Cooperative Ecosystem Studies Unit, Rocky Mountain CESU

archived

The US Geological Survey, Northern Rocky Mountain Science Center (NOROCK) is offering a funding opportunity to a CESU partner to review the integration of heterogeneous data as this topic has been identified as a pressing challenge. Applications in ecology dealing with multiple data sources and/or species types need further study and methodological developments. Typically, ecological settings and problems require spatial and temporal dependencies. With this opportunity, the Center would like to conduct a review of Bayesian methods for integrating multi-type and multi-species ecology data with an emphasis on spatial and spatiotemporal settings. Upon completion, this review should serve as a bridge for the submission of additional proposals to other partners focusing on integrated population models (IPMs) and Bayesian hierarchical models for multi-species data. IPMs provide a way to combine heterogeneous data sources for combined inferences. These models are necessary in many ecological settings where data collection can be expensive and difficult, therefore requiring the combination of data sets of different types and spatial domains. Many times in ecological settings, multi-species data sources are often incorrectly handled by failing to account for differences between species, which leads to inaccurate inferences and unidentified uncertainties. Conventional IPMs make strong assumptions about the independence of the different data sets, which may not always be reasonable. Additionally, analyzing data with spatial and spatiotemporal structures can present challenges in IPMs.

High