Anyone can send in a hay sample to a lab, and the lab is where the real science happens. Right?
Not quite. According to the National Forage Testing Association, “hay sampling produces more variation in results than does lab error.” They strongly advise against sending in hand-grab samples, and there are multiple pages of literature on proper sampling protocols. What makes these variations difficult to measure is that most folks who test hay are motivated to get certain results. Farmers want their hay to test well so that they can get more money for it. Hay buyers who bought a load of untested hay may want it to test lower so that they can negotiate a lower price on it. Hay brokers want hay to test in a range their customers are looking for to make marketing it easier. Almost everyone sending their samples to labs has some stake in how the hay tests, which makes strict following of all protocols less likely. It may even be innocent, like a farmer testing a stack and unconsciously gravitating towards bales that look greener or more leafy. It may also be as overt as putting handfuls of leaves into the sample bag or, conversely, grabbing a wad of stems from the outside of a bale.
A farmer I work with recently told me, “I think that, by picking and choosing which bales in a stack you test, you can make a test come back about as good or bad as you want it to.” We recently saw a real-world example of this, where samples pulled by two different people from the same stack went to the same lab and tested about 40 RFV points apart. In terms of nutrition and expected selling price for the farmer, this difference is huge and, depending on the amount of hay, could mean a tens of thousands of dollars discrepancy. In a year like this one, a $20,000 pay cut or raise could make or break a farmer’s, horse owner’s, or rancher’s operation. With that in mind, we must acknowledge the likelihood of someone involved in such a transaction producing a biased sample. That’s where Square Deal comes in: I have absolutely no stake in how the hay tests and get paid the same amount regardless of results.
The National Forage Testing Association’s guidelines suggest a minimum of 20 core samples per lot of hay, which should negate any variation in single bales tested–one weedy bale cored amongst 19 other bales isn’t going to skew the results. They also give a list of approved labs (https://www.foragetesting.org/_files/ugd/24f64f_229799ee5887422caf3727c40d10190c.pdf) that will grind the entire sample to divide it uniformly even though only a small amount is tested. However, cherry picking bales or sampling from fewer than 20 will alter the analysis. I can’t fault anyone for trying to make or save an extra few dollars given the drought and market conditions most farmers and livestock owners have faced over the last few years, but the demand for a service like Square Deal makes me wonder if measurable bias in hay analysis due to sampling has become more mainstream on both sides. Or maybe it’s not happening more often, but we are collectively more aware of it.