New tests for starch damage in wheat are on the horizon as researchers seek a more reliable alternative to the falling number test that is commonly used.

Those efforts got a shot in the arm last summer, when Ashley Cannon joined the USDA Agricultural Research Service in Pullman, Wash., as a research molecular biologist. Cannon studied secondary dormancy as a post-doctoral student, giving her insight into seed development and germination.

She joined USDA research plant molecular geneticist Camille Steber, who with researchers Amber Hauvermale, Andrew McCubbin and Michael Pumphrey designed a new test called an Enzyme-Linked Immunosorbent Assay, known by the acronym ELISA.

It can measure the amount of alpha-amylase in flour samples. Late maturity alpha-amylase is an enzyme that causes starch degradation in wheat.

They have teamed up with an industry partner to increase the sensitivity and speed of ELISA.

In addition, Washington State University researcher Zhiwu Zhang and former graduate student James Chen have been developing another test that combines hyperspectral imaging and machine learning.

The commonly used falling number test measures starch damage in wheat. Technicians determine how long it takes two pins to fall through a ground wheat-water slurry, measuring its viscosity. Wheat with a low falling number has starch damage and is sold at a discount because it can reduce the quality of baked goods and noodles.

Farmers were caught off guard in 2016 when 44% of soft white wheat samples and 42% of club wheat samples tested below 300, the industry standard. The industry estimated the damage that year cost farmers more than $30 million in lower wheat prices.

Low falling numbers can be caused by pre-harvest sprouting due to rain or by late maturity alpha-amylase occurring as a result of large temperature fluctuations during late grain filling.

Cannon and Steber spoke with the Capital Press March 18 via Zoom.

The interview has been edited for length and clarity.

Capital Press: What does having Ashley on board add to research efforts? What can you do now that you couldn’t do before?

Steber: When the funding first came in for falling numbers through USDA, I was asked what I needed help with. I said, “There is a lot of pressure to improve upon the falling numbers test, and I’m being stretched in the direction of biochemistry,” things I hadn’t done before.

So I’m really glad to have her expertise. When we did that job search, I was looking for a partner in crime and I think I’ve found her.

CP: Is it hard to get a handle on what falling number is?

Cannon: The test itself is actually pretty straightforward. I think we’re still learning about how those test results can actually translate to what we see in the grains themselves. …

I think the part that keeps coming up, though, is there’s so much variability in the test. That may have nothing to do with science. That’s the part that I think none of us can really get a grasp on, unfortunately.

CP: If it may have nothing to do with science, then what’s the problem?

Cannon: I think some of it has to do with users. It’s not really user error, at all. It’s just the way each of us approaches the technique. Each person who does a falling numbers test introduces some level of difference that could result in an error.

The things that are contributing to the gelling capacity are just a lot more complex than people like to let on. It’s not just, “Was there alpha-amylase or was there not?” There’s a lot more going on chemically that could lead to a lot of differences in the gelling capacity of that mixture in the test.

CP: How can a new test address what you’re talking about?

Cannon: The falling numbers test, unfortunately, although it works well to give us an answer right now, isn’t always a true reflection of end-use quality.

Steber: It’s used by the industry as a risk assessment tool.

Cannon: The research is really there to show that sprouting grain unfortunately has bad end-use quality, generally. If we use the ELISA … our confidence in the predicted end-use quality goes up. I think that’s the advantage of that test.

The other test, hyperspectral imaging … can tell us, hopefully long-term, what a predicted falling number would be, but also perhaps what is the cause of a predicted, relatively low falling number. Is it sprouting or late-maturity alpha amylase and based on that, what is the actual predicted end-use quality? Collectively, it increases our confidence that we actually know whether this grain is going to result in good or bad end-use quality, and it reduces unnecessary losses to farmers.

CP: One of the complaints we hear about the current test is that you can’t replicate it because the sample is destroyed in conducting the test. Would these other tests address that?

Cannon: Hyperspectral imaging is not destructive, so that’s really exciting. We can test the same grain more than once. That in itself increases confidence. If a farmer says, “I don’t agree,” just test the grain again.

The ELISA, unfortunately, will require destruction of the sample.

Steber: We have a struggle in that critical range from 250 to 300 seconds, there is a great deal of grain-to-grain variation.

If I say, “Here, I have this bag of wheat, and it’s 270,” if I were to go through and take out little aliquots of 10 grains each, and check the enzyme levels, I could get wildly different numbers. …

That variation is going to exist whether I’m running a falling numbers test, an immuno assay (ELISA) or a hyperspectral assay.

With the hyperspectral assay, Ashley is actually going to take things to the point where she can measure alpha-amylase on single grains.

That means we’ll literally get a picture of what’s going on, assuming that everything goes as we’ve planned.

You’ll see what the cumulative falling number alpha-amylase level is likely to be, and you’ll also get an idea of the degree of variation in that sample.

CP: Would both tests replace the existing test?

Steber: An immuno assay (ELISA) at this point is a sure-fire approach. It means we will get it to work. It’s scientifically clear we will get it to work.

The hyperspectral imaging, we are still in the exploratory stages, so I hope we will get it to work. Keep your fingers crossed for us, please.

CP: What kind of a timeline is there to move from the current test, with its issues, to something that’s more solid?

Steber: Well, there are two things that have to happen, right? The science and adoption within the industry.

We actually just wrote a pre-proposal to (the Foundation for Food and Agriculture Research) for a project that includes not just the science piece, but also the outreach piece for adoption. We think that’s going to be critical.

It’s likely that adoption is going to be gradual once we have the tool in place.

It’s going to be a long time before our foreign buyers think that they will trust a new test. It’s a matter of coming to trust it. But I think the role it can have early on is in identifying problems earlier, before we start mixing low- and high-falling number grain.

I think it will have a role at the elevators and you never know, it may even have a role for farmers when they harvest different fields, so they know that they need to be careful not to mix grain from one spot with grain from another spot, because they can see that they have wildly different falling numbers.

Cannon: Another way to think of it that I think farmers will be really happy to hear is I think breeding programs will be able to use it. It’s really going to prevent the release or even breeders pushing varieties through that have low falling number, or that have a particular cause of low falling number. Farmers can rest easy thinking that in coming years breeders are going to be releasing varieties that have more stable falling numbers.

Steber: This is a job for Ashley and I, to develop these screening techniques for the breeding programs and teach them. Because the falling numbers test requires expensive equipment, takes up a lot of space in your lab and it’s very low throughput.

Those things don’t add up to early selection against the problem in a breeding program. They add up to “OK, let’s see if we can catch it just before it goes out the door.”

By that point, a great deal of time and energy has been invested in those varieties. We need to be able to select against this earlier in the breeding cycle.

CP: So in 20 years are we going to get rid of the falling number problem?

Steber: If I were working on drought tolerance, would you ask me if in 20 years we’re going to have completely solved that problem? (laughing)

We’re going to make progress in having greater tolerance to the problem. But seeds germinate. They’re designed to germinate. If I make you a seed that couldn’t germinate, it wouldn’t be very useful in generating next year’s crop.

CP: Outlook for this coming year?

Steber (laughing): You have asked me that question before and I’m going to give you the same: I do not model the weather. I do not predict weather. I am not going to make any predictions about this year.

Last year was a much better year. And actually I think farmers are doing a better job of managing the problem. We have information out there now on which varieties are prone to the problem and which varieties aren’t.

I’m hoping that with help from their seedsmen, they’re making less risky choices for falling numbers. Although I really understand that it’s only one trait, and they have to balance that with all of the other disease packages and yield traits they’re interested in.