Texas A&M AgriLife researchers discovered they can predict corn southern rust epidemic outbreaks by utilizing unmanned aerial systems, UAS, or drones, early enough to help prevent economic damage for growers.
Outlining the work, the paper, “Phenomic Data-Facilitated Rust and Senescence Prediction in Maize Using Machine Learning Algorithms,” was selected for publication by Scientific Reportsrecently. The lead author was Aaron DeSalvio, a Department of Soil and Crop Sciences graduate student in the Genetics and Genomics program at Texas A&M University.
Southern rust is the most important foliar disease of corn in the Upper Coast region of Texas, Isakeit said. In some wetter years, when susceptible hybrids are grown, it may require a fungicide treatment to minimize yield loss.
He said southern rust is seen annually somewhere in the state, but yield-affecting epidemics occur where there is frequent rain. It will likely be seen first in the Lower Rio Grande Valley because the corn is planted there first, but it will also affect the Coastal Bend, Upper Coast, Blacklands and the High Plains, which is the last area to see it.
“Severe epidemics of this disease do not occur annually,” Isakeit said. “The sporadic occurrence of the disease makes it difficult to get good data on hybrid susceptibility from variety trials.”
Isakeit said early identification allows for informed decision making and fungicide application to prevent damages. The fungicide should be applied when there is a low severity of disease on the mid- to upper-canopy.
Murray said scientifically, this new capability to predict a corn southern rust outbreak earlier is very exciting because it can help monitor crops at critical stages where southern rust can cause the most economic damage for growers.
Through the data analysis, DeSalvio said, they revealed a positive correlation between the presence of southern rust at grain-filling time and yield, which will have practical implications for precision agricultural practices.
“As a breeder, I can never seem to find time to take notes on rust,” Murray said. “Now, with UAS/drone tools, I don’t have to, and these tools are more accurate. But most exciting to me, we can approximate the time of grain fill, which is highly correlated with grain-yield prediction. This could not practically be done before.”
Southern rust produces orange, powdery leaf spots. Severe infection results in leaf death and premature plant senescence, which prevents full grain development, DeSalvio said.
“Spores of this pathogen will be blown into the southern U.S. from the tropical areas of North America and, in wetter years, can become a problem here in Texas,” he said. “Last year happened to be one of those years. We found some visual markers scored by drones that allowed us to predict the disease before it could be seen by the eye.”
Southern rust is first seen on the lower leaves, and as the season progresses, the pathogen works its way up the plant, Isakeit said. As most of the yield is determined by middle to upper leaves, a little bit of rust on lower leaves is not a problem.
“When to take action is determined by the amount of rust and the growth stage of the plant,” he said. “Depending upon how risk-averse the grower is, the future weather outlook is also a factor. Aaron’s system can allow for a better assessment of the amount of disease.”
This study used UAS field-based high-throughput phenotyping to collect high-resolution aerial imagery of elite corn hybrids planted in the 2020 and 2021 growing seasons, with 13 UAS flights obtained from 2020 and 17 from 2021.
The team was able to extract vegetation indicators, such as color appearances, from UAS images throughout growth that were useful as predictors for southern rust scored in the field, as well as senescence as scored using UAS-acquired mosaic images.
UAS-acquired data enabled the discovery of early indicators that also allow researchers to quantitatively estimate the amount of southern rust and corn senescence before it could be detected by an expert manually scouting the field, DeSalvio said.
“The scale and resolution you can obtain with drone imagery also allow you to cover more land, more crops at a higher resolution than would be possible if you’re scoring plant death or plant disease progression by eye,” he said.
Rare opportunity as a student
DeSalvio said the discovery and publishing of research so early in his educational process is exciting.
“I was just a rotation student and a summer worker before starting on my doctorate program,” he said. “So, to get this kind of a head start, I’m definitely grateful to Dr. Murray and everyone who helped me out along the way.”
DeSalvio said when he joined the lab last year, the summer moisture levels were extremely high, contributing to the rare pathogen rise. Murray’s team was already collecting drone data on their corn experiments as part of U.S. Department of Agriculture-National Institute of Food and Agriculture and Texas Corn Producers Board projects, sometimes flying as many as 25 different times throughout the crop’s growth.
While taking measurements with drones throughout various growth stages of the corn crop, the team began to take notice. The historical images and data made it possible to look back to how plants looked earlier in the season and in past years. The comparable images and data led to their paper for Scientific Reports.
“It’s definitely an honor to say I’ve worked with such high-level researchers such as Dr. Adak, who was critical for teaching me how to do all the statistical analysis; with Dr. Isakeit for his guidance with analyzing the effects of southern rust; and with Dr. Wilde, who flew the drones for this particular data set,” DeSalvio said.
Today’s discovery leads to tomorrow’s solutions
DeSalvio said a lot of the background he discusses in his paper was inspiration from researchers who looked at rust in wheat and other corn blights and pathogens that people have tracked with drones before.
“So, we definitely took inspiration from other researchers and believe our discovery will also help researchers working on other crops,” he said.
“I think, ultimately, drones will allow producers, geneticists and research programs to make more informed decisions about how to manage their lands, how to manage their fields, which lines are more resilient and which lines are more temperately adapted to the regions that they’re grown in.”
Would you like more information from Texas A&M AgriLife?