The behavioral ecology and economics of variation, risk, and diversification among Mikea forager-farmers of Madagascar

The residents of the dry Mikea Forest of southwestern Madagascar are often incorrectly stereotyped as primitive foragers, when in fact they constitute rural households making a living in a highly variable (risky) environment by combining foraging, fishing, farming, herding retailing and wage earning activities into a diversified household "portfolio." Drawing on data and observations acquired during 22 months of ethnographic fieldwork in five Mikea communities during 1996--1999 and theory from behavioral ecology and agricultural economics, I examine some of the decisions involved in managing a risk-sensitive household portfolio. My objective is to create general explanations of human subsistence behavior that can be applied to the anthropological issue of subsistence transitions, from foraging to agriculture (or vice versa) or from non-market to market exchange. First, I find that Mikea households often chose activities that covary differently with rainfall, such as combining rain-loving maize with drought-hardy manioc. Second, I find that in the absence of large game and predators, all age/sex groups practice the same foraging activities, and everyone older than a toddler can harvest wild tubers at similar rates of proficiency. The productivity of children reduces their cost to households. Third, unlike most other hunter-gatherer groups in the world, Mikea practice very little food sharing. This may be because plant resources are abundant and predictable while wild animal resources are small, difficult to capture, and easily stored, and thus, giving away meat would incur unacceptable marginal losses to the donor (from the perspective of sharing as "tolerated theft"). Fourth, by borrowing the "missing markets" model from agricultural economics, I find that transaction costs, the costs of processing and transporting goods for sale, account for much of the variation in market behavior among households. Finally, using future discounting models I suggest that Mikea tend to under-invest in agriculture because the immediate rewards of foraging for wild tubers are often greater than the perceived future value of risky, low-yielding maize and manioc harvests. I conclude that a general model of household subsistence should parameterize (in terms of utility and fitness) household size and composition; production variance, covariance, and delay; and transaction costs.