Biocontrol in an impulsive predator–prey model

• We assume that predators are periodically released to prey on a pest population.
• This gives rise to an impulsive predator–prey model for biological pest control.
• The model possesses a pest-eradication solution.
• We find local and global stability conditions for the pest-eradication solution.
• We corroborate parts of our analysis by numerical simulations.

We study a model for biological pest control (or “biocontrol”) in which a pest population is controlled by a program of periodic releases of a fixed yield of predators that prey on the pest. Releases are represented as impulsive increases in the predator population. Between releases, predator-pest dynamics evolve according to a predator–prey model with some fairly general properties: the pest population grows logistically in the absence of predation; the predator functional response is either of Beddington–DeAngelis type or Holling type II; the predator per capita birth rate is bounded above by a constant multiple of the predator functional response; and the predator per capita death rate is allowed to be decreasing in the predator functional response and increasing in the predator population, though the special case in which it is constant is permitted too. We prove that, when the predator functional response is of Beddington–DeAngelis type and the predators are not sufficiently voracious, then the biocontrol program will fail to reduce the pest population below a particular economic threshold, regardless of the frequency or yield of the releases. We prove also that our model possesses a pest-eradication solution, which is both locally and globally stable provided that predators are sufficiently voracious and that releases occur sufficiently often. We establish, curiously, that the pest-eradication solution can be locally stable whilst not being globally stable, the upshot of which is that, if we delay a biocontrol response to a new pest invasion, then this can change the outcome of the response from pest eradication to pest persistence. Finally, we state a number of specific examples for our model, and, for one of these examples, we corroborate parts of our analysis by numerical simulations.