Stable carbon and oxygen isotope fractionation in bivalve (Placopecten magellanicus) larval aragonite

Available online at www.sciencedirect.com

Geochimica et Cosmochimica Acta 72 (2008) 4687–4698 www.elsevier.com/locate/gca

Stable carbon and oxygen isotope fractionation in bivalve (Placopecten magellanicus) larval aragonite Erin F. Owen a,*, Alan D. Wanamaker Jr. b, Scott C. Feindel c, Bernd R. Scho¨ne d, Paul D. Rawson a a School of Marine Sciences, University of Maine, Orono, ME 04469, USA Department of Earth Sciences and Climate Change Institute, University of Maine, Orono, ME 04469, USA c Darling Marine Center, University of Maine, Walpole, ME 04535, USA d Department of Applied and Analytical Paleontology and INCREMENTS Research Group, Institute of Geosciences, University of Mainz, 55128 Mainz, Germany b

Received 12 June 2007; accepted in revised form 24 June 2008; available online 24 July 2008

Abstract The relationship between stable isotope composition (d13C and d18O) in seawater and in larval shell aragonite of the sea scallop, Placopecten magellanicus, was investigated in a controlled experiment to determine whether isotopes in larval shell aragonite can be used as a reliable proxy for environmental conditions. The linear relationship between d13CDIC and d13Caragonite (r2 = 0.97, p < 0.0001, RMSE = 0.18) was: d13 CDIC ¼ 1:15ð0:05Þ  d13 Caragonite  0:85ð0:04Þ The relationship between d13CDIC and d13Caragonite described for P. magellanicus resulted in larval shell aragonite that was depleted on average by 1.82& (SD = 0.22&, range = 1.1–2.1&) from predicted equilibrium values based on the relationship calibrated for inorganic aragonite. The average contribution of metabolic carbon that resulted in this depletion was 5.4% (SD = 0.57%; range = 3.4–7.8%). Stable oxygen isotopes were deposited into the larval shell in equilibrium for most samples, and the linear relationship described by least squares regression between temperature and d18Oaragonite–d18Owater (r2 = 0.90, p < 0.0001, RMSE = 0.63) was: T ð CÞ ¼ 20:0ð0:4Þ  4:6ð0:3Þ  ðd18 Oaragonite  d18 Owater Þ However, larvae reared under ‘‘stressful” conditions were depleted from oxygen isotope equilibrium. Further studies are necessary to determine the variable contribution of metabolic carbon to the larval shell in field conditions, the potential effects of growth rate on carbon isotope composition, and the factors influencing oxygen isotope depletion in P. magellanicus larval shell before the isotope composition of larval shells can be used to reconstruct d13CDIC or temperature of the seawater in which the larvae developed. Ó 2008 Elsevier Ltd. All rights reserved.

1. INTRODUCTION Stable carbon and oxygen isotope signatures of biogenic calcium carbonate structures are indispensable tools to reconstruct the physical and chemical properties of ancient *

Corresponding author. Fax: +1 207 581 2537. E-mail address: [email protected] (E.F. Owen).

0016-7037/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.gca.2008.06.029

and modern seawater. The carbon isotope composition of calcareous skeletons of invertebrates has been used to estimate ocean circulation patterns (Duplessy et al., 1988; Lynch-Steiglitz and Fairbanks, 1994), water column productivity (Woodruff and Savin, 1985), and anthropogenic CO2 inputs (Beveridge and Shackelton, 1994), while the oxygen isotope composition of biogenic carbonates has been used to reconstruct seawater temperature (Epstein

4688

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

et al., 1953). Recent studies have investigated the application of geochemical tools to ecological questions in marine systems. For example, the carbon and oxygen isotope composition of fish otoliths and mollusk shells has been used to determine growth rates (Krantz et al., 1984), as well as spawning stocks and stock structures (Gao et al., 2001, 2005). Over the past decade, the ability to interpret stable isotope and trace metal signatures of carbonate structures from mobile larval stages of many sedentary benthic invertebrates, such as mollusks, has increased the capacity to more accurately estimate rates of dispersal of marine organisms, implement effective marine reserve design, understand the spread of invasive species, and manage fisheries resources (reviewed in Levin, 2006). The technology available for geochemical analyses of biogenic carbonates, particularly the calcite and aragonite forms of calcium carbonate, in adult structures has promising application to larval structures as well. Larval mollusks are abundant and diverse both in contemporary marine systems and in the fossil record (Hansen, 1981, 1984), and larval shells of gastropods and bivalves are readily identifiable to species from morphological characteristics (Hansen, 1984; Tremblay et al., 1987). However, studies assessing the stable isotope composition of mollusk larvae are rare (Killingley and Rex, 1985; Malchus and Steuber, 2002), and many studies on fossil larvae have focused on shell morphology as a palaeoclimatic and biostratigraphic tool (Lutz and Jablonski, 1978; Hansen, 1984) rather than exploring the utility of shell geochemistry as a potential tool for environmental reconstruction. The ability to reconstruct environmental parameters from stable isotopes in larval shell aragonite relies on establishing a predictable relationship between the stable isotope composition of the larval shell and known seawater properties. The equilibrium geochemical relationships for stable carbon and oxygen isotopes in biogenic aragonite have been calibrated for live or modern adult invertebrates or fish, including foraminifera, mollusks, scaphopods, and fish otoliths (Grossman and Ku, 1986; Thorrold et al., 1997) as well as synthetic aragonite (Romanek et al., 1992). However, deviations from equilibrium are common among many taxa and result from either kinetic or metabolic effects (McConnaughey, 1989a). Potential kinetic effects include discrimination against the heavier isotopes of carbon and oxygen during hydration and hydroxylation of CO2, as well as the disparate incorporation of lighter isotopes due to the higher diffusivity of HCO 3 molecules containing 12C or 16O (McConnaughey, 1989a,b; Kalish, 1991). Metabolic effects occur when metabolic or respiratory carbon, which is generally depleted in 13C compared with the DIC of the surrounding seawater, is incorporated into carbonate structures (McConnaughey, 1989a,b, 1997). However, the equilibrium relationships for stable oxygen and carbon isotopes have not been calibrated for mollusk larvae, although a few studies have applied the stable isotope relationships established for adults to studies on larvae (Killingley and Rex, 1985; Malchus and Steuber, 2002). Because the stable isotope composition of shell carbonate can be affected by metabolism and shell precipitation rate (e.g.

McConnaughey et al., 1997), it is possible that the equilibrium relationships for adults and larvae could be different. The objective of this study was to empirically determine the relationships between environmental conditions and stable carbon and oxygen isotopes in the larval shell aragonite of the Atlantic sea scallop, Placopecten magellanicus, to compare these relationships with those previously established for other mollusks, and to assess the utility of sea scallop larvae as an informative proxy of environmental characteristics for palaeoclimate and ecological studies. Sea scallop larvae were reared under controlled conditions and the stable isotope composition of the seawater (d13CDIC(Dissolved Inorganic Carbon) and d18Owater) and larval shell (d13Caragonite and d18Oaragonite) was monitored throughout the period of shell deposition. Fractionation relationships for stable carbon and oxygen isotopes in larval shell aragonite were established and compared with equilibrium relationships established for adult mollusk and synthetic aragonite. In addition, shell size and growth rates were monitored to determine the influence of shell deposition rate on the isotopic composition of the larval shell. 2. MATERIALS AND METHODS 2.1. Study organism The Atlantic sea scallop, Placopecten magellanicus, is a commercially important bivalve mollusk (Family Pectinidae) found exclusively in the northwest Atlantic Ocean, from the north shore of the Gulf of St. Lawrence to Cape Hatteras, North Carolina (Posgay, 1957). P. magellanicus is a filter feeder that typically occurs in dense aggregations or ‘beds’ found in benthic habitats with firm sand, gravel, shells, and cobble substrate at depths of 18–110 m, although occasional beds may be found as shallow as 2 m at the northern extent of the species’ geographic range (Hart and Chute, 2004). P. magellanicus is an iteroparous species that reproduces annually or semi-annually, with the peak spawning period from August through September in northern locations (Barber et al., 1988; DiBacco et al., 1995) and in May or June in the mid-Atlantic region (DuPaul et al., 1989). Eggs are fertilized externally and the length of the planktonic larval stage is approximately 40 days (Culliney, 1974). Sea scallops are relatively long-lived (>12 years) and often reach adult sizes great than 150 mm shell height (Naidu, 1991). As with many bivalves, P. magellanicus undergoes an ontogenetic shift in shell calcium carbonate composition during metamorphosis from the prodissoconch shell of the veliger larva to the dissoconch shell of the juvenile and adult. While the sea scallop dissoconch is composed of calcite, the prodissoconch, like that of other bivalves, is composed of aragonite (Weiss et al., 2002). Because the prodissoconch of bivalve larvae also can contain a variable proportion of amorphous calcium carbonate as a precursor for aragonite (Weiss et al., 2002), in this study powder x-ray diffraction was performed on sea scallop larval shell to confirm the composition was aragonite.

Isotope fractionation in bivalve larval aragonite

4689

Reproductively mature P. magellanicus were collected by SCUBA diver from Penobscot Bay (44°250 2000 N/ 68°520 4400 W) or Blue Hill Bay (44°240 1900 N/68°340 2000 W) in early August 2005 and 2006. Adults were transported on ice to the Darling Marine Center in Walpole, Maine and held in flowing seawater tanks at ambient water temperature (approx. 16 °C). Spawning was induced within one week of collection by immersing scallops in a shallow tray filled with sea water, filtered with a 1-lm polypropylene filter, and alternately warming and chilling the water. In 2005, eggs obtained from one female were washed with filtered sea water (FSW), resuspended in 10 L FSW, and fertilized with a mixture of sperm from five males. At an early developmental stage (2+ cells, 1.5 h of development), 300,000 embryos were randomly allocated to each of eight 22 L polycarbonate culture buckets filled with 15 L FSW, for a final density of 20 embryos ml1. Four culture buckets were assigned to each experimental block (A or B), and buckets were further assigned to each of two treatment levels of salinity and temperature within each block. One large recirculating water bath was prepared for each of the two temperature levels and culture buckets for blocks A and B at each of the two salinity levels were immersed in each water bath, for a total of four buckets in each water bath. Temperature and salinity treatments were within the range that P. magellanicus larvae would encounter during planktonic development periods from July through September in the Gulf of Maine. Salinity treatments included a 30.0 psu (High S) ambient Damariscotta River treatment, as well as a 26.2 psu (Low S) treatment in 2005. The Low S treatment was generated by mixing high salinity water with untreated well water, using adequate aeration to ensure complete mixing, following the methods in Wanamaker et al. (2006). Isotope mixing lines, or the d13CDIC and d18Owater values for each salinity treatment, were calculated. Water for each salinity treatment was prepared in advance, archived in 625 gallon polyethylene containers with lids (ACE ROTO-MOLD), and used for the duration of the experiment. All salinity measurements were made with a YSI model 85 oxygen, conductivity, salinity, and temperature system with an accuracy of ±0.1 psu (Table 1). Temperature treatments included a 15.6 °C treatment (High T) at which optimum growth for sea scallops has been demonstrated (Culliney, 1974), as well as a 12.5 °C (Med. T) treat-

ment in 2005. Experimental temperatures were obtained by partially immersing experimental culturing tanks in a temperature-regulated water bath (see Wanamaker et al., 2006 for complete description of system). Temperature loggers (iB Cod; Alpha Mach, Inc.) recorded water temperature for one culture bucket within each recirculating water bath every 30 min in 2005 for the duration of the experiment (Table 2). When sea scallop larvae had developed to the shelled Dstage (approximately day three), each treatment bucket was fed 50 ll day1 (7000 cells ml1) of concentrated Shellfish Diet 1800 (Reed Mariculture). The amount of diet was similar for all buckets and increased as the larvae developed, to 100 ll day1 (14,000 cells ml1) on days 4–6, 200 ll day1 (28,000 cells ml1) on days 7–10, and 250 ll day1 (34,000 cells ml1) on days 11–17. Shellfish Diet 1800 is a blend of five microalgae: Isochrysis sp., Pavlova sp., Tetraselmis sp., Thalassiosira weissflogii, and Nannochloropsis sp. This diet is not considered optimal for sea scallop larval growth (Gouda et al., 2006). However, the use of small volumes of a prepared diet minimized the potential effect of food addition on the isotopic composition of the water. Each treatment bucket was aerated lightly through inert glass and complete water changes were made every 2– 3 days. During water changes, larvae were sieved gently to remove metabolic wastes and any uneaten microalgae, and transferred carefully into fresh treatment water that had been pre-chilled to the experimental temperature by partial submersion in the water bath. Larvae were reared for a total of 17 days in 2005. The calibration experiment was repeated in 2006 to expand the experimental temperature and salinity ranges. The general methods for spawning and larval culturing were the same in 2006 as in 2005, but differed in the number of individuals spawned and the number of treatments. In 2006, eggs obtained from eight females were combined, washed with FSW, resuspended in 10 L FSW, and fertilized with a mixture of sperm from six males. Embryos (20 embryos ml1) were allocated to each of eighteen culture buckets and nine of these buckets were assigned to each experimental block (A or B). Within each block, buckets were assigned to each of three treatment levels each of salinity and temperature. Three large recirculating water baths, one for each temperature level, was prepared and culture buckets for blocks A and B at each of three salinity levels were placed in each water bath, for a total of six culture buckets in each water bath. Salinity treatments in 2006 in-

Table 1 Summary of seawater salinity (S) conditions for each experimental treatment

Table 2 Summary of seawater temperature (T) conditions for each experimental treatment

2.2. Larval culturing

Treatment level

Mean S (psu) ± SD

2005 Low S High S

26.2 ± 0.00 30.0 ± 0.00

2006 Low S Med. S High S

26.3 ± 0.10 28.5 ± 0.05 31.2 ± 0.06

Min. S

Max. S

Treatment level

Mean T (°C) ± SD

Min. T

Max. T

26.2 30.0

26.2 30.0

2005 Med. T High T

12.5 ± 0.8 15.6 ± 0.4

10.6 14.4

14.1 16.4

26.2 28.5 31.1

26.4 28.6 31.2

2006 Low T Med. T High T

9.2 ± 0.2 12.2 ± 1.1 14.8 ± 0.3

8.6 9.6 14.4

10.5 13.9 15.4

4690

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

cluded a 31.2 psu (High S) ambient Damariscotta River treatment, as well as a 28.5 psu (Med. S) and 26.3 psu (Low S) treatment (Table 1). Temperature treatments included a 14.8 °C treatment (High S), as well as a 12.2 °C (Med. T) and 9.2 °C (Low T) treatment. Water temperature was recorded from one culture bucket within each recirculating water bath every 15 min for the duration of the experiment (Table 2). When sea scallop larvae had developed to the shelled D-stage (approximately day three), each treatment bucket was fed an aliquot of concentrated Shellfish Diet 1800 (Reed Mariculture). The amount of diet was consistent among all buckets and increased as the larvae developed, from 50 ll day1 (7000 cells ml1) on days 3–4, to 100 ll day1 (14,000 cells ml1) on days 5–11, and 150 ll day1 (21,000 cells ml1) on days 12–17. The experiment was ended after 17 days. 2.3. Larval shell height and growth rate To compare the final shell size of larvae from each treatment, a sample of larvae from each experimental bucket was preserved in 95% ethanol at the termination of the experiment. Larval shell height (distance from hinge line to shell margin) was measured for 20 individuals from each experimental bucket in 2005 and 2006. Larvae were imaged using a SPOT camera mounted on a compound microscope and measured using the calibrated SPOT software measuring tool. If no visible shell material was present, shell height could not be measured and the larval culture was assumed to be dead. The final shell size of larvae was compared using an ANOVA for a complete random block design with shell height as the dependent variable and temperature and salinity as independent variables. When significant treatment effects were found, posthoc Tukey multiple comparisons of treatment means were used to determine the nature of the treatment effects. In addition to final shell size, growth rates were estimated for each experimental treatment in 2006. A sample of larvae was preserved from each bucket approximately every 2–3 days for the duration of the experiment, for a total of 2–5 measurements for each experimental bucket. For each experimental treatment, the slope of the least squares regression of shell height against time was used to generate a growth rate estimate with a 95% confidence interval for both blocks A and B. Confidence intervals were used to compare growth rates among treatments within each block. 2.4. Isotope measurements in water 2.4.1. d13C in DIC Water samples for d13CDIC were collected beginning on day 10, but only in the 2006 experiment. Samples were collected weekly from each archived water tank, for a total of two water samples for each tank, and every three to four days from each experimental bucket, for a total of three samples per bucket. Water was filtered through a 0.45-lm PTFE filter into 40 ml amber borosilicate glass vials (ICHEM) containing 0.5 ml of saturated mercuric chloride. Fixed water samples were stored at 4 °C until analyzed. Samples were run on an OI Analytical TIC-TOC Analyzer Model 1010. Each sample was run twice following the

methods in St-Jean (2003); the first run determined the ppmC organic/inorganic concentration and the second run determined the d13C isotope concentration. The TICTOC analyzer was interfaced to a Finnigan Mat DeltaPlus isotope ratio mass spectrometer for analysis by continuous flow, and data were normalized using internal standards. The 2r analytical precision is 2 ppb C or 2% (whichever is higher) for quantitative carbon analyses, and ± 0.2& for isotope analyses. The mean d13CDIC from each experimental bucket was used in the isotope calibration. 2.4.2. d18O in water Water samples for d18Owater measurements were collected from archived tank water every two to three days during culture water changes in 2005, for a total of five water samples for each salinity treatment. In addition, water was collected from one culture bucket per salinity treatment at the termination of the experiment. In 2006, water samples were collected from archived tank water at the beginning of the experiment and approximately weekly, for a total of four water samples from each salinity treatment. Water samples also were collected weekly from each experimental bucket during water changes, for a total of three water samples from each bucket. All water samples were stored at 4 °C in tightly sealed 15 ml polyethylene tubes until analyzed. Samples from 2006 were filtered through a 0.45-lm PTFE filter before storing. d18Owater was measured via a dual-inlet VG/Micromass SIRA (CO2–H2O equilibration method at 30 °C for 12 h), with a precision of ±0.07& based on replicate International Atomic Energy Agency (IAEA) laboratory standards [Vienna Standard Mean Ocean Water (VSMOW, 0.0&), Standard Light Antarctic Precipitation (SLAP, 55.5&), IAEA OH-1 (0.1&), IAEA OH-2 (3.3&), IAEA OH-3 (8.7&), IAEA OH-4 (15.3&), and internal laboratory standards Big Bear Brook (BBB, 8.5&), Light Antarctic Precipitation (LAP, 40.3&), Antarctic Surface Snow (ASS; 25.8&)], where one type of standard was run for every five samples. All d18Owater (d in & = [(Rsample/Rstandard)  1] * 1000; [R = 18 O/16O]) values are reported with respect to VSMOW. In 2005, the d18Owater from each salinity environment was averaged over the 17-day growing interval and used in the isotope calibration. In 2006, the measurements from the experimental buckets made it possible to use the mean d18Owater from each bucket for the isotope calibration. 2.5. Stable isotope measurements in shell and food 2.5.1. d18O and d13C in shell Larval shell samples were collected from each experimental bucket at the termination of the experiment, for a total of eight bulk samples in 2005 and 18 bulk samples in 2006. Larvae were filtered onto 47 mm membrane filters that were inserted into 50 ml polyethylene tubes and stored at 20 °C until analyzed. Prior to analysis, larvae were airdried for 1–3 days and then oven dried overnight at 40 °C. In 2005, shell carbonate analysis (d18Oaragonite and 13 d Caragonite) was performed on a dual-inlet VG/Micromass

Isotope fractionation in bivalve larval aragonite

Prism stable isotope ratio mass spectrometer, via a 30-place carousel and common phosphoric acid bath without chromium oxide (CrO3) at 90 °C, with precision of ±0.07& (d18O) and ±0.06& (d13C) based on laboratory standards. Results are reported relative to Vienna Pee-Dee Belemnite (VPBD) by calibration to the NBS-19 reference standard (d18O = 2.20& and d13C = 1.95&) at the beginning and end of each run, with a standard to sample ratio of 1:3. Average carbonate samples weighed approximately 100 lg. In 2006, each carbonate sample contained hundreds of prodissoconchs and weighed 56–127 lg. The samples were reacted with 99% anhydrous phosphoric acid at 72 °C in an automated carbonate preparation device (Gas Bench II) coupled to a Finnigan MAT-253 mass spectrometer. Oxygen and carbon isotope ratios of the samples are reported relative to the VPDB carbonate standard based on NBS-19 reference standard. For all analyses, corrections for 17O have been made according to Craig (1957). Precision of the instrument determined from replicate analyses was ±0.07& for d18O and ±0.04& for d13C. 2.5.2. d13C in food source To determine the isotopic composition (d13C) of the food (algal paste) used in this study, 30 ml of paste was dried at 50 °C for 24 h and then weighed into tin capsules. The dried samples were flash combusted at 1800 °C in an elemental analyzer (EA) [EA 1110 (CE Instruments) + Conflo III + DeltaPlus Advantage IRMS (ThermoFinnigan)] and resulting gases were carried via helium through the EA to purify and separate into N2 and CO2. Gases were carried from the EA into an isotope ratio mass spectrometer (IRMS) for isotope analysis via a Conflo interface. Based on replicate standard analyses, the analytical precision (2r) is ±0.2&. Results are reported relative to VPBD by calibration to internal reference standards. 3. RESULTS 3.1. Larval shell height and growth rate The overall effects of temperature and salinity treatments on variation in shell height and growth rate were similar across the 2005 and 2006 experiments. In both 2005 and 2006, final shell height was smallest in the lowest temperature and salinity treatment. In general, larvae reared during 2005 had a larger shell height than larvae reared during 2006. Since the experimental protocol was similar during the two years, the variation in shell height may have been due to differences in parentage. However, slight changes in experimental temperature and salinity precluded the absolute comparison of larval shell height between the two experiments. In the 2005 experiment, mean shell height ranged from approximately 105 lm in both blocks of the Med. T/Low S treatment to 124.2 lm in block A of the High T/High S treatment (Fig. 1a). Raw residuals exhibited a non-normal distribution, so data were 1/y transformed to conform to ANOVA assumptions. A significant interaction was found between block and salinity (p = 0.002). Therefore, the treatment effects were analyzed by block for subsequent analy-

4691

ses. For both blocks A and B, ANOVA indicated significant main effects of temperature (A: p = 0.004, B: p = 0.000) and salinity (A: p = 0.000, B: p = 0.000), as well as a significant interaction term (A: p = 0.038, B: p = 0.000). Post-hoc Tukey multiple comparisons of treatment means showed that within both blocks A and B, larvae reared at Med. T/Low S had smaller shell height than all other at treatments at high temperature and salinity combinations. In addition, in block A the larvae reared at High T/Low S were larger than the larvae reared at Med. T/Low S, but smaller than the larvae at both temperature levels at High S. Although the actual difference in shell sizes among treatments within each block was small (<20 lm difference between smallest and largest shells), the significant differences in estimates of treatment means reflects the low variability in shell height measurements within each treatment. Similar variation in shell height was observed among temperature and salinity treatments in the 2006 experiment. Larvae in the Low S treatments at Low T and block B of Med T died and no data were collected. The actual difference in shell heights among remaining treatments in 2006 was approximately 35 lm; mean shell height ranged from 75.6 lm in block A of the Med. T/Low S treatment to 110.6 in block A of the High T/High S treatment (Fig. 1b). In contrast to the 2005 experiment, raw residuals were distributed normally so no transformation was applied. ANOVA indicated significant main effects of temperature (p = 0.000), salinity (p = 0.000), and a significant interaction term (p = 0.003). Post-hoc Tukey multiple comparisons of treatment means showed that larvae reared at Low T/Med. S, as well as at Med. T/Low S, were smaller than larvae from all other treatments, but not significantly different from each other. Larvae reared at Low T/High S were larger than the larvae reared at Low T/Med. S, as well as Med. T/Low S, but were smaller than larvae reared at all other treatment conditions. In addition, larvae reared at High T/Low S were smaller than larvae reared at Med. T/High S, and at High T/Med. and High S. Finally, larvae reared at High T/High S were larger than larvae reared at Med. T/Med. S. The significant differences observed among estimates of treatment means, despite the relatively small differences in actual shell height, reflect the low variability in shell height measurements within each treatment. Lower temperature and salinity treatments resulted in reduced growth during the 2006 experiment (Table 3). Larvae in the Low S treatments at Low T and block B of Med. T died and no data were collected. In the remaining treatments, growth rates separated into two groups: not different from zero based on the 95% confidence interval (Low T/ Med. S, block A at Low T/High S, Med. T /Low S, and block B at High T/Low S) or different from zero but not different from each other (all other treatments). Zero growth in a culture suggests that although the larvae remained alive, most of the larvae in the treatment accumulated little or no shell material after the initial deposition of the prodissoconch as D-stage larvae. The growth rates of larvae in the remaining treatments were not significantly different from each other, and maximum growth rate esti-

4692

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

Shell Height (μm) ± SE

140 120

2005

1,2

* *

*

100 80 60 40 20 0

n/a

n/a

A

B

A

B

A

120

*5

2006 Shell Height (μm) ± SE

B

*5

*4

100

*3 *

*

80

*4

*

*3 *

60

40

20

0

ND

ND

ND

A

A

B

Low T

B

A

Med. T

B

High T

Fig. 1. Final shell height in lm (SE) for larvae (n = 20) from each experimental temperature (T) and salinity (S) treatment in 2005 (top) and 2006 (bottom). White bars, low salinity; grey bars, medium salinity; black bars, high salinity. A and B indicate experimental blocks. ND, no data because the larval culture died. *significantly smaller than all other treatments at the p < 0.05 family confidence level for pairwise comparisons. *1,2, smaller than Med. and High T/High S, but larger than Med. T/Med. S. *3, larger than Low T/Med. S and Med. T/Low S, but smaller than all other treatments. *4, smaller than Med. T/High S and High T/Med. and High S. *5, larger than Med. T/Med. S.

Table 3 Larval growth rate in lm/day (lower, upper 95% confidence interval) for each experimental temperature (T) and salinity (S) treatment in 2006, as estimated by the slope of the least squares regression line Salinity

Low S Med. S High S

Low T

Med. T

High T

A

B

A

B

A

B

ND 0.30(1.9, 2.5)* 1.3(0.04, 2.7)

ND 0.19(2.7, 3.0)* 1.6(1.3, 1.9)

0.07* 1.8(0.84, 2.7) 2.1(1.4, 2.7)

ND 1.7(1.5, 2.0) 2.3(2.0, 2.6)

1.7(0.97, 2.5) 1.8(1.4, 2.2) 2.1(1.3, 2.8)

1.1(2.1, 4.3)* 1.7(0.63, 2.9) 1.6(0.39, 2.9)

A and B indicate experimental blocks. ND, no data because the larval culture died. *Not significantly different from zero. No confidence interval is presented for block A at Low S/Med. T because the rate was estimated based on only two time points.

mates ranged from just over zero to 2.3 lm/day. The large confidence interval on many growth rate estimates is due to the cross-sectional nature of the data (destructive sampling of larvae in a culture over time, rather than tracking the growth of an individual larva) and reflects the variability in the amount each treatment grew over time, rather than the variability in the shell height measurements at each time point. It is interesting to note that differences in shell height at the end of the experiment existed among treatments that lacked measured differences in growth rate (Fig. 1b). This suggests that the differences in shell height among treat-

ments may be established by the time of initial shell deposition as D-stage larvae. 3.2. Carbon and oxygen isotope mixing lines Carbon isotope (d13CDIC) composition of experimental treatment water was directly proportional to salinity. The least squares regression (±95% confidence interval) mixing line described the tight linear relationship between d13CDIC and salinity (r2 = 0.99, p < 0.0001, root mean squared error (RMSE)=0.12; Fig. 2a) as:

Isotope fractionation in bivalve larval aragonite

4693

isotope measurements of water and shell during the 2006 experiment (Fig. 3). The linear relationship between d13CDIC and d13Caragonite in 2006 (r2 = 0.97, p < 0.0001, RMSE = 0.18) was: d13 CDIC ¼ 1:15ð0:05Þ  d13 Caragonite  0:85ð0:04Þ 13

ð3Þ

13

The relationship between d CDIC and d Caragonite described for P. magellanicus results in larval shell aragonite values that are depleted on average by 1.82& (SD = 0.22&, range = 1.1–2.1&) from predicted equilibrium values based on the relationship calibrated for inorganic aragonite (Romanek et al., 1992). The Romanek et al. (1992) relationship is estimated as: d13 CDIC ¼ 1:0  d13 Caragonite  2:7ð0:6Þ

Fig. 2. (a) Carbon isotope (d13CDIC) composition versus salinity of treatment water; (b) oxygen isotope (d18Owater) composition versus salinity of treatment water. Open circles, 2005 data; closed diamonds, 2006 data. Mixing lines are linear least squares regressions for both carbon and oxygen isotopes.

d13 CDIC ¼ 0:6ð0:03Þ  S  19:5ð0:72Þ

ð1Þ 13

The average standard deviation of d CDIC measurements across all salinity treatments (n = 8 measurements for each salinity treatment) was 0.11&, with a range of 0.08–0.13&, which is only slightly higher than the analytical error of 0.1& and reflects the low variability of d13CDIC measurements over the time period sampled. The oxygen isotope (d18Owater) composition of experimental treatment water also was directly proportional to salinity. The linear relationship between d18Owater and salinity (r2 = 0.99, p < 0.0001, RMSE = 0.05; Fig. 2b) was described by: d18 Owater ¼ 0:20ð0:01Þ  S  7:8ð0:13Þ

ð4Þ

Our relationship calibrated for P. magellanicus larval shell aragonite is between 0.10 and 0.20 higher than the slope of the Romanek et al. (1992) relationship for inorganic aragonite. Due to the close clustering of d13Caragonite measurements at each d13CDIC treatment level, final shell size or growth rate did not appear to affect the carbon isotope composition of larval shell in our experiments, although significant effects of temperature and salinity, which is related to d13CDIC, on shell size and growth rate were observed. The depletion in P. magellanicus larval shell d13Caragonite may be due to the incorporation of metabolic carbon derived from the algal food source into the shell. The d13Cfood for the prepared Shellfish Diet 1800 was 35.2&. To calculate the contribution of metabolic carbon needed to establish the observed depletion from equilibrium, we used the isotopic mixing equation described by McConnaughey et al. (1997) and modified by Gillikin et al. (2006) to assume d13Cfood was equal to d13Ctissue: d13 Caragonite  ear-b ¼ Mðd13 Cfood Þ þ ð1  MÞ  d13 CDIC

ð5Þ

where M is the percent metabolic carbon contribution and ear-b is the enrichment factor between aragonite and bicarbonate (2.7 ± 0.6& in Romanek et al., 1992; 0.90 ± 0.03 in this study). From this equation, the average contribution of metabolic carbon to the larval shell aragonite of P. mag-

ð2Þ 18

The average standard deviation of d Owater measurements across both years and all salinity treatments sampled was 0.16&, with a range of 0.11–0.21&, which is approximately twice the analytical error of 0.07&. This suggests that some evaporation or mixing may have occurred in the treatment waters, but the effects are relatively minor and similar to the variability in d18Owater observed in Wanamaker et al. (2006). 3.3. Carbon isotope calibration The relationship between d13CDIC and d13Caragonite for P. magellanicus larvae was established using direct carbon

Fig. 3. Carbon isotope composition of treatment water (d13CDIC) versus larval shell (13Caragonite) in 2006. Heavy solid line, least squares regression for the relationship. Light solid line, predicted relationship based on Romanek et al. (1992) equilibrium model for synthetic aragonite.

4694

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

ellanicus in this study is 5.4% (SD = 0.57%; range = 3.4 to 7.8%). Although the calculated contribution of metabolic carbon into the larval shell is relatively small, it is important to note that it resulted in a substantial offset of d13Caragonite values from predicted equilibrium based on Romanek et al. (1992). 3.4. Oxygen isotope calibration The relationship between temperature and oxygen isotope fractionation (d18Oaragonited18Owater) for P. magellanicus larvae conformed to the expectations of the Grossman and Ku (1986) equilibrium model for mollusk aragonite, with a few notable exceptions. Data for experiments conducted in 2005 and 2006 were combined, and the calibrated relationship was established using direct oxygen isotope measurements of water and shell (Fig. 4). For this experiment, the linear relationship described by least squares regression between temperature and d18Oaragonited18Owater (r2 = 0.90, p < 0.0001, RMSE = 0.63) was: T ð CÞ ¼ 20:0ð0:4Þ  4:6ð0:3Þ  d18 Oaragonite  d18 Owater




ð6Þ

The 95% confidence intervals for the relationship for P. magellanicus larval aragonite encompass the predicted values from the equilibrium relationship described for mollusk aragonite by Grossman and Ku (1986):
 ð7Þ T ð CÞ ¼ 20:6  4:34 d18 Oaragonite  d18 Owater  0:2

librium for the two groups of shells represents a predicted temperature overestimate that ranged from 2.3 to 4.1 °C in 2005 to and 2.1 to 5.9 °C in 2006. The groups of shells that deviate from equilibrium include samples from larvae that both grew and did not grow during the experiment based on growth rate estimates (Fig. 5a), and samples from a range of shell sizes that were among the smallest shells measured within each experimental year (Fig. 5b). 4. DISCUSSION 4.1. Carbon and oxygen isotope mixing lines The d13CDIC and d18Owater values of the experimental treatment waters were tightly related to salinity. These two mixing lines allow both d13CDIC and d18Owater to be estimated accurately and predictably from the range of salinity measurements examined in this study. However, because the waters used in this study were collected from only one location and were archived for use during the entire experiment, the relationships between salinity and d13CDIC and d18Owater should be examined for samples collected at various locations and times in the Gulf of Maine before the relationships are used to assess the isotopic composition of field-collected water samples. In particular, reliable approximations of oxygen isotope fractionation in biogenic carbonates depend on obtaining a value for d18Owater in which the carbonate

Therefore, we conclude that P. magellanicus larval aragonite is precipitated in oxygen isotopic equilibrium for most samples. However, the predicted temperature estimate based on the equilibrium relationship exceeded the experimental temperature by more than 2 °C for two groups of shells (indicated by shaded blocks; Fig. 4); data for these shells were not included in the tight linear relationship established for the other samples. The deviation from equi-

Fig. 4. Temperature (°C) versus oxygen isotope fractionation (d18Oaragonite  d18Owater). Circles, 2005 data; squares, 2006 data. Heavy solid line, least squares regression for the relationship; dashed line, 95% confidence interval for this relationship. Light solid line, predicted relationship based on Grossman and Ku (1986) equilibrium model for biogenic aragonite. Shaded blocks indicate the groups of shells in 2005 and 2006 that deviated more than 2 °C from predicted equilibrium values based on the Grossman and Ku (1986) model.

Fig. 5. Temperature deviation (°C) based on Grossman and Ku (1986) equilibrium model for biogenic aragonite versus (a) shell growth rate (lm/day) and (b) shell size. Circles, 2005 data; squares, 2006 data. Dashed line indicates zero deviation from predicted equilibrium values. Shaded blocks indicate the groups of shells in 2005 and 2006 that deviated more than 2 °C from predicted equilibrium values.

Isotope fractionation in bivalve larval aragonite

was deposited, and the ability to predict d18Owater from independent, site-specific salinity measurements strengthens the estimate of d18Oaragonited18Owater, which can be used to reconstruct temperature. 4.2. Stable carbon isotope fractionation A number of explanations for depletion from carbon isotope equilibrium in biogenic carbonates have been proposed. The carbon isotope composition of foraminifera has been observed to decrease with an increase in seawater carbonate concentration (Spero et al., 1997; Zeebe, 1999). However, because the depletion in carbon isotope composition of sea scallop larval shell aragonite was observed across the range of experimental conditions, our results do not support this explanation. In addition, depletion from carbon isotopic equilibrium can arise due to kinetic effects that are the result of enzyme discrimination against heavier isotopes of carbon, as well as to the incorporation of isotopically depleted metabolic carbon into the shell material (McConnaughey, 1989a,b). Because the enzymes that discriminate against the heavier isotopes of carbon also discriminate against the heavier isotopes of oxygen, a kinetic isotope effect usually manifests as simultaneous depletions in both carbon and oxygen isotope composition (McConnaughey, 1989a,b). We did not observe a simultaneous depletion in carbon and oxygen isotope composition of P. magellanicus larval shell (r2 = 0.62, p < 0.0001, RMSE = 0.37; Fig. 6). Therefore, we conclude that the kinetic hypothesis is not supported, and that the depletion in carbon isotope composition of sea scallop larval shell aragonite is likely due to the incorporation of metabolic carbon. Based on the metabolic carbon equation of McConnaughey et al. (1997; Eq. (5)), the metabolic carbon contribution into P. magellanicus larval shell aragonite was <6%. Previous studies also have determined that the kinetic isotope effects on mollusk shells are small, and although higher levels of metabolic carbon contribution have been reported for biogenic carbonates (McConnaughey et al., 1997; Dettman et al., 1999; Lorrain et al., 2004; Geist et al., 2005; Gillikin et al., 2006), our observations were within the range commonly detected for adults of other bivalve species.

Fig. 6. Oxygen isotope composition (d18Oaragonite) versus carbon isotope composition (d13Caragonite) of larval shell in 2006. Heavy solid line, least squares regression for the relationship.

4695

The available literature on the interpretation of the relative contribution of environmental DIC and metabolic or respiratory DIC to skeletal carbonates remains controversial (e.g. Geist et al., 2005; Geist et al., 2006; Scho¨ne et al., 2006a). As we found for the aragonite larval shell of the sea scallop, P. magellanicus, skeletal carbonates are generally depleted in 13C compared with equilibrium conditions (McConnaughey et al., 1997; McConnaughey, 2003). This depletion from equilibrium conditions is often explained by a contribution of isotopically depleted metabolic carbon into the biogenic carbonate material (Lorrain et al., 2004; Gillikin et al., 2006), although the relative contribution of metabolic carbon is highly variable (e.g. Dettman et al., 1999). Some authors infer that the relative contribution of metabolic carbon is insignificant compared with the contribution of d13C from the environment (McConnaughey et al., 1997; McConnaughey, 2003). However, other authors conclude that the highly variable contribution of metabolic carbon across seasons, rates of skeletal growth, and ontogenetic changes in the organism precludes or confounds the use of carbon isotopes in biogenic carbonates as a proxy for environmental conditions, and suggest that carbon isotopes reflect the metabolic activity of the organism (Owen et al., 2002b; Lorrain et al., 2004; Geist et al., 2005; Gillikin et al., 2006). Growth rate of shell material may also influence carbon isotope composition. In a field study of adult shell calcite of the scallop Pecten maximus, Owen et al. (2002b) found that the stable carbon isotope composition of the shell was depleted up to 2& at low growth rates, but that the magnitude and variability of the carbon isotope depletion increased at higher growth rates. It should be noted that the larval growth rates at all treatment levels in our experiment were low (0.0–2.3 lm/day; Table 3) compared with growth rates observed for P. magellanicus larvae in other studies ( 4 lm/day) (Gallager et al., 1996; Manuel et al., 1996; Gouda et al., 2006); thus, our measurements may indicate stable carbon isotope depletion under ‘‘low” growth conditions. Based on the Owen et al. (2002b) study, it is possible that carbon isotope depletion in sea scallop larval aragonite may increase at higher experimental growth rates. Further examination of P. magellanicus or other bivalve species to investigate the influence of higher growth rates on the carbon isotope composition of larval and adult shells through controlled field or laboratory studies following methods outlined here and in other studies is needed (Owen et al., 2002a,b; Wanamaker et al., 2006). In this calibration experiment, P. magellanicus larval shell aragonite reared under the controlled environmental conditions of this study accurately reflected the d13CDIC of ambient seawater, with a constant offset for metabolic carbon input, within the range of environmental conditions, diet, and growth rates examined in this study. The offset from carbon isotope equilibrium was not dependent upon growth rate. However, since the carbon isotope composition of the food source, the contribution of metabolic carbon to the shell, and growth rate likely varies in field conditions, further experiments are needed to determine whether environmental information may be interpreted from analysis of the carbon isotope composition of the larval shells.

4696

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

4.3. Oxygen isotope fractionation The experimental results for stable oxygen isotope fractionation in the aragonite larval shell of the sea scallop, P. magellanicus, indicated that fractionation for most samples was not significantly different from predicted results based on the Grossman and Ku (1986) aragonite model generated for adult gastropod and scaphopod mollusks, in which the shell is considered to be deposited in equilibrium with the oxygen isotope composition of the ambient seawater (Fig. 3). Indeed, application of the Grossman and Ku (1986) model to predict larval rearing temperatures from our experiment resulted in estimates that were very close to the actual rearing temperatures, with an accuracy of 0.63 °C. In general, mollusks are known to deposit shell material near oxygen isotope equilibrium with ambient seawater (Epstein et al., 1953), with negligible kinetic and metabolic effects on the oxygen isotope composition of the shell (McConnaughey, 1989b; McConnaughey et al., 1997). However, our results showed that oxygen isotope fractionation in P. magellanicus larval aragonite was not consistent with expectations from the Grossman and Ku (1986) model for all samples. In our experiment, two groups of shells had oxygen isotope compositions that were depleted up to 2& with respect to equilibrium conditions, which is equivalent to a predicted temperature offset up to 5.9 °C based on the Grossman and Ku (1986) model. Depletions from oxygen isotope equilibrium are often attributed to a kinetic effect when they occur at higher rates of skeletal carbonate deposition (McConnaughey, 1989a,b; McConnaughey et al., 1997). For sea scallop larvae in our experiment, depletions occurred in low temperature and salinity treatments that represented a range of growth rates and shell sizes; low growth rate or small shell size alone did not result in depletion from equilibrium. Thus, the kinetic hypothesis of discrimination against heavier isotopes at higher rates of precipitation cannot explain our results. Similarly, the mechanism by which a metabolic effect could produce an observed depletion from oxygen isotope equilibrium is not known. The enzyme carbonic anhydrase acts to equilibrate respired CO2 with seawater before the CO2 is incorporated into skeletal material, but its effect on depletion is not well understood (McConnaughey, 1989a). In addition, periodic (approx. 1/week) measurements of seawater pH showed that the water was highly buffered, with a pH of 8.0 for all salinity treatments. The pattern of oxygen isotope depletion observed in our study is unique compared with studies on adult scallop (Pecten maximus) shell calcite (Owen et al., 2002b; Chauvaud et al., 2005), which either found oxygen isotope enrichment at low growth rates or no effect of growth rate at all. We suggest that for the aragonite shell of the sea scallop larvae in our experiment, depletions from oxygen isotope equilibrium may occur when larvae are reared in combinations of low temperature and salinity, although the mechanism for this depletion is not known, and that environmental conditions might not be recorded in the shell during ‘‘extreme” or ‘‘stressful” conditions. This is similar to results found by Scho¨ne et al. (2006b) for adult mollusks and barnacles at high temperatures in the Gulf of Mexico.

However, the majority of samples in our experiment did accurately reflect environmental conditions, which indicates that the larval shell may still be useful as an environmental proxy. Because larvae reared under stressful environmental conditions are unlikely to grow well or survive to metamorphosis (e.g. Hoffman et al., 2004), we recommend using larger field-collected larvae (>200 lm shell height) or newly-settled juveniles that have not yet deposited the juvenile shell to reconstruct temperature from oxygen isotope composition of the shell. Additional laboratory experiments are needed to confirm that the oxygen isotope composition of the shell of larger larvae or newly-settled juveniles consistently reflects environmental conditions. 4.4. Implications and applications to palaeoceanography and ecological studies 4.4.1. Palaeoceanography studies Calibrated laboratory or field experiments like this one and others (e.g. Owen et al., 2002a,b; Chauvaud et al., 2005; Wanamaker et al., 2006) have the potential to improve our ability to reconstruct past ocean environments and climate changes. Understanding recent changes in the ocean environment requires a number of tools to investigate past ocean conditions (e.g. Orr et al., 2005). The incorporation of mollusk larvae into palaeoenvironmental studies adds another tool for studies of palaeoclimate, corroborates environmental reconstructions based on other species, and supplements temporally or spatially patchy environmental information available from other species (Scho¨ne et al., 2006b). In addition, studies including planktonic larvae will complement existing studies of benthic mollusks. Interpretation of stable isotope composition of larvae has the potential to expand the resolution of environmental reconstruction by offering a record of pelagic water temperature on the timescale of days or weeks that larvae are in the water column, seasonal variations in water temperature recorded in the shells of larvae spawned at different times of year, and information on larval ecology by providing evidence for larvae that occur in habitats and conditions where adults not found (e.g. Hansen, 1984). Isotope data for larval shells could also be used to confirm reproductive types of bivalves inferred from morphology, following Killingley and Rex (1985). 4.4.2. Contemporary larval ecology studies Bivalves are extremely prevalent in marine systems, and larval shells may provide a particularly practical model to examine connectivity among marine habitats via the dispersive larval stages of sedentary marine invertebrates. Examination of the carbon isotope composition of the larval shell can reveal information about the nutrient conditions, upwelling regime, and salinity in which the larvae developed, which could be used to infer the location in which the larvae developed (bay vs. ocean, coastal vs. offshore). In addition, the oxygen isotope composition of larval shells, as a proxy for temperature, can be used to indicate larval transport within eddies or along fronts, as well as to discriminate between bay or coastal development of larvae. The geochemical information contained in the stable isotope and trace metal signatures in the shells of larval and

Isotope fractionation in bivalve larval aragonite

juvenile mollusks as well as adult fish, has been used to determine spawning grounds and stock structure (Gao et al., 2001, 2005), larval dispersal trajectories (Zacherl et al., 2003), and larval neighborhood sizes (Becker et al., 2005). However, this is the first study to experimentally validate stable isotope composition of the larval shell for a marine bivalve. Adding a calibrated stable isotope tool for environmental reconstruction will enhance the power of existing geochemical proxies available to study the dispersive larval phase of many benthic marine organisms. 5. CONCLUSIONS This is the first study to empirically calibrate the stable carbon and oxygen isotope fractionation in the larval shell aragonite of a marine bivalve, the sea scallop P. magellanicus, and the results reflect the potential utility as well as the limitations of using the sea scallop larval shell as an environmental proxy. Within the range of temperature, salinity, and growth parameters investigated in this study, the stable carbon and oxygen composition of bivalve larval shell aragonite of P. magellanicus reflects the temperature and d13CDIC in which the shell was formed. However, additional studies are necessary to determine whether the carbon and oxygen isotope relationships defined in a controlled experimental setting can be applied to field studies. In particular, the contribution of metabolic carbon to the larval shell during different seasons and food regimes needs further investigation before d13Caragonite of field-collected samples can be used as a reliable proxy for environmental conditions. In addition, further studies are needed to determine the mechanism for oxygen isotope depletion in larvae reared under stressful conditions, and to assess whether the oxygen isotope composition of larger larvae or newly-settled juveniles is a consistent environmental proxy. ACKNOWLEDGEMENT We gratefully acknowledge the people at various institutions who assisted with isotope analyses. All water and shell analyses performed in 2005 were conducted at the University of Maine Stable Isotope Laboratory. In 2006, oxygen isotopes in water were measured at the University of Maine Stable Isotope Laboratory, carbon isotopes in DIC were measured at the Hatch Stable Isotope Laboratory, and carbon and oxygen isotopes in shell were measured at the University of Mainz. M. Yates conducted the X-ray diffraction analysis on larval shells. V. Moreau assisted with preparations for larval culturing in 2006. Commercial divers P. Rosen and T. Robinson collected the sea scallops used in this study. We thank D. Dettman, C. DiBacco, E. Grossman, W. Halteman, and D. McCorkle for their insightful and constructive comments that substantially improved this manuscript. Aquaculture facilities were provided by the University of Maine Stable Isotope Laboratory through a National Science Foundation Grant (NSF ATM0222553). This research was supported by the Maine Sea Grant Program under Award R-04-03 to P.D.R.

REFERENCES Barber B. J., Getchell R., Shumway S. and Schick D. (1988) Reduced fecundity in a deep-water population of the giant

4697

scallop Placopecten magellanicus in the Gulf of Maine. USA. Mar. Ecol. Prog. Ser. 42, 207–212. Becker B. J., Fodrie F. J., McMillan P. A. and Levin L. A. (2005) Spatial and temporal variation in trace elemental fingerprints of mytilid mussel shells: A precursor to invertebrate larval tracking. Limnol. Oceanogr. 50, 48–61. Beveridge N. A. S. and Shackelton N. J. (1994) Carbon isotopes in recent planktonic foraminifera—a record of anthropogenic CO2 invasion of the surface ocean. Earth Planet. Sci. Lett. 126, 259–273. Chauvaud L., Lorrain A., Dunbar R. B., Paulet Y.-M., Thouzeau G., Jean F., Guarini J.-M. and Mucciarone D. (2005) Shell of the Great Scallop Pecten maximus as a high-frequency archive of palaeoenvironmental changes. Geochem., Geophys., Geosyst. 6, Q08001. Craig H. (1957) Isotopic standards for carbon and oxygen and correction factors for mass spectrometric analysis of carbon dioxide. Geochim. Cosmochim. Acta 12, 133–149. Culliney J. L. (1974) Larval development of the giant scallop Placopecten magellanicus (Gmelin). Biol. Bull. 147, 321–332. Dettman D. L., Reische A. K. and Lohmann K. C. (1999) Controls on the stable isotope composition of seasonal growth bands in aragonitic fresh-water bivalves (Unionidae). Geochim. Cosmochim. Acta 63, 1049–1057. DiBacco C., Robert G. and Grant J. (1995) Reproductive cycle of the sea scallop, Placopecten magellanicus (Gmelin, 1791), on northeastern Georges Bank. J. Shellfish Res. 14, 59–69. DuPaul W. D., Kirkley J. E. and Schmitzer A. C. (1989) Evidence of a semiannual reproductive cycle for the sea scallop, Placopecten magellanicus (Gmelin, 1791) in the mid-Atlantic region. J. Shellfish. Res. 8, 173–178. Duplessy J. C., Shackelton N. J., Fairbanks R. G., Labeyrie L., Oppo D. and Kallel N. (1988) Deepwater source variations during the last climatic cycle and their impact on the global deepwater circulation. Paleoceanography 3, 343–360. Epstein S., Buchsbaum R., Lowenstam H. A. and Urey H. C. (1953) Revised carbonate–water temperature scale. Bull. Geol. Soc. Am. 64, 1315–1326. Gallager S. M., Manuel J. L., Manning D. A. and O’Dor R. (1996) Ontogenetic changes in the vertical distribution of giant scallop larvae, Placopecten magellanicus, in 9-m deep mesocosms as a function of light, food, and temperature stratification. Mar. Biol. 124, 679–692. Gao Y. W., Joner S. H. and Bargmann G. G. (2001) Stable isotopic composition of otoliths in identification of spawning stocks of Pacific herring (Clupea pallasi) in Puget Sound. Can. J. Fish. Aquat. Sci. 58, 2113–2120. Gao Y. W., Bargmann G. G., Brand U. and Noakes D. L. G. (2005) Stable isotopic and trace metal compositions of otoliths and the stock structure of Pacific cod, Gadus macrocephalus. Environ. Biol. Fishes 74, 335–348. Geist J., Auerswald K. and Boom A. (2005) Stable carbon isotopes in freshwater mussel shells: Environmental record or marker for metabolic activity? Geochim. Cosmochim. Acta 69, 3545–3554. Geist J., Auerswald K. and Boom A. (2006) Response to the comment by B. Scho¨ne et al. (2006) on ‘‘Stable carbon isotopes in freshwater mussel shells: Environmental record or marker for metabolic activity?”. Geochim. Cosmochim. Acta 70, 2662–2664. Gillikin D. P., Lorrain A., Boullion S., Willenz P. and Dehairs F. (2006) Stable carbon isotopic composition of Mytilus edulis shells: relation to metabolism, salinity, d13C, and phytoplankton. Org. Geochem. 37, 1371–1382. Gouda R., Kenchington E., Hatcher B. and Vercaemer B. (2006) Effects of locally-isolated micro-phytoplankton diets on growth and survival of sea scallop (Placopecten magellanicus) larvae. Aquaculture 259, 169–180.

4698

E.F. Owen et al. / Geochimica et Cosmochimica Acta 72 (2008) 4687–4698

Grossman E. L. and Ku T.-L. (1986) Oxygen and carbon isotope fractionation in biogenic aragonite: temperature effects. Chem. Geol. 59, 59–74. Hansen T. A. (1981) Fossil molluscan larvae: A new biostratigraphic tool. Science 214, 915–916. Hansen T. A. (1984) The biostratigraphic use of larval and juvenile mollusks. Geol. Soc. Am. Bull. 95, 1102–1107. Hart D. R. and Chute A. S. (2004) Essential Fish Habitat Source Document: Sea Scallop, Placopecten magellanicus, Life History and Habitat Characteristics, 2nd ed.. NOAA Tech. Mem. NMFS-NE-189, 21 pp. Hoffman E. E., Powell E. N., Bochenek E. A. and Klinck J. A. (2004) A modeling study of the influence of environment and food supply on survival of Crassostrea gigas larvae. ICES J. Mar. Sci. 61, 596–616. Kalish J. M. (1991) Oxygen and carbon stable isotopes in the otoliths of wild and laboratory-reared Australian salmon (Arripis trutta). Mar. Biol. 110, 37–47. Killingley J. S. and Rex M. A. (1985) Mode of larval development in some deep-sea gastropods indicated by oxygen-18 values of their carbonate shells. Deep-Sea Res. 32, 809–818. Krantz D. E., Jones D. S. and Williams D. F. (1984) Growth rates of the sea scallop, Placopecten magellanicus, determined from the 18O/16O record in shell calcite. Biol. Bull. 167, 186–199. Levin L. A. (2006) Recent progress in understanding larval dispersal: new directions and digressions. Integr. Comp. Biol. 46, 282–297. Lorrain A., Paulet Y.-M., Chauvaud L., Dunbar R., Mucciarone D. and Fontugne M. (2004) d13C variation in scallop shells: Increasing metabolic carbon contribution with body size? Geochim. Cosmochim. Acta 68, 3509–3519. Lutz R. A. and Jablonski D. (1978) Larval bivalve shell morphometry: a new palaeoclimatic tool? Science 202, 51–53. Lynch-Steiglitz J. and Fairbanks R. G. (1994) A conservative tracer for glacial ocean circulation from carbon-isotope and palaeonutrient measurements in benthic foraminifera. Nature 369, 308–310. Malchus N. and Steuber T. (2002) Stable isotope records (O,C) of Jurassic aragonitic shells from England and NW Poland: palaeoecologic and environmental implications. Geobios 35, 29–39. Manuel J. L., Gallager S. M., Pearce C. M., Manning D. A. and O’Dor R. K. (1996) Veligers from different populations of sea scallop Placopecten magellanicus have different vertical migration patterns. Mar. Ecol. Prog. Ser. 142, 147–163. McConnaughey T. (1989a) 13C and 18O isotopic disequilibrium in biological carbonates. I: Patterns. Geochim. Cosmochim. Acta 53, 151–162. McConnaughey T. (1989b) 13C and 18O isotopic disequilibrium in biological carbonates. II: In vitro simulation of kinetic isotope effects. Geochim. Cosmochim. Acta 53, 163–171. McConnaughey T. A., Burdett J., Whelan J. F. and Paull C. K. (1997) Carbon isotopes in biological carbonates: respiration and photosynthesis. Geochim. Cosmoschim. Acta 61, 611–622. McConnaughey T. A. (2003) Sub-equilibrium oxygen-18 and carbon-13 levels in biological carbonates: carbonate and kinetic models. Coral Reefs 22, 316–327. Naidu K. S. (1991) Sea scallop, Placopecten magellanicus. In Scallops: Biology, Ecology, and Aquaculture (ed. S. E. Shumway), Developments in Aquaculture and Fisheries Science, vol. 21. Elsevier Science Publishing, Amsterdam, pp. 861–897. Owen R. J., Kennedy H. A. and Richardson C. A. (2002a) Experimental investigation into partitioning of stable isotopes

between scallop (Pecten maximus) shell calcite and seawater. Palaeogeogr. Palaeoclimatol. Palaeoecol. 185, 163–174. Owen R., Kennedy H. and Richardson C. (2002b) Isotopic partitioning between shell calcite and seawater: effect of shell growth rate. Geochim. Cosmochim. Acta 66, 1727–1737. Orr J. C., Fabry V. J., Aumont O., Bopp L., Doney S. C., Feely R. A., Gnanadesikan A., Gruber N., Ishida A., Joos F., Key R. M., Lindsay K., Maier-Reimer E., Matear R., Monfray P., Mouchet A., Najjar R. G., Plattner G.-K., Rodgers K. B., Sabine C. L., Sarmiento J. L., Schlitzer R., Slater R. D., Totterdell I. J., Weirig M.-F., Yamanaka Y. and Yool A. (2005) Anthropogenic ocean acidification over the twenty-first century and its impact on calcifying organisms. Nature 437, 681–686. Posgay J. A. (1957) The range of the sea scallop. The Nautilus 71, 55–57. Romanek C. S., Grossman E. L. and Morse J. W. (1992) Carbon isotopic fractionation in synthetic aragonite and calcite: effects of temperature and precipitation rate. Geochim. Cosmochim. Acta 56, 419–430. Scho¨ne B. R., Rodland D. L., Surge D. M., Fiebig J., Gillikin D. P., Baier S. M. and Goewert A. (2006a) Comment on ‘‘Stable carbon isotopes in freshwater mussel shells: Environmental record or marker for metabolic activity?” by J. Geist et al. (2005). Geochim. Cosmochim. Acta 70, 2658–2661. Scho¨ne B. R., Rodland D. L., Fiebig J., Oschmann W., Goodwin D., Flessa K. W. and Dettman D. (2006b) Reliability of multitaxon, multiproxy reconstructions of environmental conditions from accretionary biogenic skeletons. J. Geol. 114, 267– 285. Spero H. J., Bijma J., Lea D. W. and Bemis B. E. (1997) Effect of seawater carbonate concentration of foraminiferal carbon and oxygen isotopes. Nature 390, 497–500. St-Jean G. (2003) Automated quantitative and isotopic (13C) analysis of dissolved inorganic carbon and dissolved organic carbon in continuous-flow using a total organic carbon analyzer. Rapid Commun. Mass Spectrom. 17, 418–428. Thorrold S. R., Campana S. E., Jones C. M. and Swart P. K. (1997) Factors determining d13C and d18O fractionation in aragonite otoliths of marine fish. Geochim. Cosmochim. Acta 61, 2909–2919. Tremblay M. J., Meade L. D. and Hurley G. V. (1987) Identification of planktonic sea scallop larvae (Placopecten magellanicus) (Gmelin). Can J. Fish. Aquat. Sci. 44, 1361–1366. Wanamaker, Jr., A. D., Kreutz K. J., Borns, Jr., H. W., Introne D. S., Feindel S. and Barber B. J. (2006) An aquaculture-based method for calibrated bivalve isotope palaeothermometry. Geochem. Geophys. Geosyst. 7, Q09011. Weiss I. M., Tuross N., Addadi L. and Weiner S. (2002) Mollusk larval shell formation: amorphous calcium carbonate is a precursor phase for aragonite. J. Exp. Zool. 293, 478–491. Woodruff F. and Savin S. M. (1985) d13C values of Miocene Pacific benthic foraminifera: correlations with sea level and biological productivity. Geology 13, 119–122. Zacherl D. C., Manrı´quez P. H., Paradis G., Day R. W., Castilla J. C., Warner R. R., Lea D. W. and Gaines S. D. (2003) Elemental fingerprinting of gastropod statoliths to study larval dispersal trajectories. Mar. Ecol. Prog. Ser. 248, 297–303. Zeebe R. E. (1999) An explanation of the effect of seawater carbonate concentration on foraminiferal oxygen isotopes. Geochim. Cosmochim. Acta 63, 2001–2007. Associate editor: Anne Cohen