Fig 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
Planktonic microorganisms in coastal waters form the foundation of food webs and biogeochemical cycles while exposed to pronounced environmental gradients, especially brackish salinities. Yet, commonplace ecological assessment overlooks most of their diversity. Here, we analyzed the protist and bacterial diversity from new and publicly available DN...
Contexts in source publication
Context 1
... gradients 97 regularly sampled throughout a year 39 . After filtering procedures, we used the data from 98 246 independent samples. They came from 18 locations, extending from Bothnian Bay 99 (minimal salinity = 2.00 PSU) to Skagerrak (maximal salinity = 33.92 PSU), sampled at 100 least six times between January 2019 and February 2020 (Fig. 1a). ...
Context 2
... Supplementary Fig. S1c-f). Additionally, in experimental conditions, bacterial 118 community assembly processes tend to converge at the family level but are more 119 stochastic at lower taxonomic levels, suggesting ecologically relevant functional 120 coherence 41 . ...
Context 3
... Proper (Fig. 1b), consistent with microscopic observations 44 . The increase in 138 cyanobacterial abundance was smaller and peaked later moving northwards within the 139 Baltic Sea. In the Baltic Proper, both dinoflagellates and Gyrista (the protist subdivision 140 including diatoms), peaked in abundance in March (Fig. 1c), as observed using 141 ...
Context 4
... Proper (Fig. 1b), consistent with microscopic observations 44 . The increase in 138 cyanobacterial abundance was smaller and peaked later moving northwards within the 139 Baltic Sea. In the Baltic Proper, both dinoflagellates and Gyrista (the protist subdivision 140 including diatoms), peaked in abundance in March (Fig. 1c), as observed using 141 microscopic methods 44,45 . The dynamics of Dinoflagellata and Gyrista differed vastly 142 between basins. However, in all cases except for Bothnian Bay, those groups increased 143 in abundance in the first half of the year. Furthermore, in Kattegat and Skagerrak, there 144 was an additional dinoflagellate peak ...
Context 5
... were abundant in the Baltic Sea, with higher abundance in the second half 147 of the year than in the first. 148 We also observed geographic differences in abundance and seasonal dynamics 149 of major non-phytoplankton groups. Bacteroidia (Fig. 1b) peaked in abundance in May 150 across the Baltic Sea, consistently with previous reports 35,46 , but displayed less clear 151 seasonal dynamics in Skagerrak and Kattegat. Actinomycetia and Acidimicrobiia were 152 abundant in the Baltic Sea but not in Kattegat/Skagerrak, consistent with Actinobacteria 153 abundance patterns from ...
Context 6
... We also observed geographic differences in abundance and seasonal dynamics 149 of major non-phytoplankton groups. Bacteroidia (Fig. 1b) peaked in abundance in May 150 across the Baltic Sea, consistently with previous reports 35,46 , but displayed less clear 151 seasonal dynamics in Skagerrak and Kattegat. Actinomycetia and Acidimicrobiia were 152 abundant in the Baltic Sea but not in Kattegat/Skagerrak, consistent with Actinobacteria 153 abundance patterns from transect-based studies [17][18][19] . ...
Context 7
... and the Baltic Sea. Compared to relative abundance-based 165 analysis, spike-in correction added ecologically relevant information, as best exemplified 166 by the March dinoflagellate bloom in Baltic Proper ( Fig. 1c) being indistinguishable 167 based on the relative abundance of Dinoflagellata ( Supplementary Fig. S2b). Still, the 168 spike-in approach entailed substantial random noise, especially in less sampled basins 169 ( Fig. 1b-c). Relative abundances better captured gradual succession patterns, 170 especially for bacteria ( Supplementary ...
Context 8
... added ecologically relevant information, as best exemplified 166 by the March dinoflagellate bloom in Baltic Proper ( Fig. 1c) being indistinguishable 167 based on the relative abundance of Dinoflagellata ( Supplementary Fig. S2b). Still, the 168 spike-in approach entailed substantial random noise, especially in less sampled basins 169 ( Fig. 1b-c). Relative abundances better captured gradual succession patterns, 170 especially for bacteria ( Supplementary Fig. S2a). These gradual changes likely reflect 171 sensitivity to factors other than the productivity of the system. Thus, the best 172 normalization approach depends on the goals of the analysis and suitable statistical 173 ...
Context 9
... for both bacterial and protist communities, there was a detectable 208 salinity divide between the Baltic Sea and Kattegat/Skagerrak, and a more gradual 209 latitudinal shift within the Baltic Sea (Fig 2c-d). The latter corresponded to decreasing 210 salinity and phosphate-to-nitrogen ratios with increasing latitude (Fig. 1d). We 211 connected those spatial changes in beta-diversity, calculated from ASV-based 212 dissimilarities, with high relative abundance of specific microbial families (Fig 2c-d). For 213 example, typically freshwater Nanopelagicaceae (acI Actinobacteria) 50 and ...
Context 10
... may best exemplify this 261 heterogeneity. Prymensiaceae were by far most abundant in summer in Baltic Proper 262 ( Fig. 3d), while Chrysochromulinaceae bloomed at a similar time in Bothnian Bay (Fig. 263 3e). These families also increased in abundance around June in Skagerrak, largely 264 explaining the increase in overall haptophyte counts (Fig. 1c). The same cannot be said 265 about Kattegat, where haptophytes were most abundant in May (Fig. 1c) . S7a, d). Still, it did not explain Kattegat's high abundance of haptophytes in May. 273 We ultimately found a May bloom of Phaeocystaceae in Kattegat ( Supplementary Fig. 274 S7e), which was short-lived, as is typical of Phaeocystis 57 ...
Context 11
... Baltic Proper 262 ( Fig. 3d), while Chrysochromulinaceae bloomed at a similar time in Bothnian Bay (Fig. 263 3e). These families also increased in abundance around June in Skagerrak, largely 264 explaining the increase in overall haptophyte counts (Fig. 1c). The same cannot be said 265 about Kattegat, where haptophytes were most abundant in May (Fig. 1c) . S7a, d). Still, it did not explain Kattegat's high abundance of haptophytes in May. 273 We ultimately found a May bloom of Phaeocystaceae in Kattegat ( Supplementary Fig. 274 S7e), which was short-lived, as is typical of Phaeocystis 57 . ...
Context 12
... family rose in abundance at the periods corresponding to a rapid decrease 297 in phosphate availability (Fig. 3f, i), as did total Bacteroidia abundance (Fig. 1c). 298 Recently, it has been suggested that Bacteroidia, and in particular Flavobacteria, gain a 299 competitive advantage under limited phosphate availability by efficiently mobilizing 300 organic phosphorus 61 62,63 . Our results further support this notion and suggest that 301 phosphorus availability contributes to brackish ...
Context 13
... distribution patterns of the exclusive co-occurrence clusters across the 485 salinity barrier were even more pronounced based on abundance (Fig 6b-d) than the 486 presence/absence of dbOTUs ( Supplementary Fig. S10), despite the co-occurrence 487 analysis being based on the latter data type. Clusters 1 and 3 showed strong 488 preferences for high and low salinities, respectively. ...
Context 14
... dbOTUs are less likely than protist dbOTUs to occur at both 528 high and low brackish salinities 529 The salinity gradient in the Baltic Sea area contains a characteristic, rapid shift across 530 the Danish straits. Consequently, our data contained geographically proximate samples 531 from lower (<9 PSU) and higher (>15 PSU) brackish salinities (Fig. 1a, Fig. 5a). Our 532 previous analyses showed the separation across this salinity barrier to be the strongest 533 structuring effect on microbial communities, though less so for protists than for bacteria 534 ( Fig. 2c-d, Supplementary Fig. S4). ...
Context 15
... further investigated whether bacteria are, in general, less likely to cross this 536 salinity barrier between lower and higher salinities. First, the proportion of protist 537 dbOTUs crossing the salinity barrier was significantly higher than the proportion of 538 bacterial dbOTUs (Fig. 7a, Supplementary Fig. S11). Based on comparable subsets of 539 samples, the proportion of both bacterial and protist dbOTUs crossing the salinity barrier 540 was higher in the samples from 2016-2017 than from 2019-2020 ( Supplementary Fig. 541 S11b-c). This suggests that the above noted differences in alpha diversity ( Fig. 4g-h) ...
Context 16
... the proportion of protist 537 dbOTUs crossing the salinity barrier was significantly higher than the proportion of 538 bacterial dbOTUs (Fig. 7a, Supplementary Fig. S11). Based on comparable subsets of 539 samples, the proportion of both bacterial and protist dbOTUs crossing the salinity barrier 540 was higher in the samples from 2016-2017 than from 2019-2020 ( Supplementary Fig. 541 S11b-c). This suggests that the above noted differences in alpha diversity ( Fig. 4g-h) 542 ...
Context 17
... proportion of barrier-crossing bacteria and protists may be affected by 549 detection biases coming from differences in relative abundance, length of vegetative 550 season, and environmental filtering by factors other than salinity. Bacteria were on 551 average more abundant in terms of rRNA gene marker copy numbers (Supplementary 552 Fig. S12a). As protists tend to have higher rRNA copy numbers per cell 128 , this 553 difference likely translates to even larger difference in cell numbers, which is likely the 554 most relevant abundance measure in the context of dispersal capabilities and detection 555 probability. Meanwhile protists were present in more samples at one side ...
Context 18
... protists tend to have higher rRNA copy numbers per cell 128 , this 553 difference likely translates to even larger difference in cell numbers, which is likely the 554 most relevant abundance measure in the context of dispersal capabilities and detection 555 probability. Meanwhile protists were present in more samples at one side of the salinity 556 barrier (Supplementary Fig. S12b). Overall, being similarly widespread on one side of 557 the salinity barrier, bacteria were more abundant ( Supplementary Fig. S12c-f). ...
Context 19
... protists were present in more samples at one side of the salinity 556 barrier (Supplementary Fig. S12b). Overall, being similarly widespread on one side of 557 the salinity barrier, bacteria were more abundant ( Supplementary Fig. S12c-f). Still, 558 ...
Context 21
... criterion 100 , we chose a version of the model assuming the probability 570 dependence on the other variables follows a different function for bacteria and for 571 protists (see Methods for details). The model predicted protists to be more likely to 572 cross the salinity barrier across the occupancy and abundance values, both on average 573 (Fig. 7e, Supplementary Fig. S12i-j) and in a majority of MCMC iterations (Fig. 7f). The 574 probability that protists are more likely to cross the salinity barrier was consistently close ...
Context 22
... comparability of results for protists and bacteria, we used only the samples 771 with data available from both 16S and 18S amplicon sequencing (excluding the under- were excluded from analyses in which spike-in normalization was used, except for 776 bacteria-only focused analyses in Fig. 1b and Fig. 3a-b, g-i. ...
Context 23
... PSU and among 831 samples with salinity <9 PSU was regarded as barrier crossing. For balancing out the 832 number of samples on both sides of the salinity barrier, apart from the four stations with 833 salinity >15 PSU (Å17, SLÄGGÖ), four out of twelve stations with salinity <9 PSU: BY2 834 ARKONA, BY5 BORNHOLMSDJ, BY15 GOTLANDSDJ, REF M1V1 (see Fig. 1a for 835 locations of the stations). First, we chose the four most saline stations from among 836 those with salinity <9 PSU. Then, we switched BCS III-10 to REF M1V1 (see Fig. 1), 837 since the latter had more samples, and, unlike other chosen stations from lower 838 salinities, was located close to the mainland coastline. For each station but ...
Context 24
... four stations with 833 salinity >15 PSU (Å17, SLÄGGÖ), four out of twelve stations with salinity <9 PSU: BY2 834 ARKONA, BY5 BORNHOLMSDJ, BY15 GOTLANDSDJ, REF M1V1 (see Fig. 1a for 835 locations of the stations). First, we chose the four most saline stations from among 836 those with salinity <9 PSU. Then, we switched BCS III-10 to REF M1V1 (see Fig. 1), 837 since the latter had more samples, and, unlike other chosen stations from lower 838 salinities, was located close to the mainland coastline. For each station but the least 839 sampled (REF M1V1), the samples were downsampled, choosing samples closest in 840 time to the ones collected at the least sampled stations. Whenever ...
Similar publications
While research has extensively investigated the dynamics of CO 2 water partial pressure (pCO 2 ) and planktonic food webs (PFWs) separately, there has been limited exploration of their potential interconnections, especially in marsh typologies. This study’s objectives were to (1) investigated if pCO 2 and atmospheric CO 2 flux can be elucidated by...
Marine ecosystems exist in a noisy and uncertain environment, not governed by deterministic laws. The development of ecological communities is significantly influenced by variability, and the interaction between nonlinearity and stochastic processes can lead to phenomena that deterministic models cannot explain. Plankton, forming the base of the ma...