For migratory landbird species, large expanses of open water or inhospitable areas provide unique challenges during migration. Research on the strategies that species use to navigate barriers can yield insights into the factors shaping the evolution of migration and facilitate the identification of critical staging areas prior to barrier crossing. One such barrier, the Caribbean Sea, has received little study but must be negotiated by ≈50 migratory landbirds as they fly from South America to North America in spring. Recent discoveries from the Gray-cheeked Thrush (Catharus minimus), which undertakes non-stop flights >3000 km across the Caribbean Sea, raises the possibility that the breadth of potential strategies has been unappreciated thus far. We calculated fuel load and potential flight range in 9985 individuals of 16 species captured over 10 years at two stopover sites in northern Colombia to 1) evaluate the likely migratory strategy of these species as they depart northern Colombia in spring, and 2) evaluate the influence of family, diet, morphology and migratory distance on potential flight range. We found considerable variation in flight ranges and therefore strategies for crossing the Caribbean Sea/Gulf of Mexico barrier complex. In addition to Gray-cheeked Thrush, non-stop flights >2500 km were possible in Yellow-billed Cuckoo (Coccyzus americanus), Yellow Warbler (Setophaga petechia) and Northern Waterthrush (Parkesia noveboracensis). The remaining species were either capable of over-water flights to the Yucatan Peninsula/Cuba (>1800 km) or shorter flights to middle Central America (>1000 km) and likely required one or more stopovers to reach North America. Predicted flight ranges were influenced by morphology but not by distance, diet or taxonomic group, providing a novel insight into the evolution of migratory strategies. Our study confirms the vital role northern Colombia performs in providing energy for migratory birds and highlights the Caribbean as a key migratory barrier for many species.

These data were collected as part of the SELVA projects “Crossing the Caribbean” and the “Neotropical Flyways Project”. They represent long term banding/ringing datasets collected in order to better understand stopover behaviour and use by migratory landbirds in strategic regions in northern Colombia. More information can be found here www.neotropicalflyways.com and here http://selva.org.co/en/research-programs/migratory-species/crossing-the-caribbean/

Data collection. Migration monitoring stations were established in two study regions, one in northwest Colombia and one in the Sierra Nevada de Santa Marta (SNSM) in the northeast. In the northwest, one station was established in Finca Las Palmeras (8.529713, -76.102434), while in the SNSM stations were established in Finca La Victoria (11.122652, -74.087351) and Quebrada Valencia (11.235270, -73.797807). Stations were run daily or every other day during spring migration between 2009 and 2018. A migration station consisted of 7 to 10 mist-nets placed to maximize captures/recaptures of migratory birds. Nets were opened at dawn and operated for an average of five hours daily between late-March and mid-May (spring). All captured individuals were marked with individually numbered rings (Porzana Ltd. reporting address www.aselva.co) and we recorded age (following Pyle 1997), fat score (Kaiser 1993), wing chord (mm) and body mass (measured to nearest 0.1 g using an electronic balance). The dataset contains capture information for16 species of Neotropical migratory landbirds selected for this study, including ring number, date, hour of capture, site, species, age, fat, muscle, wing and body mass. In addition, two processed data are included (see below).

Fuel load estimation. To calculate the fuel loads (FL) in the dataset, we first estimated lean body mass (LBM) based on individuals captured during autumn migration with no visible fat deposits (see supplementary material associated with Cano et al. 2020). FL was then calculated as the difference between the mass of birds captured on spring migration and LBM: FL = (Body mass – LBM)/LBM. We expressed FL as a percentage to facilitate comparisons between species.

Lean Body Mass: The lean body masses in the dataset were calculated from linear regressions of body mass against wing length, based on individuals with fat score 0 captured during fall migration in northern Colombia. The raw data are not included here. Lean body mass equations follow the following formula: LBM = a + b*Wing length.

This dataset does not include individuals for which body mass was not recorded or those individuals with apparent errors in the recording of body mass (i.e. extreme values).