Rice Response to Nitrogen Fertilization and Comparison of Unmanned Aerial Systems and Active Crop Canopy Sensors Vegetative Index to Estimate Rice Yield Potential
Semester of Graduation
Master of Science (MS)
School of Plant, Environmental and Soil Sciences
Nitrogen (N) fertilization is a key component in producing profitable, maximized rice grain yields because yield is directly affected by N fertilizer applications. Economical optimum N rate (EONR) is used to estimate where the N fertilization rate impacts rice grain yield but is still economically efficient. Three common response models, linear-plateau, quadratic-plateau, and quadratic models were used to determine the response of rice to N fertilizer to determine the optimum N fertilization rate. The objective of the first part of this study was to evaluate the models by assessing the coefficients of determination (R2), maximum rice grain yields each model produced, and the estimated EONRs of fertilization. Coefficients of determination (R2) of the linear-plateau, quadratic-plateau, and quadratic were found to be similar (0.77, 0.79, 0.78). Other factors beyond just R2 alone need to be taken into consideration when choosing which response model best fits a data set and should be used to estimate the EONR of fertilization for an individual variety.
Normalized difference vegetation index (NDVI) is a known indication of yield potential, one component needed to determine mid-season N requirements. The GreenSeeker has been the pre-dominant tool used to collect NDVI measurements. Unmanned aerial systems (UAS) have shown potential to collect NDVI measurements also. The objectives of the second part of this study were to: 1) evaluate the relationship between GreenSeeker (an active sensor) derived NDVI and UAS (a passive sensor) derived NDVI, and 2) evaluate the ability of GreenSeeker and UAS derived NDVI to estimate rice yield potential. This research was done in 2017 and 2018 at 5 locations in Louisiana. Remote sensor data was taken between panicle initiation and panicle differentiation using a GreenSeeker and UAS mounted remote sensor. All 5 locations showed a highly significant correlation between GreenSeeker and UAS derived NDVI. The linear relationship between GreenSeeker and UAS derived NDVI to rice grain yield were not similar. The different relationships could have been caused by the differences between ground and air-borne based sensors. More research will need to be conducted before UAS mounted sensors can be used to accurately predict mid-season N needs in rice.
Coker, Anna, "Rice Response to Nitrogen Fertilization and Comparison of Unmanned Aerial Systems and Active Crop Canopy Sensors Vegetative Index to Estimate Rice Yield Potential" (2019). LSU Master's Theses. 5020.