Small mammal and reptile occupancy across unmined and restored forest in southwestern Australia

Background/Question/Methods

The jarrah (Eucalyptus marginata) forest in southwestern Australia lies within one of the world’s 25 biodiversity hotspots. Due to this diversity there is limited knowledge about the habitat requirements of much of its native fauna. Additionally, the jarrah forest has been drastically changed by agriculture and mining, which both involve removal of the forest. Currently, mining companies apply restoration techniques intended to return the forest to a fully-functioning ecosystem. For example, Alcoa World Alumina Australia mines and restores forest patches totaling 600 ha annually creating a mosaic of seral and mature forest. Many native fauna species have been documented in restored forest, but if these areas do not meet habitat requirements, species may traverse, but not permanently occupy, restored forests. We determined attributes important for occupancy of native reptile and small mammal species through occupancy modelling using 4 years of detection histories from 20 trapping grids in unmined forest around Alcoa’s Huntly mine. We completed additional trapping surveys in 2011/2012 to validate and supplement generated models. We used this information, along with data collected on restored forest vegetation structure to predict species occupancy and the potential success of restoration over time.

Results/Conclusions

Occupancy models indicated that reptile and small mammal species’ detection probabilities were often influenced by weather covariates such as temperature, solar radiation, or survey month. These results imply that survey times greatly impact species detection and provide a guide of when, and under what conditions, to survey for specific species. However, the best models of species’ occupancy often included null models and only 2 reptile species’ occupancies were influenced by covariates (Hemiergis initialis; bare ground and Menetia greyii; 0-1 meter cover). Our recent surveys (2011/2012) were similar to older surveys in regards to species present and covariates influencing detection probability and they provide additional data for assessing occupancy. Currently, restored sites are missing slow-forming features (tree hollows, fallen trees) that develop over time, however the development of these features along with present structure, (e.g., bare ground, canopy cover, etc.) may be critical and can be directly influenced by modifying restoration processes. Determining which species are present in the restoration as it matures is essential because mining companies aim to return restored forest back to its original state. Our study supports the idea that restoration practices may need to be modified to accelerate fauna species return and increase restoration success.