Evaluating the Options for Combining Economically, Socially and Ecologically Sustainable Agriculture
Lead Research Organisation:
University of East Anglia
Department Name: Biological Sciences
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Organisations
Publications
Freckleton R
(2008)
Modelling the effects of management on population dynamics: some lessons from annual weeds
in Journal of Applied Ecology
Queenborough S
(2010)
From meso- to macroscale population dynamics: a new density-structured approach
in Methods in Ecology and Evolution
Freckleton RP
(2011)
Density-structured models for plant population dynamics.
in The American naturalist
Sutherland WJ
(2012)
Making predictive ecology more relevant to policy makers and practitioners.
in Philosophical transactions of the Royal Society of London. Series B, Biological sciences
| Description | This multi-institution research project investigated how the land management strategies of UK farmers affect the biodiversity of the countryside. We brought together social and natural scientists to understand the factors influencing farming practices and the implications of variations in practices for biodiversity. The aim of this project was to develop models of farm land-use by integrating its economic, social, biological components. We studied 48 arable farms across Bedfordshire, Lincolnshire and Norfolk. We surveyed each farmer's land-use objectives and preferences, monitored weeds and analysed field diaries for 500 fields over three years, producing valuable social data on farmers and farm management, for this and future studies. 1. Stakeholder mapping. Many groups of people have a stake in the sustainable use of farmland, but they have different degrees of influence and interest in the outcomes. We developed an interactive software tool for classifying the type and strength of interest and influence that different stakeholders, such as farmers, government, and commercial organisations, have in the management of arable farms. This has been useful in showing how stakeholders exert influence to achieve their particular interests, for example through land ownership or legislation. 2. Farmer decision making. An innovative feature of this study is that a survey of farmers' objectives and preferences was designed to produce results for direct input into a predictive model of land-use. This model determines the optimal management of a farm, given constraints such as the time and equipment available, soil type, commodity prices, labour, and input costs, taking into account the farmer's land-use objectives. We used this model to predict landscapes based on crop prices and agricultural subsidy structures, and compared our predictions with actual land-use recorded by the Farm Business Survey. We found that a model that included non-profit objectives, such as risk management and preference for the number of crops managed, improved predictive capability compared with a model based on profit-maximisation alone. 3. Weed population dynamics. Weeds are very important in determining farmland biodiversity, but good data on populations and dynamics are extremely time-consuming to collect. For the first time, we have mapped whole fields, developing a new method for assessing weed population responses to management for integration with other models. Previous studies have been conducted at the scale of a single field, but we have mapped 500 fields across 48 farms. We concentrated on six species of weeds, three of which are ecologically important, the other three being weeds that are economically significant. The resultant dataset consists of maps of weeds within ten fields per farm, over three years. This, in itself, is a unique resource. The results show that there are significant variations in density between ecologically and economically significant weeds. The latter, comprising species such as Blackgrass, tend to be dominant. Patterns of weed abundance and management are enormously variable even within the same crop. 4. Modelling bird populations. Birds suffered the most publicized losses of biodiversity due to farming practices and bird populations underlie the principal monitoring of ongoing changes. We developed statistical models of variation in bird breeding abundance with respect to cropping, field boundary characteristics and landscape composition across 880 lowland farmland 1km squares. The models integrated predicted bird population consequences with the models of farmers' cropping choices. The results showed that crop types were less important than landscape composition and field boundary characteristics in determining absolute bird abundance, but are probably more important in driving changes in numbers. A separate analysis considered how land-use heterogeneity (i.e. the mixing of arable, grass, woodland, etc.) affected bird counts. We found that more threatened species responded negatively to heterogeneity, unlike generalist species and contrary to previous research results, A second approach to predicting the bird consequences of farming practices was to use published models of habitat use by birds at the field scale to predict breeding densities on our study farms. Work to complete this approach remains to be done, but results to date support the national pattern: non-crop elements in the landscape determine habitat suitability, but some cropping changes are important. 5. Integration. This work integrates research at several scales. For a single farm we can (i) predict optimal management, (ii) measure the degree to which actual behaviour deviates from this optimum, (iii) produce average weed densities for different crops, (iv) predict weed populations and bird populations for altered management. At a landscape scale, we can predict aggregate land-use and ecological populations under different assumptions about farmer behaviour. The main result from this work was that profit maximization does not provide a good description of actual farming practice. Instead the inclusion of farmer-specific stated preferences improved models and gave predicted power. This emphasises the need to include social data in models that attempt to recreate agricultural landscapes. The enormous variability in weed and bird populations also cautions that local farm and individual farmer differences need to be considered. The work on farmer decision making has, for the first time, linked social data with farm management and ecological outcomes. In terms of farm management, we have found that, although farmers that have high stated preferences for conservation also tend to use less nitrogen, but they do not have high weed densities. It suggests that farmers are better informed or have more flexibility in nitrogen use than herbicide use. Impacts Our models contribute to answering several policy questions, including: • What would be the best policy measures to achieve the targets on bird populations set by the government? • What determines which new farming methods will be adopted by farmers? • What will be the social and economic consequences of biodiversity conservation? |
| Exploitation Route | Our models contribute to answering several policy questions, including: • What would be the best policy measures to achieve the targets on bird populations set by the government? • What determines which new farming methods will be adopted by farmers? • What will be the social and economic consequences of biodiversity conservation? By exploring how the attitudes, objectives and actions of farmers affect biodiversity, we hope to contribute to best practice models for arable farm management. Our research is particularly important given that the use of certain herbicides is set to be curtailed by the government, meaning that management practices could change considerably. |
| Sectors | Agriculture, Food and Drink,Chemicals,Environment |
| Description | This research provides the models for understanding the ecological consequences of agricultural decisions. The weed population models have been influential within their field leading to a range of other models using the same approach. |
| First Year Of Impact | 2009 |
| Sector | Environment |
| Impact Types | Economic,Policy & public services |