A fungi hotspot deep down the ocean: explaining the presence ofGjaerumia minorin Equatorial Pacific bathypelagic waters

13 14 A plant parasite associated with the white haze disease in apples, the Basidiomycota 15 Gjaerumia minor, has been found in most samples of the global bathypelagic ocean. An 16 analysis of environmental 18S rDNA sequences on 12 vertical profiles of the Malaspina 17 2010 expedition shows that the relative abundance of this cultured species actually 18 increases with depth while its distribution is remarkably different between the deep 19 waters of the Pacific and Atlantic oceans, being present in higher concentrations in the 20 former. This is evident from sequence analysis and a microscopic survey with a species-21 specific newly designed TSA-FISH probe. Several hints point to the hypothesis that G. 22 minor is transported to the deep ocean attached to particles, and the absence of G. minor 23 in bathypelagic Atlantic waters could then be explained by the absence of this organism 24 in surface waters of the equatorial Atlantic. The good correlation of G. minor biomass 25 with recalcitrant carbon and free-living prokaryotic biomass in South Pacific waters, 26 together with the identification of the observed cells as yeast and not as a resting spore 27 (teliospore), point to the possibility that once arrived at deep layer this species keeps on 28 growing and thriving. 29 30 31 32 33 34 35 36 37 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 38

39 The bathypelagic ocean is one of the largest reservoirs of carbon on the planet, both 40 particulate and dissolved. The bioavailable part of this deep carbon pool, also known as 41 labile carbon, decreases along the conveyor belt, following the aging of the water 42 masses, being higher in the Atlantic Ocean compared to the Pacific Ocean1. Despite this 43 general decrease, fast-sinking episodes, such as the collapse of a bloom, could inject 44 fresh organic carbon into the deep sea2 . The remanent part, known as refractory carbon, 45 has been operationally defined as the organic carbon that cannot be degraded at in situ 46 conditions3 and increases towards Pacific waters. This increase of refractory carbon 47 along the conveyor belt is mainly explained by the fact that microbial communities act 48 on the total bathypelagic carbon pool reshaping its composition. Among these active 49 microbes, fungi have been proposed to play an important role as decomposers of 50 organic matter, both labile and refractory 4, with a specialized skillset for the second 51 pool. Despite their importance for the microbial carbon pump, the study of bathypelagic 52 fungal community and marine fungi in general has long been overlooked. 53 54 Pelagic fungi, which are consistently detected in seawater samples, belong mainly to 55 two taxonomic divisions, Ascomycota and Basidiomycota5. Species affiliated with these 56 groups, including G. minor, survive thanks to two main trophic strategies: saprotrophy 57 and parasitism. Saprotrophic fungi degrade and recycle organic matter with extracellular 58 enzymes6, in particular, it has been reported that the abundance of CAZymes 59 (Carbohydrate Active enZymes) increases toward mesopelagic waters 7 pointing to the 60 central role of fungal catabolism in aphotic environments. A similar pattern was also 61 detected for genes related to protein degradation 8. The utilization of extracellular 62 enzymes is a more effective digestion strategy if associated with a particulate-attached 63 lifestyle and, actually, in bathypelagic waters fungi are main colonizers of marine snow, 64 exceeding in some cases the biomass of bacteria 9. As parasites, pelagic fungi are 65 associated with different hosts, including animals, macroalgae, and microplankton 10–12. 66 Several environmental variables have been proposed to shape fungi abundance in photic 67 waters, including temperature, particulate organic matter 13, salinity, depth, oxygen and 68 nitrate14,15 whereas less information is available about the drivers of their distribution in 69 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint the bathypelagic realm. It has been suggested that they increase in diversity with depth, 70 mirroring what happens with prokaryotes5. 71 72 During the Malaspina 2010 expedition, Basidiomycota were found to be one of the most 73 important microbial eukaryotic groups in the Bathypelagic Ocean 16. Their relative 74 abundance, based on 454-pyrosequencing, peaked at equatorial Pacific waters, being the 75 most abundant 18S rDNA sequence virtually identical to Gjaerumia minor. A further 76 Illumina sequencing17 confirmed that one of the ASVs (Amplicon Sequence Variants) 77 retrieved (ASV-363) has 100% similarity with Gjaerumia minor (NG_063045). A G. 78 minor MAG retrieved from the deep ocean featured an ITS rRNA sequence identical to 79 a strain isolated from a pleural effusion of a child with pneumonia (KT149771). This 80 species, previously known as Tilletiopsis minor, was recently renamed since the entire 81 taxonomy of Exobasidiomycetes has been redefined based on their phylogenetic 82 relations18. Although the most frequently reported habitat of G. minor is plant material 83 (dead or alive), through propagules in the air, this yeast is capable of colonizing other 84 niches19 , including the deep ocean. This ability to adapt to different environments is 85 well exemplified by the fact that this fungus has been found to be a pathogen both for 86 plants and humans. G. minor is one of the causes of the “white haze” in apple trees 20, a 87 post-harvest disease that flourishes with humidity, cold (4ºC) and low oxygen levels 19 88 while in literature three cases are reported of G. minor as a human pathogen, finding 89 this yeast involved in subcutaneous mycosis 21, severe pneumonia 22 and corneal 90 abscess23. The unusual plasticity of G. minor, coupled with its uneven distribution in 91 the Bathypelagic Ocean, makes this interesting species worth of deeper analyses. 92 93 Here we reanalyze published sequencing datasets from the Malaspina expedition to 94 report the distribution of the Gjaerumia minor along the vertical profiles of several 95 oceanic stations. In addition, taking advantage of the fact that a strain of Gjaerumia 96 minor is commercially available, we were able to develop and test a new TSA-FISH 97 probe with the aim of confirming the presence of this fungus in marine pelagic samples, 98 describing its morphology, quantifying its abundance and investigating a possible role 99 in the labile and refractory carbon degradation process. Finally, we aim to highlight 100 which abiotic and biotic parameters may explain this organism's uneven distribution 101 with the final goal of better characterizing the trophic dynamics of the Bathypelagic 102 Ocean. 103 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 104

105 Information about the studied stations and sampling of Malaspina 2010 expedition 106 could be found in previously published works. In particular, the methodology for 107 bacterial abundance with flow-cytometry and prokaryotic biomass calculation is 108 described in Pernice et al. (2015) 24; DNA extraction, 454-pyrosequencing, and 109 bioinformatic pipeline for samples belonging to the deep ocean are in Pernice et al. 110 (2016)16; DNA extraction and Illumina sequencing for 12 vertical profiles (5 to 4000 m) 111 are in Giner et al. (2020) 17; Illumina sequencing from samples belonging to surface of 112 the entire cruise are in Logares et al. (2020) 25; and data for Illumina sequencing in 113 bathypelagic samples are in Junger et al. (2023)26. All the Illumina raw data (12 vertical 114 profiles, global surface and global bathypelagic) has been newly analyzed here with 115 DADA227 to define the distribution of the ASV of interest (363) in the different 116 datasets. Data for fluorescence Dissolved Organic Matter (FDOM) were obtained using 117 parallel factor analyses (PARAFAC) as described in Catalá et al. (2016) 28 and DOM 118 data and AOU values are from Catalá et al. (2015)29. 119 For TSA-FISH analyses, a seawater sample of 475 mL was fixed with 25 mL of 37% 120 formaldehyde (final concentration 1.85%, at least 1h at 4ºC) and filtered on board on a 121 0.6 µm pore size polycarbonate filter (25 mm diameter). The filters were stored at -80 122 ºC until processing. The TSA-FISH probe (Gmin01, CGACCACCATGTGCCCTT 5'-123 3') to count Gjaerumia minor cells at the microscope was designed based on the target 124 sequence ASV-363 as well as sequences belonging to G. minor from NCBI. The probe 125 was checked in silico against the SILVA 138 database and results to be species-specific 126 (the closer non-target sequence had 3 mismatches). The optimization of the 127 hybridization condition was done using a commercial strain of G. minor (Tilletiopsis 128 minor Nyland fungal strain, JCM No. 8361). A first attempt with a standard TSA-FISH 129 protocol, as described in Pernice et al. (2015) 24,30 adapted from Pernthaler et al. 130 (2001)30, gave poor results (less than 25% of positive cells in the culture). The addition 131 of helpers to contiguous regions of the probe ( HelperA-Gmin1 5’-132 GCGGGCTCGCGGCGATCAAT-3’; HelperB-Gmin1 5’-ACCAAGTTTGCCCAAGTTTT-3’ ) 133 increased the results to 65% of positive cells. The use of Wheat Germ Agglutinin 134 conjugated with Alexa Fluor 594 (Thermofisher; WGA 0.01mg mL -1, 30 min at RT), a 135 fluorescent marker staining chitin, confirmed the presence of a thick wall in cells 136 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint labeled with the probe so it was decided to include an additional permeabilization step. 137 After trying several protocols, the best results were obtained using an incubation step 138 for 1 h at 30ºC in a permeabilization buffer (pH 6.5) consisting of 1x PBS, 1% SDS, 1 139 mg mL −1 chitinase and 6 mg mL −1 Glucanex (a cocktail of enzymes isolated from 140 Trichoderma harzianum that contains β -glucanase, cellulase, protease and chitinase as 141 in Priest et al. 2021) 31. With this permeabilization step we reached more than 90% of 142 positive cells (Fig. 1). Finally, the optimal hybridization stringency was determined by 143 keeping the temperature constant at 35ºC and varying formamide concentrations (0% – 144 70%) in the hybridization buffer. The optimal value was 20% formamide, the highest 145 possible concentration before probe signal intensity decreased. We then applied the 146 developed TSA-FISH protocol to a subset of 34 environmental samples from the 147 deepest layer of the Malaspina 2010 expedition. 148 Target cells (two morphotypes, rounded and elongated) were counted by inspecting a 149 transect between 30 and 80 mm (average of 54 mm/sample equivalent to 540 fields). 150 Microscopic analyses were performed on an Olympus BX61 epifluorescence 151 microscope (Olympus America Inc.) at 1000× magnification under UV for DAPI, blue 152 light for A488 (TSA-FISH) and green light for WGA. Pictures were taken on an 153 Olympus DP72 camera connected to the microscope. Biovolume for the rounded 154 morphotype was calculated by assuming a spherical cell based on the average radio 155 (100 cells measured). The biovolume of the elongated morphotype was calculated 156 assuming a prolate spheroid shape (Hillebrand et al., 1999) 32 based on the following 157 formula: V=pi/6*d2*h, where h is the largest cell dimension and d is the largest cross-158 section of h, average h and d are based on 40 measured cell. We then used the equation 159 of Menden-Deuer and Lessard (2000) 33 to convert cell biovolume to cell biomass: pgC 160 cell-1 = 0.216*(Biovolume 0.939). Within each sample, average cell biomass times cell 161 abundance counted by TSA-FISH for each morphotype was calculated and then 162 summed to obtain the total biomass of the G. minor population. 163 Statistical analyses (Regression analyses and Pearson correlation) were performed with 164 Rstudio (package Hmisc). 165 166

167 Morphotypes 168 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint Two different morphotypes have been identified as Gjaerumia minor by the TSA-FISH 169 probe (Fig. 2), one rounded with an average diameter size of 1.5 µm (panels a and b), 170 and one elongated (panels c and d) with an average h (larger cell axis) of 5.2 µm and an 171 average d (smaller cell axis) of 1.5. The rounded morphotype, which represents the 172 majority of cells retrieved, shows a single nucleus per cell, evidenced in the figure by 173 DAPI stain. This cell type never appears in culture and was not stained by WGA, as it is 174 clearly visible when comparing to the elongated shape in Figs 2g, 2h and 2i. So, it was 175 assumed that the rounded morphotype has low chitin content. Although a rounded shape 176 is often associated in congenera species with teliospore (resting spore, ticker chitin wall, 177 dikarya), our observations (cell size, lack of thick chitin, and the presence of only one 178 nucleus) lead us to consider the rounded morphotype not a resting spore but an active 179 yeast cell. The elongated morphotypes, present only in Pacific waters, are probably 180 hyphae, sometimes found in chains (Figs. 2c, 2d) but more often as single units (Figs 181 2g, 2h and 2i) with a visible nucleus in the middle as observed also in cultured cells. 182 183 Vertical distribution of the target 18S rDNA sequence 184 In order to have a general view of the distribution of G. minor in the global ocean, we 185 used already published sequencing data of 18S rDNA genes from the 0.2-3 µm size 186 fraction (Illumina tags grouped in ASVs). The relative abundance of the G. minor ASV 187 (ASV-363) was available for 12 vertical profiles from the surface till 4,000 m and for 188 124 surface samples of the entire cruise (Fig. 3). G. minor ASV tends to increase with 189 depth, as shown by the median of its relative abundance (excluding zero values), which 190 is higher in the bathypelagic ocean (median of 0.02% of tags). Sequences belonging to 191 G. minor were not found in 29% of the samples of the Bathypelagic realm, while this 192 number was 53% for Mesopelagic and 67% for Epipelagic samples, further stressing its 193 higher presence in the deeper ocean. Illumina sequencing was also performed in 22 194 bathypelagic samples belonging to a different size fraction (0.8-20 µm). This set of 195 samples shows higher values with a median of 1.73% and a maximal value of 25%. The 196 vertical distribution clearly points to the bathypelagic ocean as a preferred environment 197 for this fungus and we decided to focus on the deepest sampling point for the TSA-198 FISH analysis. 199 200 Cell abundance in the Bathypelagic region and fit with sequencing data 201 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint TSA-FISH analysis targeting G. minor was performed on the deepest sample (between 202 2600 and 4000 m) of 34 stations. Figure 4a shows the cell abundance of the two 203 morphotypes retrieved along the cruise track crossing the Atlantic, Indian, and Pacific 204 Oceans. Cells were observed in all 34 bathypelagic samples inspected, with the rounded 205 morphotype being always more abundant than the elongated one. G. minor rounded 206 cells had low abundances (less than 10 cells mL -1) both in Atlantic and Indian oceans 207 whereas they were more abundant in Pacific waters, peaking at station 91 with 107 cells 208 mL-1. Elongated cells were also more abundant (3-6 cells mL -1) where rounded cells 209 were abundant (stations 88-91-92). 210 211 We then compared the cell counts with the relative abundance of the corresponding 212 phylotype in the same samples, analyzed in the 0.8-20 µm size fraction where G. minor 213 had a larger representation, this fraction includes 20 samples of the Illumina survey 214 reported before for the vertical profile (Fig. 4b) and 26 samples from a previous report16 215 in which the same DNA extracts and the same 18S DNA region were analyzed and 454-216 pyrosequenced (Fig. 4c). As expected, since the two sequencing analyses are based on 217 the same DNA extract, the relative abundance of the specific OTU (by pyrosequencing) 218 and the specific ASV (by Illumina) have a very strong direct relationship (n=19, 219 R=0.84, p<0.001). Both sequencing datasets found an almost absence of G. minor in 220 Atlantic Ocean, an intermediate presence in the Indian ocean and maximal values in the 221 Pacific Ocean. There was a statistically significant moderate direct relationship both for 222 the specific ASV (n=18, R=0.54, p=0.02) and the specific OTU (n=26, R=0.51, 223 p=0.008) with the counts of rounded cells. These moderate fits could be explained by 224 the typical errors of each approach and the huge difference in filtered volume which was 225 ~120 L for the DNA analysis and only 500 mL for the TSA-FISH. Sequencing data also 226 allowed us to put the abundance of G. minor (in green in Figs. 4b and c) in the context 227 of other fungi, like the rest of Basidiomycota (in blue) and Ascomycota (in light 228 yellow). In the Indian and Pacific Oceans, where G. minor is abundant, this single 229 species represents almost the totality (median of 87%) of Basidiomycota. Ascomycota 230 is generally less abundant than Basidiomycota along the cruise track, but follow a 231 similar global distribution, being functionally absent in the Atlantic Ocean and abundant 232 in Pacific waters. Ascomycota maximal abundance is shifted to station 103 respect to 233 the peak of the Basidiomycota in station 88. 234 235 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint Biomass and its relation with AOU and FDOM 236 The biomass of the two morphotypes combined (Total Biomass) ranged between 237 1.74·10-5 and 2.05·10-2 µg C L -1, being globally the rounded morphotype the dominant 238 contributor to the species biomass. The distribution of biomass along the entire cruise is 239 shown in Fig. 5 (bar plot), with the highest values found in equatorial Pacific waters. 240 The points on the graph correspond to AOU (Apparent Oxygen Utilization), which is a 241 proxy for the age of the water mass, higher AOU values imply older waters. 242 Considering the entire cruise track, G.minor biomass is higher in aged waters. A similar 243 picture is obtained observing the relationship between G. minor biomass and the 244 Fluorescent Dissolved Organic Matter (FDOM) (Fig. 6). The FDOM profile for each 245 sampling point was obtained through a PARAFAC analysis28 and it is composed by four 246 peaks, two belonging to Humic-like matter (C1 and C2) that are proxies for the 247 refractory carbon, and two associated with protein-like matter (C3 and C4) that 248 represent the labile carbon. Among the four FDOM components, C2 and C4 did not 249 show a significant relationship with G. minor biomass nor bacteria (both in total 250 Bathypelagic and considering only Pacific samples) and were not considered in this 251 analysis. The correlation between fungi biomass (green dots) and FDOM other two 252 peaks (C1-C3) as well as the relationship between bacteria biomass (blue dots) and the 253 same parameters it is shown in Fig. 6, samples belonging to Pacific waters are in bold. 254 G. minor in Pacific waters has a very strong inverse relationship with the Humic-like C1 255 peak (n=11, R=-0.83, p=0.001). So, despite the presence of G. minor being higher in 256 Pacific compared to Atlantic waters, considering only the Pacific region that is richer in 257 recalcitrant carbon (C1 peak), the abundance of this fungus is higher where the C1 258 concentration is lower suggesting a possible consumption of recalcitrant carbon by 259 G.minor. It is important to note that, despite the biomass of bacteria is much higher than 260 that of G. minor , the relationship between total bacteria biomass and C1 is not 261 significant (p=0.22) in Pacific waters. Considering also the Pacific samples, G. minor 262 biomass presents also a strong inverse relationship with the protein-like C3 peak (n=9, 263 R=-0.72, p=0.027), which in this case is resembled by the prokaryotic biomass (n=9, 264 R=-0.74, p=0.023). 265 266 Abiotic and Biotic parameters 267 Abiotic parameters (temperature, salinity, conductivity and oxygen) did not have a 268 significant correlation with G. minor biomass both for the global bathypelagic ocean 269 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint and considering only Pacific samples. Higher values of biomass are within a narrow 270 range of values both for salinity (around 34.7 psu) and oxygen (3.23-3.58 mL L -1), 271 which is expected considering that they all belong to the same water mass. The values 272 of G.minor biomass are extremely low compared with the total prokaryotic biomass 273 which range between 0.12 to 4.49 µg C L -1. Nevertheless, the biomass of G.minor and 274 prokaryotes are strongly correlated (n=26, R=0.8, p<0.001). In particular, in Pacific 275 waters, the Low Nucleic Acid content bacteria (LNA) better correlated with the G. 276 minor biomass than total bacterial biomass (R2=0.86 versus R2=0.59). 277 278

279 The presence of a clear hotspot for G. minor in the bathypelagic region of the Equatorial 280 Pacific has been shown in consensus by two different sequencing analyses and a 281 species-specific TSA-FISH probe. Since the ASV retrieved from the sequencing 282 analyses was 100% similar to a cultured strain of G. minor it was possible to test the 283 probe before its application to environmental samples. The TSA-FISH technique 284 allowed us to visualize the morphotype (life-stage) of the targeted fungi, showing that 285 the majority of the G. minor cells were rounded, with a single nucleus and unstained 286 with WGA (Fig. 2). Larger elongated cells were also observed but at much lower 287 abundance. 288 289 G. minor belongs to the class ustilaginomycetes, which are usually dimorphic, 290 producing a saprobic haploid yeast phase and a parasitic dikaryotic phase 34, so 291 particular attention was given to the identification of the proper life-stage of the rounded 292 cells. If they corresponded to a resting stage (teliospore) this would mean that G. minor 293 is not thriving in the bathypelagic environment, whereas a yeast phase opens the 294 possibility that G. minor is alive and active in those waters. Three hints suggest us to 295 exclude the teliospore hypothesis: i) their size is too small; although teliospores from 296 the sister species Gjaerumia ossifragi (Bauer 2005, figs 5-14) 35 show a rounded shape, 297 they are ten times larger than the ones retrieved here, and in proportion G. minor 298 rounded cells seem too little for the size of the hyphae to be a teliospore, ii) they are not 299 dikaryotic, as a teliospore is expected to be; we did not observe two nuclei in DAPI 300 stained samples, iii) last and most relevant, rounded cells are not stained by WGA, a 301 specific stain for chitin; teliospores, on the other hand, are expected to have a very tick 302 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint chitin wall (Figs. 2 g-i). The lack of WGA staining is probably due to the fact that yeast 303 merging and budding are favored by a thinner chitin wall. There is still the possibility 304 that the rounded morphotype is a reproductive spore, such as basidiospore (sexual) or 305 basidioconidia (asexual), but although the pressure of the bathypelagic environment 306 could force a more globose shape, both of them are expected to be fusiform 35,36, 307 therefore, we also tend to exclude this hypothesis. 308 309 G. minor is known to have a strong plasticity being retrieved both as plant parasites 19,20 310 and human pathogens 21–23. In the frame of Malaspina 2010 cruise, it is often present 311 also in surface waters (51 of 136 stations), although in a lower percentage. It is clear 312 that G. minor is not an endemic species of the Bathypelagic environment but is 313 transported there from land through the surface global ocean. For prokaryotic 314 communities, based on samples from the same cruise, it has been reported a direct 315 connectivity through fast-sinking particles between surface and the bathypelagic 316 layer37,38. We hypothesized a similar mechanism of transport for G. minor, as previously 317 suggested by Bochadansky and colleagues in 2017 9. This idea is also reinforced by the 318 fact that the osmotrophic food acquisition in fungi, based on the secretion of 319 extracellular enzymes, better suits a particle-attached life-style 6. These particles could 320 be represented by sinking G. minor hosts (phytoplankton) or by fragments of them 321 (animals or macroalgae). Species taxonomically close to G. minor have been found 322 associated with sea-animals 10,11, macroalgae12 and even dinoflagellates 39,40. Related to 323 this, data of the same cruise 41 reported highest concentrations of large phytoplankton in 324 the deep sea, represented by living fast-sinking cells (81.5% of diatoms followed by 325 dinoflagellates) in the Equatorial Pacific, opening the hypothesis of undiscovered 326 microbial interactions. Interestingly, the sequencing analyses shows an absence of G. 327 minor in a large surface area of the equatorial Atlantic Ocean, from 14º N to 24º S 328 (stations 17 to 26) so, the lower presence of this species in Atlantic bathypelagic waters 329 could be explained by its absence at the surface, as the snowfall of particles at the 330 equator is similar between the two oceans42. 331 332 Despite several hints pointing to the fact that fungi arrive to the deep ocean attached to 333 particles, we have only been able to observe this phenomenon once (Fig. 2e, f, which 334 shows a large particle colonized by rounded cells), while most observed rounded cells 335 appear to be free-living. There are possible methodological reasons to explain the lack 336 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint of particles observed in our TSA-FISH filters (low volume, high filtering pressure); 337 nevertheless, even in an environment full of particles, yeasts are expected to look for 338 new resources once the ones of their transport-particle are exhausted. We propose that, 339 although the manipulation of the samples could separate some cells from particles, part 340 of the community live temporarily in a free-living state as evidenced by past results of 341 flow cytometry16, sequencing of the 0.2-0.8 µm size fraction (often used as the free-342 living size fraction for bacteria, supplementary table 1) and the TSA-FISH itself. 343 344 The biomass of G. minor yeast cells correlates very well with the biomass of free-living 345 bacteria, although fungal biomass was 3 orders of magnitude lower than the bacterial 346 one. This contrasts Bochadansky et al. (2017) 9, who found similar values for the two 347 groups. The prokaryotic free-living pool represents the endemic part of the community 348 and, in contrast with the particle attached assemblage, does not correlate with surface 349 biotic variables but with bathypelagic environmental conditions 38. This leads us to 350 propose that the good correlation between the biomass of fungi and free-living bacteria 351 is due either to a parallel response to the bathypelagic conditions and resources or to the 352 utilization by fungi of carbon processed by bacteria. In fact, both cases point to an 353 active free-living fungal community. In culture conditions, G. minor does not grow 354 without some vitamins36 ,and in particular, it needs thiamine to thrive. It is possible that 355 the prokaryotic community could be the source of these vitamins and a reliable scenario 356 is that vitamin availability coupled with recalcitrant carbon (C1 peak) created optimal 357 conditions for G. minor, which could grow and reproduce in marine deep waters. 358 359 The idea that once passively transported to the Bathypelagic layer of the Pacific, G. 360 minor is capable of establishing an actively thriving population is supported by several 361 hints. First, we know from experimental evidence in Saccharomyces cerevisiae that 362 fungi can alter their membrane composition to tolerate high hydrostatic pressures 43, 363 which could make possible a rapid colonization of deep-sea habitats by surface strains44. 364 For G. minor this colonization could be supported by a great availability of resources, 365 the pool of recalcitrant carbon. Second, considering the entire dataset, not only 366 Basidiomycota but also Ascomycota are more abundant in the aged waters. Moreover, 367 Ascomycota peak is displaced spatially forward compared to the Basidiomycota peak, 368 pointing to a different target substrate and reinforcing the hypothesis of an active fungal 369 deep community. Third and more importantly, our analysis showed a strong and 370 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint significant correlation between G. minor biomass and recalcitrant carbon (C1 peak of 371 FDOM) in Pacific waters suggesting a possible consumption of the recalcitrant pool by 372 G. minor (Fig. 6); the same relation is not significant for bacterial biomass. This 373 possibility is in accordance with Clipson et al. (2006) 45 who state that fungi are often 374 better than bacteria at breaking down recalcitrant organic material. Although the 375 biomass of G. minor is much lower than the biomass of bacteria, it is still possible that 376 they both contribute to the degradation in a mutualistic rather than antagonistic way9,46. 377 378 G. minor also relates significantly to labile carbon (C3 peak of FDOM). In fact, the 379 slope of the correlation line between the labile C3 peak and G. minor is steeper than in 380 the C1 relationship in Pacific waters, which could point to faster utilization of labile 381 than recalcitrant carbon by fungi. It has been proposed that labile matter could act as a 382 primer for the digestion of more refractory components in a dynamic known as 383 priming7,47–49 . Although this topic is still controversial for the aquatic environment 50, 384 our results could suggest a parallel consumption of Humic-like and Labile-like matter 385 operated by G. minor. 386 387 The presence of the cultured plant-parasite G. minor has been highlighted in 388 bathypelagic waters, especially in the Equatorial Pacific where its biomass peaks, both 389 by species-specific TSA-FISH probe and by high-throughput sequencing. Our data 390 suggests that in these waters G. minor feeds osmotrophically on the recalcitrant carbon 391 pool, a resource with which it correlates better than free-living bacteria. Our results 392 show, in accordance to previous works 38,51, that the Bathypelagic Ocean is not an 393 isolated environment but it is in continue connection with surface waters and with the 394 land. In this regard the fungus G. minor shows an enormous potential to be an important 395 active part of the global carbon cycle, pointing at the same time to the importance and 396 necessity of more studies both on the Bathypelagic Ocean and on its fungal diversity. 397 398

399 This project was supported by the Spanish Ministry of Economy and Competitiveness 400 through projects Consolider-Ingenio Malaspina-2010 (CSD2008–00077), ALLFLAGS 401 (CTM2016-75083-R, MINECO) and MINIME (PID2019-105775RB-I00, AEI, Spain). 402 We thank our fellow scientists, the crew of the R/V BIO-Hesperides and chief scientists 403 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint of the different cruise legs for collaboration, a special thank goes to J.M. Gasol and 404 X.A.G. Moran for sharing the bacterial abundance data. 405 406 Competing interests 407 The authors declare no competing interests. 408 409 Data Availability 410 The sequencing dataset analysed in this present work are all belonging to Malaspina 411 expedition and they are already published and publicly available at the European 412 Nucleotide Archive (ENA). Amplicon sequences from 454-pyrosequencing of the deep 413 ocean are available at www.ebi.ac.uk/ena/data/view/PRJEB9943; Illumina sequences 414 from vertical profiles are available at www.ebi.ac.uk/ena/data/view/ PRJEB23771; 415 Illumina sequences from surface samples are available at 416 www.ebi.ac.uk/ena/PRJEB23913 and Illumina sequenc es from the deep ocean are 417 available at www.ebi.ac.uk/ena/data/view/PRJEB45014. 418 419

420 1. Hansell, D. A., Carlson, C. A., Repeta, D. J. & Schlitzer, R. Dissolved organic 421 matter in the ocean, a controversy stimulates new insights. Oceanography 22, 422 202–211 (2009). 423 2. Gómez-Letona, M. et al. Deep ocean prokaryotes and fluorescent dissolved 424 organic matter reflect the history of the water masses across the Atlantic Ocean. 425 Prog Oceanogr 205, (2022). 426 3. Baltar, F. et al. What Is Refractory Organic Matter in the Ocean? Front Mar Sci 427 8, 1–7 (2021). 428 4. Li, W., Wang, M., Burgaud, G., Yu, H. & Cai, L. Fungal community 429 composition and potential depth-related driving factors impacting distribution 430 pattern and trophic modes from epi- to abyssopelagic zones of the Western 431 Pacific Ocean. Microb Ecol 78, 820–831 (2019). 432 5. Morales, S. E., Biswas, A., Herndl, G. J. & Baltar, F. Global structuring of 433 phylogenetic and functional diversity of pelagic fungi by depth and temperature. 434 Front Mar Sci 6, (2019). 435 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 6. Richards, T. A. et al. Molecular diversity and distribution of marine fungi across 436 130 european environmental samples. Proceedings of the Royal Society B: 437 Biological Sciences 282, (2015). 438 7. Baltar, F., Zhao, Z. & Herndl, G. J. Potential and expression of carbohydrate 439 utilization by marine fungi in the global ocean. Microbiome 9, (2021). 440 8. Breyer, E. & Baltar, F. The largely neglected ecological role of oceanic pelagic 441 fungi. Trends in Ecology and Evolution Preprint at 442 https://doi.org/10.1016/j.tree.2023.05.002 (2023). 443 9. Bochdansky, A. B., Clouse, M. A. & Herndl, G. J. Eukaryotic microbes, 444 principally fungi and labyrinthulomycetes, dominate biomass on bathypelagic 445 marine snow. ISME Journal 11, 362–373 (2017). 446 10. Zhang, X. Y. et al. Diversity and Antimicrobial Activity of Culturable Fungi 447 Isolated from Six Species of the South China Sea Gorgonians. Microb Ecol 64, 448 617–627 (2012). 449 11. Yue, Y. et al. Exploring the Antibacterial and Antifungal Potential of Jellyfish-450 Associated Marine Fungi by Cultivation-Dependent Approaches. PLoS One 10, 451 (2015). 452 12. Duarte, A. W. F. et al. Yeasts from macroalgae and lichens that inhabit the South 453 Shetland Islands, Antarctica. Environ Microbiol Rep 8, 874–885 (2016). 454 13. Taylor, J. D. & Cunliffe, M. Multi-year assessment of coastal planktonic fungi 455 reveals environmental drivers of diversity and abundance. ISME Journal 10, 456 2118–2128 (2016). 457 14. Jeffries, T. C. et al. Partitioning of fungal assemblages across different marine 458 habitats. Environ Microbiol Rep 8, 235–238 (2016). 459 15. Tisthammer, K. H., Cobian, G. M. & Amend, A. S. Global biogeography of 460 marine fungi is shaped by the environment. Fungal Ecol 19, 39–46 (2016). 461 16. Pernice, M. C. et al. Large variability of bathypelagic microbial eukaryotic 462 communities across the world’s oceans. ISME Journal 10, 945–958 (2016). 463 17. Giner, C. R. et al. Marked changes in diversity and relative activity of 464 picoeukaryotes with depth in the world ocean. ISME Journal 14, 437–449 465 (2020). 466 18. Wang, Q. M. et al. Multigene phylogeny and taxonomic revision of yeasts and 467 related fungi in the Ustilaginomycotina. Stud Mycol 81, 55–83 (2015). 468 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 19. Richter, C., Yurkov, A. M., Boekhout, T. & Stadler, M. Diversity of Tilletiopsis-469 like fungi in Exobasidiomycetes (Ustilaginomycotina) and description of six 470 novel species. Front Microbiol 10, (2019). 471 20. Boekhout, T. et al. Extensive colonization of apples by smut anamorphs causes a 472 new postharvest disorder. FEMS Yeast Res 6, 63–76 (2006). 473 21. Ramani, R., Kahn, B. T. & Chaturvedi, V. Tilletiopsis minor: a new etiologic 474 agent of human subcutaneous mycosis in an immunocompromised host. J Clin 475 Microbiol 35, 2992–2995 (1997). 476 22. Al-Zaydani, I. A., P. Joseph, M. R., Suheel, A. M., Al-Hakami, A. M. & E. 477 Hamid, M. Severe pneumonia with a massive pleural effusion in a child caused 478 by Tilletiopsis minor: the first case from Saudi Arabia. Ann Saudi Med 35, 475–479 478 (2015). 480 23. Kechkekian, A., Gari-Toussaint, M., Gastaud, P., Freton, A. & Gantier, J. C. 481 Kératite à Tilletiopsis minor post LASIK. J Mycol Med 17, 119–121 (2007). 482 24. Pernice, M. C. et al. Global abundance of planktonic heterotrophic protists in the 483 deep ocean. ISME Journal 9, 782–792 (2015). 484 25. Logares, R. et al. Disentangling the mechanisms shaping the surface ocean 485 microbiota. Microbiome 8, (2020). 486 26. Junger, P. C. et al. Global biogeography of the smallest plankton across ocean 487 depths. Science advances in press, (2023). 488 27. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina 489 amplicon data. Nat Methods 13, 581–583 (2016). 490 28. Catalá, T. S. et al. Drivers of fluorescent dissolved organic matter in the global 491 epipelagic ocean. Limnol Oceanogr 61, 1101–1119 (2016). 492 29. Catalá, T. S. et al. Water mass age and aging driving chromophoric dissolved 493 organic matter in the dark global ocean. Global Biogeochem Cycles 29, 917–934 494 (2015). 495 30. Pernthaler, A., Pernthaler, J. & Amann, R. Fluorescence in situ hybridization and 496 catalyzed reporter deposition for the identification of marine bacteria. Appl 497 Environ Microbiol 68, 3094–3101 (2002). 498 31. Priest, T., Fuchs, B., Amann, R. & Reich, M. Diversity and biomass dynamics of 499 unicellular marine fungi during a spring phytoplankton bloom. Environ 500 Microbiol 23, 448–463 (2021). 501 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 32. Hillebrand, H., Dürselen, C.-D., Kirschtel, D., Pollingher, U. & Zohary, T. 502 Biovolume calculation for pelagic and benthic microalgae. J Phycol 35, 403–424 503 (1999). 504 33. Menden-Deuer, S. & Lessard, E. J. Carbon to volume relationships for 505 dinoflagellates, diatoms, and other protist plankton. Limnol Oceanogr 45, 569–506 579 (2000). 507 34. Begerow, D., Stoll, M. & Bauer, R. A phylogenetic hypothesis of 508 Ustilaginomycotina based on multiple gene analyses and morphological data. 509 Mycologia 98, 906–916 (2006). 510 35. Bauer, R., Lutz, M. & Oberwinkler, F. Gjaerumia, a new genus in the 511 Georgefischeriales (Ustilaginomycetes). Mycol Res 109, 1250–1258 (2005). 512 36. Boekhout, T. & Washingtonensis Nyland, T. Tilletiopsis Derx ex Derx (1930) 513 TYPE SPECIES. in (2011). 514 37. Mestre, M. et al. Sinking particles promote vertical connectivity in the ocean 515 microbiome. Proc Natl Acad Sci U S A 115, E6799–E6807 (2018). 516 38. Ruiz-González, C. et al. Major imprint of surface plankton on deep ocean 517 prokaryotic structure and activity. Mol Ecol 29, 1820–1838 (2020). 518 39. Deng, Y., Wang, K., Hu, Z., Hu, Q. & Tang, Y. Different Geographic Strains of 519 Dinoflagellate Karlodinium veneficum Host Highly Diverse Fungal Community 520 and Potentially Serve as Possible Niche for Colonization of Fungal Endophytes. 521 Int J Mol Sci 24, (2023). 522 40. Sun, J. Y. et al. Fungal community dynamics during a marine dinoflagellate 523 (Noctiluca scintillans) bloom. Mar Environ Res 131, 183–194 (2017). 524 41. Agusti, S. et al. Ubiquitous healthy diatoms in the deep sea confirm deep carbon 525 injection by the biological pump. Nat Commun 6, (2015). 526 42. Kiko, R. et al. Biological and physical influences on marine snowfall at the 527 equator. Nat Geosci 10, 852–858 (2017). 528 43. Simonato, F. et al. Piezophilic adaptation: a genomic point of view. Journal of 529 Biotechnology vol. 126 11–25 Preprint at 530 https://doi.org/10.1016/j.jbiotec.2006.03.038 (2006). 531 44. Bass, D. et al. Yeast forms dominate fungal diversity in the deep oceans. 532 Proceedings of the Royal Society B: Biological Sciences 274, 3069–3077 (2007). 533 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 45. Clipson, N., Otte, M. & Landy, E. Biogeochemical roles of fungi in marine and 534 estuarine habitats. in Fungi in Biogeochemical Cycles (ed. Pernice, M.) 436–461 535 (Cambridge University Press, 2006). doi:10.1017/CBO9780511550522.019. 536 46. Frey-Klett, P. et al. Bacterial-Fungal Interactions: Hyphens between Agricultural, 537 Clinical, Environmental, and Food Microbiologists. Microbiology and Molecular 538 Biology Reviews 75, 583–609 (2011). 539 47. Bianchi, T. S. The role of terrestrially derived organic carbon in the coastal 540 ocean: A changing paradigm and the priming effect. Proceedings of the National 541 Academy of Sciences of the United States of America vol. 108 19473–19481 542 Preprint at https://doi.org/10.1073/pnas.1017982108 (2011). 543 48. Bengtsson, M. M., Attermeyer, K. & Catalán, N. Interactive effects on organic 544 matter processing from soils to the ocean: are priming effects relevant in aquatic 545 ecosystems? Hydrobiologia vol. 822 Preprint at https://doi.org/10.1007/s10750-546 018-3672-2 (2018). 547 49. Shen, Y. & Benner, R. Mixing it up in the ocean carbon cycle and the removal of 548 refractory dissolved organic carbon. Sci Rep 8, (2018). 549 50. Catalán, N., Kellerman, A. M., Peter, H., Carmona, F. & Tranvik, L. J. Absence 550 of a priming effect on dissolved organic carbon degradation in lake water. Limnol 551 Oceanogr 60, 159–168 (2015). 552 51. Mestre, M. & Höfer, J. The Microbial Conveyor Belt: Connecting the Globe 553 through Dispersion and Dormancy. Trends in Microbiology vol. 29 482–492 554 Preprint at https://doi.org/10.1016/j.tim.2020.10.007 (2021). 555 556 Authors contribution 557 MCP, RM, and RL ideate the paper, IF developed the protocol and counted the samples 558 at the microscope, MCP developed the figures and wrote the first draft. MCP, RM, RL 559 and IF edited the final version of the paper. 560 561 Figure Legends 562 Figure 1: Epifluorescence pictures of cultured cells of Gjaerumia minor with triple 563 staining protocols. Panels above show a single elongated cell, possible an hypha, stained 564 with a) DAPI, b) TSA-FISH (Gmin01 probe conjugated with Alexa-488), and c) WGA. 565 The panels below show an entire hyphal structure stained with d) DAPI, e) TSA-FISH, 566 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint and f) WGA. Nuclei are visible in blue in a and d, the ribosome-containing cytoplasm in 567 green in b and e, and the outer chitin cover in red in c and f. Scale bar is 5 µm. 568 569 Figure 2: Examples of cells from environmental samples. A rounded morphotype (1.06 570 µm on average, probably a yeast) from ST83 stained with DAPI (a) and the Gmin01 571 probe (b). A hyphal structure (h= 6.9 µm, d=1.16 on average) from ST88 stained with 572 DAPI (c) and the Gmin01 probe (d). A group of rounded morphotypes embedded in a 573 gel particle from ST97 stained with DAPI (e) and the Gmin01 probe (f). A field 574 including the two morphotypes from ST92, an hyphae in the low-left corner and a 575 rounded cell in the up-right corner stained with DAPI (g), the Gmin01 probe (h) and 576 WGA (i). Note that the chitin staining applies to the hypha but not to the rounded cell. 577 Scale bar is 5 µm. 578 579 580 Figure 3: Vertical distribution of Gjaerumia minor based on the relative abundance of 581 ASV-363 that is 100% similar to cultured G. minor (strain AB7-11). Sequences derive 582 from previously published Illumina datasets 17, 25 . Samples from 0.2-3 µm size-fraction 583 were grouped in three layers: Epipelagic (3-200 m), including 149 samples, 584 Mesopelagic (200-1000 m), including 32 samples, and Bathypelagic (1000-4000), 585 including 31 samples (zero values are not shown in the graph). The last boxplot (in 586 blue) shows the relative abundance in 22 Bathypelagic samples belonging to the 0.8-20 587 µm fraction (no zero values found in this dataset). 588 589 Figure 4: Distribution of Gjaerumia minor in bathypelagic waters (deepest sample in 590 each station) along the entire track of Malaspina cruise. Oceanic boundaries are 591 depicted as a red dotted line for the three graphs. a) Cell abundance of rounded cells 592 (probably yeasts, violet) and elongated cells (hyphae, yellow) stained with TSA-FISH in 593 the deepest point of 34 stations spread across Atlantic, Indian and Pacific Oceans; b) In 594 green it is shown the relative abundance of G. minor illumina tags (ASV_363), 595 abundance of tags belonging to other Basidiomycota is shown in blue whereas tags 596 belonging to Ascomycota are shown in light yellow; c) relative abundance of 454-597 pyrosequenced tags (in 26 of the 34 previous stations) published in Pernice et al. 598 (2016)16 (size fraction 0.8-20 µm), tags belonging to OTU-6359 (99.7 similar to G. 599 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint minor) are shown in green, other Basidiomycota in blue and Ascomycota in yellow. 600 Illumina and 454 sequencing were done from the same DNA extraction, asterisks show 601 stations where no sequencing data was available. 602 603 Figure 5: Distribution of Gjaerumia minor biomass and Apparent Oxygen Utilization 604 (AOU, µmol O 2 Kg-1) across the Malaspina track. The biomass (expressed as µg C l -1) 605 is visualized as a barplot (left y axis) whereas the AOU, corrected by the real value of 606 measured oxygen, is represented by the blue dot (right y axis). AOU is a measure of the 607 amount of oxygen respired in the deep ocean, higher values of AOU are typical of aged 608 waters. 609 610 Figure 6: Microbial biomass relationship with C1 peak of FDOM (Humic-like, on the 611 left) and C3 peak of FDOM (Labile-like, on the right). G. minor biomass is in green 612 whereas bacterial biomass is in blue; dots belonging to Pacific samples are darker than 613 the rest. Red lines indicate the linear relationships between biomass and FDOM values 614 using only the darker dots. 615 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint Bathypelagic 0.8-20 µm Bathypelagic 0.2-3 µm Mesopelagic 0.2-3 µm Epipelagic 0.2-3 µm 10-3 10−2 10−1 100 101 Relative abundance of itags (%) .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 0 20 40 60 st17st20st23st26st30st32st35st41st43st44st49st50st53st59st62st63st65st67st74st77st81st82st83st88st91st92st97st103st109st110st112st118st121st131 abundance % pyrotags 0 30 60 90 abundance cells ml-1 Atlantic PacificIndian abundance % itags a. TSA-FISH c. 454 pyroseq-OTUs 0 20 40 60 b. Illumina seq-ASV * * * * * * * * * * * * * * * * * * * * * * .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 0.000 0.005 0.010 0.015 0.020 0 50 100 150 200 st17st20st23st26st30st32st35st41st43st44st49st50st53st59st62st63st65st67st74st77st81st82st83st88st91st92st97st103st109st110st112st118st121st131 AOU µmol O2 Kg-1 G. minor biomass µg C L-1 .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint 10-2 10-3 10-4 G. minor biomass (µg C l-1) 100.5 10-1 10-0.5 100 Bacteria biomass (µg C l-1) 10-1.7510-1.8010-1.8510-1.90 C1(RU) Humic-like 10-2.0010-2.2510-2.50 C3 (RU) Labile-like .CC-BY-NC-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted January 27, 2024. ; https://doi.org/10.1101/2024.01.25.577184doi: bioRxiv preprint