Indication of Possible Shifts in Copepod Species Composition in St. Pauls Inlet, a Fjordal Estuary Connected to the Gulf of St. Lawrence

St. Pauls Inlet, a coastal brackish lake/estuary within Gros Morne National Park in Newfoundland and Labrador (Canada), is a semi-enclosed body of water with a narrow 80-m wide opening to the marine Gulf of St. Lawrence and with freshwater input from 24 streams. An initial biological survey of St. Pauls Inlet in 1977/78 during Park planning stages found dominant members of the mesozooplankton community to be copepods of genera Acartia, Calanus, Oithona, and Temora. Subsequent surveys of 2009, 2010, and 2019 found Calanus abundance to be much reduced or absent and low adult microcrustacean (copepods + cladocerans) densities. Recent data suggest that Calanus finmarchicus, a large energy-rich copepod, has declined across the Gulf of St. Lawrence and the wider Northwest Atlantic, with the zooplankton community shifting towards smaller, warm-water copepods This shift is coincident with rising surface sea-surface temperatures in the Atlantic. MidJuly temperatures in the inlet were higher in 2009, 2010, and 2019 compared with 1977. Shifting zooplankton abundances in St. Pauls Inlet, even with reduced exchange with the Gulf, may reflect wider trends in the Northwest Atlantic and be a useful indicator of a changing coastal environment in this national park.


Introduction
Many types of nearshore coastal waters are broadly estuarine in nature [1,2] with the intermingling of fresh and salt water. Such estuaries or brackish systems can be of great importance to a variety of ecosystems and may experience ecological alterations due to changing climate. Microcrustacean zooplankton such as copepods are key players in estuaries and other aquatic food webs, transferring energy from the microbial food web and primary producers up to higher trophic levels [3]. Although estuarine environments often have fewer species and thus lower biodiversity compared with marine or freshwater systems [4], those estuarine zooplankton that are able to tolerate the brackish salinities often show high and variable abundances and can support higher levels of tertiary (fish) productivity. Therefore, it is important to understand what type of zooplankton community is present within inlets and related estuarine systems, due to its potential to contribute to the larger marine ecosystem.
St. Pauls Inlet is a large brackish system on the west coast of the island of Newfoundland, Canada. This fjordal-type estuary [1,5] lies partially within Gros Morne National Park (GMNP) and is a semi-enclosed body of water with a narrow 80 m wide opening to the Northwest Atlantic. The basin receives freshwater input from 24 streams, the largest at the inlet head [6]. Esquiman Channel, the neritic feature of the Gulf of St. Lawrence closest to St. Pauls Inlet, has salinity concentrations ranging from 32-36‰ [7] and low nutrient concentrations compared to other regions of the Gulf [8]. The inflow of salt water into the inlet from the Gulf of St. Lawrence generally occurs only during the higher ranges of the tidal cycle. Investigation of zooplankton communities in coastal national parks such as GMNP is valuable as zooplankton can be useful bioindicators for tracking environmental changes in temperature and salinity [9].
An investigation was undertaken to determine whether St. Pauls Inlet could serve to track changes over time and correlate with environmental changes seen in the wider Atlantic. Scientific biological surveys of St. Pauls Inlet were first carried out during the summer of 1977 as well as in the spring of 1978 by Carter and MacGregor [10], following the establishment of GMNP in 1973. Zooplankton species composition and semi-quantitative abundances were determined at that time. There was no further exploration of the plankton community until we carried out our survey 31 years later in 2009 and 2010 (spring and summer of both years) [11] with additional sampling in July 2019. The purpose of our study was first to document any temporal differences in zooplankton species composition or relative abundance based on the four data sets and then evaluate likely causative environmental factors on a larger regional scale.

Materials and Methods
St. Pauls Inlet (Figure 1), located on insular Newfoundland's Great Northern Peninsula, is 11 km long and 6 km wide at the widest point. The surface area of the inlet is 34 km 2 with a maximum depth of 36 m and a mean depth of 13.6 m. The inlet can be classified as a positive fjordal-type estuary [1] with a tectonic overprint from glacial isostatic rebound [12]. There is a shallow (<6 m deep) sill at the entry, derived from a terminal moraine, a feature that marks the maximum advance of glaciation during the most recent ice age. The basin has characteristic steep-sided walls with relatively shallow outer portions exiting out into a low-lying coastal plain [6]. nutrient concentrations compared to other regions of the Gulf [8]. The inflow of salt water into the inlet from the Gulf of St. Lawrence generally occurs only during the higher ranges of the tidal cycle. Investigation of zooplankton communities in coastal national parks such as GMNP is valuable as zooplankton can be useful bioindicators for tracking environmental changes in temperature and salinity [9]. An investigation was undertaken to determine whether St. Pauls Inlet could serve to track changes over time and correlate with environmental changes seen in the wider Atlantic. Scientific biological surveys of St. Pauls Inlet were first carried out during the summer of 1977 as well as in the spring of 1978 by Carter and MacGregor [10], following the establishment of GMNP in 1973. Zooplankton species composition and semi-quantitative abundances were determined at that time. There was no further exploration of the plankton community until we carried out our survey 31 years later in 2009 and 2010 (spring and summer of both years) [11] with additional sampling in July 2019. The purpose of our study was first to document any temporal differences in zooplankton species composition or relative abundance based on the four data sets and then evaluate likely causative environmental factors on a larger regional scale.

Materials and Methods
St. Pauls Inlet (Figure 1), located on insular Newfoundland's Great Northern Peninsula, is 11 km long and 6 km wide at the widest point. The surface area of the inlet is 34 km 2 with a maximum depth of 36 m and a mean depth of 13.6 m. The inlet can be classified as a positive fjordal-type estuary [1] with a tectonic overprint from glacial isostatic rebound [12]. There is a shallow (<6 m deep) sill at the entry, derived from a terminal moraine, a feature that marks the maximum advance of glaciation during the most recent ice age. The basin has characteristic steep-sided walls with relatively shallow outer portions exiting out into a low-lying coastal plain [6].  Due to the restricted size (80 m wide) of the natural opening to the sea (St. Paul's Bay to the Gulf of St. Lawrence (GSL)), there can be a significant tidal velocity of 2 to 8 knots at the mouth of the inlet with estimated tidal amplitude of 0.6 to 0.9 m [10]. However, the tidal amplitude decreases rapidly further into the inlet meaning that most of the water body is essentially non-tidal.
In the initial zooplankton survey (methods outlined in Table 1), Carter and MacGregor sampled four open-water locations along the inlet ( Figure 1) once in summer (July 1977) and in spring (May 1978), as well as a site (Eastern Brook Beach) closer to shore. Water salinity and temperature were measured in the field with a Hydrolab meter. We sampled zooplankton at the same four open-water sites using both horizontal and vertical tows (Table 1 Mesoplankton species abundances were assessed qualitatively by Carter and MacGregor [10] through observations of the standing volume of organisms in vials of preservative. In our later surveys, zooplankton samples (pooled for replicated tows per site) were concentrated through a 25 µm filter and preserved within 4 h in 70% ethanol [14] in sterile scintillation vials. Samples were well-mixed and diluted to a known volume (20 mL) within the vial, then a 1 mL subsample was removed with a graduated pipette. Microcrustacean zooplankton were enumerated under a circular, rotating Plexiglass counting chamber at 250-500× magnification using a dissecting microscope. A minimum number of 200 individuals was counted in the samples. In some samples, it was not possible to get 200 individuals in the 1 mL subset so additional subsamples were removed until either 200 individuals or the full sample was counted. Contents of the vials were identified to the lowest taxonomic group possible using a variety of sources and dichotomous keys for cladocerans and copepods [4,[15][16][17][18][19][20][21][22][23].
In the 30+ year period since the 1979 study by Carter and MacGregor, there has been much research on copepod systematics and a number of species have been re-designated (see [20,23] to estimate the number of species expected in a random collection of individuals. Site similarity among surveys was defined statistically by grouping zooplankton samples according to their species composition, using presence/absence data which were available for all surveys. Similarity was assessed with Jaccard's coefficient of similarity and non-metric multi-dimensional scaling (NMDS) ordination was performed on the similarity matrix. Kruskal stress coefficient values, a type of goodness of fit that reflects how well the ordination summarizes observed distances among the samples [24], were examined to determine the stress level. NMDS analysis was carried out using NTSYSpc statistical software version 2.2 [25]; based on 100 iterations, the stress was type 2 and the 2D plot was chosen. An examination of the correlations of the axis values with environmental factors (bottom and surface salinity, bottom and surface temperature, depth) was done to determine what the axes might represent. Data from 2019 were not included in this classification as only one date was sampled.
Carter and MacGregor [10] ranked relative abundances of zooplankton species as Abundant, Common, or Infrequent based on settled volumes, an estimate of population density. We determined % species abundance for our zooplankton data for 2009, 2010, and 2019 based directly on population densities, setting Abundance as >70%, Common between 40-69%, Infrequent between 2-39%, and Rare as <1%. We calculated the % abundance of each species first relative to total microcrustacean numbers then just to total copepod numbers as copepods made up the majority of all microcrustacean zooplankton collected.
In order to account for seasonal variability in our analyses, we restricted comparisons of relative abundances and mean density to July dates, as all surveys (1977,2009,2010, and 2019) overlapped in having sampled during that month. Also, comparisons of relative abundances were made only for the four open-water locations that were consistent between surveys.

Species Richness
Species presence of microcrustacean zooplankton is shown in Table 2. The initial surveys from 1977/78 detected the presence of one cladoceran and 11 copepod species, while overall species richness in the later surveys consisted of two cladoceran and 7 copepod species (Table 2). Curves interpolated from the total number of individual microcrustaceans collected separately in 2009 and 2010 converged to asymptotes for both years (Figure 2). Adult microcrustacean densities (copepods + cladocerans, July data) ranged from 1.34 to 12.50 inds/L over the 4 sites, with the lowest densities observed at the Near Mouth site (Table 3). Most species sampled were Copepoda: densities of Cladocera (Evadne + Podon) averaged over the four sites were just 0.34 ± 0.17, 0.35 ± 0.16, and 0.15 ± 0.10 inds/L for 2009, 2010, and 2019, respectively.  [17].

Microcrustacean Species
Habitat Range SPI C and M

Cladocera
Evadne nordmanni Lovén, 1836 Brackish  Species composition differences were noted between survey periods. As seen in Figure 3, the NMDS ordination showed a distinct grouping separating the two time series. Axis I was significantly correlated with both surface and bottom temperature (0.663, 0.665, respectively), while Axis 2 showed correlations with surface temperature, bottom temperature, and site depth (0.860, 0.883, and −0.622, respectively). Neither axis was correlated with salinity. Mean surface salinity showed a consistent range between surveys, typically lower (~15‰) at the inlet head and higher (~21‰) at the mouth of the inlet.  Table 3. Relative abundance of the four dominant copepod species in July from the different survey periods. Carter and MacGregor in 1977 ranked species semi-quantitatively as abundant, common or infrequent. For our later surveys, we quantified relative copepod abundance as Abundant (>70% of total copepods), Common (40-69% of total copepods), Infrequent (2-39%) or Rare (<1%), based on a total of all copepod species found. Mean density (± standard deviation) refers to density of each species calculated from Vertical Tows, averaged over the 4 open-water sites. Density of adult microcrustaceans (copepods + cladocerans) refer to mean densities per site (inds/L; July data). Species composition differences were noted between survey periods. As seen in Figure 3, the NMDS ordination showed a distinct grouping separating the two time series. Axis I was significantly correlated with both surface and bottom temperature (0.663, 0.665, respectively), while Axis 2 showed correlations with surface temperature, bottom temperature, and site depth (0.860, 0.883, and −0.622, respectively). Neither axis was correlated with salinity. Mean surface salinity showed a consistent range between surveys, typically lower (~15‰) at the inlet head and higher (~21‰) at the mouth of the inlet.

Copepods
The dominant members of the zooplankton noted by Carter and MacGregor [10] in July 1977 were copepods Acartia, Calanus, Oithona, and Temora. The large Calanus finmarchicus was described as being "common to abundant at all locations and depths" ( Table  2). In contrast, the later July surveys of 2009, 2010, and 2019 found Calanus abundance to

Copepods
The dominant members of the zooplankton noted by Carter and MacGregor [10] in July 1977 were copepods Acartia, Calanus, Oithona, and Temora. The large Calanus finmarchicus was described as being "common to abundant at all locations and depths" ( Table 2). In contrast, the later July surveys of 2009, 2010, and 2019 found Calanus abundance to be much diminished with only one specimen being found in a horizontal tow in 2010 at the Mid-Inlet I site. Acartia, Oithona, and Temora were common across most sampling periods, with numerous individuals present in all samples. Johnson and Allen [21] list Acartia hudsonica and Temora longicornis as two of the species most commonly found in Atlantic coastal waters, being estuarine in nature.
Copepods comprised the majority of individuals counted (ranging from 77 to 96% of each sample) for all years. We computed relative abundances of the four dominant copepods as a percentage of the total number of all copepod species found (copepodites and nauplii excluded); data are shown in Table 3. The shift in relative abundance for Calanus finmarchicus is the most striking difference between the early and later surveys.
Carter and MacGregor did not provide density values. Table 3 shows mean density values for the four dominant copepod species from our surveys (July data). Densities of both Oithona and Temora were quite variable between years, with the small-bodied Oithona being most abundant in 2019; densities of the more abundant Acartia showed greater consistency between years.

Temperature
All of the surveys had temperature data for mid-July (period between 7-22 July) which we used as a comparison. Surface temperatures (site means with standard deviations) were 14.3 ± 0.84 • C, 16

Discussion
The mesozooplankton assemblage of St. Pauls Inlet can be characterized as estuarine and relatively depauperate. Adult microcrustacean abundances (based on July data only) seem within the lower range of those typically sampled by small-mesh nets in brackish regions of the Gulf of St. Lawrence [26]. Low mesoplankton abundance likely reflects low nutrient levels. There is little marine tidal input, and the rivers providing the freshwater influx also have very low levels of nutrients, with 0.07 ppm N measured nitrate and 0.01 ppm P measured phosphate [6].
The observed shift in species composition over a 30+ year period may be suggestive of ecological change, especially if this change consists of the almost complete absence of a previously dominant species such as C. finmarchicus. Estuarine species composition can be strongly influenced by temperature [3]. Yet given the paucity of zooplankton sampling from St. Pauls Inlet over the 30 years, it does become difficult to ascribe changes to any specific factor. However, what if observed changes in the inlet are indicative of wider regional ecosystem shifts? Indeed, recent reports from the Gulf of St. Lawrence and the Northwest Atlantic do show shifting zooplankton composition, with Calanus numbers particularly in decline [27,28].
Copepod development and spatial distribution are related to water temperature, with warmer waters more favorable for the population growth of smaller-bodied species which have lower respiration costs than larger-bodied zooplankton [29,30]. Debertin et al. [27] noted an increase in abundance of smaller zooplankton species such as Oithona that coincided with rising sea surface temperatures of 0.6 • C from 1993 to 2009 in Atlantic Canada's Northumberland Strait. Increases in water temperature may reduce hatching success in C. finmarchicus [30]. Some larger estuarine copepods such as Eurytemora affinis may find it difficult to maintain osmoregulation at warmer water temperatures [31]. E. affinis was found in St. Pauls Inlet by Carter and MacGregor [10] in early spring with colder water temperatures, and only once in summer (at Eastern Brook Beach). We did not find this copepod in any of our samples.
Large Calanus spp. are predominant members of the mesozooplankton community in the North Atlantic [22,28]. Warming sea surface temperatures (SST) may be implicated in the declining abundance of C. finmarchicus [28,32]. Sorochan et al. [33] found population levels of Calanus spp. to be lower than the long-term average since the 2000s or early 2010s throughout much of the western North Atlantic shelf system, including the Gulf of St. Lawrence (data from 1982-2016). Abundances of C. finmarchicus were found to be significantly negatively correlated with SST for the Gulf of Maine and the Scotian Shelf areas and with near-bottom temperatures for both the western and eastern parts of the Gulf of St. Lawrence [33].
Our study showed a slight increasing trend in surface water temperatures for St. Pauls Inlet, based on mid-July data. Galbraith et al. [34]  Larger spatial correlations between populations of C. finmarchicus and ocean temperatures in the Gulf and the western North Atlantic may be associated with indirect effects on changes in ocean circulation and transport of Calanus [33]. Carter and MacGregor [10] suggested that there was an intrusion of cold, oxygenated GSL sea water and nutrients into St. Pauls Inlet during winter and speculated that the inlet received an external supply of zooplankton from Gulf populations at this time. Increasingly warm sea temperatures could then indirectly limit the input of C. finmarchicus individuals into the inlet via a reduction in the supply of over-wintering diapausing copepodite stages and higher mortality resulting from subsurface warming.
Therefore, despite reduced exchange with the sea via its narrow opening, shifts in zooplankton species composition towards small-bodied copepods in St. Pauls Inlet may reflect climate changes in the wider ecosystem [28]. Without a more extensive sampling regime, however, this remains speculation. A related survey of near-shore fish in the inlet [36] also described an estuarine and depauperate fauna (15 fish species) with possible declines in abundance of marine species such as Atlantic Herring and Atlantic Cod. However, this survey only occurred in one month (August) and did not do the extensive gill-netting of Carter and MacGregor who recorded 23 fish species. Continued monitoring of the inlet can provide for an additional way to monitor and assess future changes within the Gros Morne National Park ecosystem as well as in the wider Gulf of St. Lawrence.
Author Contributions: Both authors contributed towards conceptualization, methodology, analysis, investigation; C.E.C. contributed to supervision, acquisition of funds and reviewing; E.N.S. wrote the original draft of the work and contributed to reviewing and data curation. All authors have read and agreed to the published version of the manuscript.