Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

Value of semi-natural areas as biodiversity indicators in agricultural landscapes Reija Hietala-Koivu a,c,∗ , Tiia Järvenpää b , Juha Helenius c a

b

MTT Agrifood Research Finland Systems Analysis Laboratory, Helsinki University of Technology, P.O. Box 1000, FIN-02015 Helsinki, Finland c Department of Applied Biology, University of Helsinki, P.O. Box 27, Helsinki FIN-00014, Finland Received 23 April 2002; received in revised form 22 May 2003; accepted 27 May 2003

Abstract Semi-natural habitats are key elements between cultivated and natural areas. Ditch boundaries, buffer zones, barn areas and woodland patches were analyzed in Toholampi, Yläne, Nurmijärvi and Liperi localities in Finland over the period 1954–1998 by means of indicators for semi-natural areas, based on calculations of the sum of patch densities (PDsum ) and the sum of patch areas (CAsum ). There were only a few semi-natural habitats left by the late 1990s. The sharpest decrease in PDsum indicator values, by 90%, was found for Yläne, while the corresponding drops in the figures for Nurmijärvi, Toholampi and Liperi were 77, 65 and 59%, respectively. The CAsum indicator showed a similar trend. The indicators were tested by calculating the metapopulation capacity of the barn areas. The use of CAsum is suggested for semi-natural patches as a quick, cheap and robust monitoring tool and indicator of change in biodiversity value in agricultural landscapes. The indicator PDsum may be important to interpret landscape and habitat changes when attention needs to be paid to the mosaic of landscape elements. © 2003 Elsevier B.V. All rights reserved. Keywords: Agricultural landscape; Biodiversity; Ditch boundary; Indicator; Metapopulation capacity; Semi-natural habitat

1. Introduction Indicators of sustainable agriculture may be viewed as summaries of collected information and media for further discussion on the subject, and in this sense they contribute to an ongoing learning process. Sustainability indicators aim at assessing the success of natural resource management for agricultural production (Bakkes, 1997; Duelli, 1997; Yli-Viikari, 1999; Yli-Viikari et al., 2000; Wascher, 2000; von WirénLehr, 2001). One subset of these consists of landscape indicators, which address patterns, trends and rates of ∗ Corresponding author. Tel.: +358-9-19159738; fax: +358-9-19158463. E-mail address: [email protected] (R. Hietala-Koivu).

structural change in rural land-use. These may be used, or specifically developed, for the assessment of quantitative processes such as erosion or nutrient leaching, trends in the species diversity of the natural flora or fauna, or qualitative indicators such as scenic beauty in an agricultural area. The OECD has established the Driving ForceState-Response (DSR) framework, with its emphasis on connections between causes, effects and actions, as a means of developing indicators. This framework includes indicators that reflect not just ecological, but also cultural and/or economic landscape changes including ones that measure wildlife habitats, allowing further projection to the assessment of changes in biodiversity (OECD, 1998, 1999, 2000, 2001). The OECD has coordinated its work on agri-environmental

0167-8809/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0167-8809(03)00273-1

2

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

indicators with the European Union’s statistical agency EUROSTAT (Eiden et al., 2000), and a number of other international organizations, both governmental and non-governmental (UNEP, 1996; MAFF, 1999; MMM, 1999; Naturvårdsverket, 1999; Rosenström and Palosaari, 2000; Wascher, 2000). The present work aimed to develop an indicator based on the quantitative measurement of semi-natural landscape elements to assess their biodiversity value in agricultural landscapes. It was assumed that dispersion of semi-natural habitats within an intensively farmed area improved the heterogeneity of the entire agricultural ecosystem. In general, high biodiversity and aesthetic properties are associated with heterogeneity in a landscape (Stein et al., 1999). Semi-natural elements located close to cultivated fields are rarely or not at all disturbed by farming practices (i.e. no deliberate application of chemicals; OECD, 1999). Semi-natural habitats are considered to offer better conditions for wildlife than intensively farmed land. Hay meadows, road verges, field boundaries and other marginal habitats may therefore serve as connecting corridors or stepping stones for the distribution of plants and animals in an agricultural landscape (Levin and Kerster, 1974; Hanski, 1994; Wahlberg et al., 1996; Hanski et al., 1998; Boren et al., 1999; Norderhaug et al., 2000; Pino et al., 2000). According to Andren (1994) the loss of bird and mammal species is greater than expected from habitat loss alone. In contrast Burel et al. (1998) in western France showed that agricultural intensification does not always lead to a decrease in species richness, as diversity can also be maintained by species replacement. For instance, when the open field area increased large species in carabid assemblages were replaced by smaller ones with shorter life-span. Species replacement has also been observed among wintering birds and hedgerow herb species. The theory of island biogeography (MacArthur and Wilson, 1967) predicts that species richness will decline as the patch area decreases and its isolation increases. This theory suggests that historical changes in the size and density of semi-natural habitats may be directly related to species diversity in an agricultural landscape. The theory of metapopulations (Hanski, 1994; Wahlberg et al., 1996; Hanski et al., 1998, 2000) also predicts that the probability for an endangered habitat specialist to become extinct increases

as its habitat area declines and isolation between the patches increases. Hanski and Ovaskainen (2000) developed a measure, termed the metapopulation capacity (λM ) that can be used to rank landscapes in terms of their capacity to support viable metapopulations. This paper addresses the question of how to measure biodiversity value in agricultural landscapes. It was assumed that both the patch density and the sum of the areas of semi-natural patches correlate positively with species diversity in an agricultural environment. Testing was performed for one type of semi-natural patch, barn areas, in relation to species assumed to have a preference for them. The proposed indicators are based on sums of patch density and class areas to evaluate the role of the semi-natural habitats in landscapes in the context of assessing the biodiversity value of agricultural landscapes.

2. Material and methods The material consisted of digitized aerial photographs of four localities in Finland (corner coordinates according to the Finnish grid system y; x): Toholampi (63◦ N46 , 24◦ 20 ; 2513 00;7072 00-2519 00;7076 00 according to the Finnish grid system), Yläne 60◦ N54 , 22◦ 36 (2420 60;6753 00-2427 40;6757 00), Nurmijärvi 60◦ N26 , 24◦ 38 (2532 00;6701 00-2538 00;6705 00) and Liperi 62◦ N37 , 29◦ 16 (4460 00;6946 00-4464 00;6948 00), representing the western, southwestern, southern and eastern parts of the country. The areas considered in Toholampi and Liperi were 8 km2 in size, those in Yläne and Nurmijärvi 24 km2 (Fig. 1), all of which were divided into squares of 1 km2 . Land-use habitats of >30 m2 were classified according to Hietala-Koivu (1999). The material was taken from aerial photographs (1:10 000) produced by the National Survey of Finland: Toholampi—1956, 1978 and 1995; Yläne—1958, 1977, 1993 and 1997; Nurmijärvi—1954, 1973 and 1997; Liperi—1965, 1976 and 1998. Changes in landscape composition were divided into three phases: (1) period of agricultural intensification, cattle breeding and overproduction (1954–1969) with high inputs of chemicals, many small farms in the east and north of Finland being merged to larger farms; (2) period of spatial and agricultural

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

3

Fig. 1. Location of the four areas studied, Toholampi, Yläne, Nurmijärvi and Liperi, in Finland.

specialization (1970–1994) with crop production in the south, mixed farming in the western and central parts of the country; (3) period under EU Common Agricultural Policy (CAP from 1995 to 1998) with environmentally friendly agricultural practices according to the Finnish agri-environmental scheme. Four ecologically important semi-natural land-use classes were considered, i.e. (a) ditch boundaries; (b) buffer zones along waterways; (c) woodland patches; (d) field barns. In Finland hedges are replaced by open ditches with grassy vegetation. The density of this patch type is determined by the hydrological properties of the terrain. Sloping land and good soil drainage allow for farming without ditch boundaries thought to be of great importance for maintaining species diversity in landscapes (Ruuska and Helenius, 1996; Hietala-Koivu, 1999; Ma et al., 2002). The buffer zones include both natural and artificial corridors along natural waterways. Clumps of trees and bushes up to 2000 m2 were considered as woodland patches, larger ones being treated as forests. Field barns include wooden storage barns themselves and their immediate uncultivated surroundings.

Proposed indicators of semi-natural areas: A first density indicator was obtained by calculating the sum of PD (patch density) values over the semi-natural patch types: PDsum = PD1 + · · · + PDn ,

(1)

where PD = ni /A, ni being the number of patches of type i in the landscape, and A the total area studied in km2 (McGarigal and Marks, 1995). A second indicator was based on the sum of semi-natural patch areas from i = 1–n: CAsum = CA1 + · · · + CAn .

(2)

Both indicators did not refer to connectivity between patches, but only to the number of patch types and their areas. Metapopulation capacity Semi-natural indicators are assumed to measure biodiversity value of agricultural landscapes by measuring densities and areas of the semi-natural habitats

4

Patch density, PD (n/1 km2 )

Toholampi

Nurmijärvi

Liperi

1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998

Ditch margins 232.2 Buffer areas 1.6 Barns 37.7 Woodland patches 2.8 Indicator value (Pdsum ) 292.6 Total change (%)

Yläne

193.8 1.5 22.7 2.9 238.8

80.2 1.6 8.0 3.4 102.8 −64.9

230.6 5.5 6.0 2.3 280.7

60.5 4.7 4.7 2.0 83.2

14.0 4.3 3.7 2.0 28.1 −90.0

118.7 3.8 6.5 2.7 152.8

57.6 3.3 5.2 1.9 80.6

22.0 2.8 2.7 1.8 35.4 −76.8

192.4 1.0 10.6 5.1 232.8

180.3 0.8 8.3 5.6 218.5

74.0 1.4 4.1 5.8 94.8 −59.3

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

Table 1 Patch densities (PD) of various semi-natural areas and indicator values (PDsum ) at four localities of Finland

Class areas, CA (ha)

Toholampi

Yläne

Nurmijärvi

Liperi

1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998 1950–1969 1970–1994 1995–1998 Ditch margins 49.6 Buffer areas 21.0 Barns 3.0 Woodland patches 1.8 Indicator value (Casum ) 75.4 Total change (%)

43.7 25.9 2.0 1.6 73.2

19.0 27.5 0.8 1.5 48.8 −35.3

99.6 25.2 8.1 2.3 135.2

34.4 21.5 7.0 2.5 65.4

12.8 18.3 6.3 3.2 40.6 −70.0

60.5 11.2 6.2 1.8 79.7

35.9 19.4 6.0 1.3 62.6

18.8 24.8 4.5 1.4 49,5 −37.9

44.2 9.6 1.7 1.1 56.6

42.3 9.0 1.3 1.3 53.9

19.2 6.5 0.8 1.4 27.9 −50.7

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

Table 2 Class areas (CA) of various semi-natural patches and indicator values (CAsum ) at four localities of Finland

5

6

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

Fig. 2. Location and area of barns and their surroundings in the Toholampi area (8 km2 ) in 1956, 1978 and 1995. Numbers refer to the coordinates according to the Finnish grid system y;x.

in the same way as metapopulation capacity. Metapopulation capacity of a landscape is an index that can be calculated for any animal species having a metapopulation structure. For any species, a spatially explicit habitat map is needed for calculating the metapopulation capacity. In addition, a specific parameter value describing the species dispersal ability needs to be defined. Metapopulation capacity is intended for use in conservation biology for a given species but for any landscape with known habitat map.

In this study, metapopulation capacity was used another way round. The values were calculated for one patch type over the biologically meaningful set of dispersal values, but not for any particular species. The idea was to demonstrate how changes in landscape structure of one patch type affected species depending on that patch type and having a metapopulation structure. The proposed indicators, PDsum and CAsum , were tested for barn patches against their metapopulation capacities to analyze the indicators

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

that correlated best with biologically meaningful index values. The metapopulation capacity λM of a certain fragmented landscape was defined as the leading eigenvalue of the landscape matrix M, which in turn consisted of the elements mij = exp(−αdij )Ai Aj for j = i and mii = 0, being α a species parameter, approximately the reciprocal of the migration distance of the species concerned, dij the distance between patches i and j and Ai the area of patch i (Hanski and Ovaskainen, 2000). Metapopulation capacities for barn patches were calculated using α values of 1.0, 2.0, 4.0 and 8.0, corresponding to average migration distances of 1.0, 0.5, 0.25 and 0.125 km, respectively. Heino and Hanski (2000) showed that distance dependence was weak when α = 0–2 and migration

7

propensity evolved to a relatively high level, while any increase in α implied greater effective distances between patches, so that for a very pronounced distance dependence (α = 6 or higher) metapopulations were no longer viable. Metapopulation capacities λM were calculated for barn areas in the 64 squares during the three subperiods, thereby obtaining a total of 192 squares. Both indicators were tested by analyzing the scatter plots between the calculated PDsum indicators and the metapopulation capacities, λM , and the calculated (Spearman) correlations between the CAsum values and the metapopulation capacity, λM , to assess whether these indicators were capable of describing the capacity for maintaining species with metapopulation structure in barn habitats.

Toholampi Metapopulation capacity

120 100 80

1956 1978

60

1995 40 20 0 0

10

20

30

(a)

40

50

60

70

80

PDsum indicator

Yläne Metapopulation capacity

1400 1200 1000

1958

800

1977

600

1997

400 200 0 0

(b)

2

4

6

8

10

12

14

16

18

PDsum indicator

Fig. 3. (a)–(d) Plots of metapopulation capacity (λM ) against the PDsum indicator in Toholampi, Yläne, Nurmijärvi and Liperi, respectively, in 1954–1998, based on a total of 192 squares of 1 km2 potentially occupied by a species living in barn areas (α = 1.00).

8

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

Metapopulation capacity

Nurmijärvi 900 800 700 600 500 400 300 200 100 0

1954 1973 1997

0

5

10 PDsumindicator

(c)

15

20

Liperi Metapopulation capacity

250 200

1965

150

1976 100

1998

50 0 0

(d)

2

4

6

8

10

12

14

16

18

PDsumindicator Fig. 3. (Contionued ).

3. Results An overall negative trend was observed in the sums of semi-natural patch densities (PDsum ) in all four localities, to the extent that there were only a few semi-natural habitats left by 1998. The sharpest decrease was observed in Yläne, from a PDsum value of 280.7 to 28.1, i.e. a decline of 90%. The second largest decrease in relative terms (77%) was in Nurmijärvi, whereas semi-natural patch densities for Yläne approached that for Nurmijärvi in the late 1970s (83 vs. 80). The indicator values in western and eastern parts decreased drastically during the second period, 1970–1994, by 65% in Toholampi, 59% in Liperi. Ditch boundary was by far the dominant semi-natural patch type per unit area in the period investigated, 1954–1998, and their decline contributed most to the overall decrease in the semi-natural patch density in-

dicator (PDsum ). Field barns also decreased markedly over the 44 years, whereas buffer zones and clumps of trees and bushes persisted at relatively low densities or fluctuated slightly (Table 1). The second proposed indicator (CAsum ) (Table 2) showed that the decline in area was not as drastic at the habitat type level as had been shown by the indicator based on patch densities. The most drastic decrease in area was observed in Yläne and attributable to a decline in the areas of ditch boundaries. The areas of woodland patches increased slightly in Yläne and Liperi, but decreased in Toholampi and Nurmijärvi. Buffer zones beside fields increased in Toholampi and Nurmijärvi, but decreased in the other two localities. Barn areas decreased in all four localities by 20 ha. The trend in the spatial disposition, Toholampi, where a large number of barns have been traditionally used for storing hay, is shown in Fig. 2. The most rapid

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

9

Toholampi, Yläne, Nurmijärvi and Liperi 1954-1998 Metapopulation capacity

1400 1200 1000 800 600 400 200 0 0

2000

4000

6000

8000

10000

12000

14000

16000

CAsum indicator Fig. 4. Plot of metapopulation capacity (λM ) against the CAsum indicator in Toholampi, Yläne, Nurmijärvi and Liperi in 1954–1998, based on a total of 192 squares of 1 km2 potentially occupied by a species living in barn areas (α = 1.00).

decrease was observed in 1995, but the spatial arrangement remained fairly consistent with time. The scatter plots (Fig. 3a–d) between the PDsum indicator and the corresponding metapopulation capacities (λM ) showed that the values were highly dispersed in all the areas except Toholampi (Fig. 3a). The λM values of the barn areas in Nurmijärvi and Yläne were higher than those in Toholampi and Liperi, although the PDsum indicator values for barns were lower. This is probably due to the area of the barn patches, which was larger in Nurmijärvi and Yläne than in Toholampi and Liperi. The mean patch size for a barn and its surroundings in the 1990’s was 700 m2 in Nurmijärvi and in Yläne, 200 m2 in Liperi and 100 m2 in Toholampi. The scatter plots established a trend of decreasing numbers and areas of barns and their surroundings, because the most recent values (1995–1998), were located close to the origin in all the areas, whereas the values representing the situation during the first two subperiods (1950–1969 and 1970–1994) were scattered. Therefore, the PDsum indicator did not clearly show that dispersal of patches maintained metapopulations but rather the overall dispersion of these patches. There was an obvious correlation between the CAsum indicator and the corresponding metapopulation capacities (λM ), the calculated Spearman correlations being 0.945 (P < 0.0001), 0.937 (P < 0.0001), 0.908 (P < 0.0001) and 0.844 (P < 0.0001) for α values of 1.0, 2.0, 4.0 and 8.0, respectively. The re-

lationships between the CAsum indicator and the λM values for barns at an α value of 1.0 in all four areas during 1954–1998 are shown in Fig. 4.

4. Discussion and conclusion The present results suggest that the ecological value of agricultural landscapes was reduced in Finland between the 1950s and the 1990s as it happened in Sweden (Ihse, 1995; Björklund et al., 1999). Corresponding decreases in linear field boundaries and hedgerows have also been reported in the UK, Northern Ireland and France (Le Coeur et al., 2002), yet semi-natural habitats are vital agricultural landscapes in terms of species diversity (Wahlberg et al., 1996; Hanski et al., 1998; Boren et al., 1999; Norderhaug et al., 2000). The decrease in the indicator values based on the sum of patch densities and class areas over the four semi-natural patch types describes the homogenization of the Finnish agricultural landscape. The patch density indicator, PDsum , showed a drastic change as did the CAsum indicator. The results obtained suggest a clear correlation between CAsum and metapopulation capacity, which was not the case for the PDsum indicator. This correlation is due to the fact that metapopulation capacity is sensitive to changes in areas. On the other hand, PDsum may be better suited

10

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx

for visualizing the change process itself, because it describes the dispersion of semi-natural habitats in an agricultural landscape. The results also indicate that PDsum for a given patch type serves an indicator for the capacity of the landscape to maintain the species diversity associated with that patch type. This interpretation requires that both the spatial and the size distribution of the patch type are maintained despite changes in the aggregated patch area. Thus, the metapopulation capacity interpretation of CAsum would be justified. An obvious limitation of indicators is that they do not refer to actual connectivity between semi-natural patches, but rather account for connectivity indirectly, semi-natural areas being dispersed because of traditional agricultural practices, rural history and local soil properties. This is most evident in the case of barn areas and open parcel ditches, as barns were built in the field area to store hay, while open parcel ditches were dug according to moisture content. Areas left uncultivated are usually minimal especially in the period of agricultural intensification (1950–1969). Indicators cannot act properly if semi-natural patches are located close to each other, nor if their patch size is relatively large in a particular area. Information on local agricultural structures and the physical geography of the areas concerned is needed for interpretation. CAsum may therefore be used for any semi-natural patch type as a quick, cheap and robust monitoring tool and indicator of change in biodiversity value of agricultural landscapes. The indicator PDsum may be important to interpret landscape and habitat changes when mosaics of landscape elements are to be considered.

Acknowledgements The authors would like to thank warmly the two anonymous reviewers for valuable and helpful comments in the revision of the manuscript. Thanks are extended also to Mr. Christian Eriksson (MTT Agrifood Research Finland) for advice on the statistical aspects and Mr. Malcolm Hicks for revising the language of the manuscript. Funding was provided by the Academy of Finland and the Ministry of Agriculture and Forestry under the Finnish Biodiversity

Programme (FIBRE) via the project ‘Biodiversity Implications of Agricultural Policies: An Integrated Approach’ (BIAPIA). References Andren, H., 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: a review. Oikos 71, 355–366. Bakkes, J., 1997. Research needs. In: Moldan, B., Billharz, S. (Eds.), Sustainability Indicators. Report of the Project on Indicators of Sustainable Development. Wiley, Chichester, pp. 380–386. Björklund, J., Limburg, K.E., Rydberg, T., 1999. Impact of production intensity on the ability of the agricultural landscape to generate ecosystem services: an example from Sweden. Ecol. Econ. 29, 269–291. Boren, J.C., Engle, D.M., Palmer, M.W., Masters, R.E., Criner, T., 1999. Land use change effects on breeding bird community composition. J. Range Manage. 52, 420–430. Burel, F., Baudry, J., Butet, A., Clergeau, P., Delettre, Y., Le Coeur, D., Dubs, F., Morvan, N., Paillat, G., Petit, S., Thenail, C., Brunel, E., Lefeuvre, J.-C., 1998. Comparative biodiversity along a gradient of agricultural landscapes. Acta Oecol. 19, 47–60. Duelli, P., 1997. Biodiversity evaluation in agricultural landscapes: an approach at two different scales. Ecosyst. Environ. 62, 81– 91. Eiden, G., Kayadjanian, M., Vidal, C., 2000. Capturing landscape structures: tools. http://europa.eu.int/comm/dg06/publi/ landscape/ch1.htm. Hanski, I., 1994. A practical model of metapopulation dynamics. J. Anim. Ecol. 63, 151–162. Hanski, I., Ovaskainen, O., 2000. The metapopulation capacity of a fragmented landscape. Nature 404, 755–758. Hanski, I., Lindström, J., Niemelä, J., Pietiäinen, H., Ranta, E., 1998. Ekologia. WSOY, Juva, 580 pp. ISBN 951-0-21981-9. Hanski, I., Alho, J., Moilanen, A., 2000. Estimating the parameters of survival and migration of individuals in metapopulations. Ecology 81, 239–251. Heino, M., Hanski, I., 2000. Evolution of migration rate in a spatially realistic metapopulation model. Interim Report IR00-044 for IIASA (International Institute for Applied Systems Analysis). http://www.iiasa.ac.at/Publications/Documents/IR00-044.pdf. Hietala-Koivu, R., 1999. Agricultural landscape change: a case study in Yläne, southwest Finland. Landscape Urban Plan. 46, 103–108. Ihse, M., 1995. Swedish agricultural landscape-patterns and changes during the last 50 years, studied by aerial photos. Landscape Urban Plan. 31, 21–37. Le Coeur, D., Baudry, J., Burel, F., Thenail, C., 2002. Why and how we should study field boundary biodiversity in an agrarian landscape context. Agric. Ecosyst. Environ. 89, 23–40. Levin, D.A., Kerster, H.W., 1974. Gene flow in seed plants. Evol. Biol. 7, 139–220.

R. Hietala-Koivu et al. / Agriculture, Ecosystems and Environment xxx (2003) xxx–xxx Ma, M., Tarmi, S., Helenius, J., 2002. Revisited species–area relationship in a semi-natural habitat: floral richness in agricultural buffer zones. Agricul. Ecosyst. Environ. 89, 137– 148. MacArthur, R.H., Wilson, E.O., 1967. The Theory of Island Biogeography. Princeton University Press, Princeton, NJ. MAFF, 1999. Toward Sustainable Agriculture. A Pilot Set of Indicators of the Ministry of Agriculture, Fisheries and Food. MAFF Publications, London, 72 pp. McGarigal, K., Marks, B.J., 1995. FRAGSTATS: spatial pattern analysis program for quantifying landscape structure. US Forest Service General Technical Report PNV 351. MMM, 1999. Uusiutuvien luonnonvarojen kestävän käytön yleismittarit. Maa- ja metsätalousministeriön julkaisuja 3/1999. MMM, Helsinki, Finland, 168 pp. Naturvårdsverket, 1999. Bedömningsgrunder för miljökvalitet. Odlingslandskapet. Rapport 4916, 80 pp. Norderhaug, A., Ihse, M., Pedersen, O., 2000. Biotope patterns and abundance of meadow plant species in a Norwegian rural landscape. Landscape Ecol. 15, 201–218. OECD, 1998. Report on the OECD Workshop on Agrienvironmental Indicators. Joint Working Party of the Committee for Agriculture and the Environment Policy Committee in York, September 1998. COM/AGR/CA/ENV/EPOC(98)136, 65 pp. OECD, 1999. Environmental indicators for agriculture: methods and results—the stocktaking report greenhouse gases, biodiversity, wildlife habitats. Joint Working Party of the Committee for Agriculture and the Environment Policy Committee in Paris, October 13–15, 1999. COM/AGR/CA/ ENV/EPOC(99)82, pp. 63–86. OECD, 2000. Environmental Indicators for Agriculture. Methods and Results. Executive Summary, 53 pp. OECD, 2001. Environmental Indicators for Agriculture, vol. 3. Methods and Results. http://www.oecd.org/publications/ebook/5101011e.pdf.

11

Pino, J., Roda, F., Ribas, J., Pons, X., 2000. Landscape structure and bird species richness: implications for conservation in rural areas between natural parks. Landscape Urban Plan. 49, 35–48. Rosenström, U., Palosaari, M. (Eds.), 2000. Kestävyyden mitta: Suomen kestävän kehityksen indikaattorit 2000. Suomen Ympäristö 404. Ympäristöministeriö, Helsinki, 122 pp. Ruuska, R., Helenius, J., 1996. GIS analysis of change in an agricultural landscape in central Finland. Agric. Food Sci. Finland 5, 567–576. Stein, T.V., Anderson, D.H., Kelly, T., 1999. Using stakeholders’ values to apply ecosystem management in an upper midwest landscape. Environ. Manage. 24, 399–413. UNEP (United Nations Environment Programme), 1996. Agricultural Biological Diversity. UNEP/CBD/SBSTTA/2/10. http://www.biodiv.org/sbstta2/. von Wirén-Lehr, S., 2001. Sustainability in agriculture—an evaluation of principal goal-oriented concepts to close the gap between theory and practice. Agric. Ecosyst. Environ. 84, 115– 129. Wahlberg, N., Moilanen, A., Hanski, I., 1996. Predicting the occurrence of endangered species in fragmented landscapes. Science 273, 1536–1538. Wascher, D.M. (Ed.), 2000. Agri-environmental indicators for sustainable agriculture in Europe. ECNC Publication Technical Report Series. European Centre for Nature Conservation, Tilburg, 236 pp. Yli-Viikari, A., 1999. Indicators for sustainable agriculture—a theoretical framework for classifying and assessing indicators. Agric. Food Sci. Finland 8, 265–283. Yli-Viikari, A., Hietala-Koivu, R., Risku-Norja, H., Seuri, P., Soini, K., Widbom, T., Voutilainen, P., 2000. Maatalouden kestävyyden indikaattorit. Maatalouden tutkimuskeskuksen julkaisuja. Sarja A 74. Jokioinen: Maatalouden tutkimuskeskus, 116 p. + 1 app.