Abstract

Biodiversity, landscape aesthetics and grazing livestock have considerable relevance for agricultural production, however rarely considered in public landscape preferences research. This paper studies public preferences for pasture usage by the means of a discrete choice experiment using a representative sample of 449 individuals from Germany. Graphical representations of the choice sets were used to assess the preferences for the presence of livestock and typical pasture landscape elements. To account for preference heterogeneity, the paper utilised a latent-class logit model. Four different latent classes were identified. The results showed different preferences between the latent classes, not only in terms of the magnitude of the estimated parameters, but also in terms of the parameter signs. This indicated that there are multiple types of preferred pasture landscapes. Within the groups, the preference for livestock presence did not depend on its density. Furthermore, it was found that point elements were more important for the landscape preference than linear elements. The class membership probabilities were influenced by multiple sociodemographic variables, including the individual's residence.

Keywords

Discrete choice experiment; Landscape preferences; Latent class logit model; Livestock; Pasture landscapes

Published in

Land Use Policy
2020, volume: 91, article number: 104371
Publisher: ELSEVIER SCI LTD

Authors' information