Respuestas espectrales a la fertilización con nitrógeno en el cultivo del arroz (Oryza sativa L.) (original) (raw)

2017, Universidad Nacional de Colombia - Sede Bogotá

The evaluation of nutritional status in rice cultivation through spectral responses has a great potential for decision making regarding the application of agricultural inputs, since the use of the sensors can replace the common destructive sampling in the evaluation of physiological variables, in addition, minimizes costs in the measurement of the same variables since it does not require the use of reagents, nor sophisticated laboratories for the extraction of the information of interest in the culture, which in turn could generate an efficient and specific use of fertilization as well as general management of the crop. This research was carried out in the southern region of Colombia, in the municipality of Saldaña, department of Tolima, at the Las Lagunas Experimental Center of the National Rice Federation (FEDEARROZ), in order to evaluate the relationship between rice spectral responses (Oryza sativa L.), variety Fedearroz-67 and the nutritional status associated with N in different phenological states. A simple factorial design was established in a randomized complete and randomized block design with a randomized N-dose (0, 160, 200 and 240 kg / ha) in four plots of 5m5m and four replicates (macollamiento, maximum tillering, panicle development, flowering, milky grain and mature grain), which involved the use of analysis of variance per phenological stage (since the sampling was treated as independent observations because of the destructive nature of the leaves) and the application of the Bonferroni correction. The majority of the indices had significant differences in the treatments, with clear differentiation of the control with respect to doses 160, 200 and 240 kg/ha of N, but with little or no differentiation between 160 and 200 in the first phenological stages as well as little or no differentiation for 200 and 240 in the final phenological stages. Spectral responses were measured using a FieldSpec 4 HiRes Spectroradiometer (ASD Inc.) spectrometer, from which 48 spectral indices were obtained, which were statistically modeled by regression analysis (linear and nonlinear) to establish relationships with the contents of chlorophylls and the content of N. For the tillering, the indexes with the highest determination coefficients R 2 (around 0,40) were the double difference index (DDI), the modified simple radius (mSR) and the modified simple ratio (BmSR); at the maximum tillering the older R 2 (about 0,85) were the ratios R750/R705, XII Respuestas espectrales a la fertilización con nitrógeno en el cultivo del arroz (Oryza sativa L.) R860/(R550*R708), (R790/R695-740)-1, (R790/R690-725)-1 and the simple radius at 705 nm (SR705); in the development of the panicle the indices associated to the greatest adjustment (R2 around 0,80) were the red border symmetry (RES), the normalized difference vegetation index (NDVI2), the green normalized difference vegetation index (GNDVI), DDI, ratios R740/R720, R710/R760, R695/R760, (R790/R695-740)-1, R850-R710/R850-R680, the standard red border differences (NDRE1 and NDRE2), mSR and specific normalized difference pigment (PSNDb); for flowering the indices were used for maximum values of R 2 (about 0,87) were