An agent-based model was developed to simulate the growth rate, age structure, and social system of the endangered mountain gorillas (Gorilla beringei beringei) in the Virunga Volcanoes region. The model was used to compare two types of data: 1) estimates of the overall population size, age structure, and social structure, as measured by six censuses of the entire region that were conducted in 1971-2000; and 2) information about birth rates, mortality rates, dispersal patterns, and other life history events, as measured from three to five habituated research groups since 1967. On the basis of the research-group data, the "base simulation" predicted a higher growth rate than that observed from the census data (3% vs. 1%). This was as expected, because the research groups have indeed grown faster than the overall population. Additional simulations suggested that the research groups primarily have a lower mortality rate, rather than higher birth rates, compared to the overall population. Predictions from the base simulation generally fell within the range of census values for the average group size, the percentage of multimale groups, and the distribution of females among groups. However, other discrepancies predicted from the research-group data were a higher percentage of adult males than observed, an overestimation of the number of multimale groups with more than two silverbacks, and an overestimated number of groups with only two or three members. Possible causes for such discrepancies include inaccuracies in the census techniques used, and/or limitations with the long-term demographic data set obtained from only a few research groups of a long-lived species. In particular, estimates of mortality and male dispersal obtained from the research groups may not be representative of the entire population. Our final simulation addressed these discrepancies, and provided a better basis for further studies on the complex relationships among individual life history events, group composition, population age structure, and growth rate patterns.
Copyright 2004 Wiley-Liss, Inc.