The data reported were collected during the cruise of the Korean research icebreaker Araon in the Amundsen Sea from December 2010 to January 2011. A Niskin bottle rosette equipped with conductivity—temperature—density probes was used to sample a total of 30 stations and plankton net tows were performed to assess microplankton community composition at 15 of the 30 stations sampled (Fig. 1). Stations were located in both deep offshore waters and polynya areas, considered as coastal areas of open water (free of sea ice) within the sea ice zone; areas of sea ice and concentrations were based on data from the National Snow and Ice Data Center in Boulder, Colorado, corresponding to the cruise period. The classification of stations used here follows that established in Lee et al. (2012).

For chlorophyll a determinations, water samples of 0.3–1 L were obtained from 7 to 9 discrete depths between the surface and 150 m depth. Water samples were filtered through a 0.7-µm grade GF/F glass fibre filter (Whatman, Kent, UK) and chlorophyll concentrations were determined on-board using a Trilogy fluorometer (Turner Designs, Sunnyvale, CA, USA) calibrated against commercial chl a standards.

Net tows were made with a 20-µm plankton net, 0.45 m in diameter, equipped with a calibrated flowmeter, towed from 150 m depth to the surface. A portion of the net tow material was fixed with Bouin's solution (6% final concentration) for direct microscopic examination. In the laboratory, aliquots of net tow material were sequentially diluted and examined in settling chambers using an inverted microscope equipped with differential interference contrast optics. The entire surface of the settling chamber was examined at 200× total magnification.

Species identifications were made using the monographs of Kofoid & Campbell (1929) Kofoid , 1939), Hada (1970), Balech (1973) and Alder (1999). Total sample volumes examined varied, depending on the concentrations of tintinnids and the dilutions employed varied according to concentrations of phytoplankton. However, multiple aliquots were examined until at least 50 tintinnids were enumerated for each station (tintinnids enumerated per station ranged from 52 to 358; average 172±139). The sample volumes examined were equal to 1–7 ml of the original 350 ml total net tow material, representing material from about 10–400 L filtered by the net. Nominal concentrations of organisms were calculated based on the apparent volume filtered by the net (as indicated the by net flowmeter) and the volume of net tow material examined. It should be noted that the concentrations reported here are the average values per litre for the top 150 m of the water column. In general, relatively few cells are found below the mixed layer (e.g., Dolan & Marassé 1995). Consequently, the concentrations of cells in the mixed layer were likely several times higher than those reported here, which represent the entire 150 m water column. For several stations, we also enumerated copepod nauplii as a proxy measure of copepod abundance. Simple linear regression was used to investigate the relationships between concentrations (log transformed) of tintinnids and chlorophyll as well as occurrences of different morphotypes.

The physical and biological characteristics of the offshore deep water stations differed considerably from those of the coastal polynya stations. The polynya stations, compared to the non-polynya, primarily offshore sites, were characterized by shallower mixed layer depths (Table 1), considerably higher chlorophyll a concentrations with a distinct chlorophyll maximum, and high primary production (Table 2 and Fig. 2). The offshore stations were comparatively low in concentrations of chlorophyll and showed more even vertical distributions of chlorophyll. The phytoplankton dominating the microplankton communities was distinctly different also. In the deep water offshore stations, the microplankton was dominated by diatoms while in the polynya stations, microplankton was dominated by Phaeocystis (Table 2 and Fig. 3). It should be noted that our observations concerning phytoplankton composition refer to the microplankton community, not the entire phytoplankton community.

A total of 15 species with a wide range of morphologies were encountered in the samples, as well as a variety of morphotypes of two Antarctic species, Codonellopsis gaussi and Cymatocylis affinis/convallaria (Table 3 and Fig. 4). Typically, low species-richness (2 to 4 species) characterized the polynya assemblages compared to the non-polynya assemblages (4–10 species). Furthermore, polynya assemblages were generally dominated by species endemic to the Southern Ocean (largely forms of Cymatocylis affinis/convallaria and Condonellopsis gaussi) while the offshore assemblages included many widespread tintinnid taxa (Table 3).

Overall, the abundance of tintinnids was clearly not related to bulk chlorophyll concentrations (Table 2). However, plotting chlorophyll data by the dominant type of phytoplankton (diatom, Phaeocystis, or both) suggested that within Phaeocystis-dominated communities, tintinnid abundance increased with chlorophyll concentrations (Fig. 5). Note that we have no data on the abundance of nano-phytoplankton such as the single-cell forms of Phaeocystis or other probable major prey items for tintinnid ciliates.

We found no clear relationship between the composition of the micro-phytoplankton and the occurrence of different morphologies of Condonellopsis gaussi and those of Cymatocylis affinis/convallaria. Of the many different “forms,” or morphotypes, encountered (Fig. 4), most were found in both the diatom-dominated offshore stations and the Phaeocystis-dominated polynya stations (Table 3). There was also no relationship between the number of morphotypes co-occurring and the concentration of chlorophyll, regardless of the identity of the taxa dominating the phytoplankton. However, there was a weak but significant positive relationship between the number of different morphotypes found in an assemblage and the overall concentration of C. affinis/convallaria (Fig. 6). There was also a positive trend between the numbers of C. gaussi morphotypes or forms found and the overall concentration of C. gaussi but not a significant linear relationship.

We documented a species-richness for the Amundsen Sea far in excess of that known for the adjacent Bellinghausen Sea (Table 4). There is no doubt that our results reflect the fact that we examined material from large volumes of water sampled with a relatively fine-mesh net. Interestingly, most (nine out of the 15 species) were “widespread” species, found in a variety of systems ranging from the Mediterranean Sea (Dolan 2000) to the Pacific Ocean (Dolan et al. 2007), and all were species which have been previously recorded from other Southern Ocean locations, if not the Bellinghausen Sea (see Dolan et al. 2012). Therefore, none of our species records for the Amundsen Sea are new for the Southern Ocean. However, we did establish for the first time the existence of distinct tintinnid communities in polynya and offshore deep water based on our data concerning the Amundsen Sea.

At the polynya stations, where the microplankton community was dominated by Phaeocystis, the tintinnid assemblage was a particular subset of the species catalogue. Only four species were found and three of the four are species endemic to the Southern Ocean: Cymatocylis affinis/convallaria, Codonellopsis gaussi, Laackmanniella prolongata. These Southern Ocean species are all relatively wide-mouth species, with LODs measuring from 40 to 100 µm, presumably feeding most efficiently on prey much larger than those exploited by most of the other species found in the Amundsen Sea (see Fig. 4). While there were few data points, it was only among the polynya stations that there appeared to be a relationship between the concentrations of chlorophyll and tintinnids (Fig. 5). However, it is unclear if the relationship could reflect a predator—prey relationship between Phaeocystis and tintinnids. Typical planktonic ciliates, such as oligotrichs are unable to graze on colonies of Phaeocystis and ingest only the individual free-swimming cells (e.g., Verity 2000). However, tintinnids can reportedly ingest cells from the outer layer of colonies of Phaeocystis, crop individual cells, as well as feed on the free-swimming stage (Admiraal & Venekamp 1986). An alternative, but not exclusive, explanation for a positive relationship between Phaeocystis concentration and tintinnid abundance is that predation on tintinnids is lower in such communities due to either lower predator abundance or lower predator feeding efficiency with increasing concentrations of Phaeocystis. Many polynya areas are known to be characterized by the relatively low abundance of zooplankton (e.g., Tagliabue & Arrigo 2003) and this appears to be the case for Amundsen Sea polynya also (Yager et al. 2012).

We do not have data on the abundance of predators of tintinnids in our polynya compared to non-polynya stations, only some indirect evidence in the form of copepod naupliar concentrations. Our data on copepod nauplii abundance (Table 2) suggests little difference between polynya stations and off-shore stations. However, this is weak evidence of no difference in top—down control of tintinnid composition. Regardless of the mechanism influencing tintinnid species composition found in Phaeocystis-dominated sites, the projected increase in the relative importance of Phaeocystis in the Amundsen Sea (Alderkamp et al. 2012) may be accompanied by increases in the relative importance of the endemic Antarctic species (i.e., Cymatocylis, Laackmaniella) at the expense of the widespread species found in the diatom-dominated microplankton communities.

We were also interested in examining the phenomenon of polymorphism in the species Cymatocylis affinis and Codonellopsis gaussi (Fig. 4). Our working hypothesis was that distinct morphologies represent adaptations to distinct conditions. However, we found no evidence that particular morphologies are associated with particular phytoplankton communities. A near full-range of morphologies was found in microplankton communities dominated by diatoms as well as communities dominated by Phaeocystis (Table 3). The number of morophotypes found in individual samples was also unrelated to the number of cells enumerated (data not shown). Also, the number of different morphologies found in an assemblage appeared unrelated to the concentration of chlorophyll, regardless of the type of phytoplankton assemblage (Fig. 6a, c).

An alternative explanation for polymorphism is that rather than representing adaptations to a particular environment, the different morphologies found are simply different developmental stages. In this regard, it should be noted that in tintinnids, which reproduce by cell division, the lorica is retained by the newly formed daughter cell, the opisthe, and the cell with the parental mouth, the protor, swims off and forms a new lorica (see Agatha et al. 2012). Consequently, in a population of tintinnids there are both “old” and “new” loricas. The loricas inhabited by many of the tintinnids in a population were formed not by the current inhabitants but rather by ancestral cells and not necessarily direct ancestors but possibly several generations removed.

We speculate that a population of tintinnids, if it is relatively long-lived and experiences rapid bursts of population growth, is liable to contain loricas formed by cells which divided before lorica formation was complete. These partially formed loricas could be retained in the population, inherited by successive generations. Unfortunately, we cannot directly examine this possibility. However, if polymorphism simply reflects different developmental stages, we reasoned that dense populations would exhibit more polymorphism than sparse populations as dense populations are more likely to show characteristics of rapid population growth and thus contain more “immature individuals” corresponding to different morphotypes or forms. We found some evidence in support of the hypothesis as the concentration of Cymatocylis affinis/convallaria was positively related to the number of morphotypes in the population (Fig. 6b). A positive (although not significant) trend was also apparent in Condonellopsis gaussi (Fig. 6d). Note that these trends were not dependent on the overall number of individuals examined for each station.

We found a high species-richness in the Amundsen Sea compared to the better known adjacent Bellinghausen Sea. We compared, for the first time, the species assemblages of coastal polynya sites, projected to increase in relative importance in the future, with offshore deep water sites. The polynya sites of high primary production dominated by Phaeocystis, show a lower species richness and are formed mainly of large-mouthed tintinnid species endemic to the southern Ocean. In the polymorphic species Cymatocylis affinis/convallaria and Codonellopsis gaussi, the number of types or forms found co-occurring appeared related to the overall concentration of the species, suggesting that the different morphologies may simply be developmental stages.