Estimation of direct-seeded guayule cover, crop coefficient, and yield using UAS-based multispectral and RGB data

Authors: ELSHIKHA, DIAA ELDIN - University Of Arizona; Hunsaker, Douglas - Doug; WALLER, PETER - University Of Arizona; Thorp, Kelly; DIERIG, DAVID - Bridgestone Americas Tire Operations; WANG, GUANGYAO - Bridgestone Americas Tire Operations; CRUZ, V - Bridgestone Americas Tire Operations; KATTERMAN, MATTHEW - University Of Arizona; BRONSON, KEVIN - Retired ARS Employee; Wall, Gerard - Gary; Thompson, Alison

Submitted to: Agricultural Water Management
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/7/2022
Publication Date: 2/11/2022
Citation: Elshikha, D.M., Hunsaker, D.J., Waller, P.M., Thorp, K.R., Dierig, D., Wang, G., Cruz, V.M., Katterman, M.E., Bronson, K., Wall, G.W., Thompson, A.L. 2022. Estimation of direct-seeded guayule cover, crop coefficient, and yield using UAS-based multispectral and RGB data. Agricultural Water Management. 265. Article 107540. https://doi.org/10.1016/j.agwat.2022.107540.
DOI: https://doi.org/10.1016/j.agwat.2022.107540

Interpretive Summary: In recent years, efforts have increased to commercialize guayule in the arid US Southwest for domestic supplies of natural rubber. Some information on guayule crop and irrigation management is available for guayule growers. Use of remote sensing data collected by small drones can assist growers in crop management decisions, however, such information is practically non-existent for guayule. ARS scientists in Maricopa, Arizona, evaluated several vegetation indices using remote sensing data collected by drones flown over a guayule field. The research determined that the normalized difference vegetation index (NDVI) was highly related to guayule crop growth and final yield. The research also developed NDVI relationships to estimate guayule crop evapotranspiration (ET) throughout the growing season. The ET of a crop is a most important determiner of irrigation scheduling requirements. Use and application of drone data are becoming more prevalent in assisting agricultural management. This study increases the viability of using drone-sensors in guayule crop and irrigation management in the US Southwest. The research will be of interest to the US Rubber Industry, including Tire Manufacturers, irrigation consultants, water district water managers, and other research investigators of guayule.

Technical Abstract: Guayule (Parthenium argentatum, A. Gray), a perennial desert shrub, produces high-quality natural rubber and is targeted as a domestic natural rubber source in the U.S. While commercialization efforts for guayule are on-going, crop management requires plant growth and health monitoring, irrigation requirement assessment, and final yield estimation. Such assistance for management could be provided with remote sensing (RS) data, collected by unmanned aircraft systems, such as drones. However, information on RS applications for guayule is not currently available. In this study, field and RS data from a guayule irrigation experiment conducted at Maricopa, Arizona in 2018-2020 were used. In-season field measurements included plant height (h), fractional canopy cover (fc), basal (Kcb) and single (Kc) crop coefficients, and final yields of dry biomass (DB), rubber (RY), and resin (ReY). RS data were via two drones, one carrying a multispectral sensor (MS) with bands in the red, green, blue (RGB), red edge, near infrared, and thermal, and the other carrying a built-in RGB camera. The objectives of this paper were to compare vegetations indices from MS data (normalized difference vegetative index, NDVI) and RGB data (triangular greenness index, TGI); and derive prediction models for estimating h, fc, Kcb, Kc, and yield as functions of the multispectral and RGB indices. While NDVI and TGI showed similar seasonal trends, prediction of h (r2 = 0.89) and fc (r2 = 0.91) by NDVI was better than by TGI (r2 = 0.53 and 0.38, respectively). However, prediction of fc based on RGB image evaluation was good (r2 = 0.96). The Kcb and Kc were 2nd order functions of NDVI during the first year of active guayule growth (r2 = 0.86 and 0.76, respectively) and during the following winter dormancy period (r2 = 0.88 and 0.91, respectively). However, they were linear functions of NDVI during the second year (r2 ˜ 0.64) when guayule was at full cover. Final DB, RY, and ReY were predicted by both NDVI (r2= 0.75, 0.53, and 0.70, respectively) and TGI (r2= 0.72, 0.48, and 0.65, respectively). The RS-based models enable estimation of irrigation requirements and yields in guayule production fields in the U.S.