Use of crop canopy sensors in the measurement of sugarcane parameters aiming site-specific nitrogen fertilization management

• Ano de defesa: 2015
• Autor(a) principal: Portz, Gustavo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Agricultura de precisão; Crop spatial variability; Precision agriculture; Taxa variável; Variabilidade espacial das culturas; Variable rate
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/11/11152/tde-17092015-101022/

Plant canopy sensors have emerged as a new tool for in field on-the-go spatially localized prediction of plant biomass and nitrogen (N) uptake by crops in an indirectly and plant indestructible way as base for N variable rate fertilization. Sensors based on the combination of specific reflectance bands from the electromagnetic spectrum constitute the vast majority of canopy sensors, and this principle has already been validated in many crops. Alternatively to this concept, the use of ultrasonic distance sensors to measure crop canopy height has been presented as an option to estimate biomass. Based on that, the aim of this thesis was to validate and refine canopy sensor systems on automated diagnosis of plant parameters aimed the application of N fertilizer according sugarcane needs. Therefore, it was necessary to: 1) validate the prior calibration made for the reflectance sensor (Portz et al., 2012) and to establish the best time to use the sensor over the crop; 2) test the use of the reflectance sensor in comparative strips trials of uniform and sensor based N variable rate application testing algorithms with positive and negative slope and measuring productivity at the end of the season; 3) obtain the relationship between crop canopy height with accumulated biomass and N uptake by the crop during the initial growing season; 4) explore the plant height measured with an ultrasonic sensor comparing the results of biomass and N uptake prediction with those obtained with the reflectance sensor. The experiments were conducted on commercial sugarcane fields, and in strips of the crop with N variable rate application. The experiments were installed over clayey and sandy soils in dry and rainy seasons being all evaluated with the reflectance sensor Yara N-Sensor model ALS (N-Sensor® ALS, Yara International ASA) and partly in comparison with an ultrasonic sensing system Polaroid 6500 (Polaroid, Minnetonka, MN, USA), when the crop had stalk height between 0.2 and 0.9 m. The reflectance sensor calibration fitted with the previous published data showing the interval between 0.3 - 0.5m as the most appropriate to use this sensor over sugarcane. The positive slope algorithm was superior to the negative, except in the situation of clayey soil in rainy season where the response from the negative slope algorithm was higher. The sugarcane plant height was highly correlated with biomass and N uptake by the crop, being possible to estimate the plants canopy height indirectly by the use of an ultrasonic sensor. Comparing the sensor systems, canopy reflectance was better in the early stages of crop as canopy height was more suitable for estimating the cultural parameters when the plants already covered soil in between the rows (+ 0.6 m stalk height), being the sensor systems complementary when fertilization is widely spread in the early crop growth period.