Identifying weed-tolerant corn and soybean varieties
Good weed management is important, but the variety you choose can
have a dramatic difference on how your crops fare if the season
doesn't go as planned

By Rita Seidel and Paul Hepperly
Posted September 1, 2005

Sometimes it’s just a matter of fighting preconceived notions. In two years of trials, soybean variety Iowa 3006 has shown a remarkable ability to retain high yields in the face of strong weed pressure.

Being German, I can understand the desire of our local Pennsylvania Dutch farmers to have clean, weed-free fields. After all, what will the neighbors think if they see ragweed, lambsquarters, pigweed or foxtail all over your farm?

	"What will the neighbors think if they see ragweed, lambsquarters, pigweed or foxtail all over your farm?"

Nevertheless, several questions remain:

1. Could we get even higher yields in the organic systems if we had fewer weeds?
2. Can we identify crop varieties that are particularly good at tolerating higher weed levels?

In this article we will concentrate on the second question.

Over the past few seasons, we have been conducting corn and soybean variety trials with the goal of identifying those varieties that will work well for organic farmers—i.e., those that will produce competitive yields and high quality grain in the presence of prolific weed populations.

We are now in our third year of comparing readily available, “off-the-shelf” commercial varieties of food-grade corn and soybeans classified for high yields and adaptability for large-scale production. Among those varieties, we have identified differential responses to varying weed levels across crops, fields, and years.

Both corn and soybean varieties differ widely in their ability to tolerate weed pressure.

To evaluate weed competition in each variety, we designate two areas in each plot: one that is kept entirely weed-free, and one that is treated according to our usual management of 1-2 rotary hoeings and 2-3 cultivations. While the mechanical management gets rid of most of the weeds, creating a weed-free area usually requires additional weeding by hand, especially to eliminate weeds growing within the crop rows.

At the end of the year, we compare yields from both areas and decide which varieties are worth testing further and which should be eliminated from the trial.

Findings so far

Our standard yield targets are 130 bushels per acre for corn and 40 bu/ac for soybeans. Less than optimum weather in 2003 resulted in corn and soybean yields slightly below those levels for all but two soybean varieties, Iowa 3006 and NC+ 3F43, both of which achieved 46 bu/ac.

In 2004, by contrast, the weather was favorable and both corn and soybeans did very well--soybean yields were around 46 bu/ac across all varieties, while corn yields averaged nearly 220 bu/ac. Differences among varieties in yield impact caused by weed pressure were much more pronounced in the poor-weather year than in the favorable year.

Among corn and soybean varieties, yield differences between hand-weeded and non-hand-weeded areas were highly variable, especially in 2003, as the following two graphs illustrate.

Figure 1. Corn yield increase in weeded plots in 2003 & 2004

click here for larger image Figure 1. Our goal is to identify high-yielding, food-grade corn varieties that experience less than a 10 percent yield loss under weed pressure. NC+68F32 had yield losses below that limit for two years in a row. (LG=Limagrain, GH=Great Harvest)

For corn, yield differences ranged from 0 to 46 bu/ac, meaning that for some varieties weeds had no effect on yields, while in others they reduced yields by over 60 percent. Soybean yield differences were between 0 and 18 bu/ac, translating into weed-induced yield reductions of up to 54 percent. Corn variety NC+ 68F32 and soybean variety Iowa 3006 showed the lowest yield differences (under 10 percent) between weeded and un-weeded areas for two years in a row.

Figure 2. Soybean yield increase in weeded plots in 2003 & 2004

click here for larger image Figure 2. Iowa 3006 had less than a 10 percent yield loss due to weeds in both 2003 and 2004. Needless to say, we now grow Iowa 3006 on most of our production acreage. Simple, on-farm variety trials are a good long-term investment, but bear in mind that results may vary on different soils or under different weather conditions. (GH=Great Harvest)

Weed tolerant vs. weed suppressant?

We consider varieties whose yields are minimally affected by weeds to be weed tolerant. They should not be described as weed resistant, since they were subject to weed populations just as high as less-tolerant varieties.

In addition to testing for weed tolerance, we've wondered if we could identify varieties that actually suppress weeds, for instance by strong early growth and rapid canopy closure. But although this seems theoretically possible, we have yet to identify any varieties that exhibit this effect--each year, weed biomass in the standard mechanical management subplots has been essentially the same across all varieties.

Iowa 3006 had less than a 10 percent yield loss due to weeds in both 2003 and 2004. Needless to say, we now grow Iowa 3006 on most of our production acreage.

We did, however, see differences between corn and soybeans in terms of weeds and their effect on yields. In corn, weed biomass averaged about 2,000 pounds per acre dry weight in both 2003 and 2004, but yields fluctuated by variety independently of weed pressure in a particular subplot. For example, NC+ 112E1 yielded 75 bu/ac at 2,200 lbs/ac of weeds, whereas NC+ 68F32 yielded 119 bu/ac at 2,500 lbs/ac of weeds.

Soybeans, by contrast, were generally more sensitive to weed pressure than corn. For soybeans, the higher the weed biomass, the lower the bean yields. In 2003, average weed biomass was around 1,500 lbs/ac and weed-induced yield losses averaged 35 percent across all varieties. In 2004, weed biomass was less than 300 lbs/ac and yield losses were near zero for all varieties. (Although it may sound like a lot, 300 lbs/ac of weeds is a very small amount, comparable to what you can expect with herbicide applications.) Low weed levels in the 2004 soybean plots were most likely due to preexisting low weed pressure in that particular field.

Data from our Farming Systems Trial also support the idea that economic weed thresholds are lower for soybeans than for corn. For both crops, however, significant differences in weed sensitivity among varieties are apparent.

Minimal yield losses in 2004 bring us back to the question of the economic weed threshold, the weed density at which a control practice is economically justified. As Matt Ryan wrote in a previous article, our results suggest that in some cases, weed thresholds might be higher than expected. Since weed control can be costly and time consuming, management tactics should be applied with weed thresholds in mind.

Economic analysis

So what does all this mean for your bottom line? Figures 3 and 4 show the differences between varieties and years in terms of net returns for conventional, organic feed-grade and organic food-grade corn and soybeans.

Figure 3. Net return for corn varieties under
conventional and organic production

Figure 3. Organic management and the “right” variety can lead to a significant increase in the farmer’s bottom line. While conventional corn production hovers at break-even levels, healthy returns are possible with organic production, even with substantial weed pressure. Returns are even higher with food-grade organic corn. In 2003, NC+68F32, identified as weed tolerant, produced returns twice as high as NC+112E1. Differences were smaller in 2004 because of overall high yields. Note: Y-axis net return ($/acre)

As you can see, organic food-grade and even feed-grade crops are many times more profitable on a net basis than conventional crops of the same varieties. The organic price premium simply dwarfs the reduction in yield sometimes (but not always) seen in organic systems.

Figure 4. Net return for soybean varieties under
conventional and organic production

Figure 4. Even with soybeans, with their higher per-bushel prices, economic returns of organic production with weeds present are dramatically superior to conventional, weed-free production. Note the 40 percent increase in returns of Iowa 3006 over HP 204 in 2003. Again, differences were smaller in 2004 due to overall lower weed pressure. Note: Y-axis net return ($/acre)

But the two graphs also indicate that within organic systems, your choice of corn and soybean varieties can have a big impact on year-end returns. In 2003, for instance, soybean variety Iowa 3006 gave a net return of $705/acre--40 percent higher than variety HP 204, which returned just $509/acre. Multiply that difference across 100 acres or so and you're talking real money.

Our results also suggest that while returns are higher in favorable years, the choice of variety is more critical in unfavorable years.

To conclude, the following two tables summarize the data presented in the previous four graphs, while also presenting the actual yield figures in the presence and absence of weeds, as well as the weed biomass totals.The first two columns are averages across all varieties; the remaining columns show results for two specific varieties for each year.

Economic summary for corn varieties

Keep in mind that although the weed biomass totals appear in some cases to vary a great deal between varieties (841 lbs/ac in Iowa 3006 in 2003, vs. 1635 lbs/ac in HP 204) these differences were not statistically significant across all replications. In other words, we have yet to see any consistent weed-suppressant effect by any corn or soybean variety.

In terms of yields, you can see that for the most part, plots kept entirely free of weeds did yield higher than plots in which weed management was limited to standard mechanical cultivation. In some cases the yield difference was slight; in others it was dramatic--and in one case (the soybean variety Iowa 3006 in 2004) the yield was actually higher in the weedy plot!

Looking ahead

As we move forward with these variety trials, our main criterion for choosing one variety over another will continue to be weed tolerance. Varieties that produce yields close to or above our yield targets while showing less than a 10 percent yield loss because of weeds are re-tested the following year. Poor performers are dropped from the experiment after a single year.

For soybeans we also evaluate bean quality, since split or stained beans reduce food-grade prices and therefore overall returns. It's important to keep in mind, however, that varieties with lower bean quality can still be highly competitive on the organic feed market.

For the 2005 trials we planted the following varieties:

Corn varieties.
In the second year of testing: NC+ 68F32, GH 41X2, GH 44X2.
In the first year of testing: NC+ 72H54, NC+ 4771, NC+ 60N37, LG 2603W, LG 2609W, GH 61K9, GH 61F3.

Soybean varieties.
In the second year of testing: Iowa 3006, NC+ 2F11, and NC+ 2FN93.
In the first year of testing: NC+ 2F45, NC+ 1F44, Iowa 2054, Iowa 2067, Iowa 2068, GH 292N, GH 327, Viking 2022, Latham 280, Wilken 3494, Wilken 3476N. (The Wilken varieties are feed varieties serving as a control.)

D.I.Y. Results of the RI trials suggest the utility of farmers conducting simple variety trials on their own farms, as Jeff Moyer described in his April 2005 column.

This winter we'll analyze the results of these trials and draw up a new list of varieties to test in 2006.

This work is important because it represents a new approach to variety evaluation. Conventional agricultural research has tended to assume near-complete weed control through the use of herbicides, so the ability of crop varieties to tolerate weeds has received little attention. While much work remains to be done, our results show that a tremendous opportunity exists to identify
and develop varieties specifically suited to organic growing
conditions.

This material is based upon work supported by the U.S. Department of Agriculture, under Agreement No. 58-1265-4-016.

Any opinions, findings, conclusion, or recommendations expressed in this publication are those of the author(s)
and do not necessarily reflect the view of the
U.S. Department of Agriculture.