Judging Biostimulant Field Evidence

Biostimulants occupy an awkward position in the product landscape. They are not fertilizers, because they do not supply nutrients in significant quantity. They are not pesticides, because they do not claim to control a pest. They are defined by an effect on the plant or the soil, and that definition puts the whole weight of the claim on the evidence rather than on the composition.

The consequence is that two products with almost identical compositions can make very different claims, and two products making almost identical claims can rest on very different evidence. For a grower or a distributor, the useful skill is not knowing which category a product belongs to but being able to read the trial data behind it.

Why the evidence is thinner

Fertilizer evidence is comparatively easy because the product is a nutrient with a known role and a measurable soil test. Biostimulant evidence is harder for three reasons. The active constituents are often complex mixtures whose individual components are not fully characterised. The mode of action is frequently unknown or only partly described. And the effects are usually small and strongly conditional on stress, so a trial in a good season may show nothing at all.

None of that makes the products ineffective. It does mean that a single trial, however well presented, carries very little information, and that the conditions of the trial matter as much as its result. A product that improves rooting under drought may be entirely without effect in an irrigated block, and both results can be honest.

Claims and the evidence they need

Regulatory frameworks, including the European fertilizer products regulation, tend to describe plant biostimulants through four functional claims: improving nutrient use efficiency, improving tolerance to abiotic stress, improving quality traits, and improving the availability of nutrients in the soil. Each of those implies a different trial design, and a supplier who makes several claims should be able to show evidence for each one separately.

A nutrient use efficiency claim needs a trial that measures nutrient uptake or the yield obtained per unit of nutrient applied, with the same nutrient rate in all treatments. A stress tolerance claim needs a trial conducted under the stress in question, or at least with the stress measured and reported. A quality claim needs the quality parameter measured at harvest, not inferred from a growth measurement. A soil availability claim needs soil measurements, not just plant response.

Four weaknesses in most trials

The first weakness is the absence of a true untreated control. If every plot receives a base fertilizer programme and the biostimulant is added on top, the comparison is valid only if the base programme is identical and recorded. Where the treated plots also received more nitrogen, or a foliar feed that was not applied to the control, the trial cannot separate the biostimulant from the nutrition.

The second is a single site and a single season. Biostimulant effects are conditional, so one site tells you about one combination of soil, climate and stress. Repetition across at least three sites and two seasons is the minimum for a claim that will hold up elsewhere.

The third is confounding with timing. A product applied at a convenient moment often coincides with a management operation that itself affects the crop, and the effect is attributed to the product. The fourth is selective reporting: trials that showed no effect are rarely shown, and the ones that did are presented as typical.

What good evidence looks like

Good evidence has a clear untreated control with identical base nutrition, adequate replication, randomised layout, and a stated coefficient of variation so the reader can judge whether the trial could detect the effect it reports. It runs across sites and seasons, and it reports the results that did not work as well as the ones that did.

A dose response is a particularly strong signal. If a product improves a measured parameter at one rate and improves it further at a higher rate, the pattern is much harder to attribute to chance than a single rate comparison. Where the response is flat across rates, either the lowest rate is already saturating the effect or the effect is not real, and the supplier should be able to explain which.

Mechanistic support is useful but not decisive. A demonstrated effect on root hair development, on stomatal conductance or on a microbial population helps explain why the field result occurred, and it makes the result more plausible. It does not substitute for the field result, because many mechanisms operate in a pot and not in a field.

Effect size and the economics

Most credible biostimulant effects reported in independent work are modest, in the range of a few per cent on yield or on a quality parameter, and largest where the crop is under stress. That is enough to justify a product in a high-value crop with a defined stress window, and it is often not enough in a low-margin commodity crop without one.

The arithmetic is straightforward. If a product costs USD 40 per hectare and the crop is worth USD 250 per tonne, it needs 0.16 tonnes per hectare just to break even. A three per cent gain on an eight tonne crop is 0.24 tonnes, which clears the bar; a one per cent gain does not. The trial therefore has to be able to resolve a difference of that size, which brings the discussion back to trial design: many biostimulant trials are too imprecise to detect the effect that would justify the purchase.

Reading a supplier report

Start with the control. What exactly did the untreated plots receive, and is it the same as the treated plots minus the product? Then look for the number of sites and seasons, the replication, and the coefficient of variation. Then look at whether the claim being made is the parameter that was measured. A report that measured chlorophyll and claims yield has skipped a step.

Then look at the conditions. A trial conducted in a drought year supports a stress tolerance claim and says nothing about a normal season. A trial on a soil with high organic matter says little about a degraded sandy soil. Finally, look for the failures: a report that includes a nil result is more credible than one that does not, because it shows the supplier was willing to publish what happened.

For a grower considering a first purchase, the most efficient route is a strip trial on their own field, with the same base programme on both sides and the strips replicated and randomised. Two seasons of that, on the farm where the product will be used, is worth more than any number of third-party trials conducted elsewhere.

Common questions

Are biostimulants the same as biofertilizers? No. A biofertilizer supplies nutrients, usually through a microbial population that fixes or solubilises them. A biostimulant acts on the plant's physiology or on nutrient availability without being a nutrient source in its own right, although some products sit between the two categories.

Can a biostimulant replace part of my fertilizer? Not on the evidence available for most products. Some improve nutrient use efficiency, which may allow a modest reduction over time, but the reduction has to be tested rather than assumed.

Why do results vary so much between farms? Because the effect depends on the stress the crop experiences. A product that helps a crop through a drought does little in an irrigated field, and the same product can look impressive on one farm and irrelevant on the next without either result being wrong.

Does organic certification matter? It matters if the crop is being sold as organic, because the product must be permitted under the certifier's standard. It says nothing about efficacy.

Recommendations depend on local soil, climate and crop conditions. Follow the product label and local regulations, and consult a qualified agronomist for field decisions.