Supporting sustainable soil management: Assessment of knowledge, public preferences and information to stimulate decision-making on sustainability

Soil is a crucial component of natural resources that humanity largely depends on through its contribution to many of the services delivered by ecosystems. Nevertheless, soil is experiencing increasing pressures that are often directly or indirectly connected to unsustainable human behavior. In order to preserve the wide range of ecosystem services that soil delivers, a shift towards sustainable soil-related human behavior is needed, especially in the context of ongoing climate change. Such soil-related human behavior includes several aspects and stakeholder groups that can impact soil through different pathways. Amongst these, soil management largely determines the supply of ecosystem services and has put pressure on soil in the past through intensive practices although efforts towards sustainable soil management are increasing. Additionally, soil use, demand for ecosystem services by society and related public preferences impact soil directly as well as indirectly through their effect on soil management. In this dissertation, I studied soil-related behavioral factors of managers and citizens through a number of case studies related to the two most widespread land uses in Europe, forest and agriculture. In particular, I investigated specific aspects of these behavioral factors in the soil system, focusing on knowledge and public preferences. Furthermore, I considered information as a necessary first step for behavioral change to support sustainable soil management.

Chapter 1 provides a general introduction of the intertwining between the natural soil system and the human system, referred to as the social-ecological soil system. In Chapter 2, the human system and the behaviors that it entails are discussed in more detail with a focus on knowledge and public preferences. Furthermore, this chapter introduces the two main methodologies that have been applied in this dissertation, cognitive mapping (chapter 3) and discrete choice experiments (chapters 4 to 6). In particular, the theoretical background and state-of-the-art implementation are discussed. Discrete choice experiments face a number of challenges that can impede the use of resulting values in policy making of which hypothetical bias is commonly known. Chapter 6 assesses a relatively new methodology in discrete choice experiments to reduce hypothetical bias, namely certainty calibration.

Chapter 3 is oriented towards forest managers and investigates their knowledge and understanding of the role of soil biodiversity in forest functioning as well as provides a comparison with scientists’ understanding. Moreover, chapter 3 assesses the effect of contextual factors by comparing understanding between Flanders and Romania. We found that forest managers have a basic understanding of soil biodiversity but lack in-depth understanding of ecosystem processes and soil state variables, in contrast to scientists. Moreover, understanding differed between regions especially related to drivers and pressures that influence soil biodiversity.

Chapters 4 and 5 study public preferences for soil biodiversity related forest management. While chapter 4 focuses on a comparison of public preferences between three distinct European regions (Finland, Flanders and Italy), chapter 5 assesses the effect of information on public preferences in Flanders and heterogeneity in this effect. Both similarities and differences in preferences were found between the regions, depending on the forest management practice. These differences could generally be linked to the context of the regions. Moreover, information about soil biodiversity significantly increased preferences for higher shares of old trees and dead wood and tree logging through fixed logging roads that are beneficial for soil biodiversity in all three regions. Nevertheless, additional effects of information were only found in Flanders. The in-depth assessment of the effect of information revealed heterogeneity in preferences for forest management among Flemish citizens before and after information although heterogeneity decreased after information. In particular, two third of the sample focused on aesthetics and recreation before information, while their preferences after information increased most in support of soil biodiversity, in particular related to biodiversity components, tree logging and regulating ecosystem services. Hence information was found to be most effective for this group of respondents that were less familiar with forests and soil biodiversity.

Chapter 6 provides a methodological assessment of certainty calibration to decrease hypothetical bias in choice experiments in order to improve environmental valuation. For this, it relies on a case study in Belgium that assesses consumer preferences and willingness-to-pay for potatoes produced using sustainable soil management practices. Overall, effects of certainty calibration on model fit and willingness-to-pay were relatively limited, depending on the methodology used. Simultaneously estimating choice and choice certainty, as well as taking into account variability in choice certainty within respondents resulted in (small) improvements in model fit and lower willingness-to-pay values, indicating a decrease in hypothetical bias.

Chapter 7 concludes this dissertation with an overview of the main findings and a discussion of limitations, future research and implications for policy and science.

Overall, the results suggest potential for (in-depth) information to increase awareness and attention for soils and support soils both amongst forest managers and citizens. Nevertheless, information should be considered a first step and is likely insufficient to realize a strong and durable behavioral shift towards sustainable soil-related behavior. Moreover, context was found to significantly influence managers’ understanding, public preferences and the effect of information on public preferences. Hence, policy design and the development of informational tools should consider such contextual factors to be effective.