lable at ScienceDirect

Estuarine, Coastal and Shelf Science 148 (2014) 1e13

Contents lists avai

Estuarine, Coastal and Shelf Science

journal homepage: www.elsevier .com/locate/ecss

Invited feature

Coastal ocean acidification: The other eutrophication problem

Ryan B. Wallace a, Hannes Baumann a, Jason S. Grear b, Robert C. Aller a,Christopher J. Gobler a, *

a Stony Brook University, School of Marine and Atmospheric Sciences, 239 Montauk Hwy, Southampton, NY 11968, USAb US Environmental Protection Agency, Atlantic Ecology Division, National Health and Environmental Effects Research Laboratory, Office of Research andDevelopment, 27 Tarzwell Dr, Narragansett, RI 02882, USA

a r t i c l e i n f o

Article history:Received 8 March 2014Accepted 26 May 2014Available online 5 June 2014

Keywords:acidificationpHestuaryhypoxiacalcium carbonate saturationrespiration

* Corresponding author.E-mail address: christopher.gobler@stonybrook.ed

http://dx.doi.org/10.1016/j.ecss.2014.05.0270272-7714/© 2014 Elsevier Ltd. All rights reserved.

a b s t r a c t

Increased nutrient loading into estuaries causes the accumulation of algal biomass, and microbialdegradation of this organic matter decreases oxygen levels and contributes towards hypoxia. A second,often overlooked consequence of microbial degradation of organic matter is the production of carbondioxide (CO2) and a lowering of seawater pH. To assess the potential for acidification in eutrophic es-tuaries, the levels of dissolved oxygen (DO), pH, the partial pressure of carbon dioxide (pCO2), and thesaturation state for aragonite (Uaragonite) were horizontally and vertically assessed during the onset, peak,and demise of low oxygen conditions in systems across the northeast US including Narragansett Bay (RI),Long Island Sound (CTeNY), Jamaica Bay (NY), and Hempstead Bay (NY). Low pH conditions (<7.4) weredetected in all systems during summer and fall months concurrent with the decline in DO concentra-tions. While hypoxic waters and/or regions in close proximity to sewage discharge had extremely highlevels of pCO2, (>3000 matm), were acidic pH (<7.0), and were undersaturated with regard to aragonite(Uaragonite < 1), even near-normoxic but eutrophic regions of these estuaries were often relativelyacidified (pH < 7.7) during late summer and/or early fall. The close spatial and temporal correspondencebetween DO and pH and the occurrence of extremes in these conditions in regions with the most intensenutrient loading indicated that they were primarily driven by microbial respiration. Given that coastalacidification is promoted by nutrient-enhanced organic matter loading and reaches levels that havepreviously been shown to negatively impact the growth and survival of marine organisms, it may beconsidered an additional symptom of eutrophication that warrants managerial attention.

© 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Coastal marine ecosystems are amongst the most ecologicallyand economically productive areas on the planet, providing morethan US$10 trillion in annual resources or ~40% of the globalecosystem goods and services (Costanza et al., 1997). Approxi-mately 40% of the World population lives within 100 km of acoastline, making these regions subject to a suite of anthropogenicstressors including intense nutrient loading (de Jonge et al., 2002;Valiela, 2006). Excessive nutrient loading into coastal ecosystemspromotes algal productivity and the subsequent microbial con-sumption of this organic matter reduces oxygen levels and canpromote hypoxia (Cloern, 2001; Heisler et al., 2008). The rapidacceleration of nutrient loading to coastal zones in recent decades

u (C.J. Gobler).

has contributed to a significant expansion in hypoxic zones acrossthe globe (Rabalais et al., 2002; Diaz and Rosenberg, 2008). Becausethese hypoxic zones can be lethal to aerobic marine organisms andhave contributed toward declining yields of fisheries (Vaquer-Sunyer and Duarte, 2008; Levin et al., 2009), a prime motivationof coastal zone management has been to lower nutrient loads inorder to reduce the intensity, extent, and duration of hypoxia(Scavia et al., 2004; Paerl, 2006; Scavia and Bricker, 2006).

A second, often overlooked consequence of microbial degrada-tion of organic matter in coastal zones is the production of CO2,which enters the water and forms carbonic acid (H2CO3) dissoci-ating into bicarbonate ions (HCO�

3 ), carbonate ions (CO2�3 ) and

hydrogen ions (Hþ) that cause acidification. Coastal ecosystems canalso be acidified via atmospheric carbon dioxide fluxes (Miller et al.,2009; Feely et al., 2010), the introduction of acidic river water(Salisbury et al., 2008), and/or upwelling of CO2 enriched deepwater (Feely et al., 2008). Watershed geology, climate and aciddeposition also have the potential to impact source water buffering

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e132

capacities through their effects on mineral weathering and sea icemelting (Yamamoto-Kawai et al., 2009; Wang et al., 2013). Likelydue to a combination of these processes, some recent investigationsof coastal zones have detected seasonally low pH and/or elevatedpCO2 conditions (Feely et al., 2010; Cai et al., 2011; Sunda and Cai,2012; Duarte et al., 2013; Melzner et al., 2013), while others havedocumented that some coastal regions have experienced progres-sively declining pH levels in recent decades (Waldbusser et al.,2011).

Ocean acidification has garnered much attention among scien-tists, policy makers, and the public during the past decade. Theterm has generally been used to describe the ongoing decrease insurface ocean pH due to anthropogenic increases in atmosphericCO2 (Caldeira and Wickett, 2003). Since this acidification also de-creases the availability of CO�2

3 it can have negative consequencesfor marine organisms, both pelagic and benthic, that producestructures made of calcium carbonate (CaCO3; Ries et al., 2009;Talmage and Gobler, 2009; Gazeau et al., 2013) and has some-times been deemed ‘The other CO2 problem’ (Doney et al., 2009). Asimilar perspective could be adopted with respect to eutrophica-tion processes, the literature on which generally emphasizes onlynutrient-productivity-hypoxia relationships. Beyond calcifying or-ganisms, information is mounting that acidification can also bedamaging to some finfish (Munday et al., 2010), particularly duringearly life stages (Baumann et al., 2012; Frommel et al., 2012; Murrayet al., 2014). While acidification in the open ocean is driven byexternal, atmospheric loading of CO2, in coastal zones this processmay be minor compared to internal loading processes, particularlywithin eutrophic regions where excessive nutrient loading andorganic matter production have been associated with large hypoxicevents.

The goal of this study was to characterize the spatial and tem-poral dynamics of DO, pH, pCO2, and Uaragonite in the water columnsof several representative, estuarine systems along the northern USAtlantic coast (Fig. 1). Multi-year monitoring datasets wereassessed to define seasonal patterns in pH and DO while cruises

Fig. 1. Study sites. Long Island Sound CTDEEP sampling stations represented as black dotoccupied for this study. Narragansett Bay in upper right box with black dots of inset indicatleft, with black dots within inset depicting Jamaica Bay sampling stations. Upper left inset:

were conducted to vertically and horizontally resolve spatial pat-terns of acidification during the seasonal onset, peak, and decline ofhypoxia in these estuaries. Given the excessive productivity andorganic matter loading within eutrophic estuaries, we hypothe-sized that pH and DOwould co-vary and reach extremes lower thanthose in the open ocean and most other coastal systems charac-terized to date. As a guide to possible impacts on ecosystem pro-cesses, we evaluated the saturation state of the water column withrespect to common biogenic carbonate minerals, specificallyaragonite.

2. Methods

This study characterized spatial and temporal patterns of DO,pH, pCO2, and Uaragonite in four, semi-enclosed estuarine systemsacross the Northeast US: Narragansett Bay (Rhode Island (RI),41.63N, 71.37W), Long Island Sound (New York (NY)-Connecticut(CT), 41.11N, 72.86W), Jamaica Bay (NY, 40.61N, 73.84W), andHempstead Bay (NY, 40.60N, 73.57W; Fig. 1). The Northeast US isthe most heavily urbanized and populated region in the nation andcontains some of the densest populations on the planet includingthe most populous US metropolis, New York City. Coastal waters inthis region have also been shown to have a lower buffer capacitycompared to the southeast and Gulf coasts of the US (Wang et al.,2013). We utilized three approaches for this study: 1) The anal-ysis of monthly monitoring data across Long Island Sound; 2)Vertical measurements of water column conditions across Narra-gansett Bay, Long Island Sound, and Jamaica Bay; and 3) Contin-uous, horizontal mapping of conditions across Jamaica Bay andHempstead Bay. All field work was performed during daylight.

Long Island Sound (LIS; Fig. 1) is the third largest estuary in theUS, and its western end receives more than four billion liters oftreated wastewater effluent daily from the nation's largestmetropolis, New York City. The combination of excessive waste-water loads and sluggish circulation have contributed toward theannual occurrence of hypoxia within western LIS since the middle

s whereas small black dots surrounded by circles indicate CTDEEP sampling stationsing discrete sampling stations. Jamaica Bay and Hempstead Bay in small boxes at lowerNortheast US.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 3

of the twentieth century (Parker and O'Reilly, 1991; O’Donnell et al.,2014; Suter et al., 2014). The LIS Comprehensive Conservation andManagement Planwas implemented in 1994 (USEPA,1994), leadingto a nitrogen loading reduction target of 58%. This managementplan further initiated monthly monitoring of LIS by the ConnecticutDepartment of Energy and Environmental Protection (CTDEEP). Forthis study, we compiled CTDEEP's measurements of DO and pHNBS(pH on the National Bureau of Standards scale) from the start of itsmonitoring of both parameters in August 2010 through 2012.CTDEEP conducts vertical surveys at dozens of stations in LIS fromNew York City to the coastal Atlantic Ocean using a Sea-Bird Elec-tronics© SBE 19 SEACAT CTD, transmitting data at 4 Hz. The SBE 19was outfitted with a Seabird SBE 43 DO sensor (polarographicmembrane sensor); oxygen measurements were validated viaWinkler titration measurements (Parsons et al., 1984) of discretesamples. The SBE 19 was also equipped with a Seabird 18 pH sensorequipped with a pressure balanced glass electrode/Ag/AgeClreference probe that provided pH measurements on the NBS scale.Prior to each survey, a three point calibrationwas conducted on thepH sensor using commercial buffer solutions (±0.02 pH). Finally, aWET Labs WETStar fluorometer was used to measure in vivochlorophyll a fluorescence which was converted to units of mgchlorophyll a L�1 using discrete measurements of extracted chlo-rophyll a. For analyses presented here, measurements from the 17mid-estuarine stations that span the entire 150 km longitudinalextent of LIS were used to generate surface and bottom time seriesplots (2010e2012) as well as vertical section plots using Ocean DataView (ODV 4.6.0; http://odv.awi.de/) over an annual cycle in 2011.

Following analysis of the CTDEEP data set, we performed cruisesin 2012 and 2013 based on the hypothesis that the same processespromoting hypoxia (intense microbial respiration) also acidify thewater column in coastal ecosystems. These cruises focused on thewestern half of LIS that experiences hypoxia and has the strongesteutrophication gradient within this system (O'Shea and Brosnan,2000; Gobler et al., 2006) and proceeded further towards NewYork City than cruises performed by CTDEEP. In 2012, cruises wereperformed during mid- and late August when hypoxia has beenhistorically most intense in LIS (Parker and O’Reilly, 1991; O’Sheaand Bronsan, 2000; O'Donnell et al., 2014) whereas in 2013, fivecruises were performed between July and October to capture theonset, peak, and demise of hypoxia. During each cruise, wecollected water samples from multiple depths using Niskin bottlesand vertically profiled the water column at eight stations across LIS(Fig. 1) onboard the R/V Seawolf (Stony Brook University). Stationlocations were recorded using a GPS datalogger (Sparkfun elec-tronics, model: GeoChron Blue). A Sea-Bird Electronics SBE 9plusCTD, a Contros HydroC CO2 sensor, and a Satlantic SeaFET ocean pHsensor were affixed to a standard rosette sampler to providecontinuous, vertical measurements of DO, temperature, salinity,depth, pCO2, and pHT (pH on the total Hþ scale). The SBE 9plus wasequipped with a SBE 43 dissolved oxygen sensor described above,transmitted at 24 Hz, was calibrated to 100% saturation in air priorto use, and had an analytical precision of 2.1%. The HydroC andSeaFET sensors had analytical precisions of 1.4% and 0.09%,respectively, and transmitted at 1 and 0.3 Hz, respectively. TheHydroC measures pCO2 concentrations via non-dispersive infraredspectrometry (NDIR) after dissolved gasses within seawaterpermeate a hydrophobic membrane and equilibrate with the innerpumped gas circuit (Fiedler et al., 2012; Fietzek et al., 2014). Alaminar flow pump (Sea-Bird Electronics) affixed to the HydroCcontinuously delivers seawater across the sensor membrane. TheHydroC was calibrated annually using sodium carbonate and bi-carbonate standards (200, 400, 800, 1600, 2400, 3200, 4000 matm)in quantities mimicking the CO2 buffer system of seawater (DIC:TA)and ran internal blanks every five minutes while in use (Fietzek

et al., 2014). The SeaFET pH sensor utilizes ion-selective field ef-fect transistor (ISFET) technology coupled with a solid state refer-ence electrode to measure pH on the total Hþ scale (pHT) and wascalibrated annually by the manufacturer. A comparison of pHNBSand pHT scales demonstrated a typical offset of ~0.12 during thisstudy (pHT < pHNBS). Discrete water samples for the direct mea-surement of pHT and total dissolved inorganic carbon (DIC) werecollected at multiple depths via Niskin bottles, preserved using asaturatedmercuric chloride (HgCl2) solution, and stored at 4 �C. DICmeasurements were made using an EGM-4 Environmental GasAnalyzer® (PP Systems) after acidification and separation of the gasphase from seawater using a Liqui-Cel® Membrane (Membrana).This instrument was calibrated using standards made from sodiumbicarbonate and provided a methodological precision of ±0.85% forreplicated measurements. As a quality assurance measure, certifiedreference material generated by Dr. Andrew Dickson's lab (Uni-versity of California San Diego, Scripps Institution of Oceanography;Batch 132 ¼ 2033 mmol DIC kg seawater�1) was analyzed inquadruplicate immediately before and after the analysis of everyset of samples and yielded full recovery (measuredvalues ¼ 104 ± 3.87% of certified values). Levels of pCO2 and Uar-

agonite were calculated based on pressure and measured levels ofDIC, pHT, temperature, salinity, phosphate, silicate and first andsecond dissociation constants of carbonic acid in estuarine watersaccording toMillero (2010) using the program CO2SYS (http://cdiac.ornl.gov/ftp/co2sys/). The parameter Uaragonite is a measure of thesaturation state of the water with respect to aragonite and isdefined as Uaragonite ¼ [Ca2þ][CO3

2�]/Ksp,arag, where Ksp,arag is thesolubility product of Ca2þ and CO2�

3 . There are three commonbiogenic carbonate mineral groups: aragonite, low-Mg calcite, andhigh-Mg calcite, the latter having a spectrum of compositions andsolubilites (Morse et al., 2007). We chose to use aragonite, a rela-tively soluble mineral present in many invertebrate larval stages(Weiss et al., 2002), as a representative form. During this study,levels of pCO2 measured with the HydroC were nearly identical to,and never significantly different, from levels calculated for discretesamples based on DIC and pHT, a finding consistent with our prioruse of this device (Baumann et al., 2014). Regressions of pCO2 levelsmeasured via the HydroC™ and those calculated from paralleldiscrete samples were highly significant (p < 0.001) with a corre-lation coefficient exceeding 0.9 and a slope not significantlydifferent from 1.0. Discrete pHT measurements were also made onvertically collected samples using a DuraFET III (Honeywell) ISFET-based solid state pH sensor and were made spectrophotometricallyusing m-cresol purple (Dickson et al., 2007). Measurements of pHTwith the SeaFET were nearly identical to, and never significantlydifferent, from those obtained with the DuraFET III or spectro-photometric measurements; regressions of the various pH mea-surements were highly significant (p < 0.001) with correlationcoefficients exceeding 0.96 and slopes not significantly differentfrom 1.0. Vertical section plots of DO, pHT, pCO2 and Uaragonite in LISwere made using ODV 4.6.0.

Narragansett Bay is the largest estuary in New England, USA,and is bordered by Rhode Island (RI) and Massachusetts (MA), aswell as RI's largest city, Providence. Narragansett Bay is known toexperience regionally high levels of nutrient loading and hypoxia(Nixon et al., 2008). To establish the levels of DO, pHT, pCO2, andUaragonite across Narragansett Bay during summer, monthly cruiseswere performed from June through August of 2013. The watercolumn was vertically sampled at seven geo-referenced stationsacross the complete latitudinal gradient of the estuary from thebrackish regions of the Seekonk River to the region where the Bayexchanges with Block Island Sound (Fig. 1). Cruises were conductedonboard the USEPA's R/V Coastal Explorer fromwhich Niskin bottles,the HydroC CO2 sensor, and a YSI 6920 V2-2 (Yellow Springs, Inc)

Fig. 2. Time series of DO and pHNBS, August, 2010eOctober, 2012. A) Western LongIsland Sound, bottom & B) surface. C) Eastern Long Island Sound, bottom & D) surface.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e134

sonde were used to vertically assess levels of pCO2, temperature,salinity, and DO concentrations. The 6920 V2 was equipped with aYSI 6560 conductivity/temperature probe, and a YSI 6450 ROX®

optical DO sensor with an analytical precision of 4%. Each probewasseparately calibrated prior to deployments using a 50,000 mS cm�1

conductivity solution and air saturated distilled water, respectively.Discrete water samples were collected at multiple depths at eachstation using Niskin bottles and preserved to make discrete mea-surements of DIC and pHT as described above. Chlorophyll a indiscrete samples was quantified according to Parsons et al. (1984).Vertical section plots of DO, pHT, pCO2 and Uaragonite were made asdescribed above.

Jamaica Bay is a hypereutrophic lagoonal estuary located withinthe New York City boroughs of Brooklyn and Queens, and receives~90% of its nitrogen load from four major wastewater-treatmentplants that discharge ~ one billion liters of treated effluent daily(Benotti et al., 2007). Vertical sampling of Jamaica Bay was per-formed at five stations across northeastern Jamaica Bay (Fig. 1)during the peak (August) and decline (September, November) ofhypoxia. Vertical profiles of DO, pHT, and pCO2 were generatedusing the HydroC CO2 sensor, the SeaFET pH sensor, and the YSI6920 V2-2 sonde as described above. Vertical section plots of DO,pHT, pCO2 and Uaragonite were made as described above.

Levels of DO and pCO2 were profiled horizontally acrossHempstead Bay and Jamaica Bay, NY, US, during the fall of 2011.Hempstead Bay is a eutrophic lagoon comprised of intermittent saltmarshes and dredged navigational channels. The Bay Park sewagetreatment plant receives wastewater from more than a half millionhomes in Nassau County, NY, and discharges 200 million liters oftreated effluent per day into Hempstead Bay. Both horizontalmapping cruises were performed using the HydroC CO2 sensor andYSI 6920 V2-2 sonde affixed 0.5 m below the water surface to abracket on a small vessel that proceeded ~1 m s�1 to minimizeturbulent mixing around sensors. Cruise tracks were geo-referenced using a GPS datalogger (Sparkfun electronics, Geo-Chron Blue). The cruise through Hempstead Bay began at the JonesInlet to the Atlantic Ocean, progressed past the Bay Park sewagetreatment plant outfall, and north into Hempstead Bay. In JamaicaBay, the cruise began within the northeastern extent of the Bayreceiving>300million liters per day of treatedwastewater effluent,and ended within the inlet to the Atlantic Ocean. Maps of contin-ually measured levels of DO and pCO2 were generated using ArcGIS10 (ESRI, Redlands, CA).

3. Results

Over an annual cycle, levels of pHNBS and DO were stronglycoupled and highly dynamic within the western extent of LIS,particularlywithin bottomwaterswhere peak levels of pHNBS (8.76)and DO (11.83 mg L�1) occurred in winter coincident with thewinterespring bloom and minimal values of 7.23 and 0.94 mg L�1,respectively, occurred during summer (Fig. 2A). Similar temporalpatterns occurred within surface waters of western LIS, althoughthe values were less extreme with pHNBS ranging from 8.32 inwinter to 7.48 in summer and DO ranging from 13.1 mg L�1 inwinter to 5.33 mg L�1 during summer (Fig. 2B). Within eastern LIS,similar temporal patterns were observed for DO, although therange in values was smaller for bottom (4.86e11.5 mg L�1; Fig. 2C)and surface waters (6.43e11.7 mg L�1; Fig. 2D) compared towestern LIS. pHNBS within eastern LIS also had a smaller dynamicrange (7.43e8.57), was similar between surface and bottomwaters,and displayed a less distinct seasonal pattern.

A cross sectional examination of chlorophyll a, pH and DO in LISrevealed the vertical, horizontal, and temporal evolution ofphytoplankton biomass, hypoxia, and acidification in this system.

While oxygen and pH levels were normal throughout the watercolumn in LIS duringMay (pHNBS > 8; DO > 7.5 mg L�1; Fig. 3), theseconditions changed with the onset of summer. Specifically, pHlevels were lower throughout LIS in June, while DO levels began todecline in western bottom waters (Fig. 3). During July, August, andSeptember, hypoxic and acidified conditions were present inwestern LIS as DO levels declined to <3 mg L�1 and pHNBS levelsdecreased to <7.4 (Fig. 3). These conditions developed despite acontinuous layer of dense algal biomass (chlorophyll a >30 mg L�1)in surface waters of western LIS (Fig. 3). In October, higher levels ofDO and lower levels of algal biomass were present throughout LIS,but low pH (�7.7) persisted inwestern LIS (Fig. 3). By November, pHand DO levels throughout the entire water column were similar toMay being >7.9 and >7.5 mg L�1, respectively (Fig. 3).

Given the co-occurrence of hypoxia and acidification in westernLIS during August of 2010 and 2011, a cruise focusing on thewestern half of LIS was performed to assess the levels of pCO2 andUaragonite present in August 2012. While the surface mixed layer ofeastern LIS had elevated levels of pHT and DO (>7.7, >7 mg L�1),moderate pCO2 levels (~500 matm), and was saturated with respectto aragonite (Uaragonite > 1; Fig. 4), these conditions deterioratedsignificantly to the west, particularly within bottom waters. Spe-cifically, nearly all bottom waters of LIS had pHT values <7.5, DOconcentrations <4 mg L�1, pCO2 concentrations >1500 matm, andwere undersaturated with respect to aragonite (Uaragonite < 1;Fig. 4). The most extreme of these conditions were found in the farwest where bottom DO values were <2 mg L�1 (Fig. 4), and pHlevels were low throughout the water column and in some cases

Fig. 3. Monthly vertical section plots of DO (measured via a SBE 43 DO sensor), pHNBS (measured via a Seabird 18 pH sensor), and chlorophyll a (Chl-a; measured via a WET LabsWETStar fluorometer) with temperature in �C (measured via a SBE 19 SEACAT CTD) contour lines during MayeNovember 2011 in Long Island Sound. Vertical lines indicate CTDprofiles. Depth is 0e45 m.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 5

were acidic (<7.0; Fig. 4). Concentrations of pCO2 in the west wereextremely high (>2500 matm) in both surface and bottom watersand Uaragonite was <0.5 throughout the entire water column inwestern LIS (Fig. 4). A second cruise in LIS during late August of2012 showed the spatial distribution and levels of pCO2 and pHTwere highly similar to the first cruise, while levels of DO wereslightly higher (>2 mg L�1; data not shown).

In 2013, cruises were performed July through October in LIS tobetter assess the temporal dynamics of carbonate chemistry inconjunction with hypoxia and acidification. While the spatial dy-namics of DO, pH, pCO2 and Uaragonite in 2013 were similar to 2012,the intensity of hypoxia was lower, as bottom DO levels remainedabove 2.4 mg L�1 during all sampling (Fig. 5). Despite the lesserintensity of hypoxia, low pH (<7.2) and high pCO2 (>2500 matm)conditions again developed during August of 2013 (Fig. 5). In amanner similar to 2012, the western-most region experienced

acidification and high pCO2 throughout the water column, whereasthese conditions were confined to bottom waters and were lessintense to the east (Fig. 5). While surface waters of LIS were satu-rated with respect to aragonite during July, nearly all of LIS wasundersaturated during early August (Uaragonite < 1; Fig. 5). Pro-gressing into the fall, Uaragonite gradually improved in most of LIS(Fig. 5). While LIS was well oxygenated by October of 2013(>7 mg L�1), bottom waters in the far west retained clear signs ofacidification (pH < 7.7; pCO2 > 1000 matm; Uaragonite < 1; Fig. 5).

Like many river fed estuaries, Narragansett Bay exhibits astronger vertical salinity gradient than LIS, particularly within itsnorthern extent. Regardless, the two systems displayed similarhorizontal and vertical gradients of DO, pH, pCO2, and Uaragoniteduring summer months. For example, in June much of the watercolumn within the northern, lower salinity region of the Bay hadlow DO (<4 mg L�1), low pHT (<7.5) and high pCO2 (>1000 matm),

Fig. 4. Vertical section plots of DO (measured via a SBE 43 DO sensor), pHT (measuredvia a SeaFET pH sensor), pCO2 (measured via a HydroC CO2 sensor), and Uaragonite (bothcalculated from discrete measurements of DIC and pH via a Durafet III sensor) withtemperature in �C (measured via a SBE 9plus CTD) contour lines during August of 2012in Long Island Sound. Vertical lines indicate CTD profiles. Depth is 0e40 m.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e136

resulting in water that was undersaturated with respect to arago-nite (Uaragonite < 1; Fig. 6). While oxygen levels within the southernextent of the Bay were high through nearly all of the water columnin June (>7 mg L�1), bottomwaters across the entire bay displayedsigns of acidification with pHT < 7.7, pCO2 >1000 matm, and Uar-

agonite < 1.5 (Fig. 6). During July and August, this acidificationpattern persisted in Narragansett Bay with values of pHT, pCO2, andUaragonite being similar to June, whereas DO levels across thesouthern extent of Bay declined in July (<7 mg L�1), and again inAugust (<6 mg L�1; Fig. 6).

Jamaica Bay is a lagoonal estuary that historically experienceshypoxia during summer months within its northern and easternchannels. During August of 2012, vertical and horizontal mea-surements within these regions revealed that, in addition to beinghypoxic (<3 mg L�1), much of the water columnwithin Jamaica Baywas acidified (pHT < 7.4) and had levels of pCO2 exceeding2000 matm with bottom waters displaying extremes in these con-ditions (DO < 2 mg L�1, pHT < 7.2, pCO2 > 3000 matm; Fig. 7). InSeptember, much of water column re-oxygenated to >7 mg L�1

with the exception of eastern bottom waters (<6 mg L�1; Fig. 7).The water column was slower to recover from acidification, how-ever, as September pCO2 levels ranged from 700 to 1300 matmwhilepHT levels were generally <7.7 (Fig. 7). While the water columnwas

fully oxygenated by November following Hurricane Sandy, thesystem retained slightly reduced pHT (<7.9) and mildly elevatedpCO2 levels (>700 matm; Fig. 7).

Horizontal mapping of surface waters within lagoons receivinglarge volumes of wastewater discharge further linked DO and pCO2levels in estuarine waters. For example, during a cruise across Ja-maica Bay that began within its northeastern extent and endedwithin the Atlantic Ocean inlet, levels of pCO2 steadily decreasedfrom>4000 matme~500 matm, while DO levels increased in parallelfrom <2 mg L�1 to 7 mg L�1 (Fig. 8A and B). There were strongcorrelations between concentrations of pCO2 and salinity(r2 ¼ 0.92; Fig. 8C) as well as between pCO2 and DO (r2 ¼ 0.95;Fig. 8D). Finally, mapping of Hempstead Bay surface waters furtherdemonstrated the strong influence sewage treatment plant outfallshave on pCO2 levels within an estuary. While nearly all of Hemp-stead Bay, from the Atlantic Ocean through the northern extent ofthe system, had levels of pCO2 that ranged from 450 to 700 matm,the 3 km of the cruise track to both the east and west of the BayPark sewage treatment plant outfall pipe had higher pCO2 levelsthat ranged from 700 to 1200 matm (Fig. 9). There was not a rela-tionship between pCO2 and salinity in this system. Because all ofthese measurements were made during daylight and thus periodsof photosynthetic activity, they likely represent minimum esti-mates of surface water pCO2.

4. Discussion

This study revealed that acidification is an annual feature ofeutrophic estuaries across the Northeast US that co-occurs withseasonally low oxygen. The spatial and temporal dynamics of DO,pH, pCO2, and Uaragonite suggest that they are all ultimately drivenby the same processes, namely high rates of microbial respirationfueled by nutrient enriched algal organic matter coupled withwater column stratification. The degree of acidification observed inthese systems during summer are within ranges that have beenshown to adversely impact a wide range of marine life (Ries et al.,2009; Kroeker et al., 2010; Harvey et al., 2013; Kroeker et al.,2013). As such, this study provides new insight regarding theextent of coastal acidification as well as evidence that it representsa significant, but previously underappreciated environmentalthreat.

4.1. The co-evolution of hypoxia and acidification in estuaries

The acidification of coastal systems is a seasonal phenomenonthat displays a high degree of temporal and spatial coherencewith DO levels (Cai et al., 2011; Sunda and Cai, 2012; Melzneret al., 2013; this study). It has been well established that eutro-phic coastal zones that receive large amounts of organic matterloading are vulnerable to hypoxia during warmer months as rapidmicrobial respiration rates consume oxygen faster than it isreplenished by horizontal or vertical ventilation (Rabalais et al.,2002). The northeast US has been shown to host some of thelargest nutrient loading rates in the world that promote organicmatter production by primary producers as well as rapid rates ofrespiration by heterotrophs (Anderson and Taylor, 2001; Howarth,2008). Along with oxygen consumption microbes release meta-bolic CO2, which in turn reacts with the seawater carbonate sys-tem causing acidification or a lowering of seawater pH. Allsystems presented here had pCO2 > 1000 matm during summermonths, while regions exposed to extreme eutrophication (e.g.>billion liters sewage effluent per day; LIS, Jamaica Bay) experi-enced >3000 matm pCO2 and acidic seawater (pH < 7.0). Summerchanges in temperature account for <5% of these pCO2 increases.Such pCO2 conditions are not expected to occur in the open ocean

Fig. 5. Monthly vertical section plots of DO (measured via a YSI 6450 optical DO sensor), pHT (measured via a Durafet III pH sensor), pCO2 and UAragonite (both calculated fromdiscrete measurements of DIC and pH via a Durafet III sensor) with temperature (�C, measured via a YSI 6920 V2) contour lines during July, early August, late August, September, andOctober of 2013 in Long Island Sound. Vertical lines represent CTD profiles. Depth is 0e40 m.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 7

within the next hundred years, if ever, given projected anthro-pogenic CO2 increases from fossil fuel emissions (I.P.C.C. 2007).We hypothesize that the present acidification of coastal systems isa function of the intensity of nutrient loading rates. The linksbetween acidification, hypoxia, and eutrophication were well-illustrated within LIS where regions near New York Cityreceiving the heaviest loads of nutrients, experiencing the highestnutrient levels (Supplementary Fig. 1), and hosting the mostintense algal blooms (this study; Gobler et al., 2006) also

Fig. 6. Monthly vertical section plots of DO (measured via a YSI 6450 optical DO sensor),discrete measurements of DIC and pH via a Durafet III sensor) with salinity (measured via a Yindicate CTD profiles. Depth is 0e15 m.

exhibited the lowest pH and DO levels during summer.Conversely, the eastern portions of this system have lowernutrient levels (Supplementary Fig. 1) and algal biomass (Fig. 3)and retained higher pH and DO levels through the year, even inbottom waters (pHNBS > 7.7; DO > 5 mg L�1). Similarly, withinNarragansett Bay and Jamaica Bay, the regions of the estuarieswith the most intense algal blooms and highest nutrient con-centrations (Supplementary Figs. 2 and 3) were the same regionsthat experienced the lowest levels of pH and DO during our study.

pHT (measured via a Durafet III pH sensor), pCO2 and UAragonite (both calculated fromSI 6920 V2) contour lines during the summer of 2013 in Narragansett Bay. Vertical lines

Fig. 7. Monthly vertical section plots of DO (measured via a YSI 6450 optical DO sensor), pHT (measured via a SeaFET pH sensor) and pCO2 (measured via a HydroC CO2 sensor)during August, September, and November of 2012 in Jamaica Bay. During the August and November cruises temperature stratification was not observed (25 �C and 6 �C,respectively) and contours are not included whereas September contour lines are temperature (�C, measured via a YSI 6920 V2). Vertical lines indicate CTD profiles. Depth is0e15 m.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e138

At a fundamental level eutrophication is commonly associatedwith enhanced nutrient loading that stimulates phytoplanktonblooms (Nixon 1995) and, via photosynthesis, should consume CO2and increase pH levels. Indeed, our analysis of annual pH cycles inLIS demonstrated that the winterespring phytoplankton bloomperiod was associated with highly basic pH values (>8.5; Fig. 2).These peak pH values, usually during February in LIS and othersystems (Baumann et al., 2014), are followed by a progressivedecline in pH bymore than one unit to < 7.5 during summer and fallas microbial respiration rates intensify. Therefore, while eutrophi-cation does promote algal blooms that can raise the pH and DO ofsurface waters in winter, the remineralization of this algal organicmatter during summer and fall leads to the seasonal acidification

Fig. 8. Horizontal cruise tracks across Jamaica Bay during early November, 2011 of A) surfaceDO sensor) with associated correlations between C) salinity & pCO2 and D) DO & pCO2.

detected in all four study sites, particularly in bottom waters.Furthermore, in stratified, deeper systems (LIS, Narragansett Bay),there can be a spatial decoupling of productivity, hypoxia, andacidification with high productivity confined to surface layers, andlow DO and pH found extensively across bottom waters. In othercases, however, strong summer hypoxia and acidification existed inthe presence of elevated levels of algal biomass (e.g. Fig. 3, westernLIS in September 2011; Fig. 7, Jamaica Bay in August 2012).

While comprehensive surveys of pH and pCO2 within estuariesare not common, some preliminary hypotheses can be developedregarding the types of systems that may be the most prone toacidification based on this and prior studies. Measurements ofecosystem metabolism within 42 US estuaries demonstrated that

pCO2 (measured via a HydroC CO2 sensor) and B) DO (measured via a YSI 6450 optical

Fig. 9. Surface pCO2 levels in Hempstead Bay as measured via a HydroC CO2 sensorduring a cruise conducted in late October, 2011. Levels of pCO2 > 700 matm are withinclose proximity (3 km) of a sewage treatment plant outfall.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 9

93% were net heterotrophic, experiencing an annual net con-sumption of DO (Caffrey, 2004) and thus production of pCO2 andlikely some degree of seasonal acidification. Among the 20 saltmarsh-dominated estuaries surveyed by Caffrey (2004), all werenet heterotrophic. Salt marshes host extreme levels of microbialmetabolism (Dame et al., 1986; Koch and Gobler, 2009) and the twolagoons studied here (Jamaica and Hempstead Bay) have significantstands of salt marsh. As such, beyond eutrophication-driven acid-ification, salt marshes may have also promoted the high pCO2conditions in these shallow systems. Our recent 5-yr study ofanother NY salt marsh (Flax Pond) found a highly significant cor-relation between DO and pH and extreme acidification in summer,particularly during low tides and at night (pCO2 >4000 matm;Baumann et al., 2014).

Conversely, high rates of autotrophy make seagrass meadows asignificant sink for carbon (Duarte et al., 2010). Accordingly,seagrass-dominated estuaries, which must be shallow to host suchbenthic vegetation, are often net autotrophic (Caffrey, 2004) andare likely to buffer against ocean acidification in the water column(Hendriks et al., 2014). Beyond the photosynthetic activity of thesubmerged aquatic vegetation, shallow estuaries are also morelikely to bewell-mixed, rapidly equilibrating with atmospheric CO2,and thus may be generally less vulnerable to continued respiratoryacidification than deeper systems (such as LIS) that are typicallystratified in summer and have a smaller fraction of the water col-umn within the euphotic zone. These shallow systems may, how-ever, be vulnerable to large diurnal variations in pH and DOparticularly in cases of excessive loadings of organic matter(Baumann et al., 2014). Shallow estuaries are more responsive tonutrient loading and often shift from net autotrophic to net het-erotrophic on diel timescales with the intensity of these shiftslinked to the intensity of nutrient loading (O'Boyle et al., 2013).Given that all of the surveys presented here occurred during theday, the extent of acidification in some of these systems is likelyeven more extreme at night, particularly within shallow regions.Finally, extended residence times within estuaries are likely toexacerbate acidification as poorly flushedwaters will bemore likelyto accumulate pCO2 during periods of intense organic matterremineralization.

Coastal acidification in the US has been most commonlyexamined along the Pacific coast where deep, pCO2-enriched wa-ters are seasonally upwelled (Feely et al., 2008, 2010; Hofmannet al., 2011) and negatively impact oyster populations in

hatcheries (Barton et al., 2012) and in the wild (Hettinger et al.,2013). In contrast, the enclosed, eutrophic estuaries, withextended residence times and strong summer hypoxia such asthose studied here are similar to other systems across the US EastCoast (e.g. Chesapeake Bay; Kemp et al., 2005), in Europe (e.g. BalticSea; Melzner et al., 2013; Sunda and Cai, 2012), and southeast Asia(e.g. Yangtze Estuary; Chen et al., 2007). Given the widespreadnature of hypoxia around the globe (Diaz and Rosenberg, 2008), itwould seem that eutrophication-induced acidification maycurrently be a more common phenomenon than the upwelling-induced acidification that occurs on the US West Coast. Clearly,more studies of coastal ocean acidification will be required toevaluate this hypothesis. Finally, the acidification detected duringthis study is likely confined to estuaries, as recent large scale sur-veys of the US Gulf and East Coasts determined that well-mixedshelf waters from Texas to Maine were supersaturated withrespect to aragonite during summer (Wang et al., 2013). That samestudy also found, however, that the shelf waters of the northeast UShad substantially lower pH and buffering capacity than all otherregions surveyed and inferred that this region is more susceptibleto acidification than other parts of the US Gulf and East coasts(Wang et al., 2013), a hypothesis consistent with the finding of thepresent study.

While hypoxia and acidification co-occurred in estuaries duringsummer, some observations suggest low pH conditions may persistlonger into the fall than low oxygen. For example, during 2011 and2013 in LIS and during 2012 in Jamaica Bay, waters transitionedfrom hypoxic to normoxic during early fall (October) but main-tained pHT levels < 7.7. These differences may be a function of thedifferential diffusion and solubility of O2 and CO2 as a function oftemperature (Millero et al., 2006). Cooler fall temperatures enhancegas solubility, and may contribute toward a more extended periodof acidification during fall compared to hypoxia, even in the face ofslowing microbial respiration. In the case of DO, cooler tempera-tures will lead to a diffusion of oxygen into water whereas in thecase of CO2, cooler temperatures will slow the diffusion out ofwater. An additional factor that likely enhances acid production(lower pH) during times when DO levels are increasing is theoxidation of anaerobic metabolites. The reduced constituents (e.g.,NH4

þ, HS�, Fe2þ, Mn2þ) that build up in surface sediments duringhypoxia oxidize seasonally when systems re-oxygenate (Soetaertet al., 2007). These oxidation reactions produce strong acids thattitrate alkalinity, lower pH, and could promote shell dissolution inbivalves (Green and Aller 1998). As such, the seasonal duration ofacidification as a selective pressure in estuarine organisms may belonger than the seasonal pressures imposed by hypoxia.

Beyond microbial respiration driven by internal, autochtonoussources of organic carbon, this study provides evidence that thedischarge of wastewater from sewage treatment plants is a previ-ously unappreciated, point source of acidification and high pCO2water in coastal zones. While these low pH conditions may bepartly a function of microbial respiration of the allochthonousorganic matter discharged from sewagewithin the estuary, it is alsolikely the discharged sewage itself is acidified. This was mostobvious during our horizontal cruises across the lagoons in south-west Long Island where a radius of several kilometers of estuarinesurface waters around sewage outfall sites had levels of pCO2 nearor above 1000 matm (several times supersaturated). The surfacewaters of western LIS are highly enriched in wastewater (Sweeneyand Sanudo-Wilhelmy, 2004) and had surface waters with nearlythe same low pH and high pCO2 as bottom waters during summerand fall. Due to its high organic matter load, wastewater experi-ences extreme amounts of microbial respiration and has beenpreviously shown to be low in pH (6e8) and DO (at times anoxic)and enriched in pCO2 (Gallert and Winter, 2005). A management

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e1310

goal for LIS (USEPA, 1994) and other coastal systems is to remove Nfrom wastewater to relieve the symptoms of eutrophicationincluding hypoxia (Nixon and Buckley, 2002). Given that theremoval of N is generally achieved via increased microbial activity(i.e. biological denitrification), this enhanced level of N removalmay yield effluent with lower pH and high pCO2 levels (Soetaertet al., 2007). As such, it is possible that coastal regions in closeproximity to wastewater discharge may experience more intenseacidification as treatment plants are upgradedwith higher amountsof denitrification to alleviate other symptoms of eutrophication.Given the chronically high concentrations of pCO2 in regions withinclose proximity to sewage treatment plant discharge (>1000 matm),it seems likely that these sites may have a localized impact on thesurrounding flora and fauna in a manner that parallels otherconcentrated sources of pCO2 including volcanic seeps (Fabriciuset al., 2011).

The connection between eutrophication and acidification wasfurther evidenced by the relationship between DO and pH for thethree major systems studied here: LIS, Narragansett Bay, and Ja-maica Bay (Fig. 10). For all of the sites, there was a highly significant,linear relationship between DO and pH (p < 0.0001), with JamaicaBay displaying the strongest relationship (R ¼ 0.94) and LISshowing more variance (R ¼ 0.70). The correlation for Jamaica Baywas slightly improved with a logarithmic fit (R ¼ 0.94). This vari-ability in LIS was partly a function of the depth of this system asduring summer, stratified conditions, DO levels were largely uni-form below themixed layer, while pH levels continued to decline tothe bottom. General linear regression models of the DO-pH rela-tionship for each system provided y-intercept values for JamaicaBay, LIS, and Narragansett Bay of 7.07, 7.14, and 7.25, respectively,representing the levels of pHT predicted under anoxic conditions(DO ¼ 0; Fig. 10). These values further emphasize linkages betweeneutrophication and acidification. Jamaica Bay receives ~90% of its Nload from sewage discharge (Benotti et al., 2007) and is predicted toexperience the most extreme acidification under anoxic conditions(pHT ¼ 7.07). In contrast, Narragansett Bay experiences excessivenutrient loading within its northern extent, but is a more oligo-trophic system to the south (Nixon et al., 2008), and was predictedto experience the highest pHT under anoxic conditions (7.25). LISwas similar to Jamaica Bay (anoxic pHT ¼ 7.14), but was not asextreme, again likely reflecting the more hybrid nature of this

Fig. 10. All paired measurements (n ¼ 6652) of surface pHT and DO from LIS, JamaicaBay, and Narragansett Bay. General linear regression models for each site yieldedequations and correlation coefficients of y ¼ 0.084x þ 7.06, R ¼ 0.94 for Jamaica Bay,y ¼ 0.081x þ 7.25, R ¼ 0.84 for Narragansett Bay, and y ¼ 0.085x þ 7.14, R ¼ 0.70 for LIS.

system: While its western extreme receives large nutrient inputsfrom NYC (Parker and O’Reilly, 1991; O’Shea and Bronsan, 2000)and can be acidic (pH < 7), the eastern end exchanges with theAtlantic Ocean and is fairly oligotrophic (Gobler et al., 2006).

4.2. Biological implications

This study establishes that estuarine pelagic organisms intemperate zones are commonly exposed to acidification and watersundersaturated with regard to aragonite during summer and fallmonths. Consistent with our study, Feely et al. (2010) reported thatmost of Puget Sound, WA, USA, was undersaturated with regard toaragonite during August. The summer levels of pH, pCO2, andcalculated Uaragonite present through large areas within the estu-aries studied here during summer (pH < 7.7T, pCO2 > 1000 matm,Uaragonite < 1) are within the range that have been shown in labo-ratory studies to reduce the growth and survival of early life stagemollusks (Gazeau et al., 2013) and fish (Baumann et al., 2012;Frommel et al., 2012; Chambers et al., 2013). Furthermore, theseconditions emerge at the same times that many mollusks and fishspawn in estuaries (Kennedy and Krantz, 1982; Bricelj et al., 1987;Able and Fahay, 1998; Kraeuter and Castagna, 2001), suggestingthat some early life stage fish andmollusks sensitive to acidificationmay be negatively impacted by high levels of pCO2 and low levels ofpH and carbonate. Therefore, we suggest that acidification shouldbe considered among the factors that shape fisheries yields incoastal zones. A vivid example of this may be the bay scallop(Argopecten irradians) which is known to spawn in mid-to-latesummer (July, August) and is one of the organisms most sensitiveto acidification (Talmage and Gobler, 2009, 2010, 2011), experi-encing significant declines in survival when exposed topCO2 � 750 matm for only four days (Gobler and Talmage, 2013).While the collapse of this fishery in NY waters was caused by toxicalgae blooms decades ago (Gobler et al., 2005), the failure of thispopulation to recover despite significant restoration efforts may becaused, at least in part, by the overlap of annual spawning eventswith acidified conditions in estuaries. We note that during this andprior studies (Baumann et al., 2014), there has been significantinterannual variation in the intensity and timing of acidificationwith warmer spring and summers being associated with earlieronset and more intense acidification (e.g. during 2012; this study;Baumann et al., 2014). If the impaired traits are indeed important tothe fitness of commercially viable species, then such interannualchanges in the timing and intensity of acidification may influencethe annual yields of some fisheries such as bay scallops.While thereis great variability in biological response to acidification amongspecies (Kroeker et al., 2010, 2013) the ecosystem level effects willdepend on the extent to which sensitive species or functionalgroups govern critical ecological processes (e.g filter feeding bycalcifying bivalves).

The effects of acidification on marine life will partly be a func-tion of differences between exposure levels for benthic and pelagicorganisms as well as the differential susceptibility and de-mographic importance of various life stages. During this study,bottom waters were always more acidified than surface waterssuggesting benthic organisms are subject to more extended periodsof dissolution (Green and Aller, 1998) and need to be more tolerantof acidification than pelagic species (Waldbusser and Salisbury2014). Regarding bivalves, larval stages which are generallypelagic are known to be more sensitive to acidification thanbenthic, juvenile stages (Talmage and Gobler, 2011; Gazeau et al.,2013). While larval stages are spared intense levels of acidifica-tion near or in the benthos (Green et al., 1998), in highly eutrophicsystems during summer (e.g. western LIS and Jamaica Bay), theentire water column was undersaturated with regard to aragonite,

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 11

suggesting even pelagic species and life stages will be exposed toconditions unfavorable for calcification.

We commonly observed pCO2 > 1000 matm in bottom watersduring summer months, with the most eutrophic estuarine regionsexhibiting >3000 matm pCO2 and acidic seawater (pH < 7.0). Theseobservations demonstrate that the recommendation to restrictocean acidification experiments with marine organisms to pCO2levels of 1000 matm and below to mimic future climate change(Riebesell et al., 2010) should be heeded for open ocean speciesonly. Indeed, given that estuarine organisms persisting withindeeper regions of the systems studied here may rarely experiencepCO2 below 1000 matm during summer months, experimentsdesigned to assess the realistic effects of current and future acidi-fication on coastal species will need to target pCO2 levels signifi-cantly beyond 1000 matm.

In addition to acidification, this study further demonstrates thewell-known tenet that estuarine organisms are often challenged bythe stress of hypoxia during summer (Rabalais et al., 2002; Diaz andRosenberg, 2008). Furthermore, in some cases, estuarine organismsalready existing near their upper temperature threshold mayexperience concurrent thermal stress (P€ortner 2008, 2010). Hence,it would seem the ‘hot, sour, and breathless’ (high temperature, lowpH, lowDO) conditions predicted for the future open ocean (Gruberet al., 2011) can already be found in today's coastal zones duringsummer, and especially within the benthoswhere pH and DO levelsare generally lower than the water column (Zhu et al., 2006). Priorresearch has demonstrated that each of these stressors individuallycreates a significant physiological challenge to marine organismsand that their co-occurrence can have complex, interactive effectson their performance (P€ortner 2008, 2010). Several recent studieshave examined the manner in which elevated temperatures andocean acidification impact ocean organisms (Talmage and Gobler,2011; Hiebenthal et al., 2013; Byrne and Przeslawski, 2013). Incontrast, the concurrent effects of low oxygen and acidification onmarine animals are largely unknown, as most prior studies ofhypoxia and thermal stress have not considered concurrent low pHlevels. Recently, it has been discovered that hypoxia and acidifica-tion can have additive and synergistic negative effects on thegrowth, survival, and metamorphosis of early life stage bivalves(Gobler et al., 2014). Given this finding and the frequency of low pHand low DO waters in estuaries (Cai et al., 2011; Sunda and Cai,2012; Melzner et al., 2013; this study), a comprehensive assess-ment of the effects of hypoxia and acidification on marine life isneeded to understand how coastal ecosystems will respond tothese conditions both today and under future climate changescenarios.

4.3. Management implications

The revelation that water column acidification can be intenseand widespread in estuaries may have important managementimplications. In light of the threat of hypoxia to fisheries andbiodiversity (Gray et al., 2002; Levin et al., 2009), coastal man-agement agencies often set criteria for dissolved oxygen based onlevels known to be harmful to estuarine organisms (Vaquer-Sunyerand Duarte, 2008; USEPA 2000). Given that the levels of pHdetected during this study have been shown to be harmful toseveral forms of marine life (Talmage and Gobler, 2009; Baumannet al., 2012; Gazaeu et al., 2013) and that concurrent acidificationand low oxygen can synergistically depress survival rates of bi-valves (Gobler et al., 2014), managerial criteria based strictly on DOmay not protect estuarine animals as anticipated. While signifi-cantly more research is needed to better understand the co-effectsof hypoxia and acidification on estuarine organisms, future envi-ronmental regulations developed to protect coastal organisms in

regions prone to hypoxia should consider the concurrent effects ofacidification on these animals. Further, nutrient management plansin acidified estuaries may consider the level of nutrient loadreduction required to alleviate low pH conditions and the associ-ated impacts on marine life. Finally, given that rising atmosphericCO2 levels will further depress estuarine pH levels (Miller et al.,2009), the importance of management efforts that addresscoastal acidification will continue to increase through this century.

Acknowledgments

This research was supported by NOAA's Ocean AcidificationProgram through award #NA12NOS4780148 from the NationalCenters for Coastal Ocean Science, the National Science Foundation(NSF # 1129622), and the Chicago Community Trust.

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.ecss.2014.05.027.

References

Able, K.W., Fahay, M.P., 1998. The First Year in the Life of Estuarine Fishes in theMiddle Atlantic Bight. Rutgers University Press.

Anderson, T.H., Taylor, G.T., 2001. Nutrient pulses, plankton blooms, and seasonalhypoxia in western Long Island Sound. Estuaries 24, 228e243.

Barton, A., Hales, B., Waldbusser, G.G., Langdon, C., Feely, R.A., 2012. The Pacificoyster, Crassostrea gigas, shows negative correlation to naturally elevated car-bon dioxide levels: Implications for near-term ocean acidification effects.Limnol. Oceanogr. 57, 698e710.

Baumann, H., Talmage, S.C., Gobler, C.J., 2012. Reduced early life growth and survivalin a fish in direct response to increased carbon dioxide. Nat. Clim. Chang. 2,38e41.

Baumann, H., Wallace, R., Tagliaferri, T., Gobler, C., 2014. Large natural pH, CO2 andO2 fluctuations in a temperate tidal salt marsh on diel, seasonal, and interan-nual time scales. Estuar. Coasts. http://dx.doi.org/10.1007/s12237-12014-19800-y (in press).

Benotti, M.J., Abbene, M., Terracciano, S.A., 2007. Nitrogen Loading in Jamaica Bay,Long Island, New York: Predevelopment to 2005, Scientific InvestigationsReport. USGS.

Bricelj, V.M., Epp, J., Malouf, R.E., 1987. Intraspecific variation in reproductive andsomatic growth cycles of bay scallops Argopecten irradians. Mar. Ecol. Prog. Ser.36, 123e137.

Byrne, M., Przeslawski, R., 2013. Multistressor Impacts of warming and acidificationof the ocean on Marine Invertebrates' life Histories. Integr. Comp. Biol. 53,582e596.

Caffrey, J.M., 2004. Factors controlling net ecosystem metabolism in US estuaries.Estuaries 27, 90e101.

Cai, W.J., Hu, X.P., Huang, W.J., Murrell, M.C., Lehrter, J.C., Lohrenz, S.E., Chou, W.C.,Zhai, W.D., Hollibaugh, J.T., Wang, Y.C., Zhao, P.S., Guo, X.H., Gundersen, K.,Dai, M.H., Gong, G.C., 2011. Acidification of subsurface coastal waters enhancedby eutrophication. Nat. Geosci. 4, 766e770.

Caldeira, K., Wickett, M.E., 2003. Oceanography: anthropogenic carbon and oceanpH. Nature 425, 365.

Chambers, R.C., Candelmo, A.C., Habeck, E.A., Poach, M.E., Wieczorek, D.,Cooper, K.R., Greenfield, C.E., Phelan, B.A., 2013. Ocean acidification effects inthe early life-stages of summer flounder, Paralichthys dentatus. Biogeosci.Discuss. 10, 13897e13929.

Chen, C.C., Gong, G.C., Shiah, F.K., 2007. Hypoxia in the East China Sea: one of thelargest coastal low-oxygen areas in the world. Mar. Environ. Res. 64, 399e408.

Cloern, J.E., 2001. Our evolving conceptual model of the coastal eutrophicationproblem. Mar. Ecol. Prog. Ser. 210, 223e253.

Costanza, R., dArge, R., deGroot, R., Farber, S., Grasso, M., Hannon, B., Limburg, K.,Naeem, S., Oneill, R.V., Paruelo, J., Raskin, R.G., Sutton, P., vandenBelt, M., 1997.The value of the world's ecosystem services and natural capital. Nature 387,253e260.

Dame, R., Chrzanowski, T., Bildstein, K., Kjerfve, B., McKellar, H., Nelson, D.,Spurrier, J., Stancyk, S., Stevenson, H., Vernberg, J., Zingmark, R., 1986. Theoutwelling hypothesis and North Inlet, South Carolina. Mar. Ecol. Prog. Ser. 33,217e229.

de Jonge, V.N., Elliott, M., Orive, E., 2002. Causes, historical development, effects andfuture challenges of a common environmental problem: eutrophication.Hydrobiologia 475, 1e19.

Diaz, R.J., Rosenberg, R., 2008. Spreading dead zones and consequences for marineecosystems. Science 321, 926e929.

Dickson, A.G., Sabine, C.L., Christian, J.R., 2007. Guide to best practices for ocean CO2measurements. PICES Spec. Publ. 3, 191.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e1312

Doney, S.C., Fabry, V.J., Feely, R.A., Kleypas, J.A., 2009. Ocean acidification: the otherCO2 problem. Annu. Rev. Mar. Sci., 169e192. Annual Reviews, Palo Alto.

Duarte, C.M., Hendriks, I.E., Moore, T.S., Olsen, Y.S., Steckbauer, A., Ramajo, L.,Carstensen, J., Trotter, J.A., McCulloch, M., 2013. Is ocean acidification an open-ocean syndrome? Understanding anthropogenic impacts on seawater pH.Estua. Coasts 36, 221e236.

Duarte, C.M., Marba, N., Gacia, E., Fourqurean, J.W., Beggins, J., Barron, C.,Apostolaki, E.T., 2010. Seagrass community metabolism: assessing the carbonsink capacity of seagrass meadows. Glob. Biogeochem. Cycle 24, 8.

Fabricius, K.E., Langdon, C., Uthicke, S., Humphrey, C., Noonan, S., De'ath, G.,Okazaki, R., Muehllehner, N., Glas, M.S., Lough, J.M., 2011. Losers and winners incoral reefs acclimatized to elevated carbon dioxide concentrations. Nat. Clim.Chang. 1, 165e169.

Feely, R.A., Alin, S.R., Newton, J., Sabine, C.L., Warner, M., Devol, A., Krembs, C.,Maloy, C., 2010. The combined effects of ocean acidification, mixing, andrespiration on pH and carbonate saturation in an urbanized estuary. Estuar.Coast. Shelf Sci. 88, 442e449.

Feely, R.A., Sabine, C.L., Hernandez-Ayon, J.M., Ianson, D., Hales, B., 2008. Evidencefor upwelling of corrosive “acidified” water onto the continental shelf. Science320, 1490e1492.

Fiedler, B., Fietzek, P., Vieira, N., Silva, P., Bittig, H.C., K€ortzinger, A., 2012. In situ CO2and O2 measurements on a profiling Float. J. Atmos. Ocean. Technol. 30,112e126.

Fietzek, P., Fiedler, B., Steinhoff, T., K€ortzinger, A., 2014. In situ quality assessment ofa novel underwater pCO2 sensor based on membrane equilibration and NDIRspectrometry. J. Atmos. Ocean. Technol. 31, 181e196.

Frommel, A.Y., Maneja, R., Lowe, D., Malzahn, A.M., Geffen, A.J., Folkvord, A.,Piatkowski, U., Reusch, T.B.H., Clemmesen, C., 2012. Severe tissue damage inAtlantic cod larvae under increasing ocean acidification. Nat. Clim. Chang. 2,42e46.

Gallert, C., Winter, J., 2005. Bacterial metabolism in wastewater treatment systems.In: Winter, J., Jordening, H. (Eds.), Environmental Biotechnology. Wiley-VCH,p. 488.

Gazeau, F., Parker, L.M., Comeau, S., Gattuso, J.P., O'Connor, W.A., Martin, S.,Portner, H.O., Ross, P.M., 2013. Impacts of ocean acidification on marine shelledmolluscs. Mar. Biol. 160, 2207e2245.

Gobler, C.J., Buck, N.J., Sieracki, M.E., Sanudo-Wilhelmy, S.A., 2006. Nitrogen andsilicon limitation of phytoplankton communities across an urban estuary: theEast River-Long Island Sound system. Estuar. Coast. Shelf Sci. 68, 127e138.

Gobler, C., DePasquale, E., Griffith, A., Baumann, H., 2014. Hypoxia and acidificationhave additive and synergistic negative effects on the growth, survival, andmetamorphosis of early life stage bivalves. PLoS One 9.

Gobler, C.J., Lonsdale, D.J., Boyer, G.L., 2005. A review of the causes, effects, andpotential management of harmful brown tide blooms caused by Aureococcusanophagefferens (Hargraves et Sieburth). Estuaries 28, 726e749.

Gobler, C.J., Talmage, S.C., 2013. Short- and long-term consequences of larval stageexposure to constantly and ephemerally elevated carbon dioxide for marinebivalve populations. Biogeosciences 10, 2241e2253.

Gray, J.S., Wu, R.S.S., Or, Y.Y., 2002. Effects of hypoxia and organic enrichment on thecoastal marine environment. Mar. Ecol. Prog. Ser. 238, 249e279.

Green, M.A., Aller, R.C., 1998. Seasonal patterns of carbonate diagenesis in nearshoreterrigenous muds: relation to spring phytoplankton bloom and temperature. J.Mar. Res. 56, 1097e1123.

Green, M.A., Aller, R.C., Aller, J.Y., 1998. Influence of carbonate dissolution on sur-vival of shell-bearing meiobenthos in nearshore sediments. Limnol. Oceanogr.43, 18e28.

Gruber, N., 2011. Warming up, turning sour, losing breath: ocean biogeochemistryunder global change. Philos. Trans. R. Soc. A-Math. Phys. Eng. Sci. 369,1980e1996.

Harvey, B.P., Gwynn-Jones, D., Moore, P.J., 2013. Meta-analysis reveals complexmarine biological responses to the interactive effects of ocean acidification andwarming. Ecol. Evol. 3, 1016e1030.

Heisler, J., Glibert, P.M., Burkholder, J.M., Anderson, D.M., Cochlan, W.,Dennison, W.C., Dortch, Q., Gobler, C.J., Heil, C.A., Humphries, E., Lewitus, A.,Magnien, R., Marshall, H.G., Sellner, K., Stockwell, D.A., Stoecker, D.K.,Suddleson, M., 2008. Eutrophication and harmful algal blooms: a scientificconsensus. Harmful Algae 8, 3e13.

Hendriks, I.E., Olsen, Y.S., Ramajo, L., Basso, L., Steckbauer, A., Moore, T.S., Howard, J.,Duarte, C.M., 2014. Photosynthetic activity buffers ocean acidification in sea-grass meadows. Biogeosciences 11, 333e346.

Hettinger, A., Sanford, E., Hill, T.M., Lenz, E.A., Russell, A.D., Gaylord, B., 2013. Larvalcarry-over effects from ocean acidification persist in the natural environment.Glob. Change Biol. 19, 3317e3326.

Hiebenthal, C., Philipp, E.E.R., Eisenhauer, A., Wahl, M., 2013. Effects of seawaterpCO(2) and temperature on shell growth, shell stability, condition and cellularstress of Western Baltic SeaMytilus edulis (L.) and Arctica islandica (L.). Mar. Biol.160, 2073e2087.

Hofmann, G.E., Smith, J.E., Johnson, K.S., Send, U., Levin, L.A., Micheli, F., Paytan, A.,Price, N.N., Peterson, B., Takeshita, Y., Matson, P.G., Crook, E.D., Kroeker, K.J.,Gambi, M.C., Rivest, E.B., Frieder, C.A., Yu, P.C., Martz, T.R., 2011. High-frequencydynamics of ocean pH: a multi-ecosystem comparison. PLoS One 6.

Howarth, R.W., 2008. Coastal nitrogen pollution: a review of sources and trendsglobally and regionally. Harmful Algae 8, 14e20.

Kemp, W.M., Boynton, W.R., Adolf, J.E., Boesch, D.F., Boicourt, W.C., Brush, G.,Cornwell, J.C., Fisher, T.R., Glibert, P.M., Hagy, J.D., Harding, L.W., Houde, E.D.,

Kimmel, D.G., Miller, W.D., Newell, R.I.E., Roman, M.R., Smith, E.M.,Stevenson, J.C., 2005. Eutrophication of Chesapeake Bay: historical trends andecological interactions. Mar. Ecol. Prog. Ser. 303, 1e29.

Kennedy, V., Krantz, L., 1982. Comparative gametogenic and spawning patterns ofthe oyster Crassostrea virginica (Gmelin) in central Chesapeake Bay. J. ShellfishRes. 2, 133e140.

Koch, F., Gobler, C.J., 2009. The effects of tidal Export from salt marsh ditches onestuarine water quality and Plankton communities. Estuar. Coasts 32,261e275.

Kraeuter, J.N., Castagna, M., 2001. Biology of the Hard Calm. Elsevier Science.Kroeker, K.J., Kordas, R.L., Crim, R., Hendriks, I.E., Ramajo, L., Singh, G.S., Duarte, C.M.,

Gattuso, J.P., 2013. Impacts of ocean acidification on marine organisms: quan-tifying sensitivities and interaction with warming. Glob. Change Biol. 19,1884e1896.

Kroeker, K.J., Kordas, R.L., Crim, R.N., Singh, G.G., 2010. Meta-analysis revealsnegative yet variable effects of ocean acidification on marine organisms. Ecol.Lett. 13, 1419e1434.

Levin, L.A., Ekau, W., Gooday, A.J., Jorissen, F., Middelburg, J.J., Naqvi, S.W.A.,Neira, C., Rabalais, N.N., Zhang, J., 2009. Effects of natural and human-inducedhypoxia on coastal benthos. Biogeosciences 6, 2063e2098.

Melzner, F., Thomsen, J., Koeve, W., Oschlies, A., Gutowska, M., Bange, H., Hansen, H.,K€ortzinger, A., 2013. Future ocean acidification will be amplified by hypoxia incoastal habitats. Mar. Biol. 160, 1875e1888.

Miller, A.W., Reynolds, A.C., Sobrino, C., Riedel, G.F., 2009. Shellfish face uncertainfuture in high CO2 world: influence of acidification on oyster larvae calcificationand growth in estuaries. PLoS One 4, 8.

Millero, F.J., 2010. Carbonate constants for estuarine waters. Mar. Freshw. Res. 61,139e142.

Millero, F.J., Graham, T.B., Huang, F., Bustos-Serrano, H., Pierrot, D., 2006. Dissoci-ation constants of carbonic acid in seawater as a function of salinity and tem-perature. Mar. Chem. 100, 80e94.

Morse, J.W., Arvidson, R.S., Luttge, A., 2007. Calcium carbonate formation anddissolution. Chem. Rev. 107, 342e381.

Munday, P.L., Dixson, D.L., McCormick, M.I., Meekan, M., Ferrari, M.C.O.,Chivers, D.P., 2010. Replenishment of fish populations is threatened by oceanacidification. Proc. Natl. Acad. Sci. U. S. A. 107, 12930e12934.

Murray, C.S., Malvezzi, A., Gobler, C.J., Baumann, H., 2014. Offspring sensitivity toocean acidification changes seasonally in a coastal marine fish. Mar. Ecol. Prog.Ser. 504, 1e11.

Nixon, S.W., 1995. Coastal marine eutrophication e a definition, social causes, andfuture concerns. Ophelia 41, 199e219.

Nixon, S.W., Buckley, B.A., 2002. “A strikingly rich zone” e nutrient enrichmentand secondary production in coastal marine ecosystems. Estuaries 25,782e796.

Nixon, S., Buckley, B., Granger, S., Harris, L., Oczkowski, A., Fulweiler, R., Cole, L.,2008. Nitrogen and phosphorus inputs to Narragansett Bay: past, present, andfuture. In: Desbonnet, A., Costa-Pierce, B. (Eds.), Science for Ecosystem-basedManagement. Springer, New York, pp. 101e175.

O'Donnell, J., Wilson, R.E., Lwiza, K., Whitney, M., Bohlen, W.F., Codiga, D.,Fribance, D.B., Fake, T., Bowman, M., Varekamp, J., 2014. The physical ocean-ography of Long Island Sound. In: Long Island Sound. Springer, New York,pp. 79e158.

O'Boyle, S., McDermott, G., Noklegaard, T., Wilkes, R., 2013. A simple index of trophicstatus in estuaries and coastal bays based on measurements of pH and dis-solved oxygen. Estuar. Coasts 36, 158e173.

O'Shea, M.L., Brosnan, T.M., 2000. Trends in indicators of eutrophication in WesternLong Island sound and the Hudson-Raritan Estuary. Estuaries 23, 877e901.

Paerl, H.W., 2006. Assessing and managing nutrient-enhanced eutrophication inestuarine and coastal waters: Interactive effects of human and climatic per-turbations. Ecol. Eng. 26, 40e54.

Parker, C.A., O'Reilly, J.E., 1991. Oxygen depletion in Long Island sound: a historicalperspective. Estuaries 14, 248e264.

Parsons, T.R., Maita, Y., Lalli, C.M., 1984. A Manual of Chemical and BiologicalMethods for Seawater Analysis. Pergamon Press, Oxford.

P€ortner, H.O., 2008. Ecosystem effects of ocean acidification in times of oceanwarming: a physiologist's view. Mar. Ecol. Prog. Ser. 373, 203e217.

P€ortner, H.O., 2010. Oxygen- and capacity-limitation of thermal tolerance: a matrixfor integrating climate-related stressor effects in marine ecosystems. J. Exp.Biol. 213, 881e893.

Rabalais, N.N., Turner, R.E., Dortch, Q., Justic, D., Bierman, V.J., Wiseman, W.J., 2002.Nutrient-enhanced productivity in the northern Gulf of Mexico: past, presentand future. Hydrobiologia 475, 39e63.

Riebesell, U., Fabry, V.J., Hansson, L., Gattuso, J.P., 2010. Guide to best practices forocean acidification research and data reporting. Publ. Office Eur. Union 260.

Ries, J.B., Cohen, A.L., McCorkle, D.C., 2009. Marine calcifiers exhibit mixed re-sponses to CO2-induced ocean acidification. Geology 37, 1131e1134.

Salisbury, J.E., Vandemark, D., Hunt, C.W., Campbell, J.W., McGillis, W.R.,McDowell, W.H., 2008. Seasonal observations of surface waters in two Gulf ofMaine estuary-plume systems: relationships between watershed attributes,optical measurements and surface pCO(2). Estuar. Coast. Shelf Sci. 77,245e252.

Scavia, D., Bricker, S., 2006. Coastal eutrophication assessment in the United States.Biogeochemistry 79, 187e208.

Scavia, D., Justic, D., Bierman, V.J., 2004. Reducing hypoxia in the Gulf of Mexico:advice from three models. Estuaries 27, 419e425.

R.B. Wallace et al. / Estuarine, Coastal and Shelf Science 148 (2014) 1e13 13

Soetaert, K., Hofmann, A.F., Middelburg, J.J., Meysman, F.J.R., Greenwood, J.,2007. The effect of biogeochemical processes on pH. Mar. Chem. 105,30e51.

Sunda, W.G., Cai, W.J., 2012. Eutrophication induced CO2-acidification of subsurfacecoastal waters: interactive effects of temperature, salinity, and atmospheric P-CO2. Environ. Sci. Technol. 46, 10651e10659.

Suter, E.A., Lwiza, K.M.M., Rose, J.M., Gobler, C., Taylor, G.T., 2014. Phyto-plankton assemblage changes during decadal decreases in nitrogen load-ings to the urbanized Long Island Sound estuary, USA. Mar. Ecol. Prog. Ser.497, 51e67.

Sweeney, A., Sanudo-Wilhelmy, S.A., 2004. Dissolved metal contamination in theEast River-Long Island sound system: potential biological effects. Mar. Pollut.Bull. 48, 663e670.

Talmage, S.C., Gobler, C.J., 2009. The effects of elevated carbon dioxide concentra-tions on the metamorphosis, size, and survival of larval hard clams (Mercenariamercenaria), bay scallops (Argopecten irradians), and Eastern oysters (Cras-sostrea virginica). Limnol. Oceanogr. 54, 2072e2080.

Talmage, S.C., Gobler, C.J., 2010. Effects of past, present, and future ocean carbondioxide concentrations on the growth and survival of larval shellfish. Proc. Natl.Acad. Sci. U. S. A. 107, 17246e17251.

Talmage, S.C., Gobler, C.J., 2011. Effects of elevated temperature and carbon dioxideon the growth and survival of larvae and juveniles of three species of NorthwestAtlantic bivalves. PLoS One 6.

USEPA, 1994. The Long Island Sound Study: The Comprehensive Conservation andManagement Plan. Washington, D.C.

USEPA, 2000. Ambient Aquatic Life Water Quality Criteria for Dissolved Oxygen(Saltwater): Cape Cod to Cape Hatteras. United States Environmental ProtectionAgency Report EPA-822-R-00-012, Washington DC, USA.

Valiela, I., 2006. Global Coastal Change. Blackwell Publishing, Malden, MA.Vaquer-Sunyer, R., Duarte, C.M., 2008. Thresholds of hypoxia for marine biodiver-

sity. Proc. Natl. Acad. Sci. U. S. A. 105, 15452e15457.Waldbusser, G.G., Salisbury, J.E., 2014. Ocean acidification in the coastal zone from

an organism's perspective: multiple system parameters, frequency domains,and habitats. Annu. Rev. Mar. Sci. 6, 221e247.

Weiss, I.M., Tuross, N., Addadi, L., Weiner, S., 2002. Mollusc larval shell formation:amorphous calcium carbonate is a precursor phase for aragonite. J. Exp. Zool.293, 478e491.

Waldbusser, G.G., Voigt, E.P., Bergschneider, H., Green, M.A., Newell, R.I.E., 2011.Biocalcification in the eastern oyster (Crassostrea virginica) in relation to long-term trends in Chesapeake Bay pH. Estuar. Coasts 34, 221e231.

Wang, Z.H.A., Wanninkhof, R., Cai, W.J., Byrne, R.H., Hu, X.P., Peng, T.H., Huang, W.J.,2013. The marine inorganic carbon system along the Gulf of Mexico andAtlantic coasts of the United States: Insights from a transregional coastal carbonstudy. Limnol. Oceanogr. 58, 325e342.

Yamamoto-Kawai, M., McLaughlin, F.A., Carmack, E.C., Nishino, S., Shimada, K.,2009. Aragonite undersaturation in the Arctic Ocean: effects of ocean acidifi-cation and sea ice melt. Science 326, 1098e1100.

Zhu, Q.Z., Aller, R.C., Fan, Y.Z., 2006. Two-dimensional pH distributions and dy-namics in bioturbated marine sediments. Geochim. Cosmochim. Acta 70,4933e4949.