2005-09-13

Soil carbon losses in England and Wales

A reader (in fact, the authors wife: which is why this jumps the queue :-) asks: what about Loss of soil carbon 'will speed global warming' (Tim Radford, Thursday September 8, 2005 The Guardian).

It says:

England's [Ahem. And Wales's - WMC] soils have been losing carbon at the rate of four million tonnes a year for the past 25 years - losses which will accelerate global warming and which have already offset all the cuts in Britain's industrial carbon emissions between 1990 and 2002, scientists warn today.

The research dashes hopes that more carbon dioxide emissions might mean more vegetation growth and therefore more carbon removed from the atmosphere.


The Guardian report even includes the nice "The study confirms the value of long-term research" - how very true. There is some evidence that this message has already been lost: the survey is a trend from original sampling (1978-83) and resampling (1994-2003), but only about 40% if the original sites were resampled.

A summary of the paper (Carbon losses from all soils across England and
Wales 1978–2003 Pat H. Bellamy, Peter J. Loveland, R. Ian Bradley, R. Murray Lark & Guy J. D. Kirk, Vol 437|8 September 2005|doi:10.1038/nature04038) is here, and a PDF of the Nature article is here. I wonder if Nature know that...

So: back in 1973 some far sighted individuals set about dividing England and Wales ino a 5km grid of about 6000 sites: 5662 were sampled for soil, in particular soil organic carbon, between 1978-83. Between 1994-5 the arable was resampled (only 853 of original 2578); between 1995-6 permanent grassland (771 of 1579); 2003 for non-agricultural (bogs, scrub, etc: 555 of 1505). Now this means we're drawing a trend (at each gridcell) using two points in time, and not the same two points either for all the cells. We'll quietly forget that from now on.

The paper notes that this is the only survey like it anywhere in the world. Its probably only possible in a densely populated rich country like England/Wales. Ideally we would have this globally: but there is no hope of that. In particular, I can't see any way to tie soil carbon changes to Kyoto commitments (without kludging it badly, which people wouldn't accept), because the data just isn't available.

Figure 1 shows you how closely C loses match up to C content (fig 3). Because they found no significant relations between C loss and land use, they don't believe that land use changes have caused the loss (England/Wales being heavily human-influenced, this is a natural first guess). Which is why they settle on climate change as the most likely explanation.

On the offsetting of Kyoto savings, this is less satisfactory. Firstly I'm not convinced that tying the two together makes much sense. Secondly the 13 Tg/y figure thats based on comes from two extrapolations. The survey was from Engalnd and Wales, top 15cm of soil. Firstly they say *if* the same is happening in the 15-30 cm layer; and then *if* we extrapolate to Scotland; we end up with 13 Tg/y. 13 Tg/y is the Kyoto cuts, so notice how the Grauniad quote has somewhat confused that issue. They do note, though, that Scottish soils tend to be carbon-rich and if anything this estimate might underestimate losses. But there is also a comment that they don't really know where the C from 0-15 cm has gone. Into the atmos is a reasonable guess; but leaching into lower layers is also possible. In which case the 15-30 cm layer might not be comparable at all.

The obvious thing to do would be to extrapolate these results across the whole temperate land area. I rather suspect that if you did, you would find results for C loss too large to be plausible, given the atmos C record. But thats a guess.

The Grauniads second para - about dashing hopes of inc veg - seems a bit curious. This study doesn't rule out increased forest growth, for example (it probably suggests it). The paper says "On the basis of atmospheric observations, net carbon absorption by terrestrial systems in the Northern Hemisphere has increased in recent decades" so presumably some of this C has gone into veg elsewhere.

Larger scale studies, and integrated carbon budgets, are needed.

[Update: by chance (?), Rising Atmospheric CO2 Reduces Sequestration of Root-Derived Soil Carbon by James Heath et al. has just appeared in Science, Vol 309, Issue 5741, 1711-1713 , 9 September 2005. ...We show that carbon dioxide enrichment, although causing short-term growth stimulation in a range of European tree species, also leads to an increase in soil microbial respiration and a marked decline in sequestration of root-derived carbon in the soil. These findings indicate that, should similar processes operate in forest ecosystems, the size of the annual terrestrial carbon sink may be substantially reduced, resulting in a positive feedback on the rate of increase in atmospheric carbon dioxide concentration. This is only a short term study with an artifical environment (as they note in the paper) but its in the same direction].

1 comment:

Anonymous said...

Although I have no "real" expertise in this research subject, I am in fact in the process of writing a course paper on the influences of future warming on the glogal carbon cycle. After having gone through about 30 articles published in the last 15 years, I get the distinct feeling that the "final word" on the evolution of carbon loss from soils hasn't been said yet.

A vast number of different experimental techniques have been applied (ranging from simply observing "natural" undisturbed ecosystems via in situ treatments (soil warming) to laboratory incubation studies), but unfortunately the results are far from conclusive. In addition, quite a few modelling studies have been done, attempting to couple global climate and vegetation dynamics, but here, too, the conclusions fall across a wide range, with some predicting net sinks, others net sources, and with varying global distributions of the two.

It seems to me that the main problem here is that respiration processes (responsible for emitting CO2 from ecosystems) are not only dependent on temperature but also on soil water and nutrient availability, CO2 fertilization effetcts and also, not to be forgotten, "local" climatic effects that cannot be described well by global models.

All these (often covarying) parameters are obscuring the signals and trends due to e.g. trends in temperature change. In addition, annual mean temperature and precipitation (used by many authors) may not be adequate to describe the response of soil microbes and plant roots, as there are strong seasonal variability in e.g. plant productivity or soil conditions.

What's your take on this?