Shorebirds typically spend more time fueling and refueling during migration than they spend in actual flight, and consequently their stopover ecology has important implications for fitness and conservation. We examined spatial variation and long-term changes in stopover duration of radio-tagged western sandpipers Calidris mauri over three decades on the Pacific Coast of North America. Birds were monitored during 10 periods of northward migration at four key stopover regions including: San Francisco Bay, CA; Willapa Bay/Grays Harbor, WA; Vancouver Island/Fraser River Delta, BC; and Copper River Delta, AK. We used Cormack–Jolly–Seber models to estimate the daily probability of staying at a stopover site (φ) corrected for daily detection rates (p) in relation to annual and seasonal variation. Daily probabilities of staying were then used to calculate length of stay (LOS) in days. Mean LOS decreased at stopover sites as birds progressed northward and varied from 23.1 days at San Francisco Bay to 3.6 days at the Copper River Delta. The duration of length of stay did not show a strong temporal trend by year or decade from 1992 to 2022. Effects of covariates on LOS varied by stopover site, with locally tagged birds staying 2 to 4 days longer than birds tagged farther south, females staying longer than males, and LOS being negatively correlated with mass of the bird at the time of capture with lighter birds having a longer LOS. We suggest that western sandpipers on northward migration may be limited in their ability to reduce LOS beyond the current duration over much of northern parts of the Pacific Flyway. Covariate effects demonstrate that stopover decisions of individual birds are impacted by their energy reserves, but stable lengths of stay over three decades suggest that western sandpipers may have limited flexibility to cope with environmental change in the northern part of their migration.

Birds were captured and tagged at 10 different locations (latitude – longitude): Playa Ceuta, Mexico (23.94 -106.98), Bahia Santa Maria, Mexico (25.11 -108.22); Punta Banda, Mexico (31.74 -116.63); Point Mugu, California (34.10 -119.12); three areas of San Francisco Bay, California (37.53 -122.07 , 37.62 -122.12, and 37.49 -121.96), including 10 birds in 1992 at nearby Bolinas Lagoon (37.93, -122.69); Honey Lake, California (40.16 -120.23); Grays Harbor, Washington (46.97 -123.95 and 46.89 -124.05); four coastal sites on Vancouver Island, British Columbia (49.05 -125.71, 49.11 -125.88, 49.07 -125.75, and 49.03 -125.69), and two areas of the Fraser River Estuary, British Columbia (49.22 -123.22, and 49.05 -125.69).

All birds were captured with mist nets, banded with an individually numbered metal band, weighed for body mass (± 0.5 g), and measured for three standard morphometric variables (± 1 mm), including exposed culmen, wing (both natural and flattened), and tarsus. Western Sandpipers exhibit culmen length dimorphism with males having shorter bills than females (Stein et al. 2008, Jimenez et al. 2015). We used exposed culmen length to determine sex of birds based on Page and Fearis (1971, males ≤ 24.2 mm and females ≥ 24.8 mm). For 19 birds that fell between 24.3 to 24.7 mm, we took the midpoint value of 24.5 mm and assigned birds with culmen lengths >24.5 as females and < 24.5 as males.

Days when tagged birds were detected at each stopover site were assigned a value of 1, and days they were not detected were assigned a value of 0. Days when no manual survey effort occurred were coded with a period (.) to indicate missing encounter data. Sites with Motus receivers were assumed to be constantly surveyed and were assigned a 0 when no detections were made. The encounter history period for each stopover site began on the first date a tag was deployed or detected across all years and continued until the last date of detection per site across all years (see Migration Period, Table 2). For example, the earliest detection at San Francisco Bay occurred on 7 April 2006 of a bird tagged in Point Mugu, and the detection that occurred latest in the migration period was a locally tagged bird on 12 May 2004, so the 36-day capture history period was bounded by 7 April and 12 May.