Multi‐decadal trends in biomarkers in harp seal teeth from the North Atlantic reveal the influence of prey availability on seal trophic position

Arctic food webs are being impacted by borealisation and environmental change. To quantify the impact of these multiple forcings, it is crucial to accurately determine the temporal change in key ecosystem metrics, such as trophic position of top predators. Here, we measured stable nitrogen isotopes (δ15N) in amino acids in harp seal teeth from across the North Atlantic spanning a period of 60 years to robustly assess multi‐decadal trends in harp seal trophic position, accounting for changes in δ15N at the base of the food web. We reveal long‐term variations in trophic position of harp seals which are likely to reflect fluctuations in prey availability, specifically fish‐ or invertebrate‐dominated diets. We show that the temporal trends in harp seal trophic position differ between the Northwest Atlantic, Greenland Sea and Barents Sea, suggesting divergent changes in each local ecosystem. Our results provide invaluable data for population dynamic and ecotoxicology studies.


| INTRODUC TI ON
Marine ecosystem-based management is crucial for effective marine conservation and resource management (Pikitch et al., 2004).
Top predators can act as indicators of ecosystem change, reflecting the status of the ecosystem on which they depend (Sydeman et al., 2015). Trophic position of marine top predators, which can be used as an expression of food chain length, is a fundamental property of ecological communities, reflecting direct and indirect effects of changes in food web dynamics (Post, 2002a;Post & Takimoto, 2007). Trophic position, reflecting the trophic ecology of predators, is also utilised in population dynamic studies, as body condition and reproduction of marine predators are linked to their diet (Arnould et al., 2011;Øigård et al., 2013;Ronconi et al., 2014).
In addition, trophic position is one of the key drivers of contaminant burdens (Borgå et al., 2004;Braune et al., 2015;Brown et al., 2016;Carravieri et al., 2014) and trophically derived parasite loads (Couch et al., 2022;Vestbo et al., 2019) in predators. It is therefore crucial to accurately detect decadal changes in trophic position of marine predators in order to understand how food webs are being reshaped and the consequences of these modifications on population dynamics and contaminant bioaccumulation, which could, in turn, be used in ecosystem-based management.
Here, we examined multi-decadal trends in trophic position of a generalist near-top marine predator, the harp seal (Pagophilus groenlandicus), across three locations in the North Atlantic. The North Atlantic food web is under a period of tremendous changes, including ocean warming, change in ocean circulation and change in fisheries, impacting all trophic levels from phytoplankton to predators (Meredith et al., 2019). The Barents Sea, Greenland Sea and Baffin Bay connect the Atlantic to the Arctic Ocean and have undergone several shifts in fish and zooplankton abundance and species composition resulting from fishing and changes in environmental conditions (Dalpadado et al., 2016;Dempsey et al., 2017;Fossheim et al., 2015;Nöthig et al., 2015;Oziel et al., 2020;Pedersen et al., 2017Pedersen et al., , 2020. These changes in community structure have resulted in changes to predator-prey interactions in recent decades (Dwyer et al., 2010;Haug, Bogstad, et al., 2017;Kortsch et al., 2015;Pecuchet et al., 2020;Stenson, 2012), and are likely to indirectly impact marine top predators, such as harp seals, through changes in foraging ecology (e.g. Hoover et al., 2013;Kovacs & Lydersen, 2008;Laidre et al., 2008).
Stable nitrogen isotopes are commonly used as a biomarker to reconstruct food webs and estimate trophic position of predators. The ratio between heavy ( 15 N) and light ( 14 N) stable nitrogen isotopes in bulk tissue (hereafter δ 15 N Bulk ) undergoes a fractionation between each trophic level estimated to be on average 3.4 ± 0.2‰ (Vander Zanden & Rasmussen, 2001), providing a measure of trophic position (Post, 2002b). However, δ 15 N Bulk values in consumers are heavily dependent on δ 15 N at the base of the food web, or baseline. Nitrate, underpinning the baseline of marine food web (Mariotti et al., 1981), is supplied to the Arctic Ocean by Atlantic water entering through the Barents Sea and Fram Strait, and by Pacific water crossing the Bering Strait (Torres-Valdes et al., 2013). The δ 15 N values of nitrate in Atlantic water are depleted in 15 N compared to the δ 15 N values of nitrate in Pacific water and water masses present in the high Arctic, as a result of the biological processing within the Atlantic, and Pacific oceans de la Vega et al., 2020;Knapp et al., 2008;Marconi et al., 2015;Somes et al., 2010;Tuerena, Hopkins, et al., 2021). The Arctic is experiencing changes in ocean circulation, including a simultaneous increase in the rates of Atlantic and Pacific inflows (Polyakov et al., 2020;Woodgate, 2018). Thus, spatial trends in water mass influence, alongside temporal changes in circulation, lead to a heterogenous and evolving δ 15 N baseline across the Arctic and sub-Arctic, that needs to be accounted for to reliably detect changes in trophic position of predators de la Vega et al., 2020de la Vega et al., , 2022.
Compound-specific stable nitrogen isotopes of amino acids (δ 15 N AA ) can overcome this issue. δ 15 N AA is a powerful biomarker approach that disentangles baseline and fractionating trophic effects when using δ 15 N values to estimate trophic position. The δ 15 N of the 'source' amino acid (AA) phenylalanine (δ 15 N Phe ), can conservatively trace the δ 15 N at the baseline as it experiences minimal fractionation (<1‰) during trophic transfer (McMahon & McCarthy, 2016). The δ 15 N of 'trophic' AAs (δ 15 N Trophic ), for example, glutamic acid, undergoes significant fractionation (>3‰) resulting in 15 N enrichment between each trophic step (McMahon & McCarthy, 2016). This approach simultaneously fingerprints both the δ 15 N baseline and trophic information in a predator from the analysis of predator tissue alone. This allows accurate estimation of changes in relative trophic position while accounting for variation in δ 15 N baseline, using baseline-corrected δ 15 N Trophic (corδ 15 N Trophic ; de la Vega et al., 2020).
In this study, we used harp seals as an indicator species of changes in the North Atlantic sub-Arctic food web (Laidre et al., 2008). Harp seals are the most abundant pinniped in the North Atlantic . They are found in three populations, in the Barents and White Seas (hereafter, Barents Sea population), in the Greenland Sea (hereafter, Greenland Sea population) and in the Northwest Atlantic off Newfoundland and Labrador (hereafter, Northwest Atlantic population, Figure 1; Carr et al., 2015;Sergeant, 1991). Harp seals undertake substantial annual migrations between sub-Arctic breeding and moulting areas in winter and spring, and Arctic feeding grounds in summer and autumn (Figure 1; F I G U R E 1 Map depicting the seal sampling sites, their population range, the areas from which the fish data were taken, and the main ocean circulation. Folkow et al., 2004;Lacoste & Stenson, 2000;Nordøy et al., 2008;Stenson & Sjare, 1997). Harp seal diets consist of a wide variety of fish and invertebrates, with substantial seasonal and geographical variations in their diet composition associated with changes in local prey availability and abundance (e.g. Bogstad et al., 2000;Enoksen et al., 2017;Nilssen et al., 1998;Stenson, 2012). Seal diet is therefore expected to be a strong indicator of variation in prey availability and changes in fish communities. Here, we analysed δ 15 N AA in harp seal teeth from 1949 to 2012, from the Barents Sea, the Greenland Sea and the Northwest Atlantic (Figure 1) to assess multi-decadal trends in harp seal relative trophic position across the North Atlantic sub-Arctic.
To relate temporal trends in prey availability and harp seal trophic position, we collated published temporal data on biomass of seal's prey species, including fish and invertebrates, from the Barents Sea and Northwest Atlantic. We predicted that: (1) δ 15 N Phe values, reflecting the δ 15 N baseline, will vary temporally reflecting changes in water mass circulation occurring across the North Atlantic, and (2) corδ 15 N Trophic , representing the relative trophic position of harp seals, will vary temporally reflecting changes in prey species availability across the North Atlantic.

| Seal sampling
Teeth of harp seals from the Barents Sea (n = 72), Greenland Sea The teeth were prepared following the method described in Kershaw et al. (2021). In brief, the teeth were sectioned along two planes: transverse and sagittal using a precision low speed diamond saw (Buehler, Isomet™). The transverse sections were used to identify the different annual growth layer groups (GLGs) based on the structure, width and opacity of individual layers (Bowen et al., 1983). Each GLG corresponds to 1 year of life of the individual. A 700 μm sagittal section was cut as close as possible to the central plane of the tooth and demineralised with 0.25 M HCl for between 12 and 24 h. Once softened, any remaining gum tissue and cementum were cut away from the outer edge of the tooth. The dentine samples representing the individual GLG for the second (GLG2) and third years (GLG3) of life of the seal individuals were freeze-dried and stored in plastic vials prior to stable isotope analysis. Material from GLG2 and GLG3 was combined (GLG2 + 3) in order to provide sufficient material for compound-specific stable isotope analyses for each individual (Kershaw et al., 2021).
Thus, each data point represents the isotopic signal averaged over the second and third year of life of each individual. Knowing the sampling year and the age of the individual, we were able to assign specific years represented by GLG2 and GLG3 for each individual (Supporting Information S1). GLG2 + 3 of teeth from the Barents Sea and Greenland Sea represented years from 1949-1950 to 2011-2012 and from 1945-1946 to 2007-2008 respectively (

| Stable isotope analyses of seal teeth samples
For δ 15 N Bulk , ~0.5 mg of sample was precisely weighed (±1 μg) and sealed in a tin capsule. For δ 15 N AA , ~15 mg of dentine was hydrolysed and propylated. Amino acid methyl esters were then acetylated, dissolved in dichloromethane and stored at −20°C until analysis. A detailed outline of the method is provided in Supporting Information S1.
All seal δ 15 N analyses were carried out at the Liverpool Isotope Facility for Environmental Research (LIFER), University of Liverpool, and reported in standard δ-notation (‰) relative to atmospheric N 2 (Mariotti, 1983). δ 15 N Bulk was determined using a Costech elemental analyser coupled to Thermo Fisher Delta V isotope ratio mass spectrometer (IRMS). Samples were corrected using international reference materials USGS40 and USGS41a, analysed throughout each run with a reproducibility of <0.1‰. An internal standard of ground prawn (Penaeus vannamei) with a well characterised δ 15 N value (6.8‰) was analysed every 10 samples to monitor precision, which was <0.2‰. δ 15 N AA were determined using a Trace Ultra gas chromatograph (GC) coupled to a Thermo Fisher Delta V Advantage IRMS with a ConFlo IV interface. A liquid nitrogen trap was added after the reduction oven to remove CO 2 from the sample stream. The separation of AA was achieved using a HP Innowax capillary column (Agilent). Samples were analysed in duplicate. A mixed standard prepared from eight AAs with known δ 15 N values (University of Indiana and SI Science) was used for calibration and analysed every four injections. Details of methods, precision, accuracy and δ 15 N values of all identified AA are provided in Supporting Information S1. TA B L E 1 Stable nitrogen isotopes analyses and sampling site for harp seal teeth; Ranges of the sampling years and ages of seal individuals, and year(s) represented in the tissue analysed for δ 15 N; n = number of analysed samples (=data points); All δ 15 N values represent the years assigned to GLG2 and GLG3, corresponding to the foraging at 2 and 3 years old for all individuals.

| Statistical analyses
Statistical analyses were performed in R version 3.5.1 (R Core

| Fish biomass
To qualitatively compare the relative temporal variation of fish biomass and seal trophic position, we collated estimates of fish biomass from the Barents Sea, and fish and invertebrate biomass from the Northwest Atlantic from published sources.  (Johannesen et al., 2012). These species are important prey species for harp seals in the Barents Sea Nilssen, Haug, Potelov, Stasenkov, et al., 1995;, and the interactions between capelin, cod and herring are key processes in the Barents Sea ecosystem (Hamre, 1994;Haug, Bogstad, et al., 2017;Ushakov & Prozorkevich, 2002;Wassmann et al., 2006). R 2 = 43.1%, Figure 2a; Table 2). There was no temporal trend in  Table 2).    Table 4).

| Temporal variation of fish biomass
In the Barents Sea, the biomass of capelin, dominating the pelagic schooling fish stock, fluctuated with time, with peaks of biomass in 1975, 1980, 1991, 2000    The Barents Sea, which is the main gateway between the Arctic and adjacent Atlantic Ocean in the northeast, is experiencing changes in the rate of Atlantic water inflow, nitrogen sources and in situ biogeochemical processes, influencing the δ 15 N baseline. The range of the Barents Sea harp seal population is restricted within the Barents Sea (Haug et al., 1994;Nordøy et al., 2008), harp seals migrating between their breeding and moulting areas in the southern Barents Sea in winter and spring, and Arctic feeding grounds in the northern Barents Sea in summer and autumn (Haug et al., 1994;Nordøy et al., 2008). Harp seals from the Greenland Sea gather into large and dense aggregations on the pack ice off the east coast of Greenland during the breeding and moulting periods in winter and spring (Folkow et al., 2004). The Greenland Sea has a complex hydrography hosting multiple water masses from different origins (Brakstad et al., 2019;. In addition to changes in water mass influence, other factors, such as increasing primary production and in turn benthic denitrification (Arrigo & van Dijken, 2015), or increasing bacterial activity with increasing temperature (Vaqué et al., 2019) that tend to increase the δ 15 N baseline (Granger et al., 2011;Morata et al., 2008), might also have influenced the δ 15 N Phe in harp seals from these areas. These results suggest that migratory predators, as they forage within different food webs, assimilate spatially varying isotopic baselines, which complicates the use of single-site baselines that are common in anal-

| Multi-decadal trends in harp seal trophic position
Harp seal diet consists of various prey items, including invertebrate and fish species. During summer and autumn, their diet is dominated by invertebrate species such as krill (Thysanoessa spp.) and sea iceassociated amphipods (e.g. Themisto libellula), and Arctic cod Lindstrøm et al., 2013;Nilssen et al., 2000;Ogloff et al., 2019). During winter, forage fish, such as Atlantic herring, Arctic cod and especially capelin, are the primary prey, although the proportion of capelin in the diet varies among years reflecting the local abundance of other prey species, such as Atlantic cod, Greenland halibut, haddock, sand eels (Ammodytes sp.), sculpins, redfish (Sebastes spp.), gadoids, mysids and shrimp Nilssen et al., 2000;Nilssen, Haug, Potelov, Stasenkov, et al., 1995;Ogloff et al., 2019). In general, demersal species such as Atlantic cod, Greenland halibut and haddock are at a higher trophic position, and have higher δ 15 N values than schooling fish such as capelin and Atlantic herring, which in turn are at a higher trophic position and have higher δ 15 N values than crustacean species such as amphipods and shrimp (Hansen, Hedeholm, et al., 2012;Haug et al., 2021;Haug, Falk-Petersen, et al., 2017;Lawson & Hobson, 2000;Sherwood & Rose, 2005;Tamelander et al., 2006). However, fish species often  (Dalpadado et al., 2016;Fossheim et al., 2015;Nöthig et al., 2015;Oziel et al., 2020). A change towards an invertebrate-dominated diet during years of low fish abundance might explain the low cor- reflect a diet more dominated by fish (Figure 5a). This is consistent with the higher contribution of amphipods, krill and shrimps in the diet of Atlantic cod in the late 1980s and early 1990s compared to early 2000s (Holt et al., 2019). Overall, the harp seal trophic position was higher during periods of species-rich fish community, that is, in the mid-1970s and mid-2000s when both the demersal and pelagic stocks were high (Figure 5a).
In the Northwest Atlantic, the diets of harp seals before the 1990s were dominated by fish . The slight increase in corδ 15 N Trophic values in harp seals in the mid-1980s could reflect a change from a diet dominated by capelin in the early 1980s, to a diet dominated by Arctic cod in the mid-and late 1980s Stenson, 2012). The decrease in corδ 15 N Trophic values in harp seals after the late 1980s ( Figure 2f) coincides with a shift from an ecosystem dominated by demersal fish species, to an ecosystem dominated by crustaceans such as shrimp and snow crab that occurred in the late 1980s and early 1990s in the Newfoundland and Labrador shelves (Figure 5b; Dawe et al., 2012;Koen-Alonso & Cuff, 2018;Pedersen et al., 2017;Stenson et al., 2020). This shift in community structure, concomitant with the decline in biomass of capelin ( Figure 5b) and multiple demersal species, such as Atlantic cod (Buren et al., 2019;Buren, Koen-Alonso, Pepin, et al., 2014), resulted from a history of overfishing combined with environmental change (Buren, Koen-Alonso, Pepin, et al., 2014;Stenson et al., 2020). A change from a fish-dominated diet to an invertebrate-dominated diet could explain the lower cor- sea ice associated amphipods (e.g. Themisto libellula), and Arctic cods in summer and autumn in Arctic feeding grounds, but fish, such as capelin, dominate their diet in winter and spring in sub-Arctic breeding and moulting areas (Enoksen et al., 2017;Folkow et al., 2004;Lacoste & Stenson, 2000;Lawson & Stenson, 1997;Lindstrøm et al., 2013;Nilssen, 1995;Nilssen, Haug, Potelov, Stasenkov, et al., 1995;Nordøy et al., 2008;Ogloff et al., 2019;Stenson, 2013 Sea (Enoksen et al., 2017;Haug et al., 2002;Potelov et al., 2000). Overall, these results suggest that harp seals modify their diets to adapt to change in prey availability Øigård et al., 2013), ploit local resources when these are concentrated, even when the fish or invertebrate stock overall is low Lawson et al., 1998;Lindstrøm et al., 1998Lindstrøm et al., , 2013Marshall et al., 2010).
Long time series on every important prey species, at relevant spatial scales, are therefore needed to improve our understanding about how changes in prey availability influence harp seal diet and in turn trophic position.
The plasticity in harp seal diet could render them less sensitive to borealisation (Clavel et al., 2011)  Arctic cod and negative preference for amphipods and krill , and the nutritional content of prey can have an impact on harp seal body condition with consequences for reproduction (Frie et al., 2003;Øigård et al., 2013;Stenson et al., 2016Stenson et al., , 2020.
For example, in the Barents Sea harp seal population, the mean age at sexual maturity increased and the growth rate decreased from the 1960s to the early 1990s (Frie et al., 2003;Kjellqwist et al., 1995) coinciding with indications of reduced body condition (Frie et al., 2003), and corresponding to the low relative trophic position of harp seals in the late 1980s-early 1990s (Figure 3a). The age at sexual maturity in Greenland Sea harp seals did not vary from the 1960s to the early 1990s, and was lower in the late 1980s-early 1990s than for Barents Sea harp seals (Frie et al., 2003). This suggests that Greenland Sea harp seals had access to higher quality prey items (Frie et al., 2003), and is consistent with the absence of  (Figure 3c). This supports that the temporal variations in body condition and reproduction rates of harp seals from the Barents Sea and Northwest Atlantic observed in the last decades may be associated with changes in prey availability, in addition to other factors such as population density and environmental conditions, for example, ice cover (Frie et al., 2003;Stenson et al., 2016Stenson et al., , 2020. These results could be included in studies of harp seal population dynamics to improve their predictive power given the known importance of trophic position and diet composition for body condition and reproduction. Population models that can include wider ecosystem effects have much potential to suggest future directions of change for predator populations (Smout et al., 2022).
In addition, these results are highly valuable for ecotoxicology studies, as trophic position is a key factor influencing contaminant bioaccumulation (Borgå et al., 2004). This study provides a unique 60-year record of harp seal trophic position across the North Atlantic. We thank Jim Ball for his help in the LIFER laboratory in Liverpool University.

CO N FLI C T O F I NTE R E S T S TATE M E NT
We declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available in