Sympatrically breeding congeneric seabirds (Stercorarius spp.) from Arctic Canada migrate to four oceans

Abstract Polar systems of avian migration remain unpredictable. For seabirds nesting in the Nearctic, it is often difficult to predict which of the world's oceans birds will migrate to after breeding. Here, we report on three related seabird species that migrated across four oceans following sympatric breeding at a central Canadian high Arctic nesting location. Using telemetry, we tracked pomarine jaeger (Stercorarius pomarinus, n = 1) across the Arctic Ocean to the western Pacific Ocean; parasitic jaeger (S. parasiticus, n = 4) to the western Atlantic Ocean, and long‐tailed jaeger (S. longicaudus, n = 2) to the eastern Atlantic Ocean and western Indian Ocean. We also report on extensive nomadic movements over ocean during the postbreeding period (19,002 km) and over land and ocean during the prebreeding period (5578 km) by pomarine jaeger, an irruptive species whose full migrations and nomadic behavior have been a mystery. While the small sample sizes in our study limit the ability to make generalizable inferences, our results provide a key input to the knowledge of jaeger migrations. Understanding the routes and migratory divides of birds nesting in the Arctic region has implications for understanding both the glacial refugia of the past and the Anthropocene‐driven changes in the future.


| INTRODUC TI ON
Radar and biologging technology (Alerstam et al., 2007;Egevang et al., 2010) have provided new knowledge about polar systems of avian migration. But in the Nearctic, routes and destinations remain unpredictable. For seabirds, it is often difficult to predict which of the world's oceans birds will migrate across or to from their Arctic nest sites (Davis et al., 2016;Gutowsky et al., 2020;Mehl et al., 2004). Migratory patterns are the legacy of millions of years of changing lands and seascapes, glaciation events, intra-and interspecific competition, and speciation (Johnson & Herter, 1990).
Studying the migratory connectivity of related species-the linking of migratory individuals or populations between different stages of the life cycle (Marra et al., 2019)-can provide insights into the interplay of ecological, evolutionary, and anthropogenic influences on migration patterns, habitat use, and coexistence and persistence (Weber & Strauss, 2016;Weber et al., 2017) in a changing Arctic (IPCC, 2014).
Jaegers (skuas outside of North America) are the three smallest Stercorarius species, a genus of predatory and kleptoparasitic migratory seabirds that nests in the Arctic (Furness, 1987). Jaegers are the only Stercorarius that nest in North America where they play an important regulating role on other taxa within the Arctic tundra summer food web (Gilg et al., 2003;Krebs et al., 2003). Jaegers breed sympatrically across much of the North American Arctic (Furness, 1987) and range in body size: pomarine jaeger (POJA, S. pomarinus, ~700 g [Wiley & Lee, 2020a]), parasitic jaeger (PAJA, S. parasiticus, Arctic Skua outside North America, ~450 g [Wiley & Lee, 2020b]), and long-tailed jaeger (LTJA, S. longicaudus, ~300 g [Wiley & Lee, 2020c]). Jaeger migrations connect land to sea, and the Arctic region to the tropics (Bemmelen, 2019; Gilg et al., 2013;Troy, 2007). Thus, as congeners that nest sympatrically across most of the Nearctic and then take long-distance migrations, jaegers could provide a model opportunity for studying Nearctic avian migration.
There are still many existing questions surrounding jaeger marine habitat use. Reviewing many studies of coastal and at-sea observations, Wiley and Lee (2020b) suggest that during the postbreeding period, PAJA occur more frequently in shallower coastal waters than LTJA (Wiley & Lee, 2020c) or POJA (Wiley & Lee, 2020a), but all species are found in both coastal and open ocean (Furness, 1987).

Previous results from light-level geolocator tracking of jaegers in
North America showed that LTJA from the eastern Canadian Arctic over-wintered off west and southern Africa (Seyer et al., 2021) and PAJA from eastern Greenland migrated to the Caribbean region (Bemmelen, 2019). A presumed-complete southward migration of a POJA from Alaska, USA, was recorded by satellite and this bird spent the postbreeding period off the coast of southeastern Australia (Troy, 2007). No full-annual-cycle migration data have been published for PAJA in North America west of Greenland (Bemmelen, 2019;Wiley & Lee, 2020b) or for any POJA globally (Wiley & Lee, 2020a).
The phylogenetic placement of POJA remains uncertain (Wiley & Lee, 2020a). Multiple lines of evidence suggest that POJA is more closely related to the large skuas that were once classified in a separate genus (Catharactus) than to LTJA and PAJA (Andersson, 1999;Chu et al., 2009;Cohen et al., 1997). However, POJA are more similar in size and plumage to the smaller congeners and the three species are sometimes thought of as a guild (Ruffino & Oksanen, 2014). Unlike LTJA and PAJA, POJA do not breed in Greenland or Europe (Furness, 1987;Wiley & Lee, 2020a), creating a break in their breeding distribution in Arctic areas immediately adjacent to the Atlantic Ocean.
Pomarine jaegers overwinter on both sides of the Atlantic Ocean (Brown, 1979;Starrett & Dixon, 1947) but the breeding origin of these POJA is unknown. Thus, the break in breeding distribution of POJA relative to LTJA and PAJA leads to questions about whether the three species spread across North America in the same direction(s) from the Palearctic region, their evolutionary origin, and the influences of both biogeography and evolution on contemporary migratory patterns.
Broadening the study of jaegers across the North American Arctic could help provide initial answers to these questions.
During the breeding season in North America, pomarine jaegers specialize on cyclic brown lemmings (Lemmus trimucronatus) for successful reproduction and thus POJA nesting is generally irruptive-occurring in high density at breeding sites in only some years (Andersson, 1973;Maher, 1974;Pitelka et al., 1955). This observation has implied that POJA are nomadic during the prebreeding period until they find localized areas of high lemming abundance (Wiley & Lee, 2020a), although no direct evidence for nomadic movements of individuals over large areas of the Arctic is known to be available.
We tracked the migrations of sympatrically-breeding jaeger species from a central Canadian high Arctic nesting location where both Atlantic Ocean and Pacific Ocean destinations seem equally likely.
Our goals were to: (1) describe the migratory routes and phenology of movements of the tracked jaegers following sympatric breeding; (2) describe the ocean habitats they used; and (3) provide the first direct information on full-annual-cycle movements and nomadism of a pomarine jaeger, an irruptive species whose movements are still largely a mystery (Wiley & Lee, 2020a). Like LTJA tracked from the eastern Canadian Arctic (Seyer et al., 2021), and LTJA and PAJA tracked from Greenland (Bemmelen, 2019), we expected that all three species would spend the nonbreeding period in the Atlantic Ocean. We also expected that PAJA would use shallow, coastal habitats and that LTJA and POJA would use deeper habitats in areas of consistent upwelling. Finally, we hypothesized the POJA would exhibit nomadism in the Canadian Arctic Archipelago prior to nesting and we did not expect the bird to exhibit nest site fidelity.

| ME THODS
We captured adult jaegers during incubation (late June to early July) 2018 and 2019 at Nanuit Itillinga (Polar Bear Pass) National Wildlife Area, Bathurst Island, Nunavut, Canada (NINWA, 75°43′N, 98°24′W). Birds were captured with spring-loaded bownet traps set at nests (n = 4), a handheld CO 2 powered net gun (n = 2), or noose mat (n = 1). We recorded morphometrics when possible (mass, wing chord, tarsus, bill, and total head plus bill) and fitted birds with a metal band and a color band to aid in identifying individuals.
We assessed wing and leg mobility prior to release and watched birds until they flew out of sight.
Tags were deployed in July, 2010 (n = 1), and June, 2011 (n = 1), and recovered the following year by recapturing the birds. Tags also recorded sea surface temperature (SST) when immersed for more than 120 s and stored the minimum daily value.
Area access and animal handling, banding, and tag attachment were approved by Environment and Climate Change Canada

| Tag programming and processing
Satellite tags were duty-cycled to maximize solar charging (10 h on, 48 h off). The satellite tag duty cycle resulted in nonregular location estimates during a 10-h window followed by a maximum gap of 48 h to allow for recharge of the battery via the solar panel. There were occasional gaps in transmissions ("missed duty cycle"). Positions received from Argos were preprocessed with a Kalman filter and delivered with an associated location class indicating the potential error radius and with an estimated error ellipse. Location classes 3, 2, and 1 have an estimated error of 250, 500, and 1500 m, respectively, while the accuracy of auxiliary location classes (0, A, B, and Z) is either variable or unbounded.
Given sampling irregularity and the telemetry error of position estimates, we used a model to estimate most probable paths. We applied the continuous-time random walk model of Jonsen et al. (2020) using the foieGras package in R and estimated movement paths at 24-h intervals to standardize sampling across birds tracked with different technologies (the maximum resolution of GLS tags was one position per day).
Light-level data from geolocator tags were initially processed using the manufacturer's built-in template fit algorithm to estimate locations (the raw light intensities were not stored by this tag model; only the processed position estimates and an estimated error were provided; Ekstrom, 2004). The template algorithm estimates a location once daily by fitting a model for a series of latitudes to light intensities recorded by the tag at a longitude estimated from the time of local noon using the tag's internal clock (Ekstrom, 2004). However, this method was shown to be biased south in winter and north in summer when applied to an Arctic seabird (Frederiksen et al., 2016). We therefore applied a sea surface temperature (SST) correction to further refine position estimates by comparing tagcollected SST measurements with remotely sensed SST data available for the same dates. We applied an unscented Kalman filter, a state-space model that incorporates measurement error estimation and the smoothing of the SST field directly in a single model to estimate the most probable track (Lam et al., 2008). We formulated the model with a "solstice" error structure to account for highly errone-

| RE SULTS
The three sympatrically nesting jaeger species migrated from central high Arctic Canada to postbreeding habitats in the Atlantic, Arctic, Indian, and Pacific Oceans (Figure 1). Birds departed for migration between July 22 and September 1 (Table 2). Birds were tracked to a maximum distance of 15,418 km straight-line distance from the nest site (Table 2), and for 236-887 days ( Table 1) However, the POJA migrated west across the Arctic Ocean, where it staged near Wrangel Island, Russia, before continuing to the western Pacific Ocean off Hokkaido, Japan (Figures 1 and 2).
Birds used overwintering habitats September-May (Table 2)  Our results of disparate migratory routes of sympatrically breeding congeners also raise additional questions about the evolutionary origins (Braun & Brumfield, 1998;Chu et al., 2009;Cohen et al., 1997) and biogeographic spread of North American jaegers. For  We grouped the movements off of Japan as a staging period because they proceeded a 5-month migratory loop over Micronesia and the bird also stopped in the Oyashio Current in the spring before returning to the Arctic. However, the movements off Japan could also be considered a first wintering area.  (Karl, 1999), and a place not typically mentioned as a primary overwintering habitat for the species (Furness, 1987;Wiley & Lee, 2020a). The deep water and low productivity habitats used by this POJA also contrasted with the individual Troy (2007)  This study is the first to record the full annual cycle of a pomarine jaeger (Wiley & Lee, 2020a). A presumed-complete southward migration path was reported by Troy (2007)  Distance between Canadian territories in consecutive years was similar to the mean breeding dispersal of 725 km of nine snowy owls (Therrien et al., 2014), although it is unknown whether the POJA in our study initiated a nest in its second year of tracking. In their review of POJA breeding phenology, Wiley and Lee (2020a) noted that most nesting territories in the North American Arctic were established by the third week of June, although some were not established until early July (Bathurst Island peak territory establishment 20-30 June). Maher (1974) observed that transients in Alaska (arriving from at-sea flocks) also occasionally established short-term terrestrial territories. Therefore, timing suggests that the POJA in our study arriving at Banks Island, Canada on June 23 could have initiated a nest, but it is equally plausible that these locations represented a transient terrestrial territory of a non-breeding bird.
Our study modified previous approaches to track jaegers with satellite tags (Seyer et al., 2021;Sittler et al., 2011;Troy, 2007). These modifications may have led to longer tracking durations than were previously attained (maximum 86 days for LTJA and approximately 275 days for POJA), although our small sample size limits general inferences. Rather than a backpack-style harness that loops over the wings (used previously with LTJA [Seyer et al., 2021;Sittler et al., 2011]), we used a leg-loop attachment (Mallory & Gilbert, 2008) as had been trialed for POJA (Troy, 2007). For acrobatic birds like jaegers, we felt a leg-loop harness would have the lowest risk of detrimental impact to the bird but may have a higher risk of being shed (sliding off the tail and legs of the bird possibly due to weight changes or interactions with other birds). For LTJA and POJA, we also used smaller tags than in previous studies: LTJA, 5 g instead of 9.5-10 g (Seyer et al., 2021;Sittler et al., 2011) and POJA, 9.5 g instead of 18 g (Troy, 2007). For LTJA and POJA, our 1% tag and harness to bird mass ratios were conservative in the context of conventional rules for seabird tracking studies (i.e., <3% of the body weight of the bird [Phillips et al., 2003]). To our knowledge, this was the first pilot of satellite tag and harness attachment with PAJA. The 9.5 g tag and harness combination we used for PAJA was <3% of their mass but a 5 g tag would likely have been a better choice due to weight but also due to its lower profile. We have subsequently tracked PAJA from Alaska through full annual cycles (5 of 6 birds) with 5 g tags (1% of body weight, unpublished data, Harrison A-L. 2021).
Since pelagic seabirds like jaegers spend their postbreeding period at sea, when a tag stops transmitting the reason is often unknown. In this study, the 9.5 g tag included an activity sensor to indicate whether a transmitting tag had stopped moving, but To evaluate tagging impacts, study designs that include marked but untagged individuals in the same breeding population to allow for comparison of return rates can provide additional understanding.
However, for species with low nesting site fidelity like POJA or high natural nest depredation as at Bathurst Island (making capture and resighting difficult), mark-recapture studies may yield few insights.

| CON CLUS IONS
The Arctic region is warming faster than most places on the planet (IPCC, 2014), and even closely related species may respond differently to environmental change Silva et al., 2020;Sun et al., 2017). Transformation of Arctic breeding habitats and disruption to Arctic food webs are thought to be major future threats (Gilg et al., 2012;Ims & Fuglei, 2005). Since lemmings and other small rodents depend for survival on good snow conditions in autumn/winter (Reid et al., 2012), lemming-reliant species like LTJA and POJA are at special risk of climate change; demonstrated a decline in LTJA in response to collapsing lemming cycles in Greenland.
For these three jaeger species breeding in sympatry in the Canadian Arctic, disparate patterns of nomadism and marine migratory connectivity may also have species-specific management implications (Dunn et al., 2019), although additional tracking is needed to confirm the patterns we observed from a small sample of individuals.
Trends of North American jaeger populations have been largely unknown (Gaston et al., 2009;Wiley & Lee, 2020a, 2020b, 2020c making it difficult to assess conservation status. Of the three jaeger species, PAJA is of conservation concern in some parts of its range. Globally PAJA is considered stable and is assumed to be the most abundant skua species in the world (Wiley & Lee, 2020b the coasts and the high seas to migratory seabirds (Beal et al., 2021;Harrison et al., 2018). Demonstrated links between marine biodiversity in the Arctic region and the high seas are timely to inform ongoing negotiations for an internationally binding legal instrument on the conservation and sustainable use of marine biological diversity in the areas beyond national jurisdiction (Popova et al., 2019;United Nations General Assembly, 2017;Vierros et al., 2020). Cowitz, J. Edwards, R. Johnston-Gonzalez, L. Montagano and the Nasaruvaalik Island field teams for field assistance. We thank four anonymous reviewers for helpful comments on our manuscript.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

O PE N R E S E A RCH BA D G E S
This article has earned an Open Data Badge for making publicly available the digitally-shareable data necessary to reproduce the reported results. The data is available at https://github.com/autum nlynn/ Sympa tricJ aeger s4Oceans. https://doi.org/10.5061/dryad.nk98s f7v1

DATA AVA I L A B I L I T Y S TAT E M E N T
The processed datasets resulting from this study (modeled to account for telemetry error and uncertainty associated with light-