DISCUSSION

Site 1235 is located at a shallower depth than Site 1234, which has led to a higher fraction of terrigenous siliciclastic component in the sediments at this site. However, the total sedimentation rate at Site 1235 is slightly lower than at the deeper neighboring site (696 m/m.y. vs. 788 m/m.y.) according to the preliminary age model presented in Shipboard Scientific Party (2003a, 2003b). A possible reason for this is sediment winnowing by benthic currents, as suggested by the presence of distal turbidite deposits in the sediment sequence at Site 1235. The combination of sediment winnowing and the higher degree of dilution by siliciclastic material resulted in lower concentrations of TOC and TN at Site 1235 compared to Site 1234. Consequently, the rate of organic matter decomposition (per volume of sediments) is lower for Site 1235 as well. This is consistent with the lower DIC and ammonium concentrations and deeper penetration of sulfate at this site. The average value of ~10 for the atomic TOC/TN ratios at both sites (Fig. F2C, F2D) is higher than the typical C/N Redfield ratio of marine organic matter of ~7 (Redfield et al., 1963) and might either reflect a small contribution from terrestrial organic matter, which typically has C/N >20, or represent a diagenetic effect. It is presently difficult to distinguish between these two possibilities without more specific information about the composition of organic matter in these sediments.

Ammonium profiles indicate that the net release of this metabolite appears to cease at depths below ~60 mcd at Site 1234 and ~40 mcd at Site 1235, where concentration maxima occur. Below these maxima, a modest decrease in ammonium occurs at both sites. If ammonium profiles are currently in steady state, a decrease could result from a sink for ammonium at depth. Alternatively, the decrease may represent temporal variation in the input of reactive organic matter in sediments that accumulate rapidly. At both sites, high sediment accumulation rates strongly influence solute transport through the sediment column. The Peclet number can be used to indicate the relative importance of diffusive and advective transport; values <1 indicate dominance of advection. The fast accumulation rates at these sites result in low (<1) Peclet numbers for both sites (Table T4), indicating that advection is more significant than diffusion for solute transport and in regulating the depth distribution of solutes. Consequently, it is most likely that the ammonium decreases at depth do not reflect a sink but rather are due to higher release rates near the depths of the apparent concentration maxima, in comparison to adjacent horizons. The elevated concentrations of TN near the depths of pore water ammonium maxima support this interpretation. Because of the strong influence of advection, diffusive transport is not able to homogenize the ammonium profiles at depth.

A fraction of the N_org has been converted to ammonium by diagenesis during burial. We can estimate this fraction of N_org by considering the fate of the time-averaged flux of N_org buried below 1.45 mcd (the uppermost interval where data are available). The calculation can be done by comparing N_org at 1.45 mcd to that at the depth of the ammonium maxima. At Site 1235, the decrease from 0.11% to 0.08% (Table T2) indicates ~27% is converted to ammonium during diagenesis. At Site 1234, there is not a TN measurement at 1.45 mcd, so such a calculation cannot be carried out.

Alternatively, the diagenetic transformation of N_org can be calculated from pore water ammonium profiles. The advantage of this approach is that it provides a more sensitive assessment of the N_org degradation. Both approaches rely on the assumption of steady state; their comparison helps to assess the validity of the steady-state approach. The maxima in ammonium profiles indicate the steady-state assumption is not strictly correct; however, it provides a useful approximation.

Using the pore water ammonium profile, the total N_org decomposition in a specified layer should be the sum of ammonium lost via diffusion plus advection from this layer due to burial of pore water (Berner, 1980). The net ammonium production between two horizons (P_N) can be calculated based on mass conservation:

P_N = (J_D2 – J_D1) + (J_A2 – J_A1), (1)

where

J_D1 = diffusive fluxes across the upper boundaries of the interval,

J_D2 = diffusive fluxes across the lower boundaries of the interval,

J_A1 = advective transport across the upper boundaries (Table T4), and

J_A2 = advective transport across the lower boundaries (Table T4).

Diffusive fluxes at the upper and lower boundaries were calculated by applying Fick's first law. Concentration gradients were determined as tangents to the ammonium profile at the boundaries of the considered interval, calculated from the derivative of a model curve fit to the profile (details are in Prokopenko, 2004). Advective fluxes were calculated as

, (2)

where

C_pw = concentrations at the boundaries,

K = ammonium partition coefficient,

= porosity,

w_pw = advection rate of pore fluid relative to the sediment/water
interface, and

w_s = sediment accumulation rate (Table T4).

In the absence of compaction, w_s = w_pw.

The calculated P_N (Table T4) represents the total amount of N_org degraded within the specified sediment interval. The intervals considered for Site 1234 and Site 1235 were 1.45–62 mcd and 1.45–42 mcd, respectively (the depth of our uppermost data point is 1.45 mcd). The total flux of N_org delivered to the 1.45-m horizon from above is calculated as a sum of the N_org burial flux at the bottom of the zone of active diagenesis (at 62 mcd for Site 1234 and at 42 mcd for Site 1235) and the depth-integrated ammonium production within the specified intervals. The N_org fluxes at lower boundaries at both sites were estimated based on average sedimentation rates of 788 and 696 m/m.y. (for Sites 1234 and 1235, respectively), the TN_org concentrations (measured TN concentrations corrected for adsorbed ammonium), and a porosity value of 0.65. We calculated (Table T4) that at Site 1234 the total flux of ammonium produced between 1.45 and 62 mcd is 0.025 mmol/m²d. This constitutes 19% of the 0.106 mmol/m²d time average N_org flux exported below 1.45 mcd. At Site 1235, the ammonium flux produced within the upper 42 mcd is slightly lower, 0.017 mmol/m²d. This represents 18% of the N_org burial flux at 1.45 mcd, and ~82% ± 10% remains as N_org. This is in reasonable agreement with the earlier estimates of 73% based on the measured decrease in N_org at this site. To summarize, ~20% of N_org is converted to ammonium at both of these sites during diagenesis of organic matter below 1.45 mcd.

Ammonium production fluxes in the interval of active diagenesis indicate that 70%–90% (when uncertainties associated with porosity estimates and possible nonsteady-state conditions are taken into account) of the organic matter buried below 1.45 mcd at Sites 1234 and 1235 escapes further degradation downcore. It is important to point out that these calculations do not reflect the total burial efficiency of the organic matter flux delivered to the ocean floor. Therefore, it cannot be directly compared to previously published burial efficiencies in the coastal marine sediments, though burial efficiency close to 80% has been reported for some coastal marine sediments (e.g., summary by Burdige [2005]). As a detailed discussion of factors contributing to the high degree of organic matter preservation is beyond the scope of this paper, we would like to simply point out that rapid sediment accumulation rates have been previously proposed as one of the possible mechanisms contributing to the enhanced preservation of sedimentary organic matter (Hartnett and Devol, 2003; Hartnett et al., 1998). These authors argued that one of the key factors influencing the burial efficiency is the time of organic matter exposure to aerobic degradation; therefore, in rapidly accumulating sediments, organic matter is removed from the oxic zone before it is substantially degraded.

The most prominent feature of the isotopic profiles at both sites is a noticeable similarity between the ¹⁵N of ammonium and ¹⁵N of N_org (Fig. F3). Peclet numbers <1 indicate that the high rates of advection substantially decrease the degree of openness of the diagenetic system at both sites. Thus, only a modest fraction of the ammonium released at a particular horizon is able to diffuse away (because the sink location at the sediment/water interface is moving rapidly upward due to sediment accumulation).

At Site 1235, ¹⁵N of ammonium very closely follows the ¹⁵N of N_org (Fig. F3B). The averages for both isotopic ratios are very similar at this site as well, suggesting that no isotopic fractionation occurs either during ammonium release in the upper 40–50 mcd or below, where net ammonium release ceases.

At Site 1234, the shape of the ¹⁵N of ammonium profile bears strong resemblance to that of ¹⁵N of N_org, but below ~60 mcd it appears to be offset upward relative to ¹⁵N of N_org. The offset interval thickness increases with depth, suggesting that the observed offset is depth and/or time dependent. From this, we conclude that the observed offset is probably due to the difference between the burial velocities of fluids and solids. Such differences arise from sediment compaction, which extrudes pore fluids upward as sediments are buried.

Sediment compaction results in a decrease of porosity downcore within a certain interval of compaction, below which compaction ceases and sediment porosity becomes constant. We can adjust the vertical profile of ammonium ¹⁵N values for the effect of compaction by computing the offset (_z) between pore water and solids at each depth horizon. _z can be calculated at each depth horizon (z) using

, (3)

where

v_s = burial velocity of solids,

v_w = burial velocity of pore water relative to the sediment/water
interface, and

t = time.

Next, the v_s can be recast as

so that Equation 3 becomes Equation 4:

, (4)

where

z = depth.

We assume a linear decrease in porosity through an interval of thickness b, so that porosity changes downcore according to Equation 5:

, (5)

where

,

_z = porosity at depth z,

₀ = porosity at the sediment/water interface,

_b = porosity at the base of the compaction zone, and

b = depth of the base of the compaction zone.

As shown by Berner (1980), steady-state compaction requires the following equation to be true:

_zv_w = _bV_b, (6)

where

V_b = velocity of pore water = velocity of solids below the compaction
zone, where further compaction does not occur.

For pore water, Equation 6 can be written as

(1 – )v_w = (1 – _b)V_b. (7)

Substituting Equation 7 into Equation 4, we obtain the following:

, (8)

where

.

Evaluating the integral in Equation 8 over the interval from 0 to z, we obtain

. (9)

Then _z can be found:

. (10)

Porosity measurements at Site 1234 indicate a systematic decrease from 0.65 near the surface to 0.55 at 250 mcd. Assuming that 250 mcd was depth b, the base of the compaction zone, we can calculate the offset between solids and fluids at each depth horizon. Figure F4 shows the adjusted and measured profiles of ¹⁵N ammonium to illustrate the effect of compaction.

At Site 1235, the ammonium ¹⁵N follows the sedimentary ¹⁵N more closely, and the effect of compaction seems to be missing. At this site, no porosity gradient was observed. It is possible that higher input of siliciclastic material and lower content of organic matter makes the sediments at Site 1235 less compactable than those at Site 1234.

Figure F4 demonstrates that once the impact of compaction is accounted for, ¹⁵N of ammonium reflects the isotopic composition of organic matter at Site 1234. One implication of such close similarity between ¹⁵N of ammonium and N_org is that no isotopic fractionation is associated with the diagenesis of marine organic matter at this site, similar to Site 1235. The slight difference between ¹⁵N of ammonium and N_org is probably due to the partial diffusive mixing between the horizons with variable isotopic composition of organic matter (Fig. F4) wherever local isotopic gradients are formed. Diffusion is not fast enough to homogenize the ¹⁵N of ammonium between these horizons. It is important to note also that no fractionation is observed through the part of the profile where ammonium is being presently released within the upper 50–60 mcd.

The close similarity between the ammonium and N_org ¹⁵N profiles indicate that little or no fractionation is associated with the long-term decomposition of organic matter at Sites 1234 and 1235. We reached similar conclusions based on our results from Site 1230 (ODP Leg 201) (Table T3), where despite the loss of more than 30% of the original organic matter, we found no evidence of changes in ¹⁵N of residual N_org (Prokopenko et al., 2006b). Our findings are consistent with the inference of Altabet et al. (1999a, 1999b) and Pride et al. (1999), who suggested the absence of diagenetic fractionation of nitrogen isotopes in rapidly accumulating organic-rich coastal sediments where organic matter is well preserved. As factors controlling the extent of organic matter diagenesis, these authors considered poor oxygenation of bottom water, high concentrations of organic matter, and rapid sediment accumulation rates. The O₂ concentration of the bottom water does not affect diagenesis in the sediments in the depth ranges considered in this study. These sediments do not have particularly high organic matter concentrations, and consequently, the concentration of organic matter is not as important for the high degree of preservation as are sediment accumulation rates. Sediments at Sites 1234 and 1235 contain low to moderate concentrations of organic matter (due to high degree of dilution with siliciclastic component) but accumulate at very high rates (788 m/m.y. and 696 m/m.y., respectively).

Moreover, when the results from Site 1230 are taken into consideration, it appears that the degree of organic matter preservation may not be the dominant factor behind the absence of diagenetic alteration of sedimentary ¹⁵N. Perhaps another important factor, shared by the three sites, determines the absence of any significant isotopic fractionation in N_org. The sediments of all three sites (Table T3) are characterized by relatively low atomic C/N ratios (~10). This value indicates that the sedimentary organic matter is predominantly of marine origin. Apparently, the large input of terrigenous siliciclastic component is not associated with a transport of organic matter from land at these sites. We have shown previously (Prokopenko et al., 2006a) that preferential degradation of an isotopically distinct, more labile organic fraction (such as marine vs. terrestrial organic matter) may lead to a change in the residual bulk ¹⁵N of the sediments. The difference of 2–3 was observed between ¹⁵N of pore water ammonium and N_org in the sediments of Site 1227 collected during the Leg 201. Sediments from Site 1227 (Prokopenko et al., 2006b), may contain a significant fraction of terrestrial organic matter as suggested by their high C/N ratios ranging from 12 to 18, with the average of 15, which appears to be resistant to degradation. Lithology of the sediments at Site 1227 (D'Hondt, Jørgensen, Miller, et al., 2003) also presents strong evidence for significant input of terrestrial material. Terrigenous organic matter may differ isotopically from the more labile marine N_org; in this case, the preferential degradation of the isotopically distinct marine fraction may leave the ¹⁵N of remaining bulk nitrogen diagenetically altered.

The observed absence of fractionation between ammonium and N_org at Sites 1234 and 1235 is consistent with a dominance of the marine component in the organic matter at these sites. From the discussion above, we can state that the long-term diagenesis of marine organic matter in rapidly accumulating sediments does not affect the isotopic composition of preserved N_org, as long as the terrestrial component is minor. If both of these conditions occur, then the depth-dependent variations in preserved ¹⁵N should reflect changes in input composition.