Abstract
Long-term studies of ecological restoration, within a designed randomized experimental framework, are uncommon; however, such projects provide hitherto under-utilized opportunities to inform both evidence-based management planning and action, and ecological theory. Baseline data collected prior to the application of treatments allows accurate estimation of changes taking place on the experimental units, and random allocation of treatments ensures that relations between causes and effects can be established. This is critical to effective active adaptive management. In this paper, we outline the establishment phase of a new long-term ecological restoration experiment in south-eastern Australia, that will test ways of improving critically endangered box gum grassy woodlands for biodiversity. In the experimental design, treatments include the addition of 2000 tonnes of coarse woody debris, exclusion of kangaroos and fire. Random variation in biophysical variables occurs at several levels. To facilitate accurate estimation of key main effects, selected high order interactions are partially confounded with 'random' block terms. Response variables include: plants, birds, small mammals, reptiles and invertebrates. Analysis of baseline data across selected response variables confirmed no pre-treatment effects. The experiment provides a strong inferential framework for tracking the effects of restoration treatments on woodland biodiversity over coming years. It also provides a model for other similar experiments that integrate restoration and research. A newly constructed feral animal-proof fence, that will allow reintroduction of locally extinct species, including ecosystem engineers, will provide additional opportunities to research the woodland restoration process. This experiment will become a long-term ecological research site, and an 'outdoor laboratory' for ecological restoration research, and community and student learning.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have