I’m sure everyone who follows the genetic engineering debate has heard about the report from The Organic Center which lays a net increase in pesticide usage at the feet of genetically engineered crops. So I finally found a link to the report itself [warning pdf, also 69 pages]. I’m neither a statistician nor an agronomist (despite my awesome ISU hat which has exactly that slogan), so I’m not qualified to confirm or refute the numbers they put forward. Hopefully we’ll see more detailed analysis on that end from someplace like Biofortified or Sustainablog. I now have some analysis of the methodology of the report itself, tracked down by gntis on the biofortified forums. What I can do is given a bit of the broader context about the context of their numbers and what they don’t mean. This post will be in the following format:
- The 318 million pounds in context
- Chemicals are different
- –Different Toxicity
- –Different Persistance in the Environment
- Herbicide resistant weeds
- One trait vs a technology
318 Million Pounds in Context
Example tweet:
Pesticide use has skyrocketed by 318 million lbs (in last 13 years) with use of #GMO seeds!
Let’s put that number in perspective. The total number of pounds of pesticide active ingredients used in agriculture has remained relatively flat over the past two decades. The most recent number I could find was 480 million pounds of active ingredient per year. Compared to 480 million, an increase of 318 million looks huge but it is not a fair comparison.
I swear there's an xkcd for everything
Since the increase attributed to genetically engineered crops is 318 lbs over 13 years, the honest comparison is 24 million pounds per year to 480 million pounds total every year; approx. 5% of total weight pesticide active ingredients for a given year. While 24 million pounds of ANYTHING is still a lot, I don’t think 5% can honestly be described a skyrocketing.*
Different Chemicals are Different: Toxicity
I always like to make the point that pounds of active ingredient is a deceptive measure in the first place, since toxicity between different herbicides can vary massively, and insecticides will generally be more toxic to humans than herbicides (we’re more closely related to insects than to plants, which means we share more of the same biochemical pathways and enzymes). But the report had some data that really drives it home for me. Back in 1996 when herbicide tolerant crops were new (and therefore couldn’t have yet contributed to the development of glyphosate resistant weeds), the average application was .63-.69 pounds per acre per year for various crops (table 4.1 on page 36). On the other hand, outside of glyphosate, the market has apparently (according to the report) has be moved towards the use of so-called low dose, or even very low dose herbicides. When pyraflufen ethyl, a very low dose broad leaf herbicide** was applied to fields it was applied at an average of only .003 pounds per acre (210 times less than roundup/glyphosate). Anyone think that’s because farmers needed only 1/200th the weed control power? It’s because pound for pound pyraflufen ethyl is much more lethal to weeds.***
Different Chemicals are Different: Environmental Persistence
Farmers in the US used 76 million pounds of atrazine and 135 million pounds of glyphosate (Round-up). Yet in watersheds feed by agricultural run off, atrazine is present at much MUCH higher levels. (source #1, source #2). That’s because glyphosate breaks down faster in the environmental, as well as more tightly binding to the soil in the area it’s first applied to. The half life**** of glyphosate in soil is less than seven weeks. All things being equal pesticides that don’t leave the area they were first sprayed in and degrade too fast to build up in the environment are safer than ones that are more mobile and can persist for years (for comparison the half life of DDT can be as long as 22 years, giving it lots of time to accumulate in the environment).
Herbicide Resistant Weeds:
The report spends a lot of time talking about the emergence of glyphosate resistant weeds. I’ve already addressed the development of herbicide resistance in weedy species and why they don’t by any stretch of the imagination deserve the name “superweeds” here. So I only have a couple of things to say:
- Resistance happens. It happens with antibiotics and other drugs, and while we do everything we can to slow it down, we don’t refuse to use antibiotics because we know by using them we ensure that someday diseases will no longer be susceptible to them.
- The development of resistance could be slowed down a lot if we had more herbicide/herbicide tolerance systems on the market. Right now there are only two Roundup Ready and Liberty Link, and for a lot of crops Roundup Ready was the only system available for most of the time period covered by this. Using a single herbicide over and over again speeds up the evolution of resistant weeds. When herbicides switch from year to year any resistance to to this year’s herbicide will be useless against next year’s so it isn’t selected for from generation to generation. It throws a wrench in natural selection and dramatically slows down evolution.
- Proponents of organic agriculture (like The Organic Center) who complain about the development of herbicide resistant weeds are trying to have their cake and eat it too. If they’ve already sworn off the use of synthetic herbicides, they can’t turn around and complain about weeds resistant to herbicides that they don’t want used in the first place.
That said, herbicide resistant weeds (like antibiotic resistant infections) are a real problem. The solution in both case is continued research into new alternatives, along with the adoption of better practices (rotating between different herbicides) to extend the lifespans of those new alternatives.
Trait vs Technology:
The numbers released by The Organic Center reflect both an increase in herbicide use (subject to all the caveats described above) from herbicide tolerant crops and also a decrease in insecticide use from bt crops. Total insecticide active ingredient usage is much smaller than total herbicide usage so the herbicide increase swamped out the insecticide decrease. But that doesn’t change the fact that one of two biggest traits on the market that genetic engineering has given us, had a positive effect even by the criteria used by The Organic Center. Even if you don’t buy anything else I’ve said, it doesn’t condemn a whole technology for a single use it can be put to. No more than it makes sense to condemn the entire internet for identity theft (or the lawsuits of the RIAA).
*A situation where I would apply the word “skyrocket”: Tuition at UC Berkeley is going to increase 30+% between now and next fall, and at which point it will have increased more than 280% over the past eight years (2002-2010).
**The two biggest group of plants are monocots and eudicots. Broadleaf herbicides kill eudicots, which tend to have wider, rounder leaves than the monocots people are most familiar will grasses (including grains like wheat, rice, corn, sorghum, barley, and oats). Broadleaf herbicides are sprayed on grain crops to kill non-grass weeds.
***For an example in humans, let’s use dioxin and sodium cyanide. The LD50 (the dose that kills half of the animals that receive it) for sodium cyanide is 6.4 milligrams/per kilogram which can be used to calculate that consuming even half a gram has a good chance of killing a 175 pound person. For dioxin, the LD50 is .02 milligrams per kilogram so even 1.6 milligrams (more than 300 times less) stands a good chance of killing that same person. A 10 milligram combined dose of sodium cyanide and dioxin might be completely lethal, or relatively harmless depending on how much of it was each of the chemicals. That’s why reporting total weight of pesticide active ingredients isn’t very informative by itself.
****The time it takes half of a chemical to chemical. After the half life has passed only half the chemical is left, twice the half length a quarter of it is left, after three times the half life an eighth and so on.