- Email: [email protected]
Sewage sludge impact on sediment quality and benthic assemblages off the Mediterranean coast of Israel—a long-term study
Marine Environmental Research 57 (2004) 213–233 www.elsevier.com/locate/marenvrev
Sewage sludge impact on sediment quality and benthic assemblages off the Mediterranean coast of Israel—a long-term study N. Kress*, B. Herut, B.S. Galil Israel Oceanographic and Limnological Research, National Institute of Oceanography, PO Box 8030, Haifa 31080, Israel Received 6 January 2003; received in revised form 10 July 2003; accepted 28 July 2003
Abstract The distributions of benthic assemblages, heavy metals and organic carbon (Corg) in sediments were examined during a long-term study at a sewage sludge disposal site off the Mediterranean coast of Israel. The disposal of sewage sludge has a marked but localized, seasonally dependent, impact on the benthic assemblages and sediment quality. Elevated concentrations of Corg, Hg, Cd, Cu, Zn, Pb, and to a lesser degree Ni in the sediments were detected mostly northward of the sewage outfall, in the direction of the prevalent longshore current. High concentrations of Corg and metals were reflected by elevated populations of tolerant and opportunistic polychaetes in spring and by an azoic zone in fall. The impacted area extended mainly towards the north (up to ca. 4 km) and to a lesser extent south of the outfall (up to ca. 2.5 km). No evidence of increased accumulation of sewage sludge with time was found, nor of pollutants associated with it. Principal component analysis (PCA) grouped the anthropogenic metals and Corg with infaunal abundance for the spring surveys, while biotic diversity was negatively correlated with the pollutants. In the PCA of fall surveys, abundance was negatively correlated with the pollutants, decreasing with increased concentration of Corg and anthropogenic metals. We suggest that the seasonal pattern shown by infaunal abundance, anthropogenic metals and Corg is due to the stratification of the water column from spring to fall on one-hand and winter storms on the other. Winter storms resuspend and disperse the fine organic particles, sweeping the site clean of sludge; accumulation of sludge takes place throughout the quiescent periods of the year, when stratification is reestablished. The disposal site is dispersive and the spatial extent of the impacted area varies seasonally and interannually. This monitoring study, in addition to addressing specific questions about sewage sludge impact, represents an unusually large
* Corresponding author. Tel.: +972-4-8565256; fax: +972-4-8511911. E-mail address: [email protected] (N. Kress). 0141-1136/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0141-1136(03)00081-3
214
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
and unique set of long-term measurements that will serve as a basis to evaluate the site recovery following the cessation of disposal. # 2003 Elsevier Ltd. All rights reserved. Keywords: Sewage sludge; Marine disposal; Heavy metals; Organic carbon; Benthic assemblages; Pollution; Eastern Mediterranean
1. Introduction Sewage sludge (also called biosolids) is generated mainly by the treatment of domestic waste waters. It consists primarily of organic biomass and nutrients (e.g. nitrogen, phosphorus, potassium) and may contain potentially harmful constituents such as pathogenic organisms, metals and synthetic organic compounds (US EPA, 1995; Renner, 2000). Sewage sludge can be utilized in land applications as fertilizer or soil conditioner, disposed of inland by landfilling or incineration, or disposed of in the aquatic environment by direct discharge or dumping. In the latter case, the main environmental effects are associated with the enrichment of the receiving habitat with organic matter and nutrients, accumulation of pollutants, their introduction into the food web, and changes in the biotic diversity and abundance. The treatment of domestic wastewaters in Israel generates 495,000 dry tons of sewage sludge annually, 40% of which are utilized in agriculture or disposed in landfills (D. Rubin, Ministry of the Environment, personal communication). The Dan Region Wastewater project alone, which treats the domestic sewage of the 1.5 million inhabitants of the Tel-Aviv Metropolitan area, produces 292,000 dry tons. The excess sludge produced at this plant (16,000 m3 day1) has been discharged, since 1987, through a single outfall, 5 km offshore, from a 1600 HDPE seabed pipeline placed nearly perpendicular to the shore (Fig. 1). The sludge contains approximately 1% particulate matter with metal mass emission rates (in kg year1) of: Hg— 60; Cd—430; Pb—1670; Cr—11400; Cu—19000; Zn—54000; Ni—2500 (UNEP/ WHO, 1999). Recently, reduction of metal load at the source resulted in halving the annual loads of Cd and Cr (D. Salomon, personal communication). Disposal of sewage sludge in the marine environment has been practiced globally. It affects the seafloor and its biota: contaminated sediments display increased levels of organic matter, nutrients and heavy metals, their biota are either greatly impoverished, or consist of large numbers of few opportunistic species (Bothner, Buchholtz ten Brink, & Manheim, 1998; Costello & Read, 1994; Norton & Champ, 1989; Pearson, 1987; Rodger, Davies, & Moore, 1991; Rodger, Davies, & McHenery, 1992; Studholme, O’Reilly, & Ingham, 1995; among others). In most studies, the effects of sewage sludge disposal were examined either in sediment quality parameters or in benthic assemblages, separately. Marine disposal of sewage sludge was terminated in 1992 in the USA as a result of the Ocean Disposal Ban Act of 1988 (O’Connor, 1998) and in the European Union in 1998 under the Urban Waste Water Treatment Directive (91/271/EEC). Recently, concerns were raised on the environmental safety of land application of sewage sludge such as long-term buildup of heavy metals in the soil and the presence of pathogens (Renner, 2000, Lewis & Gattie, 2002).
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
215
Fig. 1. Map of the study site. Filled circles designate sampling stations. Not shown is station 58, the reference station, located 4500 m south of the outfall.
216
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
This paper will: (a) present the environmental conditions at the disposal site after 15 years of sewage sludge discharge: heavy metals and organic carbon concentration in the sediments, composition of the benthic assemblages, spatial and temporal variations, (b) evaluate the possibility of long-term accumulation of the sewage sludge at the disposal site, and (c) combine chemical and biological variables to assess the impact of the sewage sludge on the oligotrophic environment of the eastern Mediterranean.
2. Study area The sublittoral zone off the Mediterranean coast of Israel slopes gently off shore, with isolines parallel to the shore. The Levantine basin of the eastern Mediterranean is ultra-oligotrophic (Berman, Townsend, El Sayed, Trees, & Azov, 1984; Kress & Herut, 2001); the water column at the continental shelf is mixed during winter (November–February), and stratified the rest of the year (2–2.5 sigma t units from surface to bottom) (Rosentraub, 1990). Surface water temperatures show high seasonal variations and range from 15 C during winter to 30 C in summer. The prevalent northwards current skims the shore, its average velocity less than 15 cm sl, decreasing with increasing depth in summer due to stratification, and increasing at times in winter to 90 cm s1 (Rosentraub, 1990; Rosentraub, pers. comm.). The near-bottom current has a slight westwards direction. The disposal site is located south of Tel Aviv, at depth of 38 m (Fig. 1), where the substrate was, prior to the outfall activation, mostly sand (65%) with silt (21%) and clay (14%) (Galil & Lewinsohn, 1981). A survey conducted in 1978 revealed no aberrant conditions and indicated that the outfall area was unpolluted at the advent of discharge (Galil & Lewinsohn, 1981). A monitoring program that examined the water quality, sediments and benthic fauna at a distance of 1 km north of the outfall detected no effects in 1987–1992. In 1992, 5 years after the start of discharge operations, the Ministry of the Environment authorized an expanded monitoring program that is in effect to the present (Kress & Hornung, 1996). It should be mentioned that no studies have been conducted on the effects of sewage sludge in the Mediterranean Sea except for a single survey off Barcelona (Palanques, Diaz, & Maldonado, 1991; Palanques, 1994).
3. Methods 3.1. Field sampling Twenty sampling stations were established in 1992, arrayed along two lines intersecting at the outfall, at distances of 50, 100, 200, 500 and 1000 m from it (Fig. 1). The north–south (N–S) transect runs along the 38 m isobath, subparallel to the coastline (015 ), while the east–west (E–W) transect, is subperpendicular to the shore, at depths of 35–39 m. Throughout the years the transect formation remained unaltered, though disposition of stations changed to reflect the affected area, with the farthest station at 6 km northward of the outfall. The sampling program consisted
217
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
of semiannual sampling (spring and fall) on even-numbered years, and annual sampling (fall) on odd-numbered years, allowing for the detection of biotic seasonal patterns (Galil & Lewinsohn, 1981). Sediment samples for the chemical analysis were taken in spring on even-numbered years, and in fall on odd-numbered years. This study reports the results of the surveys conducted from spring 1992 to spring 2002. At the beginning of the monitoring program, one reference site, situated 4.5 km south of the outfall was established with the rational that the peripheral stations, 1 km from the outfall (stns 5, 10, 15 and 20, Fig. 1) would reach natural levels and serve as reference stations as well. As the program evolved and the data was analyzed, it turned out that the initial assumption was correct for the eastern and western stations only (stns 10 and 20, respectively). Therefore, new peripheral stations were added mainly towards the north and formed a gradient from polluted to unaffected stations, increasing the data on the background character of the area. Consequently, data from the reference station and the peripheral background stations from all surveys were collected and served to determine the natural background of the area (Table 1). Support to these values is found in a related study (Kress, Shoam-Frider, & Shelef, 2002) that showed that the concentrations of Hg and organic carbon (Corg) in sediment cores from station 3 reached the background concentrations given in Table 1 at 15 cm sediment depth. Table 1 Descriptive statistics of Corg, heavy metal concentrations in surface sediments and macroinfaunal abundance at the disposal site in spring and fall and the natural background values in area Hg
Cd
Pb
Cu
Zn
Ni
Mn
1
mg g
0.04 0.01 25
Fe
Al
wt.%
Spring (n=134, Al and Corg, n=112) Minimum 0.02 bdl 1.66 9.40 Maximum 1.12 2.90 38.9 183 Median 0.16 0.42 16.4 36.9 Average 0.25 0.53 16.4 52.4 Stdev 0.24 0.56 7.32 37.9 CV(%) 95 104 45 72 Fall (n=113) Minimum 0.01 bdl 3.25 13.7 Maximum 1.36 3.12 43.8 242 Median 0.18 0.49 15.6 51.8 Average 0.28 0.81 16.7 73.9 Stdev 0.26 0.83 9.63 58.0 CV(%) 85 102 58 78 Natural background (n=12) Average Stdev CV(%)
Corg
0.07 0.06 85
8.79 5.12 58
21.4 6.10 29
30.0 586 97.2 149 120 80
10.8 67.9 33.9 35.4 9.94 28
257 935 631 641 94 15
0.18 10.3a 0.95 1.59 1.65 104
1.67 6.37 3.53 3.64 0.92 25
2.98 7.73 4.70 4.83 0.84 17
37.1 668 144 209 166 79
15.9 61.9 39.4 38.9 10.1 26
286 996 616 618 108 17
0.24 8.59 1.29 1.75 1.49 85
2.37 5.78 3.67 3.77 0.73 19
2.83 7.05 4.90 4.89 0.74 15
55.5 14.8 27
29.1 9.90 34
727 144 20
0.47 0.17 36
3.72 1.13 30
4.53 0.78 17
Abundance no of specimens (n=59) 224 13674 1702 2356 2345 100 (n=107) 0 458 12 55 84 152 Spring (n=22) 472 167 35
Fall (n=28) 251 104 41
Stdev—standard deviation, CV—coefficient of variance, n=number of data points, bdl—below detection limit ( <0.01 mg g1). a Without two high data points (10.3 and 9.8 wt.%), the average, stdev and median are 1.41, 1.12 and 0.96 wt.%, respectively.
218
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
A triplicate sample was collected at each station by a 0.062 m2 box-corer with an effective penetration of 40 cm (Ocean Instruments model 700 AL). A sample for chemical analysis was skimmed from a small area of the top 1 cm of the core surface of the first triplicate samples and frozen. The top 15 cm of the sediment was sieved through 0.5-mm mesh aboard ship, and preserved in 10% buffered formaldehyde. The amount of sediment sampled for the chemical analysis was minimal and did not cause a loss of organisms, shown by the good agreement of the results of the triplicate samples. An additional box-core sample was taken at selected stations along the N–S transect in order to assess sludge penetration. Sampling frequency and methods have remained constant since implementation of the improved monitoring program in 1992. 3.2. Chemical analyses in the laboratory The frozen sediment samples were lyophilized for 48 hr and then dry sieved through a 1000 mm sieve to extrude extraneous components such as seeds, broken shells, etc. The sediment was then digested with a mixture of hydrofluoric acid and aqua-regia as described in ASTM (1983) for the determination of Al, Fe, Mn, Cu and Zn. Concentrated nitric acid was used to digest the sediments for the determination of Cd, Pb, and Ni (Hornung, Krom, & Cohen, 1989). A separate digestion was performed in concentrated nitric acid for the determination of Hg. Metals in the digest solutions were analyzed by atomic absorption spectrometry, using an IL-951 or a Perkin–Elmer 1100 B spectrophotometer equipped with flame and graphite furnace modules. Only mercury was analyzed by cold vapor atomic absorption spectrometry on a Coleman Mercury Analyzer MAS-50. Quality control and quality assurance of the results was performed with standard reference materials from the USA National Institute of Standards and Technology (NIST—Estuarine Sediment— 1646, River Sediment—1645) and the National Research Council of Canada (NRCCMESS-2). The standards were digested and analyzed in the same manner as the samples, with each analytical run. All metals gave results within 5% of the certified values. Organic carbon (Corg) was determined by the potassium dichromate method following the procedure of Gaudette, Flight, Toner, and Folger (1974) and Avnimelech (1989) with slight modifications. All concentrations are reported on a dry weight basis. 3.3. Biological analysis in the laboratory The formalin-preserved biotic samples were washed, resieved thoroughly on a 0.5 mm sieve and preserved in 70% ethanol within days of sampling. Prior to sorting, the samples were stained in Rose Bengal, and the extracted organisms were identified and counted. The total number of individuals—Abundance—in the sample (0.0093 m3) is given as the average of the triplicate samples, and the ShannonWiener diversity index- H0 - (Shannon & Weaver, 1949) was calculated. Diversity indices are commonly used to assess pollution impact, where a high value denotes an unpolluted sample and a low value—a polluted one. However, statistically significant reductions in a diversity index are attended by great changes so that the index is considered a poor indicator (Bayne, Clarcke, & Gray, 1988), besides, the
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
219
interpretation of diversity indices is made difficult by natural periodic environmental changes that may produce changes that counteract or parallel those brought about by pollution. Although aware of the limitations in the use of diversity index we present it here to aid in the interpretation of the results and to compare with other sites reported in the literature. 3.4. Statistical analysis Statistical analyses were performed using the SPSS software with the procedures of general linear model (GLM) using the method of least squares, the t-test unequal variance, and Principal component analysis (PCA). Significance of < 0.05 was used.
4. Results The natural sediment in the area consists of grey-brown sandy-mud, easily distinguished from the black-colored, malodorous deposited sewage sludge sampled at the vicinity of the outfall. The depth of the sludge layer was estimated from sediment cores taken at stations northward of the outfall and ranged from no obvious sludge (i.e. 0 cm penetration) at the most distant stations (4–6 km) to 3–17 cm thick layer at the outfall area. In spring 2000, for example, the sludge layer was 17 cm deep at the outfall (station 0), 15 cm at station 3 (200 m north of the outfall), whereas 1 km north of the outfall (station 5), a layer 6.5 cm thick of sludge was covered by a 2 cm layer of sediment. No sludge was evident at station 24 located 3 km northwards of the outfall. 4.1. Heavy metals and organic carbon The ranges of Hg, Cd, Cu, Zn, Pb, Ni, Mn, Fe, Al and Corg concentration in the sediment in spring and fall, are summarized in Table 1, as well as the natural background concentrations for the area. Enrichment factor (EF), used as an index of the sewage sludge effect, is the ratio of a sample’s metal concentration to background concentration. Since water depth and sediment composition within the study area are similar throughout, there was no need for grain size normalization. Cd had the highest EF value, followed by Hg, Zn, Corg and Cu (Table 2). Pb was enriched mainly at the stations within 100 m radius from the outfall and northwards up to 1500 m from it, with a maximal value of 2.8. Mn EF values were lower than 0.9 at stations close to the outfall, indicating depletion due to formation of soluble species (Hemond & Fechner-Levy, 2000). Fe and Al, considered natural for the area, exhibited no enrichment. EF values for Ni were under 1.4, indicating that Ni was mostly natural for the area. T-test comparison between spring and fall values of Ni and background levels showed that, in spring, Ni was natural for the area (not significantly different from the background values, P one tail=0.107), however in the fall samples it was significantly higher (P one tail=0.00016) than the background concentration. In addition to the EF analysis, we compared the data set to the sediment quality guidelines, suggested by Long, MacDonald, Smith, and Calder (1995), based on
220
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
their potential to induce toxic effects in marine organisms [ERL (Effects Range Low) and ERM (Effects Range Median)]. Only Hg and Zn concentrations exceeded ERM in some of our samples (Table 3). The maximal values of Cd and Cu exceeded ERL values, while Pb concentration in all samples was lower than ERL. The natural background levels of Ni were higher than ERL values, therefore these criteria are not applicable for Ni in the study area. One of the main tracers of the presence and effects of sewage sludge in the marine environment is enrichment by organic matter, represented in this study by Corg. There was a marked positive relationship between Hg, Cd, Cu, Zn, Pb, Ni and Corg indicating the sewage sludge as the anthropogenic source of these metals to the environment (Table 4). Although no deviations from the natural background were detected for Mn it showed a negative trend with Corg, probably due to the reduction of MnO2 and formation of soluble species (Mn+2) at anoxic environments, as those Table 2 Average enrichment factors (EF) for the sampling stations Station
Corg
Hg
Cd
Pb
Cu
Zn
Ni
Mn
0 1 2 3 4 5 21 22 23 24 26 55 56 6 7 8 9 10 11 12 13 14 15 28 16 17 18 19 20 Maximum
3.2 5.7 4.8 4.9 4.4 4.3 3.3 2.4 1.5 1.2 1.0 1.2 1.7 3.7 2.7 1.9 1.7 0.9 3.0 2.1 2.9 2.3 1.6 1.7 3.0 3.2 2.3 1.2 1.3 5.7
6.5 11.3 9.8 12.0 15.2 13.0 8.8 9.2 4.6 2.3 5.3 2.2 4.9 9.8 8.8 4.6 2.8 1.4 4.9 7.9 5.5 4.7 6.5 1.9 8.0 11.7 6.0 3.0 1.3 15.2
9.3 18.4 15.7 14.5 16.7 16.2 11.7 8.5 7.2 2.7 0.9 1.0 2.3 12.5 8.1 4.9 3.2 1.2 8.0 7.9 6.6 5.4 3.0 1.4 10.2 10.7 6.5 2.2 1.3 18.4
1.6 2.8 2.4 2.3 2.3 2.3 1.9 1.7 1.4 1.2 0.9 0.9 1.1 2.3 1.8 1.4 1.3 1.1 1.8 1.8 1.6 1.4 1.3 0.9 2.0 2.1 1.5 1.3 1.3 2.8
2.7 5.6 4.5 4.6 4.5 4.8 3.9 3.2 2.2 1.4 0.9 1.0 1.6 3.6 2.6 1.7 1.4 0.9 2.4 2.3 2.2 1.9 1.5 1.4 2.8 3.1 2.3 1.3 1.3 5.6
3.4 6.1 5.2 5.5 5.5 5.1 4.0 3.1 2.2 1.4 1.0 1.1 1.7 4.7 2.7 2.0 1.5 0.9 2.8 2.6 2.6 2.0 1.6 1.3 3.3 3.5 2.5 1.3 1.2 6.1
1.2 1.3 1.2 1.3 1.3 1.3 1.2 1.2 1.0 1.0 0.7 0.8 1.0 1.4 1.2 1.0 1.0 0.8 1.2 1.1 1.1 1.0 1.0 1.1 1.1 1.1 1.0 1.1 1.2 1.4
0.82 0.72 0.72 0.78 0.78 0.79 0.80 0.83 0.80 0.80 0.75 0.78 0.90 0.82 0.83 0.82 0.83 0.83 0.80 0.83 0.82 0.82 0.88 0.94 0.78 0.81 0.86 0.96 1.09 1.1
Fe and Al concentrations were natural for the area.
221
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
found in the sediments with high organic loads (Hemond & Fechner Levy, 2000). Fe and Al, not considered pollutants in the area, showed no relationship to Corg (Table 4). The concentrations of Hg, Cd and Cu, representing the anthropogenic metals, along the south north transect across the outfall are shown in Fig. 2. There was a preferential spread (i.e. higher concentrations) northwards, in the direction of the average alongshore current. The concentrations reached background levels about 3 km north, and 1.5 km south of the outfall. The maximal values, their position within the affected region, and the distribution pattern varied with season and year (Fig. 2). The areal spread of the sewage sludge, as illustrated by Cd concentrations in spring 98 and fall 99, is depicted in Fig. 3. The sewage sludge was dispersed in the direction of the average longshore current, with limited east–west dispersion (perpendicular to the coast). The maximal concentrations were not centered at the outfall but slightly northward, consistent with the Table 3 Comparison of sediment quality criteria (Long et al., 1995) to concentrations found in the study area (Concentrations are in mg g1 dry wt.)
Cd Cu Pb Hg Ni Zn a b
ERL
ERM
Maximal concentration in the area
%a between ERL and ERM
%b above ERM
Background concentration in the area
1.2 34 46.7 0.15 20.9 150
9.6 270 218 0.71 51.6 410
3.12 242 43.8 1.36 67.9 668
17 63 0 47 – 33
0 0 0 7 – 9
0.07 21.4 8.79 0.04 29.1 55.5
Percentage of total data points (n=246) whose concentrations are between the ERL and ERM values. Percentage of total data points whose concentrations are above the ERM.
Table 4 R2 and linear regression parameters of heavy metals vs organic carbon
Hg Cd Pb Cu Zn Ni Mn Fe Al
R2
Slope
Intercept
Significance Fa
0.5375 0.6089 0.3046 0.5628 0.6730 0.3843 0.3472 0.00075 0.00016
0.119 0.360 3.078 24.18 77.62 4.027 38.24 0.0142 0.00633
0.0734 0.0779 11.198 24.633 53.77 30.53 693.29 3.75 4.88
7.22 E-39 3.8 E-47 2.9 E-19 9.29 E-42 8.56 E-56 3.56 E-25 2.5 E-22 0.683b 0.851b
Number of observations =224. a ANOVA b Not significant
222
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Fig. 2. Distribution of Cd, Hg and Cu (mg g1 dry wt.) in spring and fall along the south–north transect, from station 58 to station 61. Negative and positive distances are southwards and northwards of the outfall, respectively.
model describing the flux of settling particles out of a plume discharged from a submerged ocean outfall (Hunt, 1990). The other anthropogenic metals (Hg, Cu, Pb, Zn) were similarly distributed. 4.2. Macroinfaunal abundance and diversity The macroinfaunal abundance values for the stations along the south–north transect across the outfall, together with Corg, are shown in Fig. 4 for all the surveys. Since all stations are isobathic, the effect of depth on the infaunal community structure is annulled. We chose to present the results from stations along the south-
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
223
Fig. 3. Surficial distribution of Cd (mg g1 dry wt.) and abundance (Abn, number of specimens) at the disposal site in May 1998 and September 1999. Filled circles designate sampling stations.
224
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
north transect, because of the preferential northwards spread of the sludge. Infaunal abundance was high in the spring and low in the fall surveys (Fig. 4). In May 1992 high abundance was found at the stations in the immediate vicinity and those northwards from the outfall: 265 organisms were collected at station 4500 m northward of the outfall—higher than at the other stations. Abundance values in May 1992 were much lower than in the subsequent spring surveys, probably due to the unusual cold and stormy winter of 1991–1992. In May 1994—4238, 2251, and 406 organisms were collected at stations 4,5 and 22, respectively (Figs. 1 and 4). In May 1996, 3258, 1089 and 1371 organisms were collected at stations 4,5 and 22, respectively. In May 1998, June 2000 and May 2002, 1627, 2419 and 1961 organisms, respectively, were collected at station 22. In the fall surveys those same stations, located near the outfall, were nearly devoid of life (Fig. 4). In November 1992 only three organisms were found at station 4 and four organisms at station 5. Infaunal abundance values for the fall samples maintained the pattern of a nearly azoic zone in the immediate vicinity and northwards from the outfall, enclosed by a relatively high-abundance zone. The results for the fall of 1995 and two stations in fall 1998 were exceptional. The relatively high abundance values in 1995 resulted from the accidental rupture of the outfall pipe that reduced the amount of sludge discharged at the outfall mouth. However, even the high abundance in fall 1995 was lower than in
Fig. 4. Distribution of Corg (wt.% dry wt.) and abundance (number of specimens) in spring and fall along the south–north transect, from station 58 to station 61. Negative and positive distances are southwards and northwards of the outfall, respectively. Please note different abundance scale for spring and fall.
Distance from outfall
Station number
4500 1000 500 200 100 50 0 50 100 200 500 1000 1500 2000 3000 4500 5500
58 15 14 13 12 11 0 1 2 3 4 5 21 22 24 26 29
May 92
Nov 92
Oct 93
May 94
Nov 94
Sept 95
1.320 0.711 0.736 1.430 1.530
1.750 1.070 0.996 1.330 1.561
1.680 1.640 1.490 0.957 0.778
2.460 2.490 0.523 0.894 0.317
0.828 0.562 1.430 1.360 0.684
2.020 0.985 1.310 0.468 0.470
1.390 1.940 1.670 1.490 1.510
1.210 1.010 0.637 1.730 1.770
1.040 0.580 0.918 0.693 0.971
0.490 0.037 0.095 0.072 0.223 1.180 2.520
1.590 1.320 1.350 1.030 1.400 0.356 1.380
0.439 0.506 0.183 0.500 1.030 0.913 0.681 1.948
Negative and positive distances are southwards and northwards of the outfall, respectively.
May 96
Oct 96
Sept 97
May 98
Sept 98
Sept 99
2.190 0.873 0.218 0.163 0.129
2.050 1.660 1.470 1.080 0.939
2.160 1.910 1.120 0.902
2.130 0.769 0.696 0.301
2.180 1.450 1.150 1.099
1.590 0.528 1.280 0.000
0.146
1.030
1.050
0.166
0.693
0.995
0.346 0.486 0.214 0.205 0.331 0.678 1.930 2.450
0.561 0.775 0.796 1.470 1.040 0.975 1.110
1.350 1.250 0.857 1.650 1.230 1.740
0.420 0.374 0.263 0.354 0.524 1.290 2.050
0.171 0.093 0.206 0.114 0.476 0.792 1.910
1.210 0.000 0.000 1.050 0.685 0.901 2.080
2.060
1.823
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Table 5 Diversity index for the stations along the south–north transect from spring 1992 to fall 1999
225
226
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
the spring surveys. It seems that the two high values in the fall of 1998 (stations 21–22) were due to the transition between the azoic zone to the super abundant zone. This is supported by the lower diversity found at those stations compared to the stations further away from the outfall (Table 5). Background abundance was seasonally dependent, significantly higher in the spring than in the fall (P < 0.0001, Table 1). Infaunal diversity is presented in Table 5. In May 1992 diversity was slightly higher than in the following spring surveys, except for three stations in May 1994 (stations 14, 15 and 22, Fig. 1, Table 5). Diversity values at stations 2 and 3, in close vicinity of the outfall, were 1.94 and 1.67, respectively, higher than in May 1994 (0.037, and 0.095 at stations 2 and 3, respectively). The samples collected in May 1996, May 1998 and June 2000 exhibited low diversity values (< 1.0) at the stations up to 1000 m southwards, and 2000 m northwards from the outfall. Infaunal diversity values of the fall samples were low and only the outlaying stations—4000 m southwards and northwards from the outfall had higher diversity values. The background and peripheral stations in each survey had higher diversity than the other stations due to the representation of more polychaete families (see below). 4.3. Taxonomic groups Polychaete worms were the most abundant and diverse taxonomic group in our samples. To illustrate the faunal changes that occurred in the vicinity of the outfall, we examined the abundance of the dominant polychaete families in 1998 (spring and fall) as an example of the seasonality found during all surveys in the area (Fig. 5). In May 1998 (Fig. 5a) the effect of the sewage sludge was manifested in faunal depletion at the outfall, increased abundance in the immediate vicinity and northwards of the outfall, and decline to background levels further away. In September 1998 (Fig. 5b) the outfall and the area immediately adjacent were nearly devoid of life. Capitellid polychaetes were most abundant on the edge of the afaunal zone, followed by pilargids, chaetopterids and spionids. In both spring and fall capitellids dominated, most prominently northwards from the outfall—in spring capitellids made up more than 95% of the individuals up to 3000 m northwards from the outfall, and even in the furthest station northwards (station 26, 4500 m) they were more abundant than in the control station. Following these results, two additional stations northwards from the outfall (stations 29, 5500 m and station 61, 6000 m) were added to the monitoring program (Fig. 1). Subsequent sampling showed that the capitellids at those stations had similar abundance as the control station both in spring and fall. 4.4. Seasonal variability Seasonal variability was observed both in the sediment quality and in the infaunal abundance (Fig. 2 and 4). Statistical analysis showed that Corg, Cd, Cu, Zn, and Ni, were significantly higher in fall than in spring and infaunal abundance was significantly lower (P < 0.01). On the other hand, Hg, Pb, Mn, Fe, Al concentrations did not differ significantly in spring and fall samples. The infaunal distribution
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
227
showed a distinct seasonal pattern that conformed with the Pearson–Rosenberg model (1978), wherein increasing organic enrichment is characterized initially by increases in abundance and biomass, then by replacement of natural assemblages with guilds of opportunistic taxa, mainly polychaetes, and finally, by azoic sediments. In spring, no azoic samples were taken. The samples closest to the outfall and up to 2000 m northwards consisted of superabundant populations of opportunistic polychaetes. In the fall, the sediments in the vicinity of the outfall and up to 1000 m
Fig. 5. Distribution of the abundance (number of specimens) of polychaete families at the disposal site in May 1998 and September 1998 along the south-north transect. Negative and positive distances are southwards and northwards of the outfall, respectively.
228
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Fig. 6. Principal component analysis results of spring and fall data of the stations affected by the sewage sludge and for the background stations.
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
229
northward had high Corg concentration and were nearly devoid of infauna. Further away from the most organically enriched area the biota was composed of few pollution-tolerant taxa, including abundant populations of opportunistic polychaetes. Beyond the enriched zone, assemblages gradually approached the composition of those found in the background peripheral stations and abundance values declined.
5. Discussion Principal component analysis (PCA—Zitko, 1994) applied to the variables (Corg, heavy metals, infaunal abundance and diversity) at the affected stations, illustrated distinct seasonal differences (Fig. 6). The background reference stations, analyzed separately (see below), were excluded from this analysis in order to examine only the influence of the sewage sludge related material. In spring, the first two eigenvalues of the correlation matrix explained 80% of the variance and their plot formed three groups. The largest group included all the anthropogenic metals (Hg, Cd, Cu, Pb, Zn) and Corg, supporting the assumption of a common source. The correlation analysis showed these to be related as well (Table 4). This group included infaunal abundance, inferring a direct correlation with the anthropogenic input. The second group consisted of Fe, Al and Ni, considered natural to the area in the spring. The third group included Mn and infaunal diversity, both showing decreased values with increase in anthropogenic constituents. In PCA of fall surveys, the first two eigenvalues of the correlation matrix explained 72% of the variance. As in spring, Corg and the anthropogenic metals formed the largest group (Fig. 6). Ni in the fall was close to the Corg group indicating anthropogenic origin. The second group consisted of Fe and Al, considered natural to the area. Abundance in fall stood closer to Mn, and was negatively correlated with the pollutants, showing decreased values with increased organic carbon and anthropogenic metals enrichment, in contrast to the spring. Diversity in the fall formed a group alone, apart from the pollutants and the natural metals. PCA of background stations (Fig. 6) grouped the metals differently: Hg and Cd formed one group, Fe and Mn a second group and Ni, Cu, Zn and Al formed a third group. Pb and Corg were located separately. The first two eigenvalues of the correlation matrix explained 72% of the variance. This analysis could indicate a common source for each group but there are no other data to corroborate it. We suggest that the seasonal pattern evinced in infaunal abundance, anthropogenic metals and Corg values is due to the distinct stratification of the water column in spring and summer with its sharply defined halocline and thermocline below 25 m on one hand and winter storms on the other. Winter storms provide considerable energy for sediment transport. During winter storms, wave induced motions near the seabed rework the surface of the sediments, resuspending and widely dispersing the fine organic particles, reducing the concentration of sewage sludge. Current measurements show that daily average bottom currents of > 35 cm s1, a velocity high enough to set in motion fine sand and strong enough to qualify the site as dispersive, occur during 10–15% of the days between November and March (Norton & Champ, 1989; Rosentraub, 1990, Rosentraub, personal
230
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Table 6 Mass balance of anthropogenic metals in the study site
Hg Cd Pb Cu Zn a
UNEP/WHO (1999) kg year1
Calculated mass kg
Equivalent to years discharge
60 193a 1670 19,000 54,000
56 127 1787 9874 56,822
1 1.5 1 0.5 1
After metal reduction at source.
Table 7 North-south distances of the affected area from the outfall (in m) determined using chemical and biological criteria Criteria
Cd > background Cd > ERL Hg > background ERM > Hg > ERL Hg > ERM Corg > background Abn > background Abn background * 5b Abn
Spring
Fall
South
North
Areaa
South
North
Areaa
2000 50 2500 1500 0 2500 2000
3500 2000 3000 3000 750 3500 3000
5.5 2.05 5.5 4.5 0.75 6.0 5.0
1500 200 2000 1000 0 2500
4000 2000 4000 3500 1500 4000
5.5 2.2 6.0 4.5 1.5 6.5
3000
4000
7.0
500
2000
2.5 500
2000
2.5
Area affected (in km2) was calculated from the south–north distance assuming a west–east spread of 1 km. Heavily impacted area.
communication). A series of unusually strong storms struck the Mediterranean coast of Israel during the winter of 1991/1992, thus by May 1992 the vicinity of the outfall was only moderately affected by sludge accumulation, diverging significantly from subsequent spring surveys (Figs. 2 and 4). Undisturbed accumulation of sludge takes place through the quiescent periods of the year, when stratification is reestablished. Similarly, seasonal patterns of abundance were recorded from the Bay of Izmir, where the polluted zone was azoic or severely depleted during the summer and early fall and abounding with polychaetes in spring (Kocatas, 1981). The results show no evidence of increased accumulation with time of sewage sludge or of pollutants associated with it. The dispersive character of the site can be shown also by the mass balance of anthropogenic metals found in the area. The amount of material retained in the area after discharge was calculated as in Rodger, Davies & Topping (1992), assuming the following: the input of the metals was as reported in UNEP/WHO (1999) (Table 6) with Cd annual load halved, the area affected was 5 km2, water content and bulk
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
231
density of the sediment as 40% and 1.2 g cm3, respectively, the penetration of the sewage sludge was 17 cm (maximum measured, see above) and taking the maximal metal concentration measured. This is the worst-case scenario because of the overestimation of the affected area, penetration depth and concentrations. The calculated amounts of Hg, Pb, and Zn found at the site were equivalent to one year of discharge, and Cd and Cu equivalent of 1.5 and 0.5 years discharge, respectively (Table 6). We used here sediment chemical criteria (background and ERL-ERM) in conjunction with infaunal abundance to delimit the affected area (Table 7, Figs. 2 and 4). Assuming in all cases a west–east spread of 1000 m (Fig. 3) and depending on the criteria used, the affected area ranged between 0.75 and 7 km2. It can be seen that the extent of the influence of the sewage sludge determined by the infaunal background abundance agreed well with the extent determined by the background concentrations of Hg, Cd and Corg. The heavily impacted area was defined as the area where the infauna1 abundance was more than five times the natural background in the spring (> 2360 organisms, Tables 1 and 7) and less than five times the natural background in the fall < 50 organisms, the azoic zone, Tables 1 and 7). The southnorth distance of the heavily impacted area was on average 2500 m, from 500 to 2000 m from the outfall (Fig. 4), with an area of 2.5 km2. The heavily impacted area is slightly smaller than the area determined by ERL < Hg < ERM, similar to the area determined by Cd > ERL and larger than the area determined by Hg > ERM. The close agreement between the chemical and biological indices does not infer direct effect of Cd or Hg to the biota. At the same sediment concentrations of Hg, Cd and Corg in spring and fall, we have high abundance and azoic stations, respectively. Moreover, no bioaccumulation of Cd and Hg was found in the megabenthos species collected in the study area (our unpublished results). In conclusion, the disposal of sewage sludge off the Mediterranean coast of Israel has a marked but localized impact on the benthic assemblages and sediment quality. The disposal site is dispersive and the spatial extent of the impacted area varies seasonally and temporally. The impacted area extended mainly towards the north (up to ca. 4 km) and to a lesser extent south of the outfall (up to ca. 2.5 km). No evidence of increased accumulation of sewage sludge with time was found, nor of pollutants associated with it. Infaunal abundance together with sediment quality criteria were used to characterize the degree of impact in the study area. Long-term studies are extremely important for chronicling biotic and abiotic variables in coastal waters. This monitoring study, required by law, in addition to addressing specific questions about sewage sludge impacts, represents an unusually large and unique set of long-term measurements. Cessation of the marine disposal of sewage sludge in Israel is scheduled for 2008. Data collected during the ongoing monitoring program and shown in this study will be invaluable for following the recovery of the site.
Acknowledgements We thank the Dan Region Wastewater project (G. Zatz, General Director and D. Salomon, Head of the Environmental division) for funding this research and for
232
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
providing data on the inputs and outputs of the treatment plant. We are grateful to the captain and crew of the R/V Shikmona for their devoted work at sea. We thank Hava Hornung, Yaron Gertner, Efrat Shoham-Frider, Gerta Fainstein and Daniela Friedmann for their help in carrying out the sampling at sea and the experimental work in the laboratory, and Hana Bernhard for drawing the figures. The thorough comments of two anonymous reviewers helped to improve the manuscript and are greatly appreciated.
References ASTM. (1983). Designation- D3683-78. Standard test method for trace elements in coal and coke ash by atomic absorption. Pennsylvania, USA: American Society for Testing and Materials Publisher. Avnimelech, Y. (1989). Modeling the accumulation of organic matter in the sediments of a newly constructed reservoir. Water Research, 23, 1327–1329. Bayne, B. L., Clarcke, K. R., & Gray, J. S. (1988). Biological effects of pollution: Results of a practical workshop. Marine Ecology Progress Series, 46, 1–278. Berman, T., Townsend, D. W., El Sayed, S. Z., Trees, C. C., & Azov, Y. (1984). Optical transparency, chlorophyll and primary productivity in the Eastern Mediterranean near the Israeli coast. Oceanologica Acta, 7, 367–372. Bothner, M. H., Buchholtz ten Brink, M., & Manheim, F. T. (1998). Metal concentrations in surface sediments of Boston harbor—changes with time. Marine Environmental Research, 45, 127–155. Costello, M. I., & Read, P. (1994). Toxicity of sewage sludge to marine organisms: A Review. Marine Environmental Research, 37, 23–46. Galil, B. S., & Lewinsohn, C. (1981). Macrobenthic communities of the Eastern Mediterranean continental shelf. P.S.Z.N.I. Marine Ecology, 2, 343–352. Gaudette, H. E., Flight, W. R., Toner, L., & Folger, D. W. (1974). An inexpensive titration method for the determination of organic carbon in recent sediments. Journal of Sedimentary Petrology, 44, 249–253. Hemond, H. F., & Fechner-Levy, E. J. (2000). Chemical fate and transport in the environment. San Diego, USA: Academic Press. Hornung, H., Krom, M. D., & Cohen, Y. (1989). Trace metal distribution in sediments and benthic fauna of Haifa Bay, Israel. Estuarine Coastal and Shelf Science, 29, 43–56. Hunt, J.R (1990). Particle removal by coagulation and settling from a waste plume. In D. J. Baumgartner, & I. W. Duedall (Eds.), Oceanic processes in marine pollution V6. Physical and chemical processes: transport and transformation (pp. 109–119). Malabar, Fl, USA: R.E. Kriger Publishing. Kocatas, A. (1981). Evolution cyclique du benthos dans les zones de pollution du Golfe d’Izmir (Turquie). Journees d’ Etudes sur les Pollutions Marines, 5, 643–648. Kress, N., & Herut, B. (2001). Spatial and seasonal evolution of dissolved oxygen and nutrients in the Southern Levantine Basin (Eastern Mediterranean Sea). Chemical charaterization of the water masses and inferences on the high N:P ratio. Deep Sea Research, Part I, 48, 2347–2372. Kress, N., & Hornung, H. (1996). Disposal of sewage sludge at the continental shelf of Israel. Fresenius Environmental Bulletin, 5, 141–147. Kress, N., Shoam-Frider, E., & Shelef, G. (2002). Mercury speciation in sewage sludge and in sediments from the marine disposal site off the Mediterranean coast of Israel. Israel Oceanographic & Limnological Research Report H23/ 2002, 106. Lewis, D. L., & Gattie, D. K. (2002). Pathogen risks from applying sewage sludge to land. Environmental Science & Technology, 36(13), 287A–293A. Long, E. R., MacDonald, D. D., Smith, S. L., & Calder, F. D. (1995). Incidence of adverse biological effects within ranges of chemical concentrations in marine and estuarine sediments. Environmental Management, 19, 81–97. Norton, M.G, & Champ, M. A. (1989). The influence of site-specific characteristics on the effects of sewage-sludge dumping. Oceanic processes. In D. W. Wood, A. Schoener, & P. K. Park (Eds.), Oceanic
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
233
processes in marine pollution V4. Scientific monitoring strategies for ocean waste disposal (pp. 161–183). Malabar, Fl, USA: R.E. Kriger Publishing. O’Connor, T. P. (1998). Comparative criteria: Land application of sewage sludge and ocean disposal of dredged material. Marine Pollution Bulletin, 36, 181–184. Palanques, A., Diaz, J. I., & Maldonado, A. (1991). Impact of the sewage sludge discharged in the Barcelona continental shelf. In H. Chmaley (Ed.), Environment of epicontinental seas, Oceanologica Acta, 11 (pp. 329–336). Palanques, A. (1994). Distribution and heavy metal pollution of the suspended particulate matter on the Barcelona continental shelf (north-western Mediterranean). Environmental Pollution, 85, 205–215. Pearson, T. H. (1987). Benthic ecology in an accumulating sludge-disposal site. In J. M. Capuzzo, & D. R. Kester (Eds.), Oceanic processes in marine pollution VI. Biological processes and wastes in the ocean (pp. 195–200). Malabar, Fl, USA: R.E.Kriger Publishing. Pearson, T. H., & Rosenberg, R. (1978). Macrobenthic succession in relation to organic enrichment and pollution of the environment. Oceanography and Marine Biology Annual Reviews, 16, 229–311. Renner, R. (2000). Sewage sludge, pros and cons. Environmental Science and Technology, 34, 430 A–435A. Rodger, G. K., Davies, I. M., & Moore, D. C. (1991). Recovery of a sewage sludge dumping ground. I. Trace metal concentrations in the sediment. Marine Ecology Progress Series, 75, 293–299. Rodger, G. K., Davies, I. M., & McHenery, J. G. (1992). Effects of sewage sludge disposal on the sediment at the Garroch Head dump site, Firth of Clyde, Scotland. The Science of the Total Environment, 119, 133–156. Rodger, G. K., Davies, I. M., & Topping, G. (1992). Retention of trace metal contaminants in the sediment at an accumulating sewage sludge disposal site. Water Research, 26, 111–120. Rosentraub, Z. (1990). Investigation of the currents on the continental shelf of Israel. Israel Oceanographic and Limnological Research. Report H18/90. (in Hebrew). Shannon, C. E., & Weaver, W. (1949). The mathematical theory of communication. Urbana, Illinois: University of Illinois Press. Studholme, A.L, O’Reilly, J.E., & Ingham, M.C. (1995). Effects of cessation of sewage sludge dumping at the 12-mile site. NOAA Technical Report NMFS 124. UNEP/WHO (1999). Identification of priority hot spots and sensitive areas in the Mediterranean. MAP Technical Reports Series No. 124. UNEP, Athens. US EPA (1995). A guide to the biosolids risk assessments for the EPA Part 503 rule. EPA/832-B-93-005. Zitko, V. (1994). Principal component analysis in the evaluation of environmental data. Marine Pollution Bulletin, 28, 718–722.
Sewage sludge impact on sediment quality and benthic assemblages off the Mediterranean coast of Israel—a long-term study N. Kress*, B. Herut, B.S. Galil Israel Oceanographic and Limnological Research, National Institute of Oceanography, PO Box 8030, Haifa 31080, Israel Received 6 January 2003; received in revised form 10 July 2003; accepted 28 July 2003
Abstract The distributions of benthic assemblages, heavy metals and organic carbon (Corg) in sediments were examined during a long-term study at a sewage sludge disposal site off the Mediterranean coast of Israel. The disposal of sewage sludge has a marked but localized, seasonally dependent, impact on the benthic assemblages and sediment quality. Elevated concentrations of Corg, Hg, Cd, Cu, Zn, Pb, and to a lesser degree Ni in the sediments were detected mostly northward of the sewage outfall, in the direction of the prevalent longshore current. High concentrations of Corg and metals were reflected by elevated populations of tolerant and opportunistic polychaetes in spring and by an azoic zone in fall. The impacted area extended mainly towards the north (up to ca. 4 km) and to a lesser extent south of the outfall (up to ca. 2.5 km). No evidence of increased accumulation of sewage sludge with time was found, nor of pollutants associated with it. Principal component analysis (PCA) grouped the anthropogenic metals and Corg with infaunal abundance for the spring surveys, while biotic diversity was negatively correlated with the pollutants. In the PCA of fall surveys, abundance was negatively correlated with the pollutants, decreasing with increased concentration of Corg and anthropogenic metals. We suggest that the seasonal pattern shown by infaunal abundance, anthropogenic metals and Corg is due to the stratification of the water column from spring to fall on one-hand and winter storms on the other. Winter storms resuspend and disperse the fine organic particles, sweeping the site clean of sludge; accumulation of sludge takes place throughout the quiescent periods of the year, when stratification is reestablished. The disposal site is dispersive and the spatial extent of the impacted area varies seasonally and interannually. This monitoring study, in addition to addressing specific questions about sewage sludge impact, represents an unusually large
* Corresponding author. Tel.: +972-4-8565256; fax: +972-4-8511911. E-mail address: [email protected] (N. Kress). 0141-1136/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0141-1136(03)00081-3
214
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
and unique set of long-term measurements that will serve as a basis to evaluate the site recovery following the cessation of disposal. # 2003 Elsevier Ltd. All rights reserved. Keywords: Sewage sludge; Marine disposal; Heavy metals; Organic carbon; Benthic assemblages; Pollution; Eastern Mediterranean
1. Introduction Sewage sludge (also called biosolids) is generated mainly by the treatment of domestic waste waters. It consists primarily of organic biomass and nutrients (e.g. nitrogen, phosphorus, potassium) and may contain potentially harmful constituents such as pathogenic organisms, metals and synthetic organic compounds (US EPA, 1995; Renner, 2000). Sewage sludge can be utilized in land applications as fertilizer or soil conditioner, disposed of inland by landfilling or incineration, or disposed of in the aquatic environment by direct discharge or dumping. In the latter case, the main environmental effects are associated with the enrichment of the receiving habitat with organic matter and nutrients, accumulation of pollutants, their introduction into the food web, and changes in the biotic diversity and abundance. The treatment of domestic wastewaters in Israel generates 495,000 dry tons of sewage sludge annually, 40% of which are utilized in agriculture or disposed in landfills (D. Rubin, Ministry of the Environment, personal communication). The Dan Region Wastewater project alone, which treats the domestic sewage of the 1.5 million inhabitants of the Tel-Aviv Metropolitan area, produces 292,000 dry tons. The excess sludge produced at this plant (16,000 m3 day1) has been discharged, since 1987, through a single outfall, 5 km offshore, from a 1600 HDPE seabed pipeline placed nearly perpendicular to the shore (Fig. 1). The sludge contains approximately 1% particulate matter with metal mass emission rates (in kg year1) of: Hg— 60; Cd—430; Pb—1670; Cr—11400; Cu—19000; Zn—54000; Ni—2500 (UNEP/ WHO, 1999). Recently, reduction of metal load at the source resulted in halving the annual loads of Cd and Cr (D. Salomon, personal communication). Disposal of sewage sludge in the marine environment has been practiced globally. It affects the seafloor and its biota: contaminated sediments display increased levels of organic matter, nutrients and heavy metals, their biota are either greatly impoverished, or consist of large numbers of few opportunistic species (Bothner, Buchholtz ten Brink, & Manheim, 1998; Costello & Read, 1994; Norton & Champ, 1989; Pearson, 1987; Rodger, Davies, & Moore, 1991; Rodger, Davies, & McHenery, 1992; Studholme, O’Reilly, & Ingham, 1995; among others). In most studies, the effects of sewage sludge disposal were examined either in sediment quality parameters or in benthic assemblages, separately. Marine disposal of sewage sludge was terminated in 1992 in the USA as a result of the Ocean Disposal Ban Act of 1988 (O’Connor, 1998) and in the European Union in 1998 under the Urban Waste Water Treatment Directive (91/271/EEC). Recently, concerns were raised on the environmental safety of land application of sewage sludge such as long-term buildup of heavy metals in the soil and the presence of pathogens (Renner, 2000, Lewis & Gattie, 2002).
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
215
Fig. 1. Map of the study site. Filled circles designate sampling stations. Not shown is station 58, the reference station, located 4500 m south of the outfall.
216
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
This paper will: (a) present the environmental conditions at the disposal site after 15 years of sewage sludge discharge: heavy metals and organic carbon concentration in the sediments, composition of the benthic assemblages, spatial and temporal variations, (b) evaluate the possibility of long-term accumulation of the sewage sludge at the disposal site, and (c) combine chemical and biological variables to assess the impact of the sewage sludge on the oligotrophic environment of the eastern Mediterranean.
2. Study area The sublittoral zone off the Mediterranean coast of Israel slopes gently off shore, with isolines parallel to the shore. The Levantine basin of the eastern Mediterranean is ultra-oligotrophic (Berman, Townsend, El Sayed, Trees, & Azov, 1984; Kress & Herut, 2001); the water column at the continental shelf is mixed during winter (November–February), and stratified the rest of the year (2–2.5 sigma t units from surface to bottom) (Rosentraub, 1990). Surface water temperatures show high seasonal variations and range from 15 C during winter to 30 C in summer. The prevalent northwards current skims the shore, its average velocity less than 15 cm sl, decreasing with increasing depth in summer due to stratification, and increasing at times in winter to 90 cm s1 (Rosentraub, 1990; Rosentraub, pers. comm.). The near-bottom current has a slight westwards direction. The disposal site is located south of Tel Aviv, at depth of 38 m (Fig. 1), where the substrate was, prior to the outfall activation, mostly sand (65%) with silt (21%) and clay (14%) (Galil & Lewinsohn, 1981). A survey conducted in 1978 revealed no aberrant conditions and indicated that the outfall area was unpolluted at the advent of discharge (Galil & Lewinsohn, 1981). A monitoring program that examined the water quality, sediments and benthic fauna at a distance of 1 km north of the outfall detected no effects in 1987–1992. In 1992, 5 years after the start of discharge operations, the Ministry of the Environment authorized an expanded monitoring program that is in effect to the present (Kress & Hornung, 1996). It should be mentioned that no studies have been conducted on the effects of sewage sludge in the Mediterranean Sea except for a single survey off Barcelona (Palanques, Diaz, & Maldonado, 1991; Palanques, 1994).
3. Methods 3.1. Field sampling Twenty sampling stations were established in 1992, arrayed along two lines intersecting at the outfall, at distances of 50, 100, 200, 500 and 1000 m from it (Fig. 1). The north–south (N–S) transect runs along the 38 m isobath, subparallel to the coastline (015 ), while the east–west (E–W) transect, is subperpendicular to the shore, at depths of 35–39 m. Throughout the years the transect formation remained unaltered, though disposition of stations changed to reflect the affected area, with the farthest station at 6 km northward of the outfall. The sampling program consisted
217
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
of semiannual sampling (spring and fall) on even-numbered years, and annual sampling (fall) on odd-numbered years, allowing for the detection of biotic seasonal patterns (Galil & Lewinsohn, 1981). Sediment samples for the chemical analysis were taken in spring on even-numbered years, and in fall on odd-numbered years. This study reports the results of the surveys conducted from spring 1992 to spring 2002. At the beginning of the monitoring program, one reference site, situated 4.5 km south of the outfall was established with the rational that the peripheral stations, 1 km from the outfall (stns 5, 10, 15 and 20, Fig. 1) would reach natural levels and serve as reference stations as well. As the program evolved and the data was analyzed, it turned out that the initial assumption was correct for the eastern and western stations only (stns 10 and 20, respectively). Therefore, new peripheral stations were added mainly towards the north and formed a gradient from polluted to unaffected stations, increasing the data on the background character of the area. Consequently, data from the reference station and the peripheral background stations from all surveys were collected and served to determine the natural background of the area (Table 1). Support to these values is found in a related study (Kress, Shoam-Frider, & Shelef, 2002) that showed that the concentrations of Hg and organic carbon (Corg) in sediment cores from station 3 reached the background concentrations given in Table 1 at 15 cm sediment depth. Table 1 Descriptive statistics of Corg, heavy metal concentrations in surface sediments and macroinfaunal abundance at the disposal site in spring and fall and the natural background values in area Hg
Cd
Pb
Cu
Zn
Ni
Mn
1
mg g
0.04 0.01 25
Fe
Al
wt.%
Spring (n=134, Al and Corg, n=112) Minimum 0.02 bdl 1.66 9.40 Maximum 1.12 2.90 38.9 183 Median 0.16 0.42 16.4 36.9 Average 0.25 0.53 16.4 52.4 Stdev 0.24 0.56 7.32 37.9 CV(%) 95 104 45 72 Fall (n=113) Minimum 0.01 bdl 3.25 13.7 Maximum 1.36 3.12 43.8 242 Median 0.18 0.49 15.6 51.8 Average 0.28 0.81 16.7 73.9 Stdev 0.26 0.83 9.63 58.0 CV(%) 85 102 58 78 Natural background (n=12) Average Stdev CV(%)
Corg
0.07 0.06 85
8.79 5.12 58
21.4 6.10 29
30.0 586 97.2 149 120 80
10.8 67.9 33.9 35.4 9.94 28
257 935 631 641 94 15
0.18 10.3a 0.95 1.59 1.65 104
1.67 6.37 3.53 3.64 0.92 25
2.98 7.73 4.70 4.83 0.84 17
37.1 668 144 209 166 79
15.9 61.9 39.4 38.9 10.1 26
286 996 616 618 108 17
0.24 8.59 1.29 1.75 1.49 85
2.37 5.78 3.67 3.77 0.73 19
2.83 7.05 4.90 4.89 0.74 15
55.5 14.8 27
29.1 9.90 34
727 144 20
0.47 0.17 36
3.72 1.13 30
4.53 0.78 17
Abundance no of specimens (n=59) 224 13674 1702 2356 2345 100 (n=107) 0 458 12 55 84 152 Spring (n=22) 472 167 35
Fall (n=28) 251 104 41
Stdev—standard deviation, CV—coefficient of variance, n=number of data points, bdl—below detection limit ( <0.01 mg g1). a Without two high data points (10.3 and 9.8 wt.%), the average, stdev and median are 1.41, 1.12 and 0.96 wt.%, respectively.
218
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
A triplicate sample was collected at each station by a 0.062 m2 box-corer with an effective penetration of 40 cm (Ocean Instruments model 700 AL). A sample for chemical analysis was skimmed from a small area of the top 1 cm of the core surface of the first triplicate samples and frozen. The top 15 cm of the sediment was sieved through 0.5-mm mesh aboard ship, and preserved in 10% buffered formaldehyde. The amount of sediment sampled for the chemical analysis was minimal and did not cause a loss of organisms, shown by the good agreement of the results of the triplicate samples. An additional box-core sample was taken at selected stations along the N–S transect in order to assess sludge penetration. Sampling frequency and methods have remained constant since implementation of the improved monitoring program in 1992. 3.2. Chemical analyses in the laboratory The frozen sediment samples were lyophilized for 48 hr and then dry sieved through a 1000 mm sieve to extrude extraneous components such as seeds, broken shells, etc. The sediment was then digested with a mixture of hydrofluoric acid and aqua-regia as described in ASTM (1983) for the determination of Al, Fe, Mn, Cu and Zn. Concentrated nitric acid was used to digest the sediments for the determination of Cd, Pb, and Ni (Hornung, Krom, & Cohen, 1989). A separate digestion was performed in concentrated nitric acid for the determination of Hg. Metals in the digest solutions were analyzed by atomic absorption spectrometry, using an IL-951 or a Perkin–Elmer 1100 B spectrophotometer equipped with flame and graphite furnace modules. Only mercury was analyzed by cold vapor atomic absorption spectrometry on a Coleman Mercury Analyzer MAS-50. Quality control and quality assurance of the results was performed with standard reference materials from the USA National Institute of Standards and Technology (NIST—Estuarine Sediment— 1646, River Sediment—1645) and the National Research Council of Canada (NRCCMESS-2). The standards were digested and analyzed in the same manner as the samples, with each analytical run. All metals gave results within 5% of the certified values. Organic carbon (Corg) was determined by the potassium dichromate method following the procedure of Gaudette, Flight, Toner, and Folger (1974) and Avnimelech (1989) with slight modifications. All concentrations are reported on a dry weight basis. 3.3. Biological analysis in the laboratory The formalin-preserved biotic samples were washed, resieved thoroughly on a 0.5 mm sieve and preserved in 70% ethanol within days of sampling. Prior to sorting, the samples were stained in Rose Bengal, and the extracted organisms were identified and counted. The total number of individuals—Abundance—in the sample (0.0093 m3) is given as the average of the triplicate samples, and the ShannonWiener diversity index- H0 - (Shannon & Weaver, 1949) was calculated. Diversity indices are commonly used to assess pollution impact, where a high value denotes an unpolluted sample and a low value—a polluted one. However, statistically significant reductions in a diversity index are attended by great changes so that the index is considered a poor indicator (Bayne, Clarcke, & Gray, 1988), besides, the
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
219
interpretation of diversity indices is made difficult by natural periodic environmental changes that may produce changes that counteract or parallel those brought about by pollution. Although aware of the limitations in the use of diversity index we present it here to aid in the interpretation of the results and to compare with other sites reported in the literature. 3.4. Statistical analysis Statistical analyses were performed using the SPSS software with the procedures of general linear model (GLM) using the method of least squares, the t-test unequal variance, and Principal component analysis (PCA). Significance of < 0.05 was used.
4. Results The natural sediment in the area consists of grey-brown sandy-mud, easily distinguished from the black-colored, malodorous deposited sewage sludge sampled at the vicinity of the outfall. The depth of the sludge layer was estimated from sediment cores taken at stations northward of the outfall and ranged from no obvious sludge (i.e. 0 cm penetration) at the most distant stations (4–6 km) to 3–17 cm thick layer at the outfall area. In spring 2000, for example, the sludge layer was 17 cm deep at the outfall (station 0), 15 cm at station 3 (200 m north of the outfall), whereas 1 km north of the outfall (station 5), a layer 6.5 cm thick of sludge was covered by a 2 cm layer of sediment. No sludge was evident at station 24 located 3 km northwards of the outfall. 4.1. Heavy metals and organic carbon The ranges of Hg, Cd, Cu, Zn, Pb, Ni, Mn, Fe, Al and Corg concentration in the sediment in spring and fall, are summarized in Table 1, as well as the natural background concentrations for the area. Enrichment factor (EF), used as an index of the sewage sludge effect, is the ratio of a sample’s metal concentration to background concentration. Since water depth and sediment composition within the study area are similar throughout, there was no need for grain size normalization. Cd had the highest EF value, followed by Hg, Zn, Corg and Cu (Table 2). Pb was enriched mainly at the stations within 100 m radius from the outfall and northwards up to 1500 m from it, with a maximal value of 2.8. Mn EF values were lower than 0.9 at stations close to the outfall, indicating depletion due to formation of soluble species (Hemond & Fechner-Levy, 2000). Fe and Al, considered natural for the area, exhibited no enrichment. EF values for Ni were under 1.4, indicating that Ni was mostly natural for the area. T-test comparison between spring and fall values of Ni and background levels showed that, in spring, Ni was natural for the area (not significantly different from the background values, P one tail=0.107), however in the fall samples it was significantly higher (P one tail=0.00016) than the background concentration. In addition to the EF analysis, we compared the data set to the sediment quality guidelines, suggested by Long, MacDonald, Smith, and Calder (1995), based on
220
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
their potential to induce toxic effects in marine organisms [ERL (Effects Range Low) and ERM (Effects Range Median)]. Only Hg and Zn concentrations exceeded ERM in some of our samples (Table 3). The maximal values of Cd and Cu exceeded ERL values, while Pb concentration in all samples was lower than ERL. The natural background levels of Ni were higher than ERL values, therefore these criteria are not applicable for Ni in the study area. One of the main tracers of the presence and effects of sewage sludge in the marine environment is enrichment by organic matter, represented in this study by Corg. There was a marked positive relationship between Hg, Cd, Cu, Zn, Pb, Ni and Corg indicating the sewage sludge as the anthropogenic source of these metals to the environment (Table 4). Although no deviations from the natural background were detected for Mn it showed a negative trend with Corg, probably due to the reduction of MnO2 and formation of soluble species (Mn+2) at anoxic environments, as those Table 2 Average enrichment factors (EF) for the sampling stations Station
Corg
Hg
Cd
Pb
Cu
Zn
Ni
Mn
0 1 2 3 4 5 21 22 23 24 26 55 56 6 7 8 9 10 11 12 13 14 15 28 16 17 18 19 20 Maximum
3.2 5.7 4.8 4.9 4.4 4.3 3.3 2.4 1.5 1.2 1.0 1.2 1.7 3.7 2.7 1.9 1.7 0.9 3.0 2.1 2.9 2.3 1.6 1.7 3.0 3.2 2.3 1.2 1.3 5.7
6.5 11.3 9.8 12.0 15.2 13.0 8.8 9.2 4.6 2.3 5.3 2.2 4.9 9.8 8.8 4.6 2.8 1.4 4.9 7.9 5.5 4.7 6.5 1.9 8.0 11.7 6.0 3.0 1.3 15.2
9.3 18.4 15.7 14.5 16.7 16.2 11.7 8.5 7.2 2.7 0.9 1.0 2.3 12.5 8.1 4.9 3.2 1.2 8.0 7.9 6.6 5.4 3.0 1.4 10.2 10.7 6.5 2.2 1.3 18.4
1.6 2.8 2.4 2.3 2.3 2.3 1.9 1.7 1.4 1.2 0.9 0.9 1.1 2.3 1.8 1.4 1.3 1.1 1.8 1.8 1.6 1.4 1.3 0.9 2.0 2.1 1.5 1.3 1.3 2.8
2.7 5.6 4.5 4.6 4.5 4.8 3.9 3.2 2.2 1.4 0.9 1.0 1.6 3.6 2.6 1.7 1.4 0.9 2.4 2.3 2.2 1.9 1.5 1.4 2.8 3.1 2.3 1.3 1.3 5.6
3.4 6.1 5.2 5.5 5.5 5.1 4.0 3.1 2.2 1.4 1.0 1.1 1.7 4.7 2.7 2.0 1.5 0.9 2.8 2.6 2.6 2.0 1.6 1.3 3.3 3.5 2.5 1.3 1.2 6.1
1.2 1.3 1.2 1.3 1.3 1.3 1.2 1.2 1.0 1.0 0.7 0.8 1.0 1.4 1.2 1.0 1.0 0.8 1.2 1.1 1.1 1.0 1.0 1.1 1.1 1.1 1.0 1.1 1.2 1.4
0.82 0.72 0.72 0.78 0.78 0.79 0.80 0.83 0.80 0.80 0.75 0.78 0.90 0.82 0.83 0.82 0.83 0.83 0.80 0.83 0.82 0.82 0.88 0.94 0.78 0.81 0.86 0.96 1.09 1.1
Fe and Al concentrations were natural for the area.
221
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
found in the sediments with high organic loads (Hemond & Fechner Levy, 2000). Fe and Al, not considered pollutants in the area, showed no relationship to Corg (Table 4). The concentrations of Hg, Cd and Cu, representing the anthropogenic metals, along the south north transect across the outfall are shown in Fig. 2. There was a preferential spread (i.e. higher concentrations) northwards, in the direction of the average alongshore current. The concentrations reached background levels about 3 km north, and 1.5 km south of the outfall. The maximal values, their position within the affected region, and the distribution pattern varied with season and year (Fig. 2). The areal spread of the sewage sludge, as illustrated by Cd concentrations in spring 98 and fall 99, is depicted in Fig. 3. The sewage sludge was dispersed in the direction of the average longshore current, with limited east–west dispersion (perpendicular to the coast). The maximal concentrations were not centered at the outfall but slightly northward, consistent with the Table 3 Comparison of sediment quality criteria (Long et al., 1995) to concentrations found in the study area (Concentrations are in mg g1 dry wt.)
Cd Cu Pb Hg Ni Zn a b
ERL
ERM
Maximal concentration in the area
%a between ERL and ERM
%b above ERM
Background concentration in the area
1.2 34 46.7 0.15 20.9 150
9.6 270 218 0.71 51.6 410
3.12 242 43.8 1.36 67.9 668
17 63 0 47 – 33
0 0 0 7 – 9
0.07 21.4 8.79 0.04 29.1 55.5
Percentage of total data points (n=246) whose concentrations are between the ERL and ERM values. Percentage of total data points whose concentrations are above the ERM.
Table 4 R2 and linear regression parameters of heavy metals vs organic carbon
Hg Cd Pb Cu Zn Ni Mn Fe Al
R2
Slope
Intercept
Significance Fa
0.5375 0.6089 0.3046 0.5628 0.6730 0.3843 0.3472 0.00075 0.00016
0.119 0.360 3.078 24.18 77.62 4.027 38.24 0.0142 0.00633
0.0734 0.0779 11.198 24.633 53.77 30.53 693.29 3.75 4.88
7.22 E-39 3.8 E-47 2.9 E-19 9.29 E-42 8.56 E-56 3.56 E-25 2.5 E-22 0.683b 0.851b
Number of observations =224. a ANOVA b Not significant
222
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Fig. 2. Distribution of Cd, Hg and Cu (mg g1 dry wt.) in spring and fall along the south–north transect, from station 58 to station 61. Negative and positive distances are southwards and northwards of the outfall, respectively.
model describing the flux of settling particles out of a plume discharged from a submerged ocean outfall (Hunt, 1990). The other anthropogenic metals (Hg, Cu, Pb, Zn) were similarly distributed. 4.2. Macroinfaunal abundance and diversity The macroinfaunal abundance values for the stations along the south–north transect across the outfall, together with Corg, are shown in Fig. 4 for all the surveys. Since all stations are isobathic, the effect of depth on the infaunal community structure is annulled. We chose to present the results from stations along the south-
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
223
Fig. 3. Surficial distribution of Cd (mg g1 dry wt.) and abundance (Abn, number of specimens) at the disposal site in May 1998 and September 1999. Filled circles designate sampling stations.
224
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
north transect, because of the preferential northwards spread of the sludge. Infaunal abundance was high in the spring and low in the fall surveys (Fig. 4). In May 1992 high abundance was found at the stations in the immediate vicinity and those northwards from the outfall: 265 organisms were collected at station 4500 m northward of the outfall—higher than at the other stations. Abundance values in May 1992 were much lower than in the subsequent spring surveys, probably due to the unusual cold and stormy winter of 1991–1992. In May 1994—4238, 2251, and 406 organisms were collected at stations 4,5 and 22, respectively (Figs. 1 and 4). In May 1996, 3258, 1089 and 1371 organisms were collected at stations 4,5 and 22, respectively. In May 1998, June 2000 and May 2002, 1627, 2419 and 1961 organisms, respectively, were collected at station 22. In the fall surveys those same stations, located near the outfall, were nearly devoid of life (Fig. 4). In November 1992 only three organisms were found at station 4 and four organisms at station 5. Infaunal abundance values for the fall samples maintained the pattern of a nearly azoic zone in the immediate vicinity and northwards from the outfall, enclosed by a relatively high-abundance zone. The results for the fall of 1995 and two stations in fall 1998 were exceptional. The relatively high abundance values in 1995 resulted from the accidental rupture of the outfall pipe that reduced the amount of sludge discharged at the outfall mouth. However, even the high abundance in fall 1995 was lower than in
Fig. 4. Distribution of Corg (wt.% dry wt.) and abundance (number of specimens) in spring and fall along the south–north transect, from station 58 to station 61. Negative and positive distances are southwards and northwards of the outfall, respectively. Please note different abundance scale for spring and fall.
Distance from outfall
Station number
4500 1000 500 200 100 50 0 50 100 200 500 1000 1500 2000 3000 4500 5500
58 15 14 13 12 11 0 1 2 3 4 5 21 22 24 26 29
May 92
Nov 92
Oct 93
May 94
Nov 94
Sept 95
1.320 0.711 0.736 1.430 1.530
1.750 1.070 0.996 1.330 1.561
1.680 1.640 1.490 0.957 0.778
2.460 2.490 0.523 0.894 0.317
0.828 0.562 1.430 1.360 0.684
2.020 0.985 1.310 0.468 0.470
1.390 1.940 1.670 1.490 1.510
1.210 1.010 0.637 1.730 1.770
1.040 0.580 0.918 0.693 0.971
0.490 0.037 0.095 0.072 0.223 1.180 2.520
1.590 1.320 1.350 1.030 1.400 0.356 1.380
0.439 0.506 0.183 0.500 1.030 0.913 0.681 1.948
Negative and positive distances are southwards and northwards of the outfall, respectively.
May 96
Oct 96
Sept 97
May 98
Sept 98
Sept 99
2.190 0.873 0.218 0.163 0.129
2.050 1.660 1.470 1.080 0.939
2.160 1.910 1.120 0.902
2.130 0.769 0.696 0.301
2.180 1.450 1.150 1.099
1.590 0.528 1.280 0.000
0.146
1.030
1.050
0.166
0.693
0.995
0.346 0.486 0.214 0.205 0.331 0.678 1.930 2.450
0.561 0.775 0.796 1.470 1.040 0.975 1.110
1.350 1.250 0.857 1.650 1.230 1.740
0.420 0.374 0.263 0.354 0.524 1.290 2.050
0.171 0.093 0.206 0.114 0.476 0.792 1.910
1.210 0.000 0.000 1.050 0.685 0.901 2.080
2.060
1.823
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Table 5 Diversity index for the stations along the south–north transect from spring 1992 to fall 1999
225
226
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
the spring surveys. It seems that the two high values in the fall of 1998 (stations 21–22) were due to the transition between the azoic zone to the super abundant zone. This is supported by the lower diversity found at those stations compared to the stations further away from the outfall (Table 5). Background abundance was seasonally dependent, significantly higher in the spring than in the fall (P < 0.0001, Table 1). Infaunal diversity is presented in Table 5. In May 1992 diversity was slightly higher than in the following spring surveys, except for three stations in May 1994 (stations 14, 15 and 22, Fig. 1, Table 5). Diversity values at stations 2 and 3, in close vicinity of the outfall, were 1.94 and 1.67, respectively, higher than in May 1994 (0.037, and 0.095 at stations 2 and 3, respectively). The samples collected in May 1996, May 1998 and June 2000 exhibited low diversity values (< 1.0) at the stations up to 1000 m southwards, and 2000 m northwards from the outfall. Infaunal diversity values of the fall samples were low and only the outlaying stations—4000 m southwards and northwards from the outfall had higher diversity values. The background and peripheral stations in each survey had higher diversity than the other stations due to the representation of more polychaete families (see below). 4.3. Taxonomic groups Polychaete worms were the most abundant and diverse taxonomic group in our samples. To illustrate the faunal changes that occurred in the vicinity of the outfall, we examined the abundance of the dominant polychaete families in 1998 (spring and fall) as an example of the seasonality found during all surveys in the area (Fig. 5). In May 1998 (Fig. 5a) the effect of the sewage sludge was manifested in faunal depletion at the outfall, increased abundance in the immediate vicinity and northwards of the outfall, and decline to background levels further away. In September 1998 (Fig. 5b) the outfall and the area immediately adjacent were nearly devoid of life. Capitellid polychaetes were most abundant on the edge of the afaunal zone, followed by pilargids, chaetopterids and spionids. In both spring and fall capitellids dominated, most prominently northwards from the outfall—in spring capitellids made up more than 95% of the individuals up to 3000 m northwards from the outfall, and even in the furthest station northwards (station 26, 4500 m) they were more abundant than in the control station. Following these results, two additional stations northwards from the outfall (stations 29, 5500 m and station 61, 6000 m) were added to the monitoring program (Fig. 1). Subsequent sampling showed that the capitellids at those stations had similar abundance as the control station both in spring and fall. 4.4. Seasonal variability Seasonal variability was observed both in the sediment quality and in the infaunal abundance (Fig. 2 and 4). Statistical analysis showed that Corg, Cd, Cu, Zn, and Ni, were significantly higher in fall than in spring and infaunal abundance was significantly lower (P < 0.01). On the other hand, Hg, Pb, Mn, Fe, Al concentrations did not differ significantly in spring and fall samples. The infaunal distribution
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
227
showed a distinct seasonal pattern that conformed with the Pearson–Rosenberg model (1978), wherein increasing organic enrichment is characterized initially by increases in abundance and biomass, then by replacement of natural assemblages with guilds of opportunistic taxa, mainly polychaetes, and finally, by azoic sediments. In spring, no azoic samples were taken. The samples closest to the outfall and up to 2000 m northwards consisted of superabundant populations of opportunistic polychaetes. In the fall, the sediments in the vicinity of the outfall and up to 1000 m
Fig. 5. Distribution of the abundance (number of specimens) of polychaete families at the disposal site in May 1998 and September 1998 along the south-north transect. Negative and positive distances are southwards and northwards of the outfall, respectively.
228
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Fig. 6. Principal component analysis results of spring and fall data of the stations affected by the sewage sludge and for the background stations.
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
229
northward had high Corg concentration and were nearly devoid of infauna. Further away from the most organically enriched area the biota was composed of few pollution-tolerant taxa, including abundant populations of opportunistic polychaetes. Beyond the enriched zone, assemblages gradually approached the composition of those found in the background peripheral stations and abundance values declined.
5. Discussion Principal component analysis (PCA—Zitko, 1994) applied to the variables (Corg, heavy metals, infaunal abundance and diversity) at the affected stations, illustrated distinct seasonal differences (Fig. 6). The background reference stations, analyzed separately (see below), were excluded from this analysis in order to examine only the influence of the sewage sludge related material. In spring, the first two eigenvalues of the correlation matrix explained 80% of the variance and their plot formed three groups. The largest group included all the anthropogenic metals (Hg, Cd, Cu, Pb, Zn) and Corg, supporting the assumption of a common source. The correlation analysis showed these to be related as well (Table 4). This group included infaunal abundance, inferring a direct correlation with the anthropogenic input. The second group consisted of Fe, Al and Ni, considered natural to the area in the spring. The third group included Mn and infaunal diversity, both showing decreased values with increase in anthropogenic constituents. In PCA of fall surveys, the first two eigenvalues of the correlation matrix explained 72% of the variance. As in spring, Corg and the anthropogenic metals formed the largest group (Fig. 6). Ni in the fall was close to the Corg group indicating anthropogenic origin. The second group consisted of Fe and Al, considered natural to the area. Abundance in fall stood closer to Mn, and was negatively correlated with the pollutants, showing decreased values with increased organic carbon and anthropogenic metals enrichment, in contrast to the spring. Diversity in the fall formed a group alone, apart from the pollutants and the natural metals. PCA of background stations (Fig. 6) grouped the metals differently: Hg and Cd formed one group, Fe and Mn a second group and Ni, Cu, Zn and Al formed a third group. Pb and Corg were located separately. The first two eigenvalues of the correlation matrix explained 72% of the variance. This analysis could indicate a common source for each group but there are no other data to corroborate it. We suggest that the seasonal pattern evinced in infaunal abundance, anthropogenic metals and Corg values is due to the distinct stratification of the water column in spring and summer with its sharply defined halocline and thermocline below 25 m on one hand and winter storms on the other. Winter storms provide considerable energy for sediment transport. During winter storms, wave induced motions near the seabed rework the surface of the sediments, resuspending and widely dispersing the fine organic particles, reducing the concentration of sewage sludge. Current measurements show that daily average bottom currents of > 35 cm s1, a velocity high enough to set in motion fine sand and strong enough to qualify the site as dispersive, occur during 10–15% of the days between November and March (Norton & Champ, 1989; Rosentraub, 1990, Rosentraub, personal
230
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
Table 6 Mass balance of anthropogenic metals in the study site
Hg Cd Pb Cu Zn a
UNEP/WHO (1999) kg year1
Calculated mass kg
Equivalent to years discharge
60 193a 1670 19,000 54,000
56 127 1787 9874 56,822
1 1.5 1 0.5 1
After metal reduction at source.
Table 7 North-south distances of the affected area from the outfall (in m) determined using chemical and biological criteria Criteria
Cd > background Cd > ERL Hg > background ERM > Hg > ERL Hg > ERM Corg > background Abn > background Abn
Spring
Fall
South
North
Areaa
South
North
Areaa
2000 50 2500 1500 0 2500 2000
3500 2000 3000 3000 750 3500 3000
5.5 2.05 5.5 4.5 0.75 6.0 5.0
1500 200 2000 1000 0 2500
4000 2000 4000 3500 1500 4000
5.5 2.2 6.0 4.5 1.5 6.5
3000
4000
7.0
500
2000
2.5 500
2000
2.5
Area affected (in km2) was calculated from the south–north distance assuming a west–east spread of 1 km. Heavily impacted area.
communication). A series of unusually strong storms struck the Mediterranean coast of Israel during the winter of 1991/1992, thus by May 1992 the vicinity of the outfall was only moderately affected by sludge accumulation, diverging significantly from subsequent spring surveys (Figs. 2 and 4). Undisturbed accumulation of sludge takes place through the quiescent periods of the year, when stratification is reestablished. Similarly, seasonal patterns of abundance were recorded from the Bay of Izmir, where the polluted zone was azoic or severely depleted during the summer and early fall and abounding with polychaetes in spring (Kocatas, 1981). The results show no evidence of increased accumulation with time of sewage sludge or of pollutants associated with it. The dispersive character of the site can be shown also by the mass balance of anthropogenic metals found in the area. The amount of material retained in the area after discharge was calculated as in Rodger, Davies & Topping (1992), assuming the following: the input of the metals was as reported in UNEP/WHO (1999) (Table 6) with Cd annual load halved, the area affected was 5 km2, water content and bulk
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
231
density of the sediment as 40% and 1.2 g cm3, respectively, the penetration of the sewage sludge was 17 cm (maximum measured, see above) and taking the maximal metal concentration measured. This is the worst-case scenario because of the overestimation of the affected area, penetration depth and concentrations. The calculated amounts of Hg, Pb, and Zn found at the site were equivalent to one year of discharge, and Cd and Cu equivalent of 1.5 and 0.5 years discharge, respectively (Table 6). We used here sediment chemical criteria (background and ERL-ERM) in conjunction with infaunal abundance to delimit the affected area (Table 7, Figs. 2 and 4). Assuming in all cases a west–east spread of 1000 m (Fig. 3) and depending on the criteria used, the affected area ranged between 0.75 and 7 km2. It can be seen that the extent of the influence of the sewage sludge determined by the infaunal background abundance agreed well with the extent determined by the background concentrations of Hg, Cd and Corg. The heavily impacted area was defined as the area where the infauna1 abundance was more than five times the natural background in the spring (> 2360 organisms, Tables 1 and 7) and less than five times the natural background in the fall < 50 organisms, the azoic zone, Tables 1 and 7). The southnorth distance of the heavily impacted area was on average 2500 m, from 500 to 2000 m from the outfall (Fig. 4), with an area of 2.5 km2. The heavily impacted area is slightly smaller than the area determined by ERL < Hg < ERM, similar to the area determined by Cd > ERL and larger than the area determined by Hg > ERM. The close agreement between the chemical and biological indices does not infer direct effect of Cd or Hg to the biota. At the same sediment concentrations of Hg, Cd and Corg in spring and fall, we have high abundance and azoic stations, respectively. Moreover, no bioaccumulation of Cd and Hg was found in the megabenthos species collected in the study area (our unpublished results). In conclusion, the disposal of sewage sludge off the Mediterranean coast of Israel has a marked but localized impact on the benthic assemblages and sediment quality. The disposal site is dispersive and the spatial extent of the impacted area varies seasonally and temporally. The impacted area extended mainly towards the north (up to ca. 4 km) and to a lesser extent south of the outfall (up to ca. 2.5 km). No evidence of increased accumulation of sewage sludge with time was found, nor of pollutants associated with it. Infaunal abundance together with sediment quality criteria were used to characterize the degree of impact in the study area. Long-term studies are extremely important for chronicling biotic and abiotic variables in coastal waters. This monitoring study, required by law, in addition to addressing specific questions about sewage sludge impacts, represents an unusually large and unique set of long-term measurements. Cessation of the marine disposal of sewage sludge in Israel is scheduled for 2008. Data collected during the ongoing monitoring program and shown in this study will be invaluable for following the recovery of the site.
Acknowledgements We thank the Dan Region Wastewater project (G. Zatz, General Director and D. Salomon, Head of the Environmental division) for funding this research and for
232
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
providing data on the inputs and outputs of the treatment plant. We are grateful to the captain and crew of the R/V Shikmona for their devoted work at sea. We thank Hava Hornung, Yaron Gertner, Efrat Shoham-Frider, Gerta Fainstein and Daniela Friedmann for their help in carrying out the sampling at sea and the experimental work in the laboratory, and Hana Bernhard for drawing the figures. The thorough comments of two anonymous reviewers helped to improve the manuscript and are greatly appreciated.
References ASTM. (1983). Designation- D3683-78. Standard test method for trace elements in coal and coke ash by atomic absorption. Pennsylvania, USA: American Society for Testing and Materials Publisher. Avnimelech, Y. (1989). Modeling the accumulation of organic matter in the sediments of a newly constructed reservoir. Water Research, 23, 1327–1329. Bayne, B. L., Clarcke, K. R., & Gray, J. S. (1988). Biological effects of pollution: Results of a practical workshop. Marine Ecology Progress Series, 46, 1–278. Berman, T., Townsend, D. W., El Sayed, S. Z., Trees, C. C., & Azov, Y. (1984). Optical transparency, chlorophyll and primary productivity in the Eastern Mediterranean near the Israeli coast. Oceanologica Acta, 7, 367–372. Bothner, M. H., Buchholtz ten Brink, M., & Manheim, F. T. (1998). Metal concentrations in surface sediments of Boston harbor—changes with time. Marine Environmental Research, 45, 127–155. Costello, M. I., & Read, P. (1994). Toxicity of sewage sludge to marine organisms: A Review. Marine Environmental Research, 37, 23–46. Galil, B. S., & Lewinsohn, C. (1981). Macrobenthic communities of the Eastern Mediterranean continental shelf. P.S.Z.N.I. Marine Ecology, 2, 343–352. Gaudette, H. E., Flight, W. R., Toner, L., & Folger, D. W. (1974). An inexpensive titration method for the determination of organic carbon in recent sediments. Journal of Sedimentary Petrology, 44, 249–253. Hemond, H. F., & Fechner-Levy, E. J. (2000). Chemical fate and transport in the environment. San Diego, USA: Academic Press. Hornung, H., Krom, M. D., & Cohen, Y. (1989). Trace metal distribution in sediments and benthic fauna of Haifa Bay, Israel. Estuarine Coastal and Shelf Science, 29, 43–56. Hunt, J.R (1990). Particle removal by coagulation and settling from a waste plume. In D. J. Baumgartner, & I. W. Duedall (Eds.), Oceanic processes in marine pollution V6. Physical and chemical processes: transport and transformation (pp. 109–119). Malabar, Fl, USA: R.E. Kriger Publishing. Kocatas, A. (1981). Evolution cyclique du benthos dans les zones de pollution du Golfe d’Izmir (Turquie). Journees d’ Etudes sur les Pollutions Marines, 5, 643–648. Kress, N., & Herut, B. (2001). Spatial and seasonal evolution of dissolved oxygen and nutrients in the Southern Levantine Basin (Eastern Mediterranean Sea). Chemical charaterization of the water masses and inferences on the high N:P ratio. Deep Sea Research, Part I, 48, 2347–2372. Kress, N., & Hornung, H. (1996). Disposal of sewage sludge at the continental shelf of Israel. Fresenius Environmental Bulletin, 5, 141–147. Kress, N., Shoam-Frider, E., & Shelef, G. (2002). Mercury speciation in sewage sludge and in sediments from the marine disposal site off the Mediterranean coast of Israel. Israel Oceanographic & Limnological Research Report H23/ 2002, 106. Lewis, D. L., & Gattie, D. K. (2002). Pathogen risks from applying sewage sludge to land. Environmental Science & Technology, 36(13), 287A–293A. Long, E. R., MacDonald, D. D., Smith, S. L., & Calder, F. D. (1995). Incidence of adverse biological effects within ranges of chemical concentrations in marine and estuarine sediments. Environmental Management, 19, 81–97. Norton, M.G, & Champ, M. A. (1989). The influence of site-specific characteristics on the effects of sewage-sludge dumping. Oceanic processes. In D. W. Wood, A. Schoener, & P. K. Park (Eds.), Oceanic
N. Kress et al. / Marine Environmental Research 57 (2004) 213–233
233
processes in marine pollution V4. Scientific monitoring strategies for ocean waste disposal (pp. 161–183). Malabar, Fl, USA: R.E. Kriger Publishing. O’Connor, T. P. (1998). Comparative criteria: Land application of sewage sludge and ocean disposal of dredged material. Marine Pollution Bulletin, 36, 181–184. Palanques, A., Diaz, J. I., & Maldonado, A. (1991). Impact of the sewage sludge discharged in the Barcelona continental shelf. In H. Chmaley (Ed.), Environment of epicontinental seas, Oceanologica Acta, 11 (pp. 329–336). Palanques, A. (1994). Distribution and heavy metal pollution of the suspended particulate matter on the Barcelona continental shelf (north-western Mediterranean). Environmental Pollution, 85, 205–215. Pearson, T. H. (1987). Benthic ecology in an accumulating sludge-disposal site. In J. M. Capuzzo, & D. R. Kester (Eds.), Oceanic processes in marine pollution VI. Biological processes and wastes in the ocean (pp. 195–200). Malabar, Fl, USA: R.E.Kriger Publishing. Pearson, T. H., & Rosenberg, R. (1978). Macrobenthic succession in relation to organic enrichment and pollution of the environment. Oceanography and Marine Biology Annual Reviews, 16, 229–311. Renner, R. (2000). Sewage sludge, pros and cons. Environmental Science and Technology, 34, 430 A–435A. Rodger, G. K., Davies, I. M., & Moore, D. C. (1991). Recovery of a sewage sludge dumping ground. I. Trace metal concentrations in the sediment. Marine Ecology Progress Series, 75, 293–299. Rodger, G. K., Davies, I. M., & McHenery, J. G. (1992). Effects of sewage sludge disposal on the sediment at the Garroch Head dump site, Firth of Clyde, Scotland. The Science of the Total Environment, 119, 133–156. Rodger, G. K., Davies, I. M., & Topping, G. (1992). Retention of trace metal contaminants in the sediment at an accumulating sewage sludge disposal site. Water Research, 26, 111–120. Rosentraub, Z. (1990). Investigation of the currents on the continental shelf of Israel. Israel Oceanographic and Limnological Research. Report H18/90. (in Hebrew). Shannon, C. E., & Weaver, W. (1949). The mathematical theory of communication. Urbana, Illinois: University of Illinois Press. Studholme, A.L, O’Reilly, J.E., & Ingham, M.C. (1995). Effects of cessation of sewage sludge dumping at the 12-mile site. NOAA Technical Report NMFS 124. UNEP/WHO (1999). Identification of priority hot spots and sensitive areas in the Mediterranean. MAP Technical Reports Series No. 124. UNEP, Athens. US EPA (1995). A guide to the biosolids risk assessments for the EPA Part 503 rule. EPA/832-B-93-005. Zitko, V. (1994). Principal component analysis in the evaluation of environmental data. Marine Pollution Bulletin, 28, 718–722.