Background: Food systems are deeply complex, shaped by interconnected economic, social, and environmental forces, with implications for local economies and community health. Restaurants can serve as increasingly crucial intermediaries between consumers and locally grown, healthy foods. Yet, restaurant adoption of locally sourced, nutritious ingredients, while increasing among higher price point establishments, remains stagnant at more economically accessible levels. This challenge stems from a dynamic interplay of barriers and facilitators at multiple levels.

Addressing these complexities requires a systemic perspective that moves beyond linear cause- and-effect models to explore feedback loops and leverage points for change.

Methods: This study applies a systems dynamics approach to examine these interactions, using group model building (GMB) and human-centered design (HCD) to identify key driversand solutions. Specifically, the research aimed to develop causal loop diagram (CLD) to visualize feedback mechanisms shaping the provision of healthy, locally grown foods inrestaurants and map community-generated solutions as potential intervention points to facilitate sustainable change. The study follows an iterative, participatory approach, drawing on qualitative data from restaurant owners, chefs, local food producers, and food system organizations.

Results: Through stakeholder engagement and GMB workshops, participants collaboratively identified reinforcing and balancing feedback loops influencing restaurant decision-making. Key emergent themes include the impact of perceived customer demand, economic pressures, operational burden, costs associated with local food procurement, and supply chain coordination on restaurant participation in local food sourcing. The CLD highlights self-reinforcing barriers, such as the hesitation of restaurants to invest in local sourcing due to uncertainty about consumer interest, which in turn limits consumer exposure and demand. Conversely, potential leverage points include economic incentives, consumer educational campaigns, local food labeling regulation, recognition programs, and service provision.

Conclusions: By applying social innovation and complexity frameworks, this research advances the understanding of how food system actors navigate systemic barriers and opportunities. The findings suggest that successful interventions must address multi-level interdependencies rather than focusing on isolated behavioral change. The CLD serves as both a research tool and a practical resource, providing policymakers and practitioners with a systems-based roadmap for designing interventions that foster resilient, health-promoting food environments.
Contribution: This study contributes to discussions in social innovation and complexity by demonstrating how systems mapping can uncover hidden structures within food systems, informing more adaptive, collaborative solutions. It also underscores the importance of stakeholder-led knowledge production, positioning food system actors not only as implementers but as co-creators of positive change.