Three essays on advancing modelling of sustainable rural land use change decision making in Aotearoa New Zealand
(Lincoln University, 2025) Wang, Le
Understanding the processes driving the development of niche industries is crucial for achieving sustainable agricultural transitions. In Aotearoa New Zealand (NZ), niche agricultural products, primarily from horticultural industries, have long existed but have struggled to evolve into mainstream agricultural systems. The total land area dedicated to these alternative systems remains small compared to pastoral farming, such as dairy and sheep/beef farming. Since the 1980s, the expansion of dairy farming and a large reduction in sheep number have dominated agricultural land use change in Aotearoa NZ. However, the last decade has witnessed a shift in some of the most intensively farmed dairy regions, such as Canterbury, Marlborough, and Southland, where dairy farming has declined in total area while horticultural land use has grown. These divergent trends provide an opportunity to investigate the dynamics of rural land use change and the potential for transitioning toward more sustainable agricultural systems. This thesis explores the drivers, dynamics, and patterns of rural land use change in Aotearoa NZ through three interconnected essays, each addressing critical aspects of the transition process. Essay One employs machine learning methods, particularly artificial neural networks (ANNs), to identify and evaluate the key drivers of sustainable land use transitions. Using data from the 2021 Rural Decision Makers Survey, this essay highlights the importance of farmer and farm characteristics, values, and demographics in influencing land use decisions. Essay Two develops an opinion dynamics model to examine how farmers’ attitudes towards land use change evolve within agricultural communities. The model captures the role of social networks and peer influence in shaping decision-making processes, offering insights into mechanisms of consensus or polarization among farmers. Essay Three integrates the methodologies and findings from the first two essays into an agent-based modelling (ABM) framework. This essay simulates rural land use transitions in Canterbury over a 25-year period, considering farm-level heterogeneity, social interactions, and real-world geographical characteristics. The results reveal key patterns of land use change, including a gradual shift from sheep and beef farming to horticulture, while dairy farming shows limited transition.
Collectively, the three essays provide a comprehensive framework for understanding and modelling sustainable land use change from the perspective of decision-makers. This research contributes to the literature on agricultural land use by combining machine learning, social network analysis, and agent-based modelling to offer robust evidence and actionable insights. The findings highlight opportunities for fostering sustainable land use transitions and inform policies and interventions to support alternative land use systems in Aotearoa NZ.