A study by researchers at the University of Minnesota Twin Cities offers new insights into how alternating corn and soybean crops can help increase crop yield in a changing climate.
The research is published in the journal Global Change Biology.
Rising temperatures and weather extremes are threatening global food security, making it crucial to understand how sustainable practices like crop rotation can help improve agricultural yields and resilience.
The study found that the benefits to corn-soybean rotation, compared to continuous corn year after year, are extremely sensitive to climate. For example, corn benefits from a rotation of soybeans in colder areas, while soybean benefits from a rotation of corn in warmer regions. In addition, warming during the non-growing season reduces the corn benefits, while warming in the growing season helps boost the soybean benefits.
"Corn and soybeans may experience different crop rotation benefit changes in the future, which can help U.S. farmers make more informed decisions when facing climate warming," said Junxiong Zhou, Ph.D. candidate in University of Minnesota's Department of Bioproducts and Biosystems Engineering (BBE) and first author on the paper.
Despite climate challenges, the study suggests increasing crop rotation can still improve overall yields and highlights its potential as a climate adaptation strategy.
While most prior studies have focused on the current climate conditions, this new study takes into account a changing climate in the future and how crop rotation is impacted by it. To address this, the researchers used satellite data combined with what they call a "causal forest model," a scientific method that helps to understand cause-and-effect relationships in data. They estimated the benefits of crop rotation under varying climate conditions in the Midwest region of the United States.
"Millions of satellite observations and advanced machine learning models enable us to quantify the climate impacts on crop rotation benefits at the subfield level over the Midwest," said Zhenong Jin, an associate professor in BBE and senior author on the paper.
To analyze this enormous amount of data, the researchers used an advanced machine learning tool that teaches computers to learn from examples and improve at tasks without being explicitly programmed. This helped the team understand how crop rotation benefits corn and soybean yields in the U.S. Midwest.
"This study demonstrates the great potential of interpretable machine learning for estimating large-scale effects of agricultural management practices," said David Mulla, a professor and Larson Endowed Chair in soil and water resources at the U of M's College of Food, Agricultural and Natural Resource Sciences, and a senior researcher at the AI Institute for Climate-Land Interactions, Mitigation, Adaptation, Tradeoffs and Economy (AI-CLIMATE).
Click here to see more...