Less frequent location information restricts inference about stopover behaviour but not migration phenology in shorebirds

We tested the effects of duty cycle on the estimation of metrics commonly used to summarize migration behaviour by sub‐sampling high‐resolution data from three shorebird species with different life‐history characteristics—American Avocet Recurvirostra americana , Grey Plover (Black‐bellied Plover) Pluvialis squatarola and Hudsonian Godwit Limosa haemastica . We found that metrics related to phenology (e.g. departure and arrival date, migration duration) were robust to varying GPS collection schedule (2–24 h between locations), and inference about metrics related to stopovers was generally reduced for devices collecting less frequent information, and these results were similar when considering short‐, intermediate‐ and long‐distance migrations. Where differences in estimates within a migration metric occurred, patterns varied among species and metrics, and linear models showed significant effects of duty cycle on the number of stopovers, stopover duration and within‐stopover step length, in addition to a marginally significant effect on the proportion of time stopped over. Our results suggest that, for research questions related to stopover ecology, frequent data collection (e.g. multiple locations per day) should be prioritized in study design, but questions related to migration phenology can be addressed with less frequent information (e.g. one location per 24 h).