Understanding the social and ecological contexts of past agricultural systems in complex societies requires expansive and nuanced data sets that recognize the role of all players in the production system. Such data sets are not often available and thus, there is a tendency to generalize across polities and ecosystems and to homogenize place- and time-specific variation. Here, we bring together ethnohistoric, ethnographic, and archaeological data to explore Mā‘ohi commoner and elite involvement in the production systems of the Society Islands at the time of European contact (AD 1767). We focus our analysis on the archaeological records of five polities located in four different watersheds on the islands of Mo‘orea and Ra‘iātea. We divide the polities into those that are elite- vs. commoner-centric and those that are located in productive versus marginal agricultural landscapes. We find that elites have a greater presence and closer association with agricultural production in productive ecological settings than in the more marginal ones. Although the archaeological expression of the agricultural systems look superficially the same in all contexts, maintaining productivity in the marginal contexts would have required different knowledge and more effort on the part of the Mā‘ohi farmer than in the more productive settings. In contrast to previous summaries of Mā‘ohi agriculture that focus on elite control and seasonal shortages, we highlight the place-based knowledge of Mā‘ohi commoners that was foundational to the centuries-old production systems that provisioned both the elite and commoners alike.