Species composition and habitat preferences of the nearshore fish fauna of Bonne Bay, Newfoundland

Study area

Bonne Bay is a fjord shaped by multiple glacial events in the geological past (Shaw 2003). There are three main sources of freshwater draining off the boreal forested watershed into the bay: Deer Brook, Southeast Brook and Lomond River (Fig. 2). A shallow (14 m) sill near Norris Point separates the outer basin and South Arm from the inner basin composed of East Arm and Deer Arm (Fig. 2). A current-swept channel across the sill has a maximum depth of 26 m. The maximum depth of East Arm is 230 m (Shaw 2003; Richards and de Young 2004).

Oceanographic data collection

Water column profiles of seawater temperature and salinity were recorded at each site before deployment of fish collection equipment (Table 1). Salinity and temperature were measured using a YSI® Model 30 M-100FT water quality probe. Bottom depth was measured using a Piranha® Max 15 dual beam depth sounder. In June 2004 conductivity-temperature-depth (CTD) profiles at stations along the main axis of East Arm and South Arm were recorded using a Seabird® Electronics Inc. SBE-25 CTD (see Figure 43, 44, 49 and 50 in Richards and de Young 2004). Moorings with 300 kHz acoustic doppler current profilers (ADCP) were deployed between September 2002 and September 2006 on the sill to monitor temperature and salinity and the velocity of the water flowing across the sill (Richards and de Young 2004; de Young et al. 2005).

Methods of collecting fishes

To overcome sampling gear inefficiency and selectivity, several types of fish collection equipment were deployed (Methven et al. 2001; Hemingway and Elliott 2002). A 10 m wide by 1.5 m high beach seine with 10 mm stretch mesh throughout was used to sample intertidal waters along the shoreline at all seven sites with a tow length of 25 m. A larger beach seine measuring 25 m by 1.5 m in the wing, with 10 mm stretch mesh netting and a 5 mm stretch mesh in the cod-end, was deployed approximately 50 m from the shoreline using a boat, and pulled towards the shore by two groups of people about 17 m apart. Sampling with the 25 m beach seine was restricted to Site 2, Site 2A, Site 3, Site 4 and Site 4A (Table 2). Two gillnets were also deployed. A three-panel gillnet with stretch mesh sizes of 2.6 cm (1 in.), 3.8 cm (1.5 in.) and 5.1 cm (2 in.) was deployed in shallow water and a single panel gillnet with stretch mesh size of 7.6 cm (3 in.) was deployed in deeper water, but not exceeding a depth of 100 m. Gillnets were set during the evening, fished overnight and retrieved the following morning. Finally, a 4.9 m semi-balloon bottom trawl (5.1 m head rope and 6.4 m foot rope) with nylon twine netting (38 mm stretch mesh, with a 9 mm stretch mesh liner in the cod-end) was used to sample Site 1, Site 3 and Site 4A (Table 2). Bottom trawl sampling was conducted during daylight hours with an approximate 10 min tow duration.

Analyses of fish collections

Species richness at sampling sites

Fish species richness was highest at Site 3 (Nsp = 27). The species richness was somewhat lower at Site 2 (Nsp = 21), Site 1 (Nsp = 20), and Site 4 (Nsp = 20). The lowest species richness was observed at sites with lower sampling effort: Site 2A (Nsp = 14), Site 4A (Nsp = 13), and Site 3A (Nsp = 7). Of the 29 species of fishes collected by sampling, two species were present at all seven sites: winter flounder (Pseudopleuronectes americanus) and threespine stickleback (Gasterosteus aculeatus) (Table 3). Five species were present at all sampling sites except for the barachois at Site 3A: cunner (Tautogolabrus adspersus), brook trout (Salvelinus fontinalis), American sandlance (Ammodytes americanus), grubby (Myoxocephalus aenaeus), and shorthorn sculpin (Myoxocephalus scorpius). Three species were present at all sampling sites except for Site 2A: Atlantic salmon (Salmo salar), rainbow smelt (Osmerus mordax) and blackspotted stickleback (Gasterosteus wheatlandi). Atlantic herring (Clupea harengus) was found at all sites except for Site 4 and Site 4A at the inner end of East Arm. Atlantic cod (Gadus morhua) and longhorn sculpin (Myoxocephalus octodecemspinosus) were present at all sites except for the two low salinity sites barachois at Site 3A and the river delta at Site 4A. Five species were found only at the river delta at Site 3: silver hake (Merluccius bilinearis), Atlantic tomcod (Microgadus tomcod), alligatorfish (Aspidophoroides monopterygius), windowpane flounder (Scophthalmus aquosus), and yellowtail flounder (Limanda ferruginea). Northern pipefish (Syngnathus fuscus) was found only at the sampling sites with eelgrass beds (Zostera marina), specifically Sites 3, 4 and 4A. Radiated shanny (Ulvaria subbifurcata) was found only at Site 2 (Table 3).

Comparison of fish assemblages among sampling sites

Seventeen species of the total 29 species were observed in the intertidal zone as revealed by the sampling with the 10 m beach seine (Table 4). Individuals caught by the 10 m seine were mostly juvenile stages. Cluster analysis of the fish collections made with the 10 m beach seine indicated that spatial proximity did not explain species dissimilarity across sites (Fig. 3). Site 2A was most dissimilar (J _D  = 0.75) from the other sites, because only two species, brook trout (Salvelinus fontinalis) and threespine stickleback (Gasterosteus aculeatus), were collected there. This might be due to the steep bottom slope resulting in a narrow intertidal zone. Sites 3A and 4A cluster together and are dissimilar (J _D  = 0.39) from Sites 1, 2, 3 and 4. This is because Sites 3A and 4A were shallow low salinity environments with seven species in common (Table 4), including fourspine stickleback (Apeltes quadracus), a species found only in low salinity environments in Bonne Bay. Species from the families Gadidae, Pholidae, Ammodytidae, and Cottidae were not present at low saline intertidal Sites 3A and 4A, while they were present at Sites 1, 2, 3, and 4.

Species \ Sites	1	2	2A	3	3A	4	4A
Clupea harengus					x
Salmo salar	x			x	x	x	x
Salvelinus fontinalis	x	x	x	x	x	x	x
Osmerus mordax	x			x	x		x
Gadus morhua	x	x
Gadus ogac	x
Urophycis tenuis	x	x		x		x
Gasterosteus aculeatus	x	x	x	x	x	x	x
Gasterosteus wheatlandi	x	x		x	x		x
Apeltes quadracus				x	x		x
Tautogolabrus adspersus	x	x		x		x	x
Pholis gunnellus	x	x		x		x
Ammodytes americanus	x	x
Myoxocephalus aenaeus	x			x		x
Myoxocephalus octodecemspinosus	x	x		x
Myoxocephalus scorpius	x	x		x		x
Pseudopleuronectes americanus	x	x		x	x	x	x

Individuals from 24 species were collected using the 25 m beach seine (Table 5). Cluster analysis showed that Sites 3 and 4 were the least dissimilar (J _D  = 0.2, Fig. 4). These two sites had 17 species in common including sea raven (Hemitripterus americanus), a species not caught at other sites using the 25 m beach seine. Site 2 was the closest in species composition to the cluster formed by Sites 3 and 4. Ocean pout (Zoarces americanus) and radiated shanny (Ulvaria subbifurcata) were caught only at Site 2 and their capture explains the dissimilarity between Site 2 and the cluster of Sites 3 and 4 (J _D  = 0.35). Only 10 species were collected at Site 2A with the 25 m beach seine. This explains the dissimilarity between Site 2A and Sites 2, 3, and 4 (J _D  = 0.44). Site 4A is the most dissimilar (J _D  = 0.47) from the other sites.

Species \ Sites	2	2A	3	4	4A
Leucoraja ocellata		x	x	x	x
Salmo salar			x	x	x
Salvelinus fontinalis	x		x	x	x
Osmerus mordax	x		x	x	x
Gadus morhua	x	x	x	x
Gadus ogac		x		x
Merluccius bilinearis			x
Urophycis tenuis	x		x	x
Gasterosteus aculeatus	x	x	x	x	x
Gasterosteus wheatlandi	x		x	x	x
Apeltes quadracus					x
Syngnathus fuscus			x	x	x
Tautogolabrus adspersus	x	x	x	x	x
Zoarces americanus	x
Ulvaria subbifurcata	x
Pholis gunnellus	x		x	x
Ammodytes americanus	x	x	x	x	x
Hemitripterus americanus			x	x
Myoxocephalus aenaeus	x	x	x	x
Myoxocephalus octodecemspinosus	x	x	x	x
Myoxocephalus scorpius	x	x	x	x
Scophthalmus aquosus			x
Limanda ferruginea			x
Pseudopleuronectes americanus	x	x	x	x	x

Fish assemblages and nearshore habitats in Bonne Bay

Twenty-nine species from 17 families were collected during this baseline survey of the fish fauna of inner Bonne Bay. Fish assemblages determined at the sampling sites reflected species response to the available habitat (Able and Fahay 2010) and not the geographic distance between sites. Cluster analysis grouped together the fish assemblages from the two low salinity sampling sites: the barachois at Site 3A and the river delta at Site 4A. This was explained by the presence of Fourspine stickleback (Apeltes quadracus), a species which is found freshwater and brackish habitats, and by the absence of marine species from the families Gadidae, Zoarcidae, Stichaeidae, Pholidae, Cottidae, Agonidae, and Bothidae.

With salinity, substrate and associated flora also determined the fish assemblages at each site. For instance, the northern pipefish (Syngnathus fuscus) was found only at sites with presence of eelgrass beds (Sites 3, 4, and 4A). The long and thin body of northern pipefish is easily confounded with eelgrass leaves, thus northern pipefish might use eelgrass beds as a refugee from predator, explaining that we observed the species only at sites with eelgrass cover (Collette and Klein-MacPhee 2002). Similarly, three species, rock gunnel (Pholis gunnellus), cunner (Tautogolabrus adspersus), and shorthorn sculpin (Myoxocephalus scorpius), were found only at the four sites with rocky bottom substrate, partly explaining that these four sites grouped in the cluster analyses. Interestingly, Site 3 had the most complex substrate with rocky bottom and eelgrass beds, which might explain the higher diversity of fish observed at that site.

Comparison of the fish faunas of Bonne Bay, Newfoundland and Gilbert Bay, Labrador

While little is known about the inshore fish fauna on the west Coast of Newfoundland, a previous study used the same 10 m and 25 m beach seines and gillnets to sample the inshore fish fauna of Gilbert Bay, a Marine Protected Area in southeast Labrador (Wroblewski et al. 2007). This offers the opportunity to compare inshore fish assemblages from coastal embayment south (i.e. Bonne Bay) and north (i.e. Gilbert Bay) of the Strait of Belle Isle, a potential biogeographical boundary (Steele 1975, 1983).

Seventeen species observed in this study were also observed in Gilbert Bay: Atlantic herring (Clupea harengus), Atlantic salmon (Salmo salar), brook trout (Salvelinus fontinalis), rainbow smelt (Osmerus mordax), Atlantic cod (Gadus morhua), Greenland cod (Gadus ogac), white hake (Urophycis tenuis), threespine stickleback (Gasterosteus aculeatus), blackspotted stickleback (Gasterosteus wheatlandi), fourspine stickleback (Apeltes quadracus), ocean pout (Zoarces americanus), rock gunnel (Pholis gunnellus), American sandlance (Ammodytes americanus), sea raven (Hemitripterus americanus), shorthorn sculpin (Myoxocephalus scorpius), alligatorfish (Aspidophoroides monopterygius), and winter flounder (Pseudopleuronectes americanus).

Twelve species present in Bonne Bay were not observed in Gilbert Bay. Out of this 12 species, four have been previously documented north of the Strait of Belle Isle: thorny skate (Amblyraja radiata), Atlantic tomcod (Merluccius tomcod), Acadian redfish (Sebastes fasciatus), and yellowtail flounder (Limanda ferruginea). while eight species, winter skate (Leucoraja ocellata), northern pipefish (Syngnathus fuscus), silver hake (Merluccius bilinearis), cunner (Tautogolabrus adspersus), radiated shanny (Ulvaria subbifurcata), grubby (Myoxocephalus aenaeus), longhorn sculpin (Myoxocephalus octodecemspinosus), and windowpane flounder (Scophthalmus aquosus), have not been documented as present north of the Strait of Belle Isle (Scott and Scott 1988; Collette and Klein-MacPhee 2002).

The Strait of Belle Isle separates the subarctic waters of the Labrador Sea from the cold-temperate waters of the Gulf of St. Lawrence and this was reflected by oceanographic conditions observed in Gilbert Bay and Bonne Bay (Fig. 5). Cold (0 - 6 °C) Labrador Shelf water enters Gilbert Bay along the bottom at 5-30 m depth (Wroblewski et al. 2007). During the winter months, heat loss from the water column to the atmosphere results in seawater temperatures < 0 °C throughout the water column and sea ice forms in the Bay (Best et al. 2011). Ice formation along the shoreline begins in October and the bay is usually frozen over by the end December. Gilbert Bay is ice-covered for about five months of the year. In contrast, Gulf of St. Lawrence waters enter inner Bonne Bay by crossing along the bottom of the sill at 5-25 m depth (Richards and de Young 2004). During the late summer Gulf water crossing the sill has a temperature of 16 - 18 °C. During late winter, heat loss from Gulf surface waters to the atmosphere results in seawater temperatures < 0 °C throughout the water column. Sea ice may form in the Gulf and in Bonne Bay, depending on the severity of the winter weather.

Observation of water temperature in Gilbert Bay and Bonne Bay in 2004 and 2006, showed that during the winter of 2004 the seawater surface temperature in Bonne Bay declined to 0 °C in early February, while in Gilbert Bay the seawater surface temperature had already cooled to subzero temperatures 6 weeks earlier (Fig. 5). Seawater surface temperatures rose above 0 °C by the end of April in Bonne Bay, while the landfast ice covering Gilbert Bay only began to melt in mid-May. The ice covered period varies annually in both bays. For example, in 2006 sea ice appeared in Bonne Bay for only a few weeks in March, while Gilbert Bay was ice covered from mid-December to mid-May (authors’ personal observations).

Physiological factors limiting tolerance of low temperature may explain the observed differences. Cunner (Tautogolabrus adspersus), a species abundant in Bonne Bay but absent from Gilbert Bay, is unable to survive prolonged sub-zero seawater temperatures (Scott and Scott 1988). Seawater temperatures in Gilbert Bay are subzero for 5 months of the year (Fig. 5). Mass mortalities of cunner during winter have been observed in Newfoundland coastal waters (Green 1974). We suggest that Tautogolabrus adspersus serves as an indicator species for the biogeographical boundary between the cold-temperate waters of coastal Newfoundland and the subarctic waters of coastal Labrador.

Marine fish species of conservation concern in Bonne Bay

Our sampling surveys collected four of species of conservation concern: winter skate, thorny skate, Atlantic cod and Acadian redfish. Winter skate are extremely rare in the marine waters of the northern Gulf of St. Lawrence (Gauthier and Nozères 2016), and considered “endangered” by COSEWIC (COSEWIC 2015). Yet, we collected ten specimens of winter skate from subtidal waters with sand and gravel bottoms in Deer Arm and East Arm of the Bonne Bay fjord. Based its standard length and maturity, one specimen which we collected (Additional file 1: Photograph 1) exhibited the late-maturing trait of winter skate adapted to waters of the northern Gulf and western coastal Newfoundland (Gauthier and Nozères 2016). Only one specimen of the late-maturing type of winter skate has been recorded in fisheries stock assessment surveys of the northern Gulf (Gauthier and Nozères 2016). That winter skate was collected in a bottom trawl off the southwest coast of Newfoundland in 2008.

We collected seven specimens of thorny skate using gillnets set in subtidal waters at sampling sites in Deer Arm and East Arm. All seven were sexually mature. The thorny skate was assessed by COSEWIC and assigned a status of “special concern” because the species has declined in abundance in the southern Gulf of St. Lawrence, on the Scotian Shelf and on the Grand Banks (COSEWIC 2012d). The thorny skate has a low reproductive potential, and populations recover slowly. There is no commercial fishery in the Gulf targeting thorny skate, but the species is caught as bycatch in fisheries for groundfish using bottom trawls (Wroblewski 2013).

We collected 55 specimens of Acadian redfish using gillnets set in subtidal waters at sampling sites in the East Arm. Almost all specimens collected were sexually mature, many in spawning condition. Our data contributes to the evidence for a resident population of Acadian redfish in Bonne Bay. Genetic and morphometric analyses of specimens caught in Bonne Bay indicate that a population of Acadian redfish exists in Bonne Bay which is reproductively isolated from redfish in the Gulf of St. Lawrence (Valentin 2006). The Acadian redfish population in Bonne Bay is considered an evolutionary significant unit (Sévigny et al. 2007). The population of Acadian redfish living in Bonne Bay has been designated by COSEWIC as “special concern.” (COSEWIC 2010) Redfish are caught as bycatch in commercial seining fisheries for mackerel and capelin in Bonne Bay (Wroblewski 2013). Fish species listed as “species-at-risk” often have biological characteristics that make them susceptible to population declines, e.g. a specialized habitat or diet, a small geographic range, or a large body size easily harvested in a fishery (Powles et al. 2000; Hutchings 2001). In the case of Bonne Bay redfish population, the concern is the small geographic range. The redfish population is likely confined by the fjord sill to East Arm and Deer Arm of Bonne Bay.

Previous study of local ecological knowledge suggested that similar to redfish, Bonne Bay could support a local population of Atlantic cod (Murray et al. 2008). During our sampling we consistently collected age 0 Atlantic cod indicating that inner Bonne Bay is a nursery ground for Atlantic cod. We also observed Atlantic cod in spawning condition and a homing behaviour to South Arm Bonne Bay (Le Bris et al. 2013). This suggests that age 0 cod collected in Bonne Bay may have been spawned in the South Arm and drifted with the current flowing across the sill into East Arm. However, genetic analysis and population dynamics modelling have suggested that, while local recruitment may occur in Bonne Bay, Atlantic cod observed in Bonne Bay are part of the Northern Gulf of St. Lawrence population and do not form an isolated population (Le Bris 2014).

Finally, it can be noted that two species of wolffish, spotted wolffish (Anarhichas minor) and Atlantic wolffish (Anarhichas lupus), respectively listed as “threatened” and designed as “Special Concern” under SARA (COSEWIC 2012b) (COSEWIC 2012c) have been previously reported during SCUBA diving in Bonne Bay (Hooper 1975) and are sometimes caught in the commercial trap fishery for lobster and snow crab (authors’ personal observations). It is likely that these species were not caught in our surveys because they are often associated with boulder habitats, a type of habitat not sampled with our gears.