Meseck et al.: Effects of ocean acidification on larval Spisula solidissima from Long Island Sound 
Table 1 
Carbonate chemistry and environmental characteristics measured at low, 
medium, and high levels of partial pressure of carbon dioxide (pCO,) and 
used to investigate the effect of pCO, levels on laboratory-reared larval 
Atlantic surflclam (Spisula solidissima) in 2015. The variables include mean 
measured pH (seawater scale, pH measured at 20°C), pH in situ (seawater 
scale, pH at the temperature of the experiment), temperature (°C), salinity, 
pCO, (natm), total dissolved inorganic carbon (DIC) (umol/kg), and satura- 
tion levels of calcite (Q.icite) and aragonite (Q.,agonite) for the duration of the 
experiment. Values are means with standard errors of the mean in paren- 
theses. Asterisks (*) indicate values that were measured directly and not 
calculated by using the program CO2SYS. The adult brood stock used to rear 
individuals for this study were collected in 2014 from Long Island Sound 
near Stratford and Norwalk, Connecticut. 
Treatment level of pCO, 
Characteristic Low 
7.972 (0.003) 
8.073 (0.003) 
18.8 (0.1) 
25.72 (0.08) 
pH measured’ 
pH in situ 
Temperature 
Salinity 
pCO, 343.9 (5.9) 
DIC 1800.66 (2.35) 
Q 3.46 (0.01) 
2.18 (0.01) 
calcite 
‘aragonite 
was to begin. Samples were shipped overnight to the 
Memorial University of Newfoundland for lipid analysis, 
where they were extracted by following a modified Folch 
procedure (Parrish, 1999), and lipid classes were identified 
by using a Chromarod-Iatroscan system for thin-layer 
chromatography and flame ionization detection (Mark V, 
Iatron Laboratories, Tokyo, Japan). 
Carbonate chemistry measurements 
Seawater samples were collected from each bucket for car- 
bonate chemistry measurements immediately before and 
after water changes (n=63). One sample per bucket was 
obtained by placing a tube fit with an airstone into the 
bottom of each bucket and filling 500-mL polypropylene 
collection bottles to overflow (1.5 times) by using a peri- 
staltic pump. The airstone ensured that larvae and phy- 
toplankton were not collected with the seawater samples. 
Samples were analyzed immediately for pH and DIC. 
Total pH was determined colorimetrically by using 
m-Cresol purple indicator dye (Sigma-Aldrich, St. Louis, 
MO) (Dickson and Goyet®) with an Ocean Optics 2000+ 
UV spectrometer (Ocean Insight, Orlando, FL). All 
3 Dickson, A. G., and C. Goyet (eds.). 1994. Handbook of methods 
for the analysis of the various parameters of the carbon diox- 
ide system in sea water, version 2. U.S. Dep. Energy, Oak Ridge 
Natl. Lab., Carbon Dioxide Inf. Anal. Cent. ORNL/CDIAC-74, 
180 p. [Available from website.] 
1903.74 (2.39) 
Medium High 
7.629 (0.003) 
7.734 (0.003) 
18.9 (0.1) 
25.73 (0.08) 
820.8 (6.0) 
7.462 (0.003) 
7.568 (0.003) 
18.9 (0.1) 
25.73 (0.08) 
1242.9 (5.9) 
1953.32 (2.37) 
1.22 (0.02) 
0.77 (0.01) 
1.74 (0.02) 
1.09 (0.01) 
samples were analyzed for DIC without the addition of 
preservatives on a DIC analyzer (AS-C3, Apollo SciTech, 
LLC, Newark, DE). To ensure precision, replicate sam- 
ples (n=5) of certified reference material also were ana- 
lyzed, resulting in 1 standard deviation of 2.7 pmol/kg for 
total DIC replicates and of 0.0014 for total pH. The DIC 
instrument was part of a previous international, inter- 
laboratory exercise for comparison of DIC in test seawa- 
ter samples with low and high CO, levels; DIC during 
our study measured within 0.5% of assigned values 
(Bockmon and Dickson, 2015). Seawater temperature and 
salinity were measured daily by using a YSI Model 85 
handheld oxygen, conductivity, salinity, and temperature 
system (YSI, Inc., Yellow Springs, OH). Dissolved inor- 
ganic carbon and pH values were entered into the pro- 
gram CO2SYS (Pierrot et al., 2006) to calculate pCO, (in 
microatmospheres), calcite saturation level (Qyajcite), and 
Q.ragonite by using the following constants: K’,, K’, from 
Mehrbach et al. (1973) and refit by Dickson and Millero 
(1987), K hydrogen sulfate from Dickson (1990), and total 
boron from Uppstrém (1974). 
Statistical analyses 
The software Statgraphics Plus (vers 17.1.12, Statgraph- 
ics Technologies, The Plains, VA) was used for all statis- 
tical analyses. All data were checked for normality by 
using the Shapiro—Wilk test, with a significance level of 
0.05, and for equal variance prior to statistical analysis. 
