Landscape Ecol (2007) 22:1255–1264 DOI 10.1007/s10980-007-9105-7

RESEARCH ARTICLE

Mapping hotspots of multiple landscape functions: a case study on farmland afforestation in Scotland Alessandro Gimona Æ Dan van der Horst

Received: 13 June 2006 / Accepted: 5 April 2007 / Published online: 10 May 2007 Ó Springer Science+Business Media B.V. 2007

Abstract Many conservation and restoration efforts in developed countries are increasingly based on the premise of recognising and stimulating more ‘multifunctionality’ in agricultural landscapes. Public policy making is often a pragmatic process that involves efforts to negotiate trade-offs between the potentially conflicting demands of various stakeholders. Conservationists’ efforts to influence policy making, can therefore benefit from any tool that will help them to identify other socio-economic functions or values that coincide with good ecological conservation options. Various types of socio-economic objectives have in recent years been mapped across landscapes and so there are now important opportunities to explore the spatial heterogeneity of these diverse functions across the wider landscape in search of potential spatial synergies, i.e. ‘multiple win locations’ or multifunctional ‘hotspots’. This paper explores the potential occurrence of such synergies within the agricultural landscape of northeast

A. Gimona (&) Macaulay Institute, Craigiebuckler, Aberdeen AB15 8QH, UK e-mail: [email protected] D. van der Horst School of Geography, Earth and Environmental Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK e-mail: [email protected]

Scotland and evaluates an existing woodland planting policy using and combining three different policy objectives. Our results show that there are indeed broad areas of the studied landscape where multiple objectives (biodiversity, visual amenity and on-site recreation potential) could be achieved simultaneously (hotspots), and that the case study which we evaluate (the Farm Woodland Premium Scheme) could be much better spatially targeted with regards to each individual objective as well as with regards to these hotspots of multifunctionality. Keywords Farm woodlands
GIS
Spatial targeting
Biodiversity
Recreation
Visual amenity
Landscape
Multifunctionality

Introduction The biodiversity associated with ‘traditional’ agricultural landscapes has suffered greatly since the introduction of industrial practices in modern agriculture, which transformed many agricultural landscapes into large fields of a single, and genetically identical, crop species (e.g. Benton et al. 2003). Many governments are formulating policies to reverse this trend. Because high level policy objectives are relatively abstract (e.g. conserving biodiversity), proxy measures must be developed to track the success of the policy interventions (e.g. size of afforested area;

123

1256

population trends for endangered species etc). The use of non-monetary proxy measures is particularly important where economic valuation is problematic or controversial, and these measures are relevant for a range of different stakeholders who take an interest in the impacts of policy, as these measures allow them to evaluate the relative efficiency of policy interventions in terms of ‘value for money’. Because of the existence of diverse stakeholders, numerous conservation and restoration efforts in developed countries are now based on the premise of recognising and stimulating more ‘multi-functionality’ in these monocultural landscapes. The concept appears to be intuitively attractive as the talk of moving towards more multifunctional, or multipurpose or even multi-benefit, approaches to landscape management is now widespread amongst policy makers, various pressure groups and academics from a number of disciplines. Academic interest in ‘multifunctionality’ is expressed through dedicated conferences (e.g. Frede and Bach 2005), special issues in peer-reviewed journals (e.g. Land Use Policy 17; Journal of Environmental Planning and Management, 48(1); Landscape and Urban Planning 75) and many research projects. The notion of multi-functionality is for example at the heart of efforts to link ecological concepts of ecosystem-functions with the economic concept of total economic value (e.g. Gren et al. 1995; Costanza et al. 1998; Guo et al. 2001). Despite some academic debate about methodological and moral issues, such efforts are beneficial in a practical sense because the resulting monetary estimates of the value of ecosystems are huge and can thus help to draw more attention to the importance of conservation efforts. In pluralist democratic societies, public policy making is often a ‘communicative incremental’ process that involves negotiations between different stakeholders and the various representative structures of public authority, resulting in a compromise between various actors’ interests (Buttoud 2000). Under these circumstances conservationists’ chances to influence policy-making are greatly enhanced if they can find synergies between their interests and that of other stakeholders. The search for opportunities for collaboration could create a demand for analytical approaches that can identify the extent to which other social or economic ‘values’ or ‘benefits coincide (incidentally or systematically) with good ecological conservation options. In this

123

Landscape Ecol (2007) 22:1255–1264

paper we use the word ‘value’ to indicate a quantifiable measure of benefit, as opposed to the more restrictive economic meaning of ‘monetary value’ or ‘dollar value’ (for a discussion about the relationship between different types of economic values and ecosystem functions, see Turner et al. 2000; de Groot et al. 2002; Hein et al. 2006). This paper will focus on the development and use of a method to identify and map areas where such coincidence of multiple interests is realised, and in particular on the mapping of multiple functions across a landscape in order to identify suitable target areas for the creation small multifunctional woodlands. In recent years there has been a strong growth in the number of publications on the use of analytical (GIS) methods for the spatial targeting of conservation efforts (e.g. Church et al. 1996; Lathrop and Bognar 1998; Ferrier 2002; Wilson et al. 2006). Examples of the use of such methods to identify target areas for habitat restoration or habitat creation include work on woodlands (Lee et al. 2002; Gkaraveli et al. 2004; van der Horst and Gimona 2005), chalk grassland (Thompson et al. 1999; Lee et al. 2001), the revegetation of areas affected by dryland salinity problems (Apan et al. 2004) and more general conservation efforts at the watershed scale (Kazmiersky et al. 2004). Also there is a growing body of work on reserve selection algorithms applied to landscape zonation (e.g. Moilanen et al. 2005; Moilanen 2005). However, despite the tendency of government land use policy interventions to be justified on the basis of a range of different objectives, there has been little academic work to date which has actually sought to combine the spatial targeting of habitat restoration with other, more socio-economic rural land use objectives such as the promotion of outdoor recreation or the enhancement of the visual landscape (for recent efforts to combine socio-economic and conservation aims within a more spatially explicit concept of natural capital, see Bailey et al. 2006; Haines-Young et al. 2006; van der Horst 2006a). Since these socio-economic objectives are also being mapped across the landscape (e.g. Brainard et al. 2001; van der Horst 2006b), there are now important opportunities to explore the spatial heterogeneity of these diverse functions across the wider landscape in search of potential spatial synergies, i.e. ‘multiple win locations’ or multifunctional

Landscape Ecol (2007) 22:1255–1264

‘hotspots’. Such hotspots are relevant because they represent opportunities to pursue conservation in priority areas through coalitions between conservationists and other stakeholders. A high diversity of different values or functions in an area is also likely to be associated with a relatively high total value (and total economic value) of that area. The aims of this paper are (i) to produce a landscape zonation based on multiple criteria, (ii) to highlight the possible occurrence of any multifunctional hotspots within the agricultural landscape and (iii) to use this zonation process to evaluate the efficiency of an existing multi-functional land use policy. The following section details the method we use to map the level of landscape multifunctionality and the landscape zonations resulting from this analysis. Case study : evaluation of an existing multifunctional afforestation scheme section applies these zonations to the evaluation of the case study policy. The final section of the paper discusses the potential uses and limitations of our approach.

Mapping multifunctional hotspots Methods The benefit maps for biodiversity, visual amenity and woodland recreation

1257

lated for each grid cell of the map as the sum of weighted species preferences: X APBS ¼ S i Xi ð1Þ N

whereby N is the number of species considered; Si is the preference of species i for that cell to become afforested (whereby S can have three possible values; 1 = afforestation preferred, 1 = non-afforestation preferred, 0 = makes no difference). These preferences are elicited from a literature review of habitat requirements of each species; Xi is the weight (or importance) of the ith species. The weighting is the product of three measures, namely (a) the importance of the regional population relative to the national population; (b) the scarcity of suitable habitat relative to the distribution of the population and (c) the reliability of mapping suitable habitat. Although our index is not based on selection algorithms evaluating, for instance, the value of each single cell in the context of the whole landscape, the index is based on spatially explicit distribution models based on habitat preferences, which took into account the relative positions of landscape features within distances appropriate to the species being modelled. Potential visual amenity map

The paper makes use of previously published methods to create woodland value maps for biodiversity, visual amenity and recreation (see below and Appendix 1). These value maps are used as input in the analysis, focusing on the case study region of north– east Scotland (see Fig. 1). These raster maps attribute a ‘benefit score’ to each cell in the landscape. A brief description of the methods behind these value maps is provided below.

In van der Horst (2006b), the aggregate visual amenity score (AVAS) of each grid cell is calculated on the basis of four key variables, namely the spatial distribution of the viewing population (using travel data to present both local inhabitants and visitors), the general preference of the public for the amount of woodland they like to see in the landscape, the amount of woodland already visible in the local landscape and the actual visibility of grid cells in the landscape from (public) points of observation.

Potential biodiversity map

Potential on-site recreation map

In van der Horst and Gimona (2005), evaluation of the potential biodiversity benefits of farm woodlands is based on a species-centred approach which used spatially-explicit habitat suitability models for 16 species deemed to be priority species in a Biodiversity Action Plan. Benefit scores (APBS) are calcu-

This map was created on the basis of the analysis by Hill et al. (2003), Hill and Courtney (2006) which, in turn, adopted an existing methodology (Brainard et al. 1999, 2001). This methodology estimates the expected number of visitors at any grid cell as an inverse function of distance to where people live,

123

1258

Landscape Ecol (2007) 22:1255–1264

Fig. 1 (a) Biodiversity benefits; (b) visual amenity benefits; (c) Recreation benefits

weighted by the number of people living there (i.e. a distance decay function of weighted population centres). This method calculates distance on the basis of the existing road network and also takes account of substitute sites of woodland recreation, i.e. people living in an area with little woodland cover are assumed to be willing to travel relatively longer distances for the purpose of woodland recreation. Scores are attributed to each 500 m cell using a regression model developed and validated by Hill et al. (2003). To account for non-local visitors, areas up to 1.5 h of travel distance from each cell are included in the GIS analysis (see Appendix 1 for further details).

Each combination was then obtained applying the general formula

Combination of the benefit maps

Landscape zonation and multifunctional hotspots

To investigate multi-functionality we combined the benefit maps. Prior to combination, each benefit map was standardised to vary from 0 to 100 using the formula:

We highlighted the spatial distribution of different benefits areas, for each map, which could potentially be used as a targeting aid. We adopted the first, second and third quartile of the distribution of scores as cut-off points, and attributed different tones of grey to cells belonging to the each of the resulting four intervals. The resulting maps highlight where to

Z ¼ ((x  oldmin)  (newmax  newmin)= (oldmax  oldmin))+ newmin ð2Þ

123

Cj = w1  APBS + w2  AVAS + w3  VISITS

ð3Þ

where Cj is the particular combination and w1...w3 are the weights. Different decision makers and stakeholders might attribute the same or different priorities to the three criteria we have chosen. To simulate this, we chose four different sets of weights. In one case, all criteria were equally weighted, in the other three, each criterion was attributed double the weight as the other two. For each combination of maps weights were chosen to sum to 1. These are presented in Table 1.

Landscape Ecol (2007) 22:1255–1264

1259

Table 1 selected weightings for combining the three benefit maps Criterion

Weighting scheme a

b

c

d

Biodiversity

0.33

0.25

0.25

0.50

Visual amenity

0.33

0.25

0.50

0.25

Recreation

0.33

0.50

0.25

0.25

target planting according to each combination of criteria. We also identified areas which are consistently high/low scoring irrespective of the weighted map one might prefer. We called consistently high scoring areas ‘‘multifunctional hotspots’’. To find the hotspots, we simply queried the three weighted maps to highlight cells which have scores above/below the median in all cases.

from the point of view of all the criteria used in this study. Most of these are located in the western part, and are along river valleys and roads which follow them. These areas should be targeted for farm woodland plantings because they score high regardless of relative priorities, as defined by the weighted schemes. Conversely, there are also ‘lose-lose-lose’ areas (or ‘coldspots’), located mostly in the North-Eastern part of the study area, where woodland planting should be avoided. These are areas where planting would result in isolated woodlands and cause loss of habitat for open ground species. Also, they are relatively less well connected and accessible by visitors and have limited visual amenity value because the gently undulating landscape with a low road network density offers few opportunities to create visually attractive woodlands for a large number of people.

Results Landscape zonations and multifunctional hotspots The maps corresponding to different zonations (see also Table 1) are shown in Fig. 2. The map in Fig. 3 shows that there are some ‘‘win–win–win’’ (hotspots) areas in the landscape which offer benefits

Case study : evaluation of an existing multifunctional afforestation scheme We now use the agricultural landscape of the northeast of Scotland as a study area to illustrate one potential use of the zonations obtained above,

Fig. 2 Weighted benefits maps, with FWPS locations superimposed. Options a–d are weighted according to schemes in Table 1

123

1260

Fig. 3 Multifunctional hotspots (always above the median) and ‘coldspots’ (always below the median) in the study landscape

namely the evaluation of a multi-functional afforestation scheme.

Landscape Ecol (2007) 22:1255–1264

A relatively conventional mid-term evaluation of the FWPS (MLURI 1996) showed appreciation for the importance of spatial characteristics at the sitelevel only, which was assessed through site visits by experts who noted how the planted woodland fitted into the local landscape and how much wildlife could be observed. An evaluation of the plantings at the landscape or regional scale (as opposed to site-level) would require knowledge of the spatial variability of the benefits that the scheme is supposed to provide. As we highlighted earlier, such potential woodland ‘value maps’ or ‘benefit maps’ have already been created for biodiversity (van der Horst and Gimona 2005), visual amenity (van der Horst 2006b), and recreation (Brainard et al. 1999; Hill et al. 2003). The next section will address how these maps can be used to evaluate existing plantings with regards to each of the three separate functions.

The afforestation scheme The Farm Woodland Premium Scheme (FWPS) is a voluntary scheme, funded as an agri-environmental measure under the EU’s Common Agricultural Policy, offering farmers in Scotland annual incentive payments for the conversion of farmland to woodlands1. Biodiversity and visual amenity are the two public benefits that the woodlands planted under the FWPS were officially expected to deliver, but woodland recreation could also be important because people have the right to walk on private land in Scotland (subject to minor restrictions). The payment levels for this scheme are based on land use classifications which reflect the opportunity cost of the land, i.e. the forgone income from agricultural activities. The scheme is thus designed to off-set the spatial heterogeneity of the agricultural quality of the land to ensure that planting trees is equally financially attractive to each farmer and for each field. The scheme does not take into account the possibility that the level of potential benefits may also be spatially heterogeneous. Consequently, the scheme is offering planting subsidies that are much higher in one location than in another, without any consideration of the level of benefits in these locations.

1

This scheme has recently been replaced by the Scottish Forestry Grants Scheme: Farmland Premium (SFGS/FP), which is similar with regards to the topic of this paper.

123

Evaluation method In this section we ask whether the spatial patterns of woodlands planted under this scheme reflects the landscape heterogeneity with regards to the potential benefits outlined above, or if the woodland locations can be considered just a random sample over the study area. All woodlands planted under the scheme have been mapped by the Forestry Commission in vector GIS format. To measure to what extent the existing woodlands provide benefits according to the three (benefit) criteria mentioned above, we over-laid the map of woodland polygons (FWPS map) on each of the three benefit maps and calculated statistics for the resulting three distributions of scores. These provided a measure of how well the planted woodlands are located with respect to the potential benefits that they are supposed to deliver. We adopted several measures of success, based on the distribution of FWPS woodlands scores: 1. 2.

3.

On each of the three individual benefit maps (Fig. 1) On a multiple-benefit map which combines the three benefits at equal weights (‘a’ in Table 1 and Fig. 2) On multi-benefit maps which combine the three benefits but attributes double weight to each mapped criterion (i.e. benefit map) in turn, with

Landscape Ecol (2007) 22:1255–1264

respect to the other two mapped criteria (‘b’, ‘c’ and ‘d’ in Table 1 and Fig. 2). For each benefit map, we tested if the locations of the woodlands had scores which could be considered a random sample of the map (which is what could be expected if the uptake of this scheme has indeed been homogeneous across the landscape, rather than spatially targeted), or if scores significantly departed from randomness. In other words, given the mean calculated over all the cells of a benefit map, we asked how likely it was that the mean of the woodland scores was obtained by placing the woodlands randomly in the landscape. We compared differences between the global mean and random samples with the difference between the global mean and the mean woodland scores. Before performing this analysis, we tested whether using whole woodland polygons or using just their centroids gave different results, and found that there was no statistically significant difference in tests based on the three single-criterion benefit maps. Woodland centroids were therefore used. To avoid any assumptions about the distributions, means and variances of the scores, for each benefit map, we performed a Monte-Carlo simulation, which included the following steps: 1.

Calculate the difference between the mean of the scores of the N woodlands and the global mean, 2. Compare this difference to the differences between the scores of the same number of points randomly ‘‘thrown’’ on the map and the global mean, 3. Repeat this procedure 1,000 times and obtain a distribution of differences, 4. Check if the difference between the mean of the original woodland scores and the mean of benefit map scores exceeds the 97.5 percentile or is lower than the 2.5 percentile of such a distribution (i.e. applying a 2-tail test). The statistics calculated in the Monte Carlo simulation, and the 95% interval within which a randomly placed set of points would have scored, are presented in Tables 2 and 3, respectively. In Table 2 we report statistics for each individual benefit criterion, while in Table 3 we report statistics for each of the weighting options presented in Table 1.

1261 Table 2 Interval within which the 95% of the differences between 1,000 random samples and the global mean lie, compared to the difference between mean woodland scores and global mean Criterion

2.5 percentile

97.5 percentile

Difference

Biodiversity

0.667

0.728

0.180

Visual amenity

0.066

0.075

0.063

Recreation

0.588

0.559

2.111*

* Means the difference is significant with P = 0.05

Table 3 Interval within which the 95% of the differences between 1,000 random samples and the global mean lie, compared to the difference between mean woodland scores and global mean Weighting scheme 2.5 percentile 97.5 percentile Difference a

0.384

0.350

0.907*

b

0.424

0.381

1.337*

c

0.278

0.282

0.571*

d

0.435

0.416

0.589*

* Means the difference is significant with P = 0.05

Table 2 shows that the FWPS plantings perform as well as a random sample for biodiversity and visual amenity and it performs worse than a random sample according to the recreation map. This suggests that the woodlands are rather ineffective in providing any or all of these three benefits and that potentially great efficiency gains could be had if the FWPS had been designed to deliberately target areas of high benefit. Using as criteria each of the four weighted multibenefit maps, the results of the analysis show that the distribution of FWPS plantings is worse than that expected from a random sample. This is because of a combination of two factors. Firstly, some landscape areas, located towards the west and southwest portion of the study area, have (in general) much higher scores than average, according to the three criteria. Secondly, the effect of unequal weights on the combined benefit map is such that, while areas in the eastern and northern portion have values hovering near the average in the equally-weighted case, they receive values clearly below the average in the unequally weighted cases. As a consequence, there is an over-occurrence of FWPS woodlands in low-scoring areas which results in an apparent ‘negative targeting’. This can also be appreciated by visually

123

1262

inspecting the maps in Fig. 3, which have been classified as explained in Landscape zonations and multifunctional hotspots section. These results show that, using individual criteria, the existing FWPS plantings perform, at best, no better than randomly planted woodlands. When using weighted combinations, the existing FWPS plantings are actually located preferentially in low-benefit areas. This conclusion is robust with respect to the weighting scheme adopted.

Discussion Current policies are increasingly framed in terms of ‘multi-functionality’, which, as far as afforestation policy is concerned, suggests that every planted farm woodland should satisfy a combination of objectives. In van der Horst and Gimona (2005) we used a spatially explicit approach to species-habitat models at the landscape scale. In relation to biodiversity benefits, we discussed the fact that a site-based evaluation designed to select where to locate woodlands would have great limitations without a landscape-scale approach. In this paper, we have adopted such an approach with regard to the potential contribution of woodland plantings to three important policy objectives, namely biodiversity, visual amenity and on-site recreation. Landscape planning must take into account a number of concerns whose nature can vary according to the point of view of different stakeholders. We have considered only three possible criteria and have proposed four different ways to combine benefit maps, but this is by no means exhaustive. Nor are our three objectives necessarily exhaustive, and others could be deemed important. Some examples are the harmonisation of local scale and landscape scale objectives (Kuttila and Pukkala 2003), carbon sequestration, or reduction of nitrate leaching (Gilliams et al. 2005). These criteria could be included in future analysis. Therefore we are aware that our proposed zonation cannot have universal value. While no zonation can be considered to be ‘objective’, the criteria we adopted are of recognised importance in landscape planning and therefore we believe our methods and results can contribute constructively to the debate and offer a repeatable procedure to obtain a policy-relevant landscape zonation.

123

Landscape Ecol (2007) 22:1255–1264

Other limitations of this approach include the fact that the quartiles chosen as cut-off points to construct a zonation are reasonable but arbitrary, and the fact that a weighted summation to combine criteria assumes independence in criteria scoring (see also discussion in Bailey et al. 2006). This assumption is met in this case, but would be important to convey to stakeholders attributing scores. Also, the benefit maps clearly depend on the criteria adopted to construct them. For instance, our biodiversity map is based on a list of species based on a Biodiversity Action Plan, and therefore defensible on policy grounds, but not comprehensive. Nevertheless the analysis of our case study demonstrates that a multi-functional zonation can be a very useful tool in policy evaluation, as it can provide insights into the potential efficiency of the impacts of land use policies at the landscape level. It is important to realise that this method can add value to on-site ecological surveys: although site-level investigations had suggested a positive impact of the FWPS woodlands on biodiversity and visual amenity (MLURI 1996), our landscape-level analysis shows that a better spatial targeting of this woodland scheme could potentially provide much more benefit. Despite the different results obtained with four weighting schemes, the existence of hot and cold spots shows that, in this case, there is a broad scale structure in the landscape, with benefit criteria correlated over some fairly large areas. The existence of such multi-functional hotspots is of considerable interest because these clearly represent priority areas which would enable policy makers to obtain the maximum benefit from the same amount of subsidy to farmers: Roughly 20% of the study area would provide benefits according to all the criteria used in this study. The findings in this paper also open up the possibility to evaluate existing formal spatial targeting strategies, such as the Indicative Forestry Strategies (IFS) in Scotland (SDD 1990, 1999). The IFS, which is presented with maps detailing ‘sensitive’, ‘potential’ and ‘preferred’ areas of woodland plantings have been heavily criticised in the past (StuartMurray 1994; Stuart-Murray et al. 1999) and an analytical and spatially explicit approach to landscape multi-functionality could offer real scope for improvement. Consistent with notions of ‘sustainability science’ (Potschin and Haines-Young 2006),

Landscape Ecol (2007) 22:1255–1264

such spatial planning tools ought to be developed in consultation with a range of stakeholders (including members of the public), who can be engaged in exploration and negotiation about the type of benefits to include, the methods and spatial scale for mapping these benefits, and their relative importance (i.e. weighting). The broad zonation we have produced here can provide the spatial context for subsequent planning and analysis at a more detailed level, accounting for example for species life-histories and spatial cohesion (Vos et al. 2001; Verboom et al. 2001; Opdam et al. 2003; Moilanen et al. 2005). We propose that the type of analysis presented in this paper could be used to inform the choice between conservation-oriented approaches and other approaches which ‘‘do not pre-focus on more natural structures in the landscape’’ (Opdam et al. 2006). This paper demonstrates that these seemingly contrasting approaches can sometimes, and in some places, be reconciled. Acknowledgements We thank Dr. Gary Hill for providing reports and advice regarding the forest role in tourism, John de Groot for preliminary GIS work used in the production of the recreation benefits map and the UK Forestry Commission for providing the FWPS polygons. We also thank Prof. Paul Opdam for constructive criticism on previous versions of this manuscript. We acknowledge the Scottish Executive Environment and Rural Affairs Department (SEERAD) for financial support.

Appendix 1 Modelling day visits to forests In this work we used the number of potential visitors as an index of provision of recreation to create the ‘‘recreation’’ criterion map. We treated each 500 m cell in our map as an area of interest for which the number of potential visitors were predicted using the methods outlined below. We derived the recreation map following Hill et al. (2003), who based their approach on Brainard et al. (2001). These authors used site-specific characteristics, human population availability, and the location of existing woodlands to predict the number of visitors to forest sites in England. Hill et al. managed to explain 50% of the variation in number of visitors using:

1263

1. 2.

3.

The population within 90 minutes of travel time to a given location (Pop90) An index of substitute availability i.e. the number for woodlands larger than 10 ha within 2 h of travel time for a given location (SLRG120). An index summarising availability of facilities for visitors, such as walking paths, toilets etc. The Hill et al. (2003) equation used in this work is:

Log(No. of Visitors) = 0.247  log(Pop90) +1.539  log(SLRG120) +0.735  log(Facilities) The travel times are those giving the best-fit model among a number of different values they investigated. We assumed that the locations would have no facilities. For the other two variables we used population census data (Census Output Areas for 2001) and existing road maps (with different average speed of travel attributed to different road types) to create maps of population availability and woodland maps to create maps of substitute woodland availability, using each 500 m cell as the location of interest. References Apan AA, Raine SR, Le Brocque A, Cockfield G (2004) Spatial prioritization of revegetation sites for dryland salinity management: an analytical framework using GIS. J Environ Plan Manage 47(6):811–825 Bailey N, Lee JT, Thompson S (2006) Maximising the natural capital benefits of habitat creation: spatially targeting native woodland using GIS. Landsc Urban Plan 75:227– 243 Benton TG, Vickery JA, Wilson JD (2003) Farmland biodiversity: is habitat heterogeneity the key? Trends ecol evol 18:182–188 Brainard J, Lovett A, Bateman I (1999) Integrating geographical information systems into travel cost analysis and benefit transfer. Int J Geogr Inf Sci 13(3):227–246 Brainard J, Bateman I, Lovett A (2001) Modelling demand for recreation in English woodlands. Forestry 74:423–438 Buttoud G (2000) How can policy take into consideration the ‘full value’ of forests? Land Use Policy 17:169–175 Church RL, Stroms DM, Davis FW (1996) Reserves selection as a maximal covering location problem. Biol Conserv 76:105–112 Costanza R, d’Arge R, de Groot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, O’Neil R, Paruelo J, Raskin RG, Sutton P, van den Belt M (1998) The value of the world’s ecosystem services and natural capital. Ecol Econ 25(1):3–15

123

1264 De Groot RS, Wilson MA, Bouman RMJ (2002) A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecol Econ 41:393– 408 Ferrier S (2002) Mapping spatial pattern in biodiversity for regional conservation planning: where to go from here? Syst Biol 51:331–363 Frede HG, Bach M (conference organisers) (2005) Multifunctionality of landscapes; analysis, evaluation and decision support. International conference held 18–19 May 2005 at Justus Liebig University, Giessen, Germany Gilliams S, van Orshoven J, Muys B, Kros H, Heil GW, van Deursen W (2005) AFFOREST sDSS: a metamodel based spatial decision support system for afforestation of agricultural land. New Forests 30:33–53 Gkaraveli A, Good JEG, Williams JH (2004) Determining priority areas for native woodland expansion and restoration in Snowdonia National Park, Wales. Biol Conserv 115:395–402 Gren I, Groth K, Sylven M (1995) Economic values of Danube floodplain. J Environ Manage 45:333–345 Guo Z, Xiao X, Gan Y, Zheng Y (2001) Ecosystem functions, services and their values—a case study in Xingshan County of China. Ecol Econ 38:141–154 Haines-Young R, Watkins C, Wale C, Murdock A (2006) Modelling natural capital: the case of landscape restoration on the South Downs, England. Landsc Urban Plan 75:244–264 Hein L, van Koppen K, de Groot RS, van Ierland EC (2006) Spatial scales, stakeholdera and the valuation of ecosystem services. Ecol Econ 57:209–228 Hill G, Courtney P, Burton R, Potts J (2003) Forests’ role in tourism. Final report for the Forestry Commission, Edinburgh, Scotland Hill G, Courtney P (2006) Demand analysis projections for recreational visits to countryside woodlands in Great Britain. Forestry 79:185–200 Kuttila M, Pukkala T (2003) Combining holding-level economic goals with spatial landscape-level goals in the planning of multiple ownership. Landsc Ecol 18:529–541 Kazmierski J, Kram M, Mills E, Phemister D, Reo N, Riggs E, Tefertiller R, Erickson D (2004) Conservation planning at the landscape scale: a landscape ecology method for regional land trusts. J Environ Plan Manage 47(5):709–736 Lathrop RG, Bognar JA (1998) Applying GIS and landscape ecological principles to evaluate land conservation alternatives. Landsc Urban Plan 41:27–41 Lee JT, Woddy SJ, Thompson S (2001) Targeting sites for conservation: using a patch-based ranking scheme to assess conservation potential. J Environ Manage 61(4):367– 380 Lee JT, Bailey N, Thompson S (2002). Using geographical information systems to identify and target sites for the creation and restoration of native woodlands: a case study of the Chiltern Hills, UK. J Environ Manage 64(1):25–34 MLURI (1996) Evaluation of the farm woodland premium scheme. A report for Scottish Office Agriculture, Envi-

123

Landscape Ecol (2007) 22:1255–1264 ronment & Fisheries Department. Economics and Policy Series No 1, Macaulay Land Use Research Institute, Aberdeen Moilanen A, Franco AMA, Eary RI, Eary RI, Fox R, Wintle B, Thomas CD (2005) Prioritizing multiple-use landscapes for conservation: methods for large multi-species planning problems. Proc Roy Soc B-Biol Sci 272:1885–1891 Moilanen A (2005) Methods for reserve selection: interior point search. Biol Conserv 124:485–492 Opdam P, Steingro¨ver E, van Rooij S (2006) Ecological networks: a spatial concept for multi-actor planning of sustainable landscapes. Landsc Urban Plan 75:322–332 Opdam P, Verboom J, Pouwels R (2003) Landscape cohesion: an index for the conservation potential of landscapes for biodiversity. Landscape Ecol 18:113–126 Potschin M, Haines-Young R (2006) Rio+10’’, sustainability science and landscape ecology. Landsc Urban Plan 75:162–174 SDD (1990) Circular No 13/90 Indicative forestry strategies. The Scottish Office Development Department, Edinburgh SDD (1999) Circular 9/1999 Indicative forestry strategies. The Scottish Office Development Department, Edinburgh Stuart-Murray J (1994) Indicative forestry strategies—a critique. Scottish Forestry 48(1):16–21 Stuart-Murray J, Winterbottom SJ, Young JA (1999) Evidence for the effectiveness of forestry indicative strategies - a case study in the Scottish Borders. Scott For 53(3) Thompson S, Larcom A, Lee JT (1999). Restoring and enhancing rare and threatened habitats under agri-environmental agreements: a case study of the Chiltern Hills area of outstanding Natural Beauty, UK. Land Use Policy 16:93–105 Turner RK, van den Bergh JCJM, So¨derqvist T, Barendregt A, van der Straaten J, Maltby E, van Ierland EC (2000) Ecological-economic analysis of wetlands: scientific integration for management and policy. Ecol Econ 35:7– 23 Van der Horst D, Gimona A (2005) Where new farm woodlands support Biodiversity Action Plans: a spatial multicriteria analysis. Biol Conserv 123:421–432 Van der Horst D (2006a) Spatial cost-benefit thinking in multifunctional forestry: towards a framework for spatial targeting of policy interventions. Ecol Econ 59:71–80 Van der Horst D (2006b) A prototype method to map the potential visual amenity benefits of new farm woodlands. Environ Plan B: Plan Design 33:221–238 Verboom J, Foppen R, Chardon P, Opdam P, Luttikhuizen P (2001) Introducing the key patch approach for habitat networks with persistent populations: an example for marshland birds. Biol Conserv 100:89–101 Vos CC, Verboom J, Opdam PFM, ter Braak CJF, (2001) Towards ecologically scaled landscape indices. Am Nat 157:24–51 Wilson KA, McBride M, Bode M, Possingham HP (2006) Prioritising global conservation efforts. Nature 40:337– 340