Research Paper
Benefits of clearing forest plantations to restore nature? Evidence from a discrete choice experiment in Flanders, Belgium

https://doi.org/10.1016/j.landurbplan.2014.02.006Get rights and content

Highlights

  • We investigate WTP for nature restoration compared to forest plantations.

  • Preference heterogeneity is explored through mixed logit and latent class models.

  • People prefer landscape diversity, high biodiversity and good site accessibility.

  • We find support for small-scale conversions of forest plantations.

  • We find no distance-decay, but a significant effect of perceived substitution.

Abstract

To ensure the long-term survival of its most valuable and threatened habitats, the European Union (EU) is committing its Member States to develop a network of protected areas. Flanders (northern Belgium) is a highly urbanised region, where natural environments are scarce. Policy-makers are converting existing forest plantations (mostly former coniferous plantations) into natural areas to comply with the EU requirements about nature restoration and satisfy the growing demand for recreation and amenity spaces.

The conversion of forest plantations into higher value nature, however, sometimes meets public opposition because it often involves clearcuts and landscape modification. Regional planning authorities are looking for case studies demonstrating which type of nature restoration is valued and thus supported by citizens. Past valuation studies show that personal, site-specific and spatial characteristics influence preferences. However, little is known about the relative importance of such factors.

We conduct a discrete choice experiment to investigate preferences for nature restoration scenarios that involve forest conversion. A mixed logit and a latent class model are estimated and the influence of socio-demographic characteristics is explored. Willingness-to-pay (WTP) estimates are elicited. Though people generally prefer the forest habitat type, our results suggest that public support exists for converting forest plantations if this contributes to increasing landscape diversity and species richness. Based on our findings, we recommend small scale cuts. This in order to gently open the landscape, assist the natural regeneration process and help current species adapt to that landscape modification.

Introduction

To ensure the long-term survival of Europe's most valuable species and habitats, European Union (EU) Member States are committed to designate protected areas and considerable funds are allocated with the aim to protecting biodiversity in Europe. Several nature restoration projects are being implemented through the EU Birds and Habitats Directives and Natura 2000, a network of protected areas throughout the EU. Meeting such EU targets is not an easy task in densely populated and urbanised areas.

Flanders (northern Belgium) is a highly urbanised region with a strongly fragmented landscape, where natural environments remain scarce. Most notably, forests represent 13.1% of the Flemish territory (177,424 ha) and are mostly scattered in pieces of less than 1 ha. Biodiversity is consequently threatened. About half of the plants and animals are “red-listed” species (INBO, 2014a). The loss of suitable habitats and the decline in environmental quality explain most of this negative trend.

To date, Flanders has 62 Natura 2000 areas (i.e. 166,187 ha or 12.3% of the territory). In addition, the Flemish Government committed itself, through the Flemish Decree for Nature Conservation, to implementing an effective ecological network via two initiatives: the Flemish Ecological Network (“VEN”) and the Integral Interrelation and Support Network (“IVON”).

This nature restoration effort involves turning many agricultural lands, plantations and woodlands back to heathland or native broadleaved forest. Existing research (Liekens et al., 2013) demonstrates public preferences for converting agricultural lands into nature areas in a similar context. Conversions to forests in particular are found to be preferred over other habitat types, such as wetland or heathland. However, Flanders is witnessing a trend of clearing forests (especially coniferous plantations introduced in the late 19th century) to restore heathland (Verheyen, Lust, Carnol, Hens, & Bouma, 2006). Nowadays, this unique habitat is one of the most threatened habitats in Belgium (Maes, van Dyck, Vanreusel, & Cortens, 2003) and accommodates a number of endangered species. The willingness to pay (WTP) estimates from Liekens et al. (2013) suggest that converting forest to heathland might result in a loss of societal value. However, it is unclear if one can extrapolate those results (preferring forest over heathland) to the conversion of forest plantation into higher value nature areas.

Understanding public preferences for converting production-oriented forest stands back to heathland or native broadleaved forests is a complex matter that deserves attention and the careful consideration of its implications on land use planning decisions. Forest conversion involves clearcutting practices that traditionally meet strong opposition from the public (Bradley and Kearney, 2007, Ribe and Matteson, 2002). The size of the logged area is particularly influential. Past studies show that public opinion of small clearcut areas is usually more positive than of larger areas (Bradshaw, 1992, Tahvanainen et al., 2001). Bliss (2000) points out that people's opinion about forest clearcutting is also based on the perceived ecological benefits.

This case study has the double objective (i) to contribute to the limited literature related to public preferences for nature restoration involving forest conversion, and (ii) to inform policy-makers on how to design community-supported restoration policies. We address this question in response to a strong demand for literature on landscape preferences from policy-makers and regional planners. A discrete choice experiment (DCE) is conducted to elicit preferences for hypothetical restoration scenarios. WTP estimates are derived by means of mixed logit and latent class models that control for taste heterogeneity.

The remainder of this paper is organised as follows: the next section presents the rationale behind preferences for nature restoration. Section 3 briefly introduces the case study. Then, Section 4 describes our methodology and Section 5 outlines our modelling approach. The results of the estimated models and consequent marginal WTP are presented in Section 6. Section 7 discusses those results and Section 8 concludes the paper.

Section snippets

Public preferences for nature restoration

Public preferences are heterogeneous (Swallow, Weaver, Opaluch, & Michelman, 1994). Environmental valuation studies typically account for this by including environmental (Adamowicz, Nelson, Naidoo, Polasky, & Zhang, 2011), infrastructural (Brainard et al., 2001, Roovers et al., 2002), spatial (Geoghegan et al., 1997, Johnston et al., 2002), or individual parameters (Adamowicz, Swait, Boxall, Louviere, & Williams, 1997) to their econometric model. In our study, we investigate three dimensions of

The Drongengoed case

The case study presented in this paper relies on a survey conducted at the “Drongengoed”, Flanders (Fig. 1). With a total size of 860 ha, this site is the largest one-piece nature area in the province of East-Flanders. Protected for a large part under the Habitat Directive, this site is rich in biodiversity. A rare collection of sedges (Carex spp.), heath milkwort (Polygala serpyllifolia) and waxcap mushrooms (Hygrocybe spp.) indicate species-rich grasslands. Uncommon butterflies, such as the

Nature valuation through discrete choice experiments (DCEs)

DCE is a preference elicitation method introduced by Louviere and Hensher (1982) and commonly used in nature valuation (Hoyos, 2010). Information gathered through a survey allows modelling preferences for hypothetical nature restoration scenarios. Respondents are generally presented several choice sets. For each choice set they are asked to choose between two or more alternatives described by attributes. At least one attribute of the alternative is systematically varied across respondents so

Empirical approach

We estimate Eq. (1) using a mixed logit (MXL) model, then a latent class model (LCM). Stated preference literature indicates the need to better represent heterogeneous preferences in choice modelling (Colombo, Hanley, & Louviere, 2009). Combining MXL and LCM approaches offers the advantage to explore preference heterogeneity from two different angles.

The mixed logit (MXL) model is a generalisation of the standard multinomial logit model that allows controlling for unobserved taste heterogeneity

MXL model

The MXL results are reported in Table 4. The ASC term is positive and significant. This suggests that respondents positively value a nature restoration scenario that would convert 50 ha of the current coniferous tree cover to heathland, add more common species and maintain the accessibility level.

Most random parameters expected to influence individual's preferences, are statistically significant. Concerning changes in habitat composition, preferences diverge as expected in accordance with the

Discussion

This case study contributes to the limited literature about public preferences for nature restoration that involve the conversion of existing forests. We investigate possible reasons for preferring either to maintain existing forest sites (generally devoted to forestry rather than to conservation or recreation) or to restore nature sites to meet the requirements of the European Commission.

Public preferences are elicited by means of a DCE proposing to convert a coniferous woodland back to

Conclusion

This study supports previous findings that people are willing to pay to convert coniferous forest plantations to restore higher value nature, such as heathland or broadleaved forest. Public preferences are, however, not straightforward. They rely on site, individual and context-dependent factors. We observe that respondents favour restoration projects that involve limited forest conversion. We confirm that the WTP is highest for maximising landscape diversity by creating small glades within the

Acknowledgments

Data collection was funded by “Regionaal Landschap Meetjesland” and “Agentschap voor Natuur en Bos”. We are grateful for the “Marie Curie Actions” mobility grant co-financed by the European Commission and the Belgian Science Policy that allowed for cooperation with Dr. Wendy Chen. Thanks as well to Rik Hendrix for sharing his specific ecological knowledge about the site and to Alistair Beames revising the use of English in this article.

References (61)

  • M. Rambonilaza et al.

    Land-use planning and public preferences: What can we learn from choice experiment method?

    Landscape and Urban Planning

    (2007)
  • P. Roovers et al.

    Visitor profile, perceptions and expectations in forests from a gradient of increasing urbanisation in central Belgium

    Landscape and Urban Planning

    (2002)
  • R.S. Rosenberger et al.

    Attitudes, willingness to pay, and stated values for recreation use fees at an urban proximate forest

    Journal of Forest Economics

    (2012)
  • R. Scarpa et al.

    Valuing the recreational benefits from the creation of natures in Irish forests

    Ecological Economics

    (2000)
  • A. Stenger et al.

    Valuing environmental goods and services derived from the forests

    Journal of Forest Economics

    (2009)
  • L. Tahvanainen et al.

    Forest management and public perceptions – Visual versus verbal information

    Landscape and Urban Planning

    (2001)
  • R.K. Turner et al.

    Valuing nature: Lessons learnt and future research directions

    Ecological Economics

    (2003)
  • W. Adamowicz et al.

    Nature-based tourism and recreation values

  • Z. Aminrad et al.

    Influence of age and level of education on environmental awareness and attitude: Case study on Iranian students in Malaysian Universities

    The Social Sciences

    (2011)
  • M. Bech et al.

    Effects coding in discrete choice experiments

    Health Economics

    (2005)
  • Belgian Federal Government

    Structure de la population (Population structure)

    (2013)
  • M. Ben-Akiva et al.

    Discrete choice analysis

    (1985)
  • R.K. Blamey et al.

    Attribute causality in environmental choice modelling

    Environmental and Resource Economics

    (2002)
  • J.C. Bliss

    Public perceptions of clearcutting

    Journal of Forestry

    (2000)
  • G.A. Bradley et al.

    Public and professional responses to the visual effects of timber harvesting: Different ways of seeing

    Western Journal of Applied Forestry

    (2007)
  • J. Brainard et al.

    Modelling demand for recreation in English woodlands

    Forestry

    (2001)
  • R.T. Carson et al.

    A common nomenclature for stated preference elicitation approaches

    Environmental and Resource Economics

    (2011)
  • S. Colombo et al.

    Modeling preference heterogeneity in stated choice data: An analysis for public goods generated by agriculture

    Agricultural Economics

    (2009)
  • E. Deutskens et al.

    Response rate and response quality of internet-based surveys: An experimental study

    Marketing Letters

    (2004)
  • N. Hanley et al.

    Using choice experiments to value the environment

    Environmental and Resource Economics

    (1998)
  • Cited by (49)

    • The effect of information transfer related to soil biodiversity on Flemish citizens' preferences for forest management

      2021, Science of the Total Environment
      Citation Excerpt :

      This study uses Flanders, northern Belgium, as a case study region. Flanders is a densely populated area (487 inhabitants/km2) with high urbanization rates, a high pressure on land and a low forest cover of about 11% (Agentschap Natuur en Bos, 2019; De Valck et al., 2014; Statistiek Vlaanderen, 2020). Due to fragmentation, forest patches are relatively small and scattered and experience negative effects of neighboring land uses such as through atmospheric deposition and agricultural management (Decocq et al., 2016).

    View all citing articles on Scopus
    View full text