Nitrogen and Phosphorus Budgets of the North Atlantic Ocean and Its Watershed
Author(s): J. N. Galloway, R. W. Howarth, A. F. Michaels, S. W. Nixon, J. M. Prospero, F. J.
Dentener
Source: Biogeochemistry, Vol. 35, No. 1, Nitrogen Cycling in the North Atlantic Ocean and Its
Watersheds (Oct., 1996), pp. 3-25
Published by: Springer
Stable URL: http://www.jstor.org/stable/1469224
Accessed: 16/05/2010 01:41
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://dv1litvip.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=springer.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
Springer is collaborating with JSTOR to digitize, preserve and extend access to Biogeochemistry.
http://dv1litvip.jstor.org
Biogeochemistry35: 3-25, 1996.
@ 1996 KluwerAcademicPublishers. Printed in theNetherlands.
Nitrogen and phosphorus budgets of the North Atlantic
Oceanand its watershed
J. N. GALLOWAY',R. W. HOWARTH2,A. F. MICHAELS3,
S. W. NIXON4,J. M. PROSPERO5& F. J. DENTENER6
'EnvironmentalSciences, Universityof Virginia,Charlottesville,VA22903 USA;2Ecology
and Systematics,Cornell University,Ithaca, NY 14853 USA;3BermudaBiological Stationfor
Research,St. Georges GE-01 Bermuda;4GraduateSchool of Oceanography,Universityof
RhodeIsland, NaragansettRI02882 USA;5RSMAS,UniversityofMiami, Miami,FL 33149
USA;6DepartmentofAirquality,Wageningen,NL 6700 EV Wageningen,TheNetherlands
Received 22 March 1996; accepted22 March 1996
Abstract. Anthropogenicfood and energy productionextensively mobilize reactivenitrogen
(N) in the watershedof the NorthAtlanticOcean (NAO). There is wide spreadN distribution
by both hydrologic and atmosphericprocesses within the watershedof the NAO, resulting
in reactive N accumulationin terrestrialsystems. Net denitrificationin most estuaries and
continentalshelves exceeds the amountof N supplied to the shelves by rivers and requires
a supply of nitrate from the open ocean. Thus riverine N is only transportedto the open
ocean in a few areas with the flow from a few major rivers (e.g., Amazon). AtmosphericN
deposition to the open ocean has increasedand may increase the productivityof the surface
ocean. In addition,as a consequenceof increasedFe depositionto the open ocean (due in part
to anthropogenicprocesses), the rateof biological N-fixationmay have increasedresultingin
N accumulationin the ocean. Phosphorus(P) is also mobilized by anthropogenicprocesses
(primarilyfood production).Relative to N, more of the P is transportedacross the shelf to
the open ocean from both estuariesand majorrivers. There are several consequences of the
increased availabilityof N and P that are unique to each element. However, the control on
primaryproductivityin both coastal and open ocean ecosystems is dependenton a complex
and poorly understoodinteractionbetween N and P mobilizationand availability.
Introduction
Nitrogen(N) is a key element of many biogeochemicalprocesses and can be
a limitingelementof aquaticandterrestrialecosystem processes(Schlesinger
1991; Vitousek & Howarth 1991). However, about 99% of global N exists
as stable atmosphericN2 (Mackenzie et al. 1993) and thus is unavailable
to ecosystems unless it is converted into a reactive N species (reactive N
= NH3, NH4+, organic N, NO, NO2, HNO3, NO3, N205, HNO4, HNO2,
NO2, and NO3). Once created,one species of reactive N can be converted
into other species of reactive N by a variety of chemical and microbial
processes.In addition,reactiveN species arevery mobilevia atmosphericand
hydrologicpathways.Afterformation,reactiveN can only be convertedback
to unreactiveN2 by denitrification,an anaerobicprocess which only occurs
4
at significantrates in specific types of ecosystems. Reactive N accumulates
in the environmentif the denitrificationrateis less thanthe rateof N-fixation.
In the absenceof humans,naturalprocessescreatereactiveN by biological
N-fixation and lightning. Biological N-fixation occurs in specific microbes
when atmosphericN2 is convertedto NH3 by the enzyme nitrogenase.Lightning producesNO by the reaction of N2 and 02 at high temperatures.The
formerprocess is about two ordersof magnitudegreaterthanthe latteron a
global basis (Galloway et al. 1995). Humaninterventionin the N cycle has
increasedthe formationrateof reactiveN by fertilizerproduction,legume and
rice cultivation,and combustionof fossil fuels. Results from several recent
analyses of the global N cycle (Mackenzie et al. 1993; Ayres et al. 1994;
Galloway et al. 1995) generallyagree that anthropogenicactivities mobilize
about10 Tmol N yr-1 (Tmol-= 1012moles) andthathumanactivitiesmobilize
N at ratesequal to naturalterrestrialprocesses.
There is significant distributionof anthropogenicN by hydrologic and
atmospherictransport.Combustionof fossil fuels injects reactive N (NO)
directlyinto the atmosphere.FertilizerandcultivationN increasethe productivity of agriculturallandscapes.However, on average, no more than 50%
of the appliedN fertilizeris removed by crop harvest;the remainderis lost
to the atmosphereor hydrosphere,or stored in the soil (Howarthet al. this
volume). In addition,N fixed in cropshas a shortresidencetime; it is quickly
transformedinto human or animal waste, which also results in significant
inputs to the atmosphereand hydrosphere(Howarthet al. this volume). If
the N introducedinto the environmentby fossil fuel combustion,fertilizer
productionand cultivationis not denitrified,then reactive N accumulatesin
downwindor downstreamecosystems.
The mobilization, distribution and accumulation of anthropogenic N
impacts a number of physical and ecosystem processes. As discussed in
Howarthet al. (this volume), Nixon et al. (this volume), and Michaels et al.
(this volume), increased availabilityof reactive N increases forest productivity, and as a consequence stores atmosphericCO2, contributesto forest
decline (if soil is N-saturated)and climate change, and results in shifts
in community structureand ecosystem function. It also increases coastal
eutrophicationin estuaries(includingtheirwetlands),andresultsin increased
N supply to oligotrophicmid-ocean gyres with concomitantaffects on the
ecology of the upper ocean. In addition to being importantto ecosystems,
reactive N also affects atmosphericchemistry(Prosperoet al. this volume).
High levels of NOx (NO + NO2) play an importantrole in the photochemical
productionof 03 (Moxim et al. 1994). NH3 is a majorsourceof alkalinityin
the atmosphereand a sourceof acidity in soils (Schlesinger 1991). Although
N20 is not viewed as a reactiveform of N in the troposphere,it adsorbsIR
5
4~
4
Figure 1. The North Atlantic Ocean and its watershed.The watershedis subdividedinto its
drainagebasins. The boundarybetween the shelf and the open ocean is 200m depth.
radiationand acts as a greenhousegas. In the stratosphere,N20 impacts03
concentrations(Warneck1988). Thus, any change in the rate of formationof
reactiveN (or N20), its global distribution,or its accumulationratecan have
a fundamentalimpacton many environmentalprocesses.
Our understandingof the global distribution,fate and impactsof anthropogenic N is insufficient because of a lack of data. To partially alleviate
this problem, it is necessary to examine the N cycle on a smaller scale
where data are more available,while still keeping the salient featuresof the
N cycle - naturaland anthropogenicsources, atmosphericand hydrologic
transportprocesses, and agricultural,forested,freshwater,coastaland marine
ecosystems. Hence we focus on a watershed analysis in general, and the
North Atlantic Ocean and its watershed(NAO&W) specifically (Figure 1).
We selected the watershedscale because its physiographicand hydrological
featuresshape a biogeochemical system; these same featuresshape systems
of humaninteraction.We selected the NAO&Wbecause, comparedto other
oceanbasins, it containsa richdatarecordof the atmosphere:watershed:ocean
units on N cycling andbecauseof the high degreeof N disturbanceby human
activities.
Our analysis of the NAO&W N budget was centeredabout a workshop,
held in May 1994 at Block Island, RI, USA. The centralscientific question
of the workshopwas:
6
What are the currentsources and sinks of nitrogenin the North Atlantic
Ocean and its watershed? How might these have been changed over
naturalbackgroundlevels as a resultof humanactivity?
Since the impactof anthropogenicN on ecosystems is determinedin part
by the relative availabilityof N and P, our investigationof the N budget of
the NAO&Wincludes an analysis of the magnitudeand fate of P transferred
from continentsto the ocean.
To addressthe questionof N cycling on the scale of an entireocean basin
and its associatedwatershed,a unique groupof individualsassembled,each
with expertisein physical,chemical andbiological aspectsof marine,coastal,
freshwaterand terrestrialecosystems, the atmosphere,and the connections
among the components.Workinggroups were formed on the basis of reservoirs:atmosphere,watershed,coastal andshelf region,and open ocean. Each
working group developed an annual N cycle for their respective reservoir.
The watershedof the NAO and associated coastal segments were further
divided(Figure1) to investigatespatialvariabilityin N andP fluxes. The task
of each group was to characterizethe internalN fluxes and the hydrologic/
atmosphericexport losses for their reservoir;input fluxes to the reservoir
were the domainof the upstreamor upwind group. To enhance integration,
each groupwas allowed 'to borrow' membersfrom other groupsto address
specific issues of N exchange.
The primaryproducts of the workshop are the working group papers
on the N and P fluxes of each sub-reservoir:atmosphere- deposition to
watershedand oceans (Prosperoet al. this volume); watershed- N inputs,
fates, and riverinelosses to the coastal margins,and riverineP fluxes to the
coastal ocean (Howarthet al. this volume); coastal and shelf region-fate of
riverineandatmosphericinputs(Nixon et al. this volume);andthe open ocean
(Michaelset al. this volume). In addition,in the process of developing these
papers, it became apparentthat several topics warrantedspecial attention.
Thus accompanyingthese papersare four notes: NH3 exchangebetween the
atmosphereand the NAO (Quinn et al. this volume), N-fixationin the NAO
(LipschultzandOwens, this volume); denitrificationin the ocean boundaries
(Seitzinger and Giblin, this volume); and shelf areas of the North Atlantic
Ocean (Pilson and Seitzinger,this volume).
This paperis a synthesisof the workshopfindings.It presentsthe N andP
budgetsfor the entireNAO&W system by addressingthe following specific
topics:
* Mobilizationof reactiveN in the NAO&W
* Re-distributionof reactive N among the subsystemsof the NAO&W
* Removal of reactiveN from the NAO&Wrelativeto sources
* P distributionpatternsin the NAO&W
7
* AnthropogenicN and P impactson the NAO&W
An advantageof examiningthe NAO&Was one system is thatit provides
an overview; a disadvantageis that it masks the extensive spatialvariability
in the N and P budgets of the portions of four continents, and associated
coastal margins,that constitutethe watershedof the NAO. This variability
is extensive. Europeand NorthAmericahave significantlygreaterincreases
in the mobilization of reactive N than do Africa and South America. We
highlight some of the detailed regional findingsfrom the 14 sub-watershed
portionsof the watershedof the NAO and adjacentcoastal regions and refer
the readersto the appropriatepapersfor the rest of the detail.
The results
We illustratethe overall behaviorof N and P in the NAO&W by dividing
the NAO&W into five systems: atmosphere,watershed(26.9 x 106 km2),
estuaries, shelf (to 200m depth, 5.7 x 106 km2), and open ocean (100 x
106 km2) (e.g., Figure 2). Shaded arrows in Figure 2 representreactive N
creation;open arrowsrepresentreactive N transfersbetween the NAO&W
sub-systems. The arrow width is sized to the median flux magnitude.The
rangesaboutthe medianfluxes arefoundin the text. Units are 109moles yr-1
(Gmol yr-'). For consistency,the fluxes in this paperhave been roundedto
two significantfiguresfrom the values reportedin the accompanyingpapers.
Mobilizationof reactiveN in the NAO&W
There are two sources of reactive N - creation within, and transportto,
the NAO&W.Reactive N is createdand mobilized by biological N-fixation,
lightning, commercial N fertilizer use, fossil fuel combustion and legume
cultivation. The watershedand open ocean are the most importantsource
regionsfor reactiveN in the NAO&W;per area,the land masses of the North
Atlantic are a far greatersource of reactive N, reflectingthe greaterhuman
disturbanceon land (Figure2).
Watershed.Fertilizeruse introduces 1,600 Gmol yr-' to the watershedof
the NAO and is the most importantanthropogenicN source, followed by the
deposition of NOy (NOy = NOx + HNO3 + NO3 aerosol + other oxygencontainingN species except N20) from fossil fuel combustion (590 Gmol
yr-'), and legume cultivation(330 Gmol yr-'). The total anthropogenicN
addition to the NAO watershed is 2,500 Gmol yr-' (Howarth et al. this
volume). Figure 2 shows a watershedinput of anthropogenicNHx (NH3 +
NH+) from atmosphericdeposition(290 Gmol yr-1). We view this inputas
8
NOy dep.
anth. nat
590 76
N dep.
130
mix
denitrification 79 Ny denitrification
NHx dep.
anth. nat.
290 97
V
N dep.
620
2K NHx
36
NOy
"V
_
120
Mediteranean
major rivers 350
burial
270
330
1600
legumes
food & feilizer
fedscatch
N-fixation
5,000
South Atlantic
to
-1,400
4,600
32
fish
120
shelf & slope
sediment
46
open
ocean
sediment
Figure 2. The N budget for the North Atlantic Ocean and its watershed(Gmol yr-'). The
stippledarrowsrepresentthe introductionof new sourcesof reactiveN into the NAO&W.The
open arrowsrepresentexchange of reactiveN between the subsystemsof the NAO&W.
an internalrecycling of NHx following its introductionby fertilizeruse and
subsequentNH3 volatilizationfrom soils and animal waste (Howarthet al.
this volume).
There are two naturalsources of reactive N: lightningand terrestrialbiological N-fixation.Lightningsources are small relativeto fertilizersourcesthe global total is only about 210 Gmol yr-' (Levy et al. 1996). Naturalterrestrialbiological N-fixationfor the watershedof the NAO may be important
relative to anthropogenicN sources. Globally, the range of fifteen estimates
is 5,700 to 13,600 Gmol yr-' (as reviewed in Jaffe 1992; Mackenzieet al.
1993). Thus, naturalbiological N-fixationin the watershedof the NAO, with
18%of the earth'sland area,could be significant.However,given the probable stronglatitudinaldependenceon the N-fixation rate, the dearthof data
for many tropicaland temperateecosystems, and our focus at examiningthe
fate of anthropogenicN mobilized in the watershedof the NAO, we choose
not to estimate naturalterrestrialbiological N-fixationrate.We do point out
however, that studies more specifically focused on developing an integrated
N cycle for watershedunits need to include naturalbiological N-fixation.
9
The relativeimportanceof anthropogenicN sources varies by sub-watershed unit (as definedin Figure 1). AgriculturalN sources dominateinputsin
many regions,particularlythe Mississippiand the NorthSea sub-watersheds.
In the NE United States, NOy deposition as a consequence of fossil fuel
combustionis the majorcontrolover river N. In the temperateregion, there
is a strongcorrelationof total N flux with populationdensity.This featureis
manifestedin a strong linear correlationbetween river fluxes of total N and
the sum of anthropogenicN inputs (Howarthet al. this volume). Although
on the average, N in food and feed is exported from the watershedof the
NAO (Figure2), for some sub-watershedsit can be a significantsource.For
example, in the NE coast region, about 1/3 of the N sources are from food
and feed import(Howarthet al. this volume).
Estuaryand shelf Biological N-fixationcertainlyoccurs in estuarineecosystems but its magnitude is small relative to riverine input of reactive N
previously created in the NAO watershed. In the shelf region biological
N-fixationis estimatedto also be small, on the orderof 20 Gmol yr-' (Nixon
et al. this volume).
Ocean. Biological N-fixation in the open ocean is the most importantN
source for the NAO&W.Michaels et al. (this volume) identify a source of
nitrate of 3,700 to 6,400 Gmol yr-1 (median, 5000 Gmol yr-', Figure 2)
within the main thermoclineof the Sargasso Sea that they infer is caused
by N-fixation.This N-fixationmay be regulatedby the supply of Fe in dust
and thus may have occurred at a lower rate before the expansion of the
Saharain the early 1970s. The most conspicuous N-fixing organismsin the
North Atlantic Ocean are the filamentouscyanobacteriaTrichodesmiumsp.
Lipschultzand Owens (this volume) conclude that the measurementsof Nfixation by Trichodesmiumsp. imply a rate of only 1,100 Gmol N yr-1 for
this taxa alone. The differencebetween Lipschultzand Owens and Michaels
et al. may come from the scarcityof the data, questions about the accuracy
of the methodsand from the potentialcontributionsof other organismsthan
Trichodesmiumsp.
The second generalsourceof N to the NAO is the transportinto the region
of reactive N createdelsewhere. The large scale currentsin the NAO result
in significanthorizontaladvectionof nutrients(Michaelset al. this volume).
There is strongnortherlysurfacetransportin the Gulf Stream.There is a net
southerlyflow in the NorthAtlanticDeep Water,and net northerlytransport
in the AntarcticBottomandIntermediateWaters.Michaelset al. summarized
a numberof techniquesto estimate meridionaltransportin the NAO. From
the north they estimate that 300 to 2,800 Gmol yr-' (median: 1,500 Gmol
10
yr-1, Figure2) areadvectedinto the NAO fromthe ArcticOcean.In addition,
Michaelset al. (this volume) estimatethat50 to 200 Gmol yr-1 (median:120
Gmol yr-1, Figure2) aretransportedfrom the Mediterraneanto the NAO. To
the south, the uncertaintyof the calculationis such that both the magnitude
and directionof transportare unknown.The rangeis - 1,400 Gmol yr-1 out
of the NAO to the SouthAtlanticOcean (SAO), to 4,600 Gmol yr-1 fromthe
SAO to the NAO. The implicationsfor the overall balance of N in the NAO
are discussed in a following section.
In summary,the watershedof the NAO receives its anthropogenicreactive
N primarilyfromfertilizeruse, with smallercontributionsfromcombustionof
fossil fuels andlegume cultivation.Some regionswithin the watershedof the
NAO, especially withinthe temperatezone, are heavily impactedby not only
fossil fuel combustionand fertilizer application,but also N importationin
foods and feeds. The NAO receives its reactiveN primarilyfrom N-fixation,
although transportfrom the SAO may also be important.As discussed in
the next two sections, additionalN sources are atmosphericand hydrologic
transportof reactiveN createdin the watershed.
Re-distributionof reactiveN among the subsystemsof the NAO&W
Via atmospherictransport.NOx and NHx are emitted into the atmosphere
by naturaland anthropogenicprocesses. The most importantanthropogenic
processesareenergyproduction(NOxemissionsfromfossil fuel combustion)
and food production(NH3 emissions from fertilizeruse and animal waste)
(Prosperoet al. this volume). Once emitted, NOx and NH3 are transformed,
transporteddownwind and depositedas NOy and NHx, respectively.Unlike
ecosystems, there are no importantatmosphericreactionsthat convert NHx
to NOy and visa-versa.Thus depositionof anthropogenicNOy and NHx can
be generallylinked to energy and food production,respectively.
Total(present-daynaturalandanthropogenic)NOyatmosphericdeposition
to the watershed,shelf and open ocean regions of the NAO&W is 670 Gmol
yr-', 79 Gmol yr-' and 360 Gmol yr-', respectively(Figure2) (Prosperoet
al. this volume;Howarthet al. this volume;Nixon et al. this volume;Michaels
et al. this volume). AnthropogenicNOy deposition (590 Gmol yr-') to the
entire watershed of the NAO is about seven-fold greater than deposition
of naturallyproducedNOy (76 Gmol yr-1) (Figure 2). In regions heavily
impactedby fossil fuel combustion(i.e., NE coast of US), NOy deposition
is estimated to have increased by a factor of 30 relative to naturalrates.
For more remote portions of the watershedof the NAO (i.e., Amazon and
Tocantins basin) NOy deposition has increased about 2-3 times over natural
values (Howarth et al. this volume).
11
A similar situationexists for NHx deposition. The watershed,shelf and
open ocean regions receive total depositionof 390 Gmol yr-1, 55 Gmol yr-1
and 260 Gmol yr-1 respectively (Figure 2). For the watershedof the NAO,
currentNHx deposition is about 3 times greaterthan naturalrates for the
entirewatershedandup to 12 times greaterfor regionsof intenseagricultural
activity (i.e., NW Europeancoast) and about two times for more remote
regions (i.e., Amazon and Tocantinsbasin) (Howarth et al. this volume).
Relative to riverinesources, direct N atmosphericdeposition to estuariesis
small (Nixon et al. this volume).
While we have not estimatedthe naturaland anthropogeniccomponents
of N deposition to the shelf and open ocean portionsof the NAO, there is
likely to be a stronghumaninfluence,especially for NOy, given the disparity
between naturalNO atmosphericsources and anthropogenic.As discussed
in Prosperoet al. (this volume) and Galloway et al. (1995), the present-day
global deposition rate of NOy is about five times greaterthan pre-industrial
times largely due to emissions from energyproductionand biomassburning.
The relative increasewill be even greaterfor the NAO because it is adjacent
to region of intense fossil fuel combustion. While there is no doubt that
NHx deposition to the NAO has increased(Prosperoet al. this volume), the
magnitudeof the increaseis moredifficultto estimatebecause of a potentially
largenaturalemission of NH3fromthe ocean (Quinnet al. 1988).At this stage,
the data on NH3 emissions for the NAO are too sparse to make an estimate
of an ocean wide flux (Quinnet al. this volume). It should be noted that our
estimatesof total N depositionto the NAO&W may be underestimatesfor a
growingliteraturesuggests thatatmosphericdepositionof organicN species
may be equivalentto depositionof NOy andNHx, especially in marineregions
(Prosperoet al. this volume).
There is significant spatial variability in the NOy and NHx deposition
rates. By way of illustration,we examine NOy (Dentener& Crutzen1993)
andNHx(Dentener& Crutzen1994) depositionacrosstransectsof 450 N and
50 N at ten degree blocks of latitudeand longitude (Figure3). To show how
N depositionto the NAO comparesto the NorthPacific Ocean, we begin the
transectsat 175' W. To show how Europeanand African emissions impact
depositionto the east of the study area,we extend the transectto 350 E.
At 450 N, N deposition in NorthAmericaand Europeis about20 and 30
times, respectively,greaterthanthe Pacific Ocean (Figure 3). Downwind of
NorthAmerica there is a rapidreductionin NOy and NHx depositionto the
NAO. However, even in the middle of the NAO region, N depositionis still
about double that of the North Pacific Ocean. These model results support
previous statementsthat N deposition to the entire NAO has increasedas a
consequence of human activity (Whelpdale& Galloway 1994; Prosperoet
12
60
E
a4
3
40
30
.
i
S10
0A"
1
W1750 155( 1354 1150 950 75" 554 35" 150
North Pacific
North America
North
Ocean
Atlantic
Ocean
60
E50 250
Europe
4
b
650
3
40
Sj30
..2
20
1
--_
W175"155" 135" 115" 95"
North Pacific Ocean
150
55"
75"
C/S North35"
Atlantic
Amer
Ocean
E50 250
Africa
Longitude
from 1750 W
Figure3. NOy(openbars)and NHx (solid bars)deposition(mmol/m2/yr)
eastward to
350
E at (a) 450 N latitude and (b)
shown by dottedline.
50
N latitude. Ratio of NOy/NHx deposition
al. this volume). At the westernboundaryof Europe,N depositionincreases
significantlyandremainshigh well outside of our studyregion due to atmospherictransportof EuropeanN emissions. Withinthe NAO&Wstudyregion,
the relativeimportanceof NOy to NHx depositionvaries by abouta factorof
five. At the westernboundary,NOy/NHx is about 0.7 reflectingthe predominance of agriculturalsources of NH3. To the east, the ratio increases to
about2 due to the relativeincreasein energy over food production.The ratio
graduallydecreases over the NAO and is about 1 at the western boundary
of Europe. It then decreases to about 0.5, again due to the strengthof the
agriculturesector relativeto the energy sector.
For the 50 N transect,total deposition is significantlyless than at 450 N
(Figure3) due to the reducedimpactof NOx and NH3 sources. Nonetheless,
N depositionto the NAO at 50 N is still about2-3 times greaterthanto the
NorthPacific Ocean.In addition,because of the predominanteasterlywinds,
atmospherictransportof N is from east to west, accountingfor the gradual
reductionin depositionto the west of Africa and CentralandSouthAmerica.
With the exception of the easternportionof the Centraland South America,
13
the NOy/NHx ratio varies between about 1 and 2, reflectingthe importance
of combustionover agricultureas N sources to the atmosphere.
In summary,there is a net transferof NOy and NHx from the continental
to the marine atmosphere.A substantialfractionof the ~2,500 Gmol yr-'
of reactiveN mobilized in the watershedof the NAO by humanactivities is
deposited into the marineenvironmentfollowing atmosphericemission and
transport.
Via hydrologic transport.N is hydrologicallyexchanged between the subsystems of the NAO&W by riverine, estuarine and cross shelf processes.
2,500 Gmol yr-' of anthropogenicN are introducedinto the watershedof
the NAO. 950 Gmol yr-' of naturaland anthropogenicN are lost via rivers
to the marineenvironment;530 Gmol yr-' from the temperateportionof the
watershedand 420 Gmol yr-' from the tropical(Howarthet al. this volume).
Since we do not have estimatesof the naturalN-fixationratein the watershed,
or precise estimates of the anthropogenicN in rivers, we can only say that
the maximumloss of anthropogenicN via rivers in the temperateportionis
35%.Any amountof naturalN-fixationwithin the watershedwill reduce this
percentage.The data are too sparse to make an equivalentestimate for the
tropicalportion.
Of the 950 Gmol yr-' lost via riversfromthe entirewatershedof the NAO,
about410 Gmol yr-' arefromlargeriversthatdebouchdirectlyinto the NAO
shelf region;the remainder(540 Gmol yr-1) flow into estuaries.Ultimately,
largeriversandestuariesdeliverabout350 Gmol yr-1 (290 frommajorrivers
to shelf waters and ~61 Gmol yr-' to slope sediments)and 170-340 Gmol
yr-1 to the NAO shelf region, respectively(250 Gmol yr-1, Figure2) (Nixon
et al. this volume). From the ocean side of the shelf region, about 820 Gmol
yr-I are transportedto the shelf from the open ocean, primarilyas nitrate
(Nixon et al. this volume; Michaelset al. this volume). Thus the shelf region,
via hydrologictransport,receives about 1,400 Gmol N yr-' of N from major
rivers, estuaries and the open ocean (Figure 2). It also receives d20 Gmol
yr-1 by N-fixation(not shown on Figure2). This amount,combinedwith the
atmosphericflux of 130 Gmol yr-', resultsin an inputof ~ 1,600 Gmol yr-1
to the shelf. As discussed in the following section, most of the N deliveredto
the shelf region is denitrified.
Thereis significantspatialvariabilityin the degreeof hydrologicN transport. About two-thirdsof the riverineN is injectedfrom the westernportion
of the watershedof the NAO, aboutequallydivided between NorthAmerica
and Centraland South America. 25% of the N comes from one basin - the
Amazon(Howarthet al. this volume). Similarly,all the majorriversthatinject
N directlyto the shelf region arein the westernportionof the watershedof the
14
NAO and again, the Amazon exerts a large influence.It accountsfor about
half of the 600 Gmol yr-' of riverineN thatare injectedon to the NAO shelf
(Nixon et al. this volume).
There is also spatial variabilityin the impact that human activities have
had on hydrologictransferof N between the NAO and its watershed.Using
datafrom relativelypristineareasas an index of change, Howarthet al. (this
volume) estimate that riverine N fluxes in many of the temperateregions
have increased from pre-industrialtimes by 2 to 20 fold, although some
regions such as northernCanadaare relatively unchanged.Fluxes from the
most disturbedregion, the North Sea drainages,have increasedby 6 to 20
fold. Fluxes from the Amazon basin are also at least 2 to 5 fold greaterthan
estimatedfluxes from undisturbedtemperate-zoneregions. However, we do
not know the relative contributionof naturalvs. anthropogenicactivities.
Deforestationmay be contributingto the Amazon N flux, but tropicalregions
may also have naturallygreaterriverinefluxes due to higher amountsof Nfixationand phosphorusratherthanN limitationof net primaryproductivity
in forests (Howarthet al. this volume).
For the NAO shelf as a whole, N fluxes from rivers and estuaries (600
Gmol yr-') exceed atmosphericdeposition(130 Gmol yr-1) by about4-fold
(Figure 2), but this varies widely among regions of the shelf. For example,
on the U. S. Atlantic shelf and on the NW Europeanshelf, atmosphericN
depositionmay exceed estuarineexports.
Removalof reactiveN from the NAO&Wrelativeto sources
This section discusses the removal of reactive N from the NAO&W by
denitrification,storage in long term reservoirs (e.g., forests, sediments,
groundwater),transportout of the NAO&W via the atmosphere,the hydrosphereand food and feed. It concludes with a summaryof the N budgetfor
the NAO&W (Table 1).
Denitrificationwithin the NAO&W The conversion of reactive N to N2 by
biological denitrificationdecreasesthe reactiveN stock in the NAO&W and
can be viewed as a permanentsink (atmosphericN2 residencetime is ~ 106
yrs). Importantlocations of watersheddenitrificationare wetlands and the
sediments of freshwateraquatic ecosystems. Howarthet al. (this volume)
estimate that about >920 Gmol yr-I are denitrifiedor stored in wetlands,
lakes, streamsandrivers of the temperateportions.The equivalentvalue for
tropicalportionsis unknown.As pointedout in Howarthet al. (this volume),
there is significant spatial variability in the amount of denitrificationand
storageover the watershedof the NAO.
15
Table1. Summaryof Nitrogen Fluxes for the Watershed,Estuary,Shelf and Open Ocean
Regions of the NorthAtlanticOcean and its Watershed,Gmol yr-*
INPUTS
OUTPUTS
* TemperateWatershed
1460 River discharge(2)
520
510 Food/feed (2)
230
330 Groundwaterstorage(2)
<60
?? Foreststorage(2), [when nat. biolog
N-fix is low]
<530
Denitrification& storagein
small
N-fixationby lightning
>920
wetlands,streams,rivers(2)
2300
Totalinput
Totaloutput
2300
natural
whatever
is
in
natural
(Plus
biological N-fixation)
fixed
(w/o
biological N-fixation)
* TropicalWatershed
* TropicalWatershed
85 River discharge(2)
420
Fertilizer,excl. Africa (2)
74 Food/feed (2)
Combustion,excl. Africa (2)
??
N-fixationby legumes (2)
?? Groundwaterstorage(2)
??
Food & feedstock imports
?? Forest storage(2)
??
Naturalbiological N-fixation
?? Denitrification& storagein
N-fixationby lightning
??
wetlands,streams,rivers(2)
??
Totalinput
Totaloutput
??
??
* Estuary
* Estuary
Rivers (2)
540 Transportto shelf (3)
250
small Storagein sediment(3)
small
Atmosphericdeposition(1)
Nat. biological N-fixation(3)
small Denitrification(3, 5)
250
Totalinput
540
Totaloutput
500
* TemperateWatershed
Fertilizer(2)
Combustion(2)
N-fixationby legumes (2)
Naturalbiological N-fixation
* Shelf
* Shelf
Fromestuaries(3)
250 Denitrification(3, 5)
1400
410 Delta, shelf, slope burial(3)
180
Majorrivers(2)
130 Fish catch (3)
32
Atmosphericdepositon(1)
Nat. biological N-fixation(3)
20
Fromocean (3,4)
820
Totalinput
1630
Totaloutput
1612
* Ocean
* Ocean
Nat. biological N-fixation(4,6)
5000 Denitrification(4)
small
620 Storagein sediment(4)
46
Atmosphericdepositon(1)
FromArctic Ocean (4)
1500 To SAO (4)
??
FromMediterraneanSea (4)
120 Transportto shelf (3, 4)
820
FromSAO (4)
??
Total input
7240
Totaloutput
966
* The values in this
table are averages and are presented to give an appreciationof
the magnitudeof the fluxes. For some parameters,the range aroundthe average can be
substantial.For all values, see paperfor details. The numbersin parenthesesrefer to the
originalpaperthatdiscusses the flux: (1) Prospero;(2) Howarth;(3) Nixon; (4) Michaels;
(5) Seitzingerand Giblin; (6) Lipschultz.
16
There are three other locations where denitrificationmay be important:
estuaries,shelf and open ocean. Estuariesreceive about 540 Gmol N yr-1
via rivers, of which about half is buried or denitrified,primarilythe latter
mechanism(Nixon et al. this volume). The shelf region receives about 1,600
Gmol yr-1 from majorrivers, estuaries,atmosphericdeposition,N-fixation,
and the open ocean. It is estimated that -90% is denitrifiedin the shelf
sedimentsand the water column; the remainderis buriedin shelf and slope
sedimentsor removedin fisheriesharvest(Figure2) (Nixon et al. this volume;
Seitzinger& Giblin,this volume). Denitrificationin the openocean can occur
in the sediments and the water column. Most of the sediment denitrification occurs in the shelf, which we have alreadyaccountedfor. Open ocean
denitrificationis uncertain,but thoughtto be small because it requiresareas
with strong sub-surfaceoxygen minimumzones and anoxic basins. Anoxic
conditionsonly occur in a few marginalseas like the Cariacotrench.Even if
there is active denitrificationin anoxic microenvironmentswithin the water
column (e.g. inside particles) then oceanic rates of denitrificationmust be
smallrelativeto the N-fixationrate.Denitrificationremovesnitrateandlowers
the nitrate:phosphateratios below the nominal Redfield stoichiometriesof
ratios are anomalouslyhigh,
16N:1P.In most of the NAO, nitrate:phosphate
relative to the Redfield ratio (Fanning 1989; 1992; Michaels et al. 1994)
indicatingnet N-fixation(Michaels et al. this volume).
Storage within the NAO&W.Storage in the watershedof the NAO occurs
in biomass and groundwater.The temperatewatershedof the NAO stores
up to 530 Gmol yr-1 of the 2,500 Gmol yr-1 in forest biomass in an
unknownmixtureof anthropogenicandnaturalN; storagein groundwaterof
the temperateportionis less than a few percent(<60 Gmol yr-1) (Howarth
et al. this volume). Equivalentvalues for the tropicalforest and groundwater
areunknown(Howarthet al. this volume). Storageof reactiveN in deltas and
estuarineshelf and open ocean sedimentsis thoughtto be small; in all cases
estimated N storage is <10% of the N inputs (Figure 2) (Nixon et al. this
volume;Michaelset al. this volume). The overallbalanceof fluxes in the open
NAO indicatesthatit may be accumulatingN at a rateof 0.06-0.4% per year,
even given the considerableuncertaintyin the transportestimates(Michaels
et al. this volume). This apparentaccumulationis largely a resultof the large
inputsfrom the estimatednitrogen-fixationand a modest atmosphericdeposition, both of which may vary with time and are likely largertoday because
of humanactivity.Thereis a net flux of nitrateto the shelves to meet the shelf
denitrificationdemandthatexceeds the nitrogensupplyfromthe rivers.In the
pre-industrialworld, these riverineN fluxes would also have been lower and
the denitrificationdemandson oceanic nitratelarger.Unfortunately,the large
17
size of the total nitratepool in the oceans and our shortobservationalrecord
precludethe directobservationof changes in the total stock of reactiveN.
Transportout of the NAO&W.There are three transportmechanisms that
distributeanthropogenicN beyondthe bordersof the NAO&W:atmospheric,
hydrologicandfood/feed exports.About 10%of the anthropogenicN formed
in the watershedof the NAO is exported from the NAO&W as food and
feed (270 Gmol yr-1, Figure 2) (Howarthet al. this volume). Hydrologic
transportof N out of the region can occur via deep water flow to the South
Atlantic Ocean. There is a potentialfor transportof up to 20 to 40% of the
N fixed by biological activity in the NAO. The uncertaintyin this transport
term is such that there is also the potentialfor the SAO to be a source for N
that is equivalentto N-fixation in the NAO (Figure 2) (Michaels et al. this
volume). Atmospherictransportof N out of the region does occur; however,
the magnitudeis uncertain.Regions of significanttransportare to the east of
Europeand the west of Centraland SouthAmerica (Figure3).
N Budgetof the NAO&W.In summary,thereare threeimportantN sources to
the NAO&W:anthropogenicN introductionto the watershed,N-fixation in
the open ocean and transportof N into the NAO from adjacentwaterbodies.
Naturalbiological N-fixation in the watershedof the NAO is probablyan
important,but unknownN source. ImportantN sinks are storagewithin the
NAO&W subsystems,denitrification,and possibly transportto the SAO.
Significant uncertaintiesin some of the sinks and sources precludes an
overall budget for the NAO&W. However, we can compareN sources and
sinks for the NAO subsystems(Table 1). For the NAO temperatewatershed,
2,300 Gmol N yr-1 are introducedby fertilizer,fossil fuel combustion and
legume cultivation.Naturalbiological N-fixation is unknownbut probably
I
significant. Of this N, 520 Gmol yr- are lost via rivers; a small amount
(<60 Gmol yr-1) is storedin groundwater.Food andfeed exportsaccountfor
another230 Gmol yr-'. Storage in temperateforests is less than 530 Gmol
yr-1, storageand denitrificationin temperatewetlands,streamsand rivers is
greaterthan920 Gmol yr-1. We are not able to differentiatebetween storage
and denitrificationin wetlands, streamsand rivers. For tropicalregions we
have uncertainestimates of anthropogenicinputs and thus we are not able
to estimate either forest storage or storage and denitrificationin freshwater
ecosystems (Table 1). It is clear, nonetheless, that the tropics export more
N in rivers as a consequenceof anthropogenicactivities (Howarthet al. this
volume).
The estuaries of the NAO&W receive most of their N from riverine
discharge(540 Gmol yr-1); atmosphericdeposition and N-fixation are less
18
important.About half of the N deliveredto estuariesis denitrifiedand half is
transportedto the shelf waters (Table 1).
The NAO shelf receives about 1,600 Gmol N yr-1 from: the ocean (820
Gmol yr- 1), majorrivers(410 Gmol yr- 1), estuaries(250 Gmol yr- 1), atmosphericdeposition(130 Gmol yr-1), and N-fixation (20 Gmol yr-1). Most of
this N is denitrified,and only a small amount(<10%) is lost to delta, shelf
and slope sediments(180 Gmol yr-1) or as fish catch (32 Gmol yr-') (Table
1).
The majorN source to the NAO is biological fixation(5,000 Gmol yr-1)
(Table 1). Hydrologictransportfrom the SAO could be very important,but
both the sign and the magnitudeof the transportis very uncertain.Estimates
I
range from a loss of 1,400 Gmol yr- to the SAO to a gain of 4,600 Gmol
yr-1 from the SAO. In spite of the large uncertaintiesin all the fluxes, the
NAO appearsto be accumulatingN in today's world. Even if the estimate
of transportacross the equatoris significantlywrong, the other sub-regions
of the NAO are also accumulating N (Michaels et al. this volume). The
injectionof anthropogenicN to the NAO by cross shelf transportfrom major
rivers and atmosphericdeposition is small relative to naturalN sources in
the NAO. However,if we consideras anthropogenicthe possible increase in
N-fixationin the SargassoSea from increasedSaharandust(75%higherthan
before the drought)this would add another2,800 to 4,800 Gmol yr-' to the
anthropogenicflux (see laterdiscussion) (Michaels et al. this volume).
Overall,the greatestareasof uncertaintyin the N budgetfor the NAO&W
are(1) storagevs. denitrificationlosses in the watershed,(2) naturalN-fixation
in the watershed,(3) N-fixationin the open ocean, and (4) the exchangeof N
betweenthe NAO and the SAO. It is importantto gain additionalinformation
on all of these fluxes. Until we learn the balance between N storage vs.
denitrificationin the NAO watershedwe are unableto assess to what degree
anthropogenicN is accumulatingin continentalecosystems.In addition,until
we have betterinformationon naturalterrestrialN-fixationwe are unable to
estimate the rate at which human activities are fixing N relative to natural
processes. OuruncertaintyaboutN-fixationratesin the ocean and the degree
that the NAO is a N source or a sink relative to the SAO make it difficult
to determinenot only the naturalN cycle for the NAO but also the possible
impactof Fe inputon N-fixationin the NAO.
P distributionpatterns in the NAO&W
The distributionpatternsof N andP throughthe environmentarequitedifferent. N has a significantgas phase, and is distributedby atmosphericas well
as hydrologicprocesses. P has no stable gas phase in the atmosphereand is
only transportedon particles,limiting its atmosphericlifetime and hence its
19
6
atmospheric
deposition
Arctic
Ocean
71
120
rivers
15
4
major rivers 33
Mediteranean
20
delta
exchange with
South Atlantic
burial
-70 TO 310
2
fish
catch
22
shelf & slope
sediment
2
open ocean
sediment
Figure 4. The P budgetfor the NorthAtlanticOcean (Gmol yr- ).
global dispersion.Hence atmosphericand riverineN inputs into the marine
environmentare both important(Figure 2), while P inputs from rivers are
substantiallygreater than those from the atmosphere(Figure 4). There is
anothercriticaldifferencebetween N andP. N mobilized by humanactivities
can be "unmobilized"by both long term storage (e.g., sediments) and by
denitrification,which returnsN to an inactive state (N2). For P, there is
no process equivalentto denitrification.Once mobilized by human activity,
it accumulates in either a short term reservoir (biomass) or a long term
reservoir (sediments). Given the importanceof hydrologic transportfor P,
and its potentialfor storagein sediments,we focus our effort on the coastal,
shelf and open ocean subsystemsof the NAO&W.
Annually,about 120 Gmol yr-' P are lost from the watershedvia rivers
of which about 53 Gmol P are from large rivers (mostly the Amazon) and
the remainderfrom smaller rivers that flow into estuaries (Figure 4). Most
of the riverineP is deposited to delta, estuarine,shelf and slope sediments
(, 100 Gmol yr-', Figure 4). Only about 30 Gmol yr-' are injected into
the open ocean from rivers. Another 160-300 Gmol yr-' are injected to the
NAO from the Arctic Ocean (230 Gmol yr-1, Figure 4). It appearsthat the
NAO may be accumulatingP at a rate of about 0.1 % per year (Michael
et al. this volume). However, given the relatively large uncertaintiesaround
20
many of the estimates of P fluxes and stocks, this small differencemay not
be separablefrom a balance.
Impactof anthropogenicN and P on the NAO&W
Terrestrial systems. NOx emissions from fossil fuel combustion have
increasedthe oxidizing potentialof the atmosphereon a regionalbasis in the
watershedof the NAO with measurableimpactson humanhealth.The same
emissions have also increasedprecipitationacidity in wide spreadportions
of the watershedof the NAO, with concomitantimpacts on terrestrialand
aquaticecosystems (e.g., Grennfeltet al. 1995). The addition of fertilizer
N to the watershedof the NAO has purposefullyincreasedproductivityof
agriculturalecosystems, resulting in the fixation of more carbon and other
elementsassociatedwith plantand animalgrowth.Thereare also inadvertent
increases in productivityin other ecosystems which result in alterationsto
otherelementalcycles. Persistentatmosphericdepositionof fixed N to forest
ecosystems may eventuallylead to retrogressivechanges in ecosystem function. Aberet al. (1989) suggestedthatforestsin the northeasternUnitedStates
arebecomingN sourcesratherthanN sinksbecauseof extensive N deposition
resultingin N supply in excess of demand. Not only are the forests potentially impactedby the high rateof atmosphericdepositionbut they are also N
sources to downstreamecosystems and to the atmosphere(N20 emissions).
The emission and subsequentdepositionof N results in increasedgrowthof
forests and sequesteringof C and othernutrients(Schindler& Bayley 1993;
Schimel et al. 1995).
Coastal systems. Riverine injection of anthropogenicN and P to coastal
ecosystems has the potentialto increaseproductivity(Howarth1988; Turner
& Rabelais 1991; Vitousek& Howarth1991; NRC 1993; NRC 1994; Nixon
1995; Howarthet al. 1995). The importanceof the N versus P contribution
to increasedproductivitydependson theirrelativeinjectionrates. If the N:P
ratio is <16 ("Redfield"ratio), theoreticallyN is the limiting nutrient;if the
ratio is > 16, P is the limiting nutrient.The N:P ratios vary by region. In the
Amazon, it is 4.5:1; waters flowing into the western Gulf of Mexico have
a 271:1 ratio. However, for most of the riverinefluxes to the NAO coastal
region, the N:P ratios are within a factor of 2 or 3 of the "Redfield"ratio,
suggesting that estuaries in most of these regions have the potential to be
eitherN or P limited,dependingon the bioavailabilityof the N andP, andthe
estuarineprocesses that alterthe ratio(Howarthet al. this volume).
AnthropogenicN in rivers probablydoes not directly impact the open
ocean. Riverine N is primarilydenitrifiedin the shelf region (except for
the Amazon), and thus never reaches the open ocean. However, an indirect
21
impact is that anthropogenicN in rivers results in a decrease in the open
ocean N exportto the shelf. In the pre-industrialworld,if shelf denitrification
was unchanged from today, than supply of N from the open ocean was
perhapsabout double of what it is today. If this is correct,then the oceanic
accumulationmay have an additionalanthropogeniccause.
Open ocean. AnthropogenicN can be deposited to the open ocean, but at
rates that are generally small compared to other sources (Michaels et al.
1993). While anthropogenicN probablydoes not have a major impact on
the overall biogeochemistryof the NAO, there may be episodic impacts on
productivityin stratifiedregions of mid-oceangyre, where upwelling supply
of N is limited (Owens et al. 1992; Michaelset al. 1993). In these situations,
P limitationmay be a confoundingfactor.The N:P ratio of the deposition of
reactiveN andP fromthe atmosphereto the oceanvariessystematicallyacross
the basin.ReactiveN is primarilyfrompollutionsourcesandis maximalnear
the U.S. coastline anddecreasesrapidlyto the east. P depositionis primarily
with dust and thus has a maximum near the coast of Africa and declines
to the west. As a result, the ratio of reactive N to P varies from values of
10-30 near the African coast to values of 500-1500 near the U.S. coastline
(calculatedfrom Prosperoet al. this volume). This analysis does not include
organic N or P; however, since levels of organic phosphorousare low, the
N:P ratio is dominatedby the substantialpools of organic nitrogen. Given
that the N:P ratios are always greaterthan the Redfield ratio, it is possible
that the P deposition may regulatethe impacton surfacebiology. However,
thereare confoundingfactors;e.g., much of the depositedP is tightly bound
in insolublecompoundssuch as apatiteand,consequently,the P is not readily
availableto organisms.
Phytoplanktonin the surface waters of the Sargasso Sea are exposed to
low concentrationsof both reactive N and phosphate;in situ concentrations
are below the detection limits of standardanalysis for both nutrients.Thus,
the instantaneousgrowth rates of most phytoplanktoncould be co-limited
by both nutrients.Some organismsmay have the ability to fix atmospheric
N and these would obviously be limited by the availability of phosphate.
The atmosphericdeposition patternsmodify this picture.In nearly all cases
the deposition ratio of N and P is higher than the traditionalstoichiometry
of phytoplanktongrowth, the Redfield ratio of 16N:1P. Thus, a deposition
event, even if it adds considerablereactiveN to the surfacemixed layer,will
rarely add enough phosphateto allow complete utilization of the N. Thus,
in these deposition events, surfacepopulationsare likely phosphatelimited.
Since the required phosphate supply to sustain the N-fixation postulated
by Michaels et al. this volume, is very large comparedto the atmospheric
22
deposition, the phosphateto supportN-fixation must come from the stocks
in the thermocline.They suggest thatthe migrationbehaviorof the large and
colonial organismresponsiblefor the nitrogen-fixationallow them access to
these deeperphosphatestocks. Consequently,the nutrientlimitationpatterns
for the surfacecommunityare probablydeterminedby the dynamicsof the
phosphatetransportfrom below and not the atmosphericdeposition.
There may be an indirectimpact of human activities on the open ocean
N cycle. There is an interestingand potentially very importantinteraction
between the N cycle and Fe. It is proposed that Fe deposition to the NAO
by atmospherictransportfrom Africa is supportingN-fixation rates in the
NAO (Prosperoet al. this volume; Michaelset al. this volume). A substantial
fractionof the mineraldust carriedto the NAO appearsto be mobilized as a
consequenceof humanactivitiesin Africacompoundedby rainfallvariability
(Prosperoet al. this volume);thus, the impactof dust-borneFe on N-fixation
ratesin the NAO could be regardedas being largely an anthropogeniceffect.
Also, because dust transportis highly variable from year-to-yeardue to
changes in weatherand climate (Prospero& Nees 1986) the impact of Fe
deposition to the ocean could also vary greatly. This linkage of emission
processes in Africa to ecosystem productivityin far downwind ecosystems
is not new - it has also been proposedfor the supply of P to portionsof the
Amazon forest (Swap et al. 1992). Further,the nitrogen-fixationthat occurs
in the Sargasso Sea would require an additionaluptake of dissolved CO2
of approximately0.3-1.2 Gt yr-1 (Michaels, this volume). This carbon is
remineralizedat depthswith residence times of decades, so the process does
not resultin a long-termsequestrationof surfaceor atmosphericcarbon.However, if the rate of nitrogen-fixationis regulatedby the highly variabledust
deposition,then the resultingvariationsin N-fixationratecould yield significant interannualfluctuationsin the carbondrawdownand perhapsaffect the
interannualvariationsin atmosphericCO2as well.
Acknowledgments
We wish to acknowledgeandexpress our appreciationfor the fundingof this
projectfrom the UNEP and the Mellon Foundationto supportthe efforts of
the SCOPENitrogenProject,and from the WorldMeteorologicalOrganization to investigatethe interactionbetweenthe N cycles of the atmosphereand
the NAO&W. The Rhode Island Sea GrantProgramcontributedlogistical
supportfor the meeting. We appreciatethe dedicated and cheerful support
services providedat the meeting by Roxanne Marino (Comrnell
University),
Dennis
Ehrenfeld
University),
Swaney (Comrnell
University),
Emily
(Comrnell
and Ellen Yoder(Universityof Rhode Island). We are also thankfulfor the
23
effortsof MaryScott Kaiser(Universityof Virginia)in makingarrangements
priorto the meetingandto JudithPeatross(Universityof Virginia)in assisting
with manuscriptediting and preparation.James Galloway is gratefulto The
EcosystemsCenterof the MarineBiological Station,theWoodsHole OceanographicInstitute,andthe InternationalInstitutefor AppliedSystems Analysis
for hosting him duringhis sabbaticalyear. In addition,partialsupportfrom
NOAA (GlobalPrecipitationChemistryProject)and the NSF (AtmosphereOcean ChemistryExperiment)is gratefullyacknowledgedin the preparation
of this manuscript.This paper is a contributionto the BermudaBiological
Stationfor Research,Inc.
References
AberJD NadelhofferKJ SteudlerP & Melillo JM (1989) Nitrogensaturationin northernforest
ecosystems. BioScience 39: 378-386
AyresRU, SchlesingerWH & Socolow RH (1994) Humanimpactson the carbonandnitrogen
cycles. In: Socolow RH, Andrews C, BerkhoutR & Thomas V (Eds) IndustrialEcology
and Global Change(pp 121-155). CambridgeUniversityPress, CambridgeMass
DentenerFJ & CrutzenPJ (1993) Reaction of N205 on troposphericaerosols:Impacton the
global distributionsof NOx 03 and OH. J. Geophys. Res. 98: 7149-7163
DentenerFJ & CrutzenPJ (1994) A three-dimensionalmodel of the global ammoniacycle. J.
Atmos. Chem. 19: 331-369
Fanning KA (1989) Influence of atmosphericpollution on nutrientlimitation in the ocean.
Nature339: 460-463
FanningKA (1992) Nutrientprovinces in the sea: Concentrationrationsreaction rate ratios
and ioodeal covariation.J. Geophys.Res. C Oceans 97: 5693-5712
GallowayJN, SchlesingerWH, Levy II H, MichaelsA & SchnoorJL (1995) Nitrogen-fixation:
anthropogenicenhancement-environmentalresponse. Global Biogeochemical Cycles 9:
235-252
GrennfeltP, Rodhe H, ThrnmelifE & WisniewskiJ (1995) Acid Reign '95? Proceedingsfrom
the 5th InternationalConference on Acidic Deposition Goteborg Sweden 26-30 June.
Water,Air and Soil Pollut. 85, Nos. 1, 2, 3, & 4
HowarthRW (1988) Nutrientlimitationof net primaryproductionin marineecosystems.Ann.
Rev. Ecol. & Syst. 19: 89-110
HowarthRW, Jensen H, Marino R & Postma H (1995) Transportto and processing of P in
near-shoreand oceanic waters. In: Tiessen H (Ed) Phosphorusin the Global Environment: TransfersCycles and ManagementSCOPE 54 (pp 323-345). JohnWiley & Sons,
ChichesterUK
HowarthRW,Billen G, Swaney D, TownsendA, JaworskiN, LajthaK, Downing JA, Elmgren
R, CaracoN, JordanT, BerendseF, FreneyJ, KudeyarovV, MurdochP & Zhu Zhao-liang
(1996) Regional nitrogen budgets and riverine N and P fluxes for the drainages to the
NorthAtlanticOcean:naturaland humaninfluences.Biogeochemistry,this volume
Jaffee DA (1992) The Nitrogen Cycle. In: ButcherSS, CharlsonRJ, OriansGH & GV Wolfe
(Eds) Global Biogeochemical Cycles (pp 263-284). Pub AcademicPress
Levy II H, Moxim WJ & KasibhatlaPS (1996) A global 3-dimensional time dependent
lightningsource of troposphericNOx. J. Geophys. Res. in press
LipschultzF & Owens NJP (1996). An assessmentof nitrogen-fixationas a source of nitrogen
to the North AtlanticOcean. Biogeochemistry,this volume
24
MackenzieFT, VerLM, SabineC, LaneM & LermanA (1993)C N P S global biogeochemical
cycles andmodelingof global changeNATOASI Series Vol 14. In: WollastR, Mackenzie
FT & ChouL (Eds) Interactionsof C N P andS BiogeochemicalCycles andGlobalChange
(pp 1-61). Springer-VerlagBerlin Heidelberg
Michaels AF, Knap AH, Dow RL GundersenK, JohnsonRJ, Sorensen J, Close A, Knauer
GA, Lohrenz SE, Asper VA Tuel M & BidigareRR (1994) Seasonal patternsof ocean
biogeochemistryat the U S JGOFS BermudaAtlantic Time-series Study site. Deep-Sea
Res. 41: 1013-1038
Michaels AF, Siegel DA, JohnsonRJ, Knap AH & Galloway JN (1993) Episodic inputs of
atmosphericnitrogen to the Sargasso Sea: Contributionsto new productionand phytoplanktonblooms. Global Biogeochem. Cycles 7: 339-351
Michaels AF, Olson D, AmmermanJW, Fanning K, JahnkeR, Knap AH, Lipschultz F &
ProsperoJM (1996) Inputslosses and transformationsof nitrogenand phosphorusin the
pelagic North AtlanticOcean. Biogeochemistry,this volume
Moxim WJ, Levy II H, KasibhatlaPS & ChameidesWL (1994) The impactof anthropogenic
and naturalNOx sources on the global distributionof troposphericreactive nitrogenand
the net chemical productionof ozone. Eos TransAGU 75 Fall Meet Supple A41E
National Research Council (1993) Managing Wastewaterin Coastal Urban Areas Committee on WastewaterManagementfor Coastal Urban Areas. National Academy Press.
WashingtonDC
Nixon SW (1995) Coastal marineeutrophication:A definition social causes and futureconcerns. Ophelia 41: 199-219
Nixon SW, AmmermanJW, AtkinsonLP, BerounskyVM, Billen G, BoicourtWC, Boynton
WR, ChurchTM, DiToroDM, ElmgrenR, GarberJH, GiblinAE, JahnkeRA, Owens NJP,
Pilson MEQ & SeitzingerSP (1996) The fate of nitrogenand phosphorusat the land-sea
marginof the NorthAtlanticOcean. Biogeochemistry,this volume
Owens NJP, Galloway JN & Duce RA (1992) Episodic atmosphericnitrogen deposition to
oligotrophicoceans. Nature357: 397-399
Pilson MEQ & Seitzinger SP (1996) Areas of shallow water in the North Atlantic.Biogeochemistry,this volume
ProsperoJM & Nees RT (1986) Impactof NorthAfricandroughtandEl Nino on mineraldust
in the Barbadostradewinds. Nature320: 735-738
ProsperoJM, BarrettK, ChurchT, DentenerF, Duce RA, Galloway JN, Levy II H, Moody
J & Quinn P (1996) Atmosphericdeposition of nutrientsto the North Atlantic basin.
Biogeochemistry,this volume
Quinn PK, CharlsonRJ & Bates TS (1988) Simultaneousobservations of ammonia in the
atmosphereand ocean. Nature335: 336-338
QuinnPK, BarrettKJ, DentenerFJ, LipschultzF & KurzKD (1996) Estimationof the air/sea
exchangeof ammoniafor the North AtlanticBasin. Biogeochemistry,this volume
SchimelD, EntingI, HeimannM, Wigley TML,RaynaudD, Alves D & SiegenthalerU (1995)
CO2 and the carboncycle. In: HoughtonJ et al. (Eds) Climate Change, 1994, Radiative
Forcingof ClimateChange.CambridgeUniversityPress, England,339 p
SchlesingerWH (1991) Biogeochemistry:An Analysis of Global Change. Academic Press,
San Diego, 443 pp
SchindlerDW & Bayley SE (1993) Thebiosphereas an increasingsink for atmosphericcarbon:
Estimatesfrom increasednitrogendeposition.Global Biogeochem Cycles 7: 717-725
SeitzingerSP & Giblin AE (1996) Estimatingdenitrificationin the North Atlanticcontinental
shelf sediments.Biogeochemistry,this volume
Swap R, GarstangM, TalbotR & KakaallbergP (1992) Saharandust in the Amazon basin.
Tellus 44b: 133-149
TurnerRE & Rabalais NN (1991) Changes in Mississippi River water quality this century:
Implicationsfor coastal food webs. Bioscience 41: 140-147
VitousekPM & HowarthRW (1991) Nitrogen limitationon land and in the sea: How can it
occur?Biogeochemistry13: 87-115
25
WameckP (1988) Chemistryof the NaturalAtmosphere.AcademicPress, London
WhelpdaleDM & Galloway JN (1994) Sulfurand nitrogenoxide fluxes in the North Atlantic
atmosphere.Global Biogeochem. Cycles 8: 481-493