Seedbed: Jumping to conclusions about 'soil carbon'

Friday, February 29, 2008

It's not just promotional literature that can provide misleading information. It can even come from scientific publications, as in this example from the Journal of Environmental Quality

by KEITH REID

In January I wrote about some of the things you have to watch for in promotional literature to avoid being misled. I had fully intended to expand on the theme in this column by creating a misleading advertisement of my own, and then showing how I had tried to fool you.

I may save that idea for a future column, for I found I didn't have to make anything up this month. I was handed an example of dubious claims and misleading use of statistics on a platter - not in a product pamphlet, but in peer-reviewed scientific literature.

The December issue of the Journal of Environmental Quality, published by the American Society of Agronomy, Crop Science Society of America and Soil Science Society of America, carried an article entitled "The Myth of Nitrogen Fertilization for Carbon Sequestration" by S.A. Kahn et al.

This is a title sure to catch the interest of the non-scientific media, with its whiff of sensation and controversy. This is the first clue that an author may have sacrificed scientific rigour for the sake of a good story.

The Premise. The authors of this article have concluded that high rates of nitrogen fertilizer are responsible for declining soil organic matter levels over the past five decades, despite the increased inputs of crop residue from high production agriculture systems.

The Evidence. The Morrow plots at the University of Illinois have been maintained under various crop rotations since 1876, along with fertilizer and manure treatments. In the mid-1950s, the plots were split and high rates of nitrogen fertilizer (as urea) were applied to part of the plots whenever corn was grown.

The authors measured the soil organic carbon content of samples that had been collected from these plots in 1955, and compared them to samples collected from the same plots in 2005.

Organic matter was stable or declined in all of the treatments, with the greatest decline in the plots that had received the highest rates of nitrogen fertilizer. None of the treatments showed a statistically significant increase in soil organic carbon, despite huge increases in the amount of crop residue returned to the soil.

This appears to support the authors' conclusion.The Flaws in Logic. Closer reading of the article reveals that not all of the plots had been managed in the same way prior to 1955, and this had created large differences in the levels of soil organic carbon when the nitrogen fertilizer treatments were started.

In fact, the high nitrogen plots had been fertilized with livestock manure, resulting in soil organic carbon contents that were up to half a per cent higher than the unamended plots in the same rotation.

Soil organic carbon content is an equilibrium between the inputs of organic materials (as crop residue or manure) and the losses that occur as organic materials are consumed by soil organisms. With any change in the system, the organic carbon contents will change until a new equilibrium is established.

Changing from manure to fertilizer as the nitrogen source for the crop is a major change in the amount and nature of organic materials added to the soil, so we would expect a change in organic carbon content.

In fact, the greatest declines in soil organic carbon at the Morrow plots occurred in the plots with the highest initial organic carbon. These, incidentally, were the plots which received the highest nitrogen fertilizer rates, but it would be incorrect to assume that the nitrogen fertilizer had caused the decline.

The authors also presented a summary of results from 25 other published studies from across North America and Europe, as additional "evidence." Unfortunately, aside from supporting the contention that soil organic carbon contents are declining, there is no pattern in the data to suggest that higher nitrogen application rates have any greater organic carbon losses.

In fact, only six of the 25 papers included different nitrogen rates, and these studies showed virtually the same soil organic carbon losses with high and low N rates. A cynic might suggest that this is a classic case of trying to "baffle 'em with'..."

The Consequences. The article does present compelling evidence that adding more crop residue to the soil by increasing nitrogen rates is not an effective method of sequestering carbon in the soil. However, by going further in their conclusions than the data support, the authors have exposed us to a number of unfortunate consequences.

First, there will be significant resources diverted from other projects to try to confirm or refute the conclusions of the authors, or to uncover the nature of mechanisms that don't exist.

Second, there is the very real possibility of environmental policies being implemented, using this article as a justification. Limiting nitrogen rates to "save soil carbon" could hurt crop yields and farm incomes, without any impact on soil carbon losses.

Finally, and most importantly, the credibility of the scientific community is undermined. Disagreements among scientists are a necessary part of scientific progress, but they are most effective when they occur within the scientific community. Sensational claims that attract the attention of the mainstream media will only serve to push these disagreements into the public arena. BF

Keith Reid is soil fertility specialist with the Ontario Ministry of Agriculture and Food and Rural Affairs based in Stratford. keith.reid@ontario.ca