Dataset: Sinking particles from traps

Sampling and analytical procedures: 

Particle flux measurements from the Chief Scientist Training cruise generally followed the methods of the Hawaii Ocean Time-Series (HOT). Measurements were made at 75 m, 150 m, and 300 m using multiple cylindrical particle interceptor traps deployed on a free-floating array for approximately 3 days twice during cruise KM1910 at station ALOHA. Sediment trap design and collection methods are described in Winn et al. (1991). Three traps from each depth from both deployments were analyzed for POC at WHOI , one trap from each depth was analyzed for PIC at WHOI, three traps from each depth from first deployment were analyzed for PC, PN, d13C and d15N at Princeton lab, and three traps from each depth from the first deployment were analyzed for PP at University of Hawaii.

Acquisition Description 

The information below was adapted from the BCO-DMO HOT Dataset: Particle Flux page at https://www.bco-dmo.org/dataset/737393 (last visited on 2021-03-30) and the HOT Field & Laboratory Protocols page, found at http://hahana.soest.hawaii.edu/hot/protocols/protocols.html# (last visited on 2021-03-30). Differences between this dataset and a typical HOT particle flux data set include 1) depths sampled, 2) units in mmol m-2 d-1 (instead of mg m-2 d-1), and 3) suite of analyses (PP, PC, and PN but not Si, additionally POC (by acidifying filters) and PIC).

SUMMARY: Passively sinking particulate matter is collected using a free-floating sediment array and, after prescreening (335 μm) to remove zooplankton and micronekton carcasses, the sample materials are analyzed for P (at UH), POC and PIC (at WHOI), C and N (mmol m-2 d-1) and d13C (permil vs. VPDB) and d15N (permil vs. air-N2) (at Princeton). 

1. Principle

Sediment traps capture the downward flux of particulate matter (different from in situ suspended particles) and allow for its chemical analysis. During the KM1910 cruise we deployed a free-drifting sediment trap array with 12 individual collectors positioned at 75, 150 and 300 m. The deployment periods were 3.3875 days and 3.04027 days. The passively sinking particles are subsequently analyzed for a variety of chemical properties.

2. Precautions

Because particle fluxes in oligotrophic habitats are expected to be low, special attention must be paid to the preparation of individual sediment trap collector tubes so that they are clean and free of dust and other potentially contaminating particles. Traps should be capped immediately after filling and immediately after retrieval. Pay particular attention to airborne and/or shipboard particulate contamination sources. In addition, the time interval between trap retrieval and subsample filtration should be minimized in order to limit the inclusion of extraneous abiotic particles and the post-collection solubilization of particles.

3. Field Operations

3.1. Hardware

Our free-floating sediment trap array is patterned after the MULTITRAP system pioneered by Knauer et al. (1979) and used extensively in the decade-long VERTEX program. Twelve individual sediment trap collectors (0.0039 m2) are typically deployed at three depths (150, 300 and 500 m). The traps are affixed to a PVC cross attached to 1/2" polypropylene line. The traps are tracked using VHF radio and Argos satellite transmitters and strobelights. 

3.2. Trap solutions

Prior to deployment, each trap is cleaned with 1 M HCl, rinsed thoroughly with deionized water then filled with a high-density solution to prevent advective-diffusive loss of extractants and preservatives during the deployment period and to eliminate flushing of the traps during recovery (Knauer et al., 1979). The trap solution is prepared by adding 50 g of NaCl to each liter of surface seawater. This brine solution is pressure filtered sequentially through a 1.0 and 0.5 μm filter cartridge prior to the addition of 10 ml 100% formalin l-1. Individual traps are filled and at least 10 l of the trap solution is saved for analysis of solution blanks (see section 4.1).  Tubes for PIC and POC were not fixed i.e. did not receive an addition of formalin.  Tubes for POC used trace metal clean NaCl. 

3.3. Post-recovery processing

3.3.1.

Upon recovery, individual traps are capped and transported to the shipboard portable laboratory for analysis. Care is taken not to mix the higher density trap solutions with the overlying seawater. Trap samples are processed from deep to shallow in order to minimize potential contamination.

3.3.2.

The depth of the interface between the high density solution and overlying seawater is marked on each trap. The overlying seawater is then aspirated with a plastic tube attached to a vacuum system in order to avoid disturbing the high density solution. Because some sinking particulate material collects near the interface between the high density solution and the overlying seawater, the overlying seawater is removed only to a depth that is 5 cm above the previously identified interface.

3.3.3.

After the overlying seawater has been removed from all the traps at a given depth, the contents of each trap is passed through an acid rinsed 335 μm NitexR screen to remove contaminating zooplankton and micronekton which entered the traps in a living state and are not truly part of the passive flux. Immediately before this sieving process, the contents of each trap are mixed to disrupt large amorphous particles. The traps are rinsed with a portion of the <335 μm sample in order to recover all particulate matter, and the 335 μm NitexR screen is examined to determine whether residual material, in addition to the so-called "swimmers", is present. If so, the screens are rinsed again with a portion of the 335 μm filtrate. After all traps from a given depth have been processed, the 335 μm screen is removed and placed into a vial containing 20 ml of formalin- seawater solution, and stored at 4 °C for subsequent microscopic examination and organism identification and enumeration.

4. Data

4.1. Determination of PP flux

Followed Particulate Phosphorus HOT methods: https://hahana.soest.hawaii.edu/hot/protocols/protocols.html# Briefly, triplicate samples from each depth were combusted in 16 x 100 mm clean glass test tubes at 450°C for 4.5 hours in a muffle furnace. The samples are then allowed to cool and immersed in 10ml 0.15N HCL/vortex/ 1 hour leach/vortex/30 Minute Spin/ 1 hour color development) (5ml sample : 500ul Mixed Reagent).The liberated orthophosphate is reacted with a mixed reagent of molybdic acid, ascorbic acid and trivalent antimony to form phosphomolybdic acid. This heteropoly acid is then reduced to the colored molybdenum blue complex by ascorbic acid and the solution is measured spectrophotometrically. This procedure measures all forms of phosphorus which can be released by combustion and acid hydrolysis.

4.2. Determination of PC, POC, PIC and PN flux

The quantities of particulate PC, POC, and PN in the prescreened trap solutions are determined from three replicate traps. The samples analyzed in the Princeton lab were not acidified and represent total C, the samples analyzed at WHOI were acid-fumed to represent organic C. An equivalent volume of the time-zero sediment trap solution, filtered through the appropriate filters is used as a C or N blank.

4.3. Determination of C and N bulk isotope composition

The bulk isotope composition of total C and N sinking material in the prescreened trap solutions are determined from three replicate traps. 

4.4 Determination of PIC flux

 A single sample was analyzed for PIC at WHOI. PIC samples were analyzed on a Picarro 2101i cavity ring-down CO2 isotope analyzer system with an AutoMate prep device front end. Detailed instrument methods are established and published in Dong et al., 2019 and Subhas et al., 2019. Samples were acidified manually in 12mL exetainers and left to sit for 1h to dissolve all PIC before mounting in the autosampler rack.  Standards were run before, during, and after the samples in the same analytical session to calculate PIC concentration, using a well-characterized pure calcite standard of known isotopic composition (Iceland Spar).  Calibrated PIC quantities (in micromoles) were blank-corrected using the mean value of McLane pump dipped blank filters.  Blank-corrected quantities were then normalized to the volume of brine in each tube, length of deployment, and area of each tube, to calculate a flux in mmol/m2/d. 

Problem report:

No PIC fluxes were measured on Trap Deployment 1 because all tubes were allocated to other analyses. PP, PC, PN, d15N, d13C were not measured on Trap Deployment 2 because all tubes were allocated to other analyses.