|Chen, Ming - UNIV OF FLORIDA IFAS|
|Gilbert, Robert - UNIV OF FLORIDA IFAS|
|Daroub, Samira - UNIV OF FLORIDA IFAS|
|Barton Ii, Franklin|
|Wan, Yongshan - S. FL WATER MANAGEMENT|
Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 28, 2002
Publication Date: N/A
Interpretive Summary: Restoration of the Florida Everglades is a national priority. As part of this restoration process, there are legislative controls on the amount of phosphorus (P) that can be discharged to the natural Everglades from Florida sugarcane farms. Best management practices (BMPs) to meet these requirements cost farmers $153 per ha to install and $9 per ha to maintain. This research tested near infra red spectroscopy (NIR) as a potential tool to analyze the amount of leaf P in Florida's sugarcane. An equation was developed for the NIR that analyzed P in commercial sugarcane virtually as accurately as traditional methods. Important advantages to the NIR analyses were that they were quick, about 10 NIR samples could be processed for every sample processed by reference chemistry, and there were no hazardous chemical exposures or wastes with NIR. If these NIR prediction equations are validated in further studies, they could be used in selection programs to improve chances of identifying new cultivars that remove more soil P. Also, growers may be able to integrate NIR as a tool into precision agriculture programs aimed at improving existing or developing new BMPs for P reduction.
Technical Abstract: Sugarcane is the primary crop of the Everglades Agricultural Area (EAA) of Florida. Characterization of P uptake among sugarcane genotypes may help EAA growers maintain reduced P discharges, an important component of Everglades restoration. The purpose of this study was to evaluate near infrared reflectance spectroscopy (NIRS) as a potential tool to analyze sugarcane leaf P conc. Initial prediction equations were similar or moderately lower in accuracy to those previously reported for predicting P in other grasses with NIRS. However, further multiplicative scatter correction of spectral data and the elimination of unneeded wavelength segments by "Martens Uncertainty" regression improved the prediction regression coefficient to 0.98 for clonal material similar to commercial EAA sugarcane, but the coefficient remained at about 0.70 for predicting leaf P in large numbers of genetically unique offspring of these clones. Use of NIRS is proposed in a selection program to screen large numbers of genotypes for coarse measurements, and also for precise measurements of leaf P of commercial sugarcane.