Application of expert elicitation to estimate population trajectories for species prioritized in Australia's first threatened species strategy

Abstract

Many global and national conservation initiatives have objectives relating to reducing or reversing the rate of biodiversity decline but few have specific targets. In 2015, the Australian Government implemented a Threatened Species Strategy that aimed to improve the population trajectory of a set of 71 species (20 mammals, 21 birds and 30 plants) by 2020, relative to the trajectory before 2015. To assess the extent to which this objective had been achieved, and because there was no or limited monitoring for many of the species, we used structured expert elicitation to estimate population size for every species at 2005, 2015, 2020, and predict population size at 2025, 2035 and 2045. Elicitation was informed by a consistent collation of information for every species, including on threats, management actions, and available monitoring. Experts estimated that the population trajectories of 15 species had improved significantly (at p < 0.05, or 24 species at p < 0.1) over the period 2015–20 (the period coinciding with priority attention under the Threatened Species Strategy) relative to 2005–15, and 13 species had deteriorated significantly (at p < 0.05, or 15 at p < 0.1). The lack of recovery for a substantial component of the set of priority species was likely due to the short (5-year) period considered over which deeply ingrained threats may not be ameliorated, the impacts of drought and wildfire in this period, and the relative ineffectiveness of some management actions to mitigate the most important threats. Assessment of population trajectories was least confident and most inconsistent among experts for species with least monitoring effort. Experts predicted ongoing future declines with cessation of conservation management actions, but some longer-term recovery for most plant and mammal species with ongoing conservation management. Elicitation helped provide a consistent approach towards inferring population trends, but more ideal would be to monitor populations for all threatened species.