Paerl, Hans W., Lexia M. Valdes, Benjamin L. Peierls, Jason E. Adolf, and Lawrence W. Harding, Jr

15 Pages·2006·2.06 MB·English

Limnol.Oceanogr.,51(1,part2),2006,448–462
q2006,bytheAmericanSocietyofLimnologyandOceanography,Inc.
Anthropogenic and climatic influences on the eutrophication of large
estuarine ecosystems
Hans W. Paerl,1 Lexia M. Valdes, and Benjamin L. Peierls
University of North Carolina at Chapel Hill, Institute of Marine Sciences, Morehead City, North Carolina 28557
Jason E. Adolf2 and Lawrence W. Harding, Jr.
University of Maryland Center for Environmental Science, Horn Point Laboratory, P.O. Box 775, Cambridge,
Maryland 21613
Abstract
We examined the effects ofanthropogenic and climaticperturbationsonnutrient–phytoplanktoninteractionsand
eutrophication in the waters of the largest estuarine systems in the U.S.A., the Chesapeake Bay (CB), Maryland/
Virginia, and the Neuse River Estuary/Pamlico Sound (NRE/PS) system, North Carolina. Both systems have ex-
perienced large post-World War II increases in nitrogen (N) and phosphorus (P) loading, and nutrient reductions
have been initiated to alleviate symptoms of eutrophication. However, ecosystem-level effects of these nutrient
reductionsarestronglyaffectedbyhydrologicvariability,includingseveredroughtsandarecentincreaseinAtlantic
hurricane activity. Phytoplankton community responses to these hydrologic perturbations, including storm surges
andfloods,wereexaminedandwhenpossible,comparedforthesesystems.Inbothsystems,theresultingvariability
in water residence time strongly influenced seasonal and longer-term patterns of phytoplankton biomass and com-
munity composition. Fast-growing diatoms were favored during years of high discharge and short residence time
in CB, whereas this effect was not observed during high discharge conditions in the longer residence time NRE/
PS. In the NRE/PS, all phytoplankton groups except summer cyanobacterial populations showed decreased abun-
dance during elevated flow years when compared to low flow years. Although hurricanes affected the CB less
frequentlythantheNRE/PS,theynonethelessinfluencedfloralcompositioninbothsystems.Seasonally,hydrologic
perturbations, including droughts, floods, and storm-related deep mixing events, overwhelmed nutrientcontrolson
floralcomposition.Thisunderscoresthedifficultyinpredictingseasonalandlonger-termphytoplanktonproduction
and compositional responses to nutrient input reductions aimed at controlling eutrophication of large estuarine
ecosystems.
Nearly half the world’s human population resides within munities are being altered. Understanding how human-in-
100 km of the coast, and this proportion is expected to con- duced ecological change (e.g., declining biodiversity, water
tinue to rise in the foreseeable future (Vitousek et al. 1997; quality,andfisheriesresources)interactswithandaffectsthe
National Research Council 2000). Large increases in pollut- structure and function of large estuarine ecosystems adjoin-
ant discharge have accompanied the agricultural and urban ing coastal waters is a major research challenge. The most
developmentofcoastalwatersheds(Peierlsetal.1991;Hop- pervasiveandproblematicanthropogenicinfluencesonthese
kinson and Vallino 1995; Boesch et al. 2001). Deterioration waters include nutrient enrichment (Nixon 1995; Boesch et
ofestuarineecosystemsthatprocessthisburgeoningnutrient al. 2001) and decreases in grazer and higher-consumer pop-
load is accelerating, yet there is a paucity of information on ulations (e.g., shellfish, finfish) (Jackson et al. 2001). Al-
how primary producer and higher-ranked consumer com-
though there is considerable debate over the extent to which
‘‘bottom-up’’ (nutrient enrichment) and ‘‘top-down’’ (graz-
ing) processes influence ecological change, their combined
1Corresponding author ([email protected]).
effects have, in many instances, dramatically increased es-
2Presentaddress:CenterofMarineBiotechnology,Universityof
tuarine and coastal primary production and phytoplankton
Maryland Biotechnology Institute, Baltimore, Maryland, 21202.
biomass, promoting an excess accumulation of organic mat-
Acknowledgments
ter,oreutrophication(Nixon1995).Increasesinphytoplank-
We appreciate the technical assistance and input of J. Fear and
ton production affectnutrient(carbon,nitrogen,phosphorus,
A. Joyner.
and silicon) cycling, water quality, and ecosystem health
H.W.P.,L.M.V.andB.L.P.acknowledgesupportfromtheNation-
alScienceFoundation(DEB9815495andOCE9905723),theU.S. (Smetacek et. al 1991; Conley 1999). Increased nutrient
Dept.ofAgricultureNRIProject00–35101–9981,U.S.EPA-STAR loading to coastal waters has been the primary causative
projectsR82-5243-010andR82867701,NOAA/NorthCarolinaSea factor for increased algal blooms (Paerl 1988), decreases in
GrantProgramR/MER-43,andtheNorthCarolinaDept.ofNatural waterclarity(Cloern2001),andexpandedhypoxia(Rabalais
Resources and Community Development/UNC Water Resources
et al. 1996).
Research Institute (Neuse River Estuary Monitoring and Modeling
Estuarine and coastal ecosystems are also influenced by
Project, ModMon). L.W.H. and J.E.A. acknowledge the assistance
seasonal and multi-annual hydrologic variability (e.g.,
ofM.E.MalloneeandW.D.Miller,andfundingfromtheNSFLand
Margin Ecosystem Research (TIES), NOAA-COP and U.S. EPA- droughts, wet periods, and El Nin˜o vs. La Nin˜a years) and
STAR project R82867701. shorter-term episodic perturbations such as floods, tropical
448
Eutrophication of large estuaries 449
storms, and hurricanes,theintensityandfrequencyofwhich Thegoalofthiscontributionwastoexamineandcompare
appear to be increasing (Goldenberg et al. 2001). The com- how nutrient enrichment and climatic perturbations interact
bined effects of coastal population growth and a recent rise to control phytoplankton composition, distribution, and ac-
inAtlantic(andpossiblyPacific)hurricaneactivity(Golden- tivity in the two largest estuarine ecosystems in the U.S.A.,
berg et al. 2001) is a particularly troubling formula for the the Chesapeake Bay system (CB), Maryland–Virginia and
alteration of coastal watersheds and their receiving waters. the Neuse River Estuary-Pamlico Sound system (NRE/PS),
During 1972, Tropical Storm Agnes delivered record NorthCarolina(Fig.1).Bothsystemshavewell-documented
amounts of rainfall and caused catastrophic flooding in the histories of such anthropogenic and natural perturbations.
ChesapeakeBay(CB)(Fig.1).Theecologicalconsequences
of thisevent,includingadverseeffectsonwaterquality,dis-
ruption of fisheries habitat, and depressed finfish and shell- Anthropogenic nutrient stressors: Their interactions
fish catches, persisted for a number of years (Ruzecki et al. with hydrology
1976). More recently (fall 1999), Hurricanes Dennis, Floyd,
and Irene inundated coastal North Carolina with up to 1 m
Nitrogen (N) availability commonly controls phytoplank-
of rainfall, causing a 100–500-yr flood (depending on lo-
ton and higher plant biomass and primary production in es-
cation) in the watershed of Pamlico Sound (PS), the second
tuarineandcoastalwaters(RytherandDunstan1971;Nixon
largest estuary in the U.S.A. and a key nursery for the mid-
1992). Excessive N loading, much of it anthropogenic, is a
and southeast Atlantic fisheries (Fig. 1). Sediment- and nu-
chief causative agent of coastal eutrophication(Nixon1995;
trient-laden floodwaters displaced over 80% of the Sound’s
Paerl 1997). Symptoms include phytoplankton blooms,
volume,depressedsalinitybymorethan70%,andaccounted
whichmayaccumulateaspartiallyorungrazedorganicmat-
for half the annual nitrogen (N) load to this N-sensitivesys-
ter,providingthe‘‘fuel’’forlarge-scaleoxygenconsumption
tem (Paerl et al. 2001; Peierls et al. 2003) (Fig. 2). The
and depletion in bottom waters and sediments. This cascad-
biogeochemical and ecological effects of this sequence of
ing chain of events is particularly problematic in stratified
hurricanes included a threefold increase in algal biomass,
hypoxic (,4 mg L21 of O) bottom waters accompanied by bottomwaters,whereoxygenisnotreadilyreplenishedfrom
changes in nutrient cyclin2g, altered fish distributions, fish the atmosphere (Fig. 3). Under these conditions, persistent
catches, and increased fish disease (Paerl et al. 2001, Tester low oxygen or hypoxic conditions (,4 mg L21 of O) can
2
et al. 2003; Adams et al. 2003). Almost exactly 4 yr later alter nutrient (N, P, and trace metals) cycling (Paerl et al.
(September 2003), Hurricane Isabel made landfall in the 1998; Conley 1999; Rabalais and Turner2001)andpromote
same region (Fig. 1), crossed the PS as well as the Tide- fish disease and mortality (Diaz and Rosenberg 1995).
water–CB region to the north, breached North Carolina’s The composition, concentration, and delivery of nutrients
OuterBanks,andcausedlarge-scalehydrologic(stormsurg- depend on how the watershed has been modified by agri-
es, flooding) and nutrient perturbations (enhanced runoff, cultural,urban,andindustrialactivities(HopkinsonandVal-
washout of coastal developments, marinas, wetlands, and lino1995;Paerl1997).Futhermore,thetiming,location,and
farmland) throughout the mid-Atlantic region. intensity of storms and associated rainfall amounts also af-
Although large estuarine ecosystems exhibit a range of fect nutrient makeup and discharge to coastal waters (Paerl
biogeochemical and trophic responses to short- and longer- 1997). Freshwater discharge delivers nutrients to the coastal
term hydrologic changes, they are also affected by multiple zone and determines the hydrologic properties of the water
stressors, including nutrients and other pollutants, changes column, including vertical stratification, water residence
in light regime (turbidity), temperature, mixing, and circu-
time, salinity, turbidity, and clarity. All of these properties
lation, which are changing in time and space. Over time
interact to mediate productivity, nutrient cycling, dissolved
these stressors may alter the ecological characteristics of
oxygen dynamics, and habitat condition (Figs. 3, 4). Resi-
these large systems. For example, the delivery of anthro-
dence time plays a critical role in determining the availabil-
pogenic nutrients and other pollutants to coastal waters isin
ity and utilization of nutrients by phytoplankton. As dis-
ahighlydynamicstate,asdevelopmentandacceleratedload-
charge also controls transport of phytoplankton through
ing and management loading reductions are occurring con-
these systems, it interacts with nutrient supply to control
currently (Nixon 1995; Boesch et al. 2001; Elmgren and
growth, competition, and succession of the phytoplankton
Larsson 2001). This may result in long-term changes in es-
community. For example, high rates of freshwaterdischarge
tuarine and coastal biogeochemical processes and trophic
reduce both salinity and residence time. These conditions
function. Integrating climatic and anthropogenic perturba-
generally favor fast-growing phytoplankton such as chloro-
tionsisdifficultbutessentialtounderstandandmanagethese
phytes (green algae) and various flagellates, members of
ecosystems.
which have been shown to demonstrate optimal growth un-
Theeffectsofhumanandnaturalperturbationsarereadily
detectedatthemicrobialprimaryproducerlevel,specifically der reduced salinity conditions (Pinckney et al. 1999). In
phytoplankton, level where large amounts of ecosystem en- contrast, low discharge promotes longer water residence
ergy and nutrient flows are mediated. Phytoplankton gener- times, favoring slower-growing taxa, including dinoflagel-
ally have fast growth rates and exhibit a high degreeof sen- latesandcyanobacteria.Theinfluencesofhydrologicforcing
sitivity, and in certain cases specificity, to an array of in relation to nutrient inputs, seasonality, and climatic vari-
pollutants and environmental perturbations, making them ability were explored and compared for both of these large
useful indicators of ecological change. estuarine ecosystems.
450 Paerl et al.
Eutrophication of large estuaries 451
Hydrologic effects on estuarine phytoplankton (high flushing accompaniedbyelevatednutrientinputs).For
composition example,intheCB,diatomabundancewasincreasedduring
highflowyears,regardlessofseason,whencomparedtolow
The different effects of nutrient and hydrologic changes flow years (Fig. 6). We hypothesize that this selectiveeffect
on phytoplankton community composition were examined is due to the fast growth rates and enhanced nutrient uptake
and compared in the CB and the NRE/PS estuarine com- ratesexhibitedbythisgroupundertheseconditions(Malone
plexes (Fig. 1). Diagnostic chlorophyll and carotenoid pho- et al. 1988; Harding 1994; Pinckney et al. 1999). Contrary
topigments, measured by high-performance liquid chroma- to the CB, diatoms in the NRE were reduced in abundance
tography coupled to a photodiode array spectrophotometer, duringhighflowyears,regardlessofseason,whencompared
were used as indicators of major phytoplankton taxonomic to years characterized by reduced flow conditions. A possi-
groups (PTGs), including diatoms, dinoflagellates, chloro- ble explanation for this differential effect is that under high
phytes,cyanobacteria,andcryptomonads(Paerletal.2003). flowconditions,waterresidencetimeinvarioussegmentsof
These photopigments were measured from water samples the CB is substantially shorter (1–3 wk) than those in the
collected from the CB (1995–2000) and from the NRE/PS NRE/PS (2 wk to 2 months). Under the relatively shorter
(NRE: 1994–present; PS: 1999–present) as part of ongoing residence time conditions typifying the CB, diatoms will
monitoringstudies.Astatisticalprocedure,ChemTax(Mack- most effectively compete, as they tend to have faster dou-
ey et al. 1996), was applied to these photopigment concen- bling times (,0.7 d) than other PTGs (generally 1–2 d).
trations to partition total phytoplankton biomass, as chloro- Furthermore, this effect may be due to differences in the
phyll a (Chl a), into the PTGs and quantify the relative and native phytoplankton composition that characterize these
absolute contributions of each group. Diagnostic photopig- two systems. Diatoms are generally predominant in the CB,
ment markers included Chl b and lutein (chlorophytes), ze- whereas all five taxonomic groups are typically found in
axanthin, myxoxanthophyll and echinenone (cyanobacteria), similar proportions in the NRE (;20%).
fucoxanthin (diatoms), peridinin (dinoflagellates),andallox- Seasonality and interannual climatic differences (i.e., wet
anthin (cryptomonads). vs.dryyears)stronglyaffecthydrologicconditionsandPTG
Seasonalorhurricane-induced(orboth)variationsinriver composition in these estuarine systems. Dinoflagellates
discharge, and hence residence time, affect PTG dynamics showed greater competitive abilities in the nutrient-enriched
in the CB and NRE/PS systems as a function of the inter- springmonthsundermoderatetohighflowconditionsinthe
actions offreshwaterdischargerates,volumeandsizeofthe CB. This effect is most noticeable in spring in the CB as
receiving system (which affect water residence time), and opposed to summer and fall, when overall flows are lower
resident phytoplankton community composition and growth andresidencetimesappeartobelongenoughforotherPTGs
characteristics. Multi-annual monthly means of streamflow to increase in abundance. During summertime, flowhasless
rates for the largest tributaries of the CB (SusquehannaRiv- of an effect in promoting dominance of either diatoms or
er) and PS (Neuse River) are shown in Figure 5. Although dinoflagellatesineachsystem,whereasduringfall,highflow
SusquehannaRiverflowissignificantlygreaterinmagnitude promotes diatom dominance in CB but not in the NRE/PS.
than Neuse River flow, overall, the discharge patterns of In contrast, cyanobacteria tend to be most dominant in sum-
these two rivers were similar, reflecting regional seasonal mertime in both systems, when low flow generally predom-
wet and dry periods, including very wet spring years (e.g., inatesandthedistinctionbetweenlowandhighflowismin-
1994 and 1998) as well as severe summer droughts (e.g., imized (Fig. 6). With the exception of summertime
1997–1999, 2001–2002). However, large differences in dis- cyanobacteria populations, all PTGs in the NRE (and thus
charge patterns were also apparent, depending on the occur- total Chl a concentrations) decreased duringhighflowyears
rence of localized storm events. For example, during the when compared to low flow years. On a seasonal basis,
early fall of 1999, hurricanes Dennis and Floyd severely af- changes in flow regimes co-occur with changing irradiance
fected the Neuse River watershed while the Susquehanna and temperature patterns. In addition, zooplankton and ben-
watershed was relatively unaffected (Fig. 5). The potential thic invertebrate (shellfish) and herbivorous fish (e.g., men-
effects of high versus low flow seasons and years as wellas haden) grazing influence PTG abundance and dominance,
individual events (e.g., the hurricanes of 1999) are reflected thus creating a complex set of interactions with residence
in these freshwater flows entering the respective receiving time that control PTG community structure.
estuaries (Fig. 5). Susquehanna River flow (SRF) is an important driver of
The phytoplankton composition of the CB and the NRE/ seasonaltointerannualhydrologicvariabilityintheCB(Fig.
PS systems varied in response to nutrient enrichment, 5) (Malone et al. 1988; Fisher et al. 1988; Boynton and
droughts (reduced flushing combined with minimal nutrient Kemp2000).SummerphytoplanktondynamicsinCB,when
inputs), and elevated tropical storm and hurricane activity SRFwaslow,werecharacterizedbylowbiomass(compared
‹
Fig.1.Upperpanel:MapoftheU.S.mid-Atlanticcoastshowing(fromnorthtosouth)theSusquehannaRiver,theChesapeakeBay,the
AlbemarleSound,thePamlicoSound,andtheNeuseRiver.TheabbreviationsNJ,DC,DE,MD,VA,andNCrefertostatesormajorcities
bordering these systems. Lower panel: The tracks and intensities (Saffir–Simpson scale) of major hurricanes that have crossed this region
since1996. They include:HurricanesBerthaandFran(1996),Bonnie(1998),Dennis,Floyd,andIrene(1999),Isabel(2003),andCharley
and Alex (2004).
452 Paerl et al.
Fig. 2. (A–B) The Albemarle–Pamlico Sound Estuarine System and adjacent Atlantic Ocean coastal waters of eastern North Carolina,
U.S.A.,asobservedbyoceancolorsatelliteremotesensingsystemSeaWiFS(photoscourtesyofNASA).(A)LandfallofHurricaneFloyd
along the North Carolina coast, 16–17 September 1999. (B)Approximately 1 wk afterFloyd, 23September1999.Notethebrown-stained
floodwaters discharging into Pamlico Sound and overflowing into the coastal Atlantic Ocean. Some of the turbid, sediment-laden wateris
beingcarriedouttoseabytheGulfStream,whichpassescloselybytheNorthCarolinacoastline(fromsouthtonorth).(C)Totalnitrogen
(dissolved and particulate inorganic and organic) loading calculated from discharge and concentration at the entrance to the Neuse River
Estuary (StreetsFerry Bridge) during nonhurricaneand hurricane(1996, 1999)years.Typically,alargepercentageofannualNloadingin
nonhurricane (1984) years occurs during the rainy late-winter early-spring period from January through early-May. This is shown for a
Eutrophication of large estuaries 453
to spring) and a phytoplankton assemblage dominated by as we know it today, is in a more eutrophic statethanitwas
small and flagellated forms (Marshall and Lacouture 1986). only 50–100 years ago (c.f., Boesch et al. 2001). Eutrophi-
High SRF in the summer resulted in elevated total biomass cation of CB, as well as the NRE/PS, is strongly expressed
and an increased contribution of diatoms, which contrasted in several aspects of phytoplankton dynamics. Declining di-
with average low-flow summer conditions when diatoms atom diversity and a shift toward dominance by pelagic di-
were approximately equal in abundance with cyanobacteria, atoms preserved in CB sediment cores indicate accelerating
cryptophytes, and dinoflagellates (Fig. 6). In CB, as in the eutrophication since the time of European settlement (Coo-
NRE/PS, seasonal and interannual variability in flow pre- per and Brush 1991). Within the last 50 yr, an increasing
sents a dynamic backdrop against which indicators of an- trendinCBChlaconcentrationswasdetectedthatparalleled
thropogenic influence must be evaluated. increased Nloading thatoccurredoverthesametimeperiod
Hydrologic variability interacts with nutrient-enriched (Harding 1994; Harding and Perry 1997). These studies
conditions to control seasonal and interannual patterns and demonstrated that the effects of anthropogenic activities in
fates of phytoplankton communities. In the CB, strong sea- the CB watershed have affected the ecosystem in ways that
sonal signals in phytoplankton dynamics included accumu- are observable over long time scales (tens to hundreds of
lationsofhighbiomassduringthespringdiatombloom(Fig. years). They also emphasized the potential for natural vari-
6), much of which sediments out of the photic zone during ability to cause similar changes to the phytoplankton as
the spring–summer transition (Malone 1992; Malone et al. might be anticipated as a result of anthropogenic activities.
1996;Hardingetal.2002).Thisfluxofphytoplankton-based When large storms and hurricanes affect both the CB and
organic matter to the lower layer of the CB water column is the NRE/PS, PTG responses to these sudden and large hy-
retained by upper-Bay estuarine flow and is partially sepa- drologicperturbationscanbelargeanddistinctfromthelon-
rated from the upper water column by strong salinity strat- ger-term seasonal shifts in flow and residence time. The in-
ification (Kemp and Boynton 1984; Boicourt 1992). Micro- dividual and cumulative effects of these episodic events can
bial decomposition of this organic material supplies a be seen in both CB and the NRE/PS (Figs. 6 and 7). De-
significant fraction of nutrients that supports elevated sum- creases in the occurrence of winter–spring dinoflagellate
mer productivity and contributes to bottom-water anoxia, a blooms and increases in the abundance of chlorophytes co-
persistent feature during summer (Malone 1992). incided with an increased frequency and magnitude of trop-
Location is also an important determinant of PTG com- ical storms and hurricanes since 1996 in the NRE/PS (Fig.
positioninthesesystems;however,itappearstointeractless 7). Interestingly, this effect was not evident in CB, which
with flow and discharge than seasonality. For example, in was not nearly as severely affected by hurricanes in 1996
CB diatoms are dominant in high-flow years regardless of (Fran) or 1999 (Dennis and Floyd) (data not shown). The
location, whereas dinoflagellates increase in dominance at relatively slow growth rates of dinoflagellates may have led
both upper and mid, but not lower, estuarine locations (Ma- to their reduced abundance during the extremely high river
lone et al. 1988; Malone 1992). The lack of uniformity of discharge events that accompanied landfall of Fran and sub-
dinoflagellate dominance along the entire estuary during sequent hurricanes in the NRE/PS watershed. In addition,
low-flow years may be due to the increasing influence of changes in residence time may influence the composition
salinity under low-flow conditions (Malone et al. 1988). and interactions of PTGs and their grazers,includingmicro-
Downstream locations typically exhibit high salinity during and macrozooplankton, and larval and juvenile fish. Such
low-flow years, conditions that are unfavorable forcommon shifts will affect trophic transfer pathways and efficiencies,
bloom-forming estuarine dinoflagellate species. This effect with ramifications for food web structure, function, and car-
islessdampenedintheNRE,whichmaintainslowersalinity bon and nutrient (N, P, Si, minor nutrients)cycling(Kleppel
(,20) than CB even at its lower estuarine location, because and Burkart 1995; Turner et al. 1998, Conley 1999). Phy-
of the buffering effects of the downstream PS, which acts toplankton composition in the NRE/PS estuarine systemhas
as a large lagoonal, mesohaline reservoir. For the same rea- changedsince1995,largelyinresponsetopersistentperiods
sons, the lower NRE location is well-suited for supporting of flooding and elevated freshwaterdischargeresultingfrom
periodic dinoflagellate blooms when flow is enhanced (e.g., the 1996 and 1999 hurricanes. Such changes are not nearly
late-winter Heterocapsa triquetra blooms) (Fig. 7) (Paerl et as evident in CB.
al.1995),asthesystemtypicallymaintainsitslagoonal,non- Although these results indicate that physical–chemical
tidal nature, ensuring favorable (to dinoflagellates) oligo- to forcing features strongly influence estuarine phytoplankton
mesohaline conditions downstream. dynamicsmediatingeutrophication,canwemanipulatethese
Nutrient-enrichedconditionsinteractwithbothhydrologic features to control this process in these large ecosystems?
discharge and seasonality to determine PTG composition in The answers to this question depend on the scale of the
these systems. In CB, studies haveemphasized thattheBay, system and the availability of freshwater. Aside from aqua-
‹
relatively dry year (1994), as well as for more ‘‘normal’’ years(1996 and 1999). During droughtconditions(earlysummerof1999),very
littleNloadingtakesplace.AlsoshownareNinputsduetohurricanesthataffectedtheNeuseRiverEstuarywatershedwithheavyrainfall.
During 1996, Hurricanes Bertha (category 2) and Fran (strong category 3) affected the watershed, whereas in the summer–fall of 1999,
Dennis (category 2), Floyd (strong category 3), and Irene(category 2) sequentiallyaffectedthewatershedwithina6-wkperiod.Notethat
the amounts of N loading due to large hurricanes (Fran and Floyd) accounted for large percentagesof annual N loading to this estuary.
454 Paerl et al.
Fig.3.Upperpanel:MapoftheNeuseRiverEstuaryshowingthehistoricandcurrentlymonitored
water quality stations. Lower panel: Contour plots of water column salinity and dissolved oxygen
along a vertical axial transect of the eutrophic Neuse River Estuary, North Carolina, ranging from
the freshwater head of the estuary (left hand side) to themesohalineentranceintoPamlicoSound.
The contours were generated from profiles collected on 13 June 2000 as part of the Neuse River
Estuary Modeling and Monitoring Program,ModMon;theprofiledatapointsareindicatedbydots
on plot.
culture and retention ponds, small lagoons, and reservoirs, generallydifficulttocontrolwaterqualityandeutrophication
or a few large river systems where sufficient water is avail- solely by controlling freshwater discharge. In most coastal
able for hydrologic manipulations such as flushing of nutri- watersheds, flow controls are not likely to be technically or
ent-enriched impoundments, reservoirs, estuaries,coastalla- economically feasible and realistic because large-scale, un-
goons, and embayments (e.g., Mississippi River Delta), it is predictable weather events such as hurricanes and droughts
Eutrophication of large estuaries 455
Fig. 4. Conceptual diagram showing the various watershed and airshed anthropogenic nutrient
sources, their input to estuarine and coastal waters via freshwater discharge, the establishment of
hypoxia due to freshwater overlaying denser saltwater, and the stimulation of primary production
(eutrophication) and algal booms due to coastal nutrient enrichment. Note the linkage between
nutrient-enrichedprimaryproductionandhypoxiaasphytoplanktonsinkintostratifiedbottomwater.
Also, note the potential negative effects of hypoxia on bottom-dwelling finfish and shellfish and
submersed aquatic vegetation communities.
tend to dominate the hydrologic characteristics of thesesys- amongtheirconstituentmicrobialmembers,therebykeeping
tems. This leaves nutrient and sediment inputs as the chief nutrients from being ‘‘lost’’ by physical processes such as
controllable variables. diffusion, advection, sedimentation, or by biogeochemical
transformations like denitrification. We suggest that consor-
Biology and scaling: Their roles in estuarine and tial conservation of nutrients has evolved in response to the
coastal eutrophication pulsed, episodic manner by which nutrients are often deliv-
ered to estuarine and coastal water bodies after storms and
Primary productivity and growth responses to nutrient in- anthropogenic nutrient discharges. By using effective up-
puts and hydrologic modifications can be nonlinear, may in- take, retention (e.g., intracellular storage capabilities), and
volve long lag periods, and may be insensitive to short-term consortial exchange and recycling mechanisms, bloom spe-
(days–weeks)nutrientreductions.Inpart,thisisbecausepri- ciescanthriveandpersistonpulsedsourcesofnutrientsthat
mary producer communities are comprised of metabolically mayhaveenteredthesystemweeksormonthsbeforebloom
coupled assemblages of microorganisms and higher-ranked conditions prevailed. Conversely, reductions of nutrient in-
flora and fauna, whose interactions promote effective utili- putsduringoptimalgrowthandbloomperiodsmaynothave
zation,retention,andrecyclingofnutrientsandenergy.More the immediate desired effects, namely rapid reduction or
oftenthannot,phytoplanktonandbenthicmicroalgalspecies control of blooms and epiphytic microalgal growth. Over
exhibit maximum growth rates in the presence of bacterial, longer time scales (i.e., months, seasons, and years), persis-
protozoan,and othermicrobialconsorts(PaerlandPinckney tent nutrient reductions are likely to have beneficial effects.
1996). These mutually-beneficial, or consortial, associations Therefore, the ‘‘payoff’’ from nutrient reductions may not
appear to be the rule rather than the exception in nature be realized until subsequent years or even decades.
(PaerlandPinckney1996).Furthermore,theyplayvitalroles The homeostatic modulating roles that consortia play in
inthedevelopment,maintenance,andproliferationofplank- productivity and growth responses to variable nutrient en-
tonic algal blooms and benthic microbial ‘‘fouling’’ com- richment must be examined and evaluated in the context of
munities, common manifestations of coastal eutrophication interacting bottom-up physical–chemical drivers such ashy-
(Paerl 1988; La Pointe 1997; Paerl and Kuparinen 2002). drologic, nutrient, light, temperature regimes, and top-down
Once established, consortia retain and recycle nutrients controlsexertedbygrazingandpredation.Theseinteractions
456 Paerl et al.
Fig. 5. Monthly mean river flowrates ofthe majortributariesof
the Chesapeake Bay (Susquehanna River) and the Pamlico Sound
(Neuse River) during the 1994–2003 period. Values wereobtained
from gauging stations monitored by United States Geological Sur-
veyintheSusquehannaRiveratConowingo,Maryland(USGSSta-
tion No. 1578310) and in the Neuse River at Kinston, North Car-
olina (USGS Station No. 2089500).
influence nutrient availability, recycling, and stoichiometric
controls on production, biomass, species composition, and
resultant eutrophication dynamics on the larger ecosystem
scale.
Fig. 6. Seasonal means of phytoplankton community structure
andbiomassintheChesapeakeBay(upperpanel)andintheNeuse
P versus N controls of eutrophication in the River (lower panel) during low and high flow years. Seasons are
estuarine–coastal continuum defined by month (spring 5 March through May, summer 5 June
through August, fall 5 September through November, and winter
5 December through February). Low-flow years represent years
Because estuarine and coastal ecosystems are hydrologic
withmeanriverflowratesbelowthelong-termmean,whereashigh-
and biogeochemical continua of freshwater and marine en-
flow years represent years with mean river flow rates above the
vironments, they represent a unique and formidable chal-
long-term mean (1968–2000 forthe SusquehannaRiverand1931–
lengetoformulatingnutrientreductionsaimedatcontrolling 2002 for the Neuse River). Years defined as low-flow years in the
eutrophication. At upstream freshwater locations, P is often Chesapeake Bay were 1995, 1997, 1999, and 2000, whereas 1996
the growth-limiting macronutrient (Boynton et al. 1982; and 1998 were designated as high-flow years. In the Neuse River,
Larsson et al. 1985). At the freshwater–saltwater transition low-flowyearsincluded1994,1997,2000,2001,and2002,whereas
zone,PandNmaybothbecolimiting,whilethedownstream high-flow years were defined as 1995, 1996, 1998, and 1999. Bio-
mesohaline, polyhaline zones are usually N limited (Fisher mass (Chl a) is represented by the total length of each bar. Bar
portionsrepresentthefractionofChlaattributabletodifferenttax-
et al. 1988; Rudek et al. 1991; Elmgren and Larsson 2001).
onomic groups as determined from photopigment and ChemTax
ThespatialandtemporaloverlapsofNandPlimitationvary
analyses. ChesapeakeBay was not sampled during winter months.
widely, and are closely controlled by hydrology, morphom-
etry, geography, and climate.
Numerousstudieshaveexaminednutrientlimitationalong nutrientlimitationshiftsfromPinthefreshwatertoNinthe
the freshwater–marine continuum (c.f., Ryther and Dunstan more saline downstream waters. Starting in the 1960s and
1971;Pennocketal.1994;D’Eliaetal.1986).Thesestudies continuing throughout the 1980s, the identification of P as
haveprovidedtheimpetusandquantitativebasisfornutrient the limiting nutrient in freshwater ushered in a period of
input reduction strategies. There are numerous freshwater aggressivePreductionsinthewatershedsoflakesandrivers
‘‘successstories’’inwhichappropriateinputreductionshave draining into estuaries (Likens 1972, Schindler 1978; Vol-
helped to alleviate the most visible and problematic symp- lenweider1982).ThistrendwashighlightedbyaP-detergent
toms of eutrophication, including phytoplankton blooms, ban and advances in wastewater P treatment in the 1970s–
hypoxia, and losses of fisheriesand recreationalhabitat(Ed- 1990sinNorthAmerica,Europe,andpartsofAsia,thecom-
mondson 1970; Schindler 1978; Paerl 1988). However, the bined effect of which was marked reductions of P loading
estuarine–coastal continuum encompasses a broad range of in the headwaters of many estuaries. Parallel N reductions
varying hydrology, salinity, and nutrients, such that reduc- were generally not undertaken at that time, largely because
tion of one nutrient may not necessarily lessen eutrophica- eutrophication problems in downstream N-limited waters
tion along the entire length of an estuary. were either not recognized, detected, or ignored, i.e., the
One example is the effect of P control in estuaries where solution to pollution was dilution. Not until the mid-1980s
Eutrophication of large estuaries 457
Fig. 7. Surface concentrations of chlorophyll a (mg L21)contributedbychlorophytes,cyanobac-
teria, and dinoflagellates along the upper and middle regions of the Neuse River Estuary (1994–
2002). Values were determined from ChemTax analyses of high-performance liquid chromatogra-
phy-derived diagnostic photopigment concentrations. Data were collected biweekly and were
temporally and spatially extrapolated. Whiteareasindicateinstanceswhendatawerenotcollected.
ChemTax data were plotted along with freshwater discharge at the head of the estuary. The dates
of landfall of the four major hurricanes that have significantly affected flow since mid-1996 are
shown.
were the broader ramifications of N-driven eutrophication ban (early 1988) and improved P removal by wastewater
and benefits of N reductions widely realized (cf. Boyntonet treatment plants (mid to late 1980s). No parallel N reduc-
al. 1982; Fisher et al. 1988). Even then, there was consid- tions were pursued, even though early studies showed N to
erable reluctance to tackle the ‘‘nitrogen problem’’ as ag- becolimitingupstreamandexclusivelylimitingdownstream
gressively as that of P overenrichment in upstream fresh- of these algal blooms (Hobbie and Smith 1975;Paerl1983).
water segments. This was because, unlike P inputs, which The success of P reductions is illustrated in Fig. 8, which
are often dominated by point sources, N inputs are domi- shows significant decreases of mean annual dissolved inor-
nated by diffuse nonpoint sources that include agricultural ganic and total P concentrations during the mid- to late
runoff, groundwater, and atmospheric deposition (Paerl
1980s in both upstream and downstream portions of the
1997) that are difficult to identify, generally regional in
NRE. Effects of this selective nutrient reduction on nutrient
scale, and challenging to reduce.
stoichiometry and phytoplankton productionincludedasud-
Although P reductions have reduced eutrophication of
den increase in molar N:P ratios for total (T) and dissolved
freshwater parts of estuaries, they have had no such effect
inorganic (DI) forms (from approximately 12 DIN:DIP in
downstream.Tothecontrary,Preductionsmayhaveactually
1986 to .60 DIN:DIP by 1992 and from 15 TN:TP to 25
exacerbated eutrophication in more seaward regions of es-
TN:TP during the same time frame), and a protracted de-
tuaries (cf. Elmgren and Larsson 2001). This is because re-
crease in mean annual phytoplankton biomass (as Chl a) in
ductionsinupstreameutrophicationpotentialswillalsolow-
theupstreamfreshwatersection(Fig.8).Furtherdownstream
er the amount of biomass that assimilates, retains, and in
effect ‘‘filters’’ nutrients other than P (e.g., N, Fe, Si) that in the NRE, an increase in mean annual total and dissolved
may limit downstream productivity. inorganic N:P was also observed, although not as large as
This shift appears to have taken place in the NRE, where in the upstream region (Fig. 8).
severeeutrophicationproblemsdatingbacktothe1970sand During the decade following the large decreaseofPload-
1980s,includingfreshwatercyanobacterialblooms,hypoxia, ing (late1980s throughlate1990s),Chlareachedhighcon-
and fish kills (Paerl 1983; Christian et al. 1986), were ad- centrations in more downstream areas of the estuary (Fig.
dressed through P reductions in the form of a P-detergent 8). It appears that the Chl a maxima migrated from the P-