Detecção do efeito da adubação nitrogenada em Brachiaria decumbens utilizando técnicas de sensoriamento remoto

• Ano de defesa: 2006
• Autor(a) principal: Silva Júnior, Mário Cupertino da
• Orientador(a): Pinto, Francisco de Assis de Carvalho
• Banca de defesa: Gleriani, José Marinaldo , Ramos, Márcio Mota , Fernandes, Haroldo Carlos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Engenharia Agrícola
• Departamento: Construções rurais e ambiência; Energia na agricultura; Mecanização agrícola; Processamento de produ
• País: BR
• Palavras-chave em Português: Sensoriamento remoto; Agricultura de precisão; Forrageira; Brachiaria decumbens
• Palavras-chave em Inglês: Remote sensing; Precision agriculture; Forage; Brachiaria decumbens
• Área do conhecimento CNPq: CNPQ::CIENCIAS AGRARIAS::ENGENHARIA AGRICOLA::MAQUINAS E IMPLEMENTOS AGRICOLAS
• Link de acesso: http://locus.ufv.br/handle/123456789/3611

The present work was developed with the goal of using remote sensing techniques to detect variations in nutritional status in Brachiaria decumbens. The experimental plots were formed in a pasture area. The data was evaluated usinging a statistical model of a completely randomized design, with five doses of nitrogenous (0, 50, 100, 150 e 200 kg.ha-1), with six repetitions. Two remote sensing systems were used for the evaluations. The systems were composed of digital cameras, wires, a framegrabber and microcomputers. One system set the camera at three meters above ground using a metal framework and the other used one helium-gas balloon to lift the cameras to the heights of 15, 20, 25 and 30 meters. Since the systems can aquire images in two spectral bands simultaneously, the tested pairs of bands were: NIR (monochrome camera with a near infrared high-pass band filter) and R (red band of color camera), NIR and G (green band of color camera), NIR and R (monochrome camera with red pass band filter) and NIR and G (monochrome camera with green pass band filter). The data was acquired in two periods. The first began with the five nitrogenous dose application (N) in February, and the second began after the pasture cutting and the reapplication of the five N doses at the end of March. In the first period, the images acquisition, the SPAD measurements and the nitrogen content in the leaves were determined at 15, 21 and 32 days after fertilization (DAF). In the second period, the same evaluations were done at 28, 36, 45 and 53 DAF. In both periods the vegetation indices NDVI, GNDVI and SAVI mono and color were evaluated using the bands from monochrome cameras with filters and color cameras, respectively. The results using the metal framework in the first period showed that SPAD and leaf N content at 15 DAF and the NDVImono and SAVImono at 21 and 32 DAF were not able to detect the different nutricional status of N. The GNDVImono obtained, in general, the highest values of correlation with leaf N content, SPAD and MS (dry mass) in all dates using the metal framework. Thus, GNDVImono was considered the most indicated index to detect the nutritional variations in this period. In the second period and using the metal framework, all the tested variables were influenced by the different doses of N, and the GNDVIcolor was generally the index that obtained the best correlation with SPAD, leaf N content and MS. The balloon remote sensing system was used only at 28 DAF in the first period at 51 DAF in the second. The three indices NDVIcolor, GNDVIcolor and SAVIcolor at the four tested heights adjusted the linear models of first or second degree and therefore were capable of detecting the effects of the doses of N. The GNDVIcolor was the index with the highest values of correlation with SPAD, leaf N content and MS in the two periods of acquisition at the four tested heights. The best correlations between the indices color and SPAD, leaf N content and MS in the first experimental period were obtained with images acquired at thirty meters of height. The indices using green band showed better relationship with the leaf chlorophyll estimation, the leaf N content and the dry mass productivity than those using the red band in all the periods and tested systems. Both remote sensing systems were capable of detecting the different nutritional status of the forage in the two experimentals periods.