Cruise Report for AE2520 on R/V Atlantic Explorer

1 Cruise Report for AE2520 on R/V Atlantic Explorer September 2-15, 2025 Bermuda to Woods Hole MA Elizabeth B. Kujawinski with contributions from science party 2 3 Acknowledgements The activities of this cruise would not have been possible without the efforts of many individuals. First and foremost, we thank the National Science Foundation which funded the cruise through the Science & Technology Center grant for the Chemical Currencies of a Microbial Planet (Award #2019589). Secondly, we thank the captain and crew of the RV Atlantic Explorer, who provided a great platform for living and for doing science during our cruise. Lastly, we thank the C-CoMP staff, especially Laura Gray, the PIs, and the research groups whose efforts back on land made the sampling efforts successful. Photo (front): The science party of AE2520. Front row (left to right): Mak Saito, Natalie Graham, Fadime Stemmer, Mariana Torres, Emily Hu, Amanda Ellis, Liz Kujawinski. Second row: Mario Uchimiya, McKenzie Powers, Annika Gomez, Danny Gillissen, Eleanor Lawrence, Claire Garfield, Rachel Sandquist. Back row: Irina Koester, Mike Jakuba, Sonya Dyhrman, Justin Fujii (taken by Emily Tate, BIOS). 4 Table of Contents Table of Contents ACKNOWLEDGEMENTS ...................................................................................................................... 3 TABLE OF CONTENTS .......................................................................................................................... 4 CRUISE SUMMARY ............................................................................................................................. 6 OVERVIEW ..................................................................................................................................................... 6 STATION ACTIVITIES.......................................................................................................................................... 6 STATION CHARACTERISTICS ............................................................................................................................... 7 SAMPLE EQUIPMENT ........................................................................................................................................ 8 PARTICIPANTS AND PARAMETERS .................................................................................................... 10 LIST OF PARTICIPANTS WITH INSTITUTIONS .......................................................................................................... 10 LIST OF PARAMETERS AND RESPONSIBLE PARTIES ................................................................................................. 11 FIELD OBSERVATIONS AND SAMPLING MODALITIES .......................................................................... 12 OVERALL DATA MANAGEMENT .......................................................................................................................... 12 HYDROGRAPHY OVERVIEW CASTS ..................................................................................................................... 12 EDDY SAMPLING............................................................................................................................................ 14 LAGRANGIAN SAMPLING: SVP DRIFTERS ............................................................................................................ 16 LARGE-VOLUME SAMPLING: CLIO .................................................................................................................... 18 LARGE-VOLUME SAMPLING: UNDERWAY SYSTEM.................................................................................................. 24 TRICHO TOWS.................................................................................................................................. 25 OBSERVATIONS AND INCUBATIONS TIED TO DIEL FORCING .............................................................. 27 DIEL OBSERVATIONS: GENERAL PARAMETERS ..................................................................................................... 28 Nutrients ................................................................................................................................... 28 TOC/DOC .................................................................................................................................. 28 Bacterial abundance .................................................................................................................. 29 Particulate Organic Carbon (POC) ................................................................................................ 29 Bacterial Production: 3H-leucine incorporation ............................................................................. 29 Bacterial respiration: RSG ........................................................................................................... 32 DIEL OBSERVATIONS: META-TRANSCRIPTOMICS .................................................................................................. 32 DIEL OBSERVATIONS: METABOLOMICS .............................................................................................................. 33 DIEL OBSERVATIONS: “INVASION” EXPERIMENTS (BATS) ...................................................................................... 35 INCUBATIONS: METABOLITE DRAWDOWN ........................................................................................................... 38 Experimental design ................................................................................................................... 39 Methods .................................................................................................................................... 40 Preliminary Results .................................................................................................................... 43 INCUBATIONS: STABLE ISOTOPE PROBING .......................................................................................................... 43 Adenine metaP-SIP .................................................................................................................... 43 15N-Protein MetaP-SIP................................................................................................................. 45 Trichodesmium Adenine Protein-SIP ............................................................................................ 46 5 BIOLOGICAL DIFFERENCES BETWEEN STATIONS............................................................................... 47 PROTEASE ACTIVITY WITH DEPTH ....................................................................................................................... 47 ALKALINE PHOSPHATASE ................................................................................................................................. 50 PHYTOPLANKTON AND VIRUS ISOLATION............................................................................................................. 51 REFERENCES ................................................................................................................................... 55 CLIO OPERATIONS ON AE2520 .......................................................................................................... 57 6 Cruise summary Overview The goal of the cruise was to compare the microbial dynamics and processes in an oligotrophic region with a more productive region. This cruise is the second of two planned with the same purpose, but to compare the same regions under bloom conditions (spring; AE2504) with more stratified and nutrient-poor conditions (late summer; AE2520). We chose two stations: the Bermuda Atlantic Time-series Study site (BATS) near Bermuda and the southernmost station of the New England Shelf Long-Term Ecological Research (NES-LTER station L11) site west of Connecticut (Figure 1). At each station, we conducted diel sampling near the surface (15m) and placed these samples into context with hydrographic profiles, underway samples, targeted incubations, and large-volume samples from CLIO. We sampled each region within a Lagrangian context, enabled by an SVP drifter deployed at BATS (but not at the LTER site). In contrast to our spring cruise, we had very favorable weather. Consequently, we added a sampling site (the “Eddy”) for 3-4 days along the transect between BATS and NES-LTER-L11. At this site, we surveyed an anticyclonic eddy, completing a hydrographic survey of the eddy, and sampling the water column in the eddy center and in a gradient to its outward edges (see below). Station activities We conducted a hydrographic profile at the start of activities at BATS and NES-LTER-L11. This profile provided a snapshot of the physical, chemical, and biological dynamics of the upper 250 m. Prior to starting our diel sampling, we deployed an SVP drifter, which transmitted its position hourly. By sampling within 1km of the drifter’s position, we maintained a Lagrangian transect, allowing us to stay nearly within the same water mass throughout the sampling scheme. The drifter followed the water mass at 20m, while we sampled the water column at 15m. However, the mixed layer was always >30 m, so the drifter and our samples occurred within the same water mass. We deployed the CTD-Niskin rosette every 4 hours for approximately 72 hrs. From each diel time-point, we collected meta-transcriptomics samples, dissolved and particulate metabolomics, bacterial production and respiration, particulate organic carbon, nutrients, and total (and dissolved) organic carbon. We complemented these samples with meta-proteomics samples collected from the underway system (sampled at 2m depth) and with meta-genomics samples collected from CLIO (at 15m). We did not deploy an SVP drifter at NES-LTER-L11. To complement the field measurements, we conducted incubations using metabolite mixes from model axenic phytoplankton (Zhu, Anderson, et al. submitted). One set of incubations probed the response of the microbial community to these mixes through metabolite uptake measurements, bacterial meta-transcriptomics, and community meta-genomics. Another set of 7 incubations measured oxygen drawdown of the community to calculate respiratory quotients for different metabolites. At BATS, we examined the response of two heterotrophic bacteria, Ruegeria pomeroyi and Alteromonas macleodii, to surface seawater using bacterial invasion experiments (similar to Nowinski and Moran, 2021). At each station and in the eddy, we deployed the CLIO autonomous sampling vehicle to collect large-volume samples for meta-proteomics and meta-genomics. Station characteristics BATS. We occupied the BATS site (31° 50’N, 64° 10’W) starting on Tuesday September 2, 2025. As is typical for the summer, the mixed layer was relatively shallow (20-30m) and stable over the day. The sea surface temperature was approximately 29°C and the surface chlorophyll fluorescence was approximately 0.2 units. The subsurface chlorophyll maximum was relatively constant at approximately 120 m. We completed 72 hours of a diel sampling experiment, with approximately 24hr of hourly sampling in the middle. At the end of the diel sampling, we collected water for two incubations: (1) the metabolite draw-down and transcriptome response incubation (the “EC” incubation) and (2) an incubation with stable-isotope labeled purines and pyrimidines (the “SIP” incubation). The Dyhrman group deployed a net-tow to collect Trichodesmium colonies seven times at BATS. We completed three CLIO dives at BATS. The “Eddy”. Mara Freilich (Brown) and her team identified a cyclonic eddy (approximately 35° 53’N, 65° 36’W) for analysis and survey observations. The goal was to survey the eddy from the edges through the center. We collected water column profiles (“survey” casts) with the CTD sensors and water samples for meta-data analysis (“hydrocasts”) with the Niskin rosette bottles. Large-volume samples were collected from the DCM at approximately 1400 every day in the eddy center and at the eddy edges. One Tricho net-tow was collected. We completed two CLIO dives in the eddy; one on the edge and one in the approximate center. See below for more details. LTER. We arrived at L11 (39° 46’N, 70° 53’W) of the NES-LTER line on Friday September 12. The environment was very different from BATS (Figure 2). The sea surface temperature was lower (~20-25°C) and the fluorescence was higher (~2.5 units). We collected water for the “EC” incubation immediately to ensure the longest possible incubation time. We then completed an overview hydrocast at the same time that we initiated our diel observations (a CTD cast every 4 hours). On the second day of the diel experiment, we moved to hourly CTD casts. We completed two more CLIO dives at the LTER station. We concluded the diel observations at approximately 2100 and began our transit to Woods Hole MA. 8 Figure 1. Final cruise trajectory showing sea surface temperature and chlorophyll concentrations along the cruise track. Figure created by Arianna Krinos. Sample equipment The CTD rosette was provided by the BIOS Science Support Group and was prepared for initial deployment by marine technicians on our cruise, Emily Tate (lead) and Jace Innis, using the same set-up, sensors, and configurations as is used for the Bermuda Atlantic Time-Series Study CTD casts (see Methods from Johnson et al. 2025 for a complete description of the CTD configuration, sensors, and sampling method). The system was a SBE911+ CTD coupled with a variety of sensors and the SBE 32 Carousel Water Sampler. The sensors included: an internal Digiquartz pressure sensor, a Sea-Bird SBE-3F temperature sensor, a Sea-Bird SBE-4 conductivity cell, the Sea-Brid SBE-43 dissolved oxygen sensor, the WET Labs C-Star or Chelsea/Seatech Transmissometer, the Chelsea Aqua 3 Fluorometer, and the Biospherical/Licor PAR/Irradiance sensor. The rosette frame held 24, 12-L Niskin bottles. For planning purposes, we assumed an 11-L capacity because the Niskins rarely delivered 12 full liters. We focused our operations on the surface ocean. Every overview hydrocast was deployed to 500m to get an overview of the water column for CLIO operations; every diel hydrocast was 9 deployed to 200m to track the evolution of the deep chlorophyll maximum (DCM); and every hourly hydrocast was deployed to 50m to minimize deployment time. We used the underway seawater intake system on the AE to collect large volume samples for metaproteomics at each major station and for metaproteomics and metagenomics during the transit between stations. This system brings in water at approximately 2m below the surface at the bow of the ship and records sea surface temperature, salinity, and fluorescence. We used the tap in the AFT lab exclusively to maintain sufficient pressure in the line. Figure 2. Changes in air temperature, chlorophyll, water temperature, and wind speed along the cruise track of AE2520. Figure created by Arianna Krinos. 16 Lagrangian sampling: SVP Drifters Two 14” SVP Drifters (Pacific Gyre, Oceanside, CA, USA) were deployed and used to track water masses at 15 m during the diel sampling efforts at BATS and at the cyclonic eddy so that the same water mass could be sampled over time using a Lagrangian approach. The configuration and set-up of the SVP drifters were as follows: no six pack compression, drogue ON/OFF Sensor, no deployment status sensor, no water temperature sensor, no air pressure sensor, non-replaceable battery, 6-spoke drogue, six section drogue, fixed tether (length: standard GDP), no flasher, sampling period: 1 hour sampling. Additional sensors were not added to the drifters because they were only being used to track water movement at 15m, but we may consider adding sea surface temperature sensors in the future to encourage adoption of the drifters into the Global Drifter Program. The drifters were tested to ensure that the telemetry signal was detected. To do this, the drifter was placed outside with no obstructions surrounding it. Notably, SVP drifters were delivered ready to test without any additional assembly. The outer plastic cover was removed, while taking care to ensure the paper tape and parts securing the tether and drogue remained intact. Telemetry services were activated through the Pacific Gyre web dashboard and a pull-pin magnet was removed to turn on the transmitter. Telemetry service for WHOI_LG-SVP-0004 (Comms ID: 300534067527850) was activated on August 30th at 13:11 UTC. It took about ~2 hours after activation for the telemetry to lock onto the drifter position. The long time required could be attributed to a sudden downpour, which required the drifter to be re-wrapped in its plastic cover and moved under an overhang after which the drifter was moved back out into the open and locked onto the telemetry signal. Telemetry was activated for WHOI_LG-SVP-0001 (Comms ID: 300534067523870) on August 30th at 15:32 UTC. It took about ~28 minutes after activation for the telemetry to look onto the drifter position. The weather was very cloudy and overcast. After testing, the drifters were left activated (magnet out and telemetry activated) until deployment. Throughout testing, care was taken to ensure that the drifter was not dropped, since the shock of the drop could dislodge or tamper the drifter. The first drifter (WHOI_LG-SVP-0001) was deployed at the BATS station on September 2nd at BATS ~15 minutes before the first Diel cast (Cast 1). The drifters were left on the deck inside a large storage tote (where they could remain dry as water would dissolve the paper parts holding it together) and the telemetry signal was acquired prior to the time of deployment. The ship was directed to proceed dead-ahead into the wind and the drifter was deployed from the main deck at the stern. Telemetry services were deactivated on September 11th after 267 total hours. The second drifter (WHOI_LG-SVP-0004) was deployed at the center of the eddy (35° 42'N, 65° 30'W) on September 6th, 5 minutes before a survey hydrocast (Cast 45). The telemetry signal was acquired an hour after deployment. The same deployment strategy as the first drifter 17 was used. Telemetry services for this drifter were deactivated on September 18th after 430 total hours. Figure 4. Left photo: Eleanor Lawrence (left), Emily Hu (center), and Emily Tate (right) prepare to deploy the drifter off the stern at BATS for the diel observations). Photo by McKenzie Powers. Right photo: Eleanor (left) and Emily (right) unwrap the second drifter from its plastic cover for the eddy sampling. Photo by Claire Garfield. 18 Figure 5. Drifter positions (maroon crosses) at each hourly ping from the instrument at the BATS station. Points show the position of the ship within one hour of the sampling time as the vessel moved as close as possible to the current SVP position. Points indicating ship positions are colored by the primary chlorophyll measurement in μg/L from the underway system at that time. Black line segments indicate vessel trajectory between original ship position and drifter position at each hourly ping. Figure by Arianna Krinos. Large-volume sampling: CLIO CLIO operations. We conducted seven CLIO dives, detailed in Appendix 1. CLIO samples. We conducted a total of seven CLIO dives, all of which yielded viable samples (see CLIO report for details). The number of each sample collected at every station, are summarized in Table 3. 19 Table 3. Number of samples collected for each ‘omics analysis type from CLIO dives. * Only ¼ filter collected due to subsampling for incubations (see Section INCUBATIONS: Stable Isotope Probing). ** Only ⅛ of filter collected due to MetaT subsampling. Location MetaP [1/2 filter] MetaG [1/4 filter] MetaT [1/8 filter] Metabolomics [1/8 filter] Enzyme Essay [1/8 filter] clio055 BATS 36 36 0 36 8 clio056 BATS 36 36 0 36 14 clio057 BATS 34.5* 36 0 36 13 clio058 Eddy Edge (T3) 36 18** 36 36 14 clio059 Eddy Core (T2) 36 18** 36 36 15 clio060 LTER 34.5* 36 0 36 14 clio061 LTER 36 36 0 36 8 Omics sampling. The depths of sampling on each dive are summarized in Table 4. CLIO dives 55, 56, and 57 were conducted at BATS. The first dive (clio055) served as a test dive and thus spanned the surface in high resolution. The goal of dives clio056 and clio057 was to sample the epipelagic and mesopelagic, particularly the microbial community, and functional diversity at the surface (15 m), and around the DCM (at 150-200 m). Moreover, at dive 57, we tested CLIO’s capability to identify and automatically sample the DCM. CLIO dive 58 was used to sample the edge. Dive 59 was used to sample the core of an Eddy on transit between BATS and NES-LTER. Dives 60 and 61 were conducted at NES-LTER. On dives 58, 59, and 60, we performed adaptive sampling around the DCM. Dive clio061 was performed to test the sampling capabilities of CLIO along the seafloor. 20 Table 4. Depth of CLIO samples. SUPR / Dive + Depth [m] clio055 clio056 clio057 clio058 clio059 clio60 clio061 S1 250 1000 900 500 800 800 1608 S2 200 1000 700 300 500 600 1602.1 S3 175 800 500 200 300 400 1616 S4 150 800 300 150 200 300 1584 S5 140 600 250 135 150 200 1480.7 S6 130 600 150 125 50 150 1380.7 S7 120 400 110 50 50 100 1280.7 S8 110 250 116.9 30 50 54.9 1180.7 S9 100 200 15 15 50 54.8 1000 S10 90 150 250 500 132.2 200 1480.7 S11 80 100 80 300 118.5 100 1280.7 S12 70 80 90 200 66.7 43.2 1000 S13 60 70 100 110 60.4 162.6 54 S14 50 60 110 100 53.0 54.9 50 S15 40 50 120 90 43.5 186.4 50 S16 30 15 130 80 23.5 181.3 50 S17 15 15 140 70 30 30 50 S18 5 5 15 15 15 15 15 After each successful dive, filters were sliced and distributed as follows: ½ for metaproteomics, ¼ for metagenomics, ⅛ for protease assay, and ⅛ for metabolites, except for dives conducted in the Eddy edge (clio058) and Eddy core (clio059). Here metatranscriptomics samples were collected in addition, by splitting the DNA sample, i.e. ⅛ filter for metagenomics and ⅛ filter for metatranscriptomics. Filter slicing and packaging was done as quickly as possible and using frozen cryovial racks when possible. Metagenomics and metatranscriptomics samples were 21 flash frozen in liquid nitrogen before storage in the -80 °C freezer. Protein and metabolite samples were placed directly in the -80 °C freezer. Metaproteomics samples were extracted and are currently being analyzed on the LC-MS-MS in the Saito Lab at WHOI. A subset of metagenomics samples was sent to the Meren Lab at the Max Planck-Genome-centre Cologne (MP-GC) in Cologne, Germany for long-read sequencing. Metabolomics samples were prepared and analyzed in the Edison Lab at the University of Georgia. Metatranscriptomics samples are currently stored at the Saito Lab at WHOI but will be analyzed by the Freilich Lab at Brown University. Samples for phosphatase and protease assays were measured on board the AE within a few hours after sampling from CLIO. Incubation sampling. During CLIO dive 57, ¾ filters were taken from 15 m, 110 m, and 250 m depth and during CLIO dive 60, ¾ filters were collected from 200 m, 100 m, 54.9 m and resuspended in the respective filtered seawater. Both 0.2 µm and 51 µm filters were combined. The resulting sample slurry was used to inoculate sterile filtered surface seawater for incubation experiments as described in “INCUBATIONS: Stable Isotope Probing”. Nutrient sampling. Nutrient samples were analyzed from seawater collected in SUPR stacks during CLIO dives 56, 57, 59, 60 and 61, both before the filter (unfiltered) and after the filter (filtered). Nutrients were analyzed at the WHOI Nutrient Analytical Facility. During clio061, four samples at 50 m depth were analyzed and the mean values ± one standard deviation are summarized in Table 5. Depth profiles from all other dives are shown in Figure 6. Table 5. Summary of samples collected during CLIO dive 66 (LTER) at 50 m. Values are means from four independent samples ± one standard deviation. clio61 [NO3-+NO2] [NH4+] [PO43-] [SiO4] Unfiltered 0.15±0.10 2.4±0.6 0.14±0.20 0.7±0.3 Filtered 0.16±0.17 2.8±0.15 0.07±0.11 0.59±0.20 22 Figure 6. Depth profiles from nutrient samples collected during CLIO dives 56 (BATS), 57 (BATS), 59 (T2, Eddy core), and 60 (LTER), both unfiltered and filtered. Red: Nitrate, Blue: Ammonium, Green: Phosphate, Purple: Silicate. Figure by Fadime Stemmer. Metabolomics sampler. On CLIO dives 59, 60, and 61, we tested a new dissolved metabolomics sampler (Figure 7). The overall plan was to collect whole water into a sample bag stored within an in-situ incubation chamber. Prior to deployment, the incubation chamber was filled with freshwater. This water was displaced during pumping. Two bags were prepared but only the upper chamber pump worked during the deployments and thus three samples were collected. When the sample came on deck, the bag was removed from the incubation chamber and processed by the Kujawinski Team. Using peristaltic pumps, Teflon tubing was directly connected to the Luer Locks on the sampling bag. Water was filtered through a 0.2-um Omnipore filter, and filters were stored at -80 °C for intracellular metabolite analysis. The filtrate was acidified to pH2 and was processed as described for the PPL samples hydrocasts (see below). Different bag types were tested. The Supelco and Restek bags offered much better handling than the SKC. The straight tube extending from the Luer Lock fit perfectly into the 23 Telfon tubing, whereas the SKC fittings were narrower at the tip and prone to sucking air and leakage. The Supelco and Restek bags (provided by the Clio Team; see Table 6) also featured a simpler quarter-turn mechanism for opening and closing, while SKC bags (provided by Kujawinski Lab) were less straightforward to operate. Overall, filtration with the Supelco and Restek bags were smoother and is recommended for future use. However, only SCK bags have previously been tested by the Kujawinski lab for contamination. Table 6. Details about metabolomic samples taken from Clio. Clio Dive Pump duration Depth Bag Sample Volume 59 6min 73.5 m 0.6L Supelco Inert Foil Gas Sampling Bag Cat. No: 30279-U 600mL 60 6min 55m 3L SKC FlexFoil PLUS Sample Bag; Cat. No: 252-03 ~600mL 61 6min 54m 3L Restek Cat. No: 22951 ~2680ml 24 Figure 7. Clio samplers on AE2504. On the left is the new metabolomics sampler, with bags inserted into incubation chambers. On the right is the filter stack. Photo by Daniella Asturias. Large-volume sampling: Underway system We used the ship’s underway pump (intake at ~2 m depth) to sample for metaproteomics, metagenomics, metatranscriptomics, protease activity, phosphatase activity and metabolomics. This was done as a complement to the diel sampling efforts at BATS and LTER, and as opportunistic sampling of the transit between the two stations. Overall, we were interested in these samples for examining spatial and temporal shifts in community and/or functional characteristics at the surface. Samples for metaproteomics (metaP), metagenomes (metaG), and metabolomics were collected on 142 mm Versapor filters (51 μm and 0.2 μm). All samples except the metabolomics samples are currently stored in the Saito Lab freezers at WHOI. The metabolomics samples are stored and will be analyzed by the Edison Lab at UGA. MetaP samples were extracted and will be analyzed through LC-MS-MS in the Saito Lab at WHOI. At BATS, underway samples were collected every 2 hours between September 3rd-5th and between September 13th-14th at LTER, as well as every 3-4 hours while on station or under transit for the rest of the time. See Table 7 for a summary of underway sampling. Sampling procedure. Filters were placed inside a filter holder and connected directly to the ship’s underway faucet. Versapor filters were sliced into quarters and divided for metaP, metaG, metabolomics, and protease/phosphatase assays. Protein and metabolite samples were 25 placed directly in the -80 freezer and DNA samples were flash frozen in liquid nitrogen before transfer to the -80 freezer. Additional notes: ● We were not able to measure flowrate and thus total filtered volume on this cruise ● Note for 51 µm filters: Biomass would move around as the filters were being sliced. It is therefore likely that different slices of the same filter had different amounts of biomass on them. Table 7. Number and types of underway sampling events at BATS, during transit, at the Eddy, and LTER. metaP/metaG/metabolites sampling events Protease/Phosphatase Assay BATS 24 2 Transit 13 11 Eddy 18 17 LTER 17 17 Total 72 57 Tricho Tows Overview. A 130 μM phytoplankton net tow (General Oceanics) was deployed eight times across BATS (7 times) and Eddy (one time) sampling locations. During each deployment, the net was hauled three times at approximately 15m depth and filtered thousands of liters of water each time. After three tows, the water inside the net catch was emptied into a secondary container and immediately brought to the lab. Here, individual Trichodesmium colonies were isolated by hand using transfer pipettes and successive transfer through fresh 0.2 μm sterilefiltered local surface seawater three times to remove all but tightly associated microorganisms. Isolated colonies were then either: 1) harvested for ‘omics samples (metatranscriptomic and metagenomic) 2) transferred to single well cultures, 3) transferred to media plates, or 3) transferred to short term incubation experiments. 32 Bacterial respiration: RSG To examine diel patterns in single cell respiration rates, we collected samples at BATS using the RedoxSensor Green method described in Munson-McGee et al. (2022). Briefly, we incubated the natural microbial community from bulk seawater with 1 µl of RedoxSensor™ Green (Thermo Fisher Scientific), inverted gently several times, and incubated at 19.7 °C for 30 minutes. Samples preserved with cryopreservative (11x Tris-EDTA, 55% glycerol), incubated for one minute, and flash frozen in LN2. Samples were stored at -80 °C and transported on dry ice for further processing. Dead-cell controls were generated by amending 1 ml of sample to a final concentration of 9% cryopreservative, flash frozen in LN2, and stored at -80 °C. On shore, deadcell controls were thawed and incubated in 10% (v/v) paraformaldehyde at 4 °C for approximately 27 hours. Samples were then incubated with 1 µL RedoxSensor™ Green for 30 minutes at 4 °C, flash frozen in LN2, and stored at -80 °C. Diel Observations: Meta-Transcriptomics Overview. To track diel oscillations of microbial gene expression, we collected samples for metatranscriptomic analysis from eukaryotic and prokaryotic communities. Eukaryotic signals were size fractionated to isolate larger eukaryote (> 5 μm) signals from picoeukaryote (0.2 - 5 μm) signals, while prokaryotic metatranscriptomics were collected from the smaller size fraction (0.2 - 5 μm). Filtration. At each diel time point, 60 liters of seawater were collected from the CTD for bacterial and eukaryotic metatranscriptomic sampling. Over each 24-hour period, samples were collected every 4 hours for a total of six time points per period: three during the night (collected at 2200, 0200, and 0600 hours) under red light, and three during the day (collected at 1000, 1400, and 1800 hours) under ambient light. Seawater was collected into three 20 L carboys and processed using a peristaltic pump with two filtration lines per replicate (A,B,C) (Figure 12). Both lines for each replicate filtered seawater sequentially through a 5 μm polycarbonate (PC) filter and secondary 0.2 μm polyethersulfone (PES) Sterivex filters, with the filtrate collected into a graduated carboy. The PES filter from line 2 of each replicate was collected after 8 L was filtered through, and used for bacterial metatranscriptomics. Following filtration, 5 μm filters were gently vacuumed to remove excess water and transferred into cryovials. For the 0.2 μm Sterivex filters, excess water was expelled using a syringe before sealing the filters with end caps and putty. All samples (5 μm and 0.2 μm filters) were flash-frozen in liquid nitrogen and stored for the duration of the cruise. The average time from the start of filtration to flash freezing was approximately 50 minutes. Eukaryotic metatranscriptomics. A total of 339 eukaryotic metatranscriptomic samples (195 BATS, 144 LTER) were collected over the Diel experiments by members of the Dyhrman Lab and transported back to the Lamont Doherty Earth Observatory (LDEO) in liquid nitrogen 33 dewars. Upon arrival at LDEO the samples were transferred into long-term liquid nitrogen storage to await RNA extraction and sequencing by members of the Dyhrman Lab. The 5 μm filters will be poly-A selected prior to sequencing to explicitly detect the functional, proteincoding activity in the messenger RNA (mRNA) of the eukaryotic population. Bacterial metatranscriptomics. Bacterial metatranscriptomic samples were collected by filtering ~ 8 L sequentially through a 5 um filter and a 0.2 um polyethersulfone (PES) Sterivex filter (Millipore). Samples were flash frozen in LN2, transported on dry ice to the University of Georgia, and stored at -80 °C for further processing. Figure 12. Eukaryotic metatranscriptomic sampling time points during (left) daylight (1000, 1400, and 1800 hours) and (right) nighttime (2000, 0200, 0600 hours). Photos taken by Erin Maybach. Diel Observations: Metabolomics Overview. We collected four types of metabolomics samples for future analysis: PPL, Omni (intracellular), Benzoyl Chloride (BC), and Aniline derivatization samples (dissolved). Methods. From each time-point, we collected approximately 4.5L of seawater in two 2.5L Teflon bottles which were rinsed three times with seawater from the Niskin rosette. Samples were filtered through an in-line 0.2-μm Omnipore filter (Figure 12). For PPL extraction, the filtrate 34 was collected in acid cleaned 2.5L Teflon bottles rinsed 3x with sample filtrate. The volume of the filtrate was measured by a ruler on board and the height of the filtrate was recorded to calculate volume. Filtrate in the Teflon bottles was acidified to pH ~3 with concentrated HCl (2.5mL for 2.5L) and then extracted with PPL resin (Dittmar et al. 2008, with modifications from Longnecker 2015). PPL cartridges were eluted with methanol and stored at -20 °C for dry-down and analysis in Woods Hole. PPLs sampled at Cast 64 and after were frozen at -80 °C after drying and eluted with methanol brought on board when arriving in Woods Hole Dock. For dissolved metabolite analysis via the benzoyl chloride (BC) derivatization method (Widner et al., 2021), we collected 25mL filtrate in samples in amber 40mL EPA vials as triplicates. For the aniline derivatization method (Halloran et al., in prep), we collected ~30mL filtrate in a clear 40mL EPA vial. All EPA vials were combusted prior to the cruise and rinsed 3 times with filtrate. BC and aniline samples were stored at -20 °C. Intracellular metabolite data was also taken by collecting the 0.2um omnipore filter used in the in-line set up, and storing at -80 °C until extraction (Soule et al. 2015) at Woods Hole. At the LTER station we also collected filtrate for DOC using this set up. After collection of all filtrates, the filters were removed from the filter holders, folded cell-side inward, and placed inside cryovials, prior to storage at -80 °C. We estimate the total volume filtered by adding volumes of BC, Aniline and DOC samples to the volume measured of the PPL filtrate. The Kujawinski lab will analyze all metabolomics samples at WHOI. Total number of samples. BATS intracellular + BATS PPL + BATS BC + BATS aniline = 148; LTER intracellular + LTER PPL + LTER BC + LTER aniline = 136; Recorded single sample details for September. 35 Figure 13. (left) On ship peristaltic pump set up for filtration. (right) On ship set up for PPL solid phase extraction. Displays the vacuum manifold, teflon tubing, and 2.5L teflon bottles used in sample collection. Photos taken by Natalie Graham. Diel Observations: “Invasion” experiments (BATS) Overall goal. Bacterial ‘invasion’ studies were conducted across a diel cycle at the BATS station using the model bacteria R. pomeroyi DSS-3 and A. macleodii MIT1002. These short-term bioassays are used to identify the metabolites available in seawater via the expression of transporter and catabolic genes in the invading bacterium (Nowinski and Moran 2021). The primary goal of the invasion studies at BATS was to characterize the presence of labile metabolites across a diel cycle in the oligotrophic North Atlantic Ocean, as seen by two copiotrophic bacteria with distinct metabolic strategies and substrate preferences. Experimental design. A total of six invasion experiments were conducted at BATS across a diel cycle from September 3rd September 5th (see Table 9) using methods described in Nowinski and Moran (2021) with slight modifications (see Figure 14). Axenic cultures of R. pomeroyi and A. macleodii were prepared 38 hours before each invasion experiment. First, each bacterium was inoculated into ½ YTSS liquid medium and grown to stationary phase for 24 hours at 27 °C. Each culture was transferred into fresh ½ YTSS medium and incubated at 27 °C until they 36 reached late exponential phase (~14 hours). Cells were spun down once at 8,000 rpm for 3 minutes to remove rich medium and were resuspended in 0.22 µm filtered BATS seawater. Each bacterium was added into four replicate bottles containing 1-liter BATS whole seawater at a 1:1 ratio of invading bacterium cells to native bacterial cells. The mixtures were incubated in the dark at 27 °C for 90 minutes. Each replicate bottle was filtered through a 2 µm polycarbonate filter followed by a 0.22 µm PES Sterivex filter to collect the bacterial fraction. Filters were immediately flash frozen in liquid nitrogen followed by storage at –80 °C. Flow cytometry samples (500 ul fixed with 16% paraformaldehyde; 1% final concentration) were taken from BATS whole seawater, from each bacterial culture prior to being added into the seawater, and from the invasion incubation bottles at T-initial (0 minutes) and T-final (90 minutes). Two control experiments were conducted following the methods described above but R. pomeroyi and A. macleodii cultures were inoculated into L1 medium with either 10 mM glucose or no substrate. Post-cruise plan. There is a total of 210 flow cytometry samples, 64 bacterial RNA samples, and 46 eukaryotic RNA samples, that were collected from the six invasion experiments and two control experiments (see Table 9). All samples collected during the invasion experiments will be processed and analyzed by M. Powers (Moran Lab). Flow cytometry samples were run on the Aglient Quanteon analyzer at the University of Georgia Cytometry Shared Resource Laboratory in October. Bacterial RNA samples will be extracted using a modified protocol developed by the Moran lab in early December. Library preparation and Illumina sequencing will be completed at the University of Georgia Genomics and Bioinformatics Core (GGBC) in early December. Eukaryotic RNA samples (2 µm filters) were collected from each invasion experiment but there are currently no plans to sequence or analyze. 37 Table 9. Invasion experiment metadata and number of samples collected. Invasion Experiment Date Diel Time Point Cast Number FCM Bacterial RNA Euk RNA 1 September 3 6 AM 5 27 8 8 2 September 3 6 PM 8 27 8 8 3 September 3 10 PM 12 27 8 8 4 September 4 10 AM 24 27 8 8 5 September 5 2 AM 34 27 8 8 6 September 2 2 PM 37 27 8 8 Glucose control October 24 - - 24 8 - No substrate control October 24 - - 24 8 - Figure 14. Protocol for bacterial invasion experiments using the model marine bacteria R. pomeroyi DSS-3 and A. macleodii MIT1002 (four biological replicates per bacterium). Six bacterial invasion experiments using both bacteria were completed across a full diel cycle at BATS. Figure made by McKenzie Powers. 38 Table 10. Parameters and sample numbers from Diel casts. Parameter # samples BATS # samples LTER Responsible lab Nutrients 57** 14 Kujawinski TOC/DOC 75*** 80*** Kujawinski Bacterial abundance 19** 0 Moran/Freilich Particulate Organic Carbon (POC) 23*** 20*** Bates Bacterial respiration - RSG 52 0 Moran 16S & 18S rDNA 1 0 Freilich Prokaryotic MetaT 19** Included in LTER Eukaryotic MetaT count (0.2 size fraction) Moran/Dyhrman Eukaryotic MetaT* 132** 96** Dyhrman Bacterial production 93 55 Carlson Metabolites 148 136 Kujawinski “Invasion” experiments - bacterial abundance 210 NA Moran “Invasion experiments” - bacterial RNA 64 NA Moran “Invasion experiments” - euk RNA 48 NA Moran * Includes size fractions: 5 μm and 0.2 μm ** Samples taken as triplicates. *** Samples taken occasionally as triplicates (spot checks, usually during diel timepoints). “Metabolites” samples include PPL, Omni (intracellular), Benzoyl Chloride, and Aniline derivatization samples. Benzoyl Chloride and Aniline samples taken in triplicates, number reflects singles. Incubations: Metabolite Drawdown Experimental goal. In this experiment, we aimed to measure the microbial community’s response to metabolite mixtures designed based on culture data from (Zhu et al., 2025). Two metabolite mixes were used: one based on metabolites released by Prochlorococcus and the other based on metabolites released by Micromonas commoda (Table 11). We conducted two short-term incubation experiments lasting approximately 6 and 12 hours to assess metabolite drawdown and microbial activity. During each experiment, we collected samples for dissolved metabolite measurements to quantify assimilation rates (on a per- 39 molecule basis). At the end of each incubation, metatranscriptomic samples were collected to assess shifts in gene expression across the microbial community. In parallel, we conducted a long-term incubation to observe changes in the microbial gene pool composition in response to repeated metabolite additions. Long-term experiments at BATS and the LTER site lasted 97 and 72 hours, respectively, with recurring metabolite additions every 24 hours. At the conclusion of the long-term experiment, we harvested samples for metagenomic analysis to detect changes in the community’s functional potential in response to the metabolite treatments. Table 11. Final concentration for incubation amendments in µM carbon. Compound Prochlorococcus marinus (PM) Micromonas commoda (MC) Alanine 0.88 0.68 Aspartate 0.78 0.46 Glutamic Acid 3.68 1.17 Glycine 0.23 0 Isoleucine 0.94 0 Leucine 0.88 0.42 Lysine 0 0.40 Methionine 0.17 0 Phenylalanine 1.14 0.63 Proline 0.24 3.72 Taurine 0 1.53 Tyrosine 0 0.48 Valine 1.07 0.51 Experimental design Initial T0hr sampling and experimental setup. A total of 132 liters of seawater were collected from 15 m depth for the phytoplankton exometabolite incubation experiments. Seawater was collected in the morning (1000 hours at BATS; 0600 hours at LTER) to mimic the natural timing of morning carbon fluxes experienced by microbial communities. Initial measurements (T0hr) were taken directly from the CTD and included metatranscriptomes, metagenomes, and flow cytometry (FCM) (Fig CMG3, see Methods). At this time, 18 replicate 5 L incubation bottles were filled through a 200 μm nitex pre-filter. Three concurrent experiments were initiated: a short-term incubation (6 hours), an intermediate incubation (12 hours), and a long-term incubation (multiple days). Each experiment included triplicate 5 L bottles spiked with synthetic dissolved organic carbon (DOC) pools derived from either P. marinus or M. commoda to a final concentration of 10 μM carbon (Table 11). Additionally, triplicate 1 L bottles of 0.2 μmfiltered (PTFE/Teflon) seawater were amended with 10 μM P. marinus DOC to serve as an abiotic control. Following DOC addition, initial metabolite samples were collected from each incubation bottle (see Methods). 40 Short-term and intermediate incubation experiments. For the short-term (~6-hour) and intermediate (~12-hour) incubations and the abiotic control, metabolite and cell count (FCM) samples were collected every 3 hours to track drawdown. Metatranscriptomic samples were collected after ~6 and ~12 hours, respectively (Figure 15). Long-term incubation experiments. The long-term incubations lasted 97 hours at BATS and 72 hours at LTER. During the first 12 hours, FCM samples were collected every 3 hours alongside the shortand intermediate-term experiments. After the 24-hour mark (T24hr), each long-term incubation replicate was amended with an additional 1 mL of either P. marinus or M. commoda DOC mix. Before and after each DOC addition, metabolite samples were collected, and FCM samples were taken at each time point to monitor bacterial abundance. This amendment and sampling cycle was repeated every 24 hours thereafter. At the final time point (T97hr at BATS and T72hr at LTER), final metabolite and FCM samples were collected. All 5 L replicate bottles were then filtered for metagenomic analysis (see Methods; Figure 15). Methods Metatranscriptomics. Metatranscriptome samples from BATS were taken by filtering ~ 5 L of 200-um filtered water through a 0.2 µm PES membrane filter (Millipore) and flash-freezing in LN2. Bacterial transcriptomics samples were sent to the University of Georgia on dry ice for downstream analysis. Metatranscriptome samples from the LTER were taken by filtering ~ 5 L of 200-um filtered water through a 0.2 µm Sterivex filter and flash freezing and storage in LN2. A total of 15 metatranscriptome samples were collected by members of the Dyhrman lab and transported to Columbia University Lamont-Doherty Earth Observatory in liquid nitrogen dewars. Upon arrival at LDEO, the samples were transferred into long-term liquid nitrogen storage to await extraction and sequencing by members of the Dyhrman Lab. Metagenomics. Seawater was filtered through triplicate filtration lines (A, B, C) using a peristaltic pump for metagenome sampling. Each triplicate line filtered approximately 5 L onto a 0.2 μm polyethersulfone (PES) Sterivex filter, with the filtrate collected into a graduated carboy. Once filtering was complete, excess water was expelled from the filter containers using a syringe, and tubes were sealed with end caps and putty before being flash-frozen in liquid nitrogen and stored for the duration of the cruise. The average time from the start of filtration to flash freezing was approximately 30 minutes. A total of 18 metagenomic samples (9 BATS, 9 LTER) were collected from the incubation experiments by members of the Dyhrman Lab and transported back to the Lamont Doherty Earth Observatory (LDEO) in liquid nitrogen dewars. Upon arrival at LDEO, the samples were transferred into long-term liquid nitrogen storage to await DNA extraction and sequencing by members of the Dyhrman Lab. 41 FCM samples: BATS. Triplicate samples were fixed to a final concentration of 1% (v/v%) paraformaldehyde, incubated at 4 °C for 20 minutes, and stored at -80 °C. Samples were analyzed at CTEGD Cytometry Shared Resource Laboratory at the University of Georgia where they were stained with DAPI (1 ug/ml) and analyzed on an Agilent Quanteon flow cytometer (Acea, Biosciences Inc, San Diego CA) with a 405 nm laser with a DAPI (445/45 nm) bandpass filter and chlorophyll a (695/40 nm) bandpass filters, and a 561 nm laser with phycoerythrin (586/20 nm) bandpass filter. Heterotrophic bacteria were gated using DAPI fluorescence against chlorophyll fluorescence, Synechococcus was gated using phycoerythrin against chlorophyll, and other phytoplankton were gated using chlorophyll against forward scatter (Figure 16). Metabolite samples. Approximately 50-100 ml of incubation water was collected from each replicate bottle, resulting in biological replicates for each treatment. Samples were aliquoted into 125mL polycarbonate bottles before filtering. For extracellular profiles, the samples were filtered using teflon and an in-line 0.2-μm Omnipore filter. Biological replicates were collected as 25 mL filtrate samples in pre-combusted 40 mL amber EPA vials for dissolved metabolite analysis via the benzoyl chloride derivatization method (Widner et al., 2021). Vials were rinsed with filtrate before collecting the full 25mL sample. When sampling long term incubations, post spike samples were always taken after pre-spike samples to ensure no contamination from higher concentration media. Filtrate samples are being stored at -20 °C until processed at WHOI. 48 approximately 30-60 minutes. Details on the underway sampling can be found in section “Largevolume sampling: Underway Sampling”. At the beginning of each sampling step, sterile filtered seawater from the station was collected in 15 ml sampling tubes for background determination and sample dilution. Sample collection from CLIO. Seawater was filtered through 142 mm Versapor filters (51 µm, 0.2μm) as described in “Large-volume sampling: CLIO”. Sterile filtered seawater from each depth was collected in 15 ml falcon tubes from the clamshell filter head and utilized for background correction and sample dilution. Sample preparation. Filters were carefully removed from the filter head, cut to ⅛ filter slices and transferred to a sterile 15 ml falcon tube. To each sample tube, 5 ml of the respective seawater filtrate were added, and the tubes shaken to resuspend loose biomass. Using clean pincers the filter was removed, placed on a clean surface and residual biomass scratched off the filter with an inoculation loop. The filter and loop were placed back into the falcon tube and the tube shaken to resuspend the remaining biomass. Using a sterile syringe, approx. 2 mL of the sample were aspirated and sterile filtered through 0.2 µm syringe filters into a fresh eppendorf tube. This step is to remove biomass and isolate extracellular protease activity. Moreover, 500 µL of the cell suspension were transferred into an Eppendorf tube, fixed to a final concentration of 1% (v/v%) paraformaldehyde, incubated at 4 °C for 20 minutes, and stored at - 80 °C for downstream flow cytometry measurements. These measurements will be used to correct protease activity for biomass. Both the cell suspension as well as the sterile filtered sample were stored at 4 ˚C until sampled for protease activity. Protease Assay. The Pierce™ Fluorescent Protease Assay Kit (Thermo Scientific) was used for fluorometric protease activity determination. The workflow is illustrated in Figure 20. Of each resuspended sample, triplicates of 50 µl were diluted in a black, flat-bottom 96-well plate to a final volume of 200 µl with 0.5 µM, 1 µM, 2 µM and 3 µM of FTC-labelled casein in digestion buffer (25 mM Tris, 0.15 M NaCl, pH 7.2). The reaction was monitored on a Spectramax M3 spectrophotometer, measuring fluorescence every 30 seconds over the course of 4 hours at λexc./em.=485/538 nm and 25 ˚C. The data was processed and analyzed using a workflow written in python. In summary, samples were corrected by the respective media blank which was measured in triplicate and averaged before subtraction. The data was normalized against the first datapoint in the dataframe. A polynomial fit (R2≥0.90) was fitted against the normalized fluorescence vs. time curves, and the initial bulk proteolysis rate determined, i.e., the rate at the beginning of the experiment (time = 0, t0), where protease activity is not limited by substrate availability. The degree of the fit was chosen based on how well the fit represented the curves without oversimplification or overfitting. The proteolysis rates (PR) were converted from fluorescence intensity per minute (FI min-1) to units of fmol fluorescent substrate overturned per minute (fmol min-1) using an empirically determined conversion factor of 2.5 fmol FI-1 across samples. Samples that were not measured yet were stored at -20˚C. 49 Figure 20. Protocol for protease activity sample preparation and measurement. Filters were obtained from Underway or CLIO sampling. The filters were extracted and cut to ⅛ of the original filter size. The biomass was resuspended in sterile filtered seawater and the sample used in a fluorometric protease activity assay. Figure created in Bio Render by Fadime Stemmer. Depth profiles of protease activity from CLIO dives Fluorescence intensity was plotted against time for each Clio protease activity sample. The initial protease activity (at time t=0) was determined for each depth measured at the highest sampled FTC-casein concentration of 3 µM (Figure 21). This fluorescent substrate concentration was used since it is closest to the substrate saturation condition among the conditions tested. Figure 21. Initial bulk protease activity (V0 in fmol min-1) from Clio dives clio056 (BATS); clio057 (BATS); clio058 (Eddy Edge, T3); clio059 (Eddy core, T2); clio060 (LTER); clio061 (LTER). Rates were determined from samples measured at a FITC casein concentration of 3 μM. Measurements were made at 25˚C. Blue: 0.2 µm filter, Green: 51 µm filter. Figure created by Fadime Stemmer. 50 Alkaline phosphatase Samples were collected using the Clio AUV during the cruise of AE2520 (Table 14). Four sample types were used for the alkaline phosphatase (APase) activity measurement: ● Unfiltered seawater: Seawater samples collected from the top clamshells ● Filtered seawater: Seawater samples collected from the bottom clamshells ● 51-µm filter: 51-µm filter samples ● 0.2-µm filter: 0.2-µm filter samples For the filtered samples, a fraction (either ⅛ or ½) of each filter was scraped, and particles were resuspended in filtered seawater. For the APase measurement, 190 µL of samples were mixed with 20 µL of a substrate solution, and the absorbance at 405 nm was recorded using a plate reader for three hours. Filter samples were measured with either two or three replicates, whereas seawater samples were measured as singulates. Three blanks were prepared for each plate reader run. Figure 1 shows our preliminary results of APase distributions conducted at three stations (BATS, Eddy, and LTER). For this preliminary report, we used the maximum values of reads for individual samples as an index of APase activity. Also, for the filter samples, values were normalized by filtration volume. In the upcoming final product, the values will be expressed as absolute rate, and also normalized by total protein. Table 14. Summary of samples used for the alkaline phosphatase activity measurement during AE2520. The samples were collected using the Clio AUV. Dive Station Nutrients Filters Note 56 BATS 18 depths 14 57 BATS 8 13 58 EddyT3 0 14 59 EddyT2 18 18 Includes Volume Experiment (see below) 60 LTER11 10 18 61 LTER11 0 11 Includes Volume Experiment (see below) 51 Figure 22: Results for the alkaline phosphatase activity measurements. Note that the values are preliminary (see the text for details). Fluorescence values were provided by Mike Jakuba. Phytoplankton and virus isolation Overall goal. The overwhelming majority of marine eukaryotic viruses identified by metaomics have yet to be isolated in culture. The goal of this effort was to collect samples for downstream isolation of novel eukaryotic virus-host pairs. Viruses are obligate parasites, so having a susceptible host in culture is required for virus isolation. Representative eukaryotic phytoplankton strains from oligotrophic (BATS & eddy) and coastal (NES-LTER) sites will be isolated to serve as bait strains for virus isolation as well as experimental validation of hypotheses generated from analysis of other data types. Isolations. Five full-scale isolation efforts were performed, one at BATS, three in the eddy (edge, center, and gradient), and one at the LTER. 10-20L of whole seawater was collected from 15m (15m and DCM at the eddy sites) and serially filtered through a 200µm Nytex prefilter, 20µm, 5µm, and 0.8µm membrane filters and an 0.2µm Sterivex unit with three parallel lines (Figure 23). The 20µm, 5µm, and 0.8µm membrane filters were resuspended in 20 ml of sterile artificial seawater (ASW). All three filters of the same size were combined into a single resuspension, resulting in a 1000X concentration of each size fraction. The 20mm resuspension was aliquoted into 4 12-well microplates which contained L1/25 (BATS/Eddy) or L1/5 (LTER). The 5um and 0.8µm resuspensions were each aliquoted into 12-well microplates which contained L1/15 (BATS/Eddy) or L1/5 (LTER). Microplates were incubated at approximately 22 °C under a 12:12 light: dark cycle for the duration of the cruise. Due to incubator conditions, temperature ranged from ~20 °C at the end of the night to ~26°C during the day. Upon return, the samples were transferred to 24 °C (BATS), 22 °C (Eddy), and 18 °C (LTER) incubators. Approximately 2 weeks after return, serial dilutions were performed, with the addition of single cell picking when needed, to begin generating unialgal isolates. As of December 15th, there are 6 unialgal isolate candidates from BATS, 10 from the eddy, and 6 from the LTER (Figure 24). Downstream analysis will determine the taxonomic identity of these isolates via 18S sequencing. 52 Concentration of viral size fractions is commonly employed in marine environments to enhance virus isolation efforts (Suttle et al 1991). Giant viruses (Nucleocytoviricota > 0.2µm) were concentrated by collecting the 0.8µm filtrate on a Sterivex unit. The three filters were resuspended in 10ml of ASW, resulting in a 2000X concentration. The resuspensions were stored at 4 °C for the duration of the cruise. Viruses smaller than 0.2mm were concentrated via tangential flow filtration using Vivaflow cassettes with a molecular weight cutoff of 30kDa. Concentration factors varied but ranged from 100-600X. The concentrates were stored at 4 °C for the duration of the cruise. In addition to the full-scale isolation samples, 0.2µm filtered water was collected from the metatranscriptome sample filtrate at varying time points (Table 15). Immediate plans for virus isolation from concentrates involve screening previously isolated phytoplankton strains, predominantly diatoms, for lysis after addition of the concentrates. Future efforts will involve similar screening of novel phytoplankton isolates. Table 15. Samples collected for phytoplankton and virus isolation. Site Cast # Time (EDT) Type Volume collected Virus concentrate 0.2µm resuspension Cultures Crosslinking BATS 2 19:22 30kDa concentration 9.5L X BATS 10 20:01 Formalin cross-link 79L X BATS 33 22:06 30kDa concentration 5L X BATS 34 2:08 30kDa concentration 6L X BATS 39 18:20 Full isolation 20L X X X BATS 42 15:31 30kDa concentration 5L X Eddy 51 10:14 30kDa concentration 8L X Eddy 53 15:43 Full isolation 17.5L (Sur & DCM) X X X Eddy 54 17:27 30kDa concentration 8L X Eddy 56 14:00 Full isolation 14L (Sur & DCM) X X X Eddy 58 21:38 Full isolation 12L (Sur & DCM) X X X 53 LTER 72 16:00 Full isolation & cross-link 88L X X X X LTER 93 14:10 30kDa concentration 6L X LTER 94 17:53 30kDa concentration 6L X Figure 23. Phytoplankton isolation and virus concentration workflow. Figure created by Annika Gomez. 54 Figure 24. Haptophytes (top) and pennate diatoms (bottom) from the eddy at varying stages of isolation. Figure by Annika Gomez. 55 References Braakman, R., Satinsky, B., O’Keefe, T. J., Longnecker, K., Hogle, S. L., Becker, J. W., ... & Chisholm, S. W. (2025). Global niche partitioning of purine and pyrimidine cross-feeding among ocean microbes. Science Advances, 11(1), eadp1949. https://doi.org/10.1126/sciadv.adp1949. Cao, H., Freilich, M., Song, X., Jing, Z., Fox‐Kemper, B., Qiu, B., ... & Chen, D. (2024). Isopycnal submesoscale stirring crucially sustaining subsurface chlorophyll maximum in ocean cyclonic eddies. Geophysical Research Letters, 51(4), e2023GL105793. https://doi.org/10.1029/2023GL105793. Dittmar, T., Koch, B., Hertkorn, N., & Kattner, G. (2008). A simple and efficient method for the solid-phase extraction of dissolved organic matter (SPE-DOM) from seawater. Limnology and Oceanography Methods, 6, 230–235. https://doi.org/10.4319/lom.2008.6.230. Freilich, M. A., Poirier, C., Dever, M., Alou-Font, E., Allen, J., Cabornero, A., ... & Mahadevan, A. (2024). 3D intrusions transport active surface microbial assemblages to the dark ocean. Proceedings of the National Academy of Sciences, 121(19), e2319937121. https://doi.org/10.1073/pnas.2319937121. Frischkorn, K. R., Rouco, M., Van Mooy, B. A., & Dyhrman, S. T. (2017). Epibionts dominate metabolic functional potential of Trichodesmium colonies from the oligotrophic ocean. The ISME journal, 11(9), 2090-2101. https://doi.org/10.1038/ismej.2017.74. Halloran, K., M. C. Kido Soule, H. Anderson, Y. Zhu, E. McParland, S. Haley, S. T. Dyhrman, & E. B. Kujawinski. In prep. Quantification of dissolved carboxylate-, carbonyl-, and phosphatecontaining marine metabolites via aniline derivatization with liquid chromatography-tandem mass spectrometry. [based on Chapter 3 of PhD thesis by K. Halloran; 2024 MIT/WHOI Joint Program in Oceanography and Applied Ocean Sciences]. Johnson, R. J., Bates, N., Lethaby, P. J., Smith, D., Chambers, E. (2025) Two decibar averaged CTD profiles collected at the Bermuda Atlantic Time-series Study (BATS) site from October 1988 through July 2024. Biological and Chemical Oceanography Data Management Office (BCODMO). (Version 9) Version Date 2025-02-26. doi:10.26008/1912/bco-dmo.3918.9 [April 7, 2025]. Longnecker, K. (2015). Dissolved organic matter in newly formed sea ice and surface seawater. Geochimica et Cosmochimica Acta, 171, 39-49. https://doi.org/10.1016/j.gca.2015.08.014. Munson-McGee, J. H., Lindsay, M. R., Sintes, E., Brown, J. M., D’Angelo, T., Brown, J., Lubelczyk, L. C., Tomko, P., Emerson, D., Orcutt, B. N., Poulton, N. J., Herndl, G. J., & Stepanauskas, R. (2022). Decoupling of respiration rates and abundance in marine prokaryoplankton. Nature, 612(7941), 764-770. https://doi.org/10.1038/s41586-022-05505-3. 56 Nowinski, B., & Moran, M. A. (2021). Niche dimensions of a marine bacterium are identified using invasion studies in coastal seawater. Nature Microbiology, 6(4), 524-532. https://doi.org/10.1038/s41564-020-00851-2. Soule, M. C. K., Longnecker, K., Johnson, W. M., & Kujawinski, E. B. (2015). Environmental metabolomics: analytical strategies. Marine Chemistry, 177, 374-387. https://doi.org/10.1016/j.marchem.2015.06.029. Suttle, C. A., Chan, A. M., & Cottrell, M. T. (1991). Use of ultrafiltration to isolate viruses from seawater which are pathogens of marine phytoplankton. Applied and Environmental Microbiology, 57(3), 721–726. https://doi.org/10.1128/aem.57.3.721-726.1991. Widner, B., Kido Soule, M. C., Ferrer-González, F. X., Moran, M. A., & Kujawinski, E. B. (2021). Quantification of amine-and alcohol-containing metabolites in saline samples using preextraction benzoyl chloride derivatization and ultrahigh performance liquid chromatography tandem mass spectrometry (UHPLC MS/MS). Analytical Chemistry, 93(11), 4809-4817. https://doi.org/10.1021/acs.analchem.0c03769. Zhu, Y., Anderson, H. S., Salcedo, E., Miller, S. E., Longnecker, K., Soule, M. C. K., ... & Kujawinski, E. B. (2025). Characterization of phytoplankton-excreted metabolites mediating carbon flux through the surface ocean. bioRxiv, 2025-11. https://doi.org/10.1101/2025.11.04.686593. 57 CLIO Operations on AE2520 Figure 2: clio055: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 3: clio055: Profiling sensor data to maximum depth. 4 Figure 4: clio055: Profiling sensor data, upper water column. 5 Figure 5: clio055: Profiling sensor data vs. time, whole dive. 6 Figure 6: clio055: Control system performance. 7 Post dive notes Launch and recovery were both smooth in 15 - 18 kt winds and mild swell. Ballasting was accurately predicted despite large changes to payload. The vehicle ”skipped” one sample for reasons that are not yet clear but probably relate to a timeout (there are many). It came up 35 minutes early as a result, but acomms were solid enough to know this ahead of time. We got occasional TWTT ranges out to 900 m, and solid uplinks (MO) whenever we put the fish over the side. SUPR Tigerclaw performed well with lots of color on the deep filters (250 m max); the near-surface is oligotrophic. SUPR BushBaby had problems after the first 4 samples. The first of the four was the mostly-skipped sample, so in the end we got 3 good samples out of BB. Thereafter pump driver board gradually failed; however we believe the flow sensor to be accurate and at least some water was pumped on all except the last sample. The board was replaced with a reworked board (electrolytic capacitors replaced with tantalum), tested, and the SUPR remounted. 8 Clio 056 Dive Report Michael Jakuba, Justin Fujii, Chip Breier, Daniel Gomez-Ibanez Clio Shipboard Operations Team Leader: Michael Jakuba Summary clio056 was an upper water-column sampling dive. All samples were collected at constant depth (isobaric). Weather: Fair with light winds and long period swell. Reason for end of dive: End of planned mission. Dive Statistics Launch: 2025-09-04 20:10:34 UTC, 31 ° 30.8287’ N, 064 ° 37.1104’ W On surface: 2025-09-05 10:13:33 UTC, 31 ° 27.4253’ N, 064 ° 38.6719’ W Recovery: 2025-09-05 10:51:04 UTC, 31 ° 27.2171’ N, 064 ° 38.7149’ W Maximum depth: 1000 m Dive time: 14:00 Payload Sensors  Bushbaby SUPR  Tigerclaw SUPR  Aanderaa 4831FDW optode  SBE SeaCAT49 CTD  SBE DEEPSUNA V2 nitrate analyzer (out of calibration)  WetLabs FLUNTRTD fluorometer (Chl-a/Turbidity) 1 Figure 7: clio056: Bushbaby SUPR sampling history, valve and volume pumped versus depth station. 2 Figure 8: clio056: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 9: clio056: Profiling sensor data to maximum depth. 4 Figure 14: clio057: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 15: clio057: Profiling sensor data to maximum depth. 4 Figure 16: clio057: Profiling sensor data, upper water column. 5 Figure 17: clio057: Profiling sensor data vs. time, whole dive. 6 Figure 18: clio057: Control system performance. 7 Post dive notes Launch and recovery were both smooth in 10 kt winds and mild swell. Acomms was good during the initial descent; however, no TWTT ranges were received. The towfish was deployed for all of the descent and removed thereafter, except for a brief 10 minute dunk near 00:00 local. The vehicle was probably mostly out of range. Both SUPRs performed well; however Tigerclaw pumped significantly more volume on the deep samples than before, and somewhat less on the shallow samples. There was an error in the initialization file (duplicate adapt_id) that caused adaptive selection of the isopycnal sample to be commanded to both the nitracline and to the DCM. This was benign in that the DCM was randomly chosen. It is unclear how well the isopycnal sample worked. The T signal is dominated by occasional sharp changes (0.05 degC) and oscillations of about the same magnitude (0.05 degC) on the same period as the vehicle’s rotational period. This oscillation is always present and its source unknown. The controller clearly responded in the correct direction to T changes, but tuning was probably poor. The controller was tuned in simulation for a dT/dz of - 0.14 degC/m whereas the observed dT/dz in the vicinity of the DCM was -0.0032 degC/m, two orders of magnitude less of a gradient. Isopycnal samples during CLiOMZ were performed in much stronger gradients. 8 Clio 058 Dive Report Michael Jakuba, Justin Fujii, Chip Breier, Daniel Gomez-Ibanez Clio Shipboard Operations Team Leader: Michael Jakuba Summary clio058 was an upper water-column sampling dive at the edge of a cold-core eddy. The mission plan included adaptive selection of the depth of the DCM and deep DCM, although no deep DCM was present. Five sample depths were arrayed centered at the shallow DCM and spaced by 10 m. All adaptively selected sample depths were done in isothermal-tracking mode. The remaining sample depths were isobaric. Weather: Fair with moderate winds to 18 kt and long period swell. Reason for end of dive: End of planned mission. Dive Statistics Launch: 2025-09-08 02:14:34 UTC, 36 ° 5.1407’ N, 065 ° 41.3557’ W On surface: 2025-09-08 12:22:27 UTC, 35 ° 56.5432’ N, 065 ° 54.3527’ W Recovery: 2025-09-08 12:24:33 UTC, 35 ° 56.4686’ N, 065 ° 54.3866’ W Maximum depth: 500 m Dive time: 10:00 Payload Sensors  Bushbaby SUPR  Tigerclaw SUPR  Aanderaa 4831FDW optode  SBE SeaCAT49 CTD  SBE DEEPSUNA V2 nitrate analyzer (out of calibration)  WetLabs FLUNTRTD fluorometer (Chl-a/Turbidity) 1 Figure 19: clio058: Bushbaby SUPR sampling history, valve and volume pumped versus depth station. 2 Figure 20: clio058: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 21: clio058: Profiling sensor data to maximum depth. 4 Figure 26: clio059: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 27: clio059: Profiling sensor data to maximum depth. 4 Figure 28: clio059: Profiling sensor data, upper water column. 5 Figure 29: clio059: Profiling sensor data vs. time, whole dive. 6 Figure 30: clio059: Control system performance. 7 Post dive notes Launch was smooth in moderate winds and mild swell. Recovery went well in higher winds to 22 kt and moderate swell. The vehicle drifted 6 nmi over the course of the 15 hr dive in variable currents near the center of the eddy. Acomms was good during the initial descent though no TWTT ranges were received. Later attempts to communicate with the vehicle were unsuccessful, probably because it was out of range. The incubator test was successful insomuch as the bag was filled with water though the measured pump volume of 9 l relaive to the bag size (<1 l) suggests a leaky flow path. A mission planning error resulted in the sample being smeared across the interval of 100 m to the adaptively-selected DCM at 60 m. (The isopycnal sample type was set correctly “hold-over;” however in this case the depth-goal-type should have been set to “relative” because the depth of the preceding sample was selected adatively, however, depth-goal-type was left at the default of absolute. This caused the vehicle to first head to the absolute depth goal and then switch on isothermal tracking which brought it back to the intended depth but with the pump already running. Iridium/GPS was spotty on recovery with perhaps 50% of messages indicating “GPS Timeout.” RDF was solid at less than 2 nmi and audible to at least 6 nmi. Initially the RDF indicated a correct direction off the port side, but once the ship turned to face the signal, the RDF indicated a location on the port quarter which gradually drifted toward the bow as the ship closed distance. The isopycnal tracker lost tracking halfway through the last DCM comb sample. This is apparently because the vehicle strayed into the mixed layer on this shallowest sample of the comb. Thereafter all samples were isobaric and collected without issue. Both SUPRs performed well. The shallow DCM was correctly identified at 60 m based on a 2-m binned average profile generated mid-dive. This was used as the 2nd sample depth in a 5-sample comb per the mission specification. Roll control produced a result consistent with simulation—-a maximum achieved roll angle at speed of 3-4 deg. 8 Clio 060 Dive Report Michael Jakuba, Justin Fujii, Chip Breier, Daniel Gomez-Ibanez Clio Shipboard Operations Team Leader: Michael Jakuba Summary clio060 was an upper water-column sampling dive at LTER-11. The mission included adaptive selection of the first two peaks in Chl-a. Four additional sample depths were planned centered around the highest Chl-a peak. All samples were planned for iso-thermal mode. The dive included an array of constant depth samples at 800 m and shallower, as well as at 30 m and 15 m. Two incubation chambers were installed with samples planned for the highest Chl-a peak and at 15 m. Weather: Fair with light winds to 10 kt and medium period swell. Recovery was in 10 kt winds and medium swell. Reason for end of dive: End of planned mission. Dive Statistics Launch: 2025-09-13 00:25:33 UTC, 39 ° 46.0219’ N, 070 ° 53.0406’ W On surface: 2025-09-13 15:48:21 UTC, 39 ° 45.9871’ N, 070 ° 29.9956’ W Recovery: 2025-09-13 15:53:27 UTC, 39 ° 45.8555’ N, 070 ° 29.8786’ W Maximum depth: 800 m Dive time: 15:17 Payload Sensors  Bushbaby SUPR  Tigerclaw SUPR  Aanderaa 4831FDW optode  SBE SeaCAT49 CTD  SBE DEEPSUNA V2 nitrate analyzer (out of calibration)  WetLabs FLUNTRTD fluorometer (Chl-a/Turbidity) 1 Figure 31: clio060: Bushbaby SUPR sampling history, valve and volume pumped versus depth station. 2 Figure 32: clio060: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 33: clio060: Profiling sensor data to maximum depth. 4 Figure 38: clio061: Tigerclaw SUPR sampling history, valve and volume pumped versus depth station. 3 Figure 39: clio061: Profiling sensor data to maximum depth. 4 Figure 40: clio061: Profiling sensor data, upper water column. 5 Figure 41: clio061: Profiling sensor data vs. time, whole dive. 6 Figure 42: clio061: Control system performance. 7 Post dive notes Launch was smooth in light winds and mild swell. Recovery went well in light winds and mild swell. The vehicle drifted 8 nmi over the course of the dive. Acomms was good during the initial descent though no TWTT ranges were received. Altitude acquisition and tracking worked well. Tracking was noisier at high altitudes—this is presumably the result of noisier altitude readings from the altimeter or possibly because small attitude changes would result in larger effective altitude noise since altitude is not compensated for tilt. The sample planned for the DCM worked well—the sample bag was returned full, somewhat less than 3 l. The 15 m sample failed because of a mission planning error. The purge and sample were both set to Bushbaby port 3, a copy-and-paste error from the preceding SuPR sample. The volume pumped estimate for Bushbaby port 3 was manually corrected in the dive-planning spreadsheet post dive. 1