AbstractAbstractThe Marbled Murrelet (Brachyramphusmarmoratus) is a highprofile, federally threatened seabird, but noreliable estimates of population trends at inlandbreeding areas exist for this species. We conductedland-based radar studies of Marbled Murrelets at3–7 sites on the Olympic Peninsula,Washington, in 1996–2002 and 2004 to estimatepopulation changes and to examine relationshipsbetween our counts and oceanographic conditions,murrelet productivity, and regional at-sea countsof murrelets. Morning radar counts of murreletsvaried significantly among and within sites but didnot decline from 1996–2004, suggesting thatthe inland breeding population of murrelets isstable in this area. A retrospective power analysisindicated that we had a 25% and 56%chance of detecting 2% and 4% annualdeclines, respectively. Thus, if relatively smallannual declines did occur during our study period,there is a high probability that they would havegone undetected, even though they could add up to abiologically important decline over time. It isunlikely that murrelets on the Olympic Peninsuladeclined by ≥6% annually, however,because retrospective analyses indicated that powerto detect such declines was >88%. Therewas no significant relationship between radarcounts and at-sea counts or productivity ofmurrelets in the nearby San Juan Islands during thestudy period. We also did not detect a relationshipbetween radar counts and mean sea-surfacetemperatures or the Northern Oscillation Index,suggesting that variation in oceanographicconditions (e.g., the strong 1998 ElNiño event) was not associated withvariation in morning radar counts of MarbledMurrelets. A prospective power analysis indicatedthat small (2%–4%)annual declines could be detected with reasonablyhigh power (≥80%) with thecurrent radar sampling design by extending thestudy to 11–15 years.