A spatiotemporal model of risk and uncertainty for Hawaiian dryland agriculture and its implications for ahupua'a community formation

Abstract

The Hawaiian Islands, like many other high volcanic islands, are characterized by a diversity of ecozones, which had ramifications for the types of subsistence strategies that developed within each. Although traditional cultivation practices were highly variable, agricultural systems can be roughly split into windward and leeward forms. Leeward agriculture is differentiated from windward agriculture based on the almost complete dependence on rainfall. Dependence on rainfall, an often-unpredictable resource, creates a high level of risk and uncertainty in agricultural yields. Dryland field-systems were highly susceptible to droughts, potentially resulting in food shortages that would have had various societal consequences, such as increased intergroup conflict, community cooperation, or social inequality. The concepts of risk and uncertainty, derived from human behavioral ecology, are useful for exploring how fluctuations in the availability of resources from droughts influenced agriculturalists in the Hawaiian Islands. Using the Rainfall Atlas of Hawai‘i, a newly published rainfall archive, we investigate spatiotemporal rainfall patterns in the Leeward Kohala Field System (LKFS) on Hawai‘i Island. We employ geostatistical modeling techniques, time-series analysis, and a simulation model to quantify the intensity, frequency, and periodicity of droughts in the LKFS. Our results support previous studies and suggest a high degree of agricultural risk, particularly from ca. AD 1450–1600, with implications for Hawaiian agriculture and emerging sociocultural patterns.