Abstract
It is very time consuming and resource costly to rank forages by wet laboratory methods; therefore, near infrared reflectance spectroscopy (NIRS) was used to evaluate hybrid bermudagrass (Cynodon dactylon (L.) Pers.) cultivars for improvements In quality. Quality was assessed by the determination of acid detergent fiber (AOF), crude protein (CP), dry matter (DM), and in vitro dry matter digestibility (iVDMD). The samples were taken from plant breeding trials from 1983,1984,1985, and 1986. They included 32 cultivars, 4 cuttings, and 5 replications each for a total of 640 possible observations per year. This paper describes the process by which the NIR spectrometer was calibrated for each constituent and the effect of various methods of calibration on the performance of the NIRS system to determine quality in bermudagrass hybrids. Excellent one-year equations were obtained from calibrations using the bermudagrass parents in the 1983 samples (R2 = 0.83, 0.97, 0.70; SEC = 0.94, 0.76, 3.64, for ADF, CP, and IVDMD, respectively). Attempts to use outside populations did not produce equivalent results for the first year. The 1983 equations did not fit the 1984 and 1985 samples because of changes that were made to the spectrometer and the yearly differences in the samples. The instrument was then recalibrated with a diverse set of legumes and grass hay samples that contained some bermudagrass. These open population calibrations adequately ranked the bermudagrass entries. We documented that bermudagrass genotypes can be quickly and precisely ranked for forage quality traits via use of open population NIRS calibration equation.
Utility of Near-Infrared Reflectance Spectroscopy to Predict Nutrient Composition and In Vitro Digestibility of Total Mixed Rations
