Elsevier

Ecological Indicators

Volume 32, September 2013, Pages 259-263
Ecological Indicators

Short communication
Scale and ecological dependence of ecosystem services evaluation: Spatial extension and economic value of freshwater ecosystems in Italy

https://doi.org/10.1016/j.ecolind.2013.03.034Get rights and content

Highlights

  • Scale and “ecological” dependence of ecosystem services evaluation.

  • We tested the effectiveness of the CORINE land cover dataset to assess the spatial extension of freshwater biomes in Italy.

  • Different buffer widths are assigned to Italian rivers and streams using a georeferenced hydrographic network database.

  • Spatial data from the CORINE and the modified hydrographic network database are compared.

  • The economic value of Italian freshwaters is greatly underestimated by the CORINE dataset.

Abstract

Using land cover datasets derived from satellite imagery as a proxy for ecosystem services (ES) mapping and evaluation generates a series of technical problems mainly related to the resolution at which spatial data are acquired. Small sized and highly fragmented ecosystems such as wetlands, streams and rivers are particularly underestimated when coarse resolution land cover datasets are used. However, even at finer resolution, a large fraction of the economic value provided by freshwater ecosystems may not be adequately captured by remote sensed data because, other than technical and methodological problems, the provision of many freshwater services is strictly dependent on the interactions and contacts of these ecosystems with contiguous environments. The paper analyzes and discusses the effectiveness of the CORINE land cover (CLC) dataset to represent the spatial extension and, indirectly, the economic value of freshwater biomes in Italy, with emphasis on rivers and streams. Using a georeferenced national hydrographic network database and applying a stream buffer of variable width to more than 20,000 water courses, we demonstrated that less than 10% of the estimated surface is captured by the CLC dataset with heavy negative consequences on the overall evaluation of ES at local, regional and national scale. Our approach, easily applicable also to other geographical and spatial contexts, may contribute to improve proxy-based methodologies in ES mapping/evaluation and may help to limit errors in the estimation of freshwater services.

Introduction

The natural functions of ecological systems generate an interconnected series of processes that directly or indirectly produce goods (such as food) and services (such as climate regulation or soil formation) for the human society (MA, 2005). At the same time these processes sustain the equilibrium of the system itself and the equilibrium of the entire Biosphere (De Groot et al., 2002).

The ecosystems maintain natural capital stocks and flows which in part converge in human systems, often after various transformations. In this context, natural originated goods and services are not considered as an integrated part of the economic systems, but they are simply utilized (Daily, 1997).

Costanza et al. (1997) provided one of the most complete reviews of methods and values for the assessment of these “Ecosystem Services” (ES). Without entering in the complex debate on the methods used to assign a monetary value to non-marketable “goods and services” (De Groot et al., 2002, Spangenberg and Settele, 2010), we simply remark that once all services provided by an ecosystem or biome are identified and a monetary value (per unit area) is assigned to each service, the total value of natural capital associated to a territory is strictly dependent on the spatial extension of constituent biomes. In other words, we can use land cover attributes as a proxy for ES valuation (Konarska et al., 2002, Feng et al., 2010; but see Eigenbrod et al., 2010).

However, using land cover datasets derived from satellite imagery in ES valuation, may generate a series of technical problems mainly related to the scale dependence of the spatial extension of some biomes. For example, Konarska et al. (2002) demonstrated that the overall economic value of ES for the conterminous United States increased by 200% when comparing national land cover datasets derived from 1-km (NOAA–AVHRR) and 30-m (Landsat ETM) satellite imagery resolution. They also stated that the spatial extension of small sized, highly fragmented and almost linear extending ecosystems such as streams, rivers and wetlands is greatly underestimated at the coarser resolution (−56% for Lakes/Rivers; −5155% for wetlands).

Based on the conclusions of Konarska et al. (2002), we hypothesized that, in the context of a heterogeneous and highly fragmented land cover mosaic, the spatial extension of some biomes is still highly underrepresented also when fine-resolution satellite imageries are used. Furthermore, other than the uncertainty of the “measure” influenced by the resolution of land cover datasets, the economic evaluation of goods and services provided by rivers, streams and fresh waters in general, poses another conceptual problem. The “open system” nature (Allan, 1995) and the extreme heterogeneity in space and time of running water ecosystems suggest that their structure and functions are strictly dependent on contacts and exchanges with contiguous environments (Ward, 1989, Thorp et al., 2006). For example, the provision of many freshwater services is strictly dependent on the functional linkage with the riparian zone (Allan, 2004, Gregory et al., 1991, Naiman and Decamps, 1997, Naiman et al., 2005, Tabacchi et al., 1998). However, the exact extension of the buffer necessary to ensure a full functionality to freshwater systems is still a matter of debate (Barker et al., 2006, Lees and Peres, 2008, Parkyn, 2004, Spackman and Hughes, 1995, Vidon and Hill, 2004, Wenger, 1999), though almost all researchers agree that relatively large buffer widths (>200 m) are needed to sustain multiple ecosystem services (Jones et al., 2010).

These considerations clearly suggest that the boundaries of freshwater biomes extend well over the physical limit defined by the aquatic habitat and that their extension/evaluation is not adequately captured by satellite imagery.

These inherent and methodological problems consequently cause a documented distorted evaluation of the role of rivers and wetlands for the human well-being and particularly exposes these ecosystems to unsustainable policies and management (Thorp et al., 2010).

This paper is thus aimed (i) to test the effectiveness of a 30-m resolution land cover dataset to assess the spatial extension of Italian freshwater ecosystems with emphasis on rivers and streams, (ii) to estimate the economic value of freshwater biomes in Italy and (iii) to demonstrate that a large fraction of this natural capital is not captured by the methods currently employed in national and regional based ES evaluation.

Section snippets

Methods

We utilized the CORINE land cover 2006 dataset (CLC-2006) to represent the land cover mosaic of terrestrial biomes in Italy. The CLC-2006 dataset (EEA, 2007) is based on satellite imagery data (SPOT-4 and Landsat-7 ETM+, resolution 20/30-m) which are interpreted and converted into 44 different classes of cover types (APAT, 2005), represented and mapped at a scale of 1:100,000 with a minimum mapping unit of 25 ha and minimum width of 100 m for linear elements. The CORINE classes were grouped and

Results and discussion

From the CLC-2006 dataset, the category “Lakes/Rivers” in Italy extends for 218,730 ha, representing about 1% of the total surface of terrestrial biomes. A large fraction of this area (169,300 ha, 78%) is covered by 498 lakes and only 115 major watercourses, with a total area of 49,430 ha, are recognized by the CORINE system. Therefore, according to the CLC dataset, many regions in Italy did not have any running waters at all (Fig. 1). Clearly, this is a huge underestimation of the real spatial

Conclusions

We demonstrated that in a natural heterogeneous context, such as the Italian landscape, with a hydrographic network almost totally represented by small sized rivers and streams and highly fragmented wetlands, the CORINE land cover dataset (30-m resolution), widely used in Europe, is inadequate to estimate the areal extent of freshwater biomes. Other than the problem of scale dependence (resolution of land cover datasets) we also introduced the question of “ecological dependence” of ES

Acknowledgements

We thank P. Lombardo (Rome), F.M. Pulselli (Siena) and two anonymous reviewers for helpful comments. S. Salso (L’Aquila) helped with GIS analysis and graphics. The paper was partially funded by the Italian Ministry of University and Research (MUR ex 60%) and the Regione Abruzzo (Direzione LL.PP, Servizio Qualità delle Acque).

References (38)

  • APAT (Agenzia per la Protezione dell’Ambiente e per i servizi Tecnici)

    La realizzazione in Italia del progetto europeo CORINE Land Cover 2000

    APAT Rapporti

    (2005)
  • L.S. Barker et al.

    Use of Maryland biological stream survey data to determine effects of agricultural riparian buffers on measures of biological stream health

    Environ. Monit. Assess.

    (2006)
  • A.J. Castelle et al.

    Wetland and stream buffer size requirements – a review

    J. Environ. Qual.

    (1994)
  • R. Costanza et al.

    The value of the world's ecosystem services and natural capital

    Nature

    (1997)
  • EEA (European Environmental Agency)

    CLC-2006 Technical Guidelines

    (2007)
  • F. Eigenbrod et al.

    The impact of proxy-based methods on mapping the distribution of ecosystem services

    J. Appl. Ecol.

    (2010)
  • ESRI, 2008. ArcGIS 9 – Geoprocessing Commands Quick Reference Guide. ESRI, United States....
  • X. Feng et al.

    Remote sensing of Ecosystem Services: an opportunity for spatially explicit assessment

    Chin. Geogr. Sci.

    (2010)
  • Cited by (38)

    • Pharmaceuticals in water and sediment of small streams under the pressure of urbanization: Concentrations, interactions, and risks

      2022, Science of the Total Environment
      Citation Excerpt :

      Dendritic networks of smaller streams, which make up more than 88% of the global freshwater stream length and ensure habitat for many taxa (Maloney et al., 2020), are strongly subjected to contamination by PhACs because small, low-order streams (Strahler number 1–3; Strahler, 1957) are common recipients of treated and untreated wastewater (Casado et al., 2019; Ding et al., 2016; Mandaric et al., 2018; Sousa et al., 2019). There has been growing recognition of the importance of small-sized and highly fragmented aquatic ecosystems (Di Sabatino et al., 2013), and more research has focused on the pollution of low-order streams both in urban and rural areas (Aleksander-Kwaterczak and Plenzler, 2019; Englert et al., 2013; Kienle et al., 2019; Liu et al., 2020; Proia et al., 2016; Szöcs et al., 2017). In smaller streams, it was observed that owing to the lower dilution compared with that in rivers with a higher water flow, the PhAC concentration may be related to the effluents of wastewater treatment plants (WWTPs) (Ebele et al., 2017; Maasz et al., 2019; Paíga et al., 2019).

    View all citing articles on Scopus
    View full text