Relation Among Early-Season Canopy Reflectance Readings, Leaf N Content, Stalk N Removal Rate, and Yield of Cane Treated with Varying Sources and Rates of N Fertilizer in Louisiana Production System.

Nitrogen (N) fertilizer is unquestionably the most expensive nutrient input in sugarcane production, but can also bring big returns when managed appropriately. Remote sensing technology has been used for real time N monitoring in many field crops, specifically, grain crops. However, there is limited sensing information on sugarcane. Our study was conducted to determine the relationship between measured soil and agronomic-N related parameters to real time N monitoring based on canopy sensor readings using vegetation indices computed from different wavebands. Leaf tissue, soil samples (0-15 and 15-30cm depth), and canopy reflectance readings were collected from several N response trials established from 2013 to 2016 at LSU AgCenter Sugar Research Station in St. Gabriel, LA. Sampling was made at one, two, four, and eight weeks after N fertilization (WANF). Canopy reflectance readings collected at 660, 710, 735, and 780 nm ± 10 using a GreenSeeker® sensor were transformed to normalized difference vegetation index (NDVI660, NDVI710, and NDVI735). Plots were mechanically harvested and yield was determined by harvesting the stalks from each plot using a weigh wagon with load cells. Ten stalks were randomly selected from the middle row; shredded and dried to determine N uptake. The leaf N content had no relationship with cane tonnage and sugar yield. Nitrogen uptake had a significant relationship with both NDVI and at the red-edge position (REP) at 3 WANF in 2015. Leaf N content had a better relationship with reflectance at REP than with NDVI and red-based vegetation indices (r²= 0.36 to 0.28). Our preliminary results showed that NDVI based on reflectance at REP collected at 3 WANF was related to cane N uptake, tonnage, and sugar yield. Future work is focused on building the database to establish the model for prediction of N-related variables in cane using this vegetation index as a predictor.