A Bayesian Belief network to operationalize the concepts of Soil Quality and Health
Lead Research Organisation:
Rothamsted Research
Department Name: Sustainable Agriculture Sciences-H
Abstract
'Soil Quality' and 'Soil Health' are general terms for indicators that are associated with 'Soil Security'. None of these terms within quotation marks is easy to define, however. Neither are they easy to quantify rigorously in a way that avoids dispute. Nonetheless all three terms have traction with policy makers and with land managers and regulators. Indicators provide benchmarks for ranking different places or practices and deciding where to deploy effort to bring about change as effectively and economically as possible and they provide a means to assess afterwards whether or not and to what extent this change has actually been brought about.
As a result, indicators of this kind are attractive to stakeholders. Indicators often rely on expert opinion for their derivation, but experts differ. Even apparently objective biophysical measurements are subject to error and worse, the soil itself varies from place to place and even time to time. It is not clear how to eliminate bias or how to weight the different kinds of information - opinion and measurement.
There is therefore scope for developing a rigorous, scientific approach to SQH that incorporates expert-derived opinion alongside physically-based measurements in our understanding of Soil Quality and Health (SQH) in a scientific manner.
Bayesian Belief Networks are graph-based, directional networks that can incorporate probability distributions of these various kinds of data. Essentially the directedness leads from multiple pieces of data to a conclusion - in our case a rating of SQH. The network is self-learning in that any additional soils and data for which quality assessments are available will re-inforce the pathways that decide the quality rating. In use, SQH ratings for additional soils that contain even partial data can still be obtained if the net defaults to mean values where data is missing.
To accommodate the various functions and scales needed to operationalise SQH, will require a set of Bayesian Belief Networks that considers the interactions of soil properties with SQH but also the impact of environmental change and land use and management on soil quality. There a numerous advantages to using BBNs: they can consider and integrate biological, economic and sociological factors and have effectively been use to determine the consequence of land-management decisions in land use decision behaviour. Bayesian modelling methods are a rigorous framework in which a complete characterization of the coupling and variability of soil quality is based on physical laws, empirical relationships but can easily incorporate expert knowledge formally and other kinds of soft data.
As a result, indicators of this kind are attractive to stakeholders. Indicators often rely on expert opinion for their derivation, but experts differ. Even apparently objective biophysical measurements are subject to error and worse, the soil itself varies from place to place and even time to time. It is not clear how to eliminate bias or how to weight the different kinds of information - opinion and measurement.
There is therefore scope for developing a rigorous, scientific approach to SQH that incorporates expert-derived opinion alongside physically-based measurements in our understanding of Soil Quality and Health (SQH) in a scientific manner.
Bayesian Belief Networks are graph-based, directional networks that can incorporate probability distributions of these various kinds of data. Essentially the directedness leads from multiple pieces of data to a conclusion - in our case a rating of SQH. The network is self-learning in that any additional soils and data for which quality assessments are available will re-inforce the pathways that decide the quality rating. In use, SQH ratings for additional soils that contain even partial data can still be obtained if the net defaults to mean values where data is missing.
To accommodate the various functions and scales needed to operationalise SQH, will require a set of Bayesian Belief Networks that considers the interactions of soil properties with SQH but also the impact of environmental change and land use and management on soil quality. There a numerous advantages to using BBNs: they can consider and integrate biological, economic and sociological factors and have effectively been use to determine the consequence of land-management decisions in land use decision behaviour. Bayesian modelling methods are a rigorous framework in which a complete characterization of the coupling and variability of soil quality is based on physical laws, empirical relationships but can easily incorporate expert knowledge formally and other kinds of soft data.
Planned Impact
Understanding and estimating soil quality and health is an issue of some national and international importance as was highlighted during the 2015 International Year of the Soil. Soil health and quality are vague concepts, yet much-needed by policy-makers and planners to inform decision making, both in the UK and globally. Most of our food derives from terrestrial agriculture, so decisions and policies that preserve or enhance the quality of our soils are of tremendous strategic importance.
Many stakeholders have expressed interest in or have funded work on indicators of soil quality. Our research should be of interest to a wide range of parties: AHDB, LEAF, food retailers and NFU for the farming industry, Defra, EA, SEPA, SG, Natural England for regulation and environmental issues, the EU and other policy and regulatory bodies overseas too have funded work on SQH and will value a robust and self-consistent approach to quantifying SQH.
A joined-up series of indicators or at least methodology for doing so could allow policy-makers and regulators with different fields of management the ability to refer and compare one anothers' needs with their own. Benchmarks and indicators of change would also be a step forward for regulations.
These tools could potentially be used by land managers too who have to combine different kinds and qualities of information in order to meet differing objectives from their land. Rules, structures and guidance for how to do this and how not to would be valuable.
Many stakeholders have expressed interest in or have funded work on indicators of soil quality. Our research should be of interest to a wide range of parties: AHDB, LEAF, food retailers and NFU for the farming industry, Defra, EA, SEPA, SG, Natural England for regulation and environmental issues, the EU and other policy and regulatory bodies overseas too have funded work on SQH and will value a robust and self-consistent approach to quantifying SQH.
A joined-up series of indicators or at least methodology for doing so could allow policy-makers and regulators with different fields of management the ability to refer and compare one anothers' needs with their own. Benchmarks and indicators of change would also be a step forward for regulations.
These tools could potentially be used by land managers too who have to combine different kinds and qualities of information in order to meet differing objectives from their land. Rules, structures and guidance for how to do this and how not to would be valuable.
Organisations
Publications
Hassall K
(2019)
Facilitating the elicitation of beliefs for use in Bayesian Belief modelling
in Environmental Modelling & Software
| Description | A series of networks have been created that identify the status of soil quality and health for soils in England and Wales. These indicate better quality in semi-natural soils generally that soils under livestock or arable land uses. An opportunity map for E&W has been created which indicates which soils might be improved and by what intervention |
| Exploitation Route | the project partners have a meeting with Defra in March 2020 to see if there is any interest in making use of the methodology developed and if so how. |
| Sectors | Agriculture, Food and Drink,Environment |
| Description | Conference Workshop session on the long-term effects of reduced tillage |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | At a session of the British Society of Soil Science, Wilfred Otten, Paul Hallett and I organised a session on the long-term effects of tillage on soils report Soil Security Workshop on: How is reduced tillage really changing soils? Rationale and workshop structure The aim of the workshop was to bring together leading researchers and practitioners to discuss the impact reduced tillage has on soil properties, and to identify research gaps that need to be addressed. Tackling this problem will require an integrated approach from experts in soil biology, chemistry, physics, plant science and agronomy. During a Soil Security workshop in September 2018, the impacts of reducing tillage on soils was highlighted as a topic of importance to research. Accordingly, we invited a series of talks from keynote speakers at an open session of the recent annual meeting of the British Society of Soil Science. This series of talks was followed immediately by a panel discussion with the wider soil community where perspectives on research gaps were debated. These ideas and others were synthesised during a workshop meeting the next day with a small number of participants to prepare ideas for a NERC highlight topic as well as additional research projects that might be tackled with smaller teams. Outcome About 90 people attended the keynote presentations on the 4th September in Sheffield. It kicked off with Russell McKenzie, who is Arable Manager of D J Tebbit/John Sheard Farm and on the AHDB Cereals & Oilseeds Board. From his Nuffield Scholar project on "Success with No-till - under any conditions", Russell provided a captivating global overview of no-till agriculture across the globe, demonstrating its success in some very challenging environments. A common theme observed in Brazil, Argentina and his own farm was greater water retention. From an Argentinian experiment, he reported an increased window of 7-10 days crop water availability in no till versus conventional cultivation with ploughing. No till combined with cover cropping was found to lead to further improvements, but yield benefits and nutrient returns take 2-3 years to establish. With more extreme weather and degrading soil, Russell argued that no-till coupled with a managed flexible approach to soil cultivation produces medium to long-term benefits of reduced input costs and weed dynamics, making farming more profitable. Alistair Leake, from the Game & Wildlife Conservation Trust's Allerton Project, presented their over 30 years of experience with different tillage systems. Over 45% of UK arable farming is under reduced tillage, which is second in Europe. Reduced tillage includes zero tillage, shallow non-inversion cultivation, and strip tillage. These systems save 52 minutes/ha in labour (33% time) and reduce input costs. In addition to improved earthworm communities under reduced tillage, their work has observed increases to insects, with positive knock-on impacts to farmland bird populations. There are challenges such as compaction and slugs that can be countered by careful soil management. From one study he found inconclusive impacts of ploughing versus zero tillage on greenhouse gases, but emphasised that the dearth of evidence requires much greater research. At Agroscope, Switzerland and across Sweden, Thomas Keller has explored how soil tillage affects soil structure, productivity and farm profitability. He estimated that 4.5 billion tonnes of soil are moved by tillage in Sweden each year, but the 30 million litres of fuel that could be saved by switching to zero tillage is the same used by recreational boats. At the Oberacker field experiment, no tillage has been compared to mouldboard ploughing for 20 years. On average, crop yields do not vary between tillage approaches, with zero tillage performing better in dry years but worse in wet years. There were no observed differences in carbon stocks stored in soil between tillage systems. Duncan Cameron from the University of Sheffield presented work from his BBSRC and NERC research projects on grass-clover leys and soil cultivation. Leys offer huge potential to help reverse the £1.2 billion economic cost of soil degradation to the UK. After just 2 years of ley, soils were found to store 10% more water, have 7 x more earthworms and large increases in mycorrhizae fungi that help plants capture nutrients. Blair McKenzie from the James Hutton Institute presented research from the UK's oldest contemporary tillage experiments located at his institute and NIAB. He reported a barley cultivar specific response to soil tillage, suggesting that modern breeding on ploughed soils may not be optimising crop selection for the 45% transition to reduced tillage in the UK. Like the long-term experiment in Sweden, he found no difference in carbon storage between tillage systems after 10 years, but argued that tillage is only part of the solution. Combined with cover crops and organic amendments, he thinks reduced tillage offers huge potential to sequester carbon. A positive benefit, found from economic analysis by a team at NIAB, was greater profitability under reduced tillage, despite a smaller crop yield. These five talks set the scene for a panel discussion that brought up some interesting challenges and opportunities. Organic farming was thought to offer promising approaches that improve soil sustainability, but it was felt by the panel to not always be viable. Glyphosate was viewed as a crucial tool, with mechanical weeding under organic systems having negative impacts on soils. Rather than a one-size-fits-all tillage solution, a managed approach suitable for local conditions is required. Alongside tillage, cover crops and the use of leys could improve soils through the action of their roots. However, the underpinning mechanisms and economic viability need greater understanding. A wish-list of crop traits for reduced tillage included: roots occupying the whole soil profile, better scavenging of nutrients, learning from older varieties with vigorous root systems, root properties that deal with compaction, root screening approaches for more heterogenous soils produced by reduced tillage, and DNA markers of below-ground traits. Farmers need tools to help them select more sustainable tillage practices and crops that are suitable to this drastic change in farming practice. Policy-makers need evidence of the real benefits of changing soil cultivation practices, not just to farming but to wider environmental benefits such as biodiversity, greenhouse gas mitigation and erosion. A smaller group met the following day in a workshop to discuss research needs on the impacts of reduced tillage on soils. These ideas were formulated with input from the farming community, agronomists and specialist soil scientists, identifying both practical challenges and underpinning science that will be needed to understand the broader impacts of reduced tillage. In the arsenal of tools to address climate change challenges, land degradation and food security, changes to tillage practices and associated soil management offer considerable, untapped potential. Realising the potential of tillage, however, requires much greater understanding of how soils respond, and innovation so that farmers can change practices with confidence. The huge change towards reduced tillage in the UK is likely having a dramatic impact on biodiversity, water cycling, carbon storage, greenhouse gas emissions and farm profitability. The UK government set the ambitious target "by 2030 we want all of England's soils to be managed sustainably, and we will use natural capital thinking to develop appropriate soil metrics and management approaches." Our workshop identified many potential options, but guiding policy decisions with the existing evidence base is fraught with uncertainty. The neglected research area of soil tillage in the UK needs investment to understand the response of soils and how this affects the environment and food production. |
| Year(s) Of Engagement Activity | 2019 |
| Description | Tree of Tradeoffs - engagement with members of the public |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Public/other audiences |
| Results and Impact | As part of the Rothamsted Festival of ideas celebrating 175 years of Rothamsted, we held an event over the last weekend in June 2018 to ask members of the public about what they would like to see from Agriculture. We had a manufactured tree on which visitors were invited to hang different coloured leaves representing their 4 choices from 6 possibilities: cheap food, rural livelihoods, environment, nutritious food, farm profit or food security. Crucially visitors were limited to the 4 choices. This enabled us to engage with them and talk about what issues were more important than others. Visitors were also encouraged to write comments on the leaves that they hung on the tree. We reached over 600 people in this way, many of them children |
| Year(s) Of Engagement Activity | 2018 |
| Description | Workshop Eliciting targets for the sustainable development goals and goal 2 in particular |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Other audiences |
| Results and Impact | A workshop was convened in order to identify the views of stakeholders on the nature of agriculture in 2030 and 2050 and the ways in which agriculture might change to deliver to the UN SDGs in an full and meaningfull way as possible. Stakeholders were divided in to those representing arable, livestock or diary sectors and having identified reasonable targets by way of improving agriculture, reflected on pathways over time to reach these targets and goals |
| Year(s) Of Engagement Activity | 2018 |
| Description | Workshop on Soil Health |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Other audiences |
| Results and Impact | Views on the nature and determinants of soil quality and health were elicited from experts in the field during a 2 day workshop session. Participants were first trained with regard to the elicitation process and then introduced to the bespoke software with which we obtained their views. Networks were then constructed that expressed the interconnected way in which the components of soil quality lead to a quantitative expression of that concept |
| Year(s) Of Engagement Activity | 2018 |