NO A A PROFESSIONAL PAPER 11 



seabed on the periphery of the low D.O. area in 1976 was 

 due to the combination of higher rates of seabed oxygen 

 consumption and lower rates of total plankton respiration, 

 presumably due to oxygen limitation and other unknown 

 factors. It seems certain that the seabed received larger 

 fluxes of oxidizable carbon in 1976 than in 1977. 



NET OXYGEN DEPLETION AND 

 UTILIZATION RATES 



Atwood et al. (ch. 4) calculated oxygen depletion rates 

 for subpycnocline waters. Depletion rates are regressions 

 of observed oxygen concentrations versus time. From the 

 slope of these regressions these authors estimated an av- 

 erage oxygen depletion rate of 2.2 ml O./mVh for sub- 

 pycnocline water (segments A, Jl, J2, Ml. and H in ch. 

 1, fig. 1-9) comparable to our study area. Han et al. (ch. 

 8) calculated an average net oxygen utilization rate of 4.0 

 ml Oj/mVh for May-June 1976 and included advective in- 

 puts and outputs of oxygen and water for the various boxes 

 in their model. In this study (August-September 1976), 

 the measured rates of total plankton respiration in sub- 

 pycnocline waters (average thickness 9.4 m) were 0.1 to 

 5.0 ml Oj/mVh and averaged 1.8 ml O./mVh. Our meas- 

 ured rates of seabed oxygen consumption averaged 16.9 

 ml OJm-lh. Adding the oxygen consumed by the overlying 

 subpycnocline waters results in an average of 3.7 ml O,/ 

 mVh consumed below the pycnocline. This compares fa- 

 vorably with the net utilization rates derived from the 

 model of Han et al. (ch. 8). Despite this agreement it must 

 be stated that our measurements of oxygen consumption 

 in the water column were taken several months after the 

 time frame used in the model. In June 1977 we again 

 measured oxygen uptake both in the water column and 

 on the seabed. From these we calculated that an average 

 of 4.2 ml 0,/mVh was used below the pycnocline (27 m). 

 For comparison, Tsuji et al. (1974), also dealing with a 

 shallow (20 m) two-layered aerobic/anaerobic system, es- 

 timated an oxygen utilization rate of 6.8 ml OJmVh, based 

 on measured decreases in organic carbon over a 2-month 

 period in conjunction with a red tide in the highly eu- 

 trophic waters of Tokyo Bay. 



ANAEROBIC METABOLISM 



Our method for measuring aerobic respiration rates via 

 oxygen changes could not be used in the anoxic subpyc- 

 nocline water. However, we did observe high concentra- 

 tions of sulfide in the subpycnocline waters of the anoxic 

 area, indicating that anaerobic metabolism must have oc- 

 curred. 



Estimates of anaerobic metabolism may be made using ^ 

 Richards' model (1965) relating organic carbon, anaero- 

 bically decomposed, to the phosphate liberated. In figure ^ 

 10-1 1, soluble reactive phosphate concentrations are plot- 

 ted against oxygen concentrations for all samples collected 

 during our cruise except those samples near the estuary 

 (stations 45, 69, 101, 102, 109) and above the pycnocline 

 at stations 41 and 34. The high concentrations of phos- 

 phate plotted along the ordinate (1.35-3.6 |jlM-P/1) were 

 measured in near-bottom anoxic water having high sulfide 

 concentrations (figs. 10-3B and 10-5; stations 213, 217). 

 Excluding points representing less than 0.1 ml OJl, the 

 correlation coefficient between oxygen and phosphate is 

 -0.85 (n = 83). The functional regression line (Ricker 

 1973), drawn in figure 10-11, intercepts the y-axis at a 

 phosphate concentration of 1.03 |jlM/1. This value pro- 

 vides an estimate of the phosphate concentration imme- 

 diately before anoxia. 



Examining stations 213 and 217 (the low D.O. area), 

 we estimate that 9,175 and 7,785 p.M-P/m', respectively, 

 were more than the expected concentration of phosphate 

 (1 p.M/1) in the 6-m and 15-m subpycnocline water, re- 

 spectively, at the start of oxygen depletion. Multiplying 

 the phosphate values by an atomic ratio of 106C:1P yields 

 an estimated 12 and 10 g of organic carbon needed to 

 account for the excess phosphate mineralized. The elapsed 

 time between our visit to station 213 and the first reports 

 of oxygen depletion in the vicinity of 213 was about 45 

 days. The estimated hourly rates of anaerobic metabolism 

 of carbon for station 213 and 217 are 1.8 mg C/mVh and 

 0.6 mg C/mVh. These estimates are slightly lower than the 

 measured aerobic rates of water column carbon miner- 

 alization for subpycnocline water at stations adjacent to 

 the low D.O. area (2-4 mg C/m'h) and are conservative, 

 because some proportion of the inorganic phosphate 

 formed diffused upward through the pycnocline. 



Bacterial densities and biomass were greater in oxygen- 

 depleted areas than in adjacent regions (fig. 10-10, table 

 10-3). Different cell morphology and larger bacteria be- 

 low the pycnocline in the anoxic area (fig. 10-12) suggest 

 that different bacterial species or populations with differ- 

 ent functional capabilities and responses developed there. 

 The greater density of bacteria below the pycnocline in 

 and surrounding the anoxic area (fig. 10-10) suggests an 

 additional nutritive source, oxidizable organic carbon, 

 present as DOC and FOC. Bacterial biomass in the bot- 

 tom water of the low D.O. area in 1976 (table 10-3) was 

 about twice the value reported by Barvenik et al. (1976) 

 for spring 1974 and 1975 (48 mg C/m'). FOC and DOC 

 substrates were comparable in both areas. We might ex- 

 pect aerobic-anaerobic mineralization rates also to be 

 comparable. 



252 



