First three-dimensional tracks for the Ascension Frigatebird Fregata aquila highlight the importance of altitude for behavioural studies
* Correspondence author. Email: email@example.com
1 Environment and Sustainability Institute, University of Exeter, Penryn Campus, Cornwall TR10 9EZ, UK;
2 Centre for Ecology and Conservation, University of Exeter, Penryn Campus, Cornwall TR10 9EZ, UK;
3 Conservation and Fisheries Department, Ascension Island Government, Georgetown, ASCN 1ZZ, Ascension.
Breeding seabirds are central place foragers that can travel vast distances from the nest to feed. Many factors influence the distance travelled to forage, including species ecology (Thaxter et al. 2012; Oppel et al. 2018), prey availability (Hamer et al. 2007; Paiva et al. 2013; Thorne et al. 2015), and competition (Ashmole 1963; Lewis et al. 2001; Wakefield et al. 2013; Oppel et al. 2015; Corman et al. 2016). Our understanding of the patterns and processes involved in determining foraging behaviour is crucial to conserving seabirds both at their breeding colonies and their often-distant foraging areas (Croxall et al. 2012; Lewison et al. 2012). A key goal is to identify geographical areas or predictable environmental features associated with foraging to prioritise areas for conservation measures (BirdLife International 2010). Foraging is often classified from tracking data by identifying slow and tortuous flight as searching behaviour associated with feeding attempts (Andersson 1981; Fauchald & Tveraa 2003). However, this approach generally operates within a two-dimensional (2D) plane, whereas aerial and marine species move in three dimensions (3D) (Bailleul et al. 2010; Belant et al. 2012). Many wide-ranging species have evolved strategies such as thermalling and dynamic soaring that improve flight efficiency when travelling over large distances (Weimerskirch et al. 2003; Sachs et al. 2012; Yonehara et al. 2016). These movements may not follow the direct trajectories that many species exhibit during transit behaviour, so additional data streams may be needed to confidently interpret behavioural patterns (McClintock et al. 2017).
Identifying important areas for seabirds is particularly challenging in the tropics, where low productivity and unpredictable prey distributions mean that foraging is often diffuse and not linked to stable environmental features (Ashmole 1963; Boekelheide & Ainley 1983; Weimerskirch 2007). Frigatebirds provide an example of extreme specialisation to such sparse environments. They are the only marine animals that are physically unable to enter the water despite relying entirely on marine resources, such as flying fish and squid (Weimerskirch et al. 2003, 2010), which can be brought to the surface by other aquatic predators (Au & Pitman 1986; Miller et al. 2018). Frigatebirds may also predate on seabird chicks or turtle hatchlings and feed through kleptoparasitism, but this is more common in immature birds (Stonehouse & Stonehouse 1963; Osorno et al. 1992; Lagarde et al. 2001). As their plumage is not waterproof and they cannot reliably take off when wet (Mahoney 1984), they cannot rest on the water and must remain in flight for the duration of a foraging trip (Weimerskirch et al. 2003; De Monte et al. 2012; Weimerskirch et al. 2016). Consequently, frigatebirds have extremely low wing loadings, allowing them to remain airborne for many days with very low energetic output (Brewer & Hertel 2007). They use thermals to reach high altitudes, allowing them to glide and soar to efficiently cover large distances (Weimerskirch et al. 2016), even sleeping on the wing in rising air currents (Rattenborg et al. 2016). However, foraging can only take place when birds are near to sea level, and so altitude data is likely to provide relevant information for behavioural models.
The Ascension Frigatebird Fregata aquila is endemic to Ascension Island, an isolated peak in the central tropical Atlantic, 1,300 km from the nearest land. The species is regarded as ‘vulnerable’ due to its restricted range (BirdLife International 2018a). The Ascension Frigatebird only recently recolonised the main island from Boatswainbird Islet after the successful eradication of Feral Cats Felis catus in 2006 (Ratcliffe et al. 2008, 2010). A previous study described the at-sea foraging distributions of this species, showing them to roam over a large marine area with some trips extending up to 1,100 km from the colony (Oppel et al. 2017). However, more detailed behavioural analyses are required to locate foraging hotspots within this very large area. In this study, we use a combination of GPS and barometric altimeter data to reconstruct the first 3D foraging tracks of the Ascension Frigatebird. GPS altitude is less accurate than latitude and longitude because four satellites are required for 3D positions, compared to three for 2D positions, and the location of those satellites affects accuracy (Dussault et al. 2019). These GPS altitude errors can occur at the scale of frigatebird flight (De Monte et al. 2012). Barometric pressure loggers are less prone to large errors but are affected by changing sea level air pressure (Berberan-Santos et al. 1997). Consequently, we first compare the distribution of error in the altitude recorded at a fixed point by GPS loggers and barometric pressure loggers. We then evaluate the implications of incorporating altitude into behavioural classifications performed using hidden Markov models (HMMs): a commonly used technique for decomposing tracking data into discrete movement patterns based primarily on speed and turning angle (Michelot et al. 2016; Bennison et al. 2017). Finally, we assess whether altitudes recorded by GPS loggers can be reliably used in place of barometric pressure data in future studies of this species (e.g. Rattenborg et al. 2016; Weimerskirch et al. 2016; Parr et al. 2017).
This project was funded by the Darwin Initiative project DPLUS063, The Ascension Island Ocean Sanctuary Project (ASIOS) jointly managed by Ascension Island Government Conservation and Fisheries Department (AIGCFD) and the University of Exeter, in-kind funding from AIGCFD. Altitude loggers were previously used for a NERC GW4+ Doctoral Training Partnership studentship awarded to Bethany Clark from the Natural Environment Research Council [NE/L002434/1]. For assistance in the field, we thank Sophie Tuppen, as well as Andy Richardson, Jolene Sim, Natasha Williams, Megan Benjamin, Diane Baum and Matthew Stritch. We thank Annette Broderick and Stephen Votier for advice, and two anonymous reviewers for their valuable input. An Environmental Research Permit was provided by the Ascension Island Government, and ethical approval was granted by the University of Exeter.
Marine Important Bird Areas toolkit: Standardised techniques for identifying priority sites for the conservation of seabirds at sea. Version 1. BirdLife International, Cambridge UK
Fregata aquila, Ascension Frigatebird. The IUCN Red List of Threatened Species 2018: e.T22697728A132597828.
Fregata andrewsi, Christmas Frigatebird. The IUCN Red List of Threatened Species 2018: e.T22697742A132599384.
An ecological comparison of oceanic seabird communities of the South Pacific Ocean. Studies in Avian Biology 8, 2–23.
lnfluence of satellite geometry and differential correction on GPS location accuracy. Wildlife Society Bulletin 29: 171–179.
raster: Geographic Data Analysis and Modeling. R package version 2.8-4.
geosphere: Spherical Trigonometry. R package version 1.5-7.
caret: Classification and Regression Training. R package version 6.0-84.
R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. www.R-project.org.