PMIL offers webinars on Aflatoxin sampling protocols

By Christy Fricks
University of Georgia, Peanut & Mycotoxin Innovation Lab

Not long after aflatoxin was discovered and methods were developed for analyzing commodities for contamination levels, researchers and the industry quickly realized there was a problem.

“I would get a call from the industry explaining that they had tested the same lot three times and gotten 0, 0 and 100 parts per billion (ppb). What’s wrong with our test?” said Dr. Tom Whitaker as an introduction to the webinar on sampling variability of mycotoxins. Much of Dr. Whitaker’s 42-year career with the USDA has been aimed at resolving this issue, along with his lab mate Andrew Slate.

The extreme variance in aflatoxin testing is caused by the irregular distribution of the contaminated individual grains and the potential that those individual grains may contain highly concentrated amounts of the toxin. For example, lab results have shown that an individual peanut kernel may contain 1 million ppb, while the commercial acceptance threshold in the US is only 20ppb. Whitaker and Slate have spent years analyzing the statistical probability of that individual kernel entering into the sampling and developed sampling and testing protocols for the USDA and Codex Alimentarius that attempt to strike a balance between the risk of over or under detection.

The Peanut & Mycotoxin Innovation Lab, (PMIL) has several research projects evaluating aspects of production and post-harvest handling for their impact on controlling aflatoxin in peanuts. They have also been approached by a number of projects screening commodities for aflatoxin or developing low-aflatoxin products for certain markets and inquiring about how to test for aflatoxin.

“Our initial focus was on evaluating analytical methods, but we have quickly realized that if we don’t focus on this sampling issue, we will be generating a lot of data and will have little confidence in its accuracy. That will generate a lot of doubt in our ability to confidently make recommendations based on that data” said Dave Hoisington, Director of PMIL.

Whitaker’s research divides the testing process into three steps: sampling, sample preparation and analysis. Their results show that, depending on the sampling protocol, more than 80% of the variance is accounted for in the sampling and less than 5% is related to the actually analytical method. While testing is always a balance between cost and accuracy, this evidence suggests greater focus on sampling and sample preparation, rather than analytical methods.

On December 10, 2014, PMIL hosted an initial webinar on this topic at North Carolina State University with Whitaker and Slate presenting the fundamentals of their research on variability, sampling protocols and an online tool developed from their work that evaluates the impact of different variables on the confidence of the sampling protocol.

The majority of this research has been targeted at larger scale trade, however the new collaboration between PMIL and the USDA has already initiated research on applicability to smallholder lots and research trials. Further webinars and workshops are being developed to address these questions.

Jamie Rhoads, Assistant Director of PMIL added, “For anyone interested in testing for aflatoxins, this is essential knowledge and will help you design your sampling plan and understand your results much better.”

Watch the webinars.

Published January 6, 2015