This chapter focuses on the establishment of a long-term ecological research experiment to understand the restoration of box-gum grassy woodlands through key interventions, the reintroduction of locally extinct species, particularly ecosystem engineers like the eastern bettong. A predator-proof fence built around Mulligans Flat and Goorooyarroo Nature Reserves enabled the safe reintroduction of native species and exclusion of invasive predators. Key objectives included evaluating the ecosystem impacts of treatments, reintroducing species at different trophic levels, such as the eastern bettong, and understanding the broader dynamics of the ecosystem through long-term monitoring of various response variables, including birds, mammals and invertebrates. The project also explored the concept of whole ecosystem restoration, where the reintroduction of species, coupled with fire and grazing management, was aimed at restoring ecological processes. The research demonstrated the potential of multi-species, sequenced reintroductions as a strategy for ecosystem recovery. Ultimately, this work contributes to broader ecological restoration practices, offering insights into species interactions, trophic impacts, and long-term ecosystem sustainability in Australian woodlands.

This chapter presents two companion large-scale quasi-experiments in ecological restoration in southeastern Australia. These experiments incorporate key elements of rigorous experimental design – randomisation, replication and control – making them exemplary models in landscape ecology. They offer a strong inferential framework for evaluating the effects of restoration treatments on woodland biodiversity and enable the identification of causal relationships between experimental factors.

The design includes treatments arranged factorially (kangaroo grazing, burning, logs dispersed and logs clumped) across three vegetation classes, structured spatially. To improve estimation of main effects, selected higher-order interactions are partially confounded with random block terms. The chapter includes example analyses for both the original and modified designs.

Given the rarity of long-term quasi-experiments in restoration, this work serves as a valuable model for integrating ecological research with practical restoration. Baseline data supports accurate tracking of ecological change, and the thoughtful design ensures all treatment effects can be reliably estimated.