Population distribution changes for Australian fish could provide scientists with a useful tool in predicting the effects of climate change. Bond et al. (2011) examined 43 species of Australian freshwater fish and quantified their results into species distribution models (SDMs). The SDMs provided a useful approach for examining predicted range shifts and provided a clear way of describing the types of environment in which these species of fish will be encountered. When SDMs were combined with future climate scenarios the models predicted future population and range shifts that in some cases described total population loss. In conclusion the author’s remark on their ability to predict current and future distributions using statistical models but that the models are just a step and future efforts in mechanistic modeling and in climate scenarios will be needed to further understand the effects of climate change on fish species.
Bond, N., Thomson, J., Reich, P., Stein, J., 2011. Using species distribution models to infer potential climate change-induced range shifts of freshwater fish in south-eastern Australia. Marine and Freshwater Research 62, 1043—1061.
Bond et al. used fish distribution data from survey records drawn from the Victorian Department of Sustainability and Environment’s Aquatic Fauna Database (AFD). In gathering these data they excluded sites below large impoundments because of markedly atypical behavior created by those sites. They gathered their environmental data from a digital elevation model (DEM), which characterized stream networks all across Victoria. To enhance model sensitivity they restricted the environment data to areas that fish had been officially surveyed and recorded. For the data characterizing river flows, they used gauge data from 120 unregulated sites around Victoria that had significant records to quantify water flow patterns. Although high flow events couldn’t be modeled, they predicted that this would be offset in their model by the large effect that low flow events have on fish distribution. Their climate scenarios came from changes in temperature, precipitation, and evapotranspiration (the water put back into the atmosphere by plant respiration and evaporation). These scenarios corresponded to low, median, and high estimates for 2030, and were run in tandem with hydrologic model data. The statistic modeling was based on a system of boosted regression trees, a form of model averaging; model fit was based on residual error (R2).
Their results showed only five of their water flow models (hydrologic models) could be confidently predicted, with high water flow characteristics showing very poor residual error (R2<0.4). The SDMs for the current fish populations were extremely accurate, with only two species with inaccurate predictions. They ran the climate scenario prediction for domestic and exotic fish species and found the results differed, but overall the species showed strong and consistent range patterning.
Their overall goal of making SDM’s to describe historical distributions of the 43 specifies of fish was accomplished successfully. The climate scenarios that were found could provide a useful approach to examine future range shifts. BRTs (Boosted Regression Trees) successfully carried out the model’s predictions and their capacity to fit non-linear response functions helped describe species response to environmental changes. Bond et al.analysis suggests that the non-linear associations of water flow/climate variables is common, thus why the BRTs are such an extremely useful modeling tool.
SDMs showed the combined impacts of altered temperature and water flow patterns rising from climate change in south-eastern Australia caused distribution and population changes in freshwater fish. One of the main findings was that the fish shifted up along an elevation gradient, and also south in direction in response to the climate change scenarios. Although the results are just models the authors suggested they represent an important step in finding the long-term understanding of finding climate change impacts and their response strategies.