Acc

epte

d A

rtic

le

This article has been accepted for publication and undergone full peer review but has not

been through the copyediting, typesetting, pagination and proofreading process, which may

lead to differences between this version and the Version of Record. Please cite this article as

doi: 10.1111/gcb.12561

This article is protected by copyright. All rights reserved.

Received Date : 08-Dec-2013

Revised Date : 08-Jan-2014

Accepted Date : 19-Feb-2014

Article type : Opinion

Do grasslands act as a perpetual sink for carbon?

Pete Smith1

1 Institute of Biological & Environmental Sciences, ClimateXChange and Scottish Food

Security Alliance-Crops, University of Aberdeen, 23 St Machar Drive, Aberdeen, AB24

3UU, Scotland, UK

*Corresponding author: Prof Pete Smith, Tel: +44 (0)1224 272702, Fax: +44 (0)1224

272703, E-mail: pete.smith@abdn.ac.uk

Running head: Grassland carbon sinks

Keywords: grassland, carbon, sink, flux, stock, sequestration, soil

Paper type: Opinion

Soussana et al. (2007) published their findings of a large and careful analysis of flux measurements

over European grasslands. The nine sites at which fluxes were measured appeared to be acting as a

sink for carbon (C), with a measured flux of −240 ± 70 g C m−2 y-1, equating to a net storage of C of

104 ± 73 g C m−2 y−1 (= ~1 t C ha-1 y-1), when C imports and exports were accounted for. Although not

proposed by the authors, but perhaps resulting from the title of a later paper (Soussana et al. 2010),

this finding has increasingly been used (particularly by organisations representing livestock

producers) to suggest that grasslands are a perpetual sink for carbon, and that just maintaining

grasslands will yield a net carbon sink. In this short article, I examine this suggestion by reviewing

evidence from repeated soil surveys, long term grassland experiments and simple mass balance

calculations, before suggesting a potential explanation for the flux findings, and presenting a series

of conclusions and policy recommendations.

Eddy covariance is a powerful tool for measuring total ecosystem fluxes of carbon. It is able to detect

changes in the net ecosystem exchange (NEE) of carbon at fine temporal resolution, and enables

estimates to be made of whether given ecosystems or land management practices result in net sinks

Acc

epte

d A

rtic

le

This article is protected by copyright. All rights reserved.

or sources of carbon. There are uncertainties associated with flux measurements, but these are

discussed in detail elsewhere (Soussana et al. 2007, 2010; Osborne et al. 2010), so are not discussed

further here. The first thing to note about flux measurements is that they measure just the flux of

CO2. This is far removed from the net carbon storage (NCS) of the ecosystem as shown by the

equation proposed by Soussana et al. (2010):

where NCS is net carbon storage, FCO2 is the flux of carbon dioxide, FCH4-C is the flux of methane

carbon, FVOC is the flux of volatile organic carbon, Ffire is the flux from fire, Fharvest is the carbon

removed in harvest from the site, Fanimal products is the carbon removed from the site in animal

products (via grazing), Fmanure is the flux arising from carbon in the manure addition to the site, Fleach

is the carbon leached from the site as dissolved and particulate organic carbon and Ferosion is the

carbon lost from the site due to soil erosion. So the first question we need to ask is, do the flux

measurements really represent a measured change in NCS? The answer in the case of the careful

study of Soussana et al. (2007) is probably yes. The authors measured many of the fluxes carefully,

and for those that could not be measured, they calculated the size of the potential contribution to

the overall flux, and showed that it could not explain the size of the measured CO2 flux. So whilst it is

important to quantify all of the fluxes contributing to NCS, in the case of the study by Soussana et al.

(2007), the flux appears to be genuine, and the imports and exports were accounted for.

If this flux is real, where is the carbon going? The carbon is not found in the vegetation, so the most

likely candidate is the topsoil. Measuring a relatively small change against a very large background

stock of carbon is difficult and requires huge numbers of samples (Smith 2004). Soussana et al.

(2010) reported very variable estimates of changes from repeated soil samplings (both positive and

negative) for soil C concentration change. Soil C stock changes were reported only where land use or

management change had occurred. We can also look for other evidence of soil carbon increase in

grassland topsoils in both repeat soil samplings over large regions, and long term grassland

experiments. Using repeat soils samples, Bellamy et al. (2005) examined soil carbon change in

topsoils in England and Wales by resampling after 20 years. They found no evidence of an increase in

topsoil C in grasslands; in fact the four grassland categories (rotational grass, permanent grass, rough

grazing and upland grass) showed small to moderate C losses. In New Zealand, which is bio-

climatically similar to the UK, Schipper et al. (2010) used repeat sampling over 2-3 decades to show

that flat dairy pastures have lost 0.73±0.16 t C ha−1 y−1, but that no significant difference was

observed in flat pasture grazed by sheep or beef, or in grazed tussock grasslands (though there was a

small increase in hill pastures). Soussana et al. (2010) summarised three other studies (two in

Belgium and one in China) which show either small positive or negative changes in soil C

concentration in grasslands over periods of 18 to 50 years. Schrumpf et al. (2011), reviewing nine

studies (some of which were also included by Soussana et al., 2010), showed increases in SOC in 4,

decreases in 2, and mixed findings or no difference in 4 studies. So there is no consistent evidence in

repeat sampling studies that grasslands are gaining in topsoil C. If we turn to long term grassland

NCS = (FCO2

+ FCH4-C

+ FVOC

+ Ffire

) + (Fharvest

+ Fanimal products

– Fmanure

) + (Fleach

+ Ferosion

)

Acc

epte

d A

rtic

le

This article is protected by copyright. All rights reserved.

experiments, we also fail to find increases in topsoil C, over many decades. Examining soil C changes

in two UK long term experiments (both >100 years duration), Hopkins et al. (2009) found no

significant change in SOC over time. If the carbon is not stored in the topsoil, could be stored in

deeper soil layers? Of the studies reviewed by Soussana et al. (2010), all were topsoil studies except

for one which sampled to a metre (and showed a soil C decline), but there are too few studies to

draw firm conclusions.

What do mass-balance calculations tell us about the feasibility of a perpetual carbon sink in

grasslands? Simple mass balance approaches dispel the idea that grassland soils could be

perpetually accumulating carbon at the rates apparent from the flux measurements reported by

Soussana et al. (2007). If we take the conservative assumption that grassland soils contained no

carbon 2000 years ago, and that they accumulated carbon at a rate of 1 t C ha-1 yr-1, stocks after

2000 years would be 2000 t C ha-1. We know that pastures in the UK contain on average 160 t C ha-1

to a depth of 1m (Bradley et al. 2005), many times lower than the calculated accumulation if

grasslands were continually accumulating carbon, much higher in fact that the carbon found in UK

peatlands (1357 t C ha-1) to the same depth (Bradley et al. 2005). Grasslands can, and do, act as a

carbon sink under many circumstances (see section below), but it is simply untenable that grasslands

can perpetually sequester carbon at the rates found in the flux measurements of Soussana et al.

(2007).

How do we explain the flux measured by Soussana et al. (2007)? I propose two explanations, the

second of which I believe is the most plausible. The first relates to improved grassland management.

It is known that when moving from poor to optimal grassland management, soil carbon levels can

increase. Smith et al. (2008), in an expansion of a meta-analysis of many global studies reported by

Ogle et al. (2005), reported potential carbon sequestration rates of 0.22 t C ha-1 yr-1 (original values

reported as CO2 rather than C = 0.81 [-0.11 to 1.50 95% CI] t CO2 ha-1 yr-1) in the cool-moist

(temperate) bio-climatic region as a result of improved grassland management – a value close to

that measured by Soussana et al. (2007). If the sites had undergone significant improvement in

management regime from a poor management baseline, such carbon accumulation rates could be

observed. This is not likely however, as most of the flux sites had differing treatments which had not

changed significantly for a number of years.

The second, and most plausible, explanation relates to legacy effects of land use prior to the

beginning of the flux measurements. Figure 1 shows the increase in total soil carbon (calculated

from changes in total N using a C:N ratio of 10:1) recorded at a chronosequence of cropland to

grassland conversions at Rothamsted in the UK (Johnson et al. 2009). The graph is striking in three

ways for the purposes of this article: firstly, it shows that soil C levels are still changing up to 100

years after a land use change; secondly, that after around one hundred years, the soil C stabilises at

a new equilibrium level; and thirdly, the grasslands would be a strong carbon sink soon after

conversion (the steepest part of the curve), and this sink strength steadily declines to zero when the

soil reaches a new equilibrium after ~100 years. This suggests that effects of changes in land use (or

Acc

epte

d A

rtic

le

This article is protected by copyright. All rights reserved.

even management interventions, such as ploughing and reseeding) many years before a flux

experiment begins, could still influence to soil C and measured C flux many years later during the flux

measurement period. I propose that the high apparent carbon accumulation rates measured by flux

towers could be measuring legacy effects of land use or land management change long before the

beginning of the experiment. Further, evidence from these long-lived effects can be found in the

cropland sites at Rothamsted UK, which show that the influence of manure that was applied

between 1852-1871 and then discontinued, could still be seen in the organic matter levels over 100

years later (Johnson et al. 2009).

Are grasslands acting as a perpetual carbon sink? In conclusion, it is untenable that grasslands act as a perpetual carbon sink. Carbon will accumulate in grasslands if they move from suboptimal to optimal management, and may still be gaining carbon many years after a change from cropland, or a management intervention such as ploughing and reseeding (which would have caused a significant decline in soil C; Johnson et al. 2009). Further, even if the idea of genuine “equilibrium” is now questioned, it can be seen in the data from long term experiments and chonosequences that under constant management and environmental conditions, soil C finds relatively stable levels (Figure 1). What does this mean for the science? Firstly, it suggests that we should be very careful in extrapolating findings beyond the experiments from which they arose. It is very easy to misread conclusions, such as this from a recent paper by Senapati et al. (2014) to which I contributed. On describing the net flux from two grassland sites under different management, the paper ends with this conclusion: “The results clearly indicate temperate sown grasslands to be a carbon sink under grazing management”. If one did not read carefully that the measurements were taken in the four years following the grassland being sown in 2005, one might assume that all grasslands were a net carbon sink, rather than being in the rapid carbon accumulation carbon stage after a land use change from cropland. Secondly, these findings show that we need to have knowledge of the very long history of land use and management on the sites upon which we conduct our experiments, as the legacies of previous land use and land management interventions are very long lasting. What does this mean for management of grasslands? Firstly, this means that simply having a

grassland does not result is a carbon sink. Secondly, it means that judicious management of

previously poorly managed grasslands can increase the sink capacity (though this will decrease over

time). Thirdly, since it is easier and faster for soils to lose carbon that it is for them to gain carbon

(Johnson et al. 2009), protecting the large carbon stocks in grasslands is an important management

target, rather than necessarily trying to increase the carbon stocks.

What does this mean for policy? Firstly, there will be no carbon credits to be gained from simply

having a grassland. High carbon stocks (total storage of carbon in grasslands, mainly in the soils),

does not equate to large carbon sinks (the net annual removals of carbon from the atmosphere). On

existing grassland, only through improving the grassland can soil C be sequestered, so where

grassland management is poor, policy should seek to improve it. Secondly, since there is much more

carbon to be lost from grasslands than can be gained, protecting large grassland carbon stocks

should be a policy priority.

Acc

epte

d A

rtic

le

This article is protected by copyright. All rights reserved.

Acknowledgments:

This paper was prepared partly in response to an enquiry to Scotland’s ClimateXChange.

References

Bellamy PH, Loveland PJ, Bradley RI, Lark RM, Kirk GJD (2005) Carbon losses from all soils across

England and Wales 1978–2003. Nature, 437, 245–248.

Bradley RI, Milne R, Bell J, Lilly A, Jordan C, Higgins A (2005) A soil carbon and land use database for

the United Kingdom. Soil Use Management, 21, 363–9.

Hopkins DW, Waite IS, McNicol JW, Poulton PR, Macdonald AJ, O ’Donnell, AG (2009) Soil organic

carbon contents in long-term experimental grassland plots in the UK (Palace Leas and Park Grass)

have not changed consistently in recent decades. Global Change Biology, 15, 1739–1754.

Johnson AE, Poulton PR, Coleman K (2009) Soil organic matter: Its importance in sustainable

agriculture and carbon dioxide fluxes. Advances in Agronomy, 101, 1-57.

Ogle SM, Breidt FJ, Paustian K (2005) Agricultural management impacts on soil organic carbon

storage under moist and dry climatic conditions of temperate and tropical regions. Biogeochemistry,

72, 87-121.

Osborne B, Saunders M, Walmsley D, Jones M, Smith P (2010) Key questions and uncertainties

associated with the assessment of the cropland greenhouse gas balance. Agriculture, Ecosystems &

Environment, 139, 293-301.

Schipper LA, Parfitt RL, Ross C, Baisden WT, Claydon JJ, Fraser S (2010) Gains and losses in C and N

stocks of New Zealand pasture soils depend on land use. Agriculture, Ecosystems and Environment

139, 611–617.

Senapati N, Chabbi A, Gastal, F, Smith P, Mascher N, Loubet B, Cellier P, Naisse C (2014) Net carbon

storage potential measured in a mowed and grazed temperate sown grassland shows potential for

carbon sequestration. Carbon Management, (in press).

Smith P (2004) How long before a change in soil organic carbon can be detected? Global Change

Biology, 10, 1878-1883.

Smith P, Martino D, Cai Z, Gwary D, Janzen H.H, Kumar P, McCarl B, Ogle S, O’Mara F, Rice C, Scholes

R.J, Sirotenko O, Howden M, McAllister T, Pan G, Romanenkov V, Schneider U, Towprayoon S,

Wattenbach M, Smith J.U. (2008) Greenhouse gas mitigation in agriculture. Philosophical

Transactions of the Royal Society, B., 363, 789-813.

Soussana JF, Allard V, Pilegaard K et al. (2007) Full accounting of the greenhouse gas (CO2, N2O, CH4)

budget of nine European grassland sites. Agriculture, Ecosystems and Environment, 121, 121–134.

Soussana JF, Tallec T, Blanfort V (2010) Mitigating the greenhouse gas balance of ruminant

production systems through carbon sequestration in grasslands. Animal, 4, 334–350.

Acc

epte

d A

rtic

le

This article is protected by copyright. All rights reserved.