Disclaimer
The views expressed in this article are those of the authors. Publication in Open Research Europe does not imply endorsement of the European Commission.
Introduction
Rural land can provide important economic and social benefits; not just through employment but through the provision of ecosystem services such as mitigation of flooding and drought, sequestration and storage of carbon and food production (Gawith & Hodge, 2019; Kristensen, 2016; van Zanten et al., 2014). However, the provision of ecosystem services is dependent on having sufficient stocks of well-functioning natural capital that are geographically close to areas of demand for ecosystem services (see Mulligan, 2021). However these stocks, and in turn the economic and social benefits they enable, are threatened by a variety of activities such as inappropriate land management, urbanisation and infrastructure development (Kristensen, 2016; Plieninger & Bieling, 2012; Poulton & Johnston, 2021; Schulp et al., 2019; Winkler et al., 2021).
Macro-level policy interventions, such as publicly funded financial payments to incentivise land management actions with a public benefit, affect the stocks of natural capital (Kristensen, 2016; Springermann & Freund, 2022; Stoate et al., 2009; Vejchodska et al., 2022). Therefore there is an opportunity to use policy to steer rural lands towards ecosystem service delivery. However careful consideration needs to be given to which services are incentivised. The historical policy drive in Europe to intensify food, and the resulting subsidies applied per area of land under production, meant that food provision was prioritised over nature conservation and other ecosystem services. Hedgerows have been removed, ponds and wetlands drained, unproductive lands agriculturalised, and fertilisers, herbicides and pesticides overused. This has led to widespread agricultural soil degradation and a simplified rural landscape with reduced biodiversity and ecological function (Gerowitt et al., 2003; Hayhow et al., 2019; Mueller et al., 2021; Stoate et al., 2009). This has undermined the sustainability of agricultural production whilst also leading to increased offsite impacts such as increased flood risk and more carbon emission.
The European Union (via the Common Agricultural Policy) and England (via the Sustainable Farming Incentive within the Environmental Land Management schemes) are exploring policies which incentivise conservation and restoration of natural capital on farmland (Defra, 2021; EC, 2022; Herzon et al., 2018; Matzdorf & Lorenz, 2010). Such approaches include action- and outcome- based measures; i.e. with payment linked to particular actions (i.e. cutting a meadow once a year after a certain date) or to the benefit provided (i.e. delivery of a species-rich wet meadow) respectively.
In this evolving policy environment, there are two key challenges for outcomes-based approaches. The first is to develop methods to identify which of the possible range of land management options or nature based solutions (NBS) work best at delivering the required outcomes in a particular location. The second is to measure how well existing land management investments deliver a variety of ecosystem services, and to do so in a simple, cost-effective and scientifically rigorous way. Environmental intelligence approaches have the potential to meet both of these challenges.
Environmental intelligence (EI) brings together multiple data streams, employing human reasoning and machine learning to better understand and manage the environment (Mulligan et al., 2021). Potential data sources include monitoring data from in-situ sensors, airborne or satellite remote sensing, as well as outputs from spatial modelling (Table 1). These data streams can be combined with artificial intelligence methods (such as machine learning), and/or human reasoning. All EI techniques have limitations which can restrict their suitability for certain applications. However, combining different techniques can broaden the range of suitable applications. For example, combining in-situ sensors with satellite data and/or spatial modelling can help improve the accuracy of both approaches: in-situ sensors provide highly locally relevant data at particular points but lack spatial coverage whereas satellite data provides spatial coverage but lacks the local relevant and temporal detail of in-situ sensing.
Table 1. Glossary of terms used in this article.
| Term | Definition |
|---|
| Ecosystem services | The benefits people obtain from ecosystems |
| Nature-based solution (NBS) | The use of nature and natural processes to provide environmental, social and economic benefits. |
| NBS investor | A person or organisation financially investing in NBS. This includes governments, commercial investors, landowners and farmers. |
| Monitoring | The process of detecting, observing and taking measurements of characteristic(s) of an area over a period of time. Monitoring can take place within the area (in-situ) or at a distance (remote sensing). |
|
In-situ monitoring | Involves placing sensors within the location of interest (such as placing sensors in the soil within an agricultural field). |
| Remote sensing | A form of monitoring where observations are made from a distance. Two key forms of remote sensing are airborne and satellite. |
| Airborne remote sensing | Measurements are taken by mounting sensors on drones or aircraft to obtain observations for specific locations. |
| Satellite remote sensing | A form of remote sensing where sensors are mounted on satellites orbiting the planet. They can be used to observe specific areas at high resolution or larger areas at lower resolution. |
| Spatial modelling | An analytical approach where geographical information is explicitly included in the model. |
| Artificial intelligence | Intelligence as demonstrated by machines (in contrast to the intelligence demonstrated by humans and animals). |
| Machine learning | A form of artificial intelligence where algorithms build a model based on a sample dataset. The algorithm can then make predictions for other datasets without being explicitly programmed to do so. |
| Counterfactual | An alternate state used as a comparison to help understand the effect of an intervention such as a policy or particular land management practice. |
In this article we introduce EI and discuss the possibilities and limitations of its use to support outcomes-based rural land management. We focus on the use of in-situ monitoring, satellite remote sensing and spatial modelling because their application in this context is further developed than airborne remote sensing and artificial intelligence approaches.
Possibilities
The importance of geography
Environmental intelligent approaches are able to incorporate geography. Geography matters because management can have different impacts within and between agricultural fields as well as across larger climate, soil, topographic and geological gradients (Soane et al., 2012; Zhao et al., 2020). This spatial variability means that 'rules of thumb' on how to manage for a particular outcome are rarely effective across such gradients. Environmental intelligence offers the chance to measure and account for this spatial variability in both planning and effectiveness assessments, through use of locally specific in-situ monitoring or area-based remote sensing or modelling.
There are two principal uses for EI in supporting outcomes-based rural land policy:
For strategic planning, spatial modelling (often incorporating satellite data), can provide evidence to help answer a variety of questions related to: i) the scale of investment required (e.g. EcoActuary Investment Planner, Mulligan et al., 2023a; Mulligan et al., 2023b); ii) the optimal location for a particular type of land management activity to occur (e.g. scenario based ecosystem service assessment tools such as WaterWorld; Co$tingNature, InVEST, ARIES, LUCI and NEVO); iii) identifying beneficiaries and potential impacts of different management activities (Table 2). Spatial modelling of this kind can therefore support NBS investors to identify the scale, types and locations of investments with the greatest 'bang per buck' i.e. those which deliver the greatest ecosystem service benefits for the least financial and opportunity costs.
Table 2. Examples of policy questions answerable with environmental intelligence.
All examples provided relate to flood mitigation to help illustrate the range of insights that can be supported.
| Policy Support | EI | Example Questions | Examples |
|---|
| Planning | Spatial modelling incorporating satellite data | What level of investment - in regards to cost of, or area covered by, a particular land management action - is required to meet a specific nature conservation, climate mitigation, flood risk mitigation or other policy goal? | Over what area would regenerative agriculture need to occur to reduce flood damages by 25% to an asset at risk downstream? |
| | Where should a land management action occur in order to have the greatest ‘bang-for-buck’? | Where is the optimal location to plant 10,000 trees in a catchment in order to minimise the financial losses at a downstream asset at risk due to flooding? |
| | Which land management activity has the greatest benefit for the least capital and maintenance costs? | Should I invest in leaky dams, retention ponds or regenerative agriculture if I want to maximise downstream flood mitigation benefit with the least capital and recurrent cost?* |
| | Who will benefit from a particular management practice? | Which buildings and other infrastructure are afforded flood protection by practising regenerative agriculture at a given farm? |
| Effective-ness | In-situ monitoring | How does this management practice perform against the counterfactual (eg conventional practice) or a particular KPI? | How much extra water is stored within soil for a particular event relative to the conventional practice, and what proportion of river flow at an asset at risk does this represent?* |
| Spatial modelling incorporating in-situ monitoring | Who are the beneficiaries of a management practice? | How many, and who downstream, benefit from the observed increase in soil water storage? |
For effectiveness assessment, in-situ monitoring can provide evidence to assess the ecosystem service benefits provided by current management practices in relation to counterfactuals or Key Performance Indicators. In-situ monitoring can be combined with satellite data and spatial modelling to evaluate other aspects of performance such as identifying beneficiaries (Table 2). Therefore in-situ monitoring alone, or combined with remote sensing and spatial modelling can provide NBS investors with a means of rigorously evaluating the outcomes achieved through specific investments.
Limitations
Difficulty in measuring outcomes
Assessing outcomes, while laudable, can be difficult to do in practice. The difficulties with outcomes-based approaches include:
Time: some outcomes take a long-time to accrue. For example, it takes a long time for carbon to accumulate in soil (Cooper et al., 2021; Oelbermann & Voroney, 2007; Smith et al., 2020). This means it is not possible to rigorously and reliably measure change in soil carbon stocks in the short-term (i.e. within 5 years).
Cost: some outcomes are difficult and expensive to measure. For example the impact of land management on biodiversity is difficult to assess as it requires a highly manual and thus relatively expensive approach.
Confounding effects: some outcomes are prone to impacts of time-varying factors beyond control - such as weather extremes or nearby changes to land use or management. This makes longitudinal (before and after) analyses of change difficult. For example, biodiversity is strongly influenced by offsite factors and climatic variability which makes robustly measuring biodiversity outcomes challenging.
Starting conditions: if outcomes are designed to reward improvements in a particular outcome (i.e. incentivise restoration) then this disadvantages those who already have good starting conditions. If, for example, an increase in soil carbon is used as an outcome it will disadvantage those with already high levels and thus less potential for improvement.
Scientific and practical
Each EI approach has limitations when it comes to supporting rural land policy. The key limitations for each approach are:
In-situ monitoring can be limited by: cost of commercial equipment; cost and logistical complexities associated with equipment deployment, and maintenance and data management of large networks of sensors. Low cost, open-source monitoring initiatives such as FreeStation (www.freestation.org) have been designed to overcome these limitations and used to evaluate a variety of flood mitigation NBS (Mulligan et al., 2023a).
Spatial modelling can be limited by: lack of input data (either due to unavailability or high cost); incomplete understanding of the biophysical process of converting natural capital to ecosystem services; insufficient information to parameterise models on how management practices impact natural capital and ecosystem services.
Satellite data can be limited by the spatial resolution of data. Open-access satellite data are usually of moderate resolution (30m to 1km pixel resolution) which makes them more suitable for applications beyond the farm scale. Although very high resolution satellite data are available the cost of use can be prohibitive for many within-farm applications.
Regardless of the merits of EI, a key challenge with its uptake, as any new technology, is that it requires a change to business-as-usual. It demands a change in working practices and additional time, effort and cost. Financial barriers are however reducing due to the development of low-cost, open-source monitoring hardware and free-to-use spatial modelling software. The organisational hurdles, of changing established ways of working, remain the greater challenge.
Discussion
Environmental intelligence approaches can support assessment of outcomes-based land management. However, the ability to use outcomes-based incentivisation depends on the possibility of developing reliable indicators (measures) of outcomes and the length of time to achieve measurable outcomes.
Tension between biophysical and policy timescales
Some outcomes cannot be achieved within the timescales required for outcomes-based incentives. For example, the lag between when a management action is started and a rigorously observable change in soil carbon occurs makes change in carbon storage a poor choice for an outcomes-based approach. As policy-relevant outcomes need to be achieved quickly, a more suitable outcome is estimating the carbon fluxes of agricultural soils i.e. to estimate the carbon removed (sequestered) from the atmosphere by the soil every year. If there is a net sequestration, rather than net emission over the course of the year, then this is evidence that the management practice is working and in the long-term soil carbon will increase.
Potential to reinforce geographic inequalities
Many ecosystem service outcomes are assessed based on the number of people or the monetary value of assets benefiting from the service. This means that greater value is given to natural capital that is close to (and therefore benefits) lots of people or valuable assets. To illustrate this point take the example of improved water storage in soils and the ecosystem service of flood mitigation - if the number of people protected or the monetary value of the assets protected from a flood are used as an outcome then it puts a greater value on natural capital that influences more affluent or populated areas at the expense of natural capital that may be benefiting fewer people or assets with less monetary value. Thus, multiple values need to be considered in order for investments and incentives for more sustainable land management to be socially just (Hakkarainen et al., 2020; Kenter, 2018).
Conclusions
Environmental intelligence approaches can help deliver outcomes-based rural land management by providing context specific evidence for planning and evaluating NBS investments. We provide a series of recommendations to those considering developing or applying such approaches:
• Metrics for outcomes-based assessment need to be simple, low-cost, easily collected, rigorous and informative of the outcomes that matter.
• Metrics should be comparable between farms, between data collectors and verifiable or independent of those receiving the incentive.
• Incentivise conservation as well as restoration by considering those that have managed natural capital well and already have intact natural capital on their land.
• When evaluating outcomes, give consideration to the choice of outcome to ensure that they are realistically achievable within the necessary time period and that they are socially just.
• Consider whether measuring outcomes is the best approach, particularly in those situations when the cost of time and resources associated with evaluating outcomes is great. If it is well known that a given management action is environmentally the right choice then it might be more cost-effective and reasonable to incentivise a verifiable practice rather than an outcome.
Ethics and consent
Ethical approval and consent were not required.
Data availability
No data are associated with this article.
Author contributions
Caitlin M. S. Douglas: Conceptualization, Writing - original draft; Writing - review and editing
Mark Mulligan: Conceptualization; Writing - original draft; Writing - review and editing
Faculty Opinions recommendedReferences
- Cooper HV, Sjogersten S, Lark RM, et al.:
To till or not to till in a temperate ecosystem? implications for climate change mitigation.
Env Res.
2021; 16(5): 054022. Publisher Full Text
- Defra: Guidance environmental land management schemes: overview. Department for Environment Food and Rural Affairs and Rural Payments Agency. 2021; Accessed 4 November 2022. Reference Source
- European Council: Timeline - history of the CAP. 2022; Accessed 31 October 2022. Reference Source
- Gawith D, Hodge I:
Focus rural land policies on ecosystem services, not agriculture.
Nat Ecol Evol.
2019; 3(8): 1136–1139. PubMed Abstract
| Publisher Full Text
- Gerowitt B, Isselstein J, Marggraf R:
Rewards for ecological goods - requirements and perspectives for agricultural use.
Agr Ecosystems Env.
2003; 98(1–3): 541–547. Publisher Full Text
- Hakkarainen V, Anderson CB, Eriksson M, et al.:
Grounding IPBES experts’ views on the multiple values of nature in epistemology, knowledge and collaborative science.
Env Sci Policy.
2020; 105: 11–18. Publisher Full Text
- Hayhow DB, Eaton MA, Stanbury AJ, et al.:
The state of nature 2019. The State of Nature partnership. 2019. Reference Source
- Herzon I, Birge T, Allen B, et al.:
Time to look for evidence: results-based approach to biodiversity conservation on farmland in Europe.
Land Use Policy.
2018; 71: 347–354. Publisher Full Text
- Kenter J:
IPBES: Don’t throw out the baby whilst keeping the bathwater; put people’s values central, not nature’s contributions.
Ecosyst Serv.
2018; 33: 40–43. Publisher Full Text
- Kristensen SBP:
Agriculture and landscape interaction - landowner’s decision-making and drivers of land use change in rural Europe.
Land Use Policy.
2016; 57: 759–763. Publisher Full Text
- Matzdorf B, Lorenz J:
How cost-effective are result-orientated agri-environmental measures? - an empirical analysis in Germany.
Land Use Policy.
2010; 27(2): 535–544. Publisher Full Text
- Mueller L, Eulenstein F, Dronin N, et al.:
Agricultural landscapes: history, status and challenges. In: Mueller, L. et al. (eds), Exploring and Optimizing Agricultural Landscapes, Innovations in Landscape Research. 2021; 3–54. Publisher Full Text
- Mulligan M:
Mapping hydrological ecosystem services and impacts of scenarios for deforestation and conservation of lowland, montane and cloud-affected forests. In: Myster, W. (ed.) The Andean Cloud Forest. Springer, Cham., 2021; 189–218. Publisher Full Text
- Mulligan M, Burke S, Douglas C, et al.:
Methodologies to assess and map the biophysical effectiveness of nature based solutions. In: Lopez Gunn, E. et al. (eds) In: Greening Water Risks. Water Security in a New World. Springer, Cham., 2023a; 51–65. Publisher Full Text
- Mulligan M, Douglas C, van Soesbergen A, et al.:
Environmental intelligence for more sustainable infrastructure investment. et al.: GoodIT'21 Proceedings of the Conference on Information Technology for Social Good. 2021; 225–229. Publisher Full Text
- Mulligan M, van Soesbergen A, Douglas C, et al.:
Natural Flood Management in the Thames Basin: building evidence for what will and will not work. In: Lopez Gunn, E. et al. (eds) In: Greening Water Risks. Water Security in a New World. Springer, Cham., 2023b; 223–246. Publisher Full Text
- Oelbermann M, Voroney RP:
Carbon and nitrogen in a temperate agroforestry system: using stable isotopes as a tool to understand soil dynamics.
Ecol Eng.
2007; 29(4): 342–349. Publisher Full Text
- Plieninger T, Bieling C:
Connecting cultural landscapes to resilience. In: Plieninger, T. & Bieling, C. (eds) Resilience and the cultural landscape: understanding and managing change in human-shaped environments. Cambridge University Press, Cambridge, 2012; 3–26. Publisher Full Text
- Poulton PR, Johnston AE:
Can Long-term experiments help us understand, and manage, the wider landscapes - examples from Rothamsted, England. In: Mueller, L. et al. (eds). Exploring and Optimizing Agricultural Landscapes, Innovations in Landscape Research. 2021; 233–252. Publisher Full Text
- Schulp C, Levers C, Kuemmerle T, et al.:
Mapping and modelling past and future land use change in Europe’s cultural landscapes.
Land Use Policy.
2019; 80: 332–344. Publisher Full Text
- Smith P, Soussana J, Angers D, et al.:
How to measure, report and verify soil carbon change to realise the potential of soil carbon sequestration for atmospheric greenhouse gas removal.
Global Change Biol.
2020; 26(1): 219–241. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Soane BD, Ball BC, Arvidsson J, et al.:
No-till in northern, western and south-western Europe: a review of problems and opportunities for crop production and the environment.
Soil Tillage Res.
2012; 118: 66–87. Publisher Full Text
- Springermann M, Freund F:
Options for reforming agricultural subsidies from health, climate, and economic perspectives.
Nat Commun.
2022; 13(1): 82. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Stoate C, Báldi A, Beja P, et al.:
Ecological impacts of early 21st century agricultural change in Europe--a review.
J Environ Manage.
2009; 91(1): 22–46. PubMed Abstract
| Publisher Full Text
- van Zanten BT, Verburg PH, Espinosa M, et al.:
European agricultural landscapes, Common Agricultural Policy and ecosystem services: a review.
Agron Sustain Dev.
2014; 34: 309–325. Publisher Full Text
- Vejchodská E, Shahab S, Hartmann T:
Revisiting the purpose of land policy: efficiency and equity.
J Plan Lit.
2022; 37(4): 575–588. Publisher Full Text
- Winkler K, Fuchs R, Rounsevell M, et al.:
Global land use changes are four times greater than previously estimated.
Nat Commun.
2021; 12: 2501. PubMed Abstract
| Publisher Full Text
| Free Full Text
- Zhao J, Yang Y, Zhang K, et al.:
Does crop rotation yield more in China? a meta-analysis.
Fields Crops Res.
2020; 245: 107659. Publisher Full Text
