The practices related to the growing, harvesting, preparation, and celebration of food over millennia have given rise to diverse biocultural landscapes the world over. These landscapes – rich in biological and cultural diversity – are often characterised by persistent poverty, and, as such, are often the target of development interventions. Yet a lack of understanding of the interdependencies between human well-being, nature, and culture in these landscapes means that such interventions are often unsuccessful - and can even have adverse effects, exacerbating the poverty they were designed to address. This thesis investigates different conceptualisations of persistent poverty in rural biocultural landscapes, the consequences of these conceptualisations, and the ways in which development interventions can benefit from, rather than erode, biocultural diversity.

The thesis first reviews conceptualisations of persistent poverty and specifically, the notion of a poverty trap (Paper I), and examines the consequences of different conceptualisations of traps for efforts to alleviate poverty (Paper II). Paper I argues that the trap concept can be usefully broadened beyond a dominant development economics perspective to incorporate critical interdependencies between humans and nature. Paper II uses multi-dimensional dynamical systems models to show how nature and culture can be impacted by different development interventions, and, in turn, how the degradation of both can undermine the effectiveness of conventional poverty alleviation strategies in certain contexts.

In the second section, the thesis focuses on the effects of, and responses to, trap-like situations and development interventions in a specific context of high biocultural diversity: the Pamir Mountains of Tajikistan. Paper III advances a typology of responses to traps based around the mismatch of desires, abilities and opportunities. Observing daily practice provides a way to study social-ecological relationships as a dynamic process, as practices can embody traditional and tacit knowledge in a holistic way.  Paper IV examines the diverse effects of a development intervention on the coevolution of biocultural landscapes and the ways in which everyday practice – particularly around food – can be a source of both innovation and resilience.

Papers I-IV together combine insights from diverse disciplines and methodologies, from systematic review to dynamic systems thinking and participant observation. Paper V provides a critical analysis of the opportunities and challenges involved in pursuing such an approach in sustainability science, underscoring the need to balance methodological groundedness with epistemological agility.

Overall, the thesis contributes to understanding resilience and development, highlighting the value of viewing their interrelation as a dynamic, coevolving process. From this perspective, development should not be regarded as a normative endpoint to be achieved, but rather as a coevolving process between constantly changing ecological and social contexts. The thesis proposes that resilience can be interpreted as the active and passive filtering of practices via the constant discarding and retention of old and new, social and ecological, and endogenous and exogenous factors. This interpretation deepens understanding of resilience as the capacity to persist, adapt and transform, and ultimately shape new development pathways. The thesis also illustrates how daily practices, such as the growing, harvesting, and preparation of food, offer a powerful heuristic device for understanding this filtering process, and therefore the on-going impact of development interventions in rural landscapes across the world.