We have heard over the past several weeks many reports of "unusual" inputs that are being promoted and sold for soybean. Such products are also sold for wheat and corn and other crops, but soybean is a more likely crop for which such products can be sold because soybeans are on so many acres, new problems such as aphids show up to generate uncertainty, growth or color problems are visible, and Roundup is already being applied during the summer.
For decades, companies have sold products with claims of benefits to crops beyond what "conventional" inputs such as crop nutrients and pesticides provide. Few of these products have lasted for many years; but new ones continue to appear, and in many cases they are closely related to or identical to products that have had a "run" in the past and then faded out. These are clearly profitable products: They are often quite dilute, they may be by-products from manufacturing processes, and claims are for only beneficial effects that are rather vaguely described and so are difficult to disprove. These products, almost by definition, do not invite lawsuits over either performance or crop injury.
We (university people) are known for our opposition to the sale and use of such products, because we often seem to have a knee-jerk reaction in stating that such products are "unproven." In part this is true; such products are almost always unproven, at least from the standpoint of having had enough careful research done to show that the product can be expected to provide a solid return on investment. On the other hand, we try to keep informed about such products and would like nothing better than to find ones that actually do provide a return on investment and that we could use as a means to generate interest and profits among producers. I want Illinois crop producers to gain. But some university people who thought they had found such products in the past were badly burned when thorough research and farmer experience showed that (despite what might have been good theoretical underpinnings to support use of the product) it simply didn't worked well or consistently in the field.
Many of these products do have some theoretical basis to support their use. Plants need micronutrients, so why not put together mixes (the "multivitamin" approach) to provide whatever they might need? Soybean plants do sometimes grow tall and shade themselves or lodge--why not use a growth regulator to shorten them? Why not use a material to cut down on water loss so that plants can do better under dry conditions? Why not help "condition" heavy soils? Millions of dollars have been spent by companies trying to discover and develop such products, only to find that--once they are moved from the lab or even the greenhouse, where they work well, to the field--natural conditions mask their effects or even cause them to sometimes lower yields.
An even more common occurrence is that a product produces the desired effects sometimes but not very often, or that it produces yield increases sometimes and yield decreases sometimes, such that the expected effect is neutral or barely positive. Micronutrients that provide a positive response on a low-organic matter or low-pH soil will often provide no response on a more fertile soil with a neutral pH. This is understandable, but it's frustrating. A "reasonable" product may or may not pay off; and in some cases, it is difficult to predict where the response will be positive, such that the product can be applied only there.
If an input is relatively low cost (let's use 1% of gross crop value--$3 to $4 per acre--as an arbitrary threshold for "low cost"), then some farmers might use it as "insurance," not needing or even expecting it to pay off every time but gaining large increases in income when it's needed. Even products like these should have some pencil pushing to decide whether to use them. Take the expected percentage of time you expect the input to provide a return times the return (in crop value) when there is a response, and see whether this exceeds the average cost per acre of using the product. As an example, a product that returns $25 per acre in added income once in 5 years is an economically viable "insurance" input only if it costs less than $5 per acre, including application costs. Extra time spent to apply such inputs should be included, probably at a fairly high rate per hour, given that it will sometimes delay planting.
Inputs used as insurance reasonably should be expected to do something specific (provide disease protection, for example) and to work as advertised. What about products that are sold without an indication of what they do but that they simply "boost yields" or "improve growth" or something similar? As I mentioned previously, the companies producing these products have the real marketplace advantage in not having to provide specifics to say that the products did anything at all. These products almost never produce specific, visible effects in plants; nor do the companies producing them claim that they do, making them immune from claims that their products caused harm. Another advantage: It is virtually impossible to do definitive research to conclude that such products do not work. Indeed, it is possible that most products like this, unless they are completely inert (and some may be), do sometimes provide a response.
In general, sales of such products are supported heavily by testimonials. If university researchers have conducted careful research on such products, the fact that this happened may be part of the advertising; but results are often not specifically mentioned. It is fair to say that most neutral research on such products has failed to show a consistent response. We cannot, however, conclude from this that the product doesn't work--only that the particular set of trials we conducted failed to show a response or, at least, a consistent response.
What should the consumer of such products make of this situation? That depends largely on individual approaches to risk and to individual willingness to try new things. Many, perhaps most, producers "try" some unconventional products in their fields at least occasionally. This cautious "trying" is a good thing; but if the product is simply applied to a whole field or to a chunk of a field and then yields are compared to another field or to the other half of that field, we can't be certain that any yield change was caused by the input. The two halves of a uniformly treated field almost always yield differently. It is much more effective to make six or eight strips in the field with the new product, and to have the new product be the only thing different from "untreated" strips. Strips can alternate or be randomly assigned, but it is important to have both treated and untreated strips intermingled in the same part of a field.
There is even more power in testing the same product in a number of fields and having a number of neighbors participate, with the results then averaged over all test fields. Statistics can be run to tell us how confident we can be about the results; but even without statistics, a well-organized test can give us much better information than "selected" side-by-side comparisons that we often see in advertisements. Above all, we need to approach the use of unconventional products with an open, but healthily skeptical, mind.--Emerson Nafziger