U.S. Department of Agriculture: Animal and Plant Health Inspection Service


Date of this Version

August 2003


Published by Ecological Modelling , Volume 166, Issues 1-2 , 1 August 2003, Pages 109-121.


The management of canid populations has been at the forefront of wildlife management worldwide for much of the last century. Effective management depends on the ability to integrate species biology, the environmental aspects upon which those populations depend, and the factors controlling species abundance. Further, managing canid populations requires consideration of territoriality and dominance, which may have a significant effect on population dynamics. To better understand the effect of social structure on canid populations, we developed an individual-based computer model using Swarm to mimic natural coyote population dynamics. We selected the Swarm simulation environment because it is ideally suited for creating a system of multiple interacting agents with variable schedules and hierarchies. Swarm was a software platform that allows the user to describe generic individuals and behaviours, link those behaviours in each concurrent time step, and assemble behaviours and objects in a hierarchical framework. This model stands apart from previous modelling efforts because it explicitly incorporates behavioral features, such as dominance and territoriality, as major determinates of species demography into a simple model. Individual variation, such as status within territorial social groups and age-based reproduction are incorporated, but assumptions typically associated with most demographic models are not needed. The simple population model with few parameters not only closely resembled ‘real world’ populations but also helped us understand population dynamics that emerged from model. The sensitivity analysis revealed that the model was largely insensitive to individual parameter estimates and could be used to guide management of territorial animal populations with social structure. The model output variables closely matched the mean and range of values reported in the literature of wild populations for population size, proportion of females breeding, offspring survival and litter size. The variation of model output was similar to the variation recorded in field studies. Further, population dynamics reported from field studies emerged from the model and may help to explain the mechanisms responsible for this variation. This type of model could also provide insights into potential management alternatives for other canid species or other species with similar social structure.