Restoration potential of native forests after removal of conifer plantation: A perspective from Australia

•Natural regeneration following harvesting of Pinus radiata established on abandoned farmland.•Plant functional traits shifted towards those of native forest with time since harvesting.•Myrmecochorous shrubs returned, Ericoid heaths failed to recolonise, introduced grasses persisted.•Planting native overstorey seedlings did not promote return of native vegetation.•Additional measures required to see the return of a fuller compliment of native species. This study investigates whether undertaking a rotation of pine plantation on abandoned farmland facilitates the return of native species. Plant functional traits were used as a means of assessing the effects of land-use change on vegetation. We explored the relationships among plant traits, time since harvesting and environmental variables for the vegetation found in Pinus radiata plantations after clear-felling. Vegetation was monitored 1–3years following harvesting and repeated eight years later. We compared species richness, composition and traits to those in reference native forest and abandoned farmland. There was a significant temporal shift in attribute associations of harvested plantations towards those in native forest and a move away from plant attributes common to the pre-plantation use of abandoned farmland. Cluster analysis produced ten emergent groups of plant functional traits for 144 plant species. After harvesting, attribute associations were typical of therophytes and included high specific leaf area (SLA), small stature, arbuscular or non-mycorrhizal associations and small, wind-dispersed seed. These were predominantly introduced species that were dominant in abandoned farmland. The repeat survey identified an obvious decrease in therophytes and an increase in species with low SLA, ant dispersal, large seed and ectomycorrhizal associations. This was largely due to an increase in native Myrmecochorous shrubs. Similar attributes were also shared by Ericoid heaths but these species failed to recolonise harvested sites. Relationships among plant traits of species and environmental variables changed over time with a greater proportion of the variance in attribute associations explained by prevailing site conditions (e.g. type of ground cover, edaphic properties) with increasing time since harvesting. There were no relationships among plant attribute associations and stem density or basal area of planted overstorey species suggesting that planting of tubestock seedlings of native overstorey speci