The rothamsted experience

New approaches to soil monitoring are building on a historic legacy

There can be few sites in the world that have witnessed more soil monitoring than Rothamsted in Hertfordshire, UK. 

For 180 years scientists have been collecting and analysing soil samples from the same field where different crop plots are given varying fertiliser treatments. This makes the Broadbalk site at Rothamsted the longest running field experiment in the world.

Through astonishing foresight, the founders of Rothamsted, the Victorian philanthropist John Bennet Lawes and the chemist Henry Gilbert, kept soil and plant samples from nearly every plot they tested. The Sample Archive at Rothamsted now has almost a third of a million items, representing a continuous data set unmatched anywhere else. 

This gives Rothamsted’s scientists a priceless resource with which to study how soil quality changes over time. Old samples are now being analysed with modern techniques to provide new insights into the effects of continuous fertiliser use on soil structure and biology.

A solid legacy

Lawes and Gilbert were systematic in how they set up the experimental plots and in analysing the samples collected. A control strip with no applications at all has been maintained throughout the duration of the experiment. Likewise, an “organic” strip that has only had additions of farm-yard manure has been continuously present. Other plots have had differing applications of nitrogen, phosphorous, magnesium and potassium in an attempt to better understand the nutritional needs of plants. 

Some of the long-term findings from these efforts have been highly influential in modern thinking about soil health and security. Whilst there is currently a lively current debate about how to protect and boost levels of soil organic carbon, Lawes and Gilbert’s studies have revealed that the carbon content of some soils in the experiment has changed little in more than a century after they were first measured in 1865.

Other experiments have been used to respond to contemporary challenges in agriculture such as nitrate pollution and climate change. For instance, detailed work on nitrogen cycling has been carried out using a stable isotope of nitrogen applied to microplots within the experiments. Likewise other studies have focussed on the soil’s ability to act as a sink for methane, an important greenhouse gas.

A world-wide network

The experiments at Rothamsted have inspired similar studies elsewhere. The Morrow plots at the University of Illinois, for instance, were started in 1876. Other research institutes have followed suit.

This global proliferation led to the creation of the Global Long-Term Agricultural Experiment Network (GLTEN) in 20191, which brings together long-running experiments that span nearly two centuries and six continents, as well as representing numerous climates, environments, crop types, farming practices and land-management regimes.

The researchers responsible for these long-term experiments – defined as experiments running for at least a decade – have made their information openly available on the network’s website to help other scientists discover their studies and foster further collaborative work. 

GLTEN represents a potential treasure trove of information – over 1750 years’ worth of data in total – that will help researchers and policymakers design “the farms of the future”. The 65 sites span the globe, with about 20 in the Americas, a dozen or so in Africa, more than 10 in Europe and several others across both Asia and Australasia. 

"these long-term experiments are a really important global resource for informing future farming"

“The hope is that lessons learnt in one country might improve practices elsewhere – resulting in natural resources being used more efficiently, and in a way that produces a food supply that delivers a nutritionally balanced diet,” says Rothamsted Dr Jonathan Storkey, who helped set up the Network.

The Rothamsted Experience

“We also hope this initiative will help us uncover ‘hidden’ long-term experiments that we didn’t know about, enabling us to mine and analyse their datasets and insights.

A good example of the value of long-term experiments is our understanding the effects of man-made fertiliser use, a practice that began in Europe during the Victorian era. Fertiliser experiments that started in the UK in the 1800s have helped chart the long-term impacts of this switch not just on crop yields, but also soils, water, wildlife, human-health and climate.

“These long-term experiments are a really important global resource for informing future farming,” says Storkey. “Agriculture will face the massive challenges of feeding a growing world population without relying on the conversion of remaining natural habitats to agriculture, and all the while, reducing other impacts on the environment including the release of greenhouse gases and losses of biodiversity.”

From chemistry to biology

Much of Lawes and Gilbert’s global influence derives from their early studies focused on the chemistry of fertilisers and its relationship to plant health. However, more recently scientists have increasingly begun to realise that biology of the soil also plays a critical role. Fortunately, genetic information retained in Rothamsted samples that are now decades old can still be easily analysed and sequenced, providing new information on how the microbiome under our feet is changing.

One recent study has led to a radical new way of thinking about why adding organic material like manure to soils improves flood and drought resilience, climate control and crop yields – universal ‘ecosystem services’ that are widely recognised as worth billions to the global economy.

Using over 50 years’ worth of data from the long-term field experiments, researchers have demonstrated that common farming practices drain the soil of carbon, altering the structure of soils’ microscopic habitat and, remarkably, the genetics of microbes living within it.

The Rothamsted Experience

The team of microbiologists and physicists considered almost 9,000 genes and used X-ray imaging to look at soil pores smaller than the width of a human hair, and in concert with previous work, have started forming what they envisage will be a universal ‘Theory of Soil’.

In healthy soils, relatively low nitrogen levels limit microbes’ ability to utilise carbon compounds, so they excrete them as polymers which act as a kind of ‘glue’ – creating a porous, interconnected structure in the soil which allows water, air, and nutrients to circulate. 

The study team believe that the Victorian-era switch from manure to ammonia and phosphorous based fertilisers has caused microbes to metabolise more carbon, excrete less polymers and fundamentally alter the properties of farmland soils when compared to their original grassland state.

Lead researcher Professor Andrew Neal says: “We noticed that as carbon is lost from soil, the pores within it become smaller and less connected. This results in fundamental changes in the flow of water, nutrients and oxygen through soil and forces several significant changes to microbial behaviour and metabolism. Low carbon, poorly connected soils are much less efficient at supporting growth and recycling nutrients.”

"low carbon, poorly connected soils are much less efficient at supporting growth and recycling nutrients"

The closed soil structure also means microbes need to expend more energy on activities such as searching out and degrading less easily accessible organic matter for nutrients.

Conversely, in carbon-rich soil there is an extensive network of pores which allow for greater circulation of air, nutrients and retention of water.

“Manure is high in carbon and nitrogen, whereas ammonia-based fertilisers are devoid of carbon. Decades of such inputs – and soil processes typically act over decades – have changed the way soil microbes get their energy and nutrients, and how they respire.”

The Rothamsted Experience

Turf wars

The team followed up this study with a deeper dive into the traces of the microbial ecosystem preserved in the historic samples. They likened the impact of arable farming on soil ecosystems to creating a “gangsters’ paradise” for certain soil organisms.

This new study showed that common farming practices such as ploughing, fertilising and adding pesticide to fields results in a chaotic new (microbial) world order where nitrogen stealing archaea and killer fungi have muscled their way in at the expense of many plant beneficial fungi.

The research also found the more ‘nutritionally monotonous’ arable soil environment has led many bacteria to ‘reduce their running costs’ by jettisoning more than 600 of the genes usually needed when faced with the diverse range of food sources found in grasslands or pastures.

This means that typical measures of soil biodiversity aren’t adequately explaining what is going on in soil – as farming not only changes the number and relatedness of the species present, but also the genetic complement of the community as a whole.

Using soil samples from a long running Rothamsted plot experiment, the team together with the US Department of Energy’s Pacific Northwest National Laboratory compared arable soils with their original grassland state, as well as bare soils that have been left fallow for over 50 years.

According to Professor Neal, when grassland is converted to arable, the richness of species doesn’t change very much, but new species move in, and they don’t necessarily fill the same ecological roles.

Compared to their original grassland state, arable soils had fewer, but more varied, species of fungi. Mycorrhizal fungi, those that form mutual beneficial associations with plants and play important roles in plant nutrition, are reduced in favour of pathogenic fungi that survive by attacking insects, plants and lichen.

The Rothamsted Experience

The researchers also saw a greater variety of bacteria in arable soil, whilst the total number of species of archaea – a group of single cell organisms members of which generate the greenhouse gas nitrous oxide as a by-product of ammonia oxidation – also increased in response to fertilisation.

The results also show that the responses of these three different types of organisms varied markedly depending on the physical and chemical challenge presented by farmers.

“Farming practices cause physical disruption and alter the nutritional inputs, which means less diverse plant materials and more readily available nitrogen to soil,” says Prof Neal. “As a result, some species lose out allowing new ones – often with very different ways of making a living – to thrive. Even those that survive have had to change the way they live their lives.” 

The move from grassland to either bare soil or arable land impacted the genomes of bacteria and archaea, with organisms in arable soil having approximately 650 fewer, and those in bare soil having about 1,300 fewer genes, compared to organisms in grasslands.

“Bacteria have a habit of losing genes if there is not a good reason to keep them,” says Neal. “It’s a ‘use it or lose it’ situation. What we think is happening here is they no longer need the functions these genes code for, and the genome length is reduced as a consequence.”

Signals in the soil

Ultimately it is not the biology, chemistry or structure of soil alone that determines its viability and performance, but the interaction of all three. As Professor Neal likes to put it, “soil is a process”. So, measuring multiple facets of a given soil in-situ will be critical to our future understanding of soil health and sustainability. 

"when buried optical-fibre cables are stretched or compressed which enables detection of the changes in strain and temperature"

Consequently, Rothamsted has been fast developing its monitoring capacity across a range of soil attributes. One initiative is to develop sensor systems that measure gas concentrations in the ground and match these with strain, moisture, temperature and suction readings at different scales. This will enable scientists to provide data on the dynamics of gas flux and soil structure. 

The monitoring systems deploy new technologies in innovative ways. A good example is a distributed fibre optic sensor (DFOS) system. Originally developed for measuring strain in built structures, it can be adapted for soils and provide measurements over several metres – potentially even kilometres. When buried optical-fibre cables are stretched or compressed by surrounding soils, local anomalies are created in the fibre which enables detection of the changes in strain and temperature in the surrounding environment.

The Rothamsted Experience

The technology is currently being further developed at UC Berkeley in California, who are a partner in the monitoring project, their team have been trialling the sensor systems in the unique wind tunnel-soil facility available at the Colorado School of Mines (CSM). An experimental set up has also been installed at Rothamsted.

The experiment will manipulate soil moisture fluctuations by balancing water introduction through rainfall and losses to evaporation and evapotranspiration as controlled by atmospheric factors such as temperature, wind speed, and relative humidity. This will allow the team to make more informed predictions of soil structure changes and greenhouse gas emissions under changing climate conditions.

Once the set-up has been fully tested the researchers will be able improve the predictive understanding of how changes in atmospheric carbon are affected by soil structure changes. 

The proposed sensor development and experimental research will lead to a substantial improvement of soil carbon models such as the RothC model developed at Rothamsted in the 1990s.

Low tech solutions

High tech soil monitoring will undoubtedly be an important tool in the development of high yield, low input future farming systems. But in the shorter term more accessible and easier to use measuring systems will be needed.  

A good example is a rapid soil-phosphate measurement toolkit developed by Rothamsted.

Phosphorus application represents a particular challenge for modern agriculture. Whilst vital for healthy plant growth, globally phosphate sources are limited. Moreover, if phosphate is poorly applied it can end up polluting watercourses causing significant environmental damage. 

And it is costly. Current instability in agricultural input markets saw phosphate fertiliser prices increased by as much as 130% in 2022, according to the UK’s AHDB.

So more precise application would benefit farmers hugely. Moving from prophylactic over-use to more precise, targeted applications of phosphate would reduce the volumes and avoid unnecessary depletion of this finite natural resource. And a more measured approach would directly benefit farmers: cheaper, more frequent and accurate soil-testing would mean that they use fertiliser only where and when needed, resulting in increased yields, less work and lower costs.

Rothamsted Research produced a prototype rapid soil-phosphate measurement toolkit called Phosfield. This will enable cheaper, more frequent testing as well as more precise fertiliser applications, maximising yields and reducing pollution. 

Funded by the ERDF’s Agri-Tech Cornwall programme, the test kit provides precise results within just 20 minutes – a massive improvement on the several-day turnaround by laboratories.

'Most farmers test their soils for phosphate every three to five years,' explains Dr Susan Tandy, who until recently was a soil scientist at Rothamsted Research. 'They usually take several samples from across the field and amalgamate them to get an average reading.'

The level of phosphate varies across fields, and more accurate GPS-located testing would enable farmers to apply fertiliser at variable rates allowing them to achieve more consistent yields.

More importantly, phosphate availability can change over time and depends on the soil type, so by testing more frequently and knowing the soil type, farmers can be even more accurate in their fertiliser application.

The test has been three years in development, and has been trialled in Ghana, where it could have significant benefits.

'The technique would be extremely useful in developing countries as they have limited lab access to test their soils, meaning the application of expensive fertiliser is both financially risky and may not match crop requirement,' explains Tandy.

Having nailed down the scientific process using Cornish soil samples, the researchers worked with product design and development firm Vital Spark Creative to produce an analytical kit that would be relatively easy to use in the field.

While precision farming techniques like soil and crop scanning sand conductivity tests enable variable rate nitrogen applications, analysing phosphate will likely always require a physical soil sample to be taken, says Tandy.

The test is also extremely cost effective, according to developers. Once the kit is acquired, each test costs pence rather than pounds for a laboratory analysis.

Although the tool is not yet commercially available, the Phosfield team are seeking additional funding to bring it to market and hope to undertake further research to produce tailored fertiliser recommendations for different crops and soil types.