Dependence of population response to fertility control on the survival of sterile animals and their role in regulation.

The species for which fertility control is presently used, or for which it is being developed, range from small mammal pests, such as the house mouse (Mus domesticus), to large mammals, such as the African elephant (Loxodonta africana). However, the possibility of a population response other than a reduction in abundance proportional to the fraction of animals rendered infertile has been shown in field trials. For example, when intermediate levels of sterility were imposed on wild populations of European rabbits (Oryctolagus cuniculus), there was an increase in their abundance, on an annual basis, due to enhanced survival of juveniles and adult females. In this article, we relate intraspecific regulatory processes to the response of populations to fertility control using a set of density-dependent structured-population models. In each of the models, the population is exposed periodically to a fertility control agent that renders a fraction of fertile females sterile. Although our intention is not to predict the population response of any one particular species, the results of the models are illustrated using parameter values that are representative of populations of the European fox (Vulpes vulpes) in south-eastern Australia. When populations were regulated by density-dependent mechanisms in which sterile females did not participate, such as competition for resources among young animals or competition among fertile females for breeding sites or territories, then populations could increase in abundance for low and intermediate levels of imposed sterility. For other intraspecific regulatory mechanisms, such as competition for resources between all individuals, all levels of sterility were observed to reduce abundance. The population response was sensitive to (i) whether the survival of sterile adults was higher than that of fertile adults, (ii) whether animals could be sterilized before sexual maturity, and (iii) whether density dependence was modelled as a threshold process.