In 2006, TNC sought to develop a classification of watershed units to be used in conservation planning and conservation measures benchmarks in South America. The objectives of this project were to: Map freshwater ecological systems at a continental scale for priority setting and conservation strategy planning, build freshwater-related datasets for reference and to have a standardized dataset to perform analysis and perform parallel analyses related to threats assessments, protected area gap analyses and other applications.
The process that we followed aimed to replicate the classification methodology proposed by Higgins et al, 2005 (Figure 1). Using this hierarchical classification system, large classification units known as ecoregions and medium sized classification units known as Ecological Drainage Units, were mapped by a team of scientists in the effort published as Abel et al, 2008. The next logical step was to classify Freshwater Ecological Systems (a.k.a. Ecosystems). The Higgins et al, 2005 document presents the classification system, but not the methodology to develop the data or the classification. Because of this, we tested several options to generate the units we would use to run the classification of ecosystems. We tested the Hydrosheds (Lehner, 2008) classification system, but there was no connectivity among the units to work with this system. We tested generating watersheds at one size scale, but this only gave us one dimension of the units we needed. We then tested the Nested Sized Watershed Units Model developed by Fitzhugh (2005, see Appendix 1). Basically, this system generates watersheds at different sizes defined by the user. Once these units have been generated, we simply combined the different sizes to generate one dataset that uses attribute codes to differentiate the different sizes. We then proceeded to attribute these watershed units with geo-physical data and run a classification algorithm to find similarities among the geo-physical characteristics.