ML050960508

Comment (60) of Nancy Burton in Regard to Millstone Nuclear Power Station/Draft Environmental Impact Statement/Supplemental Comments

ML050960508
Person / Time
Site:	Millstone
Issue date:	03/28/2005
From:	Burton N Connecticut Coalition Against Millstone
To:	NRC/ADM/DAS/RDB
References
69fr71437 00060
Download: ML050960508 (55)

Text

R,- at rD fb 14-CONNECTICUJT COALITION AGAINST MILLSTONE www.mothballmillstone.org I. - I March 28, 2005 Chief C /w7 Rules and Directives Branch Division of Administrative Services Office of Administration Mailstop T-6D59 U.S. Nuclear Regulatory Commission Washington DC 20555-0001 Re: Millstone Nuclear Power Station/Draft Environmental Impact Statement/Supplemental Comments

Dear Sirs:

We enclose herewith a copy of "Trace Metals and Radionuclides Reveal Sediment Sources and Accumulation Rates in Jordan Cove, Connecticut," published in "Estuaries" in 1999.

We referred to this report in our March 16, 2005 written comments.

Thank you for your assistance.

Sincerely, Please reply to:

Nancy Burton 147 Cross Highway Redding Ridge CT 06876 Tel. 203-938-3952 h(4 , 0 -2,

/f ./--- 4Fu 6k -Aez 4--)

3-4

G . Benoit et at. Metal and radionuclide investigation of sediments Trace Metals and Radionuclides Reveal Sediment Sources and Accumulation Rates in Jordan Cove, Connecticut Gaboury Benoit*

Tim F. Rozan School ofForestry & Environmental Studies 370 Prospect Street Yale Ihbr-altvest New Haven. Connecticut 06511. USA (203) 432-5139 FAX (203) 432-3929 gaboury.benoitayvale.edu Peter C. Patton Department of Earth and Environmental Sciences Weslevan University Middletown CT 06459 Chester L. Arnold University of Connecticut oQ2peiative Extension Service Haddam aT 06422 Estuaries, in press. (voL 22, #1, 1999)

I .-

Benoit et al. - 1 1 ABSTRACT: Many small estuaries are influenced by flow restrictions resulting from 2 transportation rights-of-way and other causes. The biogeochemical functioning and history of 3 such systems can be evaluated through study of their sediments, Ten long and six short cores 4 were collected from the length of Jordan Cove, Connecticut, a Long Island Sound sub-S estuary, and analyzed for stratigraphy, radionuclides (14C, 210 Pb, "'Ra, '"Cs, and ""Co) and 6 metals (Ag, Cd, Cu, Pb, Zn, Fe, and Al). For at least 3,800 yr, rising sea level has gradually.

7 inundated Jordan Cove, filling it with mud similar to that currently being deposited there.

8 Long-term sediment accumulation in the cove averaged close to 0.1 cm ye' over the last three 9 millennia. Recent sediment accumulation rates decrease inland from 0.84 to 0.40 cm ye' and 10 are slightly faster than relative sea level rise at this site (0.3 cm yr'). Similarity of depth 11 distributions of trace metals was used to confirm relative sediment accumulation rates.

12 and Ag are derived from sources outside the cove and its watershed, 13 H a ad regional contaminated sediments, respectively. The combined data 14 thus suggest that Long Island Sound is an important source of sediment to the cove, although 15 a minor part of total sediment is also supplied from the local watershed. Trace metal levels 16 are strongly correlated with Fe, but not with either organic matter or Al. Sediment quality has 17 declined in the cove over the past 60 years, but only slightly. A novel method based on ratios 18 -of Fe-nonnalized trace metals was developed to identify and quantify sediment source areas.

'a

Benoit et al. - 2 19 Introduction 20 The Connecticut coast of Long Island Sound is characterized by a large number of 21 embayments formed from stream valleys drowned by the rise in sea level occurring since the 22 last deglaciation (Lewis and Stone 1991). Railroad and/or highway causeways have been built 23 across almost all of these systems, restricting tidal exchange with the sound. Purely anecdotal 24 evidence suggests that reduced flushing has increased the quantity and reduced the quality of 25 sediments being deposited in the embayments. Indeed, most of the coves are shallow, and 26 surface sediments are black and organic rich. In recent years, there has been significant 27 pressure to "restore' the coves, for example by replacing the railroad causeway with bridges 28 or other structures that permit nearly unrestricted flow. Before a large amount of money is 29 devoted to this purpose, it would be valuable to be able to reconstruct the conditions that 30 prevailed in the natural sedimentary environment and to understand the nature and extent of 31 human perturbation of sediment dynamics. So far, very little attention has been paid to these 32 numerous embayments (Lyons and Fitzgerald 1980), which are heavily used for recreation and 33 shelIfishing, and constitute a valuable resource. In spite of these concerns, little is known 34 about sedimentary dynamics in coastal embayments of Long Island Sound.

35 In addition, it is important to know the sources of sediments. Different management 36 strategies need to be employed depending on whether sediments are derived mainly from the 37 local watershed or from the seaward direction. In the past; trace metal ratios have been used 38 to discriminate sediment sources, but only in badly contaminated sites (Krumgalz 1993; Shine 39 et al. 1995). Here we try to develop a novel method, based on trace metal ratios, to identify 40 and quantify sources for sediments that contain only low levels of anthropogenic , .

41 contaminants.

Benoit et al. - 3 42 The goals of the current study were: 1)to determine the sedimentary history of a Long 43 Island Sound cove, 2) to evaluate both the long-term and recent rates at which sediments 44 accumulate in the cove, 3) to evaluate the sediment quality, and 4) to use trace metals to 45 identify the source(s) of those sediments.

46 Study Site 47 Jordan Cove (410 19' N, 710 09' W) is a shallow and narrow embayment formed by the 48 drowned mouth of Jordan Brook Waterford, Connecticut (Fig. 1). The cove is about 1400 m 49 wide at its mouth (Millstone Pt. to White Pt.) and narrows to less than 100 m for nearly 2 km 50 near its head. The total length of the cove is about 4.2 km and water depths are typically 2 m.

51 At its head is a small millpond behind a low dam, which prevents tidal flow. Any landward 52 sediment transport is trapped below this dam. Sediments transported seaward in Jordan 53 Brook are probably not quantitatively trapped above the dam, since we estimate that average 54 water residence time is less than one day. Jordan Brook's 9.4 km2 watershed is moderately 55 developed, and the cove's shores are mainly wooded and undeveloped except for a few 56 residences. Both a railroad embankment and a road cross the cove, and a sand spit partially 57 blocks its mouth, breaking it into three basins. The Millstone nuclear power plant is located 58 on the Long Island Sound shore just to the west of the cove.

59 The entrance to Jordan Cove has undergone significant change since the first accurate 60 map published in 1868 (Welsh and Whitlach 1980). At that time two spits extended from the 61 eastern bank and recurved to the north into the cove. The 1893 U.S. Geological Survey map 62 showed a custer of marsh islands in the position formerly occupied by the sand spit.

63 Beginning about 1917 and continuing into the 1930's a local landowner drove pilings and.

64 constructed bulkheads across the marsh islands to create a triangular structure which was

Benoit et al. - 4 65 subsequently filled with sand to create the modem day sand spit. The linear edge on the 66 northern and eastern side of the present spit is the line of the remaining bulkheads. On the 67 soundward side of the spit, erosion has exposed the edge of the marsh islands and recent 68 winter storms have cut through the western edge of the spit and established a second channel 69 into Jordan Cove. Of significance to this study is that the upper portion of Jordan Cove has 70 had a restricted connection with Long Island Sound for over 150 years.

71 The narrowness of the estuary and the existence of three separate artificial restrictions 72 make it a useful test case for the hypothesis that reduced tidal flushing has resulted in 73 accelerated rates of sediment accumulation. Also, the cove is likely to receive sediment from 74 two distinct and relatively clean sources, its watershed and tidal flushing from Long Island 75 Sound. This system provides a relatively simple case in which to evaluate the use of trace 76 metal ratios to identify sediment sources.

77 Methods 78 SEDIMENT STRATIGRAPHY AND LONG-TERM HISTORY 79 Ten vibracores (designated, JC-2, 3, 4, 5, 7, 8, 10, 11, 12, and 14), ranging from 1.4 80 to 5.9 m long, were taken from different modem environments throughout the cove (filled 81 circles on Fig. 1). Two of the cores were taken from the shallow sand flats south of the 82' modern day spit, one core (core JC-4) was taken from the base of the-spit where a peat 83 deposit is exposed, and the remaining seven cores were taken in a longitudinal transect along 84 the axis of the cove. The vibracores were taken using 7.5 cm diameter, 9 m long, aluminum 85 irrigation pipe fitted with core catchers that were vibrated into the sediment until refusal. The 86 weight of the core barrel, vibrator, and cable, is about 30 kg. The cores were cut lengthwise .

Benoit et al. - S 87 and the stratigraphy of the sediment was described. Samples were taken for grain size analysis 88 and samples of wood and shell were taken for radiocarbon analysis.

89 The recovered vibracores are commonly shorter than the total length of core vibrated 90 into the sediment. There are two possible explanations for this phenomena. The sediment 91 wthin the core may be compacted during the vibracoring process, although this requires the 92 expulsion of pore water as the sediment enters the aluminum pipe. Another, equally plausible, 93 explanation is that at some point the friction of the sediment inside the core barrel effectively 94 stops the coring process, but the plugged core barrel can be rammed into and displaces 95 sediment in front of the core. In the first case, the length of core retrieved is a minimum 96 estimate of the true stratigraphic thickness, whereas in the second case, the length of core 97 represents the true thickness of the recovered sedimentary sequence. In this paper we 98 calculate sediment accumulation rates based on both assumptions, but the cross-sections are 99 drawn using the un-adjusted thickness of the sedimentation units.

100 RECENT SEDIMENT ACCUMULATION AND TRACE METAL CONTENT 101 Short (a 30 cm) cores were taken by SCUBA divers in August 1993 at five sites along a 102 transect from just outside its mouth to the head of the embayment (JC-1, 6, 9, 13, and 15; 103 open circles on Fig. 1). Cores were collected in 12.5 cm diameter PVC pipe that had been 104 pre-sectioned at approximately I cm intervals and then reassembled with electrical tape.

210 105 Immediately after collection, cores were sectioned and measured for radionuclides ( Pb, 106 =Ra, ' 3 7Cs, 7Be," 0 Co), trace metals (Pb, Cd, A& Cu, Fe, Al), and physical characteristics 107 (porosity, loss on ignition, bulk density). A sixth core (C-16) was collected in a 4.7 cm LD.

210Pb and 137Cs were used to 108 gravity core from behind the dam and analyzed for trace metals.

109 evaluate recent sediment accumulation and mixing rates, while trace metals chronicled

Benoit et al. - 6 110 historical contaminant loading and served as fingerprints of sediment sources. Radionuclides 111 were measured by non-destructive gamma counting in sealed 100 ml aluminum cans. For 112 trace metal analysis, sediments were digested with concentrated HF and HN03 in sealed 0

113 microwavable Teflon bombs. A pressure of 725 psi and a temperature of200 C were 114 reached during a 2 min exposure in a 600 w oven (Kingston and Jassie 1988). The resultant 115 solutions were evaporated to near dryness -and then brought into solution in 0.5% HNO3 for 116 measurement by ICP-AES.

117 Results And Discussion 118 JORDAN COVE STRATIGRAPHY 119 The stratigraphy of Jordan Cove is similar to that of other small coves along the north 120 shore of Long Island Sound (Curewitz et al. 1992; Patton 1994) and reflects both the glacial 121 history of the region and the submergence of these former upland valleys during the Holocene 122 rise in sea level. Several distinct stratigraphic units are present in the vibracores and these 123 units can be readily correlated along the axis of the cove (Fig. 2).

124 At the base of several cores (JC-4, 8, 12)are oxidized sand and gravel deposits. In 125 cores 8 and 12 there is a gradation in weathering and sediment color that suggests soil 126 development. At the base of core JC-4 the sand and gravel has the poor sorting and 127 compaction characteristic ofglacial till. A till interpretation is also consistent with the 128 location of the core near the margin of the cove, adjacent to the glacial till mantled uplands.

129 The other sand and gravel deposits are better sorted and likely represent alluvial deposits 130 associated with the ancestral Jordan Brook. These sand and gravel deposits mark the 131 subaerial surface of the pre-submergence landscape.

Benoit et al. - 7 132 The oxidized sand and gravel deposits are capped by gray to brown sand and sandy 133 mud deposits. There are distinct sedimentation units with layers ranging in thickness from a 134 few centimeters to up to 50 centimeters thick for example in core JC-7. The deposits are 135 laminated and often contain interbedded layers of fine-grained organic debris. These deposits 136 are interpreted as the basal transgressive estuarine deposits. In three cores, JC-8, 11 and 12, 137 shell debris occurs stratigraphically above this transgressive sand. In cores JC-1 I and 12 the 138 shell debris is up to 40 cm thick and is predominantly oyster (Ostrea) shells. It is not clear 139 from the vibracores whether these shell layers represent reworked shell or are, in fact,' small 140 oyster beds. Radiocarbon ages of these shell layers are shown on Fig. 2 and illustrate the 141 time-transgressive nature of the submergence of the cove with younger ages for the estuarine 142 deposits found at shallower depths and more landward in the cove.

143 The predominant sedimentary unit in the cove is a thick sequence of organic-rich black 144 mud which extends from the substrate of the modem cove downward to the basal units 145 described above. The black mud unit has a maximum thickness of slightly more than 5 m in 146 Core JC-7 and thins toward the head of the cove. The mud unit is not entirely uniform, 147 occasional sand layers and thin shell layers are present within the mud sequence. The 148 radiocarbon ages on the underlying shell deposits indicates that the onset of mud deposition is 149 also time transgressive. The presence of sandy estuarine deposits beneath the black mud unit IS0 suggests that the onset of mud accumulation is, in part, related to increasing water depth in 151 the cove.

152 The mud unit also occurs in cores 4 and 5 where it is capped by the sand shoal and 153 marsh deposits at the mouth of the cove. The stratigraphy of core 4 indicates that the niarsh.

154 islands present on the 1893 map were established on the earlier sand spits present in 1868

Benoit et al. - 8 S ,

155 which, in turn, had prograded across a muddy substrate that was present in the open water of 156 the cove. Prior to 1868, mud deposition in the cove may have resulted from a protective 157 barrier bar present farther to the south, as indicated by the sand units present in core 2. The 158 medium to coarse grained gray sand unit which caps the black mud deposit in cores 2, 4, and 159 5 represents the modem sand shoal and spit environment at the mouth of the cove. Although 160 the data are not definitive, it appears that the black mud unit interfingers with and pinches out 161 against the sand deposits at the mouth of the cove.

162 Based on limited radiocarbon dating, the submergence history of the cove dates to 163 before 3,800 yrs b.p. A similar sequence of black mud deposition can be found in Quiambog 164 Cove, a small narrow cove located east of the Thames River, and 18 km to the east of Jordan 165 Cove (Curewitz 1992). At the mouth of Quiambog Cove, at a depth of 8 m below present 166 mean low water, the beginning of mud deposition dates to 5,020 + 80 yrs b.p. Farther 167 landward in Quiambog Cove, mud deposition at a depth of 5 m below mean low water began 168 slightly after 3,915 +/- 75 yrs b.p. In Jordan Cove, the radiocarbon date in core 8 indicates that 169 mud deposition at about 5 m below mean low water began at 3,780 + 70 yrs b.p. The 170 consistent ages for the onset of mud deposition between these two coves suggests that 171 regional submergence is the main control on mud deposition and not some change in the 172 sediment supply to these valleys.

173 STRATIGRAPHIC SUMMiARY 174 The submergence of Jordan Cove is marked by a sand substrate that likely resulted 175 from wave energy reworking the glacial and post-glacial deposits on the margins of the 176 expanding cove. These shallow fringing environments also concentrated shells, and thereg may 177 have been oyster beds within the cove prior to 2,500 yr b.p. As water depth increased the

Benoit et al. - 9 178 black mud unit was deposited uniformly across the open water environments of the cove. The 179 higher energy conditions at the mouth of cove prevented mud deposition.and instead sand 180 shoals, sand spits, and marshes developed there. The marsh islands present at the mouth of 181 the cove at the turn of the century, may have formed in the lee of earlier sand spits. As the 182 cove continues to be submerged, these sand dominated environments will continue to shift 183 northward toward the head of the cove. This can be seen in the coarsening upward 184 stratigraphy evident in Cores 2 - 5 at the mouth of the cove.

185 LONG-TERM SEDIMENTATION RATES 186 Long term sedimentation rates for the black mud unit can be calculated from the 187 radiocarbon ages for cores 8, I1, and 12 (Table 1). The calculated rates in each core are 188 based on a single age at one depth and assume that the measured shell or wood had a 189 radiocarbon age of zero when it was at the surface. In some cases surface sediments can 190 appear old by 4C Cbecause they contain terrestrial carbon that spent decades or even centuries 191 in soils before being eroded and transported to their final deposition site (e.g., Benoit et al.

192 1979). However, it seems unlikely that shells or wood would be subject to this kind of dating 193 artifact. The long-term sediment accumulation rates are very low, averaging less than 1 nun 194 yr , which is less than half the measured long-term sedimentation rates recorded in similar 195 cove environments (Patton 1994). Measured long-term submergence rates for the 196 Connecticut coast for most of the past 4,000 years have been approximately 2.5 mm yr", with 197 a period between 1,500 and 300 yr b. p. when the submergence rate was as low as 1.0 mm 198 yr'. This means that for the past 3,700 years, Jordan Cove has been filling with sediment at a 199 rate not greater than the submergence rate, and for the most part, significantly slower. The..

Benoit et al. - 10 200 long term sediment accumulation data argue that water depths in Jordan Cove should have 201 been increasing throughout the late Holocene.

202 RECENT SEDIMENT ACCUMULATION RATES: 2' 0Pb AND '"Cs 203 Recent sedimentation history was evaluated through measurements of210Pb and "7Cs 204 for short cores JC-I, 6, 9, 13, 15, and 16. These radionuclides can provide detailed 205 information on processes occurring in the past 100 yr. A clear pattern is that sediment 206 accumulation rates (SAR) decrease with distance from the mouth of the cove, declining by 207 nearly a factor of 4 from 0.23 to 0.84 cm yRX (Fig. 3, Table 2). Because of dissimilar depths of 208 penetration and SARs, the six cores record sedimentary history over different time periods.

209 These range from a high of 63 yr for core 15 to a low of 25 yr for core 6. No radionuclide 210 data were available from site 1, since the sandy sediments there contained immeasurably low 211 levels and seemed to be well-mixed, obscuring the sedimentary record. Site I is in the active 212 surf zone, unlike the other locations, which are more sheltered. Radionuclides were not 213 measured on core 16 from above the dam, except for "0Co, which was found to be below the 214 detection limit.

215 Bomb-produced 13Cs provides two age makers useful for calculating sediment 216 accumulation rates: first appearance in the atmosphere (1954), and peak production just 210 217 before the atmospheric weapons test ban treaty (1963). Both 3"Cs markers agree with Pb 218 results except for core 15 (Table 2). Also, only a lower limit SAR could be calculated for 219 core JC-6 with '3TCs, since the 1963 maximum was below the bottom of the core. Core 15 220 had unexplained changes in slope in the 210Pb profile, and a very poorly defined '37Cs peak.

221 Because of these anomalies and other evidence, we conclude that the SAR in this core is-near w 210 222 0.4 cm yr' (discussed later). Nevertheless, the overall concordance of the Pb and 137Cs

Benoit et al. - 11 223 dates in the other cores implies that mixing, which influences t37Cs ages differently from 2' 0Pb 224 ones, is not extensive. The agreement between radionuclides also increases confidence in the 225 reliability of the calculated SARs.

226 210Pb profiles generally show year-to-year variability that is greater than the 227 measurement uncertainty. The smoothness of data for core 15 provides an indication of the 228 resolution of this technique when the 2" 0Pb input function is steady. The much greater 229 variability for cores 6, 9, and 13, and the changes in slope for core 15, probably result from 230 real-world fluctuations rather than random errors in the data. These variations mean that 231 calculated SARs need to be interpreted cautiously.

232 Measured 210Pb inventories for cores 6, 9, 13, and 15 are 36, 40, 43, and 33 dpm cm 2 .

0 233 None of the cores contains a complete record of 210Pb deposition, since some excess 2 Pb is 234 present below the bottom of each core. Assuming 210Pb input rates were constant over the 235 last century or so, it is possible to extrapolate and estimate the total inventories that should be 236 present over the entire depth, viz. 68, 56, 55, and 41 dpm cm' 2 for cores 6, 9, 13, and 15.

237 Direct atmospheric deposition supplies about 1 dpm cnf2y tl in the northeastern U.S. (Turekian 238 et al. 1983), and this flux can support a sedimentary inventory of 32 dpm cm72. Cores 6, 9, 2 0 239 13, and 15 all contained significantly more than this amount of 210Pb. The surplus ' Pb could 240 be supplied either from lateral sediment focusing (Hilton 1985; Lehman 1975) or suspended 241 matter bearing unsupported 210Pb carried into the cove from Jordan Brook or Long Island 242 Sound.

243 If Jordan Brook is the main supplier of sediment to the cove, then the most rapid 244 sediment accumulation should occur near this source at the head of the cove, unless 4:

• 245 hydrodynamic factors intervene. The latter possibility seems unlikely since currents within the

Benoit et al. - 12 246 cove are mainly tidal and are strongest near the mouth. An alternative is that Long Island 247 Sound is the origin, and the decrease in SARs results from a diminishing supply of sediments 248 with distance from the source. In any event, the observed SAR distribution pattern seems to 249 argue against the possibility that restricted flow has reduced flushing, thereby causing river-250 borne sediments to accumulate more rapidly.

251 It is interesting to compare recent SARs in Jordan Cove to the regional rate of relative 252 sea level rise, which has been measured independently in a number of coastal salt marshes.

253 Data suggest that sea level has been rising at about 0.3 cm yrl for the past 400 yr (Nydick et 254 al. 1995; Patton 1994). Most of Jordan Cove (excluding perhaps the area near core 15) has 255 been filling at a significantly faster rate, outstripping regional inundation. If conditions 256 continue unchanged, the cove will shoal at a rate of about 50 cm/100 yr near its mouth, 30 257 cnd100 yr in its mid section, and 10 cm/100 yr near its head. These rates are too slow to have 258 been noticeable in a single human generation. Nevertheless, since the cove is no more than 2 259 m deep on average, it seems likely that the embayment will evolve into a mudflat or salt marsh 260 system within several centuries.

261 COMPARISON OF LONG-TERM AND SHORT-TERM SEDIMENT ACCUMULATION 262 Radiocarbon evaluates sediment accumulation rates over time periods of centuries, 263 while 137Cs and 'Pb measures rates over the past several decades. At some point in the 264 period between approximately 3,000 yr b.p. and the current century, sediment accumulation 265 rates in Jordan Cove increased from about I mm yr4 to nearly .cm yr', and water depths 266 switched from increasing to decreasing with time. The exact timing of the change cannot be 267 established based on the existing data, butthe possibility exists that installation of one or more 268 of the three restrictions to the cove caused the increase in sediment accumulation.

Benoit et al. - 13 269 It should be noted also that the apparent difference between short and long-term 270 sediment accumulation may not be as great ai it first appears. The absolute difference as 271 measured in units of depth per time is probably partly tempered by higher bulk densities 272 deeper in the sediments. This increase in density with depth would cause mass accumulation 273 rates to be more similar than et1h accumulation rates.

274 "0Co INVENTORIES 275 60Co was present in measurable quantities in sediment cores as far inland as the dam, 276 but not above it (Table 3). Since the '0 Co is contributed almost certainly by releases from the 277 Millstone nuclear power plant, this is compelling evidence for Long Island Sound as a source 278 of sediments in the cove. Reactor-derived 60Co has previously been used to trace sediment in 279 the Susquehanna River-Chesapeake Bay system (Donoghue et al. 1989; McLean and 280 Summers 1990). In Jordan Cove, inventories decline in a landward direction from 5.0 dpm 281 cr 2 at the mouth to 1.0 dpm cm 2 below the dam. Decay-corrected activity increases 282 monotonically with depth as far as the deepest layer where '0 Co is detectable (at an average 283 age of22 yr B.P.). Because of 0Co's short 5.3 yr half-life, all cores are longer than the depth 284 of maximum ' 0Co burial, i.e., there is a complete record of 0 Co deposition in all cores. The 285 declining input suggests that the '"Co may be the result of a release that occurred more than 286 two decades ago.

287 "0Co can be used as a tracer to evaluate the proportion of Jordan Cove sediment that is 288 derived from Long Island Sound. Assumptions in this analysis are that: 1) ' 0Co uniformly tags 289 Long Island Sound particles before they enter Jordan Cove, 2) '0 Co exchange between 290 particles and the dissolved phase within the cove does not significantly alter the "0Co .* * -

291 concentration on solids (Byrd et al. 1990; Chiffoleau et al. 1994; Flegal et al. 1991; Muller et

Benoit et at. - 14 292 al. 1994), 3) 60Co is not lost from deposited sediments, and 4) ' 0Co-tagged particles are 293 transported landward at the same rate as all other particles. If these assumptions are correct, 294 then the ratio of 0 Co inventory to SAR should be constant if all the sediment deposited in the 295 cove is derived from the Sound. Significant input of sediment from the watershed would 296 decrease the '60Co:SAR ratio. The measured ratios for cores 6, 9, 13, and 15 are actually 6.0, 297 6.7, 5.6, and 2.5, respectively. These results suggest that "Co-tagged sediments from Long 298 Island Sound provide a constant proportion of bottom sediments as far inland as station 13.

299 At station 15 there seems to be dilution by "Co-free sediments, perhaps from Jordan Brook 300 (or one or more of the assumptions is violated). The "0Co data provide strong evidence that 301 sediment in most of the cove is supplied from Long Island Sound, and that some of the 302 sediment at the head of the cove may be supplied from a terrestrial source.

303 SEDIMENT QUALITY 304 Trace metal concentrations in sediments (Table 3) were generally higher than crustal 305 abundances (normalized to Fe), suggestive that some of each metal derives from contaminant 306 sources. Metal levels in Jordan Cove can be compared with other sites in Long Island Sound 307 and the rest of the US coast (Robertson et al. 1991). Compared to samples collected for 308 NOAA's Status and Trends program, in Jordan Cove Pb and Zn bracket the national mean, 309 Ag and Cu largely fall between the mean and the mean plus 1 S.D., while only Cd is frequently 310 greater than the mean plus 1 S.D. (a 1.3 mg kg'). Compared to other sediments in the -

311 region, average Ag, Cu, Pb, and Zn levels in Jordan Cove are lower than in contaminated 312 harbors in the western Sound, and they are similar to nearby Connecticut River sediments. Of 313 metals tested, only Cd had average concentrations as high as in contaminated western harbors 314 near the influence of New York City.

Benoit et al. - 15 315 PATTERNS OF TRACE METAL DISTRIBUTION WITH DEPTH 316 Core 15 contains the longest record of metal deposition (Fig. 4), and is used as an 317 example to illustrate metal variations over time. During most of the past 60 yr, Ag and Cu 318 levels have increased, but only slightly. For both metals, the data also suggest the possibility 319 of reduced inputs over the last decade. Both Cd and Zn have risen steadily, and are currently 320 about double corresponding concentrations of a half century ago. Lead concentrations grew 321 continuously and dramatically from 1935 (the base of the core) until the mid 1970's, 322 increasing by almost a factor of 6. The peak corresponds to the time of maximum use of Pb 323 anti-knock compounds in gasoline. After 1975, Pb in core 15 declined briefly, but has 324 remained nearly constant for more than a decade at about four times the.pre-WW II value.

325 Overall, trace metals in the cove are significantly above background but are similar to regional 326 values and are not at dangerously high levels. Upward trends in Cd and Zn should be 327 monitored in the future. Comparing Fe normalized trace metals in surficial (0 - 5 cm) 328 sediments to ones collected by Hunt (1979) in the 1970s from nearby Long Island Sound 329 locations showed that Jordan Cove values for Cu, Pb, and Zn were all within 300/0 of their 330 Sound counterparts. Of metals measured by both Hunt and us, only Cd was much greater (by 331 a factor of 8), and this difference is discussed later.

332 Comparisons of trace metals show that many of the same features occur from core to 333 core. Figure 5 is a plot of Zn against depth for cores 9, 13, and 15. The depth scale has been 334 stretched for each core to align features such as maxima and minima. Clearly there is a very 335 strong correlation among Zn profiles in the various cores. Changes in total Fe explain only 336 part of the variation in Zn. Significantly, the depth scale for core 15 in relation to 9 and .13 .

337 indicates a sedimentation rate near 0.45 cm yr'. This value is close to the one derived from

Benoit et al. - 16 338 the first appearance of 137Cs in core 15, but higher than suspect rates based on the "'Cspeak 339 and the 21 0Pb depth profile C(able 2).

340 TRACE METALS AND MAJOR SEDIMENTARY COMPONENTS 341 In Jordan Cove, most of the metals correlate well with Fe and with each other (Figs. 6 342 & 7). The r2 values for regressions on Fe were 0.30, 0.74, 0.39, and 0.52 for Ag, Cu, Pb, and 343 Zn, respectively. With sample sizes over 70, P values were all much less than 0.001.

344 Cadmium was poorly correlated with Fe (r2 = 0.0015). The correlation was improved, but 345 only slightly, by excluding core 1, which had an anomalously high Cd:Fe ratio. The relation 346 between Fe and the other metals may well be causal, since Fe has been used as a surrogate for 347 fine-grained sediments, which tend to bind more metal (Morse et al. 1993; Rule 1986).

348 Metals have also been observed to correlate with sedimentary organic matter (due to surface 349 complexation) and Al, an indicator of clays (Windom et al. 1989). In Jordan Cove neither 350 organic matter nor Al could explain much of the variability in the metals. For example, e 351 values for linear regressions of Pb with Al and OM were 0.01 and 0.05 respectively (Fig. 6).

352 Interestingly, Hunt (1979) found that organic matter content had a higher correlation 353 with trace metals than did Fe for surficial sediments collected from around Long Island Sound.

354 This difference may reflect dissimilarities between areal input patterns (which are probably 355 related to water column scavenging) and down-core changes (which may be the result of 356 diagenetic processes). It could also be caused by local conditions at the Jordan Cove site.

357 For example, if substantial amounts of organic matter in the cove are supplied from the local 358 watershed, then this might mask metal-organic matter relations that operate at larger scales in 359 Long Island Sound.  :

Benoit et al. - 17 360 Trace metals correlated well with each other. For example, the average r2 value 361 among Cu, Pb, and Zn exceeded 0.4 (Fig. 7; Cu-Pb 0.42; Zn-Pb 0.34; Zn-Cu 0.47; P <<

362 0.001). Silver and Cd had weaker correlation with the other metals and with each other (Fig.

363 8; Ag-Pb e = 0.25, P << 0.001; Cd-Pb r2 = 0.07; Cd-Ag r2 = 0.08). The strong correlations 364 among Ag, Cu, Pb, and Zn in part reflect a shared dependence on the abundance of Fe, but 365 even when metals are normalized to Fe, correlations persist. In this analysis, data for cores 1 366 (surf zone sand) and 16 (sediment above the dam) were excluded, since extreme values there 367 tended to cause spuriously strong correlations. On the edited data, the correlations for Cu-Pb, 368 Zn-Pb, and Zn-Cu all have P values of 0.001 or less even when metals are first normalized to 369 Fe. In addition, a correlation between Cd and Zn at the <0.001 confidence level emerges that 370 was previously masked by Fe systematics. Silver correlations with both Cd and Pb improved, 371 but were weaker than for other metals (P > 0.01) 372 The correlation of the metals suggests either that they come from a common source, 373 or that the sediments are mixed within the cove laterally before they are buried. The latter 374 possibility seems unlikely, since it should influence all metals equally, while only a subset are 375 well-correlated. Copper, Pb, Zn, and perhaps Ag thus seem to have one source, while Cd has 376 another. Historically, contaminant Pb was derived principally from the burning of gasoline 377 containing anti-knock additives, while Cu and Zn had other sources. Their correlation implies 378 that the metals are mixed in some environmental compartment before delivery to the cove.

379 Based on the several lines of evidence already presented, this reservoir is likely to be the 380 sediments of Long Island Sound.

381 Another possible explanation for the similarity of metal profiles within cores (and by.*

• 382 inference input functions) is that the metals tend to be used, and released to the environment,

Benoit et al. - 18 383 in parallel. For example, the onset of the industrial revolution, World War II,and periods of 384 strong economic growth lead to greater use and discharge of many metals. In contrast, the 385 Great Depression and the recent era of improved environmental regulation should both be 386 times of lowered releases across all metals. Thus, broad societal trends rather than 387 biogeochemistry may explain some of the profile features.

388 TRACE METALS AS TRACERS 389 Trace metals were also evaluated for their potential to reveal sediment sources. If 390 trace metals that are bound to sediments do not react significantly during transport, metal 391 ratios can be used to indicate sediment provenance (Benoit et al. 1998, submitted; Eaton et al.

392 1980; Krumgalz 1993; Shine et al. 1995). For this method to work, each sediment source 393 must have a unique and consistent ratio of trace metals. For most of the metals, almost all of 394 the data fall near a straight line passing through the origin (Fig. 7). This pattern is consistent 395 with a model where a single source endmember sediment mixes with sediment nearly free from 396 metals.

397 Silver and Cd provide two exceptions to a single endmember source model. Silver 398 levels (compared to Fe or Pb) were much lower above the dam than in all cores in the estuary 399 (Figs. 8, 9). This deficit implies that the-source of Ag to the cove is from the Long Island 400 Sound direction, or that Ag has been preferentially rernobilized and lost from sediments behind 401 the dam. For comparison, a sediment sample from near the mouth of the Connecticut River is 402 plotted on Fig. 8 (Robertson et al. 1991) and it clearly resembles sediment from within the 403 cove. Silver has been proposed in the past as an especially sensitive indicator of 404 contamination (Benoit 1994; Benoit and Rozan 1995a; Benoit and Rozan 1995b; Benoitand4.

405 Rozan 1996; Sanudo-Wilhelmy and Flegal 1991; Sanudo-Wilhelmy and Flegal 1992).

Benoit et al. - 19 406 Because Ag's background level is so low, its presence almost always indicates a 407 contamination source. In Jordan Cove that source seems to be in the marine environment, 408 perhaps originating from regional sewage effluent. The outfall for the city of New London, 409 Connecticut, is located 9 km to the east, and currents in this region are generally from the east 410 (F. Bohlen, Univ. of Conn., pers. comm.).

411 The other exception is Cd. Both Cd:Fe and Cd:Pb ratios are much higher in sediments 412 at the mouth of the cove than at locations farther inland (Figs. 8, 9). As mentioned earlier, 413 even the inland Cd levels are high compared to nearby Long Island Sound sediments, when 414 both are normalized to Fe. The excess Cd in sediments near the cove's mouth would seem to 415 indicate the existence of an additional localized source of Cd there. Curiously, this 416 observation seems to work in direct contradiction to the known behavior of Cd in estuaries, 417 where Cd tends to be released from river-supplied sediments, apparently through formation of 418 dissolved chloro complexes (Byrd et al. 1990; Comans and van Dijk 1988). Remobilization 419 should tend to cause lower values in sediments, not elevated concentrations as were observed.

420 As was described in previous sections, Cd in Jordan Cove is higher compared to regional 421 norms than are any of the other metals. Perhaps both observations (anomalously high Cd at 422 the cove mouth, and elevated values throughout the cove) can be explained by a local source, 423 such as the fill material used there. Another possibility is that a diagenetic remobilization-424 scavenging cycle enriches Cd relative to other metals on the sand surfaces.

425 All of the trends noted in this section become more pronounced when trace metals are 426 plotted relative to each other after first being normalized to Fe (Figs. 10 - 12). This 427 manipulation reduces the covariation with Fe that may mask other important trends. On.-these 428 figures, cores 6, 9, 13, and 15 plot as a tight cluster, while cores I and 16 tend to occupy

Benoit et al. - 20 429 clearly separate domains. The cluster is especially interesting considering that it includes data 430 for all depths in the cores. It appears that trace metal ratios have remained relatively constant 431 over several decades during which total metal levels have changed substantially (e.g., Figs. 4 432 & 5). The separate domains are especially apparent for Cd in core 1 (Fig. 10) and Ag in core 433 i6 (Fig. I1). In addition, both Zn and Cu seem deficient in core 1,and Zn appears to be 434 anomalously high and Cu low in core 16 when compared to otherFe normalized metals (Fig.

435 12). An intuitively reasonable model would be if other Jordan Cove cores were derived by a 436 mixture of endmembers represented by cores 1 and 16, however that model seems to fald. The 437 within-cove cores do not generally plot as a simple linear mixture of core 16 (terrestrial 438 sediment) and core I (mouth of the cove). The latter is nearly pure sand, perhaps with metal 439 oxyhydroxide surface coatings, and may not faithfully represent a Long Island Sound 440 endmember.

441 Trace metal data are available for a core in Long Island Sound not far from Jordan 442 Cove near the mouth of the Connecticut River (Greig et al. 1977). Trace metal ratios in this 443 core are plotted as a dotted line in Figs. 11 and 12. (Since Fe data are not available, it is not 444 possible to determine where along this line the core should plot.) In all cases this line passes 445 through the cluster of data delimited by cores 6, 9,13, and 15, suggesting that sediment in the 446 main part of Jordan Cove is similar to Long Island Sound sediment. If cove sediments are 447 derived from mixing of endmenbers similar to the Long Island Sound core and Jordan Brook, 448 then a quantitative apportionment might be obtained from the relative distance that the cluster 449 falls between the Long Island Sound line and core 16. For example, a possible mixing line has 450 been drawn on the Ag-Pb graph of Fig. 11. Since the mnain cove cluster falls about 25% of the 451 distance from the Long Island Sound line to the terrestrial cluster (core 16), roughly 75% of

Benoit et al. - 21 452 these sediments seem to be derived from Long Island Sound and 25% from a local watershed 453 source. Rigorous application of this method would require more reliable data on the 454 endmembers' metal content, and an objective definition of the cluster's center, but the 455 principle would be the same as applied here.

456 CONCLUSIONS 457 1. For at least 3,800 yr, rising sea level has gradually inundated Jordan Cove, filling it with 458 mud similar to that currently being deposited there.

459 2. Long-term sediment accumulation in the cove averaged close to 1 mm ye' over the last 460 three millennia. This is slower than regional relative sea level rise and the cove 461 deepened.

462 3. Recent (past 50 yr) sediment accumulation rates in Jordan Cove decrease progressively 463 inland from the mouth, declining from 0.84 to 0.40 cm yrl.

464 4. The current SAR is faster than relative sea level rise and Jordan Cove will fill in within a 465 few centuries unless conditions change.

466 5. "0Co derived from the Millstone nuclear. power plant is present in the sediments, and its 467 inventory decreases with distance from the mouth of the cove, proportional to the local 468 sediment accumulation rate. a0Co data suggest that bottom sediments are derived 469 mainly from Long Island Sound.

470 6. In terms of trace metal content, sediment quality in the cove is degraded compared to 471 pristine sites, but metals have not reached dangerously high levels.

472 7. Similarity of trace metal distributions with depth confirmed relative sediment 473 accumulation rates among cores and was used to assign a value to an ambiguous core...

Benoit et at. - 22 474 8. Cu, Pb, and Zn data correlate strongly with Fe, but not with either organic matter or 475 aluminum.

476 9. Ratios of Ag to Fe and to trace metals suggest that Ag in the cove is derived almost 477 entirely from Long bsland Sound. This result supports the notion that Ag can serve as a 478 ' better tracer of some kinds of contamination than more common and abundant metals, 479 like Cu. Pb, and Zn.

480 10. Plots based on trace metals normalized to Fe reveal similar groupings among the cores 481 and may be useful for quantitatively apportioning the sources of estuarine sedinients 482 even in locations where sediments are only lightly'contarninated.

483 11. For several trace metals, core 16 from above the dam is unlike other cores in the cove.

484 Core 1, a sand deposit near the mouth of the cove, is different from cores within the 485 cove, above the dam, or in greater Long Island Sound.

486 ACKNOWLEDGMENTS 487 This work was conducted, in part, through a Sounds Conservancy (currently Atlantic 488 Center for the Emvironrnent) grant awarded to TFR, and a grant from the Long Island Sound 489 Research Fund, Conn. DEP, to PCP).

I-f

Benoit et al. - 23

• 490 LITERATURE CITED 491 Benoit, G. 1994. Clean technique measurement of Pb, Ag, and Cd in fresh water. A 492 redefinition of metal pollution. Environmental Science and TechnoloQy28:1987-1991.

493 Benoit, G. and T. F. Rozan, 1995a. The biogeochemistzy of silver in an estuarine system, 3rd 494 International Conference on the Transport, Fate, and Effects of Silver in the 495 Environment, pp. 213-216.

496 Benoit, G. and T. F. Rozan, 1995b. Silver biogeochemistry in river-estuary systems, 10th 497 International Conference on Heavy Metals in the Environment, Hamburg, Germany, 498 pp. 69-72.

499 Benoit, G. and T. F. Rozan. 1996. Silver as a tracer of erosion and sedimentation processes in 500 a coastal river. E9_07: S167.

501 Benoit, G., K. K. Turekian and L. K. Benninger. 1979. Radiocarbon dating of a core from 502 Long Island Sound. Estuarine Coastal and Shelf Science 9: 171-180.

503 Benoit, G., E. X Wang, W. C. Nieder, M. Levandowsky and V. Breslin. 1997, submitted.

504 Sources and history of heavy metal contamination and sediment deposition in Tivoli 505 South Bay, Hudson River, NY. Estuaries.

506 Byrd, L. T., K W. Lee, D. S. Lee, R. G. Smith and H. L. Windom. 1990. The behavior of 507 trace metals in the Geum estuary. Korea. Estuaries 13: 8-13.

508 Chiffoleau, J., D. Cossa, D. Auger and L Truquet. 1994. Trace metal distribution, partition 509 and fluxes in the Seine estuary (France) in low discharge regime. Marine Chemistry 47:

510 147-158.

511 Comans, R. N. . and C. P. J. van Dijk. 1988. Role of oomplexation processes in cadmiuqi ..

512 mobilization during estuarine mixing. Nature 336: 151-154.

Benoit et al. - 24 513 Curewitz, D. 1992. The late Quatemary stratigraphy of three coves on the north shore of 514 Fishers Island Sound. Undergraduate Honors Thesis, Wesleyan University, 515 Middletown, CT, 71 pp.

516 Curewitz, D., N. A. McLoughlin and P.. C. Patton. 1992. Post-glacial stratigraphy and rates 517 of sediment deposition in three coves bordering Fishers Island Sound. Geological 518 Society of America. Abstracts with Programs 24: 15.

519 Donoghue, J. F., 0. P. Bricker and C. R. Olsen. 1989. Particle-bourne radionuclides as tracers 520 for sediment in the Susquehanna River and Chesapeake Bay. Estuarine Coastafand 521 Shelf Science 29: 341-360.

522 Eaton, A., V. Grant and M. G. Gross.; 1980. Chemical tracers for particle transport in the 523 Chesapeake Bay. Estuarine and Coastal Marine Science 10: 74-83.

524 Flegal, A. R., G. J. Smith, G. A. Gill, S. Sanudo-Wilhelmy and L. C. D. Anderson. 1991.

525 Dissolved trace element cycles in the San Francisco Bay estuary. Marine Chemistry 36:

526 329-363.

527 Greig, R. A., R. N. Reid and D. R. WenzlofE 1977. Trace metal concentrations in the 528 sediments from Long Island Sound. Marine Pollution Bulletin 8: 183-188.

529 Hilton, L. 1985. A conceptual framework for predicting the occurrence of sediment focusing 530 and sediment redistribution in small lakes. Limnology and Oceanography 30: 1131-531 1143.

532 Hunt C. D. 1979. The role of phytoplankton and particulate organic carbon in trace metal 533 deposition in Long Island Sound. Ph.D. Thesis, Universityof Connecticut.

534 Kingston, H. M. and L. B. Jassie. 1988. Introduction to Microwave Sample Preparation. CS.C.

535 Professional Reference Book. American Chemical Society, Washington, DC, 263 pp.

Benoit et at. - 25 536 Krumgalz, B. S. 1993. Fingerprints' approach to the identification of anthropogenic trace 537 metal sources in the nearshore and estuarine environment. Estuaries 16: 488495.

538 Lehman, J. T. 1975. Reconstructing the rate of accumulation of lake sediment: The effect 539 sediment focusing. Quaternary Research :.541-550.

540 Lewis, R. S. and J. R. Stone. 1991. Late Quaternary stratigraphy and depositional history of 541 the Long Island Sound basin: Connecticut and New York. Journal of Coastal Research 542 Special Issue #11: 1-23.

543 Lyons, W. B. and W. F. Fitzgerald. 1980. Trace metal fluxes to nearshore Long Island Sound 544 sediments. Marine Pollution Bulletin 11: 157-161.

545 McLean, R. I. and J. K Summers. 1990. Evaluation of transport and storage of Co-60, Cs-546 134, Cs-137, and Zn-65 by river sediments in the Lower Susquehanna River.

547 Environmental Pollution 63: 137-153.

548 Morse, J. W., B. J. Presley, R. J. Taylor, G. Benoit and P. H. Santschi. 1993. Trace metals in 549 Galveston Bay: Water, sediments, and biota. Marine Environmental Research 36: 1-550 37.

551 Muller, F. L. L., M. Tranter and P. W. Balls. 1994. Distribution and transport ofchemical 552 constituents in the Clyde estuary. Estuarine Coastal and Shelf Science 39: 105-126.

553 Nydick, K. R, A. B. Bidwell, E. Thomas and J. C. Varekamp. 1995. A sea-level rise curve 554 from Guitford; Connecticut, USA. Marine Geolog& 124:1374159.

555 Patton, P. C. 1994. Post-glacial stratigraphy and rates of sediment accumulation in three small 556 Connecticut Coves. Final Project Report. Connecticut Department of Environmental 557 Protection, Hartford, CT. CWF266-R.  :

Benoit et al. - 26 558 Robertson, A., B. Gottholm, D. Turgeon and D. Wolfe. 1991. A comparative study of 559. contaminant levels in Long Island Sound. Estuaries 14: 290-298.

560 Rule, J. H. 1986. Assessment of trace element geochemistry of Hampton Roads Harbor and 561 lower Chesapeake Bay area sediments. Environmental Geology and Water Science 8:

562 ' 209-219.

563 Sanudo-Wilhelmy, S. A. and A. R. Flegal. 1991. Trace element distributions in coastal waters 564 along the US-Mexico boundary: relative contributions of natural processes vs.

565 anthropogenic inputs. Marine Chemistry 33: 371-392. -

566 Sanudo-Wilhelmy, S. A. and A. R. Flegal. 1992. Anthropogenic silver in Southern California 567 Bight: A new tracer of sewage in coastal waters. Environmental Science and 568 Technologev 26: 2147-215 1.

569 Shine, J. P., R.V. Ika and T. E. Ford. 1995. Multivariate statistical examination of spatial and 570 temporal patterns of heavy metal contamination in New Bedford Harbor marine 571 sediments. Environmental Science and Technology 29: 1781-1788.

572 Turekian, K. K., L. K. Benninger and E. P. Dion. 1983. 7Be and 21 0Pb total deposition fluxes 573 at New Haven, Connecticut and at Bermuda. Journal of Geophysical Research. 88:

574 5411-5415.

575 Welsh, B. L. and R B. Whitlach. 1980. Jordan Cove: Its hydraulic character, sediment and 576 macrophyte distributions as related to shoaling. Final Report to the Commission of 577 Environmental Protectection, Connecticut Department of Environmental Protection, 578 Hartford CT.

Benoit et at. - 27 579 Wtrndom, R L., S. L. Schropp, F. D. Calder, J. D. Ryan, J. Smith, R.G. and C. FL Rawlinson.

580 1989. Natural trace metal concentrations in estuarine and coastal marine sediments of 581 the southeastern United States. Environmental Science and Technology 23: 314-320.

4V

Benoit et al. - 28 382 FIGURE LEGENDS 583 Fig. 1. Map of Jordan Cove showing sampling locations and flow restrictions caused by 584 transportation arteries. Short cores (= 50 cm) measured for 137CS and 2'0 Pb are indicated 585 by open circles, and deep cores (up to 6 m) used for long-term stratigraphy are indicated by 586 ' filled circles. A nuclear power plant is located on Millstone Point. The location of Jordan 587 Cove along the Connecticut coast is shown in the inset.

588 Fig. 2. Longitudinal profile of Jordan Cove illustrating the sedimentary stratigraphy of the 589 cove. Six different depositional units are recognized: (I) the modem salt marsh teat; (LI) a 590 gray sand and gravel deposit that represents the shoal and spit complex at the mouth of the 591 cove; (III) a thick sequence of black estuarine mud; (IV) a gray laminated sand and sandy 592 mud with abundant shell deposits that represents the transgressive estuarine deposits 593 formed as sea level drowned the post-glacial valley, and (V) oxidized sand and gravel 594 deposits and (VI) diamicton which mark the post glacial surface of the valley. Radiocarbon 595 ages provide age control for the submergence of the valley.

596 Fig. 3 .2 0Pb profiles and inferred sediment accumulation rates. Rates derived in this way 597 generally matched rates derived from " 7 Cs distributions.

598 Fig. 4. Metal variations with depth in core 15, showing typical historical changes. Date axis is 599 based on a presumed SAR of 0.45 cm y' as described in the text in relation to Fig. 5.

600 Fig. 5. Zn vs. depth in cores 9, 13, and 15. Profiles all show similar patterns though at 601 different depths because of varying sediment accumulation rates. The relative depths of 602 peaks and other profile features in core 15 match those in the other cores if its SAR is 603 assumed to be 0.45 cm y'. A typical Fe distribution with depth is also illustrated for.

604 comparison with the trace metals.

Benoit et al. - 29 605 Fig. 6. Pb compared with Fe, Al, organic matter. Lead concentration is clearly related to Fe 606 but not Al or OM.

607 Fig. 7. Cu, Pb, Zn concentrations compared for all short cores. Metals fall along a single 608 trend line that passes through the origin.

609 iTig. S. Ag and Cd correlations with Pb and with each other. Correlations tend to be weaker 610 than for the other metals. A datum for a nearby Long Island Sound site falls along the 611 same trend as the main part of Jordan Cove, but is unlike sediment in core 16, which is 612 from above the dam.

613 Fig. 9. Ag and Cd compared to Fe for all short cores. Ag in core 16 is much lower compared 614 to other sites, and Cd in core I is much higher compared to other sites.

615 Fig. 10. Fe normalized Cd plotted against other Fe normalized trace metals. Most cores fall 616 within a narrow range, while Core 1 is always elevated in Cd. In core 16, Zn seems 617 elevated and Ag deficient compared to other cores. Domain shapes are drawn arbitrarily to 618 capture as many members as possible in a small area.

619 Fig. 11. Fe normalized Ag plotted against other Fe normalized trace metals. Most cores fall 620 within a narrow range, while Core 16 is always deficient in Ag. Core 1 also exhibits 621 anomalous concentrations compared to other cores, being low in Cu and Zn, and high in 622 Cd. For the Ag-Pb plot, the centroid for a cluster of data representing cores 6, 9, 13, and 623 15 is hypothesized to represent mixing between core 16 (local watershed sediments) and a 624 line representing the Ag:Pb ratio in nearby Long Island Sound sediments. A mixing line is 625 drawn as described in the text 626 Fig. 12. Fe normalized Cu, Pb, and Zn plotted against each other. Zn appears to be eleyated 627 and Cu low in core 16, and Cu and Zn deficient in core 1 compared to other cores.

fy r- - - -

I CoIECUT I I . I I I I

I Fig. 1: Jordan Cove 2*.s

• Short corms

• Long cores Millstone PoInt White PoInt U

_ N

• riga .

= r= .=;r,- P=--=m -

1, I

k..!

6

'orth South

{It m V.

sedlmcat key I peat mud sae mu -shcls u-E sand P] gravel l shell E soil I

Q lsoo crs

.T_

I

Fig. 3 CORE 6 CORE 9 10 10 me E

• 0 0M CO "o

.- a a,

0 I I 0 5 10 15 20 25 0 5 10 15 20 25 CORE 13 CORE 15 10 10 E

. ~0coCO 1 a,

I 0

0 0.1 0 5 10 15 20 25 0 5 10 15 20 25 DEPTH (cm) DEPTH (cm)

Fig. 5 Zn (mg/kg) 0 50 100 150 200 0 0 -0

5. -5 5

00-%C:>

I C.) - 10 E

0 10 I I-10 0 0- - 15 a.

w 0- 15 a w LC) 0 15 0 - 20 W 0 20 T-

[] Ilz 0

C) - 25 0 20 25

- 30 25 30 20000 30000 40000 50000 Fe (mg/kg)

I: *- &O.,.

'Fig. 6 140 140 .

120 120 IA 10 0 e = 0.01 A 100 ^

P 00-1 80 0Q 80

-E Cn 60 0-Q-

40

.0 I

" 60 40 20 20 0 0 I . . . . . . .- . . . . . I . -

0.0........

0 10000 20000 30000 0 10000 20000 30000 Fe (ppm) Al (ppm) 140 -

120 - AO Core 1 A

r2= 0.05 100 - Core 6 E R-a 80 - Core 9 0-60 - Core 13

-0 40 - Core 15 20 - Core 16 n- I I

• I

• I

• I

• I v I HI. I I 0 10- 20 30 40 Organic Matter (%).

. e

Fig. 7.

300 - 7 80 -

250 -

A6 E 60- g 200-

0. 150 - A 5 40- A 0

N 100-20 50 -

a 0 . ,.-- . I ,- . .1- .1 1111 O- . I --. I . I a I I 0 20 40 60 80 100 120 0 20 40 60 80 100 120 Pb (ppm) Pb (ppm) 250 -

A6 O Core l 200 -

0 Core 6 E 150- 0 A Core 9 v Core 13 c: 100 - Core 15 N 0 50 - Core 16 0 - ,- I- .

I

,I I

0 20 40 60 80 Cu (ppm)

Fg. 8 3 a o AO EA 16 1 oo' E E Q0 a

%Q.

fa.

Q)

-o 0

0 0 _ I Ia I I I I I I Il 0 20 40 60 80 100 120 0 20 40 60 80 101D 120 4

Pb (ppm) Pb (ppm) 4 3

o Core I 0 Core6 E

A Core 9 2

V Core 13 0-I c Core 15

• Core 16 0

0 .1 2 Ag (ppm)

Fig. 9 I

a 'E la aQ .

6m 0

0 10000 20000 30000 40000 Fe (ppm) 4 3

02 C) 1 0

. 0 10000 20000 30000 40000 Fe (ppm) Of

Fig. 10 4 4 -I 0 3 a 3-C0 0

0 0

Core l 0 0 0) 2 2-U- U- AOthers la 0 1 C. 1-16 N 0 0 . . I . - I 0 10 20 30 40 50 60 70 80 0 10 20 30 40 Pb/Fe*1 0000. Cu/Fe*1 0000 4 , *.. 4 C) 0 3- 0 0

3 - 0 0

C) m9 Core 1 0) 0 CD 2- a) 2-U- LL 16 *0 0 1- Other 0) 1-0 I I I 0 . .

I I I I I I I I I I 0 20 40 60 80 100120140 0 1 2 3 4 5 6 7 Zn/Fe*l 0000 Ag/Fe*l 00000 0 Core I 03 Core 6 A Core 9 v Core 13 To Core 15 -Uf 0 Core 16

Fig. 11 7 A 7- 0 V --..

6 a 0 0:5 5 - -"thers o4 0

o 4 0 3 A 03 U-U-

0)- ma 2

_ c .

I 1 . @nCore16 I.,. Core 16 .

0 0 10 20 30 40 50 60 70 80 0 5 1015202530354045 Pb/Fe*l 0000 Cu/Fe*1 0000 7 7 6 6 C) 0o5 0 5 o 4 o 4 a) 3 o 3 U-U-

,c' 2 I 1 l O0 0

0 20 40 60 80 1001201 40 0 1 2 3 4

-ZnIFe*l 0000 CdIFe*1 0000 O Core l o Core 6 A Core 9 V Core 13 c Core 15 II . * !

U Core 16

Fie-, . Q -

45 - . .

140 40 - .::

120 C) 35 - /0 0 SV 100

0. 30 -

V" 25 - Vothers 80 20 - ?40& HA

• 60 1:5 15 - C 40 10 - N S -

20 0 - F. T-T I I I l 0 I

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80 PbIFe*1 0,000 PbIFe*1 0,000 140 -

120 -

0 100 - 0 0

VW.

80 -

• I*

60 - 'Others 40 -

N 20 -

core 1 n_-I v

I I I I I I 1 1I 0 5 1015202530354045 CuIFe*10,000

• ..

• f

TABLE I TABLE 1. Sediment accumulation rate data for vibracores. Results are calculated assuming both no compaction (ramming) and a uniform compacti6n across all depths.

No Compaction Uniform Compaction vibracore Radiocarbon Depth Sediment Depth Sediment Age Accumulation Accumulation Rate Rate (yr b.p.) (cm) (cm y1) (cm) (cm ya)

JC-8 3,780 +/- 70 405 0.11 465 0.12 JC-11 3,130 +/-70 177 0.06 191 0.06 JC-12 2,550 +/-70 165 0.06 217 0.08

TABLE 2. Sediment accumulation rates for short cores calculated from 21 0Pb and 137 Cs I

1 37 21OPb cs 137cs CORE PEAK BOTTOM Sediment Sediment Sediment Accumulation Accumulation Accumulation Rate - Rate Rate

.(cmy') (cm Y ) (anyY)

JC-6 0.84 > 0.64 -

JC-9 0.61 0.52 0.62 JC-13 0.54 0.49 0.55 JC-15 0.23 0.25 0.40

.: I, ..

4

TABLE 3 TABLE 2. Physical characteristics and metals for short cores.

LOSS ON 210 Pb 'Ra 137CS 6OCo CORE DEPTH INTERVAL BULK DENSITY IGNITION (cm) (g cmo) (%) (dpm g9)

JC-1 0 - 1.9 1.1 1.9 - 2.9 0.5 2.9 3.9 0.6 3.9 - 4.9 0.4 4.9 - 5.9 0.4 5.9 - 6.9 0.6 JC-6 0 - 1.3 0.42 19.7 3.67 0.71 0.259 0.54 1.3 - 2.8 0.74 18.8 3.77 0.74 0.151 0.53 2.8 4.3 0.66 17.2 4.28 0.77 0.292 0.39 4.3 - 5.8 0.83 8.4 2.55 0.77 0.14 0.45 5.8 7.3 0.80 9.3 3.51 0.76 0.174 0.44 7.3 - 8.8 0.73 12.3 3.09 0.84 0.311 0.40 8.8 - 10.3 0.85 7.9 3.16 0.84 0.235 0.29 10.3 - 11.8 0.82 8.1 2.83 0.83 0.328 0.43

- 13.3 0.91 6.5 3.14 0.84 0.419 .0.45 11.8 .U 13.3 - 14.8 0.89 6.8 2.44 0.83 0.433 0.29

TABLE 3

" 0Pb 22

'Ra 13Cs OwCo CORE DEPTH INTERVAL BULK LOSS ON DENS1lY IGNITION (cm) (g cmI) (%) (dpm g9) 14.8 - 16.3 0.86 6.8 2.57 0.71 0.39 0.23 16.3 - 17.8 6.3 2.75 0.73 0.337 0.20 17.8 - 19.3 6.6 1.93 0.72 0.409 19.3 - 20.8 8.7 2.56 0.80 0.452 JC-9 0 - 1.7 0.34 15.3 4.77 0.62 0.222 0.51 1.7 - 2.7 0.23 18.6 4.38 0.75 0.172 0.47 2.7 - 3.7 0.36 26.6 3.16 0.66 0.339 0.86 3.7 _ 5.2 0.57 26.5 4.94 0.83 0.273 0.70 5.2 - 6.7 0.62 15.8 4.39 0.69 0.237 0.54 6.7 - 8.2 0.71 15.6 5.10 0.72 0.293 0.64 8.2 - 9.7 0.69 23.5 4.10 0.61 0.277 0.42 9.7 - 11.2 0.67 13.5 3.47 0.65 0.431 0.42 11.2 - 12.7 0.67 33.8 3.27 0.72 0.408 0.42 12.7 _ 14.2 0.64 20.4 3.24 0.73 0.429 0.16 14.2 - 15.7 0.69 13.8 2.33 0.74 0.44 0.17 15.7 - 17.2 0.65 19.9 3.98 0.69 0.724 17.2 - 18.7 0.68 12.3 2.94 0.85 0.322' *i, 18.7 - 20.2 12.7 3.17 0.87 0.376

TABLE 3 I

210Pb 'Ra 13 7Cs 'Co CORE DEPTH INTERVAL BULK LOSS ON DENSITY IGNITION.

(cm) (g co3 ) (%) (dpm e) 20.2 - 21.7 12.1 2.60 0.71 0.285 21.7 - 23.2 12.1 1.70 0.86 0.191 23.2 - 24.7 12.3 1.86 0.78 0.155 24.7 - 26.2 17.4 JC-13 0 - 1.7 0.25 24.3 4.44 0.53 0.155 0.35

.1.7 - 2.7 0.24 16.3 4.29 0.60 0.209 0.19 2.7 - 4.2 0.58 21.2 4.03 0.65 0.365 0.38 4.2 - 5.7 0.55 14.2 4.37 0.59 0.462 0.34 5.7 - . 7.2 0.63 13.9 5.03 0.59 0.437 0.50 7.2 _ 8.7 0.68 13.7 4.48 0.72 0.509 0.51 8.7 - 10.2 0.76 17.6 3.71 0.68 0.407 0.40 10.2 - 11.7 0.70 13.2 4.64 0.67 0.465 0.26 11.7 - 13.2 0.68 13.1 3.54 0.69 0.556 0.26 13.2 - 14.7 0.70 12.7 2.59 0.69 0.508 0.23 14.7 - 16.2 0.72 12.9 3.59 0.75 0.647 16.2 - 17.7 0.69 12.8 2.74 0.72 0.51 17.7 -. 19.2 12.8 2.69 0.69 0.333 S 19.2 - 20.7 13.3 3.32 0.75 0.389

TABLE 3 s.

210Pb 'Ra 137Cs 'Co CORE DEPTH INTERVAL BULK LOSS ON DENSITY IGNITION (cm) (g cm3 ) (%) (dpm g.i) 20.7 - 22.2 13.1 1.91 0.75 0.192 22.2 - 23.7 12.6 2.28 0.69 0.148 23.7 - 25.2 12.3 1.39 0.79 0.Z13 JC-1 5 0 - 1.1 0.27 22.8 6.30 0.82 0.45 0.16 1.1 - 2.1 0.36 21.4 5.64 0.84 0.50 0.12 2.1 - 3.1 0.36 15.2 4.86 0.76 0.414 0.25 3.1 4.6 0.63 17.4 4.56 0.82 0.429 0.17 4.6 _ 6.1 0.60 14.2 3.97 0.81 0.448 0.26 6.1 7.6 0.66 16.2 4.25 0.89 0.381 0.22 7.6 - 9.1 0.72 14.3 4.28 0.76 0.482 0.19 9.1 - 10.6 0.74 12.3 4.07 0.75 0.439 10.6 - 12.1 0.76 11.4 3.64 0.88 0.416 12.1 - 13.6 0.80 11.6 2.99 0.82 0.243 13.6 - 15.1 0.84 11.1 2.52 0.85 0.232 15.1 - 16.6 0.81 9.5 1.88 0.78 0.108 16.6 - 18.1 0.87 8.9 1.54 0.87 18.1 -. 19.6 0.49 8.3 1.80 0.80 -9 19.6 - 21.1 8.6 1.00 0.90

. I h TABLE 3

. s BULK LOSS ON 210Pb 22Ra '37Cs 6 Co CORE DEPTH INTERVAL DENSITY IGNITION (cm) (g cmn) (%) (dpm gI) 21.1 - 22.6 8.6 0.74 22.6 - 24.1 9.5 24.1 - 25.6 9.3 JC-16 0 - 4 17.8 4 7 11.2 7 - 10 -12 10 - 13 12.2 13 _ 15 13.9 15 17 15.2

*- I -f

-V

TABLE 3 CORE Ag Al Cd Cu Fe Pb Zn A

(mg kgi)

Jo-I 0.27 26700 2.9 8 10800 22 33 *.X 0.54 20000 2.4 10 8600 20 30 0.30 22000 1.9 5 8700 21 41 0.18 24400 2.2 7 11700 21 31 0.22 23600 2.1 6 11300 23 30 0.06 22200 1.6 4 7400 22 30 JC-6 0.53 25300 1.50 38 19000 38 98 0.44 25600 1.16 29 18200 48 97 0.67 25600 1.01 26 17500 46 92 0.59 22500 0.76 24 14600 38 *71 0.70 28400 1.14 41 22700 55 127 0.37 25700 1.07 37 20900 52 111 0.51 29600 1.03 40 23000 46 124 0.35 26500 1.11 43 12800 60 141 0.62 36100 1.43 50 25800 72 177 ,.

0.48 28700 0.80 38 20900 50 129

TABLE 3 4

CORE Aq Al Cd Cu Fe Pb Zn (mg kg 1 )

0.53 31100 0.92 40 19900 52 137 0.47 27700 1.97 35 18800 35 128 0.38 29000 1.84 40 22200 45 143 0.35 26700 1.60 36 18900 41 119 JC-9 1.28 27800 3.05 50 25600 76 155 1.15 22300 2.23 48 27500 65 222 1.07 23300 1.50 42 20900 55 129 0.57 23300 1.79 65 25400 48 188 1.62 23200 1.32 41 24300 63 172 0.72 21400 1.85 66 21200 68 150 0.93 21800 3.31 45 23400 167 146 0.71 25200 2.36 56 26000 105 153 0.46 20400 1.63 41 20100 53 157 0.64 26200 1.81 50 24700 55 165 0.67 26600 1.87 48 24600 61 169 0.68 22100 2.41 50 23100 55 185 1.07 27300 3.13 65 27000 87 209 -_

0.98 26700 3.17 52 24300 121 185

TABLE 3 CORE Aq Al Cd Cu Fe Pb. Zn (mg kge) 0.68 23900 2.48 43 25400 68 165 0.57 24800 2.59 44 22900 66 162 0.60 28200 2.14 45 23400 73 149 a.

0.34 25600 2.04 37 23100 51 155 JC-13 1.12 23500 1.71 69 25400 71 161 0.74 24600 1.68 67 27200 65 161 1.61 21100 1.28 60 25200 78 143 0.89 25700 1.78 77 24600 90 177 0.77 23700 1.56 70 30100 76 172 0.88 21400 1.80 63 28600 88 164 0.96 25900 2.08 72 26100 94 167 0.98 25000 1.84 69 30100- 94 173 0.87 28700 1.76 77 28700 .67 183 0.89 28800 1.89 70 29600 65 183 0.87 27700 1.96 74 27000 88 193 0.68 23100 2.30 66 27900 67 203 0.71 27600 2.06 69 26800 52 178 0.58 29200 1.78 68 27400 66 163

TABLE 3

• A CORE AP Al Cd Cu Fe Pb Zn (mg kgI) 0.64 25300 1.38 56 25600 65 135 0.61 25200 1.97 37 25600 59 140 0.54 27000 1.62 51 23700 64 118 JC-15 0.57 24600 1.77 56 32000 79 198 0.84 24700 1.58 55 28800 85 189 0.77 28100 1.61 66 31200 75 189 0.46 28300 1.40 77 32800 68 210 1.02 27900 1.71 72 31000 87 189 0.75 23000 1.40 66 27900 85 183 0.64 29500 1.11 64 29500 130 178 0.61 29000 1.15 66 31500 91 184 0.62 28000 1.25 63 28800 90 181 0.43 26700 1.48 66 29100 70 176 0.51 29200 1.41 66 27200 64 198 0.74 26600 1.38 55 25400 62 173 0.34 30400 0.89 57 26500 62- 130 0.35 28600 0.87 37 22500 54 109 I

.9 0.52 23400 0.94 41 21600 30 88

TABLE 3 CORE Ag Al Cd Cu Fe Pb Zn (mg kge) 28200 36 22900 23 81 27400 1.10 35 23400 25 87 26000 34 23800 86 JC-16 0.15 23100 0.93 25 16700 75 124 0.07 19600 1.38 74 18500 82 174 0.13 23100 2.00 20 19000 49 231 0.14 25700 1.47 21 17700 39 222 0.07 16500 1.13 21 17100 52 223 0.08 15300 1.00 16 15300 39 150

TABLE 3 CORE DEPTH INTERV A.p - LO.I. 20Pb 22

'Ra WCs$ *Co Ag I Al I Cd I Cu I Fe I Pb Zn I (c n (glcrn M - (dp - -l (M kg) _

JC_ 0 1.9 1.1 0.27 26700 2.9 8 10800 2233 1.9 2.9 0.6 - - 0.54 20000 .2.4 10 860 20 30 2.9 3.9 0.6 0.30 22000 1.9 5 8700 21 41 3.9 4.9 0.4 0.18 24400 2.2-7 11700 21 31 4.9 5.9 0.4 0.22 23600 2.1 6 11300 23 30 5.9 6.9 0.6 0.06 22200 1.6 4 7400 22 30 JC-6 0 1.3 0.42 19.7 3.671 0.71 026 Q4 0.53 25300 1.50 38 19000 38 98 1.3 2.8 0.74 18.8 3.77 0.74 0.15 0.53 0.44 25600 1.16 29 18200 48 97 2.8 4.3 0.66 17.2 4.28 0.77 029 0.39 0.67 25600 1.01 26 17500 46 92 4.3 5.8 0.83 8.4 2.55 .0.77 0.14 0.45 0.59 22500 0.76 24 14600 38 71 5.8 7.3 0.80 9.3 3.61 0.76 0.17 0.44 0.70 28400 1.14 41 22700 55 127 7.3 8.8 0.73 12.3 3.09 0.84 0.31 0.40 0.37 25700 1.07 37 20900 62 111 8.8 10.3 0.85 7.9 3.16 0.84 0.24 029 0.51 29600 1.03 40 23000 46 124 10.3 11.8 0.82 8.1 2.83 0.83 0.33 0.43 0.35 26500 1.11 43 12800 60 141 11.8 13.3 0.91 6.5 3.14 0.84 0.42 0.45 0.62 36100 1.43 50 25800 '72 177 13.3 14.8 0.89 6.8 2.44 0.83 0.43 0.29 0.48 28700 0.80 38 20900 60 129 14.8 16.3 0.86 6.8 2.57 0.71 0.39 0.23 0.53 31100 0.92 40 19900 52 137 16.3 17.8 6.3 2.75 0.73 0.34 0.20 0.47 27700 1.97 35 18800 35 128 17.8 19.3 6.6 1.93 0.72 0.41 0.38 29000 1.84 40 22200 45 143 19.3 20.8 8.7 2.56 0.80 0.45 0.35 26700 1.60 36 18900 41 119 JC-9 0 1.7 0.34 15.3 4.77 0.62 0.22 0.51 1.28 27800 3.05 50 25600 76 155 1.7 2.7 0.23 16.6 4.38 0.75 0.17 0.47 1.15 22300 2.23 48 27500 65 222 2.7 3.7 0.36 26.6 3.16 0.66 0.34 0.86 1.07 23300 1.50 42 20900 55 129 3.7 5.2 0.57 26.5 4.94 0.83 0.27 0.70 0.57 23300 1.79 65 25400 48 188 5.2 6.7 0.62 15.8 4.39 0.69 0.24 0.54 1.62 23200 1.32-41 24300 -63 172 6.7 8.2 0.71 15.6 5.10 0.72 0.29 0.64 0.72 21400 1.85 661 21200 68 150 8.2 9.7 0.69 23.5 4.10 0.61 0.28 0.42 0.93 21800 3.31 45 234D0 167 146 9.7 11.2 0.67 13.5 3.47 0.65 0.43 0.42 0.71 25200 2.36 56 26000 105 153 11.2 12.7 0.67 33.8 3.27 0.72 0.41 0.42 0.46 20400 1.63 41 20100 53 157 12.7 14.2 0.64 20.4 3.24 0.73 0.43 0.16 0.64 26200 1.81 50 24700 65 165 14.2 ,15.7 0.69 13.8 2.33 0.74 0.44 0.17 0.67 26600 1.87 481 24600 61 169 15.7 17.2 0.65 19.9 3.98 0.69 0.72 0.68 2210O 2.41 50 23100 55 185 17.2 18.7 0.68 12.3 2.94 0.85 0.32 1.07 2730O 3.13 651 27000 87 209 18.7. 20.2 12.7 3.17 0.87 0.38 0.98 26700 3.17 52 24300 121 185 20.2 21.7 12.1 2.60 0.71 029 0.68 23900 2.48 43 25400 68 165 21.7 23.2 12.1 1.70 0.86 0.19 0.57 24800 2.59 44 22900 66 162 23.2 24.7 1Z3 1.86 0.78 0.16 0_60 28200 Z14 45 23400 73 149 24.7 26.2 17.4 0.34 25600 2.04 37 23100 61 155 JC_1 0 1.7 0.25 24.3 4.44 0.53 0.16 0.35 1.12 235i0 1.71 69 25400 71 161 1.7 2.7 0.24 16.3 4.29 0.60 0.21 0.19 0.74 24600 1.68 67 27200 65 161 2.7 4.2 0.58 21.2 4.03 0.65 0.37 0.38 1.61 21100 1.28 60 25200 78 143 4.2 5.7 0.55 14.2 4.37 0.59 0.46 0.34 0.89 25700 1.78 77 24600 90 177 5.7 7.2 0.63 13.9 5.03 0.59 0.44 0.50 0.77 23700 1.56 70 30100 76 172 7.2 8.7 0.68 13.7 4.48 0.72 0.51 0.51 0.88 21400 1.80 63 28600 88 164 8.7 10.2 0.76 17.6 3.71 0.68 0.41 0.40 0.96 25900 2.08 72 26100 94 167 10.2 11.7 0.70 13.2 4.64 0.67 0.47 0.26 0.98 2500D 1.84 69 3D100 94 173

_11.7 13.2 0.68 13.1 3.54 0.69 0.56 0.26 0.87 28700 1.76 771 28700 67 183 13.2 14.7 0.70 12.7 2.59 0.69 0.51 0.23 0.89 28800 1.89 701 29800 65 183 14.7 16.2 0.721 12.9 3.59 0.75 0.65. 0.87 27700 1.96 741 27000 88 I193 4

16.2 17.7 0.69 12.8 2.74 0.72 0.51 0.68 23100 2.30 62790067 203 17.7 19.2 1 2. .9 0.69 0.33 0.71 2700 2.06 691 26800 2 17 OApP6%X IC clJ .