Fauna associated with shallow-water methane seeps in the Laptev Sea

Grab samples

Unconstrained tree with SIMPROF analysis revealed five significantly distinct groups of samples (Fig. 2). The UNCTREE parameters are shown in Supplemental Information 2. The groups partly corresponded with the station locations and presence/absence of methane seeps (Control, C15 and Oden sites). To avoid a mix-up between the station groups and seeping sites hereinafter the corresponding names are used with either –station group or –site ending.

Square root transformed biomass data are used. Samples and knots that were not significantly different at p <0.05 are connected with red dashed lines. Green lines indicate SIMPROF groups.

Characteristics of station groups

At the Control station group (Fig. 1), the bivalve Portlandia arctica comprised most of the biomass at all three stations, followed by the starfish Ctenodiscus crispatus and the bivalve Macoma calcarea. Due to the low number of samples, the species-individuals accumulation curve did not reach the saturation point (Fig. 3). Control station group had in comparison to the other groups from C15 and Oden seep sites the lowest density and species richness, whereas the evenness was the highest (Fig. 4).

The C15-seep a station group included five stations, all within the C15 seep site. In this group, biomass and diversity values were intermediate among other station groups. Dominant species in this group were the bivalve Nuculana pernula, the siboglinid Oligobrachia sp. and the polychaete Cistenides hyperborea. Species-individuals accumulation curve in this group reached saturation due to the largest number of samples (Fig. 3).

The C15-seep b station group consisted of only two stations from C15 site. This group demonstrated the highest abundance values and low biomass among all groups. Dominant species included small polychaetes Cossura longocirrata, Micronephthys minuta and Ophryotrocha sp. (Fig. 4, Supplemental Information 1).

The Oden station group included six stations, all located within the Oden seep site. Values of biomass, species richness and diversity indices in this group were the highest among all station groups (Fig. 4). The main dominant species were the siboglinid Oligobrachia sp. and the polychaetes Myriochele heeri and Nephtys ciliata.

The last group C15 background contained five stations taken within the C15 site. Taxonomical composition at these stations was similar to that in the Control group, with the bivalve Portlandia arctica being the dominant species. Bivalves Yoldiella lenticula and Y. solidula were subdominant. In this station group, the biomass values were the lowest, other general community characteristics were intermediate (Fig. 4). Similar to the Control group, C15 background did not reach the saturation point at species-individuals accumulation plot (Fig. 3).

Comparison of seep and non-seep station groups

General community characteristics in the station groups appeared different in abundance, biomass and diversity (Fig. 4, Supplemental Information 3). The abundance of several taxa varied significantly among five station groups (Control, C15 background, C15 seep a, C15 seep b and Oden) (Fig. 5). The Kruskal-Wallis test showed that differences in abundance of at least ten species are statistically reliable (Table 2). Statistics of the Dunn’s test are shown in Supplemental Information 4. The seep sites were characterized by higher densities of the polychaetes Tharyx sp. and Cistenides hyperborea and the ophiuroid Ophiocten sericeum. On the contrary, the bivalve Portlandia arctica was markedly more abundant in Control and, to a lesser extent, in C15 background station groups (Fig. 5). Notable were extreme densities of small polychaetes at some seep stations, including Cossura longocirrata and Ophtyotrocha sp. (Fig. 5A) at C15 seep b.

Certain species present at some methane seep sites were completely absent at the non-seep sites (Fig. 5). Among them, at least five species (the polychaetes Spiochaetopterus typicus and Nephtys ciliata, the siboglinid Oligobrachia sp., the bivalve Axinopsida orbiculata and the amphipod Pleusymtes pulchellus) were present only at C15 and Oden sites not randomly. At least one species, the undescribed gastropod Frigidalvania sp., was present only in Oden station group and absent in other station groups not randomly (Table 3). The estimated number of grabs required to catch the latter species was slightly lower than the number of grabs taken.

Trawl samples

The overall Bray-Curtis similarity between the two trawls was 65.6%. Species ranking graphs showed high level of dominance by abundance for both trawl stations (Fig. 6). The dominant species in both trawls was the ophiuroid Ophiocten sericeum: 37% of the total abundance at C15 and 46% at Oden. The second most abundant species at C15 was the gastropod Frigidalvania sp. (12%) and at Oden the bivalve Yoldiella solidula (11%). Ten most abundant species accounted for >70% of the total abundance in both trawls (Fig. 6).

Species richness, Pielou evenness, Hurlbert rarefaction for 100 individuals and Shannon-Wiener index are shown in Table 4. The evenness and ES (100) was higher in the Oden-trawl than in the C15-trawl, similar to results based on grab samples. However, the species richness (as well as the total amount of individuals) in the C15-trawl was higher than in the Oden-trawl (Table 4, Supplemental Information 1).

Species responsible for taxonomical difference between the two trawl samples are shown in Table 5. Most notable is a high abundance of the gastropod Frigidalvania sp. at C15. At Oden Frigidalvania sp. was also present, but in much smaller densities (only 2.3% of the total abundance). In addition, C15-sample differs from Oden by high numbers of various filter-feeders including 6 species of sponges (with Craniella polyura being most numerous), at least 6 species of cnidarians, 17 species of bryozoans and 3 species of tunicates (Table 6).

At C15 trawl sample, a large piece of carbonate crust was found. Cavities of its pores were inhabited by numerous polychaetes, also recognized from the soft sediments around the seepage area (e.g., members of families Nephthyidae, Nereididae, Oweniidae and Terebellidae, see Supplemental Information 1), and by several filter-feeders (Hydrozoa).

Comparison of C15 and Oden sites

All gears showed significant differences between the C15 and Oden sites expressed in different taxonomical composition and quantitative characteristics. The Bray-Curtis similarity between the sites according to the grab samples and trawl samples was 26.2 and 65.6, respectively. The main differences in species composition included the high abundance of the sponge Craniella polyura and the gastropod Frigidalvania sp. at C15 site and higher numbers of the ophiuroid Ophiocten sericeum at Oden site.

The grab samples indicated a high level of heterogeneity in benthic fauna distribution. Some species formed patches, for example Oligobrachia sp., Cossura longocirrata and Ophryotrocha sp., being extremely numerous at some and moderate in abundance at the neighboring grab stations. There were also species with rather uniform distribution based on combined data, for example Ophiocten sericeum. According to the cluster analysis, the C15 site is more heterogenic forming at least three different species complexes within its area (Fig. 2). Dissimilarities within the C15 and Oden sites were 64.7 and 26, respectively (Supplemental Information 2).

Dominant species were different in grab and trawl samples. The dominant species in trawls at the two methane seep sites was the ophiuroid Ophiocten sericeum, whereas for grab samples, the siboglinid Oligobrachia sp., the bivalve Nuculana pernula and the polychaete Myriochele heeri were the most abundant.

Integral community parameters: methane seep vs. non-seep

The abundance of macrofauna was higher at the seep stations compared to the background, based on the grab samples. In addition, at Oden seep site the biomass was higher compared to non-seep sites. High abundance and biomass has been reported from both hydrothermal vents and cold seeps all over the world, compared to the surrounding areas (Levin, 2005; Dando, 2010). In the Arctic, a twofold increase of biomass compared to control sites was observed at cold seeps south off Svalbard (mean values of 20.7 vs. 9.8 g ww m⁻²), the abundance increase was less prominent (770 vs. 590 ind. m⁻²) (Åström et al., 2016). At the Vestnesa Ridge the infaunal abundance and biomass was almost five times higher compared to a nearby control area (497 vs. 140 ind. m⁻² and 2.97 vs. 0.48 g ww m⁻², respectively) (Åström et al., 2018). For the Håkon Mosby mud volcano, which is very deep compared to the Laptev Sea cold seeps (1250 vs. 70 m, respectively), though similar in species composition (Oligobrachia fields near the seepage zones and Ophiocten-dominated background community) the comparison of abundance and biomass with the background is not available. In our study, the abundance at the methane seep sites C15 and Oden was more than four times higher than at the control. However, differences in biomass although pronounced were not statistically significant. Increased biomass at seep habitats is commonly explained by enhanced organic matter supply and habitat heterogeneity (Gebruk et al., 2003; Sen et al., 2018).

Pielou’s evenness was distinctly higher at the Control and C15 background station groups, which reflects the increased dominance of certain species at seep stations compared to non-seep. Many authors reported high abundance and biomass values of one to few dominant species at various cold seeps (Gebruk et al., 2003; Åström et al., 2016; Åström et al., 2018). This can be caused by conditions less favorable for some background species (due to the presence of methane and sulphides and lower oxygen level), but more favorable for symbiotrophs or grazers (Powell, Bright & Brooks, 1986; Dando et al., 1993; Dando, 2010).

The cold seeps all over the world oceans usually demonstrate lower diversity values (ES(100) and Shannon-Wiener index) compared to the background areas (summarized by Levin, 2005). However, species list from grab and trawl samples showed high ES(100) and Shannon-Wiener index values although only two trawls were sampled. Studies on the Siberian shelf using the same gear under the same conditions obtained less than 150 species per trawl (Galkin & Vedenin, 2015; Vedenin, Galkin & Kozlovskiy, 2015), while a total of 203 species were found in a single C15 sample. The unusually high diversity may reflect a higher amount of microniches within the C15 site. This is indirectly confirmed by lower similarity values observed between all C15 grab samples. Higher habitat heterogeneity at seep sites can increase the overall diversity of benthic fauna (Bergquist et al., 2003; Gebruk et al., 2003; Levin, 2005; Menot et al., 2010; Åström et al., 2018). The scale of heterogeneity is hard to assess, but based on stations coordinates and the fact that stations 5947-1 and 5947-2 from C15-site were grouped in C15 background (though planned as seep stations near the gas flares), while station 5947-2 was grouped as C15-seep a, we can assume that the scale is less than 5 m (distance between these stations) (Fig. 1; Table 1).

In addition, the diversity values (ES(100) and Shannon-Wiener index) at the Oden station group were significantly higher than at the C15-site. The reasons for this are unknown so far, since no environmental parameters measured directly at benthic stations are available. The peculiarly higher values of macrofaunal diversity (Shannon-Wiener index) within the cold seeps are known only for a few seep areas, e.g., for the Vestnesa Ridge (Åström et al., 2018), for the South-Western Barents Sea (Sen et al., 2019b) and for the Bahía Blanca estuary in Argentina (Bravo et al., 2018).

Common shelf taxa responsible for differences in station groups

The station groups revealed by UNKTREE and n-MDS analysis largely corresponded to the geographical position of the C15, Oden and control sites. Twelve common species widely distributed across the Siberian shelf (Table 2, Sirenko, 2001) were largely responsible for increased integral community parameters in our study. Among such species (based on grab samples) were the polychaetes Spiochaetopterus typicus, Cossura longocirrata and Tharyx sp., the bivalve Macoma calcarea, the amphipod Pleusymtes pulchella and the ophiuroid Ophiocten sericeum. In addition, based on trawl data, the sponge Craniella polyura was present in high densities at the C15 site, together with other filter-feeders including cnidarians and bryozoans (Table 6). Apparently the same species aggregations were visible on the video reported by Baranov et al. (in press). All these above-mentioned species have been previously reported from a wide range of areas of the Laptev Sea and adjacent regions (Sirenko et al., 2004).

The increased density of common taxa at deep-sea hydrothermal vents and cold seeps is a well-known phenomenon usually explained by increased availability of organic matter in these habitats (Hessler & Kaharl, 1995; Levin, 2005). In the Arctic, the increased biomass and abundance of common allochthonous species were reported for the Håkon Mosby mud volcano (Rybakova et al., 2013), Svalbard (Åström et al., 2016) and Vestnesa Ridge cold seeps (Åström et al., 2018). Also, a significant increase of abundance of filter feeders (especially sponges) was shown for the Aurora Seamount on the Gakkel Ridge, the only investigated hydrothermal vent in the Central Arctic Ocean (Boetius, 2015; Bünz et al., 2020). Soft corals and crinoids were reported to be abundant around the vent fields on the Mohn’s Ridge (Sweetman et al., 2013).

Taxa specific for methane seep sites

In our study, the most characteristic species from methane seep sites was the siboglinid Oligobrachia sp. (Fig. 7A). This species was present at all seep station and absent at all background/control station. This species is morphologically very close to Oligobrachia haakonmosbiensis originally described from the Håkon Mosby mud volcano from the depth of ∼1200 m (Smirnov, 2000). Colonies of O. haakonmosbiensis with the biomass reaching 350 g ww m⁻² were reported from this area (Gebruk et al., 2003). Recent phylogenetical analyses showed that the species from the Laptev Sea belongs to a separate, undescribed species of Oligobrachia (Sen et al., 2018). In the Laptev Sea, Oligobrachia sp. is known from different localities, seep and noon-seep, occurring in a wide depth range 100–2,166 m (Buzhinskaja, 2010). Our record at 63 m is the shallowest for this species, with high population density and biomass: >1000 ind. m⁻² and 45 g ww m⁻² at Stat. 2623-1 and 5953-2 (Oden site). Several specimens from 2015-samples (identified as O. haakonmosbiensis) were investigated using transmission electron microscopy (Savvichev et al., 2018). Usually the endosymbionts of siboglinids are represented by sulphide-oxidizing bacteria (Rodrigues et al., 2011; Lee et al., 2019), however in the study of Savvichev et al. (2018), the methanotrophic bacteria were found inside the trophosome.

Additionally to the siboglinids, several samples from the seep-sites were also characterized by high abundance of mollusk species. The gastropod Frigidalvania sp. (Rissoidae) occurred in high density at C15 site: up to 2340 ind. m⁻² and 25 g ww m⁻² at St. 5625-3 (Fig. 7B). According to trawl samples, this species occurs at the Oden site, but in very low numbers (only 2.4% from total abundance, see Table 5). This species is presumably new to science, based on the presence of a single spiral rib and the absence of knobs, that distinguish it from all known Arctic Frigidalvania species (Warén, 1974). Subfossils of similar unidentified Frigidalvania were found by Thomsen et al. (2019) in the sediment cores at the Vestnesa Ridge cold seeps. Large numbers of unknown rissoid gastropods were previously reported from the Håkon Mosby mud volcano, referred to as Alvania sp. in Gebruk et al. (2003). The stable isotope analysis showed that the rissoids from the HMMV actively graze on bacterial mats (Decker & Olu, 2012). Furthermore, another rissoid gastropod, Pseudosetia griegi, was observed grazing on bacterial mats at the hot vent Loki Castle on the Mohn’s Ridge (Sweetman et al., 2013). At the recently investigated Lofoten canyon seep site dense aggregations of unidentified rissoids were observed on photographs from ROV (Sen et al., 2019a). The details available from the published photo in Sen et al. (2019a) (see Fig. 4B) allow us to suggest that the gastropods are likely to belong to genus Frigidalvania, based on the shell shape and rusty-brownish periostracum. Unfortunately, in our study we were not able to identify the behavior or lifestyle of Frigidalvania sp. This species remained unnoticed in the video data due to its small size (Baranov et al., in press). However, multiple bacterial mats observed from video-transects and caught by box corer provide an opportunity for such species to graze on them (Savvichev et al., 2018; Baranov et al., in press). Other species common at the seep sites in this study, however missing in the background/control was the thyasirid bivalve Axinopsida orbiculata (Fig. 7D). Some species of thyasirids are known as symbiotrophic, however, the information on symbiotic bacteria in the gills of A. orbiculata is controversial: Zhukova, Kharlamenko & Gebruk (in press) have demonstrated the presence of bacteria in bivalve specimens from the Kraternaya Bight, the Kuril Islands, whereas according to Dufour (2005) this species lacks bacterial symbionts. It is possible that the high densitiy of A. orbiculata is attributed to increased food availability at seep sites. Overall, no bivalves restricted to cold seeps are known so far in the Arctic with the exception of two large thyasirids recently described based on few empty shells (Åström, Oliver & Carroll, 2017) and Pleistocene subfossils (e.g., Archivesica spp., Sirenko et al., 2004; Hansen et al., 2017) found deep in the sediment at formerly and/or presently active cold seeps and hydrothermal vents. The subfossils suggest that the Arctic cold seeps (and possibly hydrothermal vents) could be inhabited by richer fauna during late Pleistocene that became extinct during or after Quaternary glaciation (Discussed in Kim et al., 2006; Hansen et al., 2017; Thomsen et al., 2019).

Notable is the high density (>3600 ind. m⁻²) of the dorvilleid polychaete Ophryotrocha sp. in one grab sample at C15-seep b station group (Supplemental Information 1 (Fig. 7C)). At least 15 species of Ophtyotrocha have been described from reducing habitats (Taboada et al., 2013; Salvo et al., 2014; Ravara et al., 2015), including two species considered as obligate for cold seeps in the Kagoshima Bay, Japan (Miura, 1997). On the other hand, many species of this genus not restricted to reducing habitats are common in the Arctic seas (Sirenko, 2001).

Another taxon common in reducing habitats is tanaid crustaceans (Tanaidacea) (Larsen, 2006; Błazewicz-Paszkowycz & Bamber, 2011; Zeppilli et al., 2011). In our material three species were present (Supplemental Information 1), all widely distributed in the Arctic (Sirenko, 2001). The density of tanaids in our samples was low, although this taxon was reported in high densities from the Håkon Mosby mud volcano (Gebruk et al., 2003) and the Vestnesa Ridge (Åström et al., 2018) with several species (described as new) restricted to the methane seep habitats (Błazewicz-Paszkowycz & Bamber, 2011). It seems likely that many species of tanaids remain unidentified and diversity in this taxon remains underestimated owing to difficulties of identification of these small crustaceans (summarized by Błazewicz-Paszkowycz & Bamber, 2011). The low number of tanaids in our samples could be a result of a too large sieve mesh size used onboard (see Materials & Methods). Tanaids commonly are <0.5 mm in size and require a corresponding mesh size to be found (Pavithran et al., 2009).

Overall, considering grab and trawl data combined, all the seep-specific taxa were the same at both seep sites. The only exception is the polychaete Ophryotrocha sp., which could be missed from the Oden trawl sample due to the large sieve mesh size (Supplemental Information 1).

Presence of specific benthic communities at C15 and Oden

Up to now no distinct macrofaunal changes in response to methane seeps were reported from the Arctic Ocean at depths <80 m. In general, at depths <200 m both hydrothermal vents and cold-seeps are usually colonized by a subset of the local fauna (Tarasov et al., 2005; Dando, 2010). Some species notable at shallow-water methane seeps belong to opportunistic taxa common in various reducing habitats which include siboglinid polychaetes and thyasirid bivalves reported from Skagerrak, Kattegat, coastal areas of Florida, Japan, New Zealand, New Guinea etc. (Southward & Culter, 1986; Schmaljohann & Flügel, 1987; Schmaljohann et al., 1990; Malakhov, Obzhirov & Tarasov, 1992; Gebruk, 2002). The stable isotope data of nitrogen and carbon sources from various cold seeps all over the ocean suggest that food sources of macrofauna at shallow-water methane seeps are largely photosynthesis-based (Southward et al., 1996; Levin et al., 2000; Dando, 2001; Levin, 2005; Dando, 2010). It was suggested that the faunistic depth boundary between the deep-sea and shallow-water vents and seeps at approximately 200 m is controlled by the amount of organic matter input from the photosynthetic production (decreasing below the photic zone) and the greater number of predators at shallow depths (summarized by Dando, 2010). Seep-obligate species were not reported from depths <200 m (Tarasov et al., 2005; Dando, 2010).

At the same time, methane seep habitats even at shallow depths increase a number of microniches owing to increased organic matter availability, variety of substrates (including different types of sediment and carbonate crusts) and repeated disturbance (Sahling et al., 2003; Dando, 2010; Pohlman et al., 2017; Åström et al., 2019). Therefore, shallow cold-seeps may support greater species diversity (in terms of both richness and diversity indices) compared to the background. In our study at two methane seep sites, C15 and Oden, integral community characteristics were significantly different from those in non-seep areas partially due to presence of species obligate for reducing habitats. In addition, the communities found at C15 site formed several station groups and were more scattered at the n-MDS plot (Figs. 2A, 2B) which could indicate a larger diversity of microniches within this site. Large amount of filter-feeders (Hydrozoa and Bryozoa) found in C15-trawl indicates the presence of hard substrata (including carbonate crusts). The larger amount of microniches is partly supported by the video-data, where the landscape within the active seepages was more complex than in non-seep areas (Flint et al., 2018; Baranov et al., in press).

Overall, the impact of carbon of different origin and the methane-induced habitat heterogeneity are difficult to separate unless the detailed measurements of environmental parameters and isotope ratios from different organisms are obtained. The isotope data are currently available only for the siboglinid Oligobrachia sp., indicating the leading role of microbial methane oxidation in the nutrition of these symbiotrophic worms (Savvichev et al., 2018).

Recent studies conducted at several seep sites south of Svalbard (∼300 m depth) showed a certain input of chemosynthesis-based carbon into the largely photosynthesis-based local benthic food web (Åström et al., 2019). As an example, the diet of several heterotrophic species, also present in our samples (e.g., polychaetes Nephtys sp. and Scoletoma fragilis) consisted of chemosynthesis-derived organic at the level of 30–40% (Åström et al., 2019). Isotope-based nutritional investigations of shallow-water cold seeps in the North Sea, Northern California and Kattegat showed little or no chemosynthetic contribution to macrofaunal diet (Dando et al., 1991; Jensen et al., 1992; Levin et al., 2000). Based on the increase of density of many species at the seep sites compared to the background (Table 2), it can be assumed that the chemosynthesis-based organic may contribute to the diet of heterotrophic species, such as Nephtyidae polychaetes and other taxa.

Unfortunately, no environmental data except for the echo-sounding showing gas emissions and CTD-measurements obtained from the area of the seeps from only two stations (both away from the benthic sampling sites) are available (Flint et al., 2018; Baranov et al., in press). Nevertheless, we suggest that response of macrofauna to methane seepage at shallow depths 60–70 m can be related to very low primary productivity on the outer shelf of the Laptev Sea, dropping from ∼720 mg C m⁻² per day at Lena river delta to <100 mg C m⁻² per day at 600 km (Sorokin & Sorokin, 1996) during September. Outside short Arctic summer, these values decrease to almost zero. In these extremely oligotrophic conditions, methane as a source of energy for the methane-oxidizing bacteria stimulates the development of local patchy benthic communities even at a shallow depth. As a comparison, the specific communities with siboglinids around Svalbard located at similar latitude are developed only at depths >200 m, whereas at 80 m no response of macrofauna is observed (Åström et al., 2016). Unlike the Laptev Sea shelf, the primary production south and west off Svalbard reaches much higher values up to 1,800 mg C m⁻² per day during May blooms (Wassmann et al., 2006). Furthermore, the Barents Sea remains uncovered with ice during most of the year, while the Laptev Sea shelf is ice-free during one to two months annually. Other examples of shallow-water cold seeps with little or no macrofaunal response are also located in non-oligotrophic areas, such as the North Sea pockmark at 150 m depth with pelagic production up to 1,300 mg C m⁻² per day (Moll, 1998; Dando et al., 1991) and the Baltic Sea pockmarks with >2,000 mg C m⁻² per day in April (Lignell, 1990; Pimenov et al., 2008).

The only investigated shallow-water reducing habitat comparable in the primary production level to the Laptev Sea is the hydrothermal vent of the volcanic Deception Island, Antarctica. Environmental conditions in the inner bay (Port Foster) of the Deception Island are oligotrophic, most of the surface is ice-covered during austral winter (Bright et al., 2003; Angulo-Preckler et al., 2017). However, several fumaroles and hot springs in Port Foster, located from intertidal to ∼15 m depth enrich the surrounding waters with sulfides and methane and keep the sea surface above them ice-free (Tilbrook & Karl, 1993). Certain increase of local macrofauna around the hot springs was described recently (Angulo-Preckler et al., 2017). In addition, one taxon, apparently restricted to reducing habitat, was found in the sediment around the hydrothermal vents. The new taxon was a monocelid flat worm, covered with chemoautotrophic bacteria (Bright et al., 2003). However, no seep-specific macrobenthic communities were observed in Port Foster (Angulo-Preckler et al., 2017), presumably owing to the very shallow depth: the maximum chemoautotrophic activity and the peak abundance of monocelid worms were recorded at 1–2 m depth (Bright et al., 2003).

To confirm the relationship between the depth where the seep-specific communities can develop and the primary production, more shallow-water cold seeps at the extremely oligotrophic conditions (e.g., polar areas) should be investigated.