Article pubs.acs.org/est
Predicting the Impacts of CO2 Leakage from Subseabed Storage: Effects of Metal Accumulation and Toxicity on the Model Benthic Organism Ruditapes philippinarum Araceli Rodríguez-Romero,*,† Natalia Jiménez-Tenorio,‡ M. Dolores Basallote,‡ Manoela R. De Orte,‡,§ Julián Blasco,† and Inmaculada Riba‡ †
Departamento de Ecología y Gestión Costera, Instituto de Ciencias Marinas de Andalucía (CSIC), Campus Río San Pedro, Puerto Real, Cádiz 11510, Spain ‡ Cátedra UNESCO/UNITWIN/WiCop, Facultad de Ciencias del Mar y Ambientales, Universidad de Cádiz, Polígono Río San Pedro s/n, Puerto Real, Cádiz 11510, Spain S Supporting Information *
ABSTRACT: The urgent need to minimize the potential harm deriving from global climate change and ocean acidification has led governmental decision-makers and scientists to explore and study new strategies for reducing the levels of anthropogenic CO2. One of the mitigation measures proposed for reducing the concentration of atmospheric CO2 is the capture and storage of this gas in subseabed geological formations; this proposal is generating considerable international interest. The main risk associated with this option is the leakage of retained CO2, which could cause serious environmental perturbations, particularly acidification, in marine ecosystems. The study reported is aimed at quantifying the effects of acidification derived from CO2 leakage on marine organisms. To this end, a labscale experiment involving direct release of CO2 through marine sediment was conducted using Ruditapes philippinarum as a model benthic organism. For 10 days bivalves were exposed to 3 sediment samples with different physicochemical characteristics and at pre-established pH conditions (8.0−6.1). End points measured were: survival, burrowing activity, histopathological lesions, and metal accumulation (Fe, Al, Mn, Cu, and Zn) in whole body. Correlations analyses indicated highly significant associations (P < 0.01) between pH and the biological effects measured in R philippinarum, except for metal concentrations in tissues. Further research to understand and predict the biological and economic implications for coastal ecosystems deriving from acidification by CO2 leakages is urgently needed.
■
serious detrimental effects on the marine ecosystem.5 In this context, CO2 storage in subseabed geological formations can only be considered feasible if the environmental consequences derived from this mitigation option are significantly lower than those from the increase of CO2 in the atmosphere, which this storage is intended to prevent (e.g., OA).6,7 Previous studies have demonstrated that seawater acidification caused by an increase of CO2 levels in the marine environment may produce the release of metals from the surface of sediment to the seawater column,8 thus increasing the toxicity of sediment contaminated by metals.9,10 Modeling studies11,5 and in situ investigations of submarine eruption events12 as a “natural analogue” of CO2 leakage from marine subseabed storage sites13 suggest that these leakages will provoke local reductions in seawater pH. These alterations of
INTRODUCTION Global Climate Change (GCC) is considered an indisputable reality1 that presents an unprecedented threat to global marine biodiversity. The oceans act as a buffer against climate change by absorbing CO2 from the atmosphere; this results in a reduction of ocean pH that, in consequence, leads to ocean acidification (OA). OA presents a threat of unprecedented scale to the marine environment, with the potential to cause great adverse impacts on the numerous benefits that we obtain from the oceans.2 Currently, the capture, injection, and storage of CO2 in subseabed geological formations is proposed as one of the potential strategies for reducing the concentrations of atmospheric CO2 in order to avoid the abrupt and irreversible repercussions of GCC and OA.3 Nevertheless, practical implementation of this technological option could have significant impacts on marine ecosystems, due to possible CO2 leakages deriving from the capture facilities, transport pipelines, and offshore installations. The CO2 released would react with seawater and provoke significant changes in ocean pH.4 The extreme acidification thus caused would produce © 2014 American Chemical Society
Received: Revised: Accepted: Published: 12292
April 22, 2014 July 24, 2014 September 15, 2014 September 15, 2014 dx.doi.org/10.1021/es501939c | Environ. Sci. Technol. 2014, 48, 12292−12301
Environmental Science & Technology
■
seawater pH would be more severe in close proximity to the location of the CO2 leakage, increasing the pH values in the surrounding area. Clearly, the new pH conditions and the expansion of the plume of acidified seawater caused by such an event will depend on many factors such as the duration and rate of the leakage, tidal cycle, wind strength, regional circulation, and currents.11,13 Given the existence of cases where the anthropogenic CO2 would be captured from coal-fired power plants and industrial plants located on or near the coast,14,15 the CO2 could be transported through pipelines to the coast and then to an offshore storage site. The implication of this is that any leakage from the pipe during the transport and injection of the CO2 could severely damage coastal areas. Furthermore, many of the reports prepared on CO2 storage capacity consider deep saline aquifers situated near the coast as potential sites for the CO2 storage.15−17 Coastal environments are subjected to a high spatial and temporal heterogeneity in abiotic conditions and to a multitude of anthropogenic pressures, make them especially vulnerable to severe acidification events.18,19 Considering that acidification may have synergistic impacts with other anthropogenic factors,9,20 it is essential to evaluate how a reduction of pH resulting from a possible CO2 leakage during the transport, injection, and storage process in a subseabed site could negatively affect this environment.18,19 Coastal areas support a rich biodiversity, including species with important commercial value (fish, bivalves, crustaceans, etc.).18,21−23 Given all these issues, it is extremely urgent to predict the potential impacts of CO2 leakages deriving from OA and the implementation of this GCC mitigation measure, on species whose habitats are the shallower coastal areas of the ocean. Bivalves are considered suitable species for biomonitoring to evaluate sediment metal pollution24 and have been used previously for sediment toxicity assessment.25−29 Bivalves are capable of accumulating metal in their tissues over time, and this property provides useful information about the bioavailable metals in the environment.30 There are several factors that make bivalves particularly suitable for selection as biomonitoring species: they have a wide geographical distribution; most bivalves have adopted a sedentary life style (so they are reliable indicators of local environmental conditions); they are easy to keep in laboratory conditions; and they are available all yearround.24,31 In addition, their ecological and commercial importance in coastal ecosystems13 makes bivalves a good potential target species for assessing the impacts of acidification. In this study, we investigate the possible pH effects on sediment toxicity, over a short period of time, as a consequence of acidification, caused by CO2 leakage during its transport, injection and storage in subseabed formations, using the model organism Ruditapes philippinarum. For this purpose, we selected three sediment sites with different physicochemical characteristics, and studied the influence of pH, in an environmental range of between 8.0 and 6.0, on the accumulation of metals by R. philippinarum and resulting toxicity, after 10 days of exposure to each selected sediment. The influence of pH was assessed by comparing natural seawater pH in coastal areas (pH 8.1−7.7) with reduced pH values (pH 7−6) predicted for marine ecosystems in possible scenarios of leakages from CO2 storage sites32−34 and observed pH levels monitored during submarine eruption events.12
Article
MATERIALS AND METHODS
Sediment and Fauna Collection. Sediments samples were collected in two different areas in the Southwest of the Iberian Peninsula: the Cadiz Bay Natural Park and the Riá de Huelva estuary. “Rió San Pedro” (RSP; 36°31′52.90″ N, 6°12′48.43″ W) sediment was sampled in the Cadiz Bay Natural Park and was the source of reference sediment. “Mazagón″ (MZ; 37°08′026’’ N 6°50′215’’ W) and “Muelle Levante” (ML; 37°15′1″ N, 6°57′782″ W) sediments were collected at two sites in the Riá de Huelva, which are characterized by high metals content35,36 (see Supporting Information (SI) Figure S1). The sampling sites were selected to cover a wide range of 10 ́ physicochemical sediment properties. See Rodriguez-Romero for further information about the study areas and field sample collecting, processing, transport, and storage of sediments. Individuals of Ruditapes philippinarum (shell length 3.92 ± 0.16 cm, shell height 2.74 ± 0.16 cm; mean ± SD) were collected by hand during low tide in a clean area of Cadiz Bay and were immediately transported to the laboratory where they were acclimated to temperature and photoperiod, and kept in aerated and filtered clean seawater (pHNBS= 7.99) for 1 week. During this period, clams were fed a mixture of marine micro algae (50 mL per 30 L tank) containing Tetraselmis chuii, Isocrhysis galbana, and Chaetoceros gracilis at a concentration of 1.5 to 2 × 106 cells mL−1, 15 to 20 × 106 cells mL−1, and 5 × 106 cells mL−1, respectively (500 mL glass jar) at days 1, 3, and 7. CO2 Injection System. To assess the possible acidification effects in a relatively short period of time derived from CO2 leakages in subseabed formations, two CO2 injection systems were used for pH manipulation. The first CO2 manipulation system, described in Basallote,33 is adapted for the seawater exposure route and was used prior to the experimental exposure, for the period of acclimation of clams to the selected pH treatments. Individuals of R. philippinarum were acclimated to the selected pH treatments by gradually reducing seawater pH (by 0.5 units/day) until the pH value set for the treatment was reached.33,10 Hence, rapid changes in pH conditions that may directly provoke the death of organisms are avoided; and the acidification effects on burrowing activity can be assessed during the experimental period. The second CO2 injection system is adapted for the whole sediment exposure route and was employed for the experimental period. This system is 10 ́ described in Rodriguez-Romero and is characterized by the injection of CO2 through the sediment. Both pH manipulation systems incorporated four different pH treatments: pH 8 (control treatment, without injection of CO2) and three CO2 treatments at nominal pH values of 7.1, 6.6, and 6.1. The pH was recorded and controlled continuously over the course of the acclimation and experimental periods in each chamber. The actual pH values in seawater monitored during the period of exposure to RSP, ML, and MZ sediments in the different pH treatments are shown in SI Table S1. Treatments were performed in duplicate. All pH values recorded are on the NBS scale. Experimental Exposure. Three independent assays (one per sediment sampling site) using the model organism R. philippinarum were performed to investigate effects of CO2 leakages from subseabed storage sites on the various physicochemical properties of the sediments sampled. In each assay, sediment samples were placed in the aquariums of the CO2 injection system, which were then filled with filtered 12293
dx.doi.org/10.1021/es501939c | Environ. Sci. Technol. 2014, 48, 12292−12301
Environmental Science & Technology
Article
biomarkers of effect as a result of CO2 acidification in combination with the possible release of metals from sediment caused by the gas leakage. Clams were anaesthetized with 0.1% 2-phenoxyethanol 99% during 5−10 min. Gills and digestive gland tissues of the clams were fixed for 48 h in phosphate buffered 10% formaldehyde (pH 7.2). After dehydration in graded concentrations of ethanol, the samples were embedded in paraffin wax. Cross sections of 6 to 8 μm thickness were stained with Haematoxylin-Eosin and observed under microscope. Histopathological alterations were semiquantitatively evaluated in the specimens exposed to the different pH treatments based on the frequency of appearance of lesion in a total of four individuals from each treatment.,27,43,44 A general index of histological lesion in gills (GIG) and digestive gland (GIDG) was calculated for individuals exposed to each sediment sample at the different pH treatments after 1, 5, and 10 days of exposure. Statistical Analyses. Data obtained in respect of burrowing activity and survival tests and histopathological and metal analysis were expressed as means ± standard deviations, and were calculated using Microsoft Office Excel 2007. Data were tested for normality and homogeneity by using the Shapiro− Wilk and Levene tests, respectively. One-way analysis of variance (ANOVA) with Tukey’s posthoc test (with Bonferroni correction) was performed to test for significant differences (α = 0.05) of pH treatments in each sediment on survival, burrowing activity and metal concentrations in tissues and of physicochemical characteristics in sediment samples (see SI Tables S4−S11). When normality and homogeneity assumptions to satisfy ANOVA requirements were not met, data were logarithm-transformed (ln + 1). Pearson’s correlation was used to determine significant relationships (α = 0.05) between the data obtained, the physicochemical properties of the sediment and the pH levels in each sediment sample. Although two replicates per pH treatment gives the test design low statistical power, the very small or even non- existent variability presented between duplicates for survival rate and burrowing activity ensure enough power to detect biologically significant differences caused by pH treatments. Statistical analyses were performed with the aid of the statistical software packages SPSS 16 and GraphPad InStat version 3.00 for Windows.
natural seawater (v/v; 1:3). Once the sediment had settled, the CO2 was injected until the selected seawater pH value was reached. Subsequently, 30 acclimated clams were transferred to the corresponding pH test chamber of the CO2 injection system and were exposed during 10 days to each selected pH treatment. Water was continuously aerated and renewed every 5 days of the experiment, except for those aquariums in which mortality was detected, where water was replaced on the same day that the mortality was detected.37 The end points selected were burrowing activity after 48 h of exposure and histopathological and bioaccumulation analysis and survival (%) of R. philippinarum after 10 days of exposure. In each test chamber, the number of organisms that had burrowed completely into the sediment was counted at 15, 30, and 45 min, and 1, 1.5, 3, 6, 12, 24, and 48 h after the start of each pH test.23,25 Mortality was recorded daily until the end of the experiment. From each pH assay six individuals were sampled on days 1, 5, and 10 for histopathological and bioaccumulation analysis. Sediment samples were taken on days 0 and 10 for metal characterization, while water aliquots were sampled on days 1 and 5 from each pH treatment, for the determination of carbon system speciation (see SI Table S2). On day 5 of the exposure to each selected pH (a few hours before seawater renewal) clams were fed with a mixture of micro algae (T. chuii, I. galvana, and C. gracilis). During both the acclimation and the experimental periods, salinity (34.61 ± 1.89) was measured; temperature (17.52 ± 0.91 °C), dissolved oxygen (6.76 ± 0.56 mg L−1) and pH were controlled; and a natural photoperiod (16:8 h) was imposed. All values reported are means ± SD. The pH in sediment from the different pH treatments was monitored at days 1, 5, and 10 (see SI Table S3). Analytical Methods. Further details concerning quality assurance and analytical procedures have been reported 10 ́ previously in Rodriguez-Romero. Briefly, sediment samples were stored at −20 °C until the physical and chemical characterization was carried out. Grain size distribution followed UNE 103101:1995.38 Organic carbon (OC) content was measured in freeze-dried sediment according to Gaudette,39 modified by El-Rayis.40 The measurements were made using an automatic titrator (Metrohn. 888 Tritando). Organic matter (OM) content was estimated by loss of ignition (LOI) at 550 °C and gravimetric determination. The results obtained for grain size distribution, OC and OM are expressed as percentages. Metal concentrations (aluminum, iron, manganese, copper, and zinc) in freeze-dried total fraction of sediment samples were determined by total digestion according to the procedure of Loring and Rantala.41 Individuals of R. philippinarum sampled during the assay were placed in chambers with aerated and clean natural seawater and were depurated during 24 h to eliminate any sediment particles that may have been present in their gut. Later, they were stored at −80C prior to metals analysis. Once the experiments were completed, concentrations of aluminum, iron, manganese, copper, and zinc were determined following the procedure described by Amiard.42 Metal concentrations were measured in triplicate by optical emission spectrometry with inductioncoupled plasma (ICP-OES. Optima 2000 DV). The results are expressed as mg kg −1 dry weight. Histological Procedures. Individuals of the clam R. philippinarum were analyzed to determine histopathological damage in two different target tissues. These histological alterations were observed and examined over time as
■
RESULTS AND DISCUSSION Sediment Characterization. The results of the analyses of metals (Al, Fe, Mn, Cu, and Zn) and other sediment parameters (grain size distribution, organic carbon, and organic matter content) are shown in Table 1. The sediment samples presented variable characteristics between and within areas. In general, metal concentrations were significantly lower (P < 0.05) in the RSP sample, except for Mn concentrations. The highest values of total organic carbon and organic matter content were recorded in the sediment samples ML and MZ from the Riá de Huelva area. The three sediment samples were considered mud, the proportion of fines being higher (P < 0.05) in MZ and ML sediment samples. Previous studies have shown the importance of sediment geochemical properties (percentage of fines, OM, OC, etc.) for metal bioavailability and toxicity.45−48 Moreover, a more recent study has indicated that the physicochemical characteristics of the sediment are key factors in the indirect toxicity effects associated with a reduction in pH.10 Therefore, it is extremely important to investigate the possible effects of acidification on the toxicity and metals availability of the sediment. 12294
dx.doi.org/10.1021/es501939c | Environ. Sci. Technol. 2014, 48, 12292−12301
Environmental Science & Technology
Article
Survival and Burrowing Activity. The survival and burrowing activity of R. philippinarum decreased linearly with a reduction in seawater pH. At pH 6.1, after exposure for 10 days to sediment from the three sites, 100% mortality was observed. A similar survival rate (>95%) was observed among sediment samples in the control treatment, fulfilling the QA/ QC requirements (survival >90%).49,50 With RSP sediment, no significant differences in survival were recorded between the control and treatments at pH 7.1 and 6.6. However, with ML sediment significant differences in survival rate were observed between the control treatment and treatments at pH 7.1 and pH 6.6. Likewise, with MZ sediment significant differences in survival rate were reported between the control and treatment at pH 7.1, while 100% mortality was observed at pH 6.6 and pH 6.1. (see SI Tables S4−S9) Furthermore, differences in survival rate between RSP, MZ, and ML sediment samples were found in the treatments at pH 7.1, pH 6.6, and pH 6.1 (Figure 1).
Several articles have been published in recent years documenting similar responses in marine calcifiers to changes in seawater pH, demonstrating their vulnerability to a reduction of pH. At pH 6.1, after 5 days of exposure, 100% mortality was observed in the crab Necora puber.51 Likewise, 100% mortality was seen when the sea urchin Psammechinus miliaris was exposed to pH 6.16 for 7 days.52 Responses similar to those observed in our study were found in the clams Ruditapes decussatus and Ruditapes philippinarum, where significant mortality was recorded at pH values below 6.9 and 6.5, respectively,53,37 although in these early studies the method used for seawater acidification was the addition of a strong acid, and therefore the values obtained for the pH could be lower than their target value.54,55In recent work, individuals of the polychaete Hediste diversicolor were exposed to RSP and ML sediments with the same pH treatments as used in this study; the results showed a decrease in survival rate in line with pH reduction in ML sediment, but there were no effects on survival rate in individuals exposed to RSP sediment.10 Those findings may indicate that bivalves are more vulnerable than polychaetes when they are subjected to possible acidification scenarios caused by CO2 leakages. Similar behavior in burrowing activity was found in clams exposed to RSP, ML, and MZ sediments subjected to the different pH treatments. After exposure for 48 h to the sediment samples, in the control treatment more than 80% of individuals had burrowed into the sediment. In all three sediments, clams were noticeably less active at the lower pH values, and presented the lowest burrowing activity percentages (0−1.67%) for the pH 6.1 treatment. In individuals exposed to RSP, ML, and MZ, sediments significant differences in burrowing activity were found between the control and treatments at lower pH (see Figure 2 and SI Tables S4−S9). Burrowing activity is considered one of the most important behavioral responses in clams, since it is an escape strategy from predators.24,56 In previous publications burrowing activity has
Figure 1. Average and standard deviation of the percentage of survival of R. philippinarum after 10 days of exposure to sediments (RSP, MZ, and ML) at different pH values. Statistical differences between pH treatments for each sediment sample are indicated by lower case letters (a, b, c, d; P < 0.05), where a refers to control treatment; b refers to pH 7.1 treatment; c refers to pH 6.6 treatment. Significant differences for the pH 6.1 treatment in all sediments and for the pH 6.6 treatment in MZ are not shown because 100% mortality was recorded for all sediments.
Figure 2. Average and standard deviation of the percentage of burrowing activity of R. philippinarum after 48 h of exposure to sediments (RSP, MZ and ML) at different pH values. Statistical differences between pH treatments for each sediment sample are indicated by lower case letters (a,b,c,d; P < 0.05), where a refers to control treatment; b refers to pH 7.1 treatment; c refers to pH 6.6 treatment; and d refers to pH 6.1 treatment. Significant differences for the pH 6.1 treatment in RSP and ML sediment samples are not shown because no burrowing activity was detected.
Table 1. Summarized Results for the Physico-Chemical Parameters of Three Sediments Used in This Study sediment sand fines OC OM Al Fe Mn Cu Zn
RSP 56.12 43.89 1.06 5.07 25096 17400 342.20 19.30 161.92
± ± ± ± ± ± ± ± ±
0.18b,c 0.20b,c 0.16b 0.44 954b 302b 8.74 0.42b,c 6.07b,c
ML 47.30 52.70 2.21 7.20 15208 51537 398.61 703.62 768.34
± ± ± ± ± ± ± ± ±
1.47a 1.53a 0.11a,c 1.77 1336a 1330a,c 17.89 17.95a,c 63.95a,c
MZ 45.41 54.59 1.34 6.36 20369 30142 357.24 886.37 1034
± ± ± ± ± ± ± ± ±
0.85a 0.83a 0.16b 0.06 1703 1107b 14.32 6.25a,b 87.83a,b
a
All concentrations expressed as mg kg-1, except organic carbon (OC) content, sand (63 μm < size < 2 mm), fines (silt and clay,
