|Richard Jr, Edward|
Submitted to: American Society of Sugar Cane Technologists
Publication Type: Abstract Only
Publication Acceptance Date: May 14, 2007
Publication Date: June 11, 2007
Citation: Johnson, R.M., Viator, R.P., Richard Jr, E.P. 2007. Estimation of sugarcane sucrose and biomass with remote sensing techniques [abstract]. Journal of the American Society of Sugar Cane Technologists. 27:68. Technical Abstract: Remote sensing techniques were used to predict sucrose levels (TRS) and gross cane yield in field-grown sugarcane. To estimate sucrose levels, leaves were collected from plant-cane and first-ratoon sugarcane plants from the variety maturity studies conducted at the USDA-ARS-SRRC, Sugarcane Research Unit’s, Ardoyne Research Farm in Schriever, LA. Leaf samples were collected monthly between September 25 and November 20, 2006, from 10 varieties in the plant-cane crop, and bi-weekly between August 28 and December 5, 2006, from nine varieties in the first-ratoon crop. There were four replications in each test and four leaf samples were collected from each plot. High resolution, hyperspectral leaf reflectance measurements were made in the laboratory, shortly after sampling, using an Ocean Optics SD-2000 fiber optic spectrometer and a halogen light source. Whole-stalk samples were collected on each sampling date and analyzed for sucrose levels using the core press method. Reflectance data were condensed into 10 or 20-nm intervals and then with sucrose data were subjected to analysis of variance and linear discriminant analysis to predict TRS levels. Biomass of two first-ratoon commercial sugarcane fields was estimated using aerial imagery acquired on November 6, 2006. To provide a ground truthing data set, selected rows, from each field, were harvested in 15-meter increments with a single-row, chopper harvester, with weights determined using a weigh wagon. A total of 176 plots were harvested in this manner. Yield data were subjected to variogram analysis and block kriging to construct yield maps of each field for comparison with biomass estimates generated from the acquired aerial imagery. Multivariate analysis of leaf reflectance and TRS data from the plant-cane maturity study resulted in a 73% correct classification of TRS levels when all varieties were combined. When varieties were considered separately TRS levels could be classified correctly in 100% of the cases examined. In the first-ratoon maturity study TRS levels were correctly classified in 55% of the cases in the combined data set and in 100% of the cases when varieties were considered separately. Finally, kriged maps of sugarcane gross-cane yield, demonstrated a close agreement with estimated biomass levels for both of the fields examined. Successful implementation of these remote sensing technologies would allow producers to estimate both sucrose and gross cane yields in the field prior to harvest. This information could then be used to more effectively schedule harvest operations and identify yield limiting crop stress situations. Finally, biomass maps could also be potentially used to develop variable-rate ripener application systems.