Using fossil records to inform reintroduction of the kakapo as a refugee species

Many threatened species persist only as relict populations occupying a fraction of their former distribution, in habitats which may not be optimal for supporting viable populations. Following population growth of one such species, the kakapo (Strigops habroptilus), conservation managers are faced with the challenge of identifying suitable locations for reintroduction. Areas which support habitat conditions typical of those occupied by kakapo in the past have the greatest potential to support future populations. We collated occurrences of kakapo from recent fossil records, then used MaxEnt to model the past distribution of kakapo across New Zealand, and contemporary areas suitable for reintroductions based on extant habitat and present-day climate. We validated our models against three independent data sets of the most recent relict populations. Our models suggest that kakapo once occurred in mountain beech and Hall's totara or broadleaf forests with moderate to high precipitation and milder winters. Areas predicted to be environmentally suitable for kakapo in contemporary New Zealand include the west coast of the South Island, the west and north-east of the North Island and the southern side of Lake Taupo. Assuming that known threats of introduced predators can be managed, our study suggests that suitable kakapo habitat persists in New Zealand, and here we offer insight into locations for future population establishment. Given the finite carrying capacity of offshore islands, this is an important first step which will enable kakapo managers to prioritise focal areas and also highlights the benefits and potential pitfalls of using these modelling approaches for refugee species.