For the last 20 years, systematic conservation planning has been an invaluable tool in assigning terrestrial protected areas for biodiversity conservation. Its main goal is to represent the full variety of taxa or ecosystems in a proposed reserve design, while minimising costs and effort. We introduce this widely established technique to river systems, using fish and macroinvertebrates. Central questions in this approach include choice of targets, proper selection of algorithms and input data. Input data can range from species data to higher taxonomic levels, from predicted occurrences to GIS surrogates. Using datasets from Canada and Australia, we built and compared three sets of conservation models. All conservation plans were built using sub-watersheds as the unit of replication. The first set of models tries to assign reserve networks using real data. Using a more advanced technique recently developed for terrestrial application, the second set uses probabilities of occurrence modelled by nearest neighbour prediction to achieve reserve design targets. The third set of models assesses the representativeness of GIS surrogates alone. Possible congruence among outputs adds confidence in to the value of a single approach.
|Publication status||Published - 2005|
|Event||North American Benthological Society Annual Meeting - Louisiana, United States|
Duration: 1 Jan 2005 → …
|Conference||North American Benthological Society Annual Meeting|
|Period||1/01/05 → …|