As main primary producers, phytoplankton communities in pelagic oceans are dominated by picophytoplankton (< 2 ~ 3 µm) and nanophytoplankton (3 ~ 20 µm), which play a vital role in marine carbon sequestration and form the basis of marine food webs within euphotic zones (Buchan et al., 2014; de Vargas et al., 2015; Guidi et al., 2016). Meanwhile, more and more investigations show that phytoplankton are widely distributed throughout the water column, even in the challenger deep of Mariana Trench (Worden et al., 2015; Guo et al., 2018; Treguer et al., 2018). Strikingly, living phytoplankton cells have also been observed in the deep sea. For example, driven by internal solitary waves, abundant living Prochlorococcus cells were exported to up to 1,000 m depth at the Luzon Strait in the western Pacific Ocean (Jiao et al., 2014). Microscopy photos of various healthy diatom cells from the deep sea at depths of 2,000−4,000 m had ever been successfully obtained (Agusti et al., 2015). Indeed, the transfer of phytoplankton to deep sea (i.e., the sinking biological pump process) is affected by multiple factors that vary with different water current conditions or geographic locations (Schallenberg et al., 2018; Rocke et al., 2020; Dai et al., 2022). The distribution of phytoplankton in the whole water column is an important ecological issue that deserves in-depth study. On the one hand, it is related to the supply of organic nutrients to the microbial communities in the whole water columns (Amin et al., 2012; Turner, 2015; Durkin et al., 2016). 10%−50% of phytoplankton-derived organic carbon can be utilized by bacterioplankton and transformed to bacterial biomass for fueling microbial food web (Azam et al., 1983). On the other hand, since phyto- and bacterioplankton are inseparable in the ocean (Zhang et al., 2021a; Zhang et al., 2021b), it may have an important effect on the structure of microbial communities in the entire water columns. However, there is little knowledge about this aspect at present.

The photosynthetic nature of phytoplankton determines that they are mainly distributed in the euphotic zone of the ocean (Treguer et al., 2018; Karlusich et al., 2020), despite that living phytoplankton cells have been sporadically observed in the deep sea (Agusti et al., 2015). Compared with phytoplankton, the metabolic versatility of bacteria makes them ubiquitous in the ocean (Worden et al., 2015; Cuil et al., 2019; Alcolombri et al., 2021). As is known, phytoplankton in the epipelagic ocean can be transported to mesopelagic and deeper ocean via sinking biological pump as carbon sources for deep-sea microorganisms or long-term sequestered carbon (i.e., carbon sink) (Turner, 2015; Treguer et al., 2018). It was reported that the organic matter in the deep sea supporting bacterial productivity is mainly from phytoplankton debris (Luo et al., 2017). While in the euphotic zone, photosynthesis phytoplankton contributes to a significant proportion of organic matter by primary production, these organic matters through cascading trophic interactions are utilized by bacteria and support their growth (Thornton, 2014). It hints that no matter in the euphotic zone or the deep sea, there is a close correlation between phytoplankton and bacteria, and their correlation may vary in different depths of a water column and be influenced by the distribution of plankton communities.

Seamounts represent unique marine environments characterized by substantially enhanced currents and hydrodynamics changes (Becker et al., 2014; Ma et al., 2020). For example, eddies, isopycnal doming, and turbulent mixing occur frequently in the seamount area (Mendonca et al., 2012; Rocke et al., 2020; Sergi et al., 2020; Mohn et al., 2021). The specific topographic characteristics and complex hydrodynamics of the seamounts create distinct plankton habitats, by increasing vertical fluxes of nutrient-rich waters, enhancing primary productivity and supporting high biodiversity, especially around seamount of the oligotrophic oceans, known as “seamount effects” (Clark et al., 2010; Mendonca et al., 2012; Ma et al., 2021). In particular, the vertical flux of seawater (up-or-down-wellings) will affect the redistribution of plankton communities in the entire water columns of the seamount area (Clark et al., 2010; Levin and Sibuet, 2012; Santos et al., 2013; Xu, 2021).

Recently, a deep seamount (M5, 10.1°N, 140.2°E, the summit depth of 810 m) in the Western Pacific Ocean was discovered (Sun et al., 2020; Sun et al., 2021). We speculate that the vertical distribution of phytoplankton in this special area may have certain particularities. Meanwhile, there are usually species-specific relationships between phytoplankton and bacterial communities (Teeling et al., 2016; Kruger et al., 2019), e.g., the algae-associated bacterial communities of the diatom Nitzschia longissima were dominated by Neptuniibacter, Mesorhizobium, and Sphingorhabdus, while that of another diatom Pseudo-nitzschia were dominated by Alteromonas, Flavobacterium, and Roseobacter (Guannel et al., 2011; Behringer et al., 2018). Moreover, different Synechococcus strains in the same environment were surrounded by different microbial communities (Zhang et al., 2021b). Therefore, the vertically distributed phytoplankton may in turn have a certain regulatory effect on the bacterial community structure in the water columns. To this end, we investigated the vertical distribution of both phytoplankton (< 20 µm) and bacterioplankton communities in four water columns in this seamount area and studied their correlation networks at different depths. The results shed light on the vertical distribution and carbon sink of phytoplankton and their relationship with bacterioplankton in the deep-sea seamount.

Phytoplankton are the main primary producers in marine euphotic ecosystems (Durkin et al., 2016; Guidi et al., 2016; Treguer et al., 2018). Meanwhile, they are reported to be the most important source of organic matter in the deep sea (Luo et al., 2017). Previous sporadic studies have observed that there were diverse phytoplankton in the oceanic vertical water column (Guo et al., 2018; Richert et al., 2019), while under the influence of multiple factors, the vertical distribution of phytoplankton should have obvious geographic and temporal variations, which is poorly understood so far. Moreover, since phyto- and bacterioplankton are inseparable in the ocean, and the main organic matter required by bacterioplankton in the ocean is also mainly derived from phytoplankton (Zhang et al., 2021a; Zhang et al., 2021b), it is speculated that phytoplankton should have important impacts on the bacterioplankton community in the water columns from surface to deep oceans. In this study, we uncovered the vertical distribution of phytoplankton and bacterioplankton communities and revealed their potential associations in the water column of seamounts.

Although phytoplankton cells were only detected by flow cytometry in the water layer of < 200 m, the molecular investigation revealed that diverse phytoplankton existed throughout the water column of the seamount area ( Figure 3 ). Interestingly, there was no significant difference in the molecular community structure of phytoplankton among different water layers, which was reflected by their comparable β-diversity and similar dominant taxa in different layers ( Figure 2 ). The sinking biological carbon pump was reported to be the major force for the output of phytoplankton (including intact algal cells, algal aggregates, phytodetritus, etc.) from the euphotic zone to the deep sea and contributes significantly to oceanic carbon sinks (Agusti et al., 2015; Turner, 2015; Treguer et al., 2018; Giner et al., 2020). In the M5 seamount area, diatoms and dinoflagellates (Dinophyceae and Syndinales) had the highest relative abundance in the whole water column ( Figure 3 ), which is consistent with the previous observation in the Western Pacific Ocean and the Tara Ocean (Karlusich et al., 2020; Zhao et al., 2020). This may be due to the relatively larger cell size of diatoms and dinoflagellates than other phytoplankton in oligotrophic oceans, making them easier to sink into the deep sea (Crombet et al., 2011; Durkin et al., 2016). Furthermore, many diatoms and dinoflagellates also have mixotrophic metabolic characteristics, which can uptake organic matter and parasitize other plankton (Guillou et al., 2008; Sherr and Sherr, 2007; Marella et al., 2021), thus helping them to tolerate or survive in the dark environment. For example, a diatom (Chaetoceros sp. DS1) with the facultative heterotrophic function that was cultured from 1000m deep sea showed a very strong ability to tolerate darkness (Mou et al., 2022). Moreover, there are strong westward North Equatorial Current, reverse subcurrent, and possible upwelling in the M5 seamount area (Sun et al., 2020), the strong hydrodynamics in the seamount environment may strengthen the “biological pump” process and intensify the export of phytoplankton (even the very small picocyanobacteria cells) to the deep sea (Jiao et al., 2014), which led to the wide distribution of phytoplankton in the whole water column and the similarity of phytoplankton community structure in different water layers. The strong hydrodynamics around the deep-sea seamount regions are naturally similar to those created by the geoengineering of artificial upwelling or downwelling to increase carbon sinks (Bach and Boyd, 2021; NASEM, 2021), promoting vertical mixing of nutrients on the one hand to increase primary production, and improve the efficiency of biological carbon pump on the other hand (Casareto et al., 2017; Zhang et al., 2017; Ma et al., 2021).

However, there were also some phytoplankton groups (e.g., MAST clade belonging to Marine Stramenopile) that were only distributed in certain water layers under the impact of strong hydrodynamics. Currently, the underlying mechanism is still unclear. MAST clades have been reported to play multiple roles in microbial food webs and were driven by environmental factors, for example as symbionts with diatoms and cyanobacteria, and as heterotrophic grazers on bacteria and algae (Lin et al., 2012; Pernice et al., 2016). It was reported that the MAST clade possesses a variety of metabolic strategies adapted to different environments, making some of them to be relatively abundant in epipelagic water, while some are abundant in the deep sea (Logares et al., 2012; Massana et al., 2014; Giner et al., 2020; Obiol et al., 2020).

Unlike phytoplankton, from the epipelagic zone to deeper layers, an increase of bacterial diversity with depth have been observed in the Tara Ocean at a global scale (Sunagawa et al., 2015). The trend observed in this study is in accordance with the observation in the North Atlantic (Agogue et al., 2011) and the NW Mediterranean Sea (Mestre et al., 2017). The molecular community structure of bacterioplankton in different water layers showed great variation ( Figure 4 ), which is consistent with the phenomena observed in many previous deep-sea investigations (Li et al., 2018; Reji et al., 2020; Wang et al., 2020). Compared with phytoplankton, the cell size of bacterioplankton is much smaller, therefore their vertical output can be attached to particles and sank into the deep ocean (Mestre et al., 2018). Moreover, the bacterial community can respond more quickly to environmental changes, therefore the bacterial community is often used as an ideal biological indicator to reflect environmental disturbances (Zhang et al., 2014; Zorz et al., 2019). The huge environmental variation in different layers along the water column (Liang et al., 2017; McNichol et al., 2018; Wang et al., 2020) should be the main reason for the big differences in bacterial communities in different water layers.

From the bacterial classification, taxonomic classification showed that SAR11 clade (Alphaproteobacteria) was one of the most abundant heterotrophic microbial groups in the surface zone of this seamount ( Figure 4 ) (DeLong et al., 2006; Wang et al., 2020). However, different subclades of SAR11 showed a significantly different distribution pattern across the vertical water column of the seamount in this study. SAR11 clade Ia was overwhelmingly dominated (mean relative abundance of 24.85%) in the upper ocean (< 200 m), whereas SAR11 clade Ib had a wider distribution compared to SAR11 clade Ia and was observed in the whole water column, which mainly because these subclades of SAR11 had broader light intensity adaptabilities and nutrient metabolic capabilities (DeLong et al., 2006; Vergin et al., 2013; Traving et al., 2021). By comparison, Gammaproteobacteria, Bacteroidetes, and Firmicutes had higher relative abundance in the deep-sea compared to epipelagic waters ( Figure 4A ), similar to the investigation on deep sea of the Western Pacific (Wang et al., 2020), which may be due to the better metabolic adaptability of these taxa than others in the extreme environment of deep sea (Wu et al., 2013). For example, Alteromonas (affiliated to Gammaproteobacteria) was dominant in the mesopelagic zone of the seamount. It was reported that a member of Alteromonas (one strain of the “deep ecotype” isolated from 1,000 m deep sea) was more suitable for microaerophilic conditions, capable of degrading recalcitrant compounds and attached to relatively large particles that sink rapidly to mesopelagic depths (Ivars-Martinez et al., 2008). Bacillus (Firmicutes) and Bacteroides (Bacteroidetes) were significantly more abundant in the bathypelagic layer ( Figure 4B and Supplementary Figure S3 ). Among them, most Bacillus strains were able to form spores serving as an essential mechanism for survival in extreme environments (Cote et al., 2015; Lakhal et al., 2015). While certain Bacteroides strains could degrade difficult-to-breakdown polysaccharides and acquire carbohydrate substrates through a ‘selfish uptake’ mechanism that contributes to the efficient utilization of organic carbon due to avoiding the liberation of low molecular weight carbon (Cuskin et al., 2015; Lidbury et al., 2021; Zheng et al., 2021). This may help Bacteroides to survive in the bathypelagic layer where organic matter is barren and highly recalcitrant. Overall, the bacterial community in the seamount environment showed significant vertical heterogeneity, which was closely related to the metabolic characteristics of different bacterial groups.

Usually, phyto- and bacterioplankton have species-specific relationships due to the strong shaping effects of phytoplankton on the associated bacterial communities (Amin et al., 2012; Seymour et al., 2017). For example, when the same seawater bacterial community was co-cultivated with two different Synechococcus strains (i.e., PCC7002 and CCMP1334), the bacterial community was reshaped by the two Synechococcus strains into significantly distinct community structures (Zhang et al., 2021b). Here, a similar phenomenon occurred in Synechococcus sp. PCC7002 and P. tricornutum CCMP2561 (diatom). The same heterotrophic bacterial community in seawater was shaped by Synechococcus sp. PCC7002 or P. tricornutum CCMP2561 into significantly divergent community structures ( Figure 6 ). The difference in organic matter released by the two algae may be the reason for the difference in the bacterial community associated with the algae. Finally, at the genus level, the Synechococcus-associated bacterial communities were dominated by Algoriphagus, Dinoroseobacter, and Hyphobacterium, while the associated bacterial communities with the diatom P. tricornutum were dominated by Oceanicaulis and Cohaesibacter. ( Supplementary Figure S6 ).

Since phytoplankton can be distributed widely in the water columns and is an important organic source for bacterioplankton in both euphotic zone and deep sea (Buchan et al., 2014; Luo et al., 2017), we speculate that the close species-specific relationships between phyto- and bacterioplankton should occur in the entire water columns. Just as we expected, this phenomenon was observed in the vertical water columns of the M5 seamount area ( Figure 5 and Supplementary Figure S4 ). Notably, distinct species-specific algae-bacteria relationships occurred on different phyto- and bacterioplankton groups in different water layers ( Figure 5 and Supplementary Figure S4 ). While in the deep sea (i.e., mesopelagic and bathypelagic layers), there were more prominent algae-bacteria relationships than that in the euphotic layer. This is likely because phytoplankton-derived organic matter is the major carbon source supporting the survival of heterotrophic bacteria in the deep sea, while in the euphotic zone, the complex organic matters also come from a variety of other sources (such as protists, fungi, and zooplankton) (Netotea and Pongor, 2006; Durkin et al., 2016; Luo et al., 2017; Stock et al., 2019) and may mask the phyto- and bacterioplankton relationships, therefore, more prominent and diverse relationships were observed between phyto- and bacterioplankton in the deep sea. Furthermore, except for organic sources, the extreme physicochemical conditions in the deep sea, such as cold temperature and high pressure, can also significantly impact the microbial community structures (Hu et al., 2015; Zheng et al., 2020). It is worth mentioning that the dominant phytoplankton Bacillariophyta exhibited a species-specific relationship with Gammaproteobacteria throughout the entire water column ( Table 2 ), suggesting Bacillariophyta may have a strong enrichment effect on Gammaproteobacteria during their sinking process from the surface to the deep sea.

Here, it is difficult to define whether there is a positive or negative relationship between algae and bacteria in the vertical column, because when bacteria feed on organic matter from phytoplankton, there should be a positive correlation between phyto- and bacterioplankton, while bacterial growth comes at the expense of algal cell rupture and algal DNA loss, reflecting the negative correlation between phyto- and bacterioplankton (Samo et al., 2018; Mikhailov et al., 2019). In addition, the algae-bacteria interactions also include other aspects, such as competition due to heterotrophic growth of algae (Langwig et al., 2021), and bacterivory by some phytoplankton (Rii et al., 2016). The detailed relationship between phyto- and bacterioplankton in oceanic water columns need further in-depth study in the future. Even so, considering the species-specific relationships between phyto- and bacterioplankton that occurred throughout the water columns, we speculate that the various phytoplankton transported by biological pump to the deep sea may have an important role in shaping the structure of the bacterioplankton community in the entire water columns of the M5 seamount area.

To further verify the shaping effect of phytoplankton on bacterial communities, the quite different bacterial communities collected from different water layers were co-cultivated with Synechococcus PCC7002 and P. tricornutum CCMP2561, respectively. Interestingly, the originally distinct bacterial communities from different water layers all turned into similar bacterial communities under the impact of Synechococcus PCC7002 or P. tricornutum CCMP2561 ( Figure 6 ). However, as mentioned above, due to the species-specific relationship between algae and bacteria, the final bacterial community structures formed in the two algal cultures became significantly different ( Figure 6 and Supplementary Figure S6 ; PERMANOVA, p < 0.05). The observed phenomena in the laboratory further imply that during the export of various phytoplankton to the deep sea, the species-specific shaping effect of phytoplankton on bacterial communities (Monnich et al., 2020) may have an important impact on the community structures of bacterioplankton in the entire water columns. Here, due to the similarity in phytoplankton community structures at different water layers along the M5 seamount, we speculate that the vertically distributed phytoplankton communities may have the potential to weaken the variation of bacterial communities caused by the difference in physicochemical conditions of different water layers.

Seamount environments with strong hydrodynamics are naturally similar to those created by artificial upwelling or downwelling engineering to increase carbon sinks. Investigating such scenarios has certain implications for assessing the effects of these geoengineering strategies. The strong hydrodynamics in the seamount area may facilitate the export of diverse phytoplankton from the surface to the deep sea, i.e., increase carbon sinks through the biological pump mechanism, and lead to the similarity in the molecular community structure of phytoplankton in different water layers. The exported variety of phytoplankton may have important reshaping effects on the bacterioplankton communities in different water layers and weaken the divergence of bacterioplankton community structures caused by the variation of environmental conditions in different water layers. There is a more prominent and diverse species-specific relationship between phyto- and bacterioplankton in the deep sea as compared to the euphotic zone.

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 below: https://www.ncbi.nlm.nih.gov/, PRJNA779616.

YZ designed this study and wrote the manuscript. HZ and ZZ performed the experiments, analyzed the data, and wrote the manuscript. JZ assisted with the sample collecting. SN and SM assisted data analysis. YZ revised the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the Senior User Project of RV KEXUE (KEXUE2019GZ03) supported by the Center for Ocean Mega-Science, Chinese Academy of Sciences, the NSFC project (No. U1906216, 31700104, 41806172), the National Key R & D Program of China (2020YFA0607603), the DICP&QIBEBT (DICP&QIBEBT UN201803), the QIBEBT (QIBEBT ZZBS 201805), Dalian National Laboratory For Clean Energy (DNL), CAS.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.