RESULTS

Samples from both Site 969 and Site 974 revealed clear correlations between the three measured parameters: percent organic carbon, ¹³C_org, and ¹⁵N. Sapropel intervals of increased organic carbon content were marked by lighter ¹⁵N values and almost always by heavier ¹³C values. Table 2 and Table 3 provide compositional and isotopic data for each sample; Figure 1, Figure 2, Figure 3, Figure 4, and Figure 5 show data from representative sapropels and their surrounding sediments.

At Site 969, organic carbon content averaged 0.48% (standard deviation [S.D.] = 0.35%) during normal sedimentation and 4.69% (S.D. = 2.63%) during sapropel formation, with no consistent trends downcore. Sapropel values ranged from 2.5% to 9.2% organic carbon. Two exceptions were sapropels A1 and A5, which showed no elevation in organic carbon content despite being identified as sapropels (Shipboard Scientific Party, 1996a).

At Site 974, the organic carbon gradient between sapropel and background intervals was less pronounced, with sapropels averaging 1.61% (S.D. = 0.42%) organic carbon and surrounding sediments averaging 0.51% (S.D. = 0.19%). No downcore trends were apparent in either the sapropels or the background sediments.

Nitrogen was one to two orders of magnitude less abundant than carbon in all samples. Nonsapropel values averaged 0.018% (S.D. = 0.007%) at Site 969 and 0.030% (S.D. = 0.010%) at Site 974 while sapropel values averaged 0.179% (S.D. = 0.095%) at Site 969 and 0.088% (S.D. = 0.022%) at Site 974. Within sapropel intervals, nitrogen increased in approximate proportion to carbon, so that the C:N ratio did not change significantly across sapropel boundaries.

At Site 969, ¹⁵N averaged 3.5 (S.D. = 1.5) in background sediments, 0.2 (S.D. = 1.5) in sapropels, and ¹⁵N (¹⁵N_sapropel - ¹⁵N_background) ranged from +0.45 to-6.79. Although the mean ¹⁵N varied little from sapropel to sapropel, we found relatively large variations in ¹⁵N within narrow stratigraphic zones within the sapropels. At Site 974, ¹⁵N averaged 5.0 (S.D. = 0.6) in background sediments and 2.2 (S.D. = 0.8) in sapropels, and ¹⁵N ranged from -2.80 to-3.08.

Compared to these values, the data of Calvert et al. (1992) from the eastern Mediterranean are markedly enriched in ¹⁵N during background sedimentation (average ¹⁵N +6) and somewhat enriched during sapropel sedimentation (average ¹⁵N +1 to +3).

At Site 969, ¹³C averaged -24.7 in normal sediments, with considerable variation through the sediment column (S.D. = 1.9). However, within any single stratigraphic zone, ¹³C values were relatively constant. Sapropel ¹³C did not vary greatly, with an average value of -23.1 (S.D. = 0.97). The resulting average difference in ¹³C between sapropel and background sediments (¹³C = ¹³C_sapropel - ¹³C_background) was 1.60, a highly significant difference (single-factor ANOVA, p < 0.05). Variations in ¹³C_background through the sediment column led to ¹³C values ranging from 4.44 in sapropel A7 to -1.40 in sapropel S1; in three sapropels (S1, S6 and S8), ¹³C_sapropel was actually more negative than ¹³C_background.

The ¹³C trends were more consistent at Site 974, where the ¹³C of background sediments averaged -25.8 (S.D. = 1.0) and the ¹³C of sapropels averaged -24.5 (S.D. = 0.67). The ¹³C offset at this site varied between 0.98 and 1.66, with a mean of 1.32, a statistically significant difference (single-factor ANOVA, p < 0.05).

Virtually all samples at both sites contained less ¹³C than the eastern Mediterranean samples analyzed by Fontugne and Calvert (1992), which had a mean ¹³C of -18.8 ± 1.07 for nonsapropels and -21.0 ± 0.82 for sapropels. Furthermore, at Sites 969 and 974, ¹³C generally increased during sapropel intervals, whereas in sediments studied by Fontugne and Calvert (1992), ¹³C decreased during sapropel intervals.

A comparison of Sites 974, 969, and MD 84641 reveals a consistent east-west ¹³C gradient and spatially inconsistent trends in ¹⁵N and percent organic carbon. Table 4 presents average values of these parameters from the three sites, and Table 5 and Table 6 show east-west ¹³C and ¹⁵N gradients including data from other sites.

One-way ANOVAs were performed to determine the statistical significance of isotope differences between Sites 969 and 974. The ¹³C of organic matter was significantly lower at Site 974 than at Site 969 for both sapropel intervals (p < 0.05, F = 32.7) and background intervals (p < 0.05, F = 13.5). However, ¹³C was essentially the same in both cores (p = 0.81; F = 0.06).

The difference in ¹⁵N between Sites 969 and 974 was highly significant for both background (p < 0.05, F = 40.9) and sapropel (p < 0.05, F = 27.8) sediments, with Site 974 averaging 1.8 heavier than Site 969. However, ¹⁵N was virtually identical (p = 0.99, F < 0.001) between the two sites, with an average value of about -3.

The contrast in organic carbon content between sapropel layers and nonsapropel layers was more dramatic at Site 969 (enrichment factor 10) than at Site 974 (enrichment factor 3) (p < 0.05, F = 5.7).