Meeting Abstract
Responses of hierarchically organized biological systems (e.g., insect colonies, nervous systems) to stimuli often depend on the recruitment of individual subunits (e.g., insects, neurons) to specified tasks. We developed a Markov chain based model of the rate and extent of recruitment of individuals to tasks that require iterative stimulation to activate individuals to a task. The model posits that activation depends on four biological parameters: the frequency of individual stimulations, the probability of ‘forgetting’ previous stimuli, the cumulative number of stimuli that must be accumulated before an individual is activated to a task, and the rate of deactivation once activated. Predicted patterns of activation agree with anecdotal data and predictions of earlier models. Further predictions include: 1) All four parameters strongly affect activation under some combinations of the other parameters. 2) Different patterns of activation (rapid, slow, most vs few individuals activated) can each be achieved via several, alternate combinations of parameters. 3) The ‘size’ of the system (number of individual units) does not affect the expected rate of activation, but does affect the stochastic variation around that expectation in individual simulations. 4) Short term dynamics of the system can include deterministic oscillations in the level of activation. Biological costs (e.g., time and energy costs) may constrain the parameters of iterative stimulation for a given system.
