INTRODUCTION
The Champlain Sea (CS), North America’s most recent inland sea, occurred in the St. Lawrence Lowland covering parts of the northeastern United States and southeastern Canada from about 13,000 to 10,600 cal BP (Figure 1) (Occhietti Reference Occhietti2007; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008), that is from about 11,100 to 9500 14C yr BP. Isostatic depression caused by the enormous weight of the Laurentide Ice Sheet (LIS) depressed the area leaving the St. Lawrence Lowland below sea level. Subsequent northward retreat of the LIS allowed the Atlantic Ocean to flood this then-below-sea-level area until the region rebounded, and seawaters retreated eastward. Sedimentological evidence shows a clear transgression-regression cycle and the transition from terrestrial to marine and back to terrestrial habitats (Gadd Reference Gadd1988; Prichonnet Reference Prichonnet1988; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Rayburn et al. Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011; Belrose Reference Belrose2015; Normandeau et al. Reference Normandeau, Lajeunesse, Trottier, Poiré and Pienitz2017). These sediments are also known to contain an abundant diversity of marine and terrestrial/lacustrine invertebrate and vertebrate fossils, including taxa such as bivalve mollusks, fishes, and whales (Harington Reference Harington1977; Harington Reference Harington1988; McAllister et al. Reference McAllister, Harington, Cumbaa and Renaud1988; Harington Reference Harington2003a; Harington et al. Reference Harington, Lebel, Paiement and de Vernal2006; Feranec et al. Reference Feranec, Franzi and Kozlowski2014).
Figure 1 Map of the location and extent of the Champlain Sea overlaying present-day geography. On inset, dark gray represents current waterways while light gray represents land inundated by water during the extent of Champlain Sea. Numbers refer to location of fossil specimens presented in this study as follows: 1, Saint-Nicolas, Québec, Canada; 2, Daveluyville, Québec, Canada; 3, Charlotte, Vermont, USA; 4, Mont-Saint-Hilaire, Québec, Canada; 5, Saint-Césaire, Québec, Canada; 6, Norfolk, NY, USA; 7, Plessisville, Québec, Canada.
Establishment of the CS could present a unique opportunity to study aspects of how marine ecosystems develop themselves anew. Additionally, detailed studies of this ecosystem may be informative with respect to the responses of marine animals to on-going climatic change as the CS spanned a period of intense global warming at the Pleistocene-Holocene transition (Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Steffensen et al. Reference Steffensen, Andersen, Bigler, Clausen, Dahl-Jensen, Fischer, Goto-Azuma, Hansson, Johnsen and Jouzel2008). Here, we present radiocarbon (14C) and stable isotope data from a set of marine vertebrates (i.e., mammal and bird) from different localities across the CS with the aim of determining (1) the timing of colonization by particular fauna and if certain taxa colonize before others, (2) whether the Champlain Sea marine ecosystem was different from modern ecosystems containing similar species, and (3) what the fauna imply about specific habitats in the Champlain Sea.
Timing the colonization of the CS requires knowing the age of its inception. Inundation of the St. Lawrence Lowland by marine waters is of course dependent on geography and geomorphology and likely not synchronous for all locations. Sites nearer the mouth of the St. Lawrence necessarily would be flooded earlier than sites to the west. While this commencement date has occasionally been reported using dates from mollusks which might be affected by marine reservoir effects (Rodrigues Reference Rodrigues1988; Parent and Occhietti Reference Parent and Occhietti1999; England et al. Reference England, Dyke, Coulthard, Mcneely and Aitken2013), Richard and Occhietti (Reference Richard and Occhietti2005) present an age of 11,100 ± 100 14C yr BP (12,770–13,170 cal BP) based on a terrestrial plant date and the regression rate of the LIS at Québec, Canada. This calibrated age range for the onset of the CS has been supported by additional studies (Rayburn et al. Reference Rayburn, Franzi and Knuepfer2007; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Rayburn et al. Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011).
The waxing and waning of marine waters in the CS also resulted in changes to marine habitats at particular locations over time. Factors such as isostatic rebound, influx of meteoric and/or glacial meltwater, and the geomorphology and depth of the site, for example, would affect what fauna could survive at a particular place and time (Hillaire-Marcel Reference Hillaire-Marcel1988). Recognizing this variability in marine habitats over time, this study, in part, explores whether vertebrate ecology in the CS was similar to or different from the known modern ecology of the same species.
14C dates on bone collagen have long been used to ascertain the ages of fauna from the CS (Dyck et al. Reference Dyck, Lowdon, Fyles and Blake1966; Harington Reference Harington1977; Lowdon and Blake Reference Lowdon and Blake1981; Harington Reference Harington1988; Harington Reference Harington2003a, Reference Harington2003b; Feranec et al. Reference Feranec, Franzi and Kozlowski2014). Similarly, the analysis of stable carbon (δ13C) and nitrogen (δ15N) isotope values has been well established as a technique to understand the ecology and ecological relationships of marine fauna (Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Fry and Wainright Reference Fry and Wainright1991; Hobson and Welch Reference Hobson and Welch1992; Clementz and Koch Reference Clementz and Koch2001; Newsome et al. Reference Newsome, Clementz and Koch2010). The stable isotope analyses in this study are expressed in standard δ-notation: X = [(Rsample/Rstandard) – 1] × 1000, where X is the δ13C or δ15N value, and R = 13C/12C or 15N/14N, respectively. δ13C values are reported relative to the V-PDB standard, and δ15N values are reported relative to atmospheric N2. Bone collagen δ13C values derive from the individual’s diet and variation in the values is largely controlled by primary producers at the base of food webs (Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Koch Reference Koch1998; Clementz and Koch Reference Clementz and Koch2001; Newsome et al. Reference Newsome, Clementz and Koch2010). In terrestrial ecosystems, photosynthetic pathway (i.e., C3, C4, or CAM), primarily controls stable carbon isotope values, with C3 having the lowest values, C4 the highest, and CAM in between (O’Leary Reference O’Leary1988; Ehleringer and Monson Reference Ehleringer and Monson1993; Koch Reference Koch1998). In marine ecosystems, primary producers show a very wide range of δ13C values (Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989). Factors controlling these values include the concentration of dissolved CO2 and HCO3–, and differences in productivity, for example (Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Clementz and Koch Reference Clementz and Koch2001). Differences in productivity can also lead to geographic differences in the δ13C values of primary producers in nearshore versus offshore habitats (Burton and Koch Reference Burton and Koch1999; Clementz and Koch Reference Clementz and Koch2001; Newsome et al. Reference Newsome, Clementz and Koch2010). Nearshore habitats generally have higher productivity, and thus primary producers having higher δ13C values than offshore habitats. However, this relationship can be complicated in nearshore glacial settings as the influx of meltwater can affect salinity, pCO2, and temperature, for example, all factors that would influence δ13C values at the base of marine food webs (Hillaire-Marcel Reference Hillaire-Marcel1988; Rau et al. Reference Rau, Takahashi and Marais1989, Reference Rau, Takahashi, Marais and Sullivan1991; Clementz and Koch Reference Clementz and Koch2001; Anderson et al. Reference Anderson, Levac and Lewis2007; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Xiao et al. Reference Xiao, Wang and Cheng2011; Calleja et al. Reference Calleja, Kerhervé, Bourgeois, Kędra, Leynaert, Devred, Babin and Morata2017).
δ15N values generally relate to trophic level, with higher values being found in individuals feeding higher in the food chain (Peterson and Fry Reference Peterson and Fry1987; Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Hobson and Welch Reference Hobson and Welch1992; Newsome et al. Reference Newsome, Clementz and Koch2010). We expect primary producers to have the lowest δ15N values and apex predators to have the highest (Peterson and Fry Reference Peterson and Fry1987; Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Hobson and Welch Reference Hobson and Welch1992; Van der Zanden and Rasmussen Reference Van der Zanden and Rasmussen1999; Clementz and Koch Reference Clementz and Koch2001; Newsome et al. Reference Newsome, Clementz and Koch2010).
Methods
Fifteen specimens representing 12 individuals of CS vertebrates were analyzed. Species include the following: Uria lomvia (thick-billed murre bird), Odobenus rosmarus (walrus), Balaenoptera physalus (fin whale), Delphinapterus leucas (beluga whale), and right whale (Eubalaena cf. glacialis). These specimens are housed in three different institutions: the Musée de paléontologie et de l’évolution (MPE; Montréal, Canada), the New York State Museum (NYSM; Albany, NY, USA), and the Perkins Museum of Geology (University of Vermont, Burlington, VT, USA). Samples were taken from museum specimens and submitted to the W.M. Keck Carbon Cycle Accelerator Mass Spectrometry Laboratory at U.C. Irvine, Irvine, CA, USA (UCIAMS) for both radiocarbon and stable isotope analyses. Sample processing at the UCIAMS follows standard procedures for extracting bone collagen and can be found in more detail in Beaumont et al. (Reference Beaumont, Beverly, Southon and Taylor2010), and on the laboratory’s website (UCIAMS 2021). Generally, UCIAMS uses a modified Longin (Reference Longin1971) method for collagen extraction (Brown et al. Reference Brown, Nelson, Vogel and Southon1988). Generally, the bone samples are first decalcified with acid, then treated with a weak base to extract possible contaminating humics, and finally the remaining organic collagen is hydrolyzed to gelatin at 60ºC in a weak acid. Afterwards, the gelatin solution is filtered to remove any remaining solids, and then ultra-filtered to remove the 30 kD fraction, which is then lyophilized. This freeze-dried sample is then graphitized and analyzed for 14C.
Stable carbon (δ13C) and nitrogen (δ15N) isotope data were obtained on splits of the bone collagen. These samples were analyzed separately on a Fisons NA1500NC elemental analyzer/Finnigan Delta Plus isotope ratio mass spectrometer. The δ13C and δ15N values have a precision of <0.1‰ and <0.2‰, respectively.
RESULTS
Radiocarbon Dates
The newly acquired radiocarbon dates range from 9965 ± 35 14C yr BP for the thick-billed murre bird (Uria lomvia) to 11,930 ± 30 14C yr BP in one of the fin whale (Balaenoptera physalis) specimens (Table 1). We highlight dates from two individuals (four specimens) below due to their historical significance.
Table 1 Locality information, fraction modern, conventional 14C date, calibrated 14C date, and calibrated date with marine reservoir offset correction from Champlain Sea fossils newly analyzed in this study.
a Locality: 1, Saint-Nicolas, Québec, Canada; 2, Daveluyville, Québec, Canada; 3, Charlotte, Vermont, USA; 4, Mont-Saint-Hilaire, Québec, Canada; 5, Saint-Césaire, Québec, Canada; 6, Norfolk, NY, USA; 7, Plessisville, Québec, Canada.
b Marine reservoir effect (ΔR) of 900 ± 262 from Richard and Occhietti (Reference Richard and Occhietti2005) using the Marine20 calibration in Calib 8.1.
Beluga Whale, Delphinapterus leucas (Charlotte, Chittenden County, Vermont, USA); The Charlotte Whale—The Vermont State Marine Fossil
Bone collagen. Sample taken from ventral side of the 7th lumbar vertebra of the Charlotte Whale.
Comments: A beluga whale was discovered in August 1849 during excavation of the Rutland and Burlington (Vermont, USA) railroad. The bones were found in blue clay about 8 feet (˜2.4 m) below the present-day surface. The bones were given to Zadock Thompson, a noted scientist of Vermont’s natural history, by engineers of the railroad. Upon receiving the bones, and to prevent crumbling, Thompson immersed them in “animal glue” (Thompson Reference Thompson1853). Thompson identified the specimen as a beluga whale, and later received confirmation of that diagnosis from Louis Agassiz of Harvard University (Cambridge, MA, USA). The stable isotope analyses on this specimen are similar to the other beluga whales included in this study (Table 2). This specimen was designated as Vermont’s state fossil in 1993. In 2014, the designation was amended that this specimen is Vermont’s state marine fossil. The Mount Holly Mammoth (Mammuthus primigenius), discovered by the same railroad company in 1848, was then designated as the State’s terrestrial fossil.
Table 2 Stable isotope values of Champlain Sea fossils.
Fin Whale, Balaenoptera physalus (Daveluyville, Québec, Canada); the Daveluyville Whale
Bone collagen. Sample taken from a vertebra of the Daveluyville Whale. MPEP578.1 was previously dated to 11,400 ± 90 14C yr BP (GSC-2871) using a 760g vertebra bone core sample (sample CR-78-21; Lowdon and Blake, Reference Lowdon and Blake1981), and this date was assigned for the whole Daveluyville Whale skeleton.
Bone collagen. Sample taken from a bone fragment of the Daveluyville Whale.
Bone collagen. Sample taken from the left mandible of the Daveluyville Whale.
Comments: In 1947, the nearly complete skeleton of the Daveluyville Whale was recovered from clay sediments about 3.6 km southwest of Daveluyville, Québec, Canada (Harington Reference Harington1977). The remains of this specimen are now housed in the Musée de paléontologie et de l’évolution (Montréal, Canada), having previously been housed in a museum in Trois-Rivières (Québec, Canada) in the 2000s, and before that in its original exhibition at the Université du Québec à Trois-Rivières (UQTR), from the late 1970s to early 1990s. This specimen is catalogued under accession number MPEP717 which includes 127 fragmentary to complete bones. Amongst these bones are a partial left posterior skull, a partial left mandible, the proximal part of right mandible, left scapula, right ulna, several ribs, several vertebrae, and many partial bones and bone fragments. Examination of the vertebrae of MPEP717 shows that the epiphyseal plates are not fused indicating that this specimen is likely a juvenile/sub-adult individual. The presence of unfused vertebrae is supported by examination of an early picture of the whale skeleton (i.e., Fig. 2 in Laverdière Reference Laverdière1950). Another specimen, a lone vertebra (MPEP578.1), ascribed as belonging to the Daveluyville Whale, differs from the other vertebrae for this individual in that its epiphyseal plates are fused. As noted above, this specimen was radiocarbon dated and the date was ascribed to the whole skeleton (Lowdon and Blake, Reference Lowdon and Blake1981). Because of the lack of fused vertebrae in MPEP717 and the fact that MPEP578.1 is fused, it would appear that MPEP578.1 does not represent the same individual as the specimens of MPEP717. The three dates obtained here are statistically significantly different from each other (t(2) = 14.2963, p < 0.003), with the date from the mandible being significantly older than the other two. These three new dates are also statistically significantly older than the previously obtained date from MPEP578.1 (GSC-2871; Lowdon and Blake Reference Lowdon and Blake1981), although the three most recent 14C dates, that is UCIAMS 182296, UCIAMS 182297, and UCIAMS 182298, are close enough that the statistical differences among the dated elements may not be meaningfully different in terms of the actual age of when this individual died.
Figure 2 Stable carbon and nitrogen isotope values from specimens analyzed in this study. Symbols: black diamonds, walruses; black circles, beluga whale; open circles, fin whale; asterisk, right whale; open triangle, thick-billed murre.
Stable Isotope Analyses
Stable isotope data are presented in Table 2 and Figure 2. The δ13C values for all individuals appear similar, ranging from –12.8‰ to –15.2‰, although when species with only a single individual are excluded from the dataset a statistically significant difference is observed between the beluga whales (
${\rm{\overline X}}$
= –13.4‰, σ = 0.6‰) and walruses (
${\rm{\overline X}}$
= –14.6‰, σ = 0.5‰) with belugas showing slightly higher values (t(8) = –3.251, p < 0.012).
For δ15N values, there does appear to be differences among individuals. The total range for all specimens in this study extends from 10.9‰ to 18.9‰, but there appears to be two populations of values, those below 14‰ and those above 16‰. The fin whale, right whale, and walruses show lower δ15N values, while the belugas and thick-billed murre have higher values. Again, when species with only a single individual are excluded from the analysis, a statistically significant difference is observed between the beluga whales (
${\rm{\overline X}}$
= 18.0‰, σ = 0.7‰) and walruses (
${\rm{\overline X}}$
= 12.0‰, σ = 0.6‰), with belugas showing higher values (t(8) = –14.205, p < 0.0001).
Discussion
Calibration of the dates using the Marine20 calibration curve places some of the older 14C-dated vertebrate specimens, that is, the Charlotte Whale, the Daveluyville Whale, and the right whale, beyond or just synchronous with the earliest date recognized for the inception of the CS (i.e., about 13,000 ± 200 cal BP) (Occhietti and Richard Reference Occhietti and Richard2003; Richard and Occhietti Reference Richard and Occhietti2005; Rayburn et al. Reference Rayburn, Franzi and Knuepfer2007; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Rayburn et al. Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011; Heaton et al. Reference Heaton, Köhler, Butzin, Bard, Reimer, Austin, Ramsey, Grootes, Hughen and Kromer2020; Reimer et al. Reference Reimer, Austin, Bard, Bayliss, Blackwell, Ramsey, Butzin, Cheng, Edwards and Friedrich2020). The use of “animal glue” to preserve the Charlotte Whale might cause hesitation in accepting the 14C date as accurate for this specimen, but the glue should pull the date to the more recent, not older (Thompson Reference Thompson1853; Takahashi et al. Reference Takahashi, Nelson and Southon2002). As was done for this specimen, a simple wash in distilled water and the standard preparation of bone collagen provides adequate removal of the animal glue for stable isotope analyses (Takahashi et al. Reference Takahashi, Nelson and Southon2002), it could be that the 14C date of the Charlotte Whale is even older than what was obtained here. Similarly problematic are the three dates obtained for the Daveluyville Whale. The date on the jaw of this specimen (MPEP717.2; UCIAMS 182298—11,930 ± 30 14C yr BP) is statistically significantly different from the other two, more recently obtained, dates. Further, these three new dates are statistically significantly different from the much earlier, non-AMS, date on this specimen (GSC-2871—11,400 ± 90 14C yr BP; Lowdon and Blake, Reference Lowdon and Blake1981). Morphologically, the contrast of the fused epiphyseal plates in MPEP578.1 with the lack of fusion in the vertebrae of MPEP717 likely indicates that these specimens belong to two different individuals. Unfortunately, the 14C dates and stable isotope data do not help in clarifying the number of individuals represented by the fossils. The δ13C and δ15N values are similar enough that they could derive from a single individual (Figure 2). However, the 14C dates are more problematic in aiding in this interpretation. The new, replicate date of MPEP578.1 (UCIAMS 182297) is statistically significantly older than the previously obtained date of this same specimen (GSC-2871). Confusingly, this date is statistically similar to the date obtained on MPEP717.109 (UCIAMS-182296), but statistically significantly younger than the date obtained on MPEP717.2 (UCIAMS-182298). As mentioned above, especially for the three most recently obtained dates, the dates are close enough that they may not be meaningfully different with regards to when the Daveluyville Whale was last alive. If the significant differences are meaningful, these data would then hint at aberrant preservation of collagen in this specimen, effects of pretreatments such as conservation glue, and/or complications caused by marine reservoir effects on the 14C values, for example.
Using the standard marine calibration curve (i.e., Marine20; Table 1), all three of these newest dates for the Daveluyville Whale calibrate outside the range or possibly just synchronous with the inception of the CS. With these specimens clearly coming from CS sediments, the calibration of the dates resulting in ages likely outside of the known range for the origin of the CS points to marine reservoir effects beyond the standard marine correction for the dated vertebrate specimens presented here, and as are common and long been known for the non-vertebrate fauna of the CS (Hillaire-Marcel Reference Hillaire-Marcel1988; Rodrigues Reference Rodrigues1988). Previous studies have shown marine reservoir effects and calculated offset corrections up to 1,780 14C years in the CS, but this is variable from locality to locality (Hillaire-Marcel Reference Hillaire-Marcel1988; Occhietti et al. Reference Occhietti, Hillaire-Marcel, Cournoyer, Cumbaa and Harington2001; Richard and Occhietti Reference Richard and Occhietti2005; Rayburn et al. Reference Rayburn, Cronin, Franzi, Knuepfer and Willard2011). Interestingly, studies have also shown local marine reservoir offsets different from the average marine correction (about 400 years) in more recent Arctic marine ecosystems (Dyke et al. Reference Dyke, McNeely and Hooper1996; Furze et al. Reference Furze, Pieńkowski and Coulthard2014). While offset corrections have been calculated for some CS localities, they should be utilized with caution, particularly when applying them to marine vertebrates. First, and in general for the CS, many of the offset corrections (ΔR) were determined using deposit feeding bivalve shells, which have been shown to present unpredictable offsets up to over 2000 years from sympatric suspension feeding bivalves (England et al. Reference England, Dyke, Coulthard, Mcneely and Aitken2013). Second, and more specific to this study, ΔR is generally calculated using bivalves and plants, taxa that are not mobile. This study specifically analyzes taxa that are mobile, have large home ranges, possibly migrate in and out of the CS, and have diets of organisms (e.g., cod, capelin) that are also mobile (Dyke et al. Reference Dyke, Savelle, Szpak, Southon, Howse, Desrosiers and Kotar2019). For the species analyzed here it is difficult to reconcile what specific offset should be used even if there is a known ΔR calculated for a particular locality at a given time. For example, Richard and Occhietti (Reference Richard and Occhietti2005) present a terrestrial plant date along with a date on a foraminifera from the same level at Lake Hertel for which we can calculate a potentially appropriate ΔR. The ΔR from these samples, using the Marine20 calibration, is 900 ± 262 years. Calibration of the dates using this correction places the fauna examined here squarely within the chronology of the CS, although the effects of the correction uncertainty (i.e., 262 years) on the calibration does not help in addressing the timing of colonization (Table 1). That is, most of the calibrations range about 1500 years—about half the existence of the CS. Critically then, the accuracy and precision in the radiocarbon dates on CS vertebrate fossil specimens is limited because of the uncertainty related to ΔR. Therefore, the 14C dates obtained in this and other studies on the CS marine vertebrate fauna may be influenced to varying degrees by localized marine reservoir effects (Harington Reference Harington1977, Reference Harington2003a; Feranec et al. Reference Feranec, Franzi and Kozlowski2014). Unfortunately, without knowing the marine reservoir offset for each specimen, we believe it is currently impossible to calculate the chronology of when particular species first colonized the CS. We provide these new 14C dates with the hope that a reservoir correction and the chronology of vertebrate colonization can be calculated in future studies.
The stable carbon (δ13C) and nitrogen (δ15N) isotope results are similar to those observed for the same species in modern arctic ecosystems (Hobson and Welch Reference Hobson and Welch1992; Hobson Reference Hobson1993; Dehn et al. Reference Dehn, Sheffield, Follmann, Duffy, Thomas and O’Hara2007; Hansen et al. Reference Hansen, Hedeholm, Sünksen, Christensen and Grønkjær2012; Marcoux et al. Reference Marcoux, McMeans, Fisk and Ferguson2012), implying that they reflect the known ecology of the analyzed species. The carbon isotopes show a small, but seemingly typical, range in values for cold water marine ecosystems of the Northern Hemisphere (Hobson and Welch Reference Hobson and Welch1992; Hobson Reference Hobson1993; Dehn et al. Reference Dehn, Sheffield, Follmann, Duffy, Thomas and O’Hara2007; Hansen et al. Reference Hansen, Hedeholm, Sünksen, Christensen and Grønkjær2012; Marcoux et al. Reference Marcoux, McMeans, Fisk and Ferguson2012). The δ15N values show a much wider range indicative of individuals feeding at different trophic levels (Peterson and Fry Reference Peterson and Fry1987; Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Hobson and Welch Reference Hobson and Welch1992; Newsome et al. Reference Newsome, Clementz and Koch2010).
Of the species studied here, the walruses, fin whale, and the right whale had the lowest δ15N and δ13C values. Modern fin whales (Balaenoptera physalus) have a diet heavy in krill (Borrell et al. Reference Borrell, Abad-Oliva, Gómez-Campos, Giménez and Aguilar2012; Vighi et al. Reference Vighi, Borrell and Aguilar2016; Aguilar and García-Vernet Reference Aguilar, García-Vernet, Würsig, Thewissen and Kovacs2018). Taking the diet to bone δ15N discrimination factor into account (˜+2.0‰ for fin whales; Borrell et al., Reference Borrell, Abad-Oliva, Gómez-Campos, Giménez and Aguilar2012), the average value obtained for the Daveluyville whale (+12.7‰) results in an expected krill value (+10.7‰) that fits within the range observed for different krill species in modern arctic ecosystems (Hobson and Welch Reference Hobson and Welch1992; Borrell et al. Reference Borrell, Abad-Oliva, Gómez-Campos, Giménez and Aguilar2012; Hansen et al. Reference Hansen, Hedeholm, Sünksen, Christensen and Grønkjær2012; Agersted et al. Reference Agersted, Bode and Nielsen2014; Vighi et al. Reference Vighi, Borrell and Aguilar2016). Similarly, right whales (Eubalaena glacialis) feed exclusively on zooplankton, largely on copepods and krill, taxa having lower δ15N and δ13C values, such that we would expect right whales and fin whales to be isotopically similar (Fry and Sherr Reference Fry, Sherr, Rundel, Ehleringer and Nagy1989; Hobson and Welch Reference Hobson and Welch1992; Hansen et al. Reference Hansen, Hedeholm, Sünksen, Christensen and Grønkjær2012; Kenney Reference Kenney, Würsig, Thewissen and Kovacs2018).
Similarly, for the walruses (Odobenus rosmarus), the δ15N values are not unexpected as bivalve mollusks are a preferred food of this species in modern ecosystems. Being suspension or deposit feeders, bivalves occur lower on the food chain than most fish (Hobson and Welch Reference Hobson and Welch1992). Additionally, bivalves appear throughout CS sediments as species such as Mya truncata, Macoma balthica, and Hiatella arctica are regularly found in the sediments alongside the vertebrate fossil specimens (Thompson Reference Thompson1853; Harington and Sergeant Reference Harington and Sergeant1972; Harington Reference Harington1977, Reference Harington1988; Lowdon and Blake Reference Lowdon and Blake1981; Steadman et al. Reference Steadman, Kirchgasser and Pelkey1994). Interestingly, although walrus generally consume bivalves, they occasionally eat a variety of prey including fish and even seals and whales—prey at higher trophic levels (Fay Reference Fay1982, Reference Fay1985; Kastelein Reference Kastelein, Perrin, Würsig and Thewissen2009).
Walruses, in general, forage for bivalves at depths of about 80 m or less, which would typically put them in more nearshore environments (Fay Reference Fay1982, Reference Fay1985; Born et al. Reference Born, Rysgaard, Ehlmé, Sejr, Acquarone and Levermann2003; Kastelein Reference Kastelein, Perrin, Würsig and Thewissen2009; Jay et al. Reference Jay, Fischbach and Kochnev2012). Additionally, many individuals utilize ice thick enough to support their weight as platforms to rest in between feeding excursions (Fay Reference Fay1982; Jay et al. Reference Jay, Fischbach and Kochnev2012). The availability of fast-ice or pack-ice is also supported by the presence of ringed seals (Pusa hispida) in the CS (Cournoyer et al. Reference Cournoyer, Chartier, Dubreuil and Occhietti2006; Feranec et al. Reference Feranec, Franzi and Kozlowski2014). As compared to the beluga whales in this study, the lower δ13C values observed in walrus bone collagen could be due to the influence of glacial meltwater on the marine isotope values or simply the result of isotope values in prey taxa that occur lower on the food chain (Rau et al. Reference Rau, Mearns, Young, Olson, Schafer and Kaplan1983, Reference Rau, Takahashi and Marais1989; Hobson and Welch Reference Hobson and Welch1992; Cronin et al. Reference Cronin, Manley, Brachfeld, Manley, Willard, Guilbault, Rayburn, Thunell and Berke2008; Hansen et al. Reference Hansen, Hedeholm, Sünksen, Christensen and Grønkjær2012; Calleja et al. Reference Calleja, Kerhervé, Bourgeois, Kędra, Leynaert, Devred, Babin and Morata2017). Further assessment of nearshore to offshore productivity is not possible to address as seasonal movements are not currently known for these CS species. Due to their foraging at depths expected nearer to shore, one might expect higher δ13C values in walruses compared to the other analyzed species. However, the low δ13C of walruses and fin whale in this dataset points to ecology and position in the food web as the primary factors controlling the observed values. An assessment of nearshore-offshore foraging on an individual level might be possible with serial sampling of walrus tusk fossils, for example.
As compared to the fin whale and the walruses, the thick-billed murre (Uria lomvia) and beluga whales (Delphinapterus leucas) had higher isotopic values. Today, the thick-billed murre, a relative of the now extinct Great Auk, regularly forages on capelin (Mallotus) and cod (Gadus) (Tuck Reference Tuck1961)—fish taxa that are also regularly found within CS sediments (McAllister et al. Reference McAllister, Harington, Cumbaa and Renaud1988). Similarity in isotopic values between the thick-billed murre and beluga whales is expected as belugas are also known to have a diet high in capelin and cod (Marcoux et al. Reference Marcoux, McMeans, Fisk and Ferguson2012). The δ15N and δ13C values of capelin and cod are higher than those found in krill and bivalves, which can account for the differences observed in our dataset between the lower isotopic values of walruses, fin whale, and right whale and the higher isotopic values of thick-billed murre and belugas. Additionally, these isotopic values suggest that the individual species’ ecology and ecosystem dynamics within the CS were very similar to those observed in modern arctic ecosystems.
Unfortunately, the inability to identify an appropriate ΔR to calibrate the 14C dates for analyzed fauna makes the assessment of the effects of global climate change on the ecology of CS fauna impossible at present. It is worth noting however, that while global climate shows significant changes during the span of the CS, regionally the climate stayed colder (Anderson et al. Reference Anderson, Levac and Lewis2007; Chapdelaine and Richard Reference Chapdelaine and Richard2017). This fact may help explain why the ecosystem is isotopically similar to modern arctic ecosystems.
Conclusions
The Champlain Sea occurred in the St. Lawrence Lowlands of the northeastern USA and southeastern Canada from about 13,000 to 10,600 cal BP. Inception and termination of the CS was caused by isostatic depression and subsequent rebound of this area, a result of the weight and regression of the Laurentide Ice Sheet. This marine sea contained an abundant diversity of invertebrate and vertebrate fauna. Radiocarbon dates acquired on 15 CS vertebrate specimens imply the influence of varied marine reservoir effects on individuals, possibly relating to their mobility, and make it impossible to calculate the chronology of colonization for different species as well as the effects of global climatic change on species’ ecology without knowing the specific marine reservoir offset corrections for each individual.
Stable isotope values in the CS are similar to those observed in modern arctic ecosystems implying similar ecosystem dynamics. Specific to this study, a fin whale, right whale, and walruses had lower δ15N values indicating consumption of prey species from lower trophic levels, such as krill, copepods, and bivalve mollusks, as they do today. Additionally, the presence of walruses and ringed seals implies that fast-ice and/or pack-ice must have been available for hauling out. Higher δ13C and δ15N values observed in the belugas and thick-billed murre imply feeding on prey at higher trophic levels. Cod and capelin are likely prey for the belugas and thick-billed murre in the CS ecosystem. These fishes are prominent in their modern diets and they are regularly found as fossils in CS sediments.
Acknowledgments
We would like to thank J. Southon for help in the sample preparation and analysis of specimens at UC Irvine, and for discussion of the effects of animal glue in the Charlotte Whale. We thank C. Mehrtens and R. Hopps at the University of Vermont for access to the Charlotte Whale. We thank Pierre-Henri Fontaine (Musée du squelette, Ile Verte, Québec, Canada) and Patrice Corbeil (GREMM, Tadoussac, Québec, Canada) for identifying specimen MPEP913.1 as a vertebra belonging to a Right Whale. We thank K. Feranec for help with the figures. And, we thank P.J.H. Richard, three anonymous reviewers, and the editors for reviewing and making comments and suggestions that improved this manuscript. Funding was provided by the NY State Museum (R.S.F.), and a USGS Statemap grant to A.L.K.
