Doctoral Thesis Review – Anmeldelse av doktoravhandling

Abstract

The overarching goal of this thesis is to understand and assess the process of citizen science data collection and to investigate the usefulness and applicability of such data to make inferences about the ecology of species at different scales. The scope of the thesis can generally be placed within the field of ecology, and specifically within biodiversity conservation. While the thesis was written under the supervision of a member of staff at the Department of Geography, NTNU, Benjamin Cretois was also associated with and co-supervised by a member of staff at the Department of Terrestrial Biodiversity, Norwegian Institute for Nature Research (NINA). In his thesis, Benjamin Cretois focuses on the role of hunters in generating citizen science data, as they are one of the main groups of contributors to such data collection practices for the purpose of monitoring and management wildlife. Moreover, the data generated by hunters are geographically and thematically broad, covering a wide range of species ecology characteristics. While there are known biases in citizen science data, Cretois argues that by establishing knowledge about how the data are observed, collected, and reported, it is possible to apply statistical techniques to correct for these inherent, unavoidable biases. This can allow unbiased inferences of the ecological measures of interest. In his thesis, Cretois demonstrates this by drawing novel inferences about species ecology at different spatial scales, ranging from continental to local habitat scale, based on crowdsourced hunters data. The empirical and analytical foundations of the thesis are significantly quantitative. Cretois bases his analyses on bibliometric data, simulations, and unstructured citizen science data from the Norwegian Species Observation Service (Artsdatabanken n.d.). His analytical tools include geographic information systems (GIS), exploratory, descriptive, and prescriptive statistics, in particular spatial statistics, and various data visualizations, including a broad range of tables and map-based figures, as well as plots and other illustrations. However, the main contribution of the thesis lies in the methodology, and in particular demonstrating how Bayesian statistics can be used to fit models that account for the inherently hierarchical and biased nature of the data. In keeping with the Norwegian thesis tradition, Part I of the thesis is an overarching synopsis that first introduces the theoretical basis and empirical background, followed by a description of the overall research design and methodology, a summary of the five articles that comprise the second part of the thesis, and finally some concluding remarks and reflections on future research. Part II comprises the articles on which the thesis is based. Part I clearly puts Cretois’s work in context by introducing the main challenges, including the challenges introduced by the use of citizen science data. Thereafter, it describes the data collection processes and practices that inherently lead to different types of biases. Part I introduces the reader to the rationale behind using a Bayesian framework to tackle the challenges of citizen science data and justifies the choices that were made to narrow down the focus of the thesis to large mammals, and the different measurements and data sources used. However, the way that this part is presented creates some challenges in clearly communicating what the thesis is about. The main aim of the thesis is to contribute to the methodological development of using citizen science ecological data. The choices that have been made in narrowing down the empirical focus of the thesis are justified through pragmatic considerations rather than theoretical argumentation. While this is an acceptable approach, given the overall objective of the thesis, it nonetheless represents a challenge for communication in that it does not rely on clearly defined research questions rooted in a particular ecological theory or phenomenon. Also, the transferability of the work to other contexts may not be easily justified. Therefore, the explicit presentation of research questions, aims, and objectives expected in this part of the thesis is largely missing. The literature review in Part I could have helped to frame the work and justify this approach, but it is too brief to supply such as frame for the thesis. However, it should be mentioned that additional coverage of the relevant literature can be found in the articles in Part II,