Auxin injury on cotton, I: Aerial reflectance data, crop injury, and yield

Abstract

Synthetic auxin herbicide movement onto sensitive cotton (Gossypium hirsutum L.) cultivars has impacted many US cotton hectares. The spatial scope and severity of auxin damage in-season is typically estimated by an agronomist. The use of remote sensing technology has the potential to objectively quantify the spatial scope and severity of auxin damage. Experiments were conducted in 2019, 2020, and 2021 in Grand Junction, TN, to determine: (1) the effect of reflectance data collection timing; (2) the effect of auxin exposure timing; (3) the value of near infrared and red-edge (RE) reflectance versus reflectance within the visible spectrum data; and (4) if/how visual injury relates to aerial reflectance data. Applications of 2,4-D or dicamba were made to susceptible cotton cultivars at 1X, 1/4X, 1/16X, 1/64X, 1/256X, and 1/1024X rates at either matchhead square (MHS) or 2 weeks after first bloom (FB+2WK). Non-treated controls were also included for each application timing. Aerial reflectance data were collected 7, 14, 21, and 28 days after application. Unsupervised classification of images into pixels with and without vegetation did not increase correlations between vegetation indices (VIs) and application rate. Although Vis, which generated the strongest correlations with application rate, visual injury, and relative lint yield, were generally RE based, similar correlations were also noted with visible spectrum VIs. Correlations were greater when auxin injury occurred at MHS than FB+2WK. Results suggest reflectance measured within the visible spectrum can quantify the scope and severity of auxin injury if the injury occurs early during the growing season.