Freshwater conservation science at the landscape scale relies heavily on large data streams.
In this talk we present example applications on how freshwater data science can enhance conservation projects at local to continental scales and lead a call for the development of mechanisms to integrate disparate data streams.
The first data product is the HydroATLAS (Linke et al 2019) and its corresponding equivalent, the LakeATLAS (Lehner et al 2022) which aggregated ~250 variables to 1 million subcatchments, 8 million stream segments and 1.4 million lakes globally in one hydrologically connected framework. We illustrate the HydroATLAS framework using a current ARC funded project on modeling nutrient runoff to lakes, as well as a WWF-led conservation plan for the Yangtze River.
The second framework is an integrated species distribution modeling workbench. Based on environmental background data from the HydroATLAS and its Australian counterpart, we designed an interface that automatically connects to biodiversity databases such as the Atlas of Living Australia or GBIF and runs ensemble species distribution models that can then be extrapolated to the landscape. This will be illustrated based on a conservation assessment across Northern Australia.
While these are significant advances in conservation data science, the final section of the presentation will lament the poor integration of models produced by a generation of freshwater modelers (including the authors). We will end with a rallying call to develop standards integrating local to global datastreams and models and make them useable in conservation and restoration contexts.