Assessment of Botanical Composition of Cool-Season Pastures Via near-Infrared Reflectance Spectroscopy.

Botanical composition of mixed-species pastures can vary spatially and temporally with changing environmental and management conditions. Knowledge of concentrations of legumes and other species in swards can support nutrient and grazing management planning and allow prediction of grassland and animal performance. Methods of determining proportions of species, or species classes such as cool- and warm-season grasses, legumes, non-leguminous forbs, and live and senescent tissues, include visual estimation, measurement of sorted dry fractions in clipped samples, and laboratory forage chemical composition. Applications of these methods may be limited by their time-consuming nature or reliance on unvalidated assumptions. Near-infrared reflectance spectroscopy (NIRS) may offer an effective means of assessing botanical composition of clipped pasture samples in fresh or dried form. Our objective was to evaluate the accuracy of NIRS analysis of proportions of species and species classes in clipped pasture samples, relative to sorting and weighing dry sample fractions. Spectra (1200-2400 nm wavelength) of fresh and dried samples of individual and mixed cool-season grass, legume, and non-leguminous forb species were obtained from a range of environmental conditions and growth stages. Sample composition from sorting, drying, and weighing was paired with spectra for development of prediction equations for species and species classes. The effectiveness of NIRS for prediction of botanical composition will be compared with reference values from sorting, drying, and weighing.