Abstract
Reintroduction guidelines recommend that 'adequate' numbers of individuals be released to minimize loss of genetic diversity, but these numbers are rarely quantified. We present a framework for assessing the number of individuals required for an island reintroduction that takes account of allele loss both during the founding event and in the following establishment phase with a low population size. This is the first attempt to model release numbers for reintroductions in order to preserve alleles with a specified initial frequency, while taking post-release mortality rates, population growth rates and site carrying capacity into account. Probability of allele retention was sensitive to both release number and post-release demography. The rate of allele loss was strongly influenced by both environmental stochasticity and delayed population growth but was little affected by increasing the annual turnover rate of the breeding population. We illustrate the model's application using parameter estimates from a threatened New Zealand passerine, the mohua Mohoua ochrocephala, for which reintroduction is a common management tool. Our modelling indicates that when population growth is moderate (lambda = 1.3), c. 60 individuals would need to be released to achieve at least 95% certainty that alleles at an initial frequency of 0.05 would be retained after 20 years (five overlapping generations), which is double the number typically released in translocations of mohua and other threatened forest passerines in New Zealand.