A stochastic model for estimating human carrying capacity in Brazil's Transamazon Highway colonization area

Abstract

Human carrying capacity for tropical agricultural populations can be estimated with a computer simulation of the agroecosystem. A stochastic model is developed for estimating carrying capacity in one of the government-directed small farmer settlement projects along Brazil's Transamazon Highway. Carrying capacity is operationally defined in terms of an empirically computed relationship between population density and probability of colonist failure with respect to various criteria. When high population density leads failure probability to exceed a maximum acceptable level, population can be considered to be above carrying capacity. Colonist failure probabilities are those that are sustainable over a long period of simulated years. High variability in crop yields appears to have a strong effect on failure probability based on comparison of deterministic and stochastic runs of the simulations. Failure probabilities are raised by variability at low population densities, but are lowered at extremely high densities where most colonists would fail in an "average" year. Effects can be tested for colonists with different backgrounds or with differing agricultural practices such as fallow times. Failure probabilities in standard runs are higher than those considered acceptable to government planners at all population densities simulated in the present study's stochastic runs (lowest density 24 persons/km2), thus lending support to the informal impression of many that the carrying capacity of most of Amazonia's uplands is low. © 1985 Plenum Publishing Corporation.