Thursday, August 14, 2014

Danish Pigs Accumulate Glyphosate!

This is likely one of the worst papers ever to be published.

However, to some, it is stellar evidence.

Here's one website claiming that the link between glyphosate and piglet deformities is now CONFIRMED!   

At least that's what many in the anti-GMO community think, and a recent paper is all the proof they need.  A new steamer from Monika Kruger's group in Germany provides low-power investigation oozing with no science, heavy rhetorical flourish, and missing controls. The whole thing is designed to scare the bejezzus out of the credulous.  And it works! 

I only found this because an anti-GMO soul on a comments section pointed to it as "irrefutable evidence tying glyphosate to birth defects".  The work appears in May's J Environ Anal Toxicol. First, articles in "Omics" journals are usually pretty lame. They are a well known predatory publisher that will publish anything if the price is right.  

Second, Kruger's group always seems to find a problem, yet their reports are devoid of experimental methods and clear statistical treatments. This report has some interesting numbers, one paragraph of results (that make no sense) and a page of pictures of dead piglets!

I don't understand the second paragraph where the authors say that there were no differences in concentrations.  Then they say the highest were in the lungs. Remember, this one paragraph is the parallel to the Results section of a conventional scientific paper.

 In the usual form, this group isn't quite sure what the results of their study are exactly.  Are there differences or are there not differences?  

The data in Table 1 tell tell their story... of course, no positive or negative controls, so you don't know if the assay is just reading noise.  No standard curve, so hard to know how they can quantitate this so accurately!!   They don't say in the paper. 

Like any good paper smearing glyphosate or transgenic crops, the authors choose to conveniently omit the controls. That's my guess.  They have to know to use controls, and probably do, but they look the same as the experimental treatments and that's just not shocking (see Figure 3, Seralini et al., 2012 aka Lumpy Rats). To do this right there should be a positive control (organs spiked with glyphosate) and a negative control (pig organ never exposed to glyphosate).  Classic sloppy science!  Awful!

The clear problem is that these are all pigs from that one Danish farmer that claims the relationship between glyphosate and pork defects. If he really believed his claim and wanted independent analysis, he might send it to an independent lab for objective analysis.  Instead, he sends it to Monika Kruger's group, one of a handful of labs worldwide that somehow can always find a boogeyman tied to GM crops.

Without any controls, how do we know the farmer (with a GMO axe to grind) doesn't soak piglets in glyphosate before sending them for analysis?  That would give Kruger's lab the benefit of the doubt.  The other question is if they really detected anything at all.

There is no standard curve, and no negative control. If the detection method (ELISA; an antibody-linked colorometric assay) cross reacted with anything in the piglet juice, or if there was natural background in the assay, how would we know?

And the second page of this report is just pictures of piglets.  Not sure what it means, other than presentation for shock value.  That's it.

Here's another great example of "shock science".  Pictures of deformed piglets. They don't say that the piglets were exposed to glyphosate, they don't measure glyphosate, they don't show normal piglets.  This figure is purely there to shock and frighten... and it works! 

How does this junk get published without controls!!!

Let's talk about what they claimed to find.  The authors claim finding 80 ug/ml in lungs, maximum, but in the table it is 80 ug/g.  Are they talking about ml of extraction fluid or starting tissue?  Unclear. They also need a lesson in use of significant digits. 

Let's say that's true. 80 ug/g starting tissue.  That's the same as 80 mg/kg.  When glyphosate is sprayed on the crop it is about 20 mg/square meter which provides about 0.29 kg of soy***.  Spraying is done long before the flowers and beans are on the crop (see Inconvenient Glyphosate Math).  You see the problem.  There would have to be some sort of mystery mechanism sequestering glyphosate in the fetus, which is inconsistent with everything we know about its pharmacology.

Plus, the MSDS clearly points out that there is no evidence of teratogenicity in a suite of animals examined.

Once again, a real dud of a paper becomes part of the evidence base for anti-GM.  I would not be proud of accepting this work and claiming it as convincing evidence. I'd be embarrassed.  But the anti-GMs have a thin skill set and ambitious keyboard.  They'll argue that this is appropriate science, and excellent evidence of glyphosate's harmful effects.

And the best part is that it just keeps on coming.  The fact that this is their standard for conclusive science says a lot about their scientific tenacity... it is severely absent.

***  please note that I this value was changed on 8/20 from the original posting of 2 kg.  That number was incorrect, I forgot to add a decimal place!  This was pointed out to me in the comments section and has been corrected. It does not change the point of the paragraph. 


Tom said...

What's Kr├╝ger's story? She's been publishing a couple of sub-standard glyphosate papers in reputable journals as well (although not stunning impact factors), usually not making clear distinctions between glyphosate and RoundUp. See for instance pubmed IDs 25079171, 24268342, 23396248 and 23224412. Is she yet another nobody who suddenly got addicted to the anti-GMO limelight? Even her "real" articles are clearly flawed and/or inconclusive.

Loren Eaton said...

First, a "steamer"? I love it, can I use that?
Second, I haven't read the paper, but would I be right in assuming that many of the conclusions are based on what could be called 'incomplete' experiments? The kind that they can say 'requires more research' while retaining that ability to try and scare people?

TheOldTechnician said...

I'm beginning to think this Danish farmer is really abusing his animals (through simply poor husbandry or something more sinister) and is using anti-gmo fervor to cover it up.

Paul Anderson said...

Kevin, either your math or mine is a bit screwy.

Let's see:

4047 square meter = 1 acre

27.2 kg = 60 lb = 1 bushel soy

2 kg/sq. m = 8094 kg/acre = 297 bushel/acre. That's a bit high, considering the national average soybean yield is about 44 bushel/acre. 0.3 kg/sq. meter is more like it.
As for the amount of glyphosate sprayed on that 0.3 kg of soybeans, the standard rate is 1 lb/acre, or more than 100 mg/square meter. In the major U.S. soy growing areas, it's often sprayed at rates in excess of 150 mg/square meter, at least once after the soybeans have begun blossoming. (though usually not after pod fill begins, because that can cause yield loss)

So that's 500 mg/kg(of final grain yield) of glyphosate commonly sprayed on blossoming soybeans. It doesn't seem so unrealistic that you might find soybean grain samples with 50 mg/kg glyphosate in them.

Now, that may not be biologically significant, and my math might be messed up anyway, but the "science side" loses a lot of credibility by not using realistic numbers.

TheOldTechnician said...

Folta's numbers may have been on the low side but i would think that all but a very small fraction of what is sprayed is actually intercepted and absorbed (and even preserved) by the edible portion of the plant, to wit, the beans. The spray would cover everything in that square meter, most parts of the plants and the ground. Then there is the situation that whatever glyphosate actually finds itself in the sows would tend to be cleared by the kidneys.

Curt Hannah said...

My family has done this experiment albeit on a much larger scale. The family farrows, fattens and sells 10,000 to 13,000 hogs per year. We have been in the hog business for more than 1/2 century. We switched to Round-Up Ready beans when they came out in 1996 and to Round-Up Ready corn when it first became available. Our birth rate of hogs has never been higher and the number of hogs per litter we sell to the packing houses has never been higher. We get paid on the “dress-out” of the hog and we are always towards the top. If there were a problem with feeding Round-UP Ready soybeans and corn we (as well as virtually all the hog producers in the US for the last 18 years) would have seen it!!

Kevin M. Folta said...


Thanks for your input on this, I'm ALWAYS trying to keep it real. Good numbers are the cornerstone of credibility when critiquing a paper that has bad numbers.

You did point out a clear mistake here that was an incorrect translation from my original article, and I will fix that. Check out Inconvenient Glyphosate Math on July 5.

My number of 83 mg/square meter comes from the fact that I'm reporting the amount of active ingredient. If we add in water, inert stuff, etc it probably hits 200. I also based on .75 lb/acre rather than 1 lb, which is what came from the website application rates.

I'm not a soybean farmer, but I have checked with two of them before posting my original article on this. They both told me that glyphosate is not sprayed after a certain early stage, 4-6 weeks after germination. The labeling convention says not after "R3" stage, so it well precedes flowering. Certainly I don't doubt that some may do this later. Could happen, but it is not like it is sprayed on the stuff that is used for food.

You are EXACTLY right about the number of beans/ sq meter. The number I got in the original note was 291 g/sq meter, which is actually lower than your guesstimate. Thanks for pointing this out. I was off by a factor of 10, and it has been corrected.

Kevin M. Folta said...

Paul, sorry, hit the button too soon.

The other factor is that let's say it is a high number like you suggest. How much of that gets on the plant? How much of that goes to the understory? How much is actually there weeks later?

We don't have to guess. USDA numbers show that you can occasionally detect 20 mg/kg on finished beans, but the vast majority are 0 (detected).

Paul Anderson said...

Kevin, thanks for your response.

A couple of more points:
1. The main post currently reads: "When glyphosate is sprayed on the crop it is about 20 mg/square meter which provides about 0.29 kg of soy" But your comment says "83 mg/square meter", which would be the 0.75lb/acre rate. If your original sentence had read: "When glyphosate is sprayed on the crop it is about 83 mg/square meter which provides about 0.29 kg of soy", I wouldn't have commented in the first place. It's still the lowest use rate, and probably little used anymore, but at least it's a real use rate.

2. The "R3" growth stage of soybeans is "beginning pod fill", after most flowering is past. That follows the "R2" stage(full flower) Many fields are sprayed at the R1 stage and into the R2 stage, during flowering. (Depending on maturity for zone, and planting date, soybeans can easily be at R1 six weeks after planting)
Sorry for the nitpicking, but it takes away from the value of your work when you say that soybeans aren't sprayed after flowering has begun(or that "R3" precedes flowering). There's just no point in giving the GMO haters any room for legitimate criticism.