DISCUSSION

Studies of local adaptation are severely lacking for rainforest tree species that are key to restoration efforts. While meta-analyses are useful to discern drivers of local adaptation in studies across multiple species (Leimu & Fischer 2008, Hereford 2009, Matesanz & Ramirez-Valiente 2019), these studies often vary in methodology, site maintenance and conditions that are not easily accounted for. In this study, we used three common garden sites to investigate the effects of site, provenance, and their interaction on the growth and survival of 16 tree species commonly used in restoration plantings in the Australian Wet Tropics. Site effects were observed in 50% of species for survival, 93% for height increment, and 79% for inherent growth rate. Provenance effects were noted in 38% of species for survival, 36% for height increment, and 57% for inherent growth rate. Interestingly, the lowland provenance generally outperformed the upland provenance regardless of site. We found a negative correlation between species wood density and the magnitude and direction of provenance differentiation (mean sympatric/allopatric comparisons of local adaptation) in both survival and height increment (but not inherent growth rate). This suggests that local adaptation may be more evident in faster-growing species, with no differences or potential maladaptation for slower-growing species, but further monitoring is needed to determine if these preliminary patterns hold. These results challenge the usefulness of a ‘local is best’ provenancing strategy for upland restoration plantings in the Australian Wet Tropics and help to fill the gap in our understanding of the extent of local adaptation in tropical rainforest trees globally.