BackgroundGeolocators are useful for tracking movements of long-distance migrants, but potential negative effects on birds have not been well studied. We tested for effects of geolocators (0.8–2.0 g total, representing 0.1–3.9 % of mean body mass) on 16 species of migratory shorebirds, including five species with 2–4 subspecies each for a total of 23 study taxa. Study species spanned a range of body sizes (26–1091 g) and eight genera, and were tagged at 23 breeding and eight nonbreeding sites. We compared breeding performance and return rates of birds with geolocators to control groups while controlling for potential confounding variables.ResultsWe detected negative effects of tags for three small-bodied species. Geolocators reduced annual return rates for two of 23 taxa: by 63 % for semipalmated sandpipers and by 43 % for the arcticola subspecies of dunlin. High resighting effort for geolocator birds could have masked additional negative effects. Geolocators were more likely to negatively affect return rates if the total mass of geolocators and color markers was 2.5–5.8 % of body mass than if tags were 0.3–2.3 % of body mass. Carrying a geolocator reduced nest success by 42 % for semipalmated sandpipers and tripled the probability of partial clutch failure in semipalmated and western sandpipers. Geolocators mounted perpendicular to the leg on a flag had stronger negative effects on nest success than geolocators mounted parallel to the leg on a band. However, parallel-band geolocators were more likely to reduce return rates and cause injuries to the leg. No effects of geolocators were found on breeding movements or changes in body mass. Among-site variation in geolocator effect size was high, suggesting that local factors were important.ConclusionsNegative effects of geolocators occurred only for three of the smallest species in our dataset, but were substantial when present. Future studies could mitigate impacts of tags by reducing protruding parts and minimizing use of additional markers. Investigators could maximize recovery of tags by strategically deploying geolocators on males, previously marked individuals, and successful breeders, though targeting subsets of a population could bias the resulting migratory movement data in some species.Electronic supplementary materialThe online version of this article (doi:10.1186/s40462-016-0077-6) contains supplementary material, which is available to authorized users.

Warming in the Arctic has caused the transition from winter to summer to occur weeks earlier over the last half century, yet little is known about whether avian migrants have altered their timing of arrival on breeding areas to match this earlier seasonal transition. Over a 50-yr period, we examined trends in the timing of the first arrival for 16 avian migrant species at the terminus of their northward migration along the central Arctic coast of Alaska and compared these trends to factors potentially influencing migration phenology. Date of first arrival occurred an average of 0.12 d yr 1 or 6 d (range  3-10 d) earlier across all species and did not differ significantly among species between 1964 and 2013. Local climatic variables, particularly temperature, had a greater effect on a species first arrival date than did large-scale climatic predictors. First arrival date was 1.03 d earlier for every 1°C annual change in temperature, but there was nearly a 2-fold difference in the range of responses across species (0.69-1.33 d °C 1 ), implying that some species did better than others at timing their arrival with changing temperature. There was weak support for an influence of foraging strategy, migration distance, and flight path on timing of first arrival. Our findings, like others from temperate latitudes, indicate that avian migrants are responsive to changing environmental conditions, though some species appear to be more adaptive than others.