Introduction
Effective animal conservation first requires detailed knowledge of an organism’s spatial and temporal distributions. Seabirds present conservationists with challenges in this regard in part because of their high mobility (Norris and Pain, 2002; Oppel et al., 2018) and extensive foraging ranges, particularly when not breeding (Flint, 1991; Catry et al., 2009). In light of their perilous conservation status compared to other avian taxa (Votier and Sherley, 2017; Dias et al., 2019), characterizing seabird movements to inform conservation strategies is paramount, especially since many have undergone steep recent declines in population size worldwide (Croxall et al., 2012; Paleczny et al., 2015).
Seabirds are excellent bio-indicators of anthropogenic pressures on the marine environment (Furness and Monaghan, 1987). Their movements may also reflect the distributions of other key elements of marine food webs (Phillips et al., 2009). This is especially true of tropical seabirds, many of which have near-obligate relationships with cetaceans and predatory fish such as tuna that drive prey species to the surface where seabirds feed on them (Veit and Harrison, 2017; Reynolds et al., 2019) in so called “facilitated foraging” (Maxwell and Morgan, 2013). Understanding seabird movements is therefore of high importance for planning and assessing the effectiveness of marine reserves (Lascelles et al., 2016; Anon 2018; Oppel et al., 2018).
While at-sea distributions of far-ranging seabirds are difficult to assess through direct observation, tagging allows their movements to be studied remotely and in great detail (Fehlmann and King, 2016; Bernard et al., 2021). Here, we deploy geolocator (GLS) tags on sooty terns (Onychoprion fuscatus) nesting within one of the world’s largest marine reserves centered on Ascension Island in the Atlantic Ocean. Sooty terns are long-lived and pan-tropically distributed (Schreiber et al., 2002), and they are near-obligate commensals of tuna, cetaceans and other sub-surface predators (Au and Pitman, 1986). With a global population of 21–25 million birds, their conservation status is of “Least Concern” but under the Ascension Island Government’s (AIG’s) Schedule of the Wildlife Protection Ordinance they are a “Protected Species.” Many sooty tern populations have declined significantly since the mid-1,800s (Hughes et al., 2017a), including our study population on Ascension Island (Reynolds et al., 2019) that is currently the largest breeding colony of this species in the Atlantic Ocean. In August 2019, the entirety of Ascension Island’s 200 nautical mile Exclusive Economic Zone (EEZ)—an area of approximately 445,000 km2—was designated as an MPA that is closed to all commercial fishing. Much remains to be investigated in assessing this MPA’s benefits to marine taxa. Here, we investigate year-round movements of this iconic Ascension seabird over 5 years. Multi-month and multi-year movements of both sexes are essential to evaluate the consistency in the usage of at-sea areas. Ascension Island is the only location where sooty terns breed sub-annually (i.e., every 9.6 months; Reynolds et al., 2014) resulting in birds over many years experiencing highly variable foraging conditions that are driven seasonally (Soanes et al., 2021). We relate their at-sea locations and behaviors predominantly outside of their breeding seasons to prevailing meteorological and oceanographic conditions, season and commercial fishing activities targeting the surface-schooling skipjack tuna (Katsuwonus pelamis) and yellowfin tuna (Thunnus albacares) with which they commonly associate (Reynolds et al., 2019). We use habitat modeling to define suitability of at-sea habitats and relate this to bird distributions. Finally, we examine the extent of overlap between terns and the MPA, and assess the latter’s potential to protect the wider ecological community.
Discussion
Pelagic seabirds undergo some of the most extensive migrations of any vertebrate taxa (e.g., Shaffer et al., 2006; Egevang et al., 2010; Weimerskirch et al., 2015). Likewise, sooty terns from the Ascension Island colony undergo extensive post-breeding migration that justifies their characterization as ocean “wanderers” (Feare, 2017). Their estimated total migration distances are similar to those of post-breeding sooty terns tracked from Bird Island in the Seychelles in the western Indian Ocean (i.e., 46,984 ± 5,834 km cf. 52,503 ± 6,546 km; Jaeger et al., 2017). However, migrating Ascension birds ranged closer to their breeding location than did Seychelles’ birds (i.e., 2,904 ± 857 km cf. 4,423 ± 1,190 km; Jaeger et al., 2017). Whether this is attributable to differences in productivity and thus prey availability in foraging areas (Weimerskirch, 2007), or to intensity of interference competition from conspecifics (Oppel et al., 2015) and/or heterospecifics (Weimerskirch et al., 2005) is unclear. Duration of migration of our study birds appeared to be relatively fixed (i.e., 7.9–8.6 months; Table 1), perhaps a consequence of their sub-annual breeding (Chapin, 1954) compared with annually breeding sooty terns that may have more temporal flexibility (Reynolds et al., 2014).
Data from the geolocator (GLS) tags indicated that birds spent the vast majority of the day and night in flight, and little time, as revealed by immersion data, on the ocean surface where they likely feed on small fish driven by cetaceans and large predatory (Maxwell and Morgan, 2013; Table 1). Therefore, they may be among the most aerial of all bird species. Unusually for a seabird, sooty terns appear to have poorly waterproofed feathers because of the reduced lipid content of their preen oil compared with other tern species that land on the water (Johnston, 1979). An inability to repel water from plumage effectively can have lethal consequences (Dinsmore, 1972) as well as non-lethal ones such as negatively impacting their flight capabilities as they expend much energy attempting to take-off (Mahoney, 1984).
There were distinct differences between at-sea distributions of sooty terns according to their sex and year (Figure 1). Compared with polar and temperate waters, tropical oceanic waters are less productive and food availability is often patchy and less predictable (Weimerskirch, 2007). The distribution of avian and mammalian marine consumers therefore often reflects spatially dynamic patches of high prey availability (Moore and Kuletz, 2019). Oceanographic features such as eddies, water mass boundaries and currents can influence the at-sea distributions of seabirds (Schneider, 1991). Processes such as up- and downwelling can also result in the accumulation of planktonic organisms, providing predictable sources of prey for fish that in turn attract seabird and mammalian consumers (Catry et al., 2009; Weber et al., 2021). The highly variable at-sea distributions of our birds between years suggest complex relationships with ocean productivity and circulation. In 2013, for example, terns dispersed more widely during the significant elevation in sea surface temperature (i.e., high SSTA) and subsequent low productivity (i.e., low chlorophyll α concentration or CHLA) inside the Exclusive Economic Zone (EEZ) (Table 2 and Figure 1). Further research should therefore assess the degree of reliance of sooty terns on ocean productivity mediated through their near-obligate foraging associations with cetaceans and large predatory fish (Maxwell and Morgan, 2013). This will require detailed monitoring of movements of sub-surface predators in parallel with avian movements. Such research is especially pertinent in the face of ocean warming. Our findings in 2013 suggest that as the SSTA increases, birds appeared to migrate longer distances and thus to overlap less with the designated Ascension marine protected area (MPA). Conversely, the global abundances of yellowfin and skipjack tuna are predicted to increase in tropical waters as oceans warm (Erauskin-Extramiana et al., 2019).
Seasonal variation in the CFRs of males and females most likely reflects fluctuations in the seasonal availability of fish prey. Sooty terns in the tropical Atlantic predominantly feed on small fish such as halfbeaks (Oxyporhamphus micropterus), blue flying fishes (Exocoetus volitans), redlip blennies (Ophioblennius atlanticus), Simony’s frostfishes (Benthodesmus simonyi) and hairtails (Trichiurus spp.) (Ashmole, 1963), but also squid (Teuthida) (Reynolds et al., 2019). Significant overlap of CFRs between the sexes across years suggests that they are employing similar migratory strategies in response to food availability. There is no evidence of sex differences in their diet but future research might usefully examine such foraging strategies in greater detail such as by month (Figure 2).
In habitat suitability models explaining their core foraging ranges (CFRs), distance to colony (DCOL) featured as a major driver of sooty tern distribution in both sexes across all years of the study (Figure 3). This is unsurprising given the geometry of dispersal following breeding on the island, migration and return to the same point to breed in subsequent seasons. Fewer cells at shorter distances from the island that all birds need to pass through and a much larger potential foraging area as distance from the island increases represent the “null model” of foraging distribution of seabirds (Oppel et al., 2017). This null model reflects the random dispersal of birds away from, and return to, the island under which their core utilization distributions (UDs) will always be centered on the island. Besides DCOL, SSTA was the other main predictor of non-breeding distributions (Figure 3), with tracked birds being less likely to occur in areas of higher SSTA. A number of seabird species have been shown to respond with high sensitivity to changes in SST (Hunt et al., 1992) with foraging success of sooty terns impacted heavily by its variation (Erwin and Congdon, 2007). In the Pacific Ocean, SSTAs are precursors to intense El Niño events that can have negative impacts on temperate and tropical seabirds manifested by declines in their breeding productivity (Devney et al., 2009). These climate anomalies deflate upwelling phenomena and hence cause seasonal-scale crashes in productivity at lower trophic levels and resultant reductions in the abundance and hence availability of fish prey (Spear et al., 2001; Stenseth et al., 2002).
We found that birds occupied foraging habitat non-randomly and shared areas also occupied by fishing vessels that target surface-schooling tuna. Tropical seabird species are surface foragers and therefore they do not interact directly with long-line fishing that targets mature tuna that rarely feed at the ocean surface (Danckwerts et al., 2014). In contrast, purse-seiners target more juvenile tuna with which sooty terns may interact far more often when foraging. Their non-random association with areas of high purse-seine fishing effort in our study highlights a need for more research elaborating their foraging relationship with these predatory fish.
Given low reported direct mortality of sooty terns from fishery interactions (e.g., Gales et al., 1998; Danckwerts et al., 2014), if the Ascension MPA is to benefit this species substantially, it must protect the large predatory fish with which they commonly associate. Limited tracking data are available for such species in Ascension Island waters but satellite-tagged yellowfin tuna spent 100–200 days resident in the Ascension EEZ, much of it within 20 km of the island (Richardson et al., 2018). By alleviating the fishing pressures on semi-resident fish stocks, it is possible that the MPA, which prohibits all forms of commercial fishing, may increase the foraging success of sooty terns while they are located within the Ascension EEZ such as during breeding periods when adults appear to range only approximately 256 km from the island on average (Anon, 2018). However, it is doubtful that the Ascension MPA will enhance the regional population status of highly mobile pelagic fish stocks without being far larger (Grüss, 2014; Curnick et al., 2020), and sooty terns spend considerable periods foraging in the high seas outside the MPA boundaries during their non-breeding migrations. As for other members of the seabird community that have been tracked to date, breeding masked boobies (Sula dactylatra) appear to spend all of their time within the MPA (Oppel et al., 2015) while Ascension frigatebirds (Fregata aquila) frequently range far beyond it (Oppel et al., 2017).
We think that the population of sooty terns on Ascension Island is now stable in size, albeit markedly reduced from 2.35 million birds in 1942 (Hughes et al., 2017a). A cause of their decline was undoubtedly heavy predation pressure on adults from invasive cats (now absent due to successful cat eradication), and that from rats and mynas on chicks and eggs (ongoing) (Hughes et al., 2019). Beyond on-island predator-driven population dynamics of sooty terns, we have a greater understanding of potential at-sea drivers of their breeding population size as a result of a long-term dietary study of terns spanning 110 years that included the mid-twentieth century decline and stabilization of their population size (Reynolds et al., 2019). This indicated that birds had shifted diet from predominantly small fish prey to squid, indicating that prey availability to birds had changed partly as a result of overfishing of tuna (Reynolds et al., 2019). However, whether the shift was due to reduced fish prey availability, resulting from either overexploitation of tuna stocks in the tropical Atlantic or increased SST in recent decades, remains unclear (discussed in Reynolds et al., 2019 and references therein). Furthermore, it is not simple to estimate the relevant impacts of predation and food availability in explaining the steep decline in the sooty tern population. If we are to evaluate the conservation benefits of the MPA to sooty terns, further at-sea research must focus on movements of large predatory fish species, the extent to which their ranges are protected by the MPA, the consequences for small fish prey availability to birds, and the comparative foraging successes of birds within and outside of the MPA at critical times of their sub-annual cycle. Therefore, our study provides baseline information about post-breeding sooty tern movements that can be used to inform future studies assessing the effectiveness of the MPA. In conclusion, we argue that key foraging habitat for some species of the seabird community will be protected by the Ascension Island MPA but benefits to sooty terns outside of the breeding period will likely be more limited.