Land Use Policy 29 (2012) 502512

Contents lists available at SciVerse ScienceDirect

Land Use Policy

journal homepage: www.elsevier.com/locate/landusepol

The Neighbourhood Effect: A multidisciplinary assessment of the case for farmer co-ordination in agri-environmental programmes
Lee-Ann Sutherland a, , Doreen Gabriel b , Laura Hathaway-Jenkins c , Unai Pascual d,h,1 , Ulrich Schmutz e , Dan Rigby f , Richard Godwin c , Steven M. Sait b , Ruben Sakrabani c , William E. Kunin b , Tim G. Benton b , Sigrid Stagl g,2
a

James Hutton Institute, Craigiebuckler, Aberdeen AB15 8QH, UK Institute of Integrative & Comparative Biology, University of Leeds, Leeds LS2 9JT, UK c Department of Environmental Science and Technology, Craneld University, Craneld, Bedford MK43 0AL, UK d University of Cambridge, Department of Land Economy, 19 Silver St., Cambridge, UK e Henry Doubleday Research Association (Garden Organic), Ryton Organic Gardens, Coventry CV8 3LG, UK f University of Manchester, School of Economic Studies, Manchester M13 9PL, UK g University of Sussex, SPRU Science and Technology Policy Research, Brighton BN1 9QE, UK h IKERBASQUE, Basque Foundation for Science, Alameda Urquijo, 36-5, 48011 Bilbao Spain
b

a r t i c l e

i n f o

a b s t r a c t
In this paper we present a multi-disciplinary analysis of the potential impacts of undertaking similar environmental actions on multiple farms in a small geographic area, using organic farming as a proxy for a co-ordinated approach. Recent papers have called for more co-ordinated efforts between farmers in terms of their environmental actions, but there has been limited applied research demonstrating the environmental benets or the economic and social implications to farmers of this approach. Comparative analysis of biodiversity, soil and water, and farm protability were undertaken in England on 32 matched farms in areas of low and high organic farming concentration; qualitative interviews were also conducted with 48 farmers living in two of the eight areas. Findings demonstrate higher overall levels of biodiversity on organic farms (particularly in hotspot areas) but this was not universal across the species groups investigated. Higher water inltration rates were found in organic grasslands, which could prove to be a useful measure to combat ooding. In terms of the technical efciency of producing these environmental gains, conventional and organic farms in hotspot areas demonstrated equivalent efciency from a nancial perspective. Socio-cultural research identied the different amounts of trust farmers have in their neighbours, based in part on their performance as good farmers. We discuss the neighbourhood effect with a multi-disciplinary approach and conclude that encouraging local farmer co-ordination can have clear environmental benets without high economic cost, but must be undertaken with caution specically regarding the trade-offs between benets, local geophysical and social characteristics, and assumptions made about inter-farmer trust. 2011 Elsevier Ltd. All rights reserved.

Article history: Received 7 October 2010 Received in revised form 24 June 2011 Accepted 3 September 2011 Keywords: Organic farming Multidisciplinary research Agri-environmental schemes Landscape Farmer co-ordination Biodiversity

Introduction Policy initiatives emanating from Europe increasingly encourage the management of environmental assets beyond the farm level, e.g. through the Water Framework Directive, in which watersheds are to be managed at the catchment (rather than individual

Corresponding author. Tel.: +44 01224 395 285; fax: +44 01224 395 010. E-mail address: lee-ann.sutherland@hutton.ac.ukk (L.-A. Sutherland). 1 Present address: Basque Centre for Climate Change (BC3) Alameda Urquijo 4, 4. 48008 Bilbao Bizkaia, Basque Country, Spain. 2 Present address: Vienna University of Economics and Business, Institute for the Environment and Regional Development, Nordbergstrasse 15 (UZA4, 4B), A-1090 Vienna, Austria. 0264-8377/$ see front matter 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.landusepol.2011.09.003

farm) level (Lexartza- Artza and Wainwright, 2009). In the UK, Defra (through Natural England) has moved towards more areabased approaches, targeting specic needs through its High Level Stewardship scheme. The ecological science underpinning these approaches is well documented: although most studies have investigated the impact of farming on biodiversity at eld or farm scales (Bengtsson et al., 2005; Fuller et al., 2005; Hole et al., 2005), several have demonstrated that populations of many macroorganisms respond to the environment at substantially larger spatial scales (Chamberlain et al., 2000; Benton et al., 2002; Tscharntke et al., 2005). The hydrological consequences of farming management practices also depend on cultivation patterns at the watershed scale, which may limit the usefulness of considering management changes at the individual eld or farm scale (Hess et al., 2010). However, the social and economic implications of these

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

503

approaches are less well known: cultivation and land management practices can be expected to impact on both the protability of the land use and the prestige of the land manager (Burton et al., 2008; Sutherland and Burton, 2011). Neither is the impact of contiguous agri-environmental programming as simple as more is better: while continuous areas of arable land forming homogenous landscapes can have a negative impact on wildlife (Benton et al., 2003; Dauber et al., 2005; Gabriel et al., 2005; Tscharntke et al., 2005; Hendrickx et al., 2007; Billeter et al., 2008), agrienvironment schemes may achieve the biggest benet in those landscapes (Roschewitz et al., 2005; Rundloef and Smith, 2006; Gabriel et al., 2006; Holzschuh et al., 2007). To underpin future policy initiatives towards co-ordinated environmental initiatives, an integrated analysis is needed which looks at benets and tradeoffs in relation to environmental management across multiple farms. In this paper, we present recent ndings assessing how the ecological, soil and water, economic and cultural characteristics of agricultural land vary due to neighbourhood effects, using engagement in organic farming as a proxy for co-ordinated environmental action. It is appropriate to use organic farming as a proxy, because it enables the identication of a large number of farmers who are certied as maintaining the same minimum set of management standards and practices. Although organic farming is considerably more than an agri-environmental scheme involving alignment with an ideological approach to land management and specialist marketing channels conversion to organic farming has been included in the suite of environmental schemes supported by the EU since 1994 (Offermann et al., 2009). In the UK, 4.3% of agricultural land was certied as organic (or in conversion) in 2009 (Soil Association, 2010), close to the EU average of 4.7% (Eurosta, 2011). Similar to the research on agri-environmental schemes, analysis of the environmental impacts of organic farming has been largely limited to individual farms, or broad sectoral and national consideration of wide scale conversion (see Lampkin and Padel, 1994). However, it is reasonable to expect that impacts would vary depending on the management practices of neighbouring farms for similar reasons. Some of the biodiversity benets of organic agriculture may not be realised when only a small area of land is under organic management, as small isolated habitat fragments may be insufciently large to maintain viable populations of some species (Hole et al., 2005; Whittingham, 2007). In contrast, higher concentrations of organic farms are typically found in landscapes with certain environmental conditions; areas with lower concentrations of organic farms may differ more in soil, landscape context and topography and thus species inventory may do so too. This could lead to higher species turnover (i.e. beta diversity) between isolated organic farms (Gabriel et al., 2006), which taken together may promote similar biodiversity levels than aggregated clusters of organic farms. Different levels of local engagement in organic farming can also be expected to have social implications for farmerswhereas in the 1980s organic farmers were subject to community censure (Tate, 1994; Tovey, 1997), widespread up-take can be expected to increase social acceptability. However, high numbers of producers undertaking a specic set of environmental actionssuch as organic farmingcan lead to over-supply, and therefore have negative economic implications for the farms involved. Smith and Marsden (2004) have demonstrated that the cost-price squeeze characteristic of conventional farming is also impacting on organic farming, as the number of organic farmers has grown. The purpose of this paper is thus to evaluate the neighbourhood effect of multiple farmers adopting similar management strategies by studying matched sets of farms situated in landscapes with high and low amounts of organically farmed land. On these matched sets of farms, we specically assess:

the economic performance and relative technical efciency of producing both agricultural commodities and biodiversity, taking into account the higher prices received for organic produce, but excluding all farm subsidies; the diversity of a range of focal taxa from farmland birds, soil and above-ground invertebrates (including agricultural pests and their natural enemies, invertebrates used as food resources by birds, and pollinators) and plants (including arable weeds and species of eld margins, fallow elds and grazing lands). We examine whether these taxa respond to land management practices at different spatial scales within-eld, eld within farm, and farm-within-catchment; the social acceptability of organic farming among neighbouring farmers, in particular transitional changes in the development of prestige in farming communities, existing collaborative relationships and issues relating to the up-take of co-ordinated environmental actions by neighbouring farmers; the chemical and biological characteristics of surface and nearsurface waters. This also considers the interaction between organic management regimes and tillage practices (inversion and non-inversion tillage) on soil structure, surface water quality, tillage energy and efciency. We use our ndings to evaluate the case for multi-farmer coordination in agri-environmental action, specically addressing social, environmental and economic costs and benets.

Methods The research presented in this paper was undertaken as part of a multi-disciplinary project addressing the neighbourhood effect of organic farming.3 More specically, the research addressed how the ecological, soil and water, socio-economic and cultural aspects of organic farming vary between areas where organic farming has a strong local concentration (hotspots) as opposed to areas where there is little organically managed land (coldspots). To facilitate analysis of the effects of organic land management beyond the farm level we recruited matched conventional and organic farms in both hotspots and coldspots. This sampling approach underlies the analytical sections of the paper that follow and hence we briey outline it here. To select organic farms in different neighbourhoods, a georeferenced database of organic farms was developed using data on full postcodes and farm sizes of registered organic farms in England in 2005, which were supplied by the UK Department for Environment, Food and Rural Affairs. Organic hotspots and coldspots were identied on the basis of the area and the number of organic farms, and pairs of hot and coldspot landscapes were matched in agrienvironment conditions based on 30 variables describing climate, topography, land use, socio-economy and soil. Within each landscape, an organic and conventional farm was then selected, based on similarity in terms of: enterprise structure with dairy or mixed farms, i.e. farms with both arable and livestock farming, with similar livestock, cereal production, farm products and farm size; soil type (determined from soil survey maps and data); proximity (less than 5 km between farms); on each farm, three winter cereal elds and three permanent pastures were selected.

http://www.relu.ac.uk/research/projects/SecondCall/Author12.htm.

504

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

Soil and water methods Assessments were made in one crop and one grass eld per farm. Both soil sampling and within eld assessment were conducted in March and April 2007, when soils were at or near to the eld capacity moisture content. This is when seasonal effects of variations in soil moisture content are minimized and soil structural condition is most easily assessed allowing comparisons to be made between soils and years of sampling (Smith and Trafford, 1976; HathawayJenkins et al., 2011). At each site soil assessment was conducted and samples were collected to measure a suite of physical (texture, aggregate stability, soil organic matter) and chemical (pesticides, herbicides and nutrients) parameters. To obtain a representative sample of soil, a W shaped path sampling strategy was observed, avoiding untypical areas, taking 10 samples (0200 mm depth); which were then bulked (MAFF, 2000). At each site, inltration was recorded using a mini disk inltrometer, which because of its small surface areas was replicated 10 times. Each replicate was sampled for 30 min from which the inltration rate was calculated. One or more small pits were excavated by an experienced soil surveyor at each site to determine the visual soil structure and physical conditions of the soil using the procedures suggested by Batey (1988). In situ shear strength was measured with 30 replicates using a grid sampling method. The soil samples were prepared through air drying and homogenisation by grinding and sieving (Allen, 1989); soil texture was then determined using the pipette method (BS 7755) and soil organic matter using dichromate digestion techniques (BS 13773). The aggregate stability was determined through the wet sieving method outlined in Haynes et al. (1991) and moisture content measured by oven drying at 105 C. Soil water samples were sent to NRM laboratory to be analysed for a suite of pesticides, herbicides and nutrients. To facilitate analysis the soils were grouped into four main types (textural classes), namely: clayey, silty, medium and coarse. Statistical analysis was conducted in a similar way to the biodiversity, by using generalized linear mixed models, with sampling locations as nested hierarchical random factors. We investigated whether there were signicant differences in the soil properties between the organic and conventional management regimes; hot and coldspots; arable and grass land uses; and the soil textural classes.

Fig. 1. Location of hotspot and coldspot landscapes in two regions of lowland England.

The resulting sample comprised 8 clusters of paired hotspot and coldspot landscapes, each containing one organic and one conventional farm (32 in total) in two regions, the Central South West and the Midlands. The organic hotspots featured, on average, 17.2% organically managed land within a 10 km 10 km landscape while the organic coldspots included an average of 1.4% organic land (for further details see Gabriel et al., 2010). The methods associated with the disciplinary approaches adopted are presented here in turn. The datasets have been archived at the UK Data Archive at the University of Essex, and are currently available to other researchers for academic use. Ecological methods Assessments were made in three crop and three cereals elds per farm. Repeated surveys were undertaken to record a wide range of taxa in the eld centres, eld edges and margins during MayAugust in 2007 and 2008. Herbaceous plant species (except grasses and ferns) were identied within quadrats, earthworms were sampled using the hot mustard method, epigeal arthropods were sampled with a suction sampler, ower visitors (hoveries, bumblebees, solitary bees) were assessed using pan traps and butteries and birds were recorded by walking standardized transects across elds and farms, respectively. All taxa were surveyed in the crop eld centres, except birds, which were recorded at the farm level using standard transects. Plants and epigeal arthropods were also recorded in the eld edges and margins, earthworms in eld edges, and ower visiting insects in eld margins (survey methodology is described in full detail in Gabriel et al., 2010). Farmland biodiversity was analysed as species density for plants and birds and abundance (density of individuals) for earthworms, epigeal arthropods, butteries and insect pollinators (Fig. 1).

Economic methods Data was collected through structured interviews with the farmer or farm manager on the 32 study farms in autumn of 2006, 2007 and 2008, and data points aggregated and averaged across this time period in order to reduce outlier effects resulting from weather or market uctuations. The types of data collected are listed later in Table 4. It is important to note that subsidies (e.g. the Single Farm Payment, organic conversion grants) were not included in the calculation of economic performance. The economic performance and relative technical efciency of the study farms was investigated using farm accounting methods and Data Envelopment Analysis (DEA). DEA is a nonparametric linear programming method formally developed by Charnes et al. (1978) for the estimation of production frontiers and that can be used to empirically measure productive efciency of farms (e.g. Pascual, 2005). The focus of our work for this paper was whether there were signicant differences in the protability and technical efciency (TE) between the organic and conventional farm management systems in cereal (winter wheat) cropping and how any differences were affected by the landscape effect. To calculate technical efciency, an input-orientated perspective is used following

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

505

Shephards (1970) input distance function under a given technology set T given as: D(z; x) = max > 0 : z, x p T ,

where D(z; x) characterises an input possibility set by the maximum equi-proportional contraction of the input vector (x) without reducing the output vector (z). The reciprocal of D(z; x) reects the input-orientated efciency measure used in the analysis ( = 1/ ), for between zero and one and = 1 implying that the farm is operating on the production frontier. It follows that the maximum equi-proportional input contraction any farm can achieve without reducing output is 1 (Pascual, 2005). The technical efciency analysis undertaken considers the relationships between input and outputs in two, complementary, ways: the rst considers inputs and both outputs (cereal yield and biodiversity combined) in terms of physical quantities, while the second considers all inputs and wheat output in monetary terms. In this latter case the biodiversity output remains measured in physical terms, with no monetary equivalent available. Two sets of biodiversity indexes were used (described in The neighbourhood effect on relative technical efciency in the production of grain yield and biodiversity). Sociocultural methods The sociocultural research focused on focal farms surrounding two of the 8 cluster regions, in order to evaluate the dynamics of higher concentrations of organic farming within geographic neighbourhoods. The sites surrounded two of the matched organic hotspots and coldspots addressed by the other disciplines. In keeping with the study, one was in the English Midlands, the other in Central South-Western England; both had largely mixed farming operations (where arable crops were used primarily to feed either dairy or meat animals). Qualitative interviews were undertaken with 48 farmers across the four sites, 21 of which farmers were organic. Interviews included seven of the 32 farmers from the overall study. Potential organic farming respondents were identied through membership lists of organic farmers provided by two certication bodies (the Soil Association Certication Ltd and Organic Farmers & Growers Ltd). Almost all of the farmers on this list who were currently farming (i.e. their listing was not out of date) agreed to an interview. Conventional farmers were identied through telephone listings (a common sampling approachBurton and Wilson, 2006); all farmers listed in the local areas were contacted, leading to a response rate of approximately 25%. Interviews focused on the attitudes towards organic farming and social networks between organic and conventional farming neighbours. Issues addressed in this paper focus on inter-farmer credibility, trust, and the overlap between organic and conventional farming practices. The interviews ranged in length from 45 min to two hours. Interview recordings were transcribed in full, entered into NVIVO qualitative data analysis software and coded into nodes reecting the initial question guide and emergent themes. Note on multidisciplinary research The research presented in this paper was deliberately constructed to enable multiple disciplines to work together to address a set of research questions. This process was not without its challenges. Many of these related to the standard disciplinary practices of the researchers involved. Thirty-two farms represent an opportunity for extensive soil and water and ecological research, but

is a very low number in terms of typical economic analyses; this was addressed through using data collected over multiple years. While numbers of participants was not a problem for the qualitative study, accurately identifying a neighbourhood effect on local culture would be difcult to base on two interviews in each location. As a result, the sociocultural research focused in two of the eight locations, in order to interview the geographic neighbours in hot and coldspot locations. The selection of matched farms was particularly difcult. There were a number of constraints: farms had to be as similar as possiblein size, enterprise, environment and so onsuch that differences were more likely to be due to management than other factors. Farms needed to be close (so the environments were similar) but not too close (to allow independence, especially of biodiversity measurements). Importantly, to avoid selection bias, we recruited farmers at random, rather than by seeking volunteers. As a result of this complexity, recruitment of the 32 farms itself took a year, rather than the ve months originally planned. Inevitably, the differing requirements of the different aspects of the study led to compromises in study design and selection. For example, matching organic and conventional farms for size and enterprise means that our conventional farms did not include examples of large, specialised farms. Furthermore, distinguishing between grasslands and arable elds are important to soils and hydrology and plant and animal ecology, but it is almost impossible to determine the crop yield of grasslands in terms of amount of silage or haylage from grasslands, due to difculty gauging consumption through grazing in combination with grass fodder harvesting. For this reason, economic yield analysis was only used from the cereals elds.

Note on the scale of analysis Our study focuses on a scale larger than the farm, the neighbourhood, and it is at this scale we estimate effects. This is an important advance over many previous studies but it also omits a larger spatial scale. These larger impacts are difcult to assess but occur via two routes. One is that farming operations interface with the environment at a broader scale by, for example, buyingin cattle feed that may have ingredients produced worldwide, or the off-farm disposal of slurry, so local environmental effects are supplemented by costs and benets levied elsewhere. The second is that demand for a range of food products (and quantities) means that changing local production may affect supply chains that can have consequences on a very large spatial scale. For example, region-wide conversion of highly productive farming to less intensive, lower production farming, would potentially lead to demand for a compensatory increase in production elsewhere. In this study, we assume that demand responds to changes in the supply of organic products and concentrate only on local issues (the neighbourhood effect), merely noting the potential for broader-scale effects than we measure.

Results The neighbourhood effect on farmland birds, above- and below-ground invertebrates and plants Here we summarise the farm and landscape scale effects of organic farming on farmland biodiversity by averaging species densities or abundances across regions (Central South West vs. North Midlands), crop types (arable vs. grass) and within eld locations (centre vs. edge vs. margin). Detailed analyses for each taxonomic group are published in Gabriel et al. (2010); a summary can be found in Table 1.

506

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

Table 1 Summary statistics describing species density or abundance of different farmland taxa on 192 crop and grass elds on 32 organic and conventional farms in 16 coldspot and hotspot landscapes surveyed over 2 years. Farmland biodiversity Plantsa Earthwormsb Epigeal arthropodsc Butteriesd Hoveriese Bumblebeese Solitary beese Farmland birdsf Conventional Mean s.e.m. 5.96 15.76 14.10 1.96 1.49 0.63 0.68 8.47 0.125 2.210 0.283 0.158 0.047 0.020 0.027 0.218 Organic Mean s.e.m. 9.46 19.38 16.46 2.74 1.049 0.66 0.64 7.68 0.109 2.410 0.321 0.204 0.034 0.021 0.024 0.226 Coldspot Mean s.e.m. 7.62 17.97 14.42 2.05 1.21 0.66 0.55 8.04 0.123 2.372 0.285 0.159 0.028 0.021 0.021 0.218 Hotspot Mean s.e.m. 7.80 17.04 16.09 2.64 1.31 0.64 0.77 8.11 0.125 2.270 0.321 0.203 0.044 0.021 0.030 0.232 Org/con % 158.7 123.0 116.7 140.0 70.2 104.4 93.4 90.7 Hot/cold % 102.3 94.8 111.6 129.2 108.3 95.9 139.5 100.9 Org hot/con cold % 163.5 115.8 130.2 177.4 76.3 134.1 100.1 91.5 Org cold/con hot % 154.2 129.0 104.6 108.8 64.5 69.2 108.6 89.9

Means and standard errors are obtained by averaging either across farm or landscape type. Percentage values are given comparing species density or abundance in organic vs. conventional elds, hotspot vs. coldspot landscapes, organic elds in hotspot landscapes vs. conventional elds in coldspot landscapes and organic elds in coldspot landscapes vs. conventional elds in hotspot landscapes. a Species per transect and survey. b Individuals per eld. c Individuals per sampling station and survey. d Individuals per transect and survey. e Individuals per sampling station and survey. f Species per farm and survey.

Organic elds had on average 12.4% higher diversity levels, with plant species density 58%, buttery abundance 40%, earthworm abundance 23% and epigeal arthropods 16% higher, but a 33% lower hovery abundance and 9% lower bird species density. Fields in hotspots had on average 9.1% higher diversity levels, and butteries, solitary bees, epigeal arthropods and hoveries beneted in particular with 39%, 29% and 12% and 8% higher abundances, respectively (see Table 1). The strongest contrast in biodiversity levels occurred between conventional elds in coldspot landscapes and organic elds in hotspot landscapes with on average 22.5% higher diversity levels, whereas hardly any differences (on average 3.3%) were observed between conventional elds in hotspot landscapes and organic elds in coldspot landscapes. The neighbourhood effect on soil and water properties The principle ndings are summarised in Table 2. Overall, the analysis of the data shows that while it is possible to detect differences in the effect of soil type and land use (grassland/arable) on a number of the soil properties there are, with the exception of inltration rate, no signicant differences in the soil physical condition between organic and conventional farming practices. There was no difference between hot and coldspots, as differences in soil type were more dominant than landscape composition. The data, however, show that a. The heavier textured (higher clay content) soils have significantly higher: organic matter, aggregate stability and shear strength; while the coarse textured (higher sand content) soils have a signicantly lower eld capacity moisture content. b. Grassland has a signicantly higher: level of soil organic matter, eld capacity moisture content, aggregate stability and soil shear strength. Similarly, there were fewer low-level traces of identied pesticide and herbicides in the soil water from the organic elds compared with the conventional elds. It must be stressed that all the pesticide and herbicide levels recorded were less than the current thresholds given by no observed effect concentration (NOEC) level that are considered detrimental to the environment (PPDB Footprint Report, 2009). As would be expected the conventional arable elds had higher levels of total inorganic nitrogen (30.56 mg kg1 ) than the grassland and organic arable

(14.389.94 mg kg1 ). There were no signicant differences in total phosphorous and total potassium. However, while there was no signicant difference in the inltration rates between organic and conventional arable soils, there was evidence to support the contention that inltration rates were higher on organically managed grassland (7.6 mm h1 ) in comparison with conventional grassland (2.5 mm h1 ). Such higher inltration rates may have a benecial effect upon reducing runoff from a catchment. This was examined in more detail using the USDA Soil Conservation Service model (Mishra and Singh, 2003) to compare a conventionally dominated landscape (60% arable land, 25% grassland and 15% set-a-side) to an organically dominated landscape (45% arable, 40% grassland and 15% set-a-side) (Norton et al., 2009) which showed a predicted reduction in peak runoff of approximately 30%. There was some evidence that lower animal stocking densities could have been the cause of the difference as in this study as the organic farms averaged a stocking density of 1.1, whereas conventional farms averaged 1.3. Lower stocking density of livestock can be expected to cause less soil damage and could facilitate better inltration of water into soil. Further details of the soil and water research can be found in Hathaway-Jenkins et al. (2011). The neighbourhood effect on relative technical efciency in the production of grain yield and biodiversity Considering rst the protability of the cereal4 enterprises on organic and conventional farms, one would typically expect the yield of organic cereals to be lower than in non-organic systems, the price to be higher and the input costs to be lower. This was indeed the case, as demonstrated in the yields, prices, costs and net margins of the study farms, which are shown in Table 3. The relative protability of the two cereal enterprises will therefore depend on the scale of these differences. According to the data in our sample, conventional farming is associated with a higher yield and revenue output than organic in both hotspots and coldspots. Prices are higher, and unit costs lower, on the organic farms. All these are highly signicant differences in the sample farms (see Table 3). The resulting net margins are higher in the organic cereal systems in both hotspots and coldspots, signicant

4 The vast majority of cereal crops sampled were winter wheat but a few triticale crops (cross between wheat and rye) and spring wheat were also included in the analysis.

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

507

Table 2 Summary of the effects of soil type (texture), land use and organic/conventional management on the soil, water, pesticide, and nutrient stautus of the elds in the study. Property SOM/SOC Soil textural class Clayey and silty soils signicantly higher than medium and coarse soil Coarse soil signicantly lower than other soil textures Clayey and silty soils signicantly higher than coarse soils. Medium soil not signicantly different from the other soils No signicant effect Land use (arable/grass) Grassland signicantly higher than arable for both organic and conventional Grassland signicantly higher than arable for both organic and conventional Grassland signicantly higher than arable for both organic and conventional Management (organic/conventional) No signicant effect Remarks Agrees with the ndings of Gosling and Shepherd (2002) As expected given the SOM results As expected given the SOM results Agrees with the ndings of Stolze et al. (2000)

Field capacity

No signicant effect

Aggregate stability

No signicant effect

Plastic limit

Liquid limit Plasticity index Shear strength

Coarse soils are signicantly lower than the other soils No signicant effect Clayey soils are signicantly higher than the other soil textural groups Soil textural class

Grassland signicantly higher than arable for both organic and conventional No signicant effect No signicant effect Grassland signicantly higher than arable for both organic and conventional Land use (arable/grass)

No signicant effect

No signicant effect No signicant effect No signicant effect

Property Pesticides/herbicides

Management (organic/conventional)

As expected where soil texture and presence of roots/non tillage in grassland would maintain strength Remarks All below no observed effects concentration (NOEC) levels

Total inorganic nitrogen Total phosphorous Total potassium Inltration rate

Two organic elds showed traces (0.3 and 0.02 mg kg1 ) of organochlorine (DDE) and organonitrogen (pendimethaline) respectively. Two conventional elds showed organochlorine (DDE), nine organonitrogen (pendimethaline) and two triazoles. Levels detected below NOEC No signicant effect Conventional arable signicantly higher than all the other treatments No signicant effect No signicant effect No signicant effect No signicant effect No signicant effect No signicant effect Conventional grassland signicantly less than all the other treatments Clay and sandy loam soils signicantly higher than silty clay loam and clay loam

As expected As expected As expected

but with a larger error level (Table 3). Regarding effects at neighbourhood level, organic farming has higher yield and output in cold than in hotspots; however, at the gross and net margin level these differences between hot and coldspot organic farms are no longer signicant.

Table 3 Adjusted grain yield, revenue, prices, costs (variable and allocated xed costs) and net margins (including straw output and costs, and certication costs). Organic Grain yield (tonne/ha) 3.3 Hot Cold 3.8 All 3.5 Grain price (/tonne) 249 Hot Cold 245 All 247 Revenue (/ha) Hot Cold All Costs (/ha) Hot Cold All 900 1060 960 333 342 336 Conventional 8.2 8.5 8.3 124 140 132 1124 1255 1190 679 638 657 432 603 516 Difference 4.9*** 4.7*** 4.8*** 125*** 105*** 155*** 224*** 195*** 230*** 346*** 294*** 321*** 176** 140* 144**

In the productive (also called technical) efciency analysis the focus is broadened to include outputs beyond the physical yield and revenue from cereal production. The farms were considered as multi-output systems in which farmers use a combination of inputs to produce both cereals and biodiversity. A traditional efciency analysis would consider the efciency of the system in terms of the production of a single output, wheat.5 In this context a consideration of the efciency of farming systems in terms of both marketable and non-marketable, environmental, outputs is pertinent. In this study, we considered the ecological outputs of the farm, in terms of biodiversity. The outputs and inputs considered in the two forms of efciency analysis are summarised in Table 4. Four input-orientated TE are used in the analysis combining (i) whether farming is conducted under constant or varying returns to scale and (ii) utilizing two sets of biodiversity indexes. One is related to Principal Component Analysis using centre of eld data on biodiversity (called PCA1c) the other is related to Factor Analysis based on eld centre and edge data (called FA1FA2; the two biodiversity indexes are used together). The four TE indexes used in the analysis are associated with the following assumptions:

TE1: Assumption of constant returns to scale in production using PCA1c as only biodiversity index. TE2: Assumption of constant returns to scale in production using FA1 and FA2 as two biodiversity indexes together.

Net margin (/ha) 608 Hot Cold 743 All 660

* ** ***

Statistically signicant at 0.05 level. Statistically signicant at 0.01 level. Statistically signicant at 0.001 level.

5 We selected wheat for study because when this study was designed, wheat was the most common cereal crop in the UK, occupying 42% of total UK cropping area (Defra, 2006). The second most common was barley, representing 22% of UK crop area.

508

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

Table 4 Summary of inputs and outputs for the efciency analysis. Physical Outputs Cereal yield Biodiversity index Inputs Soil nitrogen Seed Syn fert applic Org fert applic N fertiliser Organic manure P2 O5 fertiliser K2 O fertiliser Sprays Sprays Mechanical weeding Agronomist Casual labour Cultivation and combining Cultivation and combining Qty in soil Seed rate No. of passes No. of organic manure fertiliser passes Qty synthetic and organic manure Qty synthetic and organic manure Qty synthetic and organic manure No. of spray passes Qty pesticides applied No. of passes Agronomist used (0,1) Casual labour hours in h/ha No. of all cultivation passes Cereal revenue Biodiversity index Monetary

Seed cost

Cost inc. application Cost inc. application Cost inc. application Cost inc. application Cost of sprays Cost of spray application inc labour Machinery and labour costs Agronomist cost Casual labour cost Cultivation machinery and labour costs Combine machinery and labour costs

TE3: Assumption of varying returns to scale in production using PCA1c as only biodiversity index. TE4: Assumption of varying returns to scale in production using FA1 and FA2 as two biodiversity indexes together. The results using physical indices (Tables 5a and 5b) indicate that the conventional farms outperform the organic farms in terms of technical efciency. This relationship holds in both hot and cold organic landscapes for 2 of the 4 TE indices. The neighbourhood effect on efciency differs between the two systems. There is strong evidence (all 4 TE indices) of higher efciency on hotspot, as opposed to coldspot, organic farms, while there is evidence that the reverse is true in the conventional sample, albeit for only 1 of the 4 TE indices (the other three not showing a statistically signicant difference). Next the efciency analysis is conducted using the nancial value of inputs and cereal yield in addition to on-farm biodiversity levels measured in physical terms. The results (Tables 6a and 6b) indicate that while the conventional farms still outperform the organic farms in coldspots this is no longer the case among the hotspot farms, where there is no signicant difference between the organic and conventional farms. As before, the organic farms in the hotspots outperform those in the coldspots, for three out of four TE indices. For conventional agriculture the reverse holds as in the physical analysis; however, comparing to organic farms the evidence is slightly weaker since it holds for 2 of the 4 TE indices. This implies that organic farmers are associated with greater technical efciency in landscapes where there is a greater concentration of organically managed land.

Reviewing the TE results, the conventional mixed farms had higher levels of technical efciency compared to organic farms when jointly considering the physical yield of cereals and biodiversity levels. An obvious comment on this nding is that organic farms are producing a lower mass of higher value output and hence an assessment in physical terms might be argued to penalise them. However, given that this was a multi-output assessment which included biodiversity as an output, the result is still noteworthy. The analysis using nancial values of inputs and cereal output throws further light on this. In this monetary analysis the conventional wheat systems still outperform their organic counterparts in the coldspots, but there is no difference in performance, for any of the four TE indices, in the hotspots. When discussing effects within one farming system it is interesting to note that the level of technical efciency is signicantly high where organic farmer concentration is higher under a physical multi-output setting. When nancial TE indexes are included this signicance remains for the majority of TE indexes used. Given the highly robust results, we conclude that this analysis can conrm assumptions that concentration of organic farmland could provide relatively higher physical and monetary efciencies when also considering biodiversity as an output. Neighbourhood implications for multi-farmer co-ordination Greater numbers of organic farmers in a locale can be expected to lead to greater social acceptance of organic farming, owing to its familiarity. However this was found to vary considerably between the two sites. The study site hotspot in the Midlands was formed largely due to the active recruitment efforts of an estate owner, who sought out tenants who wanted to start organic farms. Five of the seven organic farms in the immediate area were thus new to the neighbourhood. In contrast, organic farmers in the southern study site were almost entirely multi-generation farms which had been converted, some as long as 20 years prior to the study. The direct outcome in this difference in how the hotspots originated appeared to be a much higher level of credibility of organic farming among local conventional farmers in the southern hotspot. To inuence or gain the respect of neighbours, the farmer needs to have farmed well and to have done so long-term: [name] has been farming organically for I dont know, twenty years and I think she has been doing it well. And I think sort of, probably, leading up to quite a few of us around here converting. When you see a neighbouring farmer doing something well you

Table 5a Technical efciency results summary: physical indices. ORG Hot Cold HOp COp < < CONV HCp CCp

Notes: Arrows indicate direction of signicant (at 5% level) differences in technical efciency unless otherwise noted. HCp > HOp signicant for 2/4 TE indices. CCp > COp signicant for 2/4 TE indices. HOp > COp signicant for 4/4 TE indices. HCp < CCp signicant for 1/4 TE index (at 10% sign. level).

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512 Table 5b Technical efciency results parameters: physical indices (signicance levels in brackets). TE1 ORG TE2 CONV TE3 ORG TE4 CONV ORG 0.97 0.93 CONV 0.95 0.95 ORG 0.97 0.93

509

CONV 0.96 0.95

Hot 0.81 0.88 0.88 0.92 0.69 0.92 0.74 0.94 Cold Signicance level: TE1: a (1%), b (1%), c(10%), d(1%). TE2: a (1%), b(5%), d(1%). TE3: d (5%). TE4: d (5%) Notes: (a): Signicant difference (one tail-test) of TE between Conv/Org in coldspots. (b): Signicant difference (one tail-test) of TE between Conv/Org in hotspots. (c): Signicant difference (one tail-test) of TE between Hot/Cold in conventional agriculture. (d): Signicant difference (one tail-test) of TE between Hot/Cold in organic agriculture.

think well, if you basically think it is an option and you see it working, then the fact that somebody locally is making it work, it means you are more likely to give it a go. (Gary, organic farmer) The study ndings demonstrated that organic farmers clearly inuence neighbours attitudes towards organic farmingbut that this can be both positive and negative. Organic farmers in the Midlands hotspot were unlikely to be identied as good farmers by their neighbours, owing to poor livestock husbandry standards or the time spent in on-farm diversication through farm shops, whereas several of the organic farmers in the southern hotspot had widespread respect, based on longterm production of quality commodities. The implication is that hotspots grow best on their own, facilitated by highly credible neighbours. Further spontaneous development of organic farming is less likely in areas where it has been undertaken by newcomers, or is perceived to reect poor quality outcomes. Recognition that not all farmers are viewed as good farmers raises issues of trust between farming neighbours. This study did not formally assess willingness to undertake co-ordinated agri-environmental programmes. However, it did address existing collaborative arrangements with neighbouring farmers. Formal sharing of equipment and labour in the study was highly limited, with farmers generally preferring to formally contract out the larger tasks for which they did not have sufcient labour or equipment (e.g. planting or harvesting arable crops). As one organic farmer commented, [about the potential to collaborate with a neighbour who was in the process of converting to organic farming] I dont know if it will work yet because I am quite cynical about it, as to the way he approaches his farming so whether to go into some sort of business arrangement with somebody like that is a bit dodgy. (Brian, organic farmer) There is a natural reluctance on the part of any business owner to see their own efforts jeopardised by those of a collaborator. Current environmental schemes had a high level of up-take among study respondents (almost all had engaged in these at some point), which included penalties for non-compliance or early withdrawals, but these were individual-based schemes. Is it therefore highly important to the success of any co-ordinated or collaborative schemes that farmers not be required to work with specic other farmers, or penalised for action (or lack of) taken by other participants. It was also evident from study ndings that conversion to organic farming did not necessarily lead to closer collaboration with other local organic farmers; although closer collaboration was achieved in one hotspot location through joint participation in a farm shop, this farm shop was in competition with another farm shop (also organic), which led to antagonism between the two owners. Instead, relationships are based in positive neighbour interactions. Although organic farmers have more opportunities to meet other organic farmers (through membership in certication organisations, direct marketing activities and attendance at organic industry events), neighbour relationships continue to

remain important regardless of whether organic or conventional farming practices are adopted. I mean its really just your neighbours, thats all we speak to because. . .the neighbouring farms we have got something in common anyway, non organic or organic. (Frank, conventional farmer) Converts from conventional to organic farming (17 of 21 in this study) typically already have social and neighbour networks with conventional farmers, and do not lose these in the conversion process. It is therefore likely that conventional and organic farmers can collaborate in achieving environmental aims, if they have positive neighbour relationships. Discussion At a basic level, our study ndings suggest that in general, additional environmental gains can be achieved through clustering environmental action (in the form of certied organic farming), at no economic cost to the farmers involved (as long as they receive premium prices for the commodities involved). However, farmers may be reluctant to participate in co-ordinated efforts that require them to undertake activities which are not consistent with local farming standards, or make them vulnerable to defaulting by other participants. In this section we synthesize these ndings in relation to the literature and implications for multi-farmer collaboration including directions for future research, and note several limitations to study ndings. These results show that organic farming can enhance biodiversity, although the effects are mixed with some species groups responding positively, others negatively. Additional biodiversity benets are achieved when the area of organically managed land increases, especially for mobile taxa such as ower visiting insects, which experience their environment at larger scales and thus may benet from the higher oral abundance of aggregated organic farms through better pollen and nectar availability. This is corroborated by other studies (e.g. Holzschuh et al., 2008; Rundlof et al., 2008) that the clustering of organic farms may lead to additional

Table 6a Technical efciency results summary: monetary indices (). ORG Hot Cold HO CO < CONV HC CC

Notes: Arrows indicate direction of signicant (at 5% level) differences in technical efciency. HC = HO for 4/4 TE indices. CC > CO signicant for 4/4 TE indices. HO > CO signicant for 3/4 TE indices. HC < CC signicant for 2/4 TE indices.

510

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

Table 6b Technical efciency results parameters: monetary indices (signicance levels in brackets). TE1 ORG CONV TE2 ORG CONV TE3 ORG 0.90 0.84 CONV 0.92 0.93 TE4 ORG 0.92 0.84 CONV 0.92 0.93

Hot 0.69 0.73 0.80 0.76 0.68 0.83 0.72 0.84 Cold Signicance levels: TE1: a(1%), c(1%). TE2: a(1%), c(1%), d(5%). TE3: a(1%), d(1%). TE4: a(1%), d(1%). Notes: (a): Signicant difference (one tail-test) of TE between Conv/Org in coldspots. (b): Signicant difference (one tail-test) of TE between Conv/Org in hotspots. (c): Signicant difference (one tail-test) of TE between Hot/Cold in conventional agriculture. (d): Signicant difference (one tail-test) of TE between Hot/Cold in organic agriculture.

biodiversity benets. However, in this present study, farmland birds and hoveries appeared to fair better on conventional farms, albeit in organic hotspots. There are possible reasons for this: hoveries are known to prefer arable-dominated landscapes (Haenke et al., 2009; Jauker et al., 2009; Meyer et al., 2009), although in this study their larvae were more common on organic farms. This could reect a spillover effect where organic farms act as sources for the adults to spill over onto neighbouring conventional farms (Meyer et al., 2009). For birds, two recent studies have demonstrated that landscape characteristics such as the proportion of arable land and eld margin and hedge lengths, rather than farm management, are the prime determinant of numbers (Piha et al., 2007; Chamberlain et al., 2010). Furthermore, Dunn et al. (2010) indicates complex indirect effects: organic farms may be favourable to birds, but an increase in some birds (such as corvids like jays and rooks) can lead to negative effects on the birds on which they may prey. Efforts to address biodiversity through targeted schemes thus need to be undertaken at multiple scales, not solely at farm or neighbourhood level, in order to effectively manage for different species. It may also be useful to undertake co-ordinated efforts which provide a diversity of landscapes to allow for different species inventory to develop. The issue of uneven benets to biodiversity, and possible spillover effects raises an issue for evaluating the success of co-ordinated efforts. Several authors have called for changes to agri-environmental schemes in order to effectively pay farmers to produce biodiversity (Burton, 2004; Burton et al. 2008; Sutherland, 2010). There are two issues related to this: assessment and credibility. While successful production of less mobile species could be evaluated accurately on the farms involved, it is evident that successful production of highly mobile species would be very difcult to accurately assess, and therefore reward. This could also lead to problems between co-ordinating farmers, if some were rewarded and others were not. The second implication is of credibility of biodiversity measures. Farmland birds are one of the most visually evident forms of biodiversity. These were more likely to appear on conventional farms, even in organic hotspots. This could impact on the credibility of biodiversity measures as viewed by farmers, reducing their impetus to participate in associated schemes or undertake associated actions. Socio-cultural evaluation of the types of biodiversity which are recognised and valued by farmers could further advance thinking about the payment for production approach to agri-environmental schemes. The identication of a possible land management solution to the increasing problems of ooding in the UK is a highly important study nding. Overall, the analysis of the soils data shows that while it is possible to detect differences in the effect of soil type and land use (grassland/arable) on a number of the soil properties there are, with the exception of inltration rate, no signicant differences in the soil physical condition between organic and conventional farming practices. This is consistent with the earlier ndings of Stolze et al. (2000) and Gosling and Shepherd (2002) for the soil properties

where direct comparisons can be made. The analysis of water inltration rate led to an indication that if a number of farms within a neighbourhood collectively manage the grass elds with organic or less intensive conventional management this could lead to a reduction of runoff in the sub-catchment, agreeing with the ndings of Hess et al. (2010). However, Morris et al. (2010) report there is little hydrological evidence to verify this [benet], it is generally felt that policies that encourage retention of water in the landscape can contribute to ood risk mitigation especially for smaller, more frequent events. While this would be a highly useful impact of collaborative farming initiativesand an aspect that can be considered in the Catchment Sensitive Farming initiative (Defra, 2010)this effect needs further study at the sub-catchment level. If the water inltration ndings were to be supported in future research, the nancial implications for farmers would need to be considered, in order to inform the development of effective policy measures. Although it would not be necessary for farmers to formally convert to organic farming for benets to be achieved well managed6 non-organic grassland is likely to have a similar impact some form of nancial compensation may be necessary to ensure reducing stocking densities do not impact on farm protability. In addition, no differences in water inltration were found between organic and conventional arable elds. Achieving inltration benets may thus require converting arable land to grassland, an issue which has not been addressed in this study, but which certainly would impact on both farm protability and on biodiversity produced. Although the economics of grassland production were not part of the study, by denition reducing stocking capacity results in lower output, which must be compensated for through higher prices if yield value per hectare is to remain the same.7 Development of methods to economically assess the yield of grass elds would help to address the question of the cost of increasing extensive grassland within ood risk areas. This is an area for further research. The economic evaluation of producing biodiversity, alongside commodity yields, was a particularly novel research approach. It was also very promising in its indication that increased biodiversity does not necessarily have to come at an economic cost to the farmer. Further research is needed to assess other trade-offs (e.g. Hodgson et al., 2010 model the trade off between butteries and yield at the landscape level) and further strengthen the linkages between the observations made by different disciplines. Further assessment is also required on the impact of commodity price differences. In this present study, analysis included the premium prices received

6 Well managed, for these purposes, means following best soil management plans, reducing stocking densities and rotating livestock between elds to prevent trampling and poaching. 7 Ninety-nine percent of rough grazing, 93% of permanent pasture and 86% of grass (25 years old) remained untreated with pesticides in 2005 (Garthwaite et al., 2006, p. 10). Differences input costs between conventional and organic grassland are therefore likely to be minimal.

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512

511

for organic production. If premiums were not received, production of biodiversity may have come at an economic cost to the farmer. On the other hand, agri-environmental payments may provide adequate compensation. It is also important to note that organic farming (in its certied form) is only half a century old and does still suffer from a low critical mass. A signicant improvement of technical efciency was measured if land use increased from, on average, 117% (i.e. between coldspots and hotspots). It is possible that with future research, these additional premiums would not be necessary. In this paper we have not directly addressed collaborative effortactive co-operation between farmers to produce local environmental benets. However, in discussion of co-ordinated effort, the potential for active collaboration clearly follows. Findings from the research suggest that although farmers are generally willing to participate in environmental schemes as individuals, the structure of collaborative and co-operative options will need to be carefully considered. For farmers, it appears lower risk to undertake individual (rather than co-ordinated or collaborative) activities. Therefore, the rewards of co-ordinated or collaborative efforts would need to be higher (or the individual options removed) before this would become a viable prospect for most farmers. In terms of collaborative schemes, it also appears important that farmers are able to choose their collaborators (i.e. not be required to work with specic neighbours on the basis of geographic criteria such as adjacent land), owing to issues of trust. It also appears useful to build on existing hubs or hotspots of environmental activity, rather than attempting to start new activities in places where these have not been practiced in the past. Limitations There are a number of caveats to study ndings. In order to nd matched comparable farms, the sampling protocol had to be very restrictive in terms of the enterprise mix of the farms, the hotspot and coldspot concentration criteria and also the biophysical similarities of the matched landscapes. As a result the data and associated analysis provide a very rich and detailed assessment while controlling for many other potentially confounding factors such as soil and climate conditions. However, undertaking this matching approach led the farm selection towards mixed organic and conventional farms which are less typical of highly specialised farming systems which dominate other regions and landscapes of the UK. Farm selection may also have inuenced some of the research ndings. In terms of the economics of producing biodiversity, farm input costs and yields were on a like for like basis on matched farms. Although both organic and conventional farms exist at a range of acreages, conventional farms are likely to be larger than organic farms, potentially leading to economies of scale which were not reected in this assessment. Many conventional farms also classify as specialist cereal farms while the organic and conventional farms in our sample classify as mixed farms. Those specialist farms might be expected to have lower biodiversity levels than the mixed farms in the sample. Comparisons were made at eld scale between single crops (i.e. winter cereal elds and permanent pastures), but since organic farms, as in our sample, have a signicantly higher crop diversity than their conventional counterparts, additional benets may be expected at farm scale due to increased spatial and temporal land management heterogeneity in the organic system. The selection of certied organic farms also meant that premium prices were received for commodities produced. Although some conventional farms do use agri-environmental engagement as part of their marketing strategy, it cannot be assumed that

co-ordinated agri-environmental scheme participation would result in higher commodity prices without credible certication and consumers honouring the system nancially. We argue that the richness of the results derived from the very controlled sampling process has to be balanced with the caution that must be exercised when drawing general lessons from the analysis. Conclusions The study presented in this paper has led to a number of useful ndings from the perspectives of the disciplines involved: biodiversity achieved beyond farm scale, potential ood control through altered land management, the relative technical efciency of organic and conventional farms at different concentrations, and the normative impact of organic farming in different locales. However, the particular value of the study was bringing together these disciplinary perspectives on a shared research question and study sites, demonstrating a (qualied) case for multi-farmer coordination of environmental action. This is an approach which has rarely been attempted in the past. In addition to identifying trade-offs in the management of land at neighbourhood level, the research also involved trade-offs for the researchers involved. The purpose of this paper was not to evaluate multidisciplinary research, but we would be remiss in not pointing out the considerable time investment required to successfully undertake this multifaceted approach. This time is spent in seeking to understand the nature of the research conducted in the collaborating disciplines, which is ongoing throughout the research process. A key learning experience for the researchers was that it is not simply the nature of the subjects under study which differ (e.g. soil, water, insects, birds, balance sheets, social norms), use of terms and the methods by which they are evaluated (qualitative and quantitative) but also the characteristics of the laboratories in which research is conducted. It was attempting to work in each others laboratories which proved most challenging. Identifying matched study farms on the basis of the criteria important to each discipline proved an enormous task, even with the relatively small number of disciplines involved. Thus while the benets of multidisciplinary research can be said to reect the adage that the whole is greater than the sum of its parts, so too is the amount of time required to achieve this whole. Acknowledgements The research was funded as part of the UK Research Councils Rural Economy and Land Use Programme (RELU) (RES-227-250006). RELU is a collaboration between the Economic and Social Research Council, the Natural Environment Research Council and the Biotechnology and Biological Sciences Research Council, with additional funding from DEFRA and the Scottish Government. The authors also wish to thank the anonymous reviewer for comments on a previous version of this paper. References
Allen, S.E. (Ed.), 1989. Chemical Analysis of Ecological Materials. , 2nd ed. Blackwell, Oxford. Batey, T., 1988. Soil HusbandryA Practical Guide to the Use and Management of Soils. Aberdeen University Press, Aberdeen. Bengtsson, J., Ahnstrom, J., Weibull, A.C., 2005. The effects of organic agriculture on biodiversity and abundance: a meta-analysis. J. Appl. Ecol. 42, 261269. Benton, T.G., Bryant, D.M., Cole, L., Crick, H.Q.P., 2002. Linking agricultural practice to insect and bird populations: a historical study over three decades. J. Appl. Ecol. 39 (4), 673687. Benton, T.G., Vickery, J.A., Wilson, J.D., 2003. Farmland biodiversity: is habitat heterogeneity the key? Trends Ecol. Evol. 18 (4), 182188.

512

L.-A. Sutherland et al. / Land Use Policy 29 (2012) 502512 Hole, D.G., Perkins, A.J., Wilson, J.D., Alexander, I.H., Grice, P.V., Evans, A.D., 2005. Does organic farming benet biodiversity? Biol. Conserv. 122, 113130. Holzschuh, A., Steffan-Dewenter, I., Kleijn, D., Tscharntke, T., 2007. Diversity of ower-visiting bees in cereal elds: effects of farming system, landscape composition and regional context. J. Appl. Ecol. 44, 4149. Holzschuh, A., Steffan-Dewenter, I., Tscharntke, T., 2008. Agricultural landscapes with organic crops support higher pollinator diversity. Oikos 117, 354361. Jauker, F., Diekotter, T., Schwarzbach, F., Wolters, V., 2009. Pollinator dispersal in an agricultural matrix: opposing responses of wild bees and hoveries to landscape structure and distance from main habitat. Landsc. Ecol. 24, 547555. Lampkin, N., Padel, S., 1994. Economics of Organic Farming. An International Perspective. CAB International, Oxon, UK. Lexartza- Artza, I., Wainwright, J., 2009. Hydrological connectivity: linking concepts with practical implications. Catena 79 (2), 146152. MAFF, 2000. Fertiliser Recommendations for Agricultural and Horticultural Crops (RB209), 7th ed. The Stationary Ofce, London. Meyer, B., Jauker, F., Steffan-Dewenter, I., 2009. Contrasting resource-dependent responses of hovery richness and density to landscape structure. Basic Appl. Ecol. 10, 178186. Mishra, S.K., Singh, V.P., 2003. Soil Conservation Service Number (SCS-CN) Methodology. Kluwer Academic Publishers, Boston. Morris, J., Hess, T., Posthumus, H., 2010. Agricultures role in ood adaptation and mitigation: policy issues and approaches. Craneld University OECD Report. Norton, L., Johnson, P., Joys, A., Stuart, R., Chamberlain, D., Feber, R., Firbank, L., Manley, W., Wolfe, M., Hart, B., Mathews, F., Macdonald, D., Fuller, R.J., 2009. Consequences of organic and non-organic farming practices for eld, farm and landscape complexity. Agric. Ecosyst. Environ. 129, 221227. Offermann, F., Nieberg, H., Zander, K., 2009. Dependency of organic farms on direct payments in selected EU member states: today and tomorrow. Food Policy 34 (3), 273279. Pascual, U., 2005. Land-use intensication potential in slash-and-burn farming through improvements in technical efciency. Ecol. Econ. 52, 497511. Piha, M., Tiainen, J., Holopainen, J., Vepsalainen, V., 2007. Effects of land-use and landscape characteristics on avian diversity and abundance in a boreal agricultural landscape with organic and conventional farms. Biol. Conserv. 140, 5061. PPDB Footprint Report, 2009. Pesticide database accessed via http://sitem.herts. ac.uk/aeru/footprint/en/index.htm (last visited 28.06.09). Roschewitz, I., Gabriel, D., Tscharntke, T., Thies, C., 2005. The effects of landscape complexity on arable weed species diversity in organic and conventional farming. J. Appl. Ecol. 42, 873882. Rundloef, M., Smith, H.G., 2006. The effect of organic farming on buttery diversity depends on landscape context. J. Appl. Ecol. 43, 11211127. Rundlof, M., Bengtsson, J., Smith, H.G., 2008. Local and landscape effects of organic farming on buttery species richness and abundance. J. Appl. Ecol. 45, 813820. Shephard, R., 1970. Theory of Costs and Production Functions. Princeton University Press, Princeton. Smith, E., Marsden, T., 2004. Exploring the limits to growth in UK organics: beyond the statistical image. J. Rur. Stud. 20 (3), 345357. Smith, L.P., Trafford, B.D., 1976. Climate and Drainage. HMSO, London. Soil Association, 2010. Organic Market Report 2010. Available at: http://www. soilassociation.org/LinkClick.aspx?leticket=bTXno01MTtM=&tabid=116 (accessed 30.07.10). Stolze, M., Piorr, A., Hring, A., Dabbert, S., 2000. The Environmental Impacts of Organic Farming in Europe in Organic Farming in Europe. Economics and Policy, vol. 6. Hago Druck and Medien, Germany. Available at: http://orgprints.org/8400/. Sutherland, L.A., 2010. Environmental grants and regulations in strategic farm business decision-making: a case study of attitudinal behaviour in Scotland. Land Use Policy 27 (2), 415423. Sutherland, L.A., Burton, R., 2011. Good farmers, good neighbours?: the role of cultural capital in social capital development a Scottish farming community. Sociologia Ruralis 51, 238255. Tate, W.B., 1994. The development of the organic industry and market: an international perspective. In: Lampkin, N.H., Padel, S. (Eds.), Notes on Economics of Organic Farming. An International Perspective. CAB International, Oxon, UK, pp. 1126 (Chapter 2). Tovey, H., 1997. Food, environmentalism and rural sociology: on the organic farming movement in Ireland. Soc. Ruralis 37 (1), 2137. Tscharntke, T., Klein, A.M., Kruess, A., Steffan-Dewenter, I., Thies, C., 2005. Landscape perspectives on agricultural intensication and biodiversityecosystem service management. Ecol. Lett. 8, 857874. Whittingham, M.J., 2007. Will agri-environment schemes deliver substantial biodiversity gain, and if not why not? J. Appl. Ecol. 44, 15.

Billeter, R., Liira, J., Bailey, D., Bugter, R., Arens, P., Augenstein, I., 2008. Indicators for biodiversity in agricultural landscapes: a pan-European study. J. Appl. Ecol. 45, 141150. Burton, R., 2004. Seeing through the good farmers eyes: towards developing an understanding of the social symbolic value of productivist behaviour. Soc. Ruralis 44 (2), 195215. Burton, R., Kuczera, C., Schwarz, G., 2008. Exploring farmers cultural resistance to voluntary agri-environmental schemes. Soc. Ruralis 48 (1), 1637. Burton, R.F., Wilson, G.A., 2006. Injecting social psychology theory into conceptualisations of agricultural agency: towards a post-productivist farmer self-identity? J. Rur. Stud. 22 (1), 95115. Chamberlain, D.E., Fuller, R.J., Bunce, R.G.H., Duckworth, J.C., Schrubb, M., 2000. Changes in the abundance of farmland birds in relation to the timing of agricultural intensication in England and Wales. J. Appl. Ecol. 37, 771788. Chamberlain, D.E., Joys, A., Johnson, P.J., Norton, L., Feber, R.E., Fuller, R.J., 2010. Does organic farming benet farmland birds in winter? Biol. Lett. 6, 8284. Charnes, A., Cooper, W., Rhodes, E., 1978. Measuring the efciency of decisionmaking units. Eur. J. Oper. Res. 2, 429444. Dauber, J., Purtauf, T., Allspach, A., Frisch, J., Voigtlander, K., Wolters, V., 2005. Local vs. landscape controls on diversity: a test using surface-dwelling soil macroinvertebrates of differing mobility. J. Biogeogr. 14, 213221. Department for Environment and Rural Affairs (Defra), 2010. Catchment Sensitive Farming. http://www.defra.gov.uk/foodfarm/landmanage/water/csf/index.htm (accessed 27.08). Defra, 2006. Agriculture in the United Kingdom. National Statistics. Table 3.2. Dunn, J.C., Hamer, K.C., Benton, T.G., 2010. Fear for the family has negative consequences: indirect effects of nest predators on chick growth in a farmland bird. J. Appl. Ecol. 47 (5), 9941002. Eurosta. 2011. Certied organic crop area. Accessed 7 October 2011. http://appsso.eurostat.ec.europa.eu/nui/setupModifyTableLayout.do. Fuller, R.J., Norton, L.R., Feber, R.E., Johnson, P.J., Chamberlain, D.E., Joys, A.C., Mathews, F., Stuart, R.C., Townsend, M.C., Manley, W.J., Wolfe, M.S., Macdonald, D.W., Firbank, L.G., 2005. Benets of organic farming to biodiversity vary among taxa. Biol. Lett. 1, 431434. Gabriel, D., Roschewitz, I., Tscharntke, T., Thies, C., 2006. Beta diversity at different spatial scales: plant communities in organic and conventional agriculture. Ecol. Appl. 16, 20112021. Gabriel, D., Sait, S.M., Hodgson, J.A., Schmutz, U., Kunin, W.E., Benton, T.G., 2010. Scale matters: the impact of organic farming on biodiversity at different spatial scales. Ecol. Lett. 13, 858869. Gabriel, D., Thies, C., Tscharntke, T., 2005. Local diversity of arable weeds increases with landscape complexity. Perspect. Plant Ecol. Evol. Syst. 7, 8593. Garthwaite, D.G., Thomas, M.R., Anderson, H.M., Battersby, A., 2006. Pesticide Usage Survey Report 210. Grassland and Fodder Crops in Great Britain 2005 (Including Aerial Applications 20032005). Department for Environment, Food & Rural Affairs & Scottish Executive Environment & Rural Affairs Department. Available at: http://www.fera.defra.gov.uk/plants/pesticideUsage/grassland2005.pdf (accessed 30.07.10). Gosling, D., Shepherd, M., 2002. Theory and reality of organic soil fertilityorganic matter. In: Powell UK Organic Research 2002 (Ed.), Proceedings of the COR Conferences. Aberystwyth, 2628th March 2002, pp. 137138. Haenke, S., Scheid, B., Schaefer, M., Tscharntke, T., Thies, C., 2009. Increasing syrphid y diversity and density in sown ower strips within simple vs. complex landscapes. J. Appl. Ecol. 46, 11061114. Hathaway-Jenkins, L.J., Sakrabani, R., Pearce, B., Whitmore, A.P., Godwin, R.J., 2011. A comparison of soil and water properties in organic and conventional farming systems in England. Soil Use Manag. 27 (2), 133142. Haynes, R.J., Swift, R.S., Stephen, R.C., 1991. Inuence of mixed cropping rotations on organic matter content, water stable aggregation and clod porosity in a group of soils. Soil Tillage Res. 19, 7787. Hendrickx, F., Maelfait, J.P., Van Wingerden, W., Schweiger, O., Speelmans, M., Aviron, S., Augestein, I., Billeter, R., Bailey, D., Bukacek, R., Burel, R., Diekotter, T., Dirksen, J., Herzog, F., Liira, J., Roubalova, M., Vandomme, V., Bugter, R., 2007. How landscape structure, land-use intensity and habitat diversity affect components of total arthropod diversity in agricultural landscapes. J. Appl. Ecol. 44, 340351. Hess, T.M., Holman, I.P., Rose, S.C., Rosolova, Z., Parrott, A., 2010. Estimating the impact of rural land management changes on catchment runoff generation in England and Wales. Hydrol. Processes 24, 13571368. Hodgson, J., Kunin, W.E., Thomas, C.D., Benton, T.G., Gabriel, D., 2010. Comparing organic farming and land sparing: optimising yield and buttery populations at a landscape scale. Ecol. Lett., doi:10.1111/j.1461-0248.2010.01528.x.