Submitted to: Society of Photo-Optical Instrumentation Engineers
Publication Type: Proceedings
Publication Acceptance Date: 9/1/2003
Publication Date: 10/1/2003
Citation: Reddy, K.R., Zhao, D., Kakani, V.G., Read, J.J., Sailaja, K. 2003. Estimating cotton growth and developmental parameters through remote sensing. Proceedings Ecosystems Dynamics, Agricultural Remote Sensing and Modeling, and Site-Specific Agriculture. 5153:277-288. Interpretive Summary: Field studies were conducted to gather information on plant reflectance measures often cited in the literature on remote sensing of crop stress physiology. The objective was to determine responses of leaf nitrogen, leaf and canopy reflectance and yield in cotton to soil moisture stress, fertilizer nitrogen (N) treatments and application of a plant growth regulator (PIX). A symptom of low N supply was loss of chlorophyll in leaves, and this stress was detected early in plants as increased reflectance (where R is reflectance and number is waveband) in the visible wavebands of R550 and R700. Changes in leaf N concentration across treatments could be determined from a simple reflectance ratio in leaves of R517/R413. Application of PIX, a compound that retards cotton growth, indirectly affected reflectance by modifying leaf morphology, physiology, and plant architecture. Among eight reflectance measures examined, the largest correlation with lint yield was obtained from a normalized difference vegetation index (NDVI). Results indicate changes in cotton lint yield were closely related to NDVI values measured at first-flower stage in mid-July, and unrelated to NDVI measured at other times in the season. This study identified stress reflectance measures and indices in cotton that, with few exceptions, are consistent with those reported in the literature. Remote sensing of cotton stress is possible, but determining its relationship to final lint yield will likely require data that is timely and of high quality.
Technical Abstract: Crop reflectance measures sensitive to cotton (Gossypium hirsutum L.) growth and physiology are needed for remote-sensing assessment of crop productivity. This study evaluated cotton growth and yield in relation to narrow-waveband reflectance (400-2500 nm). Plants were grown in the field under four levels of nitrogen (N) fertilizer, four levels of plant growth regulator (PIX), and two levels of irrigation. Reflectance and leaf N concentration were measured weekly or biweekly, plants were harvested at different growth stages to determine plant height, main stem nodes, and dry weight, and final yield was obtained by mechanical harvest. Several reflectance indices were calculated, including simple reflectance ratios (SR) and the normalized difference vegetation index (NDVI). A linear relationship was obtained between leaf N and SR of R517/R413 (r2 = 0.70, n = 150). As expected, N deficiency increased leaf and canopy reflectance in the visible range. A SR of R750/R550 in canopies was associated with changes in plant height and number of main stem nodes (r2 = 0.63~0.68). Changes in leaf area index and biomass were related to either SR or NDVI [R935-R661) /(R935+R661)] (r2 = 0.67~0.78). Results suggest leaf reflectance ratio of R517/R413 may be used to estimate leaf N concentration. A correlation between NDVI at first flower stage and final yield (r2 = 0.64) suggests remote sensing of cotton at mid season may be useful in modeling yield on a site-by-site basis.