Recruitment and dispersal limitations in a chronosequence of secondary forests in tropical Australia

Tropical forests harbor the greatest biodiversity on the planet and play important roles regulating climate and capturing atmospheric carbon. However, they are being transformed by agricultural practices and development projects at unprecedented rates causing a considerable loss of ecosystem services and threatening thousands if not millions of species. Environmental restoration provides an important solution to mitigate the ecological damaged associated with forest loss. As human-altered lands in the tropics are abandoned due to multiple social and economic factors, the restoration of these fallow lands has gained more opportunity and relevance. This thesis aims to explore factors that affect forest recovery in a forest chronosequence in tropical Australia. Using data gathered in the field, together with greenhouse and field experiments, I examined the effects of secondary forest age on plant communities in the soil seed bank, and the seedling and adult plant cohorts. I further assessed seed germination and seedling survival and growth in direct seeding and seedling experiments. The findings provide a deeper understanding of the factors influencing forest regeneration in the Australian tropics. I reviewed restoration methods in different ecosystems by evaluating experiments into the use and relative success of direct seeding and planting of seedlings. Overall, I found that direct seeding experiments use more species than seedling studies, yet have lower survivorship. Species availability is a major barrier to the selection of species, regardless of the technique used, and is an important bias in the experiments. Although costs are extremely important when planning a restoration project, only a few studies report on this aspect. Interestingly, none of the studies addressed the impacts of climate change on restoration efforts or provided information regarding how studies should consider future shifts in the environment. Assessing the species composition and the functional traits of plant communities in secondary forests of different ages can give us insight into the recovery of these forests. I found that understorey communities in young secondary forest had lower species diversity and functional traits associated with early stages of forest succession (e.g. small-seeded species, herbs, grasses). Soil seed banks in these young forests were dominated by exotic species and existing grasses, herbs and small trees. As succession unfolds, soil seed banks, understorey and adult plant communities in these forests showed higher species diversity and a wider range of functional traits (e.g. more trees and late successional species). However, even old secondary forests had low numbers of large-seeded (< 20mm) species, a key distinguishing feature of tropical forests. This finding suggests that these secondary forests are limited by either a lack of dispersal or recruitment of these important tropical forest species. To further explore the mechanisms of recruitment and dispersal limitations, I undertook direct seeding and planting of seedlings experiments. Using species of different seed sizes and successional status, I found that species identity was more important than forest age for growth and survival in seeds and seedlings. There was no relationship between seed size and/or successional status on growth or survival of the different species. Seeds had a much lower rate of survival compared to seedlings. As all species germinated irrespective of forest age (and the variation in site conditions), I concluded that dispersal limitation is possibly a greater barrier to the recovery of secondary forests at our sites than recruitment limitations experienced during this study. Increasing awareness of the repercussions of forest lost and its implications on human well-being and ecosystems health make forest restoration a key area for current and future research, policy-making and environmental management. Results from this thesis show a delayed recovery in tropical secondary forests and the absence of key functional plant groups (e.g. large-seeded trees) even in 40 year-old forests. This finding makes restoration efforts necessary if we want to preserve local biodiversity and enhance the conservation value of these forests. Restoration efforts seeking to accelerate biodiversity recovery and the protection of the local flora should include species limited by seed dispersal. When possible, these initiatives should include multiple species and various seed sizes. Further, the inclusion of species with higher tolerance to drier conditions could improve restoration outcomes and prepare the region for future climatic scenarios. As restoration gains importance, it also creates a unique opportunity to transform the way we see ecosystems and the services they provide. Restoration approaches should include realistic goals that match the local communities' needs and expectations. It is important to take into account changes in climatic conditions and prioritize actions in a crowed world. Multidisciplinary efforts that match goals of conservation, policy and human behavior and well-being will also help improve the future of ecosystems restoration.