Article pubs.acs.org/est
Contaminant Levels in Gulf of Mexico Reef Fish after the Deepwater Horizon Oil Spill As Measured by a Fishermen-Led Testing Program Timothy P. Fitzgerald*,† and Julia M. Gohlke‡ †
Oceans Program, Environmental Defense Fund, 1875 Connecticut Avenue NW, Suite 600, Washington, DC 20009, United States Department of Environmental Health, School of Public Health, University of Alabama at Birmingham (UAB), Birmingham, Alabama 35233, United States
‡
S Supporting Information *
ABSTRACT: The BP oil disaster posed a significant threat to the U.S. seafood industry. Invertebrates (shrimp, oyster, crab) and other nearshore species comprised the majority of postspill testing by federal and state agencies. Deeper water finfish were sampled less frequently, despite population ranges that overlapped with affected waters. We report on a voluntary testing program with Gulf of Mexico commercial fishermen to ensure the safety of their catch. Seven species of reef fish were tested for polycyclic aromatic hydrocarbons, several metals, and a constituent of Corexit 9500A and 9527A dispersants. Only two of 92 samples had detectable levels of benzo(a)pyrene-equivalents (a combined measure of carcinogenic potency across 7 different PAHs), which were still below federal safety thresholds. PAH ratios for these samples suggest pyrogenic (not petrogenic) contamination − indicating potential sources other than Deepwater Horizon. Metals were largely absent (cadmium, lead) or consistent with levels previously reported (mercury, arsenic). One notable exception was tilefish, which showed mercury concentrations lower than expected. We did not detect dispersant in any of our samples, indicating that it was not present in these species during the study period. Our findings suggest minimal risk to public health from these seafoods as a result of the disaster; however, the most contaminated areas were not sampled through this program.
■
crab, oyster)5 − presumably since polycyclic aromatic hydrocarbons (PAHs) and the dispersant component dioctyl sodium sulfosuccinate (DOSS) are not expected to bioaccumulate in teleosts.6 Initially, the majority of screening was done by sensory analysis (i.e., the smell and taste test), with subsets of samples retested with more precise analytical chemistry methods to ensure that concentrations of a subset of PAHs were below levels of concern. Subsequent surveillance samples of reopened areas were chemically analyzed for PAHs. In the meantime, there were noticeable impacts on consumer perceptions of Gulf seafood, despite government assertions that no tainted fish reached the market.2,7 Surveys of consumer attitudes found broad skepticism about the safety of Gulf seafood and the veracity of government statements.8 A national poll in August 2010 found that 54% of respondents lacked confidence in the safety of Gulf seafood from areas affected by the spill.9 Yet another poll in November 2010 showed 48% of respondents reported eating less seafood due to the spill and 18% stating they would not eat any Gulf seafood.9b
INTRODUCTION The explosion and corresponding wellhead failure of the British Petroleum (BP) Deepwater Horizon drilling platform in April 2010 resulted in the largest oil spill in U.S. history. More than 4.9 million barrels of sweet Louisiana crude oil entered the Gulf of Mexico over a three month period, and almost 2 million gallons of Corexit dispersant were used to break up oil in both subsea and surface applications.1 Surveillance testing began on April 28, 2010, and the first emergency fishery closure was announced on May 2, 2010.2 It was only the second time in U.S. history that a federal fishery was closed because of an oil spill (the first was due to the North Cape spill off the coast of Rhode Island in 1996). Government efforts to ensure seafood safety quickly began with spot testing of landed seafood and progressive fishery closures as evidence of surface oil expanded from the well site. The National Oceanic and Atmospheric Administration (NOAA) soon thereafter collaborated with the Food and Drug Administration (FDA) and the five Gulf states to develop a more coordinated protocol for closing areas to fishing and sampling areas for eventual reopening, based largely on an approach presented in Yender at al. (2002)3 and summarized in Gohlke et al. (2011).4 This process involved collecting seafood from both the periphery and inside of closed areas, and a majority of testing was conducted on invertebrates (shrimp, © 2014 American Chemical Society
Received: Revised: Accepted: Published: 1993
November 19, 2013 December 31, 2013 January 8, 2014 January 8, 2014 dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
Figure 1. Map of Gulf Wild (n = 80) and NOAA (n = 143) sample locations for snapper, grouper, and tilefish. The yellow polygon shows the greatest extent of federal fisheries closures on June 2, 2010 (37% of U.S. Exclusive Economic Zone in the Gulf of Mexico). The red circle indicates the location of the Deepwater Horizon wellhead. Green circles indicate the center points of the 10 mile2 grids where Gulf Wild samples were collected. Blue triangles indicate the GPS location of reef fish samples collected by NOAA (described in Methods).
Despite the collection and screening of more than 10,000 seafood specimens for potential contamination,2 federal and state agencies tested relatively few discrete samples of Gulf reef fish species for the reopening of fisheries (n = 70), although in subsequent surveillance sampling of reopened waters, an additional 159 reef fish samples were analyzed. However these results were not made publicly available until 2011. Furthermore, federal and state testing protocols did not include heavy metal analysis in their sampling programs, despite evidence that oil spills alter the natural cycling and accumulation of toxic metals.4,8 More recent studies have even shown that the trace metal content of oil mousse increased as it moved to coastal salt marshes14 and that wetland sediments demonstrated higher concentrations of certain trace metals in heavily oiled Louisiana sites.15 Therefore, the authors designed a sampling program with Gulf reef fishermen to test their commercial catch for PAHs, DOSS, and a suite of metals. The testing protocol described below was applied to Gulf of Mexico red snapper, grouper, and tilefish. Samples were evaluated exclusively by analytical chemistry methods to determine the public health impact of the BP oil disaster − if any − on these important fisheries.
The closures had major immediate and potential economic impact on the Gulf’s biggest fisheries. Total Gulf shrimp landings decreased by 27% in 2010; in Mississippi that figure reached as high as 60%. Menhaden landings in Louisiana also decreased by approximately 80,000 metric tons.10 GNO Inc. (2010) projected a three year loss of $115−$172 million in Louisiana commercial fisheries revenue alone.11 Sumaila et al. (2012) extended these projections to include recreational fisheries, aquaculture, and associated industries, with a high-end estimate of up to $13 billion in losses over a seven year period.12 The Gulf of Mexico reef fish assemblage also supports valuable commercial and recreational fisheries. Many species have life histories that overlap in space and time with the spilled oil, making industries that rely on them particularly susceptible to the ecological and public health effects of the BP oil disaster. For example, red snapper (Lutjanus camphecanus) are distributed throughout the Gulf of Mexico, with juveniles occupying shallow, nearshore areas and adults moving to deeper water of the continental shelf.13 Spawning occurs between June and October and the commercial fishery operates in Texas, Louisiana, Alabama, and Florida, where population rebounds have translated into increasing fishing quotas and revenues since 2010. Several commercially important groupers (Epinephelus and Mycteroperca spp.) follow similar ontogenetic movements and spawning patterns as red snapper, with most species (e.g., red, black) demonstrating the highest abundance off the Florida coast.13a Tilefish generally occur in deeper water, where fishing effort is concentrated off the Florida panhandle and the south Texas coast. In total, commercial landings of this reef fish assemblage accounted for more than $30 million in exvessel revenue in 2011 (NMFS Fisheries Statistics Division, Pers. comm.). This paper outlines the results of a fishing industry-led seafood testing program designed to ensure the safety of their catch and protect the long-term health of their fishery under a branding campaign called Gulf Wild (www.myGulfWild.com).
■
RESULTS The Gulf Wild testing program and resulting data set (Table S1) provided independent sampling of areas along the Texas, Louisiana, and Florida coasts and compliments the reef fish sampling conducted by NOAA (Figure 1). Since Gulf Wild fish were obtained during commercial fishing operations, they did not include samples from the closed area. Benzo[a]pyreneequivalents (BaPE) were estimated applying the U.S. Environmental Protection Agency’s toxic equivalency factors (TEFs) to the 7 carcinogenic PAHs analyzed in this study.16 Of the 92 samples analyzed for PAHs, only one yellowedge and one gag grouper sample had detectable benzo[a]pyrene-equivalents (BaPE). The estimated BaPE of 26 and 0.90 ppb for the 1994
dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
Figure 2. Phenanthrene concentrations from NOAA reef fish samples collected after the Deepwater Horizon blowout (n = 143). A. Linear regression of PHN concentration − as a function of time (days) and distance (km) − shown as a rainbow plate. B. Stem plot map showing spatial distribution of PHN (height of blue lines indicates relative concentrations). The red filled circle indicates the location of the Deepwater Horizon wellhead.
Figure 3. Ratio crossplot of Gulf Wild (green circles) and NOAA (blue triangles) samples. Hashed lines signify thresholds for petrogenic (ANT/ (ANT+PHN) < 0.10, FLA/(FLA+PYR) < 0.50) versus pyrogenic (ANT/(ANT+PHN) > 0.10, FLA/(FLA+PYR) > 0.50) PAH sources based on Tobiszewski and Namiesnik (2012)17 shown in Table 1. Only samples with analyte concentrations above limits of detection (LOD) for either ANT or PHN and FLA or PYR are displayed (n = 7 for Gulf Wild samples, n = 4 for NOAA samples). Lines indicate potential range of ratios if ANT or FLA was below the LOD [LODs for ANT (Gulf Wild 0.5, NOAA (0.12−0.16)) and FLA (Gulf Wild 0.2, NOAA (0.16−0.21))].
suggests differences in baseline sensitivity between laboratories and prevented combined analysis of the two data sets. Relationships between LMW PAHs and time and distance from the Deepwater Horizon wellhead were further examined in the NOAA data set. Results suggested a downward trend in PHN as distance from the wellhead and time from the blowout increased (Figure 2). A regression model with distance from wellhead and time from blowout as predictors was significant (p-value 0.01), and distance from wellhead was a significant predictor (p = 0.006), but time from blowout was not. Relationships with NPH and FLU levels with time from the blowout were suggestive (p = 0.06, p = 0.13), and FLU levels decreased with distance from the wellhead (p = 0.04). No relationship was seen in the Gulf Wild data set, possibly due to the smaller variation in distance from the wellhead as well as more time elapsed since the blowout. Very few samples had detectable levels of higher molecular weight PAHs, which are thought to pose greater health risks and persist for longer in marine environments.5a Anthracene/ phenanthrene (ANT/ANT+PHN) and fluoranthene/pyrene (FLA/FLA+PYR) ratios were determined for those samples
yellowedge and gag grouper samples, respectively, are below the level of concern (LOC) for finfish set by NOAA and FDA (35 ppb) but are above a more conservative LOC estimated for a high consumption (2 seafood meals per week) child (0.60 ppb).4 All samples analyzed for DOSS (n = 57) were below the detection limit (0.05 ppm) and therefore well below the NOAA/FDA LOC set at 100 ppm for finfish. To further characterize PAH distributions in reef fish in the Gulf of Mexico during and after the Deepwater Horizon event, concentrations of LMW PAHs including naphthalene (NPH), fluorene (FLU), and phenanthrene (PHN), which were above limits of detection in many samples, were analyzed. Chemical analyses conducted by NOAA in 2010−2011 were examined along with the current Gulf Wild data set. For those samples over the limit of detection (LOD), the Gulf Wild data set had higher average estimates of LMW PAHs than the NOAA data set (e.g., PHN (1.6 ppb (STD = 1.8) vs 0.35 ppb (STD = 0.14) and NPH (5.8 ppb (STD = 8.7) vs 0.8 ppb (STD = 0.26)). NOAA LODs were lower for NPH but comparable with Gulf Wild for PHN (0.3 ppb). Only three samples had detectable levels of FLU in the Gulf Wild data set. This comparison 1995
dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
diagnostic ratios that identify emission source have been a focus of previous research and have focused on the HMW PAHs due to their relative stability in the environment. For example, the phenanthrene/anthracene ratio of >10 and fluoranthene/ pyrene ratio of 1 >10 AND
Contaminant Levels in Gulf of Mexico Reef Fish after the Deepwater Horizon Oil Spill As Measured by a Fishermen-Led Testing Program Timothy P. Fitzgerald*,† and Julia M. Gohlke‡ †
Oceans Program, Environmental Defense Fund, 1875 Connecticut Avenue NW, Suite 600, Washington, DC 20009, United States Department of Environmental Health, School of Public Health, University of Alabama at Birmingham (UAB), Birmingham, Alabama 35233, United States
‡
S Supporting Information *
ABSTRACT: The BP oil disaster posed a significant threat to the U.S. seafood industry. Invertebrates (shrimp, oyster, crab) and other nearshore species comprised the majority of postspill testing by federal and state agencies. Deeper water finfish were sampled less frequently, despite population ranges that overlapped with affected waters. We report on a voluntary testing program with Gulf of Mexico commercial fishermen to ensure the safety of their catch. Seven species of reef fish were tested for polycyclic aromatic hydrocarbons, several metals, and a constituent of Corexit 9500A and 9527A dispersants. Only two of 92 samples had detectable levels of benzo(a)pyrene-equivalents (a combined measure of carcinogenic potency across 7 different PAHs), which were still below federal safety thresholds. PAH ratios for these samples suggest pyrogenic (not petrogenic) contamination − indicating potential sources other than Deepwater Horizon. Metals were largely absent (cadmium, lead) or consistent with levels previously reported (mercury, arsenic). One notable exception was tilefish, which showed mercury concentrations lower than expected. We did not detect dispersant in any of our samples, indicating that it was not present in these species during the study period. Our findings suggest minimal risk to public health from these seafoods as a result of the disaster; however, the most contaminated areas were not sampled through this program.
■
crab, oyster)5 − presumably since polycyclic aromatic hydrocarbons (PAHs) and the dispersant component dioctyl sodium sulfosuccinate (DOSS) are not expected to bioaccumulate in teleosts.6 Initially, the majority of screening was done by sensory analysis (i.e., the smell and taste test), with subsets of samples retested with more precise analytical chemistry methods to ensure that concentrations of a subset of PAHs were below levels of concern. Subsequent surveillance samples of reopened areas were chemically analyzed for PAHs. In the meantime, there were noticeable impacts on consumer perceptions of Gulf seafood, despite government assertions that no tainted fish reached the market.2,7 Surveys of consumer attitudes found broad skepticism about the safety of Gulf seafood and the veracity of government statements.8 A national poll in August 2010 found that 54% of respondents lacked confidence in the safety of Gulf seafood from areas affected by the spill.9 Yet another poll in November 2010 showed 48% of respondents reported eating less seafood due to the spill and 18% stating they would not eat any Gulf seafood.9b
INTRODUCTION The explosion and corresponding wellhead failure of the British Petroleum (BP) Deepwater Horizon drilling platform in April 2010 resulted in the largest oil spill in U.S. history. More than 4.9 million barrels of sweet Louisiana crude oil entered the Gulf of Mexico over a three month period, and almost 2 million gallons of Corexit dispersant were used to break up oil in both subsea and surface applications.1 Surveillance testing began on April 28, 2010, and the first emergency fishery closure was announced on May 2, 2010.2 It was only the second time in U.S. history that a federal fishery was closed because of an oil spill (the first was due to the North Cape spill off the coast of Rhode Island in 1996). Government efforts to ensure seafood safety quickly began with spot testing of landed seafood and progressive fishery closures as evidence of surface oil expanded from the well site. The National Oceanic and Atmospheric Administration (NOAA) soon thereafter collaborated with the Food and Drug Administration (FDA) and the five Gulf states to develop a more coordinated protocol for closing areas to fishing and sampling areas for eventual reopening, based largely on an approach presented in Yender at al. (2002)3 and summarized in Gohlke et al. (2011).4 This process involved collecting seafood from both the periphery and inside of closed areas, and a majority of testing was conducted on invertebrates (shrimp, © 2014 American Chemical Society
Received: Revised: Accepted: Published: 1993
November 19, 2013 December 31, 2013 January 8, 2014 January 8, 2014 dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
Figure 1. Map of Gulf Wild (n = 80) and NOAA (n = 143) sample locations for snapper, grouper, and tilefish. The yellow polygon shows the greatest extent of federal fisheries closures on June 2, 2010 (37% of U.S. Exclusive Economic Zone in the Gulf of Mexico). The red circle indicates the location of the Deepwater Horizon wellhead. Green circles indicate the center points of the 10 mile2 grids where Gulf Wild samples were collected. Blue triangles indicate the GPS location of reef fish samples collected by NOAA (described in Methods).
Despite the collection and screening of more than 10,000 seafood specimens for potential contamination,2 federal and state agencies tested relatively few discrete samples of Gulf reef fish species for the reopening of fisheries (n = 70), although in subsequent surveillance sampling of reopened waters, an additional 159 reef fish samples were analyzed. However these results were not made publicly available until 2011. Furthermore, federal and state testing protocols did not include heavy metal analysis in their sampling programs, despite evidence that oil spills alter the natural cycling and accumulation of toxic metals.4,8 More recent studies have even shown that the trace metal content of oil mousse increased as it moved to coastal salt marshes14 and that wetland sediments demonstrated higher concentrations of certain trace metals in heavily oiled Louisiana sites.15 Therefore, the authors designed a sampling program with Gulf reef fishermen to test their commercial catch for PAHs, DOSS, and a suite of metals. The testing protocol described below was applied to Gulf of Mexico red snapper, grouper, and tilefish. Samples were evaluated exclusively by analytical chemistry methods to determine the public health impact of the BP oil disaster − if any − on these important fisheries.
The closures had major immediate and potential economic impact on the Gulf’s biggest fisheries. Total Gulf shrimp landings decreased by 27% in 2010; in Mississippi that figure reached as high as 60%. Menhaden landings in Louisiana also decreased by approximately 80,000 metric tons.10 GNO Inc. (2010) projected a three year loss of $115−$172 million in Louisiana commercial fisheries revenue alone.11 Sumaila et al. (2012) extended these projections to include recreational fisheries, aquaculture, and associated industries, with a high-end estimate of up to $13 billion in losses over a seven year period.12 The Gulf of Mexico reef fish assemblage also supports valuable commercial and recreational fisheries. Many species have life histories that overlap in space and time with the spilled oil, making industries that rely on them particularly susceptible to the ecological and public health effects of the BP oil disaster. For example, red snapper (Lutjanus camphecanus) are distributed throughout the Gulf of Mexico, with juveniles occupying shallow, nearshore areas and adults moving to deeper water of the continental shelf.13 Spawning occurs between June and October and the commercial fishery operates in Texas, Louisiana, Alabama, and Florida, where population rebounds have translated into increasing fishing quotas and revenues since 2010. Several commercially important groupers (Epinephelus and Mycteroperca spp.) follow similar ontogenetic movements and spawning patterns as red snapper, with most species (e.g., red, black) demonstrating the highest abundance off the Florida coast.13a Tilefish generally occur in deeper water, where fishing effort is concentrated off the Florida panhandle and the south Texas coast. In total, commercial landings of this reef fish assemblage accounted for more than $30 million in exvessel revenue in 2011 (NMFS Fisheries Statistics Division, Pers. comm.). This paper outlines the results of a fishing industry-led seafood testing program designed to ensure the safety of their catch and protect the long-term health of their fishery under a branding campaign called Gulf Wild (www.myGulfWild.com).
■
RESULTS The Gulf Wild testing program and resulting data set (Table S1) provided independent sampling of areas along the Texas, Louisiana, and Florida coasts and compliments the reef fish sampling conducted by NOAA (Figure 1). Since Gulf Wild fish were obtained during commercial fishing operations, they did not include samples from the closed area. Benzo[a]pyreneequivalents (BaPE) were estimated applying the U.S. Environmental Protection Agency’s toxic equivalency factors (TEFs) to the 7 carcinogenic PAHs analyzed in this study.16 Of the 92 samples analyzed for PAHs, only one yellowedge and one gag grouper sample had detectable benzo[a]pyrene-equivalents (BaPE). The estimated BaPE of 26 and 0.90 ppb for the 1994
dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
Figure 2. Phenanthrene concentrations from NOAA reef fish samples collected after the Deepwater Horizon blowout (n = 143). A. Linear regression of PHN concentration − as a function of time (days) and distance (km) − shown as a rainbow plate. B. Stem plot map showing spatial distribution of PHN (height of blue lines indicates relative concentrations). The red filled circle indicates the location of the Deepwater Horizon wellhead.
Figure 3. Ratio crossplot of Gulf Wild (green circles) and NOAA (blue triangles) samples. Hashed lines signify thresholds for petrogenic (ANT/ (ANT+PHN) < 0.10, FLA/(FLA+PYR) < 0.50) versus pyrogenic (ANT/(ANT+PHN) > 0.10, FLA/(FLA+PYR) > 0.50) PAH sources based on Tobiszewski and Namiesnik (2012)17 shown in Table 1. Only samples with analyte concentrations above limits of detection (LOD) for either ANT or PHN and FLA or PYR are displayed (n = 7 for Gulf Wild samples, n = 4 for NOAA samples). Lines indicate potential range of ratios if ANT or FLA was below the LOD [LODs for ANT (Gulf Wild 0.5, NOAA (0.12−0.16)) and FLA (Gulf Wild 0.2, NOAA (0.16−0.21))].
suggests differences in baseline sensitivity between laboratories and prevented combined analysis of the two data sets. Relationships between LMW PAHs and time and distance from the Deepwater Horizon wellhead were further examined in the NOAA data set. Results suggested a downward trend in PHN as distance from the wellhead and time from the blowout increased (Figure 2). A regression model with distance from wellhead and time from blowout as predictors was significant (p-value 0.01), and distance from wellhead was a significant predictor (p = 0.006), but time from blowout was not. Relationships with NPH and FLU levels with time from the blowout were suggestive (p = 0.06, p = 0.13), and FLU levels decreased with distance from the wellhead (p = 0.04). No relationship was seen in the Gulf Wild data set, possibly due to the smaller variation in distance from the wellhead as well as more time elapsed since the blowout. Very few samples had detectable levels of higher molecular weight PAHs, which are thought to pose greater health risks and persist for longer in marine environments.5a Anthracene/ phenanthrene (ANT/ANT+PHN) and fluoranthene/pyrene (FLA/FLA+PYR) ratios were determined for those samples
yellowedge and gag grouper samples, respectively, are below the level of concern (LOC) for finfish set by NOAA and FDA (35 ppb) but are above a more conservative LOC estimated for a high consumption (2 seafood meals per week) child (0.60 ppb).4 All samples analyzed for DOSS (n = 57) were below the detection limit (0.05 ppm) and therefore well below the NOAA/FDA LOC set at 100 ppm for finfish. To further characterize PAH distributions in reef fish in the Gulf of Mexico during and after the Deepwater Horizon event, concentrations of LMW PAHs including naphthalene (NPH), fluorene (FLU), and phenanthrene (PHN), which were above limits of detection in many samples, were analyzed. Chemical analyses conducted by NOAA in 2010−2011 were examined along with the current Gulf Wild data set. For those samples over the limit of detection (LOD), the Gulf Wild data set had higher average estimates of LMW PAHs than the NOAA data set (e.g., PHN (1.6 ppb (STD = 1.8) vs 0.35 ppb (STD = 0.14) and NPH (5.8 ppb (STD = 8.7) vs 0.8 ppb (STD = 0.26)). NOAA LODs were lower for NPH but comparable with Gulf Wild for PHN (0.3 ppb). Only three samples had detectable levels of FLU in the Gulf Wild data set. This comparison 1995
dx.doi.org/10.1021/es4051555 | Environ. Sci. Technol. 2014, 48, 1993−2000
Environmental Science & Technology
Article
diagnostic ratios that identify emission source have been a focus of previous research and have focused on the HMW PAHs due to their relative stability in the environment. For example, the phenanthrene/anthracene ratio of >10 and fluoranthene/ pyrene ratio of 1 >10 AND
