| Abstract: | Birth and death simulation, developed by Pielou, is a form of Markov stochastic process for describing the time evolution of populations. Applied to modelling the human element of a fishery, it expresses two features of fishing effort dynamics absent in systems of differential equations: (1) discreteness of events, such as a fishing trip or entry of a vessel into the fleet, and (2) demography stochasticity, expressed as randomness in the time occurrence of successive events. Birth and death simulation is based on randomly selecting the waiting time between events from a negative exponential distribution, derived under the assumption of Markov. Histograms from commercial landings data of waiting times between events of boats returning to port in a Nova Scotia fishery yielded good agreement with the predicted negative exponential. Algorithms are presented for stochastically modelling two processes: (1) catch and (2) the open-access hypothesis for changes in fleet size in response to changing levels of profit. The solutions qualitatively diverge from that predicted by differential equations: As the numbers of vessels and fish schools decline (i.e., as the system size scale shrinks), a birth and death formulation predicts increasing instability of the predator-prey cycle solution about the deterministically stable open-access equilibrium. Open-access models are a form of predator-prey model. In choosing the minimum wilderness preserve area needed to sustain a population of top predators, numbered in the low hundreds, a predator-prey model formulated with differential equations could underestimate instability and thus the risk of extinction, when the discreteness and randomness of predator-prey birth, death, and capture events is significant. |
