• AES;
  • Agri-Environment Scheme;
  • habitat quality;
  • hill farm;
  • in-bye;
  • moorland;
  • Peak District;
  • species richness;
  • upland;
  • UK


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Modern farmed landscapes have witnessed substantial losses in biodiversity principally driven by the ecological changes associated with agricultural intensification. The causes of declines are often well described, but current management practices seem unlikely to deliver the EU-wide policy objective of halting biodiversity losses.
  • 2
    Available evidence suggests that property-scale factors can be influential in shaping patterns of biodiversity; however, they are rarely included in studies. Using 44 upland farms in the Peak District, northern England, we investigate the roles of ecological, agricultural and socio-economic factors in determining avian species richness, for the first time incorporating information from all three influences.
  • 3
    Although we might expect that habitat quality would be the main factor affecting species richness, these variables had little influence. The landscape context of each property was unimportant in explaining any of the three measures of species richness (Total, Upland and Conservation Concern) used here. Within-property habitat quality did explain 42% of the variation in richness of upland specialist species, but had no influence on Total or Conservation Concern Richness.
  • 4
    Socio-economic circumstances of farms alone accounted for 24% of the variation in Total Richness, with land tenure and labour inputs important predictors of avian diversity. However, net income, rental value and the level of Agri-Environment Scheme (AES) payments did not play a role in predicting species richness.
  • 5
    Farm management variables, including many of the main prescriptions outlined in AES, accounted for 23% of the variation in the richness of species of Conservation Concern, but less than 10% for Total Richness. However, no farm management variable alone was shown to offer better predictive power of avian species richness than random.
  • 6
    Synthesis and applications. The agricultural landscape is managed by a mosaic of landowners, all of whom can influence biodiversity conservation. We demonstrate that variation at the property-scale in habitat, management and socio-economics can feed into determining patterns of biodiversity. Currently, farmland conservation policy largely assumes that socio-economic barriers and financial costs of implementing conservation measures are constant. Incorporating a consideration of the varying circumstances of individual properties into policy design is likely to result in substantial biodiversity gains.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Modern farmed landscapes account for nearly half of the land cover in the European Union (FAO 2003) and yet are severely depleted in both habitat heterogeneity and biodiversity (Kareiva et al. 2007). Indeed, recent decades have witnessed substantial losses in biodiversity in the wider countryside of this region, principally driven by the ecological changes associated with intensification of agricultural production (Benton et al. 2002; Robinson & Sutherland 2002; Donald et al. 2006).

Patterns of biodiversity across the landscape have been well studied (e.g. Marini et al. 2008; Rundlöf, Bengtsson & Smith 2008). The causes of declines are often well described (e.g. Chamberlain et al. 2000; Preston et al. 2002; Robinson & Sutherland 2002), and some of the mechanisms understood (e.g. Baines 1988; Wickramasinghe et al. 2003; Smith et al. 2004; Pocock & Jennings 2008). Management prescriptions are available, in the form of Agri-Environment Schemes (AES), which aim to halt those declines (Defra 2005a,b). Nonetheless, the results of AES in terms of biodiversity gain are equivocal (Kleijn & Sutherland 2003; Kleijn et al. 2006), calling into question whether current designs of AES will deliver the EU-wide policy objective of halting biodiversity loss (Whittingham 2007).

Although rarely included in biodiversity studies (but see Hudson 1992; Tharme et al. 2001), property-scale factors are likely to be highly influential in driving patterns of biodiversity, not least because management actions are undertaken at this scale. Indeed, a range of ecological and non-ecological factors will almost certainly influence between-property variation in biodiversity. Here we investigate the importance of habitat, management and socio-economic variables as predictors of biodiversity at the property scale.

This question is particularly pertinent for avian diversity in the UK uplands at this time, as these areas continue to experience widespread habitat change (Haines-Young et al. 2003). Agricultural land management resulted in a 7% increase in the area of improved grassland, indicating a continuation of agricultural intensification in the uplands that is less apparent in lowland areas. The ecological consequences of such a dramatic shift in land-use are marked, and substantial declines in upland breeding bird populations continue (Sim et al. 2005).

Differences in habitat type and quality are well known to shape the occurrence of avian species in the upland landscape (e.g. Stillman & Brown 1994; Tharme et al. 2001). We would therefore expect that the most proximal factors influencing variation in species richness at a property level would be delimited by habitat availability and quality, both on the property concerned and in the surrounding landscape. In addition, altering farm management practices (such as stocking rates or chemical use) can have a profound influence on the extent and quality of habitat and food resources available to the avian community (e.g. Peach et al. 2001; Whittingham et al. 2007). We would therefore also predict that variation in farming practice at the property level should explain variation in species richness as well. Empirical evidence also exists for coarse-scale patterns of correlation between biodiversity and socio-economic factors (e.g. Huby et al. 2006), and studies show that land ownership can influence conservation decision-making (e.g. Ando & Getzner 2006). Moreover, the willingness of land managers to engage in environmentally sensitive management practices depends on a range of socio-economic conditions (e.g. Willock et al. 1999; Vanslembouck, Van Huylenbroeck & Verbeke 2002), some of which are determined by agricultural, environmental and rural development policy. Hence, it is possible that variation in the socio-economic conditions on a farm will influence decisions that are made about farming system and management practices, which in turn can be expected to affect species richness.

In common with Europe as a whole (Donald et al. 2006), farming remains the dominant land-use in the UK uplands, although it operates on the margins of agricultural productivity. Recently, hill farm incomes in the UK have fallen dramatically in response to lower lamb and beef prices (Defra 2005c) and the viability of upland farms often depends on core subsidy support (such as the Single Farm Payment) and on AES payments (Peak District Rural Deprivation Forum 2004; Acs et al. 2008). Further, as with other rural areas, changing farming practice in the uplands has gone hand in hand with marked shifts in the socio-economic make up of rural areas, with a reduction in the number of full time agricultural workers, an increase in the age of the population, farm amalgamations and continued social deprivation with income growth lagging behind that in much of the rest of society (Defra 2004).

Upland areas represent a dynamically changing ecological and socio-economic environment where current management and knowledge has so far proven inadequate to address biodiversity losses. In this study, using the Peak District National Park of northern England as a case study, we investigate the roles of ecological, agricultural and socio-economic factors in patterning avian species richness in the uplands, incorporating information from all three influences at the property-scale for the first time. We argue that a property-scale focus is particularly valuable for understanding ecological outcomes, since this is the scale at which current agri-environmental policy operates.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

study system

This study focussed on the uplands of the Peak District National Park in northern England (Fig. 1). The Peak District is a region of hills characterized by heather Calluna vulgaris dominated moorland managed principally for pastoral farming and grouse Lagopus lagopus shooting, with almost all of the moorland receiving additional protected area designation such as the Southern Pennines Moors Special Protection Area (SPA) designated to protect the region's breeding bird interests. Farming in the Peak District operates on the margins of profitability and is reliant on agricultural subsidies (Acs et al. 2008) and the long-term economic sustainability of upland farming in its current form remains in doubt.


Figure 1. A map of the Peak District National Park in northern England. The stippled area indicates the extent of moorland; the shaded area represents a buffer 2 km wide around the major moorland blocks. All study farms had their main land-holding in this moorland fringe zone. Inset shows location of the Peak District in Britain.

Download figure to PowerPoint

Although agriculture is the dominant land-use, other drivers of changes in biodiversity in the Peak District include moorland management using controlled burning, which has altered the region dramatically in recent years (Yallop et al. 2006), aerial pollution and climate change. Historically, the Peak District has been subject to high amounts of airborne pollution (NEGTAP 2001). The impacts of these pollutants on the vegetation community have resulted in the almost complete loss of sphagnum and moss communities on moorland and bogs (Lee 1998). Climate warming is viewed as a major driver of change in all global ecosystems (Millennium Ecosystem Assessment 2005). The Peak District lies at the southern and eastern margins of climatic suitability for upland bog formation and is therefore likely to be severely impacted.

data collection

In order to characterize property-scale avian species richness, 44 farms were selected whose main landholding fell within 2 km of moorland (Fig. 1). Three properties did not fall within the National Park, and a further 11 farms were outside the boundaries of the North Peak or South West Peak Environmentally Sensitive Areas (ESA). Only five farms included farmland (as opposed to moorland) that was covered by additional protected area designations. Property maps were obtained from the farmer, and transect routes planned prior to any bird surveys being conducted, based on the size and shape of the landholding and suitable access points. To minimize the potential for recording birds outside the survey farm, transects were placed 200 m from a property boundary. Birds were only included as present if they were seen or heard within the property, irrespective of the distance from the transect. Where needed, parallel transects were placed 400 m apart to avoid double-sampling the same parts of the farm. In this situation, birds were only recorded within 200 m of the transect line. Bird surveys were carried out on two separate visits between 28 March and 5 July 2007, with the second visit at least 6 weeks after the first. To ensure that the maximum number of species was encountered, visits began between 1 and 3 h after sunrise.

A list of all bird species encountered on each farm during both visits was compiled. The number of species observed was used directly as the measure of species richness. Species were classified into two further groups with greater conservation relevance: Upland Species and Species of Conservation Concern (Supporting Information, Appendix S1). The habitat specialist Upland Species group consisted of species that have a predominantly upland breeding distribution, based on the UK Breeding Bird Atlas (Gibbons, Reid & Chapman 1993). The Conservation Concern species group comprised species that are either Amber or Red listed (Gregory et al. 2002), appear on the UK BAP list (Biodiversity Reporting and Information Group 2007) or are qualifying features for the South Pennine Moors SPA (Stroud et al. 2001).

Habitat variables were collected from surveyed fields within each farm (Table 1). These variables were those that have been shown to influence avian species richness and population size for a variety of species in the UK uplands (e.g. Baines 1988; Robson & Percival 2002; Pearce-Higgins & Yalden 2003) and for farmland birds in general (e.g. Atkinson et al. 2005; Whittingham et al. 2005). The landscape context within which each property was found was characterized by calculating the proportion of six different habitat types (moorland, woodland, arable, inland water, urban/rural developed land and grassland) based on the Land Cover Map 2000 (Haines-Young et al. 2000) in a 500-m buffer around each property.

Table 1.  Definitions of variables used in the analysis of the patterns of avian species richness across hill farms in the Peak District, northern England
VariableMedianLower–upper quartileDescription
(a) Habitat
Intensive grass0·880·72–0·97Proportion of surveyed fields that were improved
Mowed land0·250·03–0·53Proportion of surveyed fields that were cut for silage or hay
Vegetated boundaries0·080·03–0·15Proportion of field boundaries that were vegetated (hedges, woods)
Trees14262–270Total number of trees within surveyed fields
Sheep 8821–175Total number of adult sheep within surveyed fields
Cows 130–36Total number of adult cows within surveyed fields
Rush cover0·020·00–0·05Proportion of surveyed fields with rush (mainly Juncus effusus) cover
Wet features0·130·00–0·30Proportion of surveyed fields with wet features (ditches, ponds, streams)
(b) Landscape context
Arable0·010·00–0·06Proportion of arable within 500 m of the farm.
Moorland0·060·01–0·17Proportion of moorland within 500 m of the farm.
Grassland0·640·57–0·77Proportion of grassland within 500 m of the farm.
Urban/rural developed0·020·00–0·05Proportion of urban/rural developed land within 500 m of the farm.
Inland water0·000·00–0·01Proportion of inland water within 500 m of the farm.
Woodland0·090·03–0·15Proportion of woodland within 500 m of the farm.
(c) Farm management
Farm typeNot applicableNot applicableWhether the farm was a sheep (7 farms), cattle (6 farms) or mixed (31 farms) enterprise
Rough grazing12·603·00–29·95Area (ha) of the farm that the farmer stated was managed as rough grazing
LU/Ha0·980·70–1·46Density of livestock units ( ha−1) on the farm
Predator control 104·50–33Number of days of predator control a year carried out on the property by the farmer, gamekeeper (whether employed on that property or not) or other professional person.
Cutting datesNot applicableNot applicableWhether the land was cut before (18 farms) or after (21 farms) mid-July
Number of cutsNot applicableNot applicableNumber of cuts (between zero and three) taken. No cuts taken on 5 farms, one cut on 25 farms, two cuts on 12 farms and three cuts taken on two farms.
Fertilizer input43·0018·00–119·50Nitrogen input (kg ha−1) from fertilizer and manure
(d) Socio-economic
Ownership0·300·03–0·72Proportion of the farm that was owned
Farm workers1·701·06–2·29Number of workers on the farm, calculated from the questionnaire
On farm income0·820·55–1·00Proportion of farm income from farming
Net farm income6067−6758·50 to 33189·50Net farm income (£), calculated from the questionnaire
AES payment75001353·50–16285·00Total AES payment (£), taken from the questionnaire
Rental value44·1023·25–84·00Rental value of the farm (£ ha−1), calculated from overall questionnaire returns

Farm management and socio-economic characteristics of the same 44 farms were gathered using a questionnaire-based survey delivered in person during farm visits by experienced professional farm business researchers (Table 1). The purpose of the socio-economic survey was to investigate how land is managed on hill farms, what resources were available to farmers, and how land-use and resources related to farm incomes (Acs et al. 2008). The survey included a mix of closed and open-ended questions that covered the landholding, production activities (e.g. livestock numbers, labour and fertilizer use), other management activities (e.g. activities complying with different AES), and financial data such as input costs, output prices and subsidy payments received.

Initial contact with suitable farms was established through the Rural Business Research Unit at the University of Nottingham, local farming and conservation organizations, and by word of mouth. Sixty-seven farmers were contacted directly by telephone and asked whether they wished to participate in the research programme, 55 of whom indicated their interest. Of these, 47 farmers were visited and 44 were included in the economic analyses and ecological surveys. The remaining three did not actively farm themselves, with all land either rented out or used for non-agricultural activities. In total, 10 person-months were spent making contact with the farming community, establishing direct relationships with farmers and delivering the questionnaires in a face-to-face interview.

Variables recorded in the questionnaire were categorized as covering farm management practices, or describing the socio-economic status of the farm. Farm management variables pertained to the farming system, nutrient input, livestock units, areas of key land types, information relating to AES prescriptions, as well as whether predator control (known to be important for some bird species; Holt et al. 2008) was carried out. Socio-economic variables included data on patterns of land ownership, known changes in the socio-economic make-up of rural areas and information on the economic performance and subsidy uptake of each property. These factors were considered potentially to play a role in influencing property-scale management decisions, and hence, avian species richness.

Many of the farm management and socio-economic data were usable directly from the questionnaire. However, several variables required further calculation; these included density of livestock units, fertilizer input, number of farm workers, rental value of the farm and net farm income. The density of grazing livestock units was based on total livestock numbers reported by the farmer weighted by the type of livestock and proportional to the farm area. Fertilizer input represents the total nitrogen (N) use per hectare, which was based on the N content of fertilizer and manure applied. The number of farm workers was calculated according to the labour input from the farmer, family labour and hired labour working either full- or part-time. Net farm income was derived from total returns from agricultural production and subsidy payments minus variable costs, such as those for fertilizer, sprays and feed.

data analysis

All analyses were carried out at the level of the property holding, excluding any unenclosed extensively grazed moorland. Explanatory variables consisted of four broad classes: (i) Habitat, (ii) Landscape context, (iii) Farm management, and (iv) Socio-economic. Each of the response variables (Total Species Richness, Upland Species Richness and Conservation Concern Species Richness) were modelled with the four sets of variables separately in order to avoid over-specifying any one model.

The Information Theoretic approach (Burnham & Anderson 2002; Johnson & Ohmland 2004) was used to model these data based on Akaike Information Criteria (AIC). All possible subsets of the variables were modelled using a Generalized Linear Model with Poisson errors; species richness at each site was the dependant variable and explanatory variables were transformed appropriately. For the complete set of models, AIC, the difference in AIC for that model relative to the best-fitting model with the minimum AIC (termed ΔAIC), the Akaike weight (termed wi) and R2 were all calculated. The best-fitting model was defined as that with the lowest AIC. Models that differ by less than 2 AIC units have substantial support in terms of explaining the data (Burnham & Anderson 2002).

The probability of each individual explanatory variable appearing in the best-fitting model was also calculated (termed k). However, poor explanatory variables can still have high selection probabilities. A single randomly generated variable was therefore added to the existing data set (Whittingham et al. 2005). One hundred model sets were generated, and k for the random variable was calculated. Explanatory variables that do not offer predictive power significantly different from random have a probability (k) falling within the 95% confidence intervals of this random variable. All analyses were performed in R version 2·6·2 (R Development Core Team 2008).

To investigate each of the four classes of explanatory variables together, the two variables from each with the highest k were included in a further model, termed the Joint Model. In total, therefore, five best-fitting models (one for each variable type and the Joint Model) were generated for each measure of species richness. From these, the model that best predicted the pattern of avian species richness was chosen based on minimum AIC.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

There were between 13 and 45 bird species observed on each farm and a total of 97 species across all 44 study farms. Twenty-one upland specialist species were observed (range: one to eight) and 43 Conservation Concern Species (range: two to 10). On average, 95·0 ha (SD 66·7 ha) of farmland was surveyed per property, with an average 1651 m (SD 561 m) of transect walked. Surveyed farm area and transect length were positively correlated (ρ = 0·697, P < 0·001). There was no bivariate relationship between species richness and the area of farmland surveyed (e.g. for Total Species Richness, R2 = 0·01, P > 0·05), nor length of transect walked (e.g. for Upland Species Richness, R2 = 0·03, P > 0·05). Neither transect length nor farm area was therefore included in any of the models reported, to reduce model complexity. Indeed, when farm area was included in the models there were no substantive changes to the model outcomes. Colinearity between explanatory variables was investigated using correlation matrices. Although associations were apparent, they were not sufficient to preclude their inclusion into the modelling process. Within each model, Variance Inflation Factors (VIFs) were within accepted norms, indicating that problematic levels of multi-colinearity were not present (Myers 1990). One farm was excluded from data analyses, as it was an extreme outlier in terms of farm size, structure and management.

Habitat variables explained 42% of the variation in upland species richness (Table 2). For these upland specialists, the quality of the habitat was important, with fewer species where more fields were mowed for silage or hay, and more species with increasing numbers of cows and proportion of field with rush cover (Table 3; Supporting Information, Appendix S2). All three variables offered explanatory power that was better than random. Habitat characteristics did not offer any explanatory power for Conservation Concern or Total Species Richness (R2 = 0), and k for these variables did not differ from random, further indicating the relative unimportance of habitat variables in predicting Total and Conservation Concern Species Richness.

Table 2.  Information theoretic results for the relationship between species richness of three different avian species groups (Total, Upland and Conservation Concern) in the Peak District in relation to each set of explanatory variables. The proportion of the variation explained (R2), number of models appearing in the 95% set (95% set) and number of models appearing in the ΔAΙC < 2 set. The overall Best Model (based on lowest AIC) for each species group is highlighted in bold.
Explanatory variable setSpecies richnessR2AIC95% setΔAΙC < 2
  • *

    – indicates that the null model appeared in the ΔAΙC < 2 model set, suggesting that this represents a plausible alternative to the best model. The null model was also the Best Model for Conservation Concern Species Richness for the Habitat variable set and for Total and Conservation Concern Species Richness for the landscape context variable set.

(a) HabitatTotal0·00300·6116412*
Conservation Concern0·00223·62137 6*
(b) Landscape contextTotal0·00301·1041 5*
Conservation Concern0·00223·6246 7*
(c) Farm managementTotal0·09300·0182 5*
Conservation Concern0·23223·078712*
(d) Socio-economicTotal0·24291·34274
Upland0·10177·7946 7*
Conservation Concern0·18222·3743 6*
(e) Joint modelsTotal0·27289·94332
Conservation Concern0·18222·315229*
Table 3.  Akaike weights (k) for each explanatory variable included in species richness models for Total Species Richness (Total), Upland Species Richness (Upland) and Conservation Concern Species Richness (Conservation Concern) on hill farms in the Peak District. The Akaike weights indicate the probability that each variable is included in the best-fitting model. Figures in bold indicate variables that appear in the best model. For each species group, the two explanatory variables with the highest k from each variable set (excluding landscape context) were included in the Joint Model. Relationship indicates the direction of any relationship between species richness and variables featuring in the best model. For parameter estimates, see Supporting Information, Appendix S2. The random variable shows the 95% confidence intervals for the probability k that a random unrelated variable appears in the best model.
Variable setTotalUplandConservation concern
(a) Habitat
Intensive grass0·22 0·21 0·23 
Mowed land0·40 0·750·30 
Vegetated boundaries0·49+0·21 0·22 
Trees0·49+0·22 0·22 
Sheep0·23 0·26 0·41 
Cows0·28 0·89+0·27 
Rush cover0·27 0·60+0·23 
Wet features0·34 0·43 0·33 
Random Variable0·21–0·90 0·20–0·48 0·22–0·69 
(b) Landscape context
Arable0·24 0·31 0·43 
Moorland0·41 0·53+0·40 
Grassland0·30 0·38 0·38 
Urban/rural developed0·23 0·45 0·30 
Inland water0·23 0·490·28 
Woodland0·31 0·27 0·25 
Random variable0·22–0·93 0·22–0·81 0·22–0·69 
(c) Farm management
Farm type0·34 0·31 0·45Fewest species on sheep-only farms
Rough grazing0·29 0·33 0·22
LU ha−10·23 0·28 0·27 
Predator control0·50+0·52+0·43+
Cutting dates0·23 0·24 0·25 
Number of cuts0·23 0·28 0·30 
Fertilizer input0·56+0·33 0·51+
Random variable0·21–0·71 0·22–0·65 0·21–0·56 
(d) Socio-economic
Farm workers0·96+0·33 0·63+
On farm income0·550·32 0·31 
Net farm income0·22 0·24 0·23 
AES payment0·30 0·35 0·23 
Rental value0·23 0·28 0·26 
Random variable0·21–0·78 0·22–0·67 0·22–0·67 
(e) Joint model
Mowed land 0·84 
Vegetated boundaries0·69+  
Sheep  0·36 
Cows 0·90+  
Wet features  0·38 
Farm type  0·39 
Predator control0·22 0·33  
Fertilizer input0·33 0·23 0·48+
Ownership0·900·42 0·58
Farm workers0·80+ 0·48 
AES payment 0·43+ 
Random variable0·21–0·73 0·21–0·52 0·21–0·58 

Landscape context variables offered little explanatory power for all three measures of species richness. Indeed, the null model was the best-fitting model for both Total and Conservation Concern Species Richness. For Upland Species Richness, landscape context explained 13% of variation. However, no variables had a k that differed from random. As landscape context variables were poor predictors across the range of species richness measures, they were not included in the final Joint Model.

Farm management explained 23% of the variation in Conservation Concern Species Richness, this figure falling to 9% and 7% for Total and Upland Species Richness, respectively. For all three richness measures, no variable offered explanatory power that was better than random and the null model appeared in the ΔAIC < 2 model set in each case. Both results indicate the lack of confidence that should be placed in the described relationships.

Socio-economic variables explained 24% of the variation in Total Species Richness across the study farms. Variables included in the best model were the number of workers on a farm, the proportion of the farm that is owned and the proportion of farm income generated from farming. Fewer species were present where more of the farm income was generated on-farm and where more of the land area of the farm was owned. Conservation Concern Species Richness mirrored Total Species Richness with regards to the socio-economic variables, with 18% of the variation explained by ownership levels and farm workers, although no relationship was found with on-farm income. Socio-economic variables only offered limited explanatory power (10%) for Upland Species Richness. Net farm income, rental value and AES payments did not appear in any best-fitting model.

Two variables from each explanatory variable set with the highest Akaike weight for each species group were included in a Joint Model in order to investigate the relative importance of the variable sets (Table 3). The best-fitting model for the Joint Model for both Upland and Conservation Concern Species Richness had lower R2 and higher AIC than models containing only one explanatory variable type. Conversely for Total Species Richness, R2 increased and AIC decreased, indicating that explanatory power improved with the addition of the habitat variable measuring the proportion of field boundaries that were vegetated. Although additional explanatory power was gained for Total Species Richness, the increase was modest and the added variable did not offer explanatory power that was different from random.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Ecological variation across wider agricultural landscapes has been well studied (e.g. Marini et al. 2008; Rundlöf et al. 2008). However, this landscape is owned and managed by a complex mosaic of different public and private landowners, which can influence conservation management decisions (Ando & Getzner 2006). Here we demonstrate that variation at the property level in habitat, management and the socio-economic circumstances of the landowners can all feed into determining patterns of biodiversity at the property-scale.

Variation in habitat quantity and quality are the key proximal factors that influence the presence and abundance of avian species. Many of the habitat requirements of farmland birds, both in the uplands and more generally, are well known. Hence, we might anticipate that describing the composition of habitat should be sufficient to understand variation in avian species richness at the property level. Indeed, habitat is important for the suite of species that are UK upland specialists, and many of these patterns are those that would be expected based on previous studies. More upland habitat specialists were present where farms had more rush cover and fewer fields managed by cutting for hay or silage (Baines 1988; Robson & Percival 2002; Pearce-Higgins & Yalden 2003). However, for Total Species Richness and Conservation Concern Species Richness, habitat offered no explanatory power, although the variables measured included features pertaining to the extent and quality of the field boundaries, tree and woodland cover, which have been argued to be beneficial to many farmland birds (Hinsley & Bellamy 2000).

Often birds will make use of habitat over a wide area and are not restricted to a single property or habitat (e.g. Whittingham et al. 2000). We might therefore expect that the quantity of habitat types surrounding the study farms would influence species richness on each property. However, the landscape context of a farm did not account for any variation in Total or Conservation Concern Species Richness, and only had limited influence on Upland Species Richness.

Although farm management variables explained 23% of the variation in Conservation Concern Species Richness, the null model represented a plausible alternative and no individual variable was more likely than random to appear in the best-fitting model. Many of these variables are prescribed options for AES (e.g. mixed farming system to include cattle, cutting dates, number of cuts, stocking density, fertilizer input; Defra 2005a,b). Nonetheless, only one variable enhanced avian species richness. Although the extent of the influence should not be overstated, more species of conservation concern were present where a farming system included cattle. In addition, for the upland specialists, the actual numbers of cows present on a farm was positively related to species richness, although overall livestock density on farms was not found to be important. The counter-intuitive positive relationship between biodiversity and fertilizer input (see Billeter et al. 2008) may simply be a reflection of the relatively low artificial nutrient input levels used in an upland pastoral farming system. Our results are broadly in line with large-scale studies across Europe that have been unable to find consistent agricultural land-use variables that explained species richness across taxa (Billeter et al. 2008).

In the complex farming systems of the Peak District, we may expect interactions between the characteristics of individual farms, some of which may influence biodiversity. One such interaction is between levels of sheep and cattle grazing, which has been shown to affect the breeding success of an upland passerine (Evans et al. 2006). Although not included in the main modelling process, the addition of the interaction term between sheep and cattle numbers led to an increase in AIC for all three measures of species richness, indicating a lack of evidence that this particular interaction influences species richness as a whole.

In the Peak District, we were able to demonstrate a strong relationship between socio-economic circumstances and avian species richness. Species richness declined with increased ownership levels and reliance on farming for the household income. This pattern perhaps indicates that farms that are more important sources of income for their owners are managed in more intensive ways that are less beneficial to biodiversity, where this intensity is not reflected in management variables such as stocking rates and fertilizer use.

The number of farm workers can be taken as a surrogate for the amount of management effort that an individual farm receives. The positive relationship with Total Species Richness could therefore be interpreted as an effect of increased management effort where there is potentially more time available for carrying out activities that are not solely related to agricultural production and which benefit avian diversity. One result of recent policy changes to agricultural subsidy support is a predicted reduction in farm labour (Acs et al. 2008). In the heavily managed landscape of the Peak District, this decline will therefore not only result in social changes, but also impact on levels of biodiversity.

There remains little consensus as to whether or not AES in their current form will deliver substantial conservation gain (Whittingham 2007), although in five European countries, species from across a range of taxa benefited from AES (Kleijn et al. 2006). Here we found that the overall level of AES payment did not determine species richness. In part, this could be because voluntary enrolment may not lead to agreements on the highest-value properties for biodiversity (a problem of adverse selection; Hanley, Shogren & White 2007). It may also be due to the need for neighbouring farms to take joint actions, which is currently not encouraged by policy design (Parkhurst & Shogren 2007).

Although species richness is one of the simplest and most commonly discussed measurements of biodiversity, the question remains as to which species would be most relevant to conservation policy in a given region. Here we tested the richness of three different sets of species (Total, Upland specialists, Conservation Concern). Importantly, the conclusions drawn from the analyses depend very much on the measure of species richness that is used.

The farmland conservation literature has within it an implicit assumption that the socio-economic barriers and financial costs of implementing landscape-scale conservation measures are constant. However, the opportunity costs of land vary spatially (Naidoo & Iwamura 2007), and dimensions of conservation costs, such as management can also vary geographically (Balmford et al. 2000, 2003). The key policy instruments for delivering biodiversity gain in the EU are AES (EEA 2004; CEC 2006), which currently do not take account of property-level heterogeneity in farm socio-economics, factors that can influence the willingness of farmers to enter into an AES as well as the effectiveness of their management actions once in the scheme (Willock et al. 1999; Vanslembouck et al. 2002). Hence, the outcomes of any broad-brush policy at the farm level will not be the same. AES development would therefore benefit from following a similar trajectory to the conservation planning literature by incorporating tools that take into account variations in economic costs (Naidoo et al. 2006) and socio-economic and management barriers to biodiversity conservation across the farmed landscape. Such an approach is likely to lead to substantial conservation gains if future land management practices can be designed and implemented to account for the situations of individual properties.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank the farmers of the Peak District, without whom this study would not have been possible. Help identifying suitable farms was provided the National Farmers Union, the Peak District National Park Authority, the Moors for the Future Partnership, the National Trust and the RSPB. Socio-economic surveys were performed by Richard Darling, John Farrar, Nick Harpur, Helen McCoul, Phil Robertson and Robert Yates. Andrew Skinner and Richard Fuller carried out additional fieldwork. We thank Zoe Davies, Steve Redpath and an anonymous referee for comments that much improved an earlier version of the manuscript. The research was funded as part of the UK Research Councils’ Rural Economy and Land Use Programme (RELU). 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. K.J.G. holds a Royal Society Wolfson Research Merit Award.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Acs, S., Hanley, N., Dallimer, M., Gaston, K.J., Robertson, P., Wilson, R. & Armsworth, P.R. (2008) The effect of decoupling on marginal agricultural systems: implications for farm incomes, land use and upland ecology. Stirling Economics Discussion Paper 2008-18. .
  • Ando, A.W. & Getzner, M. (2006) The roles of ownership, ecology, and economics in public wetland-conservation decisions. Ecological Economics, 58, 287303.
  • Atkinson, P.W., Fuller, R.J., Vickery, J.A., Conway, G.J., Tallowin, J.R.B., Smith, R.E.N., Haysom, K.A., Ings, T.C., Asteraki, E.J. & Brown, V.K. (2005) Influence of agricultural management, sward structure and food resources on grassland field use by birds in lowland England. Journal of Applied Ecology, 42, 932942.
  • Baines, D. (1988) The effects of improvement of upland, marginal grasslands on the distribution and density of breeding wading birds (Charadriiformes) in Northern England. Biological Conservation, 45, 221236.
  • Balmford, A., Gaston, K.J., Rodrigues, A.S.L. & James, A. (2000) Integrating costs of conservation into international priority setting. Conservation Biology, 14, 597605.
  • Balmford, A., Gaston, K.J., Blyth, S., James, A. & Kapos, V. (2003) Global variation in terrestrial conservation costs, conservation benefits, and unmet conservation needs. Proceedings of the National Academy of Sciences of the United States of America, 100, 10461050.
  • 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. Journal of Applied Ecology, 39, 673687.
  • Billeter, R., Liira, J., Bailey, D., Bugter, R., Arens, P., Augenstein, I., Aviron, S., Baudry, J., Bukacek, R., Burel, F., Cerny, M., De Blust, G., De Cock, R., Diekoetter, T., Dietz, H., Dirksen, J., Dormann, C., Durka, W., Frenzel, M., Hamersky, R., Hendrickx, F., Herzog, F., Klotz, S., Koolstra, B., Lausch, A., Le Coeur, D., Maelfait, J.P., Opdam, P., Roubalova, M., Schermann, A., Schermann, N., Schmidt, T., Schweiger, O., Smulders, M.J.M., Speelmans, M., Simova, P., Verboom, J., Van Wingerden, W.K.R.E., Zobel, M. & Edwards, P.J. (2008) Indicators for biodiversity in agricultural landscapes: a pan-European study. Journal of Applied Ecology, 45, 141150.
  • Biodiversity Reporting and Information Group. (2007) Report on the Species and Habitat Review. Report to the UK Biodiversity Partnership, JNCC, Peterborough, UK.
  • Burnham, K.P. & Anderson, D.R. (2002) Model Selection and Multimodal Inference. Springer, New York.
  • CEC. (2006) Communication from the Commission. Halting the Loss of Biodiversity by 2010 and Beyond: Sustaining Ecosystem Services for Human Well-Being. Commission of the European Communities, Brussels, Belgium.
  • Chamberlain, D.E., Fuller, R.J., Bunce, R.G.H., Duckworth, J.C. & Shrubb, M. (2000) Changes in the abundance of farmland birds in relation to the timing of agricultural intensification in England and Wales. Journal of Applied Ecology, 37, 771788.
  • Defra (2004). An Assessment of the Impacts of Hill Farming in England on the Economic, Environmental and Social Sustainability of the Uplands and More Widely. Defra, London.
  • Defra (2005a). Entry Level Stewardship Handbook. Department of Environment, Food and Rural Affairs, London.
  • Defra (2005b). Higher Level Stewardship Handbook. Department of Environment, Food and Rural Affairs, London.
  • Defra (2005c). Farm Incomes in the United Kingdom. Department of Environment, Food and Rural Affairs, London.
  • Donald, P.F., Sanderson, F.J., Burfield, I.J. & Van Bommel, F.P.J. (2006) Further evidence of continent-wide impacts of agricultural intensification on European farmland birds, 1990–2000. Agriculture, Ecosystems and Environment, 116, 189196.
  • EEA. (2004) High Nature Value Farmland – Characteristics, Trends and Policy Challenges. European Environment Agency, Copenhagen, Denmark.
  • Evans, D.M., Redpath, S.M., Evans, S.A., Elston, D.A., Gardner, C.J., Dennis, P. & Pakeman, R.J. (2006) Low intensity, mixed livestock grazing improves the breeding abundance of a common insectivorous passerine. Biology Letters, 2, 636638.
  • FAO. (2003) FAOSTAT Land Use Data Archives. Accessed April 2008.
  • Gibbons, D.W., Reid, J.B. & Chapman, R.A. (1993) The New Atlas of Breeding Birds in Britain and Ireland: 1988–1991. BTO/SWC/IWC. T. & A.D. Poyser, London.
  • Gregory, R.D., Wilkinson, N.I., Noble, D.G., Robinson, J.A., Brown, A.F., Hughes, J., Procter, D.A., Gibbons, D.W. & Galbraith, C.A. (2002) The population status of birds in the United Kingdom, Channel Islands and Isle of Man: an analysis of conservation concern 2002–2007. British Birds, 95, 410450.
  • Haines-Young, R., Barr, C.J., Firbank, L.G., Furse, M., Howard, D.C., McGowan, G., Petit, S., Smart, S.M. & Watkins, J.W. (2003). Changing landscapes, habitats and vegetation diversity across Great Britain. Journal of Environmental Management, 67, 267281.
  • Haines-Young, R., Barr, C.J., Black, H.I.J., Briggs, D.J., Bunce, R.G.H., Clarke, R.T., Cooper, A., Dawson, F.H., Firbank, L.G., Fuller, R.M., Furse, M.T., Gillespie, M.K., Hill, R., Hornung, M., Howard, D.C., McCann, T., Morecroft, M.D., Petit, S., Sier, A.R.J., Smart, S.M., Smith, G.M., Stott, A.P., Stuart, R.C. & Watkins, J.W. (2000) Accounting for Nature: Assessing Habitats in the UK Countryside, Department of Environment, Transport and the Regions, London.
  • Hanley, N., Shogren, J. & White, B. (2007) Environmental Economics in Theory and Practice, 2nd edn. Palgrave Macmillan, London.
  • Hinsley, S.A. & Bellamy, P.E. (2000) The influence of hedge structure, management and landscape context on the value of hedgerows to birds: a review. Journal of Environmental Management, 60, 3349.
  • Holt, A.R., Davies, Z.G., Tyler, C. & Staddon, S. (2008). Meta-analysis of the effects of predation on animal prey abundance: evidence from UK vertebrates. PLoS One, 3, 32400.
  • Huby, M., Cinderby, S. Crowe, A.M., Gillings, S., McClean, C.J., Moran, D., Owen, A. & White, P.C.L. (2006) The association of natural, social and economic factors with bird species richness in rural England. Journal of Agricultural Economics, 57, 295312.
  • Hudson, P.J. (1992) Grouse in Space and Time. Game Conservancy Trust, Fordingbridge, UK.
  • Johnson, J.B. & Ohmland, K.S. (2004) Model selection in ecology and evolution. Trends in Ecology & Evolution, 19, 101108.
  • Kareiva, P., Watts, S., McDonald, R. & Boucher, T. (2007) Domesticated nature: shaping landscapes and ecosystems for human welfare. Science, 316, 18661869.
  • Kleijn, D. & Sutherland, W.J. (2003) How effective are European agri-environment schemes in conserving and promoting biodiversity? Journal of Applied Ecology, 40, 947969.
  • Kleijn, D., Baquero, R.A., Clough, Y., Diaz, M., De Esteban, J., Fernandez, F., Gabriel, D., Herzog, F., Holzschuh, A., Johl, R., Knop, E., Kruess, A., Marshall, E.J.P., Steffan-Dewenter, I., Tscharntke, T., Verhulst, J., West, T.M. & Yela, J.L. (2006). Mixed biodiversity benefits of agri-environment schemes in five European countries. Ecology Letters, 9, 243254.
  • Lee, J.A. (1998) Unintentional experiments with terrestrial ecosystems: ecological effects of sulphur and nitrogen pollutants. Journal of Ecology, 86, 112.
  • Marini, L., Fontana, P., Scotton, M. & Klimek, S. (2008) Vascular plant and Orthoptera diversity in relation to grassland management and landscape composition in the European Alps. Journal of Applied Ecology, 45, 361370.
  • Millennium Ecosystem Assessment. (2005) Ecosystems and Human Well-being: Biodiversity Synthesis. World Resources Institute, Washington, D.C.
  • Myers, R. (1990) Classical and Modern Regression with Applications, 2nd edn. Duxbury, Boston MA, USA.
  • Naidoo, R. & Iwamura, T. (2007) Global-scale mapping of economic benefits from agricultural lands: implications for conservation priorities. Biological Conservation, 140, 4049.
  • Naidoo, R., Balmford, A., Ferraro, P.J., Polasky, S., Ricketts, T.H. & Rouget, M. (2006). Integrating economic costs into conservation planning. Trends in Ecology & Evolution, 21, 681687.
  • NEGTAP. (2001) Transboundary Air Pollution: Acidification, Eutrophication and Ground-Level Ozone in the UK. Department of Environment, Food and Rural Affairs, London.
  • Parkhurst, G. & Shogren, J. (2007) Spatial incentives to coordinate contiguous habitat. Ecological Economics, 64, 344355.
  • Peach, W.J., Lovett, L.J., Wotton, S.R. & Jeffs, C. (2001) Countryside stewardship delivers cirl buntings (Emberiza cirlus) in Devon, UK. Biological Conservation, 101, 361373.
  • Peak District Rural Deprivation Forum. 2004. Hard Times: A Research Report into Hill Farming and Farm Families in the Peak District. Peak District Rural Deprivation Forum, Hope, Derbyshire, UK.
  • Pearce-Higgins, J.W. & Yalden, D.W. (2003) Variation in the use of pasture by breeding European golden plovers Pluvialis apricaria in relation to prey availability. Ibis, 145, 365381.
  • Pocock, M.J.O. & Jennings, N. (2008) Testing biotic indicator taxa: the sensitivity of insectivorous mammals and their prey to the intensification of lowland agriculture. Journal of Applied Ecology, 45, 151160.
  • Preston, C.D., Telfer, M.G., Arnold, H.R., Carey, P.D., Cooper, J.M., Dines, T.D., Hill, M.O., Pearman, D.A., Roy, D.B. & Smart, S.M. (2002) The Changing Flora of the UK. Department of Environment, Food and Rural Affairs, London, UK.
  • R Development Core Team. (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL
  • Robinson, R. & Sutherland, W.J. (2002) Post-war changes in arable farming and biodiversity in Great Britain. Journal of Applied Ecology, 39, 157176.
  • Robson, G., & Percival, S.M. (2002) The use of marginal farmland by curlew Numenius arquata breeding on upland moors. Aspects of Applied Biology, 67, 7584.
  • Rundlöf, M., Bengtsson, J. & Smith, H.G. (2008) Local and landscape effects of organic farming on butterfly species richness and abundance. Journal of Applied Ecology, 45, 813820.
  • Sim, I.M.W., Gregory, R.D., Hancock, M.H. & Brown, A.F. (2005) Recent changes in the abundance of British upland breeding birds. Bird Study, 52, 261275.
  • Smith, R.K., Jennings, N.V., Robinson, A. & Harris, S. (2004) Conservation of European hares Lepus europeaus in Britain: is increasing habitat heterogeneity in farmland the answer? Journal of Applied Ecology, 41, 10921102.
  • Stillman, R.A. & Brown, A.F. (1994) Population sizes and habitat associations of upland breeding birds in the south Pennines, England. Biological Conservation, 69, 307314.
  • Stroud, D.A., Chambers, D., Cook, S., Buxton, N., Fraser, B., Clement, P., Lewis, P., McLean, I., Baker, H. & Whitehead, S. (2001) The UK SPA Network: Its Scope and Content. JNCC, Peterborough, UK.
  • Tharme, A.P., Green, R.E., Baines, D., Bainbridge, I.P. & O’Brien, M. (2001) The effects of management for red grouse shooting on the population density of breeding birds on heather-dominated moorland. Journal of Applied Ecology, 38, 439457.
  • Vanslembouck, I., Van Huylenbroeck, G. & Verbeke, W. (2002) Determinants of willingness of Belgian farmers to participate in agri-environmental measures. Journal of Agricultural Economics, 53, 489511.
  • Whittingham, M.J. (2007) Will agri-environment schemes deliver substantial biodiversity gain, and if not why not? Journal of Applied Ecology, 44, 15.
  • Whittingham, M.J., Percival, S.M. & Brown, A.F. (2000) Time budgets and foraging of breeding golden plover Pluvialis apricaria. Journal of Applied Ecology, 37, 632646.
  • Whittingham, M.J., Swetnam, R.D., Wilson, J.D., Chamberlain, D.E. & Freckleton, R.P. (2005) Habitat selection by yellowhammers Emberiza citrinella on lowland farmland at two spatial scales: implications for conservation management. Journal of Applied Ecology, 42, 270280.
  • Whittingham, M.J., Krebs, J.R., Swetnam, R.D., Vickery, J.A., Wilson, J.D. & Freckleton, R.P. (2007) Should conservation strategies consider spatial generality? Farmland birds show regional not national patterns of habitat association. Ecology Letters, 10, 2535.
  • Wickramasinghe, L.P., Harris, S., Jones, G. & Jennings, N.V. (2003) Bat activity and species richness on organic and conventional farms: impact of agricultural intensification. Journal of Applied Ecology, 40, 984993.
  • Willock, J, Deary, I.J., Edwards-Jones, G., Gibson, G.J., McGregor, M.J., Sutherland, A., Dent, J.B., Morgan, O., Grieve, R. (1999) The role of attitudes and objectives in farmer decision making: business and environmentally-oriented behaviour in Scotland. Journal of Agricultural Economics, 50, 286303.
  • Yallop, A.R., Thacker, J.I., Thomas, G., Stephens, M., Clutterbuck, B., Brewer, T. & Sannier, C.A.D. (2006) The extent and intensity of management burning in the English uplands. Journal of Applied Ecology, 43, 11381148.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Appendix S1. Species list and conservation status for all birds recorded on 44 upland farms in the Peak District, Northern England

Appendix S2. Parameter estimates for generalized linear model with Poisson errors and a log link function relating the three measures of species richness to transformed explanatory variables

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

JPE_1616_sm_AppendixS1-S2.doc126KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.