Cycling of dissolved and particulate organic matter in the open ocean

,

For the A•4C analyses of the DOCu•, two 1-gal.samples were quickly thawed in warm water and 5 L introduced into a glass and quartz reactor, acidified to pH 2-3 with H3PO4, freed of DIC with N• and then saturated with oxygen.The DOC was oxidized with 1200 W UV radiation to CO• which was collected as described previously [Druffel et al., 1989a].
Blank values were _•1.5 ttM C. The dried humic materials extracted using XAD resins were combusted to CO• in double quartz tubes with CuO and silver at 850øC for at least 1 hour [Sorer, 1980].For •4C analysis of the OOCFt-htc, 100-mL seawater samples were acidified to pH 2-3 with 85% H3PO•, freed of DIC with O•, and oxidized to CO• using a modified Dohrmann Model DC-90 total carbon analyzer [Bauer et al., 1992a] The strontium chloride/ammonium hydroxide solution was decanted from each DIC sample, then heated to dryness, leaving solid strontium carbonate.CO• was liberated using 4 N H C1 and converted to acetylene gas according to Griffin and Druffel [1985].The acetylene samples were counted for 5 to 6 2-day periods in quartz gas proportional beta counters at 90.0 cm Hg pressure and 21øC.Errors are determined from counting statistics and laboratory error.DIC /i•3C was measured as above on CO• from reburned acetylene gas samples.2-4) were obtained from the manometrically observed yield of COs resulting from the UV oxidation of 5-L seawater samples as described above.DOCa,c concentrations ([DOCatc]) in the NCP were measured by Y. Suzuki [Druffel et al., 1989b] and in the Atlantic by J. Bauer [Bauer et al., 1990] and Y. Suzuki (unpublished results, 1989).The former SS values are from 100-/tL discrete seawater injections using the Dohrmann Model DC-90 analyzer in a single injection mode, and the latter from similar injections using a 3% Pt/alumina catalyst [Suzuki et al., 1992;Williams et al., 1992].The two Atlantic [DOCat•] profiles agree within 5% (Table 4).All values using the DC-90 system were corrected for injection blank values.
The concentrations of FAA and THAA were obtained using high performance liquid chromotography (HPLC) fiuorimetric determination of the o-pthaldialdehyde derivatives [Lindroth and Moppet, 1979], following acid hydrolysis of the combined amino acids with 6 N HCI [Robertson et al., 1987].These concentrations of THAA could be minimum values, since recent work by Keil and Kirchrnan [1991] using a vapor phase hydrolysis of seawater samples [Tsugita et al., 1987] indicated that the concentration of THAA in seawater may be 0.8 to 3.0 (average = 1.54 4-0.43, n = 62) times that obtained by the standard 6 N HCI hydrolysis at 120øC.
The TCHO concentrations were measured spectrophotometrically on samples hydrolyzed with 0.1 N HC1 using the procedure of Burney and Sieburth [1977] and are reported   All of the data except that for humic substances and POC from the NCP (Alcyone-5 and Eve-l) cruises are summarized in Tables 2 and 3. (Some of Alcyone-5 data in Table 2 are corrected from those previously reported by Druffel et al. [1989a].)All data from the SS (Hydros-6) cruise, except the humic substances and POC, are listed in Table 4 and are described below.Data for the POC samples are in Table 5 and the humic substances in Table 6.In this section we discuss the SS Aa4C data, compare it with that from the NCP, and then present the bulk carbon and organic constituent concentration measurements as they apply to the organic carbon cycle in the open ocean.There was no significant increase in A•4C in the 3, 20 and 85-m surface samples (Table 4) relative to the 100-m value.

Sargasso Sea
There may have been hydrocarbon contamination in the SS 3-m surface sample, judging froin the A•4C and 5•3C values found for the surface microlayer (upper 250 tim) a. nd 10-cm subsurface samples: -869 q-3 and-280 4-8%0, respectively (Table 4).This organic contamination was probably a re- These results illustrate the extensive interactions between different pools of dissolved and particulate organic matter.

Comparison of SS and NCP Ii•a C Results
The/i•aC profiles for DOCu,from the SS and NCP (Fig-

Other Sources o] Organic Carbon
In the following discussion we assume that the major source of organic carbon to the water column is directly  2 and 4, Figure 8).Assuming that these similar high concentrations prove to be valid, this suggests that the ad- To learn more about the origin and cycling of DOC in the oceanic water column, the following information is imperative: (1) characterization on a molecular level of the remaining 80-90% of DOC in seawater, including their isotopic signatures where possible, and (2) seasonal studies of suspended and sinking POC (including marine snow), to determine the origin of particles (whether surface or deep) in the water column.
. Seawater was pumped through the analyzer at ~30 mL/h and DOC was oxidized to CO• with Co/CoO-alumina or Co/CoO-alumina plus 2% Pt-faujisite catalysts (provided by M. Occelli) at 900øC in an O• stream (150 mL min-•).The resultant CO• was trapped on a molecular sieve at room temperature, desorbed at 450øC, and collected [Bauer et ol., 1992b].Blank values averaged 1.4 4-2.5 ttg C/h over all runs.The blanks were primarily due to passage of O• carrier gas through the system and secondarily to reactions of H•O within the combustion tube.The filters containing sinking and suspended POC were acidified with 1% H•PO4 for 24 hours at 20øC to remove the carbonates, dried under vacuum, and cornbusted to CO• in double quartz tubes as above.Aliquots of frozen sediment samples were acidified with 5% H•PO•, allowed to stand ~3 hours at 5øC to a constant pH of 2

AA# are the
sample numbers for the DOC•,• A•4C analyses measured at the University of Arizona TAMS facility.Event numbers are consecutive sampling periods during the cruise that represent separate casts of Gerard barrels, Yentsch pumps, nets, etc. DOCat•(Meteorological Research Institute--MRI) measurements were made by Y. Suzuki and reported previously [Draftel et al., 1989b].DOC,• values are 8% less than those reported earlier [Draftel et al., 1989a] due to a calibration correction.DOCu• A•4C values are slightly less than those reported previously [Druffel et al., 1989a] due to the correction of A14C using the actual 5laC of COs obtained from UV oxidation.DIC, alkalinity, FAA, THAA, TCHO, chlorophyll-a and phaeophytin concentrations, and DIC A14C measuremerits were reported previously [Druffel et al., 1989a].ß Initial seven samples analyzed for DOCuv A•4C that contained ~10% extra DOC than samples of DIC in the SS show the presence of bomb a•C in the main thermocline ((850 m) and in the up-
300%0 difference between DIC and DOC,,,A•4C valueswas found for both the NCP[Williams and Druffel, 1987]   (Figures3 and 4) and SS deep waters (Figure5).In both oceans, the DOC,,,pool is consistently "older" than the DIC, indicating that DOC is recycled on a longer time scale than DIC.The similarity of the differences between the average deep DOC,,v A•4C and the average deep DIC A•4C values in the Atlantic and Pacific suggests that circulation is a major factor controlling the distribution of DOC in the deep sea.However, small differences in A•4C DOC,,,• between the two oceans may reveal additional sources of young DOC to the Atlantic (see section 4).The A•4C values of suspended POC are indistinguishable from DIC A•C values in the upper 100 m from both the NCP (147 4-13% o, N = 3) and SS (125 4-7% 0, N = 6) sites (Figures 2, 3, and 6).However, profiles of A•4C in the suspended POC decrease faster with depth in the SS than in the NCP (Figure 6).Druffel and Williams [1990] suggested that a source of "old" organic carbon, perhaps DOG, was incorporated into the suspended POC.Seasonal control on the A•4C profile of suspended POC is not known; studies are ongoing to determine the influence of flux rate on the A•4C values, and hence the degree to which the deep sinking POC is surface derived.It is possible that seasonal changes in the flux rates of sinking POC somehow influence the incorporation of "old" carbon into both the suspended and sinking POC pools.The A•4C values of sinking POC at 4165-m and 5165-m depth at the NCP site during Eve-1 were 99% 0 and 136% 0, respectively, which are slightly lower than the average surface suspended POC A•4C value [Druffel and Williams, 15,650 DRUFFEL ET AL.' CYCLING OF DISSOLVED AND PAgIICULA• ORG•C h/IATIER 1990] (Table 5).The same was true of the sinking POC A•4C at 3200 m in the SS.A•4C values for sinking POC collected on the Alcyone-5 cruise during the low flux period (18•too and -39•too) were significantly lower than those obtained during Eve-l, the high flux period, suggesting that a higher fraction of "old" organic carbon was contained within the sinking POC during periods of low flux.A•4C vMues of the total sedimentary organic carbon (SOC, particle plus porewater fractions) in a gravity core taken at Station DW in the SS (bottom depth = 4460 m) were-346 -4-6•t.. (0-0.5 Clll),-372 -4-8700 (1.2-2 cm),-380 4-•7oo (3-4 cm) and-411 + •7oo (7-8 cm) (Figure 2), compared to NCP (Alcyone-5, bottom depth = 5710 m) values of-538 4-4 (0-1 cm),-607 4-4•t.o(1-2 cm),-686 4-3 (2-3 cm) and-787 4-4 (5-7 cm) (Figure 3).The A"C values for the DOC•,• 50 m above the bottom at these two sites were-391 4-5•to.(SS) and-546 4-12•to.(NCP).These close similarities between the AI•C of surficial SOC and that in the bottom water DOl•C• values, if not fortuitous, suggest that the SOC is either a source or a sink for the bottom water DOC• or that the seawater-pore water exchange of DOC• masks other organic 14C signals in the upper sedimentary column.
) measured in NCP and SS waters are plotted in Figure 10.TCHO values in the SS are approximately half those in the NCP.TCHO constitutes about 12/•M of the total 80/•M DOC•,• concentrations in the Pacific.One explanation is that dissolution of particles occurs with time, slowly increasing [TCHO] as water ages on its way to the Pacific.This presumes that this TCHO h'action is not readily utilized by heterotrophic organisms.Another possibility is that there are more polysaccharides of recent origin in the SS water column that resist hydrolysis and a, re only slowly hydrolyzed at depth to the more reactive oligomeric or •nonomeric material or HCl-hydrolyzable derivatives (J.D. Pakulski and R. Benher, personal communication, 1991).This slow conversion of polymeric to monomeric carbohydrates would result in the higher concentrations of TCHO measured in the "older" N CP deep water.The profiles of THAA and FAA in the NCP (Figures 10b and 10c) have the same general features as TCHO, with the scatter in the FAA values probably due to the extremely low concentrations measured here in deep water.The DOC equivalent of the THAA in the NCP is 1
ditional 7 /•M DOCuv in the SS may be altered enroute to the NCP, and replaced with an equal amount of DOC oxidized only by high-temperature catalytic oxidation methods.Oxidation of the extra 7-#M DOC,• in transit to the Pacific could result in increased concentrations of carboxylic acid-type compounds that are known to be among the most resistant to free radical attack such as occurs with UV and persulphate wet combustions [Peyton, 1992].All fractions of DOC that we have measured, including DOCu•, humic material and a portion of the DOC•tc, have mean ages essentially identicM to each other.Therefore we conclude that a large part of each of these fractions are old and have a similar origin, and that they compose, in effect, a very old background of organic matter which resists in situ oxidation, chemical or biochemical.Superimposed on this old fraction may be a more reactive, younger DOC component, particularly in the upper 500 m of the w•ter column, which recycles on faster time scales than we measure here.It is this fraction which may correlate with AOU.Nevertheless, it is obvious that the bulk of the DOC does not control AOU directly at these two oligotrophic ocean sites.4.5.Deep-Sea DOC and Photosynthetically Fixed Carbor• Assuming a steady state concentration of DOC and no terrestrial, rivefine or atmospheric inputs, the amount of photosynthetically fixed carbon that enters the deep sea as DOC and is concurrently oxidized completely to CO• can be estimated from the mean apparent age of the DOC pools in the NCP and SS.At 40 #M of DOC• or 110 ltM of DOC•tc, the total reservoirs of DOCu• and DOCatc in is of the order of 20% of PPt (6-30%) [Eppley and Peterson, 1979; Martin et al., 1987; Pace et al., 1987; Laws et al., 1989], or 20 x 10 as g C/yr -a, then the steady state input of DOC represents 0.12% of PPt and 0.6% of PP,• for DOC,,,,, and 0.34% of PPt and 1.7% of PP,• for DOCat•.For comparison, these inputs are about 2 g of DOC•t• (at 110/•moles L -a ).The corresponding DOC replenishment rates for a 6000-year residence time are 0.35 g C/yr -a (~1 mg C d -• for DOCu•), and 1.1 g C/yr -• (~3 mg C d -1) for DOCat•.At a sinking rate of 50 m d -• for POC (estimates vary from 30 to >150 m d the NCP.There are, however, no data for organic carbon flux measurements taken concurrently at ~1000 m and ~5800 m.The fact that the requisite DOC inputs can be attributed to flux differences does not take into consideration other DOC sources to the deep-water DOC pool, namely, mixing or advection of surface water DOC into the deep ocean, diffusion of DOC fi-oln the sediments into the over.lying water, and solubilization of suspended POC.The corresponding flux gradients required for the SS (of the organic matter in the oligotrophic NCP and SS, we have determined that most of the deep DOC has an average apparent •4C age of 4000 (SS) to 6000 (NCP) years B.P. This reflects a recycling of about 80% of the deep DOC during each deep ocean mixing cycle assuming a closed system.The apparent ages of the fractions (DOCuv, DOCFt-htc, humic materials) within an oceanic water column appear to reveal the same general values.There appears to be a significant amount of bomb •4C in the deep SS DOC pool, since the values there are slightly higher than those expected from deep circulation only.Most of the DOC in the oceans at our two sites shows little correspondence with AOU, but this does not negate the possibility that a more reactive, younger fraction of DOC exists that is oxidizable only by discrete injection high-temperature catalytic oxidation and recycles on shorter time scales than in the DOC pools measured here.