e-RA: the electronic Rothamsted Archive

Broadbalk open access data

Rothamsted Ley Arable soil organic carbon 1948-2008

Last updated 02/10/2018

Access and use conditions

Creative Commons License This work is licensed under a Creative Commons Attribution 4.0 International License.

YOU MUST CITE AS: Rothamsted Research (2018). Rothamsted ley arable soil organic carbon content. Electronic Rothamsted Archive https://doi.org/10.23637/KeyRefOARLAsoc.

Click to download chart Broadbalk SOC

Click the chart above for a PDF version. Data used for this chart and data presented as separately charts for Highfield and Fosters are available to download as an Excel Spreadsheet: SOC data and individual charts. This selected data is updated from Johnston et al, 2009. The complete data set is available from the e-RA curators.

Back to Rothamsted Ley Arable home page


Ley-arable experiments at Rothamsted, on Highfield and Fosters fields, started in Autumn 1948 (Rothamsted Report for 1948 p98) on sites previously in long-term grass (Highfield) or long-term arable (Fosters). Their purpose was to look at the effects of different cropping systems - permanent grass (now grass/clover), permanent arable and alternating ley/arable - on yield and soil organic matter (SOM). The two sites have the same soil type, a silty clay loam, but very different cropping histories: Highfield had previously been in permanent grass since at least 1838, part of a grazing experiment, and Fosters had been in arable cropping for several centuries (Lawes & Gilbert, 1885). On Highfield some plots stayed in permanent grass, some were ploughed and immediately re-seeded, others went into continuous arable cropping where 3-yr arable ‘Test’ crops followed 3-yr arable ‘Treatment’ crops. On Fosters some plots were sown to permanent grass whilst others stayed in continuous arable cropping (following the same 3-yr Treatment crops followed by 3-yr Test crops as on Highfield). On both sites some plots alternated between 3-yr leys as the ‘Treatment’ crop and 3-yr arable ‘Test crops’ Originally, six blocks, one for each of the six phases of the rotation, were put down in duplicate, making 12 blocks in each field. From 1968, only four blocks on each site continued with original treatments. The remaining eight blocks were sown to continuous wheat at the end of the test-crop cycle to study the build-up and decline of take-all; this work finished in 1985 and the blocks were no longer included in the experiment. The remaining four blocks on each site continue although yields have not been taken since 1990. The old grass plots and re-seeded grass plots on Highfield and the newly sown grass plots on Fosters were grazed until 1961. They were then split to compare grass+N or grass/clover (clover was under-sown in the grass sward); the whole plots have been treated as grass/clover since 1991. Three types of leys were originally tested; 3-yr lucerne, 3-yr grazed ley or 3-yr cut grass. These, and the arable ‘Treatment’ and ‘Test’ crops have changed over the years. See Rothamsted Experimental Station, 1966; 1970; 1978 and Johnston, 1973 for details about the early years of the experiment and Yields of the Experiments, 1990 (and before) for later details. Further details are available through e-RA.

Although we no longer measure yields we continue to monitor changes in soil organic carbon (SOC) Soil samples have been taken since the start of the experiments. For 1948, 1951, 1956 and 1961 soils were sampled to a depth of e.g. 0-5” and 5-10” or 0-6” and 6-12”; later soils were all sampled to 0-9” (0-23cm). A 0-23cm value for the earlier soils has been calculated. Using an open-ended metal box (6” x 6” x 9” deep) soils were sampled in 2000/1 to measure bulk density and soil weights; the soil weight for other years have been derived by extrapolation or interpolation. The soil weights have been used to convert the %C concentration into t/ha of SOC and has taken into account any changes over time in bulk density. The figure shows the effect of two contrasting farming systems on SOC on the two experiments at Rothamsted over a 60-year time-period 1948-2008. For each site, the data shown is the mean of replicate treatments on the two blocks which have been in the same phase of cropping since 1948. The grassland contains twice as much SOC as the arable at Highfield and two-thirds as much at Fosters. SOC increased with improved grassland management on Highfield while on Fosters it increased under newly sown grass but, after 60 years there was still less than in the permanent grassland soil on Highfield, possibly because, after 1991 and the change to grass/clover on the whole plot it is likely that yields would have been less and thus a smaller input from root exudates and dead roots. With arable cropping, large amounts of SOC were lost on Highfield but it took 50 years to decline to that on Fosters. The figure shows that after 60 years the soils appear to be reaching equilibrium following changes in the cropping system level for this farming system on this soil type. Thus, in soils ploughed out of permanent grass or put into permanent grass after arable cropping the SOC now appears relatively constant. On any one soil type, the equilibrium level will be larger with permanent grass than with continuous arable cropping.

This data demonstrates that in temperate climates, SOC changes slowly and only in long-term experiments can changes be reliably monitored, explanations sought and carbon turnover models developed and validated.


Highfield ley-arable: A site which was originally permanent grassland from 1838.
Fosters ley-arable: A site which was originally long-term arable for several centuries with no recent history of grass leys.

Highfield grass: Old permanent grassland, the original sward unploughed grass/clover with a little N, grazed by sheep. Then phased in from 1962  a split plot test of grass ley with N and grass/clover (no N). In the years when the plots were split the data shown is the mean of both halves.
Highfield arable: continuous arable since 1948; 3-yr arable ‘Treatment’ crops followed by 3-yr arable ‘Test’ crops. Both the treatment and test crops have changed during the course of the experiment.
Fosters grass: Grass sown in 1948 on old arable, grass/clover with a little N, grazed by sheep. Then phased in from 1962(CHECK) a split plot test of grass ley with N and grass/clover (no N). In the years when the plots were split the data shown is the mean of both halves.
Fosters arable: Old long-term arable; continuous arable since 1948; 3-yr arable ‘Treatment’ crops followed by 3-yr arable ‘Test’ crops. Both the treatment and test crops have changed during the experiment.

Soil: Batcombe series flinty silt or loam over clay-with-flints.


Both sites have had the same management with cultivations, drilling and harvesting done on the same days.

Ploughing: At the start of the experiment ploughing was shallow, not more than 15cm deep but the depth was gradually increased to 23 cm as more powerful tractors were introduced.
Liming: Lime has been applied as required since 1958.
Ley management - permanent and re-seeded grass: Two-years sheep grazing followed by one-years hay 1949-57. Since 1958 grazing discontinued, cut repeatedly at early silage stage. Since 1991 grass leys were cut for hay in mid-summer; the hay was baled and removed.


Analysis of soil organic carbon (SOC) was by the Tinsley technique except 2000 & 2008 which are by combustion (Total C by LECO minus CaCO3-C).
Results are given as tonnes per hectare of organic C for 0-22.5 cm depth of soil, based on measured soil weights.


Long-term experiments
Soil organic carbon
Soil organic matter
Ley farming
Arable farming
Arable soils
Sown pastures

Further information

For further information about the experiment see the key references below. For a soil map of Rothamsted see Field Plan and Soil Map of Rothamsted.

Further details can be obtained from the e-RA curators and the Rothamsted Guide to the Classical Experiments (2018), p 37.

Back to top

Key References







RRes Logo - Back to RRes