Lawes and Gilbert established several other Long-term Experiments at Rothamsted in the mid 19th century. Those which continued became known as the "Classical Experiments", one of which stopped in 1990. These are:
|Experiment||Code||Experimental Objectives||Year started|
|Barnfield||R/BN/7||Effects of fertilizers and organic manures on root crops, arable crops and grass & clover. No treatments applied or yields measured since 2001||1843-2001|
|Agdell||R/AG/6||Effects of fertilizers and organic manures on four-course rotations (cereal, root, legume, fallow). Stopped in 1990 after testing residual effects of P and K. Only archived crop and soil samples are now available||1848-1990|
|Garden Clover||R/GC/8||Factors affecting the persistence of continuous red clover (Trifolium pratense L). Including residual effects of fungicides (Benomyl)||1854|
|Exhaustion Land||R/EX/4||Residual effects of mineral fertilizer and manures applied 1876-1901 and of additional phosphate applied since 1986, on the yield of s barley (up to 1991) and winter wheat. A test of potassium was introduced in 2007. Soils now have a range of plant available P and K||1856|
For more information refer to the Key References listed at the end of the brief description of each experiment. Data from these experiments is not currently available from e-RA. For more details of what data is available and how to access it, please contact the e-RA Curators.
More recently, in the 20th century, other long-term experiments have been established both at Rothamsted and on the contrasting soils at Woburn Farm, 40km North of Rothamsted. These experiments focus on ley-arable rotations and the use of organic manures. They include:
|Experiment||Code||Experimental Objectives||Year started|
|Highfield Ley-Arable||R/RN/1||Effects of continuous arable and ley arable cropping systems on soil organic matter and fertility; originally long-term grassland (>100 years)||1948|
|Fosters Ley-Arable||R/RN/2||Effects of continuous arable and ley arable cropping systems on soil organic matter and fertility; originally long-term arable (>100 years)||1948||Highfield Bare Fallow||R/RS/1||Effects of long-term bare fallow on soil organic matter and fertility, after long-term grass (>100 years)||1959||Highfield Conversion||R/CS/767||Effects of changes in agricultural land use and management on crop production and soil physical, chemical and biological parameters, including soil organic matter and structure, soil fauna and microbial diversity; originally long-term grass (>100 years)||2008|
|Woburn Ley-Arable||W/RN/3||Effects of continuous arable and ley-arable cropping on crop production, soil organic matter dynamics and fertility in a sandy loam||1938|
|Woburn Market Garden||W/RN/4||Originally, effects of organic inputs (FYM, compost and sewage sludge) on soil and crop yield. Later the effects of heavy metals, added in the sewage sludge, were investigated. Part of the experiment was moved in 2005. The experiment was put under grass in 2006.||1942|
|Woburn Long-term sludge||W/CS/427, 428 & 439||Effects of heavy metals contained in sewage sludge on soil fertility, crop quality and microbial activity.||1994|
|Woburn Organic Manuring||W/RN/12||Effects of organic manures and grass leys on soil fertility, organic matter and crop production||1964|
|Amounts of straw||R/CS/326 & W/CS/326||Effects of straw incorporation on soil organic matter dynamics, soil structure and fertility in a silty clay loam and sandy loam soil||1987-2017|
|Continuous maize||R/CS/477 & W/CS/478||Effects of different organic matter inputs (maize tops vs cereal stubble) on soil organic matter dynamics and fertility in a silty clay loam and a sandy loam. The experiment finished in 2015||1997-2015|
|Acid strip||R/RS/9||Effects of soil acidity on soil properties under winter wheat||1850s|
|ECN plots (Rothamsted and North Wyke)||R/CS/409||Monitoring environmental change by measuring key physical, biological and chemical variables using well defined protocols. The main ECN plot is located in the south-east corner of Park Grass.||1992|
|Long-term Liming||R/CS/10 & W/CS/10||Effects of liming on crop yields in heavy soil (Rothamsted) and light sandy soil (Woburn). Since 1997 the Rothamsted plots have been in grass; the Woburn experiment is discontinued.||1962-1997|
For more details, refer to the Key References listed at the end of the brief description of each experiment. Data from these experiments is not currently available from e-RA. For more details of what data is available and how to access it, please contact the e-RA Curators.
With thanks to Andy Macdonald, Paul Poulton and Steve McGrath.
Barnfield was the first of the Classical Experiments, with treatments applied in spring 1843 for a crop of turnips sown in July. The treatments and cropping, although mainly roots, differed until 1876 when a period of continuous cropping with mangolds was started that lasted until 1959 (sugar beet were also grown, on half-plots, from 1946).
Treatments during the first two years were on long narrow plots, as on Broadbalk. After various modification, the design was settled in 1856 when the strips testing minerals and FYM, including FYM + PK, were crossed at right angles by series comparing no N fertilizer with both inorganic and organic forms of N supplying 96 kg ha-1. Before 1968 this was the only Classical in which N was applied with both FYM and FYM + PK fertilizer.
Because yields of continuous roots were declining, perhaps because of increasing numbers of cyst nematodes (Heterodera schachtii), the cropping has been progressively modified since 1959 and has included a range of arable crops, with an increased range of N dressings, and grass. From 1977 to 1983 the series that had never received N fertilizer was kept fallow. It was sown to a grass-clover ley in 1984. The remainder has been in grass since 1975.
No treatments have been applied and no yields taken since 2001, although the site is currently still available. Recently soil from the site was used in a study investigating the response of wheat roots to soil P supply (Yuan et al, 2016).
Taken from the Rothamsted Guide to the Classical Experiments 2006, page 38.
Agdell was the only Classical Experiment in which crops were originally grown in rotation. From 1848 to 1951, three different manurial combinations (none, PKNaMg and NPKNaMg plus rape cake, castor meal) were applied to the root crops of two four-course rotations. The rotations differed only in their third course - roots, barley, fallow or legume, wheat. There were only six plots and only one course of the rotation was present each year. The root crop was turnips or swedes, the legume clover or beans. From 1920, club-root (Plasmodiophora brassicae) became progressively more damaging to the root crop, especially on the NPKNaMg plots as a result of increasing soil acidity. By 1948 the produce was too small to weigh, and the four-course rotation ceased in 1951. Soil acidity was corrected and the plots were then used to evaluate the P and K reserves accumulated up to 1951. During this period the original six plots were halved and two levels of soil organic matter were established by growing leys on one half. Subsequently, the plots were further sub-divided to build up different amounts of P and K in the soil. Crop yields were then related to the reserves of P and K in the soil and the effect of adding fresh P and K. The experiment ended in 1990, only archived crop and soil samples are now available.
Taken from the Rothamsted Guide to the Classical Experiments 2006, page 39.
Current design: Four plots (each 1 x 1.4m) growing continuous red clover (Trifolium pratense), reseeded every 5-6 years. Lime and fertilizers (P, K and Mg) are applied in autumn. Vegetation is harvested three times a year.
Strengths: The only remaining Classical experiment devoted specifically to the study of a legume, and the oldest experiment of its kind in the world. Long-term data on clover yields and many archived samples (from about 1862) available. Soil samples, which show a decline in total N of >50% since 1857, are also available.
Background: the smallest and simplest of the Classical experiments, with (until 1956) only one, unmanured plot measuring 3m by 2.1m. The experiment was established in the Manor Garden in 1854 to see whether red clover could be grown continuously on a rich garden soil. Yields were very large for the first 10 years averaging about 10 t dry matter ha-1, probably because the soil was rich in nutrients and the soil-borne pests and diseases of clover were absent. Reasonable yields were obtained over the next 30 years but thereafter yields showed a marked decline and there were several complete failures.
Between 1956 and 1972 the plot was sub-divided and a sequence of tests made of K, molybdenum (Mo), formalin, N and Mg, but the the crop was usually severely damaged during the winter by clover rot (Sclerotinia trifoliorum) and was resown each spring. Since 1973 basal N, P, K, Mg and chalk have been applied.
Between 1976 and 1978 aldicarb was tested as a control for clover cyst nematode, Heterodera trifolii, which was known to be present, and the cultivar Hungaropoly, believed resistant to clover-rot, was compared with the standard susceptible variety S.123.
The plot then grew Hungaropoly only, with basal aldicarb (until 1988), and tested the fungicide benomyl from 1980-90. The cultivar was changed to Merviot in 1996. Between 1979 and 2006 the experiment has been resown seven times. A mean yield of 13t ha-1 has been achieved in this period, with up to 20 t ha-1 in some years. In 2013 the clover variety was changed to Milvus
Other than Park Grass, with its mixed herbage, this is the only Classical experiment where a non-graminaceous crop has been grown continuously. In terms of microbial diversity, its soil provides a potentially valuable contrast with those of Broadbalk and Hoosfield.
Background taken from the Rothamsted Guide to the Classical Experiments 2006, pages 37-38.
Unlike some of the other Classical experiments, which have been modified without losing the continuity of many of their treatments, the Exhaustion Land experiment has had several distinct phases since it started in 1856. From 1856 to 1901, annual dressings of N, P, K or FYM (from 1876 only) were applied. Wheat was grown initially (1856-1875) then potatoes (1876-1901). There were 10 plots from 1876 to 1901.
From 1902 to 1939 no fertilizers or manures were applied and, with a few exceptions, cereals (usually spring barley) were grown. From 1940 to 1985, spring barley was grown and N fertilizer applied to all plots every year, initially at a single rate, but in 1976 the 10 main plots were divided to test four rates of N.
In 1986, after a long period when the P residues in particular were being “exhausted”, it was decided to see how quickly this decline in soil fertility could be reversed. Annual, cumulative dressings of 0 v 44 v 87 v 131 kg P ha-1, as triple superphosphate, were tested on five of the original plots (each divided into four sub-plots). Basal N and K were applied such that these nutrients did not limit yield. Responses to fresh P were rapid. Applications of P stopped after seven years. No P was applied between 1993 and 1999, but since 2000, maintenance dressings, equivalent to offtakes by the crop, have been applied (not to the no-fresh-P sub-plots). Wheat has been grown since 1992. Typically, it showed the same response to available-P as spring barley i.e. above a critical level, on this soil, of about 12 mg kg-1 there is no further increase in yield.
On the other half of the experiment, the effects of K residues (in the presence of basal P and N) on yield are investigated. Soils are sampled regularly to follow changes in plant-available P and K.
Taken from the Rothamsted Guide to the Classical Experiments 2006, pages 35-37.
Design: Four blocks of four plots (total of 16 plots, each 300m2) each testing the effects of permanent grass, rotational grass, arable and bare fallow, were established on a long-term grassland site (>100 years). Two blocks are in winter wheat test crops each year whilst the other two are in treatment crops. Yields are no longer taken, but soils are taken about every 5 years. In 2008 two plots within each block were used to establish the Highfield Conversion Experiment (see below).
Strengths: It is the only site at Rothamsted available for studying the effects of converting long-term grass to arable and bare-fallow treatments, on SOC and fertility. It provides contrasting comparisons with the Fosters Ley-arable experiment where similar treatments were established on a long-term arable site and the Woburn Ley-arable experiment. Many archived samples and data are available.
For more details and references, see Fosters Ley-Arable, below.
Design: Four blocks of five plots (total of 20 plots) each testing the effects of permanent grass, rotational grass, arable and bare fallow, were established on a long-term arable site (>100 years). Two blocks are in winter wheat test crops each year whilst the other two are in treatment crops. Yields are no longer taken, but soils are taken about every 5 years.
Strengths: The experiment provides comparison with the Highfield Ley-arable experiment where similar treatments were established on a long-term grassland site and the Woburn Ley-arable experiment. Many archived samples and data are available.
Background: The Rothamsted Highfield and Fosters Ley-Arable experiments started in autumn 1948. The two sites have the same soil type but the cropping histories of each are very different. Highfield had been in permanent grass since 1838 (Lawes & Gilbert, 1885); on this site some plots stayed in permanent grass, others went into continuous arable cropping and some alternated between leys and arable. In contrast, Fosters has been in arable cropping for several centuries; on this site some plots stayed in continuous arable, some went into permanent grass and others alternated between leys and arable. Although yields are no longer measured, SOM continues to be monitored. Even after more than 60 years, SOM contents on these experiments have still not reached new equilibria. In soils ploughed out of permanent grass, SOM is still declining whilst it continues to increase in soils under permanent grass.
Background taken from the Rothamsted Guide to the Classical Experiments 2006, page 40.
Key References for both Highfield and Fosters Ley-Arable experiments:
In 1959 an area of permanent grass (since 1838, Lawes & Gilbert, 1885) adjacent to the Highfield Ley-Arable experiment at Rothamsted was ploughed, and has not grown a crop since. This is the Highfield Bare Fallow. It is kept free of weeds by frequent cultivation, but herbicides are used occasionally, so inputs of carbon to the soil are negligible. Soil organic carbon (SOC) is measured periodically, and this has declined substantially since the area was first ploughed out of grass. Archived soil samples are available.
Design: In 2008 two plots (continuous grass and arable) within each block of the Highfield Ley-Arable experiment were split to establish grass, arable and bare-fallow treatments within each. Adjacent long-term bare-fallow areas were also included in the experiment. Soil samples are taken regularly and yields are taken annually. Many archived samples are available. Baseline soil measurements before the Conversion are given in Hirsch et al (2009); changes that have occurred since the Conversion are described in Hirsch et al (2017).
Strengths: It is the only site at Rothamsted available for studying the effects of converting long-term grass to arable and bare-fallow treatments on SOC, fertility, soil resilience and recovery. The experiment has support from several groups within Rothamsted, providing a resource for multidisciplinary research, from analytical chemistry, soil physical structure, GHG emissions and nitrification potential, microbial and mesofauna community responses to management change. It is probably unique in that it examines soil resilience and restoration through plants.
Link to additional data for Johnston et al (2017), European Journal of Soil Science.
Current design: Five year rotation of three years in arable crops or grass leys followed by two years of test crops (wheat and barley). Total of 80 plots divided into 5 blocks. Yields are recorded every year and one block is soil sampled each year. One block limed each year.
Strengths: Cropping data and yields available since 1938. Soil samples available from 1938, and then every 5 years since the mid-1950s. Only long-term ley-arable cropping experiment based on the sandy loam at Woburn. Provides a comparison with the two ley-arable experiments at Rothamsted.
Background: started in 1938 to compare the effects of rotations with or without grass or grass-clover leys on SOM and the yield of two arable test crops. Soils at Woburn that have been in continuous arable cropping since 1876 contain about 0.9 % C, and %C is still declining slowly; soils that have alternated between 3-year leys and 2-years arable since 1938 contain about 1.2 % C. Typically, where no fertilizer N is applied, grain yields of the first wheat test crop are greater following grass leys than in the continuous arable sequence because more N is available from the mineralisation of SOM. Following grass-clover leys, yield is increased further because of the extra N being made available from the breakdown of the leguminous residues. Following the leys, a larger yield is often achieved, with less fertilizer N, compared with continuous arable cropping.
Background taken from the Rothamsted Guide to the Classical Experiments 2006, page 41.
The Market Garden experiment started in 1942, originally to look at the effects on SOM and crop yield of various organic inputs; namely FYM, compost and sewage sludge. The experiment was grass from 1974 to 1982. When concerns were expressed in the late 1970s about the heavy metal content of sewage sludges being applied to agricultural land, the experiment was “reactivated” to examine the fate of metals that had been applied in the sewage sludge between 1942 and 1961. More recently, the experiment has been used to evaluate the effectiveness of hyperaccumulator plants, i.e. plants that can naturally accumulate large amounts of metals from soils and which, potentially, could be used to “clean up” soils contaminated with heavy metals. Part of the experiment was physically moved in 2005 and in 2006 the whole experiment was sown to grass.
Taken from the Rothamsted Guide to the Classical Experiments 2006, page 42.
The Woburn long-term sludge experiments were started in 1994, at Butt Close field, to study the effects of heavy metals contained in sewage sludges on soil fertility and microbial activity. They are part of a network of nine sites throughout the UK.
Design: Three experiments in Butt Close field: 1) sewage cake experiment (CS/427), 2) liquid sludge experiment (CS/428), and 3) metal salts experiment (CS/439). Details can be found in Gibbs et al (2006 a,b) and Chaudri et al (2008).
Experiment CS/427: Five types of sewage sludge cake were applied, (including place of origin):
Sludges were applied in two different ways, as short-term (ST) treatments applied in four large applications between 1994 and 1997 or as long-term (LT) treatments of annual applications between 1994 and 2014 of approximately 1/25th of the amounts of the same sludges to realize the same total dose of metal application as in the ST treatments in 25 years.
Note that U and D refer to the original sludge treatment process. After 1997, these sludges were air-dried and stored for use during the long-term treatment period. A control receiving no sludge was included in the experiment. Treatments randomised in 3 blocks 6 x 8 m in size.
Experiments CS /428 and /439: see below, and Gibbs et al (2006 b) and Chaudri et al (2008).
Strengths: CS/427 enables comparison of different sewage sludge processing and metal concentrations and period of application (short term versus long-term) at four rates (ST treatments) up to EU limits for Zn, Cu and Cd in soils. All had the same amount of organic matter added. Wheat is grown and mineral fertilisers added every year. Samples of the sludge used and periodic soil and plant samples are available in the Rothamsted Sample Archive. This experiment is replicated on other soils in Gleadthorpe, Rosemaund, Bridgets and Hartwood.
CS/428 and 439 are small plots (3.5 x 1.2 m) surrounded by boards for comparison of additions of the bulky ex-sewage works sludge in CS/427 with respectively: a) small amounts of organic matter but the same amounts of metals in a specially made liquid sludge applied 1994-1997, or b) the same amounts of metals added as salts (no organic matter added).
Description provided by Professor Steve McGrath, Feb 2016, sponsor of these experiments.
Current design: A five year arable rotation (Wheat, Maize/C Crop, Rye, S.Barley/C Crop, Beans) with different organic amendments (FYM, Straw, Compost, None, permanent grass/clover leys). Total of 32 plots divided into four blocks. Plots are split so that N can be applied at six rates for all crops, except beans which receive no N. Yields are recorded each year and soils are taken every 5 years. Total plot size 8.83m x 8.0m Blocks 1, 2 and 4; 7.83m x 8.0m Block 3. N sub-plots are 4m wide.
Strengths: Cropping data, yields and soils available since 1964. The experiment is randomized and has adequate replication. It is the only long-term organic manure experiment on the sandy loam at Woburn and the only long-term experiment including cover crop treatments.
Background: Started in 1964 on the sandy loam soil at Woburn to test the effects of different types of organic matter inputs on soil organic matter (SOM) and crop yields. Initially six organic treatments (FYM, peat, straw, green manures and two grass leys) were compared with two fertilizer-only treatments. Arable crops were grown in rotation with an eight-level N test 1973-1980 to assess the effects of the increased levels of SOM achieved by the organic amendments. During this period, no organic manures were applied. There was another treatment phase from 1981-1986, when further organic manures were applied. Again, SOM increased with the organic treatments and the grass leys but continued to decline slowly where only fertilizers were applied. This treatment phase was followed by another test phase, 1987-1994, when six rates of N were tested on arable crops, and no further organic manures were applied. From 1995-2002 arable test cropping continued but only two rates of N were tested. In 2003 another treatment phase started. All plots, except for the permanent grass-clover leys (and beans when grown) were split to test six rates of N on arable crops grown in rotation.
Design: Duplicate straw incorporation experiments were established at Great Knott III, Rothamsted and Far Field I, Woburn in 1987; both are in continuous winter wheat: Rothamsted – Sixteen plots (each 40m2) in four blocks with different rates of straw incorporation (0, 1, 2 & 4 times normal straw yield). Woburn – twelve plots (each 43m2) in three blocks with different rates of straw incorporation (0, 1, 2 & 4 times normal straw yield). Yields of grain and straw are taken each year, and soil has been sampled after 7, 11 and 22 years of contrasting straw treatments.
Strengths: These experiments provide a comparison of soil organic matter dynamics in contrasting soil types, at Rothamsted (a silty clay loam) and at Woburn (a sandy loam). The experiments have been used recently for studies of soil structure and earthworm activity (see Sizmur et al, 2017 in Key Referenes below). The findings at Rothamsted can be compared with those from section 0 on Broadbalk where straw is also incorporated.
Yields will not be taken from these experiments in 2016 or 2017, but they will continue as a resource to be sampled. After harvest 2017 both experiments will be discontinued.
Design: Continuous maize and barley plots with different amounts of maize tops incorporated. Total of 18 plots divided into three blocks. The experiment is duplicated at Rothamsted on a silty slay loam and Woburn on a sandy loam.
Strengths: The experiment has good replication. The experiments provide a comparison between different soil types (silty clay vs sandy loam). Maize is incorporated every year so that it has contributed significantly to the soil organic C (SOC) stock The incorporation of C4 maize residues provides an opportunity to measure the contribution to the SOC from the maize and evaluate its turnover using 13C stable isotope techniques. The plots are large (225-300m2) so could be divided to include additional treatments.
Background: The experiments were started in 1997, one at Rothamsted (Hoosfield) and the other at Woburn (Stackyard). There were six crop and straw treatments: 1. Continuous maize, stubble incorporated; 2. Continuous maize, stubble plus 10t maize tops incorporated; 3. Maize after three years of spring barley, straw removed; 4. Spring barley after five years maize, stubble incorporated; 5. Continuous spring barley, straw removed plus 10 t maize tops incorporated; 6. Continuous spring barley, straw removed. Crop yields were taken each year and soils were collected in 1997, 2008 and 2015. Maize and spring barley were chosen as crops with contrasting delta 13C enrichment in their residues. This can be used to follow the fate of the C incorporated in the crop residues. The experiments were established initially for use in SOM fractionation studies, but staff and analytical funds are no longer available to support this work, so the experiment was discontinued in 2015.
An area on the North End of Hoosfield, with a soil pH gradient ranging from 3.7 to 7.8 (0-23cm), due to uneven applications of chalk in the 19th century. Spring barley was grown continuously for over 100 years. It is now sown to winter wheat each year, given only 100 kg N ha-1. The wheat starts to die out about half way along the plot, when the pH is below 5.5. This area has been used to study the relationship between soil pH, and microbial ecology and nutrient dynamics.
Design: At Rothamsted a 6m2 area located within a larger plots (10 x 10m) on the south-east corner of Park Grass is used to monitor nutrient concentrations in soil solution under long-term grass at 10 and 50cm depth. Soil nutrient analyses from plot 3d are also included as part of the ECN data. In addition, weather variables are monitored on site using an automated weather station on an hourly and daily basis. Dry and Wet deposition of potential pollutants (SO4-S, NO3-N etc) and pH of rainfall are monitored at both sites (Rothamsted and North Wyke). Surface water chemistry of the river Ver is monitored at Rothamsted. Many of these variables are monitored on a weekly basis at both sites. Vegetation surveys and insect/animal surveys have been done regularly.
Strengths: The Rothamsted ECN (Environmental Change Network) forms part of a wider network of twelve terrestrial sites in the UK in which common variable are monitored regularly. Key environmental variables have been measured on site since 1992 and the data is available to researchers at Rothamsted and externally.
For further details, see the Environmental Change Network at Rothamsted.
Two long-term liming experiments were started in 1962 at Rothamsted (Sawyers field) and Woburn (Stackyard) to study the interaction between soil pH, P and K on crop yields. The fields differ in soil type. Rothamsted is a flinty silty clay loam (Batcombe-Carstens Series) and Woburn is a sandy loam (Cottenham Series). Different amounts of lime (including zero) were applied from 1962 onwards to fields that had been acid for many years before the experiments began (Bolton, 1970). The original experimental design was four lime treatments with and without P and K (4 x 2 x 2 factorial with two replicates) within randomised blocks and with plot sizes of 6 m x 8 m. The experiments were modified in 1981 with 4 lime treatments and 4 P levels, but again only two replicates. One of the lime and one P treatment continued as “zeros”.
Yields of various crops were recorded from 1962-1996 and papers published (Bolton, 1976, 1977 a,b; McEwen et al, 1990). Soil samples have been used for experiments determining the effects of pH on the bioavailability of heavy metals (Sanders et al, 1986). Lime was applied in one year (1962) and then in small doses between 1982-1987. The plots were also used as part of a dataset to produce the Rothamsted Lime Model (RothLime; Goulding et al, 1989) which was used from 1982 onwards to apply small amounts of lime to keep the plots at the target pHs (1:2.5 soil:water) of 4.5, 5.5, 6.5 and 7.5. Since 1997 the Rothamsted plots have been in grass; the Woburn experiment is now discontinued.
See cropping details (pdf) for type and variety of crop grown each year, 1962-1996.
See data availability (pdf) for details of crop yield and pH data available, 1962-1996.
See amount of lime applied (pdf) for details of lime application dates and amounts applied, 1962-1996.