Corn is planted on about 90 million acres in the United States each year. With all this data, government agencies need months after harvest to analyze total yield and grain quality. Scientists are working to reduce these deadlines by making yield forecasts at the end of the season by the middle of the season. However, fewer researchers took to the predictions of grain quality, especially on a large scale. A new study by the University of Illinois is beginning to fill this gap.
Study published in agronomyuses a newly developed algorithm to predict yield at the end of the season and grain composition – the proportion of starch, oil and protein in the kernel – by analyzing weather conditions at three important stages of maize development. It is important to note that the predictions apply to the entire midwest maize culture in the United States, regardless of maize genotypes or production methods.
“There are several studies assessing factors affecting quality for certain genotypes or specific places, but before this study we could not make general predictions on this scale,” says Carrie Butts-Wilmsmeyer, associate professor at the Department of Plant Science at U I and co-author research.
As corn enters the grain elevators in the Midwest each season, the US Grain Council takes samples to assess the composition and quality of its annual summary reports, which are used for export sales. It is this comprehensive database of Butts-Wilmsmeyer and its colleagues that were used in the development of their new algorithm.
“We used data from 2011 to 2017, which covered years of drought, as well as record years and everything in between,” says Julianne Seebauer, Chief Research Specialist in the U Department of Plant Science I and co-author of the study.
Researchers have combined grain quality data with weather data for 2011-2017 from regions that are fed to each elevator. To build their algorithm, they concentrated on the weather for three critical periods – sprouting, peeling and filling the grain – and found that the most powerful predictor of both grain yield and composition quality was water availability during the peeling and grain filling.
The analysis has deepened, identifying conditions leading to an increase in the concentration of oil or protein – information that is important for grain buyers.
The proportion of starch, oil and protein in the corn grain is influenced by the genotype, the presence of nutrients in the soil and the weather. But the effect of the weather is not always simple when it comes to protein. In drought conditions, stressful plants lay less grain in the grain. Therefore, there is proportionally more protein in the grain than in plants that are not experiencing drought stress. Good weather can also lead to increased protein concentrations. A lot of water means that more nitrogen is transported into the plant and incorporated into proteins.
In the analysis, “the level of protein and oil in grains above average favored less leaching of nitrogen during early vegetative growth, but also higher temperatures during flowering, while higher concentrations of oil than protein resulted from lower temperatures during flowering and filling the grain, ”the authors say in the office.
The ability to better predict protein and oil concentrations in grains can affect global markets, given the growing domestic and international demand for high protein corn for use in animal feed. With the new algorithm, it is theoretically possible to make forecasts of yield and quality at the end of the season a few weeks or months before harvest, just by looking at the weather conditions.
“Other researchers have achieved real-time yield forecasts using much more complex data and models. Our approach was relatively simple, but we managed to add a quality sample and achieve decent accuracy, ”says Butts-Wilmsmeyer. "Weather variables that we consider important in this study can be used in more complex analyzes to achieve even greater accuracy in predicting future yields and quality."