At first glance the complexity of Rothamsted Research’s new long-term large-scale rotation experiment is bewildering.
Two cultivation systems, three rotations, two input regimes, each replicated twice in 60 large 24-metre square plots on two sites with different soil types.
If that was not enough, unlike some large rotational or cultivation trials where one of the variables is often set in the field each year or at least blocked in an area, each combination of cultivation, rotation, nutrient and pesticide input is carried out each year and fully randomised in the field.
Funded by the Lawes Agricultural Trust, with some support from Biotechnology and Biological Sciences Research Council Institute Strategic Programme Grants, the aim is to look at different farming systems from a wider standpoint than just yield or economic profitability, says Prof Jonathan Storkey, an agroecologist at Rothamsted Research.
Farming systems He says: “Many current farming systems are based on relatively few profitable crops and the technologies available to grow them, such as inorganic fertilisers and crop protection products combined with application techniques, such as precision agriculture.
“While those technologies have delivered great gains in terms of productivity, yield efficiency and profitability, you could argue by focusing on these outcomes we have created blind spots or unintended consequences, such as the loss of biodiversity, declining soil health and the evolution of resistance to pesticides.
“Some of these blind spots are now threatening the sustainability of the systems.” That is why, Prof Storkey says, we need to design systems which are more resilient from an environmental standpoint as well as economically.
Testing these systems needs a different approach as well, he adds.
So unlike, perhaps, Rothamsted’s most famous long-term experiment, Broadbalk, which looks at a single technological innovation – fertiliser inputs – and a single outcome, winter wheat yield, the new large-scale rotation experiment looks at how combinations of management interventions, combining technology and nature-based practices perform.
The starting point for the new long-term trial being conducted at Rothamsted’s main farm site in Harpenden and at its site at Broom’s Barn, near Bury St Edmunds in Suffolk, is three rotations, says Dr Aislinn Pearson, an agricultural ecologist at Rothamsted Research, who coordinates the scientific measurements for the project.
“The three-year rotation is wheat, wheat, oilseed rape – our business as usual.
It’s what you would grow, if you could, to make the maximum return.” The five-year ‘agronomic optimum’ rotation follows wheat with a spring legume, currently spring beans, followed by wheat, spring barley and then oilseed rape.
Finally, in the ‘radical rotation’, wheat is followed by spring barley undersown with a grass clover ley, two years of the ley, wheat, an alternative spring crop – sugar beet at Broom’s Barn, linseed at Harpenden – and a spring legume.
Dr Pearson says: “In this rotation, we wanted diverse cropping and potentially crops that don’t currently feature heavily in the British portfolio.
The hope was that the spring legume would be soya beans, but we’ve had a problem with pigeons so until we resolve that it will be spring beans.” The easiest way of thinking about this experiment is that every plot is like a field in the rotation, adds Dr Pearson.
“Every year, all the crops in every rotation get planted in both ploughed and zero-tillage plots.” This set-up is then repeated twice to compare the farm standard pesticide regime with an integrated pest management (IPM) approach, which aims to gradually reduce inputs while maintaining yields (see panel, right).
On top, half of each plot receives an organic amendment of compost either once or twice during the rotation.
This is combined with cover crops before the spring crops in the five- and seven-year rotations.
In total that means there are 24 different systems being tested – 48 if you consider the different soil types at the farms.
All this complexity means it will take 105 years for the trial to return to the initial distribution of crops, says Andrew Mead, head of statistics at Rothamsted, though analysing the data from around 15 years should provide a good assessment of the different treatment combinations.
On each site, each plot is unique every season, but the clever statistical design means there is hidden replication which provides statistical power for the comparisons of the different treatment components.
“It is complex to analyse, but the design means the different combinations occur across the trial, so we can look at the individual effects of rotation or cultivation across the other treatments as well as in combination with the other components,” he says.
Settling in phase
The long-term nature of the trial means there is a settling in phase for the plots.
“The first few years are not fully representative of consistent management practices, so we will probably need up to four years of data before we can be confident of the results.” Assessments for the project are intended to be wide-ranging, says Dr Pearson.
“Yes, we are looking at yield, but the dream for this experiment is that you can monitor more or less everything that is relevant to an agricultural site.” The wish list is long (see panel, left) and not everything will be measured from the start or each year.
“In the long-term the ideal would be to monitor multiple aspects of the plots,” says Dr Pearson.
The results will be disseminated through a number of different channels.
As a scientific institute, research papers are a focus for the Rothamsted team, but they have also set up a Twitter channel @Rothamsted_LSRE, which has already provided some interesting discussion between the researchers and growers and will be potentially considering demonstration days in the future.
“The final important pillar is using the results to provide evidence for future policy,” Dr Pearson adds.