In‐Season Nitrogen Status Sensing in Irrigated Cotton

Abstract

Pre‐plant soil NO−3–N tests and petiole NO−3–N analysis are bases for Upland cotton (Gossypium hirsutum L.) N management in the western USA. Alternative approaches include proximal multispectral reflectance sensing and chlorophyll meter readings. Our objective was to determine if spectral reflectance and chlorophyll meter measurements correlate with cotton leaf N and biomass. Urea ammonium nitrate was applied after emergence and with low energy precision (LEPA) center‐pivot, surface or subsurface drip irrigation water up to peak bloom. Multispectral reflectance readings 0.5 m above the canopy, chlorophyll meter readings, and biomass samplings were taken at early squaring, early bloom, and peak bloom for 3 site‐years in Lubbock, TX and Ropesville, TX. Green vegetative indices (GVI) and green normalized difference vegetative indices (GNDVI) calculated from reflectance data generally correlated better with leaf N and leaf N accumulation than did red vegetative indices (RVI) and red normalized difference vegetative indices (RNDVI). Biomass and lint yield correlated more often with red‐based indices than green‐based indices. Chlorophyll meter readings correlated with leaf N as often as GVI and GNDVI did. Biomass, however was poorly related to chlorophyll meter readings. These results demonstrate the effectiveness of GVI, GNDVI, and chlorophyll meter readings in assessing leaf N, and RVI and RNDVI in assessing cotton biomass. However, we recommend converting vegetative indices or chlorophyll meter readings to sufficiency indices, which are calculated from indices or readings relative to well‐fertilized plots. Sufficiency indices were able to successfully predict little or no need for in‐season N fertilizer in the low‐yielding 2000 crops (sufficiency index > 0.95), and predicted greater need of N fertilizer in the high‐yielding 2001 crop (sufficiency index < 0.95).