Hypoxia, Nitrogen, and Fisheries: Integrating Effects ...olli/eutrsem/Breitburg09.pdf · low dissolved oxygen concentrations (hypoxia) in estuaries and semienclosed oxygen-depleted

ANRV396-MA01-14 ARI 4 November 2008 8:28

Hypoxia, Nitrogen,and Fisheries: IntegratingEffects Across Localand Global LandscapesDenise L. Breitburg,1 Darryl W. Hondorp,1

Lori A. Davias,1 and Robert J. Diaz2

1Smithsonian Environmental Research Center, Edgewater, Maryland 21037;email: [email protected]; [email protected]; [email protected] Institute of Marine Science, The College of William and Mary, Gloucester Point,Virginia 23062; email: [email protected]

Annu. Rev. Mar. Sci. 2009. 1:329–49

First published online as a Review in Advance onSeptember 4, 2008

The Annual Review of Marine Science is online atmarine.annualreviews.org

This article’s doi:10.1146/annurev.marine.010908.163754

Copyright c© 2009 by Annual Reviews.All rights reserved

1941-1405/09/0115-0329$20.00

Key Words

low dissolved oxygen, eutrophication, estuaries, fish, macroinvertebrates,scale

AbstractAnthropogenic nutrient enrichment and physical characteristics result inlow dissolved oxygen concentrations (hypoxia) in estuaries and semienclosedseas throughout the world. Published research indicates that within and nearoxygen-depleted waters, finfish and mobile macroinvertebrates experiencenegative effects that range from mortality to altered trophic interactions.Chronic exposure to hypoxia and fluctuating oxygen concentrations impairreproduction, immune responses, and growth. We present an analysis of hy-poxia, nitrogen loadings, and fisheries landings in 30 estuaries and semien-closed seas worldwide. Our results suggest that hypoxia does not typicallyreduce systemwide fisheries landings below what would be predicted fromnitrogen loadings, except where raw sewage is released or particularly sen-sitive species lose critical habitat. A number of compensatory mechanismslimit the translation of local-scale effects of hypoxia to the scale of the wholesystem. Hypoxia is, however, a serious environmental challenge that shouldbe considered in fisheries management strategies and be a direct target ofenvironmental restoration.

329

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


INTRODUCTION

Low dissolved oxygen (DO) affects a wide variety of aquatic systems, ranging from the vast oxygenminimum zones of the world’s oceans (Helly & Levin 2004) to small ponds measuring less thana hectare in size. Low DO concentrations (hypoxia) result from a complex mix of hydrology,basin morphometry, nutrient supplies, and food web dynamics that vary among water bodiesand within interconnected waterways. A feature common to all affected systems, however, is thathypoxia influences the quantity and quality of habitat available to organisms dependent on aerobicrespiration, and significantly impacts physiological processes in organisms that utilize oxygen-depleted waters. The ultimate consequences of low DO concentrations include not only directeffects, but also a host of indirect effects mediated through changes in organism abundances,distributions, trophic interactions, and fisheries.

Descriptions of coastal waters in both the popular and technical press evoke contrasting imagesof massive dead zones and productive fisheries in systems affected by hypoxia. These contrastsemerge from the juxtaposition of nutrient enrichment, habitat degradation, and high populationdensities in the coastal zones of the world’s waters. Nitrogen (N) enrichment generally increasesbiological production (Nixon & Buckley 2002), whereas hypoxia reduces biomass and habitatquality and quantity (Caddy 1993, Diaz 2001, Rabalais & Turner 2001). The effects of nitrogenenrichment and hypoxia on food webs and fisheries may therefore be strongly influenced by theextent to which these two factors co-occur. The perceived effects of nutrient enrichment andhypoxia also depend on the spatial scales of interest, which range from local, sometimes severelyaffected water masses to entire fisheries ecosystems (Figure 1). The relevant spatial scales are, inturn, dependent on a range of factors such as jurisdictional boundaries, the specificity of habitat

Indirect effects

Areas of higher fishing pressure

Large marine ecosystem

Whole

estuary

Hypoxicregion

Chronicexposure

Acute mortality

Enrichedproduction

Severehypoxia

Figure 1The severity of hypoxia, nutrient enrichment of primary and secondary production, and the separate andinteracting effects of these two changes on the aquatic environment vary across the estuarine landscape. Botheffects of these stressors, and the level of management concern over consequences to living resources andtheir fisheries, can depend on how the hypoxic region, whole estuary, and larger marine ecosystem coincidewith ecological, cultural, economic, and jurisdictional boundaries.

330 Breitburg et al.

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


requirements of species or life stages of interest, and the degree of dependence of local economiesand communities on particular fishing grounds (Breitburg et al. 2008).

Our review and synthesis focuses on effects of hypoxia on finfish and mobile macroinvertebratesin estuaries, semienclosed seas, and similar water bodies. It is in these aquatic systems that multiplestressors resulting from human activities most clearly and strongly coincide, directly and indirectlyaffecting all levels of aquatic food webs. More than 400 of these systems worldwide have reportedhypoxia in recent years (Diaz & Rosenberg 2008). We focus on mobile species because theirpotential ability to avoid low DO concentrations makes understanding and predicting effects ofhypoxia on their populations and fisheries more difficult than for sessile species with no avoidancerecourse postsettlement.

In addition to reviewing published studies, with an emphasis on recently published research,we present a new analysis of data from 30 estuaries and semienclosed and coastal seas worldwideto examine the relationship between nutrient loadings and hypoxia, and the relationship betweenboth of these stressors and fisheries. We first discuss the causes and correlates of hypoxia and itseffect on organisms at the local scale, i.e., within or in close proximity to the location that hypoxiaactually occurs in affected systems. We then turn more directly to the relationship between hypoxiaand fisheries, especially at the larger whole-system scale. The relationship between hypoxia andfisheries is particularly important because these stressors potentially interact (Breitburg et al.,2008), and because correct identification of the consequences of human activities is required forthe development of appropriate management strategies. Finally we discuss issues that influence thetranslation of effects across spatial scales and cases in which hypoxia may be particularly important.Our literature review emphasizes recent publications: There has been an explosion of new researchon the occurrence and effects of hypoxia as well as several reviews during the past decade (e.g.,Breitburg 2002, Wu 2002, Domenici et al. 2007). Rather than setting a specific number to classifywaters as hypoxic (common definitions are <2 mg liter−1 or <2 ml liter−1), we define hypoxiamechanistically as oxygen concentrations that are sufficiently reduced that they affect the growth,reproduction, or survival of exposed animals, or result in avoidance behaviors. Some analyses wedescribe require the use of particular numerical criteria, which we provide where necessary.

NITROGEN ENRICHMENT AND HYPOXIA

Much of the scientific interest and management concern about hypoxia in estuaries and semien-closed seas includes a focus on the general problem of eutrophication caused by the overenrich-ment of coastal waters with anthropogenically generated nutrients. Hypoxia and anthropogenicnutrient enrichment have been identified as two critical and increasing threats to coastal waters(NRC 2000, Jeftic et al. 2006, Diaz & Rosenberg 2008). The link between nutrient enrichment(typically N in marine systems; Howarth & Marino 2006) and hypoxia is particularly importantbecause in cases where a direct cause-and-effect relationship exists, management actions can mosteffectively improve water quality. However, the susceptibility of estuaries and semienclosed seastoward the development of hypoxia and the spatial extent of hypoxia are strongly influenced byphysical characteristics such as circulation, density stratification, and flushing rates (Druon et al.2004, Gilbert et al. 2005).

Not all systems that receive high nutrient inputs develop extensive hypoxia, and not all systemswith extensive or severe hypoxia have particularly high loadings of anthropogenically generatednutrients. Our comparison of nutrient loadings, physical characteristics, and hypoxia in 30 estu-aries and semienclosed seas in Europe, North America, Australia, and Asia found no relationshipbetween either basin N loadings (watershed, point source, and direct atmospheric deposition; mea-sured as kg N km−2 surface area year−1) or total N loadings (basin plus downstream N loadings,

www.annualreviews.org • Hypoxia, Nitrogen, and Fisheries 331

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


Bo

tto

m a

rea

hyp

oxic

(%)

N loadings (log10 kgN km–2 year –1)

100

80

60

40

20

2.5 3.0 3.5 4.0 5.0

0

4.5 2.5 3.0 3.5 4.0 5.0 5.54.5

a b

Figure 2Comparison of N loadings and hypoxic extent in (a) 30 estuaries and semienclosed seas in and near Europe,North America, Australia, and Asia, and (b) the 20 systems for which we had estimates of total N loadings(basin plus downstream). Systems with high N loading rates did not have more extensive hypoxia (R2 < 0.06and P > 0.30 for linear and nonlinear regression models tested for both basin N and total N loadings).

e.g., due to tidal transport or upwelling) and the spatial extent of hypoxia (defined here as themaximum percent of bottom area with <3 mg DO liter−1; Figure 2). Shallow, well mixed estuar-ies such as Delaware Bay, Galveston Bay, and the Macleay River Estuary do not tend to develophypoxia except in marginal shallows or during severe floods in spite of high N loading rates. Incontrast, year-round density stratification and low flushing rates result in extensive and seasonallypersistent hypoxia and anoxia in deep basins of systems such as the Black and Baltic Seas in spiteof relatively low N loading.

Within many systems, however, especially those characterized by seasonally occurring hypoxia,the severity, spatial extent, and duration of hypoxia do tend to be sensitive to the magnitudeof nutrient and carbon enrichment (NRC 2000). These nutrient-sensitive hypoxic regions alsoinclude shallow portions of systems with permanently hypoxic deep basins (e.g., Mee 2006). Loss ofhighly oxygenated habitat is reached when the system becomes aerobically saturated with organicmatter and the resultant rise in microbial metabolism consumes DO (NRC 2000). In deeper waterbodies, seasonal density stratification prevents reaeration of bottom water layers. In shallow calmwaters, diel cycling oxygen concentrations with night and early morning hypoxia result from thecombination of microbial decomposition and nighttime respiration, especially by unicellular algaeand macrophytes.

Increasing use of chemical fertilizers has been accompanied by an increase in the number of sys-tems reporting hypoxia worldwide (Diaz & Rosenberg 2008). DO concentrations have improvedin some systems now receiving reduced nutrient loadings either through planned managementactions or because of socioeconomic disruptions. Implementation of primary sewage treatment,which removes organic matter but typically has only modest effects on N loadings, has resultedin marked improvements in DO concentrations, especially in waters directly downstream of dis-charges in many systems, including the Thames (Araujo et al. 2000), Delaware River (Weisberget al. 1996), Hudson River (Brosnan & O’Shea 1996), Elbe Estuary (Thiel et al. 1995), and EastRiver (Parker & O’Reilly 1991). Reductions in nutrient inputs due to the loss of fertilizer subsidiesfrom the former Soviet Union and declining agricultural output starting in 1989 led to the elimi-nation of hypoxia on the northwestern shelf of the Black Sea by 1995 (Mee 2006). Similarly, waterquality in Tokyo Bay improved in the 1940s after the collapse of the Japanese economy (Kodamaet al. 2002, 2006).


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


EFFECTS AT THE LOCAL SCALE: WITHIN OR NEAR THEOXYGEN-DEPLETED WATER MASS

Hypoxia creates physiologically relevant physical structure in the water column and alters theestuarine landscape by creating a spatial mosaic of habitats that vary in physiological suitability,prey abundance, prey vulnerability, and predator threat (Figure 1). Hypoxia fundamentally affectsthe use of space and movement through the system by altering migration pathways and transport.Hypoxia also alters spatial patterns of human use by influencing the spatial distribution and removalof fisheries resources (Selberg et al. 2001). Most negative effects tend to occur locally, withinthe region of the water body in which oxygen concentrations are reduced, but may have muchwider ranging consequences mediated through the indirect effects of altered distributions andabundances. Note that referring to these effects as local does not imply that they always occurat very small spatial scales, only that we are not considering the broader, nutrient-enriched butmore highly oxygenated waters within the larger system. For example, local-scale analyses wouldconsider the portion of the Delaware River Estuary that was severely hypoxic during the early1900s as a result of raw sewage from the city of Philadelphia (Albert 1988), whereas large-scaleanalyses would consider the Delaware River Estuary as part of the greater Delaware Bay system(Figure 1).

Fish kills are the most dramatic manifestation of low DO concentrations, and result from acombination of severe physical conditions and the failure of mobile animals to avoid or escape lethalconditions. The frequency and magnitude of fish kills have increased as nutrient-related hypoxiahas worsened. Hypoxia-related kills that involve thousands to millions of finfish and crustaceanshave been reported from systems as diverse as the Mariager Fjord in Norway, the Neuse River,a coastal plain estuary in North Carolina, and the subtropical Richmond River Estuary in NewSouth Wales (Fallesen et al. 2000, NCDWQ 2006, Walsh et al. 2004). In U.S. waters, menhaden(Brevoortia tyrannus and B. patronus) kills feature prominently in records. These fish possess swimbladders, float where they are readily seen by the public, and travel in large schools, so killsoften affect large numbers of individuals. Some menhaden kills are thought to occur when fishescaping predators concentrate in shallow embayments where their respiration reduces oxygenconcentrations to lethal levels (Sklar & Browder 1998, Wannamaker & Rice 2000). Large fishkills in some systems, including the Macleay and Richmond River Estuaries in New South Wales,result from episodic floods that simultaneously lead to extensive hypoxia and rapid salinity declines(Walsh et al. 2004). In such cases, rapid response by management agencies in the form of fisheriesclosures in affected estuaries and nearby coastal areas can permit rapid recolonization and recoveryof the systems (Steffe et al. 2007).

Hypoxia also causes substantial mortality of developing embryos that is less visible to the publicand management than the kills that involve adult and juvenile fishes. Egg/embryo mortality can bethought of as lost production potential of a population. Developing embryos are particularly sen-sitive because they lack the ability to behaviorally respond to low oxygen and because oxygen mustdiffuse across the chorion that encases the embryo (Rombough & Randall 1988). Fish embryos typ-ically have narrower temperature tolerances than other life stages (Rombough 1997). Individualsnear the upper limits of temperature tolerances require higher oxygen concentrations (Rombough1997), and may therefore be more likely to suffer hypoxia-related deformities (Breitburg 1992,Shang & Wu 2004) or mortality. In Chesapeake Bay, DO criteria are based on oxygen require-ments of bay anchovy (Anchoa mitchilli) eggs because these eggs sink into hypoxic bottom watersin the weakly stratified tributaries (Keister et al. 2000, US-EPA 2003). In the Baltic, interannualvariation in recruitment of Baltic cod (Gadus morhua) has been attributed to interannual variationin the depth of the hypoxic/anoxic bottom layer into which cod eggs sink during their prolonged


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


development (Cardinale & Modin 1999, Koster et al. 2005). However, fisheries exploitation canstrongly affect the relationship between hypoxia and embryo mortality by selectively removinglarge females who produce more buoyant eggs than smaller females (Nissling et al. 1994, Vallin& Nissling 2000). Exposure of embryos to hypoxia prior to hatching can also decrease swimmingspeeds and increase predation posthatch mortality (Roussel 2007). Susceptibility to hypoxia in lar-val fishes may be particularly high around the time of metamorphosis, when oxygen consumptionpeaks (Ishibashi et al. 2005).

Post-embryonic life stages of nearly all mobile species that have been tested to date and thatwere drawn from populations that experience hypoxia exhibit behavioral avoidance of oxygenconcentrations well above those found to be lethal. In general, fish tend to avoid oxygen con-centrations similar to those found to reduce growth rates in laboratory experiments (Breitburg2002). However, realized responses in the field may not perfectly match those required to preventmortality and reductions in growth. Rapid declines in DO associated with upwelling of hypoxicbottom waters can impose a trade-off between risk of initiating an escape when DO is still highand predators are active, and the risk of remaining in place as DO declines (Breitburg 1992).In addition, there are energetic costs and limits to precise responses to the spatial and temporalcomplexity of diel cycling and tidally mediated patterns of low oxygen (Tyler & Targett 2007).Mechanisms required to orient toward normoxic water are also imperfect. For example, althoughcrustacean oxygen sensors are bilaterally paired, water is sensed only after it is drawn into thebranchial chamber, which presumably disrupts the directional gradient (Bell et al. 2003). Somefish that are extremely tolerant of hypoxia also may attempt to wait out hypoxic events instead offleeing. In laboratory experiments, eelpout (Zoarces viviparous) held at 0 mg liter−1 exhibited noescape behavior and survived for 60 minutes (Fischer et al. 1990).

Over the past decade, there has been a growing appreciation of the consequences of chronicexposure to mild-to-moderate hypoxia that is either not avoided or ineffectively avoided. Forspecies that utilize shallow waters with diel cycling hypoxia or chronically low but only mod-erately severe oxygen depletion, effects range from reduced growth and the energetic costs ofconstant migration in response to fluctuating DO concentrations (Perez-Dominguez et al. 2006,Stierhoff et al. 2006) to impaired reproduction and immune function (Thomas et al. 2007, Landryet al. 2007). Field and laboratory evidence indicates that some fish species undergo physiologicalresponses to prolonged exposure to nonlethal hypoxia that enable them to at least partially accli-mate to hypoxic conditions, including shifts in hemoglobin isomorphs (Tun & Houston 1986) andchanges in hemoglobin concentration (Petersen & Petersen 1990). But oscillating DO, such as thatfound in shallow systems with diel fluctuations in DO concentrations, may prevent physiologicalacclimation to hypoxia (Taylor & Miller 2001).

Exposure to mild-to-moderate hypoxia that is not avoided can lead to reproductive impairment,immunosuppression, and increased pathogen-related mortality. Field exposure to hypoxia reducesoocyte maturation and sperm production of Atlantic croaker (Micropogonias undulatus), probablyby decreasing production of hormones involved in stimulation of the reproductive neuroendocrinesystem (Thomas et al. 2007). Experimental work with Gulf killifish (Fundulus grandis) suggeststhat hypoxia may delay vitellogenin uptake by oocytes and reduce 11-ketotestosterone in males(Landry et al. 2007). Experiments with the freshwater zebrafish (Danio rerio) also indicate thathypoxia can affect sex differentiation (Shane et al. 2006).

Shrimp and crabs experimentally exposed to hypoxia and pathogens suffer increased suscepti-bility to disease and mortality from bacterial infections as a result of depressed immune function(Le Moullac et al. 1998, Holman et al. 2004, Burgents et al. 2005, Tanner et al. 2006). In fish,even short-term exposure to hypoxia (0.5–96 h) can reduce bactericidal activity, antibody levels,and production of disease-fighting reactive oxygen species, although some of these effects may


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


be reversible upon restoration of normoxic conditions (Valenzuela et al. 2005, Choi et al. 2007,Welker et al. 2007). Laboratory experiments have indicated that hypoxia can increase papillo-matosis tumors in roach Rutilus rutilus (Korkea-aho et al. 2008). Field evidence also suggests thatoutbreaks of mycobacteriosis in Chesapeake Bay striped bass (Morone saxatilis) and papillomatosisin North Sea dab (Limanda limanda) may be related to low DO (Mellergaard & Nielsen 1997,Vogelbein et al. 1999).

It is important to consider that habitats avoided are functionally lost from the system (Breitburg2002, Rabalais & Turner 2001, Sagasti et al. 2001). As a result, habitat loss due to hypoxia isfar greater than would be estimated by calculations based on species recruitment or survivaltolerances. Because species vary in their oxygen requirements, sensitive predators can lose accessto prey (Breitburg et al. 1997, Nestlerode & Diaz 1998). Crowding in more highly oxygenatedrefuges can result in density-dependent growth reductions (Eby & Crowder 2002) or increasedcannibalism (Aumann et al. 2006). Because hypoxia can make deeper, cooler waters unavailable inthe summer, or overlaps with nursery habitat, the combined effects of temperature and oxygen canresult in habitat compression or squeeze that affects both pelagic and demersal species (Coutant1985, Niklitschek & Secor 2005, Pearce & Balcom 2005).

Effects of climate change on hypoxia-related mortality may be an increasingly important issue,not only because hypoxia in many systems is predicted to increase (Justic et al. 2003, Boesch et al.2007), but also because of the effect of temperature on metabolism and therefore on oxygenrequirements. Fish tend to be less tolerant of hypoxia and need higher oxygen concentrations tomeet baseline metabolic requirements at higher temperatures compared with lower temperatures(Schurmann & Steffensen 1992, Secor & Gunderson 1998, Shimps et al. 2005). Recent experimentsindicate a genetic basis for the relationship between temperature and hypoxia tolerance throughthe effect of temperature on HIF-1 (hypoxia inducible factor-1), the primary regulator of geneexpression in response to hypoxia (Rissanen et al. 2006).

Hypoxia can affect all aspects of trophic interactions including encounter rates, susceptibility topredators, the ability to capture prey, and the importance of different trophic pathways (Breitburget al. 1997, Baird et al. 2004, Craig & Crowder 2005, Seitz et al. 2003; reviewed in Breitburg 2002,Domenici et al. 2007). The spatial patterning of oxygen concentrations within any affected systemcan lead to spatial variability in the effect of hypoxia on trophic interactions across the estuarinelandscape. Hypoxia can reduce responsiveness to threat stimuli, and, to a lesser extent, locomotorycomponents of escape responses (e.g., Nestlerode & Diaz 1998, Lefrancois et al. 2005), induce useof aquatic surface respiration, which increases vulnerability to predatory birds (e.g., Poulin et al.1987), and affect interfish spacing and coordination of swimming within fish schools (Domeniciet al. 2002). Benthic prey may leave burrows and exhibit reduced escape responses to predatorthreat (Tyson & Pearson 1991). Consumers constrained to hypoxia-degraded habitat may shiftto poorer quality prey (Powers et al. 2005). Avoidance of low oxygen can also increase energeticcosts and decrease growth by extending required foraging time for species whose prey can utilizehypoxic water from which they are largely excluded (Taylor et al. 2007).

Effects of hypoxia on predator-prey interactions may be especially important where stronglydivergent taxa, such as gelatinous zooplankton and teleost fishes, occupy similar trophic positionsbut vary considerably in their tolerance to low DO concentrations (Breitburg et al. 2003). InChesapeake Bay, for example, the ctenophore Mnemiopsis leidyi, and bay anchovy (Anchoa mitchilli)are both important consumers of zooplankton, but the ctenophore can survive continuous ex-posure to 0.5 mg/L DO for >4 days, whereas bay anchovy requires substantially higher oxygenconcentrations. Most studies have found that moderate levels of hypoxia lower predation rates bypiscivorous fishes, but either increase (Breitburg et al. 1997, Shoji et al. 2005) or have no effecton (Kolesar 2006) predation on fish larvae by hypoxia-tolerant gelatinous zooplankton.


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


SCALING UP: POPULATIONS AND FISHERIES

Predicting effects of hypoxia on populations and fisheries landings of mobile species is particularlychallenging because behavioral responses of both mobile species and fishers influence ultimate ef-fects when viewed at systemwide spatial scales, and many systems with extensive hypoxia supportlarge and valuable fisheries. Because both overfishing and nutrient overenrichment are related topopulation density, these stressors tend to co-occur. Literature ranging from qualitative descrip-tions to quantitative survey or fisheries data suggests four fairly well-substantiated categories inwhich hypoxia causes or contributes to declines in species or overall biomass of fishes and mobilemacroinvertebrates: (a) blockages of migratory species caused by release of raw sewage, especiallyin tidal riverine portions of estuaries, (b) extensive hypoxia that affects a very large percentage ofbottom area, (c) combinations of hypoxia and other stressors, and (d ) elimination of specificallyrequired habitat.

Probably the clearest cases of hypoxia-related population and fisheries declines in estuaries andsemienclosed seas involve systems in which discharges of raw sewage have caused severe oxygendepletion in tidal river portions of estuaries. The location and seasonal persistence of hypoxiacaused by large discharges of raw sewage with high biological oxygen demand appears to beparticularly damaging, especially to the completion of life cycles of anadromous and catadromousspecies that require use of, and transit through, both saline and freshwater portions of estuarinesystems. Large stretches of river estuaries such as the Mersey, Thames, and Elbe have beendescribed as devoid of fish prior to implementation of primary sewage treatment (Jones 2006,Tinsley 1998, Thiel et al. 1995). These hypoxic zones have blocked upriver spawning migrationsor downriver outmigrations of species such as American shad (Alosa sapidissima) in the DelawareRiver Estuary (Weisberg et al. 1996, Albert 1988); sea trout (Salmo trutta) in the Mersey (Jones2006); and sturgeon (Acipenser sturio), Atlantic salmon (Salmo salar), allis shad (Alsoa alosa), twaitshad (Alosa fallaxx fallax) sea lamprey (Petromyzon marinus), and European eel (Anguilla anguilla)in the Scheldt (Maes et al. 2007), resulting in population declines or local extirpations.

There is also persuasive evidence of system-wide declines in fisheries landings (particularly inlandings of crustaceans; D.W. Hondorp, D.L. Breitburg & L.A. Davias, manuscript submitted)and abundances of sensitive species in systems with very high percentages of bottom hypoxia (seesection on Large Scale Effects, below). For example, abundances of demersal fish and crustaceans,including mantis shrimp (Oratosquilla oratoria), have declined in Tokyo Bay, where up to two-thirdsof the bottom becomes hypoxic during summer (Kodama et al. 2002, 2006). Hypoxia also appearsto contribute to systemwide population or fisheries declines in combination with pathogens (e.g.,American lobster Homarus americanus in Long Island Sound; Pearce & Balcom 2005) and othereffects of eutrophication (e.g., plaice recruitment in the Kattegat; Pihl et al. 2005), particularlywhere the co-occurring or interacting stressors affect organisms in highly oxygenated areas thatwould otherwise serve as refuges from hypoxia.

A few examples suggest that degradation of particularly critical habitat can affect recruitment orfisheries landings. One of the better known examples is the interannual variation in egg mortalityand fisheries landings of Baltic cod in the Baltic Sea discussed above (Nissling & Vallin 1996, Kosteret al. 2005). In the Western Atlantic, hypoxia appears to contribute to low per capita productivityand slow recovery of cod stocks in the Gulf of St. Lawrence (Dutil et al. 2007). Along the Texas shelfof the Gulf of Mexico, interannual variation in landings of brown shrimp (Farfantepenaeus aztecus)is negatively correlated with size of the hypoxic zone (O’Connor & Whitall 2007). However,Texas fisheries regulations may contribute to the apparent strength of this correlation; fishing isprohibited on the inner shelf refuge from hypoxia during much of the season when northern Gulfof Mexico hypoxia is most extensive.


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


LARGE SCALE EFFECTS: A CROSS-SYSTEMCOMPARISON OF FISHERIES

Do estuaries and semienclosed seas with hypoxia have lower or altered composition of fisherieslandings relative to other systems? What is the relationship between the spatial extent and severityof hypoxia and fisheries landings? Are local-scale effects of hypoxia reflected in the magnitude,composition, or value of fisheries landings? In spite of a rich literature that demonstrates effectsof hypoxia on physiological processes, mortality, behavior and food web interactions at the localscale, and strong effects on fisheries landings within the parts of systems most severely affectedby hypoxia, clear evidence of population-level and system-wide negative effects on mobile livingresources is still quite limited.

We examined the relationships among landings of finfish and mobile macroinvertebrates, thespatial extent of hypoxia, and N loadings in estuaries and semienclosed seas in Europe, NorthAmerica, Australia, and Asia to address the following questions:

1. What is the relationship between fisheries landings and N loadings? The agricultural modelsuggests that fisheries landings increase with increasing primary production, which respondsin turn to increased levels of N enrichment (e.g., Nixon & Buckley 2002).

2. Does hypoxia reduce fisheries landings below levels predicted by N loadings? Reducedhabitat due to oxygen depletion, as well as density- or temperature-dependent reductionsin growth or increases in mortality in oxygenated refuge habitats, could lead to decreasedfisheries landings.

3. Do fisheries landings reflect a hypoxia-caused decrease in the efficiency of production offishable biomass (e.g., Baird et al. 2004)?

4. Do high N loadings or the spatial extent of hypoxia alter the composition of fisheries land-ings? Conceptual and statistical models suggest that the ratio of pelagic planktivores todemersal species in landings should increase with increasing eutrophication because plank-tivores are more able to access enhanced prey production without experiencing the negativeeffects of hypoxia on lower water column and benthic habitat (Caddy 1993, 2000; de LeivaMoreno et al. 2000).

5. Does hypoxia or high N loadings lead to declining trends in fisheries landings? Hypoxia in-creases susceptibility to fisheries practices adapted to the spatial patterns of oxygen-depletedand refuge habitat. Fishers worldwide take advantage of spatial distributions of target speciescaused by hypoxia (e.g., Selberg et al. 2001, Craig & Crowder 2005; J. Shoji, personal com-munication), potentially increasing the efficiency of fisheries and increasing the likelihoodof overfishing. In addition, the cumulative effects of habitat degradation could result indeclining trends in species abundances not evident in multiyear averages.

Our analysis differs from previous cross-system comparisons (e.g., Caddy 2000, de LeivaMoreno et al. 2000, Nixon & Buckley 2002) by explicitly examining the link between landings,N loadings, and hypoxia in temperate and subtropical estuaries and semienclosed seas along fourcontinents, and by focusing on mobile species. N tends to be the primary driver of eutrophicationin marine systems, and is the direct target of environmental regulations and management strategiesdesigned to reduce hypoxia. We focus on mobile species because of their potential ability to avoidphysiologically stressful or lethal DO concentrations and because bivalve fisheries worldwide aretypically supplemented or sustained with hatchery production. We considered both the spatial ex-tent of bottom with oxygen concentrations below 3 mg liter−1 (a measure of the percent of habitatlost or degraded) and the severity of hypoxia to capture potential qualitative differences betweensystems that had similar spatial extents of hypoxia, but differed in seasonal persistence, minimumoxygen concentrations, or susceptibility to occasional floods that cause hypoxia. To the extent


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


possible, we exclude aquaculture from our landings estimates, because aquaculture may be less re-flective of system-wide nutrient enrichment and habitat quality than wild fisheries. We emphasizethat we are specifically asking about the relationship between hypoxia and fisheries in this analysis—fisheries may or may not reflect patterns of actual abundances, and both landings and catch per uniteffort data can be strongly influenced by fishing regulations and practices, economics, and culture.

Our database focuses on 1990–2004 and includes 30 estuaries and semienclosed seas for whichwe were able to obtain commercial landings data, recreational fishing data, the spatial extent ofhypoxia, and basin N loadings (i.e., N flux from upstream connected systems, the watershed,point sources, and direct atmospheric deposition). We also considered total N loadings (basin plusdownstream) for the 20 systems for which estimates of downstream boundary N flux was available.Our database includes N loadings and fisheries landings information from sources that range fromindividual state fisheries department records to published N flux models to FAO databases, andwas dependent on the generous help of the many people and agencies listed in the SupplementalMaterial. (Follow the Supplemental Material link from the Annual Reviews home page athttp://www.annualreviews.org.) Systems ranged in size from the 7 km2 Bellinger Estuary insoutheastern Australia to the 775,065 km2 East China Sea, and from 0% to 77% of bottom areaaffected by hypoxia. All statistical analyses and descriptions of the form of the relationships amonglandings, N loadings, and hypoxia refer to log-transformed landings and loadings data normalizedto the system surface area. In cases where it affected results, we did analyses with and without thesix Australian estuaries in our database because we did not have downstream N flux data for anyof these estuaries, and because these estuaries are substantially smaller than other systems in ourdatabase. To describe the general form of the relationship among variables in our dataset, we usedAkaike’s Information Criteria (AIC) to select among linear, asymptotic, and peaked functions thatyielded similar R2 values. We used SigmaPlot (v10, Systat Software Inc.) for curve-fitting, and SAS(v9.1, SAS Institute) for other statistical analyses. System descriptions and details on data sourcesand analyses are available in the Supplemental Material.

Total landings of finfish and mobile macroinvertebrates increased with increasing basin N load-ings to an annual rate of approximately 14 mt N km−2 year−1, the point in our dataset representedby Chesapeake Bay (Figure 3a). A sigmoidal function provided the best fit to describe the rela-tionship between fisheries landings and basin N loadings for the entire dataset, whereas a peakedcurve provided a better fit if the Australian systems were omitted from the analysis (Figure 3a).Four-parameter sigmoidal and peaked functions provided similar fits for fisheries landings versustotal N (Figure 3b). Sigmoidal and peaked functions make different predictions for the potentialeffects of nutrient management on fisheries landings if within-system temporal trajectories tend tofollow the same relationship between nutrient enrichment and fisheries landings as predicted byamong-system comparisons (Nixon & Buckley 2002). A sigmoidal relationship suggests that fora wide range of N loading rates, N and potentially negative consequences of eutrophication suchas high chlorophyll a and turbidity could be reduced without negatively affecting total fisherieslandings, but that no level of nutrient reductions will increase landings. A peaked curve impliesthat for systems with the highest levels of nutrient loadings, a reduction in N loadings may increaselandings, but that for other systems a reduction in N would likely reduce total landings. Interest-ingly, some of the systems on the downward slope of the peak that may benefit from reduced Nloadings do not have extensive hypoxia.

Analysis of residuals from the best-fit sigmoidal and Gaussian functions indicated that therelationship between N loadings and fisheries landings was not significantly affected by the spa-tial extent of hypoxia, or whether systems had minimal (<1% hypoxia), episodic, seasonal, orpersistent hypoxia (regressions between residuals and the spatial extent of hypoxia: P > 0.34 forall comparisons; 1-way ANOVA on residuals versus hypoxia severity category: P > 0.50 for all


Supplemental Material

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.

http://arjournals.annualreviews.org/doi/suppl/10.1146/annurev.marine.010908.163754


2.5

3.0

3.5

4.0

4.5Fi

sher

ies

lan

din

gs

(lo

g10

kg

km

–2 y

ear –

1 ) a

N loadings (log10 kgN km–2 year –1)2.5 3.0 3.5 4.0 5.04.5

< 2% hypoxia

4–19% hypoxia

25–31% hypoxia

39–77% hypoxia

b

2.5 3.0 3.5 4.0 5.04.5

Figure 3Relationship between N loadings, hypoxia, and fisheries landings of mobile species in estuaries andsemienclosed seas. (a) Basin N loadings and fisheries landings. A four-parameter sigmoidal function providedthe best fit when Australian systems (gray) were included (gray line in a: R2 = 0.56, p < 0.0001,f = 2.82 + 0.87/{1 + exp[–(x-3.49)/0.12]}), but peaked curves provided the best fit without the smallAustralian estuaries (black regression line = 4 parameter Gaussian function: R2 = 0.61, p < 0.0003,f = 2.83 + 0.99 · exp{–0.5 · [(x-4.08)/0.45]2}) (n = 30). (b) Four-parameter sigmoid (R2 = 0.66) andGaussian (R2 = 0.65) functions provided similar fits to the relationship between total N loadings andfisheries landings (n = 20). Systems and percent bottom area <3 mg liter−1 in a from lowest to highestN loading are: Gulf of St. Lawrence (9.3), Sea of Azov (40.0), Gulf of Bothnia (0.0), Black Sea (76.6), EastChina Sea (4.1), Northern Adriatic Sea (1.4), Irish Sea (0.0), North Sea (0.0), Gulf of Finland (25.4),Yellow/Bohai Sea (1.2), Baltic Sea proper (including the Gulf of Riga) (41.8), Kattegat/Skagerrak (9.5), SetoInland Sea (5.3), Pamlico River Estuary (39.3), Chesapeake Bay (18.0), Tampa Bay (13.3), Gulf of Mexico(18.5), Long Island Sound (31.1), Corpus Cristi Bay (10.0), Neuse River Estuary (39.3), Marmara (73.6),Sound/Belt Seas (46.0), Delaware Bay (0.0), Bellinger River Estuary (0.0∗), Clarence River Estuary (0.0∗),Manning River Estuary (0.0∗), Galveston Bay (0.0), Hastings River Estuary (0.0∗), Richmond River Estuary(0.0∗), and Macleay River Estuary (0.0∗). Asterisk denotes exclusion of years of major floods.

comparisons). Systems included in our analysis ranged from 0% to 77% bottom hypoxia. Seven ofthe ten non-Australian systems with the highest landings km−2 had seasonal or persistent hypoxiathat affected 13% or more of the bottom area. Both N loadings and landings km−2 were similar forpairs of systems such as the Azov and Gulf of Bothnia, and the Neuse River Estuary and DelawareBay, that differed greatly in the spatial extent of hypoxia (40% versus ∼0%, and 39% versus ∼0%bottom area hypoxic or anoxic, respectively; Figure 3). All five systems with ≥40% hypoxia hadfisheries landings lower than predicted by basin N loadings, but this was not true for landingspredicted for total N loadings.

Fisheries landings per unit N load were not significantly correlated with the spatial extent ofhypoxia (P > 0.05 with Australian systems included, P > 0.75 without Australian systems). Seven ofthe ten systems with the highest fisheries landings per unit N loading had seasonal or persistent hy-poxia, but the six systems with highest landings per unit N load all had <20% bottom area hypoxia(Figure 4). The same characteristics may predispose systems to both high fisheries landings andseasonal hypoxia: retention times sufficiently long to allow excess phytoplankton production to beretained in the system or be transmitted to upper trophic levels, and water columns sufficiently deepto allow bottom layer hypoxia to develop and moderate-to-large-sized fishing vessels to operate.

Comparisons of landings of pelagic planktivores, benthopelagics (bottom-oriented but do nottypically perch on the sediment), and benthic species (species without swim bladders or those thatinhabit burrows, i.e., species most strongly dependent on benthic habitat) with N loadings andhypoxia extent only partially supported the assumptions underlying models that predict shifts in


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


0

0.5

1.5

2.0

2.5

Fish

erie

s la

nd

ing

s : N

load

ing

s(m

t wet

wei

gh

t m

t N –

1 )

Bottom area hypoxic (%)0 20 40 60 10080

1.0

Figure 4Fisheries landings per unit nitrogen loading relative to the spatial extent of hypoxia. Small, shallow estuariessuch as the Australian systems (red ) had low landings relative to N loadings, but there was no relationshipbetween hypoxia and the ratio of landings to N among other systems. Both normoxic systems and the twosystems with the greatest spatial extent of hypoxia and anoxia had low fisheries yields. Percent hypoxia inAustralian systems is slightly offset in graph (plotted as 1.5 instead of 0.0) so that data from other systems arenot obscured; Australian fisheries data are for nonflood years.

pelagic:demersal ratios (Figure 5). Landings of mobile benthic species increased with increasing Nloading and tended to decrease with increasing hypoxia. Five of the eight systems in the completedataset with the lowest benthic landings have ≥40% bottom area hypoxia, and this group includesall four systems with either year-round oxygen depletion in deep basins or extensive anoxia (theBaltic, Black, Marmara, and Azov). The pattern of low benthic landings in systems with extensivehypoxia was not statistically significant, however, unless the Gulf of Bothnia, which has no hypoxiaand a fishery almost entirely comprised of herring, is removed from the analysis. Landings of ben-thic species tended to be higher than landings of pelagics at high N loadings, and highest landingsof both benthics and pelagics occurred in systems with seasonal hypoxia. A decline in landings ofpelagics at N loading rates above approximately 104 kg N km−2 year−1appears to reflect fisheriesregulations and practices that preclude the use of industrial-scale purse seining (e.g., Delaware Bay,Galveston Bay). Landings of pelagics were not significantly related to the spatial extent of hypoxia.Landings of benthopelagics (which include important fisheries species such as cod) increased withincreasing N loads and showed no significant relationship to the spatial extent of hypoxia.

Our data also do not provide support for negative effects of hypoxia on temporal trends inlandings, at least over the relatively short period of time included in our analysis. We used simplelinear regressions of standardized landings ([x-mean]/sd) to assess temporal trends in landingsduring the period 1990–2004, and compared the slope of the regressions with the spatial extent ofhypoxia in each system. Landings of demersals and all species combined declined in the majority ofsystems, but neither the likelihood of declining nor the severity of decline increased with increasingspatial extent of hypoxia or basin N loadings (Figure 6; all P > 0.25 for linear regressions). Inclusionof downstream boundary N flux to N loadings did not change our results. Declining landings incoastal systems worldwide are highlighted in a number of recent reviews (e.g., Jackson et al. 2001,Lotze et al. 2006).

TRANSLATING ACROSS SCALES: COMPENSATORY MECHANISMS

Eutrophication in general, and hypoxia specifically, can negatively affect estuaries and similarhabitats. Hypoxia can strongly alter food webs, fisheries landings, and the abundance of species


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


–1

1

3

4

5La

nd

ing

s (l

og

10 k

g k

m–2

yea

r –1 )

2

0

–1

1

3

4

2

0

2.5 3.0 3.5 4.0 4.5

0 20 40 60 80

Bottom area hypoxic (%)0 20 40 60 80 1000 20 40 60 80

5

Lan

din

gs

(lo

g10

kg

km

–2 y

ear –

1 )

2.5 3.0 3.5 4.0 4.5

N load (log10 kg km–2 year –1)2.5 3.0 3.5 4.0 4.5 5.0

BenthicsPelagic planktivores

Benthics

Benthopelagics

Pelagic planktivores

a

b

Benthopelagics

Figure 5Comparison of landings of small-to-medium sized pelagics (mostly planktivores), benthodemersals (species that utilize near-bottomhabitat or feed on benthic prey but are not entirely dependent on the benthos), and benthic species (species that lack swimbladders orthose that reside in burrows) with basin N loadings and percent bottom hypoxia. Twenty-four European, North American, and Asiansystems are included; species composition data for Australian estuaries are not in this dataset. Regressions [best fit of significant linear,polynomial, or asymptotic relationships based on Akaike’s Information Criteria (AICc) comparisons] are shown. (a) N loads: Pelagicplanktivores: y = −14.87+9.92x − 1.36x2, R2 = 0.28, P = 0.03; Benthopelagics: y = −10.44 + 6.15x − 0.72 × 2, R2 = 0.33, P =0.02; Benthics: y = −6.83 + 3.60x − 0.29x2, R2 = 0.40, P = 0.004. (b) Hypoxia: All P � 0.05, but a significant pattern of lowerbenthic landings with increasing spatial extent of hypoxia is found if the Gulf of Bothnia (lower left point), which is almost entirely aherring fishery, is deleted (y = 3.11 − 0.02x; R2 = 0.28, P = 0.01). Analyses using total N loadings yielded similar results.

on the local scale, and can negatively affect survival, growth, and reproduction of individualsexposed to physiologically stressful oxygen concentrations (Diaz 2001, Breitburg 2002, Gray et al.2002, Wu et al. 2003). Other associated negative effects of eutrophication, including the loss ofsubmerged macrophytes and development of harmful algal blooms (HABs), further reduce habitatextent and suitability (Deegan 2002, Duarte 2002, Cloern 2001). Yet our analysis of fisherieslandings indicated little relationship between the spatial extent or severity of hypoxia and whole-system fisheries landings, and our literature review revealed few unequivocal cases of population-level or whole-system hypoxia-related fisheries declines not associated with discharge of rawsewage.

Compensatory mechanisms may limit the translation of local effects of hypoxia, and negativeeffects of high N loadings in general, to systemwide reductions in fisheries landings. First, bothmobile species and fishers can potentially distribute themselves spatially to avoid oxygen-depletedhabitats and utilize prey-enriched habitats (Selberg et al. 2001, Breitburg 2002, Eby & Crowder2004). As a result, declines in diversity and biomass at the patch scale may not be reflected in anegative relationship with hypoxia at the scale of the whole system. The degree of prey enrichment,however, is a function of how nonhypoxic portions of the ecosystem respond to N loading. In the


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


–0.4

–0.2

0

0.3

0.4

–0.1

–0.3

N loadings(log10 kgN km–2 year –1)

Bottom area hypoxic (%)

Total landingsa b

0 20 40 60 80

0.2

0.1

3.0 3.5 4.0 4.5 5.02.5

Demersal landingsc

0 20 40 60 80

Bottom area hypoxic (%)

Stan

dar

diz

ed tr

end

Total landings

Figure 6Neither the spatial extent of hypoxia nor basin N loadings were significantly related to standardized trends in landings [slopes of linearregressions of (X-mean)/SD] of (a,b) all fish and mobile macroinvertebrates or (c) demersal species. The analyses included data from the21 systems in North America, Europe, and Asia for which we could acquire at least nine years of estimates of recreational catches,commercial landings, and species composition for consistent geographic boundaries.

Baltic Sea, for example, the macrobenthic biomass in shallow areas above the halocline increased3 to 5 times between 1920 and 1976—a positive effect of N loadings that likely offsets biomass lostto persistent hypoxia below the halocline (Elmgren 1989). High fishing pressure in habitats thatserve as a refuge from hypoxia could increase fishing mortality and increase the susceptibility ofstocks to overfishing. However, we found no relationship between hypoxia and trends in landingsin our dataset.

Second, fishing practices may prevent the expression or detection of negative effects of hypoxiaon fisheries landings by constraining populations to levels below habitat-determined carrying ca-pacities. Fishing pressure strongly affects population sizes of target species even in cases wherestocks are not considered technically overfished. Population sizes that generate maximum sustain-able yields are frequently used as management targets, and species with populations at predictedmaximum sustainable yields are not typically considered to be overfished (Hilborn 2007). Popula-tion sizes that generate maximum sustainable yields are, by definition, lower than carrying capac-ities. Where the spatial extent of suitable habitat is not the primary factor controlling populationsize, the loss of habitat may typically have little effect on population levels. Regulatory reduc-tions in fishing mortality that allow populations to more closely approach habitat-determinedcarrying capacities may facilitate the expression of hypoxia-related habitat loss effects onlandings.

Third, decreased water transparency associated with high nutrient loads may act as a diffuserefuge from predation that compensates for the loss of the more localized, but highly effectiverefuge provided by submerged macrophytes. Because piscivores have longer reactive distancesthan planktivores or benthivores, their feeding will be compromised by nutrient-related turbiditythat limits visual feeding (Utne-Palm 2002, De Robertis et al. 2003). Although both the prefer-ential use of submerged macrophytes and their importance as a refuge for small and juvenile preyspecies in estuaries has been clearly demonstrated (Deegan 2002), systemwide declines in speciesthat utilize macrophytes as nursery habitat have been difficult to identify. We suggest that nutrientenrichment that results in epiphyte growth and light limitation, and reduces macrophyte popula-tions, may provide a less effective but more widespread refuge from predation for these same preyspecies.


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


HABITAT DEGRADATION AND THE IMPORTANCEOF MANAGEMENT ACTION

Eutrophication is a high-priority concern for environmental and fisheries management becauseit clearly results in severe habitat degradation and local (i.e., within and near hypoxic regions)reductions in the production, diversity, and abundance of fish and invertebrates. Local is notsynonymous with unimportant; historical data that may reflect current conditions in systemsthat receive effluent from highly populated areas without primary sewage treatment illustrate thepotential for substantial economic and social costs of hypoxia-induced fishery collapses. Although80–90% of sewage in Western Europe receives at least primary treatment, in Latin America,East and Southern Asia, and West and Central Africa, 80% or more of wastewater is dischargeduntreated (Jeftic et al. 2006). In addition, the destruction and disruption of ecological systemsand services within hypoxic areas is an important concern in its own right. Without an intentionaldecision-making process to do so, we have turned our coastal waters into overfertilized agriculturalsystems. Unlike on land, sharp boundaries between fertilized and unfertilized plots cannot bedrawn to preserve the diversity and services of less enriched, more highly oxygenated ecosystems.

Nevertheless, our analysis indicates that in many, perhaps most, temperate and subtropicalestuaries and semienclosed seas, hypoxia is not likely to be the most important factor limitinglandings of finfish and mobile macroinvertebrates. Local habitats and particularly susceptiblespecies may benefit from nutrient reductions sufficient to reduce or eliminate hypoxia, but ingeneral, systemwide fisheries are not likely to increase as nutrients are reduced. An overestimateof hypoxia effects on fisheries risks drawing attention away from critical stressors that are theproximate causes of declines in exploited populations.

SUMMARY POINTS

1. Hypoxia is a widespread and increasing environmental problem that is caused by a com-bination of anthropogenic nutrient enrichment and physical characteristics of systems.Nitrogen loadings and the spatial extent of hypoxia were not, however, correlated in acomparison of 30 estuaries and semienclosed seas.

2. Hypoxia has clear consequences for individual fish and mobile shellfish, as well as for foodwebs and fisheries at the local scale, i.e., in areas immediately affected by oxygen-depletedwater.

3. Recent research has demonstrated the importance of chronic exposure to moderate levelsof hypoxia and of fluctuating oxygen concentrations to immune function, reproduction,and growth.

4. The strongest evidence for negative effects of hypoxia at the scale of populations andfisheries comes from riverine portions of estuaries that receive raw sewage, systems witha very high percentage of hypoxic habitat, and cases where interacting stressors or lossof critical habitat affect particular species.

5. Analyses of N, hypoxia, and fisheries landings in 30 estuaries and semienclosed seassuggest that hypoxia does not generally reduce fisheries landings below what would bepredicted by nitrogen loading rates. An exception is a possible decline in landings ofbenthic species in systems with ≥40% of bottom area hypoxic. Systems with seasonalhypoxia had high fisheries landings overall and high landings relative to N loading rates.


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


6. Several compensatory mechanisms—including the balance between enhanced produc-tion and habitat loss, the potential for fishing mortality to keep populations below habitat-determined carrying capacity, and the effect of nutrient-related turbidity on juvenile andforage fish survival—may limit the translation of effects of hypoxia demonstrated at thelocal scale to effects on systemwide populations or fisheries.

7. Translating effects of hypoxia across spatial scales will require better understanding ofthe compensatory mechanisms that influence hypoxia sensitivity at all levels of biologicalorganization.

FUTURE ISSUESAdvances in predicting effects of hypoxia on the abundance of living resources, the sustain-ability of fisheries that target mobile species, and fishing communities, as well as developingappropriate management strategies, require that we:

1. Improve our understanding of compensatory mechanisms that modify the effects ofnutrient-sensitive hypoxia on populations and fisheries at the whole-system scale, anddetermine whether the compensatory role of these mechanisms will continue to be pro-tective as nutrient loadings are reduced.

2. Develop predictive conceptual and quantitative models that can consider the separate aswell as combined effects of nutrient enrichment and hypoxia.

3. Quantify the economic and social costs of localized hypoxia.

4. Determine whether the effects of hypoxia on populations and fisheries are more severein nations with developing economies, especially those in low latitudes where densitystratification may persist year-round.

5. Develop improved fisheries management strategies that account for hypoxia in bothspatial management and landings regulations.

6. Improve our understanding of how hypoxia alone and in combination with nutrientoverenrichment and other stressors affects disease transmission and severity in the field.

DISCLOSURE STATEMENT

The authors are not aware of any biases that might be perceived as affecting the objectivity of thisreview.

ACKNOWLEDGMENTS

The cross-system comparison we present could not have been possible without the help of nearlyone hundred agencies and colleagues from around the world who generously provided their dataand advice, and ran additional analyses to help with this effort. These people are listed in theSupplemental Material. Funding for this project was provided by NOAA-Center for SponsoredCoastal Ocean Research and the Smithsonian Marine Science Network.


Supplemental Material

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.

http://arjournals.annualreviews.org/doi/suppl/10.1146/annurev.marine.010908.163754


LITERATURE CITED

Albert R. 1988. The historical context of water quality management for the Delaware estuary. Estuaries Coasts11:99–107

Araujo FG, Williams WP, Bailey RG. 2000. Fish assemblages as indicators of water quality in the MiddleThames Estuary, England (1980–1989). Estuaries 23:305–17

Aumann CA, Eby LA, Fagan WF. 2006. How transient patches affect population dynamics: the case of hypoxiaand blue crabs. Ecol. Monogr. 76:415–38

Baird D, Christian RR, Peterson CH, Johnson GA. 2004. Consequences of hypoxia on estuarine ecosystemfunction: energy diversion from consumers to microbes. Ecol. Appl. 14:805–22

Bell GW, Eggleston DB, Wolcott TG. 2003. Behavioral responses of free-ranging blue crabs to episodichypoxia. I. Movement. Mar. Ecol. Prog. Ser. 259:215–25

Boesch DF, Coles VJ, Kimmel DG, Miller WD. 2007. Ramifications of climate change for Chesapeake Bayhypoxia. In Regional Impacts of Climate Change: Four Case Studies in the United States. Arlington, VA: PewCenter on Global Climate Change, 54–70 pp.

Breitburg DL. 1992. Episodic hypoxia in Chesapeake Bay: interacting effects of recruitment, behavior, andphysical disturbance. Ecol. Monogr. 62:525–46

Breitburg DL. 2002. Effects of hypoxia, and the balance between hypoxia and enrichment, on coastal fishesand fisheries. Estuaries 25:767–81

Breitburg DL, Adamack A, Kolesar SE, Decker MB, Rose KA, et al. 2003. The pattern and influence of lowdissolved oxygen in the Patuxent River, a seasonally hypoxic estuary. Estuaries 26:280–97

Breitburg DL, Craig JK, Fulford RS, Rose KA, Boynton WR, et al. 2008. Nutrient enrichment and fisheriesexploitation: interactive effects on estuarine living resources and their management. Hydrobiologia. Inpress

Breitburg DL, Loher T, Pacey CA, Gerstein A. 1997. Varying effects of low dissolved oxygen on trophicinteractions in an estuarine food web. Ecol. Monogr. 67:489–507

Brosnan TM, O’Shea ML. 1996. Long-term improvements in water quality due to sewage abatement in thelower Hudson River. Estuaries Coasts 19:890–900

Burgents JE, Burnett LE, Stabb EV, Burnett KG. 2005. Localization and bacteriostasis of Vibrio introducedinto the Pacific white shrimp, Litopenaeus vannamei. Dev. Comp. Immunol. 29:681–91

Caddy JF. 1993. Toward a comparative evaluation of human impacts on fishery ecosystems of enclosed andsemienclosed seas. Rev. Fish. Sci. 1:57–95

Caddy JF. 2000. Marine catchment basin effects versus impacts of fisheries on semi-enclosed seas. ICES J.Mar. Sci. 57:628–40

Cardinale M, Modin J. 1999. Changes in size-at-maturity of Baltic cod (Gadus morhua) during a period oflarge variations in stock size and environmental conditions. Fish. Res. 41:285–95

Choi K, Lehmann DW, Harms CA, Law JM. 2007. Acute hypoxia-reperfusion triggers immunocompromisein Nile tilapia. J. Aquat. Anim. Health 19:128–40

Cloern JE. 2001. Our evolving conceptual model of the coastal eutrophication problem. Mar. Ecol. Prog. Ser.210:223–53

Coutant CC. 1985. Striped bass, temperature, and dissolved oxygen: a speculative hypothesis for environmentalrisk. Trans. Am. Fish. Soc. 114:31–61

Craig JK, Crowder LB. 2005. Hypoxia-induced habitat shifts and energetic consequences in Atlantic croakerand brown shrimp on the Gulf of Mexico shelf. Mar. Ecol. Prog. Ser. 294:79–94

Deegan LA. 2002. Lessons learned: the effects of nutrient enrichment on the support of nekton by seagrassand salt marsh ecosystems. Estuaries 25:727–42

de Leiva Moreno JI, Agostini VN, Caddy JF, Carocci F. 2000. Is the pelagic-demersal ratio from fisherylandings a useful proxy for nutrient availability? A preliminary data exploration for the semienclosed seasaround Europe. ICES J. Mar. Sci. 57:1091–102

De Robertis A, Ryer CH, Veloza A, Brodeur RD. 2003. Differential effects of turbidity on prey consumptionof piscivorous and planktivorous fish. Can. J. Fish. Aquat. Sci 60:1517–26

Diaz RJ. 2001. Overview of hypoxia around the world. J. Environ. Qual. 30:275–81


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


Diaz RJ, Rosenberg R. 2008. Spreading dead zones and consequences for marine ecosystems. Science 321:926–29

Domenici P, Ferrari RS, Steffensen JF, Batty RS. 2002. The effect of progressive hypoxia on school structureand dynamics in Atlantic herring Clupea harengus. Proc. R. Soc. London 269:2103–11

Domenici P, Lefrancois C, Shingles A. 2007. Hypoxia and the antipredator behaviours of fishes. Phil. Trans.R. Soc. Lond. 362:2105–21

Druon J-N, Schrimpf W, Dobricic S, Stips A. 2004. Comparative assessment of large-scale marine eutrophi-cation: North Sea area and Adriatic Sea as case studies. Mar. Ecol. Prog. Ser. 272:1–23

Duarte CM. 2002. The future of seagrass meadows. Environ. Conserv. 29:192–206Dutil J-D, Sylvestre E-L, Gamache L, Larocque R, Guderley H. 2007. Burst and coast use, swimming per-

formance and metabolism of Atlantic cod Gadus morhua in sublethal hypoxic conditions. J. Fish Biol.71:363–75

Eby LA, Crowder LB. 2002. Hypoxia-based habitat compression in the Neuse River Estuary: context-dependent shifts in behavioral avoidance thresholds. Can. J. Fish. Aquat. Sci. 59:952–65

Eby LA, Crowder LB. 2004. Effects of hypoxic disturbances on an estuarine nekton assemblage across multiplescales. Estuaries 27:342–51

Elmgren R. 1989. Man’s impact on the ecosystem of the Baltic Sea: energy flows today and at the turn of thecentury. Ambio 18:326–32

Fallesen G, Andersen F, Larsen B. 2000. Life, death and revival of the hypertrophic Mariager Fjord, Denmark.J. Mar. Syst. 25:313–21

Fischer P, Kils U, Walker U. 1990. In-situ investigations on the behaviour and respiration of the three-spinedstickleback (Gasterosteus aculeatus L.) and the eelpout (Zoarces viviparus L.). ICES Council Meet. (Collect.Pap.) pp. 1–10, Copenhagen, Denmark

Gilbert D, Sundby B, Gobeil C, Mucci A. 2005. A seventy-two-year record of diminishing deep-water oxygenin the St. Lawrence Estuary: the northwest Atlantic connection. Limnol. Oceanogr. 50:1654–66

Gray JS, Wu RS-S, Or YY. 2002. Effects of hypoxia and organic enrichment on the coastal marine environment.Mar. Ecol. Prog. Ser. 238:249–79

Helly JJ, Levin LA. 2004. Global distribution of naturally occurring marine hypoxia on continental margins.Deep. Sea Res. Oceanogr. Res. Pap. 51:1159–68

Hilborn R. 2007. Managing fisheries is managing people: What has been learned? Fish Fish. 8:285–96Holman JD, Burnett KG, Burnett LE. 2004. Effects of hypercapnic hypoxia on the clearance of Vibrio campbellii

in the Atlantic Blue Crab, Callinectes sapidus Rathburn. Biol. Bull. 206:188–96Howarth RW, Marino R. 2006. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosys-

tems: evolving views over three decades. Limnol. Oceanogr. 51:364–76Ishibashi Y, Inoue K, Nakatsukasa H, Ishitani Y, Miyashita S, Murata O. 2005. Ontogeny of tolerance to

hypoxia and oxygen consumption of larval and juvenile red sea bream, Pagrus major. Aquaculture 244:331–40

Jackson JBC, Kirby MX, Berger WH, Bjorndal KA, Botsford LW, et al. 2001. Historical overfishing and therecent collapse of coastal ecosystems. Science 293:629–38

Jeftic L, Matte-Baker A, Schomaker M. 2006. The State of the Marine Environment: Trends and Processes. TheHague, The Netherlands: UNEP/GPA Coordination Office

Jones PD. 2006. Water quality and fisheries in the Mersey estuary, England: a historical perspective. Mar.Pollut. Bull. 53:144–54

Justic D, Rabalais NN, Turner RE. 2003. Simulated responses of the Gulf of Mexico hypoxia to variations inclimate and anthropogenic nutrient loading. J. Mar. Syst. 42:115–26

Keister JE, Houde ED, Breitburg DL. 2000. Effects of bottom layer hypoxia on abundances and depthdistributions of organisms in Patuxent River, Chesapeake Bay. Mar. Ecol. Prog. Ser. 205:43–59

Kodama K, Aoki I, Taniuchi T, Shimizu M. 2002. Long-term changes in the assemblage of demersal fishesand invertebrates in relation to environmental variations in Tokyo Bay, Japan. Fish. Manag. Ecol. 9:303–13

Kodama K, Horiguchi T, Kume G, Nagayama S, Shimizu T, et al. 2006. Effects of hypoxia on early life historyof the stomapod Oratosquilla oratoria in a coastal sea. Mar. Ecol. Prog. Ser. 324:197–206

Kolesar SE. 2006. The effects of low dissolved oxygen on predation interactions between Mnemiopsis leidyi ctenophoresand larval fish in the Chesapeake Bay ecosystem. PhD thesis. Univ. Maryland. 172 pp.


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


Korkea-aho TL, Partanen JM, Kukkonen JVK, Taskinen J. 2008. Hypoxia increases intensity of epidermalpapillomatosis in roach Rutilus rutilus. Dis. Aquat. Org. 78:235–41

Koster FW, Mollmann C, Hinrichsen H-H, Wieland K, Tomkiewicz J, et al. 2005. Baltic cod recruitment—theimpact of climate variability on key processes. ICES J. Mar. Sci. 62:1408–25

Landry CA, Steele SL, Manning S, Cheek AO. 2007. Long term hypoxia suppresses reproductive capacity inthe estuarine fish, Fundulus grandis. Comp. Biochem. Physiol. Mol. Integr. Physiol. 148:317–23

Lefrancois C, Shingles A, Domenici P. 2005. The effect of hypoxia on locomotor performance and behaviourduring escape in Liza aurata. J. Fish Biol. 67:1711–29

Le Moullac G, Soyez C, Saulnier D, Asquer D, Avarre JC, Levy P. 1998. Effects of hypoxic stress on theimmune response and the resistance to vibriosis of the shrimp Penaeus stylirostris. Fish Shellfish Immunol.8:621–29

Lotze HK, Lenihan HS, Bourque BJ, Bradbury RH, Cooke RG, et al. 2006. Depletion, degradation, andrecovery potential of estuaries and coastal seas. Science 163:1806–9

Maes J, Stevens M, Breine J. 2007. Modelling the migration opportunities of diadromous fish species along agradient of dissolved oxygen concentration in a European tidal watershed. Estuar. Coast. Shelf Sci. 75:151–62

Mee L. 2006. Reviving dead zones. Sci. Am. 295:78–85Mellergaard S, Nielsen E. 1997. Epidemiology of lymphocystis, epidermal papilloma and skin ulcers in com-

mon dab Limanda limanda along the west coast of Denmark. Dis. Aquat. Org. 30:151–63Natl. Res. Counc. Ocean Stud. Board, Natl. Res. Counc. Water Sci. Technol. Board. 2000. Clean Coastal

Waters: Understanding and Reducing the Effects of Nutrient Pollution. Washington, DC: National AcademyPress

N.C. Div. Water Qual. (NCDWQ) 2006. Annual Report of Fish Kill Events, Div. Water Qual. Environ. Sci.Section, Raleigh, NC

Nestlerode JA, Diaz RJ. 1998. Effects of periodic environmental hypoxia on predation of a tethered polychaete,Glycera americana: implications for trophic dynamics. Mar. Ecol. Prog. Ser. 172:185–95

Niklitschek EJ, Secor DH. 2005. Modeling spatial and temporal variation of suitable nursery habitats forAtlantic sturgeon in the Chesapeake Bay. Estuar. Coast. Shelf Sci. 64:135–48

Nissling A, Kryvi H, Vallin L. 1994. Variation in egg buoyancy of Baltic cod Gadus morhua and its implicationsfor egg survival in prevailing conditions in the Baltic Sea. Mar. Ecol. Prog. Ser. 110:67–74

Nissling A, Vallin L. 1996. The ability of Baltic cod eggs to maintain neutral buoyancy and the opportunityfor survival in fluctuating conditions in the Baltic Sea. J. Fish Biol. 48:217–27

Nixon SW, Buckley BA. 2002. “A strikingly rich zone”—nutrient enrichment and secondary production incoastal marine ecosystems. Estuaries 25:782–96

O’Connor T, Whitall D. 2007. Linking hypoxia to shrimp catch in the northern Gulf of Mexico. Mar. Pollut.Bull. 54:460–63

Parker CA, O’Reilly JE. 1991. Oxygen depletion in Long Island Sound: a historical perspective. EstuariesCoasts 14:248–64

Pearce J, Balcom N. 2005. The 1999 Long Island Sound lobster mortality event: findings of the comprehensiveresearch initiative. J. Shellfish Res. 24:691–7

Perez-Dominguez R, Holt SA, Holt GJ. 2006. Environmental variability in seagrass meadows: effects ofnursery environment cycles on growth and survival in larval red drum Sciaenops ocellatus. Mar. Ecol. Prog.Ser. 321:41–53

Petersen JK, Petersen GI. 1990. Tolerance, behaviour, and oxygen consumption in the sand goby, Pomatochistusminutus (Pallas), exposed to hypoxia. J. Fish Biol. 37:921–33

Pihl L, Modin J, Wennhage H. 2005. Relating plaice (Pleuronectes platessa) recruitment to deteriorating habitatquality: effects of macroalgal blooms in coastal nursery grounds. Can. J. Fish. Aquat. Sci. 62:1184–93

Poulin R, Wolf NG, Kramer DL. 1987. The effect of hypoxia on the vulnerability of guppies (Poecilia reticulata,Poeciliidae) to an aquatic predator (Astronotus ocellatus, Cichlidae). Environ. Biol. Fishes 20:285–92

Powers SP, Peterson CH, Christian RR, Sullivan E, Powers MJ, et al. 2005. Effects of eutrophication onbottom habitat and prey resources of demersal fishes. Mar. Ecol. Prog. Ser. 302:233–43

Rabalais NN, Turner RE, eds. 2001. Coastal Hypoxia: Consequences for Living Resources and Ecosystems.Washington, D.C.: American Geophysical Union


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


Rissanen E, Tranberg HK, Sollid J, Nilsson GE, Nikinmaa M. 2006. Temperature regulates hypoxia-induciblefactor-1 (HIF-1) in a poikilothermic vertebrate, crucian carp (Carassius carassius). J. Exp. Biol. 209:994–1003

Rombough PJ. 1997. The effects of temperature on embryonic and larval development. In Global Warming:Implications for Freshwater and Marine Fish. Soc. Exp. Biol. Sem. Ser., ed. CM Wood, DG McDonald,pp. 177–223. Cambridge: Cambridge Univ. Press

Rombough PJ, Randall DJ. 1988. Respiratory gas exchange, aerobic metabolism, and effects of hypoxia duringearly life. In Fish Physiology, ed. WS Hoar, pp. 59–161. New York: Academic

Roussel J-M. 2007. Carry-over effects in brown trout (Salmo trutta): hypoxia on embryos impairs predatoravoidance by alevins in experimental channels. Can. J. Fish. Aquat. Sci. 64:786–92

Sagasti A, Schaffner LC, Duffy JE. 2001. Effects of periodic hypoxia on mortality, feeding and predation inan estuarine epifaunal community. J. Exp. Mar. Biol. Ecol. 258:257–83

Schurmann H, Steffensen JF. 1992. Lethal oxygen levels at different temperatures and the preferred temper-ature during hypoxia of the Atlantic cod, Gadus morhua L. J. Fish Biol. 41:927–34

Secor DH, Gunderson TE. 1998. Effects of hypoxia and temperature on survival, growth, and respiration ofjuvenile Atlantic sturgeon (Acipenser oxyrinchus). Fish. Bull. 96:603–13

Seitz RD, Marshall LS Jr, Hines AH, Clark KL. 2003. Effects of hypoxia on predator-prey dynamics of theblue crab Callinectes sapidus and the Baltic clam Macoma balthica in Chesapeake Bay. Mar. Ecol. Prog. Ser.257:179–88

Selberg CD, Eby LA, Crowder LB. 2001. Hypoxia in the Neuse River Estuary: responses of blue crabs andcrabbers. N. Am. J. Fish. Manag. 21:358–66

Shane MA, Nofziger C, Blazer-Yost BL. 2006. Hormonal regulation of the epithelial Na+ channel: fromamphibians to mammals. Gen. Comp. Endocrinol. 147:85–92

Shang EHH, Wu RSS. 2004. Aquatic hypoxia is a teratogen and affects fish embryonic development. Environ.Sci. Technol. 38:4763–67

Shimps EL, Rice JA, Osborne JA. 2005. Hypoxia tolerance in two juvenile estuary-dependent fishes. J. Exp.Mar. Biol. Ecol. 325:146–62

Shoji J, Masuda R, Yamashita Y, Tanaka M. 2005. Predation on fish larvae by moon jellyfish Aurelia auritaunder low dissolved oxygen concentrations. Fish. Sci. 71:748–53

Sklar FH, Browder JA. 1998. Coastal environmental impacts brought about by alterations to freshwater flowin the Gulf of Mexico. Environ. Manag. 22:547–62

Steffe AS, Macbeth WG, Murphy JJ. 2007. Status of the recreational fisheries in two Australian coastal estuariesfollowing large fish-kill events. Fish. Res. 85:258–69

Stierhoff KL, Targett TE, Miller KL. 2006. Ecophysiological responses of juvenile summer and winter flounderto hypoxia: experimental and modeling analyses of effects on estuarine nursery quality. Mar. Ecol. Prog.Ser. 325:255–66

Tanner CA, Burnett LE, Burnett KG. 2006. The effects of hypoxia and pH on phenoloxidase activity in theAtlantic blue crab, Callinectes sapidus. Comp. Biochem. Physiol. 144:218–33

Taylor JC, Miller JM. 2001. Physiological performance of juvenile southern flounder, Paralichthys lethostigma(Jordan and Gilbert, 1884), in chronic and episodic hypoxia. J. Exp. Mar. Biol. Ecol. 258:195–214

Taylor JC, Rand PS, Jenkins J. 2007. Swimming behavior of juvenile anchovies (Anchoa spp.) in an episodicallyhypoxic estuary: implications for individual energetics and trophic dynamics. Mar. Biol. 152:939–57

Thiel R, Sepulveda A, Kafemann R, Nellen W. 1995. Environmental factors as forces structuring the fishcommunity of the Elbe estuary. J. Fish Biol. 46:47–69

Thomas P, Rahman MS, Khan IA, Kummer JA. 2007. Widespread endocrine disruption and reproductiveimpairment in an estuarine fish population exposed to seasonal hypoxia. Proc. R. Soc. Lond. B Biol. Sci.274:2693–701

Tinsley D. 1998. The Thames estuary: a history of the impact of humans on the environment and a descrip-tion of the current approach to environmental management. In A Rehabilitated Estuarine Ecosystem: TheEnvironment and Ecology of the Thames Estuary, ed. MJ Attrill. London: Kluwer Academic Publishers

Tun N, Houston AH. 1986. Temperature, oxygen, photoperiod, and the hemoglobin system of the rainbowtrout, Salmo gairdneri. Can. J. Zool. 64:1883–88


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.


Tyler RM, Targett TE. 2007. Juvenile weakfish (Cynoscion regalis) distribution in relation to diel-cycling dis-solved oxygen in an estuarine tributary. Mar. Ecol. Prog. Ser. 333:257–69

Tyson RV, Pearson TH. 1991. Modern and ancient continental shelf anoxia: an overview. Geol. Soc. Lond. Spec.Publ. 58:1–24

US EPA. 2003. Ambient Water Quality Criteria for Dissolved Oxygen, Water Clarity and Chlorophyll a for ChesapeakeBay and Its Tidal Tributaries, US Environ. Prot. Agency, Reg. III, Chesapeake Bay Program Off.

Utne-Palm AC. 2002. Visual feeding of fish in a turbid environment: physical and behavioural aspects. Mar.Freshw. Behav. Physiol. 35:111–28

Valenzuela A, Silva V, Tarifeno E, Klempau A. 2005. Effect of acute hypoxia in trout (Oncorhynchus mykiss) onimmature erythrocyte release and production of oxidative radicals. Fish Physiol. Biochem. 31:65–72

Vallin L, Nissling A. 2000. Maternal effects on eggs size and egg buoyancy of Baltic cod, Gadus morhua—implications for stock structure effects on recruitment. Fish. Res. 49:21–37

Vogelbein WK, Zwerner DE, Kator H, Rhodes M, Cardinal J. 1999. Epizootic mycobacteriosis in ChesapeakeBay striped bass. Proc. 24th Annu. East. Fish Health Workshop: 8–11, Atl. Beach, NC

Walsh S, Copeland C, Westlake M, New South Wales Dept. of Primary I, Ballina Fisheries C. 2004. MajorFish Kills in the Northern Rivers of NSW in 2001: Causes, Impacts & Responses. Ballina, Aust.: NSW Dept.Primary Ind., Ballina Fish. Cent.

Wannamaker CM, Rice JA. 2000. Effects of hypoxia on movements and behavior of selected estuarine organ-isms from the southeastern United States. J. Exp. Mar. Biol. Ecol. 249:145–63

Weisberg SB, Wilson HT, Himchak P, Baum T, Allen R. 1996. Temporal trends in abundance of fish in thetidal Delaware River. Estuaries 19:723–29

Welker T, McNulty S, Klesius P. 2007. Effect of sublethal hypoxia on immune response and susceptibility ofchannel catfish, Ictalurus punctatus, to enteric speticemia. J. World Aquacult. Soc. 38:12–23

Wu RS. 2002. Hypoxia: from molecular responses to ecosystem responses. Mar. Pol. Bul. 45:35–45Wu RS, Zhou BS, Randall DJ, Woo NY, Lam PK. 2003. Aquatic hypoxia is a disrupter and impairs fish

reproduction. Environ. Sci. Tech. 37:1137–41


Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.

AR396-FM ARI 13 November 2008 17:35

Annual Review ofMarine Science

Volume 1, 2009

Contents

Wally’s Quest to Understand the Ocean’s CaCO3 CycleW.S. Broecker � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

A Decade of Satellite Ocean Color ObservationsCharles R. McClain � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �19

Chemistry of Marine Ligands and SiderophoresJulia M. Vraspir and Alison Butler � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �43

Particle AggregationAdrian B. Burd and George A. Jackson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �65

Marine Chemical Technology and Sensors for Marine Waters:Potentials and LimitsTommy S. Moore, Katherine M. Mullaugh, Rebecca R. Holyoke,Andrew S. Madison, Mustafa Yucel, and George W. Luther, III � � � � � � � � � � � � � � � � � � � � � � � � � � �91

Centuries of Human-Driven Change in Salt Marsh EcosystemsK. Bromberg Gedan, B.R. Silliman, and M.D. Bertness � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 117

Macro-Ecology of Gulf of Mexico Cold SeepsErik E. Cordes, Derk C. Bergquist, and Charles R. Fisher � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 143

Ocean Acidification: The Other CO2 ProblemScott C. Doney, Victoria J. Fabry, Richard A. Feely, and Joan A. Kleypas � � � � � � � � � � � � � � � 169

Marine Chemical Ecology: Chemical Signals and Cues StructureMarine Populations, Communities, and EcosystemsMark E. Hay � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 193

Advances in Quantifying Air-Sea Gas Exchange and EnvironmentalForcingRik Wanninkhof, William E. Asher, David T. Ho, Colm Sweeney,

and Wade R. McGillis � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 213

vi

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.

AR396-FM ARI 13 November 2008 17:35

Atmospheric Iron Deposition: Global Distribution, Variability,and Human PerturbationsNatalie M. Mahowald, Sebastian Engelstaedter, Chao Luo, Andrea Sealy,

Paulo Artaxo, Claudia Benitez-Nelson, Sophie Bonnet, Ying Chen, Patrick Y. Chuang,David D. Cohen, Francois Dulac, Barak Herut, Anne M. Johansen, Nilgun Kubilay,Remi Losno, Willy Maenhaut, Adina Paytan, Joseph M. Prospero,Lindsey M. Shank, and Ronald L. Siefert � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 245

Contributions of Long-Term Research and Time-Series Observationsto Marine Ecology and BiogeochemistryHugh W. Ducklow, Scott C. Doney, and Deborah K. Steinberg � � � � � � � � � � � � � � � � � � � � � � � � � � 279

Clathrate Hydrates in NatureKeith C. Hester and Peter G. Brewer � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 303

Hypoxia, Nitrogen, and Fisheries: Integrating Effects Across Localand Global LandscapesDenise L. Breitburg, Darryl W. Hondorp, Lori A. Davias, and Robert J. Diaz � � � � � � � � � 329

The Oceanic Vertical Pump Induced by Mesoscaleand Submesoscale TurbulencePatrice Klein and Guillaume Lapeyre � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 351

An Inconvenient Sea Truth: Spread, Steepness, and Skewnessof Surface SlopesWalter Munk � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 377

Loss of Sea Ice in the ArcticDonald K. Perovich and Jacqueline A. Richter-Menge � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 417

Larval Dispersal and Marine Population ConnectivityRobert K. Cowen and Su Sponaugle � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 443

Errata

An online log of corrections to Annual Review of Marine Science articles may be found athttp://marine.annualreviews.org/errata.shtml

Contents vii

Ann

u. R

ev. M

arin

e. S

ci. 2

009.

1:32

9-34

9. D

ownl

oade

d fr

om w

ww

.ann

ualr

evie

ws.

org

by U

nive

rsity

of

Tar

tu o

n 03

/04/

12. F

or p

erso

nal u

se o

nly.

Documents

Hypoxia, Nitrogen, and Fisheries: Integrating Effects ...olli/eutrsem/Breitburg09.pdf · low dissolved oxygen concentrations (hypoxia) in estuaries and semienclosed oxygen-depleted