Full details for collection and analyses of underway POM samples are provided by Holser et al., 2011 and Goñi et al., 2019 and Goñi et al., submitted. Brief summaries are provided below.

Samples for this study were collected aboard RV Oceanus using the surface underway scientific system.

Aboard the vessel we had access to uncontaminated seawater and collected samples at specific times that allowed us to determine location (latitude and longitude) and seawater characteristics (temperature and salinity) from the ships’ navigation and sensor panels. We used a semi-automated filtration system (SAFS) described by Goñi et al., (2019) connected to Oceanus surface underway water to collect particulate organic matter samples. Surface underway water was connected to the SAFS through a manual flow-control valve via opaque polyethylene tubing. A fly wheel flow meter was placed in-line and connected to a laptop computer using a data acquisition system to measure and record flows during the filtration stage. A switching valve with 8 ports was placed downstream from the flow meter and controlled by the laptop. Under stand-by conditions, flow was directed to the ‘waste’ port, which was fitted with unobstructed tubing that drained into one of the ship’s sinks and flowed back to sea. The 8-sample ports were fitted with tubing, quick-turn sockets and in-line stainless steel 13-mm Swinney filter holders. The flow from these filters was directed to the same sink as the ‘waste’ flow. In each holder, we placed one pre-combusted (400 oC for 3 hours) 13-mm glass fiber filter supported by a stainless steel screen and locked into place with a Teflon o-ring that prevents leakage and results in a filtration area of 78.5 mm2. Once filters were fitted in each of the sample ports, the filtration program was started to collect samples at selected intervals.

Once the filtration run was completed, the filter housings were removed from the SAFS, opened, and each individual filter folded into pre-cleaned silver capsules, which were placed into sample trays that were frozen until CN analyses. Each sample was assigned a specific time stamp (start-end of filtration process) that coincided with the ship’s clock and allowed us to retrieve location and oceanographic data for each sample, as well as determine an overall filtration volume, which was used to calculate particulate nitrogen and carbon concentrations once their contents were determined. During normal operations, we stacked two filter holders at specific positions in order to collect both particles from a sample using the first filter as well as measure blanks associated with dissolved organic matter sorption as filtered water goes through the second filter.

Carbon and nitrogen analyses were conducted using high temperature combustion/reduction according to Holser et al., 2011 and Goñi et al., 2019. Sample and blank filters were exposed to concentrated hydrochloric acid fumes to remove carbonates and run in two CN analyzers (NC2500 Thermoquest and ECS 4010 Costech) using the manufacturers’ recommendations for carbon and nitrogen analyses (e.g., specified temperatures for combustion and reduction furnaces, O2 loops/pressure, and the use of a water trap). In each run of a full auto-sampler, we typically included 6 standards (e.g., cystine, atropine) with different and known amounts of carbon and nitrogen to develop distinct calibration curves for each run. All filters were treated the same and we used the DOM blanks to correct for DOC and DN sorption. Detection limits for OC and N were 0.04 and 0.03 micromoles, respectively. Replicate analyses of selected samples yielded average standard errors for both measurements of ~2% of measured values.

References cited: Holser et al., 2011; Goñi et al., 2019; Goñi et al., submitted.