The ERS report makes very clear, in a preface to the tables, that their
data is based on means of adopters versus means of non-adopters -- not a
direct comparison of yields in comparable circumstances. RR soybean
adopters (and Bt-transgenic adopters) may include a higher percent of
large, technologically sophisticated growers; grwoers may be planting their
RR beans on their better land; RR bean agronomic practices may be completed
on a more timely basis, on average. These and other factors are not
controlled for in the average yield comparisons made by the ERS analysts,
and hence their caution to those reading the tables.
We are close to posting a detailed analysis of RR soybean yield drag,
drawing on several state soybean varietal trial data. Hopefully it will be
posted by July 9ish. The report/section on Ag BioTech InfoNet will have
links to several -- not all -- state soybean varietal trial reports for
1998. The data is absolutely clear and compelling -- the RR yield drag is
real, and for some farmers a surprisingly serious "hidden tax" on the
overall profitability of their operations. My review of the data suggests
that the Oplinger conclusions understate somewhat the degree of RR yield
drag on average, and for some growers, understate it significantly.
The recent ERS report requires a careful reading. In a nutshell, no one
(including Monsanto) ever claimed that inserting RR resistance genes in
soybeans would raise their genetic yield potential, other things being
equal. Presumably, most seed companies have sought to insert the RR
gene/trait into the best commercial varieties for various regions/maturity
groups. There are two ways that the RR bean version of the best commercial
varieties could produce higer yields (as found in the ERS table and some
varietal yield comparisons) --
* better weed control than otherwise attainable
* avoid sulfonylurea and/or other herbicide injury to soybean yields (or
weed management practice injury, like cultivation taking out too many plants)
In all likelihood, the second factor is probably the bigger factor/more
common explanation on those farms and those trials that show a RR yield
advantage. Why? Because all the new chemistry, the price reductions and
the guarentee programs have planted in farmers' minds a very high level of
expectation for weed control; most fields are being managed for weeds very
aggressively and successfully. But the actions taken can set the plant
back, indirectly reducing yields below what they otherwise might be.
Most farmers not using a Roundup based system are applying either
sulfonylureas or imidazolinone herbicides which are known to cause some
stunting/damage to soybeans earlier in the season, that reduce actual
yields below potential yield in the absence of herbicide injury. That's
one of the reasons why several companies developed IMI and STS soybean
varieties -- to avoid carryover injury from corn fields treated the year
before.
On the issue of reducing herbicide use, herbicide tolerant varieties
almost always result in increased reliance on herbicides as the major or
sole weed management tactic. This, by definition, will increase herbicide
reliance and use over time, unless the systems are so effective that they
reduce weed pressure over time (unlikely). There are two ways that
herbicide tolerant varieties could reduce "use."
First, growers switch from a acetanilide (alachlor, metolachlor) or other
"old" chemistry herbicides that are applied at .75 to 1.5 pounds a.i. per
acre to Roundup-based system involving maybe .5 to .7 pounds a.i. per acre.
This is not a reduction in reliance, it is a switch to more
biologically-active and persistent chemistry effective at lower use rates.
If the goal is/was use reduction, a RR bean system averaging in 1998 about
1 pound a.i. per acre entails almost 10 times the herbicide as some
sulfonylurea systems, and at least 5 times more a.i. than several of the
new low-dose herbicides like sulfentrazone. So any claims that RR beans
reduce herbicide use always are incomplete without the phrase "...reduced
COMPARED TO WHAT??"
Second, growers plant RR beans and spot treat fields where weeds get out
of control, using non-chemical alternatives in other parts of field, like
cultivation. While clearly a possible option, this is being used by a very
small percent of U.S. soybean growers.
Results from the field this year are showing that in many parts of the
Cornbelt, a full 32 ounce application rate of Roundup Ultra, or 1.5 pounds
a.i., is needed to control velvet leaf and ragweed species, and that in
some fields even the full rate is not working reliably. As projected by
many weed scientists, the average rate of application and average number of
applications will probably incrementally rise until weed shifts and
resistance render RR soybeans not economically attractive. This will
probably not happen for a few years because of the extraordinary price
reductions that have occurred recently, and the prospect that generic
competition in the Roundup market will drive the price per acre treated
down as low as $4.00 to $5.00 from todays about $10.00 cost.
A careful reading of the USDA data and report do not support a general
claim that RR systems have reduced herbicide use, and experience in the
field in 1999 suggests strongly that use this year will rise perhaps 15% to
25% above 1998 in terms of average pounds of Roundup per acre. This years
average rate per crop year for Roundup in beans was 0.92 pounds, 1.3
applications per acre. In 1999, look for maybe 1.6 applications per acre
and perhaps 1.2 pounds per acre on average. In 1998, the ONLY soybean acres
treated with 1.2 pounds or more of herbicide active ingredient were --
* 2% acres treated with alachlor at 1.7 pounds a.i. (and probably another
a.i.)
* 4% acres treated with metolachlor at 1.7 pounds a.i. (ditto)
* Some small portion of the 18% of acres treated with 1.0 pounds of
pendimethalin, when the amount of a second/third a.i. is added in.
* Some small portion of the 16% of acres treated with trifluralin at .91
pounds, when the amount of other a.i.s are added in.
We will do the analysis next year when the 1999 field crop chemical use
data is released, but I project that Roundup will be the major a.i. on
55-65% of the acres, and that on most of the rest of the acres, the average
rate of herbicide use will be under 0.5 pounds. On perhaps 15% to 20% of
the acres, the rate will be under .25 pounds. So COMPARED TO THESE
SYSTEMS, RR soybeans will require about a 5-fold increase in use on a
pounds applied basis.
What is really needed is an "impact adjusted" measure of use and reliance,
so that growers can more fully evaluate the consequences of their choices.
We are working on such measures.
More later.
chuck
Charles Benbrook 208-263-5236 (voice)
Benbrook Consulting Services 208-263-7342 (fax)
5085 Upper Pack River Road benbrook@hillnet.com [e-mail]
Sandpoint, Idaho 83864 http://www.pmac.net
To Unsubscribe: Email majordomo@ces.ncsu.edu with the command
"unsubscribe sanet-mg". If you receive the digest format, use the command
"unsubscribe sanet-mg-digest".
To Subscribe to Digest: Email majordomo@ces.ncsu.edu with the command
"subscribe sanet-mg-digest".
All messages to sanet-mg are archived at:
http://www.sare.org/htdocs/hypermail