Hierarchical frameworks for distributional and life history data: Implementation of a new ecoinformatics tool

Background/Question/Methods

Threats to marine and estuarine species operate over many spatial scales, from nutrient enrichment at the watershed/estuary scale to climate change at a global scale. To address this range of environmental issues, we developed a hierarchical framework for species’ abundances, distributions, habitat requirements, and life history attributes across ecological scales. A key component of the topology is to utilize hierarchical class structures to describe species’ tolerances and life history attributes.  Classes make it possible to estimate species’ tolerances from distributional data while providing a mechanism to capture qualitative attributes (e.g., reproductive type).  Advantages of the hierarchical topology include the ability to capture information at the resolution of the available data, ability to analyze the information at the appropriate resolution, and it can be expanded to include more detailed breakouts. This framework is being implemented in the Coastal Biological Risk Analysis Tool (CBRAT), a joint EPA-USGS web-based information system and decision tool. Here, we explore the utility of this framework in regard to distributions of Pacific coast soft-bottom benthos along abundance, salinity, depth, and geographical distributions.

Results/Conclusions

Salinity tolerance is an example of a hierarchical class based on quantitative thresholds, and we developed a three-level salinity classification that allows for more flexibility and finer resolution than the standard Venice system. Modeling species’ distributions showed no loss of predictive power when using these salinity classes compared to quantitative values. A key component of the framework is to map species’ global distributions, which are based on the Marine Ecoregions of the World (MEOW), a hierarchical biogeographic schema. To expand beyond presence/absence so as to assess regional-scale abundance patterns, we developed a three-level population hierarchy - from present/absent at the highest level to four classes of abundance at the finest level of resolution. Regional abundances are being estimated from the integration of quantitative surveys, qualitative descriptions in natural history texts, and online databases of occurrence records (e.g., OBIS).  When applied to quantitative benthic survey data from the Pacific Coast, we found that of the 232 known soft-bottom bivalves, 187 were recorded in the surveys, with the remaining 45 species classified as “very rare”. Comparison of these quantitative population estimates to those from natural history texts and online databases showed reasonable agreement though there were some differences due, in some cases, to sampling artifacts. Based on these encouraging results, we are applying this framework to predict species’ vulnerability to climate change.