Evolution infringes upon Monsanto's patent

NPR has a useful story today by a guy who interviewed people who were involved in Monsanto's application for release, in 1993, of "Roundup Ready" crops, that is crops that are genetically engineered to be resistant to Monsanto's best selling herbicide, glyphosate, sold as Roundup. The idea, which has made Monsanto many billions of dollars, is that if the crops are immune to the herbicide, and weeds are not, fields can be sprayed liberally. Farmers don't have to do mechanical weeding, and Monsanto gets to sell them both the patented herbicide and the patented resistant seed. This set of facts, along with Monsanto acting like the faceless corporate giant that it is in defending its patents, has made it probably the most despised company among the organic farming crowd, vilified far more than competitors like ADM or DuPont.

The point of the article is that Monsanto falsely rejected the possibility that carpeting the world in Roundup would lead to the evolution of Roundup resistant weeds.

Then the story, written by Daniel Charles, continues like so:

Oops. Since then, resistance to glyphosate has emerged in 20 different weed species.
  I called up several people who were at Monsanto at that time. Why didn't people there think resistance would happen? They all told a similar story.
First, the company had been selling Roundup for years without any problems. Second, and perhaps most important, the company's scientists had just spent more than a decade, and many millions of dollars, trying to create the Roundup-resistant plants that they desperately wanted — soybeans and cotton and corn. It had been incredibly difficult. When I interviewed former Monsanto scientists for my book on biotech crops, one of them called it the company's "Manhattan Project."
Considering how hard it had been to create those crops, "the thinking was, it would be really difficult for weeds to become tolerant" to Roundup, says Rick Cole, who is now responsible for Monsanto's efforts to deal with the problem of resistant weeds.

In case the holes in their logic haven't struck you, allow me to provide a quick lesson in how natural selection (or in this case semi-natural selection), as opposed to genetic engineering, works.

Engineers at Monsanto were surely aware of natural selection and its proclivity to producing resistant pests, but they considered the idea that there would be even a hint of heritable resistance in the weed populations to be highly unlikely. This is because they failed to consider the following facts:

A. They were testing thousands of highly targeted potential genetic alterations in the lab, on a few crop species, and found that almost none of them conferred significant resistance. They didn't consider that after global distribution of their crops, trillions of genetically distinct (although totally untargetted) genetic modifications (that is, natural mutations) in thousands of weed species would be tested for their resistance. When something potentially useful did pop up, Monsanto was again able to test hundreds or maybe thousands of slight modifications on that, while natural selection could within a few years test millions of potential modifications iteratively over several generations. So they didn't consider that nature's search for solutions would be far more exhaustive than theirs.

B. I'm sure engineers have a term for closely examining one type of failure risk while completely ignoring others. That's what Monsanto did. They were look hard at making plant tissue resistant.  From the same story:

Some weeds, Cole says, appear to keep glyphosate from entering the plant at all; others sequester the herbicide in a spot where it can't do much damage. Monsanto's genetically engineered crops use a different technique entirely.

So they didn't consider the possibility that some plants would simply shield their vital tissues from the toxin, the way many metal resistant plants do.

C. They assumed that because Roundup had been used broadly for several years already, and there were no known resistant weeds, weed populations simply had no resistance traits available for natural selection to favor. They failed to consider that with the introduction of their crops, and the resulting increase in usage, both the population size of the exposed weeds and the force of selection for resistance would increase dramatically.

The force of selection is a measure biologists use to ask the question, how much difference does a heritable change of a certain size in a trait (for example a 0.1% increase the probability of surviving a spraying with Roundup) make to the fitness of the individuals with that altered trait. So long as most individuals in a weed population were never exposed to Roundup, the force of selection for resistance to it was small. Resistance doesn't help much if you are never exposed. The seeds blowing into farmers fields were coming from unexposed sub-populations, and so were not resistant. When we started blanketing the world in Roundup, the force of selection increased, because most every weed subpopulation over huge areas was exposed. So their experience up to that point led them to underestimated the force of selection for resistance. And if there is one thing that evil empires should know, it is to never underestimate the force.