Evaluation of a dual-frame design to estimate occupancy and productivity of bald eagle nests in Kenai Fjords National Park

Thompson, William; Phillips, Larua

The Southwest Alaska Network (SWAN) is monitoring long-term trends in nest occupancy and productivity of bald eagles in five national park units in southwest Alaska, including Kenai Fjords National Park (KEFJ). The U.S. Fish and Wildlife Service (USFWS) recently proposed a dual-frame sampling design that incorporates a double-observer component as an adjustment for nests that are missed during surveys. The double-observer portion of this design was field tested in a spring occupancy survey of bald eagle nests conducted by SWAN and KEFJ staff over nearly all of the park’s coastline during May 2009. This survey also produced an updated map of occupied and empty nests in KEFJ, and estimated costs for performing this survey. As the next step in protocol development, the objectives of this study were to: 1) use information collected during the May 2009 survey to refine the sampling frame and to inform simulations for determining optimal sample unit length and sample size; 2) further evaluate the dual-frame estimator for monitoring nest occupancy and productivity of bald eagles in KEFJ; and 3) refine the techniques and standard operating procedures that will ultimately become part of SWAN’s monitoring protocol for bald eagles in KEFJ. Simulation results indicated that a sample size of 25 segments of lengths of 7.8 mi or 9.3 mi should be adequate to achieve a CV of at least 12% for the estimated number of nests with incubating bald eagles in Kenai Fjords. We detected 29 nests with incubating adults, 14 of which were newly detected nests, from a random sample of 25 segments during 9-12 May 2010. Twenty-three of these nests were in Sitka spruce (Picea sitchensis), three were on the ground, one was in a mountain hemlock (Tsuga mertensiana), and one was in a balsam poplar (Populus balsamifera). The dual-frame estimator produced an estimate of 153 occupied nests (95% confidence interval: 88-218 nests), but this estimate is highly suspect because of the rear-seat observer’s negatively biased estimate of detection probability caused by distractions related to field-testing a new data entry program. Nineteen (66%) of the 29 occupied nests contained at least one nestling during the nest productivity survey during 20-21 July 2010. Of these 19 nests, 14 had a single fledgling and five had two fledglings. There were an estimated 53 young fledged (95% Bayesian credible interval: 28-96) within the sampling frame. As an alternative to the dual-frame estimator, we propose treating our survey as a random sample of segments that contain nests with detection probabilities of one (known nests) or less than one (newly detected nests), the latter of which would be estimated via the double-observer approach. These data will be re-analyzed within a Bayesian hierarchical modeling framework with a spatial random effect added that would allow for spatially explicit estimates of nest occupancy (or nest productivity for those data).

Evaluation of a dual-frame design to estimate occupancy and productivity of bald eagle nests in Kenai Fjords National Park

Series

Collections