Oxygen dynamics in the North Atlantic subtropical gyre

Dissolved oxygen (DO) in the ocean is a tracer for most ocean biogeochemical processes including net community production and remineralization of organic matter which in turn constrains the biological carbon pump. Knowledge of oxygen dynamics in the North Atlantic Ocean is mainly derived from observations at the Bermuda Atlantic Time-series Study (BATS) site located in the western subtropical gyre which may skew our view of the biogeochemistry of the subtropical North Atlantic. This study presents and compares a 15 yr record of DO observations from ESTOC (European Station for Time-Series in the Ocean, Canary Islands) in the eastern subtropical North Atlantic with the 20 yr record at BATS. Our estimate for net community production of oxygen was 2.3±0.4 mol O₂ m⁻² yr⁻¹ and of oxygen consumption was −2.3±0.5 mol O₂ m⁻² yr⁻¹ at ESTOC, and 4 mol O₂ m⁻² yr⁻¹ and −4.4±1 mol m⁻² yr⁻¹ at BATS, respectively. These values were determined by analyzing the time-series using the Discrete Wavelet Transform (DWT) method. These flux values agree with similar estimates from in-situ observational studies but are higher than those from modeling studies. The difference in net oxygen production rates supports previous observations of a lower carbon export in the eastern compared to the western subtropical Atlantic. The inter-annual analysis showed clear annual cycles at BATS whereas longer cycles of nearly 4 years were apparent at ESTOC. The DWT analysis showed trends in DO anomalies dominated by long-term perturbations at a basin scale for the consumption zones at both sites, whereas yearly cycles dominated the production zone at BATS. The long-term perturbations found are likely associated with ventilation of the main thermocline, affecting the consumption and production zones at ESTOC.     

Seasonal POC fluxes at BATS estimated from 210Po deficits

In this study at the Bermuda Atlantic Time-series Study (BATS) site we demonstrate that the polonium–lead disequilibrium system may perform better as a tracer of organic carbon export under low-flux conditions (in this case, <2.5 mmol C m^?2 d^?1) than under bloom conditions in an oligotrophic setting. With very few exceptions, the POC flux predictions calculated from the water-column ²¹⁰Po deficit were within a factor of 2 of the POC flux caught in surface-tethered sediment traps. However, we found higher correlation between size-fractionated particulate ²¹⁰Po activity and POC concentration in November 2006 (r=0.93) than in January (r=0.79) and during the spring bloom in March 2007 (r=0.80). We suggest that this is due to the ability of polonium to distinguish between bulk mass flux and organic carbon export under oligotrophic and lithogenic-driven flux regimes. Further, we found that the POC/Po ratio on particles was largely independent of size class between 10 and 100 μm (P=0.13) during each season, supporting the notion that export in this oligotrophic system is driven by sinking aggregates of smaller cells and not by large, individual cells.     

A comparison of the quantity and composition of material caught in a neutrally buoyant versus surface-tethered sediment trap

The flux and composition of material caught using two different upper ocean sediment trap designs was compared at the Bermuda Atlantic Time-series Study site (BATS). The standard surface-tethered trap array at BATS was compared to a newly designed neutrally buoyant sediment trap (NBST). Both traps used identical cylindrical collection tubes. Of particular concern was the effect of horizontal flow on trap collection efficiency. In one experiment, mass, particulate organic carbon (POC) and particulate organic nitrogen (PON) fluxes were slightly lower (20–30%) in the NBST than in the standard BATS trap. In contrast, ²³⁴Th and fecal pellet fluxes were up to a factor of two to three lower in the NBST. In a second experiment, mass and POC fluxes decreased significantly with depth in the BATS surface-tethered trap, but not in the NBST. Different brine treatments had no measurable effect on collection efficiencies. A striking observation was that the swimmer “flux” was much larger in the standard BATS traps than in the NBST. Overall, these results show that different components of the sinking flux can be collected with differing efficiencies, depending upon how traps are deployed in the ambient environment.     

Settling fluxes in Saanich and Jervis Inlets, British Columbia, Canada: sources and seasonal patterns

We conducted a multi-year sediment-trap experiment in Saanich and Jervis Inlets, British Columbia, Canada. Moorings with traps positioned at three depths were placed near the mouth and toward the head of each fjord, and deployments were monthly. We present fluxes of total mass, biogenic silica (BSi), particulate organic carbon (POC) and aluminium (Al), as well as the δ¹³C signal of the POC, and we compare the sediment-trap fluxes to primary-production measurements made during the experiment.Diatomaceous silica and aluminosilicates were the primary components of the settling flux, while organic matter from marine (largely diatoms) and terrestrial sources was occasionally a significant portion of the sinking material. Fluxes of BSi and POC were highest in the spring and summer, tracing maxima in local primary production. These fluxes decreased, increased or remained constant with depth due to water-column remineralisation and variability in processes that cause fluxes to increase with depth. Al fluxes followed local precipitation and river runoff at the landward stations, and with remarkable faithfulness in Saanich Inlet. Near the mouths, there was little seasonality in Al flux, and the increases of flux with depth reveal sedimentary plumes at each fjord’s sill. Tidal and deepwater-renewal components of the plumes are evident, and the plume in Saanich Inlet was particularly intense. Fluxes of Al to deep sediment traps associated with renewal flows were also observed toward the head of each fjord.Marine δ¹³C endmembers are estimated from relationships between δ¹³C and BSi concentrations, and measures of soil δ¹³C from each fjord were available. These endmembers are used with the δ¹³C record to quantify marine and terrigenous contributions to the POC flux. Marine POC composed 54-72% of the total POC caught by shallow sediment traps in spring and summer, and 36-54% in fall and winter. Primary production and sediment-trap fluxes are used to estimate annually averaged export ratios (shallow-trap flux:autotrophic assimilation) for marine POC and Si. POC export ratios (0.092-0.14) were low for these productive waters, but they compare with other results based on sediment-trap fluxes from coastal waters where terrigenous OC has been subtracted. Export ratios of Si were calculated using an estimated Si:C assimilation ratio and, therefore, are susceptible to error, but the high results (>0.8) suggest that BSi was exported more efficiently than POC. The possibility that POC was preferentially lost after interception by sediment traps is also considered. Primary production and settling fluxes were higher in Saanich Inlet than in Jervis Inlet, while export ratios of OC and Si were similar in both fjords, away from the nepheloid layer near the sill of Saanich Inlet.     

Seasonal and spatial variability in plankton production and respiration in the Subtropical Gyres of the Atlantic Ocean

Euphotic zone plankton production (P) and respiration (R) were determined from the in vitro flux of dissolved oxygen during six latitudinal transects of the Atlantic Ocean, as part of the Atlantic Meridional Transect (AMT) programme. The transects traversed the North and South Atlantic Subtropical Gyres (N gyre, 18–38°N; S gyre, 11–35°S) in April–June and September–November 2003–2005. The route and timing of the cruises enabled the assessment of the seasonal variability of P, R and P/R in the N and S gyres, and the comparison of the previously unsampled N gyre centre with the more frequently sampled eastern edge of the gyre. Mean euphotic zone integrated rates (±SE) were P=63±23 (n=31), R=69±22 (n=30) mmol O₂ m⁻² d⁻¹ in the N gyre; and P=58±26 (n=30), R=62±24 (n=30) mmol O₂ m⁻² d⁻¹ in the S gyre. Overall, the N gyre was heterotrophic (R>P) and it was more heterotrophic than the S gyre, but the metabolic balance of both gyres changed with season. Both gyres were net heterotrophic in autumn, and balanced in spring. This seasonal contrast was most pronounced for the S gyre, because it was more autotrophic than the N gyre during spring. This may have arisen from differences in nitrate availability, because spring sampling in the S gyre coincided with periods of deep mixing to the nitracline, more frequently than spring sampling within the N gyre. Our results indicate that the N gyre is less heterotrophic than previous estimates suggested, and that there is an apparent decrease in R from the eastern edge to the centre of the N gyre, possibly indicative of an allochthonous organic carbon source to the east of the gyre.     

Particle fluxes associated with mesoscale eddies in the Sargasso Sea

We examined the impact of a cyclonic eddy and mode-water eddy on particle flux in the Sargasso Sea. The primary method used to quantify flux was based on measurements of the natural radionuclide, ²³⁴Th, and these flux estimates were compared to results from sediment traps in both eddies, and a ²¹⁰Po/²¹⁰Pb flux method in the mode-water eddy. Particulate organic carbon (POC) fluxes at 150 m ranged 1–4 mmol C m⁻² d⁻¹ and were comparable between methods, especially considering differences in integration times scales of each approach. Our main conclusion is that relative to summer mean conditions at the Bermuda Atlantic Time-series Study (BATS) site, eddy-driven changes in biogeochemistry did not enhance local POC fluxes during this later, more mature stage of the eddy life cycle (>6 months old). The absence of an enhancement in POC flux puts a constraint on the timing of higher POC flux events, which are thought to have caused the local O₂ minima below each eddy, and must have taken place >2 months prior to our arrival. The mode-water eddy did enhance preferentially diatom biomass in its center, where we estimated a factor of three times higher biogenic Si flux than the BATS summer average. An unexpected finding in the highly depth-resolved ²³⁴Th data sets is narrow layers of particle export and remineralization within the eddy. In particular, a strong excess ²³⁴Th signal is seen below the deep chlorophyll maxima, which we attribute to remineralization of ²³⁴Th-bearing particles. At this depth below the euphotic zone, de novo particle production in the euphotic zone has stopped, yet particle remineralization continues via consumption of labile sinking material by bacteria and/or zooplankton. These data suggest that further study of processes in ocean layers is warranted not only within, but below the euphotic zone.     

Lunar cycles and seasonal variations in deposition fluxes of planktic foraminiferal shell carbonate to the deep South Atlantic (central Walvis Ridge)

Several authors have argued that lunar reproductive cycling controls the shell fluxes of planktic foraminifera, one of the major carbonate-producing groups in the global pelagic ocean. A time-series sediment trap at 2700 m depth on the central Walvis Ridge below the South Atlantic central gyre demonstrate for the first time that shell deposition fluxes of Hastigerina pelagica are synchronous with lunar periodicity. Spectral analysis of the 6-month time-series with 8-day resolution showed a strong 30-day cyclicity in the flux maxima of H. pelagica arriving at the ocean floor on average 12.5 days after each full moon. Given a shell settling velocity of about 400 m day⁻¹, which implies about 7 days for settling, this coincides with the pronounced endogenous reproduction rhythm of 5±2 days after full moon as originally observed in laboratory-cultured isolates from off Bermuda in the North Atlantic. By contrast, no endogenous or exogenous lunar periodicity was observed in the deposition flux or size distribution of any of the 27 other shell species from austral winter (August 2000) to austral summer (February 2001). Instead, the deposition fluxes of shell species, the bulk carbonate and the total mass were dominated by a seasonal maximum during austral spring, without any periodicity in the 16–90-day domain of this study. Since H. pelagica exhibits low fluxes with a low burial efficiency, and continuous (re)production is shown by the deposition fluxes of other species, lunar reproductive cycling appears not to affect pelagic carbonate productivity and deep ocean sedimentation fluxes.     

Seasonal and interannual variability in particle fluxes of carbon,nitrogen and silicon from time series of sediment traps at Ocean Station P, 1982–1993: relationship to changes in subarctic primary productivity

An extended time series of particle fluxes at 3800 m was recorded using automated sediment traps moored at Ocean Station Papa (OSP, 50°N, 145°W) in the northeast Pacific Ocean for more than a decade (1982–1993). Time-series observations at 200 and 1000 m, and short-term measurements using surface-tethered free-drifting sediment traps also were made intermittently. We present data for fluxes of total mass (dry weight), particulate organic carbon (POC), particulate organic nitrogen (PON), biogenic Si (BSi), and particulate inorganic carbon (PIC) in calcium carbonate. Mean monthly fluxes at 3800 m showed distinct seasonality with an annual minimum during winter months (December–March), and maximum during summer and fall (April–November). Fluxes of total mass, POC, PIC and BSi showed 4-, 10-, 7- and 5-fold increases between extreme months, respectively. Mean monthly fluxes of PIC often showed two plateaus, one in May–August dominated by <63 μm particles and one in October–November, which was mainly >63 μm particles. Dominant components of the mass flux throughout the year were CaCO₃ and opal in equal amounts. The mean annual fluxes at 3800 m were 32±9 g dry weight g m⁻² yr⁻¹, 1.1±0.5 g POC m⁻² yr⁻¹, 0.15±0.07 g PON m⁻² yr⁻¹, 5.9±2.0 g BSi m⁻² yr⁻¹ and 1.7±0.6 g PIC m⁻² yr⁻¹. These biogenic fluxes clearly decreased with depth, and increased during “warm” years (1983 and 1987) of the El Niño, Southern Oscillation cycle (ENSO). Enhancement of annual mass flux rates to 3800 m was 49% in 1983 and 36% in 1987 above the decadal average, and was especially rich in biogenic Si. Biological events allowed estimates of sinking rates of detritus that range from 175 to 300 m d⁻¹, and demonstrate that, during periods of high productivity, particles sink quickly to deep ocean with less loss of organic components. Average POC flux into the deep ocean approximated the “canonical” 1% of the surface primary production.     

Nutrient gradients in the western North Atlantic Ocean: Relationship to microbial community structure and comparison to patterns in the Pacific Ocean

Studies of nitrogen and phosphorus dynamics in the oligotrophic surface waters of the western North Atlantic Ocean have been constrained because ambient concentrations are typically at or below the detection limits of standard colorometric methods, except during periods of deep vertical mixing. Here we report the application of high-sensitivity analytical methods—determinations of nitrate plus nitrite (N+N) by chemiluminescence and soluble reactive phosphorus (SRP) by the magnesium induced co-precipitation (MAGIC) protocol—to surface waters along a transect from the Sargasso Sea at 26°N through the Gulf Stream at 37°N, including sampling at the JGOFS Bermuda Atlantic Time-series Study (BATS) station. The results were compared with data from the BATS program, and the HOT station in the Pacific Ocean, permitting cross-ecosystem comparisons. Microbial populations were analyzed along the transect, and an attempt was made to interpret their distributions in the context of the measured nutrient concentrations.Surface concentrations of N+N and SRP during the March 1998 transect separated into 3 distinct regions, with the boundaries corresponding roughly to the locations of the BATS station (∼31°N) and the Gulf Stream (∼37°N). Although N+N and SRP co-varied, the [N+N] : [SRP] molar ratios increased systematically from ∼1 to 10 in the southern segment, remained relatively constant at ∼40–50 between 31°N and 37°N, then decreased again systematically to ratios <10 north of the Gulf Stream. Dissolved organic N (DON) and P (DOP) dominated (⩾90%) the total dissolved N (TDN) and P (TDP) pools except in the northern portion of the transect. The [DON] : [DOP] molar ratios were relatively invariant (∼30–60) across the entire transect.Heterotrophic prokaryotes (operationally defined as “bacteria”), Prochlorococcus, Synechococcus, ultra- and nanophytoplankton, cryptophytes, and coccolithophores were enumerated by flow cytometry. The abundance of bacteria was well correlated with the concentration of SRP, and that of the ultra- and nanophytoplankton was well correlated with the concentration of N+N. The only group whose concentration was correlated with temperature was Prochlorococcus, and its abundance was unrelated to the concentrations of nutrients measured at the surface.We combined our transect results with time-series measurements from the BATS site and data from select depth profiles, and contrasted these North Atlantic data sets with time-series of N and P nutrient measurements from a station in the North Pacific subtropical gyre near Hawaii [Hawaii Ocean Time-series (HOT) site]. Two prominent differences are readily observed from this comparison. The [N+N] : [SRP] molar ratios are much less than 16 : 1 during stratified periods in surface waters at the BATS site, as is the case at the HOT site year round. However, following deep winter mixing, this ratio is much higher than 16 : 1 at BATS. Also, SRP concentrations in the upper 100 m at BATS fall in the range 1–10 nM during stratified periods, which is at least one order of magnitude lower than at the HOT site. That two ecosystems with comparable rates of primary and export production would differ so dramatically in their nutrient dynamics is intriguing, and highlights the need for detailed cross ecosystem comparisons.     

Particle flux characterisation and sedimentation patterns of protistan plankton during the iron fertilisation experiment LOHAFEX in the Southern Ocean

The taxonomic composition and types of particles comprising the downward particle flux were examined during the mesoscale artificial iron fertilisation experiment LOHAFEX. The experiment was conducted in low-silicate waters of the Atlantic Sector of the Southern Ocean during austral summer (January–March 2009), and induced a bloom dominated by small flagellates. Downward particle flux was low throughout the experiment, and not enhanced by addition of iron; neutrally buoyant sediment traps contained mostly faecal pellets and faecal material apparently reprocessed by mesozooplankton. TEP fluxes were low, ≤5 mg GX eq. m⁻² d⁻¹, and a few phytodetrital aggregates were found in the sediment traps. Only a few per cent of the POC flux was found in the traps consisting of intact protist plankton, although remains of taxa with hard body parts (diatoms, tintinnids, thecate dinoflagellates and foraminifera) were numerous, far more so than intact specimens of these taxa. Nevertheless, many small flagellates and coccoid cells, belonging to the pico- and nanoplankton, were found in the traps, and these small, soft-bodied cells probably contributed the majority of downward POC flux via mesozooplankton grazing and faecal pellet export. TEP likely played an important role by aggregating these small cells, and making them more readily available to mesozooplankton grazers.     

Biological organic carbon export estimated from the annual carbon budget observed in the surface waters of the western subarctic and subtropical North Pacific Ocean from 2004 to 2013

The annual flux of biologically produced organic carbon from surface waters is equivalent to annual net community production (NCP) at a steady state and equals the export of particulate and dissolved organic carbon (POC and DOC, respectively) to the ocean interior. NCP was estimated from carbon budgets of salinity-normalized dissolved inorganic carbon (nDIC) inventories at two time-series stations in the western subarctic (K2) and subtropical (S1) North Pacific Ocean. By using quasi-monthly biogeochemical observations from 2004 to 2013, monthly mean nDIC inventories were integrated from the surface to the annual maximum mixed layer depth and corrected for changes due to net air–sea CO₂ exchange, net CaCO₃ production, vertical diffusion from the upper thermocline, and horizontal advection. The annual organic carbon flux at K2 (1.49 ± 0.42 mol m^?2 year^?1) was lower than S1 (2.81 ± 0.53 mol m^?2 year^?1) (p < 0.001 based on t test). These fluxes consist of three components: vertically exported POC fluxes (K2: 1.43 mol m^?2 year^?1; S1: 2.49 mol m^?2 year^?1), vertical diffusive DOC fluxes (K2: 0.03 mol m^?2 year^?1; S1: 0.25 mol m^?2 year^?1), and suspended POC fluxes (K2: 0.03 mol m^?2 year^?1; S1: 0.07 mol m^?2 year^?1). The estimated POC export flux at K2 was comparable to the sum of the POC flux observed with drifting sediment traps and active carbon flux exported by migrating zooplankton. The export fluxes at both stations were higher than those reported at other time-series sites (ALOHA, the Bermuda Atlantic Time-series Study, and Ocean Station Papa).     

Zooplankton vertical migration and the active transport of dissolved organic and inorganic carbon in the Sargasso Sea

The least known component of the “biological pump” is the active transport of carbon and nutrients by diel vertical migration of zooplankton. We measured CO₂ respiration and dissolved organic carbon (DOC) excretion by individual species of common vertically migrating zooplankton at the US JGOFS Bermuda Atlantic Time-series Study (BATS) station. The inclusion of DOC excretion in this study builds on published research on active transport by respiration of inorganic carbon and allows a direct assessment of the role of zooplankton in the production of dissolved organic matter used in midwater microbial processes. On average, excretion of DOC makes up 24% (range=5–42%) of the total C metabolized (excreted+respired) and could represent a significant augmentation to the vertical flux that has already been documented for respiratory CO₂ flux by migrant zooplankton. Migratory fluxes were compared to other transport processes at BATS. Estimates of combined active transport of CO₂ and DOC by migrators at BATS averaged 7.8% and reached 38.6% of mean sinking POC flux at 150 m, and reached 71.4% of mean sinking POC flux at 300 m. DOC export by migrator excretion averaged 1.9% and reached 13.3% of annual DOC export by physical mixing at this site. During most of the year when deep mixing does not occur, diel migration by zooplankton could provide a supply of DOC to the deeper layers that is available for use by the microbial community. A carbon budget comparing migrant zooplankton transport to the balance of fluxes in the 300–600 m depth strata at BATS shows on average that the total migrant flux supplies 37% of the organic carbon remineralized in this layer, and that migrant DOC flux is more than 3 times the DOC flux gradient by diapycnal mixing. New estimates of active transport of both organic and inorganic carbon by migrants may help resolve observed imbalances in the C budget at BATS, but the magnitude is highly dependent on the biomass of the migrating community.     

Simulation of export production and biological pump structure in the South China Sea

The export flux of particulate organic carbon (POC) consumes upwelled dissolved inorganic carbon (DIC), which hinders surplus CO₂ being released to the atmosphere. The export flux of POC is therefore crucial to the carbon and biogeochemical cycles. This study aims to model the long-term (1958–2009) variation of export flux and structure of the biological pump in the South China Sea (SCS) using a three-dimensional physical-biogeochemical coupled (ROMS-CoSiNE) model. The modeled POC export flux in the northeastern and north central SCS is high in winter and low in summer, whereas the flux in the central, southwestern and southern SCS varies following a “W” shape: two maxima in winter and summer, and two minima in spring and autumn. The pattern follows the variation of the East Asian monsoon and is consistent with observations. On the interannual scale, export flux is anti-phased with the El Niño-Southern Oscillation such that El Niño (La Niña) conditions correspond to low (high) export flux. Modeled annual mean POC export flux reaches up to 1.95 mmol m^–2 day^–1, which is underestimated comparing with field observations. The f-ratio is estimated to be ~0.4. The b value of the Martin equation for POC is 1.18±0.03. Remineralization rate of POC is greater than the classical Martin equation but is consistent with its subtropical counterparts. The modeled results indicate that the SCS is a weak source of atmospheric CO₂ with a flux estimated at 1.0 mmol m^–2 day^–1. The modeled results provide an insight of the temporal and spatial variability of the carbon cycle in this monsoon-driven, semi-enclosed basin.     

Enhancement of particulate organic carbon export flux induced by atmospheric forcing in the subtropical oligotrophic northwest Pacific Ocean

The effects of extreme atmospheric forcing on the export flux of particulate organic carbon (POC) in the warm oligotrophic nitrogen-limited northwest Pacific Ocean were examined in 2007 during the spring Asian dust storm period. Several strong northeast monsoon events (maximum sustained wind speeds approaching 16.7 m s^? 1, and gusts up to 19.0 m s^? 1) accompanied by dust storms occurred during a 1-month period. The cold stormy events decreased surface water temperature and induced strong wind-driven vertical mixing of the water column, resulting in nutrient entrainment into the mixed layer from subsurface waters. As a result, the export flux of POC ranged from 49 to 98 (average value = 71 ± 16) mg m^? 2 day^? 1, approximately 2–3 times greater than average values in other seasons. As dry and wet deposition of nitrogen attributable to Asian dust storm events does not account for the associated increases in POC stocks in this N-limited oligotrophic oceanic region, the enhancement of POC flux must have been caused by nutrient entrainment from subsurface waters because of the high winds accompanying the dust storm events.     

Export fluxes of particulate organic carbon in the Chukchi Sea: A comparative study using 234Th/238U disequilibria and drifting sediment traps

Large-volume sampling of ²³⁴Th and drifting sediment trap deployments were conducted as part of the 2004 Western Arctic Shelf–Basin Interactions (SBI) spring (May 15–June 23) and summer (July 17–August 26) process cruises in the Chukchi Sea. Measurements of ²³⁴Th and particulate organic carbon (POC) export fluxes were obtained at five stations during the spring cruise and four stations during the summer cruise along Barrow Canyon (BC) and along a parallel shelf-to-basin transect from East Hanna Shoal (EHS) to the Canada Basin. ²³⁴Th and POC fluxes obtained with in situ pumps and drifting sediment traps agreed to within a factor of 2 for 70% of the measurements. POC export fluxes measured with in situ pumps at 50 m along BC were similar in spring and summer (average = 14.0 ± 8.0 mmol C m^− 2 day^− 1 and 16.5 ± 6.5 mmol C m^− 2 day^− 1, respectively), but increased from spring to summer at the EHS transect (average = 1.9 ± 1.1 mmol C m^− 2 day^− 1 and 19.5 ± 3.3 mmol C m^− 2 day^− 1, respectively). POC fluxes measured with sediment traps at 50 m along BC were also similar in both seasons (31.3 ± 9.3 mmol C m^− 2 day^− 1 and 29.1 ± 14.2 mmol C m^− 2 day^− 1, respectively), but were approximately twice as high as POC fluxes measured with in situ pumps. Sediment trap POC fluxes measured along the EHS transect also increased from spring to summer (3.0 ± 1.9 mmol C m^− 2 day^− 1 and 13.0 ± 6.4 mmol C m^− 2 day^− 1, respectively), and these fluxes were similar to the POC fluxes obtained with in situ pumps. Discrepancies in POC export fluxes measured using in situ pumps and sediment traps may be reasonably explained by differences in the estimated POC/²³⁴Th ratios that arise from differences between the techniques, such as time-scale of measurement and size and composition of the collected particles. Despite this variability, in situ pump and sediment trap-derived POC fluxes were only significantly different at a highly productive station in BC during the spring.     

High-frequency fluxes of labile compounds in the central Ligurian Sea,northwestern Mediterranean

Sinking particles were collected every 4 h with drifting sediment traps deployed at 200 m depth in May 1995 in a 1-D vertical system during the DYNAPROC observations in the northwestern Mediterranean sea. POC, proteins, glucosamine and lipid classes were used as indicators of the intensity and quality of the particle flux. The roles of day/night cycle and wind on the particle flux were examined. The transient regime of production from late spring bloom to pre-oligotrophy determined the flux intensity and quality. POC fluxes decreased from, on average, 34 to 11 mg m⁻² d⁻¹, representing 6–14% of the primary production under late spring bloom conditions to 1–2% under pre-oligotrophic conditions. Total protein and chloroplast lipid fluxes correlated with POC and reflected the input of algal biomass into the traps. As the season proceeded, changes in the biochemical composition of the exported material were observed. The C/N ratio rose from 7.8 to 12. Increases of serine (10–28% of total proteins), total lipids (7–9 to 14–28% of POC) and reserve lipids (1–5 to 5–22% of total lipids) were noticeable, whereas total protein content in POC decreased (20–27 to 18–7%). N-acetyl glucosamine, a tracer of fecal pellet flux, showed that zooplankton grazing was a major vector of downward export during the decaying bloom. Against this background pattern, episodic events specifically increased the flux, modifying the quality and the settling velocity of particles. Day/night signals in biotracers (POC, N-acetyl glucosamine, protein and chloroplast lipids) showed that zooplankton migrations were responsible for sedimentation of fresh material through fast sinking particles (V=170–180 m d⁻¹) at night. Periodic signatures of re-processed material (high lipolysis and bacterial biomass indices) suggested that other zooplankton fecal pellets or small aggregates, probably of lower settling velocities (V<170 m d⁻¹), contributed to the flux during calm periods. At the beginning of the experiment, during the development of a prymnesiophyte bloom in the upper layers, the sterol signal with no periodicity enabled us to estimate high particle settling velocities (⩾600 m d⁻¹) likely related to large aggregate formation. A wind event increased biotracer fluxes (POC, protein, chloroplast lipids). The rapid transmission of surface signals through extremely fast sinking particles could be a general feature of particle fluxes in marine areas unaffected by horizontal advection.     

Downward particle fluxes within different productivity regimes off the Mauritanian upwelling zone (EUMELI program)

A 2-yr record of downward particle flux was obtained with moored sediment traps at several depths of the water column in two regions characterized by different primary production levels (mesotrophic and oligotrophic) of the eastern subtropical North Atlantic Ocean. Particle fluxes, of ∼71–78% biogenic origin (i.e. consisting of CaCO₃, organic matter and opal) on average, decrease about six-fold from the mesotrophic site (highest fluxes in the North Atlantic) nearer the Mauritanian margin (18°30′N, 21°00′W) to the remote, open-ocean, oligotrophic site (21°00′N, 31°00′W). This decrease largely reflects the difference in total primary production between the two sites, from ∼260 to ∼110 g organic C m⁻² yr⁻¹. At both sites, temporal variability of the downward particle flux seems to be linked to westward surface currents, which are likely to transport seaward biomass-rich water masses from regions nearer the coast. The influence of coastal upwelling is marked at the mesotrophic site. The large differences between the 1991 and 1992 records at that site, where carbon export is large, underscore the interest of long-term studies for export budget estimates. The different productivity regimes at the two sites seem to induce contrasting downward modes of transport of the particulate matter, as shown in particular by the faster settling rates and the higher E ratio (particulate organic carbon export versus total primary production) estimated at the mesotrophic site.     

Primary production export flux in Marguerite Bay (Antarctic Peninsula): Linking upper water-column production to sediment trap flux

A study was carried out to assess primary production and associated export flux in the coastal waters of the western Antarctic Peninsula at an oceanographic time-series site. New, i.e., exportable, primary production in the upper water-column was estimated in two ways; by nutrient deficit measurements, and by primary production rate measurements using separate ¹⁴C-labelled radioisotope and ¹⁵N-labelled stable isotope uptake incubations. The resulting average annual exportable primary production estimates at the time-series site from nutrient deficit and primary production rates were 13 and 16 mol C m⁻², respectively. Regenerated primary production was measured using ¹⁵N-labelled ammonium and urea uptake, and was low throughout the sampling period.The exportable primary production measurements were compared with sediment trap flux measurements from 2 locations; the time-series site and at a site 40 km away in deeper water. Results showed ∼1% of the upper mixed layer exportable primary production was exported to traps at 200 m depth at the time-series site (total water column depth 520 m). The maximum particle flux rate to sediment traps at the deeper offshore site (total water column depth 820 m) was lower than the flux at the coastal time-series site. Flux of particulate organic carbon was similar throughout the spring–summer high flux period for both sites. Remineralisation of particulate organic matter predominantly occurred in the upper water-column (<200 m depth), with minimal remineralisation below 200 m, at both sites. This highly productive region on the Western Antarctic Peninsula is therefore best characterised as ‘high recycling, low export’.     

Spatial and temporal variability of POC in the northeast Subarctic Pacific

Particulate organic carbon (POC) concentrations from 0 to 1000 m were quantified in size-fractionated particulate matter samples obtained by the multiple unit large volume in situ filtration system (MULVFS) in 1996 and 1997 along the 1600 km long “line P” transect from continental slope waters near southern Vancouver Island to Ocean Station PAPA (OSP, 50°N, 145°W). Regression of in situ POC vs. beam attenuation coefficient, c, from a simultaneously deployed 1-m pathlength SeaTech transmissometer gave slope, intercept and r² values of 6.15±0.19×10⁻⁵ m⁻¹ (nmol C l⁻¹)⁻¹, 0.363±0.003 m⁻¹, and 0.951 (n=145), respectively. This result agreed within several percent of calibrations obtained from two 2600-km-long transects of the equatorial Pacific in 1992 (Bishop, 1999). Data from other, more frequently deployed transmissometers were standardized against the 1-m instrument, and the combined optical data set was used to document POC variability at finer spatial and temporal scales than could be sampled directly using either conventional water bottle casts or MULVFS. Published bottle POC vs. c relationships show much more variability and remain problematic. Along the line P transect in the salinity-stratified upper 100 m, POC isolines shoaled from winter to summer in concert with seasonal stratification. At the same time, POC was progressively enriched in subeuphotic zone waters to depths greater than 500 m. Near-surface POC fields sampled in the winter time showed strong temporal POC variability over time scales of days as well as between years. POC concentrations at OSP in February 1996 were higher than those found at any other time of year. Less variability was found along line P in other seasons. In May 1996, kilometer-scale spatial variability of POC at OSP was small; dawn vs. dusk variations of c were used to calculate 0–100 m POC turnover times shorter than 6 d. Calculations also suggest that 25–50% of primary productivity was expressed as dissolved organic carbon at OSP in May 1996.     

Carbon and biogenic silica export influenced by the Amazon River Plume: Patterns of remineralization in deep-sea sediments

The Amazon River Plume delivers freshwater and nutrients to an otherwise oligotrophic western tropical North Atlantic (WTNA) Ocean. Plume waters create conditions favorable for carbon and nitrogen fixation, and blooms of diatoms and their diazotrophic cyanobacterial symbionts have been credited with significant CO₂ uptake from the atmosphere. The fate of the carbon, however, has been measured previously by just a few moored or drifting sediment traps, allowing only speculation about the full extent of the plume's impact on carbon flux to the deep sea. Here, we used surface (0.5 m) sediment cores collected throughout the Demerara Slope and Abyssal Plain, at depths ranging from 1800 to 5000 m, to document benthic diagenetic processes indicative of carbon flux. Pore waters were extracted from sediments using both mm- and cm-scale extraction techniques. Profiles of nitrate (NO₃) and silicate (Si(OH)₄) were modeled with a diffusion-reaction equation to determine particulate organic carbon (POC) degradation and biogenic silica (bSi) remineralization rates. Model output was used to determine the spatial patterns of POC and bSi arrival at the sea floor. Our estimates of POC and Si remineralization fluxes ranged from 0.16 to 1.92 and 0.14 to 1.35 mmol m⁻² d⁻¹, respectively. A distinct axis of POC and bSi deposition on the deep sea floor aligned with the NW axis of the plume during peak springtime flood. POC flux showed a gradient along this axis with highest fluxes closest to the river mouth. bSi had a more diffuse zone of deposition and remineralization. The impact of the Amazon plume on benthic fluxes can be detected northward to 10°N and eastward to 47°W, indicating a footprint of nearly 1 million km². We estimate that 0.15 Tmol C y⁻¹ is remineralized in abyssal sediments underlying waters influenced by the Amazon River. This constitutes a relatively high fraction (~7%) of the estimated C export from the region.; the plume thus has a demonstrable impact on C_org export in the western Atlantic. Benthic fluxes under the plume were comparable to and in some cases greater than those observed in the eastern equatorial Atlantic, the southeastern Atlantic, and the Southern Ocean.