Polar Biology

, Volume 32, Issue 9, pp 1331–1343

Dinoflagellates in a fast-ice covered inlet of the Riiser-Larsen Ice Shelf (Weddell Sea)

Authors

    • Ocean Sciences, College of Natural SciencesBangor University
    • Department of Earth and Atmospheric SciencesUniversity of Alberta
  • Fabienne Marret
    • Department of GeographyUniversity of Liverpool
  • David N. Thomas
    • Ocean Sciences, College of Natural SciencesBangor University
  • James D. Scourse
    • Ocean Sciences, College of Natural SciencesBangor University
  • Gerhard S. Dieckmann
    • Alfred Wegener Institute for Polar and Marine Research
Original Paper

DOI: 10.1007/s00300-009-0630-5

Cite this article as:
Pieńkowski, A.J., Marret, F., Thomas, D.N. et al. Polar Biol (2009) 32: 1331. doi:10.1007/s00300-009-0630-5

Abstract

A short-term (3–15 days) multiple and single sediment-trap array deployed in Drescher Inlet (Eastern Weddell Sea) during austral summer 1998 showed well preserved and relatively diverse dinoflagellate assemblages comprised of 13 taxa. Consistent with other Antarctic studies, large Protoperidinium species were dominating whereas Preperidinium and Dinophysis showed minor frequencies. Athecates were not observed, possibly due to their poor preservation status. The majority of dinoflagellates were heterotrophic species, likely feeding on previously recorded abundant diatoms at the study site. Assemblage structures varied according to depth (Protoperidinium antarcticum and P. rosaceum at 10 m depth vs. P. macrapicatum and Preperidinium granulosum at 360 m depth) and collection period (first period: P. antarcticum; second period: Protoperidinium sp. C). Sediment-trap dinoflagellates were either derived from a flux out of the overlying fast ice, platelet ice, or the water column but given their high mobility, migration between these media cannot be ruled out.

Keywords

DinoflagellatesProtozooplanktonSediment trapSouthern OceanThecatesAntarctica

Introduction

The importance of dinoflagellates as a prominent component of high latitude marine plankton has become apparent in the last few decades (e.g. Antarctic: Villafañe et al. 1995; Kopczyńska et al. 1998; Arctic: Moller and Nielsen 2000; Lovejoy et al. 2002). As phototrophic algae they contribute to the primary productivity of aquatic ecosystems (Bathmann et al. 1997), whilst as protozooplankton they provide a trophic link between small photosynthetic algae (on which they feed) and larger zooplankton (by which they are grazed) (e.g. Burkill et al. 1995; Froneman et al. 1996; Froneman 2004; reviewed in Sherr and Sherr 2007). They also occupy a significant role in carbon and nutrient cycling (Levinsen and Nielsen 2002) and may be key players in influencing atmospheric carbon dioxide and dimethyl sulphide concentrations (Scarratt et al. 2002; Steinke et al. 2002), thereby exerting a potential influence on climate (Watson and Liss 1998; Simó and Pedrós-Alló 1999). Adaptive strategies such as diel vertical migration and the formation of cysts have allowed dinoflagellates to successfully colonise extreme polar environments such as the Southern Ocean (e.g. Andreoli et al. 1995; Kopczyńska et al. 1998; Rengefors et al. 2008).

The first Southern Ocean studies on dinoflagellates were initiated in the late 1940s (Balech 1947). Pioneering descriptive and taxonomic investigations were conducted by Enrique Balech and others for the Weddell (Balech 1947; Balech and El-Sayed 1965) and Bellingshausen (Balech 1947, 1973) seas, Terre Adélie (Balech 1958), the sub-Antarctic Atlantic Ocean (Balech 1978) and for Antarctica as a whole (Balech 1968, 1970, 1975). Balech (1968) showed that armoured dinoflagellates (e.g. Protoperidinium spp.) dominate Antarctic waters (confirmed by McKenzie and Cox 1991; Andreoli et al. 1995; Economou-Amilli et al. 1998), with the polar front (PF) being the main biogeographical boundary dictating the high (80–85%) endemism of Antarctic species. Later studies have confirmed the importance of the PF as a major barrier to warmer-water northern taxa (such as Gonyaulax, Oxytoxum, Podolampas and Ceratium spp.; Hermosilla 1977, 1978; Dodge and Priddle 1987; McKenzie and Cox 1991). Unarmoured dinoflagellates (e.g. Gymnodinium spp.) have been considered rare in Antarctic waters (Balech 1959, 1968, 1975). However, recent findings of six new taxa (de Salas et al. 2008), suggest that athecates are more abundant than previously thought. This under-representation is likely the result of potentially destructive sampling and preservation techniques (Balech 1975; Taylor 1978; Steidinger and Tangen 1997).

Whilst Southern Ocean phytoplankton is primarily composed of diatoms and prymnesiophytes (e.g. Arrigo et al. 1999; Fonda Umani et al. 2005), dinoflagellates can dominate in terms of their biomass contribution (Andreoli et al. 1995; Kopczyńska et al. 2007), occupying a crucial position in the microbial pelagic food web as protozooplankton. They can dominate heterotrophic flagellate assemblages (Garrison and Gowing 1993), exerting a significant control on phytoplankton production and thus food web carbon flux (e.g. Fonda Umani et al. 2005).

Dinoflagellates are prominent components of the rich and diverse microbial assemblages inhabiting the various sub-environments created by seasonal ice-formation and decay in Antarctic waters (McConville and Wetherbee 1983; Garrison et al. 1986a; Garrison and Buck 1989, 1991; Lizotte 2003). Stoecker et al. (1991, 1992, 1993, 1998, 2000) have demonstrated that dinoflagellates (especially Polarella glacialis) are an important component in land-fast ice microbial communities. Whilst fast-ice populations usually undergo a seasonal succession from flagellates to diatoms (Stoecker et al. 1998), diatoms typically dominate pack ice year-round (Garrison et al. 1986a; Garrison and Buck 1991; Trevena et al. 2000; Lizotte 2003), although heterotrophic dinoflagellates can also be important (Garrison and Buck 1989). At the ice edge and the highly productive marginal ice zone, protozooplankton dinoflagellates can be abundant (Garrison et al. 1986b; Clarke and Ackley 1994; Kivi and Kuosa 1994).

Whilst significant advances have been made in documenting Antarctic dinoflagellate assemblages and their ecological significance, this group remains understudied. Dinoflagellates are typically less abundant than other groups such as diatoms, nannoflagellates and picoplankton in both the water column and sea-ice habitats (Kopczyńska et al. 2007). As a result, they are rarely identified to species level in broader biogeographical and ecological plankton studies (e.g. Eynaud et al. 1999; Rodriguez et al. 2002; Krell et al. 2005). More often, species are grouped together according to size, trophic level or genus. Consequently, valuable information on individual species distribution and environmental preference has been limited. In particular, dinoflagellates in sea-ice and ice-shelf environments remain understudied, despite recent work focusing on P. glacialis in fast and pack ice (Stoecker et al. 1991; Thomson et al. 2004). Logistical problems inherent to Antarctic sea-ice and ice-shelf studies are further contributing factors. Very little is known of dinoflagellates inhabiting sub-ice-shelf environments in Antarctica. Only one study has been concerned with biota beneath the Amery Ice Shelf (eastern Antarctica) (Roberts et al. 2007). This is surprising given the important role Antarctic ice shelves play in the production of Antarctic Bottom Water and the linkage between ice-shelf stability, global climate and sea-level change (Foldvik and Gammelsröd 1988; Nicholls et al. 1991). A clearer picture of sub-ice-shelf biota (dinoflagellate communities in particular) and its variability would be an important first step in characterizing such environments and ultimately improving the reconstruction of former ice-shelf extent based on microfossils such as diatoms or dinoflagellate cysts.

In this study, we present new relative abundance data on dinoflagellate communities from Drescher Inlet (Fig. 1), a rare fast-ice environment within the extensive Riiser-Larsen Ice Shelf in the eastern Weddell Sea. The inlet, characterized by a cover of multi-year sea-ice and an underlying layer of platelet ice, enables relatively easy access to sea-ice and water column habitats 20 km in from the seaward edge of the ice shelf. Here, we describe the species composition of dinoflagellate assemblages collected from an array of sediment traps deployed in the inlet. We further document the succession of those assemblages during the course of deployment over a period of several days during February 1998, prematurely curtailed by fast-ice break-up. Rich communities of diatoms and metazoans (including amphipods and copepods) have been previously reported from the study site, along with biogeochemistry data (Günther et al. 1999; Thomas et al. 2001; Schnack-Schiel et al. 2004). This study provides greater insight into the variety of biota inhabiting the Drescher Inlet and further adds to our knowledge of otherwise understudied fast ice and ice-shelf dinoflagellate communities.
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0630-5/MediaObjects/300_2009_630_Fig1_HTML.gif
Fig. 1

Map of Weddell Sea and Drescher Inlet. Average minimum sea-ice extent is based on the Sea Ice Climatic Atlas (1985)

Materials and methods

Study site

Samples were derived from an arrangement of multiple and single sediment traps (detailed in Table 1), set to varying collection times (3–15 days; 2–17 February 1998), underneath the ice cover of Drescher Inlet, an approximately 4 km wide and 20 km long opening in the Riiser-Larsen Ice Shelf, eastern Weddell Sea (72°52′S, 19°25′W) (Figs. 1, 2). Both the array and the study site are described in detail by Thomas et al. (2001) and are outlined only briefly here. The study site was a snow covered canopy of 4 m thick perennial fast ice, overlying a thick (>20 m) layer of platelet ice above a 400 m deep water column. CTD profiles showed a thermocline at 150 m which lowered to 280 m during deployment due to changing hydrographic conditions associated with a weather change (Fig. 2). A shift in current direction and strong winds in the latter half of the deployment period eventually led to ice-cover break-up.
Table 1

Details of sediment-trap samples used in this study. Sample code refers to the initial labelling used by Thomas et al. (2001)

Sample

Sample code

Collection date (February 1998)

Collection period (days)

Water depth (m)

Surf-T1

2G2

2–8

6

10

Surf-T2

2G3

8–14

6

10

Surf-T3

2G4

15–17

3

10

Mid1-T0

2K2

2–17

15

115

Mid2-T0

2K1

2–17

15

230

Deep-T1

IG2

2–8

6

360

Deep-T2

IG345

9–17

9

360

https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0630-5/MediaObjects/300_2009_630_Fig2_HTML.gif
Fig. 2

Sediment-trap array, including descriptions of trap arrangement, collection period and collection dates. Temperature and salinity profiles from the CTD at the beginning (3 February 1998) and end of the deployment period (19 February 1998) illustrate the marked changes in hydrographic conditions developed during the course of deployment (after Thomas et al. 2001)

Laboratory methods

Collected sediment-trap material had initially been fixed in hexamine-buffered formaldehyde, sub-sampled for various biogeochemical analyses and time-fractionated by Thomas et al. (2001). Therefore, data on absolute dinoflagellate abundances (i.e. cell concentrations per volume of seawater) typically applied in plankton studies (e.g. Eynaud et al. 1999) are unavailable, although distinctions between samples of high and low biomass can be made from preceding analyses on chlorophyll a (Chl a) concentrations (Thomas et al. 2001). Instead, we used relative abundances (%) in our samples as a source of information on dinoflagellate assemblage composition and succession (for example, similar to Dodge and Priddle 1987; Boltovskoy et al. 1996).

Sample material was initially re-combined and washed through a 10 μm mesh with distilled water in order to remove excess salt. Sample material was agitated, and then systematically examined by inverted microscopy, in most cases scanning the whole of the sample material. As our aim was to identify dinoflagellates to species rather than genus level, we chose to pick out any encountered cells for scanning electron microscopy (SEM) analysis rather than count cells under light or inverted microscopy, which can potentially hinder confident species-level identifications (Taylor 1978). Dinoflagellate cells were picked out from each sample using a glass pipette connected to a mouthpiece and transferred to glass cover slides, which were dried in a dehydration chamber overnight. Subsequently, cover slides were mounted on SEM stubs, sputter-coated in gold and systematically examined by SEM. This methodology allowed us to examine in detail the morphology of our specimens and therefore to confidently identify them to species level. Furthermore, it enabled the comparison of our specimens to the line drawings in our primary reference guide, Balech’s (1975) atlas of Antarctic dinoflagellates. A DVD of all SEM micrographs is available from the lead author on request.

In order to assess assemblage make-up in each sediment trap, we counted approximately 100 specimens from each sample where possible, and subsequently calculated the relative abundances (%) of individual species (similar to the method employed by Boltovskoy et al. 1996). The number of counts is consistent with other studies which recorded between 50 and 100 phytoplankton cells from their sample material (e.g. Dodge and Priddle 1987; Boltovskoy et al. 1989; Höglander et al. 2004). Although recommendations as to the optimal cell count vary greatly (reviewed in Venrick 1978), a count of 50–100 cells has been shown to be sufficiently representative (providing a 95% confidence interval of the estimate within ±20–30% deviation from the mean) by several workers using a variety of methods (Allen 1921; Lund et al. 1958; Utermöhl 1958; Frontier 1972).

Results

Thirteen dinoflagellate taxa were encountered in the sediment-trap suites (Figs. 3, 4; Table 2) with an additional category for unknown species. Fifty to 100 specimens were enumerated from all samples, with the exception of a deep water trap in which dinoflagellates were very sparse (Deep-T1, 19 cells only). The highest number of taxa was found in trap Mid1-T0 (at 115 m; 11 taxa plus unknown), followed by two surface traps (Surf-T2 and Surf-T3, both at 10 m, 9–10 taxa). The number of species encountered was generally relatively high throughout sediment-trap samples (>7 species), but was lowest in Deep-T1 (360 m) (4 species).
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0630-5/MediaObjects/300_2009_630_Fig3_HTML.jpg
Fig. 3

SEM micrographs of selected dinoflagellate species from Drescher Inlet. All scale bars denote 20 μm. Dinophysis spp. in a ventral and b lateral views; cPreperidinium meunieri in apical view; dPreperidinium granulosum in oblique ventral view; ePreperidinium cf. perlatum in oblique dorsal view; fProtoperdinium antarcticum in oblique apical view; gProtoperdinium latistriatum in ventral view; hProtoperdinium rosaceum in oblique apical view; iProtoperidinium bipatens in oblique ventral view; jProtoperidinium applanatum in oblique ventral view; kProtoperidinium cf. charcoti in oblique apical view; lProtoperidinium macrapicatum in ventral view

https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0630-5/MediaObjects/300_2009_630_Fig4_HTML.jpg
Fig. 4

SEM micrographs of two unknown dinoflagellate species from Drescher Inlet. aProtoperidinium sp. A in oblique ventral view, scale bar denotes 10 μm. bProtoperidinium sp. C in dorsal view, scale bar denotes 20 μm

Table 2

List of dinoflagellate species encountered in the present study, including taxonomic designations

Species

Trap

Surf-T1, T2, T3 (10 m)

Mid1-T0 (115 m)

Mid2-T0 (230 m)

Deep-T1, T2 (360 m)

Dinophysis spp.

x

x

  

Preperidinium granulosum (Balech) Elbrächter

x

x

x

x

Preperidinium meunieri (Pavillard) Elbrächter

x

x

  

Preperidinium cf. perlatum (Balech) Elbrächter

x

x

 

x

Protoperidinium antarcticum (Schimper) Balech

x

x

x

 

Protoperidinium applanatum (Mangin) Balech

x

x

x

x

Protoperidinium bipatens Balech

x

x

  

Protoperidinium cf. charcoti (Balech) Balech

x

 

x

x

Protoperidinium latistriatum (Balech) Balech

x

x

x

x

Protoperidinium macrapicatum (Balech) Balech

x

x

x

x

Protoperidinium rosaceum (Balech) Balech

x

 

x

x

Protoperidinium sp. A

 

x

  

Protoperidinium sp. C

x

x

x

x

Presence in sediment traps is denoted by “x”. Data for different collection periods has been combined for each sediment-trap depth. Trap details are outlined in Fig. 2 and Table 1

All encountered dinoflagellates were thecates, with a clear dominance of the genus Protoperidinium, which reached minimal abundances of 55% (Deep-T3, 360 m) and maximum abundances of 98% (Mid2-T0, 230 m), and included two unknown species (Fig. 4). Preperidinium and Dinophysis were present in minor amounts. Due to the fixed collection times, some sediment traps contained time-averaged samples and exhibit no detectable trends. For example, Mid1-T0 and Mid2-T0 (Fig. 5a), which were collecting throughout the whole deployment, were both dominated by Protoperidinium macrapicatum (both ~47%) with lesser occurrences of Protoperidinium sp. C (both ~10%). Trap Mid2-T0 also showed a relatively high occurrence of P. rosaceum (27%), which was absent in trap Mid1-T0.
https://static-content.springer.com/image/art%3A10.1007%2Fs00300-009-0630-5/MediaObjects/300_2009_630_Fig5_HTML.gif
Fig. 5

Changes in dinoflagellate species composition in sediment-trap samples a across the thermocline, b with depth during the first half of the deployment period, c with time, showing assemblage changes at 10 m water depth during the first half (Surf-T2) and latter half (Surf-T3) of the deployment period

Other traps showed differences with depth and time. For example, Surf-T1 and Deep-T1, which represent sample collection for the same time period (2–8 February; Table 1), but at distinct depths (10 vs. 360 m), exhibited a different species composition (Fig. 5b). The shallow water trap (Surf-T1) was dominated by P. antarcticum (42%), with lesser occurrences of P. rosaceum (23%) and P. macrapicatum (15%). In contrast, the latter species was most abundant in deeper water (Deep-T1) (37%), along with Preperidinium granulosum (32%), which was absent in the shallower trap. The deep trap also showed an abundance of P. applanatum (26%), which was only moderately represented at 10 m (4%).

Temporal changes in species composition were evident in two shallow water traps (Surf-T2 and Surf-T3) which recorded dinoflagellate assemblages during different time periods (8–14 and 15–17 February; Fig. 5c; Table 1). Surf-T2 exhibited an abundance of P. antarcticum (28%) and P. macrapicatum (18%), with lesser Protoperidinium sp. C (16%), P. rosaceum (12%) and P. meunieri (9%). In contrast, trap Surf-T3, which recorded assemblages during the later period of deployment, was dominated by Protoperidinium sp. C (45%), with lesser P. antarcticum (20%). P. meunieri (15%) and P. macrapicatum (12%) were moderately abundant while P. rosaceum was entirely absent.

Discussion

Drescher Inlet samples showed generally well preserved and relatively diverse dinoflagellate communities. The genus Protoperidinium dominated all Drescher Inlet assemblages. This finding is consistent with other Antarctic studies that highlight the dominance of this taxon in the Southern Ocean. In the Australian sector, Protoperidinians have been found in abundance close to Antarctica during austral spring (Kopczyńska et al. 2007) and summer (Waters et al. 2000), being numerically significant contributors to the overall plankton population in both water column and fast-ice environments, and constituting nearly two-thirds of the whole dinoflagellate community (Waters et al. 2000). In the Atlantic sector, Protoperidinium dominates dinoflagellate communities across the Weddell and Scotia seas (Balech and El-Sayed 1965; McKenzie and Cox 1991; Klaas 2001). Further east in the same sector, Froneman (2004) demonstrated that Protoperidinians are prominent at the ice edge both during winter and spring, as well as across the Antarctic PF. In the Ross Sea (Pacific sector), the genus was found to make up 55% of the whole dinoflagellate assemblage (Andreoli et al. 1995). The occurrence of Protoperidinians is not restricted to the water column and ice edge, however. Protoperidinium spp. have been reported from fast ice in the Ross (Stoecker et al. 1993) and Davis seas (Thomson et al. 2004) as well as from Weddell Sea pack ice (Garrison and Buck 1989), though they were outnumbered by other dinoflagellate taxa.

Whilst Drescher Inlet samples demonstrate the prominence of Protoperidinium similar to other Antarctic studies, specific assemblage comparisons are difficult, as species-level identification of dinoflagellates is not commonly practiced in general plankton studies (e.g. Eynaud et al. 1999, Rodriguez et al. 2002, Krell et al. 2005). Our comparisons concerning the assemblage composition are thus mostly restricted to previous studies focused specifically on dinoflagellates. The 13 taxa found in the Drescher Inlet correspond well to dinoflagellates previously reported from the Atlantic sector. Balech (1958) found abundant Protoperidinians such as P. macrapicatum, P. antarcticum and P. rosaceum in plankton tows in the vicinity of the South Orkney and South Sandwich islands, also noted to be relatively abundant in the present study. Similarly, a variety of dinoflagellates encountered in the Drescher Inlet (P. applanatum, P. antarcticum, P. minor, P. perlatum, P. granulosa, Dinophysis spp.) have been found in the vicinity of the South Shetland Islands (Hermosilla 1977, 1978). In the southern and western Weddell Sea, Balech and El-Sayed (1965) encountered Protoperidinians (P. antarcticum, P. rosaceum, P. applanatum) as well as P. granulosum (then called Diplopsalis granulosa), taxa common in the present study. The rich and diverse dinoflagellate assemblages dominated by P. antarcticum and P. applanatum reported from the southern Weddell Sea by Estrada and Delgado (1990) are also similar to the Drescher Inlet assemblages. In the present study, however, these two species exhibited a preference for specific depths: P. antarcticum was more abundant in surface waters, whilst P. applanatum was more frequent in deep water traps. This depth preference was not noted in the earlier study of Estrada and Delgado (1990).

The high relative abundance of Dinophysis antarctica prominent around South Georgia (Dodge and Priddle 1987) stands in contrast to the Drescher Inlet assemblages which show Dinophysis as a minor assemblage component. Assemblage make-up depends on factors such as seasonality and location. McKenzie and Cox (1991) found distinct inter-seasonal and regional differences in dinoflagellate assemblages in the Scotia and Weddell seas and the PF region, with a clear dominance of Protoperidinium (P. applanatum, P. rosaceum, P. turbinatum) south of the PF and warmer-water taxa (Gonyaulax, Oxytoxum, Ceratium) preferably dwelling north of the PF. These North-South assemblages also changed in terms of their species composition according to season (summer to autumn). The dinoflagellate assemblages encountered around Deception Island (South Shetland Islands) also exhibited a variation in species composition on a year-to-year basis (Hermosilla 1977, 1978). Plankton studies provide a snap-shot of the population at a specific point in time. Dinoflagellates absent or present at low relative frequencies in this study may be prominent at a different time or season.

The majority of dinoflagellates found in Drescher Inlet were heterotrophic Protoperidinians, which, in their role as protozooplankton, graze on diatoms and other primary producers (Froneman et al. 1996). Sediment-trap material did contain diatoms, in particular two species primarily associated with sea-ice and marginal ice zone environments [Fragilariopsis curta (van Heurck) Hustedt 1958 and F. cylindrus (Grunow) Krieger in Helmcke and Krieger 1954]. The low abundance of diatoms in the water column and the plenitude of small (ca. 100 μm) faecal pellets in the Drescher Inlet traps led Thomas et al. (2001) to conclude that the diatoms were likely derived from the platelet layer beneath the fast-ice canopy in the inlet, and that they had been grazed by protozooplankton. Heterotrophic dinoflagellates such as Protoperidinium spp. have been shown to efficiently graze diatoms, with grazing rates of up to 0.318 μl cell−1 h−1 in Eastern Antarctica, where heterotrophic dinoflagellates ingested 29.7 ng Chl a l−1 h−1 (Archer et al. 1996). The dominance of heterotrophic dinoflagellates reported here suggests that they may have been grazing on diatoms previously reported from Drescher Inlet.

Although present in other Antarctic studies (Balech 1958, 1975; Garrison and Buck 1989; Kang and Fryxell 1993; Stoecker et al. 1993; Scharek et al. 1994), we did not find any athecate taxa in our samples. Traditionally, athecates have been considered rare in Antarctic waters, either due to low abundances, or due to sampling, preservation, or fixation techniques (Balech 1975). Delicate unarmoured dinoflagellates can be lost due to net sampling techniques (Kopczyńska et al. 1986) whilst plankton assemblages can vary considerably according to sampling by tows or sediment traps (Boltovskoy et al. 1996). Preservation and fixation media can also affect sampled material, as no single preservation medium will optimally preserve all plankton groups (reviewed in Throndsen 1978). Lugol’s solution has been reported to cause cell shrinkages (Leakey et al. 1994) and may be an unsuitable preservative for athecates (Steidinger and Tangen 1997), whilst glutaraldehyde may better preserve external ornamentation and internal structure of plankton cells compared to hexamine-buffered formalin (Kang and Fryxell 1993). Nevertheless, naked dinoflagellates such as Gymnodinium and Gyrodinium have been reported from material preserved with a variety of media such as formalin (Balech 1958; Balech and El-Sayed 1965; Dodge and Priddle 1987), buffered formalin (Beans et al. 2008), Lugol’s solution (Garrison and Buck 1989; Estrada and Delgado 1990; Andreoli et al. 1995; Archer et al. 1996; Froneman 2004) and glutaraldehyde (Kopczyńska et al. 2007).

Samples analyzed in the present study were preserved with hexamine-buffered formaldehyde, a fixative which has been used in other studies which encountered unarmoured dinoflagellates (e.g. Fonda Umani et al. 2005). Thus, it seems likely that we should have found athecates in our samples if present in Drescher Inlet. However, because the delicate walls of athecates can potentially be destroyed or deformed by formaldehyde (Taylor 1978), the possibility of post-sampling loss of naked taxa in our samples cannot be ruled out with certainty. If this is the case, Drescher Inlet assemblages should be biased towards more robust armoured dinoflagellates. It should be noted that most recent research on Antarctic athecates demonstrating high abundance and variety, has been based on cell cultures derived from seawater samples (Gast et al. 2006, 2007; de Salas et al. 2008). Such methods rather than sediment traps or plankton tows may prove the optimal way to account for the true abundance of these fragile organisms.

The absence of dinoflagellate cysts in the present study stood in contrast to the Bellingshausen, Weddell and Scotia seas, where abundant dinocysts have been reported from sediment-trap material (Harland and Pudsey 1999). The lack of dinocysts in the Drescher Inlet samples may be due to the short deployment period (max. 15 days this study vs. 1–2 years for Harland and Pudsey 1999), the different positioning of traps (throughout the water column vs. nepheloid layer) and contrasting sea-ice regimes (year-round fast ice vs. seasonally ice-free). Only one species of dinoflagellate encountered in Drescher Inlet, P. meunieri, is hitherto known to produce cysts (cyst affinity Dubridinium capitatum, Marret and Zonneveld 2003). However, Protoperidinum spp. are known to produce a suite of dinocysts found in high latitudes (e.g. Islandinium minutum, Islandinium cezare), although the detailed, species-level affinities are not yet established (Marret and Zonneveld 2003). It is therefore possible that some of the Protoperdinium spp. encountered in the present study were cyst producers, but that the sampling period may have been too short to document any seasonal period of cyst production.

Drescher Inlet assemblages exhibited differences in dinoflagellate composition and relative abundances, variations occurring both according to depth and collection period. Several species that had been present during the first half of the deployment were absent in the later period and vice versa in shallow water (P. antarcticum vs. Protoperidinium sp. C). Differences in species composition were evident with depth: P. antarcticum and P. rosaceum were abundant in shallow water; P. macrapicatum and P. granulosum dominated deep water assemblages. Thomas et al. (2001) recorded a difference in Chl a flux between sediment traps, and thus collection times and depths, with Chl a values in shallow water nearly doubling in the latter half of the deployment period, and higher Chl a levels in deeper water. This may imply temporal and spatial changes in algal biomass, similar to other Antarctic sediment-trap studies (Leventer and Dunbar 1987; Harland and Pudsey 1999). Hydrographic factors such as water stability, current, and ice dynamics influence phytoplankton abundances and community structures to a large degree (e.g. Kopczyńska 2008). The assemblage structure changes observed in Drescher Inlet dinoflagellates may therefore reflect a community response to sea-ice break-up and to different local water masses. However, in light of the limited data, the re-combination of samples, and the short deployment period in our study, any drawing of definite causal relationships underlying these apparent changes must remain speculative.

The question of whether the dinoflagellates found in the present investigation primarily occupied the water column, or whether they were derived from a flux out of the fast ice and the ice platelet layer, is difficult to answer. At the study site, the highest concentration of diatoms as well as metazoans (such as amphipods and copepods) has been reported from the platelet ice layer (Günther et al. 1999; Schnack-Schiel et al. 2004). If the encountered dinoflagellates were grazing on the diatoms previously reported from the study site, it is likely they would be found in the platelet ice where their prey is most abundant. The only study to date on sub-ice-shelf biota (Amery Ice Shelf; Roberts et al. 2007) documented abundant diatoms contained within newly formed platelet ice. However, only one dinoflagellate genus (Preperidinium) was recorded from this environment in extremely low abundance. A direct comparison between Roberts et al. (2007) and our study is not meaningful, as the Drescher Inlet samples are from beneath multi-year sea-ice (4 m thick) as opposed to accumulating marine ice beneath several hundred metres of ice shelf.

Heterotrophic dinoflagellates have been found living in pack (Garrison and Buck 1989), and fast ice (Stoecker et al. 1998). Though recent molecular studies (Gast et al. 2004) found discrepancies between the assemblage structures of microhabitats such as slush, ice and water column, the consensus is that micro-organisms found in ice are also common in the water column and vice versa due to passive introduction through ice formation and decay (e.g. Garrison and Buck 1989). Larger protozoans, such as ciliates and heterotrophic dinoflagellates, have also been found capable of entering brine channels in high porosity ice from the water column, and to colonise the brine pockets in the upper ice (Stoecker et al. 1993). Given their high mobility, it is possible that the dinoflagellates encountered in the present study would be thus able to migrate between differing media (ice/platelet/water) to seek out their food. A further possibility is that occupation of ice versus water habitat is species specific. Our data, however, do not appear to support this hypothesis, as there are no detectable temporal or spatial trends in any specific species, which could be associated with defined sub-environment occupation.

Our study provides a snap-shot of the diverse dinoflagellate assemblages present in Drescher Inlet. Dinoflagellate communities occupying sub-ice-shelf environments, such as the one described in the present study, remain poorly documented. It is regrettable that our study is limited to one site within the Ronne-Filchner Ice Shelf. Nevertheless, our data provides an important initial description of such ice-shelf environment dinoflagellate communities and represents a necessary first step for further research. More work is needed to assess whether Drescher Inlet assemblages are typical of fast-ice covered ice-shelf inlets. If detailed enough (i.e. species-level identifications), further investigations should not only facilitate inter-study comparisons, but should also elucidate the driving factor(s) responsible for any temporal and spatial changes in dinoflagellate assemblages. Further research in Drescher Inlet (and similar environments) should also clarify any seasonal assemblage shifts, and may additionally determine whether athecates occur at such sites.

Acknowledgments

We thank J. Plötz and H. Bornemann and the captain and crew of the RV Polarstern for their support in the field. I.A.N. Lucas and G. Walker are thanked for SEM assistance. We also wish to acknowledge assistance with species identifications by K. Zonneveld and J. Lewis. We thank B. Long for assistance with lab work and M. Furze for proofreading and help with diagrams and figures. This work was partially funded by the Natural Environment Council, British Council/DAAD, the Royal Society and the Nuffield Foundation. We are grateful to Oliver Esper and one anonymous reviewer whose comments improved the manuscript.

Copyright information

© Springer-Verlag 2009