Date of Graduation

12-2019

Document Type

Thesis

Degree Name

Master of Science in Plant Pathology (MS)

Degree Level

Graduate

Department

Entomology and Plant Pathology

Advisor/Mentor

Rupe, John C.

Committee Member

Spurlock, Terry N.

Second Committee Member

Ham, Richard G.

Keywords

Fungicide resistant plant pathogens; Fungicide efficacy; Plant disease management

Abstract

Soybeans are grown on approximately 1.3 million ha in Arkansas generating an estimated $1.7 billion annually. Foliar diseases on soybean can result in economic losses. Growers spend significant time and money on disease scouting via crop consultants and often (subsequent) fungicide applications. Fungicide trials are often arranged in small plots designs. In these scenarios, spatial variability of foliar disease is minimized. While it is advantageous to minimize variance outside of treatment differences, another limitation with many small plot trials is ample disease pressure or having only lower severity. Within a commercial production field, soil types and disease severities vary. Logically, by designing trials that take advantage of sub-field variability, efficacy of foliar fungicides could be determined in multiple zones of disease severity and factors that contribute to disease incidence, severity, or product efficacy could be determined. This work sought to understand foliar diseases distributions and how fungicide product evaluation might be improved. Because of the size of these trials, it was hypothesized that aerial imagery might be useful to determine sub-field variability of plant disease or other factors that influence disease. In 2017-18, strip trials were established in nine soybean fields throughout Southeast Arkansas. Fungicides were applied between full bloom and beginning pod. Fungicide strips were georeferenced with points spread approximately equidistant throughout the length of the field. Foliar diseases were identified, and disease levels determined across the test areas. Disease distributions were mostly significantly clustered and product efficacy changed as disease severity changed.

Aerial imagery was captured on wheat, barley, and canola trials using a sUAS with visual (RGB) and near infrared sensors. Images of all test crops were captured at three different altitudes, and bloom percentage on canola and ground coverage for barley and wheat trials were assessed. Plot images were human rated and assessed using disease quantification software and plots were rated by field observations. Human rated and software quantifications of images were similar confirming plot assessment by sUAS is possible for some applications and could be useful in larger research trials such as the commercial field strip trials used in this work.

Download

Share

COinS