Canopy Hyperspectral Reflectance of Redroot Pigweed versus Okra and Super Okra Leaf Cotton

Abstract

Redroot pigweed ( Amaranthus retroflexus L . ) is a nuisance weed that affects cotton ( Gossypium hirsutum L . ) growth and yield worldwide. Being able to distinguish redroot pigweed from cotton would help producers and crop consultants better implement strategies used to suppress and control it. Hyperspectral reflectance properties of weed and crop canopies have been used to differentiate between them. Currently, no information is available on the application of hyperspectral data to distinguish redroot pigweed from cotton with different leaf shapes. Positive results will further support the exploration of remote sensing technology for distinguishing redroot pigweed from cotton. The objectives were to compare canopy hyperspectral reflectance of redroot pigweed to canopy hyperspectral reflectance of okra and super okra leaf cotton and to identify regions of the spectrum in which differences exist in their reflectance properties. Hyperspectral reflectance measurements of redroot pigweed and cotton were obtained with a spectroradimeter on May 6 and June 27, 2019. Plants grown in a greenhouse were used for this study. One-hundred and sixty-two 10-nm bands (400 - 2350 nm spectral range) were evaluated with analysis of variance ( p ≤ 0.05) and Dunnett’s test ( p ≤ 0.05) to determine the wavebands that were useful for separating redroot pigweed from okra leaf and super okra leaf cotton. The following bands were consistent in distinguishing redroot pigweed and okra leaf cotton on both dates: 420 nm, 510 - 650 nm, 690 - 740 nm, and 2000 - 2010 nm; whereas, 400 - 500 nm, 1480 - 1780 nm, and 1990 - 2350 nm were identified for both dates for separating redroot pigweed from super okra leaf cotton. Commercial imaging systems used on ground-based or airborne platforms can be easily tuned into the spectral bands listed in this study, thus providing managers with a tool to use for identifying redroot pigweed in cotton production systems.