Location: Sugarcane ResearchTitle: Integrating Near Infrared Spectroscopy (NIR) into the USDA-ARS sugarcane breeding program in Houma, LA Author
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 3/17/2015
Publication Date: N/A
Technical Abstract: Near Infrared Spectroscopy (NIRs) is a relatively new technique that has the potential to benefit Louisiana’s sugarcane industry, and is being successfully used in other parts of the world (e.g., South Africa) and even Florida. Recently, the USDA-ARS in Houma, LA purchased a NIR Cane Presentation System from Bruker OpticsTM. A modified version of this system is used for cane payments in South Africa. It is a self-contained sampling system, with a cane shredder connected to a conveyer. The cane is moved by conveyer, then “fluffed” using metal fingers, and subsequently passed through a window with a NIR detector. The NIR detector measures light reflected off the shredded cane. The machine is calibrated based on the light spectra and, with the right calibrations, can instantaneously determine sucrose, pol, fiber, moisture, and other parameters such as color and mannitol/dextran (sugarcane deterioration compounds). Other parameters such as conductivity ash and reducing sugars, can be measured with NIR and this would provide valuable processing information to the factory staff. It may even be possible to measure insoluble starch in raw juices (Eggleston et al, 2013). The machine was tested in December 2014 using samples from the basic breeding program. After only one day of use, the average sample took between 1.0 and 1.33 minutes to process. The system is simple to use, with minimal software controls, and the operator simply feeds the cane into the shredder, enters the sample ID, and presses the “start” button. This type of system has the potential to minimize laboratory errors and discrepancies. Furthermore, the speed in which samples can be processed ultimately will increase the number of samples the laboratory can handle, thus reducing sub-sampling error.