Rapid Dissemination of Regional Cross-Correlation in Waterfowl Counts Across Northeast National Wildlife Refuges

• Activity Number: 171
• Type: Topic Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Consulting
• Author(s): Charles Frost*
• Companies: U.S. Fish and Wildlife Service
• Keywords: long-term monitoring ; waterfowl ; reporting

Abstract

Long-term monitoring programs often suffer from delayed reporting that inhibits the use of data in timely decision-making. We provide an example of a flexible R package that links the steps and outputs of a comprehensive monitoring plan for waterfowl and shorebirds. The package provides immediate summaries of counts and effort, computes cross-correlation between sites and forecasts expected bird levels in the near future, streams input layers into output files formatted for more specialized analysis, including hierarchical migration models, density mapping for targeted surveillance, simulations of use of individual wetlands, and Bayesian belief networks to support management across a suite of refuge projects influenced by waterfowl and shorebird abundance and timing. We demonstrate the ease of integration of count data through R into other external software packages such as NetLogo, OpenBUGS, and Netica. The package will not only provide refuges with instant reporting, but also quantify the linkages among groups of refuges that have historically used these data independently of each other, two critical components of decision-making in the National Wildlife Refuge System.