Earth and Planetary Science Letters, 82 (1987) 289-304 289 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands
[61
Lead in corals: reconstruction of historical industrial fluxes to the surface ocean
G l e n T. Shen * and E d w a r d A. Boyle
Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 01239 (U.S.A.)
Received October 1, 1986; revised version received January 9, 1987
Twentieth century environmental lead chronologies for the western North Atlantic, Pacific, and Indian Oceans have been reconstructed from annually-banded scleractinian corals. Measurements of lattice-bound Pb in sequential coral bands reveal temporal changes in surface water Pb concentrations and Pb isotopic distributions. Perturbations are observable in all specimens studied, attesting to global augmentation of environmental Pb by industrialization.
In the western North Atlantic, Pb perturbations have occurred in direct response to the American industrial revolution and the subsequent introduction and phasing-out of alkyl Pb additives in gasoline. Surface ocean conditions near Bermuda may be reliably reconstructed from the coral data via a lead distribution coefficient of 2.3 for the species, Diploria strigosa. Based on 21°pb measurements, a similar distribution coefficient may be characteristic of corals in general. Surface Pb concentrations in the pre-industrial Sargasso Sea were about 15-20 pM. Concentrations rose to near 90 pM by 1923 as a result of metals manufacture and fossil fuel combustion. Beginning in the late 1940's, increased utilization of leaded gasoline eventually led to a peak concentration of 240 pM in 1971, representing an approximate 15-fold increase over background. Surface ocean concentrations are presently declining rapidly (128 pM in 1984) as a result of curtailed alkyl Pb usage. Lead isotopic shifts parallel the concentration record indicating that characteristic industrial and alkyl Pb source signatures have not changed appreciably in time. Industrial releases recorded in the Florida Keys reflect a weaker source and evidence of recirculated Pb (5-6 years old) from the North Atlantic subtropical gyre. An inferred background concentration of 38 pM suggests influence of shelf a n d / o r resuspended inputs of Pb to these coastal waters.
In remote areas of the South Pacific and Indian Oceans, industrial signals are fainter and the corals studied much younger than their Atlantic counterparts. Contemporary Pb concentrations implied by coral measurements (assuming K D = 2.3) are 40-50 pM for surface waters near Tutuila and Galapagos in the South Pacific, and 25-29 pM near Mauritius in the Indian Ocean. A single coral band from Fiji (1920 + 5 yr) implies a pre-industrial surface water concentration of 16-19 pM Pb for the South Pacific. In view of reported surface water measurements and the North Atlantic coral data, the Pacific coral extrapolations may be slightly high. This could be a result of small variations in K o among different coral genera, or incorporation of diagenetic Pb by corals sampled in coastal environments.
1. Introduction
Over the past two decades, convincing evidence of global environmental contamination by in- dustrial lead has accumulated. Murozumi et al. [1] documented historic increases of lead in snow strata cored in Greenland and Antarctica in 1969. More recent snow and ice core determinations from both hemispheres have confirmed their origi- nal findings (see review by Wolff and Peel [2]). Contamination continues to hamper measurement efforts in the most pristine locations, however,
particularly in the case of ancient samples. Extension of anthropogenic lead mapping to
the oceans succeeded more recently when Schaule and Patterson [3] overcame sampling difficulties in 1976. Together with newer oceanic data by Flegal, Schaule, and Patterson [4-6] and Boyle et al. [7], and atmospheric flux measurements by Settle, Patterson and coworkers [8,9] and Jickells, Church and others [10,11], effects of industrial proximity and meteorology have been observed. Eolian de- livery of stable lead parallels that of 21°Pb [7,12,13], but the stable Pb source function differs from that of 21°pb and has evolved over time. The transient nature of this flux leads to the application of stable Pb as an oceanic chemical tracer to comple-
ment findings based on chlorofluorocarbons and bomb-produced radionuclides. As discussed by Boyle et al. [7], the success of such an application hinges on construction or recovery of accurate regional source deposition records. At present, the longest synoptic Pb measurement time series (now ending its third year) is that of Boyle and co- workers in the Sargasso Sea. Continued tracking of anthropogenic Pb in the ocean is needed, since the greatest changes in the upper ocean are ex- pected over the next few years.
This paper describes the reconstruction of past surface ocean Pb concentrations and Pb isotopic distributions using the skeletal Pb content of an- nually-banded corals. Measurements on corals from the western North Atlantic, Florida Keys, Pacific, and Indian Oceans reconfirm the perva- sive nature of industrial Pb aerosols and provide detailed chronologies of 20th century worldwide industrialization.
2. Sampling and analysis
The corals selected for this study were com- prised of various scleractinian genera, all cate- gorized as zooxanthellate, constructional, and hermatypic, according to the proposed terminol- ogy of Schuhmacher and Zibrowius [14]. Pertinent species, habitat, and sampling information are given in Table 1 and Fig. 1. Dating was accom- plished by counting of annual bands on X-radio-
120 180 120 60 0 60
6C DO
120 IgO 120 60 0 60
Fig. 1. Coral sample locations: (A) = Bermuda; (B) = Florida Keys; (C)=Galapagos; (D)=Tutuila; ( E ) - F i j i ; ( F ) = Mauritius; (G)= Heron Island; ( H ) = Lisianski lsland; ( I ) = Eniwetak.
graphs according to techniques outlined by Hud- son [18] and Buddemeier et al. [17]. The Florida Keys and Bermuda chronologies are corroborated by 14C measurements by Druffel and Einick ([16,19], Bermuda data unpublished). In other cases, cross-checking was achieved by 21°pb dat- ing.
A detailed description of trace metal analysis in corals will be given elsewhere [20]. Briefly, annual bands are mapped on transparent acetate film using X-radiographs of coral slabs (0.5-1.0 cm thickness), and these are then used as sectioning templates. Samples may be cut by any of several means, depending on band contours and thick- ness: i.e. low speed rock saw (Buehler - Isomet), band saw, or jeweler's saw. Band intervals con- sisted of a high- and low-density band pair which comprised one year of growth in all cases. Sam- pling boundaries for corals from Bermuda and the Florida Keys spanned the bottom edge of a high- density band to the next lower high-density bot- tom edge. All other corals were sectioned along the top edges of the high-density bands. Dense band formation in corals has been shown to occur during the months of warmest water temperature [21,22]. Cut fragments were cleaned ultrasonically in acid and hydroxide/peroxide media to remove surface contamination associated with handling and organics. Coarse crushing in an agate mortar was followed by another preliminary cleaning se- quence, before final crushing to a 280-700 #xm particle size distribution. Where morphological considerations warranted (e.g. corals of genus Di- ploria), specific structural components were iso- lated to facilitate cleaning. These consisted of high-density, low-surface area, compound trabecu- lae which resemble spines radiating outward along the coral growth axis. The intervening network of secondary skeletal parts including loose trabeculae and synapticulae proved resistant to cleaning and usually demonstrated measurement offsets indicat- ing contamination. For concentration determina- tions, triplicate samples weighing from 60 to 120 mg were cleaned intensely in both oxidizing and reducing media in the presence of ultrasonication and occasional heating. Isotope determinations often required additional sample to yield suffi- cient Pb for mass spectrometry. Cleaning and other subsequent preparation was carried out in 1.5 ml acid-leached polyethylene centrifuge vials.
e .
..~,
,...,
8
~ m . ~ ' E 3 ,-= ~ o~ ~ ~
o o o
~ g g
z z z
o ~ o
~..~
o o o o o o
~ { z ~
o o
~ E
.- =~
o
o
291
Typical losses due to siphoning and dissolution were 30%. The cleaned samples were dissolved in 2.0N vycor-distilled HNO 3.
Lead concentration analysis was performed by graphite furnace atomic absorption spectropho- tometry (GFAAS) using a Perkin-Elmer Model 5000 spectrophotometer, Model 400 furnace, and AS-40 autosampler. A L'vov platform was used to enhance sensitivity. Due to the trace Pb levels encountered in coral aragonite (8-150 ppb), sep- aration of Pb from the calcium matrix was neces- sary for good reproducibility. This was accom- plished by cobalt-APDC chelate co-precipitation [23,24]. Precision of the reported Pb concentration data varied both as a function of the coral species and coral band age. Generally, replicate analyses of older bands were more consistent than those of younger bands. Cleaning of very recent bands (< 10 years old) occasionally proved inadequate. Measurement precision variations were probably related to a variety of factors including skeletal morphology (susceptibility to detrital contamina- tion and conversely, cleaning), turbidity of growth environment, prevalence of algal or other organic inclusions, and sample recovery/s torage/han- dling. Since a general statement of precision is not useful, the reader is referred to error bars ( lo) supplied on all data plots. Blanks were typically less than 5% of measured signals in recent Atlantic corals and 10-30% in the most pristine coral bands. Blank contributions were almost always attributable to reagents, with procedural con- tamination only a minor infrequent component.
Lead isotopic measurements of samples puri- fied by anion exchange chromatography [25] were run on 12" magnetic sector, solid source mass spectrometer (M.I.T., Department of Earth, Atmospheric and Planetary Sciences; S.R. Hart). Samples (typically 5-30 ng Pb) were loaded onto single rhenium filaments with H3PO 4 and silica gel. Blank levels generally constituted less than 1% of the total lead deposited, consequently, correc- tions were not attempted. Measurement error (95% CFL) is conservatively estimated at _+0.0010 for the reported 2°6/2°7pb ratios to allow for variable mass fractionation.
21°pb determinations on corals were performed by alpha counting of 21°po [26] on EG & G Ortec Model 576 alpha spectrometers fitted with 450 mm 2 low-background surface detectors. 21°po
292
(t]/2 = 138 days) was assumed to be in secular equilibrium with 2 ] ° p b in all samples since the youngest coral bands analyzed were 2 years old. 2°8Po was used as an autodeposition yield moni- tor. Errors due to counting statistics were gener- ally less than 5%.
3. The coral lead record
3.1. Bermuda Positioned in the southern portion of the North
American Westerly flow path [27], Bermuda is a prime site for recording historic industrial lead fallout to the Sargasso Sea. Fig. 2a contains two such depositional records in corals. The lower record reflects the Pb transient in the open ocean (seaward side of North Rock, 14 km north of Bermuda), and the upper incorporates nearshore addition of diagenetic Pb (Southern Reef Preserve, 0.5 km south of Bermuda). The conspicuous gap (1955-1957) in the North Rock data is due to an attenuation of dense structural parts (compound trabeculae) which are isolated for analysis in this particular species of brain coral. Apparently, this coral suffered unidentified environmental stress during this three-year period while the more southerly coral did not. Analysis of the lower-den- sity skeletal components in each of these affected bands resulted in variable elevated Pb concentra- tions (45-65 nmol P b / m o l Ca). The major fea- tures of the overall record are: (1) a gradual increase in skeletal Pb levels near the end of the 19th century; (2) a second more pronounced in- crease beginning in the 1950's, and (3) a dramatic decline initiated in the early 1970's. This pro- gression of events very closely parallels the devel- opment of American industries tied to possible lead emissions, as can be seen in Fig. 3. The Pb source responsible for the turn of the century rise cannot be exactly specified, since early growth patterns for most large-scale industries are similar. Nevertheless, the early perturbation must be a direct consequence of the American industrial rev- olution. The striking resemblance between the subsequent portion of the coral record and U.S. alkyl lead consumption suggests that this source gained prominence with the rise of the automo- bile. Because competing sources appear to have levelled-off after 1940, the alkyl lead pattern is accurately superimposed on the earlier coral re-
80
~6i0 E
3 E £4O
m
(8)
BERMUDA
(0 str igosa) ÷ +
+ ~- ÷ ,
C o J t h F q Ur~se 've 44-+ ~ +~ ' r [q~k~.+
+++4 Nort~ Rock
i L i , i i i i i , 18'70 ' 189 '3 1910 1930 1950 1970 199,3,
Y o a r
123~
1.2~3
~laoo
1.10C
BERMUDA
(D stP2gosa)
~ " '~ \ ~ w e s t e r n N Atlantic ' Pelagic Sediments '\
\
i i i , i i i i , i 1850 1870 IB90 t918 1930 19b0 :970 t990
(b) Year
Fig. 2. (a) Skeletal Pb concentrations in two colonies of D. strigosa cored near Bermuda: + = Southern Reef Preserve; • = North Rock. In nearly all cases, data points represent triplicate analyses or better. Error bars are 1 standard devia- tion from means. Error bars for the Southern Reef chronology have been omitted for clarity (magnitudes comparable to North Rock). Four contaminated bands (1955-1957, 1983) in the North Rock record ranging from 45 to 65 nmol Pb /mol Ca (27 analyses) are not shown (see text). (b) Lead isotopic history of Sargasso Sea surface waters as recorded by D. strigosa North Rock, Bermuda. Background field is based on Chow and Patterson [51]. 1984 and 1985 determinations are seasonally- averaged surface seawater analyses (5 in 1984; 4 in 1985).
cord. This is true to the extent that an extrapo- lated 1986 coral Pb concentration approaches the pre-alkyl Pb 1920-30's value. An approximate 1- year time lag between the peak coral Pb content in 1971 and the alkyl Pb consumption maximum in 1970 is consistent with estimates of 21°pb mixed layer residence times of 1.7-2.5 years [29,30].
The surface ocean Pb decline of 1.5-fold re- corded by the Bermuda corals between 1979 and 1984 is in agreement with the earliest time series data available for dissolved Pb and 21°pb. The ratio of Schaule and Patterson's July 1979 surface water Pb measurement to our own in June 1984 is
293
(o)1000
d S PRIMARY LEAD P~ODbC~I3N BOO
P BOO
s ~c,c
n 2°ii
:530 1958
(b) 150
1E5
100
72
58
25
~8'50 1873
L: 5 IRO[', ORE :O(DDCTi[:,N
f 1890 1910 1938 1920 1970
(cX2CO
197C 1990
IOCC
800
500
aJO
23J
I , ~ . , ~ * "~'+~
Y850 t870
L S CDAL PROO'JCTI3N
/ J 189'3
Y
1913 1958 1950 ]97: I II~3Q
(0)300
250
g zoo
12o
% 1oo
50
%So
U S ALKYL LEAD CONSUMPTION
,m
, / f ,
1930 t950 1970 1870 IBgO t910 t990 Year
Fig. 3. Possible U.S. industrial sources of Pb to the Sargasso Sea: 1980 to present. (Sources: U.S. Department of Commerce [28] and testimony by Ethyl Corporation to EPA Public Hear- ing on Proposed Regulations on the Lead Content of Gasoline, August 31, 1984.)
164 pM/103 pM or 1.6, however, extreme season- ality in surface ocean Pb places a high degree of uncertainty on this value [7]. Since Pb and 21°Pb fluxes have exhibited a close coupling over Bermuda, one could also compare a 2a°Pb-normal- ized ratio over time: (Pb/2a°Pb)jub 1 9 7 9 / ( P b /
21°Pb)1984avg.(n=4) = 1.9. A closer comparison might result if the coral Pb decrease from 1980 to 1985 were known. The absolute flux of Pb to the surface
and deep ocean has been estimated by three inde- pendent rain and sediment trap measurements by Jickells et al. [10,31] and Church et al. [11] for the period 1981-1983. These results all fall within the range 0.88-1.18 m g / m 2 yr. Settle and Patterson's [9] mean flux of 1.7 m g / m 2 yr is higher, but corresponds to surface water, rain, and dry de- position measurements from 1979 when the in- dustrial flux was higher. Integrating an annual flux of 1.0 m g / m 2 yr over the western North Atlantic between 30 ° N and 50 ° N (latitudes over which U.S. westerlies prevail) gives a rough total Pb flux of 10 × 10 l° g / y r for 1982. This represents about 20% of the total Pb consumed as alkyl Pb in the U.S. that year. However, it can be estimated (see Fig. 4) that alkyl Pb was responsible for only about 38% of the total industrial Pb flux in 1982 (based on the estimated contributions recorded by D. strigosa in 1983 which allows for a 1-year surface ocean lag time). Thus, a gasoline-derived Pb flux of approximately 3.8 × 101° g (7% of the raw additive usage) was delivered to the western North Atlantic in 1982.
The coral-based Pb source breakdown of Fig. 4 is surprisingly consistent with Nriagu's global Pb emission summary for 1975 [32]. His estimated contributions for anti-knock additives (58%), all other anthropogenic sources (37%), and back- ground (5%) compare favorably to our own 1978 estimates (again allowing for a brief time lag) of 56%, 37%, and 7% for these same respective sources. As far as the relative roles of key in- dustries and their Pb emissions, Nriagu's contem- porary analysis suggests that i ron/steel produc- tion is about 1.5 times as important as lead pro- duction, twice as important as copper production, and 3 times as important as coal combustion or zinc production.
At the opposite end of the time scale, the question of pre-industrial environmental Pb levels arises. Before addressing this question, it would be useful to translate skeletal Pb concentrations to a more meaningful measure, namely, ambient Pb concentrations in seawater. This can be accom- plished by means of a distribution coefficient ( K D) for Pb in corals relative to surface water (see Appendix 1). Based on 1983-84 seasonally-aver- aged surface water dissolved Pb concentrations, the North Rock and Southern Reef Preserve coral records (both Diploria strigosa) extrapolate to K D
Fig. 4. Reconstructed contributions of natural and anthropogenic Pb to the Sargasso Sea based on the North Rock coral chronology of Fig. 2a. A constant background coral Pb/Ca ratio of 3.5 nmol/mol is assumed. Industrial (non-alkyl Pb) inputs increase from their inception during the 19th century and are assumed constant after 1923. Alkyl Pb is presumed responsible for subsequent perturbations. Note that real-time breakdowns would anticipate the coral-based reconstruction by about 1 year due to surface ocean integration of lead inputs.
= 2 .1-2 .3 (Table 2). Thus, D. strigosa dis- c r imina tes in favor of Pb over Ca dur ing skeleto- genesis.
A p p l y i n g K D = 2.3 to the coral record reveals that as of 1890, surface waters of the western N o r t h At lan t ic con ta ined 24 p M Pb. However , p r oduc t i on of i ron ore, coal, lead, and o ther p r i m a ry meta ls had been s teadi ly increas ing the previous 2 0 - 3 0 years. If one careful ly examines the relat ive growth ra te of each poss ib le source indus t ry be tween 1890 and 1920, p r ima ry Pb pro- duc t ion matches the coral record most closely (bo th exper ienced 3.5-fold increases versus 6-fold increases in the steel, copper and coal indust r ies [26]). This compar i son , however, leads to the ex- pec ta t ion that surface water Pb fell to less than 1 p M by 1870. There are several reasons to believe this unreal is t ic : (1) the Pb decl ine impl ied by D. strigosa f rom 24 p M 1890 to the equivalent of 22 p M Pb in 1984 is too g radua l to ex t rapo la te to 1 p M by 1870; (2) Pb isotopes (sect ion 3.5) show ident ica l 2°6/2°7pb ra t ios in 1887 and 1895 which a l r eady fall wi thin the b a c k g r o u n d envelope; (3) a 1920 Fi j i coral de t e rmina t ion suggests a pre- in- dus t r ia l surface ocean Pb concen t ra t ion of 16 -19 p M in the South Pacific. A l though the la t ter de-
t e rmina t ion m a y con ta in an anomalous nearshore Pb componen t , it is p r o b a b l y not more than 50% of the total (see fol lowing sect ion on is land in- fluences). Thus, it is unl ikely that the N o r t h At lan t ic , wi th its higher fluvial and aerosol inputs, could have suppor t ed a concen t ra t ion much lower than 10 pM. If, on the o ther hand, cont ro l of the
TABLE 2
Lead distribution coefficients for two Bermuda corals
Southern Reef North Rock Preserve (D. strigosa) (D. strigosa)
Bermuda Coral [Pb] / [Ca] (nmol/mol) 1983-84 est. 42 x t 0 - 9 30 × 10- 9
early portion of the coral record was by one or a combination of the iron ore, copper, or coal in- dustries, the projected pre-Industrial Revolution surface water value would be approximately 15 pM. Flegal and Patterson [5] suggest a comparable value in estimating that 1979 North Atlantic surface waters (160 pM) were enriched by 10-fold over pre-historic concentrations. Their estimate, however, was conceived to span over two centuries of anthropogenic activity as justified by the Greenland snow strata record of Murozumi et al. [1]. This poses the question of the importance of anthropogenic Pb prior to 1850. Simple scaling of our late 18th century surface water estimates to the snow record results in a prehistoric extrapola- tion of < 1 pM, which again appears unrealistic for the reasons cited earlier. Such an extrapolation is unwarranted, however, if one considers that (a) prehistoric aerosol fluxes to Camp Century, Greenland were very small relative to those reach- ing the Sargasso Sea; and (b) the Sargasso Sea may have received substantial fluvial inputs of Pb, whereas Greenland ice did not. Thus, an estimated Sargasso Sea surface water value of 15-20 pM Pb for the year 1850 is probably also a good prehis- toric approximation.
295
If one accepts that the coral Pb record is equiv- alent to a record of dissolved Pb concentrations (differing only by the factor Ko) , it is straightfor- ward to interpret the coral data alternatively, as a flux history. The assumption must be made that the dissolved Pb concentration at a given location is at any time proportional to the incoming Pb flux. Conversion can then be accomplished simply by scaling the coral data to a suitable flux mea- surement (i.e. 1982 total Pb flux to Sargasso Sea - 1.0 m g / m 2 yr; resultant 1983 coral skeletal Pb level = 30 n m o l / m o l Ca).
Island influences at Bermuda. Lead of local origin may influence coral-based reconstructions via habitat pollution or dissolved-particulate interac- tions in waters overlying reefs. Coastal Pb pollu- tion resulting from urban effluents and mine tail- ings run-off has been documented by Patterson et al. [33], Stukas and Wong [34], and Bruland et al. [35] among others. A survey of Bermudian water- ways (Fig. 5) reveals that neither large populations nor heavy industries are required to grossly con- taminate inshore surface waters. The progression in dissolved Pb concentrations from well-flushed coastal zones to Hamilton Harbor yields values
BERHUDA
Nearshore Lead Concen t ra t i ons
32°20'N
~970 (2550)
North Rock
(I OkmJ
///~reorgs 400 ~ y - . . . ~ 276
Qg/ , ,~ 754 u L k~
,',,f t 4 5 ~ c" J?~7#; ? / ® co,-a~
(9 235 Saap i Jng Site
looo (1 ooo)
(2250)
64' 45'W
] TotaJ Pb (pM) I OJssoJved Pb (pM) J
Fig. 5. Total and dissolved ( < 0.4 ~m) lead in waterways and coastal areas around Bermuda (April 1984). Four samplings at point A spanned 6 / 8 3 - 6 / 8 4 (see Table 2).
296
1.5 to 20 times open ocean concentrations. A similar situation has been documented for St. Croix, U.S. Virgin Islands from coral analyses of Dodge and Gilbert [36]. Particulate Pb in nearshore waters can also be observed in Fig. 5 to account for up to 85% (inner Hamilton Harbor) of the total measured Pb. This contrasts with levels of 10% or less encountered in the open ocean
[5,291. In the coastal zone of Bermuda, an interesting
phenomenon is revealed by seasonal seawater measurements and comparison of the two Bermuda coral Pb records. Point A in Fig. 5 represents four separate water samplings accumulated over three different seasons from June 1983 to June 1984. The average dissolved Pb concentration here shows little seasonal variation and amounts to 198 _+ 15 pM as compared to a seasonally-averaged oceanic value of 128 pM for 1984. The Southern Reef Preserve coral of Fig. 2a was collected only 2 km northeast of point A in similar waters. The con- stant skeletal Pb offset (approximately 10 nmol Pb /mol Ca) exhibited relative to the North Rock specimen suggests that an elevated dissolved Pb concentration is sustained by the sediment inven- tory or particulate Pb burden in nearshore waters. Additionally, absence of seasonality in the dis- solved Pb measurements at point A suggests that
these nearshore sources buffer the dissolved Pb concentration. The extent of this coastal effect was assessed in two transects heading away from Bermuda (Fig. 6). The first transect was directed toward North Rock, 9 miles north of the island. The second transect headed toward Station "S" (formerly known as the Panulirus Station), 20 miles away, where the bulk of our hydrographic work [7] is done. The results show convergence of total Pb concentrations to prevailing ocean values within 5 miles from shore. Dissolved Pb, the more relevant measure in terms of coral Pb uptake, appears to stabilize within 1.5 miles from the island. Given that Bermuda is well-populated by remote island standards, a 1.5 mile coral sampling boundary seems a reasonable rule for future col- lections targeted for trace elements. Seawater sam- ples collected within 5 miles must be filtered be- fore analysis in order to accurately assess ambient dissolved Pb concentrations.
3.2. Florida Keys The general pattern of Pb increase in the Florida
Keys over time (Fig. 7) is depressed relative to the Bermuda records. In addition, subtle distinctions suggest a modified source term from that affecting Bermuda. For example, the turn of the century rise is absent as levels appear to have remained
500
400
30O ]E
i-i 200
100
BERMUDA - North
Transect
k \ z~ Unfiltered \ / • Fi l tered
'°° I 400[ ~ '",, ,,,,,J 300]- BERMUDA - S o u t h e a s t
A\A Transect
/ / ~ ~ Unfiltered
2001 / II . Filtered
t i t I t i I 01 I t t t t t t O0 I 2 3 4 5 6 7 i 2 3 4 5 6 7 B
Miles from Bermuda Miles from Bermuda
Fig. 6. Local transect seawater Pb data heading from Bermuda toward: (a) North Rock (April 13, 1984); (b) Station "S" (Panulirus Station) (0-3 miles; June 12, 1984; 3-7 miles: December 11, 1984). Dashed lines depict dissolved Pb concentrations measured at Station "S" (20 miles southeast of Bermuda) during the same time periods.
FLORIOA KEYS
(~ annularis) 5 O
~ 4 0
Eq ~ 8
~20
60
++
~ / I 1890 1910 ' 19'30 ' 19'50 ' 19'70 ' 1990
Year'
Fig. 7. Skeletal Pb concen t ra t ions in M. annularis, Flor ida Straits. Annua l de te rmina t ions are in most cases, based on
t r ip l ica te analyses or better. Error bars are l o f rom means.
fixed at 10 nmol Pb /mo l Ca back to the year 1698. Also, the post-World War II increase is moderated and the 1970's peak occurs several years later. Since Pb sources to the Florida Straits are borne by easterly winds and surface waters whirh have transited the Caribbean Sea, dissim- ilarities between records are expected. Episodic influence of Florida Bay discharges at this site has also been implicated by stress band studies of Hudson et al. [37]. Heavy industrial emissions originating in the northeastern U.S. should play a lesser role in the Florida Keys, with greater repre- sentation of Pb aerosols from the southern U.S. and perhaps nations encircling the Caribbean. Riverine inputs of Pb to the Florida Straits and Florida Bay, however, appear unimportant as dis- solved Pb undergoes large-scale uptake onto sus- pended particles in the Mississippi River and estuary (K n >/105; Lee and Boyle, personal com- munication).
The existence of a peak in the coral record at 1977 has important implications. This maximum identifies the presence of U.S. emissions in the area since use of octane-boosting alkyl leads in Latin America and Caribbean nations has not been curtailed to the extent legislated in the U.S. More importantly, though, the timing of this max- imum is 6 years delayed relative to the signal at Bermuda. If horizontal transport of a North Atlantic source is responsible for this delay, this implies an average surface ocean recirculation velocity (assuming a 6000 km travel path from Bermuda, through the Antilles, and around the Caribbean) of 3.2 cm/s . This is comparable to
297
surface Ekman transport estimates and wind drift speeds observed in the center of the subtropical gyre [38]. There are some qualifications to this interpretation, though. Since Pb is stripped from oligotrophic surface waters within 2-3 years, sim- ple recirculation of a mixed layer source originat- ing from northerly latitudes is unrealistic. Ad- ditionally, while subsurface waters (> 100 m) complete an anti-cyclonic trajectory through the Caribbean, the northern branch of the North Equatorial Current hinders southward passage of North Atlantic surface waters. Thus, if the Florida Keys signal is derived from long-range horizontal transport, it must consist of Pb which has been mixed from deeper in the thermocline. The sill depths of the Anegada and Jungfern Passages separating the Caribbean from the North Atlantic are sufficiently deep (1500-2000 m) to permit passage of thermocline waters containing longer- lived Pb. Re-enrichment of surface waters may occur by mixing in the Yucatan Channel and Florida Straits as sill depths shoal from 1600 to 800 m.
Additional insight can be drawn from data published by Dodge and Gilbert [36] for a coral (also M. annularis) collected 1850 km southeast of Florida at St, Croix, U.S. Virgin Islands. Their 26-year record (1954-1980) at Buck Island shows a gradual rise from 10.3 nmol Pb /mol Ca in 1954 to a maximum of 24.5 nmol Pb /mol Ca in 1976. Tlais maximum, which occurs between those re- corded at Bermuda and the Florida Keys, appears to accord with a horizontally transported U.S. Pb source. The oldest concentration approaches the pre-industrial value measured in the Florida Keys. 1954, however, is relatively recent as far as in- dustrial emissions are concerned, so an even lower pre-industrial value may be likely for this location.
3.3. South Pacific and Indian Ocean corals In comparing the above data with coral mea-
surements from the more remote locations of Galapagos, Tutuila, Fiji, and Mauritius (Fig. 8), the tremendous influence of the United States as an industrial source can be appreciated. Skeletal Pb levels here do note exceed 15 nmol Pb /mol Ca. As a consequence of these lower levels, it is more difficult to ascertain historic patterns. This problem is compounded by the scarcity of old sample cores in these outlying regions. The single
298
30 PACIFIC AN[} INDIAN OCEANS
' ....... i +
~ 5 H 14icroconos) (6alapagos P
o (Mauritius 8 Flji ~ ~ a ) ~ 2 o - - ; . . . .
E /I I\ / ,CC 15 Tutuila
MBuPitius
019~"55'5~-40 ,~ ' 19'65 1g'70 19'75 1g'80 1985
Year
Fig. 8. Skeletal Pb concentrations in corals collected from the Galapagos Islands, Tutuila, Mauritius, and Fiji. Annual de- terminations are triplicate analyses or better. Error bars are la from means.
de t e rmina t ion f rom Fij i (4.1 + 0.4 nmol P b / m o l Ca) poin ts out a poss ib le t r ip l ing or more of Pb be tween 1920 and 1960 in t ropical reef waters of the centra l South Pacific. Despi te the overal l low levels of Pb encountered , however, concerns re- ga rd ing is land inf luences on these r emote ly s i tuated corals are addressed in the next section. The Ind ian Ocean spec imen f rom Maur i t ius was too young to furnish a useful t ime series, but c o n t e m p o r a r y concen t ra t ions are roughly half those found in the South Pacific. Occas ional un-
usually high values, particularly among recent coral bands, are probably a result of a refractory contaminant Pb phase as discussed earlier.
3.4. A survey o f Zl°Pb in corals
Though a shor tage of si te-specific dissolved Pb measurements prec ludes add i t iona l es t imates of s table Pb d i s t r ibu t ion coefficients for the corals descr ibed, 21°pb da t a are more plentiful . Ci t ing open ocean values from the l i terature, a suite of 21°Pb-based es t imates of K D is r epor ted in Table 3. It is immedia t e ly evident that these es t imates are consis tent ly higher than the value based on s table Pb in Bermuda ( K D = 2 . 3 ) . This is true even for several specimens f rom N o r t h Rock, Be rmuda ( including the one analyzed in Fig. 2) for which an accurate dissolved 21°pb concen t ra t ion can be deduced f rom seasonal measurements . Mos t of these inconsis tencies are expla inable .
In the case of the Bermuda corals where convergence with the s table Pb result is most expected, the d i sc repancy is l ikely due to two factors. Firs t , 21°pb was de te rmined from whole coral crushings while s table Pb analyses were pe r fo rmed only on the c o m p o u n d t rabeculae of the Diplor ia corals. A t t empt s to de te rmine s table Pb on f ragments of D. labyr in th i formis which in-
TABLE 3
Estimated lead distribution coefficients for various corals based on 21°Pb
Location Genus/species Initial unsupported 21°pb (dpm/100 g coral)
Observed surface 2mpb (dpm/100 1)
Reference Impl i ed K D
North Atlantic Bermuda D. strigosa 56 Bermuda D. labyrinthiformis 63 Bermuda M. annularis 65 Florida Keys M. annularis 15
North Pacific Eniwetak F. speciosa 75 Lisianski P. lobata 69
Equatorial Pacific Galapagos P. clavus 33
South Pacific Tutuila H. microconos 30 Heron Island P. a~traliens~ 18
Indian Mauritius P. rustica 32
19.7_+3.4 (n = 5)
9
13 2O
9
10 11 7.6
9-11
[7]
[39] [30]
[40]
[41] [41]
[42]
2.8 3.2 3.3 ?
5.8 3.4
3.7
2.8 2.4
3.2
cluded septal trabeculae and synapticulae (low- density, high surface area structures), resulted in +10-20 nmol P b /mo l Ca offsets from the D. strigosa data and were marked by poor reproduci- bility [20]. This extraneous Pb amounts to ap- proximately 30% of the true lattice-bound Pb pool. If 21°Pb is distributed analogously to stable Pb as one might expect, this would effectively explain the difference in calculated distribution coeffi- cients. The second factor is a less stringent clean- ing protocol adopted for 21°pb assay compared with the stable Pb procedure (the initial assump- tion having been that in the absence of handling contamination for 21°pb, natural contaminants would constitute a small easily removed compo- nent). Coupled with a 21°pb sample bias toward very young bands (which are often difficult to clean), these natural contaminants are probably also responsible for part of the K D inconsistency. M. annularis would have been especially suscepti- ble to a 2.5 and 6.5 years old. Additional evidence for a Pb distribution coefficient near 2 is sug- gested by Dodge and Thompson's [43] determina- tion of 21°pb in a specimen of D. labyrinthiformis from Castle Harbor, Bermuda. Refinement of their estimate depends, however, on actual measure- ment of dissolved 21°pb within this semi-enclosed basin.
As far as any of the Pacific and Indian Ocean K D estimates are concerned, it is first necessary to point out that the cited dissolved 21°pb measure- ments represent single season determinations of waters in the general vicinity of the sampled coral reefs. Sargasso Sea surface 21°pb concentrations have been observed to change by as much as 50% in three months, depending on rainfall and mixed layer integrity [7]. A small bias may actually be transmitted to the coral record through the inter- play of seasonal surface water variation and skeletal mass accumulation. However, since coral growth is known to be continuous [44] and sea- sonal band width and density generally vary in- versely [43,45], the effect of seasonally-dependent accretion is probably limited. Apart from the pos- sibility of applying a non-representative surface water 2a°pb concentration in calculating KD, it should also be noted that neither of the two North Pacific corals in Table 3 produced useable stable Pb data. If this was a consequence of unfavorable morphology rather than poor sample storage (a
299
strong possibility in the cases of Porites and Favia speciosa), the Zl°Pb results may also be suspect and should be considered as upper limits. In the Galapagos Islands case, the coral measurements are probably not so much at fault as the choice of a surface water 21°pb value. Eastern Equatorial Pacific waters are distinguished by variable up- welling and associated high- and low-productivity and scavenging regimes. The only surface water 21°pb measurements available in the area are from the Peru Basin at 11°S (9.1 dpm/100 kg) and 19°S (3.0 dpm/100 kg), non-upwelling and up- welling locations, respectively [40]. Although coral cadmium measurements indicate that the Gala- pagos site at San Cristobal Island is nutrient en- riched by upwelling [46], a relatively high dis- solved Za°Pb concentration (> 11 dpm/100 kg) is required to yield K D < 3.0. This condition would depart from the conclusion reached by Thomson and Turekian [40] that high-productivity zones exhibit depleted 21°Pb relative to parent 226Ra, particularly in surface waters. With regard to the remaining cases, discrepancies between 21°pb and stable Pb KD'S appear small or attributable to mechanisms previously discussed.
In view of the above discussion, the actual range of Pb distribution coefficients exhibited by eight species (seven genera) is probably very small, perhaps 2.3-3.0. This is a desirable outcome in that coral Pb measurements can be interpreted directly, without the need to normalize according to genus as in the case of 180 coral paleothermom- eters [47].
If one accepts, tentatively, that K D dependence on genus is negligible, plots such as those depicted in Fig. 9 can be constructed giving surface ocean Pb concentrations at any site, at any time in the past. The 1984 reconstructions (Fig. 9b) appear reasonable in light of SEAREX and WATOX atmospheric fluxes [8,9,11], but a pair of surface water measurements near American Samoa of 17 and 21 pM by Flegal and Patterson [5] suggests that the Tutuila coral may have been subject to island influences. On the "pre-industrial" plot (Fig. 9a), the 1920 _+ 5 Fiji coral determination translates to a surface water dissolved Pb value of 16-19 pM (12-15 pM if K D -- 3.0), which may be slightly high if one accepts the South Pacific seawater data and a strong anthropogenic inva- sion to this region since 1920. Preliminary mass
300
50 ' P r e - I n d u s t n J a ] '
5 2~ ca_
4~
40
3O
2O
lO Fi)i
North RocK, Bermuda [est)
1 2 3 4 5 6 7 8
orida
Io
2OO
== 160
%' 120
8O
I 9 8 4
Florida
10 20 30 40
Bermuda SPreserve
5O
C0ral Pb/Ca (nm0l/m0l)
Fig. 9. Estimated dissolved lead concentrations in surface seawater corresponding to various coral sampling sites for (a) pre-industrial time and (b) 1984. Estimates are based on a universal K D - 2.3.
spectrometric measurements of the same coral sample (acid leached to 40% of initial mass) have yielded a Pb concentration of 4.4 nmol /mol Ca, confirming our own results (J. Chen C.I.T., per- sonal communication). If Flegal and Patterson are indeed correct in arguing a maximum pre- industrial Pacific surface water concentration of about 10 pM, then nearshore diagenetic Pb must be invoked as the source of the coral offset. By analogy to Bermuda, however, dissolved Pb anomalies 2.5 km away from Fiji are expected to be small. Therefore, the South Pacific prehistoric surface water Pb concentration was probably not much lower than 10 pM.
In contrast to the South Pacific, surface waters in the Florida Keys remained near 38 pM throughout the 18th and 19th centuries. Presuma- bly, this is a reflection of the importance of shelf a n d / o r resuspended Pb in waters overlying this coral reef.
3.5. The coral lead isotope record Industrial leads are isotopically very heteroge-
neous due to radiogenic production (from 23~U, 235U, and 232Th) a n d a multiplicity of lead ore genesis pathways [48]. The range of North Atlantic ratios exhibited in Fig. 2b (2°6/2°vPb=l.184- 1.215) represents only a few percent of the range bounded by the anomalously radiogenic Missis- sippi Valley-type ores of the U.S. (2°6/2°7pb ~ 1.30) and the non-radiogenic ores of Australia's Broken Hill Mine (2 °6 /2°7pb ~ ] . 00 ) . Yet, there is suffi- cient analytical precision within the range of his- toric American emissions to monitor changes in corals and seawater. Fig. 2b is the isotopic com- plement to the North Rock, Bermuda concentra- tion record of Fig. 2a. Eighty-five years ago, the 2°6/2°7pb isotopic signature of the surface ocean, 1.2147, was indistinguishable from the back- ground value [51], suggesting that an anthropo- genic component was non-existent or coinciden- tally bore and background signature. Industrial growth then prompted a migration toward less radiogenic values until the 1940's. None of the possible source industries discussed earlier can be dismissed on the basis of Pb isotopes since iso- topic data for iron and copper ores are not availa- ble, and coal and lead ore signatures can both be found in the required 1.180-1.190 2°6/2°7pb range [49,50]. Subsequently, due to changing feedstocks for alkyl lead and efficient Pb dispersal from internal combustion engines, the isotopic progres- sion was reversed. Introduction of more radio- genic Pb, possibly by admixture of Mississippi Valley ores, is implicated. In the mid-1970's, a second reversal occurred, roughly in parallel with the U.S. phase-out of leaded gasoline. As in the case of the Pb concentration record, the present- day isotopic value appears to roughly coincide with the pre-alkyl lead (ca. 1930-1940) value. A single cross-check is afforded by an isotopic Pb measurement of 1979 Sargasso Sea surface water by Ng et al. (in [4]). Their reported ratios of 1.199 ( 2 0 6 / 2 0 7 Pb) and 0.492 (2°6/2°8 Pb) agree closely with ratios measured in the 1980 band of D. strigosa: 1.203 and 0.492.
Due to the difficulty of the measurement, only partial records are presented for the other sample sites (Fig. 10). The results thus far concur with a global blanketing of industrial Pb. Anthropogenic influence in these remote regions is more reliably identified via isotopic shifts than by concentration anomalies because exact pre-anthropogenic con-
301
[ SEOIWEN~S/ wa MQOUtE£ :
i 2 ~ 0 Mauritius I rbri~a
5traits Gnl~aees N Zealana
/ daoan
~ 1170
1 i% I ! 130
-- ~ Straits
eGalaPaSos
~a u r i t ius
fu tu i la / /
V - T~
/ / I 18'80 IC30 15~0 157,2,
Y £ a P
!993
Fig. 10. Preliminary Pb isotopic histories for other sample sites. Left margin: background ranges estimated from [51,52]. Right margin: industrial emission data are from [6,50] and Fig. 2b.
centrations are as yet unknown. For example, the waters around Mauritius are clearly industrially influenced, despite very low skeletal Pb levels in P. rustica. The magnitude of these perturbations cannot be assessed from Pb isotopes until historic source signatures have been better characterized. Point determinations in time of industrial emis- sion signatures shown in the right margin of Fig. 10 only indicate that the measured departures from background are in the expected directions. The Tutuila record displays a curious trend away from Broken Hill-type Pb in recent years, which appears to be supported by a single surface water m e a s u r e m e n t ( 2 ° 6 / 2 ° 7 p b = 1.176) by Flegal et al. at 15°0 'S , 1 5 0 ° 0 ' W in January, 1980 [6]. Intro- duction of unleaded fuels in Australia over the last several years, however, has been too recent to account for the isotopic shift. Also, the coral Pb concentration record from Tutuila (Fig. 5) shows no concomitant change. A large-scale conversion in ore usage either regionally or locally is probably responsible. The Florida Keys data suggest that as late as 1929, input of industrial lead to these waters was minimal. This was also the turning point in the concentration record after which levels began to increase. The subsequent isotopic shift as indicated by the lone measurement at 1974 resem- bles that which occurred at Bermuda.
4. Conclusions
(1) This survey of stable Pb and Pb isotopes in corals from four major ocean basins confirms (by
independent means) the previously-inferred an- thropogenic dominance of Pb found throughout the surface ocean today, and over the past cen- tury.
(2) Direct lattice substitution of lead in coral- line aragonite is documented by (a) constancy of replicate measurements, (b) low background con- centrations, and (c) consistency of the temporal and regional response to known industrial fluxes. A Bermuda-based coral:seawater Pb distribution coefficient of 2.3 is estimated for the species, Diploria strigosa. K n dependence on species and genus is estimated to be small (estimated range = 2.3-3.0) based on similarity of 2i°pb K D estimates and plausibility of inferred seawater dissolved Pb concentrations.
(3) In view of the above conclusions, historic surface ocean conditions may be reconstructed as follows:
In Sargasso Sea surface waters, the pre-anthro- pogenic dissolved Pb concentration was probably between 15 and 20 pM (2°6/2°7pb-1.215). This value rose to near 90 pM in the 1920's as a result of the American industrial revolution (2°6/2°7 Pb = 1.184-1.190). Subsequent combustion of alkyl leads pushed the surface value to a maximum of 240 pM in 1971 (2°6/2°vpb = 1.202). A one-year time lag between this maximum and peak U.S. alkyl Pb combustion in 1970 reflects the brief mixed layer residence time of Pb. Since then, phasing-out of leaded gasoline has caused a pre- cipitous return to levels near those of the 1920's ( 2°6/2°7pb = 1.186)
The Florida Keys maintained a surface water concentration of 38 pM Pb until about 1930, which was probably supported by shelf/resus- pended Pb inputs. Levels grew gradually to peak of 190 pM in 1977, followed by a decline to 142 pM in 1982. Relative to the Bermuda records, the Florida coral lacks a strong industrial revolution signal and exhibits a moderated post-World War II Pb increase and muted maximum. These pat- terns reflect dilution of U.S. Pb sources and de- layed response due to long-range horizontal trans- port.
In the South Pacific, a single coral measure- ment from Fiji (1920 _+ 5 yr) implies a pre-in- dustrial surface water Pb concentration of 16-19 pM (12-15 pM if K D = 3.0). Although coastal diagenetic Pb may have biased this result, such a
302
c o n t r i b u t i o n is e x p e c t e d to be small . The re fo r e ,
S o u t h Pac i f ic p r e - i n d u s t r i a l su r face wa te r s were
p r o b a b l y n o t m u c h l ower t h a n 10 p M Pb. R e c e n t
cora l s f r o m T u t u i l a a n d G a l a p a g o s i m p l y h ighe r
c o n c e n t r a t i o n s o f 4 0 - 5 0 p M resu l t ing f r o m re-
g iona l i ndus t r i a l i n f l u e n c e a n d poss ib le loca l ef-
fects .
T h e co ra l f r o m M a u r i t i u s was too y o u n g to
fu rn i sh h i s to r i ca l pe r spec t ive , b u t c o n t e m p o r a r y
su r face wa te r Pb va lues are e s t i m a t e d at 2 5 - 2 9
p M . T h e Pb i so top i c s igna tu re of this cora l is
c lea r ly i nd i ca t i ve o f a n t h r o p o g e n i c p e r t u r b a t i o n .
(4) A survey of n e a r a n d i n sho re wa te r s a r o u n d
B e r m u d a suggests the f o l l o w i n g genera l t race ele-
m e n t s a m p l i n g s t ra teg ies for r e m o t e l y s i tua ted is-
lands . U n l e s s loca l p o l l u t a n t s ignals a re the s t udy
objec t , enc lo sed a n d pa r t i a l ly enc lo sed w a t e r w a y s
m u s t be avo ided . C o a s t a l s a m p l i n g m a y also be
p r o n e to local ou t fa l l and a d d i t i o n of Pb f r o m
s e d i m e n t s a n d re suspens ion . T h e l imi ts of these
a n o m a l i e s wil l v a r y f r o m i s land to is land, b u t the
B e r m u d a resul ts sugges t tha t 2.5 k m is p r o b a b l y a
safe m i n i m u m s a m p l i n g d i s t ance wi th r e spec t to
d i s so lved Pb c o n c e n t r a t i o n s .
(5) T h e p a l e o - c h e m i c a l r e c o r d i n g capab i l i t y of
cora l s of fers a sens i t ive m e a n s of r e c o n s t r u c t i n g
h i s to r i c Pb f luxes a n d i so top i c labels to the t em-
p e r a t e / t r o p i c a l ocean . F u t u r e t r a n s p o r t m o d e l i n g
o f Pb in the o c e a n can be ba sed on such r e c o n -
s t ruc t ions o f r eg iona l Pb t rans ien ts .
Acknowledgements
W e wish to t h a n k E l l en Dru f fe l , R i c k Fa i r -
banks , Bob B u d d e m e i e r , S teve Smi th , T e d M c -
C o n n a u g h e y , a n d Russ F r i t h for the i r gene ros i t y
in p r o v i d i n g cora l samples for this s tudy. S t an
H a r t p r o v i d e d r e a d y access to the mass spec-
t r o m e t e r u sed fo r all Pb i so tope d e t e r m i n a t i o n s .
R o b b i e S m i t h is t h a n k e d for his s a m p l i n g e f fo r t s
a n d ass i s tance whi l e wa t e r s a m p l i n g in B e r m u d a .
She i la G r i f f i n ' s expe r t i s e in s a m p l e p r e p a r a t i o n
was i n s t r u m e n t a l in p r o d u c i n g the A t l a n t i c cora l
records . J o h n G o d d a r d a n d R a n d i S c h n e i d e r a re
a l so t h a n k e d for thei r he lp in se lec t ing cora l sec-
t ions f r o m the Pac i f ic and I n d i a n Oceans . Susan
C h a p n i c k ' s ass i s tance in m e a s u r e m e n t o f d i s so lved
l e a d is g rea t ly app rec i a t ed . T h e m a n u s c r i p t be-
n e f i t t e d f r o m c o m m e n t s by Cla i r Pa t t e r son , M i k e
Bacon , and an a n o n y m o u s reviewer . Th is w o r k
was s u p p o r t e d by N S F g ran t O C E 8416382.
Appendix 1-- The lead distribution coefficient in aragonite
The standard means of expressing an enrichment or deple- tion of a lattice-bound element relative to seawater is em- bodied in the distribution coefficient KD:
([M]/[Ca])lamce K D -- ( 1 )
( [M]/[Cal) . . . . . . . .
Earlier, it was established by means of stable lead and 2mPb-based distribution coefficients that corals discriminate mildly in favor of Pb over Ca during skeletogenesis. A thermo- dynamic basis for this preference can be invoked if the solid solution behavior of PbCO 3 (cerrusite) and aragonite are ex- amined. Both of these mineral phases generally occur in near end-member composition, but limited solid solution with Pb and Ca (up to 3 tool %) has been reported [53]. The equi- librium constant for Pb substitution in aragonite:
CaCO 3 (s) + Pb 2~ ~ PBCO 3 (s) + Ca 2 + (2)
Xpbco3fPbCO , [Ca 2+ ] fCa 2+
K -- XCaCO,fCaC03 [ pb2 + ] fPb 2" (3)
is equivalent to equation (1) if the ratios of solid and aqueous phase total activity coefficients (including species complexing) are equal to unity. In this case, K D is given simply by the ratio of C a C O 3 (aragonite) and P b C O 3 solubility products in surface seawater. Since Ksp (aragonite) in surface seawater (25°C) is reported by Morse et al. [54] as 10 -618 mol2/kg 2, and Ksp
(PbCO3) (10 13.1 mol2/kg2 @ 25oC Smith and Martell [55]) can be corrected for seawater ionic strength by the Davies equation to give 10 11.8 molZ/kg2, a theoretical K D of 4 × 105 results. While Ca 2+ and Pb 2+ activity coefficients are ex- pected to be similar ( = 0.23), however, species complexing for the two elements are very different. Garrels and Thompson [56] calculated that 91% of dissolved Ca in seawater exists as the free C a 2+ ion. In contrast, CO 2 and C1 complexing of Pb reduces the free Pb 2+ concentration to only about 3% of the total Pb [57]. Thus, the effect of differential ion-association is to decrease K D to about 104. This value, nearly 6000 times higher than the observed K D resulting from biogenic precipita- tion of aragonite, suggests that the activity coefficient of the solid solute, fPbco3, must be very high--i.e, near 6000. A simple laboratory experiment demonstrated that inorganic pre- cipitation of aragonite gives results similar to coral-mediated precipitation. Clean seawater was spiked to a Pb concentration of 8.2 nM, seeded with a few clean coral crystals, and allowed to precipitate at 60 ° C for one month. Two such solutions were then filtered and the precipitates immersed in 0.1N HNO 3 to dissolve only the new crystal growth surrounding the coral seeds. Due to uncertainties stemming from adsorption of Pb onto container walls, the experimental K D ranged from 20 to 35; about an order of magnitude higher than the coral-based K D. Part of the enhanced discrimination for Pb is likely due to increased precipitation rate induced by the higher temperature and presence of seed crystals, as described by Lorens [58] for Sr precipitation in calcite. The experimental result, nonethe- less, supports preferential uptake of Pb over Ca during arago- nite precipitation.
More generally, it has become increasingly apparent that lattice-compatible trace elements are incorporated into coral aragonite in the same ratios to calcium as exist in seawater, regardless of differences in chemical speciation. Distribution coefficients for Ra, Ba, Nd, Sr, Cd, Co, and Zn in corals are all surprisingly close to unity, despite varying degrees of ion association (see [20]). While this pattern bespeaks simplicity, a thermodynamic basis is not obvious. One possibility is that in the supersaturated, possibly reducing environment of the coral polyp, carbonate complexes prevail, and these are indis- criminately coprecipitated with CaCO3 in a kinetically-con- trolled non-equilibrium process. The preference manifested toward Pb is a matter of speculation.
References
1 M. Murozumi, T.J. Chow and C. Patterson, Chemical con- centrations of pollutant lead aerosols, terrestrial dusts, and sea salts in Greenland and Antarctic snow strata, Geochim. Cosmochim. Acta 33, 1247-1294, 1969.
2 E.W. Wolff and D.A. Peel, The record of global pollution in polar snow and ice, Nature 313, 535-540, 1985.
3 B.K. Schaule and C.C. Patterson, Lead concentrations in the northeast Pacific: evidence for global anthropogenic perturbations, Earth Planet. Sci. Lett. 54, 97-116, 1981.
4 B.K. Schaule and C.C. Patterson, Perturbations of the natural lead depth profile in the Sargasso Sea by industrial lead, in: Trace Elements in Seawater, C.S. Wong, E. Boyle, K. Bruland D. Burton and E.D. Goldberg, eds., pp. 487-503, Plenum, New York, N.Y., 1983.
5 A.R. Flegal and C.C. Patterson, Vertical concentration profiles of lead in the central Pacific at 15 °N and 20 o S, Earth Planet. Sci. Lett. 64. 19-32, 1983.
6 A.R. Flegal, B.K. Schaule and C.C. Patterson, Stable iso- tope ratios of lead in surface waters of the central Pacific, Mar. Chem. 14, 281-287, 1984.
7 E.A. Boyle, S.D. Chapnick and G.T. Shen, Temporal varia- bility of lead in the western North Atlantic Ocean, Geo- chim. Cosmochim. Acta 91, 8573-8593, 1986.
8 D.M. Settle, C.C. Patterson, K.K. Turekian and J.K. Cochran, Lead precipitation fluxes at tropical ocean sites determined from 21°pb measurements, J. Geophys. Res. 87, 1239-1245, 1982.
9 D.M. Settle and C.C. Patterson, Magnitudes and sources of precipitation and dry deposition fluxes of industrial and natural leads to the North Pacific at Eniwetak, J. Geophys. Res. 87, 8857-8869, 1982.
10 T.D. Jickells, A.H. Knap and T.M. Church, Trace metals in Bermuda rainwater, J. Geophys. Res. 89, 1423-1428, 1984.
11 T.M. Church, J.M. Tramontano, J.R. Scudlark, T.D. Jickells, J.J. Tokos, A.H. Knap, and J.N. Galloway, The wet deposi- tion of trace metals to the coastal and western Atlantic Ocean, Atmos. Environ. 18, 2657-2664, 1984.
12 R.W. Talbot and A.W. Andren, Relationships between Pb and 21°pb in aerosol and precipitation at a semiremote site in northern Wisconsin, J. Geophys. Res. 88, 6752-6760, 1983.
13 K.K. Turekian, Y. Nozaki and L.K. Benninger, Geochem-
303
istry of atmospheric radon and radon products, Annu. Rev. Earth Planet. Sci. 5, 227-255, 1977.
14 H. Schuhmacher and H. Zibrowius, What is hermatypic? A redefinition of ecological groups in corals and other organisms, Coral Reefs 4, 1-9, 1985.
15 C. Emiliani, J.H. Hudson, B. Lidz, E.A. Shinn and R.Y. George, Oxygen and carbon isotopic record of growth in a reef coral from the Florida Keys and a deep-sea coral from Blake Plateau, Science 202, 627-629, 1978.
16 E.M. Druffel and T.W. Linick, Radiocarbon in annual coral tings of Florida, Geophys. Res. Lett. 5, 913-916, 1978.
17 R.W. Buddemeier, J.E. Maragos and D.W. Knutson, Ra- diographic studies of reef coral exoskeletons: rates and patterns of coral growth, J. Exp. Mar. Biol. Ecol. 14, 177-200, 1974.
18 J.H. Hudson, Growth rates in Montastrea annularis: a record of environmental change in Key Largo Coral Reef Marine Sanctuary, Florida, Bull. Mar. Sci. 31, 444-457, 1981.
19 E.M. Druffel, Radiocarbon in annual coral rings of Belize and Florida, Radiocarbon 22, 363-371, 1980.
20 G,T. Shen and E.A. Boyle, Determination of lead, cadmium and other trace metals in aragonitic annually-banded corals, Chem. Geol., submitted.
21 J.N. Weber, E.W. White and P.H. Weber, Correlation of density banding in reef coral skeletons with environmental parameters: the basis for interpretations of chronological records preserved in the corella of corals, Paleobiology 1, 137-149, 1975.
22 R.G. Fairbanks and R.E. Dodge, Annual periodicity of the 180/160 and 13C/12C ratios in the coral Montastrea an-
nularis, Geochim. Cosmochim. Acta 43, 1009-1020, 1979. 23 E.A. Boyle and J.M. Edmond, Determination of trace
metals in aqueous solution by APDC chelate co-precipita- tion, in: Advances in Chemistry, Vol. 147, pp. 44-55, American Chemical Society, Washington, D.C., 1975.
24 E.A. Boyle and J.M. Edmond, Determination of copper, nickel and cadmium in seawater by APDC Chelate coprecipitation and flameless atomic absorption spectrome- try, Anal. Claim. Acta 91, 189-197, 1977.
25 G. Manhes, J.F. Minster and C.J. All~gre, Comparative uranium-thorium-lead and rubidium-strontium study of the Saint Severin Amphoterite: Consequences for early solar system chronology, Earth Planet. Sci. Lett. 39, 14-24, 1978.
26 W.W. Flynn, The determination of low levels of polonium- 210 in environmental materials, Anal. Chim. Acta 43, 221-227, 1968.
27 G. Schott, Geographie des Atlantischen Ozean, 1944. 28 Bureau of the Census of the U.S. Department of Com-
merce, The Statistical History of the United States from Colonial Times to the Present, Vols. 1965-1984, Fairfield Publishers, Inc., Stanford, Conn.
29 M.P. Bacon, D.W. Spencer and P.G. Brewer, 21°pb/226Ra and 210 Po/210 Pb disequilibrium in seawater and suspended particulate matter, Earth Planet. Sci. Lett. 32, 277-296, 1976.
30 Y. Nozaki, J. Thomson and K.K. Turekian, The distribu- tion of 21°pb and 21°po in the surface waters of the Pacific Ocean, Earth Planet. Sci. Lett. 32, 304-312, 1976.
304
31 T.D. Jickells, W.G. Deuser and A.H. Knap, The sedimenta- tion rates of trace elements in the Sargasso Sea measured by sediment trap, Deep Sea Res. 31, 1169-1178, 1984.
32 J.O. Nriagu, Global inventory of natural and anthropo- genic emissions of trace metals to the atmosphere, Nature, 279, 409-411, 1979.
33 C.C. Patterson, D. Settle and B. Glover, Analysis of lead in polluted coastal seawater, Mar. Chem. 4, 305-319, 1976.
34 V.J. Stukas and C.S. Wong, Stable lead isotopes as a tracer in coastal waters, Science 211, 1424 1427, 1981.
35 K.W. Bruland, K. Bertine, M. Koide and E.D. Goldberg, History of metal pollution in southern California coastal zone, Environ. Sci. Technol. 8, 425-432, 1974.
36 R.E. Dodge and T.R. Gilbert, Chronology of lead pollution contained in banded coral skeletons, Mar. Biol. 82, 9-13, 1984.
37 J.H. Hudson, E.A. Shinn, R.B. Halley and B. Lidz, Sclerochronology: A tool for interpreting past environ- ments, Geol. 4, 361-364, 1976.
38 J.L. Sarmiento and K. Bryan, An ocean transport model for the North Atlantic, J, Geophys. Res. 87, 394-408, 1982.
39 Y. Nozaki and S. Tsunogai, Lead-210 in the North Pacific and the transport of terrestrial material through the atmo- sphere, Earth Planet. Sci. Lett. 20, 88-92, 1973.
40 J. Thomson and K.K. Turekian, 21°po and 2wPb distribu- tions in ocean water profiles from the Eastern South Pacific. Earth Planet. Sci. Lett. 32, 297-303, 1976.
41 S. Tsunogai and Y. Nozaki, Lead-210 and polonium-210 in the surface water of the Pacific, Geochem. J. 5, 165-173, 1971.
42 J.K. Cochran, M.P. Bacon, S. Krishnaswami and K.K. Turekian, 21°po and 21°Pb distributions in the central and eastern Indian Ocean, Earth Planet. Sci. Lett. 65, 433-452, 1984.
43 R.E. Dodge and J. Thompson, The natural radiochemical and growth records in contemporary hermatypic corals from the Atlantic and Caribbean, Earth Planet. Sci. Lett. 23, 313-322, 1974.
44 R.W. Buddemeier, Environmental controls over annual and lunar monthly cycles in hermatypic coral calcification, in: Proc. 2nd Int. Coral Reef Symp. Brisbane, Vol. 2, pp. 259-267, 1974.
45 R.E. Dodge and G.W. Brass, Skeletal extension, density and calcification of the reef coral, Montastrea annularis: St. Croix, U.S. Virgin Islands, Bull. Mar. Sci. 34, 288-307.
46 G.T. Shen, E.A. Boyle and D.W. Lea, Cadmium in corals: chronicles of historic upwelling and industrial fallout, Na- ture, submitted.
47 J.N. Weber and P.M.J. Woodhead, Temperature depen- dence of oxygen-18 concentration in reef coral carbonates, J. Geophys. Res. 77, 463-473, 1972.
48 J.S. Brown, Oceanic lead isotopes and ore genesis, Econ. Geol. 66, 47-68, 1965.
49 T.J. Chow and J.L. Earl, Lead isotopes in North American coals, Science 176, 510-511, 1972.
50 B.R. Doe, Lead Isotopes, Springer Verlag, New York, N.Y., 1970.
51 T.J. Chow and C.C. Patterson, The occurrence and signifi- cance of lead isotopes in pelagic sediments, Geochim. Cos- mochim. Acta 26, 263-308, 1962.
52 S,S. Sun, Lead isotopic study of young volcanic rocks from mid-ocean ridges, ocean islands and island arcs, Philos. Trans. R. Soc. London A 297, 409-445, 1980.
53 J.A. Speer, Crystal chemistry and phase relations of ortho- rhombic carbonates, in: Carbonates: Mineralogy and Chemistry, pp. 145-189, Mineralogical Society of America, Chelsea, Mich., 1983.
54 J.W. Morse, A. Mucci and F.J. Millero, The solubility of calcite and aragonite in seawater at 35% salinity at 2 5 ° C and atmospheric pressure, Geochim. Cosmochim. Acta 44, 85-94, 1980.
55 R.M. Smith and A.E. Martell, Critical Stability Constants, Vol. 4, Plenum, New York, N.Y., 1976.
56 R.M. Garrels and M.E. Thompson, A chemical model for sea water at 25 ° C and one atmosphere total pressure, Am. J. Sci. 260, 57-66, 1962.
57 P.R. Turner, M. Whitfield and A.G. Dickson, The equi- librium speciation of dissolved components in freshwater and seawater at 25 ° C and 1 atm pressure, Geochim. Cos- mochim. Acta 45, 855-881, 1981.
58 R.B. Lorens, Sr, Cd, Mn and Co distribution coefficients in calcite as a function of calcite precipitation rate, Geochim. Cosmochim. Acta 45, 553-561, 1981.
