Comparison of Desktop and Handheld Near-Infrared Spectroscopy Devices to Determine Forage Nutritive Value.

Near-infrared (NIR) spectroscopy is a non-destructive analysis technology that can be used for rapid determination of constituents of various materials. The publication by Norris et al. (1976) marked a milestone for the utilization of NIR-spectra to predict forage nutritive value and animal intake. The corollary is that NIR-spectroscopy applications are only as good as the prediction models developed out of meaningful biological or chemical entities, hence a spectro-chemical model. Therefore, the quality of predictions for plant-, soil- and animal-responses from NIR spectra depend on accurate and precise chemometric modeling. The first step in development of NIR-models is collection of spectral data. There are several devices available in the market to collect spectral data with a wide range of specifications (e.g., spectral range and resolution, desktop vs. portable) and price. Traditional NIR devices have a desktop-type configuration, high resolution and associated high cost, and they have been mainly used by trained personnel working in commercial and research laboratories. Recent technological advances have brought to market more affordable handheld NIR devices that could potentially be a tool in the hand of farmers for on-site and real- or almost-real time analysis. In this poster, we present the results of 1) development of NIR spectroscopy models using three NIR devices of contrasting specifications to predict crude protein (CP), acid detergent fiber (ADF), neutral detergent fiber (NDF), and in-vitro true dry matter digestibility (IVTD) for bermudagrass [Cynodon dactylon (L.) Pers] and switchgrass(Panicum virgatum L.); and 2) the comparison of predictions among the three NIR devices.