COS 187-8 - Optimal fire histories for conserving biodiversity

Friday, August 10, 2012: 10:30 AM
E144, Oregon Convention Center
Luke T. Kelly, School of Botany, The University of Melbourne, Parkville, Australia, Andrew F. Bennett, School of Life and Environmental Sciences, Deakin University, Burwood, Australia, Michael F. Clarke, Department of Zoology, La Trobe University, Bundoora, Australia and Michael A. McCarthy, School of BioSciences, The University of Melbourne, Australia

Fire is an important driver of ecosystem structure and function worldwide. It is also widely used as a tool to achieve conservation goals. A common management objective is to maintain a diversity of post-fire successional stages. However, in many cases, it is unclear what distribution of fire age-classes enhances conservation efforts. Here, we present a method of determining the optimal fire history of a given area using the geometric mean of species abundance. The geometric mean is a useful indicator of biodiversity: it has heuristic properties that make it sensitive to changes in the abundance of sets of species and it can be related to the viability of species populations. Our case study is based on extensive field surveys of birds (560 sites), small mammals (280 sites), and reptiles (280 sites), completed in the Murray Mallee region (104, 000km²) of semi-arid Australia. First, we developed statistical models of species responses to time since fire. Second, we determined the relative abundance of species in three post-fire age-classes: early (0-9 years), mid (10-34 years) and late (≥35 years). Finally, we calculated the proportion of each age-class in the landscape that maximized the geometric mean of species abundance, for each taxonomic group.


Several species of birds, small mammals and reptiles were influenced by fire history: for example, 16 of 30 bird species were associated with the post-fire age of vegetation. The fire history that maximizes the geometric mean abundance is dependent on the number of species that prefer particular successional stages and the relative abundance of those species in each post-fire age-class. In this case, the optimal allocation of fire age-classes for birds was strongly weighted towards vegetation ≥35 years post-fire. This reflects the large number of species associated with older vegetation in the study region. This study demonstrates that the geometric mean can be used to determine landscape fire histories that maintain biodiversity, and provides a strong basis for ecological fire management.