Impact of land ownership and altitude on biodiversity evaluated by indicators at the landscape level in Central Italy

Land Use Policy 45 (2015) 43–51

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Impact of land ownership and altitude on biodiversity evaluated by indicators at the landscape level in Central Italy Roberto Mancinelli, Vincenzo Di Felice, Emanuele Radicetti, Enio Campiglia ∗ Department of Agriculture, Forests, Nature and Energy, University of Tuscia, Via San Camillo De Lellis, 01100 Viterbo, Italy

a r t i c l e

i n f o

Article history: Received 4 April 2014 Received in revised form 10 December 2014 Accepted 4 January 2015 Keywords: Indicators Landscape Sustainability Agroecosystem Biodiversity

a b s t r a c t This study analyzed the ecological role of agroecosystems in maintaining landscape diversity and sustainability using selected indicators and indexes in Central Italy. The study focuses on analysing ecological systems at the landscape level to assess the environmental sustainability in terms of land cover composition and diversity. We used geographical information systems, landscape fragmentation and heterogeneity to determine the ecotone types and patch characteristics and gather useful information based on ownership (private and public areas) and altitude (plain, hilly, and mountain areas). The results showed that private areas had major landscape heterogeneity, which increased both the presence of ecotones and land cover diversification. There was a higher level of landscape heterogeneity and land cover diversification in plain and hilly areas, where agriculture was mostly present. There was also an increase of margins and/or hedgerows. Publicly owned areas generally had permanent land cover types, such as woods and pastures with low human influence, that induce biodiversity conservation. Privately owned areas were mainly characterized by cultivated land with high human influence that requires careful biodiversity management. However, most important ecological structures, such as ecotones, occurred where human land use intensity was at its highest level (plain-hilly areas and private ownership areas), and this could help to maintain and increase the biodiversity of the agricultural landscape. In Central Italy, there is an acceptable level of landscape sustainability achieved by integrating both private and public types of management. © 2015 Elsevier Ltd. All rights reserved.

Introduction Over the last century, many environmental problems have emerged, and agriculture is becoming a disturbing factor for environmental sustainability in developed countries. Modern intensive and conventional agriculture violates the ecological principles of closed cycles, the spatial and temporal dimensions of cycles, and the diversity and complexity of agroecosystems (Jorgensen and Nielsen, 1996; Lin, 2011). To reverse this trend, it is advisable to change from a reductionist approach to an ecological approach (holistic, systemic, and interdisciplinary) when choosing a type of agricultural management. Agriculture plays an active and positive role in preserving and protecting agroecosystem sustainability (Caporali et al., 2003) by properly organizing a territory (Ghersa and Leòn, 1999). Biodiversity means the variability among living organisms from all sources, including inter alia diversity of ecosystems within

∗ Corresponding author. Tel.: +39 0761 357538; fax: +39 0761 357 558. E-mail address: [email protected] (E. Campiglia). http://dx.doi.org/10.1016/j.landusepol.2015.01.008 0264-8377/© 2015 Elsevier Ltd. All rights reserved.

species and between species (UNCED, 1992). Therefore, it is possible to associate biological diversity with ecosystem diversity and landscape diversity. Biodiversity contributes to security, resiliency, social relations, health, and freedom of choice and actions (OECD, 2001, 2002). One way to increase biodiversity in agroecosystems where natural vegetation patches have been virtually eliminated is to establish different types of field margin vegetation and/or hedgerows that serve as biological corridors to guarantee the sustainability of the systems (Altieri, 1999). Land management choices directly and indirectly affect biodiversity and environmental sustainability (e.g., in time and space crop diversification favours an equilibrium in associated biodiversity, such as weeds, insects, etc., that increases the integrity of the ecosystem) (Altieri, 1999; Forman, 1995a, 1995b). Furthermore, important ecosystem services are provided by biodiversity in farm systems and associated agricultural landscapes (Tscharntke et al., 2012). Landscape and biodiversity are ecologically connected and mutually dependent (Thies and Tscharntke, 1999). Additionally, there are mutual interdependencies among land cover, land use and biodiversity (Haines-Young, 2009). Biodiversity exists in a matrix

44

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

of habitat patches, including managed and natural environments. Thus, habitat change is the most important cause of biodiversity change and loss in the ecosystem (MEA, 2005). Activities that support and increase landscape heterogeneity, biodiversity and ecological sustainability in agroecosystems include the following: the development of biocentres (permanent existence of flora and fauna), biocorridors (allow migration of flora and fauna among biocentres), heterogeneous crops, crop structuring (organization of growing areas), polyculture and intercropping, crop rotation, biocontrol introduction, and modification of pesticide use (Caporali et al., 2010). It is known that habitat fragmentation affects biodiversity and is generally defined as a landscape-scale process involving both habitat loss and separation. However, habitat fragmentation per se does not necessarily have a strong impact on biodiversity. Therefore, if agriculture is managed in a sustainable way, it can play an important role in landscape sustainability because biodiversity should be preserved in protected agricultural areas (Donald and Evans, 2006). There are areas in the Mediterranean where the evolution of agroecosystems has produced a typical and diversified landscape, such as in Central Italy. Indicators are essential for evaluating biodiversity. The “indicators” and the “indexes” (UNCED, 1992) are able to translate large quantities of information into simple numbers and permit the quantification of biodiversity, evaluation of sustainability and quick comparisons of various situations far in space and/or in time (Ghersa et al., 2002). In the scientific literature, the words index and indicators are occasionally confused. In this research, an index is considered to be a numerical synthesis aggregating several indicators to gain further information. Consequently, the indicator is a measure of an environmentally relevant phenomena (direct indicators considered interesting for the aim of the study). The index is the indirect quantification obtained through the aggregation of several indicators of significant phenomena relating to the state of the agroecosystem. The use of biodiversity indicators and indexes enables us to evaluate agroecosystem sustainability at the landscape level (Moonen and Bàrberi, 2008). Ecosystem biodiversity indicators can be observed and measured in terms of composition (species richness and distribution) and variation in structure (e.g., fragmentation, ecotones, number and size of the patches) and function (flow paths relative to system processes at scale level) (Noss, 1990). In this study, the biodiversity of an area of Central Italy was evaluated at the landscape level using indicators elaborated from data collected by remote sensing images, cartography, and fieldwork. The main objectives of this research were the following: (i) to evaluate biodiversity at the landscape agroecological level using structural biodiversity indicators and indexes and (ii) to determine the influence of territorial characteristics, such as ownership and altitude in land cover patterns.

Materials and methods Approach adopted The methodology used in this study was based on the analysis of landscape hierarchical levels according to the concepts of landscape (Halffter, 1998), ecomosaics (Forman, 1995b) and integrated organization levels (Caporali et al., 2010). The landscape analysis was based on photograph interpretation, cartographical analysis, and fieldwork. All of the information obtained was used to build a database to determine the core set of selected indexes and indicators (Table 1) needed to evaluate agroecosystem sustainability in terms of biodiversity.

The aerial photography was classified according to vegetation type (described below) and provides an excellent source of data for performing the structural studies of a landscape (Sachs et al., 1998). Simple measurements of patterns, such as the number and size (area and perimeter) of patches, were taken because they can indicate the functionality of a land cover type better than the total area of the cover alone (Forman, 1995a). High-resolution aerial imagery is also effective for extracting individual objects and permits the detailed classification of lands. These types of images have previously been used to analyze landscape patterns (Dunbar et al., 2003). In this study, the GIS analysis technique was applied (Balram et al., 2004). Aerial photos (imagery acquired in the year 2000 – 1 pixel = 1 m2 ) were used to construct a land cover map (at the same resolution) for the entire study area (Forman, 1995a, 1995b). After photograph interpretation, the data (4% of mapped units) were verified directly on the field according to a method used by Congalton and Green (1999) and Cozzi and Ferretti (2003). The patches are small ecological landscape features that represent relatively homogeneous, spatially explicit landscape functional units. The number, area, and perimeter of the patches and the length of other landscape characteristics, such as streets and rivers, were used in this study according to Odum and Turner (1990) and Forman (1995a, 1995b). Each patch was classified into categories based on land cover. The land cover classes (LCs) selected for the study area were herbaceous crops (HC), tree crops (TC), woods (W), hedges (H), grassland (G), shrub-grassland (SG) (areas with shrubs scattered or widely distributed over fields of native herbaceous species), and buildings (B) in terms of surface and rivers (Wa) and roads (Ro) in terms of length. Consequently, it was possible to assess some important features of the landscape, such as heterogeneity, starting from structural information and described using the landscape indicators reported by Magurran (1988). The structural information refers to the use of natural resources, such as land, and affects the composition and the organization of these system components. The landscape analysis was also distinguished by the ownership of the patches (private and public) and the altitude of the patches (plain, hill and mountain). These parameters were defined by national maps because they are a principal driving force of landscape composition and management. Diversity consists of richness and abundance components. Diversity can be measured by one of these components or by measures incorporating both factors. Moreover, methods for evaluating the diversity of a high ecological hierarchical level, such as landscape, are closely related to techniques for measuring species diversity (Magurran, 1988). In this study, the selected core set includes both new and existing indicators and indexes. Ecotones are transition zones between different ecosystems or patches and are generally characterized by a large variety of species and properties that at times do not exist in either of the adjacent systems (Odum, 1993). Particular attention is given to the core set construction. According to Duelli (1997), ecotones with high structural heterogeneity, such as forest edges and hedgerows, enhance regional biodiversity and the abundance and diversity of beneficial organisms. According to Marshall and Moonen (2002), ecotones (or field margins) in farmland play an important role in biodiversity by interacting with adjacent farmlands. Each indicator refers to a specific characteristic. Several indicators and indexes were combined to understand the various agroecological aspects and different relationships among the components and important agroecosystem processes. The indicators and indexes used were the following (Table 1): the complementary of the Simpson index [S ], Margalef index [MAR], Menhinick index [MER], McIntosh index [U], Berger–Parker index [BP ], patch complexity index [PC], ecotope diversity index [d] (patch as ecotope)

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

45

Table 1 Core set of selected indicators and indexes. Indicators and indexes [symbol]

Formulae

Simpson [SI ]

1−

References

 s   p2j



j=1

[8,14,19,23,26]

s 



Margalef [MAR]

(s − 1) /

Menhinick [MER]

⎛ ⎞   s s/ ⎝ nj ⎠

ln

nj

[6,13,14,19]

j=1

[14,19]

j=1

McIntosh [U]

   s 2 nj

[17]

j=1

Berger–Parker [BP ] Patch complexity [PC] Ecotope diversity [d]

pmax (∀: pmax ≥ pj ) P/A

 s 

ln

2

(pj )

1/2

j=1

Connectivity index [RSi] log series ˛ [˛]

LCI  J /(pj ·b) s 

[4,14,24] [1,9,24] /ln(ε)



nj (1 − x)

[18] [24]

/x

[11,19]

j=1

Ecotone density [ED]

nj  e

i

[21,25]

ai

 nj  

i=1

Ecotone intensity [EI]

nj /

eij

 nj i=1  eij

Mean patch ecotone [MPE]

  

[12,21]

/nj

[7,20]

/nj

[5,16,21,22]

i=1 nj

aij

Mean patch size [MPS]

Patch density [PD]

nj /

i=1   nj 

aij

[5,15,21,22]

i=1

Hydrogeological risk [HR] Protect area [PPA] Road density [RD] River density [RiD] Forested areas [] Arable land/woods ecotone ratio [ˇ] Arable land/woods area ratio [] Herbaceous crops/tree crops ecotone ratio [ı] Herbaceous crops/tree crops area ratio [] Arable land/hedge ecotones ratio [] Arable land/hedge area ratio [] Total ecotones/arable land ecotones ratio [] Total area/arable land area ratio [ ] Woods/total vegetate ecotone ratio [ ]

HRA/A·100 PA/A ·100 RL/A R/A Wa/TVa (HCe + TCe)/We (HCa + TCa)/Wa HCe/TCe HCa/TCa (HCe + TCe)/He (HCa + TCa)/Ha Te/(HCe + TCe) Ta/(HCa + TCa) We/Te

[8] [2] [2] [3] [3,10]

Legend: A = total area; a = patch’s area; e = patch’s perimeter; HRA = hydrogeological risk area in hectares; i = patch; j = land cover class; n = number of patches; p = proportion of the land cover class; P = total perimeter of the patches; PA = protected area in hectares; R = rivers length; RL = roads length in metres; s = number of land cover class; ε = 1/(e + b) [where e = 271,828 and b = area of studied area in hectare]. References: [1] Baker and Cai, 1992; [2] Fammler et al., 2000; [3] Steel et al., 2004; [4] Berger and Parker, 1970; [5] Caporali et al., 2003; [6] Clifford and Stephenson, 1975; [7] Corona et al., 2004; [8] Crimella et al., 2001; [9] EC, 2005; [10] EEA, 2003; [11] Fischer et al., 1943; [12] Forman, 1995a; [13] Legendre and Legendre, 1983; [14] Magurran, 1988; [15] McCarigal and Marks, 1995; [16] McCarigal et al., 2002; [17] McIntosh, 1967; [18] Yue et al., 2004; [19] Pielou, 1975; [20] Riitters et al., 1995; [21] Rutledge, 2003; [22] Saura and Martinez-Millan, 2001; [23] Simpson, 1949; [24] Turner et al., 1989, 2001; [25] Ward and Tockner, 2001 and [26] Krebs, 1989.

(Naveh and Lieberman, 1994), connectivity index [RSi], log series ˛ index [˛] (Kempton and Taylor, 1976), ecotone density indicators [ED], ecotone intensity indicators [EI], mean patch ecotone indicators [MPE], mean patch size indicator [MPS], patch density indicator [PD], hydro-geological risk indicator [HR], protected area indicator [PPA], road density indicator [RD], river density indicator [RiD], and forested area indicator [].

The other new indicators adopted in this study that completed the core set and improved its agroecological significance were the following (Table 1): the ecotone ratio between arable land and woods [ˇ], area ratio between arable land and woods [], ecotone ratio between herbaceous crops and orchards [ı], area ratio between herbaceous crops and orchards [], ecotone ratio between arable land and hedges [], area ratio between arable land and

46

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

hedges [], ecotone ratio between total ecotones and arable land [], total area ratio between ground surface and arable land [ ], ecotone ratio between woods and total vegetation [ ]. To simplify the reading and explanation of the results, we performed a statistical normalization (range 0–10 per class of land cover) of the mean patch size (MPS) indicator and mean patch ecotone (MPE) indicator and graphically aggregated the results in spider graphs. The aim is to compare the score of the selected sub-systems simultaneously. Several indicators (mean patch size, mean patch ecotone, and ecotone density) were calculated using the standard errors of the mean values. Study area The study area represents an example of landscape management and human interference in an area located in Central Italy (Fig. 1) between the Tyrrhenian coast of the Mediterranean sea (West coast of Italy) and the Apennine mountains (41◦ 28 38 –41◦ 39 16 N and 12◦ 55 00 –13◦ 09 51 E) from 10 to 1500 m a.s.l. The area investigated is approximately 16,166 ha and is 24% plain (2952 patches and 3410 ha within the classified classes), 27% hill (2529 patches and 4525 ha within the classified classes) and 49% mountain (1269 patches and 7439 ha within the classified classes). The study area includes three municipalities (Carpineto Romano, Sermoneta, and Bassiano) with approximately 13,000 inhabitants. The borders of the municipalities represent the physical limits of the area. The lands are 58% private and 42% public. The climatic characteristics are heterogeneous and range from Mediterranean to temperate climates, which causes a high level of biological diversity. The annual rainfall ranges from 830 to 1530 mm, and there is a dry period during the summer. Although rainfall generally increases according to altitude, there is no strict relationship. Factors such as local orientation (for instance facing north or south), orography and so on seem to play an important role in landscape diversity. On the plains, the summer climate is typically semi-arid. The soil is mainly calcareous in the mountain area and there are heterogeneous characteristics (various types of deposits approximately 10 m deep) on the plains (Sevink et al., 1984). The data show that 74% of the study area is classified by the Italian Ministry of Environmental Protection as being at hydro-geological risk. Protected areas represent 24% of the total area, and these areas are mainly situated in the mountains. Results The landscape is characterized by agricultural activity (herbaceous and tree crops) on the plain (86.5%) and private property (45.8%) areas (Table 2). Agricultural activity accounts for 21.5%

Fig. 1. Area location, altimetric (a) and ownership (b) classification.

and 26.6% of the total area in terms of number and area of patches, respectively. The public ownership is characterized by less anthropised LCs (woods, grassland and shrub-grasslands), both in terms of the number of patches (69.0% of total) and of the area (98.6% of total). Hedges are mostly present below 600 m a.s.l. (plain and hilly areas) and in private ownership areas. The woodland increases with altitude in terms of area (3.4% on the plain, 54.9% in the hill and 59.7% in the mountain areas), and it is larger in the public (20.3%) than in the private (8.1%) ownership areas. Other small LCs, such as private gardens, artificial lakes, and sport areas, together with roads and rivers only account for 4.5% of the area. Therefore, they are not discussed in this study due to their

Table 2 Percentage of land cover classes on the number and area of patches. Indicators

Plain (%)

Hill (%)

Mountain (%)

Public (%)

Private (%)

Total (%)

On number of patches Herbaceous crops Tree crops Woods Hedges Grasslands Shrub-grassland Buildings

13.5 9.2 2.6 12.2 5.7 4.1 52.7

15.9 12.6 12.2 13.0 8.3 10.1 27.9

3.5 1.3 23.3 9.0 26.6 28.1 8.2

5.2 4.5 20.3 13.6 19.9 28.8 7.7

13.9 9.8 8.1 11.6 8.8 7.4 40.4

12.5 9.0 10.0 11.9 10.6 10.9 35.1

On area of patches Herbaceous crops Tree crops Woods Hedges Grasslands Shrub-grassland Buildings

72.3 14.2 3.4 1.3 4.1 2.5 2.2

8.0 15.3 54.9 1.1 8.8 10.8 1.1

0.8 0.2 59.7 0.3 22.8 16.2 0.0

0.4 0.6 64.5 0.3 20.4 13.7 0.1

32.6 13.2 31.6 1.1 10.1 9.9 1.5

18.8 7.8 45.7 0.8 14.5 11.6 0.8

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

minimum spatial dimensions and because they were not the main objective of this study. On the plain (Fig. 2a), where wood plots are closely connected to the agricultural matrix, these wood plots (even if the size of the wood patches is very small) are more capable of developing ecotones than the wood patches in mountain areas. This characteristic is also observed for grasslands. The data in Fig. 2 show how woods on plains generate more ecotones compared to herbaceous crops. On the plain, woods with 1.51 ha of MPS have 1257 m of MPE and herbaceous crops with 6.11 ha of MPS (four times larger) have only

47

1114 m of MPE. In the mountain areas, there is a different situation and woods with 15 ha of MPS are only able to sustain 2128 m of MPE. As expected, the MPS of less anthropised LCs increase significantly with increased altitude. The results indicate differences between the public and private ownerships in the wood LCs (Fig. 2b). The MPS is significantly larger in public ownerships than in private ownerships. However, this difference does not have any effect on the ability of developing ecotones (MPE). The MPS and MPE indicator values were larger in cultivated plots in private ownerships and larger less anthropised

7

HC = herbaceous crops, TC = tree crops, W = woods, H = hedges, G = grassland, SG = shrubs and grassland

MPS-TC

MPE-HC

6

MPS-HC 6

[f=± 0.01] [h=± 0.01] [m=± 0.04]

MPE-TC

[f=± 0.02] [h=± 0.02] [m=± 0.05]

4

[f=± 0.02] [h=± 0.02] [m=± 0.06]

5

[f=± 0.01] [h=± 0.01] [m=± 0.04]

2

3

4

2

MPS-W

MPE-W

[f=± 0.03] [h=± 0.02] [m=± 0.02]

0

1

[f=± 0.10] [h=± 0.05] [m=± 0.06]

MPE-H

MPS-H

[f=± 0.01] [h=± 0.02] [m=± 0.03]

[f=± 0.05] [h=± 0.05] [m=± 0.08]

MPE-G

MPS-G

(a)

[f=± 0.31] [h=± 0.19] [m=± 0.15]

Plain (f) Flat (f)

MPE-SG

[f=± 0.22] [h=± 0.18] [m=± 0.15] 7

Hill (h) Mountain (m)

MPS-SG

[f=± 0.15] [h=± 0.18] [m=± 0.14]

[f=± 0.24] [h=± 0.19] [m=± 0.16]

MPS-HC 6

MPE-HC 6

5

[pu=± 0.04] [pr=± 0.01]

[pu=± 0.04] [pr=± 0.01]

MPS-TC 4

MPE-TC 4

[pu=± 0.04] [pr=± 0.01]

3

[pu=± 0.03] [pr=± 0.01]

2

2

MPE-W [pu=± 0.05] [pr=± 0.03]

1

MPS-W

0

[pu=± 0.05] [pr=± 0.03]

MPS-H

MPE-H

[pu=± 0.06] [pr=± 0.03]

[pu=± 0.06] [pr=± 0.03]

MPS-G

(b)

MPE-G [pu=± 0.21] [pr=± 0.13]

[pu=± 0.19] [pr=± 0.13]

MPS-SG

Public (pu) [pu=± 0.16] Private (pr) [pr=± 0.14]

MPE-SG

[pu=± 0.16] [pr=± 0.14]

Fig. 2. Indexes of diversity at the landscape hierarchical level (MPS and MPE) for local comparisons between altitudes (a) and ownership types (b) (standard error values are reported in brackets).

48

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

Table 3 Patch density (PD) and ecotone intensity (EI) indicators. Indicators

Plain

Hill

Mountain

Public

Private

Total

PDa Herbaceous crops Tree crops Woods Hedges Grasslands Shrub-grassland Total vegetate

Number ha−1 0.16 0.56 0.66 8.16 1.17 1.41 0.42

1.11 0.46 0.12 6.72 0.52 0.53 0.41

0.78 0.89 0.07 4.91 0.19 0.30 0.16

2.04 1.29 0.05 7.04 0.15 0.35 0.15

0.28 0.48 0.16 6.88 0.55 0.48 0.39

0.29 0.51 0.10 6.91 0.31 0.41 0.29

EIb Herbaceous crops Tree crops Woods Hedges Grasslands Shrub-grassland Total vegetate

Number km−1 0.90 1.67 0.84 4.51 0.99 1.87 1.38

2.09 1.40 0.55 4.57 1.38 1.46 1.32

1.71 2.08 0.47 3.70 0.77 1.00 0.78

3.35 2.55 0.67 1.04 0.69 1.06 0.85

1.22 1.45 1.22 2.16 1.11 1.39 1.25

1.28 1.53 1.09 1.96 0.94 1.23 1.13

a b

Number of patches per hectare of land. Number of ecotones per kilometre of ecotones.

Table 4 Edge density (ED) of the land cover classes (standard error values are reported in brackets).

HC TC W H G SG

Plain (m ha−1 )

Hill (m ha−1 )

Mountain (m ha−1 )

Public (m ha−1 )

Private (m ha−1 )

Total (m ha−1 )

864 (±45) 960 (±46) 989 (±48) 2241 (±56) 1519 (±60) 1177 (±44)

944 (±30) 772 (±31) 590 (±20) 2289 (±97) 802 (±28) 730 (±26)

752 (±72) 714 (±114) 483 (±18) 5626 (±1801) 642 (±21) 588 (±17)

1939 (±193) 1071 (±113) 470 (±21) 2641 (±197) 661 (±29) 644 (±22)

819 (±22) 835 (±27) 647 (±18) 2764 (±319) 998 (±30) 808 (±23)

896 (±26) 854 (±27) 589 (±14) 2741 (±262) 896 (±23) 737 (±16)

plots (woods, grasslands, and shrub-grasslands) in public ownerships. Table 3 shows the results of patch density (PD) and the ecotone intensity (EI) indicators. In terms of total area results, the hedge PD and EI indicators are the highest values (6.91 hedges per hectare and 1.96 hedge ecotones in a kilometre of hedge ecotones, respectively). In this study, PD hedge features are more related to the variation in altitude (8.16 on plain, 6.72 in hill, and 4.91 in mountain) than to ownership (7.04 and 6.88, in public and private ownerships, respectively). Conversely, EI hedge features are more related to the two

types of ownership (1.04 and 2.16, in public and private ownerships, respectively) than altitude (4.51 on plain, 4.57 in hill and in 3.70 mountain). Table 4 shows the results of the edge density index (ED). This index characterizes the landscape and quantifies the edge length of LCs per hectare of LCs. On cultivated lands, herbaceous crops show the highest ED values (896 vs 854 for HC and TC, respectively), and in hilly areas (944 vs 772 for HC and TC, respectively) on plains, the tree crop ED values are higher (864 vs 960 for HC and TC, respectively).

Table 5 Indexes and indicators of sustainability at the landscape level. Indexes

Plain

Hill

Mountain

Public

Private

Total

SI MAR MER U BP PC d RSiHC RSiTC RSiW RSiH RSiNH RSiES ˛ ˇ  ı     

0.428 0.415 0.107 0.248 0.260 0.089 0.252 0.740 0.145 0.035 0.013 0.041 0.025 0.505 6.96 25.36 2.71 3.56 7.58 66.84 1.66 1.16 0.02 0.03

0.642 0.400 0.094 0.407 0.445 0.076 0.265 0.081 0.155 0.555 0.011 0.089 0.109 0.486 0.75 0.43 0.85 0.52 5.82 21.55 3.28 4.28 0.15 0.57

0.566 0.425 0.118 0.345 0.404 0.067 0.199 0.008 0.003 0.596 0.003 0.228 0.162 0.519 0.06 0.02 2.93 2.77 1.18 3.48 41.03 92.87 0.16 0.61

0.523 0.434 0.126 0.313 0.354 0.059 0.197 0.004 0.006 0.646 0.003 0.204 0.137 0.530 0.07 0.02 0.88 0.73 1.19 3.17 32.10 98.29 0.43 0.64

0.749 0.369 0.069 0.504 0.669 0.099 0.276 0.331 0.134 0.321 0.011 0.103 0.101 0.448 1.14 1.40 1.68 2.47 6.80 42.3 2.60 2.18 0.27 0.32

0.710 0.358 0.061 0.465 0.539 0.099 0.265 0.190 0.079 0.461 0.008 0.146 0.117 0.434 0.78 0.57 1.66 2.41 5.81 35.17 3.64 3.76 0.34 0.44

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

The results of the other indicators and indexes that provide useful information on landscape diversity are reported in Table 5. The results of the SI , U, BP and d indexes (referring to areas with patches of vegetation) suggest a higher level of biodiversity in hilly areas compared to plain and mountain areas. Additionally, there is more biodiversity in private ownerships compared to public ownerships. Hilly areas show good land use concerning human activities and the natural environment. There is more intensive land on the plains than in the mountains where the use of the land is more conservative. The values of the MAR, MER and ˛ indexes (referring to number of patches of vegetation) show the highest level of biodiversity in mountain areas compared to plains and hills and in public ownerships compared to private ownerships. It is important to note that mountain and public areas have a lower number of vegetation patches compared to the other sub-systems (17% fewer patches of vegetation in mountains than on plains, 36% fewer patches in mountains than in hilly areas and 69% fewer patches of vegetation in public ownership than private ownership), and this could influence the index results. The results of MAR, MER and ˛ indexes reported in Table 5 show a higher level of diversity on the plain compared to the hilly areas (MAR = 0.415 and 0.400; MER = 0.107 and 0.094; ˛ = 0.505 and 0.486; in plain and hill, respectively). The PC index describes the complexity of the ecomosaic and is higher on private property (0.099) and on the plains (0.089). The landscape connectivity measured with the RSi index is higher for herbaceous crops on the plains (0.740), followed by woods in the mountains (0.596), private herbaceous crops (0.331) and public woods (0.321). The hedges have the lowest RSi (0.008, reported to total) and a lower level of connectivity than the other LCs. This finding indicates the agroecological function (e.g., refuge for wild animals). The indexes concerning the ecotone ratio (ˇ) and the area ratio () of the arable land and woods show the highest values on the plain (ˇ = 6.96 and  = 25.36). These results indicate a lack of woodland below 200 m a.s.l. and a tendency of the cultivated lands to reduce the amount of ecotones. The ˇ and  values show a higher level of ecotones on cultivated lands over 200 m a.s.l. (ˇ = 0.75 and 0.06,  = 0.43 and 0.02 in hill and mountain, respectively). There is an equal amount of cultivated land and woodland (1.14) in the private ownership areas, while in public ownerships, the woods cover most of the area (0.07). The indexes concerning the ecotone ratio (ı) and the area ratio () between herbaceous crops and tree crops, which suggests a higher capacity of tree crops to generate ecotones on plains (ı = 2.71 and 0.06,  = 3.56). In the hilly areas, the situation is different (ı = 0.85 and 0.06,  = 0.52). and this is probably due to the different patch complexity. The ecotone ratio () and the area ratio () between the arable land and hedges emphasize the important role of the hedges for the sustainability of ecosystems and especially agroecosystems. The results show higher values both for  (7.58, 5.82 and 1.18 in plain, hill and mountain, respectively) and  (66.84, 21.55 and 3.48 in plain, hill and mountain, respectively) on the plain. The ecotone intensity and patch density are only slightly influenced by the altitude (hill and mountain areas). Regarding the role played by cultivated land in the study area, the indexes of the ecotone () and area ( ) ratios between the total amount of ecotones and the arable land ecotones underline the positive influence of cultivated lands in favouring the design of a more complicated landscape. On the plain, cultivated land ( = 1.66 and = 1.16) plays a more important role than in hilly areas ( = 3.28 and = 4.28). The indexes of the forest area indicator ( ) and the ratio of wood and total vegetation ecotones () show that woodland sustains fewer ecotones (3%, 57% and 61% of woodland sustains only 2%, 15% and 16% of the total amount of ecotones in plain, hill and mountain areas, respectively). These results confirm the role of woods as relatively important and widespread elements in Central

49

Italian landscape management, and this is more common in public ownerships in mountain areas ( = 0.64 and  = 0.43). The road density (RD), protected areas (PPA) and river density (RiD) indexes (data not reported in the tables) enable us to complete the information and the description of the study area. The road density indicator shows a limited presence of roads (0.031, 0.028 and 0.007 km km−2 in plain, hill and mountain areas, respectively), while the national average is 0.57 km km−2 . The protected area covers 24.5% of the total study area and mainly consists of forested areas over 600 m a.s.l. The river density indicator (RiD) is, on average, 0.56 m km−2 (1.10, 0.78, and 0.17 m km−2 on the plains, hills, and mountains, respectively). The hydro-geological risk indicator (HR) shows an important environmental and social sustainability issue in the study area (74.14%), which is much higher than the average national level (7.1%) and is directly linked to the conservation of the soil and biological diversity of the landscape.

Discussion The land cover map produced in this study supplies basic information in terms of landscape indicators and enabled us to determine the various land cover classes (LCs) in terms of the number and area of patches. We also accounted for the ownership and altitude sub-systems of the landscape. The core set indicators and indexes show that the main aspects of landscape biodiversity in Central Italy is characterized by small farms concentrated mainly on the plains. Therefore, the design of the landscape is “patch-worked”. The landscape is characterized by a high level of agricultural activity, mainly on the plains, and is closely linked with hedges, which represent important agroecological infrastructures with structural and functional characteristics (e.g., wind protection, habitat for wildlife), especially where the landscape is subject to intensive farming practices (Caporali et al., 2003, 2010; Walker et al., 2006). It is important to highlight the influence of hedges for increasing the fragmentation and heterogeneity of agroecosystems when describing landscape diversity and its influence on a lower hierarchical level of biodiversity. The woodland area is an important indicator that is connected to biodiversity protection aspects (Agbogidi and Ofuoku, 2009). On the plains, the presence of woodland is observed in areas with particular conditions or positions (e.g., steep slopes for agricultural management, low productive soils, road or river proximity), which cause the irregular shapes of the woodland patches. This situation could depend on the farmers’ choices concerning the use of their land, such as the allocation of tree crops. Vineyards and kiwis have regular shapes and usually grow on plains and low hilly areas with low-medium slopes. Olive trees tend to have irregular shapes and usually grow on hills with medium-high slopes. However, this situation highlights the role of the hedges in increasing landscape heterogeneity and complexity in all territories and sub-systems. Hedges provide suitable conditions for the sustainable agricultural management of an area with good agroecological values (Sklenick, 2006). The high complexity of the ecomosaic observed on the plain and on private properties is most likely due to the result of historical management choices of the territory. These choices depended on the environmental, economical, and social aspects of these areas. The higher level of landscape fragmentation in the private compared to the public management areas is certainly due to the small size of the farms and fields, which is a typical aspect of the rural territory in Central Italy. Landscape heterogeneity is a multi-faceted term with no strict definition. In human managed territories and in areas with agricultural cultivated land, a more heterogeneous landscape has a higher level of diversity (Cabell and Oelofse, 2012) among cover types and a higher complexity in the spatial patterning

50

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51

of the cover types. The homogeneous scenarios or areas with limited heterogeneity and complexity may lack a positive link between diversity and plot heterogeneity (Di Falco et al., 2010). In the mountain areas where the cultivations are rare and there is a low level of land use, the ownership is mainly public and there are many areas with permanent cover types such as woods and pastures. This condition of low influence of human activity enhances biodiversity conservation.

Conclusions Our data show a more fragmented landscape in the private management area agricultural lands compared to the public management areas, and there is an elevated level of heterogeneity on plains and hills. This heterogeneity increases the amount of ecotones that improves agroecosystem diversity and the presence of biocorridors, which are the areas where such structural features are particularly important. The landscape heterogeneity could be an issue in the plain and hilly areas where agriculture is more intensive. Therefore, maintaining and increasing biodiversity could be a difficult task. Sustainable agricultural land cover also requires the maintenance of high-level biodiversity on the landscape (Heywood, 1999a, 1999b; Tscharntke et al., 2012). In the area of Central Italy examined, agriculture caused a high level of fragmentation of the original landscape and generated a characteristic landscape with a high number of small fields that are cropped and managed differently. This type of management increases both heterogeneity and ecotones, and according to Stoate et al. (2001), it can also increase biodiversity. This study suggests that the levels and characteristic anthropic pressure on the land are determined by different management approaches. These pressures affect the land cover and use and are key drivers of biodiversity and ecosystem services, which is consistent with Haines-Young (2009). The outcomes of the study also suggest that the overall habitat loss has a higher effect on biodiversity than the spatial arrangement of the agroecosystems. In Central Italy, there is an integration of the private–public matrix that allows for biodiversity integrity. Public ownership provides habitat conservation, while private ownership confers a spatial arrangement of the agroecosystems.

Acknowledgments The authors would especially like to thank Bosa Bachisio and Claudio Stefanoni for their valuable help with the fieldwork required for this study.

References Agbogidi, O.M., Ofuoku, A.U., 2009. Forestry extension: implications for forest protection. Int. J. Biodivers. Conserv. 1, 98–104. Altieri, M.A., 1999. The ecological role of biodiversity in agroecosystems. Agric. Ecosyst. Environ. 74, 19–31. Baker, W.L., Cai, Y., 1992. The rule programs for multiscale analysis of landscape structure using the GRASS geographical information system. Landsc. Ecol. 7, 291–302. ´ ´ S., Meredith, T., 2004. A collaborative GIS method for inteBalram, S., Dragicevi c, grating local and technical knowledge in establishing biodiversity conservation priorities. Biodivers. Conserv. 13, 1195–1208. Fammler, H., Veidemane, K., Ruskule, A., Simanovska, J., Indriksone, D., Kipper, K., 2000. 2nd Baltic State of the Environment Report: Based on Environmental Indicators. Baltic Environmental Forum, Gandrs Ltd., Latvia. Berger, W.H., Parker, F.L., 1970. Diversity of planctonic Foraminifera in deep sea sediments. Science 168, 1345–1347. Cabell, J.F., Oelofse, M., 2012. An indicator framework for assessing agroecosystem resilience. Ecol. Soc. 17 (1–18), 1–13.

Caporali, F., Mancinelli, R., Campiglia, E., 2003. Indicators of cropping system diversity in organic and conventional farms in Central Italy. Int. J. Agric. Sustain. 1, 67–72. Caporali, F., Campiglia, E., Mancinelli, R., 2010. Agroecologia: Teoria e pratica degli agroecosistemi (Agroecology: Theory and Practice of the Agroecosystems). De Agostini Scuola SpA, Novara, ISBN 978-88-251-7352-9, 222 pp. Clifford, H.T., Stephenson, W., 1975. Introduction to Numerical Classification. Academic Press, London. Congalton, R.G., Green, K., 1999. Assessing the Accuracy of Remotely Sensed Data: Principles and Practices. Lewis Publisher, New York. Corona, P., Chirici, G., Travaglino, D., 2004. Forest ecotone survey by line intersect sampling. Can. J. For. Res. 34 (8), 1776–1783. Cozzi, A., Ferretti, M., 2003. Procedure di controllo di qualità dei dati di prima fase. Inventario nazionale delle foreste e dei serbatoi forestali di carbonio (Procedures for Quality Control of Data in First Phase. National Inventory of Forests and Forest Carbon Stocks). MiPAF – Direzione Generale per le Risorse Forestali Montane e Idriche, Corpo Forestale dello Stato, CRA-ISAFA, Trento. Crimella, A., Pareglio, S., Valentinelli, A., Venuta, M.L., 2001. Agenda 21: Indicatori di sostenibilità locale (Agenda 21: Indicators of Local Sustainability). Fondazione Lombardia per l’Ambiente, Milano. Di Falco, S., Penov, I., Aleksiev, A., van Rensburgc, T.M., 2010. Agrobiodiversity, farm profits and land fragmentation: evidence from Bulgaria. Land Use Policy 27, 763–771. Donald, P.F., Evans, A.D., 2006. Habitat connectivity and matrix restoration: the wider implications of agri-environment schemes. J. Appl. Ecol. 43, 209–218. Duelli, P., 1997. Biodiversity evaluation in agricultural landscapes: an approach at two different scales. Agric. Ecosyst. Environ. 62, 81–91. Dunbar, M.D., Moskal, L.M., Jakubauskas, M.E., Dobson, J.E., Martinko, E.A., 2003. Computer visualization of forest cover change: human impacts in Northeastern Kansas and natural disturbance in Yellowstone National Park. In: Proceedings American Society for Photogrammetry and Remote Sensing Annual Conference, Anchorage, AK, May 2003. EEA, 2003. An Inventory of Biodiversity Indicators in Europe 2002. European Environment Agency, Copenhagen. EC, 2005. Trends of Some Agri-Environmental Indicators in the European Union: Report EUR 21565 EN. European commission, Joint Research Centre, Bruxelles. Forman, R.T.T., 1995a. Some general principles of landscape and regional ecology. Landsc. Ecol. 10, 133–142. Forman, R.T.T., 1995b. Land Mosaics. The Ecology of Landscapes and Regions. Cambridge University Press, Cambridge. Ghersa, C.M., Leòn, R.J., 1999. Landscape changes induced by human activities in the rolling pampas grassland. In: Eldridge, D., Freudenberger, D. (Eds.), Proceedings of the 6th International Rangelands Congress, vol. 2. , pp. 624–629. Ghersa, C.M., Ferraro, D.O., Omacini, M., Martìnez-Ghersa, M.A., Perelman, S., Satorre, E.H., Soriano, A., 2002. Farm and landscape level variables as indicators of sustainable land-use in the Argentine Inland-Pampa. Agric. Ecosyst. Environ. 93, 279–293. Haines-Young, R., 2009. Land use and biodiversity relationships. Land Use Policy 26S, S178–S186. Halffter, G., 1998. A strategy for measuring landscape biodiversity. Biol. Int. 36, 3–17. Heywood, V.H., 1999a. Conservation of wild relatives of native European crops. In: Janick, J. (Ed.), Perspectives on New Crops and New Uses. ASHS Press, Alexandria, pp. 146–147. Heywood, V.H., 1999b. Trends in agricultural biodiversity. In: Janick, J. (Ed.), Perspectives on New Crops and New Uses. ASHS Press, Alexandria, pp. 2–14. Jorgensen, S.E., Nielsen, S.N., 1996. Application of ecological engineering principles in agriculture. Ecol. Eng. 7, 373–381. Kempton, R.A., Taylor, L.R., 1976. Models and statistics for species diversity. Nature 262, 818–820. Krebs, C.J., 1989. Ecological Methodology. Harper Row, New York. Legendre, L., Legendre, P., 1983. Numerical Ecology: Developments in Environmental Modelling, vol. 3. Elsevier Scientific Publishing Co., Amsterdam. Lin, B.B., 2011. Crop diversification: adaptive management for environmental change. Bioscience 61, 183–193. Magurran, A.E., 1988. Ecological Diversity and its Measurement. Princeton University Press, Princeton. Marshall, E.J.P., Moonen, A.C., 2002. Field margins in northern Europe: their functions and interactions with agriculture. Agric. Ecosyst. Environ. 89, 5–21. McCarigal, K., Marks, B.J., 1995. FRAGSTATS: Spatial Pattern Analysis Program for Quantifying Landscape Structure. Coastal Oregon Productivity Enhancement Program, Portland. McCarigal, K., Cushman, S.A., Neel, M.C., Ene, E., 2002. FRAGSTATS: Spatial Pattern Analysis Program for Categorical Maps, Version 3.0. Amherst, University of Massachusetts. McIntosh, R.P., 1967. An index of diversity and the relation of certain concepts to diversity. Ecology 48, 392–404. MEA, 2005. Millennium Ecosystem Assessment. Island Press. Moonen, A.-C., Bàrberi, P., 2008. Functional biodiversity: an agroecosystem approach. Agric. Ecosyst. Environ. 127, 7–21. Naveh, Z., Lieberman, A.S., 1994. Landscape Ecology: Theory and Application. Springer-Verlag, New York. Noss, R.F., 1990. Indicators for monitoring biodiversity: a hierarchical approach. Conserv. Biol. 4, 355–364. Odum, E.P., 1993. Ecology and Our Endangered Life Support Systems. Sinauer, Sunderland.

R. Mancinelli et al. / Land Use Policy 45 (2015) 43–51 Odum, E.P., Turner, M.G., 1990. The Georgia landscape: a changing resource. In: Zonneveld, I.S., Forman, R.T.T. (Eds.), Changing Landscapes: An Ecological Perspective. Springer-Verlag, New York, pp. 137–164. OECD, 2001. Valuation of Biodiversity benefits: Selected Studies. Organization for Economic Co-operation and Development, Paris. OECD, 2002. Handbook of Biodiversity Valuation: A Guide For Policy Makers. Organization for Economic Co-operation and Development, Paris. Pielou, E.C., 1975. Ecological Diversity. Wiley, New York. Riitters, K.H., O’Neill, R.V., Hunsaker, C.T., Wickam, J.D., Yankee, D.H., Timmins, S.P., Jones, K.B., Jackson, B.L., 1995. A factor analysis of landscape pattern and structure metrics. Landsc. Ecol. 10, 23–29. Rutledge, D., 2003. Landscape Indices as Measures of the Effects of Fragmentation: Can Pattern Reflect Process?, vol. 98. Department of Conservation, Science Internal Series, Wellington. Sachs, D.L., Sollins, P., Cohen, W.B., 1998. Detecting landscape changes in the interior of British Columbia from 1975 to 1992 using satellite imagery. Can. J. For. Res. 28, 23–36. Saura, S., Martinez-Millan, J., 2001. Sensitivity of landscape pattern metrics to map spatial extent. Photogramm. Eng. Remote Sens. 67, 1027–1036. Sevink, J., Remmelzwaal, A., Spaargaren, O.C., 1984. The soil of southern Lazio and adjacent Campania. Universiteit van Amsterdam, Amsterdam. Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688. Sklenick, P., 2006. Applying evaluation criteria for the land consolidation effect to three contrasting study areas in the Czech Republic. Land Use Policy 23, 502–510.

51

Stoate, C., Boatman, N.D., Borralho, R.J., Carvalho, C.R., de Snoo, G.R., Eden, P., 2001. Ecological impacts of arable intensification in Europe. J. Environ. Manag. 63, 337–365. Thies, C., Tscharntke, T., 1999. Landscape structure and biological control in agroecosystems. Science 285, 893–895. Tscharntke, T., Clough, Y., Wanger, T.C., Jackson, L., Motzke, I., Perfecto, I., Vandermeer, J., Whitbread, A., 2012. Global food security, biodiversity conservation and the future of agricultural intensification. Biol. Conserv. 151, 53–59. Turner, M.G., O’Neill, R.V., Gardner, R.H., Milne, B.T., 1989. Effects of changing spatial scale on the analysis of landscape pattern. Landsc. Ecol. 3, 153–162. Turner, M.G., Gardner, R.H., O’Neill., R.V., 2001. Landscape Ecology in Theory and Practice. Springer-Verlag, New York. UNCED, 1992. Agenda 21: Program for Action for Sustainable Development. United Nations, New York. Steel, E.A., Feist, B.E., Jensen, D.V., Pess, G.R., Sheer, M.B., Brauner, J.B., Bilby, R.E., 2004. Landscape models to understand steelhead (Oncorhynchus mykiss) distribution and help prioritize barrier removals in the Willamett basin, Oregon, USA. Can. J. Fish. Aquat. Sci. 61, 999–1011. Walker, M.P., Dover, J.W., Sparks, T.H., Hinsley, S.A., 2006. Hedges and green lanes: vegetation composition and structure. Biodivers. Conserv. 15 (8), 2595–2610. Ward, J.V., Tockner, K., 2001. Biodiversity: towards a unifying theme for river ecology. Freshw. Biol. 46, 807–819. Yue, T.X., Xu, B., Liu, J.Y., 2004. A patch connectivity index and its change in relation to new wetland at the Yellow River Delta. Int. J. Remote Sens. 25, 4617–4628.