Modeling rangelands as spatially-explicit complex adaptive systems

摘要

Rangelands cover one third of the earth's land area, provide livelihoods for one billion persons, and most have been degraded by overgrazing of domestic livestock. Recent debate about best management practices often has centered on comparison of continuous grazing and rotational grazing. Resolution to this debate may lie in viewing rangelands as complex adaptive systems. We describe a spatially-structured, individual-based model of rangelands that embodies this perspective, and simulate forage dynamics and cattle production under semi-arid rangeland conditions typical of the southern Great Plains of the USA employing both continuous and rotational grazing. Relative "success" of simulated grazing schemes depended primarily on the evaluation metric used (e.g., rangeland ecological condition, sale weight of cattle, secondary production efficiency) and the particular manage scheme employed, and neither continuous nor rotational grazing schemes were uniformly more successful. Our results demonstrate that solution of the grazing systems debate is unlikely to be found in a single group of grazing schemes, but, rather, in adaptive management of feedbacks among system components. The present work provides an example of how modeling rangelands as complex adaptive systems can aid in the evaluation of management schemes.