Geochemistry of Mercury in Palos Verdes Sediments Robert P. Eganhouse’’, David R. Young, and Joseph N. Johnson Southern California Coastal Water Research Project, 1500 E. Imperial Highway, El Segundo, Calif. 90245 Sediments affected by a major submarine wastewater outfall on the Palos Verdes shelf were analyzed for total mercury and organic mercury. In addition, chemical leaching studies were performed to determine the phase partitioning of mercury in these sediments. A statistically significant decline in surface sediment concentrations of total mercury during 1972-1975 may be due to several factors including the reduced emission of wastewater solids. Organic mercury levels up to 21.3 pg/dry kg and 2% of the total mercury were found in sediments near t,he outfalls; however, the sulfide-rich sediments in the immediate vicinity of the outfalls contained low or undetectable levels of organic mercury. Most of the mercury in these sediments is fixed in a refractory phase, although organic-associated mercury increases in prominence away from the outfalls. In the most contaminated sediments, mercury appears to be accumulating in this refractory phase and, hence, is largely unavailable for introduction into the tissues of local marine life. The perturbation of certain coastal areas in southern California as a result of submarine wastewater discharge is a well-documented fact (1-4). In particular, the sediments surrounding Los Angeles County Sanitation District’s outfall system off Palos Verdes peninsula have been the subject of several investigations concerning the ultimate fate of sewage-derived organic carbon and heavy metals (5-7). As such, our knowledge of the geochemical cycling of trace metals in this area has been based largely upon bulk sediment properties. Clearly, a complete understanding of the fate of heavy metals depends on detailed information of the behavior of individual sewage constituents (and their natural counterparts) prior to, during, and subsequent to sedimentation. Temporal variations may be expected to play an important role as well. Once derived, information of this nature can be used to estimate the potential harm that might result when marine animals are exposed to anthropogenic wastes. The purpose of this study was to address specific aspects of the gebchemistry of mercury in Palos Verdes sediments and, thereby, assess its bioavailability to local marine life. S t u d y Area The distribution of sewage-derived particulates and associated pollutants over the Palos Verdes shelf is undoubtedly influenced by prevailing local hydrologic conditions. Figure 1demonstrates the distribution of total mercury in this region.
These contours are typical of all major effluent constituents ( 3 ) ,and the general features have not changed significantly since 1972. The pronounced extension of high surface sediment concentrations to the northwest of the outfall terminii and centered roughly along the 60-m isobath is caused by northward-flowing subsurface currents in this area (8).In the case of mercury, relatively high concentrations (>5 mg/dry kg) are found within the top 20 cm of these sediments until at depths below 20-30 cm, an apparent background level of approximately 0.05 mg/dry kg is reached. Young et al. (9)have estimated that roughly 4 metric tons of anthropogenic mercury are lodged in the upper 30 cm of these sediments. During the period from 1972 to 1975 some distinct changes were observed in the characteristics of Palos Verdes shelf sediments. As a part of their routine monitoring activities, the Los Angeles County Sanitation District’s Ocean Monitoring Group (LACSD-OMG)has measured H2S odor, Eh, ZS2-, and pH (IO)on the pore waters of surface sediments at 40 stations since 1972. The data demonstrated a progressive contraction of the hydrogen sulfide field from an area of 16 km2 in 1972 to small patchy regions (20 ccg/dry kg). Phase Partitioning Study. For the present study, extractions were carried out on the entire suite of 1975 sediments collected along the 60-m contour, the onshwe-offshore transect 6, and effluent particulates collected in 1976. Table I11 lists the analytibal results for the sequential extractions. For the moment, attention is confined to results for the 60-m contour plotted in Figure 5. For all samples tested, the water and MgC12 extractions were ineffective, indicating that measurable quantities of mercury are not present either in the form of readily soluble molecules or exchangeable ions (with
respect to magnesium). Small portions of the total mercury extracted were recovered in organic acid (humic and fulvic acids) extractions, percentages generally dropping off with distance from the outfalls and being low or nil for sewage particulates and sediments a t station 7C. The percentage of mercury yielded upon mild oxidation (3% H202) appears to be highest a t stations farthest from the outfalls; this feature decreases rapidly with proximity to the site of discharge. The rigorous oxidation seemed to mainly affect sediments taken from intermediate stations; the effluent particulates and sediments from stations 0,10, and 7 had little, if any, mercury in this phase. I t should be noted that the levels of mercury in the oxidizable organic phase (Le., in the 3% HzOz and 30% HzOz leaches, Table 111) and the concentration of "total organic mercury" (Figure 4) were markedly depressed or undetectable for stations 6C and 7C. By far, the most effective extraction was the aqua regia treatment. With the exception of the anomalous results for station 3C, the proportion of mercury recovered in this refractory phase was highest in the effluent and 7C sediment. With increasing distance from the outfalls the percentages drop off in a monotonic fashion, reaching a t the lowest, 35%. Data for the onshore/offshore distributions (Table 111) indicated that the mercury was concentrated primarily in the refractory phase of the sediments with a small percentage found in the humic fulvic acid fractions.
+
Discussion The data on mercury content of surface sediments revealed a statistically significant, albeit small, decline in mercury during the interval 1972-1975. Differences in sampling techniques could be invoked to explain the decrease between 1972-1975; however, the 1972/1973 results argue against this. Considering the magnitude of the change and the apparent lack of correlation between sample location and percent decrease, we are left to speculate as to the true cause(s) of the decline. Among the possibilities are: changes in the effluent characteristics, enhanced sedimentation of nonsewage particulate matter, resuspension and transport of the contaminated sediments away from the outfall area by currents, increased biological mixing of sediments, mobilization of mercury from the surface sediments to the overlying waters or underlying sediments, or some combination of these events. At first glance it would seem that the reduction in discharged effluent solids from 1971 to 1975 (Table IV) could
Table 111. Concentrations of Mercury (mg/Dry kg) in Various Chemical Phases of Sediments Collected from Stations on the Palos Verdes Shelf, 1975 stallon
oc 1c 2c 3c 4c 5c 6C 7cc
dC 9c 1oc 6A 6B
6D
H20 soluble
MgCl2
0.2 N NaOH
0.005 N NaOH
ND ND ND ND ND ND ND ND ND ND ND ND ND ND
ND ND ND ND ND ND ND ND ND ND ND ND ND ND
0.021 0.081 0.039 0.094 0.065 0.092 0.114 ND 0.070 0.056 0.016 0.054 0.129 0.017
0.029 0.173 0.332 0.246 0.529 0.584 1.23
0.007 0.179 0.044 0.332 0.035 0.228 0.040 0.010 0.048 ND 0.016 ND 0.067 ND ND ND ND 0.393 0.024 ND 0.037 ND 0.137 ND 0.091 0.021 0.034 ND 0.287 0.069 ND 0.617 0.001 ND 0.030 J WPCP effluent solids 31 261 78
ND
ND
0.485
0.209
1N
0.005 N HOAc
0.022
3%
Hz02
0.044
30 % H202
aqua regla
ND 0.557 0.629 0.058 1.16 0.481 0.054 ND 0.376 0.158 0.018 0.129 ND ND
0.144 0.649 0.695 2.68 1.28 1.96 3.84 4.75 1.55 0.935 0.202 1.05 2.92 0.234
0.380 1.84 1.96 3.13 3.08 3.13 4.75 2.41 1.32 0.348 1.55 3.74 0.282
0.360 2.28 2.13 3.04 3.24 4.37 4.73 4.36 3.69 1.50 0.394 1.55 3.50 0.368
4.95
5.71
5.60
ND
total ext'd Hg
5.30
*
total Hg
a This is the total mercury extracted in the eight-step scheme in terms of concentration based upon the original solids weight. This is the concentration of "total mercury" determined by the aqua regia digestion. Results for station 7C have been summarized by averaging the yields in the aqua regia step for samples 1-4 (see Table IV).
1154
Environmental Science & Technology
'ti 70
WITER SOLUBLE
ION EXCHANGEABLE
60
t
x
'.i 2"
10
0 I 2 3 4 5 6 7 E 8 910
ORGINIC
ORGANIC
ICIDS. I
ACIDS,
0 1 2 3 4 5 6 7 E 8 9 , O
3%
30%
H2 02
n*02
ihL STATIONS a 9 10 (60m 0 1contour) 2 3 4 5 6 7
0 ,2 3 4 5 6 7 E
TI
0 I 2 3 4 1 6 7 E 6 910
0 , 2 3 4 1 6 7 E 6 910
E 6 9 l O
Figure 5. Mercury released in chemical extractions of Palos Verdes surface sediments along 60-m contour and JWPCP effluent particulates (shaded bars)
have been a factor--the result was a lowering in the mass emission rate for mercury by 29%. However, the data do not show a greater decline for sediments at the perimeter of the sediment field than for sediments near the outfalls as would be expected. For the same reason, these results do not support the hypothesis that anaerobic conditions might act as the sole or major controlling factor for the metals. In short, no single explanation provides for the result obtained. We therefore propose that a combination of factors including the change in mass emission of mercury from the outfalls have been responsible for the observed decline. Until further studies are conducted, interpretations must be viewed with caution. If, as a point of departure, the residual concentrations of organic mercury in sediments are taken as being representative of the methylating activities of the indigenous microbiota, the organic mercury distribution may be interpreted. Previous work (21, 22) has shown that methylation rates in aquatic environments are profoundly influenced by temperature, pH, redox potential, sulfide ion activity, mercuric ion concentration, and organic matter concentration. In light of this, the rise in sediment concentrations of organic mercury with proximity to the outfalls may be due to increased organic matter. The absence of any such trends for other depths may either be due to the loss of volatile organomercurials during the sampling a t greater depths or, in the case of 30-m stations, to lower levels of mercury and organic matter which would tend to limit formation of methylmercury. The marked depression a t stations 6C and 7C may be related to some effect(s) of ,the re maining H2S field.Sincemercury in the sedimentat these stations was found to be highly refractory, and, hence, biologically unavailable, methylation would be unlikely. The presence of toxic H2S could also have had a debilitating effect on methylating microorganisms. Computation of the percent organic mercury in sediments along the 60-m isobath shows that the highest values are reached at the farthest stations (0,lO)where the percentage of mercury associated with labile organic matter (3% H202 leach) is at its peak. Hence, the conversion of inorganic mercury to organic forms is most effective on the perimeter of the sediment field. Mobilization of organomercurials to overlying water is unlikely owing to the probable presence of sulfhydryl-functional organic matter of sewage
Table IV. Mean Annual Flow, Total Suspended Solids, and Mercury Data on Joint Water Pollution Control Plant (JWPCP) Effluent, 1971-1975 a 1971
flow (mgd) total suspended solids (mg/L) total mercury (mg/L) particulate mzrcury (mg/dry Wb mass emission rate for total suspended solids (lo6kg/yr) mass emission rate for mercury (lo3kglyr)
371 325
1972
351 293
1973
359 258
1974
346 276
1975
341 278
0.0014 0.0011 0.0012 0.0011 0.0011 4.24
3.75
4.64
3.98
3.96
193.9
165.4
148.9
152.8
152.4
0.84
0.62
0.69
0.61
0.60
a Data from Los Angeles County Sanitation Districts annual summaries for Joint Water Pollution Control Plant Effluent. Calculated based upon mercury being 100% in the solid phase (cf. refs. 19 and 20).
origin for which methylmercury has a strong affinity ( 2 4 ) . Were mobilization of organic mercury occurring to any great extent, one would expect to see evidence of enhanced uptake in the tissue of indigenous benthic animals (25).Data obtained so far indicates that this is not occurring (26). The phase partitioning experiment showed that the majority of the mercury in near-outfall sediments and effluent particulates is bound in a refractory phase. With increasing distance from the outfalls, mercury becomes distributed among progressively more degradable organic phases. Two questions emerge: Do the observed phase distribution patterns reflect simple mixing of sewage particulates with natural particulate matter, alteration of the sewage particulates during and after deposition, or a combination of these effects? If significant alteration of effluent solids does occur, to what extent do postdepositional changes contribute to the overall process? Volume 12,Number 10,October 1978
1155
I
-s v)
0 1
70
60
-
I
I
X JH?CP EFFLUENT 11975) JwPcP EFFLUENT 11974)
I
I
I
I
b
-
POSSIBLE DIAGENETIC PATHWAYS
A
4C
0 v)
w
-I--I Q
-I
0
> -l
-
Q IO I-
-
T O T A L M E R C U R Y (mg/dry kg) Figure 6. Total volatile solids vs. total mercury concentration for Palos Verdes sediments, sediment trap particulates,and JWPCP effluent partic-
ulates Open symbols in 6b represent 1974 trap particulates and solid symbols are for 1975 surface sediments
Some clarification is provided by the work of Sweeney e t al. (27) who have developed a graphical model for the analysis of outfall-affected sediments. The method relies on the use of two properties that can be used as tracers for sewage, one of which must be conservative with respect to sewage particulate matter. In the present case, we will consider the concentration of total mercury, [Hg], and total volatile solids, [TVS],as the two properties of interest. If natural particulates and sewage were simply mixed, one could construct a line on a plot of [TVS] vs. [Hg] which represented all possible proportions of naturalhewage particulate mixture compositions. Line A in Figure 6a demonstrates this idea. Here the endmembers are composed of JWPCP sewage particulates and a suitable natural sediment sample. In essence, this is an ideal mixing curve for these two properties, and any deviation from this line by a sediment sample represents an enrichment or loss of one component relative to the other. Consider now the hypothetical degradation of pure sewage particulates, that is, loss of organic matter (TVS). If this was carried out without loss of mercury, for example, a particulate could be generated whose [TVS] and [Hg] coordinates would fall on line B in Figure 6a; its exact position would depend on the degree of degradation effected. (Note: Line B is not a vertical line because there is a loss of mass in the degradation process, causing the mercury concentration to increase.) Alternatively, if one were to specify a degree of degradation for the particulate organic matter (without loss of mercury), a line could be formed that would be the locus of points for mixed natural/sewage particulates which have undergone a given degree of degradation without loss of mercury. An example is the hypothetical line labeled rf = 0.75 in Figure 6a. Here, the value of 0.75 refers to the undegraded fraction of the mixed particulate organic matter. The significance of this particular line derives from the experiments of Myers (5) who found the limit of reduction in organic carbon values for JWPCP effluent particulates exposed to seawater to be 25% after 6 weeks. Consequently, this line represents a “revised” ideal mixing curve where the refractory organic matter is now being used as the conservative component. By simply plotting the data for sediments from the Palos Verdes shelf, one can visualize the apparent behavior of mercury relative to the refractory organic matter and, thus, assess whether it is enriched, lost, or conserved. Inspection of the 1975 sediment data reveals that most points fall below the line A, and in the case of sediments containing the highest mercury concentrations, 1156
Environmental Science & Technology
the points are even below the hypothetical line for rf = 0.75. A linear regression analysis of the sediment data yields line C which intercepts with the degradation curve (line B)a t an r f = 0.58. Interpretation of these results demands caution because the sediment data reflect a net result of sedimentation, scouring, diagenisis, and bioturbation. The data do suggest the 1975 sediments were enriched with mercury relative to the organic matter. Comparison of sewage particulates, trapped particulate material, and deposited sediments allows some speculation as to the pathway taken by particulate-associated mercury discharged from the outfalls. Figure 6b illustrates analytical results for these samples. The data for trapped particulates generally fall close to the ideal mixing curve, and all of them lie above the rf = 0.75 line. We can conclude that, while they are suspended in the water column, sewage particulates mix with natural sedimenting particles and either experience very little loss of organic matter and mercury, or release both in near-constant proportion. In contrast, the sediment data fall to the right and below the sediment trap data for each respective station. This indicates that settling particulates undergo substantial changes once they become incorporated into the bottom sediments. Hypothetical two-step paths leading from the trap particulates to sediments can be envisioned in which organic matter is degraded in the first step, and mercury is accumulated in the second. Paths of this type drawn in Figure 6b indicate that sediments closer to the outfalls are apparently more enriched in mercury than those away from the outfalls by comparison with the suspended trap particulates from which they were presumably derived. These results demonstrate that mercury is effectively fixed in a refractory phase of the Palos Verdes sediments, actual enrichment of mercury increasing with proximity to the outfalls. A similar enrichment was noticed previously for Cd and P b by Morel et al. ( 7 ) ;however, Figure 6b shows that for Hg, flocculation of Hg-rich particulates of nonsewage origin does not play an important role in the enrichment process. Rather, postdepositional changes dominate. One cannot discount the possibility that the formation of H2S in anaerobic regions may have caused some enhanced stripping of mercury (and certain other metals) from overlying waters and precipitation of the insoluble sulfide(s). Under the conditions of a contracting sulfide field, enrichment would thus be greatest for sediments that produced H2S for the longest interval; these sediments are located in the immediate outfall area.
Conclusions The present results demonstrate that a small decrease in surface sediment concentrations of mercury off Palos Verdes has occurred during 1972-1975. If genuine, this change may have involved several factors including the reduced output of effluent solids. There is a lack of correlation between the Hg concentration decreases and the sediment location. This finding is inconsistent with the hypothesis that metal mobilization is related primarily to the occurrence of anaerobic sediment conditions. Quantities of organic mercury up to 21.3 pg/dry kg were also found in these sediments. Highest concentrations were recovered near the outfalls; however, little organic Hg was detected in the immediate vicinity of the outfall diffusers where surface sediments still produce H2S.The proportion of organic mercury produced (relative to total mercury) seems to be related to the availability of mercury in degradable organic phases. Consequently, the highest organic mercury percentages occurred at the perimeter of the sediment field where the proportions of mercury bound in the most labile organic phases were largest. Mobilization does not appear to be biologically significant in view of the low levels found in indigenous benthic organisms ( 2 5 ) . The chemical leaching experiments showed that a large proportion of the mercury in Palos Verdes surface sediments and effluent solids is contained in a refractory phase and is not principally associated with labile organic matter. This may explain the chronic uptake response in tissues of M . californianus which were exposed to Palos Verdes waters for 24 weeks (28).With distance from the outfalls the phase distribution patterns increasingly favor partitioning of mercury in the chemically degradable organic phases. In the immediate vicinity of the outfalls, however, all of the mercury is associated with the refractory phase. By simply comparing the effluent solids, settling trap, and surface sediment data, it becomes apparent that the sewage particulates undergo some alteration prior to incorporation into the sediments. With deposition, there is additional destruction of organic matter and some enrichment of mercury, this latter effect intensifying with proximity to the outfalls. The enrichment appears to be a diagenetic feature and not related primarily to the sedimentation process. I t may be due to scavenging of mercury from overlying water by sulfide produced in outfall sediments; however, this is speculation. In general, this investigation demonstrates the utility of relatively simple geochemical methods in studying the potential for biological utilization and/or uptake of anthropogenic metals in coastal areas. Acknowledgment We gratefully acknowledge D. Hotchkiss and the crew of
the Sea-S-Dea and also P. Hershelman, N. Kettenring, H. Schafer, and R. Sweeney. Literature Cited (1) Klein, D. H., Goldberg, E. D., Enuiron. Sci. Technol., 4, 765 (1970). (2) Galloway, J. N., PhD thesis, University of California, San Diego, Calif., 1972. (3) “The Ecology of the Southern California Bight: Implications for Water Quality Management”, Southern California Coastal Water Research Project Rep., Tech. Rep. 104, Mar. 1973. (4) Bruland, K. W., Bertine, K., Koide, M., Goldberg, E. D., Enuiron. Sci. Technol., 8,425 (1974). (5) Myers, E. D., PhD thesis, California Institute of Technology, Pasadena, Calif., 1974. (6) Hendricks. T. J.. Younn. D. R.. “Modeline the Fates of Metals in Ocean-Discharged Wastewaters”, Southern California Coastal Water Research Project Rep., Tech. Mem. 208, Jan. 1974. ( 7 ) Morel, F.M.M., Westall, J. C., O’Melia, C. R., Morgan, J. J., Enciron. Sci. Technol., 9, 756 (197?). (8)Jones, J. H., “General Circulation and Water Characteristics in the Southern California Bight”, Southern California Coastal Water Research Project Rep., Tech. Rep. 101, 1971. (9) Young, D. R., McDermott, D. J., Heesen,T. C., J a n , T . K., “Pollution Inputs and Distribution Off Southern California”, in “Marine Chemistry in the Coastal Environment”, T. M. Church, Ed., p 424, American Chemical Society, Washington, D.C., 1975. (10) Kalil, E. K., “Rapid Pore Water Analyses for Sediments Adjacent to Reactor Discharges”, International Atomic Energy Agency Publ. 180/37, 1974. (11) Eganhouse, R. P., “The Measurement of Total and Organic Mercury in Marine Sediments, Organisms, and Waters”, Southern California Coastal Water Research Project Rep., Tech. Mem. 221, Sept. 1975. (12) Hatch, W. R., Ott, W. L., Anal. Chem., 40, 2085 (1968). (13) Bisogni, J. J., Lawrence, A. W., Enuiron. Sci. Technol., 8, 850 (1974). (14) Walters, L. J., Wolery, T. J., “Transfer of Heavy Metal Pollutants from Lake Erie Bottom Sediments to the Overlying Water”, prepared for the Office of Water Resources Research, Jan. 1974. (15) Gupta, S. K., Chen, K. Y., Enuiron. Lett., 10, 129 (1975). (16) “Standard Methods for the Examination of Water and Wastewater”, 14th ed., Part 208E, p 95, 1975. (17) Olson, B. H., Cooper, R. C., Nature, 252,682 (1974). (18) Andren, A. W., Harriss, R. C., ibid.,245, 256 (1973). (19) Chen, K. Y., Young, C. S., Jan, T. K., Rohatgi, N., J . Water Pollut Control Fed., 46, 2663 (1974). (20) Lingle, .J. W., Hermann, E. R., ibid., 47,466 (1975). (21) Bisogni, J. J., Lawrence, A . W., ibid., p 135. (22) Langley, D. G., ibid., 45, 44 (1973). (23) Billen, G., Joiris, C., Wollast, R., Water Res., 8, 219 (1974). (24) Reimers, R. S., Krenkel, P. A,, J . Water Pollut. Control Fed., 46,352 (1974). (25) Jernelov, A , , Landner, L., Larsson, T., ibid., 47,810 (1975). (26) Eeanhouse. R. P.. Younn. D. R.. Bull. Enuiron. Contam. Toxicol.. in press (1978). (27) Sweeney, R. E., Kalil. E. K., K a d a n . I. R., Marine Chem.. in press (1978). (28) Eganhouse, R. P., Young, D. R., Mar. Pollut. Bull., in press (1978).
-
Received for reuieu; August 9, 1977. Accepted April 27, 1978
Volume 12, Number 10, October 1978
1157
