ( Schizachyrium scoparium var. littorale Nash K.N. Gandhi & F.E. Smeins), and slender grama ( Bouteloua repens Kunth Scribn. & Merr.) using a DJI Matrice 210 drone equipped with a MicaSense Altum multispectral sensor capturing five spectral bands (blue, green, red, red-edge, near-infrared). Vegetation samples were collected monthly and analyzed for crude protein and digestible organic matter using near-infrared spectroscopy. We performed exploratory regression analyses to establish a relationship between spectral data and nutritional parameters. Our results show that pink pappusgrass crude protein ranged from 6% to ​14% (root mean square error RMSE = 1.86) and digestible organic matter reached 70% (RMSE = 7.92). Seacoast bluestem exhibited crude protein levels of 3–8% (RMSE = 1.36) and digestible organic matter up to 65% (RMSE = 5.22), while slender grama crude protein ranged between 5% and ​10% (RMSE = 1.47) and digestible organic matter between 50% and ​60% (RMSE = 4.03). Our regression analyses between spectral data and nutritional content show R ² values ranging between 0.48 and 0.80. Normalized Red and Enhanced Normalized Difference Vegetation Index indices are consistently explaining the variability of crude protein, whereas the blue, red, and NIR bands explain the variability for digestible organic matter for all three species. Multispectral imagery spectral information can potentially be used to assess the nutritional value of rangeland species in semi-arid landscapes. This work shows novel insights for integrating remote sensing technology with traditional rangeland monitoring methods in semi-arid landscapes.