Methane, grazing, and our credibility gap


[mea_culpa] This piece is not fair. I’m critiquing the work of one person, Christine Jones of Amazing Carbon, to illustrate an issue that is much bigger than one person. I’ve been sitting on this piece for over a year now, because I’m not excited about picking on one person in particular, and because I’ve hoped to put this critique in the context of a larger discussion. But that larger work, being outside the scope of things I actually get paid for, continues to elude me. And in the meantime I see my allies continue to cite some pretty questionable materials. So I’m sharing this now. Try and zoom out to the larger questions of how we can all do better, and think of this piece as an abbreviated case study. [/mea_culpa]

Grazing advocates and regenerative agriculture folks frequently draw on the work of Christine Jones when addressing the power of grazing systems to mitigate climate change – especially when the question of methane comes up. The most popular paper to cite is also the most problematic. It’s a 2010 article called “SOIL CARBON – CAN IT SAVE AGRICULTURE’S BACON?” Here is the juicy bit:

“Recent research undertaken by Professor Mark Adams, Dean of the Faculty of Agriculture at Sydney University, found that one hectare of pasture land could oxidise as much methane as emitted by 162 head of cattle in an entire year (Cawood 2009). The highest methane oxidation rate recorded in soil to date has been 137mg/m2/day (Dunfield 2007) which, over one hectare, equates to the absorption of the methane produced by approximately 1000 head of cattle.”

There are some problems with these claims at several levels. Let’s look at them in turn.

(1) “Recent research undertaken by Professor Mark Adams, Dean of the Faculty of Agriculture at Sydney University, found that one hectare of pasture land could oxidise as much methane as emitted by 162 head of cattle in an entire year (Cawood 2009).”

Cawood 2009 is an article from the online magazine The Land, and consists of an interview with researcher Mark Adams. No reference to the original research anywhere to be found in that article or any of Jones’ writings, and you’ll find nothing topical published under Adams’ name. If you keep digging, eventually you find references to a correction and retraction, and lot’s of dead links. Turns out that a grad student badly displaced a decimal point – turning micrograms to milligrams – three orders of magnitude! This hullabaloo is discussed in detail here, and that discussion summarized in another useful article here. A tidbit from the second article:

“A figure in micrograms had mistakenly been represented as milligrams within the calculations, meaning that the original “preliminary research” had overstated the relevant land’s methane absorption rate by a factor of 1,000. The result was that the high country soil’s methane oxidisation rate was only 8.76 kg per hectare per year, rather than 8,760 kg. That hectare of land would not support 162 cows in a carbon neutral [ed. note: actually methane neutral] manner, but 0.162 of a cow. That is: .76kg/54kg = 0.162 (Corrected)”

So, first claim: wrong by three orders of magnitude.

Now to the second claim from Jones’ article:

(2) “The highest methane oxidation rate recorded in soil to date has been 137 mg/m2/day (Dunfield 2007) which, over one hectare, equates to the  absorption of the methane produced by approximately 1000 head of cattle.”

Let’s start with arithmetic. We’ll continue to assume 54 kg / year of methane (though the Savory Institute’s review puts it at 60 to 71 kg / year). We need to convert mg to kg, m2 to hectare, and day to year.

original: 137 mg/m2/day
mg/m2 → kg/ha: 1.37 kg/hectare/day
day  year: ~500  kg/hectare/year
= less than 10 cows

In other words, 2nd claim: wrong by more than two orders magnitude.

I’m not trying to condemn anyone for bad luck or honest blunders. Mistakes are mistakes. The first one was not even Jones’ mistake – she just had the misfortune to cite someone else’s error. Please note that to her credit, Jones did post a revised version of the PDF – though not until four years later. The paragraph in the revised version reads:

“Recent research undertaken by Professor Mark Adams, Dean of the Faculty of Agriculture at Sydney University, has found that biologically active soils can oxidise the methane emitted by cattle at low stocking rates.The highest methane oxidation rate recorded in soil to date has been 13.7mg/m2/day (Dunfield 2007) which, over one hectare, equates to the absorption of the methane produced by approximately one livestock unit (LSU)”

These are much more reasonable claims! My appreciation, however, is limited by the fact that the revised PDF doesn’t actually mention anything about the correction, nor is it mentioned anywhere on Jones’ website. Jones doesn’t appear to have gone out of her way to admit the error, to publicize the correction, or address the fact that the original PDF is still in circulation, still getting linked to and re-posted, and turning up fairly high in search results. You can find it on PermacultureNews.org, the Australian Farm Institute, Regeneration International (pointing back to the Permaculture News post) – and maybe most distressingly, on Jones’ own website. She only links to the revised version from her list of publications, but the old one still turns up in search results.

Making the correction is good, but owning up to the mistakes and dealing with the fallout would be much, much better.

[EDIT: Ironically, I screwed up my own unit conversions for point (2) in the original version of this piece. Unironically, I’ve corrected the error and I’m telling you about it. Thanks to commenter properly managed livestock for catching it and letting me know. See their comment below for details.]

Having dealt with the quantitative questions, I got curious about that Dunfield 2007 citation in the second claim. It is in fact a very credible scientific reference – a chapter called “The Soil Methane Sink” in the book Greenhouse gas sinks from CABI.

But – here is what it actually says in Dunfield’s chapter:

“The highest methane oxidation rates have been measured in pristine forests, and the record is 13.7mg/m2/day measured in tropical forests of India (Singh et al., 1997).” (p. 153, emphasis mine).

Here is where my ability to assume good faith is seriously strained. Even after correcting for the 1000-fold exaggeration, the figure being used to inspire faith in rangeland methanotroph activity is taken from pristine forests in the humid tropics. Jones used this figure in an article as recently as 2014, though no longer cites the source. Nowhere does the highly significant context of that number get mentioned.

In conclusion: we have a credibility problem in the regenerative agriculture movement. We know this, but we usually chalk it up to how radical and holistic we are, and how reductionist scientists are in the pocket of corporate interests. I think those explanations are accurate  – some of the time. Some of the responsibility is ours alone.

[Aside: Jones makes other arguments about the lack of correlation between atmospheric methane and livestock that are, at least, interesting. They deserve more attention than I have right now.]