Foraging pit location provides valuable insights into critical habitat requirements of soil engineers.

Context: Successful conservation requires a detailed understanding of critical habitats required to support different plant and animal populations. This is particularly important for rewilding programs where locally extinct species are often introduced into novel or reintroduced into areas that have been dramatically altered since they were extirpated. Objectives: Here, we explored changes in the volume of foraging pits of three engineers under different vegetation patch types and different landforms in an arid woodland in Australia based on field experiment. Methods: In this study, we examined the foraging habitat of three ecosystem engineers in the arid eastern Australia: a reintroduced mammal, the greater Bilby (Macrotis lagotis), the Small-beaked echidna (Tachyglossus aculeatus) and a varanid, Gould's sand goanna (Varanus gouldii). The soil disturbing activities of these three animals has been shown to lead to marked and persistent changes in a range of critical ecosystem functions and services. We tracked the density, size and habitat location of 4102 foraging pits over 3 years, as the measure of habitat favourability and the time taken for these foraging pits to infill (pit longevity). Results: We found that foraging was non-random, and animals tended to forage more within shrub patches when accounting for the relative cover of shrubs. There were also some differences among different habitat patch types, with generally larger echidna foraging pits beneath shrubs and trees than in the open interspaces. Pit longevity also varied among animal and patch types, with no differences for bilby or goanna pits across all patches, but greater longevity of echidna foraging pits beneath shrubs. Larger pits tended to persist for longer, and those beneath trees were closer together. Conclusion: Our work shows that foraging, and therefore the ecological effects of these animals vary with the target engineer and the biophysical features of the rewilded habitat. Land restoration programs that aim to restore degraded habitat and ecosystem functions must be cognisant of the idiosyncratic effects of different reintroduced engineers, and the makeup of different patch elements within the target restoration area. [ABSTRACT FROM AUTHOR]