Agriculture relies on plant breeding and genetics to deliver the best crop varieties and agronomic practices that will deploy results across nearly a billion hectares of cropland. We regularly see industry delivering on this objective, although optimized for their goals and limited geographies. However, many public sector efforts responsible for feeding most of the planet are not meeting these objectives or working effectively across disciplines. A key example is repeated publications on single genes, multiple gene constructs, or similar edits that claim to confer incredible yield increases. Many of these publications are flawed in how they measure and report yield. Often, they lack replication across environments, have low sample sizes, compare yield estimates in non-commercially competitive germplasm, and report on experiment-specific yields that massively underperform as compared to real-world local or global yields. Here, we detail common issues that arise and describe how such findings, when published in high-profile journals, skew global agricultural funding away from proven plant breeding methods.
To address these issues, we suggest approaches for researchers and reviewers to use when evaluating the impact of single genes on crop yield, including:
Accurate measurement and reporting of crop productivity have drastic consequences for feeding the planet. Due to the impact that many of these problematic yield studies have on setting global food policy and subsequent economic investments, there is no more important time to collaborate as a community to accurately measure, improve, and deliver on crop resilience and yield.