Optimized Survey Design for Monitoring Protocols: A Case Study of Waterfowl Abundance

Abstract

Nationwide monitoring programs are important tools that quantify the status and trends of natural resources providing important information for management and conservation decisions. These programs operate at large spatial scales with standardized protocols requiring wide-spread participation. However, resource limitations can reduce participation, which can then compromise the spatial replication needed for nationwide inference. The Integrated Waterbird Management and Monitoring program is an example of a national monitoring program that could benefit from a reduction in sampling effort to facilitate increased participation and ultimately broader inference. Therefore, we examined various sampling schemes to determine if it is possible to reduce the sampling effort while maintaining the statistical accuracy needed to support management. We found that instead of needing to census a National Wildlife Refuge, sampling effort could be reduced while accurately estimating waterfowl abundance to within 10% of the census count by surveying just 2/3 of all the sample units or 3/4 of the total survey area. Not only did this guideline apply to our five pilot National Wildlife Refuges, but it was also further validated by applying it to four additional National Wildlife Refuges. We hope that by applying this finding to other National Wildlife Refuges, we can increase participation in the program by reducing the logistical and financial burden of sampling.