Table 1.9. The Relative Intensity of Herbicide Use Along the Distribution of All Soybean Fields Surveyed in 1998, No Till Systems

Source: USDA Economic Research Service Special Tabulation Number 3, based on soybean field-level sample data collected as a part of the "Agricultural Chemical Usage" survey (National Agricultural Statistics Service, 1999.)

Significant shifts have already occurred in herbicide use on fields planted to Roundup Ready soybeans since their commercial introduction in 1996. Several factors have driven the changes, most triggered in one way or another by the remarkable commercial success of this technology.

Rapid increases in the acreage planted to RR soybeans forced other herbicide manufacturers to cut their prices and look for ways to formulate their existing herbicide product lines into combination products that were compatible with RR soybeans and convenient for the farmers planting them. Today, there are more than a dozen new combination products on the market specifically marketed for RR soybean producers.

As noted above, the popularity of RR soybean systems forced other herbicide companies to lower prices, making it possible for farmers to make an additional spray or add in a new active ingredient without increasing per acre herbicide costs. The generally lower prices today have encouraged heavier reliance on herbicides. In the early 1990s in states like Iowa, many farmers were open to sustainable agriculture systems and methods, largely because of potential to lower per acre cash costs. The costs of seed plus herbicides were growing the fastest of any major category of production input (Benbrook, 2000). Up through about 1993 the acreage of row crops planted under ridge till and/or treated with banded (in the row only) applications of herbicides in conjunction with mechanical cultivation had risen steadily.

The introduction of several new soybean herbicides in the mid-1990s, and then RR soybeans in 1996, quickly refocused most farmers on largely herbicide-dependent systems. In the last few years, the percent of soybean acres managed under multitactic weed management systems with lessened reliance on herbicides has shrunk back to a fraction of the level in 1993. The falling cost per acre of herbicide-dependent systems and the simplicity of the RR system have been the major reason why.

Well before the introduction of RR soybeans, it was known that heavy reliance on any single herbicide, class of herbicides, or weed management tactic in a given field will trigger a shift in the composition of weeds commonly found (Ghersa et al., 1994). Roundup Ready soybean systems are no exception.

Recurrent applications of glyphosate in many corn-soybean production regions in the U.S. have brought about a shift in weed species (Owen, 1999; Hartzler, 1999). Waterhemp, velvetleaf, horseweed, yellow nutsedge and nightshade are more common and difficult to control, especially in RR fields. (Scientists at Iowa State University have done an excellent job tracking and explaining the factors giving rise to weed shifts. These factors include the time period over which weed seeds in the soil are able to germinate and how susceptible a weed is to glyphosate. For more information see http://www.weeds.iastate.edu/

Some weeds have developed resistance to glyphosate (Horstmeier, April 2001) and others are displaying rising tolerance (Hartzler, 1999). As a result, farmers are compensating by adding additional herbicide active ingredients into their control programs, while others are increasing the rates of Roundup applied in the hope of getting ahead of even tough to control weeds. The dramatic price reductions in recent years have accommodated increased rates without much, if any increase in per acre herbicide expenditures. (For more on resistance to herbicides, see the "International Survey of Herbicide Resistant Weeds" accessible at http://www.weedscience.org/in.asp or several items on Ag BioTech InfoNet at http://www.biotech-info.net/herbicide-tolerance.html#soy .

As a result of weed shifts and slipping efficacy of Roundup in the control of some weeds, most farmers growing RR soybeans now apply one to three additional active ingredients. An effective pre-plant burndown application is critical in no-till and conservation tillage systems to give RR soybeans a good jump on weeds. Cost-conscious farmers typically include about 0.5 pounds of 2,4-D in a pre-plant or at plant tank mix. The 2,4-D helps manage broadleaf weeds. Another product is typically applied to provide some residual grass control. Popular products include pendimethalin, imazethapry, and treflan. Table 1.10 displays just a few of the popular combinations of products used on conventional and RR soybean varieties. Among post-application programs on conventional soybeans, farmers applying Classic and Assure use only 0.08 pounds of active ingredient at a cost of $24.51 per acre.

Table 1.10 Popular Soybean Herbicide Control Programs Used on Conventional and Roundup Ready Soybean Varieties Under Conventional Tillage. 2000-2001

Notes: In "POST" planting in "PRE" systems. incorporated) s stem. systems, all herbicides are applied at or after Herbicides are worked into the soil planting. All herbicides are before planting in a "PPI" applied before (pre-plant

The cost of this very-low dose program actually compares favorably to a Roundup-based program with RR varieties when the technology fee is counted as a cost of the herbicide program. Under the best of circumstances, farmers in 2001 might get through the season with two applications of Roundup, the second at a reduced rate. This program will cost about $23.00 with the technology fee ($16.77 plus about $6.00 for the technology fee) and results in the application of 1.3 pounds of active ingredient. A typical PRE/POST program in RR soybeans would include two applications of glyphosate and a single application of pendimethalin. This program costs about $30.00 with the technology fee and results in application of about 2.3 pounds of herbicides.

While the "best case scenario" RR system requires less herbicide than the highest-rate conventional systems, it is clear that most RR soybeans will be sprayed with about 0.5 pounds more herbicide than most conventional soybeans in crop season 2001.

There will be exceptions, but the number of conventional, non-GMO acres sprayed with very low rates of herbicides will almost certainly exceed the number of RR soybean acres treated with less than 1.0 pound of herbicides.

In the last few years Monsanto, the biotechnology industry, and the U.S. Department of Agriculture have claimed repeatedly that Roundup Ready soybeans reduce herbicide use. As the data cited above shows clearly, this is certainly not the case on the majority of RR soybean acres grown in the United States, nor is it true "on average." Plus, extensive evidence shows that the effectiveness of the Roundup applied in the RR soybean system is slipping. This technology is, to a large extent, a victim of its own success.

In the first few years of commercial RR soybean use, many farmers got through the season with a single application of just one herbicide - Roundup. Between 0.75 and 1.1 pounds of glyphosate active ingredient were applied per acre, clearly not a low rate compared to sulfonylurea or imidazolinone weed management systems requiring between 0.1 and 0.3 pounds of herbicide active ingredient, but about mid-range across all systems. Four years later almost no farmer can get by with just one application of Roundup.

Farmers who applied one application of Roundup on RR beans in 1996 and 1997 are likely to be malting two or three in crop year 2001. They will also be applying at least one, and more likely two additional herbicide active ingredients. Some are applying three additional herbicides. Why? Again, the evidence is voluminous, consistent and compellingly clear. Heavy reliance on Roundup in RR soybeans has --

• Triggered significant weed species shifts, favoring those weeds that are not as sensitive to Roundup, as well as those that tend to emerge over extended periods of time, so that some weeds emerge outside the window of time when Roundup applications deliver good control.

• The emergence of resistance in some of the nation's most common, tough to control soybean weeds like waterhemp, coupled with modest to moderate slippage in efficacy in a growing number of other weeds. Slipping efficacy increases the number of escapes and then requires higher application rates to knock back the escaped weeds when a subsequent application is made.

Despite these widely recognized facts, it is still common to encounter claims by Monsanto, the biotechnology industry, the U.S. Department of Agriculture (USDA), and others that RR soybeans reduce herbicide use. How can major companies and a government agency get away with malting such claims? It takes a certain amount of care coupled with a little misinformation and a major dose of missing information.

A November 30, 1999 Monsanto document entitled "Chemical Reduction Benefits of Biotechnology Crops" was prepared for the press, political leaders, and PR purposes (Monsanto, 1999). It states that --

"In a Sparks Commodities, Inc. study conducted in 1996 and 1997, in-season herbicide use in Roundup Ready soybean fields was shown to be less than traditional soybean varieties by an average of 26 percent and 22 percent respectively, over four regions of the United States." [Emphasis added]

No doubt Sparks Commodities had access to data supporting the above-stated conclusion. Still, this statement falls somewhere between misleading and dishonest. Clearly, the statement leaves much to the imagination. Unless a person knows a lot about contemporary soybean herbicide use patterns, one would conclude from such a statement that RR soybeans make it possible for farmers to reduce per acre herbicide use by about one-quarter on a per acre basis.

But that is not what the statement actually says. Note that the reduced herbicide use claim is based on a comparison to "traditional soybean varieties." Even this caveat is less than truthful. What Sparks Commodities and Monsanto really mean is that herbicide use in RR soybean fields was 22 to 26 percent less than a selected number of other fields producing conventional soybean varieties.

But not any random set of fields producing conventional soybean varieties, nor even the average field producing conventional varieties; they really mean, in all likelihood, fields planted to conventional varieties on which farmers primarily used conventional, high-dose rate herbicides. Only on such fields would there be a 22 to 25 percent reduction in herbicide use. They surely do not mean the approximate 20 percent (see above data) of fields treated predominantly with combinations of modern, low-dose herbicides applied at a rate of 0.5 pounds or less of herbicide active ingredient per acre (see Table 1.6).

Nor do they mean the approximately 25 percent of RR soybean fields under a Roundup-only program that will, according to Monsanto itself, likely require three

applications of Roundup at 32 ounces per acre to achieve satisfactory control (Dunn, 1998). Such a program in 1999 cost about $30.00 per acre for the herbicide and resulted in application of 3.0 pounds of Roundup.

The above data show clearly that much more herbicide is applied to the average RR soybean field compared to the 20 percent of fields reliant largely on low-dose products. Indeed, when compared to soybean weed management systems utilizing the really low-dose herbicides, Roundup Ready fields require more than 10 times the herbicide. But it is inappropriate and misleading to pass off such a selective comparison as representative of the average field, at least in the view of this analyst.

An April 2000 USDA report, Genetically Engineered Crops for Pest Management in U .S. Agriculture: Farm-Level Effects (Femandez-Comejo, et al., 2000), makes the following statement in its abstract

"...increases in adoption of herbicide-tolerant soybeans led to small but significant increases in yields, no changes in returns, and significant decreases in herbicide use."

It is widely recognized that adopters of RR soybeans are large-scale operators who are aggressive managers. The land they farm is, on average, more productive than land managed by those who are slower to try new technologies. Hence, it is no surprise that on average, RR soybean adopters harvested more bushels per acre than non-adopters. They harvested more bushels before RR varieties hit the market, as well (Economic Research Service, 1999; Miller, 2000; Fernandez-Comejo et al., 2000; Duffy, 1999). It is unfounded to equate the slightly higher soybean yield on GMO acres as a sign of a "yield advantage," or evidence suggesting the absence of a genetic yield drag.

The slightly higher yields are largely driven by differences in management skills and soil productivity. A third factor is the likely higher degree of herbicide injury on some farms where conventional soybeans are planted and farmers apply modern, lowdose herbicides without adequate care in calibrating equipment to assure that maximum, safe application rates for a given farm's soils are not exceeded.

The claimed "significant decrease in herbicide use" is based on two measures. The first-a decline in herbicide acre-treatments -- has nothing to do with pounds applied. The second measure is the net change in conventional and herbicide-tolerant application rates over time, taking into account the increase in average rates of glyphosate use per acre and the decrease in use of other herbicides. But as explained in more detail below, this comparison encompasses so many changing variables that it is impossible to tell exactly what it means.

The claim founded on reduction in herbicide acre-treatments is fleshed out in a summary article published in Agricultural Outlook, one of USDA's most widely read publications. The August 2000 article states that-

"In 1998, adopters of herbicide-tolerant soybeans accounted for the largest share of the difference in acre-treatments (54 percent [decrease]), with most of the reduction occurring in the Heartland region." (Agricultural Outlook, August 2000, page 13-14).

This decline is the result of one of the major advantages of RR soybeans - the simplicity of the RR system and its reliance on a single herbicide for multiple weed management challenges. But it has little to do with changes in the pounds of herbicides applied per acre, since different soybean herbicides are sprayed at such different rates. In the USDA report, the authors state correctly -

"...since average application rates vary across pesticide active ingredients, the net effect of substituting one for another may be an increase or decrease in total pounds used." (Agricultural Outlook, August 2000, page 15).

On the key question of whether herbicide-tolerant soybeans reduced herbicide use, the August 2000 article states -

"...as adoption of herbicide-tolerant soybean varieties increased from 7 to 45 percent, the average annual rate of glyphosate application increased from 0.17 pounds per acre in 1996 to 0.43 pounds per acre in 1998, while all other herbicides combined dropped from about 1 pound per acre to 0.57 pounds per acre. That translates into a decline of nearly 10 percent in the overall rate of herbicide use on soybeans during that period." (Agricultural Outlook, August 2000, page 14-15).

This statement does not mean that RR soybeans reduce per acre herbicide use by nearly 10 percent. It refers to aggregate estimates of total herbicide use, not clean comparisons of an acre planted to RR soybeans in contrast to conventional varieties planted on similar soils under the same tillage system. It also does not correct for the timeliness of field operators and the quality of management, nor differences in soil quality.

It also mixes together RR soybeans and two other types of herbicide-tolerant varieties - those engineered to be resistant to the very low-dose sulfonylurea herbicides and those resistant to the low-moderate dose imidazolinone herbicides. While these other herbicide-tolerant varieties account for a relatively small share of total herbicide-tolerant acres, they clearly improve the average performance of all herbicide tolerant varieties in terms of reducing average rates of herbicide use.

Last, the above USDA estimate of a nearly 10 percent decline includes a myriad of changes in herbicide use on the approximate 50 percent of acres not planted to any herbicide tolerant variety.

If reducing the pounds of herbicides applied per acre was among the important goals shaping U.S. soybean weed management systems in the 1990s, the introduction of RR soybean varieties was a major step backwards. It is clear that the average pounds of herbicides applied on soybeans in the U.S. would have dropped by far more than 10 percent from 1995 through 1998 in the absence of RR soybeans. This is because it is likely that the majority of farmers planting RR soybeans - typically top-notch, aggressive managers-would have planted either other varieties tolerant to much lower dose herbicides, or conventional beans in conjunction with mixtures of low- and moderate dose products, or mixtures of low-dose and higher dose "standbys."

With the wide selection of today's very competitively priced low-dose soybean herbicides, farmers could easily reduce average application rates to no more than 0.5 pounds per acre, if there were incentives offered to do so. This would cut average soybean herbicide rates about two-thirds from today's levels and would indeed be a major accomplishment. It also would probably not prove sustainable nor would it prove beneficial because of other agronomic and environmental problems associated with use of many of today's low-dose herbicides.

< Table of Contents | Next Section >