Statistical models for data on recreational fishing

Recreational fishing is a globally significant activity, often involving angling but also encompassing various other methods in both marine and freshwater environments. While it provides notable health benefits, economic contributions, and conservation support, it can also exert substantial pressures on fish populations and ecosystems. To effectively manage and sustain this activity, it is crucial to quantify the scale, benefits, and impacts of recreational fishing. Traditional methods for monitoring, such as onsite surveys and recall surveys, face challenges in terms of cost, time, and data reliability, especially in accounting for the growing sector of angling tourism. Recent advancements in technology have introduced alternative data collection methods, particularly through smartphone applications. These apps, like Fishbrain, allow anglers to record their catches, offering a vast and continuous stream of data that could significantly enhance recreational fisheries management. However, to harness these data effectively, it is essential to develop robust statistical models to understand their strengths and limitations.

This thesis aims to advance the understanding and management of recreational fishing through the development and application of statistical models using data from traditional face-to-face surveys and app-based records.

Chapter 2 focuses on modelling marine recreational fishing data from a 2012-2013 UK survey, employing zero-inflated Poisson models with shrinkage methods to identify key predictors for catch rates. This chapter introduces a grid-based search algorithm for determining shrinkage penalties, and considers data on a number of key UK species.

Chapter 3 considers Fishbrain data from 2018 to 2021 to study the spatiotemporal patterns of recorded catches for four key marine species in the UK and Ireland. This analysis uses integrated Laplace approximation methods to develop models that provide a framework for visualising large-scale catch data, comparing different error structures, and producing predictive maps for each species.

Chapter 4 extends the analysis to a global scale, examining angling tourism patterns using Fishbrain data. This chapter utilises network models and clustering methods to analyse over 100,000 catches, focusing on international travel and its changes in response to the COVID-19 pandemic. It identifies communities of countries with strong angling connections and explores the implications of these patterns for fisheries management.

Finally, Chapter 5 discusses the broader implications of the research and suggests future directions. The findings underscore the potential of app-based data to complement traditional survey methods, offering a rich resource for sustainable recreational fisheries management. This work contributes to the evidence base needed for informed marine policy and highlights the importance of integrating diverse data sources to support productive and sustainable fisheries.