Long-term demographic responses to historical logging and ongoing climate change in an upland subtropical rainforest

Subtropical rainforests in Australia are key refugia for Gondwana biodiversity. Although much of this forest is now protected, historically many areas were selectively logged, leaving a legacy that still poses challenges to land managers, especially considering the uncertainty of ongoing climate change. We analysed a rare long-term dataset from upland rainforests in southeast Queensland, comprising logged and intact plots established in 1962 and censused ten times between 1964 and 2015. We used generalised linear mixed-effects models to examine recruitment, survival, and growth of tree species in relation to logging treatment, climate variables, species functional traits, and stem diameter. Recruitment was higher in the logged plot, favouring species with pioneer traits (particularly low wood density). Recruits in the logged plot also contributed more to the plot basal area than those in the intact plot. Survival declined with increasing minimum temperatures, especially in the logged plot, which also suffered damage during an extreme storm event. Diameter growth was faster in the logged plot but declined over time as basal area increased. In the intact plot, growth was concentrated in the largest trees positioned to intercept cloud water, while in the logged plot growth was distributed more widely across stem sizes. Even after more than 50 years, the logged plot has not fully recovered in forest structure and composition and shows greater sensitivity to warming and extreme weather. Nevertheless, recovery is progressing slowly.