-
Biogeosciences, 6, 2985–3008,
2009www.biogeosciences.net/6/2985/2009/© Author(s) 2009. This work
is distributed underthe Creative Commons Attribution 3.0
License.
Biogeosciences
Temporal responses of coastal hypoxia to nutrient loading
andphysical controls
W. M. Kemp1, J. M. Testa1, D. J. Conley2, D. Gilbert3, and J. D.
Hagy4
1University of Maryland, Center for Environmental Science, Horn
Point Laboratory, Cambridge, Maryland, USA2GeoBiosphere Science
Centre, Department of Geology, Lund University, Lund,
Sweden3Maurice-Lamontagne Institute, Department of Fisheries and
Oceans, Mont-Joli, Québec G5H 3Z4, Canada4US Environmental
Protection Agency, National Health and Environmental Effects
Laboratory, Gulf Ecology Division,Gulf Breeze, Florida, USA
Received: 26 June 2009 – Published in Biogeosciences Discuss.:
14 July 2009Revised: 25 November 2009 – Accepted: 6 December 2009 –
Published: 15 December 2009
Abstract. The incidence and intensity of hypoxic watersin
coastal aquatic ecosystems has been expanding in re-cent decades
coincident with eutrophication of the coastalzone. Worldwide, there
is strong interest in reducing the sizeand duration of hypoxia in
coastal waters, because hypoxiacauses negative effects for many
organisms and ecosystemprocesses. Although strategies to reduce
hypoxia by de-creasing nutrient loading are predicated on the
assumptionthat this action would reverse eutrophication, recent
analy-ses of historical data from European and North
Americancoastal systems suggest little evidence for simple linear
re-sponse trajectories. We review published parallel
time-seriesdata on hypoxia and loading rates for inorganic
nutrientsand labile organic matter to analyze trajectories of
oxygen(O2) response to nutrient loading. We also assess
existingknowledge of physical and ecological factors regulating
O2in coastal marine waters to facilitate analysis of hypoxia
re-sponses to reductions in nutrient (and/or organic matter)
in-puts. Of the 24 systems identified where concurrent timeseries
of loading and O2 were available, half displayed rela-tively clear
and direct recoveries following remediation. Weexplored in detail 5
well-studied systems that have exhibitedcomplex, non-linear
responses to variations in loading, in-cluding apparent “regime
shifts”. A summary of these analy-ses suggests that O2 conditions
improved rapidly and linearlyin systems where remediation focused
on organic inputs fromsewage treatment plants, which were the
primary drivers ofhypoxia. In larger more open systems where
diffuse nutrientloads are more important in fueling O2 depletion
and where
Correspondence to:W. M. Kemp([email protected])
climatic influences are pronounced, responses to
remediationtended to follow non-linear trends that may include
hystere-sis and time-lags. Improved understanding of hypoxia
re-mediation requires that future studies use comparative
ap-proaches and consider multiple regulating factors. Theseanalyses
should consider: (1) the dominant temporal scalesof the hypoxia,
(2) the relative contributions of inorganic andorganic nutrients,
(3) the influence of shifts in climatic andoceanographic processes,
and (4) the roles of feedback in-teractions whereby O2-sensitive
biogeochemistry, trophic in-teractions, and habitat conditions
influence the nutrient andalgal dynamics that regulate O2
levels.
1 Introduction
Depletion of dissolved oxygen from coastal waters is awidespread
phenomenon that appears to be growing glob-ally (Dı́az and
Rosenberg, 2008; Gilbert et al., 2009; Rabal-ais and Gilbert,
2009). There is considerable interest in thisphenomenon because low
oxygen causes physiological stressfor most marine metazoans. Oxygen
concentrations belowapproximately 30% saturation (“hypoxia”=O2
-
2986 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
alter the oxidation-reduction balance in marine sedimentsand
associated biogeochemical processes, including
couplednitrification-denitrification (e.g., Kemp et al., 1990) and
in-organic phosphorus sorption to metal oxide-hydroxide com-plexes
(Slomp and van Cappellen, 2007; Middelburg andLevin, 2009).
Oxygen depletion occurs at various time and space scalesfrom an
imbalance between biological and physical sourcesand sinks for O2.
In very shallow (1–5 m) tidal rivers andlagoons with well-mixed
water columns that are enrichedwith inorganic nutrients, hypoxic
conditions tend to appearand disappear on short (hours-days)
time-scales (D’Avanzoand Kremer, 1994; Tyler et al., 2009).
Slightly deeper (3–8 m) microtidal systems typically experience
periodic strati-fication that may allow episodic hypoxia to occur
on daily-to-weekly scales with changes in wind-driven mixing
(Parket al., 2007). In deeper (10–50 m) estuaries and shelf
sys-tems with stratified water columns, hypoxia often occurs
dur-ing much of the summer (2–4 mo) season (e.g., Kemp et al.,1992;
Rabalais and Gilbert, 2008). In much deeper (>100 m)coastal seas
and fjords, strongly stratified water columns re-sult in virtually
permanent hypoxia/anoxia, changing in sizeand position with
decadal-scale variations in circulation (e.g.,Zill én et al.,
2008).
There is mounting evidence that eutrophication (i.e.,
an-thropogenic nutrient and organic enrichment of waters)
iscontributing to the expansion of occurrence, intensity,
andduration of hypoxic conditions in coastal waters worldwide(e.g.,
D́ıaz and Rosenberg, 2008; Gilbert et al., 2009; Ra-balais and
Gilbert, 2009). Nutrient additions tend to fertilizegrowth, sinking
and decomposition of phytoplankton in bot-tom waters of estuaries,
bays, and inland seas. For manycoastal systems in the
industrialized regions of the world,there have been major
socio-economic commitments to re-mediate hypoxic zones by reducing
nutrient loading fromthe adjacent catchment and overlying
atmosphere (Boesch,2002; Carstensen et al., 2006). Although
reductions in an-thropogenic nutrient inputs to coastal systems is
the primarymeans that has been employed to remediate hypoxia
associ-ated with eutrophication, biomanipulation approaches
havealso been suggested, including re-establishment of dimin-ished
populations of benthic filter-feeding bivalves that con-sume
phytoplankton directly (e.g., Cerco and Noel, 2007;Petersen et al.,
2008). Engineering solutions (including en-hanced vertical mixing,
increased horizontal exchange, andmechanical air-bubbling) have
also been discussed as optionsfor mitigating human-induced coastal
hypoxia (Stigebrandtand Gustafsson, 2007; Conley et al., 2009c).
Although sub-stantial socio-economic investments have been made to
re-duce hypoxia in many regions worldwide (e.g., Kronvang etal.,
2005), recent analyses of historical data from Europeanand North
American coastal systems suggest little evidencefor simple and
straightforward responses of hypoxia to reme-diation actions
(Duarte et al., 2009; Conley et al., 2009b).
The purpose of this paper is to review published concur-rent
time-series data (or proxies) on hypoxia and inputs ofnutrients and
labile organic matter to analyze trajectories ofO2 response to
reductions in nutrient loading. Where avail-able, we also review
parallel time-series data on key physicaland ecological processes
that might also affect changes O2conditions in coastal marine
waters. To minimize questionsabout data quality, we limited this
review to information pub-lished in peer-reviewed literature. While
most of the data se-ries in case-studies reviewed here include
extended periods(>10 yr) of declining nutrient inputs, a few
were character-ized more by strong interannual variations rather
than long-term trends. The review is structured into seven
sections:(1) introduction, (2) external drivers (3) internal
processes(4) theoretical response trajectories, (5) hypoxia
recovery, (6)complex responses, and (7) concluding comments.
Sections2 and 3 provide background information needed to
interpretobserved responses in terms of physical, biogeochemical
andecological controls. In Section 4 we describe theoretical
tra-jectories along which hypoxia might be expected to respondto
changes in nutrient loading. In the next two sections, wedescribe
case studies of systems showing relatively clear re-sponses to
decreased nutrient loading (Sect. 5), and otherswhere hypoxia has
exhibited complex responses to fluctua-tions and changes in
nutrient loading (Sect. 6). Section 7provides conclusions and
considers implications for remedi-ating and managing low O2 coastal
waters.
2 External drivers
Although recent decades have seen widespread observationsof
hypoxia in diverse coastal systems (e.g., Dı́az and Rosen-berg,
2008), low O2 conditions in different systems varyacross a wide
range of temporal (as well as spatial) scales.Here we provide a
modified scheme for categorizing hypoxicsystems according to scales
of variability, and we discuss themajor external drivers that
regulate O2 conditions by con-trolling physical transport and
mixing as well as ecologicalproduction and consumption of organic
matter.
2.1 Topology of coastal hypoxia
Drawing from previous hypoxia classification schemes basedon
duration and dominant time-scales of low O2 (e.g., D́ıazand
Rosenberg, 2008), we define four broad categories of hy-poxia: (1)
permanent, (2) persistent seasonal, both stratifiedand vertically
mixed, (3) episodic, and (4) diel.Permanenthypoxia occurs primarily
in shelf regions, large fjords, andinland seas in which strong
stratification isolates the bottomlayer of deep water columns
(>100 m), leading to bottomwater hypoxia/anoxia (e.g., Helly and
Levin, 2004; Gilbertet al., 2005) that tends to change only in size
and positionwith annual-to-decadal scale variations in circulation
(e.g.,Helly and Levin, 2004; Zilĺen et al., 2008; Chen et al.,
2007).
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2987
Persistent seasonalhypoxia occurs in many stratified tem-perate
estuarine and shelf regions where the combination ofspring river
discharge and summer heat strengthen stratifi-cation, promote
phytoplankton growth and stimulate respi-ration of sinking organic
matter (e.g., Rabalais and Gilbert,2009). Seasonal hypoxia may also
occur throughout the wa-ter column in shallow well-mixed estuaries
and tidal riversthat are heavily loaded with large inputs of labile
organicmaterial that are respired in warmer months (e.g.,
Soertaertet al., 2006).Episodichypoxia tends to occur at irregular
in-tervals (weeks to decades) in productive, shallow (5–15
m),weakly-stratified microtidal coastal systems that are gener-ally
subjected to wind mixing. These systems are susceptibleto
occasional hypoxic conditions that are terminated by windevents
(e.g., Stanley and Nixon, 1992); however, they maybe prolonged by
extended warm calm weather (Møhlenberg,1999) or exacerbated by
major storm events that deliver largepulses of organic loading
(Peierls et al., 2003).Diel hypoxiatends to occur in shallow
productive lagoons and bays, whennight-time respiration of organic
matter produced during theday exceeds O2 replenishment via air-sea
exchange. Typi-cally, daytime O2 levels in these shallow systems
are high(often supersaturated) because of strong photosynthetic
O2production. Although this paper considers all types of hy-poxia,
it generally focuses on coastal systems withseasonalhypoxia
(persistent or variable) because these systems tendto be
well-studied and are often heavily influenced by an-thropogenic
nutrient enrichment. In contrast, systems withdiel hypoxia are less
studied, while systems withpermanenthypoxia tend to be dominated by
natural processes that aredifficult or impossible to remediate.
2.2 Factors driving physical and ecological processes
In many coastal systems, density stratification is sufficient
tocreate a bottom layer isolated from surface waters and im-pede
downward mixing of O2 from surface waters, therebyreducing physical
replenishment and allowing depletion ofbottom water O2 through
aerobic respiration of accumulatedcarbon (e.g., Kemp et al., 1992,
2005). Buoyancy of theupper layer is increased and stratification
is strengthened byseasonal inputs of fresh water (Boicourt, 1992)
and warmingof surface waters (e.g., Welsh and Eller, 1991).
Relativelyweak stratification in systems such as the Neuse River
estu-ary, Long Island Sound, and Mobile Bay can be disrupted
bytypical summer wind events (e.g., Turner et al., 1987; Stan-ley
and Nixon, 1992; O’Donnell et al., 2008). In any givenyear,
stronger stratification, created by larger freshwater in-put or
warmer surface water, is more resistant to disruptionby wind events
(Lin et al., 2008). Ventilation of bottom-water hypoxia may involve
relatively complex mechanisms,where for example wind stress induces
the straining of den-sity fields (e.g., Scully et al., 2005),
lateral tilting of the py-cnocline (Malone et al., 1986),
alteration of far-field coastalcirculation (e.g., Wiseman et al.,
1997), or interaction with
spring-neap tidal cycles (Sharples et al., 1994). In
stratifiedsystems with estuarine circulation, bottom-water O2
poolsare also replenished by landward transport of O2-rich
waterfrom downstream sources or offshore (e.g., Kuo et al.,
1991;Kemp et al., 1992; Wiseman et al., 2004). Because hypoxiain
stratified coastal systems is confined to the bottom
layer,respiration must be fueled by labile organic matter,
typicallyorganic particles sinking from the upper water column
(e.g.,Hagy et al., 2005; Chen et al., 2007, 2009).
Although water-column stratification is a key control
onpersistent seasonalhypoxia for many systems, other well-mixed
coastal waters experience intermittent or persistent hy-poxic
conditions that are confined to the warm season. Forexample,
vertically mixed shallow brackish tidal rivers andsaline lagoons
may experience relatively continuous sum-mertime low O2
concentrations if they are receiving heavyloads of labile organic
wastes. In industrialized regions ofthe world prior to 1990, and
even today in densely populateddeveloping countries, large
discharges of organic wastes cancreate high rates of O2 demand that
often lead to hypoxicconditions throughout the water column (e.g.,
Andrews andRickard, 1980; Soetaert et al., 2006; Dı́az and
Rosenburg,2008; Yin et al., 2008). If these systems are relatively
turbiddue to suspended sediment inputs and resuspension,
photo-synthesis and associated O2 production would be
severelylight-limited. In this case, vertical mixing of the water
col-umn is typically induced by winds and/or tidal turbulence,and
hypoxia results from a sink-source imbalance wherecommunity
respiration exceeds the rate of O2 replenishmentvia air-water
exchange. In contrast, when shallow, clear-water coastal systems
(e.g., lagoons) receive substantial in-puts of inorganic nutrients,
photosynthetic production (of-ten dominated by benthic plants)
represents an important O2source, leading to diel-scale cycling
between supersaturatedO2 concentrations during the day and hypoxic
conditions atnight (e.g., MacPherson et al., 2007; Tyler et al.,
2009). Al-thoughdiel hypoxiais generally confined to the warmer
sum-mer months, its occurrence and intensity tends to vary
ondaily-to-weekly time-scales associated with periodic
fluctu-ations in sunlight and tides, as well as rain and wind
events(e.g., Shen et al., 2008). There are surprisingly few
reportsof diel-scale hypoxia in the scientific literature; however,
re-cent evidence suggests that this phenomenon is widespreadin
shallow eutrophic waters (e.g., Wenner et al., 2004).
Key ecological controls on seasonal hypoxia in coastalwaters
involve the production and delivery of labile organicmatter to the
region of O2 depletion. The origin of the or-ganic matter that
fuels respiratory O2 sinks can either be fromsources within the
aquatic system or from external sources,including the adjacent
watershed or ocean (Bianchi, 2007).Major external sources of
organic material to coastal wa-ters can be derived from runoff of
terrestrial plant debris,phytoplankton biomass from adjacent
river-borne or oceanic-upwelling sources, and anthropogenic inputs
of particulateand dissolved organics, (e.g., sewage effluents; see
Sect. 5).
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
2988 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
For non-stratified coastal systems, respiration and hypoxiamay
be driven by inputs of dissolved organic matter (e.g.,Andrews and
Rickard, 1980; MacPherson et al., 2007). Tofuel bottom respiration
in stratified waters, however, organicmatter must be in the form of
particles capable of sinking tothe bottom layer. Most bottom water
hypoxia is fueled bysinking of living and detrital phytoplankton
cells, whetherthey are transported from external sources or
produced in-ternally in overlying waters. The high rates of
particulateorganic input generally needed to support bottom-layer
hy-poxia, however, tend to be from algal production in
overlyingwaters driven by inputs of inorganic nutrients from
adjacentsources. Recent reviews of anthropogenic hypoxia
suggestthat O2 depletion in stratified coastal waters is most
oftendriven by nutrient-stimulated production of organic
matter(e.g., D́ıaz and Rosenburg, 2008).
Long-term trends and decadal-scale cycles in climaticforcing can
also exert control over O2 concentrations in bot-tom waters via
changes in temperature, salinity, freshwaterinputs, and wind stress
(Rabalais et al., 2009). For exam-ple, recent increases in water
temperature (e.g., Nixon et al.,2004), which are expected to
continue with increases in at-mospheric CO2 concentrations, will
have direct and indirectconsequences for hypoxia. The direct
effects include de-creased solubility of O2 in water and enhanced
respirationrates, while indirect effects include changes in food
webs re-sulting from spatial and temporal shifts in species
distribu-tion and abundance (e.g., Najjar et al., 2000, 2009; Pyke
etal., 2008). In addition, long-term increases in relative sealevel
occurring in many coastal regions worldwide (Holgateand Woodworth,
2004) may result in elevated bottom wa-ter salinities (Hilton et
al., 2008), thus potentially enhancingstratification and reducing
ventilation of deep waters. Long-term increases or decreases in
freshwater input caused byglobal climate change will influence
coastal hypoxia in manycoastal systems by increasing or decreasing
(respectively) thestratification strength and nutrient delivery
rate (e.g., Justićet al., 2003; Arnell, 1999). Lastly, long-term
trends anddecadal-scale shifts in atmospheric pressure fields and
circu-lation (e.g., Ogi et al., 2003) may alter the magnitude and
di-rection of wind stress, causing changes in vertical mixing
andoxygenation of O2-depleted bottom waters in coastal
systems(e.g., Wilson et al., 2008, Scully 2009; see Sect. 6).
3 Internal ecological processes
Although external forcing of physical and biological pro-cesses
strongly influences coastal ecosystem dynamics andhypoxia
development, internal ecosystem structure and as-sociated processes
are also important. For example, internalprocesses regulate key
biogeochemical fluxes, including pro-duction and consumption of
organic carbon and cycling ofinorganic nutrients. These processes,
which create positiveand negative feedbacks within the ecological
system, can in-
fluence O2 dynamics in coastal water columns (e.g., Kempet al.,
2005). In this section we review key internal biogeo-chemical and
ecological processes that interact with and reg-ulate hypoxic
conditions. This background information isessential for
interpreting and predicting how zones of coastalhypoxia will
respond to changes in external nutrient loading.
Oxygen depletion in most stratified coastal systems is
ul-timately supported by surface layer phytoplankton produc-tion
and particulate sinking to the bottom layer. Herbivorousgrazing in
the upper water column tends to impede sink-ing of algal cells and
detritus to the lower layer. However,most marine zooplankton are
relatively less effective graz-ers compared to large-bodied
cladocerans in lakes, whichcan strongly control phytoplankton
biomass (e.g., Jeppesenet al., 2007). Marine zooplankton (primarily
copepods) areless effective because of lower filtering efficiency
and strongtop-down control by planktivores (e.g., Roman and
Gauzens,1997; Stock and Dunne, 2009). On the other hand, ma-rine
suspension-feeding benthic bivalves can effectively con-trol
phytoplankton growth, especially in shallow coastal sys-tems (e.g.,
Prins et al., 1998; Dame and Olenin, 2005), lead-ing to the
suggestion that mussels, oysters and other reef-forming benthic
bivalves could potentially regulate phyto-plankton sufficiently to
reduce hypoxia in eutrophic coastalsystems (e.g., Officer et al.,
1982; Newell and Ott, 1999).A requirement for this to be effective
is that benthic grazersmust have access to upper mixed layer water
where they cangraze rapidly growing cells and retain organic matter
in theshallow aerobic waters (e.g., Pomeroy et al., 2006; Newellet
al., 2007). Although field-scale documentation of benthicgrazing
impacts mitigating coastal hypoxia is limited, sev-eral modeling
studies have demonstrated potential effective-ness (e.g., Cerco and
Noel, 2007; Banas et al., 2007). Theobservation that substantial
reduction in nutrient loading tocoastal waters can lead to
food-limited conditions for ben-thic bivalves (e.g., Dame and
Prins, 1998) suggests that ben-thic bivalve consumption of excess
phytoplankton produc-tion might help retard development and
maintenance of hy-poxia. Variations in climatic conditions (e.g.,
increased tem-perature and rainfall, can initiate low O2 events
that weakenbenthic filter-feeders, thereby reducing control on
phyto-plankton and further intensifying hypoxia (e.g., Fallesen
etal., 2000; Petersen et al., 2008). Benthic bivalves thus
rep-resent potential negative feedback control on
phytoplanktonwhereby hypoxia tends to increase with declines in
bivalvepopulations (see Sect. 6, Chesapeake Bay).
Bottom water O2 concentrations can influence the balancebetween
decomposition and preservation of organic matterdeposited on the
seafloor through a variety of complex in-teractions (e.g.,
Middelburg and Levin, 2009). The fractionof sinking organic matter
that is incorporated into the sed-iments tends to increase with
deposition rate, possibly be-cause high rates of organic input fuel
O2 depletion, whichretards decomposition. This makes it challenging
to resolvethe relative importance of hypoxia, per se, as a control
on
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2989
Table 1. Characteristic categories of coastal hypoxia defined by
time-scales of variability and partially controlled by water column
depth.See Zhang et al. (2009) for a broader classification.
Table 1. Characteristic categories of coastal hypoxia defined by
time-scales of variability and partially controlled by water column
depth. See Zhang et al. (2009) for a broader classification.
Hypoxia Type Dominant Depth System Types Key Controls Example
Time-Scale Range (m) Systems* (1) “Permanent” Decades- > 300 m
Silled Fjord, Depth, mixing, Black Sea1 Centuries Inland Sea
Stratification Baltic Sea2 Organic input flushing (2) “Persistent
Seasonal”
-Stratified Months ~ 10-100 m Estuary, River flow, Chesapeake
Bay3 Shelf plume Temperature, Pensacola Bay4 Organic input
Changjiang plume5 -Mixed Months ~ 5-15 m Tidal river River flow,
Thames Estuary6 Tidal range, Scheldt Estuary7 Organic input
(3) “Episodic” Weeks- ~ 5-20 m Lagoons, Wind & Tides, Mobile
Bay8 (Intermittent) Years Bays Storms, Neuse Estuary9 Organic input
(4) “Diel” Hours- ~ 1-5 m Lagoons, Wind, Light, DE Inland Bays10
Days Bays Nutrient input, Waquoit Bay11 Organic input *References:
1Mee (2006), 2Conley et al. (2007), 3Kemp et al. (1992), 10Hagy and
Murrell (2007), 5Chen et al. (2007), 6Andrews and Rickard (1980),
7Soertaert et al. (2006), 8Turner et al. (1987), 9Borsuk et al.
(2001), 10Tyler et al. (2009), 11D’Avanzo and Kremer (1994).
44
*References:1 Mee (2006),2 Conley et al. (2007),3 Kemp et al.
(1992),4 Hagy and Murrell (2007),5 Chen et al. (2007),6 Andrews
andRickard (1980),7 Soertaert et al. (2006),8 Turner et al.
(1987),9 Borsuk et al. (2001),10 Tyler et al. (2009),11 D’Avanzo
and Kremer (1994).
decomposition versus physical effects of rapid of burial
(e.g.,Hedges and Keil, 1995). Numerous experiments where natu-ral
organic matter is allowed to decompose under controlledconditions
with and without O2 have been generally incon-clusive (e.g.,
Westrich and Berner, 1984); however, more re-cent laboratory and
field investigations tend to support theidea that decomposition
rates are retarded by absence of O2due to a range of mechanisms
including loss of macrofaunaactivity and sulfide inhibition of
microbial metabolism (e.g.,Middelburg and Levin, 2009). Recent
papers (see Sect. 6,northern Gulf of Mexico (NGOM) hypoxia) have
speculatedthat relatively labile organic matter produced in one
yearcould be buried and preserved under seasonally hypoxic
con-ditions, until it is exposed by subsequent physical
distur-bance in the following year, when decomposition (and O2
de-mand) would increase under aerobic conditions (e.g., Turneret
al., 2008; Bianchi et al., 2008).
Sediment biogeochemical processes, porewater chemistry,and
nutrient recycling are clearly influenced by low watercolumn O2 and
associated sediment oxidation-reduction (re-dox) profiles. For both
nitrogen (N) and phosphorus (P), ben-thic recycling efficiency (the
fraction of inputs of organic N
and P to sediments that efflux back to overlying water) tendsto
increase with decreasing bottom water O2 concentrations(e.g., Kemp
et al., 2005). Particulate organic nitrogen deliv-ered to bottom
water and the sediment surface is decomposedvia hydrolysis
reactions using one of several available termi-nal electron
acceptors (e.g., O2, NO
−
3 , Mn (III, IV), Fe (III),and SO2−4 ), generating inorganic
ions of nitrogen (NH
+
4 ) andphosphorus (PO3−4 ) as end-products (Middelburg and
Levin,2009). In the presence of O2, NH
+
4 tends to be oxidized com-pletely to NO−3 (or to NO
−
2 and N2O) by chemoautotrophicnitrifying bacteria, resulting in
O2 consumption. Althoughnitrification may be limited by NH+4
availability in sedi-ments with low % organics, rates in eutrophic
coastal sys-tems are more often controlled by depth of O2
penetrationinto NH+4 -rich fine-grain organic sediments (e.g.,
Henrik-sen and Kemp, 1988). A substantial fraction of the
NO−3generated in nitrification is generally reduced in
surroundinganaerobic zones via denitrification to gaseous N2 (or
N2O) –forms that are virtually unavailable for assimilation by
plants(e.g., Seitzinger, 1988).
Under conditions with hypoxic overlying water, sedimentswith low
redox levels and high sulfide concentrations favor
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
2990 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
dissimilatory reduction of NO−3 back to NH+
4 over denitri-fication (e.g., Tiedje, 1987) and strongly
inhibit nitrifica-tion (e.g., Joye and Hollibaugh, 1995). Although
anammox(anaerobic oxidation of NH+4 to N2 with NO
−
2 ) may occurwith anoxia, it is limited by availability of NO−2
, and ratestend to be substantially lower than denitrification in
mostcoastal sediments (Revsbech et al., 2006). Thus, hypoxic
andanoxic bottom waters greatly suppress nitrification and cou-pled
nitrification-denitrification rates, causing a higher % oftotal
nitrogen to be recycled to overlying water as NH+4 (e.g.,Kemp et
al., 1990; see Sect. 6). Similar dynamics involvinghypoxia and
PO3−4 recycling are attributable to totally dif-ferent mechanisms.
Under normoxic conditions, dissolvedPO3−4 binds to oxides and
hydroxides of Fe and Mn, formingamorphous solid-phase substances
that are retained in sedi-ments (Froelich, 1988). In contrast,
hypoxic conditions pro-mote reduction of Fe and Mn to soluble
states, thereby re-leasing bound PO3−4 (Froelich et al., 1982). The
presence offree sulfide, which has a very high affinity for binding
siteson Fe and Mn, further promotes rapid release PO3−4 and ef-flux
to overlying waters (e.g., Caraco et al., 1989).
Many benthic invertebrate macrofauna (e.g.,
polychaetes,bivalves, amphipods) are highly susceptible to
physiologicalstresses or mortality from bottom-water hypoxia and
anoxia(e.g., D́ıaz and Rosenberg, 1995; Levin, 2003). Healthy
ben-thic faunal communities can, however, exert strong influenceon
N and P cycling in coastal marine sediments (e.g., Aller,1982).
Although direct excretion by these organisms tendsto increase
nutrient recycling, activities of many species alsoretard recycling
of NH+4 and PO
3−4 by enhanced O2 advec-
tion into sediment porewaters. Macrofauna burrows, tunnelsand
tubes that penetrate (0.2–10 cm) into sediments are ven-tilated by
natural circulation and by active animal pumpingof overlying water
(e.g., Aller, 1982). Macrofaunal ventila-tion tends to stimulate
sediment nitrification and strengthenits coupling to
denitrification by increasing the effective areaof oxic-anoxic
interfaces (e.g., Pelegri and Blackburn, 1995).Enhanced O2
penetration into coastal sediments also retardsdissolution of
Fe-Mn-oxide-hydroxide complexes, promot-ing burial of PO3−4 rather
than release to overlying waters(e.g., Welsh, 2003; Middelburg and
Levin, 2009). Feedingactivities of other benthic fauna can
dramatically alter sedi-ment biogeochemistry by homogenizing or
vertically trans-porting particles within the upper (0–30 cm)
sediment col-umn (e.g., Francois et al., 2001). Field observations
andmodeling studies suggest that vertical mixing of
P-boundparticles can reduce PO3−4 release from sediments to
over-lying water in summer (e.g., DiToro, 2001). In summary,hypoxia
and anoxia can further stimulate NH+4 and PO
3−4 re-
cycling to overlying waters by reducing benthic
macrofaunabioturbation.
Meadows of tidal marsh and seagrass plants effectivelymitigate
eutrophication and hypoxia along the coastal mar-gins through
dissolved nutrient uptake and particulate nutri-
ent trapping (e.g., Kemp et al., 2005). Plant biomass
accu-mulation in marshes and seagrass beds can store 103 moredry
weight (dw) than phytoplankton, with plant stands some-times
exceeding 1000 g dw m−2 (e.g., Valiela, 1995). Inte-grated nutrient
pools contained in these macrophytes planttissues and associated
sediments can dominate coastal bioticnutrient budgets (e.g.,
Bricker and Stevenson, 1996; Kempet al., 2005). These plants can
respond to N and P enrich-ment by incorporating higher nutrient
concentrations intotheir leaves (e.g., Duarte, 1990). In addition,
denitrifica-tion rates in marsh and seagrass sediments are often
muchhigher than those in nearby unvegetated sediments, becauseof
enhanced nitrification associated with O2 transported byroots into
sediments and interception of nitrate-rich ground-water flux from
watersheds (Bricker and Stevenson, 1996).The largest impact that
these plants have on coastal N and Pbudgets is derived from their
intense trapping of suspendednutrient-rich particles (e.g., Kemp et
al., 2005; Boynton etal., 2008). As with benthic macrofauna,
however, marsh andespecially seagrass plants are also highly
vulnerable to nega-tive effects of coastal eutrophication,
including reduced wa-ter clarity (e.g., Orth et al., 2006) and
undermining of below-ground tissue (Darby and Turner, 2008).
4 Theoretical response trajectories
To our knowledge, there are very few mechanistic modelsthat have
effectively predicted responses of eutrophication-induced coastal
hypoxia to remediation, particularly reduc-tions in nutrient
loading. Numerical models, which are oftenused to provide
quantitative guidance to the mitigation pro-cess, generally predict
simple linear reductions in hypoxia inresponse to reduced nutrient
loading (Arhonditsis and Brett,2004) though recent data suggest
that coastal ecosystem re-sponses to nutrient reduction are often
more complex (Duarteet al., 2009). Although relatively simple
models have beenused to hind-cast responses of shallow lakes to
such reme-diation efforts (e.g., Scheffer and Jeppesen, 2007), few
nu-merical forecasts have been documented for coastal
systems(Soetaert and Middelburg, 2009). Detailed retrospective
ob-servations showing how hypoxia has changed with eutroph-ication
abatement is limited to a few coastal systems (e.g.,Dı́az and
Rosenburg, 2008).
A broad range of possible aquatic ecosystem responses tochanges
in nutrient loading (Fig. 1) have been defined fromtheory and
observation (e.g., Scheffer et al., 2001; Zhang etal., 2009). In
the simplest case, responses of hypoxia andother eutrophication
effects might be relatively smooth, con-tinuous and linear, where
effects increase and decrease alongthe same pathway in lock-step
with changes in nutrient in-puts (Fig. 1a). Alternatively, hypoxia
might exhibit littleresponse to an initial increase or decrease in
nutrients untilthe system approaches a “threshold” where relatively
smallchanges in nutrient input cause an abrupt system change
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2991
Increased Nutrient Load
Incr
ease
d H
ypox
ia In
tens
ity
(a) Linear Recovery
(d) Hysteresis with Threshold(c) Hysteresis
(b) Recovery with Threshold
(f ) Shifting Baseline, Hysteresis, Threshold(e) Shifting
Baseline
Fig. 1. Six hypothetical response trajectories of oxygen
conditionsin relation to changes in nutrient load. Trajectories
include:(a)linear relationship of hypoxia to load with immediate
responses,(b)direct threshold response to nutrient load,(c) delayed
(hysteretic)response to nutrient load,(d) threshold response with
hysteresis,(e)linear response with shifted baseline due to changes
in other forcingvariables, and(f) threshold response with
hysteresis with a shiftedbaseline.
(Fig. 1b). In this case, hypoxia again follows the same ba-sic
pathway in response to nutrient increase (eutrophication)and
nutrient decrease (oligotrophication). If, however, nutri-ent
increases change the fundamental ecosystem character –including
trophic structure, habitat conditions, and biogeo-chemical cycles –
the system may follow a distinctly differ-ent trajectory and reach
different endpoints in response to nu-trient input declines (Fig.
1c, d, f). These altered ecosystemsbecome resistant to a change in
state, and relatively largernutrient reductions and longer recovery
times are requiredto induce a complete reversal of
eutrophication-induced hy-poxia and a return to the original state
(Fig. 1, Lanthropand Carpenter, 2008). Many coastal ecosystems are
alsoexperiencing disturbance from other factors (e.g.,
climatechange, fishing harvest, species invasion) that can alter
hy-poxia responses to loading. Thus, these other factors mayhave
caused the “baseline conditions” to change during longintervals
between periods of nutrient increase and decrease.Such baseline
shifts can lead to situations where completerecovery to
pre-eutrophication conditions cannot be read-ily achieved simply
with reduced nutrient loading (Fig. 1e,f), because of other
important environmental changes (e.g.,Duarte et al., 2009).
5 Hypoxia remediation with nutrient reduction
Perhaps the most important assumption of the
hypotheticalresponse trajectories presented above is that nutrient
loadingis the primary driver of hypoxia. Whereas several
publishedcases report time series data documenting
contemporaneousincreases or interannual variations in hypoxia and
nutrientloading (e.g., Hagy et al., 2004; Turner et al., 2006b;
Kaup-pila et al., 2005), there are remarkably few examples withtime
series data that cover periods of both increasing and de-creasing
nutrient loads. For stratified coastal systems, strongexperimental
evidence links nutrient loading, phytoplanktonproductivity, organic
particle sinking, and bottom water O2consumption (e.g., de Vries et
al., 1998); however, the rel-ative balance between this ecological
pathway and bottomlayer ventilation by vertical mixing is seldom
well described.In contrast, the role of labile organic inputs to
shallow well-mixed coastal systems in regulating hypoxia has been
gener-ally well described (Andrews and Rickard, 1980; Soetaert
etal., 2006).
We compiled from the published literature a number ofparallel
time series of both hypoxia indices and nutrient (andorganic
matter) loading (or proxies) for several coastal sys-tems to test
theoretical expectations of system response toremediation.
Available case studies include systems withhypoxia of varying
duration (seasonal, episodic, diel), withdifferent anthropogenic
inputs fueling hypoxia (inorganic ni-trogen or phosphorus, and
labile organic matter), and in dif-ferent system types (well-mixed
tidal rivers and shallow la-goons, as well as stratified estuaries,
inland seas, and conti-nental shelves). Comparisons of observed
responses with hy-pothetical trajectories described above are made
where possi-ble. In general, this analysis suggests that O2
conditions im-proved rapidly and linearly in systems with large
reductionsin discharges of labile organic matter from point sources
thathad been sustaining O2 consumption and hypoxia (Table 2).In
larger stratified systems where diffuse input of inorganicnutrients
was the primary driver of hypoxia through growth,sinking and
decomposition of algal cells, the response to re-mediation tended
to exhibit more complex non-linear behav-iors (Table 2).
Improved and more widely applied secondary sewagetreatment in
the 1960s, 1970s, and 1980s led to major re-ductions in loads of
dissolved and particulate labile organicmaterial (or biochemical
oxygen demand, BOD) to coastalwaters (e.g., Smith et al., 1987).
One striking example isthe inner Thames estuary, which received
high loads of nu-trients and organic matter from two major London
sewagetreatment plants (STP) through the 1960s and 1970s,
causingsummer dissolved O2 to remain well below saturation lev-els
(e.g., Andrews and Rickard, 1980) for a stretch of river(>20 km)
seaward of London Bridge (Tinsley, 1998). In-stallation of
secondary treatment at the major STPs reducedBOD loads by 80% in
the early 1970s, resulting in a returnof non-hypoxic summer O2
levels (Fig. 2), followed by a
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
2992 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
Table 2. Summary of reported changes in dissolved oxygen for
coastal ecosystems in the peer-reviewed literature, including the
dominantloading source, target pollutant, and suggested trajectory
of the response.
Table 2: Summary of reported changes in dissolved oxygen for
coastal ecosystems in the peer-reviewed literature, including the
dominant loading source, target pollutant, and suggested trajectory
of the response.
Trajectory System Hypoxia Type Loading Source Target1 Response 2
Type Reference Baltic Sea proper Permanently Stratified Diffuse/
Point Nut. No Red. None Conley et al., 2009a Boston Harbor Seasonal
Stratified Point Nut./BOD + Unknown3 Díaz et al., 2008 Charlotte
Harbor Seasonal Stratified Point Nutrients + Unknown3 Turner et
al., 2006a Chesapeake Bay Seasonal Stratified Diffuse/ Point Nut. -
Reg. Shift4 Hagy et al., 2004 Danish Coastal Waters Seasonal
Stratified Diffuse/Point Nut. None Reg. Shift4 Conley et al., 2007
N. Gulf of Mexico Seasonal Stratified Diffuse Nut. No Red. Reg.
Shift4 Turner et al., 2008 Delaware estuary Seasonal Mixed Point
Nut./BOD + Linear5 Patrick, 1988 East River Seasonal Mixed Point
Nut./BOD + Linear5 Parker and O’Reily, 1991 Forth estuary Seasonal
Mixed Point Nut./BOD + Unknown5 Balls et al., 1996 Lajaalahti Bay
Seasonal Mixed Point Nut./BOD + Linear5 Kauppila et al., 2005 Los
Angeles Harbor Seasonal Mixed Point Nut./BOD + Unknown3 Reish, 2000
Lower Hudson Seasonal Stratified Point Nut./BOD + Linear5 Brosnan
and O’Shea, 1996 Lower Patuxent estuary Seasonal Stratified Point
Nut. - Unknown3 Testa et al., 2008 Mersey estuary Seasonal Mixed
Point BOD + Unknown3 Jones, 2006 New River estuary Seasonal
Stratified Point Nut. + Unknown3 Mallin et al., 2006 New York
Harbor Seasonal Stratified Point Nut./BOD + Linear5 Parker and
O’Reily, 1991 Nervión estuary Seasonal Stratified Point Nut./BOD +
Unknown3 Borja et al., 2006 NW Shelf Black Sea Seasonal Stratified
Diffuse/ Point Nut. + Hysteresis Mee, 2006 Raritan Bay Seasonal
Mixed Point Nut./BOD + Linear5 Parker and O’Reily, 1991 Scheldt
estuary Seasonal Mixed Diffuse/ Point Nut./BOD + Linear5 Soetaert
et al., 2006 Thames estuary Seasonal Mixed Point BOD + Threshold
Andrews and Rickard, 1980 Upper Patuxent estuary Seasonal
Stratified Point Nut. + Linear5 This study Upper Potomac estuary
Seasonal Stratified Point Nut + Linear5 Kemp et al., 2005 Western
LIS Seasonal Stratified Point Nut./BOD - Hysteresis Wilson et al.,
2008
1Nut. = Nutrients, BOD = biochemical oxygen demand 2+ =
improvement, - = degradation, None = no response, No Red. = No
reduction in load 3Unknown = data too limited to reveal both
degradation and recovery trajectory 4See Conley et al. 2009b for
regime shift analysis
45
1 Nut. = Nutrients, BOD = biochemical oxygen demand.2 + =
improvement,− = degradation, None = no response, No Red. = No
reduction in load.3 Unknown = data too limited to reveal both
degradation and recovery trajectory.4 See Conley et al., 2009b for
regime shift analysis.5 Relationship of O2 and nutrient loading was
significantly and linearly related or publication cited a rapid
response to remediation.
recovery of fish, benthic macroinvertebrates, and benthic al-gal
communities (Andrews and Rickard, 1980). Although nodata were
available to describe the time course of degrada-tion (hypoxia
development), the remediation response of O2(% saturation) versus
BOD load suggests threshold behavior,where O2 conditions improved
slowly until∼70% of the loadwas removed, followed by rapid response
to the final 30% ofBOD removal (Fig. 2). The explanation for this
threshold isunclear; however, it may reflect that community
respiration
had been saturated with respect to organic loading, whereO2
levels began to increase only after loading decreased suf-ficiently
for respiration to become substrate-limited. Alter-natively, a
decline in turbidity following waste load reduc-tions may have been
sufficient to allow net photosynthesisby benthic algae, which would
augment O2 replenishmentduring summer (Andrews and Rickard, 1980).
A final pos-sibility is that BOD loading from other, upstream STPs
re-mained high (and kept O2 low) during the initial phase of
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2993
0
100
200
300
400
500
600
1955 1960 1965 1970 1975
Effe
ctiv
e B
OD
Loa
d (to
nnes
/d)
0
10
20
30
40
50
O2
(% S
atur
atio
n)BOD Load
O2 (% Saturation)
50
60
70
80
90
100
110
120
0 100 200 300 400 500 600Total Effective BOD Load (tonnes/d)
O2 D
efic
it (%
of S
atur
atio
n)
19551971
1978
(a)
(b)
Fig. 2. Time series (1955–1980) of observations in the
upperThames River estuary (England) for(a) BOD load from
majorsewage treatment plants and summer O2 % saturation, and(b)
rela-tionship of O2 % saturation deficit (concentration units below
meansaturation) to BOD load showing threshold response of O2 to
re-duced BOD load. Data are from Andrews and Rickard (1980).
the remediation. Several other recently published time se-ries
data from shallow well-mixed estuaries have also docu-mented
hypoxia responses to increases and decreases in BODloading from
STPs (Table 2). These case studies, includingthe Delaware River
estuary (Patrick, 1988), the lower Hud-son River and adjacent
estuaries (Brosnan and O’Shea, 1996;O’Shea and Brosnan, 2000), and
the Mersey estuary (Jones,2006), have generally reported relatively
positive and rapidresponses to reduced BOD inputs.
The Scheldt estuary is another example of a shallow, tur-bid,
and eutrophic upper estuarine system that respondedstrongly to
changes in both nutrient and organic matter load-ing (Soetaert et
al., 2006). The tidal Scheldt is a macroti-dal, relatively shallow
(∼10–12 m) estuary that received in-creasing nutrient and organic
loads through the 1970s. De-clining O2 levels during this period
were linearly related toloading and associated both with oxidation
of NH+4 (nitrifi-cation) and respiration of organic matter (Fig.
3). It appearsthat nutrient-stimulated algal production was not an
impor-tant source of organic matter to fuel O2 depletion in this
sys-tem because of high turbidity and associated light
limitation.When improved sewage treatment reduced BOD loads in
the
20
40
60
80
100
20 40 60 80 100DIN (% MAX)
O2 D
efic
it In
dex
(% M
AX) ‘68-’70
‘65-’67
‘71-’73
‘80-’82
‘77-’79‘74-’76
‘86-’97
‘83-’85
‘01-’02
‘98-’00
O2 deficit
DIN
BOD
Year1965 1975 1985 1995 2005
0
20
40
60
80
100
120
O2 D
efic
it or
5-D
ay B
OD
(% M
AX)
0
20
40
60
80
100
120
DIN
or
NH4+
Fra
ctio
n of
DIN
(% M
AX)
NH4+ Fraction
(a)
(c)
(b)
Fig. 3. Time series (1965–2002) of observations at a brackish
waterstation near the Dutch-Belgian border in the Scheldt River
estuary(Netherlands) for(a) BOD and O2 % saturation deficit and
for(b)DIN concentrations and fraction of DIN as NH+4 and(c)
relation-ship of O2 deficit index to DIN concentration, showing a
favorablelinear shifting baseline response of O2 to reduced DIN and
BODloads. Data are from Soetaert et al. (2006).
mid-1970s, O2 returned to pre-load levels (Fig. 3) over a
20-year period. We explored relationships between O2 deficitsand
DIN concentrations (proxy for nutrient loading) in theupper Scheldt
to examine the response trajectory of O2 to nu-trient loading (Fig.
3). Although data on N loading were notavailable (Soetaert et al.,
2006), both DIN and phytoplanktonchlorophyll-a declined in the
upper Scheldt during the periodof reported nutrient load reduction.
Plots of O2 deficit ver-sus DIN concentration revealed that the
slope of O2 increasewith DIN reduction was flatter than the slope
of O2 declineduring increasing DIN (Fig. 3). This trajectory
reveals a “fa-vorable” shifted baseline scenario that appears to be
related,in part, to a declining ratio of NH+4 to NO
−
3 loading and con-comitant BOD loading reductions (Fig. 3)
during this period
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
2994 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
(Soetaert et al., 2006). A direct but more complex responseto
nutrient load reduction has also been reported for the tidalfresh
region of the Scheldt (Cox et al., 2009). This success
ofremediation in the Scheldt despite complicating changes
innutrient ratios, biogeochemistry, and climate (Soetaert et
al.,2006) underscores the ability to control low O2 conditions
inshallow tidal estuaries dominated by point source inputs
ofnutrients and labile organics.
Laajalahti Bay is a shallow (∼2.4 m), well-mixed, semi-enclosed
estuary adjacent to Helsinki and connected to theGulf of Finland by
a series of straits and sounds (Kauppilaet al., 2005). Although the
well-mixed nature of the estu-ary generally prevented anoxia from
occurring in this estuary,hypoxic conditions were relatively common
during the mid1960s when nutrient and BOD loads were highest (Fig.
4).Improved sewage treatment in the late 1960s led to a
steepdecline in nutrient and BOD loading to the estuary, and
O2levels increased linearly following a decline in BOD loadingthen
stabilized in the early 1980s (Fig. 4). Further remedia-tion
occurred when STP effluents were diverted to the outerarchipelago
in the mid-1980s, resulting in a second increasein O2 to near
saturation levels (Fig. 4). This second O2 in-crease was
significantly and linearly correlated with a declinein
chlorophyll-a, which was in turn related to decreased TNloading.
Improvements in O2 were, however, likely also af-fected by removal
of the remaining BOD loads following thediversion (Kauppila et al.,
2005). Both phases of remedia-tion in Laajalahti Bay caused a
linear increase in O2, one viareduced BOD input and a second linked
primarily to reducedinputs of inorganic nutrients (but also BOD).
These examplessuggest that low O2 conditions in shallow and
well-mixedcoastal systems respond positively and rapidly to
reductionsin both inorganic nutrients and labile organic
matter.
In coastal systems where only nutrient loads were reduced,few
examples exist where data show that hypoxia decreasedmarkedly with
decreased nutrient loading. Where positiveO2 responses have been
documented (e.g., Mallin et al.,2005), increases were relatively
small despite significant de-clines in nutrient concentrations. To
improve O2 conditions,reductions in nutrient loading must first
cause decreases inthe phytoplankton biomass and production that
fuels O2 con-sumption. Although non-linear responses of
phytoplanktonbiomass to nutrient loading reduction have been
reportedfor many coastal systems (e.g., Duarte et al., 2009),
thereis a growing number of examples where reductions in
algalbiomass have been linearly correlated with decreasing
nu-trient loading (e.g., Henriksen, 2009; van Beusekom et
al.,2009). In many large stratified coastal systems, physical
pro-cesses (e.g., wind stress, river flow, and tidal mixing)
playkey roles in O2 depletion, where variations in ventilation
ofbottom waters may dominate the O2 balance and control hy-poxia
formation. Thus, climate-induced changes in circula-tion and mixing
at decadal or longer scales might mask hy-poxia responses to
decreases nutrient loading, even if organicproduction and ecosystem
respiration decline significantly.
6 Complex responses of hypoxia to nutrients
Recent studies have revealed complex, dynamic relation-ships
between hypoxia, nutrient loading, food webs, and cli-mate for a
number of well-studied coastal systems includingChesapeake Bay and
its tributaries (e.g., Hagy et al., 2004;Testa et al., 2008), the
northern Gulf of Mexico (Turner etal., 2009), the Black Sea (Oguz
and Gilbert, 2007), the BalticSea (Conley et al., 2009a), Long
Island Sound (Wilson et al.,2008), the Danish Coastal straits
(Conley et al., 2007), andthe shelf region off the northwestern
United States (Chan etal., 2008). In several cases, the complexity
of responses wasmanifested in terms of the extent of hypoxia water
generatedper unit nitrogen loading. For example, the Patuxent
Riverestuary, Chesapeake Bay, and the northern Gulf of Mexicohave
all exhibited relatively abrupt increases (by more than 2-fold) in
hypoxia per N-loading occurring between 1983 and1993 (Fig. 5).
Although these increases may be caused bydifferent factors in
different systems, the changing hypoxia-loading relationship
underscores the importance of interac-tions among multiple
ecological and physical factors in regu-lating coastal hypoxia. In
this section, we review and analyzeselected case studies toward
improved understanding of hy-poxia responses to remediation in
large coastal ecosystems.
6.1 Patuxent River estuary
The Patuxent River estuary, which is a tributary system
ofChesapeake Bay, is characterized by two-layered
circulation(seaward-flowing surface layer and landward-flowing
bottomlayer) during most of the year. The mesohaline region
hasbroad shoals (10 m) withbottom waters that experience hypoxia
each year in summer(May–September). Large interannual variations in
hypoxiavolume and duration are driven largely by changes in
fresh-water flow and associated nutrient loading and
stratification(Testa et al., 2008). Although episodic low-O2 bottom
wa-ter has been reported since 1940s (Newcombe et al., 1939),large
persistent bottom water hypoxic zones have been evi-dent for the
last 5 decades following increased urbanizationin the upper
watershed (D’Elia et al., 2003).
In an effort to decrease hypoxia and other
eutrophicationeffects, upgrades to sewage treatment reduced loads
of TN(via Biological Nitrogen Removal, or BNR) and TP (viachemical
precipitation) from point sources, thereby reduc-ing total N and P
inputs to the estuary by 25–30% in the mid1980s for P and early
1990s for N (Fig. 6). Associated withnutrient loading reductions,
there have been significant de-creases in DIN and DIP
concentrations throughout the estu-ary, as well as declines in
phytoplankton chlorophyll-a (chl-a) and light-saturated carbon
fixation for the upper estuary(Testa et al., 2008). Bottom water O2
in the upper estuary in-creased rapidly following the nutrient load
reductions, wherehypoxia now rarely occurs during summer (Fig. 6).
Despitethe improvements in the upper estuary and reduced
transport
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2995
0
200
400
600
800
1000
1200
BO
D o
r TN
Loa
d (to
nnes
/yr)
1965 1970 1975 1980 1985 1990 1995 20000
20
40
60
80
100
O2 (
% S
atur
atio
n)or
Chl
-a (m
g/m
3 )
y = -0.06x + 75.88 r2 = 0.77
0
20
40
60
80
100
0 200 400 600 800 1000 1200BOD Load (tonnes/d)
y = -0.36x + 86.61 r2 = 0.51
40
50
60
70
80
90
100
0 20 40 60 80Chl-a (mg/m3)
Dis
solv
ed O
xyge
n (%
Sat
urat
ion)
p
-
2996 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
Patuxent River estuary
Northern Gulf of Mexico
Chesapeake Bay
0
2
4
6H
ypox
ia p
er N
Loa
d (1
09 m
3 y-
1 10
3 kg
N-1
)
0
50
100
150
200
Hyp
oxia
per
N L
oad
(km
2)
0
0.1
0.2
0.3
0.4
0.5
0.6
Hyp
oxia
per
N L
oad
(km3
Gg
N-1
)
1980 1990 2000 2010
1970 1980 1990 2000 2010
1950 1970 1990 2010Year
0
5
10
15
20
25
0 100 200NO3
- Load (106 kg N)
Hyp
oxia
(103
km
2)
0
4
8
12
Jan-May NO3- Load (Gg)
Hyp
oxia
(km
3 )
02468
10
0 1.5 33-NO Load (103 kg d-1)
Hyp
oxia
(109
m3 d
y-1
)
1985-2007
1950-1984
1985-1992
1993-2004
1980-1993
1994-2007
10 20 30 40 50
106 k
g N
-1(a)
(b)
(c)
Fig. 5. Multi-decadal time series data for hypoxia per unit
nitro-gen load in the(a) Patuxent River estuary, USA,(b) northern
Gulfof Mexico, USA, and(c) Chesapeake Bay, USA. Inset figures
arerelationships between N load and hypoxia for each system dur-ing
periods before and after statistically significant change
points(vertical dashed lines) in time-series of hypoxia per unit N
load.Patuxent data are from Testa et al. (2008), Gulf of Mexico
data arefrom Turner et al. (2008), and Chesapeake data are from
Hagy etal. (2004).
been linked to elevated phytoplankton biomass, which in-creased
from the early 1970s to a peak in the 1980s withincreased nitrogen
and phosphorus fertilizer use and loadsfrom the Danube River (Fig.
7, Mee, 2006). Following thecollapse of the former Soviet Union and
Warsaw Pact gov-ernments in Eastern Europe in the late 1980s,
agriculturalsubsidies were greatly reduced, causing abrupt and
substan-tial declines in fertilizer use and animal agriculture. The
re-sulting large decrease in Danube watershed’s N and P fertil-
izer use was quickly followed by the virtual disappearanceof the
associated hypoxic area by 1993 (Fig. 7, Mee, 2006;Oguz and
Gilbert, 2007). Parallel time-series of hypoxia andboth N and P
fertilizer application (proxies for nutrient load-ing) suggest a
threshold response of hypoxia to increasingnutrient input, where
nutrient loads increased for more thana decade before hypoxia
appeared (Fig. 7). While the ex-planation for this threshold is
unclear, Danube River nutrientloading, as well as chl-a in the open
Black Sea (Oguz andGilbert, 2007), have generally followed trends
in fertilizeruse (Behrendt et al., 2005).
As fertilizer use began to decline in the late 1980s, hy-poxic
areas were still observed for∼5 years (1989–1993).This brief
hysteretic time-lag may be linked multiple factors:First, a period
of sustained phytoplankton production on thenorthwestern shelf
occurred in concert with depressed zoo-plankton grazing and
enhanced upward mixing of nutrient-rich deep water into the photic
zone in the adjacent, openBlack Sea (Oguz and Gilbert, 2007).
Secondly, the observedshift in the N:P ratio in the loading may
have caused a shiftfrom N- to P-limitation for phytoplankton growth
on the shelf(Fig. 7; T. Oguz, personal communication, 2009).
Finally,sustained N and P loading may have occurred despite
re-duced fertilizer application, due to residual nutrient
(espe-cially NO−3 ) pools in groundwater and soils (Behrendt et
al.,2005). By the mid-1990s, continued nutrient loading
reduc-tions, recovered zooplankton and benthic communities,
andreduced wintertime mixing appear to have helped decreasethe
extent of hypoxia (Mee, 2006; Oguz and Gilbert, 2007).Although
physical and ecological factors modulated the ef-fects of nutrient
loading on hypoxia in the Black Sea NWshelf, this system represents
a relatively rare example of howa major reduction in diffuse
nutrient loading can dramaticallyreduce or eliminate hypoxia.
6.3 Baltic Sea
The Baltic Sea is a large and permanently stratified estuaryin
northern Europe that receives freshwater inflow from anexpansive
watershed and salt water inflow with North Seaintrusions via the
Kattegat, Skagerrak, and Danish straits.Hypoxia has been present in
the Baltic Sea to varying de-grees since the Holocene, when the
previously freshwaterlake was connected to the adjacent ocean, and
brackish con-ditions were established (Zillén et al., 2008).
Evidence sug-gests that the extent of hypoxic water in the Baltic
has, how-ever, increased since 1950 (Jonsson et al., 1990).
Expandinghypoxic areas appear to have accompanied increased
nutri-ent loads derived from the large and populous Baltic
water-shed, which includes parts of Scandinavia and Eastern Eu-rope
(Conley et al., 2009a). Concerted international effortsto reduce
eutrophication and hypoxia in the Baltic Sea havebeen unsuccessful
thus far. Below we explore the factors re-sponsible for controlling
the extent of hypoxia in this ecosys-tem.
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2997
R2 = 0.32
50
100
p < 0.05
0
2
4
6
8
15
Mnemiopsis Biovolume
ml m
-3R2 = 0.19
R2 = 0.35
10
20
30
40
50
10
20
103
ind.
m-3
85 87 89 91 93 95 97 99 01 03 Year
10
5
µMµg
l-1
Chl-a
DINR2 = 0.61p < 0.01
p < 0.05
p < 0.1A. tonsa
2 -1
0
1
103
kg N
d-1
DIN Load from Ches. BayR2 = 0.37p < 0.05
85 87 89 91 93 95 97 99 01 03 Year
(109
m3 d
y-1
)
0
1
2
3
4
5
(mg
l-1)
Summer Bottom O2
Patuxent Hypoxia
Nutrient Load from WWTP
103 k
g d-
1
(a)
(f )
(e)
(d)
(c)
(b)
(h)
(g)
0
1
2
3
0
0.1
0.2
0.3
0.4
TP
TN
Fig. 6. Time series and trend lines (1985 to 2003) of annual
mean June-August concentrations(a) Mnemiopsissp. biovolume,(b)
Acartiatonsaabundance,(c) chlorophyll-a, and(d) DIN, and(e)mean
annual TN (left axis) and TP (right axis) loads from upstream
sewage plants,(f) upper Patuxent summer (June–August) mean
concentrations of bottom water O2, (g) hypoxic volume days in the
entire Patuxent Riverestuary, and(h) mean annual DIN inputs from
Chesapeake Bay to the Patuxent River estuary. Data are from Testa
et al. (2008).
The time-series data describing Baltic Sea hypoxia spanfrom 1960
to the present (Fig. 8, Conley et al., 2009a). Asteady decline in
the extent of hypoxia from 1970 through1993 is easily apparent
against a small background of inter-annual variability (Fig. 8).
The latter decade of the decline(1985–1995) is associated with a
decrease in total phospho-rus (TP) load and concentration (Fig. 8),
which upon firstinspection would suggest that nutrient load
remediation ef-forts were successful. However, a steady increase in
hy-poxic area and dissolved inorganic phosphorus (DIP) oc-curred
from 1993 to 2000, despite no apparent increase in TPload (Fig. 8).
Increased DIP levels appear to be attributableboth to sustained
external P inputs and to increased internal Ploading, where
iron-bound P was presumably released fromsediments in larger
quantities during the expansion of hy-poxic bottom area from the
mid-1990s to early 2000s (Con-ley et al., 2002; see Sect. 3).
Shifts in climate and physical circulation appear also tohave
been involved in both the decrease in hypoxic area priorto 1993,
and the subsequent hypoxia increase (Fig. 8). TheBaltic Sea proper
is permanently stratified and has fairly lim-
ited exchange with adjacent oceanic waters. Thus, mech-anisms
replenishing O2 to bottom waters via both verticalcross-pycnocline
mixing and water renewal via the Danishstraits are highly
restricted. These processes, which are con-trolled by physical
circulation, may set an upper limit onBaltic hypoxia (Conley et
al., 2009a). The 1993 hypoxiaminimum followed a 10-year period of
low salt water inflowsthrough the Danish straits (Fig. 8). This
served to reducestratification and enhance downward mixing of O2
(Conleyet al., 2009a), as inferred by a steady decline in deep
wa-ter (200 m) salinity in the Gotland Deep through the 1980sto a
minimum in 1992 (Fonselius and Valderrama, 2003).A return of
inflowing salt water, documented in the early1990s, probably
continued and caused increasing deep watersalinity from 1992–2000
(Fig. 8). This would be expectedto increase stratification strength
(Fonselius and Valderrama,2003), contributing to an expanding
hypoxic area (Fig. 8).It appears that changes in water renewal rate
and strati-fication combined with increasing temperature
(Fonseliusand Valderrama, 2003; Omstedt et al., 2004),
persistentlyhigh nutrient loads (Fig. 8), and enhanced P release
from
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
2998 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
Hyp
oxic
Are
a (%
of M
ax) 20
40
60
80
20 40 60 80 100
Danube Basin N or P Fertilizer Use (% of Max)
61-63
85-87
73-75
79-81
00-02
76-78
88-90
64-66
91-93
70-72
0
10
20
30
40
1960 1970 1980 19900
0.5
1
1.5
2
2.5
3
3.5
Hypoxic Area
N Fertilizer Use
Dan
ube
Bas
in N
or P
Fer
tiliz
er U
se
(10
6 ton
s/yr
)
Hyp
oxic
Are
a (1
03 k
m2 )
(a)
(b)
P Fertilizer Use
2000
0
20
40
60
800
(c)
Hypoxia vs. N Fertilizer
Hypoxia vs. P Fertilizer
61-63
85-87
73-75
79-81
00-02
76-78
88-90
64-66
91-93
70-7220 40 60 80 1000
Fig. 7. (a) Time series (1960–2001) of Danube River
watershednitrogen and phosphorus fertilizer use and summer hypoxic
areain the northwest shelf of the Black Sea and response trajectory
ofhypoxic area on the northwest shelf of the Black Sea to
interannualchanges in(b) nitrogen and(c) phosphorus fertilizer use
derivedfrom time series data. Data are from Mee (2006).
sediments (Conley et al., 2002, 2009a) have all contributed
tostable but generally increasing hypoxic zones in this
system.Despite the importance of abiotic controls on current
hypoxiain the Baltic, modeling studies indicate that effective
water-shed nutrient management will help to reduce
eutrophicationand hypoxia (Wulff et al., 2007). The water volume
and com-plex physical circulation of the Baltic suggest, however,
thatsuch a recovery would follow hysteretic pathways with
longtime-lags.
6.4 Chesapeake Bay
Chesapeake Bay is a large estuary in the United States
andreceives freshwater, nutrient, and organic matter inputs
fromseveral rivers, the largest of which is the Susquehanna.
Dra-matic ecological changes have occurred in Chesapeake Bayduring
the past century as a result of nutrient enrichment
Hyp
oxic
Are
a (1
03 k
m2 )
Maj
or B
altic
Inflo
w E
vent
s
D
IP (µ
M)
T
P Lo
ad (1
03 to
n y-
1 )no data
10
20
30
40
2
3
4
5
0
2
4
6
8
10
12
14
0
20
40
60
80
1970 1975 1980 1985 1990 1995 2000
0
10
20
30
40
50
60
70
10
11
12
13
DIN
(µM
)Sa
linity
DIN
DIP
Salinity
Inflows
(a)
(d)
(c)
(b)
Fig. 8. Time-series (1970 to 2000) of(a) annual TP loads to
theBaltic Sea (Conley et al., 2002),(b) annual concentration of
DIP(solid circles) and DIN (open circles) in the Baltic Proper for
depths
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 2999
Because Chesapeake Bay has a large ratio of watershed–to-estuary
area, a deep channel isolated from the atmosphereduring summer
stratification, and a long water-residence-time, it is particularly
susceptible to hypoxia and related ef-fects of eutrophication (Kemp
et al., 2005). Although inci-dents of hypoxia were reported as
early as the 1930s (New-combe and Horne, 1938), direct measurements
and geochem-ical indicators suggest that intense and recurrent
hypoxia andanoxia were uncommon before a rapid increase in
nutrientloading starting in the 1950s (Hagy et al., 2004).
Exten-sive efforts to curb nutrient enrichment of Chesapeake Bayare
reflected in the stabilization and slight decline in springnitrogen
loading from the Susquehanna River since∼1990(Fig. 9). Here we
focus on N-loading because it is viewed asthe primary limiting
nutrient for phytoplankton production(e.g., Fisher et al., 1992,
1999). During this period of de-clining N loading, however, hypoxia
volume has continuedto rise (Hagy et al., 2004), resulting in an
abrupt doubling ofthe hypoxia volume per unit spring N load (Fig.
5).
The cause for this unexpected shift in hypoxia volume
perN-loading is uncertain. However, it generally coincided
withnotable changes in several key factors that may have
con-tributed to the hypoxia increase, including a sharp increasein
water temperature, a potential decrease in nutrient reten-tion in
the upper Bay, and a rapid decrease in oyster abun-dance (indexed
to harvest) and associated filtration capac-ity (Fig. 9). The rapid
increase in surface water tempera-tures (∼0.7◦C) that occurred over
two decades around 1985(Kaushal et al., in press) would be
sufficient to reduce O2saturation level by∼0.20 mg l−1 and possibly
increase respi-ration by∼5–10% (Sampou and Kemp, 1994). The
relativeabundance and filtering capacity of the eastern oyster
(Cras-sostera virginica) in Chesapeake Bay have declined by al-most
100-fold over the past 150 years due to over-fishing andtwo disease
outbreaks (Newell, 1988; Newell and Ott, 1999;Newell et al., 2007).
The drought-induced final decline inoyster harvest during the 10–15
yr around 1985 was∼10% ofthis overall drop between 1900 and the
present (Fig. 9, Kim-mel and Newell, 2007). Recent modeling (Cerco
and Noel,2007) studies have concluded that 10-fold increases in
oys-ter filtration (equivalent to reversing the decline in the
1980s)would induce a 0.3 mg l−1 increase in average mid-Bay bot-tom
water O2 concentration. The absence of any signal ofincreased
phytoplankton corresponding to the oyster decline,however, raises
some doubt about this explanation for the hy-poxia shift. Other
large changes in the Bay, including lossesof marshes and seagrass
beds, were important for Bay nu-trient budgets (Kemp et al., 2005);
however, they are out ofphase with this abrupt increase in
hypoxia.
While many of these changes in the ecosystem may havecontributed
to increased hypoxia in Chesapeake Bay, noneappear to be
sufficiently large and synchronous with hypoxiatrends to have
caused this hypoxia regime shift. The sub-pycnocline recycling of
nitrogen is a key biogeochemicalprocess that has, however,
exhibited a time-course of change
that parallels the hypoxia regime shift trajectory.
Previousstudies have suggested that as O2 decreases, a larger
fractionof the total nitrogen load is recycled from sediments as
NH+4and mixed vertically into the euphotic zone where it can
stim-ulate further algal growth (see Sect. 3). This increase in
NH+4recycling arises because coupled
nitrification-denitrificationis restricted under hypoxic conditions
(Kemp et al., 1990),especially with anoxia-induced loss of benthic
macrofauna(Kemp et al., 2005), which otherwise increase O2
penetrationinto sediments though ventilation of their tube and
burrowhabitats (e.g., Mayer et al., 1995). Preliminary analyses
oftime-series data from mid-Chesapeake Bay suggest an
abruptincrease in mean summer bottom water NH+4 concentrationper
unit spring TN loading that closely follows the trend inhypoxia per
N loading. While this analysis suggests the po-tential for hypoxia
enhancement via benthic N recycling, thissignal is restricted to a
small area of the bay and thus it is dif-ficult to imagine how this
mechanism could have driven theobserved regime shift in hypoxia per
unit N loading.
On the other hand, abrupt changes in atmospheric forcingor
continental shelf circulation might be strong enough to al-ter
vertical or horizontal replenishment of bottom water O2in the Bay.
For example, recent modeling and data analysissuggest that sea
level rise tends to cause increases in salt fluxand bottom-layer
salinity in Chesapeake Bay (Hilton et al.,2008), which could have
increased stratification. Other ev-idence suggests an increase in
the latitude of the north wallof the Gulf Stream since the 1980s
(Taylor and Stephens,1998) that may have reinforced the trend
associated with sealevel rise by causing an increase in salinity at
the Bay mouth(Lee and Lwiza, 2008). A shift from negative to
positive val-ues for the winter North Atlantic Oscillation (NAO)
index inthe late 1970s
(http://www.cgd.ucar.edu/cas/jhurrell/indices.html, Fig. 9) may be
related to the change in the Gulf Streamposition (Taylor and
Stephens, 1998). Such a shift in NAOmight also lead to changes in
the prevailing wind directionand intensity during summer (Ogi et
al., 2003; Scully, 2009),which could affect the strength of
stratification and associ-ated ventilation of hypoxic bottom waters
in summer (e.g.,Malone et al., 1986; Scully et al., 2005). Although
manyof the ecological and biogeochemical factors discussed heremay
have contributed to the initiation and resilience of thishypoxia
regime shift, we conclude that physical factors arelikely involved
in the initiation of this change in hypoxia perN loading.
6.5 Northern Gulf of Mexico
A large region of the Gulf of Mexico’s northwest continen-tal
shelf is highly influenced by outflow of the Mississippiand
Atchafalaya Rivers, with associated buoyancy causingstratification
and river-borne nutrient loads supporting richphytoplankton
productivity (e.g., Bianchi et al., 2008). Un-der present
conditions, bottom-water hypoxia occurs duringsummer months in this
northern Gulf of Mexico (NGOM)
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
http://www.cgd.ucar.edu/cas/jhurrell/indices.htmlhttp://www.cgd.ucar.edu/cas/jhurrell/indices.html
-
3000 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
0
5
10
15
20
14
16
18
0
2
4
6
8
10
12
14
1940 1950 1960 1970 1980 1990 2000 2010 1940 1950 1960 1970 1980
1990 2000 2010
17
15
(wet
mea
t wei
ght 1
06 k
g)
(
o C)
0
10
20
30
40
(km
3 ) (1
02 m
3 s-1
)
0
10
20
30
40
50
(a) Jan-May Susquehanna River Flow
(b) Jan-May Susquehanna River Nitrate Load
(c) July Hypoxic Volume
(Gg)
(f ) MD+VA Oyster Harvest
(e) Jan-March NAO
(d) Water Temp. @ Solomons, MD
Year Year
-1.5
-1
-0.5
0
0.5
1
1.5
2
Fig. 9. Time-series (1945–2007) of(a) Susquehanna River Flow at
Conowingo, MD,(b) January–May NO−3 loads to Chesapeake Bay,(c) July
hypoxic volume in Chesapeake Bay,(d) mean annual water temperature
at Solomons, MD,(e) January to March North AtlanticOscillation,
and(f) annual Maryland (MD) and Virginia (VA) oyster harvest. All
data are from Kemp et al. (2005), except temperature data(Kaushal
et al., 2009) and NAO data
(http://www.cgd.ucar.edu/cas/jhurrell/indices.html).
region. This low-O2 zone, which was first noted in the
early1970s, has been mapped annually since the mid-1980s
(e.g.,Rabalais et al., 2007). Fossil records of
low-oxygen-tolerantforaminifera and modeling studies, however,
suggest that in-frequent hypoxia probably occurred back into the
1950s (Os-terman et al., 2005; Greene et al., 2009), but that
large-scalehypoxic zones likely did not occur prior to the 1970s
(Scaviaet al., 2003; Justić et al., 2005; Turner et al., 2006b;
Greeneet al., 2009). A significant trend of increasing areal extent
ofNGOM hypoxia corresponds strongly with a parallel trendof
increasing nitrogen (N) loads from the Mississippi River(Turner et
al., 2005), driven largely by the growth of intensiveagricultural
production in the Midwest region of the USA(Alexander et al.,
2008). Since the mid-1980s, the July hy-poxic zone has ranged in
size from 40 to 22 000 km2 andhas been correlated with May TN and
NO−3 loading from theMississippi River (Scavia et al., 2004; Turner
et al., 2008). Asmall (17%) decrease in TN loading between 1997 and
2005(Turner et al., 2007), however, produced no significant
de-crease in the extent of July hypoxia (Rabalais et al.,
2007).Recent analyses have shown that, much like the situation
inChesapeake Bay, the extent of hypoxia increased substan-tially in
1993 and thereafter, while N loading has changedvery little,
resulting in an increase in the area of hypoxia ob-served per unit
of N loading (Turner et al., 2008; Greene etal., 2009).
Although there is no clear explanation for this change inhypoxia
per unit N loading, a range of physical and biolog-ical factors may
have contributed to this hypoxia “regimeshift” (Rabalais et al.,
2007). Turner et al. (2008) suggestedthat increased sediment oxygen
demand (SOD) could havecaused the observed rapid increase in
hypoxia. They hypoth-esized that intensified SOD in recent years
was derived fromorganic matter produced in previous years but
preserved andstored in sediments. They further suggested that this
storedorganic matter would be exposed (e.g., through resuspen-sion
events) in subsequent years to fuel enhanced respira-tion, thereby
causing larger than expected hypoxic regionsper unit N loading in a
given year. Empirical support for thishypothesis (e.g., long-term
records of SOD) is limited, andprevious observations from warm
marine systems suggestthat the vast majority of labile organic
matter produced inany year is generally respired within months
(e.g., Burdige,1991). A related hypothesis is that there has been a
recentincrease in organic matter inputs from sources other than
theMississippi-Atchafalaya River System (Bianchi et al., 2008).The
primary postulated source for this increased organic mat-ter input
is cross-shelf transport of material from Louisiana’seroding
coastal wetlands (e.g., Barras et al., 2008). Al-though the general
magnitude of eroding wetland organicmatter fluxes have been
estimated, they have not been di-rectly linked to Gulf hypoxia
(Dagg et al., 2007). While
Biogeosciences, 6, 2985–3008, 2009
www.biogeosciences.net/6/2985/2009/
http://www.cgd.ucar.edu/cas/jhurrell/indices.html
-
W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls 3001
hypoxia per N loading has been increasing, the rate of
coastalland loss has been relatively constant (apart from
hurricane-driven increases in 2005) for the last several decades
(Barraset al., 2008). Others have emphasized the role of
phospho-rus (P) inputs in regulating NGOM hypoxia (e.g., Sylvan
etal., 2006; Dagg et al., 2007). Although statistical
analysisreveals that N loading was by far the best hypoxia
predictor,inclusion of winter P levels in a multiple regression
modeldid improve summer hypoxia predictions (e.g., Greene et
al.,2009). However, the absence of long-term trends in P
loading(Sylvan et al., 2006) indicates that it is not likely
responsiblefor the observed hypoxia shift.
The general open nature of the NGOM system, with un-restricted
boundaries to the south along the Gulf and to thewest across the
broad continental shelf, makes hypoxia de-velopment and control
highly sensitive to physical circula-tion processes. Key physical
controls on NGOM hypoxiainclude freshwater inflow, local
meteorological forcing, andinteractions with the open Gulf
circulation. Of these, onlythe impact of freshwater flow has been
considered in analy-ses of changes and long-term trends in the
extent of hypoxia(e.g., Justíc et al., 2003; Turner et al., 2005;
Greene et al.,2009). One way that a change in other physical
factors couldimpact the extent of hypoxia (and its ratio to N
loading), is bymodulating the flux of freshwater and nutrients
between theshelf and open Gulf (e.g., Bianchi et al., 2006).
Current es-timates indicate that only 43% of the Mississippi River
flowand associated nutrients enters the western shelf, whereas
theremainder is directed eastward or otherwise lost to the openGulf
(Etter et al., 2004). Any change in water or nutrienttransport at
the shelf margin (e.g., Bianchi et al., 2006) wouldlikely impact
the extent of hypoxia independently from theriverine N loading
rate. Summer winds are also impor-tant in regulating NGOM hypoxia,
where up-coast (south-westerly) winds favor retention of freshwater
(plus nutri-ents and plankton) on the Louisiana-Texas shelf and
enhancestratification and hypoxia, until September when
down-coastwinds (northeasterly) promote vertical mixing and
Gulf-shelfexchange (Cochrane and Kelly, 1986). Although the ex-tent
of hypoxia can be reduced quickly when frontal windevents reduce
stratification (Rabalais and Turner, 2006), suchevents are
relatively rare in summer. Clearly, changes inthese physical forces
are important in controlling NGOM hy-poxia; however, there is no
evidence to suggest decadal scalechanges needed to explain the
observed shift in hypoxia perunit N loading.
In summary, a long-term increase in the areal extent ofhypoxia
in the northern Gulf of Mexico since the early1980s is linked
principally to increases in N loading fromthe
Mississippi-Atchafalaya River System. Since 1993, ob-served hypoxic
areas have been larger than expected fromspring N loading alone,
suggesting changes in other eco-logical or physical drivers of
hypoxia. While many alterna-tive explanations have been suggested,
no empirical evidenceand/or modeling analyses have been presented
to support
these hypotheses. Prior to 1993, the statistical
relationshipbetween hypoxia and N-inputs might suggest a
relativelylinear response to proposed reductions in Mississippi
RiverN-loading. However, the recently described and yet
unex-plained shift in this hypoxia versus N relationship now
indi-cates that more complex response trajectories with
hystere-sis, shifting baselines, and time-lags might be
anticipated.
7 Concluding comments
In response to the recently documented worldwide increasein
eutrophication-induced cases of coastal hypoxia, we con-ducted this
review of published case studies, focusing ontemporal trajectories
of hypoxia responses to changes in nu-trient loading for a range of
coastal systems. The purposeof this effort was to expand the
knowledge-base toward im-proved (1) understanding of factors
controlling hypoxia, (2)simulation modeling of observed hypoxia
response trajecto-ries, and (3) strategies for nutrient management
to remediatecoastal hypoxia. For 24 case studies published in the
sci-entific literature, we compared observed trajectories of
hy-poxia response to nutrient loading to diverse patterns
derivedfrom ecological theory. Although our sample of hypoxic
sys-tems with published time series data on hypoxia, nutrientsand
other factors provides a rich and diverse set of examples,it
appears that appropriate temporal data series may also exist(in
unpublished monitoring reports) for other hypoxic
coastalsystems.
In general, we found that shallow well-mixed coastal sys-tems
tended to be more responsive to changes in inorganicand organic
nutrient loading compared to deeper stratifiedestuaries and shelf
regions. Most of the clearest examplesof hypoxia response to
loading reductions are from systemswhere nutrient and organic
matter inputs were dominatedby point sources (e.g., from sewage
treatment plants). Hy-poxia in vertically well-mixed coastal
systems has generallyresponded strongly along linear or threshold
trajectories tolarge reductions in loading of labile dissolved
organic matter(BOD). Other shallow well-mixed systems have also
exhib-ited declines in hypoxia following decreased loading of
inor-ganic nutrients; however, responses were sometimes delayedby
5–10 years. Even for shallow well-mixed systems, thereare few
published cases demonstrating hypoxia responsesto reduced nutrient
loading from watersheds with predomi-nantly diffuse sources.
For deep stratified estuaries, seas or shelf systems, thereare
surprisingly few examples in the published literature doc-umenting
temporal responses of hypoxia to nutrient remedi-ation. The Black
Sea is the one case that is most often citedto demonstrate that
large rapid reductions in nutrient load-ing can elicit rapid
dramatic decreases in bottom-water hy-poxia. Unfortunately,
however, response trajectories for thisexample are based on proxies
for N and P loading and rel-atively sparse measurements of O2
profiles in a very large
www.biogeosciences.net/6/2985/2009/ Biogeosciences, 6,
2985–3008, 2009
-
3002 W. M. Kemp et al.: Coastal hypoxia, nutrient loading and
physical controls
system. Other deep, stratified systems, where hypoxia re-sponses
to reduced nutrient loading have been monitoredgenerally suggest
recalcitrance with respect to hypoxia im-provement. Many of the
best documented time series for hy-poxia and nutrient loading for
these stratified systems haverevealed unexpected complex non-linear
relationships be-tween hypoxia extent and nutrient loading. These
responsesare often characterized by abrupt increases in hypoxia
perunit nutrient loading that have been speculatively attributedto
a range of altered internal ecosystem processes. In manyinstances,
however, evidence is presented suggesting thatchanges in climatic
and/or hydrodynamic conditions coin-cided with the observed
increase in hypoxia and that thesephysical changes may have altered
ventilation of hypoxicbottom waters or transport of nutrients to
the overlying eu-photic zone.
Lessons learned from this comparative analysis of
hypoxiaresponses to changes in nutrient loading will be useful
forcoastal researchers and managers alike. Improved under-standing
of the dominant physical and biological processesis needed to
characterize the natural susceptibility of specificcoastal systems
to hypoxia. Further information is neededconcerning interactions
among climatic trends and cycles,hydrodynamic circulation and
mixing, biogeochemical cy-cles, and food-web relationships.
Analyses should involveapplication of sophisticated statistical
methods and diagnos-tic studies using biophysical numerical models.
Remediationin smaller well-mixed systems dominated by point
sourcesof nutrients are more likely to yield clear and
immediateresults compared to larger stratified systems with
predomi-nantly diffuse nutrient sources. In any case, managers
willneed to adopt realistic expectations regarding the speed,
ex-tent, and nature of hypoxia responses to remediation.
Whatappears to be unsuccessful remediation may simply resultfrom
time-lags in ecological response to nutrient input re-ductions.
Successful remediation may require combinationsof nutrient loading
reductions, restoration of key ecologi-cal habitats, and hydrologic
engineering solutions (e.g., riverflow controls). Effective
sustained hypoxia remediation mayrequire open communication among
researchers, managersand stakeholders.
Acknowledgements.This work was conducted as part of
theactivities of the Scientific Committee on Oceanic Research
(SCOR)Working Group 128 (Natural and Human-Induced Hypoxia
andConsequences for Coastal Areas). We are indebted to the
manyindividuals and agencies responsible for coastal aquatic
monitoringprograms, whose data make such comparative analyses
possible.We would also like to thank Nancy Rabalais, Robert Dı́az,
andJesper Andersen for their thoughtful reviews that greatly
improvedthis manuscrip