Down but not out: long-term monitoring reveals possum population changes and associations in the southern jarrah forest

Context Long-term monitoring is critical for detecting population trends and informing conservation, yet such programs are rare in Australian forests. The critically endangered ngwayir (western ringtail possum, Pseudocheirus occidentalis) and the declining koomal (common brushtail possum, Trichosurus vulpecula hypoleucus) are the most prominent arboreal marsupials in the southern jarrah (Eucalyptus marginata) forest, where threats include timber harvesting, fire, predation and climate change. Aims This study aims to quantify long-term changes in possum abundance, assess whether detection probability varies with forest management history, and evaluate whether detection rates provide a reliable index of abundance. Methods We analysed 30 years (1995–2024) of vehicle-based spotlight surveys along three permanent transects in the Greater Kingston area, spanning a range of timber harvest and fire histories. Distance sampling was applied to estimate detection probability and density from 2012. Mixed-effects models tested associations between detection metrics and covariates (harvest type, time since harvest, burn age, season). Detection rates (detections km−1) were used to infer abundance trends across the full dataset. Key results Ngwayir numbers collapsed by >99% between 1999 and 2002, followed by a decade of near absence and only partial recovery to ~11% of pre-decline levels by 2024. Koomal numbers increased markedly during 1995–2005, declined by ~60% over the next decade, and have since stabilised at ~75% of peak levels. Detection probability was largely unaffected by management history, except for weak effects of burn age and shelterwood harvesting on koomal. Median density estimates for koomal and ngwayir between 2012 and 2024 were 0.140 and 0.011 ha−1 respectively. Conclusions Ngwayir remain at critically low densities despite decades of predator control and fire management, whereas koomal exhibit eruptive dynamics consistent with predator-release effects. Drivers of ngwayir decline and limited recovery remain unclear but likely involve interacting threats beyond local habitat disturbance. Implications This study highlights the irreplaceable value of long-term, fit-for-purpose monitoring for diagnosing declines and informing adaptive management. For ngwayir, urgent action is needed to identify and mitigate recovery-limiting factors, including predation and climate stress. More broadly, conservation outcomes depend on integrating rigorous monitoring with proactive, evidence-based management in native forests.