In the heart of Mexico’s agricultural innovation, researchers are challenging conventional wisdom about the use of vegetation indices in protected agriculture systems. Edgar Vladimir Gutiérrez-Castorena, from the Facultad de Agronomía, Universidad Autónoma de Nuevo León, has led a groundbreaking study that questions the reliability of NDVI (Normalized Difference Vegetation Index) in predicting yields for indeterminate crops like tomatoes. The findings, published in Horticulturae, could reshape how agritech companies approach crop monitoring and yield prediction in controlled environments.
The study, conducted over two agricultural cycles in 2022 and 2023, used the Green Seeker® sensor to monitor NDVI and agronomic variables in a protected agriculture system. The results were surprising: while NDVI curves showed high significance levels during all phenological stages, yield predictive models fell short, particularly in the later stages of production. “The maximum values recorded for NDVI inside the greenhouse did not correlate with the yield prediction obtained from the 18th week of harvest,” Gutiérrez-Castorena explained. “This discrepancy suggests that NDVI may not be a reliable index for predicting yield in indeterminate crops under controlled conditions.”
The implications for the agritech industry are significant. For years, NDVI has been a go-to metric for assessing crop health and predicting yields. However, this research suggests that in protected agriculture systems, where environmental conditions are tightly controlled, NDVI may not tell the whole story. “The constant optimal development in response to controlled environmental conditions, nutrient status, and water supply inside the greenhouse does not necessarily translate to sustained yield,” Gutiérrez-Castorena noted. This finding could prompt agritech companies to rethink their reliance on NDVI and explore additional sensors and variables for more accurate yield predictions.
The study also highlights the economic challenges of protected agriculture. While these systems offer controlled conditions that can optimize yields, they come with high energy costs and the need for continuous monitoring. The research underscores the importance of integrating multiple data points and sensors to create more robust predictive models. “Evaluating the models between NDVI and agronomic variables is not an index that offers certainty in predicting yield in indeterminate crops in protected agriculture production systems,” Gutiérrez-Castorena stated. This insight could drive innovation in sensor technology and data analytics, leading to more sustainable and cost-effective agricultural practices.
As the global demand for food continues to rise, the need for efficient and sustainable agricultural practices becomes ever more pressing. This research, published in Horticulturae, offers a critical perspective on the limitations of current technologies and paves the way for future developments in agritech. By challenging the status quo, Gutiérrez-Castorena and his team have opened the door to new possibilities in crop monitoring and yield prediction, potentially revolutionizing the way we approach protected agriculture.