Adapting to climate change: A risk assessment and decision making framework for managing groundwater dependent ecosystems with declining water levels. Supporting document 3: Identifying thresholds for responses of amphibians to groundwater and rainfall decline
|Title||Adapting to climate change: A risk assessment and decision making framework for managing groundwater dependent ecosystems with declining water levels. Supporting document 3: Identifying thresholds for responses of amphibians to groundwater and rainfall decline|
|Year of Publication||2013|
|Authors||Mitchell, N, Sommer, B, Speldewinde, P|
|Institution||National Climate Change Adaptation Research Facility|
Amphibians are key indicators of wetland health and under conditions of hydrological change it is important to be able to predict changes in their diversity and abundance. Here we develop a methodology for predicting how nine amphibian species with different life histories could potentially respond to declining groundwater, and declining rainfall – both of which are occurring on the Gnangara Mound north of Perth in southwestern Australia. This example can be considered as ‘data-poor’ because although we had access to data on amphibian assemblages and relative abundance from systematic surveys, we could derive few insights into the environmental drivers of such patterns using multivariate analysis. Instead we turned to amphibian biologists with relevant expertise on Gnangara Mound amphibians to derive conceptual models of the most important factors explaining whether populations would persist or decline under anticipated environmental change. These models were constructed as Bayesian Belief Networks (BBNs) and were developed for each of three reproductive guilds: amphibians that breed in water, amphibians that breed in terrestrial nests that are later flooded, and one entirely terrestrial-breeding species. The models showed that aquatic breeding species were most sensitive to changes in the hydroperiods, but were unable to demonstrate an impact of groundwater level because we lacked the information to derive an empirical link between groundwater and hydroperiods. The terrestrial-aquatic breeding species were also sensitive to hydroperiod length, with declines in autumn and winter rainfall further decreasing the probabilities of population persistence. The BBN developed for the Turtle frog – the terrestrial-breeding species – assumed that groundwater levels would influence soil moisture and that this species would experience population declines if groundwater levels were either high or very low. Hence, situations where groundwater-dependent wetlands transition to terrestrial ecosystems under groundwater decline could potentially suit Turtle frogs provided that an appropriate vegetation community develops. While the BBNs developed here are largely conceptual, and thus preliminary, they demonstrate a promising approach for anticipating the impacts of groundwater decline on amphibians, provided that the links between hydrological processes and amphibian thresholds can be determined empirically or mechanistically. Our models also demonstrate the key role that seasonal rainfall events play in triggering reproduction, and suggest that the drier climates projected for south-western Australia will place additional stress on amphibian communities.