Experts’ opinions on threats to Leach’s Storm-Petrels ( Hydrobates leucorhous ) across their global range

. Seabirds are declining globally, though the threats they face differ among and within species and populations. Following substantial population declines at several breeding colonies, Leach’s Storm-Petrel ( Hydrobates leucorhous ) was uplisted from Least Concern to Vulnerable by the International Union for Conservation of Nature (IUCN) in 2016. Reasons for these declines are unclear


INTRODUCTION
Human activities have grown at unprecedented rates in recent decades, affecting all ecosystems (Marques et al. 2019, Baud et al. 2021. Marine ecosystems, in particular, are deteriorating, with degradation driven mainly by coastal development, offshore energy production, fisheries, and pollution (Halpern et al. , 2019. Perhaps it is not surprising, then, that declines in seabird numbers are more pronounced than in most other bird groups (Paleczny et al. 2015, Dias et al. 2019). Rodríguez et al. (2019) reviewed threats (defined as humaninduced or natural actions or events that negatively affect a species) for shearwaters and petrels (Procellariidae), highlighting that interspecific differences in threats are influenced by numerous behavioural, geographic, and life-history factors. Much attention has focused on population declines and threats to larger seabirds (Baker et al. 2002, Phillips et al. 2016, with fisheries bycatch being a major issue for divers (e.g., alcids and diving duck; Zydelis et al. 2013) and vessel-attracted scavenging species such as albatrosses (Anderson et al. 2011). For smaller species, such as storm-petrels, invasive mammals, especially rodents, are raising the most concern , Bolton et al. 2014, Dias et al. 2019, whereas fisheries bycatch is less of an issue (Bugoni et al. 2008, Jiménez et al. 2011; but see Pott and Wiedenfeld 2017). Approximately 44% of storm-petrel species are listed on the International Union for Conservation of Nature (IUCN) Red List as Near Threatened, Vulnerable, Endangered, or Critically Endangered, and data are insufficient to determine a threat category for an additional 7% of species (n = 27; BirdLife International 2021; Table A1.S1).
Leach's Storm-Petrel (Hydrobates leucorhous) was recently uplisted from Least Concern to Vulnerable globally (BirdLife International 2018), mainly as a consequence of population declines at northwest Atlantic colonies. The eastern (i.e., Atlantic) Canadian population was designated as Threatened by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC; COSEWIC 2020), the UK population was moved from Amber to Red by the UK Birds of Conservation Concern 5 (Stanbury et al. 2021), and the European population from Near Threatened to Vulnerable in the European Red List of Birds (Birdlife International 2021; Deakin et al. 2021). In general, storm-petrels are difficult to monitor and, therefore, colony abundance and population trends are either estimated infrequently or not available (Olivier and Wotherspoon 2006;Insley et al. 2014). However, Leach's Storm-Petrel is one of the most studied storm-petrel species, breeding from the Gulf of Maine to southern Labrador in the western North Atlantic, to Iceland, the British Isles, and coastal Norway in the eastern North Atlantic, and from Hokkaido to the Aleutians and Baja California in the Pacific . In western North Atlantic colonies, apparent adult survival rates for Leach's Storm-Petrels range from 79%-86% at a number of colonies, which is lower than expected for population stability (Fife et al. 2015, COSEWIC 2020, whereas limited data using a short time series from two Pacific Ocean colonies suggest a higher (94%-99%) apparent survival rate (Rennie et al. 2020). Concomitant with low adult survival rate, population sizes have been decreasing at several large Atlantic colonies (Wilhelm et al. 2015, 2019, 2019, d'Entremont et al. 2020, Deakin et al. 2021. Thus, reduced adult survival in western North Atlantic colonies could be a key driver of population decline. Following the listing of Leach's Storm-Petrel as Vulnerable by the IUCN and Threatened by COSEWIC, a list of threats was compiled (BirdLife International 2018, COSEWIC 2020). Yet, understanding the impact of threats is complicated by the enormous spatial range of Leach's Storm-Petrels throughout their life cycle. Leach's Storm-Petrels forage hundreds of kilometers from their colony during the breeding season (Halpin et al. 2018, Hedd et al. 2018 and are trans-equatorial migrants during the non-breeding season (Halpin et al. 2018. Given that this species is globally distributed across a seascape spanning a variety of threats and stressors, conservation planning must also consider spatial and temporal variation in risk. Current knowledge suggests that populations in the Pacific may be faring better than those in the Atlantic, potentially because of a different suite of threats in each ocean. Spatial variation is invaluable for assessing cumulative risk (but see Lieske et al. 2020), but it is often difficult to estimate given that multiple threats can influence ecosystems in a variety of ways. As such, solicitation of expert opinion is an effective approach to synthesize the current state of knowledge regarding sensitivity of marine birds to threats (Lieske et al. 2019). In this context, we created a survey with which Leach's Storm-Petrel researchers ranked threats for different colonies to assess spatial variation in risk, with the goal of supporting the selection of appropriate conservation strategies across the global range. We also identified knowledge gaps to be addressed to assist conservation of storm-petrels, and to inform conservation planning processes.

METHODS
The lead author contacted 48 Leach's Storm-Petrel researchers from nine countries or territories (Canada, Faroe Islands, Iceland, Japan, Mexico, Russia, St. Pierre et Miquelon, United Kingdom, and United States) covering the species' global breeding range to complete a survey (Appendix 1). The selected researchers have published articles on Leach's Storm-Petrels and were at different stages of their careers. The survey first asked researchers about their professional affiliations, how long they had been working with the species, and the location of the breeding colony(ies) where they conducted their research. Hence, threats that had caused extirpation of colonies were excluded from this survey. Participants were then asked to rank the importance of seven terrestrial and five at-sea threats (during both the breeding and non-breeding seasons; Table 1) at colony(ies) where they conduct Weather events Offshore light attraction and collisions Predation at offshore platforms Bycatch Prey depletion Spatial shift in prey items research or have in-depth knowledge. The threats were chosen based on the COSEWIC assessment and status report (COSEWIC 2020). For each ranking question, survey participants could propose additional threats not provided in the survey, and researchers were not required to rank all threats. It was assumed that, although most research on Leach's Storm-Petrels is colony-based, researchers could also evaluate at-sea threats.

DATA ANALYSIS
To determine which threat was the most important, weighted ordinal values were determined by the number of threat options in each question, with the top threat given the most weight. A top-ranked terrestrial threat was given a value of seven and a topranked at-sea threat was given a value of five. A lowest-rank terrestrial or at-sea threat was given a value of one. When a survey participant did not rank all proposed threats, we gave top values to the ranked threats and assumed the others were negligible (Table A1.S2). A score for each threat was calculated by summing all weighted values at three different geographic scales, as follows: global, ocean basin sector (western Atlantic, eastern Atlantic, western Pacific, and eastern Pacific), and jurisdiction (state, province, country). The number of participants was unequal among regions (jurisdiction and basin scales); therefore, score values varied greatly among regions. Hence, we determined for each threat what percent of the total weighted score it represented in each region.

RESULTS
Thirty-seven researchers (the authors) responded to the survey, representing eight of the nine countries where Leach's Storm-Petrels breed (we were unable to secure responses from Russia), with two, eight, twenty-four, and five responses for the western and eastern North Pacific, and the western and eastern North Atlantic, respectively. Most participants were government employees (n = 16), followed by university researchers (n = 15), but the group also included non-governmental organization employees (n = 5) and a museum employee (n = 1). More than two-thirds of participants (n = 26) had studied Leach's Storm-Petrels for more than 11 years and their combined experience represented a minimum of 433 years.

Terrestrial threats
Globally, survey participants ranked the three most important terrestrial threats that currently affect existing colonies as avian predators, mammalian predators, and onshore light attraction, with rankings differing among regions ( Fig. 1; Table 2). Avian predators were the top threat in all four ocean basin sectors, tied with mammalian predation in the eastern Pacific. At the jurisdiction level, avian predators had the highest or second highest summed weighted scores in all but one case, the Faroe Islands, where they ranked fourth. Mammalian predators were ranked first, second, or third in the eastern Pacific, western Atlantic, and eastern Atlantic, respectively, but were not perceived as a threat in the western Pacific. At the jurisdiction level, mammalian predators ranked among the top three threats in nine of fourteen cases (64%). They were not ranked as a threat in four jurisdictions (Japan, Mexico, St. Pierre et Miquelon, and Iceland). Onshore light attraction received the third highest weight, ranking third in the western Atlantic and eastern Pacific, but only fifth in the western Pacific and eastern Atlantic. At the jurisdiction level, onshore light attraction was ranked second in Newfoundland and Labrador, New Brunswick, California (tied with avian predation), and St. Pierre et Miquelon (tied with avian predation and habitat loss), third in Alaska, British Columbia, and Mexico, and fourth to sixth elsewhere (Fig. 1). There was a high consistency in the responses from the survey participants at the jurisdiction level. This concordance of expert opinion was not present at the basin or global scales (Table A1.S3). The heterogeneity in perceived threats is potentially the result of the combined effect of the different specific issues faced at their colonies and the perceptions of the survey participants.

At-sea threats during breeding season
The top four at-sea threats during the breeding season were offshore light attraction, spatial shifts in prey, mercury, and pesticides and other contaminants ( Fig. 2; Table 3). Offshore light attraction ranked first in the east Pacific, spatial shifts in prey ranked first in the east and west Atlantic, mercury ranked second in the west Atlantic and fourth in the eastern Pacific and the eastern Atlantic. Pesticides and other contaminants were either third (east Pacific, east Atlantic) or fourth (west Atlantic; Fig. 2). At the jurisdiction level, spatial shifts in prey items were ranked  (17) 14.6 (14) 9.3 (11) 6.9 (12) 7.5 (9) 0.0 (0) East Atlantic 30.5 (5) 19.1 (4) 8.6 (3) 8.6 (2) 20.9 (4) 11.4 (3) 0.9 (1) 0.0 (0) Overall 28.5 18.5 16.5 12.1 9.0 7.9 6.2 1.3 first for seven jurisdictions, offshore light attraction was ranked first for three jurisdictions, and mercury and pesticides and other contaminants each ranked first for one jurisdiction (Fig. 2). Survey participants were in agreement with the most important threats within each jurisdiction but not at the basin or global scale (Table A1.S4).

At-sea threats during non-breeding season
The top four offshore threats during the non-breeding season were the same as during the breeding season, namely offshore light attraction, spatial shifts in prey, mercury contamination, and pesticide contamination ( Fig. 3; Table 3). However, the order at the ocean-basin scale was somewhat different. Offshore light attraction ranked first as a threat in the east Pacific and west Atlantic but only seventh in the east Atlantic, where spatial shifts in prey items ranked first ( Fig. 3; Table 3). At the jurisdiction level, responses from survey participants were fairly consistent, with spatial shifts in prey items ranked first in five jurisdictions, offshore light attraction ranked first for four jurisdictions, mercury ranked first for one jurisdiction, and pesticides and other contaminants and prey depletion each ranked first for one jurisdiction. The agreement among survey participants was slightly more pronounced at the jurisdiction scale than at the basin scale (Table A1.S5).  (Massaro et al. 2000), such as storm-petrels (Stenhouse and Montevecchi 1999). However, at Atlantic colonies, gulls breed earlier than Leach's Storm-Petrels. When juvenile storm-petrels leave their burrows, gulls are not attending colonies, which presumably reduces this direct predation pressure on fledging individuals (Hoeg et al. 2021). However, there is considerable predation of fledglings by gulls during onshore wrecks and at illuminated coastal facilities (Burt 2022).

Avian predators
In eastern Canada, Herring Gull and Great Black-backed Gull populations decreased following the collapse of the northern cod fishery that provided gulls with a large food source from offal (Regular et al. 2013, Weseloh et al. 2020. We might therefore expect avian predation on storm-petrels to decrease, although some individual gulls specialize in feeding on storm-petrels (Pierotti andAnnett 1991, Hey et al. 2019). In skuas, specializing on seabirds and/or prey-switching from discards to seabirds (including storm-petrels) results in high levels of stormpetrel predation (Votier et al. 2006). Changes in regional fisheries policy could also affect those behaviors in the future (Votier et al. 2004, Bicknell et al. 2013).
Avian predation pressure may also be influenced by small mammal population dynamics on islands, where avian predators are initially attracted by the mammals but switch to seabirds once the nesting season begins. In cases where conservation strategies include removal of problematic mammals (e.g., mice, voles, and hares), there is potential for unintended consequences of predator diet-switching that inadvertently increases predation pressure on storm-petrels (Rayner et al. 2007, Hervías et al. 2013. Dynamics among avian predators and predation on storm-petrels and other prey may be complex and differ among regions and colonies (Stenhouse and Montevecchi 1999, Steenweg et al. 2011, Thomsen et al. 2018. In some cases, removal of just a few specialist avian predators has some positive effect on prey species (Sanz-Aguilar et al. 2009, Scopel andDiamond 2018). Where human activities have exacerbated avian predation, management strategies should be carefully considered. These strategies must also have clear objectives for both predator and prey, with an evaluation plan to assess if targets are met and the implementation of adaptive management when objectives are not reached (Libois et al. 2012, Fuentes et al. 2014, Bourgeois et al. 2015. In other cases of avian predation from a native predator, management might not be advisable.

Mammalian predators
Seabirds tend to breed on islands free of mammalian predators and, as a result, may lack anti-predator defense mechanisms (e.g., Buxton et al. 2016). Therefore, seabirds are highly vulnerable if mammalian predators are introduced to a colony (Borrelle et al. 2018), and they rarely coexist with introduced mammalian predators (De . Historical presence of introduced mammalian predators (mostly rats) has decimated storm-petrel colonies (McClelland et al. 2008), making mammalian predators the top threat for storm-petrels (Dias et al. 2019). Mammalian predators can also naturally occur on seabird islands, and measures to deal with naturally occurring mammals will be different than for introduced species. Co-occurrences of mammalian predators with storm-petrels will have different outcomes depending on the density of other species, and it is not necessarily a sustainable situation (Towns et al. 2006; but see Tuck 1987, Hammer andBond 2022 various periods of time and are potential predators of eggs, nestlings, and/or adults (Skelpkovych and Montevecchi 1996, Bicknell et al. 2009, Hoeg et al. 2021). For example, over 900 Leach's Storm-Petrels were found in larders of red foxes over six years on Baccalieu Island, Newfoundland and Labrador, Canada, the species' largest colony, which is now free of these predators (Sklepkovych and Montevecchi 1996). Red (Roesler et al. 2012). In contrast, introduced predators usually occur because of some accidental or deliberate anthropogenic intervention and may have greater impacts on seabird colonies. Introduced mammalian predators are a major global problem at seabird colonies (Dias et al. 2019), and both prevention and early detection of introduced predators should be of high priority because costs associated with At-sea threats during the non-breeding season West Pacific

Mammalian herbivores
The presence of mammalian herbivores is often the result of deliberate introduction by humans. Mammalian herbivores are a threat to Leach's Storm-Petrels through soil erosion and compaction, burrow destruction, and changes in vegetation composition and structure (Peterson et al. 2005). For example, in the western North Atlantic, the most commonly introduced species are snowshoe hares (Lepus americanus), historically introduced as food for island caretakers or as sources of pelts (Wheelwright 2016), and sheep (Ovis spp.), introduced for seasonal or year-round pasturing. White-tailed deer (Odocoileus virginianus) are not introduced in Nova Scotia but often swim to breeding colonies (I. Pollet, personal observation). Although they are herbivores, sheep and deer regularly consume eggs of groundnesting birds and have been documented occasionally biting off legs, wings, or heads of young seabirds (Furness 1988). However, the presence of mammalian herbivores is a far lower mortality risk to storm-petrels than is the presence of mammalian predators.

Onshore light attraction and collisions
Leach's Storm-Petrels travel to and from their colonies at night, presumably to avoid diurnal predators, especially gulls (Watanuki 1986. They likely use moonlight to visualize landscape cues, but it is not their only cue (Yoda et al. 2017, Wynn et al. 2020. Juveniles are especially attracted to light , Burt 2022) and, when onshore, birds are susceptible to collisions with buildings and vehicles, more prone to predation, and can become disoriented and grounded (Troy et al. 2013, Rodríguez et al. 2015. Effects of this threat depend on the proximity of a colony to onshore light sources and the type, color, and direction of lights (Miles et al. 2010, Rodríguez et al. 2017, Syposz et al. 2021). More than 1900 stranded Leach's Storm-Petrels were found on the island of Newfoundland during autumn 2018 and 2019 ), but these onshore stranding events have not been monitored systematically (but see Burt 2022); therefore, the total impact of onshore strandings is not well known. From a conservation perspective, onshore light attraction is probably one of the easiest threats to mitigate and reduce. For example, people living on island communities are invited to turn off their lights during peak fledgling seasons of endemic seabirds (https://web.archive.org/web/20221019115210/https://www.lesjoursdelanuit. re/; https://web.archive.org/web/20221019115223/https://birdlifemalta. org/wp-content/uploads/2020/07/Guidelines-for-Ecologically-Responsible-Lighting.pdf), and Kaua'i (Hawai'i, USA) residents are encouraged to place stranded shearwaters in "Shearwater Aid Stations" for care and safe release (Telfer et al. 1987). Such measures could complement information campaigns to reduce stranding events and rescue stranded birds (Ainley et al. 2001, Le Corre et al. 2003).

Disturbance
Recreational disturbance was considered a threat, and some survey respondents added researcher disturbance to the threat list. Effects of disturbance can be difficult to quantify. Nonetheless, visitors and people handling birds at colonies of burrow-nesting seabirds can negatively affect fledgling body mass, cause nestling mortality, or provoke nest abandonment by incubating or provisioning adults (Blackmer et al. 2004, Albores-Barajas et al. 2009, Watson et al. 2014. Active burrows can collapse from trampling. Some seabird colonies are difficult to access, which limits human disturbance, but some colonies, such as in the Faroe Islands, are situated near human settlements and experience seasonal influxes of tourists (A. Ausems, personal observation). Using paleo-ecological records, Duda et al. (2020) documented a severe decline in a North American colony once European settlers arrived in the area in the early 1800s, and the colony has yet to fully recover.
Recently, unmanned aerial vehicles (drones) have been used to limit disturbance from researchers surveying surface-nesting Breeding status of burrow-nesting birds can be assessed with burrow-scopes (i.e., endoscopic cameras), but human presence is still required within colonies to operate them (Surman and Nicholson 2009). Researchers are attempting to reduce disturbance (trampling, repeated grubbing) in study plots where long-term monitoring of population demography occurs by developing remote methods such as Passive Integrated Transponders (PIT) tag burrow-monitoring technology (Zangmeister et al. 2009;D. Fifield, personal communication) and acoustic monitoring (Orben et al. 2019).

Offshore light attraction and collisions
Lights at offshore oil and gas platforms are responsible for seabird strandings or collisions at structures or incineration from contact with flames (Wiese et al. 2001, Montevecchi 2006, Ronconi et al. 2015Davis et al. 2017). In Nova Scotia and Newfoundland and Labrador, Canada, offshore structures occur within the foraging ranges of major Leach's Storm-Petrel colonies (Hedd et al. 2018), and Leach's Storm-Petrels are by far the most numerous species reported stranded, with 6920 individuals reported between 1998 and 2018 (87% of all reported birds), which undoubtedly is an underestimate of the total numbers stranded (Gjerdrum et al. 2021). Leach's Storm-Petrels breeding at Gull Island, Newfoundland and Labrador, flew within the light catch-basin of an oil platform in 17.5% of trips, although they tended to transit rapidly past platforms during the day (when light attraction may be minimal), whereas exposure to oil platforms at night occurred in only 1.1% of trips but represents a very large number of individual trips considering the number trips each bird takes during a season (Collins et al. 2022).
Lighting at offshore operations is important for worker safety and for maritime and air navigation, so mitigation strategies must not increase human risks. In Canadian offshore industries, 76% of stranded storm-petrels are found alive (Gjerdrum et al. 2021), presenting opportunities for mitigation through search and release. In addition to offshore oil and gas platforms, offshore lights can also originate from emerging offshore wind facilities, vessels in transit or working offshore, as well as cruise and cargo ships. Squid-fishing fleets in particular use powerful lights to attract squid to the surface (Waluda et al. 2004), and the location and timing of squid-fishing season can coincide with either the breeding or the migration of storm-petrels (McIver et al. 2016). The extent and magnitude of strandings on vessels is largely unknown with respect to potential population-level impacts.

Spatial shifts in prey and climate change
Spatial shifts in seabird prey can be an indirect consequence of climate change. Changing environmental conditions, such as rising sea surface temperature, can induce a spatial shift of marine species because they gravitate to their preferred thermal preferences Boulinier 2009, Kleisner et al. 2016). During marine heat waves, some Leach's Storm-Petrels can temporarily shift their feeding habits and forage on or near shore (D'Entremont et al. 2021). During breeding seasons, seabirds in general are central place foragers and thus limited in their foraging range by the necessity to return to their colony (Elliott et al. 2009). The foraging range of Leach's Storm-Petrels is 400-800 km depending on the colony (Hedd et al. 2018, Mauck et al. 2022, which is a great distance considering they only weigh ~45g. If preferred prey species shift outside of maximum viable foraging range, seabirds must switch to alternative prey species that might not provide sufficient nutrition or extend their foraging trips, with the associated energy expenditure potentially leading to breeding failure (Ponchon et al. 2014, Fayet et al. 2020. Some dietary shifts in Leach's Storm-Petrels have been observed at various timescales (Hedd et al. 2009, Fairhurst et al. 2015. Ecological impacts of these dietary shifts are not clear, though they may be related to decreases in breeding success . Some survey participants included prey depletion and climate change as additional threats, and it is difficult to differentiate spatial shifts in prey from prey depletion. Extreme weather events were not provided as a threat category in the survey but were added by several participants. Severe storms during the breeding season may flood burrows and drown or chill chicks. Moreover, the start of fall migration and the fledging of storm-petrel chicks coincides with the peak of the hurricane season in the Atlantic, and hurricanes and onshore winds may force storm-petrels onshore (Boyd 1954, Teixeira 1987).

Mercury
Mercury is a globally distributed, toxic metal that, at sufficient concentrations, can have negative effects on neuroendocrine systems, lead to reduced reproductive success, and induce motor and behavioral problems (Wolfe et al. 1998, Scheuhammer et al. 2015, Evers 2018. In aquatic environments, mercury is transformed by bacteria into methylmercury (Lehnherr 2014). Methylmercury biomagnifies in food webs, and, during periods of stress or poor body condition, may be released from tissue reserves at concentrations that create physiological, physical, or behavioural problems (Fort et al. 2015). Storm-petrels, and Procellariiformes in general, have some of the highest tissue mercury levels among seabirds (Carravieri et al. 2014, Becker et al. 2016. Mercury exposure also varies spatially, and in the Gulf of Maine, Leach's Storm-Petrels have high mercury concentrations compared to other sympatric seabirds (Goodale et al. 2008, Bond andDiamond 2009). Mercury exposure in the western North Atlantic, particularly in deep offshore waters, appears to be high for some species, including Leach's Storm-Petrels (Goodale et al. 2008N. Burgess et al. 2019, personal communication) and Little Auks (Alle alle), another planktivore (Fort et al. 2014). Mercury has been considered a potential problem for Leach's Storm-Petrels for decades (Pearce et al. 1979, Elliott et al. 1992), but to date there is little evidence of deleterious effects (Pollet et al. 2017, submitted;Krug et al. 2021).
In recent years, reductions of mercury emissions in some regions (North America and the European Union) and industrial sectors (e.g., energy production) have been encouraging yet, in 2015, global emissions were still 20% higher than in 2010 (UN Environment Programme 2019). Increasing mercury emissions and legacy mercury already in soil, sediments, and aquatic systems will continue to produce methylmercury for millennia; therefore, monitoring concentrations in biota should continue.

Pesticides and other contaminants
This broad category could include hydrocarbons, trace elements, hydrophobic persistent organic pollutants, and plastics. Most of these substances are poorly studied in Leach's Storm-Petrels, but, in Atlantic Canada, Elliott et al. (1992) found high levels of selenium and cadmium in Leach's Storm-Petrels relative to other sympatric seabirds. However, the high levels of metallothionein that Leach's Storm-Petrels produce endogenously may enable them to limit effects of high tissue concentrations of heavy metals (Osborn 1978, Elliott et al. 1992. Organic compounds and organochlorines also vary spatially in concentration (Megson et al. 2014) and were at low concentrations in eggs of Fork-tailed Storm-Petrels (Hydrobates furcata) and Leach's Storm-Petrels in Alaska, with PCB concentrations below toxicity thresholds (D. D. Rudis and B. L. Slater 2009, personal communication). None of these contaminant studies evaluated toxicological effects of measured concentrations.
Like many seabird species, Leach's Storm-Petrels ingest plastic (Bond and Lavers 2013). Krug et al. (2021) found high frequencies of occurrence (87.5%) of plastic debris in recently fledged stormpetrels in Atlantic Canada, and when using emetics (which recover all plastic in the gastrointestinal tract), 48% of adults sampled at Gull Island, Newfoundland and Labrador had also ingested plastic (Bond and Lavers 2013). Leach's Storm-Petrels spend most of their time at sea in contact with the ocean surface, either to feed or rest. This makes them vulnerable to hydrocarbons discharged from vessels or offshore platforms and to more voluminous oil spills (Fraser et al. 2006, Wilhelm et al. 2007, Ellis et al. 2013, Morandin and O'Hara 2016.

Bycatch
The unintentional killing of seabirds in fishing gear (seabird bycatch), causes significant global mortality for medium-large tubenose species (such as albatrosses and shearwaters) and divers (e.g., alcids and diving ducks), killing hundreds of thousands of individuals per year (Anderson et al. 2011, Croxall et al. 2012, Zydelis et al. 2013, though bycatch appears rare for Leach's Storm-Petrels (Hedd et al. 2016, Jannot et al. 2021. Presumably this is related to their foraging mode and dietary preferences and a lack of strong attraction to vessels, fishing bait, or prey in nets.

KNOWLEDGE GAPS AND PRIORITIES FOR FUTURE RESEARCH
Survey participants agreed that predation and light attraction represent major population threats to Leach's Storm-Petrels across their breeding range, with perhaps less importance in the eastern North Atlantic. With relatively easy access to colonies by researchers or predators (compared to following birds at sea), documenting predation seems an obvious and useful task, and carcasses collected at colonies can provide clear evidence Shutler 2019, Hoeg et al. 2021). Because storm-petrels spend approximately 90% of their life at sea , there is significant impetus for quantifying impacts of threats away from colonies. Offshore light attraction appears to be an important threat, but the combined ranking of climate change effects (the spatial shift and depletion in prey items and the weather events) would rank higher than light attraction in many jurisdictions. Low apparent adult annual survival and population declines at several colonies sharing the same wintering areas suggest that threats during the non-breeding season are highly important and are areas of active investigation. Some participants were unable to identify at-sea threats, but that does not mean threats are not present, and, in the Pacific, more research is required to evaluate the population status and to identify threats faced by Leach's Storm-Petrels ( Figs. 1 and 2). Threats at sea are more difficult to observe, and the effects of some threats may not be independent of other threats (e.g., Tartu et al. 2013). Increasingly, however, development of miniature tracking technology is enabling researchers to study at-sea distribution and foraging ranges (Pollet et al. 2014a, 2014b, 2019, Hedd et al. 2018 and to begin identifying threats to storm-petrels in pelagic environments (Collins et al. 2022). Studying Leach's Storm-Petrels during the non-breeding season would involve international collaboration because of the widespread distribution of the species during that time (Halpin et al. 2018. Differences in perceived threats in different jurisdictions and the agreement of survey participants within each jurisdiction but not within each basin (Tables A1.S3, A1.S4, and A1.S5) highlight the need to carefully tailor conservation measures in a contextdependent way. Terrestrial threats are probably colony-specific, and conservation measures implemented at a global scale might have different results depending on the colony.
In the western North Atlantic, tracking studies show little overlap in foraging areas for Leach's Storm-Petrels breeding at adjacent colonies (Hedd et al. 2018), so threats encountered at sea during the breeding season by birds from one colony might differ from those of a neighboring colony. The spatial distribution of risk varies in the seascape, and compared to other seabird species in eastern Canada, Leach's Storm-Petrels have high cumulative risks from threats, which include marine traffic, light pollution, and ship-source oil pollution (Lieske et al. 2020). During migration, birds from different colonies within the same ocean basin may follow similar routes, thereby being exposed to the same threats during this portion of their annual cycle . Similarly, the degree of migratory connectivity in over-wintering areas may provide common, or divergent, exposure to risk at both the individual and colony levels (González-Solís et al. 2007, Frederiksen et al. 2012, 2016.
Seabirds have a diverse range of foraging strategies, body sizes, and diets. Thus, threats will affect each species or guild differently. For example, Lieske et al. (2019) concluded that Leach's Storm-Petrels had the highest species-specific risk score for sensitivity to light pollution in the western North Atlantic (i.e., most sensitive of all seabird species in the study). In contrast, in the same study, Leach's Storm-Petrels were ranked least sensitive to fisheries bycatch. Survey participants in the present study ranked offshore lights as the top offshore threat (Table 3). In our study, each threat was considered separately, but, of course, threats are not mutually exclusive and can have additive, synergistic, or antagonistic effects (Crain et al. 2008, Piggott et al. 2015, Dias et al. 2019, Lieske et al. 2020. For example, in a given breeding colony, Leach's Storm-Petrels could suffer from both predation and difficulties finding enough prey to feed their chicks. The current assessment of spatial variation in threats affecting Leach's Storm-Petrels could be applied to some other stormpetrel species living sympatrically because they share many lifehistory traits. This includes European Storm-Petrels (Hydrobates pelagicus) in the eastern North Atlantic and Fork-tailed Storm-Petrels, Ashy Storm-Petrel (H. homochroa), and Least (H. microsoma) and Black Storm-Petrels (H. melania) in the Pacific , Halpin et al. 2018, Ausems et al. 2021, Bedolla-Guzmán et al. 2021. In cases where a conservation planning framework is developed to implement and monitor conservation initiatives, spatial variation in threats should be considered to direct conservation actions and biosecurity measures that are most relevant to a particular colony or population (Russell et al. 2008). In this way, cumulative threats can be mitigated on a caseby-case basis, contributing to an overall conservation strategy of cumulative actions that maximize positive benefits to the global Leach's Storm-Petrel population.   Southern stormpetrels 9 2 (22) 5 (56) 0 (0.0) 0 (0) 1 (11) 1 (11) Total 27 2 (7.5) 13 (48) 3 (11) 4 (15) 3 (11) 2 (7.5) Table A1.S2. Example of hypothetical weighted score calculations with five threats and two survey participants. Participant 1 scored threat B the highest, so the score is the maximum out of the five threats = 5. Participant 2 scored only four of the five threats. For each threat, scores from both participants are added (i.e.: score = 9 for threat A). The sum of the scores is 29. The percent score for threat A is 9/29 (i.e.: 31.03%).