Integrated Crop Management

Fungicides: Plant health fungicide applications

Seventh in a series. See Part 6 [1]

If you take a look at the current distribution of soybean rust in the United States and listen carefully to the experts on the chances of rust making it to Iowa, you have to be encouraged. Despite the good news about soybean rust not spreading quickly (or hardly at all), there have been several reports of chemical reps from major fungicide manufacturers trying to convince growers to purchase fungicides and apply them to soybean to enhance plant health, leading to higher crop yields; suggested treatments involve QoI-containing fungicides, such as Headline® or Quadris®.

During the next 30 days, many soybean acres in Iowa will move from flowering to the early pod fill, which is the targeted period for "plant health" applications of fungicides. In the absence of soybean rust and significant risk of infection by the rust pathogen, the decision to apply a "plant health" fungicide application is not as straightforward as some might believe. There is not general agreement among university and industry scientists regarding the potential for "plant health" applications to result in economical benefits to producers--economical means that dollars return per acre exceeds dollars invested in fungicide/application costs.

Arguments for not spraying fungicides for "plant health"

In 2005, fungicide studies were set up across the Midwest and northeastern United States. In all, 62 plots with different fungicide/timing combinations were completed. Significant yield results were encountered in 27 percent of these 62 replicated experiments. These experiments involved both large and small plot tests. In a separate fungicide study at the University of Kentucky during 2003-2005, plant pathologist Don Hershman looked specifically at "plant health" applications and if they led to economic benefits to producers. Over the three years of study, significant yield results were encountered in six of 24 experiments.

All of the experiments referenced so far involved replication of treatments, which allowed treatments to be statistically analyzed. The only way to know with any degree of certainty if one treatment produced a different result than another treatment is to analyze results with accepted statistical procedures. If no statistics are done, anything less results in uncertainty whether or not the treatment really had an impact or if apparent differences were merely the result of chance and/or inherent experiment variability.

Arguments for spraying fungicides for "plant health"

In contrast to the above, results of "plant health" fungicide applications based on a large number of strip plot, side-by-side comparisons in grower fields have been much more favorable to potential yield benefits. For example, 162 side-by-side comparisons, summarized by BASF from Indiana and Ohio in 2005, showed an average yield response to "plant health" fungicide applications of 5.6 bu/acre. Comparisons summarized from 192 fields across the Midsouth in 2005 were even more favorable, with an average yield response of 8.5 bu/acre. Supposedly, Syngenta can produce similar results from their on-farm trials over the last few years. Keep in mind that a 3 to 4 bu/acre yield increase is considered to be the economically break-even point, depending on the fungicide and application cost and the price received per bushel of soybean.

There are problems with nonreplicated, side-by-side, strip-plot comparisons. One major concern is that results typically do not include detailed information on disease activity. Thus, one cannot know if a disease was the reason for any potential yield increase. In addition, treatments in a field cannot be statistically analyzed because there is no in-field treatment replication. Still, large numbers of strip trials with similar results have meaning and value, despite serious questions about the data generated.

How to interpret these different results?

What is one to make of the apparent discrepancy between the results of 86 replicated experiments, conducted by numerous scientists, and the results of nonreplicated, but numerous, on-farm treatment comparisons conducted by industry?

The great disparity in results should be a "red flag" when considering making applications of foliar fungicides to soybean in the absence of specific fungal target diseases. Very high treatment yields in many strip-plot comparisons (and some replicated experiments) over the last few years make it clear that something significant is going on. There have been enough positive results to conclude that "plant health" applications of fungicides to soybean cannot be discarded. Yet, there also have been enough negative results to know that there are many, many scenarios where "plant health" fungicide applications simply will not pay for themselves.

Side effects

Applications of any pesticide in the absence of a specific target pest(s) is contrary to integrated pest management principles. There may be unintended, long-term implications (environmental, regulatory, and/or sociological) associated with applying fungicides to soybean in the absence of specific target pests.

One consequence could be the suppression of native fungi that naturally keep insect populations in check. There is already some evidence that indiscriminate use of fungicides in soybean has resulted in increased insect/mite activity in some states. Another consequence could be stricter future pesticide laws and regulations.


Producers who are considering making fungicide applications for "plant health" benefits should make those applications with the awareness that a significant economic response is not assured. At this point, university and industry scientists do not have enough information to be able to recommend which fields/situations have a high probability of an economic response and which fields have a low probability of response.

Much of this article was excerpted from an article in Kentucky Pest News written by Don Hershman.

On to Part 8. [2]

This article originally appeared on pages 189-190 of the IC-496(17) -- June 26, 2006 issue.

Source URL: