|SAHA, UTTAM - University Of Georgia|
|VANN, RACHEL - North Carolina State University|
|REBERG-HORTON, CHRIS - North Carolina State University|
|CASTILLO, MIGUEL - North Carolina State University|
|SONON, LETICIA - University Of Georgia|
Submitted to: Journal of the Science of Food and Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/1/2018
Publication Date: 2/9/2018
Citation: Saha, U., Vann, R., Reberg-Horton, C., Castillo, M., Mcgee, R.J., Mirsky, S.B., Sonon, L. 2018. Near-Infrared Spectroscopic models for analysis of cover crop and forage quality constitutents of winter pea (Pisum sativum L.). Journal of the Science of Food and Agriculture. 98)11_:4253-4267. https://doi.org/10.1002/jsfa.8947.
Interpretive Summary: Winter pea (Pisum sativum L.), a cool season legume, grows well in a wide geographic region both as forage and green manure and cover crops. However, analysis of forage and cover crop quality constituents by conventional wet chemistry methods is laborious, slow, and costly. In contrast, Near Infrared Reflectance Spectroscopy (NIRS) is precise, accurate, rapid, and cheap, and could be appropriate. We developed and validated NIRS calibration models for analysis of this crop. These models, without any further updating, can be reliably applied in the routine analysis of 11 forage and cover crop quality parameters of winter pea.
Technical Abstract: Winter pea (Pisum sativum L.) is a legume that has desirable attributes for use as a cover crop in the Southeast. Winter pea can produce high biomass in this region, allowing for substantial N delivery to the following cash crop. Beyond biomass production and N provision, winter pea has a high nutritive concentration; 18-30% protein, 35-50% starch, and high concentrations of the amino acid lysine, which is deficient in small grains. Pea can be grazed, cut for hay, and/or ensiled. Pea silage is more degradable in ruminant animals than barley silage. For these reasons, winter pea is also a desirable forage crop. Therefore, there is a great need for analysis of this crop for its cover crop quality and forage quality constituents, namely moisture, dry-matter (DM), total nitrogen (TN), crude protein (CP), acid detergent fiber (ADF), neutral detergent fiber (NDF), lignin, cellulose, and non-fibrous carbohydrate (NFC). However, conventional laboratory reference methods for these analyses are laborious, time consuming, and costly. In contrast, the near infra-red spectroscopic (NIRS) technique, if based on properly developed and validated calibration models, could be an excellent alternative to provide rapid and low cost analyses without sacrificing precision and accuracy. In order to develop and validate NIRS calibration models, samples were collected from research plots of 18 winter pea genotypes grown in Maryland and North Carolina during the 2015-2016 and 2016-2017 growing seasons. Using conventional wet chemistry methods of the Association of Official Analytical Chemistry and the National Forage Testing Association, DM, TN, CP, ADF, NDF, lignin, cellulose and NFC were determined. Subsamples (5 g) from the wet chemistry subjected to 32 successive scans in both visible and NIR regions (400 to 2498 nm) at 2 nm intervals. Absorption of radiation in the 400-2498nm region was used to develop calibration models for moisture, DM, TN, CP, ADF, NDF, ash, lignin, cellulose, hemicellulose, and NFC. Development of calibration models and validation of the developed models were performed using the global program in WinISI software. The predictability of the NIRS calibration models for moisture, DM, TN, CP, ADF, NDF, lignin, cellulose, ash, and NFC was tested through external validation, using 46 samples for moisture and DM, and 49 samples for each of the other 9 parameters. The chemometric techniques established valid quantitative relationships between the visible-NIR light absorption characteristics and laboratory measured values of 11 forage and cover crop quality parameters of winter pea thereby developing acceptable NIRS calibration prediction models. All 11 calibration models developed prediction abilities with acceptable accuracy; nine (moisture, DM, TN, CP, ADF, NDF, lignin, cellulose, NFC) appeared suitable for quantitative prediction and the other two (ash, hemicellulose) for at least qualitative screening, if not for quantitative prediction. This method will provide rapid and high throughput analysis, and will replace the time-consuming traditional laboratory methods. This method is also considered a “cheap” and “green chemistry” because it does not involve any chemicals, nor does it generate any hazardous waste.