Due to climate change, the Arctic is one of the fastest changing environments in the world (Overpeck et al., 1997; McBean et al., 2005; Rantanen et al., 2022). This has already affected the Arctic biosphere and will lead to further changes in the future (Hoegh-Guldberg and Bruno, 2010). The base of the complex marine pelagic food web consists of unicellular organisms, such as bacteria and eukaryotic unicellular plankton (protists) occupying different ecological niches and providing food for higher trophic levels.

Because of their crucial role in the marine ecosystem, marine protists are frequent study objects. Community studies of Arctic pelagic waters often focus on transect or snapshot studies (Baggesen et al., 2012; Tillmann et al., 2014; Elferink et al., 2017, 2020), which do not properly display the temporal dynamics. To our knowledge, only a limited set of studies investigated seasonal dynamics, focusing on mesopelagic water species (Terrado et al., 2009), or seasonal sea ice and its impact on the protist community (Terrado et al., 2011; Massicotte et al., 2020). Until now, the pelagic winter protist community in the Arctic has been characterized as most likely heterotrophic (Marquardt et al., 2016; Kubiszyn et al., 2017). Some studies indicate that phototrophic diatoms, which are typical spring bloom formers in the Arctic, may survive unfavorable conditions as resting stages (Zhang et al., 1998; McQuoid and Hobson, 2008; Hegseth and Tverberg, 2013; Kvernvik et al., 2018). Other studies have found that diatoms can recover their light harvesting abilities very quickly when it becomes brighter (Lacour et al., 2019), which hints at their activity also during the dark winter period.

The periods with ice cover have been declining during the past decades due climate change. This is expected to impact the timing and dynamics of the spring bloom, and the trophic modes of the protist community (Alexander and Niebauer, 1981; Hunt et al., 2002). Phytoplankton blooms have occasionally been found to develop before the sea ice melts (Arrigo et al., 2012; Spall et al., 2014; Massicotte et al., 2020), and some studies have recognized the abundance of parasitic and mixotrophic protists in the presence of sea-ice (Terrado et al., 2011; Clarke et al., 2019; Søgaard et al., 2021). The seeding of the pelagic phototrophic spring bloom event by sea ice algae has also been considered as an option (Ardyna et al., 2020). While the pattern of phototrophic dominance during the spring bloom event is comparatively well-described (Tammilehto et al., 2017; Lafond et al., 2019; Luostarinen et al., 2020), the community structure of the winter community and its transition toward the vernal bloom is less investigated (Kubiszyn et al., 2017), especially in relation to seasonal sea ice. With this study, we aim to provide insights into the temporal changes in protist community composition from winter to spring, generating a better understanding of the Arctic marine ecosystem. Therefore, we discuss the impact of the occurrence of seasonal sea ice and other abiotic parameters in their interplay with the protist community composition transition from winter to spring in Disko Bay, with special focus on the functional (trophic) groups of the observed organisms.

The method of size fractionation might have impacted the integrity of more fragile cells, which could have fragmentized under the pressure of the vacuum filtration. The reads of a very few fragile larger taxa such as Strombidium spp. (Figure 4A) in the picoplankton size fraction may have been the result of this method. On the other hand, these findings could also hint at the presence of considerably smaller gametes. Additionally, this method allows, e.g., for insights in seasonal colony formation of the haptophyte Phaeocystis pouchetii, and was successfully applied in other field studies several times (Krock et al., 2009; Elferink et al., 2020; Bruhn et al., 2021), which was the reason why it was used for this study.

The winter communities were dominated by parasites, heterotrophs and mixotrophs during February (Figure 4A). In more temperate coastal regions, where more light is available, small heterotrophic protists were also dominating the winter population (Morán et al., 2018). However, especially the picoplankton and nanoplankton size fractions revealed a high relative abundance of parasitic organisms during winter, and not only general heterotrophs. At times, the picoplankton fraction consisted almost entirely of parasites and heterotrophs, which underlines the importance of these two trophic modes for the winter community. Most marine parasitic protists are relatively small and target considerably larger cells as host organisms (Gómez et al., 2009; Alacid et al., 2015; John et al., 2019), indicating that most of the parasitic protists detected in the study were probably in their free-living stage, showing up in the picoplankton fraction. Very few parasites were detected in the microplankton fraction, further supporting the conjecture that few of the parasites were inside microplankton host cells, unless these cells were broken up by, or external parasites that have been dislodged during the filtration process. On the other hand, the low abundance of reads assigned to parasites in the microplankton size fraction could also be accounted for by a higher DNA content of larger cells, which in turn could dominate over the comparably low relative abundance of endoparasites in the PCR-based method. In Antarctic waters, parasitic protists have been detected as being surprisingly prevalent in winter (Cleary and Durbin, 2016), probably associated with the sea ice lead, i.e., long openings in the sea ice cover (Clarke et al., 2019). Most of the parasitic organisms were dinoflagellates, specifically Syndiniales. These often infect ciliates, dinoflagellates, cercozoons, and crustaceans (Guillou et al., 2008), i.e., groups of mixotrophic and heterotrophic organisms, but apparently not or only rarely diatoms (Tillmann et al., 1999). In Disko Bay, heterotrophic and especially mixotrophic dinoflagellates were detected in all size fractions. The overall biomass (assessed as POC) was, however, extremely low (Table 1). Little is known about the autecology of parasitic dinoflagellates in the ocean, in particular because they are difficult to maintain in culture. Parasitic protists usually do not stay alive for prolonged periods of time without their host organisms, and they typically complete their free-living stages within a few hours to days (Alacid et al., 2015; Reñé et al., 2017; John et al., 2019). The existing laboratory experiments suggest that they are not fit to live without their host organisms for an extended period of time (Alacid et al., 2015; John et al., 2019). Resting spores as an overwintering strategy for parasites have not been described yet, although such a strategy is a possibility (Guillou et al., 2008). It is possible that the parasitic organisms observed were simply very successful in finding their host organisms and completing their life cycles with an output of many new individual cells (dinospores), but we cannot exclude alternative survival strategies.

The presence of mixotrophic organisms, mainly CMs, may be related to them having had an advantage over organisms which are less flexible in their trophic mode, because they gain energy from both harvesting the little light available and additional food uptake.

Also later, during the early stages of the spring bloom, mixotrophs, especially CM dinoflagellates, contributed substantially to the total photosynthetic protist community in the pico- and nanoplankton size fractions (March). This may have been a response to the slightly increased day length (Figure 2B), although the light reaching into the water was still negligible (Figure 1A). Similar observations in the community structure have been made in the Young Sound fjord in Northeast Greenland. Here, a bloom of mixotrophic haptophytes developed in ice covered surface waters during early spring (Søgaard et al., 2021). The two locations differ considerably with regard to salinity and nutrient concentrations. Nevertheless, mixotrophs seemed to have had an advantage at both locations, because they compensate for low levels of photosynthesis with their ability to ingest other organisms. The mixotrophic ability seems to give them the flexibility to quickly adapt to increasing light availability, thereby giving them an advantage over pure photoautotrophs at this seasonal time point. It is even possible that mixotrophy dominates the pelagic food web during much of the year in the Arctic, due to this increased persistence (Stoecker and Lavrentyev, 2018).

Chlorophyll a concentration followed an exponential growth curve throughout the sampling period (Figure 2A). Additionally, an increase of phototrophs in relative abundance in all size fractions can be seen (Figure 4A). These two changes over time showed the spring bloom initiation. A large relative abundance of photosynthetic diatoms, especially in the nanoplankton and microplankton size fractions, mainly characterized the spring bloom community in April. In the dark winter period in the Arctic, light as the primary source of energy for phototrophs is naturally lacking, while inorganic nutrients are sufficient. One known possible overwintering strategy for diatoms are resting spores, which can germinate when the conditions are more favorable (McQuoid and Hobson, 1995; Tsukazaki et al., 2019; Luostarinen et al., 2020). Another strategy for fast adaptation to better conditions of phototrophs, mainly diatoms, is the quick photosynthetic reactivation of resting cells after a period of darkness (Kvernvik et al., 2018), as dormant cells only display a much-reduced photosynthetic capacity. The presence of diatoms throughout all months, although at times in small proportions, suggests the utilization of the latter or both strategies. As stated before, diatoms are usually not the primary target of the parasitic Syndiniales. Diatoms seem to combine the advantages of the ability to photosynthesize and efficient nutrient uptake. Being r-strategists, surviving as resting cells and with not being targeted prominently by parasitic Syndiniales organisms, they might be avoiding much biotic pressure at the start, thus possibly giving them the critical advantage for overgrowing the other organisms both proportionally and in absolute abundance, leading to the spring bloom event.

Diatoms are typical spring bloom organisms and are often the dominant group in Arctic spring blooms (Tammilehto et al., 2017; Krause et al., 2018; Lafond et al., 2019; Bruhn et al., 2021). The genera Thalassiosira spp. and Navicula spp. Have previously been detected as important spring bloom species in the Baffin Bay area, not far from the sampled position, although much later in the year and two years prior in 2016 (Lafond et al., 2019). Porosira glacialis is also a cold-water diatom, commonly found in the northern hemisphere (McMinn et al., 2005; Svenning et al., 2019), and it was also one of the dominating phototrophs in the microplankton size fraction (Figure 5C).

The haptophyte Phaeocystis spp., generally P. pouchetii, is an important Arctic phytoplankton species especially toward the late spring bloom and summer (Marquardt et al., 2016), but can also be found in Arctic waters during the entire winter (Vader et al., 2015; Marquardt et al., 2016). On occasion, they have been detected in under-ice blooms as well (Pavlov et al., 2017; Ardyna et al., 2020). Phaeocystis spp. are considered a less desirable food source for zooplankton compared to other phytoplankton taxa (Weisse et al., 1994; Nejstgaard et al., 2007). In the presented study, Phaeocystis pouchetii started as solitary cells in February and March (in the picoplankton fraction) making them potential prey for microplankton (Figure 5E). This is in accordance with findings, e.g., in Svalbard (Hegseth et al., 2019), where Phaeocystis spp. also was initially present as single cells and progressing to colonies later on. Later in April, toward the bloom, it started to form larger colonies. This is similar to Wassmann et al. (2005) where P. pouchetii started forming colonies during sea ice retreat. The colony formation observed here may have been a defense mechanism against smaller copepod species (Nejstgaard et al., 2007). While larger copepods, such as Calanus spp., are able to graze on these colonies, Phaeocystis spp. do not appear to make up a significant part of Calanus spp. diet (Ray et al., 2016). Phaeocystis spp. have an advantage over diatoms because they are not dependent on silicate concentrations, which diminish quickly during the spring bloom (Bruhn et al., 2021). Compared to some other Arctic phytoplankton species, Phaeocystis spp. have a wider tolerance toward temperature, as they are also commonly found in the Atlantic (Hoppe et al., 2018), but may only be able to compete against temperate species to a limited degree (Ahme et al., 2023). This increased fitness makes them a possible candidate for gaining importance in the Arctic spring bloom event in the future. We can confirm the presence of P. pouchetii in the winter community in Disko Bay, as also shown close to Svalbard (Vader et al., 2015), underlining the far distribution of this predominantly Arctic species.

The diversity analyses based on metabarcoding and resulting ASVs showed that the community in winter was generally more diverse than toward and during the spring bloom event (Figure 4B). Interestingly, the smaller the organisms, the more similar the communities of the different months were in terms of presence or absence of ASVs (Figure 3). The largest community differences were thus seen in the microplankton size fraction, in which only 21.8% of ASVs were shared among all sampling months. These findings are similar to a comparative study of ASVs from Iceland and Greenland (Elferink et al., 2020), in which the microplankton size fraction was most dissimilar in the different regions compared to smaller size fractions. These findings are related to plasticity and their boundaries of species and ecotype forming (Wohlrab et al., 2018; Elferink et al., 2020). Smaller species have smaller genomes and higher generation times and therefore might evolve faster and differentiate more rapidly into distinct ecotypes, i.e., different ecotypes but same species. Hence, the same species can progress over the seasonal transition, as in this study we do have more species but less changes over time providing a different strategy of some adaptational flexibility (Wohlrab et al., 2018; Elferink et al., 2020). Ecotype forming within species complexes have been well documented, e.g., for prokaryotes as Synechococcus and Prochlorococcus (Sohm et al., 2015). Genome size and its inverse correlation to fitness has been documented, e.g., for the eukaryotic alga Dunaliella tertiolecta (Malerba et al., 2020). In addition, parasites have been documented to reduce their genomes by gene loss and elimination of any secondary DNA to achieve higher fitness in response to their hosts (as also reviewed by Thomas Cavalier-Smith, 2005).

In a global context, it has been shown that the highest phytoplankton diversity is often detected at intermediate biomasses, while especially high and low biomass correlate with lower diversity (Irigoien et al., 2004). In our case, we found that the low biomass winter community was surprisingly diverse (Figure 3B), and that the diversity, by means of ASVs and Shannon diversity index, decreased with the onset of the spring bloom. This suggests a highly diverse winter community followed by a spring bloom, in which fewer species and mostly diatom ASVs started to dominate the community in both relative and absolute abundance, as the conditions became favorable for them. Additionally, as described above, the overall less diverse microplankton size fraction show larger changes by means of community shifts to a changing environmental conditions over the sampling period than the smaller size fractions, supporting the hypothesis that larger celled species have a larger niche space, but cannot rapidly adapt to the seasonal changes due to their reaction norm limits. In temperate regions, food-web shifts are also discussed as a possible factor for plankton bloom initiations (Behrenfeld and Boss, 2014), which could be supported by the trophic shifts in the community of this study for this Arctic environment as well.

Studies in the Arctic have been investigating the phytoplankton spring bloom both in areas with sea ice (Terrado et al., 2008; Massicotte et al., 2020) and without sea ice (Kubiszyn et al., 2017). The ice cover has often been discussed as a factor involved in the initiation of the spring bloom because snow and ice cover will lower the penetration of light into the water column, depriving phototrophs of their energy source (Terrado et al., 2008; Leu et al., 2015). However, the transition from a sea ice covered surface water environment to surface waters without sea ice cover has rarely been studied. Here, we present data on the bloom dynamics starting in the dark winter period to the sea ice break-up and formation of a spring bloom. The slow increase in Chl a shows the initiation of the spring bloom event at a time when the sea ice was still largely covering the Bay (Figure 2A, Table 1). Biomass is, at this time, not yet strongly increasing, but when taking POC into consideration, the amount of phototrophs (measured as Chl a, Table 1) is starting to dominate the total amount of biomass, showing the imminence of the spring bloom.

Several studies have shown that phytoplankton growth is possible under very low light conditions, as often observed in surface waters under sea ice (Arrigo et al., 2012; Spall et al., 2014; Massicotte et al., 2020). It has also been shown that once light is available after the dark season, photosynthetic capabilities are quickly reactivated, usually within a few hours to a day (Kvernvik et al., 2018; Lacour et al., 2019). In the present study, the photon flux per day above water changed considerably (Figure 2B), but the light penetrating the ice was still extremely limited at the time of increasing photosynthetic activity (Figures 1A,C). The fluorescence measurement shows that photosynthetic cells were first gathering at a depth of approximately 55 m, which coincided with the approximate halocline at that time (Figures 1B,C), and which contrasts with previous findings of photosynthetic cells higher in the water column (Rat’kova and Wassmann, 2002). It is likely that the photosynthetic cells accumulated at this layer, e.g., due to increased density beyond this depth, marking the lower border of the ocean’s mixed layer. The decrease of this mixed layer depth over time (Figures 1B,D) could have possibly led to the cells being able to stay closer to the surface for a longer amount of time to be able to harvest sufficient photons for growth. Our study suggests that the pelagic spring bloom was not seeded from the sea ice or from the bottom of the sea ice as pennate diatoms typically dominate sea ice communities. Instead, we observed typical centric pelagic bloom species, similar to the findings of Arrigo et al. (2008, 2017). This suggests the seeding of the bloom from algal cells that were already present in the water column (as considered by Ardyna et al., 2020), but accumulated at the mixed layer depth. It is possible that the ongoing sea ice break-up in the vicinity could have additionally led to increased turbulences in the upper ocean layers. This could support non-motile cells such as diatoms to stay in the illuminated layers of the ocean, increasing the amount of possibly absorbed photons due to residence in lighter areas of the ocean, eventually enabling their growth. During the initiation of the spring bloom, the local area was still completely covered with sea ice. However, open patches further away from the sampling area may have been sufficient to increase the mixing in the suggested way and to lead to advective effects.

During winter, the low biomass but highly diverse protist community mostly consisted of parasites, heterotrophs, and mixotrophs, which is probably a natural adaptation to a life at low light availability (Clarke et al., 2019; Søgaard et al., 2021). The transitional period was characterized by a high relative abundance of mixotrophs, which most likely have a trophic advantage due to their trophic flexibility. The community shift toward a spring bloom community already started before the sea ice retreated. Past studies have forecasted and shown an increase in primary productivity after sea ice retreat, based on satellite data (Arrigo et al., 2008; Renaut et al., 2018). However, in situ studies, such as ours, confirm that blooms of microbial plankton not only occur (Arrigo et al., 2012; Spall et al., 2014; Pavlov et al., 2017; Massicotte et al., 2020; Søgaard et al., 2021), but also start growing while ice is still covering the surface waters. We also show that the period prior to the phytoplankton spring bloom is most likely not a period of dormancy, but that changes in the low biomass community are occurring (as also observed by, e.g., Marquardt et al., 2016). This suggests that sea ice retreat may not be the major factor in initiating the phytoplankton spring bloom in the Arctic. Rather, an interplay of light intensity, spectral composition and day-length, as well as oceanographic factors such as nutrient availability and halocline depth, i.e., mixed layer depth, are involved, making the spring bloom initiation and the shift from the winter community a multifactorial event (as also suggested by Ardyna et al., 2020).

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

CB: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation. NL: Writing – review & editing, Writing – original draft, Supervision. PH: Writing – review & editing, Writing – original draft, Methodology. SW: Writing – review & editing, Writing – original draft, Methodology, Formal analysis. UJ: Writing – review & editing, Writing – original draft, Supervision, Resources, Project administration, Funding acquisition.