Auburn University researchers are investigating methods to predict meat spoilage based on microbial activity, which would enable producers to affix more accurate “best by” or “sell by” dates, thereby reducing food waste.

How ‘Sell By’ Dates Contribute to Food Waste

According to the United Nations (UN), approximately 10 percent of all meat is discarded by retailers or consumers due to it reaching its “sell by” or “best by” date. These dates are conservative values that are typically based on when meat will begin to lose its color (about four days past the date of packaging), rather than being an indicator of when the product would be unsafe for consumption. Meat turning from pink to brown in its packaging is a food quality issue, rather than a food safety risk.

As consumers often make decisions about when to throw away meat based on discoloration and the “sell by” or “best by” date, overly conservative dates exacerbate the problem of food waste.

Wasted food translates into wasted resources—including the emissions, water, energy, and land required for production—presenting a pressing sustainability challenge.

Predicting Meat Spoilage Based on Patterns in Microbial Activity

Led by Isabella Gafanha, M.S. during her time as a master’s student in the Auburn College of Agriculture, a team of Auburn University researchers sought to identify changes in microbial communities over the shelf life of packaged ground beef and associate those changes with meat quality degradation and spoilage indicators.

The researchers tracked the activity of microbial communities on ground beef over a 14-day span, collecting data on the meat’s objective color, lipid oxidation, microbial plate counts, and microbiome. A machine learning model was used to predict meat spoilage based on the data, and its predictions were compared against Ms. Gafanha’s findings from her experiments with ground beef samples.

Quality analysis results showed that the meat was microbiologically spoiled after six days. The microbiome shifted over the 14 days in a predictable succession pattern, with Rhodobacteraceae and Enterobacterales present within the samples during the early stages of shelf life, and Pseudomonadaceae and Carnobacteriaceae increasing to detectable levels by the end of the 14 days.

In other words, aerobic bacteria multiplied first, depleting the oxygen present in the package. Subsequently, anaerobic bacteria grew, thriving in the oxygen-depleted environment.

The machine learning models were able to accurately predict spoilage based on changes in microbial communities, demonstrating that Carnobacterium, among other organisms, are important for predicting day of spoilage.

”Trends followed spoilage to a ‘T’ … [which is] exactly what we wanted to see,” said Ms. Gafanha.

Next Steps: Using the Data to Derive Accurate ‘Sell By’ Dates

The study, published in the American Meat Science Association’s Meat and Muscle Biology, was part of a larger initiative led by Aeriel Belk, Ph.D., Assistant Professor of Animal Sciences at Auburn University. The research was funded by a $10,000 grant from the Alabama Beef Checkoff program.

Dr. Belk plans to replicate the study with additional samples and sessions, as well as look at other markers of spoilage. In the future, the research team’s model will be used as a tool to derive more accurate “sell by” dates to prevent meat from being unnecessarily discarded.

Ms. Gafanha was joined on the study by Dianna Bourassa, Ph.D., Associate Professor and Extension Specialist, and Amit Morey, Ph.D., Associate Professor, from the Auburn Department of Poultry Science; as well as Barney Wilborn, M.S., Associate Director, Agricultural Research and Extension Center, Auburn Department of Animal Sciences.